| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 100, | |
| "global_step": 784, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002551020408163265, | |
| "grad_norm": 1.0885527168794946, | |
| "learning_rate": 9.999959857256764e-06, | |
| "loss": 0.1967, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00510204081632653, | |
| "grad_norm": 1.3738268211294016, | |
| "learning_rate": 9.999839429671632e-06, | |
| "loss": 0.2404, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007653061224489796, | |
| "grad_norm": 1.2030714088342824, | |
| "learning_rate": 9.99963871917832e-06, | |
| "loss": 0.2047, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01020408163265306, | |
| "grad_norm": 0.9283104826448032, | |
| "learning_rate": 9.999357728999657e-06, | |
| "loss": 0.1697, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012755102040816327, | |
| "grad_norm": 0.8955433527027652, | |
| "learning_rate": 9.99899646364753e-06, | |
| "loss": 0.2107, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.015306122448979591, | |
| "grad_norm": 0.839499734603853, | |
| "learning_rate": 9.99855492892281e-06, | |
| "loss": 0.1967, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 0.7667610964068909, | |
| "learning_rate": 9.998033131915266e-06, | |
| "loss": 0.1599, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02040816326530612, | |
| "grad_norm": 0.7717571071599194, | |
| "learning_rate": 9.99743108100344e-06, | |
| "loss": 0.1708, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02295918367346939, | |
| "grad_norm": 0.8407497770261014, | |
| "learning_rate": 9.996748785854524e-06, | |
| "loss": 0.1666, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.025510204081632654, | |
| "grad_norm": 0.8443185283822576, | |
| "learning_rate": 9.9959862574242e-06, | |
| "loss": 0.1631, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02806122448979592, | |
| "grad_norm": 0.7685166109923904, | |
| "learning_rate": 9.995143507956456e-06, | |
| "loss": 0.1821, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.030612244897959183, | |
| "grad_norm": 0.8594670816954412, | |
| "learning_rate": 9.994220550983404e-06, | |
| "loss": 0.1814, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03316326530612245, | |
| "grad_norm": 0.7096373833961109, | |
| "learning_rate": 9.993217401325057e-06, | |
| "loss": 0.1743, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 0.7098228947181078, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.1639, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03826530612244898, | |
| "grad_norm": 0.7563234672158685, | |
| "learning_rate": 9.99097058967056e-06, | |
| "loss": 0.1746, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 0.7772366871916015, | |
| "learning_rate": 9.989726963751683e-06, | |
| "loss": 0.1894, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04336734693877551, | |
| "grad_norm": 0.6199203281059847, | |
| "learning_rate": 9.988403217301477e-06, | |
| "loss": 0.1636, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04591836734693878, | |
| "grad_norm": 0.6852105358993172, | |
| "learning_rate": 9.986999371575465e-06, | |
| "loss": 0.1643, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04846938775510204, | |
| "grad_norm": 0.7736501072385444, | |
| "learning_rate": 9.985515449115338e-06, | |
| "loss": 0.1729, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05102040816326531, | |
| "grad_norm": 0.7089838632803512, | |
| "learning_rate": 9.983951473748579e-06, | |
| "loss": 0.1764, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 0.7063856891605839, | |
| "learning_rate": 9.982307470588097e-06, | |
| "loss": 0.1733, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05612244897959184, | |
| "grad_norm": 0.6555043059144575, | |
| "learning_rate": 9.980583466031808e-06, | |
| "loss": 0.1614, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.058673469387755105, | |
| "grad_norm": 0.689488625298009, | |
| "learning_rate": 9.978779487762221e-06, | |
| "loss": 0.1767, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.061224489795918366, | |
| "grad_norm": 0.6773533739024657, | |
| "learning_rate": 9.976895564745993e-06, | |
| "loss": 0.154, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06377551020408163, | |
| "grad_norm": 0.7140495522183441, | |
| "learning_rate": 9.974931727233454e-06, | |
| "loss": 0.1815, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0663265306122449, | |
| "grad_norm": 0.6802708469583214, | |
| "learning_rate": 9.97288800675814e-06, | |
| "loss": 0.1734, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06887755102040816, | |
| "grad_norm": 0.6558035514810944, | |
| "learning_rate": 9.970764436136266e-06, | |
| "loss": 0.1607, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 0.6255417776467211, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.1573, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07397959183673469, | |
| "grad_norm": 0.6604445618965782, | |
| "learning_rate": 9.966277882127975e-06, | |
| "loss": 0.163, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07653061224489796, | |
| "grad_norm": 0.8657610543676655, | |
| "learning_rate": 9.963914970782594e-06, | |
| "loss": 0.2162, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07908163265306123, | |
| "grad_norm": 0.683836315319945, | |
| "learning_rate": 9.961472353371564e-06, | |
| "loss": 0.186, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 0.6923067383095874, | |
| "learning_rate": 9.95895006911623e-06, | |
| "loss": 0.1994, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08418367346938775, | |
| "grad_norm": 0.6564772259446939, | |
| "learning_rate": 9.956348158517162e-06, | |
| "loss": 0.155, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08673469387755102, | |
| "grad_norm": 0.7093161053652716, | |
| "learning_rate": 9.953666663353485e-06, | |
| "loss": 0.1755, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 0.835449363527211, | |
| "learning_rate": 9.950905626682229e-06, | |
| "loss": 0.2058, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09183673469387756, | |
| "grad_norm": 0.6276558861634733, | |
| "learning_rate": 9.948065092837631e-06, | |
| "loss": 0.1499, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09438775510204081, | |
| "grad_norm": 0.7029471015624225, | |
| "learning_rate": 9.945145107430416e-06, | |
| "loss": 0.1852, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09693877551020408, | |
| "grad_norm": 0.6194669679342504, | |
| "learning_rate": 9.942145717347077e-06, | |
| "loss": 0.1652, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09948979591836735, | |
| "grad_norm": 0.6471970281746303, | |
| "learning_rate": 9.939066970749112e-06, | |
| "loss": 0.1614, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 0.7172164141492537, | |
| "learning_rate": 9.935908917072253e-06, | |
| "loss": 0.1933, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10459183673469388, | |
| "grad_norm": 0.6810647853743311, | |
| "learning_rate": 9.932671607025673e-06, | |
| "loss": 0.1777, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 0.6124857370200141, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.1471, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1096938775510204, | |
| "grad_norm": 0.6506073417099824, | |
| "learning_rate": 9.925959427022366e-06, | |
| "loss": 0.1791, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11224489795918367, | |
| "grad_norm": 0.7867658827361405, | |
| "learning_rate": 9.922484664843763e-06, | |
| "loss": 0.1924, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11479591836734694, | |
| "grad_norm": 0.6579496186128595, | |
| "learning_rate": 9.918930861849965e-06, | |
| "loss": 0.1825, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11734693877551021, | |
| "grad_norm": 0.6295603187995811, | |
| "learning_rate": 9.915298075104735e-06, | |
| "loss": 0.1748, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.11989795918367346, | |
| "grad_norm": 0.6001949141128546, | |
| "learning_rate": 9.911586362940082e-06, | |
| "loss": 0.1625, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 0.638398314084115, | |
| "learning_rate": 9.907795784955327e-06, | |
| "loss": 0.1517, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.6399210795033582, | |
| "learning_rate": 9.903926402016153e-06, | |
| "loss": 0.153, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12755102040816327, | |
| "grad_norm": 0.7153964945511394, | |
| "learning_rate": 9.899978276253617e-06, | |
| "loss": 0.1715, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13010204081632654, | |
| "grad_norm": 0.6490936443465146, | |
| "learning_rate": 9.895951471063156e-06, | |
| "loss": 0.1783, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1326530612244898, | |
| "grad_norm": 0.6495112832522, | |
| "learning_rate": 9.891846051103578e-06, | |
| "loss": 0.1746, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.13520408163265307, | |
| "grad_norm": 0.756147962576476, | |
| "learning_rate": 9.887662082296006e-06, | |
| "loss": 0.1973, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.1377551020408163, | |
| "grad_norm": 0.6587202327840592, | |
| "learning_rate": 9.883399631822836e-06, | |
| "loss": 0.1624, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14030612244897958, | |
| "grad_norm": 0.7693655467061048, | |
| "learning_rate": 9.87905876812665e-06, | |
| "loss": 0.2131, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 0.7303583152250236, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.1953, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.14540816326530612, | |
| "grad_norm": 0.6899593083786243, | |
| "learning_rate": 9.870142081129883e-06, | |
| "loss": 0.1684, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.14795918367346939, | |
| "grad_norm": 0.693858594186068, | |
| "learning_rate": 9.86556640100541e-06, | |
| "loss": 0.1923, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.15051020408163265, | |
| "grad_norm": 0.665392217093543, | |
| "learning_rate": 9.860912594007847e-06, | |
| "loss": 0.1796, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.15306122448979592, | |
| "grad_norm": 0.6903415551256747, | |
| "learning_rate": 9.85618073486382e-06, | |
| "loss": 0.1856, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1556122448979592, | |
| "grad_norm": 0.6652802573305568, | |
| "learning_rate": 9.851370899553254e-06, | |
| "loss": 0.1713, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.15816326530612246, | |
| "grad_norm": 0.6650490539593145, | |
| "learning_rate": 9.846483165308142e-06, | |
| "loss": 0.172, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 0.6483494869026171, | |
| "learning_rate": 9.841517610611309e-06, | |
| "loss": 0.1665, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 0.7366504472267554, | |
| "learning_rate": 9.836474315195148e-06, | |
| "loss": 0.2066, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.16581632653061223, | |
| "grad_norm": 0.7433091492751206, | |
| "learning_rate": 9.831353360040347e-06, | |
| "loss": 0.1822, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1683673469387755, | |
| "grad_norm": 0.5733277219690146, | |
| "learning_rate": 9.826154827374578e-06, | |
| "loss": 0.1398, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.17091836734693877, | |
| "grad_norm": 0.6510749686231305, | |
| "learning_rate": 9.82087880067119e-06, | |
| "loss": 0.1683, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.17346938775510204, | |
| "grad_norm": 0.6369611187446915, | |
| "learning_rate": 9.815525364647853e-06, | |
| "loss": 0.1698, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1760204081632653, | |
| "grad_norm": 0.6249985108358178, | |
| "learning_rate": 9.810094605265212e-06, | |
| "loss": 0.1598, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 0.7226036716768888, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.191, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18112244897959184, | |
| "grad_norm": 0.6414207048608993, | |
| "learning_rate": 9.799001466471134e-06, | |
| "loss": 0.1874, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1836734693877551, | |
| "grad_norm": 0.7694659731472804, | |
| "learning_rate": 9.793339265183303e-06, | |
| "loss": 0.2056, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.18622448979591838, | |
| "grad_norm": 0.6898939009314069, | |
| "learning_rate": 9.787600096780529e-06, | |
| "loss": 0.1793, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.18877551020408162, | |
| "grad_norm": 0.6117605046259157, | |
| "learning_rate": 9.781784053417192e-06, | |
| "loss": 0.1452, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1913265306122449, | |
| "grad_norm": 0.7207508689727239, | |
| "learning_rate": 9.775891228482068e-06, | |
| "loss": 0.1862, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19387755102040816, | |
| "grad_norm": 0.6195808967801664, | |
| "learning_rate": 9.76992171659682e-06, | |
| "loss": 0.1538, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 0.6630826534576278, | |
| "learning_rate": 9.763875613614482e-06, | |
| "loss": 0.1648, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1989795918367347, | |
| "grad_norm": 0.702756061791453, | |
| "learning_rate": 9.757753016617917e-06, | |
| "loss": 0.1884, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.20153061224489796, | |
| "grad_norm": 0.597140282770509, | |
| "learning_rate": 9.751554023918261e-06, | |
| "loss": 0.1509, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 0.7398956578868144, | |
| "learning_rate": 9.745278735053345e-06, | |
| "loss": 0.1893, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2066326530612245, | |
| "grad_norm": 0.6596462460468865, | |
| "learning_rate": 9.738927250786088e-06, | |
| "loss": 0.1704, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.20918367346938777, | |
| "grad_norm": 0.6869452338425555, | |
| "learning_rate": 9.732499673102895e-06, | |
| "loss": 0.1892, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.21173469387755103, | |
| "grad_norm": 0.6390745720379519, | |
| "learning_rate": 9.725996105212002e-06, | |
| "loss": 0.1781, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 0.6284608424012191, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.1643, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.21683673469387754, | |
| "grad_norm": 0.6422070407982153, | |
| "learning_rate": 9.712761417739326e-06, | |
| "loss": 0.1832, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2193877551020408, | |
| "grad_norm": 0.6950581393056733, | |
| "learning_rate": 9.706030510668202e-06, | |
| "loss": 0.1794, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.22193877551020408, | |
| "grad_norm": 0.604792640333442, | |
| "learning_rate": 9.6992240384073e-06, | |
| "loss": 0.1716, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.22448979591836735, | |
| "grad_norm": 0.7157986039856136, | |
| "learning_rate": 9.692342110248802e-06, | |
| "loss": 0.2274, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.22704081632653061, | |
| "grad_norm": 0.6239441335723747, | |
| "learning_rate": 9.685384836696503e-06, | |
| "loss": 0.1747, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.22959183673469388, | |
| "grad_norm": 0.6341863503525225, | |
| "learning_rate": 9.678352329464018e-06, | |
| "loss": 0.1771, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 0.6494114016675228, | |
| "learning_rate": 9.671244701472999e-06, | |
| "loss": 0.1769, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.23469387755102042, | |
| "grad_norm": 0.6473038344264327, | |
| "learning_rate": 9.664062066851325e-06, | |
| "loss": 0.1714, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2372448979591837, | |
| "grad_norm": 0.6552409387122857, | |
| "learning_rate": 9.656804540931256e-06, | |
| "loss": 0.1684, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.23979591836734693, | |
| "grad_norm": 0.7059100317343475, | |
| "learning_rate": 9.649472240247588e-06, | |
| "loss": 0.2126, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2423469387755102, | |
| "grad_norm": 0.7101604206793758, | |
| "learning_rate": 9.642065282535795e-06, | |
| "loss": 0.1786, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 0.6055224147282258, | |
| "learning_rate": 9.63458378673011e-06, | |
| "loss": 0.1487, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.24744897959183673, | |
| "grad_norm": 0.6630323155965017, | |
| "learning_rate": 9.627027872961642e-06, | |
| "loss": 0.1802, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.6943625213974632, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.1662, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.25255102040816324, | |
| "grad_norm": 0.6707618830728986, | |
| "learning_rate": 9.611693278033517e-06, | |
| "loss": 0.1722, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "grad_norm": 0.63430666549655, | |
| "learning_rate": 9.603914843102941e-06, | |
| "loss": 0.1755, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "eval_loss": 0.17825095355510712, | |
| "eval_runtime": 2.5188, | |
| "eval_samples_per_second": 12.704, | |
| "eval_steps_per_second": 3.176, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2576530612244898, | |
| "grad_norm": 0.6301792150542433, | |
| "learning_rate": 9.596062482663796e-06, | |
| "loss": 0.1647, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2602040816326531, | |
| "grad_norm": 0.7439256267733433, | |
| "learning_rate": 9.588136322802194e-06, | |
| "loss": 0.1843, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2627551020408163, | |
| "grad_norm": 0.6241120007830788, | |
| "learning_rate": 9.580136490789255e-06, | |
| "loss": 0.1774, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "grad_norm": 0.6589666329393432, | |
| "learning_rate": 9.572063115079063e-06, | |
| "loss": 0.1685, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 0.628815378883589, | |
| "learning_rate": 9.563916325306595e-06, | |
| "loss": 0.1595, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.27040816326530615, | |
| "grad_norm": 0.7047293090329506, | |
| "learning_rate": 9.555696252285648e-06, | |
| "loss": 0.1952, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2729591836734694, | |
| "grad_norm": 0.6557715899844985, | |
| "learning_rate": 9.547403028006734e-06, | |
| "loss": 0.1661, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2755102040816326, | |
| "grad_norm": 0.7357553062387641, | |
| "learning_rate": 9.539036785634961e-06, | |
| "loss": 0.2147, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2780612244897959, | |
| "grad_norm": 0.604934655741894, | |
| "learning_rate": 9.530597659507898e-06, | |
| "loss": 0.1478, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.28061224489795916, | |
| "grad_norm": 0.6963912461652633, | |
| "learning_rate": 9.522085785133415e-06, | |
| "loss": 0.1732, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.28316326530612246, | |
| "grad_norm": 0.6913597853134738, | |
| "learning_rate": 9.513501299187506e-06, | |
| "loss": 0.1802, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.6061097854231792, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.1459, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.288265306122449, | |
| "grad_norm": 0.6784991397247085, | |
| "learning_rate": 9.49611504511283e-06, | |
| "loss": 0.1577, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.29081632653061223, | |
| "grad_norm": 0.6820299760172144, | |
| "learning_rate": 9.48731355615684e-06, | |
| "loss": 0.181, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.29336734693877553, | |
| "grad_norm": 0.6080242625892597, | |
| "learning_rate": 9.478440013970485e-06, | |
| "loss": 0.1438, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.29591836734693877, | |
| "grad_norm": 0.7274055237100493, | |
| "learning_rate": 9.469494561037097e-06, | |
| "loss": 0.1861, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.29846938775510207, | |
| "grad_norm": 0.681863192199185, | |
| "learning_rate": 9.46047734099469e-06, | |
| "loss": 0.1714, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3010204081632653, | |
| "grad_norm": 0.6493118569066717, | |
| "learning_rate": 9.451388498633635e-06, | |
| "loss": 0.1673, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 0.6750599274016734, | |
| "learning_rate": 9.442228179894362e-06, | |
| "loss": 0.1819, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 0.6385238805634963, | |
| "learning_rate": 9.432996531865001e-06, | |
| "loss": 0.1657, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3086734693877551, | |
| "grad_norm": 0.6134862321136166, | |
| "learning_rate": 9.423693702779022e-06, | |
| "loss": 0.1626, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3112244897959184, | |
| "grad_norm": 0.6711298307545004, | |
| "learning_rate": 9.414319842012855e-06, | |
| "loss": 0.1804, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3137755102040816, | |
| "grad_norm": 0.6460358850190799, | |
| "learning_rate": 9.404875100083497e-06, | |
| "loss": 0.1799, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3163265306122449, | |
| "grad_norm": 0.6229317835591002, | |
| "learning_rate": 9.395359628646087e-06, | |
| "loss": 0.1726, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.31887755102040816, | |
| "grad_norm": 0.7105507007735112, | |
| "learning_rate": 9.385773580491476e-06, | |
| "loss": 0.1773, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 0.652326436562633, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.1696, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3239795918367347, | |
| "grad_norm": 0.6846427621008662, | |
| "learning_rate": 9.366390370857863e-06, | |
| "loss": 0.1586, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 0.6988027763071323, | |
| "learning_rate": 9.356593520616948e-06, | |
| "loss": 0.1939, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.32908163265306123, | |
| "grad_norm": 0.5989815910486795, | |
| "learning_rate": 9.34672671613e-06, | |
| "loss": 0.1508, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.33163265306122447, | |
| "grad_norm": 0.6157873448382567, | |
| "learning_rate": 9.336790115829255e-06, | |
| "loss": 0.1652, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.33418367346938777, | |
| "grad_norm": 0.6104056434176193, | |
| "learning_rate": 9.326783879267678e-06, | |
| "loss": 0.1538, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.336734693877551, | |
| "grad_norm": 0.7468762943061447, | |
| "learning_rate": 9.316708167116377e-06, | |
| "loss": 0.2006, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 0.6338257023806732, | |
| "learning_rate": 9.306563141162046e-06, | |
| "loss": 0.1759, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.34183673469387754, | |
| "grad_norm": 0.623029871752068, | |
| "learning_rate": 9.296348964304351e-06, | |
| "loss": 0.1563, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.34438775510204084, | |
| "grad_norm": 0.616534494076029, | |
| "learning_rate": 9.286065800553327e-06, | |
| "loss": 0.1612, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3469387755102041, | |
| "grad_norm": 0.6567894577284148, | |
| "learning_rate": 9.275713815026732e-06, | |
| "loss": 0.1794, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3494897959183674, | |
| "grad_norm": 0.6929529884982644, | |
| "learning_rate": 9.265293173947404e-06, | |
| "loss": 0.183, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3520408163265306, | |
| "grad_norm": 0.5590498656811169, | |
| "learning_rate": 9.254804044640596e-06, | |
| "loss": 0.1383, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.35459183673469385, | |
| "grad_norm": 0.6259188867432768, | |
| "learning_rate": 9.244246595531272e-06, | |
| "loss": 0.15, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.6374491966087964, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.1724, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3596938775510204, | |
| "grad_norm": 0.6775149835839037, | |
| "learning_rate": 9.22292741708733e-06, | |
| "loss": 0.16, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3622448979591837, | |
| "grad_norm": 0.6507427637455202, | |
| "learning_rate": 9.212166030076832e-06, | |
| "loss": 0.1791, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3647959183673469, | |
| "grad_norm": 0.5787991775867446, | |
| "learning_rate": 9.20133700790657e-06, | |
| "loss": 0.1436, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3673469387755102, | |
| "grad_norm": 0.6577565969598809, | |
| "learning_rate": 9.190440524459203e-06, | |
| "loss": 0.1778, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.36989795918367346, | |
| "grad_norm": 0.63281714426237, | |
| "learning_rate": 9.179476754700632e-06, | |
| "loss": 0.1579, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.37244897959183676, | |
| "grad_norm": 0.620787271558136, | |
| "learning_rate": 9.168445874677168e-06, | |
| "loss": 0.1523, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.604356995426742, | |
| "learning_rate": 9.157348061512728e-06, | |
| "loss": 0.1556, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.37755102040816324, | |
| "grad_norm": 0.6941039323941053, | |
| "learning_rate": 9.146183493405976e-06, | |
| "loss": 0.2001, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.38010204081632654, | |
| "grad_norm": 0.6710751485897907, | |
| "learning_rate": 9.13495234962747e-06, | |
| "loss": 0.1997, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3826530612244898, | |
| "grad_norm": 0.6080309117589121, | |
| "learning_rate": 9.12365481051678e-06, | |
| "loss": 0.1665, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3852040816326531, | |
| "grad_norm": 0.621319557269394, | |
| "learning_rate": 9.112291057479586e-06, | |
| "loss": 0.1584, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3877551020408163, | |
| "grad_norm": 0.6304671673258588, | |
| "learning_rate": 9.10086127298478e-06, | |
| "loss": 0.1681, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3903061224489796, | |
| "grad_norm": 0.643668035685642, | |
| "learning_rate": 9.089365640561523e-06, | |
| "loss": 0.1652, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 0.6295795105784965, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.1632, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.39540816326530615, | |
| "grad_norm": 0.6212503718092719, | |
| "learning_rate": 9.066177571329969e-06, | |
| "loss": 0.1555, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3979591836734694, | |
| "grad_norm": 0.5942667214942609, | |
| "learning_rate": 9.054485506854756e-06, | |
| "loss": 0.1642, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4005102040816326, | |
| "grad_norm": 0.6468545759854039, | |
| "learning_rate": 9.04272833911128e-06, | |
| "loss": 0.1654, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.4030612244897959, | |
| "grad_norm": 0.6066009285810315, | |
| "learning_rate": 9.030906256885528e-06, | |
| "loss": 0.1448, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.40561224489795916, | |
| "grad_norm": 0.6277372050287073, | |
| "learning_rate": 9.019019450005826e-06, | |
| "loss": 0.1627, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 0.6086742021736391, | |
| "learning_rate": 9.007068109339783e-06, | |
| "loss": 0.1574, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 0.671251680756004, | |
| "learning_rate": 8.995052426791247e-06, | |
| "loss": 0.1912, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.413265306122449, | |
| "grad_norm": 0.6714069477587861, | |
| "learning_rate": 8.982972595297195e-06, | |
| "loss": 0.186, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.41581632653061223, | |
| "grad_norm": 0.642109869112752, | |
| "learning_rate": 8.970828808824659e-06, | |
| "loss": 0.1708, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.41836734693877553, | |
| "grad_norm": 0.6168564976222412, | |
| "learning_rate": 8.9586212623676e-06, | |
| "loss": 0.1669, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.42091836734693877, | |
| "grad_norm": 0.6695625235286308, | |
| "learning_rate": 8.946350151943778e-06, | |
| "loss": 0.1909, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.42346938775510207, | |
| "grad_norm": 0.6036159435553136, | |
| "learning_rate": 8.93401567459161e-06, | |
| "loss": 0.1687, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4260204081632653, | |
| "grad_norm": 0.6616845609083462, | |
| "learning_rate": 8.921618028366997e-06, | |
| "loss": 0.1684, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.6019985636648711, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.164, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.43112244897959184, | |
| "grad_norm": 0.6318313353599098, | |
| "learning_rate": 8.896634026592393e-06, | |
| "loss": 0.1707, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4336734693877551, | |
| "grad_norm": 0.6343404996339642, | |
| "learning_rate": 8.884048072212952e-06, | |
| "loss": 0.1552, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4362244897959184, | |
| "grad_norm": 0.6073430634182778, | |
| "learning_rate": 8.871399751295717e-06, | |
| "loss": 0.1523, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4387755102040816, | |
| "grad_norm": 0.684776429861673, | |
| "learning_rate": 8.85868926693601e-06, | |
| "loss": 0.1624, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4413265306122449, | |
| "grad_norm": 0.6020383554217804, | |
| "learning_rate": 8.845916823227315e-06, | |
| "loss": 0.1625, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.44387755102040816, | |
| "grad_norm": 0.6548300342133035, | |
| "learning_rate": 8.833082625258003e-06, | |
| "loss": 0.1562, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 0.6479894794435125, | |
| "learning_rate": 8.820186879108038e-06, | |
| "loss": 0.1649, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4489795918367347, | |
| "grad_norm": 0.6764426315561457, | |
| "learning_rate": 8.807229791845673e-06, | |
| "loss": 0.1633, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.45153061224489793, | |
| "grad_norm": 0.6239624169344059, | |
| "learning_rate": 8.794211571524119e-06, | |
| "loss": 0.1557, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.45408163265306123, | |
| "grad_norm": 0.6474988134364261, | |
| "learning_rate": 8.781132427178203e-06, | |
| "loss": 0.1818, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.45663265306122447, | |
| "grad_norm": 0.6938478733121906, | |
| "learning_rate": 8.767992568821021e-06, | |
| "loss": 0.1924, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.45918367346938777, | |
| "grad_norm": 0.6179549390370858, | |
| "learning_rate": 8.754792207440557e-06, | |
| "loss": 0.1544, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.461734693877551, | |
| "grad_norm": 0.6256602273247317, | |
| "learning_rate": 8.741531554996298e-06, | |
| "loss": 0.1627, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 0.6269721018978721, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.1502, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.46683673469387754, | |
| "grad_norm": 0.6746598533142751, | |
| "learning_rate": 8.714830229591418e-06, | |
| "loss": 0.1946, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.46938775510204084, | |
| "grad_norm": 0.7113907278859266, | |
| "learning_rate": 8.701389985376578e-06, | |
| "loss": 0.1984, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4719387755102041, | |
| "grad_norm": 0.6364152878079946, | |
| "learning_rate": 8.687890307582618e-06, | |
| "loss": 0.171, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4744897959183674, | |
| "grad_norm": 0.659485503945726, | |
| "learning_rate": 8.674331412975178e-06, | |
| "loss": 0.166, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4770408163265306, | |
| "grad_norm": 0.6273274956447809, | |
| "learning_rate": 8.660713519270749e-06, | |
| "loss": 0.1549, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.47959183673469385, | |
| "grad_norm": 0.6683355477651822, | |
| "learning_rate": 8.647036845133171e-06, | |
| "loss": 0.1715, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 0.7035769520046447, | |
| "learning_rate": 8.633301610170136e-06, | |
| "loss": 0.1676, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4846938775510204, | |
| "grad_norm": 0.665388776003314, | |
| "learning_rate": 8.619508034929646e-06, | |
| "loss": 0.1745, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4872448979591837, | |
| "grad_norm": 0.6263473288405155, | |
| "learning_rate": 8.60565634089648e-06, | |
| "loss": 0.168, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 0.6454061053639517, | |
| "learning_rate": 8.591746750488639e-06, | |
| "loss": 0.1793, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4923469387755102, | |
| "grad_norm": 0.595965931581441, | |
| "learning_rate": 8.577779487053767e-06, | |
| "loss": 0.1438, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.49489795918367346, | |
| "grad_norm": 0.585416860329295, | |
| "learning_rate": 8.563754774865574e-06, | |
| "loss": 0.152, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.49744897959183676, | |
| "grad_norm": 0.65139162533389, | |
| "learning_rate": 8.549672839120227e-06, | |
| "loss": 0.1721, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6021792322784167, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.167, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5025510204081632, | |
| "grad_norm": 0.6974220008021444, | |
| "learning_rate": 8.521338202333335e-06, | |
| "loss": 0.183, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5051020408163265, | |
| "grad_norm": 0.6710913460442544, | |
| "learning_rate": 8.507085956263808e-06, | |
| "loss": 0.19, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5076530612244898, | |
| "grad_norm": 0.6602730300692774, | |
| "learning_rate": 8.492777396573863e-06, | |
| "loss": 0.1822, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 0.6096784561422763, | |
| "learning_rate": 8.478412753017433e-06, | |
| "loss": 0.1688, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "eval_loss": 0.17530770599842072, | |
| "eval_runtime": 2.3928, | |
| "eval_samples_per_second": 13.374, | |
| "eval_steps_per_second": 3.343, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5127551020408163, | |
| "grad_norm": 0.5777967437570939, | |
| "learning_rate": 8.463992256248997e-06, | |
| "loss": 0.1515, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5153061224489796, | |
| "grad_norm": 0.6469736667234415, | |
| "learning_rate": 8.449516137819875e-06, | |
| "loss": 0.1716, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 0.6042565412596849, | |
| "learning_rate": 8.43498463017451e-06, | |
| "loss": 0.1498, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.5204081632653061, | |
| "grad_norm": 0.6453530466355405, | |
| "learning_rate": 8.420397966646732e-06, | |
| "loss": 0.1716, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5229591836734694, | |
| "grad_norm": 0.6434693539199635, | |
| "learning_rate": 8.405756381456017e-06, | |
| "loss": 0.1812, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5255102040816326, | |
| "grad_norm": 0.7389376263189975, | |
| "learning_rate": 8.391060109703725e-06, | |
| "loss": 0.1781, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5280612244897959, | |
| "grad_norm": 0.6155812259074535, | |
| "learning_rate": 8.376309387369318e-06, | |
| "loss": 0.1649, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5306122448979592, | |
| "grad_norm": 0.650794631172212, | |
| "learning_rate": 8.361504451306585e-06, | |
| "loss": 0.175, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5331632653061225, | |
| "grad_norm": 0.6309566465508609, | |
| "learning_rate": 8.34664553923982e-06, | |
| "loss": 0.1722, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.6153722780738814, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.151, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5382653061224489, | |
| "grad_norm": 0.6213867045748396, | |
| "learning_rate": 8.316766742321052e-06, | |
| "loss": 0.1709, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5408163265306123, | |
| "grad_norm": 0.6231378594046826, | |
| "learning_rate": 8.301747337235798e-06, | |
| "loss": 0.1726, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5433673469387755, | |
| "grad_norm": 0.6347367176192658, | |
| "learning_rate": 8.286674915672308e-06, | |
| "loss": 0.1675, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5459183673469388, | |
| "grad_norm": 0.6293391795307883, | |
| "learning_rate": 8.271549719649923e-06, | |
| "loss": 0.16, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.548469387755102, | |
| "grad_norm": 0.6270775654657718, | |
| "learning_rate": 8.256371992035384e-06, | |
| "loss": 0.1679, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5510204081632653, | |
| "grad_norm": 0.5994977548727174, | |
| "learning_rate": 8.241141976538944e-06, | |
| "loss": 0.1321, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 0.6676987808194949, | |
| "learning_rate": 8.22585991771044e-06, | |
| "loss": 0.1678, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5561224489795918, | |
| "grad_norm": 0.6632408731790466, | |
| "learning_rate": 8.210526060935377e-06, | |
| "loss": 0.1797, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5586734693877551, | |
| "grad_norm": 0.6700416358253908, | |
| "learning_rate": 8.19514065243099e-06, | |
| "loss": 0.1862, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5612244897959183, | |
| "grad_norm": 0.6307204362512712, | |
| "learning_rate": 8.179703939242276e-06, | |
| "loss": 0.1737, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5637755102040817, | |
| "grad_norm": 0.6441334257769658, | |
| "learning_rate": 8.164216169238043e-06, | |
| "loss": 0.1599, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5663265306122449, | |
| "grad_norm": 0.5816475384619316, | |
| "learning_rate": 8.148677591106919e-06, | |
| "loss": 0.1509, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5688775510204082, | |
| "grad_norm": 0.6348006825878267, | |
| "learning_rate": 8.133088454353366e-06, | |
| "loss": 0.1697, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.6582403387620182, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.1797, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5739795918367347, | |
| "grad_norm": 0.6270714690286511, | |
| "learning_rate": 8.101759507051919e-06, | |
| "loss": 0.1625, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.576530612244898, | |
| "grad_norm": 0.6652022735952716, | |
| "learning_rate": 8.08602019955598e-06, | |
| "loss": 0.1622, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5790816326530612, | |
| "grad_norm": 0.693450200019601, | |
| "learning_rate": 8.070231339533442e-06, | |
| "loss": 0.188, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5816326530612245, | |
| "grad_norm": 0.611727450765603, | |
| "learning_rate": 8.054393180507572e-06, | |
| "loss": 0.1587, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5841836734693877, | |
| "grad_norm": 0.592663870640422, | |
| "learning_rate": 8.038505976793226e-06, | |
| "loss": 0.1452, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5867346938775511, | |
| "grad_norm": 0.6302845415061488, | |
| "learning_rate": 8.022569983492781e-06, | |
| "loss": 0.1672, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 0.5804051065068835, | |
| "learning_rate": 8.00658545649203e-06, | |
| "loss": 0.1471, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5918367346938775, | |
| "grad_norm": 0.6273355896398363, | |
| "learning_rate": 7.99055265245608e-06, | |
| "loss": 0.1721, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5943877551020408, | |
| "grad_norm": 0.65405300250257, | |
| "learning_rate": 7.974471828825227e-06, | |
| "loss": 0.1698, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5969387755102041, | |
| "grad_norm": 0.611434007618611, | |
| "learning_rate": 7.958343243810818e-06, | |
| "loss": 0.1507, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5994897959183674, | |
| "grad_norm": 0.6897532176595209, | |
| "learning_rate": 7.942167156391111e-06, | |
| "loss": 0.1768, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6020408163265306, | |
| "grad_norm": 0.6147188964397875, | |
| "learning_rate": 7.925943826307119e-06, | |
| "loss": 0.1502, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6045918367346939, | |
| "grad_norm": 0.6057461789141037, | |
| "learning_rate": 7.909673514058428e-06, | |
| "loss": 0.1538, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 0.5842486053218805, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.1473, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6096938775510204, | |
| "grad_norm": 0.6124541145552258, | |
| "learning_rate": 7.876992988833107e-06, | |
| "loss": 0.1713, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 0.6918344738184035, | |
| "learning_rate": 7.860583300610849e-06, | |
| "loss": 0.1741, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6147959183673469, | |
| "grad_norm": 0.5869529235754722, | |
| "learning_rate": 7.844127679724213e-06, | |
| "loss": 0.1643, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6173469387755102, | |
| "grad_norm": 0.5917786409568101, | |
| "learning_rate": 7.827626390402707e-06, | |
| "loss": 0.1323, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6198979591836735, | |
| "grad_norm": 0.649081051443825, | |
| "learning_rate": 7.811079697609137e-06, | |
| "loss": 0.1642, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6224489795918368, | |
| "grad_norm": 0.6068854372103691, | |
| "learning_rate": 7.794487867035358e-06, | |
| "loss": 0.1497, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.6921605466960177, | |
| "learning_rate": 7.777851165098012e-06, | |
| "loss": 0.1685, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6275510204081632, | |
| "grad_norm": 0.6336841548041962, | |
| "learning_rate": 7.761169858934238e-06, | |
| "loss": 0.1741, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6301020408163265, | |
| "grad_norm": 0.7562805886149276, | |
| "learning_rate": 7.744444216397393e-06, | |
| "loss": 0.202, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6326530612244898, | |
| "grad_norm": 0.6315316001771077, | |
| "learning_rate": 7.727674506052744e-06, | |
| "loss": 0.1571, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6352040816326531, | |
| "grad_norm": 0.6322335288478561, | |
| "learning_rate": 7.710860997173164e-06, | |
| "loss": 0.1648, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6377551020408163, | |
| "grad_norm": 0.6990439392046193, | |
| "learning_rate": 7.694003959734802e-06, | |
| "loss": 0.1621, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6403061224489796, | |
| "grad_norm": 0.5892549294977432, | |
| "learning_rate": 7.677103664412747e-06, | |
| "loss": 0.1666, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.7152881541118469, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.1912, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6454081632653061, | |
| "grad_norm": 0.6123697518954341, | |
| "learning_rate": 7.643174386286539e-06, | |
| "loss": 0.1705, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6479591836734694, | |
| "grad_norm": 0.6404458075559971, | |
| "learning_rate": 7.626145948288107e-06, | |
| "loss": 0.1445, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6505102040816326, | |
| "grad_norm": 0.5989704014552829, | |
| "learning_rate": 7.609075342008674e-06, | |
| "loss": 0.1518, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 0.6166584428376881, | |
| "learning_rate": 7.591962841552627e-06, | |
| "loss": 0.1646, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6556122448979592, | |
| "grad_norm": 0.6023893275125043, | |
| "learning_rate": 7.574808721697047e-06, | |
| "loss": 0.1594, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6581632653061225, | |
| "grad_norm": 0.5980117337714704, | |
| "learning_rate": 7.55761325788731e-06, | |
| "loss": 0.1629, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 0.5957147716396232, | |
| "learning_rate": 7.540376726232648e-06, | |
| "loss": 0.1562, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6632653061224489, | |
| "grad_norm": 0.6352476407461065, | |
| "learning_rate": 7.52309940350173e-06, | |
| "loss": 0.1611, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6658163265306123, | |
| "grad_norm": 0.6705656721875611, | |
| "learning_rate": 7.505781567118204e-06, | |
| "loss": 0.1717, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6683673469387755, | |
| "grad_norm": 0.6416780470208198, | |
| "learning_rate": 7.488423495156258e-06, | |
| "loss": 0.1765, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6709183673469388, | |
| "grad_norm": 0.6396852156623875, | |
| "learning_rate": 7.471025466336139e-06, | |
| "loss": 0.1608, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.673469387755102, | |
| "grad_norm": 0.6315942638591941, | |
| "learning_rate": 7.453587760019691e-06, | |
| "loss": 0.1608, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6760204081632653, | |
| "grad_norm": 0.6445458308180846, | |
| "learning_rate": 7.436110656205859e-06, | |
| "loss": 0.1623, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 0.6064991852375194, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.1557, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6811224489795918, | |
| "grad_norm": 0.7041372391725068, | |
| "learning_rate": 7.401039379240373e-06, | |
| "loss": 0.1889, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6836734693877551, | |
| "grad_norm": 0.652310000895515, | |
| "learning_rate": 7.383445769231628e-06, | |
| "loss": 0.1842, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6862244897959183, | |
| "grad_norm": 0.5766488734278856, | |
| "learning_rate": 7.365813888002269e-06, | |
| "loss": 0.1515, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6887755102040817, | |
| "grad_norm": 0.6210407194072641, | |
| "learning_rate": 7.348144018669129e-06, | |
| "loss": 0.1566, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6913265306122449, | |
| "grad_norm": 0.5587746621446245, | |
| "learning_rate": 7.3304364449590215e-06, | |
| "loss": 0.14, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6938775510204082, | |
| "grad_norm": 0.6355280238362423, | |
| "learning_rate": 7.312691451204178e-06, | |
| "loss": 0.1726, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 0.642365945745617, | |
| "learning_rate": 7.294909322337689e-06, | |
| "loss": 0.1833, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6989795918367347, | |
| "grad_norm": 0.6283138953382016, | |
| "learning_rate": 7.277090343888931e-06, | |
| "loss": 0.1688, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.701530612244898, | |
| "grad_norm": 0.6011352110745912, | |
| "learning_rate": 7.259234801978971e-06, | |
| "loss": 0.1574, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7040816326530612, | |
| "grad_norm": 0.6354286801860528, | |
| "learning_rate": 7.241342983315985e-06, | |
| "loss": 0.1739, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7066326530612245, | |
| "grad_norm": 0.6106322866032395, | |
| "learning_rate": 7.223415175190647e-06, | |
| "loss": 0.1707, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7091836734693877, | |
| "grad_norm": 0.7159138748192706, | |
| "learning_rate": 7.205451665471515e-06, | |
| "loss": 0.2022, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7117346938775511, | |
| "grad_norm": 0.5823787309582513, | |
| "learning_rate": 7.187452742600409e-06, | |
| "loss": 0.1568, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.6151783594340412, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.1635, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7168367346938775, | |
| "grad_norm": 0.6130620738320395, | |
| "learning_rate": 7.151349814008104e-06, | |
| "loss": 0.1327, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7193877551020408, | |
| "grad_norm": 0.6748859373806838, | |
| "learning_rate": 7.1332463879951404e-06, | |
| "loss": 0.1811, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7219387755102041, | |
| "grad_norm": 0.5834476885398581, | |
| "learning_rate": 7.1151087082373705e-06, | |
| "loss": 0.1478, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7244897959183674, | |
| "grad_norm": 0.6169796809060963, | |
| "learning_rate": 7.096937065973285e-06, | |
| "loss": 0.1631, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7270408163265306, | |
| "grad_norm": 0.6959780203274529, | |
| "learning_rate": 7.078731752986709e-06, | |
| "loss": 0.1616, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7295918367346939, | |
| "grad_norm": 0.6979567612851677, | |
| "learning_rate": 7.060493061602128e-06, | |
| "loss": 0.187, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 0.5732766031573223, | |
| "learning_rate": 7.042221284679982e-06, | |
| "loss": 0.1529, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7346938775510204, | |
| "grad_norm": 0.6156530993396518, | |
| "learning_rate": 7.023916715611969e-06, | |
| "loss": 0.1498, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7372448979591837, | |
| "grad_norm": 0.7132345983065509, | |
| "learning_rate": 7.00557964831634e-06, | |
| "loss": 0.1617, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7397959183673469, | |
| "grad_norm": 0.6232249158259106, | |
| "learning_rate": 6.987210377233165e-06, | |
| "loss": 0.149, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7423469387755102, | |
| "grad_norm": 0.7277934128266045, | |
| "learning_rate": 6.968809197319618e-06, | |
| "loss": 0.1864, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7448979591836735, | |
| "grad_norm": 0.7164985831395658, | |
| "learning_rate": 6.950376404045235e-06, | |
| "loss": 0.1664, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7474489795918368, | |
| "grad_norm": 0.6722545299275654, | |
| "learning_rate": 6.931912293387172e-06, | |
| "loss": 0.1808, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.6442507964460605, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.1583, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7525510204081632, | |
| "grad_norm": 0.645133746906539, | |
| "learning_rate": 6.894891306338195e-06, | |
| "loss": 0.1696, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7551020408163265, | |
| "grad_norm": 0.5410055981651295, | |
| "learning_rate": 6.876335024396872e-06, | |
| "loss": 0.1293, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7576530612244898, | |
| "grad_norm": 0.688793218054993, | |
| "learning_rate": 6.857748613961503e-06, | |
| "loss": 0.1906, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7602040816326531, | |
| "grad_norm": 0.6133050542728484, | |
| "learning_rate": 6.839132373475894e-06, | |
| "loss": 0.1592, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7627551020408163, | |
| "grad_norm": 0.6599036669431815, | |
| "learning_rate": 6.820486601862826e-06, | |
| "loss": 0.1783, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7653061224489796, | |
| "grad_norm": 0.6391559096522518, | |
| "learning_rate": 6.801811598519268e-06, | |
| "loss": 0.1684, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7653061224489796, | |
| "eval_loss": 0.1719779670238495, | |
| "eval_runtime": 2.1687, | |
| "eval_samples_per_second": 14.755, | |
| "eval_steps_per_second": 3.689, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 0.6092171680985036, | |
| "learning_rate": 6.783107663311566e-06, | |
| "loss": 0.1599, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.7704081632653061, | |
| "grad_norm": 0.6151859069141618, | |
| "learning_rate": 6.764375096570628e-06, | |
| "loss": 0.1577, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7729591836734694, | |
| "grad_norm": 0.6557762414524845, | |
| "learning_rate": 6.7456141990871e-06, | |
| "loss": 0.1877, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7755102040816326, | |
| "grad_norm": 0.6277047181654297, | |
| "learning_rate": 6.726825272106539e-06, | |
| "loss": 0.1623, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.7780612244897959, | |
| "grad_norm": 0.6430715807058038, | |
| "learning_rate": 6.708008617324573e-06, | |
| "loss": 0.1491, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7806122448979592, | |
| "grad_norm": 0.644153170652738, | |
| "learning_rate": 6.689164536882059e-06, | |
| "loss": 0.1791, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7831632653061225, | |
| "grad_norm": 0.6160033596879607, | |
| "learning_rate": 6.670293333360229e-06, | |
| "loss": 0.1427, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.5649228032258986, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.1362, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7882653061224489, | |
| "grad_norm": 0.6371148038872716, | |
| "learning_rate": 6.632470769576283e-06, | |
| "loss": 0.171, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7908163265306123, | |
| "grad_norm": 0.5975125963921615, | |
| "learning_rate": 6.6135200166347505e-06, | |
| "loss": 0.149, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7933673469387755, | |
| "grad_norm": 0.5408720399231929, | |
| "learning_rate": 6.594543355245324e-06, | |
| "loss": 0.1216, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7959183673469388, | |
| "grad_norm": 0.7069571267475502, | |
| "learning_rate": 6.575541090118105e-06, | |
| "loss": 0.1837, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.798469387755102, | |
| "grad_norm": 0.6603211346802758, | |
| "learning_rate": 6.556513526374307e-06, | |
| "loss": 0.1894, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8010204081632653, | |
| "grad_norm": 0.5956839299454775, | |
| "learning_rate": 6.537460969541378e-06, | |
| "loss": 0.148, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 0.6123273259294997, | |
| "learning_rate": 6.518383725548074e-06, | |
| "loss": 0.1495, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8061224489795918, | |
| "grad_norm": 0.6038963523853219, | |
| "learning_rate": 6.499282100719558e-06, | |
| "loss": 0.1528, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8086734693877551, | |
| "grad_norm": 0.6358364483640969, | |
| "learning_rate": 6.4801564017724804e-06, | |
| "loss": 0.1539, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.8112244897959183, | |
| "grad_norm": 0.6181985396343684, | |
| "learning_rate": 6.461006935810048e-06, | |
| "loss": 0.1548, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8137755102040817, | |
| "grad_norm": 0.6209023492344633, | |
| "learning_rate": 6.441834010317097e-06, | |
| "loss": 0.1424, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 0.6560248537666912, | |
| "learning_rate": 6.4226379331551625e-06, | |
| "loss": 0.1605, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8188775510204082, | |
| "grad_norm": 0.7033920181013527, | |
| "learning_rate": 6.40341901255752e-06, | |
| "loss": 0.1863, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 0.6117156078888957, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.1604, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8239795918367347, | |
| "grad_norm": 0.6141997763943361, | |
| "learning_rate": 6.364913875817267e-06, | |
| "loss": 0.1475, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.826530612244898, | |
| "grad_norm": 0.546093227403362, | |
| "learning_rate": 6.345628277955384e-06, | |
| "loss": 0.1285, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8290816326530612, | |
| "grad_norm": 0.6952190901409022, | |
| "learning_rate": 6.326321073209321e-06, | |
| "loss": 0.1736, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8316326530612245, | |
| "grad_norm": 0.6457827737823263, | |
| "learning_rate": 6.306992571596742e-06, | |
| "loss": 0.1822, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.8341836734693877, | |
| "grad_norm": 0.5815248212460418, | |
| "learning_rate": 6.287643083477275e-06, | |
| "loss": 0.1303, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.8367346938775511, | |
| "grad_norm": 0.6366648152395168, | |
| "learning_rate": 6.268272919547537e-06, | |
| "loss": 0.1522, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 0.5957035573721868, | |
| "learning_rate": 6.248882390836135e-06, | |
| "loss": 0.1411, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8418367346938775, | |
| "grad_norm": 0.628234311427476, | |
| "learning_rate": 6.229471808698673e-06, | |
| "loss": 0.165, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8443877551020408, | |
| "grad_norm": 0.5801673506686771, | |
| "learning_rate": 6.2100414848127586e-06, | |
| "loss": 0.1496, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8469387755102041, | |
| "grad_norm": 0.6150706660795319, | |
| "learning_rate": 6.1905917311729915e-06, | |
| "loss": 0.1414, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8494897959183674, | |
| "grad_norm": 0.7390081360716291, | |
| "learning_rate": 6.17112286008596e-06, | |
| "loss": 0.1872, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8520408163265306, | |
| "grad_norm": 0.6878106569070934, | |
| "learning_rate": 6.151635184165219e-06, | |
| "loss": 0.1876, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8545918367346939, | |
| "grad_norm": 0.5683915193981929, | |
| "learning_rate": 6.132129016326279e-06, | |
| "loss": 0.1437, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.6162512785667651, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.1583, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8596938775510204, | |
| "grad_norm": 0.6164623578503621, | |
| "learning_rate": 6.093062458035433e-06, | |
| "loss": 0.1521, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.8622448979591837, | |
| "grad_norm": 0.6110050489294294, | |
| "learning_rate": 6.073502694879059e-06, | |
| "loss": 0.1462, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8647959183673469, | |
| "grad_norm": 0.5658668675870233, | |
| "learning_rate": 6.053925694385464e-06, | |
| "loss": 0.1415, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8673469387755102, | |
| "grad_norm": 0.6183851263339577, | |
| "learning_rate": 6.034331770904455e-06, | |
| "loss": 0.1678, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8698979591836735, | |
| "grad_norm": 0.6380267035308205, | |
| "learning_rate": 6.014721239057565e-06, | |
| "loss": 0.1647, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.8724489795918368, | |
| "grad_norm": 0.6475786946852516, | |
| "learning_rate": 5.9950944137330125e-06, | |
| "loss": 0.1626, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.7178296798128999, | |
| "learning_rate": 5.975451610080643e-06, | |
| "loss": 0.176, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.8775510204081632, | |
| "grad_norm": 0.548799383239884, | |
| "learning_rate": 5.955793143506863e-06, | |
| "loss": 0.132, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.8801020408163265, | |
| "grad_norm": 0.6140832996638459, | |
| "learning_rate": 5.936119329669584e-06, | |
| "loss": 0.1622, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8826530612244898, | |
| "grad_norm": 0.6117216867875228, | |
| "learning_rate": 5.916430484473149e-06, | |
| "loss": 0.166, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.8852040816326531, | |
| "grad_norm": 0.6438861515674844, | |
| "learning_rate": 5.896726924063263e-06, | |
| "loss": 0.1484, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.8877551020408163, | |
| "grad_norm": 0.684727641108475, | |
| "learning_rate": 5.877008964821909e-06, | |
| "loss": 0.1713, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.8903061224489796, | |
| "grad_norm": 0.6263934617347773, | |
| "learning_rate": 5.8572769233622794e-06, | |
| "loss": 0.1617, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.6512663586367785, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.1833, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8954081632653061, | |
| "grad_norm": 0.6446733318710774, | |
| "learning_rate": 5.81777186136646e-06, | |
| "loss": 0.1627, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8979591836734694, | |
| "grad_norm": 0.6157294973622298, | |
| "learning_rate": 5.797999475166897e-06, | |
| "loss": 0.1365, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9005102040816326, | |
| "grad_norm": 0.607524650157665, | |
| "learning_rate": 5.778214275412118e-06, | |
| "loss": 0.1541, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9030612244897959, | |
| "grad_norm": 0.5654449477539436, | |
| "learning_rate": 5.7584165797950055e-06, | |
| "loss": 0.1354, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9056122448979592, | |
| "grad_norm": 0.6257348947364124, | |
| "learning_rate": 5.738606706209079e-06, | |
| "loss": 0.1476, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9081632653061225, | |
| "grad_norm": 0.6459117447837404, | |
| "learning_rate": 5.71878497274341e-06, | |
| "loss": 0.155, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 0.5840351587262443, | |
| "learning_rate": 5.698951697677498e-06, | |
| "loss": 0.1407, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9132653061224489, | |
| "grad_norm": 0.6739427176791648, | |
| "learning_rate": 5.679107199476174e-06, | |
| "loss": 0.1667, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9158163265306123, | |
| "grad_norm": 0.6041549005412797, | |
| "learning_rate": 5.659251796784474e-06, | |
| "loss": 0.1697, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 0.6116781652359067, | |
| "learning_rate": 5.6393858084225305e-06, | |
| "loss": 0.1532, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9209183673469388, | |
| "grad_norm": 0.6077294016328789, | |
| "learning_rate": 5.6195095533804546e-06, | |
| "loss": 0.1675, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.923469387755102, | |
| "grad_norm": 0.6230191584366697, | |
| "learning_rate": 5.599623350813202e-06, | |
| "loss": 0.1595, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9260204081632653, | |
| "grad_norm": 0.6494202373448055, | |
| "learning_rate": 5.579727520035469e-06, | |
| "loss": 0.1687, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.678388379140103, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.196, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.9311224489795918, | |
| "grad_norm": 0.6795163128070991, | |
| "learning_rate": 5.53990825187518e-06, | |
| "loss": 0.1937, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9336734693877551, | |
| "grad_norm": 0.631118915172661, | |
| "learning_rate": 5.5199854538744905e-06, | |
| "loss": 0.1684, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9362244897959183, | |
| "grad_norm": 0.5111362970956352, | |
| "learning_rate": 5.500054306416776e-06, | |
| "loss": 0.135, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.9387755102040817, | |
| "grad_norm": 0.7200957437333302, | |
| "learning_rate": 5.480115129538409e-06, | |
| "loss": 0.197, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9413265306122449, | |
| "grad_norm": 0.5990796224723971, | |
| "learning_rate": 5.460168243404696e-06, | |
| "loss": 0.1404, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9438775510204082, | |
| "grad_norm": 0.6508909764184223, | |
| "learning_rate": 5.440213968304728e-06, | |
| "loss": 0.1653, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 0.7058087802665988, | |
| "learning_rate": 5.420252624646238e-06, | |
| "loss": 0.1889, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.9489795918367347, | |
| "grad_norm": 0.62520078208343, | |
| "learning_rate": 5.4002845329504675e-06, | |
| "loss": 0.151, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.951530612244898, | |
| "grad_norm": 0.6175896170477174, | |
| "learning_rate": 5.380310013847006e-06, | |
| "loss": 0.1617, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9540816326530612, | |
| "grad_norm": 0.6241321981735418, | |
| "learning_rate": 5.360329388068649e-06, | |
| "loss": 0.1738, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9566326530612245, | |
| "grad_norm": 0.5387011325868518, | |
| "learning_rate": 5.340342976446251e-06, | |
| "loss": 0.1327, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9591836734693877, | |
| "grad_norm": 0.6360363000351817, | |
| "learning_rate": 5.320351099903565e-06, | |
| "loss": 0.1798, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.9617346938775511, | |
| "grad_norm": 0.6197073671288826, | |
| "learning_rate": 5.3003540794521005e-06, | |
| "loss": 0.1536, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.5982694119598283, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.1427, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9668367346938775, | |
| "grad_norm": 0.5902423939987766, | |
| "learning_rate": 5.260345891276685e-06, | |
| "loss": 0.1399, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.9693877551020408, | |
| "grad_norm": 0.6648465322813785, | |
| "learning_rate": 5.240335365968104e-06, | |
| "loss": 0.1691, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9719387755102041, | |
| "grad_norm": 0.6128401036727914, | |
| "learning_rate": 5.220320981571168e-06, | |
| "loss": 0.1575, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.9744897959183674, | |
| "grad_norm": 0.6072504634016694, | |
| "learning_rate": 5.2003030594587964e-06, | |
| "loss": 0.1677, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9770408163265306, | |
| "grad_norm": 0.5721919519862779, | |
| "learning_rate": 5.18028192106071e-06, | |
| "loss": 0.1444, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 0.7081920183600142, | |
| "learning_rate": 5.160257887858278e-06, | |
| "loss": 0.2072, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 0.6065745788364745, | |
| "learning_rate": 5.140231281379345e-06, | |
| "loss": 0.1522, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9846938775510204, | |
| "grad_norm": 0.6371181525494265, | |
| "learning_rate": 5.120202423193085e-06, | |
| "loss": 0.1406, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.9872448979591837, | |
| "grad_norm": 0.6505788223343205, | |
| "learning_rate": 5.100171634904821e-06, | |
| "loss": 0.1683, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.9897959183673469, | |
| "grad_norm": 0.6419015901383764, | |
| "learning_rate": 5.080139238150869e-06, | |
| "loss": 0.1739, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.9923469387755102, | |
| "grad_norm": 0.5741359343804769, | |
| "learning_rate": 5.060105554593374e-06, | |
| "loss": 0.1432, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.9948979591836735, | |
| "grad_norm": 0.5944648027143441, | |
| "learning_rate": 5.040070905915139e-06, | |
| "loss": 0.1578, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9974489795918368, | |
| "grad_norm": 0.5745910641764089, | |
| "learning_rate": 5.020035613814469e-06, | |
| "loss": 0.1449, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6947779028476239, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1125, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.0025510204081634, | |
| "grad_norm": 0.5323047103078277, | |
| "learning_rate": 4.979964386185531e-06, | |
| "loss": 0.097, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.0051020408163265, | |
| "grad_norm": 0.563964010101689, | |
| "learning_rate": 4.959929094084862e-06, | |
| "loss": 0.135, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.0076530612244898, | |
| "grad_norm": 0.6137555160835548, | |
| "learning_rate": 4.9398944454066275e-06, | |
| "loss": 0.1438, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.010204081632653, | |
| "grad_norm": 0.5584169281749033, | |
| "learning_rate": 4.919860761849132e-06, | |
| "loss": 0.1259, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.0127551020408163, | |
| "grad_norm": 0.5655350454746906, | |
| "learning_rate": 4.89982836509518e-06, | |
| "loss": 0.1227, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.0153061224489797, | |
| "grad_norm": 0.5932329728601095, | |
| "learning_rate": 4.879797576806915e-06, | |
| "loss": 0.132, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.0178571428571428, | |
| "grad_norm": 0.6072074491772425, | |
| "learning_rate": 4.859768718620656e-06, | |
| "loss": 0.1224, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.0204081632653061, | |
| "grad_norm": 0.5867196136790125, | |
| "learning_rate": 4.839742112141725e-06, | |
| "loss": 0.1183, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0204081632653061, | |
| "eval_loss": 0.17335140705108643, | |
| "eval_runtime": 2.1717, | |
| "eval_samples_per_second": 14.735, | |
| "eval_steps_per_second": 3.684, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0229591836734695, | |
| "grad_norm": 0.6842931691283886, | |
| "learning_rate": 4.8197180789392904e-06, | |
| "loss": 0.1289, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.0255102040816326, | |
| "grad_norm": 0.5843835873726522, | |
| "learning_rate": 4.799696940541204e-06, | |
| "loss": 0.1224, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.028061224489796, | |
| "grad_norm": 0.6303376367215173, | |
| "learning_rate": 4.779679018428832e-06, | |
| "loss": 0.1159, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.030612244897959, | |
| "grad_norm": 0.6128285299025862, | |
| "learning_rate": 4.759664634031897e-06, | |
| "loss": 0.1109, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.0331632653061225, | |
| "grad_norm": 0.6997351244068603, | |
| "learning_rate": 4.739654108723317e-06, | |
| "loss": 0.1211, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 0.6504032679810356, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.1299, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.038265306122449, | |
| "grad_norm": 0.609002998941228, | |
| "learning_rate": 4.699645920547901e-06, | |
| "loss": 0.1227, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.0408163265306123, | |
| "grad_norm": 0.6418661123328896, | |
| "learning_rate": 4.679648900096436e-06, | |
| "loss": 0.1301, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.0433673469387754, | |
| "grad_norm": 0.6541209431022946, | |
| "learning_rate": 4.659657023553752e-06, | |
| "loss": 0.1149, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.0459183673469388, | |
| "grad_norm": 0.5416993630355218, | |
| "learning_rate": 4.6396706119313526e-06, | |
| "loss": 0.0949, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0484693877551021, | |
| "grad_norm": 0.6686437193372081, | |
| "learning_rate": 4.619689986152996e-06, | |
| "loss": 0.139, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.0510204081632653, | |
| "grad_norm": 0.6680097009304201, | |
| "learning_rate": 4.599715467049534e-06, | |
| "loss": 0.1446, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.0535714285714286, | |
| "grad_norm": 0.5528151642549715, | |
| "learning_rate": 4.579747375353763e-06, | |
| "loss": 0.1075, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.0561224489795917, | |
| "grad_norm": 0.5916503836624794, | |
| "learning_rate": 4.559786031695275e-06, | |
| "loss": 0.1113, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.058673469387755, | |
| "grad_norm": 0.5999571410820965, | |
| "learning_rate": 4.5398317565953055e-06, | |
| "loss": 0.1168, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.0612244897959184, | |
| "grad_norm": 0.6152629066291727, | |
| "learning_rate": 4.5198848704615915e-06, | |
| "loss": 0.1263, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.0637755102040816, | |
| "grad_norm": 0.5434823415058014, | |
| "learning_rate": 4.499945693583227e-06, | |
| "loss": 0.1077, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.066326530612245, | |
| "grad_norm": 0.6220819462707029, | |
| "learning_rate": 4.480014546125511e-06, | |
| "loss": 0.1306, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.068877551020408, | |
| "grad_norm": 0.5835461892516719, | |
| "learning_rate": 4.4600917481248205e-06, | |
| "loss": 0.1117, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.633213180463133, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.1165, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0739795918367347, | |
| "grad_norm": 0.6041200748663155, | |
| "learning_rate": 4.420272479964532e-06, | |
| "loss": 0.1143, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.0765306122448979, | |
| "grad_norm": 0.6969958031590414, | |
| "learning_rate": 4.4003766491867984e-06, | |
| "loss": 0.1212, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.0790816326530612, | |
| "grad_norm": 0.6445563858005139, | |
| "learning_rate": 4.380490446619547e-06, | |
| "loss": 0.1389, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.0816326530612246, | |
| "grad_norm": 0.6125954009368678, | |
| "learning_rate": 4.3606141915774695e-06, | |
| "loss": 0.1092, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.0841836734693877, | |
| "grad_norm": 0.8037138319949373, | |
| "learning_rate": 4.340748203215528e-06, | |
| "loss": 0.1489, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.086734693877551, | |
| "grad_norm": 0.5725494006977168, | |
| "learning_rate": 4.320892800523827e-06, | |
| "loss": 0.1162, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.0892857142857142, | |
| "grad_norm": 0.6096824707475266, | |
| "learning_rate": 4.3010483023225045e-06, | |
| "loss": 0.1123, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.0918367346938775, | |
| "grad_norm": 0.648005078531451, | |
| "learning_rate": 4.281215027256592e-06, | |
| "loss": 0.1198, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.094387755102041, | |
| "grad_norm": 0.5945617987155178, | |
| "learning_rate": 4.261393293790922e-06, | |
| "loss": 0.1079, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.096938775510204, | |
| "grad_norm": 0.5850588276218499, | |
| "learning_rate": 4.241583420204998e-06, | |
| "loss": 0.101, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0994897959183674, | |
| "grad_norm": 0.6791345011904737, | |
| "learning_rate": 4.221785724587883e-06, | |
| "loss": 0.1417, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.1020408163265305, | |
| "grad_norm": 0.637553585890781, | |
| "learning_rate": 4.2020005248331056e-06, | |
| "loss": 0.1233, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.1045918367346939, | |
| "grad_norm": 0.6509522793471568, | |
| "learning_rate": 4.182228138633541e-06, | |
| "loss": 0.1199, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 0.6443343768927554, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.0941, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.1096938775510203, | |
| "grad_norm": 0.6027022006132112, | |
| "learning_rate": 4.142723076637723e-06, | |
| "loss": 0.1131, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.1122448979591837, | |
| "grad_norm": 0.6397623750787707, | |
| "learning_rate": 4.122991035178093e-06, | |
| "loss": 0.1347, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.114795918367347, | |
| "grad_norm": 0.6329399298444501, | |
| "learning_rate": 4.103273075936739e-06, | |
| "loss": 0.1328, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.1173469387755102, | |
| "grad_norm": 0.5764546644609387, | |
| "learning_rate": 4.083569515526853e-06, | |
| "loss": 0.1107, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.1198979591836735, | |
| "grad_norm": 0.6606104418167631, | |
| "learning_rate": 4.063880670330417e-06, | |
| "loss": 0.128, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.1224489795918366, | |
| "grad_norm": 0.6106517768536024, | |
| "learning_rate": 4.04420685649314e-06, | |
| "loss": 0.1152, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 0.6184011733260752, | |
| "learning_rate": 4.02454838991936e-06, | |
| "loss": 0.131, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.1275510204081634, | |
| "grad_norm": 0.6214106226786603, | |
| "learning_rate": 4.004905586266988e-06, | |
| "loss": 0.1218, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.1301020408163265, | |
| "grad_norm": 0.5864515890113116, | |
| "learning_rate": 3.985278760942437e-06, | |
| "loss": 0.1131, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.1326530612244898, | |
| "grad_norm": 0.6318618768163554, | |
| "learning_rate": 3.965668229095546e-06, | |
| "loss": 0.1052, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.135204081632653, | |
| "grad_norm": 0.6627304810877596, | |
| "learning_rate": 3.946074305614537e-06, | |
| "loss": 0.1296, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.1377551020408163, | |
| "grad_norm": 0.587382351188069, | |
| "learning_rate": 3.926497305120943e-06, | |
| "loss": 0.1045, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.1403061224489797, | |
| "grad_norm": 0.7385589854862445, | |
| "learning_rate": 3.906937541964567e-06, | |
| "loss": 0.1396, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.6636112117501785, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.1233, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.1454081632653061, | |
| "grad_norm": 0.589345932247423, | |
| "learning_rate": 3.8678709836737225e-06, | |
| "loss": 0.1168, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.1479591836734695, | |
| "grad_norm": 0.6799593595492233, | |
| "learning_rate": 3.848364815834782e-06, | |
| "loss": 0.1466, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1505102040816326, | |
| "grad_norm": 0.6091206654350524, | |
| "learning_rate": 3.828877139914042e-06, | |
| "loss": 0.1138, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.153061224489796, | |
| "grad_norm": 0.5823166449307083, | |
| "learning_rate": 3.809408268827009e-06, | |
| "loss": 0.1189, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.155612244897959, | |
| "grad_norm": 0.6079458200187238, | |
| "learning_rate": 3.7899585151872444e-06, | |
| "loss": 0.1302, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.1581632653061225, | |
| "grad_norm": 0.6287432593849783, | |
| "learning_rate": 3.7705281913013286e-06, | |
| "loss": 0.1315, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.1607142857142858, | |
| "grad_norm": 0.6137365328324991, | |
| "learning_rate": 3.751117609163865e-06, | |
| "loss": 0.126, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.163265306122449, | |
| "grad_norm": 0.6115580022695268, | |
| "learning_rate": 3.731727080452464e-06, | |
| "loss": 0.1161, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.1658163265306123, | |
| "grad_norm": 0.5932705584378828, | |
| "learning_rate": 3.7123569165227263e-06, | |
| "loss": 0.1059, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.1683673469387754, | |
| "grad_norm": 0.7061338946181202, | |
| "learning_rate": 3.6930074284032613e-06, | |
| "loss": 0.1492, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.1709183673469388, | |
| "grad_norm": 0.6614069888583459, | |
| "learning_rate": 3.6736789267906803e-06, | |
| "loss": 0.1287, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.1734693877551021, | |
| "grad_norm": 0.6099727489081356, | |
| "learning_rate": 3.654371722044616e-06, | |
| "loss": 0.1233, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.1760204081632653, | |
| "grad_norm": 0.5734538026453548, | |
| "learning_rate": 3.635086124182734e-06, | |
| "loss": 0.1141, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 0.6570242670654317, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.1339, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.181122448979592, | |
| "grad_norm": 0.6261495037300829, | |
| "learning_rate": 3.5965809874424817e-06, | |
| "loss": 0.1189, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.183673469387755, | |
| "grad_norm": 0.6587312495933207, | |
| "learning_rate": 3.5773620668448384e-06, | |
| "loss": 0.1341, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.1862244897959184, | |
| "grad_norm": 0.564185159561041, | |
| "learning_rate": 3.5581659896829024e-06, | |
| "loss": 0.115, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.1887755102040816, | |
| "grad_norm": 0.5477375739868198, | |
| "learning_rate": 3.538993064189954e-06, | |
| "loss": 0.1016, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.191326530612245, | |
| "grad_norm": 0.5636030369867936, | |
| "learning_rate": 3.519843598227521e-06, | |
| "loss": 0.1091, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.193877551020408, | |
| "grad_norm": 0.6017788620294022, | |
| "learning_rate": 3.500717899280442e-06, | |
| "loss": 0.1091, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.1964285714285714, | |
| "grad_norm": 0.5961071194795863, | |
| "learning_rate": 3.4816162744519266e-06, | |
| "loss": 0.1228, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.1989795918367347, | |
| "grad_norm": 0.6155967911771543, | |
| "learning_rate": 3.4625390304586224e-06, | |
| "loss": 0.1209, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.2015306122448979, | |
| "grad_norm": 0.5994417641613636, | |
| "learning_rate": 3.4434864736256936e-06, | |
| "loss": 0.1126, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.2040816326530612, | |
| "grad_norm": 0.5962055760170724, | |
| "learning_rate": 3.424458909881897e-06, | |
| "loss": 0.1238, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.2066326530612246, | |
| "grad_norm": 0.6212245734208236, | |
| "learning_rate": 3.4054566447546755e-06, | |
| "loss": 0.1203, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.2091836734693877, | |
| "grad_norm": 0.6674392281469284, | |
| "learning_rate": 3.386479983365251e-06, | |
| "loss": 0.1292, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.211734693877551, | |
| "grad_norm": 0.619719014082779, | |
| "learning_rate": 3.3675292304237184e-06, | |
| "loss": 0.123, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 0.6787663179071107, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.1297, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.2168367346938775, | |
| "grad_norm": 0.615581005728474, | |
| "learning_rate": 3.3297066666397715e-06, | |
| "loss": 0.1261, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.219387755102041, | |
| "grad_norm": 0.6772704385194932, | |
| "learning_rate": 3.310835463117942e-06, | |
| "loss": 0.1124, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.221938775510204, | |
| "grad_norm": 0.5788924276971349, | |
| "learning_rate": 3.291991382675429e-06, | |
| "loss": 0.1085, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.2244897959183674, | |
| "grad_norm": 0.6627991978200464, | |
| "learning_rate": 3.273174727893463e-06, | |
| "loss": 0.1342, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.2270408163265305, | |
| "grad_norm": 0.5795534308305166, | |
| "learning_rate": 3.2543858009129028e-06, | |
| "loss": 0.1168, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.2295918367346939, | |
| "grad_norm": 0.5802342946393737, | |
| "learning_rate": 3.235624903429374e-06, | |
| "loss": 0.1084, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.2321428571428572, | |
| "grad_norm": 0.6081875215495436, | |
| "learning_rate": 3.216892336688435e-06, | |
| "loss": 0.1255, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.2346938775510203, | |
| "grad_norm": 0.5606290959505179, | |
| "learning_rate": 3.198188401480734e-06, | |
| "loss": 0.1054, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.2372448979591837, | |
| "grad_norm": 0.6254832539000906, | |
| "learning_rate": 3.179513398137176e-06, | |
| "loss": 0.1229, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.239795918367347, | |
| "grad_norm": 0.6016469759559874, | |
| "learning_rate": 3.160867626524107e-06, | |
| "loss": 0.1238, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.2423469387755102, | |
| "grad_norm": 0.6212628039144623, | |
| "learning_rate": 3.1422513860384972e-06, | |
| "loss": 0.1251, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.2448979591836735, | |
| "grad_norm": 0.6001544256782972, | |
| "learning_rate": 3.12366497560313e-06, | |
| "loss": 0.1107, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.2474489795918366, | |
| "grad_norm": 0.7212798836278922, | |
| "learning_rate": 3.105108693661807e-06, | |
| "loss": 0.1378, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.5877635178170327, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.1235, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.2525510204081631, | |
| "grad_norm": 0.6030584719864545, | |
| "learning_rate": 3.0680877066128287e-06, | |
| "loss": 0.1154, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.2551020408163265, | |
| "grad_norm": 0.5769299888240651, | |
| "learning_rate": 3.049623595954766e-06, | |
| "loss": 0.1066, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.2576530612244898, | |
| "grad_norm": 0.640929927230285, | |
| "learning_rate": 3.031190802680383e-06, | |
| "loss": 0.1272, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.260204081632653, | |
| "grad_norm": 0.6131724938331422, | |
| "learning_rate": 3.0127896227668367e-06, | |
| "loss": 0.136, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.2627551020408163, | |
| "grad_norm": 0.6420939632710524, | |
| "learning_rate": 2.9944203516836614e-06, | |
| "loss": 0.1303, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.2653061224489797, | |
| "grad_norm": 0.5545302124781333, | |
| "learning_rate": 2.976083284388031e-06, | |
| "loss": 0.1128, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.2678571428571428, | |
| "grad_norm": 0.612470227896181, | |
| "learning_rate": 2.95777871532002e-06, | |
| "loss": 0.1118, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.2704081632653061, | |
| "grad_norm": 0.6065384958439005, | |
| "learning_rate": 2.9395069383978725e-06, | |
| "loss": 0.125, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.2729591836734695, | |
| "grad_norm": 0.667310023102144, | |
| "learning_rate": 2.9212682470132904e-06, | |
| "loss": 0.1334, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.2755102040816326, | |
| "grad_norm": 0.5680061294949651, | |
| "learning_rate": 2.9030629340267165e-06, | |
| "loss": 0.1013, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2755102040816326, | |
| "eval_loss": 0.17167198657989502, | |
| "eval_runtime": 2.1684, | |
| "eval_samples_per_second": 14.757, | |
| "eval_steps_per_second": 3.689, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.278061224489796, | |
| "grad_norm": 0.5741423355706899, | |
| "learning_rate": 2.8848912917626295e-06, | |
| "loss": 0.1096, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.280612244897959, | |
| "grad_norm": 0.5925872438345459, | |
| "learning_rate": 2.8667536120048616e-06, | |
| "loss": 0.1157, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.2831632653061225, | |
| "grad_norm": 0.6383357759073394, | |
| "learning_rate": 2.8486501859918967e-06, | |
| "loss": 0.1206, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.6120389722512302, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.1172, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.288265306122449, | |
| "grad_norm": 0.6437124638460001, | |
| "learning_rate": 2.8125472573995903e-06, | |
| "loss": 0.1302, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.2908163265306123, | |
| "grad_norm": 0.6071220260605691, | |
| "learning_rate": 2.794548334528486e-06, | |
| "loss": 0.1202, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.2933673469387754, | |
| "grad_norm": 0.5954473533080002, | |
| "learning_rate": 2.776584824809355e-06, | |
| "loss": 0.1141, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.2959183673469388, | |
| "grad_norm": 0.6173913926365706, | |
| "learning_rate": 2.7586570166840154e-06, | |
| "loss": 0.1091, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.2984693877551021, | |
| "grad_norm": 0.5459335138030598, | |
| "learning_rate": 2.74076519802103e-06, | |
| "loss": 0.1035, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.3010204081632653, | |
| "grad_norm": 0.6207207997242377, | |
| "learning_rate": 2.7229096561110703e-06, | |
| "loss": 0.1234, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.3035714285714286, | |
| "grad_norm": 0.5489479626562238, | |
| "learning_rate": 2.705090677662311e-06, | |
| "loss": 0.1039, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.306122448979592, | |
| "grad_norm": 0.5972773116742518, | |
| "learning_rate": 2.687308548795825e-06, | |
| "loss": 0.1102, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.308673469387755, | |
| "grad_norm": 0.6008948675073799, | |
| "learning_rate": 2.6695635550409806e-06, | |
| "loss": 0.1103, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.3112244897959184, | |
| "grad_norm": 0.6672021707421822, | |
| "learning_rate": 2.651855981330872e-06, | |
| "loss": 0.1274, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.3137755102040816, | |
| "grad_norm": 0.6088405581814523, | |
| "learning_rate": 2.634186111997734e-06, | |
| "loss": 0.1102, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.316326530612245, | |
| "grad_norm": 0.6482576902690778, | |
| "learning_rate": 2.6165542307683744e-06, | |
| "loss": 0.1376, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.318877551020408, | |
| "grad_norm": 0.6204897670785255, | |
| "learning_rate": 2.5989606207596262e-06, | |
| "loss": 0.1148, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 0.573934588312968, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.1062, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.3239795918367347, | |
| "grad_norm": 0.65506919752253, | |
| "learning_rate": 2.5638893437941426e-06, | |
| "loss": 0.1274, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.3265306122448979, | |
| "grad_norm": 0.5654354860920429, | |
| "learning_rate": 2.5464122399803126e-06, | |
| "loss": 0.0995, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.3290816326530612, | |
| "grad_norm": 0.6193610227103749, | |
| "learning_rate": 2.5289745336638634e-06, | |
| "loss": 0.1106, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.3316326530612246, | |
| "grad_norm": 0.5918931034127528, | |
| "learning_rate": 2.5115765048437445e-06, | |
| "loss": 0.112, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.3341836734693877, | |
| "grad_norm": 0.5669060325477613, | |
| "learning_rate": 2.494218432881797e-06, | |
| "loss": 0.1059, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.336734693877551, | |
| "grad_norm": 0.5976675877158825, | |
| "learning_rate": 2.4769005964982718e-06, | |
| "loss": 0.1081, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.3392857142857144, | |
| "grad_norm": 0.6899118506088527, | |
| "learning_rate": 2.4596232737673544e-06, | |
| "loss": 0.1313, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.3418367346938775, | |
| "grad_norm": 0.6341327579512294, | |
| "learning_rate": 2.4423867421126923e-06, | |
| "loss": 0.1225, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.344387755102041, | |
| "grad_norm": 0.5935000252838433, | |
| "learning_rate": 2.425191278302954e-06, | |
| "loss": 0.1218, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.346938775510204, | |
| "grad_norm": 0.577233550155536, | |
| "learning_rate": 2.408037158447375e-06, | |
| "loss": 0.1117, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.3494897959183674, | |
| "grad_norm": 0.5910734869924921, | |
| "learning_rate": 2.3909246579913265e-06, | |
| "loss": 0.1052, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.3520408163265305, | |
| "grad_norm": 0.6022665804010583, | |
| "learning_rate": 2.3738540517118953e-06, | |
| "loss": 0.1092, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.3545918367346939, | |
| "grad_norm": 0.6087219965715591, | |
| "learning_rate": 2.356825613713463e-06, | |
| "loss": 0.1262, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 0.5961121795834876, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.1148, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.3596938775510203, | |
| "grad_norm": 0.5652951454072906, | |
| "learning_rate": 2.322896335587255e-06, | |
| "loss": 0.1009, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.3622448979591837, | |
| "grad_norm": 0.6288949087425784, | |
| "learning_rate": 2.305996040265198e-06, | |
| "loss": 0.1168, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.364795918367347, | |
| "grad_norm": 0.6079352763616246, | |
| "learning_rate": 2.289139002826835e-06, | |
| "loss": 0.1212, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.3673469387755102, | |
| "grad_norm": 0.6069904187749375, | |
| "learning_rate": 2.272325493947257e-06, | |
| "loss": 0.1126, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.3698979591836735, | |
| "grad_norm": 0.5730921239886521, | |
| "learning_rate": 2.255555783602609e-06, | |
| "loss": 0.1112, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.3724489795918369, | |
| "grad_norm": 0.6388291813483298, | |
| "learning_rate": 2.238830141065765e-06, | |
| "loss": 0.1283, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.5797940980374893, | |
| "learning_rate": 2.2221488349019903e-06, | |
| "loss": 0.1112, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.3775510204081631, | |
| "grad_norm": 0.7076481363035177, | |
| "learning_rate": 2.2055121329646416e-06, | |
| "loss": 0.1351, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.3801020408163265, | |
| "grad_norm": 0.624306986743108, | |
| "learning_rate": 2.1889203023908655e-06, | |
| "loss": 0.1185, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.3826530612244898, | |
| "grad_norm": 0.6470359831455992, | |
| "learning_rate": 2.1723736095972946e-06, | |
| "loss": 0.1351, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.385204081632653, | |
| "grad_norm": 0.6481944902344723, | |
| "learning_rate": 2.155872320275789e-06, | |
| "loss": 0.1276, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.3877551020408163, | |
| "grad_norm": 0.602073181996085, | |
| "learning_rate": 2.139416699389153e-06, | |
| "loss": 0.1186, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.3903061224489797, | |
| "grad_norm": 0.6084186098092307, | |
| "learning_rate": 2.123007011166894e-06, | |
| "loss": 0.1193, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 0.6633754617972489, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.136, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.3954081632653061, | |
| "grad_norm": 0.6598906975031021, | |
| "learning_rate": 2.090326485941571e-06, | |
| "loss": 0.1355, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.3979591836734695, | |
| "grad_norm": 0.6040777362158246, | |
| "learning_rate": 2.074056173692881e-06, | |
| "loss": 0.1165, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.4005102040816326, | |
| "grad_norm": 0.5450778208953521, | |
| "learning_rate": 2.0578328436088903e-06, | |
| "loss": 0.1043, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.403061224489796, | |
| "grad_norm": 0.5919992077560257, | |
| "learning_rate": 2.041656756189184e-06, | |
| "loss": 0.1078, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.405612244897959, | |
| "grad_norm": 0.5568670148515367, | |
| "learning_rate": 2.025528171174775e-06, | |
| "loss": 0.1043, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.4081632653061225, | |
| "grad_norm": 0.6076808020159301, | |
| "learning_rate": 2.00944734754392e-06, | |
| "loss": 0.1255, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.4107142857142856, | |
| "grad_norm": 0.589238565781633, | |
| "learning_rate": 1.9934145435079705e-06, | |
| "loss": 0.1199, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.413265306122449, | |
| "grad_norm": 0.6444093306560709, | |
| "learning_rate": 1.977430016507222e-06, | |
| "loss": 0.1272, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.4158163265306123, | |
| "grad_norm": 0.601713977214125, | |
| "learning_rate": 1.9614940232067758e-06, | |
| "loss": 0.1169, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.4183673469387754, | |
| "grad_norm": 1.1093633597999613, | |
| "learning_rate": 1.945606819492429e-06, | |
| "loss": 0.1201, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.4209183673469388, | |
| "grad_norm": 0.6666149410482589, | |
| "learning_rate": 1.929768660466557e-06, | |
| "loss": 0.1114, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.4234693877551021, | |
| "grad_norm": 0.6082880380250021, | |
| "learning_rate": 1.913979800444021e-06, | |
| "loss": 0.1153, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.4260204081632653, | |
| "grad_norm": 0.6431693821610938, | |
| "learning_rate": 1.898240492948083e-06, | |
| "loss": 0.1253, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.6278068786376103, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.123, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.431122448979592, | |
| "grad_norm": 0.6946774708929673, | |
| "learning_rate": 1.866911545646637e-06, | |
| "loss": 0.1289, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.433673469387755, | |
| "grad_norm": 0.6384174590542929, | |
| "learning_rate": 1.8513224088930814e-06, | |
| "loss": 0.1227, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.4362244897959184, | |
| "grad_norm": 0.5938784568604658, | |
| "learning_rate": 1.8357838307619574e-06, | |
| "loss": 0.1244, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.4387755102040816, | |
| "grad_norm": 0.6319464389681908, | |
| "learning_rate": 1.8202960607577246e-06, | |
| "loss": 0.1282, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.441326530612245, | |
| "grad_norm": 0.6527056924637811, | |
| "learning_rate": 1.8048593475690112e-06, | |
| "loss": 0.1361, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.443877551020408, | |
| "grad_norm": 0.6763003101411658, | |
| "learning_rate": 1.7894739390646227e-06, | |
| "loss": 0.1341, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.4464285714285714, | |
| "grad_norm": 0.6141066682445433, | |
| "learning_rate": 1.7741400822895633e-06, | |
| "loss": 0.1157, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.4489795918367347, | |
| "grad_norm": 0.6216654523919399, | |
| "learning_rate": 1.7588580234610592e-06, | |
| "loss": 0.1109, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.4515306122448979, | |
| "grad_norm": 0.6008945770620517, | |
| "learning_rate": 1.7436280079646172e-06, | |
| "loss": 0.1192, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.4540816326530612, | |
| "grad_norm": 0.6118854666306467, | |
| "learning_rate": 1.728450280350079e-06, | |
| "loss": 0.1167, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.4566326530612246, | |
| "grad_norm": 0.6283008080395546, | |
| "learning_rate": 1.7133250843276928e-06, | |
| "loss": 0.1277, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.4591836734693877, | |
| "grad_norm": 0.6654587725209296, | |
| "learning_rate": 1.6982526627642043e-06, | |
| "loss": 0.1276, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.461734693877551, | |
| "grad_norm": 0.7397859180981282, | |
| "learning_rate": 1.6832332576789501e-06, | |
| "loss": 0.131, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 0.6487672371648604, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.1356, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.4668367346938775, | |
| "grad_norm": 0.6205070557254059, | |
| "learning_rate": 1.6533544607601815e-06, | |
| "loss": 0.1234, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.469387755102041, | |
| "grad_norm": 0.6493137431186743, | |
| "learning_rate": 1.6384955486934157e-06, | |
| "loss": 0.1188, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.471938775510204, | |
| "grad_norm": 0.6479947647795395, | |
| "learning_rate": 1.6236906126306824e-06, | |
| "loss": 0.1404, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.4744897959183674, | |
| "grad_norm": 0.5133523501462218, | |
| "learning_rate": 1.6089398902962767e-06, | |
| "loss": 0.0822, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.4770408163265305, | |
| "grad_norm": 0.6734740471049492, | |
| "learning_rate": 1.5942436185439842e-06, | |
| "loss": 0.0994, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.4795918367346939, | |
| "grad_norm": 0.616152641405367, | |
| "learning_rate": 1.5796020333532696e-06, | |
| "loss": 0.1246, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.4821428571428572, | |
| "grad_norm": 0.6141253226645143, | |
| "learning_rate": 1.5650153698254916e-06, | |
| "loss": 0.1282, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.4846938775510203, | |
| "grad_norm": 0.6364488406850194, | |
| "learning_rate": 1.5504838621801272e-06, | |
| "loss": 0.1233, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.4872448979591837, | |
| "grad_norm": 0.622428601485364, | |
| "learning_rate": 1.5360077437510046e-06, | |
| "loss": 0.1147, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.489795918367347, | |
| "grad_norm": 0.6015605938585638, | |
| "learning_rate": 1.5215872469825682e-06, | |
| "loss": 0.1085, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.4923469387755102, | |
| "grad_norm": 0.6073886538256179, | |
| "learning_rate": 1.5072226034261373e-06, | |
| "loss": 0.1218, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.4948979591836735, | |
| "grad_norm": 0.6260867939739607, | |
| "learning_rate": 1.4929140437361916e-06, | |
| "loss": 0.1307, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.4974489795918369, | |
| "grad_norm": 0.6184344057757138, | |
| "learning_rate": 1.4786617976666674e-06, | |
| "loss": 0.1249, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.6501373896364471, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.1303, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.5025510204081631, | |
| "grad_norm": 0.59898207205143, | |
| "learning_rate": 1.4503271608797741e-06, | |
| "loss": 0.1252, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.5051020408163265, | |
| "grad_norm": 0.5658331177409838, | |
| "learning_rate": 1.4362452251344283e-06, | |
| "loss": 0.1066, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.5076530612244898, | |
| "grad_norm": 0.5963824135058199, | |
| "learning_rate": 1.4222205129462347e-06, | |
| "loss": 0.11, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.510204081632653, | |
| "grad_norm": 0.6320736861925936, | |
| "learning_rate": 1.4082532495113627e-06, | |
| "loss": 0.1187, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.5127551020408163, | |
| "grad_norm": 0.608135373271551, | |
| "learning_rate": 1.3943436591035208e-06, | |
| "loss": 0.1191, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.5153061224489797, | |
| "grad_norm": 0.5958622972293384, | |
| "learning_rate": 1.3804919650703551e-06, | |
| "loss": 0.1112, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.5178571428571428, | |
| "grad_norm": 0.5984589902887925, | |
| "learning_rate": 1.3666983898298659e-06, | |
| "loss": 0.1063, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.5204081632653061, | |
| "grad_norm": 0.5813429342860311, | |
| "learning_rate": 1.3529631548668298e-06, | |
| "loss": 0.1092, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.5229591836734695, | |
| "grad_norm": 0.5785015266459514, | |
| "learning_rate": 1.3392864807292533e-06, | |
| "loss": 0.1192, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.5255102040816326, | |
| "grad_norm": 0.6096834391748477, | |
| "learning_rate": 1.3256685870248227e-06, | |
| "loss": 0.1178, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.5280612244897958, | |
| "grad_norm": 0.6385790013683148, | |
| "learning_rate": 1.3121096924173822e-06, | |
| "loss": 0.135, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.5306122448979593, | |
| "grad_norm": 0.5985384578843902, | |
| "learning_rate": 1.298610014623423e-06, | |
| "loss": 0.1174, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5306122448979593, | |
| "eval_loss": 0.1697322279214859, | |
| "eval_runtime": 2.1699, | |
| "eval_samples_per_second": 14.747, | |
| "eval_steps_per_second": 3.687, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5331632653061225, | |
| "grad_norm": 0.5750684962923132, | |
| "learning_rate": 1.2851697704085835e-06, | |
| "loss": 0.1163, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 0.5726286154930482, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.1014, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.538265306122449, | |
| "grad_norm": 0.5978740659319679, | |
| "learning_rate": 1.2584684450037033e-06, | |
| "loss": 0.1243, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.5408163265306123, | |
| "grad_norm": 0.5756266287690696, | |
| "learning_rate": 1.2452077925594435e-06, | |
| "loss": 0.122, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.5433673469387754, | |
| "grad_norm": 0.6125015524156492, | |
| "learning_rate": 1.2320074311789804e-06, | |
| "loss": 0.1175, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.5459183673469388, | |
| "grad_norm": 0.6519146321418777, | |
| "learning_rate": 1.2188675728217986e-06, | |
| "loss": 0.1239, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.5484693877551021, | |
| "grad_norm": 0.5822060099198982, | |
| "learning_rate": 1.2057884284758831e-06, | |
| "loss": 0.1104, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.5510204081632653, | |
| "grad_norm": 0.5734092094057847, | |
| "learning_rate": 1.1927702081543279e-06, | |
| "loss": 0.1177, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.5535714285714286, | |
| "grad_norm": 0.6496922741781745, | |
| "learning_rate": 1.1798131208919628e-06, | |
| "loss": 0.1256, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.556122448979592, | |
| "grad_norm": 0.5759657676358546, | |
| "learning_rate": 1.166917374742e-06, | |
| "loss": 0.1065, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.558673469387755, | |
| "grad_norm": 0.5897621812551755, | |
| "learning_rate": 1.1540831767726868e-06, | |
| "loss": 0.1107, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.5612244897959182, | |
| "grad_norm": 0.6155690329879021, | |
| "learning_rate": 1.141310733063991e-06, | |
| "loss": 0.1297, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.5637755102040818, | |
| "grad_norm": 0.5972777096073331, | |
| "learning_rate": 1.1286002487042858e-06, | |
| "loss": 0.1104, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.566326530612245, | |
| "grad_norm": 0.5545756200377255, | |
| "learning_rate": 1.1159519277870507e-06, | |
| "loss": 0.1063, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.568877551020408, | |
| "grad_norm": 0.6473303351126259, | |
| "learning_rate": 1.1033659734076069e-06, | |
| "loss": 0.1357, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.6289627879632347, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.1223, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.5739795918367347, | |
| "grad_norm": 0.6548826721422767, | |
| "learning_rate": 1.078381971633004e-06, | |
| "loss": 0.1141, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.5765306122448979, | |
| "grad_norm": 0.6073010852184507, | |
| "learning_rate": 1.0659843254083919e-06, | |
| "loss": 0.1141, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.5790816326530612, | |
| "grad_norm": 0.624859565552834, | |
| "learning_rate": 1.0536498480562224e-06, | |
| "loss": 0.1037, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.5816326530612246, | |
| "grad_norm": 0.6145027976623533, | |
| "learning_rate": 1.041378737632402e-06, | |
| "loss": 0.1141, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.5841836734693877, | |
| "grad_norm": 0.6121850726408201, | |
| "learning_rate": 1.0291711911753426e-06, | |
| "loss": 0.1176, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.586734693877551, | |
| "grad_norm": 0.6281282406543977, | |
| "learning_rate": 1.0170274047028068e-06, | |
| "loss": 0.1341, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.5892857142857144, | |
| "grad_norm": 0.6207735780436981, | |
| "learning_rate": 1.004947573208756e-06, | |
| "loss": 0.1267, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.5918367346938775, | |
| "grad_norm": 0.593259088117168, | |
| "learning_rate": 9.929318906602176e-07, | |
| "loss": 0.1121, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.5943877551020407, | |
| "grad_norm": 0.6268891283106383, | |
| "learning_rate": 9.809805499941766e-07, | |
| "loss": 0.1163, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.5969387755102042, | |
| "grad_norm": 0.5988029304619094, | |
| "learning_rate": 9.690937431144725e-07, | |
| "loss": 0.1108, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.5994897959183674, | |
| "grad_norm": 0.5841571355523898, | |
| "learning_rate": 9.572716608887206e-07, | |
| "loss": 0.12, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.6020408163265305, | |
| "grad_norm": 0.6290182145775021, | |
| "learning_rate": 9.455144931452459e-07, | |
| "loss": 0.1154, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.6045918367346939, | |
| "grad_norm": 0.6042795874115894, | |
| "learning_rate": 9.338224286700331e-07, | |
| "loss": 0.1031, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 0.6567260772251352, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.1366, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.6096938775510203, | |
| "grad_norm": 0.6078750345148625, | |
| "learning_rate": 9.106343594384781e-07, | |
| "loss": 0.1097, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.6122448979591837, | |
| "grad_norm": 0.6804350031515204, | |
| "learning_rate": 8.991387270152202e-07, | |
| "loss": 0.1274, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.614795918367347, | |
| "grad_norm": 0.5317896654689815, | |
| "learning_rate": 8.877089425204139e-07, | |
| "loss": 0.0822, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.6173469387755102, | |
| "grad_norm": 0.6041659502330885, | |
| "learning_rate": 8.76345189483222e-07, | |
| "loss": 0.1231, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.6198979591836735, | |
| "grad_norm": 0.5840192411770336, | |
| "learning_rate": 8.650476503725302e-07, | |
| "loss": 0.1046, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.6224489795918369, | |
| "grad_norm": 0.6028973067898996, | |
| "learning_rate": 8.538165065940263e-07, | |
| "loss": 0.1121, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 0.5931218547558655, | |
| "learning_rate": 8.426519384872733e-07, | |
| "loss": 0.1192, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.6275510204081631, | |
| "grad_norm": 0.5938652446965892, | |
| "learning_rate": 8.315541253228332e-07, | |
| "loss": 0.1237, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.6301020408163265, | |
| "grad_norm": 0.5798559230674245, | |
| "learning_rate": 8.205232452993705e-07, | |
| "loss": 0.1023, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.6326530612244898, | |
| "grad_norm": 0.5604806286202133, | |
| "learning_rate": 8.095594755407971e-07, | |
| "loss": 0.1075, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.635204081632653, | |
| "grad_norm": 0.5927466172637318, | |
| "learning_rate": 7.986629920934325e-07, | |
| "loss": 0.1147, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.6377551020408163, | |
| "grad_norm": 0.7286784566346942, | |
| "learning_rate": 7.878339699231702e-07, | |
| "loss": 0.1396, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.6403061224489797, | |
| "grad_norm": 0.6117769245456077, | |
| "learning_rate": 7.770725829126729e-07, | |
| "loss": 0.1145, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.6722856470291241, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.1379, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.6454081632653061, | |
| "grad_norm": 0.6754780154914234, | |
| "learning_rate": 7.557534044687293e-07, | |
| "loss": 0.1367, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.6479591836734695, | |
| "grad_norm": 0.6734627155633237, | |
| "learning_rate": 7.451959553594051e-07, | |
| "loss": 0.1353, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.6505102040816326, | |
| "grad_norm": 0.6588713762790792, | |
| "learning_rate": 7.347068260525963e-07, | |
| "loss": 0.1386, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.6530612244897958, | |
| "grad_norm": 0.57992012984708, | |
| "learning_rate": 7.242861849732696e-07, | |
| "loss": 0.1171, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.6556122448979593, | |
| "grad_norm": 0.6147459621798507, | |
| "learning_rate": 7.139341994466742e-07, | |
| "loss": 0.1258, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.6581632653061225, | |
| "grad_norm": 0.6233471241208175, | |
| "learning_rate": 7.036510356956494e-07, | |
| "loss": 0.1108, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.6607142857142856, | |
| "grad_norm": 0.6273261120821847, | |
| "learning_rate": 6.934368588379553e-07, | |
| "loss": 0.1105, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.663265306122449, | |
| "grad_norm": 0.5674331409773821, | |
| "learning_rate": 6.832918328836247e-07, | |
| "loss": 0.1008, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.6658163265306123, | |
| "grad_norm": 0.6514909353266433, | |
| "learning_rate": 6.73216120732324e-07, | |
| "loss": 0.1369, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.6683673469387754, | |
| "grad_norm": 0.5875385008989574, | |
| "learning_rate": 6.632098841707458e-07, | |
| "loss": 0.1062, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.6709183673469388, | |
| "grad_norm": 0.6013339921903986, | |
| "learning_rate": 6.532732838700023e-07, | |
| "loss": 0.1053, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.6734693877551021, | |
| "grad_norm": 0.6317221666265858, | |
| "learning_rate": 6.43406479383053e-07, | |
| "loss": 0.121, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.6760204081632653, | |
| "grad_norm": 0.6159119707750507, | |
| "learning_rate": 6.336096291421379e-07, | |
| "loss": 0.117, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 0.6540774604197073, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.1372, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.681122448979592, | |
| "grad_norm": 0.5855990129942852, | |
| "learning_rate": 6.142264195085273e-07, | |
| "loss": 0.1124, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.683673469387755, | |
| "grad_norm": 0.5905644474117571, | |
| "learning_rate": 6.04640371353914e-07, | |
| "loss": 0.1103, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.6862244897959182, | |
| "grad_norm": 0.5828354953054207, | |
| "learning_rate": 5.951248999165032e-07, | |
| "loss": 0.1152, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.6887755102040818, | |
| "grad_norm": 0.6090515161337641, | |
| "learning_rate": 5.856801579871457e-07, | |
| "loss": 0.1186, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.691326530612245, | |
| "grad_norm": 0.6893535410804377, | |
| "learning_rate": 5.763062972209793e-07, | |
| "loss": 0.1313, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.693877551020408, | |
| "grad_norm": 0.6124288616769757, | |
| "learning_rate": 5.670034681349995e-07, | |
| "loss": 0.1233, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.6964285714285714, | |
| "grad_norm": 0.6003579181797115, | |
| "learning_rate": 5.577718201056392e-07, | |
| "loss": 0.1138, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.6989795918367347, | |
| "grad_norm": 0.6545089312294674, | |
| "learning_rate": 5.486115013663668e-07, | |
| "loss": 0.1265, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.7015306122448979, | |
| "grad_norm": 0.6202843082223166, | |
| "learning_rate": 5.395226590053126e-07, | |
| "loss": 0.1187, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.7040816326530612, | |
| "grad_norm": 0.6030705142860276, | |
| "learning_rate": 5.305054389629022e-07, | |
| "loss": 0.1031, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.7066326530612246, | |
| "grad_norm": 0.5826774737854693, | |
| "learning_rate": 5.215599860295162e-07, | |
| "loss": 0.1133, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.7091836734693877, | |
| "grad_norm": 0.6377804413011416, | |
| "learning_rate": 5.126864438431628e-07, | |
| "loss": 0.122, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.711734693877551, | |
| "grad_norm": 0.7283982000793318, | |
| "learning_rate": 5.038849548871705e-07, | |
| "loss": 0.1439, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.7141564642792316, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.1521, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.7168367346938775, | |
| "grad_norm": 0.593372869293056, | |
| "learning_rate": 4.864987008124949e-07, | |
| "loss": 0.1152, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.7193877551020407, | |
| "grad_norm": 0.5778699052985155, | |
| "learning_rate": 4.779142148665855e-07, | |
| "loss": 0.1167, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.7219387755102042, | |
| "grad_norm": 0.6047519754038335, | |
| "learning_rate": 4.6940234049210277e-07, | |
| "loss": 0.1144, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.7244897959183674, | |
| "grad_norm": 0.6089775983442018, | |
| "learning_rate": 4.6096321436504e-07, | |
| "loss": 0.1222, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.7270408163265305, | |
| "grad_norm": 0.6215389881009823, | |
| "learning_rate": 4.5259697199326714e-07, | |
| "loss": 0.1243, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.7295918367346939, | |
| "grad_norm": 0.5968710527528669, | |
| "learning_rate": 4.4430374771435245e-07, | |
| "loss": 0.1182, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.7321428571428572, | |
| "grad_norm": 0.5602996301369507, | |
| "learning_rate": 4.3608367469340553e-07, | |
| "loss": 0.1137, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.7346938775510203, | |
| "grad_norm": 0.6177976764551923, | |
| "learning_rate": 4.279368849209381e-07, | |
| "loss": 0.1171, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.7372448979591837, | |
| "grad_norm": 0.6362139473836395, | |
| "learning_rate": 4.198635092107456e-07, | |
| "loss": 0.1357, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.739795918367347, | |
| "grad_norm": 0.616733264645363, | |
| "learning_rate": 4.1186367719780737e-07, | |
| "loss": 0.1216, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.7423469387755102, | |
| "grad_norm": 0.6113272277992989, | |
| "learning_rate": 4.039375173362053e-07, | |
| "loss": 0.1247, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.7448979591836735, | |
| "grad_norm": 0.5859466621068377, | |
| "learning_rate": 3.960851568970586e-07, | |
| "loss": 0.1156, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.7474489795918369, | |
| "grad_norm": 0.5768103705810081, | |
| "learning_rate": 3.883067219664838e-07, | |
| "loss": 0.1146, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.5900740212059127, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.1183, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.7525510204081631, | |
| "grad_norm": 0.6071582238903932, | |
| "learning_rate": 3.729721270383585e-07, | |
| "loss": 0.1068, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.7551020408163265, | |
| "grad_norm": 0.579043109644617, | |
| "learning_rate": 3.6541621326989183e-07, | |
| "loss": 0.1133, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.7576530612244898, | |
| "grad_norm": 0.5317309897500903, | |
| "learning_rate": 3.579347174642073e-07, | |
| "loss": 0.095, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.760204081632653, | |
| "grad_norm": 0.5792381004031227, | |
| "learning_rate": 3.5052775975241203e-07, | |
| "loss": 0.119, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.7627551020408163, | |
| "grad_norm": 0.5806092547850344, | |
| "learning_rate": 3.431954590687464e-07, | |
| "loss": 0.1027, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.7653061224489797, | |
| "grad_norm": 0.5720600883457903, | |
| "learning_rate": 3.359379331486762e-07, | |
| "loss": 0.1076, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.7678571428571428, | |
| "grad_norm": 0.6301152495971145, | |
| "learning_rate": 3.287552985270015e-07, | |
| "loss": 0.1226, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.7704081632653061, | |
| "grad_norm": 0.6366540857894387, | |
| "learning_rate": 3.216476705359839e-07, | |
| "loss": 0.1302, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.7729591836734695, | |
| "grad_norm": 0.5689854208147107, | |
| "learning_rate": 3.146151633034983e-07, | |
| "loss": 0.0909, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.7755102040816326, | |
| "grad_norm": 0.624163406400808, | |
| "learning_rate": 3.076578897511978e-07, | |
| "loss": 0.1166, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.7780612244897958, | |
| "grad_norm": 0.605898818231271, | |
| "learning_rate": 3.0077596159270086e-07, | |
| "loss": 0.1137, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.7806122448979593, | |
| "grad_norm": 0.5928268623308518, | |
| "learning_rate": 2.939694893317979e-07, | |
| "loss": 0.112, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.7831632653061225, | |
| "grad_norm": 0.565357840643717, | |
| "learning_rate": 2.8723858226067493e-07, | |
| "loss": 0.0942, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.6756531287422569, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.1391, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "eval_loss": 0.16841819882392883, | |
| "eval_runtime": 2.4116, | |
| "eval_samples_per_second": 13.269, | |
| "eval_steps_per_second": 3.317, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.788265306122449, | |
| "grad_norm": 0.5735443897062006, | |
| "learning_rate": 2.7400389478799696e-07, | |
| "loss": 0.1065, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.7908163265306123, | |
| "grad_norm": 0.5810012198351104, | |
| "learning_rate": 2.6750032689710604e-07, | |
| "loss": 0.1082, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.7933673469387754, | |
| "grad_norm": 0.5958481412831177, | |
| "learning_rate": 2.610727492139131e-07, | |
| "loss": 0.1119, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.7959183673469388, | |
| "grad_norm": 0.5840809015166184, | |
| "learning_rate": 2.547212649466568e-07, | |
| "loss": 0.1036, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.7984693877551021, | |
| "grad_norm": 0.5888202219012438, | |
| "learning_rate": 2.4844597608173925e-07, | |
| "loss": 0.114, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.8010204081632653, | |
| "grad_norm": 0.6092819061535076, | |
| "learning_rate": 2.4224698338208344e-07, | |
| "loss": 0.1189, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.8035714285714286, | |
| "grad_norm": 0.6314320254586682, | |
| "learning_rate": 2.3612438638551837e-07, | |
| "loss": 0.1191, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.806122448979592, | |
| "grad_norm": 0.5897736239110092, | |
| "learning_rate": 2.3007828340318117e-07, | |
| "loss": 0.1081, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.808673469387755, | |
| "grad_norm": 0.6275390646683443, | |
| "learning_rate": 2.2410877151793354e-07, | |
| "loss": 0.1275, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.8112244897959182, | |
| "grad_norm": 0.5797864579749057, | |
| "learning_rate": 2.1821594658280932e-07, | |
| "loss": 0.1167, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.8137755102040818, | |
| "grad_norm": 0.5530302763446563, | |
| "learning_rate": 2.123999032194729e-07, | |
| "loss": 0.111, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.816326530612245, | |
| "grad_norm": 0.5920894147609517, | |
| "learning_rate": 2.0666073481669714e-07, | |
| "loss": 0.1173, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.818877551020408, | |
| "grad_norm": 0.5578466613023159, | |
| "learning_rate": 2.0099853352886721e-07, | |
| "loss": 0.1063, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 0.6118756903721626, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.1142, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.8239795918367347, | |
| "grad_norm": 0.5906387766424285, | |
| "learning_rate": 1.8990539473478896e-07, | |
| "loss": 0.1165, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.8265306122448979, | |
| "grad_norm": 0.573214883177142, | |
| "learning_rate": 1.8447463535214872e-07, | |
| "loss": 0.1143, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.8290816326530612, | |
| "grad_norm": 0.5543354201963542, | |
| "learning_rate": 1.7912119932881167e-07, | |
| "loss": 0.1022, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.8316326530612246, | |
| "grad_norm": 0.6540919785157406, | |
| "learning_rate": 1.7384517262542255e-07, | |
| "loss": 0.1284, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.8341836734693877, | |
| "grad_norm": 0.5724171162277051, | |
| "learning_rate": 1.686466399596548e-07, | |
| "loss": 0.1097, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.836734693877551, | |
| "grad_norm": 0.5988377208759277, | |
| "learning_rate": 1.6352568480485277e-07, | |
| "loss": 0.1109, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.8392857142857144, | |
| "grad_norm": 0.5946198760051452, | |
| "learning_rate": 1.5848238938869332e-07, | |
| "loss": 0.1218, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.8418367346938775, | |
| "grad_norm": 0.6610304566003866, | |
| "learning_rate": 1.5351683469185973e-07, | |
| "loss": 0.1247, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.8443877551020407, | |
| "grad_norm": 0.6351539530682201, | |
| "learning_rate": 1.4862910044674772e-07, | |
| "loss": 0.1292, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.8469387755102042, | |
| "grad_norm": 0.6207785896757562, | |
| "learning_rate": 1.4381926513618139e-07, | |
| "loss": 0.1214, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.8494897959183674, | |
| "grad_norm": 0.5778796347145898, | |
| "learning_rate": 1.390874059921543e-07, | |
| "loss": 0.1137, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.8520408163265305, | |
| "grad_norm": 0.6289894556783412, | |
| "learning_rate": 1.3443359899458997e-07, | |
| "loss": 0.1306, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.8545918367346939, | |
| "grad_norm": 0.5920906901307591, | |
| "learning_rate": 1.298579188701188e-07, | |
| "loss": 0.1162, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.6488560446877133, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.1326, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.8596938775510203, | |
| "grad_norm": 0.5940363576120224, | |
| "learning_rate": 1.2094123187335104e-07, | |
| "loss": 0.1111, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.8622448979591837, | |
| "grad_norm": 0.6319328716661287, | |
| "learning_rate": 1.1660036817716492e-07, | |
| "loss": 0.1324, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.864795918367347, | |
| "grad_norm": 0.6077519215437209, | |
| "learning_rate": 1.1233791770399516e-07, | |
| "loss": 0.1155, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.8673469387755102, | |
| "grad_norm": 0.6250261912069963, | |
| "learning_rate": 1.0815394889642339e-07, | |
| "loss": 0.1286, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.8698979591836735, | |
| "grad_norm": 0.5491011558690894, | |
| "learning_rate": 1.0404852893684425e-07, | |
| "loss": 0.099, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.8724489795918369, | |
| "grad_norm": 0.5744764104571831, | |
| "learning_rate": 1.0002172374638519e-07, | |
| "loss": 0.1079, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 0.6679548709900088, | |
| "learning_rate": 9.607359798384785e-08, | |
| "loss": 0.1237, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.8775510204081631, | |
| "grad_norm": 0.6108632547487393, | |
| "learning_rate": 9.22042150446728e-08, | |
| "loss": 0.1228, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.8801020408163265, | |
| "grad_norm": 0.5435663894998473, | |
| "learning_rate": 8.841363705991924e-08, | |
| "loss": 0.0981, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.8826530612244898, | |
| "grad_norm": 0.6268626044693186, | |
| "learning_rate": 8.470192489526519e-08, | |
| "loss": 0.1132, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.885204081632653, | |
| "grad_norm": 0.6193334720497722, | |
| "learning_rate": 8.106913815003503e-08, | |
| "loss": 0.0943, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.8877551020408163, | |
| "grad_norm": 0.6074530776682006, | |
| "learning_rate": 7.7515335156238e-08, | |
| "loss": 0.1194, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.8903061224489797, | |
| "grad_norm": 0.5440145745307361, | |
| "learning_rate": 7.4040572977635e-08, | |
| "loss": 0.0938, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 0.6266963631247235, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.1254, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.8954081632653061, | |
| "grad_norm": 0.585333945184069, | |
| "learning_rate": 6.732839297432736e-08, | |
| "loss": 0.1125, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.8979591836734695, | |
| "grad_norm": 0.6102667577399207, | |
| "learning_rate": 6.409108292774912e-08, | |
| "loss": 0.1186, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.9005102040816326, | |
| "grad_norm": 0.5757285795331835, | |
| "learning_rate": 6.093302925088974e-08, | |
| "loss": 0.1113, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.9030612244897958, | |
| "grad_norm": 0.6411162442844834, | |
| "learning_rate": 5.785428265292381e-08, | |
| "loss": 0.1246, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.9056122448979593, | |
| "grad_norm": 0.6282690667914622, | |
| "learning_rate": 5.485489256958465e-08, | |
| "loss": 0.1232, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.9081632653061225, | |
| "grad_norm": 0.6100309844718854, | |
| "learning_rate": 5.1934907162370374e-08, | |
| "loss": 0.1287, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.9107142857142856, | |
| "grad_norm": 0.6013903017998977, | |
| "learning_rate": 4.909437331777178e-08, | |
| "loss": 0.1108, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.913265306122449, | |
| "grad_norm": 0.5687648257186603, | |
| "learning_rate": 4.63333366465174e-08, | |
| "loss": 0.1066, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.9158163265306123, | |
| "grad_norm": 0.5666552452982463, | |
| "learning_rate": 4.3651841482840185e-08, | |
| "loss": 0.0968, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.9183673469387754, | |
| "grad_norm": 0.6465617962350463, | |
| "learning_rate": 4.104993088376974e-08, | |
| "loss": 0.1218, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.9209183673469388, | |
| "grad_norm": 0.5822884224730042, | |
| "learning_rate": 3.8527646628437334e-08, | |
| "loss": 0.1132, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.9234693877551021, | |
| "grad_norm": 0.5758652097322462, | |
| "learning_rate": 3.608502921740753e-08, | |
| "loss": 0.1068, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.9260204081632653, | |
| "grad_norm": 0.7245821946400494, | |
| "learning_rate": 3.372211787202484e-08, | |
| "loss": 0.1202, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.6115149197447761, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.1221, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.931122448979592, | |
| "grad_norm": 0.6143567966734291, | |
| "learning_rate": 2.9235563863734274e-08, | |
| "loss": 0.1215, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.933673469387755, | |
| "grad_norm": 0.6825501189925092, | |
| "learning_rate": 2.7111993241860646e-08, | |
| "loss": 0.1281, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.9362244897959182, | |
| "grad_norm": 0.57001567551457, | |
| "learning_rate": 2.5068272766545776e-08, | |
| "loss": 0.1072, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.9387755102040818, | |
| "grad_norm": 0.6273070677346755, | |
| "learning_rate": 2.3104435254008852e-08, | |
| "loss": 0.1217, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.941326530612245, | |
| "grad_norm": 0.6108930925346002, | |
| "learning_rate": 2.1220512237779544e-08, | |
| "loss": 0.1267, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.943877551020408, | |
| "grad_norm": 0.6167546366354988, | |
| "learning_rate": 1.9416533968193428e-08, | |
| "loss": 0.124, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.9464285714285714, | |
| "grad_norm": 0.5886499101171274, | |
| "learning_rate": 1.769252941190458e-08, | |
| "loss": 0.11, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.9489795918367347, | |
| "grad_norm": 0.6360123535606697, | |
| "learning_rate": 1.6048526251421502e-08, | |
| "loss": 0.1269, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.9515306122448979, | |
| "grad_norm": 0.6104208596036047, | |
| "learning_rate": 1.4484550884664162e-08, | |
| "loss": 0.1196, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.9540816326530612, | |
| "grad_norm": 0.6460129093291147, | |
| "learning_rate": 1.3000628424535978e-08, | |
| "loss": 0.1265, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.9566326530612246, | |
| "grad_norm": 0.5905020194158532, | |
| "learning_rate": 1.1596782698524712e-08, | |
| "loss": 0.1142, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.9591836734693877, | |
| "grad_norm": 0.5472224778815454, | |
| "learning_rate": 1.0273036248318325e-08, | |
| "loss": 0.1081, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.961734693877551, | |
| "grad_norm": 0.5782839724577236, | |
| "learning_rate": 9.029410329441379e-09, | |
| "loss": 0.1058, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 0.5777960470695286, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.1094, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.9668367346938775, | |
| "grad_norm": 0.6786006991385576, | |
| "learning_rate": 6.782598674943686e-09, | |
| "loss": 0.1159, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.9693877551020407, | |
| "grad_norm": 0.6711169694004527, | |
| "learning_rate": 5.779449016595773e-09, | |
| "loss": 0.1335, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.9719387755102042, | |
| "grad_norm": 0.6530310832242592, | |
| "learning_rate": 4.856492043545102e-09, | |
| "loss": 0.1443, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.9744897959183674, | |
| "grad_norm": 0.6253745584184208, | |
| "learning_rate": 4.0137425758018935e-09, | |
| "loss": 0.136, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.9770408163265305, | |
| "grad_norm": 0.5727994729661025, | |
| "learning_rate": 3.2512141454760227e-09, | |
| "loss": 0.1042, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.9795918367346939, | |
| "grad_norm": 0.6480065878060496, | |
| "learning_rate": 2.568918996560532e-09, | |
| "loss": 0.1247, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.9821428571428572, | |
| "grad_norm": 0.5819417407468576, | |
| "learning_rate": 1.9668680847356735e-09, | |
| "loss": 0.1029, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.9846938775510203, | |
| "grad_norm": 0.6137075701886557, | |
| "learning_rate": 1.4450710771907184e-09, | |
| "loss": 0.1133, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.9872448979591837, | |
| "grad_norm": 0.6333191914868241, | |
| "learning_rate": 1.0035363524713015e-09, | |
| "loss": 0.1235, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.989795918367347, | |
| "grad_norm": 0.5929959113660918, | |
| "learning_rate": 6.422710003439747e-10, | |
| "loss": 0.1075, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.9923469387755102, | |
| "grad_norm": 0.5584038012931548, | |
| "learning_rate": 3.6128082168074286e-10, | |
| "loss": 0.1037, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.9948979591836735, | |
| "grad_norm": 0.5336444844217255, | |
| "learning_rate": 1.6057032836913623e-10, | |
| "loss": 0.0918, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.9974489795918369, | |
| "grad_norm": 0.557285241591839, | |
| "learning_rate": 4.0142743236160075e-11, | |
| "loss": 0.1073, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.49667553817630916, | |
| "learning_rate": 0.0, | |
| "loss": 0.0785, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 784, | |
| "total_flos": 44668733620224.0, | |
| "train_loss": 0.14301018513815136, | |
| "train_runtime": 1428.0634, | |
| "train_samples_per_second": 4.382, | |
| "train_steps_per_second": 0.549 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 784, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 800000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 44668733620224.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |