| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0005646527385659, | |
| "eval_steps": 222, | |
| "global_step": 443, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002258610954263128, | |
| "grad_norm": 0.696092426776886, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 2.1698, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002258610954263128, | |
| "eval_loss": 2.083223819732666, | |
| "eval_runtime": 250.3739, | |
| "eval_samples_per_second": 2.98, | |
| "eval_steps_per_second": 0.375, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004517221908526256, | |
| "grad_norm": 0.7258424758911133, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 2.2383, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006775832862789385, | |
| "grad_norm": 0.999031662940979, | |
| "learning_rate": 1e-05, | |
| "loss": 1.9518, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.009034443817052512, | |
| "grad_norm": 0.8558158278465271, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 2.3031, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01129305477131564, | |
| "grad_norm": 1.0158196687698364, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.3383, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01355166572557877, | |
| "grad_norm": 1.3310142755508423, | |
| "learning_rate": 2e-05, | |
| "loss": 2.3352, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.015810276679841896, | |
| "grad_norm": 0.7580591440200806, | |
| "learning_rate": 2.3333333333333336e-05, | |
| "loss": 1.6814, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.018068887634105024, | |
| "grad_norm": 0.7518572211265564, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 1.7568, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.020327498588368152, | |
| "grad_norm": 2.006964683532715, | |
| "learning_rate": 3e-05, | |
| "loss": 2.6346, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.02258610954263128, | |
| "grad_norm": 0.6915990114212036, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 2.3084, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.024844720496894408, | |
| "grad_norm": 0.8575783967971802, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 2.0648, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02710333145115754, | |
| "grad_norm": 1.1422725915908813, | |
| "learning_rate": 4e-05, | |
| "loss": 2.1607, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.029361942405420668, | |
| "grad_norm": 1.0447956323623657, | |
| "learning_rate": 4.3333333333333334e-05, | |
| "loss": 1.9577, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03162055335968379, | |
| "grad_norm": 1.0838192701339722, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 1.8649, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03387916431394692, | |
| "grad_norm": 1.2682744264602661, | |
| "learning_rate": 5e-05, | |
| "loss": 2.0382, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03613777526821005, | |
| "grad_norm": 1.5851482152938843, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 1.9882, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.038396386222473176, | |
| "grad_norm": 1.1369619369506836, | |
| "learning_rate": 5.666666666666667e-05, | |
| "loss": 1.609, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.040654997176736304, | |
| "grad_norm": 1.489020824432373, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6515, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04291360813099943, | |
| "grad_norm": 1.0247050523757935, | |
| "learning_rate": 6.333333333333333e-05, | |
| "loss": 1.7473, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04517221908526256, | |
| "grad_norm": 0.913925051689148, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.7019, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04743083003952569, | |
| "grad_norm": 1.1576547622680664, | |
| "learning_rate": 7e-05, | |
| "loss": 1.5537, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.049689440993788817, | |
| "grad_norm": 1.4580681324005127, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 1.6705, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 1.6036103963851929, | |
| "learning_rate": 7.666666666666667e-05, | |
| "loss": 1.6114, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05420666290231508, | |
| "grad_norm": 1.218241572380066, | |
| "learning_rate": 8e-05, | |
| "loss": 1.8225, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.05646527385657821, | |
| "grad_norm": 1.334812879562378, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 1.6176, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.058723884810841336, | |
| "grad_norm": 1.0912553071975708, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 1.7103, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.060982495765104464, | |
| "grad_norm": 0.8983953595161438, | |
| "learning_rate": 9e-05, | |
| "loss": 1.826, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06324110671936758, | |
| "grad_norm": 1.1796810626983643, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 1.3892, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06549971767363072, | |
| "grad_norm": 0.9814470410346985, | |
| "learning_rate": 9.666666666666667e-05, | |
| "loss": 1.4582, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06775832862789384, | |
| "grad_norm": 1.072553038597107, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4753, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07001693958215698, | |
| "grad_norm": 1.1693260669708252, | |
| "learning_rate": 9.999855343632036e-05, | |
| "loss": 1.6425, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0722755505364201, | |
| "grad_norm": 0.8432520031929016, | |
| "learning_rate": 9.999421382898329e-05, | |
| "loss": 1.6913, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07453416149068323, | |
| "grad_norm": 0.6651790738105774, | |
| "learning_rate": 9.998698142908953e-05, | |
| "loss": 1.2049, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07679277244494635, | |
| "grad_norm": 1.2236829996109009, | |
| "learning_rate": 9.997685665512418e-05, | |
| "loss": 1.3681, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07905138339920949, | |
| "grad_norm": 0.5604356527328491, | |
| "learning_rate": 9.99638400929324e-05, | |
| "loss": 1.2572, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08130999435347261, | |
| "grad_norm": 0.5676224827766418, | |
| "learning_rate": 9.994793249568569e-05, | |
| "loss": 1.3594, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08356860530773574, | |
| "grad_norm": 0.5219910144805908, | |
| "learning_rate": 9.99291347838381e-05, | |
| "loss": 1.597, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08582721626199886, | |
| "grad_norm": 0.7226600646972656, | |
| "learning_rate": 9.990744804507315e-05, | |
| "loss": 1.3577, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.088085827216262, | |
| "grad_norm": 0.5216960310935974, | |
| "learning_rate": 9.988287353424077e-05, | |
| "loss": 1.7175, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09034443817052512, | |
| "grad_norm": 0.5965225100517273, | |
| "learning_rate": 9.985541267328477e-05, | |
| "loss": 1.5252, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09260304912478826, | |
| "grad_norm": 0.5782445669174194, | |
| "learning_rate": 9.98250670511605e-05, | |
| "loss": 1.9063, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09486166007905138, | |
| "grad_norm": 0.9046345353126526, | |
| "learning_rate": 9.979183842374293e-05, | |
| "loss": 1.371, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09712027103331451, | |
| "grad_norm": 0.7153818607330322, | |
| "learning_rate": 9.975572871372513e-05, | |
| "loss": 1.3616, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.09937888198757763, | |
| "grad_norm": 1.479604959487915, | |
| "learning_rate": 9.971674001050686e-05, | |
| "loss": 1.4922, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10163749294184077, | |
| "grad_norm": 0.7710794806480408, | |
| "learning_rate": 9.967487457007381e-05, | |
| "loss": 1.0812, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 0.792734682559967, | |
| "learning_rate": 9.963013481486703e-05, | |
| "loss": 1.5412, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.10615471485036702, | |
| "grad_norm": 0.6719285249710083, | |
| "learning_rate": 9.958252333364267e-05, | |
| "loss": 1.1906, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.10841332580463016, | |
| "grad_norm": 1.0701615810394287, | |
| "learning_rate": 9.953204288132234e-05, | |
| "loss": 1.3426, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11067193675889328, | |
| "grad_norm": 1.067221760749817, | |
| "learning_rate": 9.947869637883358e-05, | |
| "loss": 1.435, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11293054771315642, | |
| "grad_norm": 0.6624462008476257, | |
| "learning_rate": 9.942248691294093e-05, | |
| "loss": 1.148, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11518915866741954, | |
| "grad_norm": 1.10640549659729, | |
| "learning_rate": 9.936341773606723e-05, | |
| "loss": 1.4428, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11744776962168267, | |
| "grad_norm": 0.7549408674240112, | |
| "learning_rate": 9.930149226610554e-05, | |
| "loss": 1.4182, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.11970638057594579, | |
| "grad_norm": 0.6992340683937073, | |
| "learning_rate": 9.923671408622129e-05, | |
| "loss": 1.3721, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12196499153020893, | |
| "grad_norm": 0.6394023895263672, | |
| "learning_rate": 9.916908694464492e-05, | |
| "loss": 1.5456, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.12422360248447205, | |
| "grad_norm": 0.8388894200325012, | |
| "learning_rate": 9.909861475445517e-05, | |
| "loss": 1.6266, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12648221343873517, | |
| "grad_norm": 0.5841717720031738, | |
| "learning_rate": 9.902530159335243e-05, | |
| "loss": 1.6483, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1287408243929983, | |
| "grad_norm": 0.635645866394043, | |
| "learning_rate": 9.894915170342295e-05, | |
| "loss": 1.2798, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.13099943534726144, | |
| "grad_norm": 0.8329370617866516, | |
| "learning_rate": 9.887016949089333e-05, | |
| "loss": 1.5148, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.13325804630152457, | |
| "grad_norm": 0.7651770710945129, | |
| "learning_rate": 9.878835952587559e-05, | |
| "loss": 1.4382, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13551665725578768, | |
| "grad_norm": 0.6698804497718811, | |
| "learning_rate": 9.870372654210265e-05, | |
| "loss": 1.4922, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13777526821005082, | |
| "grad_norm": 0.807618260383606, | |
| "learning_rate": 9.861627543665456e-05, | |
| "loss": 1.8907, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.14003387916431395, | |
| "grad_norm": 0.610428512096405, | |
| "learning_rate": 9.852601126967502e-05, | |
| "loss": 1.4187, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1422924901185771, | |
| "grad_norm": 0.5798942446708679, | |
| "learning_rate": 9.843293926407866e-05, | |
| "loss": 1.4308, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1445511010728402, | |
| "grad_norm": 0.5696601271629333, | |
| "learning_rate": 9.833706480524878e-05, | |
| "loss": 1.6111, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14680971202710333, | |
| "grad_norm": 0.5774890780448914, | |
| "learning_rate": 9.82383934407258e-05, | |
| "loss": 1.2417, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14906832298136646, | |
| "grad_norm": 0.6529080271720886, | |
| "learning_rate": 9.81369308798862e-05, | |
| "loss": 1.3103, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1513269339356296, | |
| "grad_norm": 0.47429534792900085, | |
| "learning_rate": 9.803268299361217e-05, | |
| "loss": 1.2813, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1535855448898927, | |
| "grad_norm": 0.9300480484962463, | |
| "learning_rate": 9.7925655813952e-05, | |
| "loss": 1.5553, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 0.6896660327911377, | |
| "learning_rate": 9.781585553377085e-05, | |
| "loss": 1.3391, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.15810276679841898, | |
| "grad_norm": 0.513652503490448, | |
| "learning_rate": 9.770328850639268e-05, | |
| "loss": 1.2952, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1603613777526821, | |
| "grad_norm": 0.6096103191375732, | |
| "learning_rate": 9.758796124523239e-05, | |
| "loss": 1.6856, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.16261998870694522, | |
| "grad_norm": 0.9341354370117188, | |
| "learning_rate": 9.746988042341906e-05, | |
| "loss": 1.2378, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16487859966120835, | |
| "grad_norm": 0.616295337677002, | |
| "learning_rate": 9.734905287340985e-05, | |
| "loss": 1.4577, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1671372106154715, | |
| "grad_norm": 0.5131121277809143, | |
| "learning_rate": 9.722548558659457e-05, | |
| "loss": 1.6515, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.16939582156973462, | |
| "grad_norm": 0.7272030711174011, | |
| "learning_rate": 9.709918571289114e-05, | |
| "loss": 1.3936, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17165443252399773, | |
| "grad_norm": 0.6986638903617859, | |
| "learning_rate": 9.697016056033201e-05, | |
| "loss": 1.5857, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.5578714609146118, | |
| "learning_rate": 9.683841759464113e-05, | |
| "loss": 1.4122, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.176171654432524, | |
| "grad_norm": 0.49430081248283386, | |
| "learning_rate": 9.670396443880208e-05, | |
| "loss": 1.5742, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17843026538678713, | |
| "grad_norm": 0.7834444642066956, | |
| "learning_rate": 9.656680887261693e-05, | |
| "loss": 1.4708, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.18068887634105024, | |
| "grad_norm": 0.7123986482620239, | |
| "learning_rate": 9.64269588322561e-05, | |
| "loss": 1.6196, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18294748729531338, | |
| "grad_norm": 0.7276789546012878, | |
| "learning_rate": 9.628442240979916e-05, | |
| "loss": 1.38, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.1852060982495765, | |
| "grad_norm": 0.580315351486206, | |
| "learning_rate": 9.613920785276656e-05, | |
| "loss": 1.62, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18746470920383965, | |
| "grad_norm": 0.8134167790412903, | |
| "learning_rate": 9.599132356364247e-05, | |
| "loss": 1.3458, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.18972332015810275, | |
| "grad_norm": 0.9041950702667236, | |
| "learning_rate": 9.584077809938855e-05, | |
| "loss": 1.3943, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1919819311123659, | |
| "grad_norm": 0.6255788207054138, | |
| "learning_rate": 9.568758017094883e-05, | |
| "loss": 1.2289, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19424054206662902, | |
| "grad_norm": 0.5172642469406128, | |
| "learning_rate": 9.553173864274567e-05, | |
| "loss": 1.4261, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19649915302089216, | |
| "grad_norm": 0.4685104489326477, | |
| "learning_rate": 9.537326253216685e-05, | |
| "loss": 1.4084, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.19875776397515527, | |
| "grad_norm": 0.5454607605934143, | |
| "learning_rate": 9.521216100904378e-05, | |
| "loss": 1.5901, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2010163749294184, | |
| "grad_norm": 0.5021257996559143, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 1.4189, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.20327498588368154, | |
| "grad_norm": 0.6921396851539612, | |
| "learning_rate": 9.488211916351656e-05, | |
| "loss": 1.3857, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20553359683794467, | |
| "grad_norm": 0.6917213797569275, | |
| "learning_rate": 9.471319793817426e-05, | |
| "loss": 1.6815, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 1.099507451057434, | |
| "learning_rate": 9.454168949330645e-05, | |
| "loss": 1.4097, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2100508187464709, | |
| "grad_norm": 0.5121451616287231, | |
| "learning_rate": 9.436760375282859e-05, | |
| "loss": 1.4267, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.21230942970073405, | |
| "grad_norm": 0.5218266248703003, | |
| "learning_rate": 9.419095078978506e-05, | |
| "loss": 1.2902, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.21456804065499718, | |
| "grad_norm": 0.5271095633506775, | |
| "learning_rate": 9.40117408257663e-05, | |
| "loss": 1.402, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21682665160926032, | |
| "grad_norm": 0.5236210227012634, | |
| "learning_rate": 9.382998423031727e-05, | |
| "loss": 1.7724, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.21908526256352343, | |
| "grad_norm": 0.5875529646873474, | |
| "learning_rate": 9.364569152033756e-05, | |
| "loss": 1.7372, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.22134387351778656, | |
| "grad_norm": 0.6892306208610535, | |
| "learning_rate": 9.345887335947281e-05, | |
| "loss": 1.6303, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2236024844720497, | |
| "grad_norm": 0.5231241583824158, | |
| "learning_rate": 9.326954055749767e-05, | |
| "loss": 1.618, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.22586109542631283, | |
| "grad_norm": 0.6638787388801575, | |
| "learning_rate": 9.30777040696903e-05, | |
| "loss": 1.1699, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22811970638057594, | |
| "grad_norm": 0.604011058807373, | |
| "learning_rate": 9.288337499619857e-05, | |
| "loss": 1.3338, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.23037831733483907, | |
| "grad_norm": 0.5968809127807617, | |
| "learning_rate": 9.268656458139762e-05, | |
| "loss": 1.4609, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2326369282891022, | |
| "grad_norm": 0.5389347672462463, | |
| "learning_rate": 9.248728421323941e-05, | |
| "loss": 1.5215, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.23489553924336534, | |
| "grad_norm": 0.712431013584137, | |
| "learning_rate": 9.22855454225936e-05, | |
| "loss": 1.0865, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.23715415019762845, | |
| "grad_norm": 0.7650911808013916, | |
| "learning_rate": 9.208135988258051e-05, | |
| "loss": 1.478, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23941276115189158, | |
| "grad_norm": 0.9736595153808594, | |
| "learning_rate": 9.187473940789557e-05, | |
| "loss": 1.4201, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.24167137210615472, | |
| "grad_norm": 0.703115701675415, | |
| "learning_rate": 9.166569595412575e-05, | |
| "loss": 1.3264, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.24392998306041785, | |
| "grad_norm": 0.9081875085830688, | |
| "learning_rate": 9.145424161705776e-05, | |
| "loss": 1.3541, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24618859401468096, | |
| "grad_norm": 0.5371877551078796, | |
| "learning_rate": 9.124038863197818e-05, | |
| "loss": 1.4782, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2484472049689441, | |
| "grad_norm": 0.5240530967712402, | |
| "learning_rate": 9.10241493729654e-05, | |
| "loss": 1.5567, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25070581592320723, | |
| "grad_norm": 0.6024327874183655, | |
| "learning_rate": 9.08055363521738e-05, | |
| "loss": 1.565, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.25296442687747034, | |
| "grad_norm": 0.6612941026687622, | |
| "learning_rate": 9.058456221910956e-05, | |
| "loss": 1.489, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2552230378317335, | |
| "grad_norm": 0.6889235377311707, | |
| "learning_rate": 9.036123975989892e-05, | |
| "loss": 1.4708, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2574816487859966, | |
| "grad_norm": 0.611516535282135, | |
| "learning_rate": 9.013558189654819e-05, | |
| "loss": 1.4802, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 0.5951672196388245, | |
| "learning_rate": 8.990760168619615e-05, | |
| "loss": 1.2703, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2619988706945229, | |
| "grad_norm": 0.670804500579834, | |
| "learning_rate": 8.967731232035847e-05, | |
| "loss": 1.8547, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.264257481648786, | |
| "grad_norm": 0.4999150037765503, | |
| "learning_rate": 8.944472712416447e-05, | |
| "loss": 1.44, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.26651609260304915, | |
| "grad_norm": 0.5740485191345215, | |
| "learning_rate": 8.9209859555586e-05, | |
| "loss": 1.2195, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.26877470355731226, | |
| "grad_norm": 0.5336365103721619, | |
| "learning_rate": 8.897272320465887e-05, | |
| "loss": 1.4808, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.27103331451157536, | |
| "grad_norm": 0.5398093461990356, | |
| "learning_rate": 8.873333179269635e-05, | |
| "loss": 1.4662, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2732919254658385, | |
| "grad_norm": 0.5843866467475891, | |
| "learning_rate": 8.849169917149531e-05, | |
| "loss": 1.6004, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.27555053642010163, | |
| "grad_norm": 0.5099990963935852, | |
| "learning_rate": 8.82478393225347e-05, | |
| "loss": 1.6265, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.27780914737436474, | |
| "grad_norm": 0.6847774386405945, | |
| "learning_rate": 8.800176635616657e-05, | |
| "loss": 1.7536, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2800677583286279, | |
| "grad_norm": 0.6712337732315063, | |
| "learning_rate": 8.775349451079948e-05, | |
| "loss": 1.3252, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.282326369282891, | |
| "grad_norm": 0.7772756814956665, | |
| "learning_rate": 8.750303815207486e-05, | |
| "loss": 1.6567, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2845849802371542, | |
| "grad_norm": 0.6570346355438232, | |
| "learning_rate": 8.725041177203554e-05, | |
| "loss": 1.2232, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2868435911914173, | |
| "grad_norm": 0.6152611374855042, | |
| "learning_rate": 8.699562998828738e-05, | |
| "loss": 1.42, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2891022021456804, | |
| "grad_norm": 0.5262449979782104, | |
| "learning_rate": 8.673870754315336e-05, | |
| "loss": 1.3769, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.29136081309994355, | |
| "grad_norm": 0.644555926322937, | |
| "learning_rate": 8.647965930282059e-05, | |
| "loss": 1.4069, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.29361942405420666, | |
| "grad_norm": 0.74615079164505, | |
| "learning_rate": 8.621850025648009e-05, | |
| "loss": 1.2758, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29587803500846976, | |
| "grad_norm": 1.079114317893982, | |
| "learning_rate": 8.59552455154595e-05, | |
| "loss": 1.7415, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.2981366459627329, | |
| "grad_norm": 0.522380530834198, | |
| "learning_rate": 8.56899103123487e-05, | |
| "loss": 1.3716, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.30039525691699603, | |
| "grad_norm": 0.6177389621734619, | |
| "learning_rate": 8.54225100001184e-05, | |
| "loss": 1.2745, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3026538678712592, | |
| "grad_norm": 0.5246083736419678, | |
| "learning_rate": 8.51530600512318e-05, | |
| "loss": 1.4917, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3049124788255223, | |
| "grad_norm": 1.4917463064193726, | |
| "learning_rate": 8.488157605674925e-05, | |
| "loss": 1.2877, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3071710897797854, | |
| "grad_norm": 0.5848981738090515, | |
| "learning_rate": 8.460807372542618e-05, | |
| "loss": 1.2985, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3094297007340486, | |
| "grad_norm": 0.8174607157707214, | |
| "learning_rate": 8.43325688828042e-05, | |
| "loss": 1.183, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 0.467540442943573, | |
| "learning_rate": 8.405507747029523e-05, | |
| "loss": 1.3237, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.31394692264257484, | |
| "grad_norm": 0.43935471773147583, | |
| "learning_rate": 8.377561554425922e-05, | |
| "loss": 1.1914, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.31620553359683795, | |
| "grad_norm": 0.5460615754127502, | |
| "learning_rate": 8.349419927507505e-05, | |
| "loss": 1.474, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31846414455110106, | |
| "grad_norm": 0.5589008331298828, | |
| "learning_rate": 8.321084494620488e-05, | |
| "loss": 1.5669, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3207227555053642, | |
| "grad_norm": 0.5428789854049683, | |
| "learning_rate": 8.292556895325194e-05, | |
| "loss": 1.2317, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.32298136645962733, | |
| "grad_norm": 0.5516197085380554, | |
| "learning_rate": 8.263838780301182e-05, | |
| "loss": 1.374, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.32523997741389044, | |
| "grad_norm": 0.6941014528274536, | |
| "learning_rate": 8.234931811251739e-05, | |
| "loss": 1.1839, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3274985883681536, | |
| "grad_norm": 0.7662340402603149, | |
| "learning_rate": 8.205837660807725e-05, | |
| "loss": 1.6056, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3297571993224167, | |
| "grad_norm": 0.6064298748970032, | |
| "learning_rate": 8.176558012430791e-05, | |
| "loss": 1.2736, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.33201581027667987, | |
| "grad_norm": 0.7128156423568726, | |
| "learning_rate": 8.147094560315977e-05, | |
| "loss": 1.156, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.334274421230943, | |
| "grad_norm": 0.5706272721290588, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 1.3529, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3365330321852061, | |
| "grad_norm": 0.4535754919052124, | |
| "learning_rate": 8.08762307473096e-05, | |
| "loss": 1.2878, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.33879164313946925, | |
| "grad_norm": 0.5142524242401123, | |
| "learning_rate": 8.057618482432399e-05, | |
| "loss": 1.4953, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.34105025409373235, | |
| "grad_norm": 0.4920322895050049, | |
| "learning_rate": 8.027436968540123e-05, | |
| "loss": 1.4318, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.34330886504799546, | |
| "grad_norm": 0.5119665861129761, | |
| "learning_rate": 7.997080279433402e-05, | |
| "loss": 1.2697, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3455674760022586, | |
| "grad_norm": 0.6550598740577698, | |
| "learning_rate": 7.966550171627592e-05, | |
| "loss": 1.3713, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.47998011112213135, | |
| "learning_rate": 7.9358484116725e-05, | |
| "loss": 1.3111, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3500846979107849, | |
| "grad_norm": 0.535916268825531, | |
| "learning_rate": 7.904976776050156e-05, | |
| "loss": 1.5861, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.352343308865048, | |
| "grad_norm": 0.4098435640335083, | |
| "learning_rate": 7.873937051072035e-05, | |
| "loss": 1.1488, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3546019198193111, | |
| "grad_norm": 0.6115076541900635, | |
| "learning_rate": 7.842731032775687e-05, | |
| "loss": 1.4248, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.35686053077357427, | |
| "grad_norm": 0.6232655644416809, | |
| "learning_rate": 7.81136052682082e-05, | |
| "loss": 1.3005, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3591191417278374, | |
| "grad_norm": 0.9366422295570374, | |
| "learning_rate": 7.779827348384813e-05, | |
| "loss": 1.6395, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3613777526821005, | |
| "grad_norm": 0.709642231464386, | |
| "learning_rate": 7.748133322057693e-05, | |
| "loss": 1.5304, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.5911598205566406, | |
| "learning_rate": 7.716280281736551e-05, | |
| "loss": 1.6141, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.36589497459062675, | |
| "grad_norm": 0.45050862431526184, | |
| "learning_rate": 7.68427007051944e-05, | |
| "loss": 1.3448, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3681535855448899, | |
| "grad_norm": 0.5454220771789551, | |
| "learning_rate": 7.652104540598712e-05, | |
| "loss": 1.2553, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.370412196499153, | |
| "grad_norm": 0.5132555365562439, | |
| "learning_rate": 7.619785553153864e-05, | |
| "loss": 1.4322, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.37267080745341613, | |
| "grad_norm": 0.48642510175704956, | |
| "learning_rate": 7.58731497824383e-05, | |
| "loss": 1.2141, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3749294184076793, | |
| "grad_norm": 0.6442435383796692, | |
| "learning_rate": 7.554694694698784e-05, | |
| "loss": 1.3274, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3771880293619424, | |
| "grad_norm": 0.6116953492164612, | |
| "learning_rate": 7.521926590011418e-05, | |
| "loss": 1.5006, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3794466403162055, | |
| "grad_norm": 3.9714138507843018, | |
| "learning_rate": 7.489012560227742e-05, | |
| "loss": 1.193, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.38170525127046867, | |
| "grad_norm": 0.5512414574623108, | |
| "learning_rate": 7.455954509837352e-05, | |
| "loss": 1.4702, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3839638622247318, | |
| "grad_norm": 0.5597122311592102, | |
| "learning_rate": 7.422754351663252e-05, | |
| "loss": 1.4636, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38622247317899494, | |
| "grad_norm": 0.4637336730957031, | |
| "learning_rate": 7.389414006751158e-05, | |
| "loss": 1.4661, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.38848108413325805, | |
| "grad_norm": 1.08772873878479, | |
| "learning_rate": 7.355935404258354e-05, | |
| "loss": 1.4758, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.39073969508752115, | |
| "grad_norm": 0.537563681602478, | |
| "learning_rate": 7.322320481342054e-05, | |
| "loss": 1.4603, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3929983060417843, | |
| "grad_norm": 0.6120302677154541, | |
| "learning_rate": 7.288571183047322e-05, | |
| "loss": 1.1336, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3952569169960474, | |
| "grad_norm": 0.5834792852401733, | |
| "learning_rate": 7.254689462194522e-05, | |
| "loss": 1.1359, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.39751552795031053, | |
| "grad_norm": 0.5482034683227539, | |
| "learning_rate": 7.220677279266327e-05, | |
| "loss": 1.416, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3997741389045737, | |
| "grad_norm": 0.6085582375526428, | |
| "learning_rate": 7.186536602294278e-05, | |
| "loss": 1.5104, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4020327498588368, | |
| "grad_norm": 0.5022709369659424, | |
| "learning_rate": 7.152269406744903e-05, | |
| "loss": 1.8106, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.40429136081309996, | |
| "grad_norm": 0.6673846244812012, | |
| "learning_rate": 7.117877675405427e-05, | |
| "loss": 1.4334, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.40654997176736307, | |
| "grad_norm": 0.6298004388809204, | |
| "learning_rate": 7.083363398269022e-05, | |
| "loss": 1.4593, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4088085827216262, | |
| "grad_norm": 0.5097036361694336, | |
| "learning_rate": 7.04872857241968e-05, | |
| "loss": 1.3969, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.41106719367588934, | |
| "grad_norm": 0.5180112719535828, | |
| "learning_rate": 7.013975201916648e-05, | |
| "loss": 1.7094, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.41332580463015245, | |
| "grad_norm": 0.803783655166626, | |
| "learning_rate": 6.979105297678462e-05, | |
| "loss": 1.4688, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 0.5442876219749451, | |
| "learning_rate": 6.944120877366604e-05, | |
| "loss": 1.5643, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4178430265386787, | |
| "grad_norm": 0.6098489165306091, | |
| "learning_rate": 6.909023965268746e-05, | |
| "loss": 1.3152, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4201016374929418, | |
| "grad_norm": 0.8846858143806458, | |
| "learning_rate": 6.873816592181617e-05, | |
| "loss": 1.1858, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.422360248447205, | |
| "grad_norm": 0.5681151747703552, | |
| "learning_rate": 6.838500795293505e-05, | |
| "loss": 1.3645, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4246188594014681, | |
| "grad_norm": 0.5397717356681824, | |
| "learning_rate": 6.803078618066378e-05, | |
| "loss": 1.4444, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4268774703557312, | |
| "grad_norm": 0.5300354361534119, | |
| "learning_rate": 6.767552110117631e-05, | |
| "loss": 1.4324, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.42913608130999437, | |
| "grad_norm": 0.4856337010860443, | |
| "learning_rate": 6.73192332710151e-05, | |
| "loss": 1.3462, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4313946922642575, | |
| "grad_norm": 0.6010419130325317, | |
| "learning_rate": 6.696194330590151e-05, | |
| "loss": 1.6966, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.43365330321852064, | |
| "grad_norm": 0.5775107145309448, | |
| "learning_rate": 6.660367187954304e-05, | |
| "loss": 1.4229, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.43591191417278374, | |
| "grad_norm": 0.6146912574768066, | |
| "learning_rate": 6.624443972243698e-05, | |
| "loss": 1.2737, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.43817052512704685, | |
| "grad_norm": 1.4060486555099487, | |
| "learning_rate": 6.5884267620671e-05, | |
| "loss": 1.2221, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.44042913608131, | |
| "grad_norm": 1.0087642669677734, | |
| "learning_rate": 6.552317641472026e-05, | |
| "loss": 1.504, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4426877470355731, | |
| "grad_norm": 0.4778720438480377, | |
| "learning_rate": 6.516118699824178e-05, | |
| "loss": 1.2763, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4449463579898362, | |
| "grad_norm": 0.5498337745666504, | |
| "learning_rate": 6.479832031686521e-05, | |
| "loss": 1.4852, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.4472049689440994, | |
| "grad_norm": 0.5609453916549683, | |
| "learning_rate": 6.443459736698105e-05, | |
| "loss": 1.6017, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4494635798983625, | |
| "grad_norm": 0.5456724166870117, | |
| "learning_rate": 6.407003919452564e-05, | |
| "loss": 1.2885, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.45172219085262566, | |
| "grad_norm": 0.7034818530082703, | |
| "learning_rate": 6.370466689376342e-05, | |
| "loss": 1.4856, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45398080180688877, | |
| "grad_norm": 0.5647463202476501, | |
| "learning_rate": 6.33385016060664e-05, | |
| "loss": 1.745, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4562394127611519, | |
| "grad_norm": 0.5447326302528381, | |
| "learning_rate": 6.297156451869082e-05, | |
| "loss": 1.407, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.45849802371541504, | |
| "grad_norm": 0.5139984488487244, | |
| "learning_rate": 6.260387686355121e-05, | |
| "loss": 1.3284, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.46075663466967814, | |
| "grad_norm": 0.44981297850608826, | |
| "learning_rate": 6.223545991599184e-05, | |
| "loss": 1.4727, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.46301524562394125, | |
| "grad_norm": 0.508630096912384, | |
| "learning_rate": 6.186633499355576e-05, | |
| "loss": 1.6505, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4652738565782044, | |
| "grad_norm": 0.5300900340080261, | |
| "learning_rate": 6.149652345475118e-05, | |
| "loss": 1.3005, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 0.8527065515518188, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 1.4615, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.4697910784867307, | |
| "grad_norm": 0.6106131672859192, | |
| "learning_rate": 6.075492615947823e-05, | |
| "loss": 1.3441, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.4720496894409938, | |
| "grad_norm": 0.4431338608264923, | |
| "learning_rate": 6.038318331371836e-05, | |
| "loss": 1.5369, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.4743083003952569, | |
| "grad_norm": 0.6845422983169556, | |
| "learning_rate": 6.001083967052408e-05, | |
| "loss": 1.6926, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.47656691134952006, | |
| "grad_norm": 0.48047032952308655, | |
| "learning_rate": 5.963791677464696e-05, | |
| "loss": 1.7314, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.47882552230378317, | |
| "grad_norm": 0.7335491180419922, | |
| "learning_rate": 5.9264436204355724e-05, | |
| "loss": 1.2773, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.4810841332580463, | |
| "grad_norm": 0.7190502285957336, | |
| "learning_rate": 5.889041957018745e-05, | |
| "loss": 1.2519, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.48334274421230944, | |
| "grad_norm": 0.547556459903717, | |
| "learning_rate": 5.85158885136973e-05, | |
| "loss": 1.1781, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.48560135516657255, | |
| "grad_norm": 0.6162766814231873, | |
| "learning_rate": 5.81408647062062e-05, | |
| "loss": 1.4605, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4878599661208357, | |
| "grad_norm": 0.7362288236618042, | |
| "learning_rate": 5.7765369847546916e-05, | |
| "loss": 1.5264, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4901185770750988, | |
| "grad_norm": 0.8551615476608276, | |
| "learning_rate": 5.7389425664808396e-05, | |
| "loss": 1.7111, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4923771880293619, | |
| "grad_norm": 0.5673311948776245, | |
| "learning_rate": 5.7013053911078677e-05, | |
| "loss": 1.4621, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.4946357989836251, | |
| "grad_norm": 0.4883188307285309, | |
| "learning_rate": 5.6636276364186105e-05, | |
| "loss": 1.4759, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.4968944099378882, | |
| "grad_norm": 0.5014625191688538, | |
| "learning_rate": 5.6259114825439275e-05, | |
| "loss": 1.5336, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4991530208921513, | |
| "grad_norm": 0.5976287722587585, | |
| "learning_rate": 5.588159111836553e-05, | |
| "loss": 1.3577, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5014116318464145, | |
| "grad_norm": 0.5646213889122009, | |
| "learning_rate": 5.550372708744815e-05, | |
| "loss": 1.4191, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5014116318464145, | |
| "eval_loss": 1.4101738929748535, | |
| "eval_runtime": 96.1988, | |
| "eval_samples_per_second": 7.755, | |
| "eval_steps_per_second": 0.977, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5036702428006776, | |
| "grad_norm": 0.48547643423080444, | |
| "learning_rate": 5.51255445968625e-05, | |
| "loss": 1.4118, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5059288537549407, | |
| "grad_norm": 0.7442102432250977, | |
| "learning_rate": 5.4747065529210736e-05, | |
| "loss": 1.3571, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5081874647092038, | |
| "grad_norm": 0.6088778376579285, | |
| "learning_rate": 5.436831178425582e-05, | |
| "loss": 1.3432, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.510446075663467, | |
| "grad_norm": 0.8041794896125793, | |
| "learning_rate": 5.3989305277654156e-05, | |
| "loss": 1.4741, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.51270468661773, | |
| "grad_norm": 0.47421059012413025, | |
| "learning_rate": 5.361006793968764e-05, | |
| "loss": 1.2545, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5149632975719932, | |
| "grad_norm": 0.6154189109802246, | |
| "learning_rate": 5.32306217139946e-05, | |
| "loss": 1.4776, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5172219085262564, | |
| "grad_norm": 0.5990487337112427, | |
| "learning_rate": 5.28509885563002e-05, | |
| "loss": 1.2825, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 0.6563726663589478, | |
| "learning_rate": 5.247119043314592e-05, | |
| "loss": 1.4692, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.5759412050247192, | |
| "learning_rate": 5.209124932061862e-05, | |
| "loss": 1.599, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5239977413890458, | |
| "grad_norm": 0.5122321248054504, | |
| "learning_rate": 5.1711187203078824e-05, | |
| "loss": 1.4495, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5262563523433089, | |
| "grad_norm": 0.5724780559539795, | |
| "learning_rate": 5.133102607188874e-05, | |
| "loss": 1.4343, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.528514963297572, | |
| "grad_norm": 0.6793734431266785, | |
| "learning_rate": 5.0950787924139764e-05, | |
| "loss": 1.2671, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5307735742518351, | |
| "grad_norm": 0.5128283500671387, | |
| "learning_rate": 5.057049476137967e-05, | |
| "loss": 1.1848, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5330321852060983, | |
| "grad_norm": 0.5823889970779419, | |
| "learning_rate": 5.0190168588339536e-05, | |
| "loss": 1.2724, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5352907961603613, | |
| "grad_norm": 0.41200748085975647, | |
| "learning_rate": 4.9809831411660476e-05, | |
| "loss": 1.3234, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5375494071146245, | |
| "grad_norm": 0.5784906148910522, | |
| "learning_rate": 4.942950523862033e-05, | |
| "loss": 1.4, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5398080180688877, | |
| "grad_norm": 0.9113706946372986, | |
| "learning_rate": 4.904921207586024e-05, | |
| "loss": 1.3569, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5420666290231507, | |
| "grad_norm": 0.5041924118995667, | |
| "learning_rate": 4.866897392811126e-05, | |
| "loss": 1.1772, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5443252399774139, | |
| "grad_norm": 0.7096860408782959, | |
| "learning_rate": 4.828881279692119e-05, | |
| "loss": 1.6993, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.546583850931677, | |
| "grad_norm": 0.4959520101547241, | |
| "learning_rate": 4.7908750679381384e-05, | |
| "loss": 1.6176, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5488424618859401, | |
| "grad_norm": 0.5700163841247559, | |
| "learning_rate": 4.752880956685407e-05, | |
| "loss": 1.0817, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5511010728402033, | |
| "grad_norm": 0.5470147132873535, | |
| "learning_rate": 4.7149011443699814e-05, | |
| "loss": 1.4854, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5533596837944664, | |
| "grad_norm": 0.7049497961997986, | |
| "learning_rate": 4.676937828600542e-05, | |
| "loss": 1.4135, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5556182947487295, | |
| "grad_norm": 0.7252426743507385, | |
| "learning_rate": 4.638993206031237e-05, | |
| "loss": 1.4154, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5578769057029926, | |
| "grad_norm": 0.575930118560791, | |
| "learning_rate": 4.601069472234584e-05, | |
| "loss": 1.0767, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5601355166572558, | |
| "grad_norm": 1.1477127075195312, | |
| "learning_rate": 4.56316882157442e-05, | |
| "loss": 1.332, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.562394127611519, | |
| "grad_norm": 0.5191376209259033, | |
| "learning_rate": 4.525293447078927e-05, | |
| "loss": 1.7038, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.564652738565782, | |
| "grad_norm": 0.554302990436554, | |
| "learning_rate": 4.4874455403137514e-05, | |
| "loss": 1.7356, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5669113495200452, | |
| "grad_norm": 1.1342941522598267, | |
| "learning_rate": 4.449627291255184e-05, | |
| "loss": 1.6412, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5691699604743083, | |
| "grad_norm": 0.5467344522476196, | |
| "learning_rate": 4.411840888163449e-05, | |
| "loss": 1.3031, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.4619009494781494, | |
| "learning_rate": 4.3740885174560736e-05, | |
| "loss": 1.4958, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5736871823828346, | |
| "grad_norm": 0.6281998157501221, | |
| "learning_rate": 4.336372363581391e-05, | |
| "loss": 1.4351, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5759457933370977, | |
| "grad_norm": 0.6288211345672607, | |
| "learning_rate": 4.298694608892134e-05, | |
| "loss": 1.7786, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5782044042913608, | |
| "grad_norm": 1.0841307640075684, | |
| "learning_rate": 4.2610574335191615e-05, | |
| "loss": 1.4286, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.5804630152456239, | |
| "grad_norm": 0.7780132293701172, | |
| "learning_rate": 4.2234630152453116e-05, | |
| "loss": 1.385, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.5827216261998871, | |
| "grad_norm": 0.49743711948394775, | |
| "learning_rate": 4.185913529379381e-05, | |
| "loss": 1.1707, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.5849802371541502, | |
| "grad_norm": 0.7499518394470215, | |
| "learning_rate": 4.1484111486302704e-05, | |
| "loss": 1.2754, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5872388481084133, | |
| "grad_norm": 0.5472561717033386, | |
| "learning_rate": 4.110958042981255e-05, | |
| "loss": 1.4632, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5894974590626765, | |
| "grad_norm": 0.6489593386650085, | |
| "learning_rate": 4.0735563795644294e-05, | |
| "loss": 1.5454, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5917560700169395, | |
| "grad_norm": 0.5738951563835144, | |
| "learning_rate": 4.0362083225353046e-05, | |
| "loss": 1.4356, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5940146809712027, | |
| "grad_norm": 0.4810039699077606, | |
| "learning_rate": 3.998916032947594e-05, | |
| "loss": 1.3287, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5962732919254659, | |
| "grad_norm": 0.6687359809875488, | |
| "learning_rate": 3.961681668628164e-05, | |
| "loss": 1.2772, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.598531902879729, | |
| "grad_norm": 0.5497633814811707, | |
| "learning_rate": 3.9245073840521765e-05, | |
| "loss": 1.4254, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6007905138339921, | |
| "grad_norm": 0.5534753203392029, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 1.3866, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6030491247882552, | |
| "grad_norm": 1.2714121341705322, | |
| "learning_rate": 3.850347654524883e-05, | |
| "loss": 1.4241, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6053077357425184, | |
| "grad_norm": 0.6475468277931213, | |
| "learning_rate": 3.8133665006444255e-05, | |
| "loss": 1.2423, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6075663466967814, | |
| "grad_norm": 0.6580026745796204, | |
| "learning_rate": 3.776454008400816e-05, | |
| "loss": 1.7248, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6098249576510446, | |
| "grad_norm": 0.47581014037132263, | |
| "learning_rate": 3.7396123136448824e-05, | |
| "loss": 1.5206, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6120835686053078, | |
| "grad_norm": 0.6214271783828735, | |
| "learning_rate": 3.70284354813092e-05, | |
| "loss": 1.3408, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6143421795595708, | |
| "grad_norm": 0.6450148224830627, | |
| "learning_rate": 3.666149839393361e-05, | |
| "loss": 1.6774, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.616600790513834, | |
| "grad_norm": 0.4316481947898865, | |
| "learning_rate": 3.629533310623658e-05, | |
| "loss": 1.0687, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6188594014680971, | |
| "grad_norm": 0.6752032041549683, | |
| "learning_rate": 3.592996080547438e-05, | |
| "loss": 1.7115, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6211180124223602, | |
| "grad_norm": 0.6112974882125854, | |
| "learning_rate": 3.556540263301896e-05, | |
| "loss": 1.5235, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 0.5716705322265625, | |
| "learning_rate": 3.520167968313479e-05, | |
| "loss": 1.4077, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6256352343308865, | |
| "grad_norm": 0.443218469619751, | |
| "learning_rate": 3.483881300175823e-05, | |
| "loss": 1.5901, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6278938452851497, | |
| "grad_norm": 0.6423050761222839, | |
| "learning_rate": 3.447682358527974e-05, | |
| "loss": 1.2472, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6301524562394127, | |
| "grad_norm": 0.6637226939201355, | |
| "learning_rate": 3.411573237932904e-05, | |
| "loss": 1.416, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6324110671936759, | |
| "grad_norm": 0.6266383528709412, | |
| "learning_rate": 3.3755560277563023e-05, | |
| "loss": 1.3596, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6346696781479391, | |
| "grad_norm": 0.4592457711696625, | |
| "learning_rate": 3.339632812045696e-05, | |
| "loss": 1.4775, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6369282891022021, | |
| "grad_norm": 0.4435870945453644, | |
| "learning_rate": 3.303805669409848e-05, | |
| "loss": 1.2606, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6391869000564653, | |
| "grad_norm": 0.8418070673942566, | |
| "learning_rate": 3.268076672898492e-05, | |
| "loss": 1.292, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6414455110107284, | |
| "grad_norm": 0.46599528193473816, | |
| "learning_rate": 3.2324478898823705e-05, | |
| "loss": 1.4636, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6437041219649915, | |
| "grad_norm": 0.5560009479522705, | |
| "learning_rate": 3.196921381933624e-05, | |
| "loss": 1.3654, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6459627329192547, | |
| "grad_norm": 0.5951889753341675, | |
| "learning_rate": 3.1614992047064945e-05, | |
| "loss": 1.2783, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6482213438735178, | |
| "grad_norm": 0.5692819952964783, | |
| "learning_rate": 3.126183407818384e-05, | |
| "loss": 1.3873, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6504799548277809, | |
| "grad_norm": 0.7618367671966553, | |
| "learning_rate": 3.090976034731257e-05, | |
| "loss": 1.3041, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.652738565782044, | |
| "grad_norm": 0.6365513205528259, | |
| "learning_rate": 3.055879122633397e-05, | |
| "loss": 1.3736, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6549971767363072, | |
| "grad_norm": 0.6634312272071838, | |
| "learning_rate": 3.020894702321539e-05, | |
| "loss": 1.5358, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6572557876905702, | |
| "grad_norm": 0.4850894510746002, | |
| "learning_rate": 2.9860247980833532e-05, | |
| "loss": 1.2746, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.6595143986448334, | |
| "grad_norm": 0.6311604976654053, | |
| "learning_rate": 2.951271427580321e-05, | |
| "loss": 1.3607, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.6617730095990966, | |
| "grad_norm": 0.6195625066757202, | |
| "learning_rate": 2.91663660173098e-05, | |
| "loss": 1.3412, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.6640316205533597, | |
| "grad_norm": 0.6908369660377502, | |
| "learning_rate": 2.882122324594575e-05, | |
| "loss": 1.5332, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.6662902315076228, | |
| "grad_norm": 0.9951730966567993, | |
| "learning_rate": 2.847730593255097e-05, | |
| "loss": 1.4562, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.668548842461886, | |
| "grad_norm": 0.4664572775363922, | |
| "learning_rate": 2.8134633977057235e-05, | |
| "loss": 1.608, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.6708074534161491, | |
| "grad_norm": 0.5350282788276672, | |
| "learning_rate": 2.779322720733673e-05, | |
| "loss": 1.6665, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.6730660643704122, | |
| "grad_norm": 0.4838089346885681, | |
| "learning_rate": 2.745310537805479e-05, | |
| "loss": 1.3432, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 0.5749621987342834, | |
| "learning_rate": 2.7114288169526793e-05, | |
| "loss": 1.5841, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.6775832862789385, | |
| "grad_norm": 0.44221848249435425, | |
| "learning_rate": 2.6776795186579468e-05, | |
| "loss": 1.2465, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6798418972332015, | |
| "grad_norm": 0.5502551794052124, | |
| "learning_rate": 2.6440645957416484e-05, | |
| "loss": 1.7499, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6821005081874647, | |
| "grad_norm": 0.51673823595047, | |
| "learning_rate": 2.610585993248843e-05, | |
| "loss": 1.4352, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.6843591191417279, | |
| "grad_norm": 0.5909664034843445, | |
| "learning_rate": 2.5772456483367497e-05, | |
| "loss": 1.2685, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6866177300959909, | |
| "grad_norm": 0.5207455158233643, | |
| "learning_rate": 2.5440454901626486e-05, | |
| "loss": 1.4422, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6888763410502541, | |
| "grad_norm": 0.6264783143997192, | |
| "learning_rate": 2.510987439772261e-05, | |
| "loss": 1.1306, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6911349520045172, | |
| "grad_norm": 0.5614155530929565, | |
| "learning_rate": 2.4780734099885833e-05, | |
| "loss": 1.9378, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.6933935629587803, | |
| "grad_norm": 0.5099676251411438, | |
| "learning_rate": 2.4453053053012187e-05, | |
| "loss": 1.4991, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.6968526244163513, | |
| "learning_rate": 2.4126850217561698e-05, | |
| "loss": 1.5211, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.6979107848673066, | |
| "grad_norm": 0.4605090320110321, | |
| "learning_rate": 2.3802144468461367e-05, | |
| "loss": 1.5865, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7001693958215698, | |
| "grad_norm": 0.8055828809738159, | |
| "learning_rate": 2.347895459401288e-05, | |
| "loss": 1.4216, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7024280067758328, | |
| "grad_norm": 0.5503789186477661, | |
| "learning_rate": 2.3157299294805613e-05, | |
| "loss": 1.378, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.704686617730096, | |
| "grad_norm": 0.6064462661743164, | |
| "learning_rate": 2.2837197182634483e-05, | |
| "loss": 1.6254, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7069452286843592, | |
| "grad_norm": 0.7875195741653442, | |
| "learning_rate": 2.2518666779423074e-05, | |
| "loss": 1.459, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7092038396386222, | |
| "grad_norm": 0.5678854584693909, | |
| "learning_rate": 2.2201726516151882e-05, | |
| "loss": 1.4943, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7114624505928854, | |
| "grad_norm": 0.6135299801826477, | |
| "learning_rate": 2.1886394731791816e-05, | |
| "loss": 1.5494, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7137210615471485, | |
| "grad_norm": 0.49695029854774475, | |
| "learning_rate": 2.157268967224314e-05, | |
| "loss": 1.4126, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7159796725014116, | |
| "grad_norm": 0.5455291271209717, | |
| "learning_rate": 2.126062948927966e-05, | |
| "loss": 1.5505, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7182382834556748, | |
| "grad_norm": 0.6816684603691101, | |
| "learning_rate": 2.0950232239498446e-05, | |
| "loss": 1.1783, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7204968944099379, | |
| "grad_norm": 0.548568069934845, | |
| "learning_rate": 2.064151588327501e-05, | |
| "loss": 1.5532, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.722755505364201, | |
| "grad_norm": 0.44020459055900574, | |
| "learning_rate": 2.0334498283724078e-05, | |
| "loss": 1.7366, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7250141163184641, | |
| "grad_norm": 0.5120103359222412, | |
| "learning_rate": 2.002919720566599e-05, | |
| "loss": 1.2849, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.1308592557907104, | |
| "learning_rate": 1.9725630314598782e-05, | |
| "loss": 1.3336, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7295313382269905, | |
| "grad_norm": 0.7465949058532715, | |
| "learning_rate": 1.9423815175676025e-05, | |
| "loss": 1.4263, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7317899491812535, | |
| "grad_norm": 0.6979694962501526, | |
| "learning_rate": 1.912376925269041e-05, | |
| "loss": 1.2292, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7340485601355167, | |
| "grad_norm": 0.6051463484764099, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 1.6008, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7363071710897798, | |
| "grad_norm": 0.5719790458679199, | |
| "learning_rate": 1.8529054396840234e-05, | |
| "loss": 1.5861, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7385657820440429, | |
| "grad_norm": 0.4965897798538208, | |
| "learning_rate": 1.8234419875692105e-05, | |
| "loss": 1.9165, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.740824392998306, | |
| "grad_norm": 0.6010226011276245, | |
| "learning_rate": 1.7941623391922772e-05, | |
| "loss": 1.2716, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7430830039525692, | |
| "grad_norm": 0.6560564637184143, | |
| "learning_rate": 1.7650681887482628e-05, | |
| "loss": 1.5902, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.7453416149068323, | |
| "grad_norm": 0.7512021660804749, | |
| "learning_rate": 1.7361612196988174e-05, | |
| "loss": 1.4901, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7476002258610954, | |
| "grad_norm": 0.5139881372451782, | |
| "learning_rate": 1.7074431046748075e-05, | |
| "loss": 1.366, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.7498588368153586, | |
| "grad_norm": 0.5443575978279114, | |
| "learning_rate": 1.678915505379513e-05, | |
| "loss": 1.2283, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.7521174477696216, | |
| "grad_norm": 0.5868397951126099, | |
| "learning_rate": 1.650580072492496e-05, | |
| "loss": 1.396, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.7543760587238848, | |
| "grad_norm": 0.5796262621879578, | |
| "learning_rate": 1.6224384455740788e-05, | |
| "loss": 1.522, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.756634669678148, | |
| "grad_norm": 0.5410485863685608, | |
| "learning_rate": 1.5944922529704777e-05, | |
| "loss": 1.3733, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.758893280632411, | |
| "grad_norm": 0.5935060977935791, | |
| "learning_rate": 1.5667431117195814e-05, | |
| "loss": 1.4833, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.7611518915866742, | |
| "grad_norm": 0.6086990833282471, | |
| "learning_rate": 1.539192627457382e-05, | |
| "loss": 1.2748, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.7634105025409373, | |
| "grad_norm": 0.9184845089912415, | |
| "learning_rate": 1.5118423943250771e-05, | |
| "loss": 1.5052, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.7656691134952005, | |
| "grad_norm": 0.5554947853088379, | |
| "learning_rate": 1.4846939948768218e-05, | |
| "loss": 1.8114, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.7679277244494636, | |
| "grad_norm": 0.5822046399116516, | |
| "learning_rate": 1.45774899998816e-05, | |
| "loss": 1.1491, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7701863354037267, | |
| "grad_norm": 0.5082180500030518, | |
| "learning_rate": 1.4310089687651301e-05, | |
| "loss": 1.2931, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.7724449463579899, | |
| "grad_norm": 0.7087529897689819, | |
| "learning_rate": 1.40447544845405e-05, | |
| "loss": 1.4946, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.7747035573122529, | |
| "grad_norm": 0.566061794757843, | |
| "learning_rate": 1.378149974351991e-05, | |
| "loss": 1.31, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.7769621682665161, | |
| "grad_norm": 0.48755723237991333, | |
| "learning_rate": 1.3520340697179406e-05, | |
| "loss": 1.5299, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 0.6160857677459717, | |
| "learning_rate": 1.3261292456846647e-05, | |
| "loss": 1.468, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7814793901750423, | |
| "grad_norm": 0.9709001779556274, | |
| "learning_rate": 1.3004370011712624e-05, | |
| "loss": 1.3858, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.7837380011293055, | |
| "grad_norm": 0.5348103046417236, | |
| "learning_rate": 1.2749588227964465e-05, | |
| "loss": 1.2159, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.7859966120835686, | |
| "grad_norm": 0.6455698609352112, | |
| "learning_rate": 1.2496961847925153e-05, | |
| "loss": 1.4438, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.7882552230378317, | |
| "grad_norm": 0.7605583071708679, | |
| "learning_rate": 1.2246505489200532e-05, | |
| "loss": 1.2351, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.7905138339920948, | |
| "grad_norm": 0.6536300182342529, | |
| "learning_rate": 1.1998233643833457e-05, | |
| "loss": 1.3756, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.792772444946358, | |
| "grad_norm": 0.5800183415412903, | |
| "learning_rate": 1.1752160677465286e-05, | |
| "loss": 1.3849, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.7950310559006211, | |
| "grad_norm": 0.5287428498268127, | |
| "learning_rate": 1.150830082850468e-05, | |
| "loss": 1.4647, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.7972896668548842, | |
| "grad_norm": 0.4410305619239807, | |
| "learning_rate": 1.126666820730366e-05, | |
| "loss": 1.3785, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.7995482778091474, | |
| "grad_norm": 0.4867897927761078, | |
| "learning_rate": 1.1027276795341135e-05, | |
| "loss": 1.4087, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8018068887634106, | |
| "grad_norm": 0.7169702053070068, | |
| "learning_rate": 1.0790140444414e-05, | |
| "loss": 1.1061, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8040654997176736, | |
| "grad_norm": 1.4617806673049927, | |
| "learning_rate": 1.0555272875835537e-05, | |
| "loss": 1.3367, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8063241106719368, | |
| "grad_norm": 0.37601158022880554, | |
| "learning_rate": 1.0322687679641523e-05, | |
| "loss": 1.3613, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8085827216261999, | |
| "grad_norm": 0.6595029830932617, | |
| "learning_rate": 1.0092398313803863e-05, | |
| "loss": 1.6219, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.810841332580463, | |
| "grad_norm": 0.5519759058952332, | |
| "learning_rate": 9.864418103451828e-06, | |
| "loss": 1.5668, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8130999435347261, | |
| "grad_norm": 0.5467275977134705, | |
| "learning_rate": 9.638760240101102e-06, | |
| "loss": 1.2981, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8153585544889893, | |
| "grad_norm": 0.5513888597488403, | |
| "learning_rate": 9.415437780890451e-06, | |
| "loss": 1.319, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8176171654432524, | |
| "grad_norm": 0.6388846039772034, | |
| "learning_rate": 9.194463647826223e-06, | |
| "loss": 1.6887, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8198757763975155, | |
| "grad_norm": 0.6353502869606018, | |
| "learning_rate": 8.975850627034604e-06, | |
| "loss": 1.3972, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8221343873517787, | |
| "grad_norm": 0.6129746437072754, | |
| "learning_rate": 8.759611368021831e-06, | |
| "loss": 1.3317, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8243929983060417, | |
| "grad_norm": 0.6045680046081543, | |
| "learning_rate": 8.545758382942232e-06, | |
| "loss": 1.2371, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8266516092603049, | |
| "grad_norm": 0.5147614479064941, | |
| "learning_rate": 8.334304045874247e-06, | |
| "loss": 1.498, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8289102202145681, | |
| "grad_norm": 0.41877540946006775, | |
| "learning_rate": 8.125260592104445e-06, | |
| "loss": 1.4604, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 0.5674257278442383, | |
| "learning_rate": 7.918640117419507e-06, | |
| "loss": 1.2845, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8334274421230943, | |
| "grad_norm": 0.7951464056968689, | |
| "learning_rate": 7.71445457740641e-06, | |
| "loss": 1.3372, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8356860530773574, | |
| "grad_norm": 0.6295720934867859, | |
| "learning_rate": 7.512715786760605e-06, | |
| "loss": 1.6711, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8379446640316206, | |
| "grad_norm": 0.5587788224220276, | |
| "learning_rate": 7.313435418602388e-06, | |
| "loss": 1.4419, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8402032749858837, | |
| "grad_norm": 0.8322708010673523, | |
| "learning_rate": 7.116625003801436e-06, | |
| "loss": 1.3559, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8424618859401468, | |
| "grad_norm": 0.6834149956703186, | |
| "learning_rate": 6.922295930309691e-06, | |
| "loss": 1.5552, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.84472049689441, | |
| "grad_norm": 0.7518057823181152, | |
| "learning_rate": 6.730459442502329e-06, | |
| "loss": 1.5805, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.846979107848673, | |
| "grad_norm": 0.5238639712333679, | |
| "learning_rate": 6.541126640527195e-06, | |
| "loss": 1.6161, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8492377188029362, | |
| "grad_norm": 0.5006189346313477, | |
| "learning_rate": 6.354308479662446e-06, | |
| "loss": 1.5392, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.8514963297571994, | |
| "grad_norm": 0.5511265397071838, | |
| "learning_rate": 6.170015769682741e-06, | |
| "loss": 1.6469, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.8537549407114624, | |
| "grad_norm": 0.5680528283119202, | |
| "learning_rate": 5.988259174233713e-06, | |
| "loss": 1.3312, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.8560135516657256, | |
| "grad_norm": 0.773024320602417, | |
| "learning_rate": 5.80904921021494e-06, | |
| "loss": 1.4814, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.8582721626199887, | |
| "grad_norm": 0.47784173488616943, | |
| "learning_rate": 5.6323962471714286e-06, | |
| "loss": 1.3431, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8605307735742518, | |
| "grad_norm": 0.5493645668029785, | |
| "learning_rate": 5.458310506693571e-06, | |
| "loss": 1.5127, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.862789384528515, | |
| "grad_norm": 0.5539724826812744, | |
| "learning_rate": 5.286802061825752e-06, | |
| "loss": 1.4401, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.8650479954827781, | |
| "grad_norm": 0.4766329824924469, | |
| "learning_rate": 5.117880836483452e-06, | |
| "loss": 1.5344, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.8673066064370413, | |
| "grad_norm": 0.46693822741508484, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 1.3009, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.7873904705047607, | |
| "learning_rate": 4.7878389909562285e-06, | |
| "loss": 1.4413, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8718238283455675, | |
| "grad_norm": 0.5490197539329529, | |
| "learning_rate": 4.62673746783317e-06, | |
| "loss": 1.2131, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.8740824392998306, | |
| "grad_norm": 0.4545706808567047, | |
| "learning_rate": 4.468261357254339e-06, | |
| "loss": 1.1735, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.8763410502540937, | |
| "grad_norm": 0.5005091428756714, | |
| "learning_rate": 4.312419829051173e-06, | |
| "loss": 1.5943, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.8785996612083569, | |
| "grad_norm": 0.546002984046936, | |
| "learning_rate": 4.15922190061146e-06, | |
| "loss": 1.3671, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.88085827216262, | |
| "grad_norm": 0.6055895686149597, | |
| "learning_rate": 4.008676436357539e-06, | |
| "loss": 1.4737, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 0.7526850700378418, | |
| "learning_rate": 3.86079214723345e-06, | |
| "loss": 1.6545, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.8853754940711462, | |
| "grad_norm": 0.6517372727394104, | |
| "learning_rate": 3.7155775902008526e-06, | |
| "loss": 1.2377, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.8876341050254094, | |
| "grad_norm": 0.590649425983429, | |
| "learning_rate": 3.5730411677439125e-06, | |
| "loss": 1.5282, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.8898927159796725, | |
| "grad_norm": 0.7884031534194946, | |
| "learning_rate": 3.4331911273830784e-06, | |
| "loss": 1.2815, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.8921513269339356, | |
| "grad_norm": 0.4688451886177063, | |
| "learning_rate": 3.2960355611979245e-06, | |
| "loss": 1.2642, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8944099378881988, | |
| "grad_norm": 0.5359352231025696, | |
| "learning_rate": 3.161582405358876e-06, | |
| "loss": 1.4675, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.8966685488424618, | |
| "grad_norm": 0.5633996725082397, | |
| "learning_rate": 3.029839439668003e-06, | |
| "loss": 1.5731, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.898927159796725, | |
| "grad_norm": 0.6047767400741577, | |
| "learning_rate": 2.9008142871088663e-06, | |
| "loss": 1.4189, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9011857707509882, | |
| "grad_norm": 0.47979751229286194, | |
| "learning_rate": 2.7745144134054433e-06, | |
| "loss": 1.3879, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.9034443817052513, | |
| "grad_norm": 0.4920281171798706, | |
| "learning_rate": 2.6509471265901477e-06, | |
| "loss": 1.423, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9057029926595144, | |
| "grad_norm": 0.5504844784736633, | |
| "learning_rate": 2.530119576580936e-06, | |
| "loss": 1.3404, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.9079616036137775, | |
| "grad_norm": 0.4826345145702362, | |
| "learning_rate": 2.412038754767626e-06, | |
| "loss": 1.2597, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.9102202145680407, | |
| "grad_norm": 0.7517184615135193, | |
| "learning_rate": 2.296711493607334e-06, | |
| "loss": 1.3037, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.9124788255223037, | |
| "grad_norm": 0.6532472372055054, | |
| "learning_rate": 2.1841444662291543e-06, | |
| "loss": 1.4865, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9147374364765669, | |
| "grad_norm": 0.5889557600021362, | |
| "learning_rate": 2.074344186048022e-06, | |
| "loss": 1.2332, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9169960474308301, | |
| "grad_norm": 0.4623706638813019, | |
| "learning_rate": 1.967317006387831e-06, | |
| "loss": 1.5374, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.9192546583850931, | |
| "grad_norm": 0.6362307071685791, | |
| "learning_rate": 1.863069120113814e-06, | |
| "loss": 1.0415, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.9215132693393563, | |
| "grad_norm": 0.5328913927078247, | |
| "learning_rate": 1.7616065592742038e-06, | |
| "loss": 1.3665, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.9237718802936195, | |
| "grad_norm": 0.61063551902771, | |
| "learning_rate": 1.6629351947512195e-06, | |
| "loss": 1.4575, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.9260304912478825, | |
| "grad_norm": 0.5676048994064331, | |
| "learning_rate": 1.567060735921344e-06, | |
| "loss": 1.5069, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9282891022021457, | |
| "grad_norm": 0.6710496544837952, | |
| "learning_rate": 1.4739887303249877e-06, | |
| "loss": 1.4816, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.9305477131564088, | |
| "grad_norm": 0.5965588092803955, | |
| "learning_rate": 1.383724563345451e-06, | |
| "loss": 1.6784, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.932806324110672, | |
| "grad_norm": 0.6151901483535767, | |
| "learning_rate": 1.2962734578973568e-06, | |
| "loss": 1.4018, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 0.6532612442970276, | |
| "learning_rate": 1.2116404741244203e-06, | |
| "loss": 1.415, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.9373235460191982, | |
| "grad_norm": 0.5151341557502747, | |
| "learning_rate": 1.1298305091066664e-06, | |
| "loss": 1.3436, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9395821569734614, | |
| "grad_norm": 0.6843656301498413, | |
| "learning_rate": 1.0508482965770505e-06, | |
| "loss": 1.5971, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.9418407679277244, | |
| "grad_norm": 0.5822469592094421, | |
| "learning_rate": 9.746984066475729e-07, | |
| "loss": 1.3158, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.9440993788819876, | |
| "grad_norm": 0.8204763531684875, | |
| "learning_rate": 9.013852455448335e-07, | |
| "loss": 1.2683, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.9463579898362507, | |
| "grad_norm": 0.9668586254119873, | |
| "learning_rate": 8.309130553550815e-07, | |
| "loss": 1.4966, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.9486166007905138, | |
| "grad_norm": 0.3911222815513611, | |
| "learning_rate": 7.63285913778733e-07, | |
| "loss": 1.3486, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.950875211744777, | |
| "grad_norm": 0.5275930166244507, | |
| "learning_rate": 6.985077338944657e-07, | |
| "loss": 1.044, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.9531338226990401, | |
| "grad_norm": 0.5202364921569824, | |
| "learning_rate": 6.365822639327723e-07, | |
| "loss": 1.262, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.9553924336533032, | |
| "grad_norm": 0.5621113181114197, | |
| "learning_rate": 5.775130870590783e-07, | |
| "loss": 1.3838, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.9576510446075663, | |
| "grad_norm": 0.5562776923179626, | |
| "learning_rate": 5.213036211664191e-07, | |
| "loss": 1.4821, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.9599096555618295, | |
| "grad_norm": 0.5257914066314697, | |
| "learning_rate": 4.6795711867766436e-07, | |
| "loss": 1.6874, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9621682665160926, | |
| "grad_norm": 0.7182251811027527, | |
| "learning_rate": 4.1747666635733597e-07, | |
| "loss": 1.3114, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.9644268774703557, | |
| "grad_norm": 0.7395084500312805, | |
| "learning_rate": 3.698651851329837e-07, | |
| "loss": 1.4869, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.9666854884246189, | |
| "grad_norm": 0.5292602777481079, | |
| "learning_rate": 3.251254299261874e-07, | |
| "loss": 1.6935, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.968944099378882, | |
| "grad_norm": 0.7703830599784851, | |
| "learning_rate": 2.8325998949314536e-07, | |
| "loss": 1.581, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.9712027103331451, | |
| "grad_norm": 0.4652191400527954, | |
| "learning_rate": 2.442712862748775e-07, | |
| "loss": 1.7907, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9734613212874083, | |
| "grad_norm": 0.7090250849723816, | |
| "learning_rate": 2.0816157625706545e-07, | |
| "loss": 1.3354, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.9757199322416714, | |
| "grad_norm": 0.6182876825332642, | |
| "learning_rate": 1.749329488395124e-07, | |
| "loss": 1.6426, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.9779785431959345, | |
| "grad_norm": 0.42393407225608826, | |
| "learning_rate": 1.4458732671523977e-07, | |
| "loss": 1.2865, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.9802371541501976, | |
| "grad_norm": 0.7182570695877075, | |
| "learning_rate": 1.1712646575922637e-07, | |
| "loss": 1.3558, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.9824957651044608, | |
| "grad_norm": 0.5384182929992676, | |
| "learning_rate": 9.255195492685609e-08, | |
| "loss": 1.4792, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9847543760587238, | |
| "grad_norm": 0.6862145662307739, | |
| "learning_rate": 7.086521616190279e-08, | |
| "loss": 1.4509, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 0.9429339170455933, | |
| "learning_rate": 5.2067504314323723e-08, | |
| "loss": 1.2235, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.9892715979672502, | |
| "grad_norm": 0.48626038432121277, | |
| "learning_rate": 3.6159907067601085e-08, | |
| "loss": 1.3711, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.9915302089215132, | |
| "grad_norm": 0.6334356665611267, | |
| "learning_rate": 2.3143344875831142e-08, | |
| "loss": 1.3192, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.9937888198757764, | |
| "grad_norm": 0.49483799934387207, | |
| "learning_rate": 1.3018570910466877e-08, | |
| "loss": 1.1924, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9960474308300395, | |
| "grad_norm": 0.5049090385437012, | |
| "learning_rate": 5.786171016708419e-09, | |
| "loss": 1.5949, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.9983060417843026, | |
| "grad_norm": 0.6922277808189392, | |
| "learning_rate": 1.446563679641244e-09, | |
| "loss": 1.4329, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.0005646527385659, | |
| "grad_norm": 0.656385064125061, | |
| "learning_rate": 0.0, | |
| "loss": 1.5656, | |
| "step": 443 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 443, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.630732306700042e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |