| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 465, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0064516129032258064, | |
| "grad_norm": 0.15169227525554954, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.1803, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012903225806451613, | |
| "grad_norm": 0.1108587617034759, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.1441, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01935483870967742, | |
| "grad_norm": 0.08345105115839851, | |
| "learning_rate": 1.2765957446808511e-05, | |
| "loss": 0.1118, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025806451612903226, | |
| "grad_norm": 0.1426234633379846, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "loss": 0.1313, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03225806451612903, | |
| "grad_norm": 0.14217530777409554, | |
| "learning_rate": 2.1276595744680852e-05, | |
| "loss": 0.1195, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03870967741935484, | |
| "grad_norm": 0.1459141077043439, | |
| "learning_rate": 2.5531914893617022e-05, | |
| "loss": 0.1453, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04516129032258064, | |
| "grad_norm": 0.10195797727709278, | |
| "learning_rate": 2.9787234042553192e-05, | |
| "loss": 0.1119, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05161290322580645, | |
| "grad_norm": 0.13553076939195616, | |
| "learning_rate": 3.4042553191489365e-05, | |
| "loss": 0.1791, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05806451612903226, | |
| "grad_norm": 0.15732711391586654, | |
| "learning_rate": 3.829787234042553e-05, | |
| "loss": 0.1282, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 0.1192108991071351, | |
| "learning_rate": 4.2553191489361704e-05, | |
| "loss": 0.1288, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07096774193548387, | |
| "grad_norm": 0.1534053229135478, | |
| "learning_rate": 4.680851063829788e-05, | |
| "loss": 0.15, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07741935483870968, | |
| "grad_norm": 0.11808236646609899, | |
| "learning_rate": 5.1063829787234044e-05, | |
| "loss": 0.1098, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08387096774193549, | |
| "grad_norm": 0.1127350100020763, | |
| "learning_rate": 5.531914893617022e-05, | |
| "loss": 0.1214, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09032258064516129, | |
| "grad_norm": 0.13802666747672132, | |
| "learning_rate": 5.9574468085106384e-05, | |
| "loss": 0.1555, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0967741935483871, | |
| "grad_norm": 0.19187355993638458, | |
| "learning_rate": 6.382978723404256e-05, | |
| "loss": 0.1474, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1032258064516129, | |
| "grad_norm": 0.16430813087649648, | |
| "learning_rate": 6.808510638297873e-05, | |
| "loss": 0.1461, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.10967741935483871, | |
| "grad_norm": 0.13789502094335715, | |
| "learning_rate": 7.23404255319149e-05, | |
| "loss": 0.1226, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11612903225806452, | |
| "grad_norm": 0.173980091579583, | |
| "learning_rate": 7.659574468085106e-05, | |
| "loss": 0.1558, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12258064516129032, | |
| "grad_norm": 0.1418687450951512, | |
| "learning_rate": 8.085106382978723e-05, | |
| "loss": 0.1379, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 0.1765388069700401, | |
| "learning_rate": 8.510638297872341e-05, | |
| "loss": 0.149, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13548387096774195, | |
| "grad_norm": 0.17659794510341198, | |
| "learning_rate": 8.936170212765958e-05, | |
| "loss": 0.1101, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14193548387096774, | |
| "grad_norm": 0.22633201272355616, | |
| "learning_rate": 9.361702127659576e-05, | |
| "loss": 0.1656, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.14838709677419354, | |
| "grad_norm": 0.16004849032165383, | |
| "learning_rate": 9.787234042553192e-05, | |
| "loss": 0.1616, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15483870967741936, | |
| "grad_norm": 0.16981049690586422, | |
| "learning_rate": 0.00010212765957446809, | |
| "loss": 0.1321, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16129032258064516, | |
| "grad_norm": 0.13581353274905067, | |
| "learning_rate": 0.00010638297872340425, | |
| "loss": 0.1122, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16774193548387098, | |
| "grad_norm": 0.2089185781750498, | |
| "learning_rate": 0.00011063829787234043, | |
| "loss": 0.1527, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17419354838709677, | |
| "grad_norm": 0.16125655225492164, | |
| "learning_rate": 0.00011489361702127661, | |
| "loss": 0.1263, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18064516129032257, | |
| "grad_norm": 0.13753328983503424, | |
| "learning_rate": 0.00011914893617021277, | |
| "loss": 0.1224, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1870967741935484, | |
| "grad_norm": 0.19089590529877826, | |
| "learning_rate": 0.00012340425531914893, | |
| "loss": 0.168, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 0.15419681175807715, | |
| "learning_rate": 0.00012765957446808513, | |
| "loss": 0.138, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.18178665024271073, | |
| "learning_rate": 0.00013191489361702127, | |
| "loss": 0.1955, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2064516129032258, | |
| "grad_norm": 0.11760523051627117, | |
| "learning_rate": 0.00013617021276595746, | |
| "loss": 0.1367, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2129032258064516, | |
| "grad_norm": 0.10853469531105876, | |
| "learning_rate": 0.00014042553191489363, | |
| "loss": 0.1208, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.21935483870967742, | |
| "grad_norm": 0.12890024443339382, | |
| "learning_rate": 0.0001446808510638298, | |
| "loss": 0.142, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.22580645161290322, | |
| "grad_norm": 0.12692879603118554, | |
| "learning_rate": 0.00014893617021276596, | |
| "loss": 0.1268, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.23225806451612904, | |
| "grad_norm": 0.1529612097417899, | |
| "learning_rate": 0.00015319148936170213, | |
| "loss": 0.1194, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.23870967741935484, | |
| "grad_norm": 0.129432114467059, | |
| "learning_rate": 0.00015744680851063832, | |
| "loss": 0.128, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.24516129032258063, | |
| "grad_norm": 0.11343138573627701, | |
| "learning_rate": 0.00016170212765957446, | |
| "loss": 0.102, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.25161290322580643, | |
| "grad_norm": 0.16577469112721435, | |
| "learning_rate": 0.00016595744680851065, | |
| "loss": 0.1691, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 0.13614466586953358, | |
| "learning_rate": 0.00017021276595744682, | |
| "loss": 0.1348, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2645161290322581, | |
| "grad_norm": 0.1524812917659128, | |
| "learning_rate": 0.00017446808510638298, | |
| "loss": 0.1281, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2709677419354839, | |
| "grad_norm": 0.14519269708506746, | |
| "learning_rate": 0.00017872340425531915, | |
| "loss": 0.1349, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.27741935483870966, | |
| "grad_norm": 0.18300481897670345, | |
| "learning_rate": 0.00018297872340425532, | |
| "loss": 0.1694, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2838709677419355, | |
| "grad_norm": 0.11929331561330575, | |
| "learning_rate": 0.0001872340425531915, | |
| "loss": 0.1331, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2903225806451613, | |
| "grad_norm": 0.12604932986068976, | |
| "learning_rate": 0.00019148936170212768, | |
| "loss": 0.0949, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2967741935483871, | |
| "grad_norm": 0.14226790630565908, | |
| "learning_rate": 0.00019574468085106384, | |
| "loss": 0.1477, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3032258064516129, | |
| "grad_norm": 0.1465980952326029, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1408, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3096774193548387, | |
| "grad_norm": 0.14535685898764863, | |
| "learning_rate": 0.0001999971756719333, | |
| "loss": 0.1474, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3161290322580645, | |
| "grad_norm": 0.1559112597851861, | |
| "learning_rate": 0.00019998870284726968, | |
| "loss": 0.1568, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 0.14392375991483491, | |
| "learning_rate": 0.00019997458200460993, | |
| "loss": 0.1526, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.32903225806451614, | |
| "grad_norm": 0.18920444771524633, | |
| "learning_rate": 0.00019995481394159188, | |
| "loss": 0.1384, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.33548387096774196, | |
| "grad_norm": 0.17943388616343298, | |
| "learning_rate": 0.0001999293997748454, | |
| "loss": 0.1555, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3419354838709677, | |
| "grad_norm": 0.16752646291327727, | |
| "learning_rate": 0.00019989834093992945, | |
| "loss": 0.1634, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.34838709677419355, | |
| "grad_norm": 0.1484740777328073, | |
| "learning_rate": 0.00019986163919125075, | |
| "loss": 0.124, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3548387096774194, | |
| "grad_norm": 0.15268630472434244, | |
| "learning_rate": 0.00019981929660196492, | |
| "loss": 0.1228, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.36129032258064514, | |
| "grad_norm": 0.19583236181584973, | |
| "learning_rate": 0.0001997713155638592, | |
| "loss": 0.1227, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.36774193548387096, | |
| "grad_norm": 0.1469999186659826, | |
| "learning_rate": 0.00019971769878721743, | |
| "loss": 0.1188, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3741935483870968, | |
| "grad_norm": 0.10845193424043034, | |
| "learning_rate": 0.000199658449300667, | |
| "loss": 0.1177, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.38064516129032255, | |
| "grad_norm": 0.1845188187089657, | |
| "learning_rate": 0.00019959357045100764, | |
| "loss": 0.1726, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 0.14769124665651473, | |
| "learning_rate": 0.00019952306590302247, | |
| "loss": 0.1634, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3935483870967742, | |
| "grad_norm": 0.16408468392163889, | |
| "learning_rate": 0.00019944693963927092, | |
| "loss": 0.1584, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.15156206723802879, | |
| "learning_rate": 0.00019936519595986394, | |
| "loss": 0.1454, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4064516129032258, | |
| "grad_norm": 0.15835599927161748, | |
| "learning_rate": 0.00019927783948222084, | |
| "loss": 0.1621, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4129032258064516, | |
| "grad_norm": 0.1333411665662589, | |
| "learning_rate": 0.00019918487514080865, | |
| "loss": 0.1293, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.41935483870967744, | |
| "grad_norm": 0.1589874051481937, | |
| "learning_rate": 0.00019908630818686338, | |
| "loss": 0.1391, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4258064516129032, | |
| "grad_norm": 0.12399538770065353, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 0.1208, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.432258064516129, | |
| "grad_norm": 0.16949673743147955, | |
| "learning_rate": 0.00019887238902836448, | |
| "loss": 0.1139, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.43870967741935485, | |
| "grad_norm": 0.1875430265393267, | |
| "learning_rate": 0.00019875704890736853, | |
| "loss": 0.1856, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.44516129032258067, | |
| "grad_norm": 0.14927152386929934, | |
| "learning_rate": 0.00019863613034027224, | |
| "loss": 0.1516, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 0.17501653378304205, | |
| "learning_rate": 0.0001985096401573497, | |
| "loss": 0.161, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.45806451612903226, | |
| "grad_norm": 0.17582762670350804, | |
| "learning_rate": 0.00019837758550359636, | |
| "loss": 0.1564, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4645161290322581, | |
| "grad_norm": 0.15692049884405931, | |
| "learning_rate": 0.0001982399738383255, | |
| "loss": 0.1282, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.47096774193548385, | |
| "grad_norm": 0.14436392088728167, | |
| "learning_rate": 0.00019809681293474693, | |
| "loss": 0.1299, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4774193548387097, | |
| "grad_norm": 0.19048463762976417, | |
| "learning_rate": 0.0001979481108795278, | |
| "loss": 0.1983, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4838709677419355, | |
| "grad_norm": 0.13475205452089994, | |
| "learning_rate": 0.00019779387607233586, | |
| "loss": 0.156, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.49032258064516127, | |
| "grad_norm": 0.14145398172929924, | |
| "learning_rate": 0.00019763411722536502, | |
| "loss": 0.1355, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4967741935483871, | |
| "grad_norm": 0.14400838677113517, | |
| "learning_rate": 0.00019746884336284317, | |
| "loss": 0.1371, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5032258064516129, | |
| "grad_norm": 0.13421206031331137, | |
| "learning_rate": 0.00019729806382052248, | |
| "loss": 0.1156, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5096774193548387, | |
| "grad_norm": 0.11613915473665105, | |
| "learning_rate": 0.00019712178824515212, | |
| "loss": 0.1293, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 0.12472786830602234, | |
| "learning_rate": 0.00019694002659393305, | |
| "loss": 0.1189, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5225806451612903, | |
| "grad_norm": 0.11050712097688373, | |
| "learning_rate": 0.00019675278913395606, | |
| "loss": 0.12, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5290322580645161, | |
| "grad_norm": 0.1303276052231771, | |
| "learning_rate": 0.0001965600864416213, | |
| "loss": 0.1428, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.535483870967742, | |
| "grad_norm": 0.13774570876008593, | |
| "learning_rate": 0.00019636192940204134, | |
| "loss": 0.1438, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5419354838709678, | |
| "grad_norm": 0.14810394982940484, | |
| "learning_rate": 0.00019615832920842586, | |
| "loss": 0.1404, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5483870967741935, | |
| "grad_norm": 0.144275852100491, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 0.1357, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5548387096774193, | |
| "grad_norm": 0.16879600090790034, | |
| "learning_rate": 0.0001957348456686032, | |
| "loss": 0.1578, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5612903225806452, | |
| "grad_norm": 0.1588074767274021, | |
| "learning_rate": 0.00019551498624352496, | |
| "loss": 0.1228, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.567741935483871, | |
| "grad_norm": 0.15067916918276408, | |
| "learning_rate": 0.00019528973150531787, | |
| "loss": 0.1323, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5741935483870968, | |
| "grad_norm": 0.16266341913656662, | |
| "learning_rate": 0.00019505909417784754, | |
| "loss": 0.13, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 0.121529537729675, | |
| "learning_rate": 0.00019482308728902356, | |
| "loss": 0.1067, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5870967741935483, | |
| "grad_norm": 0.1740468182649888, | |
| "learning_rate": 0.00019458172417006347, | |
| "loss": 0.1513, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5935483870967742, | |
| "grad_norm": 0.14062538318374462, | |
| "learning_rate": 0.00019433501845473995, | |
| "loss": 0.1438, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.17690034130801896, | |
| "learning_rate": 0.00019408298407861042, | |
| "loss": 0.1356, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6064516129032258, | |
| "grad_norm": 0.18789482750619546, | |
| "learning_rate": 0.00019382563527823026, | |
| "loss": 0.1758, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6129032258064516, | |
| "grad_norm": 0.16993067663839118, | |
| "learning_rate": 0.00019356298659034817, | |
| "loss": 0.1599, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6193548387096774, | |
| "grad_norm": 0.16495058136550117, | |
| "learning_rate": 0.00019329505285108542, | |
| "loss": 0.1283, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6258064516129033, | |
| "grad_norm": 0.15497395377626808, | |
| "learning_rate": 0.00019302184919509755, | |
| "loss": 0.1493, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.632258064516129, | |
| "grad_norm": 0.13660921526912856, | |
| "learning_rate": 0.00019274339105471971, | |
| "loss": 0.1307, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6387096774193548, | |
| "grad_norm": 0.18246231884688152, | |
| "learning_rate": 0.00019245969415909465, | |
| "loss": 0.1598, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 0.11344749529118914, | |
| "learning_rate": 0.00019217077453328449, | |
| "loss": 0.1304, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6516129032258065, | |
| "grad_norm": 0.11682283731326468, | |
| "learning_rate": 0.0001918766484973654, | |
| "loss": 0.0977, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6580645161290323, | |
| "grad_norm": 0.14494532290577813, | |
| "learning_rate": 0.00019157733266550575, | |
| "loss": 0.1338, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6645161290322581, | |
| "grad_norm": 0.12095275202026515, | |
| "learning_rate": 0.0001912728439450276, | |
| "loss": 0.1513, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6709677419354839, | |
| "grad_norm": 0.13423702981009097, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 0.1335, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6774193548387096, | |
| "grad_norm": 0.1670132860946028, | |
| "learning_rate": 0.0001906484169275263, | |
| "loss": 0.1607, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6838709677419355, | |
| "grad_norm": 0.14053854153152684, | |
| "learning_rate": 0.00019032851390223812, | |
| "loss": 0.1365, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6903225806451613, | |
| "grad_norm": 0.1399807021991922, | |
| "learning_rate": 0.00019000350852980909, | |
| "loss": 0.1589, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6967741935483871, | |
| "grad_norm": 0.15473299551506894, | |
| "learning_rate": 0.00018967341916867518, | |
| "loss": 0.166, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7032258064516129, | |
| "grad_norm": 0.1536872117526864, | |
| "learning_rate": 0.00018933826446444933, | |
| "loss": 0.1657, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 0.1282004975556196, | |
| "learning_rate": 0.0001889980633488683, | |
| "loss": 0.1212, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7161290322580646, | |
| "grad_norm": 0.16458328388975405, | |
| "learning_rate": 0.00018865283503872324, | |
| "loss": 0.1655, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7225806451612903, | |
| "grad_norm": 0.1505113828615181, | |
| "learning_rate": 0.00018830259903477426, | |
| "loss": 0.1571, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7290322580645161, | |
| "grad_norm": 0.16575595088070239, | |
| "learning_rate": 0.0001879473751206489, | |
| "loss": 0.1504, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7354838709677419, | |
| "grad_norm": 0.1484230902451611, | |
| "learning_rate": 0.0001875871833617246, | |
| "loss": 0.1498, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7419354838709677, | |
| "grad_norm": 0.2059907492830938, | |
| "learning_rate": 0.0001872220441039952, | |
| "loss": 0.134, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7483870967741936, | |
| "grad_norm": 0.1491671097000444, | |
| "learning_rate": 0.0001868519779729218, | |
| "loss": 0.1374, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7548387096774194, | |
| "grad_norm": 0.1727675086328308, | |
| "learning_rate": 0.0001864770058722676, | |
| "loss": 0.1624, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7612903225806451, | |
| "grad_norm": 0.16357318423846662, | |
| "learning_rate": 0.00018609714898291718, | |
| "loss": 0.1528, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7677419354838709, | |
| "grad_norm": 0.1584052674932312, | |
| "learning_rate": 0.00018571242876167996, | |
| "loss": 0.1321, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 0.21714396600094343, | |
| "learning_rate": 0.0001853228669400784, | |
| "loss": 0.1748, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7806451612903226, | |
| "grad_norm": 0.13743651841776636, | |
| "learning_rate": 0.00018492848552312014, | |
| "loss": 0.1493, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7870967741935484, | |
| "grad_norm": 0.1541126978032927, | |
| "learning_rate": 0.00018452930678805536, | |
| "loss": 0.1331, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7935483870967742, | |
| "grad_norm": 0.1571822882709535, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 0.1427, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.17825934340851243, | |
| "learning_rate": 0.00018371664782625287, | |
| "loss": 0.1871, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8064516129032258, | |
| "grad_norm": 0.1745800846806893, | |
| "learning_rate": 0.00018330321350382544, | |
| "loss": 0.1672, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8129032258064516, | |
| "grad_norm": 0.13970218689990957, | |
| "learning_rate": 0.00018288507366931905, | |
| "loss": 0.1715, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8193548387096774, | |
| "grad_norm": 0.15100292169016535, | |
| "learning_rate": 0.00018246225194201517, | |
| "loss": 0.1411, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8258064516129032, | |
| "grad_norm": 0.1538126122981586, | |
| "learning_rate": 0.00018203477220565912, | |
| "loss": 0.1516, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.832258064516129, | |
| "grad_norm": 0.15630735898296536, | |
| "learning_rate": 0.00018160265860711134, | |
| "loss": 0.1636, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 0.1427718560771215, | |
| "learning_rate": 0.00018116593555498307, | |
| "loss": 0.1297, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8451612903225807, | |
| "grad_norm": 0.11911885929622754, | |
| "learning_rate": 0.0001807246277182578, | |
| "loss": 0.1115, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8516129032258064, | |
| "grad_norm": 0.13702745725576418, | |
| "learning_rate": 0.0001802787600248977, | |
| "loss": 0.157, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8580645161290322, | |
| "grad_norm": 0.17220644805792673, | |
| "learning_rate": 0.0001798283576604356, | |
| "loss": 0.1561, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.864516129032258, | |
| "grad_norm": 0.196867226472361, | |
| "learning_rate": 0.0001793734460665523, | |
| "loss": 0.1657, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8709677419354839, | |
| "grad_norm": 0.1759359784948508, | |
| "learning_rate": 0.00017891405093963938, | |
| "loss": 0.1909, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8774193548387097, | |
| "grad_norm": 0.16063163300244113, | |
| "learning_rate": 0.0001784501982293479, | |
| "loss": 0.1552, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8838709677419355, | |
| "grad_norm": 0.17034183209183734, | |
| "learning_rate": 0.00017798191413712243, | |
| "loss": 0.1502, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8903225806451613, | |
| "grad_norm": 0.14694529779128243, | |
| "learning_rate": 0.0001775092251147211, | |
| "loss": 0.1277, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.896774193548387, | |
| "grad_norm": 0.13174927545002138, | |
| "learning_rate": 0.0001770321578627213, | |
| "loss": 0.1277, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 0.15784079444219237, | |
| "learning_rate": 0.00017655073932901168, | |
| "loss": 0.1534, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9096774193548387, | |
| "grad_norm": 0.19691370909219638, | |
| "learning_rate": 0.0001760649967072697, | |
| "loss": 0.1688, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9161290322580645, | |
| "grad_norm": 0.15697904965484202, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 0.1523, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9225806451612903, | |
| "grad_norm": 0.1759980765103477, | |
| "learning_rate": 0.00017508064919411344, | |
| "loss": 0.15, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9290322580645162, | |
| "grad_norm": 0.19254834048997346, | |
| "learning_rate": 0.00017458209990510527, | |
| "loss": 0.1474, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9354838709677419, | |
| "grad_norm": 0.19220369870461818, | |
| "learning_rate": 0.00017407933772973637, | |
| "loss": 0.1678, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9419354838709677, | |
| "grad_norm": 0.14789056250556576, | |
| "learning_rate": 0.00017357239106731317, | |
| "loss": 0.1634, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9483870967741935, | |
| "grad_norm": 0.15823815803270533, | |
| "learning_rate": 0.00017306128855350942, | |
| "loss": 0.1744, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9548387096774194, | |
| "grad_norm": 0.159128130793647, | |
| "learning_rate": 0.0001725460590587486, | |
| "loss": 0.1732, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9612903225806452, | |
| "grad_norm": 0.1420932941022579, | |
| "learning_rate": 0.00017202673168657318, | |
| "loss": 0.1193, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 0.1581041276537875, | |
| "learning_rate": 0.0001715033357720006, | |
| "loss": 0.157, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9741935483870968, | |
| "grad_norm": 0.13409040520330398, | |
| "learning_rate": 0.00017097590087986633, | |
| "loss": 0.1187, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9806451612903225, | |
| "grad_norm": 0.15236261967510367, | |
| "learning_rate": 0.00017044445680315372, | |
| "loss": 0.1541, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9870967741935484, | |
| "grad_norm": 0.15028826750982388, | |
| "learning_rate": 0.00016990903356131124, | |
| "loss": 0.1462, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9935483870967742, | |
| "grad_norm": 0.14170487741522195, | |
| "learning_rate": 0.00016936966139855663, | |
| "loss": 0.1275, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.14288119090775778, | |
| "learning_rate": 0.00016882637078216868, | |
| "loss": 0.1316, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.14944089949131012, | |
| "eval_runtime": 27.6083, | |
| "eval_samples_per_second": 4.745, | |
| "eval_steps_per_second": 0.616, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0064516129032257, | |
| "grad_norm": 0.08453811084356862, | |
| "learning_rate": 0.0001682791924007661, | |
| "loss": 0.0732, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0129032258064516, | |
| "grad_norm": 0.1163343033490188, | |
| "learning_rate": 0.00016772815716257412, | |
| "loss": 0.0955, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0193548387096774, | |
| "grad_norm": 0.12774969124655083, | |
| "learning_rate": 0.0001671732961936785, | |
| "loss": 0.1154, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0258064516129033, | |
| "grad_norm": 0.10149504758995384, | |
| "learning_rate": 0.00016661464083626734, | |
| "loss": 0.0834, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.032258064516129, | |
| "grad_norm": 0.1558767636437416, | |
| "learning_rate": 0.00016605222264686086, | |
| "loss": 0.0978, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.038709677419355, | |
| "grad_norm": 0.14800857594022712, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 0.0783, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0451612903225806, | |
| "grad_norm": 0.10427437820954576, | |
| "learning_rate": 0.00016491622505909482, | |
| "loss": 0.0714, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0516129032258064, | |
| "grad_norm": 0.16509881361577539, | |
| "learning_rate": 0.00016434270982933273, | |
| "loss": 0.0971, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0580645161290323, | |
| "grad_norm": 0.15528612822142446, | |
| "learning_rate": 0.0001637655601011454, | |
| "loss": 0.0806, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.064516129032258, | |
| "grad_norm": 0.1578175598943513, | |
| "learning_rate": 0.00016318480847573642, | |
| "loss": 0.0962, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.070967741935484, | |
| "grad_norm": 0.2591491854389689, | |
| "learning_rate": 0.00016260048775776804, | |
| "loss": 0.1107, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0774193548387097, | |
| "grad_norm": 0.1557337965418426, | |
| "learning_rate": 0.00016201263095350833, | |
| "loss": 0.0707, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.0838709677419356, | |
| "grad_norm": 0.24212581528206514, | |
| "learning_rate": 0.0001614212712689668, | |
| "loss": 0.0964, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0903225806451613, | |
| "grad_norm": 0.12600203218602474, | |
| "learning_rate": 0.00016082644210801844, | |
| "loss": 0.0649, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.096774193548387, | |
| "grad_norm": 0.14377456483205683, | |
| "learning_rate": 0.00016022817707051724, | |
| "loss": 0.0778, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.103225806451613, | |
| "grad_norm": 0.1828183442232092, | |
| "learning_rate": 0.00015962650995039783, | |
| "loss": 0.1068, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1096774193548387, | |
| "grad_norm": 0.1695078607562437, | |
| "learning_rate": 0.00015902147473376694, | |
| "loss": 0.0938, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1161290322580646, | |
| "grad_norm": 0.10525262692642562, | |
| "learning_rate": 0.00015841310559698343, | |
| "loss": 0.0645, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1225806451612903, | |
| "grad_norm": 0.15379971536703851, | |
| "learning_rate": 0.0001578014369047279, | |
| "loss": 0.0752, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.129032258064516, | |
| "grad_norm": 0.1352511385337785, | |
| "learning_rate": 0.00015718650320806142, | |
| "loss": 0.0803, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.135483870967742, | |
| "grad_norm": 0.1708537982318491, | |
| "learning_rate": 0.00015656833924247398, | |
| "loss": 0.0908, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1419354838709677, | |
| "grad_norm": 0.16847128771716718, | |
| "learning_rate": 0.00015594697992592232, | |
| "loss": 0.0697, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1483870967741936, | |
| "grad_norm": 0.14044376651199897, | |
| "learning_rate": 0.00015532246035685756, | |
| "loss": 0.0707, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1548387096774193, | |
| "grad_norm": 0.13752732543534765, | |
| "learning_rate": 0.00015469481581224272, | |
| "loss": 0.0769, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1612903225806452, | |
| "grad_norm": 0.20084535405957962, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 0.1114, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.167741935483871, | |
| "grad_norm": 0.16882912932677738, | |
| "learning_rate": 0.0001534302937848073, | |
| "loss": 0.0977, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1741935483870969, | |
| "grad_norm": 0.16088758279122345, | |
| "learning_rate": 0.00015279348773048786, | |
| "loss": 0.0862, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1806451612903226, | |
| "grad_norm": 0.1610420842518173, | |
| "learning_rate": 0.00015215369955358566, | |
| "loss": 0.0909, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1870967741935483, | |
| "grad_norm": 0.1567858759713509, | |
| "learning_rate": 0.0001515109653935348, | |
| "loss": 0.0988, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.1935483870967742, | |
| "grad_norm": 0.12400211503217344, | |
| "learning_rate": 0.00015086532155617784, | |
| "loss": 0.0813, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.12785181430507606, | |
| "learning_rate": 0.00015021680451171498, | |
| "loss": 0.0802, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.206451612903226, | |
| "grad_norm": 0.14298258948499543, | |
| "learning_rate": 0.00014956545089264407, | |
| "loss": 0.0843, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2129032258064516, | |
| "grad_norm": 0.20148810623427083, | |
| "learning_rate": 0.0001489112974916912, | |
| "loss": 0.0942, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2193548387096773, | |
| "grad_norm": 0.14657915199625932, | |
| "learning_rate": 0.00014825438125973264, | |
| "loss": 0.0829, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2258064516129032, | |
| "grad_norm": 0.16147059815020365, | |
| "learning_rate": 0.00014759473930370736, | |
| "loss": 0.0756, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.232258064516129, | |
| "grad_norm": 0.17956881851269735, | |
| "learning_rate": 0.0001469324088845212, | |
| "loss": 0.1007, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.238709677419355, | |
| "grad_norm": 0.1578318532174886, | |
| "learning_rate": 0.00014626742741494206, | |
| "loss": 0.0983, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2451612903225806, | |
| "grad_norm": 0.1649792762044239, | |
| "learning_rate": 0.00014559983245748638, | |
| "loss": 0.0905, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2516129032258063, | |
| "grad_norm": 0.1455653603561888, | |
| "learning_rate": 0.00014492966172229777, | |
| "loss": 0.0791, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2580645161290323, | |
| "grad_norm": 0.1312841252501999, | |
| "learning_rate": 0.00014425695306501658, | |
| "loss": 0.0613, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2645161290322582, | |
| "grad_norm": 0.1361347125426188, | |
| "learning_rate": 0.00014358174448464154, | |
| "loss": 0.0773, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.270967741935484, | |
| "grad_norm": 0.17551022044504175, | |
| "learning_rate": 0.00014290407412138366, | |
| "loss": 0.0985, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2774193548387096, | |
| "grad_norm": 0.1456899710914213, | |
| "learning_rate": 0.00014222398025451135, | |
| "loss": 0.0827, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.2838709677419355, | |
| "grad_norm": 0.19482064326323745, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.0974, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2903225806451613, | |
| "grad_norm": 0.13544624920378107, | |
| "learning_rate": 0.0001408566758093048, | |
| "loss": 0.0742, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2967741935483872, | |
| "grad_norm": 0.12685823273525554, | |
| "learning_rate": 0.00014016954246529696, | |
| "loss": 0.0677, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.303225806451613, | |
| "grad_norm": 0.15457862761329938, | |
| "learning_rate": 0.00013948014008196487, | |
| "loss": 0.0815, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.3096774193548386, | |
| "grad_norm": 0.1482256655702007, | |
| "learning_rate": 0.0001387885076012785, | |
| "loss": 0.0961, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.3161290322580645, | |
| "grad_norm": 0.20896278322687534, | |
| "learning_rate": 0.00013809468409117846, | |
| "loss": 0.1049, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.3225806451612903, | |
| "grad_norm": 0.175019480492322, | |
| "learning_rate": 0.00013739870874336898, | |
| "loss": 0.0884, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3290322580645162, | |
| "grad_norm": 0.14363356864891735, | |
| "learning_rate": 0.00013670062087110422, | |
| "loss": 0.0864, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.335483870967742, | |
| "grad_norm": 0.1397081258858762, | |
| "learning_rate": 0.00013600045990696762, | |
| "loss": 0.0797, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3419354838709676, | |
| "grad_norm": 0.1667674455506937, | |
| "learning_rate": 0.0001352982654006444, | |
| "loss": 0.0977, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3483870967741935, | |
| "grad_norm": 0.1409898972781143, | |
| "learning_rate": 0.00013459407701668763, | |
| "loss": 0.0923, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3548387096774195, | |
| "grad_norm": 0.16290520924324786, | |
| "learning_rate": 0.00013388793453227767, | |
| "loss": 0.1033, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3612903225806452, | |
| "grad_norm": 0.14967784709530377, | |
| "learning_rate": 0.0001331798778349752, | |
| "loss": 0.0843, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.367741935483871, | |
| "grad_norm": 0.14370370779202904, | |
| "learning_rate": 0.00013246994692046836, | |
| "loss": 0.0845, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.3741935483870968, | |
| "grad_norm": 0.15288920332301245, | |
| "learning_rate": 0.00013175818189031327, | |
| "loss": 0.0962, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3806451612903226, | |
| "grad_norm": 0.16907810212754584, | |
| "learning_rate": 0.00013104462294966896, | |
| "loss": 0.1033, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.3870967741935485, | |
| "grad_norm": 0.17023880837738012, | |
| "learning_rate": 0.00013032931040502627, | |
| "loss": 0.1155, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3935483870967742, | |
| "grad_norm": 0.12189251069561345, | |
| "learning_rate": 0.00012961228466193116, | |
| "loss": 0.0692, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.18108512316647296, | |
| "learning_rate": 0.00012889358622270223, | |
| "loss": 0.086, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4064516129032258, | |
| "grad_norm": 0.15804455793477432, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 0.0751, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.4129032258064516, | |
| "grad_norm": 0.13668879316730062, | |
| "learning_rate": 0.00012745133373524853, | |
| "loss": 0.0786, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.4193548387096775, | |
| "grad_norm": 0.18610949095399393, | |
| "learning_rate": 0.0001267278611549073, | |
| "loss": 0.0732, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4258064516129032, | |
| "grad_norm": 0.186875270696186, | |
| "learning_rate": 0.00012600287880959763, | |
| "loss": 0.1051, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.432258064516129, | |
| "grad_norm": 0.15558801012557805, | |
| "learning_rate": 0.0001252764276510792, | |
| "loss": 0.0879, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4387096774193548, | |
| "grad_norm": 0.16188457902393685, | |
| "learning_rate": 0.00012454854871407994, | |
| "loss": 0.0887, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.4451612903225808, | |
| "grad_norm": 0.1332789968563102, | |
| "learning_rate": 0.00012381928311397806, | |
| "loss": 0.0819, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4516129032258065, | |
| "grad_norm": 0.13104439295337186, | |
| "learning_rate": 0.0001230886720444796, | |
| "loss": 0.0992, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4580645161290322, | |
| "grad_norm": 0.1741921210517873, | |
| "learning_rate": 0.00012235675677529158, | |
| "loss": 0.1049, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.4645161290322581, | |
| "grad_norm": 0.17686059237450052, | |
| "learning_rate": 0.00012162357864979072, | |
| "loss": 0.0942, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4709677419354839, | |
| "grad_norm": 0.14208255193217198, | |
| "learning_rate": 0.00012088917908268821, | |
| "loss": 0.0932, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4774193548387098, | |
| "grad_norm": 0.16207984639669018, | |
| "learning_rate": 0.00012015359955769021, | |
| "loss": 0.0952, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4838709677419355, | |
| "grad_norm": 0.10567516069189406, | |
| "learning_rate": 0.00011941688162515467, | |
| "loss": 0.0698, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4903225806451612, | |
| "grad_norm": 0.14315458863752037, | |
| "learning_rate": 0.00011867906689974428, | |
| "loss": 0.0706, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.4967741935483871, | |
| "grad_norm": 0.14644874259664967, | |
| "learning_rate": 0.00011794019705807584, | |
| "loss": 0.0954, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.5032258064516129, | |
| "grad_norm": 0.1522258926853921, | |
| "learning_rate": 0.00011720031383636585, | |
| "loss": 0.1026, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.5096774193548388, | |
| "grad_norm": 0.2389589012648026, | |
| "learning_rate": 0.00011645945902807341, | |
| "loss": 0.0981, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.5161290322580645, | |
| "grad_norm": 0.1539160607687386, | |
| "learning_rate": 0.00011571767448153901, | |
| "loss": 0.074, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5225806451612902, | |
| "grad_norm": 0.17340052794011998, | |
| "learning_rate": 0.00011497500209762102, | |
| "loss": 0.0943, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.5290322580645161, | |
| "grad_norm": 0.11294207936842038, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 0.0644, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.535483870967742, | |
| "grad_norm": 0.14391081206665055, | |
| "learning_rate": 0.00011348716166945195, | |
| "loss": 0.0774, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5419354838709678, | |
| "grad_norm": 0.16298875474891092, | |
| "learning_rate": 0.0001127420776681905, | |
| "loss": 0.0786, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5483870967741935, | |
| "grad_norm": 0.1325076119120941, | |
| "learning_rate": 0.00011199627391077732, | |
| "loss": 0.0889, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5548387096774192, | |
| "grad_norm": 0.10971482413094012, | |
| "learning_rate": 0.00011124979252510208, | |
| "loss": 0.0737, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.5612903225806452, | |
| "grad_norm": 0.14614299703062, | |
| "learning_rate": 0.0001105026756773314, | |
| "loss": 0.0682, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.567741935483871, | |
| "grad_norm": 0.22690104820583093, | |
| "learning_rate": 0.00010975496556952682, | |
| "loss": 0.1094, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5741935483870968, | |
| "grad_norm": 0.2561727438038473, | |
| "learning_rate": 0.00010900670443726135, | |
| "loss": 0.0851, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5806451612903225, | |
| "grad_norm": 0.15371305701947427, | |
| "learning_rate": 0.00010825793454723325, | |
| "loss": 0.0923, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5870967741935482, | |
| "grad_norm": 0.16267574566743875, | |
| "learning_rate": 0.00010750869819487883, | |
| "loss": 0.1036, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.5935483870967742, | |
| "grad_norm": 0.15591155698404394, | |
| "learning_rate": 0.00010675903770198333, | |
| "loss": 0.0893, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.14338972962339533, | |
| "learning_rate": 0.00010600899541429004, | |
| "loss": 0.0837, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.6064516129032258, | |
| "grad_norm": 0.12387607320751257, | |
| "learning_rate": 0.00010525861369910877, | |
| "loss": 0.0755, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.6129032258064515, | |
| "grad_norm": 0.16606169294386383, | |
| "learning_rate": 0.00010450793494292224, | |
| "loss": 0.1043, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6193548387096774, | |
| "grad_norm": 0.1795920159350681, | |
| "learning_rate": 0.00010375700154899208, | |
| "loss": 0.1008, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.6258064516129034, | |
| "grad_norm": 0.13025097291519463, | |
| "learning_rate": 0.00010300585593496348, | |
| "loss": 0.0851, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.632258064516129, | |
| "grad_norm": 0.14349816154023654, | |
| "learning_rate": 0.00010225454053046921, | |
| "loss": 0.0807, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6387096774193548, | |
| "grad_norm": 0.16695448888874226, | |
| "learning_rate": 0.00010150309777473306, | |
| "loss": 0.1117, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6451612903225805, | |
| "grad_norm": 0.15743712528269815, | |
| "learning_rate": 0.0001007515701141722, | |
| "loss": 0.086, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6516129032258065, | |
| "grad_norm": 0.16704335210894908, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0886, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6580645161290324, | |
| "grad_norm": 0.10245531520994122, | |
| "learning_rate": 9.924842988582782e-05, | |
| "loss": 0.0678, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.664516129032258, | |
| "grad_norm": 0.1688033836669086, | |
| "learning_rate": 9.849690222526698e-05, | |
| "loss": 0.0958, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6709677419354838, | |
| "grad_norm": 0.1338126992775965, | |
| "learning_rate": 9.77454594695308e-05, | |
| "loss": 0.0782, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6774193548387095, | |
| "grad_norm": 0.1332820451604909, | |
| "learning_rate": 9.699414406503654e-05, | |
| "loss": 0.0856, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.6838709677419355, | |
| "grad_norm": 0.1449401720605745, | |
| "learning_rate": 9.624299845100795e-05, | |
| "loss": 0.0882, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6903225806451614, | |
| "grad_norm": 0.19395703151279187, | |
| "learning_rate": 9.549206505707777e-05, | |
| "loss": 0.1148, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.696774193548387, | |
| "grad_norm": 0.13482277559178169, | |
| "learning_rate": 9.474138630089124e-05, | |
| "loss": 0.0644, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.7032258064516128, | |
| "grad_norm": 0.18530593840863338, | |
| "learning_rate": 9.399100458570997e-05, | |
| "loss": 0.1074, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.7096774193548387, | |
| "grad_norm": 0.1897995033582595, | |
| "learning_rate": 9.324096229801674e-05, | |
| "loss": 0.0867, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.7161290322580647, | |
| "grad_norm": 0.16472780150681127, | |
| "learning_rate": 9.249130180512118e-05, | |
| "loss": 0.0896, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.7225806451612904, | |
| "grad_norm": 0.15525608679066774, | |
| "learning_rate": 9.174206545276677e-05, | |
| "loss": 0.0865, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.729032258064516, | |
| "grad_norm": 0.15426999703424252, | |
| "learning_rate": 9.099329556273866e-05, | |
| "loss": 0.0853, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7354838709677418, | |
| "grad_norm": 0.19103115451320254, | |
| "learning_rate": 9.024503443047319e-05, | |
| "loss": 0.0993, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.7419354838709677, | |
| "grad_norm": 0.12323460303269068, | |
| "learning_rate": 8.949732432266866e-05, | |
| "loss": 0.0723, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7483870967741937, | |
| "grad_norm": 0.1339668030381976, | |
| "learning_rate": 8.875020747489794e-05, | |
| "loss": 0.0852, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.7548387096774194, | |
| "grad_norm": 0.1385020226698345, | |
| "learning_rate": 8.800372608922271e-05, | |
| "loss": 0.0773, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.761290322580645, | |
| "grad_norm": 0.21835470061774626, | |
| "learning_rate": 8.72579223318095e-05, | |
| "loss": 0.1167, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7677419354838708, | |
| "grad_norm": 0.1425322245524885, | |
| "learning_rate": 8.651283833054809e-05, | |
| "loss": 0.0801, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.7741935483870968, | |
| "grad_norm": 0.1794495686810833, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 0.0784, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7806451612903227, | |
| "grad_norm": 0.15688310999446023, | |
| "learning_rate": 8.5024997902379e-05, | |
| "loss": 0.1001, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.7870967741935484, | |
| "grad_norm": 0.15417421931142297, | |
| "learning_rate": 8.428232551846101e-05, | |
| "loss": 0.0898, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.793548387096774, | |
| "grad_norm": 0.17958092903510822, | |
| "learning_rate": 8.35405409719266e-05, | |
| "loss": 0.0921, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.165061388870699, | |
| "learning_rate": 8.279968616363418e-05, | |
| "loss": 0.0935, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.8064516129032258, | |
| "grad_norm": 0.1116000947584692, | |
| "learning_rate": 8.205980294192421e-05, | |
| "loss": 0.0684, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.8129032258064517, | |
| "grad_norm": 0.12370745355237868, | |
| "learning_rate": 8.132093310025571e-05, | |
| "loss": 0.0764, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.8193548387096774, | |
| "grad_norm": 0.1724137684896053, | |
| "learning_rate": 8.058311837484535e-05, | |
| "loss": 0.0969, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.8258064516129031, | |
| "grad_norm": 0.16125193051943912, | |
| "learning_rate": 7.984640044230983e-05, | |
| "loss": 0.0868, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.832258064516129, | |
| "grad_norm": 0.13929471780512187, | |
| "learning_rate": 7.911082091731181e-05, | |
| "loss": 0.0701, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.838709677419355, | |
| "grad_norm": 0.1172859356132756, | |
| "learning_rate": 7.837642135020929e-05, | |
| "loss": 0.0705, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.8451612903225807, | |
| "grad_norm": 0.1416371336298496, | |
| "learning_rate": 7.764324322470841e-05, | |
| "loss": 0.0683, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8516129032258064, | |
| "grad_norm": 0.19198949443360017, | |
| "learning_rate": 7.691132795552043e-05, | |
| "loss": 0.0894, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8580645161290321, | |
| "grad_norm": 0.19384526534149363, | |
| "learning_rate": 7.618071688602199e-05, | |
| "loss": 0.0954, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.864516129032258, | |
| "grad_norm": 0.21508649416468206, | |
| "learning_rate": 7.54514512859201e-05, | |
| "loss": 0.0986, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.870967741935484, | |
| "grad_norm": 0.1952670486171591, | |
| "learning_rate": 7.472357234892082e-05, | |
| "loss": 0.1049, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8774193548387097, | |
| "grad_norm": 0.13259304301639457, | |
| "learning_rate": 7.399712119040238e-05, | |
| "loss": 0.0784, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.8838709677419354, | |
| "grad_norm": 0.1367442063153438, | |
| "learning_rate": 7.327213884509272e-05, | |
| "loss": 0.0828, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.8903225806451613, | |
| "grad_norm": 0.16389743747526248, | |
| "learning_rate": 7.254866626475152e-05, | |
| "loss": 0.0927, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.896774193548387, | |
| "grad_norm": 0.22089121957380548, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 0.1077, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.903225806451613, | |
| "grad_norm": 0.14882797154919192, | |
| "learning_rate": 7.110641377729778e-05, | |
| "loss": 0.0879, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.9096774193548387, | |
| "grad_norm": 0.12794357419440644, | |
| "learning_rate": 7.038771533806884e-05, | |
| "loss": 0.0719, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.9161290322580644, | |
| "grad_norm": 0.1846824594193936, | |
| "learning_rate": 6.967068959497376e-05, | |
| "loss": 0.0895, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.9225806451612903, | |
| "grad_norm": 0.17646276082900025, | |
| "learning_rate": 6.895537705033108e-05, | |
| "loss": 0.0996, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9290322580645163, | |
| "grad_norm": 0.15604696338212207, | |
| "learning_rate": 6.824181810968675e-05, | |
| "loss": 0.087, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.935483870967742, | |
| "grad_norm": 0.15478000466177547, | |
| "learning_rate": 6.753005307953167e-05, | |
| "loss": 0.0831, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9419354838709677, | |
| "grad_norm": 0.19642059718505833, | |
| "learning_rate": 6.682012216502484e-05, | |
| "loss": 0.1073, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.9483870967741934, | |
| "grad_norm": 0.17255606883063224, | |
| "learning_rate": 6.611206546772237e-05, | |
| "loss": 0.1018, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.9548387096774194, | |
| "grad_norm": 0.1393602083174125, | |
| "learning_rate": 6.54059229833124e-05, | |
| "loss": 0.0859, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9612903225806453, | |
| "grad_norm": 0.16652643955266602, | |
| "learning_rate": 6.47017345993556e-05, | |
| "loss": 0.0899, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.967741935483871, | |
| "grad_norm": 0.16878660937501272, | |
| "learning_rate": 6.39995400930324e-05, | |
| "loss": 0.0998, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9741935483870967, | |
| "grad_norm": 0.12969061293334938, | |
| "learning_rate": 6.329937912889582e-05, | |
| "loss": 0.0706, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.9806451612903224, | |
| "grad_norm": 0.15739894009886643, | |
| "learning_rate": 6.260129125663106e-05, | |
| "loss": 0.0888, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.9870967741935484, | |
| "grad_norm": 0.1752067977727585, | |
| "learning_rate": 6.190531590882159e-05, | |
| "loss": 0.0826, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.9935483870967743, | |
| "grad_norm": 0.21002280616145647, | |
| "learning_rate": 6.121149239872151e-05, | |
| "loss": 0.123, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.13850366944373244, | |
| "learning_rate": 6.051985991803517e-05, | |
| "loss": 0.073, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.16690203547477722, | |
| "eval_runtime": 25.4682, | |
| "eval_samples_per_second": 5.144, | |
| "eval_steps_per_second": 0.668, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.0064516129032257, | |
| "grad_norm": 0.1124990090834071, | |
| "learning_rate": 5.983045753470308e-05, | |
| "loss": 0.0485, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.0129032258064514, | |
| "grad_norm": 0.08246086368829486, | |
| "learning_rate": 5.9143324190695196e-05, | |
| "loss": 0.0411, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.0193548387096776, | |
| "grad_norm": 0.11824209839228844, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 0.0578, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.0258064516129033, | |
| "grad_norm": 0.09183702414760068, | |
| "learning_rate": 5.777601974548866e-05, | |
| "loss": 0.0405, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.032258064516129, | |
| "grad_norm": 0.10138990147374953, | |
| "learning_rate": 5.709592587861637e-05, | |
| "loss": 0.045, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.0387096774193547, | |
| "grad_norm": 0.13174518232765414, | |
| "learning_rate": 5.6418255515358486e-05, | |
| "loss": 0.0417, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.0451612903225804, | |
| "grad_norm": 0.11134581768568716, | |
| "learning_rate": 5.574304693498346e-05, | |
| "loss": 0.0351, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.0516129032258066, | |
| "grad_norm": 0.10372734844639428, | |
| "learning_rate": 5.507033827770225e-05, | |
| "loss": 0.0458, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.0580645161290323, | |
| "grad_norm": 0.09893496819685892, | |
| "learning_rate": 5.4400167542513636e-05, | |
| "loss": 0.0421, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.064516129032258, | |
| "grad_norm": 0.1544849048513997, | |
| "learning_rate": 5.3732572585057974e-05, | |
| "loss": 0.0449, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0709677419354837, | |
| "grad_norm": 0.12211629106983518, | |
| "learning_rate": 5.306759111547881e-05, | |
| "loss": 0.0341, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.07741935483871, | |
| "grad_norm": 0.1519039459170261, | |
| "learning_rate": 5.240526069629265e-05, | |
| "loss": 0.0477, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.0838709677419356, | |
| "grad_norm": 0.10882973801143421, | |
| "learning_rate": 5.174561874026741e-05, | |
| "loss": 0.0376, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.0903225806451613, | |
| "grad_norm": 0.09785322895617321, | |
| "learning_rate": 5.108870250830882e-05, | |
| "loss": 0.0305, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.096774193548387, | |
| "grad_norm": 0.14990301799413153, | |
| "learning_rate": 5.0434549107355944e-05, | |
| "loss": 0.0409, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.1032258064516127, | |
| "grad_norm": 0.1915279670852206, | |
| "learning_rate": 4.978319548828504e-05, | |
| "loss": 0.045, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.109677419354839, | |
| "grad_norm": 0.15445162243894608, | |
| "learning_rate": 4.9134678443822166e-05, | |
| "loss": 0.0334, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.1161290322580646, | |
| "grad_norm": 0.34562522599625123, | |
| "learning_rate": 4.8489034606465225e-05, | |
| "loss": 0.0395, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.1225806451612903, | |
| "grad_norm": 0.3183911745576579, | |
| "learning_rate": 4.784630044641435e-05, | |
| "loss": 0.0546, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.129032258064516, | |
| "grad_norm": 0.1619196710614726, | |
| "learning_rate": 4.7206512269512124e-05, | |
| "loss": 0.0446, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.135483870967742, | |
| "grad_norm": 0.24657637060341897, | |
| "learning_rate": 4.65697062151927e-05, | |
| "loss": 0.0349, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.141935483870968, | |
| "grad_norm": 0.11723613581448042, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 0.0337, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.1483870967741936, | |
| "grad_norm": 0.12413977056977685, | |
| "learning_rate": 4.530518418775733e-05, | |
| "loss": 0.0355, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.1548387096774193, | |
| "grad_norm": 0.1368012687546921, | |
| "learning_rate": 4.4677539643142454e-05, | |
| "loss": 0.0369, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.161290322580645, | |
| "grad_norm": 0.132092425970093, | |
| "learning_rate": 4.40530200740777e-05, | |
| "loss": 0.0346, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.167741935483871, | |
| "grad_norm": 0.17255652388608525, | |
| "learning_rate": 4.343166075752605e-05, | |
| "loss": 0.0425, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.174193548387097, | |
| "grad_norm": 0.10537625462278236, | |
| "learning_rate": 4.281349679193861e-05, | |
| "loss": 0.0363, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.1806451612903226, | |
| "grad_norm": 0.08446560264535163, | |
| "learning_rate": 4.2198563095272116e-05, | |
| "loss": 0.0303, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.1870967741935483, | |
| "grad_norm": 0.11155624490505092, | |
| "learning_rate": 4.158689440301657e-05, | |
| "loss": 0.0424, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.193548387096774, | |
| "grad_norm": 0.10614591478252769, | |
| "learning_rate": 4.097852526623307e-05, | |
| "loss": 0.0374, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.15380955207075353, | |
| "learning_rate": 4.0373490049602204e-05, | |
| "loss": 0.0506, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.206451612903226, | |
| "grad_norm": 0.11905414571153279, | |
| "learning_rate": 3.977182292948283e-05, | |
| "loss": 0.0517, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.2129032258064516, | |
| "grad_norm": 0.14425425935781702, | |
| "learning_rate": 3.9173557891981573e-05, | |
| "loss": 0.0467, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.2193548387096773, | |
| "grad_norm": 0.12453230757386329, | |
| "learning_rate": 3.857872873103322e-05, | |
| "loss": 0.0435, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.225806451612903, | |
| "grad_norm": 0.10038267160538758, | |
| "learning_rate": 3.7987369046491684e-05, | |
| "loss": 0.0387, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.232258064516129, | |
| "grad_norm": 0.1438025015993235, | |
| "learning_rate": 3.7399512242231995e-05, | |
| "loss": 0.0346, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.238709677419355, | |
| "grad_norm": 0.26072179652296745, | |
| "learning_rate": 3.6815191524263624e-05, | |
| "loss": 0.0545, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.2451612903225806, | |
| "grad_norm": 0.08770511437142381, | |
| "learning_rate": 3.623443989885462e-05, | |
| "loss": 0.0327, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.2516129032258063, | |
| "grad_norm": 0.12301181296016188, | |
| "learning_rate": 3.565729017066729e-05, | |
| "loss": 0.039, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.258064516129032, | |
| "grad_norm": 0.13814667341858822, | |
| "learning_rate": 3.508377494090521e-05, | |
| "loss": 0.0356, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.264516129032258, | |
| "grad_norm": 0.12413533495298362, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 0.0415, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.270967741935484, | |
| "grad_norm": 0.11031185879435731, | |
| "learning_rate": 3.394777735313919e-05, | |
| "loss": 0.0363, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.2774193548387096, | |
| "grad_norm": 0.11353520726841541, | |
| "learning_rate": 3.338535916373266e-05, | |
| "loss": 0.0376, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.2838709677419353, | |
| "grad_norm": 0.23000368159173218, | |
| "learning_rate": 3.2826703806321525e-05, | |
| "loss": 0.0444, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.2903225806451615, | |
| "grad_norm": 0.10390022981327145, | |
| "learning_rate": 3.227184283742591e-05, | |
| "loss": 0.0308, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.296774193548387, | |
| "grad_norm": 0.14881153611607173, | |
| "learning_rate": 3.17208075992339e-05, | |
| "loss": 0.0428, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.303225806451613, | |
| "grad_norm": 0.22847373170988944, | |
| "learning_rate": 3.117362921783134e-05, | |
| "loss": 0.0449, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.3096774193548386, | |
| "grad_norm": 0.1261660214676178, | |
| "learning_rate": 3.063033860144339e-05, | |
| "loss": 0.0354, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.3161290322580643, | |
| "grad_norm": 0.16508383182402034, | |
| "learning_rate": 3.0090966438688772e-05, | |
| "loss": 0.0404, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.3225806451612905, | |
| "grad_norm": 0.14003607495149023, | |
| "learning_rate": 2.9555543196846292e-05, | |
| "loss": 0.0417, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.329032258064516, | |
| "grad_norm": 0.14067522061651452, | |
| "learning_rate": 2.9024099120133673e-05, | |
| "loss": 0.0325, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.335483870967742, | |
| "grad_norm": 0.17575236098692995, | |
| "learning_rate": 2.8496664227999415e-05, | |
| "loss": 0.0533, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.3419354838709676, | |
| "grad_norm": 0.11116834491408094, | |
| "learning_rate": 2.7973268313426837e-05, | |
| "loss": 0.0347, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.3483870967741938, | |
| "grad_norm": 0.16113939577201203, | |
| "learning_rate": 2.745394094125141e-05, | |
| "loss": 0.0472, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.3548387096774195, | |
| "grad_norm": 0.10955543396615161, | |
| "learning_rate": 2.6938711446490606e-05, | |
| "loss": 0.0357, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.361290322580645, | |
| "grad_norm": 0.17969798691111394, | |
| "learning_rate": 2.6427608932686843e-05, | |
| "loss": 0.0584, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.367741935483871, | |
| "grad_norm": 0.08542932983952008, | |
| "learning_rate": 2.5920662270263653e-05, | |
| "loss": 0.032, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.3741935483870966, | |
| "grad_norm": 0.14296192462496796, | |
| "learning_rate": 2.5417900094894744e-05, | |
| "loss": 0.0463, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.3806451612903228, | |
| "grad_norm": 0.11857064948287055, | |
| "learning_rate": 2.4919350805886577e-05, | |
| "loss": 0.0328, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.3870967741935485, | |
| "grad_norm": 0.1614783419773078, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 0.0398, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.393548387096774, | |
| "grad_norm": 0.12389155943230613, | |
| "learning_rate": 2.3935003292730296e-05, | |
| "loss": 0.0389, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.11380344182842446, | |
| "learning_rate": 2.344926067098836e-05, | |
| "loss": 0.0412, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.4064516129032256, | |
| "grad_norm": 0.09336252258103142, | |
| "learning_rate": 2.2967842137278706e-05, | |
| "loss": 0.0311, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.412903225806452, | |
| "grad_norm": 0.1312469591783424, | |
| "learning_rate": 2.2490774885278908e-05, | |
| "loss": 0.0363, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.4193548387096775, | |
| "grad_norm": 0.1928829483525128, | |
| "learning_rate": 2.201808586287757e-05, | |
| "loss": 0.0447, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.425806451612903, | |
| "grad_norm": 0.1407497782508695, | |
| "learning_rate": 2.15498017706521e-05, | |
| "loss": 0.0425, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.432258064516129, | |
| "grad_norm": 0.12625194439970722, | |
| "learning_rate": 2.1085949060360654e-05, | |
| "loss": 0.0394, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.4387096774193546, | |
| "grad_norm": 0.1650711686265322, | |
| "learning_rate": 2.0626553933447734e-05, | |
| "loss": 0.0356, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.445161290322581, | |
| "grad_norm": 0.203794329158885, | |
| "learning_rate": 2.01716423395644e-05, | |
| "loss": 0.046, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.4516129032258065, | |
| "grad_norm": 0.0983865374506401, | |
| "learning_rate": 1.9721239975102313e-05, | |
| "loss": 0.0383, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.458064516129032, | |
| "grad_norm": 0.11475237817508949, | |
| "learning_rate": 1.9275372281742242e-05, | |
| "loss": 0.0341, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.464516129032258, | |
| "grad_norm": 0.11696076741124051, | |
| "learning_rate": 1.8834064445016953e-05, | |
| "loss": 0.0351, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.4709677419354836, | |
| "grad_norm": 0.11440255447742229, | |
| "learning_rate": 1.839734139288868e-05, | |
| "loss": 0.0344, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.47741935483871, | |
| "grad_norm": 0.12722166866147971, | |
| "learning_rate": 1.7965227794340877e-05, | |
| "loss": 0.0327, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.4838709677419355, | |
| "grad_norm": 0.11977054711252778, | |
| "learning_rate": 1.753774805798486e-05, | |
| "loss": 0.0374, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.490322580645161, | |
| "grad_norm": 0.19041811914083892, | |
| "learning_rate": 1.7114926330680957e-05, | |
| "loss": 0.0448, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.496774193548387, | |
| "grad_norm": 0.12908385372034334, | |
| "learning_rate": 1.6696786496174578e-05, | |
| "loss": 0.0421, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.5032258064516126, | |
| "grad_norm": 0.11345958947234494, | |
| "learning_rate": 1.6283352173747145e-05, | |
| "loss": 0.0337, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.509677419354839, | |
| "grad_norm": 0.13882462080577593, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 0.0399, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.5161290322580645, | |
| "grad_norm": 0.15241110627485502, | |
| "learning_rate": 1.5470693211944643e-05, | |
| "loss": 0.0331, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.52258064516129, | |
| "grad_norm": 0.10975692065298487, | |
| "learning_rate": 1.5071514476879878e-05, | |
| "loss": 0.0379, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.5290322580645164, | |
| "grad_norm": 0.1379096296948562, | |
| "learning_rate": 1.4677133059921632e-05, | |
| "loss": 0.0362, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.535483870967742, | |
| "grad_norm": 0.10190107251103887, | |
| "learning_rate": 1.4287571238320053e-05, | |
| "loss": 0.0371, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.541935483870968, | |
| "grad_norm": 0.0875699618751271, | |
| "learning_rate": 1.3902851017082864e-05, | |
| "loss": 0.0335, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.5483870967741935, | |
| "grad_norm": 0.13655405710076232, | |
| "learning_rate": 1.3522994127732414e-05, | |
| "loss": 0.0388, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.554838709677419, | |
| "grad_norm": 0.11073272357568731, | |
| "learning_rate": 1.3148022027078222e-05, | |
| "loss": 0.0366, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.5612903225806454, | |
| "grad_norm": 0.11393589351391524, | |
| "learning_rate": 1.2777955896004812e-05, | |
| "loss": 0.0375, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.567741935483871, | |
| "grad_norm": 0.1294793091961239, | |
| "learning_rate": 1.2412816638275404e-05, | |
| "loss": 0.0447, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.574193548387097, | |
| "grad_norm": 0.12252957295031998, | |
| "learning_rate": 1.2052624879351104e-05, | |
| "loss": 0.036, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.5806451612903225, | |
| "grad_norm": 0.10701768631052949, | |
| "learning_rate": 1.1697400965225747e-05, | |
| "loss": 0.0375, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.587096774193548, | |
| "grad_norm": 0.1826126683217801, | |
| "learning_rate": 1.134716496127679e-05, | |
| "loss": 0.0398, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.5935483870967744, | |
| "grad_norm": 0.12179126729794809, | |
| "learning_rate": 1.1001936651131717e-05, | |
| "loss": 0.0396, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.11972903116479075, | |
| "learning_rate": 1.0661735535550666e-05, | |
| "loss": 0.0336, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.606451612903226, | |
| "grad_norm": 0.09797391900877583, | |
| "learning_rate": 1.0326580831324817e-05, | |
| "loss": 0.0291, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.6129032258064515, | |
| "grad_norm": 0.08807259670414254, | |
| "learning_rate": 9.996491470190917e-06, | |
| "loss": 0.0318, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.6193548387096772, | |
| "grad_norm": 0.12986706205285126, | |
| "learning_rate": 9.671486097761917e-06, | |
| "loss": 0.0408, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.6258064516129034, | |
| "grad_norm": 0.1331746828106328, | |
| "learning_rate": 9.351583072473713e-06, | |
| "loss": 0.0427, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.632258064516129, | |
| "grad_norm": 0.14912986532560119, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 0.0464, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.638709677419355, | |
| "grad_norm": 0.1440292521357826, | |
| "learning_rate": 8.727156054972374e-06, | |
| "loss": 0.0448, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.6451612903225805, | |
| "grad_norm": 0.16571651133316453, | |
| "learning_rate": 8.422667334494249e-06, | |
| "loss": 0.0449, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.6516129032258062, | |
| "grad_norm": 0.15006924502881824, | |
| "learning_rate": 8.123351502634625e-06, | |
| "loss": 0.0371, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.6580645161290324, | |
| "grad_norm": 0.3766926409227814, | |
| "learning_rate": 7.82922546671555e-06, | |
| "loss": 0.0522, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.664516129032258, | |
| "grad_norm": 0.1693245533307095, | |
| "learning_rate": 7.54030584090537e-06, | |
| "loss": 0.0523, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.670967741935484, | |
| "grad_norm": 0.16933380856734512, | |
| "learning_rate": 7.256608945280319e-06, | |
| "loss": 0.0448, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.6774193548387095, | |
| "grad_norm": 0.12372656470631721, | |
| "learning_rate": 6.97815080490245e-06, | |
| "loss": 0.041, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.6838709677419352, | |
| "grad_norm": 0.1139014768011305, | |
| "learning_rate": 6.704947148914609e-06, | |
| "loss": 0.034, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.6903225806451614, | |
| "grad_norm": 0.14509781428626975, | |
| "learning_rate": 6.437013409651849e-06, | |
| "loss": 0.0289, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.696774193548387, | |
| "grad_norm": 0.1403482292705624, | |
| "learning_rate": 6.174364721769743e-06, | |
| "loss": 0.0493, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.703225806451613, | |
| "grad_norm": 0.12152856591733673, | |
| "learning_rate": 5.917015921389568e-06, | |
| "loss": 0.0343, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.709677419354839, | |
| "grad_norm": 0.10494471511565585, | |
| "learning_rate": 5.664981545260073e-06, | |
| "loss": 0.03, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.7161290322580647, | |
| "grad_norm": 0.2176951472197835, | |
| "learning_rate": 5.418275829936537e-06, | |
| "loss": 0.0486, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.7225806451612904, | |
| "grad_norm": 0.09677627092271344, | |
| "learning_rate": 5.176912710976467e-06, | |
| "loss": 0.0318, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.729032258064516, | |
| "grad_norm": 0.14416553291024004, | |
| "learning_rate": 4.940905822152453e-06, | |
| "loss": 0.0356, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.735483870967742, | |
| "grad_norm": 0.10141360651760384, | |
| "learning_rate": 4.710268494682146e-06, | |
| "loss": 0.0346, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.741935483870968, | |
| "grad_norm": 0.13463566674635727, | |
| "learning_rate": 4.485013756475076e-06, | |
| "loss": 0.0443, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.7483870967741937, | |
| "grad_norm": 0.26842778961550184, | |
| "learning_rate": 4.2651543313968145e-06, | |
| "loss": 0.0445, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.7548387096774194, | |
| "grad_norm": 0.16014595507517537, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 0.0469, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.761290322580645, | |
| "grad_norm": 0.2690022963645026, | |
| "learning_rate": 3.841670791574137e-06, | |
| "loss": 0.0423, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.767741935483871, | |
| "grad_norm": 0.2323526833237967, | |
| "learning_rate": 3.638070597958665e-06, | |
| "loss": 0.035, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.774193548387097, | |
| "grad_norm": 0.11643458974488918, | |
| "learning_rate": 3.4399135583787043e-06, | |
| "loss": 0.0401, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.7806451612903227, | |
| "grad_norm": 0.12278976372135998, | |
| "learning_rate": 3.2472108660439706e-06, | |
| "loss": 0.04, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.7870967741935484, | |
| "grad_norm": 0.11369002878299092, | |
| "learning_rate": 3.059973406066963e-06, | |
| "loss": 0.0335, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.793548387096774, | |
| "grad_norm": 0.1426764505452139, | |
| "learning_rate": 2.878211754847926e-06, | |
| "loss": 0.0431, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.17614309627774977, | |
| "learning_rate": 2.7019361794775156e-06, | |
| "loss": 0.0336, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.806451612903226, | |
| "grad_norm": 0.13119646329419238, | |
| "learning_rate": 2.5311566371568507e-06, | |
| "loss": 0.0388, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.8129032258064517, | |
| "grad_norm": 0.2274358694692015, | |
| "learning_rate": 2.365882774634998e-06, | |
| "loss": 0.0483, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.8193548387096774, | |
| "grad_norm": 0.16809646480030987, | |
| "learning_rate": 2.206123927664161e-06, | |
| "loss": 0.0428, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.825806451612903, | |
| "grad_norm": 0.08861174696327136, | |
| "learning_rate": 2.0518891204722168e-06, | |
| "loss": 0.0297, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.832258064516129, | |
| "grad_norm": 0.1516310249747229, | |
| "learning_rate": 1.903187065253076e-06, | |
| "loss": 0.0486, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.838709677419355, | |
| "grad_norm": 0.1154092981393085, | |
| "learning_rate": 1.7600261616745106e-06, | |
| "loss": 0.044, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.8451612903225807, | |
| "grad_norm": 0.14626098217303432, | |
| "learning_rate": 1.6224144964036681e-06, | |
| "loss": 0.045, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.8516129032258064, | |
| "grad_norm": 0.09272906884038755, | |
| "learning_rate": 1.4903598426503241e-06, | |
| "loss": 0.0305, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.858064516129032, | |
| "grad_norm": 0.1688562089326058, | |
| "learning_rate": 1.3638696597277679e-06, | |
| "loss": 0.0402, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.864516129032258, | |
| "grad_norm": 0.11628209549950047, | |
| "learning_rate": 1.2429510926314836e-06, | |
| "loss": 0.0309, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.870967741935484, | |
| "grad_norm": 0.14018105443781761, | |
| "learning_rate": 1.1276109716355287e-06, | |
| "loss": 0.0466, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.8774193548387097, | |
| "grad_norm": 0.12354234788520546, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 0.0315, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.8838709677419354, | |
| "grad_norm": 0.15472531322652747, | |
| "learning_rate": 9.136918131366412e-07, | |
| "loss": 0.0436, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.8903225806451616, | |
| "grad_norm": 0.16727260037004013, | |
| "learning_rate": 8.151248591913518e-07, | |
| "loss": 0.0413, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.896774193548387, | |
| "grad_norm": 0.15358557638143366, | |
| "learning_rate": 7.221605177791691e-07, | |
| "loss": 0.0477, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.903225806451613, | |
| "grad_norm": 0.11899732193116695, | |
| "learning_rate": 6.348040401360833e-07, | |
| "loss": 0.0337, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.9096774193548387, | |
| "grad_norm": 0.19561410636845064, | |
| "learning_rate": 5.530603607290851e-07, | |
| "loss": 0.0556, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.9161290322580644, | |
| "grad_norm": 0.14981745610583072, | |
| "learning_rate": 4.76934096977566e-07, | |
| "loss": 0.0372, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.9225806451612906, | |
| "grad_norm": 0.1039480982798509, | |
| "learning_rate": 4.0642954899238197e-07, | |
| "loss": 0.0343, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.9290322580645163, | |
| "grad_norm": 0.11333580055218463, | |
| "learning_rate": 3.415506993330153e-07, | |
| "loss": 0.0388, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.935483870967742, | |
| "grad_norm": 0.10481858385197364, | |
| "learning_rate": 2.8230121278257637e-07, | |
| "loss": 0.0313, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.9419354838709677, | |
| "grad_norm": 0.15064675470229316, | |
| "learning_rate": 2.2868443614082469e-07, | |
| "loss": 0.0449, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.9483870967741934, | |
| "grad_norm": 0.13918413557689208, | |
| "learning_rate": 1.8070339803509807e-07, | |
| "loss": 0.0414, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.9548387096774196, | |
| "grad_norm": 0.15721864030858848, | |
| "learning_rate": 1.3836080874926049e-07, | |
| "loss": 0.0342, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.9612903225806453, | |
| "grad_norm": 0.15411456561887094, | |
| "learning_rate": 1.0165906007056914e-07, | |
| "loss": 0.0418, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.967741935483871, | |
| "grad_norm": 0.1513944041768792, | |
| "learning_rate": 7.060022515460451e-08, | |
| "loss": 0.0364, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.9741935483870967, | |
| "grad_norm": 0.22512284396579382, | |
| "learning_rate": 4.518605840815315e-08, | |
| "loss": 0.0447, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.9806451612903224, | |
| "grad_norm": 0.1529295717768871, | |
| "learning_rate": 2.5417995390086824e-08, | |
| "loss": 0.0449, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.9870967741935486, | |
| "grad_norm": 0.13942712092352366, | |
| "learning_rate": 1.129715273033849e-08, | |
| "loss": 0.0367, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.9935483870967743, | |
| "grad_norm": 0.13569899513050301, | |
| "learning_rate": 2.824328066730608e-09, | |
| "loss": 0.0398, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.11004047385534296, | |
| "learning_rate": 0.0, | |
| "loss": 0.0305, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.2107405662536621, | |
| "eval_runtime": 25.5157, | |
| "eval_samples_per_second": 5.134, | |
| "eval_steps_per_second": 0.666, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 465, | |
| "total_flos": 261374226563072.0, | |
| "train_loss": 0.09038178783751304, | |
| "train_runtime": 3542.1658, | |
| "train_samples_per_second": 2.097, | |
| "train_steps_per_second": 0.131 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 465, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 261374226563072.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |