| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 7.2531029160060845, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.157, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.969179457167951, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.1171, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 8.038628883674138, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.2447, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 7.181284079485226, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.1856, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.112433344504312, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.1559, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.327113207619511, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.0448, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.465135685237857, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 1.0978, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.090888995306241, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9938, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 3.803896660128775, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 1.0318, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.6371472402264415, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9638, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.130642297497438, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 1.0166, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.3363564333285844, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.964, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.9280567347298754, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9003, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.5765742280096355, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.8629, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.244586230255915, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.8612, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.080737770345006, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.826, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.69897868957997, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.8079, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.0877396316585757, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7754, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.7936724065212142, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.8014, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6190503973836212, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7875, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.40854608886397, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.7444, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.3425668727675473, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.6596, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.3047129082935371, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.6612, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.1867424045056345, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.6806, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.1629015401642369, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.5862, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.2456712323806793, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.661, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.2343964304901094, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6788, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.0950402630889828, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6348, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.1158623959231708, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6418, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0799049840363595, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.5757, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.4340456623061018, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.4831, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.0767137939930178, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.5122, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 1.0640548990604022, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.4774, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 1.1292770560560779, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.4725, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 1.2830936948054563, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.4837, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.9714465641206192, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.4707, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 1.0139798872633365, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.4958, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.9414245892225763, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.5107, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.9508877861916378, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.452, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.077782334534896, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.443, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.9333972228759309, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.3766, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.9498345058338064, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.4221, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.8157186697432257, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.3539, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.9722855413289595, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.4064, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 1.046666112362766, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.3123, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 1.0318384917287478, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.3632, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.8701545298941484, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.351, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.9347847107663282, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.3594, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 1.0722442973361932, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.3511, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.9222871276707566, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.2802, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.8354667963739475, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.3242, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.7965683394359753, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2873, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.8088074596012741, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.2863, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.9363984502803274, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.2483, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.985795735598309, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.2687, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.8624139999745933, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.2809, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.922663278241632, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.2739, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.8255190193389739, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.2339, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.9231770862356226, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.2951, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.8612512913670906, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.1876, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.7235066708899192, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.1553, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.6892401918602734, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.2397, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.7919752712617845, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.1707, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 1.0019000154582967, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.2422, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.8595605346344654, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.1948, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.6868269015185005, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.2079, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.8201404657810867, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.2142, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.8858223845196631, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.1753, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.8564676917129438, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.2719, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.7481106289081885, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.2098, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.7557827659733048, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.1587, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.6178849364723907, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.1926, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.6917464826789695, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.1451, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.946288703629972, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.2055, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.7094569211291132, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.1693, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.7693312986002137, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.1381, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.5613704861389744, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.1605, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.8194236890861043, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.1806, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.7935539855186317, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.0962, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.709075887870442, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.2206, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.5973061314978039, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.1653, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.42683571712390966, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.0885, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.730721005416514, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.1024, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.6232407215074838, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.1749, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.6519531899982641, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.1421, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.5542510394037562, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.1319, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.7642027393646142, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.1755, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.6792213822912663, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.1519, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.6608649694457307, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.1402, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.39741301608753127, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.091, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.5550256285445277, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0827, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.7281341762576158, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.1305, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.48702701719547203, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.072, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.6692713682690057, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.1218, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.4820457166622916, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.1264, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.675203535202114, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.15, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.5635087151972364, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.098, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.8211452136343707, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.1517, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.4928534295482402, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.0969, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.6325667838690089, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.1355, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.320138165399711, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.0391, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.3475163628118662, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.1106, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.39052004347760355, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.1052, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.41830287424801427, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.1319, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.3903057657315893, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.1034, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.423367878773483, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.0851, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.4322152458302539, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.0931, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.3455706095976376, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.1037, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.4085126056084587, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1289, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.37706609226811444, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.1238, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.32695401506246546, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.1133, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.33195413336425694, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1142, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.3271667723305574, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.0774, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.29749672655824966, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.0911, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.32982425629246453, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.0846, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.341363149670028, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.0847, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.33000693337255044, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1033, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.33560643585719374, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.0809, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.396398350896948, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.1438, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.296896628046753, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.0711, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.31842649417369, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1147, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.2593391035603889, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.0725, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.3227774159496283, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.1093, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.2944522467405997, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.0863, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.323083196292361, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.1132, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.3207721540722538, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.1052, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.30665854482697075, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.0705, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.3605682843311115, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.1106, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.30815304279016853, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.075, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.32051726773859623, | |
| "learning_rate": 0.0, | |
| "loss": 0.1035, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 3.698694612320256e+16, | |
| "train_loss": 0.35700673948113737, | |
| "train_runtime": 2838.6013, | |
| "train_samples_per_second": 1.447, | |
| "train_steps_per_second": 0.046 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.698694612320256e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |