{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "eval_steps": 500, "global_step": 1955, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0025575447570332483, "grad_norm": 2.956157423011394, "learning_rate": 4.0816326530612243e-07, "loss": 1.5213, "step": 1 }, { "epoch": 0.005115089514066497, "grad_norm": 2.957156524653666, "learning_rate": 8.163265306122449e-07, "loss": 1.4742, "step": 2 }, { "epoch": 0.0076726342710997444, "grad_norm": 3.013201374781629, "learning_rate": 1.2244897959183673e-06, "loss": 1.4946, "step": 3 }, { "epoch": 0.010230179028132993, "grad_norm": 2.9332409751747672, "learning_rate": 1.6326530612244897e-06, "loss": 1.4837, "step": 4 }, { "epoch": 0.01278772378516624, "grad_norm": 2.8958487377498088, "learning_rate": 2.0408163265306125e-06, "loss": 1.4922, "step": 5 }, { "epoch": 0.015345268542199489, "grad_norm": 2.7113697659439957, "learning_rate": 2.4489795918367347e-06, "loss": 1.4997, "step": 6 }, { "epoch": 0.017902813299232736, "grad_norm": 2.55930320591955, "learning_rate": 2.8571428571428573e-06, "loss": 1.4709, "step": 7 }, { "epoch": 0.020460358056265986, "grad_norm": 2.0258165853063708, "learning_rate": 3.2653061224489794e-06, "loss": 1.4673, "step": 8 }, { "epoch": 0.023017902813299233, "grad_norm": 1.9427534891501914, "learning_rate": 3.6734693877551024e-06, "loss": 1.5002, "step": 9 }, { "epoch": 0.02557544757033248, "grad_norm": 1.8741278853652108, "learning_rate": 4.081632653061225e-06, "loss": 1.4897, "step": 10 }, { "epoch": 0.028132992327365727, "grad_norm": 1.8672035113437018, "learning_rate": 4.489795918367348e-06, "loss": 1.4669, "step": 11 }, { "epoch": 0.030690537084398978, "grad_norm": 1.8231697254158397, "learning_rate": 4.897959183673469e-06, "loss": 1.4282, "step": 12 }, { "epoch": 0.03324808184143223, "grad_norm": 1.7349261636368631, "learning_rate": 5.306122448979593e-06, "loss": 1.4664, "step": 13 }, { "epoch": 0.03580562659846547, "grad_norm": 1.4416073688922821, "learning_rate": 5.7142857142857145e-06, "loss": 1.3861, "step": 14 }, { "epoch": 0.03836317135549872, "grad_norm": 1.804185644593906, "learning_rate": 6.122448979591837e-06, "loss": 1.4201, "step": 15 }, { "epoch": 0.04092071611253197, "grad_norm": 1.78311427201953, "learning_rate": 6.530612244897959e-06, "loss": 1.3764, "step": 16 }, { "epoch": 0.043478260869565216, "grad_norm": 1.5180464959128221, "learning_rate": 6.938775510204082e-06, "loss": 1.3926, "step": 17 }, { "epoch": 0.04603580562659847, "grad_norm": 1.2280933246833325, "learning_rate": 7.346938775510205e-06, "loss": 1.3886, "step": 18 }, { "epoch": 0.04859335038363171, "grad_norm": 0.978036927475401, "learning_rate": 7.755102040816327e-06, "loss": 1.3666, "step": 19 }, { "epoch": 0.05115089514066496, "grad_norm": 1.087810311630016, "learning_rate": 8.16326530612245e-06, "loss": 1.3479, "step": 20 }, { "epoch": 0.05370843989769821, "grad_norm": 0.890366570770121, "learning_rate": 8.571428571428571e-06, "loss": 1.3275, "step": 21 }, { "epoch": 0.056265984654731455, "grad_norm": 0.8853934440384114, "learning_rate": 8.979591836734695e-06, "loss": 1.3511, "step": 22 }, { "epoch": 0.058823529411764705, "grad_norm": 0.7917639991185591, "learning_rate": 9.387755102040818e-06, "loss": 1.3045, "step": 23 }, { "epoch": 0.061381074168797956, "grad_norm": 0.6651063674631976, "learning_rate": 9.795918367346939e-06, "loss": 1.3239, "step": 24 }, { "epoch": 0.0639386189258312, "grad_norm": 0.6825326754949458, "learning_rate": 1.0204081632653063e-05, "loss": 1.2989, "step": 25 }, { "epoch": 0.06649616368286446, "grad_norm": 0.7207844273872734, "learning_rate": 1.0612244897959186e-05, "loss": 1.3023, "step": 26 }, { "epoch": 0.06905370843989769, "grad_norm": 0.587801356926753, "learning_rate": 1.1020408163265306e-05, "loss": 1.2843, "step": 27 }, { "epoch": 0.07161125319693094, "grad_norm": 0.49770805079425884, "learning_rate": 1.1428571428571429e-05, "loss": 1.3075, "step": 28 }, { "epoch": 0.0741687979539642, "grad_norm": 0.545355697285009, "learning_rate": 1.1836734693877552e-05, "loss": 1.2835, "step": 29 }, { "epoch": 0.07672634271099744, "grad_norm": 0.5693171639034866, "learning_rate": 1.2244897959183674e-05, "loss": 1.2971, "step": 30 }, { "epoch": 0.0792838874680307, "grad_norm": 0.5161243281317084, "learning_rate": 1.2653061224489798e-05, "loss": 1.2911, "step": 31 }, { "epoch": 0.08184143222506395, "grad_norm": 0.46725528465384736, "learning_rate": 1.3061224489795918e-05, "loss": 1.2638, "step": 32 }, { "epoch": 0.08439897698209718, "grad_norm": 0.41798797083329553, "learning_rate": 1.3469387755102042e-05, "loss": 1.2964, "step": 33 }, { "epoch": 0.08695652173913043, "grad_norm": 0.4623409185375301, "learning_rate": 1.3877551020408165e-05, "loss": 1.3138, "step": 34 }, { "epoch": 0.08951406649616368, "grad_norm": 0.4374009191444938, "learning_rate": 1.4285714285714287e-05, "loss": 1.2555, "step": 35 }, { "epoch": 0.09207161125319693, "grad_norm": 0.36271282660839976, "learning_rate": 1.469387755102041e-05, "loss": 1.2684, "step": 36 }, { "epoch": 0.09462915601023018, "grad_norm": 0.39662907683429816, "learning_rate": 1.510204081632653e-05, "loss": 1.2828, "step": 37 }, { "epoch": 0.09718670076726342, "grad_norm": 0.3621344734960456, "learning_rate": 1.5510204081632655e-05, "loss": 1.2585, "step": 38 }, { "epoch": 0.09974424552429667, "grad_norm": 0.38753977253082816, "learning_rate": 1.5918367346938776e-05, "loss": 1.2739, "step": 39 }, { "epoch": 0.10230179028132992, "grad_norm": 0.32331110740235386, "learning_rate": 1.63265306122449e-05, "loss": 1.2148, "step": 40 }, { "epoch": 0.10485933503836317, "grad_norm": 0.3235269735931933, "learning_rate": 1.673469387755102e-05, "loss": 1.242, "step": 41 }, { "epoch": 0.10741687979539642, "grad_norm": 0.3438080240782346, "learning_rate": 1.7142857142857142e-05, "loss": 1.2409, "step": 42 }, { "epoch": 0.10997442455242967, "grad_norm": 0.2916611186028452, "learning_rate": 1.7551020408163266e-05, "loss": 1.2136, "step": 43 }, { "epoch": 0.11253196930946291, "grad_norm": 0.3008986538459429, "learning_rate": 1.795918367346939e-05, "loss": 1.2245, "step": 44 }, { "epoch": 0.11508951406649616, "grad_norm": 0.3060253220280668, "learning_rate": 1.836734693877551e-05, "loss": 1.2252, "step": 45 }, { "epoch": 0.11764705882352941, "grad_norm": 0.31820392904760986, "learning_rate": 1.8775510204081636e-05, "loss": 1.2464, "step": 46 }, { "epoch": 0.12020460358056266, "grad_norm": 0.2865033138020395, "learning_rate": 1.9183673469387756e-05, "loss": 1.2466, "step": 47 }, { "epoch": 0.12276214833759591, "grad_norm": 0.26555310624821354, "learning_rate": 1.9591836734693877e-05, "loss": 1.2171, "step": 48 }, { "epoch": 0.12531969309462915, "grad_norm": 0.2876989657538828, "learning_rate": 2e-05, "loss": 1.2317, "step": 49 }, { "epoch": 0.1278772378516624, "grad_norm": 0.247990267920054, "learning_rate": 2.0408163265306126e-05, "loss": 1.201, "step": 50 }, { "epoch": 0.13043478260869565, "grad_norm": 0.2570347387829217, "learning_rate": 2.0816326530612247e-05, "loss": 1.2083, "step": 51 }, { "epoch": 0.1329923273657289, "grad_norm": 0.2502935337530908, "learning_rate": 2.122448979591837e-05, "loss": 1.189, "step": 52 }, { "epoch": 0.13554987212276215, "grad_norm": 0.25819296392357916, "learning_rate": 2.1632653061224492e-05, "loss": 1.1904, "step": 53 }, { "epoch": 0.13810741687979539, "grad_norm": 0.2557655926265301, "learning_rate": 2.2040816326530613e-05, "loss": 1.2268, "step": 54 }, { "epoch": 0.14066496163682865, "grad_norm": 0.23483854657893372, "learning_rate": 2.2448979591836734e-05, "loss": 1.2391, "step": 55 }, { "epoch": 0.1432225063938619, "grad_norm": 0.24717465180571466, "learning_rate": 2.2857142857142858e-05, "loss": 1.227, "step": 56 }, { "epoch": 0.14578005115089515, "grad_norm": 0.2172921037220085, "learning_rate": 2.3265306122448982e-05, "loss": 1.2093, "step": 57 }, { "epoch": 0.1483375959079284, "grad_norm": 0.26820055891479894, "learning_rate": 2.3673469387755103e-05, "loss": 1.2253, "step": 58 }, { "epoch": 0.15089514066496162, "grad_norm": 0.21516549373211177, "learning_rate": 2.4081632653061227e-05, "loss": 1.191, "step": 59 }, { "epoch": 0.1534526854219949, "grad_norm": 0.27883891563143004, "learning_rate": 2.448979591836735e-05, "loss": 1.2151, "step": 60 }, { "epoch": 0.15601023017902813, "grad_norm": 0.23221139463765425, "learning_rate": 2.4897959183673473e-05, "loss": 1.2032, "step": 61 }, { "epoch": 0.1585677749360614, "grad_norm": 0.2902807098346473, "learning_rate": 2.5306122448979597e-05, "loss": 1.2352, "step": 62 }, { "epoch": 0.16112531969309463, "grad_norm": 0.25945252836481614, "learning_rate": 2.5714285714285718e-05, "loss": 1.2182, "step": 63 }, { "epoch": 0.1636828644501279, "grad_norm": 0.3154426361540951, "learning_rate": 2.6122448979591835e-05, "loss": 1.2128, "step": 64 }, { "epoch": 0.16624040920716113, "grad_norm": 0.3917230985512171, "learning_rate": 2.653061224489796e-05, "loss": 1.2052, "step": 65 }, { "epoch": 0.16879795396419436, "grad_norm": 0.4945933154303559, "learning_rate": 2.6938775510204084e-05, "loss": 1.1671, "step": 66 }, { "epoch": 0.17135549872122763, "grad_norm": 0.6037374924582795, "learning_rate": 2.7346938775510205e-05, "loss": 1.1915, "step": 67 }, { "epoch": 0.17391304347826086, "grad_norm": 0.5830383636720183, "learning_rate": 2.775510204081633e-05, "loss": 1.1683, "step": 68 }, { "epoch": 0.17647058823529413, "grad_norm": 0.4049022994573582, "learning_rate": 2.816326530612245e-05, "loss": 1.152, "step": 69 }, { "epoch": 0.17902813299232737, "grad_norm": 0.36129854662345895, "learning_rate": 2.8571428571428574e-05, "loss": 1.1638, "step": 70 }, { "epoch": 0.1815856777493606, "grad_norm": 0.5309473427051697, "learning_rate": 2.89795918367347e-05, "loss": 1.1747, "step": 71 }, { "epoch": 0.18414322250639387, "grad_norm": 0.6377735187370677, "learning_rate": 2.938775510204082e-05, "loss": 1.1894, "step": 72 }, { "epoch": 0.1867007672634271, "grad_norm": 0.648301783776856, "learning_rate": 2.9795918367346944e-05, "loss": 1.2237, "step": 73 }, { "epoch": 0.18925831202046037, "grad_norm": 0.5452770096862445, "learning_rate": 3.020408163265306e-05, "loss": 1.1695, "step": 74 }, { "epoch": 0.1918158567774936, "grad_norm": 0.47917361255424545, "learning_rate": 3.061224489795918e-05, "loss": 1.1671, "step": 75 }, { "epoch": 0.19437340153452684, "grad_norm": 0.46650181032832944, "learning_rate": 3.102040816326531e-05, "loss": 1.1705, "step": 76 }, { "epoch": 0.1969309462915601, "grad_norm": 0.39177334859844115, "learning_rate": 3.142857142857143e-05, "loss": 1.1808, "step": 77 }, { "epoch": 0.19948849104859334, "grad_norm": 0.499101882627103, "learning_rate": 3.183673469387755e-05, "loss": 1.1616, "step": 78 }, { "epoch": 0.2020460358056266, "grad_norm": 0.5788889023503736, "learning_rate": 3.224489795918368e-05, "loss": 1.1554, "step": 79 }, { "epoch": 0.20460358056265984, "grad_norm": 0.5961045731637039, "learning_rate": 3.26530612244898e-05, "loss": 1.1725, "step": 80 }, { "epoch": 0.2071611253196931, "grad_norm": 0.47119669685447696, "learning_rate": 3.306122448979592e-05, "loss": 1.1561, "step": 81 }, { "epoch": 0.20971867007672634, "grad_norm": 0.4703018870884221, "learning_rate": 3.346938775510204e-05, "loss": 1.153, "step": 82 }, { "epoch": 0.21227621483375958, "grad_norm": 0.5556907967995176, "learning_rate": 3.387755102040817e-05, "loss": 1.1677, "step": 83 }, { "epoch": 0.21483375959079284, "grad_norm": 0.6373677375142429, "learning_rate": 3.4285714285714284e-05, "loss": 1.1634, "step": 84 }, { "epoch": 0.21739130434782608, "grad_norm": 0.825864360069245, "learning_rate": 3.469387755102041e-05, "loss": 1.2139, "step": 85 }, { "epoch": 0.21994884910485935, "grad_norm": 1.1333633511553514, "learning_rate": 3.510204081632653e-05, "loss": 1.1666, "step": 86 }, { "epoch": 0.22250639386189258, "grad_norm": 1.1545216752131862, "learning_rate": 3.551020408163265e-05, "loss": 1.1727, "step": 87 }, { "epoch": 0.22506393861892582, "grad_norm": 0.842741588438428, "learning_rate": 3.591836734693878e-05, "loss": 1.1958, "step": 88 }, { "epoch": 0.22762148337595908, "grad_norm": 0.5836488090866901, "learning_rate": 3.63265306122449e-05, "loss": 1.1561, "step": 89 }, { "epoch": 0.23017902813299232, "grad_norm": 0.6807274898079663, "learning_rate": 3.673469387755102e-05, "loss": 1.1866, "step": 90 }, { "epoch": 0.23273657289002558, "grad_norm": 0.978642798883616, "learning_rate": 3.714285714285715e-05, "loss": 1.1831, "step": 91 }, { "epoch": 0.23529411764705882, "grad_norm": 1.1688573758554635, "learning_rate": 3.755102040816327e-05, "loss": 1.1865, "step": 92 }, { "epoch": 0.23785166240409208, "grad_norm": 0.6824215678902069, "learning_rate": 3.795918367346939e-05, "loss": 1.1676, "step": 93 }, { "epoch": 0.24040920716112532, "grad_norm": 0.44788443667257866, "learning_rate": 3.836734693877551e-05, "loss": 1.1884, "step": 94 }, { "epoch": 0.24296675191815856, "grad_norm": 0.5868828512762089, "learning_rate": 3.8775510204081634e-05, "loss": 1.1359, "step": 95 }, { "epoch": 0.24552429667519182, "grad_norm": 0.9092466565103363, "learning_rate": 3.9183673469387755e-05, "loss": 1.1769, "step": 96 }, { "epoch": 0.24808184143222506, "grad_norm": 1.0055631954693776, "learning_rate": 3.959183673469388e-05, "loss": 1.18, "step": 97 }, { "epoch": 0.2506393861892583, "grad_norm": 0.8661368356493842, "learning_rate": 4e-05, "loss": 1.1508, "step": 98 }, { "epoch": 0.2531969309462916, "grad_norm": 0.6456496755830919, "learning_rate": 4.0408163265306124e-05, "loss": 1.1473, "step": 99 }, { "epoch": 0.2557544757033248, "grad_norm": 0.8111263397131059, "learning_rate": 4.081632653061225e-05, "loss": 1.1413, "step": 100 }, { "epoch": 0.25831202046035806, "grad_norm": 1.2262439163561427, "learning_rate": 4.1224489795918366e-05, "loss": 1.1647, "step": 101 }, { "epoch": 0.2608695652173913, "grad_norm": 0.895523164890546, "learning_rate": 4.1632653061224494e-05, "loss": 1.1698, "step": 102 }, { "epoch": 0.26342710997442453, "grad_norm": 0.4128663944963409, "learning_rate": 4.2040816326530615e-05, "loss": 1.1374, "step": 103 }, { "epoch": 0.2659846547314578, "grad_norm": 0.6875687358413721, "learning_rate": 4.244897959183674e-05, "loss": 1.1295, "step": 104 }, { "epoch": 0.26854219948849106, "grad_norm": 1.2301351848291924, "learning_rate": 4.2857142857142856e-05, "loss": 1.1397, "step": 105 }, { "epoch": 0.2710997442455243, "grad_norm": 0.7010073231328794, "learning_rate": 4.3265306122448984e-05, "loss": 1.1273, "step": 106 }, { "epoch": 0.27365728900255754, "grad_norm": 0.5227686699195527, "learning_rate": 4.3673469387755105e-05, "loss": 1.1544, "step": 107 }, { "epoch": 0.27621483375959077, "grad_norm": 0.5470695476248038, "learning_rate": 4.4081632653061226e-05, "loss": 1.134, "step": 108 }, { "epoch": 0.27877237851662406, "grad_norm": 0.7660262395726699, "learning_rate": 4.4489795918367353e-05, "loss": 1.1582, "step": 109 }, { "epoch": 0.2813299232736573, "grad_norm": 1.1805124576184864, "learning_rate": 4.489795918367347e-05, "loss": 1.1614, "step": 110 }, { "epoch": 0.28388746803069054, "grad_norm": 0.8289200241114665, "learning_rate": 4.5306122448979595e-05, "loss": 1.162, "step": 111 }, { "epoch": 0.2864450127877238, "grad_norm": 0.6363878509379256, "learning_rate": 4.5714285714285716e-05, "loss": 1.1508, "step": 112 }, { "epoch": 0.289002557544757, "grad_norm": 0.8223651947759109, "learning_rate": 4.6122448979591844e-05, "loss": 1.1246, "step": 113 }, { "epoch": 0.2915601023017903, "grad_norm": 0.9658149004434372, "learning_rate": 4.6530612244897965e-05, "loss": 1.155, "step": 114 }, { "epoch": 0.29411764705882354, "grad_norm": 1.1429259285624525, "learning_rate": 4.6938775510204086e-05, "loss": 1.1396, "step": 115 }, { "epoch": 0.2966751918158568, "grad_norm": 1.0902172962065055, "learning_rate": 4.7346938775510206e-05, "loss": 1.146, "step": 116 }, { "epoch": 0.29923273657289, "grad_norm": 0.988988857628424, "learning_rate": 4.7755102040816334e-05, "loss": 1.1366, "step": 117 }, { "epoch": 0.30179028132992325, "grad_norm": 0.7402542318116936, "learning_rate": 4.8163265306122455e-05, "loss": 1.1863, "step": 118 }, { "epoch": 0.30434782608695654, "grad_norm": 0.9283436178370745, "learning_rate": 4.857142857142857e-05, "loss": 1.1481, "step": 119 }, { "epoch": 0.3069053708439898, "grad_norm": 1.1139736386114893, "learning_rate": 4.89795918367347e-05, "loss": 1.1574, "step": 120 }, { "epoch": 0.309462915601023, "grad_norm": 0.6424105578724688, "learning_rate": 4.938775510204082e-05, "loss": 1.1701, "step": 121 }, { "epoch": 0.31202046035805625, "grad_norm": 0.5840585035243322, "learning_rate": 4.9795918367346945e-05, "loss": 1.1204, "step": 122 }, { "epoch": 0.3145780051150895, "grad_norm": 0.8914478021127121, "learning_rate": 5.0204081632653066e-05, "loss": 1.1424, "step": 123 }, { "epoch": 0.3171355498721228, "grad_norm": 0.8050516376303483, "learning_rate": 5.0612244897959194e-05, "loss": 1.1374, "step": 124 }, { "epoch": 0.319693094629156, "grad_norm": 0.7768422507226259, "learning_rate": 5.102040816326531e-05, "loss": 1.1181, "step": 125 }, { "epoch": 0.32225063938618925, "grad_norm": 0.9832467250312055, "learning_rate": 5.1428571428571436e-05, "loss": 1.1755, "step": 126 }, { "epoch": 0.3248081841432225, "grad_norm": 1.0832409134965832, "learning_rate": 5.1836734693877557e-05, "loss": 1.12, "step": 127 }, { "epoch": 0.3273657289002558, "grad_norm": 1.05095046390139, "learning_rate": 5.224489795918367e-05, "loss": 1.1371, "step": 128 }, { "epoch": 0.329923273657289, "grad_norm": 0.8504844773142398, "learning_rate": 5.26530612244898e-05, "loss": 1.1289, "step": 129 }, { "epoch": 0.33248081841432225, "grad_norm": 1.013882978247601, "learning_rate": 5.306122448979592e-05, "loss": 1.1588, "step": 130 }, { "epoch": 0.3350383631713555, "grad_norm": 1.4983721735915714, "learning_rate": 5.346938775510205e-05, "loss": 1.1252, "step": 131 }, { "epoch": 0.3375959079283887, "grad_norm": 0.6395636406832796, "learning_rate": 5.387755102040817e-05, "loss": 1.1789, "step": 132 }, { "epoch": 0.340153452685422, "grad_norm": 1.1119427951585954, "learning_rate": 5.4285714285714295e-05, "loss": 1.1548, "step": 133 }, { "epoch": 0.34271099744245526, "grad_norm": 1.50419051921151, "learning_rate": 5.469387755102041e-05, "loss": 1.1259, "step": 134 }, { "epoch": 0.3452685421994885, "grad_norm": 0.8868212687779602, "learning_rate": 5.510204081632654e-05, "loss": 1.1825, "step": 135 }, { "epoch": 0.34782608695652173, "grad_norm": 1.1455650604061842, "learning_rate": 5.551020408163266e-05, "loss": 1.1453, "step": 136 }, { "epoch": 0.35038363171355497, "grad_norm": 1.1928108885641358, "learning_rate": 5.5918367346938786e-05, "loss": 1.1554, "step": 137 }, { "epoch": 0.35294117647058826, "grad_norm": 1.2946027592370013, "learning_rate": 5.63265306122449e-05, "loss": 1.1429, "step": 138 }, { "epoch": 0.3554987212276215, "grad_norm": 0.7456369454357881, "learning_rate": 5.673469387755102e-05, "loss": 1.116, "step": 139 }, { "epoch": 0.35805626598465473, "grad_norm": 0.9145391630298629, "learning_rate": 5.714285714285715e-05, "loss": 1.176, "step": 140 }, { "epoch": 0.36061381074168797, "grad_norm": 1.0539461991771215, "learning_rate": 5.755102040816327e-05, "loss": 1.0897, "step": 141 }, { "epoch": 0.3631713554987212, "grad_norm": 1.1574195644796725, "learning_rate": 5.79591836734694e-05, "loss": 1.1411, "step": 142 }, { "epoch": 0.3657289002557545, "grad_norm": 1.0444683774173515, "learning_rate": 5.836734693877551e-05, "loss": 1.1689, "step": 143 }, { "epoch": 0.36828644501278773, "grad_norm": 0.8301811817561374, "learning_rate": 5.877551020408164e-05, "loss": 1.1345, "step": 144 }, { "epoch": 0.37084398976982097, "grad_norm": 0.812249217700949, "learning_rate": 5.918367346938776e-05, "loss": 1.1466, "step": 145 }, { "epoch": 0.3734015345268542, "grad_norm": 0.7307569711274879, "learning_rate": 5.959183673469389e-05, "loss": 1.1303, "step": 146 }, { "epoch": 0.37595907928388744, "grad_norm": 0.8056769718133762, "learning_rate": 6.000000000000001e-05, "loss": 1.1675, "step": 147 }, { "epoch": 0.37851662404092073, "grad_norm": 1.2641845208914293, "learning_rate": 6.040816326530612e-05, "loss": 1.13, "step": 148 }, { "epoch": 0.38107416879795397, "grad_norm": 1.0107711010817608, "learning_rate": 6.081632653061225e-05, "loss": 1.1145, "step": 149 }, { "epoch": 0.3836317135549872, "grad_norm": 1.1073005971254093, "learning_rate": 6.122448979591836e-05, "loss": 1.1266, "step": 150 }, { "epoch": 0.38618925831202044, "grad_norm": 0.8625795523903089, "learning_rate": 6.163265306122449e-05, "loss": 1.129, "step": 151 }, { "epoch": 0.3887468030690537, "grad_norm": 1.0401585969587819, "learning_rate": 6.204081632653062e-05, "loss": 1.1361, "step": 152 }, { "epoch": 0.391304347826087, "grad_norm": 1.0394281409109958, "learning_rate": 6.244897959183675e-05, "loss": 1.1266, "step": 153 }, { "epoch": 0.3938618925831202, "grad_norm": 0.9953073831065966, "learning_rate": 6.285714285714286e-05, "loss": 1.1343, "step": 154 }, { "epoch": 0.39641943734015345, "grad_norm": 1.2177849301315296, "learning_rate": 6.326530612244899e-05, "loss": 1.1463, "step": 155 }, { "epoch": 0.3989769820971867, "grad_norm": 0.7596280219343059, "learning_rate": 6.36734693877551e-05, "loss": 1.1454, "step": 156 }, { "epoch": 0.40153452685422, "grad_norm": 1.0754506244483104, "learning_rate": 6.408163265306123e-05, "loss": 1.1616, "step": 157 }, { "epoch": 0.4040920716112532, "grad_norm": 1.1630866426285293, "learning_rate": 6.448979591836736e-05, "loss": 1.1545, "step": 158 }, { "epoch": 0.40664961636828645, "grad_norm": 0.6988890630556159, "learning_rate": 6.489795918367347e-05, "loss": 1.1052, "step": 159 }, { "epoch": 0.4092071611253197, "grad_norm": 0.7715767301674425, "learning_rate": 6.53061224489796e-05, "loss": 1.1264, "step": 160 }, { "epoch": 0.4117647058823529, "grad_norm": 0.7984070291716113, "learning_rate": 6.571428571428571e-05, "loss": 1.12, "step": 161 }, { "epoch": 0.4143222506393862, "grad_norm": 0.8560213949536746, "learning_rate": 6.612244897959184e-05, "loss": 1.1171, "step": 162 }, { "epoch": 0.41687979539641945, "grad_norm": 0.9369116479027433, "learning_rate": 6.653061224489797e-05, "loss": 1.1196, "step": 163 }, { "epoch": 0.4194373401534527, "grad_norm": 0.9843108307123993, "learning_rate": 6.693877551020408e-05, "loss": 1.1316, "step": 164 }, { "epoch": 0.4219948849104859, "grad_norm": 1.0791992478077372, "learning_rate": 6.734693877551021e-05, "loss": 1.107, "step": 165 }, { "epoch": 0.42455242966751916, "grad_norm": 1.2221610713261708, "learning_rate": 6.775510204081634e-05, "loss": 1.1294, "step": 166 }, { "epoch": 0.42710997442455245, "grad_norm": 0.9019331212289963, "learning_rate": 6.816326530612245e-05, "loss": 1.1245, "step": 167 }, { "epoch": 0.4296675191815857, "grad_norm": 0.9269989033015092, "learning_rate": 6.857142857142857e-05, "loss": 1.1338, "step": 168 }, { "epoch": 0.4322250639386189, "grad_norm": 1.0981263529686658, "learning_rate": 6.89795918367347e-05, "loss": 1.1332, "step": 169 }, { "epoch": 0.43478260869565216, "grad_norm": 0.9888336583144302, "learning_rate": 6.938775510204082e-05, "loss": 1.1127, "step": 170 }, { "epoch": 0.4373401534526854, "grad_norm": 1.3074455063450379, "learning_rate": 6.979591836734695e-05, "loss": 1.1285, "step": 171 }, { "epoch": 0.4398976982097187, "grad_norm": 0.6818203503765217, "learning_rate": 7.020408163265306e-05, "loss": 1.1223, "step": 172 }, { "epoch": 0.4424552429667519, "grad_norm": 0.8192921172095765, "learning_rate": 7.061224489795919e-05, "loss": 1.142, "step": 173 }, { "epoch": 0.44501278772378516, "grad_norm": 0.8114896652522754, "learning_rate": 7.10204081632653e-05, "loss": 1.151, "step": 174 }, { "epoch": 0.4475703324808184, "grad_norm": 0.8520710303534221, "learning_rate": 7.142857142857143e-05, "loss": 1.1235, "step": 175 }, { "epoch": 0.45012787723785164, "grad_norm": 1.1274861056526013, "learning_rate": 7.183673469387756e-05, "loss": 1.1553, "step": 176 }, { "epoch": 0.45268542199488493, "grad_norm": 1.0575599991469968, "learning_rate": 7.224489795918368e-05, "loss": 1.1281, "step": 177 }, { "epoch": 0.45524296675191817, "grad_norm": 0.9955941287548709, "learning_rate": 7.26530612244898e-05, "loss": 1.1267, "step": 178 }, { "epoch": 0.4578005115089514, "grad_norm": 1.334626282193364, "learning_rate": 7.306122448979592e-05, "loss": 1.1147, "step": 179 }, { "epoch": 0.46035805626598464, "grad_norm": 1.0475795735421864, "learning_rate": 7.346938775510205e-05, "loss": 1.1197, "step": 180 }, { "epoch": 0.4629156010230179, "grad_norm": 1.5007093885023364, "learning_rate": 7.387755102040817e-05, "loss": 1.1482, "step": 181 }, { "epoch": 0.46547314578005117, "grad_norm": 0.7288699412052543, "learning_rate": 7.42857142857143e-05, "loss": 1.1206, "step": 182 }, { "epoch": 0.4680306905370844, "grad_norm": 1.3525278294620537, "learning_rate": 7.469387755102041e-05, "loss": 1.1366, "step": 183 }, { "epoch": 0.47058823529411764, "grad_norm": 0.9907423022995694, "learning_rate": 7.510204081632654e-05, "loss": 1.1294, "step": 184 }, { "epoch": 0.4731457800511509, "grad_norm": 1.4564946558231437, "learning_rate": 7.551020408163266e-05, "loss": 1.1395, "step": 185 }, { "epoch": 0.47570332480818417, "grad_norm": 0.7257811934036715, "learning_rate": 7.591836734693878e-05, "loss": 1.1332, "step": 186 }, { "epoch": 0.4782608695652174, "grad_norm": 1.0803795052811809, "learning_rate": 7.63265306122449e-05, "loss": 1.1187, "step": 187 }, { "epoch": 0.48081841432225064, "grad_norm": 1.0690865063674833, "learning_rate": 7.673469387755103e-05, "loss": 1.1428, "step": 188 }, { "epoch": 0.4833759590792839, "grad_norm": 1.0845046051039753, "learning_rate": 7.714285714285715e-05, "loss": 1.126, "step": 189 }, { "epoch": 0.4859335038363171, "grad_norm": 0.9227022064148079, "learning_rate": 7.755102040816327e-05, "loss": 1.1262, "step": 190 }, { "epoch": 0.4884910485933504, "grad_norm": 1.031390374218521, "learning_rate": 7.79591836734694e-05, "loss": 1.0769, "step": 191 }, { "epoch": 0.49104859335038364, "grad_norm": 0.9817678976684439, "learning_rate": 7.836734693877551e-05, "loss": 1.1177, "step": 192 }, { "epoch": 0.4936061381074169, "grad_norm": 1.0891373841172232, "learning_rate": 7.877551020408164e-05, "loss": 1.0895, "step": 193 }, { "epoch": 0.4961636828644501, "grad_norm": 0.8992782943686848, "learning_rate": 7.918367346938776e-05, "loss": 1.0864, "step": 194 }, { "epoch": 0.49872122762148335, "grad_norm": 1.0356054173418965, "learning_rate": 7.959183673469389e-05, "loss": 1.1628, "step": 195 }, { "epoch": 0.5012787723785166, "grad_norm": 1.054829292229048, "learning_rate": 8e-05, "loss": 1.1215, "step": 196 }, { "epoch": 0.5038363171355499, "grad_norm": 1.1952728526697924, "learning_rate": 7.999993620333937e-05, "loss": 1.1072, "step": 197 }, { "epoch": 0.5063938618925832, "grad_norm": 1.4069207531654522, "learning_rate": 7.999974481356096e-05, "loss": 1.1389, "step": 198 }, { "epoch": 0.5089514066496164, "grad_norm": 1.0226573738564007, "learning_rate": 7.999942583127525e-05, "loss": 1.1039, "step": 199 }, { "epoch": 0.5115089514066496, "grad_norm": 1.237262581205958, "learning_rate": 7.999897925749977e-05, "loss": 1.1205, "step": 200 }, { "epoch": 0.5140664961636828, "grad_norm": 0.713505858861466, "learning_rate": 7.9998405093659e-05, "loss": 1.0991, "step": 201 }, { "epoch": 0.5166240409207161, "grad_norm": 0.8851648472236339, "learning_rate": 7.999770334158445e-05, "loss": 1.0939, "step": 202 }, { "epoch": 0.5191815856777494, "grad_norm": 1.2619572731510282, "learning_rate": 7.999687400351458e-05, "loss": 1.1263, "step": 203 }, { "epoch": 0.5217391304347826, "grad_norm": 0.650360402034941, "learning_rate": 7.999591708209482e-05, "loss": 1.0945, "step": 204 }, { "epoch": 0.5242966751918159, "grad_norm": 1.0524079760247882, "learning_rate": 7.99948325803776e-05, "loss": 1.1189, "step": 205 }, { "epoch": 0.5268542199488491, "grad_norm": 1.1840200840428867, "learning_rate": 7.999362050182232e-05, "loss": 1.1079, "step": 206 }, { "epoch": 0.5294117647058824, "grad_norm": 0.9689049896840791, "learning_rate": 7.999228085029529e-05, "loss": 1.1252, "step": 207 }, { "epoch": 0.5319693094629157, "grad_norm": 1.3635698843008814, "learning_rate": 7.999081363006977e-05, "loss": 1.1119, "step": 208 }, { "epoch": 0.5345268542199488, "grad_norm": 0.7404269657383984, "learning_rate": 7.998921884582596e-05, "loss": 1.0998, "step": 209 }, { "epoch": 0.5370843989769821, "grad_norm": 0.901502531204049, "learning_rate": 7.998749650265096e-05, "loss": 1.0953, "step": 210 }, { "epoch": 0.5396419437340153, "grad_norm": 0.8855993913877257, "learning_rate": 7.998564660603873e-05, "loss": 1.1146, "step": 211 }, { "epoch": 0.5421994884910486, "grad_norm": 0.6583772614283104, "learning_rate": 7.998366916189016e-05, "loss": 1.0821, "step": 212 }, { "epoch": 0.5447570332480819, "grad_norm": 0.6745326521828335, "learning_rate": 7.998156417651297e-05, "loss": 1.1258, "step": 213 }, { "epoch": 0.5473145780051151, "grad_norm": 0.9038994334777463, "learning_rate": 7.997933165662167e-05, "loss": 1.0766, "step": 214 }, { "epoch": 0.5498721227621484, "grad_norm": 0.9031283752629393, "learning_rate": 7.997697160933767e-05, "loss": 1.0825, "step": 215 }, { "epoch": 0.5524296675191815, "grad_norm": 1.3656136727514145, "learning_rate": 7.99744840421891e-05, "loss": 1.1183, "step": 216 }, { "epoch": 0.5549872122762148, "grad_norm": 0.9662280283806104, "learning_rate": 7.997186896311091e-05, "loss": 1.0898, "step": 217 }, { "epoch": 0.5575447570332481, "grad_norm": 1.1227766801563628, "learning_rate": 7.996912638044474e-05, "loss": 1.0641, "step": 218 }, { "epoch": 0.5601023017902813, "grad_norm": 0.5938083971975682, "learning_rate": 7.996625630293898e-05, "loss": 1.087, "step": 219 }, { "epoch": 0.5626598465473146, "grad_norm": 0.830004505260378, "learning_rate": 7.996325873974871e-05, "loss": 1.1414, "step": 220 }, { "epoch": 0.5652173913043478, "grad_norm": 0.688470818535481, "learning_rate": 7.996013370043562e-05, "loss": 1.0817, "step": 221 }, { "epoch": 0.5677749360613811, "grad_norm": 0.9773236602394363, "learning_rate": 7.995688119496811e-05, "loss": 1.0851, "step": 222 }, { "epoch": 0.5703324808184144, "grad_norm": 1.584952668299985, "learning_rate": 7.99535012337211e-05, "loss": 1.0933, "step": 223 }, { "epoch": 0.5728900255754475, "grad_norm": 0.8022915092257511, "learning_rate": 7.99499938274761e-05, "loss": 1.1196, "step": 224 }, { "epoch": 0.5754475703324808, "grad_norm": 1.3582138669069936, "learning_rate": 7.994635898742118e-05, "loss": 1.1199, "step": 225 }, { "epoch": 0.578005115089514, "grad_norm": 1.3157395032889772, "learning_rate": 7.994259672515083e-05, "loss": 1.0922, "step": 226 }, { "epoch": 0.5805626598465473, "grad_norm": 1.0747983662822311, "learning_rate": 7.993870705266608e-05, "loss": 1.1012, "step": 227 }, { "epoch": 0.5831202046035806, "grad_norm": 0.9000554867539409, "learning_rate": 7.99346899823743e-05, "loss": 1.086, "step": 228 }, { "epoch": 0.5856777493606138, "grad_norm": 1.1498952956983293, "learning_rate": 7.993054552708931e-05, "loss": 1.1069, "step": 229 }, { "epoch": 0.5882352941176471, "grad_norm": 0.9821705283797745, "learning_rate": 7.99262737000312e-05, "loss": 1.1105, "step": 230 }, { "epoch": 0.5907928388746803, "grad_norm": 1.2035932083418026, "learning_rate": 7.992187451482642e-05, "loss": 1.1132, "step": 231 }, { "epoch": 0.5933503836317136, "grad_norm": 0.9030392431196859, "learning_rate": 7.991734798550758e-05, "loss": 1.0944, "step": 232 }, { "epoch": 0.5959079283887468, "grad_norm": 0.901349070838302, "learning_rate": 7.99126941265136e-05, "loss": 1.1008, "step": 233 }, { "epoch": 0.59846547314578, "grad_norm": 0.9137700738163895, "learning_rate": 7.990791295268949e-05, "loss": 1.0579, "step": 234 }, { "epoch": 0.6010230179028133, "grad_norm": 0.9571473627952674, "learning_rate": 7.99030044792864e-05, "loss": 1.0766, "step": 235 }, { "epoch": 0.6035805626598465, "grad_norm": 0.9060175971731932, "learning_rate": 7.989796872196156e-05, "loss": 1.1069, "step": 236 }, { "epoch": 0.6061381074168798, "grad_norm": 0.9882670857555712, "learning_rate": 7.989280569677817e-05, "loss": 1.0935, "step": 237 }, { "epoch": 0.6086956521739131, "grad_norm": 0.8147636024776135, "learning_rate": 7.988751542020543e-05, "loss": 1.0755, "step": 238 }, { "epoch": 0.6112531969309463, "grad_norm": 0.9201311976288161, "learning_rate": 7.988209790911844e-05, "loss": 1.1183, "step": 239 }, { "epoch": 0.6138107416879796, "grad_norm": 1.0312318256040205, "learning_rate": 7.987655318079816e-05, "loss": 1.084, "step": 240 }, { "epoch": 0.6163682864450127, "grad_norm": 1.0473461685811893, "learning_rate": 7.987088125293133e-05, "loss": 1.0765, "step": 241 }, { "epoch": 0.618925831202046, "grad_norm": 1.1909721824510273, "learning_rate": 7.986508214361047e-05, "loss": 1.1047, "step": 242 }, { "epoch": 0.6214833759590793, "grad_norm": 0.6527590513693277, "learning_rate": 7.985915587133375e-05, "loss": 1.0975, "step": 243 }, { "epoch": 0.6240409207161125, "grad_norm": 0.6644980276462075, "learning_rate": 7.985310245500502e-05, "loss": 1.1448, "step": 244 }, { "epoch": 0.6265984654731458, "grad_norm": 0.8238984428442946, "learning_rate": 7.984692191393364e-05, "loss": 1.1053, "step": 245 }, { "epoch": 0.629156010230179, "grad_norm": 0.9188194695868873, "learning_rate": 7.984061426783452e-05, "loss": 1.1014, "step": 246 }, { "epoch": 0.6317135549872123, "grad_norm": 0.7164005648194159, "learning_rate": 7.9834179536828e-05, "loss": 1.0916, "step": 247 }, { "epoch": 0.6342710997442456, "grad_norm": 0.6529981781161239, "learning_rate": 7.982761774143978e-05, "loss": 1.1379, "step": 248 }, { "epoch": 0.6368286445012787, "grad_norm": 0.7893937427043366, "learning_rate": 7.98209289026009e-05, "loss": 1.1201, "step": 249 }, { "epoch": 0.639386189258312, "grad_norm": 0.6695991863515521, "learning_rate": 7.981411304164765e-05, "loss": 1.0988, "step": 250 }, { "epoch": 0.6419437340153452, "grad_norm": 0.8732562498954997, "learning_rate": 7.980717018032147e-05, "loss": 1.1052, "step": 251 }, { "epoch": 0.6445012787723785, "grad_norm": 1.1886830926779333, "learning_rate": 7.980010034076893e-05, "loss": 1.1121, "step": 252 }, { "epoch": 0.6470588235294118, "grad_norm": 0.7817538943389751, "learning_rate": 7.979290354554166e-05, "loss": 1.089, "step": 253 }, { "epoch": 0.649616368286445, "grad_norm": 0.7904438046538128, "learning_rate": 7.978557981759622e-05, "loss": 1.1395, "step": 254 }, { "epoch": 0.6521739130434783, "grad_norm": 0.7725929256674784, "learning_rate": 7.977812918029407e-05, "loss": 1.0977, "step": 255 }, { "epoch": 0.6547314578005116, "grad_norm": 0.7520483951987087, "learning_rate": 7.977055165740151e-05, "loss": 1.0859, "step": 256 }, { "epoch": 0.6572890025575447, "grad_norm": 0.6629038851395048, "learning_rate": 7.976284727308958e-05, "loss": 1.1208, "step": 257 }, { "epoch": 0.659846547314578, "grad_norm": 0.7911526422920417, "learning_rate": 7.975501605193396e-05, "loss": 1.0834, "step": 258 }, { "epoch": 0.6624040920716112, "grad_norm": 0.9171702660893679, "learning_rate": 7.974705801891496e-05, "loss": 1.1104, "step": 259 }, { "epoch": 0.6649616368286445, "grad_norm": 1.0462534089765212, "learning_rate": 7.973897319941736e-05, "loss": 1.0931, "step": 260 }, { "epoch": 0.6675191815856778, "grad_norm": 1.1787829597415223, "learning_rate": 7.97307616192304e-05, "loss": 1.0654, "step": 261 }, { "epoch": 0.670076726342711, "grad_norm": 0.8204297503121395, "learning_rate": 7.972242330454762e-05, "loss": 1.0806, "step": 262 }, { "epoch": 0.6726342710997443, "grad_norm": 0.8044857551775063, "learning_rate": 7.971395828196688e-05, "loss": 1.1243, "step": 263 }, { "epoch": 0.6751918158567775, "grad_norm": 0.7741196371916483, "learning_rate": 7.970536657849018e-05, "loss": 1.102, "step": 264 }, { "epoch": 0.6777493606138107, "grad_norm": 0.7513622426525168, "learning_rate": 7.969664822152362e-05, "loss": 1.1313, "step": 265 }, { "epoch": 0.680306905370844, "grad_norm": 0.7139289605653666, "learning_rate": 7.96878032388773e-05, "loss": 1.0976, "step": 266 }, { "epoch": 0.6828644501278772, "grad_norm": 0.658091891766218, "learning_rate": 7.967883165876524e-05, "loss": 1.0522, "step": 267 }, { "epoch": 0.6854219948849105, "grad_norm": 0.5913576930217892, "learning_rate": 7.966973350980528e-05, "loss": 1.0948, "step": 268 }, { "epoch": 0.6879795396419437, "grad_norm": 0.5132302401343739, "learning_rate": 7.9660508821019e-05, "loss": 1.0707, "step": 269 }, { "epoch": 0.690537084398977, "grad_norm": 0.5780733190718835, "learning_rate": 7.965115762183163e-05, "loss": 1.1105, "step": 270 }, { "epoch": 0.6930946291560103, "grad_norm": 0.6018356145185099, "learning_rate": 7.96416799420719e-05, "loss": 1.1096, "step": 271 }, { "epoch": 0.6956521739130435, "grad_norm": 0.6205282943178246, "learning_rate": 7.963207581197207e-05, "loss": 1.0721, "step": 272 }, { "epoch": 0.6982097186700768, "grad_norm": 0.6061678652911838, "learning_rate": 7.962234526216767e-05, "loss": 1.0879, "step": 273 }, { "epoch": 0.7007672634271099, "grad_norm": 0.6945934405726988, "learning_rate": 7.961248832369754e-05, "loss": 1.1085, "step": 274 }, { "epoch": 0.7033248081841432, "grad_norm": 0.5760159589741866, "learning_rate": 7.96025050280037e-05, "loss": 1.0864, "step": 275 }, { "epoch": 0.7058823529411765, "grad_norm": 0.5565501271336987, "learning_rate": 7.959239540693115e-05, "loss": 1.1034, "step": 276 }, { "epoch": 0.7084398976982097, "grad_norm": 0.6676422629214439, "learning_rate": 7.958215949272791e-05, "loss": 1.1176, "step": 277 }, { "epoch": 0.710997442455243, "grad_norm": 0.7153664297383551, "learning_rate": 7.957179731804485e-05, "loss": 1.0847, "step": 278 }, { "epoch": 0.7135549872122762, "grad_norm": 0.5930553863066046, "learning_rate": 7.956130891593556e-05, "loss": 1.114, "step": 279 }, { "epoch": 0.7161125319693095, "grad_norm": 0.5888284152918564, "learning_rate": 7.955069431985632e-05, "loss": 1.0802, "step": 280 }, { "epoch": 0.7186700767263428, "grad_norm": 0.7930110052547374, "learning_rate": 7.953995356366588e-05, "loss": 1.0739, "step": 281 }, { "epoch": 0.7212276214833759, "grad_norm": 0.6696001227538381, "learning_rate": 7.952908668162549e-05, "loss": 1.1003, "step": 282 }, { "epoch": 0.7237851662404092, "grad_norm": 0.4757426938641106, "learning_rate": 7.951809370839867e-05, "loss": 1.0299, "step": 283 }, { "epoch": 0.7263427109974424, "grad_norm": 0.5474713187788883, "learning_rate": 7.950697467905118e-05, "loss": 1.0951, "step": 284 }, { "epoch": 0.7289002557544757, "grad_norm": 0.5071386457506744, "learning_rate": 7.949572962905086e-05, "loss": 1.0772, "step": 285 }, { "epoch": 0.731457800511509, "grad_norm": 0.5247168975203097, "learning_rate": 7.948435859426755e-05, "loss": 1.1039, "step": 286 }, { "epoch": 0.7340153452685422, "grad_norm": 0.6300459744494643, "learning_rate": 7.947286161097294e-05, "loss": 1.1019, "step": 287 }, { "epoch": 0.7365728900255755, "grad_norm": 0.8793621293918455, "learning_rate": 7.94612387158405e-05, "loss": 1.0641, "step": 288 }, { "epoch": 0.7391304347826086, "grad_norm": 1.354355970537, "learning_rate": 7.944948994594532e-05, "loss": 1.0976, "step": 289 }, { "epoch": 0.7416879795396419, "grad_norm": 0.811793129507913, "learning_rate": 7.943761533876402e-05, "loss": 1.1066, "step": 290 }, { "epoch": 0.7442455242966752, "grad_norm": 0.6985981549179727, "learning_rate": 7.94256149321746e-05, "loss": 1.0889, "step": 291 }, { "epoch": 0.7468030690537084, "grad_norm": 0.6979422293870231, "learning_rate": 7.941348876445636e-05, "loss": 1.0538, "step": 292 }, { "epoch": 0.7493606138107417, "grad_norm": 0.5854843946594773, "learning_rate": 7.940123687428975e-05, "loss": 1.0817, "step": 293 }, { "epoch": 0.7519181585677749, "grad_norm": 0.7081721412507801, "learning_rate": 7.938885930075628e-05, "loss": 1.0781, "step": 294 }, { "epoch": 0.7544757033248082, "grad_norm": 0.7189613701722457, "learning_rate": 7.937635608333829e-05, "loss": 1.0709, "step": 295 }, { "epoch": 0.7570332480818415, "grad_norm": 0.9612156787986404, "learning_rate": 7.936372726191899e-05, "loss": 1.0792, "step": 296 }, { "epoch": 0.7595907928388747, "grad_norm": 1.5095702200934382, "learning_rate": 7.935097287678222e-05, "loss": 1.1203, "step": 297 }, { "epoch": 0.7621483375959079, "grad_norm": 0.49810286262739045, "learning_rate": 7.933809296861232e-05, "loss": 1.0951, "step": 298 }, { "epoch": 0.7647058823529411, "grad_norm": 1.094271200719579, "learning_rate": 7.932508757849403e-05, "loss": 1.1087, "step": 299 }, { "epoch": 0.7672634271099744, "grad_norm": 1.6990794876777922, "learning_rate": 7.931195674791242e-05, "loss": 1.0713, "step": 300 }, { "epoch": 0.7698209718670077, "grad_norm": 0.6778774182184969, "learning_rate": 7.929870051875261e-05, "loss": 1.1025, "step": 301 }, { "epoch": 0.7723785166240409, "grad_norm": 1.9523378206319744, "learning_rate": 7.928531893329977e-05, "loss": 1.1099, "step": 302 }, { "epoch": 0.7749360613810742, "grad_norm": 1.0174747598205622, "learning_rate": 7.927181203423893e-05, "loss": 1.0761, "step": 303 }, { "epoch": 0.7774936061381074, "grad_norm": 2.097344028961776, "learning_rate": 7.925817986465483e-05, "loss": 1.0964, "step": 304 }, { "epoch": 0.7800511508951407, "grad_norm": 1.7588955371199144, "learning_rate": 7.924442246803183e-05, "loss": 1.1114, "step": 305 }, { "epoch": 0.782608695652174, "grad_norm": 1.1038887334502365, "learning_rate": 7.923053988825372e-05, "loss": 1.0863, "step": 306 }, { "epoch": 0.7851662404092071, "grad_norm": 1.2204166281857975, "learning_rate": 7.92165321696036e-05, "loss": 1.0609, "step": 307 }, { "epoch": 0.7877237851662404, "grad_norm": 0.6613406186209091, "learning_rate": 7.920239935676377e-05, "loss": 1.0832, "step": 308 }, { "epoch": 0.7902813299232737, "grad_norm": 0.9099805345560761, "learning_rate": 7.918814149481553e-05, "loss": 1.1062, "step": 309 }, { "epoch": 0.7928388746803069, "grad_norm": 0.8966794823734242, "learning_rate": 7.917375862923908e-05, "loss": 1.0699, "step": 310 }, { "epoch": 0.7953964194373402, "grad_norm": 0.6431878954941771, "learning_rate": 7.915925080591338e-05, "loss": 1.0999, "step": 311 }, { "epoch": 0.7979539641943734, "grad_norm": 0.7980040266876918, "learning_rate": 7.914461807111595e-05, "loss": 1.1169, "step": 312 }, { "epoch": 0.8005115089514067, "grad_norm": 0.5792659205085796, "learning_rate": 7.912986047152278e-05, "loss": 1.0783, "step": 313 }, { "epoch": 0.80306905370844, "grad_norm": 0.7059169343501349, "learning_rate": 7.911497805420811e-05, "loss": 1.0823, "step": 314 }, { "epoch": 0.8056265984654731, "grad_norm": 0.5933585988646929, "learning_rate": 7.909997086664443e-05, "loss": 1.0748, "step": 315 }, { "epoch": 0.8081841432225064, "grad_norm": 0.6845366174297771, "learning_rate": 7.908483895670209e-05, "loss": 1.071, "step": 316 }, { "epoch": 0.8107416879795396, "grad_norm": 0.6043958213664847, "learning_rate": 7.906958237264942e-05, "loss": 1.0984, "step": 317 }, { "epoch": 0.8132992327365729, "grad_norm": 0.8065799320997168, "learning_rate": 7.905420116315234e-05, "loss": 1.0908, "step": 318 }, { "epoch": 0.8158567774936062, "grad_norm": 0.9613882476117077, "learning_rate": 7.903869537727437e-05, "loss": 1.0503, "step": 319 }, { "epoch": 0.8184143222506394, "grad_norm": 0.8851238240180742, "learning_rate": 7.902306506447634e-05, "loss": 1.0783, "step": 320 }, { "epoch": 0.8209718670076727, "grad_norm": 0.9671778018065025, "learning_rate": 7.900731027461635e-05, "loss": 1.057, "step": 321 }, { "epoch": 0.8235294117647058, "grad_norm": 1.0461389584721013, "learning_rate": 7.899143105794958e-05, "loss": 1.087, "step": 322 }, { "epoch": 0.8260869565217391, "grad_norm": 0.8134854113246186, "learning_rate": 7.897542746512804e-05, "loss": 1.1064, "step": 323 }, { "epoch": 0.8286445012787724, "grad_norm": 0.7093110104105061, "learning_rate": 7.895929954720053e-05, "loss": 1.1046, "step": 324 }, { "epoch": 0.8312020460358056, "grad_norm": 0.4793356093430287, "learning_rate": 7.894304735561243e-05, "loss": 1.0652, "step": 325 }, { "epoch": 0.8337595907928389, "grad_norm": 0.5002076354784437, "learning_rate": 7.89266709422055e-05, "loss": 1.0756, "step": 326 }, { "epoch": 0.8363171355498721, "grad_norm": 0.5414464668464288, "learning_rate": 7.891017035921778e-05, "loss": 1.1155, "step": 327 }, { "epoch": 0.8388746803069054, "grad_norm": 0.6820350461047787, "learning_rate": 7.889354565928337e-05, "loss": 1.0879, "step": 328 }, { "epoch": 0.8414322250639387, "grad_norm": 0.8316642019332523, "learning_rate": 7.887679689543227e-05, "loss": 1.1265, "step": 329 }, { "epoch": 0.8439897698209718, "grad_norm": 0.8446501746399396, "learning_rate": 7.885992412109026e-05, "loss": 1.0971, "step": 330 }, { "epoch": 0.8465473145780051, "grad_norm": 0.8372575100501372, "learning_rate": 7.884292739007866e-05, "loss": 1.0698, "step": 331 }, { "epoch": 0.8491048593350383, "grad_norm": 0.751312270457101, "learning_rate": 7.882580675661422e-05, "loss": 1.0612, "step": 332 }, { "epoch": 0.8516624040920716, "grad_norm": 0.8153412859957982, "learning_rate": 7.88085622753089e-05, "loss": 1.1092, "step": 333 }, { "epoch": 0.8542199488491049, "grad_norm": 0.7887032728971717, "learning_rate": 7.879119400116969e-05, "loss": 1.0411, "step": 334 }, { "epoch": 0.8567774936061381, "grad_norm": 0.7667290672650495, "learning_rate": 7.877370198959852e-05, "loss": 1.0671, "step": 335 }, { "epoch": 0.8593350383631714, "grad_norm": 0.6040917626312119, "learning_rate": 7.875608629639195e-05, "loss": 1.067, "step": 336 }, { "epoch": 0.8618925831202046, "grad_norm": 0.5108422350504948, "learning_rate": 7.873834697774114e-05, "loss": 1.0896, "step": 337 }, { "epoch": 0.8644501278772379, "grad_norm": 0.6073358430679844, "learning_rate": 7.872048409023153e-05, "loss": 1.0945, "step": 338 }, { "epoch": 0.8670076726342711, "grad_norm": 0.5811775734604341, "learning_rate": 7.870249769084274e-05, "loss": 1.0599, "step": 339 }, { "epoch": 0.8695652173913043, "grad_norm": 0.4942443213793738, "learning_rate": 7.868438783694841e-05, "loss": 1.0984, "step": 340 }, { "epoch": 0.8721227621483376, "grad_norm": 0.44691873435841717, "learning_rate": 7.866615458631591e-05, "loss": 1.0551, "step": 341 }, { "epoch": 0.8746803069053708, "grad_norm": 0.49127713203197954, "learning_rate": 7.864779799710631e-05, "loss": 1.0821, "step": 342 }, { "epoch": 0.8772378516624041, "grad_norm": 0.42857623809659445, "learning_rate": 7.862931812787404e-05, "loss": 1.0802, "step": 343 }, { "epoch": 0.8797953964194374, "grad_norm": 0.5052119789372929, "learning_rate": 7.86107150375668e-05, "loss": 1.0876, "step": 344 }, { "epoch": 0.8823529411764706, "grad_norm": 0.38100935660159413, "learning_rate": 7.859198878552536e-05, "loss": 1.068, "step": 345 }, { "epoch": 0.8849104859335039, "grad_norm": 0.44991333074166684, "learning_rate": 7.85731394314833e-05, "loss": 1.1331, "step": 346 }, { "epoch": 0.887468030690537, "grad_norm": 0.45213307645529244, "learning_rate": 7.855416703556693e-05, "loss": 1.0161, "step": 347 }, { "epoch": 0.8900255754475703, "grad_norm": 0.4159819987382006, "learning_rate": 7.853507165829504e-05, "loss": 1.0811, "step": 348 }, { "epoch": 0.8925831202046036, "grad_norm": 0.5486443431245583, "learning_rate": 7.851585336057868e-05, "loss": 1.0653, "step": 349 }, { "epoch": 0.8951406649616368, "grad_norm": 0.6862683732800593, "learning_rate": 7.849651220372101e-05, "loss": 1.0867, "step": 350 }, { "epoch": 0.8976982097186701, "grad_norm": 0.6981991056672471, "learning_rate": 7.847704824941709e-05, "loss": 1.105, "step": 351 }, { "epoch": 0.9002557544757033, "grad_norm": 0.785234109714766, "learning_rate": 7.84574615597537e-05, "loss": 1.0596, "step": 352 }, { "epoch": 0.9028132992327366, "grad_norm": 0.9770924073440606, "learning_rate": 7.84377521972091e-05, "loss": 1.0823, "step": 353 }, { "epoch": 0.9053708439897699, "grad_norm": 1.1898436603817308, "learning_rate": 7.841792022465286e-05, "loss": 1.0763, "step": 354 }, { "epoch": 0.907928388746803, "grad_norm": 0.8386259285895766, "learning_rate": 7.839796570534566e-05, "loss": 1.0542, "step": 355 }, { "epoch": 0.9104859335038363, "grad_norm": 0.6389869741905022, "learning_rate": 7.83778887029391e-05, "loss": 1.0668, "step": 356 }, { "epoch": 0.9130434782608695, "grad_norm": 0.5583652456462365, "learning_rate": 7.835768928147545e-05, "loss": 1.0759, "step": 357 }, { "epoch": 0.9156010230179028, "grad_norm": 0.571315368558641, "learning_rate": 7.83373675053875e-05, "loss": 1.0976, "step": 358 }, { "epoch": 0.9181585677749361, "grad_norm": 0.6447121021663385, "learning_rate": 7.831692343949832e-05, "loss": 1.0792, "step": 359 }, { "epoch": 0.9207161125319693, "grad_norm": 0.7085586190843103, "learning_rate": 7.829635714902108e-05, "loss": 1.0461, "step": 360 }, { "epoch": 0.9232736572890026, "grad_norm": 0.8729805375627833, "learning_rate": 7.827566869955878e-05, "loss": 1.0957, "step": 361 }, { "epoch": 0.9258312020460358, "grad_norm": 1.1209022936512785, "learning_rate": 7.825485815710415e-05, "loss": 1.0246, "step": 362 }, { "epoch": 0.928388746803069, "grad_norm": 1.0159476142523682, "learning_rate": 7.823392558803933e-05, "loss": 1.0859, "step": 363 }, { "epoch": 0.9309462915601023, "grad_norm": 0.8822113574613138, "learning_rate": 7.821287105913571e-05, "loss": 1.069, "step": 364 }, { "epoch": 0.9335038363171355, "grad_norm": 0.8276272022971064, "learning_rate": 7.819169463755376e-05, "loss": 1.0857, "step": 365 }, { "epoch": 0.9360613810741688, "grad_norm": 0.814599705940635, "learning_rate": 7.817039639084271e-05, "loss": 1.1084, "step": 366 }, { "epoch": 0.9386189258312021, "grad_norm": 0.9444127014112077, "learning_rate": 7.814897638694039e-05, "loss": 1.0864, "step": 367 }, { "epoch": 0.9411764705882353, "grad_norm": 0.9252499734817574, "learning_rate": 7.812743469417306e-05, "loss": 1.0797, "step": 368 }, { "epoch": 0.9437340153452686, "grad_norm": 0.8304614201741923, "learning_rate": 7.810577138125512e-05, "loss": 1.074, "step": 369 }, { "epoch": 0.9462915601023018, "grad_norm": 0.7597853417158138, "learning_rate": 7.808398651728891e-05, "loss": 1.0704, "step": 370 }, { "epoch": 0.948849104859335, "grad_norm": 0.6492633689174822, "learning_rate": 7.806208017176453e-05, "loss": 1.0978, "step": 371 }, { "epoch": 0.9514066496163683, "grad_norm": 0.7103188739776854, "learning_rate": 7.804005241455954e-05, "loss": 1.0799, "step": 372 }, { "epoch": 0.9539641943734015, "grad_norm": 0.8080045645046198, "learning_rate": 7.801790331593882e-05, "loss": 1.0847, "step": 373 }, { "epoch": 0.9565217391304348, "grad_norm": 0.7941792943654038, "learning_rate": 7.799563294655429e-05, "loss": 1.0843, "step": 374 }, { "epoch": 0.959079283887468, "grad_norm": 0.5866783302080361, "learning_rate": 7.797324137744472e-05, "loss": 1.1051, "step": 375 }, { "epoch": 0.9616368286445013, "grad_norm": 0.4359647133720881, "learning_rate": 7.795072868003546e-05, "loss": 1.0541, "step": 376 }, { "epoch": 0.9641943734015346, "grad_norm": 0.5857374136808916, "learning_rate": 7.792809492613828e-05, "loss": 1.0867, "step": 377 }, { "epoch": 0.9667519181585678, "grad_norm": 0.6385407081997835, "learning_rate": 7.790534018795106e-05, "loss": 1.0925, "step": 378 }, { "epoch": 0.969309462915601, "grad_norm": 0.6434152574400142, "learning_rate": 7.788246453805762e-05, "loss": 1.0609, "step": 379 }, { "epoch": 0.9718670076726342, "grad_norm": 0.6915297988962398, "learning_rate": 7.785946804942746e-05, "loss": 1.0443, "step": 380 }, { "epoch": 0.9744245524296675, "grad_norm": 0.7443780294295859, "learning_rate": 7.783635079541553e-05, "loss": 1.0742, "step": 381 }, { "epoch": 0.9769820971867008, "grad_norm": 0.7984905238389998, "learning_rate": 7.781311284976204e-05, "loss": 1.0784, "step": 382 }, { "epoch": 0.979539641943734, "grad_norm": 0.7209084804370026, "learning_rate": 7.778975428659213e-05, "loss": 1.0828, "step": 383 }, { "epoch": 0.9820971867007673, "grad_norm": 0.5272873910092764, "learning_rate": 7.776627518041573e-05, "loss": 1.0359, "step": 384 }, { "epoch": 0.9846547314578005, "grad_norm": 0.5081491988480245, "learning_rate": 7.774267560612725e-05, "loss": 1.1162, "step": 385 }, { "epoch": 0.9872122762148338, "grad_norm": 0.47763924105204525, "learning_rate": 7.771895563900542e-05, "loss": 1.0707, "step": 386 }, { "epoch": 0.989769820971867, "grad_norm": 0.4293473992712172, "learning_rate": 7.769511535471294e-05, "loss": 1.0835, "step": 387 }, { "epoch": 0.9923273657289002, "grad_norm": 0.4033861632452256, "learning_rate": 7.767115482929638e-05, "loss": 1.0974, "step": 388 }, { "epoch": 0.9948849104859335, "grad_norm": 0.5275929751722469, "learning_rate": 7.764707413918578e-05, "loss": 1.0806, "step": 389 }, { "epoch": 0.9974424552429667, "grad_norm": 0.5525995164607961, "learning_rate": 7.762287336119453e-05, "loss": 1.0609, "step": 390 }, { "epoch": 1.0, "grad_norm": 0.573425625533648, "learning_rate": 7.759855257251907e-05, "loss": 1.0896, "step": 391 }, { "epoch": 1.0025575447570332, "grad_norm": 0.6121282952092544, "learning_rate": 7.757411185073866e-05, "loss": 1.0565, "step": 392 }, { "epoch": 1.0051150895140666, "grad_norm": 0.6487069044059379, "learning_rate": 7.754955127381512e-05, "loss": 1.0719, "step": 393 }, { "epoch": 1.0076726342710998, "grad_norm": 0.7649726943677306, "learning_rate": 7.75248709200926e-05, "loss": 1.044, "step": 394 }, { "epoch": 1.010230179028133, "grad_norm": 1.0004112774028269, "learning_rate": 7.750007086829729e-05, "loss": 1.0534, "step": 395 }, { "epoch": 1.0127877237851663, "grad_norm": 1.3267806534444309, "learning_rate": 7.747515119753721e-05, "loss": 1.0731, "step": 396 }, { "epoch": 1.0153452685421995, "grad_norm": 0.535485816275368, "learning_rate": 7.745011198730196e-05, "loss": 1.0712, "step": 397 }, { "epoch": 1.0179028132992327, "grad_norm": 0.6246263933300512, "learning_rate": 7.742495331746246e-05, "loss": 1.0622, "step": 398 }, { "epoch": 1.020460358056266, "grad_norm": 1.2351151058960912, "learning_rate": 7.739967526827063e-05, "loss": 1.0357, "step": 399 }, { "epoch": 1.0230179028132993, "grad_norm": 0.9704077261446196, "learning_rate": 7.737427792035923e-05, "loss": 1.0885, "step": 400 }, { "epoch": 1.0255754475703325, "grad_norm": 0.8543383222043499, "learning_rate": 7.73487613547416e-05, "loss": 1.0677, "step": 401 }, { "epoch": 1.0281329923273657, "grad_norm": 0.5879183387167717, "learning_rate": 7.732312565281128e-05, "loss": 1.0844, "step": 402 }, { "epoch": 1.030690537084399, "grad_norm": 0.4537293576835104, "learning_rate": 7.72973708963419e-05, "loss": 1.0497, "step": 403 }, { "epoch": 1.0332480818414322, "grad_norm": 0.5939375825104586, "learning_rate": 7.727149716748681e-05, "loss": 1.0572, "step": 404 }, { "epoch": 1.0358056265984654, "grad_norm": 0.6995057421533399, "learning_rate": 7.724550454877892e-05, "loss": 1.0549, "step": 405 }, { "epoch": 1.0383631713554988, "grad_norm": 0.7494681155888331, "learning_rate": 7.721939312313033e-05, "loss": 1.0869, "step": 406 }, { "epoch": 1.040920716112532, "grad_norm": 0.7643538285835613, "learning_rate": 7.71931629738321e-05, "loss": 1.022, "step": 407 }, { "epoch": 1.0434782608695652, "grad_norm": 0.6356534653615626, "learning_rate": 7.716681418455409e-05, "loss": 1.0529, "step": 408 }, { "epoch": 1.0460358056265984, "grad_norm": 0.5605358498748841, "learning_rate": 7.714034683934446e-05, "loss": 1.034, "step": 409 }, { "epoch": 1.0485933503836318, "grad_norm": 0.5583980180495551, "learning_rate": 7.711376102262968e-05, "loss": 1.0282, "step": 410 }, { "epoch": 1.051150895140665, "grad_norm": 0.6738348730192517, "learning_rate": 7.708705681921405e-05, "loss": 1.0553, "step": 411 }, { "epoch": 1.0537084398976981, "grad_norm": 0.5298166396458623, "learning_rate": 7.706023431427951e-05, "loss": 1.0365, "step": 412 }, { "epoch": 1.0562659846547315, "grad_norm": 0.4983596119303216, "learning_rate": 7.703329359338538e-05, "loss": 1.0437, "step": 413 }, { "epoch": 1.0588235294117647, "grad_norm": 0.5960572452353167, "learning_rate": 7.700623474246806e-05, "loss": 1.0898, "step": 414 }, { "epoch": 1.061381074168798, "grad_norm": 0.5540140213615853, "learning_rate": 7.697905784784077e-05, "loss": 1.061, "step": 415 }, { "epoch": 1.0639386189258313, "grad_norm": 0.6096623152113402, "learning_rate": 7.695176299619327e-05, "loss": 1.0433, "step": 416 }, { "epoch": 1.0664961636828645, "grad_norm": 0.5729999229219644, "learning_rate": 7.692435027459157e-05, "loss": 1.0268, "step": 417 }, { "epoch": 1.0690537084398977, "grad_norm": 0.516351992871556, "learning_rate": 7.689681977047767e-05, "loss": 1.0372, "step": 418 }, { "epoch": 1.0716112531969308, "grad_norm": 0.5409419169165403, "learning_rate": 7.68691715716693e-05, "loss": 1.0924, "step": 419 }, { "epoch": 1.0741687979539642, "grad_norm": 0.481066894118381, "learning_rate": 7.684140576635959e-05, "loss": 1.0415, "step": 420 }, { "epoch": 1.0767263427109974, "grad_norm": 0.4798293644502134, "learning_rate": 7.68135224431168e-05, "loss": 1.0556, "step": 421 }, { "epoch": 1.0792838874680306, "grad_norm": 0.4778706340999981, "learning_rate": 7.678552169088411e-05, "loss": 1.0458, "step": 422 }, { "epoch": 1.081841432225064, "grad_norm": 0.47353441131918134, "learning_rate": 7.675740359897923e-05, "loss": 1.0314, "step": 423 }, { "epoch": 1.0843989769820972, "grad_norm": 0.6103257451747757, "learning_rate": 7.672916825709417e-05, "loss": 1.0779, "step": 424 }, { "epoch": 1.0869565217391304, "grad_norm": 0.6787803255539844, "learning_rate": 7.670081575529499e-05, "loss": 1.0412, "step": 425 }, { "epoch": 1.0895140664961638, "grad_norm": 0.5996475603666427, "learning_rate": 7.667234618402139e-05, "loss": 1.0312, "step": 426 }, { "epoch": 1.092071611253197, "grad_norm": 0.6420838832118385, "learning_rate": 7.664375963408657e-05, "loss": 1.0559, "step": 427 }, { "epoch": 1.0946291560102301, "grad_norm": 0.6257260302049759, "learning_rate": 7.661505619667685e-05, "loss": 1.0367, "step": 428 }, { "epoch": 1.0971867007672633, "grad_norm": 0.6078916092596967, "learning_rate": 7.658623596335141e-05, "loss": 1.0662, "step": 429 }, { "epoch": 1.0997442455242967, "grad_norm": 0.6869715214996625, "learning_rate": 7.655729902604198e-05, "loss": 1.0426, "step": 430 }, { "epoch": 1.10230179028133, "grad_norm": 0.7202490835032408, "learning_rate": 7.652824547705255e-05, "loss": 1.0724, "step": 431 }, { "epoch": 1.104859335038363, "grad_norm": 0.6798833175024812, "learning_rate": 7.64990754090591e-05, "loss": 1.0899, "step": 432 }, { "epoch": 1.1074168797953965, "grad_norm": 0.5492648806120208, "learning_rate": 7.646978891510927e-05, "loss": 1.0671, "step": 433 }, { "epoch": 1.1099744245524297, "grad_norm": 0.502990998641414, "learning_rate": 7.64403860886221e-05, "loss": 1.0558, "step": 434 }, { "epoch": 1.1125319693094629, "grad_norm": 0.37487945047582993, "learning_rate": 7.641086702338766e-05, "loss": 1.0477, "step": 435 }, { "epoch": 1.1150895140664963, "grad_norm": 0.41781785412216255, "learning_rate": 7.638123181356688e-05, "loss": 1.046, "step": 436 }, { "epoch": 1.1176470588235294, "grad_norm": 0.5447678799969146, "learning_rate": 7.635148055369111e-05, "loss": 1.0268, "step": 437 }, { "epoch": 1.1202046035805626, "grad_norm": 0.7354119493315371, "learning_rate": 7.63216133386619e-05, "loss": 1.0407, "step": 438 }, { "epoch": 1.1227621483375958, "grad_norm": 0.8468997724198893, "learning_rate": 7.62916302637507e-05, "loss": 1.0613, "step": 439 }, { "epoch": 1.1253196930946292, "grad_norm": 0.90648052854961, "learning_rate": 7.62615314245985e-05, "loss": 1.0389, "step": 440 }, { "epoch": 1.1278772378516624, "grad_norm": 1.090719930551356, "learning_rate": 7.623131691721556e-05, "loss": 1.0604, "step": 441 }, { "epoch": 1.1304347826086956, "grad_norm": 0.8711758817781659, "learning_rate": 7.620098683798113e-05, "loss": 1.0396, "step": 442 }, { "epoch": 1.132992327365729, "grad_norm": 0.7941323029712526, "learning_rate": 7.617054128364308e-05, "loss": 1.064, "step": 443 }, { "epoch": 1.1355498721227621, "grad_norm": 0.7358238997014726, "learning_rate": 7.613998035131767e-05, "loss": 1.0302, "step": 444 }, { "epoch": 1.1381074168797953, "grad_norm": 0.8267368055242302, "learning_rate": 7.610930413848916e-05, "loss": 1.0499, "step": 445 }, { "epoch": 1.1406649616368287, "grad_norm": 0.8808186893039806, "learning_rate": 7.607851274300954e-05, "loss": 1.0241, "step": 446 }, { "epoch": 1.143222506393862, "grad_norm": 0.844367546052809, "learning_rate": 7.604760626309822e-05, "loss": 1.0136, "step": 447 }, { "epoch": 1.145780051150895, "grad_norm": 0.9008836613292239, "learning_rate": 7.601658479734172e-05, "loss": 1.0478, "step": 448 }, { "epoch": 1.1483375959079285, "grad_norm": 0.8890437058493534, "learning_rate": 7.598544844469335e-05, "loss": 1.0371, "step": 449 }, { "epoch": 1.1508951406649617, "grad_norm": 0.7574790514791548, "learning_rate": 7.595419730447283e-05, "loss": 1.0871, "step": 450 }, { "epoch": 1.1534526854219949, "grad_norm": 0.6622645508305952, "learning_rate": 7.592283147636613e-05, "loss": 1.0584, "step": 451 }, { "epoch": 1.156010230179028, "grad_norm": 0.5240209699484857, "learning_rate": 7.589135106042497e-05, "loss": 1.061, "step": 452 }, { "epoch": 1.1585677749360614, "grad_norm": 0.7258085512815178, "learning_rate": 7.585975615706664e-05, "loss": 1.0782, "step": 453 }, { "epoch": 1.1611253196930946, "grad_norm": 0.8692692426989976, "learning_rate": 7.582804686707361e-05, "loss": 1.0487, "step": 454 }, { "epoch": 1.1636828644501278, "grad_norm": 0.7256777326774219, "learning_rate": 7.57962232915932e-05, "loss": 1.0443, "step": 455 }, { "epoch": 1.1662404092071612, "grad_norm": 0.5531298720327349, "learning_rate": 7.576428553213731e-05, "loss": 1.0366, "step": 456 }, { "epoch": 1.1687979539641944, "grad_norm": 0.6217533009743852, "learning_rate": 7.573223369058206e-05, "loss": 1.0295, "step": 457 }, { "epoch": 1.1713554987212276, "grad_norm": 0.49232292611041006, "learning_rate": 7.570006786916749e-05, "loss": 1.0329, "step": 458 }, { "epoch": 1.1739130434782608, "grad_norm": 0.52993398976404, "learning_rate": 7.566778817049718e-05, "loss": 1.0402, "step": 459 }, { "epoch": 1.1764705882352942, "grad_norm": 0.5585620019915892, "learning_rate": 7.563539469753798e-05, "loss": 1.069, "step": 460 }, { "epoch": 1.1790281329923273, "grad_norm": 0.6428206102008164, "learning_rate": 7.560288755361967e-05, "loss": 1.0908, "step": 461 }, { "epoch": 1.1815856777493605, "grad_norm": 0.6280277113340915, "learning_rate": 7.557026684243459e-05, "loss": 1.0348, "step": 462 }, { "epoch": 1.184143222506394, "grad_norm": 0.43288917306833347, "learning_rate": 7.55375326680374e-05, "loss": 1.0241, "step": 463 }, { "epoch": 1.186700767263427, "grad_norm": 0.42841127990041067, "learning_rate": 7.550468513484461e-05, "loss": 1.0514, "step": 464 }, { "epoch": 1.1892583120204603, "grad_norm": 0.4254341549287906, "learning_rate": 7.54717243476344e-05, "loss": 1.0606, "step": 465 }, { "epoch": 1.1918158567774937, "grad_norm": 0.4171053731115308, "learning_rate": 7.543865041154616e-05, "loss": 1.0563, "step": 466 }, { "epoch": 1.1943734015345269, "grad_norm": 0.4236528757379926, "learning_rate": 7.540546343208021e-05, "loss": 1.0336, "step": 467 }, { "epoch": 1.19693094629156, "grad_norm": 0.4994661910864934, "learning_rate": 7.537216351509752e-05, "loss": 1.0844, "step": 468 }, { "epoch": 1.1994884910485935, "grad_norm": 0.6198060558124957, "learning_rate": 7.533875076681923e-05, "loss": 1.0492, "step": 469 }, { "epoch": 1.2020460358056266, "grad_norm": 0.583735235841795, "learning_rate": 7.530522529382641e-05, "loss": 1.071, "step": 470 }, { "epoch": 1.2046035805626598, "grad_norm": 0.5659582359875482, "learning_rate": 7.527158720305975e-05, "loss": 1.0429, "step": 471 }, { "epoch": 1.207161125319693, "grad_norm": 0.5451963873404054, "learning_rate": 7.523783660181915e-05, "loss": 1.0021, "step": 472 }, { "epoch": 1.2097186700767264, "grad_norm": 0.4423684081873438, "learning_rate": 7.520397359776338e-05, "loss": 1.0574, "step": 473 }, { "epoch": 1.2122762148337596, "grad_norm": 0.5198395882214115, "learning_rate": 7.516999829890976e-05, "loss": 1.0409, "step": 474 }, { "epoch": 1.2148337595907928, "grad_norm": 0.5360908221928974, "learning_rate": 7.513591081363384e-05, "loss": 1.0555, "step": 475 }, { "epoch": 1.2173913043478262, "grad_norm": 0.40726862330103586, "learning_rate": 7.5101711250669e-05, "loss": 1.0253, "step": 476 }, { "epoch": 1.2199488491048593, "grad_norm": 0.554526695474263, "learning_rate": 7.506739971910613e-05, "loss": 1.0792, "step": 477 }, { "epoch": 1.2225063938618925, "grad_norm": 0.6837744946054213, "learning_rate": 7.503297632839328e-05, "loss": 1.0618, "step": 478 }, { "epoch": 1.2250639386189257, "grad_norm": 0.787840350339849, "learning_rate": 7.499844118833533e-05, "loss": 1.0711, "step": 479 }, { "epoch": 1.227621483375959, "grad_norm": 0.9184717384704346, "learning_rate": 7.496379440909361e-05, "loss": 1.0225, "step": 480 }, { "epoch": 1.2301790281329923, "grad_norm": 0.98861238646143, "learning_rate": 7.492903610118556e-05, "loss": 1.0322, "step": 481 }, { "epoch": 1.2327365728900257, "grad_norm": 0.8602509324185822, "learning_rate": 7.489416637548436e-05, "loss": 1.0511, "step": 482 }, { "epoch": 1.2352941176470589, "grad_norm": 0.7497024524996155, "learning_rate": 7.485918534321866e-05, "loss": 1.0439, "step": 483 }, { "epoch": 1.237851662404092, "grad_norm": 0.6658522387073332, "learning_rate": 7.482409311597206e-05, "loss": 1.0597, "step": 484 }, { "epoch": 1.2404092071611252, "grad_norm": 0.6041271689399434, "learning_rate": 7.478888980568294e-05, "loss": 1.0622, "step": 485 }, { "epoch": 1.2429667519181586, "grad_norm": 0.5141728116418924, "learning_rate": 7.475357552464397e-05, "loss": 1.0559, "step": 486 }, { "epoch": 1.2455242966751918, "grad_norm": 0.3924962212572937, "learning_rate": 7.47181503855018e-05, "loss": 1.0283, "step": 487 }, { "epoch": 1.248081841432225, "grad_norm": 0.4006024520158971, "learning_rate": 7.468261450125674e-05, "loss": 1.0463, "step": 488 }, { "epoch": 1.2506393861892584, "grad_norm": 0.3732442212525654, "learning_rate": 7.464696798526229e-05, "loss": 1.0553, "step": 489 }, { "epoch": 1.2531969309462916, "grad_norm": 0.4036756007900454, "learning_rate": 7.461121095122491e-05, "loss": 1.0754, "step": 490 }, { "epoch": 1.2557544757033248, "grad_norm": 0.4521255607620733, "learning_rate": 7.457534351320356e-05, "loss": 1.0475, "step": 491 }, { "epoch": 1.258312020460358, "grad_norm": 0.4480587687694347, "learning_rate": 7.453936578560937e-05, "loss": 1.0724, "step": 492 }, { "epoch": 1.2608695652173914, "grad_norm": 0.4064497930563765, "learning_rate": 7.450327788320529e-05, "loss": 1.0411, "step": 493 }, { "epoch": 1.2634271099744245, "grad_norm": 0.32223473130563357, "learning_rate": 7.446707992110571e-05, "loss": 1.0391, "step": 494 }, { "epoch": 1.265984654731458, "grad_norm": 0.33489725476820903, "learning_rate": 7.443077201477609e-05, "loss": 1.0631, "step": 495 }, { "epoch": 1.2685421994884911, "grad_norm": 0.34146623369427676, "learning_rate": 7.439435428003256e-05, "loss": 1.0523, "step": 496 }, { "epoch": 1.2710997442455243, "grad_norm": 0.29613834050231674, "learning_rate": 7.435782683304164e-05, "loss": 1.0446, "step": 497 }, { "epoch": 1.2736572890025575, "grad_norm": 0.3841724281833824, "learning_rate": 7.43211897903198e-05, "loss": 1.0306, "step": 498 }, { "epoch": 1.2762148337595907, "grad_norm": 0.4433757479687236, "learning_rate": 7.428444326873303e-05, "loss": 1.0448, "step": 499 }, { "epoch": 1.278772378516624, "grad_norm": 0.36337315524962616, "learning_rate": 7.424758738549667e-05, "loss": 1.0201, "step": 500 }, { "epoch": 1.2813299232736572, "grad_norm": 0.29686335739489117, "learning_rate": 7.421062225817478e-05, "loss": 1.0397, "step": 501 }, { "epoch": 1.2838874680306906, "grad_norm": 0.41385410051793525, "learning_rate": 7.417354800467997e-05, "loss": 1.0449, "step": 502 }, { "epoch": 1.2864450127877238, "grad_norm": 0.41885148772105985, "learning_rate": 7.413636474327292e-05, "loss": 1.0642, "step": 503 }, { "epoch": 1.289002557544757, "grad_norm": 0.43853476074138853, "learning_rate": 7.409907259256202e-05, "loss": 1.0397, "step": 504 }, { "epoch": 1.2915601023017902, "grad_norm": 0.6686928411130255, "learning_rate": 7.4061671671503e-05, "loss": 1.0542, "step": 505 }, { "epoch": 1.2941176470588236, "grad_norm": 0.607563092071724, "learning_rate": 7.402416209939856e-05, "loss": 1.0304, "step": 506 }, { "epoch": 1.2966751918158568, "grad_norm": 0.580772361466558, "learning_rate": 7.398654399589795e-05, "loss": 1.0632, "step": 507 }, { "epoch": 1.29923273657289, "grad_norm": 0.7281319746142754, "learning_rate": 7.394881748099668e-05, "loss": 1.026, "step": 508 }, { "epoch": 1.3017902813299234, "grad_norm": 0.782767959332906, "learning_rate": 7.391098267503599e-05, "loss": 1.0584, "step": 509 }, { "epoch": 1.3043478260869565, "grad_norm": 0.7334069729083356, "learning_rate": 7.387303969870262e-05, "loss": 1.0265, "step": 510 }, { "epoch": 1.3069053708439897, "grad_norm": 0.8443472328520687, "learning_rate": 7.383498867302835e-05, "loss": 1.0362, "step": 511 }, { "epoch": 1.309462915601023, "grad_norm": 0.9845489824701952, "learning_rate": 7.379682971938953e-05, "loss": 1.0658, "step": 512 }, { "epoch": 1.3120204603580563, "grad_norm": 1.0456842320775128, "learning_rate": 7.375856295950693e-05, "loss": 1.0222, "step": 513 }, { "epoch": 1.3145780051150895, "grad_norm": 0.872370978809925, "learning_rate": 7.372018851544507e-05, "loss": 1.0431, "step": 514 }, { "epoch": 1.317135549872123, "grad_norm": 0.5883115072059536, "learning_rate": 7.368170650961203e-05, "loss": 1.0349, "step": 515 }, { "epoch": 1.319693094629156, "grad_norm": 0.4266657814616918, "learning_rate": 7.364311706475901e-05, "loss": 1.0304, "step": 516 }, { "epoch": 1.3222506393861893, "grad_norm": 0.5955652733904395, "learning_rate": 7.360442030397987e-05, "loss": 1.0267, "step": 517 }, { "epoch": 1.3248081841432224, "grad_norm": 0.6659451706135621, "learning_rate": 7.356561635071082e-05, "loss": 1.0437, "step": 518 }, { "epoch": 1.3273657289002558, "grad_norm": 0.6313803070585998, "learning_rate": 7.352670532872999e-05, "loss": 1.0342, "step": 519 }, { "epoch": 1.329923273657289, "grad_norm": 0.6906340675596294, "learning_rate": 7.348768736215706e-05, "loss": 1.0784, "step": 520 }, { "epoch": 1.3324808184143222, "grad_norm": 0.6387943436906747, "learning_rate": 7.34485625754528e-05, "loss": 1.0785, "step": 521 }, { "epoch": 1.3350383631713556, "grad_norm": 0.6214194954347501, "learning_rate": 7.340933109341878e-05, "loss": 1.0554, "step": 522 }, { "epoch": 1.3375959079283888, "grad_norm": 0.6783392090338549, "learning_rate": 7.336999304119685e-05, "loss": 1.0461, "step": 523 }, { "epoch": 1.340153452685422, "grad_norm": 0.7773583234292399, "learning_rate": 7.333054854426884e-05, "loss": 1.0714, "step": 524 }, { "epoch": 1.3427109974424551, "grad_norm": 0.7619691171883715, "learning_rate": 7.329099772845611e-05, "loss": 1.0427, "step": 525 }, { "epoch": 1.3452685421994885, "grad_norm": 0.6796465920444608, "learning_rate": 7.325134071991915e-05, "loss": 1.0673, "step": 526 }, { "epoch": 1.3478260869565217, "grad_norm": 0.5764938509906341, "learning_rate": 7.321157764515721e-05, "loss": 1.0172, "step": 527 }, { "epoch": 1.350383631713555, "grad_norm": 0.5703898248580829, "learning_rate": 7.317170863100786e-05, "loss": 1.0814, "step": 528 }, { "epoch": 1.3529411764705883, "grad_norm": 0.543928921118232, "learning_rate": 7.313173380464658e-05, "loss": 1.0427, "step": 529 }, { "epoch": 1.3554987212276215, "grad_norm": 0.403264231625384, "learning_rate": 7.309165329358639e-05, "loss": 1.0141, "step": 530 }, { "epoch": 1.3580562659846547, "grad_norm": 0.37834189295305315, "learning_rate": 7.305146722567746e-05, "loss": 1.0518, "step": 531 }, { "epoch": 1.3606138107416879, "grad_norm": 0.4423087622812266, "learning_rate": 7.301117572910661e-05, "loss": 1.0374, "step": 532 }, { "epoch": 1.3631713554987213, "grad_norm": 0.6513039474923614, "learning_rate": 7.297077893239699e-05, "loss": 1.0751, "step": 533 }, { "epoch": 1.3657289002557544, "grad_norm": 0.7901236558090899, "learning_rate": 7.293027696440763e-05, "loss": 1.0429, "step": 534 }, { "epoch": 1.3682864450127878, "grad_norm": 0.8028952131258908, "learning_rate": 7.288966995433309e-05, "loss": 1.0531, "step": 535 }, { "epoch": 1.370843989769821, "grad_norm": 0.7531252236472497, "learning_rate": 7.284895803170288e-05, "loss": 1.0875, "step": 536 }, { "epoch": 1.3734015345268542, "grad_norm": 0.7178997040327029, "learning_rate": 7.280814132638129e-05, "loss": 1.0439, "step": 537 }, { "epoch": 1.3759590792838874, "grad_norm": 0.5847814083136553, "learning_rate": 7.276721996856677e-05, "loss": 1.0554, "step": 538 }, { "epoch": 1.3785166240409208, "grad_norm": 0.49255490697808374, "learning_rate": 7.272619408879164e-05, "loss": 1.0404, "step": 539 }, { "epoch": 1.381074168797954, "grad_norm": 0.45239860846158764, "learning_rate": 7.268506381792158e-05, "loss": 1.049, "step": 540 }, { "epoch": 1.3836317135549872, "grad_norm": 0.68450243672809, "learning_rate": 7.26438292871553e-05, "loss": 1.076, "step": 541 }, { "epoch": 1.3861892583120206, "grad_norm": 0.6927411824014928, "learning_rate": 7.260249062802407e-05, "loss": 1.0677, "step": 542 }, { "epoch": 1.3887468030690537, "grad_norm": 0.6632064448071592, "learning_rate": 7.256104797239132e-05, "loss": 1.0227, "step": 543 }, { "epoch": 1.391304347826087, "grad_norm": 0.6518144933302445, "learning_rate": 7.251950145245218e-05, "loss": 1.044, "step": 544 }, { "epoch": 1.39386189258312, "grad_norm": 0.5815850660081673, "learning_rate": 7.247785120073313e-05, "loss": 1.0333, "step": 545 }, { "epoch": 1.3964194373401535, "grad_norm": 0.47544855954948456, "learning_rate": 7.24360973500915e-05, "loss": 1.0128, "step": 546 }, { "epoch": 1.3989769820971867, "grad_norm": 0.5253477384563784, "learning_rate": 7.239424003371511e-05, "loss": 1.0519, "step": 547 }, { "epoch": 1.40153452685422, "grad_norm": 0.54424332711931, "learning_rate": 7.235227938512182e-05, "loss": 1.0421, "step": 548 }, { "epoch": 1.4040920716112533, "grad_norm": 0.4163026169587667, "learning_rate": 7.231021553815909e-05, "loss": 1.0647, "step": 549 }, { "epoch": 1.4066496163682864, "grad_norm": 0.43975874335531806, "learning_rate": 7.226804862700357e-05, "loss": 1.006, "step": 550 }, { "epoch": 1.4092071611253196, "grad_norm": 0.5676277493519121, "learning_rate": 7.222577878616065e-05, "loss": 1.018, "step": 551 }, { "epoch": 1.4117647058823528, "grad_norm": 0.4945104420529325, "learning_rate": 7.218340615046407e-05, "loss": 1.0609, "step": 552 }, { "epoch": 1.4143222506393862, "grad_norm": 0.42595338834103746, "learning_rate": 7.214093085507549e-05, "loss": 1.0455, "step": 553 }, { "epoch": 1.4168797953964194, "grad_norm": 0.3651457329411855, "learning_rate": 7.209835303548396e-05, "loss": 1.0558, "step": 554 }, { "epoch": 1.4194373401534528, "grad_norm": 0.48372716214113715, "learning_rate": 7.205567282750567e-05, "loss": 1.069, "step": 555 }, { "epoch": 1.421994884910486, "grad_norm": 0.483272224628568, "learning_rate": 7.201289036728331e-05, "loss": 1.0536, "step": 556 }, { "epoch": 1.4245524296675192, "grad_norm": 0.4770924027986762, "learning_rate": 7.197000579128582e-05, "loss": 1.0403, "step": 557 }, { "epoch": 1.4271099744245523, "grad_norm": 0.448305597034261, "learning_rate": 7.192701923630781e-05, "loss": 1.0328, "step": 558 }, { "epoch": 1.4296675191815857, "grad_norm": 0.495021183649144, "learning_rate": 7.188393083946923e-05, "loss": 1.0258, "step": 559 }, { "epoch": 1.432225063938619, "grad_norm": 0.4705177505313109, "learning_rate": 7.184074073821487e-05, "loss": 1.0174, "step": 560 }, { "epoch": 1.434782608695652, "grad_norm": 0.43574757389209073, "learning_rate": 7.179744907031394e-05, "loss": 1.0287, "step": 561 }, { "epoch": 1.4373401534526855, "grad_norm": 0.3635773785423228, "learning_rate": 7.175405597385964e-05, "loss": 1.0412, "step": 562 }, { "epoch": 1.4398976982097187, "grad_norm": 0.553156783226794, "learning_rate": 7.171056158726869e-05, "loss": 1.0688, "step": 563 }, { "epoch": 1.4424552429667519, "grad_norm": 0.6900930611820969, "learning_rate": 7.166696604928091e-05, "loss": 1.0668, "step": 564 }, { "epoch": 1.445012787723785, "grad_norm": 0.85914870330665, "learning_rate": 7.162326949895882e-05, "loss": 1.0752, "step": 565 }, { "epoch": 1.4475703324808185, "grad_norm": 1.0719699916305456, "learning_rate": 7.157947207568708e-05, "loss": 1.038, "step": 566 }, { "epoch": 1.4501278772378516, "grad_norm": 0.8876743742200847, "learning_rate": 7.153557391917221e-05, "loss": 1.0186, "step": 567 }, { "epoch": 1.452685421994885, "grad_norm": 0.6329728177218358, "learning_rate": 7.149157516944195e-05, "loss": 1.0496, "step": 568 }, { "epoch": 1.4552429667519182, "grad_norm": 0.4320509800831513, "learning_rate": 7.144747596684498e-05, "loss": 1.0401, "step": 569 }, { "epoch": 1.4578005115089514, "grad_norm": 0.3695887057019756, "learning_rate": 7.14032764520504e-05, "loss": 1.0298, "step": 570 }, { "epoch": 1.4603580562659846, "grad_norm": 0.4557157139603631, "learning_rate": 7.135897676604728e-05, "loss": 1.0504, "step": 571 }, { "epoch": 1.4629156010230178, "grad_norm": 0.5697580796103671, "learning_rate": 7.131457705014424e-05, "loss": 1.0553, "step": 572 }, { "epoch": 1.4654731457800512, "grad_norm": 0.562535535478939, "learning_rate": 7.127007744596891e-05, "loss": 1.0402, "step": 573 }, { "epoch": 1.4680306905370843, "grad_norm": 0.5386864181747942, "learning_rate": 7.122547809546766e-05, "loss": 1.0266, "step": 574 }, { "epoch": 1.4705882352941178, "grad_norm": 0.5604159637885537, "learning_rate": 7.118077914090492e-05, "loss": 1.0593, "step": 575 }, { "epoch": 1.473145780051151, "grad_norm": 0.5345575941108697, "learning_rate": 7.113598072486291e-05, "loss": 1.0628, "step": 576 }, { "epoch": 1.4757033248081841, "grad_norm": 0.42108117394775474, "learning_rate": 7.10910829902411e-05, "loss": 1.0545, "step": 577 }, { "epoch": 1.4782608695652173, "grad_norm": 0.42679316673017587, "learning_rate": 7.104608608025578e-05, "loss": 1.0441, "step": 578 }, { "epoch": 1.4808184143222507, "grad_norm": 0.4632742996319085, "learning_rate": 7.100099013843956e-05, "loss": 1.0233, "step": 579 }, { "epoch": 1.4833759590792839, "grad_norm": 0.41059416177562724, "learning_rate": 7.095579530864097e-05, "loss": 1.0393, "step": 580 }, { "epoch": 1.485933503836317, "grad_norm": 0.49854832145650435, "learning_rate": 7.091050173502397e-05, "loss": 1.0578, "step": 581 }, { "epoch": 1.4884910485933505, "grad_norm": 0.6121395366510051, "learning_rate": 7.08651095620675e-05, "loss": 1.0463, "step": 582 }, { "epoch": 1.4910485933503836, "grad_norm": 0.6431404220907431, "learning_rate": 7.081961893456502e-05, "loss": 1.0715, "step": 583 }, { "epoch": 1.4936061381074168, "grad_norm": 0.8536051526625087, "learning_rate": 7.077402999762401e-05, "loss": 1.0774, "step": 584 }, { "epoch": 1.49616368286445, "grad_norm": 0.9386047615779075, "learning_rate": 7.07283428966656e-05, "loss": 1.0167, "step": 585 }, { "epoch": 1.4987212276214834, "grad_norm": 0.8795122294030159, "learning_rate": 7.068255777742399e-05, "loss": 1.0325, "step": 586 }, { "epoch": 1.5012787723785166, "grad_norm": 0.6822762592618459, "learning_rate": 7.063667478594609e-05, "loss": 1.0408, "step": 587 }, { "epoch": 1.50383631713555, "grad_norm": 0.5845658862279202, "learning_rate": 7.059069406859096e-05, "loss": 1.0671, "step": 588 }, { "epoch": 1.5063938618925832, "grad_norm": 0.5244238010393297, "learning_rate": 7.05446157720294e-05, "loss": 1.0071, "step": 589 }, { "epoch": 1.5089514066496164, "grad_norm": 0.5405468625462244, "learning_rate": 7.049844004324352e-05, "loss": 1.0364, "step": 590 }, { "epoch": 1.5115089514066495, "grad_norm": 0.5321510176297898, "learning_rate": 7.045216702952615e-05, "loss": 1.0099, "step": 591 }, { "epoch": 1.5140664961636827, "grad_norm": 0.6304136952687458, "learning_rate": 7.040579687848051e-05, "loss": 1.0629, "step": 592 }, { "epoch": 1.5166240409207161, "grad_norm": 0.6788574527638351, "learning_rate": 7.035932973801961e-05, "loss": 0.9933, "step": 593 }, { "epoch": 1.5191815856777495, "grad_norm": 0.5664383321834381, "learning_rate": 7.031276575636588e-05, "loss": 1.0581, "step": 594 }, { "epoch": 1.5217391304347827, "grad_norm": 0.4764617199761682, "learning_rate": 7.026610508205065e-05, "loss": 1.03, "step": 595 }, { "epoch": 1.5242966751918159, "grad_norm": 0.4115926551003442, "learning_rate": 7.021934786391369e-05, "loss": 1.0404, "step": 596 }, { "epoch": 1.526854219948849, "grad_norm": 0.5008700498030078, "learning_rate": 7.017249425110268e-05, "loss": 1.0477, "step": 597 }, { "epoch": 1.5294117647058822, "grad_norm": 0.5614325700629363, "learning_rate": 7.012554439307288e-05, "loss": 1.0496, "step": 598 }, { "epoch": 1.5319693094629157, "grad_norm": 0.6424393529952258, "learning_rate": 7.007849843958646e-05, "loss": 1.027, "step": 599 }, { "epoch": 1.5345268542199488, "grad_norm": 0.7601254190908516, "learning_rate": 7.003135654071216e-05, "loss": 1.0771, "step": 600 }, { "epoch": 1.5370843989769822, "grad_norm": 0.6683027372767937, "learning_rate": 6.998411884682476e-05, "loss": 1.0245, "step": 601 }, { "epoch": 1.5396419437340154, "grad_norm": 0.4292563464334402, "learning_rate": 6.993678550860464e-05, "loss": 1.0208, "step": 602 }, { "epoch": 1.5421994884910486, "grad_norm": 0.37531967359267043, "learning_rate": 6.988935667703722e-05, "loss": 1.0341, "step": 603 }, { "epoch": 1.5447570332480818, "grad_norm": 0.5020503230133389, "learning_rate": 6.984183250341257e-05, "loss": 1.0454, "step": 604 }, { "epoch": 1.547314578005115, "grad_norm": 0.46091810824436635, "learning_rate": 6.979421313932488e-05, "loss": 1.0315, "step": 605 }, { "epoch": 1.5498721227621484, "grad_norm": 0.4736527457196331, "learning_rate": 6.974649873667194e-05, "loss": 1.0086, "step": 606 }, { "epoch": 1.5524296675191815, "grad_norm": 0.4999676394044973, "learning_rate": 6.969868944765474e-05, "loss": 1.0323, "step": 607 }, { "epoch": 1.554987212276215, "grad_norm": 0.5697900185173456, "learning_rate": 6.965078542477693e-05, "loss": 1.0575, "step": 608 }, { "epoch": 1.5575447570332481, "grad_norm": 0.6296338412635226, "learning_rate": 6.960278682084436e-05, "loss": 1.0537, "step": 609 }, { "epoch": 1.5601023017902813, "grad_norm": 0.564117474441083, "learning_rate": 6.955469378896456e-05, "loss": 1.0231, "step": 610 }, { "epoch": 1.5626598465473145, "grad_norm": 0.5743418608288569, "learning_rate": 6.950650648254623e-05, "loss": 1.0572, "step": 611 }, { "epoch": 1.5652173913043477, "grad_norm": 0.48086239097871125, "learning_rate": 6.945822505529887e-05, "loss": 1.0466, "step": 612 }, { "epoch": 1.567774936061381, "grad_norm": 0.3756345397327383, "learning_rate": 6.940984966123217e-05, "loss": 1.0366, "step": 613 }, { "epoch": 1.5703324808184145, "grad_norm": 0.4472187990537581, "learning_rate": 6.936138045465556e-05, "loss": 1.0415, "step": 614 }, { "epoch": 1.5728900255754477, "grad_norm": 0.4852565244536602, "learning_rate": 6.93128175901777e-05, "loss": 1.0591, "step": 615 }, { "epoch": 1.5754475703324808, "grad_norm": 0.4059570141509962, "learning_rate": 6.926416122270604e-05, "loss": 1.0144, "step": 616 }, { "epoch": 1.578005115089514, "grad_norm": 0.3681367090287205, "learning_rate": 6.921541150744624e-05, "loss": 1.0473, "step": 617 }, { "epoch": 1.5805626598465472, "grad_norm": 0.3430601437581608, "learning_rate": 6.916656859990179e-05, "loss": 1.0409, "step": 618 }, { "epoch": 1.5831202046035806, "grad_norm": 0.3419778416901883, "learning_rate": 6.911763265587338e-05, "loss": 1.0274, "step": 619 }, { "epoch": 1.5856777493606138, "grad_norm": 0.3796918250923331, "learning_rate": 6.90686038314585e-05, "loss": 1.015, "step": 620 }, { "epoch": 1.5882352941176472, "grad_norm": 0.49070616775356396, "learning_rate": 6.901948228305094e-05, "loss": 1.0446, "step": 621 }, { "epoch": 1.5907928388746804, "grad_norm": 0.448562317355006, "learning_rate": 6.89702681673402e-05, "loss": 1.0583, "step": 622 }, { "epoch": 1.5933503836317136, "grad_norm": 0.41319179373282156, "learning_rate": 6.892096164131113e-05, "loss": 1.031, "step": 623 }, { "epoch": 1.5959079283887467, "grad_norm": 0.4625806799698519, "learning_rate": 6.887156286224329e-05, "loss": 1.0698, "step": 624 }, { "epoch": 1.59846547314578, "grad_norm": 0.5803778153984633, "learning_rate": 6.882207198771056e-05, "loss": 1.0904, "step": 625 }, { "epoch": 1.6010230179028133, "grad_norm": 0.6358263157928301, "learning_rate": 6.877248917558055e-05, "loss": 1.0178, "step": 626 }, { "epoch": 1.6035805626598465, "grad_norm": 0.6408849160156058, "learning_rate": 6.872281458401414e-05, "loss": 1.028, "step": 627 }, { "epoch": 1.60613810741688, "grad_norm": 0.6318504394160956, "learning_rate": 6.867304837146501e-05, "loss": 1.0669, "step": 628 }, { "epoch": 1.608695652173913, "grad_norm": 0.6132529357479547, "learning_rate": 6.862319069667904e-05, "loss": 1.0215, "step": 629 }, { "epoch": 1.6112531969309463, "grad_norm": 0.5149229199094381, "learning_rate": 6.857324171869393e-05, "loss": 1.0268, "step": 630 }, { "epoch": 1.6138107416879794, "grad_norm": 0.3860072737272255, "learning_rate": 6.852320159683853e-05, "loss": 1.0342, "step": 631 }, { "epoch": 1.6163682864450126, "grad_norm": 0.33909485027918096, "learning_rate": 6.84730704907325e-05, "loss": 1.0283, "step": 632 }, { "epoch": 1.618925831202046, "grad_norm": 0.4191834960800194, "learning_rate": 6.842284856028571e-05, "loss": 1.0114, "step": 633 }, { "epoch": 1.6214833759590794, "grad_norm": 0.5347405775672891, "learning_rate": 6.83725359656977e-05, "loss": 1.0589, "step": 634 }, { "epoch": 1.6240409207161126, "grad_norm": 0.6355103271925497, "learning_rate": 6.832213286745727e-05, "loss": 1.0119, "step": 635 }, { "epoch": 1.6265984654731458, "grad_norm": 0.7345332659340343, "learning_rate": 6.82716394263419e-05, "loss": 1.058, "step": 636 }, { "epoch": 1.629156010230179, "grad_norm": 0.6965270461851645, "learning_rate": 6.82210558034172e-05, "loss": 1.0517, "step": 637 }, { "epoch": 1.6317135549872122, "grad_norm": 0.582639080193431, "learning_rate": 6.817038216003649e-05, "loss": 1.0361, "step": 638 }, { "epoch": 1.6342710997442456, "grad_norm": 0.49231339614396796, "learning_rate": 6.811961865784026e-05, "loss": 1.0691, "step": 639 }, { "epoch": 1.6368286445012787, "grad_norm": 0.4590551715359045, "learning_rate": 6.806876545875558e-05, "loss": 1.004, "step": 640 }, { "epoch": 1.6393861892583121, "grad_norm": 0.4934679803392225, "learning_rate": 6.801782272499567e-05, "loss": 1.0249, "step": 641 }, { "epoch": 1.6419437340153453, "grad_norm": 0.4855827703329849, "learning_rate": 6.796679061905936e-05, "loss": 1.0275, "step": 642 }, { "epoch": 1.6445012787723785, "grad_norm": 0.5663891915421606, "learning_rate": 6.791566930373051e-05, "loss": 1.0431, "step": 643 }, { "epoch": 1.6470588235294117, "grad_norm": 0.5449325271192005, "learning_rate": 6.786445894207763e-05, "loss": 1.088, "step": 644 }, { "epoch": 1.6496163682864449, "grad_norm": 0.4404635331433356, "learning_rate": 6.781315969745319e-05, "loss": 0.9978, "step": 645 }, { "epoch": 1.6521739130434783, "grad_norm": 0.44006079141319454, "learning_rate": 6.776177173349322e-05, "loss": 1.0542, "step": 646 }, { "epoch": 1.6547314578005117, "grad_norm": 0.43703882239072955, "learning_rate": 6.771029521411674e-05, "loss": 1.0127, "step": 647 }, { "epoch": 1.6572890025575449, "grad_norm": 0.5232507591272976, "learning_rate": 6.765873030352525e-05, "loss": 1.0307, "step": 648 }, { "epoch": 1.659846547314578, "grad_norm": 0.5907572767491934, "learning_rate": 6.760707716620224e-05, "loss": 1.0364, "step": 649 }, { "epoch": 1.6624040920716112, "grad_norm": 0.5978645336642385, "learning_rate": 6.755533596691255e-05, "loss": 1.0197, "step": 650 }, { "epoch": 1.6649616368286444, "grad_norm": 0.5721851679271525, "learning_rate": 6.7503506870702e-05, "loss": 1.0456, "step": 651 }, { "epoch": 1.6675191815856778, "grad_norm": 0.526587743406477, "learning_rate": 6.74515900428967e-05, "loss": 1.0698, "step": 652 }, { "epoch": 1.670076726342711, "grad_norm": 0.45666448861954223, "learning_rate": 6.739958564910274e-05, "loss": 1.0444, "step": 653 }, { "epoch": 1.6726342710997444, "grad_norm": 0.33495402182030853, "learning_rate": 6.734749385520539e-05, "loss": 1.0494, "step": 654 }, { "epoch": 1.6751918158567776, "grad_norm": 0.48053582090285774, "learning_rate": 6.72953148273688e-05, "loss": 1.0384, "step": 655 }, { "epoch": 1.6777493606138107, "grad_norm": 0.6618699544458244, "learning_rate": 6.724304873203538e-05, "loss": 1.0252, "step": 656 }, { "epoch": 1.680306905370844, "grad_norm": 0.7825491070996107, "learning_rate": 6.71906957359252e-05, "loss": 1.0279, "step": 657 }, { "epoch": 1.682864450127877, "grad_norm": 0.7241348855375171, "learning_rate": 6.71382560060356e-05, "loss": 1.0492, "step": 658 }, { "epoch": 1.6854219948849105, "grad_norm": 0.49266363223214565, "learning_rate": 6.708572970964054e-05, "loss": 1.0496, "step": 659 }, { "epoch": 1.6879795396419437, "grad_norm": 0.3589459784191605, "learning_rate": 6.703311701429019e-05, "loss": 1.0657, "step": 660 }, { "epoch": 1.690537084398977, "grad_norm": 0.5098638866479447, "learning_rate": 6.698041808781021e-05, "loss": 1.045, "step": 661 }, { "epoch": 1.6930946291560103, "grad_norm": 0.5960811936716889, "learning_rate": 6.69276330983014e-05, "loss": 1.0372, "step": 662 }, { "epoch": 1.6956521739130435, "grad_norm": 0.6441771191728006, "learning_rate": 6.687476221413907e-05, "loss": 1.0591, "step": 663 }, { "epoch": 1.6982097186700766, "grad_norm": 0.7198375562017726, "learning_rate": 6.682180560397248e-05, "loss": 1.045, "step": 664 }, { "epoch": 1.7007672634271098, "grad_norm": 0.7219444637646196, "learning_rate": 6.676876343672441e-05, "loss": 1.0487, "step": 665 }, { "epoch": 1.7033248081841432, "grad_norm": 0.5227925388389713, "learning_rate": 6.67156358815905e-05, "loss": 1.0198, "step": 666 }, { "epoch": 1.7058823529411766, "grad_norm": 0.465506768448557, "learning_rate": 6.666242310803879e-05, "loss": 1.0403, "step": 667 }, { "epoch": 1.7084398976982098, "grad_norm": 0.4532826345672689, "learning_rate": 6.660912528580912e-05, "loss": 1.0297, "step": 668 }, { "epoch": 1.710997442455243, "grad_norm": 0.45327509709330577, "learning_rate": 6.655574258491267e-05, "loss": 1.0554, "step": 669 }, { "epoch": 1.7135549872122762, "grad_norm": 0.5641203058724333, "learning_rate": 6.650227517563134e-05, "loss": 1.0345, "step": 670 }, { "epoch": 1.7161125319693094, "grad_norm": 0.5934846994434492, "learning_rate": 6.644872322851722e-05, "loss": 1.0512, "step": 671 }, { "epoch": 1.7186700767263428, "grad_norm": 0.5243971153118981, "learning_rate": 6.63950869143921e-05, "loss": 1.0531, "step": 672 }, { "epoch": 1.721227621483376, "grad_norm": 0.47550551440418576, "learning_rate": 6.634136640434685e-05, "loss": 1.0305, "step": 673 }, { "epoch": 1.7237851662404093, "grad_norm": 0.4645714276383756, "learning_rate": 6.628756186974093e-05, "loss": 1.0734, "step": 674 }, { "epoch": 1.7263427109974425, "grad_norm": 0.5045165698461913, "learning_rate": 6.623367348220184e-05, "loss": 1.044, "step": 675 }, { "epoch": 1.7289002557544757, "grad_norm": 0.55969672109987, "learning_rate": 6.61797014136245e-05, "loss": 1.01, "step": 676 }, { "epoch": 1.7314578005115089, "grad_norm": 0.5692361967437805, "learning_rate": 6.612564583617085e-05, "loss": 1.0271, "step": 677 }, { "epoch": 1.734015345268542, "grad_norm": 0.4887973521051894, "learning_rate": 6.607150692226912e-05, "loss": 1.0289, "step": 678 }, { "epoch": 1.7365728900255755, "grad_norm": 0.4760009295478612, "learning_rate": 6.601728484461342e-05, "loss": 1.0301, "step": 679 }, { "epoch": 1.7391304347826086, "grad_norm": 0.48489041804369454, "learning_rate": 6.59629797761631e-05, "loss": 1.0309, "step": 680 }, { "epoch": 1.741687979539642, "grad_norm": 0.38491676259975743, "learning_rate": 6.59085918901423e-05, "loss": 1.0192, "step": 681 }, { "epoch": 1.7442455242966752, "grad_norm": 0.33184552078799817, "learning_rate": 6.585412136003927e-05, "loss": 1.0649, "step": 682 }, { "epoch": 1.7468030690537084, "grad_norm": 0.29047480084769817, "learning_rate": 6.579956835960592e-05, "loss": 1.0573, "step": 683 }, { "epoch": 1.7493606138107416, "grad_norm": 0.43218015798550996, "learning_rate": 6.574493306285719e-05, "loss": 1.0425, "step": 684 }, { "epoch": 1.7519181585677748, "grad_norm": 0.39350320437790215, "learning_rate": 6.569021564407057e-05, "loss": 1.0349, "step": 685 }, { "epoch": 1.7544757033248082, "grad_norm": 0.28656671953450064, "learning_rate": 6.56354162777855e-05, "loss": 1.0648, "step": 686 }, { "epoch": 1.7570332480818416, "grad_norm": 0.44037380340676446, "learning_rate": 6.558053513880278e-05, "loss": 1.0399, "step": 687 }, { "epoch": 1.7595907928388748, "grad_norm": 0.48759264629519933, "learning_rate": 6.552557240218412e-05, "loss": 1.0274, "step": 688 }, { "epoch": 1.762148337595908, "grad_norm": 0.47090986696853737, "learning_rate": 6.547052824325142e-05, "loss": 1.0105, "step": 689 }, { "epoch": 1.7647058823529411, "grad_norm": 0.47925643310416566, "learning_rate": 6.541540283758642e-05, "loss": 1.0363, "step": 690 }, { "epoch": 1.7672634271099743, "grad_norm": 0.3174804641375395, "learning_rate": 6.53601963610299e-05, "loss": 1.0469, "step": 691 }, { "epoch": 1.7698209718670077, "grad_norm": 0.34413914838661525, "learning_rate": 6.530490898968135e-05, "loss": 1.0405, "step": 692 }, { "epoch": 1.772378516624041, "grad_norm": 0.44861200909145516, "learning_rate": 6.524954089989822e-05, "loss": 1.0554, "step": 693 }, { "epoch": 1.7749360613810743, "grad_norm": 0.4663028003033403, "learning_rate": 6.519409226829549e-05, "loss": 1.0071, "step": 694 }, { "epoch": 1.7774936061381075, "grad_norm": 0.4344266413537487, "learning_rate": 6.513856327174501e-05, "loss": 1.0219, "step": 695 }, { "epoch": 1.7800511508951407, "grad_norm": 0.44323943000930544, "learning_rate": 6.508295408737505e-05, "loss": 1.0553, "step": 696 }, { "epoch": 1.7826086956521738, "grad_norm": 0.4656359741585452, "learning_rate": 6.50272648925696e-05, "loss": 1.0331, "step": 697 }, { "epoch": 1.785166240409207, "grad_norm": 0.4298294147098899, "learning_rate": 6.49714958649679e-05, "loss": 1.0474, "step": 698 }, { "epoch": 1.7877237851662404, "grad_norm": 0.373839672554522, "learning_rate": 6.491564718246381e-05, "loss": 1.0372, "step": 699 }, { "epoch": 1.7902813299232738, "grad_norm": 0.3633684264946787, "learning_rate": 6.485971902320534e-05, "loss": 1.0545, "step": 700 }, { "epoch": 1.792838874680307, "grad_norm": 0.374369128686114, "learning_rate": 6.480371156559395e-05, "loss": 1.0207, "step": 701 }, { "epoch": 1.7953964194373402, "grad_norm": 0.38960484038509224, "learning_rate": 6.47476249882841e-05, "loss": 1.0395, "step": 702 }, { "epoch": 1.7979539641943734, "grad_norm": 0.41076940243434146, "learning_rate": 6.469145947018261e-05, "loss": 1.0423, "step": 703 }, { "epoch": 1.8005115089514065, "grad_norm": 0.4037378076235256, "learning_rate": 6.463521519044808e-05, "loss": 1.0228, "step": 704 }, { "epoch": 1.80306905370844, "grad_norm": 0.3340754340147289, "learning_rate": 6.457889232849039e-05, "loss": 1.0534, "step": 705 }, { "epoch": 1.8056265984654731, "grad_norm": 0.2796291247118141, "learning_rate": 6.452249106397004e-05, "loss": 1.0043, "step": 706 }, { "epoch": 1.8081841432225065, "grad_norm": 0.29928606836625965, "learning_rate": 6.446601157679768e-05, "loss": 1.0086, "step": 707 }, { "epoch": 1.8107416879795397, "grad_norm": 0.32398878662153513, "learning_rate": 6.440945404713344e-05, "loss": 1.0346, "step": 708 }, { "epoch": 1.813299232736573, "grad_norm": 0.3448476257962995, "learning_rate": 6.435281865538638e-05, "loss": 0.9851, "step": 709 }, { "epoch": 1.815856777493606, "grad_norm": 0.3437352152692172, "learning_rate": 6.429610558221395e-05, "loss": 1.0297, "step": 710 }, { "epoch": 1.8184143222506393, "grad_norm": 0.3285991493351655, "learning_rate": 6.423931500852139e-05, "loss": 1.0827, "step": 711 }, { "epoch": 1.8209718670076727, "grad_norm": 0.2915476971624437, "learning_rate": 6.418244711546113e-05, "loss": 1.0311, "step": 712 }, { "epoch": 1.8235294117647058, "grad_norm": 0.3038318588786024, "learning_rate": 6.412550208443227e-05, "loss": 1.0402, "step": 713 }, { "epoch": 1.8260869565217392, "grad_norm": 0.312782111489241, "learning_rate": 6.406848009707995e-05, "loss": 1.0283, "step": 714 }, { "epoch": 1.8286445012787724, "grad_norm": 0.2822241296244193, "learning_rate": 6.401138133529478e-05, "loss": 1.0125, "step": 715 }, { "epoch": 1.8312020460358056, "grad_norm": 0.407257179302736, "learning_rate": 6.395420598121229e-05, "loss": 1.0293, "step": 716 }, { "epoch": 1.8337595907928388, "grad_norm": 0.4664636629971284, "learning_rate": 6.389695421721231e-05, "loss": 0.9997, "step": 717 }, { "epoch": 1.836317135549872, "grad_norm": 0.4720160006464846, "learning_rate": 6.383962622591838e-05, "loss": 1.0498, "step": 718 }, { "epoch": 1.8388746803069054, "grad_norm": 0.45150296457653805, "learning_rate": 6.378222219019726e-05, "loss": 1.0491, "step": 719 }, { "epoch": 1.8414322250639388, "grad_norm": 0.43237175184442106, "learning_rate": 6.372474229315823e-05, "loss": 1.0556, "step": 720 }, { "epoch": 1.843989769820972, "grad_norm": 0.3555614837810917, "learning_rate": 6.366718671815255e-05, "loss": 1.0566, "step": 721 }, { "epoch": 1.8465473145780051, "grad_norm": 0.4043617128578834, "learning_rate": 6.360955564877289e-05, "loss": 1.0128, "step": 722 }, { "epoch": 1.8491048593350383, "grad_norm": 0.36791381894407194, "learning_rate": 6.355184926885275e-05, "loss": 1.0398, "step": 723 }, { "epoch": 1.8516624040920715, "grad_norm": 0.41285311729480534, "learning_rate": 6.349406776246587e-05, "loss": 1.0303, "step": 724 }, { "epoch": 1.854219948849105, "grad_norm": 0.47956295629579015, "learning_rate": 6.343621131392558e-05, "loss": 1.0613, "step": 725 }, { "epoch": 1.856777493606138, "grad_norm": 0.35715402007757124, "learning_rate": 6.337828010778427e-05, "loss": 0.9846, "step": 726 }, { "epoch": 1.8593350383631715, "grad_norm": 0.30788741050485563, "learning_rate": 6.332027432883287e-05, "loss": 1.01, "step": 727 }, { "epoch": 1.8618925831202047, "grad_norm": 0.36516318391298225, "learning_rate": 6.326219416210009e-05, "loss": 1.015, "step": 728 }, { "epoch": 1.8644501278772379, "grad_norm": 0.4172573664919271, "learning_rate": 6.320403979285197e-05, "loss": 1.0721, "step": 729 }, { "epoch": 1.867007672634271, "grad_norm": 0.357721021972882, "learning_rate": 6.314581140659124e-05, "loss": 1.027, "step": 730 }, { "epoch": 1.8695652173913042, "grad_norm": 0.3079770159596423, "learning_rate": 6.308750918905674e-05, "loss": 0.9984, "step": 731 }, { "epoch": 1.8721227621483376, "grad_norm": 0.3516516243175642, "learning_rate": 6.302913332622279e-05, "loss": 1.0354, "step": 732 }, { "epoch": 1.8746803069053708, "grad_norm": 0.5749314259929272, "learning_rate": 6.297068400429866e-05, "loss": 1.0471, "step": 733 }, { "epoch": 1.8772378516624042, "grad_norm": 0.6805580946328815, "learning_rate": 6.291216140972793e-05, "loss": 1.0069, "step": 734 }, { "epoch": 1.8797953964194374, "grad_norm": 0.7991805134495602, "learning_rate": 6.28535657291879e-05, "loss": 1.0364, "step": 735 }, { "epoch": 1.8823529411764706, "grad_norm": 0.9755837794916613, "learning_rate": 6.279489714958902e-05, "loss": 1.0685, "step": 736 }, { "epoch": 1.8849104859335037, "grad_norm": 1.0407798737927882, "learning_rate": 6.273615585807423e-05, "loss": 1.0413, "step": 737 }, { "epoch": 1.887468030690537, "grad_norm": 0.6895467560527573, "learning_rate": 6.267734204201848e-05, "loss": 1.0468, "step": 738 }, { "epoch": 1.8900255754475703, "grad_norm": 0.3662552887287085, "learning_rate": 6.261845588902799e-05, "loss": 1.0311, "step": 739 }, { "epoch": 1.8925831202046037, "grad_norm": 0.5316160201986302, "learning_rate": 6.255949758693977e-05, "loss": 1.0032, "step": 740 }, { "epoch": 1.895140664961637, "grad_norm": 0.6379429419889909, "learning_rate": 6.250046732382099e-05, "loss": 1.0131, "step": 741 }, { "epoch": 1.89769820971867, "grad_norm": 0.5804871548161847, "learning_rate": 6.244136528796827e-05, "loss": 1.0191, "step": 742 }, { "epoch": 1.9002557544757033, "grad_norm": 0.43651704808841196, "learning_rate": 6.238219166790728e-05, "loss": 1.0563, "step": 743 }, { "epoch": 1.9028132992327365, "grad_norm": 0.4032559565923256, "learning_rate": 6.232294665239199e-05, "loss": 1.012, "step": 744 }, { "epoch": 1.9053708439897699, "grad_norm": 0.4809968163709181, "learning_rate": 6.22636304304041e-05, "loss": 1.0094, "step": 745 }, { "epoch": 1.907928388746803, "grad_norm": 0.6033128328270763, "learning_rate": 6.220424319115243e-05, "loss": 1.0333, "step": 746 }, { "epoch": 1.9104859335038364, "grad_norm": 0.47176355179184654, "learning_rate": 6.214478512407238e-05, "loss": 1.025, "step": 747 }, { "epoch": 1.9130434782608696, "grad_norm": 0.49617831540264234, "learning_rate": 6.208525641882527e-05, "loss": 1.0401, "step": 748 }, { "epoch": 1.9156010230179028, "grad_norm": 0.4499870092411898, "learning_rate": 6.20256572652977e-05, "loss": 1.0626, "step": 749 }, { "epoch": 1.918158567774936, "grad_norm": 0.3753432726335465, "learning_rate": 6.196598785360104e-05, "loss": 1.0152, "step": 750 }, { "epoch": 1.9207161125319692, "grad_norm": 0.524059742940959, "learning_rate": 6.190624837407075e-05, "loss": 1.0711, "step": 751 }, { "epoch": 1.9232736572890026, "grad_norm": 0.37911876119146704, "learning_rate": 6.184643901726578e-05, "loss": 1.0184, "step": 752 }, { "epoch": 1.9258312020460358, "grad_norm": 0.5046505349197199, "learning_rate": 6.178655997396801e-05, "loss": 1.0341, "step": 753 }, { "epoch": 1.9283887468030692, "grad_norm": 0.5897426305004047, "learning_rate": 6.172661143518158e-05, "loss": 1.0288, "step": 754 }, { "epoch": 1.9309462915601023, "grad_norm": 0.5277155522871224, "learning_rate": 6.16665935921323e-05, "loss": 0.9989, "step": 755 }, { "epoch": 1.9335038363171355, "grad_norm": 0.4455680271469787, "learning_rate": 6.16065066362671e-05, "loss": 1.0358, "step": 756 }, { "epoch": 1.9360613810741687, "grad_norm": 0.37516675425752755, "learning_rate": 6.154635075925333e-05, "loss": 1.0328, "step": 757 }, { "epoch": 1.938618925831202, "grad_norm": 0.31134972244002584, "learning_rate": 6.148612615297819e-05, "loss": 1.0143, "step": 758 }, { "epoch": 1.9411764705882353, "grad_norm": 0.30116788123121785, "learning_rate": 6.14258330095481e-05, "loss": 1.0603, "step": 759 }, { "epoch": 1.9437340153452687, "grad_norm": 0.33789017729539433, "learning_rate": 6.136547152128817e-05, "loss": 1.0642, "step": 760 }, { "epoch": 1.9462915601023019, "grad_norm": 0.3723406702535513, "learning_rate": 6.130504188074141e-05, "loss": 1.0637, "step": 761 }, { "epoch": 1.948849104859335, "grad_norm": 0.3074090795104566, "learning_rate": 6.12445442806683e-05, "loss": 1.0411, "step": 762 }, { "epoch": 1.9514066496163682, "grad_norm": 0.29952175422386457, "learning_rate": 6.118397891404612e-05, "loss": 1.0311, "step": 763 }, { "epoch": 1.9539641943734014, "grad_norm": 0.34505379977188627, "learning_rate": 6.112334597406823e-05, "loss": 1.0455, "step": 764 }, { "epoch": 1.9565217391304348, "grad_norm": 0.3773271960134726, "learning_rate": 6.106264565414363e-05, "loss": 0.978, "step": 765 }, { "epoch": 1.959079283887468, "grad_norm": 0.3577711544939887, "learning_rate": 6.100187814789617e-05, "loss": 1.0441, "step": 766 }, { "epoch": 1.9616368286445014, "grad_norm": 0.3957345794906749, "learning_rate": 6.094104364916405e-05, "loss": 1.0126, "step": 767 }, { "epoch": 1.9641943734015346, "grad_norm": 0.34753096665022937, "learning_rate": 6.088014235199918e-05, "loss": 1.0374, "step": 768 }, { "epoch": 1.9667519181585678, "grad_norm": 0.31284752029157914, "learning_rate": 6.081917445066651e-05, "loss": 1.0469, "step": 769 }, { "epoch": 1.969309462915601, "grad_norm": 0.42998501605296235, "learning_rate": 6.07581401396435e-05, "loss": 1.0298, "step": 770 }, { "epoch": 1.9718670076726341, "grad_norm": 0.4320264632598454, "learning_rate": 6.0697039613619376e-05, "loss": 1.0209, "step": 771 }, { "epoch": 1.9744245524296675, "grad_norm": 0.45573994194726763, "learning_rate": 6.063587306749463e-05, "loss": 1.0364, "step": 772 }, { "epoch": 1.976982097186701, "grad_norm": 0.4479493314499644, "learning_rate": 6.057464069638032e-05, "loss": 1.0454, "step": 773 }, { "epoch": 1.979539641943734, "grad_norm": 0.3489694286760604, "learning_rate": 6.051334269559751e-05, "loss": 1.0121, "step": 774 }, { "epoch": 1.9820971867007673, "grad_norm": 0.2796056957493086, "learning_rate": 6.045197926067657e-05, "loss": 1.0229, "step": 775 }, { "epoch": 1.9846547314578005, "grad_norm": 0.2899329917764029, "learning_rate": 6.03905505873566e-05, "loss": 1.04, "step": 776 }, { "epoch": 1.9872122762148337, "grad_norm": 0.31537011362535095, "learning_rate": 6.032905687158483e-05, "loss": 1.0201, "step": 777 }, { "epoch": 1.989769820971867, "grad_norm": 0.3419075043049475, "learning_rate": 6.026749830951595e-05, "loss": 1.0604, "step": 778 }, { "epoch": 1.9923273657289002, "grad_norm": 0.3779426657587405, "learning_rate": 6.0205875097511474e-05, "loss": 1.042, "step": 779 }, { "epoch": 1.9948849104859336, "grad_norm": 0.31387221943228044, "learning_rate": 6.0144187432139185e-05, "loss": 1.024, "step": 780 }, { "epoch": 1.9974424552429668, "grad_norm": 0.3760802263824882, "learning_rate": 6.008243551017241e-05, "loss": 1.0303, "step": 781 }, { "epoch": 2.0, "grad_norm": 0.48112235112341956, "learning_rate": 6.002061952858949e-05, "loss": 1.0061, "step": 782 }, { "epoch": 2.002557544757033, "grad_norm": 0.4843347641579091, "learning_rate": 5.995873968457307e-05, "loss": 1.0042, "step": 783 }, { "epoch": 2.0051150895140664, "grad_norm": 0.43280419351794497, "learning_rate": 5.989679617550953e-05, "loss": 0.9913, "step": 784 }, { "epoch": 2.0076726342710995, "grad_norm": 0.33817333447982634, "learning_rate": 5.983478919898832e-05, "loss": 0.9942, "step": 785 }, { "epoch": 2.010230179028133, "grad_norm": 0.46364451856775074, "learning_rate": 5.9772718952801335e-05, "loss": 1.077, "step": 786 }, { "epoch": 2.0127877237851663, "grad_norm": 0.6456172502250567, "learning_rate": 5.9710585634942304e-05, "loss": 1.0141, "step": 787 }, { "epoch": 2.0153452685421995, "grad_norm": 0.7087434988801616, "learning_rate": 5.9648389443606136e-05, "loss": 1.024, "step": 788 }, { "epoch": 2.0179028132992327, "grad_norm": 0.6206523201919865, "learning_rate": 5.958613057718828e-05, "loss": 1.008, "step": 789 }, { "epoch": 2.020460358056266, "grad_norm": 0.48723855559372675, "learning_rate": 5.9523809234284155e-05, "loss": 0.9975, "step": 790 }, { "epoch": 2.023017902813299, "grad_norm": 0.3479935955066568, "learning_rate": 5.9461425613688416e-05, "loss": 0.995, "step": 791 }, { "epoch": 2.0255754475703327, "grad_norm": 0.40412834234775397, "learning_rate": 5.9398979914394416e-05, "loss": 0.9994, "step": 792 }, { "epoch": 2.028132992327366, "grad_norm": 0.4719705013531893, "learning_rate": 5.933647233559349e-05, "loss": 0.9927, "step": 793 }, { "epoch": 2.030690537084399, "grad_norm": 0.39255651120492613, "learning_rate": 5.9273903076674374e-05, "loss": 0.9919, "step": 794 }, { "epoch": 2.0332480818414322, "grad_norm": 0.3148451039709822, "learning_rate": 5.921127233722258e-05, "loss": 0.9973, "step": 795 }, { "epoch": 2.0358056265984654, "grad_norm": 0.2961510487834574, "learning_rate": 5.91485803170197e-05, "loss": 1.0086, "step": 796 }, { "epoch": 2.0383631713554986, "grad_norm": 0.318553289880585, "learning_rate": 5.908582721604281e-05, "loss": 1.0301, "step": 797 }, { "epoch": 2.040920716112532, "grad_norm": 0.4333673733070479, "learning_rate": 5.902301323446381e-05, "loss": 1.0039, "step": 798 }, { "epoch": 2.0434782608695654, "grad_norm": 0.5115830801046926, "learning_rate": 5.896013857264884e-05, "loss": 1.0043, "step": 799 }, { "epoch": 2.0460358056265986, "grad_norm": 0.42276666456237705, "learning_rate": 5.889720343115755e-05, "loss": 1.0487, "step": 800 }, { "epoch": 2.0485933503836318, "grad_norm": 0.33643099281393857, "learning_rate": 5.8834208010742534e-05, "loss": 1.0009, "step": 801 }, { "epoch": 2.051150895140665, "grad_norm": 0.3504564236510021, "learning_rate": 5.8771152512348694e-05, "loss": 1.0075, "step": 802 }, { "epoch": 2.053708439897698, "grad_norm": 0.5251143269792232, "learning_rate": 5.870803713711251e-05, "loss": 1.0153, "step": 803 }, { "epoch": 2.0562659846547313, "grad_norm": 0.6328580629452392, "learning_rate": 5.86448620863615e-05, "loss": 1.0436, "step": 804 }, { "epoch": 2.0588235294117645, "grad_norm": 0.6211486830413784, "learning_rate": 5.8581627561613536e-05, "loss": 0.9909, "step": 805 }, { "epoch": 2.061381074168798, "grad_norm": 0.6716809355299344, "learning_rate": 5.8518333764576176e-05, "loss": 1.0042, "step": 806 }, { "epoch": 2.0639386189258313, "grad_norm": 0.7599675406888277, "learning_rate": 5.8454980897146094e-05, "loss": 1.0454, "step": 807 }, { "epoch": 2.0664961636828645, "grad_norm": 0.6684391810626017, "learning_rate": 5.839156916140832e-05, "loss": 1.0176, "step": 808 }, { "epoch": 2.0690537084398977, "grad_norm": 0.41255587095603724, "learning_rate": 5.832809875963573e-05, "loss": 0.9945, "step": 809 }, { "epoch": 2.071611253196931, "grad_norm": 0.4429337990160458, "learning_rate": 5.826456989428832e-05, "loss": 0.9978, "step": 810 }, { "epoch": 2.074168797953964, "grad_norm": 0.5017870623183658, "learning_rate": 5.820098276801252e-05, "loss": 1.0212, "step": 811 }, { "epoch": 2.0767263427109977, "grad_norm": 0.33363936794703103, "learning_rate": 5.813733758364069e-05, "loss": 0.9955, "step": 812 }, { "epoch": 2.079283887468031, "grad_norm": 0.3184226812318926, "learning_rate": 5.8073634544190315e-05, "loss": 0.9961, "step": 813 }, { "epoch": 2.081841432225064, "grad_norm": 0.38065698509998974, "learning_rate": 5.800987385286346e-05, "loss": 1.0569, "step": 814 }, { "epoch": 2.084398976982097, "grad_norm": 0.34937290071950106, "learning_rate": 5.79460557130461e-05, "loss": 0.9854, "step": 815 }, { "epoch": 2.0869565217391304, "grad_norm": 0.30190537548927504, "learning_rate": 5.788218032830742e-05, "loss": 0.9973, "step": 816 }, { "epoch": 2.0895140664961636, "grad_norm": 0.3785973016391384, "learning_rate": 5.781824790239924e-05, "loss": 1.0084, "step": 817 }, { "epoch": 2.0920716112531967, "grad_norm": 0.3560048267667031, "learning_rate": 5.7754258639255334e-05, "loss": 0.9879, "step": 818 }, { "epoch": 2.0946291560102304, "grad_norm": 0.27318601205797294, "learning_rate": 5.769021274299077e-05, "loss": 0.9916, "step": 819 }, { "epoch": 2.0971867007672635, "grad_norm": 0.27970625507687286, "learning_rate": 5.7626110417901236e-05, "loss": 0.9926, "step": 820 }, { "epoch": 2.0997442455242967, "grad_norm": 0.34794308760067744, "learning_rate": 5.7561951868462476e-05, "loss": 1.0049, "step": 821 }, { "epoch": 2.10230179028133, "grad_norm": 0.30884358202442075, "learning_rate": 5.749773729932954e-05, "loss": 0.99, "step": 822 }, { "epoch": 2.104859335038363, "grad_norm": 0.2890559493826192, "learning_rate": 5.743346691533617e-05, "loss": 1.0209, "step": 823 }, { "epoch": 2.1074168797953963, "grad_norm": 0.3795109731603794, "learning_rate": 5.73691409214942e-05, "loss": 1.0133, "step": 824 }, { "epoch": 2.10997442455243, "grad_norm": 0.3707355948082582, "learning_rate": 5.730475952299275e-05, "loss": 1.0144, "step": 825 }, { "epoch": 2.112531969309463, "grad_norm": 0.30752005319876746, "learning_rate": 5.724032292519777e-05, "loss": 1.0104, "step": 826 }, { "epoch": 2.1150895140664963, "grad_norm": 0.38345043941403895, "learning_rate": 5.717583133365125e-05, "loss": 1.0029, "step": 827 }, { "epoch": 2.1176470588235294, "grad_norm": 0.39070708528958853, "learning_rate": 5.7111284954070576e-05, "loss": 1.0131, "step": 828 }, { "epoch": 2.1202046035805626, "grad_norm": 0.45836567684648993, "learning_rate": 5.704668399234795e-05, "loss": 1.017, "step": 829 }, { "epoch": 2.122762148337596, "grad_norm": 0.5856980096882749, "learning_rate": 5.6982028654549624e-05, "loss": 1.0435, "step": 830 }, { "epoch": 2.125319693094629, "grad_norm": 0.419392306323121, "learning_rate": 5.6917319146915354e-05, "loss": 1.0343, "step": 831 }, { "epoch": 2.1278772378516626, "grad_norm": 0.30877657000728803, "learning_rate": 5.685255567585766e-05, "loss": 1.0317, "step": 832 }, { "epoch": 2.130434782608696, "grad_norm": 0.4491035337879607, "learning_rate": 5.678773844796118e-05, "loss": 1.0305, "step": 833 }, { "epoch": 2.132992327365729, "grad_norm": 0.5632576746306452, "learning_rate": 5.6722867669982084e-05, "loss": 0.9911, "step": 834 }, { "epoch": 2.135549872122762, "grad_norm": 0.6141512327447454, "learning_rate": 5.66579435488473e-05, "loss": 1.0223, "step": 835 }, { "epoch": 2.1381074168797953, "grad_norm": 0.4763196937292799, "learning_rate": 5.659296629165393e-05, "loss": 1.0383, "step": 836 }, { "epoch": 2.1406649616368285, "grad_norm": 0.30729108731019755, "learning_rate": 5.652793610566859e-05, "loss": 0.9768, "step": 837 }, { "epoch": 2.1432225063938617, "grad_norm": 0.3227450801791841, "learning_rate": 5.646285319832671e-05, "loss": 1.0356, "step": 838 }, { "epoch": 2.1457800511508953, "grad_norm": 0.38794788918322937, "learning_rate": 5.639771777723189e-05, "loss": 1.0125, "step": 839 }, { "epoch": 2.1483375959079285, "grad_norm": 0.4348976014763873, "learning_rate": 5.633253005015526e-05, "loss": 1.0369, "step": 840 }, { "epoch": 2.1508951406649617, "grad_norm": 0.41931164227914425, "learning_rate": 5.6267290225034775e-05, "loss": 1.0142, "step": 841 }, { "epoch": 2.153452685421995, "grad_norm": 0.27541715665063, "learning_rate": 5.620199850997459e-05, "loss": 1.0262, "step": 842 }, { "epoch": 2.156010230179028, "grad_norm": 0.36736598694146594, "learning_rate": 5.6136655113244386e-05, "loss": 1.0015, "step": 843 }, { "epoch": 2.1585677749360612, "grad_norm": 0.43457711232669183, "learning_rate": 5.607126024327867e-05, "loss": 1.0493, "step": 844 }, { "epoch": 2.1611253196930944, "grad_norm": 0.3685703029484094, "learning_rate": 5.600581410867614e-05, "loss": 1.0086, "step": 845 }, { "epoch": 2.163682864450128, "grad_norm": 0.36431748824284954, "learning_rate": 5.5940316918199084e-05, "loss": 1.0048, "step": 846 }, { "epoch": 2.166240409207161, "grad_norm": 0.35086550909201497, "learning_rate": 5.587476888077258e-05, "loss": 0.9661, "step": 847 }, { "epoch": 2.1687979539641944, "grad_norm": 0.47340488623900007, "learning_rate": 5.580917020548391e-05, "loss": 1.0578, "step": 848 }, { "epoch": 2.1713554987212276, "grad_norm": 0.41591848145887617, "learning_rate": 5.5743521101581934e-05, "loss": 1.0163, "step": 849 }, { "epoch": 2.1739130434782608, "grad_norm": 0.2646123185537355, "learning_rate": 5.56778217784763e-05, "loss": 1.0171, "step": 850 }, { "epoch": 2.176470588235294, "grad_norm": 0.3511706100428371, "learning_rate": 5.561207244573688e-05, "loss": 1.0219, "step": 851 }, { "epoch": 2.1790281329923276, "grad_norm": 0.38155483809818286, "learning_rate": 5.554627331309308e-05, "loss": 1.0355, "step": 852 }, { "epoch": 2.1815856777493607, "grad_norm": 0.44428569276837054, "learning_rate": 5.548042459043313e-05, "loss": 0.9744, "step": 853 }, { "epoch": 2.184143222506394, "grad_norm": 0.3228503410573803, "learning_rate": 5.541452648780348e-05, "loss": 0.9585, "step": 854 }, { "epoch": 2.186700767263427, "grad_norm": 0.3380049356254192, "learning_rate": 5.5348579215408064e-05, "loss": 1.0107, "step": 855 }, { "epoch": 2.1892583120204603, "grad_norm": 0.29475974621755224, "learning_rate": 5.528258298360767e-05, "loss": 1.029, "step": 856 }, { "epoch": 2.1918158567774935, "grad_norm": 0.3037628643580443, "learning_rate": 5.521653800291927e-05, "loss": 1.0448, "step": 857 }, { "epoch": 2.1943734015345266, "grad_norm": 0.29688565337946315, "learning_rate": 5.51504444840153e-05, "loss": 1.0376, "step": 858 }, { "epoch": 2.1969309462915603, "grad_norm": 0.323945767250843, "learning_rate": 5.508430263772308e-05, "loss": 1.0336, "step": 859 }, { "epoch": 2.1994884910485935, "grad_norm": 0.3284522396502332, "learning_rate": 5.501811267502403e-05, "loss": 1.0302, "step": 860 }, { "epoch": 2.2020460358056266, "grad_norm": 0.3592349215963694, "learning_rate": 5.49518748070531e-05, "loss": 1.023, "step": 861 }, { "epoch": 2.20460358056266, "grad_norm": 0.3423858035574432, "learning_rate": 5.488558924509802e-05, "loss": 0.9957, "step": 862 }, { "epoch": 2.207161125319693, "grad_norm": 0.2626938661984674, "learning_rate": 5.481925620059866e-05, "loss": 1.0192, "step": 863 }, { "epoch": 2.209718670076726, "grad_norm": 0.31225095162840705, "learning_rate": 5.4752875885146354e-05, "loss": 1.0212, "step": 864 }, { "epoch": 2.21227621483376, "grad_norm": 0.39268950929217, "learning_rate": 5.468644851048324e-05, "loss": 0.9785, "step": 865 }, { "epoch": 2.214833759590793, "grad_norm": 0.3479708324986449, "learning_rate": 5.461997428850154e-05, "loss": 1.0092, "step": 866 }, { "epoch": 2.217391304347826, "grad_norm": 0.30125519235337805, "learning_rate": 5.4553453431242936e-05, "loss": 1.0514, "step": 867 }, { "epoch": 2.2199488491048593, "grad_norm": 0.3569041941007821, "learning_rate": 5.448688615089784e-05, "loss": 0.9922, "step": 868 }, { "epoch": 2.2225063938618925, "grad_norm": 0.3578777931289321, "learning_rate": 5.442027265980478e-05, "loss": 1.0292, "step": 869 }, { "epoch": 2.2250639386189257, "grad_norm": 0.43031174528074956, "learning_rate": 5.4353613170449636e-05, "loss": 1.0144, "step": 870 }, { "epoch": 2.227621483375959, "grad_norm": 0.3034147546996222, "learning_rate": 5.42869078954651e-05, "loss": 0.995, "step": 871 }, { "epoch": 2.2301790281329925, "grad_norm": 0.3265430475750126, "learning_rate": 5.422015704762981e-05, "loss": 1.016, "step": 872 }, { "epoch": 2.2327365728900257, "grad_norm": 0.4632802224731344, "learning_rate": 5.415336083986786e-05, "loss": 1.0321, "step": 873 }, { "epoch": 2.235294117647059, "grad_norm": 0.42905428916009136, "learning_rate": 5.408651948524799e-05, "loss": 1.0206, "step": 874 }, { "epoch": 2.237851662404092, "grad_norm": 0.35791980721264094, "learning_rate": 5.401963319698295e-05, "loss": 0.9878, "step": 875 }, { "epoch": 2.2404092071611252, "grad_norm": 0.3518593213090125, "learning_rate": 5.395270218842886e-05, "loss": 1.0073, "step": 876 }, { "epoch": 2.2429667519181584, "grad_norm": 0.274369572647955, "learning_rate": 5.388572667308442e-05, "loss": 0.9968, "step": 877 }, { "epoch": 2.2455242966751916, "grad_norm": 0.4023661915440134, "learning_rate": 5.381870686459038e-05, "loss": 1.0038, "step": 878 }, { "epoch": 2.2480818414322252, "grad_norm": 0.4463692936387662, "learning_rate": 5.375164297672872e-05, "loss": 1.0037, "step": 879 }, { "epoch": 2.2506393861892584, "grad_norm": 0.3917913866024726, "learning_rate": 5.368453522342204e-05, "loss": 0.9793, "step": 880 }, { "epoch": 2.2531969309462916, "grad_norm": 0.2696435570953871, "learning_rate": 5.361738381873289e-05, "loss": 0.9997, "step": 881 }, { "epoch": 2.2557544757033248, "grad_norm": 0.2954089772852502, "learning_rate": 5.355018897686302e-05, "loss": 0.9663, "step": 882 }, { "epoch": 2.258312020460358, "grad_norm": 0.30149260615056417, "learning_rate": 5.348295091215276e-05, "loss": 1.0061, "step": 883 }, { "epoch": 2.260869565217391, "grad_norm": 0.2802190777854028, "learning_rate": 5.341566983908031e-05, "loss": 1.0006, "step": 884 }, { "epoch": 2.2634271099744243, "grad_norm": 0.2655762208217684, "learning_rate": 5.3348345972261065e-05, "loss": 1.0199, "step": 885 }, { "epoch": 2.265984654731458, "grad_norm": 0.2351284398309919, "learning_rate": 5.328097952644691e-05, "loss": 1.028, "step": 886 }, { "epoch": 2.268542199488491, "grad_norm": 0.2557435656189813, "learning_rate": 5.321357071652557e-05, "loss": 0.9823, "step": 887 }, { "epoch": 2.2710997442455243, "grad_norm": 0.267763753999149, "learning_rate": 5.3146119757519875e-05, "loss": 0.9887, "step": 888 }, { "epoch": 2.2736572890025575, "grad_norm": 0.27860205774229696, "learning_rate": 5.3078626864587144e-05, "loss": 1.0133, "step": 889 }, { "epoch": 2.2762148337595907, "grad_norm": 0.2987558728085059, "learning_rate": 5.3011092253018434e-05, "loss": 1.0467, "step": 890 }, { "epoch": 2.2787723785166243, "grad_norm": 0.269660583834978, "learning_rate": 5.294351613823787e-05, "loss": 1.0236, "step": 891 }, { "epoch": 2.2813299232736575, "grad_norm": 0.2682105777894826, "learning_rate": 5.287589873580196e-05, "loss": 1.0073, "step": 892 }, { "epoch": 2.2838874680306906, "grad_norm": 0.26935713463820965, "learning_rate": 5.280824026139896e-05, "loss": 1.0137, "step": 893 }, { "epoch": 2.286445012787724, "grad_norm": 0.2632358936931599, "learning_rate": 5.2740540930848094e-05, "loss": 1.0335, "step": 894 }, { "epoch": 2.289002557544757, "grad_norm": 0.26738226375067914, "learning_rate": 5.267280096009892e-05, "loss": 1.0228, "step": 895 }, { "epoch": 2.29156010230179, "grad_norm": 0.2780005314658838, "learning_rate": 5.260502056523063e-05, "loss": 1.0098, "step": 896 }, { "epoch": 2.2941176470588234, "grad_norm": 0.2741454954614826, "learning_rate": 5.253719996245137e-05, "loss": 1.0088, "step": 897 }, { "epoch": 2.296675191815857, "grad_norm": 0.3539765842461498, "learning_rate": 5.246933936809754e-05, "loss": 0.9849, "step": 898 }, { "epoch": 2.29923273657289, "grad_norm": 0.3350858712179409, "learning_rate": 5.240143899863312e-05, "loss": 1.0194, "step": 899 }, { "epoch": 2.3017902813299234, "grad_norm": 0.28258375421118737, "learning_rate": 5.233349907064892e-05, "loss": 1.0043, "step": 900 }, { "epoch": 2.3043478260869565, "grad_norm": 0.2746324008943725, "learning_rate": 5.2265519800861996e-05, "loss": 0.9957, "step": 901 }, { "epoch": 2.3069053708439897, "grad_norm": 0.284357215500465, "learning_rate": 5.219750140611485e-05, "loss": 1.0013, "step": 902 }, { "epoch": 2.309462915601023, "grad_norm": 0.3486207256426231, "learning_rate": 5.212944410337482e-05, "loss": 1.0235, "step": 903 }, { "epoch": 2.312020460358056, "grad_norm": 0.29063234065847066, "learning_rate": 5.206134810973332e-05, "loss": 0.9498, "step": 904 }, { "epoch": 2.3145780051150897, "grad_norm": 0.2783227068727512, "learning_rate": 5.19932136424052e-05, "loss": 0.9856, "step": 905 }, { "epoch": 2.317135549872123, "grad_norm": 0.255357028205985, "learning_rate": 5.192504091872807e-05, "loss": 0.9942, "step": 906 }, { "epoch": 2.319693094629156, "grad_norm": 0.28410051410619885, "learning_rate": 5.185683015616149e-05, "loss": 0.9985, "step": 907 }, { "epoch": 2.3222506393861893, "grad_norm": 0.27168522625319097, "learning_rate": 5.178858157228643e-05, "loss": 1.0386, "step": 908 }, { "epoch": 2.3248081841432224, "grad_norm": 0.2974602944219359, "learning_rate": 5.1720295384804466e-05, "loss": 0.9756, "step": 909 }, { "epoch": 2.3273657289002556, "grad_norm": 0.4194104880758461, "learning_rate": 5.165197181153713e-05, "loss": 1.0509, "step": 910 }, { "epoch": 2.329923273657289, "grad_norm": 0.34398031740039026, "learning_rate": 5.158361107042524e-05, "loss": 1.0279, "step": 911 }, { "epoch": 2.3324808184143224, "grad_norm": 0.4011339549083505, "learning_rate": 5.1515213379528115e-05, "loss": 0.9884, "step": 912 }, { "epoch": 2.3350383631713556, "grad_norm": 0.3280230569283143, "learning_rate": 5.1446778957022984e-05, "loss": 1.0194, "step": 913 }, { "epoch": 2.337595907928389, "grad_norm": 0.2805049932148967, "learning_rate": 5.137830802120422e-05, "loss": 1.0192, "step": 914 }, { "epoch": 2.340153452685422, "grad_norm": 0.28082514853803286, "learning_rate": 5.130980079048269e-05, "loss": 1.0164, "step": 915 }, { "epoch": 2.342710997442455, "grad_norm": 0.3252478527316345, "learning_rate": 5.124125748338501e-05, "loss": 1.0374, "step": 916 }, { "epoch": 2.3452685421994883, "grad_norm": 0.36012680682896564, "learning_rate": 5.117267831855289e-05, "loss": 0.9791, "step": 917 }, { "epoch": 2.3478260869565215, "grad_norm": 0.3562744865725541, "learning_rate": 5.110406351474241e-05, "loss": 1.0071, "step": 918 }, { "epoch": 2.350383631713555, "grad_norm": 0.32552496986950086, "learning_rate": 5.103541329082334e-05, "loss": 0.9929, "step": 919 }, { "epoch": 2.3529411764705883, "grad_norm": 0.28878679693944026, "learning_rate": 5.096672786577843e-05, "loss": 0.9936, "step": 920 }, { "epoch": 2.3554987212276215, "grad_norm": 0.33555234908502024, "learning_rate": 5.089800745870273e-05, "loss": 0.9976, "step": 921 }, { "epoch": 2.3580562659846547, "grad_norm": 0.2852116174307783, "learning_rate": 5.082925228880284e-05, "loss": 0.9972, "step": 922 }, { "epoch": 2.360613810741688, "grad_norm": 0.2858096756505163, "learning_rate": 5.076046257539632e-05, "loss": 0.9994, "step": 923 }, { "epoch": 2.363171355498721, "grad_norm": 0.27416120487035023, "learning_rate": 5.0691638537910806e-05, "loss": 1.0077, "step": 924 }, { "epoch": 2.3657289002557547, "grad_norm": 0.2524911331080638, "learning_rate": 5.0622780395883526e-05, "loss": 0.9952, "step": 925 }, { "epoch": 2.368286445012788, "grad_norm": 0.43007425455994347, "learning_rate": 5.055388836896045e-05, "loss": 0.9719, "step": 926 }, { "epoch": 2.370843989769821, "grad_norm": 0.3162275053374832, "learning_rate": 5.0484962676895624e-05, "loss": 1.0277, "step": 927 }, { "epoch": 2.373401534526854, "grad_norm": 0.3630514647001877, "learning_rate": 5.041600353955054e-05, "loss": 1.0331, "step": 928 }, { "epoch": 2.3759590792838874, "grad_norm": 0.288338048580676, "learning_rate": 5.0347011176893284e-05, "loss": 1.003, "step": 929 }, { "epoch": 2.3785166240409206, "grad_norm": 0.3257712128860222, "learning_rate": 5.0277985808998e-05, "loss": 1.0268, "step": 930 }, { "epoch": 2.381074168797954, "grad_norm": 0.2742545893923996, "learning_rate": 5.020892765604407e-05, "loss": 0.9939, "step": 931 }, { "epoch": 2.3836317135549874, "grad_norm": 0.34791994805077125, "learning_rate": 5.013983693831549e-05, "loss": 1.0188, "step": 932 }, { "epoch": 2.3861892583120206, "grad_norm": 0.29742817188293785, "learning_rate": 5.007071387620009e-05, "loss": 1.0014, "step": 933 }, { "epoch": 2.3887468030690537, "grad_norm": 0.28763416890386856, "learning_rate": 5.000155869018893e-05, "loss": 0.9896, "step": 934 }, { "epoch": 2.391304347826087, "grad_norm": 0.27895542194014733, "learning_rate": 4.993237160087549e-05, "loss": 0.9879, "step": 935 }, { "epoch": 2.39386189258312, "grad_norm": 0.2729422261234032, "learning_rate": 4.986315282895502e-05, "loss": 1.0062, "step": 936 }, { "epoch": 2.3964194373401533, "grad_norm": 0.27170320699542444, "learning_rate": 4.979390259522386e-05, "loss": 1.0087, "step": 937 }, { "epoch": 2.398976982097187, "grad_norm": 0.30230312142914184, "learning_rate": 4.972462112057868e-05, "loss": 1.0173, "step": 938 }, { "epoch": 2.40153452685422, "grad_norm": 0.3184224115524826, "learning_rate": 4.9655308626015835e-05, "loss": 1.0014, "step": 939 }, { "epoch": 2.4040920716112533, "grad_norm": 0.3172883163212683, "learning_rate": 4.958596533263061e-05, "loss": 1.0102, "step": 940 }, { "epoch": 2.4066496163682864, "grad_norm": 0.303355088910324, "learning_rate": 4.9516591461616505e-05, "loss": 0.999, "step": 941 }, { "epoch": 2.4092071611253196, "grad_norm": 0.24585768714005402, "learning_rate": 4.9447187234264615e-05, "loss": 1.0091, "step": 942 }, { "epoch": 2.411764705882353, "grad_norm": 0.3093500884677307, "learning_rate": 4.937775287196282e-05, "loss": 1.0122, "step": 943 }, { "epoch": 2.414322250639386, "grad_norm": 0.41304662460185293, "learning_rate": 4.930828859619515e-05, "loss": 1.0119, "step": 944 }, { "epoch": 2.4168797953964196, "grad_norm": 0.4451137669910999, "learning_rate": 4.923879462854105e-05, "loss": 0.9659, "step": 945 }, { "epoch": 2.419437340153453, "grad_norm": 0.32994372437876596, "learning_rate": 4.9169271190674656e-05, "loss": 0.9856, "step": 946 }, { "epoch": 2.421994884910486, "grad_norm": 0.22818907249672446, "learning_rate": 4.909971850436414e-05, "loss": 1.0361, "step": 947 }, { "epoch": 2.424552429667519, "grad_norm": 0.33962850938503897, "learning_rate": 4.9030136791470976e-05, "loss": 0.9944, "step": 948 }, { "epoch": 2.4271099744245523, "grad_norm": 0.40954174881553046, "learning_rate": 4.896052627394917e-05, "loss": 1.0093, "step": 949 }, { "epoch": 2.4296675191815855, "grad_norm": 0.4257393909425225, "learning_rate": 4.889088717384468e-05, "loss": 0.9912, "step": 950 }, { "epoch": 2.4322250639386187, "grad_norm": 0.30412969316946686, "learning_rate": 4.8821219713294595e-05, "loss": 1.0071, "step": 951 }, { "epoch": 2.4347826086956523, "grad_norm": 0.35293329411505, "learning_rate": 4.8751524114526486e-05, "loss": 0.9719, "step": 952 }, { "epoch": 2.4373401534526855, "grad_norm": 0.4117484756819428, "learning_rate": 4.8681800599857687e-05, "loss": 1.0234, "step": 953 }, { "epoch": 2.4398976982097187, "grad_norm": 0.437336308379765, "learning_rate": 4.861204939169454e-05, "loss": 1.0258, "step": 954 }, { "epoch": 2.442455242966752, "grad_norm": 0.31336295315356344, "learning_rate": 4.854227071253177e-05, "loss": 1.0007, "step": 955 }, { "epoch": 2.445012787723785, "grad_norm": 0.3033170124494707, "learning_rate": 4.847246478495172e-05, "loss": 1.0077, "step": 956 }, { "epoch": 2.4475703324808182, "grad_norm": 0.37625594485986924, "learning_rate": 4.840263183162363e-05, "loss": 0.9793, "step": 957 }, { "epoch": 2.4501278772378514, "grad_norm": 0.4110181159588999, "learning_rate": 4.833277207530298e-05, "loss": 0.9742, "step": 958 }, { "epoch": 2.452685421994885, "grad_norm": 0.3664571890685922, "learning_rate": 4.826288573883071e-05, "loss": 0.9968, "step": 959 }, { "epoch": 2.455242966751918, "grad_norm": 0.27080091028479425, "learning_rate": 4.8192973045132575e-05, "loss": 0.9728, "step": 960 }, { "epoch": 2.4578005115089514, "grad_norm": 0.38098722077365216, "learning_rate": 4.812303421721837e-05, "loss": 1.0399, "step": 961 }, { "epoch": 2.4603580562659846, "grad_norm": 0.46341954095412086, "learning_rate": 4.8053069478181315e-05, "loss": 1.0156, "step": 962 }, { "epoch": 2.4629156010230178, "grad_norm": 0.32489657486295764, "learning_rate": 4.798307905119723e-05, "loss": 1.023, "step": 963 }, { "epoch": 2.4654731457800514, "grad_norm": 0.2790931892033216, "learning_rate": 4.7913063159523896e-05, "loss": 1.0285, "step": 964 }, { "epoch": 2.4680306905370846, "grad_norm": 0.28666393779100274, "learning_rate": 4.784302202650031e-05, "loss": 0.9802, "step": 965 }, { "epoch": 2.4705882352941178, "grad_norm": 0.2756414211620851, "learning_rate": 4.777295587554597e-05, "loss": 0.9958, "step": 966 }, { "epoch": 2.473145780051151, "grad_norm": 0.254767874308196, "learning_rate": 4.770286493016025e-05, "loss": 1.0056, "step": 967 }, { "epoch": 2.475703324808184, "grad_norm": 0.27565052471697443, "learning_rate": 4.7632749413921524e-05, "loss": 1.0109, "step": 968 }, { "epoch": 2.4782608695652173, "grad_norm": 2.156916489907195, "learning_rate": 4.756260955048658e-05, "loss": 1.0193, "step": 969 }, { "epoch": 2.4808184143222505, "grad_norm": 0.34706047060118994, "learning_rate": 4.74924455635899e-05, "loss": 0.9803, "step": 970 }, { "epoch": 2.483375959079284, "grad_norm": 0.4005154747079565, "learning_rate": 4.742225767704286e-05, "loss": 1.0025, "step": 971 }, { "epoch": 2.4859335038363173, "grad_norm": 0.3356815450895581, "learning_rate": 4.7352046114733116e-05, "loss": 1.0224, "step": 972 }, { "epoch": 2.4884910485933505, "grad_norm": 0.34686857253960623, "learning_rate": 4.728181110062383e-05, "loss": 1.0031, "step": 973 }, { "epoch": 2.4910485933503836, "grad_norm": 0.26287696238858543, "learning_rate": 4.721155285875294e-05, "loss": 1.0016, "step": 974 }, { "epoch": 2.493606138107417, "grad_norm": 0.34276165243054646, "learning_rate": 4.714127161323255e-05, "loss": 0.9816, "step": 975 }, { "epoch": 2.49616368286445, "grad_norm": 0.28389546601802473, "learning_rate": 4.707096758824807e-05, "loss": 1.0099, "step": 976 }, { "epoch": 2.498721227621483, "grad_norm": 0.2515028036738449, "learning_rate": 4.7000641008057605e-05, "loss": 0.999, "step": 977 }, { "epoch": 2.501278772378517, "grad_norm": 0.2828216002216222, "learning_rate": 4.6930292096991214e-05, "loss": 1.0151, "step": 978 }, { "epoch": 2.50383631713555, "grad_norm": 0.3728810365049237, "learning_rate": 4.685992107945016e-05, "loss": 1.0189, "step": 979 }, { "epoch": 2.506393861892583, "grad_norm": 0.31696277982756876, "learning_rate": 4.6789528179906254e-05, "loss": 0.9779, "step": 980 }, { "epoch": 2.5089514066496164, "grad_norm": 0.2802033478255871, "learning_rate": 4.67191136229011e-05, "loss": 0.987, "step": 981 }, { "epoch": 2.5115089514066495, "grad_norm": 0.3340397189248965, "learning_rate": 4.664867763304536e-05, "loss": 0.9693, "step": 982 }, { "epoch": 2.5140664961636827, "grad_norm": 0.31946713078301153, "learning_rate": 4.657822043501807e-05, "loss": 0.98, "step": 983 }, { "epoch": 2.516624040920716, "grad_norm": 0.2827099960450771, "learning_rate": 4.6507742253565964e-05, "loss": 1.0107, "step": 984 }, { "epoch": 2.5191815856777495, "grad_norm": 0.27277020153203635, "learning_rate": 4.643724331350265e-05, "loss": 1.0085, "step": 985 }, { "epoch": 2.5217391304347827, "grad_norm": 0.2950110771549243, "learning_rate": 4.6366723839707976e-05, "loss": 0.9817, "step": 986 }, { "epoch": 2.524296675191816, "grad_norm": 0.28737673459457397, "learning_rate": 4.629618405712729e-05, "loss": 0.9931, "step": 987 }, { "epoch": 2.526854219948849, "grad_norm": 0.328976885705382, "learning_rate": 4.6225624190770725e-05, "loss": 1.0179, "step": 988 }, { "epoch": 2.5294117647058822, "grad_norm": 0.28874722164604466, "learning_rate": 4.615504446571248e-05, "loss": 1.0176, "step": 989 }, { "epoch": 2.531969309462916, "grad_norm": 0.2514991876721953, "learning_rate": 4.6084445107090076e-05, "loss": 1.0257, "step": 990 }, { "epoch": 2.5345268542199486, "grad_norm": 0.25531502605497386, "learning_rate": 4.601382634010367e-05, "loss": 1.0067, "step": 991 }, { "epoch": 2.5370843989769822, "grad_norm": 0.3276255896510962, "learning_rate": 4.594318839001539e-05, "loss": 1.0238, "step": 992 }, { "epoch": 2.5396419437340154, "grad_norm": 0.26779804167719207, "learning_rate": 4.587253148214845e-05, "loss": 1.006, "step": 993 }, { "epoch": 2.5421994884910486, "grad_norm": 0.37061278544032084, "learning_rate": 4.580185584188661e-05, "loss": 0.9882, "step": 994 }, { "epoch": 2.544757033248082, "grad_norm": 0.2609641658380784, "learning_rate": 4.573116169467336e-05, "loss": 0.9815, "step": 995 }, { "epoch": 2.547314578005115, "grad_norm": 0.38096136166812233, "learning_rate": 4.5660449266011214e-05, "loss": 0.9984, "step": 996 }, { "epoch": 2.5498721227621486, "grad_norm": 0.33636150127134645, "learning_rate": 4.558971878146104e-05, "loss": 1.0026, "step": 997 }, { "epoch": 2.5524296675191813, "grad_norm": 0.27604661429078775, "learning_rate": 4.551897046664124e-05, "loss": 0.9993, "step": 998 }, { "epoch": 2.554987212276215, "grad_norm": 0.31994197051269924, "learning_rate": 4.5448204547227154e-05, "loss": 0.9949, "step": 999 }, { "epoch": 2.557544757033248, "grad_norm": 0.3511508004578517, "learning_rate": 4.537742124895024e-05, "loss": 0.9982, "step": 1000 }, { "epoch": 2.5601023017902813, "grad_norm": 0.29090223993859365, "learning_rate": 4.530662079759738e-05, "loss": 0.9997, "step": 1001 }, { "epoch": 2.5626598465473145, "grad_norm": 0.2462037300168939, "learning_rate": 4.523580341901023e-05, "loss": 0.9904, "step": 1002 }, { "epoch": 2.5652173913043477, "grad_norm": 0.2746318643511379, "learning_rate": 4.516496933908438e-05, "loss": 1.0181, "step": 1003 }, { "epoch": 2.5677749360613813, "grad_norm": 0.31680827605801093, "learning_rate": 4.509411878376873e-05, "loss": 1.0241, "step": 1004 }, { "epoch": 2.5703324808184145, "grad_norm": 0.34253425537117993, "learning_rate": 4.50232519790647e-05, "loss": 1.0167, "step": 1005 }, { "epoch": 2.5728900255754477, "grad_norm": 0.26338348935218686, "learning_rate": 4.495236915102559e-05, "loss": 0.9547, "step": 1006 }, { "epoch": 2.575447570332481, "grad_norm": 0.33449339157822405, "learning_rate": 4.488147052575577e-05, "loss": 1.0314, "step": 1007 }, { "epoch": 2.578005115089514, "grad_norm": 0.42495311822077536, "learning_rate": 4.481055632941003e-05, "loss": 1.0195, "step": 1008 }, { "epoch": 2.580562659846547, "grad_norm": 0.3543777063618769, "learning_rate": 4.4739626788192806e-05, "loss": 1.0179, "step": 1009 }, { "epoch": 2.5831202046035804, "grad_norm": 0.2734074675391571, "learning_rate": 4.4668682128357474e-05, "loss": 1.0057, "step": 1010 }, { "epoch": 2.585677749360614, "grad_norm": 0.3724948524313399, "learning_rate": 4.45977225762057e-05, "loss": 1.0057, "step": 1011 }, { "epoch": 2.588235294117647, "grad_norm": 0.3795718033122553, "learning_rate": 4.4526748358086566e-05, "loss": 1.0112, "step": 1012 }, { "epoch": 2.5907928388746804, "grad_norm": 0.30997955818137274, "learning_rate": 4.445575970039599e-05, "loss": 0.9937, "step": 1013 }, { "epoch": 2.5933503836317136, "grad_norm": 0.27956010657525027, "learning_rate": 4.438475682957595e-05, "loss": 1.0112, "step": 1014 }, { "epoch": 2.5959079283887467, "grad_norm": 0.3860164174815654, "learning_rate": 4.431373997211372e-05, "loss": 0.9941, "step": 1015 }, { "epoch": 2.59846547314578, "grad_norm": 0.43208138507444893, "learning_rate": 4.424270935454124e-05, "loss": 0.9932, "step": 1016 }, { "epoch": 2.601023017902813, "grad_norm": 0.2756081486534577, "learning_rate": 4.4171665203434336e-05, "loss": 1.0022, "step": 1017 }, { "epoch": 2.6035805626598467, "grad_norm": 0.2986676945368832, "learning_rate": 4.4100607745411947e-05, "loss": 1.0067, "step": 1018 }, { "epoch": 2.60613810741688, "grad_norm": 0.3273660950180001, "learning_rate": 4.4029537207135535e-05, "loss": 0.9788, "step": 1019 }, { "epoch": 2.608695652173913, "grad_norm": 0.28650532791024097, "learning_rate": 4.395845381530823e-05, "loss": 0.9874, "step": 1020 }, { "epoch": 2.6112531969309463, "grad_norm": 0.246812531877351, "learning_rate": 4.3887357796674174e-05, "loss": 0.9998, "step": 1021 }, { "epoch": 2.6138107416879794, "grad_norm": 0.24906583542460717, "learning_rate": 4.381624937801784e-05, "loss": 0.9925, "step": 1022 }, { "epoch": 2.6163682864450126, "grad_norm": 0.2738737665378058, "learning_rate": 4.374512878616315e-05, "loss": 1.0151, "step": 1023 }, { "epoch": 2.618925831202046, "grad_norm": 0.29728516470216987, "learning_rate": 4.3673996247972966e-05, "loss": 0.9945, "step": 1024 }, { "epoch": 2.6214833759590794, "grad_norm": 0.23273105269614583, "learning_rate": 4.360285199034819e-05, "loss": 0.9899, "step": 1025 }, { "epoch": 2.6240409207161126, "grad_norm": 0.29094117249252427, "learning_rate": 4.353169624022713e-05, "loss": 0.9982, "step": 1026 }, { "epoch": 2.626598465473146, "grad_norm": 0.29039727074129185, "learning_rate": 4.346052922458473e-05, "loss": 1.005, "step": 1027 }, { "epoch": 2.629156010230179, "grad_norm": 0.2914886029801254, "learning_rate": 4.3389351170431914e-05, "loss": 1.0143, "step": 1028 }, { "epoch": 2.631713554987212, "grad_norm": 0.25123267713852904, "learning_rate": 4.331816230481478e-05, "loss": 1.0175, "step": 1029 }, { "epoch": 2.634271099744246, "grad_norm": 0.2761513425595422, "learning_rate": 4.3246962854813904e-05, "loss": 1.0032, "step": 1030 }, { "epoch": 2.6368286445012785, "grad_norm": 0.2952763733952678, "learning_rate": 4.317575304754367e-05, "loss": 1.0079, "step": 1031 }, { "epoch": 2.639386189258312, "grad_norm": 0.2396174674261751, "learning_rate": 4.310453311015146e-05, "loss": 0.9751, "step": 1032 }, { "epoch": 2.6419437340153453, "grad_norm": 0.30588512401846724, "learning_rate": 4.3033303269817e-05, "loss": 1.0328, "step": 1033 }, { "epoch": 2.6445012787723785, "grad_norm": 0.35350331623413067, "learning_rate": 4.296206375375156e-05, "loss": 1.011, "step": 1034 }, { "epoch": 2.6470588235294117, "grad_norm": 0.28796718761109585, "learning_rate": 4.289081478919731e-05, "loss": 1.0149, "step": 1035 }, { "epoch": 2.649616368286445, "grad_norm": 0.2383147053710307, "learning_rate": 4.281955660342656e-05, "loss": 0.9874, "step": 1036 }, { "epoch": 2.6521739130434785, "grad_norm": 0.27157160483463727, "learning_rate": 4.274828942374103e-05, "loss": 0.9949, "step": 1037 }, { "epoch": 2.6547314578005117, "grad_norm": 0.2534384580666098, "learning_rate": 4.2677013477471104e-05, "loss": 0.9997, "step": 1038 }, { "epoch": 2.657289002557545, "grad_norm": 0.22567843700077125, "learning_rate": 4.260572899197517e-05, "loss": 1.0078, "step": 1039 }, { "epoch": 2.659846547314578, "grad_norm": 0.29838728779169416, "learning_rate": 4.253443619463882e-05, "loss": 1.0185, "step": 1040 }, { "epoch": 2.662404092071611, "grad_norm": 0.25199620198627026, "learning_rate": 4.246313531287418e-05, "loss": 0.9965, "step": 1041 }, { "epoch": 2.6649616368286444, "grad_norm": 0.25813433899181243, "learning_rate": 4.2391826574119156e-05, "loss": 1.0276, "step": 1042 }, { "epoch": 2.6675191815856776, "grad_norm": 0.2213842255776451, "learning_rate": 4.2320510205836714e-05, "loss": 0.9618, "step": 1043 }, { "epoch": 2.670076726342711, "grad_norm": 0.34212252067273924, "learning_rate": 4.2249186435514184e-05, "loss": 1.0035, "step": 1044 }, { "epoch": 2.6726342710997444, "grad_norm": 0.23979377486467365, "learning_rate": 4.217785549066245e-05, "loss": 0.9896, "step": 1045 }, { "epoch": 2.6751918158567776, "grad_norm": 0.2138184617430876, "learning_rate": 4.2106517598815337e-05, "loss": 0.9743, "step": 1046 }, { "epoch": 2.6777493606138107, "grad_norm": 0.2659130978637147, "learning_rate": 4.203517298752881e-05, "loss": 1.0077, "step": 1047 }, { "epoch": 2.680306905370844, "grad_norm": 0.30697679317607074, "learning_rate": 4.1963821884380246e-05, "loss": 1.008, "step": 1048 }, { "epoch": 2.682864450127877, "grad_norm": 0.22999291962370588, "learning_rate": 4.1892464516967776e-05, "loss": 0.9824, "step": 1049 }, { "epoch": 2.6854219948849103, "grad_norm": 0.26998959283360185, "learning_rate": 4.182110111290948e-05, "loss": 1.0083, "step": 1050 }, { "epoch": 2.687979539641944, "grad_norm": 0.29486316071332647, "learning_rate": 4.17497318998427e-05, "loss": 1.0048, "step": 1051 }, { "epoch": 2.690537084398977, "grad_norm": 0.2986425273868719, "learning_rate": 4.1678357105423306e-05, "loss": 0.9973, "step": 1052 }, { "epoch": 2.6930946291560103, "grad_norm": 0.25347000823131294, "learning_rate": 4.1606976957324976e-05, "loss": 1.01, "step": 1053 }, { "epoch": 2.6956521739130435, "grad_norm": 0.25469040517907776, "learning_rate": 4.153559168323847e-05, "loss": 0.9737, "step": 1054 }, { "epoch": 2.6982097186700766, "grad_norm": 0.2468042191556332, "learning_rate": 4.146420151087089e-05, "loss": 0.9875, "step": 1055 }, { "epoch": 2.70076726342711, "grad_norm": 0.3168078654851793, "learning_rate": 4.1392806667944965e-05, "loss": 1.0072, "step": 1056 }, { "epoch": 2.703324808184143, "grad_norm": 0.2577802930863439, "learning_rate": 4.132140738219831e-05, "loss": 1.0074, "step": 1057 }, { "epoch": 2.7058823529411766, "grad_norm": 0.22240516058092002, "learning_rate": 4.125000388138276e-05, "loss": 0.9964, "step": 1058 }, { "epoch": 2.70843989769821, "grad_norm": 0.2899408076786431, "learning_rate": 4.117859639326352e-05, "loss": 1.0246, "step": 1059 }, { "epoch": 2.710997442455243, "grad_norm": 0.24816779206581488, "learning_rate": 4.1107185145618574e-05, "loss": 0.9832, "step": 1060 }, { "epoch": 2.713554987212276, "grad_norm": 0.35105713611659695, "learning_rate": 4.10357703662379e-05, "loss": 1.0324, "step": 1061 }, { "epoch": 2.7161125319693094, "grad_norm": 0.23582025700520193, "learning_rate": 4.096435228292268e-05, "loss": 1.0105, "step": 1062 }, { "epoch": 2.718670076726343, "grad_norm": 0.224766329004601, "learning_rate": 4.089293112348469e-05, "loss": 1.043, "step": 1063 }, { "epoch": 2.7212276214833757, "grad_norm": 0.2607269768142398, "learning_rate": 4.082150711574552e-05, "loss": 1.007, "step": 1064 }, { "epoch": 2.7237851662404093, "grad_norm": 0.22627614919460648, "learning_rate": 4.0750080487535807e-05, "loss": 0.9762, "step": 1065 }, { "epoch": 2.7263427109974425, "grad_norm": 0.2541558036254898, "learning_rate": 4.067865146669459e-05, "loss": 0.9949, "step": 1066 }, { "epoch": 2.7289002557544757, "grad_norm": 0.23805701026486054, "learning_rate": 4.060722028106849e-05, "loss": 1.0013, "step": 1067 }, { "epoch": 2.731457800511509, "grad_norm": 0.26429490418987045, "learning_rate": 4.053578715851109e-05, "loss": 1.0213, "step": 1068 }, { "epoch": 2.734015345268542, "grad_norm": 0.24156276865731632, "learning_rate": 4.046435232688211e-05, "loss": 0.9954, "step": 1069 }, { "epoch": 2.7365728900255757, "grad_norm": 0.2882909987018262, "learning_rate": 4.0392916014046735e-05, "loss": 1.0414, "step": 1070 }, { "epoch": 2.7391304347826084, "grad_norm": 0.20412597587654827, "learning_rate": 4.0321478447874895e-05, "loss": 0.9911, "step": 1071 }, { "epoch": 2.741687979539642, "grad_norm": 0.24979733236444654, "learning_rate": 4.0250039856240476e-05, "loss": 1.0219, "step": 1072 }, { "epoch": 2.7442455242966752, "grad_norm": 0.24634392488290774, "learning_rate": 4.0178600467020664e-05, "loss": 1.0048, "step": 1073 }, { "epoch": 2.7468030690537084, "grad_norm": 0.26158148075151116, "learning_rate": 4.010716050809517e-05, "loss": 0.9903, "step": 1074 }, { "epoch": 2.7493606138107416, "grad_norm": 0.2179061990457055, "learning_rate": 4.0035720207345556e-05, "loss": 1.0435, "step": 1075 }, { "epoch": 2.7519181585677748, "grad_norm": 0.2569251284054396, "learning_rate": 3.9964279792654464e-05, "loss": 1.0001, "step": 1076 }, { "epoch": 2.7544757033248084, "grad_norm": 0.2834854817803271, "learning_rate": 3.989283949190484e-05, "loss": 1.0211, "step": 1077 }, { "epoch": 2.7570332480818416, "grad_norm": 0.24647968373722906, "learning_rate": 3.982139953297935e-05, "loss": 0.9836, "step": 1078 }, { "epoch": 2.7595907928388748, "grad_norm": 0.2603116983015655, "learning_rate": 3.974996014375954e-05, "loss": 0.9624, "step": 1079 }, { "epoch": 2.762148337595908, "grad_norm": 0.26774223660199403, "learning_rate": 3.967852155212511e-05, "loss": 1.0087, "step": 1080 }, { "epoch": 2.764705882352941, "grad_norm": 0.25111070048702194, "learning_rate": 3.9607083985953265e-05, "loss": 0.9716, "step": 1081 }, { "epoch": 2.7672634271099743, "grad_norm": 0.27537157062354006, "learning_rate": 3.95356476731179e-05, "loss": 1.009, "step": 1082 }, { "epoch": 2.7698209718670075, "grad_norm": 0.26627222201056977, "learning_rate": 3.946421284148892e-05, "loss": 1.0071, "step": 1083 }, { "epoch": 2.772378516624041, "grad_norm": 0.2647256382334767, "learning_rate": 3.939277971893152e-05, "loss": 1.0099, "step": 1084 }, { "epoch": 2.7749360613810743, "grad_norm": 0.3161240038333297, "learning_rate": 3.932134853330542e-05, "loss": 0.9844, "step": 1085 }, { "epoch": 2.7774936061381075, "grad_norm": 0.24673702901545966, "learning_rate": 3.92499195124642e-05, "loss": 1.0384, "step": 1086 }, { "epoch": 2.7800511508951407, "grad_norm": 0.2627376460002893, "learning_rate": 3.91784928842545e-05, "loss": 0.9644, "step": 1087 }, { "epoch": 2.782608695652174, "grad_norm": 0.25015861128564604, "learning_rate": 3.9107068876515314e-05, "loss": 0.9907, "step": 1088 }, { "epoch": 2.785166240409207, "grad_norm": 0.25505938008111867, "learning_rate": 3.9035647717077335e-05, "loss": 1.0117, "step": 1089 }, { "epoch": 2.78772378516624, "grad_norm": 0.2885186518776429, "learning_rate": 3.896422963376211e-05, "loss": 0.9782, "step": 1090 }, { "epoch": 2.790281329923274, "grad_norm": 0.24772701208848075, "learning_rate": 3.8892814854381425e-05, "loss": 0.9543, "step": 1091 }, { "epoch": 2.792838874680307, "grad_norm": 0.3043825900169142, "learning_rate": 3.882140360673648e-05, "loss": 1.0135, "step": 1092 }, { "epoch": 2.79539641943734, "grad_norm": 0.26534262469691305, "learning_rate": 3.8749996118617256e-05, "loss": 1.0158, "step": 1093 }, { "epoch": 2.7979539641943734, "grad_norm": 0.2439645917934899, "learning_rate": 3.86785926178017e-05, "loss": 0.9991, "step": 1094 }, { "epoch": 2.8005115089514065, "grad_norm": 0.2963910857686297, "learning_rate": 3.860719333205505e-05, "loss": 0.9949, "step": 1095 }, { "epoch": 2.80306905370844, "grad_norm": 0.3347355595394114, "learning_rate": 3.853579848912912e-05, "loss": 1.043, "step": 1096 }, { "epoch": 2.805626598465473, "grad_norm": 0.23698370137396493, "learning_rate": 3.846440831676153e-05, "loss": 1.0109, "step": 1097 }, { "epoch": 2.8081841432225065, "grad_norm": 0.2553667731091932, "learning_rate": 3.839302304267504e-05, "loss": 1.0243, "step": 1098 }, { "epoch": 2.8107416879795397, "grad_norm": 0.25553599743585725, "learning_rate": 3.832164289457671e-05, "loss": 1.0148, "step": 1099 }, { "epoch": 2.813299232736573, "grad_norm": 0.20938583263905827, "learning_rate": 3.8250268100157304e-05, "loss": 0.9899, "step": 1100 }, { "epoch": 2.815856777493606, "grad_norm": 0.2803140720481866, "learning_rate": 3.817889888709053e-05, "loss": 1.0172, "step": 1101 }, { "epoch": 2.8184143222506393, "grad_norm": 0.26697443095676276, "learning_rate": 3.8107535483032224e-05, "loss": 0.999, "step": 1102 }, { "epoch": 2.820971867007673, "grad_norm": 0.24132691739947107, "learning_rate": 3.803617811561976e-05, "loss": 1.0137, "step": 1103 }, { "epoch": 2.8235294117647056, "grad_norm": 0.28295370026487926, "learning_rate": 3.796482701247121e-05, "loss": 0.9906, "step": 1104 }, { "epoch": 2.8260869565217392, "grad_norm": 0.2363477732559459, "learning_rate": 3.789348240118468e-05, "loss": 0.9807, "step": 1105 }, { "epoch": 2.8286445012787724, "grad_norm": 0.2870494002943626, "learning_rate": 3.782214450933756e-05, "loss": 1.0106, "step": 1106 }, { "epoch": 2.8312020460358056, "grad_norm": 0.24656039873041008, "learning_rate": 3.775081356448583e-05, "loss": 1.0069, "step": 1107 }, { "epoch": 2.833759590792839, "grad_norm": 0.24824680931599624, "learning_rate": 3.7679489794163286e-05, "loss": 1.017, "step": 1108 }, { "epoch": 2.836317135549872, "grad_norm": 0.25584210376463995, "learning_rate": 3.760817342588086e-05, "loss": 1.0209, "step": 1109 }, { "epoch": 2.8388746803069056, "grad_norm": 0.23768463941469484, "learning_rate": 3.753686468712583e-05, "loss": 1.002, "step": 1110 }, { "epoch": 2.8414322250639388, "grad_norm": 0.2140982655494908, "learning_rate": 3.746556380536119e-05, "loss": 1.0018, "step": 1111 }, { "epoch": 2.843989769820972, "grad_norm": 0.2322699559270119, "learning_rate": 3.7394271008024834e-05, "loss": 1.0102, "step": 1112 }, { "epoch": 2.846547314578005, "grad_norm": 0.2541795231409938, "learning_rate": 3.7322986522528896e-05, "loss": 0.9998, "step": 1113 }, { "epoch": 2.8491048593350383, "grad_norm": 0.2711876988209337, "learning_rate": 3.7251710576258966e-05, "loss": 1.0001, "step": 1114 }, { "epoch": 2.8516624040920715, "grad_norm": 0.23402242185528463, "learning_rate": 3.7180443396573445e-05, "loss": 1.0051, "step": 1115 }, { "epoch": 2.8542199488491047, "grad_norm": 0.225618095882578, "learning_rate": 3.71091852108027e-05, "loss": 0.997, "step": 1116 }, { "epoch": 2.8567774936061383, "grad_norm": 0.2373879922383219, "learning_rate": 3.7037936246248455e-05, "loss": 0.9823, "step": 1117 }, { "epoch": 2.8593350383631715, "grad_norm": 0.2179619293216389, "learning_rate": 3.696669673018302e-05, "loss": 1.014, "step": 1118 }, { "epoch": 2.8618925831202047, "grad_norm": 0.23271028414622047, "learning_rate": 3.689546688984854e-05, "loss": 0.9867, "step": 1119 }, { "epoch": 2.864450127877238, "grad_norm": 0.22959905790536927, "learning_rate": 3.6824246952456346e-05, "loss": 1.0419, "step": 1120 }, { "epoch": 2.867007672634271, "grad_norm": 0.22427029568274529, "learning_rate": 3.675303714518611e-05, "loss": 1.0103, "step": 1121 }, { "epoch": 2.869565217391304, "grad_norm": 0.23497147356312345, "learning_rate": 3.6681837695185235e-05, "loss": 0.9924, "step": 1122 }, { "epoch": 2.8721227621483374, "grad_norm": 0.2313618533257831, "learning_rate": 3.66106488295681e-05, "loss": 0.9621, "step": 1123 }, { "epoch": 2.874680306905371, "grad_norm": 0.21491888385362423, "learning_rate": 3.653947077541527e-05, "loss": 1.0429, "step": 1124 }, { "epoch": 2.877237851662404, "grad_norm": 0.2142173562919566, "learning_rate": 3.646830375977288e-05, "loss": 1.0197, "step": 1125 }, { "epoch": 2.8797953964194374, "grad_norm": 0.22758855581505263, "learning_rate": 3.639714800965183e-05, "loss": 1.0112, "step": 1126 }, { "epoch": 2.8823529411764706, "grad_norm": 0.2037855866672826, "learning_rate": 3.632600375202704e-05, "loss": 0.9783, "step": 1127 }, { "epoch": 2.8849104859335037, "grad_norm": 0.245178868595634, "learning_rate": 3.625487121383686e-05, "loss": 0.9839, "step": 1128 }, { "epoch": 2.887468030690537, "grad_norm": 0.24685844906961088, "learning_rate": 3.6183750621982176e-05, "loss": 0.9899, "step": 1129 }, { "epoch": 2.89002557544757, "grad_norm": 0.30768839678374543, "learning_rate": 3.611264220332583e-05, "loss": 1.0155, "step": 1130 }, { "epoch": 2.8925831202046037, "grad_norm": 0.27604427834359535, "learning_rate": 3.60415461846918e-05, "loss": 1.0281, "step": 1131 }, { "epoch": 2.895140664961637, "grad_norm": 0.2953227328216093, "learning_rate": 3.5970462792864486e-05, "loss": 0.9653, "step": 1132 }, { "epoch": 2.89769820971867, "grad_norm": 0.2776299135622876, "learning_rate": 3.589939225458807e-05, "loss": 0.9844, "step": 1133 }, { "epoch": 2.9002557544757033, "grad_norm": 0.25982730487682326, "learning_rate": 3.582833479656568e-05, "loss": 0.9857, "step": 1134 }, { "epoch": 2.9028132992327365, "grad_norm": 0.272622930597586, "learning_rate": 3.575729064545876e-05, "loss": 1.01, "step": 1135 }, { "epoch": 2.90537084398977, "grad_norm": 0.28024133642600774, "learning_rate": 3.568626002788628e-05, "loss": 0.9812, "step": 1136 }, { "epoch": 2.907928388746803, "grad_norm": 0.21927470498778384, "learning_rate": 3.5615243170424074e-05, "loss": 1.0438, "step": 1137 }, { "epoch": 2.9104859335038364, "grad_norm": 0.2407505627513316, "learning_rate": 3.5544240299604024e-05, "loss": 0.9827, "step": 1138 }, { "epoch": 2.9130434782608696, "grad_norm": 0.2583021004717082, "learning_rate": 3.547325164191345e-05, "loss": 0.9976, "step": 1139 }, { "epoch": 2.915601023017903, "grad_norm": 0.23084176420891026, "learning_rate": 3.5402277423794316e-05, "loss": 0.9897, "step": 1140 }, { "epoch": 2.918158567774936, "grad_norm": 0.258034589423058, "learning_rate": 3.5331317871642525e-05, "loss": 1.0057, "step": 1141 }, { "epoch": 2.920716112531969, "grad_norm": 0.3037754487191455, "learning_rate": 3.526037321180722e-05, "loss": 1.0152, "step": 1142 }, { "epoch": 2.923273657289003, "grad_norm": 0.20675604578569703, "learning_rate": 3.518944367058999e-05, "loss": 1.0117, "step": 1143 }, { "epoch": 2.9258312020460355, "grad_norm": 0.278991208465292, "learning_rate": 3.511852947424424e-05, "loss": 0.99, "step": 1144 }, { "epoch": 2.928388746803069, "grad_norm": 0.24889238741757388, "learning_rate": 3.504763084897442e-05, "loss": 0.9982, "step": 1145 }, { "epoch": 2.9309462915601023, "grad_norm": 0.2497728135979749, "learning_rate": 3.4976748020935305e-05, "loss": 0.9852, "step": 1146 }, { "epoch": 2.9335038363171355, "grad_norm": 0.27461155792883407, "learning_rate": 3.4905881216231284e-05, "loss": 1.0029, "step": 1147 }, { "epoch": 2.9360613810741687, "grad_norm": 0.22344834081280318, "learning_rate": 3.4835030660915634e-05, "loss": 1.0008, "step": 1148 }, { "epoch": 2.938618925831202, "grad_norm": 0.25313066405291, "learning_rate": 3.476419658098978e-05, "loss": 1.0166, "step": 1149 }, { "epoch": 2.9411764705882355, "grad_norm": 0.2368466744031409, "learning_rate": 3.469337920240263e-05, "loss": 1.0058, "step": 1150 }, { "epoch": 2.9437340153452687, "grad_norm": 0.260205430555691, "learning_rate": 3.462257875104977e-05, "loss": 1.0211, "step": 1151 }, { "epoch": 2.946291560102302, "grad_norm": 0.24750655038016556, "learning_rate": 3.455179545277286e-05, "loss": 0.9922, "step": 1152 }, { "epoch": 2.948849104859335, "grad_norm": 0.26061134046269646, "learning_rate": 3.4481029533358775e-05, "loss": 1.0137, "step": 1153 }, { "epoch": 2.9514066496163682, "grad_norm": 0.23020961645395976, "learning_rate": 3.441028121853898e-05, "loss": 1.0183, "step": 1154 }, { "epoch": 2.9539641943734014, "grad_norm": 0.22552716653712218, "learning_rate": 3.43395507339888e-05, "loss": 1.0343, "step": 1155 }, { "epoch": 2.9565217391304346, "grad_norm": 0.22338059564503102, "learning_rate": 3.426883830532665e-05, "loss": 1.0078, "step": 1156 }, { "epoch": 2.959079283887468, "grad_norm": 0.26129239518427483, "learning_rate": 3.4198144158113405e-05, "loss": 0.9989, "step": 1157 }, { "epoch": 2.9616368286445014, "grad_norm": 0.2292830652258008, "learning_rate": 3.412746851785156e-05, "loss": 0.9938, "step": 1158 }, { "epoch": 2.9641943734015346, "grad_norm": 0.24649745683965166, "learning_rate": 3.405681160998463e-05, "loss": 0.9934, "step": 1159 }, { "epoch": 2.9667519181585678, "grad_norm": 0.23007278451720706, "learning_rate": 3.3986173659896335e-05, "loss": 1.015, "step": 1160 }, { "epoch": 2.969309462915601, "grad_norm": 0.1758081567513317, "learning_rate": 3.3915554892909944e-05, "loss": 0.985, "step": 1161 }, { "epoch": 2.971867007672634, "grad_norm": 0.2255832579224465, "learning_rate": 3.384495553428754e-05, "loss": 0.9894, "step": 1162 }, { "epoch": 2.9744245524296673, "grad_norm": 0.21165101777080628, "learning_rate": 3.3774375809229275e-05, "loss": 1.0311, "step": 1163 }, { "epoch": 2.976982097186701, "grad_norm": 0.23815079825461055, "learning_rate": 3.370381594287273e-05, "loss": 0.985, "step": 1164 }, { "epoch": 2.979539641943734, "grad_norm": 0.2196988840192207, "learning_rate": 3.3633276160292044e-05, "loss": 0.9935, "step": 1165 }, { "epoch": 2.9820971867007673, "grad_norm": 0.20489499569969788, "learning_rate": 3.3562756686497364e-05, "loss": 1.0254, "step": 1166 }, { "epoch": 2.9846547314578005, "grad_norm": 0.19784059419234884, "learning_rate": 3.349225774643405e-05, "loss": 1.0012, "step": 1167 }, { "epoch": 2.9872122762148337, "grad_norm": 0.18918392336436632, "learning_rate": 3.342177956498193e-05, "loss": 1.0156, "step": 1168 }, { "epoch": 2.9897698209718673, "grad_norm": 0.20588207839076775, "learning_rate": 3.3351322366954656e-05, "loss": 0.9923, "step": 1169 }, { "epoch": 2.9923273657289, "grad_norm": 0.19126183243148004, "learning_rate": 3.328088637709892e-05, "loss": 0.9903, "step": 1170 }, { "epoch": 2.9948849104859336, "grad_norm": 0.20591684079974318, "learning_rate": 3.321047182009375e-05, "loss": 0.9872, "step": 1171 }, { "epoch": 2.997442455242967, "grad_norm": 0.20972832124324786, "learning_rate": 3.314007892054985e-05, "loss": 1.0294, "step": 1172 }, { "epoch": 3.0, "grad_norm": 0.19979271297295348, "learning_rate": 3.30697079030088e-05, "loss": 1.0199, "step": 1173 }, { "epoch": 3.002557544757033, "grad_norm": 0.20927717303693083, "learning_rate": 3.29993589919424e-05, "loss": 1.0065, "step": 1174 }, { "epoch": 3.0051150895140664, "grad_norm": 0.1894544897464941, "learning_rate": 3.292903241175195e-05, "loss": 1.0053, "step": 1175 }, { "epoch": 3.0076726342710995, "grad_norm": 0.19263685473563882, "learning_rate": 3.285872838676746e-05, "loss": 0.9849, "step": 1176 }, { "epoch": 3.010230179028133, "grad_norm": 0.20499176394344015, "learning_rate": 3.278844714124707e-05, "loss": 1.0211, "step": 1177 }, { "epoch": 3.0127877237851663, "grad_norm": 0.2267745823697339, "learning_rate": 3.2718188899376183e-05, "loss": 1.0096, "step": 1178 }, { "epoch": 3.0153452685421995, "grad_norm": 0.19329973297024897, "learning_rate": 3.264795388526689e-05, "loss": 0.9608, "step": 1179 }, { "epoch": 3.0179028132992327, "grad_norm": 0.2456406558753071, "learning_rate": 3.257774232295714e-05, "loss": 0.9707, "step": 1180 }, { "epoch": 3.020460358056266, "grad_norm": 0.22184901987844155, "learning_rate": 3.250755443641012e-05, "loss": 0.9833, "step": 1181 }, { "epoch": 3.023017902813299, "grad_norm": 0.20437847819940258, "learning_rate": 3.243739044951343e-05, "loss": 0.9764, "step": 1182 }, { "epoch": 3.0255754475703327, "grad_norm": 0.20756135809272994, "learning_rate": 3.2367250586078496e-05, "loss": 0.9979, "step": 1183 }, { "epoch": 3.028132992327366, "grad_norm": 0.21183456404253598, "learning_rate": 3.2297135069839766e-05, "loss": 0.9648, "step": 1184 }, { "epoch": 3.030690537084399, "grad_norm": 0.19385697959377585, "learning_rate": 3.2227044124454024e-05, "loss": 0.979, "step": 1185 }, { "epoch": 3.0332480818414322, "grad_norm": 0.24603982366166482, "learning_rate": 3.215697797349972e-05, "loss": 0.9618, "step": 1186 }, { "epoch": 3.0358056265984654, "grad_norm": 0.1921193878500632, "learning_rate": 3.2086936840476124e-05, "loss": 0.9852, "step": 1187 }, { "epoch": 3.0383631713554986, "grad_norm": 0.22131221807833543, "learning_rate": 3.201692094880278e-05, "loss": 0.9512, "step": 1188 }, { "epoch": 3.040920716112532, "grad_norm": 0.26400713025589573, "learning_rate": 3.194693052181869e-05, "loss": 0.9841, "step": 1189 }, { "epoch": 3.0434782608695654, "grad_norm": 0.18604109405773514, "learning_rate": 3.1876965782781626e-05, "loss": 0.9725, "step": 1190 }, { "epoch": 3.0460358056265986, "grad_norm": 0.32633685255896494, "learning_rate": 3.1807026954867445e-05, "loss": 0.9523, "step": 1191 }, { "epoch": 3.0485933503836318, "grad_norm": 0.21116781754958366, "learning_rate": 3.173711426116931e-05, "loss": 0.9898, "step": 1192 }, { "epoch": 3.051150895140665, "grad_norm": 0.2464885736886057, "learning_rate": 3.1667227924697033e-05, "loss": 0.9975, "step": 1193 }, { "epoch": 3.053708439897698, "grad_norm": 0.2174756211201157, "learning_rate": 3.1597368168376376e-05, "loss": 0.9695, "step": 1194 }, { "epoch": 3.0562659846547313, "grad_norm": 0.21679630916030693, "learning_rate": 3.1527535215048284e-05, "loss": 0.9757, "step": 1195 }, { "epoch": 3.0588235294117645, "grad_norm": 0.19385880112351023, "learning_rate": 3.1457729287468225e-05, "loss": 0.9874, "step": 1196 }, { "epoch": 3.061381074168798, "grad_norm": 0.2307908579955019, "learning_rate": 3.138795060830549e-05, "loss": 0.9883, "step": 1197 }, { "epoch": 3.0639386189258313, "grad_norm": 0.19918037708419395, "learning_rate": 3.131819940014234e-05, "loss": 0.9853, "step": 1198 }, { "epoch": 3.0664961636828645, "grad_norm": 0.27259883681321456, "learning_rate": 3.1248475885473534e-05, "loss": 1.0029, "step": 1199 }, { "epoch": 3.0690537084398977, "grad_norm": 0.21850928888775867, "learning_rate": 3.117878028670542e-05, "loss": 0.9627, "step": 1200 }, { "epoch": 3.071611253196931, "grad_norm": 0.26311180789408495, "learning_rate": 3.110911282615533e-05, "loss": 0.9749, "step": 1201 }, { "epoch": 3.074168797953964, "grad_norm": 0.2365413768196383, "learning_rate": 3.103947372605084e-05, "loss": 0.9971, "step": 1202 }, { "epoch": 3.0767263427109977, "grad_norm": 0.20574394116778202, "learning_rate": 3.0969863208529044e-05, "loss": 1.0029, "step": 1203 }, { "epoch": 3.079283887468031, "grad_norm": 0.2285897523783578, "learning_rate": 3.090028149563587e-05, "loss": 1.0025, "step": 1204 }, { "epoch": 3.081841432225064, "grad_norm": 0.2062668987784313, "learning_rate": 3.083072880932535e-05, "loss": 1.0031, "step": 1205 }, { "epoch": 3.084398976982097, "grad_norm": 0.19032654177423064, "learning_rate": 3.0761205371458953e-05, "loss": 0.974, "step": 1206 }, { "epoch": 3.0869565217391304, "grad_norm": 0.21207164634258652, "learning_rate": 3.069171140380485e-05, "loss": 0.9952, "step": 1207 }, { "epoch": 3.0895140664961636, "grad_norm": 0.21972821827355038, "learning_rate": 3.06222471280372e-05, "loss": 1.0226, "step": 1208 }, { "epoch": 3.0920716112531967, "grad_norm": 0.20237512777327865, "learning_rate": 3.055281276573539e-05, "loss": 0.9692, "step": 1209 }, { "epoch": 3.0946291560102304, "grad_norm": 0.2216867068993571, "learning_rate": 3.04834085383835e-05, "loss": 0.9928, "step": 1210 }, { "epoch": 3.0971867007672635, "grad_norm": 0.19513855102804212, "learning_rate": 3.04140346673694e-05, "loss": 0.9933, "step": 1211 }, { "epoch": 3.0997442455242967, "grad_norm": 0.21612553792582045, "learning_rate": 3.034469137398417e-05, "loss": 0.9588, "step": 1212 }, { "epoch": 3.10230179028133, "grad_norm": 0.197236734236815, "learning_rate": 3.0275378879421313e-05, "loss": 1.0255, "step": 1213 }, { "epoch": 3.104859335038363, "grad_norm": 0.20305992275112705, "learning_rate": 3.0206097404776158e-05, "loss": 0.9745, "step": 1214 }, { "epoch": 3.1074168797953963, "grad_norm": 0.20883263282580242, "learning_rate": 3.0136847171044996e-05, "loss": 1.0162, "step": 1215 }, { "epoch": 3.10997442455243, "grad_norm": 0.1988099822661825, "learning_rate": 3.0067628399124523e-05, "loss": 0.9678, "step": 1216 }, { "epoch": 3.112531969309463, "grad_norm": 0.19802134385759912, "learning_rate": 2.999844130981108e-05, "loss": 1.0046, "step": 1217 }, { "epoch": 3.1150895140664963, "grad_norm": 0.21886668903308013, "learning_rate": 2.9929286123799906e-05, "loss": 0.9786, "step": 1218 }, { "epoch": 3.1176470588235294, "grad_norm": 0.20270734092286255, "learning_rate": 2.986016306168453e-05, "loss": 0.9952, "step": 1219 }, { "epoch": 3.1202046035805626, "grad_norm": 0.19771293400631856, "learning_rate": 2.9791072343955946e-05, "loss": 0.9832, "step": 1220 }, { "epoch": 3.122762148337596, "grad_norm": 0.1998240112481755, "learning_rate": 2.972201419100202e-05, "loss": 0.9803, "step": 1221 }, { "epoch": 3.125319693094629, "grad_norm": 0.21631234902354732, "learning_rate": 2.9652988823106736e-05, "loss": 0.9296, "step": 1222 }, { "epoch": 3.1278772378516626, "grad_norm": 0.2001811610318111, "learning_rate": 2.9583996460449474e-05, "loss": 0.9911, "step": 1223 }, { "epoch": 3.130434782608696, "grad_norm": 0.223830854725341, "learning_rate": 2.9515037323104373e-05, "loss": 0.9603, "step": 1224 }, { "epoch": 3.132992327365729, "grad_norm": 0.18521322384491026, "learning_rate": 2.9446111631039572e-05, "loss": 0.9667, "step": 1225 }, { "epoch": 3.135549872122762, "grad_norm": 0.21059872337735092, "learning_rate": 2.9377219604116484e-05, "loss": 0.9842, "step": 1226 }, { "epoch": 3.1381074168797953, "grad_norm": 0.22516840522850048, "learning_rate": 2.930836146208921e-05, "loss": 0.9838, "step": 1227 }, { "epoch": 3.1406649616368285, "grad_norm": 0.17870913096115748, "learning_rate": 2.92395374246037e-05, "loss": 1.0012, "step": 1228 }, { "epoch": 3.1432225063938617, "grad_norm": 0.22974849107011203, "learning_rate": 2.9170747711197154e-05, "loss": 1.0068, "step": 1229 }, { "epoch": 3.1457800511508953, "grad_norm": 0.18080075081074998, "learning_rate": 2.9101992541297288e-05, "loss": 0.9905, "step": 1230 }, { "epoch": 3.1483375959079285, "grad_norm": 0.22879808464892484, "learning_rate": 2.9033272134221575e-05, "loss": 0.999, "step": 1231 }, { "epoch": 3.1508951406649617, "grad_norm": 0.22112889389313278, "learning_rate": 2.8964586709176674e-05, "loss": 0.9777, "step": 1232 }, { "epoch": 3.153452685421995, "grad_norm": 0.19632604879440582, "learning_rate": 2.88959364852576e-05, "loss": 1.0168, "step": 1233 }, { "epoch": 3.156010230179028, "grad_norm": 0.21685788275652715, "learning_rate": 2.882732168144712e-05, "loss": 0.9994, "step": 1234 }, { "epoch": 3.1585677749360612, "grad_norm": 0.2179611102756312, "learning_rate": 2.8758742516614994e-05, "loss": 0.9893, "step": 1235 }, { "epoch": 3.1611253196930944, "grad_norm": 0.19368842885976295, "learning_rate": 2.8690199209517322e-05, "loss": 0.9934, "step": 1236 }, { "epoch": 3.163682864450128, "grad_norm": 0.19759500447000294, "learning_rate": 2.8621691978795793e-05, "loss": 0.9822, "step": 1237 }, { "epoch": 3.166240409207161, "grad_norm": 0.19295856916803417, "learning_rate": 2.8553221042977022e-05, "loss": 0.9868, "step": 1238 }, { "epoch": 3.1687979539641944, "grad_norm": 0.25414221486911515, "learning_rate": 2.84847866204719e-05, "loss": 0.9566, "step": 1239 }, { "epoch": 3.1713554987212276, "grad_norm": 0.1883908256135399, "learning_rate": 2.8416388929574768e-05, "loss": 0.9772, "step": 1240 }, { "epoch": 3.1739130434782608, "grad_norm": 0.2326630917084391, "learning_rate": 2.8348028188462876e-05, "loss": 0.9663, "step": 1241 }, { "epoch": 3.176470588235294, "grad_norm": 0.201683374235404, "learning_rate": 2.8279704615195554e-05, "loss": 0.9727, "step": 1242 }, { "epoch": 3.1790281329923276, "grad_norm": 0.1971824727000338, "learning_rate": 2.8211418427713584e-05, "loss": 0.9812, "step": 1243 }, { "epoch": 3.1815856777493607, "grad_norm": 0.19508872782415485, "learning_rate": 2.8143169843838522e-05, "loss": 0.9663, "step": 1244 }, { "epoch": 3.184143222506394, "grad_norm": 0.18504489219850168, "learning_rate": 2.807495908127194e-05, "loss": 0.944, "step": 1245 }, { "epoch": 3.186700767263427, "grad_norm": 0.18984378821009806, "learning_rate": 2.8006786357594795e-05, "loss": 0.9938, "step": 1246 }, { "epoch": 3.1892583120204603, "grad_norm": 0.1875579903849718, "learning_rate": 2.7938651890266695e-05, "loss": 1.0008, "step": 1247 }, { "epoch": 3.1918158567774935, "grad_norm": 0.18441764733962582, "learning_rate": 2.7870555896625196e-05, "loss": 0.9826, "step": 1248 }, { "epoch": 3.1943734015345266, "grad_norm": 0.1805205042347647, "learning_rate": 2.7802498593885155e-05, "loss": 0.9715, "step": 1249 }, { "epoch": 3.1969309462915603, "grad_norm": 0.1793075530333337, "learning_rate": 2.7734480199138007e-05, "loss": 0.981, "step": 1250 }, { "epoch": 3.1994884910485935, "grad_norm": 0.21281110414647764, "learning_rate": 2.7666500929351082e-05, "loss": 0.9591, "step": 1251 }, { "epoch": 3.2020460358056266, "grad_norm": 0.1916993153560797, "learning_rate": 2.7598561001366905e-05, "loss": 0.9972, "step": 1252 }, { "epoch": 3.20460358056266, "grad_norm": 0.19622262403295782, "learning_rate": 2.7530660631902467e-05, "loss": 0.9543, "step": 1253 }, { "epoch": 3.207161125319693, "grad_norm": 0.17506135270655868, "learning_rate": 2.7462800037548643e-05, "loss": 0.9787, "step": 1254 }, { "epoch": 3.209718670076726, "grad_norm": 0.19641653510944684, "learning_rate": 2.7394979434769378e-05, "loss": 0.9786, "step": 1255 }, { "epoch": 3.21227621483376, "grad_norm": 0.191620221833112, "learning_rate": 2.7327199039901092e-05, "loss": 0.9703, "step": 1256 }, { "epoch": 3.214833759590793, "grad_norm": 0.18678920887787637, "learning_rate": 2.725945906915191e-05, "loss": 0.9748, "step": 1257 }, { "epoch": 3.217391304347826, "grad_norm": 0.18895520158914417, "learning_rate": 2.7191759738601052e-05, "loss": 0.9708, "step": 1258 }, { "epoch": 3.2199488491048593, "grad_norm": 0.17852521291354706, "learning_rate": 2.712410126419805e-05, "loss": 1.0167, "step": 1259 }, { "epoch": 3.2225063938618925, "grad_norm": 0.1975194067747155, "learning_rate": 2.705648386176215e-05, "loss": 0.9843, "step": 1260 }, { "epoch": 3.2250639386189257, "grad_norm": 0.19210567469178344, "learning_rate": 2.698890774698158e-05, "loss": 0.9679, "step": 1261 }, { "epoch": 3.227621483375959, "grad_norm": 0.17550111816818414, "learning_rate": 2.6921373135412852e-05, "loss": 0.9504, "step": 1262 }, { "epoch": 3.2301790281329925, "grad_norm": 0.19205803836193247, "learning_rate": 2.685388024248013e-05, "loss": 0.9719, "step": 1263 }, { "epoch": 3.2327365728900257, "grad_norm": 0.19464925988695062, "learning_rate": 2.6786429283474454e-05, "loss": 0.9774, "step": 1264 }, { "epoch": 3.235294117647059, "grad_norm": 0.18695245921746287, "learning_rate": 2.67190204735531e-05, "loss": 0.9719, "step": 1265 }, { "epoch": 3.237851662404092, "grad_norm": 0.18689844980516787, "learning_rate": 2.665165402773895e-05, "loss": 0.9793, "step": 1266 }, { "epoch": 3.2404092071611252, "grad_norm": 0.1793689613933117, "learning_rate": 2.6584330160919693e-05, "loss": 0.9725, "step": 1267 }, { "epoch": 3.2429667519181584, "grad_norm": 0.19239691492067834, "learning_rate": 2.6517049087847245e-05, "loss": 0.9351, "step": 1268 }, { "epoch": 3.2455242966751916, "grad_norm": 0.18288592509102963, "learning_rate": 2.6449811023137e-05, "loss": 0.9962, "step": 1269 }, { "epoch": 3.2480818414322252, "grad_norm": 0.231930212219846, "learning_rate": 2.6382616181267125e-05, "loss": 0.973, "step": 1270 }, { "epoch": 3.2506393861892584, "grad_norm": 0.17786106163226348, "learning_rate": 2.6315464776577974e-05, "loss": 0.9895, "step": 1271 }, { "epoch": 3.2531969309462916, "grad_norm": 0.2054873075026308, "learning_rate": 2.6248357023271292e-05, "loss": 0.9731, "step": 1272 }, { "epoch": 3.2557544757033248, "grad_norm": 0.19487399212179374, "learning_rate": 2.6181293135409628e-05, "loss": 0.9727, "step": 1273 }, { "epoch": 3.258312020460358, "grad_norm": 0.18326903320186239, "learning_rate": 2.611427332691558e-05, "loss": 0.9847, "step": 1274 }, { "epoch": 3.260869565217391, "grad_norm": 0.18796016490886536, "learning_rate": 2.6047297811571163e-05, "loss": 0.9929, "step": 1275 }, { "epoch": 3.2634271099744243, "grad_norm": 0.17566753019414313, "learning_rate": 2.5980366803017063e-05, "loss": 0.9486, "step": 1276 }, { "epoch": 3.265984654731458, "grad_norm": 0.20427285085857036, "learning_rate": 2.591348051475202e-05, "loss": 1.0034, "step": 1277 }, { "epoch": 3.268542199488491, "grad_norm": 0.1896070148851372, "learning_rate": 2.5846639160132152e-05, "loss": 0.9725, "step": 1278 }, { "epoch": 3.2710997442455243, "grad_norm": 0.185712139910448, "learning_rate": 2.5779842952370193e-05, "loss": 1.0223, "step": 1279 }, { "epoch": 3.2736572890025575, "grad_norm": 0.2064687121716206, "learning_rate": 2.5713092104534926e-05, "loss": 0.9905, "step": 1280 }, { "epoch": 3.2762148337595907, "grad_norm": 0.18607952395957236, "learning_rate": 2.5646386829550378e-05, "loss": 0.9996, "step": 1281 }, { "epoch": 3.2787723785166243, "grad_norm": 0.20062219284524274, "learning_rate": 2.557972734019524e-05, "loss": 0.9859, "step": 1282 }, { "epoch": 3.2813299232736575, "grad_norm": 0.2154776394176252, "learning_rate": 2.5513113849102168e-05, "loss": 0.9847, "step": 1283 }, { "epoch": 3.2838874680306906, "grad_norm": 0.18368848032395593, "learning_rate": 2.5446546568757074e-05, "loss": 0.9804, "step": 1284 }, { "epoch": 3.286445012787724, "grad_norm": 0.20334568862048236, "learning_rate": 2.5380025711498463e-05, "loss": 0.9838, "step": 1285 }, { "epoch": 3.289002557544757, "grad_norm": 0.18953720629273552, "learning_rate": 2.5313551489516773e-05, "loss": 0.9803, "step": 1286 }, { "epoch": 3.29156010230179, "grad_norm": 0.18747324088248227, "learning_rate": 2.5247124114853656e-05, "loss": 0.972, "step": 1287 }, { "epoch": 3.2941176470588234, "grad_norm": 0.18560605380576226, "learning_rate": 2.518074379940136e-05, "loss": 0.9585, "step": 1288 }, { "epoch": 3.296675191815857, "grad_norm": 0.18595879737381238, "learning_rate": 2.5114410754901992e-05, "loss": 0.9584, "step": 1289 }, { "epoch": 3.29923273657289, "grad_norm": 0.18911753423327268, "learning_rate": 2.5048125192946912e-05, "loss": 0.9567, "step": 1290 }, { "epoch": 3.3017902813299234, "grad_norm": 0.1957348195686964, "learning_rate": 2.4981887324975988e-05, "loss": 0.969, "step": 1291 }, { "epoch": 3.3043478260869565, "grad_norm": 0.19191922140197118, "learning_rate": 2.4915697362276935e-05, "loss": 0.9958, "step": 1292 }, { "epoch": 3.3069053708439897, "grad_norm": 0.17974364713669186, "learning_rate": 2.484955551598471e-05, "loss": 0.9392, "step": 1293 }, { "epoch": 3.309462915601023, "grad_norm": 0.20101316164789793, "learning_rate": 2.4783461997080745e-05, "loss": 0.9697, "step": 1294 }, { "epoch": 3.312020460358056, "grad_norm": 0.1958479453724846, "learning_rate": 2.471741701639234e-05, "loss": 0.9814, "step": 1295 }, { "epoch": 3.3145780051150897, "grad_norm": 0.19206314348892373, "learning_rate": 2.4651420784591942e-05, "loss": 0.9924, "step": 1296 }, { "epoch": 3.317135549872123, "grad_norm": 0.20836822041907688, "learning_rate": 2.4585473512196537e-05, "loss": 0.9493, "step": 1297 }, { "epoch": 3.319693094629156, "grad_norm": 0.18897201432176333, "learning_rate": 2.4519575409566888e-05, "loss": 0.9686, "step": 1298 }, { "epoch": 3.3222506393861893, "grad_norm": 0.24239885455652932, "learning_rate": 2.445372668690694e-05, "loss": 0.9934, "step": 1299 }, { "epoch": 3.3248081841432224, "grad_norm": 0.2116091122884991, "learning_rate": 2.4387927554263136e-05, "loss": 1.0145, "step": 1300 }, { "epoch": 3.3273657289002556, "grad_norm": 0.22172159721754445, "learning_rate": 2.432217822152371e-05, "loss": 0.9585, "step": 1301 }, { "epoch": 3.329923273657289, "grad_norm": 0.20040699370284384, "learning_rate": 2.425647889841808e-05, "loss": 0.9883, "step": 1302 }, { "epoch": 3.3324808184143224, "grad_norm": 0.20978963710409357, "learning_rate": 2.4190829794516093e-05, "loss": 0.9681, "step": 1303 }, { "epoch": 3.3350383631713556, "grad_norm": 0.2099270440664411, "learning_rate": 2.412523111922743e-05, "loss": 0.9839, "step": 1304 }, { "epoch": 3.337595907928389, "grad_norm": 0.19154406605739202, "learning_rate": 2.405968308180093e-05, "loss": 0.9592, "step": 1305 }, { "epoch": 3.340153452685422, "grad_norm": 0.2041595809244005, "learning_rate": 2.399418589132386e-05, "loss": 0.9928, "step": 1306 }, { "epoch": 3.342710997442455, "grad_norm": 0.18133483534377134, "learning_rate": 2.392873975672135e-05, "loss": 0.9465, "step": 1307 }, { "epoch": 3.3452685421994883, "grad_norm": 0.1932151711889084, "learning_rate": 2.386334488675564e-05, "loss": 0.9885, "step": 1308 }, { "epoch": 3.3478260869565215, "grad_norm": 0.16981875008819527, "learning_rate": 2.3798001490025417e-05, "loss": 0.9736, "step": 1309 }, { "epoch": 3.350383631713555, "grad_norm": 0.19797950201310727, "learning_rate": 2.3732709774965238e-05, "loss": 1.0126, "step": 1310 }, { "epoch": 3.3529411764705883, "grad_norm": 0.18722532558513932, "learning_rate": 2.3667469949844744e-05, "loss": 0.9801, "step": 1311 }, { "epoch": 3.3554987212276215, "grad_norm": 0.2168260181647949, "learning_rate": 2.3602282222768117e-05, "loss": 0.9735, "step": 1312 }, { "epoch": 3.3580562659846547, "grad_norm": 0.19677536362501252, "learning_rate": 2.353714680167331e-05, "loss": 0.9759, "step": 1313 }, { "epoch": 3.360613810741688, "grad_norm": 0.19959150858815647, "learning_rate": 2.3472063894331423e-05, "loss": 0.9853, "step": 1314 }, { "epoch": 3.363171355498721, "grad_norm": 0.19210281532956738, "learning_rate": 2.340703370834608e-05, "loss": 0.956, "step": 1315 }, { "epoch": 3.3657289002557547, "grad_norm": 0.1866860991602217, "learning_rate": 2.3342056451152713e-05, "loss": 0.9786, "step": 1316 }, { "epoch": 3.368286445012788, "grad_norm": 0.18937893295508668, "learning_rate": 2.327713233001793e-05, "loss": 0.9896, "step": 1317 }, { "epoch": 3.370843989769821, "grad_norm": 0.1771423252358527, "learning_rate": 2.3212261552038827e-05, "loss": 0.9856, "step": 1318 }, { "epoch": 3.373401534526854, "grad_norm": 0.2235030616821708, "learning_rate": 2.314744432414236e-05, "loss": 0.9846, "step": 1319 }, { "epoch": 3.3759590792838874, "grad_norm": 0.1700881379670313, "learning_rate": 2.3082680853084663e-05, "loss": 0.96, "step": 1320 }, { "epoch": 3.3785166240409206, "grad_norm": 0.2032024302743889, "learning_rate": 2.301797134545039e-05, "loss": 0.9602, "step": 1321 }, { "epoch": 3.381074168797954, "grad_norm": 0.18105668532410607, "learning_rate": 2.2953316007652058e-05, "loss": 0.9701, "step": 1322 }, { "epoch": 3.3836317135549874, "grad_norm": 0.19642096592335476, "learning_rate": 2.2888715045929427e-05, "loss": 0.9633, "step": 1323 }, { "epoch": 3.3861892583120206, "grad_norm": 0.179999924216972, "learning_rate": 2.2824168666348775e-05, "loss": 0.9771, "step": 1324 }, { "epoch": 3.3887468030690537, "grad_norm": 0.18186351034819495, "learning_rate": 2.2759677074802237e-05, "loss": 0.9786, "step": 1325 }, { "epoch": 3.391304347826087, "grad_norm": 0.19979550514976036, "learning_rate": 2.269524047700726e-05, "loss": 0.981, "step": 1326 }, { "epoch": 3.39386189258312, "grad_norm": 0.20152802687590698, "learning_rate": 2.2630859078505824e-05, "loss": 0.9674, "step": 1327 }, { "epoch": 3.3964194373401533, "grad_norm": 0.2070039830486493, "learning_rate": 2.256653308466382e-05, "loss": 0.9556, "step": 1328 }, { "epoch": 3.398976982097187, "grad_norm": 0.2135037631517176, "learning_rate": 2.2502262700670457e-05, "loss": 1.0158, "step": 1329 }, { "epoch": 3.40153452685422, "grad_norm": 0.20419304849123474, "learning_rate": 2.243804813153754e-05, "loss": 1.0003, "step": 1330 }, { "epoch": 3.4040920716112533, "grad_norm": 0.18452079803418459, "learning_rate": 2.2373889582098767e-05, "loss": 0.9725, "step": 1331 }, { "epoch": 3.4066496163682864, "grad_norm": 0.21656282916679845, "learning_rate": 2.2309787257009243e-05, "loss": 0.9907, "step": 1332 }, { "epoch": 3.4092071611253196, "grad_norm": 0.17662188950107724, "learning_rate": 2.224574136074467e-05, "loss": 0.968, "step": 1333 }, { "epoch": 3.411764705882353, "grad_norm": 0.22073774520396994, "learning_rate": 2.218175209760077e-05, "loss": 0.9957, "step": 1334 }, { "epoch": 3.414322250639386, "grad_norm": 0.1952236355606058, "learning_rate": 2.2117819671692596e-05, "loss": 0.9802, "step": 1335 }, { "epoch": 3.4168797953964196, "grad_norm": 0.2123002319054059, "learning_rate": 2.2053944286953925e-05, "loss": 0.9802, "step": 1336 }, { "epoch": 3.419437340153453, "grad_norm": 0.17112900395184183, "learning_rate": 2.199012614713655e-05, "loss": 0.9756, "step": 1337 }, { "epoch": 3.421994884910486, "grad_norm": 0.18105726277890447, "learning_rate": 2.1926365455809684e-05, "loss": 0.9646, "step": 1338 }, { "epoch": 3.424552429667519, "grad_norm": 0.20192420986526344, "learning_rate": 2.1862662416359316e-05, "loss": 0.9712, "step": 1339 }, { "epoch": 3.4271099744245523, "grad_norm": 0.2191925792086017, "learning_rate": 2.1799017231987483e-05, "loss": 1.0229, "step": 1340 }, { "epoch": 3.4296675191815855, "grad_norm": 0.20301795120999852, "learning_rate": 2.1735430105711695e-05, "loss": 1.012, "step": 1341 }, { "epoch": 3.4322250639386187, "grad_norm": 0.23205083833577947, "learning_rate": 2.1671901240364272e-05, "loss": 0.9772, "step": 1342 }, { "epoch": 3.4347826086956523, "grad_norm": 0.22514218245005138, "learning_rate": 2.1608430838591685e-05, "loss": 0.9893, "step": 1343 }, { "epoch": 3.4373401534526855, "grad_norm": 0.20385071246928765, "learning_rate": 2.1545019102853927e-05, "loss": 0.9717, "step": 1344 }, { "epoch": 3.4398976982097187, "grad_norm": 0.22705339818003886, "learning_rate": 2.148166623542382e-05, "loss": 0.9616, "step": 1345 }, { "epoch": 3.442455242966752, "grad_norm": 0.2084596120525907, "learning_rate": 2.1418372438386484e-05, "loss": 1.0011, "step": 1346 }, { "epoch": 3.445012787723785, "grad_norm": 0.23886059421463335, "learning_rate": 2.1355137913638515e-05, "loss": 0.9772, "step": 1347 }, { "epoch": 3.4475703324808182, "grad_norm": 0.18940104107641928, "learning_rate": 2.1291962862887498e-05, "loss": 0.9979, "step": 1348 }, { "epoch": 3.4501278772378514, "grad_norm": 0.24592659749162, "learning_rate": 2.1228847487651312e-05, "loss": 1.0035, "step": 1349 }, { "epoch": 3.452685421994885, "grad_norm": 0.16598425333535824, "learning_rate": 2.116579198925747e-05, "loss": 0.9917, "step": 1350 }, { "epoch": 3.455242966751918, "grad_norm": 0.2205308902985363, "learning_rate": 2.110279656884246e-05, "loss": 0.9865, "step": 1351 }, { "epoch": 3.4578005115089514, "grad_norm": 0.18936884708047907, "learning_rate": 2.103986142735118e-05, "loss": 0.9941, "step": 1352 }, { "epoch": 3.4603580562659846, "grad_norm": 0.2113422925950507, "learning_rate": 2.0976986765536202e-05, "loss": 0.9584, "step": 1353 }, { "epoch": 3.4629156010230178, "grad_norm": 0.1616728561002516, "learning_rate": 2.0914172783957213e-05, "loss": 0.9804, "step": 1354 }, { "epoch": 3.4654731457800514, "grad_norm": 0.16257254424814638, "learning_rate": 2.08514196829803e-05, "loss": 0.9722, "step": 1355 }, { "epoch": 3.4680306905370846, "grad_norm": 0.1834674640457499, "learning_rate": 2.078872766277742e-05, "loss": 0.9788, "step": 1356 }, { "epoch": 3.4705882352941178, "grad_norm": 0.1685618446023564, "learning_rate": 2.072609692332564e-05, "loss": 0.9819, "step": 1357 }, { "epoch": 3.473145780051151, "grad_norm": 0.17275999233879516, "learning_rate": 2.0663527664406518e-05, "loss": 0.9845, "step": 1358 }, { "epoch": 3.475703324808184, "grad_norm": 0.17003829121965705, "learning_rate": 2.0601020085605598e-05, "loss": 1.0034, "step": 1359 }, { "epoch": 3.4782608695652173, "grad_norm": 0.18552163039343017, "learning_rate": 2.053857438631159e-05, "loss": 0.992, "step": 1360 }, { "epoch": 3.4808184143222505, "grad_norm": 0.1656628875654245, "learning_rate": 2.0476190765715858e-05, "loss": 0.9748, "step": 1361 }, { "epoch": 3.483375959079284, "grad_norm": 0.18199082040271558, "learning_rate": 2.041386942281171e-05, "loss": 0.9809, "step": 1362 }, { "epoch": 3.4859335038363173, "grad_norm": 0.16923865119691642, "learning_rate": 2.0351610556393887e-05, "loss": 0.9971, "step": 1363 }, { "epoch": 3.4884910485933505, "grad_norm": 0.1628005115603011, "learning_rate": 2.0289414365057716e-05, "loss": 0.9669, "step": 1364 }, { "epoch": 3.4910485933503836, "grad_norm": 0.16855632039381577, "learning_rate": 2.022728104719867e-05, "loss": 1.004, "step": 1365 }, { "epoch": 3.493606138107417, "grad_norm": 0.16112536512929426, "learning_rate": 2.0165210801011684e-05, "loss": 1.0225, "step": 1366 }, { "epoch": 3.49616368286445, "grad_norm": 0.17493102634964577, "learning_rate": 2.010320382449048e-05, "loss": 1.0037, "step": 1367 }, { "epoch": 3.498721227621483, "grad_norm": 0.16471041553434848, "learning_rate": 2.0041260315426933e-05, "loss": 0.9444, "step": 1368 }, { "epoch": 3.501278772378517, "grad_norm": 0.16903247119557196, "learning_rate": 1.997938047141052e-05, "loss": 0.9926, "step": 1369 }, { "epoch": 3.50383631713555, "grad_norm": 0.16181490376970553, "learning_rate": 1.9917564489827595e-05, "loss": 0.9941, "step": 1370 }, { "epoch": 3.506393861892583, "grad_norm": 0.16844937271379926, "learning_rate": 1.9855812567860832e-05, "loss": 0.9904, "step": 1371 }, { "epoch": 3.5089514066496164, "grad_norm": 0.1955609613101622, "learning_rate": 1.979412490248852e-05, "loss": 1.0092, "step": 1372 }, { "epoch": 3.5115089514066495, "grad_norm": 0.17584894207107374, "learning_rate": 1.9732501690484054e-05, "loss": 0.9686, "step": 1373 }, { "epoch": 3.5140664961636827, "grad_norm": 0.1709859307794947, "learning_rate": 1.9670943128415185e-05, "loss": 0.9905, "step": 1374 }, { "epoch": 3.516624040920716, "grad_norm": 0.18078474312646597, "learning_rate": 1.9609449412643403e-05, "loss": 0.9622, "step": 1375 }, { "epoch": 3.5191815856777495, "grad_norm": 0.1939597215141259, "learning_rate": 1.9548020739323443e-05, "loss": 0.9684, "step": 1376 }, { "epoch": 3.5217391304347827, "grad_norm": 0.16884122381730668, "learning_rate": 1.94866573044025e-05, "loss": 0.9961, "step": 1377 }, { "epoch": 3.524296675191816, "grad_norm": 0.17059684327689034, "learning_rate": 1.9425359303619685e-05, "loss": 0.9624, "step": 1378 }, { "epoch": 3.526854219948849, "grad_norm": 0.17751695012672875, "learning_rate": 1.9364126932505383e-05, "loss": 1.0249, "step": 1379 }, { "epoch": 3.5294117647058822, "grad_norm": 0.15312460755115656, "learning_rate": 1.9302960386380637e-05, "loss": 0.9665, "step": 1380 }, { "epoch": 3.531969309462916, "grad_norm": 0.18097753824526422, "learning_rate": 1.924185986035652e-05, "loss": 1.021, "step": 1381 }, { "epoch": 3.5345268542199486, "grad_norm": 0.16847855508114354, "learning_rate": 1.9180825549333483e-05, "loss": 0.9914, "step": 1382 }, { "epoch": 3.5370843989769822, "grad_norm": 0.17393073349900104, "learning_rate": 1.9119857648000827e-05, "loss": 0.9762, "step": 1383 }, { "epoch": 3.5396419437340154, "grad_norm": 0.17339771912019677, "learning_rate": 1.9058956350835952e-05, "loss": 1.0047, "step": 1384 }, { "epoch": 3.5421994884910486, "grad_norm": 0.18250020868636335, "learning_rate": 1.8998121852103843e-05, "loss": 0.9844, "step": 1385 }, { "epoch": 3.544757033248082, "grad_norm": 0.168632483834106, "learning_rate": 1.8937354345856383e-05, "loss": 1.0024, "step": 1386 }, { "epoch": 3.547314578005115, "grad_norm": 0.16967759514899067, "learning_rate": 1.8876654025931767e-05, "loss": 0.9568, "step": 1387 }, { "epoch": 3.5498721227621486, "grad_norm": 0.16759243143730015, "learning_rate": 1.881602108595389e-05, "loss": 0.9762, "step": 1388 }, { "epoch": 3.5524296675191813, "grad_norm": 0.15832481897103784, "learning_rate": 1.875545571933169e-05, "loss": 1.0031, "step": 1389 }, { "epoch": 3.554987212276215, "grad_norm": 0.1811084645197821, "learning_rate": 1.8694958119258612e-05, "loss": 1.0111, "step": 1390 }, { "epoch": 3.557544757033248, "grad_norm": 0.1648706646392026, "learning_rate": 1.863452847871186e-05, "loss": 0.948, "step": 1391 }, { "epoch": 3.5601023017902813, "grad_norm": 0.16239454243103976, "learning_rate": 1.8574166990451898e-05, "loss": 0.9727, "step": 1392 }, { "epoch": 3.5626598465473145, "grad_norm": 0.17942259610790612, "learning_rate": 1.8513873847021823e-05, "loss": 0.9908, "step": 1393 }, { "epoch": 3.5652173913043477, "grad_norm": 0.17137935333375814, "learning_rate": 1.8453649240746675e-05, "loss": 0.9932, "step": 1394 }, { "epoch": 3.5677749360613813, "grad_norm": 0.17038531828743989, "learning_rate": 1.8393493363732905e-05, "loss": 0.9989, "step": 1395 }, { "epoch": 3.5703324808184145, "grad_norm": 0.16665000610754513, "learning_rate": 1.8333406407867712e-05, "loss": 0.9777, "step": 1396 }, { "epoch": 3.5728900255754477, "grad_norm": 0.1578795644839414, "learning_rate": 1.8273388564818443e-05, "loss": 0.9587, "step": 1397 }, { "epoch": 3.575447570332481, "grad_norm": 0.17474995545729136, "learning_rate": 1.8213440026032013e-05, "loss": 0.9979, "step": 1398 }, { "epoch": 3.578005115089514, "grad_norm": 0.15791533543568376, "learning_rate": 1.815356098273422e-05, "loss": 0.9993, "step": 1399 }, { "epoch": 3.580562659846547, "grad_norm": 0.1521732452347076, "learning_rate": 1.8093751625929252e-05, "loss": 0.9605, "step": 1400 }, { "epoch": 3.5831202046035804, "grad_norm": 0.1592417054700969, "learning_rate": 1.8034012146398972e-05, "loss": 0.9697, "step": 1401 }, { "epoch": 3.585677749360614, "grad_norm": 0.15829811384256975, "learning_rate": 1.79743427347023e-05, "loss": 0.9891, "step": 1402 }, { "epoch": 3.588235294117647, "grad_norm": 0.16118804739509424, "learning_rate": 1.791474358117474e-05, "loss": 0.9841, "step": 1403 }, { "epoch": 3.5907928388746804, "grad_norm": 0.16191134259771192, "learning_rate": 1.785521487592762e-05, "loss": 0.9899, "step": 1404 }, { "epoch": 3.5933503836317136, "grad_norm": 0.1526578683810162, "learning_rate": 1.7795756808847583e-05, "loss": 0.9807, "step": 1405 }, { "epoch": 3.5959079283887467, "grad_norm": 0.1700342189970984, "learning_rate": 1.773636956959591e-05, "loss": 0.9837, "step": 1406 }, { "epoch": 3.59846547314578, "grad_norm": 0.17079530240041818, "learning_rate": 1.7677053347608026e-05, "loss": 0.9929, "step": 1407 }, { "epoch": 3.601023017902813, "grad_norm": 0.16946592118696485, "learning_rate": 1.7617808332092733e-05, "loss": 0.9937, "step": 1408 }, { "epoch": 3.6035805626598467, "grad_norm": 0.16721155279112196, "learning_rate": 1.7558634712031733e-05, "loss": 0.9716, "step": 1409 }, { "epoch": 3.60613810741688, "grad_norm": 0.15812399028035554, "learning_rate": 1.7499532676179027e-05, "loss": 0.9463, "step": 1410 }, { "epoch": 3.608695652173913, "grad_norm": 0.18933737997751232, "learning_rate": 1.744050241306023e-05, "loss": 0.9873, "step": 1411 }, { "epoch": 3.6112531969309463, "grad_norm": 0.1513273347708644, "learning_rate": 1.738154411097202e-05, "loss": 1.0049, "step": 1412 }, { "epoch": 3.6138107416879794, "grad_norm": 0.16160857563728534, "learning_rate": 1.7322657957981543e-05, "loss": 0.9629, "step": 1413 }, { "epoch": 3.6163682864450126, "grad_norm": 0.1575626021393853, "learning_rate": 1.726384414192578e-05, "loss": 0.9817, "step": 1414 }, { "epoch": 3.618925831202046, "grad_norm": 0.14742706913576417, "learning_rate": 1.7205102850411e-05, "loss": 0.9608, "step": 1415 }, { "epoch": 3.6214833759590794, "grad_norm": 0.15385433784590577, "learning_rate": 1.71464342708121e-05, "loss": 0.9834, "step": 1416 }, { "epoch": 3.6240409207161126, "grad_norm": 0.15976764455036044, "learning_rate": 1.7087838590272067e-05, "loss": 0.9599, "step": 1417 }, { "epoch": 3.626598465473146, "grad_norm": 0.17485245098862065, "learning_rate": 1.7029315995701355e-05, "loss": 0.9952, "step": 1418 }, { "epoch": 3.629156010230179, "grad_norm": 0.16888749566043196, "learning_rate": 1.6970866673777217e-05, "loss": 0.9879, "step": 1419 }, { "epoch": 3.631713554987212, "grad_norm": 0.17530855499982936, "learning_rate": 1.6912490810943273e-05, "loss": 0.9886, "step": 1420 }, { "epoch": 3.634271099744246, "grad_norm": 0.17867965330900965, "learning_rate": 1.685418859340877e-05, "loss": 1.004, "step": 1421 }, { "epoch": 3.6368286445012785, "grad_norm": 0.16102945276705097, "learning_rate": 1.6795960207148047e-05, "loss": 0.9585, "step": 1422 }, { "epoch": 3.639386189258312, "grad_norm": 0.1740372821884824, "learning_rate": 1.673780583789993e-05, "loss": 0.9864, "step": 1423 }, { "epoch": 3.6419437340153453, "grad_norm": 0.1599072680335884, "learning_rate": 1.6679725671167153e-05, "loss": 0.9778, "step": 1424 }, { "epoch": 3.6445012787723785, "grad_norm": 0.17915224537175092, "learning_rate": 1.662171989221574e-05, "loss": 1.0061, "step": 1425 }, { "epoch": 3.6470588235294117, "grad_norm": 0.15032289661695453, "learning_rate": 1.6563788686074437e-05, "loss": 0.9656, "step": 1426 }, { "epoch": 3.649616368286445, "grad_norm": 0.16252679935892114, "learning_rate": 1.6505932237534135e-05, "loss": 0.9766, "step": 1427 }, { "epoch": 3.6521739130434785, "grad_norm": 0.16372638209261026, "learning_rate": 1.6448150731147245e-05, "loss": 0.971, "step": 1428 }, { "epoch": 3.6547314578005117, "grad_norm": 0.15286966641481936, "learning_rate": 1.6390444351227116e-05, "loss": 0.9615, "step": 1429 }, { "epoch": 3.657289002557545, "grad_norm": 0.17422830046285157, "learning_rate": 1.6332813281847468e-05, "loss": 0.9918, "step": 1430 }, { "epoch": 3.659846547314578, "grad_norm": 0.1536114632035703, "learning_rate": 1.6275257706841787e-05, "loss": 0.9853, "step": 1431 }, { "epoch": 3.662404092071611, "grad_norm": 0.15498834822619906, "learning_rate": 1.6217777809802752e-05, "loss": 0.9889, "step": 1432 }, { "epoch": 3.6649616368286444, "grad_norm": 0.1525756152661097, "learning_rate": 1.6160373774081616e-05, "loss": 0.9645, "step": 1433 }, { "epoch": 3.6675191815856776, "grad_norm": 0.163535134810249, "learning_rate": 1.610304578278772e-05, "loss": 0.9704, "step": 1434 }, { "epoch": 3.670076726342711, "grad_norm": 0.1439706547451718, "learning_rate": 1.604579401878773e-05, "loss": 0.9661, "step": 1435 }, { "epoch": 3.6726342710997444, "grad_norm": 0.18467413525193094, "learning_rate": 1.5988618664705223e-05, "loss": 0.9958, "step": 1436 }, { "epoch": 3.6751918158567776, "grad_norm": 0.14597437711546657, "learning_rate": 1.593151990292006e-05, "loss": 0.9723, "step": 1437 }, { "epoch": 3.6777493606138107, "grad_norm": 0.19323217406252066, "learning_rate": 1.587449791556774e-05, "loss": 1.0298, "step": 1438 }, { "epoch": 3.680306905370844, "grad_norm": 0.15130412791127118, "learning_rate": 1.5817552884538883e-05, "loss": 0.9652, "step": 1439 }, { "epoch": 3.682864450127877, "grad_norm": 0.18744157138750245, "learning_rate": 1.5760684991478634e-05, "loss": 0.9658, "step": 1440 }, { "epoch": 3.6854219948849103, "grad_norm": 0.19424967162069548, "learning_rate": 1.5703894417786066e-05, "loss": 0.9659, "step": 1441 }, { "epoch": 3.687979539641944, "grad_norm": 0.16583222812083506, "learning_rate": 1.564718134461364e-05, "loss": 0.9634, "step": 1442 }, { "epoch": 3.690537084398977, "grad_norm": 0.20374960351527888, "learning_rate": 1.5590545952866565e-05, "loss": 0.9473, "step": 1443 }, { "epoch": 3.6930946291560103, "grad_norm": 0.16010311366530502, "learning_rate": 1.553398842320232e-05, "loss": 0.974, "step": 1444 }, { "epoch": 3.6956521739130435, "grad_norm": 0.18480335391304883, "learning_rate": 1.5477508936029978e-05, "loss": 0.9945, "step": 1445 }, { "epoch": 3.6982097186700766, "grad_norm": 0.1640514417582284, "learning_rate": 1.542110767150963e-05, "loss": 0.9586, "step": 1446 }, { "epoch": 3.70076726342711, "grad_norm": 0.15448350271744538, "learning_rate": 1.5364784809551938e-05, "loss": 0.9853, "step": 1447 }, { "epoch": 3.703324808184143, "grad_norm": 0.17847282007082363, "learning_rate": 1.5308540529817405e-05, "loss": 1.0062, "step": 1448 }, { "epoch": 3.7058823529411766, "grad_norm": 0.15577197474274293, "learning_rate": 1.5252375011715894e-05, "loss": 0.9602, "step": 1449 }, { "epoch": 3.70843989769821, "grad_norm": 0.18403129017994174, "learning_rate": 1.5196288434406045e-05, "loss": 0.9729, "step": 1450 }, { "epoch": 3.710997442455243, "grad_norm": 0.16034660354421298, "learning_rate": 1.514028097679468e-05, "loss": 0.9534, "step": 1451 }, { "epoch": 3.713554987212276, "grad_norm": 0.14891427160648704, "learning_rate": 1.5084352817536205e-05, "loss": 0.9829, "step": 1452 }, { "epoch": 3.7161125319693094, "grad_norm": 0.1660110312736295, "learning_rate": 1.5028504135032113e-05, "loss": 0.9666, "step": 1453 }, { "epoch": 3.718670076726343, "grad_norm": 0.14924967707687345, "learning_rate": 1.4972735107430407e-05, "loss": 1.0091, "step": 1454 }, { "epoch": 3.7212276214833757, "grad_norm": 0.17536986017606493, "learning_rate": 1.4917045912624954e-05, "loss": 0.9656, "step": 1455 }, { "epoch": 3.7237851662404093, "grad_norm": 0.1545050307942204, "learning_rate": 1.4861436728254992e-05, "loss": 0.9626, "step": 1456 }, { "epoch": 3.7263427109974425, "grad_norm": 0.16567326268668936, "learning_rate": 1.480590773170453e-05, "loss": 1.0058, "step": 1457 }, { "epoch": 3.7289002557544757, "grad_norm": 0.159986133602208, "learning_rate": 1.47504591001018e-05, "loss": 0.965, "step": 1458 }, { "epoch": 3.731457800511509, "grad_norm": 0.1636516639657428, "learning_rate": 1.469509101031866e-05, "loss": 0.9992, "step": 1459 }, { "epoch": 3.734015345268542, "grad_norm": 0.15877712329888719, "learning_rate": 1.4639803638970098e-05, "loss": 0.9818, "step": 1460 }, { "epoch": 3.7365728900255757, "grad_norm": 0.15177768229747693, "learning_rate": 1.4584597162413592e-05, "loss": 0.9596, "step": 1461 }, { "epoch": 3.7391304347826084, "grad_norm": 0.16539728281480964, "learning_rate": 1.4529471756748579e-05, "loss": 0.9654, "step": 1462 }, { "epoch": 3.741687979539642, "grad_norm": 0.15013106984230754, "learning_rate": 1.4474427597815898e-05, "loss": 0.9749, "step": 1463 }, { "epoch": 3.7442455242966752, "grad_norm": 0.1607063503736631, "learning_rate": 1.4419464861197225e-05, "loss": 0.9936, "step": 1464 }, { "epoch": 3.7468030690537084, "grad_norm": 0.15822776373381015, "learning_rate": 1.4364583722214521e-05, "loss": 0.9816, "step": 1465 }, { "epoch": 3.7493606138107416, "grad_norm": 0.16495768276016773, "learning_rate": 1.4309784355929429e-05, "loss": 0.9784, "step": 1466 }, { "epoch": 3.7519181585677748, "grad_norm": 0.16464946701490385, "learning_rate": 1.4255066937142816e-05, "loss": 0.9801, "step": 1467 }, { "epoch": 3.7544757033248084, "grad_norm": 0.16145775604523005, "learning_rate": 1.4200431640394103e-05, "loss": 0.9749, "step": 1468 }, { "epoch": 3.7570332480818416, "grad_norm": 0.16615520065079115, "learning_rate": 1.4145878639960735e-05, "loss": 0.9978, "step": 1469 }, { "epoch": 3.7595907928388748, "grad_norm": 0.13971867189011447, "learning_rate": 1.4091408109857706e-05, "loss": 0.9557, "step": 1470 }, { "epoch": 3.762148337595908, "grad_norm": 0.14959140214161223, "learning_rate": 1.4037020223836897e-05, "loss": 0.9881, "step": 1471 }, { "epoch": 3.764705882352941, "grad_norm": 0.1682587608263464, "learning_rate": 1.3982715155386592e-05, "loss": 0.982, "step": 1472 }, { "epoch": 3.7672634271099743, "grad_norm": 0.15458557583904273, "learning_rate": 1.3928493077730889e-05, "loss": 0.9652, "step": 1473 }, { "epoch": 3.7698209718670075, "grad_norm": 0.1788403999805433, "learning_rate": 1.3874354163829158e-05, "loss": 0.9742, "step": 1474 }, { "epoch": 3.772378516624041, "grad_norm": 0.15175695136886533, "learning_rate": 1.3820298586375502e-05, "loss": 0.9873, "step": 1475 }, { "epoch": 3.7749360613810743, "grad_norm": 0.16076876387469338, "learning_rate": 1.3766326517798168e-05, "loss": 0.9799, "step": 1476 }, { "epoch": 3.7774936061381075, "grad_norm": 0.13980858682129033, "learning_rate": 1.3712438130259073e-05, "loss": 0.9489, "step": 1477 }, { "epoch": 3.7800511508951407, "grad_norm": 0.15469469304691397, "learning_rate": 1.3658633595653159e-05, "loss": 0.9741, "step": 1478 }, { "epoch": 3.782608695652174, "grad_norm": 0.16083095141615023, "learning_rate": 1.3604913085607914e-05, "loss": 0.959, "step": 1479 }, { "epoch": 3.785166240409207, "grad_norm": 0.1527879996311441, "learning_rate": 1.3551276771482788e-05, "loss": 0.9692, "step": 1480 }, { "epoch": 3.78772378516624, "grad_norm": 0.1501151428939216, "learning_rate": 1.3497724824368672e-05, "loss": 0.9696, "step": 1481 }, { "epoch": 3.790281329923274, "grad_norm": 0.15367672935484297, "learning_rate": 1.3444257415087339e-05, "loss": 0.9879, "step": 1482 }, { "epoch": 3.792838874680307, "grad_norm": 0.16397124262551233, "learning_rate": 1.3390874714190876e-05, "loss": 1.0394, "step": 1483 }, { "epoch": 3.79539641943734, "grad_norm": 0.1590628803184472, "learning_rate": 1.3337576891961228e-05, "loss": 0.9854, "step": 1484 }, { "epoch": 3.7979539641943734, "grad_norm": 0.18225039937660117, "learning_rate": 1.3284364118409512e-05, "loss": 1.0181, "step": 1485 }, { "epoch": 3.8005115089514065, "grad_norm": 0.15539862711036953, "learning_rate": 1.3231236563275593e-05, "loss": 1.0083, "step": 1486 }, { "epoch": 3.80306905370844, "grad_norm": 0.1702708590595948, "learning_rate": 1.317819439602752e-05, "loss": 0.993, "step": 1487 }, { "epoch": 3.805626598465473, "grad_norm": 0.17494004955408202, "learning_rate": 1.3125237785860945e-05, "loss": 0.9749, "step": 1488 }, { "epoch": 3.8081841432225065, "grad_norm": 0.1525516290905405, "learning_rate": 1.3072366901698606e-05, "loss": 0.9819, "step": 1489 }, { "epoch": 3.8107416879795397, "grad_norm": 0.16157170912735652, "learning_rate": 1.3019581912189794e-05, "loss": 0.9852, "step": 1490 }, { "epoch": 3.813299232736573, "grad_norm": 0.17274917566336145, "learning_rate": 1.2966882985709824e-05, "loss": 0.9839, "step": 1491 }, { "epoch": 3.815856777493606, "grad_norm": 0.17196613842795713, "learning_rate": 1.291427029035946e-05, "loss": 0.9797, "step": 1492 }, { "epoch": 3.8184143222506393, "grad_norm": 0.17260762571717933, "learning_rate": 1.2861743993964409e-05, "loss": 0.9868, "step": 1493 }, { "epoch": 3.820971867007673, "grad_norm": 0.17169441039113015, "learning_rate": 1.2809304264074807e-05, "loss": 0.9641, "step": 1494 }, { "epoch": 3.8235294117647056, "grad_norm": 0.15100733833702012, "learning_rate": 1.2756951267964644e-05, "loss": 1.0115, "step": 1495 }, { "epoch": 3.8260869565217392, "grad_norm": 0.16076257621863999, "learning_rate": 1.2704685172631193e-05, "loss": 0.9817, "step": 1496 }, { "epoch": 3.8286445012787724, "grad_norm": 0.1574213450598427, "learning_rate": 1.2652506144794612e-05, "loss": 0.9941, "step": 1497 }, { "epoch": 3.8312020460358056, "grad_norm": 0.14534492595767098, "learning_rate": 1.2600414350897272e-05, "loss": 0.9637, "step": 1498 }, { "epoch": 3.833759590792839, "grad_norm": 0.14698706161648437, "learning_rate": 1.2548409957103296e-05, "loss": 0.9718, "step": 1499 }, { "epoch": 3.836317135549872, "grad_norm": 0.15605018007996246, "learning_rate": 1.2496493129298011e-05, "loss": 0.9997, "step": 1500 }, { "epoch": 3.8388746803069056, "grad_norm": 0.1477242511131571, "learning_rate": 1.2444664033087466e-05, "loss": 0.9797, "step": 1501 }, { "epoch": 3.8414322250639388, "grad_norm": 0.13847001744552098, "learning_rate": 1.2392922833797783e-05, "loss": 0.9881, "step": 1502 }, { "epoch": 3.843989769820972, "grad_norm": 0.14462911397711195, "learning_rate": 1.2341269696474747e-05, "loss": 0.9535, "step": 1503 }, { "epoch": 3.846547314578005, "grad_norm": 0.15004821476381025, "learning_rate": 1.2289704785883272e-05, "loss": 0.9979, "step": 1504 }, { "epoch": 3.8491048593350383, "grad_norm": 0.14681550872608987, "learning_rate": 1.2238228266506798e-05, "loss": 0.9988, "step": 1505 }, { "epoch": 3.8516624040920715, "grad_norm": 0.13903328529411774, "learning_rate": 1.2186840302546826e-05, "loss": 0.9756, "step": 1506 }, { "epoch": 3.8542199488491047, "grad_norm": 0.1450706267912665, "learning_rate": 1.213554105792238e-05, "loss": 0.9945, "step": 1507 }, { "epoch": 3.8567774936061383, "grad_norm": 0.14188979334960153, "learning_rate": 1.2084330696269491e-05, "loss": 0.9766, "step": 1508 }, { "epoch": 3.8593350383631715, "grad_norm": 0.1504358702702587, "learning_rate": 1.2033209380940654e-05, "loss": 1.004, "step": 1509 }, { "epoch": 3.8618925831202047, "grad_norm": 0.14446185909795273, "learning_rate": 1.198217727500432e-05, "loss": 0.982, "step": 1510 }, { "epoch": 3.864450127877238, "grad_norm": 0.14918074072958132, "learning_rate": 1.1931234541244416e-05, "loss": 0.9572, "step": 1511 }, { "epoch": 3.867007672634271, "grad_norm": 0.1476126427894347, "learning_rate": 1.1880381342159754e-05, "loss": 0.9429, "step": 1512 }, { "epoch": 3.869565217391304, "grad_norm": 0.14658231627960866, "learning_rate": 1.1829617839963508e-05, "loss": 0.9706, "step": 1513 }, { "epoch": 3.8721227621483374, "grad_norm": 0.1463434207088499, "learning_rate": 1.1778944196582814e-05, "loss": 0.9799, "step": 1514 }, { "epoch": 3.874680306905371, "grad_norm": 0.14270360583380973, "learning_rate": 1.172836057365812e-05, "loss": 0.9826, "step": 1515 }, { "epoch": 3.877237851662404, "grad_norm": 0.15081164075513045, "learning_rate": 1.1677867132542735e-05, "loss": 0.993, "step": 1516 }, { "epoch": 3.8797953964194374, "grad_norm": 0.14549864989116654, "learning_rate": 1.162746403430231e-05, "loss": 1.0152, "step": 1517 }, { "epoch": 3.8823529411764706, "grad_norm": 0.1436310825646862, "learning_rate": 1.157715143971431e-05, "loss": 0.9881, "step": 1518 }, { "epoch": 3.8849104859335037, "grad_norm": 0.14506305658975452, "learning_rate": 1.1526929509267513e-05, "loss": 1.0059, "step": 1519 }, { "epoch": 3.887468030690537, "grad_norm": 0.13957687924429218, "learning_rate": 1.1476798403161475e-05, "loss": 0.9961, "step": 1520 }, { "epoch": 3.89002557544757, "grad_norm": 0.15982963311960688, "learning_rate": 1.1426758281306087e-05, "loss": 0.9726, "step": 1521 }, { "epoch": 3.8925831202046037, "grad_norm": 0.14440211669907038, "learning_rate": 1.1376809303320964e-05, "loss": 0.943, "step": 1522 }, { "epoch": 3.895140664961637, "grad_norm": 0.15463895891888774, "learning_rate": 1.1326951628535006e-05, "loss": 0.9789, "step": 1523 }, { "epoch": 3.89769820971867, "grad_norm": 0.14858430492843056, "learning_rate": 1.1277185415985871e-05, "loss": 0.9748, "step": 1524 }, { "epoch": 3.9002557544757033, "grad_norm": 0.15150163420283375, "learning_rate": 1.1227510824419468e-05, "loss": 0.9964, "step": 1525 }, { "epoch": 3.9028132992327365, "grad_norm": 0.16315553122177956, "learning_rate": 1.1177928012289455e-05, "loss": 0.9832, "step": 1526 }, { "epoch": 3.90537084398977, "grad_norm": 0.13639222004794507, "learning_rate": 1.1128437137756705e-05, "loss": 0.9744, "step": 1527 }, { "epoch": 3.907928388746803, "grad_norm": 0.15405033084464928, "learning_rate": 1.1079038358688882e-05, "loss": 0.9882, "step": 1528 }, { "epoch": 3.9104859335038364, "grad_norm": 0.1691974950328252, "learning_rate": 1.1029731832659811e-05, "loss": 0.9538, "step": 1529 }, { "epoch": 3.9130434782608696, "grad_norm": 0.1378352496795574, "learning_rate": 1.0980517716949075e-05, "loss": 0.9822, "step": 1530 }, { "epoch": 3.915601023017903, "grad_norm": 0.14209307859417014, "learning_rate": 1.0931396168541509e-05, "loss": 0.9734, "step": 1531 }, { "epoch": 3.918158567774936, "grad_norm": 0.160326538775092, "learning_rate": 1.088236734412664e-05, "loss": 0.9877, "step": 1532 }, { "epoch": 3.920716112531969, "grad_norm": 0.14428090720467784, "learning_rate": 1.0833431400098227e-05, "loss": 0.9846, "step": 1533 }, { "epoch": 3.923273657289003, "grad_norm": 0.1465209742252091, "learning_rate": 1.0784588492553768e-05, "loss": 0.9964, "step": 1534 }, { "epoch": 3.9258312020460355, "grad_norm": 0.16030862232460416, "learning_rate": 1.0735838777293979e-05, "loss": 0.9858, "step": 1535 }, { "epoch": 3.928388746803069, "grad_norm": 0.13686581434508327, "learning_rate": 1.0687182409822311e-05, "loss": 0.9716, "step": 1536 }, { "epoch": 3.9309462915601023, "grad_norm": 0.1625327040079931, "learning_rate": 1.0638619545344447e-05, "loss": 0.9699, "step": 1537 }, { "epoch": 3.9335038363171355, "grad_norm": 0.1525199965864014, "learning_rate": 1.0590150338767828e-05, "loss": 0.9877, "step": 1538 }, { "epoch": 3.9360613810741687, "grad_norm": 0.13380670697300262, "learning_rate": 1.054177494470114e-05, "loss": 0.97, "step": 1539 }, { "epoch": 3.938618925831202, "grad_norm": 0.17163164533417571, "learning_rate": 1.0493493517453777e-05, "loss": 0.98, "step": 1540 }, { "epoch": 3.9411764705882355, "grad_norm": 0.14388648975792426, "learning_rate": 1.044530621103546e-05, "loss": 0.9714, "step": 1541 }, { "epoch": 3.9437340153452687, "grad_norm": 0.15724845081283767, "learning_rate": 1.0397213179155647e-05, "loss": 0.9482, "step": 1542 }, { "epoch": 3.946291560102302, "grad_norm": 0.14205243895810854, "learning_rate": 1.034921457522307e-05, "loss": 0.9727, "step": 1543 }, { "epoch": 3.948849104859335, "grad_norm": 0.14157720210833158, "learning_rate": 1.0301310552345258e-05, "loss": 0.9548, "step": 1544 }, { "epoch": 3.9514066496163682, "grad_norm": 0.1457113622126631, "learning_rate": 1.0253501263328075e-05, "loss": 0.9902, "step": 1545 }, { "epoch": 3.9539641943734014, "grad_norm": 0.16390432154458898, "learning_rate": 1.020578686067514e-05, "loss": 0.9627, "step": 1546 }, { "epoch": 3.9565217391304346, "grad_norm": 0.141405745046886, "learning_rate": 1.0158167496587428e-05, "loss": 0.9909, "step": 1547 }, { "epoch": 3.959079283887468, "grad_norm": 0.14806685880079012, "learning_rate": 1.0110643322962779e-05, "loss": 0.9693, "step": 1548 }, { "epoch": 3.9616368286445014, "grad_norm": 0.15773747192555593, "learning_rate": 1.006321449139537e-05, "loss": 1.0323, "step": 1549 }, { "epoch": 3.9641943734015346, "grad_norm": 0.140456219803408, "learning_rate": 1.0015881153175244e-05, "loss": 1.0, "step": 1550 }, { "epoch": 3.9667519181585678, "grad_norm": 0.15067874532052797, "learning_rate": 9.968643459287856e-06, "loss": 0.9892, "step": 1551 }, { "epoch": 3.969309462915601, "grad_norm": 0.14479984414191183, "learning_rate": 9.921501560413555e-06, "loss": 0.9901, "step": 1552 }, { "epoch": 3.971867007672634, "grad_norm": 0.15329653150211797, "learning_rate": 9.874455606927134e-06, "loss": 0.9889, "step": 1553 }, { "epoch": 3.9744245524296673, "grad_norm": 0.15721815567916902, "learning_rate": 9.827505748897308e-06, "loss": 0.9767, "step": 1554 }, { "epoch": 3.976982097186701, "grad_norm": 0.13746284270048956, "learning_rate": 9.780652136086322e-06, "loss": 0.9987, "step": 1555 }, { "epoch": 3.979539641943734, "grad_norm": 0.1445235042786382, "learning_rate": 9.733894917949365e-06, "loss": 1.0099, "step": 1556 }, { "epoch": 3.9820971867007673, "grad_norm": 0.15925854158907532, "learning_rate": 9.687234243634127e-06, "loss": 0.9791, "step": 1557 }, { "epoch": 3.9846547314578005, "grad_norm": 0.1469893002397056, "learning_rate": 9.640670261980403e-06, "loss": 0.9931, "step": 1558 }, { "epoch": 3.9872122762148337, "grad_norm": 0.14758630633936268, "learning_rate": 9.594203121519508e-06, "loss": 0.9732, "step": 1559 }, { "epoch": 3.9897698209718673, "grad_norm": 0.14169937812952016, "learning_rate": 9.547832970473858e-06, "loss": 0.9874, "step": 1560 }, { "epoch": 3.9923273657289, "grad_norm": 0.13845067281699075, "learning_rate": 9.501559956756496e-06, "loss": 0.9869, "step": 1561 }, { "epoch": 3.9948849104859336, "grad_norm": 0.1347356255658209, "learning_rate": 9.45538422797061e-06, "loss": 1.0008, "step": 1562 }, { "epoch": 3.997442455242967, "grad_norm": 0.1446889101806559, "learning_rate": 9.409305931409066e-06, "loss": 0.9778, "step": 1563 }, { "epoch": 4.0, "grad_norm": 0.1456124424171873, "learning_rate": 9.36332521405392e-06, "loss": 0.9869, "step": 1564 }, { "epoch": 4.002557544757034, "grad_norm": 0.1694434470987983, "learning_rate": 9.317442222576015e-06, "loss": 0.9557, "step": 1565 }, { "epoch": 4.005115089514066, "grad_norm": 0.14659146019169372, "learning_rate": 9.27165710333441e-06, "loss": 0.9685, "step": 1566 }, { "epoch": 4.0076726342711, "grad_norm": 0.1469630437189336, "learning_rate": 9.225970002375994e-06, "loss": 0.9439, "step": 1567 }, { "epoch": 4.010230179028133, "grad_norm": 0.17386186743500315, "learning_rate": 9.180381065435e-06, "loss": 0.9612, "step": 1568 }, { "epoch": 4.012787723785166, "grad_norm": 0.15084057811060614, "learning_rate": 9.13489043793251e-06, "loss": 0.9726, "step": 1569 }, { "epoch": 4.015345268542199, "grad_norm": 0.15247338496330756, "learning_rate": 9.089498264976044e-06, "loss": 0.9744, "step": 1570 }, { "epoch": 4.017902813299233, "grad_norm": 0.1489429891129685, "learning_rate": 9.044204691359035e-06, "loss": 0.9713, "step": 1571 }, { "epoch": 4.020460358056266, "grad_norm": 0.1480263481725032, "learning_rate": 8.99900986156046e-06, "loss": 0.9579, "step": 1572 }, { "epoch": 4.023017902813299, "grad_norm": 0.1608202224818316, "learning_rate": 8.953913919744237e-06, "loss": 0.9623, "step": 1573 }, { "epoch": 4.025575447570333, "grad_norm": 0.14602910658372553, "learning_rate": 8.908917009758902e-06, "loss": 0.956, "step": 1574 }, { "epoch": 4.028132992327365, "grad_norm": 0.14751895138775156, "learning_rate": 8.864019275137096e-06, "loss": 0.9611, "step": 1575 }, { "epoch": 4.030690537084399, "grad_norm": 0.13940784210831325, "learning_rate": 8.819220859095096e-06, "loss": 0.9734, "step": 1576 }, { "epoch": 4.033248081841432, "grad_norm": 0.15771825616041407, "learning_rate": 8.774521904532349e-06, "loss": 0.9448, "step": 1577 }, { "epoch": 4.035805626598465, "grad_norm": 0.14987870095750866, "learning_rate": 8.7299225540311e-06, "loss": 0.9735, "step": 1578 }, { "epoch": 4.038363171355499, "grad_norm": 0.13597639144343102, "learning_rate": 8.685422949855785e-06, "loss": 0.9733, "step": 1579 }, { "epoch": 4.040920716112532, "grad_norm": 0.14348384657538388, "learning_rate": 8.64102323395272e-06, "loss": 0.9915, "step": 1580 }, { "epoch": 4.043478260869565, "grad_norm": 0.13911803653848545, "learning_rate": 8.596723547949603e-06, "loss": 0.9659, "step": 1581 }, { "epoch": 4.046035805626598, "grad_norm": 0.1303425224909242, "learning_rate": 8.552524033155025e-06, "loss": 0.952, "step": 1582 }, { "epoch": 4.048593350383632, "grad_norm": 0.13127570779064995, "learning_rate": 8.508424830558075e-06, "loss": 0.9458, "step": 1583 }, { "epoch": 4.051150895140665, "grad_norm": 0.13605211935620362, "learning_rate": 8.464426080827808e-06, "loss": 0.9638, "step": 1584 }, { "epoch": 4.053708439897698, "grad_norm": 0.1353571143519125, "learning_rate": 8.42052792431292e-06, "loss": 0.9592, "step": 1585 }, { "epoch": 4.056265984654732, "grad_norm": 0.13743810323417613, "learning_rate": 8.376730501041197e-06, "loss": 1.0004, "step": 1586 }, { "epoch": 4.0588235294117645, "grad_norm": 0.1335028249124241, "learning_rate": 8.333033950719089e-06, "loss": 0.9268, "step": 1587 }, { "epoch": 4.061381074168798, "grad_norm": 0.14308735627127328, "learning_rate": 8.28943841273132e-06, "loss": 0.9524, "step": 1588 }, { "epoch": 4.063938618925831, "grad_norm": 0.13927706340176157, "learning_rate": 8.24594402614038e-06, "loss": 0.9931, "step": 1589 }, { "epoch": 4.0664961636828645, "grad_norm": 0.13527098384462513, "learning_rate": 8.202550929686061e-06, "loss": 0.9591, "step": 1590 }, { "epoch": 4.069053708439898, "grad_norm": 0.13623666624179986, "learning_rate": 8.15925926178513e-06, "loss": 0.9966, "step": 1591 }, { "epoch": 4.071611253196931, "grad_norm": 0.13999973460956114, "learning_rate": 8.116069160530772e-06, "loss": 0.9862, "step": 1592 }, { "epoch": 4.0741687979539645, "grad_norm": 0.13171292461081588, "learning_rate": 8.072980763692197e-06, "loss": 0.9959, "step": 1593 }, { "epoch": 4.076726342710997, "grad_norm": 0.1367650747219921, "learning_rate": 8.029994208714198e-06, "loss": 0.9983, "step": 1594 }, { "epoch": 4.079283887468031, "grad_norm": 0.13227052084210594, "learning_rate": 7.987109632716703e-06, "loss": 0.9592, "step": 1595 }, { "epoch": 4.081841432225064, "grad_norm": 0.14444449340962193, "learning_rate": 7.94432717249435e-06, "loss": 0.9853, "step": 1596 }, { "epoch": 4.084398976982097, "grad_norm": 0.13614148393534858, "learning_rate": 7.901646964516039e-06, "loss": 0.9715, "step": 1597 }, { "epoch": 4.086956521739131, "grad_norm": 0.12749269393727558, "learning_rate": 7.859069144924528e-06, "loss": 0.9395, "step": 1598 }, { "epoch": 4.089514066496164, "grad_norm": 0.13805044128867539, "learning_rate": 7.816593849535935e-06, "loss": 0.9437, "step": 1599 }, { "epoch": 4.092071611253197, "grad_norm": 0.12754607705203092, "learning_rate": 7.77422121383936e-06, "loss": 0.9752, "step": 1600 }, { "epoch": 4.09462915601023, "grad_norm": 0.1351624415394795, "learning_rate": 7.731951372996445e-06, "loss": 0.9709, "step": 1601 }, { "epoch": 4.0971867007672635, "grad_norm": 0.12306767716486365, "learning_rate": 7.689784461840916e-06, "loss": 0.957, "step": 1602 }, { "epoch": 4.099744245524296, "grad_norm": 0.12976229511348925, "learning_rate": 7.647720614878182e-06, "loss": 0.9433, "step": 1603 }, { "epoch": 4.10230179028133, "grad_norm": 0.13817478561667218, "learning_rate": 7.605759966284885e-06, "loss": 0.9843, "step": 1604 }, { "epoch": 4.1048593350383635, "grad_norm": 0.12832872660123104, "learning_rate": 7.563902649908513e-06, "loss": 0.9695, "step": 1605 }, { "epoch": 4.107416879795396, "grad_norm": 0.1412483260552206, "learning_rate": 7.522148799266888e-06, "loss": 0.9728, "step": 1606 }, { "epoch": 4.10997442455243, "grad_norm": 0.1365069279515938, "learning_rate": 7.480498547547825e-06, "loss": 0.9508, "step": 1607 }, { "epoch": 4.112531969309463, "grad_norm": 0.13711455047388862, "learning_rate": 7.438952027608688e-06, "loss": 0.9579, "step": 1608 }, { "epoch": 4.115089514066496, "grad_norm": 0.12701385871827708, "learning_rate": 7.397509371975929e-06, "loss": 0.9757, "step": 1609 }, { "epoch": 4.117647058823529, "grad_norm": 0.15072117557519477, "learning_rate": 7.356170712844703e-06, "loss": 0.9704, "step": 1610 }, { "epoch": 4.120204603580563, "grad_norm": 0.13660469487293564, "learning_rate": 7.314936182078427e-06, "loss": 0.9625, "step": 1611 }, { "epoch": 4.122762148337596, "grad_norm": 0.1348626532127348, "learning_rate": 7.273805911208374e-06, "loss": 0.9565, "step": 1612 }, { "epoch": 4.125319693094629, "grad_norm": 0.1558370983441284, "learning_rate": 7.23278003143324e-06, "loss": 0.9816, "step": 1613 }, { "epoch": 4.127877237851663, "grad_norm": 0.1442560358532163, "learning_rate": 7.191858673618713e-06, "loss": 0.9693, "step": 1614 }, { "epoch": 4.130434782608695, "grad_norm": 0.12076587398680057, "learning_rate": 7.1510419682971235e-06, "loss": 0.9704, "step": 1615 }, { "epoch": 4.132992327365729, "grad_norm": 0.13613335771644464, "learning_rate": 7.110330045666938e-06, "loss": 0.9686, "step": 1616 }, { "epoch": 4.135549872122763, "grad_norm": 0.15273830211192063, "learning_rate": 7.069723035592368e-06, "loss": 0.9643, "step": 1617 }, { "epoch": 4.138107416879795, "grad_norm": 0.13829841062139175, "learning_rate": 7.02922106760302e-06, "loss": 0.9638, "step": 1618 }, { "epoch": 4.140664961636829, "grad_norm": 0.1312114107225528, "learning_rate": 6.988824270893402e-06, "loss": 0.9392, "step": 1619 }, { "epoch": 4.143222506393862, "grad_norm": 0.1411742619531022, "learning_rate": 6.948532774322552e-06, "loss": 0.9921, "step": 1620 }, { "epoch": 4.145780051150895, "grad_norm": 0.1472736684669091, "learning_rate": 6.908346706413605e-06, "loss": 0.9759, "step": 1621 }, { "epoch": 4.148337595907928, "grad_norm": 0.13211483942081645, "learning_rate": 6.868266195353439e-06, "loss": 0.9725, "step": 1622 }, { "epoch": 4.150895140664962, "grad_norm": 0.1328326493254342, "learning_rate": 6.828291368992159e-06, "loss": 0.9881, "step": 1623 }, { "epoch": 4.153452685421995, "grad_norm": 0.13520210264575472, "learning_rate": 6.7884223548427915e-06, "loss": 0.9523, "step": 1624 }, { "epoch": 4.156010230179028, "grad_norm": 0.12812477279560075, "learning_rate": 6.748659280080852e-06, "loss": 0.9646, "step": 1625 }, { "epoch": 4.158567774936062, "grad_norm": 0.1447636884254579, "learning_rate": 6.709002271543896e-06, "loss": 0.9764, "step": 1626 }, { "epoch": 4.161125319693094, "grad_norm": 0.12397606004342086, "learning_rate": 6.669451455731168e-06, "loss": 0.9688, "step": 1627 }, { "epoch": 4.163682864450128, "grad_norm": 0.12250284994573661, "learning_rate": 6.630006958803159e-06, "loss": 0.9832, "step": 1628 }, { "epoch": 4.166240409207161, "grad_norm": 0.12522111574524233, "learning_rate": 6.590668906581235e-06, "loss": 0.9865, "step": 1629 }, { "epoch": 4.168797953964194, "grad_norm": 0.14426664986318974, "learning_rate": 6.551437424547207e-06, "loss": 0.9745, "step": 1630 }, { "epoch": 4.171355498721228, "grad_norm": 0.1364036728911068, "learning_rate": 6.5123126378429455e-06, "loss": 0.9926, "step": 1631 }, { "epoch": 4.173913043478261, "grad_norm": 0.12516777388845976, "learning_rate": 6.473294671270011e-06, "loss": 0.9718, "step": 1632 }, { "epoch": 4.176470588235294, "grad_norm": 0.1255663623374594, "learning_rate": 6.4343836492891974e-06, "loss": 0.9668, "step": 1633 }, { "epoch": 4.179028132992327, "grad_norm": 0.12758144062263563, "learning_rate": 6.395579696020138e-06, "loss": 0.9631, "step": 1634 }, { "epoch": 4.181585677749361, "grad_norm": 0.13530428494260122, "learning_rate": 6.3568829352409975e-06, "loss": 0.98, "step": 1635 }, { "epoch": 4.1841432225063935, "grad_norm": 0.13016949790316, "learning_rate": 6.3182934903879725e-06, "loss": 1.0123, "step": 1636 }, { "epoch": 4.186700767263427, "grad_norm": 0.14112114746254337, "learning_rate": 6.279811484554943e-06, "loss": 0.9655, "step": 1637 }, { "epoch": 4.189258312020461, "grad_norm": 0.1275629640487005, "learning_rate": 6.241437040493088e-06, "loss": 0.9812, "step": 1638 }, { "epoch": 4.1918158567774935, "grad_norm": 0.14094316110593494, "learning_rate": 6.2031702806104735e-06, "loss": 0.9639, "step": 1639 }, { "epoch": 4.194373401534527, "grad_norm": 0.13295300129747156, "learning_rate": 6.165011326971679e-06, "loss": 0.9572, "step": 1640 }, { "epoch": 4.19693094629156, "grad_norm": 0.12981205516840236, "learning_rate": 6.1269603012973795e-06, "loss": 0.9358, "step": 1641 }, { "epoch": 4.1994884910485935, "grad_norm": 0.12438813099797683, "learning_rate": 6.089017324964017e-06, "loss": 0.9838, "step": 1642 }, { "epoch": 4.202046035805626, "grad_norm": 0.12621995618989404, "learning_rate": 6.051182519003336e-06, "loss": 0.9413, "step": 1643 }, { "epoch": 4.20460358056266, "grad_norm": 0.12388214081758457, "learning_rate": 6.013456004102058e-06, "loss": 0.9833, "step": 1644 }, { "epoch": 4.207161125319693, "grad_norm": 0.13864147929412118, "learning_rate": 5.97583790060146e-06, "loss": 0.9699, "step": 1645 }, { "epoch": 4.209718670076726, "grad_norm": 0.12152527637941755, "learning_rate": 5.938328328497016e-06, "loss": 0.9504, "step": 1646 }, { "epoch": 4.21227621483376, "grad_norm": 0.1228917559817376, "learning_rate": 5.900927407437995e-06, "loss": 0.9521, "step": 1647 }, { "epoch": 4.2148337595907925, "grad_norm": 0.1281261751626848, "learning_rate": 5.8636352567270804e-06, "loss": 0.964, "step": 1648 }, { "epoch": 4.217391304347826, "grad_norm": 0.13928877396090505, "learning_rate": 5.8264519953200375e-06, "loss": 0.9479, "step": 1649 }, { "epoch": 4.21994884910486, "grad_norm": 0.12140483592869106, "learning_rate": 5.7893777418252325e-06, "loss": 0.9512, "step": 1650 }, { "epoch": 4.2225063938618925, "grad_norm": 0.12832906273182243, "learning_rate": 5.75241261450334e-06, "loss": 0.9714, "step": 1651 }, { "epoch": 4.225063938618926, "grad_norm": 0.12325256644456352, "learning_rate": 5.715556731266971e-06, "loss": 0.9621, "step": 1652 }, { "epoch": 4.227621483375959, "grad_norm": 0.1231212443972644, "learning_rate": 5.678810209680219e-06, "loss": 0.9744, "step": 1653 }, { "epoch": 4.2301790281329925, "grad_norm": 0.12843613548840424, "learning_rate": 5.642173166958365e-06, "loss": 0.9681, "step": 1654 }, { "epoch": 4.232736572890025, "grad_norm": 0.12908136661093464, "learning_rate": 5.605645719967445e-06, "loss": 0.9931, "step": 1655 }, { "epoch": 4.235294117647059, "grad_norm": 0.1221393745688676, "learning_rate": 5.5692279852239286e-06, "loss": 0.968, "step": 1656 }, { "epoch": 4.2378516624040925, "grad_norm": 0.1283209209042707, "learning_rate": 5.532920078894299e-06, "loss": 0.9548, "step": 1657 }, { "epoch": 4.240409207161125, "grad_norm": 0.13088853822976476, "learning_rate": 5.496722116794715e-06, "loss": 0.989, "step": 1658 }, { "epoch": 4.242966751918159, "grad_norm": 0.12637460822527927, "learning_rate": 5.4606342143906386e-06, "loss": 0.9722, "step": 1659 }, { "epoch": 4.245524296675192, "grad_norm": 0.12414353807978612, "learning_rate": 5.424656486796465e-06, "loss": 0.976, "step": 1660 }, { "epoch": 4.248081841432225, "grad_norm": 0.1215866811292329, "learning_rate": 5.388789048775103e-06, "loss": 0.9828, "step": 1661 }, { "epoch": 4.250639386189258, "grad_norm": 0.12501721399744686, "learning_rate": 5.353032014737718e-06, "loss": 0.9775, "step": 1662 }, { "epoch": 4.253196930946292, "grad_norm": 0.12240136677562505, "learning_rate": 5.317385498743273e-06, "loss": 0.9585, "step": 1663 }, { "epoch": 4.255754475703325, "grad_norm": 0.12452692032290599, "learning_rate": 5.2818496144982025e-06, "loss": 0.9567, "step": 1664 }, { "epoch": 4.258312020460358, "grad_norm": 0.12715363573233318, "learning_rate": 5.246424475356034e-06, "loss": 0.9615, "step": 1665 }, { "epoch": 4.260869565217392, "grad_norm": 0.12536021121142926, "learning_rate": 5.211110194317073e-06, "loss": 0.9626, "step": 1666 }, { "epoch": 4.263427109974424, "grad_norm": 0.11961524409682861, "learning_rate": 5.17590688402795e-06, "loss": 0.9741, "step": 1667 }, { "epoch": 4.265984654731458, "grad_norm": 0.12506582590821555, "learning_rate": 5.140814656781352e-06, "loss": 0.9226, "step": 1668 }, { "epoch": 4.268542199488491, "grad_norm": 0.14073939782978875, "learning_rate": 5.105833624515639e-06, "loss": 0.9648, "step": 1669 }, { "epoch": 4.271099744245524, "grad_norm": 0.12278625115353788, "learning_rate": 5.070963898814456e-06, "loss": 0.9779, "step": 1670 }, { "epoch": 4.273657289002558, "grad_norm": 0.12666237685804943, "learning_rate": 5.036205590906403e-06, "loss": 0.9539, "step": 1671 }, { "epoch": 4.276214833759591, "grad_norm": 0.13431493290109958, "learning_rate": 5.0015588116646865e-06, "loss": 0.9636, "step": 1672 }, { "epoch": 4.278772378516624, "grad_norm": 0.12811446773007196, "learning_rate": 4.967023671606734e-06, "loss": 0.9201, "step": 1673 }, { "epoch": 4.281329923273657, "grad_norm": 0.12424895319204726, "learning_rate": 4.932600280893889e-06, "loss": 0.9603, "step": 1674 }, { "epoch": 4.283887468030691, "grad_norm": 0.13835712175563045, "learning_rate": 4.8982887493310085e-06, "loss": 0.9763, "step": 1675 }, { "epoch": 4.286445012787723, "grad_norm": 0.1266202682434657, "learning_rate": 4.8640891863661614e-06, "loss": 0.9603, "step": 1676 }, { "epoch": 4.289002557544757, "grad_norm": 0.12820404642448693, "learning_rate": 4.830001701090248e-06, "loss": 0.977, "step": 1677 }, { "epoch": 4.291560102301791, "grad_norm": 0.11736882208459962, "learning_rate": 4.796026402236624e-06, "loss": 0.9769, "step": 1678 }, { "epoch": 4.294117647058823, "grad_norm": 0.12866380807856084, "learning_rate": 4.762163398180852e-06, "loss": 0.9639, "step": 1679 }, { "epoch": 4.296675191815857, "grad_norm": 0.13603267144210404, "learning_rate": 4.728412796940251e-06, "loss": 0.9822, "step": 1680 }, { "epoch": 4.29923273657289, "grad_norm": 0.12397632361985814, "learning_rate": 4.694774706173602e-06, "loss": 0.9545, "step": 1681 }, { "epoch": 4.301790281329923, "grad_norm": 0.13248904474625434, "learning_rate": 4.661249233180787e-06, "loss": 1.0015, "step": 1682 }, { "epoch": 4.304347826086957, "grad_norm": 0.13056497746259793, "learning_rate": 4.627836484902495e-06, "loss": 0.9803, "step": 1683 }, { "epoch": 4.30690537084399, "grad_norm": 0.12577847853997326, "learning_rate": 4.59453656791979e-06, "loss": 1.0028, "step": 1684 }, { "epoch": 4.309462915601023, "grad_norm": 0.11751400707268564, "learning_rate": 4.561349588453849e-06, "loss": 0.9534, "step": 1685 }, { "epoch": 4.312020460358056, "grad_norm": 0.1298510057754835, "learning_rate": 4.528275652365608e-06, "loss": 0.9848, "step": 1686 }, { "epoch": 4.31457800511509, "grad_norm": 0.12778091258561974, "learning_rate": 4.495314865155393e-06, "loss": 0.9469, "step": 1687 }, { "epoch": 4.3171355498721224, "grad_norm": 0.12331926980523211, "learning_rate": 4.462467331962614e-06, "loss": 0.9438, "step": 1688 }, { "epoch": 4.319693094629156, "grad_norm": 0.11808697416694176, "learning_rate": 4.4297331575654165e-06, "loss": 0.9493, "step": 1689 }, { "epoch": 4.322250639386189, "grad_norm": 0.1192941326588608, "learning_rate": 4.397112446380347e-06, "loss": 1.01, "step": 1690 }, { "epoch": 4.324808184143222, "grad_norm": 0.1167465157397439, "learning_rate": 4.3646053024620375e-06, "loss": 0.9614, "step": 1691 }, { "epoch": 4.327365728900256, "grad_norm": 0.12977635130536203, "learning_rate": 4.33221182950283e-06, "loss": 0.9589, "step": 1692 }, { "epoch": 4.329923273657289, "grad_norm": 0.13176862781260584, "learning_rate": 4.299932130832516e-06, "loss": 0.9734, "step": 1693 }, { "epoch": 4.332480818414322, "grad_norm": 0.13475421457283548, "learning_rate": 4.267766309417947e-06, "loss": 0.9647, "step": 1694 }, { "epoch": 4.335038363171355, "grad_norm": 0.11898434149307777, "learning_rate": 4.235714467862697e-06, "loss": 0.9531, "step": 1695 }, { "epoch": 4.337595907928389, "grad_norm": 0.12202881260318174, "learning_rate": 4.2037767084068105e-06, "loss": 0.9931, "step": 1696 }, { "epoch": 4.340153452685422, "grad_norm": 0.1359310771287758, "learning_rate": 4.171953132926403e-06, "loss": 0.9611, "step": 1697 }, { "epoch": 4.342710997442455, "grad_norm": 0.12498482699675276, "learning_rate": 4.140243842933353e-06, "loss": 0.9384, "step": 1698 }, { "epoch": 4.345268542199489, "grad_norm": 0.12177790637791187, "learning_rate": 4.1086489395750326e-06, "loss": 0.9747, "step": 1699 }, { "epoch": 4.3478260869565215, "grad_norm": 0.12407774991295589, "learning_rate": 4.077168523633881e-06, "loss": 0.9708, "step": 1700 }, { "epoch": 4.350383631713555, "grad_norm": 0.12887036467594995, "learning_rate": 4.045802695527177e-06, "loss": 0.9354, "step": 1701 }, { "epoch": 4.352941176470588, "grad_norm": 0.12989274173197723, "learning_rate": 4.014551555306665e-06, "loss": 0.9849, "step": 1702 }, { "epoch": 4.3554987212276215, "grad_norm": 0.12421979367055937, "learning_rate": 3.983415202658281e-06, "loss": 0.9995, "step": 1703 }, { "epoch": 4.358056265984655, "grad_norm": 0.1198682959010047, "learning_rate": 3.952393736901785e-06, "loss": 0.9533, "step": 1704 }, { "epoch": 4.360613810741688, "grad_norm": 0.128521264353093, "learning_rate": 3.921487256990473e-06, "loss": 0.9708, "step": 1705 }, { "epoch": 4.3631713554987215, "grad_norm": 0.11847073942111394, "learning_rate": 3.890695861510856e-06, "loss": 0.9814, "step": 1706 }, { "epoch": 4.365728900255754, "grad_norm": 0.12769887613282743, "learning_rate": 3.8600196486823405e-06, "loss": 0.9717, "step": 1707 }, { "epoch": 4.368286445012788, "grad_norm": 0.12999690542001419, "learning_rate": 3.8294587163569194e-06, "loss": 0.9598, "step": 1708 }, { "epoch": 4.370843989769821, "grad_norm": 0.11857254485023354, "learning_rate": 3.799013162018881e-06, "loss": 0.9837, "step": 1709 }, { "epoch": 4.373401534526854, "grad_norm": 0.12046803154933006, "learning_rate": 3.7686830827844544e-06, "loss": 0.9721, "step": 1710 }, { "epoch": 4.375959079283888, "grad_norm": 0.12149366728466514, "learning_rate": 3.7384685754015126e-06, "loss": 0.9946, "step": 1711 }, { "epoch": 4.378516624040921, "grad_norm": 0.11870216605190843, "learning_rate": 3.7083697362493066e-06, "loss": 0.9938, "step": 1712 }, { "epoch": 4.381074168797954, "grad_norm": 0.11844452341376038, "learning_rate": 3.6783866613381025e-06, "loss": 0.9597, "step": 1713 }, { "epoch": 4.383631713554987, "grad_norm": 0.1243613960126632, "learning_rate": 3.6485194463089022e-06, "loss": 0.9816, "step": 1714 }, { "epoch": 4.3861892583120206, "grad_norm": 0.11603761858752591, "learning_rate": 3.6187681864331235e-06, "loss": 0.9774, "step": 1715 }, { "epoch": 4.388746803069053, "grad_norm": 0.11736635034857253, "learning_rate": 3.5891329766123505e-06, "loss": 0.9663, "step": 1716 }, { "epoch": 4.391304347826087, "grad_norm": 0.11962011769251431, "learning_rate": 3.5596139113779215e-06, "loss": 1.0028, "step": 1717 }, { "epoch": 4.3938618925831205, "grad_norm": 0.114948006498465, "learning_rate": 3.5302110848907336e-06, "loss": 0.9469, "step": 1718 }, { "epoch": 4.396419437340153, "grad_norm": 0.10841359436318306, "learning_rate": 3.500924590940904e-06, "loss": 0.9587, "step": 1719 }, { "epoch": 4.398976982097187, "grad_norm": 0.11627559315730779, "learning_rate": 3.4717545229474524e-06, "loss": 0.9788, "step": 1720 }, { "epoch": 4.40153452685422, "grad_norm": 0.1152200142947816, "learning_rate": 3.442700973958024e-06, "loss": 0.9962, "step": 1721 }, { "epoch": 4.404092071611253, "grad_norm": 0.1154591706617095, "learning_rate": 3.4137640366485924e-06, "loss": 0.9824, "step": 1722 }, { "epoch": 4.406649616368286, "grad_norm": 0.11858603589980796, "learning_rate": 3.384943803323153e-06, "loss": 0.9865, "step": 1723 }, { "epoch": 4.40920716112532, "grad_norm": 0.12077395065465756, "learning_rate": 3.356240365913439e-06, "loss": 0.9551, "step": 1724 }, { "epoch": 4.411764705882353, "grad_norm": 0.11389523914601839, "learning_rate": 3.3276538159786197e-06, "loss": 0.9706, "step": 1725 }, { "epoch": 4.414322250639386, "grad_norm": 0.11263773696411303, "learning_rate": 3.299184244705025e-06, "loss": 0.9697, "step": 1726 }, { "epoch": 4.41687979539642, "grad_norm": 0.12406578721816504, "learning_rate": 3.2708317429058334e-06, "loss": 0.9589, "step": 1727 }, { "epoch": 4.419437340153452, "grad_norm": 0.12286088725943455, "learning_rate": 3.2425964010207765e-06, "loss": 0.9588, "step": 1728 }, { "epoch": 4.421994884910486, "grad_norm": 0.11915716329986083, "learning_rate": 3.214478309115894e-06, "loss": 0.9641, "step": 1729 }, { "epoch": 4.42455242966752, "grad_norm": 0.12652222267615001, "learning_rate": 3.1864775568832027e-06, "loss": 1.0013, "step": 1730 }, { "epoch": 4.427109974424552, "grad_norm": 0.11389511903367709, "learning_rate": 3.1585942336404262e-06, "loss": 0.9535, "step": 1731 }, { "epoch": 4.429667519181586, "grad_norm": 0.12438352231529569, "learning_rate": 3.13082842833071e-06, "loss": 0.9435, "step": 1732 }, { "epoch": 4.432225063938619, "grad_norm": 0.1175440012829412, "learning_rate": 3.1031802295223355e-06, "loss": 0.9757, "step": 1733 }, { "epoch": 4.434782608695652, "grad_norm": 0.1258322406199311, "learning_rate": 3.075649725408445e-06, "loss": 0.9593, "step": 1734 }, { "epoch": 4.437340153452685, "grad_norm": 0.1181435995087277, "learning_rate": 3.048237003806738e-06, "loss": 0.9811, "step": 1735 }, { "epoch": 4.439897698209719, "grad_norm": 0.12072878038936147, "learning_rate": 3.0209421521592318e-06, "loss": 1.0008, "step": 1736 }, { "epoch": 4.442455242966752, "grad_norm": 0.1199404755824827, "learning_rate": 2.9937652575319444e-06, "loss": 0.9636, "step": 1737 }, { "epoch": 4.445012787723785, "grad_norm": 0.11517484111123064, "learning_rate": 2.966706406614628e-06, "loss": 0.9639, "step": 1738 }, { "epoch": 4.447570332480819, "grad_norm": 0.11599655109609232, "learning_rate": 2.939765685720497e-06, "loss": 0.983, "step": 1739 }, { "epoch": 4.450127877237851, "grad_norm": 0.11702394553897594, "learning_rate": 2.912943180785961e-06, "loss": 0.9378, "step": 1740 }, { "epoch": 4.452685421994885, "grad_norm": 0.11662021025404341, "learning_rate": 2.886238977370326e-06, "loss": 0.9564, "step": 1741 }, { "epoch": 4.455242966751918, "grad_norm": 0.11861552530992366, "learning_rate": 2.8596531606555377e-06, "loss": 1.0017, "step": 1742 }, { "epoch": 4.457800511508951, "grad_norm": 0.11486045991602647, "learning_rate": 2.833185815445929e-06, "loss": 0.956, "step": 1743 }, { "epoch": 4.460358056265985, "grad_norm": 0.1164478351871581, "learning_rate": 2.806837026167899e-06, "loss": 0.9627, "step": 1744 }, { "epoch": 4.462915601023018, "grad_norm": 0.11604852293545323, "learning_rate": 2.780606876869678e-06, "loss": 0.9723, "step": 1745 }, { "epoch": 4.465473145780051, "grad_norm": 0.11757102549840029, "learning_rate": 2.7544954512210842e-06, "loss": 0.9769, "step": 1746 }, { "epoch": 4.468030690537084, "grad_norm": 0.12255097610890886, "learning_rate": 2.7285028325131892e-06, "loss": 0.963, "step": 1747 }, { "epoch": 4.470588235294118, "grad_norm": 0.11399478345208364, "learning_rate": 2.702629103658114e-06, "loss": 0.9682, "step": 1748 }, { "epoch": 4.4731457800511505, "grad_norm": 0.11142062474858554, "learning_rate": 2.6768743471887337e-06, "loss": 0.9715, "step": 1749 }, { "epoch": 4.475703324808184, "grad_norm": 0.11410463536786239, "learning_rate": 2.651238645258416e-06, "loss": 0.9953, "step": 1750 }, { "epoch": 4.478260869565218, "grad_norm": 0.12100539423129517, "learning_rate": 2.6257220796407756e-06, "loss": 0.9606, "step": 1751 }, { "epoch": 4.4808184143222505, "grad_norm": 0.11758059360916463, "learning_rate": 2.6003247317293847e-06, "loss": 0.9759, "step": 1752 }, { "epoch": 4.483375959079284, "grad_norm": 0.11144724361162288, "learning_rate": 2.5750466825375585e-06, "loss": 0.9872, "step": 1753 }, { "epoch": 4.485933503836317, "grad_norm": 0.11347669660230812, "learning_rate": 2.5498880126980518e-06, "loss": 0.988, "step": 1754 }, { "epoch": 4.4884910485933505, "grad_norm": 0.11874354177271917, "learning_rate": 2.5248488024628024e-06, "loss": 0.9636, "step": 1755 }, { "epoch": 4.491048593350383, "grad_norm": 0.11925275787467719, "learning_rate": 2.499929131702734e-06, "loss": 0.9773, "step": 1756 }, { "epoch": 4.493606138107417, "grad_norm": 0.11090730679921106, "learning_rate": 2.4751290799074213e-06, "loss": 0.9769, "step": 1757 }, { "epoch": 4.4961636828644505, "grad_norm": 0.11558893311677444, "learning_rate": 2.450448726184891e-06, "loss": 0.9523, "step": 1758 }, { "epoch": 4.498721227621483, "grad_norm": 0.1273459674627714, "learning_rate": 2.425888149261346e-06, "loss": 0.9389, "step": 1759 }, { "epoch": 4.501278772378517, "grad_norm": 0.11864458292723912, "learning_rate": 2.401447427480941e-06, "loss": 0.9895, "step": 1760 }, { "epoch": 4.5038363171355495, "grad_norm": 0.10951592315079048, "learning_rate": 2.3771266388054846e-06, "loss": 0.9817, "step": 1761 }, { "epoch": 4.506393861892583, "grad_norm": 0.11357909677815746, "learning_rate": 2.3529258608142326e-06, "loss": 0.9939, "step": 1762 }, { "epoch": 4.508951406649617, "grad_norm": 0.11971018322164677, "learning_rate": 2.3288451707036288e-06, "loss": 0.9793, "step": 1763 }, { "epoch": 4.5115089514066495, "grad_norm": 0.1122150483378089, "learning_rate": 2.3048846452870597e-06, "loss": 0.9554, "step": 1764 }, { "epoch": 4.514066496163683, "grad_norm": 0.11638676977027419, "learning_rate": 2.281044360994589e-06, "loss": 0.9705, "step": 1765 }, { "epoch": 4.516624040920716, "grad_norm": 0.11255736814038485, "learning_rate": 2.257324393872753e-06, "loss": 0.9569, "step": 1766 }, { "epoch": 4.5191815856777495, "grad_norm": 0.1183417374581666, "learning_rate": 2.233724819584282e-06, "loss": 0.9875, "step": 1767 }, { "epoch": 4.521739130434782, "grad_norm": 0.11525721410180423, "learning_rate": 2.210245713407879e-06, "loss": 0.9851, "step": 1768 }, { "epoch": 4.524296675191816, "grad_norm": 0.11672302497611096, "learning_rate": 2.186887150237964e-06, "loss": 0.9434, "step": 1769 }, { "epoch": 4.526854219948849, "grad_norm": 0.11462568134574831, "learning_rate": 2.163649204584464e-06, "loss": 0.9676, "step": 1770 }, { "epoch": 4.529411764705882, "grad_norm": 0.11637157916960862, "learning_rate": 2.140531950572551e-06, "loss": 0.9625, "step": 1771 }, { "epoch": 4.531969309462916, "grad_norm": 0.11317862758953197, "learning_rate": 2.117535461942386e-06, "loss": 0.9667, "step": 1772 }, { "epoch": 4.534526854219949, "grad_norm": 0.1126947562619317, "learning_rate": 2.094659812048945e-06, "loss": 0.9548, "step": 1773 }, { "epoch": 4.537084398976982, "grad_norm": 0.1135661694913091, "learning_rate": 2.071905073861724e-06, "loss": 0.9586, "step": 1774 }, { "epoch": 4.539641943734015, "grad_norm": 0.11250229002588859, "learning_rate": 2.049271319964543e-06, "loss": 0.9482, "step": 1775 }, { "epoch": 4.542199488491049, "grad_norm": 0.11649750185398619, "learning_rate": 2.0267586225552892e-06, "loss": 0.9571, "step": 1776 }, { "epoch": 4.544757033248082, "grad_norm": 0.12069116537202378, "learning_rate": 2.0043670534457194e-06, "loss": 0.9737, "step": 1777 }, { "epoch": 4.547314578005115, "grad_norm": 0.11274024393050174, "learning_rate": 1.982096684061192e-06, "loss": 0.9848, "step": 1778 }, { "epoch": 4.549872122762149, "grad_norm": 0.11791521579577226, "learning_rate": 1.959947585440469e-06, "loss": 0.9695, "step": 1779 }, { "epoch": 4.552429667519181, "grad_norm": 0.11087965419915281, "learning_rate": 1.9379198282354794e-06, "loss": 0.9621, "step": 1780 }, { "epoch": 4.554987212276215, "grad_norm": 0.11400022805723166, "learning_rate": 1.916013482711092e-06, "loss": 0.9935, "step": 1781 }, { "epoch": 4.557544757033249, "grad_norm": 0.11289106885368878, "learning_rate": 1.8942286187448865e-06, "loss": 0.9679, "step": 1782 }, { "epoch": 4.560102301790281, "grad_norm": 0.11665726379138751, "learning_rate": 1.8725653058269433e-06, "loss": 1.0187, "step": 1783 }, { "epoch": 4.562659846547315, "grad_norm": 0.1067411918098924, "learning_rate": 1.8510236130596126e-06, "loss": 0.9438, "step": 1784 }, { "epoch": 4.565217391304348, "grad_norm": 0.11467062095804521, "learning_rate": 1.8296036091573022e-06, "loss": 0.9453, "step": 1785 }, { "epoch": 4.567774936061381, "grad_norm": 0.1141598096289312, "learning_rate": 1.8083053624462365e-06, "loss": 0.9531, "step": 1786 }, { "epoch": 4.570332480818414, "grad_norm": 0.11829288762440843, "learning_rate": 1.7871289408642889e-06, "loss": 0.9635, "step": 1787 }, { "epoch": 4.572890025575448, "grad_norm": 0.11375107040879473, "learning_rate": 1.7660744119606876e-06, "loss": 0.9531, "step": 1788 }, { "epoch": 4.57544757033248, "grad_norm": 0.11425273488833146, "learning_rate": 1.745141842895861e-06, "loss": 1.0048, "step": 1789 }, { "epoch": 4.578005115089514, "grad_norm": 0.1136362533257648, "learning_rate": 1.7243313004412288e-06, "loss": 0.9783, "step": 1790 }, { "epoch": 4.580562659846548, "grad_norm": 0.11456227986474488, "learning_rate": 1.7036428509789394e-06, "loss": 0.9669, "step": 1791 }, { "epoch": 4.58312020460358, "grad_norm": 0.11073420030884291, "learning_rate": 1.6830765605016842e-06, "loss": 0.9688, "step": 1792 }, { "epoch": 4.585677749360614, "grad_norm": 0.11285691919453725, "learning_rate": 1.6626324946125061e-06, "loss": 0.962, "step": 1793 }, { "epoch": 4.588235294117647, "grad_norm": 0.11211314756286417, "learning_rate": 1.6423107185245557e-06, "loss": 0.972, "step": 1794 }, { "epoch": 4.59079283887468, "grad_norm": 0.11365304032190994, "learning_rate": 1.6221112970609088e-06, "loss": 0.9526, "step": 1795 }, { "epoch": 4.593350383631714, "grad_norm": 0.10847895079352249, "learning_rate": 1.60203429465434e-06, "loss": 0.9635, "step": 1796 }, { "epoch": 4.595907928388747, "grad_norm": 0.11376227683469585, "learning_rate": 1.5820797753471495e-06, "loss": 0.971, "step": 1797 }, { "epoch": 4.59846547314578, "grad_norm": 0.10877615039280635, "learning_rate": 1.562247802790915e-06, "loss": 0.9482, "step": 1798 }, { "epoch": 4.601023017902813, "grad_norm": 0.1152456202929066, "learning_rate": 1.542538440246304e-06, "loss": 0.9568, "step": 1799 }, { "epoch": 4.603580562659847, "grad_norm": 0.11554187769428996, "learning_rate": 1.5229517505829105e-06, "loss": 0.9892, "step": 1800 }, { "epoch": 4.6061381074168795, "grad_norm": 0.11514254996593944, "learning_rate": 1.5034877962789973e-06, "loss": 0.9494, "step": 1801 }, { "epoch": 4.608695652173913, "grad_norm": 0.11224106372638573, "learning_rate": 1.4841466394213311e-06, "loss": 0.9392, "step": 1802 }, { "epoch": 4.611253196930946, "grad_norm": 0.11385256614913049, "learning_rate": 1.4649283417049653e-06, "loss": 0.9633, "step": 1803 }, { "epoch": 4.6138107416879794, "grad_norm": 0.10775692637459648, "learning_rate": 1.445832964433076e-06, "loss": 0.9312, "step": 1804 }, { "epoch": 4.616368286445013, "grad_norm": 0.10984407586190349, "learning_rate": 1.4268605685167125e-06, "loss": 0.9746, "step": 1805 }, { "epoch": 4.618925831202046, "grad_norm": 0.11487256938413119, "learning_rate": 1.4080112144746516e-06, "loss": 0.9383, "step": 1806 }, { "epoch": 4.621483375959079, "grad_norm": 0.11240309733019481, "learning_rate": 1.3892849624331973e-06, "loss": 0.9678, "step": 1807 }, { "epoch": 4.624040920716112, "grad_norm": 0.1169186360063392, "learning_rate": 1.3706818721259584e-06, "loss": 1.0104, "step": 1808 }, { "epoch": 4.626598465473146, "grad_norm": 0.11728481679961328, "learning_rate": 1.35220200289369e-06, "loss": 0.9806, "step": 1809 }, { "epoch": 4.629156010230179, "grad_norm": 0.11281383249587894, "learning_rate": 1.3338454136840872e-06, "loss": 0.9677, "step": 1810 }, { "epoch": 4.631713554987212, "grad_norm": 0.10499750834126508, "learning_rate": 1.315612163051605e-06, "loss": 0.9676, "step": 1811 }, { "epoch": 4.634271099744246, "grad_norm": 0.1110910923804026, "learning_rate": 1.2975023091572657e-06, "loss": 0.994, "step": 1812 }, { "epoch": 4.6368286445012785, "grad_norm": 0.11326212934873885, "learning_rate": 1.279515909768483e-06, "loss": 0.9345, "step": 1813 }, { "epoch": 4.639386189258312, "grad_norm": 0.11405643543560509, "learning_rate": 1.2616530222588685e-06, "loss": 1.0019, "step": 1814 }, { "epoch": 4.641943734015345, "grad_norm": 0.1158242012128305, "learning_rate": 1.2439137036080573e-06, "loss": 0.9927, "step": 1815 }, { "epoch": 4.6445012787723785, "grad_norm": 0.1149985852845067, "learning_rate": 1.2262980104014921e-06, "loss": 0.9925, "step": 1816 }, { "epoch": 4.647058823529412, "grad_norm": 0.11465565149918291, "learning_rate": 1.208805998830318e-06, "loss": 0.9895, "step": 1817 }, { "epoch": 4.649616368286445, "grad_norm": 0.11220682285996932, "learning_rate": 1.1914377246911158e-06, "loss": 0.9783, "step": 1818 }, { "epoch": 4.6521739130434785, "grad_norm": 0.11250586266208777, "learning_rate": 1.174193243385786e-06, "loss": 0.9549, "step": 1819 }, { "epoch": 4.654731457800511, "grad_norm": 0.10899951533184332, "learning_rate": 1.1570726099213414e-06, "loss": 0.943, "step": 1820 }, { "epoch": 4.657289002557545, "grad_norm": 0.11251333624607752, "learning_rate": 1.1400758789097499e-06, "loss": 0.9443, "step": 1821 }, { "epoch": 4.659846547314578, "grad_norm": 0.10854566000476132, "learning_rate": 1.1232031045677405e-06, "loss": 0.9987, "step": 1822 }, { "epoch": 4.662404092071611, "grad_norm": 0.10722540031517447, "learning_rate": 1.106454340716643e-06, "loss": 0.9463, "step": 1823 }, { "epoch": 4.664961636828645, "grad_norm": 0.11013365048748615, "learning_rate": 1.089829640782223e-06, "loss": 0.9599, "step": 1824 }, { "epoch": 4.667519181585678, "grad_norm": 0.10882148143721988, "learning_rate": 1.073329057794501e-06, "loss": 0.9626, "step": 1825 }, { "epoch": 4.670076726342711, "grad_norm": 0.10362710867323117, "learning_rate": 1.0569526443875745e-06, "loss": 0.9586, "step": 1826 }, { "epoch": 4.672634271099744, "grad_norm": 0.10858009010596835, "learning_rate": 1.0407004527994701e-06, "loss": 0.9603, "step": 1827 }, { "epoch": 4.675191815856778, "grad_norm": 0.1044741695809889, "learning_rate": 1.0245725348719681e-06, "loss": 0.9526, "step": 1828 }, { "epoch": 4.677749360613811, "grad_norm": 0.11369566676381282, "learning_rate": 1.0085689420504274e-06, "loss": 0.958, "step": 1829 }, { "epoch": 4.680306905370844, "grad_norm": 0.11199398881890137, "learning_rate": 9.926897253836442e-07, "loss": 0.9825, "step": 1830 }, { "epoch": 4.6828644501278776, "grad_norm": 0.10991238024490946, "learning_rate": 9.769349355236702e-07, "loss": 0.9671, "step": 1831 }, { "epoch": 4.68542199488491, "grad_norm": 0.1085409176039824, "learning_rate": 9.613046227256472e-07, "loss": 0.9883, "step": 1832 }, { "epoch": 4.687979539641944, "grad_norm": 0.10448681879944104, "learning_rate": 9.457988368476579e-07, "loss": 0.9578, "step": 1833 }, { "epoch": 4.690537084398977, "grad_norm": 0.11285313196545035, "learning_rate": 9.304176273505861e-07, "loss": 1.0142, "step": 1834 }, { "epoch": 4.69309462915601, "grad_norm": 0.11325534614886543, "learning_rate": 9.151610432979097e-07, "loss": 0.9775, "step": 1835 }, { "epoch": 4.695652173913043, "grad_norm": 0.11030129813920701, "learning_rate": 9.000291333555888e-07, "loss": 0.9895, "step": 1836 }, { "epoch": 4.698209718670077, "grad_norm": 0.10981087858061067, "learning_rate": 8.850219457918929e-07, "loss": 0.973, "step": 1837 }, { "epoch": 4.70076726342711, "grad_norm": 0.10905450057218977, "learning_rate": 8.701395284772407e-07, "loss": 0.9425, "step": 1838 }, { "epoch": 4.703324808184143, "grad_norm": 0.10782067684161685, "learning_rate": 8.553819288840537e-07, "loss": 0.939, "step": 1839 }, { "epoch": 4.705882352941177, "grad_norm": 0.10938807520129998, "learning_rate": 8.40749194086623e-07, "loss": 0.9976, "step": 1840 }, { "epoch": 4.708439897698209, "grad_norm": 0.10882193167069305, "learning_rate": 8.262413707609185e-07, "loss": 0.9576, "step": 1841 }, { "epoch": 4.710997442455243, "grad_norm": 0.11301673326793968, "learning_rate": 8.118585051844774e-07, "loss": 0.9554, "step": 1842 }, { "epoch": 4.713554987212277, "grad_norm": 0.10441867915032454, "learning_rate": 7.976006432362449e-07, "loss": 0.9571, "step": 1843 }, { "epoch": 4.716112531969309, "grad_norm": 0.10956276651492997, "learning_rate": 7.834678303964094e-07, "loss": 0.9781, "step": 1844 }, { "epoch": 4.718670076726343, "grad_norm": 0.10620345574636991, "learning_rate": 7.694601117462919e-07, "loss": 0.9743, "step": 1845 }, { "epoch": 4.721227621483376, "grad_norm": 0.11385253611249209, "learning_rate": 7.555775319681724e-07, "loss": 0.9642, "step": 1846 }, { "epoch": 4.723785166240409, "grad_norm": 0.10874528997426589, "learning_rate": 7.41820135345166e-07, "loss": 0.9681, "step": 1847 }, { "epoch": 4.726342710997442, "grad_norm": 0.10852960725439142, "learning_rate": 7.281879657610758e-07, "loss": 0.9974, "step": 1848 }, { "epoch": 4.728900255754476, "grad_norm": 0.10745125077889328, "learning_rate": 7.146810667002291e-07, "loss": 0.9972, "step": 1849 }, { "epoch": 4.731457800511509, "grad_norm": 0.10513964134651423, "learning_rate": 7.012994812473928e-07, "loss": 0.965, "step": 1850 }, { "epoch": 4.734015345268542, "grad_norm": 0.10468761014491187, "learning_rate": 6.880432520875868e-07, "loss": 0.9666, "step": 1851 }, { "epoch": 4.736572890025576, "grad_norm": 0.10729241645271993, "learning_rate": 6.749124215059733e-07, "loss": 0.9584, "step": 1852 }, { "epoch": 4.739130434782608, "grad_norm": 0.11359362236943447, "learning_rate": 6.619070313877007e-07, "loss": 0.9999, "step": 1853 }, { "epoch": 4.741687979539642, "grad_norm": 0.10962775985169017, "learning_rate": 6.490271232177981e-07, "loss": 0.9938, "step": 1854 }, { "epoch": 4.744245524296675, "grad_norm": 0.10854938801338014, "learning_rate": 6.362727380810185e-07, "loss": 0.97, "step": 1855 }, { "epoch": 4.746803069053708, "grad_norm": 0.1068077939245722, "learning_rate": 6.236439166617204e-07, "loss": 0.9377, "step": 1856 }, { "epoch": 4.749360613810742, "grad_norm": 0.10646672365841833, "learning_rate": 6.111406992437374e-07, "loss": 0.9583, "step": 1857 }, { "epoch": 4.751918158567775, "grad_norm": 0.10549076785830483, "learning_rate": 5.98763125710251e-07, "loss": 0.9879, "step": 1858 }, { "epoch": 4.754475703324808, "grad_norm": 0.10927589356320379, "learning_rate": 5.865112355436431e-07, "loss": 0.9596, "step": 1859 }, { "epoch": 4.757033248081841, "grad_norm": 0.10616809908251783, "learning_rate": 5.74385067825407e-07, "loss": 0.9843, "step": 1860 }, { "epoch": 4.759590792838875, "grad_norm": 0.1069862495687352, "learning_rate": 5.623846612359884e-07, "loss": 0.9781, "step": 1861 }, { "epoch": 4.762148337595908, "grad_norm": 0.11193531922331344, "learning_rate": 5.505100540546827e-07, "loss": 0.97, "step": 1862 }, { "epoch": 4.764705882352941, "grad_norm": 0.10748221296390185, "learning_rate": 5.387612841595014e-07, "loss": 0.9848, "step": 1863 }, { "epoch": 4.767263427109975, "grad_norm": 0.11368280666790082, "learning_rate": 5.271383890270621e-07, "loss": 0.9915, "step": 1864 }, { "epoch": 4.7698209718670075, "grad_norm": 0.10555727887171781, "learning_rate": 5.156414057324588e-07, "loss": 0.958, "step": 1865 }, { "epoch": 4.772378516624041, "grad_norm": 0.11117163332286255, "learning_rate": 5.042703709491425e-07, "loss": 0.9827, "step": 1866 }, { "epoch": 4.774936061381074, "grad_norm": 0.1117593310611823, "learning_rate": 4.930253209488278e-07, "loss": 0.9591, "step": 1867 }, { "epoch": 4.7774936061381075, "grad_norm": 0.10350421985765804, "learning_rate": 4.819062916013329e-07, "loss": 0.9598, "step": 1868 }, { "epoch": 4.78005115089514, "grad_norm": 0.10831994521473463, "learning_rate": 4.7091331837451736e-07, "loss": 0.974, "step": 1869 }, { "epoch": 4.782608695652174, "grad_norm": 0.10688454108423966, "learning_rate": 4.600464363341228e-07, "loss": 0.9639, "step": 1870 }, { "epoch": 4.7851662404092075, "grad_norm": 0.10740390940878632, "learning_rate": 4.493056801436879e-07, "loss": 1.0149, "step": 1871 }, { "epoch": 4.78772378516624, "grad_norm": 0.1076013560128204, "learning_rate": 4.386910840644376e-07, "loss": 0.9489, "step": 1872 }, { "epoch": 4.790281329923274, "grad_norm": 0.11083779030015688, "learning_rate": 4.2820268195515434e-07, "loss": 0.9543, "step": 1873 }, { "epoch": 4.792838874680307, "grad_norm": 0.10701125626461934, "learning_rate": 4.178405072720937e-07, "loss": 0.9422, "step": 1874 }, { "epoch": 4.79539641943734, "grad_norm": 0.10561417083050885, "learning_rate": 4.076045930688599e-07, "loss": 0.9719, "step": 1875 }, { "epoch": 4.797953964194374, "grad_norm": 0.10517634037309377, "learning_rate": 3.9749497199631724e-07, "loss": 0.9398, "step": 1876 }, { "epoch": 4.8005115089514065, "grad_norm": 0.10322694629271818, "learning_rate": 3.875116763024611e-07, "loss": 0.9813, "step": 1877 }, { "epoch": 4.80306905370844, "grad_norm": 0.11224775443862717, "learning_rate": 3.7765473783233807e-07, "loss": 0.9674, "step": 1878 }, { "epoch": 4.805626598465473, "grad_norm": 0.10356896266412557, "learning_rate": 3.6792418802794386e-07, "loss": 0.952, "step": 1879 }, { "epoch": 4.8081841432225065, "grad_norm": 0.1064632218381248, "learning_rate": 3.583200579280943e-07, "loss": 0.9479, "step": 1880 }, { "epoch": 4.810741687979539, "grad_norm": 0.10668965210736751, "learning_rate": 3.4884237816837696e-07, "loss": 0.979, "step": 1881 }, { "epoch": 4.813299232736573, "grad_norm": 0.11234432110553998, "learning_rate": 3.394911789809996e-07, "loss": 0.9489, "step": 1882 }, { "epoch": 4.8158567774936065, "grad_norm": 0.10305811442615184, "learning_rate": 3.3026649019471946e-07, "loss": 0.9401, "step": 1883 }, { "epoch": 4.818414322250639, "grad_norm": 0.10655521657426903, "learning_rate": 3.2116834123476326e-07, "loss": 0.9613, "step": 1884 }, { "epoch": 4.820971867007673, "grad_norm": 0.10913574692316319, "learning_rate": 3.1219676112270723e-07, "loss": 0.962, "step": 1885 }, { "epoch": 4.823529411764706, "grad_norm": 0.10237237844134321, "learning_rate": 3.033517784763884e-07, "loss": 0.9778, "step": 1886 }, { "epoch": 4.826086956521739, "grad_norm": 0.10794182864566411, "learning_rate": 2.94633421509829e-07, "loss": 0.9765, "step": 1887 }, { "epoch": 4.828644501278772, "grad_norm": 0.10761359006799004, "learning_rate": 2.8604171803312986e-07, "loss": 0.9748, "step": 1888 }, { "epoch": 4.831202046035806, "grad_norm": 0.10699503894718207, "learning_rate": 2.775766954523906e-07, "loss": 0.9611, "step": 1889 }, { "epoch": 4.833759590792839, "grad_norm": 0.10534467355372792, "learning_rate": 2.692383807696164e-07, "loss": 0.941, "step": 1890 }, { "epoch": 4.836317135549872, "grad_norm": 0.10931285876643325, "learning_rate": 2.6102680058264664e-07, "loss": 0.977, "step": 1891 }, { "epoch": 4.838874680306906, "grad_norm": 0.10276275524204001, "learning_rate": 2.5294198108504865e-07, "loss": 0.9424, "step": 1892 }, { "epoch": 4.841432225063938, "grad_norm": 0.10844571118942553, "learning_rate": 2.449839480660421e-07, "loss": 0.9217, "step": 1893 }, { "epoch": 4.843989769820972, "grad_norm": 0.10833347969842211, "learning_rate": 2.371527269104279e-07, "loss": 0.9712, "step": 1894 }, { "epoch": 4.846547314578006, "grad_norm": 0.11577418267184579, "learning_rate": 2.2944834259849058e-07, "loss": 0.961, "step": 1895 }, { "epoch": 4.849104859335038, "grad_norm": 0.10873181639986831, "learning_rate": 2.2187081970593605e-07, "loss": 0.9934, "step": 1896 }, { "epoch": 4.851662404092072, "grad_norm": 0.10735020699640388, "learning_rate": 2.1442018240378948e-07, "loss": 0.9582, "step": 1897 }, { "epoch": 4.854219948849105, "grad_norm": 0.10907777804447821, "learning_rate": 2.0709645445834646e-07, "loss": 0.9714, "step": 1898 }, { "epoch": 4.856777493606138, "grad_norm": 0.1096334655026699, "learning_rate": 1.998996592310709e-07, "loss": 1.0043, "step": 1899 }, { "epoch": 4.859335038363171, "grad_norm": 0.11043084155630652, "learning_rate": 1.928298196785372e-07, "loss": 0.9557, "step": 1900 }, { "epoch": 4.861892583120205, "grad_norm": 0.10553296566509356, "learning_rate": 1.8588695835235926e-07, "loss": 0.9563, "step": 1901 }, { "epoch": 4.864450127877237, "grad_norm": 0.10821191093088821, "learning_rate": 1.790710973991061e-07, "loss": 0.9618, "step": 1902 }, { "epoch": 4.867007672634271, "grad_norm": 0.11019549559652472, "learning_rate": 1.723822585602264e-07, "loss": 0.9745, "step": 1903 }, { "epoch": 4.869565217391305, "grad_norm": 0.10707545546640006, "learning_rate": 1.658204631720084e-07, "loss": 0.9847, "step": 1904 }, { "epoch": 4.872122762148337, "grad_norm": 0.10806355656419639, "learning_rate": 1.5938573216547792e-07, "loss": 0.9789, "step": 1905 }, { "epoch": 4.874680306905371, "grad_norm": 0.10651089976914624, "learning_rate": 1.5307808606635832e-07, "loss": 0.9662, "step": 1906 }, { "epoch": 4.877237851662404, "grad_norm": 0.10490157621191196, "learning_rate": 1.4689754499498166e-07, "loss": 0.9715, "step": 1907 }, { "epoch": 4.879795396419437, "grad_norm": 0.10593964096398936, "learning_rate": 1.408441286662443e-07, "loss": 0.9947, "step": 1908 }, { "epoch": 4.882352941176471, "grad_norm": 0.10970227452154628, "learning_rate": 1.3491785638953592e-07, "loss": 1.0021, "step": 1909 }, { "epoch": 4.884910485933504, "grad_norm": 0.10530476523192196, "learning_rate": 1.291187470686728e-07, "loss": 0.9644, "step": 1910 }, { "epoch": 4.887468030690537, "grad_norm": 0.10894549145015038, "learning_rate": 1.2344681920184455e-07, "loss": 0.9445, "step": 1911 }, { "epoch": 4.89002557544757, "grad_norm": 0.10932447169022794, "learning_rate": 1.1790209088155646e-07, "loss": 0.987, "step": 1912 }, { "epoch": 4.892583120204604, "grad_norm": 0.10669046030097515, "learning_rate": 1.1248457979456729e-07, "loss": 0.9693, "step": 1913 }, { "epoch": 4.8951406649616365, "grad_norm": 0.10317584987321615, "learning_rate": 1.0719430322183144e-07, "loss": 0.9783, "step": 1914 }, { "epoch": 4.89769820971867, "grad_norm": 0.10544431476402813, "learning_rate": 1.0203127803844137e-07, "loss": 0.9698, "step": 1915 }, { "epoch": 4.900255754475703, "grad_norm": 0.1059139406234035, "learning_rate": 9.699552071359642e-08, "loss": 0.9935, "step": 1916 }, { "epoch": 4.9028132992327365, "grad_norm": 0.10842738215390242, "learning_rate": 9.208704731051399e-08, "loss": 0.9509, "step": 1917 }, { "epoch": 4.90537084398977, "grad_norm": 0.10719633245807475, "learning_rate": 8.73058734864074e-08, "loss": 0.9255, "step": 1918 }, { "epoch": 4.907928388746803, "grad_norm": 0.10801419457325277, "learning_rate": 8.265201449242367e-08, "loss": 0.9974, "step": 1919 }, { "epoch": 4.910485933503836, "grad_norm": 0.10217294119273099, "learning_rate": 7.812548517359908e-08, "loss": 0.9625, "step": 1920 }, { "epoch": 4.913043478260869, "grad_norm": 0.10904336094462923, "learning_rate": 7.37262999687971e-08, "loss": 0.9778, "step": 1921 }, { "epoch": 4.915601023017903, "grad_norm": 0.10284703646879777, "learning_rate": 6.945447291069052e-08, "loss": 0.9712, "step": 1922 }, { "epoch": 4.918158567774936, "grad_norm": 0.106153765145594, "learning_rate": 6.531001762569044e-08, "loss": 0.9557, "step": 1923 }, { "epoch": 4.920716112531969, "grad_norm": 0.1074054901476557, "learning_rate": 6.129294733391966e-08, "loss": 0.9581, "step": 1924 }, { "epoch": 4.923273657289003, "grad_norm": 0.10969417040676402, "learning_rate": 5.740327484916819e-08, "loss": 0.9918, "step": 1925 }, { "epoch": 4.9258312020460355, "grad_norm": 0.10657300074043048, "learning_rate": 5.364101257883114e-08, "loss": 0.9898, "step": 1926 }, { "epoch": 4.928388746803069, "grad_norm": 0.10352327904399762, "learning_rate": 5.000617252389983e-08, "loss": 0.9926, "step": 1927 }, { "epoch": 4.930946291560103, "grad_norm": 0.10995132445135233, "learning_rate": 4.649876627890848e-08, "loss": 0.9555, "step": 1928 }, { "epoch": 4.9335038363171355, "grad_norm": 0.10681061794840668, "learning_rate": 4.311880503189869e-08, "loss": 0.9761, "step": 1929 }, { "epoch": 4.936061381074169, "grad_norm": 0.10628340512917848, "learning_rate": 3.986629956437948e-08, "loss": 0.9981, "step": 1930 }, { "epoch": 4.938618925831202, "grad_norm": 0.10958658587579229, "learning_rate": 3.674126025130509e-08, "loss": 0.9968, "step": 1931 }, { "epoch": 4.9411764705882355, "grad_norm": 0.1078617025355873, "learning_rate": 3.374369706102609e-08, "loss": 0.978, "step": 1932 }, { "epoch": 4.943734015345268, "grad_norm": 0.10200017383938678, "learning_rate": 3.0873619555267244e-08, "loss": 0.9466, "step": 1933 }, { "epoch": 4.946291560102302, "grad_norm": 0.10404264076360889, "learning_rate": 2.8131036889096352e-08, "loss": 0.9722, "step": 1934 }, { "epoch": 4.948849104859335, "grad_norm": 0.10619021083501325, "learning_rate": 2.5515957810897663e-08, "loss": 0.956, "step": 1935 }, { "epoch": 4.951406649616368, "grad_norm": 0.10572319413582357, "learning_rate": 2.302839066233631e-08, "loss": 0.9641, "step": 1936 }, { "epoch": 4.953964194373402, "grad_norm": 0.10543925313697497, "learning_rate": 2.066834337833168e-08, "loss": 0.9671, "step": 1937 }, { "epoch": 4.956521739130435, "grad_norm": 0.10450226460470871, "learning_rate": 1.843582348704409e-08, "loss": 0.9409, "step": 1938 }, { "epoch": 4.959079283887468, "grad_norm": 0.10771140839710609, "learning_rate": 1.6330838109839266e-08, "loss": 0.9782, "step": 1939 }, { "epoch": 4.961636828644501, "grad_norm": 0.24834469255449557, "learning_rate": 1.4353393961270556e-08, "loss": 0.9874, "step": 1940 }, { "epoch": 4.964194373401535, "grad_norm": 0.10753233482520143, "learning_rate": 1.2503497349047877e-08, "loss": 0.9653, "step": 1941 }, { "epoch": 4.966751918158568, "grad_norm": 0.11018037804047005, "learning_rate": 1.078115417404213e-08, "loss": 0.977, "step": 1942 }, { "epoch": 4.969309462915601, "grad_norm": 0.10439962360210749, "learning_rate": 9.18636993023192e-09, "loss": 0.9729, "step": 1943 }, { "epoch": 4.971867007672635, "grad_norm": 0.10236667855362652, "learning_rate": 7.719149704716878e-09, "loss": 0.9452, "step": 1944 }, { "epoch": 4.974424552429667, "grad_norm": 0.10489163848161775, "learning_rate": 6.379498177686572e-09, "loss": 0.9638, "step": 1945 }, { "epoch": 4.976982097186701, "grad_norm": 0.11428348011942069, "learning_rate": 5.167419622402747e-09, "loss": 1.0006, "step": 1946 }, { "epoch": 4.979539641943734, "grad_norm": 0.1067592882173923, "learning_rate": 4.082917905194883e-09, "loss": 0.9504, "step": 1947 }, { "epoch": 4.982097186700767, "grad_norm": 0.10497889316311938, "learning_rate": 3.1259964854379877e-09, "loss": 0.9662, "step": 1948 }, { "epoch": 4.9846547314578, "grad_norm": 0.10799110314474006, "learning_rate": 2.2966584155570403e-09, "loss": 0.9928, "step": 1949 }, { "epoch": 4.987212276214834, "grad_norm": 0.10574572409471053, "learning_rate": 1.5949063410003462e-09, "loss": 0.9916, "step": 1950 }, { "epoch": 4.989769820971867, "grad_norm": 0.10267660554290474, "learning_rate": 1.0207425002395355e-09, "loss": 0.9555, "step": 1951 }, { "epoch": 4.9923273657289, "grad_norm": 0.10468023161937465, "learning_rate": 5.74168724760682e-10, "loss": 0.9803, "step": 1952 }, { "epoch": 4.994884910485934, "grad_norm": 0.10454422969262668, "learning_rate": 2.551864390598624e-10, "loss": 0.9432, "step": 1953 }, { "epoch": 4.997442455242966, "grad_norm": 0.10571613867824278, "learning_rate": 6.379666063871525e-11, "loss": 0.9505, "step": 1954 }, { "epoch": 5.0, "grad_norm": 0.10885932868024188, "learning_rate": 0.0, "loss": 0.971, "step": 1955 }, { "epoch": 5.0, "step": 1955, "total_flos": 7122204608430080.0, "train_loss": 1.0287959775656386, "train_runtime": 35452.5316, "train_samples_per_second": 14.103, "train_steps_per_second": 0.055 } ], "logging_steps": 1.0, "max_steps": 1955, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 7122204608430080.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }