{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 6621, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00015103458692040478, "grad_norm": 0.2868419909573293, "learning_rate": 9.999999937461124e-06, "loss": 0.0006, "step": 1 }, { "epoch": 0.00030206917384080956, "grad_norm": 7.150025292186746, "learning_rate": 9.9999997498445e-06, "loss": 0.0641, "step": 2 }, { "epoch": 0.00045310376076121433, "grad_norm": 9.630996758224255, "learning_rate": 9.999999437150131e-06, "loss": 0.205, "step": 3 }, { "epoch": 0.0006041383476816191, "grad_norm": 4.792281582109639, "learning_rate": 9.999998999378025e-06, "loss": 0.0995, "step": 4 }, { "epoch": 0.0007551729346020238, "grad_norm": 5.4445592757361245, "learning_rate": 9.999998436528191e-06, "loss": 0.1739, "step": 5 }, { "epoch": 0.0009062075215224287, "grad_norm": 7.962774642926837, "learning_rate": 9.999997748600647e-06, "loss": 0.168, "step": 6 }, { "epoch": 0.0010572421084428334, "grad_norm": 5.326356509983322, "learning_rate": 9.999996935595409e-06, "loss": 0.1454, "step": 7 }, { "epoch": 0.0012082766953632382, "grad_norm": 5.534416296459608, "learning_rate": 9.999995997512494e-06, "loss": 0.1442, "step": 8 }, { "epoch": 0.001359311282283643, "grad_norm": 5.863248631922456, "learning_rate": 9.999994934351932e-06, "loss": 0.1108, "step": 9 }, { "epoch": 0.0015103458692040477, "grad_norm": 4.732105007597219, "learning_rate": 9.999993746113744e-06, "loss": 0.1387, "step": 10 }, { "epoch": 0.0016613804561244525, "grad_norm": 4.870567636868542, "learning_rate": 9.99999243279796e-06, "loss": 0.1447, "step": 11 }, { "epoch": 0.0018124150430448573, "grad_norm": 3.6484887536294304, "learning_rate": 9.999990994404615e-06, "loss": 0.1343, "step": 12 }, { "epoch": 0.001963449629965262, "grad_norm": 3.853197206117984, "learning_rate": 9.999989430933744e-06, "loss": 0.1289, "step": 13 }, { "epoch": 0.002114484216885667, "grad_norm": 4.8005853925559885, "learning_rate": 9.999987742385386e-06, "loss": 0.1704, "step": 14 }, { "epoch": 0.0022655188038060714, "grad_norm": 3.521350055162576, "learning_rate": 9.999985928759585e-06, "loss": 0.1099, "step": 15 }, { "epoch": 0.0024165533907264765, "grad_norm": 4.232169519369242, "learning_rate": 9.999983990056386e-06, "loss": 0.1402, "step": 16 }, { "epoch": 0.002567587977646881, "grad_norm": 4.778411732151405, "learning_rate": 9.999981926275834e-06, "loss": 0.1286, "step": 17 }, { "epoch": 0.002718622564567286, "grad_norm": 3.835591849502601, "learning_rate": 9.999979737417983e-06, "loss": 0.1462, "step": 18 }, { "epoch": 0.0028696571514876907, "grad_norm": 4.33314569184108, "learning_rate": 9.999977423482892e-06, "loss": 0.1501, "step": 19 }, { "epoch": 0.0030206917384080953, "grad_norm": 4.354693390625557, "learning_rate": 9.99997498447061e-06, "loss": 0.1764, "step": 20 }, { "epoch": 0.0031717263253285004, "grad_norm": 4.056769098497268, "learning_rate": 9.999972420381206e-06, "loss": 0.1544, "step": 21 }, { "epoch": 0.003322760912248905, "grad_norm": 4.174515263350388, "learning_rate": 9.99996973121474e-06, "loss": 0.1616, "step": 22 }, { "epoch": 0.0034737954991693096, "grad_norm": 5.008735750369672, "learning_rate": 9.99996691697128e-06, "loss": 0.1478, "step": 23 }, { "epoch": 0.0036248300860897147, "grad_norm": 3.4395826871125394, "learning_rate": 9.999963977650896e-06, "loss": 0.1294, "step": 24 }, { "epoch": 0.0037758646730101193, "grad_norm": 4.308385338239042, "learning_rate": 9.999960913253664e-06, "loss": 0.1353, "step": 25 }, { "epoch": 0.003926899259930524, "grad_norm": 3.6835353025662956, "learning_rate": 9.999957723779657e-06, "loss": 0.1218, "step": 26 }, { "epoch": 0.0040779338468509285, "grad_norm": 4.380284784733959, "learning_rate": 9.999954409228956e-06, "loss": 0.1428, "step": 27 }, { "epoch": 0.004228968433771334, "grad_norm": 3.985598434309975, "learning_rate": 9.999950969601646e-06, "loss": 0.1366, "step": 28 }, { "epoch": 0.004380003020691739, "grad_norm": 3.8299116064410947, "learning_rate": 9.999947404897811e-06, "loss": 0.1494, "step": 29 }, { "epoch": 0.004531037607612143, "grad_norm": 3.2649771960859075, "learning_rate": 9.99994371511754e-06, "loss": 0.1251, "step": 30 }, { "epoch": 0.004682072194532548, "grad_norm": 3.1434309654894785, "learning_rate": 9.999939900260927e-06, "loss": 0.1177, "step": 31 }, { "epoch": 0.004833106781452953, "grad_norm": 4.969866635427274, "learning_rate": 9.999935960328066e-06, "loss": 0.1595, "step": 32 }, { "epoch": 0.004984141368373357, "grad_norm": 2.9851550939086082, "learning_rate": 9.999931895319057e-06, "loss": 0.1134, "step": 33 }, { "epoch": 0.005135175955293762, "grad_norm": 4.452935094182824, "learning_rate": 9.999927705234e-06, "loss": 0.1362, "step": 34 }, { "epoch": 0.005286210542214167, "grad_norm": 5.132671680536512, "learning_rate": 9.999923390073e-06, "loss": 0.1599, "step": 35 }, { "epoch": 0.005437245129134572, "grad_norm": 3.5522982672840198, "learning_rate": 9.999918949836165e-06, "loss": 0.1326, "step": 36 }, { "epoch": 0.005588279716054976, "grad_norm": 6.122079071898288, "learning_rate": 9.999914384523607e-06, "loss": 0.1426, "step": 37 }, { "epoch": 0.0057393143029753815, "grad_norm": 4.311422465770554, "learning_rate": 9.999909694135442e-06, "loss": 0.1571, "step": 38 }, { "epoch": 0.0058903488898957865, "grad_norm": 6.003208223517402, "learning_rate": 9.999904878671783e-06, "loss": 0.1859, "step": 39 }, { "epoch": 0.006041383476816191, "grad_norm": 4.226886962679195, "learning_rate": 9.999899938132753e-06, "loss": 0.1532, "step": 40 }, { "epoch": 0.006192418063736596, "grad_norm": 4.071972050968826, "learning_rate": 9.999894872518474e-06, "loss": 0.1477, "step": 41 }, { "epoch": 0.006343452650657001, "grad_norm": 4.350954701305775, "learning_rate": 9.999889681829077e-06, "loss": 0.1519, "step": 42 }, { "epoch": 0.006494487237577405, "grad_norm": 4.5771237268635785, "learning_rate": 9.999884366064685e-06, "loss": 0.1591, "step": 43 }, { "epoch": 0.00664552182449781, "grad_norm": 3.9502640600821204, "learning_rate": 9.999878925225436e-06, "loss": 0.1378, "step": 44 }, { "epoch": 0.006796556411418215, "grad_norm": 3.9985349730548996, "learning_rate": 9.999873359311464e-06, "loss": 0.1375, "step": 45 }, { "epoch": 0.006947590998338619, "grad_norm": 3.6149651481771845, "learning_rate": 9.99986766832291e-06, "loss": 0.1348, "step": 46 }, { "epoch": 0.007098625585259024, "grad_norm": 5.6513364383881095, "learning_rate": 9.999861852259915e-06, "loss": 0.1572, "step": 47 }, { "epoch": 0.007249660172179429, "grad_norm": 3.231800797281274, "learning_rate": 9.999855911122625e-06, "loss": 0.1534, "step": 48 }, { "epoch": 0.0074006947590998335, "grad_norm": 3.3120487314647478, "learning_rate": 9.999849844911188e-06, "loss": 0.1354, "step": 49 }, { "epoch": 0.007551729346020239, "grad_norm": 3.6076782372342424, "learning_rate": 9.999843653625759e-06, "loss": 0.1336, "step": 50 }, { "epoch": 0.007702763932940644, "grad_norm": 3.826526666243839, "learning_rate": 9.999837337266486e-06, "loss": 0.1301, "step": 51 }, { "epoch": 0.007853798519861049, "grad_norm": 4.239963726371498, "learning_rate": 9.999830895833533e-06, "loss": 0.1583, "step": 52 }, { "epoch": 0.008004833106781453, "grad_norm": 29.88861679975928, "learning_rate": 9.99982432932706e-06, "loss": 0.1368, "step": 53 }, { "epoch": 0.008155867693701857, "grad_norm": 5.528875764249062, "learning_rate": 9.999817637747229e-06, "loss": 0.1572, "step": 54 }, { "epoch": 0.008306902280622263, "grad_norm": 5.310565226215291, "learning_rate": 9.999810821094211e-06, "loss": 0.1575, "step": 55 }, { "epoch": 0.008457936867542667, "grad_norm": 4.567144116270126, "learning_rate": 9.999803879368173e-06, "loss": 0.1677, "step": 56 }, { "epoch": 0.008608971454463071, "grad_norm": 3.818173130004099, "learning_rate": 9.999796812569289e-06, "loss": 0.164, "step": 57 }, { "epoch": 0.008760006041383477, "grad_norm": 4.049266411314044, "learning_rate": 9.999789620697738e-06, "loss": 0.1737, "step": 58 }, { "epoch": 0.008911040628303881, "grad_norm": 4.548552629813011, "learning_rate": 9.9997823037537e-06, "loss": 0.1434, "step": 59 }, { "epoch": 0.009062075215224286, "grad_norm": 3.4068256841148816, "learning_rate": 9.999774861737355e-06, "loss": 0.1393, "step": 60 }, { "epoch": 0.009213109802144692, "grad_norm": 3.5153558857471623, "learning_rate": 9.999767294648893e-06, "loss": 0.1524, "step": 61 }, { "epoch": 0.009364144389065096, "grad_norm": 3.720880803436565, "learning_rate": 9.9997596024885e-06, "loss": 0.1192, "step": 62 }, { "epoch": 0.0095151789759855, "grad_norm": 3.5449902332097016, "learning_rate": 9.99975178525637e-06, "loss": 0.1285, "step": 63 }, { "epoch": 0.009666213562905906, "grad_norm": 3.5510606600164687, "learning_rate": 9.999743842952696e-06, "loss": 0.1254, "step": 64 }, { "epoch": 0.00981724814982631, "grad_norm": 3.6837491969724696, "learning_rate": 9.99973577557768e-06, "loss": 0.1417, "step": 65 }, { "epoch": 0.009968282736746714, "grad_norm": 3.3023488297900694, "learning_rate": 9.999727583131524e-06, "loss": 0.1307, "step": 66 }, { "epoch": 0.01011931732366712, "grad_norm": 4.323205838639129, "learning_rate": 9.999719265614432e-06, "loss": 0.142, "step": 67 }, { "epoch": 0.010270351910587524, "grad_norm": 5.286678158354002, "learning_rate": 9.99971082302661e-06, "loss": 0.1774, "step": 68 }, { "epoch": 0.01042138649750793, "grad_norm": 3.963664338880426, "learning_rate": 9.99970225536827e-06, "loss": 0.1374, "step": 69 }, { "epoch": 0.010572421084428334, "grad_norm": 3.6703737295407537, "learning_rate": 9.99969356263963e-06, "loss": 0.1281, "step": 70 }, { "epoch": 0.010723455671348739, "grad_norm": 5.664536916389655, "learning_rate": 9.999684744840904e-06, "loss": 0.1525, "step": 71 }, { "epoch": 0.010874490258269144, "grad_norm": 3.805135065969202, "learning_rate": 9.999675801972312e-06, "loss": 0.1395, "step": 72 }, { "epoch": 0.011025524845189549, "grad_norm": 4.949116221318011, "learning_rate": 9.999666734034082e-06, "loss": 0.1584, "step": 73 }, { "epoch": 0.011176559432109953, "grad_norm": 4.316638459762137, "learning_rate": 9.999657541026434e-06, "loss": 0.1503, "step": 74 }, { "epoch": 0.011327594019030359, "grad_norm": 4.144318158706904, "learning_rate": 9.999648222949605e-06, "loss": 0.1373, "step": 75 }, { "epoch": 0.011478628605950763, "grad_norm": 3.9150064780769616, "learning_rate": 9.999638779803824e-06, "loss": 0.1319, "step": 76 }, { "epoch": 0.011629663192871167, "grad_norm": 3.2601115576853985, "learning_rate": 9.999629211589328e-06, "loss": 0.134, "step": 77 }, { "epoch": 0.011780697779791573, "grad_norm": 3.41571001947362, "learning_rate": 9.999619518306356e-06, "loss": 0.1629, "step": 78 }, { "epoch": 0.011931732366711977, "grad_norm": 3.060166016257363, "learning_rate": 9.99960969995515e-06, "loss": 0.1385, "step": 79 }, { "epoch": 0.012082766953632381, "grad_norm": 3.659990736667174, "learning_rate": 9.999599756535958e-06, "loss": 0.1297, "step": 80 }, { "epoch": 0.012233801540552787, "grad_norm": 3.0860781219154494, "learning_rate": 9.999589688049027e-06, "loss": 0.1193, "step": 81 }, { "epoch": 0.012384836127473191, "grad_norm": 3.3933257306611675, "learning_rate": 9.999579494494612e-06, "loss": 0.1271, "step": 82 }, { "epoch": 0.012535870714393596, "grad_norm": 6.507510066559226, "learning_rate": 9.999569175872961e-06, "loss": 0.1513, "step": 83 }, { "epoch": 0.012686905301314002, "grad_norm": 3.7887190546734986, "learning_rate": 9.999558732184338e-06, "loss": 0.1355, "step": 84 }, { "epoch": 0.012837939888234406, "grad_norm": 3.8889893510192, "learning_rate": 9.999548163429003e-06, "loss": 0.1421, "step": 85 }, { "epoch": 0.01298897447515481, "grad_norm": 3.3325614179015495, "learning_rate": 9.999537469607218e-06, "loss": 0.1314, "step": 86 }, { "epoch": 0.013140009062075216, "grad_norm": 4.1730669048229325, "learning_rate": 9.999526650719254e-06, "loss": 0.1367, "step": 87 }, { "epoch": 0.01329104364899562, "grad_norm": 3.871998765157889, "learning_rate": 9.99951570676538e-06, "loss": 0.146, "step": 88 }, { "epoch": 0.013442078235916024, "grad_norm": 4.475668944672202, "learning_rate": 9.99950463774587e-06, "loss": 0.1505, "step": 89 }, { "epoch": 0.01359311282283643, "grad_norm": 3.125577901560911, "learning_rate": 9.999493443661001e-06, "loss": 0.1195, "step": 90 }, { "epoch": 0.013744147409756834, "grad_norm": 5.1229200110494775, "learning_rate": 9.999482124511054e-06, "loss": 0.123, "step": 91 }, { "epoch": 0.013895181996677239, "grad_norm": 3.693781110337074, "learning_rate": 9.999470680296308e-06, "loss": 0.1356, "step": 92 }, { "epoch": 0.014046216583597644, "grad_norm": 5.327602008360415, "learning_rate": 9.999459111017053e-06, "loss": 0.1601, "step": 93 }, { "epoch": 0.014197251170518049, "grad_norm": 4.97621170735378, "learning_rate": 9.999447416673579e-06, "loss": 0.1613, "step": 94 }, { "epoch": 0.014348285757438453, "grad_norm": 3.286527750303076, "learning_rate": 9.999435597266176e-06, "loss": 0.1399, "step": 95 }, { "epoch": 0.014499320344358859, "grad_norm": 3.2297498088324152, "learning_rate": 9.999423652795142e-06, "loss": 0.1483, "step": 96 }, { "epoch": 0.014650354931279263, "grad_norm": 4.2063556652565905, "learning_rate": 9.999411583260775e-06, "loss": 0.1551, "step": 97 }, { "epoch": 0.014801389518199667, "grad_norm": 6.8591767752607495, "learning_rate": 9.999399388663374e-06, "loss": 0.1407, "step": 98 }, { "epoch": 0.014952424105120073, "grad_norm": 4.683776887642205, "learning_rate": 9.999387069003248e-06, "loss": 0.1325, "step": 99 }, { "epoch": 0.015103458692040477, "grad_norm": 8.142918244765532, "learning_rate": 9.999374624280704e-06, "loss": 0.2018, "step": 100 }, { "epoch": 0.015254493278960881, "grad_norm": 4.333804188415226, "learning_rate": 9.999362054496054e-06, "loss": 0.1726, "step": 101 }, { "epoch": 0.015405527865881287, "grad_norm": 5.625846091271469, "learning_rate": 9.99934935964961e-06, "loss": 0.1379, "step": 102 }, { "epoch": 0.015556562452801691, "grad_norm": 5.182541675338578, "learning_rate": 9.999336539741692e-06, "loss": 0.1404, "step": 103 }, { "epoch": 0.015707597039722097, "grad_norm": 3.0595136160560226, "learning_rate": 9.99932359477262e-06, "loss": 0.1217, "step": 104 }, { "epoch": 0.0158586316266425, "grad_norm": 3.065250373829727, "learning_rate": 9.999310524742716e-06, "loss": 0.1363, "step": 105 }, { "epoch": 0.016009666213562906, "grad_norm": 3.319964039587066, "learning_rate": 9.99929732965231e-06, "loss": 0.1265, "step": 106 }, { "epoch": 0.01616070080048331, "grad_norm": 4.432887669430643, "learning_rate": 9.999284009501732e-06, "loss": 0.1684, "step": 107 }, { "epoch": 0.016311735387403714, "grad_norm": 7.924999946949084, "learning_rate": 9.999270564291312e-06, "loss": 0.1602, "step": 108 }, { "epoch": 0.01646276997432412, "grad_norm": 8.009097849442968, "learning_rate": 9.999256994021389e-06, "loss": 0.1394, "step": 109 }, { "epoch": 0.016613804561244526, "grad_norm": 3.003509077194938, "learning_rate": 9.999243298692301e-06, "loss": 0.1357, "step": 110 }, { "epoch": 0.01676483914816493, "grad_norm": 3.888710342090013, "learning_rate": 9.999229478304393e-06, "loss": 0.1552, "step": 111 }, { "epoch": 0.016915873735085334, "grad_norm": 4.557752429130758, "learning_rate": 9.999215532858007e-06, "loss": 0.1597, "step": 112 }, { "epoch": 0.01706690832200574, "grad_norm": 4.450421409644834, "learning_rate": 9.999201462353497e-06, "loss": 0.1604, "step": 113 }, { "epoch": 0.017217942908926143, "grad_norm": 3.758801941740709, "learning_rate": 9.999187266791208e-06, "loss": 0.1408, "step": 114 }, { "epoch": 0.01736897749584655, "grad_norm": 6.921412703959984, "learning_rate": 9.9991729461715e-06, "loss": 0.1775, "step": 115 }, { "epoch": 0.017520012082766954, "grad_norm": 4.863173883353016, "learning_rate": 9.999158500494732e-06, "loss": 0.1664, "step": 116 }, { "epoch": 0.017671046669687357, "grad_norm": 4.384325702809859, "learning_rate": 9.999143929761263e-06, "loss": 0.1468, "step": 117 }, { "epoch": 0.017822081256607763, "grad_norm": 3.3679286668639277, "learning_rate": 9.999129233971455e-06, "loss": 0.1307, "step": 118 }, { "epoch": 0.01797311584352817, "grad_norm": 3.860685136936363, "learning_rate": 9.999114413125684e-06, "loss": 0.1563, "step": 119 }, { "epoch": 0.01812415043044857, "grad_norm": 5.264869938036361, "learning_rate": 9.999099467224312e-06, "loss": 0.1768, "step": 120 }, { "epoch": 0.018275185017368977, "grad_norm": 3.2981374892109243, "learning_rate": 9.999084396267717e-06, "loss": 0.1338, "step": 121 }, { "epoch": 0.018426219604289383, "grad_norm": 3.0274376038161344, "learning_rate": 9.999069200256276e-06, "loss": 0.1302, "step": 122 }, { "epoch": 0.018577254191209785, "grad_norm": 3.2410930429382354, "learning_rate": 9.999053879190369e-06, "loss": 0.1317, "step": 123 }, { "epoch": 0.01872828877813019, "grad_norm": 4.019713286356711, "learning_rate": 9.999038433070376e-06, "loss": 0.1438, "step": 124 }, { "epoch": 0.018879323365050597, "grad_norm": 3.585149776650965, "learning_rate": 9.999022861896687e-06, "loss": 0.1244, "step": 125 }, { "epoch": 0.019030357951971, "grad_norm": 3.236486652571585, "learning_rate": 9.999007165669691e-06, "loss": 0.1406, "step": 126 }, { "epoch": 0.019181392538891406, "grad_norm": 5.631056574214319, "learning_rate": 9.998991344389782e-06, "loss": 0.173, "step": 127 }, { "epoch": 0.01933242712581181, "grad_norm": 3.5274168703075803, "learning_rate": 9.998975398057354e-06, "loss": 0.1377, "step": 128 }, { "epoch": 0.019483461712732214, "grad_norm": 3.7195919381206903, "learning_rate": 9.998959326672803e-06, "loss": 0.1336, "step": 129 }, { "epoch": 0.01963449629965262, "grad_norm": 6.630727151438578, "learning_rate": 9.998943130236536e-06, "loss": 0.1611, "step": 130 }, { "epoch": 0.019785530886573026, "grad_norm": 3.8303931609033155, "learning_rate": 9.998926808748957e-06, "loss": 0.1602, "step": 131 }, { "epoch": 0.01993656547349343, "grad_norm": 6.004730409627928, "learning_rate": 9.998910362210473e-06, "loss": 0.1722, "step": 132 }, { "epoch": 0.020087600060413834, "grad_norm": 4.991418213217509, "learning_rate": 9.998893790621496e-06, "loss": 0.1778, "step": 133 }, { "epoch": 0.02023863464733424, "grad_norm": 4.624046469955102, "learning_rate": 9.99887709398244e-06, "loss": 0.1788, "step": 134 }, { "epoch": 0.020389669234254643, "grad_norm": 3.8897159021390566, "learning_rate": 9.998860272293722e-06, "loss": 0.161, "step": 135 }, { "epoch": 0.02054070382117505, "grad_norm": 3.624799547369644, "learning_rate": 9.998843325555766e-06, "loss": 0.14, "step": 136 }, { "epoch": 0.020691738408095454, "grad_norm": 3.175336615931592, "learning_rate": 9.998826253768992e-06, "loss": 0.1446, "step": 137 }, { "epoch": 0.02084277299501586, "grad_norm": 4.852824017271855, "learning_rate": 9.998809056933829e-06, "loss": 0.1577, "step": 138 }, { "epoch": 0.020993807581936263, "grad_norm": 4.949332962487865, "learning_rate": 9.998791735050708e-06, "loss": 0.1697, "step": 139 }, { "epoch": 0.02114484216885667, "grad_norm": 4.119476232644044, "learning_rate": 9.998774288120062e-06, "loss": 0.151, "step": 140 }, { "epoch": 0.021295876755777075, "grad_norm": 2.6572155533911337, "learning_rate": 9.998756716142326e-06, "loss": 0.1117, "step": 141 }, { "epoch": 0.021446911342697477, "grad_norm": 3.8276305818965626, "learning_rate": 9.99873901911794e-06, "loss": 0.142, "step": 142 }, { "epoch": 0.021597945929617883, "grad_norm": 3.281902125615535, "learning_rate": 9.998721197047349e-06, "loss": 0.1387, "step": 143 }, { "epoch": 0.02174898051653829, "grad_norm": 7.217157896257923, "learning_rate": 9.998703249930996e-06, "loss": 0.1805, "step": 144 }, { "epoch": 0.02190001510345869, "grad_norm": 3.4181748560639917, "learning_rate": 9.998685177769329e-06, "loss": 0.1419, "step": 145 }, { "epoch": 0.022051049690379097, "grad_norm": 3.845002936293022, "learning_rate": 9.998666980562806e-06, "loss": 0.1628, "step": 146 }, { "epoch": 0.022202084277299503, "grad_norm": 4.405049609893203, "learning_rate": 9.998648658311875e-06, "loss": 0.1487, "step": 147 }, { "epoch": 0.022353118864219906, "grad_norm": 4.082324340329364, "learning_rate": 9.998630211016998e-06, "loss": 0.1524, "step": 148 }, { "epoch": 0.02250415345114031, "grad_norm": 3.6833931627250025, "learning_rate": 9.998611638678637e-06, "loss": 0.1393, "step": 149 }, { "epoch": 0.022655188038060717, "grad_norm": 3.432520678421747, "learning_rate": 9.998592941297254e-06, "loss": 0.1569, "step": 150 }, { "epoch": 0.02280622262498112, "grad_norm": 4.747599358563046, "learning_rate": 9.99857411887332e-06, "loss": 0.149, "step": 151 }, { "epoch": 0.022957257211901526, "grad_norm": 3.9770001258124914, "learning_rate": 9.998555171407305e-06, "loss": 0.1656, "step": 152 }, { "epoch": 0.023108291798821932, "grad_norm": 3.9143172669716697, "learning_rate": 9.99853609889968e-06, "loss": 0.1555, "step": 153 }, { "epoch": 0.023259326385742334, "grad_norm": 3.5484143216515767, "learning_rate": 9.998516901350925e-06, "loss": 0.148, "step": 154 }, { "epoch": 0.02341036097266274, "grad_norm": 3.556122766216046, "learning_rate": 9.998497578761522e-06, "loss": 0.1607, "step": 155 }, { "epoch": 0.023561395559583146, "grad_norm": 4.4893654505828025, "learning_rate": 9.998478131131948e-06, "loss": 0.1669, "step": 156 }, { "epoch": 0.02371243014650355, "grad_norm": 3.413186361801123, "learning_rate": 9.998458558462696e-06, "loss": 0.1575, "step": 157 }, { "epoch": 0.023863464733423954, "grad_norm": 4.056972309368031, "learning_rate": 9.998438860754252e-06, "loss": 0.1344, "step": 158 }, { "epoch": 0.02401449932034436, "grad_norm": 3.541359544150984, "learning_rate": 9.99841903800711e-06, "loss": 0.1595, "step": 159 }, { "epoch": 0.024165533907264763, "grad_norm": 3.5481689043997435, "learning_rate": 9.998399090221765e-06, "loss": 0.1453, "step": 160 }, { "epoch": 0.02431656849418517, "grad_norm": 4.968158158127261, "learning_rate": 9.998379017398717e-06, "loss": 0.1928, "step": 161 }, { "epoch": 0.024467603081105575, "grad_norm": 3.564083777190636, "learning_rate": 9.998358819538469e-06, "loss": 0.1309, "step": 162 }, { "epoch": 0.024618637668025977, "grad_norm": 4.21831081226782, "learning_rate": 9.998338496641524e-06, "loss": 0.1499, "step": 163 }, { "epoch": 0.024769672254946383, "grad_norm": 6.562685025610138, "learning_rate": 9.99831804870839e-06, "loss": 0.1764, "step": 164 }, { "epoch": 0.02492070684186679, "grad_norm": 3.564210882772642, "learning_rate": 9.998297475739582e-06, "loss": 0.1737, "step": 165 }, { "epoch": 0.02507174142878719, "grad_norm": 4.837040388148813, "learning_rate": 9.998276777735613e-06, "loss": 0.1661, "step": 166 }, { "epoch": 0.025222776015707597, "grad_norm": 4.088005083904095, "learning_rate": 9.998255954696999e-06, "loss": 0.179, "step": 167 }, { "epoch": 0.025373810602628003, "grad_norm": 3.719520851229798, "learning_rate": 9.998235006624263e-06, "loss": 0.1537, "step": 168 }, { "epoch": 0.025524845189548406, "grad_norm": 5.50002170116797, "learning_rate": 9.99821393351793e-06, "loss": 0.1668, "step": 169 }, { "epoch": 0.02567587977646881, "grad_norm": 3.7634400895123394, "learning_rate": 9.998192735378521e-06, "loss": 0.1631, "step": 170 }, { "epoch": 0.025826914363389217, "grad_norm": 2.918517575048816, "learning_rate": 9.998171412206574e-06, "loss": 0.1383, "step": 171 }, { "epoch": 0.02597794895030962, "grad_norm": 4.7045308258669225, "learning_rate": 9.998149964002618e-06, "loss": 0.1902, "step": 172 }, { "epoch": 0.026128983537230026, "grad_norm": 4.714579259959394, "learning_rate": 9.99812839076719e-06, "loss": 0.1804, "step": 173 }, { "epoch": 0.02628001812415043, "grad_norm": 3.4203674453310144, "learning_rate": 9.998106692500832e-06, "loss": 0.1541, "step": 174 }, { "epoch": 0.026431052711070834, "grad_norm": 3.9611248515664164, "learning_rate": 9.998084869204085e-06, "loss": 0.1506, "step": 175 }, { "epoch": 0.02658208729799124, "grad_norm": 3.920189711261336, "learning_rate": 9.998062920877494e-06, "loss": 0.1673, "step": 176 }, { "epoch": 0.026733121884911646, "grad_norm": 2.861399459463572, "learning_rate": 9.99804084752161e-06, "loss": 0.1242, "step": 177 }, { "epoch": 0.02688415647183205, "grad_norm": 4.43204944618275, "learning_rate": 9.998018649136985e-06, "loss": 0.1644, "step": 178 }, { "epoch": 0.027035191058752454, "grad_norm": 3.3545019659490927, "learning_rate": 9.99799632572417e-06, "loss": 0.1396, "step": 179 }, { "epoch": 0.02718622564567286, "grad_norm": 7.210520108664242, "learning_rate": 9.99797387728373e-06, "loss": 0.1612, "step": 180 }, { "epoch": 0.027337260232593263, "grad_norm": 3.6489572323613695, "learning_rate": 9.997951303816225e-06, "loss": 0.1512, "step": 181 }, { "epoch": 0.02748829481951367, "grad_norm": 3.6838259994562574, "learning_rate": 9.997928605322215e-06, "loss": 0.1793, "step": 182 }, { "epoch": 0.027639329406434075, "grad_norm": 3.7636275015908383, "learning_rate": 9.997905781802272e-06, "loss": 0.1631, "step": 183 }, { "epoch": 0.027790363993354477, "grad_norm": 3.36738139815884, "learning_rate": 9.997882833256968e-06, "loss": 0.1673, "step": 184 }, { "epoch": 0.027941398580274883, "grad_norm": 2.9618777487578085, "learning_rate": 9.997859759686875e-06, "loss": 0.1417, "step": 185 }, { "epoch": 0.02809243316719529, "grad_norm": 5.52549360590678, "learning_rate": 9.997836561092567e-06, "loss": 0.175, "step": 186 }, { "epoch": 0.02824346775411569, "grad_norm": 3.039024628090466, "learning_rate": 9.997813237474631e-06, "loss": 0.1385, "step": 187 }, { "epoch": 0.028394502341036097, "grad_norm": 2.9616741328642324, "learning_rate": 9.997789788833645e-06, "loss": 0.1404, "step": 188 }, { "epoch": 0.028545536927956503, "grad_norm": 5.1014358019422055, "learning_rate": 9.997766215170199e-06, "loss": 0.1454, "step": 189 }, { "epoch": 0.028696571514876906, "grad_norm": 3.4754044726875573, "learning_rate": 9.997742516484881e-06, "loss": 0.1604, "step": 190 }, { "epoch": 0.02884760610179731, "grad_norm": 4.276468904054172, "learning_rate": 9.997718692778283e-06, "loss": 0.1751, "step": 191 }, { "epoch": 0.028998640688717717, "grad_norm": 6.425844404074553, "learning_rate": 9.997694744051003e-06, "loss": 0.1547, "step": 192 }, { "epoch": 0.02914967527563812, "grad_norm": 4.5601135771627, "learning_rate": 9.99767067030364e-06, "loss": 0.1487, "step": 193 }, { "epoch": 0.029300709862558526, "grad_norm": 3.6196928587397865, "learning_rate": 9.997646471536796e-06, "loss": 0.1599, "step": 194 }, { "epoch": 0.02945174444947893, "grad_norm": 3.5780028415931646, "learning_rate": 9.997622147751073e-06, "loss": 0.1453, "step": 195 }, { "epoch": 0.029602779036399334, "grad_norm": 4.407273881248349, "learning_rate": 9.997597698947084e-06, "loss": 0.1849, "step": 196 }, { "epoch": 0.02975381362331974, "grad_norm": 5.360342651633241, "learning_rate": 9.99757312512544e-06, "loss": 0.1781, "step": 197 }, { "epoch": 0.029904848210240146, "grad_norm": 3.681111427851151, "learning_rate": 9.997548426286753e-06, "loss": 0.1667, "step": 198 }, { "epoch": 0.03005588279716055, "grad_norm": 3.3102575614599328, "learning_rate": 9.997523602431643e-06, "loss": 0.1372, "step": 199 }, { "epoch": 0.030206917384080954, "grad_norm": 3.5865651596479817, "learning_rate": 9.99749865356073e-06, "loss": 0.1486, "step": 200 }, { "epoch": 0.03035795197100136, "grad_norm": 2.9643268914773886, "learning_rate": 9.99747357967464e-06, "loss": 0.1398, "step": 201 }, { "epoch": 0.030508986557921763, "grad_norm": 3.73022234365957, "learning_rate": 9.997448380773998e-06, "loss": 0.1428, "step": 202 }, { "epoch": 0.03066002114484217, "grad_norm": 4.851395302290496, "learning_rate": 9.997423056859434e-06, "loss": 0.166, "step": 203 }, { "epoch": 0.030811055731762575, "grad_norm": 4.07295669091537, "learning_rate": 9.997397607931584e-06, "loss": 0.1603, "step": 204 }, { "epoch": 0.030962090318682977, "grad_norm": 3.357133402048586, "learning_rate": 9.997372033991083e-06, "loss": 0.1583, "step": 205 }, { "epoch": 0.031113124905603383, "grad_norm": 2.95273354937129, "learning_rate": 9.99734633503857e-06, "loss": 0.1435, "step": 206 }, { "epoch": 0.031264159492523785, "grad_norm": 4.059779666305551, "learning_rate": 9.997320511074691e-06, "loss": 0.1599, "step": 207 }, { "epoch": 0.031415194079444195, "grad_norm": 4.311495321313036, "learning_rate": 9.997294562100088e-06, "loss": 0.1685, "step": 208 }, { "epoch": 0.0315662286663646, "grad_norm": 4.03224942220964, "learning_rate": 9.997268488115415e-06, "loss": 0.1621, "step": 209 }, { "epoch": 0.031717263253285, "grad_norm": 5.855913546724843, "learning_rate": 9.997242289121318e-06, "loss": 0.1798, "step": 210 }, { "epoch": 0.03186829784020541, "grad_norm": 4.76750528625266, "learning_rate": 9.997215965118457e-06, "loss": 0.1996, "step": 211 }, { "epoch": 0.03201933242712581, "grad_norm": 3.089316865624052, "learning_rate": 9.997189516107488e-06, "loss": 0.1672, "step": 212 }, { "epoch": 0.032170367014046214, "grad_norm": 4.615971774360618, "learning_rate": 9.997162942089076e-06, "loss": 0.1502, "step": 213 }, { "epoch": 0.03232140160096662, "grad_norm": 3.265893265440729, "learning_rate": 9.997136243063883e-06, "loss": 0.1485, "step": 214 }, { "epoch": 0.032472436187887026, "grad_norm": 4.173653559938493, "learning_rate": 9.997109419032575e-06, "loss": 0.1826, "step": 215 }, { "epoch": 0.03262347077480743, "grad_norm": 3.4978285969582092, "learning_rate": 9.99708246999583e-06, "loss": 0.1777, "step": 216 }, { "epoch": 0.03277450536172784, "grad_norm": 3.3616699334158517, "learning_rate": 9.997055395954314e-06, "loss": 0.1593, "step": 217 }, { "epoch": 0.03292553994864824, "grad_norm": 8.99864420621868, "learning_rate": 9.997028196908708e-06, "loss": 0.1844, "step": 218 }, { "epoch": 0.03307657453556864, "grad_norm": 3.4677252707471538, "learning_rate": 9.997000872859693e-06, "loss": 0.1735, "step": 219 }, { "epoch": 0.03322760912248905, "grad_norm": 3.4969364280962005, "learning_rate": 9.996973423807952e-06, "loss": 0.1552, "step": 220 }, { "epoch": 0.033378643709409454, "grad_norm": 3.8246443280217153, "learning_rate": 9.996945849754172e-06, "loss": 0.1939, "step": 221 }, { "epoch": 0.03352967829632986, "grad_norm": 3.2404652485997274, "learning_rate": 9.996918150699042e-06, "loss": 0.1422, "step": 222 }, { "epoch": 0.033680712883250266, "grad_norm": 3.284886864816569, "learning_rate": 9.996890326643255e-06, "loss": 0.1426, "step": 223 }, { "epoch": 0.03383174747017067, "grad_norm": 3.040115256164735, "learning_rate": 9.996862377587508e-06, "loss": 0.1432, "step": 224 }, { "epoch": 0.03398278205709107, "grad_norm": 3.410949839684935, "learning_rate": 9.996834303532498e-06, "loss": 0.1562, "step": 225 }, { "epoch": 0.03413381664401148, "grad_norm": 3.714074891854224, "learning_rate": 9.99680610447893e-06, "loss": 0.177, "step": 226 }, { "epoch": 0.03428485123093188, "grad_norm": 3.1921021647970114, "learning_rate": 9.996777780427507e-06, "loss": 0.1395, "step": 227 }, { "epoch": 0.034435885817852285, "grad_norm": 3.031436176768711, "learning_rate": 9.99674933137894e-06, "loss": 0.1402, "step": 228 }, { "epoch": 0.034586920404772695, "grad_norm": 6.391260085891436, "learning_rate": 9.996720757333937e-06, "loss": 0.1722, "step": 229 }, { "epoch": 0.0347379549916931, "grad_norm": 3.9724187464680316, "learning_rate": 9.996692058293216e-06, "loss": 0.1725, "step": 230 }, { "epoch": 0.0348889895786135, "grad_norm": 4.2253423371679615, "learning_rate": 9.996663234257493e-06, "loss": 0.1747, "step": 231 }, { "epoch": 0.03504002416553391, "grad_norm": 5.150124203651883, "learning_rate": 9.996634285227491e-06, "loss": 0.1774, "step": 232 }, { "epoch": 0.03519105875245431, "grad_norm": 4.121624159171427, "learning_rate": 9.996605211203934e-06, "loss": 0.1696, "step": 233 }, { "epoch": 0.035342093339374714, "grad_norm": 3.947874026533828, "learning_rate": 9.996576012187548e-06, "loss": 0.1553, "step": 234 }, { "epoch": 0.03549312792629512, "grad_norm": 4.307823082387146, "learning_rate": 9.996546688179064e-06, "loss": 0.1564, "step": 235 }, { "epoch": 0.035644162513215526, "grad_norm": 3.6838023884135267, "learning_rate": 9.996517239179216e-06, "loss": 0.1434, "step": 236 }, { "epoch": 0.03579519710013593, "grad_norm": 3.143145321185023, "learning_rate": 9.996487665188739e-06, "loss": 0.1537, "step": 237 }, { "epoch": 0.03594623168705634, "grad_norm": 4.114114850499709, "learning_rate": 9.996457966208374e-06, "loss": 0.1678, "step": 238 }, { "epoch": 0.03609726627397674, "grad_norm": 3.7368062307704983, "learning_rate": 9.996428142238866e-06, "loss": 0.137, "step": 239 }, { "epoch": 0.03624830086089714, "grad_norm": 4.021500354099638, "learning_rate": 9.996398193280958e-06, "loss": 0.1604, "step": 240 }, { "epoch": 0.03639933544781755, "grad_norm": 3.4703494322945247, "learning_rate": 9.9963681193354e-06, "loss": 0.1621, "step": 241 }, { "epoch": 0.036550370034737954, "grad_norm": 4.7813491528966985, "learning_rate": 9.996337920402944e-06, "loss": 0.177, "step": 242 }, { "epoch": 0.03670140462165836, "grad_norm": 3.844028481949023, "learning_rate": 9.996307596484346e-06, "loss": 0.1762, "step": 243 }, { "epoch": 0.036852439208578766, "grad_norm": 4.186161028024291, "learning_rate": 9.996277147580365e-06, "loss": 0.1677, "step": 244 }, { "epoch": 0.03700347379549917, "grad_norm": 3.5141468416257355, "learning_rate": 9.996246573691761e-06, "loss": 0.1596, "step": 245 }, { "epoch": 0.03715450838241957, "grad_norm": 3.6657676622073865, "learning_rate": 9.996215874819301e-06, "loss": 0.1652, "step": 246 }, { "epoch": 0.03730554296933998, "grad_norm": 3.3853845609725366, "learning_rate": 9.996185050963753e-06, "loss": 0.1484, "step": 247 }, { "epoch": 0.03745657755626038, "grad_norm": 3.845643845047163, "learning_rate": 9.996154102125885e-06, "loss": 0.16, "step": 248 }, { "epoch": 0.037607612143180785, "grad_norm": 3.6551645290627364, "learning_rate": 9.996123028306474e-06, "loss": 0.1625, "step": 249 }, { "epoch": 0.037758646730101195, "grad_norm": 3.058782042766959, "learning_rate": 9.996091829506298e-06, "loss": 0.1415, "step": 250 }, { "epoch": 0.0379096813170216, "grad_norm": 4.067185610499594, "learning_rate": 9.996060505726135e-06, "loss": 0.1705, "step": 251 }, { "epoch": 0.038060715903942, "grad_norm": 3.9703631367975003, "learning_rate": 9.996029056966769e-06, "loss": 0.1434, "step": 252 }, { "epoch": 0.03821175049086241, "grad_norm": 3.6390043966181342, "learning_rate": 9.99599748322899e-06, "loss": 0.1535, "step": 253 }, { "epoch": 0.03836278507778281, "grad_norm": 3.786047540047175, "learning_rate": 9.995965784513582e-06, "loss": 0.1777, "step": 254 }, { "epoch": 0.038513819664703214, "grad_norm": 4.475812510317071, "learning_rate": 9.995933960821343e-06, "loss": 0.1678, "step": 255 }, { "epoch": 0.03866485425162362, "grad_norm": 3.520028297553961, "learning_rate": 9.995902012153067e-06, "loss": 0.1616, "step": 256 }, { "epoch": 0.038815888838544026, "grad_norm": 3.8098724976041884, "learning_rate": 9.995869938509553e-06, "loss": 0.1346, "step": 257 }, { "epoch": 0.03896692342546443, "grad_norm": 4.379195950111815, "learning_rate": 9.995837739891604e-06, "loss": 0.1715, "step": 258 }, { "epoch": 0.03911795801238484, "grad_norm": 3.4548294149863747, "learning_rate": 9.995805416300024e-06, "loss": 0.1624, "step": 259 }, { "epoch": 0.03926899259930524, "grad_norm": 3.487736965900251, "learning_rate": 9.995772967735624e-06, "loss": 0.1372, "step": 260 }, { "epoch": 0.03942002718622564, "grad_norm": 3.8668814968565917, "learning_rate": 9.995740394199214e-06, "loss": 0.1498, "step": 261 }, { "epoch": 0.03957106177314605, "grad_norm": 4.073400723504865, "learning_rate": 9.99570769569161e-06, "loss": 0.1735, "step": 262 }, { "epoch": 0.039722096360066454, "grad_norm": 3.2932124808839625, "learning_rate": 9.99567487221363e-06, "loss": 0.1474, "step": 263 }, { "epoch": 0.03987313094698686, "grad_norm": 6.773514729391781, "learning_rate": 9.995641923766092e-06, "loss": 0.1805, "step": 264 }, { "epoch": 0.040024165533907266, "grad_norm": 3.2617464173645785, "learning_rate": 9.995608850349825e-06, "loss": 0.1597, "step": 265 }, { "epoch": 0.04017520012082767, "grad_norm": 3.6836122087750662, "learning_rate": 9.995575651965653e-06, "loss": 0.1631, "step": 266 }, { "epoch": 0.04032623470774807, "grad_norm": 3.1582493921381745, "learning_rate": 9.995542328614409e-06, "loss": 0.1466, "step": 267 }, { "epoch": 0.04047726929466848, "grad_norm": 3.381418770487439, "learning_rate": 9.995508880296922e-06, "loss": 0.1605, "step": 268 }, { "epoch": 0.04062830388158888, "grad_norm": 7.052696398962445, "learning_rate": 9.995475307014035e-06, "loss": 0.1481, "step": 269 }, { "epoch": 0.040779338468509285, "grad_norm": 4.422287317113572, "learning_rate": 9.995441608766584e-06, "loss": 0.1577, "step": 270 }, { "epoch": 0.040930373055429695, "grad_norm": 3.3050329792009085, "learning_rate": 9.995407785555411e-06, "loss": 0.1614, "step": 271 }, { "epoch": 0.0410814076423501, "grad_norm": 3.4006996284590834, "learning_rate": 9.995373837381366e-06, "loss": 0.1484, "step": 272 }, { "epoch": 0.0412324422292705, "grad_norm": 5.88486440680623, "learning_rate": 9.995339764245296e-06, "loss": 0.1778, "step": 273 }, { "epoch": 0.04138347681619091, "grad_norm": 3.646029826397492, "learning_rate": 9.995305566148051e-06, "loss": 0.1504, "step": 274 }, { "epoch": 0.04153451140311131, "grad_norm": 4.352940927647637, "learning_rate": 9.99527124309049e-06, "loss": 0.2259, "step": 275 }, { "epoch": 0.04168554599003172, "grad_norm": 4.082542548954162, "learning_rate": 9.99523679507347e-06, "loss": 0.1845, "step": 276 }, { "epoch": 0.04183658057695212, "grad_norm": 3.8380451653213736, "learning_rate": 9.995202222097855e-06, "loss": 0.1593, "step": 277 }, { "epoch": 0.041987615163872526, "grad_norm": 3.2189931082399874, "learning_rate": 9.995167524164508e-06, "loss": 0.1663, "step": 278 }, { "epoch": 0.042138649750792935, "grad_norm": 3.702439782355108, "learning_rate": 9.995132701274296e-06, "loss": 0.1613, "step": 279 }, { "epoch": 0.04228968433771334, "grad_norm": 2.9786938448939715, "learning_rate": 9.99509775342809e-06, "loss": 0.1379, "step": 280 }, { "epoch": 0.04244071892463374, "grad_norm": 2.9957414993361997, "learning_rate": 9.995062680626767e-06, "loss": 0.1372, "step": 281 }, { "epoch": 0.04259175351155415, "grad_norm": 3.275380494215229, "learning_rate": 9.995027482871202e-06, "loss": 0.1592, "step": 282 }, { "epoch": 0.04274278809847455, "grad_norm": 4.536255471054455, "learning_rate": 9.994992160162278e-06, "loss": 0.1976, "step": 283 }, { "epoch": 0.042893822685394954, "grad_norm": 5.214731402662311, "learning_rate": 9.994956712500876e-06, "loss": 0.2313, "step": 284 }, { "epoch": 0.043044857272315364, "grad_norm": 3.5294122240560983, "learning_rate": 9.994921139887881e-06, "loss": 0.1489, "step": 285 }, { "epoch": 0.043195891859235766, "grad_norm": 4.646016274713446, "learning_rate": 9.994885442324188e-06, "loss": 0.1598, "step": 286 }, { "epoch": 0.04334692644615617, "grad_norm": 2.8324967991377186, "learning_rate": 9.994849619810688e-06, "loss": 0.1628, "step": 287 }, { "epoch": 0.04349796103307658, "grad_norm": 2.819186075524487, "learning_rate": 9.994813672348275e-06, "loss": 0.1231, "step": 288 }, { "epoch": 0.04364899561999698, "grad_norm": 3.9794036327507643, "learning_rate": 9.994777599937851e-06, "loss": 0.1651, "step": 289 }, { "epoch": 0.04380003020691738, "grad_norm": 4.163183709295682, "learning_rate": 9.994741402580316e-06, "loss": 0.1572, "step": 290 }, { "epoch": 0.04395106479383779, "grad_norm": 3.7742165721128256, "learning_rate": 9.994705080276578e-06, "loss": 0.1812, "step": 291 }, { "epoch": 0.044102099380758195, "grad_norm": 3.068512738029114, "learning_rate": 9.994668633027544e-06, "loss": 0.1416, "step": 292 }, { "epoch": 0.0442531339676786, "grad_norm": 3.5838045797818956, "learning_rate": 9.994632060834127e-06, "loss": 0.1623, "step": 293 }, { "epoch": 0.044404168554599006, "grad_norm": 4.150682833918057, "learning_rate": 9.994595363697239e-06, "loss": 0.1478, "step": 294 }, { "epoch": 0.04455520314151941, "grad_norm": 3.3530389245121386, "learning_rate": 9.994558541617802e-06, "loss": 0.1515, "step": 295 }, { "epoch": 0.04470623772843981, "grad_norm": 4.072437650256459, "learning_rate": 9.994521594596734e-06, "loss": 0.1835, "step": 296 }, { "epoch": 0.04485727231536022, "grad_norm": 5.5055579761223745, "learning_rate": 9.99448452263496e-06, "loss": 0.1741, "step": 297 }, { "epoch": 0.04500830690228062, "grad_norm": 3.39427401552872, "learning_rate": 9.994447325733409e-06, "loss": 0.1666, "step": 298 }, { "epoch": 0.045159341489201026, "grad_norm": 5.210276357142116, "learning_rate": 9.994410003893009e-06, "loss": 0.168, "step": 299 }, { "epoch": 0.045310376076121435, "grad_norm": 3.105925433217789, "learning_rate": 9.994372557114695e-06, "loss": 0.1645, "step": 300 }, { "epoch": 0.04546141066304184, "grad_norm": 3.493710108218223, "learning_rate": 9.994334985399403e-06, "loss": 0.1738, "step": 301 }, { "epoch": 0.04561244524996224, "grad_norm": 2.9860991624511226, "learning_rate": 9.994297288748075e-06, "loss": 0.1433, "step": 302 }, { "epoch": 0.04576347983688265, "grad_norm": 3.225914289997123, "learning_rate": 9.994259467161652e-06, "loss": 0.1413, "step": 303 }, { "epoch": 0.04591451442380305, "grad_norm": 2.9885658477905124, "learning_rate": 9.99422152064108e-06, "loss": 0.1408, "step": 304 }, { "epoch": 0.046065549010723454, "grad_norm": 3.4969865497174997, "learning_rate": 9.99418344918731e-06, "loss": 0.1485, "step": 305 }, { "epoch": 0.046216583597643864, "grad_norm": 3.233354818644672, "learning_rate": 9.994145252801292e-06, "loss": 0.1277, "step": 306 }, { "epoch": 0.046367618184564266, "grad_norm": 2.959647058037847, "learning_rate": 9.994106931483981e-06, "loss": 0.1468, "step": 307 }, { "epoch": 0.04651865277148467, "grad_norm": 3.6756719702610523, "learning_rate": 9.994068485236341e-06, "loss": 0.1775, "step": 308 }, { "epoch": 0.04666968735840508, "grad_norm": 4.3341785015364005, "learning_rate": 9.994029914059329e-06, "loss": 0.1634, "step": 309 }, { "epoch": 0.04682072194532548, "grad_norm": 4.393493653898047, "learning_rate": 9.993991217953912e-06, "loss": 0.166, "step": 310 }, { "epoch": 0.04697175653224588, "grad_norm": 2.8213660304599353, "learning_rate": 9.993952396921054e-06, "loss": 0.133, "step": 311 }, { "epoch": 0.04712279111916629, "grad_norm": 2.846660224381184, "learning_rate": 9.993913450961732e-06, "loss": 0.1434, "step": 312 }, { "epoch": 0.047273825706086695, "grad_norm": 4.0987063280442255, "learning_rate": 9.993874380076916e-06, "loss": 0.1704, "step": 313 }, { "epoch": 0.0474248602930071, "grad_norm": 6.001039228662546, "learning_rate": 9.993835184267584e-06, "loss": 0.1879, "step": 314 }, { "epoch": 0.047575894879927506, "grad_norm": 3.2717563957719324, "learning_rate": 9.993795863534718e-06, "loss": 0.1409, "step": 315 }, { "epoch": 0.04772692946684791, "grad_norm": 3.7363193551537943, "learning_rate": 9.9937564178793e-06, "loss": 0.1681, "step": 316 }, { "epoch": 0.04787796405376831, "grad_norm": 4.036477482273475, "learning_rate": 9.99371684730232e-06, "loss": 0.1749, "step": 317 }, { "epoch": 0.04802899864068872, "grad_norm": 3.3292943905938657, "learning_rate": 9.993677151804765e-06, "loss": 0.1619, "step": 318 }, { "epoch": 0.04818003322760912, "grad_norm": 2.895209388493885, "learning_rate": 9.993637331387627e-06, "loss": 0.1373, "step": 319 }, { "epoch": 0.048331067814529526, "grad_norm": 3.734882629150427, "learning_rate": 9.993597386051905e-06, "loss": 0.1672, "step": 320 }, { "epoch": 0.048482102401449935, "grad_norm": 3.1572279633833076, "learning_rate": 9.993557315798597e-06, "loss": 0.13, "step": 321 }, { "epoch": 0.04863313698837034, "grad_norm": 3.367547736366355, "learning_rate": 9.993517120628705e-06, "loss": 0.1682, "step": 322 }, { "epoch": 0.04878417157529074, "grad_norm": 3.7500860247312358, "learning_rate": 9.993476800543233e-06, "loss": 0.163, "step": 323 }, { "epoch": 0.04893520616221115, "grad_norm": 4.020590117853299, "learning_rate": 9.993436355543193e-06, "loss": 0.1682, "step": 324 }, { "epoch": 0.04908624074913155, "grad_norm": 3.5405040177109366, "learning_rate": 9.993395785629596e-06, "loss": 0.1696, "step": 325 }, { "epoch": 0.049237275336051954, "grad_norm": 4.299333114914034, "learning_rate": 9.993355090803457e-06, "loss": 0.1709, "step": 326 }, { "epoch": 0.049388309922972363, "grad_norm": 3.333749620485609, "learning_rate": 9.993314271065791e-06, "loss": 0.1417, "step": 327 }, { "epoch": 0.049539344509892766, "grad_norm": 5.941372006360486, "learning_rate": 9.99327332641762e-06, "loss": 0.1876, "step": 328 }, { "epoch": 0.04969037909681317, "grad_norm": 3.3069435239122256, "learning_rate": 9.993232256859972e-06, "loss": 0.1633, "step": 329 }, { "epoch": 0.04984141368373358, "grad_norm": 3.5608794887856097, "learning_rate": 9.99319106239387e-06, "loss": 0.1515, "step": 330 }, { "epoch": 0.04999244827065398, "grad_norm": 3.401214085212386, "learning_rate": 9.993149743020347e-06, "loss": 0.1593, "step": 331 }, { "epoch": 0.05014348285757438, "grad_norm": 3.0027231633132296, "learning_rate": 9.993108298740438e-06, "loss": 0.1335, "step": 332 }, { "epoch": 0.05029451744449479, "grad_norm": 3.5551613853136512, "learning_rate": 9.993066729555174e-06, "loss": 0.1724, "step": 333 }, { "epoch": 0.050445552031415195, "grad_norm": 4.726759359362944, "learning_rate": 9.9930250354656e-06, "loss": 0.2099, "step": 334 }, { "epoch": 0.0505965866183356, "grad_norm": 4.240254602978863, "learning_rate": 9.992983216472759e-06, "loss": 0.1733, "step": 335 }, { "epoch": 0.050747621205256006, "grad_norm": 3.6603692017960885, "learning_rate": 9.992941272577694e-06, "loss": 0.1811, "step": 336 }, { "epoch": 0.05089865579217641, "grad_norm": 3.864622844309328, "learning_rate": 9.992899203781458e-06, "loss": 0.1655, "step": 337 }, { "epoch": 0.05104969037909681, "grad_norm": 4.299468672671151, "learning_rate": 9.992857010085099e-06, "loss": 0.2022, "step": 338 }, { "epoch": 0.05120072496601722, "grad_norm": 2.9621267429993834, "learning_rate": 9.992814691489677e-06, "loss": 0.1449, "step": 339 }, { "epoch": 0.05135175955293762, "grad_norm": 3.594579970185378, "learning_rate": 9.992772247996248e-06, "loss": 0.1367, "step": 340 }, { "epoch": 0.051502794139858026, "grad_norm": 4.204918616359229, "learning_rate": 9.992729679605873e-06, "loss": 0.1731, "step": 341 }, { "epoch": 0.051653828726778435, "grad_norm": 4.019134801743246, "learning_rate": 9.992686986319618e-06, "loss": 0.1683, "step": 342 }, { "epoch": 0.05180486331369884, "grad_norm": 3.701512755241209, "learning_rate": 9.992644168138553e-06, "loss": 0.1653, "step": 343 }, { "epoch": 0.05195589790061924, "grad_norm": 4.649791317007214, "learning_rate": 9.992601225063745e-06, "loss": 0.1632, "step": 344 }, { "epoch": 0.05210693248753965, "grad_norm": 4.352648475677026, "learning_rate": 9.992558157096272e-06, "loss": 0.1799, "step": 345 }, { "epoch": 0.05225796707446005, "grad_norm": 3.132472527572296, "learning_rate": 9.992514964237209e-06, "loss": 0.1531, "step": 346 }, { "epoch": 0.052409001661380454, "grad_norm": 3.347166272669838, "learning_rate": 9.992471646487636e-06, "loss": 0.1438, "step": 347 }, { "epoch": 0.05256003624830086, "grad_norm": 3.2419950180811337, "learning_rate": 9.99242820384864e-06, "loss": 0.1215, "step": 348 }, { "epoch": 0.052711070835221266, "grad_norm": 3.9360675473948707, "learning_rate": 9.992384636321307e-06, "loss": 0.162, "step": 349 }, { "epoch": 0.05286210542214167, "grad_norm": 4.5498332723415, "learning_rate": 9.992340943906722e-06, "loss": 0.1872, "step": 350 }, { "epoch": 0.05301314000906208, "grad_norm": 3.1150761941980467, "learning_rate": 9.992297126605983e-06, "loss": 0.163, "step": 351 }, { "epoch": 0.05316417459598248, "grad_norm": 3.4092004207483435, "learning_rate": 9.992253184420185e-06, "loss": 0.1683, "step": 352 }, { "epoch": 0.05331520918290288, "grad_norm": 3.147208249748109, "learning_rate": 9.992209117350427e-06, "loss": 0.1613, "step": 353 }, { "epoch": 0.05346624376982329, "grad_norm": 5.5025715357951, "learning_rate": 9.99216492539781e-06, "loss": 0.191, "step": 354 }, { "epoch": 0.053617278356743694, "grad_norm": 2.98826835456538, "learning_rate": 9.992120608563442e-06, "loss": 0.1744, "step": 355 }, { "epoch": 0.0537683129436641, "grad_norm": 2.8622408208989953, "learning_rate": 9.99207616684843e-06, "loss": 0.1516, "step": 356 }, { "epoch": 0.053919347530584506, "grad_norm": 3.2781138509762946, "learning_rate": 9.992031600253886e-06, "loss": 0.1556, "step": 357 }, { "epoch": 0.05407038211750491, "grad_norm": 3.4769593596272763, "learning_rate": 9.991986908780924e-06, "loss": 0.1689, "step": 358 }, { "epoch": 0.05422141670442531, "grad_norm": 4.245962675198921, "learning_rate": 9.991942092430664e-06, "loss": 0.1498, "step": 359 }, { "epoch": 0.05437245129134572, "grad_norm": 4.5534885754522305, "learning_rate": 9.991897151204226e-06, "loss": 0.1533, "step": 360 }, { "epoch": 0.05452348587826612, "grad_norm": 3.1805650581679963, "learning_rate": 9.991852085102734e-06, "loss": 0.131, "step": 361 }, { "epoch": 0.054674520465186525, "grad_norm": 2.4601288241254555, "learning_rate": 9.991806894127316e-06, "loss": 0.1314, "step": 362 }, { "epoch": 0.054825555052106935, "grad_norm": 4.1450417695132336, "learning_rate": 9.991761578279102e-06, "loss": 0.1737, "step": 363 }, { "epoch": 0.05497658963902734, "grad_norm": 3.2890393535391365, "learning_rate": 9.991716137559225e-06, "loss": 0.1399, "step": 364 }, { "epoch": 0.05512762422594774, "grad_norm": 3.743395511770964, "learning_rate": 9.991670571968824e-06, "loss": 0.183, "step": 365 }, { "epoch": 0.05527865881286815, "grad_norm": 3.8105454850440443, "learning_rate": 9.991624881509035e-06, "loss": 0.1708, "step": 366 }, { "epoch": 0.05542969339978855, "grad_norm": 3.3126587471105937, "learning_rate": 9.991579066181004e-06, "loss": 0.1315, "step": 367 }, { "epoch": 0.055580727986708954, "grad_norm": 7.619888700439525, "learning_rate": 9.991533125985877e-06, "loss": 0.1907, "step": 368 }, { "epoch": 0.05573176257362936, "grad_norm": 4.223028793629431, "learning_rate": 9.991487060924802e-06, "loss": 0.1972, "step": 369 }, { "epoch": 0.055882797160549766, "grad_norm": 2.911777533126204, "learning_rate": 9.991440870998931e-06, "loss": 0.1437, "step": 370 }, { "epoch": 0.05603383174747017, "grad_norm": 4.441007097650985, "learning_rate": 9.991394556209422e-06, "loss": 0.1952, "step": 371 }, { "epoch": 0.05618486633439058, "grad_norm": 3.379950609777028, "learning_rate": 9.99134811655743e-06, "loss": 0.1472, "step": 372 }, { "epoch": 0.05633590092131098, "grad_norm": 3.7070907344244564, "learning_rate": 9.991301552044121e-06, "loss": 0.1807, "step": 373 }, { "epoch": 0.05648693550823138, "grad_norm": 3.541801978271261, "learning_rate": 9.991254862670656e-06, "loss": 0.1414, "step": 374 }, { "epoch": 0.05663797009515179, "grad_norm": 3.2819293626036137, "learning_rate": 9.991208048438205e-06, "loss": 0.1605, "step": 375 }, { "epoch": 0.056789004682072194, "grad_norm": 3.772721004688516, "learning_rate": 9.991161109347937e-06, "loss": 0.1852, "step": 376 }, { "epoch": 0.0569400392689926, "grad_norm": 3.0372637545389614, "learning_rate": 9.991114045401029e-06, "loss": 0.1453, "step": 377 }, { "epoch": 0.057091073855913006, "grad_norm": 3.314962407505794, "learning_rate": 9.991066856598658e-06, "loss": 0.1654, "step": 378 }, { "epoch": 0.05724210844283341, "grad_norm": 3.865997443941885, "learning_rate": 9.991019542942002e-06, "loss": 0.1777, "step": 379 }, { "epoch": 0.05739314302975381, "grad_norm": 4.578356584742831, "learning_rate": 9.990972104432247e-06, "loss": 0.1755, "step": 380 }, { "epoch": 0.05754417761667422, "grad_norm": 5.133820630069038, "learning_rate": 9.99092454107058e-06, "loss": 0.1969, "step": 381 }, { "epoch": 0.05769521220359462, "grad_norm": 4.359878563475751, "learning_rate": 9.990876852858186e-06, "loss": 0.1911, "step": 382 }, { "epoch": 0.057846246790515025, "grad_norm": 3.261962276742379, "learning_rate": 9.990829039796264e-06, "loss": 0.1565, "step": 383 }, { "epoch": 0.057997281377435435, "grad_norm": 4.314052141455931, "learning_rate": 9.990781101886008e-06, "loss": 0.1834, "step": 384 }, { "epoch": 0.05814831596435584, "grad_norm": 3.6251770647369024, "learning_rate": 9.990733039128615e-06, "loss": 0.1584, "step": 385 }, { "epoch": 0.05829935055127624, "grad_norm": 3.1080444128098037, "learning_rate": 9.99068485152529e-06, "loss": 0.1554, "step": 386 }, { "epoch": 0.05845038513819665, "grad_norm": 3.089339567315305, "learning_rate": 9.990636539077239e-06, "loss": 0.1532, "step": 387 }, { "epoch": 0.05860141972511705, "grad_norm": 3.4468492268875517, "learning_rate": 9.990588101785668e-06, "loss": 0.1835, "step": 388 }, { "epoch": 0.058752454312037454, "grad_norm": 2.870893349352697, "learning_rate": 9.99053953965179e-06, "loss": 0.1519, "step": 389 }, { "epoch": 0.05890348889895786, "grad_norm": 3.3846941366941663, "learning_rate": 9.990490852676819e-06, "loss": 0.1485, "step": 390 }, { "epoch": 0.059054523485878266, "grad_norm": 2.8321085744073877, "learning_rate": 9.990442040861974e-06, "loss": 0.1507, "step": 391 }, { "epoch": 0.05920555807279867, "grad_norm": 4.050581098946334, "learning_rate": 9.990393104208476e-06, "loss": 0.1678, "step": 392 }, { "epoch": 0.05935659265971908, "grad_norm": 3.380817582175604, "learning_rate": 9.990344042717547e-06, "loss": 0.1623, "step": 393 }, { "epoch": 0.05950762724663948, "grad_norm": 3.94980423214413, "learning_rate": 9.990294856390419e-06, "loss": 0.1687, "step": 394 }, { "epoch": 0.05965866183355988, "grad_norm": 3.893282626631903, "learning_rate": 9.990245545228316e-06, "loss": 0.182, "step": 395 }, { "epoch": 0.05980969642048029, "grad_norm": 5.202399446364915, "learning_rate": 9.990196109232477e-06, "loss": 0.1562, "step": 396 }, { "epoch": 0.059960731007400694, "grad_norm": 3.3559619318695035, "learning_rate": 9.990146548404138e-06, "loss": 0.1634, "step": 397 }, { "epoch": 0.0601117655943211, "grad_norm": 3.7808566381125917, "learning_rate": 9.990096862744536e-06, "loss": 0.1627, "step": 398 }, { "epoch": 0.060262800181241506, "grad_norm": 3.5107076375790185, "learning_rate": 9.990047052254917e-06, "loss": 0.1465, "step": 399 }, { "epoch": 0.06041383476816191, "grad_norm": 4.363856890165701, "learning_rate": 9.989997116936523e-06, "loss": 0.1909, "step": 400 }, { "epoch": 0.06056486935508231, "grad_norm": 3.3689840526755885, "learning_rate": 9.989947056790607e-06, "loss": 0.1639, "step": 401 }, { "epoch": 0.06071590394200272, "grad_norm": 3.6570053981490984, "learning_rate": 9.98989687181842e-06, "loss": 0.1794, "step": 402 }, { "epoch": 0.06086693852892312, "grad_norm": 4.2513762721533395, "learning_rate": 9.989846562021217e-06, "loss": 0.1754, "step": 403 }, { "epoch": 0.061017973115843525, "grad_norm": 4.02252064401009, "learning_rate": 9.989796127400258e-06, "loss": 0.182, "step": 404 }, { "epoch": 0.061169007702763935, "grad_norm": 3.690658919584816, "learning_rate": 9.989745567956801e-06, "loss": 0.1857, "step": 405 }, { "epoch": 0.06132004228968434, "grad_norm": 6.823644592702338, "learning_rate": 9.989694883692115e-06, "loss": 0.1529, "step": 406 }, { "epoch": 0.06147107687660474, "grad_norm": 3.5724434655891395, "learning_rate": 9.989644074607467e-06, "loss": 0.162, "step": 407 }, { "epoch": 0.06162211146352515, "grad_norm": 3.044571162282378, "learning_rate": 9.989593140704125e-06, "loss": 0.1687, "step": 408 }, { "epoch": 0.06177314605044555, "grad_norm": 2.5573480731193863, "learning_rate": 9.989542081983368e-06, "loss": 0.1327, "step": 409 }, { "epoch": 0.061924180637365954, "grad_norm": 3.395629283903432, "learning_rate": 9.989490898446467e-06, "loss": 0.1804, "step": 410 }, { "epoch": 0.06207521522428636, "grad_norm": 2.8870750313083606, "learning_rate": 9.98943959009471e-06, "loss": 0.1339, "step": 411 }, { "epoch": 0.062226249811206766, "grad_norm": 3.649455070921834, "learning_rate": 9.989388156929374e-06, "loss": 0.1514, "step": 412 }, { "epoch": 0.06237728439812717, "grad_norm": 4.220168469858137, "learning_rate": 9.989336598951748e-06, "loss": 0.1723, "step": 413 }, { "epoch": 0.06252831898504757, "grad_norm": 3.5853453832194955, "learning_rate": 9.989284916163122e-06, "loss": 0.1638, "step": 414 }, { "epoch": 0.06267935357196798, "grad_norm": 2.788179536138111, "learning_rate": 9.98923310856479e-06, "loss": 0.1469, "step": 415 }, { "epoch": 0.06283038815888839, "grad_norm": 2.9941663214257783, "learning_rate": 9.989181176158046e-06, "loss": 0.1544, "step": 416 }, { "epoch": 0.06298142274580878, "grad_norm": 3.2570708689541568, "learning_rate": 9.98912911894419e-06, "loss": 0.147, "step": 417 }, { "epoch": 0.0631324573327292, "grad_norm": 3.192454533030965, "learning_rate": 9.989076936924524e-06, "loss": 0.1501, "step": 418 }, { "epoch": 0.0632834919196496, "grad_norm": 4.238059087073455, "learning_rate": 9.989024630100352e-06, "loss": 0.1818, "step": 419 }, { "epoch": 0.06343452650657, "grad_norm": 2.7454838125314853, "learning_rate": 9.988972198472986e-06, "loss": 0.1409, "step": 420 }, { "epoch": 0.06358556109349041, "grad_norm": 4.958090400938061, "learning_rate": 9.988919642043736e-06, "loss": 0.1783, "step": 421 }, { "epoch": 0.06373659568041082, "grad_norm": 3.2001628867058662, "learning_rate": 9.988866960813915e-06, "loss": 0.1558, "step": 422 }, { "epoch": 0.06388763026733121, "grad_norm": 3.9448941584672563, "learning_rate": 9.98881415478484e-06, "loss": 0.1882, "step": 423 }, { "epoch": 0.06403866485425162, "grad_norm": 2.9840747970520223, "learning_rate": 9.988761223957837e-06, "loss": 0.1416, "step": 424 }, { "epoch": 0.06418969944117203, "grad_norm": 3.8890111909724254, "learning_rate": 9.988708168334227e-06, "loss": 0.1873, "step": 425 }, { "epoch": 0.06434073402809243, "grad_norm": 4.054387408525083, "learning_rate": 9.988654987915337e-06, "loss": 0.1287, "step": 426 }, { "epoch": 0.06449176861501284, "grad_norm": 4.507138101322779, "learning_rate": 9.988601682702496e-06, "loss": 0.1665, "step": 427 }, { "epoch": 0.06464280320193325, "grad_norm": 2.6948624426540033, "learning_rate": 9.98854825269704e-06, "loss": 0.1527, "step": 428 }, { "epoch": 0.06479383778885364, "grad_norm": 3.83896456019734, "learning_rate": 9.988494697900308e-06, "loss": 0.151, "step": 429 }, { "epoch": 0.06494487237577405, "grad_norm": 3.2879683743299806, "learning_rate": 9.988441018313633e-06, "loss": 0.1738, "step": 430 }, { "epoch": 0.06509590696269446, "grad_norm": 3.116943715782655, "learning_rate": 9.988387213938363e-06, "loss": 0.1473, "step": 431 }, { "epoch": 0.06524694154961486, "grad_norm": 2.5833291947920394, "learning_rate": 9.988333284775842e-06, "loss": 0.1322, "step": 432 }, { "epoch": 0.06539797613653527, "grad_norm": 2.995660761072866, "learning_rate": 9.988279230827418e-06, "loss": 0.1535, "step": 433 }, { "epoch": 0.06554901072345568, "grad_norm": 3.0365364352507456, "learning_rate": 9.988225052094446e-06, "loss": 0.1568, "step": 434 }, { "epoch": 0.06570004531037607, "grad_norm": 4.688945115567106, "learning_rate": 9.98817074857828e-06, "loss": 0.1478, "step": 435 }, { "epoch": 0.06585107989729648, "grad_norm": 3.356779058624426, "learning_rate": 9.988116320280276e-06, "loss": 0.1775, "step": 436 }, { "epoch": 0.06600211448421689, "grad_norm": 3.1576333337695344, "learning_rate": 9.9880617672018e-06, "loss": 0.1654, "step": 437 }, { "epoch": 0.06615314907113728, "grad_norm": 2.323738327223142, "learning_rate": 9.988007089344214e-06, "loss": 0.1028, "step": 438 }, { "epoch": 0.0663041836580577, "grad_norm": 2.5402349415940777, "learning_rate": 9.987952286708886e-06, "loss": 0.1218, "step": 439 }, { "epoch": 0.0664552182449781, "grad_norm": 3.6891500336339913, "learning_rate": 9.987897359297187e-06, "loss": 0.1828, "step": 440 }, { "epoch": 0.0666062528318985, "grad_norm": 4.227161377108587, "learning_rate": 9.987842307110492e-06, "loss": 0.1733, "step": 441 }, { "epoch": 0.06675728741881891, "grad_norm": 3.238928137207802, "learning_rate": 9.987787130150176e-06, "loss": 0.1438, "step": 442 }, { "epoch": 0.06690832200573932, "grad_norm": 2.5884622263408157, "learning_rate": 9.987731828417622e-06, "loss": 0.1303, "step": 443 }, { "epoch": 0.06705935659265971, "grad_norm": 3.492699490449795, "learning_rate": 9.98767640191421e-06, "loss": 0.1666, "step": 444 }, { "epoch": 0.06721039117958012, "grad_norm": 3.706168424026975, "learning_rate": 9.98762085064133e-06, "loss": 0.1875, "step": 445 }, { "epoch": 0.06736142576650053, "grad_norm": 4.486910189290135, "learning_rate": 9.987565174600372e-06, "loss": 0.1932, "step": 446 }, { "epoch": 0.06751246035342093, "grad_norm": 3.232799351827479, "learning_rate": 9.987509373792726e-06, "loss": 0.1583, "step": 447 }, { "epoch": 0.06766349494034134, "grad_norm": 3.3625662396498863, "learning_rate": 9.987453448219787e-06, "loss": 0.1566, "step": 448 }, { "epoch": 0.06781452952726175, "grad_norm": 3.9859890650685794, "learning_rate": 9.987397397882958e-06, "loss": 0.1926, "step": 449 }, { "epoch": 0.06796556411418214, "grad_norm": 4.511828381827101, "learning_rate": 9.987341222783637e-06, "loss": 0.1875, "step": 450 }, { "epoch": 0.06811659870110255, "grad_norm": 5.478576423516532, "learning_rate": 9.987284922923233e-06, "loss": 0.1716, "step": 451 }, { "epoch": 0.06826763328802296, "grad_norm": 4.095367996088305, "learning_rate": 9.987228498303152e-06, "loss": 0.1681, "step": 452 }, { "epoch": 0.06841866787494336, "grad_norm": 2.966574695498381, "learning_rate": 9.987171948924808e-06, "loss": 0.1481, "step": 453 }, { "epoch": 0.06856970246186377, "grad_norm": 4.17573200625637, "learning_rate": 9.987115274789612e-06, "loss": 0.1779, "step": 454 }, { "epoch": 0.06872073704878418, "grad_norm": 3.4420064342006826, "learning_rate": 9.987058475898983e-06, "loss": 0.1743, "step": 455 }, { "epoch": 0.06887177163570457, "grad_norm": 4.131202136397824, "learning_rate": 9.987001552254344e-06, "loss": 0.155, "step": 456 }, { "epoch": 0.06902280622262498, "grad_norm": 3.5993657912913237, "learning_rate": 9.986944503857116e-06, "loss": 0.1536, "step": 457 }, { "epoch": 0.06917384080954539, "grad_norm": 3.494647730644006, "learning_rate": 9.986887330708727e-06, "loss": 0.182, "step": 458 }, { "epoch": 0.06932487539646578, "grad_norm": 4.5226726706862435, "learning_rate": 9.986830032810608e-06, "loss": 0.1572, "step": 459 }, { "epoch": 0.0694759099833862, "grad_norm": 7.267477319979273, "learning_rate": 9.986772610164192e-06, "loss": 0.1811, "step": 460 }, { "epoch": 0.0696269445703066, "grad_norm": 3.4485185721452236, "learning_rate": 9.986715062770918e-06, "loss": 0.1769, "step": 461 }, { "epoch": 0.069777979157227, "grad_norm": 3.027364534624024, "learning_rate": 9.98665739063222e-06, "loss": 0.1561, "step": 462 }, { "epoch": 0.06992901374414741, "grad_norm": 3.226600271546299, "learning_rate": 9.986599593749544e-06, "loss": 0.1595, "step": 463 }, { "epoch": 0.07008004833106782, "grad_norm": 4.528752363671887, "learning_rate": 9.986541672124337e-06, "loss": 0.2035, "step": 464 }, { "epoch": 0.07023108291798821, "grad_norm": 3.204816820226396, "learning_rate": 9.986483625758045e-06, "loss": 0.1579, "step": 465 }, { "epoch": 0.07038211750490862, "grad_norm": 3.4343978225783767, "learning_rate": 9.986425454652124e-06, "loss": 0.1786, "step": 466 }, { "epoch": 0.07053315209182903, "grad_norm": 4.242469312891193, "learning_rate": 9.986367158808023e-06, "loss": 0.2091, "step": 467 }, { "epoch": 0.07068418667874943, "grad_norm": 3.9596149933856326, "learning_rate": 9.986308738227208e-06, "loss": 0.1515, "step": 468 }, { "epoch": 0.07083522126566984, "grad_norm": 3.2830769103356063, "learning_rate": 9.986250192911133e-06, "loss": 0.1633, "step": 469 }, { "epoch": 0.07098625585259025, "grad_norm": 3.7014782007300506, "learning_rate": 9.986191522861268e-06, "loss": 0.1582, "step": 470 }, { "epoch": 0.07113729043951064, "grad_norm": 4.387727291310029, "learning_rate": 9.986132728079079e-06, "loss": 0.1851, "step": 471 }, { "epoch": 0.07128832502643105, "grad_norm": 3.6330048537865007, "learning_rate": 9.986073808566036e-06, "loss": 0.1793, "step": 472 }, { "epoch": 0.07143935961335146, "grad_norm": 3.086911398479002, "learning_rate": 9.986014764323613e-06, "loss": 0.1611, "step": 473 }, { "epoch": 0.07159039420027186, "grad_norm": 4.469862677592981, "learning_rate": 9.985955595353288e-06, "loss": 0.1897, "step": 474 }, { "epoch": 0.07174142878719227, "grad_norm": 2.723721481663779, "learning_rate": 9.98589630165654e-06, "loss": 0.1374, "step": 475 }, { "epoch": 0.07189246337411268, "grad_norm": 5.234594437288907, "learning_rate": 9.985836883234851e-06, "loss": 0.1735, "step": 476 }, { "epoch": 0.07204349796103307, "grad_norm": 3.282473983443536, "learning_rate": 9.98577734008971e-06, "loss": 0.1723, "step": 477 }, { "epoch": 0.07219453254795348, "grad_norm": 3.233986754795171, "learning_rate": 9.985717672222608e-06, "loss": 0.1651, "step": 478 }, { "epoch": 0.07234556713487389, "grad_norm": 3.923735755935844, "learning_rate": 9.985657879635033e-06, "loss": 0.1947, "step": 479 }, { "epoch": 0.07249660172179428, "grad_norm": 4.320528919796565, "learning_rate": 9.985597962328483e-06, "loss": 0.1825, "step": 480 }, { "epoch": 0.0726476363087147, "grad_norm": 3.7410208258731226, "learning_rate": 9.98553792030446e-06, "loss": 0.1791, "step": 481 }, { "epoch": 0.0727986708956351, "grad_norm": 3.6553951154309807, "learning_rate": 9.985477753564459e-06, "loss": 0.1895, "step": 482 }, { "epoch": 0.0729497054825555, "grad_norm": 3.7251559571333925, "learning_rate": 9.985417462109992e-06, "loss": 0.1869, "step": 483 }, { "epoch": 0.07310074006947591, "grad_norm": 3.5086302337184256, "learning_rate": 9.985357045942562e-06, "loss": 0.1838, "step": 484 }, { "epoch": 0.07325177465639632, "grad_norm": 3.3017709691380035, "learning_rate": 9.985296505063684e-06, "loss": 0.1835, "step": 485 }, { "epoch": 0.07340280924331671, "grad_norm": 3.2583208121696674, "learning_rate": 9.98523583947487e-06, "loss": 0.161, "step": 486 }, { "epoch": 0.07355384383023712, "grad_norm": 3.552849246955639, "learning_rate": 9.98517504917764e-06, "loss": 0.1618, "step": 487 }, { "epoch": 0.07370487841715753, "grad_norm": 4.1210467018405215, "learning_rate": 9.985114134173512e-06, "loss": 0.1742, "step": 488 }, { "epoch": 0.07385591300407793, "grad_norm": 3.2516954698906515, "learning_rate": 9.985053094464013e-06, "loss": 0.1585, "step": 489 }, { "epoch": 0.07400694759099834, "grad_norm": 3.1980631029873643, "learning_rate": 9.984991930050666e-06, "loss": 0.1515, "step": 490 }, { "epoch": 0.07415798217791875, "grad_norm": 3.538899331361703, "learning_rate": 9.984930640935003e-06, "loss": 0.1737, "step": 491 }, { "epoch": 0.07430901676483914, "grad_norm": 3.052811208165519, "learning_rate": 9.98486922711856e-06, "loss": 0.1488, "step": 492 }, { "epoch": 0.07446005135175955, "grad_norm": 4.017253113318992, "learning_rate": 9.984807688602867e-06, "loss": 0.1937, "step": 493 }, { "epoch": 0.07461108593867996, "grad_norm": 3.690748321928705, "learning_rate": 9.984746025389469e-06, "loss": 0.1834, "step": 494 }, { "epoch": 0.07476212052560036, "grad_norm": 2.7703048956513263, "learning_rate": 9.984684237479907e-06, "loss": 0.1534, "step": 495 }, { "epoch": 0.07491315511252077, "grad_norm": 2.9019835294476213, "learning_rate": 9.984622324875726e-06, "loss": 0.1524, "step": 496 }, { "epoch": 0.07506418969944118, "grad_norm": 3.239513284587166, "learning_rate": 9.984560287578473e-06, "loss": 0.1582, "step": 497 }, { "epoch": 0.07521522428636157, "grad_norm": 3.3246875574816763, "learning_rate": 9.984498125589704e-06, "loss": 0.1573, "step": 498 }, { "epoch": 0.07536625887328198, "grad_norm": 3.0916293949547167, "learning_rate": 9.984435838910971e-06, "loss": 0.1743, "step": 499 }, { "epoch": 0.07551729346020239, "grad_norm": 3.6697373245310985, "learning_rate": 9.984373427543834e-06, "loss": 0.165, "step": 500 }, { "epoch": 0.07566832804712278, "grad_norm": 3.2122753701838582, "learning_rate": 9.984310891489854e-06, "loss": 0.1263, "step": 501 }, { "epoch": 0.0758193626340432, "grad_norm": 3.0509048566469175, "learning_rate": 9.984248230750592e-06, "loss": 0.1345, "step": 502 }, { "epoch": 0.0759703972209636, "grad_norm": 3.517833010902483, "learning_rate": 9.984185445327622e-06, "loss": 0.181, "step": 503 }, { "epoch": 0.076121431807884, "grad_norm": 3.2207625493667793, "learning_rate": 9.984122535222506e-06, "loss": 0.1537, "step": 504 }, { "epoch": 0.07627246639480441, "grad_norm": 3.19342473826223, "learning_rate": 9.984059500436827e-06, "loss": 0.1697, "step": 505 }, { "epoch": 0.07642350098172482, "grad_norm": 3.4343600614680025, "learning_rate": 9.983996340972154e-06, "loss": 0.144, "step": 506 }, { "epoch": 0.07657453556864521, "grad_norm": 2.565972794026625, "learning_rate": 9.983933056830073e-06, "loss": 0.1307, "step": 507 }, { "epoch": 0.07672557015556562, "grad_norm": 2.9437125257594663, "learning_rate": 9.983869648012163e-06, "loss": 0.1586, "step": 508 }, { "epoch": 0.07687660474248603, "grad_norm": 4.781477152957989, "learning_rate": 9.983806114520012e-06, "loss": 0.1523, "step": 509 }, { "epoch": 0.07702763932940643, "grad_norm": 3.1217655413745753, "learning_rate": 9.98374245635521e-06, "loss": 0.1624, "step": 510 }, { "epoch": 0.07717867391632684, "grad_norm": 3.2995755920425194, "learning_rate": 9.983678673519346e-06, "loss": 0.1701, "step": 511 }, { "epoch": 0.07732970850324725, "grad_norm": 4.426703258342277, "learning_rate": 9.983614766014021e-06, "loss": 0.1899, "step": 512 }, { "epoch": 0.07748074309016764, "grad_norm": 3.7128832584262663, "learning_rate": 9.983550733840828e-06, "loss": 0.1967, "step": 513 }, { "epoch": 0.07763177767708805, "grad_norm": 3.2597791835939005, "learning_rate": 9.983486577001371e-06, "loss": 0.1623, "step": 514 }, { "epoch": 0.07778281226400846, "grad_norm": 3.2633728468221923, "learning_rate": 9.983422295497258e-06, "loss": 0.146, "step": 515 }, { "epoch": 0.07793384685092886, "grad_norm": 2.7317650860244265, "learning_rate": 9.983357889330094e-06, "loss": 0.1406, "step": 516 }, { "epoch": 0.07808488143784927, "grad_norm": 3.294174095366431, "learning_rate": 9.98329335850149e-06, "loss": 0.1862, "step": 517 }, { "epoch": 0.07823591602476967, "grad_norm": 2.776744898824888, "learning_rate": 9.98322870301306e-06, "loss": 0.1534, "step": 518 }, { "epoch": 0.07838695061169007, "grad_norm": 3.6411535797126615, "learning_rate": 9.983163922866423e-06, "loss": 0.1655, "step": 519 }, { "epoch": 0.07853798519861048, "grad_norm": 4.597817894233564, "learning_rate": 9.983099018063198e-06, "loss": 0.1835, "step": 520 }, { "epoch": 0.07868901978553089, "grad_norm": 5.644668349786358, "learning_rate": 9.98303398860501e-06, "loss": 0.1698, "step": 521 }, { "epoch": 0.07884005437245128, "grad_norm": 3.0278169507583175, "learning_rate": 9.982968834493485e-06, "loss": 0.1674, "step": 522 }, { "epoch": 0.0789910889593717, "grad_norm": 3.157947865142883, "learning_rate": 9.982903555730253e-06, "loss": 0.1563, "step": 523 }, { "epoch": 0.0791421235462921, "grad_norm": 2.5231340606310035, "learning_rate": 9.982838152316947e-06, "loss": 0.1361, "step": 524 }, { "epoch": 0.0792931581332125, "grad_norm": 2.817691022672766, "learning_rate": 9.982772624255204e-06, "loss": 0.1323, "step": 525 }, { "epoch": 0.07944419272013291, "grad_norm": 3.078735435297512, "learning_rate": 9.98270697154666e-06, "loss": 0.1637, "step": 526 }, { "epoch": 0.07959522730705332, "grad_norm": 7.148394811690192, "learning_rate": 9.982641194192963e-06, "loss": 0.1744, "step": 527 }, { "epoch": 0.07974626189397371, "grad_norm": 4.752136655758901, "learning_rate": 9.982575292195753e-06, "loss": 0.1908, "step": 528 }, { "epoch": 0.07989729648089412, "grad_norm": 3.2430926465394245, "learning_rate": 9.982509265556682e-06, "loss": 0.1858, "step": 529 }, { "epoch": 0.08004833106781453, "grad_norm": 3.8483619179291013, "learning_rate": 9.982443114277398e-06, "loss": 0.1761, "step": 530 }, { "epoch": 0.08019936565473493, "grad_norm": 3.0272536828928582, "learning_rate": 9.982376838359559e-06, "loss": 0.1557, "step": 531 }, { "epoch": 0.08035040024165534, "grad_norm": 3.8165230387600104, "learning_rate": 9.982310437804823e-06, "loss": 0.1824, "step": 532 }, { "epoch": 0.08050143482857575, "grad_norm": 3.342803471636902, "learning_rate": 9.98224391261485e-06, "loss": 0.1677, "step": 533 }, { "epoch": 0.08065246941549614, "grad_norm": 4.34182659226055, "learning_rate": 9.982177262791304e-06, "loss": 0.1933, "step": 534 }, { "epoch": 0.08080350400241655, "grad_norm": 2.8781620991228465, "learning_rate": 9.982110488335852e-06, "loss": 0.1658, "step": 535 }, { "epoch": 0.08095453858933696, "grad_norm": 3.8375366209728057, "learning_rate": 9.982043589250165e-06, "loss": 0.186, "step": 536 }, { "epoch": 0.08110557317625736, "grad_norm": 3.643845248504087, "learning_rate": 9.981976565535918e-06, "loss": 0.1769, "step": 537 }, { "epoch": 0.08125660776317777, "grad_norm": 2.4835264548084335, "learning_rate": 9.981909417194785e-06, "loss": 0.1535, "step": 538 }, { "epoch": 0.08140764235009817, "grad_norm": 2.9881465807811356, "learning_rate": 9.981842144228447e-06, "loss": 0.1479, "step": 539 }, { "epoch": 0.08155867693701857, "grad_norm": 4.0629777804652765, "learning_rate": 9.981774746638586e-06, "loss": 0.2043, "step": 540 }, { "epoch": 0.08170971152393898, "grad_norm": 5.286370536581238, "learning_rate": 9.98170722442689e-06, "loss": 0.1822, "step": 541 }, { "epoch": 0.08186074611085939, "grad_norm": 3.5551095605982965, "learning_rate": 9.981639577595048e-06, "loss": 0.1954, "step": 542 }, { "epoch": 0.08201178069777978, "grad_norm": 2.8090029401223857, "learning_rate": 9.98157180614475e-06, "loss": 0.1684, "step": 543 }, { "epoch": 0.0821628152847002, "grad_norm": 3.001589137743043, "learning_rate": 9.98150391007769e-06, "loss": 0.1444, "step": 544 }, { "epoch": 0.0823138498716206, "grad_norm": 2.7835982713480645, "learning_rate": 9.981435889395572e-06, "loss": 0.1629, "step": 545 }, { "epoch": 0.082464884458541, "grad_norm": 3.853667469417395, "learning_rate": 9.981367744100093e-06, "loss": 0.1869, "step": 546 }, { "epoch": 0.08261591904546141, "grad_norm": 2.9994158070807773, "learning_rate": 9.981299474192958e-06, "loss": 0.1574, "step": 547 }, { "epoch": 0.08276695363238182, "grad_norm": 4.178811647252809, "learning_rate": 9.98123107967588e-06, "loss": 0.1723, "step": 548 }, { "epoch": 0.08291798821930221, "grad_norm": 2.7528248716463515, "learning_rate": 9.981162560550561e-06, "loss": 0.1566, "step": 549 }, { "epoch": 0.08306902280622262, "grad_norm": 3.490602934795476, "learning_rate": 9.981093916818722e-06, "loss": 0.1701, "step": 550 }, { "epoch": 0.08322005739314303, "grad_norm": 2.842606132633464, "learning_rate": 9.98102514848208e-06, "loss": 0.1549, "step": 551 }, { "epoch": 0.08337109198006344, "grad_norm": 2.732524675954087, "learning_rate": 9.980956255542352e-06, "loss": 0.1492, "step": 552 }, { "epoch": 0.08352212656698384, "grad_norm": 4.711953307105513, "learning_rate": 9.980887238001263e-06, "loss": 0.2145, "step": 553 }, { "epoch": 0.08367316115390425, "grad_norm": 2.5358077037767686, "learning_rate": 9.980818095860538e-06, "loss": 0.1587, "step": 554 }, { "epoch": 0.08382419574082466, "grad_norm": 3.269623426566454, "learning_rate": 9.98074882912191e-06, "loss": 0.1735, "step": 555 }, { "epoch": 0.08397523032774505, "grad_norm": 5.7533082788297305, "learning_rate": 9.980679437787107e-06, "loss": 0.1604, "step": 556 }, { "epoch": 0.08412626491466546, "grad_norm": 4.94915222896149, "learning_rate": 9.98060992185787e-06, "loss": 0.2039, "step": 557 }, { "epoch": 0.08427729950158587, "grad_norm": 3.05537754784103, "learning_rate": 9.980540281335934e-06, "loss": 0.1594, "step": 558 }, { "epoch": 0.08442833408850627, "grad_norm": 2.8358843346993554, "learning_rate": 9.980470516223042e-06, "loss": 0.1495, "step": 559 }, { "epoch": 0.08457936867542667, "grad_norm": 5.520070970471085, "learning_rate": 9.98040062652094e-06, "loss": 0.1957, "step": 560 }, { "epoch": 0.08473040326234708, "grad_norm": 4.187922389266999, "learning_rate": 9.980330612231378e-06, "loss": 0.1724, "step": 561 }, { "epoch": 0.08488143784926748, "grad_norm": 2.44229127156493, "learning_rate": 9.980260473356103e-06, "loss": 0.1345, "step": 562 }, { "epoch": 0.08503247243618789, "grad_norm": 2.9945685843260677, "learning_rate": 9.980190209896874e-06, "loss": 0.1436, "step": 563 }, { "epoch": 0.0851835070231083, "grad_norm": 2.9009621153705343, "learning_rate": 9.980119821855447e-06, "loss": 0.169, "step": 564 }, { "epoch": 0.0853345416100287, "grad_norm": 3.5712791786794242, "learning_rate": 9.980049309233582e-06, "loss": 0.1731, "step": 565 }, { "epoch": 0.0854855761969491, "grad_norm": 3.09813838896804, "learning_rate": 9.979978672033043e-06, "loss": 0.1791, "step": 566 }, { "epoch": 0.08563661078386951, "grad_norm": 2.9368886974735746, "learning_rate": 9.979907910255598e-06, "loss": 0.1556, "step": 567 }, { "epoch": 0.08578764537078991, "grad_norm": 4.180430378719176, "learning_rate": 9.979837023903016e-06, "loss": 0.139, "step": 568 }, { "epoch": 0.08593867995771032, "grad_norm": 3.563057316912559, "learning_rate": 9.979766012977073e-06, "loss": 0.1637, "step": 569 }, { "epoch": 0.08608971454463073, "grad_norm": 3.7082233006822127, "learning_rate": 9.97969487747954e-06, "loss": 0.1877, "step": 570 }, { "epoch": 0.08624074913155112, "grad_norm": 3.212509985051406, "learning_rate": 9.979623617412203e-06, "loss": 0.1575, "step": 571 }, { "epoch": 0.08639178371847153, "grad_norm": 3.12150452119328, "learning_rate": 9.97955223277684e-06, "loss": 0.166, "step": 572 }, { "epoch": 0.08654281830539194, "grad_norm": 3.509581334884807, "learning_rate": 9.979480723575239e-06, "loss": 0.1744, "step": 573 }, { "epoch": 0.08669385289231234, "grad_norm": 3.6684674390722627, "learning_rate": 9.979409089809187e-06, "loss": 0.1719, "step": 574 }, { "epoch": 0.08684488747923275, "grad_norm": 3.1995606215996055, "learning_rate": 9.979337331480478e-06, "loss": 0.1505, "step": 575 }, { "epoch": 0.08699592206615316, "grad_norm": 3.811620326741347, "learning_rate": 9.979265448590904e-06, "loss": 0.182, "step": 576 }, { "epoch": 0.08714695665307355, "grad_norm": 4.793319335379477, "learning_rate": 9.979193441142267e-06, "loss": 0.1667, "step": 577 }, { "epoch": 0.08729799123999396, "grad_norm": 2.467265480380921, "learning_rate": 9.979121309136366e-06, "loss": 0.1313, "step": 578 }, { "epoch": 0.08744902582691437, "grad_norm": 2.9195419051049067, "learning_rate": 9.979049052575008e-06, "loss": 0.1514, "step": 579 }, { "epoch": 0.08760006041383477, "grad_norm": 3.4395809126582697, "learning_rate": 9.978976671459995e-06, "loss": 0.186, "step": 580 }, { "epoch": 0.08775109500075517, "grad_norm": 3.635341968702123, "learning_rate": 9.978904165793145e-06, "loss": 0.1634, "step": 581 }, { "epoch": 0.08790212958767558, "grad_norm": 3.8548987948446882, "learning_rate": 9.978831535576264e-06, "loss": 0.1657, "step": 582 }, { "epoch": 0.08805316417459598, "grad_norm": 8.184861066535039, "learning_rate": 9.978758780811175e-06, "loss": 0.1543, "step": 583 }, { "epoch": 0.08820419876151639, "grad_norm": 3.033753266803461, "learning_rate": 9.978685901499696e-06, "loss": 0.1659, "step": 584 }, { "epoch": 0.0883552333484368, "grad_norm": 3.4394360498608085, "learning_rate": 9.978612897643648e-06, "loss": 0.1472, "step": 585 }, { "epoch": 0.0885062679353572, "grad_norm": 4.984547376359847, "learning_rate": 9.978539769244861e-06, "loss": 0.1733, "step": 586 }, { "epoch": 0.0886573025222776, "grad_norm": 4.0369266310138485, "learning_rate": 9.97846651630516e-06, "loss": 0.1803, "step": 587 }, { "epoch": 0.08880833710919801, "grad_norm": 3.936696234973551, "learning_rate": 9.978393138826383e-06, "loss": 0.1856, "step": 588 }, { "epoch": 0.08895937169611841, "grad_norm": 2.8888323203428783, "learning_rate": 9.978319636810362e-06, "loss": 0.1675, "step": 589 }, { "epoch": 0.08911040628303882, "grad_norm": 3.4115810544434666, "learning_rate": 9.978246010258936e-06, "loss": 0.1562, "step": 590 }, { "epoch": 0.08926144086995923, "grad_norm": 3.8451845253372245, "learning_rate": 9.978172259173946e-06, "loss": 0.1539, "step": 591 }, { "epoch": 0.08941247545687962, "grad_norm": 3.156241419846115, "learning_rate": 9.978098383557238e-06, "loss": 0.1706, "step": 592 }, { "epoch": 0.08956351004380003, "grad_norm": 3.5770041425364147, "learning_rate": 9.97802438341066e-06, "loss": 0.1828, "step": 593 }, { "epoch": 0.08971454463072044, "grad_norm": 2.669301549255329, "learning_rate": 9.977950258736065e-06, "loss": 0.1419, "step": 594 }, { "epoch": 0.08986557921764084, "grad_norm": 3.2502144398970896, "learning_rate": 9.977876009535303e-06, "loss": 0.1799, "step": 595 }, { "epoch": 0.09001661380456125, "grad_norm": 3.5648277529881875, "learning_rate": 9.977801635810236e-06, "loss": 0.1727, "step": 596 }, { "epoch": 0.09016764839148166, "grad_norm": 3.12372717268673, "learning_rate": 9.977727137562721e-06, "loss": 0.1587, "step": 597 }, { "epoch": 0.09031868297840205, "grad_norm": 3.849847441998453, "learning_rate": 9.977652514794624e-06, "loss": 0.1644, "step": 598 }, { "epoch": 0.09046971756532246, "grad_norm": 2.968902288425904, "learning_rate": 9.977577767507808e-06, "loss": 0.1637, "step": 599 }, { "epoch": 0.09062075215224287, "grad_norm": 3.573661838378854, "learning_rate": 9.97750289570415e-06, "loss": 0.1883, "step": 600 }, { "epoch": 0.09077178673916327, "grad_norm": 3.3368246794471306, "learning_rate": 9.977427899385515e-06, "loss": 0.1695, "step": 601 }, { "epoch": 0.09092282132608367, "grad_norm": 3.8660287400591336, "learning_rate": 9.977352778553783e-06, "loss": 0.16, "step": 602 }, { "epoch": 0.09107385591300408, "grad_norm": 3.9974974265157592, "learning_rate": 9.977277533210832e-06, "loss": 0.1607, "step": 603 }, { "epoch": 0.09122489049992448, "grad_norm": 2.979483728849965, "learning_rate": 9.977202163358548e-06, "loss": 0.1601, "step": 604 }, { "epoch": 0.09137592508684489, "grad_norm": 2.5926756046998194, "learning_rate": 9.977126668998811e-06, "loss": 0.1435, "step": 605 }, { "epoch": 0.0915269596737653, "grad_norm": 3.5330311021022243, "learning_rate": 9.977051050133513e-06, "loss": 0.1524, "step": 606 }, { "epoch": 0.0916779942606857, "grad_norm": 3.1564750896708995, "learning_rate": 9.976975306764543e-06, "loss": 0.1614, "step": 607 }, { "epoch": 0.0918290288476061, "grad_norm": 4.126408307827154, "learning_rate": 9.9768994388938e-06, "loss": 0.1996, "step": 608 }, { "epoch": 0.09198006343452651, "grad_norm": 4.545191149395787, "learning_rate": 9.976823446523176e-06, "loss": 0.1913, "step": 609 }, { "epoch": 0.09213109802144691, "grad_norm": 2.8034500568629204, "learning_rate": 9.976747329654577e-06, "loss": 0.1528, "step": 610 }, { "epoch": 0.09228213260836732, "grad_norm": 2.829239624445306, "learning_rate": 9.976671088289906e-06, "loss": 0.1585, "step": 611 }, { "epoch": 0.09243316719528773, "grad_norm": 2.645696057360248, "learning_rate": 9.976594722431069e-06, "loss": 0.1632, "step": 612 }, { "epoch": 0.09258420178220812, "grad_norm": 2.7970512742482923, "learning_rate": 9.976518232079976e-06, "loss": 0.1536, "step": 613 }, { "epoch": 0.09273523636912853, "grad_norm": 2.8402894931184894, "learning_rate": 9.976441617238544e-06, "loss": 0.1564, "step": 614 }, { "epoch": 0.09288627095604894, "grad_norm": 3.377121524634637, "learning_rate": 9.976364877908684e-06, "loss": 0.1828, "step": 615 }, { "epoch": 0.09303730554296934, "grad_norm": 2.9809217319102057, "learning_rate": 9.97628801409232e-06, "loss": 0.1689, "step": 616 }, { "epoch": 0.09318834012988975, "grad_norm": 3.3794417306025117, "learning_rate": 9.976211025791374e-06, "loss": 0.1553, "step": 617 }, { "epoch": 0.09333937471681016, "grad_norm": 3.3253696156563493, "learning_rate": 9.97613391300777e-06, "loss": 0.1658, "step": 618 }, { "epoch": 0.09349040930373055, "grad_norm": 3.3971068689566883, "learning_rate": 9.976056675743438e-06, "loss": 0.1606, "step": 619 }, { "epoch": 0.09364144389065096, "grad_norm": 3.0619196044239305, "learning_rate": 9.975979314000313e-06, "loss": 0.1517, "step": 620 }, { "epoch": 0.09379247847757137, "grad_norm": 3.942243098034354, "learning_rate": 9.975901827780325e-06, "loss": 0.1791, "step": 621 }, { "epoch": 0.09394351306449177, "grad_norm": 2.855151957217576, "learning_rate": 9.975824217085416e-06, "loss": 0.1508, "step": 622 }, { "epoch": 0.09409454765141217, "grad_norm": 2.983311097333986, "learning_rate": 9.975746481917526e-06, "loss": 0.1721, "step": 623 }, { "epoch": 0.09424558223833258, "grad_norm": 3.106761264002703, "learning_rate": 9.9756686222786e-06, "loss": 0.1833, "step": 624 }, { "epoch": 0.09439661682525298, "grad_norm": 3.2231018461948504, "learning_rate": 9.975590638170586e-06, "loss": 0.1564, "step": 625 }, { "epoch": 0.09454765141217339, "grad_norm": 3.368526247616104, "learning_rate": 9.975512529595437e-06, "loss": 0.1696, "step": 626 }, { "epoch": 0.0946986859990938, "grad_norm": 4.056141172724535, "learning_rate": 9.975434296555102e-06, "loss": 0.1695, "step": 627 }, { "epoch": 0.0948497205860142, "grad_norm": 2.9786707167738515, "learning_rate": 9.975355939051542e-06, "loss": 0.1663, "step": 628 }, { "epoch": 0.0950007551729346, "grad_norm": 3.1871329785579943, "learning_rate": 9.975277457086713e-06, "loss": 0.1476, "step": 629 }, { "epoch": 0.09515178975985501, "grad_norm": 2.9684046794221115, "learning_rate": 9.975198850662584e-06, "loss": 0.1632, "step": 630 }, { "epoch": 0.09530282434677541, "grad_norm": 3.5329348986700975, "learning_rate": 9.975120119781118e-06, "loss": 0.1661, "step": 631 }, { "epoch": 0.09545385893369582, "grad_norm": 3.5601183020438727, "learning_rate": 9.975041264444285e-06, "loss": 0.1786, "step": 632 }, { "epoch": 0.09560489352061623, "grad_norm": 3.556283328435932, "learning_rate": 9.974962284654054e-06, "loss": 0.1678, "step": 633 }, { "epoch": 0.09575592810753662, "grad_norm": 3.5823714455592666, "learning_rate": 9.974883180412409e-06, "loss": 0.1698, "step": 634 }, { "epoch": 0.09590696269445703, "grad_norm": 3.9499363930478446, "learning_rate": 9.974803951721322e-06, "loss": 0.165, "step": 635 }, { "epoch": 0.09605799728137744, "grad_norm": 4.204881453173944, "learning_rate": 9.974724598582777e-06, "loss": 0.1918, "step": 636 }, { "epoch": 0.09620903186829784, "grad_norm": 2.9211840199644294, "learning_rate": 9.974645120998759e-06, "loss": 0.1502, "step": 637 }, { "epoch": 0.09636006645521825, "grad_norm": 3.959490324837134, "learning_rate": 9.974565518971258e-06, "loss": 0.195, "step": 638 }, { "epoch": 0.09651110104213866, "grad_norm": 3.1881635236245893, "learning_rate": 9.97448579250226e-06, "loss": 0.1719, "step": 639 }, { "epoch": 0.09666213562905905, "grad_norm": 3.0826991520188716, "learning_rate": 9.974405941593765e-06, "loss": 0.1616, "step": 640 }, { "epoch": 0.09681317021597946, "grad_norm": 3.670566770155845, "learning_rate": 9.974325966247768e-06, "loss": 0.1937, "step": 641 }, { "epoch": 0.09696420480289987, "grad_norm": 2.8391570348776662, "learning_rate": 9.974245866466272e-06, "loss": 0.1657, "step": 642 }, { "epoch": 0.09711523938982027, "grad_norm": 3.5146798588744814, "learning_rate": 9.974165642251278e-06, "loss": 0.1822, "step": 643 }, { "epoch": 0.09726627397674067, "grad_norm": 2.73353912667958, "learning_rate": 9.974085293604793e-06, "loss": 0.1401, "step": 644 }, { "epoch": 0.09741730856366108, "grad_norm": 2.530470685078373, "learning_rate": 9.974004820528828e-06, "loss": 0.1309, "step": 645 }, { "epoch": 0.09756834315058148, "grad_norm": 2.560629167854562, "learning_rate": 9.973924223025395e-06, "loss": 0.1412, "step": 646 }, { "epoch": 0.09771937773750189, "grad_norm": 4.880339454391561, "learning_rate": 9.973843501096511e-06, "loss": 0.1852, "step": 647 }, { "epoch": 0.0978704123244223, "grad_norm": 4.448101667669772, "learning_rate": 9.973762654744197e-06, "loss": 0.2009, "step": 648 }, { "epoch": 0.0980214469113427, "grad_norm": 4.9598048715801495, "learning_rate": 9.97368168397047e-06, "loss": 0.1782, "step": 649 }, { "epoch": 0.0981724814982631, "grad_norm": 3.3732103808493434, "learning_rate": 9.973600588777364e-06, "loss": 0.1876, "step": 650 }, { "epoch": 0.09832351608518351, "grad_norm": 2.855124771257598, "learning_rate": 9.9735193691669e-06, "loss": 0.1664, "step": 651 }, { "epoch": 0.09847455067210391, "grad_norm": 6.680821263428072, "learning_rate": 9.973438025141113e-06, "loss": 0.1807, "step": 652 }, { "epoch": 0.09862558525902432, "grad_norm": 3.6739559061524543, "learning_rate": 9.973356556702037e-06, "loss": 0.18, "step": 653 }, { "epoch": 0.09877661984594473, "grad_norm": 5.923257215370391, "learning_rate": 9.97327496385171e-06, "loss": 0.1961, "step": 654 }, { "epoch": 0.09892765443286512, "grad_norm": 3.8260753252342234, "learning_rate": 9.973193246592174e-06, "loss": 0.2076, "step": 655 }, { "epoch": 0.09907868901978553, "grad_norm": 3.4599615135722304, "learning_rate": 9.973111404925472e-06, "loss": 0.1574, "step": 656 }, { "epoch": 0.09922972360670594, "grad_norm": 2.439587937018964, "learning_rate": 9.973029438853654e-06, "loss": 0.1527, "step": 657 }, { "epoch": 0.09938075819362634, "grad_norm": 3.176014250294739, "learning_rate": 9.972947348378767e-06, "loss": 0.1719, "step": 658 }, { "epoch": 0.09953179278054675, "grad_norm": 3.5755428497296102, "learning_rate": 9.972865133502867e-06, "loss": 0.1858, "step": 659 }, { "epoch": 0.09968282736746716, "grad_norm": 3.2647787550313665, "learning_rate": 9.972782794228008e-06, "loss": 0.1735, "step": 660 }, { "epoch": 0.09983386195438755, "grad_norm": 4.471245964467144, "learning_rate": 9.972700330556253e-06, "loss": 0.1518, "step": 661 }, { "epoch": 0.09998489654130796, "grad_norm": 4.084261571785137, "learning_rate": 9.972617742489662e-06, "loss": 0.161, "step": 662 }, { "epoch": 0.10013593112822837, "grad_norm": 3.83715327044534, "learning_rate": 9.972535030030303e-06, "loss": 0.1787, "step": 663 }, { "epoch": 0.10028696571514877, "grad_norm": 3.9797454294513948, "learning_rate": 9.972452193180245e-06, "loss": 0.1735, "step": 664 }, { "epoch": 0.10043800030206917, "grad_norm": 3.293012631761966, "learning_rate": 9.97236923194156e-06, "loss": 0.1571, "step": 665 }, { "epoch": 0.10058903488898958, "grad_norm": 3.8939090062444204, "learning_rate": 9.972286146316322e-06, "loss": 0.1627, "step": 666 }, { "epoch": 0.10074006947590998, "grad_norm": 3.141389089382669, "learning_rate": 9.97220293630661e-06, "loss": 0.1646, "step": 667 }, { "epoch": 0.10089110406283039, "grad_norm": 3.512421009983053, "learning_rate": 9.972119601914505e-06, "loss": 0.147, "step": 668 }, { "epoch": 0.1010421386497508, "grad_norm": 2.8721608814378374, "learning_rate": 9.972036143142092e-06, "loss": 0.1616, "step": 669 }, { "epoch": 0.1011931732366712, "grad_norm": 2.931711391988398, "learning_rate": 9.971952559991461e-06, "loss": 0.1707, "step": 670 }, { "epoch": 0.1013442078235916, "grad_norm": 3.047805406448608, "learning_rate": 9.971868852464701e-06, "loss": 0.1546, "step": 671 }, { "epoch": 0.10149524241051201, "grad_norm": 4.205221656782718, "learning_rate": 9.971785020563907e-06, "loss": 0.1894, "step": 672 }, { "epoch": 0.10164627699743241, "grad_norm": 4.905713724626514, "learning_rate": 9.971701064291172e-06, "loss": 0.1985, "step": 673 }, { "epoch": 0.10179731158435282, "grad_norm": 3.597340558327726, "learning_rate": 9.971616983648603e-06, "loss": 0.1638, "step": 674 }, { "epoch": 0.10194834617127323, "grad_norm": 3.2584852193574205, "learning_rate": 9.971532778638298e-06, "loss": 0.162, "step": 675 }, { "epoch": 0.10209938075819362, "grad_norm": 3.0534300157033636, "learning_rate": 9.971448449262366e-06, "loss": 0.1765, "step": 676 }, { "epoch": 0.10225041534511403, "grad_norm": 5.515298193793144, "learning_rate": 9.971363995522917e-06, "loss": 0.2077, "step": 677 }, { "epoch": 0.10240144993203444, "grad_norm": 3.1833179496443025, "learning_rate": 9.97127941742206e-06, "loss": 0.1814, "step": 678 }, { "epoch": 0.10255248451895484, "grad_norm": 3.866381516173686, "learning_rate": 9.971194714961914e-06, "loss": 0.1917, "step": 679 }, { "epoch": 0.10270351910587525, "grad_norm": 3.764789437243979, "learning_rate": 9.971109888144597e-06, "loss": 0.191, "step": 680 }, { "epoch": 0.10285455369279566, "grad_norm": 3.419851331692132, "learning_rate": 9.971024936972231e-06, "loss": 0.182, "step": 681 }, { "epoch": 0.10300558827971605, "grad_norm": 3.0150093990923685, "learning_rate": 9.97093986144694e-06, "loss": 0.1561, "step": 682 }, { "epoch": 0.10315662286663646, "grad_norm": 3.3051423203499732, "learning_rate": 9.970854661570857e-06, "loss": 0.1736, "step": 683 }, { "epoch": 0.10330765745355687, "grad_norm": 2.784329203395719, "learning_rate": 9.970769337346107e-06, "loss": 0.1583, "step": 684 }, { "epoch": 0.10345869204047727, "grad_norm": 3.295882369633486, "learning_rate": 9.970683888774827e-06, "loss": 0.1685, "step": 685 }, { "epoch": 0.10360972662739767, "grad_norm": 2.9255793610380927, "learning_rate": 9.970598315859154e-06, "loss": 0.1925, "step": 686 }, { "epoch": 0.10376076121431808, "grad_norm": 3.734052582735421, "learning_rate": 9.970512618601231e-06, "loss": 0.169, "step": 687 }, { "epoch": 0.10391179580123848, "grad_norm": 2.6900975964135423, "learning_rate": 9.970426797003202e-06, "loss": 0.1581, "step": 688 }, { "epoch": 0.10406283038815889, "grad_norm": 3.3091374692706497, "learning_rate": 9.970340851067209e-06, "loss": 0.1818, "step": 689 }, { "epoch": 0.1042138649750793, "grad_norm": 2.7628721090103654, "learning_rate": 9.970254780795404e-06, "loss": 0.1671, "step": 690 }, { "epoch": 0.1043648995619997, "grad_norm": 3.341632101456738, "learning_rate": 9.970168586189943e-06, "loss": 0.1617, "step": 691 }, { "epoch": 0.1045159341489201, "grad_norm": 2.9354720833171, "learning_rate": 9.97008226725298e-06, "loss": 0.1715, "step": 692 }, { "epoch": 0.10466696873584051, "grad_norm": 2.946403367751692, "learning_rate": 9.969995823986674e-06, "loss": 0.1612, "step": 693 }, { "epoch": 0.10481800332276091, "grad_norm": 3.4449770159006086, "learning_rate": 9.969909256393189e-06, "loss": 0.1779, "step": 694 }, { "epoch": 0.10496903790968132, "grad_norm": 3.7778134698644923, "learning_rate": 9.969822564474688e-06, "loss": 0.1796, "step": 695 }, { "epoch": 0.10512007249660173, "grad_norm": 3.6054075111694064, "learning_rate": 9.969735748233342e-06, "loss": 0.1943, "step": 696 }, { "epoch": 0.10527110708352212, "grad_norm": 3.795881238303369, "learning_rate": 9.969648807671321e-06, "loss": 0.1773, "step": 697 }, { "epoch": 0.10542214167044253, "grad_norm": 3.0673021584598796, "learning_rate": 9.969561742790803e-06, "loss": 0.1891, "step": 698 }, { "epoch": 0.10557317625736294, "grad_norm": 3.382168342749969, "learning_rate": 9.969474553593964e-06, "loss": 0.1712, "step": 699 }, { "epoch": 0.10572421084428334, "grad_norm": 3.694327127849169, "learning_rate": 9.969387240082982e-06, "loss": 0.174, "step": 700 }, { "epoch": 0.10587524543120375, "grad_norm": 2.510744458529531, "learning_rate": 9.969299802260046e-06, "loss": 0.1481, "step": 701 }, { "epoch": 0.10602628001812416, "grad_norm": 3.271268014211583, "learning_rate": 9.96921224012734e-06, "loss": 0.1715, "step": 702 }, { "epoch": 0.10617731460504455, "grad_norm": 3.8396111662153762, "learning_rate": 9.969124553687057e-06, "loss": 0.2256, "step": 703 }, { "epoch": 0.10632834919196496, "grad_norm": 2.9132992778841458, "learning_rate": 9.96903674294139e-06, "loss": 0.1741, "step": 704 }, { "epoch": 0.10647938377888537, "grad_norm": 4.028533357231575, "learning_rate": 9.968948807892532e-06, "loss": 0.1761, "step": 705 }, { "epoch": 0.10663041836580577, "grad_norm": 3.1812753267687492, "learning_rate": 9.968860748542688e-06, "loss": 0.1793, "step": 706 }, { "epoch": 0.10678145295272617, "grad_norm": 2.7318795059325276, "learning_rate": 9.968772564894059e-06, "loss": 0.1609, "step": 707 }, { "epoch": 0.10693248753964658, "grad_norm": 3.3979797104396874, "learning_rate": 9.968684256948848e-06, "loss": 0.1647, "step": 708 }, { "epoch": 0.10708352212656698, "grad_norm": 3.6444549610602355, "learning_rate": 9.968595824709268e-06, "loss": 0.1724, "step": 709 }, { "epoch": 0.10723455671348739, "grad_norm": 3.1336430881966293, "learning_rate": 9.96850726817753e-06, "loss": 0.1698, "step": 710 }, { "epoch": 0.1073855913004078, "grad_norm": 3.731298347794092, "learning_rate": 9.968418587355848e-06, "loss": 0.165, "step": 711 }, { "epoch": 0.1075366258873282, "grad_norm": 4.0293203800513675, "learning_rate": 9.968329782246443e-06, "loss": 0.2507, "step": 712 }, { "epoch": 0.1076876604742486, "grad_norm": 3.8766305996974677, "learning_rate": 9.968240852851533e-06, "loss": 0.1679, "step": 713 }, { "epoch": 0.10783869506116901, "grad_norm": 2.872066768076661, "learning_rate": 9.968151799173346e-06, "loss": 0.1894, "step": 714 }, { "epoch": 0.10798972964808941, "grad_norm": 3.1443097529320974, "learning_rate": 9.968062621214109e-06, "loss": 0.1595, "step": 715 }, { "epoch": 0.10814076423500982, "grad_norm": 3.9841130426843194, "learning_rate": 9.967973318976051e-06, "loss": 0.1826, "step": 716 }, { "epoch": 0.10829179882193023, "grad_norm": 2.9802510850824095, "learning_rate": 9.967883892461408e-06, "loss": 0.1768, "step": 717 }, { "epoch": 0.10844283340885062, "grad_norm": 2.749124446668053, "learning_rate": 9.967794341672415e-06, "loss": 0.1479, "step": 718 }, { "epoch": 0.10859386799577103, "grad_norm": 2.7757885732143395, "learning_rate": 9.967704666611313e-06, "loss": 0.1389, "step": 719 }, { "epoch": 0.10874490258269144, "grad_norm": 3.6335195683964594, "learning_rate": 9.967614867280346e-06, "loss": 0.1878, "step": 720 }, { "epoch": 0.10889593716961184, "grad_norm": 5.078349128807621, "learning_rate": 9.96752494368176e-06, "loss": 0.2094, "step": 721 }, { "epoch": 0.10904697175653225, "grad_norm": 3.540488842053732, "learning_rate": 9.967434895817805e-06, "loss": 0.1614, "step": 722 }, { "epoch": 0.10919800634345266, "grad_norm": 2.815398696457693, "learning_rate": 9.967344723690732e-06, "loss": 0.1708, "step": 723 }, { "epoch": 0.10934904093037305, "grad_norm": 2.7676541984207077, "learning_rate": 9.967254427302798e-06, "loss": 0.1587, "step": 724 }, { "epoch": 0.10950007551729346, "grad_norm": 3.3081032480251937, "learning_rate": 9.967164006656263e-06, "loss": 0.1519, "step": 725 }, { "epoch": 0.10965111010421387, "grad_norm": 3.5294311002227543, "learning_rate": 9.967073461753385e-06, "loss": 0.1658, "step": 726 }, { "epoch": 0.10980214469113427, "grad_norm": 2.9254324052919247, "learning_rate": 9.966982792596432e-06, "loss": 0.1595, "step": 727 }, { "epoch": 0.10995317927805467, "grad_norm": 3.651139492528313, "learning_rate": 9.966891999187673e-06, "loss": 0.1668, "step": 728 }, { "epoch": 0.11010421386497508, "grad_norm": 3.2234126900422284, "learning_rate": 9.966801081529376e-06, "loss": 0.1622, "step": 729 }, { "epoch": 0.11025524845189548, "grad_norm": 2.560779575532902, "learning_rate": 9.96671003962382e-06, "loss": 0.1616, "step": 730 }, { "epoch": 0.11040628303881589, "grad_norm": 3.392278854919292, "learning_rate": 9.966618873473277e-06, "loss": 0.189, "step": 731 }, { "epoch": 0.1105573176257363, "grad_norm": 3.0385836979730425, "learning_rate": 9.966527583080033e-06, "loss": 0.1593, "step": 732 }, { "epoch": 0.1107083522126567, "grad_norm": 3.313468278598065, "learning_rate": 9.966436168446366e-06, "loss": 0.1739, "step": 733 }, { "epoch": 0.1108593867995771, "grad_norm": 2.9671740145469583, "learning_rate": 9.966344629574567e-06, "loss": 0.1525, "step": 734 }, { "epoch": 0.11101042138649751, "grad_norm": 3.4675136901705073, "learning_rate": 9.966252966466926e-06, "loss": 0.1829, "step": 735 }, { "epoch": 0.11116145597341791, "grad_norm": 4.4113904460236535, "learning_rate": 9.966161179125731e-06, "loss": 0.1845, "step": 736 }, { "epoch": 0.11131249056033832, "grad_norm": 3.5032980074204216, "learning_rate": 9.966069267553284e-06, "loss": 0.1633, "step": 737 }, { "epoch": 0.11146352514725873, "grad_norm": 2.834233026032157, "learning_rate": 9.965977231751883e-06, "loss": 0.1647, "step": 738 }, { "epoch": 0.11161455973417912, "grad_norm": 3.399020680415306, "learning_rate": 9.965885071723829e-06, "loss": 0.1691, "step": 739 }, { "epoch": 0.11176559432109953, "grad_norm": 2.9572854661960815, "learning_rate": 9.965792787471427e-06, "loss": 0.1751, "step": 740 }, { "epoch": 0.11191662890801994, "grad_norm": 3.33828551858083, "learning_rate": 9.965700378996984e-06, "loss": 0.1672, "step": 741 }, { "epoch": 0.11206766349494034, "grad_norm": 2.489517811314558, "learning_rate": 9.965607846302817e-06, "loss": 0.1538, "step": 742 }, { "epoch": 0.11221869808186075, "grad_norm": 2.9668745009842454, "learning_rate": 9.965515189391237e-06, "loss": 0.1748, "step": 743 }, { "epoch": 0.11236973266878116, "grad_norm": 2.8372229165378164, "learning_rate": 9.965422408264563e-06, "loss": 0.1514, "step": 744 }, { "epoch": 0.11252076725570155, "grad_norm": 2.966506908467527, "learning_rate": 9.965329502925115e-06, "loss": 0.1529, "step": 745 }, { "epoch": 0.11267180184262196, "grad_norm": 3.6983160899536607, "learning_rate": 9.965236473375217e-06, "loss": 0.1642, "step": 746 }, { "epoch": 0.11282283642954237, "grad_norm": 2.5472517660385137, "learning_rate": 9.965143319617196e-06, "loss": 0.1535, "step": 747 }, { "epoch": 0.11297387101646277, "grad_norm": 3.1531041872959604, "learning_rate": 9.965050041653386e-06, "loss": 0.1727, "step": 748 }, { "epoch": 0.11312490560338317, "grad_norm": 4.125081710908682, "learning_rate": 9.964956639486114e-06, "loss": 0.2019, "step": 749 }, { "epoch": 0.11327594019030358, "grad_norm": 3.028775799685097, "learning_rate": 9.964863113117722e-06, "loss": 0.1733, "step": 750 }, { "epoch": 0.11342697477722398, "grad_norm": 3.14475915068562, "learning_rate": 9.964769462550547e-06, "loss": 0.1775, "step": 751 }, { "epoch": 0.11357800936414439, "grad_norm": 2.972614701364103, "learning_rate": 9.964675687786933e-06, "loss": 0.1769, "step": 752 }, { "epoch": 0.1137290439510648, "grad_norm": 3.1220526817612133, "learning_rate": 9.964581788829225e-06, "loss": 0.1797, "step": 753 }, { "epoch": 0.1138800785379852, "grad_norm": 3.221397353013783, "learning_rate": 9.964487765679772e-06, "loss": 0.1731, "step": 754 }, { "epoch": 0.1140311131249056, "grad_norm": 2.8230443862805106, "learning_rate": 9.964393618340926e-06, "loss": 0.1725, "step": 755 }, { "epoch": 0.11418214771182601, "grad_norm": 3.9327060552850606, "learning_rate": 9.964299346815041e-06, "loss": 0.2093, "step": 756 }, { "epoch": 0.11433318229874641, "grad_norm": 2.8521701080521815, "learning_rate": 9.964204951104477e-06, "loss": 0.1554, "step": 757 }, { "epoch": 0.11448421688566682, "grad_norm": 2.895940907348357, "learning_rate": 9.964110431211596e-06, "loss": 0.1389, "step": 758 }, { "epoch": 0.11463525147258723, "grad_norm": 2.9537550350377546, "learning_rate": 9.96401578713876e-06, "loss": 0.1467, "step": 759 }, { "epoch": 0.11478628605950762, "grad_norm": 2.8850442831743894, "learning_rate": 9.963921018888339e-06, "loss": 0.1397, "step": 760 }, { "epoch": 0.11493732064642803, "grad_norm": 3.2583555299837212, "learning_rate": 9.963826126462702e-06, "loss": 0.1909, "step": 761 }, { "epoch": 0.11508835523334844, "grad_norm": 3.0828165867268384, "learning_rate": 9.963731109864223e-06, "loss": 0.1373, "step": 762 }, { "epoch": 0.11523938982026884, "grad_norm": 2.8421008699447246, "learning_rate": 9.96363596909528e-06, "loss": 0.165, "step": 763 }, { "epoch": 0.11539042440718925, "grad_norm": 3.7944168989386604, "learning_rate": 9.963540704158252e-06, "loss": 0.199, "step": 764 }, { "epoch": 0.11554145899410966, "grad_norm": 3.785510362282745, "learning_rate": 9.963445315055523e-06, "loss": 0.2062, "step": 765 }, { "epoch": 0.11569249358103005, "grad_norm": 2.6038295324151663, "learning_rate": 9.963349801789478e-06, "loss": 0.1382, "step": 766 }, { "epoch": 0.11584352816795046, "grad_norm": 3.489049656158465, "learning_rate": 9.963254164362507e-06, "loss": 0.1808, "step": 767 }, { "epoch": 0.11599456275487087, "grad_norm": 2.9132895353936443, "learning_rate": 9.963158402777002e-06, "loss": 0.1494, "step": 768 }, { "epoch": 0.11614559734179127, "grad_norm": 2.8847679020918306, "learning_rate": 9.96306251703536e-06, "loss": 0.1667, "step": 769 }, { "epoch": 0.11629663192871167, "grad_norm": 3.241722983619171, "learning_rate": 9.962966507139975e-06, "loss": 0.1804, "step": 770 }, { "epoch": 0.11644766651563208, "grad_norm": 3.8128722150158962, "learning_rate": 9.962870373093257e-06, "loss": 0.1782, "step": 771 }, { "epoch": 0.11659870110255248, "grad_norm": 3.707801277288662, "learning_rate": 9.962774114897601e-06, "loss": 0.1968, "step": 772 }, { "epoch": 0.11674973568947289, "grad_norm": 3.6599002668599, "learning_rate": 9.962677732555425e-06, "loss": 0.1611, "step": 773 }, { "epoch": 0.1169007702763933, "grad_norm": 2.5825907344918617, "learning_rate": 9.962581226069133e-06, "loss": 0.1513, "step": 774 }, { "epoch": 0.1170518048633137, "grad_norm": 4.8417526195546, "learning_rate": 9.962484595441139e-06, "loss": 0.1851, "step": 775 }, { "epoch": 0.1172028394502341, "grad_norm": 3.6465946302711103, "learning_rate": 9.962387840673867e-06, "loss": 0.1949, "step": 776 }, { "epoch": 0.11735387403715451, "grad_norm": 2.9910695535699467, "learning_rate": 9.96229096176973e-06, "loss": 0.1648, "step": 777 }, { "epoch": 0.11750490862407491, "grad_norm": 3.664583495676884, "learning_rate": 9.962193958731155e-06, "loss": 0.1462, "step": 778 }, { "epoch": 0.11765594321099532, "grad_norm": 2.528860200034072, "learning_rate": 9.962096831560567e-06, "loss": 0.1585, "step": 779 }, { "epoch": 0.11780697779791573, "grad_norm": 2.463875882732881, "learning_rate": 9.961999580260398e-06, "loss": 0.1455, "step": 780 }, { "epoch": 0.11795801238483612, "grad_norm": 2.684518494654107, "learning_rate": 9.961902204833077e-06, "loss": 0.1616, "step": 781 }, { "epoch": 0.11810904697175653, "grad_norm": 3.1560923369859504, "learning_rate": 9.961804705281044e-06, "loss": 0.166, "step": 782 }, { "epoch": 0.11826008155867694, "grad_norm": 2.8241786715215795, "learning_rate": 9.961707081606738e-06, "loss": 0.168, "step": 783 }, { "epoch": 0.11841111614559734, "grad_norm": 3.5122630381726223, "learning_rate": 9.961609333812599e-06, "loss": 0.1745, "step": 784 }, { "epoch": 0.11856215073251775, "grad_norm": 3.1818016669242266, "learning_rate": 9.96151146190107e-06, "loss": 0.1909, "step": 785 }, { "epoch": 0.11871318531943816, "grad_norm": 3.9300642166463664, "learning_rate": 9.961413465874604e-06, "loss": 0.1675, "step": 786 }, { "epoch": 0.11886421990635855, "grad_norm": 4.000056907985667, "learning_rate": 9.961315345735649e-06, "loss": 0.1819, "step": 787 }, { "epoch": 0.11901525449327896, "grad_norm": 3.3551049280672802, "learning_rate": 9.961217101486662e-06, "loss": 0.1641, "step": 788 }, { "epoch": 0.11916628908019937, "grad_norm": 3.1260363341278525, "learning_rate": 9.9611187331301e-06, "loss": 0.1827, "step": 789 }, { "epoch": 0.11931732366711977, "grad_norm": 2.9530467359295796, "learning_rate": 9.961020240668421e-06, "loss": 0.1835, "step": 790 }, { "epoch": 0.11946835825404017, "grad_norm": 3.4454956310617337, "learning_rate": 9.960921624104095e-06, "loss": 0.1904, "step": 791 }, { "epoch": 0.11961939284096058, "grad_norm": 4.278854592247018, "learning_rate": 9.960822883439582e-06, "loss": 0.1934, "step": 792 }, { "epoch": 0.11977042742788098, "grad_norm": 2.8766593118051857, "learning_rate": 9.960724018677356e-06, "loss": 0.1604, "step": 793 }, { "epoch": 0.11992146201480139, "grad_norm": 3.5200821393544777, "learning_rate": 9.96062502981989e-06, "loss": 0.183, "step": 794 }, { "epoch": 0.1200724966017218, "grad_norm": 3.1192234875975724, "learning_rate": 9.960525916869659e-06, "loss": 0.1738, "step": 795 }, { "epoch": 0.1202235311886422, "grad_norm": 3.605701886343441, "learning_rate": 9.960426679829143e-06, "loss": 0.1921, "step": 796 }, { "epoch": 0.1203745657755626, "grad_norm": 3.6338876107452687, "learning_rate": 9.960327318700824e-06, "loss": 0.1809, "step": 797 }, { "epoch": 0.12052560036248301, "grad_norm": 2.7006457775277277, "learning_rate": 9.960227833487188e-06, "loss": 0.1522, "step": 798 }, { "epoch": 0.12067663494940341, "grad_norm": 2.5022270588347455, "learning_rate": 9.960128224190723e-06, "loss": 0.1354, "step": 799 }, { "epoch": 0.12082766953632382, "grad_norm": 2.665007472266893, "learning_rate": 9.960028490813923e-06, "loss": 0.1571, "step": 800 }, { "epoch": 0.12097870412324423, "grad_norm": 3.4601639344548736, "learning_rate": 9.959928633359282e-06, "loss": 0.1626, "step": 801 }, { "epoch": 0.12112973871016462, "grad_norm": 3.1293093523969784, "learning_rate": 9.959828651829296e-06, "loss": 0.1629, "step": 802 }, { "epoch": 0.12128077329708503, "grad_norm": 3.2003125730142945, "learning_rate": 9.959728546226466e-06, "loss": 0.1598, "step": 803 }, { "epoch": 0.12143180788400544, "grad_norm": 2.935515461323718, "learning_rate": 9.959628316553297e-06, "loss": 0.1779, "step": 804 }, { "epoch": 0.12158284247092584, "grad_norm": 2.9299054781231657, "learning_rate": 9.959527962812301e-06, "loss": 0.1504, "step": 805 }, { "epoch": 0.12173387705784625, "grad_norm": 3.065569903550972, "learning_rate": 9.959427485005982e-06, "loss": 0.1619, "step": 806 }, { "epoch": 0.12188491164476666, "grad_norm": 2.9284492362018124, "learning_rate": 9.959326883136855e-06, "loss": 0.1637, "step": 807 }, { "epoch": 0.12203594623168705, "grad_norm": 2.7791780428509925, "learning_rate": 9.959226157207439e-06, "loss": 0.1528, "step": 808 }, { "epoch": 0.12218698081860746, "grad_norm": 2.7771746803923323, "learning_rate": 9.95912530722025e-06, "loss": 0.1434, "step": 809 }, { "epoch": 0.12233801540552787, "grad_norm": 3.253378790295068, "learning_rate": 9.959024333177816e-06, "loss": 0.1564, "step": 810 }, { "epoch": 0.12248904999244827, "grad_norm": 3.94136030329455, "learning_rate": 9.958923235082658e-06, "loss": 0.1808, "step": 811 }, { "epoch": 0.12264008457936867, "grad_norm": 2.59876450969108, "learning_rate": 9.958822012937307e-06, "loss": 0.1714, "step": 812 }, { "epoch": 0.12279111916628908, "grad_norm": 3.540894908810077, "learning_rate": 9.958720666744295e-06, "loss": 0.1844, "step": 813 }, { "epoch": 0.12294215375320948, "grad_norm": 2.650526831730403, "learning_rate": 9.958619196506157e-06, "loss": 0.1673, "step": 814 }, { "epoch": 0.12309318834012989, "grad_norm": 3.254798024004384, "learning_rate": 9.958517602225432e-06, "loss": 0.1715, "step": 815 }, { "epoch": 0.1232442229270503, "grad_norm": 2.748086782422205, "learning_rate": 9.958415883904661e-06, "loss": 0.1568, "step": 816 }, { "epoch": 0.1233952575139707, "grad_norm": 3.9662082568790877, "learning_rate": 9.95831404154639e-06, "loss": 0.1683, "step": 817 }, { "epoch": 0.1235462921008911, "grad_norm": 2.9009119313978537, "learning_rate": 9.958212075153163e-06, "loss": 0.152, "step": 818 }, { "epoch": 0.12369732668781151, "grad_norm": 2.9845169422161404, "learning_rate": 9.958109984727533e-06, "loss": 0.1431, "step": 819 }, { "epoch": 0.12384836127473191, "grad_norm": 3.4612660122178798, "learning_rate": 9.958007770272055e-06, "loss": 0.2011, "step": 820 }, { "epoch": 0.12399939586165232, "grad_norm": 3.8888105287069776, "learning_rate": 9.957905431789284e-06, "loss": 0.174, "step": 821 }, { "epoch": 0.12415043044857273, "grad_norm": 4.1307463698272695, "learning_rate": 9.95780296928178e-06, "loss": 0.1613, "step": 822 }, { "epoch": 0.12430146503549312, "grad_norm": 3.363266791089063, "learning_rate": 9.95770038275211e-06, "loss": 0.165, "step": 823 }, { "epoch": 0.12445249962241353, "grad_norm": 2.8910292051535897, "learning_rate": 9.957597672202834e-06, "loss": 0.1758, "step": 824 }, { "epoch": 0.12460353420933394, "grad_norm": 3.902557815425972, "learning_rate": 9.957494837636526e-06, "loss": 0.178, "step": 825 }, { "epoch": 0.12475456879625434, "grad_norm": 2.9354591964600543, "learning_rate": 9.957391879055757e-06, "loss": 0.1745, "step": 826 }, { "epoch": 0.12490560338317475, "grad_norm": 6.331046486842062, "learning_rate": 9.957288796463103e-06, "loss": 0.1945, "step": 827 }, { "epoch": 0.12505663797009514, "grad_norm": 2.814484291005432, "learning_rate": 9.95718558986114e-06, "loss": 0.1538, "step": 828 }, { "epoch": 0.12520767255701556, "grad_norm": 3.780559066556261, "learning_rate": 9.957082259252454e-06, "loss": 0.1608, "step": 829 }, { "epoch": 0.12535870714393596, "grad_norm": 3.592950924177683, "learning_rate": 9.956978804639627e-06, "loss": 0.165, "step": 830 }, { "epoch": 0.12550974173085636, "grad_norm": 2.5468225416922183, "learning_rate": 9.95687522602525e-06, "loss": 0.1415, "step": 831 }, { "epoch": 0.12566077631777678, "grad_norm": 4.10230005520567, "learning_rate": 9.95677152341191e-06, "loss": 0.1752, "step": 832 }, { "epoch": 0.12581181090469717, "grad_norm": 3.1354088086787417, "learning_rate": 9.956667696802203e-06, "loss": 0.1775, "step": 833 }, { "epoch": 0.12596284549161757, "grad_norm": 3.0278659355244297, "learning_rate": 9.956563746198728e-06, "loss": 0.1775, "step": 834 }, { "epoch": 0.126113880078538, "grad_norm": 3.068878455638462, "learning_rate": 9.956459671604082e-06, "loss": 0.1831, "step": 835 }, { "epoch": 0.1262649146654584, "grad_norm": 2.7251089398887767, "learning_rate": 9.95635547302087e-06, "loss": 0.1615, "step": 836 }, { "epoch": 0.12641594925237878, "grad_norm": 5.818223657710914, "learning_rate": 9.9562511504517e-06, "loss": 0.1704, "step": 837 }, { "epoch": 0.1265669838392992, "grad_norm": 2.9829782632528548, "learning_rate": 9.956146703899182e-06, "loss": 0.1864, "step": 838 }, { "epoch": 0.1267180184262196, "grad_norm": 3.1933195354089747, "learning_rate": 9.956042133365926e-06, "loss": 0.1757, "step": 839 }, { "epoch": 0.12686905301314, "grad_norm": 2.998751559579723, "learning_rate": 9.95593743885455e-06, "loss": 0.1715, "step": 840 }, { "epoch": 0.12702008760006042, "grad_norm": 2.6626403813241555, "learning_rate": 9.95583262036767e-06, "loss": 0.1591, "step": 841 }, { "epoch": 0.12717112218698082, "grad_norm": 2.8420365520233877, "learning_rate": 9.955727677907912e-06, "loss": 0.1674, "step": 842 }, { "epoch": 0.1273221567739012, "grad_norm": 3.5651603370607035, "learning_rate": 9.9556226114779e-06, "loss": 0.19, "step": 843 }, { "epoch": 0.12747319136082164, "grad_norm": 2.5100316246380294, "learning_rate": 9.95551742108026e-06, "loss": 0.1276, "step": 844 }, { "epoch": 0.12762422594774203, "grad_norm": 3.5469626964698526, "learning_rate": 9.955412106717628e-06, "loss": 0.2, "step": 845 }, { "epoch": 0.12777526053466243, "grad_norm": 4.197717366212537, "learning_rate": 9.955306668392632e-06, "loss": 0.1792, "step": 846 }, { "epoch": 0.12792629512158285, "grad_norm": 4.050051419286757, "learning_rate": 9.955201106107916e-06, "loss": 0.2355, "step": 847 }, { "epoch": 0.12807732970850325, "grad_norm": 3.2335898728352124, "learning_rate": 9.955095419866116e-06, "loss": 0.1702, "step": 848 }, { "epoch": 0.12822836429542364, "grad_norm": 8.347842787399053, "learning_rate": 9.95498960966988e-06, "loss": 0.2017, "step": 849 }, { "epoch": 0.12837939888234406, "grad_norm": 4.656356282014654, "learning_rate": 9.95488367552185e-06, "loss": 0.1842, "step": 850 }, { "epoch": 0.12853043346926446, "grad_norm": 2.5097236730419157, "learning_rate": 9.954777617424682e-06, "loss": 0.1536, "step": 851 }, { "epoch": 0.12868146805618486, "grad_norm": 3.010140291233602, "learning_rate": 9.954671435381025e-06, "loss": 0.1713, "step": 852 }, { "epoch": 0.12883250264310528, "grad_norm": 2.8866977989402804, "learning_rate": 9.954565129393534e-06, "loss": 0.1678, "step": 853 }, { "epoch": 0.12898353723002567, "grad_norm": 2.908419394846907, "learning_rate": 9.954458699464872e-06, "loss": 0.1648, "step": 854 }, { "epoch": 0.12913457181694607, "grad_norm": 2.8138505517880255, "learning_rate": 9.954352145597699e-06, "loss": 0.1515, "step": 855 }, { "epoch": 0.1292856064038665, "grad_norm": 4.1902926958438425, "learning_rate": 9.95424546779468e-06, "loss": 0.1735, "step": 856 }, { "epoch": 0.1294366409907869, "grad_norm": 9.083435552842484, "learning_rate": 9.954138666058489e-06, "loss": 0.1947, "step": 857 }, { "epoch": 0.12958767557770728, "grad_norm": 4.298421291322096, "learning_rate": 9.95403174039179e-06, "loss": 0.1578, "step": 858 }, { "epoch": 0.1297387101646277, "grad_norm": 2.4352286654414166, "learning_rate": 9.953924690797261e-06, "loss": 0.15, "step": 859 }, { "epoch": 0.1298897447515481, "grad_norm": 3.1158907057181127, "learning_rate": 9.953817517277582e-06, "loss": 0.1716, "step": 860 }, { "epoch": 0.1300407793384685, "grad_norm": 3.661321984647423, "learning_rate": 9.953710219835431e-06, "loss": 0.1745, "step": 861 }, { "epoch": 0.13019181392538892, "grad_norm": 3.0063376488967126, "learning_rate": 9.953602798473493e-06, "loss": 0.1696, "step": 862 }, { "epoch": 0.13034284851230932, "grad_norm": 3.0334288778315117, "learning_rate": 9.953495253194457e-06, "loss": 0.1355, "step": 863 }, { "epoch": 0.1304938830992297, "grad_norm": 2.9030152750301172, "learning_rate": 9.953387584001012e-06, "loss": 0.1533, "step": 864 }, { "epoch": 0.13064491768615014, "grad_norm": 2.960848786461452, "learning_rate": 9.95327979089585e-06, "loss": 0.1676, "step": 865 }, { "epoch": 0.13079595227307053, "grad_norm": 2.7529385777891924, "learning_rate": 9.95317187388167e-06, "loss": 0.1712, "step": 866 }, { "epoch": 0.13094698685999093, "grad_norm": 3.4937675058390663, "learning_rate": 9.95306383296117e-06, "loss": 0.1584, "step": 867 }, { "epoch": 0.13109802144691135, "grad_norm": 3.284310839243933, "learning_rate": 9.952955668137052e-06, "loss": 0.1791, "step": 868 }, { "epoch": 0.13124905603383175, "grad_norm": 3.0761815824079495, "learning_rate": 9.952847379412023e-06, "loss": 0.194, "step": 869 }, { "epoch": 0.13140009062075214, "grad_norm": 3.5682332964117354, "learning_rate": 9.952738966788792e-06, "loss": 0.1817, "step": 870 }, { "epoch": 0.13155112520767256, "grad_norm": 2.468289581806737, "learning_rate": 9.952630430270071e-06, "loss": 0.1514, "step": 871 }, { "epoch": 0.13170215979459296, "grad_norm": 3.12223927857685, "learning_rate": 9.952521769858575e-06, "loss": 0.1286, "step": 872 }, { "epoch": 0.13185319438151336, "grad_norm": 2.397851831392024, "learning_rate": 9.952412985557023e-06, "loss": 0.1429, "step": 873 }, { "epoch": 0.13200422896843378, "grad_norm": 2.803929688306375, "learning_rate": 9.952304077368134e-06, "loss": 0.1617, "step": 874 }, { "epoch": 0.13215526355535417, "grad_norm": 2.6776324698095895, "learning_rate": 9.952195045294634e-06, "loss": 0.1589, "step": 875 }, { "epoch": 0.13230629814227457, "grad_norm": 3.5029077867772687, "learning_rate": 9.95208588933925e-06, "loss": 0.191, "step": 876 }, { "epoch": 0.132457332729195, "grad_norm": 2.9071013222554765, "learning_rate": 9.951976609504714e-06, "loss": 0.1556, "step": 877 }, { "epoch": 0.1326083673161154, "grad_norm": 2.4260950971468382, "learning_rate": 9.951867205793758e-06, "loss": 0.155, "step": 878 }, { "epoch": 0.13275940190303578, "grad_norm": 2.905680471112644, "learning_rate": 9.951757678209117e-06, "loss": 0.1722, "step": 879 }, { "epoch": 0.1329104364899562, "grad_norm": 4.69396266565464, "learning_rate": 9.951648026753537e-06, "loss": 0.191, "step": 880 }, { "epoch": 0.1330614710768766, "grad_norm": 2.9687722923550774, "learning_rate": 9.951538251429755e-06, "loss": 0.1486, "step": 881 }, { "epoch": 0.133212505663797, "grad_norm": 2.900452521489787, "learning_rate": 9.95142835224052e-06, "loss": 0.1499, "step": 882 }, { "epoch": 0.13336354025071742, "grad_norm": 4.418583331015882, "learning_rate": 9.951318329188579e-06, "loss": 0.1979, "step": 883 }, { "epoch": 0.13351457483763782, "grad_norm": 2.8191317464322583, "learning_rate": 9.951208182276688e-06, "loss": 0.164, "step": 884 }, { "epoch": 0.1336656094245582, "grad_norm": 3.303813174023111, "learning_rate": 9.9510979115076e-06, "loss": 0.1519, "step": 885 }, { "epoch": 0.13381664401147864, "grad_norm": 2.7685553560923206, "learning_rate": 9.950987516884071e-06, "loss": 0.1598, "step": 886 }, { "epoch": 0.13396767859839903, "grad_norm": 3.1367495908902967, "learning_rate": 9.95087699840887e-06, "loss": 0.168, "step": 887 }, { "epoch": 0.13411871318531943, "grad_norm": 2.8921733365970717, "learning_rate": 9.950766356084753e-06, "loss": 0.136, "step": 888 }, { "epoch": 0.13426974777223985, "grad_norm": 3.4286510376449137, "learning_rate": 9.950655589914493e-06, "loss": 0.1704, "step": 889 }, { "epoch": 0.13442078235916025, "grad_norm": 3.462151744441248, "learning_rate": 9.950544699900859e-06, "loss": 0.1844, "step": 890 }, { "epoch": 0.13457181694608064, "grad_norm": 3.7251752761651016, "learning_rate": 9.950433686046627e-06, "loss": 0.2049, "step": 891 }, { "epoch": 0.13472285153300106, "grad_norm": 3.0288278734397394, "learning_rate": 9.950322548354572e-06, "loss": 0.1712, "step": 892 }, { "epoch": 0.13487388611992146, "grad_norm": 2.990017574508584, "learning_rate": 9.950211286827474e-06, "loss": 0.1463, "step": 893 }, { "epoch": 0.13502492070684186, "grad_norm": 2.5499475550908346, "learning_rate": 9.950099901468118e-06, "loss": 0.1462, "step": 894 }, { "epoch": 0.13517595529376228, "grad_norm": 2.4072020751995495, "learning_rate": 9.949988392279288e-06, "loss": 0.149, "step": 895 }, { "epoch": 0.13532698988068267, "grad_norm": 2.9107535530984485, "learning_rate": 9.949876759263777e-06, "loss": 0.1443, "step": 896 }, { "epoch": 0.13547802446760307, "grad_norm": 3.0605839176960505, "learning_rate": 9.949765002424374e-06, "loss": 0.1611, "step": 897 }, { "epoch": 0.1356290590545235, "grad_norm": 2.7664293837931893, "learning_rate": 9.949653121763877e-06, "loss": 0.1717, "step": 898 }, { "epoch": 0.1357800936414439, "grad_norm": 3.6780754080170737, "learning_rate": 9.949541117285082e-06, "loss": 0.1958, "step": 899 }, { "epoch": 0.13593112822836428, "grad_norm": 3.8658797599787005, "learning_rate": 9.949428988990795e-06, "loss": 0.1903, "step": 900 }, { "epoch": 0.1360821628152847, "grad_norm": 3.3469429718000168, "learning_rate": 9.949316736883817e-06, "loss": 0.2034, "step": 901 }, { "epoch": 0.1362331974022051, "grad_norm": 2.722729309764317, "learning_rate": 9.949204360966959e-06, "loss": 0.1531, "step": 902 }, { "epoch": 0.1363842319891255, "grad_norm": 2.5197771063564205, "learning_rate": 9.94909186124303e-06, "loss": 0.1417, "step": 903 }, { "epoch": 0.13653526657604592, "grad_norm": 2.6527363006765396, "learning_rate": 9.948979237714847e-06, "loss": 0.1687, "step": 904 }, { "epoch": 0.13668630116296632, "grad_norm": 3.0047064970919286, "learning_rate": 9.948866490385223e-06, "loss": 0.1756, "step": 905 }, { "epoch": 0.1368373357498867, "grad_norm": 3.1614990667437857, "learning_rate": 9.94875361925698e-06, "loss": 0.1764, "step": 906 }, { "epoch": 0.13698837033680714, "grad_norm": 2.915649761128876, "learning_rate": 9.948640624332947e-06, "loss": 0.1671, "step": 907 }, { "epoch": 0.13713940492372753, "grad_norm": 2.8881082104250853, "learning_rate": 9.948527505615942e-06, "loss": 0.1669, "step": 908 }, { "epoch": 0.13729043951064793, "grad_norm": 3.2771850403366076, "learning_rate": 9.9484142631088e-06, "loss": 0.1972, "step": 909 }, { "epoch": 0.13744147409756835, "grad_norm": 2.14097000697349, "learning_rate": 9.948300896814354e-06, "loss": 0.1242, "step": 910 }, { "epoch": 0.13759250868448875, "grad_norm": 2.871204648637047, "learning_rate": 9.948187406735437e-06, "loss": 0.1698, "step": 911 }, { "epoch": 0.13774354327140914, "grad_norm": 2.582766771492872, "learning_rate": 9.948073792874889e-06, "loss": 0.1594, "step": 912 }, { "epoch": 0.13789457785832956, "grad_norm": 3.23487598937935, "learning_rate": 9.947960055235553e-06, "loss": 0.1932, "step": 913 }, { "epoch": 0.13804561244524996, "grad_norm": 3.0198547459389715, "learning_rate": 9.947846193820274e-06, "loss": 0.1769, "step": 914 }, { "epoch": 0.13819664703217036, "grad_norm": 3.113931959234162, "learning_rate": 9.947732208631901e-06, "loss": 0.1535, "step": 915 }, { "epoch": 0.13834768161909078, "grad_norm": 3.120188809531193, "learning_rate": 9.947618099673284e-06, "loss": 0.1599, "step": 916 }, { "epoch": 0.13849871620601117, "grad_norm": 2.73255128300267, "learning_rate": 9.947503866947277e-06, "loss": 0.1663, "step": 917 }, { "epoch": 0.13864975079293157, "grad_norm": 16.03223153824161, "learning_rate": 9.94738951045674e-06, "loss": 0.2126, "step": 918 }, { "epoch": 0.138800785379852, "grad_norm": 2.713457539180447, "learning_rate": 9.947275030204533e-06, "loss": 0.1512, "step": 919 }, { "epoch": 0.1389518199667724, "grad_norm": 3.015663855028896, "learning_rate": 9.947160426193519e-06, "loss": 0.1851, "step": 920 }, { "epoch": 0.13910285455369278, "grad_norm": 3.0882349494061936, "learning_rate": 9.947045698426563e-06, "loss": 0.1694, "step": 921 }, { "epoch": 0.1392538891406132, "grad_norm": 2.603368283582819, "learning_rate": 9.94693084690654e-06, "loss": 0.1511, "step": 922 }, { "epoch": 0.1394049237275336, "grad_norm": 3.0796421144532844, "learning_rate": 9.946815871636317e-06, "loss": 0.178, "step": 923 }, { "epoch": 0.139555958314454, "grad_norm": 3.054475474834138, "learning_rate": 9.946700772618774e-06, "loss": 0.1796, "step": 924 }, { "epoch": 0.13970699290137442, "grad_norm": 6.937675161145046, "learning_rate": 9.94658554985679e-06, "loss": 0.207, "step": 925 }, { "epoch": 0.13985802748829482, "grad_norm": 3.25522586585317, "learning_rate": 9.946470203353247e-06, "loss": 0.166, "step": 926 }, { "epoch": 0.1400090620752152, "grad_norm": 4.162262692807451, "learning_rate": 9.946354733111032e-06, "loss": 0.1629, "step": 927 }, { "epoch": 0.14016009666213564, "grad_norm": 3.3520675425086983, "learning_rate": 9.946239139133029e-06, "loss": 0.1626, "step": 928 }, { "epoch": 0.14031113124905603, "grad_norm": 3.483034560974812, "learning_rate": 9.946123421422135e-06, "loss": 0.1876, "step": 929 }, { "epoch": 0.14046216583597643, "grad_norm": 2.718478755818476, "learning_rate": 9.94600757998124e-06, "loss": 0.1606, "step": 930 }, { "epoch": 0.14061320042289685, "grad_norm": 2.662967038550155, "learning_rate": 9.945891614813245e-06, "loss": 0.1513, "step": 931 }, { "epoch": 0.14076423500981725, "grad_norm": 2.8097960679983074, "learning_rate": 9.945775525921052e-06, "loss": 0.154, "step": 932 }, { "epoch": 0.14091526959673764, "grad_norm": 2.8882251317244085, "learning_rate": 9.945659313307561e-06, "loss": 0.1685, "step": 933 }, { "epoch": 0.14106630418365806, "grad_norm": 8.21058896208773, "learning_rate": 9.94554297697568e-06, "loss": 0.2221, "step": 934 }, { "epoch": 0.14121733877057846, "grad_norm": 2.4426506859724335, "learning_rate": 9.945426516928323e-06, "loss": 0.1448, "step": 935 }, { "epoch": 0.14136837335749886, "grad_norm": 3.4619021810289627, "learning_rate": 9.9453099331684e-06, "loss": 0.1647, "step": 936 }, { "epoch": 0.14151940794441928, "grad_norm": 2.4630495636127447, "learning_rate": 9.945193225698825e-06, "loss": 0.1542, "step": 937 }, { "epoch": 0.14167044253133967, "grad_norm": 5.798493592337076, "learning_rate": 9.945076394522526e-06, "loss": 0.1927, "step": 938 }, { "epoch": 0.14182147711826007, "grad_norm": 3.2611263149864214, "learning_rate": 9.944959439642417e-06, "loss": 0.1752, "step": 939 }, { "epoch": 0.1419725117051805, "grad_norm": 3.2011566003822196, "learning_rate": 9.944842361061426e-06, "loss": 0.1824, "step": 940 }, { "epoch": 0.1421235462921009, "grad_norm": 2.9434221961797093, "learning_rate": 9.944725158782485e-06, "loss": 0.1785, "step": 941 }, { "epoch": 0.14227458087902128, "grad_norm": 2.6564950580638675, "learning_rate": 9.944607832808524e-06, "loss": 0.1693, "step": 942 }, { "epoch": 0.1424256154659417, "grad_norm": 3.2037385409655887, "learning_rate": 9.944490383142476e-06, "loss": 0.1757, "step": 943 }, { "epoch": 0.1425766500528621, "grad_norm": 3.0746766742213048, "learning_rate": 9.94437280978728e-06, "loss": 0.1681, "step": 944 }, { "epoch": 0.1427276846397825, "grad_norm": 2.592582720173778, "learning_rate": 9.94425511274588e-06, "loss": 0.1498, "step": 945 }, { "epoch": 0.14287871922670292, "grad_norm": 3.0423262075026662, "learning_rate": 9.944137292021217e-06, "loss": 0.1863, "step": 946 }, { "epoch": 0.14302975381362332, "grad_norm": 3.6524048390794075, "learning_rate": 9.944019347616238e-06, "loss": 0.197, "step": 947 }, { "epoch": 0.1431807884005437, "grad_norm": 5.466703655904563, "learning_rate": 9.943901279533896e-06, "loss": 0.1834, "step": 948 }, { "epoch": 0.14333182298746414, "grad_norm": 2.762107731816732, "learning_rate": 9.943783087777143e-06, "loss": 0.1663, "step": 949 }, { "epoch": 0.14348285757438453, "grad_norm": 2.6674400153882996, "learning_rate": 9.943664772348937e-06, "loss": 0.1648, "step": 950 }, { "epoch": 0.14363389216130493, "grad_norm": 2.6587753922324215, "learning_rate": 9.943546333252236e-06, "loss": 0.1481, "step": 951 }, { "epoch": 0.14378492674822535, "grad_norm": 2.6729077660750247, "learning_rate": 9.943427770490004e-06, "loss": 0.1593, "step": 952 }, { "epoch": 0.14393596133514575, "grad_norm": 3.429096575152585, "learning_rate": 9.943309084065206e-06, "loss": 0.1969, "step": 953 }, { "epoch": 0.14408699592206614, "grad_norm": 3.4062772708010267, "learning_rate": 9.943190273980812e-06, "loss": 0.1835, "step": 954 }, { "epoch": 0.14423803050898656, "grad_norm": 4.040315164531905, "learning_rate": 9.943071340239791e-06, "loss": 0.209, "step": 955 }, { "epoch": 0.14438906509590696, "grad_norm": 4.013653091925079, "learning_rate": 9.942952282845124e-06, "loss": 0.1889, "step": 956 }, { "epoch": 0.14454009968282736, "grad_norm": 3.12075308932894, "learning_rate": 9.942833101799785e-06, "loss": 0.1795, "step": 957 }, { "epoch": 0.14469113426974778, "grad_norm": 3.609961732094463, "learning_rate": 9.942713797106756e-06, "loss": 0.168, "step": 958 }, { "epoch": 0.14484216885666817, "grad_norm": 4.49624270036934, "learning_rate": 9.942594368769023e-06, "loss": 0.1822, "step": 959 }, { "epoch": 0.14499320344358857, "grad_norm": 2.8588509372896493, "learning_rate": 9.94247481678957e-06, "loss": 0.1717, "step": 960 }, { "epoch": 0.145144238030509, "grad_norm": 2.342375435864953, "learning_rate": 9.942355141171393e-06, "loss": 0.1263, "step": 961 }, { "epoch": 0.1452952726174294, "grad_norm": 5.38701987353534, "learning_rate": 9.942235341917482e-06, "loss": 0.2151, "step": 962 }, { "epoch": 0.14544630720434978, "grad_norm": 3.05166610279403, "learning_rate": 9.942115419030835e-06, "loss": 0.1783, "step": 963 }, { "epoch": 0.1455973417912702, "grad_norm": 2.923592522334313, "learning_rate": 9.94199537251445e-06, "loss": 0.1839, "step": 964 }, { "epoch": 0.1457483763781906, "grad_norm": 2.5301861851600562, "learning_rate": 9.941875202371333e-06, "loss": 0.1657, "step": 965 }, { "epoch": 0.145899410965111, "grad_norm": 5.82421989016126, "learning_rate": 9.94175490860449e-06, "loss": 0.2215, "step": 966 }, { "epoch": 0.14605044555203142, "grad_norm": 3.470135565299302, "learning_rate": 9.941634491216927e-06, "loss": 0.1996, "step": 967 }, { "epoch": 0.14620148013895182, "grad_norm": 3.0089952868554644, "learning_rate": 9.941513950211659e-06, "loss": 0.1607, "step": 968 }, { "epoch": 0.1463525147258722, "grad_norm": 7.220198801445822, "learning_rate": 9.9413932855917e-06, "loss": 0.1856, "step": 969 }, { "epoch": 0.14650354931279264, "grad_norm": 3.4043065530874177, "learning_rate": 9.94127249736007e-06, "loss": 0.1731, "step": 970 }, { "epoch": 0.14665458389971303, "grad_norm": 3.2627394917419035, "learning_rate": 9.94115158551979e-06, "loss": 0.191, "step": 971 }, { "epoch": 0.14680561848663343, "grad_norm": 3.394846953645669, "learning_rate": 9.941030550073884e-06, "loss": 0.1855, "step": 972 }, { "epoch": 0.14695665307355385, "grad_norm": 5.088088786353871, "learning_rate": 9.940909391025382e-06, "loss": 0.2484, "step": 973 }, { "epoch": 0.14710768766047425, "grad_norm": 3.5567024184280793, "learning_rate": 9.94078810837731e-06, "loss": 0.184, "step": 974 }, { "epoch": 0.14725872224739464, "grad_norm": 2.677959771226615, "learning_rate": 9.940666702132708e-06, "loss": 0.1438, "step": 975 }, { "epoch": 0.14740975683431506, "grad_norm": 2.8909097373659707, "learning_rate": 9.940545172294608e-06, "loss": 0.1785, "step": 976 }, { "epoch": 0.14756079142123546, "grad_norm": 3.022694135725918, "learning_rate": 9.940423518866053e-06, "loss": 0.1579, "step": 977 }, { "epoch": 0.14771182600815586, "grad_norm": 4.5111865158440265, "learning_rate": 9.940301741850086e-06, "loss": 0.2254, "step": 978 }, { "epoch": 0.14786286059507628, "grad_norm": 3.0989405296693198, "learning_rate": 9.940179841249751e-06, "loss": 0.1929, "step": 979 }, { "epoch": 0.14801389518199667, "grad_norm": 6.04761825400586, "learning_rate": 9.9400578170681e-06, "loss": 0.1914, "step": 980 }, { "epoch": 0.14816492976891707, "grad_norm": 2.7128277624768327, "learning_rate": 9.939935669308188e-06, "loss": 0.1627, "step": 981 }, { "epoch": 0.1483159643558375, "grad_norm": 2.682484172381204, "learning_rate": 9.939813397973063e-06, "loss": 0.1654, "step": 982 }, { "epoch": 0.1484669989427579, "grad_norm": 3.6504088259554828, "learning_rate": 9.93969100306579e-06, "loss": 0.2077, "step": 983 }, { "epoch": 0.14861803352967828, "grad_norm": 3.1060411195984865, "learning_rate": 9.939568484589427e-06, "loss": 0.1957, "step": 984 }, { "epoch": 0.1487690681165987, "grad_norm": 2.548057874860104, "learning_rate": 9.939445842547042e-06, "loss": 0.1452, "step": 985 }, { "epoch": 0.1489201027035191, "grad_norm": 3.0980449489524076, "learning_rate": 9.939323076941701e-06, "loss": 0.1822, "step": 986 }, { "epoch": 0.1490711372904395, "grad_norm": 2.2895362518383666, "learning_rate": 9.939200187776477e-06, "loss": 0.1425, "step": 987 }, { "epoch": 0.14922217187735992, "grad_norm": 2.7723354304666317, "learning_rate": 9.93907717505444e-06, "loss": 0.1819, "step": 988 }, { "epoch": 0.14937320646428032, "grad_norm": 3.1534101087543016, "learning_rate": 9.938954038778673e-06, "loss": 0.1356, "step": 989 }, { "epoch": 0.1495242410512007, "grad_norm": 3.6117400236385753, "learning_rate": 9.938830778952253e-06, "loss": 0.1624, "step": 990 }, { "epoch": 0.14967527563812114, "grad_norm": 2.87387451067728, "learning_rate": 9.938707395578262e-06, "loss": 0.1482, "step": 991 }, { "epoch": 0.14982631022504153, "grad_norm": 3.6972374905271246, "learning_rate": 9.93858388865979e-06, "loss": 0.1526, "step": 992 }, { "epoch": 0.14997734481196193, "grad_norm": 2.9800074456821437, "learning_rate": 9.938460258199924e-06, "loss": 0.1505, "step": 993 }, { "epoch": 0.15012837939888235, "grad_norm": 3.413664852504228, "learning_rate": 9.938336504201757e-06, "loss": 0.1773, "step": 994 }, { "epoch": 0.15027941398580275, "grad_norm": 2.354861104613853, "learning_rate": 9.938212626668385e-06, "loss": 0.1494, "step": 995 }, { "epoch": 0.15043044857272314, "grad_norm": 2.584561755103005, "learning_rate": 9.938088625602908e-06, "loss": 0.1705, "step": 996 }, { "epoch": 0.15058148315964356, "grad_norm": 2.8278713255595744, "learning_rate": 9.937964501008428e-06, "loss": 0.1595, "step": 997 }, { "epoch": 0.15073251774656396, "grad_norm": 3.4034959670004894, "learning_rate": 9.937840252888047e-06, "loss": 0.1649, "step": 998 }, { "epoch": 0.15088355233348436, "grad_norm": 2.791817531880729, "learning_rate": 9.937715881244876e-06, "loss": 0.1683, "step": 999 }, { "epoch": 0.15103458692040478, "grad_norm": 3.009720714067996, "learning_rate": 9.937591386082027e-06, "loss": 0.1576, "step": 1000 }, { "epoch": 0.15118562150732517, "grad_norm": 2.998917119914494, "learning_rate": 9.937466767402612e-06, "loss": 0.1715, "step": 1001 }, { "epoch": 0.15133665609424557, "grad_norm": 3.426218348223617, "learning_rate": 9.937342025209749e-06, "loss": 0.1752, "step": 1002 }, { "epoch": 0.151487690681166, "grad_norm": 2.6634302202648663, "learning_rate": 9.937217159506558e-06, "loss": 0.1433, "step": 1003 }, { "epoch": 0.1516387252680864, "grad_norm": 2.387809137519754, "learning_rate": 9.937092170296165e-06, "loss": 0.1436, "step": 1004 }, { "epoch": 0.15178975985500678, "grad_norm": 3.64178360722794, "learning_rate": 9.936967057581694e-06, "loss": 0.2052, "step": 1005 }, { "epoch": 0.1519407944419272, "grad_norm": 2.5650151287284877, "learning_rate": 9.936841821366275e-06, "loss": 0.1532, "step": 1006 }, { "epoch": 0.1520918290288476, "grad_norm": 3.4858222664891936, "learning_rate": 9.936716461653042e-06, "loss": 0.1764, "step": 1007 }, { "epoch": 0.152242863615768, "grad_norm": 2.4438728691660296, "learning_rate": 9.936590978445132e-06, "loss": 0.1291, "step": 1008 }, { "epoch": 0.15239389820268842, "grad_norm": 2.738301470143143, "learning_rate": 9.936465371745682e-06, "loss": 0.1755, "step": 1009 }, { "epoch": 0.15254493278960882, "grad_norm": 3.283939998462775, "learning_rate": 9.936339641557833e-06, "loss": 0.1938, "step": 1010 }, { "epoch": 0.1526959673765292, "grad_norm": 2.5255181547062673, "learning_rate": 9.936213787884734e-06, "loss": 0.1561, "step": 1011 }, { "epoch": 0.15284700196344964, "grad_norm": 2.8005198915457115, "learning_rate": 9.93608781072953e-06, "loss": 0.175, "step": 1012 }, { "epoch": 0.15299803655037003, "grad_norm": 3.930185281317105, "learning_rate": 9.935961710095375e-06, "loss": 0.1799, "step": 1013 }, { "epoch": 0.15314907113729043, "grad_norm": 2.9425671176461816, "learning_rate": 9.935835485985421e-06, "loss": 0.1769, "step": 1014 }, { "epoch": 0.15330010572421085, "grad_norm": 6.078209910811355, "learning_rate": 9.935709138402825e-06, "loss": 0.1772, "step": 1015 }, { "epoch": 0.15345114031113125, "grad_norm": 2.924102076335692, "learning_rate": 9.935582667350751e-06, "loss": 0.1653, "step": 1016 }, { "epoch": 0.15360217489805164, "grad_norm": 3.5989659971161925, "learning_rate": 9.935456072832361e-06, "loss": 0.1814, "step": 1017 }, { "epoch": 0.15375320948497206, "grad_norm": 3.5428681786351532, "learning_rate": 9.935329354850822e-06, "loss": 0.1851, "step": 1018 }, { "epoch": 0.15390424407189246, "grad_norm": 3.1210824280022607, "learning_rate": 9.935202513409304e-06, "loss": 0.197, "step": 1019 }, { "epoch": 0.15405527865881286, "grad_norm": 3.0652889781186317, "learning_rate": 9.93507554851098e-06, "loss": 0.1836, "step": 1020 }, { "epoch": 0.15420631324573328, "grad_norm": 2.8175558889126795, "learning_rate": 9.934948460159024e-06, "loss": 0.1725, "step": 1021 }, { "epoch": 0.15435734783265367, "grad_norm": 2.933140598167281, "learning_rate": 9.934821248356619e-06, "loss": 0.1681, "step": 1022 }, { "epoch": 0.15450838241957407, "grad_norm": 2.401765280358054, "learning_rate": 9.934693913106946e-06, "loss": 0.1554, "step": 1023 }, { "epoch": 0.1546594170064945, "grad_norm": 3.3064162119048635, "learning_rate": 9.934566454413188e-06, "loss": 0.1756, "step": 1024 }, { "epoch": 0.1548104515934149, "grad_norm": 3.6386661352486813, "learning_rate": 9.934438872278535e-06, "loss": 0.171, "step": 1025 }, { "epoch": 0.15496148618033528, "grad_norm": 2.7695925199576577, "learning_rate": 9.934311166706178e-06, "loss": 0.1461, "step": 1026 }, { "epoch": 0.1551125207672557, "grad_norm": 2.9184813175520534, "learning_rate": 9.934183337699314e-06, "loss": 0.1625, "step": 1027 }, { "epoch": 0.1552635553541761, "grad_norm": 3.293546073587853, "learning_rate": 9.934055385261139e-06, "loss": 0.153, "step": 1028 }, { "epoch": 0.1554145899410965, "grad_norm": 2.528295571851272, "learning_rate": 9.933927309394854e-06, "loss": 0.163, "step": 1029 }, { "epoch": 0.15556562452801692, "grad_norm": 2.968822086330793, "learning_rate": 9.933799110103662e-06, "loss": 0.1771, "step": 1030 }, { "epoch": 0.15571665911493732, "grad_norm": 2.7803911565214214, "learning_rate": 9.933670787390771e-06, "loss": 0.1651, "step": 1031 }, { "epoch": 0.1558676937018577, "grad_norm": 2.9263288413260407, "learning_rate": 9.93354234125939e-06, "loss": 0.182, "step": 1032 }, { "epoch": 0.15601872828877814, "grad_norm": 2.73331445705435, "learning_rate": 9.933413771712734e-06, "loss": 0.1486, "step": 1033 }, { "epoch": 0.15616976287569853, "grad_norm": 3.8302330591926914, "learning_rate": 9.933285078754017e-06, "loss": 0.1954, "step": 1034 }, { "epoch": 0.15632079746261893, "grad_norm": 2.2028518140450415, "learning_rate": 9.933156262386461e-06, "loss": 0.1459, "step": 1035 }, { "epoch": 0.15647183204953935, "grad_norm": 3.0489605978719836, "learning_rate": 9.933027322613286e-06, "loss": 0.1774, "step": 1036 }, { "epoch": 0.15662286663645975, "grad_norm": 4.33102605188305, "learning_rate": 9.93289825943772e-06, "loss": 0.2086, "step": 1037 }, { "epoch": 0.15677390122338014, "grad_norm": 3.0077164166121, "learning_rate": 9.932769072862988e-06, "loss": 0.1789, "step": 1038 }, { "epoch": 0.15692493581030056, "grad_norm": 2.9436524448612307, "learning_rate": 9.932639762892325e-06, "loss": 0.1815, "step": 1039 }, { "epoch": 0.15707597039722096, "grad_norm": 3.5804809216960436, "learning_rate": 9.932510329528962e-06, "loss": 0.1951, "step": 1040 }, { "epoch": 0.15722700498414136, "grad_norm": 3.2350705746458646, "learning_rate": 9.93238077277614e-06, "loss": 0.1708, "step": 1041 }, { "epoch": 0.15737803957106178, "grad_norm": 3.065218093296157, "learning_rate": 9.932251092637102e-06, "loss": 0.1821, "step": 1042 }, { "epoch": 0.15752907415798217, "grad_norm": 2.7768041844203624, "learning_rate": 9.932121289115086e-06, "loss": 0.1817, "step": 1043 }, { "epoch": 0.15768010874490257, "grad_norm": 2.396990686761492, "learning_rate": 9.931991362213343e-06, "loss": 0.1526, "step": 1044 }, { "epoch": 0.157831143331823, "grad_norm": 4.122186728988527, "learning_rate": 9.931861311935123e-06, "loss": 0.1505, "step": 1045 }, { "epoch": 0.1579821779187434, "grad_norm": 2.5381348096848075, "learning_rate": 9.93173113828368e-06, "loss": 0.1473, "step": 1046 }, { "epoch": 0.15813321250566378, "grad_norm": 2.788937733382081, "learning_rate": 9.931600841262268e-06, "loss": 0.1476, "step": 1047 }, { "epoch": 0.1582842470925842, "grad_norm": 2.5274298575775815, "learning_rate": 9.931470420874147e-06, "loss": 0.1533, "step": 1048 }, { "epoch": 0.1584352816795046, "grad_norm": 3.019133600261103, "learning_rate": 9.93133987712258e-06, "loss": 0.1445, "step": 1049 }, { "epoch": 0.158586316266425, "grad_norm": 2.780354997678927, "learning_rate": 9.931209210010835e-06, "loss": 0.1636, "step": 1050 }, { "epoch": 0.15873735085334542, "grad_norm": 2.9223742343244052, "learning_rate": 9.931078419542177e-06, "loss": 0.1921, "step": 1051 }, { "epoch": 0.15888838544026582, "grad_norm": 3.582967219279585, "learning_rate": 9.93094750571988e-06, "loss": 0.1586, "step": 1052 }, { "epoch": 0.1590394200271862, "grad_norm": 2.9126130850022065, "learning_rate": 9.930816468547215e-06, "loss": 0.1708, "step": 1053 }, { "epoch": 0.15919045461410664, "grad_norm": 2.993162282576796, "learning_rate": 9.930685308027465e-06, "loss": 0.1547, "step": 1054 }, { "epoch": 0.15934148920102703, "grad_norm": 2.8494582438839546, "learning_rate": 9.93055402416391e-06, "loss": 0.1575, "step": 1055 }, { "epoch": 0.15949252378794743, "grad_norm": 3.4290753778346117, "learning_rate": 9.930422616959835e-06, "loss": 0.1801, "step": 1056 }, { "epoch": 0.15964355837486785, "grad_norm": 3.2682305685445807, "learning_rate": 9.930291086418522e-06, "loss": 0.1805, "step": 1057 }, { "epoch": 0.15979459296178825, "grad_norm": 2.6189284692833916, "learning_rate": 9.930159432543268e-06, "loss": 0.1493, "step": 1058 }, { "epoch": 0.15994562754870864, "grad_norm": 3.1710481313960606, "learning_rate": 9.93002765533736e-06, "loss": 0.1822, "step": 1059 }, { "epoch": 0.16009666213562906, "grad_norm": 2.53752838507505, "learning_rate": 9.9298957548041e-06, "loss": 0.1455, "step": 1060 }, { "epoch": 0.16024769672254946, "grad_norm": 3.3876470886220225, "learning_rate": 9.929763730946783e-06, "loss": 0.1849, "step": 1061 }, { "epoch": 0.16039873130946986, "grad_norm": 2.6864224757528246, "learning_rate": 9.929631583768718e-06, "loss": 0.1557, "step": 1062 }, { "epoch": 0.16054976589639028, "grad_norm": 2.23588181145321, "learning_rate": 9.929499313273204e-06, "loss": 0.1313, "step": 1063 }, { "epoch": 0.16070080048331067, "grad_norm": 3.0758283157527733, "learning_rate": 9.929366919463552e-06, "loss": 0.1477, "step": 1064 }, { "epoch": 0.16085183507023107, "grad_norm": 2.6055179447849577, "learning_rate": 9.929234402343075e-06, "loss": 0.1757, "step": 1065 }, { "epoch": 0.1610028696571515, "grad_norm": 2.7625401701478656, "learning_rate": 9.929101761915088e-06, "loss": 0.1748, "step": 1066 }, { "epoch": 0.1611539042440719, "grad_norm": 4.426546974451077, "learning_rate": 9.928968998182905e-06, "loss": 0.1704, "step": 1067 }, { "epoch": 0.16130493883099228, "grad_norm": 2.4221787557402132, "learning_rate": 9.928836111149856e-06, "loss": 0.1418, "step": 1068 }, { "epoch": 0.1614559734179127, "grad_norm": 3.065163976139544, "learning_rate": 9.928703100819257e-06, "loss": 0.2144, "step": 1069 }, { "epoch": 0.1616070080048331, "grad_norm": 3.122036885904642, "learning_rate": 9.928569967194437e-06, "loss": 0.1875, "step": 1070 }, { "epoch": 0.1617580425917535, "grad_norm": 2.8022628014772564, "learning_rate": 9.92843671027873e-06, "loss": 0.1353, "step": 1071 }, { "epoch": 0.16190907717867392, "grad_norm": 2.586758977126138, "learning_rate": 9.928303330075466e-06, "loss": 0.1492, "step": 1072 }, { "epoch": 0.16206011176559432, "grad_norm": 3.118207460088234, "learning_rate": 9.928169826587984e-06, "loss": 0.1748, "step": 1073 }, { "epoch": 0.1622111463525147, "grad_norm": 2.4988687847418696, "learning_rate": 9.928036199819621e-06, "loss": 0.1582, "step": 1074 }, { "epoch": 0.16236218093943514, "grad_norm": 2.3834121617246082, "learning_rate": 9.927902449773722e-06, "loss": 0.1534, "step": 1075 }, { "epoch": 0.16251321552635553, "grad_norm": 3.5895779156551746, "learning_rate": 9.92776857645363e-06, "loss": 0.1683, "step": 1076 }, { "epoch": 0.16266425011327593, "grad_norm": 2.891953671292175, "learning_rate": 9.927634579862697e-06, "loss": 0.1788, "step": 1077 }, { "epoch": 0.16281528470019635, "grad_norm": 3.309995587467627, "learning_rate": 9.927500460004274e-06, "loss": 0.1777, "step": 1078 }, { "epoch": 0.16296631928711675, "grad_norm": 4.266577301150299, "learning_rate": 9.927366216881715e-06, "loss": 0.1633, "step": 1079 }, { "epoch": 0.16311735387403714, "grad_norm": 7.803866414175067, "learning_rate": 9.92723185049838e-06, "loss": 0.2662, "step": 1080 }, { "epoch": 0.16326838846095756, "grad_norm": 3.314464721426022, "learning_rate": 9.92709736085763e-06, "loss": 0.1947, "step": 1081 }, { "epoch": 0.16341942304787796, "grad_norm": 2.8931955345782394, "learning_rate": 9.926962747962828e-06, "loss": 0.1622, "step": 1082 }, { "epoch": 0.16357045763479836, "grad_norm": 2.8496234175636803, "learning_rate": 9.92682801181734e-06, "loss": 0.187, "step": 1083 }, { "epoch": 0.16372149222171878, "grad_norm": 2.645406532721558, "learning_rate": 9.92669315242454e-06, "loss": 0.1541, "step": 1084 }, { "epoch": 0.16387252680863917, "grad_norm": 3.4181252175778414, "learning_rate": 9.926558169787801e-06, "loss": 0.166, "step": 1085 }, { "epoch": 0.16402356139555957, "grad_norm": 2.8441587378777147, "learning_rate": 9.9264230639105e-06, "loss": 0.1675, "step": 1086 }, { "epoch": 0.16417459598248, "grad_norm": 2.9770583230803367, "learning_rate": 9.926287834796013e-06, "loss": 0.1597, "step": 1087 }, { "epoch": 0.1643256305694004, "grad_norm": 2.969727240073377, "learning_rate": 9.926152482447726e-06, "loss": 0.1605, "step": 1088 }, { "epoch": 0.16447666515632078, "grad_norm": 3.0004967101587643, "learning_rate": 9.926017006869025e-06, "loss": 0.1682, "step": 1089 }, { "epoch": 0.1646276997432412, "grad_norm": 2.8016755669208147, "learning_rate": 9.925881408063297e-06, "loss": 0.1629, "step": 1090 }, { "epoch": 0.1647787343301616, "grad_norm": 2.962478196407774, "learning_rate": 9.925745686033936e-06, "loss": 0.1803, "step": 1091 }, { "epoch": 0.164929768917082, "grad_norm": 3.202193401611355, "learning_rate": 9.925609840784336e-06, "loss": 0.1869, "step": 1092 }, { "epoch": 0.16508080350400242, "grad_norm": 2.4536892453297514, "learning_rate": 9.925473872317897e-06, "loss": 0.1534, "step": 1093 }, { "epoch": 0.16523183809092282, "grad_norm": 3.9699929445068345, "learning_rate": 9.925337780638018e-06, "loss": 0.1736, "step": 1094 }, { "epoch": 0.1653828726778432, "grad_norm": 2.842417823632091, "learning_rate": 9.925201565748105e-06, "loss": 0.1744, "step": 1095 }, { "epoch": 0.16553390726476364, "grad_norm": 2.4754100332917077, "learning_rate": 9.925065227651565e-06, "loss": 0.1663, "step": 1096 }, { "epoch": 0.16568494185168403, "grad_norm": 2.911355335618653, "learning_rate": 9.92492876635181e-06, "loss": 0.1811, "step": 1097 }, { "epoch": 0.16583597643860443, "grad_norm": 3.0913273760226203, "learning_rate": 9.92479218185225e-06, "loss": 0.1603, "step": 1098 }, { "epoch": 0.16598701102552485, "grad_norm": 2.7424806280240133, "learning_rate": 9.924655474156305e-06, "loss": 0.1727, "step": 1099 }, { "epoch": 0.16613804561244525, "grad_norm": 2.3001897897497225, "learning_rate": 9.924518643267393e-06, "loss": 0.1301, "step": 1100 }, { "epoch": 0.16628908019936564, "grad_norm": 2.992393634185184, "learning_rate": 9.92438168918894e-06, "loss": 0.156, "step": 1101 }, { "epoch": 0.16644011478628606, "grad_norm": 3.003684419361946, "learning_rate": 9.924244611924368e-06, "loss": 0.1663, "step": 1102 }, { "epoch": 0.16659114937320646, "grad_norm": 4.7370659345776165, "learning_rate": 9.924107411477106e-06, "loss": 0.2023, "step": 1103 }, { "epoch": 0.16674218396012688, "grad_norm": 2.5269252627430805, "learning_rate": 9.923970087850591e-06, "loss": 0.1491, "step": 1104 }, { "epoch": 0.16689321854704728, "grad_norm": 2.3955180603916473, "learning_rate": 9.923832641048254e-06, "loss": 0.1452, "step": 1105 }, { "epoch": 0.16704425313396767, "grad_norm": 2.9904790025546353, "learning_rate": 9.923695071073534e-06, "loss": 0.1765, "step": 1106 }, { "epoch": 0.1671952877208881, "grad_norm": 2.902872442913884, "learning_rate": 9.923557377929872e-06, "loss": 0.188, "step": 1107 }, { "epoch": 0.1673463223078085, "grad_norm": 2.547760141552346, "learning_rate": 9.923419561620716e-06, "loss": 0.1706, "step": 1108 }, { "epoch": 0.1674973568947289, "grad_norm": 2.854360379997613, "learning_rate": 9.923281622149508e-06, "loss": 0.1787, "step": 1109 }, { "epoch": 0.1676483914816493, "grad_norm": 2.610913518107109, "learning_rate": 9.923143559519701e-06, "loss": 0.1595, "step": 1110 }, { "epoch": 0.1677994260685697, "grad_norm": 2.674139546584187, "learning_rate": 9.923005373734752e-06, "loss": 0.1611, "step": 1111 }, { "epoch": 0.1679504606554901, "grad_norm": 4.761333745799679, "learning_rate": 9.922867064798113e-06, "loss": 0.3881, "step": 1112 }, { "epoch": 0.16810149524241053, "grad_norm": 2.9330542641771364, "learning_rate": 9.922728632713246e-06, "loss": 0.1727, "step": 1113 }, { "epoch": 0.16825252982933092, "grad_norm": 3.817458521697432, "learning_rate": 9.922590077483614e-06, "loss": 0.186, "step": 1114 }, { "epoch": 0.16840356441625132, "grad_norm": 2.9809120057753793, "learning_rate": 9.922451399112683e-06, "loss": 0.1833, "step": 1115 }, { "epoch": 0.16855459900317174, "grad_norm": 2.840345965181265, "learning_rate": 9.922312597603922e-06, "loss": 0.1849, "step": 1116 }, { "epoch": 0.16870563359009214, "grad_norm": 2.299019007311814, "learning_rate": 9.922173672960803e-06, "loss": 0.155, "step": 1117 }, { "epoch": 0.16885666817701253, "grad_norm": 3.272986329828003, "learning_rate": 9.922034625186801e-06, "loss": 0.1673, "step": 1118 }, { "epoch": 0.16900770276393295, "grad_norm": 2.785000910546995, "learning_rate": 9.921895454285396e-06, "loss": 0.1539, "step": 1119 }, { "epoch": 0.16915873735085335, "grad_norm": 3.014629684951083, "learning_rate": 9.921756160260067e-06, "loss": 0.1761, "step": 1120 }, { "epoch": 0.16930977193777375, "grad_norm": 2.6466907826190447, "learning_rate": 9.9216167431143e-06, "loss": 0.1564, "step": 1121 }, { "epoch": 0.16946080652469417, "grad_norm": 3.9876831312533865, "learning_rate": 9.921477202851584e-06, "loss": 0.1679, "step": 1122 }, { "epoch": 0.16961184111161456, "grad_norm": 3.0333925017872776, "learning_rate": 9.921337539475405e-06, "loss": 0.1764, "step": 1123 }, { "epoch": 0.16976287569853496, "grad_norm": 2.7494486750593765, "learning_rate": 9.921197752989262e-06, "loss": 0.1607, "step": 1124 }, { "epoch": 0.16991391028545538, "grad_norm": 2.7609532243611388, "learning_rate": 9.921057843396648e-06, "loss": 0.1707, "step": 1125 }, { "epoch": 0.17006494487237578, "grad_norm": 3.998858403608429, "learning_rate": 9.920917810701065e-06, "loss": 0.175, "step": 1126 }, { "epoch": 0.17021597945929617, "grad_norm": 3.381217316925807, "learning_rate": 9.920777654906016e-06, "loss": 0.1415, "step": 1127 }, { "epoch": 0.1703670140462166, "grad_norm": 3.2480428306039175, "learning_rate": 9.920637376015007e-06, "loss": 0.169, "step": 1128 }, { "epoch": 0.170518048633137, "grad_norm": 3.3121490521266446, "learning_rate": 9.920496974031546e-06, "loss": 0.1646, "step": 1129 }, { "epoch": 0.1706690832200574, "grad_norm": 2.7864478523233, "learning_rate": 9.920356448959145e-06, "loss": 0.1759, "step": 1130 }, { "epoch": 0.1708201178069778, "grad_norm": 3.0731582478438013, "learning_rate": 9.920215800801321e-06, "loss": 0.1589, "step": 1131 }, { "epoch": 0.1709711523938982, "grad_norm": 2.547596382282077, "learning_rate": 9.920075029561594e-06, "loss": 0.1579, "step": 1132 }, { "epoch": 0.1711221869808186, "grad_norm": 2.5736494795203995, "learning_rate": 9.91993413524348e-06, "loss": 0.1598, "step": 1133 }, { "epoch": 0.17127322156773903, "grad_norm": 2.6854343562576246, "learning_rate": 9.919793117850506e-06, "loss": 0.1695, "step": 1134 }, { "epoch": 0.17142425615465942, "grad_norm": 2.6641303600936546, "learning_rate": 9.919651977386204e-06, "loss": 0.1566, "step": 1135 }, { "epoch": 0.17157529074157982, "grad_norm": 3.1332874979877854, "learning_rate": 9.919510713854099e-06, "loss": 0.1834, "step": 1136 }, { "epoch": 0.17172632532850024, "grad_norm": 2.941863911409242, "learning_rate": 9.919369327257726e-06, "loss": 0.1705, "step": 1137 }, { "epoch": 0.17187735991542064, "grad_norm": 3.321358879653198, "learning_rate": 9.919227817600625e-06, "loss": 0.1666, "step": 1138 }, { "epoch": 0.17202839450234103, "grad_norm": 4.828360132582347, "learning_rate": 9.919086184886331e-06, "loss": 0.1918, "step": 1139 }, { "epoch": 0.17217942908926145, "grad_norm": 2.682339919099411, "learning_rate": 9.918944429118392e-06, "loss": 0.1376, "step": 1140 }, { "epoch": 0.17233046367618185, "grad_norm": 3.673682818582982, "learning_rate": 9.918802550300348e-06, "loss": 0.1852, "step": 1141 }, { "epoch": 0.17248149826310225, "grad_norm": 2.6343689183585717, "learning_rate": 9.918660548435755e-06, "loss": 0.1579, "step": 1142 }, { "epoch": 0.17263253285002267, "grad_norm": 2.898534120238209, "learning_rate": 9.918518423528162e-06, "loss": 0.1626, "step": 1143 }, { "epoch": 0.17278356743694306, "grad_norm": 2.89100629989136, "learning_rate": 9.918376175581123e-06, "loss": 0.1903, "step": 1144 }, { "epoch": 0.17293460202386346, "grad_norm": 4.027041466616999, "learning_rate": 9.918233804598199e-06, "loss": 0.2042, "step": 1145 }, { "epoch": 0.17308563661078388, "grad_norm": 2.4189886131576257, "learning_rate": 9.918091310582951e-06, "loss": 0.146, "step": 1146 }, { "epoch": 0.17323667119770428, "grad_norm": 2.7450345107283214, "learning_rate": 9.91794869353894e-06, "loss": 0.1397, "step": 1147 }, { "epoch": 0.17338770578462467, "grad_norm": 2.628001079521027, "learning_rate": 9.91780595346974e-06, "loss": 0.1467, "step": 1148 }, { "epoch": 0.1735387403715451, "grad_norm": 2.6441414772081346, "learning_rate": 9.917663090378917e-06, "loss": 0.1651, "step": 1149 }, { "epoch": 0.1736897749584655, "grad_norm": 3.8496364035013952, "learning_rate": 9.917520104270046e-06, "loss": 0.1668, "step": 1150 }, { "epoch": 0.1738408095453859, "grad_norm": 3.4970370168513076, "learning_rate": 9.917376995146704e-06, "loss": 0.181, "step": 1151 }, { "epoch": 0.1739918441323063, "grad_norm": 9.872332947537764, "learning_rate": 9.917233763012472e-06, "loss": 0.2508, "step": 1152 }, { "epoch": 0.1741428787192267, "grad_norm": 3.8093766672276437, "learning_rate": 9.917090407870931e-06, "loss": 0.1692, "step": 1153 }, { "epoch": 0.1742939133061471, "grad_norm": 2.9940890447356323, "learning_rate": 9.916946929725668e-06, "loss": 0.1816, "step": 1154 }, { "epoch": 0.17444494789306753, "grad_norm": 2.400423322253169, "learning_rate": 9.916803328580273e-06, "loss": 0.1571, "step": 1155 }, { "epoch": 0.17459598247998792, "grad_norm": 2.777092120471612, "learning_rate": 9.916659604438336e-06, "loss": 0.1661, "step": 1156 }, { "epoch": 0.17474701706690832, "grad_norm": 4.927651048162132, "learning_rate": 9.916515757303456e-06, "loss": 0.1906, "step": 1157 }, { "epoch": 0.17489805165382874, "grad_norm": 3.437275140038844, "learning_rate": 9.916371787179227e-06, "loss": 0.2071, "step": 1158 }, { "epoch": 0.17504908624074914, "grad_norm": 2.9465498244009334, "learning_rate": 9.916227694069253e-06, "loss": 0.1872, "step": 1159 }, { "epoch": 0.17520012082766953, "grad_norm": 2.5432306547915204, "learning_rate": 9.916083477977141e-06, "loss": 0.1566, "step": 1160 }, { "epoch": 0.17535115541458995, "grad_norm": 3.7914739712737706, "learning_rate": 9.915939138906494e-06, "loss": 0.1861, "step": 1161 }, { "epoch": 0.17550219000151035, "grad_norm": 2.8977409032791597, "learning_rate": 9.915794676860924e-06, "loss": 0.1534, "step": 1162 }, { "epoch": 0.17565322458843075, "grad_norm": 3.3514455783742627, "learning_rate": 9.915650091844047e-06, "loss": 0.1866, "step": 1163 }, { "epoch": 0.17580425917535117, "grad_norm": 3.1568500941758, "learning_rate": 9.91550538385948e-06, "loss": 0.1814, "step": 1164 }, { "epoch": 0.17595529376227156, "grad_norm": 2.7950800200663086, "learning_rate": 9.915360552910837e-06, "loss": 0.1611, "step": 1165 }, { "epoch": 0.17610632834919196, "grad_norm": 2.852035289712551, "learning_rate": 9.915215599001748e-06, "loss": 0.1543, "step": 1166 }, { "epoch": 0.17625736293611238, "grad_norm": 3.417975868831882, "learning_rate": 9.915070522135834e-06, "loss": 0.1848, "step": 1167 }, { "epoch": 0.17640839752303278, "grad_norm": 2.5823708898367705, "learning_rate": 9.91492532231673e-06, "loss": 0.164, "step": 1168 }, { "epoch": 0.17655943210995317, "grad_norm": 4.301937309465748, "learning_rate": 9.914779999548063e-06, "loss": 0.1617, "step": 1169 }, { "epoch": 0.1767104666968736, "grad_norm": 3.2399992281351278, "learning_rate": 9.914634553833471e-06, "loss": 0.1588, "step": 1170 }, { "epoch": 0.176861501283794, "grad_norm": 2.8865933937005512, "learning_rate": 9.914488985176591e-06, "loss": 0.1669, "step": 1171 }, { "epoch": 0.1770125358707144, "grad_norm": 3.599648493307587, "learning_rate": 9.914343293581066e-06, "loss": 0.1806, "step": 1172 }, { "epoch": 0.1771635704576348, "grad_norm": 4.178868427728367, "learning_rate": 9.914197479050538e-06, "loss": 0.1926, "step": 1173 }, { "epoch": 0.1773146050445552, "grad_norm": 2.764610500611124, "learning_rate": 9.914051541588657e-06, "loss": 0.15, "step": 1174 }, { "epoch": 0.1774656396314756, "grad_norm": 3.1994443815270373, "learning_rate": 9.913905481199074e-06, "loss": 0.1639, "step": 1175 }, { "epoch": 0.17761667421839603, "grad_norm": 4.046405698144353, "learning_rate": 9.91375929788544e-06, "loss": 0.1838, "step": 1176 }, { "epoch": 0.17776770880531642, "grad_norm": 3.2957466280497094, "learning_rate": 9.913612991651416e-06, "loss": 0.1665, "step": 1177 }, { "epoch": 0.17791874339223682, "grad_norm": 3.081681341360573, "learning_rate": 9.913466562500658e-06, "loss": 0.1673, "step": 1178 }, { "epoch": 0.17806977797915724, "grad_norm": 4.331232415281952, "learning_rate": 9.913320010436832e-06, "loss": 0.2041, "step": 1179 }, { "epoch": 0.17822081256607764, "grad_norm": 3.144841284695798, "learning_rate": 9.913173335463599e-06, "loss": 0.1746, "step": 1180 }, { "epoch": 0.17837184715299803, "grad_norm": 3.1375678683928094, "learning_rate": 9.913026537584636e-06, "loss": 0.1719, "step": 1181 }, { "epoch": 0.17852288173991845, "grad_norm": 2.9944419021691293, "learning_rate": 9.912879616803607e-06, "loss": 0.1529, "step": 1182 }, { "epoch": 0.17867391632683885, "grad_norm": 2.6658825437295994, "learning_rate": 9.912732573124195e-06, "loss": 0.1719, "step": 1183 }, { "epoch": 0.17882495091375925, "grad_norm": 3.4933031750398085, "learning_rate": 9.912585406550073e-06, "loss": 0.174, "step": 1184 }, { "epoch": 0.17897598550067967, "grad_norm": 2.823116098373148, "learning_rate": 9.912438117084925e-06, "loss": 0.1624, "step": 1185 }, { "epoch": 0.17912702008760006, "grad_norm": 2.7446309026507256, "learning_rate": 9.912290704732435e-06, "loss": 0.1596, "step": 1186 }, { "epoch": 0.17927805467452046, "grad_norm": 2.9021445083817157, "learning_rate": 9.912143169496289e-06, "loss": 0.1778, "step": 1187 }, { "epoch": 0.17942908926144088, "grad_norm": 2.50304087299115, "learning_rate": 9.91199551138018e-06, "loss": 0.1428, "step": 1188 }, { "epoch": 0.17958012384836128, "grad_norm": 2.78337018366671, "learning_rate": 9.911847730387799e-06, "loss": 0.1735, "step": 1189 }, { "epoch": 0.17973115843528167, "grad_norm": 3.4576458673053385, "learning_rate": 9.911699826522847e-06, "loss": 0.176, "step": 1190 }, { "epoch": 0.1798821930222021, "grad_norm": 2.621816957171426, "learning_rate": 9.911551799789018e-06, "loss": 0.1567, "step": 1191 }, { "epoch": 0.1800332276091225, "grad_norm": 3.5008985537069677, "learning_rate": 9.911403650190022e-06, "loss": 0.1774, "step": 1192 }, { "epoch": 0.1801842621960429, "grad_norm": 2.926077754931342, "learning_rate": 9.91125537772956e-06, "loss": 0.1664, "step": 1193 }, { "epoch": 0.1803352967829633, "grad_norm": 2.6535639816087326, "learning_rate": 9.911106982411342e-06, "loss": 0.161, "step": 1194 }, { "epoch": 0.1804863313698837, "grad_norm": 2.8736194205063232, "learning_rate": 9.910958464239082e-06, "loss": 0.1544, "step": 1195 }, { "epoch": 0.1806373659568041, "grad_norm": 2.446482575359826, "learning_rate": 9.910809823216493e-06, "loss": 0.1508, "step": 1196 }, { "epoch": 0.18078840054372453, "grad_norm": 2.7800267070694895, "learning_rate": 9.910661059347295e-06, "loss": 0.1699, "step": 1197 }, { "epoch": 0.18093943513064492, "grad_norm": 3.1684068592343593, "learning_rate": 9.910512172635209e-06, "loss": 0.1737, "step": 1198 }, { "epoch": 0.18109046971756532, "grad_norm": 2.3968323272640304, "learning_rate": 9.910363163083957e-06, "loss": 0.1707, "step": 1199 }, { "epoch": 0.18124150430448574, "grad_norm": 2.8141525885655456, "learning_rate": 9.910214030697272e-06, "loss": 0.1839, "step": 1200 }, { "epoch": 0.18139253889140614, "grad_norm": 2.489264913573276, "learning_rate": 9.91006477547888e-06, "loss": 0.1587, "step": 1201 }, { "epoch": 0.18154357347832653, "grad_norm": 2.6199983443570853, "learning_rate": 9.909915397432515e-06, "loss": 0.1823, "step": 1202 }, { "epoch": 0.18169460806524695, "grad_norm": 3.1760244810442932, "learning_rate": 9.909765896561917e-06, "loss": 0.1579, "step": 1203 }, { "epoch": 0.18184564265216735, "grad_norm": 3.0772620357308402, "learning_rate": 9.909616272870823e-06, "loss": 0.1539, "step": 1204 }, { "epoch": 0.18199667723908775, "grad_norm": 2.3468258633883385, "learning_rate": 9.909466526362976e-06, "loss": 0.1541, "step": 1205 }, { "epoch": 0.18214771182600817, "grad_norm": 3.568719321923356, "learning_rate": 9.909316657042123e-06, "loss": 0.1696, "step": 1206 }, { "epoch": 0.18229874641292856, "grad_norm": 2.3732586558065574, "learning_rate": 9.909166664912014e-06, "loss": 0.1348, "step": 1207 }, { "epoch": 0.18244978099984896, "grad_norm": 2.585640527185471, "learning_rate": 9.909016549976399e-06, "loss": 0.1636, "step": 1208 }, { "epoch": 0.18260081558676938, "grad_norm": 3.104440465044141, "learning_rate": 9.908866312239033e-06, "loss": 0.1662, "step": 1209 }, { "epoch": 0.18275185017368978, "grad_norm": 2.4262572535882905, "learning_rate": 9.908715951703676e-06, "loss": 0.1488, "step": 1210 }, { "epoch": 0.18290288476061017, "grad_norm": 3.1349518731533306, "learning_rate": 9.90856546837409e-06, "loss": 0.1792, "step": 1211 }, { "epoch": 0.1830539193475306, "grad_norm": 2.7663734292579543, "learning_rate": 9.908414862254036e-06, "loss": 0.1466, "step": 1212 }, { "epoch": 0.183204953934451, "grad_norm": 3.0073275171069525, "learning_rate": 9.908264133347285e-06, "loss": 0.1776, "step": 1213 }, { "epoch": 0.1833559885213714, "grad_norm": 2.860269496391725, "learning_rate": 9.908113281657606e-06, "loss": 0.1726, "step": 1214 }, { "epoch": 0.1835070231082918, "grad_norm": 3.14641031862686, "learning_rate": 9.907962307188772e-06, "loss": 0.15, "step": 1215 }, { "epoch": 0.1836580576952122, "grad_norm": 2.5755204636534508, "learning_rate": 9.907811209944562e-06, "loss": 0.1524, "step": 1216 }, { "epoch": 0.1838090922821326, "grad_norm": 3.5098039591278463, "learning_rate": 9.907659989928753e-06, "loss": 0.2034, "step": 1217 }, { "epoch": 0.18396012686905303, "grad_norm": 3.3236971117820366, "learning_rate": 9.90750864714513e-06, "loss": 0.1586, "step": 1218 }, { "epoch": 0.18411116145597342, "grad_norm": 3.2391021061165333, "learning_rate": 9.907357181597477e-06, "loss": 0.1673, "step": 1219 }, { "epoch": 0.18426219604289382, "grad_norm": 2.953876367671559, "learning_rate": 9.907205593289585e-06, "loss": 0.173, "step": 1220 }, { "epoch": 0.18441323062981424, "grad_norm": 3.211871334646318, "learning_rate": 9.907053882225245e-06, "loss": 0.1857, "step": 1221 }, { "epoch": 0.18456426521673464, "grad_norm": 2.644126480345143, "learning_rate": 9.906902048408254e-06, "loss": 0.1951, "step": 1222 }, { "epoch": 0.18471529980365503, "grad_norm": 3.4769135816121475, "learning_rate": 9.906750091842405e-06, "loss": 0.1891, "step": 1223 }, { "epoch": 0.18486633439057545, "grad_norm": 3.240746234712207, "learning_rate": 9.906598012531506e-06, "loss": 0.1455, "step": 1224 }, { "epoch": 0.18501736897749585, "grad_norm": 3.6711973188986375, "learning_rate": 9.906445810479356e-06, "loss": 0.17, "step": 1225 }, { "epoch": 0.18516840356441625, "grad_norm": 2.612691785271652, "learning_rate": 9.906293485689766e-06, "loss": 0.1566, "step": 1226 }, { "epoch": 0.18531943815133667, "grad_norm": 3.212480648433158, "learning_rate": 9.906141038166543e-06, "loss": 0.1606, "step": 1227 }, { "epoch": 0.18547047273825706, "grad_norm": 2.3935393300531467, "learning_rate": 9.905988467913504e-06, "loss": 0.1463, "step": 1228 }, { "epoch": 0.18562150732517746, "grad_norm": 8.117931539011868, "learning_rate": 9.905835774934465e-06, "loss": 0.2131, "step": 1229 }, { "epoch": 0.18577254191209788, "grad_norm": 5.65168193876282, "learning_rate": 9.905682959233244e-06, "loss": 0.2196, "step": 1230 }, { "epoch": 0.18592357649901828, "grad_norm": 3.1356094735554687, "learning_rate": 9.905530020813664e-06, "loss": 0.1944, "step": 1231 }, { "epoch": 0.18607461108593867, "grad_norm": 3.448802762390671, "learning_rate": 9.905376959679552e-06, "loss": 0.1882, "step": 1232 }, { "epoch": 0.1862256456728591, "grad_norm": 2.9611653295228364, "learning_rate": 9.905223775834737e-06, "loss": 0.1884, "step": 1233 }, { "epoch": 0.1863766802597795, "grad_norm": 3.279458115694425, "learning_rate": 9.90507046928305e-06, "loss": 0.168, "step": 1234 }, { "epoch": 0.1865277148466999, "grad_norm": 2.72466905032855, "learning_rate": 9.904917040028326e-06, "loss": 0.1637, "step": 1235 }, { "epoch": 0.1866787494336203, "grad_norm": 3.0132840974565993, "learning_rate": 9.904763488074405e-06, "loss": 0.194, "step": 1236 }, { "epoch": 0.1868297840205407, "grad_norm": 2.592221195920695, "learning_rate": 9.904609813425126e-06, "loss": 0.1569, "step": 1237 }, { "epoch": 0.1869808186074611, "grad_norm": 4.861956735808422, "learning_rate": 9.904456016084333e-06, "loss": 0.1806, "step": 1238 }, { "epoch": 0.18713185319438153, "grad_norm": 3.0741613191507375, "learning_rate": 9.904302096055876e-06, "loss": 0.1553, "step": 1239 }, { "epoch": 0.18728288778130192, "grad_norm": 3.2936727051803136, "learning_rate": 9.904148053343603e-06, "loss": 0.1799, "step": 1240 }, { "epoch": 0.18743392236822232, "grad_norm": 2.6943284376150776, "learning_rate": 9.903993887951369e-06, "loss": 0.178, "step": 1241 }, { "epoch": 0.18758495695514274, "grad_norm": 2.601701091640637, "learning_rate": 9.903839599883028e-06, "loss": 0.1473, "step": 1242 }, { "epoch": 0.18773599154206314, "grad_norm": 2.44682575594342, "learning_rate": 9.903685189142442e-06, "loss": 0.1652, "step": 1243 }, { "epoch": 0.18788702612898353, "grad_norm": 4.336322313094945, "learning_rate": 9.903530655733473e-06, "loss": 0.1904, "step": 1244 }, { "epoch": 0.18803806071590395, "grad_norm": 5.206116891165799, "learning_rate": 9.903375999659988e-06, "loss": 0.2168, "step": 1245 }, { "epoch": 0.18818909530282435, "grad_norm": 3.3191132273454107, "learning_rate": 9.903221220925853e-06, "loss": 0.1765, "step": 1246 }, { "epoch": 0.18834012988974475, "grad_norm": 2.2414203616269677, "learning_rate": 9.903066319534942e-06, "loss": 0.14, "step": 1247 }, { "epoch": 0.18849116447666517, "grad_norm": 4.297723543018357, "learning_rate": 9.902911295491128e-06, "loss": 0.1848, "step": 1248 }, { "epoch": 0.18864219906358556, "grad_norm": 8.05087798394847, "learning_rate": 9.902756148798292e-06, "loss": 0.2149, "step": 1249 }, { "epoch": 0.18879323365050596, "grad_norm": 3.467736438569819, "learning_rate": 9.902600879460313e-06, "loss": 0.1511, "step": 1250 }, { "epoch": 0.18894426823742638, "grad_norm": 3.454768763274536, "learning_rate": 9.902445487481076e-06, "loss": 0.1907, "step": 1251 }, { "epoch": 0.18909530282434678, "grad_norm": 3.2980077925374327, "learning_rate": 9.902289972864469e-06, "loss": 0.1514, "step": 1252 }, { "epoch": 0.18924633741126717, "grad_norm": 2.850570014473061, "learning_rate": 9.90213433561438e-06, "loss": 0.1535, "step": 1253 }, { "epoch": 0.1893973719981876, "grad_norm": 3.271699455040283, "learning_rate": 9.901978575734702e-06, "loss": 0.1488, "step": 1254 }, { "epoch": 0.189548406585108, "grad_norm": 2.866330548802409, "learning_rate": 9.901822693229334e-06, "loss": 0.1619, "step": 1255 }, { "epoch": 0.1896994411720284, "grad_norm": 2.965203799862044, "learning_rate": 9.901666688102174e-06, "loss": 0.1576, "step": 1256 }, { "epoch": 0.1898504757589488, "grad_norm": 2.763688752624667, "learning_rate": 9.901510560357125e-06, "loss": 0.1596, "step": 1257 }, { "epoch": 0.1900015103458692, "grad_norm": 3.2249652436366762, "learning_rate": 9.901354309998093e-06, "loss": 0.1858, "step": 1258 }, { "epoch": 0.1901525449327896, "grad_norm": 3.133902886191092, "learning_rate": 9.901197937028986e-06, "loss": 0.1765, "step": 1259 }, { "epoch": 0.19030357951971003, "grad_norm": 2.4355030881048805, "learning_rate": 9.901041441453715e-06, "loss": 0.1475, "step": 1260 }, { "epoch": 0.19045461410663042, "grad_norm": 2.702595402794447, "learning_rate": 9.900884823276197e-06, "loss": 0.1761, "step": 1261 }, { "epoch": 0.19060564869355082, "grad_norm": 2.807466576767471, "learning_rate": 9.90072808250035e-06, "loss": 0.1531, "step": 1262 }, { "epoch": 0.19075668328047124, "grad_norm": 2.750237324889081, "learning_rate": 9.900571219130092e-06, "loss": 0.1676, "step": 1263 }, { "epoch": 0.19090771786739164, "grad_norm": 3.522784298807298, "learning_rate": 9.900414233169349e-06, "loss": 0.1929, "step": 1264 }, { "epoch": 0.19105875245431203, "grad_norm": 2.6022046225637157, "learning_rate": 9.900257124622048e-06, "loss": 0.1703, "step": 1265 }, { "epoch": 0.19120978704123245, "grad_norm": 8.0987949924497, "learning_rate": 9.900099893492117e-06, "loss": 0.238, "step": 1266 }, { "epoch": 0.19136082162815285, "grad_norm": 3.4162540812165147, "learning_rate": 9.899942539783494e-06, "loss": 0.1881, "step": 1267 }, { "epoch": 0.19151185621507325, "grad_norm": 3.8274485703415357, "learning_rate": 9.899785063500112e-06, "loss": 0.1648, "step": 1268 }, { "epoch": 0.19166289080199367, "grad_norm": 2.729699655890492, "learning_rate": 9.89962746464591e-06, "loss": 0.1618, "step": 1269 }, { "epoch": 0.19181392538891406, "grad_norm": 2.899831819015068, "learning_rate": 9.89946974322483e-06, "loss": 0.1612, "step": 1270 }, { "epoch": 0.19196495997583446, "grad_norm": 3.4595410319760673, "learning_rate": 9.89931189924082e-06, "loss": 0.2083, "step": 1271 }, { "epoch": 0.19211599456275488, "grad_norm": 4.8951342663241215, "learning_rate": 9.899153932697827e-06, "loss": 0.2276, "step": 1272 }, { "epoch": 0.19226702914967528, "grad_norm": 3.0886743559671768, "learning_rate": 9.898995843599802e-06, "loss": 0.1511, "step": 1273 }, { "epoch": 0.19241806373659567, "grad_norm": 2.8933536874099564, "learning_rate": 9.898837631950701e-06, "loss": 0.1681, "step": 1274 }, { "epoch": 0.1925690983235161, "grad_norm": 2.9208746120366524, "learning_rate": 9.898679297754483e-06, "loss": 0.169, "step": 1275 }, { "epoch": 0.1927201329104365, "grad_norm": 2.43413101575708, "learning_rate": 9.898520841015105e-06, "loss": 0.1646, "step": 1276 }, { "epoch": 0.1928711674973569, "grad_norm": 3.4739524445357177, "learning_rate": 9.898362261736534e-06, "loss": 0.1723, "step": 1277 }, { "epoch": 0.1930222020842773, "grad_norm": 3.535387939188254, "learning_rate": 9.898203559922737e-06, "loss": 0.1675, "step": 1278 }, { "epoch": 0.1931732366711977, "grad_norm": 3.021829717865429, "learning_rate": 9.898044735577682e-06, "loss": 0.1785, "step": 1279 }, { "epoch": 0.1933242712581181, "grad_norm": 2.9075188455706904, "learning_rate": 9.897885788705343e-06, "loss": 0.177, "step": 1280 }, { "epoch": 0.19347530584503853, "grad_norm": 2.168230061057665, "learning_rate": 9.897726719309694e-06, "loss": 0.1557, "step": 1281 }, { "epoch": 0.19362634043195892, "grad_norm": 3.1524809272324896, "learning_rate": 9.89756752739472e-06, "loss": 0.192, "step": 1282 }, { "epoch": 0.19377737501887932, "grad_norm": 5.685836873331989, "learning_rate": 9.8974082129644e-06, "loss": 0.1968, "step": 1283 }, { "epoch": 0.19392840960579974, "grad_norm": 2.9229167139193377, "learning_rate": 9.897248776022715e-06, "loss": 0.1649, "step": 1284 }, { "epoch": 0.19407944419272014, "grad_norm": 2.6641650853514576, "learning_rate": 9.89708921657366e-06, "loss": 0.1716, "step": 1285 }, { "epoch": 0.19423047877964053, "grad_norm": 2.7216220805911857, "learning_rate": 9.896929534621224e-06, "loss": 0.1819, "step": 1286 }, { "epoch": 0.19438151336656095, "grad_norm": 2.913812275329237, "learning_rate": 9.8967697301694e-06, "loss": 0.1683, "step": 1287 }, { "epoch": 0.19453254795348135, "grad_norm": 3.250695829112208, "learning_rate": 9.896609803222188e-06, "loss": 0.1697, "step": 1288 }, { "epoch": 0.19468358254040175, "grad_norm": 2.5937783553723603, "learning_rate": 9.896449753783588e-06, "loss": 0.1524, "step": 1289 }, { "epoch": 0.19483461712732217, "grad_norm": 3.5262140849100754, "learning_rate": 9.896289581857603e-06, "loss": 0.195, "step": 1290 }, { "epoch": 0.19498565171424256, "grad_norm": 3.4415276859205712, "learning_rate": 9.89612928744824e-06, "loss": 0.1634, "step": 1291 }, { "epoch": 0.19513668630116296, "grad_norm": 3.2606553130777667, "learning_rate": 9.89596887055951e-06, "loss": 0.1823, "step": 1292 }, { "epoch": 0.19528772088808338, "grad_norm": 2.557335068722074, "learning_rate": 9.895808331195425e-06, "loss": 0.1653, "step": 1293 }, { "epoch": 0.19543875547500378, "grad_norm": 2.622220874728092, "learning_rate": 9.89564766936e-06, "loss": 0.1749, "step": 1294 }, { "epoch": 0.19558979006192417, "grad_norm": 2.19881381942841, "learning_rate": 9.895486885057254e-06, "loss": 0.1577, "step": 1295 }, { "epoch": 0.1957408246488446, "grad_norm": 2.926711840539242, "learning_rate": 9.89532597829121e-06, "loss": 0.1804, "step": 1296 }, { "epoch": 0.195891859235765, "grad_norm": 2.3905251436740813, "learning_rate": 9.895164949065897e-06, "loss": 0.1522, "step": 1297 }, { "epoch": 0.1960428938226854, "grad_norm": 3.1703519542668293, "learning_rate": 9.895003797385337e-06, "loss": 0.1867, "step": 1298 }, { "epoch": 0.1961939284096058, "grad_norm": 2.5468282164735454, "learning_rate": 9.894842523253563e-06, "loss": 0.1644, "step": 1299 }, { "epoch": 0.1963449629965262, "grad_norm": 2.692539998654574, "learning_rate": 9.89468112667461e-06, "loss": 0.1796, "step": 1300 }, { "epoch": 0.1964959975834466, "grad_norm": 2.3546894750847667, "learning_rate": 9.894519607652517e-06, "loss": 0.138, "step": 1301 }, { "epoch": 0.19664703217036703, "grad_norm": 2.5915843001699597, "learning_rate": 9.894357966191322e-06, "loss": 0.155, "step": 1302 }, { "epoch": 0.19679806675728742, "grad_norm": 3.3258909773899044, "learning_rate": 9.89419620229507e-06, "loss": 0.1837, "step": 1303 }, { "epoch": 0.19694910134420782, "grad_norm": 2.679968966317782, "learning_rate": 9.894034315967808e-06, "loss": 0.1723, "step": 1304 }, { "epoch": 0.19710013593112824, "grad_norm": 2.671915929344966, "learning_rate": 9.893872307213585e-06, "loss": 0.1519, "step": 1305 }, { "epoch": 0.19725117051804864, "grad_norm": 2.774555314530525, "learning_rate": 9.893710176036452e-06, "loss": 0.1751, "step": 1306 }, { "epoch": 0.19740220510496903, "grad_norm": 2.5979374317540294, "learning_rate": 9.893547922440467e-06, "loss": 0.1516, "step": 1307 }, { "epoch": 0.19755323969188945, "grad_norm": 3.5594345522125126, "learning_rate": 9.893385546429689e-06, "loss": 0.186, "step": 1308 }, { "epoch": 0.19770427427880985, "grad_norm": 3.370270848936061, "learning_rate": 9.893223048008179e-06, "loss": 0.1787, "step": 1309 }, { "epoch": 0.19785530886573025, "grad_norm": 2.250161796374347, "learning_rate": 9.893060427180002e-06, "loss": 0.1485, "step": 1310 }, { "epoch": 0.19800634345265067, "grad_norm": 3.104637464176949, "learning_rate": 9.892897683949228e-06, "loss": 0.1786, "step": 1311 }, { "epoch": 0.19815737803957106, "grad_norm": 2.6155081363668438, "learning_rate": 9.892734818319924e-06, "loss": 0.151, "step": 1312 }, { "epoch": 0.19830841262649146, "grad_norm": 2.130932019475206, "learning_rate": 9.892571830296167e-06, "loss": 0.1454, "step": 1313 }, { "epoch": 0.19845944721341188, "grad_norm": 2.4749189904075255, "learning_rate": 9.892408719882034e-06, "loss": 0.157, "step": 1314 }, { "epoch": 0.19861048180033228, "grad_norm": 2.5810718233009684, "learning_rate": 9.892245487081605e-06, "loss": 0.1728, "step": 1315 }, { "epoch": 0.19876151638725267, "grad_norm": 2.472293834373759, "learning_rate": 9.892082131898963e-06, "loss": 0.1638, "step": 1316 }, { "epoch": 0.1989125509741731, "grad_norm": 5.093092359947358, "learning_rate": 9.891918654338197e-06, "loss": 0.2049, "step": 1317 }, { "epoch": 0.1990635855610935, "grad_norm": 3.0837247532990912, "learning_rate": 9.891755054403392e-06, "loss": 0.1993, "step": 1318 }, { "epoch": 0.1992146201480139, "grad_norm": 2.5307995292059347, "learning_rate": 9.891591332098644e-06, "loss": 0.1556, "step": 1319 }, { "epoch": 0.1993656547349343, "grad_norm": 3.329333629264179, "learning_rate": 9.891427487428047e-06, "loss": 0.1948, "step": 1320 }, { "epoch": 0.1995166893218547, "grad_norm": 2.731229987962899, "learning_rate": 9.891263520395701e-06, "loss": 0.1664, "step": 1321 }, { "epoch": 0.1996677239087751, "grad_norm": 2.9505049123968456, "learning_rate": 9.891099431005707e-06, "loss": 0.1873, "step": 1322 }, { "epoch": 0.19981875849569553, "grad_norm": 3.9966608268390615, "learning_rate": 9.890935219262167e-06, "loss": 0.1928, "step": 1323 }, { "epoch": 0.19996979308261592, "grad_norm": 3.707343132109054, "learning_rate": 9.890770885169194e-06, "loss": 0.1707, "step": 1324 }, { "epoch": 0.20012082766953632, "grad_norm": 2.7888728270196057, "learning_rate": 9.890606428730896e-06, "loss": 0.1794, "step": 1325 }, { "epoch": 0.20027186225645674, "grad_norm": 2.9110970888647936, "learning_rate": 9.890441849951387e-06, "loss": 0.1743, "step": 1326 }, { "epoch": 0.20042289684337714, "grad_norm": 2.5546998388753277, "learning_rate": 9.890277148834785e-06, "loss": 0.1469, "step": 1327 }, { "epoch": 0.20057393143029753, "grad_norm": 2.845100298953577, "learning_rate": 9.89011232538521e-06, "loss": 0.1499, "step": 1328 }, { "epoch": 0.20072496601721795, "grad_norm": 3.1757610260510303, "learning_rate": 9.889947379606784e-06, "loss": 0.2015, "step": 1329 }, { "epoch": 0.20087600060413835, "grad_norm": 2.725388268079473, "learning_rate": 9.889782311503633e-06, "loss": 0.1619, "step": 1330 }, { "epoch": 0.20102703519105874, "grad_norm": 2.444780494529352, "learning_rate": 9.889617121079888e-06, "loss": 0.1769, "step": 1331 }, { "epoch": 0.20117806977797917, "grad_norm": 2.896823466142177, "learning_rate": 9.889451808339681e-06, "loss": 0.1807, "step": 1332 }, { "epoch": 0.20132910436489956, "grad_norm": 2.6033839529024845, "learning_rate": 9.889286373287146e-06, "loss": 0.1692, "step": 1333 }, { "epoch": 0.20148013895181996, "grad_norm": 2.6677460555197556, "learning_rate": 9.889120815926423e-06, "loss": 0.1496, "step": 1334 }, { "epoch": 0.20163117353874038, "grad_norm": 2.5597869735540377, "learning_rate": 9.888955136261651e-06, "loss": 0.1691, "step": 1335 }, { "epoch": 0.20178220812566078, "grad_norm": 2.483665091584859, "learning_rate": 9.888789334296978e-06, "loss": 0.171, "step": 1336 }, { "epoch": 0.20193324271258117, "grad_norm": 2.8819179686073295, "learning_rate": 9.888623410036549e-06, "loss": 0.1822, "step": 1337 }, { "epoch": 0.2020842772995016, "grad_norm": 2.5186209711556273, "learning_rate": 9.888457363484515e-06, "loss": 0.1657, "step": 1338 }, { "epoch": 0.202235311886422, "grad_norm": 3.363505289326104, "learning_rate": 9.888291194645034e-06, "loss": 0.1661, "step": 1339 }, { "epoch": 0.2023863464733424, "grad_norm": 2.693113020870346, "learning_rate": 9.888124903522256e-06, "loss": 0.16, "step": 1340 }, { "epoch": 0.2025373810602628, "grad_norm": 4.037764995599374, "learning_rate": 9.887958490120346e-06, "loss": 0.2072, "step": 1341 }, { "epoch": 0.2026884156471832, "grad_norm": 3.8112198315519357, "learning_rate": 9.887791954443465e-06, "loss": 0.2026, "step": 1342 }, { "epoch": 0.2028394502341036, "grad_norm": 5.5447725369192975, "learning_rate": 9.887625296495777e-06, "loss": 0.1831, "step": 1343 }, { "epoch": 0.20299048482102403, "grad_norm": 3.651304124658771, "learning_rate": 9.887458516281453e-06, "loss": 0.1968, "step": 1344 }, { "epoch": 0.20314151940794442, "grad_norm": 2.4723389278698034, "learning_rate": 9.887291613804668e-06, "loss": 0.1707, "step": 1345 }, { "epoch": 0.20329255399486482, "grad_norm": 4.010699698811202, "learning_rate": 9.887124589069593e-06, "loss": 0.1711, "step": 1346 }, { "epoch": 0.20344358858178524, "grad_norm": 3.603982009382552, "learning_rate": 9.886957442080406e-06, "loss": 0.1825, "step": 1347 }, { "epoch": 0.20359462316870564, "grad_norm": 2.7099835031574537, "learning_rate": 9.886790172841292e-06, "loss": 0.1771, "step": 1348 }, { "epoch": 0.20374565775562603, "grad_norm": 2.2598992412483545, "learning_rate": 9.886622781356431e-06, "loss": 0.1441, "step": 1349 }, { "epoch": 0.20389669234254645, "grad_norm": 2.7372993073874983, "learning_rate": 9.886455267630014e-06, "loss": 0.1645, "step": 1350 }, { "epoch": 0.20404772692946685, "grad_norm": 3.573390169962759, "learning_rate": 9.88628763166623e-06, "loss": 0.1632, "step": 1351 }, { "epoch": 0.20419876151638724, "grad_norm": 7.290400738078803, "learning_rate": 9.886119873469273e-06, "loss": 0.1856, "step": 1352 }, { "epoch": 0.20434979610330767, "grad_norm": 2.677652004418074, "learning_rate": 9.885951993043337e-06, "loss": 0.1753, "step": 1353 }, { "epoch": 0.20450083069022806, "grad_norm": 3.038480127042222, "learning_rate": 9.885783990392626e-06, "loss": 0.1584, "step": 1354 }, { "epoch": 0.20465186527714846, "grad_norm": 2.688855948947413, "learning_rate": 9.885615865521339e-06, "loss": 0.1697, "step": 1355 }, { "epoch": 0.20480289986406888, "grad_norm": 2.138546813509331, "learning_rate": 9.885447618433682e-06, "loss": 0.1327, "step": 1356 }, { "epoch": 0.20495393445098928, "grad_norm": 3.7264370882610995, "learning_rate": 9.885279249133866e-06, "loss": 0.1702, "step": 1357 }, { "epoch": 0.20510496903790967, "grad_norm": 2.4105825766885762, "learning_rate": 9.885110757626103e-06, "loss": 0.1767, "step": 1358 }, { "epoch": 0.2052560036248301, "grad_norm": 3.565538813076652, "learning_rate": 9.884942143914605e-06, "loss": 0.1889, "step": 1359 }, { "epoch": 0.2054070382117505, "grad_norm": 2.77296059004634, "learning_rate": 9.884773408003592e-06, "loss": 0.1884, "step": 1360 }, { "epoch": 0.2055580727986709, "grad_norm": 3.657193752822568, "learning_rate": 9.884604549897284e-06, "loss": 0.1784, "step": 1361 }, { "epoch": 0.2057091073855913, "grad_norm": 2.710371221040804, "learning_rate": 9.884435569599906e-06, "loss": 0.1568, "step": 1362 }, { "epoch": 0.2058601419725117, "grad_norm": 2.3921637623976926, "learning_rate": 9.884266467115687e-06, "loss": 0.1393, "step": 1363 }, { "epoch": 0.2060111765594321, "grad_norm": 2.9933517825441145, "learning_rate": 9.884097242448852e-06, "loss": 0.1719, "step": 1364 }, { "epoch": 0.20616221114635253, "grad_norm": 3.139763271627463, "learning_rate": 9.883927895603638e-06, "loss": 0.1971, "step": 1365 }, { "epoch": 0.20631324573327292, "grad_norm": 4.816123051718904, "learning_rate": 9.883758426584282e-06, "loss": 0.261, "step": 1366 }, { "epoch": 0.20646428032019332, "grad_norm": 3.13605374260052, "learning_rate": 9.88358883539502e-06, "loss": 0.1698, "step": 1367 }, { "epoch": 0.20661531490711374, "grad_norm": 2.7589473330462773, "learning_rate": 9.883419122040098e-06, "loss": 0.1735, "step": 1368 }, { "epoch": 0.20676634949403414, "grad_norm": 2.5277977436299808, "learning_rate": 9.883249286523758e-06, "loss": 0.155, "step": 1369 }, { "epoch": 0.20691738408095453, "grad_norm": 2.904444065097221, "learning_rate": 9.883079328850252e-06, "loss": 0.1753, "step": 1370 }, { "epoch": 0.20706841866787495, "grad_norm": 3.388800780780785, "learning_rate": 9.882909249023829e-06, "loss": 0.161, "step": 1371 }, { "epoch": 0.20721945325479535, "grad_norm": 2.5753180248029746, "learning_rate": 9.882739047048746e-06, "loss": 0.1588, "step": 1372 }, { "epoch": 0.20737048784171574, "grad_norm": 2.7722668186894066, "learning_rate": 9.882568722929258e-06, "loss": 0.1606, "step": 1373 }, { "epoch": 0.20752152242863617, "grad_norm": 6.396882302302419, "learning_rate": 9.882398276669628e-06, "loss": 0.1765, "step": 1374 }, { "epoch": 0.20767255701555656, "grad_norm": 3.65911685939138, "learning_rate": 9.882227708274116e-06, "loss": 0.1425, "step": 1375 }, { "epoch": 0.20782359160247696, "grad_norm": 2.957387118475303, "learning_rate": 9.882057017746996e-06, "loss": 0.1836, "step": 1376 }, { "epoch": 0.20797462618939738, "grad_norm": 2.7620046650990315, "learning_rate": 9.88188620509253e-06, "loss": 0.1729, "step": 1377 }, { "epoch": 0.20812566077631778, "grad_norm": 2.223085742473691, "learning_rate": 9.881715270314995e-06, "loss": 0.1355, "step": 1378 }, { "epoch": 0.20827669536323817, "grad_norm": 2.180605988659474, "learning_rate": 9.881544213418668e-06, "loss": 0.1619, "step": 1379 }, { "epoch": 0.2084277299501586, "grad_norm": 3.499794056586733, "learning_rate": 9.881373034407826e-06, "loss": 0.1913, "step": 1380 }, { "epoch": 0.208578764537079, "grad_norm": 3.1234619679988844, "learning_rate": 9.881201733286753e-06, "loss": 0.1717, "step": 1381 }, { "epoch": 0.2087297991239994, "grad_norm": 5.505660321025835, "learning_rate": 9.881030310059732e-06, "loss": 0.1679, "step": 1382 }, { "epoch": 0.2088808337109198, "grad_norm": 2.587480711702894, "learning_rate": 9.880858764731051e-06, "loss": 0.1571, "step": 1383 }, { "epoch": 0.2090318682978402, "grad_norm": 2.834710289945592, "learning_rate": 9.880687097305005e-06, "loss": 0.18, "step": 1384 }, { "epoch": 0.2091829028847606, "grad_norm": 2.9264938792040276, "learning_rate": 9.880515307785885e-06, "loss": 0.1706, "step": 1385 }, { "epoch": 0.20933393747168103, "grad_norm": 5.212124692166129, "learning_rate": 9.880343396177989e-06, "loss": 0.1918, "step": 1386 }, { "epoch": 0.20948497205860142, "grad_norm": 2.779075978067428, "learning_rate": 9.880171362485618e-06, "loss": 0.1819, "step": 1387 }, { "epoch": 0.20963600664552182, "grad_norm": 2.684090408203458, "learning_rate": 9.879999206713076e-06, "loss": 0.1544, "step": 1388 }, { "epoch": 0.20978704123244224, "grad_norm": 2.926216498757306, "learning_rate": 9.879826928864669e-06, "loss": 0.1841, "step": 1389 }, { "epoch": 0.20993807581936264, "grad_norm": 3.0134342099732945, "learning_rate": 9.879654528944705e-06, "loss": 0.1734, "step": 1390 }, { "epoch": 0.21008911040628303, "grad_norm": 2.2956899609283283, "learning_rate": 9.8794820069575e-06, "loss": 0.1546, "step": 1391 }, { "epoch": 0.21024014499320345, "grad_norm": 3.161959169228782, "learning_rate": 9.879309362907365e-06, "loss": 0.185, "step": 1392 }, { "epoch": 0.21039117958012385, "grad_norm": 2.6730984080925406, "learning_rate": 9.879136596798624e-06, "loss": 0.1574, "step": 1393 }, { "epoch": 0.21054221416704424, "grad_norm": 2.4998336237793946, "learning_rate": 9.878963708635596e-06, "loss": 0.1755, "step": 1394 }, { "epoch": 0.21069324875396467, "grad_norm": 2.856716210159613, "learning_rate": 9.878790698422606e-06, "loss": 0.1556, "step": 1395 }, { "epoch": 0.21084428334088506, "grad_norm": 3.7355578777064626, "learning_rate": 9.878617566163983e-06, "loss": 0.1742, "step": 1396 }, { "epoch": 0.21099531792780546, "grad_norm": 2.633195837920462, "learning_rate": 9.878444311864057e-06, "loss": 0.169, "step": 1397 }, { "epoch": 0.21114635251472588, "grad_norm": 3.40280354042744, "learning_rate": 9.878270935527163e-06, "loss": 0.1463, "step": 1398 }, { "epoch": 0.21129738710164628, "grad_norm": 2.528769775695909, "learning_rate": 9.878097437157637e-06, "loss": 0.1731, "step": 1399 }, { "epoch": 0.21144842168856667, "grad_norm": 2.4575590485691072, "learning_rate": 9.87792381675982e-06, "loss": 0.1579, "step": 1400 }, { "epoch": 0.2115994562754871, "grad_norm": 1.7403047514109404, "learning_rate": 9.877750074338055e-06, "loss": 0.1113, "step": 1401 }, { "epoch": 0.2117504908624075, "grad_norm": 4.5977296426106165, "learning_rate": 9.877576209896688e-06, "loss": 0.2257, "step": 1402 }, { "epoch": 0.2119015254493279, "grad_norm": 3.0796815633430343, "learning_rate": 9.87740222344007e-06, "loss": 0.1801, "step": 1403 }, { "epoch": 0.2120525600362483, "grad_norm": 2.489042640249363, "learning_rate": 9.87722811497255e-06, "loss": 0.1521, "step": 1404 }, { "epoch": 0.2122035946231687, "grad_norm": 2.761412650507982, "learning_rate": 9.877053884498484e-06, "loss": 0.1543, "step": 1405 }, { "epoch": 0.2123546292100891, "grad_norm": 2.772459996731734, "learning_rate": 9.876879532022234e-06, "loss": 0.183, "step": 1406 }, { "epoch": 0.21250566379700953, "grad_norm": 3.057832835724708, "learning_rate": 9.876705057548159e-06, "loss": 0.2134, "step": 1407 }, { "epoch": 0.21265669838392992, "grad_norm": 2.3918446071200976, "learning_rate": 9.876530461080623e-06, "loss": 0.1692, "step": 1408 }, { "epoch": 0.21280773297085032, "grad_norm": 3.428778347640393, "learning_rate": 9.876355742623995e-06, "loss": 0.188, "step": 1409 }, { "epoch": 0.21295876755777074, "grad_norm": 3.1417455382405395, "learning_rate": 9.876180902182645e-06, "loss": 0.1847, "step": 1410 }, { "epoch": 0.21310980214469114, "grad_norm": 2.6905885530883924, "learning_rate": 9.876005939760946e-06, "loss": 0.1417, "step": 1411 }, { "epoch": 0.21326083673161153, "grad_norm": 4.498980816838148, "learning_rate": 9.875830855363277e-06, "loss": 0.2061, "step": 1412 }, { "epoch": 0.21341187131853195, "grad_norm": 2.255912481972632, "learning_rate": 9.875655648994016e-06, "loss": 0.1639, "step": 1413 }, { "epoch": 0.21356290590545235, "grad_norm": 2.9139776483552455, "learning_rate": 9.875480320657547e-06, "loss": 0.1536, "step": 1414 }, { "epoch": 0.21371394049237274, "grad_norm": 2.61391557700871, "learning_rate": 9.875304870358254e-06, "loss": 0.1639, "step": 1415 }, { "epoch": 0.21386497507929317, "grad_norm": 5.294544696743941, "learning_rate": 9.875129298100529e-06, "loss": 0.2611, "step": 1416 }, { "epoch": 0.21401600966621356, "grad_norm": 3.222798841394195, "learning_rate": 9.874953603888761e-06, "loss": 0.1741, "step": 1417 }, { "epoch": 0.21416704425313396, "grad_norm": 2.6434869118906494, "learning_rate": 9.874777787727347e-06, "loss": 0.175, "step": 1418 }, { "epoch": 0.21431807884005438, "grad_norm": 3.889857158906113, "learning_rate": 9.874601849620683e-06, "loss": 0.1787, "step": 1419 }, { "epoch": 0.21446911342697478, "grad_norm": 2.498400665096606, "learning_rate": 9.874425789573175e-06, "loss": 0.1595, "step": 1420 }, { "epoch": 0.21462014801389517, "grad_norm": 3.297181951712556, "learning_rate": 9.874249607589222e-06, "loss": 0.1543, "step": 1421 }, { "epoch": 0.2147711826008156, "grad_norm": 2.732138205041404, "learning_rate": 9.874073303673233e-06, "loss": 0.1604, "step": 1422 }, { "epoch": 0.214922217187736, "grad_norm": 8.3870681814734, "learning_rate": 9.873896877829618e-06, "loss": 0.2494, "step": 1423 }, { "epoch": 0.2150732517746564, "grad_norm": 2.365635993225761, "learning_rate": 9.873720330062793e-06, "loss": 0.1477, "step": 1424 }, { "epoch": 0.2152242863615768, "grad_norm": 2.850730846456062, "learning_rate": 9.873543660377171e-06, "loss": 0.1814, "step": 1425 }, { "epoch": 0.2153753209484972, "grad_norm": 3.0574033623950947, "learning_rate": 9.873366868777172e-06, "loss": 0.1641, "step": 1426 }, { "epoch": 0.2155263555354176, "grad_norm": 2.9156715771837436, "learning_rate": 9.873189955267221e-06, "loss": 0.1748, "step": 1427 }, { "epoch": 0.21567739012233803, "grad_norm": 4.84726736437559, "learning_rate": 9.873012919851743e-06, "loss": 0.179, "step": 1428 }, { "epoch": 0.21582842470925842, "grad_norm": 2.9138784851010526, "learning_rate": 9.872835762535165e-06, "loss": 0.1927, "step": 1429 }, { "epoch": 0.21597945929617882, "grad_norm": 2.307821533186042, "learning_rate": 9.872658483321917e-06, "loss": 0.1588, "step": 1430 }, { "epoch": 0.21613049388309924, "grad_norm": 2.841770237773955, "learning_rate": 9.872481082216439e-06, "loss": 0.1833, "step": 1431 }, { "epoch": 0.21628152847001963, "grad_norm": 3.9020355403394102, "learning_rate": 9.872303559223165e-06, "loss": 0.1964, "step": 1432 }, { "epoch": 0.21643256305694003, "grad_norm": 2.106283997098011, "learning_rate": 9.872125914346537e-06, "loss": 0.1409, "step": 1433 }, { "epoch": 0.21658359764386045, "grad_norm": 3.1451757561051004, "learning_rate": 9.871948147590997e-06, "loss": 0.177, "step": 1434 }, { "epoch": 0.21673463223078085, "grad_norm": 3.041366745087603, "learning_rate": 9.871770258960995e-06, "loss": 0.1552, "step": 1435 }, { "epoch": 0.21688566681770124, "grad_norm": 2.517201814204237, "learning_rate": 9.87159224846098e-06, "loss": 0.1394, "step": 1436 }, { "epoch": 0.21703670140462167, "grad_norm": 3.6135901778264663, "learning_rate": 9.871414116095403e-06, "loss": 0.1796, "step": 1437 }, { "epoch": 0.21718773599154206, "grad_norm": 2.9652449964478986, "learning_rate": 9.871235861868723e-06, "loss": 0.1979, "step": 1438 }, { "epoch": 0.21733877057846246, "grad_norm": 3.7735350162929944, "learning_rate": 9.871057485785398e-06, "loss": 0.201, "step": 1439 }, { "epoch": 0.21748980516538288, "grad_norm": 3.168766788832821, "learning_rate": 9.870878987849888e-06, "loss": 0.1964, "step": 1440 }, { "epoch": 0.21764083975230328, "grad_norm": 2.456658472012387, "learning_rate": 9.870700368066663e-06, "loss": 0.1683, "step": 1441 }, { "epoch": 0.21779187433922367, "grad_norm": 2.5275252153168815, "learning_rate": 9.870521626440186e-06, "loss": 0.1758, "step": 1442 }, { "epoch": 0.2179429089261441, "grad_norm": 2.3920739337243395, "learning_rate": 9.870342762974932e-06, "loss": 0.146, "step": 1443 }, { "epoch": 0.2180939435130645, "grad_norm": 2.6345390035924834, "learning_rate": 9.870163777675375e-06, "loss": 0.1704, "step": 1444 }, { "epoch": 0.2182449780999849, "grad_norm": 4.21368414840786, "learning_rate": 9.869984670545989e-06, "loss": 0.1715, "step": 1445 }, { "epoch": 0.2183960126869053, "grad_norm": 2.679388002282753, "learning_rate": 9.869805441591259e-06, "loss": 0.1583, "step": 1446 }, { "epoch": 0.2185470472738257, "grad_norm": 2.3776929723963276, "learning_rate": 9.869626090815666e-06, "loss": 0.1608, "step": 1447 }, { "epoch": 0.2186980818607461, "grad_norm": 3.3161491290527136, "learning_rate": 9.869446618223697e-06, "loss": 0.1614, "step": 1448 }, { "epoch": 0.21884911644766653, "grad_norm": 3.3552141971556333, "learning_rate": 9.869267023819842e-06, "loss": 0.21, "step": 1449 }, { "epoch": 0.21900015103458692, "grad_norm": 2.3883671603862715, "learning_rate": 9.869087307608593e-06, "loss": 0.1234, "step": 1450 }, { "epoch": 0.21915118562150732, "grad_norm": 2.7787814026950075, "learning_rate": 9.868907469594446e-06, "loss": 0.1836, "step": 1451 }, { "epoch": 0.21930222020842774, "grad_norm": 3.3262219778622453, "learning_rate": 9.868727509781898e-06, "loss": 0.1976, "step": 1452 }, { "epoch": 0.21945325479534813, "grad_norm": 2.4091471324500957, "learning_rate": 9.868547428175457e-06, "loss": 0.1629, "step": 1453 }, { "epoch": 0.21960428938226853, "grad_norm": 3.0877084034420283, "learning_rate": 9.86836722477962e-06, "loss": 0.1502, "step": 1454 }, { "epoch": 0.21975532396918895, "grad_norm": 2.545798890473065, "learning_rate": 9.868186899598899e-06, "loss": 0.1619, "step": 1455 }, { "epoch": 0.21990635855610935, "grad_norm": 2.563609756672713, "learning_rate": 9.868006452637804e-06, "loss": 0.1585, "step": 1456 }, { "epoch": 0.22005739314302974, "grad_norm": 3.1087046077900364, "learning_rate": 9.86782588390085e-06, "loss": 0.1821, "step": 1457 }, { "epoch": 0.22020842772995017, "grad_norm": 2.219088130502891, "learning_rate": 9.867645193392553e-06, "loss": 0.1422, "step": 1458 }, { "epoch": 0.22035946231687056, "grad_norm": 2.816243222260662, "learning_rate": 9.867464381117433e-06, "loss": 0.1738, "step": 1459 }, { "epoch": 0.22051049690379096, "grad_norm": 2.657208131872758, "learning_rate": 9.867283447080013e-06, "loss": 0.1418, "step": 1460 }, { "epoch": 0.22066153149071138, "grad_norm": 2.279931787774598, "learning_rate": 9.867102391284821e-06, "loss": 0.1561, "step": 1461 }, { "epoch": 0.22081256607763178, "grad_norm": 2.1557166311595335, "learning_rate": 9.866921213736385e-06, "loss": 0.1215, "step": 1462 }, { "epoch": 0.22096360066455217, "grad_norm": 3.1104182322733873, "learning_rate": 9.866739914439236e-06, "loss": 0.1672, "step": 1463 }, { "epoch": 0.2211146352514726, "grad_norm": 3.433006540638066, "learning_rate": 9.866558493397911e-06, "loss": 0.1843, "step": 1464 }, { "epoch": 0.221265669838393, "grad_norm": 2.887643629483961, "learning_rate": 9.866376950616947e-06, "loss": 0.1611, "step": 1465 }, { "epoch": 0.2214167044253134, "grad_norm": 2.556247551157445, "learning_rate": 9.866195286100886e-06, "loss": 0.1641, "step": 1466 }, { "epoch": 0.2215677390122338, "grad_norm": 2.872904092143059, "learning_rate": 9.866013499854273e-06, "loss": 0.1937, "step": 1467 }, { "epoch": 0.2217187735991542, "grad_norm": 2.7745236570953526, "learning_rate": 9.865831591881656e-06, "loss": 0.1624, "step": 1468 }, { "epoch": 0.2218698081860746, "grad_norm": 5.229027493322167, "learning_rate": 9.865649562187583e-06, "loss": 0.1933, "step": 1469 }, { "epoch": 0.22202084277299503, "grad_norm": 2.4007707864157957, "learning_rate": 9.86546741077661e-06, "loss": 0.1554, "step": 1470 }, { "epoch": 0.22217187735991542, "grad_norm": 2.378399753891899, "learning_rate": 9.865285137653292e-06, "loss": 0.1632, "step": 1471 }, { "epoch": 0.22232291194683582, "grad_norm": 2.4448002950161176, "learning_rate": 9.865102742822191e-06, "loss": 0.1743, "step": 1472 }, { "epoch": 0.22247394653375624, "grad_norm": 2.4132355936445373, "learning_rate": 9.864920226287866e-06, "loss": 0.1642, "step": 1473 }, { "epoch": 0.22262498112067663, "grad_norm": 3.045622282425265, "learning_rate": 9.864737588054885e-06, "loss": 0.1758, "step": 1474 }, { "epoch": 0.22277601570759703, "grad_norm": 2.4297954633849272, "learning_rate": 9.864554828127819e-06, "loss": 0.1493, "step": 1475 }, { "epoch": 0.22292705029451745, "grad_norm": 2.8539468998950115, "learning_rate": 9.864371946511235e-06, "loss": 0.1849, "step": 1476 }, { "epoch": 0.22307808488143785, "grad_norm": 3.5287531292479573, "learning_rate": 9.864188943209712e-06, "loss": 0.187, "step": 1477 }, { "epoch": 0.22322911946835824, "grad_norm": 2.9990232726949975, "learning_rate": 9.864005818227826e-06, "loss": 0.2013, "step": 1478 }, { "epoch": 0.22338015405527867, "grad_norm": 2.7102050432821314, "learning_rate": 9.86382257157016e-06, "loss": 0.1681, "step": 1479 }, { "epoch": 0.22353118864219906, "grad_norm": 3.4993406138579384, "learning_rate": 9.863639203241293e-06, "loss": 0.2044, "step": 1480 }, { "epoch": 0.22368222322911946, "grad_norm": 3.197836277386952, "learning_rate": 9.863455713245816e-06, "loss": 0.1817, "step": 1481 }, { "epoch": 0.22383325781603988, "grad_norm": 2.6177599097653172, "learning_rate": 9.86327210158832e-06, "loss": 0.1666, "step": 1482 }, { "epoch": 0.22398429240296028, "grad_norm": 3.14208212782568, "learning_rate": 9.863088368273394e-06, "loss": 0.1989, "step": 1483 }, { "epoch": 0.22413532698988067, "grad_norm": 2.4066819085840345, "learning_rate": 9.86290451330564e-06, "loss": 0.149, "step": 1484 }, { "epoch": 0.2242863615768011, "grad_norm": 2.335746506487486, "learning_rate": 9.862720536689653e-06, "loss": 0.1591, "step": 1485 }, { "epoch": 0.2244373961637215, "grad_norm": 2.8833693546675967, "learning_rate": 9.862536438430035e-06, "loss": 0.1656, "step": 1486 }, { "epoch": 0.2245884307506419, "grad_norm": 3.6175181964521195, "learning_rate": 9.862352218531394e-06, "loss": 0.1976, "step": 1487 }, { "epoch": 0.2247394653375623, "grad_norm": 3.1373825669151585, "learning_rate": 9.862167876998337e-06, "loss": 0.1604, "step": 1488 }, { "epoch": 0.2248904999244827, "grad_norm": 2.933439184937988, "learning_rate": 9.861983413835475e-06, "loss": 0.1906, "step": 1489 }, { "epoch": 0.2250415345114031, "grad_norm": 2.471011889475836, "learning_rate": 9.861798829047423e-06, "loss": 0.1876, "step": 1490 }, { "epoch": 0.22519256909832353, "grad_norm": 3.037078080702591, "learning_rate": 9.861614122638799e-06, "loss": 0.1679, "step": 1491 }, { "epoch": 0.22534360368524392, "grad_norm": 6.364887420162631, "learning_rate": 9.861429294614222e-06, "loss": 0.1974, "step": 1492 }, { "epoch": 0.22549463827216432, "grad_norm": 2.681358757958284, "learning_rate": 9.861244344978315e-06, "loss": 0.1745, "step": 1493 }, { "epoch": 0.22564567285908474, "grad_norm": 3.7913909416782285, "learning_rate": 9.86105927373571e-06, "loss": 0.1885, "step": 1494 }, { "epoch": 0.22579670744600513, "grad_norm": 4.527903740848301, "learning_rate": 9.86087408089103e-06, "loss": 0.2407, "step": 1495 }, { "epoch": 0.22594774203292553, "grad_norm": 2.5390797621719736, "learning_rate": 9.86068876644891e-06, "loss": 0.1559, "step": 1496 }, { "epoch": 0.22609877661984595, "grad_norm": 2.240418059425955, "learning_rate": 9.860503330413986e-06, "loss": 0.1459, "step": 1497 }, { "epoch": 0.22624981120676635, "grad_norm": 2.226481340191797, "learning_rate": 9.8603177727909e-06, "loss": 0.1494, "step": 1498 }, { "epoch": 0.22640084579368674, "grad_norm": 4.304439528313828, "learning_rate": 9.860132093584288e-06, "loss": 0.1906, "step": 1499 }, { "epoch": 0.22655188038060717, "grad_norm": 2.1850999725735036, "learning_rate": 9.8599462927988e-06, "loss": 0.1499, "step": 1500 }, { "epoch": 0.22670291496752756, "grad_norm": 2.2716748800726627, "learning_rate": 9.859760370439077e-06, "loss": 0.129, "step": 1501 }, { "epoch": 0.22685394955444796, "grad_norm": 2.7062822976016987, "learning_rate": 9.859574326509778e-06, "loss": 0.1846, "step": 1502 }, { "epoch": 0.22700498414136838, "grad_norm": 2.5803193452560507, "learning_rate": 9.859388161015552e-06, "loss": 0.1667, "step": 1503 }, { "epoch": 0.22715601872828878, "grad_norm": 5.5554087535347, "learning_rate": 9.85920187396106e-06, "loss": 0.2183, "step": 1504 }, { "epoch": 0.22730705331520917, "grad_norm": 2.4773427483664228, "learning_rate": 9.85901546535096e-06, "loss": 0.1584, "step": 1505 }, { "epoch": 0.2274580879021296, "grad_norm": 2.4877214740563716, "learning_rate": 9.858828935189912e-06, "loss": 0.1644, "step": 1506 }, { "epoch": 0.22760912248905, "grad_norm": 2.2956924088429314, "learning_rate": 9.858642283482585e-06, "loss": 0.161, "step": 1507 }, { "epoch": 0.2277601570759704, "grad_norm": 2.924616100764954, "learning_rate": 9.85845551023365e-06, "loss": 0.1739, "step": 1508 }, { "epoch": 0.2279111916628908, "grad_norm": 2.3106706857694803, "learning_rate": 9.858268615447774e-06, "loss": 0.157, "step": 1509 }, { "epoch": 0.2280622262498112, "grad_norm": 3.250844051283217, "learning_rate": 9.858081599129639e-06, "loss": 0.1782, "step": 1510 }, { "epoch": 0.2282132608367316, "grad_norm": 3.565392455303298, "learning_rate": 9.857894461283917e-06, "loss": 0.1936, "step": 1511 }, { "epoch": 0.22836429542365203, "grad_norm": 2.3598535518218005, "learning_rate": 9.857707201915293e-06, "loss": 0.1338, "step": 1512 }, { "epoch": 0.22851533001057242, "grad_norm": 3.27302900888808, "learning_rate": 9.857519821028453e-06, "loss": 0.179, "step": 1513 }, { "epoch": 0.22866636459749282, "grad_norm": 2.5383303051015833, "learning_rate": 9.857332318628081e-06, "loss": 0.1418, "step": 1514 }, { "epoch": 0.22881739918441324, "grad_norm": 2.3955435069841005, "learning_rate": 9.857144694718867e-06, "loss": 0.1649, "step": 1515 }, { "epoch": 0.22896843377133363, "grad_norm": 2.7593900413896164, "learning_rate": 9.856956949305508e-06, "loss": 0.1649, "step": 1516 }, { "epoch": 0.22911946835825403, "grad_norm": 3.1169779745448034, "learning_rate": 9.856769082392698e-06, "loss": 0.2003, "step": 1517 }, { "epoch": 0.22927050294517445, "grad_norm": 2.591780547095585, "learning_rate": 9.856581093985137e-06, "loss": 0.152, "step": 1518 }, { "epoch": 0.22942153753209485, "grad_norm": 2.56531203682555, "learning_rate": 9.856392984087529e-06, "loss": 0.166, "step": 1519 }, { "epoch": 0.22957257211901524, "grad_norm": 2.459044092251835, "learning_rate": 9.856204752704576e-06, "loss": 0.1695, "step": 1520 }, { "epoch": 0.22972360670593567, "grad_norm": 3.495233328752902, "learning_rate": 9.85601639984099e-06, "loss": 0.173, "step": 1521 }, { "epoch": 0.22987464129285606, "grad_norm": 3.307982348833221, "learning_rate": 9.855827925501483e-06, "loss": 0.1892, "step": 1522 }, { "epoch": 0.23002567587977646, "grad_norm": 3.740599599056339, "learning_rate": 9.855639329690769e-06, "loss": 0.1982, "step": 1523 }, { "epoch": 0.23017671046669688, "grad_norm": 2.940918555123479, "learning_rate": 9.855450612413565e-06, "loss": 0.159, "step": 1524 }, { "epoch": 0.23032774505361728, "grad_norm": 3.339189583868376, "learning_rate": 9.855261773674592e-06, "loss": 0.1726, "step": 1525 }, { "epoch": 0.23047877964053767, "grad_norm": 2.727667298159679, "learning_rate": 9.855072813478576e-06, "loss": 0.1508, "step": 1526 }, { "epoch": 0.2306298142274581, "grad_norm": 2.4149867917765415, "learning_rate": 9.85488373183024e-06, "loss": 0.1506, "step": 1527 }, { "epoch": 0.2307808488143785, "grad_norm": 2.9397945516969077, "learning_rate": 9.85469452873432e-06, "loss": 0.1814, "step": 1528 }, { "epoch": 0.2309318834012989, "grad_norm": 2.476223228813141, "learning_rate": 9.854505204195541e-06, "loss": 0.1427, "step": 1529 }, { "epoch": 0.2310829179882193, "grad_norm": 2.345126181522387, "learning_rate": 9.854315758218645e-06, "loss": 0.1691, "step": 1530 }, { "epoch": 0.2312339525751397, "grad_norm": 3.047455952109798, "learning_rate": 9.854126190808372e-06, "loss": 0.1801, "step": 1531 }, { "epoch": 0.2313849871620601, "grad_norm": 2.836638315351471, "learning_rate": 9.853936501969457e-06, "loss": 0.141, "step": 1532 }, { "epoch": 0.23153602174898053, "grad_norm": 4.30085730178912, "learning_rate": 9.853746691706652e-06, "loss": 0.2156, "step": 1533 }, { "epoch": 0.23168705633590092, "grad_norm": 2.8654214020750373, "learning_rate": 9.853556760024704e-06, "loss": 0.1692, "step": 1534 }, { "epoch": 0.23183809092282132, "grad_norm": 2.9724635198715457, "learning_rate": 9.853366706928363e-06, "loss": 0.1666, "step": 1535 }, { "epoch": 0.23198912550974174, "grad_norm": 2.34202448009326, "learning_rate": 9.853176532422381e-06, "loss": 0.1764, "step": 1536 }, { "epoch": 0.23214016009666213, "grad_norm": 3.906423927578254, "learning_rate": 9.852986236511521e-06, "loss": 0.1832, "step": 1537 }, { "epoch": 0.23229119468358253, "grad_norm": 2.559830851426968, "learning_rate": 9.852795819200539e-06, "loss": 0.1618, "step": 1538 }, { "epoch": 0.23244222927050295, "grad_norm": 2.8008127583244042, "learning_rate": 9.852605280494198e-06, "loss": 0.1685, "step": 1539 }, { "epoch": 0.23259326385742335, "grad_norm": 2.4538752684903944, "learning_rate": 9.852414620397269e-06, "loss": 0.1536, "step": 1540 }, { "epoch": 0.23274429844434374, "grad_norm": 2.227558383197046, "learning_rate": 9.852223838914516e-06, "loss": 0.135, "step": 1541 }, { "epoch": 0.23289533303126417, "grad_norm": 3.005208340068727, "learning_rate": 9.852032936050715e-06, "loss": 0.1621, "step": 1542 }, { "epoch": 0.23304636761818456, "grad_norm": 2.6744519376143465, "learning_rate": 9.85184191181064e-06, "loss": 0.1816, "step": 1543 }, { "epoch": 0.23319740220510496, "grad_norm": 3.8618640403890674, "learning_rate": 9.851650766199073e-06, "loss": 0.1674, "step": 1544 }, { "epoch": 0.23334843679202538, "grad_norm": 2.8209298700454166, "learning_rate": 9.851459499220789e-06, "loss": 0.1747, "step": 1545 }, { "epoch": 0.23349947137894578, "grad_norm": 3.4093391890828593, "learning_rate": 9.851268110880578e-06, "loss": 0.1768, "step": 1546 }, { "epoch": 0.23365050596586617, "grad_norm": 3.5248714375319916, "learning_rate": 9.851076601183226e-06, "loss": 0.1995, "step": 1547 }, { "epoch": 0.2338015405527866, "grad_norm": 2.6132179484569007, "learning_rate": 9.850884970133523e-06, "loss": 0.1549, "step": 1548 }, { "epoch": 0.233952575139707, "grad_norm": 2.565485300184851, "learning_rate": 9.850693217736265e-06, "loss": 0.1465, "step": 1549 }, { "epoch": 0.2341036097266274, "grad_norm": 2.3744700682190834, "learning_rate": 9.850501343996246e-06, "loss": 0.1672, "step": 1550 }, { "epoch": 0.2342546443135478, "grad_norm": 3.423988067042892, "learning_rate": 9.850309348918268e-06, "loss": 0.2129, "step": 1551 }, { "epoch": 0.2344056789004682, "grad_norm": 2.775507056641517, "learning_rate": 9.850117232507133e-06, "loss": 0.1518, "step": 1552 }, { "epoch": 0.2345567134873886, "grad_norm": 3.4650741752720453, "learning_rate": 9.849924994767646e-06, "loss": 0.1801, "step": 1553 }, { "epoch": 0.23470774807430902, "grad_norm": 2.5016371758827938, "learning_rate": 9.849732635704616e-06, "loss": 0.1656, "step": 1554 }, { "epoch": 0.23485878266122942, "grad_norm": 2.773565792649538, "learning_rate": 9.849540155322858e-06, "loss": 0.1713, "step": 1555 }, { "epoch": 0.23500981724814982, "grad_norm": 3.465154507181112, "learning_rate": 9.849347553627185e-06, "loss": 0.1803, "step": 1556 }, { "epoch": 0.23516085183507024, "grad_norm": 3.3061878348604563, "learning_rate": 9.849154830622412e-06, "loss": 0.1653, "step": 1557 }, { "epoch": 0.23531188642199063, "grad_norm": 2.2737649826764192, "learning_rate": 9.848961986313365e-06, "loss": 0.1559, "step": 1558 }, { "epoch": 0.23546292100891103, "grad_norm": 3.376698717132528, "learning_rate": 9.848769020704866e-06, "loss": 0.1923, "step": 1559 }, { "epoch": 0.23561395559583145, "grad_norm": 2.761319965302814, "learning_rate": 9.848575933801741e-06, "loss": 0.1663, "step": 1560 }, { "epoch": 0.23576499018275185, "grad_norm": 3.3854050005123217, "learning_rate": 9.848382725608822e-06, "loss": 0.1704, "step": 1561 }, { "epoch": 0.23591602476967224, "grad_norm": 2.55497664609255, "learning_rate": 9.848189396130943e-06, "loss": 0.1782, "step": 1562 }, { "epoch": 0.23606705935659267, "grad_norm": 2.978845170000291, "learning_rate": 9.847995945372938e-06, "loss": 0.1579, "step": 1563 }, { "epoch": 0.23621809394351306, "grad_norm": 3.63663438385162, "learning_rate": 9.847802373339646e-06, "loss": 0.1935, "step": 1564 }, { "epoch": 0.23636912853043346, "grad_norm": 3.6570380251903236, "learning_rate": 9.847608680035911e-06, "loss": 0.1949, "step": 1565 }, { "epoch": 0.23652016311735388, "grad_norm": 3.301475335246788, "learning_rate": 9.847414865466578e-06, "loss": 0.2069, "step": 1566 }, { "epoch": 0.23667119770427428, "grad_norm": 3.6597094167763005, "learning_rate": 9.847220929636496e-06, "loss": 0.1974, "step": 1567 }, { "epoch": 0.23682223229119467, "grad_norm": 3.0375136657095516, "learning_rate": 9.847026872550513e-06, "loss": 0.1892, "step": 1568 }, { "epoch": 0.2369732668781151, "grad_norm": 5.306156231454977, "learning_rate": 9.846832694213488e-06, "loss": 0.2713, "step": 1569 }, { "epoch": 0.2371243014650355, "grad_norm": 2.9134018463526887, "learning_rate": 9.846638394630276e-06, "loss": 0.1815, "step": 1570 }, { "epoch": 0.2372753360519559, "grad_norm": 2.460759078544831, "learning_rate": 9.846443973805736e-06, "loss": 0.1556, "step": 1571 }, { "epoch": 0.2374263706388763, "grad_norm": 2.9890361774609113, "learning_rate": 9.846249431744736e-06, "loss": 0.1791, "step": 1572 }, { "epoch": 0.2375774052257967, "grad_norm": 3.572574802371502, "learning_rate": 9.84605476845214e-06, "loss": 0.1924, "step": 1573 }, { "epoch": 0.2377284398127171, "grad_norm": 3.541862610509322, "learning_rate": 9.845859983932817e-06, "loss": 0.1797, "step": 1574 }, { "epoch": 0.23787947439963752, "grad_norm": 2.5977180743383546, "learning_rate": 9.845665078191641e-06, "loss": 0.144, "step": 1575 }, { "epoch": 0.23803050898655792, "grad_norm": 2.9428707766345994, "learning_rate": 9.845470051233486e-06, "loss": 0.2013, "step": 1576 }, { "epoch": 0.23818154357347832, "grad_norm": 3.133227785687709, "learning_rate": 9.845274903063232e-06, "loss": 0.1616, "step": 1577 }, { "epoch": 0.23833257816039874, "grad_norm": 2.7852144941864068, "learning_rate": 9.84507963368576e-06, "loss": 0.1803, "step": 1578 }, { "epoch": 0.23848361274731913, "grad_norm": 2.9139911884655816, "learning_rate": 9.844884243105956e-06, "loss": 0.1727, "step": 1579 }, { "epoch": 0.23863464733423953, "grad_norm": 2.3159855452415012, "learning_rate": 9.844688731328707e-06, "loss": 0.1492, "step": 1580 }, { "epoch": 0.23878568192115995, "grad_norm": 2.7921758620755823, "learning_rate": 9.844493098358904e-06, "loss": 0.1599, "step": 1581 }, { "epoch": 0.23893671650808035, "grad_norm": 2.552575712520867, "learning_rate": 9.84429734420144e-06, "loss": 0.1732, "step": 1582 }, { "epoch": 0.23908775109500074, "grad_norm": 2.8669222366701606, "learning_rate": 9.844101468861214e-06, "loss": 0.1751, "step": 1583 }, { "epoch": 0.23923878568192117, "grad_norm": 2.4652706814102245, "learning_rate": 9.843905472343123e-06, "loss": 0.1839, "step": 1584 }, { "epoch": 0.23938982026884156, "grad_norm": 3.189557437361195, "learning_rate": 9.843709354652072e-06, "loss": 0.1604, "step": 1585 }, { "epoch": 0.23954085485576196, "grad_norm": 2.771867412077143, "learning_rate": 9.843513115792967e-06, "loss": 0.1437, "step": 1586 }, { "epoch": 0.23969188944268238, "grad_norm": 2.567910652505609, "learning_rate": 9.843316755770715e-06, "loss": 0.1801, "step": 1587 }, { "epoch": 0.23984292402960278, "grad_norm": 2.3033375641888285, "learning_rate": 9.84312027459023e-06, "loss": 0.1572, "step": 1588 }, { "epoch": 0.23999395861652317, "grad_norm": 2.5190989936168275, "learning_rate": 9.842923672256428e-06, "loss": 0.1612, "step": 1589 }, { "epoch": 0.2401449932034436, "grad_norm": 4.606298762171361, "learning_rate": 9.842726948774224e-06, "loss": 0.1902, "step": 1590 }, { "epoch": 0.240296027790364, "grad_norm": 2.6522557832495326, "learning_rate": 9.842530104148542e-06, "loss": 0.1926, "step": 1591 }, { "epoch": 0.2404470623772844, "grad_norm": 3.535713493725515, "learning_rate": 9.842333138384305e-06, "loss": 0.1953, "step": 1592 }, { "epoch": 0.2405980969642048, "grad_norm": 2.817313040842201, "learning_rate": 9.842136051486442e-06, "loss": 0.1582, "step": 1593 }, { "epoch": 0.2407491315511252, "grad_norm": 2.933402694132232, "learning_rate": 9.84193884345988e-06, "loss": 0.1784, "step": 1594 }, { "epoch": 0.2409001661380456, "grad_norm": 3.0545295580204885, "learning_rate": 9.841741514309552e-06, "loss": 0.1733, "step": 1595 }, { "epoch": 0.24105120072496602, "grad_norm": 2.1931176098933496, "learning_rate": 9.8415440640404e-06, "loss": 0.1627, "step": 1596 }, { "epoch": 0.24120223531188642, "grad_norm": 3.3820997909668193, "learning_rate": 9.841346492657355e-06, "loss": 0.1888, "step": 1597 }, { "epoch": 0.24135326989880682, "grad_norm": 2.071686568068893, "learning_rate": 9.841148800165368e-06, "loss": 0.1315, "step": 1598 }, { "epoch": 0.24150430448572724, "grad_norm": 3.405814150761333, "learning_rate": 9.840950986569378e-06, "loss": 0.1928, "step": 1599 }, { "epoch": 0.24165533907264763, "grad_norm": 3.058555885503471, "learning_rate": 9.840753051874336e-06, "loss": 0.1779, "step": 1600 }, { "epoch": 0.24180637365956803, "grad_norm": 2.6983495501833876, "learning_rate": 9.840554996085192e-06, "loss": 0.1561, "step": 1601 }, { "epoch": 0.24195740824648845, "grad_norm": 4.084334964498333, "learning_rate": 9.840356819206901e-06, "loss": 0.1955, "step": 1602 }, { "epoch": 0.24210844283340885, "grad_norm": 2.8530840502046173, "learning_rate": 9.840158521244424e-06, "loss": 0.1882, "step": 1603 }, { "epoch": 0.24225947742032924, "grad_norm": 3.0371457391641354, "learning_rate": 9.839960102202717e-06, "loss": 0.1416, "step": 1604 }, { "epoch": 0.24241051200724967, "grad_norm": 2.5669514628506547, "learning_rate": 9.839761562086744e-06, "loss": 0.1736, "step": 1605 }, { "epoch": 0.24256154659417006, "grad_norm": 2.3668266390186794, "learning_rate": 9.839562900901475e-06, "loss": 0.1529, "step": 1606 }, { "epoch": 0.24271258118109046, "grad_norm": 2.5835345518861055, "learning_rate": 9.839364118651876e-06, "loss": 0.1751, "step": 1607 }, { "epoch": 0.24286361576801088, "grad_norm": 3.0088959200564447, "learning_rate": 9.83916521534292e-06, "loss": 0.1648, "step": 1608 }, { "epoch": 0.24301465035493128, "grad_norm": 4.1777896951050035, "learning_rate": 9.838966190979586e-06, "loss": 0.1844, "step": 1609 }, { "epoch": 0.24316568494185167, "grad_norm": 2.609409359140186, "learning_rate": 9.838767045566848e-06, "loss": 0.1553, "step": 1610 }, { "epoch": 0.2433167195287721, "grad_norm": 2.514590884065175, "learning_rate": 9.838567779109691e-06, "loss": 0.1541, "step": 1611 }, { "epoch": 0.2434677541156925, "grad_norm": 3.0161065145118515, "learning_rate": 9.8383683916131e-06, "loss": 0.1697, "step": 1612 }, { "epoch": 0.2436187887026129, "grad_norm": 3.546970806522019, "learning_rate": 9.83816888308206e-06, "loss": 0.1705, "step": 1613 }, { "epoch": 0.2437698232895333, "grad_norm": 2.308281612211673, "learning_rate": 9.837969253521564e-06, "loss": 0.1446, "step": 1614 }, { "epoch": 0.2439208578764537, "grad_norm": 3.711137190295816, "learning_rate": 9.837769502936606e-06, "loss": 0.1971, "step": 1615 }, { "epoch": 0.2440718924633741, "grad_norm": 1.8020934831632252, "learning_rate": 9.837569631332183e-06, "loss": 0.1215, "step": 1616 }, { "epoch": 0.24422292705029452, "grad_norm": 2.374993696000421, "learning_rate": 9.837369638713292e-06, "loss": 0.1441, "step": 1617 }, { "epoch": 0.24437396163721492, "grad_norm": 2.7099669394716988, "learning_rate": 9.837169525084938e-06, "loss": 0.1752, "step": 1618 }, { "epoch": 0.24452499622413532, "grad_norm": 2.574663342891234, "learning_rate": 9.836969290452127e-06, "loss": 0.1577, "step": 1619 }, { "epoch": 0.24467603081105574, "grad_norm": 2.565511076318593, "learning_rate": 9.836768934819868e-06, "loss": 0.1606, "step": 1620 }, { "epoch": 0.24482706539797613, "grad_norm": 2.9900914479129543, "learning_rate": 9.836568458193175e-06, "loss": 0.1542, "step": 1621 }, { "epoch": 0.24497809998489653, "grad_norm": 2.9177710853996466, "learning_rate": 9.83636786057706e-06, "loss": 0.164, "step": 1622 }, { "epoch": 0.24512913457181695, "grad_norm": 2.9605780262541233, "learning_rate": 9.83616714197654e-06, "loss": 0.1743, "step": 1623 }, { "epoch": 0.24528016915873735, "grad_norm": 2.710086641560201, "learning_rate": 9.83596630239664e-06, "loss": 0.1671, "step": 1624 }, { "epoch": 0.24543120374565774, "grad_norm": 3.4221070544733285, "learning_rate": 9.83576534184238e-06, "loss": 0.1787, "step": 1625 }, { "epoch": 0.24558223833257817, "grad_norm": 2.472976941829442, "learning_rate": 9.835564260318791e-06, "loss": 0.1548, "step": 1626 }, { "epoch": 0.24573327291949856, "grad_norm": 2.772415942591639, "learning_rate": 9.835363057830903e-06, "loss": 0.1688, "step": 1627 }, { "epoch": 0.24588430750641896, "grad_norm": 3.036322165534609, "learning_rate": 9.835161734383745e-06, "loss": 0.1735, "step": 1628 }, { "epoch": 0.24603534209333938, "grad_norm": 2.6898230502492053, "learning_rate": 9.834960289982357e-06, "loss": 0.1689, "step": 1629 }, { "epoch": 0.24618637668025978, "grad_norm": 2.5256811106418855, "learning_rate": 9.834758724631777e-06, "loss": 0.1468, "step": 1630 }, { "epoch": 0.24633741126718017, "grad_norm": 2.1203659036838887, "learning_rate": 9.834557038337048e-06, "loss": 0.1427, "step": 1631 }, { "epoch": 0.2464884458541006, "grad_norm": 2.688964610792455, "learning_rate": 9.834355231103215e-06, "loss": 0.1649, "step": 1632 }, { "epoch": 0.246639480441021, "grad_norm": 2.036424597776862, "learning_rate": 9.834153302935326e-06, "loss": 0.1367, "step": 1633 }, { "epoch": 0.2467905150279414, "grad_norm": 3.012883808820083, "learning_rate": 9.83395125383843e-06, "loss": 0.1809, "step": 1634 }, { "epoch": 0.2469415496148618, "grad_norm": 3.4879362154369664, "learning_rate": 9.833749083817586e-06, "loss": 0.1904, "step": 1635 }, { "epoch": 0.2470925842017822, "grad_norm": 2.4568423032490525, "learning_rate": 9.833546792877849e-06, "loss": 0.1504, "step": 1636 }, { "epoch": 0.2472436187887026, "grad_norm": 2.424973891782285, "learning_rate": 9.83334438102428e-06, "loss": 0.1404, "step": 1637 }, { "epoch": 0.24739465337562302, "grad_norm": 2.652726139804566, "learning_rate": 9.83314184826194e-06, "loss": 0.1522, "step": 1638 }, { "epoch": 0.24754568796254342, "grad_norm": 2.66340835721018, "learning_rate": 9.8329391945959e-06, "loss": 0.183, "step": 1639 }, { "epoch": 0.24769672254946382, "grad_norm": 2.380935871272131, "learning_rate": 9.832736420031226e-06, "loss": 0.1403, "step": 1640 }, { "epoch": 0.24784775713638424, "grad_norm": 2.6125883907486114, "learning_rate": 9.832533524572991e-06, "loss": 0.1856, "step": 1641 }, { "epoch": 0.24799879172330463, "grad_norm": 2.1230972418355583, "learning_rate": 9.832330508226271e-06, "loss": 0.1317, "step": 1642 }, { "epoch": 0.24814982631022503, "grad_norm": 2.987312520814484, "learning_rate": 9.832127370996146e-06, "loss": 0.1812, "step": 1643 }, { "epoch": 0.24830086089714545, "grad_norm": 3.7060224311765886, "learning_rate": 9.831924112887695e-06, "loss": 0.2091, "step": 1644 }, { "epoch": 0.24845189548406585, "grad_norm": 2.3916051951025774, "learning_rate": 9.831720733906002e-06, "loss": 0.1574, "step": 1645 }, { "epoch": 0.24860293007098624, "grad_norm": 3.643307853907183, "learning_rate": 9.831517234056161e-06, "loss": 0.1828, "step": 1646 }, { "epoch": 0.24875396465790667, "grad_norm": 2.512592807806666, "learning_rate": 9.831313613343255e-06, "loss": 0.1788, "step": 1647 }, { "epoch": 0.24890499924482706, "grad_norm": 2.8084364737468546, "learning_rate": 9.83110987177238e-06, "loss": 0.1845, "step": 1648 }, { "epoch": 0.24905603383174746, "grad_norm": 2.1710177239573767, "learning_rate": 9.830906009348636e-06, "loss": 0.1398, "step": 1649 }, { "epoch": 0.24920706841866788, "grad_norm": 3.2397261333698073, "learning_rate": 9.83070202607712e-06, "loss": 0.1686, "step": 1650 }, { "epoch": 0.24935810300558828, "grad_norm": 8.880433885233593, "learning_rate": 9.830497921962932e-06, "loss": 0.2729, "step": 1651 }, { "epoch": 0.24950913759250867, "grad_norm": 2.5534378056375724, "learning_rate": 9.830293697011183e-06, "loss": 0.1551, "step": 1652 }, { "epoch": 0.2496601721794291, "grad_norm": 2.56391381079493, "learning_rate": 9.83008935122698e-06, "loss": 0.1732, "step": 1653 }, { "epoch": 0.2498112067663495, "grad_norm": 2.5356565413219863, "learning_rate": 9.829884884615433e-06, "loss": 0.1614, "step": 1654 }, { "epoch": 0.2499622413532699, "grad_norm": 2.75914827238624, "learning_rate": 9.829680297181659e-06, "loss": 0.153, "step": 1655 }, { "epoch": 0.2501132759401903, "grad_norm": 2.354318417659074, "learning_rate": 9.829475588930774e-06, "loss": 0.1451, "step": 1656 }, { "epoch": 0.25026431052711073, "grad_norm": 2.8375721322099112, "learning_rate": 9.829270759867901e-06, "loss": 0.1801, "step": 1657 }, { "epoch": 0.25041534511403113, "grad_norm": 2.4237761430506866, "learning_rate": 9.829065809998164e-06, "loss": 0.1644, "step": 1658 }, { "epoch": 0.2505663797009515, "grad_norm": 2.80858318538068, "learning_rate": 9.828860739326688e-06, "loss": 0.1785, "step": 1659 }, { "epoch": 0.2507174142878719, "grad_norm": 2.773056989202051, "learning_rate": 9.828655547858603e-06, "loss": 0.1663, "step": 1660 }, { "epoch": 0.2508684488747923, "grad_norm": 5.146318445538299, "learning_rate": 9.828450235599044e-06, "loss": 0.1834, "step": 1661 }, { "epoch": 0.2510194834617127, "grad_norm": 2.3988949075393573, "learning_rate": 9.828244802553144e-06, "loss": 0.1626, "step": 1662 }, { "epoch": 0.25117051804863316, "grad_norm": 2.528356743102957, "learning_rate": 9.828039248726046e-06, "loss": 0.1661, "step": 1663 }, { "epoch": 0.25132155263555356, "grad_norm": 3.1844372712682367, "learning_rate": 9.827833574122889e-06, "loss": 0.1897, "step": 1664 }, { "epoch": 0.25147258722247395, "grad_norm": 2.5575448417529785, "learning_rate": 9.827627778748816e-06, "loss": 0.186, "step": 1665 }, { "epoch": 0.25162362180939435, "grad_norm": 3.174575817134861, "learning_rate": 9.827421862608981e-06, "loss": 0.1807, "step": 1666 }, { "epoch": 0.25177465639631474, "grad_norm": 5.7768638776748436, "learning_rate": 9.827215825708533e-06, "loss": 0.1994, "step": 1667 }, { "epoch": 0.25192569098323514, "grad_norm": 2.4142327061441256, "learning_rate": 9.827009668052623e-06, "loss": 0.1369, "step": 1668 }, { "epoch": 0.2520767255701556, "grad_norm": 2.327108678674163, "learning_rate": 9.826803389646411e-06, "loss": 0.1649, "step": 1669 }, { "epoch": 0.252227760157076, "grad_norm": 2.5253764832668786, "learning_rate": 9.826596990495057e-06, "loss": 0.1445, "step": 1670 }, { "epoch": 0.2523787947439964, "grad_norm": 3.6285705949839797, "learning_rate": 9.826390470603724e-06, "loss": 0.1999, "step": 1671 }, { "epoch": 0.2525298293309168, "grad_norm": 2.97895934951601, "learning_rate": 9.826183829977576e-06, "loss": 0.1711, "step": 1672 }, { "epoch": 0.2526808639178372, "grad_norm": 2.79252963347811, "learning_rate": 9.825977068621786e-06, "loss": 0.1692, "step": 1673 }, { "epoch": 0.25283189850475757, "grad_norm": 4.744348692778767, "learning_rate": 9.825770186541523e-06, "loss": 0.2161, "step": 1674 }, { "epoch": 0.252982933091678, "grad_norm": 3.0960844136190624, "learning_rate": 9.825563183741965e-06, "loss": 0.186, "step": 1675 }, { "epoch": 0.2531339676785984, "grad_norm": 2.6581128176112094, "learning_rate": 9.825356060228288e-06, "loss": 0.1743, "step": 1676 }, { "epoch": 0.2532850022655188, "grad_norm": 2.550510073935248, "learning_rate": 9.825148816005674e-06, "loss": 0.1745, "step": 1677 }, { "epoch": 0.2534360368524392, "grad_norm": 4.604110511950623, "learning_rate": 9.824941451079309e-06, "loss": 0.1953, "step": 1678 }, { "epoch": 0.2535870714393596, "grad_norm": 2.901677681185062, "learning_rate": 9.824733965454379e-06, "loss": 0.1698, "step": 1679 }, { "epoch": 0.25373810602628, "grad_norm": 2.6872791642515597, "learning_rate": 9.824526359136073e-06, "loss": 0.1826, "step": 1680 }, { "epoch": 0.25388914061320045, "grad_norm": 4.344715011743739, "learning_rate": 9.824318632129585e-06, "loss": 0.2102, "step": 1681 }, { "epoch": 0.25404017520012084, "grad_norm": 3.169238859518044, "learning_rate": 9.824110784440115e-06, "loss": 0.1681, "step": 1682 }, { "epoch": 0.25419120978704124, "grad_norm": 2.7534787318528933, "learning_rate": 9.823902816072856e-06, "loss": 0.1498, "step": 1683 }, { "epoch": 0.25434224437396163, "grad_norm": 2.6290230345923873, "learning_rate": 9.823694727033018e-06, "loss": 0.1519, "step": 1684 }, { "epoch": 0.25449327896088203, "grad_norm": 2.795846450946764, "learning_rate": 9.8234865173258e-06, "loss": 0.1824, "step": 1685 }, { "epoch": 0.2546443135478024, "grad_norm": 3.8068448529941272, "learning_rate": 9.823278186956412e-06, "loss": 0.1896, "step": 1686 }, { "epoch": 0.2547953481347229, "grad_norm": 2.3361996146192525, "learning_rate": 9.823069735930068e-06, "loss": 0.1645, "step": 1687 }, { "epoch": 0.25494638272164327, "grad_norm": 2.515806846848759, "learning_rate": 9.82286116425198e-06, "loss": 0.159, "step": 1688 }, { "epoch": 0.25509741730856367, "grad_norm": 2.4422454303884416, "learning_rate": 9.822652471927366e-06, "loss": 0.1679, "step": 1689 }, { "epoch": 0.25524845189548406, "grad_norm": 2.68679060019578, "learning_rate": 9.822443658961448e-06, "loss": 0.1824, "step": 1690 }, { "epoch": 0.25539948648240446, "grad_norm": 2.733505908585312, "learning_rate": 9.822234725359448e-06, "loss": 0.1725, "step": 1691 }, { "epoch": 0.25555052106932485, "grad_norm": 3.074188947646695, "learning_rate": 9.822025671126593e-06, "loss": 0.1844, "step": 1692 }, { "epoch": 0.2557015556562453, "grad_norm": 2.799866418341489, "learning_rate": 9.821816496268115e-06, "loss": 0.1947, "step": 1693 }, { "epoch": 0.2558525902431657, "grad_norm": 3.9324822914571307, "learning_rate": 9.821607200789243e-06, "loss": 0.2037, "step": 1694 }, { "epoch": 0.2560036248300861, "grad_norm": 2.8210989926287104, "learning_rate": 9.821397784695213e-06, "loss": 0.1639, "step": 1695 }, { "epoch": 0.2561546594170065, "grad_norm": 3.9899164008291033, "learning_rate": 9.821188247991266e-06, "loss": 0.1981, "step": 1696 }, { "epoch": 0.2563056940039269, "grad_norm": 2.7459697065225255, "learning_rate": 9.820978590682642e-06, "loss": 0.1852, "step": 1697 }, { "epoch": 0.2564567285908473, "grad_norm": 2.614204206272215, "learning_rate": 9.820768812774586e-06, "loss": 0.1626, "step": 1698 }, { "epoch": 0.25660776317776773, "grad_norm": 3.2436418651431365, "learning_rate": 9.820558914272344e-06, "loss": 0.1804, "step": 1699 }, { "epoch": 0.25675879776468813, "grad_norm": 3.03245640365945, "learning_rate": 9.820348895181171e-06, "loss": 0.1751, "step": 1700 }, { "epoch": 0.2569098323516085, "grad_norm": 3.78231024942884, "learning_rate": 9.820138755506318e-06, "loss": 0.201, "step": 1701 }, { "epoch": 0.2570608669385289, "grad_norm": 2.697869616119837, "learning_rate": 9.819928495253041e-06, "loss": 0.1719, "step": 1702 }, { "epoch": 0.2572119015254493, "grad_norm": 4.136532204101267, "learning_rate": 9.819718114426602e-06, "loss": 0.1819, "step": 1703 }, { "epoch": 0.2573629361123697, "grad_norm": 2.680504926917412, "learning_rate": 9.819507613032263e-06, "loss": 0.1463, "step": 1704 }, { "epoch": 0.25751397069929016, "grad_norm": 2.9254618700199733, "learning_rate": 9.819296991075286e-06, "loss": 0.1693, "step": 1705 }, { "epoch": 0.25766500528621056, "grad_norm": 2.192197866102021, "learning_rate": 9.819086248560945e-06, "loss": 0.1592, "step": 1706 }, { "epoch": 0.25781603987313095, "grad_norm": 2.962361707323138, "learning_rate": 9.818875385494511e-06, "loss": 0.1528, "step": 1707 }, { "epoch": 0.25796707446005135, "grad_norm": 4.265850456809082, "learning_rate": 9.818664401881258e-06, "loss": 0.1515, "step": 1708 }, { "epoch": 0.25811810904697174, "grad_norm": 2.6774255415660155, "learning_rate": 9.818453297726462e-06, "loss": 0.162, "step": 1709 }, { "epoch": 0.25826914363389214, "grad_norm": 3.1600617184765385, "learning_rate": 9.818242073035406e-06, "loss": 0.1656, "step": 1710 }, { "epoch": 0.2584201782208126, "grad_norm": 3.2904889464743237, "learning_rate": 9.818030727813375e-06, "loss": 0.1813, "step": 1711 }, { "epoch": 0.258571212807733, "grad_norm": 3.184536697710786, "learning_rate": 9.817819262065654e-06, "loss": 0.1741, "step": 1712 }, { "epoch": 0.2587222473946534, "grad_norm": 2.2059041135815387, "learning_rate": 9.817607675797534e-06, "loss": 0.1324, "step": 1713 }, { "epoch": 0.2588732819815738, "grad_norm": 2.607657120727859, "learning_rate": 9.817395969014306e-06, "loss": 0.1872, "step": 1714 }, { "epoch": 0.2590243165684942, "grad_norm": 3.262609183998837, "learning_rate": 9.817184141721268e-06, "loss": 0.1906, "step": 1715 }, { "epoch": 0.25917535115541457, "grad_norm": 3.4310223273925833, "learning_rate": 9.816972193923719e-06, "loss": 0.1809, "step": 1716 }, { "epoch": 0.259326385742335, "grad_norm": 2.7996033094624844, "learning_rate": 9.81676012562696e-06, "loss": 0.1879, "step": 1717 }, { "epoch": 0.2594774203292554, "grad_norm": 2.5917503118778695, "learning_rate": 9.816547936836295e-06, "loss": 0.1728, "step": 1718 }, { "epoch": 0.2596284549161758, "grad_norm": 2.8746095985741995, "learning_rate": 9.816335627557034e-06, "loss": 0.1784, "step": 1719 }, { "epoch": 0.2597794895030962, "grad_norm": 3.1977012726029517, "learning_rate": 9.816123197794487e-06, "loss": 0.1772, "step": 1720 }, { "epoch": 0.2599305240900166, "grad_norm": 3.0740559237475296, "learning_rate": 9.815910647553969e-06, "loss": 0.1647, "step": 1721 }, { "epoch": 0.260081558676937, "grad_norm": 2.3967882280361965, "learning_rate": 9.815697976840797e-06, "loss": 0.155, "step": 1722 }, { "epoch": 0.26023259326385745, "grad_norm": 2.8524672214260853, "learning_rate": 9.81548518566029e-06, "loss": 0.1747, "step": 1723 }, { "epoch": 0.26038362785077784, "grad_norm": 2.6500227563309733, "learning_rate": 9.815272274017771e-06, "loss": 0.1668, "step": 1724 }, { "epoch": 0.26053466243769824, "grad_norm": 2.8651481663910556, "learning_rate": 9.815059241918566e-06, "loss": 0.1744, "step": 1725 }, { "epoch": 0.26068569702461863, "grad_norm": 3.573772199483827, "learning_rate": 9.814846089368007e-06, "loss": 0.1889, "step": 1726 }, { "epoch": 0.26083673161153903, "grad_norm": 2.926149703299907, "learning_rate": 9.814632816371422e-06, "loss": 0.1901, "step": 1727 }, { "epoch": 0.2609877661984594, "grad_norm": 2.3966468074438687, "learning_rate": 9.81441942293415e-06, "loss": 0.1572, "step": 1728 }, { "epoch": 0.2611388007853799, "grad_norm": 2.8426412544872726, "learning_rate": 9.814205909061524e-06, "loss": 0.1662, "step": 1729 }, { "epoch": 0.26128983537230027, "grad_norm": 2.6306367454366364, "learning_rate": 9.813992274758892e-06, "loss": 0.1697, "step": 1730 }, { "epoch": 0.26144086995922067, "grad_norm": 2.389667471046895, "learning_rate": 9.813778520031594e-06, "loss": 0.1671, "step": 1731 }, { "epoch": 0.26159190454614106, "grad_norm": 3.044725793945912, "learning_rate": 9.813564644884977e-06, "loss": 0.184, "step": 1732 }, { "epoch": 0.26174293913306146, "grad_norm": 2.4799336284626086, "learning_rate": 9.813350649324392e-06, "loss": 0.1723, "step": 1733 }, { "epoch": 0.26189397371998185, "grad_norm": 2.7031172246307356, "learning_rate": 9.813136533355193e-06, "loss": 0.1756, "step": 1734 }, { "epoch": 0.2620450083069023, "grad_norm": 2.405367871345457, "learning_rate": 9.812922296982734e-06, "loss": 0.1543, "step": 1735 }, { "epoch": 0.2621960428938227, "grad_norm": 2.664497299862766, "learning_rate": 9.812707940212376e-06, "loss": 0.1714, "step": 1736 }, { "epoch": 0.2623470774807431, "grad_norm": 3.24685512975681, "learning_rate": 9.812493463049482e-06, "loss": 0.1849, "step": 1737 }, { "epoch": 0.2624981120676635, "grad_norm": 2.448531721167046, "learning_rate": 9.812278865499416e-06, "loss": 0.1478, "step": 1738 }, { "epoch": 0.2626491466545839, "grad_norm": 3.675582934710728, "learning_rate": 9.812064147567546e-06, "loss": 0.1923, "step": 1739 }, { "epoch": 0.2628001812415043, "grad_norm": 2.8565755247295463, "learning_rate": 9.811849309259243e-06, "loss": 0.1641, "step": 1740 }, { "epoch": 0.26295121582842473, "grad_norm": 2.831574422218359, "learning_rate": 9.811634350579884e-06, "loss": 0.1811, "step": 1741 }, { "epoch": 0.26310225041534513, "grad_norm": 2.7210099028836527, "learning_rate": 9.811419271534844e-06, "loss": 0.1732, "step": 1742 }, { "epoch": 0.2632532850022655, "grad_norm": 3.345122864672488, "learning_rate": 9.811204072129502e-06, "loss": 0.1854, "step": 1743 }, { "epoch": 0.2634043195891859, "grad_norm": 2.3534809437124906, "learning_rate": 9.810988752369243e-06, "loss": 0.1653, "step": 1744 }, { "epoch": 0.2635553541761063, "grad_norm": 4.62384512734575, "learning_rate": 9.810773312259455e-06, "loss": 0.1859, "step": 1745 }, { "epoch": 0.2637063887630267, "grad_norm": 2.654290246923394, "learning_rate": 9.810557751805525e-06, "loss": 0.1859, "step": 1746 }, { "epoch": 0.26385742334994716, "grad_norm": 1.9430305754932713, "learning_rate": 9.810342071012847e-06, "loss": 0.1293, "step": 1747 }, { "epoch": 0.26400845793686756, "grad_norm": 3.5331670506669313, "learning_rate": 9.810126269886814e-06, "loss": 0.213, "step": 1748 }, { "epoch": 0.26415949252378795, "grad_norm": 3.1140568333858143, "learning_rate": 9.809910348432827e-06, "loss": 0.1747, "step": 1749 }, { "epoch": 0.26431052711070835, "grad_norm": 2.753721178165355, "learning_rate": 9.809694306656285e-06, "loss": 0.1794, "step": 1750 }, { "epoch": 0.26446156169762874, "grad_norm": 2.641579184043524, "learning_rate": 9.809478144562594e-06, "loss": 0.1856, "step": 1751 }, { "epoch": 0.26461259628454914, "grad_norm": 3.053548476252888, "learning_rate": 9.809261862157162e-06, "loss": 0.1571, "step": 1752 }, { "epoch": 0.2647636308714696, "grad_norm": 2.2885340359322788, "learning_rate": 9.809045459445397e-06, "loss": 0.1429, "step": 1753 }, { "epoch": 0.26491466545839, "grad_norm": 3.365371369159781, "learning_rate": 9.808828936432714e-06, "loss": 0.1901, "step": 1754 }, { "epoch": 0.2650657000453104, "grad_norm": 3.060514091230801, "learning_rate": 9.80861229312453e-06, "loss": 0.1729, "step": 1755 }, { "epoch": 0.2652167346322308, "grad_norm": 2.203092378890647, "learning_rate": 9.808395529526266e-06, "loss": 0.1438, "step": 1756 }, { "epoch": 0.2653677692191512, "grad_norm": 2.733722789448055, "learning_rate": 9.808178645643341e-06, "loss": 0.1851, "step": 1757 }, { "epoch": 0.26551880380607157, "grad_norm": 2.6989500782772367, "learning_rate": 9.80796164148118e-06, "loss": 0.1832, "step": 1758 }, { "epoch": 0.265669838392992, "grad_norm": 2.983078175297036, "learning_rate": 9.807744517045214e-06, "loss": 0.1733, "step": 1759 }, { "epoch": 0.2658208729799124, "grad_norm": 2.456963507597587, "learning_rate": 9.807527272340874e-06, "loss": 0.1635, "step": 1760 }, { "epoch": 0.2659719075668328, "grad_norm": 3.423209111250897, "learning_rate": 9.807309907373594e-06, "loss": 0.1909, "step": 1761 }, { "epoch": 0.2661229421537532, "grad_norm": 2.613483737420027, "learning_rate": 9.807092422148811e-06, "loss": 0.1652, "step": 1762 }, { "epoch": 0.2662739767406736, "grad_norm": 2.4933011566481285, "learning_rate": 9.806874816671968e-06, "loss": 0.1826, "step": 1763 }, { "epoch": 0.266425011327594, "grad_norm": 2.7328474164004426, "learning_rate": 9.806657090948505e-06, "loss": 0.1602, "step": 1764 }, { "epoch": 0.26657604591451445, "grad_norm": 2.2583561082058186, "learning_rate": 9.806439244983871e-06, "loss": 0.1552, "step": 1765 }, { "epoch": 0.26672708050143484, "grad_norm": 2.4684952405415386, "learning_rate": 9.806221278783514e-06, "loss": 0.1493, "step": 1766 }, { "epoch": 0.26687811508835524, "grad_norm": 2.6898606802188376, "learning_rate": 9.806003192352889e-06, "loss": 0.179, "step": 1767 }, { "epoch": 0.26702914967527563, "grad_norm": 3.005940839500678, "learning_rate": 9.805784985697448e-06, "loss": 0.1763, "step": 1768 }, { "epoch": 0.26718018426219603, "grad_norm": 3.219743880942522, "learning_rate": 9.805566658822653e-06, "loss": 0.2085, "step": 1769 }, { "epoch": 0.2673312188491164, "grad_norm": 3.3752660489282387, "learning_rate": 9.805348211733963e-06, "loss": 0.1802, "step": 1770 }, { "epoch": 0.2674822534360369, "grad_norm": 3.1610051847211946, "learning_rate": 9.805129644436845e-06, "loss": 0.1919, "step": 1771 }, { "epoch": 0.26763328802295727, "grad_norm": 2.3346451776982633, "learning_rate": 9.804910956936764e-06, "loss": 0.1571, "step": 1772 }, { "epoch": 0.26778432260987767, "grad_norm": 2.905774341368695, "learning_rate": 9.804692149239192e-06, "loss": 0.1573, "step": 1773 }, { "epoch": 0.26793535719679806, "grad_norm": 3.2845941269532957, "learning_rate": 9.804473221349601e-06, "loss": 0.1723, "step": 1774 }, { "epoch": 0.26808639178371846, "grad_norm": 2.9683561038854465, "learning_rate": 9.804254173273472e-06, "loss": 0.1857, "step": 1775 }, { "epoch": 0.26823742637063885, "grad_norm": 2.8895952705734973, "learning_rate": 9.80403500501628e-06, "loss": 0.1744, "step": 1776 }, { "epoch": 0.2683884609575593, "grad_norm": 2.9995738238375145, "learning_rate": 9.803815716583509e-06, "loss": 0.177, "step": 1777 }, { "epoch": 0.2685394955444797, "grad_norm": 2.509750192208786, "learning_rate": 9.803596307980645e-06, "loss": 0.1496, "step": 1778 }, { "epoch": 0.2686905301314001, "grad_norm": 2.5315612188435654, "learning_rate": 9.803376779213177e-06, "loss": 0.167, "step": 1779 }, { "epoch": 0.2688415647183205, "grad_norm": 2.5462162010399245, "learning_rate": 9.803157130286595e-06, "loss": 0.1448, "step": 1780 }, { "epoch": 0.2689925993052409, "grad_norm": 2.6965972387053228, "learning_rate": 9.802937361206396e-06, "loss": 0.1624, "step": 1781 }, { "epoch": 0.2691436338921613, "grad_norm": 4.820509205643074, "learning_rate": 9.802717471978075e-06, "loss": 0.2134, "step": 1782 }, { "epoch": 0.26929466847908173, "grad_norm": 4.132856945631213, "learning_rate": 9.802497462607134e-06, "loss": 0.2264, "step": 1783 }, { "epoch": 0.26944570306600213, "grad_norm": 4.8681034466279955, "learning_rate": 9.802277333099077e-06, "loss": 0.2026, "step": 1784 }, { "epoch": 0.2695967376529225, "grad_norm": 3.2392368408517687, "learning_rate": 9.802057083459411e-06, "loss": 0.1849, "step": 1785 }, { "epoch": 0.2697477722398429, "grad_norm": 2.511896752628236, "learning_rate": 9.801836713693644e-06, "loss": 0.1678, "step": 1786 }, { "epoch": 0.2698988068267633, "grad_norm": 3.1332642373250468, "learning_rate": 9.80161622380729e-06, "loss": 0.1522, "step": 1787 }, { "epoch": 0.2700498414136837, "grad_norm": 2.6594988197771046, "learning_rate": 9.801395613805865e-06, "loss": 0.1667, "step": 1788 }, { "epoch": 0.27020087600060416, "grad_norm": 2.462633899551131, "learning_rate": 9.801174883694885e-06, "loss": 0.1717, "step": 1789 }, { "epoch": 0.27035191058752456, "grad_norm": 2.1039264648900198, "learning_rate": 9.800954033479876e-06, "loss": 0.1674, "step": 1790 }, { "epoch": 0.27050294517444495, "grad_norm": 2.3639335691612, "learning_rate": 9.800733063166358e-06, "loss": 0.1583, "step": 1791 }, { "epoch": 0.27065397976136535, "grad_norm": 3.113240984110691, "learning_rate": 9.800511972759862e-06, "loss": 0.1987, "step": 1792 }, { "epoch": 0.27080501434828574, "grad_norm": 2.195022964365766, "learning_rate": 9.800290762265921e-06, "loss": 0.1595, "step": 1793 }, { "epoch": 0.27095604893520614, "grad_norm": 2.8750403500154595, "learning_rate": 9.800069431690062e-06, "loss": 0.1751, "step": 1794 }, { "epoch": 0.2711070835221266, "grad_norm": 3.825352437893192, "learning_rate": 9.799847981037826e-06, "loss": 0.1679, "step": 1795 }, { "epoch": 0.271258118109047, "grad_norm": 2.5619806315703464, "learning_rate": 9.799626410314752e-06, "loss": 0.1696, "step": 1796 }, { "epoch": 0.2714091526959674, "grad_norm": 3.0606395983586947, "learning_rate": 9.799404719526381e-06, "loss": 0.2004, "step": 1797 }, { "epoch": 0.2715601872828878, "grad_norm": 2.311682738238735, "learning_rate": 9.799182908678264e-06, "loss": 0.1521, "step": 1798 }, { "epoch": 0.2717112218698082, "grad_norm": 2.6561073971043387, "learning_rate": 9.798960977775941e-06, "loss": 0.153, "step": 1799 }, { "epoch": 0.27186225645672857, "grad_norm": 2.7517270589164258, "learning_rate": 9.798738926824973e-06, "loss": 0.1726, "step": 1800 }, { "epoch": 0.272013291043649, "grad_norm": 2.860725673562827, "learning_rate": 9.798516755830909e-06, "loss": 0.1762, "step": 1801 }, { "epoch": 0.2721643256305694, "grad_norm": 2.4022875154660155, "learning_rate": 9.798294464799309e-06, "loss": 0.1724, "step": 1802 }, { "epoch": 0.2723153602174898, "grad_norm": 2.5199240784715315, "learning_rate": 9.798072053735732e-06, "loss": 0.1752, "step": 1803 }, { "epoch": 0.2724663948044102, "grad_norm": 3.0370697606232557, "learning_rate": 9.797849522645744e-06, "loss": 0.1777, "step": 1804 }, { "epoch": 0.2726174293913306, "grad_norm": 2.748087302960482, "learning_rate": 9.797626871534909e-06, "loss": 0.1764, "step": 1805 }, { "epoch": 0.272768463978251, "grad_norm": 3.4806257676598915, "learning_rate": 9.797404100408798e-06, "loss": 0.1982, "step": 1806 }, { "epoch": 0.27291949856517145, "grad_norm": 3.807949077579366, "learning_rate": 9.797181209272984e-06, "loss": 0.2067, "step": 1807 }, { "epoch": 0.27307053315209184, "grad_norm": 3.2561568349162124, "learning_rate": 9.796958198133044e-06, "loss": 0.193, "step": 1808 }, { "epoch": 0.27322156773901224, "grad_norm": 3.0250654718568963, "learning_rate": 9.796735066994554e-06, "loss": 0.1545, "step": 1809 }, { "epoch": 0.27337260232593263, "grad_norm": 2.268312179648648, "learning_rate": 9.796511815863097e-06, "loss": 0.1387, "step": 1810 }, { "epoch": 0.27352363691285303, "grad_norm": 2.0068342984689864, "learning_rate": 9.79628844474426e-06, "loss": 0.1292, "step": 1811 }, { "epoch": 0.2736746714997734, "grad_norm": 2.3719616882730206, "learning_rate": 9.796064953643627e-06, "loss": 0.1737, "step": 1812 }, { "epoch": 0.2738257060866939, "grad_norm": 2.544673405441549, "learning_rate": 9.79584134256679e-06, "loss": 0.1668, "step": 1813 }, { "epoch": 0.27397674067361427, "grad_norm": 2.7847940018212998, "learning_rate": 9.795617611519344e-06, "loss": 0.159, "step": 1814 }, { "epoch": 0.27412777526053467, "grad_norm": 3.089465552161271, "learning_rate": 9.795393760506885e-06, "loss": 0.2052, "step": 1815 }, { "epoch": 0.27427880984745506, "grad_norm": 4.128130166317505, "learning_rate": 9.795169789535013e-06, "loss": 0.1885, "step": 1816 }, { "epoch": 0.27442984443437546, "grad_norm": 3.98541566365405, "learning_rate": 9.794945698609329e-06, "loss": 0.2074, "step": 1817 }, { "epoch": 0.27458087902129585, "grad_norm": 3.2504263266027453, "learning_rate": 9.79472148773544e-06, "loss": 0.1775, "step": 1818 }, { "epoch": 0.2747319136082163, "grad_norm": 2.6421045539771755, "learning_rate": 9.794497156918956e-06, "loss": 0.1664, "step": 1819 }, { "epoch": 0.2748829481951367, "grad_norm": 2.636328214253757, "learning_rate": 9.794272706165488e-06, "loss": 0.1437, "step": 1820 }, { "epoch": 0.2750339827820571, "grad_norm": 3.0093864594625055, "learning_rate": 9.794048135480651e-06, "loss": 0.1443, "step": 1821 }, { "epoch": 0.2751850173689775, "grad_norm": 2.691571532281293, "learning_rate": 9.793823444870061e-06, "loss": 0.1945, "step": 1822 }, { "epoch": 0.2753360519558979, "grad_norm": 2.864726759909635, "learning_rate": 9.79359863433934e-06, "loss": 0.1679, "step": 1823 }, { "epoch": 0.2754870865428183, "grad_norm": 2.285458854821602, "learning_rate": 9.793373703894113e-06, "loss": 0.1672, "step": 1824 }, { "epoch": 0.27563812112973873, "grad_norm": 2.1813338849079322, "learning_rate": 9.793148653540004e-06, "loss": 0.1584, "step": 1825 }, { "epoch": 0.27578915571665913, "grad_norm": 2.336313591592244, "learning_rate": 9.792923483282646e-06, "loss": 0.1463, "step": 1826 }, { "epoch": 0.2759401903035795, "grad_norm": 3.3333810785726175, "learning_rate": 9.792698193127667e-06, "loss": 0.19, "step": 1827 }, { "epoch": 0.2760912248904999, "grad_norm": 3.2435678812483832, "learning_rate": 9.79247278308071e-06, "loss": 0.1627, "step": 1828 }, { "epoch": 0.2762422594774203, "grad_norm": 3.8095861735257355, "learning_rate": 9.792247253147405e-06, "loss": 0.1903, "step": 1829 }, { "epoch": 0.2763932940643407, "grad_norm": 2.736126497763216, "learning_rate": 9.7920216033334e-06, "loss": 0.1798, "step": 1830 }, { "epoch": 0.27654432865126116, "grad_norm": 2.968826464417154, "learning_rate": 9.79179583364434e-06, "loss": 0.168, "step": 1831 }, { "epoch": 0.27669536323818156, "grad_norm": 2.245769010276041, "learning_rate": 9.791569944085869e-06, "loss": 0.1659, "step": 1832 }, { "epoch": 0.27684639782510195, "grad_norm": 2.237043323056987, "learning_rate": 9.79134393466364e-06, "loss": 0.1501, "step": 1833 }, { "epoch": 0.27699743241202235, "grad_norm": 2.819933677276339, "learning_rate": 9.791117805383306e-06, "loss": 0.1608, "step": 1834 }, { "epoch": 0.27714846699894274, "grad_norm": 2.6915017782929134, "learning_rate": 9.790891556250524e-06, "loss": 0.1598, "step": 1835 }, { "epoch": 0.27729950158586314, "grad_norm": 3.073676774083332, "learning_rate": 9.790665187270953e-06, "loss": 0.1459, "step": 1836 }, { "epoch": 0.2774505361727836, "grad_norm": 2.3573942674391115, "learning_rate": 9.790438698450257e-06, "loss": 0.1465, "step": 1837 }, { "epoch": 0.277601570759704, "grad_norm": 8.730538951398504, "learning_rate": 9.790212089794103e-06, "loss": 0.208, "step": 1838 }, { "epoch": 0.2777526053466244, "grad_norm": 2.120320998106101, "learning_rate": 9.789985361308157e-06, "loss": 0.157, "step": 1839 }, { "epoch": 0.2779036399335448, "grad_norm": 3.0562365824854303, "learning_rate": 9.78975851299809e-06, "loss": 0.1646, "step": 1840 }, { "epoch": 0.2780546745204652, "grad_norm": 5.183088422275223, "learning_rate": 9.789531544869581e-06, "loss": 0.2237, "step": 1841 }, { "epoch": 0.27820570910738557, "grad_norm": 2.6022843305987475, "learning_rate": 9.789304456928304e-06, "loss": 0.1915, "step": 1842 }, { "epoch": 0.278356743694306, "grad_norm": 2.204911017533324, "learning_rate": 9.789077249179942e-06, "loss": 0.1621, "step": 1843 }, { "epoch": 0.2785077782812264, "grad_norm": 2.409531068505289, "learning_rate": 9.788849921630178e-06, "loss": 0.1484, "step": 1844 }, { "epoch": 0.2786588128681468, "grad_norm": 3.4515167306325005, "learning_rate": 9.788622474284694e-06, "loss": 0.1782, "step": 1845 }, { "epoch": 0.2788098474550672, "grad_norm": 4.436362612066002, "learning_rate": 9.788394907149187e-06, "loss": 0.1621, "step": 1846 }, { "epoch": 0.2789608820419876, "grad_norm": 2.524022236111638, "learning_rate": 9.78816722022935e-06, "loss": 0.1717, "step": 1847 }, { "epoch": 0.279111916628908, "grad_norm": 2.4881246186114874, "learning_rate": 9.787939413530872e-06, "loss": 0.1694, "step": 1848 }, { "epoch": 0.27926295121582845, "grad_norm": 2.759164514233403, "learning_rate": 9.787711487059455e-06, "loss": 0.1945, "step": 1849 }, { "epoch": 0.27941398580274884, "grad_norm": 2.5248451358773023, "learning_rate": 9.787483440820802e-06, "loss": 0.177, "step": 1850 }, { "epoch": 0.27956502038966924, "grad_norm": 3.4933729615326317, "learning_rate": 9.787255274820618e-06, "loss": 0.1946, "step": 1851 }, { "epoch": 0.27971605497658963, "grad_norm": 2.5074795322602585, "learning_rate": 9.787026989064607e-06, "loss": 0.1552, "step": 1852 }, { "epoch": 0.27986708956351003, "grad_norm": 2.764596377421029, "learning_rate": 9.786798583558484e-06, "loss": 0.1895, "step": 1853 }, { "epoch": 0.2800181241504304, "grad_norm": 2.6147069411392567, "learning_rate": 9.78657005830796e-06, "loss": 0.1653, "step": 1854 }, { "epoch": 0.2801691587373509, "grad_norm": 2.4577636397700604, "learning_rate": 9.786341413318751e-06, "loss": 0.1467, "step": 1855 }, { "epoch": 0.28032019332427127, "grad_norm": 3.345669785422091, "learning_rate": 9.78611264859658e-06, "loss": 0.1566, "step": 1856 }, { "epoch": 0.28047122791119167, "grad_norm": 3.4516294198900557, "learning_rate": 9.785883764147166e-06, "loss": 0.1891, "step": 1857 }, { "epoch": 0.28062226249811206, "grad_norm": 5.011751507923939, "learning_rate": 9.785654759976239e-06, "loss": 0.1939, "step": 1858 }, { "epoch": 0.28077329708503246, "grad_norm": 3.2319862274773574, "learning_rate": 9.785425636089524e-06, "loss": 0.2084, "step": 1859 }, { "epoch": 0.28092433167195285, "grad_norm": 2.8148560416073365, "learning_rate": 9.785196392492753e-06, "loss": 0.1699, "step": 1860 }, { "epoch": 0.2810753662588733, "grad_norm": 2.5919641793461703, "learning_rate": 9.78496702919166e-06, "loss": 0.1677, "step": 1861 }, { "epoch": 0.2812264008457937, "grad_norm": 2.1593383391403895, "learning_rate": 9.784737546191987e-06, "loss": 0.1725, "step": 1862 }, { "epoch": 0.2813774354327141, "grad_norm": 4.192759957932776, "learning_rate": 9.784507943499471e-06, "loss": 0.1783, "step": 1863 }, { "epoch": 0.2815284700196345, "grad_norm": 2.4618902441016433, "learning_rate": 9.784278221119855e-06, "loss": 0.1565, "step": 1864 }, { "epoch": 0.2816795046065549, "grad_norm": 5.839332766460608, "learning_rate": 9.78404837905889e-06, "loss": 0.1322, "step": 1865 }, { "epoch": 0.2818305391934753, "grad_norm": 2.9265690890561515, "learning_rate": 9.78381841732232e-06, "loss": 0.1987, "step": 1866 }, { "epoch": 0.28198157378039573, "grad_norm": 2.168335572198409, "learning_rate": 9.7835883359159e-06, "loss": 0.1447, "step": 1867 }, { "epoch": 0.28213260836731613, "grad_norm": 2.209309368264921, "learning_rate": 9.783358134845389e-06, "loss": 0.1772, "step": 1868 }, { "epoch": 0.2822836429542365, "grad_norm": 2.0415603943120946, "learning_rate": 9.78312781411654e-06, "loss": 0.1538, "step": 1869 }, { "epoch": 0.2824346775411569, "grad_norm": 2.9639729018277317, "learning_rate": 9.782897373735117e-06, "loss": 0.181, "step": 1870 }, { "epoch": 0.2825857121280773, "grad_norm": 3.0312441376384283, "learning_rate": 9.782666813706886e-06, "loss": 0.1857, "step": 1871 }, { "epoch": 0.2827367467149977, "grad_norm": 2.7907611894708797, "learning_rate": 9.782436134037612e-06, "loss": 0.1741, "step": 1872 }, { "epoch": 0.28288778130191816, "grad_norm": 2.8364549419651497, "learning_rate": 9.782205334733067e-06, "loss": 0.17, "step": 1873 }, { "epoch": 0.28303881588883856, "grad_norm": 2.854346492840786, "learning_rate": 9.781974415799025e-06, "loss": 0.1559, "step": 1874 }, { "epoch": 0.28318985047575895, "grad_norm": 2.385536844161919, "learning_rate": 9.781743377241262e-06, "loss": 0.1605, "step": 1875 }, { "epoch": 0.28334088506267935, "grad_norm": 2.832576189632856, "learning_rate": 9.781512219065557e-06, "loss": 0.1756, "step": 1876 }, { "epoch": 0.28349191964959974, "grad_norm": 2.324480020603526, "learning_rate": 9.781280941277691e-06, "loss": 0.1452, "step": 1877 }, { "epoch": 0.28364295423652014, "grad_norm": 3.607430561176278, "learning_rate": 9.781049543883454e-06, "loss": 0.1746, "step": 1878 }, { "epoch": 0.2837939888234406, "grad_norm": 2.6766482559375326, "learning_rate": 9.780818026888633e-06, "loss": 0.1697, "step": 1879 }, { "epoch": 0.283945023410361, "grad_norm": 2.876684058324615, "learning_rate": 9.780586390299017e-06, "loss": 0.1904, "step": 1880 }, { "epoch": 0.2840960579972814, "grad_norm": 2.796818416110566, "learning_rate": 9.780354634120404e-06, "loss": 0.1843, "step": 1881 }, { "epoch": 0.2842470925842018, "grad_norm": 3.23144253593625, "learning_rate": 9.780122758358588e-06, "loss": 0.1831, "step": 1882 }, { "epoch": 0.2843981271711222, "grad_norm": 2.4659553481974568, "learning_rate": 9.779890763019371e-06, "loss": 0.1667, "step": 1883 }, { "epoch": 0.28454916175804257, "grad_norm": 2.513130086421199, "learning_rate": 9.779658648108558e-06, "loss": 0.1357, "step": 1884 }, { "epoch": 0.284700196344963, "grad_norm": 2.7160664723469368, "learning_rate": 9.779426413631953e-06, "loss": 0.1298, "step": 1885 }, { "epoch": 0.2848512309318834, "grad_norm": 2.1150399159379347, "learning_rate": 9.779194059595368e-06, "loss": 0.1356, "step": 1886 }, { "epoch": 0.2850022655188038, "grad_norm": 2.4726716748954884, "learning_rate": 9.778961586004612e-06, "loss": 0.1507, "step": 1887 }, { "epoch": 0.2851533001057242, "grad_norm": 2.6613539931137824, "learning_rate": 9.778728992865504e-06, "loss": 0.1701, "step": 1888 }, { "epoch": 0.2853043346926446, "grad_norm": 2.2271284726421574, "learning_rate": 9.778496280183861e-06, "loss": 0.1354, "step": 1889 }, { "epoch": 0.285455369279565, "grad_norm": 2.952805786270208, "learning_rate": 9.778263447965503e-06, "loss": 0.1858, "step": 1890 }, { "epoch": 0.28560640386648545, "grad_norm": 2.787590478540486, "learning_rate": 9.778030496216257e-06, "loss": 0.1886, "step": 1891 }, { "epoch": 0.28575743845340584, "grad_norm": 2.884434211519765, "learning_rate": 9.77779742494195e-06, "loss": 0.165, "step": 1892 }, { "epoch": 0.28590847304032624, "grad_norm": 2.6113366989055558, "learning_rate": 9.777564234148412e-06, "loss": 0.1601, "step": 1893 }, { "epoch": 0.28605950762724663, "grad_norm": 4.828985228502667, "learning_rate": 9.777330923841476e-06, "loss": 0.1635, "step": 1894 }, { "epoch": 0.28621054221416703, "grad_norm": 2.9041277201611924, "learning_rate": 9.777097494026976e-06, "loss": 0.1679, "step": 1895 }, { "epoch": 0.2863615768010874, "grad_norm": 2.537629526181458, "learning_rate": 9.776863944710756e-06, "loss": 0.1655, "step": 1896 }, { "epoch": 0.2865126113880079, "grad_norm": 2.253110298490645, "learning_rate": 9.776630275898656e-06, "loss": 0.1592, "step": 1897 }, { "epoch": 0.28666364597492827, "grad_norm": 4.842473784033201, "learning_rate": 9.77639648759652e-06, "loss": 0.1548, "step": 1898 }, { "epoch": 0.28681468056184867, "grad_norm": 2.519299543876463, "learning_rate": 9.7761625798102e-06, "loss": 0.1598, "step": 1899 }, { "epoch": 0.28696571514876906, "grad_norm": 2.4599967211705844, "learning_rate": 9.775928552545544e-06, "loss": 0.1472, "step": 1900 }, { "epoch": 0.28711674973568946, "grad_norm": 2.6852760899530517, "learning_rate": 9.775694405808407e-06, "loss": 0.1662, "step": 1901 }, { "epoch": 0.28726778432260985, "grad_norm": 3.0295878640001166, "learning_rate": 9.775460139604649e-06, "loss": 0.1854, "step": 1902 }, { "epoch": 0.2874188189095303, "grad_norm": 2.3662410196066808, "learning_rate": 9.775225753940126e-06, "loss": 0.1545, "step": 1903 }, { "epoch": 0.2875698534964507, "grad_norm": 2.5857727483666952, "learning_rate": 9.774991248820705e-06, "loss": 0.1806, "step": 1904 }, { "epoch": 0.2877208880833711, "grad_norm": 2.483541691860185, "learning_rate": 9.77475662425225e-06, "loss": 0.1393, "step": 1905 }, { "epoch": 0.2878719226702915, "grad_norm": 3.3057981038826645, "learning_rate": 9.77452188024063e-06, "loss": 0.197, "step": 1906 }, { "epoch": 0.2880229572572119, "grad_norm": 3.508686767718275, "learning_rate": 9.774287016791718e-06, "loss": 0.1558, "step": 1907 }, { "epoch": 0.2881739918441323, "grad_norm": 2.3445184767220235, "learning_rate": 9.774052033911391e-06, "loss": 0.1721, "step": 1908 }, { "epoch": 0.28832502643105273, "grad_norm": 2.4353614666355403, "learning_rate": 9.773816931605526e-06, "loss": 0.1674, "step": 1909 }, { "epoch": 0.28847606101797313, "grad_norm": 3.4111806143026198, "learning_rate": 9.773581709880004e-06, "loss": 0.1448, "step": 1910 }, { "epoch": 0.2886270956048935, "grad_norm": 2.1536726901041483, "learning_rate": 9.773346368740709e-06, "loss": 0.1446, "step": 1911 }, { "epoch": 0.2887781301918139, "grad_norm": 4.219825305142911, "learning_rate": 9.773110908193529e-06, "loss": 0.1815, "step": 1912 }, { "epoch": 0.2889291647787343, "grad_norm": 2.0933699295375927, "learning_rate": 9.772875328244352e-06, "loss": 0.1337, "step": 1913 }, { "epoch": 0.2890801993656547, "grad_norm": 1.9054298910878753, "learning_rate": 9.772639628899072e-06, "loss": 0.1234, "step": 1914 }, { "epoch": 0.28923123395257516, "grad_norm": 4.66258648813113, "learning_rate": 9.772403810163586e-06, "loss": 0.1979, "step": 1915 }, { "epoch": 0.28938226853949556, "grad_norm": 2.3770998134899632, "learning_rate": 9.772167872043795e-06, "loss": 0.1617, "step": 1916 }, { "epoch": 0.28953330312641595, "grad_norm": 2.4576862350932696, "learning_rate": 9.771931814545597e-06, "loss": 0.1596, "step": 1917 }, { "epoch": 0.28968433771333635, "grad_norm": 2.4527814997648743, "learning_rate": 9.7716956376749e-06, "loss": 0.1655, "step": 1918 }, { "epoch": 0.28983537230025674, "grad_norm": 2.30774805697476, "learning_rate": 9.771459341437611e-06, "loss": 0.1423, "step": 1919 }, { "epoch": 0.28998640688717714, "grad_norm": 2.846474595505282, "learning_rate": 9.771222925839642e-06, "loss": 0.1715, "step": 1920 }, { "epoch": 0.2901374414740976, "grad_norm": 2.316768747348402, "learning_rate": 9.770986390886905e-06, "loss": 0.1698, "step": 1921 }, { "epoch": 0.290288476061018, "grad_norm": 2.118868591348078, "learning_rate": 9.77074973658532e-06, "loss": 0.1417, "step": 1922 }, { "epoch": 0.2904395106479384, "grad_norm": 2.635916257251866, "learning_rate": 9.770512962940804e-06, "loss": 0.1847, "step": 1923 }, { "epoch": 0.2905905452348588, "grad_norm": 2.349652375339953, "learning_rate": 9.770276069959283e-06, "loss": 0.1523, "step": 1924 }, { "epoch": 0.2907415798217792, "grad_norm": 2.666011395407425, "learning_rate": 9.77003905764668e-06, "loss": 0.1548, "step": 1925 }, { "epoch": 0.29089261440869957, "grad_norm": 3.2026151099603895, "learning_rate": 9.769801926008927e-06, "loss": 0.1928, "step": 1926 }, { "epoch": 0.29104364899562, "grad_norm": 2.717241578440762, "learning_rate": 9.769564675051954e-06, "loss": 0.1695, "step": 1927 }, { "epoch": 0.2911946835825404, "grad_norm": 2.7452055947040286, "learning_rate": 9.769327304781698e-06, "loss": 0.1709, "step": 1928 }, { "epoch": 0.2913457181694608, "grad_norm": 2.7596440254579218, "learning_rate": 9.769089815204091e-06, "loss": 0.1742, "step": 1929 }, { "epoch": 0.2914967527563812, "grad_norm": 2.890380085226697, "learning_rate": 9.768852206325082e-06, "loss": 0.1836, "step": 1930 }, { "epoch": 0.2916477873433016, "grad_norm": 2.3294844071359204, "learning_rate": 9.76861447815061e-06, "loss": 0.1612, "step": 1931 }, { "epoch": 0.291798821930222, "grad_norm": 2.201476662051099, "learning_rate": 9.768376630686625e-06, "loss": 0.1598, "step": 1932 }, { "epoch": 0.29194985651714245, "grad_norm": 2.36869816229337, "learning_rate": 9.768138663939072e-06, "loss": 0.1542, "step": 1933 }, { "epoch": 0.29210089110406284, "grad_norm": 2.9581317015316277, "learning_rate": 9.767900577913909e-06, "loss": 0.148, "step": 1934 }, { "epoch": 0.29225192569098324, "grad_norm": 2.5152201316046896, "learning_rate": 9.767662372617089e-06, "loss": 0.1584, "step": 1935 }, { "epoch": 0.29240296027790363, "grad_norm": 4.470672648051218, "learning_rate": 9.767424048054572e-06, "loss": 0.143, "step": 1936 }, { "epoch": 0.29255399486482403, "grad_norm": 2.694113936783201, "learning_rate": 9.76718560423232e-06, "loss": 0.1721, "step": 1937 }, { "epoch": 0.2927050294517444, "grad_norm": 2.141227769972905, "learning_rate": 9.766947041156296e-06, "loss": 0.1567, "step": 1938 }, { "epoch": 0.2928560640386649, "grad_norm": 2.4501904557123186, "learning_rate": 9.766708358832468e-06, "loss": 0.1651, "step": 1939 }, { "epoch": 0.29300709862558527, "grad_norm": 3.0312674066561707, "learning_rate": 9.766469557266809e-06, "loss": 0.1631, "step": 1940 }, { "epoch": 0.29315813321250567, "grad_norm": 2.9496584001264963, "learning_rate": 9.766230636465293e-06, "loss": 0.1458, "step": 1941 }, { "epoch": 0.29330916779942606, "grad_norm": 2.1022429378279988, "learning_rate": 9.765991596433892e-06, "loss": 0.1484, "step": 1942 }, { "epoch": 0.29346020238634646, "grad_norm": 3.489988564183853, "learning_rate": 9.765752437178593e-06, "loss": 0.1975, "step": 1943 }, { "epoch": 0.29361123697326685, "grad_norm": 2.7424498448917376, "learning_rate": 9.76551315870537e-06, "loss": 0.1898, "step": 1944 }, { "epoch": 0.2937622715601873, "grad_norm": 3.564328570440423, "learning_rate": 9.765273761020218e-06, "loss": 0.1699, "step": 1945 }, { "epoch": 0.2939133061471077, "grad_norm": 2.8403424907403125, "learning_rate": 9.765034244129119e-06, "loss": 0.1834, "step": 1946 }, { "epoch": 0.2940643407340281, "grad_norm": 2.3918479424217485, "learning_rate": 9.764794608038067e-06, "loss": 0.166, "step": 1947 }, { "epoch": 0.2942153753209485, "grad_norm": 2.510592750255119, "learning_rate": 9.764554852753056e-06, "loss": 0.1846, "step": 1948 }, { "epoch": 0.2943664099078689, "grad_norm": 2.1833423185744274, "learning_rate": 9.764314978280083e-06, "loss": 0.1575, "step": 1949 }, { "epoch": 0.2945174444947893, "grad_norm": 1.974776413217818, "learning_rate": 9.764074984625151e-06, "loss": 0.135, "step": 1950 }, { "epoch": 0.29466847908170973, "grad_norm": 2.9668425671368244, "learning_rate": 9.76383487179426e-06, "loss": 0.1837, "step": 1951 }, { "epoch": 0.29481951366863013, "grad_norm": 2.646564404563105, "learning_rate": 9.76359463979342e-06, "loss": 0.1627, "step": 1952 }, { "epoch": 0.2949705482555505, "grad_norm": 2.1601727050562087, "learning_rate": 9.76335428862864e-06, "loss": 0.1423, "step": 1953 }, { "epoch": 0.2951215828424709, "grad_norm": 2.3654064324144253, "learning_rate": 9.763113818305932e-06, "loss": 0.166, "step": 1954 }, { "epoch": 0.2952726174293913, "grad_norm": 2.8448022715750962, "learning_rate": 9.76287322883131e-06, "loss": 0.1919, "step": 1955 }, { "epoch": 0.2954236520163117, "grad_norm": 3.6377081355329075, "learning_rate": 9.762632520210793e-06, "loss": 0.1788, "step": 1956 }, { "epoch": 0.29557468660323216, "grad_norm": 2.371452185069091, "learning_rate": 9.762391692450405e-06, "loss": 0.1522, "step": 1957 }, { "epoch": 0.29572572119015256, "grad_norm": 2.7141967533534976, "learning_rate": 9.762150745556167e-06, "loss": 0.1846, "step": 1958 }, { "epoch": 0.29587675577707295, "grad_norm": 5.021941900064746, "learning_rate": 9.761909679534108e-06, "loss": 0.2007, "step": 1959 }, { "epoch": 0.29602779036399335, "grad_norm": 2.536381482764258, "learning_rate": 9.761668494390259e-06, "loss": 0.1818, "step": 1960 }, { "epoch": 0.29617882495091374, "grad_norm": 3.0750400968933485, "learning_rate": 9.761427190130653e-06, "loss": 0.1844, "step": 1961 }, { "epoch": 0.29632985953783414, "grad_norm": 2.278267928661324, "learning_rate": 9.761185766761325e-06, "loss": 0.138, "step": 1962 }, { "epoch": 0.2964808941247546, "grad_norm": 2.556434818270394, "learning_rate": 9.760944224288316e-06, "loss": 0.1765, "step": 1963 }, { "epoch": 0.296631928711675, "grad_norm": 2.8446078868157083, "learning_rate": 9.760702562717668e-06, "loss": 0.1748, "step": 1964 }, { "epoch": 0.2967829632985954, "grad_norm": 2.14623857340957, "learning_rate": 9.760460782055428e-06, "loss": 0.1342, "step": 1965 }, { "epoch": 0.2969339978855158, "grad_norm": 2.5787281549098067, "learning_rate": 9.760218882307639e-06, "loss": 0.1504, "step": 1966 }, { "epoch": 0.2970850324724362, "grad_norm": 3.802669959840623, "learning_rate": 9.759976863480358e-06, "loss": 0.1986, "step": 1967 }, { "epoch": 0.29723606705935657, "grad_norm": 2.865023687989109, "learning_rate": 9.759734725579634e-06, "loss": 0.1832, "step": 1968 }, { "epoch": 0.297387101646277, "grad_norm": 3.3445406638919697, "learning_rate": 9.75949246861153e-06, "loss": 0.1665, "step": 1969 }, { "epoch": 0.2975381362331974, "grad_norm": 2.08020764044881, "learning_rate": 9.759250092582102e-06, "loss": 0.1297, "step": 1970 }, { "epoch": 0.2976891708201178, "grad_norm": 2.215450848126458, "learning_rate": 9.759007597497414e-06, "loss": 0.1424, "step": 1971 }, { "epoch": 0.2978402054070382, "grad_norm": 2.657661203610989, "learning_rate": 9.758764983363534e-06, "loss": 0.1497, "step": 1972 }, { "epoch": 0.2979912399939586, "grad_norm": 2.3376731059234634, "learning_rate": 9.75852225018653e-06, "loss": 0.1596, "step": 1973 }, { "epoch": 0.298142274580879, "grad_norm": 2.471102048953651, "learning_rate": 9.75827939797247e-06, "loss": 0.1788, "step": 1974 }, { "epoch": 0.29829330916779945, "grad_norm": 2.3394765650182556, "learning_rate": 9.758036426727435e-06, "loss": 0.1536, "step": 1975 }, { "epoch": 0.29844434375471984, "grad_norm": 2.7181775261307446, "learning_rate": 9.757793336457503e-06, "loss": 0.1797, "step": 1976 }, { "epoch": 0.29859537834164024, "grad_norm": 4.156577546320162, "learning_rate": 9.75755012716875e-06, "loss": 0.2082, "step": 1977 }, { "epoch": 0.29874641292856063, "grad_norm": 2.542549688550352, "learning_rate": 9.757306798867266e-06, "loss": 0.177, "step": 1978 }, { "epoch": 0.29889744751548103, "grad_norm": 2.165092429144584, "learning_rate": 9.757063351559133e-06, "loss": 0.1511, "step": 1979 }, { "epoch": 0.2990484821024014, "grad_norm": 2.822565537478327, "learning_rate": 9.756819785250444e-06, "loss": 0.177, "step": 1980 }, { "epoch": 0.2991995166893219, "grad_norm": 2.675681079509183, "learning_rate": 9.75657609994729e-06, "loss": 0.1677, "step": 1981 }, { "epoch": 0.29935055127624227, "grad_norm": 2.141356307200772, "learning_rate": 9.756332295655767e-06, "loss": 0.1432, "step": 1982 }, { "epoch": 0.29950158586316267, "grad_norm": 2.8193173660728803, "learning_rate": 9.756088372381976e-06, "loss": 0.1506, "step": 1983 }, { "epoch": 0.29965262045008306, "grad_norm": 2.134668390131682, "learning_rate": 9.755844330132016e-06, "loss": 0.1362, "step": 1984 }, { "epoch": 0.29980365503700346, "grad_norm": 3.110832838023987, "learning_rate": 9.755600168911995e-06, "loss": 0.1759, "step": 1985 }, { "epoch": 0.29995468962392385, "grad_norm": 2.4228210655832765, "learning_rate": 9.75535588872802e-06, "loss": 0.1624, "step": 1986 }, { "epoch": 0.3001057242108443, "grad_norm": 2.0657045469655317, "learning_rate": 9.7551114895862e-06, "loss": 0.1536, "step": 1987 }, { "epoch": 0.3002567587977647, "grad_norm": 2.394867963341991, "learning_rate": 9.754866971492649e-06, "loss": 0.1573, "step": 1988 }, { "epoch": 0.3004077933846851, "grad_norm": 2.382230203165415, "learning_rate": 9.754622334453485e-06, "loss": 0.1716, "step": 1989 }, { "epoch": 0.3005588279716055, "grad_norm": 2.5517122886389787, "learning_rate": 9.754377578474827e-06, "loss": 0.1735, "step": 1990 }, { "epoch": 0.3007098625585259, "grad_norm": 2.9167117737499604, "learning_rate": 9.7541327035628e-06, "loss": 0.1759, "step": 1991 }, { "epoch": 0.3008608971454463, "grad_norm": 2.051179791501208, "learning_rate": 9.753887709723525e-06, "loss": 0.1432, "step": 1992 }, { "epoch": 0.30101193173236673, "grad_norm": 3.086939377740631, "learning_rate": 9.753642596963135e-06, "loss": 0.1387, "step": 1993 }, { "epoch": 0.30116296631928713, "grad_norm": 2.1391993391695268, "learning_rate": 9.75339736528776e-06, "loss": 0.1449, "step": 1994 }, { "epoch": 0.3013140009062075, "grad_norm": 2.779060867812634, "learning_rate": 9.753152014703535e-06, "loss": 0.1984, "step": 1995 }, { "epoch": 0.3014650354931279, "grad_norm": 2.426318643099375, "learning_rate": 9.752906545216596e-06, "loss": 0.1495, "step": 1996 }, { "epoch": 0.3016160700800483, "grad_norm": 2.5255109046926694, "learning_rate": 9.752660956833086e-06, "loss": 0.1591, "step": 1997 }, { "epoch": 0.3017671046669687, "grad_norm": 3.1803062849773114, "learning_rate": 9.752415249559147e-06, "loss": 0.1779, "step": 1998 }, { "epoch": 0.30191813925388916, "grad_norm": 2.7282913062930527, "learning_rate": 9.752169423400926e-06, "loss": 0.1935, "step": 1999 }, { "epoch": 0.30206917384080956, "grad_norm": 2.90839174079624, "learning_rate": 9.751923478364571e-06, "loss": 0.2042, "step": 2000 }, { "epoch": 0.30222020842772995, "grad_norm": 2.8795700618317572, "learning_rate": 9.751677414456235e-06, "loss": 0.1854, "step": 2001 }, { "epoch": 0.30237124301465035, "grad_norm": 2.78189614882119, "learning_rate": 9.751431231682076e-06, "loss": 0.1959, "step": 2002 }, { "epoch": 0.30252227760157074, "grad_norm": 2.7574821317184632, "learning_rate": 9.75118493004825e-06, "loss": 0.1487, "step": 2003 }, { "epoch": 0.30267331218849114, "grad_norm": 3.6958614611951157, "learning_rate": 9.750938509560919e-06, "loss": 0.1739, "step": 2004 }, { "epoch": 0.3028243467754116, "grad_norm": 2.5140699819967374, "learning_rate": 9.750691970226248e-06, "loss": 0.1542, "step": 2005 }, { "epoch": 0.302975381362332, "grad_norm": 2.306273339469973, "learning_rate": 9.750445312050401e-06, "loss": 0.1596, "step": 2006 }, { "epoch": 0.3031264159492524, "grad_norm": 2.7446019606533376, "learning_rate": 9.750198535039554e-06, "loss": 0.1979, "step": 2007 }, { "epoch": 0.3032774505361728, "grad_norm": 3.013578675465058, "learning_rate": 9.749951639199875e-06, "loss": 0.1764, "step": 2008 }, { "epoch": 0.3034284851230932, "grad_norm": 1.9787302348141091, "learning_rate": 9.749704624537543e-06, "loss": 0.132, "step": 2009 }, { "epoch": 0.30357951971001357, "grad_norm": 2.0740580162729167, "learning_rate": 9.749457491058736e-06, "loss": 0.1515, "step": 2010 }, { "epoch": 0.303730554296934, "grad_norm": 2.6233156136064535, "learning_rate": 9.749210238769637e-06, "loss": 0.1619, "step": 2011 }, { "epoch": 0.3038815888838544, "grad_norm": 2.9077133922739735, "learning_rate": 9.748962867676431e-06, "loss": 0.1597, "step": 2012 }, { "epoch": 0.3040326234707748, "grad_norm": 4.456746903489802, "learning_rate": 9.748715377785307e-06, "loss": 0.1907, "step": 2013 }, { "epoch": 0.3041836580576952, "grad_norm": 2.689954134249373, "learning_rate": 9.748467769102455e-06, "loss": 0.1841, "step": 2014 }, { "epoch": 0.3043346926446156, "grad_norm": 3.0852599111754775, "learning_rate": 9.748220041634067e-06, "loss": 0.1653, "step": 2015 }, { "epoch": 0.304485727231536, "grad_norm": 2.217363241757122, "learning_rate": 9.747972195386344e-06, "loss": 0.1642, "step": 2016 }, { "epoch": 0.30463676181845645, "grad_norm": 3.2146244374093254, "learning_rate": 9.747724230365483e-06, "loss": 0.1903, "step": 2017 }, { "epoch": 0.30478779640537684, "grad_norm": 2.2174552326946393, "learning_rate": 9.747476146577688e-06, "loss": 0.1505, "step": 2018 }, { "epoch": 0.30493883099229724, "grad_norm": 2.330920348465638, "learning_rate": 9.747227944029168e-06, "loss": 0.1734, "step": 2019 }, { "epoch": 0.30508986557921763, "grad_norm": 2.174465984739428, "learning_rate": 9.746979622726127e-06, "loss": 0.1545, "step": 2020 }, { "epoch": 0.30524090016613803, "grad_norm": 2.7250698116229395, "learning_rate": 9.746731182674779e-06, "loss": 0.149, "step": 2021 }, { "epoch": 0.3053919347530584, "grad_norm": 2.4650399852322233, "learning_rate": 9.746482623881338e-06, "loss": 0.1508, "step": 2022 }, { "epoch": 0.3055429693399789, "grad_norm": 3.6596882581410277, "learning_rate": 9.746233946352024e-06, "loss": 0.1886, "step": 2023 }, { "epoch": 0.30569400392689927, "grad_norm": 3.283413455454362, "learning_rate": 9.745985150093056e-06, "loss": 0.186, "step": 2024 }, { "epoch": 0.30584503851381967, "grad_norm": 2.9469627548820134, "learning_rate": 9.745736235110658e-06, "loss": 0.1618, "step": 2025 }, { "epoch": 0.30599607310074006, "grad_norm": 2.9052141923357833, "learning_rate": 9.745487201411056e-06, "loss": 0.1782, "step": 2026 }, { "epoch": 0.30614710768766046, "grad_norm": 2.045794203242728, "learning_rate": 9.745238049000483e-06, "loss": 0.1319, "step": 2027 }, { "epoch": 0.30629814227458085, "grad_norm": 2.7349644858818856, "learning_rate": 9.74498877788517e-06, "loss": 0.1783, "step": 2028 }, { "epoch": 0.3064491768615013, "grad_norm": 2.344220413979339, "learning_rate": 9.74473938807135e-06, "loss": 0.1525, "step": 2029 }, { "epoch": 0.3066002114484217, "grad_norm": 2.181622113927258, "learning_rate": 9.744489879565263e-06, "loss": 0.1544, "step": 2030 }, { "epoch": 0.3067512460353421, "grad_norm": 2.3507840514909106, "learning_rate": 9.744240252373153e-06, "loss": 0.1708, "step": 2031 }, { "epoch": 0.3069022806222625, "grad_norm": 3.1242291012238192, "learning_rate": 9.743990506501262e-06, "loss": 0.177, "step": 2032 }, { "epoch": 0.3070533152091829, "grad_norm": 3.6672722645116878, "learning_rate": 9.74374064195584e-06, "loss": 0.1929, "step": 2033 }, { "epoch": 0.3072043497961033, "grad_norm": 2.592941705775127, "learning_rate": 9.743490658743134e-06, "loss": 0.1708, "step": 2034 }, { "epoch": 0.30735538438302373, "grad_norm": 3.644509861302582, "learning_rate": 9.743240556869398e-06, "loss": 0.1811, "step": 2035 }, { "epoch": 0.30750641896994413, "grad_norm": 2.497347512963412, "learning_rate": 9.742990336340892e-06, "loss": 0.1778, "step": 2036 }, { "epoch": 0.3076574535568645, "grad_norm": 2.0629390466133866, "learning_rate": 9.742739997163873e-06, "loss": 0.1559, "step": 2037 }, { "epoch": 0.3078084881437849, "grad_norm": 4.53061639755818, "learning_rate": 9.742489539344602e-06, "loss": 0.1647, "step": 2038 }, { "epoch": 0.3079595227307053, "grad_norm": 3.830442329650619, "learning_rate": 9.742238962889348e-06, "loss": 0.1682, "step": 2039 }, { "epoch": 0.3081105573176257, "grad_norm": 2.283909032093183, "learning_rate": 9.741988267804375e-06, "loss": 0.1525, "step": 2040 }, { "epoch": 0.30826159190454616, "grad_norm": 2.3491463924658924, "learning_rate": 9.741737454095959e-06, "loss": 0.1744, "step": 2041 }, { "epoch": 0.30841262649146656, "grad_norm": 4.340069864136351, "learning_rate": 9.74148652177037e-06, "loss": 0.2066, "step": 2042 }, { "epoch": 0.30856366107838695, "grad_norm": 3.432291228682933, "learning_rate": 9.741235470833887e-06, "loss": 0.1957, "step": 2043 }, { "epoch": 0.30871469566530735, "grad_norm": 2.7641243474756054, "learning_rate": 9.74098430129279e-06, "loss": 0.1806, "step": 2044 }, { "epoch": 0.30886573025222774, "grad_norm": 2.6037416586505056, "learning_rate": 9.740733013153363e-06, "loss": 0.187, "step": 2045 }, { "epoch": 0.30901676483914814, "grad_norm": 2.65225442042699, "learning_rate": 9.740481606421889e-06, "loss": 0.1731, "step": 2046 }, { "epoch": 0.3091677994260686, "grad_norm": 2.490710131009152, "learning_rate": 9.740230081104662e-06, "loss": 0.1542, "step": 2047 }, { "epoch": 0.309318834012989, "grad_norm": 2.4055836917811155, "learning_rate": 9.739978437207972e-06, "loss": 0.1699, "step": 2048 }, { "epoch": 0.3094698685999094, "grad_norm": 2.30871869571439, "learning_rate": 9.739726674738113e-06, "loss": 0.1539, "step": 2049 }, { "epoch": 0.3096209031868298, "grad_norm": 2.953315799290089, "learning_rate": 9.739474793701382e-06, "loss": 0.1955, "step": 2050 }, { "epoch": 0.3097719377737502, "grad_norm": 2.506984395083607, "learning_rate": 9.739222794104081e-06, "loss": 0.1498, "step": 2051 }, { "epoch": 0.30992297236067057, "grad_norm": 2.291049677786471, "learning_rate": 9.738970675952516e-06, "loss": 0.1756, "step": 2052 }, { "epoch": 0.310074006947591, "grad_norm": 2.9225075863073404, "learning_rate": 9.738718439252992e-06, "loss": 0.1835, "step": 2053 }, { "epoch": 0.3102250415345114, "grad_norm": 2.547750068227092, "learning_rate": 9.738466084011818e-06, "loss": 0.1696, "step": 2054 }, { "epoch": 0.3103760761214318, "grad_norm": 2.492361786951654, "learning_rate": 9.738213610235309e-06, "loss": 0.1708, "step": 2055 }, { "epoch": 0.3105271107083522, "grad_norm": 2.7010804091601917, "learning_rate": 9.737961017929777e-06, "loss": 0.174, "step": 2056 }, { "epoch": 0.3106781452952726, "grad_norm": 3.0630085793231094, "learning_rate": 9.737708307101548e-06, "loss": 0.1882, "step": 2057 }, { "epoch": 0.310829179882193, "grad_norm": 3.0948605327422922, "learning_rate": 9.737455477756935e-06, "loss": 0.2083, "step": 2058 }, { "epoch": 0.31098021446911345, "grad_norm": 2.3535784969503992, "learning_rate": 9.737202529902268e-06, "loss": 0.164, "step": 2059 }, { "epoch": 0.31113124905603384, "grad_norm": 2.5419974034401482, "learning_rate": 9.736949463543873e-06, "loss": 0.1821, "step": 2060 }, { "epoch": 0.31128228364295424, "grad_norm": 2.3630098578202605, "learning_rate": 9.736696278688082e-06, "loss": 0.1625, "step": 2061 }, { "epoch": 0.31143331822987463, "grad_norm": 2.5588097562704055, "learning_rate": 9.736442975341227e-06, "loss": 0.1765, "step": 2062 }, { "epoch": 0.31158435281679503, "grad_norm": 2.3187506940861513, "learning_rate": 9.736189553509646e-06, "loss": 0.1738, "step": 2063 }, { "epoch": 0.3117353874037154, "grad_norm": 2.17849699800243, "learning_rate": 9.735936013199673e-06, "loss": 0.1548, "step": 2064 }, { "epoch": 0.3118864219906359, "grad_norm": 4.258067977038643, "learning_rate": 9.73568235441766e-06, "loss": 0.2067, "step": 2065 }, { "epoch": 0.31203745657755627, "grad_norm": 2.4735447079687045, "learning_rate": 9.735428577169946e-06, "loss": 0.1501, "step": 2066 }, { "epoch": 0.31218849116447667, "grad_norm": 2.2965124432622, "learning_rate": 9.73517468146288e-06, "loss": 0.1551, "step": 2067 }, { "epoch": 0.31233952575139706, "grad_norm": 6.077726624629743, "learning_rate": 9.734920667302814e-06, "loss": 0.2255, "step": 2068 }, { "epoch": 0.31249056033831746, "grad_norm": 2.063938612419326, "learning_rate": 9.734666534696103e-06, "loss": 0.1457, "step": 2069 }, { "epoch": 0.31264159492523785, "grad_norm": 3.961139438147891, "learning_rate": 9.734412283649102e-06, "loss": 0.2139, "step": 2070 }, { "epoch": 0.3127926295121583, "grad_norm": 2.2459740371642094, "learning_rate": 9.734157914168174e-06, "loss": 0.1674, "step": 2071 }, { "epoch": 0.3129436640990787, "grad_norm": 2.812186292000179, "learning_rate": 9.733903426259681e-06, "loss": 0.1703, "step": 2072 }, { "epoch": 0.3130946986859991, "grad_norm": 2.2730066136950726, "learning_rate": 9.733648819929991e-06, "loss": 0.1535, "step": 2073 }, { "epoch": 0.3132457332729195, "grad_norm": 2.5250849582903228, "learning_rate": 9.733394095185469e-06, "loss": 0.1543, "step": 2074 }, { "epoch": 0.3133967678598399, "grad_norm": 2.2062921877668504, "learning_rate": 9.73313925203249e-06, "loss": 0.1554, "step": 2075 }, { "epoch": 0.3135478024467603, "grad_norm": 2.7691778401925933, "learning_rate": 9.73288429047743e-06, "loss": 0.1581, "step": 2076 }, { "epoch": 0.31369883703368073, "grad_norm": 2.247541110007363, "learning_rate": 9.732629210526663e-06, "loss": 0.1495, "step": 2077 }, { "epoch": 0.31384987162060113, "grad_norm": 2.9969535539592864, "learning_rate": 9.732374012186574e-06, "loss": 0.1657, "step": 2078 }, { "epoch": 0.3140009062075215, "grad_norm": 2.469495405175924, "learning_rate": 9.732118695463544e-06, "loss": 0.146, "step": 2079 }, { "epoch": 0.3141519407944419, "grad_norm": 2.6622312970205653, "learning_rate": 9.731863260363963e-06, "loss": 0.1538, "step": 2080 }, { "epoch": 0.3143029753813623, "grad_norm": 2.498316289958268, "learning_rate": 9.731607706894217e-06, "loss": 0.1513, "step": 2081 }, { "epoch": 0.3144540099682827, "grad_norm": 2.2786025893445085, "learning_rate": 9.731352035060703e-06, "loss": 0.1637, "step": 2082 }, { "epoch": 0.31460504455520316, "grad_norm": 2.5631292716179725, "learning_rate": 9.731096244869813e-06, "loss": 0.145, "step": 2083 }, { "epoch": 0.31475607914212356, "grad_norm": 3.743266733060591, "learning_rate": 9.73084033632795e-06, "loss": 0.183, "step": 2084 }, { "epoch": 0.31490711372904395, "grad_norm": 3.0491617855341544, "learning_rate": 9.73058430944151e-06, "loss": 0.2032, "step": 2085 }, { "epoch": 0.31505814831596435, "grad_norm": 2.5967000730403256, "learning_rate": 9.730328164216902e-06, "loss": 0.1707, "step": 2086 }, { "epoch": 0.31520918290288474, "grad_norm": 2.2630253141855365, "learning_rate": 9.730071900660532e-06, "loss": 0.1397, "step": 2087 }, { "epoch": 0.31536021748980514, "grad_norm": 3.159185999008745, "learning_rate": 9.72981551877881e-06, "loss": 0.1557, "step": 2088 }, { "epoch": 0.3155112520767256, "grad_norm": 3.0921792312174388, "learning_rate": 9.729559018578154e-06, "loss": 0.1561, "step": 2089 }, { "epoch": 0.315662286663646, "grad_norm": 2.266316727345954, "learning_rate": 9.729302400064975e-06, "loss": 0.1654, "step": 2090 }, { "epoch": 0.3158133212505664, "grad_norm": 2.541264649614601, "learning_rate": 9.729045663245693e-06, "loss": 0.1621, "step": 2091 }, { "epoch": 0.3159643558374868, "grad_norm": 3.02032401626209, "learning_rate": 9.728788808126733e-06, "loss": 0.1742, "step": 2092 }, { "epoch": 0.31611539042440717, "grad_norm": 2.4831006069499564, "learning_rate": 9.728531834714518e-06, "loss": 0.1732, "step": 2093 }, { "epoch": 0.31626642501132757, "grad_norm": 2.1784766940064935, "learning_rate": 9.72827474301548e-06, "loss": 0.1565, "step": 2094 }, { "epoch": 0.316417459598248, "grad_norm": 2.371785619518727, "learning_rate": 9.728017533036045e-06, "loss": 0.1569, "step": 2095 }, { "epoch": 0.3165684941851684, "grad_norm": 2.1708290310311855, "learning_rate": 9.727760204782652e-06, "loss": 0.1478, "step": 2096 }, { "epoch": 0.3167195287720888, "grad_norm": 1.9829369023746553, "learning_rate": 9.727502758261736e-06, "loss": 0.1455, "step": 2097 }, { "epoch": 0.3168705633590092, "grad_norm": 2.2469894853568966, "learning_rate": 9.727245193479737e-06, "loss": 0.1712, "step": 2098 }, { "epoch": 0.3170215979459296, "grad_norm": 2.4973149018094185, "learning_rate": 9.7269875104431e-06, "loss": 0.1605, "step": 2099 }, { "epoch": 0.31717263253285, "grad_norm": 2.3219128552033537, "learning_rate": 9.726729709158268e-06, "loss": 0.1678, "step": 2100 }, { "epoch": 0.31732366711977045, "grad_norm": 3.041920546243804, "learning_rate": 9.726471789631693e-06, "loss": 0.1924, "step": 2101 }, { "epoch": 0.31747470170669084, "grad_norm": 2.8051226934589164, "learning_rate": 9.726213751869825e-06, "loss": 0.1764, "step": 2102 }, { "epoch": 0.31762573629361124, "grad_norm": 2.996624741250484, "learning_rate": 9.725955595879118e-06, "loss": 0.1887, "step": 2103 }, { "epoch": 0.31777677088053163, "grad_norm": 2.757960412420056, "learning_rate": 9.725697321666035e-06, "loss": 0.1598, "step": 2104 }, { "epoch": 0.31792780546745203, "grad_norm": 2.2343921649419385, "learning_rate": 9.72543892923703e-06, "loss": 0.1414, "step": 2105 }, { "epoch": 0.3180788400543724, "grad_norm": 2.681215497519312, "learning_rate": 9.725180418598571e-06, "loss": 0.1849, "step": 2106 }, { "epoch": 0.3182298746412929, "grad_norm": 4.127254550178982, "learning_rate": 9.724921789757126e-06, "loss": 0.1385, "step": 2107 }, { "epoch": 0.31838090922821327, "grad_norm": 2.5967284780307143, "learning_rate": 9.724663042719162e-06, "loss": 0.1804, "step": 2108 }, { "epoch": 0.31853194381513367, "grad_norm": 2.4608343052092367, "learning_rate": 9.724404177491152e-06, "loss": 0.1637, "step": 2109 }, { "epoch": 0.31868297840205406, "grad_norm": 2.8079520798746227, "learning_rate": 9.724145194079572e-06, "loss": 0.1899, "step": 2110 }, { "epoch": 0.31883401298897446, "grad_norm": 2.5361199183731946, "learning_rate": 9.723886092490902e-06, "loss": 0.1547, "step": 2111 }, { "epoch": 0.31898504757589485, "grad_norm": 2.3503986927663196, "learning_rate": 9.72362687273162e-06, "loss": 0.1675, "step": 2112 }, { "epoch": 0.3191360821628153, "grad_norm": 2.436546916676474, "learning_rate": 9.723367534808214e-06, "loss": 0.1406, "step": 2113 }, { "epoch": 0.3192871167497357, "grad_norm": 2.6295401966981706, "learning_rate": 9.723108078727171e-06, "loss": 0.1538, "step": 2114 }, { "epoch": 0.3194381513366561, "grad_norm": 2.383816860085621, "learning_rate": 9.72284850449498e-06, "loss": 0.1386, "step": 2115 }, { "epoch": 0.3195891859235765, "grad_norm": 2.3989368378368834, "learning_rate": 9.722588812118137e-06, "loss": 0.1686, "step": 2116 }, { "epoch": 0.3197402205104969, "grad_norm": 7.344731450095502, "learning_rate": 9.722329001603135e-06, "loss": 0.2047, "step": 2117 }, { "epoch": 0.3198912550974173, "grad_norm": 2.88904630096637, "learning_rate": 9.722069072956474e-06, "loss": 0.1819, "step": 2118 }, { "epoch": 0.32004228968433773, "grad_norm": 3.3059224539639933, "learning_rate": 9.721809026184658e-06, "loss": 0.1734, "step": 2119 }, { "epoch": 0.32019332427125813, "grad_norm": 2.6345811036236655, "learning_rate": 9.721548861294192e-06, "loss": 0.1595, "step": 2120 }, { "epoch": 0.3203443588581785, "grad_norm": 2.5346191892045136, "learning_rate": 9.72128857829158e-06, "loss": 0.158, "step": 2121 }, { "epoch": 0.3204953934450989, "grad_norm": 2.2280842271097163, "learning_rate": 9.72102817718334e-06, "loss": 0.1787, "step": 2122 }, { "epoch": 0.3206464280320193, "grad_norm": 2.9353246203504693, "learning_rate": 9.720767657975983e-06, "loss": 0.183, "step": 2123 }, { "epoch": 0.3207974626189397, "grad_norm": 2.50933297980397, "learning_rate": 9.720507020676024e-06, "loss": 0.1827, "step": 2124 }, { "epoch": 0.32094849720586016, "grad_norm": 2.3041310881860775, "learning_rate": 9.720246265289985e-06, "loss": 0.1633, "step": 2125 }, { "epoch": 0.32109953179278056, "grad_norm": 2.7245537083206055, "learning_rate": 9.71998539182439e-06, "loss": 0.1791, "step": 2126 }, { "epoch": 0.32125056637970095, "grad_norm": 2.187959957679703, "learning_rate": 9.719724400285762e-06, "loss": 0.1596, "step": 2127 }, { "epoch": 0.32140160096662135, "grad_norm": 2.731692063359516, "learning_rate": 9.719463290680631e-06, "loss": 0.1829, "step": 2128 }, { "epoch": 0.32155263555354174, "grad_norm": 5.005554519085273, "learning_rate": 9.71920206301553e-06, "loss": 0.1777, "step": 2129 }, { "epoch": 0.32170367014046214, "grad_norm": 3.6941999666732492, "learning_rate": 9.718940717296992e-06, "loss": 0.1736, "step": 2130 }, { "epoch": 0.3218547047273826, "grad_norm": 2.711898725120484, "learning_rate": 9.718679253531557e-06, "loss": 0.1668, "step": 2131 }, { "epoch": 0.322005739314303, "grad_norm": 4.718729288111024, "learning_rate": 9.718417671725763e-06, "loss": 0.1826, "step": 2132 }, { "epoch": 0.3221567739012234, "grad_norm": 2.802309128372694, "learning_rate": 9.718155971886156e-06, "loss": 0.1724, "step": 2133 }, { "epoch": 0.3223078084881438, "grad_norm": 2.690829619199325, "learning_rate": 9.717894154019281e-06, "loss": 0.1657, "step": 2134 }, { "epoch": 0.32245884307506417, "grad_norm": 2.7156123301028154, "learning_rate": 9.717632218131688e-06, "loss": 0.1803, "step": 2135 }, { "epoch": 0.32260987766198457, "grad_norm": 3.2642862868006253, "learning_rate": 9.71737016422993e-06, "loss": 0.1724, "step": 2136 }, { "epoch": 0.322760912248905, "grad_norm": 2.225027276323307, "learning_rate": 9.71710799232056e-06, "loss": 0.1334, "step": 2137 }, { "epoch": 0.3229119468358254, "grad_norm": 3.7740362754066408, "learning_rate": 9.71684570241014e-06, "loss": 0.2015, "step": 2138 }, { "epoch": 0.3230629814227458, "grad_norm": 3.7232382865260503, "learning_rate": 9.716583294505229e-06, "loss": 0.1663, "step": 2139 }, { "epoch": 0.3232140160096662, "grad_norm": 2.3138576554931007, "learning_rate": 9.716320768612393e-06, "loss": 0.17, "step": 2140 }, { "epoch": 0.3233650505965866, "grad_norm": 3.0491020258853427, "learning_rate": 9.716058124738195e-06, "loss": 0.2023, "step": 2141 }, { "epoch": 0.323516085183507, "grad_norm": 2.6526684382403998, "learning_rate": 9.71579536288921e-06, "loss": 0.1817, "step": 2142 }, { "epoch": 0.32366711977042745, "grad_norm": 2.3608347938384457, "learning_rate": 9.71553248307201e-06, "loss": 0.1655, "step": 2143 }, { "epoch": 0.32381815435734784, "grad_norm": 2.761332748761188, "learning_rate": 9.71526948529317e-06, "loss": 0.175, "step": 2144 }, { "epoch": 0.32396918894426824, "grad_norm": 2.1697852401624576, "learning_rate": 9.71500636955927e-06, "loss": 0.144, "step": 2145 }, { "epoch": 0.32412022353118863, "grad_norm": 2.826595691407284, "learning_rate": 9.714743135876892e-06, "loss": 0.1608, "step": 2146 }, { "epoch": 0.32427125811810903, "grad_norm": 2.0507692196308844, "learning_rate": 9.714479784252618e-06, "loss": 0.1413, "step": 2147 }, { "epoch": 0.3244222927050294, "grad_norm": 2.7026019799126373, "learning_rate": 9.71421631469304e-06, "loss": 0.1575, "step": 2148 }, { "epoch": 0.3245733272919499, "grad_norm": 2.712031700795439, "learning_rate": 9.713952727204747e-06, "loss": 0.1856, "step": 2149 }, { "epoch": 0.32472436187887027, "grad_norm": 2.5520378665306223, "learning_rate": 9.713689021794332e-06, "loss": 0.1606, "step": 2150 }, { "epoch": 0.32487539646579067, "grad_norm": 4.953685134567533, "learning_rate": 9.713425198468394e-06, "loss": 0.17, "step": 2151 }, { "epoch": 0.32502643105271106, "grad_norm": 2.3913206319125364, "learning_rate": 9.71316125723353e-06, "loss": 0.1619, "step": 2152 }, { "epoch": 0.32517746563963146, "grad_norm": 2.8601278273879855, "learning_rate": 9.712897198096346e-06, "loss": 0.1744, "step": 2153 }, { "epoch": 0.32532850022655185, "grad_norm": 3.183305321824166, "learning_rate": 9.712633021063444e-06, "loss": 0.2031, "step": 2154 }, { "epoch": 0.3254795348134723, "grad_norm": 2.509527191549596, "learning_rate": 9.712368726141435e-06, "loss": 0.1893, "step": 2155 }, { "epoch": 0.3256305694003927, "grad_norm": 2.2208544391057288, "learning_rate": 9.712104313336929e-06, "loss": 0.1413, "step": 2156 }, { "epoch": 0.3257816039873131, "grad_norm": 2.6830531130339086, "learning_rate": 9.71183978265654e-06, "loss": 0.1671, "step": 2157 }, { "epoch": 0.3259326385742335, "grad_norm": 2.6834966776949165, "learning_rate": 9.711575134106888e-06, "loss": 0.1857, "step": 2158 }, { "epoch": 0.3260836731611539, "grad_norm": 2.77693545823736, "learning_rate": 9.71131036769459e-06, "loss": 0.1821, "step": 2159 }, { "epoch": 0.3262347077480743, "grad_norm": 2.50420679664558, "learning_rate": 9.711045483426274e-06, "loss": 0.1691, "step": 2160 }, { "epoch": 0.32638574233499473, "grad_norm": 2.104331889766809, "learning_rate": 9.710780481308561e-06, "loss": 0.1567, "step": 2161 }, { "epoch": 0.32653677692191513, "grad_norm": 2.3852683827893393, "learning_rate": 9.710515361348083e-06, "loss": 0.1654, "step": 2162 }, { "epoch": 0.3266878115088355, "grad_norm": 2.8398947505190697, "learning_rate": 9.71025012355147e-06, "loss": 0.1751, "step": 2163 }, { "epoch": 0.3268388460957559, "grad_norm": 2.215611257168738, "learning_rate": 9.70998476792536e-06, "loss": 0.1529, "step": 2164 }, { "epoch": 0.3269898806826763, "grad_norm": 2.1934742376422887, "learning_rate": 9.709719294476391e-06, "loss": 0.1672, "step": 2165 }, { "epoch": 0.3271409152695967, "grad_norm": 2.7215706714440664, "learning_rate": 9.7094537032112e-06, "loss": 0.1612, "step": 2166 }, { "epoch": 0.32729194985651716, "grad_norm": 2.38594447798294, "learning_rate": 9.709187994136435e-06, "loss": 0.1478, "step": 2167 }, { "epoch": 0.32744298444343756, "grad_norm": 2.506340636069281, "learning_rate": 9.70892216725874e-06, "loss": 0.1612, "step": 2168 }, { "epoch": 0.32759401903035795, "grad_norm": 2.3354346679103783, "learning_rate": 9.708656222584768e-06, "loss": 0.1435, "step": 2169 }, { "epoch": 0.32774505361727835, "grad_norm": 2.591449560723565, "learning_rate": 9.70839016012117e-06, "loss": 0.1827, "step": 2170 }, { "epoch": 0.32789608820419874, "grad_norm": 2.6749293571111887, "learning_rate": 9.708123979874599e-06, "loss": 0.1787, "step": 2171 }, { "epoch": 0.32804712279111914, "grad_norm": 3.449113035857043, "learning_rate": 9.707857681851719e-06, "loss": 0.1874, "step": 2172 }, { "epoch": 0.3281981573780396, "grad_norm": 2.198256669648097, "learning_rate": 9.707591266059187e-06, "loss": 0.1468, "step": 2173 }, { "epoch": 0.32834919196496, "grad_norm": 2.840594790968199, "learning_rate": 9.707324732503671e-06, "loss": 0.1854, "step": 2174 }, { "epoch": 0.3285002265518804, "grad_norm": 2.0843859027013116, "learning_rate": 9.707058081191836e-06, "loss": 0.1559, "step": 2175 }, { "epoch": 0.3286512611388008, "grad_norm": 3.5593989571469185, "learning_rate": 9.706791312130352e-06, "loss": 0.1804, "step": 2176 }, { "epoch": 0.32880229572572117, "grad_norm": 2.3001052758042118, "learning_rate": 9.706524425325897e-06, "loss": 0.153, "step": 2177 }, { "epoch": 0.32895333031264157, "grad_norm": 2.826043710002056, "learning_rate": 9.706257420785141e-06, "loss": 0.1587, "step": 2178 }, { "epoch": 0.329104364899562, "grad_norm": 2.704046941003026, "learning_rate": 9.705990298514767e-06, "loss": 0.1634, "step": 2179 }, { "epoch": 0.3292553994864824, "grad_norm": 2.8499029474012993, "learning_rate": 9.705723058521456e-06, "loss": 0.1755, "step": 2180 }, { "epoch": 0.3294064340734028, "grad_norm": 3.0421537605034965, "learning_rate": 9.705455700811895e-06, "loss": 0.2001, "step": 2181 }, { "epoch": 0.3295574686603232, "grad_norm": 3.5062140543114912, "learning_rate": 9.70518822539277e-06, "loss": 0.1863, "step": 2182 }, { "epoch": 0.3297085032472436, "grad_norm": 1.9938839832359616, "learning_rate": 9.704920632270773e-06, "loss": 0.1271, "step": 2183 }, { "epoch": 0.329859537834164, "grad_norm": 3.690339342646474, "learning_rate": 9.704652921452597e-06, "loss": 0.1927, "step": 2184 }, { "epoch": 0.33001057242108445, "grad_norm": 2.2725872979622435, "learning_rate": 9.704385092944941e-06, "loss": 0.1706, "step": 2185 }, { "epoch": 0.33016160700800484, "grad_norm": 2.9261227268664363, "learning_rate": 9.704117146754502e-06, "loss": 0.1924, "step": 2186 }, { "epoch": 0.33031264159492524, "grad_norm": 2.6867305293424946, "learning_rate": 9.703849082887984e-06, "loss": 0.1489, "step": 2187 }, { "epoch": 0.33046367618184563, "grad_norm": 2.367653749720036, "learning_rate": 9.703580901352094e-06, "loss": 0.1419, "step": 2188 }, { "epoch": 0.33061471076876603, "grad_norm": 2.4610907732439804, "learning_rate": 9.70331260215354e-06, "loss": 0.159, "step": 2189 }, { "epoch": 0.3307657453556864, "grad_norm": 2.241349685983203, "learning_rate": 9.703044185299035e-06, "loss": 0.1441, "step": 2190 }, { "epoch": 0.3309167799426069, "grad_norm": 1.919529097427449, "learning_rate": 9.702775650795291e-06, "loss": 0.1383, "step": 2191 }, { "epoch": 0.33106781452952727, "grad_norm": 2.9010434185609983, "learning_rate": 9.702506998649026e-06, "loss": 0.1824, "step": 2192 }, { "epoch": 0.33121884911644767, "grad_norm": 2.0275739827891703, "learning_rate": 9.70223822886696e-06, "loss": 0.1462, "step": 2193 }, { "epoch": 0.33136988370336806, "grad_norm": 3.122494474898684, "learning_rate": 9.70196934145582e-06, "loss": 0.1885, "step": 2194 }, { "epoch": 0.33152091829028846, "grad_norm": 2.684012423557912, "learning_rate": 9.701700336422329e-06, "loss": 0.1615, "step": 2195 }, { "epoch": 0.33167195287720885, "grad_norm": 2.306228378238809, "learning_rate": 9.701431213773216e-06, "loss": 0.1566, "step": 2196 }, { "epoch": 0.3318229874641293, "grad_norm": 2.0876887577343686, "learning_rate": 9.701161973515217e-06, "loss": 0.14, "step": 2197 }, { "epoch": 0.3319740220510497, "grad_norm": 2.2098317803628693, "learning_rate": 9.700892615655063e-06, "loss": 0.1505, "step": 2198 }, { "epoch": 0.3321250566379701, "grad_norm": 2.7808820234170164, "learning_rate": 9.700623140199494e-06, "loss": 0.1743, "step": 2199 }, { "epoch": 0.3322760912248905, "grad_norm": 2.3040287168484657, "learning_rate": 9.700353547155251e-06, "loss": 0.1673, "step": 2200 }, { "epoch": 0.3324271258118109, "grad_norm": 3.0918125751417977, "learning_rate": 9.700083836529077e-06, "loss": 0.1537, "step": 2201 }, { "epoch": 0.3325781603987313, "grad_norm": 2.3149405766276483, "learning_rate": 9.699814008327721e-06, "loss": 0.1613, "step": 2202 }, { "epoch": 0.33272919498565173, "grad_norm": 2.4252693597649793, "learning_rate": 9.699544062557933e-06, "loss": 0.1547, "step": 2203 }, { "epoch": 0.33288022957257213, "grad_norm": 3.7288574746027, "learning_rate": 9.699273999226462e-06, "loss": 0.1981, "step": 2204 }, { "epoch": 0.3330312641594925, "grad_norm": 3.435223867409875, "learning_rate": 9.699003818340067e-06, "loss": 0.17, "step": 2205 }, { "epoch": 0.3331822987464129, "grad_norm": 4.309103955567857, "learning_rate": 9.698733519905507e-06, "loss": 0.17, "step": 2206 }, { "epoch": 0.3333333333333333, "grad_norm": 2.381367061357787, "learning_rate": 9.698463103929542e-06, "loss": 0.1661, "step": 2207 }, { "epoch": 0.33348436792025377, "grad_norm": 2.4294189413642147, "learning_rate": 9.698192570418939e-06, "loss": 0.1594, "step": 2208 }, { "epoch": 0.33363540250717416, "grad_norm": 2.142265456986441, "learning_rate": 9.697921919380462e-06, "loss": 0.1553, "step": 2209 }, { "epoch": 0.33378643709409456, "grad_norm": 2.5768817003719393, "learning_rate": 9.697651150820884e-06, "loss": 0.1727, "step": 2210 }, { "epoch": 0.33393747168101495, "grad_norm": 2.3938914506859867, "learning_rate": 9.697380264746978e-06, "loss": 0.1348, "step": 2211 }, { "epoch": 0.33408850626793535, "grad_norm": 2.35236183525933, "learning_rate": 9.69710926116552e-06, "loss": 0.1492, "step": 2212 }, { "epoch": 0.33423954085485574, "grad_norm": 2.433573152515671, "learning_rate": 9.696838140083289e-06, "loss": 0.1684, "step": 2213 }, { "epoch": 0.3343905754417762, "grad_norm": 2.6925340318920568, "learning_rate": 9.69656690150707e-06, "loss": 0.168, "step": 2214 }, { "epoch": 0.3345416100286966, "grad_norm": 2.7589404898882863, "learning_rate": 9.696295545443643e-06, "loss": 0.1491, "step": 2215 }, { "epoch": 0.334692644615617, "grad_norm": 2.9490897675050642, "learning_rate": 9.696024071899801e-06, "loss": 0.1823, "step": 2216 }, { "epoch": 0.3348436792025374, "grad_norm": 2.9326272187577787, "learning_rate": 9.695752480882332e-06, "loss": 0.1824, "step": 2217 }, { "epoch": 0.3349947137894578, "grad_norm": 2.690279472673437, "learning_rate": 9.69548077239803e-06, "loss": 0.1815, "step": 2218 }, { "epoch": 0.33514574837637817, "grad_norm": 3.6310832847333905, "learning_rate": 9.695208946453695e-06, "loss": 0.1884, "step": 2219 }, { "epoch": 0.3352967829632986, "grad_norm": 2.5955152948535734, "learning_rate": 9.694937003056126e-06, "loss": 0.1539, "step": 2220 }, { "epoch": 0.335447817550219, "grad_norm": 3.351990288127552, "learning_rate": 9.694664942212122e-06, "loss": 0.1969, "step": 2221 }, { "epoch": 0.3355988521371394, "grad_norm": 3.0808991434364352, "learning_rate": 9.694392763928493e-06, "loss": 0.1688, "step": 2222 }, { "epoch": 0.3357498867240598, "grad_norm": 2.383748935392356, "learning_rate": 9.694120468212046e-06, "loss": 0.1664, "step": 2223 }, { "epoch": 0.3359009213109802, "grad_norm": 2.2536759291250394, "learning_rate": 9.693848055069593e-06, "loss": 0.1444, "step": 2224 }, { "epoch": 0.3360519558979006, "grad_norm": 2.886289901774464, "learning_rate": 9.693575524507947e-06, "loss": 0.202, "step": 2225 }, { "epoch": 0.33620299048482105, "grad_norm": 2.7864761739831048, "learning_rate": 9.693302876533929e-06, "loss": 0.159, "step": 2226 }, { "epoch": 0.33635402507174145, "grad_norm": 3.4365044894916625, "learning_rate": 9.693030111154356e-06, "loss": 0.1877, "step": 2227 }, { "epoch": 0.33650505965866184, "grad_norm": 2.787770597383624, "learning_rate": 9.692757228376053e-06, "loss": 0.1826, "step": 2228 }, { "epoch": 0.33665609424558224, "grad_norm": 3.3905673004561825, "learning_rate": 9.692484228205845e-06, "loss": 0.1545, "step": 2229 }, { "epoch": 0.33680712883250263, "grad_norm": 2.7451129051649725, "learning_rate": 9.692211110650564e-06, "loss": 0.17, "step": 2230 }, { "epoch": 0.33695816341942303, "grad_norm": 2.760139004602134, "learning_rate": 9.69193787571704e-06, "loss": 0.167, "step": 2231 }, { "epoch": 0.3371091980063435, "grad_norm": 2.833974422579342, "learning_rate": 9.691664523412108e-06, "loss": 0.153, "step": 2232 }, { "epoch": 0.3372602325932639, "grad_norm": 2.812669720072746, "learning_rate": 9.691391053742605e-06, "loss": 0.1907, "step": 2233 }, { "epoch": 0.33741126718018427, "grad_norm": 2.524853715358271, "learning_rate": 9.691117466715377e-06, "loss": 0.1733, "step": 2234 }, { "epoch": 0.33756230176710467, "grad_norm": 2.7708986036391243, "learning_rate": 9.69084376233726e-06, "loss": 0.1755, "step": 2235 }, { "epoch": 0.33771333635402506, "grad_norm": 3.4517310280727513, "learning_rate": 9.69056994061511e-06, "loss": 0.1648, "step": 2236 }, { "epoch": 0.33786437094094546, "grad_norm": 2.1477659187376013, "learning_rate": 9.690296001555771e-06, "loss": 0.1471, "step": 2237 }, { "epoch": 0.3380154055278659, "grad_norm": 2.762020608118704, "learning_rate": 9.690021945166095e-06, "loss": 0.1798, "step": 2238 }, { "epoch": 0.3381664401147863, "grad_norm": 2.711986497371396, "learning_rate": 9.689747771452943e-06, "loss": 0.1563, "step": 2239 }, { "epoch": 0.3383174747017067, "grad_norm": 4.19578476865431, "learning_rate": 9.689473480423166e-06, "loss": 0.1988, "step": 2240 }, { "epoch": 0.3384685092886271, "grad_norm": 2.6728804117776637, "learning_rate": 9.689199072083633e-06, "loss": 0.1721, "step": 2241 }, { "epoch": 0.3386195438755475, "grad_norm": 2.3376437787190283, "learning_rate": 9.688924546441204e-06, "loss": 0.179, "step": 2242 }, { "epoch": 0.3387705784624679, "grad_norm": 2.4831102450710762, "learning_rate": 9.688649903502746e-06, "loss": 0.1842, "step": 2243 }, { "epoch": 0.33892161304938834, "grad_norm": 3.4103032484579763, "learning_rate": 9.688375143275134e-06, "loss": 0.2015, "step": 2244 }, { "epoch": 0.33907264763630873, "grad_norm": 2.8969985492309367, "learning_rate": 9.688100265765237e-06, "loss": 0.1804, "step": 2245 }, { "epoch": 0.33922368222322913, "grad_norm": 2.1468536124753195, "learning_rate": 9.687825270979932e-06, "loss": 0.1575, "step": 2246 }, { "epoch": 0.3393747168101495, "grad_norm": 2.310427813075007, "learning_rate": 9.6875501589261e-06, "loss": 0.1534, "step": 2247 }, { "epoch": 0.3395257513970699, "grad_norm": 2.4255912387726837, "learning_rate": 9.687274929610621e-06, "loss": 0.1811, "step": 2248 }, { "epoch": 0.3396767859839903, "grad_norm": 2.193328520456946, "learning_rate": 9.68699958304038e-06, "loss": 0.1619, "step": 2249 }, { "epoch": 0.33982782057091077, "grad_norm": 2.4881683477766634, "learning_rate": 9.686724119222266e-06, "loss": 0.1566, "step": 2250 }, { "epoch": 0.33997885515783116, "grad_norm": 2.448369693385396, "learning_rate": 9.68644853816317e-06, "loss": 0.1486, "step": 2251 }, { "epoch": 0.34012988974475156, "grad_norm": 2.6603545279057297, "learning_rate": 9.686172839869985e-06, "loss": 0.1632, "step": 2252 }, { "epoch": 0.34028092433167195, "grad_norm": 2.389369466968686, "learning_rate": 9.685897024349609e-06, "loss": 0.1622, "step": 2253 }, { "epoch": 0.34043195891859235, "grad_norm": 2.2467869293565523, "learning_rate": 9.68562109160894e-06, "loss": 0.1695, "step": 2254 }, { "epoch": 0.34058299350551274, "grad_norm": 2.5604291506342394, "learning_rate": 9.68534504165488e-06, "loss": 0.1981, "step": 2255 }, { "epoch": 0.3407340280924332, "grad_norm": 3.1526054607624294, "learning_rate": 9.685068874494337e-06, "loss": 0.185, "step": 2256 }, { "epoch": 0.3408850626793536, "grad_norm": 2.7110339034353492, "learning_rate": 9.68479259013422e-06, "loss": 0.171, "step": 2257 }, { "epoch": 0.341036097266274, "grad_norm": 3.127580514579895, "learning_rate": 9.684516188581436e-06, "loss": 0.2168, "step": 2258 }, { "epoch": 0.3411871318531944, "grad_norm": 2.0049297223604903, "learning_rate": 9.684239669842904e-06, "loss": 0.1557, "step": 2259 }, { "epoch": 0.3413381664401148, "grad_norm": 2.8653670636841784, "learning_rate": 9.68396303392554e-06, "loss": 0.1943, "step": 2260 }, { "epoch": 0.34148920102703517, "grad_norm": 2.4819111024745752, "learning_rate": 9.683686280836262e-06, "loss": 0.1731, "step": 2261 }, { "epoch": 0.3416402356139556, "grad_norm": 2.215000861615687, "learning_rate": 9.683409410581995e-06, "loss": 0.1755, "step": 2262 }, { "epoch": 0.341791270200876, "grad_norm": 2.5277613868113065, "learning_rate": 9.683132423169666e-06, "loss": 0.1867, "step": 2263 }, { "epoch": 0.3419423047877964, "grad_norm": 2.238842303943618, "learning_rate": 9.682855318606202e-06, "loss": 0.1496, "step": 2264 }, { "epoch": 0.3420933393747168, "grad_norm": 2.071243496692368, "learning_rate": 9.682578096898534e-06, "loss": 0.1443, "step": 2265 }, { "epoch": 0.3422443739616372, "grad_norm": 2.19724490348089, "learning_rate": 9.6823007580536e-06, "loss": 0.1497, "step": 2266 }, { "epoch": 0.3423954085485576, "grad_norm": 2.5209070202622916, "learning_rate": 9.682023302078337e-06, "loss": 0.1735, "step": 2267 }, { "epoch": 0.34254644313547805, "grad_norm": 2.299992655366491, "learning_rate": 9.681745728979684e-06, "loss": 0.1395, "step": 2268 }, { "epoch": 0.34269747772239845, "grad_norm": 2.6386721438241474, "learning_rate": 9.681468038764588e-06, "loss": 0.1622, "step": 2269 }, { "epoch": 0.34284851230931884, "grad_norm": 2.658169525810094, "learning_rate": 9.68119023143999e-06, "loss": 0.1657, "step": 2270 }, { "epoch": 0.34299954689623924, "grad_norm": 2.9215130750632476, "learning_rate": 9.680912307012843e-06, "loss": 0.1695, "step": 2271 }, { "epoch": 0.34315058148315963, "grad_norm": 2.4904686894444987, "learning_rate": 9.680634265490104e-06, "loss": 0.1715, "step": 2272 }, { "epoch": 0.34330161607008003, "grad_norm": 2.4058597896599814, "learning_rate": 9.680356106878718e-06, "loss": 0.1606, "step": 2273 }, { "epoch": 0.3434526506570005, "grad_norm": 2.047709967981168, "learning_rate": 9.680077831185651e-06, "loss": 0.1414, "step": 2274 }, { "epoch": 0.3436036852439209, "grad_norm": 2.49445367129809, "learning_rate": 9.679799438417862e-06, "loss": 0.1776, "step": 2275 }, { "epoch": 0.34375471983084127, "grad_norm": 2.376004696152282, "learning_rate": 9.679520928582316e-06, "loss": 0.1792, "step": 2276 }, { "epoch": 0.34390575441776167, "grad_norm": 2.3564456327961936, "learning_rate": 9.679242301685979e-06, "loss": 0.1816, "step": 2277 }, { "epoch": 0.34405678900468206, "grad_norm": 2.9709158426195654, "learning_rate": 9.67896355773582e-06, "loss": 0.1916, "step": 2278 }, { "epoch": 0.34420782359160246, "grad_norm": 2.2481098052107296, "learning_rate": 9.678684696738815e-06, "loss": 0.1582, "step": 2279 }, { "epoch": 0.3443588581785229, "grad_norm": 3.112060956075879, "learning_rate": 9.678405718701937e-06, "loss": 0.2119, "step": 2280 }, { "epoch": 0.3445098927654433, "grad_norm": 2.677019135891741, "learning_rate": 9.678126623632167e-06, "loss": 0.154, "step": 2281 }, { "epoch": 0.3446609273523637, "grad_norm": 3.207216597343383, "learning_rate": 9.677847411536484e-06, "loss": 0.1793, "step": 2282 }, { "epoch": 0.3448119619392841, "grad_norm": 2.8613121694054837, "learning_rate": 9.677568082421875e-06, "loss": 0.1471, "step": 2283 }, { "epoch": 0.3449629965262045, "grad_norm": 2.2953325354806133, "learning_rate": 9.677288636295328e-06, "loss": 0.1494, "step": 2284 }, { "epoch": 0.3451140311131249, "grad_norm": 2.600915305357451, "learning_rate": 9.67700907316383e-06, "loss": 0.1657, "step": 2285 }, { "epoch": 0.34526506570004534, "grad_norm": 2.657072856295686, "learning_rate": 9.676729393034378e-06, "loss": 0.167, "step": 2286 }, { "epoch": 0.34541610028696573, "grad_norm": 3.256841025468264, "learning_rate": 9.676449595913969e-06, "loss": 0.1587, "step": 2287 }, { "epoch": 0.34556713487388613, "grad_norm": 3.0672578853118884, "learning_rate": 9.676169681809597e-06, "loss": 0.1887, "step": 2288 }, { "epoch": 0.3457181694608065, "grad_norm": 2.2167365564100345, "learning_rate": 9.675889650728269e-06, "loss": 0.1577, "step": 2289 }, { "epoch": 0.3458692040477269, "grad_norm": 2.605387224702682, "learning_rate": 9.67560950267699e-06, "loss": 0.1561, "step": 2290 }, { "epoch": 0.3460202386346473, "grad_norm": 2.082227934533779, "learning_rate": 9.675329237662766e-06, "loss": 0.1448, "step": 2291 }, { "epoch": 0.34617127322156777, "grad_norm": 2.328244228390361, "learning_rate": 9.675048855692608e-06, "loss": 0.1516, "step": 2292 }, { "epoch": 0.34632230780848816, "grad_norm": 3.19204460736859, "learning_rate": 9.67476835677353e-06, "loss": 0.185, "step": 2293 }, { "epoch": 0.34647334239540856, "grad_norm": 3.5059052410445624, "learning_rate": 9.67448774091255e-06, "loss": 0.1947, "step": 2294 }, { "epoch": 0.34662437698232895, "grad_norm": 2.703132226310503, "learning_rate": 9.67420700811669e-06, "loss": 0.1646, "step": 2295 }, { "epoch": 0.34677541156924935, "grad_norm": 2.5229698926706967, "learning_rate": 9.673926158392967e-06, "loss": 0.1517, "step": 2296 }, { "epoch": 0.34692644615616974, "grad_norm": 2.0350550017939932, "learning_rate": 9.673645191748411e-06, "loss": 0.1427, "step": 2297 }, { "epoch": 0.3470774807430902, "grad_norm": 2.5406704638371096, "learning_rate": 9.673364108190049e-06, "loss": 0.1695, "step": 2298 }, { "epoch": 0.3472285153300106, "grad_norm": 3.090563239535391, "learning_rate": 9.67308290772491e-06, "loss": 0.1946, "step": 2299 }, { "epoch": 0.347379549916931, "grad_norm": 2.843362237550477, "learning_rate": 9.672801590360034e-06, "loss": 0.1884, "step": 2300 }, { "epoch": 0.3475305845038514, "grad_norm": 2.4625684529257317, "learning_rate": 9.672520156102454e-06, "loss": 0.1581, "step": 2301 }, { "epoch": 0.3476816190907718, "grad_norm": 2.367995172197763, "learning_rate": 9.672238604959212e-06, "loss": 0.1634, "step": 2302 }, { "epoch": 0.34783265367769217, "grad_norm": 2.3991704604284956, "learning_rate": 9.67195693693735e-06, "loss": 0.1615, "step": 2303 }, { "epoch": 0.3479836882646126, "grad_norm": 3.501703433303054, "learning_rate": 9.671675152043916e-06, "loss": 0.171, "step": 2304 }, { "epoch": 0.348134722851533, "grad_norm": 2.4810788464192957, "learning_rate": 9.671393250285957e-06, "loss": 0.1778, "step": 2305 }, { "epoch": 0.3482857574384534, "grad_norm": 2.5582273327936975, "learning_rate": 9.671111231670524e-06, "loss": 0.181, "step": 2306 }, { "epoch": 0.3484367920253738, "grad_norm": 2.9174589864700153, "learning_rate": 9.670829096204677e-06, "loss": 0.1726, "step": 2307 }, { "epoch": 0.3485878266122942, "grad_norm": 2.3162742667551277, "learning_rate": 9.670546843895467e-06, "loss": 0.1513, "step": 2308 }, { "epoch": 0.3487388611992146, "grad_norm": 2.7201870701741515, "learning_rate": 9.67026447474996e-06, "loss": 0.1529, "step": 2309 }, { "epoch": 0.34888989578613505, "grad_norm": 2.9358350772762587, "learning_rate": 9.669981988775217e-06, "loss": 0.1815, "step": 2310 }, { "epoch": 0.34904093037305545, "grad_norm": 2.645604090688502, "learning_rate": 9.669699385978303e-06, "loss": 0.1674, "step": 2311 }, { "epoch": 0.34919196495997584, "grad_norm": 2.0033775705160686, "learning_rate": 9.669416666366293e-06, "loss": 0.1434, "step": 2312 }, { "epoch": 0.34934299954689624, "grad_norm": 2.9633098662284376, "learning_rate": 9.669133829946255e-06, "loss": 0.1826, "step": 2313 }, { "epoch": 0.34949403413381663, "grad_norm": 3.1616335531693314, "learning_rate": 9.668850876725263e-06, "loss": 0.1718, "step": 2314 }, { "epoch": 0.34964506872073703, "grad_norm": 2.893813791485139, "learning_rate": 9.668567806710402e-06, "loss": 0.1655, "step": 2315 }, { "epoch": 0.3497961033076575, "grad_norm": 3.004689790959727, "learning_rate": 9.668284619908745e-06, "loss": 0.1774, "step": 2316 }, { "epoch": 0.3499471378945779, "grad_norm": 2.3039415411309205, "learning_rate": 9.668001316327382e-06, "loss": 0.154, "step": 2317 }, { "epoch": 0.35009817248149827, "grad_norm": 2.346179962679111, "learning_rate": 9.667717895973397e-06, "loss": 0.1523, "step": 2318 }, { "epoch": 0.35024920706841867, "grad_norm": 2.7372614000168047, "learning_rate": 9.66743435885388e-06, "loss": 0.1915, "step": 2319 }, { "epoch": 0.35040024165533906, "grad_norm": 2.729415663892963, "learning_rate": 9.667150704975925e-06, "loss": 0.1898, "step": 2320 }, { "epoch": 0.35055127624225946, "grad_norm": 2.866293021934122, "learning_rate": 9.666866934346626e-06, "loss": 0.159, "step": 2321 }, { "epoch": 0.3507023108291799, "grad_norm": 2.7409448020787184, "learning_rate": 9.666583046973085e-06, "loss": 0.1887, "step": 2322 }, { "epoch": 0.3508533454161003, "grad_norm": 2.3557299240398804, "learning_rate": 9.6662990428624e-06, "loss": 0.1627, "step": 2323 }, { "epoch": 0.3510043800030207, "grad_norm": 2.207472670859635, "learning_rate": 9.666014922021679e-06, "loss": 0.1407, "step": 2324 }, { "epoch": 0.3511554145899411, "grad_norm": 3.3876559911170214, "learning_rate": 9.665730684458026e-06, "loss": 0.174, "step": 2325 }, { "epoch": 0.3513064491768615, "grad_norm": 2.791987803594905, "learning_rate": 9.665446330178552e-06, "loss": 0.1943, "step": 2326 }, { "epoch": 0.3514574837637819, "grad_norm": 2.548971616486758, "learning_rate": 9.665161859190372e-06, "loss": 0.1719, "step": 2327 }, { "epoch": 0.35160851835070234, "grad_norm": 2.095004547215813, "learning_rate": 9.6648772715006e-06, "loss": 0.1475, "step": 2328 }, { "epoch": 0.35175955293762273, "grad_norm": 2.8023158067414173, "learning_rate": 9.66459256711636e-06, "loss": 0.1805, "step": 2329 }, { "epoch": 0.35191058752454313, "grad_norm": 2.242593249007611, "learning_rate": 9.664307746044767e-06, "loss": 0.1625, "step": 2330 }, { "epoch": 0.3520616221114635, "grad_norm": 2.9756602183533865, "learning_rate": 9.664022808292953e-06, "loss": 0.1732, "step": 2331 }, { "epoch": 0.3522126566983839, "grad_norm": 1.8241480584387288, "learning_rate": 9.663737753868039e-06, "loss": 0.1482, "step": 2332 }, { "epoch": 0.3523636912853043, "grad_norm": 2.4401984720669443, "learning_rate": 9.66345258277716e-06, "loss": 0.1876, "step": 2333 }, { "epoch": 0.35251472587222477, "grad_norm": 2.3352174673087, "learning_rate": 9.66316729502745e-06, "loss": 0.1477, "step": 2334 }, { "epoch": 0.35266576045914516, "grad_norm": 2.5725637647111794, "learning_rate": 9.662881890626043e-06, "loss": 0.1537, "step": 2335 }, { "epoch": 0.35281679504606556, "grad_norm": 2.4752807227294618, "learning_rate": 9.66259636958008e-06, "loss": 0.1808, "step": 2336 }, { "epoch": 0.35296782963298595, "grad_norm": 2.1976042708195718, "learning_rate": 9.662310731896705e-06, "loss": 0.1488, "step": 2337 }, { "epoch": 0.35311886421990635, "grad_norm": 2.1611604169072867, "learning_rate": 9.662024977583061e-06, "loss": 0.156, "step": 2338 }, { "epoch": 0.35326989880682674, "grad_norm": 4.0277636724074934, "learning_rate": 9.661739106646298e-06, "loss": 0.2028, "step": 2339 }, { "epoch": 0.3534209333937472, "grad_norm": 2.3797230500132853, "learning_rate": 9.661453119093565e-06, "loss": 0.1795, "step": 2340 }, { "epoch": 0.3535719679806676, "grad_norm": 2.949310165364204, "learning_rate": 9.661167014932018e-06, "loss": 0.1632, "step": 2341 }, { "epoch": 0.353723002567588, "grad_norm": 2.4383682567227187, "learning_rate": 9.660880794168813e-06, "loss": 0.1705, "step": 2342 }, { "epoch": 0.3538740371545084, "grad_norm": 2.4870667113831746, "learning_rate": 9.660594456811112e-06, "loss": 0.1592, "step": 2343 }, { "epoch": 0.3540250717414288, "grad_norm": 2.8775042348652673, "learning_rate": 9.660308002866075e-06, "loss": 0.1461, "step": 2344 }, { "epoch": 0.35417610632834917, "grad_norm": 2.5004180000628793, "learning_rate": 9.66002143234087e-06, "loss": 0.155, "step": 2345 }, { "epoch": 0.3543271409152696, "grad_norm": 2.3917795562113606, "learning_rate": 9.659734745242663e-06, "loss": 0.1499, "step": 2346 }, { "epoch": 0.35447817550219, "grad_norm": 3.041421792938887, "learning_rate": 9.65944794157863e-06, "loss": 0.1816, "step": 2347 }, { "epoch": 0.3546292100891104, "grad_norm": 2.0495166580371134, "learning_rate": 9.659161021355942e-06, "loss": 0.1314, "step": 2348 }, { "epoch": 0.3547802446760308, "grad_norm": 5.231894448559906, "learning_rate": 9.658873984581778e-06, "loss": 0.1537, "step": 2349 }, { "epoch": 0.3549312792629512, "grad_norm": 3.017498592935146, "learning_rate": 9.658586831263317e-06, "loss": 0.186, "step": 2350 }, { "epoch": 0.3550823138498716, "grad_norm": 2.0590006030097885, "learning_rate": 9.658299561407745e-06, "loss": 0.1578, "step": 2351 }, { "epoch": 0.35523334843679205, "grad_norm": 2.703283502434135, "learning_rate": 9.658012175022246e-06, "loss": 0.1554, "step": 2352 }, { "epoch": 0.35538438302371245, "grad_norm": 2.326393329186139, "learning_rate": 9.657724672114008e-06, "loss": 0.151, "step": 2353 }, { "epoch": 0.35553541761063284, "grad_norm": 2.4347128439795287, "learning_rate": 9.657437052690228e-06, "loss": 0.1606, "step": 2354 }, { "epoch": 0.35568645219755324, "grad_norm": 3.067755802059363, "learning_rate": 9.657149316758094e-06, "loss": 0.1719, "step": 2355 }, { "epoch": 0.35583748678447363, "grad_norm": 3.8291191212024756, "learning_rate": 9.656861464324809e-06, "loss": 0.1718, "step": 2356 }, { "epoch": 0.35598852137139403, "grad_norm": 2.7194905961305733, "learning_rate": 9.656573495397573e-06, "loss": 0.1662, "step": 2357 }, { "epoch": 0.3561395559583145, "grad_norm": 2.728763734915162, "learning_rate": 9.656285409983587e-06, "loss": 0.1767, "step": 2358 }, { "epoch": 0.3562905905452349, "grad_norm": 2.741400180161744, "learning_rate": 9.655997208090063e-06, "loss": 0.1683, "step": 2359 }, { "epoch": 0.35644162513215527, "grad_norm": 2.2699738915928593, "learning_rate": 9.655708889724204e-06, "loss": 0.1522, "step": 2360 }, { "epoch": 0.35659265971907567, "grad_norm": 2.441141054848911, "learning_rate": 9.655420454893227e-06, "loss": 0.1884, "step": 2361 }, { "epoch": 0.35674369430599606, "grad_norm": 2.7382410238444974, "learning_rate": 9.655131903604345e-06, "loss": 0.18, "step": 2362 }, { "epoch": 0.35689472889291646, "grad_norm": 3.310201044963322, "learning_rate": 9.654843235864778e-06, "loss": 0.1621, "step": 2363 }, { "epoch": 0.3570457634798369, "grad_norm": 2.1796235305249034, "learning_rate": 9.654554451681745e-06, "loss": 0.1519, "step": 2364 }, { "epoch": 0.3571967980667573, "grad_norm": 2.2428188899911436, "learning_rate": 9.654265551062474e-06, "loss": 0.1384, "step": 2365 }, { "epoch": 0.3573478326536777, "grad_norm": 2.7955126346211534, "learning_rate": 9.653976534014187e-06, "loss": 0.1727, "step": 2366 }, { "epoch": 0.3574988672405981, "grad_norm": 2.4461908197984576, "learning_rate": 9.653687400544118e-06, "loss": 0.1415, "step": 2367 }, { "epoch": 0.3576499018275185, "grad_norm": 2.6833948477097604, "learning_rate": 9.653398150659498e-06, "loss": 0.1596, "step": 2368 }, { "epoch": 0.3578009364144389, "grad_norm": 1.9791757545355813, "learning_rate": 9.653108784367564e-06, "loss": 0.1484, "step": 2369 }, { "epoch": 0.35795197100135934, "grad_norm": 2.633507666600223, "learning_rate": 9.652819301675552e-06, "loss": 0.1518, "step": 2370 }, { "epoch": 0.35810300558827973, "grad_norm": 2.2087289025520858, "learning_rate": 9.652529702590705e-06, "loss": 0.1584, "step": 2371 }, { "epoch": 0.35825404017520013, "grad_norm": 2.9797754128368443, "learning_rate": 9.652239987120268e-06, "loss": 0.2025, "step": 2372 }, { "epoch": 0.3584050747621205, "grad_norm": 2.77373226370562, "learning_rate": 9.651950155271488e-06, "loss": 0.1562, "step": 2373 }, { "epoch": 0.3585561093490409, "grad_norm": 2.616748836727735, "learning_rate": 9.651660207051617e-06, "loss": 0.151, "step": 2374 }, { "epoch": 0.3587071439359613, "grad_norm": 2.266157781448491, "learning_rate": 9.651370142467904e-06, "loss": 0.1575, "step": 2375 }, { "epoch": 0.35885817852288177, "grad_norm": 2.4929671972117746, "learning_rate": 9.651079961527608e-06, "loss": 0.1725, "step": 2376 }, { "epoch": 0.35900921310980216, "grad_norm": 5.6898878702274605, "learning_rate": 9.65078966423799e-06, "loss": 0.1944, "step": 2377 }, { "epoch": 0.35916024769672256, "grad_norm": 3.113774350529292, "learning_rate": 9.650499250606308e-06, "loss": 0.1629, "step": 2378 }, { "epoch": 0.35931128228364295, "grad_norm": 2.5212833536029042, "learning_rate": 9.650208720639829e-06, "loss": 0.1606, "step": 2379 }, { "epoch": 0.35946231687056335, "grad_norm": 2.4886512501931803, "learning_rate": 9.64991807434582e-06, "loss": 0.1787, "step": 2380 }, { "epoch": 0.35961335145748374, "grad_norm": 5.3312312515584095, "learning_rate": 9.649627311731552e-06, "loss": 0.1965, "step": 2381 }, { "epoch": 0.3597643860444042, "grad_norm": 2.656444174635714, "learning_rate": 9.6493364328043e-06, "loss": 0.1734, "step": 2382 }, { "epoch": 0.3599154206313246, "grad_norm": 2.0045387677933446, "learning_rate": 9.649045437571337e-06, "loss": 0.1483, "step": 2383 }, { "epoch": 0.360066455218245, "grad_norm": 2.8454259800717887, "learning_rate": 9.648754326039945e-06, "loss": 0.1586, "step": 2384 }, { "epoch": 0.3602174898051654, "grad_norm": 3.2349999307426445, "learning_rate": 9.648463098217407e-06, "loss": 0.1707, "step": 2385 }, { "epoch": 0.3603685243920858, "grad_norm": 2.959617646374647, "learning_rate": 9.648171754111006e-06, "loss": 0.1806, "step": 2386 }, { "epoch": 0.36051955897900617, "grad_norm": 6.713906611932972, "learning_rate": 9.64788029372803e-06, "loss": 0.1944, "step": 2387 }, { "epoch": 0.3606705935659266, "grad_norm": 2.6320599636481554, "learning_rate": 9.647588717075775e-06, "loss": 0.1859, "step": 2388 }, { "epoch": 0.360821628152847, "grad_norm": 2.450349521891589, "learning_rate": 9.647297024161528e-06, "loss": 0.1691, "step": 2389 }, { "epoch": 0.3609726627397674, "grad_norm": 2.3485319082593903, "learning_rate": 9.64700521499259e-06, "loss": 0.1461, "step": 2390 }, { "epoch": 0.3611236973266878, "grad_norm": 2.0449615923300692, "learning_rate": 9.646713289576259e-06, "loss": 0.1515, "step": 2391 }, { "epoch": 0.3612747319136082, "grad_norm": 1.9739119008365775, "learning_rate": 9.646421247919839e-06, "loss": 0.1441, "step": 2392 }, { "epoch": 0.3614257665005286, "grad_norm": 2.506972915746566, "learning_rate": 9.646129090030636e-06, "loss": 0.2043, "step": 2393 }, { "epoch": 0.36157680108744905, "grad_norm": 2.8544384899845574, "learning_rate": 9.645836815915957e-06, "loss": 0.1814, "step": 2394 }, { "epoch": 0.36172783567436945, "grad_norm": 2.567836546728615, "learning_rate": 9.645544425583112e-06, "loss": 0.1745, "step": 2395 }, { "epoch": 0.36187887026128984, "grad_norm": 3.0011539574849695, "learning_rate": 9.64525191903942e-06, "loss": 0.1936, "step": 2396 }, { "epoch": 0.36202990484821024, "grad_norm": 2.565299612699341, "learning_rate": 9.644959296292194e-06, "loss": 0.1621, "step": 2397 }, { "epoch": 0.36218093943513063, "grad_norm": 2.336214981215356, "learning_rate": 9.644666557348755e-06, "loss": 0.1767, "step": 2398 }, { "epoch": 0.36233197402205103, "grad_norm": 2.3102710218913276, "learning_rate": 9.644373702216427e-06, "loss": 0.1482, "step": 2399 }, { "epoch": 0.3624830086089715, "grad_norm": 2.8165795191902143, "learning_rate": 9.644080730902536e-06, "loss": 0.1513, "step": 2400 }, { "epoch": 0.3626340431958919, "grad_norm": 2.553999518246005, "learning_rate": 9.643787643414411e-06, "loss": 0.1505, "step": 2401 }, { "epoch": 0.36278507778281227, "grad_norm": 1.9569241350656585, "learning_rate": 9.643494439759383e-06, "loss": 0.1477, "step": 2402 }, { "epoch": 0.36293611236973267, "grad_norm": 2.3795559711532563, "learning_rate": 9.643201119944785e-06, "loss": 0.1962, "step": 2403 }, { "epoch": 0.36308714695665306, "grad_norm": 2.851719926405392, "learning_rate": 9.642907683977958e-06, "loss": 0.197, "step": 2404 }, { "epoch": 0.36323818154357346, "grad_norm": 2.425274650090626, "learning_rate": 9.64261413186624e-06, "loss": 0.1674, "step": 2405 }, { "epoch": 0.3633892161304939, "grad_norm": 4.5029050031450915, "learning_rate": 9.642320463616974e-06, "loss": 0.1636, "step": 2406 }, { "epoch": 0.3635402507174143, "grad_norm": 4.992299066425689, "learning_rate": 9.642026679237509e-06, "loss": 0.2104, "step": 2407 }, { "epoch": 0.3636912853043347, "grad_norm": 2.447297794999584, "learning_rate": 9.641732778735194e-06, "loss": 0.1522, "step": 2408 }, { "epoch": 0.3638423198912551, "grad_norm": 2.32722444565933, "learning_rate": 9.641438762117377e-06, "loss": 0.1783, "step": 2409 }, { "epoch": 0.3639933544781755, "grad_norm": 2.7662151462958677, "learning_rate": 9.641144629391416e-06, "loss": 0.1725, "step": 2410 }, { "epoch": 0.3641443890650959, "grad_norm": 2.4544925774643724, "learning_rate": 9.640850380564669e-06, "loss": 0.1646, "step": 2411 }, { "epoch": 0.36429542365201634, "grad_norm": 2.2548440830074723, "learning_rate": 9.640556015644497e-06, "loss": 0.1737, "step": 2412 }, { "epoch": 0.36444645823893673, "grad_norm": 2.991399035987006, "learning_rate": 9.640261534638262e-06, "loss": 0.1957, "step": 2413 }, { "epoch": 0.36459749282585713, "grad_norm": 2.2951260958097106, "learning_rate": 9.639966937553333e-06, "loss": 0.1485, "step": 2414 }, { "epoch": 0.3647485274127775, "grad_norm": 2.3622364332213053, "learning_rate": 9.639672224397077e-06, "loss": 0.1509, "step": 2415 }, { "epoch": 0.3648995619996979, "grad_norm": 2.3007809544421622, "learning_rate": 9.639377395176867e-06, "loss": 0.152, "step": 2416 }, { "epoch": 0.3650505965866183, "grad_norm": 2.5413774521226906, "learning_rate": 9.639082449900081e-06, "loss": 0.1705, "step": 2417 }, { "epoch": 0.36520163117353877, "grad_norm": 4.611780317797153, "learning_rate": 9.638787388574094e-06, "loss": 0.2036, "step": 2418 }, { "epoch": 0.36535266576045916, "grad_norm": 3.4466378535401594, "learning_rate": 9.638492211206289e-06, "loss": 0.1805, "step": 2419 }, { "epoch": 0.36550370034737956, "grad_norm": 2.1263865108641853, "learning_rate": 9.638196917804049e-06, "loss": 0.1506, "step": 2420 }, { "epoch": 0.36565473493429995, "grad_norm": 5.108092678351778, "learning_rate": 9.63790150837476e-06, "loss": 0.1676, "step": 2421 }, { "epoch": 0.36580576952122035, "grad_norm": 2.359841894336533, "learning_rate": 9.637605982925814e-06, "loss": 0.1676, "step": 2422 }, { "epoch": 0.36595680410814074, "grad_norm": 2.2530736934181204, "learning_rate": 9.637310341464601e-06, "loss": 0.1647, "step": 2423 }, { "epoch": 0.3661078386950612, "grad_norm": 3.5008836387459503, "learning_rate": 9.63701458399852e-06, "loss": 0.1719, "step": 2424 }, { "epoch": 0.3662588732819816, "grad_norm": 2.3924508428962294, "learning_rate": 9.636718710534966e-06, "loss": 0.1748, "step": 2425 }, { "epoch": 0.366409907868902, "grad_norm": 2.2935020581494334, "learning_rate": 9.636422721081345e-06, "loss": 0.1564, "step": 2426 }, { "epoch": 0.3665609424558224, "grad_norm": 2.2192254491167147, "learning_rate": 9.636126615645058e-06, "loss": 0.1551, "step": 2427 }, { "epoch": 0.3667119770427428, "grad_norm": 2.5323815547333415, "learning_rate": 9.635830394233513e-06, "loss": 0.177, "step": 2428 }, { "epoch": 0.36686301162966317, "grad_norm": 2.7226464772322716, "learning_rate": 9.63553405685412e-06, "loss": 0.1691, "step": 2429 }, { "epoch": 0.3670140462165836, "grad_norm": 2.685659473773309, "learning_rate": 9.635237603514292e-06, "loss": 0.16, "step": 2430 }, { "epoch": 0.367165080803504, "grad_norm": 2.271046737445561, "learning_rate": 9.634941034221444e-06, "loss": 0.1577, "step": 2431 }, { "epoch": 0.3673161153904244, "grad_norm": 2.352510658791862, "learning_rate": 9.634644348982998e-06, "loss": 0.1695, "step": 2432 }, { "epoch": 0.3674671499773448, "grad_norm": 2.1618333879779708, "learning_rate": 9.63434754780637e-06, "loss": 0.1373, "step": 2433 }, { "epoch": 0.3676181845642652, "grad_norm": 3.0310817335833073, "learning_rate": 9.634050630698992e-06, "loss": 0.1623, "step": 2434 }, { "epoch": 0.3677692191511856, "grad_norm": 2.703959414419936, "learning_rate": 9.633753597668288e-06, "loss": 0.184, "step": 2435 }, { "epoch": 0.36792025373810605, "grad_norm": 2.4928023421740115, "learning_rate": 9.633456448721685e-06, "loss": 0.149, "step": 2436 }, { "epoch": 0.36807128832502645, "grad_norm": 1.9708587051833841, "learning_rate": 9.633159183866623e-06, "loss": 0.1316, "step": 2437 }, { "epoch": 0.36822232291194684, "grad_norm": 2.085322590978173, "learning_rate": 9.632861803110532e-06, "loss": 0.1672, "step": 2438 }, { "epoch": 0.36837335749886724, "grad_norm": 2.203657265365976, "learning_rate": 9.632564306460855e-06, "loss": 0.1507, "step": 2439 }, { "epoch": 0.36852439208578763, "grad_norm": 2.326796852172633, "learning_rate": 9.632266693925035e-06, "loss": 0.1792, "step": 2440 }, { "epoch": 0.36867542667270803, "grad_norm": 2.3812277352731646, "learning_rate": 9.631968965510511e-06, "loss": 0.1529, "step": 2441 }, { "epoch": 0.3688264612596285, "grad_norm": 2.3714195583002824, "learning_rate": 9.631671121224738e-06, "loss": 0.1631, "step": 2442 }, { "epoch": 0.3689774958465489, "grad_norm": 2.4618278854845688, "learning_rate": 9.631373161075163e-06, "loss": 0.173, "step": 2443 }, { "epoch": 0.36912853043346927, "grad_norm": 2.1830950171453383, "learning_rate": 9.63107508506924e-06, "loss": 0.1535, "step": 2444 }, { "epoch": 0.36927956502038967, "grad_norm": 2.471699263433678, "learning_rate": 9.630776893214426e-06, "loss": 0.1719, "step": 2445 }, { "epoch": 0.36943059960731006, "grad_norm": 2.6690463751877904, "learning_rate": 9.63047858551818e-06, "loss": 0.1923, "step": 2446 }, { "epoch": 0.36958163419423046, "grad_norm": 4.084388335365834, "learning_rate": 9.630180161987965e-06, "loss": 0.1868, "step": 2447 }, { "epoch": 0.3697326687811509, "grad_norm": 2.7188686190144646, "learning_rate": 9.629881622631245e-06, "loss": 0.1672, "step": 2448 }, { "epoch": 0.3698837033680713, "grad_norm": 2.2599535354752196, "learning_rate": 9.629582967455488e-06, "loss": 0.1564, "step": 2449 }, { "epoch": 0.3700347379549917, "grad_norm": 2.611453712817095, "learning_rate": 9.629284196468168e-06, "loss": 0.1698, "step": 2450 }, { "epoch": 0.3701857725419121, "grad_norm": 2.8936651368277504, "learning_rate": 9.628985309676754e-06, "loss": 0.1492, "step": 2451 }, { "epoch": 0.3703368071288325, "grad_norm": 2.734458153543529, "learning_rate": 9.628686307088729e-06, "loss": 0.1873, "step": 2452 }, { "epoch": 0.3704878417157529, "grad_norm": 2.4308774677796707, "learning_rate": 9.628387188711567e-06, "loss": 0.1565, "step": 2453 }, { "epoch": 0.37063887630267334, "grad_norm": 3.124642973043458, "learning_rate": 9.628087954552755e-06, "loss": 0.1695, "step": 2454 }, { "epoch": 0.37078991088959373, "grad_norm": 2.3523661603711976, "learning_rate": 9.627788604619773e-06, "loss": 0.1635, "step": 2455 }, { "epoch": 0.37094094547651413, "grad_norm": 3.160994196604337, "learning_rate": 9.627489138920116e-06, "loss": 0.1722, "step": 2456 }, { "epoch": 0.3710919800634345, "grad_norm": 2.499653660459174, "learning_rate": 9.627189557461271e-06, "loss": 0.1536, "step": 2457 }, { "epoch": 0.3712430146503549, "grad_norm": 2.997954381444046, "learning_rate": 9.626889860250733e-06, "loss": 0.1652, "step": 2458 }, { "epoch": 0.3713940492372753, "grad_norm": 5.494249292679732, "learning_rate": 9.626590047296e-06, "loss": 0.1767, "step": 2459 }, { "epoch": 0.37154508382419577, "grad_norm": 3.080521732770279, "learning_rate": 9.626290118604571e-06, "loss": 0.1779, "step": 2460 }, { "epoch": 0.37169611841111616, "grad_norm": 2.9358389999121184, "learning_rate": 9.62599007418395e-06, "loss": 0.1908, "step": 2461 }, { "epoch": 0.37184715299803656, "grad_norm": 2.246747066170693, "learning_rate": 9.625689914041642e-06, "loss": 0.1432, "step": 2462 }, { "epoch": 0.37199818758495695, "grad_norm": 3.0432066079770115, "learning_rate": 9.625389638185154e-06, "loss": 0.1741, "step": 2463 }, { "epoch": 0.37214922217187735, "grad_norm": 2.0650244370209667, "learning_rate": 9.625089246622001e-06, "loss": 0.1551, "step": 2464 }, { "epoch": 0.37230025675879774, "grad_norm": 2.7036944761811297, "learning_rate": 9.624788739359696e-06, "loss": 0.1788, "step": 2465 }, { "epoch": 0.3724512913457182, "grad_norm": 2.6492840050746578, "learning_rate": 9.624488116405755e-06, "loss": 0.1564, "step": 2466 }, { "epoch": 0.3726023259326386, "grad_norm": 3.1757106765627543, "learning_rate": 9.6241873777677e-06, "loss": 0.1834, "step": 2467 }, { "epoch": 0.372753360519559, "grad_norm": 2.4903471111086875, "learning_rate": 9.623886523453054e-06, "loss": 0.1684, "step": 2468 }, { "epoch": 0.3729043951064794, "grad_norm": 2.318439164118098, "learning_rate": 9.623585553469344e-06, "loss": 0.1557, "step": 2469 }, { "epoch": 0.3730554296933998, "grad_norm": 2.358440749100013, "learning_rate": 9.623284467824094e-06, "loss": 0.1768, "step": 2470 }, { "epoch": 0.37320646428032017, "grad_norm": 2.533828697719371, "learning_rate": 9.622983266524842e-06, "loss": 0.1699, "step": 2471 }, { "epoch": 0.3733574988672406, "grad_norm": 2.3330715289257697, "learning_rate": 9.622681949579118e-06, "loss": 0.1541, "step": 2472 }, { "epoch": 0.373508533454161, "grad_norm": 2.661351132101582, "learning_rate": 9.622380516994462e-06, "loss": 0.174, "step": 2473 }, { "epoch": 0.3736595680410814, "grad_norm": 2.387893307584853, "learning_rate": 9.622078968778416e-06, "loss": 0.1637, "step": 2474 }, { "epoch": 0.3738106026280018, "grad_norm": 1.8570343892013905, "learning_rate": 9.62177730493852e-06, "loss": 0.151, "step": 2475 }, { "epoch": 0.3739616372149222, "grad_norm": 2.510483599799123, "learning_rate": 9.621475525482323e-06, "loss": 0.1855, "step": 2476 }, { "epoch": 0.3741126718018426, "grad_norm": 2.2111318770736594, "learning_rate": 9.621173630417373e-06, "loss": 0.1514, "step": 2477 }, { "epoch": 0.37426370638876305, "grad_norm": 6.330385450159322, "learning_rate": 9.62087161975122e-06, "loss": 0.1822, "step": 2478 }, { "epoch": 0.37441474097568345, "grad_norm": 3.5145020806233256, "learning_rate": 9.620569493491423e-06, "loss": 0.1733, "step": 2479 }, { "epoch": 0.37456577556260384, "grad_norm": 2.782252520666591, "learning_rate": 9.620267251645536e-06, "loss": 0.1576, "step": 2480 }, { "epoch": 0.37471681014952424, "grad_norm": 2.3896587140131054, "learning_rate": 9.619964894221125e-06, "loss": 0.145, "step": 2481 }, { "epoch": 0.37486784473644463, "grad_norm": 2.1133119782135728, "learning_rate": 9.619662421225747e-06, "loss": 0.1577, "step": 2482 }, { "epoch": 0.37501887932336503, "grad_norm": 2.1656389698527736, "learning_rate": 9.619359832666973e-06, "loss": 0.1621, "step": 2483 }, { "epoch": 0.3751699139102855, "grad_norm": 2.0446580709335143, "learning_rate": 9.61905712855237e-06, "loss": 0.1366, "step": 2484 }, { "epoch": 0.3753209484972059, "grad_norm": 2.190708773503336, "learning_rate": 9.618754308889513e-06, "loss": 0.1408, "step": 2485 }, { "epoch": 0.37547198308412627, "grad_norm": 2.1151150852232243, "learning_rate": 9.618451373685978e-06, "loss": 0.153, "step": 2486 }, { "epoch": 0.37562301767104667, "grad_norm": 2.315742709991345, "learning_rate": 9.618148322949336e-06, "loss": 0.1375, "step": 2487 }, { "epoch": 0.37577405225796706, "grad_norm": 3.2052677217695535, "learning_rate": 9.617845156687176e-06, "loss": 0.165, "step": 2488 }, { "epoch": 0.37592508684488746, "grad_norm": 2.3543407829656013, "learning_rate": 9.617541874907078e-06, "loss": 0.1337, "step": 2489 }, { "epoch": 0.3760761214318079, "grad_norm": 2.4224626346218714, "learning_rate": 9.61723847761663e-06, "loss": 0.1829, "step": 2490 }, { "epoch": 0.3762271560187283, "grad_norm": 2.0623647587970577, "learning_rate": 9.61693496482342e-06, "loss": 0.1568, "step": 2491 }, { "epoch": 0.3763781906056487, "grad_norm": 2.754942657243111, "learning_rate": 9.616631336535042e-06, "loss": 0.1667, "step": 2492 }, { "epoch": 0.3765292251925691, "grad_norm": 2.206185810580343, "learning_rate": 9.616327592759091e-06, "loss": 0.1312, "step": 2493 }, { "epoch": 0.3766802597794895, "grad_norm": 2.192127899121956, "learning_rate": 9.616023733503165e-06, "loss": 0.16, "step": 2494 }, { "epoch": 0.3768312943664099, "grad_norm": 2.748331711733583, "learning_rate": 9.615719758774866e-06, "loss": 0.1554, "step": 2495 }, { "epoch": 0.37698232895333034, "grad_norm": 2.3133438987289927, "learning_rate": 9.615415668581798e-06, "loss": 0.1622, "step": 2496 }, { "epoch": 0.37713336354025073, "grad_norm": 2.245452541544213, "learning_rate": 9.615111462931568e-06, "loss": 0.1646, "step": 2497 }, { "epoch": 0.37728439812717113, "grad_norm": 3.025673823884799, "learning_rate": 9.614807141831783e-06, "loss": 0.1599, "step": 2498 }, { "epoch": 0.3774354327140915, "grad_norm": 2.3853973217071247, "learning_rate": 9.61450270529006e-06, "loss": 0.1794, "step": 2499 }, { "epoch": 0.3775864673010119, "grad_norm": 2.5527493362807907, "learning_rate": 9.614198153314014e-06, "loss": 0.1614, "step": 2500 }, { "epoch": 0.3777375018879323, "grad_norm": 3.382663757271757, "learning_rate": 9.613893485911262e-06, "loss": 0.1764, "step": 2501 }, { "epoch": 0.37788853647485277, "grad_norm": 3.373962129584339, "learning_rate": 9.613588703089426e-06, "loss": 0.1679, "step": 2502 }, { "epoch": 0.37803957106177316, "grad_norm": 2.8667833229996074, "learning_rate": 9.61328380485613e-06, "loss": 0.1959, "step": 2503 }, { "epoch": 0.37819060564869356, "grad_norm": 2.704079346798208, "learning_rate": 9.612978791219001e-06, "loss": 0.1645, "step": 2504 }, { "epoch": 0.37834164023561395, "grad_norm": 2.911284305347494, "learning_rate": 9.61267366218567e-06, "loss": 0.1842, "step": 2505 }, { "epoch": 0.37849267482253435, "grad_norm": 2.0858872944863096, "learning_rate": 9.612368417763769e-06, "loss": 0.1557, "step": 2506 }, { "epoch": 0.37864370940945474, "grad_norm": 3.548571567871613, "learning_rate": 9.612063057960934e-06, "loss": 0.1879, "step": 2507 }, { "epoch": 0.3787947439963752, "grad_norm": 2.002793578193465, "learning_rate": 9.611757582784805e-06, "loss": 0.1389, "step": 2508 }, { "epoch": 0.3789457785832956, "grad_norm": 2.0237121714859083, "learning_rate": 9.611451992243021e-06, "loss": 0.1413, "step": 2509 }, { "epoch": 0.379096813170216, "grad_norm": 2.1566712237293584, "learning_rate": 9.61114628634323e-06, "loss": 0.1617, "step": 2510 }, { "epoch": 0.3792478477571364, "grad_norm": 2.9720301844921497, "learning_rate": 9.610840465093076e-06, "loss": 0.1694, "step": 2511 }, { "epoch": 0.3793988823440568, "grad_norm": 2.585165027809346, "learning_rate": 9.610534528500212e-06, "loss": 0.1629, "step": 2512 }, { "epoch": 0.37954991693097717, "grad_norm": 1.8292106124637104, "learning_rate": 9.61022847657229e-06, "loss": 0.1174, "step": 2513 }, { "epoch": 0.3797009515178976, "grad_norm": 2.7142873018041125, "learning_rate": 9.609922309316965e-06, "loss": 0.1844, "step": 2514 }, { "epoch": 0.379851986104818, "grad_norm": 2.8333656842467443, "learning_rate": 9.6096160267419e-06, "loss": 0.2001, "step": 2515 }, { "epoch": 0.3800030206917384, "grad_norm": 1.9262357158831855, "learning_rate": 9.609309628854752e-06, "loss": 0.1403, "step": 2516 }, { "epoch": 0.3801540552786588, "grad_norm": 2.337136205651524, "learning_rate": 9.609003115663186e-06, "loss": 0.1576, "step": 2517 }, { "epoch": 0.3803050898655792, "grad_norm": 2.666033980927666, "learning_rate": 9.608696487174875e-06, "loss": 0.1569, "step": 2518 }, { "epoch": 0.3804561244524996, "grad_norm": 2.2264098190415615, "learning_rate": 9.608389743397483e-06, "loss": 0.161, "step": 2519 }, { "epoch": 0.38060715903942005, "grad_norm": 2.462557909968665, "learning_rate": 9.608082884338687e-06, "loss": 0.1513, "step": 2520 }, { "epoch": 0.38075819362634045, "grad_norm": 2.3168188223789046, "learning_rate": 9.60777591000616e-06, "loss": 0.1586, "step": 2521 }, { "epoch": 0.38090922821326084, "grad_norm": 2.2645488046540763, "learning_rate": 9.607468820407586e-06, "loss": 0.1531, "step": 2522 }, { "epoch": 0.38106026280018124, "grad_norm": 3.199747141874504, "learning_rate": 9.607161615550644e-06, "loss": 0.1931, "step": 2523 }, { "epoch": 0.38121129738710163, "grad_norm": 2.3679996096054787, "learning_rate": 9.60685429544302e-06, "loss": 0.1395, "step": 2524 }, { "epoch": 0.38136233197402203, "grad_norm": 2.8383967556726506, "learning_rate": 9.606546860092399e-06, "loss": 0.1551, "step": 2525 }, { "epoch": 0.3815133665609425, "grad_norm": 2.4745186448392658, "learning_rate": 9.606239309506476e-06, "loss": 0.16, "step": 2526 }, { "epoch": 0.3816644011478629, "grad_norm": 4.312882127155532, "learning_rate": 9.605931643692941e-06, "loss": 0.1695, "step": 2527 }, { "epoch": 0.38181543573478327, "grad_norm": 3.3948438117543356, "learning_rate": 9.605623862659492e-06, "loss": 0.197, "step": 2528 }, { "epoch": 0.38196647032170367, "grad_norm": 2.7291661662192825, "learning_rate": 9.605315966413829e-06, "loss": 0.1818, "step": 2529 }, { "epoch": 0.38211750490862406, "grad_norm": 2.2203992064183566, "learning_rate": 9.605007954963652e-06, "loss": 0.1744, "step": 2530 }, { "epoch": 0.38226853949554446, "grad_norm": 2.5115747290262775, "learning_rate": 9.604699828316667e-06, "loss": 0.1733, "step": 2531 }, { "epoch": 0.3824195740824649, "grad_norm": 2.384634531405125, "learning_rate": 9.604391586480584e-06, "loss": 0.1643, "step": 2532 }, { "epoch": 0.3825706086693853, "grad_norm": 2.345761843879842, "learning_rate": 9.604083229463111e-06, "loss": 0.1671, "step": 2533 }, { "epoch": 0.3827216432563057, "grad_norm": 2.6460486191269594, "learning_rate": 9.603774757271961e-06, "loss": 0.1607, "step": 2534 }, { "epoch": 0.3828726778432261, "grad_norm": 3.193093061417217, "learning_rate": 9.603466169914857e-06, "loss": 0.1849, "step": 2535 }, { "epoch": 0.3830237124301465, "grad_norm": 2.0726393532437886, "learning_rate": 9.603157467399511e-06, "loss": 0.1346, "step": 2536 }, { "epoch": 0.3831747470170669, "grad_norm": 2.2273808734348397, "learning_rate": 9.602848649733648e-06, "loss": 0.159, "step": 2537 }, { "epoch": 0.38332578160398734, "grad_norm": 2.712095552562746, "learning_rate": 9.602539716924995e-06, "loss": 0.1494, "step": 2538 }, { "epoch": 0.38347681619090773, "grad_norm": 2.8437378978776566, "learning_rate": 9.602230668981278e-06, "loss": 0.1771, "step": 2539 }, { "epoch": 0.3836278507778281, "grad_norm": 2.732368627164089, "learning_rate": 9.601921505910229e-06, "loss": 0.1601, "step": 2540 }, { "epoch": 0.3837788853647485, "grad_norm": 1.9218124600130937, "learning_rate": 9.60161222771958e-06, "loss": 0.1394, "step": 2541 }, { "epoch": 0.3839299199516689, "grad_norm": 2.649041058040757, "learning_rate": 9.601302834417072e-06, "loss": 0.1945, "step": 2542 }, { "epoch": 0.3840809545385893, "grad_norm": 2.8657383488970223, "learning_rate": 9.60099332601044e-06, "loss": 0.1819, "step": 2543 }, { "epoch": 0.38423198912550977, "grad_norm": 2.7854433586521754, "learning_rate": 9.60068370250743e-06, "loss": 0.149, "step": 2544 }, { "epoch": 0.38438302371243016, "grad_norm": 2.4118516035711357, "learning_rate": 9.600373963915784e-06, "loss": 0.1702, "step": 2545 }, { "epoch": 0.38453405829935056, "grad_norm": 2.7619295718531127, "learning_rate": 9.600064110243254e-06, "loss": 0.1656, "step": 2546 }, { "epoch": 0.38468509288627095, "grad_norm": 2.6756973321938173, "learning_rate": 9.599754141497587e-06, "loss": 0.1667, "step": 2547 }, { "epoch": 0.38483612747319135, "grad_norm": 3.2133064648874905, "learning_rate": 9.599444057686542e-06, "loss": 0.1531, "step": 2548 }, { "epoch": 0.38498716206011174, "grad_norm": 2.097996449928792, "learning_rate": 9.599133858817872e-06, "loss": 0.1566, "step": 2549 }, { "epoch": 0.3851381966470322, "grad_norm": 1.9295660885126094, "learning_rate": 9.598823544899339e-06, "loss": 0.1351, "step": 2550 }, { "epoch": 0.3852892312339526, "grad_norm": 2.581911296909144, "learning_rate": 9.598513115938704e-06, "loss": 0.175, "step": 2551 }, { "epoch": 0.385440265820873, "grad_norm": 2.708941053030446, "learning_rate": 9.598202571943734e-06, "loss": 0.1763, "step": 2552 }, { "epoch": 0.3855913004077934, "grad_norm": 2.2470181931520843, "learning_rate": 9.597891912922196e-06, "loss": 0.1432, "step": 2553 }, { "epoch": 0.3857423349947138, "grad_norm": 2.1977670593065994, "learning_rate": 9.597581138881862e-06, "loss": 0.151, "step": 2554 }, { "epoch": 0.38589336958163417, "grad_norm": 2.296796527133286, "learning_rate": 9.597270249830506e-06, "loss": 0.1806, "step": 2555 }, { "epoch": 0.3860444041685546, "grad_norm": 2.7100295748660956, "learning_rate": 9.596959245775906e-06, "loss": 0.1669, "step": 2556 }, { "epoch": 0.386195438755475, "grad_norm": 2.344358346976762, "learning_rate": 9.596648126725841e-06, "loss": 0.176, "step": 2557 }, { "epoch": 0.3863464733423954, "grad_norm": 2.9166639982271043, "learning_rate": 9.596336892688093e-06, "loss": 0.1797, "step": 2558 }, { "epoch": 0.3864975079293158, "grad_norm": 2.0585463329057223, "learning_rate": 9.59602554367045e-06, "loss": 0.1541, "step": 2559 }, { "epoch": 0.3866485425162362, "grad_norm": 3.3363880088549154, "learning_rate": 9.595714079680698e-06, "loss": 0.1951, "step": 2560 }, { "epoch": 0.3867995771031566, "grad_norm": 2.3564645120425847, "learning_rate": 9.595402500726632e-06, "loss": 0.1488, "step": 2561 }, { "epoch": 0.38695061169007705, "grad_norm": 4.084607843273676, "learning_rate": 9.595090806816043e-06, "loss": 0.199, "step": 2562 }, { "epoch": 0.38710164627699745, "grad_norm": 2.601186742104694, "learning_rate": 9.594778997956729e-06, "loss": 0.1707, "step": 2563 }, { "epoch": 0.38725268086391784, "grad_norm": 2.1067164352103087, "learning_rate": 9.59446707415649e-06, "loss": 0.1465, "step": 2564 }, { "epoch": 0.38740371545083824, "grad_norm": 1.8933905667503486, "learning_rate": 9.59415503542313e-06, "loss": 0.1377, "step": 2565 }, { "epoch": 0.38755475003775863, "grad_norm": 2.0903793502910792, "learning_rate": 9.593842881764454e-06, "loss": 0.1527, "step": 2566 }, { "epoch": 0.38770578462467903, "grad_norm": 2.512357362294946, "learning_rate": 9.59353061318827e-06, "loss": 0.1503, "step": 2567 }, { "epoch": 0.3878568192115995, "grad_norm": 2.220377122706177, "learning_rate": 9.593218229702392e-06, "loss": 0.155, "step": 2568 }, { "epoch": 0.3880078537985199, "grad_norm": 2.4393742848202202, "learning_rate": 9.592905731314631e-06, "loss": 0.1638, "step": 2569 }, { "epoch": 0.38815888838544027, "grad_norm": 2.6208640461182875, "learning_rate": 9.592593118032807e-06, "loss": 0.1943, "step": 2570 }, { "epoch": 0.38830992297236067, "grad_norm": 2.841743029535938, "learning_rate": 9.592280389864739e-06, "loss": 0.1892, "step": 2571 }, { "epoch": 0.38846095755928106, "grad_norm": 2.2244069917069864, "learning_rate": 9.59196754681825e-06, "loss": 0.1519, "step": 2572 }, { "epoch": 0.38861199214620146, "grad_norm": 2.7154473653915936, "learning_rate": 9.591654588901167e-06, "loss": 0.1799, "step": 2573 }, { "epoch": 0.3887630267331219, "grad_norm": 2.3131858045093785, "learning_rate": 9.591341516121318e-06, "loss": 0.1738, "step": 2574 }, { "epoch": 0.3889140613200423, "grad_norm": 1.8797786395990548, "learning_rate": 9.591028328486536e-06, "loss": 0.1433, "step": 2575 }, { "epoch": 0.3890650959069627, "grad_norm": 2.225775683704239, "learning_rate": 9.590715026004654e-06, "loss": 0.1555, "step": 2576 }, { "epoch": 0.3892161304938831, "grad_norm": 2.8040985814052513, "learning_rate": 9.590401608683509e-06, "loss": 0.184, "step": 2577 }, { "epoch": 0.3893671650808035, "grad_norm": 2.418441507385423, "learning_rate": 9.590088076530942e-06, "loss": 0.158, "step": 2578 }, { "epoch": 0.3895181996677239, "grad_norm": 2.0187550484713577, "learning_rate": 9.589774429554797e-06, "loss": 0.1591, "step": 2579 }, { "epoch": 0.38966923425464434, "grad_norm": 2.6456145887781326, "learning_rate": 9.589460667762917e-06, "loss": 0.1728, "step": 2580 }, { "epoch": 0.38982026884156473, "grad_norm": 2.7088030278587576, "learning_rate": 9.589146791163156e-06, "loss": 0.1963, "step": 2581 }, { "epoch": 0.3899713034284851, "grad_norm": 2.38289593701632, "learning_rate": 9.588832799763364e-06, "loss": 0.1476, "step": 2582 }, { "epoch": 0.3901223380154055, "grad_norm": 2.0933682236161504, "learning_rate": 9.588518693571391e-06, "loss": 0.1475, "step": 2583 }, { "epoch": 0.3902733726023259, "grad_norm": 2.1730226567920314, "learning_rate": 9.588204472595103e-06, "loss": 0.1639, "step": 2584 }, { "epoch": 0.3904244071892463, "grad_norm": 1.9282851116665038, "learning_rate": 9.587890136842351e-06, "loss": 0.1469, "step": 2585 }, { "epoch": 0.39057544177616677, "grad_norm": 2.606213698854238, "learning_rate": 9.587575686321006e-06, "loss": 0.1516, "step": 2586 }, { "epoch": 0.39072647636308716, "grad_norm": 3.0424972342955243, "learning_rate": 9.58726112103893e-06, "loss": 0.2002, "step": 2587 }, { "epoch": 0.39087751095000756, "grad_norm": 2.5842949695666007, "learning_rate": 9.586946441003993e-06, "loss": 0.1659, "step": 2588 }, { "epoch": 0.39102854553692795, "grad_norm": 2.2002240911291424, "learning_rate": 9.586631646224067e-06, "loss": 0.1464, "step": 2589 }, { "epoch": 0.39117958012384835, "grad_norm": 3.3854576377181105, "learning_rate": 9.586316736707028e-06, "loss": 0.1545, "step": 2590 }, { "epoch": 0.39133061471076874, "grad_norm": 3.059898421260511, "learning_rate": 9.58600171246075e-06, "loss": 0.1888, "step": 2591 }, { "epoch": 0.3914816492976892, "grad_norm": 2.657195742009482, "learning_rate": 9.585686573493119e-06, "loss": 0.182, "step": 2592 }, { "epoch": 0.3916326838846096, "grad_norm": 2.2387239574776725, "learning_rate": 9.585371319812013e-06, "loss": 0.1725, "step": 2593 }, { "epoch": 0.39178371847153, "grad_norm": 2.388696928420016, "learning_rate": 9.585055951425322e-06, "loss": 0.1656, "step": 2594 }, { "epoch": 0.3919347530584504, "grad_norm": 1.981032234233877, "learning_rate": 9.584740468340935e-06, "loss": 0.1441, "step": 2595 }, { "epoch": 0.3920857876453708, "grad_norm": 2.116448973617012, "learning_rate": 9.58442487056674e-06, "loss": 0.1528, "step": 2596 }, { "epoch": 0.39223682223229117, "grad_norm": 2.8761009663024404, "learning_rate": 9.584109158110634e-06, "loss": 0.1769, "step": 2597 }, { "epoch": 0.3923878568192116, "grad_norm": 4.622639401234404, "learning_rate": 9.583793330980518e-06, "loss": 0.2401, "step": 2598 }, { "epoch": 0.392538891406132, "grad_norm": 3.1913382690350756, "learning_rate": 9.583477389184288e-06, "loss": 0.1685, "step": 2599 }, { "epoch": 0.3926899259930524, "grad_norm": 2.2535109308103327, "learning_rate": 9.583161332729851e-06, "loss": 0.1518, "step": 2600 }, { "epoch": 0.3928409605799728, "grad_norm": 2.119850218519153, "learning_rate": 9.58284516162511e-06, "loss": 0.1548, "step": 2601 }, { "epoch": 0.3929919951668932, "grad_norm": 3.3477812485812923, "learning_rate": 9.582528875877977e-06, "loss": 0.1708, "step": 2602 }, { "epoch": 0.3931430297538136, "grad_norm": 2.7252709251825453, "learning_rate": 9.582212475496362e-06, "loss": 0.1684, "step": 2603 }, { "epoch": 0.39329406434073405, "grad_norm": 1.875407325868065, "learning_rate": 9.581895960488181e-06, "loss": 0.124, "step": 2604 }, { "epoch": 0.39344509892765445, "grad_norm": 2.4077640036118653, "learning_rate": 9.58157933086135e-06, "loss": 0.1883, "step": 2605 }, { "epoch": 0.39359613351457484, "grad_norm": 3.011425586820055, "learning_rate": 9.581262586623795e-06, "loss": 0.1639, "step": 2606 }, { "epoch": 0.39374716810149524, "grad_norm": 2.525412195709707, "learning_rate": 9.580945727783432e-06, "loss": 0.1851, "step": 2607 }, { "epoch": 0.39389820268841563, "grad_norm": 2.0989005843340043, "learning_rate": 9.580628754348194e-06, "loss": 0.1586, "step": 2608 }, { "epoch": 0.39404923727533603, "grad_norm": 2.36491967447477, "learning_rate": 9.580311666326005e-06, "loss": 0.1803, "step": 2609 }, { "epoch": 0.3942002718622565, "grad_norm": 3.035383108458528, "learning_rate": 9.579994463724802e-06, "loss": 0.1825, "step": 2610 }, { "epoch": 0.3943513064491769, "grad_norm": 2.569374816174548, "learning_rate": 9.579677146552517e-06, "loss": 0.1758, "step": 2611 }, { "epoch": 0.39450234103609727, "grad_norm": 2.329024957887069, "learning_rate": 9.579359714817087e-06, "loss": 0.1617, "step": 2612 }, { "epoch": 0.39465337562301767, "grad_norm": 4.136337459239983, "learning_rate": 9.579042168526455e-06, "loss": 0.1815, "step": 2613 }, { "epoch": 0.39480441020993806, "grad_norm": 2.9023859783559183, "learning_rate": 9.578724507688564e-06, "loss": 0.1869, "step": 2614 }, { "epoch": 0.39495544479685846, "grad_norm": 2.061903816172431, "learning_rate": 9.57840673231136e-06, "loss": 0.1502, "step": 2615 }, { "epoch": 0.3951064793837789, "grad_norm": 6.1659317025537845, "learning_rate": 9.578088842402794e-06, "loss": 0.1869, "step": 2616 }, { "epoch": 0.3952575139706993, "grad_norm": 2.6513710547731444, "learning_rate": 9.577770837970814e-06, "loss": 0.1596, "step": 2617 }, { "epoch": 0.3954085485576197, "grad_norm": 2.3389896473702287, "learning_rate": 9.577452719023378e-06, "loss": 0.1583, "step": 2618 }, { "epoch": 0.3955595831445401, "grad_norm": 2.297518771925311, "learning_rate": 9.577134485568447e-06, "loss": 0.1648, "step": 2619 }, { "epoch": 0.3957106177314605, "grad_norm": 5.086063520384343, "learning_rate": 9.576816137613975e-06, "loss": 0.188, "step": 2620 }, { "epoch": 0.3958616523183809, "grad_norm": 2.890237456230777, "learning_rate": 9.576497675167933e-06, "loss": 0.1883, "step": 2621 }, { "epoch": 0.39601268690530134, "grad_norm": 3.2349191991071646, "learning_rate": 9.57617909823828e-06, "loss": 0.1782, "step": 2622 }, { "epoch": 0.39616372149222173, "grad_norm": 2.4303785897043193, "learning_rate": 9.57586040683299e-06, "loss": 0.1545, "step": 2623 }, { "epoch": 0.3963147560791421, "grad_norm": 2.80890382805801, "learning_rate": 9.575541600960036e-06, "loss": 0.1609, "step": 2624 }, { "epoch": 0.3964657906660625, "grad_norm": 2.117126445462171, "learning_rate": 9.57522268062739e-06, "loss": 0.1358, "step": 2625 }, { "epoch": 0.3966168252529829, "grad_norm": 2.753333277103633, "learning_rate": 9.57490364584303e-06, "loss": 0.1933, "step": 2626 }, { "epoch": 0.3967678598399033, "grad_norm": 2.041926247374833, "learning_rate": 9.57458449661494e-06, "loss": 0.155, "step": 2627 }, { "epoch": 0.39691889442682377, "grad_norm": 2.221494982999489, "learning_rate": 9.574265232951103e-06, "loss": 0.1702, "step": 2628 }, { "epoch": 0.39706992901374416, "grad_norm": 2.8952550278601414, "learning_rate": 9.573945854859503e-06, "loss": 0.1579, "step": 2629 }, { "epoch": 0.39722096360066456, "grad_norm": 1.9042184387652392, "learning_rate": 9.57362636234813e-06, "loss": 0.1388, "step": 2630 }, { "epoch": 0.39737199818758495, "grad_norm": 4.785706160134142, "learning_rate": 9.573306755424979e-06, "loss": 0.1541, "step": 2631 }, { "epoch": 0.39752303277450535, "grad_norm": 2.7033053882924754, "learning_rate": 9.572987034098042e-06, "loss": 0.1934, "step": 2632 }, { "epoch": 0.39767406736142574, "grad_norm": 2.2930030126065577, "learning_rate": 9.572667198375318e-06, "loss": 0.1391, "step": 2633 }, { "epoch": 0.3978251019483462, "grad_norm": 2.294566959212646, "learning_rate": 9.57234724826481e-06, "loss": 0.1628, "step": 2634 }, { "epoch": 0.3979761365352666, "grad_norm": 2.4632736824566366, "learning_rate": 9.572027183774518e-06, "loss": 0.1496, "step": 2635 }, { "epoch": 0.398127171122187, "grad_norm": 1.7413534584064079, "learning_rate": 9.571707004912452e-06, "loss": 0.1291, "step": 2636 }, { "epoch": 0.3982782057091074, "grad_norm": 2.2470332378850397, "learning_rate": 9.57138671168662e-06, "loss": 0.1653, "step": 2637 }, { "epoch": 0.3984292402960278, "grad_norm": 1.8802093255272523, "learning_rate": 9.571066304105032e-06, "loss": 0.134, "step": 2638 }, { "epoch": 0.39858027488294817, "grad_norm": 2.39944715599045, "learning_rate": 9.570745782175708e-06, "loss": 0.1662, "step": 2639 }, { "epoch": 0.3987313094698686, "grad_norm": 2.7558505031110396, "learning_rate": 9.57042514590666e-06, "loss": 0.1769, "step": 2640 }, { "epoch": 0.398882344056789, "grad_norm": 2.504645686500055, "learning_rate": 9.570104395305915e-06, "loss": 0.1688, "step": 2641 }, { "epoch": 0.3990333786437094, "grad_norm": 3.1694396412644816, "learning_rate": 9.569783530381494e-06, "loss": 0.1851, "step": 2642 }, { "epoch": 0.3991844132306298, "grad_norm": 2.5413642838057844, "learning_rate": 9.569462551141425e-06, "loss": 0.1691, "step": 2643 }, { "epoch": 0.3993354478175502, "grad_norm": 2.1603149482145896, "learning_rate": 9.569141457593734e-06, "loss": 0.139, "step": 2644 }, { "epoch": 0.3994864824044706, "grad_norm": 2.069129760279608, "learning_rate": 9.568820249746455e-06, "loss": 0.1489, "step": 2645 }, { "epoch": 0.39963751699139105, "grad_norm": 2.022984613915903, "learning_rate": 9.568498927607625e-06, "loss": 0.1608, "step": 2646 }, { "epoch": 0.39978855157831145, "grad_norm": 2.245426491445231, "learning_rate": 9.568177491185282e-06, "loss": 0.174, "step": 2647 }, { "epoch": 0.39993958616523184, "grad_norm": 2.361016772958071, "learning_rate": 9.567855940487463e-06, "loss": 0.1591, "step": 2648 }, { "epoch": 0.40009062075215224, "grad_norm": 2.7105944585012858, "learning_rate": 9.567534275522216e-06, "loss": 0.1776, "step": 2649 }, { "epoch": 0.40024165533907263, "grad_norm": 2.535007374280877, "learning_rate": 9.567212496297585e-06, "loss": 0.1687, "step": 2650 }, { "epoch": 0.40039268992599303, "grad_norm": 2.970872296602646, "learning_rate": 9.566890602821621e-06, "loss": 0.2141, "step": 2651 }, { "epoch": 0.4005437245129135, "grad_norm": 3.1264820661549844, "learning_rate": 9.566568595102377e-06, "loss": 0.1829, "step": 2652 }, { "epoch": 0.4006947590998339, "grad_norm": 2.5746317990254894, "learning_rate": 9.566246473147906e-06, "loss": 0.1786, "step": 2653 }, { "epoch": 0.40084579368675427, "grad_norm": 2.8927011289257196, "learning_rate": 9.565924236966268e-06, "loss": 0.1594, "step": 2654 }, { "epoch": 0.40099682827367467, "grad_norm": 2.4306993796917773, "learning_rate": 9.565601886565521e-06, "loss": 0.177, "step": 2655 }, { "epoch": 0.40114786286059506, "grad_norm": 2.009466645839368, "learning_rate": 9.565279421953732e-06, "loss": 0.1504, "step": 2656 }, { "epoch": 0.40129889744751546, "grad_norm": 2.6468590899004867, "learning_rate": 9.564956843138968e-06, "loss": 0.194, "step": 2657 }, { "epoch": 0.4014499320344359, "grad_norm": 2.9722789727595047, "learning_rate": 9.564634150129295e-06, "loss": 0.1703, "step": 2658 }, { "epoch": 0.4016009666213563, "grad_norm": 2.5226458541552623, "learning_rate": 9.564311342932789e-06, "loss": 0.1809, "step": 2659 }, { "epoch": 0.4017520012082767, "grad_norm": 2.8483990749598282, "learning_rate": 9.563988421557523e-06, "loss": 0.1622, "step": 2660 }, { "epoch": 0.4019030357951971, "grad_norm": 2.3557662537329334, "learning_rate": 9.563665386011573e-06, "loss": 0.1501, "step": 2661 }, { "epoch": 0.4020540703821175, "grad_norm": 2.651556376911863, "learning_rate": 9.563342236303024e-06, "loss": 0.175, "step": 2662 }, { "epoch": 0.4022051049690379, "grad_norm": 2.705971151385237, "learning_rate": 9.563018972439959e-06, "loss": 0.1674, "step": 2663 }, { "epoch": 0.40235613955595834, "grad_norm": 2.447657243521863, "learning_rate": 9.562695594430464e-06, "loss": 0.172, "step": 2664 }, { "epoch": 0.40250717414287873, "grad_norm": 2.3920431576464103, "learning_rate": 9.562372102282629e-06, "loss": 0.1717, "step": 2665 }, { "epoch": 0.4026582087297991, "grad_norm": 2.1701617641324105, "learning_rate": 9.562048496004544e-06, "loss": 0.1692, "step": 2666 }, { "epoch": 0.4028092433167195, "grad_norm": 2.9394937488257695, "learning_rate": 9.561724775604306e-06, "loss": 0.1541, "step": 2667 }, { "epoch": 0.4029602779036399, "grad_norm": 2.5783055659083747, "learning_rate": 9.561400941090013e-06, "loss": 0.1665, "step": 2668 }, { "epoch": 0.4031113124905603, "grad_norm": 2.121199751694028, "learning_rate": 9.561076992469766e-06, "loss": 0.154, "step": 2669 }, { "epoch": 0.40326234707748077, "grad_norm": 3.1700730022289005, "learning_rate": 9.56075292975167e-06, "loss": 0.1572, "step": 2670 }, { "epoch": 0.40341338166440116, "grad_norm": 2.854552795670787, "learning_rate": 9.560428752943827e-06, "loss": 0.1629, "step": 2671 }, { "epoch": 0.40356441625132156, "grad_norm": 2.90665176627692, "learning_rate": 9.560104462054351e-06, "loss": 0.1778, "step": 2672 }, { "epoch": 0.40371545083824195, "grad_norm": 2.779344108235319, "learning_rate": 9.559780057091354e-06, "loss": 0.161, "step": 2673 }, { "epoch": 0.40386648542516235, "grad_norm": 2.29458704852372, "learning_rate": 9.559455538062949e-06, "loss": 0.1748, "step": 2674 }, { "epoch": 0.40401752001208274, "grad_norm": 2.3037952062933225, "learning_rate": 9.559130904977257e-06, "loss": 0.159, "step": 2675 }, { "epoch": 0.4041685545990032, "grad_norm": 2.0754634695354026, "learning_rate": 9.558806157842393e-06, "loss": 0.1648, "step": 2676 }, { "epoch": 0.4043195891859236, "grad_norm": 2.4840056966231723, "learning_rate": 9.558481296666488e-06, "loss": 0.1683, "step": 2677 }, { "epoch": 0.404470623772844, "grad_norm": 2.141469714197182, "learning_rate": 9.558156321457664e-06, "loss": 0.1404, "step": 2678 }, { "epoch": 0.4046216583597644, "grad_norm": 2.2395837541990034, "learning_rate": 9.557831232224054e-06, "loss": 0.1441, "step": 2679 }, { "epoch": 0.4047726929466848, "grad_norm": 2.377471418698027, "learning_rate": 9.557506028973784e-06, "loss": 0.155, "step": 2680 }, { "epoch": 0.40492372753360517, "grad_norm": 2.4639645494959646, "learning_rate": 9.557180711714996e-06, "loss": 0.1552, "step": 2681 }, { "epoch": 0.4050747621205256, "grad_norm": 2.376698921683739, "learning_rate": 9.556855280455824e-06, "loss": 0.1581, "step": 2682 }, { "epoch": 0.405225796707446, "grad_norm": 2.3242744039267706, "learning_rate": 9.556529735204412e-06, "loss": 0.1751, "step": 2683 }, { "epoch": 0.4053768312943664, "grad_norm": 3.143056242832101, "learning_rate": 9.5562040759689e-06, "loss": 0.1715, "step": 2684 }, { "epoch": 0.4055278658812868, "grad_norm": 2.7849253739002173, "learning_rate": 9.555878302757435e-06, "loss": 0.1657, "step": 2685 }, { "epoch": 0.4056789004682072, "grad_norm": 2.84780961844127, "learning_rate": 9.55555241557817e-06, "loss": 0.1696, "step": 2686 }, { "epoch": 0.4058299350551276, "grad_norm": 2.538503900083751, "learning_rate": 9.555226414439254e-06, "loss": 0.1812, "step": 2687 }, { "epoch": 0.40598096964204805, "grad_norm": 2.2850823219055787, "learning_rate": 9.554900299348842e-06, "loss": 0.1583, "step": 2688 }, { "epoch": 0.40613200422896845, "grad_norm": 2.637290918303119, "learning_rate": 9.554574070315094e-06, "loss": 0.1678, "step": 2689 }, { "epoch": 0.40628303881588884, "grad_norm": 1.9693446069571687, "learning_rate": 9.55424772734617e-06, "loss": 0.1284, "step": 2690 }, { "epoch": 0.40643407340280924, "grad_norm": 2.30687360403661, "learning_rate": 9.553921270450233e-06, "loss": 0.1746, "step": 2691 }, { "epoch": 0.40658510798972963, "grad_norm": 2.348180662786458, "learning_rate": 9.553594699635448e-06, "loss": 0.1559, "step": 2692 }, { "epoch": 0.40673614257665003, "grad_norm": 2.358935696273506, "learning_rate": 9.55326801490999e-06, "loss": 0.1638, "step": 2693 }, { "epoch": 0.4068871771635705, "grad_norm": 2.9577840221951766, "learning_rate": 9.552941216282026e-06, "loss": 0.189, "step": 2694 }, { "epoch": 0.4070382117504909, "grad_norm": 3.050205339889225, "learning_rate": 9.55261430375973e-06, "loss": 0.1425, "step": 2695 }, { "epoch": 0.40718924633741127, "grad_norm": 2.9401724870131978, "learning_rate": 9.552287277351284e-06, "loss": 0.1825, "step": 2696 }, { "epoch": 0.40734028092433167, "grad_norm": 2.443103686406497, "learning_rate": 9.551960137064869e-06, "loss": 0.1907, "step": 2697 }, { "epoch": 0.40749131551125206, "grad_norm": 2.4314592844460203, "learning_rate": 9.551632882908665e-06, "loss": 0.1405, "step": 2698 }, { "epoch": 0.40764235009817246, "grad_norm": 2.2305491286557704, "learning_rate": 9.551305514890859e-06, "loss": 0.168, "step": 2699 }, { "epoch": 0.4077933846850929, "grad_norm": 2.4915468806347936, "learning_rate": 9.550978033019643e-06, "loss": 0.1944, "step": 2700 }, { "epoch": 0.4079444192720133, "grad_norm": 3.4435544412054164, "learning_rate": 9.550650437303207e-06, "loss": 0.1745, "step": 2701 }, { "epoch": 0.4080954538589337, "grad_norm": 2.0903932104982785, "learning_rate": 9.550322727749748e-06, "loss": 0.1482, "step": 2702 }, { "epoch": 0.4082464884458541, "grad_norm": 2.17135575062928, "learning_rate": 9.54999490436746e-06, "loss": 0.1674, "step": 2703 }, { "epoch": 0.4083975230327745, "grad_norm": 2.2801150870560476, "learning_rate": 9.549666967164548e-06, "loss": 0.1356, "step": 2704 }, { "epoch": 0.4085485576196949, "grad_norm": 1.9976250427617814, "learning_rate": 9.549338916149212e-06, "loss": 0.1538, "step": 2705 }, { "epoch": 0.40869959220661534, "grad_norm": 2.8230689191129525, "learning_rate": 9.549010751329661e-06, "loss": 0.1791, "step": 2706 }, { "epoch": 0.40885062679353573, "grad_norm": 2.6192652761385635, "learning_rate": 9.548682472714102e-06, "loss": 0.1604, "step": 2707 }, { "epoch": 0.4090016613804561, "grad_norm": 2.3827665954132486, "learning_rate": 9.54835408031075e-06, "loss": 0.1769, "step": 2708 }, { "epoch": 0.4091526959673765, "grad_norm": 2.400118493715247, "learning_rate": 9.548025574127818e-06, "loss": 0.1725, "step": 2709 }, { "epoch": 0.4093037305542969, "grad_norm": 2.756338752296965, "learning_rate": 9.547696954173524e-06, "loss": 0.1564, "step": 2710 }, { "epoch": 0.4094547651412173, "grad_norm": 2.7664220818982375, "learning_rate": 9.547368220456086e-06, "loss": 0.1931, "step": 2711 }, { "epoch": 0.40960579972813776, "grad_norm": 2.834979195703777, "learning_rate": 9.547039372983733e-06, "loss": 0.172, "step": 2712 }, { "epoch": 0.40975683431505816, "grad_norm": 3.0133587302816016, "learning_rate": 9.546710411764686e-06, "loss": 0.1912, "step": 2713 }, { "epoch": 0.40990786890197856, "grad_norm": 2.243325543117922, "learning_rate": 9.546381336807178e-06, "loss": 0.1454, "step": 2714 }, { "epoch": 0.41005890348889895, "grad_norm": 3.333769535119997, "learning_rate": 9.54605214811944e-06, "loss": 0.1935, "step": 2715 }, { "epoch": 0.41020993807581935, "grad_norm": 2.999759983245315, "learning_rate": 9.545722845709707e-06, "loss": 0.181, "step": 2716 }, { "epoch": 0.41036097266273974, "grad_norm": 2.589963232715892, "learning_rate": 9.545393429586214e-06, "loss": 0.1888, "step": 2717 }, { "epoch": 0.4105120072496602, "grad_norm": 2.37033020009293, "learning_rate": 9.545063899757202e-06, "loss": 0.1658, "step": 2718 }, { "epoch": 0.4106630418365806, "grad_norm": 3.646311561524267, "learning_rate": 9.544734256230919e-06, "loss": 0.1937, "step": 2719 }, { "epoch": 0.410814076423501, "grad_norm": 2.197706954408947, "learning_rate": 9.544404499015607e-06, "loss": 0.1425, "step": 2720 }, { "epoch": 0.4109651110104214, "grad_norm": 2.6247445152925817, "learning_rate": 9.544074628119514e-06, "loss": 0.1724, "step": 2721 }, { "epoch": 0.4111161455973418, "grad_norm": 2.768156361251191, "learning_rate": 9.543744643550898e-06, "loss": 0.1425, "step": 2722 }, { "epoch": 0.41126718018426217, "grad_norm": 2.045872363304879, "learning_rate": 9.543414545318008e-06, "loss": 0.1338, "step": 2723 }, { "epoch": 0.4114182147711826, "grad_norm": 2.413571456029521, "learning_rate": 9.543084333429102e-06, "loss": 0.1588, "step": 2724 }, { "epoch": 0.411569249358103, "grad_norm": 2.4551550131531044, "learning_rate": 9.542754007892443e-06, "loss": 0.1643, "step": 2725 }, { "epoch": 0.4117202839450234, "grad_norm": 2.601847065944098, "learning_rate": 9.542423568716295e-06, "loss": 0.1769, "step": 2726 }, { "epoch": 0.4118713185319438, "grad_norm": 3.1198023955064347, "learning_rate": 9.542093015908919e-06, "loss": 0.1813, "step": 2727 }, { "epoch": 0.4120223531188642, "grad_norm": 2.908826757882265, "learning_rate": 9.54176234947859e-06, "loss": 0.1838, "step": 2728 }, { "epoch": 0.4121733877057846, "grad_norm": 2.321928247766814, "learning_rate": 9.541431569433575e-06, "loss": 0.1653, "step": 2729 }, { "epoch": 0.41232442229270505, "grad_norm": 2.564122454892943, "learning_rate": 9.541100675782151e-06, "loss": 0.1721, "step": 2730 }, { "epoch": 0.41247545687962545, "grad_norm": 2.691133800962865, "learning_rate": 9.540769668532594e-06, "loss": 0.1688, "step": 2731 }, { "epoch": 0.41262649146654584, "grad_norm": 1.9278181943228259, "learning_rate": 9.540438547693188e-06, "loss": 0.1282, "step": 2732 }, { "epoch": 0.41277752605346624, "grad_norm": 1.79089965917439, "learning_rate": 9.540107313272211e-06, "loss": 0.1362, "step": 2733 }, { "epoch": 0.41292856064038663, "grad_norm": 2.9682082431373624, "learning_rate": 9.539775965277954e-06, "loss": 0.194, "step": 2734 }, { "epoch": 0.41307959522730703, "grad_norm": 2.217560931218267, "learning_rate": 9.539444503718702e-06, "loss": 0.1464, "step": 2735 }, { "epoch": 0.4132306298142275, "grad_norm": 2.585421233763812, "learning_rate": 9.539112928602748e-06, "loss": 0.1465, "step": 2736 }, { "epoch": 0.4133816644011479, "grad_norm": 2.585080706646738, "learning_rate": 9.538781239938387e-06, "loss": 0.1744, "step": 2737 }, { "epoch": 0.41353269898806827, "grad_norm": 2.2788135083499794, "learning_rate": 9.538449437733917e-06, "loss": 0.1643, "step": 2738 }, { "epoch": 0.41368373357498867, "grad_norm": 2.3243875700715546, "learning_rate": 9.538117521997636e-06, "loss": 0.1655, "step": 2739 }, { "epoch": 0.41383476816190906, "grad_norm": 2.2614885060984182, "learning_rate": 9.537785492737849e-06, "loss": 0.149, "step": 2740 }, { "epoch": 0.41398580274882946, "grad_norm": 2.644074932685391, "learning_rate": 9.537453349962861e-06, "loss": 0.1808, "step": 2741 }, { "epoch": 0.4141368373357499, "grad_norm": 3.5349273211921703, "learning_rate": 9.537121093680981e-06, "loss": 0.1978, "step": 2742 }, { "epoch": 0.4142878719226703, "grad_norm": 3.4246710178928113, "learning_rate": 9.53678872390052e-06, "loss": 0.1823, "step": 2743 }, { "epoch": 0.4144389065095907, "grad_norm": 2.182336760340176, "learning_rate": 9.536456240629794e-06, "loss": 0.1539, "step": 2744 }, { "epoch": 0.4145899410965111, "grad_norm": 2.2169083092729887, "learning_rate": 9.536123643877117e-06, "loss": 0.1543, "step": 2745 }, { "epoch": 0.4147409756834315, "grad_norm": 3.108906709728173, "learning_rate": 9.535790933650815e-06, "loss": 0.181, "step": 2746 }, { "epoch": 0.4148920102703519, "grad_norm": 2.126858513771055, "learning_rate": 9.535458109959204e-06, "loss": 0.1357, "step": 2747 }, { "epoch": 0.41504304485727234, "grad_norm": 2.4450872907344445, "learning_rate": 9.535125172810615e-06, "loss": 0.1661, "step": 2748 }, { "epoch": 0.41519407944419273, "grad_norm": 2.393544088321813, "learning_rate": 9.534792122213373e-06, "loss": 0.1687, "step": 2749 }, { "epoch": 0.4153451140311131, "grad_norm": 2.5754785420654875, "learning_rate": 9.534458958175811e-06, "loss": 0.1609, "step": 2750 }, { "epoch": 0.4154961486180335, "grad_norm": 2.1980896606541767, "learning_rate": 9.534125680706265e-06, "loss": 0.1347, "step": 2751 }, { "epoch": 0.4156471832049539, "grad_norm": 2.3672257164168173, "learning_rate": 9.533792289813068e-06, "loss": 0.1757, "step": 2752 }, { "epoch": 0.4157982177918743, "grad_norm": 2.176701277057895, "learning_rate": 9.533458785504565e-06, "loss": 0.1438, "step": 2753 }, { "epoch": 0.41594925237879476, "grad_norm": 2.5123428886566246, "learning_rate": 9.533125167789096e-06, "loss": 0.1785, "step": 2754 }, { "epoch": 0.41610028696571516, "grad_norm": 2.5845334593378637, "learning_rate": 9.532791436675005e-06, "loss": 0.1498, "step": 2755 }, { "epoch": 0.41625132155263556, "grad_norm": 2.5247960377662966, "learning_rate": 9.532457592170643e-06, "loss": 0.1546, "step": 2756 }, { "epoch": 0.41640235613955595, "grad_norm": 2.5697353275415686, "learning_rate": 9.53212363428436e-06, "loss": 0.1574, "step": 2757 }, { "epoch": 0.41655339072647635, "grad_norm": 2.2183473600379586, "learning_rate": 9.531789563024513e-06, "loss": 0.1634, "step": 2758 }, { "epoch": 0.41670442531339674, "grad_norm": 1.862957393772819, "learning_rate": 9.531455378399454e-06, "loss": 0.1359, "step": 2759 }, { "epoch": 0.4168554599003172, "grad_norm": 2.223284574298462, "learning_rate": 9.531121080417547e-06, "loss": 0.1619, "step": 2760 }, { "epoch": 0.4170064944872376, "grad_norm": 2.614922037467149, "learning_rate": 9.530786669087153e-06, "loss": 0.1651, "step": 2761 }, { "epoch": 0.417157529074158, "grad_norm": 2.2768474322104746, "learning_rate": 9.530452144416637e-06, "loss": 0.1717, "step": 2762 }, { "epoch": 0.4173085636610784, "grad_norm": 2.280455504854231, "learning_rate": 9.530117506414369e-06, "loss": 0.1517, "step": 2763 }, { "epoch": 0.4174595982479988, "grad_norm": 3.003075755265155, "learning_rate": 9.529782755088719e-06, "loss": 0.1626, "step": 2764 }, { "epoch": 0.41761063283491917, "grad_norm": 2.9100508148488613, "learning_rate": 9.52944789044806e-06, "loss": 0.1515, "step": 2765 }, { "epoch": 0.4177616674218396, "grad_norm": 2.442255638661107, "learning_rate": 9.52911291250077e-06, "loss": 0.187, "step": 2766 }, { "epoch": 0.41791270200876, "grad_norm": 2.4545144817160653, "learning_rate": 9.52877782125523e-06, "loss": 0.1618, "step": 2767 }, { "epoch": 0.4180637365956804, "grad_norm": 4.184521399995222, "learning_rate": 9.52844261671982e-06, "loss": 0.17, "step": 2768 }, { "epoch": 0.4182147711826008, "grad_norm": 2.748128502951613, "learning_rate": 9.528107298902928e-06, "loss": 0.1931, "step": 2769 }, { "epoch": 0.4183658057695212, "grad_norm": 2.4875460736677364, "learning_rate": 9.527771867812938e-06, "loss": 0.1737, "step": 2770 }, { "epoch": 0.4185168403564416, "grad_norm": 2.2709791276160254, "learning_rate": 9.527436323458245e-06, "loss": 0.1745, "step": 2771 }, { "epoch": 0.41866787494336205, "grad_norm": 2.4932180788169247, "learning_rate": 9.527100665847242e-06, "loss": 0.1811, "step": 2772 }, { "epoch": 0.41881890953028245, "grad_norm": 2.563108502986208, "learning_rate": 9.526764894988326e-06, "loss": 0.1686, "step": 2773 }, { "epoch": 0.41896994411720284, "grad_norm": 2.7049904809305425, "learning_rate": 9.526429010889893e-06, "loss": 0.1594, "step": 2774 }, { "epoch": 0.41912097870412324, "grad_norm": 1.8630968232502216, "learning_rate": 9.526093013560348e-06, "loss": 0.1501, "step": 2775 }, { "epoch": 0.41927201329104363, "grad_norm": 2.5877747141121032, "learning_rate": 9.525756903008097e-06, "loss": 0.185, "step": 2776 }, { "epoch": 0.41942304787796403, "grad_norm": 2.8597357048289496, "learning_rate": 9.525420679241546e-06, "loss": 0.1493, "step": 2777 }, { "epoch": 0.4195740824648845, "grad_norm": 2.6464755936846927, "learning_rate": 9.525084342269108e-06, "loss": 0.1511, "step": 2778 }, { "epoch": 0.4197251170518049, "grad_norm": 2.5415515565247535, "learning_rate": 9.524747892099195e-06, "loss": 0.1692, "step": 2779 }, { "epoch": 0.41987615163872527, "grad_norm": 2.1669199506376917, "learning_rate": 9.524411328740224e-06, "loss": 0.1476, "step": 2780 }, { "epoch": 0.42002718622564567, "grad_norm": 2.9923125940574873, "learning_rate": 9.524074652200614e-06, "loss": 0.1698, "step": 2781 }, { "epoch": 0.42017822081256606, "grad_norm": 2.5616093546864533, "learning_rate": 9.523737862488789e-06, "loss": 0.1506, "step": 2782 }, { "epoch": 0.42032925539948646, "grad_norm": 2.6160376656527817, "learning_rate": 9.52340095961317e-06, "loss": 0.174, "step": 2783 }, { "epoch": 0.4204802899864069, "grad_norm": 2.497783685166077, "learning_rate": 9.52306394358219e-06, "loss": 0.1808, "step": 2784 }, { "epoch": 0.4206313245733273, "grad_norm": 2.5820331151305513, "learning_rate": 9.522726814404275e-06, "loss": 0.1725, "step": 2785 }, { "epoch": 0.4207823591602477, "grad_norm": 1.9095144352459816, "learning_rate": 9.52238957208786e-06, "loss": 0.157, "step": 2786 }, { "epoch": 0.4209333937471681, "grad_norm": 2.765068744929758, "learning_rate": 9.522052216641382e-06, "loss": 0.1511, "step": 2787 }, { "epoch": 0.4210844283340885, "grad_norm": 2.1713658157691618, "learning_rate": 9.52171474807328e-06, "loss": 0.1302, "step": 2788 }, { "epoch": 0.4212354629210089, "grad_norm": 2.945740861877681, "learning_rate": 9.521377166391996e-06, "loss": 0.185, "step": 2789 }, { "epoch": 0.42138649750792934, "grad_norm": 1.957277225374505, "learning_rate": 9.521039471605976e-06, "loss": 0.1404, "step": 2790 }, { "epoch": 0.42153753209484973, "grad_norm": 2.419760780447007, "learning_rate": 9.520701663723663e-06, "loss": 0.1592, "step": 2791 }, { "epoch": 0.4216885666817701, "grad_norm": 2.049878063673132, "learning_rate": 9.520363742753512e-06, "loss": 0.151, "step": 2792 }, { "epoch": 0.4218396012686905, "grad_norm": 2.328889683269986, "learning_rate": 9.520025708703975e-06, "loss": 0.1566, "step": 2793 }, { "epoch": 0.4219906358556109, "grad_norm": 2.4289959977366475, "learning_rate": 9.519687561583509e-06, "loss": 0.1532, "step": 2794 }, { "epoch": 0.4221416704425313, "grad_norm": 2.285104689118902, "learning_rate": 9.51934930140057e-06, "loss": 0.1471, "step": 2795 }, { "epoch": 0.42229270502945176, "grad_norm": 2.375544324796631, "learning_rate": 9.519010928163624e-06, "loss": 0.1549, "step": 2796 }, { "epoch": 0.42244373961637216, "grad_norm": 3.3709796235072327, "learning_rate": 9.518672441881131e-06, "loss": 0.1937, "step": 2797 }, { "epoch": 0.42259477420329256, "grad_norm": 2.8086063630531184, "learning_rate": 9.518333842561562e-06, "loss": 0.1721, "step": 2798 }, { "epoch": 0.42274580879021295, "grad_norm": 2.722023685477234, "learning_rate": 9.517995130213386e-06, "loss": 0.1741, "step": 2799 }, { "epoch": 0.42289684337713335, "grad_norm": 2.5934724543687446, "learning_rate": 9.517656304845076e-06, "loss": 0.1833, "step": 2800 }, { "epoch": 0.42304787796405374, "grad_norm": 2.2718009670694483, "learning_rate": 9.517317366465106e-06, "loss": 0.1564, "step": 2801 }, { "epoch": 0.4231989125509742, "grad_norm": 10.144869431432841, "learning_rate": 9.516978315081959e-06, "loss": 0.2229, "step": 2802 }, { "epoch": 0.4233499471378946, "grad_norm": 2.5401390146195997, "learning_rate": 9.516639150704112e-06, "loss": 0.1674, "step": 2803 }, { "epoch": 0.423500981724815, "grad_norm": 2.4646044946590067, "learning_rate": 9.516299873340054e-06, "loss": 0.1793, "step": 2804 }, { "epoch": 0.4236520163117354, "grad_norm": 2.4497713611667193, "learning_rate": 9.515960482998269e-06, "loss": 0.1662, "step": 2805 }, { "epoch": 0.4238030508986558, "grad_norm": 2.2337470844813523, "learning_rate": 9.515620979687249e-06, "loss": 0.1553, "step": 2806 }, { "epoch": 0.42395408548557617, "grad_norm": 2.3682187939764012, "learning_rate": 9.515281363415483e-06, "loss": 0.1674, "step": 2807 }, { "epoch": 0.4241051200724966, "grad_norm": 1.962987063247365, "learning_rate": 9.51494163419147e-06, "loss": 0.1299, "step": 2808 }, { "epoch": 0.424256154659417, "grad_norm": 2.5352071093118345, "learning_rate": 9.51460179202371e-06, "loss": 0.1824, "step": 2809 }, { "epoch": 0.4244071892463374, "grad_norm": 3.253147999935852, "learning_rate": 9.5142618369207e-06, "loss": 0.1658, "step": 2810 }, { "epoch": 0.4245582238332578, "grad_norm": 2.5047796414474726, "learning_rate": 9.513921768890947e-06, "loss": 0.1751, "step": 2811 }, { "epoch": 0.4247092584201782, "grad_norm": 2.2885549562700547, "learning_rate": 9.51358158794296e-06, "loss": 0.1632, "step": 2812 }, { "epoch": 0.4248602930070986, "grad_norm": 2.283574044703842, "learning_rate": 9.513241294085243e-06, "loss": 0.1745, "step": 2813 }, { "epoch": 0.42501132759401905, "grad_norm": 2.2241678426201457, "learning_rate": 9.512900887326313e-06, "loss": 0.1387, "step": 2814 }, { "epoch": 0.42516236218093945, "grad_norm": 2.8115575200449134, "learning_rate": 9.512560367674683e-06, "loss": 0.1657, "step": 2815 }, { "epoch": 0.42531339676785984, "grad_norm": 2.591534283179618, "learning_rate": 9.512219735138874e-06, "loss": 0.1581, "step": 2816 }, { "epoch": 0.42546443135478024, "grad_norm": 2.4208918449421013, "learning_rate": 9.511878989727405e-06, "loss": 0.1828, "step": 2817 }, { "epoch": 0.42561546594170063, "grad_norm": 2.040331707785777, "learning_rate": 9.511538131448803e-06, "loss": 0.1431, "step": 2818 }, { "epoch": 0.42576650052862103, "grad_norm": 2.584621095124379, "learning_rate": 9.51119716031159e-06, "loss": 0.1476, "step": 2819 }, { "epoch": 0.4259175351155415, "grad_norm": 2.4274583263210103, "learning_rate": 9.510856076324298e-06, "loss": 0.1574, "step": 2820 }, { "epoch": 0.4260685697024619, "grad_norm": 2.5504801943889253, "learning_rate": 9.510514879495461e-06, "loss": 0.1476, "step": 2821 }, { "epoch": 0.42621960428938227, "grad_norm": 2.228950114819406, "learning_rate": 9.510173569833611e-06, "loss": 0.1739, "step": 2822 }, { "epoch": 0.42637063887630267, "grad_norm": 2.383374801135129, "learning_rate": 9.50983214734729e-06, "loss": 0.1712, "step": 2823 }, { "epoch": 0.42652167346322306, "grad_norm": 1.927103490096801, "learning_rate": 9.509490612045034e-06, "loss": 0.1556, "step": 2824 }, { "epoch": 0.42667270805014346, "grad_norm": 3.672771285146693, "learning_rate": 9.50914896393539e-06, "loss": 0.1613, "step": 2825 }, { "epoch": 0.4268237426370639, "grad_norm": 2.3889041507779187, "learning_rate": 9.508807203026903e-06, "loss": 0.1622, "step": 2826 }, { "epoch": 0.4269747772239843, "grad_norm": 2.80468282377415, "learning_rate": 9.508465329328124e-06, "loss": 0.1652, "step": 2827 }, { "epoch": 0.4271258118109047, "grad_norm": 1.7285493969392478, "learning_rate": 9.508123342847605e-06, "loss": 0.1221, "step": 2828 }, { "epoch": 0.4272768463978251, "grad_norm": 3.0381416171442073, "learning_rate": 9.5077812435939e-06, "loss": 0.1566, "step": 2829 }, { "epoch": 0.4274278809847455, "grad_norm": 3.45466714825477, "learning_rate": 9.507439031575565e-06, "loss": 0.1825, "step": 2830 }, { "epoch": 0.4275789155716659, "grad_norm": 2.4390884961437735, "learning_rate": 9.507096706801164e-06, "loss": 0.1616, "step": 2831 }, { "epoch": 0.42772995015858634, "grad_norm": 2.7393265911720133, "learning_rate": 9.50675426927926e-06, "loss": 0.1465, "step": 2832 }, { "epoch": 0.42788098474550673, "grad_norm": 2.2882866017490007, "learning_rate": 9.506411719018416e-06, "loss": 0.1506, "step": 2833 }, { "epoch": 0.4280320193324271, "grad_norm": 2.491582660784161, "learning_rate": 9.506069056027205e-06, "loss": 0.1613, "step": 2834 }, { "epoch": 0.4281830539193475, "grad_norm": 2.8321615778247495, "learning_rate": 9.505726280314197e-06, "loss": 0.1896, "step": 2835 }, { "epoch": 0.4283340885062679, "grad_norm": 2.4646708258327776, "learning_rate": 9.505383391887968e-06, "loss": 0.1712, "step": 2836 }, { "epoch": 0.4284851230931883, "grad_norm": 2.7110225180173373, "learning_rate": 9.505040390757094e-06, "loss": 0.183, "step": 2837 }, { "epoch": 0.42863615768010876, "grad_norm": 2.9538253190019477, "learning_rate": 9.504697276930155e-06, "loss": 0.1794, "step": 2838 }, { "epoch": 0.42878719226702916, "grad_norm": 2.9594344083715955, "learning_rate": 9.504354050415736e-06, "loss": 0.148, "step": 2839 }, { "epoch": 0.42893822685394956, "grad_norm": 3.0148290429049824, "learning_rate": 9.504010711222422e-06, "loss": 0.1943, "step": 2840 }, { "epoch": 0.42908926144086995, "grad_norm": 2.3107101756986816, "learning_rate": 9.503667259358802e-06, "loss": 0.1554, "step": 2841 }, { "epoch": 0.42924029602779035, "grad_norm": 2.7891479560436863, "learning_rate": 9.503323694833468e-06, "loss": 0.1785, "step": 2842 }, { "epoch": 0.42939133061471074, "grad_norm": 2.351872902048178, "learning_rate": 9.502980017655015e-06, "loss": 0.155, "step": 2843 }, { "epoch": 0.4295423652016312, "grad_norm": 3.3016303340257753, "learning_rate": 9.502636227832037e-06, "loss": 0.1952, "step": 2844 }, { "epoch": 0.4296933997885516, "grad_norm": 1.7854277568383117, "learning_rate": 9.502292325373136e-06, "loss": 0.1196, "step": 2845 }, { "epoch": 0.429844434375472, "grad_norm": 2.1734944078342835, "learning_rate": 9.50194831028692e-06, "loss": 0.1376, "step": 2846 }, { "epoch": 0.4299954689623924, "grad_norm": 1.996728454913964, "learning_rate": 9.501604182581985e-06, "loss": 0.1417, "step": 2847 }, { "epoch": 0.4301465035493128, "grad_norm": 2.118828728768492, "learning_rate": 9.501259942266946e-06, "loss": 0.1529, "step": 2848 }, { "epoch": 0.43029753813623317, "grad_norm": 3.5975126462365847, "learning_rate": 9.500915589350414e-06, "loss": 0.1678, "step": 2849 }, { "epoch": 0.4304485727231536, "grad_norm": 2.2464609010590584, "learning_rate": 9.500571123841002e-06, "loss": 0.1708, "step": 2850 }, { "epoch": 0.430599607310074, "grad_norm": 2.004045529236699, "learning_rate": 9.500226545747326e-06, "loss": 0.1393, "step": 2851 }, { "epoch": 0.4307506418969944, "grad_norm": 2.070183853175175, "learning_rate": 9.499881855078009e-06, "loss": 0.1417, "step": 2852 }, { "epoch": 0.4309016764839148, "grad_norm": 2.4727384902328198, "learning_rate": 9.49953705184167e-06, "loss": 0.1868, "step": 2853 }, { "epoch": 0.4310527110708352, "grad_norm": 2.8305656840638354, "learning_rate": 9.499192136046936e-06, "loss": 0.1843, "step": 2854 }, { "epoch": 0.4312037456577556, "grad_norm": 2.440066083573402, "learning_rate": 9.498847107702436e-06, "loss": 0.1647, "step": 2855 }, { "epoch": 0.43135478024467605, "grad_norm": 4.641916607813762, "learning_rate": 9.4985019668168e-06, "loss": 0.1909, "step": 2856 }, { "epoch": 0.43150581483159645, "grad_norm": 1.9334394490614784, "learning_rate": 9.498156713398661e-06, "loss": 0.1559, "step": 2857 }, { "epoch": 0.43165684941851684, "grad_norm": 3.1180985631234432, "learning_rate": 9.49781134745666e-06, "loss": 0.179, "step": 2858 }, { "epoch": 0.43180788400543724, "grad_norm": 2.751761594322883, "learning_rate": 9.497465868999432e-06, "loss": 0.1871, "step": 2859 }, { "epoch": 0.43195891859235763, "grad_norm": 2.4681759734856135, "learning_rate": 9.497120278035618e-06, "loss": 0.1584, "step": 2860 }, { "epoch": 0.43210995317927803, "grad_norm": 2.167525378266776, "learning_rate": 9.496774574573869e-06, "loss": 0.1583, "step": 2861 }, { "epoch": 0.4322609877661985, "grad_norm": 2.948746666985831, "learning_rate": 9.49642875862283e-06, "loss": 0.2017, "step": 2862 }, { "epoch": 0.4324120223531189, "grad_norm": 2.1335249929638014, "learning_rate": 9.496082830191152e-06, "loss": 0.1491, "step": 2863 }, { "epoch": 0.43256305694003927, "grad_norm": 2.3952229160679877, "learning_rate": 9.495736789287487e-06, "loss": 0.1798, "step": 2864 }, { "epoch": 0.43271409152695967, "grad_norm": 2.866164677629776, "learning_rate": 9.495390635920492e-06, "loss": 0.1633, "step": 2865 }, { "epoch": 0.43286512611388006, "grad_norm": 2.660836274512189, "learning_rate": 9.495044370098827e-06, "loss": 0.1743, "step": 2866 }, { "epoch": 0.43301616070080046, "grad_norm": 2.118488239700756, "learning_rate": 9.494697991831154e-06, "loss": 0.1329, "step": 2867 }, { "epoch": 0.4331671952877209, "grad_norm": 2.895114621855141, "learning_rate": 9.49435150112614e-06, "loss": 0.1519, "step": 2868 }, { "epoch": 0.4333182298746413, "grad_norm": 4.900402600714372, "learning_rate": 9.494004897992449e-06, "loss": 0.1597, "step": 2869 }, { "epoch": 0.4334692644615617, "grad_norm": 2.2176992013649794, "learning_rate": 9.493658182438752e-06, "loss": 0.1671, "step": 2870 }, { "epoch": 0.4336202990484821, "grad_norm": 2.475471200897973, "learning_rate": 9.493311354473725e-06, "loss": 0.1688, "step": 2871 }, { "epoch": 0.4337713336354025, "grad_norm": 3.0935396806806197, "learning_rate": 9.492964414106041e-06, "loss": 0.1609, "step": 2872 }, { "epoch": 0.4339223682223229, "grad_norm": 3.142184174321951, "learning_rate": 9.49261736134438e-06, "loss": 0.1537, "step": 2873 }, { "epoch": 0.43407340280924334, "grad_norm": 2.4767502829279593, "learning_rate": 9.492270196197424e-06, "loss": 0.1467, "step": 2874 }, { "epoch": 0.43422443739616373, "grad_norm": 2.358537354029172, "learning_rate": 9.491922918673857e-06, "loss": 0.1683, "step": 2875 }, { "epoch": 0.4343754719830841, "grad_norm": 3.07842375930715, "learning_rate": 9.491575528782368e-06, "loss": 0.2055, "step": 2876 }, { "epoch": 0.4345265065700045, "grad_norm": 2.43340515316373, "learning_rate": 9.491228026531645e-06, "loss": 0.1633, "step": 2877 }, { "epoch": 0.4346775411569249, "grad_norm": 2.262704409939446, "learning_rate": 9.490880411930382e-06, "loss": 0.1643, "step": 2878 }, { "epoch": 0.4348285757438453, "grad_norm": 2.6670834079513113, "learning_rate": 9.490532684987275e-06, "loss": 0.1998, "step": 2879 }, { "epoch": 0.43497961033076576, "grad_norm": 2.7725707964186643, "learning_rate": 9.49018484571102e-06, "loss": 0.174, "step": 2880 }, { "epoch": 0.43513064491768616, "grad_norm": 2.295420239616843, "learning_rate": 9.489836894110324e-06, "loss": 0.1502, "step": 2881 }, { "epoch": 0.43528167950460656, "grad_norm": 2.2225295625014385, "learning_rate": 9.489488830193887e-06, "loss": 0.1329, "step": 2882 }, { "epoch": 0.43543271409152695, "grad_norm": 2.568785653135949, "learning_rate": 9.489140653970416e-06, "loss": 0.1607, "step": 2883 }, { "epoch": 0.43558374867844735, "grad_norm": 2.7066362672959667, "learning_rate": 9.488792365448622e-06, "loss": 0.1649, "step": 2884 }, { "epoch": 0.43573478326536774, "grad_norm": 1.9417063697771102, "learning_rate": 9.488443964637217e-06, "loss": 0.1455, "step": 2885 }, { "epoch": 0.4358858178522882, "grad_norm": 2.6264881935770115, "learning_rate": 9.488095451544917e-06, "loss": 0.1619, "step": 2886 }, { "epoch": 0.4360368524392086, "grad_norm": 6.8239000944682315, "learning_rate": 9.48774682618044e-06, "loss": 0.1971, "step": 2887 }, { "epoch": 0.436187887026129, "grad_norm": 2.688963531375313, "learning_rate": 9.487398088552507e-06, "loss": 0.1898, "step": 2888 }, { "epoch": 0.4363389216130494, "grad_norm": 2.972945867784992, "learning_rate": 9.48704923866984e-06, "loss": 0.1444, "step": 2889 }, { "epoch": 0.4364899561999698, "grad_norm": 3.1287869889106603, "learning_rate": 9.48670027654117e-06, "loss": 0.1717, "step": 2890 }, { "epoch": 0.43664099078689017, "grad_norm": 2.296110434021422, "learning_rate": 9.486351202175223e-06, "loss": 0.1628, "step": 2891 }, { "epoch": 0.4367920253738106, "grad_norm": 2.993650604568296, "learning_rate": 9.486002015580733e-06, "loss": 0.1533, "step": 2892 }, { "epoch": 0.436943059960731, "grad_norm": 2.8217440446439332, "learning_rate": 9.485652716766434e-06, "loss": 0.171, "step": 2893 }, { "epoch": 0.4370940945476514, "grad_norm": 2.3371701868447308, "learning_rate": 9.485303305741062e-06, "loss": 0.165, "step": 2894 }, { "epoch": 0.4372451291345718, "grad_norm": 2.5300341000578213, "learning_rate": 9.484953782513365e-06, "loss": 0.1819, "step": 2895 }, { "epoch": 0.4373961637214922, "grad_norm": 2.705799786611232, "learning_rate": 9.484604147092077e-06, "loss": 0.1556, "step": 2896 }, { "epoch": 0.4375471983084126, "grad_norm": 2.1378122997683775, "learning_rate": 9.48425439948595e-06, "loss": 0.1574, "step": 2897 }, { "epoch": 0.43769823289533305, "grad_norm": 2.4514169683371048, "learning_rate": 9.483904539703734e-06, "loss": 0.1845, "step": 2898 }, { "epoch": 0.43784926748225345, "grad_norm": 2.71243064672052, "learning_rate": 9.483554567754176e-06, "loss": 0.1563, "step": 2899 }, { "epoch": 0.43800030206917384, "grad_norm": 2.1325784902732607, "learning_rate": 9.483204483646035e-06, "loss": 0.1497, "step": 2900 }, { "epoch": 0.43815133665609424, "grad_norm": 2.136070504941807, "learning_rate": 9.482854287388068e-06, "loss": 0.1602, "step": 2901 }, { "epoch": 0.43830237124301463, "grad_norm": 2.680211510740247, "learning_rate": 9.482503978989033e-06, "loss": 0.1606, "step": 2902 }, { "epoch": 0.43845340582993503, "grad_norm": 2.6579752423604566, "learning_rate": 9.482153558457696e-06, "loss": 0.1875, "step": 2903 }, { "epoch": 0.4386044404168555, "grad_norm": 2.6782590561851487, "learning_rate": 9.48180302580282e-06, "loss": 0.1588, "step": 2904 }, { "epoch": 0.4387554750037759, "grad_norm": 2.137426130271736, "learning_rate": 9.481452381033178e-06, "loss": 0.1461, "step": 2905 }, { "epoch": 0.43890650959069627, "grad_norm": 2.2816372501099282, "learning_rate": 9.481101624157538e-06, "loss": 0.1409, "step": 2906 }, { "epoch": 0.43905754417761667, "grad_norm": 2.467453692329133, "learning_rate": 9.480750755184674e-06, "loss": 0.1557, "step": 2907 }, { "epoch": 0.43920857876453706, "grad_norm": 2.274986767328205, "learning_rate": 9.480399774123365e-06, "loss": 0.157, "step": 2908 }, { "epoch": 0.43935961335145746, "grad_norm": 3.000273024300605, "learning_rate": 9.480048680982392e-06, "loss": 0.1868, "step": 2909 }, { "epoch": 0.4395106479383779, "grad_norm": 2.5373864610530057, "learning_rate": 9.479697475770534e-06, "loss": 0.1709, "step": 2910 }, { "epoch": 0.4396616825252983, "grad_norm": 2.1172918695158307, "learning_rate": 9.479346158496581e-06, "loss": 0.138, "step": 2911 }, { "epoch": 0.4398127171122187, "grad_norm": 3.371685800387069, "learning_rate": 9.47899472916932e-06, "loss": 0.1924, "step": 2912 }, { "epoch": 0.4399637516991391, "grad_norm": 2.8557231840106856, "learning_rate": 9.478643187797539e-06, "loss": 0.1709, "step": 2913 }, { "epoch": 0.4401147862860595, "grad_norm": 2.478006854475944, "learning_rate": 9.478291534390036e-06, "loss": 0.2075, "step": 2914 }, { "epoch": 0.4402658208729799, "grad_norm": 2.4476351379551, "learning_rate": 9.477939768955606e-06, "loss": 0.1757, "step": 2915 }, { "epoch": 0.44041685545990034, "grad_norm": 2.1360726396913625, "learning_rate": 9.47758789150305e-06, "loss": 0.1452, "step": 2916 }, { "epoch": 0.44056789004682073, "grad_norm": 2.2962397005935413, "learning_rate": 9.477235902041168e-06, "loss": 0.1774, "step": 2917 }, { "epoch": 0.4407189246337411, "grad_norm": 2.5810476993724993, "learning_rate": 9.476883800578766e-06, "loss": 0.1529, "step": 2918 }, { "epoch": 0.4408699592206615, "grad_norm": 1.9202729574984543, "learning_rate": 9.476531587124653e-06, "loss": 0.1398, "step": 2919 }, { "epoch": 0.4410209938075819, "grad_norm": 1.9486942521088866, "learning_rate": 9.476179261687641e-06, "loss": 0.1475, "step": 2920 }, { "epoch": 0.4411720283945023, "grad_norm": 2.307021220610515, "learning_rate": 9.475826824276541e-06, "loss": 0.1729, "step": 2921 }, { "epoch": 0.44132306298142276, "grad_norm": 4.097696485095745, "learning_rate": 9.475474274900169e-06, "loss": 0.1794, "step": 2922 }, { "epoch": 0.44147409756834316, "grad_norm": 3.382004011428012, "learning_rate": 9.475121613567347e-06, "loss": 0.1936, "step": 2923 }, { "epoch": 0.44162513215526356, "grad_norm": 2.5623629021655367, "learning_rate": 9.474768840286896e-06, "loss": 0.1882, "step": 2924 }, { "epoch": 0.44177616674218395, "grad_norm": 1.999838992906664, "learning_rate": 9.474415955067639e-06, "loss": 0.1401, "step": 2925 }, { "epoch": 0.44192720132910435, "grad_norm": 2.5122477719301437, "learning_rate": 9.474062957918405e-06, "loss": 0.1688, "step": 2926 }, { "epoch": 0.44207823591602474, "grad_norm": 2.1034744075933953, "learning_rate": 9.473709848848027e-06, "loss": 0.1737, "step": 2927 }, { "epoch": 0.4422292705029452, "grad_norm": 3.128369412823823, "learning_rate": 9.473356627865335e-06, "loss": 0.1581, "step": 2928 }, { "epoch": 0.4423803050898656, "grad_norm": 2.2927910361706925, "learning_rate": 9.473003294979163e-06, "loss": 0.1601, "step": 2929 }, { "epoch": 0.442531339676786, "grad_norm": 2.4901864468659887, "learning_rate": 9.472649850198357e-06, "loss": 0.1554, "step": 2930 }, { "epoch": 0.4426823742637064, "grad_norm": 2.5224512272989985, "learning_rate": 9.472296293531751e-06, "loss": 0.1461, "step": 2931 }, { "epoch": 0.4428334088506268, "grad_norm": 2.766253092047908, "learning_rate": 9.471942624988193e-06, "loss": 0.1618, "step": 2932 }, { "epoch": 0.44298444343754717, "grad_norm": 2.379679120268621, "learning_rate": 9.47158884457653e-06, "loss": 0.1548, "step": 2933 }, { "epoch": 0.4431354780244676, "grad_norm": 2.255289606679209, "learning_rate": 9.471234952305612e-06, "loss": 0.1617, "step": 2934 }, { "epoch": 0.443286512611388, "grad_norm": 3.1652841077493328, "learning_rate": 9.470880948184292e-06, "loss": 0.1476, "step": 2935 }, { "epoch": 0.4434375471983084, "grad_norm": 3.0044197907428156, "learning_rate": 9.470526832221422e-06, "loss": 0.1664, "step": 2936 }, { "epoch": 0.4435885817852288, "grad_norm": 2.0636027257648752, "learning_rate": 9.470172604425867e-06, "loss": 0.1606, "step": 2937 }, { "epoch": 0.4437396163721492, "grad_norm": 2.247462007381648, "learning_rate": 9.469818264806485e-06, "loss": 0.1496, "step": 2938 }, { "epoch": 0.4438906509590696, "grad_norm": 2.272450870013867, "learning_rate": 9.469463813372138e-06, "loss": 0.1531, "step": 2939 }, { "epoch": 0.44404168554599005, "grad_norm": 1.9444524169558222, "learning_rate": 9.469109250131696e-06, "loss": 0.1453, "step": 2940 }, { "epoch": 0.44419272013291045, "grad_norm": 2.4559643051012277, "learning_rate": 9.468754575094027e-06, "loss": 0.1596, "step": 2941 }, { "epoch": 0.44434375471983084, "grad_norm": 3.3755351761364616, "learning_rate": 9.468399788268003e-06, "loss": 0.2146, "step": 2942 }, { "epoch": 0.44449478930675124, "grad_norm": 2.8473839156072906, "learning_rate": 9.4680448896625e-06, "loss": 0.1766, "step": 2943 }, { "epoch": 0.44464582389367163, "grad_norm": 2.4796300158630764, "learning_rate": 9.467689879286397e-06, "loss": 0.1426, "step": 2944 }, { "epoch": 0.4447968584805921, "grad_norm": 2.3710703144517926, "learning_rate": 9.467334757148572e-06, "loss": 0.1656, "step": 2945 }, { "epoch": 0.4449478930675125, "grad_norm": 2.6003755692921904, "learning_rate": 9.46697952325791e-06, "loss": 0.1579, "step": 2946 }, { "epoch": 0.4450989276544329, "grad_norm": 2.8653151642071273, "learning_rate": 9.466624177623298e-06, "loss": 0.2078, "step": 2947 }, { "epoch": 0.44524996224135327, "grad_norm": 2.389612281430796, "learning_rate": 9.466268720253625e-06, "loss": 0.1747, "step": 2948 }, { "epoch": 0.44540099682827367, "grad_norm": 2.017366299474137, "learning_rate": 9.465913151157782e-06, "loss": 0.137, "step": 2949 }, { "epoch": 0.44555203141519406, "grad_norm": 2.856040644488685, "learning_rate": 9.465557470344665e-06, "loss": 0.1991, "step": 2950 }, { "epoch": 0.4457030660021145, "grad_norm": 2.8568713903365244, "learning_rate": 9.46520167782317e-06, "loss": 0.2028, "step": 2951 }, { "epoch": 0.4458541005890349, "grad_norm": 2.2479624170645476, "learning_rate": 9.4648457736022e-06, "loss": 0.1853, "step": 2952 }, { "epoch": 0.4460051351759553, "grad_norm": 2.3371091683970824, "learning_rate": 9.464489757690655e-06, "loss": 0.1841, "step": 2953 }, { "epoch": 0.4461561697628757, "grad_norm": 2.7552113604428086, "learning_rate": 9.464133630097443e-06, "loss": 0.1544, "step": 2954 }, { "epoch": 0.4463072043497961, "grad_norm": 2.4392994062736326, "learning_rate": 9.463777390831472e-06, "loss": 0.1682, "step": 2955 }, { "epoch": 0.4464582389367165, "grad_norm": 2.359042797507817, "learning_rate": 9.463421039901654e-06, "loss": 0.166, "step": 2956 }, { "epoch": 0.44660927352363694, "grad_norm": 2.3069904657862104, "learning_rate": 9.463064577316902e-06, "loss": 0.1553, "step": 2957 }, { "epoch": 0.44676030811055734, "grad_norm": 2.109910651373422, "learning_rate": 9.462708003086134e-06, "loss": 0.1626, "step": 2958 }, { "epoch": 0.44691134269747773, "grad_norm": 4.005710900073862, "learning_rate": 9.46235131721827e-06, "loss": 0.2124, "step": 2959 }, { "epoch": 0.4470623772843981, "grad_norm": 2.6591270818183372, "learning_rate": 9.461994519722236e-06, "loss": 0.1758, "step": 2960 }, { "epoch": 0.4472134118713185, "grad_norm": 2.6651848515225245, "learning_rate": 9.46163761060695e-06, "loss": 0.1558, "step": 2961 }, { "epoch": 0.4473644464582389, "grad_norm": 2.6242756097488398, "learning_rate": 9.461280589881345e-06, "loss": 0.156, "step": 2962 }, { "epoch": 0.44751548104515937, "grad_norm": 2.809231105585134, "learning_rate": 9.460923457554351e-06, "loss": 0.1808, "step": 2963 }, { "epoch": 0.44766651563207976, "grad_norm": 2.8580035545141014, "learning_rate": 9.460566213634905e-06, "loss": 0.1788, "step": 2964 }, { "epoch": 0.44781755021900016, "grad_norm": 2.200620954018476, "learning_rate": 9.460208858131938e-06, "loss": 0.16, "step": 2965 }, { "epoch": 0.44796858480592056, "grad_norm": 2.219095551884271, "learning_rate": 9.459851391054395e-06, "loss": 0.1464, "step": 2966 }, { "epoch": 0.44811961939284095, "grad_norm": 2.4646760370934278, "learning_rate": 9.459493812411213e-06, "loss": 0.1483, "step": 2967 }, { "epoch": 0.44827065397976135, "grad_norm": 3.5116981808593875, "learning_rate": 9.459136122211343e-06, "loss": 0.1966, "step": 2968 }, { "epoch": 0.4484216885666818, "grad_norm": 2.4147040667275768, "learning_rate": 9.458778320463726e-06, "loss": 0.1563, "step": 2969 }, { "epoch": 0.4485727231536022, "grad_norm": 1.9891446924887939, "learning_rate": 9.458420407177319e-06, "loss": 0.1626, "step": 2970 }, { "epoch": 0.4487237577405226, "grad_norm": 2.4656387691775623, "learning_rate": 9.45806238236107e-06, "loss": 0.1584, "step": 2971 }, { "epoch": 0.448874792327443, "grad_norm": 2.4017779410464737, "learning_rate": 9.45770424602394e-06, "loss": 0.164, "step": 2972 }, { "epoch": 0.4490258269143634, "grad_norm": 2.3005798973167813, "learning_rate": 9.457345998174885e-06, "loss": 0.1697, "step": 2973 }, { "epoch": 0.4491768615012838, "grad_norm": 2.3115686345138102, "learning_rate": 9.456987638822866e-06, "loss": 0.157, "step": 2974 }, { "epoch": 0.4493278960882042, "grad_norm": 2.242067478597751, "learning_rate": 9.45662916797685e-06, "loss": 0.1572, "step": 2975 }, { "epoch": 0.4494789306751246, "grad_norm": 2.3368105384171405, "learning_rate": 9.456270585645804e-06, "loss": 0.1714, "step": 2976 }, { "epoch": 0.449629965262045, "grad_norm": 2.52977162205854, "learning_rate": 9.455911891838697e-06, "loss": 0.1551, "step": 2977 }, { "epoch": 0.4497809998489654, "grad_norm": 1.95850748607203, "learning_rate": 9.455553086564502e-06, "loss": 0.1463, "step": 2978 }, { "epoch": 0.4499320344358858, "grad_norm": 2.7298233950200332, "learning_rate": 9.455194169832196e-06, "loss": 0.1812, "step": 2979 }, { "epoch": 0.4500830690228062, "grad_norm": 2.7044966731620748, "learning_rate": 9.454835141650755e-06, "loss": 0.1562, "step": 2980 }, { "epoch": 0.45023410360972665, "grad_norm": 2.3045746723815563, "learning_rate": 9.454476002029162e-06, "loss": 0.15, "step": 2981 }, { "epoch": 0.45038513819664705, "grad_norm": 2.3072966384252602, "learning_rate": 9.454116750976403e-06, "loss": 0.1697, "step": 2982 }, { "epoch": 0.45053617278356745, "grad_norm": 2.439912809954086, "learning_rate": 9.453757388501461e-06, "loss": 0.1886, "step": 2983 }, { "epoch": 0.45068720737048784, "grad_norm": 3.230301926767873, "learning_rate": 9.453397914613329e-06, "loss": 0.1745, "step": 2984 }, { "epoch": 0.45083824195740824, "grad_norm": 2.226497304134881, "learning_rate": 9.453038329320995e-06, "loss": 0.1628, "step": 2985 }, { "epoch": 0.45098927654432863, "grad_norm": 2.4336057722878848, "learning_rate": 9.45267863263346e-06, "loss": 0.1362, "step": 2986 }, { "epoch": 0.4511403111312491, "grad_norm": 2.363523595018171, "learning_rate": 9.452318824559718e-06, "loss": 0.1579, "step": 2987 }, { "epoch": 0.4512913457181695, "grad_norm": 2.672884256389289, "learning_rate": 9.451958905108772e-06, "loss": 0.1681, "step": 2988 }, { "epoch": 0.4514423803050899, "grad_norm": 2.495397003280486, "learning_rate": 9.451598874289624e-06, "loss": 0.1699, "step": 2989 }, { "epoch": 0.45159341489201027, "grad_norm": 1.903626874474732, "learning_rate": 9.45123873211128e-06, "loss": 0.1336, "step": 2990 }, { "epoch": 0.45174444947893067, "grad_norm": 1.999747968107678, "learning_rate": 9.450878478582752e-06, "loss": 0.1545, "step": 2991 }, { "epoch": 0.45189548406585106, "grad_norm": 2.0209188604327304, "learning_rate": 9.450518113713048e-06, "loss": 0.1403, "step": 2992 }, { "epoch": 0.4520465186527715, "grad_norm": 2.623277539979502, "learning_rate": 9.450157637511186e-06, "loss": 0.167, "step": 2993 }, { "epoch": 0.4521975532396919, "grad_norm": 2.134376316446282, "learning_rate": 9.449797049986181e-06, "loss": 0.1531, "step": 2994 }, { "epoch": 0.4523485878266123, "grad_norm": 2.8572798622902686, "learning_rate": 9.449436351147056e-06, "loss": 0.1981, "step": 2995 }, { "epoch": 0.4524996224135327, "grad_norm": 3.422460319961676, "learning_rate": 9.449075541002832e-06, "loss": 0.1579, "step": 2996 }, { "epoch": 0.4526506570004531, "grad_norm": 2.1734431317065, "learning_rate": 9.448714619562535e-06, "loss": 0.1452, "step": 2997 }, { "epoch": 0.4528016915873735, "grad_norm": 2.2282243002769824, "learning_rate": 9.448353586835196e-06, "loss": 0.1597, "step": 2998 }, { "epoch": 0.45295272617429394, "grad_norm": 1.95855378537138, "learning_rate": 9.447992442829843e-06, "loss": 0.1558, "step": 2999 }, { "epoch": 0.45310376076121434, "grad_norm": 4.5092899472882815, "learning_rate": 9.447631187555511e-06, "loss": 0.2537, "step": 3000 }, { "epoch": 0.45325479534813473, "grad_norm": 2.7975961013795123, "learning_rate": 9.447269821021238e-06, "loss": 0.191, "step": 3001 }, { "epoch": 0.4534058299350551, "grad_norm": 2.270376014995287, "learning_rate": 9.446908343236064e-06, "loss": 0.1392, "step": 3002 }, { "epoch": 0.4535568645219755, "grad_norm": 2.165744561568729, "learning_rate": 9.446546754209032e-06, "loss": 0.1674, "step": 3003 }, { "epoch": 0.4537078991088959, "grad_norm": 2.3613708804467235, "learning_rate": 9.446185053949185e-06, "loss": 0.1679, "step": 3004 }, { "epoch": 0.45385893369581637, "grad_norm": 2.573978169499503, "learning_rate": 9.445823242465573e-06, "loss": 0.1351, "step": 3005 }, { "epoch": 0.45400996828273676, "grad_norm": 2.2849926033941834, "learning_rate": 9.445461319767248e-06, "loss": 0.1612, "step": 3006 }, { "epoch": 0.45416100286965716, "grad_norm": 2.0951380368646397, "learning_rate": 9.44509928586326e-06, "loss": 0.158, "step": 3007 }, { "epoch": 0.45431203745657756, "grad_norm": 2.384123630536586, "learning_rate": 9.44473714076267e-06, "loss": 0.1558, "step": 3008 }, { "epoch": 0.45446307204349795, "grad_norm": 3.521618189789225, "learning_rate": 9.444374884474532e-06, "loss": 0.1624, "step": 3009 }, { "epoch": 0.45461410663041835, "grad_norm": 2.283916740088682, "learning_rate": 9.444012517007914e-06, "loss": 0.1394, "step": 3010 }, { "epoch": 0.4547651412173388, "grad_norm": 7.031277597384395, "learning_rate": 9.443650038371877e-06, "loss": 0.197, "step": 3011 }, { "epoch": 0.4549161758042592, "grad_norm": 3.9200381895289684, "learning_rate": 9.443287448575487e-06, "loss": 0.1647, "step": 3012 }, { "epoch": 0.4550672103911796, "grad_norm": 2.181026925503647, "learning_rate": 9.44292474762782e-06, "loss": 0.1558, "step": 3013 }, { "epoch": 0.4552182449781, "grad_norm": 2.4244846039922154, "learning_rate": 9.442561935537944e-06, "loss": 0.1704, "step": 3014 }, { "epoch": 0.4553692795650204, "grad_norm": 2.9243412601989056, "learning_rate": 9.442199012314937e-06, "loss": 0.1784, "step": 3015 }, { "epoch": 0.4555203141519408, "grad_norm": 2.0370487525816463, "learning_rate": 9.441835977967879e-06, "loss": 0.1556, "step": 3016 }, { "epoch": 0.4556713487388612, "grad_norm": 2.7900053319945584, "learning_rate": 9.441472832505849e-06, "loss": 0.1737, "step": 3017 }, { "epoch": 0.4558223833257816, "grad_norm": 2.055443372103988, "learning_rate": 9.441109575937933e-06, "loss": 0.1474, "step": 3018 }, { "epoch": 0.455973417912702, "grad_norm": 2.1141165775061457, "learning_rate": 9.440746208273216e-06, "loss": 0.1608, "step": 3019 }, { "epoch": 0.4561244524996224, "grad_norm": 2.3944545366223653, "learning_rate": 9.440382729520791e-06, "loss": 0.1563, "step": 3020 }, { "epoch": 0.4562754870865428, "grad_norm": 2.6150087200296155, "learning_rate": 9.44001913968975e-06, "loss": 0.1893, "step": 3021 }, { "epoch": 0.4564265216734632, "grad_norm": 2.4818112523932556, "learning_rate": 9.439655438789183e-06, "loss": 0.1614, "step": 3022 }, { "epoch": 0.45657755626038365, "grad_norm": 2.215795976528233, "learning_rate": 9.439291626828196e-06, "loss": 0.1672, "step": 3023 }, { "epoch": 0.45672859084730405, "grad_norm": 2.6120277461966177, "learning_rate": 9.438927703815887e-06, "loss": 0.1932, "step": 3024 }, { "epoch": 0.45687962543422445, "grad_norm": 2.524218476129623, "learning_rate": 9.438563669761359e-06, "loss": 0.1664, "step": 3025 }, { "epoch": 0.45703066002114484, "grad_norm": 2.2164423876628496, "learning_rate": 9.438199524673717e-06, "loss": 0.1679, "step": 3026 }, { "epoch": 0.45718169460806524, "grad_norm": 2.4972259767506584, "learning_rate": 9.437835268562073e-06, "loss": 0.1698, "step": 3027 }, { "epoch": 0.45733272919498563, "grad_norm": 2.2850065903143895, "learning_rate": 9.437470901435538e-06, "loss": 0.1568, "step": 3028 }, { "epoch": 0.4574837637819061, "grad_norm": 2.6710490125050397, "learning_rate": 9.437106423303224e-06, "loss": 0.1455, "step": 3029 }, { "epoch": 0.4576347983688265, "grad_norm": 2.275935076294426, "learning_rate": 9.436741834174255e-06, "loss": 0.1563, "step": 3030 }, { "epoch": 0.4577858329557469, "grad_norm": 2.870487066240661, "learning_rate": 9.436377134057747e-06, "loss": 0.1495, "step": 3031 }, { "epoch": 0.45793686754266727, "grad_norm": 2.7517979991552632, "learning_rate": 9.436012322962824e-06, "loss": 0.1503, "step": 3032 }, { "epoch": 0.45808790212958767, "grad_norm": 2.167634131260022, "learning_rate": 9.435647400898614e-06, "loss": 0.1584, "step": 3033 }, { "epoch": 0.45823893671650806, "grad_norm": 2.6147261227445244, "learning_rate": 9.43528236787424e-06, "loss": 0.1689, "step": 3034 }, { "epoch": 0.4583899713034285, "grad_norm": 2.5986029154689594, "learning_rate": 9.434917223898841e-06, "loss": 0.1551, "step": 3035 }, { "epoch": 0.4585410058903489, "grad_norm": 2.812143230806351, "learning_rate": 9.434551968981545e-06, "loss": 0.1717, "step": 3036 }, { "epoch": 0.4586920404772693, "grad_norm": 2.679012226625913, "learning_rate": 9.434186603131493e-06, "loss": 0.1488, "step": 3037 }, { "epoch": 0.4588430750641897, "grad_norm": 2.239540801184452, "learning_rate": 9.433821126357822e-06, "loss": 0.1641, "step": 3038 }, { "epoch": 0.4589941096511101, "grad_norm": 4.290523295106972, "learning_rate": 9.433455538669678e-06, "loss": 0.1649, "step": 3039 }, { "epoch": 0.4591451442380305, "grad_norm": 1.9686368097922249, "learning_rate": 9.433089840076204e-06, "loss": 0.1507, "step": 3040 }, { "epoch": 0.45929617882495094, "grad_norm": 2.2323317902635997, "learning_rate": 9.432724030586547e-06, "loss": 0.183, "step": 3041 }, { "epoch": 0.45944721341187134, "grad_norm": 2.566389876516711, "learning_rate": 9.43235811020986e-06, "loss": 0.1741, "step": 3042 }, { "epoch": 0.45959824799879173, "grad_norm": 2.454838726449345, "learning_rate": 9.431992078955297e-06, "loss": 0.1997, "step": 3043 }, { "epoch": 0.4597492825857121, "grad_norm": 2.814561590920498, "learning_rate": 9.431625936832014e-06, "loss": 0.1831, "step": 3044 }, { "epoch": 0.4599003171726325, "grad_norm": 2.3991133124174837, "learning_rate": 9.43125968384917e-06, "loss": 0.1531, "step": 3045 }, { "epoch": 0.4600513517595529, "grad_norm": 2.081119620399646, "learning_rate": 9.430893320015926e-06, "loss": 0.1514, "step": 3046 }, { "epoch": 0.46020238634647337, "grad_norm": 2.148145477795726, "learning_rate": 9.430526845341448e-06, "loss": 0.1683, "step": 3047 }, { "epoch": 0.46035342093339376, "grad_norm": 2.0773428074131877, "learning_rate": 9.430160259834902e-06, "loss": 0.1438, "step": 3048 }, { "epoch": 0.46050445552031416, "grad_norm": 3.3691030522500243, "learning_rate": 9.429793563505463e-06, "loss": 0.2066, "step": 3049 }, { "epoch": 0.46065549010723456, "grad_norm": 2.8058847921553327, "learning_rate": 9.429426756362298e-06, "loss": 0.1629, "step": 3050 }, { "epoch": 0.46080652469415495, "grad_norm": 2.436133419784794, "learning_rate": 9.429059838414586e-06, "loss": 0.1529, "step": 3051 }, { "epoch": 0.46095755928107535, "grad_norm": 2.715249046524794, "learning_rate": 9.428692809671506e-06, "loss": 0.1674, "step": 3052 }, { "epoch": 0.4611085938679958, "grad_norm": 2.0390833989570765, "learning_rate": 9.428325670142238e-06, "loss": 0.1443, "step": 3053 }, { "epoch": 0.4612596284549162, "grad_norm": 2.3497896153660403, "learning_rate": 9.427958419835967e-06, "loss": 0.1661, "step": 3054 }, { "epoch": 0.4614106630418366, "grad_norm": 2.5564044347557626, "learning_rate": 9.42759105876188e-06, "loss": 0.1758, "step": 3055 }, { "epoch": 0.461561697628757, "grad_norm": 2.3946776866005584, "learning_rate": 9.427223586929168e-06, "loss": 0.1653, "step": 3056 }, { "epoch": 0.4617127322156774, "grad_norm": 2.830388188935523, "learning_rate": 9.42685600434702e-06, "loss": 0.1811, "step": 3057 }, { "epoch": 0.4618637668025978, "grad_norm": 1.67231131804339, "learning_rate": 9.426488311024636e-06, "loss": 0.115, "step": 3058 }, { "epoch": 0.4620148013895182, "grad_norm": 2.0635780095102834, "learning_rate": 9.426120506971209e-06, "loss": 0.167, "step": 3059 }, { "epoch": 0.4621658359764386, "grad_norm": 2.1280797185466347, "learning_rate": 9.425752592195942e-06, "loss": 0.1486, "step": 3060 }, { "epoch": 0.462316870563359, "grad_norm": 3.000344182158473, "learning_rate": 9.42538456670804e-06, "loss": 0.166, "step": 3061 }, { "epoch": 0.4624679051502794, "grad_norm": 2.333104105292057, "learning_rate": 9.425016430516709e-06, "loss": 0.1677, "step": 3062 }, { "epoch": 0.4626189397371998, "grad_norm": 2.302371707581049, "learning_rate": 9.424648183631155e-06, "loss": 0.152, "step": 3063 }, { "epoch": 0.4627699743241202, "grad_norm": 2.2866166908238172, "learning_rate": 9.424279826060594e-06, "loss": 0.157, "step": 3064 }, { "epoch": 0.46292100891104065, "grad_norm": 2.2494196827501294, "learning_rate": 9.42391135781424e-06, "loss": 0.1752, "step": 3065 }, { "epoch": 0.46307204349796105, "grad_norm": 2.9477822749011215, "learning_rate": 9.423542778901307e-06, "loss": 0.193, "step": 3066 }, { "epoch": 0.46322307808488145, "grad_norm": 1.9845941678487613, "learning_rate": 9.423174089331019e-06, "loss": 0.1344, "step": 3067 }, { "epoch": 0.46337411267180184, "grad_norm": 2.5897715382620614, "learning_rate": 9.422805289112598e-06, "loss": 0.1607, "step": 3068 }, { "epoch": 0.46352514725872224, "grad_norm": 2.4740842776033247, "learning_rate": 9.422436378255268e-06, "loss": 0.1875, "step": 3069 }, { "epoch": 0.46367618184564263, "grad_norm": 2.0956397244645637, "learning_rate": 9.422067356768259e-06, "loss": 0.1554, "step": 3070 }, { "epoch": 0.4638272164325631, "grad_norm": 2.3987437556793174, "learning_rate": 9.421698224660803e-06, "loss": 0.1573, "step": 3071 }, { "epoch": 0.4639782510194835, "grad_norm": 2.394508645993544, "learning_rate": 9.421328981942132e-06, "loss": 0.1438, "step": 3072 }, { "epoch": 0.4641292856064039, "grad_norm": 2.3286688694991033, "learning_rate": 9.420959628621485e-06, "loss": 0.1701, "step": 3073 }, { "epoch": 0.46428032019332427, "grad_norm": 2.254928983959413, "learning_rate": 9.4205901647081e-06, "loss": 0.1748, "step": 3074 }, { "epoch": 0.46443135478024467, "grad_norm": 3.454065242134329, "learning_rate": 9.420220590211219e-06, "loss": 0.1785, "step": 3075 }, { "epoch": 0.46458238936716506, "grad_norm": 2.269391876352089, "learning_rate": 9.41985090514009e-06, "loss": 0.1569, "step": 3076 }, { "epoch": 0.4647334239540855, "grad_norm": 2.573756803255579, "learning_rate": 9.419481109503957e-06, "loss": 0.168, "step": 3077 }, { "epoch": 0.4648844585410059, "grad_norm": 2.803430810038323, "learning_rate": 9.419111203312071e-06, "loss": 0.153, "step": 3078 }, { "epoch": 0.4650354931279263, "grad_norm": 2.2503887041294823, "learning_rate": 9.41874118657369e-06, "loss": 0.1509, "step": 3079 }, { "epoch": 0.4651865277148467, "grad_norm": 2.2744840656632084, "learning_rate": 9.418371059298066e-06, "loss": 0.1439, "step": 3080 }, { "epoch": 0.4653375623017671, "grad_norm": 2.306244965773245, "learning_rate": 9.418000821494457e-06, "loss": 0.1609, "step": 3081 }, { "epoch": 0.4654885968886875, "grad_norm": 2.3380769685567073, "learning_rate": 9.417630473172128e-06, "loss": 0.1753, "step": 3082 }, { "epoch": 0.46563963147560794, "grad_norm": 2.0738613126624994, "learning_rate": 9.417260014340343e-06, "loss": 0.1547, "step": 3083 }, { "epoch": 0.46579066606252834, "grad_norm": 2.84555281860221, "learning_rate": 9.416889445008369e-06, "loss": 0.1789, "step": 3084 }, { "epoch": 0.46594170064944873, "grad_norm": 2.221979312798778, "learning_rate": 9.416518765185475e-06, "loss": 0.1727, "step": 3085 }, { "epoch": 0.4660927352363691, "grad_norm": 2.449018986435257, "learning_rate": 9.416147974880932e-06, "loss": 0.158, "step": 3086 }, { "epoch": 0.4662437698232895, "grad_norm": 2.098476043899509, "learning_rate": 9.41577707410402e-06, "loss": 0.1566, "step": 3087 }, { "epoch": 0.4663948044102099, "grad_norm": 2.6980250376880144, "learning_rate": 9.415406062864015e-06, "loss": 0.1822, "step": 3088 }, { "epoch": 0.46654583899713037, "grad_norm": 1.8410291781170405, "learning_rate": 9.415034941170195e-06, "loss": 0.1529, "step": 3089 }, { "epoch": 0.46669687358405076, "grad_norm": 2.2832594470107095, "learning_rate": 9.414663709031851e-06, "loss": 0.1392, "step": 3090 }, { "epoch": 0.46684790817097116, "grad_norm": 2.7115917590853327, "learning_rate": 9.414292366458261e-06, "loss": 0.165, "step": 3091 }, { "epoch": 0.46699894275789156, "grad_norm": 3.3721443583268482, "learning_rate": 9.413920913458722e-06, "loss": 0.1438, "step": 3092 }, { "epoch": 0.46714997734481195, "grad_norm": 3.4968071746877993, "learning_rate": 9.413549350042521e-06, "loss": 0.2136, "step": 3093 }, { "epoch": 0.46730101193173235, "grad_norm": 2.7373731603448412, "learning_rate": 9.413177676218957e-06, "loss": 0.1681, "step": 3094 }, { "epoch": 0.4674520465186528, "grad_norm": 2.3324669059916405, "learning_rate": 9.412805891997323e-06, "loss": 0.1639, "step": 3095 }, { "epoch": 0.4676030811055732, "grad_norm": 2.2228265922444805, "learning_rate": 9.412433997386923e-06, "loss": 0.1567, "step": 3096 }, { "epoch": 0.4677541156924936, "grad_norm": 2.6855273236096813, "learning_rate": 9.412061992397058e-06, "loss": 0.1369, "step": 3097 }, { "epoch": 0.467905150279414, "grad_norm": 2.756877509679327, "learning_rate": 9.411689877037036e-06, "loss": 0.1775, "step": 3098 }, { "epoch": 0.4680561848663344, "grad_norm": 2.534442800420768, "learning_rate": 9.411317651316163e-06, "loss": 0.1622, "step": 3099 }, { "epoch": 0.4682072194532548, "grad_norm": 1.8517077993777917, "learning_rate": 9.410945315243752e-06, "loss": 0.1482, "step": 3100 }, { "epoch": 0.4683582540401752, "grad_norm": 4.8068347095758295, "learning_rate": 9.410572868829117e-06, "loss": 0.2319, "step": 3101 }, { "epoch": 0.4685092886270956, "grad_norm": 1.9971386398334476, "learning_rate": 9.410200312081576e-06, "loss": 0.141, "step": 3102 }, { "epoch": 0.468660323214016, "grad_norm": 3.283020700824815, "learning_rate": 9.409827645010448e-06, "loss": 0.1693, "step": 3103 }, { "epoch": 0.4688113578009364, "grad_norm": 2.7369461391419314, "learning_rate": 9.409454867625055e-06, "loss": 0.1806, "step": 3104 }, { "epoch": 0.4689623923878568, "grad_norm": 2.3014529445596286, "learning_rate": 9.40908197993472e-06, "loss": 0.1792, "step": 3105 }, { "epoch": 0.4691134269747772, "grad_norm": 2.3587960382530513, "learning_rate": 9.408708981948776e-06, "loss": 0.1609, "step": 3106 }, { "epoch": 0.46926446156169765, "grad_norm": 2.292227762860851, "learning_rate": 9.40833587367655e-06, "loss": 0.1525, "step": 3107 }, { "epoch": 0.46941549614861805, "grad_norm": 2.5076782783475555, "learning_rate": 9.407962655127377e-06, "loss": 0.1664, "step": 3108 }, { "epoch": 0.46956653073553845, "grad_norm": 2.526789609870176, "learning_rate": 9.407589326310593e-06, "loss": 0.1696, "step": 3109 }, { "epoch": 0.46971756532245884, "grad_norm": 2.4651336162732203, "learning_rate": 9.407215887235536e-06, "loss": 0.1866, "step": 3110 }, { "epoch": 0.46986859990937924, "grad_norm": 3.0518641755269633, "learning_rate": 9.40684233791155e-06, "loss": 0.1836, "step": 3111 }, { "epoch": 0.47001963449629963, "grad_norm": 2.585885150198552, "learning_rate": 9.406468678347977e-06, "loss": 0.1365, "step": 3112 }, { "epoch": 0.4701706690832201, "grad_norm": 2.1260644206791555, "learning_rate": 9.406094908554168e-06, "loss": 0.1492, "step": 3113 }, { "epoch": 0.4703217036701405, "grad_norm": 2.883239435471735, "learning_rate": 9.40572102853947e-06, "loss": 0.169, "step": 3114 }, { "epoch": 0.4704727382570609, "grad_norm": 2.5946425589307145, "learning_rate": 9.405347038313235e-06, "loss": 0.1473, "step": 3115 }, { "epoch": 0.47062377284398127, "grad_norm": 4.741795488689868, "learning_rate": 9.40497293788482e-06, "loss": 0.1988, "step": 3116 }, { "epoch": 0.47077480743090167, "grad_norm": 2.7778593086219776, "learning_rate": 9.404598727263585e-06, "loss": 0.1829, "step": 3117 }, { "epoch": 0.47092584201782206, "grad_norm": 2.14565021182244, "learning_rate": 9.404224406458888e-06, "loss": 0.1502, "step": 3118 }, { "epoch": 0.4710768766047425, "grad_norm": 2.3060097807529214, "learning_rate": 9.403849975480095e-06, "loss": 0.1655, "step": 3119 }, { "epoch": 0.4712279111916629, "grad_norm": 2.097904443936051, "learning_rate": 9.403475434336571e-06, "loss": 0.1599, "step": 3120 }, { "epoch": 0.4713789457785833, "grad_norm": 1.6968919443279737, "learning_rate": 9.403100783037687e-06, "loss": 0.1272, "step": 3121 }, { "epoch": 0.4715299803655037, "grad_norm": 2.403466944381608, "learning_rate": 9.402726021592815e-06, "loss": 0.1507, "step": 3122 }, { "epoch": 0.4716810149524241, "grad_norm": 3.0719451543228926, "learning_rate": 9.402351150011328e-06, "loss": 0.1795, "step": 3123 }, { "epoch": 0.4718320495393445, "grad_norm": 3.2121470474327882, "learning_rate": 9.401976168302604e-06, "loss": 0.1744, "step": 3124 }, { "epoch": 0.47198308412626494, "grad_norm": 2.2172573942706295, "learning_rate": 9.401601076476026e-06, "loss": 0.1692, "step": 3125 }, { "epoch": 0.47213411871318534, "grad_norm": 1.5775201369840752, "learning_rate": 9.401225874540975e-06, "loss": 0.1125, "step": 3126 }, { "epoch": 0.47228515330010573, "grad_norm": 2.3563004250769537, "learning_rate": 9.400850562506837e-06, "loss": 0.1608, "step": 3127 }, { "epoch": 0.4724361878870261, "grad_norm": 2.307165259304035, "learning_rate": 9.400475140383001e-06, "loss": 0.1613, "step": 3128 }, { "epoch": 0.4725872224739465, "grad_norm": 2.1265974021548715, "learning_rate": 9.400099608178859e-06, "loss": 0.1438, "step": 3129 }, { "epoch": 0.4727382570608669, "grad_norm": 2.9352087883372175, "learning_rate": 9.399723965903802e-06, "loss": 0.152, "step": 3130 }, { "epoch": 0.47288929164778737, "grad_norm": 2.181365888984092, "learning_rate": 9.399348213567231e-06, "loss": 0.1607, "step": 3131 }, { "epoch": 0.47304032623470776, "grad_norm": 2.3070497231117875, "learning_rate": 9.398972351178543e-06, "loss": 0.1749, "step": 3132 }, { "epoch": 0.47319136082162816, "grad_norm": 2.8059036811784623, "learning_rate": 9.398596378747143e-06, "loss": 0.1718, "step": 3133 }, { "epoch": 0.47334239540854856, "grad_norm": 2.504069166674323, "learning_rate": 9.398220296282434e-06, "loss": 0.1598, "step": 3134 }, { "epoch": 0.47349342999546895, "grad_norm": 2.1202597308802464, "learning_rate": 9.397844103793824e-06, "loss": 0.1536, "step": 3135 }, { "epoch": 0.47364446458238935, "grad_norm": 2.1169877876929046, "learning_rate": 9.397467801290723e-06, "loss": 0.1429, "step": 3136 }, { "epoch": 0.4737954991693098, "grad_norm": 2.075422503308984, "learning_rate": 9.397091388782547e-06, "loss": 0.173, "step": 3137 }, { "epoch": 0.4739465337562302, "grad_norm": 2.906946366917232, "learning_rate": 9.396714866278709e-06, "loss": 0.168, "step": 3138 }, { "epoch": 0.4740975683431506, "grad_norm": 2.0857754476433747, "learning_rate": 9.396338233788631e-06, "loss": 0.1474, "step": 3139 }, { "epoch": 0.474248602930071, "grad_norm": 2.5971686597827084, "learning_rate": 9.395961491321732e-06, "loss": 0.1846, "step": 3140 }, { "epoch": 0.4743996375169914, "grad_norm": 2.064485610098695, "learning_rate": 9.39558463888744e-06, "loss": 0.1501, "step": 3141 }, { "epoch": 0.4745506721039118, "grad_norm": 2.140965289041653, "learning_rate": 9.395207676495178e-06, "loss": 0.1411, "step": 3142 }, { "epoch": 0.4747017066908322, "grad_norm": 3.1034895302108128, "learning_rate": 9.394830604154377e-06, "loss": 0.1796, "step": 3143 }, { "epoch": 0.4748527412777526, "grad_norm": 2.2066719959103374, "learning_rate": 9.394453421874473e-06, "loss": 0.1433, "step": 3144 }, { "epoch": 0.475003775864673, "grad_norm": 2.8574330346977175, "learning_rate": 9.394076129664896e-06, "loss": 0.1893, "step": 3145 }, { "epoch": 0.4751548104515934, "grad_norm": 2.199334832708802, "learning_rate": 9.393698727535088e-06, "loss": 0.1756, "step": 3146 }, { "epoch": 0.4753058450385138, "grad_norm": 2.535334784914728, "learning_rate": 9.39332121549449e-06, "loss": 0.1687, "step": 3147 }, { "epoch": 0.4754568796254342, "grad_norm": 6.574134923077675, "learning_rate": 9.392943593552545e-06, "loss": 0.164, "step": 3148 }, { "epoch": 0.47560791421235465, "grad_norm": 2.5640937746336636, "learning_rate": 9.392565861718696e-06, "loss": 0.1669, "step": 3149 }, { "epoch": 0.47575894879927505, "grad_norm": 2.4558341374779027, "learning_rate": 9.392188020002397e-06, "loss": 0.1826, "step": 3150 }, { "epoch": 0.47590998338619545, "grad_norm": 2.1014144902897587, "learning_rate": 9.391810068413099e-06, "loss": 0.1505, "step": 3151 }, { "epoch": 0.47606101797311584, "grad_norm": 3.0520351441118896, "learning_rate": 9.391432006960255e-06, "loss": 0.1635, "step": 3152 }, { "epoch": 0.47621205256003624, "grad_norm": 2.2399038893298107, "learning_rate": 9.391053835653323e-06, "loss": 0.1564, "step": 3153 }, { "epoch": 0.47636308714695663, "grad_norm": 2.207836055447323, "learning_rate": 9.390675554501762e-06, "loss": 0.1525, "step": 3154 }, { "epoch": 0.4765141217338771, "grad_norm": 3.185897825615928, "learning_rate": 9.390297163515039e-06, "loss": 0.1571, "step": 3155 }, { "epoch": 0.4766651563207975, "grad_norm": 1.8117537526044123, "learning_rate": 9.389918662702615e-06, "loss": 0.1322, "step": 3156 }, { "epoch": 0.4768161909077179, "grad_norm": 5.020964556400165, "learning_rate": 9.389540052073963e-06, "loss": 0.1784, "step": 3157 }, { "epoch": 0.47696722549463827, "grad_norm": 2.174511420998066, "learning_rate": 9.389161331638547e-06, "loss": 0.1671, "step": 3158 }, { "epoch": 0.47711826008155866, "grad_norm": 2.2392255361511744, "learning_rate": 9.388782501405849e-06, "loss": 0.1551, "step": 3159 }, { "epoch": 0.47726929466847906, "grad_norm": 2.223952391890241, "learning_rate": 9.388403561385341e-06, "loss": 0.157, "step": 3160 }, { "epoch": 0.4774203292553995, "grad_norm": 2.229435096087771, "learning_rate": 9.388024511586504e-06, "loss": 0.153, "step": 3161 }, { "epoch": 0.4775713638423199, "grad_norm": 1.9824885440373106, "learning_rate": 9.387645352018818e-06, "loss": 0.1477, "step": 3162 }, { "epoch": 0.4777223984292403, "grad_norm": 2.4389904721016467, "learning_rate": 9.38726608269177e-06, "loss": 0.145, "step": 3163 }, { "epoch": 0.4778734330161607, "grad_norm": 3.3154724584921245, "learning_rate": 9.386886703614848e-06, "loss": 0.1794, "step": 3164 }, { "epoch": 0.4780244676030811, "grad_norm": 2.746738526751078, "learning_rate": 9.386507214797542e-06, "loss": 0.1809, "step": 3165 }, { "epoch": 0.4781755021900015, "grad_norm": 2.677421472161293, "learning_rate": 9.386127616249343e-06, "loss": 0.1747, "step": 3166 }, { "epoch": 0.47832653677692194, "grad_norm": 2.2241835842189412, "learning_rate": 9.385747907979749e-06, "loss": 0.17, "step": 3167 }, { "epoch": 0.47847757136384234, "grad_norm": 2.3262508493604153, "learning_rate": 9.385368089998258e-06, "loss": 0.1647, "step": 3168 }, { "epoch": 0.47862860595076273, "grad_norm": 2.752337941749457, "learning_rate": 9.38498816231437e-06, "loss": 0.1757, "step": 3169 }, { "epoch": 0.4787796405376831, "grad_norm": 2.126063623469065, "learning_rate": 9.384608124937594e-06, "loss": 0.1632, "step": 3170 }, { "epoch": 0.4789306751246035, "grad_norm": 2.310846645674285, "learning_rate": 9.38422797787743e-06, "loss": 0.1546, "step": 3171 }, { "epoch": 0.4790817097115239, "grad_norm": 2.3126848669359497, "learning_rate": 9.383847721143394e-06, "loss": 0.1705, "step": 3172 }, { "epoch": 0.47923274429844437, "grad_norm": 2.6946034105622196, "learning_rate": 9.383467354744992e-06, "loss": 0.1829, "step": 3173 }, { "epoch": 0.47938377888536476, "grad_norm": 2.728172218839751, "learning_rate": 9.383086878691742e-06, "loss": 0.1628, "step": 3174 }, { "epoch": 0.47953481347228516, "grad_norm": 2.283188014247287, "learning_rate": 9.382706292993166e-06, "loss": 0.1587, "step": 3175 }, { "epoch": 0.47968584805920556, "grad_norm": 2.0922743806136213, "learning_rate": 9.382325597658779e-06, "loss": 0.1518, "step": 3176 }, { "epoch": 0.47983688264612595, "grad_norm": 2.5249758719840854, "learning_rate": 9.381944792698106e-06, "loss": 0.1585, "step": 3177 }, { "epoch": 0.47998791723304635, "grad_norm": 2.6545833617554053, "learning_rate": 9.381563878120671e-06, "loss": 0.1706, "step": 3178 }, { "epoch": 0.4801389518199668, "grad_norm": 2.3639267565740205, "learning_rate": 9.381182853936008e-06, "loss": 0.1658, "step": 3179 }, { "epoch": 0.4802899864068872, "grad_norm": 2.424657881498885, "learning_rate": 9.380801720153644e-06, "loss": 0.172, "step": 3180 }, { "epoch": 0.4804410209938076, "grad_norm": 1.9517421502032815, "learning_rate": 9.380420476783113e-06, "loss": 0.1602, "step": 3181 }, { "epoch": 0.480592055580728, "grad_norm": 2.551245415307415, "learning_rate": 9.380039123833957e-06, "loss": 0.1792, "step": 3182 }, { "epoch": 0.4807430901676484, "grad_norm": 2.375731137060149, "learning_rate": 9.37965766131571e-06, "loss": 0.188, "step": 3183 }, { "epoch": 0.4808941247545688, "grad_norm": 2.4529385726230846, "learning_rate": 9.379276089237919e-06, "loss": 0.1779, "step": 3184 }, { "epoch": 0.4810451593414892, "grad_norm": 3.2190022435489043, "learning_rate": 9.378894407610126e-06, "loss": 0.1868, "step": 3185 }, { "epoch": 0.4811961939284096, "grad_norm": 2.679680190309039, "learning_rate": 9.37851261644188e-06, "loss": 0.17, "step": 3186 }, { "epoch": 0.48134722851533, "grad_norm": 2.799735342855553, "learning_rate": 9.378130715742733e-06, "loss": 0.1706, "step": 3187 }, { "epoch": 0.4814982631022504, "grad_norm": 2.3815866338052363, "learning_rate": 9.377748705522236e-06, "loss": 0.1567, "step": 3188 }, { "epoch": 0.4816492976891708, "grad_norm": 2.4521558501084697, "learning_rate": 9.377366585789946e-06, "loss": 0.1666, "step": 3189 }, { "epoch": 0.4818003322760912, "grad_norm": 2.68885324563275, "learning_rate": 9.376984356555424e-06, "loss": 0.1622, "step": 3190 }, { "epoch": 0.48195136686301165, "grad_norm": 2.2656472782943906, "learning_rate": 9.376602017828229e-06, "loss": 0.1766, "step": 3191 }, { "epoch": 0.48210240144993205, "grad_norm": 2.2160171376634925, "learning_rate": 9.376219569617928e-06, "loss": 0.1429, "step": 3192 }, { "epoch": 0.48225343603685245, "grad_norm": 2.3267134705227455, "learning_rate": 9.375837011934083e-06, "loss": 0.1663, "step": 3193 }, { "epoch": 0.48240447062377284, "grad_norm": 1.9345057340790175, "learning_rate": 9.375454344786272e-06, "loss": 0.146, "step": 3194 }, { "epoch": 0.48255550521069324, "grad_norm": 2.614537517539277, "learning_rate": 9.375071568184062e-06, "loss": 0.1602, "step": 3195 }, { "epoch": 0.48270653979761363, "grad_norm": 2.558136759541689, "learning_rate": 9.374688682137028e-06, "loss": 0.1832, "step": 3196 }, { "epoch": 0.4828575743845341, "grad_norm": 4.103584085911953, "learning_rate": 9.37430568665475e-06, "loss": 0.22, "step": 3197 }, { "epoch": 0.4830086089714545, "grad_norm": 2.40872053395393, "learning_rate": 9.37392258174681e-06, "loss": 0.1576, "step": 3198 }, { "epoch": 0.4831596435583749, "grad_norm": 7.3198225839898985, "learning_rate": 9.373539367422788e-06, "loss": 0.1962, "step": 3199 }, { "epoch": 0.48331067814529527, "grad_norm": 3.2286695251504494, "learning_rate": 9.373156043692273e-06, "loss": 0.1718, "step": 3200 }, { "epoch": 0.48346171273221566, "grad_norm": 2.355650570376001, "learning_rate": 9.372772610564855e-06, "loss": 0.1675, "step": 3201 }, { "epoch": 0.48361274731913606, "grad_norm": 2.324977101686656, "learning_rate": 9.372389068050122e-06, "loss": 0.1652, "step": 3202 }, { "epoch": 0.4837637819060565, "grad_norm": 2.1786772949627293, "learning_rate": 9.372005416157675e-06, "loss": 0.1715, "step": 3203 }, { "epoch": 0.4839148164929769, "grad_norm": 2.862421927604699, "learning_rate": 9.371621654897102e-06, "loss": 0.1651, "step": 3204 }, { "epoch": 0.4840658510798973, "grad_norm": 2.1599651956102233, "learning_rate": 9.371237784278011e-06, "loss": 0.1455, "step": 3205 }, { "epoch": 0.4842168856668177, "grad_norm": 2.4618686106414525, "learning_rate": 9.370853804310001e-06, "loss": 0.1687, "step": 3206 }, { "epoch": 0.4843679202537381, "grad_norm": 2.681249727467442, "learning_rate": 9.370469715002679e-06, "loss": 0.1627, "step": 3207 }, { "epoch": 0.4845189548406585, "grad_norm": 2.272885988286878, "learning_rate": 9.370085516365653e-06, "loss": 0.1543, "step": 3208 }, { "epoch": 0.48466998942757894, "grad_norm": 2.257465114803614, "learning_rate": 9.36970120840853e-06, "loss": 0.1579, "step": 3209 }, { "epoch": 0.48482102401449934, "grad_norm": 3.1340114219707047, "learning_rate": 9.36931679114093e-06, "loss": 0.171, "step": 3210 }, { "epoch": 0.48497205860141973, "grad_norm": 2.373601624557538, "learning_rate": 9.368932264572466e-06, "loss": 0.1755, "step": 3211 }, { "epoch": 0.4851230931883401, "grad_norm": 2.5386895643811314, "learning_rate": 9.368547628712757e-06, "loss": 0.1612, "step": 3212 }, { "epoch": 0.4852741277752605, "grad_norm": 3.60617874974608, "learning_rate": 9.368162883571425e-06, "loss": 0.1537, "step": 3213 }, { "epoch": 0.4854251623621809, "grad_norm": 2.358095378290024, "learning_rate": 9.367778029158096e-06, "loss": 0.1643, "step": 3214 }, { "epoch": 0.48557619694910137, "grad_norm": 2.3758518034109897, "learning_rate": 9.367393065482398e-06, "loss": 0.1841, "step": 3215 }, { "epoch": 0.48572723153602176, "grad_norm": 2.3493696996885416, "learning_rate": 9.367007992553957e-06, "loss": 0.1578, "step": 3216 }, { "epoch": 0.48587826612294216, "grad_norm": 2.53752634429946, "learning_rate": 9.366622810382409e-06, "loss": 0.1692, "step": 3217 }, { "epoch": 0.48602930070986256, "grad_norm": 2.3595204913434062, "learning_rate": 9.36623751897739e-06, "loss": 0.1424, "step": 3218 }, { "epoch": 0.48618033529678295, "grad_norm": 2.598806738492424, "learning_rate": 9.365852118348536e-06, "loss": 0.1604, "step": 3219 }, { "epoch": 0.48633136988370335, "grad_norm": 2.368909049311771, "learning_rate": 9.36546660850549e-06, "loss": 0.1541, "step": 3220 }, { "epoch": 0.4864824044706238, "grad_norm": 1.9284077757992897, "learning_rate": 9.365080989457895e-06, "loss": 0.1444, "step": 3221 }, { "epoch": 0.4866334390575442, "grad_norm": 1.8926962095800968, "learning_rate": 9.364695261215396e-06, "loss": 0.16, "step": 3222 }, { "epoch": 0.4867844736444646, "grad_norm": 2.0583911934156722, "learning_rate": 9.364309423787645e-06, "loss": 0.1317, "step": 3223 }, { "epoch": 0.486935508231385, "grad_norm": 2.4262880022272677, "learning_rate": 9.363923477184291e-06, "loss": 0.1786, "step": 3224 }, { "epoch": 0.4870865428183054, "grad_norm": 2.6744313173032137, "learning_rate": 9.363537421414992e-06, "loss": 0.164, "step": 3225 }, { "epoch": 0.4872375774052258, "grad_norm": 3.207337750228931, "learning_rate": 9.363151256489403e-06, "loss": 0.1636, "step": 3226 }, { "epoch": 0.4873886119921462, "grad_norm": 2.299451898463178, "learning_rate": 9.362764982417186e-06, "loss": 0.1656, "step": 3227 }, { "epoch": 0.4875396465790666, "grad_norm": 2.596280127751634, "learning_rate": 9.362378599208e-06, "loss": 0.1665, "step": 3228 }, { "epoch": 0.487690681165987, "grad_norm": 2.428632955409802, "learning_rate": 9.361992106871513e-06, "loss": 0.1712, "step": 3229 }, { "epoch": 0.4878417157529074, "grad_norm": 2.290358669287138, "learning_rate": 9.361605505417396e-06, "loss": 0.152, "step": 3230 }, { "epoch": 0.4879927503398278, "grad_norm": 1.9923523029724117, "learning_rate": 9.361218794855317e-06, "loss": 0.1562, "step": 3231 }, { "epoch": 0.4881437849267482, "grad_norm": 3.0642111356999324, "learning_rate": 9.360831975194948e-06, "loss": 0.1706, "step": 3232 }, { "epoch": 0.48829481951366865, "grad_norm": 2.6061291090828655, "learning_rate": 9.36044504644597e-06, "loss": 0.1707, "step": 3233 }, { "epoch": 0.48844585410058905, "grad_norm": 2.3864478026594815, "learning_rate": 9.360058008618061e-06, "loss": 0.1626, "step": 3234 }, { "epoch": 0.48859688868750945, "grad_norm": 2.1178351160922957, "learning_rate": 9.3596708617209e-06, "loss": 0.1589, "step": 3235 }, { "epoch": 0.48874792327442984, "grad_norm": 2.1025125141754866, "learning_rate": 9.359283605764173e-06, "loss": 0.1635, "step": 3236 }, { "epoch": 0.48889895786135024, "grad_norm": 2.518256970813844, "learning_rate": 9.35889624075757e-06, "loss": 0.1907, "step": 3237 }, { "epoch": 0.48904999244827063, "grad_norm": 2.3545266323156966, "learning_rate": 9.358508766710778e-06, "loss": 0.1421, "step": 3238 }, { "epoch": 0.4892010270351911, "grad_norm": 2.7733308960728587, "learning_rate": 9.358121183633492e-06, "loss": 0.1737, "step": 3239 }, { "epoch": 0.4893520616221115, "grad_norm": 3.313112441828787, "learning_rate": 9.357733491535409e-06, "loss": 0.1991, "step": 3240 }, { "epoch": 0.4895030962090319, "grad_norm": 1.7239988276729303, "learning_rate": 9.357345690426222e-06, "loss": 0.1275, "step": 3241 }, { "epoch": 0.48965413079595227, "grad_norm": 2.53602473121975, "learning_rate": 9.356957780315636e-06, "loss": 0.1604, "step": 3242 }, { "epoch": 0.48980516538287266, "grad_norm": 2.568533114461194, "learning_rate": 9.356569761213353e-06, "loss": 0.2048, "step": 3243 }, { "epoch": 0.48995619996979306, "grad_norm": 2.5408983269989727, "learning_rate": 9.356181633129083e-06, "loss": 0.1809, "step": 3244 }, { "epoch": 0.4901072345567135, "grad_norm": 1.957075744328577, "learning_rate": 9.35579339607253e-06, "loss": 0.1444, "step": 3245 }, { "epoch": 0.4902582691436339, "grad_norm": 2.1331841726184333, "learning_rate": 9.35540505005341e-06, "loss": 0.1557, "step": 3246 }, { "epoch": 0.4904093037305543, "grad_norm": 1.8317577724872545, "learning_rate": 9.355016595081435e-06, "loss": 0.1197, "step": 3247 }, { "epoch": 0.4905603383174747, "grad_norm": 1.9573236042793656, "learning_rate": 9.354628031166325e-06, "loss": 0.1384, "step": 3248 }, { "epoch": 0.4907113729043951, "grad_norm": 2.1706278742797727, "learning_rate": 9.354239358317797e-06, "loss": 0.1461, "step": 3249 }, { "epoch": 0.4908624074913155, "grad_norm": 2.849966783162814, "learning_rate": 9.35385057654558e-06, "loss": 0.1688, "step": 3250 }, { "epoch": 0.49101344207823594, "grad_norm": 2.6713108028370303, "learning_rate": 9.353461685859392e-06, "loss": 0.1626, "step": 3251 }, { "epoch": 0.49116447666515634, "grad_norm": 2.628772133607984, "learning_rate": 9.353072686268964e-06, "loss": 0.1705, "step": 3252 }, { "epoch": 0.49131551125207673, "grad_norm": 2.8154845493096414, "learning_rate": 9.35268357778403e-06, "loss": 0.193, "step": 3253 }, { "epoch": 0.4914665458389971, "grad_norm": 2.0386475209922743, "learning_rate": 9.352294360414322e-06, "loss": 0.1377, "step": 3254 }, { "epoch": 0.4916175804259175, "grad_norm": 2.4438968477970846, "learning_rate": 9.351905034169573e-06, "loss": 0.1373, "step": 3255 }, { "epoch": 0.4917686150128379, "grad_norm": 2.224969270812923, "learning_rate": 9.351515599059527e-06, "loss": 0.1678, "step": 3256 }, { "epoch": 0.49191964959975837, "grad_norm": 2.2844723744619633, "learning_rate": 9.351126055093924e-06, "loss": 0.159, "step": 3257 }, { "epoch": 0.49207068418667876, "grad_norm": 2.1391330744772086, "learning_rate": 9.350736402282508e-06, "loss": 0.1634, "step": 3258 }, { "epoch": 0.49222171877359916, "grad_norm": 2.2565899614098917, "learning_rate": 9.350346640635028e-06, "loss": 0.1771, "step": 3259 }, { "epoch": 0.49237275336051955, "grad_norm": 1.993621714923478, "learning_rate": 9.349956770161234e-06, "loss": 0.1402, "step": 3260 }, { "epoch": 0.49252378794743995, "grad_norm": 3.186764295359229, "learning_rate": 9.349566790870876e-06, "loss": 0.2102, "step": 3261 }, { "epoch": 0.49267482253436035, "grad_norm": 2.061093451217089, "learning_rate": 9.349176702773714e-06, "loss": 0.1429, "step": 3262 }, { "epoch": 0.4928258571212808, "grad_norm": 2.0650097529899125, "learning_rate": 9.348786505879502e-06, "loss": 0.1362, "step": 3263 }, { "epoch": 0.4929768917082012, "grad_norm": 7.267566058541256, "learning_rate": 9.348396200198001e-06, "loss": 0.1683, "step": 3264 }, { "epoch": 0.4931279262951216, "grad_norm": 2.205210092350434, "learning_rate": 9.34800578573898e-06, "loss": 0.1466, "step": 3265 }, { "epoch": 0.493278960882042, "grad_norm": 2.284291916676737, "learning_rate": 9.347615262512201e-06, "loss": 0.165, "step": 3266 }, { "epoch": 0.4934299954689624, "grad_norm": 2.2185365439720197, "learning_rate": 9.347224630527435e-06, "loss": 0.1613, "step": 3267 }, { "epoch": 0.4935810300558828, "grad_norm": 2.848650175274451, "learning_rate": 9.346833889794452e-06, "loss": 0.2052, "step": 3268 }, { "epoch": 0.4937320646428032, "grad_norm": 2.5576034332391147, "learning_rate": 9.346443040323027e-06, "loss": 0.1726, "step": 3269 }, { "epoch": 0.4938830992297236, "grad_norm": 2.257792318199067, "learning_rate": 9.34605208212294e-06, "loss": 0.1437, "step": 3270 }, { "epoch": 0.494034133816644, "grad_norm": 2.979929020306279, "learning_rate": 9.345661015203968e-06, "loss": 0.1854, "step": 3271 }, { "epoch": 0.4941851684035644, "grad_norm": 5.027565059933791, "learning_rate": 9.345269839575895e-06, "loss": 0.2123, "step": 3272 }, { "epoch": 0.4943362029904848, "grad_norm": 2.1340270855589987, "learning_rate": 9.344878555248504e-06, "loss": 0.1577, "step": 3273 }, { "epoch": 0.4944872375774052, "grad_norm": 2.0522882482009672, "learning_rate": 9.344487162231588e-06, "loss": 0.1494, "step": 3274 }, { "epoch": 0.49463827216432565, "grad_norm": 2.3971731426118628, "learning_rate": 9.344095660534934e-06, "loss": 0.1756, "step": 3275 }, { "epoch": 0.49478930675124605, "grad_norm": 2.155520816298284, "learning_rate": 9.343704050168338e-06, "loss": 0.1615, "step": 3276 }, { "epoch": 0.49494034133816645, "grad_norm": 2.1390905269244493, "learning_rate": 9.343312331141593e-06, "loss": 0.1647, "step": 3277 }, { "epoch": 0.49509137592508684, "grad_norm": 2.8286468699986607, "learning_rate": 9.342920503464503e-06, "loss": 0.1936, "step": 3278 }, { "epoch": 0.49524241051200724, "grad_norm": 2.7436777818882043, "learning_rate": 9.342528567146867e-06, "loss": 0.1889, "step": 3279 }, { "epoch": 0.49539344509892763, "grad_norm": 2.438653223396555, "learning_rate": 9.342136522198489e-06, "loss": 0.1908, "step": 3280 }, { "epoch": 0.4955444796858481, "grad_norm": 2.040925730760593, "learning_rate": 9.341744368629175e-06, "loss": 0.1493, "step": 3281 }, { "epoch": 0.4956955142727685, "grad_norm": 2.7468246471385624, "learning_rate": 9.341352106448738e-06, "loss": 0.1569, "step": 3282 }, { "epoch": 0.4958465488596889, "grad_norm": 4.566706645215724, "learning_rate": 9.340959735666992e-06, "loss": 0.2234, "step": 3283 }, { "epoch": 0.49599758344660927, "grad_norm": 2.389533915270842, "learning_rate": 9.340567256293747e-06, "loss": 0.165, "step": 3284 }, { "epoch": 0.49614861803352966, "grad_norm": 2.823244605347512, "learning_rate": 9.340174668338824e-06, "loss": 0.1952, "step": 3285 }, { "epoch": 0.49629965262045006, "grad_norm": 2.203540462635843, "learning_rate": 9.339781971812045e-06, "loss": 0.1632, "step": 3286 }, { "epoch": 0.4964506872073705, "grad_norm": 2.2545707600315397, "learning_rate": 9.339389166723232e-06, "loss": 0.1636, "step": 3287 }, { "epoch": 0.4966017217942909, "grad_norm": 2.536930859452923, "learning_rate": 9.33899625308221e-06, "loss": 0.1838, "step": 3288 }, { "epoch": 0.4967527563812113, "grad_norm": 1.8985750920849087, "learning_rate": 9.33860323089881e-06, "loss": 0.1393, "step": 3289 }, { "epoch": 0.4969037909681317, "grad_norm": 2.0476206425832686, "learning_rate": 9.338210100182865e-06, "loss": 0.1255, "step": 3290 }, { "epoch": 0.4970548255550521, "grad_norm": 3.208237335810491, "learning_rate": 9.337816860944206e-06, "loss": 0.1861, "step": 3291 }, { "epoch": 0.4972058601419725, "grad_norm": 2.3058352943454525, "learning_rate": 9.337423513192671e-06, "loss": 0.1503, "step": 3292 }, { "epoch": 0.49735689472889294, "grad_norm": 2.708649547180269, "learning_rate": 9.337030056938101e-06, "loss": 0.1925, "step": 3293 }, { "epoch": 0.49750792931581334, "grad_norm": 2.218906868834313, "learning_rate": 9.336636492190339e-06, "loss": 0.1395, "step": 3294 }, { "epoch": 0.49765896390273373, "grad_norm": 2.5052619143340045, "learning_rate": 9.336242818959226e-06, "loss": 0.1913, "step": 3295 }, { "epoch": 0.4978099984896541, "grad_norm": 1.9609086205449946, "learning_rate": 9.335849037254617e-06, "loss": 0.1445, "step": 3296 }, { "epoch": 0.4979610330765745, "grad_norm": 2.9289230960199673, "learning_rate": 9.335455147086356e-06, "loss": 0.1784, "step": 3297 }, { "epoch": 0.4981120676634949, "grad_norm": 2.123611769028261, "learning_rate": 9.335061148464301e-06, "loss": 0.1756, "step": 3298 }, { "epoch": 0.49826310225041537, "grad_norm": 2.331080437540746, "learning_rate": 9.334667041398304e-06, "loss": 0.1651, "step": 3299 }, { "epoch": 0.49841413683733576, "grad_norm": 1.7824870222469824, "learning_rate": 9.334272825898228e-06, "loss": 0.1447, "step": 3300 }, { "epoch": 0.49856517142425616, "grad_norm": 3.5421264352594677, "learning_rate": 9.333878501973932e-06, "loss": 0.1818, "step": 3301 }, { "epoch": 0.49871620601117655, "grad_norm": 2.7239574353779488, "learning_rate": 9.333484069635281e-06, "loss": 0.1633, "step": 3302 }, { "epoch": 0.49886724059809695, "grad_norm": 2.488332225456823, "learning_rate": 9.333089528892141e-06, "loss": 0.1607, "step": 3303 }, { "epoch": 0.49901827518501735, "grad_norm": 2.4477614929551463, "learning_rate": 9.332694879754382e-06, "loss": 0.1434, "step": 3304 }, { "epoch": 0.4991693097719378, "grad_norm": 2.628943459538166, "learning_rate": 9.33230012223188e-06, "loss": 0.1803, "step": 3305 }, { "epoch": 0.4993203443588582, "grad_norm": 3.2529360798986704, "learning_rate": 9.331905256334504e-06, "loss": 0.1744, "step": 3306 }, { "epoch": 0.4994713789457786, "grad_norm": 2.497348970936653, "learning_rate": 9.331510282072134e-06, "loss": 0.1599, "step": 3307 }, { "epoch": 0.499622413532699, "grad_norm": 2.2715321217067705, "learning_rate": 9.331115199454652e-06, "loss": 0.1698, "step": 3308 }, { "epoch": 0.4997734481196194, "grad_norm": 2.050697276873344, "learning_rate": 9.33072000849194e-06, "loss": 0.1339, "step": 3309 }, { "epoch": 0.4999244827065398, "grad_norm": 2.6299518328995983, "learning_rate": 9.330324709193885e-06, "loss": 0.1828, "step": 3310 }, { "epoch": 0.5000755172934602, "grad_norm": 3.0681777695443864, "learning_rate": 9.329929301570375e-06, "loss": 0.1731, "step": 3311 }, { "epoch": 0.5002265518803806, "grad_norm": 2.634544711538818, "learning_rate": 9.3295337856313e-06, "loss": 0.1887, "step": 3312 }, { "epoch": 0.500377586467301, "grad_norm": 2.123193317828186, "learning_rate": 9.329138161386557e-06, "loss": 0.148, "step": 3313 }, { "epoch": 0.5005286210542215, "grad_norm": 2.335557426426904, "learning_rate": 9.32874242884604e-06, "loss": 0.1663, "step": 3314 }, { "epoch": 0.5006796556411418, "grad_norm": 2.0547070523286983, "learning_rate": 9.32834658801965e-06, "loss": 0.1482, "step": 3315 }, { "epoch": 0.5008306902280623, "grad_norm": 3.3922771632010518, "learning_rate": 9.327950638917287e-06, "loss": 0.1672, "step": 3316 }, { "epoch": 0.5009817248149826, "grad_norm": 2.219713140298311, "learning_rate": 9.327554581548858e-06, "loss": 0.1398, "step": 3317 }, { "epoch": 0.501132759401903, "grad_norm": 2.5718924836106223, "learning_rate": 9.327158415924273e-06, "loss": 0.1436, "step": 3318 }, { "epoch": 0.5012837939888234, "grad_norm": 3.431471672769986, "learning_rate": 9.326762142053435e-06, "loss": 0.1907, "step": 3319 }, { "epoch": 0.5014348285757438, "grad_norm": 2.91786480065686, "learning_rate": 9.326365759946265e-06, "loss": 0.153, "step": 3320 }, { "epoch": 0.5015858631626643, "grad_norm": 2.085761209101573, "learning_rate": 9.325969269612673e-06, "loss": 0.1512, "step": 3321 }, { "epoch": 0.5017368977495846, "grad_norm": 1.9563553131920635, "learning_rate": 9.325572671062582e-06, "loss": 0.1322, "step": 3322 }, { "epoch": 0.5018879323365051, "grad_norm": 2.0619317136023807, "learning_rate": 9.325175964305907e-06, "loss": 0.16, "step": 3323 }, { "epoch": 0.5020389669234254, "grad_norm": 2.0309850465484893, "learning_rate": 9.324779149352578e-06, "loss": 0.1454, "step": 3324 }, { "epoch": 0.5021900015103459, "grad_norm": 2.0107347153273585, "learning_rate": 9.324382226212521e-06, "loss": 0.1218, "step": 3325 }, { "epoch": 0.5023410360972663, "grad_norm": 2.8947187617077437, "learning_rate": 9.32398519489566e-06, "loss": 0.2383, "step": 3326 }, { "epoch": 0.5024920706841867, "grad_norm": 2.3081162455293076, "learning_rate": 9.323588055411932e-06, "loss": 0.1681, "step": 3327 }, { "epoch": 0.5026431052711071, "grad_norm": 2.370195134985187, "learning_rate": 9.32319080777127e-06, "loss": 0.1517, "step": 3328 }, { "epoch": 0.5027941398580275, "grad_norm": 1.8971390395198242, "learning_rate": 9.32279345198361e-06, "loss": 0.1351, "step": 3329 }, { "epoch": 0.5029451744449479, "grad_norm": 2.329407111001686, "learning_rate": 9.322395988058894e-06, "loss": 0.1261, "step": 3330 }, { "epoch": 0.5030962090318682, "grad_norm": 3.2372491220573174, "learning_rate": 9.321998416007065e-06, "loss": 0.1822, "step": 3331 }, { "epoch": 0.5032472436187887, "grad_norm": 2.874939688429943, "learning_rate": 9.321600735838066e-06, "loss": 0.1907, "step": 3332 }, { "epoch": 0.5033982782057091, "grad_norm": 2.8235306435241205, "learning_rate": 9.32120294756185e-06, "loss": 0.1828, "step": 3333 }, { "epoch": 0.5035493127926295, "grad_norm": 2.449110223116736, "learning_rate": 9.320805051188362e-06, "loss": 0.1391, "step": 3334 }, { "epoch": 0.5037003473795499, "grad_norm": 2.6515175466710708, "learning_rate": 9.320407046727559e-06, "loss": 0.1725, "step": 3335 }, { "epoch": 0.5038513819664703, "grad_norm": 2.982365846470432, "learning_rate": 9.320008934189396e-06, "loss": 0.1911, "step": 3336 }, { "epoch": 0.5040024165533907, "grad_norm": 3.197034795872646, "learning_rate": 9.319610713583833e-06, "loss": 0.1671, "step": 3337 }, { "epoch": 0.5041534511403112, "grad_norm": 2.172018399373988, "learning_rate": 9.319212384920833e-06, "loss": 0.1479, "step": 3338 }, { "epoch": 0.5043044857272315, "grad_norm": 2.4146447777844346, "learning_rate": 9.318813948210357e-06, "loss": 0.1489, "step": 3339 }, { "epoch": 0.504455520314152, "grad_norm": 2.897506287131292, "learning_rate": 9.318415403462374e-06, "loss": 0.1488, "step": 3340 }, { "epoch": 0.5046065549010723, "grad_norm": 2.152642684940875, "learning_rate": 9.318016750686854e-06, "loss": 0.1452, "step": 3341 }, { "epoch": 0.5047575894879928, "grad_norm": 2.729836297293674, "learning_rate": 9.31761798989377e-06, "loss": 0.1691, "step": 3342 }, { "epoch": 0.5049086240749131, "grad_norm": 2.4517987997517072, "learning_rate": 9.317219121093095e-06, "loss": 0.1652, "step": 3343 }, { "epoch": 0.5050596586618336, "grad_norm": 2.1408099763059307, "learning_rate": 9.316820144294809e-06, "loss": 0.1537, "step": 3344 }, { "epoch": 0.505210693248754, "grad_norm": 2.3504334413393044, "learning_rate": 9.316421059508893e-06, "loss": 0.1714, "step": 3345 }, { "epoch": 0.5053617278356743, "grad_norm": 2.1152500063634063, "learning_rate": 9.316021866745328e-06, "loss": 0.1631, "step": 3346 }, { "epoch": 0.5055127624225948, "grad_norm": 2.666132496896591, "learning_rate": 9.315622566014102e-06, "loss": 0.1346, "step": 3347 }, { "epoch": 0.5056637970095151, "grad_norm": 2.6215866765362574, "learning_rate": 9.315223157325202e-06, "loss": 0.1697, "step": 3348 }, { "epoch": 0.5058148315964356, "grad_norm": 2.259644421047587, "learning_rate": 9.314823640688622e-06, "loss": 0.1449, "step": 3349 }, { "epoch": 0.505965866183356, "grad_norm": 2.1749845259844074, "learning_rate": 9.314424016114354e-06, "loss": 0.1636, "step": 3350 }, { "epoch": 0.5061169007702764, "grad_norm": 2.2980821797641933, "learning_rate": 9.314024283612396e-06, "loss": 0.1463, "step": 3351 }, { "epoch": 0.5062679353571968, "grad_norm": 2.1956420732365753, "learning_rate": 9.313624443192746e-06, "loss": 0.1412, "step": 3352 }, { "epoch": 0.5064189699441172, "grad_norm": 2.290268392944364, "learning_rate": 9.313224494865408e-06, "loss": 0.1577, "step": 3353 }, { "epoch": 0.5065700045310376, "grad_norm": 2.7581304871600323, "learning_rate": 9.312824438640385e-06, "loss": 0.1643, "step": 3354 }, { "epoch": 0.506721039117958, "grad_norm": 2.593690529343445, "learning_rate": 9.312424274527688e-06, "loss": 0.1718, "step": 3355 }, { "epoch": 0.5068720737048784, "grad_norm": 3.1662345142838184, "learning_rate": 9.312024002537322e-06, "loss": 0.1787, "step": 3356 }, { "epoch": 0.5070231082917989, "grad_norm": 2.2154071979209458, "learning_rate": 9.311623622679304e-06, "loss": 0.1435, "step": 3357 }, { "epoch": 0.5071741428787192, "grad_norm": 2.650100138599345, "learning_rate": 9.311223134963651e-06, "loss": 0.1704, "step": 3358 }, { "epoch": 0.5073251774656397, "grad_norm": 2.217835097921592, "learning_rate": 9.310822539400377e-06, "loss": 0.1759, "step": 3359 }, { "epoch": 0.50747621205256, "grad_norm": 3.7490614677623153, "learning_rate": 9.310421835999507e-06, "loss": 0.1661, "step": 3360 }, { "epoch": 0.5076272466394804, "grad_norm": 2.033141366258788, "learning_rate": 9.310021024771062e-06, "loss": 0.1385, "step": 3361 }, { "epoch": 0.5077782812264009, "grad_norm": 2.1788740126055073, "learning_rate": 9.30962010572507e-06, "loss": 0.1475, "step": 3362 }, { "epoch": 0.5079293158133212, "grad_norm": 3.0617687172936354, "learning_rate": 9.30921907887156e-06, "loss": 0.1541, "step": 3363 }, { "epoch": 0.5080803504002417, "grad_norm": 1.97522223491614, "learning_rate": 9.308817944220564e-06, "loss": 0.1343, "step": 3364 }, { "epoch": 0.508231384987162, "grad_norm": 2.5339822587114127, "learning_rate": 9.308416701782117e-06, "loss": 0.1386, "step": 3365 }, { "epoch": 0.5083824195740825, "grad_norm": 3.2264934509151644, "learning_rate": 9.308015351566253e-06, "loss": 0.1714, "step": 3366 }, { "epoch": 0.5085334541610028, "grad_norm": 2.508607420832539, "learning_rate": 9.307613893583018e-06, "loss": 0.1648, "step": 3367 }, { "epoch": 0.5086844887479233, "grad_norm": 2.0350513084419553, "learning_rate": 9.30721232784245e-06, "loss": 0.1494, "step": 3368 }, { "epoch": 0.5088355233348437, "grad_norm": 2.42574101368247, "learning_rate": 9.306810654354596e-06, "loss": 0.1612, "step": 3369 }, { "epoch": 0.5089865579217641, "grad_norm": 2.9254018357460074, "learning_rate": 9.306408873129504e-06, "loss": 0.148, "step": 3370 }, { "epoch": 0.5091375925086845, "grad_norm": 2.61356607144658, "learning_rate": 9.306006984177224e-06, "loss": 0.1732, "step": 3371 }, { "epoch": 0.5092886270956049, "grad_norm": 2.2324176012576546, "learning_rate": 9.30560498750781e-06, "loss": 0.1621, "step": 3372 }, { "epoch": 0.5094396616825253, "grad_norm": 2.115106548224253, "learning_rate": 9.305202883131319e-06, "loss": 0.1577, "step": 3373 }, { "epoch": 0.5095906962694458, "grad_norm": 1.896186949694356, "learning_rate": 9.30480067105781e-06, "loss": 0.119, "step": 3374 }, { "epoch": 0.5097417308563661, "grad_norm": 2.691299144884139, "learning_rate": 9.304398351297343e-06, "loss": 0.1817, "step": 3375 }, { "epoch": 0.5098927654432865, "grad_norm": 2.2597046127106113, "learning_rate": 9.303995923859983e-06, "loss": 0.1697, "step": 3376 }, { "epoch": 0.5100438000302069, "grad_norm": 2.429728395107791, "learning_rate": 9.303593388755796e-06, "loss": 0.1609, "step": 3377 }, { "epoch": 0.5101948346171273, "grad_norm": 2.2691834341547983, "learning_rate": 9.303190745994853e-06, "loss": 0.1693, "step": 3378 }, { "epoch": 0.5103458692040477, "grad_norm": 2.3164798911334032, "learning_rate": 9.302787995587227e-06, "loss": 0.164, "step": 3379 }, { "epoch": 0.5104969037909681, "grad_norm": 2.8400815476413523, "learning_rate": 9.30238513754299e-06, "loss": 0.1777, "step": 3380 }, { "epoch": 0.5106479383778886, "grad_norm": 2.356476983168311, "learning_rate": 9.301982171872224e-06, "loss": 0.1549, "step": 3381 }, { "epoch": 0.5107989729648089, "grad_norm": 1.8562182880610612, "learning_rate": 9.301579098585004e-06, "loss": 0.1284, "step": 3382 }, { "epoch": 0.5109500075517294, "grad_norm": 2.5008115838825997, "learning_rate": 9.30117591769142e-06, "loss": 0.1593, "step": 3383 }, { "epoch": 0.5111010421386497, "grad_norm": 2.6417976996662738, "learning_rate": 9.300772629201551e-06, "loss": 0.1527, "step": 3384 }, { "epoch": 0.5112520767255702, "grad_norm": 2.690157054269085, "learning_rate": 9.300369233125488e-06, "loss": 0.1758, "step": 3385 }, { "epoch": 0.5114031113124906, "grad_norm": 2.5008117088876642, "learning_rate": 9.299965729473324e-06, "loss": 0.1565, "step": 3386 }, { "epoch": 0.511554145899411, "grad_norm": 2.3760194225047426, "learning_rate": 9.299562118255152e-06, "loss": 0.1614, "step": 3387 }, { "epoch": 0.5117051804863314, "grad_norm": 2.5003043345034857, "learning_rate": 9.299158399481068e-06, "loss": 0.1724, "step": 3388 }, { "epoch": 0.5118562150732517, "grad_norm": 3.555465380902336, "learning_rate": 9.29875457316117e-06, "loss": 0.2011, "step": 3389 }, { "epoch": 0.5120072496601722, "grad_norm": 2.075725742488133, "learning_rate": 9.298350639305561e-06, "loss": 0.1537, "step": 3390 }, { "epoch": 0.5121582842470925, "grad_norm": 2.901195658819624, "learning_rate": 9.297946597924346e-06, "loss": 0.1603, "step": 3391 }, { "epoch": 0.512309318834013, "grad_norm": 2.054198869864008, "learning_rate": 9.297542449027634e-06, "loss": 0.1219, "step": 3392 }, { "epoch": 0.5124603534209334, "grad_norm": 2.262905442889386, "learning_rate": 9.297138192625531e-06, "loss": 0.1261, "step": 3393 }, { "epoch": 0.5126113880078538, "grad_norm": 2.452429344789097, "learning_rate": 9.296733828728152e-06, "loss": 0.1685, "step": 3394 }, { "epoch": 0.5127624225947742, "grad_norm": 1.9732995730658462, "learning_rate": 9.296329357345613e-06, "loss": 0.1407, "step": 3395 }, { "epoch": 0.5129134571816946, "grad_norm": 2.568688004391325, "learning_rate": 9.295924778488032e-06, "loss": 0.17, "step": 3396 }, { "epoch": 0.513064491768615, "grad_norm": 2.7234413642909803, "learning_rate": 9.295520092165529e-06, "loss": 0.2095, "step": 3397 }, { "epoch": 0.5132155263555355, "grad_norm": 1.9544589330306765, "learning_rate": 9.295115298388226e-06, "loss": 0.1552, "step": 3398 }, { "epoch": 0.5133665609424558, "grad_norm": 2.1193082616465824, "learning_rate": 9.294710397166251e-06, "loss": 0.1462, "step": 3399 }, { "epoch": 0.5135175955293763, "grad_norm": 2.3333374448892394, "learning_rate": 9.294305388509732e-06, "loss": 0.1529, "step": 3400 }, { "epoch": 0.5136686301162966, "grad_norm": 2.5997396228484178, "learning_rate": 9.293900272428802e-06, "loss": 0.1686, "step": 3401 }, { "epoch": 0.513819664703217, "grad_norm": 2.5607513877596424, "learning_rate": 9.293495048933595e-06, "loss": 0.1537, "step": 3402 }, { "epoch": 0.5139706992901374, "grad_norm": 2.1290561045707808, "learning_rate": 9.293089718034247e-06, "loss": 0.1332, "step": 3403 }, { "epoch": 0.5141217338770578, "grad_norm": 3.1535744311222715, "learning_rate": 9.292684279740896e-06, "loss": 0.2041, "step": 3404 }, { "epoch": 0.5142727684639783, "grad_norm": 2.799277278449591, "learning_rate": 9.292278734063686e-06, "loss": 0.1749, "step": 3405 }, { "epoch": 0.5144238030508986, "grad_norm": 3.269741866990183, "learning_rate": 9.291873081012763e-06, "loss": 0.1577, "step": 3406 }, { "epoch": 0.5145748376378191, "grad_norm": 3.0724430032747057, "learning_rate": 9.291467320598273e-06, "loss": 0.1695, "step": 3407 }, { "epoch": 0.5147258722247394, "grad_norm": 2.9710977251446304, "learning_rate": 9.291061452830368e-06, "loss": 0.1815, "step": 3408 }, { "epoch": 0.5148769068116599, "grad_norm": 4.169022573474104, "learning_rate": 9.290655477719197e-06, "loss": 0.1759, "step": 3409 }, { "epoch": 0.5150279413985803, "grad_norm": 2.9196586226160415, "learning_rate": 9.290249395274922e-06, "loss": 0.1526, "step": 3410 }, { "epoch": 0.5151789759855007, "grad_norm": 2.157048445694284, "learning_rate": 9.289843205507696e-06, "loss": 0.142, "step": 3411 }, { "epoch": 0.5153300105724211, "grad_norm": 1.846788956110989, "learning_rate": 9.289436908427682e-06, "loss": 0.1411, "step": 3412 }, { "epoch": 0.5154810451593415, "grad_norm": 2.337566705172713, "learning_rate": 9.289030504045042e-06, "loss": 0.1527, "step": 3413 }, { "epoch": 0.5156320797462619, "grad_norm": 2.4841514139523944, "learning_rate": 9.288623992369948e-06, "loss": 0.1695, "step": 3414 }, { "epoch": 0.5157831143331822, "grad_norm": 2.085547271250797, "learning_rate": 9.288217373412562e-06, "loss": 0.1472, "step": 3415 }, { "epoch": 0.5159341489201027, "grad_norm": 2.193126089292869, "learning_rate": 9.28781064718306e-06, "loss": 0.1637, "step": 3416 }, { "epoch": 0.5160851835070231, "grad_norm": 2.640299397424062, "learning_rate": 9.287403813691616e-06, "loss": 0.1785, "step": 3417 }, { "epoch": 0.5162362180939435, "grad_norm": 3.7939473945897086, "learning_rate": 9.286996872948407e-06, "loss": 0.2126, "step": 3418 }, { "epoch": 0.5163872526808639, "grad_norm": 2.302909187512576, "learning_rate": 9.286589824963612e-06, "loss": 0.1737, "step": 3419 }, { "epoch": 0.5165382872677843, "grad_norm": 2.321096777173033, "learning_rate": 9.286182669747413e-06, "loss": 0.1578, "step": 3420 }, { "epoch": 0.5166893218547047, "grad_norm": 2.2839521773865687, "learning_rate": 9.285775407309997e-06, "loss": 0.1706, "step": 3421 }, { "epoch": 0.5168403564416252, "grad_norm": 2.0614872967859497, "learning_rate": 9.285368037661553e-06, "loss": 0.1647, "step": 3422 }, { "epoch": 0.5169913910285455, "grad_norm": 2.703035785261727, "learning_rate": 9.284960560812267e-06, "loss": 0.1865, "step": 3423 }, { "epoch": 0.517142425615466, "grad_norm": 2.2248698733239674, "learning_rate": 9.284552976772336e-06, "loss": 0.1614, "step": 3424 }, { "epoch": 0.5172934602023863, "grad_norm": 1.6446227185078017, "learning_rate": 9.284145285551955e-06, "loss": 0.1196, "step": 3425 }, { "epoch": 0.5174444947893068, "grad_norm": 2.010142027996488, "learning_rate": 9.283737487161323e-06, "loss": 0.1699, "step": 3426 }, { "epoch": 0.5175955293762271, "grad_norm": 2.4359161082299887, "learning_rate": 9.28332958161064e-06, "loss": 0.1801, "step": 3427 }, { "epoch": 0.5177465639631476, "grad_norm": 2.2866746352761873, "learning_rate": 9.282921568910112e-06, "loss": 0.1361, "step": 3428 }, { "epoch": 0.517897598550068, "grad_norm": 3.316858746683991, "learning_rate": 9.282513449069945e-06, "loss": 0.1525, "step": 3429 }, { "epoch": 0.5180486331369883, "grad_norm": 2.2751894043014453, "learning_rate": 9.282105222100346e-06, "loss": 0.1532, "step": 3430 }, { "epoch": 0.5181996677239088, "grad_norm": 1.9204206445298106, "learning_rate": 9.281696888011529e-06, "loss": 0.1353, "step": 3431 }, { "epoch": 0.5183507023108291, "grad_norm": 2.0835048065931714, "learning_rate": 9.28128844681371e-06, "loss": 0.1448, "step": 3432 }, { "epoch": 0.5185017368977496, "grad_norm": 1.9459282405853517, "learning_rate": 9.280879898517103e-06, "loss": 0.1432, "step": 3433 }, { "epoch": 0.51865277148467, "grad_norm": 2.081108332749978, "learning_rate": 9.280471243131932e-06, "loss": 0.1576, "step": 3434 }, { "epoch": 0.5188038060715904, "grad_norm": 2.3098821425979166, "learning_rate": 9.280062480668416e-06, "loss": 0.1655, "step": 3435 }, { "epoch": 0.5189548406585108, "grad_norm": 3.628959363731646, "learning_rate": 9.279653611136784e-06, "loss": 0.2082, "step": 3436 }, { "epoch": 0.5191058752454312, "grad_norm": 2.427240766213095, "learning_rate": 9.279244634547261e-06, "loss": 0.1601, "step": 3437 }, { "epoch": 0.5192569098323516, "grad_norm": 2.8159669311498905, "learning_rate": 9.278835550910079e-06, "loss": 0.1802, "step": 3438 }, { "epoch": 0.519407944419272, "grad_norm": 2.0904381454590926, "learning_rate": 9.278426360235472e-06, "loss": 0.1632, "step": 3439 }, { "epoch": 0.5195589790061924, "grad_norm": 2.2393226525018664, "learning_rate": 9.278017062533676e-06, "loss": 0.174, "step": 3440 }, { "epoch": 0.5197100135931129, "grad_norm": 1.9442589209667709, "learning_rate": 9.27760765781493e-06, "loss": 0.13, "step": 3441 }, { "epoch": 0.5198610481800332, "grad_norm": 2.440650730976218, "learning_rate": 9.277198146089473e-06, "loss": 0.1749, "step": 3442 }, { "epoch": 0.5200120827669537, "grad_norm": 2.3663276407751055, "learning_rate": 9.276788527367552e-06, "loss": 0.1429, "step": 3443 }, { "epoch": 0.520163117353874, "grad_norm": 16.94952970469343, "learning_rate": 9.276378801659413e-06, "loss": 0.3005, "step": 3444 }, { "epoch": 0.5203141519407944, "grad_norm": 2.0634690526649737, "learning_rate": 9.275968968975303e-06, "loss": 0.1569, "step": 3445 }, { "epoch": 0.5204651865277149, "grad_norm": 2.7215889400290543, "learning_rate": 9.275559029325481e-06, "loss": 0.1865, "step": 3446 }, { "epoch": 0.5206162211146352, "grad_norm": 1.8955284674556683, "learning_rate": 9.275148982720195e-06, "loss": 0.1533, "step": 3447 }, { "epoch": 0.5207672557015557, "grad_norm": 2.3871197958422434, "learning_rate": 9.274738829169705e-06, "loss": 0.1537, "step": 3448 }, { "epoch": 0.520918290288476, "grad_norm": 2.420111048576695, "learning_rate": 9.274328568684272e-06, "loss": 0.1728, "step": 3449 }, { "epoch": 0.5210693248753965, "grad_norm": 2.465417317983724, "learning_rate": 9.273918201274156e-06, "loss": 0.1722, "step": 3450 }, { "epoch": 0.5212203594623168, "grad_norm": 1.8192876513815217, "learning_rate": 9.27350772694963e-06, "loss": 0.1324, "step": 3451 }, { "epoch": 0.5213713940492373, "grad_norm": 2.5423878946367306, "learning_rate": 9.27309714572095e-06, "loss": 0.1908, "step": 3452 }, { "epoch": 0.5215224286361577, "grad_norm": 2.635825390132972, "learning_rate": 9.2726864575984e-06, "loss": 0.1688, "step": 3453 }, { "epoch": 0.5216734632230781, "grad_norm": 2.4867684228017612, "learning_rate": 9.272275662592242e-06, "loss": 0.1648, "step": 3454 }, { "epoch": 0.5218244978099985, "grad_norm": 2.1375109407560133, "learning_rate": 9.271864760712761e-06, "loss": 0.151, "step": 3455 }, { "epoch": 0.5219755323969189, "grad_norm": 2.3980655799344848, "learning_rate": 9.271453751970232e-06, "loss": 0.1697, "step": 3456 }, { "epoch": 0.5221265669838393, "grad_norm": 1.9100565436605799, "learning_rate": 9.271042636374938e-06, "loss": 0.148, "step": 3457 }, { "epoch": 0.5222776015707598, "grad_norm": 2.1745244094945826, "learning_rate": 9.270631413937162e-06, "loss": 0.1571, "step": 3458 }, { "epoch": 0.5224286361576801, "grad_norm": 3.1754605020628044, "learning_rate": 9.270220084667192e-06, "loss": 0.1818, "step": 3459 }, { "epoch": 0.5225796707446005, "grad_norm": 2.7888826295244167, "learning_rate": 9.269808648575316e-06, "loss": 0.1654, "step": 3460 }, { "epoch": 0.5227307053315209, "grad_norm": 2.149178762580731, "learning_rate": 9.269397105671827e-06, "loss": 0.1531, "step": 3461 }, { "epoch": 0.5228817399184413, "grad_norm": 2.791753021519446, "learning_rate": 9.268985455967022e-06, "loss": 0.1677, "step": 3462 }, { "epoch": 0.5230327745053617, "grad_norm": 2.216391968760741, "learning_rate": 9.268573699471194e-06, "loss": 0.1702, "step": 3463 }, { "epoch": 0.5231838090922821, "grad_norm": 1.9665312138793847, "learning_rate": 9.268161836194651e-06, "loss": 0.1323, "step": 3464 }, { "epoch": 0.5233348436792026, "grad_norm": 3.1126137899810513, "learning_rate": 9.267749866147689e-06, "loss": 0.1707, "step": 3465 }, { "epoch": 0.5234858782661229, "grad_norm": 2.2257300654746204, "learning_rate": 9.267337789340616e-06, "loss": 0.1708, "step": 3466 }, { "epoch": 0.5236369128530434, "grad_norm": 2.5830218000743272, "learning_rate": 9.26692560578374e-06, "loss": 0.1781, "step": 3467 }, { "epoch": 0.5237879474399637, "grad_norm": 2.028371863328805, "learning_rate": 9.266513315487373e-06, "loss": 0.1392, "step": 3468 }, { "epoch": 0.5239389820268842, "grad_norm": 2.4198295732403916, "learning_rate": 9.266100918461827e-06, "loss": 0.1497, "step": 3469 }, { "epoch": 0.5240900166138046, "grad_norm": 2.019734166503747, "learning_rate": 9.26568841471742e-06, "loss": 0.1411, "step": 3470 }, { "epoch": 0.524241051200725, "grad_norm": 2.6573399336941415, "learning_rate": 9.26527580426447e-06, "loss": 0.1549, "step": 3471 }, { "epoch": 0.5243920857876454, "grad_norm": 3.9300334801881682, "learning_rate": 9.264863087113299e-06, "loss": 0.1515, "step": 3472 }, { "epoch": 0.5245431203745657, "grad_norm": 2.4968772259616996, "learning_rate": 9.264450263274233e-06, "loss": 0.1589, "step": 3473 }, { "epoch": 0.5246941549614862, "grad_norm": 2.399865730464934, "learning_rate": 9.264037332757596e-06, "loss": 0.1756, "step": 3474 }, { "epoch": 0.5248451895484065, "grad_norm": 2.9018224077689223, "learning_rate": 9.263624295573719e-06, "loss": 0.1707, "step": 3475 }, { "epoch": 0.524996224135327, "grad_norm": 3.024845552366074, "learning_rate": 9.263211151732933e-06, "loss": 0.1521, "step": 3476 }, { "epoch": 0.5251472587222474, "grad_norm": 3.0060538493907134, "learning_rate": 9.262797901245577e-06, "loss": 0.1849, "step": 3477 }, { "epoch": 0.5252982933091678, "grad_norm": 2.89913018726887, "learning_rate": 9.262384544121986e-06, "loss": 0.165, "step": 3478 }, { "epoch": 0.5254493278960882, "grad_norm": 2.232598987345142, "learning_rate": 9.261971080372498e-06, "loss": 0.1632, "step": 3479 }, { "epoch": 0.5256003624830086, "grad_norm": 2.4827199244854885, "learning_rate": 9.26155751000746e-06, "loss": 0.1418, "step": 3480 }, { "epoch": 0.525751397069929, "grad_norm": 2.5038851073916977, "learning_rate": 9.261143833037216e-06, "loss": 0.1551, "step": 3481 }, { "epoch": 0.5259024316568495, "grad_norm": 1.9067619609591002, "learning_rate": 9.260730049472116e-06, "loss": 0.1395, "step": 3482 }, { "epoch": 0.5260534662437698, "grad_norm": 2.0771680105846158, "learning_rate": 9.260316159322507e-06, "loss": 0.1511, "step": 3483 }, { "epoch": 0.5262045008306903, "grad_norm": 2.2180611633079574, "learning_rate": 9.259902162598746e-06, "loss": 0.1613, "step": 3484 }, { "epoch": 0.5263555354176106, "grad_norm": 3.287854247938939, "learning_rate": 9.25948805931119e-06, "loss": 0.1688, "step": 3485 }, { "epoch": 0.526506570004531, "grad_norm": 2.2367825608872907, "learning_rate": 9.259073849470197e-06, "loss": 0.159, "step": 3486 }, { "epoch": 0.5266576045914514, "grad_norm": 2.6681246665897826, "learning_rate": 9.258659533086128e-06, "loss": 0.172, "step": 3487 }, { "epoch": 0.5268086391783718, "grad_norm": 2.532015199750587, "learning_rate": 9.258245110169346e-06, "loss": 0.1466, "step": 3488 }, { "epoch": 0.5269596737652923, "grad_norm": 2.4172461463323103, "learning_rate": 9.25783058073022e-06, "loss": 0.1681, "step": 3489 }, { "epoch": 0.5271107083522126, "grad_norm": 2.64341543640398, "learning_rate": 9.25741594477912e-06, "loss": 0.1766, "step": 3490 }, { "epoch": 0.5272617429391331, "grad_norm": 2.2863224635989874, "learning_rate": 9.257001202326417e-06, "loss": 0.1517, "step": 3491 }, { "epoch": 0.5274127775260534, "grad_norm": 2.071621391459188, "learning_rate": 9.256586353382488e-06, "loss": 0.1538, "step": 3492 }, { "epoch": 0.5275638121129739, "grad_norm": 2.36835941690657, "learning_rate": 9.256171397957709e-06, "loss": 0.1506, "step": 3493 }, { "epoch": 0.5277148466998943, "grad_norm": 2.2133969071501762, "learning_rate": 9.25575633606246e-06, "loss": 0.1488, "step": 3494 }, { "epoch": 0.5278658812868147, "grad_norm": 2.6207864419012674, "learning_rate": 9.255341167707126e-06, "loss": 0.1542, "step": 3495 }, { "epoch": 0.5280169158737351, "grad_norm": 2.5036996606386333, "learning_rate": 9.25492589290209e-06, "loss": 0.1835, "step": 3496 }, { "epoch": 0.5281679504606555, "grad_norm": 1.9815127587363626, "learning_rate": 9.254510511657743e-06, "loss": 0.1299, "step": 3497 }, { "epoch": 0.5283189850475759, "grad_norm": 2.333966046616833, "learning_rate": 9.254095023984475e-06, "loss": 0.1574, "step": 3498 }, { "epoch": 0.5284700196344962, "grad_norm": 2.481448434475699, "learning_rate": 9.253679429892679e-06, "loss": 0.1581, "step": 3499 }, { "epoch": 0.5286210542214167, "grad_norm": 2.153420854462058, "learning_rate": 9.25326372939275e-06, "loss": 0.1484, "step": 3500 }, { "epoch": 0.5287720888083371, "grad_norm": 3.1732968483409323, "learning_rate": 9.25284792249509e-06, "loss": 0.1657, "step": 3501 }, { "epoch": 0.5289231233952575, "grad_norm": 2.307058843082498, "learning_rate": 9.2524320092101e-06, "loss": 0.1642, "step": 3502 }, { "epoch": 0.5290741579821779, "grad_norm": 1.9902042699170475, "learning_rate": 9.252015989548182e-06, "loss": 0.1394, "step": 3503 }, { "epoch": 0.5292251925690983, "grad_norm": 2.412537233070491, "learning_rate": 9.251599863519746e-06, "loss": 0.1782, "step": 3504 }, { "epoch": 0.5293762271560187, "grad_norm": 5.899811912908721, "learning_rate": 9.251183631135198e-06, "loss": 0.1451, "step": 3505 }, { "epoch": 0.5295272617429392, "grad_norm": 2.4570259918100237, "learning_rate": 9.250767292404955e-06, "loss": 0.1691, "step": 3506 }, { "epoch": 0.5296782963298595, "grad_norm": 1.7890336941007887, "learning_rate": 9.25035084733943e-06, "loss": 0.1269, "step": 3507 }, { "epoch": 0.52982933091678, "grad_norm": 2.6765390774770634, "learning_rate": 9.249934295949037e-06, "loss": 0.1649, "step": 3508 }, { "epoch": 0.5299803655037003, "grad_norm": 2.089245894549386, "learning_rate": 9.249517638244201e-06, "loss": 0.1516, "step": 3509 }, { "epoch": 0.5301314000906208, "grad_norm": 2.7589119117948986, "learning_rate": 9.249100874235344e-06, "loss": 0.1716, "step": 3510 }, { "epoch": 0.5302824346775411, "grad_norm": 2.405414315002588, "learning_rate": 9.248684003932889e-06, "loss": 0.1762, "step": 3511 }, { "epoch": 0.5304334692644616, "grad_norm": 2.3745181260808956, "learning_rate": 9.248267027347267e-06, "loss": 0.1486, "step": 3512 }, { "epoch": 0.530584503851382, "grad_norm": 2.091947978012036, "learning_rate": 9.24784994448891e-06, "loss": 0.1303, "step": 3513 }, { "epoch": 0.5307355384383023, "grad_norm": 2.137172952115537, "learning_rate": 9.247432755368246e-06, "loss": 0.1446, "step": 3514 }, { "epoch": 0.5308865730252228, "grad_norm": 3.935711385104786, "learning_rate": 9.247015459995717e-06, "loss": 0.1968, "step": 3515 }, { "epoch": 0.5310376076121431, "grad_norm": 2.148077196480298, "learning_rate": 9.246598058381758e-06, "loss": 0.1453, "step": 3516 }, { "epoch": 0.5311886421990636, "grad_norm": 2.6129786004242064, "learning_rate": 9.246180550536813e-06, "loss": 0.1707, "step": 3517 }, { "epoch": 0.531339676785984, "grad_norm": 2.1761607825681946, "learning_rate": 9.245762936471324e-06, "loss": 0.1539, "step": 3518 }, { "epoch": 0.5314907113729044, "grad_norm": 1.8908976912353537, "learning_rate": 9.24534521619574e-06, "loss": 0.1444, "step": 3519 }, { "epoch": 0.5316417459598248, "grad_norm": 2.362294242093243, "learning_rate": 9.24492738972051e-06, "loss": 0.1626, "step": 3520 }, { "epoch": 0.5317927805467452, "grad_norm": 2.5535567815609386, "learning_rate": 9.244509457056086e-06, "loss": 0.1646, "step": 3521 }, { "epoch": 0.5319438151336656, "grad_norm": 1.9580708130928974, "learning_rate": 9.244091418212921e-06, "loss": 0.1516, "step": 3522 }, { "epoch": 0.532094849720586, "grad_norm": 2.2613090363261614, "learning_rate": 9.243673273201477e-06, "loss": 0.1647, "step": 3523 }, { "epoch": 0.5322458843075064, "grad_norm": 2.561627645241278, "learning_rate": 9.24325502203221e-06, "loss": 0.1638, "step": 3524 }, { "epoch": 0.5323969188944269, "grad_norm": 2.4739438573761174, "learning_rate": 9.242836664715583e-06, "loss": 0.1605, "step": 3525 }, { "epoch": 0.5325479534813472, "grad_norm": 2.104936463194844, "learning_rate": 9.242418201262063e-06, "loss": 0.1486, "step": 3526 }, { "epoch": 0.5326989880682677, "grad_norm": 1.7436167667882354, "learning_rate": 9.241999631682116e-06, "loss": 0.1357, "step": 3527 }, { "epoch": 0.532850022655188, "grad_norm": 1.833414340227033, "learning_rate": 9.241580955986216e-06, "loss": 0.1313, "step": 3528 }, { "epoch": 0.5330010572421084, "grad_norm": 2.1363337688607484, "learning_rate": 9.241162174184834e-06, "loss": 0.1651, "step": 3529 }, { "epoch": 0.5331520918290289, "grad_norm": 2.2834692778511836, "learning_rate": 9.240743286288448e-06, "loss": 0.1623, "step": 3530 }, { "epoch": 0.5333031264159492, "grad_norm": 2.294140728569715, "learning_rate": 9.240324292307534e-06, "loss": 0.1669, "step": 3531 }, { "epoch": 0.5334541610028697, "grad_norm": 2.228456781965948, "learning_rate": 9.239905192252576e-06, "loss": 0.1458, "step": 3532 }, { "epoch": 0.53360519558979, "grad_norm": 2.366987764803969, "learning_rate": 9.239485986134056e-06, "loss": 0.1605, "step": 3533 }, { "epoch": 0.5337562301767105, "grad_norm": 2.3561517150138584, "learning_rate": 9.239066673962462e-06, "loss": 0.1596, "step": 3534 }, { "epoch": 0.5339072647636308, "grad_norm": 2.360722168455503, "learning_rate": 9.238647255748282e-06, "loss": 0.1758, "step": 3535 }, { "epoch": 0.5340582993505513, "grad_norm": 2.0047172773654163, "learning_rate": 9.23822773150201e-06, "loss": 0.1268, "step": 3536 }, { "epoch": 0.5342093339374717, "grad_norm": 2.1083725539752196, "learning_rate": 9.237808101234139e-06, "loss": 0.1526, "step": 3537 }, { "epoch": 0.5343603685243921, "grad_norm": 2.576759361897152, "learning_rate": 9.237388364955166e-06, "loss": 0.143, "step": 3538 }, { "epoch": 0.5345114031113125, "grad_norm": 2.013474494284868, "learning_rate": 9.236968522675592e-06, "loss": 0.1583, "step": 3539 }, { "epoch": 0.5346624376982329, "grad_norm": 2.37963784701845, "learning_rate": 9.236548574405918e-06, "loss": 0.1669, "step": 3540 }, { "epoch": 0.5348134722851533, "grad_norm": 2.0148638523449, "learning_rate": 9.236128520156652e-06, "loss": 0.1581, "step": 3541 }, { "epoch": 0.5349645068720738, "grad_norm": 2.4394298323974364, "learning_rate": 9.235708359938303e-06, "loss": 0.133, "step": 3542 }, { "epoch": 0.5351155414589941, "grad_norm": 2.579674980138323, "learning_rate": 9.235288093761375e-06, "loss": 0.1752, "step": 3543 }, { "epoch": 0.5352665760459145, "grad_norm": 2.0440858776900748, "learning_rate": 9.234867721636387e-06, "loss": 0.1397, "step": 3544 }, { "epoch": 0.5354176106328349, "grad_norm": 2.6873075447826276, "learning_rate": 9.234447243573853e-06, "loss": 0.1886, "step": 3545 }, { "epoch": 0.5355686452197553, "grad_norm": 2.209214542183019, "learning_rate": 9.23402665958429e-06, "loss": 0.1581, "step": 3546 }, { "epoch": 0.5357196798066757, "grad_norm": 1.8709731482477088, "learning_rate": 9.233605969678222e-06, "loss": 0.1412, "step": 3547 }, { "epoch": 0.5358707143935961, "grad_norm": 2.852798088456748, "learning_rate": 9.23318517386617e-06, "loss": 0.18, "step": 3548 }, { "epoch": 0.5360217489805166, "grad_norm": 2.3754020961386266, "learning_rate": 9.232764272158661e-06, "loss": 0.1631, "step": 3549 }, { "epoch": 0.5361727835674369, "grad_norm": 3.9786396316391497, "learning_rate": 9.232343264566225e-06, "loss": 0.1951, "step": 3550 }, { "epoch": 0.5363238181543574, "grad_norm": 2.0479749655808916, "learning_rate": 9.231922151099396e-06, "loss": 0.1344, "step": 3551 }, { "epoch": 0.5364748527412777, "grad_norm": 2.286926028508894, "learning_rate": 9.231500931768705e-06, "loss": 0.1675, "step": 3552 }, { "epoch": 0.5366258873281982, "grad_norm": 2.8462220653795214, "learning_rate": 9.23107960658469e-06, "loss": 0.213, "step": 3553 }, { "epoch": 0.5367769219151186, "grad_norm": 4.096813748042624, "learning_rate": 9.23065817555789e-06, "loss": 0.1447, "step": 3554 }, { "epoch": 0.536927956502039, "grad_norm": 1.878412694107271, "learning_rate": 9.23023663869885e-06, "loss": 0.1535, "step": 3555 }, { "epoch": 0.5370789910889594, "grad_norm": 1.8828871981779407, "learning_rate": 9.229814996018112e-06, "loss": 0.1498, "step": 3556 }, { "epoch": 0.5372300256758797, "grad_norm": 2.2278143373643333, "learning_rate": 9.229393247526223e-06, "loss": 0.1441, "step": 3557 }, { "epoch": 0.5373810602628002, "grad_norm": 2.1603854795473256, "learning_rate": 9.228971393233736e-06, "loss": 0.1455, "step": 3558 }, { "epoch": 0.5375320948497205, "grad_norm": 1.863239153278031, "learning_rate": 9.228549433151205e-06, "loss": 0.1353, "step": 3559 }, { "epoch": 0.537683129436641, "grad_norm": 2.720237357231114, "learning_rate": 9.22812736728918e-06, "loss": 0.1729, "step": 3560 }, { "epoch": 0.5378341640235614, "grad_norm": 2.454828946721301, "learning_rate": 9.227705195658223e-06, "loss": 0.1539, "step": 3561 }, { "epoch": 0.5379851986104818, "grad_norm": 2.5378169779793964, "learning_rate": 9.227282918268896e-06, "loss": 0.1909, "step": 3562 }, { "epoch": 0.5381362331974022, "grad_norm": 1.9169248650297888, "learning_rate": 9.22686053513176e-06, "loss": 0.1549, "step": 3563 }, { "epoch": 0.5382872677843226, "grad_norm": 2.2012420556012087, "learning_rate": 9.226438046257383e-06, "loss": 0.1072, "step": 3564 }, { "epoch": 0.538438302371243, "grad_norm": 2.47111625686753, "learning_rate": 9.226015451656333e-06, "loss": 0.1524, "step": 3565 }, { "epoch": 0.5385893369581635, "grad_norm": 1.7990947120211833, "learning_rate": 9.225592751339182e-06, "loss": 0.151, "step": 3566 }, { "epoch": 0.5387403715450838, "grad_norm": 3.037139736964382, "learning_rate": 9.2251699453165e-06, "loss": 0.1714, "step": 3567 }, { "epoch": 0.5388914061320043, "grad_norm": 2.990240446439394, "learning_rate": 9.224747033598872e-06, "loss": 0.1852, "step": 3568 }, { "epoch": 0.5390424407189246, "grad_norm": 2.4889766227552514, "learning_rate": 9.224324016196869e-06, "loss": 0.1562, "step": 3569 }, { "epoch": 0.539193475305845, "grad_norm": 2.248597063815566, "learning_rate": 9.223900893121075e-06, "loss": 0.1588, "step": 3570 }, { "epoch": 0.5393445098927654, "grad_norm": 2.207033271645952, "learning_rate": 9.22347766438208e-06, "loss": 0.1547, "step": 3571 }, { "epoch": 0.5394955444796858, "grad_norm": 2.6662816988956743, "learning_rate": 9.223054329990466e-06, "loss": 0.161, "step": 3572 }, { "epoch": 0.5396465790666063, "grad_norm": 2.2896098548123454, "learning_rate": 9.222630889956826e-06, "loss": 0.1502, "step": 3573 }, { "epoch": 0.5397976136535266, "grad_norm": 2.032298301565045, "learning_rate": 9.222207344291748e-06, "loss": 0.1547, "step": 3574 }, { "epoch": 0.5399486482404471, "grad_norm": 4.384837629065052, "learning_rate": 9.221783693005832e-06, "loss": 0.1691, "step": 3575 }, { "epoch": 0.5400996828273674, "grad_norm": 2.1646308262821368, "learning_rate": 9.221359936109674e-06, "loss": 0.1365, "step": 3576 }, { "epoch": 0.5402507174142879, "grad_norm": 2.391284404710628, "learning_rate": 9.220936073613874e-06, "loss": 0.1594, "step": 3577 }, { "epoch": 0.5404017520012083, "grad_norm": 1.9949420663002881, "learning_rate": 9.220512105529037e-06, "loss": 0.1512, "step": 3578 }, { "epoch": 0.5405527865881287, "grad_norm": 2.838571639727046, "learning_rate": 9.220088031865766e-06, "loss": 0.1739, "step": 3579 }, { "epoch": 0.5407038211750491, "grad_norm": 2.358148951859338, "learning_rate": 9.219663852634672e-06, "loss": 0.167, "step": 3580 }, { "epoch": 0.5408548557619695, "grad_norm": 2.7464640393963546, "learning_rate": 9.219239567846366e-06, "loss": 0.1618, "step": 3581 }, { "epoch": 0.5410058903488899, "grad_norm": 2.467649683718033, "learning_rate": 9.218815177511458e-06, "loss": 0.1627, "step": 3582 }, { "epoch": 0.5411569249358102, "grad_norm": 2.897950275727778, "learning_rate": 9.21839068164057e-06, "loss": 0.1813, "step": 3583 }, { "epoch": 0.5413079595227307, "grad_norm": 1.9582472298928257, "learning_rate": 9.217966080244316e-06, "loss": 0.1511, "step": 3584 }, { "epoch": 0.5414589941096511, "grad_norm": 2.783748617103837, "learning_rate": 9.21754137333332e-06, "loss": 0.1573, "step": 3585 }, { "epoch": 0.5416100286965715, "grad_norm": 2.0730772899099454, "learning_rate": 9.217116560918208e-06, "loss": 0.1435, "step": 3586 }, { "epoch": 0.5417610632834919, "grad_norm": 1.9887099069000034, "learning_rate": 9.216691643009603e-06, "loss": 0.1344, "step": 3587 }, { "epoch": 0.5419120978704123, "grad_norm": 1.9804978272269222, "learning_rate": 9.216266619618139e-06, "loss": 0.148, "step": 3588 }, { "epoch": 0.5420631324573327, "grad_norm": 2.151008828300826, "learning_rate": 9.215841490754443e-06, "loss": 0.1361, "step": 3589 }, { "epoch": 0.5422141670442532, "grad_norm": 2.4453869099312042, "learning_rate": 9.215416256429156e-06, "loss": 0.1698, "step": 3590 }, { "epoch": 0.5423652016311735, "grad_norm": 2.3273427726319524, "learning_rate": 9.214990916652908e-06, "loss": 0.1584, "step": 3591 }, { "epoch": 0.542516236218094, "grad_norm": 2.1508042986964475, "learning_rate": 9.214565471436345e-06, "loss": 0.154, "step": 3592 }, { "epoch": 0.5426672708050143, "grad_norm": 1.986772418843, "learning_rate": 9.214139920790108e-06, "loss": 0.1462, "step": 3593 }, { "epoch": 0.5428183053919348, "grad_norm": 1.9350719434480732, "learning_rate": 9.213714264724843e-06, "loss": 0.1336, "step": 3594 }, { "epoch": 0.5429693399788551, "grad_norm": 3.7536218992198824, "learning_rate": 9.213288503251194e-06, "loss": 0.1737, "step": 3595 }, { "epoch": 0.5431203745657756, "grad_norm": 2.647779133737349, "learning_rate": 9.212862636379818e-06, "loss": 0.1544, "step": 3596 }, { "epoch": 0.543271409152696, "grad_norm": 2.3103195905876017, "learning_rate": 9.212436664121365e-06, "loss": 0.1606, "step": 3597 }, { "epoch": 0.5434224437396163, "grad_norm": 2.9979109607670495, "learning_rate": 9.21201058648649e-06, "loss": 0.1486, "step": 3598 }, { "epoch": 0.5435734783265368, "grad_norm": 2.5643840841843204, "learning_rate": 9.211584403485853e-06, "loss": 0.1445, "step": 3599 }, { "epoch": 0.5437245129134571, "grad_norm": 3.6514045338108363, "learning_rate": 9.211158115130117e-06, "loss": 0.1634, "step": 3600 }, { "epoch": 0.5438755475003776, "grad_norm": 2.300017493602284, "learning_rate": 9.210731721429941e-06, "loss": 0.1703, "step": 3601 }, { "epoch": 0.544026582087298, "grad_norm": 1.811906561454033, "learning_rate": 9.210305222395996e-06, "loss": 0.1493, "step": 3602 }, { "epoch": 0.5441776166742184, "grad_norm": 2.014211604181974, "learning_rate": 9.209878618038948e-06, "loss": 0.1597, "step": 3603 }, { "epoch": 0.5443286512611388, "grad_norm": 2.6086198136527448, "learning_rate": 9.20945190836947e-06, "loss": 0.1668, "step": 3604 }, { "epoch": 0.5444796858480592, "grad_norm": 2.5050428354854652, "learning_rate": 9.209025093398238e-06, "loss": 0.1741, "step": 3605 }, { "epoch": 0.5446307204349796, "grad_norm": 2.557149079341852, "learning_rate": 9.208598173135928e-06, "loss": 0.1661, "step": 3606 }, { "epoch": 0.5447817550219, "grad_norm": 1.908907966127251, "learning_rate": 9.208171147593218e-06, "loss": 0.1527, "step": 3607 }, { "epoch": 0.5449327896088204, "grad_norm": 2.051879394593891, "learning_rate": 9.20774401678079e-06, "loss": 0.1418, "step": 3608 }, { "epoch": 0.5450838241957409, "grad_norm": 2.39466825727864, "learning_rate": 9.207316780709334e-06, "loss": 0.1534, "step": 3609 }, { "epoch": 0.5452348587826612, "grad_norm": 2.8041204737477035, "learning_rate": 9.20688943938953e-06, "loss": 0.1788, "step": 3610 }, { "epoch": 0.5453858933695817, "grad_norm": 3.340347075695943, "learning_rate": 9.206461992832074e-06, "loss": 0.1794, "step": 3611 }, { "epoch": 0.545536927956502, "grad_norm": 2.0152148377955044, "learning_rate": 9.206034441047658e-06, "loss": 0.1501, "step": 3612 }, { "epoch": 0.5456879625434224, "grad_norm": 2.4927800189344844, "learning_rate": 9.205606784046974e-06, "loss": 0.1544, "step": 3613 }, { "epoch": 0.5458389971303429, "grad_norm": 2.2051150294546935, "learning_rate": 9.205179021840723e-06, "loss": 0.1586, "step": 3614 }, { "epoch": 0.5459900317172632, "grad_norm": 2.0852190773192842, "learning_rate": 9.204751154439607e-06, "loss": 0.1393, "step": 3615 }, { "epoch": 0.5461410663041837, "grad_norm": 2.130907372529793, "learning_rate": 9.204323181854324e-06, "loss": 0.1543, "step": 3616 }, { "epoch": 0.546292100891104, "grad_norm": 1.9768008542081745, "learning_rate": 9.203895104095585e-06, "loss": 0.1456, "step": 3617 }, { "epoch": 0.5464431354780245, "grad_norm": 2.3012933098922104, "learning_rate": 9.203466921174098e-06, "loss": 0.1622, "step": 3618 }, { "epoch": 0.5465941700649448, "grad_norm": 2.3442190100681937, "learning_rate": 9.203038633100572e-06, "loss": 0.1534, "step": 3619 }, { "epoch": 0.5467452046518653, "grad_norm": 2.2137968788366753, "learning_rate": 9.202610239885724e-06, "loss": 0.1707, "step": 3620 }, { "epoch": 0.5468962392387857, "grad_norm": 2.4135494159555377, "learning_rate": 9.202181741540265e-06, "loss": 0.1656, "step": 3621 }, { "epoch": 0.5470472738257061, "grad_norm": 2.851083724000443, "learning_rate": 9.20175313807492e-06, "loss": 0.2002, "step": 3622 }, { "epoch": 0.5471983084126265, "grad_norm": 2.2298750772089138, "learning_rate": 9.20132442950041e-06, "loss": 0.167, "step": 3623 }, { "epoch": 0.5473493429995469, "grad_norm": 2.9442000689583074, "learning_rate": 9.200895615827456e-06, "loss": 0.1451, "step": 3624 }, { "epoch": 0.5475003775864673, "grad_norm": 2.7006268012245034, "learning_rate": 9.200466697066785e-06, "loss": 0.1613, "step": 3625 }, { "epoch": 0.5476514121733878, "grad_norm": 1.7725745045707082, "learning_rate": 9.200037673229131e-06, "loss": 0.1338, "step": 3626 }, { "epoch": 0.5478024467603081, "grad_norm": 2.1958349476148755, "learning_rate": 9.199608544325223e-06, "loss": 0.1489, "step": 3627 }, { "epoch": 0.5479534813472285, "grad_norm": 2.7544859248215365, "learning_rate": 9.199179310365798e-06, "loss": 0.1868, "step": 3628 }, { "epoch": 0.5481045159341489, "grad_norm": 2.372379479440418, "learning_rate": 9.198749971361591e-06, "loss": 0.1899, "step": 3629 }, { "epoch": 0.5482555505210693, "grad_norm": 1.971242231822713, "learning_rate": 9.198320527323344e-06, "loss": 0.1368, "step": 3630 }, { "epoch": 0.5484065851079897, "grad_norm": 2.663920162115881, "learning_rate": 9.197890978261798e-06, "loss": 0.1792, "step": 3631 }, { "epoch": 0.5485576196949101, "grad_norm": 2.932710230724745, "learning_rate": 9.1974613241877e-06, "loss": 0.1627, "step": 3632 }, { "epoch": 0.5487086542818306, "grad_norm": 3.089727573036253, "learning_rate": 9.197031565111798e-06, "loss": 0.172, "step": 3633 }, { "epoch": 0.5488596888687509, "grad_norm": 2.078004085534844, "learning_rate": 9.19660170104484e-06, "loss": 0.1768, "step": 3634 }, { "epoch": 0.5490107234556714, "grad_norm": 2.407728017879293, "learning_rate": 9.196171731997585e-06, "loss": 0.1597, "step": 3635 }, { "epoch": 0.5491617580425917, "grad_norm": 1.886802042197193, "learning_rate": 9.195741657980783e-06, "loss": 0.1348, "step": 3636 }, { "epoch": 0.5493127926295122, "grad_norm": 2.203295927922647, "learning_rate": 9.195311479005195e-06, "loss": 0.1701, "step": 3637 }, { "epoch": 0.5494638272164326, "grad_norm": 2.4058970198153156, "learning_rate": 9.194881195081583e-06, "loss": 0.1793, "step": 3638 }, { "epoch": 0.549614861803353, "grad_norm": 1.8351278882932631, "learning_rate": 9.194450806220709e-06, "loss": 0.136, "step": 3639 }, { "epoch": 0.5497658963902734, "grad_norm": 2.578047185922308, "learning_rate": 9.194020312433342e-06, "loss": 0.1375, "step": 3640 }, { "epoch": 0.5499169309771937, "grad_norm": 2.0923969590202014, "learning_rate": 9.193589713730247e-06, "loss": 0.1543, "step": 3641 }, { "epoch": 0.5500679655641142, "grad_norm": 2.1643192464400993, "learning_rate": 9.1931590101222e-06, "loss": 0.1592, "step": 3642 }, { "epoch": 0.5502190001510345, "grad_norm": 2.1157715277027505, "learning_rate": 9.192728201619973e-06, "loss": 0.1638, "step": 3643 }, { "epoch": 0.550370034737955, "grad_norm": 1.8611728122945266, "learning_rate": 9.192297288234341e-06, "loss": 0.1337, "step": 3644 }, { "epoch": 0.5505210693248754, "grad_norm": 2.0863818339511235, "learning_rate": 9.191866269976088e-06, "loss": 0.1508, "step": 3645 }, { "epoch": 0.5506721039117958, "grad_norm": 1.9555202533886449, "learning_rate": 9.191435146855995e-06, "loss": 0.1508, "step": 3646 }, { "epoch": 0.5508231384987162, "grad_norm": 2.3833251680879224, "learning_rate": 9.191003918884845e-06, "loss": 0.1762, "step": 3647 }, { "epoch": 0.5509741730856366, "grad_norm": 1.8391408571594, "learning_rate": 9.190572586073424e-06, "loss": 0.1494, "step": 3648 }, { "epoch": 0.551125207672557, "grad_norm": 3.0538489232075543, "learning_rate": 9.190141148432525e-06, "loss": 0.1646, "step": 3649 }, { "epoch": 0.5512762422594775, "grad_norm": 2.392985534891734, "learning_rate": 9.189709605972941e-06, "loss": 0.1396, "step": 3650 }, { "epoch": 0.5514272768463978, "grad_norm": 2.226297658575812, "learning_rate": 9.189277958705465e-06, "loss": 0.1585, "step": 3651 }, { "epoch": 0.5515783114333183, "grad_norm": 2.434129185388931, "learning_rate": 9.188846206640896e-06, "loss": 0.1743, "step": 3652 }, { "epoch": 0.5517293460202386, "grad_norm": 2.5457239807784764, "learning_rate": 9.188414349790033e-06, "loss": 0.1763, "step": 3653 }, { "epoch": 0.551880380607159, "grad_norm": 3.6552706820267513, "learning_rate": 9.187982388163683e-06, "loss": 0.1443, "step": 3654 }, { "epoch": 0.5520314151940794, "grad_norm": 2.2369381222304825, "learning_rate": 9.18755032177265e-06, "loss": 0.1411, "step": 3655 }, { "epoch": 0.5521824497809998, "grad_norm": 1.7172012677379767, "learning_rate": 9.187118150627738e-06, "loss": 0.121, "step": 3656 }, { "epoch": 0.5523334843679203, "grad_norm": 1.8669385046937543, "learning_rate": 9.186685874739765e-06, "loss": 0.1274, "step": 3657 }, { "epoch": 0.5524845189548406, "grad_norm": 2.283306996584451, "learning_rate": 9.18625349411954e-06, "loss": 0.1655, "step": 3658 }, { "epoch": 0.5526355535417611, "grad_norm": 2.181013999672435, "learning_rate": 9.18582100877788e-06, "loss": 0.14, "step": 3659 }, { "epoch": 0.5527865881286814, "grad_norm": 2.241707246250028, "learning_rate": 9.185388418725605e-06, "loss": 0.146, "step": 3660 }, { "epoch": 0.5529376227156019, "grad_norm": 3.408247909523271, "learning_rate": 9.184955723973537e-06, "loss": 0.1679, "step": 3661 }, { "epoch": 0.5530886573025223, "grad_norm": 2.585950036390463, "learning_rate": 9.184522924532498e-06, "loss": 0.1692, "step": 3662 }, { "epoch": 0.5532396918894427, "grad_norm": 2.3928629158458863, "learning_rate": 9.184090020413316e-06, "loss": 0.1472, "step": 3663 }, { "epoch": 0.5533907264763631, "grad_norm": 1.9028258070802864, "learning_rate": 9.18365701162682e-06, "loss": 0.1393, "step": 3664 }, { "epoch": 0.5535417610632835, "grad_norm": 2.7161695759600266, "learning_rate": 9.183223898183842e-06, "loss": 0.1669, "step": 3665 }, { "epoch": 0.5536927956502039, "grad_norm": 2.340224980704433, "learning_rate": 9.182790680095217e-06, "loss": 0.1633, "step": 3666 }, { "epoch": 0.5538438302371242, "grad_norm": 2.189662036152442, "learning_rate": 9.18235735737178e-06, "loss": 0.1491, "step": 3667 }, { "epoch": 0.5539948648240447, "grad_norm": 2.0056509679019374, "learning_rate": 9.181923930024374e-06, "loss": 0.1366, "step": 3668 }, { "epoch": 0.5541458994109651, "grad_norm": 2.9409871635763647, "learning_rate": 9.181490398063841e-06, "loss": 0.1875, "step": 3669 }, { "epoch": 0.5542969339978855, "grad_norm": 2.2922237763691635, "learning_rate": 9.181056761501024e-06, "loss": 0.1434, "step": 3670 }, { "epoch": 0.5544479685848059, "grad_norm": 2.1699877713866664, "learning_rate": 9.180623020346771e-06, "loss": 0.156, "step": 3671 }, { "epoch": 0.5545990031717263, "grad_norm": 1.967930926280596, "learning_rate": 9.180189174611934e-06, "loss": 0.1507, "step": 3672 }, { "epoch": 0.5547500377586467, "grad_norm": 2.148931326333162, "learning_rate": 9.179755224307365e-06, "loss": 0.1692, "step": 3673 }, { "epoch": 0.5549010723455672, "grad_norm": 2.8903826085817097, "learning_rate": 9.17932116944392e-06, "loss": 0.175, "step": 3674 }, { "epoch": 0.5550521069324875, "grad_norm": 2.36724634400547, "learning_rate": 9.178887010032455e-06, "loss": 0.1643, "step": 3675 }, { "epoch": 0.555203141519408, "grad_norm": 2.064684478288793, "learning_rate": 9.178452746083832e-06, "loss": 0.1391, "step": 3676 }, { "epoch": 0.5553541761063283, "grad_norm": 3.3399677100393212, "learning_rate": 9.178018377608914e-06, "loss": 0.1691, "step": 3677 }, { "epoch": 0.5555052106932488, "grad_norm": 1.9988839257113773, "learning_rate": 9.17758390461857e-06, "loss": 0.1445, "step": 3678 }, { "epoch": 0.5556562452801692, "grad_norm": 2.1721115927347294, "learning_rate": 9.177149327123665e-06, "loss": 0.1647, "step": 3679 }, { "epoch": 0.5558072798670896, "grad_norm": 2.218919642135805, "learning_rate": 9.176714645135072e-06, "loss": 0.1461, "step": 3680 }, { "epoch": 0.55595831445401, "grad_norm": 2.2859581981954644, "learning_rate": 9.176279858663663e-06, "loss": 0.1551, "step": 3681 }, { "epoch": 0.5561093490409303, "grad_norm": 1.953481592246942, "learning_rate": 9.175844967720316e-06, "loss": 0.1321, "step": 3682 }, { "epoch": 0.5562603836278508, "grad_norm": 1.98413902048455, "learning_rate": 9.17540997231591e-06, "loss": 0.1578, "step": 3683 }, { "epoch": 0.5564114182147711, "grad_norm": 2.2532389426498365, "learning_rate": 9.174974872461325e-06, "loss": 0.1446, "step": 3684 }, { "epoch": 0.5565624528016916, "grad_norm": 2.681799898738688, "learning_rate": 9.174539668167448e-06, "loss": 0.2151, "step": 3685 }, { "epoch": 0.556713487388612, "grad_norm": 2.1970164862858286, "learning_rate": 9.174104359445163e-06, "loss": 0.1564, "step": 3686 }, { "epoch": 0.5568645219755324, "grad_norm": 3.3376901991509564, "learning_rate": 9.173668946305362e-06, "loss": 0.1629, "step": 3687 }, { "epoch": 0.5570155565624528, "grad_norm": 2.359437326815895, "learning_rate": 9.173233428758936e-06, "loss": 0.1611, "step": 3688 }, { "epoch": 0.5571665911493732, "grad_norm": 1.9482032153066358, "learning_rate": 9.172797806816778e-06, "loss": 0.1293, "step": 3689 }, { "epoch": 0.5573176257362936, "grad_norm": 3.4551510741578637, "learning_rate": 9.17236208048979e-06, "loss": 0.1553, "step": 3690 }, { "epoch": 0.5574686603232141, "grad_norm": 1.8803628253492193, "learning_rate": 9.171926249788866e-06, "loss": 0.1385, "step": 3691 }, { "epoch": 0.5576196949101344, "grad_norm": 2.8261011144714554, "learning_rate": 9.171490314724913e-06, "loss": 0.1716, "step": 3692 }, { "epoch": 0.5577707294970549, "grad_norm": 2.131652761992478, "learning_rate": 9.171054275308834e-06, "loss": 0.1475, "step": 3693 }, { "epoch": 0.5579217640839752, "grad_norm": 2.1173017278581487, "learning_rate": 9.17061813155154e-06, "loss": 0.1426, "step": 3694 }, { "epoch": 0.5580727986708957, "grad_norm": 2.0478031776632166, "learning_rate": 9.170181883463933e-06, "loss": 0.1632, "step": 3695 }, { "epoch": 0.558223833257816, "grad_norm": 2.1158035918371536, "learning_rate": 9.169745531056936e-06, "loss": 0.1361, "step": 3696 }, { "epoch": 0.5583748678447364, "grad_norm": 2.5278628905239713, "learning_rate": 9.169309074341461e-06, "loss": 0.1775, "step": 3697 }, { "epoch": 0.5585259024316569, "grad_norm": 1.6334800092824318, "learning_rate": 9.168872513328424e-06, "loss": 0.1351, "step": 3698 }, { "epoch": 0.5586769370185772, "grad_norm": 1.917200134098871, "learning_rate": 9.168435848028747e-06, "loss": 0.1237, "step": 3699 }, { "epoch": 0.5588279716054977, "grad_norm": 3.017469720939847, "learning_rate": 9.167999078453354e-06, "loss": 0.193, "step": 3700 }, { "epoch": 0.558979006192418, "grad_norm": 2.5453941532440174, "learning_rate": 9.167562204613171e-06, "loss": 0.1773, "step": 3701 }, { "epoch": 0.5591300407793385, "grad_norm": 2.28061987729615, "learning_rate": 9.167125226519127e-06, "loss": 0.1664, "step": 3702 }, { "epoch": 0.5592810753662589, "grad_norm": 1.9310872713102287, "learning_rate": 9.166688144182152e-06, "loss": 0.1286, "step": 3703 }, { "epoch": 0.5594321099531793, "grad_norm": 2.5699729556976667, "learning_rate": 9.16625095761318e-06, "loss": 0.1719, "step": 3704 }, { "epoch": 0.5595831445400997, "grad_norm": 4.4127234793086805, "learning_rate": 9.16581366682315e-06, "loss": 0.2005, "step": 3705 }, { "epoch": 0.5597341791270201, "grad_norm": 2.30100319915763, "learning_rate": 9.165376271823e-06, "loss": 0.1627, "step": 3706 }, { "epoch": 0.5598852137139405, "grad_norm": 2.6260710250400394, "learning_rate": 9.164938772623668e-06, "loss": 0.186, "step": 3707 }, { "epoch": 0.5600362483008609, "grad_norm": 2.1494479284580015, "learning_rate": 9.164501169236102e-06, "loss": 0.1702, "step": 3708 }, { "epoch": 0.5601872828877813, "grad_norm": 1.899902880018554, "learning_rate": 9.164063461671247e-06, "loss": 0.1507, "step": 3709 }, { "epoch": 0.5603383174747018, "grad_norm": 3.1535435275192865, "learning_rate": 9.163625649940054e-06, "loss": 0.1759, "step": 3710 }, { "epoch": 0.5604893520616221, "grad_norm": 2.54953801017305, "learning_rate": 9.163187734053475e-06, "loss": 0.1745, "step": 3711 }, { "epoch": 0.5606403866485425, "grad_norm": 1.9466930716611002, "learning_rate": 9.162749714022464e-06, "loss": 0.1326, "step": 3712 }, { "epoch": 0.5607914212354629, "grad_norm": 2.4743159211190773, "learning_rate": 9.16231158985798e-06, "loss": 0.1853, "step": 3713 }, { "epoch": 0.5609424558223833, "grad_norm": 2.532734138182259, "learning_rate": 9.16187336157098e-06, "loss": 0.1695, "step": 3714 }, { "epoch": 0.5610934904093038, "grad_norm": 4.7810016913027775, "learning_rate": 9.161435029172427e-06, "loss": 0.242, "step": 3715 }, { "epoch": 0.5612445249962241, "grad_norm": 2.1145010321913285, "learning_rate": 9.16099659267329e-06, "loss": 0.1498, "step": 3716 }, { "epoch": 0.5613955595831446, "grad_norm": 2.16190339829092, "learning_rate": 9.160558052084532e-06, "loss": 0.1611, "step": 3717 }, { "epoch": 0.5615465941700649, "grad_norm": 1.7519957322582984, "learning_rate": 9.160119407417124e-06, "loss": 0.124, "step": 3718 }, { "epoch": 0.5616976287569854, "grad_norm": 2.464940078149108, "learning_rate": 9.159680658682042e-06, "loss": 0.1707, "step": 3719 }, { "epoch": 0.5618486633439057, "grad_norm": 2.3673821445079635, "learning_rate": 9.15924180589026e-06, "loss": 0.1616, "step": 3720 }, { "epoch": 0.5619996979308262, "grad_norm": 2.191156046887686, "learning_rate": 9.158802849052755e-06, "loss": 0.166, "step": 3721 }, { "epoch": 0.5621507325177466, "grad_norm": 1.9350278074261433, "learning_rate": 9.15836378818051e-06, "loss": 0.1493, "step": 3722 }, { "epoch": 0.562301767104667, "grad_norm": 2.189627370664278, "learning_rate": 9.157924623284505e-06, "loss": 0.1772, "step": 3723 }, { "epoch": 0.5624528016915874, "grad_norm": 1.9340248564312013, "learning_rate": 9.157485354375728e-06, "loss": 0.147, "step": 3724 }, { "epoch": 0.5626038362785077, "grad_norm": 2.075521009849117, "learning_rate": 9.157045981465167e-06, "loss": 0.1492, "step": 3725 }, { "epoch": 0.5627548708654282, "grad_norm": 2.0294887331698286, "learning_rate": 9.156606504563814e-06, "loss": 0.1425, "step": 3726 }, { "epoch": 0.5629059054523486, "grad_norm": 2.435635907227161, "learning_rate": 9.156166923682664e-06, "loss": 0.18, "step": 3727 }, { "epoch": 0.563056940039269, "grad_norm": 2.416448314601212, "learning_rate": 9.15572723883271e-06, "loss": 0.1947, "step": 3728 }, { "epoch": 0.5632079746261894, "grad_norm": 2.321268400488711, "learning_rate": 9.155287450024952e-06, "loss": 0.1347, "step": 3729 }, { "epoch": 0.5633590092131098, "grad_norm": 2.2205967265441555, "learning_rate": 9.154847557270394e-06, "loss": 0.1592, "step": 3730 }, { "epoch": 0.5635100438000302, "grad_norm": 2.9584803914517925, "learning_rate": 9.154407560580037e-06, "loss": 0.1954, "step": 3731 }, { "epoch": 0.5636610783869506, "grad_norm": 1.9755854877567869, "learning_rate": 9.15396745996489e-06, "loss": 0.1409, "step": 3732 }, { "epoch": 0.563812112973871, "grad_norm": 1.9037359173197221, "learning_rate": 9.153527255435962e-06, "loss": 0.1492, "step": 3733 }, { "epoch": 0.5639631475607915, "grad_norm": 2.437182739297785, "learning_rate": 9.153086947004264e-06, "loss": 0.171, "step": 3734 }, { "epoch": 0.5641141821477118, "grad_norm": 2.2098219175216602, "learning_rate": 9.15264653468081e-06, "loss": 0.1486, "step": 3735 }, { "epoch": 0.5642652167346323, "grad_norm": 2.8081881352262976, "learning_rate": 9.15220601847662e-06, "loss": 0.1753, "step": 3736 }, { "epoch": 0.5644162513215526, "grad_norm": 2.280631558888315, "learning_rate": 9.15176539840271e-06, "loss": 0.1512, "step": 3737 }, { "epoch": 0.564567285908473, "grad_norm": 2.016210750413351, "learning_rate": 9.151324674470105e-06, "loss": 0.154, "step": 3738 }, { "epoch": 0.5647183204953935, "grad_norm": 3.20407043402595, "learning_rate": 9.15088384668983e-06, "loss": 0.1451, "step": 3739 }, { "epoch": 0.5648693550823138, "grad_norm": 2.332013313315058, "learning_rate": 9.15044291507291e-06, "loss": 0.1862, "step": 3740 }, { "epoch": 0.5650203896692343, "grad_norm": 2.552505707552089, "learning_rate": 9.15000187963038e-06, "loss": 0.1587, "step": 3741 }, { "epoch": 0.5651714242561546, "grad_norm": 2.5375171627401385, "learning_rate": 9.149560740373266e-06, "loss": 0.1938, "step": 3742 }, { "epoch": 0.5653224588430751, "grad_norm": 2.220051209948698, "learning_rate": 9.149119497312609e-06, "loss": 0.1596, "step": 3743 }, { "epoch": 0.5654734934299954, "grad_norm": 2.4471646115325476, "learning_rate": 9.148678150459444e-06, "loss": 0.149, "step": 3744 }, { "epoch": 0.5656245280169159, "grad_norm": 2.986794081308482, "learning_rate": 9.148236699824815e-06, "loss": 0.1686, "step": 3745 }, { "epoch": 0.5657755626038363, "grad_norm": 2.139181205683906, "learning_rate": 9.14779514541976e-06, "loss": 0.154, "step": 3746 }, { "epoch": 0.5659265971907567, "grad_norm": 2.653199598199458, "learning_rate": 9.147353487255328e-06, "loss": 0.1745, "step": 3747 }, { "epoch": 0.5660776317776771, "grad_norm": 1.866149931482679, "learning_rate": 9.146911725342567e-06, "loss": 0.1306, "step": 3748 }, { "epoch": 0.5662286663645975, "grad_norm": 2.428362403511417, "learning_rate": 9.146469859692525e-06, "loss": 0.1765, "step": 3749 }, { "epoch": 0.5663797009515179, "grad_norm": 2.2185593326560773, "learning_rate": 9.146027890316262e-06, "loss": 0.1681, "step": 3750 }, { "epoch": 0.5665307355384384, "grad_norm": 2.8992420867743216, "learning_rate": 9.14558581722483e-06, "loss": 0.2005, "step": 3751 }, { "epoch": 0.5666817701253587, "grad_norm": 2.5453238282237036, "learning_rate": 9.145143640429286e-06, "loss": 0.1509, "step": 3752 }, { "epoch": 0.5668328047122791, "grad_norm": 3.0509380377780024, "learning_rate": 9.144701359940693e-06, "loss": 0.1955, "step": 3753 }, { "epoch": 0.5669838392991995, "grad_norm": 2.0690858651828035, "learning_rate": 9.144258975770115e-06, "loss": 0.1629, "step": 3754 }, { "epoch": 0.5671348738861199, "grad_norm": 2.9904104697491736, "learning_rate": 9.14381648792862e-06, "loss": 0.1508, "step": 3755 }, { "epoch": 0.5672859084730403, "grad_norm": 1.9657660066722538, "learning_rate": 9.143373896427275e-06, "loss": 0.1381, "step": 3756 }, { "epoch": 0.5674369430599607, "grad_norm": 2.237580428979254, "learning_rate": 9.142931201277153e-06, "loss": 0.1601, "step": 3757 }, { "epoch": 0.5675879776468812, "grad_norm": 2.1280598432420312, "learning_rate": 9.142488402489326e-06, "loss": 0.1483, "step": 3758 }, { "epoch": 0.5677390122338015, "grad_norm": 2.228203989062616, "learning_rate": 9.142045500074874e-06, "loss": 0.1499, "step": 3759 }, { "epoch": 0.567890046820722, "grad_norm": 2.1614761542543053, "learning_rate": 9.141602494044873e-06, "loss": 0.1606, "step": 3760 }, { "epoch": 0.5680410814076423, "grad_norm": 1.974217887315363, "learning_rate": 9.141159384410409e-06, "loss": 0.1407, "step": 3761 }, { "epoch": 0.5681921159945628, "grad_norm": 2.36639708745851, "learning_rate": 9.140716171182563e-06, "loss": 0.16, "step": 3762 }, { "epoch": 0.5683431505814832, "grad_norm": 2.1422580584137054, "learning_rate": 9.140272854372425e-06, "loss": 0.1475, "step": 3763 }, { "epoch": 0.5684941851684036, "grad_norm": 3.8756056497191516, "learning_rate": 9.139829433991082e-06, "loss": 0.1581, "step": 3764 }, { "epoch": 0.568645219755324, "grad_norm": 2.78174517306959, "learning_rate": 9.13938591004963e-06, "loss": 0.1869, "step": 3765 }, { "epoch": 0.5687962543422443, "grad_norm": 2.6843488543809237, "learning_rate": 9.138942282559161e-06, "loss": 0.1896, "step": 3766 }, { "epoch": 0.5689472889291648, "grad_norm": 2.1028081710461106, "learning_rate": 9.138498551530772e-06, "loss": 0.1593, "step": 3767 }, { "epoch": 0.5690983235160851, "grad_norm": 2.20712232966232, "learning_rate": 9.138054716975565e-06, "loss": 0.1503, "step": 3768 }, { "epoch": 0.5692493581030056, "grad_norm": 2.169822358651931, "learning_rate": 9.137610778904645e-06, "loss": 0.1417, "step": 3769 }, { "epoch": 0.569400392689926, "grad_norm": 3.406200313393253, "learning_rate": 9.137166737329114e-06, "loss": 0.1831, "step": 3770 }, { "epoch": 0.5695514272768464, "grad_norm": 2.2951671947619148, "learning_rate": 9.13672259226008e-06, "loss": 0.1585, "step": 3771 }, { "epoch": 0.5697024618637668, "grad_norm": 2.610212494437288, "learning_rate": 9.136278343708656e-06, "loss": 0.1872, "step": 3772 }, { "epoch": 0.5698534964506872, "grad_norm": 2.134334618125195, "learning_rate": 9.135833991685951e-06, "loss": 0.1736, "step": 3773 }, { "epoch": 0.5700045310376076, "grad_norm": 2.376658573648182, "learning_rate": 9.135389536203085e-06, "loss": 0.1354, "step": 3774 }, { "epoch": 0.5701555656245281, "grad_norm": 2.2941793247385265, "learning_rate": 9.134944977271175e-06, "loss": 0.1444, "step": 3775 }, { "epoch": 0.5703066002114484, "grad_norm": 2.1077604489553403, "learning_rate": 9.13450031490134e-06, "loss": 0.1489, "step": 3776 }, { "epoch": 0.5704576347983689, "grad_norm": 1.9431623072903552, "learning_rate": 9.134055549104703e-06, "loss": 0.1308, "step": 3777 }, { "epoch": 0.5706086693852892, "grad_norm": 2.1344760511893788, "learning_rate": 9.133610679892395e-06, "loss": 0.1388, "step": 3778 }, { "epoch": 0.5707597039722097, "grad_norm": 2.661573342211187, "learning_rate": 9.13316570727554e-06, "loss": 0.1658, "step": 3779 }, { "epoch": 0.57091073855913, "grad_norm": 2.0012873320070934, "learning_rate": 9.132720631265271e-06, "loss": 0.1558, "step": 3780 }, { "epoch": 0.5710617731460504, "grad_norm": 2.518632259120001, "learning_rate": 9.132275451872723e-06, "loss": 0.1577, "step": 3781 }, { "epoch": 0.5712128077329709, "grad_norm": 2.475250369667986, "learning_rate": 9.13183016910903e-06, "loss": 0.1709, "step": 3782 }, { "epoch": 0.5713638423198912, "grad_norm": 2.7525718259906724, "learning_rate": 9.13138478298533e-06, "loss": 0.1628, "step": 3783 }, { "epoch": 0.5715148769068117, "grad_norm": 2.6168305460739507, "learning_rate": 9.13093929351277e-06, "loss": 0.1734, "step": 3784 }, { "epoch": 0.571665911493732, "grad_norm": 2.311326896707662, "learning_rate": 9.130493700702489e-06, "loss": 0.1678, "step": 3785 }, { "epoch": 0.5718169460806525, "grad_norm": 1.877771080432676, "learning_rate": 9.130048004565637e-06, "loss": 0.1452, "step": 3786 }, { "epoch": 0.5719679806675729, "grad_norm": 3.3015173882241475, "learning_rate": 9.129602205113358e-06, "loss": 0.1998, "step": 3787 }, { "epoch": 0.5721190152544933, "grad_norm": 1.9737867701272762, "learning_rate": 9.129156302356812e-06, "loss": 0.1576, "step": 3788 }, { "epoch": 0.5722700498414137, "grad_norm": 2.1181215143404284, "learning_rate": 9.128710296307146e-06, "loss": 0.1552, "step": 3789 }, { "epoch": 0.5724210844283341, "grad_norm": 1.8864822549984175, "learning_rate": 9.128264186975523e-06, "loss": 0.1538, "step": 3790 }, { "epoch": 0.5725721190152545, "grad_norm": 2.1201783091897153, "learning_rate": 9.127817974373098e-06, "loss": 0.1624, "step": 3791 }, { "epoch": 0.5727231536021749, "grad_norm": 2.248917346902855, "learning_rate": 9.127371658511037e-06, "loss": 0.1486, "step": 3792 }, { "epoch": 0.5728741881890953, "grad_norm": 1.9276360342132923, "learning_rate": 9.126925239400502e-06, "loss": 0.1371, "step": 3793 }, { "epoch": 0.5730252227760158, "grad_norm": 2.2360862543313935, "learning_rate": 9.126478717052661e-06, "loss": 0.1476, "step": 3794 }, { "epoch": 0.5731762573629361, "grad_norm": 2.513507205476285, "learning_rate": 9.126032091478685e-06, "loss": 0.169, "step": 3795 }, { "epoch": 0.5733272919498565, "grad_norm": 2.966978067153093, "learning_rate": 9.125585362689747e-06, "loss": 0.2105, "step": 3796 }, { "epoch": 0.5734783265367769, "grad_norm": 2.6716635246602105, "learning_rate": 9.125138530697022e-06, "loss": 0.1757, "step": 3797 }, { "epoch": 0.5736293611236973, "grad_norm": 2.4521448908024075, "learning_rate": 9.124691595511685e-06, "loss": 0.1636, "step": 3798 }, { "epoch": 0.5737803957106178, "grad_norm": 2.7682126493823485, "learning_rate": 9.124244557144918e-06, "loss": 0.1538, "step": 3799 }, { "epoch": 0.5739314302975381, "grad_norm": 2.4484658649310926, "learning_rate": 9.123797415607906e-06, "loss": 0.1988, "step": 3800 }, { "epoch": 0.5740824648844586, "grad_norm": 2.3789190706164853, "learning_rate": 9.123350170911831e-06, "loss": 0.1511, "step": 3801 }, { "epoch": 0.5742334994713789, "grad_norm": 2.1123210380266064, "learning_rate": 9.122902823067886e-06, "loss": 0.1487, "step": 3802 }, { "epoch": 0.5743845340582994, "grad_norm": 2.900441779244232, "learning_rate": 9.122455372087255e-06, "loss": 0.1842, "step": 3803 }, { "epoch": 0.5745355686452197, "grad_norm": 2.0820513294917125, "learning_rate": 9.122007817981136e-06, "loss": 0.1899, "step": 3804 }, { "epoch": 0.5746866032321402, "grad_norm": 2.0938722969428145, "learning_rate": 9.121560160760722e-06, "loss": 0.1585, "step": 3805 }, { "epoch": 0.5748376378190606, "grad_norm": 3.049889820773422, "learning_rate": 9.121112400437215e-06, "loss": 0.1921, "step": 3806 }, { "epoch": 0.574988672405981, "grad_norm": 2.3766523580810817, "learning_rate": 9.120664537021815e-06, "loss": 0.1523, "step": 3807 }, { "epoch": 0.5751397069929014, "grad_norm": 2.076136881421085, "learning_rate": 9.120216570525721e-06, "loss": 0.1468, "step": 3808 }, { "epoch": 0.5752907415798217, "grad_norm": 16.077387318139092, "learning_rate": 9.119768500960145e-06, "loss": 0.1892, "step": 3809 }, { "epoch": 0.5754417761667422, "grad_norm": 2.1037428377800187, "learning_rate": 9.119320328336292e-06, "loss": 0.142, "step": 3810 }, { "epoch": 0.5755928107536626, "grad_norm": 2.3711650529786144, "learning_rate": 9.118872052665375e-06, "loss": 0.1524, "step": 3811 }, { "epoch": 0.575743845340583, "grad_norm": 2.3496747240500566, "learning_rate": 9.118423673958607e-06, "loss": 0.1816, "step": 3812 }, { "epoch": 0.5758948799275034, "grad_norm": 2.6032363343715694, "learning_rate": 9.117975192227207e-06, "loss": 0.1733, "step": 3813 }, { "epoch": 0.5760459145144238, "grad_norm": 3.058259971008921, "learning_rate": 9.11752660748239e-06, "loss": 0.1788, "step": 3814 }, { "epoch": 0.5761969491013442, "grad_norm": 2.138051256085586, "learning_rate": 9.11707791973538e-06, "loss": 0.1484, "step": 3815 }, { "epoch": 0.5763479836882646, "grad_norm": 1.9285939249590984, "learning_rate": 9.1166291289974e-06, "loss": 0.1399, "step": 3816 }, { "epoch": 0.576499018275185, "grad_norm": 2.449683328645932, "learning_rate": 9.116180235279677e-06, "loss": 0.1733, "step": 3817 }, { "epoch": 0.5766500528621055, "grad_norm": 2.2386626668972034, "learning_rate": 9.115731238593441e-06, "loss": 0.1483, "step": 3818 }, { "epoch": 0.5768010874490258, "grad_norm": 2.0102464071678843, "learning_rate": 9.115282138949925e-06, "loss": 0.1591, "step": 3819 }, { "epoch": 0.5769521220359463, "grad_norm": 2.662914800741061, "learning_rate": 9.114832936360361e-06, "loss": 0.1932, "step": 3820 }, { "epoch": 0.5771031566228666, "grad_norm": 2.0367370388684134, "learning_rate": 9.114383630835987e-06, "loss": 0.1517, "step": 3821 }, { "epoch": 0.577254191209787, "grad_norm": 2.0695167759258353, "learning_rate": 9.113934222388044e-06, "loss": 0.1554, "step": 3822 }, { "epoch": 0.5774052257967075, "grad_norm": 2.310643715341719, "learning_rate": 9.113484711027772e-06, "loss": 0.1419, "step": 3823 }, { "epoch": 0.5775562603836278, "grad_norm": 2.702169861963506, "learning_rate": 9.113035096766416e-06, "loss": 0.1611, "step": 3824 }, { "epoch": 0.5777072949705483, "grad_norm": 2.8905106458099024, "learning_rate": 9.112585379615224e-06, "loss": 0.1737, "step": 3825 }, { "epoch": 0.5778583295574686, "grad_norm": 2.2340267229477613, "learning_rate": 9.112135559585447e-06, "loss": 0.1769, "step": 3826 }, { "epoch": 0.5780093641443891, "grad_norm": 2.611378795660935, "learning_rate": 9.111685636688337e-06, "loss": 0.174, "step": 3827 }, { "epoch": 0.5781603987313094, "grad_norm": 3.5418437056819156, "learning_rate": 9.111235610935147e-06, "loss": 0.185, "step": 3828 }, { "epoch": 0.5783114333182299, "grad_norm": 2.4974118697241354, "learning_rate": 9.110785482337137e-06, "loss": 0.1642, "step": 3829 }, { "epoch": 0.5784624679051503, "grad_norm": 2.3192302012407797, "learning_rate": 9.110335250905566e-06, "loss": 0.1627, "step": 3830 }, { "epoch": 0.5786135024920707, "grad_norm": 2.9308266954863575, "learning_rate": 9.109884916651698e-06, "loss": 0.1415, "step": 3831 }, { "epoch": 0.5787645370789911, "grad_norm": 2.075768281639753, "learning_rate": 9.109434479586798e-06, "loss": 0.1556, "step": 3832 }, { "epoch": 0.5789155716659115, "grad_norm": 2.381427648281898, "learning_rate": 9.108983939722132e-06, "loss": 0.1653, "step": 3833 }, { "epoch": 0.5790666062528319, "grad_norm": 2.3429287485614294, "learning_rate": 9.108533297068972e-06, "loss": 0.1641, "step": 3834 }, { "epoch": 0.5792176408397524, "grad_norm": 3.141248136847111, "learning_rate": 9.108082551638593e-06, "loss": 0.1515, "step": 3835 }, { "epoch": 0.5793686754266727, "grad_norm": 2.5070956617710727, "learning_rate": 9.107631703442267e-06, "loss": 0.183, "step": 3836 }, { "epoch": 0.5795197100135931, "grad_norm": 2.6005716126267355, "learning_rate": 9.107180752491273e-06, "loss": 0.1868, "step": 3837 }, { "epoch": 0.5796707446005135, "grad_norm": 2.55813167191339, "learning_rate": 9.106729698796894e-06, "loss": 0.1547, "step": 3838 }, { "epoch": 0.5798217791874339, "grad_norm": 2.307806729949188, "learning_rate": 9.106278542370413e-06, "loss": 0.1429, "step": 3839 }, { "epoch": 0.5799728137743543, "grad_norm": 2.0786134378855174, "learning_rate": 9.105827283223114e-06, "loss": 0.1528, "step": 3840 }, { "epoch": 0.5801238483612747, "grad_norm": 2.036170405287027, "learning_rate": 9.105375921366288e-06, "loss": 0.1508, "step": 3841 }, { "epoch": 0.5802748829481952, "grad_norm": 2.6743051219126865, "learning_rate": 9.104924456811222e-06, "loss": 0.1481, "step": 3842 }, { "epoch": 0.5804259175351155, "grad_norm": 2.096329757751671, "learning_rate": 9.104472889569213e-06, "loss": 0.1616, "step": 3843 }, { "epoch": 0.580576952122036, "grad_norm": 1.9371859546580747, "learning_rate": 9.104021219651558e-06, "loss": 0.1306, "step": 3844 }, { "epoch": 0.5807279867089563, "grad_norm": 1.8371371679030006, "learning_rate": 9.103569447069553e-06, "loss": 0.1324, "step": 3845 }, { "epoch": 0.5808790212958768, "grad_norm": 2.1551797856488557, "learning_rate": 9.1031175718345e-06, "loss": 0.1663, "step": 3846 }, { "epoch": 0.5810300558827972, "grad_norm": 1.9702480433056055, "learning_rate": 9.102665593957705e-06, "loss": 0.1539, "step": 3847 }, { "epoch": 0.5811810904697176, "grad_norm": 2.11283063118937, "learning_rate": 9.10221351345047e-06, "loss": 0.1265, "step": 3848 }, { "epoch": 0.581332125056638, "grad_norm": 3.2429386527760022, "learning_rate": 9.101761330324111e-06, "loss": 0.1878, "step": 3849 }, { "epoch": 0.5814831596435583, "grad_norm": 2.0087172069958896, "learning_rate": 9.101309044589932e-06, "loss": 0.1682, "step": 3850 }, { "epoch": 0.5816341942304788, "grad_norm": 2.3930368607769674, "learning_rate": 9.100856656259251e-06, "loss": 0.168, "step": 3851 }, { "epoch": 0.5817852288173991, "grad_norm": 1.988854110966821, "learning_rate": 9.100404165343385e-06, "loss": 0.1443, "step": 3852 }, { "epoch": 0.5819362634043196, "grad_norm": 6.150642612944946, "learning_rate": 9.099951571853654e-06, "loss": 0.1885, "step": 3853 }, { "epoch": 0.58208729799124, "grad_norm": 2.304055697056619, "learning_rate": 9.099498875801379e-06, "loss": 0.1661, "step": 3854 }, { "epoch": 0.5822383325781604, "grad_norm": 2.692542780125573, "learning_rate": 9.099046077197883e-06, "loss": 0.1721, "step": 3855 }, { "epoch": 0.5823893671650808, "grad_norm": 2.4336823967825625, "learning_rate": 9.098593176054492e-06, "loss": 0.1608, "step": 3856 }, { "epoch": 0.5825404017520012, "grad_norm": 3.356711368252825, "learning_rate": 9.098140172382541e-06, "loss": 0.1613, "step": 3857 }, { "epoch": 0.5826914363389216, "grad_norm": 2.407118072373805, "learning_rate": 9.097687066193359e-06, "loss": 0.128, "step": 3858 }, { "epoch": 0.5828424709258421, "grad_norm": 2.233635855670535, "learning_rate": 9.097233857498278e-06, "loss": 0.1457, "step": 3859 }, { "epoch": 0.5829935055127624, "grad_norm": 2.6365241084336, "learning_rate": 9.096780546308639e-06, "loss": 0.1856, "step": 3860 }, { "epoch": 0.5831445400996829, "grad_norm": 1.9578609292154185, "learning_rate": 9.09632713263578e-06, "loss": 0.1521, "step": 3861 }, { "epoch": 0.5832955746866032, "grad_norm": 2.9094214913242973, "learning_rate": 9.095873616491045e-06, "loss": 0.1964, "step": 3862 }, { "epoch": 0.5834466092735237, "grad_norm": 1.6586620367391554, "learning_rate": 9.095419997885778e-06, "loss": 0.1387, "step": 3863 }, { "epoch": 0.583597643860444, "grad_norm": 2.3338777753791864, "learning_rate": 9.094966276831325e-06, "loss": 0.154, "step": 3864 }, { "epoch": 0.5837486784473644, "grad_norm": 2.074035389190972, "learning_rate": 9.094512453339039e-06, "loss": 0.1229, "step": 3865 }, { "epoch": 0.5838997130342849, "grad_norm": 1.8755483107358757, "learning_rate": 9.09405852742027e-06, "loss": 0.1434, "step": 3866 }, { "epoch": 0.5840507476212052, "grad_norm": 1.8434916402764616, "learning_rate": 9.093604499086374e-06, "loss": 0.1445, "step": 3867 }, { "epoch": 0.5842017822081257, "grad_norm": 3.6386305306450617, "learning_rate": 9.09315036834871e-06, "loss": 0.1954, "step": 3868 }, { "epoch": 0.584352816795046, "grad_norm": 1.8799951888334014, "learning_rate": 9.09269613521864e-06, "loss": 0.142, "step": 3869 }, { "epoch": 0.5845038513819665, "grad_norm": 3.0528290660898345, "learning_rate": 9.092241799707522e-06, "loss": 0.1769, "step": 3870 }, { "epoch": 0.5846548859688869, "grad_norm": 3.5169502983491965, "learning_rate": 9.091787361826723e-06, "loss": 0.1608, "step": 3871 }, { "epoch": 0.5848059205558073, "grad_norm": 2.0497427287764975, "learning_rate": 9.091332821587613e-06, "loss": 0.1441, "step": 3872 }, { "epoch": 0.5849569551427277, "grad_norm": 2.4661859387288083, "learning_rate": 9.090878179001561e-06, "loss": 0.1627, "step": 3873 }, { "epoch": 0.5851079897296481, "grad_norm": 2.1537315041533027, "learning_rate": 9.09042343407994e-06, "loss": 0.1325, "step": 3874 }, { "epoch": 0.5852590243165685, "grad_norm": 2.3825482509841875, "learning_rate": 9.089968586834128e-06, "loss": 0.1504, "step": 3875 }, { "epoch": 0.5854100589034889, "grad_norm": 2.4459517501568646, "learning_rate": 9.0895136372755e-06, "loss": 0.1718, "step": 3876 }, { "epoch": 0.5855610934904093, "grad_norm": 1.8806014304117569, "learning_rate": 9.089058585415438e-06, "loss": 0.1447, "step": 3877 }, { "epoch": 0.5857121280773298, "grad_norm": 1.9033208944794793, "learning_rate": 9.088603431265329e-06, "loss": 0.1697, "step": 3878 }, { "epoch": 0.5858631626642501, "grad_norm": 2.5570726125098537, "learning_rate": 9.08814817483655e-06, "loss": 0.1668, "step": 3879 }, { "epoch": 0.5860141972511705, "grad_norm": 1.7020768675115905, "learning_rate": 9.0876928161405e-06, "loss": 0.1204, "step": 3880 }, { "epoch": 0.5861652318380909, "grad_norm": 2.41270825762533, "learning_rate": 9.087237355188563e-06, "loss": 0.1798, "step": 3881 }, { "epoch": 0.5863162664250113, "grad_norm": 1.86559771876864, "learning_rate": 9.086781791992137e-06, "loss": 0.154, "step": 3882 }, { "epoch": 0.5864673010119318, "grad_norm": 2.2212538882133854, "learning_rate": 9.086326126562612e-06, "loss": 0.1613, "step": 3883 }, { "epoch": 0.5866183355988521, "grad_norm": 2.6726677208024903, "learning_rate": 9.085870358911394e-06, "loss": 0.1628, "step": 3884 }, { "epoch": 0.5867693701857726, "grad_norm": 1.893599711641354, "learning_rate": 9.085414489049879e-06, "loss": 0.1413, "step": 3885 }, { "epoch": 0.5869204047726929, "grad_norm": 1.9974803560772687, "learning_rate": 9.084958516989474e-06, "loss": 0.1454, "step": 3886 }, { "epoch": 0.5870714393596134, "grad_norm": 1.9091013447212335, "learning_rate": 9.084502442741583e-06, "loss": 0.158, "step": 3887 }, { "epoch": 0.5872224739465337, "grad_norm": 2.4378613809710528, "learning_rate": 9.084046266317617e-06, "loss": 0.1666, "step": 3888 }, { "epoch": 0.5873735085334542, "grad_norm": 2.94238932565561, "learning_rate": 9.083589987728986e-06, "loss": 0.1722, "step": 3889 }, { "epoch": 0.5875245431203746, "grad_norm": 2.1792201769258273, "learning_rate": 9.083133606987106e-06, "loss": 0.1637, "step": 3890 }, { "epoch": 0.587675577707295, "grad_norm": 1.9956856905597293, "learning_rate": 9.082677124103392e-06, "loss": 0.1305, "step": 3891 }, { "epoch": 0.5878266122942154, "grad_norm": 2.2233838765942995, "learning_rate": 9.082220539089263e-06, "loss": 0.1529, "step": 3892 }, { "epoch": 0.5879776468811357, "grad_norm": 2.304831881569778, "learning_rate": 9.08176385195614e-06, "loss": 0.1683, "step": 3893 }, { "epoch": 0.5881286814680562, "grad_norm": 2.360073382337796, "learning_rate": 9.081307062715449e-06, "loss": 0.1699, "step": 3894 }, { "epoch": 0.5882797160549766, "grad_norm": 3.9461277226218057, "learning_rate": 9.080850171378617e-06, "loss": 0.1761, "step": 3895 }, { "epoch": 0.588430750641897, "grad_norm": 1.7628411445758867, "learning_rate": 9.080393177957073e-06, "loss": 0.1477, "step": 3896 }, { "epoch": 0.5885817852288174, "grad_norm": 2.0248128497423967, "learning_rate": 9.079936082462248e-06, "loss": 0.1366, "step": 3897 }, { "epoch": 0.5887328198157378, "grad_norm": 2.570023545879603, "learning_rate": 9.079478884905577e-06, "loss": 0.1887, "step": 3898 }, { "epoch": 0.5888838544026582, "grad_norm": 2.8400494842742723, "learning_rate": 9.079021585298499e-06, "loss": 0.173, "step": 3899 }, { "epoch": 0.5890348889895786, "grad_norm": 2.0014424884133057, "learning_rate": 9.07856418365245e-06, "loss": 0.1609, "step": 3900 }, { "epoch": 0.589185923576499, "grad_norm": 2.619044852525568, "learning_rate": 9.078106679978873e-06, "loss": 0.181, "step": 3901 }, { "epoch": 0.5893369581634195, "grad_norm": 1.5513511706242973, "learning_rate": 9.077649074289216e-06, "loss": 0.1202, "step": 3902 }, { "epoch": 0.5894879927503398, "grad_norm": 2.924484555497876, "learning_rate": 9.077191366594922e-06, "loss": 0.176, "step": 3903 }, { "epoch": 0.5896390273372603, "grad_norm": 1.8727961749797872, "learning_rate": 9.076733556907443e-06, "loss": 0.1369, "step": 3904 }, { "epoch": 0.5897900619241806, "grad_norm": 2.312658316698318, "learning_rate": 9.076275645238233e-06, "loss": 0.1619, "step": 3905 }, { "epoch": 0.589941096511101, "grad_norm": 2.1718342558165125, "learning_rate": 9.075817631598742e-06, "loss": 0.1474, "step": 3906 }, { "epoch": 0.5900921310980215, "grad_norm": 2.2722194533053353, "learning_rate": 9.075359516000432e-06, "loss": 0.1788, "step": 3907 }, { "epoch": 0.5902431656849418, "grad_norm": 2.3660761517897475, "learning_rate": 9.074901298454762e-06, "loss": 0.1586, "step": 3908 }, { "epoch": 0.5903942002718623, "grad_norm": 2.130785272252903, "learning_rate": 9.074442978973193e-06, "loss": 0.1505, "step": 3909 }, { "epoch": 0.5905452348587826, "grad_norm": 2.104864686764231, "learning_rate": 9.073984557567192e-06, "loss": 0.1418, "step": 3910 }, { "epoch": 0.5906962694457031, "grad_norm": 2.2369327497369706, "learning_rate": 9.073526034248225e-06, "loss": 0.1603, "step": 3911 }, { "epoch": 0.5908473040326234, "grad_norm": 2.0741592668083633, "learning_rate": 9.073067409027764e-06, "loss": 0.1627, "step": 3912 }, { "epoch": 0.5909983386195439, "grad_norm": 2.2886414194517815, "learning_rate": 9.07260868191728e-06, "loss": 0.1667, "step": 3913 }, { "epoch": 0.5911493732064643, "grad_norm": 2.608428445587872, "learning_rate": 9.072149852928249e-06, "loss": 0.1733, "step": 3914 }, { "epoch": 0.5913004077933847, "grad_norm": 2.308932850545607, "learning_rate": 9.07169092207215e-06, "loss": 0.1432, "step": 3915 }, { "epoch": 0.5914514423803051, "grad_norm": 1.9411141158461105, "learning_rate": 9.071231889360461e-06, "loss": 0.1288, "step": 3916 }, { "epoch": 0.5916024769672255, "grad_norm": 2.45345415349523, "learning_rate": 9.070772754804666e-06, "loss": 0.2029, "step": 3917 }, { "epoch": 0.5917535115541459, "grad_norm": 2.0135433758056926, "learning_rate": 9.070313518416252e-06, "loss": 0.1677, "step": 3918 }, { "epoch": 0.5919045461410664, "grad_norm": 2.3757766185456566, "learning_rate": 9.069854180206705e-06, "loss": 0.1615, "step": 3919 }, { "epoch": 0.5920555807279867, "grad_norm": 2.16541666892117, "learning_rate": 9.069394740187519e-06, "loss": 0.1641, "step": 3920 }, { "epoch": 0.5922066153149071, "grad_norm": 2.259232584454808, "learning_rate": 9.068935198370181e-06, "loss": 0.1658, "step": 3921 }, { "epoch": 0.5923576499018275, "grad_norm": 1.9105187733050093, "learning_rate": 9.068475554766192e-06, "loss": 0.1575, "step": 3922 }, { "epoch": 0.5925086844887479, "grad_norm": 4.991940298630906, "learning_rate": 9.068015809387049e-06, "loss": 0.2654, "step": 3923 }, { "epoch": 0.5926597190756683, "grad_norm": 2.114253578853436, "learning_rate": 9.06755596224425e-06, "loss": 0.141, "step": 3924 }, { "epoch": 0.5928107536625887, "grad_norm": 1.854621121870983, "learning_rate": 9.067096013349304e-06, "loss": 0.1323, "step": 3925 }, { "epoch": 0.5929617882495092, "grad_norm": 3.602783912504149, "learning_rate": 9.066635962713712e-06, "loss": 0.1804, "step": 3926 }, { "epoch": 0.5931128228364295, "grad_norm": 2.05918648561385, "learning_rate": 9.066175810348983e-06, "loss": 0.1578, "step": 3927 }, { "epoch": 0.59326385742335, "grad_norm": 3.0653959278439937, "learning_rate": 9.06571555626663e-06, "loss": 0.1565, "step": 3928 }, { "epoch": 0.5934148920102703, "grad_norm": 2.410723750993003, "learning_rate": 9.065255200478165e-06, "loss": 0.1613, "step": 3929 }, { "epoch": 0.5935659265971908, "grad_norm": 2.1150560352973633, "learning_rate": 9.064794742995104e-06, "loss": 0.1605, "step": 3930 }, { "epoch": 0.5937169611841112, "grad_norm": 2.140093650311971, "learning_rate": 9.064334183828967e-06, "loss": 0.1624, "step": 3931 }, { "epoch": 0.5938679957710316, "grad_norm": 2.3244823783652957, "learning_rate": 9.063873522991275e-06, "loss": 0.1583, "step": 3932 }, { "epoch": 0.594019030357952, "grad_norm": 2.5535453323185493, "learning_rate": 9.06341276049355e-06, "loss": 0.1789, "step": 3933 }, { "epoch": 0.5941700649448723, "grad_norm": 2.077874184956061, "learning_rate": 9.062951896347318e-06, "loss": 0.1672, "step": 3934 }, { "epoch": 0.5943210995317928, "grad_norm": 2.2584048208533627, "learning_rate": 9.06249093056411e-06, "loss": 0.1651, "step": 3935 }, { "epoch": 0.5944721341187131, "grad_norm": 1.9124235930518014, "learning_rate": 9.062029863155455e-06, "loss": 0.1433, "step": 3936 }, { "epoch": 0.5946231687056336, "grad_norm": 1.9423622192539582, "learning_rate": 9.06156869413289e-06, "loss": 0.1521, "step": 3937 }, { "epoch": 0.594774203292554, "grad_norm": 1.7108596115567394, "learning_rate": 9.061107423507947e-06, "loss": 0.1269, "step": 3938 }, { "epoch": 0.5949252378794744, "grad_norm": 2.6124642238991598, "learning_rate": 9.06064605129217e-06, "loss": 0.1434, "step": 3939 }, { "epoch": 0.5950762724663948, "grad_norm": 2.1760227955843194, "learning_rate": 9.060184577497097e-06, "loss": 0.1441, "step": 3940 }, { "epoch": 0.5952273070533152, "grad_norm": 2.3048589191641278, "learning_rate": 9.059723002134273e-06, "loss": 0.1685, "step": 3941 }, { "epoch": 0.5953783416402356, "grad_norm": 1.9035331257633845, "learning_rate": 9.059261325215242e-06, "loss": 0.1577, "step": 3942 }, { "epoch": 0.5955293762271561, "grad_norm": 2.004844127600623, "learning_rate": 9.05879954675156e-06, "loss": 0.1367, "step": 3943 }, { "epoch": 0.5956804108140764, "grad_norm": 2.4372650249985615, "learning_rate": 9.05833766675477e-06, "loss": 0.1604, "step": 3944 }, { "epoch": 0.5958314454009969, "grad_norm": 2.69925391702319, "learning_rate": 9.057875685236434e-06, "loss": 0.1346, "step": 3945 }, { "epoch": 0.5959824799879172, "grad_norm": 2.2843315991858137, "learning_rate": 9.057413602208102e-06, "loss": 0.1484, "step": 3946 }, { "epoch": 0.5961335145748377, "grad_norm": 2.12294039262153, "learning_rate": 9.056951417681339e-06, "loss": 0.1503, "step": 3947 }, { "epoch": 0.596284549161758, "grad_norm": 3.0268299324993384, "learning_rate": 9.056489131667703e-06, "loss": 0.1657, "step": 3948 }, { "epoch": 0.5964355837486784, "grad_norm": 2.2018001325805656, "learning_rate": 9.056026744178758e-06, "loss": 0.1663, "step": 3949 }, { "epoch": 0.5965866183355989, "grad_norm": 2.1680431574527974, "learning_rate": 9.055564255226075e-06, "loss": 0.1654, "step": 3950 }, { "epoch": 0.5967376529225192, "grad_norm": 2.128834706799947, "learning_rate": 9.05510166482122e-06, "loss": 0.1475, "step": 3951 }, { "epoch": 0.5968886875094397, "grad_norm": 1.6537995431940957, "learning_rate": 9.054638972975764e-06, "loss": 0.1186, "step": 3952 }, { "epoch": 0.59703972209636, "grad_norm": 1.8622993402252368, "learning_rate": 9.054176179701285e-06, "loss": 0.1472, "step": 3953 }, { "epoch": 0.5971907566832805, "grad_norm": 2.1770006053158473, "learning_rate": 9.053713285009358e-06, "loss": 0.157, "step": 3954 }, { "epoch": 0.5973417912702009, "grad_norm": 1.8505906008558661, "learning_rate": 9.053250288911563e-06, "loss": 0.1528, "step": 3955 }, { "epoch": 0.5974928258571213, "grad_norm": 2.4500778121440043, "learning_rate": 9.05278719141948e-06, "loss": 0.172, "step": 3956 }, { "epoch": 0.5976438604440417, "grad_norm": 3.0566974558100353, "learning_rate": 9.052323992544698e-06, "loss": 0.1518, "step": 3957 }, { "epoch": 0.5977948950309621, "grad_norm": 2.071586293273783, "learning_rate": 9.051860692298803e-06, "loss": 0.1472, "step": 3958 }, { "epoch": 0.5979459296178825, "grad_norm": 1.6978673402709417, "learning_rate": 9.05139729069338e-06, "loss": 0.1301, "step": 3959 }, { "epoch": 0.5980969642048029, "grad_norm": 2.4143136139048496, "learning_rate": 9.050933787740026e-06, "loss": 0.1668, "step": 3960 }, { "epoch": 0.5982479987917233, "grad_norm": 2.303139402100339, "learning_rate": 9.050470183450335e-06, "loss": 0.1649, "step": 3961 }, { "epoch": 0.5983990333786438, "grad_norm": 2.686837546593851, "learning_rate": 9.050006477835903e-06, "loss": 0.1559, "step": 3962 }, { "epoch": 0.5985500679655641, "grad_norm": 2.060889941047801, "learning_rate": 9.049542670908331e-06, "loss": 0.1415, "step": 3963 }, { "epoch": 0.5987011025524845, "grad_norm": 2.3103617889492365, "learning_rate": 9.04907876267922e-06, "loss": 0.169, "step": 3964 }, { "epoch": 0.5988521371394049, "grad_norm": 2.4513862627664462, "learning_rate": 9.048614753160179e-06, "loss": 0.1608, "step": 3965 }, { "epoch": 0.5990031717263253, "grad_norm": 2.1471091505557163, "learning_rate": 9.04815064236281e-06, "loss": 0.1521, "step": 3966 }, { "epoch": 0.5991542063132458, "grad_norm": 2.162612273574919, "learning_rate": 9.047686430298726e-06, "loss": 0.1278, "step": 3967 }, { "epoch": 0.5993052409001661, "grad_norm": 1.8971104463835309, "learning_rate": 9.047222116979538e-06, "loss": 0.1497, "step": 3968 }, { "epoch": 0.5994562754870866, "grad_norm": 2.2867227213233816, "learning_rate": 9.046757702416863e-06, "loss": 0.1482, "step": 3969 }, { "epoch": 0.5996073100740069, "grad_norm": 3.2448343332998815, "learning_rate": 9.046293186622317e-06, "loss": 0.156, "step": 3970 }, { "epoch": 0.5997583446609274, "grad_norm": 1.9109807024699073, "learning_rate": 9.045828569607521e-06, "loss": 0.1429, "step": 3971 }, { "epoch": 0.5999093792478477, "grad_norm": 2.665861918239592, "learning_rate": 9.045363851384097e-06, "loss": 0.16, "step": 3972 }, { "epoch": 0.6000604138347682, "grad_norm": 2.963613140677246, "learning_rate": 9.044899031963672e-06, "loss": 0.1882, "step": 3973 }, { "epoch": 0.6002114484216886, "grad_norm": 1.9169630398958728, "learning_rate": 9.04443411135787e-06, "loss": 0.1448, "step": 3974 }, { "epoch": 0.600362483008609, "grad_norm": 2.3696889098875404, "learning_rate": 9.043969089578325e-06, "loss": 0.1568, "step": 3975 }, { "epoch": 0.6005135175955294, "grad_norm": 2.399777162938639, "learning_rate": 9.043503966636668e-06, "loss": 0.1581, "step": 3976 }, { "epoch": 0.6006645521824497, "grad_norm": 2.4983429234828067, "learning_rate": 9.043038742544534e-06, "loss": 0.1557, "step": 3977 }, { "epoch": 0.6008155867693702, "grad_norm": 2.35584747394662, "learning_rate": 9.042573417313561e-06, "loss": 0.1589, "step": 3978 }, { "epoch": 0.6009666213562906, "grad_norm": 3.0277519382962534, "learning_rate": 9.04210799095539e-06, "loss": 0.1597, "step": 3979 }, { "epoch": 0.601117655943211, "grad_norm": 2.3600326768822764, "learning_rate": 9.041642463481667e-06, "loss": 0.1577, "step": 3980 }, { "epoch": 0.6012686905301314, "grad_norm": 2.0871637137340744, "learning_rate": 9.04117683490403e-06, "loss": 0.1494, "step": 3981 }, { "epoch": 0.6014197251170518, "grad_norm": 2.260289048627219, "learning_rate": 9.040711105234133e-06, "loss": 0.171, "step": 3982 }, { "epoch": 0.6015707597039722, "grad_norm": 2.5874746288834958, "learning_rate": 9.040245274483624e-06, "loss": 0.1331, "step": 3983 }, { "epoch": 0.6017217942908926, "grad_norm": 3.3267521176946224, "learning_rate": 9.039779342664155e-06, "loss": 0.171, "step": 3984 }, { "epoch": 0.601872828877813, "grad_norm": 1.7789020199337637, "learning_rate": 9.039313309787385e-06, "loss": 0.1296, "step": 3985 }, { "epoch": 0.6020238634647335, "grad_norm": 2.580638626262454, "learning_rate": 9.03884717586497e-06, "loss": 0.1655, "step": 3986 }, { "epoch": 0.6021748980516538, "grad_norm": 2.450297494338703, "learning_rate": 9.038380940908569e-06, "loss": 0.1775, "step": 3987 }, { "epoch": 0.6023259326385743, "grad_norm": 3.284462073563801, "learning_rate": 9.037914604929849e-06, "loss": 0.1904, "step": 3988 }, { "epoch": 0.6024769672254946, "grad_norm": 1.9648610797259625, "learning_rate": 9.037448167940473e-06, "loss": 0.1436, "step": 3989 }, { "epoch": 0.602628001812415, "grad_norm": 4.022150292483688, "learning_rate": 9.03698162995211e-06, "loss": 0.144, "step": 3990 }, { "epoch": 0.6027790363993355, "grad_norm": 2.1216371490917063, "learning_rate": 9.036514990976428e-06, "loss": 0.1342, "step": 3991 }, { "epoch": 0.6029300709862558, "grad_norm": 2.1518547401524057, "learning_rate": 9.036048251025103e-06, "loss": 0.1917, "step": 3992 }, { "epoch": 0.6030811055731763, "grad_norm": 2.9649895681743437, "learning_rate": 9.035581410109812e-06, "loss": 0.177, "step": 3993 }, { "epoch": 0.6032321401600966, "grad_norm": 2.0830847939504644, "learning_rate": 9.035114468242232e-06, "loss": 0.1403, "step": 3994 }, { "epoch": 0.6033831747470171, "grad_norm": 2.05667450271556, "learning_rate": 9.034647425434043e-06, "loss": 0.1493, "step": 3995 }, { "epoch": 0.6035342093339374, "grad_norm": 2.0891677469023984, "learning_rate": 9.034180281696928e-06, "loss": 0.1753, "step": 3996 }, { "epoch": 0.6036852439208579, "grad_norm": 2.0549944996514666, "learning_rate": 9.033713037042575e-06, "loss": 0.1425, "step": 3997 }, { "epoch": 0.6038362785077783, "grad_norm": 1.9295376420889363, "learning_rate": 9.03324569148267e-06, "loss": 0.1426, "step": 3998 }, { "epoch": 0.6039873130946987, "grad_norm": 2.4838019115983534, "learning_rate": 9.032778245028906e-06, "loss": 0.1676, "step": 3999 }, { "epoch": 0.6041383476816191, "grad_norm": 2.7319157600674826, "learning_rate": 9.032310697692974e-06, "loss": 0.1888, "step": 4000 }, { "epoch": 0.6042893822685395, "grad_norm": 2.6255586382049807, "learning_rate": 9.031843049486572e-06, "loss": 0.1531, "step": 4001 }, { "epoch": 0.6044404168554599, "grad_norm": 2.0458442573495894, "learning_rate": 9.031375300421398e-06, "loss": 0.1365, "step": 4002 }, { "epoch": 0.6045914514423804, "grad_norm": 1.8465215723700563, "learning_rate": 9.030907450509151e-06, "loss": 0.1356, "step": 4003 }, { "epoch": 0.6047424860293007, "grad_norm": 2.692706825525375, "learning_rate": 9.030439499761539e-06, "loss": 0.182, "step": 4004 }, { "epoch": 0.6048935206162211, "grad_norm": 1.8163679398927388, "learning_rate": 9.029971448190264e-06, "loss": 0.145, "step": 4005 }, { "epoch": 0.6050445552031415, "grad_norm": 2.33796684205281, "learning_rate": 9.029503295807036e-06, "loss": 0.1474, "step": 4006 }, { "epoch": 0.6051955897900619, "grad_norm": 4.812839338001444, "learning_rate": 9.029035042623567e-06, "loss": 0.1755, "step": 4007 }, { "epoch": 0.6053466243769823, "grad_norm": 2.9572534679907965, "learning_rate": 9.02856668865157e-06, "loss": 0.1709, "step": 4008 }, { "epoch": 0.6054976589639027, "grad_norm": 1.9020528175404592, "learning_rate": 9.028098233902758e-06, "loss": 0.1466, "step": 4009 }, { "epoch": 0.6056486935508232, "grad_norm": 1.9186515588979596, "learning_rate": 9.027629678388854e-06, "loss": 0.1491, "step": 4010 }, { "epoch": 0.6057997281377435, "grad_norm": 2.004604560761134, "learning_rate": 9.027161022121579e-06, "loss": 0.1373, "step": 4011 }, { "epoch": 0.605950762724664, "grad_norm": 1.8940824519417403, "learning_rate": 9.026692265112653e-06, "loss": 0.1519, "step": 4012 }, { "epoch": 0.6061017973115843, "grad_norm": 2.8112612796268324, "learning_rate": 9.026223407373806e-06, "loss": 0.146, "step": 4013 }, { "epoch": 0.6062528318985048, "grad_norm": 2.613943225909374, "learning_rate": 9.025754448916764e-06, "loss": 0.1578, "step": 4014 }, { "epoch": 0.6064038664854252, "grad_norm": 2.5938331732456144, "learning_rate": 9.025285389753261e-06, "loss": 0.2079, "step": 4015 }, { "epoch": 0.6065549010723456, "grad_norm": 2.111157608248993, "learning_rate": 9.024816229895029e-06, "loss": 0.1364, "step": 4016 }, { "epoch": 0.606705935659266, "grad_norm": 2.720411262018249, "learning_rate": 9.024346969353804e-06, "loss": 0.2156, "step": 4017 }, { "epoch": 0.6068569702461863, "grad_norm": 2.0496240585552874, "learning_rate": 9.023877608141325e-06, "loss": 0.144, "step": 4018 }, { "epoch": 0.6070080048331068, "grad_norm": 2.4127710004141565, "learning_rate": 9.023408146269335e-06, "loss": 0.1661, "step": 4019 }, { "epoch": 0.6071590394200271, "grad_norm": 2.663233216867802, "learning_rate": 9.022938583749575e-06, "loss": 0.1783, "step": 4020 }, { "epoch": 0.6073100740069476, "grad_norm": 2.4296705319229592, "learning_rate": 9.022468920593794e-06, "loss": 0.1788, "step": 4021 }, { "epoch": 0.607461108593868, "grad_norm": 2.3610721740083997, "learning_rate": 9.02199915681374e-06, "loss": 0.1599, "step": 4022 }, { "epoch": 0.6076121431807884, "grad_norm": 2.6366622628435556, "learning_rate": 9.021529292421163e-06, "loss": 0.2073, "step": 4023 }, { "epoch": 0.6077631777677088, "grad_norm": 1.9857950356565874, "learning_rate": 9.02105932742782e-06, "loss": 0.1611, "step": 4024 }, { "epoch": 0.6079142123546292, "grad_norm": 1.970191200746481, "learning_rate": 9.020589261845465e-06, "loss": 0.1616, "step": 4025 }, { "epoch": 0.6080652469415496, "grad_norm": 2.3584012776420016, "learning_rate": 9.020119095685855e-06, "loss": 0.1592, "step": 4026 }, { "epoch": 0.6082162815284701, "grad_norm": 1.684007886443082, "learning_rate": 9.019648828960756e-06, "loss": 0.1201, "step": 4027 }, { "epoch": 0.6083673161153904, "grad_norm": 2.725792830122289, "learning_rate": 9.01917846168193e-06, "loss": 0.1761, "step": 4028 }, { "epoch": 0.6085183507023109, "grad_norm": 2.1687316910419394, "learning_rate": 9.018707993861145e-06, "loss": 0.1516, "step": 4029 }, { "epoch": 0.6086693852892312, "grad_norm": 1.8516871844201908, "learning_rate": 9.018237425510166e-06, "loss": 0.127, "step": 4030 }, { "epoch": 0.6088204198761517, "grad_norm": 1.9902567948357115, "learning_rate": 9.017766756640766e-06, "loss": 0.142, "step": 4031 }, { "epoch": 0.608971454463072, "grad_norm": 2.1545148911457295, "learning_rate": 9.017295987264722e-06, "loss": 0.1575, "step": 4032 }, { "epoch": 0.6091224890499924, "grad_norm": 2.3823629385717804, "learning_rate": 9.01682511739381e-06, "loss": 0.1575, "step": 4033 }, { "epoch": 0.6092735236369129, "grad_norm": 2.275652926897981, "learning_rate": 9.016354147039805e-06, "loss": 0.1374, "step": 4034 }, { "epoch": 0.6094245582238332, "grad_norm": 2.1318797720527587, "learning_rate": 9.015883076214493e-06, "loss": 0.1444, "step": 4035 }, { "epoch": 0.6095755928107537, "grad_norm": 2.7694006033828353, "learning_rate": 9.015411904929653e-06, "loss": 0.1667, "step": 4036 }, { "epoch": 0.609726627397674, "grad_norm": 2.0066515434835397, "learning_rate": 9.014940633197078e-06, "loss": 0.1468, "step": 4037 }, { "epoch": 0.6098776619845945, "grad_norm": 2.5148021573205446, "learning_rate": 9.014469261028554e-06, "loss": 0.1615, "step": 4038 }, { "epoch": 0.6100286965715149, "grad_norm": 2.195175407786468, "learning_rate": 9.013997788435873e-06, "loss": 0.1546, "step": 4039 }, { "epoch": 0.6101797311584353, "grad_norm": 2.202626838494507, "learning_rate": 9.013526215430828e-06, "loss": 0.1481, "step": 4040 }, { "epoch": 0.6103307657453557, "grad_norm": 2.409425761450817, "learning_rate": 9.013054542025215e-06, "loss": 0.1426, "step": 4041 }, { "epoch": 0.6104818003322761, "grad_norm": 3.527447871308372, "learning_rate": 9.012582768230838e-06, "loss": 0.1677, "step": 4042 }, { "epoch": 0.6106328349191965, "grad_norm": 2.1303170604387955, "learning_rate": 9.012110894059492e-06, "loss": 0.1615, "step": 4043 }, { "epoch": 0.6107838695061168, "grad_norm": 1.8573593464498186, "learning_rate": 9.011638919522987e-06, "loss": 0.1364, "step": 4044 }, { "epoch": 0.6109349040930373, "grad_norm": 2.1408267247516664, "learning_rate": 9.011166844633126e-06, "loss": 0.1543, "step": 4045 }, { "epoch": 0.6110859386799578, "grad_norm": 1.996502133354256, "learning_rate": 9.010694669401719e-06, "loss": 0.1373, "step": 4046 }, { "epoch": 0.6112369732668781, "grad_norm": 2.4621871329384994, "learning_rate": 9.01022239384058e-06, "loss": 0.1783, "step": 4047 }, { "epoch": 0.6113880078537985, "grad_norm": 1.6758302480650131, "learning_rate": 9.009750017961521e-06, "loss": 0.1315, "step": 4048 }, { "epoch": 0.6115390424407189, "grad_norm": 2.4572006268523046, "learning_rate": 9.009277541776358e-06, "loss": 0.1368, "step": 4049 }, { "epoch": 0.6116900770276393, "grad_norm": 2.15061918443058, "learning_rate": 9.00880496529691e-06, "loss": 0.1661, "step": 4050 }, { "epoch": 0.6118411116145598, "grad_norm": 2.0195455723997906, "learning_rate": 9.008332288535003e-06, "loss": 0.1422, "step": 4051 }, { "epoch": 0.6119921462014801, "grad_norm": 1.9526779714525346, "learning_rate": 9.007859511502457e-06, "loss": 0.1494, "step": 4052 }, { "epoch": 0.6121431807884006, "grad_norm": 2.1562906935558988, "learning_rate": 9.007386634211101e-06, "loss": 0.1905, "step": 4053 }, { "epoch": 0.6122942153753209, "grad_norm": 2.828106656205294, "learning_rate": 9.006913656672762e-06, "loss": 0.1626, "step": 4054 }, { "epoch": 0.6124452499622414, "grad_norm": 2.51617817219016, "learning_rate": 9.006440578899276e-06, "loss": 0.1658, "step": 4055 }, { "epoch": 0.6125962845491617, "grad_norm": 2.997831558724931, "learning_rate": 9.005967400902474e-06, "loss": 0.1814, "step": 4056 }, { "epoch": 0.6127473191360822, "grad_norm": 1.8165589616278126, "learning_rate": 9.00549412269419e-06, "loss": 0.1427, "step": 4057 }, { "epoch": 0.6128983537230026, "grad_norm": 2.5528649610834435, "learning_rate": 9.005020744286271e-06, "loss": 0.1828, "step": 4058 }, { "epoch": 0.613049388309923, "grad_norm": 1.9188431125165324, "learning_rate": 9.004547265690551e-06, "loss": 0.1403, "step": 4059 }, { "epoch": 0.6132004228968434, "grad_norm": 2.3111611795207274, "learning_rate": 9.004073686918879e-06, "loss": 0.1543, "step": 4060 }, { "epoch": 0.6133514574837637, "grad_norm": 1.8595981993281208, "learning_rate": 9.003600007983102e-06, "loss": 0.1339, "step": 4061 }, { "epoch": 0.6135024920706842, "grad_norm": 2.662150332631584, "learning_rate": 9.003126228895067e-06, "loss": 0.1445, "step": 4062 }, { "epoch": 0.6136535266576046, "grad_norm": 2.137223551021557, "learning_rate": 9.002652349666627e-06, "loss": 0.1506, "step": 4063 }, { "epoch": 0.613804561244525, "grad_norm": 2.199652426498521, "learning_rate": 9.002178370309636e-06, "loss": 0.169, "step": 4064 }, { "epoch": 0.6139555958314454, "grad_norm": 4.111963527500941, "learning_rate": 9.001704290835954e-06, "loss": 0.1495, "step": 4065 }, { "epoch": 0.6141066304183658, "grad_norm": 4.286685297388094, "learning_rate": 9.001230111257434e-06, "loss": 0.1857, "step": 4066 }, { "epoch": 0.6142576650052862, "grad_norm": 2.0674053153039766, "learning_rate": 9.000755831585942e-06, "loss": 0.1596, "step": 4067 }, { "epoch": 0.6144086995922066, "grad_norm": 2.4851463441156585, "learning_rate": 9.000281451833341e-06, "loss": 0.1619, "step": 4068 }, { "epoch": 0.614559734179127, "grad_norm": 3.1189710622433116, "learning_rate": 8.9998069720115e-06, "loss": 0.1572, "step": 4069 }, { "epoch": 0.6147107687660475, "grad_norm": 2.078115228152165, "learning_rate": 8.999332392132285e-06, "loss": 0.1705, "step": 4070 }, { "epoch": 0.6148618033529678, "grad_norm": 1.7396792898513447, "learning_rate": 8.99885771220757e-06, "loss": 0.115, "step": 4071 }, { "epoch": 0.6150128379398883, "grad_norm": 2.8564076708513415, "learning_rate": 8.998382932249231e-06, "loss": 0.154, "step": 4072 }, { "epoch": 0.6151638725268086, "grad_norm": 1.9335808867714022, "learning_rate": 8.997908052269144e-06, "loss": 0.1521, "step": 4073 }, { "epoch": 0.615314907113729, "grad_norm": 2.2396608342271658, "learning_rate": 8.997433072279185e-06, "loss": 0.1514, "step": 4074 }, { "epoch": 0.6154659417006495, "grad_norm": 2.152438635734398, "learning_rate": 8.996957992291238e-06, "loss": 0.1553, "step": 4075 }, { "epoch": 0.6156169762875698, "grad_norm": 2.9345023739378173, "learning_rate": 8.99648281231719e-06, "loss": 0.1474, "step": 4076 }, { "epoch": 0.6157680108744903, "grad_norm": 2.574596627368039, "learning_rate": 8.996007532368924e-06, "loss": 0.1548, "step": 4077 }, { "epoch": 0.6159190454614106, "grad_norm": 2.239688872112976, "learning_rate": 8.99553215245833e-06, "loss": 0.1543, "step": 4078 }, { "epoch": 0.6160700800483311, "grad_norm": 2.124920912386263, "learning_rate": 8.995056672597304e-06, "loss": 0.156, "step": 4079 }, { "epoch": 0.6162211146352514, "grad_norm": 2.077079456695254, "learning_rate": 8.994581092797735e-06, "loss": 0.1433, "step": 4080 }, { "epoch": 0.6163721492221719, "grad_norm": 2.0871266419904004, "learning_rate": 8.994105413071524e-06, "loss": 0.1508, "step": 4081 }, { "epoch": 0.6165231838090923, "grad_norm": 2.41513809305707, "learning_rate": 8.993629633430565e-06, "loss": 0.1793, "step": 4082 }, { "epoch": 0.6166742183960127, "grad_norm": 1.9247180711693541, "learning_rate": 8.993153753886766e-06, "loss": 0.1414, "step": 4083 }, { "epoch": 0.6168252529829331, "grad_norm": 2.866611804655844, "learning_rate": 8.992677774452029e-06, "loss": 0.1613, "step": 4084 }, { "epoch": 0.6169762875698535, "grad_norm": 1.8143537540949322, "learning_rate": 8.99220169513826e-06, "loss": 0.1437, "step": 4085 }, { "epoch": 0.6171273221567739, "grad_norm": 2.260535030286418, "learning_rate": 8.99172551595737e-06, "loss": 0.168, "step": 4086 }, { "epoch": 0.6172783567436944, "grad_norm": 1.9707324519546179, "learning_rate": 8.991249236921268e-06, "loss": 0.1611, "step": 4087 }, { "epoch": 0.6174293913306147, "grad_norm": 2.122334962276443, "learning_rate": 8.990772858041872e-06, "loss": 0.1535, "step": 4088 }, { "epoch": 0.6175804259175351, "grad_norm": 3.821027354754168, "learning_rate": 8.990296379331097e-06, "loss": 0.178, "step": 4089 }, { "epoch": 0.6177314605044555, "grad_norm": 1.7679998763948161, "learning_rate": 8.989819800800863e-06, "loss": 0.1455, "step": 4090 }, { "epoch": 0.6178824950913759, "grad_norm": 2.2428331030713804, "learning_rate": 8.98934312246309e-06, "loss": 0.1549, "step": 4091 }, { "epoch": 0.6180335296782963, "grad_norm": 2.2073789236055923, "learning_rate": 8.988866344329705e-06, "loss": 0.156, "step": 4092 }, { "epoch": 0.6181845642652167, "grad_norm": 2.346875657725054, "learning_rate": 8.988389466412633e-06, "loss": 0.1763, "step": 4093 }, { "epoch": 0.6183355988521372, "grad_norm": 2.4672867816530464, "learning_rate": 8.987912488723804e-06, "loss": 0.161, "step": 4094 }, { "epoch": 0.6184866334390575, "grad_norm": 2.1019218593730566, "learning_rate": 8.987435411275152e-06, "loss": 0.1433, "step": 4095 }, { "epoch": 0.618637668025978, "grad_norm": 1.8622158360593652, "learning_rate": 8.98695823407861e-06, "loss": 0.1449, "step": 4096 }, { "epoch": 0.6187887026128983, "grad_norm": 2.16691242467165, "learning_rate": 8.98648095714611e-06, "loss": 0.1384, "step": 4097 }, { "epoch": 0.6189397371998188, "grad_norm": 2.297747366200714, "learning_rate": 8.986003580489597e-06, "loss": 0.1413, "step": 4098 }, { "epoch": 0.6190907717867392, "grad_norm": 3.2575902297482227, "learning_rate": 8.985526104121012e-06, "loss": 0.1759, "step": 4099 }, { "epoch": 0.6192418063736596, "grad_norm": 2.6203589108210044, "learning_rate": 8.985048528052299e-06, "loss": 0.1498, "step": 4100 }, { "epoch": 0.61939284096058, "grad_norm": 2.113874093734047, "learning_rate": 8.984570852295404e-06, "loss": 0.165, "step": 4101 }, { "epoch": 0.6195438755475003, "grad_norm": 2.076445618456364, "learning_rate": 8.984093076862277e-06, "loss": 0.14, "step": 4102 }, { "epoch": 0.6196949101344208, "grad_norm": 2.3609096637638185, "learning_rate": 8.98361520176487e-06, "loss": 0.158, "step": 4103 }, { "epoch": 0.6198459447213411, "grad_norm": 2.7983202675960497, "learning_rate": 8.983137227015137e-06, "loss": 0.1923, "step": 4104 }, { "epoch": 0.6199969793082616, "grad_norm": 2.446342332173013, "learning_rate": 8.982659152625035e-06, "loss": 0.1572, "step": 4105 }, { "epoch": 0.620148013895182, "grad_norm": 3.7183088070671295, "learning_rate": 8.982180978606521e-06, "loss": 0.1739, "step": 4106 }, { "epoch": 0.6202990484821024, "grad_norm": 2.1854937356082926, "learning_rate": 8.98170270497156e-06, "loss": 0.1607, "step": 4107 }, { "epoch": 0.6204500830690228, "grad_norm": 1.7776619945198144, "learning_rate": 8.981224331732116e-06, "loss": 0.1296, "step": 4108 }, { "epoch": 0.6206011176559432, "grad_norm": 2.624712646971327, "learning_rate": 8.980745858900153e-06, "loss": 0.1597, "step": 4109 }, { "epoch": 0.6207521522428636, "grad_norm": 2.1608862080599947, "learning_rate": 8.980267286487645e-06, "loss": 0.1469, "step": 4110 }, { "epoch": 0.6209031868297841, "grad_norm": 2.538807518727895, "learning_rate": 8.979788614506559e-06, "loss": 0.1885, "step": 4111 }, { "epoch": 0.6210542214167044, "grad_norm": 2.1468176747516785, "learning_rate": 8.979309842968872e-06, "loss": 0.1388, "step": 4112 }, { "epoch": 0.6212052560036249, "grad_norm": 2.540681245073126, "learning_rate": 8.978830971886559e-06, "loss": 0.134, "step": 4113 }, { "epoch": 0.6213562905905452, "grad_norm": 2.271492712710375, "learning_rate": 8.9783520012716e-06, "loss": 0.1764, "step": 4114 }, { "epoch": 0.6215073251774657, "grad_norm": 2.47589444331829, "learning_rate": 8.977872931135976e-06, "loss": 0.1764, "step": 4115 }, { "epoch": 0.621658359764386, "grad_norm": 2.6497811650737826, "learning_rate": 8.977393761491675e-06, "loss": 0.1597, "step": 4116 }, { "epoch": 0.6218093943513064, "grad_norm": 2.6542232837248267, "learning_rate": 8.97691449235068e-06, "loss": 0.1567, "step": 4117 }, { "epoch": 0.6219604289382269, "grad_norm": 3.9178890671382725, "learning_rate": 8.976435123724978e-06, "loss": 0.1689, "step": 4118 }, { "epoch": 0.6221114635251472, "grad_norm": 2.0566467990560535, "learning_rate": 8.975955655626565e-06, "loss": 0.1386, "step": 4119 }, { "epoch": 0.6222624981120677, "grad_norm": 2.484504383583782, "learning_rate": 8.975476088067434e-06, "loss": 0.1868, "step": 4120 }, { "epoch": 0.622413532698988, "grad_norm": 2.2138061389667185, "learning_rate": 8.974996421059582e-06, "loss": 0.1582, "step": 4121 }, { "epoch": 0.6225645672859085, "grad_norm": 1.6070804884672085, "learning_rate": 8.974516654615006e-06, "loss": 0.118, "step": 4122 }, { "epoch": 0.6227156018728289, "grad_norm": 2.101654939633018, "learning_rate": 8.974036788745711e-06, "loss": 0.1382, "step": 4123 }, { "epoch": 0.6228666364597493, "grad_norm": 1.963596715813692, "learning_rate": 8.973556823463698e-06, "loss": 0.1472, "step": 4124 }, { "epoch": 0.6230176710466697, "grad_norm": 1.9048330055242768, "learning_rate": 8.973076758780976e-06, "loss": 0.1444, "step": 4125 }, { "epoch": 0.6231687056335901, "grad_norm": 1.8472126145168513, "learning_rate": 8.97259659470955e-06, "loss": 0.148, "step": 4126 }, { "epoch": 0.6233197402205105, "grad_norm": 1.7221919298590598, "learning_rate": 8.972116331261439e-06, "loss": 0.1183, "step": 4127 }, { "epoch": 0.6234707748074308, "grad_norm": 2.7198747490888344, "learning_rate": 8.971635968448649e-06, "loss": 0.1475, "step": 4128 }, { "epoch": 0.6236218093943513, "grad_norm": 1.7942716566801542, "learning_rate": 8.9711555062832e-06, "loss": 0.1384, "step": 4129 }, { "epoch": 0.6237728439812718, "grad_norm": 1.9199188425167113, "learning_rate": 8.970674944777113e-06, "loss": 0.1244, "step": 4130 }, { "epoch": 0.6239238785681921, "grad_norm": 2.0044699645605335, "learning_rate": 8.970194283942406e-06, "loss": 0.1482, "step": 4131 }, { "epoch": 0.6240749131551125, "grad_norm": 2.363427816372637, "learning_rate": 8.969713523791106e-06, "loss": 0.1775, "step": 4132 }, { "epoch": 0.6242259477420329, "grad_norm": 2.1846944893142126, "learning_rate": 8.969232664335237e-06, "loss": 0.1666, "step": 4133 }, { "epoch": 0.6243769823289533, "grad_norm": 2.314075733113248, "learning_rate": 8.96875170558683e-06, "loss": 0.177, "step": 4134 }, { "epoch": 0.6245280169158738, "grad_norm": 2.7003888815576436, "learning_rate": 8.968270647557914e-06, "loss": 0.1691, "step": 4135 }, { "epoch": 0.6246790515027941, "grad_norm": 1.8306330226960263, "learning_rate": 8.967789490260526e-06, "loss": 0.1348, "step": 4136 }, { "epoch": 0.6248300860897146, "grad_norm": 2.382841092983686, "learning_rate": 8.9673082337067e-06, "loss": 0.1596, "step": 4137 }, { "epoch": 0.6249811206766349, "grad_norm": 2.327999268885375, "learning_rate": 8.966826877908477e-06, "loss": 0.1575, "step": 4138 }, { "epoch": 0.6251321552635554, "grad_norm": 2.7921559657904895, "learning_rate": 8.966345422877896e-06, "loss": 0.2079, "step": 4139 }, { "epoch": 0.6252831898504757, "grad_norm": 1.9872353576141313, "learning_rate": 8.965863868627002e-06, "loss": 0.1529, "step": 4140 }, { "epoch": 0.6254342244373962, "grad_norm": 1.7812436870073645, "learning_rate": 8.965382215167842e-06, "loss": 0.1459, "step": 4141 }, { "epoch": 0.6255852590243166, "grad_norm": 2.6255527433091967, "learning_rate": 8.964900462512462e-06, "loss": 0.1775, "step": 4142 }, { "epoch": 0.625736293611237, "grad_norm": 1.9455133012976271, "learning_rate": 8.964418610672917e-06, "loss": 0.1378, "step": 4143 }, { "epoch": 0.6258873281981574, "grad_norm": 2.0057091957882704, "learning_rate": 8.96393665966126e-06, "loss": 0.1492, "step": 4144 }, { "epoch": 0.6260383627850777, "grad_norm": 2.4842368969242066, "learning_rate": 8.963454609489546e-06, "loss": 0.1644, "step": 4145 }, { "epoch": 0.6261893973719982, "grad_norm": 2.8212107672813183, "learning_rate": 8.962972460169834e-06, "loss": 0.1683, "step": 4146 }, { "epoch": 0.6263404319589186, "grad_norm": 3.295929944030436, "learning_rate": 8.962490211714185e-06, "loss": 0.2124, "step": 4147 }, { "epoch": 0.626491466545839, "grad_norm": 2.0451409677582166, "learning_rate": 8.962007864134663e-06, "loss": 0.1321, "step": 4148 }, { "epoch": 0.6266425011327594, "grad_norm": 1.8449571188237393, "learning_rate": 8.961525417443336e-06, "loss": 0.129, "step": 4149 }, { "epoch": 0.6267935357196798, "grad_norm": 2.1349733635863792, "learning_rate": 8.961042871652271e-06, "loss": 0.1613, "step": 4150 }, { "epoch": 0.6269445703066002, "grad_norm": 2.038259110999849, "learning_rate": 8.960560226773538e-06, "loss": 0.1524, "step": 4151 }, { "epoch": 0.6270956048935206, "grad_norm": 1.9775773412533897, "learning_rate": 8.960077482819212e-06, "loss": 0.1562, "step": 4152 }, { "epoch": 0.627246639480441, "grad_norm": 2.0518195537826442, "learning_rate": 8.95959463980137e-06, "loss": 0.141, "step": 4153 }, { "epoch": 0.6273976740673615, "grad_norm": 2.17731539720689, "learning_rate": 8.959111697732088e-06, "loss": 0.1497, "step": 4154 }, { "epoch": 0.6275487086542818, "grad_norm": 2.6996702990842, "learning_rate": 8.95862865662345e-06, "loss": 0.1862, "step": 4155 }, { "epoch": 0.6276997432412023, "grad_norm": 3.00152331666034, "learning_rate": 8.958145516487536e-06, "loss": 0.1711, "step": 4156 }, { "epoch": 0.6278507778281226, "grad_norm": 2.523064702179048, "learning_rate": 8.957662277336436e-06, "loss": 0.1631, "step": 4157 }, { "epoch": 0.628001812415043, "grad_norm": 2.5892315533622323, "learning_rate": 8.957178939182237e-06, "loss": 0.1779, "step": 4158 }, { "epoch": 0.6281528470019635, "grad_norm": 1.7913076462473045, "learning_rate": 8.956695502037027e-06, "loss": 0.1392, "step": 4159 }, { "epoch": 0.6283038815888838, "grad_norm": 4.166863044056374, "learning_rate": 8.956211965912904e-06, "loss": 0.1957, "step": 4160 }, { "epoch": 0.6284549161758043, "grad_norm": 1.9333096466653519, "learning_rate": 8.955728330821963e-06, "loss": 0.14, "step": 4161 }, { "epoch": 0.6286059507627246, "grad_norm": 1.6282955562610157, "learning_rate": 8.9552445967763e-06, "loss": 0.1187, "step": 4162 }, { "epoch": 0.6287569853496451, "grad_norm": 1.9395357817936068, "learning_rate": 8.954760763788018e-06, "loss": 0.1532, "step": 4163 }, { "epoch": 0.6289080199365654, "grad_norm": 4.488881596962528, "learning_rate": 8.954276831869219e-06, "loss": 0.1595, "step": 4164 }, { "epoch": 0.6290590545234859, "grad_norm": 2.5446036822881983, "learning_rate": 8.953792801032008e-06, "loss": 0.1698, "step": 4165 }, { "epoch": 0.6292100891104063, "grad_norm": 2.4247493325966833, "learning_rate": 8.953308671288496e-06, "loss": 0.1488, "step": 4166 }, { "epoch": 0.6293611236973267, "grad_norm": 3.2503053934027366, "learning_rate": 8.952824442650794e-06, "loss": 0.1605, "step": 4167 }, { "epoch": 0.6295121582842471, "grad_norm": 1.8652416810160815, "learning_rate": 8.952340115131012e-06, "loss": 0.1413, "step": 4168 }, { "epoch": 0.6296631928711675, "grad_norm": 2.7765049476569654, "learning_rate": 8.951855688741268e-06, "loss": 0.1965, "step": 4169 }, { "epoch": 0.6298142274580879, "grad_norm": 2.3923196854280695, "learning_rate": 8.951371163493679e-06, "loss": 0.15, "step": 4170 }, { "epoch": 0.6299652620450084, "grad_norm": 1.846309064388343, "learning_rate": 8.950886539400367e-06, "loss": 0.1485, "step": 4171 }, { "epoch": 0.6301162966319287, "grad_norm": 2.3923460980358424, "learning_rate": 8.950401816473455e-06, "loss": 0.1712, "step": 4172 }, { "epoch": 0.6302673312188491, "grad_norm": 1.9032689639028801, "learning_rate": 8.949916994725067e-06, "loss": 0.1456, "step": 4173 }, { "epoch": 0.6304183658057695, "grad_norm": 2.241440106178778, "learning_rate": 8.949432074167332e-06, "loss": 0.1635, "step": 4174 }, { "epoch": 0.6305694003926899, "grad_norm": 2.0465276844105817, "learning_rate": 8.948947054812383e-06, "loss": 0.1424, "step": 4175 }, { "epoch": 0.6307204349796103, "grad_norm": 1.9351990082045336, "learning_rate": 8.948461936672348e-06, "loss": 0.1395, "step": 4176 }, { "epoch": 0.6308714695665307, "grad_norm": 1.9897638078862256, "learning_rate": 8.947976719759366e-06, "loss": 0.1613, "step": 4177 }, { "epoch": 0.6310225041534512, "grad_norm": 1.7118000109453329, "learning_rate": 8.947491404085574e-06, "loss": 0.1182, "step": 4178 }, { "epoch": 0.6311735387403715, "grad_norm": 2.258432350647609, "learning_rate": 8.947005989663114e-06, "loss": 0.1731, "step": 4179 }, { "epoch": 0.631324573327292, "grad_norm": 2.494312757687663, "learning_rate": 8.946520476504125e-06, "loss": 0.1594, "step": 4180 }, { "epoch": 0.6314756079142123, "grad_norm": 2.532745048811284, "learning_rate": 8.946034864620758e-06, "loss": 0.1491, "step": 4181 }, { "epoch": 0.6316266425011328, "grad_norm": 1.9784386075307785, "learning_rate": 8.945549154025156e-06, "loss": 0.1513, "step": 4182 }, { "epoch": 0.6317776770880532, "grad_norm": 2.2019603034361315, "learning_rate": 8.945063344729469e-06, "loss": 0.1716, "step": 4183 }, { "epoch": 0.6319287116749736, "grad_norm": 1.6664066352977043, "learning_rate": 8.944577436745855e-06, "loss": 0.1385, "step": 4184 }, { "epoch": 0.632079746261894, "grad_norm": 2.223458878947008, "learning_rate": 8.944091430086466e-06, "loss": 0.1445, "step": 4185 }, { "epoch": 0.6322307808488143, "grad_norm": 2.1085702188577216, "learning_rate": 8.94360532476346e-06, "loss": 0.1487, "step": 4186 }, { "epoch": 0.6323818154357348, "grad_norm": 2.198385768871063, "learning_rate": 8.943119120788995e-06, "loss": 0.1721, "step": 4187 }, { "epoch": 0.6325328500226551, "grad_norm": 1.44215751907074, "learning_rate": 8.942632818175237e-06, "loss": 0.1045, "step": 4188 }, { "epoch": 0.6326838846095756, "grad_norm": 2.7429645315776736, "learning_rate": 8.94214641693435e-06, "loss": 0.1755, "step": 4189 }, { "epoch": 0.632834919196496, "grad_norm": 1.963220671683408, "learning_rate": 8.9416599170785e-06, "loss": 0.1524, "step": 4190 }, { "epoch": 0.6329859537834164, "grad_norm": 1.9231966889640078, "learning_rate": 8.94117331861986e-06, "loss": 0.1415, "step": 4191 }, { "epoch": 0.6331369883703368, "grad_norm": 2.004516828283572, "learning_rate": 8.9406866215706e-06, "loss": 0.1493, "step": 4192 }, { "epoch": 0.6332880229572572, "grad_norm": 2.2553267250101294, "learning_rate": 8.940199825942898e-06, "loss": 0.1466, "step": 4193 }, { "epoch": 0.6334390575441776, "grad_norm": 1.9773506729601602, "learning_rate": 8.939712931748927e-06, "loss": 0.1382, "step": 4194 }, { "epoch": 0.6335900921310981, "grad_norm": 2.043779723511122, "learning_rate": 8.93922593900087e-06, "loss": 0.1517, "step": 4195 }, { "epoch": 0.6337411267180184, "grad_norm": 2.0595613319265746, "learning_rate": 8.938738847710911e-06, "loss": 0.1488, "step": 4196 }, { "epoch": 0.6338921613049389, "grad_norm": 2.7206266343313645, "learning_rate": 8.93825165789123e-06, "loss": 0.1964, "step": 4197 }, { "epoch": 0.6340431958918592, "grad_norm": 2.621174105407693, "learning_rate": 8.937764369554018e-06, "loss": 0.1497, "step": 4198 }, { "epoch": 0.6341942304787797, "grad_norm": 2.604426829763241, "learning_rate": 8.937276982711463e-06, "loss": 0.1642, "step": 4199 }, { "epoch": 0.6343452650657, "grad_norm": 2.0889763865635116, "learning_rate": 8.93678949737576e-06, "loss": 0.1329, "step": 4200 }, { "epoch": 0.6344962996526204, "grad_norm": 2.0130390490875953, "learning_rate": 8.9363019135591e-06, "loss": 0.1344, "step": 4201 }, { "epoch": 0.6346473342395409, "grad_norm": 1.9061808361591168, "learning_rate": 8.935814231273683e-06, "loss": 0.1509, "step": 4202 }, { "epoch": 0.6347983688264612, "grad_norm": 2.161720553707755, "learning_rate": 8.935326450531705e-06, "loss": 0.1491, "step": 4203 }, { "epoch": 0.6349494034133817, "grad_norm": 2.1008179093027466, "learning_rate": 8.934838571345373e-06, "loss": 0.1718, "step": 4204 }, { "epoch": 0.635100438000302, "grad_norm": 1.9151704324041, "learning_rate": 8.934350593726887e-06, "loss": 0.1333, "step": 4205 }, { "epoch": 0.6352514725872225, "grad_norm": 2.013604791640087, "learning_rate": 8.93386251768846e-06, "loss": 0.1663, "step": 4206 }, { "epoch": 0.6354025071741429, "grad_norm": 2.218703633876006, "learning_rate": 8.933374343242296e-06, "loss": 0.137, "step": 4207 }, { "epoch": 0.6355535417610633, "grad_norm": 2.0123643845860215, "learning_rate": 8.932886070400609e-06, "loss": 0.1495, "step": 4208 }, { "epoch": 0.6357045763479837, "grad_norm": 3.0548925957648274, "learning_rate": 8.93239769917561e-06, "loss": 0.1595, "step": 4209 }, { "epoch": 0.6358556109349041, "grad_norm": 2.7711273179317937, "learning_rate": 8.931909229579522e-06, "loss": 0.1432, "step": 4210 }, { "epoch": 0.6360066455218245, "grad_norm": 2.1682737006803796, "learning_rate": 8.93142066162456e-06, "loss": 0.1258, "step": 4211 }, { "epoch": 0.6361576801087448, "grad_norm": 2.4824668273562156, "learning_rate": 8.930931995322949e-06, "loss": 0.1499, "step": 4212 }, { "epoch": 0.6363087146956653, "grad_norm": 1.9589425084850511, "learning_rate": 8.930443230686909e-06, "loss": 0.1323, "step": 4213 }, { "epoch": 0.6364597492825858, "grad_norm": 2.47027579600357, "learning_rate": 8.92995436772867e-06, "loss": 0.1727, "step": 4214 }, { "epoch": 0.6366107838695061, "grad_norm": 2.32888949145439, "learning_rate": 8.92946540646046e-06, "loss": 0.1841, "step": 4215 }, { "epoch": 0.6367618184564265, "grad_norm": 1.8626950158273405, "learning_rate": 8.928976346894512e-06, "loss": 0.1416, "step": 4216 }, { "epoch": 0.6369128530433469, "grad_norm": 2.298997654287189, "learning_rate": 8.928487189043058e-06, "loss": 0.1691, "step": 4217 }, { "epoch": 0.6370638876302673, "grad_norm": 2.649558097916303, "learning_rate": 8.927997932918336e-06, "loss": 0.1414, "step": 4218 }, { "epoch": 0.6372149222171878, "grad_norm": 1.8830235541811806, "learning_rate": 8.927508578532584e-06, "loss": 0.1267, "step": 4219 }, { "epoch": 0.6373659568041081, "grad_norm": 2.105217606550932, "learning_rate": 8.927019125898044e-06, "loss": 0.1448, "step": 4220 }, { "epoch": 0.6375169913910286, "grad_norm": 3.020454993778208, "learning_rate": 8.92652957502696e-06, "loss": 0.1562, "step": 4221 }, { "epoch": 0.6376680259779489, "grad_norm": 2.1540074606090913, "learning_rate": 8.92603992593158e-06, "loss": 0.1603, "step": 4222 }, { "epoch": 0.6378190605648694, "grad_norm": 2.258690648306076, "learning_rate": 8.925550178624149e-06, "loss": 0.1545, "step": 4223 }, { "epoch": 0.6379700951517897, "grad_norm": 1.9501049557322654, "learning_rate": 8.92506033311692e-06, "loss": 0.162, "step": 4224 }, { "epoch": 0.6381211297387102, "grad_norm": 2.0663942889723423, "learning_rate": 8.92457038942215e-06, "loss": 0.1569, "step": 4225 }, { "epoch": 0.6382721643256306, "grad_norm": 3.113956833195268, "learning_rate": 8.924080347552089e-06, "loss": 0.1659, "step": 4226 }, { "epoch": 0.638423198912551, "grad_norm": 1.890514249168562, "learning_rate": 8.923590207519001e-06, "loss": 0.1212, "step": 4227 }, { "epoch": 0.6385742334994714, "grad_norm": 2.652317286247439, "learning_rate": 8.923099969335145e-06, "loss": 0.1881, "step": 4228 }, { "epoch": 0.6387252680863917, "grad_norm": 2.2336880635502645, "learning_rate": 8.922609633012786e-06, "loss": 0.1745, "step": 4229 }, { "epoch": 0.6388763026733122, "grad_norm": 2.295691211211716, "learning_rate": 8.922119198564187e-06, "loss": 0.1765, "step": 4230 }, { "epoch": 0.6390273372602326, "grad_norm": 1.9572444676278191, "learning_rate": 8.92162866600162e-06, "loss": 0.1538, "step": 4231 }, { "epoch": 0.639178371847153, "grad_norm": 1.7744558560190244, "learning_rate": 8.921138035337353e-06, "loss": 0.1204, "step": 4232 }, { "epoch": 0.6393294064340734, "grad_norm": 3.253910941287437, "learning_rate": 8.92064730658366e-06, "loss": 0.1718, "step": 4233 }, { "epoch": 0.6394804410209938, "grad_norm": 1.9152176134854133, "learning_rate": 8.92015647975282e-06, "loss": 0.132, "step": 4234 }, { "epoch": 0.6396314756079142, "grad_norm": 1.9381449481661623, "learning_rate": 8.919665554857107e-06, "loss": 0.1452, "step": 4235 }, { "epoch": 0.6397825101948346, "grad_norm": 3.1129984773310935, "learning_rate": 8.919174531908803e-06, "loss": 0.1669, "step": 4236 }, { "epoch": 0.639933544781755, "grad_norm": 2.2800338575689323, "learning_rate": 8.918683410920192e-06, "loss": 0.1453, "step": 4237 }, { "epoch": 0.6400845793686755, "grad_norm": 2.0397396398904935, "learning_rate": 8.918192191903561e-06, "loss": 0.1671, "step": 4238 }, { "epoch": 0.6402356139555958, "grad_norm": 1.98584396486289, "learning_rate": 8.917700874871194e-06, "loss": 0.157, "step": 4239 }, { "epoch": 0.6403866485425163, "grad_norm": 2.0404857646300996, "learning_rate": 8.917209459835386e-06, "loss": 0.141, "step": 4240 }, { "epoch": 0.6405376831294366, "grad_norm": 2.518580538307042, "learning_rate": 8.916717946808428e-06, "loss": 0.1703, "step": 4241 }, { "epoch": 0.640688717716357, "grad_norm": 2.2730505403214587, "learning_rate": 8.916226335802616e-06, "loss": 0.1505, "step": 4242 }, { "epoch": 0.6408397523032775, "grad_norm": 2.18069197958625, "learning_rate": 8.915734626830246e-06, "loss": 0.1672, "step": 4243 }, { "epoch": 0.6409907868901978, "grad_norm": 2.8643539029572254, "learning_rate": 8.915242819903621e-06, "loss": 0.1984, "step": 4244 }, { "epoch": 0.6411418214771183, "grad_norm": 2.6816745331718934, "learning_rate": 8.914750915035042e-06, "loss": 0.1791, "step": 4245 }, { "epoch": 0.6412928560640386, "grad_norm": 2.4780952001590206, "learning_rate": 8.914258912236816e-06, "loss": 0.1739, "step": 4246 }, { "epoch": 0.6414438906509591, "grad_norm": 2.7064929493046486, "learning_rate": 8.913766811521248e-06, "loss": 0.1589, "step": 4247 }, { "epoch": 0.6415949252378794, "grad_norm": 3.4276552264643803, "learning_rate": 8.91327461290065e-06, "loss": 0.1805, "step": 4248 }, { "epoch": 0.6417459598247999, "grad_norm": 2.050413770147378, "learning_rate": 8.912782316387337e-06, "loss": 0.1461, "step": 4249 }, { "epoch": 0.6418969944117203, "grad_norm": 2.552885874380327, "learning_rate": 8.91228992199362e-06, "loss": 0.1674, "step": 4250 }, { "epoch": 0.6420480289986407, "grad_norm": 1.8667674843833721, "learning_rate": 8.911797429731819e-06, "loss": 0.1558, "step": 4251 }, { "epoch": 0.6421990635855611, "grad_norm": 1.9378842891034025, "learning_rate": 8.911304839614251e-06, "loss": 0.1423, "step": 4252 }, { "epoch": 0.6423500981724815, "grad_norm": 1.7766892617414813, "learning_rate": 8.91081215165324e-06, "loss": 0.1365, "step": 4253 }, { "epoch": 0.6425011327594019, "grad_norm": 2.4536268999183073, "learning_rate": 8.910319365861115e-06, "loss": 0.1832, "step": 4254 }, { "epoch": 0.6426521673463224, "grad_norm": 2.5851491579070345, "learning_rate": 8.909826482250197e-06, "loss": 0.1721, "step": 4255 }, { "epoch": 0.6428032019332427, "grad_norm": 2.4070498472517996, "learning_rate": 8.90933350083282e-06, "loss": 0.1556, "step": 4256 }, { "epoch": 0.6429542365201631, "grad_norm": 2.4822078758457238, "learning_rate": 8.908840421621312e-06, "loss": 0.1511, "step": 4257 }, { "epoch": 0.6431052711070835, "grad_norm": 1.8203552924616235, "learning_rate": 8.908347244628012e-06, "loss": 0.1372, "step": 4258 }, { "epoch": 0.6432563056940039, "grad_norm": 2.136307899415612, "learning_rate": 8.907853969865255e-06, "loss": 0.1478, "step": 4259 }, { "epoch": 0.6434073402809243, "grad_norm": 3.3232431789295918, "learning_rate": 8.90736059734538e-06, "loss": 0.1768, "step": 4260 }, { "epoch": 0.6435583748678447, "grad_norm": 2.2782687078833517, "learning_rate": 8.906867127080732e-06, "loss": 0.1246, "step": 4261 }, { "epoch": 0.6437094094547652, "grad_norm": 1.8416449274184896, "learning_rate": 8.90637355908365e-06, "loss": 0.117, "step": 4262 }, { "epoch": 0.6438604440416855, "grad_norm": 2.698307446561345, "learning_rate": 8.905879893366486e-06, "loss": 0.2165, "step": 4263 }, { "epoch": 0.644011478628606, "grad_norm": 2.7000818162402678, "learning_rate": 8.90538612994159e-06, "loss": 0.1551, "step": 4264 }, { "epoch": 0.6441625132155263, "grad_norm": 1.9426735798027612, "learning_rate": 8.904892268821308e-06, "loss": 0.1366, "step": 4265 }, { "epoch": 0.6443135478024468, "grad_norm": 2.3697386579142985, "learning_rate": 8.904398310017998e-06, "loss": 0.1485, "step": 4266 }, { "epoch": 0.6444645823893672, "grad_norm": 2.5060715124788215, "learning_rate": 8.903904253544016e-06, "loss": 0.1708, "step": 4267 }, { "epoch": 0.6446156169762876, "grad_norm": 2.367799066788, "learning_rate": 8.903410099411721e-06, "loss": 0.1651, "step": 4268 }, { "epoch": 0.644766651563208, "grad_norm": 2.760325219867596, "learning_rate": 8.902915847633475e-06, "loss": 0.1792, "step": 4269 }, { "epoch": 0.6449176861501283, "grad_norm": 2.3188471415345515, "learning_rate": 8.902421498221643e-06, "loss": 0.1783, "step": 4270 }, { "epoch": 0.6450687207370488, "grad_norm": 2.3102787692542317, "learning_rate": 8.901927051188589e-06, "loss": 0.1752, "step": 4271 }, { "epoch": 0.6452197553239691, "grad_norm": 2.138113845132557, "learning_rate": 8.901432506546682e-06, "loss": 0.1543, "step": 4272 }, { "epoch": 0.6453707899108896, "grad_norm": 2.5789307281623692, "learning_rate": 8.900937864308295e-06, "loss": 0.172, "step": 4273 }, { "epoch": 0.64552182449781, "grad_norm": 2.7018214717946236, "learning_rate": 8.900443124485803e-06, "loss": 0.1625, "step": 4274 }, { "epoch": 0.6456728590847304, "grad_norm": 2.2904000992644376, "learning_rate": 8.899948287091577e-06, "loss": 0.164, "step": 4275 }, { "epoch": 0.6458238936716508, "grad_norm": 2.037101267126868, "learning_rate": 8.899453352138001e-06, "loss": 0.1274, "step": 4276 }, { "epoch": 0.6459749282585712, "grad_norm": 2.8246852477391293, "learning_rate": 8.898958319637453e-06, "loss": 0.201, "step": 4277 }, { "epoch": 0.6461259628454916, "grad_norm": 2.890849443775828, "learning_rate": 8.898463189602318e-06, "loss": 0.121, "step": 4278 }, { "epoch": 0.6462769974324121, "grad_norm": 2.1410468907369453, "learning_rate": 8.897967962044981e-06, "loss": 0.154, "step": 4279 }, { "epoch": 0.6464280320193324, "grad_norm": 1.9922157323276528, "learning_rate": 8.89747263697783e-06, "loss": 0.1435, "step": 4280 }, { "epoch": 0.6465790666062529, "grad_norm": 3.255061068062084, "learning_rate": 8.896977214413258e-06, "loss": 0.1608, "step": 4281 }, { "epoch": 0.6467301011931732, "grad_norm": 2.364908171328864, "learning_rate": 8.896481694363656e-06, "loss": 0.1598, "step": 4282 }, { "epoch": 0.6468811357800937, "grad_norm": 2.149718877521335, "learning_rate": 8.895986076841424e-06, "loss": 0.153, "step": 4283 }, { "epoch": 0.647032170367014, "grad_norm": 1.8044845008314834, "learning_rate": 8.895490361858952e-06, "loss": 0.1276, "step": 4284 }, { "epoch": 0.6471832049539344, "grad_norm": 2.2752837786612, "learning_rate": 8.894994549428647e-06, "loss": 0.1342, "step": 4285 }, { "epoch": 0.6473342395408549, "grad_norm": 2.051870318361392, "learning_rate": 8.89449863956291e-06, "loss": 0.151, "step": 4286 }, { "epoch": 0.6474852741277752, "grad_norm": 2.27582090524114, "learning_rate": 8.894002632274147e-06, "loss": 0.1531, "step": 4287 }, { "epoch": 0.6476363087146957, "grad_norm": 1.7747613037200753, "learning_rate": 8.893506527574766e-06, "loss": 0.134, "step": 4288 }, { "epoch": 0.647787343301616, "grad_norm": 1.8748408683405882, "learning_rate": 8.893010325477179e-06, "loss": 0.144, "step": 4289 }, { "epoch": 0.6479383778885365, "grad_norm": 2.098778382359459, "learning_rate": 8.892514025993793e-06, "loss": 0.1365, "step": 4290 }, { "epoch": 0.6480894124754569, "grad_norm": 2.3918802278249234, "learning_rate": 8.892017629137029e-06, "loss": 0.162, "step": 4291 }, { "epoch": 0.6482404470623773, "grad_norm": 1.8855518643671734, "learning_rate": 8.891521134919303e-06, "loss": 0.1273, "step": 4292 }, { "epoch": 0.6483914816492977, "grad_norm": 2.496241206996064, "learning_rate": 8.891024543353032e-06, "loss": 0.1509, "step": 4293 }, { "epoch": 0.6485425162362181, "grad_norm": 2.2649933436337104, "learning_rate": 8.890527854450645e-06, "loss": 0.1594, "step": 4294 }, { "epoch": 0.6486935508231385, "grad_norm": 2.2735274673372015, "learning_rate": 8.890031068224561e-06, "loss": 0.1572, "step": 4295 }, { "epoch": 0.6488445854100588, "grad_norm": 2.203274317891883, "learning_rate": 8.88953418468721e-06, "loss": 0.1309, "step": 4296 }, { "epoch": 0.6489956199969793, "grad_norm": 1.977699059156529, "learning_rate": 8.88903720385102e-06, "loss": 0.1301, "step": 4297 }, { "epoch": 0.6491466545838998, "grad_norm": 2.2258362751186884, "learning_rate": 8.888540125728427e-06, "loss": 0.1707, "step": 4298 }, { "epoch": 0.6492976891708201, "grad_norm": 2.068963997898545, "learning_rate": 8.888042950331863e-06, "loss": 0.1399, "step": 4299 }, { "epoch": 0.6494487237577405, "grad_norm": 2.2305711322753408, "learning_rate": 8.887545677673763e-06, "loss": 0.1593, "step": 4300 }, { "epoch": 0.6495997583446609, "grad_norm": 2.4536841347057847, "learning_rate": 8.887048307766571e-06, "loss": 0.1724, "step": 4301 }, { "epoch": 0.6497507929315813, "grad_norm": 2.0786031247521772, "learning_rate": 8.886550840622728e-06, "loss": 0.1395, "step": 4302 }, { "epoch": 0.6499018275185018, "grad_norm": 2.165136312631865, "learning_rate": 8.886053276254676e-06, "loss": 0.1423, "step": 4303 }, { "epoch": 0.6500528621054221, "grad_norm": 2.4785494051905226, "learning_rate": 8.885555614674861e-06, "loss": 0.1646, "step": 4304 }, { "epoch": 0.6502038966923426, "grad_norm": 2.2058381755874423, "learning_rate": 8.885057855895737e-06, "loss": 0.1523, "step": 4305 }, { "epoch": 0.6503549312792629, "grad_norm": 2.0441436114322955, "learning_rate": 8.884559999929752e-06, "loss": 0.1656, "step": 4306 }, { "epoch": 0.6505059658661834, "grad_norm": 1.7815020715530618, "learning_rate": 8.884062046789363e-06, "loss": 0.1283, "step": 4307 }, { "epoch": 0.6506570004531037, "grad_norm": 2.371928958592013, "learning_rate": 8.883563996487023e-06, "loss": 0.1685, "step": 4308 }, { "epoch": 0.6508080350400242, "grad_norm": 1.8929249376337638, "learning_rate": 8.883065849035195e-06, "loss": 0.139, "step": 4309 }, { "epoch": 0.6509590696269446, "grad_norm": 1.7685388094908696, "learning_rate": 8.882567604446336e-06, "loss": 0.1393, "step": 4310 }, { "epoch": 0.651110104213865, "grad_norm": 2.1958130105577545, "learning_rate": 8.882069262732914e-06, "loss": 0.1431, "step": 4311 }, { "epoch": 0.6512611388007854, "grad_norm": 2.3401164384301847, "learning_rate": 8.881570823907392e-06, "loss": 0.1432, "step": 4312 }, { "epoch": 0.6514121733877057, "grad_norm": 2.7934435486148304, "learning_rate": 8.881072287982243e-06, "loss": 0.1845, "step": 4313 }, { "epoch": 0.6515632079746262, "grad_norm": 2.2263353891842343, "learning_rate": 8.880573654969933e-06, "loss": 0.1536, "step": 4314 }, { "epoch": 0.6517142425615466, "grad_norm": 2.4813335998859314, "learning_rate": 8.880074924882938e-06, "loss": 0.1446, "step": 4315 }, { "epoch": 0.651865277148467, "grad_norm": 2.038400082584396, "learning_rate": 8.879576097733736e-06, "loss": 0.1487, "step": 4316 }, { "epoch": 0.6520163117353874, "grad_norm": 2.7404652962256995, "learning_rate": 8.879077173534801e-06, "loss": 0.1656, "step": 4317 }, { "epoch": 0.6521673463223078, "grad_norm": 2.6065522209347893, "learning_rate": 8.878578152298616e-06, "loss": 0.1859, "step": 4318 }, { "epoch": 0.6523183809092282, "grad_norm": 3.237838153476075, "learning_rate": 8.878079034037665e-06, "loss": 0.1467, "step": 4319 }, { "epoch": 0.6524694154961486, "grad_norm": 1.9540361477270016, "learning_rate": 8.877579818764434e-06, "loss": 0.1602, "step": 4320 }, { "epoch": 0.652620450083069, "grad_norm": 2.2799589534556866, "learning_rate": 8.877080506491412e-06, "loss": 0.1654, "step": 4321 }, { "epoch": 0.6527714846699895, "grad_norm": 2.5178023301368775, "learning_rate": 8.876581097231085e-06, "loss": 0.1782, "step": 4322 }, { "epoch": 0.6529225192569098, "grad_norm": 1.6170132393830927, "learning_rate": 8.876081590995949e-06, "loss": 0.1285, "step": 4323 }, { "epoch": 0.6530735538438303, "grad_norm": 1.7034576959134697, "learning_rate": 8.875581987798502e-06, "loss": 0.1241, "step": 4324 }, { "epoch": 0.6532245884307506, "grad_norm": 1.8313253514615582, "learning_rate": 8.875082287651238e-06, "loss": 0.1276, "step": 4325 }, { "epoch": 0.653375623017671, "grad_norm": 2.5852777368059043, "learning_rate": 8.874582490566659e-06, "loss": 0.1739, "step": 4326 }, { "epoch": 0.6535266576045915, "grad_norm": 2.622565818635353, "learning_rate": 8.874082596557266e-06, "loss": 0.1728, "step": 4327 }, { "epoch": 0.6536776921915118, "grad_norm": 1.9723374643525744, "learning_rate": 8.873582605635564e-06, "loss": 0.1216, "step": 4328 }, { "epoch": 0.6538287267784323, "grad_norm": 2.1708122096079423, "learning_rate": 8.873082517814064e-06, "loss": 0.161, "step": 4329 }, { "epoch": 0.6539797613653526, "grad_norm": 2.8256778454560307, "learning_rate": 8.872582333105274e-06, "loss": 0.1714, "step": 4330 }, { "epoch": 0.6541307959522731, "grad_norm": 2.386742987617883, "learning_rate": 8.872082051521706e-06, "loss": 0.1745, "step": 4331 }, { "epoch": 0.6542818305391934, "grad_norm": 2.23270779539249, "learning_rate": 8.871581673075874e-06, "loss": 0.1556, "step": 4332 }, { "epoch": 0.6544328651261139, "grad_norm": 1.698946158923877, "learning_rate": 8.871081197780298e-06, "loss": 0.1447, "step": 4333 }, { "epoch": 0.6545838997130343, "grad_norm": 2.6805968395847812, "learning_rate": 8.870580625647496e-06, "loss": 0.1466, "step": 4334 }, { "epoch": 0.6547349342999547, "grad_norm": 1.6808313233632308, "learning_rate": 8.87007995668999e-06, "loss": 0.1307, "step": 4335 }, { "epoch": 0.6548859688868751, "grad_norm": 1.8705985625025014, "learning_rate": 8.869579190920303e-06, "loss": 0.1552, "step": 4336 }, { "epoch": 0.6550370034737955, "grad_norm": 1.4759107048848423, "learning_rate": 8.869078328350964e-06, "loss": 0.1101, "step": 4337 }, { "epoch": 0.6551880380607159, "grad_norm": 2.3656859631934752, "learning_rate": 8.868577368994502e-06, "loss": 0.1438, "step": 4338 }, { "epoch": 0.6553390726476364, "grad_norm": 2.009151613581905, "learning_rate": 8.868076312863449e-06, "loss": 0.1663, "step": 4339 }, { "epoch": 0.6554901072345567, "grad_norm": 2.200931048944445, "learning_rate": 8.867575159970339e-06, "loss": 0.1459, "step": 4340 }, { "epoch": 0.6556411418214771, "grad_norm": 4.10245194641948, "learning_rate": 8.867073910327708e-06, "loss": 0.1844, "step": 4341 }, { "epoch": 0.6557921764083975, "grad_norm": 2.662001790774113, "learning_rate": 8.866572563948095e-06, "loss": 0.1596, "step": 4342 }, { "epoch": 0.6559432109953179, "grad_norm": 2.289798288013484, "learning_rate": 8.866071120844041e-06, "loss": 0.1525, "step": 4343 }, { "epoch": 0.6560942455822383, "grad_norm": 2.375620181622449, "learning_rate": 8.865569581028095e-06, "loss": 0.1384, "step": 4344 }, { "epoch": 0.6562452801691587, "grad_norm": 2.271473388293234, "learning_rate": 8.865067944512794e-06, "loss": 0.1419, "step": 4345 }, { "epoch": 0.6563963147560792, "grad_norm": 2.427302807263998, "learning_rate": 8.864566211310695e-06, "loss": 0.1313, "step": 4346 }, { "epoch": 0.6565473493429995, "grad_norm": 2.707028398023797, "learning_rate": 8.864064381434346e-06, "loss": 0.1574, "step": 4347 }, { "epoch": 0.65669838392992, "grad_norm": 2.2003273945084, "learning_rate": 8.8635624548963e-06, "loss": 0.1725, "step": 4348 }, { "epoch": 0.6568494185168403, "grad_norm": 1.9463519246534262, "learning_rate": 8.863060431709112e-06, "loss": 0.1495, "step": 4349 }, { "epoch": 0.6570004531037608, "grad_norm": 4.167272871825305, "learning_rate": 8.862558311885345e-06, "loss": 0.1709, "step": 4350 }, { "epoch": 0.6571514876906812, "grad_norm": 2.284000429912845, "learning_rate": 8.862056095437554e-06, "loss": 0.1664, "step": 4351 }, { "epoch": 0.6573025222776016, "grad_norm": 1.9422533501293686, "learning_rate": 8.861553782378306e-06, "loss": 0.1449, "step": 4352 }, { "epoch": 0.657453556864522, "grad_norm": 1.8622771871787542, "learning_rate": 8.861051372720163e-06, "loss": 0.1668, "step": 4353 }, { "epoch": 0.6576045914514423, "grad_norm": 1.8828253129542012, "learning_rate": 8.8605488664757e-06, "loss": 0.1305, "step": 4354 }, { "epoch": 0.6577556260383628, "grad_norm": 2.6801067679345953, "learning_rate": 8.860046263657481e-06, "loss": 0.179, "step": 4355 }, { "epoch": 0.6579066606252831, "grad_norm": 2.1932838709841334, "learning_rate": 8.859543564278079e-06, "loss": 0.1533, "step": 4356 }, { "epoch": 0.6580576952122036, "grad_norm": 1.901542113438867, "learning_rate": 8.859040768350074e-06, "loss": 0.1523, "step": 4357 }, { "epoch": 0.658208729799124, "grad_norm": 2.1206775267751974, "learning_rate": 8.85853787588604e-06, "loss": 0.1535, "step": 4358 }, { "epoch": 0.6583597643860444, "grad_norm": 2.447912294756573, "learning_rate": 8.858034886898558e-06, "loss": 0.1698, "step": 4359 }, { "epoch": 0.6585107989729648, "grad_norm": 1.9974538528135077, "learning_rate": 8.85753180140021e-06, "loss": 0.146, "step": 4360 }, { "epoch": 0.6586618335598852, "grad_norm": 2.4968862515991606, "learning_rate": 8.857028619403583e-06, "loss": 0.1501, "step": 4361 }, { "epoch": 0.6588128681468056, "grad_norm": 2.427432336635907, "learning_rate": 8.856525340921262e-06, "loss": 0.1562, "step": 4362 }, { "epoch": 0.6589639027337261, "grad_norm": 2.1284587463167637, "learning_rate": 8.856021965965838e-06, "loss": 0.1383, "step": 4363 }, { "epoch": 0.6591149373206464, "grad_norm": 2.3294275116384253, "learning_rate": 8.855518494549901e-06, "loss": 0.178, "step": 4364 }, { "epoch": 0.6592659719075669, "grad_norm": 2.399664928570391, "learning_rate": 8.85501492668605e-06, "loss": 0.1738, "step": 4365 }, { "epoch": 0.6594170064944872, "grad_norm": 2.051766288707087, "learning_rate": 8.854511262386878e-06, "loss": 0.1556, "step": 4366 }, { "epoch": 0.6595680410814077, "grad_norm": 2.3875611630958797, "learning_rate": 8.854007501664988e-06, "loss": 0.149, "step": 4367 }, { "epoch": 0.659719075668328, "grad_norm": 2.0188176227269197, "learning_rate": 8.85350364453298e-06, "loss": 0.1531, "step": 4368 }, { "epoch": 0.6598701102552484, "grad_norm": 2.0548658040993044, "learning_rate": 8.852999691003456e-06, "loss": 0.161, "step": 4369 }, { "epoch": 0.6600211448421689, "grad_norm": 2.0090351529191164, "learning_rate": 8.852495641089026e-06, "loss": 0.1329, "step": 4370 }, { "epoch": 0.6601721794290892, "grad_norm": 2.258326102538151, "learning_rate": 8.851991494802299e-06, "loss": 0.1381, "step": 4371 }, { "epoch": 0.6603232140160097, "grad_norm": 4.6302154700305715, "learning_rate": 8.851487252155882e-06, "loss": 0.2062, "step": 4372 }, { "epoch": 0.66047424860293, "grad_norm": 2.489192920890474, "learning_rate": 8.850982913162396e-06, "loss": 0.1731, "step": 4373 }, { "epoch": 0.6606252831898505, "grad_norm": 2.7529591992537266, "learning_rate": 8.850478477834453e-06, "loss": 0.162, "step": 4374 }, { "epoch": 0.6607763177767709, "grad_norm": 2.1159912409705233, "learning_rate": 8.84997394618467e-06, "loss": 0.1502, "step": 4375 }, { "epoch": 0.6609273523636913, "grad_norm": 2.0699070389994416, "learning_rate": 8.849469318225671e-06, "loss": 0.1566, "step": 4376 }, { "epoch": 0.6610783869506117, "grad_norm": 2.3066975096076066, "learning_rate": 8.848964593970082e-06, "loss": 0.1629, "step": 4377 }, { "epoch": 0.6612294215375321, "grad_norm": 2.472352128915734, "learning_rate": 8.848459773430522e-06, "loss": 0.17, "step": 4378 }, { "epoch": 0.6613804561244525, "grad_norm": 2.2870069329150766, "learning_rate": 8.847954856619625e-06, "loss": 0.1589, "step": 4379 }, { "epoch": 0.6615314907113728, "grad_norm": 2.788300150595566, "learning_rate": 8.847449843550022e-06, "loss": 0.1675, "step": 4380 }, { "epoch": 0.6616825252982933, "grad_norm": 1.8672800671527914, "learning_rate": 8.846944734234342e-06, "loss": 0.1299, "step": 4381 }, { "epoch": 0.6618335598852138, "grad_norm": 2.74813198861976, "learning_rate": 8.846439528685225e-06, "loss": 0.1474, "step": 4382 }, { "epoch": 0.6619845944721341, "grad_norm": 2.5094903373887245, "learning_rate": 8.845934226915304e-06, "loss": 0.1678, "step": 4383 }, { "epoch": 0.6621356290590545, "grad_norm": 2.522555838480101, "learning_rate": 8.845428828937223e-06, "loss": 0.161, "step": 4384 }, { "epoch": 0.6622866636459749, "grad_norm": 2.222571974957588, "learning_rate": 8.844923334763626e-06, "loss": 0.1636, "step": 4385 }, { "epoch": 0.6624376982328953, "grad_norm": 2.115316470375366, "learning_rate": 8.844417744407155e-06, "loss": 0.1495, "step": 4386 }, { "epoch": 0.6625887328198158, "grad_norm": 1.7497839650480207, "learning_rate": 8.84391205788046e-06, "loss": 0.1451, "step": 4387 }, { "epoch": 0.6627397674067361, "grad_norm": 2.0553625303340914, "learning_rate": 8.843406275196189e-06, "loss": 0.1511, "step": 4388 }, { "epoch": 0.6628908019936566, "grad_norm": 1.9457619678851557, "learning_rate": 8.842900396366995e-06, "loss": 0.1436, "step": 4389 }, { "epoch": 0.6630418365805769, "grad_norm": 2.0633965543289605, "learning_rate": 8.842394421405534e-06, "loss": 0.1627, "step": 4390 }, { "epoch": 0.6631928711674974, "grad_norm": 2.086194053846992, "learning_rate": 8.841888350324463e-06, "loss": 0.139, "step": 4391 }, { "epoch": 0.6633439057544177, "grad_norm": 2.2150312514194788, "learning_rate": 8.84138218313644e-06, "loss": 0.1533, "step": 4392 }, { "epoch": 0.6634949403413382, "grad_norm": 2.1354937767262223, "learning_rate": 8.840875919854128e-06, "loss": 0.1701, "step": 4393 }, { "epoch": 0.6636459749282586, "grad_norm": 2.654309080316272, "learning_rate": 8.840369560490191e-06, "loss": 0.1535, "step": 4394 }, { "epoch": 0.663797009515179, "grad_norm": 2.1100934322910847, "learning_rate": 8.839863105057299e-06, "loss": 0.1401, "step": 4395 }, { "epoch": 0.6639480441020994, "grad_norm": 1.8665230653051808, "learning_rate": 8.839356553568118e-06, "loss": 0.1394, "step": 4396 }, { "epoch": 0.6640990786890197, "grad_norm": 1.7821429678459415, "learning_rate": 8.838849906035319e-06, "loss": 0.1377, "step": 4397 }, { "epoch": 0.6642501132759402, "grad_norm": 2.0778514651280076, "learning_rate": 8.838343162471579e-06, "loss": 0.1499, "step": 4398 }, { "epoch": 0.6644011478628606, "grad_norm": 1.724667207848538, "learning_rate": 8.837836322889573e-06, "loss": 0.1347, "step": 4399 }, { "epoch": 0.664552182449781, "grad_norm": 2.049443636939261, "learning_rate": 8.837329387301978e-06, "loss": 0.1511, "step": 4400 }, { "epoch": 0.6647032170367014, "grad_norm": 2.0836318621507983, "learning_rate": 8.836822355721478e-06, "loss": 0.1546, "step": 4401 }, { "epoch": 0.6648542516236218, "grad_norm": 1.7870728755726624, "learning_rate": 8.836315228160757e-06, "loss": 0.1413, "step": 4402 }, { "epoch": 0.6650052862105422, "grad_norm": 1.8967067537777553, "learning_rate": 8.835808004632497e-06, "loss": 0.1309, "step": 4403 }, { "epoch": 0.6651563207974626, "grad_norm": 1.9788798716741787, "learning_rate": 8.835300685149393e-06, "loss": 0.1427, "step": 4404 }, { "epoch": 0.665307355384383, "grad_norm": 1.9073497522793377, "learning_rate": 8.834793269724128e-06, "loss": 0.1581, "step": 4405 }, { "epoch": 0.6654583899713035, "grad_norm": 2.0718329777799003, "learning_rate": 8.834285758369403e-06, "loss": 0.1329, "step": 4406 }, { "epoch": 0.6656094245582238, "grad_norm": 2.2103306819683906, "learning_rate": 8.83377815109791e-06, "loss": 0.1557, "step": 4407 }, { "epoch": 0.6657604591451443, "grad_norm": 2.202836197527056, "learning_rate": 8.833270447922346e-06, "loss": 0.1532, "step": 4408 }, { "epoch": 0.6659114937320646, "grad_norm": 1.9405246316666622, "learning_rate": 8.832762648855412e-06, "loss": 0.1449, "step": 4409 }, { "epoch": 0.666062528318985, "grad_norm": 1.9538078168235915, "learning_rate": 8.832254753909812e-06, "loss": 0.1458, "step": 4410 }, { "epoch": 0.6662135629059055, "grad_norm": 1.9054084060564722, "learning_rate": 8.831746763098252e-06, "loss": 0.1221, "step": 4411 }, { "epoch": 0.6663645974928258, "grad_norm": 2.1238595495403576, "learning_rate": 8.831238676433437e-06, "loss": 0.1601, "step": 4412 }, { "epoch": 0.6665156320797463, "grad_norm": 2.301415528685143, "learning_rate": 8.830730493928078e-06, "loss": 0.1597, "step": 4413 }, { "epoch": 0.6666666666666666, "grad_norm": 2.356016333116334, "learning_rate": 8.83022221559489e-06, "loss": 0.1577, "step": 4414 }, { "epoch": 0.6668177012535871, "grad_norm": 1.843554706945789, "learning_rate": 8.829713841446585e-06, "loss": 0.1404, "step": 4415 }, { "epoch": 0.6669687358405075, "grad_norm": 2.6425917766726696, "learning_rate": 8.829205371495883e-06, "loss": 0.151, "step": 4416 }, { "epoch": 0.6671197704274279, "grad_norm": 1.7643698626272746, "learning_rate": 8.828696805755498e-06, "loss": 0.1431, "step": 4417 }, { "epoch": 0.6672708050143483, "grad_norm": 3.3372626406152235, "learning_rate": 8.828188144238161e-06, "loss": 0.1479, "step": 4418 }, { "epoch": 0.6674218396012687, "grad_norm": 2.6263136559699563, "learning_rate": 8.827679386956586e-06, "loss": 0.1764, "step": 4419 }, { "epoch": 0.6675728741881891, "grad_norm": 2.1046041785316554, "learning_rate": 8.827170533923509e-06, "loss": 0.1474, "step": 4420 }, { "epoch": 0.6677239087751095, "grad_norm": 2.1221163589685133, "learning_rate": 8.826661585151654e-06, "loss": 0.1587, "step": 4421 }, { "epoch": 0.6678749433620299, "grad_norm": 1.8304915308879046, "learning_rate": 8.826152540653753e-06, "loss": 0.1322, "step": 4422 }, { "epoch": 0.6680259779489504, "grad_norm": 2.0416835351272367, "learning_rate": 8.825643400442542e-06, "loss": 0.1476, "step": 4423 }, { "epoch": 0.6681770125358707, "grad_norm": 3.2193224809052867, "learning_rate": 8.825134164530757e-06, "loss": 0.1442, "step": 4424 }, { "epoch": 0.6683280471227911, "grad_norm": 2.4651279416842624, "learning_rate": 8.824624832931135e-06, "loss": 0.1879, "step": 4425 }, { "epoch": 0.6684790817097115, "grad_norm": 2.075456166975253, "learning_rate": 8.82411540565642e-06, "loss": 0.1367, "step": 4426 }, { "epoch": 0.6686301162966319, "grad_norm": 1.7317459401694708, "learning_rate": 8.823605882719353e-06, "loss": 0.129, "step": 4427 }, { "epoch": 0.6687811508835524, "grad_norm": 2.037749933891493, "learning_rate": 8.823096264132682e-06, "loss": 0.1545, "step": 4428 }, { "epoch": 0.6689321854704727, "grad_norm": 2.5653932385700364, "learning_rate": 8.822586549909154e-06, "loss": 0.1953, "step": 4429 }, { "epoch": 0.6690832200573932, "grad_norm": 2.1905027285276284, "learning_rate": 8.82207674006152e-06, "loss": 0.1613, "step": 4430 }, { "epoch": 0.6692342546443135, "grad_norm": 2.359244427976905, "learning_rate": 8.821566834602534e-06, "loss": 0.1519, "step": 4431 }, { "epoch": 0.669385289231234, "grad_norm": 2.254472303061926, "learning_rate": 8.821056833544953e-06, "loss": 0.176, "step": 4432 }, { "epoch": 0.6695363238181543, "grad_norm": 2.091551367760987, "learning_rate": 8.820546736901529e-06, "loss": 0.1578, "step": 4433 }, { "epoch": 0.6696873584050748, "grad_norm": 1.8172618371580669, "learning_rate": 8.820036544685028e-06, "loss": 0.1492, "step": 4434 }, { "epoch": 0.6698383929919952, "grad_norm": 2.545501996327014, "learning_rate": 8.81952625690821e-06, "loss": 0.191, "step": 4435 }, { "epoch": 0.6699894275789156, "grad_norm": 1.9770219980349972, "learning_rate": 8.819015873583845e-06, "loss": 0.1492, "step": 4436 }, { "epoch": 0.670140462165836, "grad_norm": 2.4846900021955998, "learning_rate": 8.818505394724693e-06, "loss": 0.1814, "step": 4437 }, { "epoch": 0.6702914967527563, "grad_norm": 2.1216304437532476, "learning_rate": 8.817994820343526e-06, "loss": 0.1459, "step": 4438 }, { "epoch": 0.6704425313396768, "grad_norm": 2.4051133085392498, "learning_rate": 8.817484150453122e-06, "loss": 0.1773, "step": 4439 }, { "epoch": 0.6705935659265972, "grad_norm": 2.6397247551052234, "learning_rate": 8.816973385066251e-06, "loss": 0.1735, "step": 4440 }, { "epoch": 0.6707446005135176, "grad_norm": 2.0768929700784224, "learning_rate": 8.816462524195691e-06, "loss": 0.1481, "step": 4441 }, { "epoch": 0.670895635100438, "grad_norm": 2.3861399390264224, "learning_rate": 8.81595156785422e-06, "loss": 0.1482, "step": 4442 }, { "epoch": 0.6710466696873584, "grad_norm": 2.3888354799493237, "learning_rate": 8.81544051605462e-06, "loss": 0.1581, "step": 4443 }, { "epoch": 0.6711977042742788, "grad_norm": 2.008237019329116, "learning_rate": 8.814929368809677e-06, "loss": 0.1624, "step": 4444 }, { "epoch": 0.6713487388611992, "grad_norm": 2.0583354860078953, "learning_rate": 8.814418126132177e-06, "loss": 0.1653, "step": 4445 }, { "epoch": 0.6714997734481196, "grad_norm": 2.068693725208457, "learning_rate": 8.813906788034909e-06, "loss": 0.1186, "step": 4446 }, { "epoch": 0.6716508080350401, "grad_norm": 1.9452046481417657, "learning_rate": 8.813395354530665e-06, "loss": 0.1479, "step": 4447 }, { "epoch": 0.6718018426219604, "grad_norm": 1.9493820209786579, "learning_rate": 8.812883825632236e-06, "loss": 0.1297, "step": 4448 }, { "epoch": 0.6719528772088809, "grad_norm": 2.069266498072143, "learning_rate": 8.812372201352422e-06, "loss": 0.1348, "step": 4449 }, { "epoch": 0.6721039117958012, "grad_norm": 2.5279998461446596, "learning_rate": 8.81186048170402e-06, "loss": 0.1758, "step": 4450 }, { "epoch": 0.6722549463827217, "grad_norm": 2.210549757591047, "learning_rate": 8.81134866669983e-06, "loss": 0.1437, "step": 4451 }, { "epoch": 0.6724059809696421, "grad_norm": 2.4854849333853335, "learning_rate": 8.810836756352657e-06, "loss": 0.1666, "step": 4452 }, { "epoch": 0.6725570155565624, "grad_norm": 2.799743233527777, "learning_rate": 8.810324750675306e-06, "loss": 0.1464, "step": 4453 }, { "epoch": 0.6727080501434829, "grad_norm": 2.028121387308886, "learning_rate": 8.809812649680584e-06, "loss": 0.1485, "step": 4454 }, { "epoch": 0.6728590847304032, "grad_norm": 2.194937599436245, "learning_rate": 8.809300453381302e-06, "loss": 0.1793, "step": 4455 }, { "epoch": 0.6730101193173237, "grad_norm": 2.7599389933836536, "learning_rate": 8.808788161790274e-06, "loss": 0.1649, "step": 4456 }, { "epoch": 0.673161153904244, "grad_norm": 2.320480414512147, "learning_rate": 8.808275774920314e-06, "loss": 0.1613, "step": 4457 }, { "epoch": 0.6733121884911645, "grad_norm": 2.4681248492823338, "learning_rate": 8.807763292784242e-06, "loss": 0.1686, "step": 4458 }, { "epoch": 0.6734632230780849, "grad_norm": 1.851620384580475, "learning_rate": 8.807250715394874e-06, "loss": 0.1272, "step": 4459 }, { "epoch": 0.6736142576650053, "grad_norm": 2.7867401561109437, "learning_rate": 8.806738042765036e-06, "loss": 0.1674, "step": 4460 }, { "epoch": 0.6737652922519257, "grad_norm": 2.1651487902775175, "learning_rate": 8.80622527490755e-06, "loss": 0.1659, "step": 4461 }, { "epoch": 0.6739163268388461, "grad_norm": 2.384805383305368, "learning_rate": 8.805712411835246e-06, "loss": 0.1589, "step": 4462 }, { "epoch": 0.6740673614257665, "grad_norm": 2.181385120471587, "learning_rate": 8.805199453560951e-06, "loss": 0.1711, "step": 4463 }, { "epoch": 0.674218396012687, "grad_norm": 2.024933971875939, "learning_rate": 8.804686400097499e-06, "loss": 0.1472, "step": 4464 }, { "epoch": 0.6743694305996073, "grad_norm": 1.9696246223543141, "learning_rate": 8.804173251457723e-06, "loss": 0.1722, "step": 4465 }, { "epoch": 0.6745204651865278, "grad_norm": 2.162457991298113, "learning_rate": 8.80366000765446e-06, "loss": 0.1479, "step": 4466 }, { "epoch": 0.6746714997734481, "grad_norm": 2.5664344262432532, "learning_rate": 8.803146668700548e-06, "loss": 0.1634, "step": 4467 }, { "epoch": 0.6748225343603685, "grad_norm": 2.063877141677791, "learning_rate": 8.802633234608832e-06, "loss": 0.1476, "step": 4468 }, { "epoch": 0.6749735689472889, "grad_norm": 2.3517483487183792, "learning_rate": 8.80211970539215e-06, "loss": 0.1526, "step": 4469 }, { "epoch": 0.6751246035342093, "grad_norm": 2.0824381938955643, "learning_rate": 8.801606081063356e-06, "loss": 0.1519, "step": 4470 }, { "epoch": 0.6752756381211298, "grad_norm": 2.4764430901839365, "learning_rate": 8.801092361635293e-06, "loss": 0.1888, "step": 4471 }, { "epoch": 0.6754266727080501, "grad_norm": 2.4908047796275636, "learning_rate": 8.800578547120812e-06, "loss": 0.1706, "step": 4472 }, { "epoch": 0.6755777072949706, "grad_norm": 1.8317108321036228, "learning_rate": 8.800064637532769e-06, "loss": 0.1386, "step": 4473 }, { "epoch": 0.6757287418818909, "grad_norm": 1.9844706094167555, "learning_rate": 8.799550632884015e-06, "loss": 0.1437, "step": 4474 }, { "epoch": 0.6758797764688114, "grad_norm": 2.1000907746768416, "learning_rate": 8.799036533187415e-06, "loss": 0.1584, "step": 4475 }, { "epoch": 0.6760308110557318, "grad_norm": 2.1475060508899504, "learning_rate": 8.798522338455823e-06, "loss": 0.1526, "step": 4476 }, { "epoch": 0.6761818456426522, "grad_norm": 1.6692496426301102, "learning_rate": 8.798008048702107e-06, "loss": 0.1134, "step": 4477 }, { "epoch": 0.6763328802295726, "grad_norm": 2.5132114752363406, "learning_rate": 8.79749366393913e-06, "loss": 0.1831, "step": 4478 }, { "epoch": 0.676483914816493, "grad_norm": 2.2125304408614217, "learning_rate": 8.796979184179759e-06, "loss": 0.1522, "step": 4479 }, { "epoch": 0.6766349494034134, "grad_norm": 2.6687192614029147, "learning_rate": 8.796464609436864e-06, "loss": 0.1564, "step": 4480 }, { "epoch": 0.6767859839903337, "grad_norm": 2.2204817634362004, "learning_rate": 8.79594993972332e-06, "loss": 0.157, "step": 4481 }, { "epoch": 0.6769370185772542, "grad_norm": 1.97196273556498, "learning_rate": 8.795435175051998e-06, "loss": 0.1484, "step": 4482 }, { "epoch": 0.6770880531641746, "grad_norm": 2.1582706723755103, "learning_rate": 8.794920315435777e-06, "loss": 0.149, "step": 4483 }, { "epoch": 0.677239087751095, "grad_norm": 2.183501912903484, "learning_rate": 8.794405360887537e-06, "loss": 0.1383, "step": 4484 }, { "epoch": 0.6773901223380154, "grad_norm": 2.346992741955997, "learning_rate": 8.793890311420159e-06, "loss": 0.1588, "step": 4485 }, { "epoch": 0.6775411569249358, "grad_norm": 3.1900250642987684, "learning_rate": 8.793375167046526e-06, "loss": 0.1984, "step": 4486 }, { "epoch": 0.6776921915118562, "grad_norm": 1.9471417909645847, "learning_rate": 8.792859927779527e-06, "loss": 0.1343, "step": 4487 }, { "epoch": 0.6778432260987767, "grad_norm": 2.020708902034869, "learning_rate": 8.792344593632053e-06, "loss": 0.1696, "step": 4488 }, { "epoch": 0.677994260685697, "grad_norm": 2.997433858369008, "learning_rate": 8.791829164616988e-06, "loss": 0.1713, "step": 4489 }, { "epoch": 0.6781452952726175, "grad_norm": 1.8210263521205, "learning_rate": 8.791313640747232e-06, "loss": 0.154, "step": 4490 }, { "epoch": 0.6782963298595378, "grad_norm": 2.2544045009896685, "learning_rate": 8.79079802203568e-06, "loss": 0.1655, "step": 4491 }, { "epoch": 0.6784473644464583, "grad_norm": 1.8643623249755148, "learning_rate": 8.790282308495229e-06, "loss": 0.1287, "step": 4492 }, { "epoch": 0.6785983990333786, "grad_norm": 1.889127829398544, "learning_rate": 8.789766500138781e-06, "loss": 0.143, "step": 4493 }, { "epoch": 0.678749433620299, "grad_norm": 2.6596399026271613, "learning_rate": 8.789250596979238e-06, "loss": 0.1747, "step": 4494 }, { "epoch": 0.6789004682072195, "grad_norm": 2.2447048195198787, "learning_rate": 8.788734599029507e-06, "loss": 0.1697, "step": 4495 }, { "epoch": 0.6790515027941398, "grad_norm": 1.957194267342663, "learning_rate": 8.788218506302497e-06, "loss": 0.1412, "step": 4496 }, { "epoch": 0.6792025373810603, "grad_norm": 2.69223798455558, "learning_rate": 8.787702318811115e-06, "loss": 0.1708, "step": 4497 }, { "epoch": 0.6793535719679806, "grad_norm": 1.87534282088877, "learning_rate": 8.787186036568275e-06, "loss": 0.1539, "step": 4498 }, { "epoch": 0.6795046065549011, "grad_norm": 1.8841335919052264, "learning_rate": 8.786669659586896e-06, "loss": 0.1455, "step": 4499 }, { "epoch": 0.6796556411418215, "grad_norm": 2.207277988791209, "learning_rate": 8.786153187879889e-06, "loss": 0.1493, "step": 4500 }, { "epoch": 0.6798066757287419, "grad_norm": 3.1356179559113313, "learning_rate": 8.785636621460177e-06, "loss": 0.1254, "step": 4501 }, { "epoch": 0.6799577103156623, "grad_norm": 2.2722157766869304, "learning_rate": 8.785119960340682e-06, "loss": 0.181, "step": 4502 }, { "epoch": 0.6801087449025827, "grad_norm": 2.1990722526602613, "learning_rate": 8.78460320453433e-06, "loss": 0.1443, "step": 4503 }, { "epoch": 0.6802597794895031, "grad_norm": 2.4077447201826283, "learning_rate": 8.784086354054046e-06, "loss": 0.1346, "step": 4504 }, { "epoch": 0.6804108140764235, "grad_norm": 1.8922491748915395, "learning_rate": 8.783569408912759e-06, "loss": 0.1558, "step": 4505 }, { "epoch": 0.6805618486633439, "grad_norm": 3.1346788181973237, "learning_rate": 8.783052369123402e-06, "loss": 0.1465, "step": 4506 }, { "epoch": 0.6807128832502644, "grad_norm": 2.0031509464765156, "learning_rate": 8.78253523469891e-06, "loss": 0.1601, "step": 4507 }, { "epoch": 0.6808639178371847, "grad_norm": 1.9965127891172612, "learning_rate": 8.782018005652214e-06, "loss": 0.1471, "step": 4508 }, { "epoch": 0.6810149524241051, "grad_norm": 3.8160265352729894, "learning_rate": 8.78150068199626e-06, "loss": 0.1416, "step": 4509 }, { "epoch": 0.6811659870110255, "grad_norm": 2.1643009699605544, "learning_rate": 8.780983263743986e-06, "loss": 0.1389, "step": 4510 }, { "epoch": 0.6813170215979459, "grad_norm": 2.020540057265469, "learning_rate": 8.780465750908334e-06, "loss": 0.1541, "step": 4511 }, { "epoch": 0.6814680561848664, "grad_norm": 1.9249589054537948, "learning_rate": 8.779948143502252e-06, "loss": 0.1289, "step": 4512 }, { "epoch": 0.6816190907717867, "grad_norm": 2.1575037234868204, "learning_rate": 8.779430441538688e-06, "loss": 0.1565, "step": 4513 }, { "epoch": 0.6817701253587072, "grad_norm": 2.3556196331627097, "learning_rate": 8.778912645030591e-06, "loss": 0.1638, "step": 4514 }, { "epoch": 0.6819211599456275, "grad_norm": 2.037041017629922, "learning_rate": 8.778394753990917e-06, "loss": 0.156, "step": 4515 }, { "epoch": 0.682072194532548, "grad_norm": 2.509607258937539, "learning_rate": 8.777876768432617e-06, "loss": 0.1712, "step": 4516 }, { "epoch": 0.6822232291194683, "grad_norm": 2.0214479526547087, "learning_rate": 8.777358688368652e-06, "loss": 0.1492, "step": 4517 }, { "epoch": 0.6823742637063888, "grad_norm": 4.021626850232264, "learning_rate": 8.776840513811981e-06, "loss": 0.1698, "step": 4518 }, { "epoch": 0.6825252982933092, "grad_norm": 2.317337132856148, "learning_rate": 8.776322244775565e-06, "loss": 0.1641, "step": 4519 }, { "epoch": 0.6826763328802296, "grad_norm": 1.816400056981668, "learning_rate": 8.775803881272374e-06, "loss": 0.1394, "step": 4520 }, { "epoch": 0.68282736746715, "grad_norm": 1.8227869205215286, "learning_rate": 8.775285423315369e-06, "loss": 0.1447, "step": 4521 }, { "epoch": 0.6829784020540703, "grad_norm": 1.7087978577799985, "learning_rate": 8.774766870917523e-06, "loss": 0.1449, "step": 4522 }, { "epoch": 0.6831294366409908, "grad_norm": 2.068467607445932, "learning_rate": 8.774248224091805e-06, "loss": 0.1383, "step": 4523 }, { "epoch": 0.6832804712279112, "grad_norm": 2.678863453849656, "learning_rate": 8.773729482851194e-06, "loss": 0.1805, "step": 4524 }, { "epoch": 0.6834315058148316, "grad_norm": 2.5180782491133216, "learning_rate": 8.773210647208661e-06, "loss": 0.1586, "step": 4525 }, { "epoch": 0.683582540401752, "grad_norm": 1.8896135764808484, "learning_rate": 8.77269171717719e-06, "loss": 0.1497, "step": 4526 }, { "epoch": 0.6837335749886724, "grad_norm": 2.118681316752945, "learning_rate": 8.772172692769757e-06, "loss": 0.1516, "step": 4527 }, { "epoch": 0.6838846095755928, "grad_norm": 1.7940727000830046, "learning_rate": 8.77165357399935e-06, "loss": 0.1581, "step": 4528 }, { "epoch": 0.6840356441625132, "grad_norm": 1.8663410980767408, "learning_rate": 8.771134360878954e-06, "loss": 0.1442, "step": 4529 }, { "epoch": 0.6841866787494336, "grad_norm": 1.8503873005563556, "learning_rate": 8.770615053421558e-06, "loss": 0.139, "step": 4530 }, { "epoch": 0.6843377133363541, "grad_norm": 2.126687033552443, "learning_rate": 8.770095651640152e-06, "loss": 0.1553, "step": 4531 }, { "epoch": 0.6844887479232744, "grad_norm": 2.484751648934944, "learning_rate": 8.769576155547727e-06, "loss": 0.155, "step": 4532 }, { "epoch": 0.6846397825101949, "grad_norm": 3.543441495086544, "learning_rate": 8.76905656515728e-06, "loss": 0.1792, "step": 4533 }, { "epoch": 0.6847908170971152, "grad_norm": 1.8317785946261862, "learning_rate": 8.768536880481812e-06, "loss": 0.145, "step": 4534 }, { "epoch": 0.6849418516840357, "grad_norm": 2.523441867717961, "learning_rate": 8.768017101534318e-06, "loss": 0.1532, "step": 4535 }, { "epoch": 0.6850928862709561, "grad_norm": 1.9879995938087445, "learning_rate": 8.767497228327804e-06, "loss": 0.1439, "step": 4536 }, { "epoch": 0.6852439208578764, "grad_norm": 2.102196981183225, "learning_rate": 8.766977260875275e-06, "loss": 0.1526, "step": 4537 }, { "epoch": 0.6853949554447969, "grad_norm": 3.3792886817239713, "learning_rate": 8.766457199189737e-06, "loss": 0.1689, "step": 4538 }, { "epoch": 0.6855459900317172, "grad_norm": 2.207053747890099, "learning_rate": 8.765937043284198e-06, "loss": 0.1676, "step": 4539 }, { "epoch": 0.6856970246186377, "grad_norm": 2.2020799395534247, "learning_rate": 8.765416793171675e-06, "loss": 0.1418, "step": 4540 }, { "epoch": 0.685848059205558, "grad_norm": 2.6666680078375533, "learning_rate": 8.764896448865177e-06, "loss": 0.1674, "step": 4541 }, { "epoch": 0.6859990937924785, "grad_norm": 2.1024421751481, "learning_rate": 8.764376010377724e-06, "loss": 0.1534, "step": 4542 }, { "epoch": 0.6861501283793989, "grad_norm": 2.7511289165394905, "learning_rate": 8.763855477722331e-06, "loss": 0.1446, "step": 4543 }, { "epoch": 0.6863011629663193, "grad_norm": 1.9478783374106008, "learning_rate": 8.763334850912027e-06, "loss": 0.1409, "step": 4544 }, { "epoch": 0.6864521975532397, "grad_norm": 1.895814291865702, "learning_rate": 8.762814129959826e-06, "loss": 0.1504, "step": 4545 }, { "epoch": 0.6866032321401601, "grad_norm": 3.485898088143928, "learning_rate": 8.762293314878764e-06, "loss": 0.1612, "step": 4546 }, { "epoch": 0.6867542667270805, "grad_norm": 1.968635468748224, "learning_rate": 8.761772405681863e-06, "loss": 0.145, "step": 4547 }, { "epoch": 0.686905301314001, "grad_norm": 1.6614299916638817, "learning_rate": 8.761251402382154e-06, "loss": 0.1246, "step": 4548 }, { "epoch": 0.6870563359009213, "grad_norm": 2.2930830169105354, "learning_rate": 8.760730304992674e-06, "loss": 0.143, "step": 4549 }, { "epoch": 0.6872073704878418, "grad_norm": 2.8306386142697924, "learning_rate": 8.760209113526455e-06, "loss": 0.1497, "step": 4550 }, { "epoch": 0.6873584050747621, "grad_norm": 2.182389897220221, "learning_rate": 8.759687827996537e-06, "loss": 0.1589, "step": 4551 }, { "epoch": 0.6875094396616825, "grad_norm": 1.8464684249449657, "learning_rate": 8.759166448415959e-06, "loss": 0.1304, "step": 4552 }, { "epoch": 0.6876604742486029, "grad_norm": 2.3057142147966627, "learning_rate": 8.758644974797764e-06, "loss": 0.1653, "step": 4553 }, { "epoch": 0.6878115088355233, "grad_norm": 2.3472438560069095, "learning_rate": 8.758123407154995e-06, "loss": 0.1521, "step": 4554 }, { "epoch": 0.6879625434224438, "grad_norm": 2.382523890025646, "learning_rate": 8.757601745500704e-06, "loss": 0.1446, "step": 4555 }, { "epoch": 0.6881135780093641, "grad_norm": 2.2979811673336994, "learning_rate": 8.757079989847936e-06, "loss": 0.1571, "step": 4556 }, { "epoch": 0.6882646125962846, "grad_norm": 1.987106105898636, "learning_rate": 8.756558140209747e-06, "loss": 0.1512, "step": 4557 }, { "epoch": 0.6884156471832049, "grad_norm": 1.8959662087493705, "learning_rate": 8.756036196599189e-06, "loss": 0.1697, "step": 4558 }, { "epoch": 0.6885666817701254, "grad_norm": 1.9614227884858841, "learning_rate": 8.755514159029317e-06, "loss": 0.139, "step": 4559 }, { "epoch": 0.6887177163570458, "grad_norm": 2.0678685623074657, "learning_rate": 8.754992027513194e-06, "loss": 0.1505, "step": 4560 }, { "epoch": 0.6888687509439662, "grad_norm": 2.768610602140076, "learning_rate": 8.754469802063879e-06, "loss": 0.1551, "step": 4561 }, { "epoch": 0.6890197855308866, "grad_norm": 2.3983269351332663, "learning_rate": 8.753947482694435e-06, "loss": 0.1542, "step": 4562 }, { "epoch": 0.689170820117807, "grad_norm": 1.828505532493076, "learning_rate": 8.75342506941793e-06, "loss": 0.1256, "step": 4563 }, { "epoch": 0.6893218547047274, "grad_norm": 2.158419084231102, "learning_rate": 8.752902562247432e-06, "loss": 0.1583, "step": 4564 }, { "epoch": 0.6894728892916477, "grad_norm": 2.1068545806527523, "learning_rate": 8.752379961196011e-06, "loss": 0.1621, "step": 4565 }, { "epoch": 0.6896239238785682, "grad_norm": 2.003727132129498, "learning_rate": 8.751857266276742e-06, "loss": 0.1448, "step": 4566 }, { "epoch": 0.6897749584654886, "grad_norm": 2.1358250261089133, "learning_rate": 8.751334477502698e-06, "loss": 0.139, "step": 4567 }, { "epoch": 0.689925993052409, "grad_norm": 2.282779934011106, "learning_rate": 8.75081159488696e-06, "loss": 0.1407, "step": 4568 }, { "epoch": 0.6900770276393294, "grad_norm": 2.2390813856092833, "learning_rate": 8.750288618442604e-06, "loss": 0.1742, "step": 4569 }, { "epoch": 0.6902280622262498, "grad_norm": 2.0298170054838716, "learning_rate": 8.749765548182716e-06, "loss": 0.1788, "step": 4570 }, { "epoch": 0.6903790968131702, "grad_norm": 1.6932875176777018, "learning_rate": 8.749242384120379e-06, "loss": 0.1284, "step": 4571 }, { "epoch": 0.6905301314000907, "grad_norm": 1.7969496115464387, "learning_rate": 8.74871912626868e-06, "loss": 0.1244, "step": 4572 }, { "epoch": 0.690681165987011, "grad_norm": 2.325299971634737, "learning_rate": 8.74819577464071e-06, "loss": 0.1808, "step": 4573 }, { "epoch": 0.6908322005739315, "grad_norm": 2.869072092996391, "learning_rate": 8.747672329249563e-06, "loss": 0.1725, "step": 4574 }, { "epoch": 0.6909832351608518, "grad_norm": 2.150994541896629, "learning_rate": 8.747148790108328e-06, "loss": 0.1462, "step": 4575 }, { "epoch": 0.6911342697477723, "grad_norm": 3.154417135860244, "learning_rate": 8.746625157230107e-06, "loss": 0.1409, "step": 4576 }, { "epoch": 0.6912853043346926, "grad_norm": 2.0348561366845312, "learning_rate": 8.746101430627993e-06, "loss": 0.1521, "step": 4577 }, { "epoch": 0.691436338921613, "grad_norm": 2.040665205041113, "learning_rate": 8.745577610315093e-06, "loss": 0.1458, "step": 4578 }, { "epoch": 0.6915873735085335, "grad_norm": 1.9706515774001883, "learning_rate": 8.745053696304507e-06, "loss": 0.1547, "step": 4579 }, { "epoch": 0.6917384080954538, "grad_norm": 2.4193605158991964, "learning_rate": 8.744529688609343e-06, "loss": 0.1673, "step": 4580 }, { "epoch": 0.6918894426823743, "grad_norm": 2.3954559113495644, "learning_rate": 8.74400558724271e-06, "loss": 0.1536, "step": 4581 }, { "epoch": 0.6920404772692946, "grad_norm": 2.09910829171619, "learning_rate": 8.743481392217713e-06, "loss": 0.1616, "step": 4582 }, { "epoch": 0.6921915118562151, "grad_norm": 1.9081500649328482, "learning_rate": 8.742957103547474e-06, "loss": 0.1251, "step": 4583 }, { "epoch": 0.6923425464431355, "grad_norm": 2.5819512129334754, "learning_rate": 8.7424327212451e-06, "loss": 0.1608, "step": 4584 }, { "epoch": 0.6924935810300559, "grad_norm": 2.272610389299132, "learning_rate": 8.741908245323715e-06, "loss": 0.1556, "step": 4585 }, { "epoch": 0.6926446156169763, "grad_norm": 1.8772794187451622, "learning_rate": 8.741383675796435e-06, "loss": 0.1496, "step": 4586 }, { "epoch": 0.6927956502038967, "grad_norm": 2.205322177176732, "learning_rate": 8.740859012676386e-06, "loss": 0.1566, "step": 4587 }, { "epoch": 0.6929466847908171, "grad_norm": 1.9489542848743742, "learning_rate": 8.740334255976689e-06, "loss": 0.1349, "step": 4588 }, { "epoch": 0.6930977193777375, "grad_norm": 2.0557938177002986, "learning_rate": 8.739809405710472e-06, "loss": 0.1513, "step": 4589 }, { "epoch": 0.6932487539646579, "grad_norm": 2.687866952084783, "learning_rate": 8.739284461890865e-06, "loss": 0.1578, "step": 4590 }, { "epoch": 0.6933997885515784, "grad_norm": 3.492591479773203, "learning_rate": 8.738759424531002e-06, "loss": 0.1673, "step": 4591 }, { "epoch": 0.6935508231384987, "grad_norm": 2.0169453542062574, "learning_rate": 8.738234293644012e-06, "loss": 0.14, "step": 4592 }, { "epoch": 0.6937018577254191, "grad_norm": 2.680624053145373, "learning_rate": 8.737709069243037e-06, "loss": 0.151, "step": 4593 }, { "epoch": 0.6938528923123395, "grad_norm": 2.9359165005750785, "learning_rate": 8.737183751341211e-06, "loss": 0.1648, "step": 4594 }, { "epoch": 0.6940039268992599, "grad_norm": 2.159135983422571, "learning_rate": 8.73665833995168e-06, "loss": 0.1533, "step": 4595 }, { "epoch": 0.6941549614861804, "grad_norm": 2.4048691116284058, "learning_rate": 8.736132835087582e-06, "loss": 0.1674, "step": 4596 }, { "epoch": 0.6943059960731007, "grad_norm": 3.003276652852155, "learning_rate": 8.735607236762067e-06, "loss": 0.1522, "step": 4597 }, { "epoch": 0.6944570306600212, "grad_norm": 2.0245770888157186, "learning_rate": 8.735081544988281e-06, "loss": 0.1491, "step": 4598 }, { "epoch": 0.6946080652469415, "grad_norm": 2.0185274682263716, "learning_rate": 8.734555759779375e-06, "loss": 0.1705, "step": 4599 }, { "epoch": 0.694759099833862, "grad_norm": 1.813105420494603, "learning_rate": 8.734029881148503e-06, "loss": 0.1255, "step": 4600 }, { "epoch": 0.6949101344207823, "grad_norm": 2.510623908309334, "learning_rate": 8.733503909108817e-06, "loss": 0.1667, "step": 4601 }, { "epoch": 0.6950611690077028, "grad_norm": 2.0607771842784524, "learning_rate": 8.732977843673479e-06, "loss": 0.1441, "step": 4602 }, { "epoch": 0.6952122035946232, "grad_norm": 1.9676883296323717, "learning_rate": 8.732451684855646e-06, "loss": 0.1361, "step": 4603 }, { "epoch": 0.6953632381815436, "grad_norm": 1.9084225293342547, "learning_rate": 8.73192543266848e-06, "loss": 0.1402, "step": 4604 }, { "epoch": 0.695514272768464, "grad_norm": 2.585886460813935, "learning_rate": 8.731399087125145e-06, "loss": 0.1695, "step": 4605 }, { "epoch": 0.6956653073553843, "grad_norm": 2.2533979796954724, "learning_rate": 8.73087264823881e-06, "loss": 0.1519, "step": 4606 }, { "epoch": 0.6958163419423048, "grad_norm": 1.9068833060154762, "learning_rate": 8.730346116022644e-06, "loss": 0.145, "step": 4607 }, { "epoch": 0.6959673765292252, "grad_norm": 2.037608759642697, "learning_rate": 8.729819490489815e-06, "loss": 0.1436, "step": 4608 }, { "epoch": 0.6961184111161456, "grad_norm": 3.1310870468031387, "learning_rate": 8.729292771653501e-06, "loss": 0.1664, "step": 4609 }, { "epoch": 0.696269445703066, "grad_norm": 2.0318346141051102, "learning_rate": 8.728765959526874e-06, "loss": 0.1446, "step": 4610 }, { "epoch": 0.6964204802899864, "grad_norm": 4.276580715320186, "learning_rate": 8.728239054123117e-06, "loss": 0.1741, "step": 4611 }, { "epoch": 0.6965715148769068, "grad_norm": 2.215464995870003, "learning_rate": 8.72771205545541e-06, "loss": 0.1627, "step": 4612 }, { "epoch": 0.6967225494638272, "grad_norm": 2.2927036217056003, "learning_rate": 8.727184963536933e-06, "loss": 0.1578, "step": 4613 }, { "epoch": 0.6968735840507476, "grad_norm": 2.528033181762702, "learning_rate": 8.726657778380874e-06, "loss": 0.161, "step": 4614 }, { "epoch": 0.6970246186376681, "grad_norm": 1.808215346138575, "learning_rate": 8.72613050000042e-06, "loss": 0.1455, "step": 4615 }, { "epoch": 0.6971756532245884, "grad_norm": 1.8240322634754795, "learning_rate": 8.725603128408763e-06, "loss": 0.1453, "step": 4616 }, { "epoch": 0.6973266878115089, "grad_norm": 2.515071750479156, "learning_rate": 8.725075663619093e-06, "loss": 0.1707, "step": 4617 }, { "epoch": 0.6974777223984292, "grad_norm": 2.1120905601051736, "learning_rate": 8.724548105644604e-06, "loss": 0.1362, "step": 4618 }, { "epoch": 0.6976287569853497, "grad_norm": 1.8267592265043497, "learning_rate": 8.724020454498496e-06, "loss": 0.1506, "step": 4619 }, { "epoch": 0.6977797915722701, "grad_norm": 2.1073580309173754, "learning_rate": 8.723492710193968e-06, "loss": 0.1489, "step": 4620 }, { "epoch": 0.6979308261591904, "grad_norm": 2.012457781592766, "learning_rate": 8.722964872744222e-06, "loss": 0.1506, "step": 4621 }, { "epoch": 0.6980818607461109, "grad_norm": 1.6316673167857583, "learning_rate": 8.72243694216246e-06, "loss": 0.1134, "step": 4622 }, { "epoch": 0.6982328953330312, "grad_norm": 2.068202384308958, "learning_rate": 8.72190891846189e-06, "loss": 0.1511, "step": 4623 }, { "epoch": 0.6983839299199517, "grad_norm": 1.9489606891634192, "learning_rate": 8.721380801655721e-06, "loss": 0.1484, "step": 4624 }, { "epoch": 0.698534964506872, "grad_norm": 2.1513498429687954, "learning_rate": 8.720852591757165e-06, "loss": 0.1647, "step": 4625 }, { "epoch": 0.6986859990937925, "grad_norm": 2.1872796794294502, "learning_rate": 8.720324288779431e-06, "loss": 0.1596, "step": 4626 }, { "epoch": 0.6988370336807129, "grad_norm": 1.8833095039272678, "learning_rate": 8.71979589273574e-06, "loss": 0.1395, "step": 4627 }, { "epoch": 0.6989880682676333, "grad_norm": 2.0293113461335026, "learning_rate": 8.719267403639308e-06, "loss": 0.139, "step": 4628 }, { "epoch": 0.6991391028545537, "grad_norm": 1.868584046235681, "learning_rate": 8.718738821503355e-06, "loss": 0.1278, "step": 4629 }, { "epoch": 0.6992901374414741, "grad_norm": 2.8756641472193616, "learning_rate": 8.718210146341106e-06, "loss": 0.192, "step": 4630 }, { "epoch": 0.6994411720283945, "grad_norm": 2.1185817108605245, "learning_rate": 8.717681378165783e-06, "loss": 0.1659, "step": 4631 }, { "epoch": 0.699592206615315, "grad_norm": 2.089098604376863, "learning_rate": 8.717152516990614e-06, "loss": 0.1663, "step": 4632 }, { "epoch": 0.6997432412022353, "grad_norm": 2.573560785117386, "learning_rate": 8.716623562828831e-06, "loss": 0.1727, "step": 4633 }, { "epoch": 0.6998942757891558, "grad_norm": 3.353706290592279, "learning_rate": 8.716094515693665e-06, "loss": 0.1466, "step": 4634 }, { "epoch": 0.7000453103760761, "grad_norm": 3.3032963367822417, "learning_rate": 8.715565375598349e-06, "loss": 0.1532, "step": 4635 }, { "epoch": 0.7001963449629965, "grad_norm": 2.0171155610832536, "learning_rate": 8.715036142556122e-06, "loss": 0.1375, "step": 4636 }, { "epoch": 0.7003473795499169, "grad_norm": 2.3229853862785492, "learning_rate": 8.714506816580221e-06, "loss": 0.1435, "step": 4637 }, { "epoch": 0.7004984141368373, "grad_norm": 4.3969409465834355, "learning_rate": 8.713977397683888e-06, "loss": 0.2096, "step": 4638 }, { "epoch": 0.7006494487237578, "grad_norm": 2.215676443584127, "learning_rate": 8.713447885880369e-06, "loss": 0.1559, "step": 4639 }, { "epoch": 0.7008004833106781, "grad_norm": 2.447174573761201, "learning_rate": 8.712918281182906e-06, "loss": 0.171, "step": 4640 }, { "epoch": 0.7009515178975986, "grad_norm": 2.334240917036291, "learning_rate": 8.712388583604749e-06, "loss": 0.1597, "step": 4641 }, { "epoch": 0.7011025524845189, "grad_norm": 1.8454402321694723, "learning_rate": 8.71185879315915e-06, "loss": 0.1512, "step": 4642 }, { "epoch": 0.7012535870714394, "grad_norm": 4.0991538781351595, "learning_rate": 8.711328909859362e-06, "loss": 0.1846, "step": 4643 }, { "epoch": 0.7014046216583598, "grad_norm": 2.3177807079258126, "learning_rate": 8.710798933718637e-06, "loss": 0.1511, "step": 4644 }, { "epoch": 0.7015556562452802, "grad_norm": 1.967232447433289, "learning_rate": 8.710268864750238e-06, "loss": 0.1469, "step": 4645 }, { "epoch": 0.7017066908322006, "grad_norm": 2.0928239411045366, "learning_rate": 8.709738702967418e-06, "loss": 0.1421, "step": 4646 }, { "epoch": 0.701857725419121, "grad_norm": 2.3188483155304374, "learning_rate": 8.709208448383446e-06, "loss": 0.1539, "step": 4647 }, { "epoch": 0.7020087600060414, "grad_norm": 3.2732495068167724, "learning_rate": 8.708678101011584e-06, "loss": 0.1557, "step": 4648 }, { "epoch": 0.7021597945929617, "grad_norm": 1.7536892308049035, "learning_rate": 8.708147660865098e-06, "loss": 0.1372, "step": 4649 }, { "epoch": 0.7023108291798822, "grad_norm": 2.114030391689677, "learning_rate": 8.707617127957258e-06, "loss": 0.177, "step": 4650 }, { "epoch": 0.7024618637668026, "grad_norm": 2.3305854806304898, "learning_rate": 8.707086502301336e-06, "loss": 0.1583, "step": 4651 }, { "epoch": 0.702612898353723, "grad_norm": 1.9268837180771892, "learning_rate": 8.706555783910606e-06, "loss": 0.1632, "step": 4652 }, { "epoch": 0.7027639329406434, "grad_norm": 1.9485080016182361, "learning_rate": 8.706024972798343e-06, "loss": 0.1364, "step": 4653 }, { "epoch": 0.7029149675275638, "grad_norm": 2.0164928198743044, "learning_rate": 8.705494068977827e-06, "loss": 0.146, "step": 4654 }, { "epoch": 0.7030660021144842, "grad_norm": 1.9057255262502883, "learning_rate": 8.704963072462336e-06, "loss": 0.1369, "step": 4655 }, { "epoch": 0.7032170367014047, "grad_norm": 2.12075461870559, "learning_rate": 8.704431983265158e-06, "loss": 0.1491, "step": 4656 }, { "epoch": 0.703368071288325, "grad_norm": 1.8782459544285088, "learning_rate": 8.703900801399574e-06, "loss": 0.1243, "step": 4657 }, { "epoch": 0.7035191058752455, "grad_norm": 1.6423540370216994, "learning_rate": 8.703369526878873e-06, "loss": 0.1207, "step": 4658 }, { "epoch": 0.7036701404621658, "grad_norm": 2.2145539724136927, "learning_rate": 8.702838159716348e-06, "loss": 0.1549, "step": 4659 }, { "epoch": 0.7038211750490863, "grad_norm": 1.9997567669158116, "learning_rate": 8.702306699925287e-06, "loss": 0.1582, "step": 4660 }, { "epoch": 0.7039722096360066, "grad_norm": 1.9326316080331287, "learning_rate": 8.701775147518989e-06, "loss": 0.1297, "step": 4661 }, { "epoch": 0.704123244222927, "grad_norm": 2.2216388369887787, "learning_rate": 8.701243502510748e-06, "loss": 0.1729, "step": 4662 }, { "epoch": 0.7042742788098475, "grad_norm": 2.3723400544898547, "learning_rate": 8.700711764913863e-06, "loss": 0.1505, "step": 4663 }, { "epoch": 0.7044253133967678, "grad_norm": 2.310970338053252, "learning_rate": 8.70017993474164e-06, "loss": 0.151, "step": 4664 }, { "epoch": 0.7045763479836883, "grad_norm": 2.070560056361176, "learning_rate": 8.699648012007378e-06, "loss": 0.1458, "step": 4665 }, { "epoch": 0.7047273825706086, "grad_norm": 2.330858867844005, "learning_rate": 8.699115996724386e-06, "loss": 0.164, "step": 4666 }, { "epoch": 0.7048784171575291, "grad_norm": 1.8668211631492986, "learning_rate": 8.698583888905972e-06, "loss": 0.1558, "step": 4667 }, { "epoch": 0.7050294517444495, "grad_norm": 2.155213539344162, "learning_rate": 8.698051688565446e-06, "loss": 0.1465, "step": 4668 }, { "epoch": 0.7051804863313699, "grad_norm": 2.3212448520048605, "learning_rate": 8.697519395716123e-06, "loss": 0.1608, "step": 4669 }, { "epoch": 0.7053315209182903, "grad_norm": 4.688212688751288, "learning_rate": 8.69698701037132e-06, "loss": 0.1637, "step": 4670 }, { "epoch": 0.7054825555052107, "grad_norm": 1.9225318788103358, "learning_rate": 8.69645453254435e-06, "loss": 0.1425, "step": 4671 }, { "epoch": 0.7056335900921311, "grad_norm": 2.18243715878646, "learning_rate": 8.695921962248537e-06, "loss": 0.1338, "step": 4672 }, { "epoch": 0.7057846246790515, "grad_norm": 2.135901109641134, "learning_rate": 8.695389299497204e-06, "loss": 0.147, "step": 4673 }, { "epoch": 0.7059356592659719, "grad_norm": 2.8265226812234387, "learning_rate": 8.694856544303673e-06, "loss": 0.16, "step": 4674 }, { "epoch": 0.7060866938528924, "grad_norm": 1.9847238688453945, "learning_rate": 8.694323696681272e-06, "loss": 0.1496, "step": 4675 }, { "epoch": 0.7062377284398127, "grad_norm": 2.2061531525792537, "learning_rate": 8.693790756643333e-06, "loss": 0.1239, "step": 4676 }, { "epoch": 0.7063887630267331, "grad_norm": 2.1828087793546023, "learning_rate": 8.693257724203186e-06, "loss": 0.1665, "step": 4677 }, { "epoch": 0.7065397976136535, "grad_norm": 2.223150336189225, "learning_rate": 8.692724599374162e-06, "loss": 0.1482, "step": 4678 }, { "epoch": 0.7066908322005739, "grad_norm": 2.5925522849702585, "learning_rate": 8.692191382169604e-06, "loss": 0.171, "step": 4679 }, { "epoch": 0.7068418667874944, "grad_norm": 1.9184605303140305, "learning_rate": 8.691658072602845e-06, "loss": 0.1288, "step": 4680 }, { "epoch": 0.7069929013744147, "grad_norm": 2.0283920816829966, "learning_rate": 8.69112467068723e-06, "loss": 0.1514, "step": 4681 }, { "epoch": 0.7071439359613352, "grad_norm": 1.8094421683920079, "learning_rate": 8.690591176436099e-06, "loss": 0.1236, "step": 4682 }, { "epoch": 0.7072949705482555, "grad_norm": 1.6293718710794316, "learning_rate": 8.6900575898628e-06, "loss": 0.1304, "step": 4683 }, { "epoch": 0.707446005135176, "grad_norm": 2.0406098397807493, "learning_rate": 8.689523910980679e-06, "loss": 0.1604, "step": 4684 }, { "epoch": 0.7075970397220963, "grad_norm": 2.213048166347823, "learning_rate": 8.688990139803088e-06, "loss": 0.1474, "step": 4685 }, { "epoch": 0.7077480743090168, "grad_norm": 4.085012216973311, "learning_rate": 8.688456276343379e-06, "loss": 0.1911, "step": 4686 }, { "epoch": 0.7078991088959372, "grad_norm": 2.621771668418654, "learning_rate": 8.687922320614907e-06, "loss": 0.1854, "step": 4687 }, { "epoch": 0.7080501434828576, "grad_norm": 1.8093622602265789, "learning_rate": 8.68738827263103e-06, "loss": 0.1346, "step": 4688 }, { "epoch": 0.708201178069778, "grad_norm": 2.125200216463993, "learning_rate": 8.686854132405104e-06, "loss": 0.1495, "step": 4689 }, { "epoch": 0.7083522126566983, "grad_norm": 1.948482286212535, "learning_rate": 8.686319899950494e-06, "loss": 0.1497, "step": 4690 }, { "epoch": 0.7085032472436188, "grad_norm": 1.664378172046026, "learning_rate": 8.685785575280566e-06, "loss": 0.1167, "step": 4691 }, { "epoch": 0.7086542818305392, "grad_norm": 2.276802636940786, "learning_rate": 8.685251158408682e-06, "loss": 0.1466, "step": 4692 }, { "epoch": 0.7088053164174596, "grad_norm": 1.7160841665717417, "learning_rate": 8.684716649348214e-06, "loss": 0.1334, "step": 4693 }, { "epoch": 0.70895635100438, "grad_norm": 2.7806298626053123, "learning_rate": 8.68418204811253e-06, "loss": 0.168, "step": 4694 }, { "epoch": 0.7091073855913004, "grad_norm": 1.8853086256271658, "learning_rate": 8.683647354715006e-06, "loss": 0.1485, "step": 4695 }, { "epoch": 0.7092584201782208, "grad_norm": 3.2439005900443085, "learning_rate": 8.683112569169016e-06, "loss": 0.1978, "step": 4696 }, { "epoch": 0.7094094547651412, "grad_norm": 2.6625868424286865, "learning_rate": 8.68257769148794e-06, "loss": 0.1732, "step": 4697 }, { "epoch": 0.7095604893520616, "grad_norm": 2.2624233636693645, "learning_rate": 8.682042721685155e-06, "loss": 0.1743, "step": 4698 }, { "epoch": 0.7097115239389821, "grad_norm": 2.670256323828651, "learning_rate": 8.681507659774046e-06, "loss": 0.1729, "step": 4699 }, { "epoch": 0.7098625585259024, "grad_norm": 2.7725846731973345, "learning_rate": 8.680972505767998e-06, "loss": 0.1471, "step": 4700 }, { "epoch": 0.7100135931128229, "grad_norm": 3.6637458602321917, "learning_rate": 8.680437259680398e-06, "loss": 0.1566, "step": 4701 }, { "epoch": 0.7101646276997432, "grad_norm": 2.5350930712639013, "learning_rate": 8.679901921524634e-06, "loss": 0.1792, "step": 4702 }, { "epoch": 0.7103156622866637, "grad_norm": 2.356421127285559, "learning_rate": 8.6793664913141e-06, "loss": 0.1669, "step": 4703 }, { "epoch": 0.7104666968735841, "grad_norm": 2.5030604773069505, "learning_rate": 8.678830969062188e-06, "loss": 0.182, "step": 4704 }, { "epoch": 0.7106177314605044, "grad_norm": 2.4297478475420617, "learning_rate": 8.678295354782295e-06, "loss": 0.1439, "step": 4705 }, { "epoch": 0.7107687660474249, "grad_norm": 1.7938380073604394, "learning_rate": 8.677759648487821e-06, "loss": 0.14, "step": 4706 }, { "epoch": 0.7109198006343452, "grad_norm": 2.1891944406953927, "learning_rate": 8.677223850192162e-06, "loss": 0.1556, "step": 4707 }, { "epoch": 0.7110708352212657, "grad_norm": 2.1620178091128843, "learning_rate": 8.67668795990873e-06, "loss": 0.1451, "step": 4708 }, { "epoch": 0.711221869808186, "grad_norm": 2.039906451018772, "learning_rate": 8.676151977650923e-06, "loss": 0.157, "step": 4709 }, { "epoch": 0.7113729043951065, "grad_norm": 2.3532194242326514, "learning_rate": 8.675615903432153e-06, "loss": 0.1589, "step": 4710 }, { "epoch": 0.7115239389820269, "grad_norm": 1.8073773919129936, "learning_rate": 8.675079737265828e-06, "loss": 0.1468, "step": 4711 }, { "epoch": 0.7116749735689473, "grad_norm": 1.8801508584112592, "learning_rate": 8.674543479165363e-06, "loss": 0.1363, "step": 4712 }, { "epoch": 0.7118260081558677, "grad_norm": 2.7034711430661007, "learning_rate": 8.674007129144169e-06, "loss": 0.1662, "step": 4713 }, { "epoch": 0.7119770427427881, "grad_norm": 4.209780597960342, "learning_rate": 8.673470687215666e-06, "loss": 0.1967, "step": 4714 }, { "epoch": 0.7121280773297085, "grad_norm": 2.202590146827295, "learning_rate": 8.672934153393273e-06, "loss": 0.1467, "step": 4715 }, { "epoch": 0.712279111916629, "grad_norm": 2.149357777775997, "learning_rate": 8.672397527690412e-06, "loss": 0.1458, "step": 4716 }, { "epoch": 0.7124301465035493, "grad_norm": 2.108433323557576, "learning_rate": 8.671860810120505e-06, "loss": 0.1417, "step": 4717 }, { "epoch": 0.7125811810904698, "grad_norm": 2.009669992999982, "learning_rate": 8.67132400069698e-06, "loss": 0.1827, "step": 4718 }, { "epoch": 0.7127322156773901, "grad_norm": 2.0073684885249037, "learning_rate": 8.670787099433265e-06, "loss": 0.155, "step": 4719 }, { "epoch": 0.7128832502643105, "grad_norm": 2.434878828093774, "learning_rate": 8.670250106342792e-06, "loss": 0.134, "step": 4720 }, { "epoch": 0.7130342848512309, "grad_norm": 2.050630587686164, "learning_rate": 8.669713021438992e-06, "loss": 0.142, "step": 4721 }, { "epoch": 0.7131853194381513, "grad_norm": 2.3783047834205315, "learning_rate": 8.669175844735303e-06, "loss": 0.1629, "step": 4722 }, { "epoch": 0.7133363540250718, "grad_norm": 2.2559540449847866, "learning_rate": 8.668638576245158e-06, "loss": 0.1578, "step": 4723 }, { "epoch": 0.7134873886119921, "grad_norm": 1.8726492475240804, "learning_rate": 8.668101215982006e-06, "loss": 0.157, "step": 4724 }, { "epoch": 0.7136384231989126, "grad_norm": 2.60631316456085, "learning_rate": 8.66756376395928e-06, "loss": 0.1694, "step": 4725 }, { "epoch": 0.7137894577858329, "grad_norm": 2.309442405964809, "learning_rate": 8.667026220190429e-06, "loss": 0.154, "step": 4726 }, { "epoch": 0.7139404923727534, "grad_norm": 2.346872977497484, "learning_rate": 8.666488584688901e-06, "loss": 0.1513, "step": 4727 }, { "epoch": 0.7140915269596738, "grad_norm": 1.983053579338368, "learning_rate": 8.665950857468142e-06, "loss": 0.147, "step": 4728 }, { "epoch": 0.7142425615465942, "grad_norm": 4.1188812616365205, "learning_rate": 8.665413038541607e-06, "loss": 0.1578, "step": 4729 }, { "epoch": 0.7143935961335146, "grad_norm": 2.3275765089805014, "learning_rate": 8.664875127922747e-06, "loss": 0.1727, "step": 4730 }, { "epoch": 0.714544630720435, "grad_norm": 2.0296830363650304, "learning_rate": 8.664337125625019e-06, "loss": 0.1568, "step": 4731 }, { "epoch": 0.7146956653073554, "grad_norm": 2.0195105908270023, "learning_rate": 8.663799031661884e-06, "loss": 0.1459, "step": 4732 }, { "epoch": 0.7148466998942757, "grad_norm": 1.7933042694875776, "learning_rate": 8.663260846046799e-06, "loss": 0.1219, "step": 4733 }, { "epoch": 0.7149977344811962, "grad_norm": 1.8944574747794753, "learning_rate": 8.662722568793227e-06, "loss": 0.1322, "step": 4734 }, { "epoch": 0.7151487690681166, "grad_norm": 3.524684248893905, "learning_rate": 8.662184199914635e-06, "loss": 0.1428, "step": 4735 }, { "epoch": 0.715299803655037, "grad_norm": 3.24259862093761, "learning_rate": 8.661645739424492e-06, "loss": 0.1607, "step": 4736 }, { "epoch": 0.7154508382419574, "grad_norm": 2.1238102267389993, "learning_rate": 8.661107187336266e-06, "loss": 0.1619, "step": 4737 }, { "epoch": 0.7156018728288778, "grad_norm": 1.738455095644376, "learning_rate": 8.660568543663429e-06, "loss": 0.1251, "step": 4738 }, { "epoch": 0.7157529074157982, "grad_norm": 1.8055149811124858, "learning_rate": 8.660029808419456e-06, "loss": 0.1383, "step": 4739 }, { "epoch": 0.7159039420027187, "grad_norm": 1.7693471807780832, "learning_rate": 8.659490981617823e-06, "loss": 0.133, "step": 4740 }, { "epoch": 0.716054976589639, "grad_norm": 1.9674222520823714, "learning_rate": 8.658952063272011e-06, "loss": 0.1488, "step": 4741 }, { "epoch": 0.7162060111765595, "grad_norm": 1.6610467933009072, "learning_rate": 8.658413053395497e-06, "loss": 0.1226, "step": 4742 }, { "epoch": 0.7163570457634798, "grad_norm": 10.031716934697254, "learning_rate": 8.65787395200177e-06, "loss": 0.153, "step": 4743 }, { "epoch": 0.7165080803504003, "grad_norm": 2.638685451317582, "learning_rate": 8.657334759104313e-06, "loss": 0.1483, "step": 4744 }, { "epoch": 0.7166591149373206, "grad_norm": 1.9804277018639103, "learning_rate": 8.656795474716615e-06, "loss": 0.1506, "step": 4745 }, { "epoch": 0.716810149524241, "grad_norm": 3.689720676239371, "learning_rate": 8.656256098852165e-06, "loss": 0.1997, "step": 4746 }, { "epoch": 0.7169611841111615, "grad_norm": 1.8490819957679518, "learning_rate": 8.65571663152446e-06, "loss": 0.1462, "step": 4747 }, { "epoch": 0.7171122186980818, "grad_norm": 2.056924798379539, "learning_rate": 8.65517707274699e-06, "loss": 0.161, "step": 4748 }, { "epoch": 0.7172632532850023, "grad_norm": 1.7728185222767237, "learning_rate": 8.654637422533254e-06, "loss": 0.1272, "step": 4749 }, { "epoch": 0.7174142878719226, "grad_norm": 1.872642452381913, "learning_rate": 8.654097680896753e-06, "loss": 0.1312, "step": 4750 }, { "epoch": 0.7175653224588431, "grad_norm": 2.1277034487136524, "learning_rate": 8.653557847850987e-06, "loss": 0.1763, "step": 4751 }, { "epoch": 0.7177163570457635, "grad_norm": 2.1868494302934733, "learning_rate": 8.653017923409462e-06, "loss": 0.1359, "step": 4752 }, { "epoch": 0.7178673916326839, "grad_norm": 2.918537637504054, "learning_rate": 8.652477907585685e-06, "loss": 0.18, "step": 4753 }, { "epoch": 0.7180184262196043, "grad_norm": 1.660077300807047, "learning_rate": 8.651937800393162e-06, "loss": 0.1302, "step": 4754 }, { "epoch": 0.7181694608065247, "grad_norm": 2.534698188058763, "learning_rate": 8.651397601845407e-06, "loss": 0.1605, "step": 4755 }, { "epoch": 0.7183204953934451, "grad_norm": 1.9962790784101527, "learning_rate": 8.650857311955929e-06, "loss": 0.1367, "step": 4756 }, { "epoch": 0.7184715299803655, "grad_norm": 3.592335992383102, "learning_rate": 8.650316930738249e-06, "loss": 0.1555, "step": 4757 }, { "epoch": 0.7186225645672859, "grad_norm": 2.109232746438264, "learning_rate": 8.649776458205884e-06, "loss": 0.1474, "step": 4758 }, { "epoch": 0.7187735991542064, "grad_norm": 3.2426526958917963, "learning_rate": 8.649235894372349e-06, "loss": 0.1495, "step": 4759 }, { "epoch": 0.7189246337411267, "grad_norm": 4.352544732470391, "learning_rate": 8.648695239251173e-06, "loss": 0.1543, "step": 4760 }, { "epoch": 0.7190756683280471, "grad_norm": 1.9186845230584526, "learning_rate": 8.648154492855879e-06, "loss": 0.1437, "step": 4761 }, { "epoch": 0.7192267029149675, "grad_norm": 1.8670332180531735, "learning_rate": 8.64761365519999e-06, "loss": 0.1282, "step": 4762 }, { "epoch": 0.7193777375018879, "grad_norm": 2.025717272437892, "learning_rate": 8.647072726297041e-06, "loss": 0.1369, "step": 4763 }, { "epoch": 0.7195287720888084, "grad_norm": 2.2673075733304997, "learning_rate": 8.646531706160559e-06, "loss": 0.13, "step": 4764 }, { "epoch": 0.7196798066757287, "grad_norm": 1.7477294485973123, "learning_rate": 8.645990594804082e-06, "loss": 0.1417, "step": 4765 }, { "epoch": 0.7198308412626492, "grad_norm": 1.9333186508456492, "learning_rate": 8.645449392241143e-06, "loss": 0.1584, "step": 4766 }, { "epoch": 0.7199818758495695, "grad_norm": 2.60683013824048, "learning_rate": 8.644908098485283e-06, "loss": 0.1814, "step": 4767 }, { "epoch": 0.72013291043649, "grad_norm": 1.9455558642464956, "learning_rate": 8.644366713550042e-06, "loss": 0.1316, "step": 4768 }, { "epoch": 0.7202839450234103, "grad_norm": 2.494811481121864, "learning_rate": 8.643825237448961e-06, "loss": 0.1631, "step": 4769 }, { "epoch": 0.7204349796103308, "grad_norm": 3.9031893278900553, "learning_rate": 8.643283670195586e-06, "loss": 0.173, "step": 4770 }, { "epoch": 0.7205860141972512, "grad_norm": 2.4731659339817345, "learning_rate": 8.642742011803466e-06, "loss": 0.1588, "step": 4771 }, { "epoch": 0.7207370487841716, "grad_norm": 2.6177362558911965, "learning_rate": 8.642200262286153e-06, "loss": 0.1577, "step": 4772 }, { "epoch": 0.720888083371092, "grad_norm": 2.100330879622607, "learning_rate": 8.641658421657195e-06, "loss": 0.1594, "step": 4773 }, { "epoch": 0.7210391179580123, "grad_norm": 2.536564417635219, "learning_rate": 8.641116489930147e-06, "loss": 0.1727, "step": 4774 }, { "epoch": 0.7211901525449328, "grad_norm": 2.273625954025132, "learning_rate": 8.640574467118566e-06, "loss": 0.1515, "step": 4775 }, { "epoch": 0.7213411871318532, "grad_norm": 2.3032264903874595, "learning_rate": 8.640032353236015e-06, "loss": 0.1372, "step": 4776 }, { "epoch": 0.7214922217187736, "grad_norm": 2.1593624177830826, "learning_rate": 8.639490148296048e-06, "loss": 0.1788, "step": 4777 }, { "epoch": 0.721643256305694, "grad_norm": 1.8473768448541177, "learning_rate": 8.638947852312236e-06, "loss": 0.1405, "step": 4778 }, { "epoch": 0.7217942908926144, "grad_norm": 1.9212475042794426, "learning_rate": 8.638405465298138e-06, "loss": 0.1408, "step": 4779 }, { "epoch": 0.7219453254795348, "grad_norm": 1.8437755815412016, "learning_rate": 8.637862987267327e-06, "loss": 0.1315, "step": 4780 }, { "epoch": 0.7220963600664552, "grad_norm": 1.7985675080012955, "learning_rate": 8.637320418233372e-06, "loss": 0.1447, "step": 4781 }, { "epoch": 0.7222473946533756, "grad_norm": 3.8910935675437224, "learning_rate": 8.636777758209846e-06, "loss": 0.1745, "step": 4782 }, { "epoch": 0.7223984292402961, "grad_norm": 2.1269336962096728, "learning_rate": 8.636235007210323e-06, "loss": 0.1429, "step": 4783 }, { "epoch": 0.7225494638272164, "grad_norm": 2.296635145662473, "learning_rate": 8.63569216524838e-06, "loss": 0.1555, "step": 4784 }, { "epoch": 0.7227004984141369, "grad_norm": 2.44306156493011, "learning_rate": 8.635149232337599e-06, "loss": 0.1648, "step": 4785 }, { "epoch": 0.7228515330010572, "grad_norm": 2.5992631453368964, "learning_rate": 8.634606208491558e-06, "loss": 0.1694, "step": 4786 }, { "epoch": 0.7230025675879777, "grad_norm": 2.027688929597973, "learning_rate": 8.634063093723843e-06, "loss": 0.1494, "step": 4787 }, { "epoch": 0.7231536021748981, "grad_norm": 2.14813706000838, "learning_rate": 8.633519888048041e-06, "loss": 0.1706, "step": 4788 }, { "epoch": 0.7233046367618184, "grad_norm": 1.8839427366717358, "learning_rate": 8.63297659147774e-06, "loss": 0.1366, "step": 4789 }, { "epoch": 0.7234556713487389, "grad_norm": 2.6302806625867627, "learning_rate": 8.63243320402653e-06, "loss": 0.1623, "step": 4790 }, { "epoch": 0.7236067059356592, "grad_norm": 1.8310885325008646, "learning_rate": 8.631889725708005e-06, "loss": 0.1322, "step": 4791 }, { "epoch": 0.7237577405225797, "grad_norm": 1.9089813502742954, "learning_rate": 8.63134615653576e-06, "loss": 0.1306, "step": 4792 }, { "epoch": 0.7239087751095, "grad_norm": 2.298584059376361, "learning_rate": 8.630802496523395e-06, "loss": 0.1779, "step": 4793 }, { "epoch": 0.7240598096964205, "grad_norm": 1.9915997538006955, "learning_rate": 8.630258745684506e-06, "loss": 0.1388, "step": 4794 }, { "epoch": 0.7242108442833409, "grad_norm": 2.692280885663234, "learning_rate": 8.629714904032696e-06, "loss": 0.1579, "step": 4795 }, { "epoch": 0.7243618788702613, "grad_norm": 2.219522757081923, "learning_rate": 8.629170971581571e-06, "loss": 0.1398, "step": 4796 }, { "epoch": 0.7245129134571817, "grad_norm": 1.9358610268947924, "learning_rate": 8.628626948344739e-06, "loss": 0.1443, "step": 4797 }, { "epoch": 0.7246639480441021, "grad_norm": 3.4695902779938805, "learning_rate": 8.628082834335807e-06, "loss": 0.1625, "step": 4798 }, { "epoch": 0.7248149826310225, "grad_norm": 2.1033209607578485, "learning_rate": 8.627538629568386e-06, "loss": 0.1514, "step": 4799 }, { "epoch": 0.724966017217943, "grad_norm": 2.0552849731932805, "learning_rate": 8.626994334056092e-06, "loss": 0.163, "step": 4800 }, { "epoch": 0.7251170518048633, "grad_norm": 3.578132578180454, "learning_rate": 8.62644994781254e-06, "loss": 0.1431, "step": 4801 }, { "epoch": 0.7252680863917838, "grad_norm": 2.251851898041612, "learning_rate": 8.625905470851343e-06, "loss": 0.1767, "step": 4802 }, { "epoch": 0.7254191209787041, "grad_norm": 2.0059218559197007, "learning_rate": 8.625360903186128e-06, "loss": 0.153, "step": 4803 }, { "epoch": 0.7255701555656245, "grad_norm": 2.1042893828762903, "learning_rate": 8.624816244830516e-06, "loss": 0.1631, "step": 4804 }, { "epoch": 0.7257211901525449, "grad_norm": 1.8602959254730798, "learning_rate": 8.62427149579813e-06, "loss": 0.1367, "step": 4805 }, { "epoch": 0.7258722247394653, "grad_norm": 2.341357010572072, "learning_rate": 8.6237266561026e-06, "loss": 0.1531, "step": 4806 }, { "epoch": 0.7260232593263858, "grad_norm": 2.1932362644856895, "learning_rate": 8.623181725757552e-06, "loss": 0.1305, "step": 4807 }, { "epoch": 0.7261742939133061, "grad_norm": 1.9535625500010911, "learning_rate": 8.622636704776621e-06, "loss": 0.1513, "step": 4808 }, { "epoch": 0.7263253285002266, "grad_norm": 1.9271251613300102, "learning_rate": 8.62209159317344e-06, "loss": 0.1514, "step": 4809 }, { "epoch": 0.7264763630871469, "grad_norm": 2.1743649809121735, "learning_rate": 8.621546390961644e-06, "loss": 0.1569, "step": 4810 }, { "epoch": 0.7266273976740674, "grad_norm": 3.1247578051399607, "learning_rate": 8.621001098154873e-06, "loss": 0.1868, "step": 4811 }, { "epoch": 0.7267784322609878, "grad_norm": 1.6781296134822998, "learning_rate": 8.620455714766767e-06, "loss": 0.1174, "step": 4812 }, { "epoch": 0.7269294668479082, "grad_norm": 1.8145861216686745, "learning_rate": 8.61991024081097e-06, "loss": 0.1372, "step": 4813 }, { "epoch": 0.7270805014348286, "grad_norm": 1.967134356474115, "learning_rate": 8.619364676301125e-06, "loss": 0.14, "step": 4814 }, { "epoch": 0.727231536021749, "grad_norm": 1.9371419422928204, "learning_rate": 8.618819021250883e-06, "loss": 0.143, "step": 4815 }, { "epoch": 0.7273825706086694, "grad_norm": 2.925815221921508, "learning_rate": 8.61827327567389e-06, "loss": 0.1351, "step": 4816 }, { "epoch": 0.7275336051955897, "grad_norm": 1.8102485332422402, "learning_rate": 8.617727439583804e-06, "loss": 0.1387, "step": 4817 }, { "epoch": 0.7276846397825102, "grad_norm": 1.9048762394513459, "learning_rate": 8.617181512994272e-06, "loss": 0.1466, "step": 4818 }, { "epoch": 0.7278356743694306, "grad_norm": 1.9848171166054216, "learning_rate": 8.616635495918955e-06, "loss": 0.1475, "step": 4819 }, { "epoch": 0.727986708956351, "grad_norm": 1.8481603459699545, "learning_rate": 8.616089388371512e-06, "loss": 0.1424, "step": 4820 }, { "epoch": 0.7281377435432714, "grad_norm": 1.9352495775164438, "learning_rate": 8.615543190365603e-06, "loss": 0.1349, "step": 4821 }, { "epoch": 0.7282887781301918, "grad_norm": 2.1480045743083407, "learning_rate": 8.614996901914892e-06, "loss": 0.1573, "step": 4822 }, { "epoch": 0.7284398127171122, "grad_norm": 2.1009825820756003, "learning_rate": 8.614450523033045e-06, "loss": 0.1515, "step": 4823 }, { "epoch": 0.7285908473040327, "grad_norm": 2.002919553412824, "learning_rate": 8.613904053733729e-06, "loss": 0.1456, "step": 4824 }, { "epoch": 0.728741881890953, "grad_norm": 2.57885592287924, "learning_rate": 8.613357494030615e-06, "loss": 0.185, "step": 4825 }, { "epoch": 0.7288929164778735, "grad_norm": 2.3943631327562733, "learning_rate": 8.612810843937375e-06, "loss": 0.1577, "step": 4826 }, { "epoch": 0.7290439510647938, "grad_norm": 1.842377341141638, "learning_rate": 8.612264103467684e-06, "loss": 0.1288, "step": 4827 }, { "epoch": 0.7291949856517143, "grad_norm": 2.264823955781415, "learning_rate": 8.611717272635219e-06, "loss": 0.1666, "step": 4828 }, { "epoch": 0.7293460202386346, "grad_norm": 2.2712470857685165, "learning_rate": 8.61117035145366e-06, "loss": 0.161, "step": 4829 }, { "epoch": 0.729497054825555, "grad_norm": 2.0489477409223826, "learning_rate": 8.610623339936685e-06, "loss": 0.1489, "step": 4830 }, { "epoch": 0.7296480894124755, "grad_norm": 3.336470784969859, "learning_rate": 8.610076238097983e-06, "loss": 0.1742, "step": 4831 }, { "epoch": 0.7297991239993958, "grad_norm": 2.8159399097052185, "learning_rate": 8.60952904595124e-06, "loss": 0.1634, "step": 4832 }, { "epoch": 0.7299501585863163, "grad_norm": 2.285166149905388, "learning_rate": 8.608981763510138e-06, "loss": 0.1628, "step": 4833 }, { "epoch": 0.7301011931732366, "grad_norm": 2.1649901153469635, "learning_rate": 8.608434390788373e-06, "loss": 0.1533, "step": 4834 }, { "epoch": 0.7302522277601571, "grad_norm": 2.2523412177284072, "learning_rate": 8.607886927799635e-06, "loss": 0.1735, "step": 4835 }, { "epoch": 0.7304032623470775, "grad_norm": 1.8909048173982086, "learning_rate": 8.607339374557622e-06, "loss": 0.1395, "step": 4836 }, { "epoch": 0.7305542969339979, "grad_norm": 2.2959846827088284, "learning_rate": 8.60679173107603e-06, "loss": 0.1606, "step": 4837 }, { "epoch": 0.7307053315209183, "grad_norm": 2.2798334737409545, "learning_rate": 8.606243997368558e-06, "loss": 0.1654, "step": 4838 }, { "epoch": 0.7308563661078387, "grad_norm": 3.634829806094518, "learning_rate": 8.605696173448907e-06, "loss": 0.1888, "step": 4839 }, { "epoch": 0.7310074006947591, "grad_norm": 2.533840096158156, "learning_rate": 8.605148259330785e-06, "loss": 0.1636, "step": 4840 }, { "epoch": 0.7311584352816795, "grad_norm": 2.121600649292606, "learning_rate": 8.604600255027892e-06, "loss": 0.1515, "step": 4841 }, { "epoch": 0.7313094698685999, "grad_norm": 2.1887192283039716, "learning_rate": 8.604052160553945e-06, "loss": 0.1595, "step": 4842 }, { "epoch": 0.7314605044555204, "grad_norm": 2.46163651502346, "learning_rate": 8.603503975922647e-06, "loss": 0.144, "step": 4843 }, { "epoch": 0.7316115390424407, "grad_norm": 2.1295937052563993, "learning_rate": 8.602955701147715e-06, "loss": 0.1523, "step": 4844 }, { "epoch": 0.7317625736293611, "grad_norm": 1.8422701121642235, "learning_rate": 8.602407336242864e-06, "loss": 0.1535, "step": 4845 }, { "epoch": 0.7319136082162815, "grad_norm": 1.9122126286908963, "learning_rate": 8.601858881221813e-06, "loss": 0.1523, "step": 4846 }, { "epoch": 0.7320646428032019, "grad_norm": 2.1040918113241407, "learning_rate": 8.601310336098278e-06, "loss": 0.1379, "step": 4847 }, { "epoch": 0.7322156773901224, "grad_norm": 2.7390073078562636, "learning_rate": 8.600761700885985e-06, "loss": 0.1457, "step": 4848 }, { "epoch": 0.7323667119770427, "grad_norm": 1.981129078353284, "learning_rate": 8.600212975598656e-06, "loss": 0.1585, "step": 4849 }, { "epoch": 0.7325177465639632, "grad_norm": 3.211105519490653, "learning_rate": 8.59966416025002e-06, "loss": 0.2021, "step": 4850 }, { "epoch": 0.7326687811508835, "grad_norm": 2.2550728031847056, "learning_rate": 8.599115254853802e-06, "loss": 0.1412, "step": 4851 }, { "epoch": 0.732819815737804, "grad_norm": 1.793043794778488, "learning_rate": 8.598566259423737e-06, "loss": 0.1178, "step": 4852 }, { "epoch": 0.7329708503247243, "grad_norm": 2.0450235023345766, "learning_rate": 8.598017173973557e-06, "loss": 0.1483, "step": 4853 }, { "epoch": 0.7331218849116448, "grad_norm": 1.823495087763403, "learning_rate": 8.597467998516999e-06, "loss": 0.1365, "step": 4854 }, { "epoch": 0.7332729194985652, "grad_norm": 2.494930648995959, "learning_rate": 8.596918733067798e-06, "loss": 0.1474, "step": 4855 }, { "epoch": 0.7334239540854856, "grad_norm": 1.8323927071760115, "learning_rate": 8.596369377639697e-06, "loss": 0.1426, "step": 4856 }, { "epoch": 0.733574988672406, "grad_norm": 2.3273897826265277, "learning_rate": 8.595819932246436e-06, "loss": 0.1629, "step": 4857 }, { "epoch": 0.7337260232593263, "grad_norm": 1.972882864452221, "learning_rate": 8.595270396901762e-06, "loss": 0.1248, "step": 4858 }, { "epoch": 0.7338770578462468, "grad_norm": 1.8012686048245707, "learning_rate": 8.594720771619421e-06, "loss": 0.1316, "step": 4859 }, { "epoch": 0.7340280924331672, "grad_norm": 2.010153045769762, "learning_rate": 8.594171056413164e-06, "loss": 0.1196, "step": 4860 }, { "epoch": 0.7341791270200876, "grad_norm": 2.1207975188967763, "learning_rate": 8.593621251296736e-06, "loss": 0.1571, "step": 4861 }, { "epoch": 0.734330161607008, "grad_norm": 2.194316757923857, "learning_rate": 8.593071356283899e-06, "loss": 0.1327, "step": 4862 }, { "epoch": 0.7344811961939284, "grad_norm": 2.714086640400457, "learning_rate": 8.592521371388403e-06, "loss": 0.1975, "step": 4863 }, { "epoch": 0.7346322307808488, "grad_norm": 2.017591057713576, "learning_rate": 8.591971296624008e-06, "loss": 0.1431, "step": 4864 }, { "epoch": 0.7347832653677692, "grad_norm": 2.498275473002373, "learning_rate": 8.591421132004477e-06, "loss": 0.1621, "step": 4865 }, { "epoch": 0.7349342999546896, "grad_norm": 1.7595253643360498, "learning_rate": 8.59087087754357e-06, "loss": 0.1411, "step": 4866 }, { "epoch": 0.7350853345416101, "grad_norm": 1.6263913204193334, "learning_rate": 8.59032053325505e-06, "loss": 0.1307, "step": 4867 }, { "epoch": 0.7352363691285304, "grad_norm": 1.753877297665552, "learning_rate": 8.589770099152689e-06, "loss": 0.1307, "step": 4868 }, { "epoch": 0.7353874037154509, "grad_norm": 1.8892387909682404, "learning_rate": 8.589219575250252e-06, "loss": 0.1382, "step": 4869 }, { "epoch": 0.7355384383023712, "grad_norm": 1.89774509043458, "learning_rate": 8.588668961561514e-06, "loss": 0.152, "step": 4870 }, { "epoch": 0.7356894728892917, "grad_norm": 1.804633833238593, "learning_rate": 8.588118258100247e-06, "loss": 0.1313, "step": 4871 }, { "epoch": 0.7358405074762121, "grad_norm": 2.0801634142060466, "learning_rate": 8.587567464880227e-06, "loss": 0.1701, "step": 4872 }, { "epoch": 0.7359915420631324, "grad_norm": 2.181938958782894, "learning_rate": 8.587016581915233e-06, "loss": 0.1396, "step": 4873 }, { "epoch": 0.7361425766500529, "grad_norm": 1.7970321066463433, "learning_rate": 8.586465609219046e-06, "loss": 0.1376, "step": 4874 }, { "epoch": 0.7362936112369732, "grad_norm": 2.095704796998284, "learning_rate": 8.58591454680545e-06, "loss": 0.1453, "step": 4875 }, { "epoch": 0.7364446458238937, "grad_norm": 2.146970838110404, "learning_rate": 8.585363394688227e-06, "loss": 0.1347, "step": 4876 }, { "epoch": 0.736595680410814, "grad_norm": 2.239582605122242, "learning_rate": 8.584812152881165e-06, "loss": 0.1686, "step": 4877 }, { "epoch": 0.7367467149977345, "grad_norm": 2.3522875910762187, "learning_rate": 8.584260821398057e-06, "loss": 0.1522, "step": 4878 }, { "epoch": 0.7368977495846549, "grad_norm": 1.950079493921937, "learning_rate": 8.583709400252692e-06, "loss": 0.1316, "step": 4879 }, { "epoch": 0.7370487841715753, "grad_norm": 2.4202864453218593, "learning_rate": 8.583157889458863e-06, "loss": 0.1738, "step": 4880 }, { "epoch": 0.7371998187584957, "grad_norm": 1.723354716633088, "learning_rate": 8.582606289030369e-06, "loss": 0.1471, "step": 4881 }, { "epoch": 0.7373508533454161, "grad_norm": 2.1371539239783073, "learning_rate": 8.582054598981007e-06, "loss": 0.1441, "step": 4882 }, { "epoch": 0.7375018879323365, "grad_norm": 1.8493932217961713, "learning_rate": 8.58150281932458e-06, "loss": 0.1319, "step": 4883 }, { "epoch": 0.737652922519257, "grad_norm": 2.1612778637952736, "learning_rate": 8.580950950074888e-06, "loss": 0.1508, "step": 4884 }, { "epoch": 0.7378039571061773, "grad_norm": 1.7680307773597264, "learning_rate": 8.580398991245737e-06, "loss": 0.1259, "step": 4885 }, { "epoch": 0.7379549916930978, "grad_norm": 1.7562929216401206, "learning_rate": 8.579846942850938e-06, "loss": 0.1366, "step": 4886 }, { "epoch": 0.7381060262800181, "grad_norm": 2.015774537667382, "learning_rate": 8.579294804904296e-06, "loss": 0.1437, "step": 4887 }, { "epoch": 0.7382570608669385, "grad_norm": 2.2730320783770144, "learning_rate": 8.578742577419626e-06, "loss": 0.1571, "step": 4888 }, { "epoch": 0.7384080954538589, "grad_norm": 2.80236887203594, "learning_rate": 8.57819026041074e-06, "loss": 0.1645, "step": 4889 }, { "epoch": 0.7385591300407793, "grad_norm": 2.2126409410260455, "learning_rate": 8.577637853891458e-06, "loss": 0.1661, "step": 4890 }, { "epoch": 0.7387101646276998, "grad_norm": 2.331590035512323, "learning_rate": 8.577085357875595e-06, "loss": 0.1443, "step": 4891 }, { "epoch": 0.7388611992146201, "grad_norm": 2.7850019671375326, "learning_rate": 8.576532772376975e-06, "loss": 0.1464, "step": 4892 }, { "epoch": 0.7390122338015406, "grad_norm": 2.0980252833736244, "learning_rate": 8.575980097409418e-06, "loss": 0.1718, "step": 4893 }, { "epoch": 0.7391632683884609, "grad_norm": 2.020945307166062, "learning_rate": 8.575427332986752e-06, "loss": 0.1499, "step": 4894 }, { "epoch": 0.7393143029753814, "grad_norm": 2.2472605212576227, "learning_rate": 8.574874479122806e-06, "loss": 0.1545, "step": 4895 }, { "epoch": 0.7394653375623018, "grad_norm": 2.2338405376198853, "learning_rate": 8.574321535831407e-06, "loss": 0.1615, "step": 4896 }, { "epoch": 0.7396163721492222, "grad_norm": 1.7802940180779756, "learning_rate": 8.573768503126388e-06, "loss": 0.1556, "step": 4897 }, { "epoch": 0.7397674067361426, "grad_norm": 1.7943292407762492, "learning_rate": 8.573215381021583e-06, "loss": 0.1282, "step": 4898 }, { "epoch": 0.739918441323063, "grad_norm": 2.2664127082314742, "learning_rate": 8.57266216953083e-06, "loss": 0.1323, "step": 4899 }, { "epoch": 0.7400694759099834, "grad_norm": 2.3753779138001136, "learning_rate": 8.572108868667967e-06, "loss": 0.1614, "step": 4900 }, { "epoch": 0.7402205104969037, "grad_norm": 2.5043107881673397, "learning_rate": 8.571555478446836e-06, "loss": 0.1943, "step": 4901 }, { "epoch": 0.7403715450838242, "grad_norm": 1.9649112692042563, "learning_rate": 8.571001998881278e-06, "loss": 0.1771, "step": 4902 }, { "epoch": 0.7405225796707446, "grad_norm": 1.9237364912961101, "learning_rate": 8.57044842998514e-06, "loss": 0.1455, "step": 4903 }, { "epoch": 0.740673614257665, "grad_norm": 2.150600679803608, "learning_rate": 8.569894771772272e-06, "loss": 0.1729, "step": 4904 }, { "epoch": 0.7408246488445854, "grad_norm": 2.1440182507700105, "learning_rate": 8.569341024256521e-06, "loss": 0.1327, "step": 4905 }, { "epoch": 0.7409756834315058, "grad_norm": 2.1153758919011754, "learning_rate": 8.568787187451742e-06, "loss": 0.1576, "step": 4906 }, { "epoch": 0.7411267180184262, "grad_norm": 2.0220452199475063, "learning_rate": 8.568233261371786e-06, "loss": 0.148, "step": 4907 }, { "epoch": 0.7412777526053467, "grad_norm": 1.980811800625024, "learning_rate": 8.567679246030513e-06, "loss": 0.1479, "step": 4908 }, { "epoch": 0.741428787192267, "grad_norm": 2.142453018945474, "learning_rate": 8.56712514144178e-06, "loss": 0.1571, "step": 4909 }, { "epoch": 0.7415798217791875, "grad_norm": 2.224116112318038, "learning_rate": 8.56657094761945e-06, "loss": 0.1479, "step": 4910 }, { "epoch": 0.7417308563661078, "grad_norm": 1.76829286281624, "learning_rate": 8.566016664577383e-06, "loss": 0.1388, "step": 4911 }, { "epoch": 0.7418818909530283, "grad_norm": 2.035336706837117, "learning_rate": 8.56546229232945e-06, "loss": 0.1448, "step": 4912 }, { "epoch": 0.7420329255399486, "grad_norm": 2.5699057819680964, "learning_rate": 8.564907830889515e-06, "loss": 0.1436, "step": 4913 }, { "epoch": 0.742183960126869, "grad_norm": 2.0520908973392236, "learning_rate": 8.564353280271449e-06, "loss": 0.1573, "step": 4914 }, { "epoch": 0.7423349947137895, "grad_norm": 2.185472304586548, "learning_rate": 8.563798640489125e-06, "loss": 0.1248, "step": 4915 }, { "epoch": 0.7424860293007098, "grad_norm": 1.894141126717199, "learning_rate": 8.563243911556416e-06, "loss": 0.1304, "step": 4916 }, { "epoch": 0.7426370638876303, "grad_norm": 2.0413623593665133, "learning_rate": 8.562689093487201e-06, "loss": 0.1583, "step": 4917 }, { "epoch": 0.7427880984745506, "grad_norm": 1.8887804117202267, "learning_rate": 8.562134186295358e-06, "loss": 0.1566, "step": 4918 }, { "epoch": 0.7429391330614711, "grad_norm": 1.9077488211317186, "learning_rate": 8.56157918999477e-06, "loss": 0.1394, "step": 4919 }, { "epoch": 0.7430901676483915, "grad_norm": 2.970434610607679, "learning_rate": 8.561024104599317e-06, "loss": 0.1744, "step": 4920 }, { "epoch": 0.7432412022353119, "grad_norm": 1.9990891740297174, "learning_rate": 8.560468930122886e-06, "loss": 0.1375, "step": 4921 }, { "epoch": 0.7433922368222323, "grad_norm": 2.348695521634085, "learning_rate": 8.559913666579367e-06, "loss": 0.1561, "step": 4922 }, { "epoch": 0.7435432714091527, "grad_norm": 2.0848173159590386, "learning_rate": 8.559358313982648e-06, "loss": 0.13, "step": 4923 }, { "epoch": 0.7436943059960731, "grad_norm": 1.9073159938430015, "learning_rate": 8.558802872346624e-06, "loss": 0.1524, "step": 4924 }, { "epoch": 0.7438453405829935, "grad_norm": 2.6867372167066885, "learning_rate": 8.558247341685189e-06, "loss": 0.175, "step": 4925 }, { "epoch": 0.7439963751699139, "grad_norm": 2.384766076060272, "learning_rate": 8.557691722012238e-06, "loss": 0.158, "step": 4926 }, { "epoch": 0.7441474097568344, "grad_norm": 2.636138100727845, "learning_rate": 8.557136013341668e-06, "loss": 0.1655, "step": 4927 }, { "epoch": 0.7442984443437547, "grad_norm": 2.530482271831566, "learning_rate": 8.556580215687386e-06, "loss": 0.1648, "step": 4928 }, { "epoch": 0.7444494789306751, "grad_norm": 2.334351993985367, "learning_rate": 8.556024329063296e-06, "loss": 0.1709, "step": 4929 }, { "epoch": 0.7446005135175955, "grad_norm": 1.9382855830601522, "learning_rate": 8.555468353483295e-06, "loss": 0.1403, "step": 4930 }, { "epoch": 0.7447515481045159, "grad_norm": 2.0688639002341627, "learning_rate": 8.554912288961302e-06, "loss": 0.1163, "step": 4931 }, { "epoch": 0.7449025826914364, "grad_norm": 1.9792754163749866, "learning_rate": 8.55435613551122e-06, "loss": 0.1366, "step": 4932 }, { "epoch": 0.7450536172783567, "grad_norm": 1.7634859588404583, "learning_rate": 8.553799893146964e-06, "loss": 0.1481, "step": 4933 }, { "epoch": 0.7452046518652772, "grad_norm": 1.8115356705746748, "learning_rate": 8.553243561882449e-06, "loss": 0.1502, "step": 4934 }, { "epoch": 0.7453556864521975, "grad_norm": 2.0281044683395653, "learning_rate": 8.55268714173159e-06, "loss": 0.1452, "step": 4935 }, { "epoch": 0.745506721039118, "grad_norm": 1.9388739299036475, "learning_rate": 8.552130632708308e-06, "loss": 0.1457, "step": 4936 }, { "epoch": 0.7456577556260383, "grad_norm": 2.2624191251826797, "learning_rate": 8.551574034826524e-06, "loss": 0.1459, "step": 4937 }, { "epoch": 0.7458087902129588, "grad_norm": 2.2761023275506784, "learning_rate": 8.551017348100161e-06, "loss": 0.1606, "step": 4938 }, { "epoch": 0.7459598247998792, "grad_norm": 2.3773105212120864, "learning_rate": 8.550460572543147e-06, "loss": 0.1554, "step": 4939 }, { "epoch": 0.7461108593867996, "grad_norm": 1.9367424117789014, "learning_rate": 8.549903708169406e-06, "loss": 0.1534, "step": 4940 }, { "epoch": 0.74626189397372, "grad_norm": 1.7763926327976602, "learning_rate": 8.54934675499287e-06, "loss": 0.1456, "step": 4941 }, { "epoch": 0.7464129285606403, "grad_norm": 2.165856048473013, "learning_rate": 8.548789713027477e-06, "loss": 0.1227, "step": 4942 }, { "epoch": 0.7465639631475608, "grad_norm": 2.3965840168519934, "learning_rate": 8.548232582287153e-06, "loss": 0.1501, "step": 4943 }, { "epoch": 0.7467149977344812, "grad_norm": 1.9105827368770834, "learning_rate": 8.547675362785839e-06, "loss": 0.1496, "step": 4944 }, { "epoch": 0.7468660323214016, "grad_norm": 2.0509334541823656, "learning_rate": 8.547118054537472e-06, "loss": 0.1483, "step": 4945 }, { "epoch": 0.747017066908322, "grad_norm": 1.8719796222107332, "learning_rate": 8.546560657556e-06, "loss": 0.1497, "step": 4946 }, { "epoch": 0.7471681014952424, "grad_norm": 2.2083604466449915, "learning_rate": 8.546003171855359e-06, "loss": 0.1356, "step": 4947 }, { "epoch": 0.7473191360821628, "grad_norm": 1.5887289184409636, "learning_rate": 8.545445597449498e-06, "loss": 0.124, "step": 4948 }, { "epoch": 0.7474701706690832, "grad_norm": 1.9403095886750947, "learning_rate": 8.544887934352364e-06, "loss": 0.1431, "step": 4949 }, { "epoch": 0.7476212052560036, "grad_norm": 2.3797642168631086, "learning_rate": 8.544330182577909e-06, "loss": 0.1651, "step": 4950 }, { "epoch": 0.7477722398429241, "grad_norm": 3.156016291590645, "learning_rate": 8.543772342140083e-06, "loss": 0.1404, "step": 4951 }, { "epoch": 0.7479232744298444, "grad_norm": 2.3360170400463707, "learning_rate": 8.543214413052843e-06, "loss": 0.1555, "step": 4952 }, { "epoch": 0.7480743090167649, "grad_norm": 2.4351091247371084, "learning_rate": 8.542656395330145e-06, "loss": 0.1476, "step": 4953 }, { "epoch": 0.7482253436036852, "grad_norm": 2.2836428541506435, "learning_rate": 8.542098288985948e-06, "loss": 0.1387, "step": 4954 }, { "epoch": 0.7483763781906057, "grad_norm": 1.7887741799329901, "learning_rate": 8.541540094034214e-06, "loss": 0.1274, "step": 4955 }, { "epoch": 0.7485274127775261, "grad_norm": 2.4314424653370676, "learning_rate": 8.540981810488905e-06, "loss": 0.1659, "step": 4956 }, { "epoch": 0.7486784473644464, "grad_norm": 2.2715019815445747, "learning_rate": 8.540423438363989e-06, "loss": 0.1513, "step": 4957 }, { "epoch": 0.7488294819513669, "grad_norm": 1.9631213033419346, "learning_rate": 8.539864977673432e-06, "loss": 0.155, "step": 4958 }, { "epoch": 0.7489805165382872, "grad_norm": 2.815913793243631, "learning_rate": 8.539306428431204e-06, "loss": 0.1586, "step": 4959 }, { "epoch": 0.7491315511252077, "grad_norm": 2.124861105882016, "learning_rate": 8.53874779065128e-06, "loss": 0.1656, "step": 4960 }, { "epoch": 0.749282585712128, "grad_norm": 1.7965989537998914, "learning_rate": 8.538189064347633e-06, "loss": 0.1549, "step": 4961 }, { "epoch": 0.7494336202990485, "grad_norm": 2.5925993951045783, "learning_rate": 8.537630249534239e-06, "loss": 0.1678, "step": 4962 }, { "epoch": 0.7495846548859689, "grad_norm": 2.1888158367700403, "learning_rate": 8.537071346225077e-06, "loss": 0.1496, "step": 4963 }, { "epoch": 0.7497356894728893, "grad_norm": 3.1654863948533154, "learning_rate": 8.536512354434132e-06, "loss": 0.1554, "step": 4964 }, { "epoch": 0.7498867240598097, "grad_norm": 1.8749108613650514, "learning_rate": 8.535953274175382e-06, "loss": 0.144, "step": 4965 }, { "epoch": 0.7500377586467301, "grad_norm": 1.907684848036438, "learning_rate": 8.535394105462817e-06, "loss": 0.1404, "step": 4966 }, { "epoch": 0.7501887932336505, "grad_norm": 2.2555819524686402, "learning_rate": 8.534834848310423e-06, "loss": 0.1713, "step": 4967 }, { "epoch": 0.750339827820571, "grad_norm": 2.110828818287335, "learning_rate": 8.53427550273219e-06, "loss": 0.1563, "step": 4968 }, { "epoch": 0.7504908624074913, "grad_norm": 2.4765240820291616, "learning_rate": 8.53371606874211e-06, "loss": 0.1716, "step": 4969 }, { "epoch": 0.7506418969944118, "grad_norm": 1.747217343529524, "learning_rate": 8.53315654635418e-06, "loss": 0.1411, "step": 4970 }, { "epoch": 0.7507929315813321, "grad_norm": 2.3411012552118495, "learning_rate": 8.532596935582394e-06, "loss": 0.1513, "step": 4971 }, { "epoch": 0.7509439661682525, "grad_norm": 2.014976682884051, "learning_rate": 8.532037236440752e-06, "loss": 0.1692, "step": 4972 }, { "epoch": 0.7510950007551729, "grad_norm": 2.122144911127506, "learning_rate": 8.531477448943256e-06, "loss": 0.166, "step": 4973 }, { "epoch": 0.7512460353420933, "grad_norm": 2.364506202617792, "learning_rate": 8.530917573103908e-06, "loss": 0.1281, "step": 4974 }, { "epoch": 0.7513970699290138, "grad_norm": 2.4467699761336448, "learning_rate": 8.530357608936715e-06, "loss": 0.1644, "step": 4975 }, { "epoch": 0.7515481045159341, "grad_norm": 1.9647334048784166, "learning_rate": 8.529797556455682e-06, "loss": 0.1331, "step": 4976 }, { "epoch": 0.7516991391028546, "grad_norm": 1.9701520969551267, "learning_rate": 8.529237415674823e-06, "loss": 0.1433, "step": 4977 }, { "epoch": 0.7518501736897749, "grad_norm": 1.8589278118179862, "learning_rate": 8.528677186608147e-06, "loss": 0.124, "step": 4978 }, { "epoch": 0.7520012082766954, "grad_norm": 3.0590744409986934, "learning_rate": 8.528116869269672e-06, "loss": 0.1798, "step": 4979 }, { "epoch": 0.7521522428636158, "grad_norm": 1.915174206482815, "learning_rate": 8.52755646367341e-06, "loss": 0.1356, "step": 4980 }, { "epoch": 0.7523032774505362, "grad_norm": 1.7359074826734564, "learning_rate": 8.526995969833383e-06, "loss": 0.1331, "step": 4981 }, { "epoch": 0.7524543120374566, "grad_norm": 2.0060772870794277, "learning_rate": 8.52643538776361e-06, "loss": 0.1469, "step": 4982 }, { "epoch": 0.752605346624377, "grad_norm": 1.6586732964354969, "learning_rate": 8.525874717478118e-06, "loss": 0.1193, "step": 4983 }, { "epoch": 0.7527563812112974, "grad_norm": 2.758023178410298, "learning_rate": 8.525313958990928e-06, "loss": 0.1712, "step": 4984 }, { "epoch": 0.7529074157982177, "grad_norm": 2.5790443929582536, "learning_rate": 8.52475311231607e-06, "loss": 0.1684, "step": 4985 }, { "epoch": 0.7530584503851382, "grad_norm": 2.364860355670052, "learning_rate": 8.524192177467574e-06, "loss": 0.174, "step": 4986 }, { "epoch": 0.7532094849720586, "grad_norm": 2.5838307028908067, "learning_rate": 8.523631154459472e-06, "loss": 0.1668, "step": 4987 }, { "epoch": 0.753360519558979, "grad_norm": 1.9298161156467584, "learning_rate": 8.523070043305797e-06, "loss": 0.1372, "step": 4988 }, { "epoch": 0.7535115541458994, "grad_norm": 1.792931615709487, "learning_rate": 8.522508844020587e-06, "loss": 0.1175, "step": 4989 }, { "epoch": 0.7536625887328198, "grad_norm": 2.2288385199116236, "learning_rate": 8.521947556617882e-06, "loss": 0.1458, "step": 4990 }, { "epoch": 0.7538136233197402, "grad_norm": 2.431482538988341, "learning_rate": 8.521386181111719e-06, "loss": 0.1611, "step": 4991 }, { "epoch": 0.7539646579066607, "grad_norm": 2.5031456478281284, "learning_rate": 8.520824717516143e-06, "loss": 0.159, "step": 4992 }, { "epoch": 0.754115692493581, "grad_norm": 2.0644406334936796, "learning_rate": 8.520263165845201e-06, "loss": 0.1453, "step": 4993 }, { "epoch": 0.7542667270805015, "grad_norm": 1.9173173315553262, "learning_rate": 8.519701526112938e-06, "loss": 0.129, "step": 4994 }, { "epoch": 0.7544177616674218, "grad_norm": 2.5936409537198166, "learning_rate": 8.519139798333404e-06, "loss": 0.1469, "step": 4995 }, { "epoch": 0.7545687962543423, "grad_norm": 2.0054440525610935, "learning_rate": 8.518577982520653e-06, "loss": 0.1501, "step": 4996 }, { "epoch": 0.7547198308412626, "grad_norm": 1.9412833955604325, "learning_rate": 8.518016078688738e-06, "loss": 0.1434, "step": 4997 }, { "epoch": 0.754870865428183, "grad_norm": 2.074630514881935, "learning_rate": 8.517454086851716e-06, "loss": 0.1411, "step": 4998 }, { "epoch": 0.7550219000151035, "grad_norm": 2.197596123235772, "learning_rate": 8.516892007023642e-06, "loss": 0.1479, "step": 4999 }, { "epoch": 0.7551729346020238, "grad_norm": 3.49955150890958, "learning_rate": 8.516329839218582e-06, "loss": 0.1951, "step": 5000 }, { "epoch": 0.7553239691889443, "grad_norm": 1.704912070286995, "learning_rate": 8.515767583450596e-06, "loss": 0.1415, "step": 5001 }, { "epoch": 0.7554750037758646, "grad_norm": 2.6559873250476342, "learning_rate": 8.51520523973375e-06, "loss": 0.174, "step": 5002 }, { "epoch": 0.7556260383627851, "grad_norm": 1.9098919037029498, "learning_rate": 8.514642808082108e-06, "loss": 0.119, "step": 5003 }, { "epoch": 0.7557770729497055, "grad_norm": 1.8565323880553661, "learning_rate": 8.514080288509744e-06, "loss": 0.1343, "step": 5004 }, { "epoch": 0.7559281075366259, "grad_norm": 2.3613261716368728, "learning_rate": 8.51351768103073e-06, "loss": 0.1642, "step": 5005 }, { "epoch": 0.7560791421235463, "grad_norm": 2.70327903181652, "learning_rate": 8.512954985659136e-06, "loss": 0.2038, "step": 5006 }, { "epoch": 0.7562301767104667, "grad_norm": 1.9737286667090364, "learning_rate": 8.512392202409041e-06, "loss": 0.1339, "step": 5007 }, { "epoch": 0.7563812112973871, "grad_norm": 2.01138972261974, "learning_rate": 8.511829331294523e-06, "loss": 0.1372, "step": 5008 }, { "epoch": 0.7565322458843075, "grad_norm": 1.9009632729476367, "learning_rate": 8.51126637232966e-06, "loss": 0.1519, "step": 5009 }, { "epoch": 0.7566832804712279, "grad_norm": 2.149579461749239, "learning_rate": 8.510703325528539e-06, "loss": 0.1551, "step": 5010 }, { "epoch": 0.7568343150581484, "grad_norm": 2.504259194870727, "learning_rate": 8.510140190905242e-06, "loss": 0.1375, "step": 5011 }, { "epoch": 0.7569853496450687, "grad_norm": 1.855207456909146, "learning_rate": 8.509576968473857e-06, "loss": 0.1313, "step": 5012 }, { "epoch": 0.7571363842319891, "grad_norm": 2.723496412801528, "learning_rate": 8.509013658248473e-06, "loss": 0.1637, "step": 5013 }, { "epoch": 0.7572874188189095, "grad_norm": 2.3780341263159004, "learning_rate": 8.508450260243182e-06, "loss": 0.1748, "step": 5014 }, { "epoch": 0.7574384534058299, "grad_norm": 1.9321748067586666, "learning_rate": 8.507886774472077e-06, "loss": 0.133, "step": 5015 }, { "epoch": 0.7575894879927504, "grad_norm": 2.0664779108439735, "learning_rate": 8.507323200949253e-06, "loss": 0.1349, "step": 5016 }, { "epoch": 0.7577405225796707, "grad_norm": 2.6909393729928848, "learning_rate": 8.506759539688812e-06, "loss": 0.1353, "step": 5017 }, { "epoch": 0.7578915571665912, "grad_norm": 2.7760792611033436, "learning_rate": 8.50619579070485e-06, "loss": 0.1726, "step": 5018 }, { "epoch": 0.7580425917535115, "grad_norm": 1.9911750870818516, "learning_rate": 8.505631954011472e-06, "loss": 0.1525, "step": 5019 }, { "epoch": 0.758193626340432, "grad_norm": 2.455000621975255, "learning_rate": 8.50506802962278e-06, "loss": 0.1587, "step": 5020 }, { "epoch": 0.7583446609273523, "grad_norm": 2.2784333968019514, "learning_rate": 8.504504017552886e-06, "loss": 0.1509, "step": 5021 }, { "epoch": 0.7584956955142728, "grad_norm": 2.2011775177390622, "learning_rate": 8.503939917815892e-06, "loss": 0.1431, "step": 5022 }, { "epoch": 0.7586467301011932, "grad_norm": 3.4036206842953116, "learning_rate": 8.503375730425917e-06, "loss": 0.1635, "step": 5023 }, { "epoch": 0.7587977646881136, "grad_norm": 2.648111757361365, "learning_rate": 8.502811455397069e-06, "loss": 0.1689, "step": 5024 }, { "epoch": 0.758948799275034, "grad_norm": 2.1344555363929065, "learning_rate": 8.502247092743464e-06, "loss": 0.1693, "step": 5025 }, { "epoch": 0.7590998338619543, "grad_norm": 1.9666597692619416, "learning_rate": 8.501682642479223e-06, "loss": 0.1255, "step": 5026 }, { "epoch": 0.7592508684488748, "grad_norm": 1.8579256958727515, "learning_rate": 8.501118104618462e-06, "loss": 0.141, "step": 5027 }, { "epoch": 0.7594019030357952, "grad_norm": 1.97748461741409, "learning_rate": 8.500553479175306e-06, "loss": 0.1469, "step": 5028 }, { "epoch": 0.7595529376227156, "grad_norm": 2.165496515479349, "learning_rate": 8.49998876616388e-06, "loss": 0.1628, "step": 5029 }, { "epoch": 0.759703972209636, "grad_norm": 2.059273106203058, "learning_rate": 8.499423965598308e-06, "loss": 0.1676, "step": 5030 }, { "epoch": 0.7598550067965564, "grad_norm": 2.457731053060718, "learning_rate": 8.498859077492722e-06, "loss": 0.1696, "step": 5031 }, { "epoch": 0.7600060413834768, "grad_norm": 1.8925384335070652, "learning_rate": 8.498294101861248e-06, "loss": 0.1396, "step": 5032 }, { "epoch": 0.7601570759703972, "grad_norm": 1.9634682195617037, "learning_rate": 8.497729038718026e-06, "loss": 0.1395, "step": 5033 }, { "epoch": 0.7603081105573176, "grad_norm": 2.36019465426681, "learning_rate": 8.497163888077185e-06, "loss": 0.1552, "step": 5034 }, { "epoch": 0.7604591451442381, "grad_norm": 1.9043151998994767, "learning_rate": 8.496598649952865e-06, "loss": 0.1423, "step": 5035 }, { "epoch": 0.7606101797311584, "grad_norm": 2.2927561431135497, "learning_rate": 8.496033324359207e-06, "loss": 0.1352, "step": 5036 }, { "epoch": 0.7607612143180789, "grad_norm": 1.9587061016467309, "learning_rate": 8.495467911310354e-06, "loss": 0.1321, "step": 5037 }, { "epoch": 0.7609122489049992, "grad_norm": 2.163209255922597, "learning_rate": 8.494902410820445e-06, "loss": 0.1494, "step": 5038 }, { "epoch": 0.7610632834919197, "grad_norm": 1.7373032248413083, "learning_rate": 8.49433682290363e-06, "loss": 0.1486, "step": 5039 }, { "epoch": 0.7612143180788401, "grad_norm": 2.5652950089807915, "learning_rate": 8.493771147574057e-06, "loss": 0.1674, "step": 5040 }, { "epoch": 0.7613653526657604, "grad_norm": 2.070210362063369, "learning_rate": 8.493205384845875e-06, "loss": 0.1462, "step": 5041 }, { "epoch": 0.7615163872526809, "grad_norm": 2.234082761490726, "learning_rate": 8.49263953473324e-06, "loss": 0.152, "step": 5042 }, { "epoch": 0.7616674218396012, "grad_norm": 2.1419626819066044, "learning_rate": 8.492073597250306e-06, "loss": 0.1521, "step": 5043 }, { "epoch": 0.7618184564265217, "grad_norm": 1.8314983591930343, "learning_rate": 8.491507572411228e-06, "loss": 0.1296, "step": 5044 }, { "epoch": 0.761969491013442, "grad_norm": 3.1940332100569324, "learning_rate": 8.490941460230169e-06, "loss": 0.1756, "step": 5045 }, { "epoch": 0.7621205256003625, "grad_norm": 2.6859678629282104, "learning_rate": 8.490375260721287e-06, "loss": 0.1703, "step": 5046 }, { "epoch": 0.7622715601872829, "grad_norm": 2.771846854803047, "learning_rate": 8.489808973898748e-06, "loss": 0.1748, "step": 5047 }, { "epoch": 0.7624225947742033, "grad_norm": 1.9800140625384002, "learning_rate": 8.489242599776717e-06, "loss": 0.1371, "step": 5048 }, { "epoch": 0.7625736293611237, "grad_norm": 1.8355872925623336, "learning_rate": 8.488676138369362e-06, "loss": 0.1255, "step": 5049 }, { "epoch": 0.7627246639480441, "grad_norm": 2.3793782436079005, "learning_rate": 8.488109589690855e-06, "loss": 0.143, "step": 5050 }, { "epoch": 0.7628756985349645, "grad_norm": 1.972356774356095, "learning_rate": 8.487542953755368e-06, "loss": 0.15, "step": 5051 }, { "epoch": 0.763026733121885, "grad_norm": 2.4573818804162353, "learning_rate": 8.486976230577074e-06, "loss": 0.1492, "step": 5052 }, { "epoch": 0.7631777677088053, "grad_norm": 2.078862263223112, "learning_rate": 8.486409420170152e-06, "loss": 0.1564, "step": 5053 }, { "epoch": 0.7633288022957258, "grad_norm": 2.1499398974668917, "learning_rate": 8.48584252254878e-06, "loss": 0.1387, "step": 5054 }, { "epoch": 0.7634798368826461, "grad_norm": 2.1869392109712673, "learning_rate": 8.485275537727138e-06, "loss": 0.142, "step": 5055 }, { "epoch": 0.7636308714695665, "grad_norm": 2.1403656848882746, "learning_rate": 8.484708465719413e-06, "loss": 0.1502, "step": 5056 }, { "epoch": 0.7637819060564869, "grad_norm": 1.6377637293328369, "learning_rate": 8.484141306539786e-06, "loss": 0.1297, "step": 5057 }, { "epoch": 0.7639329406434073, "grad_norm": 1.5942642438854489, "learning_rate": 8.48357406020245e-06, "loss": 0.122, "step": 5058 }, { "epoch": 0.7640839752303278, "grad_norm": 4.350946940648535, "learning_rate": 8.483006726721592e-06, "loss": 0.1874, "step": 5059 }, { "epoch": 0.7642350098172481, "grad_norm": 2.1344529704279083, "learning_rate": 8.482439306111402e-06, "loss": 0.1526, "step": 5060 }, { "epoch": 0.7643860444041686, "grad_norm": 2.0142586576895414, "learning_rate": 8.481871798386078e-06, "loss": 0.1414, "step": 5061 }, { "epoch": 0.7645370789910889, "grad_norm": 2.1880937190165715, "learning_rate": 8.481304203559814e-06, "loss": 0.1504, "step": 5062 }, { "epoch": 0.7646881135780094, "grad_norm": 1.8679217453196162, "learning_rate": 8.480736521646813e-06, "loss": 0.1228, "step": 5063 }, { "epoch": 0.7648391481649298, "grad_norm": 2.020696984915641, "learning_rate": 8.480168752661271e-06, "loss": 0.1607, "step": 5064 }, { "epoch": 0.7649901827518502, "grad_norm": 1.773506048590794, "learning_rate": 8.479600896617393e-06, "loss": 0.1414, "step": 5065 }, { "epoch": 0.7651412173387706, "grad_norm": 4.7362915985216745, "learning_rate": 8.479032953529384e-06, "loss": 0.1471, "step": 5066 }, { "epoch": 0.765292251925691, "grad_norm": 2.497026825126742, "learning_rate": 8.478464923411451e-06, "loss": 0.1475, "step": 5067 }, { "epoch": 0.7654432865126114, "grad_norm": 2.185308398089561, "learning_rate": 8.477896806277806e-06, "loss": 0.1289, "step": 5068 }, { "epoch": 0.7655943210995317, "grad_norm": 1.7637456836791707, "learning_rate": 8.477328602142657e-06, "loss": 0.1296, "step": 5069 }, { "epoch": 0.7657453556864522, "grad_norm": 1.9750044946127177, "learning_rate": 8.47676031102022e-06, "loss": 0.1534, "step": 5070 }, { "epoch": 0.7658963902733726, "grad_norm": 1.7089811143405147, "learning_rate": 8.476191932924711e-06, "loss": 0.1291, "step": 5071 }, { "epoch": 0.766047424860293, "grad_norm": 3.7431494518301505, "learning_rate": 8.475623467870348e-06, "loss": 0.1783, "step": 5072 }, { "epoch": 0.7661984594472134, "grad_norm": 1.9367678814367753, "learning_rate": 8.475054915871353e-06, "loss": 0.1624, "step": 5073 }, { "epoch": 0.7663494940341338, "grad_norm": 1.9317751082433667, "learning_rate": 8.474486276941947e-06, "loss": 0.1428, "step": 5074 }, { "epoch": 0.7665005286210542, "grad_norm": 2.0649194336809353, "learning_rate": 8.473917551096353e-06, "loss": 0.1642, "step": 5075 }, { "epoch": 0.7666515632079747, "grad_norm": 2.173762151258399, "learning_rate": 8.473348738348802e-06, "loss": 0.1433, "step": 5076 }, { "epoch": 0.766802597794895, "grad_norm": 2.4948977620001167, "learning_rate": 8.47277983871352e-06, "loss": 0.1281, "step": 5077 }, { "epoch": 0.7669536323818155, "grad_norm": 2.359031197031239, "learning_rate": 8.472210852204741e-06, "loss": 0.1396, "step": 5078 }, { "epoch": 0.7671046669687358, "grad_norm": 1.9583462453255285, "learning_rate": 8.471641778836698e-06, "loss": 0.1455, "step": 5079 }, { "epoch": 0.7672557015556563, "grad_norm": 1.8720747940575473, "learning_rate": 8.471072618623624e-06, "loss": 0.1284, "step": 5080 }, { "epoch": 0.7674067361425766, "grad_norm": 2.394205907819114, "learning_rate": 8.470503371579759e-06, "loss": 0.15, "step": 5081 }, { "epoch": 0.767557770729497, "grad_norm": 2.532832886327164, "learning_rate": 8.469934037719342e-06, "loss": 0.165, "step": 5082 }, { "epoch": 0.7677088053164175, "grad_norm": 3.339233884328545, "learning_rate": 8.46936461705662e-06, "loss": 0.1724, "step": 5083 }, { "epoch": 0.7678598399033378, "grad_norm": 1.7479106210111541, "learning_rate": 8.468795109605829e-06, "loss": 0.1373, "step": 5084 }, { "epoch": 0.7680108744902583, "grad_norm": 1.8557994280417018, "learning_rate": 8.46822551538122e-06, "loss": 0.1307, "step": 5085 }, { "epoch": 0.7681619090771786, "grad_norm": 2.0804781407377795, "learning_rate": 8.467655834397042e-06, "loss": 0.1375, "step": 5086 }, { "epoch": 0.7683129436640991, "grad_norm": 2.0082829179629145, "learning_rate": 8.467086066667547e-06, "loss": 0.1426, "step": 5087 }, { "epoch": 0.7684639782510195, "grad_norm": 2.344150396892425, "learning_rate": 8.466516212206984e-06, "loss": 0.1479, "step": 5088 }, { "epoch": 0.7686150128379399, "grad_norm": 1.8414681328220046, "learning_rate": 8.465946271029612e-06, "loss": 0.1349, "step": 5089 }, { "epoch": 0.7687660474248603, "grad_norm": 2.073459224338276, "learning_rate": 8.465376243149687e-06, "loss": 0.136, "step": 5090 }, { "epoch": 0.7689170820117807, "grad_norm": 2.652967874560771, "learning_rate": 8.464806128581469e-06, "loss": 0.1788, "step": 5091 }, { "epoch": 0.7690681165987011, "grad_norm": 2.514056751299902, "learning_rate": 8.46423592733922e-06, "loss": 0.1431, "step": 5092 }, { "epoch": 0.7692191511856215, "grad_norm": 1.6295982769152257, "learning_rate": 8.463665639437202e-06, "loss": 0.1241, "step": 5093 }, { "epoch": 0.7693701857725419, "grad_norm": 2.1027649828935293, "learning_rate": 8.463095264889683e-06, "loss": 0.152, "step": 5094 }, { "epoch": 0.7695212203594624, "grad_norm": 1.909385609205081, "learning_rate": 8.46252480371093e-06, "loss": 0.1379, "step": 5095 }, { "epoch": 0.7696722549463827, "grad_norm": 2.4030735659806277, "learning_rate": 8.461954255915215e-06, "loss": 0.1305, "step": 5096 }, { "epoch": 0.7698232895333031, "grad_norm": 2.1048015076950626, "learning_rate": 8.461383621516808e-06, "loss": 0.1314, "step": 5097 }, { "epoch": 0.7699743241202235, "grad_norm": 2.04204909344493, "learning_rate": 8.460812900529986e-06, "loss": 0.1438, "step": 5098 }, { "epoch": 0.7701253587071439, "grad_norm": 2.2573786164075127, "learning_rate": 8.460242092969024e-06, "loss": 0.1737, "step": 5099 }, { "epoch": 0.7702763932940644, "grad_norm": 2.574082174306891, "learning_rate": 8.459671198848203e-06, "loss": 0.1603, "step": 5100 }, { "epoch": 0.7704274278809847, "grad_norm": 2.2042743128441935, "learning_rate": 8.459100218181806e-06, "loss": 0.1572, "step": 5101 }, { "epoch": 0.7705784624679052, "grad_norm": 1.879539483373305, "learning_rate": 8.45852915098411e-06, "loss": 0.1403, "step": 5102 }, { "epoch": 0.7707294970548255, "grad_norm": 2.503522423005131, "learning_rate": 8.457957997269406e-06, "loss": 0.1684, "step": 5103 }, { "epoch": 0.770880531641746, "grad_norm": 2.0077422277737584, "learning_rate": 8.457386757051981e-06, "loss": 0.1405, "step": 5104 }, { "epoch": 0.7710315662286663, "grad_norm": 1.9676002109379855, "learning_rate": 8.456815430346123e-06, "loss": 0.157, "step": 5105 }, { "epoch": 0.7711826008155868, "grad_norm": 2.365768401956094, "learning_rate": 8.456244017166126e-06, "loss": 0.1545, "step": 5106 }, { "epoch": 0.7713336354025072, "grad_norm": 2.5326420035575015, "learning_rate": 8.455672517526281e-06, "loss": 0.1655, "step": 5107 }, { "epoch": 0.7714846699894276, "grad_norm": 1.934005819659821, "learning_rate": 8.45510093144089e-06, "loss": 0.1332, "step": 5108 }, { "epoch": 0.771635704576348, "grad_norm": 2.4191669621559835, "learning_rate": 8.454529258924246e-06, "loss": 0.147, "step": 5109 }, { "epoch": 0.7717867391632683, "grad_norm": 1.9851029328882075, "learning_rate": 8.453957499990653e-06, "loss": 0.1478, "step": 5110 }, { "epoch": 0.7719377737501888, "grad_norm": 1.655851409186764, "learning_rate": 8.453385654654411e-06, "loss": 0.1163, "step": 5111 }, { "epoch": 0.7720888083371092, "grad_norm": 2.0596636497905796, "learning_rate": 8.452813722929829e-06, "loss": 0.1506, "step": 5112 }, { "epoch": 0.7722398429240296, "grad_norm": 2.0751443676884342, "learning_rate": 8.452241704831212e-06, "loss": 0.1315, "step": 5113 }, { "epoch": 0.77239087751095, "grad_norm": 2.075066864457546, "learning_rate": 8.451669600372868e-06, "loss": 0.174, "step": 5114 }, { "epoch": 0.7725419120978704, "grad_norm": 1.808079832660701, "learning_rate": 8.451097409569108e-06, "loss": 0.1416, "step": 5115 }, { "epoch": 0.7726929466847908, "grad_norm": 1.9273357830258449, "learning_rate": 8.45052513243425e-06, "loss": 0.1295, "step": 5116 }, { "epoch": 0.7728439812717112, "grad_norm": 1.832984965079411, "learning_rate": 8.449952768982608e-06, "loss": 0.1369, "step": 5117 }, { "epoch": 0.7729950158586316, "grad_norm": 1.959634281367554, "learning_rate": 8.4493803192285e-06, "loss": 0.1513, "step": 5118 }, { "epoch": 0.7731460504455521, "grad_norm": 1.6077314040756525, "learning_rate": 8.448807783186243e-06, "loss": 0.1344, "step": 5119 }, { "epoch": 0.7732970850324724, "grad_norm": 2.1126094801510793, "learning_rate": 8.448235160870163e-06, "loss": 0.1585, "step": 5120 }, { "epoch": 0.7734481196193929, "grad_norm": 2.796516132327696, "learning_rate": 8.447662452294582e-06, "loss": 0.154, "step": 5121 }, { "epoch": 0.7735991542063132, "grad_norm": 2.413680965097096, "learning_rate": 8.447089657473829e-06, "loss": 0.138, "step": 5122 }, { "epoch": 0.7737501887932337, "grad_norm": 1.9954729698676725, "learning_rate": 8.446516776422233e-06, "loss": 0.1612, "step": 5123 }, { "epoch": 0.7739012233801541, "grad_norm": 4.55161565489598, "learning_rate": 8.445943809154121e-06, "loss": 0.1752, "step": 5124 }, { "epoch": 0.7740522579670744, "grad_norm": 1.433899349242588, "learning_rate": 8.445370755683829e-06, "loss": 0.11, "step": 5125 }, { "epoch": 0.7742032925539949, "grad_norm": 1.738697521216369, "learning_rate": 8.444797616025693e-06, "loss": 0.1331, "step": 5126 }, { "epoch": 0.7743543271409152, "grad_norm": 1.9324874297054415, "learning_rate": 8.444224390194049e-06, "loss": 0.1452, "step": 5127 }, { "epoch": 0.7745053617278357, "grad_norm": 1.8554395486908155, "learning_rate": 8.443651078203236e-06, "loss": 0.1377, "step": 5128 }, { "epoch": 0.774656396314756, "grad_norm": 2.018745843400344, "learning_rate": 8.443077680067597e-06, "loss": 0.1655, "step": 5129 }, { "epoch": 0.7748074309016765, "grad_norm": 6.1141355794550725, "learning_rate": 8.442504195801478e-06, "loss": 0.1569, "step": 5130 }, { "epoch": 0.7749584654885969, "grad_norm": 1.6612906655058828, "learning_rate": 8.44193062541922e-06, "loss": 0.1178, "step": 5131 }, { "epoch": 0.7751095000755173, "grad_norm": 2.5216750628273568, "learning_rate": 8.441356968935173e-06, "loss": 0.1658, "step": 5132 }, { "epoch": 0.7752605346624377, "grad_norm": 2.001554844242392, "learning_rate": 8.44078322636369e-06, "loss": 0.1493, "step": 5133 }, { "epoch": 0.7754115692493581, "grad_norm": 2.3604626434169727, "learning_rate": 8.44020939771912e-06, "loss": 0.159, "step": 5134 }, { "epoch": 0.7755626038362785, "grad_norm": 2.0488889066749425, "learning_rate": 8.43963548301582e-06, "loss": 0.1397, "step": 5135 }, { "epoch": 0.775713638423199, "grad_norm": 3.99460107737905, "learning_rate": 8.439061482268144e-06, "loss": 0.1922, "step": 5136 }, { "epoch": 0.7758646730101193, "grad_norm": 3.091525087848795, "learning_rate": 8.438487395490456e-06, "loss": 0.1431, "step": 5137 }, { "epoch": 0.7760157075970397, "grad_norm": 1.8229904419182494, "learning_rate": 8.437913222697112e-06, "loss": 0.1419, "step": 5138 }, { "epoch": 0.7761667421839601, "grad_norm": 3.1122123858117114, "learning_rate": 8.437338963902478e-06, "loss": 0.1648, "step": 5139 }, { "epoch": 0.7763177767708805, "grad_norm": 1.8629114747368325, "learning_rate": 8.436764619120916e-06, "loss": 0.1455, "step": 5140 }, { "epoch": 0.7764688113578009, "grad_norm": 2.3112160270391158, "learning_rate": 8.4361901883668e-06, "loss": 0.1418, "step": 5141 }, { "epoch": 0.7766198459447213, "grad_norm": 2.571671780607768, "learning_rate": 8.435615671654493e-06, "loss": 0.1534, "step": 5142 }, { "epoch": 0.7767708805316418, "grad_norm": 1.8882487514018018, "learning_rate": 8.43504106899837e-06, "loss": 0.1336, "step": 5143 }, { "epoch": 0.7769219151185621, "grad_norm": 2.1676718201382847, "learning_rate": 8.434466380412807e-06, "loss": 0.1326, "step": 5144 }, { "epoch": 0.7770729497054826, "grad_norm": 2.406228762299424, "learning_rate": 8.433891605912177e-06, "loss": 0.1825, "step": 5145 }, { "epoch": 0.7772239842924029, "grad_norm": 2.0253835296889964, "learning_rate": 8.433316745510858e-06, "loss": 0.1649, "step": 5146 }, { "epoch": 0.7773750188793234, "grad_norm": 1.8834303600248228, "learning_rate": 8.432741799223232e-06, "loss": 0.152, "step": 5147 }, { "epoch": 0.7775260534662438, "grad_norm": 2.2633152317967666, "learning_rate": 8.432166767063681e-06, "loss": 0.1495, "step": 5148 }, { "epoch": 0.7776770880531642, "grad_norm": 8.178480956492292, "learning_rate": 8.431591649046591e-06, "loss": 0.1524, "step": 5149 }, { "epoch": 0.7778281226400846, "grad_norm": 2.51198410051908, "learning_rate": 8.431016445186348e-06, "loss": 0.1628, "step": 5150 }, { "epoch": 0.777979157227005, "grad_norm": 1.9156230192859336, "learning_rate": 8.43044115549734e-06, "loss": 0.1396, "step": 5151 }, { "epoch": 0.7781301918139254, "grad_norm": 2.4963652473635136, "learning_rate": 8.42986577999396e-06, "loss": 0.1508, "step": 5152 }, { "epoch": 0.7782812264008458, "grad_norm": 2.1578897273466664, "learning_rate": 8.429290318690598e-06, "loss": 0.1544, "step": 5153 }, { "epoch": 0.7784322609877662, "grad_norm": 2.271202147365994, "learning_rate": 8.428714771601654e-06, "loss": 0.1412, "step": 5154 }, { "epoch": 0.7785832955746866, "grad_norm": 2.4547654057590904, "learning_rate": 8.428139138741522e-06, "loss": 0.1524, "step": 5155 }, { "epoch": 0.778734330161607, "grad_norm": 1.8870693251034334, "learning_rate": 8.427563420124606e-06, "loss": 0.1355, "step": 5156 }, { "epoch": 0.7788853647485274, "grad_norm": 1.9701125079343154, "learning_rate": 8.426987615765302e-06, "loss": 0.1543, "step": 5157 }, { "epoch": 0.7790363993354478, "grad_norm": 2.2180923859590247, "learning_rate": 8.426411725678018e-06, "loss": 0.1513, "step": 5158 }, { "epoch": 0.7791874339223682, "grad_norm": 1.9143064908337681, "learning_rate": 8.425835749877163e-06, "loss": 0.1407, "step": 5159 }, { "epoch": 0.7793384685092887, "grad_norm": 1.841364078822237, "learning_rate": 8.425259688377139e-06, "loss": 0.1379, "step": 5160 }, { "epoch": 0.779489503096209, "grad_norm": 1.8101310177014134, "learning_rate": 8.424683541192358e-06, "loss": 0.1504, "step": 5161 }, { "epoch": 0.7796405376831295, "grad_norm": 1.566403497807243, "learning_rate": 8.424107308337235e-06, "loss": 0.109, "step": 5162 }, { "epoch": 0.7797915722700498, "grad_norm": 2.1395635143928744, "learning_rate": 8.423530989826185e-06, "loss": 0.1726, "step": 5163 }, { "epoch": 0.7799426068569703, "grad_norm": 1.862342604017711, "learning_rate": 8.422954585673623e-06, "loss": 0.1229, "step": 5164 }, { "epoch": 0.7800936414438907, "grad_norm": 2.0248155034531097, "learning_rate": 8.422378095893968e-06, "loss": 0.1451, "step": 5165 }, { "epoch": 0.780244676030811, "grad_norm": 2.0353516243279115, "learning_rate": 8.421801520501643e-06, "loss": 0.1509, "step": 5166 }, { "epoch": 0.7803957106177315, "grad_norm": 1.9836637207918233, "learning_rate": 8.42122485951107e-06, "loss": 0.1504, "step": 5167 }, { "epoch": 0.7805467452046518, "grad_norm": 1.766269402108638, "learning_rate": 8.420648112936673e-06, "loss": 0.1398, "step": 5168 }, { "epoch": 0.7806977797915723, "grad_norm": 1.7949945551354614, "learning_rate": 8.420071280792884e-06, "loss": 0.1264, "step": 5169 }, { "epoch": 0.7808488143784926, "grad_norm": 2.4623420722124743, "learning_rate": 8.419494363094129e-06, "loss": 0.1494, "step": 5170 }, { "epoch": 0.7809998489654131, "grad_norm": 2.755820768224055, "learning_rate": 8.418917359854841e-06, "loss": 0.1635, "step": 5171 }, { "epoch": 0.7811508835523335, "grad_norm": 2.8656592073618694, "learning_rate": 8.418340271089456e-06, "loss": 0.1656, "step": 5172 }, { "epoch": 0.7813019181392539, "grad_norm": 2.187175492384709, "learning_rate": 8.417763096812406e-06, "loss": 0.1273, "step": 5173 }, { "epoch": 0.7814529527261743, "grad_norm": 2.0625395973132425, "learning_rate": 8.417185837038133e-06, "loss": 0.1583, "step": 5174 }, { "epoch": 0.7816039873130947, "grad_norm": 2.2393830531970855, "learning_rate": 8.416608491781075e-06, "loss": 0.14, "step": 5175 }, { "epoch": 0.7817550219000151, "grad_norm": 1.9671009030341498, "learning_rate": 8.416031061055677e-06, "loss": 0.1304, "step": 5176 }, { "epoch": 0.7819060564869356, "grad_norm": 2.299802598716901, "learning_rate": 8.415453544876382e-06, "loss": 0.1548, "step": 5177 }, { "epoch": 0.7820570910738559, "grad_norm": 2.4471038402309233, "learning_rate": 8.414875943257638e-06, "loss": 0.1617, "step": 5178 }, { "epoch": 0.7822081256607764, "grad_norm": 2.1552656079109784, "learning_rate": 8.414298256213892e-06, "loss": 0.1694, "step": 5179 }, { "epoch": 0.7823591602476967, "grad_norm": 1.8697308281973628, "learning_rate": 8.413720483759598e-06, "loss": 0.1271, "step": 5180 }, { "epoch": 0.7825101948346171, "grad_norm": 2.3404706281384224, "learning_rate": 8.413142625909207e-06, "loss": 0.1341, "step": 5181 }, { "epoch": 0.7826612294215375, "grad_norm": 2.0090177135748433, "learning_rate": 8.412564682677176e-06, "loss": 0.1359, "step": 5182 }, { "epoch": 0.7828122640084579, "grad_norm": 2.2247752156680023, "learning_rate": 8.411986654077961e-06, "loss": 0.1584, "step": 5183 }, { "epoch": 0.7829632985953784, "grad_norm": 2.530834802624412, "learning_rate": 8.411408540126023e-06, "loss": 0.1652, "step": 5184 }, { "epoch": 0.7831143331822987, "grad_norm": 2.08998030566527, "learning_rate": 8.410830340835824e-06, "loss": 0.161, "step": 5185 }, { "epoch": 0.7832653677692192, "grad_norm": 2.5645627445742822, "learning_rate": 8.410252056221826e-06, "loss": 0.1491, "step": 5186 }, { "epoch": 0.7834164023561395, "grad_norm": 2.174645075547037, "learning_rate": 8.4096736862985e-06, "loss": 0.1577, "step": 5187 }, { "epoch": 0.78356743694306, "grad_norm": 1.8361709365482195, "learning_rate": 8.409095231080307e-06, "loss": 0.124, "step": 5188 }, { "epoch": 0.7837184715299804, "grad_norm": 2.2824430018158015, "learning_rate": 8.408516690581721e-06, "loss": 0.1363, "step": 5189 }, { "epoch": 0.7838695061169008, "grad_norm": 2.187400089509381, "learning_rate": 8.407938064817216e-06, "loss": 0.1368, "step": 5190 }, { "epoch": 0.7840205407038212, "grad_norm": 2.5990574921356555, "learning_rate": 8.407359353801266e-06, "loss": 0.1358, "step": 5191 }, { "epoch": 0.7841715752907416, "grad_norm": 1.9184916809724977, "learning_rate": 8.406780557548346e-06, "loss": 0.1426, "step": 5192 }, { "epoch": 0.784322609877662, "grad_norm": 2.060504025595185, "learning_rate": 8.406201676072935e-06, "loss": 0.1564, "step": 5193 }, { "epoch": 0.7844736444645823, "grad_norm": 2.5826053965647726, "learning_rate": 8.405622709389517e-06, "loss": 0.146, "step": 5194 }, { "epoch": 0.7846246790515028, "grad_norm": 2.6342140126140055, "learning_rate": 8.405043657512572e-06, "loss": 0.1486, "step": 5195 }, { "epoch": 0.7847757136384232, "grad_norm": 2.7868039399553926, "learning_rate": 8.404464520456588e-06, "loss": 0.1675, "step": 5196 }, { "epoch": 0.7849267482253436, "grad_norm": 2.939344224723795, "learning_rate": 8.40388529823605e-06, "loss": 0.1389, "step": 5197 }, { "epoch": 0.785077782812264, "grad_norm": 2.358489588253979, "learning_rate": 8.403305990865447e-06, "loss": 0.1692, "step": 5198 }, { "epoch": 0.7852288173991844, "grad_norm": 1.6145853014238705, "learning_rate": 8.402726598359274e-06, "loss": 0.1269, "step": 5199 }, { "epoch": 0.7853798519861048, "grad_norm": 1.9379079918439088, "learning_rate": 8.402147120732021e-06, "loss": 0.1339, "step": 5200 }, { "epoch": 0.7855308865730253, "grad_norm": 2.419443289359307, "learning_rate": 8.401567557998188e-06, "loss": 0.1627, "step": 5201 }, { "epoch": 0.7856819211599456, "grad_norm": 1.8573353225714375, "learning_rate": 8.40098791017227e-06, "loss": 0.1306, "step": 5202 }, { "epoch": 0.7858329557468661, "grad_norm": 1.6218955743080945, "learning_rate": 8.400408177268768e-06, "loss": 0.1356, "step": 5203 }, { "epoch": 0.7859839903337864, "grad_norm": 2.230758522824415, "learning_rate": 8.399828359302185e-06, "loss": 0.1548, "step": 5204 }, { "epoch": 0.7861350249207069, "grad_norm": 1.6020181429281053, "learning_rate": 8.399248456287025e-06, "loss": 0.1075, "step": 5205 }, { "epoch": 0.7862860595076272, "grad_norm": 2.707418285980251, "learning_rate": 8.398668468237796e-06, "loss": 0.1781, "step": 5206 }, { "epoch": 0.7864370940945477, "grad_norm": 2.5137612722875757, "learning_rate": 8.398088395169002e-06, "loss": 0.1796, "step": 5207 }, { "epoch": 0.7865881286814681, "grad_norm": 1.8779245060755325, "learning_rate": 8.39750823709516e-06, "loss": 0.1357, "step": 5208 }, { "epoch": 0.7867391632683884, "grad_norm": 1.9939047525276556, "learning_rate": 8.396927994030778e-06, "loss": 0.1348, "step": 5209 }, { "epoch": 0.7868901978553089, "grad_norm": 2.522380906204141, "learning_rate": 8.396347665990376e-06, "loss": 0.1782, "step": 5210 }, { "epoch": 0.7870412324422292, "grad_norm": 2.2862684332199534, "learning_rate": 8.395767252988467e-06, "loss": 0.1412, "step": 5211 }, { "epoch": 0.7871922670291497, "grad_norm": 1.8590191050865703, "learning_rate": 8.395186755039571e-06, "loss": 0.1511, "step": 5212 }, { "epoch": 0.7873433016160701, "grad_norm": 2.212716050969119, "learning_rate": 8.394606172158212e-06, "loss": 0.1424, "step": 5213 }, { "epoch": 0.7874943362029905, "grad_norm": 1.8905007952054065, "learning_rate": 8.394025504358912e-06, "loss": 0.1337, "step": 5214 }, { "epoch": 0.7876453707899109, "grad_norm": 1.9634889625732885, "learning_rate": 8.393444751656194e-06, "loss": 0.1484, "step": 5215 }, { "epoch": 0.7877964053768313, "grad_norm": 2.2238648833401427, "learning_rate": 8.392863914064591e-06, "loss": 0.1262, "step": 5216 }, { "epoch": 0.7879474399637517, "grad_norm": 1.7566528516508437, "learning_rate": 8.392282991598629e-06, "loss": 0.141, "step": 5217 }, { "epoch": 0.7880984745506721, "grad_norm": 1.884361622846155, "learning_rate": 8.391701984272841e-06, "loss": 0.1479, "step": 5218 }, { "epoch": 0.7882495091375925, "grad_norm": 1.9886570946265474, "learning_rate": 8.391120892101763e-06, "loss": 0.1496, "step": 5219 }, { "epoch": 0.788400543724513, "grad_norm": 4.35234397751012, "learning_rate": 8.390539715099931e-06, "loss": 0.1638, "step": 5220 }, { "epoch": 0.7885515783114333, "grad_norm": 2.2793382851781576, "learning_rate": 8.389958453281882e-06, "loss": 0.1684, "step": 5221 }, { "epoch": 0.7887026128983537, "grad_norm": 2.4805927596882835, "learning_rate": 8.389377106662155e-06, "loss": 0.1564, "step": 5222 }, { "epoch": 0.7888536474852741, "grad_norm": 3.1404476736586195, "learning_rate": 8.388795675255297e-06, "loss": 0.1378, "step": 5223 }, { "epoch": 0.7890046820721945, "grad_norm": 2.065542444921722, "learning_rate": 8.38821415907585e-06, "loss": 0.146, "step": 5224 }, { "epoch": 0.789155716659115, "grad_norm": 1.7333229082836914, "learning_rate": 8.387632558138363e-06, "loss": 0.1398, "step": 5225 }, { "epoch": 0.7893067512460353, "grad_norm": 2.212850972380492, "learning_rate": 8.387050872457381e-06, "loss": 0.1337, "step": 5226 }, { "epoch": 0.7894577858329558, "grad_norm": 1.654115121498465, "learning_rate": 8.38646910204746e-06, "loss": 0.1344, "step": 5227 }, { "epoch": 0.7896088204198761, "grad_norm": 1.967689076719385, "learning_rate": 8.385887246923152e-06, "loss": 0.1407, "step": 5228 }, { "epoch": 0.7897598550067966, "grad_norm": 2.3168216461885733, "learning_rate": 8.38530530709901e-06, "loss": 0.1765, "step": 5229 }, { "epoch": 0.7899108895937169, "grad_norm": 2.8931928892733985, "learning_rate": 8.384723282589592e-06, "loss": 0.1469, "step": 5230 }, { "epoch": 0.7900619241806374, "grad_norm": 2.43268644341388, "learning_rate": 8.384141173409461e-06, "loss": 0.1666, "step": 5231 }, { "epoch": 0.7902129587675578, "grad_norm": 2.242282506784333, "learning_rate": 8.383558979573175e-06, "loss": 0.1556, "step": 5232 }, { "epoch": 0.7903639933544782, "grad_norm": 2.1499108240083076, "learning_rate": 8.3829767010953e-06, "loss": 0.147, "step": 5233 }, { "epoch": 0.7905150279413986, "grad_norm": 2.219854091796997, "learning_rate": 8.382394337990403e-06, "loss": 0.1516, "step": 5234 }, { "epoch": 0.790666062528319, "grad_norm": 2.2850908974329105, "learning_rate": 8.38181189027305e-06, "loss": 0.1572, "step": 5235 }, { "epoch": 0.7908170971152394, "grad_norm": 1.942904619181691, "learning_rate": 8.38122935795781e-06, "loss": 0.1591, "step": 5236 }, { "epoch": 0.7909681317021598, "grad_norm": 2.6010227862291693, "learning_rate": 8.38064674105926e-06, "loss": 0.1671, "step": 5237 }, { "epoch": 0.7911191662890802, "grad_norm": 2.233142536000535, "learning_rate": 8.38006403959197e-06, "loss": 0.1713, "step": 5238 }, { "epoch": 0.7912702008760006, "grad_norm": 2.032337265682637, "learning_rate": 8.379481253570519e-06, "loss": 0.1447, "step": 5239 }, { "epoch": 0.791421235462921, "grad_norm": 2.073722216936039, "learning_rate": 8.378898383009484e-06, "loss": 0.1501, "step": 5240 }, { "epoch": 0.7915722700498414, "grad_norm": 2.265877772554198, "learning_rate": 8.378315427923447e-06, "loss": 0.1637, "step": 5241 }, { "epoch": 0.7917233046367618, "grad_norm": 1.7835976569948901, "learning_rate": 8.377732388326992e-06, "loss": 0.125, "step": 5242 }, { "epoch": 0.7918743392236822, "grad_norm": 2.0514669199927784, "learning_rate": 8.377149264234703e-06, "loss": 0.1485, "step": 5243 }, { "epoch": 0.7920253738106027, "grad_norm": 2.3452470261178053, "learning_rate": 8.376566055661167e-06, "loss": 0.1632, "step": 5244 }, { "epoch": 0.792176408397523, "grad_norm": 1.9905395271008275, "learning_rate": 8.375982762620972e-06, "loss": 0.147, "step": 5245 }, { "epoch": 0.7923274429844435, "grad_norm": 1.6822990657560228, "learning_rate": 8.37539938512871e-06, "loss": 0.1211, "step": 5246 }, { "epoch": 0.7924784775713638, "grad_norm": 2.383532176344015, "learning_rate": 8.374815923198977e-06, "loss": 0.1499, "step": 5247 }, { "epoch": 0.7926295121582843, "grad_norm": 2.1453803261115025, "learning_rate": 8.374232376846367e-06, "loss": 0.1499, "step": 5248 }, { "epoch": 0.7927805467452047, "grad_norm": 2.0872929183448066, "learning_rate": 8.373648746085479e-06, "loss": 0.1391, "step": 5249 }, { "epoch": 0.792931581332125, "grad_norm": 1.8078036969464877, "learning_rate": 8.37306503093091e-06, "loss": 0.1376, "step": 5250 }, { "epoch": 0.7930826159190455, "grad_norm": 1.7419490124398183, "learning_rate": 8.372481231397263e-06, "loss": 0.1248, "step": 5251 }, { "epoch": 0.7932336505059658, "grad_norm": 2.1293566269871547, "learning_rate": 8.371897347499141e-06, "loss": 0.132, "step": 5252 }, { "epoch": 0.7933846850928863, "grad_norm": 2.1026736797881753, "learning_rate": 8.371313379251153e-06, "loss": 0.1711, "step": 5253 }, { "epoch": 0.7935357196798066, "grad_norm": 2.098204534775286, "learning_rate": 8.370729326667908e-06, "loss": 0.1486, "step": 5254 }, { "epoch": 0.7936867542667271, "grad_norm": 1.6527926293131556, "learning_rate": 8.370145189764011e-06, "loss": 0.1147, "step": 5255 }, { "epoch": 0.7938377888536475, "grad_norm": 2.113774694868925, "learning_rate": 8.36956096855408e-06, "loss": 0.1231, "step": 5256 }, { "epoch": 0.7939888234405679, "grad_norm": 1.9020783555013117, "learning_rate": 8.368976663052726e-06, "loss": 0.1499, "step": 5257 }, { "epoch": 0.7941398580274883, "grad_norm": 2.026719975886632, "learning_rate": 8.368392273274568e-06, "loss": 0.1522, "step": 5258 }, { "epoch": 0.7942908926144087, "grad_norm": 1.7124594368005932, "learning_rate": 8.367807799234225e-06, "loss": 0.1441, "step": 5259 }, { "epoch": 0.7944419272013291, "grad_norm": 1.9081062955805783, "learning_rate": 8.367223240946314e-06, "loss": 0.157, "step": 5260 }, { "epoch": 0.7945929617882496, "grad_norm": 2.4611886080462977, "learning_rate": 8.366638598425463e-06, "loss": 0.1879, "step": 5261 }, { "epoch": 0.7947439963751699, "grad_norm": 2.384719926164827, "learning_rate": 8.366053871686296e-06, "loss": 0.1883, "step": 5262 }, { "epoch": 0.7948950309620904, "grad_norm": 2.4690702669204994, "learning_rate": 8.365469060743438e-06, "loss": 0.1445, "step": 5263 }, { "epoch": 0.7950460655490107, "grad_norm": 2.565847795579378, "learning_rate": 8.36488416561152e-06, "loss": 0.1527, "step": 5264 }, { "epoch": 0.7951971001359311, "grad_norm": 2.1517230409762007, "learning_rate": 8.364299186305174e-06, "loss": 0.1507, "step": 5265 }, { "epoch": 0.7953481347228515, "grad_norm": 2.098803142213556, "learning_rate": 8.363714122839033e-06, "loss": 0.1549, "step": 5266 }, { "epoch": 0.7954991693097719, "grad_norm": 2.048406639348931, "learning_rate": 8.363128975227733e-06, "loss": 0.1541, "step": 5267 }, { "epoch": 0.7956502038966924, "grad_norm": 1.876139151670438, "learning_rate": 8.36254374348591e-06, "loss": 0.133, "step": 5268 }, { "epoch": 0.7958012384836127, "grad_norm": 1.7305675113146324, "learning_rate": 8.361958427628203e-06, "loss": 0.1321, "step": 5269 }, { "epoch": 0.7959522730705332, "grad_norm": 1.9507861566674471, "learning_rate": 8.361373027669258e-06, "loss": 0.1323, "step": 5270 }, { "epoch": 0.7961033076574535, "grad_norm": 2.240475521102372, "learning_rate": 8.360787543623719e-06, "loss": 0.1601, "step": 5271 }, { "epoch": 0.796254342244374, "grad_norm": 1.7376571148818745, "learning_rate": 8.360201975506228e-06, "loss": 0.1319, "step": 5272 }, { "epoch": 0.7964053768312944, "grad_norm": 1.9762104680250987, "learning_rate": 8.359616323331438e-06, "loss": 0.1437, "step": 5273 }, { "epoch": 0.7965564114182148, "grad_norm": 1.6634037326718925, "learning_rate": 8.359030587113997e-06, "loss": 0.1246, "step": 5274 }, { "epoch": 0.7967074460051352, "grad_norm": 1.5342010127857602, "learning_rate": 8.358444766868556e-06, "loss": 0.1143, "step": 5275 }, { "epoch": 0.7968584805920556, "grad_norm": 2.3575662027070883, "learning_rate": 8.357858862609773e-06, "loss": 0.1697, "step": 5276 }, { "epoch": 0.797009515178976, "grad_norm": 3.2298443605906724, "learning_rate": 8.357272874352303e-06, "loss": 0.144, "step": 5277 }, { "epoch": 0.7971605497658963, "grad_norm": 2.363909664159229, "learning_rate": 8.356686802110803e-06, "loss": 0.1791, "step": 5278 }, { "epoch": 0.7973115843528168, "grad_norm": 1.9235771652581404, "learning_rate": 8.356100645899939e-06, "loss": 0.1412, "step": 5279 }, { "epoch": 0.7974626189397372, "grad_norm": 1.876706617056202, "learning_rate": 8.355514405734369e-06, "loss": 0.166, "step": 5280 }, { "epoch": 0.7976136535266576, "grad_norm": 1.7335679498821581, "learning_rate": 8.354928081628758e-06, "loss": 0.1292, "step": 5281 }, { "epoch": 0.797764688113578, "grad_norm": 2.05161500382401, "learning_rate": 8.354341673597779e-06, "loss": 0.1464, "step": 5282 }, { "epoch": 0.7979157227004984, "grad_norm": 1.8076525245576682, "learning_rate": 8.353755181656095e-06, "loss": 0.1352, "step": 5283 }, { "epoch": 0.7980667572874188, "grad_norm": 2.4494263383747312, "learning_rate": 8.35316860581838e-06, "loss": 0.1389, "step": 5284 }, { "epoch": 0.7982177918743393, "grad_norm": 4.599723906921978, "learning_rate": 8.352581946099308e-06, "loss": 0.2035, "step": 5285 }, { "epoch": 0.7983688264612596, "grad_norm": 1.6130670207649624, "learning_rate": 8.351995202513554e-06, "loss": 0.1333, "step": 5286 }, { "epoch": 0.7985198610481801, "grad_norm": 2.227951202162024, "learning_rate": 8.351408375075796e-06, "loss": 0.1699, "step": 5287 }, { "epoch": 0.7986708956351004, "grad_norm": 1.9086670809122153, "learning_rate": 8.350821463800714e-06, "loss": 0.1368, "step": 5288 }, { "epoch": 0.7988219302220209, "grad_norm": 2.3120224163587757, "learning_rate": 8.350234468702988e-06, "loss": 0.175, "step": 5289 }, { "epoch": 0.7989729648089412, "grad_norm": 2.8282307295875557, "learning_rate": 8.349647389797304e-06, "loss": 0.1448, "step": 5290 }, { "epoch": 0.7991239993958617, "grad_norm": 1.9484333529720086, "learning_rate": 8.349060227098347e-06, "loss": 0.1486, "step": 5291 }, { "epoch": 0.7992750339827821, "grad_norm": 4.833205459111938, "learning_rate": 8.348472980620807e-06, "loss": 0.1691, "step": 5292 }, { "epoch": 0.7994260685697024, "grad_norm": 1.7511266207313283, "learning_rate": 8.347885650379372e-06, "loss": 0.1311, "step": 5293 }, { "epoch": 0.7995771031566229, "grad_norm": 1.605783541729805, "learning_rate": 8.347298236388737e-06, "loss": 0.132, "step": 5294 }, { "epoch": 0.7997281377435432, "grad_norm": 2.0980675882438393, "learning_rate": 8.346710738663594e-06, "loss": 0.1505, "step": 5295 }, { "epoch": 0.7998791723304637, "grad_norm": 2.71374556988463, "learning_rate": 8.34612315721864e-06, "loss": 0.1377, "step": 5296 }, { "epoch": 0.8000302069173841, "grad_norm": 2.5566882952769894, "learning_rate": 8.345535492068575e-06, "loss": 0.18, "step": 5297 }, { "epoch": 0.8001812415043045, "grad_norm": 1.7298638029568647, "learning_rate": 8.3449477432281e-06, "loss": 0.1322, "step": 5298 }, { "epoch": 0.8003322760912249, "grad_norm": 1.9145987102999222, "learning_rate": 8.344359910711914e-06, "loss": 0.1586, "step": 5299 }, { "epoch": 0.8004833106781453, "grad_norm": 2.0868452492447913, "learning_rate": 8.343771994534728e-06, "loss": 0.1352, "step": 5300 }, { "epoch": 0.8006343452650657, "grad_norm": 5.821660879052488, "learning_rate": 8.343183994711244e-06, "loss": 0.1521, "step": 5301 }, { "epoch": 0.8007853798519861, "grad_norm": 2.3640576736729395, "learning_rate": 8.342595911256173e-06, "loss": 0.1476, "step": 5302 }, { "epoch": 0.8009364144389065, "grad_norm": 2.443123589008131, "learning_rate": 8.342007744184226e-06, "loss": 0.1365, "step": 5303 }, { "epoch": 0.801087449025827, "grad_norm": 2.48339274632438, "learning_rate": 8.341419493510116e-06, "loss": 0.1536, "step": 5304 }, { "epoch": 0.8012384836127473, "grad_norm": 2.440338710900628, "learning_rate": 8.34083115924856e-06, "loss": 0.1466, "step": 5305 }, { "epoch": 0.8013895181996677, "grad_norm": 2.861048792708165, "learning_rate": 8.340242741414275e-06, "loss": 0.1793, "step": 5306 }, { "epoch": 0.8015405527865881, "grad_norm": 1.4747204442474717, "learning_rate": 8.33965424002198e-06, "loss": 0.1109, "step": 5307 }, { "epoch": 0.8016915873735085, "grad_norm": 2.0425268862102373, "learning_rate": 8.339065655086395e-06, "loss": 0.1352, "step": 5308 }, { "epoch": 0.801842621960429, "grad_norm": 4.103593044537323, "learning_rate": 8.338476986622247e-06, "loss": 0.1915, "step": 5309 }, { "epoch": 0.8019936565473493, "grad_norm": 2.5235824032281653, "learning_rate": 8.33788823464426e-06, "loss": 0.1811, "step": 5310 }, { "epoch": 0.8021446911342698, "grad_norm": 1.966319173653265, "learning_rate": 8.337299399167161e-06, "loss": 0.1372, "step": 5311 }, { "epoch": 0.8022957257211901, "grad_norm": 2.0755852636500243, "learning_rate": 8.336710480205683e-06, "loss": 0.1427, "step": 5312 }, { "epoch": 0.8024467603081106, "grad_norm": 2.7217716129694693, "learning_rate": 8.336121477774555e-06, "loss": 0.1791, "step": 5313 }, { "epoch": 0.8025977948950309, "grad_norm": 2.282550618380362, "learning_rate": 8.335532391888515e-06, "loss": 0.1498, "step": 5314 }, { "epoch": 0.8027488294819514, "grad_norm": 2.095018196524542, "learning_rate": 8.334943222562294e-06, "loss": 0.1311, "step": 5315 }, { "epoch": 0.8028998640688718, "grad_norm": 2.122648838553249, "learning_rate": 8.334353969810636e-06, "loss": 0.154, "step": 5316 }, { "epoch": 0.8030508986557922, "grad_norm": 1.7732031538195967, "learning_rate": 8.333764633648278e-06, "loss": 0.1344, "step": 5317 }, { "epoch": 0.8032019332427126, "grad_norm": 1.811947268871892, "learning_rate": 8.333175214089963e-06, "loss": 0.1464, "step": 5318 }, { "epoch": 0.803352967829633, "grad_norm": 1.8864926999068263, "learning_rate": 8.332585711150437e-06, "loss": 0.1344, "step": 5319 }, { "epoch": 0.8035040024165534, "grad_norm": 2.0128167987112837, "learning_rate": 8.331996124844445e-06, "loss": 0.1349, "step": 5320 }, { "epoch": 0.8036550370034738, "grad_norm": 2.1398832323933967, "learning_rate": 8.331406455186738e-06, "loss": 0.1393, "step": 5321 }, { "epoch": 0.8038060715903942, "grad_norm": 3.3674096524514665, "learning_rate": 8.330816702192064e-06, "loss": 0.1881, "step": 5322 }, { "epoch": 0.8039571061773146, "grad_norm": 2.0926676238922477, "learning_rate": 8.330226865875178e-06, "loss": 0.1409, "step": 5323 }, { "epoch": 0.804108140764235, "grad_norm": 1.813458671566494, "learning_rate": 8.329636946250835e-06, "loss": 0.1266, "step": 5324 }, { "epoch": 0.8042591753511554, "grad_norm": 2.0198770955938747, "learning_rate": 8.329046943333794e-06, "loss": 0.1606, "step": 5325 }, { "epoch": 0.8044102099380758, "grad_norm": 2.0620826996354653, "learning_rate": 8.328456857138809e-06, "loss": 0.1328, "step": 5326 }, { "epoch": 0.8045612445249962, "grad_norm": 1.984783240393048, "learning_rate": 8.327866687680645e-06, "loss": 0.1452, "step": 5327 }, { "epoch": 0.8047122791119167, "grad_norm": 1.9477745301447331, "learning_rate": 8.327276434974065e-06, "loss": 0.1479, "step": 5328 }, { "epoch": 0.804863313698837, "grad_norm": 2.5410805873934987, "learning_rate": 8.326686099033835e-06, "loss": 0.1412, "step": 5329 }, { "epoch": 0.8050143482857575, "grad_norm": 2.415458060723016, "learning_rate": 8.32609567987472e-06, "loss": 0.1661, "step": 5330 }, { "epoch": 0.8051653828726778, "grad_norm": 2.0724878249065024, "learning_rate": 8.325505177511492e-06, "loss": 0.154, "step": 5331 }, { "epoch": 0.8053164174595983, "grad_norm": 2.313646755146059, "learning_rate": 8.324914591958925e-06, "loss": 0.1601, "step": 5332 }, { "epoch": 0.8054674520465187, "grad_norm": 2.3013427616902815, "learning_rate": 8.324323923231788e-06, "loss": 0.1603, "step": 5333 }, { "epoch": 0.805618486633439, "grad_norm": 2.7632317404130213, "learning_rate": 8.323733171344858e-06, "loss": 0.1598, "step": 5334 }, { "epoch": 0.8057695212203595, "grad_norm": 1.8739456064575206, "learning_rate": 8.323142336312917e-06, "loss": 0.122, "step": 5335 }, { "epoch": 0.8059205558072798, "grad_norm": 2.08281737528909, "learning_rate": 8.32255141815074e-06, "loss": 0.166, "step": 5336 }, { "epoch": 0.8060715903942003, "grad_norm": 1.957536357258332, "learning_rate": 8.321960416873111e-06, "loss": 0.1357, "step": 5337 }, { "epoch": 0.8062226249811206, "grad_norm": 2.013547222720117, "learning_rate": 8.321369332494816e-06, "loss": 0.12, "step": 5338 }, { "epoch": 0.8063736595680411, "grad_norm": 2.522422880973609, "learning_rate": 8.32077816503064e-06, "loss": 0.1621, "step": 5339 }, { "epoch": 0.8065246941549615, "grad_norm": 1.8653251570085958, "learning_rate": 8.32018691449537e-06, "loss": 0.1532, "step": 5340 }, { "epoch": 0.8066757287418819, "grad_norm": 1.796457013662872, "learning_rate": 8.319595580903796e-06, "loss": 0.1253, "step": 5341 }, { "epoch": 0.8068267633288023, "grad_norm": 2.3601142078714727, "learning_rate": 8.319004164270714e-06, "loss": 0.1107, "step": 5342 }, { "epoch": 0.8069777979157227, "grad_norm": 2.118764409741539, "learning_rate": 8.318412664610917e-06, "loss": 0.1633, "step": 5343 }, { "epoch": 0.8071288325026431, "grad_norm": 2.177289050275393, "learning_rate": 8.3178210819392e-06, "loss": 0.1484, "step": 5344 }, { "epoch": 0.8072798670895636, "grad_norm": 2.008618339647114, "learning_rate": 8.317229416270364e-06, "loss": 0.1613, "step": 5345 }, { "epoch": 0.8074309016764839, "grad_norm": 1.4598218616540029, "learning_rate": 8.31663766761921e-06, "loss": 0.1103, "step": 5346 }, { "epoch": 0.8075819362634044, "grad_norm": 1.739306999453904, "learning_rate": 8.316045836000539e-06, "loss": 0.1415, "step": 5347 }, { "epoch": 0.8077329708503247, "grad_norm": 1.813177539967365, "learning_rate": 8.315453921429155e-06, "loss": 0.1366, "step": 5348 }, { "epoch": 0.8078840054372451, "grad_norm": 3.1793952193827746, "learning_rate": 8.31486192391987e-06, "loss": 0.1602, "step": 5349 }, { "epoch": 0.8080350400241655, "grad_norm": 2.112659767835216, "learning_rate": 8.314269843487489e-06, "loss": 0.1632, "step": 5350 }, { "epoch": 0.8081860746110859, "grad_norm": 2.90464793663306, "learning_rate": 8.313677680146825e-06, "loss": 0.1359, "step": 5351 }, { "epoch": 0.8083371091980064, "grad_norm": 2.343902216948462, "learning_rate": 8.313085433912689e-06, "loss": 0.1433, "step": 5352 }, { "epoch": 0.8084881437849267, "grad_norm": 1.9521499156488051, "learning_rate": 8.3124931047999e-06, "loss": 0.1358, "step": 5353 }, { "epoch": 0.8086391783718472, "grad_norm": 2.181687125778404, "learning_rate": 8.311900692823272e-06, "loss": 0.1447, "step": 5354 }, { "epoch": 0.8087902129587675, "grad_norm": 2.0982166910530706, "learning_rate": 8.311308197997627e-06, "loss": 0.1442, "step": 5355 }, { "epoch": 0.808941247545688, "grad_norm": 2.0155431187993806, "learning_rate": 8.310715620337785e-06, "loss": 0.1434, "step": 5356 }, { "epoch": 0.8090922821326084, "grad_norm": 1.7531644592215272, "learning_rate": 8.310122959858571e-06, "loss": 0.1377, "step": 5357 }, { "epoch": 0.8092433167195288, "grad_norm": 2.395790763103377, "learning_rate": 8.309530216574808e-06, "loss": 0.1614, "step": 5358 }, { "epoch": 0.8093943513064492, "grad_norm": 1.929276994671551, "learning_rate": 8.308937390501327e-06, "loss": 0.1429, "step": 5359 }, { "epoch": 0.8095453858933696, "grad_norm": 2.5304931803411903, "learning_rate": 8.308344481652956e-06, "loss": 0.1511, "step": 5360 }, { "epoch": 0.80969642048029, "grad_norm": 2.020191758718576, "learning_rate": 8.30775149004453e-06, "loss": 0.1377, "step": 5361 }, { "epoch": 0.8098474550672103, "grad_norm": 1.8697093479348545, "learning_rate": 8.307158415690878e-06, "loss": 0.137, "step": 5362 }, { "epoch": 0.8099984896541308, "grad_norm": 2.0420233568148607, "learning_rate": 8.306565258606838e-06, "loss": 0.1482, "step": 5363 }, { "epoch": 0.8101495242410512, "grad_norm": 2.078393562128612, "learning_rate": 8.305972018807252e-06, "loss": 0.1362, "step": 5364 }, { "epoch": 0.8103005588279716, "grad_norm": 1.9199353889639932, "learning_rate": 8.305378696306955e-06, "loss": 0.1406, "step": 5365 }, { "epoch": 0.810451593414892, "grad_norm": 2.282969498941226, "learning_rate": 8.30478529112079e-06, "loss": 0.1713, "step": 5366 }, { "epoch": 0.8106026280018124, "grad_norm": 1.8954572305820307, "learning_rate": 8.304191803263603e-06, "loss": 0.1438, "step": 5367 }, { "epoch": 0.8107536625887328, "grad_norm": 1.9840529135157725, "learning_rate": 8.303598232750243e-06, "loss": 0.1325, "step": 5368 }, { "epoch": 0.8109046971756533, "grad_norm": 1.901748183760559, "learning_rate": 8.303004579595553e-06, "loss": 0.1441, "step": 5369 }, { "epoch": 0.8110557317625736, "grad_norm": 2.076172553227144, "learning_rate": 8.302410843814387e-06, "loss": 0.1492, "step": 5370 }, { "epoch": 0.8112067663494941, "grad_norm": 2.38741217522588, "learning_rate": 8.301817025421597e-06, "loss": 0.1555, "step": 5371 }, { "epoch": 0.8113578009364144, "grad_norm": 2.097454120269727, "learning_rate": 8.301223124432037e-06, "loss": 0.1608, "step": 5372 }, { "epoch": 0.8115088355233349, "grad_norm": 1.8625165608506011, "learning_rate": 8.300629140860565e-06, "loss": 0.1238, "step": 5373 }, { "epoch": 0.8116598701102552, "grad_norm": 2.0228348485367618, "learning_rate": 8.300035074722037e-06, "loss": 0.154, "step": 5374 }, { "epoch": 0.8118109046971757, "grad_norm": 2.400616542346102, "learning_rate": 8.29944092603132e-06, "loss": 0.1566, "step": 5375 }, { "epoch": 0.8119619392840961, "grad_norm": 2.7533397533096906, "learning_rate": 8.298846694803268e-06, "loss": 0.1515, "step": 5376 }, { "epoch": 0.8121129738710164, "grad_norm": 2.35259181655217, "learning_rate": 8.298252381052754e-06, "loss": 0.1642, "step": 5377 }, { "epoch": 0.8122640084579369, "grad_norm": 1.9829086582584328, "learning_rate": 8.29765798479464e-06, "loss": 0.1455, "step": 5378 }, { "epoch": 0.8124150430448572, "grad_norm": 1.8510715088379432, "learning_rate": 8.2970635060438e-06, "loss": 0.1288, "step": 5379 }, { "epoch": 0.8125660776317777, "grad_norm": 1.8943642805440628, "learning_rate": 8.2964689448151e-06, "loss": 0.1587, "step": 5380 }, { "epoch": 0.8127171122186981, "grad_norm": 2.253137717860372, "learning_rate": 8.295874301123415e-06, "loss": 0.1445, "step": 5381 }, { "epoch": 0.8128681468056185, "grad_norm": 1.9647121083755472, "learning_rate": 8.295279574983623e-06, "loss": 0.133, "step": 5382 }, { "epoch": 0.8130191813925389, "grad_norm": 1.9032296324622193, "learning_rate": 8.2946847664106e-06, "loss": 0.1421, "step": 5383 }, { "epoch": 0.8131702159794593, "grad_norm": 1.9492969648693823, "learning_rate": 8.294089875419223e-06, "loss": 0.1488, "step": 5384 }, { "epoch": 0.8133212505663797, "grad_norm": 1.9046182404907992, "learning_rate": 8.293494902024376e-06, "loss": 0.1388, "step": 5385 }, { "epoch": 0.8134722851533001, "grad_norm": 2.019557400341214, "learning_rate": 8.29289984624094e-06, "loss": 0.142, "step": 5386 }, { "epoch": 0.8136233197402205, "grad_norm": 1.943976515994665, "learning_rate": 8.292304708083805e-06, "loss": 0.1411, "step": 5387 }, { "epoch": 0.813774354327141, "grad_norm": 2.0289770376434637, "learning_rate": 8.291709487567857e-06, "loss": 0.1337, "step": 5388 }, { "epoch": 0.8139253889140613, "grad_norm": 2.5420514864681514, "learning_rate": 8.291114184707983e-06, "loss": 0.1613, "step": 5389 }, { "epoch": 0.8140764235009817, "grad_norm": 2.587167317475122, "learning_rate": 8.290518799519078e-06, "loss": 0.1474, "step": 5390 }, { "epoch": 0.8142274580879021, "grad_norm": 2.2514905378235213, "learning_rate": 8.289923332016034e-06, "loss": 0.1482, "step": 5391 }, { "epoch": 0.8143784926748225, "grad_norm": 1.6370620978568096, "learning_rate": 8.289327782213747e-06, "loss": 0.121, "step": 5392 }, { "epoch": 0.814529527261743, "grad_norm": 2.28495353894177, "learning_rate": 8.288732150127118e-06, "loss": 0.167, "step": 5393 }, { "epoch": 0.8146805618486633, "grad_norm": 1.7385243359322196, "learning_rate": 8.288136435771044e-06, "loss": 0.1193, "step": 5394 }, { "epoch": 0.8148315964355838, "grad_norm": 1.8579568510235684, "learning_rate": 8.287540639160428e-06, "loss": 0.1503, "step": 5395 }, { "epoch": 0.8149826310225041, "grad_norm": 1.8804342831133432, "learning_rate": 8.286944760310174e-06, "loss": 0.1445, "step": 5396 }, { "epoch": 0.8151336656094246, "grad_norm": 1.8277427482305493, "learning_rate": 8.286348799235188e-06, "loss": 0.1419, "step": 5397 }, { "epoch": 0.8152847001963449, "grad_norm": 2.079232827247737, "learning_rate": 8.28575275595038e-06, "loss": 0.1506, "step": 5398 }, { "epoch": 0.8154357347832654, "grad_norm": 1.800778996087322, "learning_rate": 8.285156630470656e-06, "loss": 0.1395, "step": 5399 }, { "epoch": 0.8155867693701858, "grad_norm": 2.108299416068304, "learning_rate": 8.284560422810935e-06, "loss": 0.1654, "step": 5400 }, { "epoch": 0.8157378039571062, "grad_norm": 2.2492224459531034, "learning_rate": 8.283964132986125e-06, "loss": 0.1281, "step": 5401 }, { "epoch": 0.8158888385440266, "grad_norm": 2.2782082786412263, "learning_rate": 8.283367761011147e-06, "loss": 0.1553, "step": 5402 }, { "epoch": 0.816039873130947, "grad_norm": 3.2147687010137638, "learning_rate": 8.282771306900917e-06, "loss": 0.1546, "step": 5403 }, { "epoch": 0.8161909077178674, "grad_norm": 2.3038646177119917, "learning_rate": 8.282174770670358e-06, "loss": 0.1455, "step": 5404 }, { "epoch": 0.8163419423047878, "grad_norm": 1.7664544961707511, "learning_rate": 8.28157815233439e-06, "loss": 0.1264, "step": 5405 }, { "epoch": 0.8164929768917082, "grad_norm": 2.027677144657719, "learning_rate": 8.28098145190794e-06, "loss": 0.1472, "step": 5406 }, { "epoch": 0.8166440114786286, "grad_norm": 2.1848959658837446, "learning_rate": 8.280384669405933e-06, "loss": 0.1652, "step": 5407 }, { "epoch": 0.816795046065549, "grad_norm": 2.0017099009706985, "learning_rate": 8.279787804843298e-06, "loss": 0.147, "step": 5408 }, { "epoch": 0.8169460806524694, "grad_norm": 2.191493746466064, "learning_rate": 8.279190858234966e-06, "loss": 0.1562, "step": 5409 }, { "epoch": 0.8170971152393898, "grad_norm": 3.5701702907694726, "learning_rate": 8.278593829595873e-06, "loss": 0.1523, "step": 5410 }, { "epoch": 0.8172481498263102, "grad_norm": 1.8892197442297882, "learning_rate": 8.27799671894095e-06, "loss": 0.1274, "step": 5411 }, { "epoch": 0.8173991844132307, "grad_norm": 2.6608139439497256, "learning_rate": 8.277399526285137e-06, "loss": 0.1482, "step": 5412 }, { "epoch": 0.817550219000151, "grad_norm": 3.0673885625881234, "learning_rate": 8.27680225164337e-06, "loss": 0.1809, "step": 5413 }, { "epoch": 0.8177012535870715, "grad_norm": 1.9656232024398468, "learning_rate": 8.27620489503059e-06, "loss": 0.128, "step": 5414 }, { "epoch": 0.8178522881739918, "grad_norm": 1.9421546846220628, "learning_rate": 8.275607456461747e-06, "loss": 0.14, "step": 5415 }, { "epoch": 0.8180033227609123, "grad_norm": 2.172074040065623, "learning_rate": 8.275009935951778e-06, "loss": 0.1693, "step": 5416 }, { "epoch": 0.8181543573478327, "grad_norm": 1.7955824715540218, "learning_rate": 8.274412333515633e-06, "loss": 0.1325, "step": 5417 }, { "epoch": 0.818305391934753, "grad_norm": 2.0139227875181276, "learning_rate": 8.273814649168263e-06, "loss": 0.1664, "step": 5418 }, { "epoch": 0.8184564265216735, "grad_norm": 2.603881063756068, "learning_rate": 8.27321688292462e-06, "loss": 0.1584, "step": 5419 }, { "epoch": 0.8186074611085938, "grad_norm": 1.974770236579096, "learning_rate": 8.272619034799652e-06, "loss": 0.1548, "step": 5420 }, { "epoch": 0.8187584956955143, "grad_norm": 1.9686112013600854, "learning_rate": 8.27202110480832e-06, "loss": 0.1617, "step": 5421 }, { "epoch": 0.8189095302824346, "grad_norm": 2.8071028671592266, "learning_rate": 8.27142309296558e-06, "loss": 0.1767, "step": 5422 }, { "epoch": 0.8190605648693551, "grad_norm": 1.9449272116920056, "learning_rate": 8.270824999286394e-06, "loss": 0.1389, "step": 5423 }, { "epoch": 0.8192115994562755, "grad_norm": 2.387436480176195, "learning_rate": 8.270226823785717e-06, "loss": 0.1665, "step": 5424 }, { "epoch": 0.8193626340431959, "grad_norm": 1.7514214951764084, "learning_rate": 8.269628566478518e-06, "loss": 0.1111, "step": 5425 }, { "epoch": 0.8195136686301163, "grad_norm": 2.005676404435623, "learning_rate": 8.269030227379763e-06, "loss": 0.13, "step": 5426 }, { "epoch": 0.8196647032170367, "grad_norm": 2.2123567369992245, "learning_rate": 8.268431806504415e-06, "loss": 0.1454, "step": 5427 }, { "epoch": 0.8198157378039571, "grad_norm": 2.3146877264250123, "learning_rate": 8.26783330386745e-06, "loss": 0.1621, "step": 5428 }, { "epoch": 0.8199667723908776, "grad_norm": 2.1498842710791997, "learning_rate": 8.267234719483837e-06, "loss": 0.139, "step": 5429 }, { "epoch": 0.8201178069777979, "grad_norm": 1.8627638122136225, "learning_rate": 8.266636053368549e-06, "loss": 0.1283, "step": 5430 }, { "epoch": 0.8202688415647184, "grad_norm": 2.1912843275024523, "learning_rate": 8.266037305536561e-06, "loss": 0.1416, "step": 5431 }, { "epoch": 0.8204198761516387, "grad_norm": 2.123663223747669, "learning_rate": 8.265438476002854e-06, "loss": 0.1439, "step": 5432 }, { "epoch": 0.8205709107385591, "grad_norm": 2.32990138222652, "learning_rate": 8.264839564782409e-06, "loss": 0.1404, "step": 5433 }, { "epoch": 0.8207219453254795, "grad_norm": 1.7649744346044174, "learning_rate": 8.264240571890204e-06, "loss": 0.1407, "step": 5434 }, { "epoch": 0.8208729799123999, "grad_norm": 1.8901116640226825, "learning_rate": 8.263641497341226e-06, "loss": 0.1374, "step": 5435 }, { "epoch": 0.8210240144993204, "grad_norm": 2.3259721325454703, "learning_rate": 8.26304234115046e-06, "loss": 0.162, "step": 5436 }, { "epoch": 0.8211750490862407, "grad_norm": 2.2859313372693437, "learning_rate": 8.262443103332893e-06, "loss": 0.1304, "step": 5437 }, { "epoch": 0.8213260836731612, "grad_norm": 2.334341707914066, "learning_rate": 8.26184378390352e-06, "loss": 0.1438, "step": 5438 }, { "epoch": 0.8214771182600815, "grad_norm": 2.538754975513185, "learning_rate": 8.261244382877327e-06, "loss": 0.176, "step": 5439 }, { "epoch": 0.821628152847002, "grad_norm": 2.1885690199033028, "learning_rate": 8.260644900269314e-06, "loss": 0.1475, "step": 5440 }, { "epoch": 0.8217791874339224, "grad_norm": 1.8202470272199327, "learning_rate": 8.260045336094473e-06, "loss": 0.1414, "step": 5441 }, { "epoch": 0.8219302220208428, "grad_norm": 2.0100765603455693, "learning_rate": 8.259445690367804e-06, "loss": 0.1301, "step": 5442 }, { "epoch": 0.8220812566077632, "grad_norm": 1.8065369239594729, "learning_rate": 8.258845963104309e-06, "loss": 0.1475, "step": 5443 }, { "epoch": 0.8222322911946836, "grad_norm": 2.321306034392668, "learning_rate": 8.25824615431899e-06, "loss": 0.162, "step": 5444 }, { "epoch": 0.822383325781604, "grad_norm": 2.1720204459859356, "learning_rate": 8.257646264026849e-06, "loss": 0.1772, "step": 5445 }, { "epoch": 0.8225343603685243, "grad_norm": 1.9568012058461473, "learning_rate": 8.257046292242894e-06, "loss": 0.157, "step": 5446 }, { "epoch": 0.8226853949554448, "grad_norm": 1.980082429800694, "learning_rate": 8.256446238982134e-06, "loss": 0.1621, "step": 5447 }, { "epoch": 0.8228364295423652, "grad_norm": 3.7994621415291694, "learning_rate": 8.255846104259582e-06, "loss": 0.1637, "step": 5448 }, { "epoch": 0.8229874641292856, "grad_norm": 1.8396838114716467, "learning_rate": 8.255245888090247e-06, "loss": 0.1353, "step": 5449 }, { "epoch": 0.823138498716206, "grad_norm": 2.243615982731391, "learning_rate": 8.254645590489143e-06, "loss": 0.172, "step": 5450 }, { "epoch": 0.8232895333031264, "grad_norm": 2.551736487888975, "learning_rate": 8.254045211471292e-06, "loss": 0.1533, "step": 5451 }, { "epoch": 0.8234405678900468, "grad_norm": 1.6996247394613484, "learning_rate": 8.253444751051709e-06, "loss": 0.1333, "step": 5452 }, { "epoch": 0.8235916024769673, "grad_norm": 1.584279917332289, "learning_rate": 8.252844209245414e-06, "loss": 0.1211, "step": 5453 }, { "epoch": 0.8237426370638876, "grad_norm": 5.460308657543578, "learning_rate": 8.252243586067432e-06, "loss": 0.1551, "step": 5454 }, { "epoch": 0.8238936716508081, "grad_norm": 1.8917141395218646, "learning_rate": 8.251642881532788e-06, "loss": 0.151, "step": 5455 }, { "epoch": 0.8240447062377284, "grad_norm": 2.3541829629018065, "learning_rate": 8.251042095656507e-06, "loss": 0.1595, "step": 5456 }, { "epoch": 0.8241957408246489, "grad_norm": 1.992660975828303, "learning_rate": 8.25044122845362e-06, "loss": 0.1601, "step": 5457 }, { "epoch": 0.8243467754115692, "grad_norm": 2.4888571951414287, "learning_rate": 8.249840279939159e-06, "loss": 0.1448, "step": 5458 }, { "epoch": 0.8244978099984897, "grad_norm": 2.3694564719330904, "learning_rate": 8.249239250128154e-06, "loss": 0.1379, "step": 5459 }, { "epoch": 0.8246488445854101, "grad_norm": 2.2763710359067275, "learning_rate": 8.24863813903564e-06, "loss": 0.1306, "step": 5460 }, { "epoch": 0.8247998791723304, "grad_norm": 2.2032350748977847, "learning_rate": 8.248036946676658e-06, "loss": 0.1393, "step": 5461 }, { "epoch": 0.8249509137592509, "grad_norm": 2.0827183650521124, "learning_rate": 8.247435673066243e-06, "loss": 0.1561, "step": 5462 }, { "epoch": 0.8251019483461712, "grad_norm": 2.011045795745961, "learning_rate": 8.246834318219438e-06, "loss": 0.1478, "step": 5463 }, { "epoch": 0.8252529829330917, "grad_norm": 3.0951371063570368, "learning_rate": 8.246232882151286e-06, "loss": 0.1871, "step": 5464 }, { "epoch": 0.8254040175200121, "grad_norm": 3.3402019657151447, "learning_rate": 8.245631364876834e-06, "loss": 0.1614, "step": 5465 }, { "epoch": 0.8255550521069325, "grad_norm": 2.7532015816281836, "learning_rate": 8.245029766411127e-06, "loss": 0.1727, "step": 5466 }, { "epoch": 0.8257060866938529, "grad_norm": 1.8773159924631133, "learning_rate": 8.244428086769213e-06, "loss": 0.1371, "step": 5467 }, { "epoch": 0.8258571212807733, "grad_norm": 2.2708707104445307, "learning_rate": 8.243826325966147e-06, "loss": 0.1711, "step": 5468 }, { "epoch": 0.8260081558676937, "grad_norm": 2.6886120651069345, "learning_rate": 8.243224484016981e-06, "loss": 0.134, "step": 5469 }, { "epoch": 0.8261591904546141, "grad_norm": 1.7655703779306682, "learning_rate": 8.242622560936771e-06, "loss": 0.1381, "step": 5470 }, { "epoch": 0.8263102250415345, "grad_norm": 1.8748299222870195, "learning_rate": 8.242020556740572e-06, "loss": 0.1531, "step": 5471 }, { "epoch": 0.826461259628455, "grad_norm": 3.4868657487648047, "learning_rate": 8.241418471443446e-06, "loss": 0.1434, "step": 5472 }, { "epoch": 0.8266122942153753, "grad_norm": 1.635818111604559, "learning_rate": 8.240816305060452e-06, "loss": 0.1109, "step": 5473 }, { "epoch": 0.8267633288022957, "grad_norm": 2.2022510218543805, "learning_rate": 8.240214057606656e-06, "loss": 0.1509, "step": 5474 }, { "epoch": 0.8269143633892161, "grad_norm": 2.1216854554950717, "learning_rate": 8.239611729097123e-06, "loss": 0.1696, "step": 5475 }, { "epoch": 0.8270653979761365, "grad_norm": 2.1110053812029883, "learning_rate": 8.23900931954692e-06, "loss": 0.1332, "step": 5476 }, { "epoch": 0.827216432563057, "grad_norm": 2.6543416803363953, "learning_rate": 8.238406828971115e-06, "loss": 0.1543, "step": 5477 }, { "epoch": 0.8273674671499773, "grad_norm": 2.3390868142474304, "learning_rate": 8.237804257384783e-06, "loss": 0.1631, "step": 5478 }, { "epoch": 0.8275185017368978, "grad_norm": 2.7103869006303754, "learning_rate": 8.237201604802997e-06, "loss": 0.1714, "step": 5479 }, { "epoch": 0.8276695363238181, "grad_norm": 1.979014506657392, "learning_rate": 8.23659887124083e-06, "loss": 0.1387, "step": 5480 }, { "epoch": 0.8278205709107386, "grad_norm": 1.9210534071073369, "learning_rate": 8.235996056713362e-06, "loss": 0.1428, "step": 5481 }, { "epoch": 0.8279716054976589, "grad_norm": 7.070301877210198, "learning_rate": 8.235393161235673e-06, "loss": 0.1873, "step": 5482 }, { "epoch": 0.8281226400845794, "grad_norm": 2.065724377979385, "learning_rate": 8.234790184822844e-06, "loss": 0.1498, "step": 5483 }, { "epoch": 0.8282736746714998, "grad_norm": 2.4777574181558313, "learning_rate": 8.234187127489957e-06, "loss": 0.1441, "step": 5484 }, { "epoch": 0.8284247092584202, "grad_norm": 2.0016467859863436, "learning_rate": 8.2335839892521e-06, "loss": 0.1528, "step": 5485 }, { "epoch": 0.8285757438453406, "grad_norm": 1.8784659692243053, "learning_rate": 8.23298077012436e-06, "loss": 0.14, "step": 5486 }, { "epoch": 0.828726778432261, "grad_norm": 2.05705408798112, "learning_rate": 8.23237747012183e-06, "loss": 0.1505, "step": 5487 }, { "epoch": 0.8288778130191814, "grad_norm": 2.007756532915438, "learning_rate": 8.231774089259599e-06, "loss": 0.1359, "step": 5488 }, { "epoch": 0.8290288476061018, "grad_norm": 2.06019453702073, "learning_rate": 8.231170627552758e-06, "loss": 0.1527, "step": 5489 }, { "epoch": 0.8291798821930222, "grad_norm": 2.858673095000776, "learning_rate": 8.230567085016407e-06, "loss": 0.1774, "step": 5490 }, { "epoch": 0.8293309167799426, "grad_norm": 2.3435139623758348, "learning_rate": 8.229963461665645e-06, "loss": 0.1337, "step": 5491 }, { "epoch": 0.829481951366863, "grad_norm": 1.9408260805372315, "learning_rate": 8.229359757515567e-06, "loss": 0.1537, "step": 5492 }, { "epoch": 0.8296329859537834, "grad_norm": 1.9455444634870933, "learning_rate": 8.228755972581279e-06, "loss": 0.1438, "step": 5493 }, { "epoch": 0.8297840205407038, "grad_norm": 1.8131452314747094, "learning_rate": 8.228152106877884e-06, "loss": 0.1332, "step": 5494 }, { "epoch": 0.8299350551276242, "grad_norm": 2.1774770251277937, "learning_rate": 8.227548160420488e-06, "loss": 0.1481, "step": 5495 }, { "epoch": 0.8300860897145447, "grad_norm": 1.6216943543398095, "learning_rate": 8.2269441332242e-06, "loss": 0.1246, "step": 5496 }, { "epoch": 0.830237124301465, "grad_norm": 2.1943353805580066, "learning_rate": 8.226340025304129e-06, "loss": 0.1375, "step": 5497 }, { "epoch": 0.8303881588883855, "grad_norm": 2.8489176201007673, "learning_rate": 8.225735836675387e-06, "loss": 0.1646, "step": 5498 }, { "epoch": 0.8305391934753058, "grad_norm": 1.556885729963384, "learning_rate": 8.225131567353087e-06, "loss": 0.1227, "step": 5499 }, { "epoch": 0.8306902280622263, "grad_norm": 3.2588498604730405, "learning_rate": 8.224527217352345e-06, "loss": 0.1686, "step": 5500 }, { "epoch": 0.8308412626491467, "grad_norm": 2.3541504423034807, "learning_rate": 8.223922786688283e-06, "loss": 0.1722, "step": 5501 }, { "epoch": 0.830992297236067, "grad_norm": 1.787152700383822, "learning_rate": 8.223318275376017e-06, "loss": 0.1232, "step": 5502 }, { "epoch": 0.8311433318229875, "grad_norm": 3.013346175081442, "learning_rate": 8.222713683430674e-06, "loss": 0.1155, "step": 5503 }, { "epoch": 0.8312943664099078, "grad_norm": 2.617154588380394, "learning_rate": 8.222109010867372e-06, "loss": 0.18, "step": 5504 }, { "epoch": 0.8314454009968283, "grad_norm": 2.2427290912171896, "learning_rate": 8.221504257701241e-06, "loss": 0.1464, "step": 5505 }, { "epoch": 0.8315964355837486, "grad_norm": 2.035034977320431, "learning_rate": 8.220899423947409e-06, "loss": 0.1502, "step": 5506 }, { "epoch": 0.8317474701706691, "grad_norm": 1.9340586731533382, "learning_rate": 8.220294509621005e-06, "loss": 0.134, "step": 5507 }, { "epoch": 0.8318985047575895, "grad_norm": 4.328480094511842, "learning_rate": 8.219689514737164e-06, "loss": 0.1639, "step": 5508 }, { "epoch": 0.8320495393445099, "grad_norm": 1.863473881904576, "learning_rate": 8.219084439311018e-06, "loss": 0.1308, "step": 5509 }, { "epoch": 0.8322005739314303, "grad_norm": 1.8304973819478283, "learning_rate": 8.218479283357702e-06, "loss": 0.1284, "step": 5510 }, { "epoch": 0.8323516085183507, "grad_norm": 2.3424912199641645, "learning_rate": 8.217874046892356e-06, "loss": 0.1365, "step": 5511 }, { "epoch": 0.8325026431052711, "grad_norm": 2.4664496928964206, "learning_rate": 8.217268729930121e-06, "loss": 0.1881, "step": 5512 }, { "epoch": 0.8326536776921916, "grad_norm": 2.7723246863915465, "learning_rate": 8.216663332486138e-06, "loss": 0.1721, "step": 5513 }, { "epoch": 0.8328047122791119, "grad_norm": 2.121436052883297, "learning_rate": 8.216057854575552e-06, "loss": 0.1509, "step": 5514 }, { "epoch": 0.8329557468660324, "grad_norm": 2.2633047914302478, "learning_rate": 8.21545229621351e-06, "loss": 0.1519, "step": 5515 }, { "epoch": 0.8331067814529527, "grad_norm": 2.055533535274603, "learning_rate": 8.21484665741516e-06, "loss": 0.1565, "step": 5516 }, { "epoch": 0.8332578160398731, "grad_norm": 1.9430729201651191, "learning_rate": 8.21424093819565e-06, "loss": 0.1628, "step": 5517 }, { "epoch": 0.8334088506267935, "grad_norm": 2.7558612284385986, "learning_rate": 8.213635138570138e-06, "loss": 0.1674, "step": 5518 }, { "epoch": 0.8335598852137139, "grad_norm": 2.1957681042881645, "learning_rate": 8.213029258553772e-06, "loss": 0.1509, "step": 5519 }, { "epoch": 0.8337109198006344, "grad_norm": 1.7492578116956339, "learning_rate": 8.212423298161712e-06, "loss": 0.1332, "step": 5520 }, { "epoch": 0.8338619543875547, "grad_norm": 1.833332297108785, "learning_rate": 8.211817257409116e-06, "loss": 0.1329, "step": 5521 }, { "epoch": 0.8340129889744752, "grad_norm": 3.3185426589968015, "learning_rate": 8.211211136311144e-06, "loss": 0.164, "step": 5522 }, { "epoch": 0.8341640235613955, "grad_norm": 1.7944247241773696, "learning_rate": 8.210604934882957e-06, "loss": 0.111, "step": 5523 }, { "epoch": 0.834315058148316, "grad_norm": 1.6002023495657867, "learning_rate": 8.209998653139723e-06, "loss": 0.1081, "step": 5524 }, { "epoch": 0.8344660927352364, "grad_norm": 1.9004497732815526, "learning_rate": 8.209392291096607e-06, "loss": 0.14, "step": 5525 }, { "epoch": 0.8346171273221568, "grad_norm": 2.204525154559175, "learning_rate": 8.208785848768776e-06, "loss": 0.1364, "step": 5526 }, { "epoch": 0.8347681619090772, "grad_norm": 2.087629491652422, "learning_rate": 8.208179326171402e-06, "loss": 0.1462, "step": 5527 }, { "epoch": 0.8349191964959976, "grad_norm": 2.1753221148953474, "learning_rate": 8.207572723319656e-06, "loss": 0.1537, "step": 5528 }, { "epoch": 0.835070231082918, "grad_norm": 1.9007990528353391, "learning_rate": 8.206966040228715e-06, "loss": 0.1533, "step": 5529 }, { "epoch": 0.8352212656698383, "grad_norm": 1.8697564975214969, "learning_rate": 8.206359276913754e-06, "loss": 0.1327, "step": 5530 }, { "epoch": 0.8353723002567588, "grad_norm": 1.8157600089725674, "learning_rate": 8.205752433389951e-06, "loss": 0.1495, "step": 5531 }, { "epoch": 0.8355233348436792, "grad_norm": 1.8901480903384196, "learning_rate": 8.205145509672486e-06, "loss": 0.1241, "step": 5532 }, { "epoch": 0.8356743694305996, "grad_norm": 1.8813642434461058, "learning_rate": 8.204538505776544e-06, "loss": 0.1239, "step": 5533 }, { "epoch": 0.83582540401752, "grad_norm": 2.4237418124505496, "learning_rate": 8.203931421717309e-06, "loss": 0.1559, "step": 5534 }, { "epoch": 0.8359764386044404, "grad_norm": 2.1161886803023244, "learning_rate": 8.203324257509965e-06, "loss": 0.1479, "step": 5535 }, { "epoch": 0.8361274731913608, "grad_norm": 2.0608968667905816, "learning_rate": 8.2027170131697e-06, "loss": 0.1479, "step": 5536 }, { "epoch": 0.8362785077782813, "grad_norm": 1.7651861181290383, "learning_rate": 8.202109688711712e-06, "loss": 0.1559, "step": 5537 }, { "epoch": 0.8364295423652016, "grad_norm": 1.8281634628674857, "learning_rate": 8.201502284151186e-06, "loss": 0.1448, "step": 5538 }, { "epoch": 0.8365805769521221, "grad_norm": 1.8492073517624896, "learning_rate": 8.200894799503318e-06, "loss": 0.1325, "step": 5539 }, { "epoch": 0.8367316115390424, "grad_norm": 1.9033008304068098, "learning_rate": 8.200287234783306e-06, "loss": 0.1342, "step": 5540 }, { "epoch": 0.8368826461259629, "grad_norm": 2.0585746611686497, "learning_rate": 8.199679590006349e-06, "loss": 0.1564, "step": 5541 }, { "epoch": 0.8370336807128832, "grad_norm": 1.9453118844302295, "learning_rate": 8.199071865187645e-06, "loss": 0.1125, "step": 5542 }, { "epoch": 0.8371847152998037, "grad_norm": 1.706057963783321, "learning_rate": 8.1984640603424e-06, "loss": 0.1326, "step": 5543 }, { "epoch": 0.8373357498867241, "grad_norm": 2.1573834390923876, "learning_rate": 8.197856175485817e-06, "loss": 0.1413, "step": 5544 }, { "epoch": 0.8374867844736444, "grad_norm": 1.9585401167682235, "learning_rate": 8.197248210633099e-06, "loss": 0.1473, "step": 5545 }, { "epoch": 0.8376378190605649, "grad_norm": 2.4535953994864665, "learning_rate": 8.19664016579946e-06, "loss": 0.1558, "step": 5546 }, { "epoch": 0.8377888536474852, "grad_norm": 1.7784924854844455, "learning_rate": 8.196032041000108e-06, "loss": 0.13, "step": 5547 }, { "epoch": 0.8379398882344057, "grad_norm": 2.429366219315052, "learning_rate": 8.195423836250258e-06, "loss": 0.1478, "step": 5548 }, { "epoch": 0.8380909228213261, "grad_norm": 2.865150543197724, "learning_rate": 8.19481555156512e-06, "loss": 0.1545, "step": 5549 }, { "epoch": 0.8382419574082465, "grad_norm": 1.7250244877655658, "learning_rate": 8.194207186959915e-06, "loss": 0.1185, "step": 5550 }, { "epoch": 0.8383929919951669, "grad_norm": 2.136284859401028, "learning_rate": 8.19359874244986e-06, "loss": 0.1638, "step": 5551 }, { "epoch": 0.8385440265820873, "grad_norm": 2.2215211495466396, "learning_rate": 8.192990218050174e-06, "loss": 0.1435, "step": 5552 }, { "epoch": 0.8386950611690077, "grad_norm": 3.0935969968436083, "learning_rate": 8.192381613776082e-06, "loss": 0.1515, "step": 5553 }, { "epoch": 0.8388460957559281, "grad_norm": 1.78659939806429, "learning_rate": 8.191772929642807e-06, "loss": 0.1348, "step": 5554 }, { "epoch": 0.8389971303428485, "grad_norm": 1.9549719588746708, "learning_rate": 8.191164165665576e-06, "loss": 0.1502, "step": 5555 }, { "epoch": 0.839148164929769, "grad_norm": 2.6034685236577637, "learning_rate": 8.190555321859617e-06, "loss": 0.1744, "step": 5556 }, { "epoch": 0.8392991995166893, "grad_norm": 1.86325163083891, "learning_rate": 8.189946398240162e-06, "loss": 0.1276, "step": 5557 }, { "epoch": 0.8394502341036097, "grad_norm": 2.283766419185674, "learning_rate": 8.189337394822442e-06, "loss": 0.1366, "step": 5558 }, { "epoch": 0.8396012686905301, "grad_norm": 3.5538735623067805, "learning_rate": 8.188728311621693e-06, "loss": 0.1772, "step": 5559 }, { "epoch": 0.8397523032774505, "grad_norm": 2.693088437072799, "learning_rate": 8.188119148653151e-06, "loss": 0.1845, "step": 5560 }, { "epoch": 0.839903337864371, "grad_norm": 1.6804343478471067, "learning_rate": 8.187509905932055e-06, "loss": 0.1325, "step": 5561 }, { "epoch": 0.8400543724512913, "grad_norm": 2.1920892616578143, "learning_rate": 8.186900583473643e-06, "loss": 0.153, "step": 5562 }, { "epoch": 0.8402054070382118, "grad_norm": 7.645509325633119, "learning_rate": 8.18629118129316e-06, "loss": 0.2663, "step": 5563 }, { "epoch": 0.8403564416251321, "grad_norm": 2.3659984444421926, "learning_rate": 8.18568169940585e-06, "loss": 0.1551, "step": 5564 }, { "epoch": 0.8405074762120526, "grad_norm": 3.249639044500344, "learning_rate": 8.18507213782696e-06, "loss": 0.1672, "step": 5565 }, { "epoch": 0.8406585107989729, "grad_norm": 2.4263767697919296, "learning_rate": 8.18446249657174e-06, "loss": 0.1792, "step": 5566 }, { "epoch": 0.8408095453858934, "grad_norm": 2.1657679311598668, "learning_rate": 8.183852775655436e-06, "loss": 0.1477, "step": 5567 }, { "epoch": 0.8409605799728138, "grad_norm": 3.0568151727606683, "learning_rate": 8.183242975093304e-06, "loss": 0.1953, "step": 5568 }, { "epoch": 0.8411116145597342, "grad_norm": 2.053121762867845, "learning_rate": 8.182633094900598e-06, "loss": 0.1431, "step": 5569 }, { "epoch": 0.8412626491466546, "grad_norm": 2.683486307071432, "learning_rate": 8.182023135092572e-06, "loss": 0.1554, "step": 5570 }, { "epoch": 0.841413683733575, "grad_norm": 1.8755851011198206, "learning_rate": 8.181413095684487e-06, "loss": 0.1472, "step": 5571 }, { "epoch": 0.8415647183204954, "grad_norm": 1.8525125295821876, "learning_rate": 8.180802976691604e-06, "loss": 0.1333, "step": 5572 }, { "epoch": 0.8417157529074158, "grad_norm": 2.2296642471691754, "learning_rate": 8.180192778129185e-06, "loss": 0.1564, "step": 5573 }, { "epoch": 0.8418667874943362, "grad_norm": 2.560292308871769, "learning_rate": 8.179582500012495e-06, "loss": 0.1404, "step": 5574 }, { "epoch": 0.8420178220812566, "grad_norm": 1.6061444898436295, "learning_rate": 8.178972142356796e-06, "loss": 0.1191, "step": 5575 }, { "epoch": 0.842168856668177, "grad_norm": 2.268412469987364, "learning_rate": 8.178361705177363e-06, "loss": 0.1068, "step": 5576 }, { "epoch": 0.8423198912550974, "grad_norm": 1.770057336857915, "learning_rate": 8.177751188489462e-06, "loss": 0.1366, "step": 5577 }, { "epoch": 0.8424709258420178, "grad_norm": 1.6992921551147304, "learning_rate": 8.177140592308364e-06, "loss": 0.1359, "step": 5578 }, { "epoch": 0.8426219604289382, "grad_norm": 2.0876367629606243, "learning_rate": 8.17652991664935e-06, "loss": 0.1529, "step": 5579 }, { "epoch": 0.8427729950158587, "grad_norm": 1.5660671691965058, "learning_rate": 8.17591916152769e-06, "loss": 0.1251, "step": 5580 }, { "epoch": 0.842924029602779, "grad_norm": 2.405945360630544, "learning_rate": 8.175308326958666e-06, "loss": 0.1871, "step": 5581 }, { "epoch": 0.8430750641896995, "grad_norm": 1.9703584902523696, "learning_rate": 8.174697412957556e-06, "loss": 0.1333, "step": 5582 }, { "epoch": 0.8432260987766198, "grad_norm": 1.8202477363429237, "learning_rate": 8.174086419539644e-06, "loss": 0.1359, "step": 5583 }, { "epoch": 0.8433771333635403, "grad_norm": 2.07190193152742, "learning_rate": 8.173475346720215e-06, "loss": 0.1484, "step": 5584 }, { "epoch": 0.8435281679504607, "grad_norm": 2.5020849245805667, "learning_rate": 8.172864194514554e-06, "loss": 0.1354, "step": 5585 }, { "epoch": 0.843679202537381, "grad_norm": 1.8098045371094318, "learning_rate": 8.172252962937947e-06, "loss": 0.1305, "step": 5586 }, { "epoch": 0.8438302371243015, "grad_norm": 1.6977261946611293, "learning_rate": 8.17164165200569e-06, "loss": 0.1446, "step": 5587 }, { "epoch": 0.8439812717112218, "grad_norm": 10.569085154975745, "learning_rate": 8.171030261733071e-06, "loss": 0.1501, "step": 5588 }, { "epoch": 0.8441323062981423, "grad_norm": 2.0666141276873566, "learning_rate": 8.170418792135383e-06, "loss": 0.1408, "step": 5589 }, { "epoch": 0.8442833408850626, "grad_norm": 1.8729043064243587, "learning_rate": 8.169807243227927e-06, "loss": 0.1337, "step": 5590 }, { "epoch": 0.8444343754719831, "grad_norm": 2.404642553708087, "learning_rate": 8.169195615026e-06, "loss": 0.1632, "step": 5591 }, { "epoch": 0.8445854100589035, "grad_norm": 1.931092937813774, "learning_rate": 8.168583907544898e-06, "loss": 0.128, "step": 5592 }, { "epoch": 0.8447364446458239, "grad_norm": 2.626618928456587, "learning_rate": 8.167972120799928e-06, "loss": 0.1821, "step": 5593 }, { "epoch": 0.8448874792327443, "grad_norm": 2.2832606279148004, "learning_rate": 8.16736025480639e-06, "loss": 0.1629, "step": 5594 }, { "epoch": 0.8450385138196647, "grad_norm": 2.754250288361314, "learning_rate": 8.166748309579596e-06, "loss": 0.1809, "step": 5595 }, { "epoch": 0.8451895484065851, "grad_norm": 1.790623573100094, "learning_rate": 8.16613628513485e-06, "loss": 0.141, "step": 5596 }, { "epoch": 0.8453405829935056, "grad_norm": 2.519886475263267, "learning_rate": 8.16552418148746e-06, "loss": 0.1727, "step": 5597 }, { "epoch": 0.8454916175804259, "grad_norm": 2.1815158130618997, "learning_rate": 8.164911998652743e-06, "loss": 0.1567, "step": 5598 }, { "epoch": 0.8456426521673464, "grad_norm": 1.6827518262526935, "learning_rate": 8.16429973664601e-06, "loss": 0.1243, "step": 5599 }, { "epoch": 0.8457936867542667, "grad_norm": 2.201650205796813, "learning_rate": 8.16368739548258e-06, "loss": 0.1315, "step": 5600 }, { "epoch": 0.8459447213411871, "grad_norm": 2.024476559206982, "learning_rate": 8.163074975177767e-06, "loss": 0.1351, "step": 5601 }, { "epoch": 0.8460957559281075, "grad_norm": 2.0580614045720744, "learning_rate": 8.162462475746895e-06, "loss": 0.1456, "step": 5602 }, { "epoch": 0.8462467905150279, "grad_norm": 2.1683396007632147, "learning_rate": 8.161849897205283e-06, "loss": 0.1504, "step": 5603 }, { "epoch": 0.8463978251019484, "grad_norm": 2.916675593824838, "learning_rate": 8.161237239568257e-06, "loss": 0.1854, "step": 5604 }, { "epoch": 0.8465488596888687, "grad_norm": 1.9944737254874372, "learning_rate": 8.160624502851141e-06, "loss": 0.1476, "step": 5605 }, { "epoch": 0.8466998942757892, "grad_norm": 1.702641870207312, "learning_rate": 8.160011687069267e-06, "loss": 0.1255, "step": 5606 }, { "epoch": 0.8468509288627095, "grad_norm": 2.2227777592776996, "learning_rate": 8.159398792237959e-06, "loss": 0.1667, "step": 5607 }, { "epoch": 0.84700196344963, "grad_norm": 2.015157280690527, "learning_rate": 8.158785818372552e-06, "loss": 0.1393, "step": 5608 }, { "epoch": 0.8471529980365504, "grad_norm": 1.7820534720521348, "learning_rate": 8.158172765488382e-06, "loss": 0.123, "step": 5609 }, { "epoch": 0.8473040326234708, "grad_norm": 1.6304245344835528, "learning_rate": 8.157559633600781e-06, "loss": 0.1305, "step": 5610 }, { "epoch": 0.8474550672103912, "grad_norm": 1.9855949358601623, "learning_rate": 8.15694642272509e-06, "loss": 0.1492, "step": 5611 }, { "epoch": 0.8476061017973116, "grad_norm": 3.727227716321512, "learning_rate": 8.156333132876644e-06, "loss": 0.1901, "step": 5612 }, { "epoch": 0.847757136384232, "grad_norm": 2.343968121981516, "learning_rate": 8.15571976407079e-06, "loss": 0.1572, "step": 5613 }, { "epoch": 0.8479081709711523, "grad_norm": 2.1726711770997373, "learning_rate": 8.155106316322871e-06, "loss": 0.1485, "step": 5614 }, { "epoch": 0.8480592055580728, "grad_norm": 1.7432127985065051, "learning_rate": 8.154492789648231e-06, "loss": 0.128, "step": 5615 }, { "epoch": 0.8482102401449932, "grad_norm": 1.986080339522332, "learning_rate": 8.153879184062219e-06, "loss": 0.1443, "step": 5616 }, { "epoch": 0.8483612747319136, "grad_norm": 1.7582640498215583, "learning_rate": 8.153265499580181e-06, "loss": 0.1338, "step": 5617 }, { "epoch": 0.848512309318834, "grad_norm": 2.1749485532747297, "learning_rate": 8.152651736217473e-06, "loss": 0.1429, "step": 5618 }, { "epoch": 0.8486633439057544, "grad_norm": 1.9220092150471573, "learning_rate": 8.152037893989449e-06, "loss": 0.1454, "step": 5619 }, { "epoch": 0.8488143784926748, "grad_norm": 2.5879186055844396, "learning_rate": 8.151423972911461e-06, "loss": 0.2207, "step": 5620 }, { "epoch": 0.8489654130795953, "grad_norm": 2.159556580864669, "learning_rate": 8.150809972998869e-06, "loss": 0.1454, "step": 5621 }, { "epoch": 0.8491164476665156, "grad_norm": 2.0107649413747177, "learning_rate": 8.150195894267033e-06, "loss": 0.1366, "step": 5622 }, { "epoch": 0.8492674822534361, "grad_norm": 1.60113518104336, "learning_rate": 8.149581736731311e-06, "loss": 0.1172, "step": 5623 }, { "epoch": 0.8494185168403564, "grad_norm": 1.8591989834687586, "learning_rate": 8.14896750040707e-06, "loss": 0.1315, "step": 5624 }, { "epoch": 0.8495695514272769, "grad_norm": 2.037712568066148, "learning_rate": 8.148353185309675e-06, "loss": 0.1482, "step": 5625 }, { "epoch": 0.8497205860141972, "grad_norm": 3.482024822053955, "learning_rate": 8.147738791454493e-06, "loss": 0.153, "step": 5626 }, { "epoch": 0.8498716206011176, "grad_norm": 2.577460924618208, "learning_rate": 8.147124318856894e-06, "loss": 0.1607, "step": 5627 }, { "epoch": 0.8500226551880381, "grad_norm": 2.1412122943905563, "learning_rate": 8.146509767532246e-06, "loss": 0.1594, "step": 5628 }, { "epoch": 0.8501736897749584, "grad_norm": 1.710602803698743, "learning_rate": 8.145895137495927e-06, "loss": 0.1497, "step": 5629 }, { "epoch": 0.8503247243618789, "grad_norm": 2.977225746505502, "learning_rate": 8.14528042876331e-06, "loss": 0.1576, "step": 5630 }, { "epoch": 0.8504757589487992, "grad_norm": 1.7830559555005487, "learning_rate": 8.14466564134977e-06, "loss": 0.1305, "step": 5631 }, { "epoch": 0.8506267935357197, "grad_norm": 3.488387196956335, "learning_rate": 8.144050775270693e-06, "loss": 0.1628, "step": 5632 }, { "epoch": 0.8507778281226401, "grad_norm": 2.0970027049458517, "learning_rate": 8.143435830541453e-06, "loss": 0.1647, "step": 5633 }, { "epoch": 0.8509288627095605, "grad_norm": 2.037055713912173, "learning_rate": 8.142820807177439e-06, "loss": 0.1665, "step": 5634 }, { "epoch": 0.8510798972964809, "grad_norm": 1.7955846549651533, "learning_rate": 8.142205705194031e-06, "loss": 0.13, "step": 5635 }, { "epoch": 0.8512309318834013, "grad_norm": 3.0951361715760384, "learning_rate": 8.141590524606618e-06, "loss": 0.1936, "step": 5636 }, { "epoch": 0.8513819664703217, "grad_norm": 1.8683847800364015, "learning_rate": 8.14097526543059e-06, "loss": 0.1373, "step": 5637 }, { "epoch": 0.8515330010572421, "grad_norm": 2.3923133501403764, "learning_rate": 8.14035992768134e-06, "loss": 0.1667, "step": 5638 }, { "epoch": 0.8516840356441625, "grad_norm": 2.0404147685342022, "learning_rate": 8.139744511374256e-06, "loss": 0.1397, "step": 5639 }, { "epoch": 0.851835070231083, "grad_norm": 1.9856208927139565, "learning_rate": 8.139129016524736e-06, "loss": 0.1456, "step": 5640 }, { "epoch": 0.8519861048180033, "grad_norm": 2.322658399072204, "learning_rate": 8.138513443148177e-06, "loss": 0.1526, "step": 5641 }, { "epoch": 0.8521371394049237, "grad_norm": 1.965427721972455, "learning_rate": 8.137897791259979e-06, "loss": 0.1482, "step": 5642 }, { "epoch": 0.8522881739918441, "grad_norm": 2.685607161510655, "learning_rate": 8.137282060875538e-06, "loss": 0.1823, "step": 5643 }, { "epoch": 0.8524392085787645, "grad_norm": 2.107031492092812, "learning_rate": 8.13666625201026e-06, "loss": 0.1538, "step": 5644 }, { "epoch": 0.852590243165685, "grad_norm": 1.824450847568573, "learning_rate": 8.136050364679552e-06, "loss": 0.1309, "step": 5645 }, { "epoch": 0.8527412777526053, "grad_norm": 3.707942382897375, "learning_rate": 8.135434398898821e-06, "loss": 0.1455, "step": 5646 }, { "epoch": 0.8528923123395258, "grad_norm": 2.1229334603699623, "learning_rate": 8.134818354683469e-06, "loss": 0.165, "step": 5647 }, { "epoch": 0.8530433469264461, "grad_norm": 2.0101647365122126, "learning_rate": 8.134202232048914e-06, "loss": 0.1439, "step": 5648 }, { "epoch": 0.8531943815133666, "grad_norm": 2.4473941293818755, "learning_rate": 8.133586031010564e-06, "loss": 0.1476, "step": 5649 }, { "epoch": 0.8533454161002869, "grad_norm": 1.7277359009771744, "learning_rate": 8.132969751583835e-06, "loss": 0.1313, "step": 5650 }, { "epoch": 0.8534964506872074, "grad_norm": 1.9392583140249862, "learning_rate": 8.132353393784145e-06, "loss": 0.1518, "step": 5651 }, { "epoch": 0.8536474852741278, "grad_norm": 7.009079823192682, "learning_rate": 8.131736957626913e-06, "loss": 0.1429, "step": 5652 }, { "epoch": 0.8537985198610482, "grad_norm": 1.749218443713567, "learning_rate": 8.131120443127555e-06, "loss": 0.1221, "step": 5653 }, { "epoch": 0.8539495544479686, "grad_norm": 2.165939115710217, "learning_rate": 8.130503850301498e-06, "loss": 0.1622, "step": 5654 }, { "epoch": 0.854100589034889, "grad_norm": 2.3566246730301468, "learning_rate": 8.129887179164164e-06, "loss": 0.1392, "step": 5655 }, { "epoch": 0.8542516236218094, "grad_norm": 2.2659587402669645, "learning_rate": 8.12927042973098e-06, "loss": 0.1754, "step": 5656 }, { "epoch": 0.8544026582087298, "grad_norm": 1.97326304435121, "learning_rate": 8.128653602017375e-06, "loss": 0.1448, "step": 5657 }, { "epoch": 0.8545536927956502, "grad_norm": 1.9743590232399257, "learning_rate": 8.12803669603878e-06, "loss": 0.1494, "step": 5658 }, { "epoch": 0.8547047273825706, "grad_norm": 1.938470271169818, "learning_rate": 8.127419711810625e-06, "loss": 0.1411, "step": 5659 }, { "epoch": 0.854855761969491, "grad_norm": 2.4016888992067895, "learning_rate": 8.126802649348344e-06, "loss": 0.1766, "step": 5660 }, { "epoch": 0.8550067965564114, "grad_norm": 2.149023697128591, "learning_rate": 8.126185508667375e-06, "loss": 0.1436, "step": 5661 }, { "epoch": 0.8551578311433318, "grad_norm": 1.9046090247955383, "learning_rate": 8.125568289783156e-06, "loss": 0.137, "step": 5662 }, { "epoch": 0.8553088657302522, "grad_norm": 2.0276287787355307, "learning_rate": 8.124950992711128e-06, "loss": 0.153, "step": 5663 }, { "epoch": 0.8554599003171727, "grad_norm": 1.6289016128694835, "learning_rate": 8.124333617466728e-06, "loss": 0.1314, "step": 5664 }, { "epoch": 0.855610934904093, "grad_norm": 1.6109733033070328, "learning_rate": 8.123716164065407e-06, "loss": 0.1163, "step": 5665 }, { "epoch": 0.8557619694910135, "grad_norm": 2.163329021982913, "learning_rate": 8.123098632522605e-06, "loss": 0.17, "step": 5666 }, { "epoch": 0.8559130040779338, "grad_norm": 2.052790545741618, "learning_rate": 8.122481022853774e-06, "loss": 0.139, "step": 5667 }, { "epoch": 0.8560640386648543, "grad_norm": 1.870996819444729, "learning_rate": 8.12186333507436e-06, "loss": 0.1356, "step": 5668 }, { "epoch": 0.8562150732517747, "grad_norm": 2.1887260800531374, "learning_rate": 8.12124556919982e-06, "loss": 0.1551, "step": 5669 }, { "epoch": 0.856366107838695, "grad_norm": 2.2456540519234833, "learning_rate": 8.120627725245603e-06, "loss": 0.1459, "step": 5670 }, { "epoch": 0.8565171424256155, "grad_norm": 1.554275765532452, "learning_rate": 8.120009803227166e-06, "loss": 0.1188, "step": 5671 }, { "epoch": 0.8566681770125358, "grad_norm": 1.9584317865227752, "learning_rate": 8.119391803159969e-06, "loss": 0.1581, "step": 5672 }, { "epoch": 0.8568192115994563, "grad_norm": 2.2075688718651083, "learning_rate": 8.11877372505947e-06, "loss": 0.1599, "step": 5673 }, { "epoch": 0.8569702461863766, "grad_norm": 1.5761700013104558, "learning_rate": 8.11815556894113e-06, "loss": 0.1299, "step": 5674 }, { "epoch": 0.8571212807732971, "grad_norm": 1.9047570756440246, "learning_rate": 8.11753733482041e-06, "loss": 0.129, "step": 5675 }, { "epoch": 0.8572723153602175, "grad_norm": 1.961667046040436, "learning_rate": 8.11691902271278e-06, "loss": 0.1423, "step": 5676 }, { "epoch": 0.8574233499471379, "grad_norm": 1.651315203591381, "learning_rate": 8.116300632633706e-06, "loss": 0.1121, "step": 5677 }, { "epoch": 0.8575743845340583, "grad_norm": 1.7346536829976014, "learning_rate": 8.115682164598658e-06, "loss": 0.1411, "step": 5678 }, { "epoch": 0.8577254191209787, "grad_norm": 1.9149876017052399, "learning_rate": 8.115063618623105e-06, "loss": 0.1399, "step": 5679 }, { "epoch": 0.8578764537078991, "grad_norm": 2.1823080990207866, "learning_rate": 8.114444994722522e-06, "loss": 0.1511, "step": 5680 }, { "epoch": 0.8580274882948196, "grad_norm": 2.207069533625652, "learning_rate": 8.113826292912385e-06, "loss": 0.153, "step": 5681 }, { "epoch": 0.8581785228817399, "grad_norm": 1.970630778018197, "learning_rate": 8.11320751320817e-06, "loss": 0.1399, "step": 5682 }, { "epoch": 0.8583295574686604, "grad_norm": 1.9333359415641593, "learning_rate": 8.112588655625354e-06, "loss": 0.1472, "step": 5683 }, { "epoch": 0.8584805920555807, "grad_norm": 2.498862037692961, "learning_rate": 8.111969720179423e-06, "loss": 0.1562, "step": 5684 }, { "epoch": 0.8586316266425011, "grad_norm": 2.0741486007786447, "learning_rate": 8.111350706885854e-06, "loss": 0.16, "step": 5685 }, { "epoch": 0.8587826612294215, "grad_norm": 2.2642652881177496, "learning_rate": 8.110731615760139e-06, "loss": 0.1744, "step": 5686 }, { "epoch": 0.8589336958163419, "grad_norm": 1.9212392036467458, "learning_rate": 8.110112446817759e-06, "loss": 0.1474, "step": 5687 }, { "epoch": 0.8590847304032624, "grad_norm": 1.933983412405605, "learning_rate": 8.109493200074205e-06, "loss": 0.1482, "step": 5688 }, { "epoch": 0.8592357649901827, "grad_norm": 1.9739468935403688, "learning_rate": 8.108873875544968e-06, "loss": 0.1432, "step": 5689 }, { "epoch": 0.8593867995771032, "grad_norm": 1.868968853300492, "learning_rate": 8.10825447324554e-06, "loss": 0.1182, "step": 5690 }, { "epoch": 0.8595378341640235, "grad_norm": 2.1052714639736, "learning_rate": 8.107634993191416e-06, "loss": 0.1429, "step": 5691 }, { "epoch": 0.859688868750944, "grad_norm": 1.9548332531076762, "learning_rate": 8.107015435398094e-06, "loss": 0.1357, "step": 5692 }, { "epoch": 0.8598399033378644, "grad_norm": 2.0590171451133377, "learning_rate": 8.106395799881072e-06, "loss": 0.1626, "step": 5693 }, { "epoch": 0.8599909379247848, "grad_norm": 2.3463480024140355, "learning_rate": 8.105776086655849e-06, "loss": 0.1494, "step": 5694 }, { "epoch": 0.8601419725117052, "grad_norm": 1.9944940569929168, "learning_rate": 8.105156295737928e-06, "loss": 0.1479, "step": 5695 }, { "epoch": 0.8602930070986256, "grad_norm": 2.3939573098292404, "learning_rate": 8.104536427142813e-06, "loss": 0.1558, "step": 5696 }, { "epoch": 0.860444041685546, "grad_norm": 1.9983740166250104, "learning_rate": 8.103916480886013e-06, "loss": 0.153, "step": 5697 }, { "epoch": 0.8605950762724663, "grad_norm": 1.8525612392333748, "learning_rate": 8.103296456983035e-06, "loss": 0.1471, "step": 5698 }, { "epoch": 0.8607461108593868, "grad_norm": 1.8519340014215862, "learning_rate": 8.102676355449386e-06, "loss": 0.1435, "step": 5699 }, { "epoch": 0.8608971454463072, "grad_norm": 2.0760823128345973, "learning_rate": 8.102056176300582e-06, "loss": 0.1384, "step": 5700 }, { "epoch": 0.8610481800332276, "grad_norm": 2.542240970880873, "learning_rate": 8.101435919552136e-06, "loss": 0.1613, "step": 5701 }, { "epoch": 0.861199214620148, "grad_norm": 2.261961499287164, "learning_rate": 8.100815585219564e-06, "loss": 0.1678, "step": 5702 }, { "epoch": 0.8613502492070684, "grad_norm": 1.8564065584641407, "learning_rate": 8.100195173318386e-06, "loss": 0.1302, "step": 5703 }, { "epoch": 0.8615012837939888, "grad_norm": 1.9372358390361202, "learning_rate": 8.099574683864117e-06, "loss": 0.1484, "step": 5704 }, { "epoch": 0.8616523183809093, "grad_norm": 2.588207451641574, "learning_rate": 8.098954116872284e-06, "loss": 0.1728, "step": 5705 }, { "epoch": 0.8618033529678296, "grad_norm": 2.3885925987632324, "learning_rate": 8.098333472358408e-06, "loss": 0.1587, "step": 5706 }, { "epoch": 0.8619543875547501, "grad_norm": 1.8125964539772583, "learning_rate": 8.097712750338016e-06, "loss": 0.1378, "step": 5707 }, { "epoch": 0.8621054221416704, "grad_norm": 1.8153318096114073, "learning_rate": 8.097091950826634e-06, "loss": 0.1325, "step": 5708 }, { "epoch": 0.8622564567285909, "grad_norm": 1.9700062689622484, "learning_rate": 8.096471073839794e-06, "loss": 0.1465, "step": 5709 }, { "epoch": 0.8624074913155112, "grad_norm": 1.7823117006573415, "learning_rate": 8.095850119393027e-06, "loss": 0.144, "step": 5710 }, { "epoch": 0.8625585259024316, "grad_norm": 2.332104822374863, "learning_rate": 8.095229087501864e-06, "loss": 0.1618, "step": 5711 }, { "epoch": 0.8627095604893521, "grad_norm": 1.456188826423383, "learning_rate": 8.094607978181845e-06, "loss": 0.1112, "step": 5712 }, { "epoch": 0.8628605950762724, "grad_norm": 1.9743613269560751, "learning_rate": 8.093986791448503e-06, "loss": 0.1387, "step": 5713 }, { "epoch": 0.8630116296631929, "grad_norm": 2.392626617205051, "learning_rate": 8.093365527317378e-06, "loss": 0.1531, "step": 5714 }, { "epoch": 0.8631626642501132, "grad_norm": 2.517416836231689, "learning_rate": 8.092744185804014e-06, "loss": 0.1467, "step": 5715 }, { "epoch": 0.8633136988370337, "grad_norm": 1.9039776681494365, "learning_rate": 8.092122766923953e-06, "loss": 0.1321, "step": 5716 }, { "epoch": 0.8634647334239541, "grad_norm": 2.9001166063605464, "learning_rate": 8.09150127069274e-06, "loss": 0.144, "step": 5717 }, { "epoch": 0.8636157680108745, "grad_norm": 1.82133330934333, "learning_rate": 8.09087969712592e-06, "loss": 0.1403, "step": 5718 }, { "epoch": 0.8637668025977949, "grad_norm": 2.0246068579501717, "learning_rate": 8.090258046239044e-06, "loss": 0.1484, "step": 5719 }, { "epoch": 0.8639178371847153, "grad_norm": 1.9357833238365976, "learning_rate": 8.089636318047663e-06, "loss": 0.1387, "step": 5720 }, { "epoch": 0.8640688717716357, "grad_norm": 3.0245882819672123, "learning_rate": 8.089014512567331e-06, "loss": 0.1735, "step": 5721 }, { "epoch": 0.8642199063585561, "grad_norm": 2.8336846459119402, "learning_rate": 8.0883926298136e-06, "loss": 0.1541, "step": 5722 }, { "epoch": 0.8643709409454765, "grad_norm": 2.9726587041028565, "learning_rate": 8.087770669802027e-06, "loss": 0.1954, "step": 5723 }, { "epoch": 0.864521975532397, "grad_norm": 2.0182090157786003, "learning_rate": 8.087148632548175e-06, "loss": 0.1525, "step": 5724 }, { "epoch": 0.8646730101193173, "grad_norm": 2.7899102361471724, "learning_rate": 8.0865265180676e-06, "loss": 0.1556, "step": 5725 }, { "epoch": 0.8648240447062377, "grad_norm": 2.03452451784295, "learning_rate": 8.085904326375865e-06, "loss": 0.1695, "step": 5726 }, { "epoch": 0.8649750792931581, "grad_norm": 2.127846007534868, "learning_rate": 8.085282057488536e-06, "loss": 0.1484, "step": 5727 }, { "epoch": 0.8651261138800785, "grad_norm": 2.300592211698999, "learning_rate": 8.084659711421179e-06, "loss": 0.1647, "step": 5728 }, { "epoch": 0.865277148466999, "grad_norm": 2.0183030378132987, "learning_rate": 8.08403728818936e-06, "loss": 0.1486, "step": 5729 }, { "epoch": 0.8654281830539193, "grad_norm": 2.5671659240162312, "learning_rate": 8.083414787808654e-06, "loss": 0.1774, "step": 5730 }, { "epoch": 0.8655792176408398, "grad_norm": 1.6691656346218033, "learning_rate": 8.08279221029463e-06, "loss": 0.1258, "step": 5731 }, { "epoch": 0.8657302522277601, "grad_norm": 2.376952126385312, "learning_rate": 8.082169555662864e-06, "loss": 0.1796, "step": 5732 }, { "epoch": 0.8658812868146806, "grad_norm": 1.9657522341606828, "learning_rate": 8.081546823928929e-06, "loss": 0.1656, "step": 5733 }, { "epoch": 0.8660323214016009, "grad_norm": 2.200056442029496, "learning_rate": 8.080924015108405e-06, "loss": 0.1612, "step": 5734 }, { "epoch": 0.8661833559885214, "grad_norm": 1.8835436753170132, "learning_rate": 8.080301129216872e-06, "loss": 0.1413, "step": 5735 }, { "epoch": 0.8663343905754418, "grad_norm": 2.085122782091124, "learning_rate": 8.079678166269911e-06, "loss": 0.1304, "step": 5736 }, { "epoch": 0.8664854251623622, "grad_norm": 2.205637074141168, "learning_rate": 8.079055126283106e-06, "loss": 0.1521, "step": 5737 }, { "epoch": 0.8666364597492826, "grad_norm": 2.394809115282051, "learning_rate": 8.078432009272044e-06, "loss": 0.1468, "step": 5738 }, { "epoch": 0.866787494336203, "grad_norm": 3.096902018244816, "learning_rate": 8.077808815252311e-06, "loss": 0.1342, "step": 5739 }, { "epoch": 0.8669385289231234, "grad_norm": 1.6152934476734782, "learning_rate": 8.077185544239496e-06, "loss": 0.1192, "step": 5740 }, { "epoch": 0.8670895635100438, "grad_norm": 2.21326336683047, "learning_rate": 8.076562196249193e-06, "loss": 0.1525, "step": 5741 }, { "epoch": 0.8672405980969642, "grad_norm": 1.998244027564651, "learning_rate": 8.075938771296995e-06, "loss": 0.1675, "step": 5742 }, { "epoch": 0.8673916326838846, "grad_norm": 1.9989036741764012, "learning_rate": 8.075315269398493e-06, "loss": 0.1248, "step": 5743 }, { "epoch": 0.867542667270805, "grad_norm": 1.8815780658657553, "learning_rate": 8.07469169056929e-06, "loss": 0.1405, "step": 5744 }, { "epoch": 0.8676937018577254, "grad_norm": 2.269239700003006, "learning_rate": 8.074068034824982e-06, "loss": 0.1669, "step": 5745 }, { "epoch": 0.8678447364446458, "grad_norm": 2.17250171317976, "learning_rate": 8.073444302181172e-06, "loss": 0.1347, "step": 5746 }, { "epoch": 0.8679957710315662, "grad_norm": 1.8691620153759918, "learning_rate": 8.07282049265346e-06, "loss": 0.1468, "step": 5747 }, { "epoch": 0.8681468056184867, "grad_norm": 1.870414359407089, "learning_rate": 8.072196606257455e-06, "loss": 0.1295, "step": 5748 }, { "epoch": 0.868297840205407, "grad_norm": 1.9173643984623066, "learning_rate": 8.071572643008759e-06, "loss": 0.1405, "step": 5749 }, { "epoch": 0.8684488747923275, "grad_norm": 2.3736198141636926, "learning_rate": 8.070948602922984e-06, "loss": 0.1393, "step": 5750 }, { "epoch": 0.8685999093792478, "grad_norm": 2.1141033340511446, "learning_rate": 8.07032448601574e-06, "loss": 0.139, "step": 5751 }, { "epoch": 0.8687509439661683, "grad_norm": 2.317892495417363, "learning_rate": 8.06970029230264e-06, "loss": 0.1406, "step": 5752 }, { "epoch": 0.8689019785530887, "grad_norm": 2.4163509843857542, "learning_rate": 8.069076021799297e-06, "loss": 0.1357, "step": 5753 }, { "epoch": 0.869053013140009, "grad_norm": 2.2936221637581675, "learning_rate": 8.068451674521332e-06, "loss": 0.1504, "step": 5754 }, { "epoch": 0.8692040477269295, "grad_norm": 1.6797878557404118, "learning_rate": 8.067827250484358e-06, "loss": 0.1324, "step": 5755 }, { "epoch": 0.8693550823138498, "grad_norm": 1.9028814776421283, "learning_rate": 8.067202749703998e-06, "loss": 0.126, "step": 5756 }, { "epoch": 0.8695061169007703, "grad_norm": 2.4864939477974524, "learning_rate": 8.066578172195872e-06, "loss": 0.2019, "step": 5757 }, { "epoch": 0.8696571514876906, "grad_norm": 2.086182672173682, "learning_rate": 8.065953517975608e-06, "loss": 0.1599, "step": 5758 }, { "epoch": 0.8698081860746111, "grad_norm": 1.7367111821841583, "learning_rate": 8.065328787058829e-06, "loss": 0.1271, "step": 5759 }, { "epoch": 0.8699592206615315, "grad_norm": 2.399400204145638, "learning_rate": 8.064703979461164e-06, "loss": 0.1554, "step": 5760 }, { "epoch": 0.8701102552484519, "grad_norm": 2.3408668949351314, "learning_rate": 8.064079095198242e-06, "loss": 0.1283, "step": 5761 }, { "epoch": 0.8702612898353723, "grad_norm": 1.758914381887841, "learning_rate": 8.063454134285696e-06, "loss": 0.1346, "step": 5762 }, { "epoch": 0.8704123244222927, "grad_norm": 1.8812528909101076, "learning_rate": 8.062829096739158e-06, "loss": 0.1358, "step": 5763 }, { "epoch": 0.8705633590092131, "grad_norm": 1.9070223936924253, "learning_rate": 8.062203982574269e-06, "loss": 0.1436, "step": 5764 }, { "epoch": 0.8707143935961336, "grad_norm": 1.7552515858640392, "learning_rate": 8.061578791806659e-06, "loss": 0.124, "step": 5765 }, { "epoch": 0.8708654281830539, "grad_norm": 1.9797078168676183, "learning_rate": 8.060953524451972e-06, "loss": 0.1335, "step": 5766 }, { "epoch": 0.8710164627699744, "grad_norm": 2.343179467400081, "learning_rate": 8.060328180525848e-06, "loss": 0.1332, "step": 5767 }, { "epoch": 0.8711674973568947, "grad_norm": 1.8074688674851742, "learning_rate": 8.059702760043931e-06, "loss": 0.147, "step": 5768 }, { "epoch": 0.8713185319438151, "grad_norm": 3.6365394521408287, "learning_rate": 8.059077263021866e-06, "loss": 0.206, "step": 5769 }, { "epoch": 0.8714695665307355, "grad_norm": 2.438307515906323, "learning_rate": 8.058451689475302e-06, "loss": 0.1604, "step": 5770 }, { "epoch": 0.8716206011176559, "grad_norm": 2.3998569805516965, "learning_rate": 8.057826039419884e-06, "loss": 0.1384, "step": 5771 }, { "epoch": 0.8717716357045764, "grad_norm": 2.010237796388969, "learning_rate": 8.057200312871265e-06, "loss": 0.1531, "step": 5772 }, { "epoch": 0.8719226702914967, "grad_norm": 3.7524191933351156, "learning_rate": 8.0565745098451e-06, "loss": 0.1302, "step": 5773 }, { "epoch": 0.8720737048784172, "grad_norm": 1.897096016540436, "learning_rate": 8.05594863035704e-06, "loss": 0.156, "step": 5774 }, { "epoch": 0.8722247394653375, "grad_norm": 1.9697303808104414, "learning_rate": 8.055322674422744e-06, "loss": 0.1503, "step": 5775 }, { "epoch": 0.872375774052258, "grad_norm": 2.033790685904405, "learning_rate": 8.054696642057872e-06, "loss": 0.1387, "step": 5776 }, { "epoch": 0.8725268086391784, "grad_norm": 1.865391218521309, "learning_rate": 8.054070533278082e-06, "loss": 0.1491, "step": 5777 }, { "epoch": 0.8726778432260988, "grad_norm": 1.8712540406309015, "learning_rate": 8.053444348099037e-06, "loss": 0.1397, "step": 5778 }, { "epoch": 0.8728288778130192, "grad_norm": 2.556067346265433, "learning_rate": 8.0528180865364e-06, "loss": 0.1583, "step": 5779 }, { "epoch": 0.8729799123999396, "grad_norm": 1.9611184157538408, "learning_rate": 8.052191748605843e-06, "loss": 0.1406, "step": 5780 }, { "epoch": 0.87313094698686, "grad_norm": 1.9317946844023095, "learning_rate": 8.051565334323025e-06, "loss": 0.1363, "step": 5781 }, { "epoch": 0.8732819815737803, "grad_norm": 1.9498386614960614, "learning_rate": 8.050938843703625e-06, "loss": 0.1475, "step": 5782 }, { "epoch": 0.8734330161607008, "grad_norm": 3.0903373341171125, "learning_rate": 8.05031227676331e-06, "loss": 0.217, "step": 5783 }, { "epoch": 0.8735840507476212, "grad_norm": 2.287438524889415, "learning_rate": 8.049685633517755e-06, "loss": 0.146, "step": 5784 }, { "epoch": 0.8737350853345416, "grad_norm": 1.709805060827801, "learning_rate": 8.049058913982637e-06, "loss": 0.1255, "step": 5785 }, { "epoch": 0.873886119921462, "grad_norm": 2.1037896213321994, "learning_rate": 8.048432118173632e-06, "loss": 0.1378, "step": 5786 }, { "epoch": 0.8740371545083824, "grad_norm": 2.2149240860970205, "learning_rate": 8.047805246106419e-06, "loss": 0.1571, "step": 5787 }, { "epoch": 0.8741881890953028, "grad_norm": 1.990893851409491, "learning_rate": 8.047178297796682e-06, "loss": 0.1607, "step": 5788 }, { "epoch": 0.8743392236822233, "grad_norm": 1.884179401484276, "learning_rate": 8.046551273260104e-06, "loss": 0.1334, "step": 5789 }, { "epoch": 0.8744902582691436, "grad_norm": 1.858714339398235, "learning_rate": 8.045924172512369e-06, "loss": 0.1431, "step": 5790 }, { "epoch": 0.8746412928560641, "grad_norm": 1.7780617054777443, "learning_rate": 8.045296995569165e-06, "loss": 0.1175, "step": 5791 }, { "epoch": 0.8747923274429844, "grad_norm": 3.2615626078215896, "learning_rate": 8.04466974244618e-06, "loss": 0.1623, "step": 5792 }, { "epoch": 0.8749433620299049, "grad_norm": 1.7715571281205067, "learning_rate": 8.044042413159108e-06, "loss": 0.1234, "step": 5793 }, { "epoch": 0.8750943966168252, "grad_norm": 2.9562075156682766, "learning_rate": 8.043415007723639e-06, "loss": 0.1567, "step": 5794 }, { "epoch": 0.8752454312037456, "grad_norm": 1.649453748764671, "learning_rate": 8.042787526155469e-06, "loss": 0.1162, "step": 5795 }, { "epoch": 0.8753964657906661, "grad_norm": 2.0305534560163627, "learning_rate": 8.042159968470296e-06, "loss": 0.1338, "step": 5796 }, { "epoch": 0.8755475003775864, "grad_norm": 2.239499851905395, "learning_rate": 8.041532334683817e-06, "loss": 0.1556, "step": 5797 }, { "epoch": 0.8756985349645069, "grad_norm": 2.369362552774695, "learning_rate": 8.040904624811735e-06, "loss": 0.181, "step": 5798 }, { "epoch": 0.8758495695514272, "grad_norm": 2.178345184562108, "learning_rate": 8.040276838869748e-06, "loss": 0.1488, "step": 5799 }, { "epoch": 0.8760006041383477, "grad_norm": 1.7623007052365744, "learning_rate": 8.039648976873564e-06, "loss": 0.1268, "step": 5800 }, { "epoch": 0.8761516387252681, "grad_norm": 1.7856302512057405, "learning_rate": 8.03902103883889e-06, "loss": 0.1304, "step": 5801 }, { "epoch": 0.8763026733121885, "grad_norm": 2.273181549809605, "learning_rate": 8.03839302478143e-06, "loss": 0.1502, "step": 5802 }, { "epoch": 0.8764537078991089, "grad_norm": 1.9166791608567348, "learning_rate": 8.037764934716898e-06, "loss": 0.1202, "step": 5803 }, { "epoch": 0.8766047424860293, "grad_norm": 2.380169219853225, "learning_rate": 8.037136768661005e-06, "loss": 0.1516, "step": 5804 }, { "epoch": 0.8767557770729497, "grad_norm": 2.341519570243771, "learning_rate": 8.036508526629464e-06, "loss": 0.1369, "step": 5805 }, { "epoch": 0.8769068116598701, "grad_norm": 2.2136441826296887, "learning_rate": 8.035880208637992e-06, "loss": 0.1612, "step": 5806 }, { "epoch": 0.8770578462467905, "grad_norm": 2.093765207853851, "learning_rate": 8.035251814702308e-06, "loss": 0.1476, "step": 5807 }, { "epoch": 0.877208880833711, "grad_norm": 2.7501319017374786, "learning_rate": 8.034623344838126e-06, "loss": 0.1421, "step": 5808 }, { "epoch": 0.8773599154206313, "grad_norm": 1.8879636455255517, "learning_rate": 8.033994799061175e-06, "loss": 0.1436, "step": 5809 }, { "epoch": 0.8775109500075517, "grad_norm": 1.8858059224973218, "learning_rate": 8.033366177387172e-06, "loss": 0.1484, "step": 5810 }, { "epoch": 0.8776619845944721, "grad_norm": 2.0488215560931984, "learning_rate": 8.032737479831847e-06, "loss": 0.143, "step": 5811 }, { "epoch": 0.8778130191813925, "grad_norm": 2.4418159213657145, "learning_rate": 8.032108706410927e-06, "loss": 0.1398, "step": 5812 }, { "epoch": 0.877964053768313, "grad_norm": 1.83819176439509, "learning_rate": 8.031479857140134e-06, "loss": 0.1506, "step": 5813 }, { "epoch": 0.8781150883552333, "grad_norm": 2.6958417189722623, "learning_rate": 8.03085093203521e-06, "loss": 0.1493, "step": 5814 }, { "epoch": 0.8782661229421538, "grad_norm": 1.8046028272440606, "learning_rate": 8.03022193111188e-06, "loss": 0.1296, "step": 5815 }, { "epoch": 0.8784171575290741, "grad_norm": 2.229286211133249, "learning_rate": 8.029592854385881e-06, "loss": 0.1624, "step": 5816 }, { "epoch": 0.8785681921159946, "grad_norm": 2.4441703013238025, "learning_rate": 8.02896370187295e-06, "loss": 0.1408, "step": 5817 }, { "epoch": 0.8787192267029149, "grad_norm": 1.8991161208971017, "learning_rate": 8.028334473588825e-06, "loss": 0.1281, "step": 5818 }, { "epoch": 0.8788702612898354, "grad_norm": 1.8263789901000582, "learning_rate": 8.027705169549248e-06, "loss": 0.1305, "step": 5819 }, { "epoch": 0.8790212958767558, "grad_norm": 2.303972397791404, "learning_rate": 8.02707578976996e-06, "loss": 0.1577, "step": 5820 }, { "epoch": 0.8791723304636762, "grad_norm": 2.6911895965352115, "learning_rate": 8.026446334266707e-06, "loss": 0.1667, "step": 5821 }, { "epoch": 0.8793233650505966, "grad_norm": 1.916600095466764, "learning_rate": 8.025816803055231e-06, "loss": 0.1567, "step": 5822 }, { "epoch": 0.879474399637517, "grad_norm": 1.9155999884468784, "learning_rate": 8.025187196151286e-06, "loss": 0.1463, "step": 5823 }, { "epoch": 0.8796254342244374, "grad_norm": 2.055042329214425, "learning_rate": 8.024557513570616e-06, "loss": 0.1552, "step": 5824 }, { "epoch": 0.8797764688113578, "grad_norm": 1.8844966006837263, "learning_rate": 8.023927755328976e-06, "loss": 0.1412, "step": 5825 }, { "epoch": 0.8799275033982782, "grad_norm": 2.3082962727607206, "learning_rate": 8.023297921442122e-06, "loss": 0.1809, "step": 5826 }, { "epoch": 0.8800785379851986, "grad_norm": 2.2201312036450638, "learning_rate": 8.022668011925803e-06, "loss": 0.1304, "step": 5827 }, { "epoch": 0.880229572572119, "grad_norm": 1.6204278411863413, "learning_rate": 8.022038026795783e-06, "loss": 0.143, "step": 5828 }, { "epoch": 0.8803806071590394, "grad_norm": 2.0661047222173523, "learning_rate": 8.02140796606782e-06, "loss": 0.1429, "step": 5829 }, { "epoch": 0.8805316417459598, "grad_norm": 1.8020930062050085, "learning_rate": 8.020777829757673e-06, "loss": 0.1313, "step": 5830 }, { "epoch": 0.8806826763328802, "grad_norm": 1.9009847293759583, "learning_rate": 8.020147617881107e-06, "loss": 0.1289, "step": 5831 }, { "epoch": 0.8808337109198007, "grad_norm": 1.9247795108924843, "learning_rate": 8.019517330453887e-06, "loss": 0.1429, "step": 5832 }, { "epoch": 0.880984745506721, "grad_norm": 3.3084341157954613, "learning_rate": 8.018886967491778e-06, "loss": 0.1705, "step": 5833 }, { "epoch": 0.8811357800936415, "grad_norm": 2.1898880751166994, "learning_rate": 8.018256529010553e-06, "loss": 0.1631, "step": 5834 }, { "epoch": 0.8812868146805618, "grad_norm": 1.8339677404417936, "learning_rate": 8.017626015025979e-06, "loss": 0.1364, "step": 5835 }, { "epoch": 0.8814378492674823, "grad_norm": 1.810311020286865, "learning_rate": 8.016995425553831e-06, "loss": 0.1416, "step": 5836 }, { "epoch": 0.8815888838544027, "grad_norm": 2.1340883714795917, "learning_rate": 8.01636476060988e-06, "loss": 0.1543, "step": 5837 }, { "epoch": 0.881739918441323, "grad_norm": 1.662987472941924, "learning_rate": 8.015734020209906e-06, "loss": 0.1469, "step": 5838 }, { "epoch": 0.8818909530282435, "grad_norm": 1.9592439836567543, "learning_rate": 8.015103204369687e-06, "loss": 0.134, "step": 5839 }, { "epoch": 0.8820419876151638, "grad_norm": 1.8701084423162058, "learning_rate": 8.014472313105003e-06, "loss": 0.1506, "step": 5840 }, { "epoch": 0.8821930222020843, "grad_norm": 1.9752725738681538, "learning_rate": 8.013841346431634e-06, "loss": 0.1422, "step": 5841 }, { "epoch": 0.8823440567890046, "grad_norm": 1.8780695842768653, "learning_rate": 8.013210304365366e-06, "loss": 0.1399, "step": 5842 }, { "epoch": 0.8824950913759251, "grad_norm": 2.5776878683449396, "learning_rate": 8.012579186921985e-06, "loss": 0.1751, "step": 5843 }, { "epoch": 0.8826461259628455, "grad_norm": 1.900856130990448, "learning_rate": 8.011947994117277e-06, "loss": 0.1392, "step": 5844 }, { "epoch": 0.8827971605497659, "grad_norm": 1.6601706792819924, "learning_rate": 8.011316725967034e-06, "loss": 0.1282, "step": 5845 }, { "epoch": 0.8829481951366863, "grad_norm": 2.5955282184077046, "learning_rate": 8.010685382487046e-06, "loss": 0.1432, "step": 5846 }, { "epoch": 0.8830992297236067, "grad_norm": 2.1979381353213103, "learning_rate": 8.010053963693106e-06, "loss": 0.1562, "step": 5847 }, { "epoch": 0.8832502643105271, "grad_norm": 2.0433764855775305, "learning_rate": 8.00942246960101e-06, "loss": 0.1422, "step": 5848 }, { "epoch": 0.8834012988974476, "grad_norm": 1.6375713654777049, "learning_rate": 8.008790900226555e-06, "loss": 0.1447, "step": 5849 }, { "epoch": 0.8835523334843679, "grad_norm": 1.6603779412940847, "learning_rate": 8.008159255585541e-06, "loss": 0.1235, "step": 5850 }, { "epoch": 0.8837033680712884, "grad_norm": 1.6624826704569382, "learning_rate": 8.00752753569377e-06, "loss": 0.1284, "step": 5851 }, { "epoch": 0.8838544026582087, "grad_norm": 1.8424018243272737, "learning_rate": 8.00689574056704e-06, "loss": 0.1506, "step": 5852 }, { "epoch": 0.8840054372451291, "grad_norm": 1.7092801717591788, "learning_rate": 8.006263870221159e-06, "loss": 0.1364, "step": 5853 }, { "epoch": 0.8841564718320495, "grad_norm": 1.8968087460343506, "learning_rate": 8.005631924671935e-06, "loss": 0.1232, "step": 5854 }, { "epoch": 0.8843075064189699, "grad_norm": 2.211976152629292, "learning_rate": 8.004999903935172e-06, "loss": 0.1379, "step": 5855 }, { "epoch": 0.8844585410058904, "grad_norm": 1.7897938460763967, "learning_rate": 8.004367808026684e-06, "loss": 0.1397, "step": 5856 }, { "epoch": 0.8846095755928107, "grad_norm": 1.9512718282968098, "learning_rate": 8.003735636962285e-06, "loss": 0.1377, "step": 5857 }, { "epoch": 0.8847606101797312, "grad_norm": 1.8251250863713353, "learning_rate": 8.003103390757782e-06, "loss": 0.1396, "step": 5858 }, { "epoch": 0.8849116447666515, "grad_norm": 1.5109862897213895, "learning_rate": 8.002471069429e-06, "loss": 0.1178, "step": 5859 }, { "epoch": 0.885062679353572, "grad_norm": 1.9232703814488619, "learning_rate": 8.001838672991747e-06, "loss": 0.1287, "step": 5860 }, { "epoch": 0.8852137139404924, "grad_norm": 1.7160271920431345, "learning_rate": 8.001206201461852e-06, "loss": 0.148, "step": 5861 }, { "epoch": 0.8853647485274128, "grad_norm": 2.139034296595588, "learning_rate": 8.000573654855131e-06, "loss": 0.1227, "step": 5862 }, { "epoch": 0.8855157831143332, "grad_norm": 1.9660783570120837, "learning_rate": 7.99994103318741e-06, "loss": 0.1347, "step": 5863 }, { "epoch": 0.8856668177012536, "grad_norm": 2.102828013068391, "learning_rate": 7.999308336474512e-06, "loss": 0.1553, "step": 5864 }, { "epoch": 0.885817852288174, "grad_norm": 2.45968814298277, "learning_rate": 7.998675564732269e-06, "loss": 0.1554, "step": 5865 }, { "epoch": 0.8859688868750943, "grad_norm": 2.055974634031816, "learning_rate": 7.998042717976504e-06, "loss": 0.1316, "step": 5866 }, { "epoch": 0.8861199214620148, "grad_norm": 1.7876814419965896, "learning_rate": 7.99740979622305e-06, "loss": 0.1502, "step": 5867 }, { "epoch": 0.8862709560489352, "grad_norm": 1.7869944325126452, "learning_rate": 7.996776799487745e-06, "loss": 0.1317, "step": 5868 }, { "epoch": 0.8864219906358556, "grad_norm": 1.9576837638567586, "learning_rate": 7.996143727786414e-06, "loss": 0.1343, "step": 5869 }, { "epoch": 0.886573025222776, "grad_norm": 2.009664680376608, "learning_rate": 7.995510581134903e-06, "loss": 0.1362, "step": 5870 }, { "epoch": 0.8867240598096964, "grad_norm": 3.658583686068294, "learning_rate": 7.994877359549046e-06, "loss": 0.1429, "step": 5871 }, { "epoch": 0.8868750943966168, "grad_norm": 2.346792713386527, "learning_rate": 7.994244063044683e-06, "loss": 0.1504, "step": 5872 }, { "epoch": 0.8870261289835373, "grad_norm": 1.6174779498842258, "learning_rate": 7.993610691637658e-06, "loss": 0.1188, "step": 5873 }, { "epoch": 0.8871771635704576, "grad_norm": 2.3274535231914695, "learning_rate": 7.992977245343816e-06, "loss": 0.1683, "step": 5874 }, { "epoch": 0.8873281981573781, "grad_norm": 2.502301983596035, "learning_rate": 7.992343724179e-06, "loss": 0.1706, "step": 5875 }, { "epoch": 0.8874792327442984, "grad_norm": 1.8909866835113216, "learning_rate": 7.99171012815906e-06, "loss": 0.1489, "step": 5876 }, { "epoch": 0.8876302673312189, "grad_norm": 2.539467352585576, "learning_rate": 7.991076457299844e-06, "loss": 0.1567, "step": 5877 }, { "epoch": 0.8877813019181392, "grad_norm": 2.8750286105936085, "learning_rate": 7.990442711617206e-06, "loss": 0.157, "step": 5878 }, { "epoch": 0.8879323365050596, "grad_norm": 2.579792371671565, "learning_rate": 7.989808891126999e-06, "loss": 0.1546, "step": 5879 }, { "epoch": 0.8880833710919801, "grad_norm": 1.943579356760257, "learning_rate": 7.989174995845074e-06, "loss": 0.1545, "step": 5880 }, { "epoch": 0.8882344056789004, "grad_norm": 2.245104104012461, "learning_rate": 7.988541025787295e-06, "loss": 0.1424, "step": 5881 }, { "epoch": 0.8883854402658209, "grad_norm": 2.0591794383114816, "learning_rate": 7.987906980969515e-06, "loss": 0.1513, "step": 5882 }, { "epoch": 0.8885364748527412, "grad_norm": 3.3376584756315255, "learning_rate": 7.987272861407602e-06, "loss": 0.1518, "step": 5883 }, { "epoch": 0.8886875094396617, "grad_norm": 1.988691578797846, "learning_rate": 7.986638667117413e-06, "loss": 0.1642, "step": 5884 }, { "epoch": 0.8888385440265821, "grad_norm": 2.0463670354984616, "learning_rate": 7.986004398114813e-06, "loss": 0.1609, "step": 5885 }, { "epoch": 0.8889895786135025, "grad_norm": 1.721463220721476, "learning_rate": 7.985370054415671e-06, "loss": 0.1266, "step": 5886 }, { "epoch": 0.8891406132004229, "grad_norm": 1.7028341370924431, "learning_rate": 7.984735636035855e-06, "loss": 0.1305, "step": 5887 }, { "epoch": 0.8892916477873433, "grad_norm": 5.407479508343774, "learning_rate": 7.984101142991235e-06, "loss": 0.1763, "step": 5888 }, { "epoch": 0.8894426823742637, "grad_norm": 2.0411268363225803, "learning_rate": 7.983466575297682e-06, "loss": 0.1608, "step": 5889 }, { "epoch": 0.8895937169611842, "grad_norm": 2.4674523204507444, "learning_rate": 7.982831932971073e-06, "loss": 0.1481, "step": 5890 }, { "epoch": 0.8897447515481045, "grad_norm": 1.491021360842656, "learning_rate": 7.982197216027281e-06, "loss": 0.1121, "step": 5891 }, { "epoch": 0.889895786135025, "grad_norm": 2.5533133674375605, "learning_rate": 7.981562424482185e-06, "loss": 0.1463, "step": 5892 }, { "epoch": 0.8900468207219453, "grad_norm": 1.827833085227739, "learning_rate": 7.980927558351663e-06, "loss": 0.1372, "step": 5893 }, { "epoch": 0.8901978553088657, "grad_norm": 2.1926314165249776, "learning_rate": 7.980292617651602e-06, "loss": 0.1651, "step": 5894 }, { "epoch": 0.8903488898957861, "grad_norm": 1.835623935719081, "learning_rate": 7.979657602397877e-06, "loss": 0.1587, "step": 5895 }, { "epoch": 0.8904999244827065, "grad_norm": 2.010663875022064, "learning_rate": 7.979022512606379e-06, "loss": 0.153, "step": 5896 }, { "epoch": 0.890650959069627, "grad_norm": 2.3168133351403064, "learning_rate": 7.978387348292995e-06, "loss": 0.1513, "step": 5897 }, { "epoch": 0.8908019936565473, "grad_norm": 2.0239204906959407, "learning_rate": 7.977752109473612e-06, "loss": 0.1465, "step": 5898 }, { "epoch": 0.8909530282434678, "grad_norm": 1.9275195152348314, "learning_rate": 7.977116796164122e-06, "loss": 0.1597, "step": 5899 }, { "epoch": 0.8911040628303881, "grad_norm": 2.3302923302626577, "learning_rate": 7.976481408380417e-06, "loss": 0.1319, "step": 5900 }, { "epoch": 0.8912550974173086, "grad_norm": 1.5677608904748674, "learning_rate": 7.975845946138391e-06, "loss": 0.118, "step": 5901 }, { "epoch": 0.891406132004229, "grad_norm": 1.9418905339104962, "learning_rate": 7.975210409453943e-06, "loss": 0.1436, "step": 5902 }, { "epoch": 0.8915571665911494, "grad_norm": 1.6129376805504179, "learning_rate": 7.97457479834297e-06, "loss": 0.1093, "step": 5903 }, { "epoch": 0.8917082011780698, "grad_norm": 2.1441796151315113, "learning_rate": 7.97393911282137e-06, "loss": 0.1375, "step": 5904 }, { "epoch": 0.8918592357649902, "grad_norm": 1.8250166824790786, "learning_rate": 7.973303352905048e-06, "loss": 0.1258, "step": 5905 }, { "epoch": 0.8920102703519106, "grad_norm": 2.067192937377114, "learning_rate": 7.972667518609905e-06, "loss": 0.1463, "step": 5906 }, { "epoch": 0.892161304938831, "grad_norm": 1.7374690239352082, "learning_rate": 7.97203160995185e-06, "loss": 0.1245, "step": 5907 }, { "epoch": 0.8923123395257514, "grad_norm": 2.038602571031635, "learning_rate": 7.97139562694679e-06, "loss": 0.1435, "step": 5908 }, { "epoch": 0.8924633741126718, "grad_norm": 1.7147211936612323, "learning_rate": 7.97075956961063e-06, "loss": 0.1284, "step": 5909 }, { "epoch": 0.8926144086995922, "grad_norm": 1.996941400791868, "learning_rate": 7.97012343795929e-06, "loss": 0.1376, "step": 5910 }, { "epoch": 0.8927654432865126, "grad_norm": 2.066509264571151, "learning_rate": 7.969487232008675e-06, "loss": 0.1557, "step": 5911 }, { "epoch": 0.892916477873433, "grad_norm": 2.5201711109870963, "learning_rate": 7.968850951774703e-06, "loss": 0.1627, "step": 5912 }, { "epoch": 0.8930675124603534, "grad_norm": 1.7212160747816718, "learning_rate": 7.968214597273291e-06, "loss": 0.1414, "step": 5913 }, { "epoch": 0.8932185470472739, "grad_norm": 1.8730986410755097, "learning_rate": 7.967578168520359e-06, "loss": 0.1479, "step": 5914 }, { "epoch": 0.8933695816341942, "grad_norm": 2.6107513899495616, "learning_rate": 7.966941665531825e-06, "loss": 0.167, "step": 5915 }, { "epoch": 0.8935206162211147, "grad_norm": 2.0114743597615354, "learning_rate": 7.966305088323614e-06, "loss": 0.134, "step": 5916 }, { "epoch": 0.893671650808035, "grad_norm": 2.1138953724897904, "learning_rate": 7.965668436911647e-06, "loss": 0.137, "step": 5917 }, { "epoch": 0.8938226853949555, "grad_norm": 2.5194891605790253, "learning_rate": 7.965031711311854e-06, "loss": 0.1357, "step": 5918 }, { "epoch": 0.8939737199818758, "grad_norm": 1.8025379224101459, "learning_rate": 7.964394911540162e-06, "loss": 0.1394, "step": 5919 }, { "epoch": 0.8941247545687963, "grad_norm": 2.133801665760954, "learning_rate": 7.9637580376125e-06, "loss": 0.1344, "step": 5920 }, { "epoch": 0.8942757891557167, "grad_norm": 1.7836059029028029, "learning_rate": 7.963121089544799e-06, "loss": 0.1403, "step": 5921 }, { "epoch": 0.894426823742637, "grad_norm": 1.692484518725735, "learning_rate": 7.962484067352995e-06, "loss": 0.1326, "step": 5922 }, { "epoch": 0.8945778583295575, "grad_norm": 3.258190459020176, "learning_rate": 7.961846971053019e-06, "loss": 0.1673, "step": 5923 }, { "epoch": 0.8947288929164778, "grad_norm": 1.816508892158131, "learning_rate": 7.961209800660815e-06, "loss": 0.1232, "step": 5924 }, { "epoch": 0.8948799275033983, "grad_norm": 1.9152145774437974, "learning_rate": 7.960572556192318e-06, "loss": 0.1448, "step": 5925 }, { "epoch": 0.8950309620903187, "grad_norm": 1.7601421850558412, "learning_rate": 7.959935237663468e-06, "loss": 0.1358, "step": 5926 }, { "epoch": 0.8951819966772391, "grad_norm": 1.7123269587295649, "learning_rate": 7.959297845090211e-06, "loss": 0.1289, "step": 5927 }, { "epoch": 0.8953330312641595, "grad_norm": 1.884444333227132, "learning_rate": 7.95866037848849e-06, "loss": 0.1463, "step": 5928 }, { "epoch": 0.8954840658510799, "grad_norm": 1.9362928374180186, "learning_rate": 7.95802283787425e-06, "loss": 0.1437, "step": 5929 }, { "epoch": 0.8956351004380003, "grad_norm": 2.911857450248201, "learning_rate": 7.957385223263445e-06, "loss": 0.1835, "step": 5930 }, { "epoch": 0.8957861350249207, "grad_norm": 2.1691284532473625, "learning_rate": 7.956747534672018e-06, "loss": 0.1414, "step": 5931 }, { "epoch": 0.8959371696118411, "grad_norm": 2.1303849576782605, "learning_rate": 7.956109772115925e-06, "loss": 0.1378, "step": 5932 }, { "epoch": 0.8960882041987616, "grad_norm": 2.3637670567535727, "learning_rate": 7.95547193561112e-06, "loss": 0.157, "step": 5933 }, { "epoch": 0.8962392387856819, "grad_norm": 2.3653883549677315, "learning_rate": 7.954834025173562e-06, "loss": 0.1574, "step": 5934 }, { "epoch": 0.8963902733726024, "grad_norm": 2.364946737709193, "learning_rate": 7.9541960408192e-06, "loss": 0.1565, "step": 5935 }, { "epoch": 0.8965413079595227, "grad_norm": 2.4450692843084236, "learning_rate": 7.953557982564e-06, "loss": 0.1734, "step": 5936 }, { "epoch": 0.8966923425464431, "grad_norm": 2.425237317822405, "learning_rate": 7.952919850423923e-06, "loss": 0.1603, "step": 5937 }, { "epoch": 0.8968433771333636, "grad_norm": 2.1708212182534483, "learning_rate": 7.952281644414931e-06, "loss": 0.1653, "step": 5938 }, { "epoch": 0.8969944117202839, "grad_norm": 2.0153868232708327, "learning_rate": 7.951643364552988e-06, "loss": 0.1323, "step": 5939 }, { "epoch": 0.8971454463072044, "grad_norm": 2.1812592942533913, "learning_rate": 7.951005010854064e-06, "loss": 0.1541, "step": 5940 }, { "epoch": 0.8972964808941247, "grad_norm": 1.8966928137952008, "learning_rate": 7.950366583334126e-06, "loss": 0.1395, "step": 5941 }, { "epoch": 0.8974475154810452, "grad_norm": 1.7473478795755277, "learning_rate": 7.949728082009145e-06, "loss": 0.1338, "step": 5942 }, { "epoch": 0.8975985500679655, "grad_norm": 1.901304860077575, "learning_rate": 7.949089506895092e-06, "loss": 0.1611, "step": 5943 }, { "epoch": 0.897749584654886, "grad_norm": 2.4505286275187275, "learning_rate": 7.948450858007943e-06, "loss": 0.1367, "step": 5944 }, { "epoch": 0.8979006192418064, "grad_norm": 3.863303506875271, "learning_rate": 7.947812135363673e-06, "loss": 0.1721, "step": 5945 }, { "epoch": 0.8980516538287268, "grad_norm": 2.5136353947825185, "learning_rate": 7.94717333897826e-06, "loss": 0.1395, "step": 5946 }, { "epoch": 0.8982026884156472, "grad_norm": 1.5818070069034902, "learning_rate": 7.946534468867685e-06, "loss": 0.1339, "step": 5947 }, { "epoch": 0.8983537230025676, "grad_norm": 2.417388623354156, "learning_rate": 7.94589552504793e-06, "loss": 0.143, "step": 5948 }, { "epoch": 0.898504757589488, "grad_norm": 1.904453281272548, "learning_rate": 7.945256507534976e-06, "loss": 0.1418, "step": 5949 }, { "epoch": 0.8986557921764085, "grad_norm": 2.1372927905824772, "learning_rate": 7.944617416344812e-06, "loss": 0.1593, "step": 5950 }, { "epoch": 0.8988068267633288, "grad_norm": 2.1913173597078193, "learning_rate": 7.94397825149342e-06, "loss": 0.1513, "step": 5951 }, { "epoch": 0.8989578613502492, "grad_norm": 3.0987557700285726, "learning_rate": 7.943339012996796e-06, "loss": 0.1708, "step": 5952 }, { "epoch": 0.8991088959371696, "grad_norm": 2.265933072095879, "learning_rate": 7.942699700870926e-06, "loss": 0.1648, "step": 5953 }, { "epoch": 0.89925993052409, "grad_norm": 2.0918092866704945, "learning_rate": 7.942060315131802e-06, "loss": 0.1424, "step": 5954 }, { "epoch": 0.8994109651110104, "grad_norm": 9.07662587137702, "learning_rate": 7.941420855795423e-06, "loss": 0.2237, "step": 5955 }, { "epoch": 0.8995619996979308, "grad_norm": 1.9885794119249194, "learning_rate": 7.940781322877782e-06, "loss": 0.1484, "step": 5956 }, { "epoch": 0.8997130342848513, "grad_norm": 1.8104354906914468, "learning_rate": 7.940141716394877e-06, "loss": 0.1137, "step": 5957 }, { "epoch": 0.8998640688717716, "grad_norm": 1.9152601565253125, "learning_rate": 7.93950203636271e-06, "loss": 0.1314, "step": 5958 }, { "epoch": 0.9000151034586921, "grad_norm": 1.6540232207805265, "learning_rate": 7.938862282797285e-06, "loss": 0.1207, "step": 5959 }, { "epoch": 0.9001661380456124, "grad_norm": 2.1762291855263793, "learning_rate": 7.938222455714598e-06, "loss": 0.1599, "step": 5960 }, { "epoch": 0.9003171726325329, "grad_norm": 3.25677585521181, "learning_rate": 7.937582555130663e-06, "loss": 0.1563, "step": 5961 }, { "epoch": 0.9004682072194533, "grad_norm": 1.7824211060038901, "learning_rate": 7.936942581061485e-06, "loss": 0.1384, "step": 5962 }, { "epoch": 0.9006192418063736, "grad_norm": 2.5565521980538937, "learning_rate": 7.93630253352307e-06, "loss": 0.1374, "step": 5963 }, { "epoch": 0.9007702763932941, "grad_norm": 2.165429440038684, "learning_rate": 7.935662412531432e-06, "loss": 0.1581, "step": 5964 }, { "epoch": 0.9009213109802144, "grad_norm": 1.5660356310997354, "learning_rate": 7.935022218102584e-06, "loss": 0.1241, "step": 5965 }, { "epoch": 0.9010723455671349, "grad_norm": 2.2175981715283233, "learning_rate": 7.934381950252541e-06, "loss": 0.1469, "step": 5966 }, { "epoch": 0.9012233801540552, "grad_norm": 2.4698073314583224, "learning_rate": 7.933741608997318e-06, "loss": 0.1508, "step": 5967 }, { "epoch": 0.9013744147409757, "grad_norm": 1.8258359866773233, "learning_rate": 7.933101194352934e-06, "loss": 0.123, "step": 5968 }, { "epoch": 0.9015254493278961, "grad_norm": 2.208606685228427, "learning_rate": 7.932460706335413e-06, "loss": 0.1401, "step": 5969 }, { "epoch": 0.9016764839148165, "grad_norm": 2.2422996979822107, "learning_rate": 7.93182014496077e-06, "loss": 0.1434, "step": 5970 }, { "epoch": 0.9018275185017369, "grad_norm": 1.8465393039708633, "learning_rate": 7.931179510245035e-06, "loss": 0.1454, "step": 5971 }, { "epoch": 0.9019785530886573, "grad_norm": 2.6693501081970292, "learning_rate": 7.93053880220423e-06, "loss": 0.1339, "step": 5972 }, { "epoch": 0.9021295876755777, "grad_norm": 2.2832086562134357, "learning_rate": 7.929898020854388e-06, "loss": 0.1269, "step": 5973 }, { "epoch": 0.9022806222624982, "grad_norm": 2.1090057412184926, "learning_rate": 7.929257166211533e-06, "loss": 0.1483, "step": 5974 }, { "epoch": 0.9024316568494185, "grad_norm": 1.5620064963991134, "learning_rate": 7.928616238291697e-06, "loss": 0.1188, "step": 5975 }, { "epoch": 0.902582691436339, "grad_norm": 1.8413947380914912, "learning_rate": 7.927975237110918e-06, "loss": 0.132, "step": 5976 }, { "epoch": 0.9027337260232593, "grad_norm": 2.1101535484982734, "learning_rate": 7.927334162685225e-06, "loss": 0.1627, "step": 5977 }, { "epoch": 0.9028847606101797, "grad_norm": 2.3171272633215305, "learning_rate": 7.926693015030658e-06, "loss": 0.1503, "step": 5978 }, { "epoch": 0.9030357951971001, "grad_norm": 1.6395644972458225, "learning_rate": 7.926051794163254e-06, "loss": 0.1322, "step": 5979 }, { "epoch": 0.9031868297840205, "grad_norm": 2.282506571636089, "learning_rate": 7.925410500099056e-06, "loss": 0.1277, "step": 5980 }, { "epoch": 0.903337864370941, "grad_norm": 2.1947961231581314, "learning_rate": 7.924769132854102e-06, "loss": 0.1583, "step": 5981 }, { "epoch": 0.9034888989578613, "grad_norm": 1.6243020261377772, "learning_rate": 7.924127692444442e-06, "loss": 0.1299, "step": 5982 }, { "epoch": 0.9036399335447818, "grad_norm": 1.9623398121380753, "learning_rate": 7.923486178886118e-06, "loss": 0.1356, "step": 5983 }, { "epoch": 0.9037909681317021, "grad_norm": 2.172845969242011, "learning_rate": 7.922844592195177e-06, "loss": 0.1372, "step": 5984 }, { "epoch": 0.9039420027186226, "grad_norm": 2.022663642508013, "learning_rate": 7.922202932387672e-06, "loss": 0.1373, "step": 5985 }, { "epoch": 0.904093037305543, "grad_norm": 1.701775748512791, "learning_rate": 7.921561199479653e-06, "loss": 0.1292, "step": 5986 }, { "epoch": 0.9042440718924634, "grad_norm": 2.060318536326836, "learning_rate": 7.920919393487175e-06, "loss": 0.1444, "step": 5987 }, { "epoch": 0.9043951064793838, "grad_norm": 1.745568337171143, "learning_rate": 7.920277514426289e-06, "loss": 0.1184, "step": 5988 }, { "epoch": 0.9045461410663042, "grad_norm": 3.276017035019672, "learning_rate": 7.919635562313053e-06, "loss": 0.1434, "step": 5989 }, { "epoch": 0.9046971756532246, "grad_norm": 2.9579638596303113, "learning_rate": 7.918993537163529e-06, "loss": 0.1556, "step": 5990 }, { "epoch": 0.904848210240145, "grad_norm": 2.217230212716455, "learning_rate": 7.918351438993774e-06, "loss": 0.1473, "step": 5991 }, { "epoch": 0.9049992448270654, "grad_norm": 2.2072039318221846, "learning_rate": 7.917709267819854e-06, "loss": 0.1379, "step": 5992 }, { "epoch": 0.9051502794139858, "grad_norm": 2.5635266621414257, "learning_rate": 7.91706702365783e-06, "loss": 0.1582, "step": 5993 }, { "epoch": 0.9053013140009062, "grad_norm": 2.032182841052765, "learning_rate": 7.916424706523771e-06, "loss": 0.1379, "step": 5994 }, { "epoch": 0.9054523485878266, "grad_norm": 1.8860436949837092, "learning_rate": 7.915782316433742e-06, "loss": 0.1374, "step": 5995 }, { "epoch": 0.905603383174747, "grad_norm": 1.791715950184525, "learning_rate": 7.915139853403816e-06, "loss": 0.1247, "step": 5996 }, { "epoch": 0.9057544177616674, "grad_norm": 2.5970421745004986, "learning_rate": 7.91449731745006e-06, "loss": 0.1219, "step": 5997 }, { "epoch": 0.9059054523485879, "grad_norm": 1.6185415047437113, "learning_rate": 7.913854708588551e-06, "loss": 0.1072, "step": 5998 }, { "epoch": 0.9060564869355082, "grad_norm": 1.8667069084969685, "learning_rate": 7.913212026835364e-06, "loss": 0.1302, "step": 5999 }, { "epoch": 0.9062075215224287, "grad_norm": 2.484463537638858, "learning_rate": 7.912569272206573e-06, "loss": 0.1471, "step": 6000 }, { "epoch": 0.906358556109349, "grad_norm": 2.6363283632244348, "learning_rate": 7.911926444718261e-06, "loss": 0.1585, "step": 6001 }, { "epoch": 0.9065095906962695, "grad_norm": 2.1087619967372753, "learning_rate": 7.911283544386508e-06, "loss": 0.1439, "step": 6002 }, { "epoch": 0.9066606252831898, "grad_norm": 1.8157263556572536, "learning_rate": 7.910640571227394e-06, "loss": 0.1115, "step": 6003 }, { "epoch": 0.9068116598701103, "grad_norm": 2.2236892076264407, "learning_rate": 7.909997525257005e-06, "loss": 0.1633, "step": 6004 }, { "epoch": 0.9069626944570307, "grad_norm": 1.7777284756415688, "learning_rate": 7.909354406491428e-06, "loss": 0.1355, "step": 6005 }, { "epoch": 0.907113729043951, "grad_norm": 2.5237774572436438, "learning_rate": 7.908711214946748e-06, "loss": 0.1619, "step": 6006 }, { "epoch": 0.9072647636308715, "grad_norm": 2.132501423065914, "learning_rate": 7.908067950639056e-06, "loss": 0.1596, "step": 6007 }, { "epoch": 0.9074157982177918, "grad_norm": 2.277305061949232, "learning_rate": 7.907424613584445e-06, "loss": 0.1414, "step": 6008 }, { "epoch": 0.9075668328047123, "grad_norm": 1.7653850439633172, "learning_rate": 7.906781203799007e-06, "loss": 0.1224, "step": 6009 }, { "epoch": 0.9077178673916327, "grad_norm": 2.0605883832722447, "learning_rate": 7.906137721298841e-06, "loss": 0.1596, "step": 6010 }, { "epoch": 0.9078689019785531, "grad_norm": 1.9455944020997453, "learning_rate": 7.905494166100038e-06, "loss": 0.1292, "step": 6011 }, { "epoch": 0.9080199365654735, "grad_norm": 2.022409025895877, "learning_rate": 7.9048505382187e-06, "loss": 0.147, "step": 6012 }, { "epoch": 0.9081709711523939, "grad_norm": 1.980190459706239, "learning_rate": 7.904206837670929e-06, "loss": 0.1227, "step": 6013 }, { "epoch": 0.9083220057393143, "grad_norm": 1.9081338384615805, "learning_rate": 7.903563064472825e-06, "loss": 0.1275, "step": 6014 }, { "epoch": 0.9084730403262347, "grad_norm": 1.8664443638433765, "learning_rate": 7.902919218640492e-06, "loss": 0.1371, "step": 6015 }, { "epoch": 0.9086240749131551, "grad_norm": 2.4692621356702844, "learning_rate": 7.90227530019004e-06, "loss": 0.159, "step": 6016 }, { "epoch": 0.9087751095000756, "grad_norm": 1.4473536881080633, "learning_rate": 7.901631309137573e-06, "loss": 0.1092, "step": 6017 }, { "epoch": 0.9089261440869959, "grad_norm": 1.8915432328303028, "learning_rate": 7.900987245499204e-06, "loss": 0.1519, "step": 6018 }, { "epoch": 0.9090771786739164, "grad_norm": 2.4789923394486504, "learning_rate": 7.900343109291042e-06, "loss": 0.1476, "step": 6019 }, { "epoch": 0.9092282132608367, "grad_norm": 2.3995576583143072, "learning_rate": 7.899698900529203e-06, "loss": 0.1518, "step": 6020 }, { "epoch": 0.9093792478477571, "grad_norm": 2.2556849758388506, "learning_rate": 7.899054619229798e-06, "loss": 0.176, "step": 6021 }, { "epoch": 0.9095302824346776, "grad_norm": 1.7562965668172188, "learning_rate": 7.898410265408947e-06, "loss": 0.1283, "step": 6022 }, { "epoch": 0.9096813170215979, "grad_norm": 1.810057929118145, "learning_rate": 7.897765839082769e-06, "loss": 0.1332, "step": 6023 }, { "epoch": 0.9098323516085184, "grad_norm": 2.551052588623302, "learning_rate": 7.897121340267386e-06, "loss": 0.1434, "step": 6024 }, { "epoch": 0.9099833861954387, "grad_norm": 1.6395365325730507, "learning_rate": 7.896476768978916e-06, "loss": 0.1294, "step": 6025 }, { "epoch": 0.9101344207823592, "grad_norm": 2.3918038175708896, "learning_rate": 7.895832125233487e-06, "loss": 0.1534, "step": 6026 }, { "epoch": 0.9102854553692795, "grad_norm": 1.6663528020851595, "learning_rate": 7.895187409047224e-06, "loss": 0.1269, "step": 6027 }, { "epoch": 0.9104364899562, "grad_norm": 1.8842851049134908, "learning_rate": 7.894542620436255e-06, "loss": 0.1507, "step": 6028 }, { "epoch": 0.9105875245431204, "grad_norm": 1.9386302008696887, "learning_rate": 7.89389775941671e-06, "loss": 0.1392, "step": 6029 }, { "epoch": 0.9107385591300408, "grad_norm": 2.210470170772007, "learning_rate": 7.89325282600472e-06, "loss": 0.1379, "step": 6030 }, { "epoch": 0.9108895937169612, "grad_norm": 2.212856198834704, "learning_rate": 7.89260782021642e-06, "loss": 0.1507, "step": 6031 }, { "epoch": 0.9110406283038816, "grad_norm": 1.9157890166139746, "learning_rate": 7.891962742067943e-06, "loss": 0.132, "step": 6032 }, { "epoch": 0.911191662890802, "grad_norm": 1.6417898640123374, "learning_rate": 7.891317591575425e-06, "loss": 0.1183, "step": 6033 }, { "epoch": 0.9113426974777225, "grad_norm": 1.8381920995214787, "learning_rate": 7.89067236875501e-06, "loss": 0.1289, "step": 6034 }, { "epoch": 0.9114937320646428, "grad_norm": 2.319428954389125, "learning_rate": 7.890027073622833e-06, "loss": 0.1601, "step": 6035 }, { "epoch": 0.9116447666515632, "grad_norm": 3.041459123591642, "learning_rate": 7.88938170619504e-06, "loss": 0.1629, "step": 6036 }, { "epoch": 0.9117958012384836, "grad_norm": 1.7128084784493331, "learning_rate": 7.888736266487773e-06, "loss": 0.1404, "step": 6037 }, { "epoch": 0.911946835825404, "grad_norm": 2.3184406949434253, "learning_rate": 7.888090754517179e-06, "loss": 0.1449, "step": 6038 }, { "epoch": 0.9120978704123244, "grad_norm": 2.8827337832054676, "learning_rate": 7.887445170299406e-06, "loss": 0.1379, "step": 6039 }, { "epoch": 0.9122489049992448, "grad_norm": 2.1887209040893483, "learning_rate": 7.886799513850602e-06, "loss": 0.1276, "step": 6040 }, { "epoch": 0.9123999395861653, "grad_norm": 2.4283232124421663, "learning_rate": 7.886153785186924e-06, "loss": 0.1321, "step": 6041 }, { "epoch": 0.9125509741730856, "grad_norm": 9.643016011827312, "learning_rate": 7.885507984324516e-06, "loss": 0.1359, "step": 6042 }, { "epoch": 0.9127020087600061, "grad_norm": 2.734240202827107, "learning_rate": 7.88486211127954e-06, "loss": 0.1634, "step": 6043 }, { "epoch": 0.9128530433469264, "grad_norm": 1.7637363552738425, "learning_rate": 7.884216166068152e-06, "loss": 0.1356, "step": 6044 }, { "epoch": 0.9130040779338469, "grad_norm": 2.035765089987848, "learning_rate": 7.88357014870651e-06, "loss": 0.1405, "step": 6045 }, { "epoch": 0.9131551125207673, "grad_norm": 1.664777917018216, "learning_rate": 7.882924059210772e-06, "loss": 0.1173, "step": 6046 }, { "epoch": 0.9133061471076876, "grad_norm": 4.692614678092296, "learning_rate": 7.882277897597104e-06, "loss": 0.1753, "step": 6047 }, { "epoch": 0.9134571816946081, "grad_norm": 1.911866588128902, "learning_rate": 7.88163166388167e-06, "loss": 0.1443, "step": 6048 }, { "epoch": 0.9136082162815284, "grad_norm": 2.1774770002537855, "learning_rate": 7.880985358080633e-06, "loss": 0.1662, "step": 6049 }, { "epoch": 0.9137592508684489, "grad_norm": 1.6214249267778444, "learning_rate": 7.880338980210161e-06, "loss": 0.1177, "step": 6050 }, { "epoch": 0.9139102854553692, "grad_norm": 1.6916446868548018, "learning_rate": 7.879692530286425e-06, "loss": 0.1177, "step": 6051 }, { "epoch": 0.9140613200422897, "grad_norm": 2.7048122609664755, "learning_rate": 7.879046008325596e-06, "loss": 0.1543, "step": 6052 }, { "epoch": 0.9142123546292101, "grad_norm": 2.3505527839909752, "learning_rate": 7.878399414343848e-06, "loss": 0.1224, "step": 6053 }, { "epoch": 0.9143633892161305, "grad_norm": 1.786207459517632, "learning_rate": 7.877752748357354e-06, "loss": 0.1454, "step": 6054 }, { "epoch": 0.9145144238030509, "grad_norm": 2.309775855964906, "learning_rate": 7.877106010382294e-06, "loss": 0.1477, "step": 6055 }, { "epoch": 0.9146654583899713, "grad_norm": 2.2723667376337153, "learning_rate": 7.876459200434842e-06, "loss": 0.1368, "step": 6056 }, { "epoch": 0.9148164929768917, "grad_norm": 2.2024347010021845, "learning_rate": 7.875812318531181e-06, "loss": 0.1631, "step": 6057 }, { "epoch": 0.9149675275638122, "grad_norm": 2.2684861521274224, "learning_rate": 7.875165364687492e-06, "loss": 0.1698, "step": 6058 }, { "epoch": 0.9151185621507325, "grad_norm": 2.9221074646584704, "learning_rate": 7.87451833891996e-06, "loss": 0.1485, "step": 6059 }, { "epoch": 0.915269596737653, "grad_norm": 2.0464434475190543, "learning_rate": 7.87387124124477e-06, "loss": 0.1527, "step": 6060 }, { "epoch": 0.9154206313245733, "grad_norm": 1.894721241724609, "learning_rate": 7.873224071678112e-06, "loss": 0.1259, "step": 6061 }, { "epoch": 0.9155716659114937, "grad_norm": 1.8992932780076945, "learning_rate": 7.87257683023617e-06, "loss": 0.1352, "step": 6062 }, { "epoch": 0.9157227004984141, "grad_norm": 1.902471912588703, "learning_rate": 7.87192951693514e-06, "loss": 0.1352, "step": 6063 }, { "epoch": 0.9158737350853345, "grad_norm": 1.8104481784293152, "learning_rate": 7.871282131791211e-06, "loss": 0.113, "step": 6064 }, { "epoch": 0.916024769672255, "grad_norm": 3.724395329888364, "learning_rate": 7.870634674820582e-06, "loss": 0.1833, "step": 6065 }, { "epoch": 0.9161758042591753, "grad_norm": 2.012317798319616, "learning_rate": 7.869987146039446e-06, "loss": 0.146, "step": 6066 }, { "epoch": 0.9163268388460958, "grad_norm": 1.8280102892512868, "learning_rate": 7.869339545464001e-06, "loss": 0.1536, "step": 6067 }, { "epoch": 0.9164778734330161, "grad_norm": 2.004019515607894, "learning_rate": 7.86869187311045e-06, "loss": 0.1522, "step": 6068 }, { "epoch": 0.9166289080199366, "grad_norm": 1.9444788338420385, "learning_rate": 7.868044128994993e-06, "loss": 0.1492, "step": 6069 }, { "epoch": 0.916779942606857, "grad_norm": 2.1941585750570654, "learning_rate": 7.867396313133836e-06, "loss": 0.1472, "step": 6070 }, { "epoch": 0.9169309771937774, "grad_norm": 2.1689022373901006, "learning_rate": 7.866748425543181e-06, "loss": 0.1401, "step": 6071 }, { "epoch": 0.9170820117806978, "grad_norm": 1.5582904269474314, "learning_rate": 7.866100466239237e-06, "loss": 0.1385, "step": 6072 }, { "epoch": 0.9172330463676182, "grad_norm": 1.7727003319783659, "learning_rate": 7.865452435238213e-06, "loss": 0.116, "step": 6073 }, { "epoch": 0.9173840809545386, "grad_norm": 1.6639940752616427, "learning_rate": 7.86480433255632e-06, "loss": 0.1296, "step": 6074 }, { "epoch": 0.917535115541459, "grad_norm": 1.8325126304886437, "learning_rate": 7.86415615820977e-06, "loss": 0.1167, "step": 6075 }, { "epoch": 0.9176861501283794, "grad_norm": 2.120454908661456, "learning_rate": 7.863507912214778e-06, "loss": 0.1339, "step": 6076 }, { "epoch": 0.9178371847152998, "grad_norm": 1.6320509819590325, "learning_rate": 7.86285959458756e-06, "loss": 0.1238, "step": 6077 }, { "epoch": 0.9179882193022202, "grad_norm": 4.739780330032091, "learning_rate": 7.862211205344335e-06, "loss": 0.189, "step": 6078 }, { "epoch": 0.9181392538891406, "grad_norm": 2.1226601454759275, "learning_rate": 7.86156274450132e-06, "loss": 0.1468, "step": 6079 }, { "epoch": 0.918290288476061, "grad_norm": 2.12393497517424, "learning_rate": 7.860914212074739e-06, "loss": 0.1506, "step": 6080 }, { "epoch": 0.9184413230629814, "grad_norm": 1.8306111284619089, "learning_rate": 7.860265608080817e-06, "loss": 0.1416, "step": 6081 }, { "epoch": 0.9185923576499019, "grad_norm": 1.7905179945159626, "learning_rate": 7.859616932535776e-06, "loss": 0.1346, "step": 6082 }, { "epoch": 0.9187433922368222, "grad_norm": 2.193839492799032, "learning_rate": 7.858968185455844e-06, "loss": 0.1395, "step": 6083 }, { "epoch": 0.9188944268237427, "grad_norm": 1.662788110582665, "learning_rate": 7.85831936685725e-06, "loss": 0.1236, "step": 6084 }, { "epoch": 0.919045461410663, "grad_norm": 1.934913729392646, "learning_rate": 7.857670476756224e-06, "loss": 0.1381, "step": 6085 }, { "epoch": 0.9191964959975835, "grad_norm": 1.8087125892017635, "learning_rate": 7.857021515168998e-06, "loss": 0.1317, "step": 6086 }, { "epoch": 0.9193475305845038, "grad_norm": 2.722278129380145, "learning_rate": 7.856372482111807e-06, "loss": 0.1624, "step": 6087 }, { "epoch": 0.9194985651714243, "grad_norm": 2.3144098913683258, "learning_rate": 7.855723377600887e-06, "loss": 0.1445, "step": 6088 }, { "epoch": 0.9196495997583447, "grad_norm": 1.6581792184759352, "learning_rate": 7.855074201652477e-06, "loss": 0.1322, "step": 6089 }, { "epoch": 0.919800634345265, "grad_norm": 2.030732114579827, "learning_rate": 7.854424954282814e-06, "loss": 0.1448, "step": 6090 }, { "epoch": 0.9199516689321855, "grad_norm": 1.5753172764223786, "learning_rate": 7.85377563550814e-06, "loss": 0.1153, "step": 6091 }, { "epoch": 0.9201027035191058, "grad_norm": 2.068814981578311, "learning_rate": 7.853126245344698e-06, "loss": 0.1551, "step": 6092 }, { "epoch": 0.9202537381060263, "grad_norm": 1.8869467594008482, "learning_rate": 7.852476783808733e-06, "loss": 0.133, "step": 6093 }, { "epoch": 0.9204047726929467, "grad_norm": 2.2111534157502764, "learning_rate": 7.851827250916494e-06, "loss": 0.1279, "step": 6094 }, { "epoch": 0.9205558072798671, "grad_norm": 1.4425887276300906, "learning_rate": 7.851177646684226e-06, "loss": 0.1025, "step": 6095 }, { "epoch": 0.9207068418667875, "grad_norm": 2.4595672499085857, "learning_rate": 7.85052797112818e-06, "loss": 0.1327, "step": 6096 }, { "epoch": 0.9208578764537079, "grad_norm": 2.9133068402025373, "learning_rate": 7.84987822426461e-06, "loss": 0.1672, "step": 6097 }, { "epoch": 0.9210089110406283, "grad_norm": 2.255234433393727, "learning_rate": 7.849228406109766e-06, "loss": 0.1408, "step": 6098 }, { "epoch": 0.9211599456275487, "grad_norm": 2.122055986833665, "learning_rate": 7.848578516679906e-06, "loss": 0.1353, "step": 6099 }, { "epoch": 0.9213109802144691, "grad_norm": 1.696858614569823, "learning_rate": 7.847928555991288e-06, "loss": 0.1076, "step": 6100 }, { "epoch": 0.9214620148013896, "grad_norm": 2.9169515559401256, "learning_rate": 7.84727852406017e-06, "loss": 0.1439, "step": 6101 }, { "epoch": 0.9216130493883099, "grad_norm": 1.8581571419445615, "learning_rate": 7.846628420902815e-06, "loss": 0.1562, "step": 6102 }, { "epoch": 0.9217640839752304, "grad_norm": 2.057443279891777, "learning_rate": 7.845978246535482e-06, "loss": 0.1517, "step": 6103 }, { "epoch": 0.9219151185621507, "grad_norm": 2.2125324128145847, "learning_rate": 7.845328000974438e-06, "loss": 0.1352, "step": 6104 }, { "epoch": 0.9220661531490711, "grad_norm": 1.6438295523017437, "learning_rate": 7.844677684235946e-06, "loss": 0.1242, "step": 6105 }, { "epoch": 0.9222171877359916, "grad_norm": 1.989417599589489, "learning_rate": 7.84402729633628e-06, "loss": 0.1396, "step": 6106 }, { "epoch": 0.9223682223229119, "grad_norm": 2.5486277426439194, "learning_rate": 7.843376837291704e-06, "loss": 0.1582, "step": 6107 }, { "epoch": 0.9225192569098324, "grad_norm": 2.1307339017148803, "learning_rate": 7.842726307118493e-06, "loss": 0.1487, "step": 6108 }, { "epoch": 0.9226702914967527, "grad_norm": 2.596758208588988, "learning_rate": 7.842075705832918e-06, "loss": 0.1476, "step": 6109 }, { "epoch": 0.9228213260836732, "grad_norm": 2.2360632608457243, "learning_rate": 7.841425033451258e-06, "loss": 0.1588, "step": 6110 }, { "epoch": 0.9229723606705935, "grad_norm": 2.4278958859231263, "learning_rate": 7.840774289989787e-06, "loss": 0.162, "step": 6111 }, { "epoch": 0.923123395257514, "grad_norm": 1.8726036816473162, "learning_rate": 7.840123475464783e-06, "loss": 0.1504, "step": 6112 }, { "epoch": 0.9232744298444344, "grad_norm": 3.329075849595682, "learning_rate": 7.839472589892528e-06, "loss": 0.14, "step": 6113 }, { "epoch": 0.9234254644313548, "grad_norm": 1.8394484825099855, "learning_rate": 7.838821633289304e-06, "loss": 0.136, "step": 6114 }, { "epoch": 0.9235764990182752, "grad_norm": 1.4282771967464394, "learning_rate": 7.838170605671395e-06, "loss": 0.1229, "step": 6115 }, { "epoch": 0.9237275336051955, "grad_norm": 2.705183861481049, "learning_rate": 7.837519507055087e-06, "loss": 0.1499, "step": 6116 }, { "epoch": 0.923878568192116, "grad_norm": 2.460620361540222, "learning_rate": 7.836868337456666e-06, "loss": 0.1486, "step": 6117 }, { "epoch": 0.9240296027790365, "grad_norm": 1.9082264594967395, "learning_rate": 7.836217096892425e-06, "loss": 0.1419, "step": 6118 }, { "epoch": 0.9241806373659568, "grad_norm": 2.0054192108269695, "learning_rate": 7.835565785378652e-06, "loss": 0.1623, "step": 6119 }, { "epoch": 0.9243316719528772, "grad_norm": 2.4575983749888257, "learning_rate": 7.834914402931639e-06, "loss": 0.1617, "step": 6120 }, { "epoch": 0.9244827065397976, "grad_norm": 1.7989320387163448, "learning_rate": 7.834262949567683e-06, "loss": 0.1318, "step": 6121 }, { "epoch": 0.924633741126718, "grad_norm": 1.8472875414742154, "learning_rate": 7.833611425303082e-06, "loss": 0.1088, "step": 6122 }, { "epoch": 0.9247847757136384, "grad_norm": 2.272850491351311, "learning_rate": 7.832959830154129e-06, "loss": 0.1284, "step": 6123 }, { "epoch": 0.9249358103005588, "grad_norm": 2.4948313492547864, "learning_rate": 7.83230816413713e-06, "loss": 0.1596, "step": 6124 }, { "epoch": 0.9250868448874793, "grad_norm": 2.1587743868739806, "learning_rate": 7.831656427268384e-06, "loss": 0.1517, "step": 6125 }, { "epoch": 0.9252378794743996, "grad_norm": 2.9095187901938275, "learning_rate": 7.831004619564193e-06, "loss": 0.1513, "step": 6126 }, { "epoch": 0.9253889140613201, "grad_norm": 2.332190614628971, "learning_rate": 7.830352741040864e-06, "loss": 0.1555, "step": 6127 }, { "epoch": 0.9255399486482404, "grad_norm": 1.431970396260049, "learning_rate": 7.829700791714705e-06, "loss": 0.099, "step": 6128 }, { "epoch": 0.9256909832351609, "grad_norm": 1.8058162950995396, "learning_rate": 7.829048771602024e-06, "loss": 0.1299, "step": 6129 }, { "epoch": 0.9258420178220813, "grad_norm": 2.051255889814663, "learning_rate": 7.82839668071913e-06, "loss": 0.1476, "step": 6130 }, { "epoch": 0.9259930524090016, "grad_norm": 1.6081475056475578, "learning_rate": 7.827744519082336e-06, "loss": 0.112, "step": 6131 }, { "epoch": 0.9261440869959221, "grad_norm": 2.6903264830883673, "learning_rate": 7.827092286707959e-06, "loss": 0.1794, "step": 6132 }, { "epoch": 0.9262951215828424, "grad_norm": 2.262117407217989, "learning_rate": 7.826439983612313e-06, "loss": 0.1618, "step": 6133 }, { "epoch": 0.9264461561697629, "grad_norm": 3.036791663804331, "learning_rate": 7.825787609811714e-06, "loss": 0.1549, "step": 6134 }, { "epoch": 0.9265971907566832, "grad_norm": 1.8820239347156231, "learning_rate": 7.825135165322484e-06, "loss": 0.1491, "step": 6135 }, { "epoch": 0.9267482253436037, "grad_norm": 1.8941899276274672, "learning_rate": 7.824482650160945e-06, "loss": 0.1436, "step": 6136 }, { "epoch": 0.9268992599305241, "grad_norm": 1.9890060386262038, "learning_rate": 7.823830064343416e-06, "loss": 0.1405, "step": 6137 }, { "epoch": 0.9270502945174445, "grad_norm": 2.002761467433589, "learning_rate": 7.823177407886226e-06, "loss": 0.1479, "step": 6138 }, { "epoch": 0.9272013291043649, "grad_norm": 2.640595164695947, "learning_rate": 7.822524680805698e-06, "loss": 0.1346, "step": 6139 }, { "epoch": 0.9273523636912853, "grad_norm": 1.9229459118483518, "learning_rate": 7.821871883118165e-06, "loss": 0.1367, "step": 6140 }, { "epoch": 0.9275033982782057, "grad_norm": 2.0150692852079652, "learning_rate": 7.821219014839952e-06, "loss": 0.1533, "step": 6141 }, { "epoch": 0.9276544328651262, "grad_norm": 2.0332213614569614, "learning_rate": 7.820566075987395e-06, "loss": 0.1263, "step": 6142 }, { "epoch": 0.9278054674520465, "grad_norm": 1.8906538393560934, "learning_rate": 7.819913066576824e-06, "loss": 0.1447, "step": 6143 }, { "epoch": 0.927956502038967, "grad_norm": 2.5108241804788403, "learning_rate": 7.819259986624576e-06, "loss": 0.1807, "step": 6144 }, { "epoch": 0.9281075366258873, "grad_norm": 2.2572822181321417, "learning_rate": 7.81860683614699e-06, "loss": 0.1397, "step": 6145 }, { "epoch": 0.9282585712128077, "grad_norm": 2.468020690176771, "learning_rate": 7.817953615160402e-06, "loss": 0.1692, "step": 6146 }, { "epoch": 0.9284096057997281, "grad_norm": 2.9214358945663945, "learning_rate": 7.817300323681154e-06, "loss": 0.1711, "step": 6147 }, { "epoch": 0.9285606403866485, "grad_norm": 2.5880892843017542, "learning_rate": 7.816646961725589e-06, "loss": 0.1636, "step": 6148 }, { "epoch": 0.928711674973569, "grad_norm": 3.256244826822802, "learning_rate": 7.81599352931005e-06, "loss": 0.197, "step": 6149 }, { "epoch": 0.9288627095604893, "grad_norm": 2.252850044705775, "learning_rate": 7.815340026450885e-06, "loss": 0.1683, "step": 6150 }, { "epoch": 0.9290137441474098, "grad_norm": 1.5378968816007805, "learning_rate": 7.814686453164438e-06, "loss": 0.1165, "step": 6151 }, { "epoch": 0.9291647787343301, "grad_norm": 1.9511127605976586, "learning_rate": 7.81403280946706e-06, "loss": 0.1543, "step": 6152 }, { "epoch": 0.9293158133212506, "grad_norm": 1.988208295667253, "learning_rate": 7.813379095375105e-06, "loss": 0.1443, "step": 6153 }, { "epoch": 0.929466847908171, "grad_norm": 2.1517533944887406, "learning_rate": 7.812725310904925e-06, "loss": 0.1437, "step": 6154 }, { "epoch": 0.9296178824950914, "grad_norm": 2.3845301055856303, "learning_rate": 7.812071456072871e-06, "loss": 0.1432, "step": 6155 }, { "epoch": 0.9297689170820118, "grad_norm": 1.9139527209181386, "learning_rate": 7.811417530895304e-06, "loss": 0.1427, "step": 6156 }, { "epoch": 0.9299199516689322, "grad_norm": 1.6326745689461961, "learning_rate": 7.81076353538858e-06, "loss": 0.1245, "step": 6157 }, { "epoch": 0.9300709862558526, "grad_norm": 2.0181391886216535, "learning_rate": 7.81010946956906e-06, "loss": 0.1317, "step": 6158 }, { "epoch": 0.930222020842773, "grad_norm": 2.365333671565175, "learning_rate": 7.809455333453104e-06, "loss": 0.1389, "step": 6159 }, { "epoch": 0.9303730554296934, "grad_norm": 3.7916938161240945, "learning_rate": 7.80880112705708e-06, "loss": 0.1707, "step": 6160 }, { "epoch": 0.9305240900166138, "grad_norm": 2.362770807298771, "learning_rate": 7.808146850397348e-06, "loss": 0.134, "step": 6161 }, { "epoch": 0.9306751246035342, "grad_norm": 2.231593246496411, "learning_rate": 7.80749250349028e-06, "loss": 0.1414, "step": 6162 }, { "epoch": 0.9308261591904546, "grad_norm": 2.753725147545256, "learning_rate": 7.80683808635224e-06, "loss": 0.1754, "step": 6163 }, { "epoch": 0.930977193777375, "grad_norm": 1.9440055442394655, "learning_rate": 7.806183598999601e-06, "loss": 0.1226, "step": 6164 }, { "epoch": 0.9311282283642954, "grad_norm": 1.9485226159408624, "learning_rate": 7.805529041448737e-06, "loss": 0.147, "step": 6165 }, { "epoch": 0.9312792629512159, "grad_norm": 1.857348768229614, "learning_rate": 7.804874413716019e-06, "loss": 0.1421, "step": 6166 }, { "epoch": 0.9314302975381362, "grad_norm": 2.0364678218755254, "learning_rate": 7.804219715817826e-06, "loss": 0.1456, "step": 6167 }, { "epoch": 0.9315813321250567, "grad_norm": 2.0481895045467593, "learning_rate": 7.803564947770532e-06, "loss": 0.1398, "step": 6168 }, { "epoch": 0.931732366711977, "grad_norm": 2.0731594494010754, "learning_rate": 7.80291010959052e-06, "loss": 0.1437, "step": 6169 }, { "epoch": 0.9318834012988975, "grad_norm": 2.158974070788487, "learning_rate": 7.802255201294166e-06, "loss": 0.1652, "step": 6170 }, { "epoch": 0.9320344358858178, "grad_norm": 1.6301756531763625, "learning_rate": 7.80160022289786e-06, "loss": 0.1277, "step": 6171 }, { "epoch": 0.9321854704727383, "grad_norm": 2.414708302273297, "learning_rate": 7.800945174417984e-06, "loss": 0.1551, "step": 6172 }, { "epoch": 0.9323365050596587, "grad_norm": 1.65446432675827, "learning_rate": 7.80029005587092e-06, "loss": 0.1112, "step": 6173 }, { "epoch": 0.932487539646579, "grad_norm": 2.5257034031101133, "learning_rate": 7.79963486727306e-06, "loss": 0.155, "step": 6174 }, { "epoch": 0.9326385742334995, "grad_norm": 1.7655660814985423, "learning_rate": 7.798979608640795e-06, "loss": 0.15, "step": 6175 }, { "epoch": 0.9327896088204198, "grad_norm": 1.9027927816029353, "learning_rate": 7.798324279990515e-06, "loss": 0.1519, "step": 6176 }, { "epoch": 0.9329406434073403, "grad_norm": 2.8900760312249005, "learning_rate": 7.79766888133861e-06, "loss": 0.1442, "step": 6177 }, { "epoch": 0.9330916779942607, "grad_norm": 2.1547461474658647, "learning_rate": 7.797013412701485e-06, "loss": 0.1477, "step": 6178 }, { "epoch": 0.9332427125811811, "grad_norm": 1.8096744926757204, "learning_rate": 7.796357874095526e-06, "loss": 0.1245, "step": 6179 }, { "epoch": 0.9333937471681015, "grad_norm": 1.8521865605545804, "learning_rate": 7.795702265537137e-06, "loss": 0.1298, "step": 6180 }, { "epoch": 0.9335447817550219, "grad_norm": 1.7163577525882612, "learning_rate": 7.795046587042718e-06, "loss": 0.1233, "step": 6181 }, { "epoch": 0.9336958163419423, "grad_norm": 2.704596536663153, "learning_rate": 7.794390838628672e-06, "loss": 0.1594, "step": 6182 }, { "epoch": 0.9338468509288627, "grad_norm": 2.1705925271676727, "learning_rate": 7.7937350203114e-06, "loss": 0.1128, "step": 6183 }, { "epoch": 0.9339978855157831, "grad_norm": 2.49395943389743, "learning_rate": 7.793079132107308e-06, "loss": 0.1669, "step": 6184 }, { "epoch": 0.9341489201027036, "grad_norm": 2.376541097536307, "learning_rate": 7.792423174032809e-06, "loss": 0.1633, "step": 6185 }, { "epoch": 0.9342999546896239, "grad_norm": 1.5983394435522917, "learning_rate": 7.791767146104305e-06, "loss": 0.1032, "step": 6186 }, { "epoch": 0.9344509892765444, "grad_norm": 5.3275262522370905, "learning_rate": 7.79111104833821e-06, "loss": 0.1809, "step": 6187 }, { "epoch": 0.9346020238634647, "grad_norm": 1.5936738079907362, "learning_rate": 7.790454880750937e-06, "loss": 0.1027, "step": 6188 }, { "epoch": 0.9347530584503851, "grad_norm": 2.8717003419165743, "learning_rate": 7.789798643358899e-06, "loss": 0.1588, "step": 6189 }, { "epoch": 0.9349040930373056, "grad_norm": 1.6776876565768757, "learning_rate": 7.789142336178515e-06, "loss": 0.1341, "step": 6190 }, { "epoch": 0.9350551276242259, "grad_norm": 1.9583746942744902, "learning_rate": 7.788485959226199e-06, "loss": 0.1509, "step": 6191 }, { "epoch": 0.9352061622111464, "grad_norm": 2.4912766391991408, "learning_rate": 7.787829512518374e-06, "loss": 0.1333, "step": 6192 }, { "epoch": 0.9353571967980667, "grad_norm": 1.4999731011953925, "learning_rate": 7.787172996071458e-06, "loss": 0.1197, "step": 6193 }, { "epoch": 0.9355082313849872, "grad_norm": 1.8367072085144127, "learning_rate": 7.786516409901877e-06, "loss": 0.1433, "step": 6194 }, { "epoch": 0.9356592659719075, "grad_norm": 2.1613030540033993, "learning_rate": 7.785859754026055e-06, "loss": 0.1527, "step": 6195 }, { "epoch": 0.935810300558828, "grad_norm": 1.8267232445699153, "learning_rate": 7.785203028460418e-06, "loss": 0.1355, "step": 6196 }, { "epoch": 0.9359613351457484, "grad_norm": 2.0169961208573786, "learning_rate": 7.784546233221395e-06, "loss": 0.1442, "step": 6197 }, { "epoch": 0.9361123697326688, "grad_norm": 2.25675449676405, "learning_rate": 7.783889368325416e-06, "loss": 0.1476, "step": 6198 }, { "epoch": 0.9362634043195892, "grad_norm": 1.7191022607225137, "learning_rate": 7.783232433788913e-06, "loss": 0.1305, "step": 6199 }, { "epoch": 0.9364144389065095, "grad_norm": 1.4602372156158883, "learning_rate": 7.78257542962832e-06, "loss": 0.1044, "step": 6200 }, { "epoch": 0.93656547349343, "grad_norm": 1.9196552930338526, "learning_rate": 7.78191835586007e-06, "loss": 0.1484, "step": 6201 }, { "epoch": 0.9367165080803505, "grad_norm": 1.8897251427705528, "learning_rate": 7.781261212500603e-06, "loss": 0.1651, "step": 6202 }, { "epoch": 0.9368675426672708, "grad_norm": 2.0858874277443036, "learning_rate": 7.780603999566357e-06, "loss": 0.1465, "step": 6203 }, { "epoch": 0.9370185772541912, "grad_norm": 2.3976825439899643, "learning_rate": 7.77994671707377e-06, "loss": 0.1556, "step": 6204 }, { "epoch": 0.9371696118411116, "grad_norm": 2.1193758813424473, "learning_rate": 7.779289365039288e-06, "loss": 0.1228, "step": 6205 }, { "epoch": 0.937320646428032, "grad_norm": 1.9459891367092648, "learning_rate": 7.778631943479353e-06, "loss": 0.1457, "step": 6206 }, { "epoch": 0.9374716810149524, "grad_norm": 2.0344919800931645, "learning_rate": 7.77797445241041e-06, "loss": 0.1297, "step": 6207 }, { "epoch": 0.9376227156018728, "grad_norm": 2.09478931776698, "learning_rate": 7.77731689184891e-06, "loss": 0.157, "step": 6208 }, { "epoch": 0.9377737501887933, "grad_norm": 2.5247150974237638, "learning_rate": 7.776659261811298e-06, "loss": 0.148, "step": 6209 }, { "epoch": 0.9379247847757136, "grad_norm": 1.9996031822490905, "learning_rate": 7.776001562314029e-06, "loss": 0.1391, "step": 6210 }, { "epoch": 0.9380758193626341, "grad_norm": 3.580171479018156, "learning_rate": 7.775343793373552e-06, "loss": 0.1449, "step": 6211 }, { "epoch": 0.9382268539495544, "grad_norm": 1.837758628996855, "learning_rate": 7.774685955006322e-06, "loss": 0.1402, "step": 6212 }, { "epoch": 0.9383778885364749, "grad_norm": 1.8115406036798747, "learning_rate": 7.774028047228797e-06, "loss": 0.1483, "step": 6213 }, { "epoch": 0.9385289231233953, "grad_norm": 1.7593262810953973, "learning_rate": 7.773370070057435e-06, "loss": 0.1453, "step": 6214 }, { "epoch": 0.9386799577103156, "grad_norm": 1.4920481847056108, "learning_rate": 7.772712023508693e-06, "loss": 0.1188, "step": 6215 }, { "epoch": 0.9388309922972361, "grad_norm": 2.1711302159963966, "learning_rate": 7.772053907599035e-06, "loss": 0.1429, "step": 6216 }, { "epoch": 0.9389820268841564, "grad_norm": 2.1565450280739054, "learning_rate": 7.771395722344925e-06, "loss": 0.1568, "step": 6217 }, { "epoch": 0.9391330614710769, "grad_norm": 3.860243167592441, "learning_rate": 7.770737467762825e-06, "loss": 0.1673, "step": 6218 }, { "epoch": 0.9392840960579972, "grad_norm": 1.8501345369457383, "learning_rate": 7.770079143869203e-06, "loss": 0.1446, "step": 6219 }, { "epoch": 0.9394351306449177, "grad_norm": 1.7077554445669008, "learning_rate": 7.769420750680526e-06, "loss": 0.1298, "step": 6220 }, { "epoch": 0.9395861652318381, "grad_norm": 2.1037096163909736, "learning_rate": 7.768762288213266e-06, "loss": 0.1251, "step": 6221 }, { "epoch": 0.9397371998187585, "grad_norm": 2.1751696417268827, "learning_rate": 7.768103756483893e-06, "loss": 0.1402, "step": 6222 }, { "epoch": 0.9398882344056789, "grad_norm": 2.404047489856199, "learning_rate": 7.767445155508884e-06, "loss": 0.1504, "step": 6223 }, { "epoch": 0.9400392689925993, "grad_norm": 1.8030046040644794, "learning_rate": 7.76678648530471e-06, "loss": 0.1341, "step": 6224 }, { "epoch": 0.9401903035795197, "grad_norm": 2.1027306441160394, "learning_rate": 7.766127745887851e-06, "loss": 0.1431, "step": 6225 }, { "epoch": 0.9403413381664402, "grad_norm": 1.8553631027460982, "learning_rate": 7.765468937274783e-06, "loss": 0.1355, "step": 6226 }, { "epoch": 0.9404923727533605, "grad_norm": 3.8657156208946164, "learning_rate": 7.764810059481989e-06, "loss": 0.1377, "step": 6227 }, { "epoch": 0.940643407340281, "grad_norm": 3.0940925211907206, "learning_rate": 7.764151112525949e-06, "loss": 0.1721, "step": 6228 }, { "epoch": 0.9407944419272013, "grad_norm": 2.0724890385656627, "learning_rate": 7.763492096423148e-06, "loss": 0.1349, "step": 6229 }, { "epoch": 0.9409454765141217, "grad_norm": 1.957924058481581, "learning_rate": 7.762833011190074e-06, "loss": 0.1509, "step": 6230 }, { "epoch": 0.9410965111010421, "grad_norm": 1.8656126963546975, "learning_rate": 7.762173856843211e-06, "loss": 0.1414, "step": 6231 }, { "epoch": 0.9412475456879625, "grad_norm": 2.027195897637355, "learning_rate": 7.76151463339905e-06, "loss": 0.1379, "step": 6232 }, { "epoch": 0.941398580274883, "grad_norm": 2.138538827028259, "learning_rate": 7.760855340874079e-06, "loss": 0.1229, "step": 6233 }, { "epoch": 0.9415496148618033, "grad_norm": 2.2856756272829024, "learning_rate": 7.760195979284793e-06, "loss": 0.1453, "step": 6234 }, { "epoch": 0.9417006494487238, "grad_norm": 1.7935428450148896, "learning_rate": 7.759536548647686e-06, "loss": 0.1429, "step": 6235 }, { "epoch": 0.9418516840356441, "grad_norm": 2.0153225107463775, "learning_rate": 7.758877048979257e-06, "loss": 0.112, "step": 6236 }, { "epoch": 0.9420027186225646, "grad_norm": 1.178820529428031, "learning_rate": 7.758217480295997e-06, "loss": 0.0883, "step": 6237 }, { "epoch": 0.942153753209485, "grad_norm": 2.048715664413856, "learning_rate": 7.757557842614409e-06, "loss": 0.1558, "step": 6238 }, { "epoch": 0.9423047877964054, "grad_norm": 1.7615349048019537, "learning_rate": 7.756898135950997e-06, "loss": 0.1416, "step": 6239 }, { "epoch": 0.9424558223833258, "grad_norm": 2.778098556322793, "learning_rate": 7.756238360322259e-06, "loss": 0.1405, "step": 6240 }, { "epoch": 0.9426068569702462, "grad_norm": 1.8523517613444065, "learning_rate": 7.755578515744705e-06, "loss": 0.1321, "step": 6241 }, { "epoch": 0.9427578915571666, "grad_norm": 1.9573603676138607, "learning_rate": 7.754918602234835e-06, "loss": 0.1403, "step": 6242 }, { "epoch": 0.942908926144087, "grad_norm": 1.8499643511736292, "learning_rate": 7.754258619809162e-06, "loss": 0.1357, "step": 6243 }, { "epoch": 0.9430599607310074, "grad_norm": 2.585260086470824, "learning_rate": 7.753598568484193e-06, "loss": 0.1472, "step": 6244 }, { "epoch": 0.9432109953179278, "grad_norm": 1.8165686041927682, "learning_rate": 7.752938448276443e-06, "loss": 0.1282, "step": 6245 }, { "epoch": 0.9433620299048482, "grad_norm": 1.9749545652402654, "learning_rate": 7.752278259202421e-06, "loss": 0.1316, "step": 6246 }, { "epoch": 0.9435130644917686, "grad_norm": 3.654569105293357, "learning_rate": 7.751618001278645e-06, "loss": 0.1784, "step": 6247 }, { "epoch": 0.943664099078689, "grad_norm": 2.0721799459414387, "learning_rate": 7.750957674521631e-06, "loss": 0.1288, "step": 6248 }, { "epoch": 0.9438151336656094, "grad_norm": 1.9795428589875956, "learning_rate": 7.750297278947898e-06, "loss": 0.1359, "step": 6249 }, { "epoch": 0.9439661682525299, "grad_norm": 1.716127340021652, "learning_rate": 7.749636814573965e-06, "loss": 0.1312, "step": 6250 }, { "epoch": 0.9441172028394502, "grad_norm": 5.433221664250206, "learning_rate": 7.748976281416352e-06, "loss": 0.1612, "step": 6251 }, { "epoch": 0.9442682374263707, "grad_norm": 1.847830616672005, "learning_rate": 7.748315679491586e-06, "loss": 0.1331, "step": 6252 }, { "epoch": 0.944419272013291, "grad_norm": 1.6473186872701955, "learning_rate": 7.747655008816192e-06, "loss": 0.1256, "step": 6253 }, { "epoch": 0.9445703066002115, "grad_norm": 2.244044764063548, "learning_rate": 7.746994269406695e-06, "loss": 0.1639, "step": 6254 }, { "epoch": 0.9447213411871318, "grad_norm": 2.4241849528671304, "learning_rate": 7.746333461279626e-06, "loss": 0.1822, "step": 6255 }, { "epoch": 0.9448723757740523, "grad_norm": 2.151803404899984, "learning_rate": 7.745672584451513e-06, "loss": 0.1493, "step": 6256 }, { "epoch": 0.9450234103609727, "grad_norm": 2.035042405599797, "learning_rate": 7.745011638938893e-06, "loss": 0.145, "step": 6257 }, { "epoch": 0.945174444947893, "grad_norm": 1.9255202558326316, "learning_rate": 7.744350624758293e-06, "loss": 0.1429, "step": 6258 }, { "epoch": 0.9453254795348135, "grad_norm": 1.6681773343205477, "learning_rate": 7.743689541926254e-06, "loss": 0.1258, "step": 6259 }, { "epoch": 0.9454765141217338, "grad_norm": 1.890095145451446, "learning_rate": 7.74302839045931e-06, "loss": 0.1509, "step": 6260 }, { "epoch": 0.9456275487086543, "grad_norm": 2.624414907122292, "learning_rate": 7.742367170374003e-06, "loss": 0.1665, "step": 6261 }, { "epoch": 0.9457785832955747, "grad_norm": 2.0930396129673428, "learning_rate": 7.741705881686872e-06, "loss": 0.1339, "step": 6262 }, { "epoch": 0.9459296178824951, "grad_norm": 2.032484307447612, "learning_rate": 7.74104452441446e-06, "loss": 0.1445, "step": 6263 }, { "epoch": 0.9460806524694155, "grad_norm": 2.415984127401505, "learning_rate": 7.740383098573312e-06, "loss": 0.1597, "step": 6264 }, { "epoch": 0.9462316870563359, "grad_norm": 2.396770984064722, "learning_rate": 7.739721604179972e-06, "loss": 0.1669, "step": 6265 }, { "epoch": 0.9463827216432563, "grad_norm": 2.1512028593356267, "learning_rate": 7.739060041250988e-06, "loss": 0.1413, "step": 6266 }, { "epoch": 0.9465337562301767, "grad_norm": 2.620676124009259, "learning_rate": 7.73839840980291e-06, "loss": 0.1343, "step": 6267 }, { "epoch": 0.9466847908170971, "grad_norm": 2.221250106332895, "learning_rate": 7.737736709852291e-06, "loss": 0.1393, "step": 6268 }, { "epoch": 0.9468358254040176, "grad_norm": 1.7414255197150426, "learning_rate": 7.73707494141568e-06, "loss": 0.1249, "step": 6269 }, { "epoch": 0.9469868599909379, "grad_norm": 1.842299176859912, "learning_rate": 7.736413104509635e-06, "loss": 0.132, "step": 6270 }, { "epoch": 0.9471378945778584, "grad_norm": 2.471310742830751, "learning_rate": 7.73575119915071e-06, "loss": 0.1632, "step": 6271 }, { "epoch": 0.9472889291647787, "grad_norm": 2.4670405669213182, "learning_rate": 7.735089225355463e-06, "loss": 0.1588, "step": 6272 }, { "epoch": 0.9474399637516991, "grad_norm": 2.0948124493520446, "learning_rate": 7.734427183140455e-06, "loss": 0.1313, "step": 6273 }, { "epoch": 0.9475909983386196, "grad_norm": 1.8305329006866875, "learning_rate": 7.733765072522246e-06, "loss": 0.1359, "step": 6274 }, { "epoch": 0.9477420329255399, "grad_norm": 2.4422073220890006, "learning_rate": 7.7331028935174e-06, "loss": 0.1483, "step": 6275 }, { "epoch": 0.9478930675124604, "grad_norm": 2.25085735218011, "learning_rate": 7.73244064614248e-06, "loss": 0.1424, "step": 6276 }, { "epoch": 0.9480441020993807, "grad_norm": 1.6096220238129144, "learning_rate": 7.731778330414055e-06, "loss": 0.104, "step": 6277 }, { "epoch": 0.9481951366863012, "grad_norm": 1.8786662594980712, "learning_rate": 7.731115946348693e-06, "loss": 0.1312, "step": 6278 }, { "epoch": 0.9483461712732215, "grad_norm": 2.9276311616689643, "learning_rate": 7.730453493962962e-06, "loss": 0.1497, "step": 6279 }, { "epoch": 0.948497205860142, "grad_norm": 2.2280696836033598, "learning_rate": 7.729790973273434e-06, "loss": 0.1655, "step": 6280 }, { "epoch": 0.9486482404470624, "grad_norm": 1.7143040628531931, "learning_rate": 7.729128384296683e-06, "loss": 0.1175, "step": 6281 }, { "epoch": 0.9487992750339828, "grad_norm": 2.984244900025774, "learning_rate": 7.728465727049285e-06, "loss": 0.1429, "step": 6282 }, { "epoch": 0.9489503096209032, "grad_norm": 2.003587147811294, "learning_rate": 7.727803001547816e-06, "loss": 0.1447, "step": 6283 }, { "epoch": 0.9491013442078235, "grad_norm": 2.4020200178174473, "learning_rate": 7.727140207808851e-06, "loss": 0.1614, "step": 6284 }, { "epoch": 0.949252378794744, "grad_norm": 2.058836819552213, "learning_rate": 7.726477345848977e-06, "loss": 0.1452, "step": 6285 }, { "epoch": 0.9494034133816645, "grad_norm": 2.145957402832711, "learning_rate": 7.72581441568477e-06, "loss": 0.147, "step": 6286 }, { "epoch": 0.9495544479685848, "grad_norm": 2.5783232640879454, "learning_rate": 7.725151417332817e-06, "loss": 0.1441, "step": 6287 }, { "epoch": 0.9497054825555052, "grad_norm": 2.467043553584293, "learning_rate": 7.7244883508097e-06, "loss": 0.1535, "step": 6288 }, { "epoch": 0.9498565171424256, "grad_norm": 2.694516808356515, "learning_rate": 7.723825216132009e-06, "loss": 0.1853, "step": 6289 }, { "epoch": 0.950007551729346, "grad_norm": 3.061479847709262, "learning_rate": 7.723162013316332e-06, "loss": 0.134, "step": 6290 }, { "epoch": 0.9501585863162664, "grad_norm": 2.588873555392299, "learning_rate": 7.722498742379259e-06, "loss": 0.1334, "step": 6291 }, { "epoch": 0.9503096209031868, "grad_norm": 1.7762997384279213, "learning_rate": 7.72183540333738e-06, "loss": 0.1371, "step": 6292 }, { "epoch": 0.9504606554901073, "grad_norm": 1.960869118855331, "learning_rate": 7.721171996207293e-06, "loss": 0.132, "step": 6293 }, { "epoch": 0.9506116900770276, "grad_norm": 1.862694192186607, "learning_rate": 7.72050852100559e-06, "loss": 0.1309, "step": 6294 }, { "epoch": 0.9507627246639481, "grad_norm": 1.7426130960898174, "learning_rate": 7.71984497774887e-06, "loss": 0.1472, "step": 6295 }, { "epoch": 0.9509137592508684, "grad_norm": 2.0104542993508954, "learning_rate": 7.719181366453731e-06, "loss": 0.1585, "step": 6296 }, { "epoch": 0.9510647938377889, "grad_norm": 1.9373424723285957, "learning_rate": 7.718517687136775e-06, "loss": 0.1533, "step": 6297 }, { "epoch": 0.9512158284247093, "grad_norm": 2.143862964316114, "learning_rate": 7.717853939814603e-06, "loss": 0.1623, "step": 6298 }, { "epoch": 0.9513668630116296, "grad_norm": 1.5766832698245103, "learning_rate": 7.71719012450382e-06, "loss": 0.1215, "step": 6299 }, { "epoch": 0.9515178975985501, "grad_norm": 1.9526105273097634, "learning_rate": 7.71652624122103e-06, "loss": 0.1397, "step": 6300 }, { "epoch": 0.9516689321854704, "grad_norm": 2.1759153126870547, "learning_rate": 7.715862289982842e-06, "loss": 0.1509, "step": 6301 }, { "epoch": 0.9518199667723909, "grad_norm": 2.1493997697218576, "learning_rate": 7.715198270805864e-06, "loss": 0.1623, "step": 6302 }, { "epoch": 0.9519710013593112, "grad_norm": 2.344412338535928, "learning_rate": 7.714534183706708e-06, "loss": 0.1422, "step": 6303 }, { "epoch": 0.9521220359462317, "grad_norm": 2.3934921631554356, "learning_rate": 7.713870028701986e-06, "loss": 0.1815, "step": 6304 }, { "epoch": 0.9522730705331521, "grad_norm": 1.455479814547663, "learning_rate": 7.71320580580831e-06, "loss": 0.1041, "step": 6305 }, { "epoch": 0.9524241051200725, "grad_norm": 2.491819449376073, "learning_rate": 7.7125415150423e-06, "loss": 0.1407, "step": 6306 }, { "epoch": 0.9525751397069929, "grad_norm": 1.8559433456763585, "learning_rate": 7.711877156420572e-06, "loss": 0.147, "step": 6307 }, { "epoch": 0.9527261742939133, "grad_norm": 1.7676260586366004, "learning_rate": 7.711212729959746e-06, "loss": 0.1418, "step": 6308 }, { "epoch": 0.9528772088808337, "grad_norm": 1.5405410378042264, "learning_rate": 7.710548235676438e-06, "loss": 0.1117, "step": 6309 }, { "epoch": 0.9530282434677542, "grad_norm": 1.8759324500003836, "learning_rate": 7.709883673587277e-06, "loss": 0.146, "step": 6310 }, { "epoch": 0.9531792780546745, "grad_norm": 2.8348227522128826, "learning_rate": 7.709219043708885e-06, "loss": 0.1695, "step": 6311 }, { "epoch": 0.953330312641595, "grad_norm": 2.021643988983232, "learning_rate": 7.708554346057886e-06, "loss": 0.1446, "step": 6312 }, { "epoch": 0.9534813472285153, "grad_norm": 1.815709896064738, "learning_rate": 7.707889580650912e-06, "loss": 0.1308, "step": 6313 }, { "epoch": 0.9536323818154357, "grad_norm": 2.3602129620798697, "learning_rate": 7.707224747504589e-06, "loss": 0.1855, "step": 6314 }, { "epoch": 0.9537834164023561, "grad_norm": 2.008610510026111, "learning_rate": 7.706559846635551e-06, "loss": 0.1478, "step": 6315 }, { "epoch": 0.9539344509892765, "grad_norm": 2.150602284495439, "learning_rate": 7.705894878060427e-06, "loss": 0.1489, "step": 6316 }, { "epoch": 0.954085485576197, "grad_norm": 2.698044463537122, "learning_rate": 7.705229841795854e-06, "loss": 0.1472, "step": 6317 }, { "epoch": 0.9542365201631173, "grad_norm": 3.136224198119438, "learning_rate": 7.704564737858471e-06, "loss": 0.1598, "step": 6318 }, { "epoch": 0.9543875547500378, "grad_norm": 2.1003021190953146, "learning_rate": 7.703899566264912e-06, "loss": 0.151, "step": 6319 }, { "epoch": 0.9545385893369581, "grad_norm": 2.332263083964025, "learning_rate": 7.703234327031815e-06, "loss": 0.1406, "step": 6320 }, { "epoch": 0.9546896239238786, "grad_norm": 1.5850572928523443, "learning_rate": 7.702569020175827e-06, "loss": 0.1121, "step": 6321 }, { "epoch": 0.954840658510799, "grad_norm": 1.7219933237007934, "learning_rate": 7.701903645713587e-06, "loss": 0.1198, "step": 6322 }, { "epoch": 0.9549916930977194, "grad_norm": 1.7164121035533806, "learning_rate": 7.701238203661741e-06, "loss": 0.1245, "step": 6323 }, { "epoch": 0.9551427276846398, "grad_norm": 1.8648342100631319, "learning_rate": 7.700572694036934e-06, "loss": 0.1475, "step": 6324 }, { "epoch": 0.9552937622715602, "grad_norm": 1.9769184250628828, "learning_rate": 7.699907116855816e-06, "loss": 0.1466, "step": 6325 }, { "epoch": 0.9554447968584806, "grad_norm": 2.0522557756602313, "learning_rate": 7.699241472135036e-06, "loss": 0.1435, "step": 6326 }, { "epoch": 0.955595831445401, "grad_norm": 2.731090731868627, "learning_rate": 7.698575759891244e-06, "loss": 0.1664, "step": 6327 }, { "epoch": 0.9557468660323214, "grad_norm": 1.8986390997444837, "learning_rate": 7.697909980141098e-06, "loss": 0.1246, "step": 6328 }, { "epoch": 0.9558979006192418, "grad_norm": 2.109393962806471, "learning_rate": 7.697244132901247e-06, "loss": 0.1324, "step": 6329 }, { "epoch": 0.9560489352061622, "grad_norm": 2.0143328779532004, "learning_rate": 7.69657821818835e-06, "loss": 0.1583, "step": 6330 }, { "epoch": 0.9561999697930826, "grad_norm": 2.3798561825540627, "learning_rate": 7.695912236019064e-06, "loss": 0.1262, "step": 6331 }, { "epoch": 0.956351004380003, "grad_norm": 1.838984332150822, "learning_rate": 7.695246186410054e-06, "loss": 0.1364, "step": 6332 }, { "epoch": 0.9565020389669234, "grad_norm": 1.9124468080310872, "learning_rate": 7.694580069377976e-06, "loss": 0.1489, "step": 6333 }, { "epoch": 0.9566530735538439, "grad_norm": 2.5808780393184603, "learning_rate": 7.693913884939494e-06, "loss": 0.1318, "step": 6334 }, { "epoch": 0.9568041081407642, "grad_norm": 1.840059903487448, "learning_rate": 7.693247633111274e-06, "loss": 0.1323, "step": 6335 }, { "epoch": 0.9569551427276847, "grad_norm": 2.0480373421734575, "learning_rate": 7.692581313909983e-06, "loss": 0.1193, "step": 6336 }, { "epoch": 0.957106177314605, "grad_norm": 1.8485713761035776, "learning_rate": 7.69191492735229e-06, "loss": 0.1185, "step": 6337 }, { "epoch": 0.9572572119015255, "grad_norm": 1.8089397970406518, "learning_rate": 7.691248473454863e-06, "loss": 0.1353, "step": 6338 }, { "epoch": 0.9574082464884458, "grad_norm": 1.9177938184844008, "learning_rate": 7.690581952234375e-06, "loss": 0.1237, "step": 6339 }, { "epoch": 0.9575592810753663, "grad_norm": 2.3760525667708547, "learning_rate": 7.6899153637075e-06, "loss": 0.1464, "step": 6340 }, { "epoch": 0.9577103156622867, "grad_norm": 1.9652787323629735, "learning_rate": 7.68924870789091e-06, "loss": 0.1388, "step": 6341 }, { "epoch": 0.957861350249207, "grad_norm": 2.566055613871269, "learning_rate": 7.688581984801285e-06, "loss": 0.1564, "step": 6342 }, { "epoch": 0.9580123848361275, "grad_norm": 1.5949634689861454, "learning_rate": 7.687915194455303e-06, "loss": 0.1206, "step": 6343 }, { "epoch": 0.9581634194230478, "grad_norm": 2.006424746303927, "learning_rate": 7.687248336869644e-06, "loss": 0.1442, "step": 6344 }, { "epoch": 0.9583144540099683, "grad_norm": 2.440983849754053, "learning_rate": 7.686581412060988e-06, "loss": 0.1247, "step": 6345 }, { "epoch": 0.9584654885968887, "grad_norm": 1.8954841335492363, "learning_rate": 7.68591442004602e-06, "loss": 0.1379, "step": 6346 }, { "epoch": 0.9586165231838091, "grad_norm": 2.0795691058035604, "learning_rate": 7.685247360841425e-06, "loss": 0.1307, "step": 6347 }, { "epoch": 0.9587675577707295, "grad_norm": 2.109020347540415, "learning_rate": 7.68458023446389e-06, "loss": 0.1312, "step": 6348 }, { "epoch": 0.9589185923576499, "grad_norm": 2.1488723451080207, "learning_rate": 7.683913040930104e-06, "loss": 0.1472, "step": 6349 }, { "epoch": 0.9590696269445703, "grad_norm": 2.0421515971951796, "learning_rate": 7.683245780256757e-06, "loss": 0.1423, "step": 6350 }, { "epoch": 0.9592206615314907, "grad_norm": 1.4557633904827951, "learning_rate": 7.68257845246054e-06, "loss": 0.1101, "step": 6351 }, { "epoch": 0.9593716961184111, "grad_norm": 2.234067635934867, "learning_rate": 7.681911057558144e-06, "loss": 0.1379, "step": 6352 }, { "epoch": 0.9595227307053316, "grad_norm": 1.7932765234939332, "learning_rate": 7.681243595566271e-06, "loss": 0.123, "step": 6353 }, { "epoch": 0.9596737652922519, "grad_norm": 1.7820593677652137, "learning_rate": 7.680576066501615e-06, "loss": 0.1429, "step": 6354 }, { "epoch": 0.9598247998791724, "grad_norm": 1.944607093533593, "learning_rate": 7.679908470380871e-06, "loss": 0.144, "step": 6355 }, { "epoch": 0.9599758344660927, "grad_norm": 1.6970604521804613, "learning_rate": 7.679240807220744e-06, "loss": 0.1199, "step": 6356 }, { "epoch": 0.9601268690530131, "grad_norm": 2.4207464694784218, "learning_rate": 7.678573077037932e-06, "loss": 0.1345, "step": 6357 }, { "epoch": 0.9602779036399336, "grad_norm": 1.9409323161627134, "learning_rate": 7.677905279849142e-06, "loss": 0.1307, "step": 6358 }, { "epoch": 0.9604289382268539, "grad_norm": 2.145497415892822, "learning_rate": 7.67723741567108e-06, "loss": 0.1293, "step": 6359 }, { "epoch": 0.9605799728137744, "grad_norm": 1.8572280368540892, "learning_rate": 7.67656948452045e-06, "loss": 0.1553, "step": 6360 }, { "epoch": 0.9607310074006947, "grad_norm": 1.9192931882746809, "learning_rate": 7.675901486413959e-06, "loss": 0.1081, "step": 6361 }, { "epoch": 0.9608820419876152, "grad_norm": 2.2548826411651013, "learning_rate": 7.675233421368323e-06, "loss": 0.1459, "step": 6362 }, { "epoch": 0.9610330765745355, "grad_norm": 2.507802824306766, "learning_rate": 7.674565289400248e-06, "loss": 0.1552, "step": 6363 }, { "epoch": 0.961184111161456, "grad_norm": 1.671588712462686, "learning_rate": 7.673897090526453e-06, "loss": 0.1393, "step": 6364 }, { "epoch": 0.9613351457483764, "grad_norm": 1.8236207659204104, "learning_rate": 7.673228824763653e-06, "loss": 0.1459, "step": 6365 }, { "epoch": 0.9614861803352968, "grad_norm": 2.1193800896773736, "learning_rate": 7.67256049212856e-06, "loss": 0.1376, "step": 6366 }, { "epoch": 0.9616372149222172, "grad_norm": 1.8643975423792551, "learning_rate": 7.671892092637896e-06, "loss": 0.1395, "step": 6367 }, { "epoch": 0.9617882495091375, "grad_norm": 1.9830965156722853, "learning_rate": 7.671223626308382e-06, "loss": 0.1383, "step": 6368 }, { "epoch": 0.961939284096058, "grad_norm": 1.9532205427483993, "learning_rate": 7.67055509315674e-06, "loss": 0.14, "step": 6369 }, { "epoch": 0.9620903186829785, "grad_norm": 1.8415979543699894, "learning_rate": 7.669886493199691e-06, "loss": 0.1204, "step": 6370 }, { "epoch": 0.9622413532698988, "grad_norm": 1.854327980573512, "learning_rate": 7.669217826453966e-06, "loss": 0.138, "step": 6371 }, { "epoch": 0.9623923878568192, "grad_norm": 2.66324008793751, "learning_rate": 7.668549092936286e-06, "loss": 0.121, "step": 6372 }, { "epoch": 0.9625434224437396, "grad_norm": 2.0478678924509586, "learning_rate": 7.667880292663382e-06, "loss": 0.1614, "step": 6373 }, { "epoch": 0.96269445703066, "grad_norm": 2.3631113705967617, "learning_rate": 7.667211425651984e-06, "loss": 0.1409, "step": 6374 }, { "epoch": 0.9628454916175804, "grad_norm": 1.6860833834739815, "learning_rate": 7.666542491918828e-06, "loss": 0.1284, "step": 6375 }, { "epoch": 0.9629965262045008, "grad_norm": 2.827286722502377, "learning_rate": 7.66587349148064e-06, "loss": 0.182, "step": 6376 }, { "epoch": 0.9631475607914213, "grad_norm": 3.2324553753073895, "learning_rate": 7.665204424354162e-06, "loss": 0.1661, "step": 6377 }, { "epoch": 0.9632985953783416, "grad_norm": 2.2677037778636864, "learning_rate": 7.66453529055613e-06, "loss": 0.165, "step": 6378 }, { "epoch": 0.9634496299652621, "grad_norm": 2.2362963013647796, "learning_rate": 7.66386609010328e-06, "loss": 0.1506, "step": 6379 }, { "epoch": 0.9636006645521824, "grad_norm": 1.9797454926220628, "learning_rate": 7.663196823012355e-06, "loss": 0.1418, "step": 6380 }, { "epoch": 0.9637516991391029, "grad_norm": 2.050922623353572, "learning_rate": 7.662527489300095e-06, "loss": 0.1491, "step": 6381 }, { "epoch": 0.9639027337260233, "grad_norm": 2.2890994705724688, "learning_rate": 7.661858088983248e-06, "loss": 0.1572, "step": 6382 }, { "epoch": 0.9640537683129436, "grad_norm": 1.9035314762902331, "learning_rate": 7.661188622078555e-06, "loss": 0.1233, "step": 6383 }, { "epoch": 0.9642048028998641, "grad_norm": 1.935180675196995, "learning_rate": 7.660519088602763e-06, "loss": 0.1423, "step": 6384 }, { "epoch": 0.9643558374867844, "grad_norm": 1.8996750550491048, "learning_rate": 7.659849488572622e-06, "loss": 0.1421, "step": 6385 }, { "epoch": 0.9645068720737049, "grad_norm": 1.876940077001697, "learning_rate": 7.659179822004884e-06, "loss": 0.1371, "step": 6386 }, { "epoch": 0.9646579066606252, "grad_norm": 1.6782357337904659, "learning_rate": 7.6585100889163e-06, "loss": 0.131, "step": 6387 }, { "epoch": 0.9648089412475457, "grad_norm": 1.9691951311670015, "learning_rate": 7.657840289323621e-06, "loss": 0.1166, "step": 6388 }, { "epoch": 0.9649599758344661, "grad_norm": 1.5600172355727526, "learning_rate": 7.657170423243607e-06, "loss": 0.106, "step": 6389 }, { "epoch": 0.9651110104213865, "grad_norm": 2.263126835448898, "learning_rate": 7.656500490693014e-06, "loss": 0.1243, "step": 6390 }, { "epoch": 0.9652620450083069, "grad_norm": 2.1074757348098943, "learning_rate": 7.655830491688596e-06, "loss": 0.1513, "step": 6391 }, { "epoch": 0.9654130795952273, "grad_norm": 2.805461299165616, "learning_rate": 7.65516042624712e-06, "loss": 0.1652, "step": 6392 }, { "epoch": 0.9655641141821477, "grad_norm": 2.0330191225944603, "learning_rate": 7.654490294385344e-06, "loss": 0.1422, "step": 6393 }, { "epoch": 0.9657151487690682, "grad_norm": 2.2489346961694845, "learning_rate": 7.653820096120033e-06, "loss": 0.1537, "step": 6394 }, { "epoch": 0.9658661833559885, "grad_norm": 3.6503534364269594, "learning_rate": 7.653149831467951e-06, "loss": 0.1528, "step": 6395 }, { "epoch": 0.966017217942909, "grad_norm": 2.0311129680713518, "learning_rate": 7.652479500445869e-06, "loss": 0.1508, "step": 6396 }, { "epoch": 0.9661682525298293, "grad_norm": 2.5173099392329803, "learning_rate": 7.65180910307055e-06, "loss": 0.1495, "step": 6397 }, { "epoch": 0.9663192871167497, "grad_norm": 1.9278445345400643, "learning_rate": 7.651138639358769e-06, "loss": 0.1413, "step": 6398 }, { "epoch": 0.9664703217036701, "grad_norm": 3.3454294430920575, "learning_rate": 7.650468109327296e-06, "loss": 0.1329, "step": 6399 }, { "epoch": 0.9666213562905905, "grad_norm": 2.2999484833385693, "learning_rate": 7.649797512992903e-06, "loss": 0.127, "step": 6400 }, { "epoch": 0.966772390877511, "grad_norm": 2.3504749121101183, "learning_rate": 7.64912685037237e-06, "loss": 0.147, "step": 6401 }, { "epoch": 0.9669234254644313, "grad_norm": 2.2817730789983726, "learning_rate": 7.648456121482471e-06, "loss": 0.125, "step": 6402 }, { "epoch": 0.9670744600513518, "grad_norm": 1.6450194316761741, "learning_rate": 7.647785326339984e-06, "loss": 0.1213, "step": 6403 }, { "epoch": 0.9672254946382721, "grad_norm": 1.609480902226146, "learning_rate": 7.647114464961689e-06, "loss": 0.1344, "step": 6404 }, { "epoch": 0.9673765292251926, "grad_norm": 2.4048846959114107, "learning_rate": 7.646443537364372e-06, "loss": 0.1614, "step": 6405 }, { "epoch": 0.967527563812113, "grad_norm": 2.2717076425752625, "learning_rate": 7.645772543564814e-06, "loss": 0.189, "step": 6406 }, { "epoch": 0.9676785983990334, "grad_norm": 1.6610808782714037, "learning_rate": 7.645101483579797e-06, "loss": 0.1182, "step": 6407 }, { "epoch": 0.9678296329859538, "grad_norm": 1.987433614138115, "learning_rate": 7.644430357426114e-06, "loss": 0.1625, "step": 6408 }, { "epoch": 0.9679806675728742, "grad_norm": 1.8937551583080432, "learning_rate": 7.643759165120548e-06, "loss": 0.1468, "step": 6409 }, { "epoch": 0.9681317021597946, "grad_norm": 2.04850883185278, "learning_rate": 7.643087906679894e-06, "loss": 0.1449, "step": 6410 }, { "epoch": 0.968282736746715, "grad_norm": 2.107153301016087, "learning_rate": 7.64241658212094e-06, "loss": 0.1349, "step": 6411 }, { "epoch": 0.9684337713336354, "grad_norm": 2.055046145207259, "learning_rate": 7.641745191460483e-06, "loss": 0.1527, "step": 6412 }, { "epoch": 0.9685848059205558, "grad_norm": 2.1549921893192323, "learning_rate": 7.641073734715316e-06, "loss": 0.1333, "step": 6413 }, { "epoch": 0.9687358405074762, "grad_norm": 2.372999370927327, "learning_rate": 7.640402211902237e-06, "loss": 0.1543, "step": 6414 }, { "epoch": 0.9688868750943966, "grad_norm": 2.3477511687974966, "learning_rate": 7.639730623038042e-06, "loss": 0.1786, "step": 6415 }, { "epoch": 0.969037909681317, "grad_norm": 2.420449095129016, "learning_rate": 7.639058968139536e-06, "loss": 0.1447, "step": 6416 }, { "epoch": 0.9691889442682374, "grad_norm": 2.307428799553186, "learning_rate": 7.638387247223515e-06, "loss": 0.1405, "step": 6417 }, { "epoch": 0.9693399788551579, "grad_norm": 2.860837065607224, "learning_rate": 7.637715460306786e-06, "loss": 0.1632, "step": 6418 }, { "epoch": 0.9694910134420782, "grad_norm": 1.8257554911270184, "learning_rate": 7.637043607406157e-06, "loss": 0.1421, "step": 6419 }, { "epoch": 0.9696420480289987, "grad_norm": 1.939936636829656, "learning_rate": 7.636371688538427e-06, "loss": 0.1576, "step": 6420 }, { "epoch": 0.969793082615919, "grad_norm": 1.9354168295872634, "learning_rate": 7.63569970372041e-06, "loss": 0.1449, "step": 6421 }, { "epoch": 0.9699441172028395, "grad_norm": 1.4568336254124943, "learning_rate": 7.635027652968916e-06, "loss": 0.113, "step": 6422 }, { "epoch": 0.9700951517897598, "grad_norm": 2.06676671912117, "learning_rate": 7.634355536300755e-06, "loss": 0.1286, "step": 6423 }, { "epoch": 0.9702461863766803, "grad_norm": 2.315327932979462, "learning_rate": 7.633683353732739e-06, "loss": 0.1723, "step": 6424 }, { "epoch": 0.9703972209636007, "grad_norm": 1.8999907333621553, "learning_rate": 7.633011105281687e-06, "loss": 0.1383, "step": 6425 }, { "epoch": 0.970548255550521, "grad_norm": 1.545491550801559, "learning_rate": 7.632338790964413e-06, "loss": 0.1183, "step": 6426 }, { "epoch": 0.9706992901374415, "grad_norm": 1.6247813649316567, "learning_rate": 7.631666410797734e-06, "loss": 0.1302, "step": 6427 }, { "epoch": 0.9708503247243618, "grad_norm": 2.4563410340120004, "learning_rate": 7.630993964798475e-06, "loss": 0.1579, "step": 6428 }, { "epoch": 0.9710013593112823, "grad_norm": 2.0618651000507455, "learning_rate": 7.630321452983452e-06, "loss": 0.1316, "step": 6429 }, { "epoch": 0.9711523938982027, "grad_norm": 1.7608787099166876, "learning_rate": 7.629648875369491e-06, "loss": 0.139, "step": 6430 }, { "epoch": 0.9713034284851231, "grad_norm": 2.155683886432801, "learning_rate": 7.628976231973417e-06, "loss": 0.1388, "step": 6431 }, { "epoch": 0.9714544630720435, "grad_norm": 1.9980489329011002, "learning_rate": 7.628303522812055e-06, "loss": 0.1466, "step": 6432 }, { "epoch": 0.9716054976589639, "grad_norm": 1.8213571875837133, "learning_rate": 7.627630747902236e-06, "loss": 0.161, "step": 6433 }, { "epoch": 0.9717565322458843, "grad_norm": 1.4347148201662756, "learning_rate": 7.626957907260788e-06, "loss": 0.1166, "step": 6434 }, { "epoch": 0.9719075668328047, "grad_norm": 2.80790432765619, "learning_rate": 7.626285000904542e-06, "loss": 0.1643, "step": 6435 }, { "epoch": 0.9720586014197251, "grad_norm": 1.6810667834885848, "learning_rate": 7.625612028850331e-06, "loss": 0.1187, "step": 6436 }, { "epoch": 0.9722096360066456, "grad_norm": 1.891620107910592, "learning_rate": 7.624938991114992e-06, "loss": 0.1175, "step": 6437 }, { "epoch": 0.9723606705935659, "grad_norm": 2.010236565665505, "learning_rate": 7.624265887715359e-06, "loss": 0.1273, "step": 6438 }, { "epoch": 0.9725117051804864, "grad_norm": 2.448600465064593, "learning_rate": 7.623592718668271e-06, "loss": 0.1709, "step": 6439 }, { "epoch": 0.9726627397674067, "grad_norm": 2.290870606954682, "learning_rate": 7.622919483990568e-06, "loss": 0.1383, "step": 6440 }, { "epoch": 0.9728137743543271, "grad_norm": 2.2343256802092597, "learning_rate": 7.622246183699091e-06, "loss": 0.1458, "step": 6441 }, { "epoch": 0.9729648089412476, "grad_norm": 2.273199740141112, "learning_rate": 7.6215728178106825e-06, "loss": 0.1407, "step": 6442 }, { "epoch": 0.9731158435281679, "grad_norm": 2.2193035142764277, "learning_rate": 7.620899386342188e-06, "loss": 0.1274, "step": 6443 }, { "epoch": 0.9732668781150884, "grad_norm": 2.1520572075589883, "learning_rate": 7.620225889310453e-06, "loss": 0.1529, "step": 6444 }, { "epoch": 0.9734179127020087, "grad_norm": 5.490255610832153, "learning_rate": 7.619552326732327e-06, "loss": 0.1548, "step": 6445 }, { "epoch": 0.9735689472889292, "grad_norm": 2.190143793097073, "learning_rate": 7.618878698624657e-06, "loss": 0.1711, "step": 6446 }, { "epoch": 0.9737199818758495, "grad_norm": 2.7536668237467814, "learning_rate": 7.618205005004296e-06, "loss": 0.1625, "step": 6447 }, { "epoch": 0.97387101646277, "grad_norm": 1.8083170818109924, "learning_rate": 7.617531245888095e-06, "loss": 0.1451, "step": 6448 }, { "epoch": 0.9740220510496904, "grad_norm": 1.9268528228760362, "learning_rate": 7.61685742129291e-06, "loss": 0.1384, "step": 6449 }, { "epoch": 0.9741730856366108, "grad_norm": 1.5658715379431078, "learning_rate": 7.616183531235597e-06, "loss": 0.1171, "step": 6450 }, { "epoch": 0.9743241202235312, "grad_norm": 1.780862910742839, "learning_rate": 7.6155095757330146e-06, "loss": 0.1359, "step": 6451 }, { "epoch": 0.9744751548104515, "grad_norm": 2.320675547142392, "learning_rate": 7.614835554802021e-06, "loss": 0.145, "step": 6452 }, { "epoch": 0.974626189397372, "grad_norm": 2.040022428826484, "learning_rate": 7.6141614684594775e-06, "loss": 0.131, "step": 6453 }, { "epoch": 0.9747772239842925, "grad_norm": 2.7420104702222274, "learning_rate": 7.613487316722246e-06, "loss": 0.1728, "step": 6454 }, { "epoch": 0.9749282585712128, "grad_norm": 1.8279873619055536, "learning_rate": 7.61281309960719e-06, "loss": 0.131, "step": 6455 }, { "epoch": 0.9750792931581332, "grad_norm": 1.6479252057393288, "learning_rate": 7.612138817131178e-06, "loss": 0.1252, "step": 6456 }, { "epoch": 0.9752303277450536, "grad_norm": 3.086149354017242, "learning_rate": 7.611464469311076e-06, "loss": 0.1972, "step": 6457 }, { "epoch": 0.975381362331974, "grad_norm": 2.2066274368102103, "learning_rate": 7.610790056163755e-06, "loss": 0.1635, "step": 6458 }, { "epoch": 0.9755323969188944, "grad_norm": 3.550711644362843, "learning_rate": 7.610115577706083e-06, "loss": 0.1629, "step": 6459 }, { "epoch": 0.9756834315058148, "grad_norm": 1.9541987688077345, "learning_rate": 7.609441033954933e-06, "loss": 0.1335, "step": 6460 }, { "epoch": 0.9758344660927353, "grad_norm": 1.7814012168726836, "learning_rate": 7.608766424927181e-06, "loss": 0.1326, "step": 6461 }, { "epoch": 0.9759855006796556, "grad_norm": 2.184520463505056, "learning_rate": 7.608091750639701e-06, "loss": 0.1522, "step": 6462 }, { "epoch": 0.9761365352665761, "grad_norm": 1.617031611099065, "learning_rate": 7.607417011109372e-06, "loss": 0.1173, "step": 6463 }, { "epoch": 0.9762875698534964, "grad_norm": 2.102484297293465, "learning_rate": 7.606742206353071e-06, "loss": 0.1283, "step": 6464 }, { "epoch": 0.9764386044404169, "grad_norm": 1.7960902958531362, "learning_rate": 7.606067336387679e-06, "loss": 0.1469, "step": 6465 }, { "epoch": 0.9765896390273373, "grad_norm": 2.1462087110556842, "learning_rate": 7.605392401230079e-06, "loss": 0.1339, "step": 6466 }, { "epoch": 0.9767406736142576, "grad_norm": 2.3669286847341966, "learning_rate": 7.604717400897155e-06, "loss": 0.1646, "step": 6467 }, { "epoch": 0.9768917082011781, "grad_norm": 1.5775880771995703, "learning_rate": 7.604042335405793e-06, "loss": 0.1187, "step": 6468 }, { "epoch": 0.9770427427880984, "grad_norm": 1.738023753053049, "learning_rate": 7.603367204772877e-06, "loss": 0.1344, "step": 6469 }, { "epoch": 0.9771937773750189, "grad_norm": 2.0546625324545236, "learning_rate": 7.6026920090153e-06, "loss": 0.1472, "step": 6470 }, { "epoch": 0.9773448119619392, "grad_norm": 1.8621312995443124, "learning_rate": 7.602016748149947e-06, "loss": 0.1396, "step": 6471 }, { "epoch": 0.9774958465488597, "grad_norm": 1.8992981598877126, "learning_rate": 7.601341422193717e-06, "loss": 0.128, "step": 6472 }, { "epoch": 0.9776468811357801, "grad_norm": 2.1868475290543943, "learning_rate": 7.6006660311634994e-06, "loss": 0.1526, "step": 6473 }, { "epoch": 0.9777979157227005, "grad_norm": 2.6126862449617385, "learning_rate": 7.59999057507619e-06, "loss": 0.1644, "step": 6474 }, { "epoch": 0.9779489503096209, "grad_norm": 2.1374331851041646, "learning_rate": 7.599315053948686e-06, "loss": 0.1443, "step": 6475 }, { "epoch": 0.9780999848965413, "grad_norm": 2.513863332298229, "learning_rate": 7.598639467797886e-06, "loss": 0.1482, "step": 6476 }, { "epoch": 0.9782510194834617, "grad_norm": 2.0918347691725194, "learning_rate": 7.597963816640688e-06, "loss": 0.1373, "step": 6477 }, { "epoch": 0.9784020540703822, "grad_norm": 3.7175906794262223, "learning_rate": 7.597288100493998e-06, "loss": 0.1501, "step": 6478 }, { "epoch": 0.9785530886573025, "grad_norm": 2.093543150721899, "learning_rate": 7.596612319374715e-06, "loss": 0.1467, "step": 6479 }, { "epoch": 0.978704123244223, "grad_norm": 2.09785024756247, "learning_rate": 7.59593647329975e-06, "loss": 0.1469, "step": 6480 }, { "epoch": 0.9788551578311433, "grad_norm": 1.958441461927899, "learning_rate": 7.595260562286002e-06, "loss": 0.1425, "step": 6481 }, { "epoch": 0.9790061924180637, "grad_norm": 2.0165151550669442, "learning_rate": 7.594584586350384e-06, "loss": 0.1404, "step": 6482 }, { "epoch": 0.9791572270049841, "grad_norm": 2.3323784168714092, "learning_rate": 7.593908545509805e-06, "loss": 0.1642, "step": 6483 }, { "epoch": 0.9793082615919045, "grad_norm": 2.09787624563458, "learning_rate": 7.593232439781177e-06, "loss": 0.1584, "step": 6484 }, { "epoch": 0.979459296178825, "grad_norm": 2.267166071629668, "learning_rate": 7.592556269181414e-06, "loss": 0.1517, "step": 6485 }, { "epoch": 0.9796103307657453, "grad_norm": 1.951434798218732, "learning_rate": 7.591880033727428e-06, "loss": 0.1296, "step": 6486 }, { "epoch": 0.9797613653526658, "grad_norm": 1.821511965073111, "learning_rate": 7.5912037334361366e-06, "loss": 0.1541, "step": 6487 }, { "epoch": 0.9799123999395861, "grad_norm": 2.634931964476399, "learning_rate": 7.590527368324458e-06, "loss": 0.1575, "step": 6488 }, { "epoch": 0.9800634345265066, "grad_norm": 2.02170871504368, "learning_rate": 7.589850938409313e-06, "loss": 0.134, "step": 6489 }, { "epoch": 0.980214469113427, "grad_norm": 1.679697086098546, "learning_rate": 7.589174443707622e-06, "loss": 0.1314, "step": 6490 }, { "epoch": 0.9803655037003474, "grad_norm": 2.3007647297349445, "learning_rate": 7.588497884236308e-06, "loss": 0.1476, "step": 6491 }, { "epoch": 0.9805165382872678, "grad_norm": 1.6704630318226292, "learning_rate": 7.587821260012295e-06, "loss": 0.1341, "step": 6492 }, { "epoch": 0.9806675728741882, "grad_norm": 1.8515235883074137, "learning_rate": 7.587144571052509e-06, "loss": 0.1293, "step": 6493 }, { "epoch": 0.9808186074611086, "grad_norm": 1.6400391274660642, "learning_rate": 7.586467817373877e-06, "loss": 0.1311, "step": 6494 }, { "epoch": 0.980969642048029, "grad_norm": 1.9255211972195416, "learning_rate": 7.585790998993331e-06, "loss": 0.1281, "step": 6495 }, { "epoch": 0.9811206766349494, "grad_norm": 2.0061641862460884, "learning_rate": 7.585114115927799e-06, "loss": 0.1338, "step": 6496 }, { "epoch": 0.9812717112218698, "grad_norm": 3.013326291331573, "learning_rate": 7.584437168194216e-06, "loss": 0.1568, "step": 6497 }, { "epoch": 0.9814227458087902, "grad_norm": 2.4250359340242915, "learning_rate": 7.583760155809517e-06, "loss": 0.1475, "step": 6498 }, { "epoch": 0.9815737803957106, "grad_norm": 2.518232583227365, "learning_rate": 7.583083078790634e-06, "loss": 0.1891, "step": 6499 }, { "epoch": 0.981724814982631, "grad_norm": 2.3762999540459337, "learning_rate": 7.582405937154507e-06, "loss": 0.1536, "step": 6500 }, { "epoch": 0.9818758495695514, "grad_norm": 2.3849681045156146, "learning_rate": 7.581728730918074e-06, "loss": 0.1662, "step": 6501 }, { "epoch": 0.9820268841564719, "grad_norm": 1.7040039622021925, "learning_rate": 7.581051460098278e-06, "loss": 0.1188, "step": 6502 }, { "epoch": 0.9821779187433922, "grad_norm": 1.9024031878815892, "learning_rate": 7.5803741247120574e-06, "loss": 0.1437, "step": 6503 }, { "epoch": 0.9823289533303127, "grad_norm": 1.8471611790642857, "learning_rate": 7.57969672477636e-06, "loss": 0.1204, "step": 6504 }, { "epoch": 0.982479987917233, "grad_norm": 1.5778570312052305, "learning_rate": 7.579019260308129e-06, "loss": 0.1271, "step": 6505 }, { "epoch": 0.9826310225041535, "grad_norm": 1.7279831330320679, "learning_rate": 7.5783417313243135e-06, "loss": 0.1308, "step": 6506 }, { "epoch": 0.9827820570910738, "grad_norm": 1.775193010203476, "learning_rate": 7.57766413784186e-06, "loss": 0.1379, "step": 6507 }, { "epoch": 0.9829330916779943, "grad_norm": 1.9623066795425579, "learning_rate": 7.5769864798777195e-06, "loss": 0.149, "step": 6508 }, { "epoch": 0.9830841262649147, "grad_norm": 2.1657913618912383, "learning_rate": 7.576308757448846e-06, "loss": 0.1634, "step": 6509 }, { "epoch": 0.983235160851835, "grad_norm": 2.3997278281099703, "learning_rate": 7.57563097057219e-06, "loss": 0.1378, "step": 6510 }, { "epoch": 0.9833861954387555, "grad_norm": 1.9378868081108822, "learning_rate": 7.574953119264707e-06, "loss": 0.1458, "step": 6511 }, { "epoch": 0.9835372300256758, "grad_norm": 1.6981874569031314, "learning_rate": 7.574275203543357e-06, "loss": 0.119, "step": 6512 }, { "epoch": 0.9836882646125963, "grad_norm": 1.7096922647937598, "learning_rate": 7.573597223425096e-06, "loss": 0.1396, "step": 6513 }, { "epoch": 0.9838392991995167, "grad_norm": 2.118246889362941, "learning_rate": 7.572919178926884e-06, "loss": 0.1582, "step": 6514 }, { "epoch": 0.9839903337864371, "grad_norm": 2.4906062909061832, "learning_rate": 7.572241070065683e-06, "loss": 0.1367, "step": 6515 }, { "epoch": 0.9841413683733575, "grad_norm": 1.6139451895834902, "learning_rate": 7.5715628968584575e-06, "loss": 0.1098, "step": 6516 }, { "epoch": 0.9842924029602779, "grad_norm": 1.7988713308577606, "learning_rate": 7.570884659322171e-06, "loss": 0.1201, "step": 6517 }, { "epoch": 0.9844434375471983, "grad_norm": 2.714129080765319, "learning_rate": 7.570206357473788e-06, "loss": 0.1359, "step": 6518 }, { "epoch": 0.9845944721341187, "grad_norm": 2.046845798764715, "learning_rate": 7.569527991330281e-06, "loss": 0.1374, "step": 6519 }, { "epoch": 0.9847455067210391, "grad_norm": 2.347771790285904, "learning_rate": 7.568849560908619e-06, "loss": 0.1607, "step": 6520 }, { "epoch": 0.9848965413079596, "grad_norm": 2.5180807395754132, "learning_rate": 7.568171066225769e-06, "loss": 0.1577, "step": 6521 }, { "epoch": 0.9850475758948799, "grad_norm": 1.851793631552283, "learning_rate": 7.567492507298708e-06, "loss": 0.1214, "step": 6522 }, { "epoch": 0.9851986104818004, "grad_norm": 3.032925249534164, "learning_rate": 7.566813884144408e-06, "loss": 0.1543, "step": 6523 }, { "epoch": 0.9853496450687207, "grad_norm": 2.2964375221337474, "learning_rate": 7.566135196779849e-06, "loss": 0.1497, "step": 6524 }, { "epoch": 0.9855006796556411, "grad_norm": 2.189108221636898, "learning_rate": 7.565456445222005e-06, "loss": 0.1631, "step": 6525 }, { "epoch": 0.9856517142425616, "grad_norm": 1.937804334372778, "learning_rate": 7.564777629487856e-06, "loss": 0.1473, "step": 6526 }, { "epoch": 0.9858027488294819, "grad_norm": 1.9879446319398302, "learning_rate": 7.564098749594384e-06, "loss": 0.15, "step": 6527 }, { "epoch": 0.9859537834164024, "grad_norm": 2.092107685417178, "learning_rate": 7.563419805558571e-06, "loss": 0.1492, "step": 6528 }, { "epoch": 0.9861048180033227, "grad_norm": 1.9491440828251787, "learning_rate": 7.562740797397401e-06, "loss": 0.1316, "step": 6529 }, { "epoch": 0.9862558525902432, "grad_norm": 1.9838486079041353, "learning_rate": 7.562061725127858e-06, "loss": 0.154, "step": 6530 }, { "epoch": 0.9864068871771635, "grad_norm": 2.5282034062431147, "learning_rate": 7.561382588766934e-06, "loss": 0.1765, "step": 6531 }, { "epoch": 0.986557921764084, "grad_norm": 1.7036511137653674, "learning_rate": 7.560703388331616e-06, "loss": 0.1252, "step": 6532 }, { "epoch": 0.9867089563510044, "grad_norm": 1.8769336183876357, "learning_rate": 7.560024123838891e-06, "loss": 0.1115, "step": 6533 }, { "epoch": 0.9868599909379248, "grad_norm": 1.7402292088027034, "learning_rate": 7.559344795305757e-06, "loss": 0.1288, "step": 6534 }, { "epoch": 0.9870110255248452, "grad_norm": 2.530277499896953, "learning_rate": 7.558665402749202e-06, "loss": 0.1532, "step": 6535 }, { "epoch": 0.9871620601117655, "grad_norm": 2.9860333667983148, "learning_rate": 7.557985946186226e-06, "loss": 0.1561, "step": 6536 }, { "epoch": 0.987313094698686, "grad_norm": 2.295911358219224, "learning_rate": 7.557306425633823e-06, "loss": 0.1554, "step": 6537 }, { "epoch": 0.9874641292856065, "grad_norm": 2.1262411229814084, "learning_rate": 7.556626841108994e-06, "loss": 0.1505, "step": 6538 }, { "epoch": 0.9876151638725268, "grad_norm": 2.1823218897806855, "learning_rate": 7.5559471926287365e-06, "loss": 0.1457, "step": 6539 }, { "epoch": 0.9877661984594472, "grad_norm": 1.710463421604133, "learning_rate": 7.555267480210053e-06, "loss": 0.1329, "step": 6540 }, { "epoch": 0.9879172330463676, "grad_norm": 1.4621898124037154, "learning_rate": 7.554587703869949e-06, "loss": 0.1065, "step": 6541 }, { "epoch": 0.988068267633288, "grad_norm": 2.9038630676911406, "learning_rate": 7.55390786362543e-06, "loss": 0.1734, "step": 6542 }, { "epoch": 0.9882193022202084, "grad_norm": 2.0536805236224436, "learning_rate": 7.553227959493499e-06, "loss": 0.1466, "step": 6543 }, { "epoch": 0.9883703368071288, "grad_norm": 2.239498355320195, "learning_rate": 7.552547991491163e-06, "loss": 0.1711, "step": 6544 }, { "epoch": 0.9885213713940493, "grad_norm": 2.0867075931556176, "learning_rate": 7.551867959635438e-06, "loss": 0.1175, "step": 6545 }, { "epoch": 0.9886724059809696, "grad_norm": 2.3717080113513473, "learning_rate": 7.551187863943331e-06, "loss": 0.1524, "step": 6546 }, { "epoch": 0.9888234405678901, "grad_norm": 1.746706027331134, "learning_rate": 7.550507704431856e-06, "loss": 0.1149, "step": 6547 }, { "epoch": 0.9889744751548104, "grad_norm": 1.5300534309468152, "learning_rate": 7.549827481118028e-06, "loss": 0.1059, "step": 6548 }, { "epoch": 0.9891255097417309, "grad_norm": 2.1780419713471106, "learning_rate": 7.549147194018862e-06, "loss": 0.1553, "step": 6549 }, { "epoch": 0.9892765443286513, "grad_norm": 1.7577223330340985, "learning_rate": 7.5484668431513766e-06, "loss": 0.1362, "step": 6550 }, { "epoch": 0.9894275789155716, "grad_norm": 1.6389590077588316, "learning_rate": 7.547786428532591e-06, "loss": 0.1229, "step": 6551 }, { "epoch": 0.9895786135024921, "grad_norm": 1.6857137420609747, "learning_rate": 7.547105950179525e-06, "loss": 0.1467, "step": 6552 }, { "epoch": 0.9897296480894124, "grad_norm": 2.048493179045975, "learning_rate": 7.546425408109204e-06, "loss": 0.1325, "step": 6553 }, { "epoch": 0.9898806826763329, "grad_norm": 1.8948129918969414, "learning_rate": 7.545744802338649e-06, "loss": 0.1257, "step": 6554 }, { "epoch": 0.9900317172632532, "grad_norm": 1.8667699229626542, "learning_rate": 7.5450641328848884e-06, "loss": 0.1315, "step": 6555 }, { "epoch": 0.9901827518501737, "grad_norm": 2.6935319488309917, "learning_rate": 7.5443833997649476e-06, "loss": 0.1304, "step": 6556 }, { "epoch": 0.9903337864370941, "grad_norm": 1.6969383089785814, "learning_rate": 7.543702602995855e-06, "loss": 0.1103, "step": 6557 }, { "epoch": 0.9904848210240145, "grad_norm": 1.8626149413267787, "learning_rate": 7.543021742594644e-06, "loss": 0.1413, "step": 6558 }, { "epoch": 0.9906358556109349, "grad_norm": 2.1125650267062097, "learning_rate": 7.542340818578344e-06, "loss": 0.1552, "step": 6559 }, { "epoch": 0.9907868901978553, "grad_norm": 1.3120810521394335, "learning_rate": 7.541659830963991e-06, "loss": 0.0993, "step": 6560 }, { "epoch": 0.9909379247847757, "grad_norm": 2.0283540859418787, "learning_rate": 7.540978779768618e-06, "loss": 0.1405, "step": 6561 }, { "epoch": 0.9910889593716962, "grad_norm": 1.6920678249196834, "learning_rate": 7.5402976650092605e-06, "loss": 0.1232, "step": 6562 }, { "epoch": 0.9912399939586165, "grad_norm": 1.8340182313401843, "learning_rate": 7.539616486702963e-06, "loss": 0.1374, "step": 6563 }, { "epoch": 0.991391028545537, "grad_norm": 2.1028924758537237, "learning_rate": 7.53893524486676e-06, "loss": 0.1293, "step": 6564 }, { "epoch": 0.9915420631324573, "grad_norm": 1.8250304286325185, "learning_rate": 7.538253939517695e-06, "loss": 0.1346, "step": 6565 }, { "epoch": 0.9916930977193777, "grad_norm": 1.8441982569716857, "learning_rate": 7.537572570672811e-06, "loss": 0.1234, "step": 6566 }, { "epoch": 0.9918441323062981, "grad_norm": 2.155404749054103, "learning_rate": 7.536891138349154e-06, "loss": 0.121, "step": 6567 }, { "epoch": 0.9919951668932185, "grad_norm": 1.668877612947988, "learning_rate": 7.536209642563767e-06, "loss": 0.1255, "step": 6568 }, { "epoch": 0.992146201480139, "grad_norm": 2.101837233951652, "learning_rate": 7.535528083333703e-06, "loss": 0.1416, "step": 6569 }, { "epoch": 0.9922972360670593, "grad_norm": 1.9767332790974332, "learning_rate": 7.534846460676008e-06, "loss": 0.1275, "step": 6570 }, { "epoch": 0.9924482706539798, "grad_norm": 2.1148520301243465, "learning_rate": 7.534164774607735e-06, "loss": 0.1309, "step": 6571 }, { "epoch": 0.9925993052409001, "grad_norm": 3.597785027544572, "learning_rate": 7.533483025145933e-06, "loss": 0.1209, "step": 6572 }, { "epoch": 0.9927503398278206, "grad_norm": 2.4284872747079556, "learning_rate": 7.5328012123076625e-06, "loss": 0.1456, "step": 6573 }, { "epoch": 0.992901374414741, "grad_norm": 1.8048637236299343, "learning_rate": 7.532119336109975e-06, "loss": 0.1261, "step": 6574 }, { "epoch": 0.9930524090016614, "grad_norm": 2.5998249529585085, "learning_rate": 7.531437396569928e-06, "loss": 0.1546, "step": 6575 }, { "epoch": 0.9932034435885818, "grad_norm": 2.5328267763902934, "learning_rate": 7.530755393704583e-06, "loss": 0.1602, "step": 6576 }, { "epoch": 0.9933544781755022, "grad_norm": 2.0062628671816274, "learning_rate": 7.530073327530999e-06, "loss": 0.122, "step": 6577 }, { "epoch": 0.9935055127624226, "grad_norm": 2.364559267563497, "learning_rate": 7.5293911980662396e-06, "loss": 0.1387, "step": 6578 }, { "epoch": 0.993656547349343, "grad_norm": 1.8752638080774098, "learning_rate": 7.528709005327366e-06, "loss": 0.1616, "step": 6579 }, { "epoch": 0.9938075819362634, "grad_norm": 2.2165011224914766, "learning_rate": 7.528026749331445e-06, "loss": 0.1327, "step": 6580 }, { "epoch": 0.9939586165231838, "grad_norm": 1.9713276526870593, "learning_rate": 7.5273444300955465e-06, "loss": 0.1338, "step": 6581 }, { "epoch": 0.9941096511101042, "grad_norm": 1.587137992115594, "learning_rate": 7.5266620476367345e-06, "loss": 0.1286, "step": 6582 }, { "epoch": 0.9942606856970246, "grad_norm": 3.3136640516294587, "learning_rate": 7.525979601972082e-06, "loss": 0.1672, "step": 6583 }, { "epoch": 0.994411720283945, "grad_norm": 1.7854383701160537, "learning_rate": 7.52529709311866e-06, "loss": 0.1245, "step": 6584 }, { "epoch": 0.9945627548708654, "grad_norm": 2.368786803142681, "learning_rate": 7.5246145210935425e-06, "loss": 0.1447, "step": 6585 }, { "epoch": 0.9947137894577859, "grad_norm": 2.772631945173343, "learning_rate": 7.5239318859138025e-06, "loss": 0.1465, "step": 6586 }, { "epoch": 0.9948648240447062, "grad_norm": 1.7551732595512874, "learning_rate": 7.523249187596517e-06, "loss": 0.1448, "step": 6587 }, { "epoch": 0.9950158586316267, "grad_norm": 2.1641396101809613, "learning_rate": 7.5225664261587684e-06, "loss": 0.1714, "step": 6588 }, { "epoch": 0.995166893218547, "grad_norm": 1.6055486004502888, "learning_rate": 7.52188360161763e-06, "loss": 0.1182, "step": 6589 }, { "epoch": 0.9953179278054675, "grad_norm": 2.554572773903796, "learning_rate": 7.5212007139901845e-06, "loss": 0.1373, "step": 6590 }, { "epoch": 0.9954689623923878, "grad_norm": 1.8040188276935027, "learning_rate": 7.520517763293518e-06, "loss": 0.1247, "step": 6591 }, { "epoch": 0.9956199969793083, "grad_norm": 3.0320109425529806, "learning_rate": 7.519834749544712e-06, "loss": 0.1497, "step": 6592 }, { "epoch": 0.9957710315662287, "grad_norm": 1.734794554178413, "learning_rate": 7.519151672760854e-06, "loss": 0.1107, "step": 6593 }, { "epoch": 0.995922066153149, "grad_norm": 1.9115369499037855, "learning_rate": 7.518468532959031e-06, "loss": 0.137, "step": 6594 }, { "epoch": 0.9960731007400695, "grad_norm": 2.058956161076635, "learning_rate": 7.517785330156331e-06, "loss": 0.1562, "step": 6595 }, { "epoch": 0.9962241353269898, "grad_norm": 1.7494902776572214, "learning_rate": 7.517102064369847e-06, "loss": 0.1328, "step": 6596 }, { "epoch": 0.9963751699139103, "grad_norm": 2.2673407380067956, "learning_rate": 7.516418735616668e-06, "loss": 0.1715, "step": 6597 }, { "epoch": 0.9965262045008307, "grad_norm": 2.0385674236963376, "learning_rate": 7.5157353439138925e-06, "loss": 0.1391, "step": 6598 }, { "epoch": 0.9966772390877511, "grad_norm": 2.0561524096340555, "learning_rate": 7.515051889278612e-06, "loss": 0.1583, "step": 6599 }, { "epoch": 0.9968282736746715, "grad_norm": 2.258890428039315, "learning_rate": 7.514368371727925e-06, "loss": 0.1353, "step": 6600 }, { "epoch": 0.9969793082615919, "grad_norm": 2.759254821914357, "learning_rate": 7.513684791278928e-06, "loss": 0.1588, "step": 6601 }, { "epoch": 0.9971303428485123, "grad_norm": 2.040217968428254, "learning_rate": 7.513001147948724e-06, "loss": 0.1254, "step": 6602 }, { "epoch": 0.9972813774354327, "grad_norm": 1.719989758588653, "learning_rate": 7.512317441754414e-06, "loss": 0.1267, "step": 6603 }, { "epoch": 0.9974324120223531, "grad_norm": 2.289540221596505, "learning_rate": 7.511633672713101e-06, "loss": 0.1481, "step": 6604 }, { "epoch": 0.9975834466092736, "grad_norm": 2.2577331971202597, "learning_rate": 7.5109498408418905e-06, "loss": 0.1537, "step": 6605 }, { "epoch": 0.9977344811961939, "grad_norm": 2.1073144218638613, "learning_rate": 7.510265946157887e-06, "loss": 0.1607, "step": 6606 }, { "epoch": 0.9978855157831144, "grad_norm": 1.6093343786500012, "learning_rate": 7.5095819886781985e-06, "loss": 0.1237, "step": 6607 }, { "epoch": 0.9980365503700347, "grad_norm": 1.9820242040876679, "learning_rate": 7.508897968419937e-06, "loss": 0.12, "step": 6608 }, { "epoch": 0.9981875849569551, "grad_norm": 2.9073089667576837, "learning_rate": 7.508213885400213e-06, "loss": 0.1586, "step": 6609 }, { "epoch": 0.9983386195438756, "grad_norm": 6.6106037320836535, "learning_rate": 7.507529739636139e-06, "loss": 0.1614, "step": 6610 }, { "epoch": 0.9984896541307959, "grad_norm": 1.8884130497204712, "learning_rate": 7.506845531144828e-06, "loss": 0.1441, "step": 6611 }, { "epoch": 0.9986406887177164, "grad_norm": 1.9373955049667715, "learning_rate": 7.506161259943398e-06, "loss": 0.145, "step": 6612 }, { "epoch": 0.9987917233046367, "grad_norm": 1.8646167935850462, "learning_rate": 7.505476926048964e-06, "loss": 0.1385, "step": 6613 }, { "epoch": 0.9989427578915572, "grad_norm": 1.8832799757221395, "learning_rate": 7.5047925294786465e-06, "loss": 0.138, "step": 6614 }, { "epoch": 0.9990937924784775, "grad_norm": 2.3898872387921752, "learning_rate": 7.504108070249565e-06, "loss": 0.13, "step": 6615 }, { "epoch": 0.999244827065398, "grad_norm": 2.6121261116377092, "learning_rate": 7.5034235483788434e-06, "loss": 0.1316, "step": 6616 }, { "epoch": 0.9993958616523184, "grad_norm": 2.272447903305152, "learning_rate": 7.502738963883605e-06, "loss": 0.1585, "step": 6617 }, { "epoch": 0.9995468962392388, "grad_norm": 1.9958142322115593, "learning_rate": 7.5020543167809724e-06, "loss": 0.1466, "step": 6618 }, { "epoch": 0.9996979308261592, "grad_norm": 5.494499652722891, "learning_rate": 7.501369607088076e-06, "loss": 0.1633, "step": 6619 }, { "epoch": 0.9998489654130795, "grad_norm": 1.788642907281398, "learning_rate": 7.500684834822041e-06, "loss": 0.1322, "step": 6620 }, { "epoch": 1.0, "grad_norm": 1.9742092946646308, "learning_rate": 7.500000000000001e-06, "loss": 0.1549, "step": 6621 } ], "logging_steps": 1, "max_steps": 19863, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 153503327477760.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }