diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,5913 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 19.591836734693878, + "eval_steps": 500, + "global_step": 840, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.023323615160349854, + "grad_norm": 481.58927975506265, + "learning_rate": 1.9230769230769234e-07, + "loss": 4.1544, + "step": 1 + }, + { + "epoch": 0.04664723032069971, + "grad_norm": 480.04985092696364, + "learning_rate": 3.846153846153847e-07, + "loss": 4.1893, + "step": 2 + }, + { + "epoch": 0.06997084548104957, + "grad_norm": 464.6078319400137, + "learning_rate": 5.76923076923077e-07, + "loss": 3.9817, + "step": 3 + }, + { + "epoch": 0.09329446064139942, + "grad_norm": 313.6186353479334, + "learning_rate": 7.692307692307694e-07, + "loss": 3.5082, + "step": 4 + }, + { + "epoch": 0.11661807580174927, + "grad_norm": 130.28683050533158, + "learning_rate": 9.615384615384617e-07, + "loss": 3.1032, + "step": 5 + }, + { + "epoch": 0.13994169096209913, + "grad_norm": 108.72214082812204, + "learning_rate": 1.153846153846154e-06, + "loss": 2.9264, + "step": 6 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 90.00734833282364, + "learning_rate": 1.3461538461538462e-06, + "loss": 2.5326, + "step": 7 + }, + { + "epoch": 0.18658892128279883, + "grad_norm": 80.17560377949299, + "learning_rate": 1.5384615384615387e-06, + "loss": 2.2113, + "step": 8 + }, + { + "epoch": 0.2099125364431487, + "grad_norm": 256.89238079884734, + "learning_rate": 1.7307692307692308e-06, + "loss": 2.3292, + "step": 9 + }, + { + "epoch": 0.23323615160349853, + "grad_norm": 40.345129720791576, + "learning_rate": 1.9230769230769234e-06, + "loss": 1.8276, + "step": 10 + }, + { + "epoch": 0.2565597667638484, + "grad_norm": 37.352128701784686, + "learning_rate": 2.1153846153846155e-06, + "loss": 1.732, + "step": 11 + }, + { + "epoch": 0.27988338192419826, + "grad_norm": 39.91551747323652, + "learning_rate": 2.307692307692308e-06, + "loss": 1.558, + "step": 12 + }, + { + "epoch": 0.3032069970845481, + "grad_norm": 46.301839613156844, + "learning_rate": 2.5e-06, + "loss": 1.6342, + "step": 13 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 40.14073751930512, + "learning_rate": 2.6923076923076923e-06, + "loss": 1.4451, + "step": 14 + }, + { + "epoch": 0.3498542274052478, + "grad_norm": 42.4956914176308, + "learning_rate": 2.8846153846153845e-06, + "loss": 1.2808, + "step": 15 + }, + { + "epoch": 0.37317784256559766, + "grad_norm": 33.67740673818837, + "learning_rate": 3.0769230769230774e-06, + "loss": 1.0825, + "step": 16 + }, + { + "epoch": 0.3965014577259475, + "grad_norm": 129.23064107923477, + "learning_rate": 3.2692307692307696e-06, + "loss": 1.0782, + "step": 17 + }, + { + "epoch": 0.4198250728862974, + "grad_norm": 36.39439855381251, + "learning_rate": 3.4615384615384617e-06, + "loss": 0.6994, + "step": 18 + }, + { + "epoch": 0.44314868804664725, + "grad_norm": 15.512433879259516, + "learning_rate": 3.653846153846154e-06, + "loss": 0.5511, + "step": 19 + }, + { + "epoch": 0.46647230320699706, + "grad_norm": 18.30849165549196, + "learning_rate": 3.846153846153847e-06, + "loss": 0.5391, + "step": 20 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 20.61130656343632, + "learning_rate": 4.0384615384615385e-06, + "loss": 0.5451, + "step": 21 + }, + { + "epoch": 0.5131195335276968, + "grad_norm": 14.982450264158022, + "learning_rate": 4.230769230769231e-06, + "loss": 0.4925, + "step": 22 + }, + { + "epoch": 0.5364431486880467, + "grad_norm": 13.209296703805986, + "learning_rate": 4.423076923076924e-06, + "loss": 0.4292, + "step": 23 + }, + { + "epoch": 0.5597667638483965, + "grad_norm": 8.1438483301815, + "learning_rate": 4.615384615384616e-06, + "loss": 0.3799, + "step": 24 + }, + { + "epoch": 0.5830903790087464, + "grad_norm": 19.941972093465644, + "learning_rate": 4.807692307692308e-06, + "loss": 0.3865, + "step": 25 + }, + { + "epoch": 0.6064139941690962, + "grad_norm": 11.016048515240694, + "learning_rate": 5e-06, + "loss": 0.3728, + "step": 26 + }, + { + "epoch": 0.6297376093294461, + "grad_norm": 15.074313900645539, + "learning_rate": 4.999981380826043e-06, + "loss": 0.3462, + "step": 27 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 8.543797917151428, + "learning_rate": 4.999925523581508e-06, + "loss": 0.3476, + "step": 28 + }, + { + "epoch": 0.6763848396501457, + "grad_norm": 13.753135185003897, + "learning_rate": 4.99983242909841e-06, + "loss": 0.3278, + "step": 29 + }, + { + "epoch": 0.6997084548104956, + "grad_norm": 8.608638115725363, + "learning_rate": 4.999702098763422e-06, + "loss": 0.2821, + "step": 30 + }, + { + "epoch": 0.7230320699708455, + "grad_norm": 6.19614153747254, + "learning_rate": 4.9995345345178584e-06, + "loss": 0.2994, + "step": 31 + }, + { + "epoch": 0.7463556851311953, + "grad_norm": 7.678155942302579, + "learning_rate": 4.999329738857645e-06, + "loss": 0.2916, + "step": 32 + }, + { + "epoch": 0.7696793002915452, + "grad_norm": 8.051557345234135, + "learning_rate": 4.999087714833284e-06, + "loss": 0.277, + "step": 33 + }, + { + "epoch": 0.793002915451895, + "grad_norm": 6.334602262299757, + "learning_rate": 4.998808466049803e-06, + "loss": 0.2597, + "step": 34 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 5.092340192351995, + "learning_rate": 4.998491996666711e-06, + "loss": 0.2621, + "step": 35 + }, + { + "epoch": 0.8396501457725948, + "grad_norm": 12.200374434931867, + "learning_rate": 4.998138311397924e-06, + "loss": 0.2351, + "step": 36 + }, + { + "epoch": 0.8629737609329446, + "grad_norm": 5.732076823622836, + "learning_rate": 4.997747415511705e-06, + "loss": 0.202, + "step": 37 + }, + { + "epoch": 0.8862973760932945, + "grad_norm": 5.650767667488515, + "learning_rate": 4.9973193148305784e-06, + "loss": 0.1888, + "step": 38 + }, + { + "epoch": 0.9096209912536443, + "grad_norm": 8.206628251201462, + "learning_rate": 4.996854015731253e-06, + "loss": 0.1775, + "step": 39 + }, + { + "epoch": 0.9329446064139941, + "grad_norm": 8.145125128228061, + "learning_rate": 4.996351525144515e-06, + "loss": 0.1743, + "step": 40 + }, + { + "epoch": 0.956268221574344, + "grad_norm": 9.558683543841621, + "learning_rate": 4.995811850555131e-06, + "loss": 0.1706, + "step": 41 + }, + { + "epoch": 0.9795918367346939, + "grad_norm": 21.870848517891986, + "learning_rate": 4.995235000001739e-06, + "loss": 0.192, + "step": 42 + }, + { + "epoch": 1.0029154518950438, + "grad_norm": 258.46039046147865, + "learning_rate": 4.994620982076721e-06, + "loss": 0.4032, + "step": 43 + }, + { + "epoch": 1.0262390670553936, + "grad_norm": 26.567296212210675, + "learning_rate": 4.993969805926085e-06, + "loss": 0.5322, + "step": 44 + }, + { + "epoch": 1.0495626822157433, + "grad_norm": 25.494805498804556, + "learning_rate": 4.99328148124932e-06, + "loss": 0.6008, + "step": 45 + }, + { + "epoch": 1.0728862973760933, + "grad_norm": 15.470490452118884, + "learning_rate": 4.992556018299255e-06, + "loss": 0.4806, + "step": 46 + }, + { + "epoch": 1.096209912536443, + "grad_norm": 14.107985790378562, + "learning_rate": 4.9917934278819055e-06, + "loss": 0.4627, + "step": 47 + }, + { + "epoch": 1.119533527696793, + "grad_norm": 12.124582814383823, + "learning_rate": 4.990993721356317e-06, + "loss": 0.4231, + "step": 48 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 8.12927884627824, + "learning_rate": 4.990156910634387e-06, + "loss": 0.3367, + "step": 49 + }, + { + "epoch": 1.1661807580174928, + "grad_norm": 8.363524360811459, + "learning_rate": 4.989283008180697e-06, + "loss": 0.3325, + "step": 50 + }, + { + "epoch": 1.1895043731778425, + "grad_norm": 8.263812437870355, + "learning_rate": 4.988372027012319e-06, + "loss": 0.3007, + "step": 51 + }, + { + "epoch": 1.2128279883381925, + "grad_norm": 7.277527999797857, + "learning_rate": 4.987423980698627e-06, + "loss": 0.294, + "step": 52 + }, + { + "epoch": 1.2361516034985423, + "grad_norm": 12.255979985711312, + "learning_rate": 4.986438883361092e-06, + "loss": 0.3, + "step": 53 + }, + { + "epoch": 1.259475218658892, + "grad_norm": 20.381703777578387, + "learning_rate": 4.985416749673075e-06, + "loss": 0.2891, + "step": 54 + }, + { + "epoch": 1.282798833819242, + "grad_norm": 11.379464750926473, + "learning_rate": 4.9843575948596e-06, + "loss": 0.279, + "step": 55 + }, + { + "epoch": 1.306122448979592, + "grad_norm": 3.571428587777274, + "learning_rate": 4.983261434697141e-06, + "loss": 0.2591, + "step": 56 + }, + { + "epoch": 1.3294460641399417, + "grad_norm": 3.424764867354762, + "learning_rate": 4.982128285513373e-06, + "loss": 0.2327, + "step": 57 + }, + { + "epoch": 1.3527696793002915, + "grad_norm": 4.582571324669013, + "learning_rate": 4.98095816418694e-06, + "loss": 0.243, + "step": 58 + }, + { + "epoch": 1.3760932944606414, + "grad_norm": 6.075638838721734, + "learning_rate": 4.979751088147192e-06, + "loss": 0.2754, + "step": 59 + }, + { + "epoch": 1.3994169096209912, + "grad_norm": 6.412677796633595, + "learning_rate": 4.97850707537394e-06, + "loss": 0.261, + "step": 60 + }, + { + "epoch": 1.4227405247813412, + "grad_norm": 5.315104126793953, + "learning_rate": 4.977226144397174e-06, + "loss": 0.2357, + "step": 61 + }, + { + "epoch": 1.446064139941691, + "grad_norm": 6.211582673349952, + "learning_rate": 4.9759083142967965e-06, + "loss": 0.2632, + "step": 62 + }, + { + "epoch": 1.469387755102041, + "grad_norm": 5.980458863173571, + "learning_rate": 4.974553604702332e-06, + "loss": 0.251, + "step": 63 + }, + { + "epoch": 1.4927113702623906, + "grad_norm": 5.652339700672357, + "learning_rate": 4.973162035792641e-06, + "loss": 0.2476, + "step": 64 + }, + { + "epoch": 1.5160349854227406, + "grad_norm": 4.840728604985413, + "learning_rate": 4.971733628295614e-06, + "loss": 0.235, + "step": 65 + }, + { + "epoch": 1.5393586005830904, + "grad_norm": 3.658991641610298, + "learning_rate": 4.970268403487866e-06, + "loss": 0.2321, + "step": 66 + }, + { + "epoch": 1.5626822157434401, + "grad_norm": 3.487525781242212, + "learning_rate": 4.9687663831944156e-06, + "loss": 0.2429, + "step": 67 + }, + { + "epoch": 1.58600583090379, + "grad_norm": 3.749615204460126, + "learning_rate": 4.967227589788365e-06, + "loss": 0.2394, + "step": 68 + }, + { + "epoch": 1.60932944606414, + "grad_norm": 4.290007606979775, + "learning_rate": 4.965652046190565e-06, + "loss": 0.2362, + "step": 69 + }, + { + "epoch": 1.6326530612244898, + "grad_norm": 4.69637810774778, + "learning_rate": 4.964039775869271e-06, + "loss": 0.2208, + "step": 70 + }, + { + "epoch": 1.6559766763848396, + "grad_norm": 3.6428908472078065, + "learning_rate": 4.962390802839797e-06, + "loss": 0.234, + "step": 71 + }, + { + "epoch": 1.6793002915451893, + "grad_norm": 2.7063967799319495, + "learning_rate": 4.960705151664155e-06, + "loss": 0.2171, + "step": 72 + }, + { + "epoch": 1.7026239067055393, + "grad_norm": 2.9521513448960213, + "learning_rate": 4.9589828474506906e-06, + "loss": 0.2166, + "step": 73 + }, + { + "epoch": 1.7259475218658893, + "grad_norm": 3.575771040565363, + "learning_rate": 4.9572239158537095e-06, + "loss": 0.2117, + "step": 74 + }, + { + "epoch": 1.749271137026239, + "grad_norm": 4.218830249164411, + "learning_rate": 4.955428383073094e-06, + "loss": 0.2067, + "step": 75 + }, + { + "epoch": 1.7725947521865888, + "grad_norm": 5.195007590617856, + "learning_rate": 4.9535962758539155e-06, + "loss": 0.2222, + "step": 76 + }, + { + "epoch": 1.7959183673469388, + "grad_norm": 4.3676585044114, + "learning_rate": 4.951727621486031e-06, + "loss": 0.1933, + "step": 77 + }, + { + "epoch": 1.8192419825072887, + "grad_norm": 3.8250350003105953, + "learning_rate": 4.949822447803681e-06, + "loss": 0.1821, + "step": 78 + }, + { + "epoch": 1.8425655976676385, + "grad_norm": 3.7621739192349652, + "learning_rate": 4.947880783185074e-06, + "loss": 0.1907, + "step": 79 + }, + { + "epoch": 1.8658892128279883, + "grad_norm": 3.2426278056631124, + "learning_rate": 4.945902656551964e-06, + "loss": 0.1862, + "step": 80 + }, + { + "epoch": 1.8892128279883382, + "grad_norm": 3.0607040744667753, + "learning_rate": 4.943888097369216e-06, + "loss": 0.1903, + "step": 81 + }, + { + "epoch": 1.9125364431486882, + "grad_norm": 1.9997107177378481, + "learning_rate": 4.941837135644374e-06, + "loss": 0.1699, + "step": 82 + }, + { + "epoch": 1.935860058309038, + "grad_norm": 2.9208613794769716, + "learning_rate": 4.939749801927208e-06, + "loss": 0.1733, + "step": 83 + }, + { + "epoch": 1.9591836734693877, + "grad_norm": 2.446872104141256, + "learning_rate": 4.9376261273092614e-06, + "loss": 0.1755, + "step": 84 + }, + { + "epoch": 1.9825072886297375, + "grad_norm": 2.3668168284457636, + "learning_rate": 4.935466143423389e-06, + "loss": 0.1675, + "step": 85 + }, + { + "epoch": 2.0058309037900877, + "grad_norm": 2.2402757742265083, + "learning_rate": 4.933269882443281e-06, + "loss": 0.1691, + "step": 86 + }, + { + "epoch": 2.0291545189504374, + "grad_norm": 1.3263354642981466, + "learning_rate": 4.9310373770829925e-06, + "loss": 0.1523, + "step": 87 + }, + { + "epoch": 2.052478134110787, + "grad_norm": 2.836622255323569, + "learning_rate": 4.928768660596446e-06, + "loss": 0.156, + "step": 88 + }, + { + "epoch": 2.075801749271137, + "grad_norm": 2.479270537521849, + "learning_rate": 4.926463766776945e-06, + "loss": 0.1613, + "step": 89 + }, + { + "epoch": 2.0991253644314867, + "grad_norm": 2.975995828348743, + "learning_rate": 4.924122729956662e-06, + "loss": 0.1495, + "step": 90 + }, + { + "epoch": 2.122448979591837, + "grad_norm": 1.9541545138221803, + "learning_rate": 4.921745585006135e-06, + "loss": 0.1527, + "step": 91 + }, + { + "epoch": 2.1457725947521866, + "grad_norm": 4.856180692789397, + "learning_rate": 4.919332367333748e-06, + "loss": 0.1459, + "step": 92 + }, + { + "epoch": 2.1690962099125364, + "grad_norm": 4.550523899643916, + "learning_rate": 4.916883112885192e-06, + "loss": 0.1442, + "step": 93 + }, + { + "epoch": 2.192419825072886, + "grad_norm": 4.080918564709067, + "learning_rate": 4.914397858142945e-06, + "loss": 0.145, + "step": 94 + }, + { + "epoch": 2.2157434402332363, + "grad_norm": 3.177257762352267, + "learning_rate": 4.911876640125719e-06, + "loss": 0.1464, + "step": 95 + }, + { + "epoch": 2.239067055393586, + "grad_norm": 3.2919310612465007, + "learning_rate": 4.909319496387911e-06, + "loss": 0.1405, + "step": 96 + }, + { + "epoch": 2.262390670553936, + "grad_norm": 2.4487248703045426, + "learning_rate": 4.9067264650190436e-06, + "loss": 0.1401, + "step": 97 + }, + { + "epoch": 2.2857142857142856, + "grad_norm": 3.2548940929053702, + "learning_rate": 4.904097584643201e-06, + "loss": 0.1425, + "step": 98 + }, + { + "epoch": 2.3090379008746353, + "grad_norm": 2.3394224777927675, + "learning_rate": 4.901432894418446e-06, + "loss": 0.144, + "step": 99 + }, + { + "epoch": 2.3323615160349855, + "grad_norm": 1.987074930092349, + "learning_rate": 4.8987324340362445e-06, + "loss": 0.1271, + "step": 100 + }, + { + "epoch": 2.3556851311953353, + "grad_norm": 2.269137691444028, + "learning_rate": 4.895996243720868e-06, + "loss": 0.1273, + "step": 101 + }, + { + "epoch": 2.379008746355685, + "grad_norm": 3.6904521698177315, + "learning_rate": 4.8932243642288e-06, + "loss": 0.1418, + "step": 102 + }, + { + "epoch": 2.402332361516035, + "grad_norm": 7.437555393044772, + "learning_rate": 4.890416836848128e-06, + "loss": 0.1359, + "step": 103 + }, + { + "epoch": 2.425655976676385, + "grad_norm": 6.088458615126284, + "learning_rate": 4.887573703397921e-06, + "loss": 0.1386, + "step": 104 + }, + { + "epoch": 2.4489795918367347, + "grad_norm": 1.9457975954726117, + "learning_rate": 4.884695006227619e-06, + "loss": 0.1227, + "step": 105 + }, + { + "epoch": 2.4723032069970845, + "grad_norm": 2.127917923004136, + "learning_rate": 4.8817807882163904e-06, + "loss": 0.1284, + "step": 106 + }, + { + "epoch": 2.4956268221574343, + "grad_norm": 4.11663293451448, + "learning_rate": 4.878831092772501e-06, + "loss": 0.129, + "step": 107 + }, + { + "epoch": 2.518950437317784, + "grad_norm": 6.08617442775337, + "learning_rate": 4.875845963832667e-06, + "loss": 0.1305, + "step": 108 + }, + { + "epoch": 2.542274052478134, + "grad_norm": 1.6921004000832196, + "learning_rate": 4.872825445861395e-06, + "loss": 0.1356, + "step": 109 + }, + { + "epoch": 2.565597667638484, + "grad_norm": 10.251083316775865, + "learning_rate": 4.869769583850324e-06, + "loss": 0.1247, + "step": 110 + }, + { + "epoch": 2.5889212827988337, + "grad_norm": 8.60721468259536, + "learning_rate": 4.8666784233175566e-06, + "loss": 0.1315, + "step": 111 + }, + { + "epoch": 2.612244897959184, + "grad_norm": 6.1560755671076235, + "learning_rate": 4.863552010306976e-06, + "loss": 0.1321, + "step": 112 + }, + { + "epoch": 2.6355685131195337, + "grad_norm": 9.37865041235299, + "learning_rate": 4.860390391387566e-06, + "loss": 0.1327, + "step": 113 + }, + { + "epoch": 2.6588921282798834, + "grad_norm": 4.762581475629758, + "learning_rate": 4.857193613652711e-06, + "loss": 0.1255, + "step": 114 + }, + { + "epoch": 2.682215743440233, + "grad_norm": 6.745590182330897, + "learning_rate": 4.8539617247195e-06, + "loss": 0.1446, + "step": 115 + }, + { + "epoch": 2.705539358600583, + "grad_norm": 3.3151863289735295, + "learning_rate": 4.850694772728015e-06, + "loss": 0.125, + "step": 116 + }, + { + "epoch": 2.7288629737609327, + "grad_norm": 3.6648399975915424, + "learning_rate": 4.847392806340615e-06, + "loss": 0.1308, + "step": 117 + }, + { + "epoch": 2.752186588921283, + "grad_norm": 3.9684370022599276, + "learning_rate": 4.844055874741208e-06, + "loss": 0.1276, + "step": 118 + }, + { + "epoch": 2.7755102040816326, + "grad_norm": 3.9099707621386135, + "learning_rate": 4.8406840276345215e-06, + "loss": 0.1353, + "step": 119 + }, + { + "epoch": 2.7988338192419824, + "grad_norm": 2.4159538489348478, + "learning_rate": 4.837277315245364e-06, + "loss": 0.121, + "step": 120 + }, + { + "epoch": 2.8221574344023326, + "grad_norm": 3.3742191683006513, + "learning_rate": 4.833835788317869e-06, + "loss": 0.1338, + "step": 121 + }, + { + "epoch": 2.8454810495626823, + "grad_norm": 2.2056302375331316, + "learning_rate": 4.830359498114749e-06, + "loss": 0.1257, + "step": 122 + }, + { + "epoch": 2.868804664723032, + "grad_norm": 2.4034843900335123, + "learning_rate": 4.826848496416526e-06, + "loss": 0.1348, + "step": 123 + }, + { + "epoch": 2.892128279883382, + "grad_norm": 2.2255115906707745, + "learning_rate": 4.82330283552076e-06, + "loss": 0.137, + "step": 124 + }, + { + "epoch": 2.9154518950437316, + "grad_norm": 3.043279002757231, + "learning_rate": 4.819722568241274e-06, + "loss": 0.1339, + "step": 125 + }, + { + "epoch": 2.938775510204082, + "grad_norm": 2.172615748264224, + "learning_rate": 4.816107747907362e-06, + "loss": 0.1245, + "step": 126 + }, + { + "epoch": 2.9620991253644315, + "grad_norm": 3.9756197608178, + "learning_rate": 4.812458428362999e-06, + "loss": 0.1256, + "step": 127 + }, + { + "epoch": 2.9854227405247813, + "grad_norm": 1.6228193910931026, + "learning_rate": 4.808774663966039e-06, + "loss": 0.1268, + "step": 128 + }, + { + "epoch": 3.008746355685131, + "grad_norm": 2.666938281134312, + "learning_rate": 4.8050565095874e-06, + "loss": 0.1239, + "step": 129 + }, + { + "epoch": 3.0320699708454812, + "grad_norm": 2.427622711977247, + "learning_rate": 4.801304020610255e-06, + "loss": 0.1271, + "step": 130 + }, + { + "epoch": 3.055393586005831, + "grad_norm": 3.238407882588861, + "learning_rate": 4.7975172529291965e-06, + "loss": 0.1279, + "step": 131 + }, + { + "epoch": 3.0787172011661808, + "grad_norm": 4.385283020351036, + "learning_rate": 4.793696262949417e-06, + "loss": 0.1246, + "step": 132 + }, + { + "epoch": 3.1020408163265305, + "grad_norm": 1.7220244687210526, + "learning_rate": 4.789841107585855e-06, + "loss": 0.1151, + "step": 133 + }, + { + "epoch": 3.1253644314868803, + "grad_norm": 2.19739951848995, + "learning_rate": 4.785951844262361e-06, + "loss": 0.1215, + "step": 134 + }, + { + "epoch": 3.1486880466472305, + "grad_norm": 3.1169982699749537, + "learning_rate": 4.782028530910827e-06, + "loss": 0.1241, + "step": 135 + }, + { + "epoch": 3.17201166180758, + "grad_norm": 3.16924293190244, + "learning_rate": 4.77807122597034e-06, + "loss": 0.1209, + "step": 136 + }, + { + "epoch": 3.19533527696793, + "grad_norm": 3.8621927083561602, + "learning_rate": 4.7740799883862966e-06, + "loss": 0.1173, + "step": 137 + }, + { + "epoch": 3.2186588921282797, + "grad_norm": 2.9142580350261293, + "learning_rate": 4.770054877609534e-06, + "loss": 0.1171, + "step": 138 + }, + { + "epoch": 3.24198250728863, + "grad_norm": 5.773960446017771, + "learning_rate": 4.765995953595446e-06, + "loss": 0.1298, + "step": 139 + }, + { + "epoch": 3.2653061224489797, + "grad_norm": 3.8584285919316486, + "learning_rate": 4.761903276803078e-06, + "loss": 0.1254, + "step": 140 + }, + { + "epoch": 3.2886297376093294, + "grad_norm": 5.500769127907038, + "learning_rate": 4.7577769081942425e-06, + "loss": 0.1245, + "step": 141 + }, + { + "epoch": 3.311953352769679, + "grad_norm": 3.679145141285145, + "learning_rate": 4.753616909232597e-06, + "loss": 0.124, + "step": 142 + }, + { + "epoch": 3.335276967930029, + "grad_norm": 4.3867652734342855, + "learning_rate": 4.749423341882738e-06, + "loss": 0.1182, + "step": 143 + }, + { + "epoch": 3.358600583090379, + "grad_norm": 2.816608564541258, + "learning_rate": 4.745196268609274e-06, + "loss": 0.1267, + "step": 144 + }, + { + "epoch": 3.381924198250729, + "grad_norm": 11.096184949547764, + "learning_rate": 4.740935752375893e-06, + "loss": 0.1297, + "step": 145 + }, + { + "epoch": 3.4052478134110786, + "grad_norm": 1.649738611844767, + "learning_rate": 4.736641856644431e-06, + "loss": 0.1211, + "step": 146 + }, + { + "epoch": 3.4285714285714284, + "grad_norm": 8.186265226248914, + "learning_rate": 4.732314645373922e-06, + "loss": 0.1353, + "step": 147 + }, + { + "epoch": 3.4518950437317786, + "grad_norm": 5.788547837910415, + "learning_rate": 4.727954183019644e-06, + "loss": 0.125, + "step": 148 + }, + { + "epoch": 3.4752186588921283, + "grad_norm": 7.463202678882185, + "learning_rate": 4.723560534532163e-06, + "loss": 0.1214, + "step": 149 + }, + { + "epoch": 3.498542274052478, + "grad_norm": 9.046619892641317, + "learning_rate": 4.7191337653563635e-06, + "loss": 0.1272, + "step": 150 + }, + { + "epoch": 3.521865889212828, + "grad_norm": 4.648476069955531, + "learning_rate": 4.714673941430474e-06, + "loss": 0.1204, + "step": 151 + }, + { + "epoch": 3.5451895043731776, + "grad_norm": 6.417386827886447, + "learning_rate": 4.710181129185085e-06, + "loss": 0.1299, + "step": 152 + }, + { + "epoch": 3.568513119533528, + "grad_norm": 5.286844754933566, + "learning_rate": 4.705655395542158e-06, + "loss": 0.1263, + "step": 153 + }, + { + "epoch": 3.5918367346938775, + "grad_norm": 2.578749240092962, + "learning_rate": 4.70109680791403e-06, + "loss": 0.1209, + "step": 154 + }, + { + "epoch": 3.6151603498542273, + "grad_norm": 9.466600835150404, + "learning_rate": 4.696505434202411e-06, + "loss": 0.1281, + "step": 155 + }, + { + "epoch": 3.6384839650145775, + "grad_norm": 1.9708225301025741, + "learning_rate": 4.691881342797368e-06, + "loss": 0.1187, + "step": 156 + }, + { + "epoch": 3.6618075801749272, + "grad_norm": 3.655745903542407, + "learning_rate": 4.6872246025763135e-06, + "loss": 0.1158, + "step": 157 + }, + { + "epoch": 3.685131195335277, + "grad_norm": 4.730977164178507, + "learning_rate": 4.6825352829029705e-06, + "loss": 0.1267, + "step": 158 + }, + { + "epoch": 3.7084548104956268, + "grad_norm": 4.191588270744791, + "learning_rate": 4.677813453626347e-06, + "loss": 0.1249, + "step": 159 + }, + { + "epoch": 3.7317784256559765, + "grad_norm": 2.8625748870249113, + "learning_rate": 4.67305918507969e-06, + "loss": 0.1103, + "step": 160 + }, + { + "epoch": 3.7551020408163263, + "grad_norm": 7.161616682671691, + "learning_rate": 4.668272548079445e-06, + "loss": 0.1147, + "step": 161 + }, + { + "epoch": 3.7784256559766765, + "grad_norm": 1.9603442628349634, + "learning_rate": 4.66345361392419e-06, + "loss": 0.1146, + "step": 162 + }, + { + "epoch": 3.801749271137026, + "grad_norm": 3.434376077417512, + "learning_rate": 4.658602454393584e-06, + "loss": 0.1186, + "step": 163 + }, + { + "epoch": 3.825072886297376, + "grad_norm": 5.614931038002465, + "learning_rate": 4.653719141747297e-06, + "loss": 0.1335, + "step": 164 + }, + { + "epoch": 3.848396501457726, + "grad_norm": 3.7351012222076014, + "learning_rate": 4.648803748723925e-06, + "loss": 0.1182, + "step": 165 + }, + { + "epoch": 3.871720116618076, + "grad_norm": 4.285033398991045, + "learning_rate": 4.643856348539913e-06, + "loss": 0.1285, + "step": 166 + }, + { + "epoch": 3.8950437317784257, + "grad_norm": 5.9516731153780995, + "learning_rate": 4.638877014888468e-06, + "loss": 0.1204, + "step": 167 + }, + { + "epoch": 3.9183673469387754, + "grad_norm": 2.98542040345268, + "learning_rate": 4.633865821938449e-06, + "loss": 0.1184, + "step": 168 + }, + { + "epoch": 3.941690962099125, + "grad_norm": 3.50350218349349, + "learning_rate": 4.6288228443332786e-06, + "loss": 0.1166, + "step": 169 + }, + { + "epoch": 3.9650145772594754, + "grad_norm": 4.810586127656542, + "learning_rate": 4.623748157189817e-06, + "loss": 0.1217, + "step": 170 + }, + { + "epoch": 3.988338192419825, + "grad_norm": 2.569654107918901, + "learning_rate": 4.61864183609725e-06, + "loss": 0.1212, + "step": 171 + }, + { + "epoch": 4.011661807580175, + "grad_norm": 1.9366280839570846, + "learning_rate": 4.613503957115963e-06, + "loss": 0.1188, + "step": 172 + }, + { + "epoch": 4.034985422740525, + "grad_norm": 3.4762352264320815, + "learning_rate": 4.608334596776406e-06, + "loss": 0.1222, + "step": 173 + }, + { + "epoch": 4.058309037900875, + "grad_norm": 4.344331551462011, + "learning_rate": 4.603133832077953e-06, + "loss": 0.1239, + "step": 174 + }, + { + "epoch": 4.081632653061225, + "grad_norm": 2.095141275963881, + "learning_rate": 4.597901740487761e-06, + "loss": 0.1167, + "step": 175 + }, + { + "epoch": 4.104956268221574, + "grad_norm": 2.8477141867605917, + "learning_rate": 4.592638399939606e-06, + "loss": 0.1131, + "step": 176 + }, + { + "epoch": 4.128279883381924, + "grad_norm": 3.481611858306002, + "learning_rate": 4.587343888832732e-06, + "loss": 0.117, + "step": 177 + }, + { + "epoch": 4.151603498542274, + "grad_norm": 2.8284834323753216, + "learning_rate": 4.582018286030677e-06, + "loss": 0.1093, + "step": 178 + }, + { + "epoch": 4.174927113702624, + "grad_norm": 2.1278869988762494, + "learning_rate": 4.576661670860102e-06, + "loss": 0.1205, + "step": 179 + }, + { + "epoch": 4.198250728862973, + "grad_norm": 4.183833780535433, + "learning_rate": 4.571274123109606e-06, + "loss": 0.1121, + "step": 180 + }, + { + "epoch": 4.221574344023324, + "grad_norm": 2.1166025177053305, + "learning_rate": 4.565855723028539e-06, + "loss": 0.1044, + "step": 181 + }, + { + "epoch": 4.244897959183674, + "grad_norm": 2.479368816009341, + "learning_rate": 4.560406551325811e-06, + "loss": 0.1161, + "step": 182 + }, + { + "epoch": 4.2682215743440235, + "grad_norm": 2.987320649577329, + "learning_rate": 4.55492668916868e-06, + "loss": 0.1211, + "step": 183 + }, + { + "epoch": 4.291545189504373, + "grad_norm": 1.9760419439041783, + "learning_rate": 4.5494162181815515e-06, + "loss": 0.1204, + "step": 184 + }, + { + "epoch": 4.314868804664723, + "grad_norm": 4.117652128521018, + "learning_rate": 4.543875220444761e-06, + "loss": 0.1175, + "step": 185 + }, + { + "epoch": 4.338192419825073, + "grad_norm": 3.226732595915632, + "learning_rate": 4.5383037784933494e-06, + "loss": 0.115, + "step": 186 + }, + { + "epoch": 4.3615160349854225, + "grad_norm": 2.9884740031788435, + "learning_rate": 4.532701975315832e-06, + "loss": 0.1163, + "step": 187 + }, + { + "epoch": 4.384839650145772, + "grad_norm": 4.084097628525921, + "learning_rate": 4.52706989435297e-06, + "loss": 0.1098, + "step": 188 + }, + { + "epoch": 4.408163265306122, + "grad_norm": 2.097393907448701, + "learning_rate": 4.521407619496517e-06, + "loss": 0.1168, + "step": 189 + }, + { + "epoch": 4.431486880466473, + "grad_norm": 2.798836132392157, + "learning_rate": 4.515715235087979e-06, + "loss": 0.1085, + "step": 190 + }, + { + "epoch": 4.454810495626822, + "grad_norm": 5.007862394325352, + "learning_rate": 4.509992825917352e-06, + "loss": 0.1186, + "step": 191 + }, + { + "epoch": 4.478134110787172, + "grad_norm": 1.7983696733373176, + "learning_rate": 4.504240477221861e-06, + "loss": 0.1163, + "step": 192 + }, + { + "epoch": 4.501457725947522, + "grad_norm": 3.8700602640590405, + "learning_rate": 4.498458274684691e-06, + "loss": 0.1175, + "step": 193 + }, + { + "epoch": 4.524781341107872, + "grad_norm": 2.6360213197995006, + "learning_rate": 4.492646304433711e-06, + "loss": 0.1214, + "step": 194 + }, + { + "epoch": 4.548104956268221, + "grad_norm": 1.6210900731791886, + "learning_rate": 4.486804653040187e-06, + "loss": 0.1175, + "step": 195 + }, + { + "epoch": 4.571428571428571, + "grad_norm": 4.3135489694582185, + "learning_rate": 4.4809334075175e-06, + "loss": 0.1199, + "step": 196 + }, + { + "epoch": 4.594752186588921, + "grad_norm": 1.5436096405592212, + "learning_rate": 4.475032655319842e-06, + "loss": 0.1111, + "step": 197 + }, + { + "epoch": 4.618075801749271, + "grad_norm": 2.1458534931112387, + "learning_rate": 4.469102484340919e-06, + "loss": 0.113, + "step": 198 + }, + { + "epoch": 4.641399416909621, + "grad_norm": 1.8801712567873288, + "learning_rate": 4.463142982912638e-06, + "loss": 0.1092, + "step": 199 + }, + { + "epoch": 4.664723032069971, + "grad_norm": 2.376405199853165, + "learning_rate": 4.457154239803796e-06, + "loss": 0.1186, + "step": 200 + }, + { + "epoch": 4.688046647230321, + "grad_norm": 2.029847820196563, + "learning_rate": 4.451136344218751e-06, + "loss": 0.1197, + "step": 201 + }, + { + "epoch": 4.711370262390671, + "grad_norm": 1.947337066043897, + "learning_rate": 4.445089385796099e-06, + "loss": 0.1178, + "step": 202 + }, + { + "epoch": 4.73469387755102, + "grad_norm": 4.016814890510145, + "learning_rate": 4.439013454607338e-06, + "loss": 0.1127, + "step": 203 + }, + { + "epoch": 4.75801749271137, + "grad_norm": 3.0189359998562453, + "learning_rate": 4.432908641155522e-06, + "loss": 0.1215, + "step": 204 + }, + { + "epoch": 4.78134110787172, + "grad_norm": 4.180011943453285, + "learning_rate": 4.4267750363739195e-06, + "loss": 0.1166, + "step": 205 + }, + { + "epoch": 4.80466472303207, + "grad_norm": 2.289589362328887, + "learning_rate": 4.420612731624652e-06, + "loss": 0.108, + "step": 206 + }, + { + "epoch": 4.827988338192419, + "grad_norm": 6.041807431740632, + "learning_rate": 4.414421818697341e-06, + "loss": 0.1171, + "step": 207 + }, + { + "epoch": 4.85131195335277, + "grad_norm": 2.4860177037352154, + "learning_rate": 4.408202389807733e-06, + "loss": 0.119, + "step": 208 + }, + { + "epoch": 4.87463556851312, + "grad_norm": 2.635976458014569, + "learning_rate": 4.401954537596332e-06, + "loss": 0.109, + "step": 209 + }, + { + "epoch": 4.8979591836734695, + "grad_norm": 3.1423969489220687, + "learning_rate": 4.395678355127013e-06, + "loss": 0.1217, + "step": 210 + }, + { + "epoch": 4.921282798833819, + "grad_norm": 2.382465975219246, + "learning_rate": 4.3893739358856465e-06, + "loss": 0.1074, + "step": 211 + }, + { + "epoch": 4.944606413994169, + "grad_norm": 2.057841386063691, + "learning_rate": 4.383041373778691e-06, + "loss": 0.1147, + "step": 212 + }, + { + "epoch": 4.967930029154519, + "grad_norm": 3.155060950172627, + "learning_rate": 4.376680763131811e-06, + "loss": 0.116, + "step": 213 + }, + { + "epoch": 4.9912536443148685, + "grad_norm": 2.865124701004675, + "learning_rate": 4.3702921986884576e-06, + "loss": 0.1147, + "step": 214 + }, + { + "epoch": 5.014577259475218, + "grad_norm": 1.9051999811650222, + "learning_rate": 4.363875775608464e-06, + "loss": 0.1101, + "step": 215 + }, + { + "epoch": 5.037900874635569, + "grad_norm": 2.938649054269692, + "learning_rate": 4.357431589466629e-06, + "loss": 0.1107, + "step": 216 + }, + { + "epoch": 5.061224489795919, + "grad_norm": 3.6046541064451407, + "learning_rate": 4.350959736251291e-06, + "loss": 0.1098, + "step": 217 + }, + { + "epoch": 5.084548104956268, + "grad_norm": 2.1379404428463866, + "learning_rate": 4.344460312362899e-06, + "loss": 0.1066, + "step": 218 + }, + { + "epoch": 5.107871720116618, + "grad_norm": 4.314344150035587, + "learning_rate": 4.337933414612576e-06, + "loss": 0.1089, + "step": 219 + }, + { + "epoch": 5.131195335276968, + "grad_norm": 3.577256942798804, + "learning_rate": 4.3313791402206765e-06, + "loss": 0.1202, + "step": 220 + }, + { + "epoch": 5.154518950437318, + "grad_norm": 3.4475350682589148, + "learning_rate": 4.324797586815342e-06, + "loss": 0.1112, + "step": 221 + }, + { + "epoch": 5.177842565597667, + "grad_norm": 4.006896987066902, + "learning_rate": 4.318188852431043e-06, + "loss": 0.1069, + "step": 222 + }, + { + "epoch": 5.201166180758017, + "grad_norm": 3.903254344498221, + "learning_rate": 4.311553035507118e-06, + "loss": 0.1081, + "step": 223 + }, + { + "epoch": 5.224489795918367, + "grad_norm": 3.22338459454092, + "learning_rate": 4.3048902348863116e-06, + "loss": 0.1146, + "step": 224 + }, + { + "epoch": 5.247813411078718, + "grad_norm": 3.776843948198735, + "learning_rate": 4.2982005498133e-06, + "loss": 0.1108, + "step": 225 + }, + { + "epoch": 5.271137026239067, + "grad_norm": 2.350737873867638, + "learning_rate": 4.291484079933208e-06, + "loss": 0.0966, + "step": 226 + }, + { + "epoch": 5.294460641399417, + "grad_norm": 3.09101608162593, + "learning_rate": 4.284740925290136e-06, + "loss": 0.1102, + "step": 227 + }, + { + "epoch": 5.317784256559767, + "grad_norm": 4.080414695608127, + "learning_rate": 4.277971186325658e-06, + "loss": 0.1114, + "step": 228 + }, + { + "epoch": 5.341107871720117, + "grad_norm": 2.4749981966879084, + "learning_rate": 4.271174963877331e-06, + "loss": 0.1114, + "step": 229 + }, + { + "epoch": 5.364431486880466, + "grad_norm": 6.2340152034335485, + "learning_rate": 4.264352359177196e-06, + "loss": 0.1203, + "step": 230 + }, + { + "epoch": 5.387755102040816, + "grad_norm": 2.998631324089385, + "learning_rate": 4.2575034738502615e-06, + "loss": 0.104, + "step": 231 + }, + { + "epoch": 5.411078717201166, + "grad_norm": 4.196171450636836, + "learning_rate": 4.250628409912998e-06, + "loss": 0.1154, + "step": 232 + }, + { + "epoch": 5.4344023323615165, + "grad_norm": 2.2916532573885626, + "learning_rate": 4.243727269771815e-06, + "loss": 0.1098, + "step": 233 + }, + { + "epoch": 5.457725947521866, + "grad_norm": 3.5946716453629075, + "learning_rate": 4.236800156221536e-06, + "loss": 0.1124, + "step": 234 + }, + { + "epoch": 5.481049562682216, + "grad_norm": 2.0708027960855855, + "learning_rate": 4.229847172443866e-06, + "loss": 0.1147, + "step": 235 + }, + { + "epoch": 5.504373177842566, + "grad_norm": 3.110015343140659, + "learning_rate": 4.222868422005856e-06, + "loss": 0.1092, + "step": 236 + }, + { + "epoch": 5.5276967930029155, + "grad_norm": 1.9738246969785502, + "learning_rate": 4.2158640088583625e-06, + "loss": 0.1106, + "step": 237 + }, + { + "epoch": 5.551020408163265, + "grad_norm": 2.3185763558818153, + "learning_rate": 4.208834037334494e-06, + "loss": 0.1114, + "step": 238 + }, + { + "epoch": 5.574344023323615, + "grad_norm": 3.0247875422752126, + "learning_rate": 4.20177861214806e-06, + "loss": 0.111, + "step": 239 + }, + { + "epoch": 5.597667638483965, + "grad_norm": 1.4817220169187448, + "learning_rate": 4.194697838392013e-06, + "loss": 0.1082, + "step": 240 + }, + { + "epoch": 5.6209912536443145, + "grad_norm": 3.4112665874767045, + "learning_rate": 4.1875918215368785e-06, + "loss": 0.1087, + "step": 241 + }, + { + "epoch": 5.644314868804665, + "grad_norm": 1.788023601218338, + "learning_rate": 4.180460667429188e-06, + "loss": 0.1158, + "step": 242 + }, + { + "epoch": 5.667638483965015, + "grad_norm": 2.4010457820126123, + "learning_rate": 4.1733044822899016e-06, + "loss": 0.1082, + "step": 243 + }, + { + "epoch": 5.690962099125365, + "grad_norm": 1.9745487615109616, + "learning_rate": 4.166123372712823e-06, + "loss": 0.1077, + "step": 244 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 2.7086581891198156, + "learning_rate": 4.158917445663014e-06, + "loss": 0.1158, + "step": 245 + }, + { + "epoch": 5.737609329446064, + "grad_norm": 3.2519661751333495, + "learning_rate": 4.151686808475204e-06, + "loss": 0.1075, + "step": 246 + }, + { + "epoch": 5.760932944606414, + "grad_norm": 2.5247034257513272, + "learning_rate": 4.144431568852185e-06, + "loss": 0.1098, + "step": 247 + }, + { + "epoch": 5.784256559766764, + "grad_norm": 2.1809660193845204, + "learning_rate": 4.137151834863213e-06, + "loss": 0.1041, + "step": 248 + }, + { + "epoch": 5.807580174927113, + "grad_norm": 3.8326422994514417, + "learning_rate": 4.1298477149423935e-06, + "loss": 0.1004, + "step": 249 + }, + { + "epoch": 5.830903790087463, + "grad_norm": 1.598467781622798, + "learning_rate": 4.122519317887072e-06, + "loss": 0.1085, + "step": 250 + }, + { + "epoch": 5.854227405247814, + "grad_norm": 2.765625862078344, + "learning_rate": 4.115166752856206e-06, + "loss": 0.1046, + "step": 251 + }, + { + "epoch": 5.877551020408164, + "grad_norm": 1.6978018630207297, + "learning_rate": 4.1077901293687464e-06, + "loss": 0.1068, + "step": 252 + }, + { + "epoch": 5.900874635568513, + "grad_norm": 1.7221455411691755, + "learning_rate": 4.100389557302001e-06, + "loss": 0.108, + "step": 253 + }, + { + "epoch": 5.924198250728863, + "grad_norm": 3.352657539369917, + "learning_rate": 4.092965146890002e-06, + "loss": 0.1098, + "step": 254 + }, + { + "epoch": 5.947521865889213, + "grad_norm": 2.23141689157229, + "learning_rate": 4.085517008721861e-06, + "loss": 0.1066, + "step": 255 + }, + { + "epoch": 5.970845481049563, + "grad_norm": 2.1918226856159384, + "learning_rate": 4.078045253740121e-06, + "loss": 0.1126, + "step": 256 + }, + { + "epoch": 5.994169096209912, + "grad_norm": 2.4817818597720085, + "learning_rate": 4.070549993239106e-06, + "loss": 0.1117, + "step": 257 + }, + { + "epoch": 6.017492711370262, + "grad_norm": 2.701936405356819, + "learning_rate": 4.0630313388632645e-06, + "loss": 0.0974, + "step": 258 + }, + { + "epoch": 6.040816326530612, + "grad_norm": 2.094145552436028, + "learning_rate": 4.055489402605504e-06, + "loss": 0.1021, + "step": 259 + }, + { + "epoch": 6.0641399416909625, + "grad_norm": 2.539235740033433, + "learning_rate": 4.047924296805522e-06, + "loss": 0.0984, + "step": 260 + }, + { + "epoch": 6.087463556851312, + "grad_norm": 2.5824401808658113, + "learning_rate": 4.040336134148136e-06, + "loss": 0.1089, + "step": 261 + }, + { + "epoch": 6.110787172011662, + "grad_norm": 3.596662585020769, + "learning_rate": 4.032725027661601e-06, + "loss": 0.1055, + "step": 262 + }, + { + "epoch": 6.134110787172012, + "grad_norm": 2.457646473335156, + "learning_rate": 4.025091090715931e-06, + "loss": 0.1096, + "step": 263 + }, + { + "epoch": 6.1574344023323615, + "grad_norm": 1.7432745035698631, + "learning_rate": 4.017434437021206e-06, + "loss": 0.0979, + "step": 264 + }, + { + "epoch": 6.180758017492711, + "grad_norm": 2.936327111671702, + "learning_rate": 4.009755180625878e-06, + "loss": 0.1071, + "step": 265 + }, + { + "epoch": 6.204081632653061, + "grad_norm": 1.7878964551347618, + "learning_rate": 4.002053435915078e-06, + "loss": 0.094, + "step": 266 + }, + { + "epoch": 6.227405247813411, + "grad_norm": 3.0620629134937523, + "learning_rate": 3.9943293176089036e-06, + "loss": 0.1029, + "step": 267 + }, + { + "epoch": 6.2507288629737605, + "grad_norm": 1.9573953560132995, + "learning_rate": 3.986582940760717e-06, + "loss": 0.1013, + "step": 268 + }, + { + "epoch": 6.274052478134111, + "grad_norm": 3.61949766140591, + "learning_rate": 3.978814420755429e-06, + "loss": 0.1105, + "step": 269 + }, + { + "epoch": 6.297376093294461, + "grad_norm": 2.0103439818635613, + "learning_rate": 3.971023873307781e-06, + "loss": 0.0997, + "step": 270 + }, + { + "epoch": 6.320699708454811, + "grad_norm": 3.672405448096318, + "learning_rate": 3.963211414460618e-06, + "loss": 0.1008, + "step": 271 + }, + { + "epoch": 6.34402332361516, + "grad_norm": 2.144377933990508, + "learning_rate": 3.955377160583165e-06, + "loss": 0.1027, + "step": 272 + }, + { + "epoch": 6.36734693877551, + "grad_norm": 2.7273008554626452, + "learning_rate": 3.94752122836929e-06, + "loss": 0.1015, + "step": 273 + }, + { + "epoch": 6.39067055393586, + "grad_norm": 2.13308524040233, + "learning_rate": 3.939643734835768e-06, + "loss": 0.1033, + "step": 274 + }, + { + "epoch": 6.41399416909621, + "grad_norm": 2.0125298207602946, + "learning_rate": 3.931744797320538e-06, + "loss": 0.1083, + "step": 275 + }, + { + "epoch": 6.437317784256559, + "grad_norm": 1.7281564802782643, + "learning_rate": 3.9238245334809525e-06, + "loss": 0.1013, + "step": 276 + }, + { + "epoch": 6.460641399416909, + "grad_norm": 2.7281331345129924, + "learning_rate": 3.915883061292027e-06, + "loss": 0.1001, + "step": 277 + }, + { + "epoch": 6.48396501457726, + "grad_norm": 1.9362860537497555, + "learning_rate": 3.907920499044684e-06, + "loss": 0.1033, + "step": 278 + }, + { + "epoch": 6.50728862973761, + "grad_norm": 3.094024646012348, + "learning_rate": 3.899936965343989e-06, + "loss": 0.1088, + "step": 279 + }, + { + "epoch": 6.530612244897959, + "grad_norm": 3.0693157412634937, + "learning_rate": 3.891932579107384e-06, + "loss": 0.1002, + "step": 280 + }, + { + "epoch": 6.553935860058309, + "grad_norm": 1.811589768832765, + "learning_rate": 3.883907459562916e-06, + "loss": 0.097, + "step": 281 + }, + { + "epoch": 6.577259475218659, + "grad_norm": 2.5573629665858912, + "learning_rate": 3.875861726247464e-06, + "loss": 0.1061, + "step": 282 + }, + { + "epoch": 6.600583090379009, + "grad_norm": 3.1377447025271676, + "learning_rate": 3.867795499004954e-06, + "loss": 0.1006, + "step": 283 + }, + { + "epoch": 6.623906705539358, + "grad_norm": 2.754614339929124, + "learning_rate": 3.859708897984575e-06, + "loss": 0.105, + "step": 284 + }, + { + "epoch": 6.647230320699708, + "grad_norm": 2.6639057012599148, + "learning_rate": 3.8516020436389945e-06, + "loss": 0.1049, + "step": 285 + }, + { + "epoch": 6.670553935860058, + "grad_norm": 3.6023054370319465, + "learning_rate": 3.843475056722555e-06, + "loss": 0.1028, + "step": 286 + }, + { + "epoch": 6.6938775510204085, + "grad_norm": 3.266025062404138, + "learning_rate": 3.835328058289486e-06, + "loss": 0.1078, + "step": 287 + }, + { + "epoch": 6.717201166180758, + "grad_norm": 4.453189889535449, + "learning_rate": 3.82716116969209e-06, + "loss": 0.1047, + "step": 288 + }, + { + "epoch": 6.740524781341108, + "grad_norm": 5.333045812645013, + "learning_rate": 3.818974512578943e-06, + "loss": 0.1071, + "step": 289 + }, + { + "epoch": 6.763848396501458, + "grad_norm": 3.2006720850408077, + "learning_rate": 3.8107682088930797e-06, + "loss": 0.1065, + "step": 290 + }, + { + "epoch": 6.7871720116618075, + "grad_norm": 5.145686385595828, + "learning_rate": 3.802542380870177e-06, + "loss": 0.1045, + "step": 291 + }, + { + "epoch": 6.810495626822157, + "grad_norm": 3.6186906562643433, + "learning_rate": 3.794297151036732e-06, + "loss": 0.0996, + "step": 292 + }, + { + "epoch": 6.833819241982507, + "grad_norm": 5.063709820898958, + "learning_rate": 3.7860326422082417e-06, + "loss": 0.107, + "step": 293 + }, + { + "epoch": 6.857142857142857, + "grad_norm": 4.260079716397743, + "learning_rate": 3.777748977487366e-06, + "loss": 0.1068, + "step": 294 + }, + { + "epoch": 6.8804664723032065, + "grad_norm": 2.772148240995808, + "learning_rate": 3.7694462802621025e-06, + "loss": 0.1039, + "step": 295 + }, + { + "epoch": 6.903790087463557, + "grad_norm": 2.468481411386976, + "learning_rate": 3.76112467420394e-06, + "loss": 0.101, + "step": 296 + }, + { + "epoch": 6.927113702623907, + "grad_norm": 2.7704651284383757, + "learning_rate": 3.7527842832660244e-06, + "loss": 0.107, + "step": 297 + }, + { + "epoch": 6.950437317784257, + "grad_norm": 3.261999572725812, + "learning_rate": 3.744425231681308e-06, + "loss": 0.1054, + "step": 298 + }, + { + "epoch": 6.973760932944606, + "grad_norm": 3.8333346048988775, + "learning_rate": 3.7360476439606984e-06, + "loss": 0.1059, + "step": 299 + }, + { + "epoch": 6.997084548104956, + "grad_norm": 3.781950704421763, + "learning_rate": 3.727651644891207e-06, + "loss": 0.1038, + "step": 300 + }, + { + "epoch": 7.020408163265306, + "grad_norm": 3.670002951140088, + "learning_rate": 3.719237359534087e-06, + "loss": 0.1056, + "step": 301 + }, + { + "epoch": 7.043731778425656, + "grad_norm": 4.066117070274146, + "learning_rate": 3.710804913222972e-06, + "loss": 0.0988, + "step": 302 + }, + { + "epoch": 7.067055393586005, + "grad_norm": 3.10789584915452, + "learning_rate": 3.702354431562011e-06, + "loss": 0.1054, + "step": 303 + }, + { + "epoch": 7.090379008746356, + "grad_norm": 2.740386719387549, + "learning_rate": 3.693886040423994e-06, + "loss": 0.1001, + "step": 304 + }, + { + "epoch": 7.113702623906706, + "grad_norm": 3.473383377895931, + "learning_rate": 3.6853998659484784e-06, + "loss": 0.0988, + "step": 305 + }, + { + "epoch": 7.137026239067056, + "grad_norm": 3.3568900920476814, + "learning_rate": 3.676896034539913e-06, + "loss": 0.1002, + "step": 306 + }, + { + "epoch": 7.160349854227405, + "grad_norm": 5.428025332926568, + "learning_rate": 3.668374672865749e-06, + "loss": 0.1043, + "step": 307 + }, + { + "epoch": 7.183673469387755, + "grad_norm": 4.99524787142711, + "learning_rate": 3.6598359078545597e-06, + "loss": 0.0998, + "step": 308 + }, + { + "epoch": 7.206997084548105, + "grad_norm": 4.265821291700417, + "learning_rate": 3.6512798666941457e-06, + "loss": 0.0973, + "step": 309 + }, + { + "epoch": 7.230320699708455, + "grad_norm": 4.835380559104649, + "learning_rate": 3.6427066768296425e-06, + "loss": 0.1047, + "step": 310 + }, + { + "epoch": 7.253644314868804, + "grad_norm": 3.131377312362004, + "learning_rate": 3.634116465961621e-06, + "loss": 0.0918, + "step": 311 + }, + { + "epoch": 7.276967930029155, + "grad_norm": 3.518209092623886, + "learning_rate": 3.6255093620441835e-06, + "loss": 0.0972, + "step": 312 + }, + { + "epoch": 7.300291545189505, + "grad_norm": 4.999014757360847, + "learning_rate": 3.6168854932830643e-06, + "loss": 0.0989, + "step": 313 + }, + { + "epoch": 7.3236151603498545, + "grad_norm": 2.90469532063187, + "learning_rate": 3.6082449881337132e-06, + "loss": 0.102, + "step": 314 + }, + { + "epoch": 7.346938775510204, + "grad_norm": 1.9017338019692047, + "learning_rate": 3.5995879752993846e-06, + "loss": 0.0919, + "step": 315 + }, + { + "epoch": 7.370262390670554, + "grad_norm": 4.74281611039079, + "learning_rate": 3.5909145837292207e-06, + "loss": 0.1062, + "step": 316 + }, + { + "epoch": 7.393586005830904, + "grad_norm": 2.7998177162508595, + "learning_rate": 3.5822249426163302e-06, + "loss": 0.0975, + "step": 317 + }, + { + "epoch": 7.4169096209912535, + "grad_norm": 5.268248676962246, + "learning_rate": 3.5735191813958657e-06, + "loss": 0.0952, + "step": 318 + }, + { + "epoch": 7.440233236151603, + "grad_norm": 2.514032938174696, + "learning_rate": 3.564797429743093e-06, + "loss": 0.0954, + "step": 319 + }, + { + "epoch": 7.463556851311953, + "grad_norm": 4.346782915358994, + "learning_rate": 3.5560598175714617e-06, + "loss": 0.0905, + "step": 320 + }, + { + "epoch": 7.486880466472304, + "grad_norm": 2.662355743225369, + "learning_rate": 3.5473064750306675e-06, + "loss": 0.1002, + "step": 321 + }, + { + "epoch": 7.510204081632653, + "grad_norm": 2.6526705690815224, + "learning_rate": 3.5385375325047167e-06, + "loss": 0.0941, + "step": 322 + }, + { + "epoch": 7.533527696793003, + "grad_norm": 3.1377681055301125, + "learning_rate": 3.529753120609982e-06, + "loss": 0.1013, + "step": 323 + }, + { + "epoch": 7.556851311953353, + "grad_norm": 2.8516907545536796, + "learning_rate": 3.520953370193259e-06, + "loss": 0.0943, + "step": 324 + }, + { + "epoch": 7.580174927113703, + "grad_norm": 2.2692649277972192, + "learning_rate": 3.5121384123298137e-06, + "loss": 0.0987, + "step": 325 + }, + { + "epoch": 7.603498542274052, + "grad_norm": 3.201499414823122, + "learning_rate": 3.5033083783214334e-06, + "loss": 0.0941, + "step": 326 + }, + { + "epoch": 7.626822157434402, + "grad_norm": 2.3873360143137075, + "learning_rate": 3.4944633996944705e-06, + "loss": 0.0975, + "step": 327 + }, + { + "epoch": 7.650145772594752, + "grad_norm": 3.9439172676028953, + "learning_rate": 3.48560360819788e-06, + "loss": 0.0965, + "step": 328 + }, + { + "epoch": 7.673469387755102, + "grad_norm": 3.0373265032413754, + "learning_rate": 3.4767291358012623e-06, + "loss": 0.0925, + "step": 329 + }, + { + "epoch": 7.696793002915452, + "grad_norm": 2.539015737249666, + "learning_rate": 3.467840114692894e-06, + "loss": 0.0926, + "step": 330 + }, + { + "epoch": 7.720116618075802, + "grad_norm": 3.7437097563196575, + "learning_rate": 3.4589366772777578e-06, + "loss": 0.0946, + "step": 331 + }, + { + "epoch": 7.743440233236152, + "grad_norm": 2.9377775669851274, + "learning_rate": 3.450018956175575e-06, + "loss": 0.0987, + "step": 332 + }, + { + "epoch": 7.766763848396502, + "grad_norm": 2.98638049741519, + "learning_rate": 3.441087084218826e-06, + "loss": 0.0954, + "step": 333 + }, + { + "epoch": 7.790087463556851, + "grad_norm": 3.609350559432606, + "learning_rate": 3.432141194450772e-06, + "loss": 0.1007, + "step": 334 + }, + { + "epoch": 7.813411078717201, + "grad_norm": 3.2372500407812317, + "learning_rate": 3.4231814201234763e-06, + "loss": 0.0984, + "step": 335 + }, + { + "epoch": 7.836734693877551, + "grad_norm": 4.127296964202223, + "learning_rate": 3.414207894695816e-06, + "loss": 0.1012, + "step": 336 + }, + { + "epoch": 7.860058309037901, + "grad_norm": 3.1118203520319603, + "learning_rate": 3.405220751831495e-06, + "loss": 0.1003, + "step": 337 + }, + { + "epoch": 7.88338192419825, + "grad_norm": 3.507953190690225, + "learning_rate": 3.3962201253970563e-06, + "loss": 0.1025, + "step": 338 + }, + { + "epoch": 7.906705539358601, + "grad_norm": 2.2012231071081474, + "learning_rate": 3.387206149459882e-06, + "loss": 0.0965, + "step": 339 + }, + { + "epoch": 7.930029154518951, + "grad_norm": 2.2871527403019467, + "learning_rate": 3.3781789582862e-06, + "loss": 0.0998, + "step": 340 + }, + { + "epoch": 7.9533527696793005, + "grad_norm": 2.4730155888616405, + "learning_rate": 3.369138686339087e-06, + "loss": 0.0945, + "step": 341 + }, + { + "epoch": 7.97667638483965, + "grad_norm": 2.1207877653231675, + "learning_rate": 3.3600854682764585e-06, + "loss": 0.0938, + "step": 342 + }, + { + "epoch": 8.0, + "grad_norm": 2.6214986792515047, + "learning_rate": 3.351019438949068e-06, + "loss": 0.0963, + "step": 343 + }, + { + "epoch": 8.02332361516035, + "grad_norm": 3.084487613063804, + "learning_rate": 3.3419407333984972e-06, + "loss": 0.0839, + "step": 344 + }, + { + "epoch": 8.0466472303207, + "grad_norm": 2.0003748780822432, + "learning_rate": 3.3328494868551444e-06, + "loss": 0.0971, + "step": 345 + }, + { + "epoch": 8.06997084548105, + "grad_norm": 3.204166745374767, + "learning_rate": 3.3237458347362106e-06, + "loss": 0.0931, + "step": 346 + }, + { + "epoch": 8.093294460641399, + "grad_norm": 2.2486814238019472, + "learning_rate": 3.314629912643682e-06, + "loss": 0.0896, + "step": 347 + }, + { + "epoch": 8.11661807580175, + "grad_norm": 2.2590096543685405, + "learning_rate": 3.3055018563623085e-06, + "loss": 0.0857, + "step": 348 + }, + { + "epoch": 8.139941690962099, + "grad_norm": 3.2766070259645264, + "learning_rate": 3.296361801857587e-06, + "loss": 0.0916, + "step": 349 + }, + { + "epoch": 8.16326530612245, + "grad_norm": 2.2460306971448745, + "learning_rate": 3.2872098852737274e-06, + "loss": 0.0839, + "step": 350 + }, + { + "epoch": 8.186588921282798, + "grad_norm": 2.7558318636228667, + "learning_rate": 3.2780462429316317e-06, + "loss": 0.0878, + "step": 351 + }, + { + "epoch": 8.209912536443149, + "grad_norm": 3.1544483443701856, + "learning_rate": 3.268871011326861e-06, + "loss": 0.0867, + "step": 352 + }, + { + "epoch": 8.2332361516035, + "grad_norm": 2.236180676259461, + "learning_rate": 3.2596843271276003e-06, + "loss": 0.0888, + "step": 353 + }, + { + "epoch": 8.256559766763848, + "grad_norm": 4.011686657384414, + "learning_rate": 3.2504863271726284e-06, + "loss": 0.0821, + "step": 354 + }, + { + "epoch": 8.279883381924199, + "grad_norm": 2.5203855027411937, + "learning_rate": 3.241277148469274e-06, + "loss": 0.0942, + "step": 355 + }, + { + "epoch": 8.303206997084548, + "grad_norm": 3.540099945597143, + "learning_rate": 3.232056928191376e-06, + "loss": 0.0866, + "step": 356 + }, + { + "epoch": 8.326530612244898, + "grad_norm": 3.43864726948626, + "learning_rate": 3.2228258036772443e-06, + "loss": 0.0894, + "step": 357 + }, + { + "epoch": 8.349854227405247, + "grad_norm": 4.373034444419023, + "learning_rate": 3.213583912427609e-06, + "loss": 0.089, + "step": 358 + }, + { + "epoch": 8.373177842565598, + "grad_norm": 3.7748671956733397, + "learning_rate": 3.2043313921035747e-06, + "loss": 0.0857, + "step": 359 + }, + { + "epoch": 8.396501457725947, + "grad_norm": 4.151535917347032, + "learning_rate": 3.195068380524569e-06, + "loss": 0.0915, + "step": 360 + }, + { + "epoch": 8.419825072886297, + "grad_norm": 4.7744553100527005, + "learning_rate": 3.185795015666292e-06, + "loss": 0.0932, + "step": 361 + }, + { + "epoch": 8.443148688046648, + "grad_norm": 3.8892753742055044, + "learning_rate": 3.1765114356586592e-06, + "loss": 0.0848, + "step": 362 + }, + { + "epoch": 8.466472303206997, + "grad_norm": 3.993048468689458, + "learning_rate": 3.1672177787837404e-06, + "loss": 0.0863, + "step": 363 + }, + { + "epoch": 8.489795918367347, + "grad_norm": 3.4051738753996252, + "learning_rate": 3.1579141834737106e-06, + "loss": 0.0843, + "step": 364 + }, + { + "epoch": 8.513119533527696, + "grad_norm": 4.425378712111029, + "learning_rate": 3.1486007883087754e-06, + "loss": 0.0951, + "step": 365 + }, + { + "epoch": 8.536443148688047, + "grad_norm": 2.674656541112342, + "learning_rate": 3.139277732015114e-06, + "loss": 0.087, + "step": 366 + }, + { + "epoch": 8.559766763848396, + "grad_norm": 4.3155606231196995, + "learning_rate": 3.1299451534628134e-06, + "loss": 0.0886, + "step": 367 + }, + { + "epoch": 8.583090379008746, + "grad_norm": 2.4973769732503333, + "learning_rate": 3.1206031916637956e-06, + "loss": 0.0886, + "step": 368 + }, + { + "epoch": 8.606413994169095, + "grad_norm": 4.782817708224153, + "learning_rate": 3.1112519857697505e-06, + "loss": 0.0923, + "step": 369 + }, + { + "epoch": 8.629737609329446, + "grad_norm": 2.5062794976151435, + "learning_rate": 3.1018916750700618e-06, + "loss": 0.0911, + "step": 370 + }, + { + "epoch": 8.653061224489797, + "grad_norm": 5.905723588685558, + "learning_rate": 3.092522398989732e-06, + "loss": 0.0986, + "step": 371 + }, + { + "epoch": 8.676384839650146, + "grad_norm": 3.567470095214398, + "learning_rate": 3.0831442970873044e-06, + "loss": 0.0909, + "step": 372 + }, + { + "epoch": 8.699708454810496, + "grad_norm": 3.8365983879273293, + "learning_rate": 3.0737575090527892e-06, + "loss": 0.0881, + "step": 373 + }, + { + "epoch": 8.723032069970845, + "grad_norm": 5.570185292036708, + "learning_rate": 3.0643621747055785e-06, + "loss": 0.0901, + "step": 374 + }, + { + "epoch": 8.746355685131196, + "grad_norm": 4.025989027119222, + "learning_rate": 3.054958433992362e-06, + "loss": 0.0855, + "step": 375 + }, + { + "epoch": 8.769679300291545, + "grad_norm": 4.030985505646047, + "learning_rate": 3.045546426985049e-06, + "loss": 0.0852, + "step": 376 + }, + { + "epoch": 8.793002915451895, + "grad_norm": 4.022543561567968, + "learning_rate": 3.036126293878674e-06, + "loss": 0.093, + "step": 377 + }, + { + "epoch": 8.816326530612244, + "grad_norm": 4.828598838689659, + "learning_rate": 3.026698174989316e-06, + "loss": 0.0953, + "step": 378 + }, + { + "epoch": 8.839650145772595, + "grad_norm": 4.2944264596746455, + "learning_rate": 3.0172622107520033e-06, + "loss": 0.089, + "step": 379 + }, + { + "epoch": 8.862973760932945, + "grad_norm": 4.049298242123209, + "learning_rate": 3.0078185417186245e-06, + "loss": 0.0978, + "step": 380 + }, + { + "epoch": 8.886297376093294, + "grad_norm": 3.046707696478407, + "learning_rate": 2.9983673085558306e-06, + "loss": 0.0924, + "step": 381 + }, + { + "epoch": 8.909620991253645, + "grad_norm": 4.541947988960665, + "learning_rate": 2.988908652042948e-06, + "loss": 0.0902, + "step": 382 + }, + { + "epoch": 8.932944606413994, + "grad_norm": 2.7874887337906, + "learning_rate": 2.979442713069871e-06, + "loss": 0.0892, + "step": 383 + }, + { + "epoch": 8.956268221574344, + "grad_norm": 4.904390407872027, + "learning_rate": 2.9699696326349726e-06, + "loss": 0.0928, + "step": 384 + }, + { + "epoch": 8.979591836734693, + "grad_norm": 4.009865753514206, + "learning_rate": 2.960489551842998e-06, + "loss": 0.0899, + "step": 385 + }, + { + "epoch": 9.002915451895044, + "grad_norm": 3.724178039187428, + "learning_rate": 2.9510026119029655e-06, + "loss": 0.0899, + "step": 386 + }, + { + "epoch": 9.026239067055394, + "grad_norm": 4.384561172850345, + "learning_rate": 2.9415089541260645e-06, + "loss": 0.0854, + "step": 387 + }, + { + "epoch": 9.049562682215743, + "grad_norm": 2.1881553758061685, + "learning_rate": 2.9320087199235463e-06, + "loss": 0.0771, + "step": 388 + }, + { + "epoch": 9.072886297376094, + "grad_norm": 4.409817597299811, + "learning_rate": 2.9225020508046233e-06, + "loss": 0.0832, + "step": 389 + }, + { + "epoch": 9.096209912536443, + "grad_norm": 3.108449827027684, + "learning_rate": 2.9129890883743544e-06, + "loss": 0.0796, + "step": 390 + }, + { + "epoch": 9.119533527696793, + "grad_norm": 3.7741277785995506, + "learning_rate": 2.9034699743315418e-06, + "loss": 0.0751, + "step": 391 + }, + { + "epoch": 9.142857142857142, + "grad_norm": 5.538458658079067, + "learning_rate": 2.893944850466619e-06, + "loss": 0.0891, + "step": 392 + }, + { + "epoch": 9.166180758017493, + "grad_norm": 4.107660784555083, + "learning_rate": 2.8844138586595354e-06, + "loss": 0.0809, + "step": 393 + }, + { + "epoch": 9.189504373177842, + "grad_norm": 5.161974324578949, + "learning_rate": 2.8748771408776467e-06, + "loss": 0.0766, + "step": 394 + }, + { + "epoch": 9.212827988338192, + "grad_norm": 3.5131532923575977, + "learning_rate": 2.8653348391735996e-06, + "loss": 0.0808, + "step": 395 + }, + { + "epoch": 9.236151603498543, + "grad_norm": 7.6199699445316975, + "learning_rate": 2.8557870956832135e-06, + "loss": 0.0853, + "step": 396 + }, + { + "epoch": 9.259475218658892, + "grad_norm": 4.24358624464005, + "learning_rate": 2.8462340526233657e-06, + "loss": 0.082, + "step": 397 + }, + { + "epoch": 9.282798833819243, + "grad_norm": 5.131989899420963, + "learning_rate": 2.8366758522898737e-06, + "loss": 0.0797, + "step": 398 + }, + { + "epoch": 9.306122448979592, + "grad_norm": 4.042970637076854, + "learning_rate": 2.8271126370553727e-06, + "loss": 0.0766, + "step": 399 + }, + { + "epoch": 9.329446064139942, + "grad_norm": 6.451199962567526, + "learning_rate": 2.817544549367197e-06, + "loss": 0.0812, + "step": 400 + }, + { + "epoch": 9.352769679300291, + "grad_norm": 4.460602275057708, + "learning_rate": 2.8079717317452582e-06, + "loss": 0.0835, + "step": 401 + }, + { + "epoch": 9.376093294460642, + "grad_norm": 4.474716365072937, + "learning_rate": 2.798394326779922e-06, + "loss": 0.0828, + "step": 402 + }, + { + "epoch": 9.39941690962099, + "grad_norm": 5.622483601900293, + "learning_rate": 2.788812477129883e-06, + "loss": 0.0843, + "step": 403 + }, + { + "epoch": 9.422740524781341, + "grad_norm": 3.953290973541459, + "learning_rate": 2.7792263255200406e-06, + "loss": 0.0844, + "step": 404 + }, + { + "epoch": 9.446064139941692, + "grad_norm": 5.3449177497837255, + "learning_rate": 2.769636014739377e-06, + "loss": 0.0873, + "step": 405 + }, + { + "epoch": 9.46938775510204, + "grad_norm": 4.633708487675383, + "learning_rate": 2.760041687638822e-06, + "loss": 0.0816, + "step": 406 + }, + { + "epoch": 9.492711370262391, + "grad_norm": 3.38289326842269, + "learning_rate": 2.7504434871291317e-06, + "loss": 0.083, + "step": 407 + }, + { + "epoch": 9.51603498542274, + "grad_norm": 5.207711063087272, + "learning_rate": 2.7408415561787587e-06, + "loss": 0.0797, + "step": 408 + }, + { + "epoch": 9.53935860058309, + "grad_norm": 4.7282328530949, + "learning_rate": 2.7312360378117214e-06, + "loss": 0.0791, + "step": 409 + }, + { + "epoch": 9.56268221574344, + "grad_norm": 3.9839978806287037, + "learning_rate": 2.721627075105473e-06, + "loss": 0.0789, + "step": 410 + }, + { + "epoch": 9.58600583090379, + "grad_norm": 4.795922597965258, + "learning_rate": 2.7120148111887732e-06, + "loss": 0.078, + "step": 411 + }, + { + "epoch": 9.60932944606414, + "grad_norm": 3.7271565537628444, + "learning_rate": 2.7023993892395523e-06, + "loss": 0.0809, + "step": 412 + }, + { + "epoch": 9.63265306122449, + "grad_norm": 4.174873788719055, + "learning_rate": 2.6927809524827815e-06, + "loss": 0.0769, + "step": 413 + }, + { + "epoch": 9.65597667638484, + "grad_norm": 4.259366205196579, + "learning_rate": 2.6831596441883388e-06, + "loss": 0.0825, + "step": 414 + }, + { + "epoch": 9.67930029154519, + "grad_norm": 5.37835783645601, + "learning_rate": 2.6735356076688744e-06, + "loss": 0.0827, + "step": 415 + }, + { + "epoch": 9.70262390670554, + "grad_norm": 3.7147173132975277, + "learning_rate": 2.6639089862776763e-06, + "loss": 0.0821, + "step": 416 + }, + { + "epoch": 9.725947521865889, + "grad_norm": 5.772966378570507, + "learning_rate": 2.6542799234065354e-06, + "loss": 0.0811, + "step": 417 + }, + { + "epoch": 9.74927113702624, + "grad_norm": 4.34521503579229, + "learning_rate": 2.644648562483608e-06, + "loss": 0.0874, + "step": 418 + }, + { + "epoch": 9.772594752186588, + "grad_norm": 3.6215374951442034, + "learning_rate": 2.6350150469712824e-06, + "loss": 0.0815, + "step": 419 + }, + { + "epoch": 9.795918367346939, + "grad_norm": 3.4450949256342502, + "learning_rate": 2.625379520364041e-06, + "loss": 0.0768, + "step": 420 + }, + { + "epoch": 9.819241982507288, + "grad_norm": 3.95042321074671, + "learning_rate": 2.615742126186319e-06, + "loss": 0.0776, + "step": 421 + }, + { + "epoch": 9.842565597667638, + "grad_norm": 3.0236686992714827, + "learning_rate": 2.606103007990371e-06, + "loss": 0.0796, + "step": 422 + }, + { + "epoch": 9.86588921282799, + "grad_norm": 4.361007111450639, + "learning_rate": 2.5964623093541326e-06, + "loss": 0.0857, + "step": 423 + }, + { + "epoch": 9.889212827988338, + "grad_norm": 3.0417992201536803, + "learning_rate": 2.5868201738790798e-06, + "loss": 0.0793, + "step": 424 + }, + { + "epoch": 9.912536443148689, + "grad_norm": 4.234482429521092, + "learning_rate": 2.5771767451880908e-06, + "loss": 0.0844, + "step": 425 + }, + { + "epoch": 9.935860058309038, + "grad_norm": 3.2532366261674546, + "learning_rate": 2.567532166923306e-06, + "loss": 0.0845, + "step": 426 + }, + { + "epoch": 9.959183673469388, + "grad_norm": 5.25009409502266, + "learning_rate": 2.557886582743991e-06, + "loss": 0.085, + "step": 427 + }, + { + "epoch": 9.982507288629737, + "grad_norm": 3.3120422046939297, + "learning_rate": 2.548240136324392e-06, + "loss": 0.0855, + "step": 428 + }, + { + "epoch": 10.005830903790088, + "grad_norm": 3.6242571102413876, + "learning_rate": 2.5385929713516006e-06, + "loss": 0.0795, + "step": 429 + }, + { + "epoch": 10.029154518950437, + "grad_norm": 3.484857500137509, + "learning_rate": 2.5289452315234125e-06, + "loss": 0.0743, + "step": 430 + }, + { + "epoch": 10.052478134110787, + "grad_norm": 2.9146847303493293, + "learning_rate": 2.5192970605461835e-06, + "loss": 0.0716, + "step": 431 + }, + { + "epoch": 10.075801749271138, + "grad_norm": 4.090638820313027, + "learning_rate": 2.509648602132692e-06, + "loss": 0.071, + "step": 432 + }, + { + "epoch": 10.099125364431487, + "grad_norm": 2.7367985039975364, + "learning_rate": 2.5e-06, + "loss": 0.0653, + "step": 433 + }, + { + "epoch": 10.122448979591837, + "grad_norm": 3.1514711365816175, + "learning_rate": 2.4903513978673076e-06, + "loss": 0.0724, + "step": 434 + }, + { + "epoch": 10.145772594752186, + "grad_norm": 3.1212207066540483, + "learning_rate": 2.480702939453818e-06, + "loss": 0.066, + "step": 435 + }, + { + "epoch": 10.169096209912537, + "grad_norm": 3.5360489366260692, + "learning_rate": 2.4710547684765884e-06, + "loss": 0.0726, + "step": 436 + }, + { + "epoch": 10.192419825072886, + "grad_norm": 4.199846655453183, + "learning_rate": 2.4614070286483994e-06, + "loss": 0.0751, + "step": 437 + }, + { + "epoch": 10.215743440233236, + "grad_norm": 4.2344916633778835, + "learning_rate": 2.451759863675609e-06, + "loss": 0.0673, + "step": 438 + }, + { + "epoch": 10.239067055393585, + "grad_norm": 5.237854487183302, + "learning_rate": 2.44211341725601e-06, + "loss": 0.0785, + "step": 439 + }, + { + "epoch": 10.262390670553936, + "grad_norm": 4.301828643181056, + "learning_rate": 2.432467833076694e-06, + "loss": 0.0773, + "step": 440 + }, + { + "epoch": 10.285714285714286, + "grad_norm": 4.474031968078861, + "learning_rate": 2.42282325481191e-06, + "loss": 0.0737, + "step": 441 + }, + { + "epoch": 10.309037900874635, + "grad_norm": 6.017173349843788, + "learning_rate": 2.4131798261209206e-06, + "loss": 0.078, + "step": 442 + }, + { + "epoch": 10.332361516034986, + "grad_norm": 3.7621797495056417, + "learning_rate": 2.4035376906458674e-06, + "loss": 0.0704, + "step": 443 + }, + { + "epoch": 10.355685131195335, + "grad_norm": 5.9349864257040235, + "learning_rate": 2.39389699200963e-06, + "loss": 0.0719, + "step": 444 + }, + { + "epoch": 10.379008746355685, + "grad_norm": 4.943076738697602, + "learning_rate": 2.3842578738136816e-06, + "loss": 0.07, + "step": 445 + }, + { + "epoch": 10.402332361516034, + "grad_norm": 4.132956720442613, + "learning_rate": 2.374620479635959e-06, + "loss": 0.0707, + "step": 446 + }, + { + "epoch": 10.425655976676385, + "grad_norm": 3.519182634605719, + "learning_rate": 2.364984953028718e-06, + "loss": 0.0669, + "step": 447 + }, + { + "epoch": 10.448979591836734, + "grad_norm": 4.880062116008721, + "learning_rate": 2.3553514375163926e-06, + "loss": 0.0716, + "step": 448 + }, + { + "epoch": 10.472303206997085, + "grad_norm": 3.6519392636148584, + "learning_rate": 2.3457200765934654e-06, + "loss": 0.0702, + "step": 449 + }, + { + "epoch": 10.495626822157435, + "grad_norm": 3.9433212845253585, + "learning_rate": 2.3360910137223246e-06, + "loss": 0.0711, + "step": 450 + }, + { + "epoch": 10.518950437317784, + "grad_norm": 4.1091062508662635, + "learning_rate": 2.326464392331126e-06, + "loss": 0.0719, + "step": 451 + }, + { + "epoch": 10.542274052478135, + "grad_norm": 4.652424924512233, + "learning_rate": 2.3168403558116612e-06, + "loss": 0.0723, + "step": 452 + }, + { + "epoch": 10.565597667638484, + "grad_norm": 3.4455140070040255, + "learning_rate": 2.3072190475172194e-06, + "loss": 0.07, + "step": 453 + }, + { + "epoch": 10.588921282798834, + "grad_norm": 4.649570871876516, + "learning_rate": 2.297600610760448e-06, + "loss": 0.0696, + "step": 454 + }, + { + "epoch": 10.612244897959183, + "grad_norm": 4.559645922784148, + "learning_rate": 2.287985188811228e-06, + "loss": 0.0746, + "step": 455 + }, + { + "epoch": 10.635568513119534, + "grad_norm": 3.5444974618957277, + "learning_rate": 2.2783729248945275e-06, + "loss": 0.0684, + "step": 456 + }, + { + "epoch": 10.658892128279884, + "grad_norm": 3.993887164367774, + "learning_rate": 2.2687639621882795e-06, + "loss": 0.0712, + "step": 457 + }, + { + "epoch": 10.682215743440233, + "grad_norm": 3.1108405718415604, + "learning_rate": 2.259158443821242e-06, + "loss": 0.0744, + "step": 458 + }, + { + "epoch": 10.705539358600584, + "grad_norm": 3.3980853018828223, + "learning_rate": 2.249556512870869e-06, + "loss": 0.0686, + "step": 459 + }, + { + "epoch": 10.728862973760933, + "grad_norm": 4.093233731162892, + "learning_rate": 2.2399583123611788e-06, + "loss": 0.0788, + "step": 460 + }, + { + "epoch": 10.752186588921283, + "grad_norm": 3.8432637969379626, + "learning_rate": 2.2303639852606244e-06, + "loss": 0.0694, + "step": 461 + }, + { + "epoch": 10.775510204081632, + "grad_norm": 3.883530609839273, + "learning_rate": 2.2207736744799598e-06, + "loss": 0.0703, + "step": 462 + }, + { + "epoch": 10.798833819241983, + "grad_norm": 4.4013742294915374, + "learning_rate": 2.2111875228701175e-06, + "loss": 0.0765, + "step": 463 + }, + { + "epoch": 10.822157434402332, + "grad_norm": 4.486608288079545, + "learning_rate": 2.2016056732200794e-06, + "loss": 0.0733, + "step": 464 + }, + { + "epoch": 10.845481049562682, + "grad_norm": 3.687365383988495, + "learning_rate": 2.192028268254742e-06, + "loss": 0.0683, + "step": 465 + }, + { + "epoch": 10.868804664723033, + "grad_norm": 4.698752566431009, + "learning_rate": 2.1824554506328033e-06, + "loss": 0.0715, + "step": 466 + }, + { + "epoch": 10.892128279883382, + "grad_norm": 3.1651675708156075, + "learning_rate": 2.1728873629446277e-06, + "loss": 0.0674, + "step": 467 + }, + { + "epoch": 10.915451895043732, + "grad_norm": 3.9151451456147153, + "learning_rate": 2.163324147710127e-06, + "loss": 0.0784, + "step": 468 + }, + { + "epoch": 10.938775510204081, + "grad_norm": 3.5638025144670378, + "learning_rate": 2.1537659473766347e-06, + "loss": 0.0713, + "step": 469 + }, + { + "epoch": 10.962099125364432, + "grad_norm": 3.6805884033912597, + "learning_rate": 2.1442129043167877e-06, + "loss": 0.0732, + "step": 470 + }, + { + "epoch": 10.98542274052478, + "grad_norm": 4.068187078159665, + "learning_rate": 2.1346651608264012e-06, + "loss": 0.0717, + "step": 471 + }, + { + "epoch": 11.008746355685131, + "grad_norm": 3.5401808969431285, + "learning_rate": 2.1251228591223533e-06, + "loss": 0.0697, + "step": 472 + }, + { + "epoch": 11.03206997084548, + "grad_norm": 4.771005463583638, + "learning_rate": 2.1155861413404655e-06, + "loss": 0.0684, + "step": 473 + }, + { + "epoch": 11.055393586005831, + "grad_norm": 4.2335670516810175, + "learning_rate": 2.1060551495333816e-06, + "loss": 0.0652, + "step": 474 + }, + { + "epoch": 11.078717201166182, + "grad_norm": 3.978324814398319, + "learning_rate": 2.096530025668458e-06, + "loss": 0.0653, + "step": 475 + }, + { + "epoch": 11.10204081632653, + "grad_norm": 3.45924049949899, + "learning_rate": 2.087010911625647e-06, + "loss": 0.0601, + "step": 476 + }, + { + "epoch": 11.125364431486881, + "grad_norm": 3.9286305881356, + "learning_rate": 2.077497949195378e-06, + "loss": 0.0602, + "step": 477 + }, + { + "epoch": 11.14868804664723, + "grad_norm": 4.2209211625918135, + "learning_rate": 2.0679912800764537e-06, + "loss": 0.0682, + "step": 478 + }, + { + "epoch": 11.17201166180758, + "grad_norm": 4.0900836863268735, + "learning_rate": 2.0584910458739367e-06, + "loss": 0.0628, + "step": 479 + }, + { + "epoch": 11.19533527696793, + "grad_norm": 4.350925145481364, + "learning_rate": 2.048997388097035e-06, + "loss": 0.0586, + "step": 480 + }, + { + "epoch": 11.21865889212828, + "grad_norm": 4.567833696107266, + "learning_rate": 2.0395104481570025e-06, + "loss": 0.0649, + "step": 481 + }, + { + "epoch": 11.241982507288629, + "grad_norm": 4.3906584497446906, + "learning_rate": 2.0300303673650286e-06, + "loss": 0.0573, + "step": 482 + }, + { + "epoch": 11.26530612244898, + "grad_norm": 4.47237052171397, + "learning_rate": 2.0205572869301295e-06, + "loss": 0.0574, + "step": 483 + }, + { + "epoch": 11.28862973760933, + "grad_norm": 4.007831536260833, + "learning_rate": 2.0110913479570525e-06, + "loss": 0.061, + "step": 484 + }, + { + "epoch": 11.31195335276968, + "grad_norm": 4.571950452681269, + "learning_rate": 2.00163269144417e-06, + "loss": 0.0579, + "step": 485 + }, + { + "epoch": 11.33527696793003, + "grad_norm": 5.218817430620227, + "learning_rate": 1.9921814582813763e-06, + "loss": 0.0669, + "step": 486 + }, + { + "epoch": 11.358600583090379, + "grad_norm": 3.9881175696012776, + "learning_rate": 1.9827377892479967e-06, + "loss": 0.0601, + "step": 487 + }, + { + "epoch": 11.38192419825073, + "grad_norm": 4.1143174999008325, + "learning_rate": 1.973301825010685e-06, + "loss": 0.062, + "step": 488 + }, + { + "epoch": 11.405247813411078, + "grad_norm": 3.9485244216076754, + "learning_rate": 1.963873706121327e-06, + "loss": 0.061, + "step": 489 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 3.873171959654352, + "learning_rate": 1.9544535730149524e-06, + "loss": 0.0673, + "step": 490 + }, + { + "epoch": 11.451895043731778, + "grad_norm": 5.007394091761698, + "learning_rate": 1.9450415660076388e-06, + "loss": 0.0666, + "step": 491 + }, + { + "epoch": 11.475218658892128, + "grad_norm": 3.8132922490669965, + "learning_rate": 1.9356378252944223e-06, + "loss": 0.0653, + "step": 492 + }, + { + "epoch": 11.498542274052479, + "grad_norm": 4.60560810890988, + "learning_rate": 1.9262424909472103e-06, + "loss": 0.0672, + "step": 493 + }, + { + "epoch": 11.521865889212828, + "grad_norm": 4.209183987798544, + "learning_rate": 1.9168557029126965e-06, + "loss": 0.0644, + "step": 494 + }, + { + "epoch": 11.545189504373178, + "grad_norm": 5.595398388299023, + "learning_rate": 1.9074776010102693e-06, + "loss": 0.0666, + "step": 495 + }, + { + "epoch": 11.568513119533527, + "grad_norm": 4.004167531490067, + "learning_rate": 1.8981083249299393e-06, + "loss": 0.0661, + "step": 496 + }, + { + "epoch": 11.591836734693878, + "grad_norm": 4.686857255738534, + "learning_rate": 1.88874801423025e-06, + "loss": 0.0578, + "step": 497 + }, + { + "epoch": 11.615160349854227, + "grad_norm": 4.644190590495262, + "learning_rate": 1.879396808336205e-06, + "loss": 0.0681, + "step": 498 + }, + { + "epoch": 11.638483965014577, + "grad_norm": 5.0449793417652256, + "learning_rate": 1.8700548465371877e-06, + "loss": 0.0604, + "step": 499 + }, + { + "epoch": 11.661807580174926, + "grad_norm": 4.855446751760304, + "learning_rate": 1.860722267984887e-06, + "loss": 0.0648, + "step": 500 + }, + { + "epoch": 11.685131195335277, + "grad_norm": 5.429052345943996, + "learning_rate": 1.8513992116912254e-06, + "loss": 0.0625, + "step": 501 + }, + { + "epoch": 11.708454810495628, + "grad_norm": 4.2187151872116875, + "learning_rate": 1.8420858165262905e-06, + "loss": 0.0598, + "step": 502 + }, + { + "epoch": 11.731778425655977, + "grad_norm": 6.771490853490556, + "learning_rate": 1.8327822212162598e-06, + "loss": 0.0674, + "step": 503 + }, + { + "epoch": 11.755102040816327, + "grad_norm": 5.9975504643367215, + "learning_rate": 1.823488564341342e-06, + "loss": 0.0599, + "step": 504 + }, + { + "epoch": 11.778425655976676, + "grad_norm": 4.1364213442135265, + "learning_rate": 1.814204984333709e-06, + "loss": 0.067, + "step": 505 + }, + { + "epoch": 11.801749271137027, + "grad_norm": 5.912170026737122, + "learning_rate": 1.8049316194754319e-06, + "loss": 0.0668, + "step": 506 + }, + { + "epoch": 11.825072886297376, + "grad_norm": 5.308482008228855, + "learning_rate": 1.7956686078964257e-06, + "loss": 0.0626, + "step": 507 + }, + { + "epoch": 11.848396501457726, + "grad_norm": 6.158712911027675, + "learning_rate": 1.7864160875723916e-06, + "loss": 0.0675, + "step": 508 + }, + { + "epoch": 11.871720116618075, + "grad_norm": 7.051006045152656, + "learning_rate": 1.7771741963227563e-06, + "loss": 0.0678, + "step": 509 + }, + { + "epoch": 11.895043731778426, + "grad_norm": 4.332760088246268, + "learning_rate": 1.7679430718086244e-06, + "loss": 0.0633, + "step": 510 + }, + { + "epoch": 11.918367346938776, + "grad_norm": 4.451530090621097, + "learning_rate": 1.7587228515307273e-06, + "loss": 0.0623, + "step": 511 + }, + { + "epoch": 11.941690962099125, + "grad_norm": 4.5406241430713665, + "learning_rate": 1.7495136728273722e-06, + "loss": 0.0629, + "step": 512 + }, + { + "epoch": 11.965014577259476, + "grad_norm": 3.310623933238498, + "learning_rate": 1.7403156728724003e-06, + "loss": 0.0532, + "step": 513 + }, + { + "epoch": 11.988338192419825, + "grad_norm": 5.673693320795326, + "learning_rate": 1.7311289886731408e-06, + "loss": 0.0648, + "step": 514 + }, + { + "epoch": 12.011661807580175, + "grad_norm": 3.6331444373300714, + "learning_rate": 1.7219537570683692e-06, + "loss": 0.0605, + "step": 515 + }, + { + "epoch": 12.034985422740524, + "grad_norm": 4.178597167218763, + "learning_rate": 1.7127901147262732e-06, + "loss": 0.0533, + "step": 516 + }, + { + "epoch": 12.058309037900875, + "grad_norm": 4.026501838603133, + "learning_rate": 1.7036381981424141e-06, + "loss": 0.0542, + "step": 517 + }, + { + "epoch": 12.081632653061224, + "grad_norm": 5.886567435949283, + "learning_rate": 1.6944981436376917e-06, + "loss": 0.0576, + "step": 518 + }, + { + "epoch": 12.104956268221574, + "grad_norm": 4.151810189217474, + "learning_rate": 1.6853700873563188e-06, + "loss": 0.0551, + "step": 519 + }, + { + "epoch": 12.128279883381925, + "grad_norm": 5.477245076043875, + "learning_rate": 1.6762541652637904e-06, + "loss": 0.0581, + "step": 520 + }, + { + "epoch": 12.151603498542274, + "grad_norm": 5.0774749809336175, + "learning_rate": 1.6671505131448562e-06, + "loss": 0.056, + "step": 521 + }, + { + "epoch": 12.174927113702624, + "grad_norm": 3.601653058616095, + "learning_rate": 1.6580592666015034e-06, + "loss": 0.052, + "step": 522 + }, + { + "epoch": 12.198250728862973, + "grad_norm": 4.176157150264005, + "learning_rate": 1.648980561050933e-06, + "loss": 0.0528, + "step": 523 + }, + { + "epoch": 12.221574344023324, + "grad_norm": 3.556338687588662, + "learning_rate": 1.6399145317235422e-06, + "loss": 0.0538, + "step": 524 + }, + { + "epoch": 12.244897959183673, + "grad_norm": 3.880914911372885, + "learning_rate": 1.630861313660913e-06, + "loss": 0.0493, + "step": 525 + }, + { + "epoch": 12.268221574344023, + "grad_norm": 4.8554685535875866, + "learning_rate": 1.6218210417138007e-06, + "loss": 0.0491, + "step": 526 + }, + { + "epoch": 12.291545189504372, + "grad_norm": 4.2655409360993195, + "learning_rate": 1.6127938505401192e-06, + "loss": 0.0531, + "step": 527 + }, + { + "epoch": 12.314868804664723, + "grad_norm": 3.9912333981353574, + "learning_rate": 1.6037798746029444e-06, + "loss": 0.0565, + "step": 528 + }, + { + "epoch": 12.338192419825074, + "grad_norm": 3.8208239628390213, + "learning_rate": 1.5947792481685054e-06, + "loss": 0.0479, + "step": 529 + }, + { + "epoch": 12.361516034985423, + "grad_norm": 4.109928456447175, + "learning_rate": 1.5857921053041852e-06, + "loss": 0.0555, + "step": 530 + }, + { + "epoch": 12.384839650145773, + "grad_norm": 4.093633984967496, + "learning_rate": 1.5768185798765246e-06, + "loss": 0.0497, + "step": 531 + }, + { + "epoch": 12.408163265306122, + "grad_norm": 4.939878832300203, + "learning_rate": 1.5678588055492289e-06, + "loss": 0.0547, + "step": 532 + }, + { + "epoch": 12.431486880466473, + "grad_norm": 5.319609608201244, + "learning_rate": 1.558912915781175e-06, + "loss": 0.0528, + "step": 533 + }, + { + "epoch": 12.454810495626822, + "grad_norm": 3.872044205066462, + "learning_rate": 1.5499810438244251e-06, + "loss": 0.0563, + "step": 534 + }, + { + "epoch": 12.478134110787172, + "grad_norm": 5.036716500588434, + "learning_rate": 1.5410633227222433e-06, + "loss": 0.0552, + "step": 535 + }, + { + "epoch": 12.501457725947521, + "grad_norm": 4.09470175824976, + "learning_rate": 1.5321598853071068e-06, + "loss": 0.0513, + "step": 536 + }, + { + "epoch": 12.524781341107872, + "grad_norm": 4.375511139574889, + "learning_rate": 1.5232708641987388e-06, + "loss": 0.0491, + "step": 537 + }, + { + "epoch": 12.548104956268222, + "grad_norm": 5.0017332887433765, + "learning_rate": 1.514396391802121e-06, + "loss": 0.0571, + "step": 538 + }, + { + "epoch": 12.571428571428571, + "grad_norm": 3.908119425718553, + "learning_rate": 1.5055366003055305e-06, + "loss": 0.0503, + "step": 539 + }, + { + "epoch": 12.594752186588922, + "grad_norm": 4.2271482164051255, + "learning_rate": 1.496691621678567e-06, + "loss": 0.0564, + "step": 540 + }, + { + "epoch": 12.61807580174927, + "grad_norm": 6.6548142295032715, + "learning_rate": 1.487861587670187e-06, + "loss": 0.0576, + "step": 541 + }, + { + "epoch": 12.641399416909621, + "grad_norm": 6.231794774644079, + "learning_rate": 1.4790466298067415e-06, + "loss": 0.0573, + "step": 542 + }, + { + "epoch": 12.66472303206997, + "grad_norm": 4.344863419575536, + "learning_rate": 1.4702468793900187e-06, + "loss": 0.0534, + "step": 543 + }, + { + "epoch": 12.68804664723032, + "grad_norm": 4.213349000402689, + "learning_rate": 1.4614624674952843e-06, + "loss": 0.0518, + "step": 544 + }, + { + "epoch": 12.71137026239067, + "grad_norm": 7.569114629654876, + "learning_rate": 1.4526935249693336e-06, + "loss": 0.0531, + "step": 545 + }, + { + "epoch": 12.73469387755102, + "grad_norm": 6.067248185900788, + "learning_rate": 1.4439401824285387e-06, + "loss": 0.0548, + "step": 546 + }, + { + "epoch": 12.758017492711371, + "grad_norm": 4.958285270051922, + "learning_rate": 1.4352025702569066e-06, + "loss": 0.0546, + "step": 547 + }, + { + "epoch": 12.78134110787172, + "grad_norm": 5.615875642730984, + "learning_rate": 1.426480818604134e-06, + "loss": 0.0594, + "step": 548 + }, + { + "epoch": 12.80466472303207, + "grad_norm": 4.328415403022571, + "learning_rate": 1.4177750573836702e-06, + "loss": 0.0533, + "step": 549 + }, + { + "epoch": 12.82798833819242, + "grad_norm": 5.138897730272723, + "learning_rate": 1.4090854162707801e-06, + "loss": 0.0496, + "step": 550 + }, + { + "epoch": 12.85131195335277, + "grad_norm": 4.839504029895119, + "learning_rate": 1.4004120247006159e-06, + "loss": 0.0501, + "step": 551 + }, + { + "epoch": 12.874635568513119, + "grad_norm": 3.8022100697475603, + "learning_rate": 1.3917550118662876e-06, + "loss": 0.0513, + "step": 552 + }, + { + "epoch": 12.89795918367347, + "grad_norm": 6.104926711883666, + "learning_rate": 1.383114506716936e-06, + "loss": 0.0523, + "step": 553 + }, + { + "epoch": 12.921282798833818, + "grad_norm": 3.748559929588339, + "learning_rate": 1.3744906379558165e-06, + "loss": 0.0473, + "step": 554 + }, + { + "epoch": 12.944606413994169, + "grad_norm": 5.338422115817136, + "learning_rate": 1.3658835340383809e-06, + "loss": 0.0556, + "step": 555 + }, + { + "epoch": 12.96793002915452, + "grad_norm": 4.161310313444585, + "learning_rate": 1.3572933231703582e-06, + "loss": 0.0544, + "step": 556 + }, + { + "epoch": 12.991253644314869, + "grad_norm": 6.0576026248464725, + "learning_rate": 1.3487201333058543e-06, + "loss": 0.0564, + "step": 557 + }, + { + "epoch": 13.01457725947522, + "grad_norm": 3.6315615233113174, + "learning_rate": 1.3401640921454411e-06, + "loss": 0.0414, + "step": 558 + }, + { + "epoch": 13.037900874635568, + "grad_norm": 3.9125541171392704, + "learning_rate": 1.3316253271342517e-06, + "loss": 0.0441, + "step": 559 + }, + { + "epoch": 13.061224489795919, + "grad_norm": 3.563470139501611, + "learning_rate": 1.3231039654600875e-06, + "loss": 0.0451, + "step": 560 + }, + { + "epoch": 13.084548104956268, + "grad_norm": 4.0771176785890635, + "learning_rate": 1.3146001340515224e-06, + "loss": 0.0468, + "step": 561 + }, + { + "epoch": 13.107871720116618, + "grad_norm": 4.509589468124622, + "learning_rate": 1.306113959576007e-06, + "loss": 0.0473, + "step": 562 + }, + { + "epoch": 13.131195335276967, + "grad_norm": 3.9113416523129456, + "learning_rate": 1.2976455684379895e-06, + "loss": 0.0473, + "step": 563 + }, + { + "epoch": 13.154518950437318, + "grad_norm": 4.7449900658586905, + "learning_rate": 1.2891950867770291e-06, + "loss": 0.0413, + "step": 564 + }, + { + "epoch": 13.177842565597668, + "grad_norm": 4.4217248897886625, + "learning_rate": 1.2807626404659144e-06, + "loss": 0.0472, + "step": 565 + }, + { + "epoch": 13.201166180758017, + "grad_norm": 4.584718298458554, + "learning_rate": 1.272348355108794e-06, + "loss": 0.0462, + "step": 566 + }, + { + "epoch": 13.224489795918368, + "grad_norm": 5.466004298175304, + "learning_rate": 1.263952356039302e-06, + "loss": 0.0476, + "step": 567 + }, + { + "epoch": 13.247813411078717, + "grad_norm": 4.81734965320916, + "learning_rate": 1.2555747683186922e-06, + "loss": 0.0455, + "step": 568 + }, + { + "epoch": 13.271137026239067, + "grad_norm": 5.756185107533244, + "learning_rate": 1.2472157167339754e-06, + "loss": 0.0453, + "step": 569 + }, + { + "epoch": 13.294460641399416, + "grad_norm": 3.822016985198266, + "learning_rate": 1.2388753257960611e-06, + "loss": 0.041, + "step": 570 + }, + { + "epoch": 13.317784256559767, + "grad_norm": 4.691368439559332, + "learning_rate": 1.2305537197378985e-06, + "loss": 0.0426, + "step": 571 + }, + { + "epoch": 13.341107871720117, + "grad_norm": 5.311689068104946, + "learning_rate": 1.222251022512634e-06, + "loss": 0.0445, + "step": 572 + }, + { + "epoch": 13.364431486880466, + "grad_norm": 4.8844713978886185, + "learning_rate": 1.2139673577917594e-06, + "loss": 0.0496, + "step": 573 + }, + { + "epoch": 13.387755102040817, + "grad_norm": 5.524811871808131, + "learning_rate": 1.2057028489632682e-06, + "loss": 0.0471, + "step": 574 + }, + { + "epoch": 13.411078717201166, + "grad_norm": 3.8990317830084007, + "learning_rate": 1.1974576191298235e-06, + "loss": 0.0474, + "step": 575 + }, + { + "epoch": 13.434402332361516, + "grad_norm": 5.0817642083111325, + "learning_rate": 1.1892317911069212e-06, + "loss": 0.0437, + "step": 576 + }, + { + "epoch": 13.457725947521865, + "grad_norm": 6.175434409239747, + "learning_rate": 1.1810254874210578e-06, + "loss": 0.0415, + "step": 577 + }, + { + "epoch": 13.481049562682216, + "grad_norm": 4.105620202665507, + "learning_rate": 1.1728388303079114e-06, + "loss": 0.0398, + "step": 578 + }, + { + "epoch": 13.504373177842565, + "grad_norm": 5.399997160169066, + "learning_rate": 1.1646719417105152e-06, + "loss": 0.0443, + "step": 579 + }, + { + "epoch": 13.527696793002915, + "grad_norm": 5.360588270045755, + "learning_rate": 1.156524943277445e-06, + "loss": 0.0477, + "step": 580 + }, + { + "epoch": 13.551020408163264, + "grad_norm": 3.9622517649396145, + "learning_rate": 1.148397956361007e-06, + "loss": 0.0414, + "step": 581 + }, + { + "epoch": 13.574344023323615, + "grad_norm": 4.5985837663622045, + "learning_rate": 1.1402911020154258e-06, + "loss": 0.0444, + "step": 582 + }, + { + "epoch": 13.597667638483966, + "grad_norm": 4.261820119741356, + "learning_rate": 1.1322045009950475e-06, + "loss": 0.0423, + "step": 583 + }, + { + "epoch": 13.620991253644315, + "grad_norm": 3.7606986971242136, + "learning_rate": 1.1241382737525367e-06, + "loss": 0.0445, + "step": 584 + }, + { + "epoch": 13.644314868804665, + "grad_norm": 5.001124064931982, + "learning_rate": 1.1160925404370851e-06, + "loss": 0.0416, + "step": 585 + }, + { + "epoch": 13.667638483965014, + "grad_norm": 4.126856299517844, + "learning_rate": 1.1080674208926173e-06, + "loss": 0.0398, + "step": 586 + }, + { + "epoch": 13.690962099125365, + "grad_norm": 4.164557622563336, + "learning_rate": 1.1000630346560118e-06, + "loss": 0.0428, + "step": 587 + }, + { + "epoch": 13.714285714285714, + "grad_norm": 4.26993463256259, + "learning_rate": 1.0920795009553164e-06, + "loss": 0.0417, + "step": 588 + }, + { + "epoch": 13.737609329446064, + "grad_norm": 4.325947899111832, + "learning_rate": 1.0841169387079734e-06, + "loss": 0.044, + "step": 589 + }, + { + "epoch": 13.760932944606415, + "grad_norm": 5.762006083937098, + "learning_rate": 1.0761754665190486e-06, + "loss": 0.0447, + "step": 590 + }, + { + "epoch": 13.784256559766764, + "grad_norm": 3.5402602301601247, + "learning_rate": 1.0682552026794624e-06, + "loss": 0.0431, + "step": 591 + }, + { + "epoch": 13.807580174927114, + "grad_norm": 5.245561994585335, + "learning_rate": 1.0603562651642318e-06, + "loss": 0.0443, + "step": 592 + }, + { + "epoch": 13.830903790087463, + "grad_norm": 4.92187887524649, + "learning_rate": 1.0524787716307107e-06, + "loss": 0.0395, + "step": 593 + }, + { + "epoch": 13.854227405247814, + "grad_norm": 5.371003905498567, + "learning_rate": 1.0446228394168356e-06, + "loss": 0.0436, + "step": 594 + }, + { + "epoch": 13.877551020408163, + "grad_norm": 4.753556175162423, + "learning_rate": 1.0367885855393823e-06, + "loss": 0.048, + "step": 595 + }, + { + "epoch": 13.900874635568513, + "grad_norm": 4.399646978955097, + "learning_rate": 1.02897612669222e-06, + "loss": 0.0404, + "step": 596 + }, + { + "epoch": 13.924198250728862, + "grad_norm": 4.0872273205438185, + "learning_rate": 1.0211855792445712e-06, + "loss": 0.0421, + "step": 597 + }, + { + "epoch": 13.947521865889213, + "grad_norm": 4.404736630969894, + "learning_rate": 1.0134170592392837e-06, + "loss": 0.0445, + "step": 598 + }, + { + "epoch": 13.970845481049563, + "grad_norm": 4.508551419683562, + "learning_rate": 1.005670682391098e-06, + "loss": 0.0425, + "step": 599 + }, + { + "epoch": 13.994169096209912, + "grad_norm": 5.648490346236622, + "learning_rate": 9.97946564084923e-07, + "loss": 0.0513, + "step": 600 + }, + { + "epoch": 14.017492711370263, + "grad_norm": 3.6058075684685664, + "learning_rate": 9.90244819374122e-07, + "loss": 0.0382, + "step": 601 + }, + { + "epoch": 14.040816326530612, + "grad_norm": 3.4872878690349833, + "learning_rate": 9.825655629787952e-07, + "loss": 0.0408, + "step": 602 + }, + { + "epoch": 14.064139941690962, + "grad_norm": 3.9437709705018924, + "learning_rate": 9.749089092840694e-07, + "loss": 0.0358, + "step": 603 + }, + { + "epoch": 14.087463556851311, + "grad_norm": 3.098542479921476, + "learning_rate": 9.672749723383993e-07, + "loss": 0.0314, + "step": 604 + }, + { + "epoch": 14.110787172011662, + "grad_norm": 4.3034719422876035, + "learning_rate": 9.596638658518644e-07, + "loss": 0.0403, + "step": 605 + }, + { + "epoch": 14.13411078717201, + "grad_norm": 3.9472960669595554, + "learning_rate": 9.520757031944786e-07, + "loss": 0.0371, + "step": 606 + }, + { + "epoch": 14.157434402332362, + "grad_norm": 3.634289513039296, + "learning_rate": 9.445105973944962e-07, + "loss": 0.0357, + "step": 607 + }, + { + "epoch": 14.180758017492712, + "grad_norm": 4.147703633888312, + "learning_rate": 9.369686611367354e-07, + "loss": 0.0341, + "step": 608 + }, + { + "epoch": 14.204081632653061, + "grad_norm": 4.857802746929539, + "learning_rate": 9.294500067608941e-07, + "loss": 0.0345, + "step": 609 + }, + { + "epoch": 14.227405247813412, + "grad_norm": 4.887916863266706, + "learning_rate": 9.219547462598796e-07, + "loss": 0.0376, + "step": 610 + }, + { + "epoch": 14.25072886297376, + "grad_norm": 4.601389028391382, + "learning_rate": 9.1448299127814e-07, + "loss": 0.0362, + "step": 611 + }, + { + "epoch": 14.274052478134111, + "grad_norm": 4.6368102761212, + "learning_rate": 9.070348531099982e-07, + "loss": 0.0363, + "step": 612 + }, + { + "epoch": 14.29737609329446, + "grad_norm": 5.32549464834129, + "learning_rate": 8.99610442697999e-07, + "loss": 0.0361, + "step": 613 + }, + { + "epoch": 14.32069970845481, + "grad_norm": 4.278979290973843, + "learning_rate": 8.922098706312548e-07, + "loss": 0.0356, + "step": 614 + }, + { + "epoch": 14.34402332361516, + "grad_norm": 5.638938419185062, + "learning_rate": 8.848332471437948e-07, + "loss": 0.0369, + "step": 615 + }, + { + "epoch": 14.36734693877551, + "grad_norm": 4.43808624263548, + "learning_rate": 8.774806821129286e-07, + "loss": 0.0362, + "step": 616 + }, + { + "epoch": 14.39067055393586, + "grad_norm": 5.872231987967421, + "learning_rate": 8.701522850576069e-07, + "loss": 0.037, + "step": 617 + }, + { + "epoch": 14.41399416909621, + "grad_norm": 5.542707570154857, + "learning_rate": 8.628481651367876e-07, + "loss": 0.0342, + "step": 618 + }, + { + "epoch": 14.43731778425656, + "grad_norm": 3.9740951467872248, + "learning_rate": 8.555684311478158e-07, + "loss": 0.0301, + "step": 619 + }, + { + "epoch": 14.46064139941691, + "grad_norm": 4.520913165862787, + "learning_rate": 8.483131915247969e-07, + "loss": 0.0369, + "step": 620 + }, + { + "epoch": 14.48396501457726, + "grad_norm": 6.336511249646333, + "learning_rate": 8.410825543369866e-07, + "loss": 0.0381, + "step": 621 + }, + { + "epoch": 14.507288629737609, + "grad_norm": 6.351571809168301, + "learning_rate": 8.33876627287179e-07, + "loss": 0.0411, + "step": 622 + }, + { + "epoch": 14.53061224489796, + "grad_norm": 5.359115371865932, + "learning_rate": 8.266955177100997e-07, + "loss": 0.0302, + "step": 623 + }, + { + "epoch": 14.55393586005831, + "grad_norm": 4.855515888749458, + "learning_rate": 8.195393325708123e-07, + "loss": 0.0349, + "step": 624 + }, + { + "epoch": 14.577259475218659, + "grad_norm": 5.699262591396659, + "learning_rate": 8.124081784631219e-07, + "loss": 0.0293, + "step": 625 + }, + { + "epoch": 14.60058309037901, + "grad_norm": 4.567521558688794, + "learning_rate": 8.053021616079873e-07, + "loss": 0.0328, + "step": 626 + }, + { + "epoch": 14.623906705539358, + "grad_norm": 4.609803260692903, + "learning_rate": 7.982213878519405e-07, + "loss": 0.0317, + "step": 627 + }, + { + "epoch": 14.647230320699709, + "grad_norm": 6.458495632306298, + "learning_rate": 7.911659626655066e-07, + "loss": 0.0354, + "step": 628 + }, + { + "epoch": 14.670553935860058, + "grad_norm": 4.150607071561942, + "learning_rate": 7.841359911416377e-07, + "loss": 0.0388, + "step": 629 + }, + { + "epoch": 14.693877551020408, + "grad_norm": 5.1220569999529095, + "learning_rate": 7.771315779941435e-07, + "loss": 0.0347, + "step": 630 + }, + { + "epoch": 14.717201166180757, + "grad_norm": 5.132376038270946, + "learning_rate": 7.701528275561349e-07, + "loss": 0.0365, + "step": 631 + }, + { + "epoch": 14.740524781341108, + "grad_norm": 3.5918722803395284, + "learning_rate": 7.631998437784644e-07, + "loss": 0.035, + "step": 632 + }, + { + "epoch": 14.763848396501459, + "grad_norm": 4.178545815525575, + "learning_rate": 7.56272730228185e-07, + "loss": 0.0327, + "step": 633 + }, + { + "epoch": 14.787172011661808, + "grad_norm": 4.4614095087413235, + "learning_rate": 7.493715900870027e-07, + "loss": 0.0368, + "step": 634 + }, + { + "epoch": 14.810495626822158, + "grad_norm": 4.519648571024788, + "learning_rate": 7.424965261497394e-07, + "loss": 0.0372, + "step": 635 + }, + { + "epoch": 14.833819241982507, + "grad_norm": 5.416749669930699, + "learning_rate": 7.356476408228047e-07, + "loss": 0.0355, + "step": 636 + }, + { + "epoch": 14.857142857142858, + "grad_norm": 5.91713297264535, + "learning_rate": 7.288250361226692e-07, + "loss": 0.0398, + "step": 637 + }, + { + "epoch": 14.880466472303207, + "grad_norm": 4.330248150135053, + "learning_rate": 7.220288136743428e-07, + "loss": 0.0349, + "step": 638 + }, + { + "epoch": 14.903790087463557, + "grad_norm": 5.328988729592788, + "learning_rate": 7.152590747098645e-07, + "loss": 0.0396, + "step": 639 + }, + { + "epoch": 14.927113702623906, + "grad_norm": 5.698975775535172, + "learning_rate": 7.085159200667929e-07, + "loss": 0.0371, + "step": 640 + }, + { + "epoch": 14.950437317784257, + "grad_norm": 4.510406963992824, + "learning_rate": 7.017994501867017e-07, + "loss": 0.032, + "step": 641 + }, + { + "epoch": 14.973760932944607, + "grad_norm": 5.494979127214271, + "learning_rate": 6.95109765113689e-07, + "loss": 0.0347, + "step": 642 + }, + { + "epoch": 14.997084548104956, + "grad_norm": 5.2319532571156975, + "learning_rate": 6.884469644928834e-07, + "loss": 0.0392, + "step": 643 + }, + { + "epoch": 15.020408163265307, + "grad_norm": 3.3408845406351566, + "learning_rate": 6.818111475689585e-07, + "loss": 0.0305, + "step": 644 + }, + { + "epoch": 15.043731778425656, + "grad_norm": 3.5179721711116914, + "learning_rate": 6.752024131846586e-07, + "loss": 0.0263, + "step": 645 + }, + { + "epoch": 15.067055393586006, + "grad_norm": 4.349263595862402, + "learning_rate": 6.68620859779324e-07, + "loss": 0.0315, + "step": 646 + }, + { + "epoch": 15.090379008746355, + "grad_norm": 3.617996290180757, + "learning_rate": 6.620665853874252e-07, + "loss": 0.0286, + "step": 647 + }, + { + "epoch": 15.113702623906706, + "grad_norm": 3.675480407662625, + "learning_rate": 6.555396876371017e-07, + "loss": 0.0277, + "step": 648 + }, + { + "epoch": 15.137026239067055, + "grad_norm": 4.156256424748324, + "learning_rate": 6.490402637487093e-07, + "loss": 0.0293, + "step": 649 + }, + { + "epoch": 15.160349854227405, + "grad_norm": 4.293695334540458, + "learning_rate": 6.425684105333713e-07, + "loss": 0.029, + "step": 650 + }, + { + "epoch": 15.183673469387756, + "grad_norm": 4.138107513358523, + "learning_rate": 6.361242243915364e-07, + "loss": 0.0243, + "step": 651 + }, + { + "epoch": 15.206997084548105, + "grad_norm": 3.686422675568326, + "learning_rate": 6.297078013115435e-07, + "loss": 0.0265, + "step": 652 + }, + { + "epoch": 15.230320699708455, + "grad_norm": 4.378657093156888, + "learning_rate": 6.23319236868189e-07, + "loss": 0.0292, + "step": 653 + }, + { + "epoch": 15.253644314868804, + "grad_norm": 4.339007965833144, + "learning_rate": 6.169586262213081e-07, + "loss": 0.0302, + "step": 654 + }, + { + "epoch": 15.276967930029155, + "grad_norm": 6.16875137282707, + "learning_rate": 6.106260641143547e-07, + "loss": 0.0297, + "step": 655 + }, + { + "epoch": 15.300291545189504, + "grad_norm": 4.435156404818784, + "learning_rate": 6.043216448729869e-07, + "loss": 0.0255, + "step": 656 + }, + { + "epoch": 15.323615160349854, + "grad_norm": 3.8681138903824595, + "learning_rate": 5.980454624036696e-07, + "loss": 0.0286, + "step": 657 + }, + { + "epoch": 15.346938775510203, + "grad_norm": 3.945328695433962, + "learning_rate": 5.917976101922676e-07, + "loss": 0.0309, + "step": 658 + }, + { + "epoch": 15.370262390670554, + "grad_norm": 3.941723934561367, + "learning_rate": 5.855781813026595e-07, + "loss": 0.0242, + "step": 659 + }, + { + "epoch": 15.393586005830905, + "grad_norm": 3.615879555645832, + "learning_rate": 5.79387268375349e-07, + "loss": 0.0322, + "step": 660 + }, + { + "epoch": 15.416909620991254, + "grad_norm": 4.258040507799856, + "learning_rate": 5.73224963626082e-07, + "loss": 0.0253, + "step": 661 + }, + { + "epoch": 15.440233236151604, + "grad_norm": 4.208961490294516, + "learning_rate": 5.670913588444788e-07, + "loss": 0.0344, + "step": 662 + }, + { + "epoch": 15.463556851311953, + "grad_norm": 4.070003446535343, + "learning_rate": 5.609865453926635e-07, + "loss": 0.0286, + "step": 663 + }, + { + "epoch": 15.486880466472304, + "grad_norm": 3.942814222935444, + "learning_rate": 5.549106142039018e-07, + "loss": 0.0305, + "step": 664 + }, + { + "epoch": 15.510204081632653, + "grad_norm": 3.9970858210363684, + "learning_rate": 5.488636557812499e-07, + "loss": 0.0239, + "step": 665 + }, + { + "epoch": 15.533527696793003, + "grad_norm": 4.50085123276231, + "learning_rate": 5.428457601962048e-07, + "loss": 0.0274, + "step": 666 + }, + { + "epoch": 15.556851311953352, + "grad_norm": 3.6513940892295973, + "learning_rate": 5.368570170873618e-07, + "loss": 0.0237, + "step": 667 + }, + { + "epoch": 15.580174927113703, + "grad_norm": 3.110499959666367, + "learning_rate": 5.30897515659082e-07, + "loss": 0.0251, + "step": 668 + }, + { + "epoch": 15.603498542274053, + "grad_norm": 4.673667130206186, + "learning_rate": 5.249673446801584e-07, + "loss": 0.0309, + "step": 669 + }, + { + "epoch": 15.626822157434402, + "grad_norm": 4.138706900217531, + "learning_rate": 5.190665924825005e-07, + "loss": 0.027, + "step": 670 + }, + { + "epoch": 15.650145772594753, + "grad_norm": 3.9353264758772837, + "learning_rate": 5.13195346959813e-07, + "loss": 0.0272, + "step": 671 + }, + { + "epoch": 15.673469387755102, + "grad_norm": 4.734505931693013, + "learning_rate": 5.0735369556629e-07, + "loss": 0.0279, + "step": 672 + }, + { + "epoch": 15.696793002915452, + "grad_norm": 4.469816067386837, + "learning_rate": 5.015417253153093e-07, + "loss": 0.0272, + "step": 673 + }, + { + "epoch": 15.720116618075801, + "grad_norm": 4.5790698017199345, + "learning_rate": 4.957595227781395e-07, + "loss": 0.0325, + "step": 674 + }, + { + "epoch": 15.743440233236152, + "grad_norm": 3.9610773848110963, + "learning_rate": 4.90007174082649e-07, + "loss": 0.0267, + "step": 675 + }, + { + "epoch": 15.7667638483965, + "grad_norm": 4.231649387279768, + "learning_rate": 4.842847649120216e-07, + "loss": 0.0262, + "step": 676 + }, + { + "epoch": 15.790087463556851, + "grad_norm": 4.758712007026724, + "learning_rate": 4.785923805034831e-07, + "loss": 0.0278, + "step": 677 + }, + { + "epoch": 15.813411078717202, + "grad_norm": 3.245691083958514, + "learning_rate": 4.729301056470309e-07, + "loss": 0.0244, + "step": 678 + }, + { + "epoch": 15.83673469387755, + "grad_norm": 3.6139691332125388, + "learning_rate": 4.67298024684168e-07, + "loss": 0.0246, + "step": 679 + }, + { + "epoch": 15.860058309037901, + "grad_norm": 4.715830525039419, + "learning_rate": 4.616962215066512e-07, + "loss": 0.0262, + "step": 680 + }, + { + "epoch": 15.88338192419825, + "grad_norm": 4.003030344359909, + "learning_rate": 4.5612477955523953e-07, + "loss": 0.0293, + "step": 681 + }, + { + "epoch": 15.906705539358601, + "grad_norm": 4.0002567685684305, + "learning_rate": 4.505837818184489e-07, + "loss": 0.0265, + "step": 682 + }, + { + "epoch": 15.93002915451895, + "grad_norm": 4.240186693651765, + "learning_rate": 4.4507331083132074e-07, + "loss": 0.0335, + "step": 683 + }, + { + "epoch": 15.9533527696793, + "grad_norm": 3.8489095902442956, + "learning_rate": 4.3959344867419015e-07, + "loss": 0.0259, + "step": 684 + }, + { + "epoch": 15.97667638483965, + "grad_norm": 5.316228836989733, + "learning_rate": 4.34144276971461e-07, + "loss": 0.0323, + "step": 685 + }, + { + "epoch": 16.0, + "grad_norm": 4.182654694810523, + "learning_rate": 4.2872587689039486e-07, + "loss": 0.0235, + "step": 686 + }, + { + "epoch": 16.02332361516035, + "grad_norm": 4.422215964176046, + "learning_rate": 4.2333832913989845e-07, + "loss": 0.0249, + "step": 687 + }, + { + "epoch": 16.0466472303207, + "grad_norm": 5.1425082111061275, + "learning_rate": 4.179817139693232e-07, + "loss": 0.0241, + "step": 688 + }, + { + "epoch": 16.06997084548105, + "grad_norm": 5.364149630400351, + "learning_rate": 4.126561111672689e-07, + "loss": 0.0253, + "step": 689 + }, + { + "epoch": 16.0932944606414, + "grad_norm": 3.1119300657519378, + "learning_rate": 4.073616000603947e-07, + "loss": 0.021, + "step": 690 + }, + { + "epoch": 16.11661807580175, + "grad_norm": 4.5740999097233175, + "learning_rate": 4.0209825951223996e-07, + "loss": 0.0229, + "step": 691 + }, + { + "epoch": 16.1399416909621, + "grad_norm": 4.6084190282505135, + "learning_rate": 3.9686616792204677e-07, + "loss": 0.0211, + "step": 692 + }, + { + "epoch": 16.163265306122447, + "grad_norm": 5.429734571513587, + "learning_rate": 3.91665403223595e-07, + "loss": 0.0238, + "step": 693 + }, + { + "epoch": 16.186588921282798, + "grad_norm": 3.273517325436302, + "learning_rate": 3.864960428840375e-07, + "loss": 0.0273, + "step": 694 + }, + { + "epoch": 16.20991253644315, + "grad_norm": 4.565062756812091, + "learning_rate": 3.8135816390275027e-07, + "loss": 0.0228, + "step": 695 + }, + { + "epoch": 16.2332361516035, + "grad_norm": 5.08354503420065, + "learning_rate": 3.7625184281018385e-07, + "loss": 0.0253, + "step": 696 + }, + { + "epoch": 16.25655976676385, + "grad_norm": 4.670042973714779, + "learning_rate": 3.711771556667218e-07, + "loss": 0.0255, + "step": 697 + }, + { + "epoch": 16.279883381924197, + "grad_norm": 4.075983998200786, + "learning_rate": 3.6613417806155126e-07, + "loss": 0.0236, + "step": 698 + }, + { + "epoch": 16.303206997084548, + "grad_norm": 4.274402795314428, + "learning_rate": 3.611229851115333e-07, + "loss": 0.0225, + "step": 699 + }, + { + "epoch": 16.3265306122449, + "grad_norm": 4.209429808941616, + "learning_rate": 3.561436514600866e-07, + "loss": 0.0203, + "step": 700 + }, + { + "epoch": 16.34985422740525, + "grad_norm": 4.145068738710229, + "learning_rate": 3.511962512760758e-07, + "loss": 0.0196, + "step": 701 + }, + { + "epoch": 16.373177842565596, + "grad_norm": 3.934032790714013, + "learning_rate": 3.46280858252703e-07, + "loss": 0.0205, + "step": 702 + }, + { + "epoch": 16.396501457725947, + "grad_norm": 3.836792145595831, + "learning_rate": 3.413975456064153e-07, + "loss": 0.0235, + "step": 703 + }, + { + "epoch": 16.419825072886297, + "grad_norm": 4.405125467719681, + "learning_rate": 3.365463860758114e-07, + "loss": 0.0234, + "step": 704 + }, + { + "epoch": 16.443148688046648, + "grad_norm": 3.49860367850944, + "learning_rate": 3.3172745192055656e-07, + "loss": 0.0224, + "step": 705 + }, + { + "epoch": 16.466472303207, + "grad_norm": 4.330738586794788, + "learning_rate": 3.2694081492031017e-07, + "loss": 0.0239, + "step": 706 + }, + { + "epoch": 16.489795918367346, + "grad_norm": 4.006534960248732, + "learning_rate": 3.2218654637365373e-07, + "loss": 0.0197, + "step": 707 + }, + { + "epoch": 16.513119533527696, + "grad_norm": 3.823420159690705, + "learning_rate": 3.1746471709702963e-07, + "loss": 0.0246, + "step": 708 + }, + { + "epoch": 16.536443148688047, + "grad_norm": 4.483000541351714, + "learning_rate": 3.1277539742368636e-07, + "loss": 0.0203, + "step": 709 + }, + { + "epoch": 16.559766763848398, + "grad_norm": 4.425099413773361, + "learning_rate": 3.081186572026318e-07, + "loss": 0.0204, + "step": 710 + }, + { + "epoch": 16.583090379008745, + "grad_norm": 5.020044774011588, + "learning_rate": 3.034945657975899e-07, + "loss": 0.0211, + "step": 711 + }, + { + "epoch": 16.606413994169095, + "grad_norm": 3.481525436271193, + "learning_rate": 2.9890319208597056e-07, + "loss": 0.0209, + "step": 712 + }, + { + "epoch": 16.629737609329446, + "grad_norm": 5.084619307319498, + "learning_rate": 2.943446044578432e-07, + "loss": 0.0204, + "step": 713 + }, + { + "epoch": 16.653061224489797, + "grad_norm": 3.942713963829196, + "learning_rate": 2.898188708149158e-07, + "loss": 0.0231, + "step": 714 + }, + { + "epoch": 16.676384839650147, + "grad_norm": 3.8706390698936537, + "learning_rate": 2.8532605856952613e-07, + "loss": 0.023, + "step": 715 + }, + { + "epoch": 16.699708454810494, + "grad_norm": 4.072339627624961, + "learning_rate": 2.808662346436372e-07, + "loss": 0.0247, + "step": 716 + }, + { + "epoch": 16.723032069970845, + "grad_norm": 3.7024539387765794, + "learning_rate": 2.764394654678379e-07, + "loss": 0.0228, + "step": 717 + }, + { + "epoch": 16.746355685131196, + "grad_norm": 3.2166705775525393, + "learning_rate": 2.720458169803569e-07, + "loss": 0.02, + "step": 718 + }, + { + "epoch": 16.769679300291546, + "grad_norm": 3.80832130435794, + "learning_rate": 2.676853546260791e-07, + "loss": 0.025, + "step": 719 + }, + { + "epoch": 16.793002915451893, + "grad_norm": 2.807860298577788, + "learning_rate": 2.6335814335556933e-07, + "loss": 0.0177, + "step": 720 + }, + { + "epoch": 16.816326530612244, + "grad_norm": 4.087349117189053, + "learning_rate": 2.590642476241076e-07, + "loss": 0.0219, + "step": 721 + }, + { + "epoch": 16.839650145772595, + "grad_norm": 4.537123457993102, + "learning_rate": 2.5480373139072784e-07, + "loss": 0.0225, + "step": 722 + }, + { + "epoch": 16.862973760932945, + "grad_norm": 4.516624049775102, + "learning_rate": 2.505766581172628e-07, + "loss": 0.0219, + "step": 723 + }, + { + "epoch": 16.886297376093296, + "grad_norm": 3.8164973331073773, + "learning_rate": 2.4638309076740364e-07, + "loss": 0.0202, + "step": 724 + }, + { + "epoch": 16.909620991253643, + "grad_norm": 3.655723859963662, + "learning_rate": 2.422230918057586e-07, + "loss": 0.024, + "step": 725 + }, + { + "epoch": 16.932944606413994, + "grad_norm": 4.001000040929099, + "learning_rate": 2.3809672319692224e-07, + "loss": 0.0208, + "step": 726 + }, + { + "epoch": 16.956268221574344, + "grad_norm": 3.858770160925135, + "learning_rate": 2.340040464045548e-07, + "loss": 0.0224, + "step": 727 + }, + { + "epoch": 16.979591836734695, + "grad_norm": 3.482575640490216, + "learning_rate": 2.2994512239046557e-07, + "loss": 0.0198, + "step": 728 + }, + { + "epoch": 17.002915451895042, + "grad_norm": 3.323685241274939, + "learning_rate": 2.2592001161370392e-07, + "loss": 0.0207, + "step": 729 + }, + { + "epoch": 17.026239067055393, + "grad_norm": 3.4118874471688594, + "learning_rate": 2.219287740296605e-07, + "loss": 0.0165, + "step": 730 + }, + { + "epoch": 17.049562682215743, + "grad_norm": 2.917633078186951, + "learning_rate": 2.1797146908917295e-07, + "loss": 0.019, + "step": 731 + }, + { + "epoch": 17.072886297376094, + "grad_norm": 3.0512810565097785, + "learning_rate": 2.1404815573764004e-07, + "loss": 0.0189, + "step": 732 + }, + { + "epoch": 17.096209912536445, + "grad_norm": 2.9573534206720615, + "learning_rate": 2.1015889241414456e-07, + "loss": 0.0175, + "step": 733 + }, + { + "epoch": 17.11953352769679, + "grad_norm": 3.6324930030036007, + "learning_rate": 2.0630373705058408e-07, + "loss": 0.0148, + "step": 734 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 3.430174912422043, + "learning_rate": 2.0248274707080378e-07, + "loss": 0.0184, + "step": 735 + }, + { + "epoch": 17.166180758017493, + "grad_norm": 3.3850899093610507, + "learning_rate": 1.9869597938974611e-07, + "loss": 0.0162, + "step": 736 + }, + { + "epoch": 17.189504373177844, + "grad_norm": 3.042707047575919, + "learning_rate": 1.9494349041260018e-07, + "loss": 0.0204, + "step": 737 + }, + { + "epoch": 17.21282798833819, + "grad_norm": 3.596583037536939, + "learning_rate": 1.9122533603396165e-07, + "loss": 0.0188, + "step": 738 + }, + { + "epoch": 17.23615160349854, + "grad_norm": 2.6161715685027778, + "learning_rate": 1.875415716370013e-07, + "loss": 0.0158, + "step": 739 + }, + { + "epoch": 17.259475218658892, + "grad_norm": 3.1256243272829085, + "learning_rate": 1.8389225209263867e-07, + "loss": 0.0194, + "step": 740 + }, + { + "epoch": 17.282798833819243, + "grad_norm": 3.3380374775697406, + "learning_rate": 1.8027743175872663e-07, + "loss": 0.026, + "step": 741 + }, + { + "epoch": 17.306122448979593, + "grad_norm": 2.8879281325234887, + "learning_rate": 1.7669716447924062e-07, + "loss": 0.0155, + "step": 742 + }, + { + "epoch": 17.32944606413994, + "grad_norm": 3.3041476627170545, + "learning_rate": 1.7315150358347492e-07, + "loss": 0.0199, + "step": 743 + }, + { + "epoch": 17.35276967930029, + "grad_norm": 3.36979570737397, + "learning_rate": 1.696405018852515e-07, + "loss": 0.0186, + "step": 744 + }, + { + "epoch": 17.37609329446064, + "grad_norm": 3.0199209655120867, + "learning_rate": 1.6616421168213215e-07, + "loss": 0.0174, + "step": 745 + }, + { + "epoch": 17.399416909620992, + "grad_norm": 2.929456045544688, + "learning_rate": 1.627226847546376e-07, + "loss": 0.0197, + "step": 746 + }, + { + "epoch": 17.42274052478134, + "grad_norm": 2.806790809090712, + "learning_rate": 1.5931597236547886e-07, + "loss": 0.0173, + "step": 747 + }, + { + "epoch": 17.44606413994169, + "grad_norm": 3.7535778779294233, + "learning_rate": 1.5594412525879288e-07, + "loss": 0.0131, + "step": 748 + }, + { + "epoch": 17.46938775510204, + "grad_norm": 3.469150966804552, + "learning_rate": 1.5260719365938537e-07, + "loss": 0.0151, + "step": 749 + }, + { + "epoch": 17.49271137026239, + "grad_norm": 3.916355337119822, + "learning_rate": 1.4930522727198472e-07, + "loss": 0.0187, + "step": 750 + }, + { + "epoch": 17.516034985422742, + "grad_norm": 3.5350399778012864, + "learning_rate": 1.4603827528050035e-07, + "loss": 0.0172, + "step": 751 + }, + { + "epoch": 17.53935860058309, + "grad_norm": 3.898385148853154, + "learning_rate": 1.428063863472895e-07, + "loss": 0.0276, + "step": 752 + }, + { + "epoch": 17.56268221574344, + "grad_norm": 3.3704370807229593, + "learning_rate": 1.396096086124346e-07, + "loss": 0.0218, + "step": 753 + }, + { + "epoch": 17.58600583090379, + "grad_norm": 3.9975639316201232, + "learning_rate": 1.3644798969302403e-07, + "loss": 0.0155, + "step": 754 + }, + { + "epoch": 17.60932944606414, + "grad_norm": 4.277717636698487, + "learning_rate": 1.333215766824436e-07, + "loss": 0.0206, + "step": 755 + }, + { + "epoch": 17.632653061224488, + "grad_norm": 3.5163862294354113, + "learning_rate": 1.302304161496759e-07, + "loss": 0.0216, + "step": 756 + }, + { + "epoch": 17.65597667638484, + "grad_norm": 3.3342619715001827, + "learning_rate": 1.2717455413860574e-07, + "loss": 0.0222, + "step": 757 + }, + { + "epoch": 17.67930029154519, + "grad_norm": 3.3972292100389683, + "learning_rate": 1.2415403616733322e-07, + "loss": 0.0167, + "step": 758 + }, + { + "epoch": 17.70262390670554, + "grad_norm": 3.6025702996578777, + "learning_rate": 1.211689072274988e-07, + "loss": 0.015, + "step": 759 + }, + { + "epoch": 17.72594752186589, + "grad_norm": 3.587962524158811, + "learning_rate": 1.1821921178361062e-07, + "loss": 0.0199, + "step": 760 + }, + { + "epoch": 17.749271137026238, + "grad_norm": 3.2190849352151916, + "learning_rate": 1.1530499377238213e-07, + "loss": 0.0181, + "step": 761 + }, + { + "epoch": 17.77259475218659, + "grad_norm": 3.2480067963458583, + "learning_rate": 1.1242629660207922e-07, + "loss": 0.0191, + "step": 762 + }, + { + "epoch": 17.79591836734694, + "grad_norm": 4.435559560404031, + "learning_rate": 1.0958316315187289e-07, + "loss": 0.0167, + "step": 763 + }, + { + "epoch": 17.81924198250729, + "grad_norm": 3.296628825321176, + "learning_rate": 1.0677563577119965e-07, + "loss": 0.0144, + "step": 764 + }, + { + "epoch": 17.842565597667637, + "grad_norm": 3.718101380786653, + "learning_rate": 1.0400375627913268e-07, + "loss": 0.0201, + "step": 765 + }, + { + "epoch": 17.865889212827987, + "grad_norm": 3.313696447261319, + "learning_rate": 1.0126756596375687e-07, + "loss": 0.0139, + "step": 766 + }, + { + "epoch": 17.889212827988338, + "grad_norm": 2.9657404050088982, + "learning_rate": 9.856710558155458e-08, + "loss": 0.022, + "step": 767 + }, + { + "epoch": 17.91253644314869, + "grad_norm": 3.339676273898552, + "learning_rate": 9.590241535679973e-08, + "loss": 0.0191, + "step": 768 + }, + { + "epoch": 17.93586005830904, + "grad_norm": 3.3890064788873495, + "learning_rate": 9.32735349809566e-08, + "loss": 0.0173, + "step": 769 + }, + { + "epoch": 17.959183673469386, + "grad_norm": 3.733632544607117, + "learning_rate": 9.068050361209002e-08, + "loss": 0.0165, + "step": 770 + }, + { + "epoch": 17.982507288629737, + "grad_norm": 3.167836742731714, + "learning_rate": 8.8123359874282e-08, + "loss": 0.019, + "step": 771 + }, + { + "epoch": 18.005830903790088, + "grad_norm": 3.1147726930617843, + "learning_rate": 8.560214185705573e-08, + "loss": 0.0201, + "step": 772 + }, + { + "epoch": 18.02915451895044, + "grad_norm": 2.7408804971223417, + "learning_rate": 8.311688711480858e-08, + "loss": 0.0202, + "step": 773 + }, + { + "epoch": 18.05247813411079, + "grad_norm": 2.573615274038812, + "learning_rate": 8.066763266625283e-08, + "loss": 0.0157, + "step": 774 + }, + { + "epoch": 18.075801749271136, + "grad_norm": 3.2240784622186216, + "learning_rate": 7.825441499386437e-08, + "loss": 0.0156, + "step": 775 + }, + { + "epoch": 18.099125364431487, + "grad_norm": 3.053222148681091, + "learning_rate": 7.587727004333856e-08, + "loss": 0.0148, + "step": 776 + }, + { + "epoch": 18.122448979591837, + "grad_norm": 3.3140019304710755, + "learning_rate": 7.353623322305603e-08, + "loss": 0.0164, + "step": 777 + }, + { + "epoch": 18.145772594752188, + "grad_norm": 3.4693247001133947, + "learning_rate": 7.123133940355409e-08, + "loss": 0.0174, + "step": 778 + }, + { + "epoch": 18.169096209912535, + "grad_norm": 2.8945333915836375, + "learning_rate": 6.89626229170079e-08, + "loss": 0.0195, + "step": 779 + }, + { + "epoch": 18.192419825072886, + "grad_norm": 3.957731796869534, + "learning_rate": 6.673011755671921e-08, + "loss": 0.0156, + "step": 780 + }, + { + "epoch": 18.215743440233236, + "grad_norm": 3.5887708586136795, + "learning_rate": 6.45338565766121e-08, + "loss": 0.0185, + "step": 781 + }, + { + "epoch": 18.239067055393587, + "grad_norm": 2.484618744799056, + "learning_rate": 6.237387269073913e-08, + "loss": 0.0172, + "step": 782 + }, + { + "epoch": 18.262390670553934, + "grad_norm": 2.7343133537972837, + "learning_rate": 6.025019807279292e-08, + "loss": 0.0135, + "step": 783 + }, + { + "epoch": 18.285714285714285, + "grad_norm": 3.6152761859835745, + "learning_rate": 5.8162864355626736e-08, + "loss": 0.0171, + "step": 784 + }, + { + "epoch": 18.309037900874635, + "grad_norm": 3.003376649917478, + "learning_rate": 5.6111902630784643e-08, + "loss": 0.0164, + "step": 785 + }, + { + "epoch": 18.332361516034986, + "grad_norm": 3.015679551654015, + "learning_rate": 5.409734344803713e-08, + "loss": 0.0142, + "step": 786 + }, + { + "epoch": 18.355685131195337, + "grad_norm": 2.8544846703309332, + "learning_rate": 5.211921681492649e-08, + "loss": 0.0155, + "step": 787 + }, + { + "epoch": 18.379008746355684, + "grad_norm": 3.284336491294599, + "learning_rate": 5.017755219631964e-08, + "loss": 0.013, + "step": 788 + }, + { + "epoch": 18.402332361516034, + "grad_norm": 2.8660825008489135, + "learning_rate": 4.827237851396965e-08, + "loss": 0.0184, + "step": 789 + }, + { + "epoch": 18.425655976676385, + "grad_norm": 2.5316419886612014, + "learning_rate": 4.640372414608491e-08, + "loss": 0.0113, + "step": 790 + }, + { + "epoch": 18.448979591836736, + "grad_norm": 3.9166951415196976, + "learning_rate": 4.457161692690587e-08, + "loss": 0.0156, + "step": 791 + }, + { + "epoch": 18.472303206997086, + "grad_norm": 2.7355307616015683, + "learning_rate": 4.277608414629097e-08, + "loss": 0.0165, + "step": 792 + }, + { + "epoch": 18.495626822157433, + "grad_norm": 2.8313800773465614, + "learning_rate": 4.101715254930999e-08, + "loss": 0.018, + "step": 793 + }, + { + "epoch": 18.518950437317784, + "grad_norm": 2.7990361393935856, + "learning_rate": 3.929484833584546e-08, + "loss": 0.0145, + "step": 794 + }, + { + "epoch": 18.542274052478135, + "grad_norm": 2.7540460645747307, + "learning_rate": 3.760919716020328e-08, + "loss": 0.015, + "step": 795 + }, + { + "epoch": 18.565597667638485, + "grad_norm": 2.940018568434434, + "learning_rate": 3.596022413072886e-08, + "loss": 0.0201, + "step": 796 + }, + { + "epoch": 18.588921282798832, + "grad_norm": 3.1727857362746867, + "learning_rate": 3.4347953809434894e-08, + "loss": 0.0175, + "step": 797 + }, + { + "epoch": 18.612244897959183, + "grad_norm": 3.5560591178845913, + "learning_rate": 3.277241021163502e-08, + "loss": 0.0166, + "step": 798 + }, + { + "epoch": 18.635568513119534, + "grad_norm": 3.203525425142765, + "learning_rate": 3.1233616805584896e-08, + "loss": 0.0167, + "step": 799 + }, + { + "epoch": 18.658892128279884, + "grad_norm": 2.797864600345931, + "learning_rate": 2.9731596512134753e-08, + "loss": 0.0165, + "step": 800 + }, + { + "epoch": 18.682215743440235, + "grad_norm": 3.571162483657132, + "learning_rate": 2.826637170438601e-08, + "loss": 0.0176, + "step": 801 + }, + { + "epoch": 18.705539358600582, + "grad_norm": 3.4360879945778358, + "learning_rate": 2.683796420735907e-08, + "loss": 0.0158, + "step": 802 + }, + { + "epoch": 18.728862973760933, + "grad_norm": 3.3669790393708277, + "learning_rate": 2.544639529766829e-08, + "loss": 0.0165, + "step": 803 + }, + { + "epoch": 18.752186588921283, + "grad_norm": 3.344597824849103, + "learning_rate": 2.409168570320447e-08, + "loss": 0.0145, + "step": 804 + }, + { + "epoch": 18.775510204081634, + "grad_norm": 3.0129341414147985, + "learning_rate": 2.2773855602826188e-08, + "loss": 0.018, + "step": 805 + }, + { + "epoch": 18.79883381924198, + "grad_norm": 3.8055786091894377, + "learning_rate": 2.149292462606034e-08, + "loss": 0.0189, + "step": 806 + }, + { + "epoch": 18.82215743440233, + "grad_norm": 2.7134696282155604, + "learning_rate": 2.0248911852807918e-08, + "loss": 0.0152, + "step": 807 + }, + { + "epoch": 18.845481049562682, + "grad_norm": 3.205023821475907, + "learning_rate": 1.9041835813061182e-08, + "loss": 0.0174, + "step": 808 + }, + { + "epoch": 18.868804664723033, + "grad_norm": 3.0937133170371, + "learning_rate": 1.7871714486626944e-08, + "loss": 0.0188, + "step": 809 + }, + { + "epoch": 18.892128279883384, + "grad_norm": 3.3376923352088217, + "learning_rate": 1.6738565302859276e-08, + "loss": 0.0151, + "step": 810 + }, + { + "epoch": 18.91545189504373, + "grad_norm": 2.8935337587308427, + "learning_rate": 1.5642405140399998e-08, + "loss": 0.0147, + "step": 811 + }, + { + "epoch": 18.93877551020408, + "grad_norm": 2.940204917098264, + "learning_rate": 1.458325032692609e-08, + "loss": 0.019, + "step": 812 + }, + { + "epoch": 18.962099125364432, + "grad_norm": 2.970374375711716, + "learning_rate": 1.3561116638907968e-08, + "loss": 0.0153, + "step": 813 + }, + { + "epoch": 18.985422740524783, + "grad_norm": 3.290703370989595, + "learning_rate": 1.2576019301373533e-08, + "loss": 0.0182, + "step": 814 + }, + { + "epoch": 19.00874635568513, + "grad_norm": 3.8527255826600477, + "learning_rate": 1.1627972987681702e-08, + "loss": 0.0153, + "step": 815 + }, + { + "epoch": 19.03206997084548, + "grad_norm": 2.452166133013113, + "learning_rate": 1.071699181930369e-08, + "loss": 0.0116, + "step": 816 + }, + { + "epoch": 19.05539358600583, + "grad_norm": 3.2554283291545842, + "learning_rate": 9.843089365613178e-09, + "loss": 0.0182, + "step": 817 + }, + { + "epoch": 19.07871720116618, + "grad_norm": 3.575695829098365, + "learning_rate": 9.006278643683697e-09, + "loss": 0.0195, + "step": 818 + }, + { + "epoch": 19.102040816326532, + "grad_norm": 2.3868383199705696, + "learning_rate": 8.206572118094614e-09, + "loss": 0.0116, + "step": 819 + }, + { + "epoch": 19.12536443148688, + "grad_norm": 3.0586927453946693, + "learning_rate": 7.4439817007460056e-09, + "loss": 0.0161, + "step": 820 + }, + { + "epoch": 19.14868804664723, + "grad_norm": 2.6031425191557647, + "learning_rate": 6.7185187506804695e-09, + "loss": 0.0139, + "step": 821 + }, + { + "epoch": 19.17201166180758, + "grad_norm": 2.897274472903437, + "learning_rate": 6.030194073914919e-09, + "loss": 0.0166, + "step": 822 + }, + { + "epoch": 19.19533527696793, + "grad_norm": 2.6842476438213736, + "learning_rate": 5.379017923278773e-09, + "loss": 0.0152, + "step": 823 + }, + { + "epoch": 19.21865889212828, + "grad_norm": 3.2130258270849827, + "learning_rate": 4.764999998262132e-09, + "loss": 0.0194, + "step": 824 + }, + { + "epoch": 19.24198250728863, + "grad_norm": 2.6256504161306493, + "learning_rate": 4.1881494448695046e-09, + "loss": 0.0198, + "step": 825 + }, + { + "epoch": 19.26530612244898, + "grad_norm": 3.0907940855917286, + "learning_rate": 3.6484748554857508e-09, + "loss": 0.0157, + "step": 826 + }, + { + "epoch": 19.28862973760933, + "grad_norm": 3.0455037479230396, + "learning_rate": 3.145984268747293e-09, + "loss": 0.0171, + "step": 827 + }, + { + "epoch": 19.31195335276968, + "grad_norm": 2.913053854478876, + "learning_rate": 2.680685169421382e-09, + "loss": 0.0153, + "step": 828 + }, + { + "epoch": 19.335276967930028, + "grad_norm": 2.841583600512857, + "learning_rate": 2.252584488296461e-09, + "loss": 0.0173, + "step": 829 + }, + { + "epoch": 19.35860058309038, + "grad_norm": 3.5529614628152104, + "learning_rate": 1.8616886020766944e-09, + "loss": 0.0152, + "step": 830 + }, + { + "epoch": 19.38192419825073, + "grad_norm": 2.7916265076742643, + "learning_rate": 1.5080033332892652e-09, + "loss": 0.0139, + "step": 831 + }, + { + "epoch": 19.40524781341108, + "grad_norm": 2.8798786731171506, + "learning_rate": 1.1915339501963885e-09, + "loss": 0.0159, + "step": 832 + }, + { + "epoch": 19.428571428571427, + "grad_norm": 2.3887949402642494, + "learning_rate": 9.122851667164867e-10, + "loss": 0.0145, + "step": 833 + }, + { + "epoch": 19.451895043731778, + "grad_norm": 3.258576154970029, + "learning_rate": 6.702611423550775e-10, + "loss": 0.0132, + "step": 834 + }, + { + "epoch": 19.47521865889213, + "grad_norm": 3.171290301471705, + "learning_rate": 4.654654821420468e-10, + "loss": 0.0151, + "step": 835 + }, + { + "epoch": 19.49854227405248, + "grad_norm": 2.941285636222306, + "learning_rate": 2.9790123657835733e-10, + "loss": 0.0196, + "step": 836 + }, + { + "epoch": 19.52186588921283, + "grad_norm": 2.9155196159591314, + "learning_rate": 1.6757090159025268e-10, + "loss": 0.0169, + "step": 837 + }, + { + "epoch": 19.545189504373177, + "grad_norm": 3.4301587869352153, + "learning_rate": 7.447641849206433e-11, + "loss": 0.0165, + "step": 838 + }, + { + "epoch": 19.568513119533527, + "grad_norm": 3.040206732998616, + "learning_rate": 1.8619173957623583e-11, + "loss": 0.0166, + "step": 839 + }, + { + "epoch": 19.591836734693878, + "grad_norm": 2.972491827489143, + "learning_rate": 0.0, + "loss": 0.016, + "step": 840 + } + ], + "logging_steps": 1.0, + "max_steps": 840, + "num_input_tokens_seen": 0, + "num_train_epochs": 20, + "save_steps": 5000.0, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 623803706515456.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}