| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9856, | |
| "eval_steps": 500, | |
| "global_step": 468, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 2.451486825942993, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 1.5704, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 1.047212839126587, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 1.3347, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 1.2422709465026855, | |
| "learning_rate": 6.3829787234042555e-06, | |
| "loss": 1.4116, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 2.0192582607269287, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 1.5004, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.6840580701828003, | |
| "learning_rate": 1.0638297872340426e-05, | |
| "loss": 1.478, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 1.680579662322998, | |
| "learning_rate": 1.2765957446808511e-05, | |
| "loss": 1.3965, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 1.8126763105392456, | |
| "learning_rate": 1.4893617021276596e-05, | |
| "loss": 1.4944, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 2.288001537322998, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "loss": 1.4845, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 1.4367578029632568, | |
| "learning_rate": 1.9148936170212766e-05, | |
| "loss": 1.3447, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 1.576250433921814, | |
| "learning_rate": 2.1276595744680852e-05, | |
| "loss": 1.3918, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 1.6590297222137451, | |
| "learning_rate": 2.340425531914894e-05, | |
| "loss": 1.3442, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.338789701461792, | |
| "learning_rate": 2.5531914893617022e-05, | |
| "loss": 1.7413, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 3.2776601314544678, | |
| "learning_rate": 2.765957446808511e-05, | |
| "loss": 1.7825, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.463249921798706, | |
| "learning_rate": 2.9787234042553192e-05, | |
| "loss": 1.2387, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.5417563915252686, | |
| "learning_rate": 3.191489361702128e-05, | |
| "loss": 1.5182, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 1.448764681816101, | |
| "learning_rate": 3.4042553191489365e-05, | |
| "loss": 1.1844, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 1.5268467664718628, | |
| "learning_rate": 3.617021276595745e-05, | |
| "loss": 1.3816, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 2.561483383178711, | |
| "learning_rate": 3.829787234042553e-05, | |
| "loss": 1.3909, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 1.2603929042816162, | |
| "learning_rate": 4.0425531914893614e-05, | |
| "loss": 1.241, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.7591760158538818, | |
| "learning_rate": 4.2553191489361704e-05, | |
| "loss": 1.1528, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 1.6874586343765259, | |
| "learning_rate": 4.468085106382979e-05, | |
| "loss": 1.3256, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 1.5209052562713623, | |
| "learning_rate": 4.680851063829788e-05, | |
| "loss": 1.3031, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 0.5414145588874817, | |
| "learning_rate": 4.893617021276596e-05, | |
| "loss": 1.192, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 1.071704626083374, | |
| "learning_rate": 5.1063829787234044e-05, | |
| "loss": 1.0929, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.5227977633476257, | |
| "learning_rate": 5.319148936170213e-05, | |
| "loss": 1.1455, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 0.5635082721710205, | |
| "learning_rate": 5.531914893617022e-05, | |
| "loss": 1.2057, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 0.6073344945907593, | |
| "learning_rate": 5.744680851063831e-05, | |
| "loss": 1.2319, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.3395496606826782, | |
| "learning_rate": 5.9574468085106384e-05, | |
| "loss": 1.3086, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 0.5415111184120178, | |
| "learning_rate": 6.170212765957447e-05, | |
| "loss": 1.2653, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 0.7579400539398193, | |
| "learning_rate": 6.382978723404256e-05, | |
| "loss": 1.0565, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 0.4650651514530182, | |
| "learning_rate": 6.595744680851063e-05, | |
| "loss": 1.3408, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 0.31583133339881897, | |
| "learning_rate": 6.808510638297873e-05, | |
| "loss": 1.1341, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 0.3642672896385193, | |
| "learning_rate": 7.021276595744681e-05, | |
| "loss": 1.2263, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 0.40529701113700867, | |
| "learning_rate": 7.23404255319149e-05, | |
| "loss": 1.2229, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 0.2421422302722931, | |
| "learning_rate": 7.446808510638298e-05, | |
| "loss": 0.9803, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.3055451810359955, | |
| "learning_rate": 7.659574468085106e-05, | |
| "loss": 1.0436, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 0.26383817195892334, | |
| "learning_rate": 7.872340425531916e-05, | |
| "loss": 0.9907, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.3040568232536316, | |
| "learning_rate": 8.085106382978723e-05, | |
| "loss": 1.1131, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 0.20183025300502777, | |
| "learning_rate": 8.297872340425533e-05, | |
| "loss": 0.955, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.24510295689105988, | |
| "learning_rate": 8.510638297872341e-05, | |
| "loss": 1.0317, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 0.22165146470069885, | |
| "learning_rate": 8.723404255319149e-05, | |
| "loss": 0.9927, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.18096181750297546, | |
| "learning_rate": 8.936170212765958e-05, | |
| "loss": 0.9169, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 0.2374049425125122, | |
| "learning_rate": 9.148936170212766e-05, | |
| "loss": 0.9629, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.28989022970199585, | |
| "learning_rate": 9.361702127659576e-05, | |
| "loss": 1.0841, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.25799044966697693, | |
| "learning_rate": 9.574468085106384e-05, | |
| "loss": 1.0746, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.22019030153751373, | |
| "learning_rate": 9.787234042553192e-05, | |
| "loss": 0.9699, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 0.18696561455726624, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8615, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.2744862735271454, | |
| "learning_rate": 9.999860789001946e-05, | |
| "loss": 1.0763, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 0.19078461825847626, | |
| "learning_rate": 9.999443163759668e-05, | |
| "loss": 0.8894, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.19642512500286102, | |
| "learning_rate": 9.998747147528374e-05, | |
| "loss": 0.8941, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.1972423791885376, | |
| "learning_rate": 9.997772779065312e-05, | |
| "loss": 0.9058, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.19428911805152893, | |
| "learning_rate": 9.996520112627602e-05, | |
| "loss": 0.9221, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 0.20967088639736176, | |
| "learning_rate": 9.994989217969224e-05, | |
| "loss": 0.9024, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.19962789118289948, | |
| "learning_rate": 9.993180180337126e-05, | |
| "loss": 0.9491, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.1546671986579895, | |
| "learning_rate": 9.991093100466482e-05, | |
| "loss": 0.7734, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.15869870781898499, | |
| "learning_rate": 9.988728094575082e-05, | |
| "loss": 0.8626, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.2690005302429199, | |
| "learning_rate": 9.986085294356857e-05, | |
| "loss": 0.905, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.265375018119812, | |
| "learning_rate": 9.983164846974548e-05, | |
| "loss": 0.9522, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 0.1881999671459198, | |
| "learning_rate": 9.979966915051517e-05, | |
| "loss": 0.8813, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.1687404066324234, | |
| "learning_rate": 9.97649167666268e-05, | |
| "loss": 0.7873, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 0.16878949105739594, | |
| "learning_rate": 9.972739325324596e-05, | |
| "loss": 0.8416, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.1852271556854248, | |
| "learning_rate": 9.968710069984698e-05, | |
| "loss": 0.8725, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.20418129861354828, | |
| "learning_rate": 9.964404135009648e-05, | |
| "loss": 0.8924, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.17017708718776703, | |
| "learning_rate": 9.95982176017285e-05, | |
| "loss": 0.8465, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.200260192155838, | |
| "learning_rate": 9.95496320064109e-05, | |
| "loss": 0.897, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.19314530491828918, | |
| "learning_rate": 9.94982872696034e-05, | |
| "loss": 0.8314, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 0.16122443974018097, | |
| "learning_rate": 9.94441862504068e-05, | |
| "loss": 0.8258, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.1727280616760254, | |
| "learning_rate": 9.938733196140386e-05, | |
| "loss": 0.852, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.1809166669845581, | |
| "learning_rate": 9.932772756849153e-05, | |
| "loss": 0.7956, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.13888217508792877, | |
| "learning_rate": 9.926537639070457e-05, | |
| "loss": 0.7456, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 0.1737329661846161, | |
| "learning_rate": 9.92002819000309e-05, | |
| "loss": 0.7735, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.21496297419071198, | |
| "learning_rate": 9.91324477212181e-05, | |
| "loss": 0.8925, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 0.1533055156469345, | |
| "learning_rate": 9.906187763157168e-05, | |
| "loss": 0.7513, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.17420215904712677, | |
| "learning_rate": 9.898857556074468e-05, | |
| "loss": 0.82, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.17036575078964233, | |
| "learning_rate": 9.891254559051885e-05, | |
| "loss": 0.8445, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.19612562656402588, | |
| "learning_rate": 9.883379195457746e-05, | |
| "loss": 0.7657, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 0.18145954608917236, | |
| "learning_rate": 9.875231903826936e-05, | |
| "loss": 0.8554, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.19653621315956116, | |
| "learning_rate": 9.866813137836499e-05, | |
| "loss": 0.8247, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 0.15563753247261047, | |
| "learning_rate": 9.858123366280358e-05, | |
| "loss": 0.8427, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.18898862600326538, | |
| "learning_rate": 9.849163073043223e-05, | |
| "loss": 0.8435, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.2034827470779419, | |
| "learning_rate": 9.839932757073638e-05, | |
| "loss": 0.8755, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.18574374914169312, | |
| "learning_rate": 9.830432932356206e-05, | |
| "loss": 0.7832, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 0.1965630203485489, | |
| "learning_rate": 9.820664127882957e-05, | |
| "loss": 0.8227, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.1712363362312317, | |
| "learning_rate": 9.8106268876239e-05, | |
| "loss": 0.7677, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.1882326453924179, | |
| "learning_rate": 9.800321770496726e-05, | |
| "loss": 0.8642, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.21462184190750122, | |
| "learning_rate": 9.789749350335693e-05, | |
| "loss": 0.8289, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.16833196580410004, | |
| "learning_rate": 9.778910215859667e-05, | |
| "loss": 0.7127, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.23049375414848328, | |
| "learning_rate": 9.767804970639339e-05, | |
| "loss": 0.8281, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 0.22619804739952087, | |
| "learning_rate": 9.756434233063616e-05, | |
| "loss": 0.8528, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.17294900119304657, | |
| "learning_rate": 9.744798636305188e-05, | |
| "loss": 0.7163, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 0.21446748077869415, | |
| "learning_rate": 9.732898828285273e-05, | |
| "loss": 0.76, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.19360435009002686, | |
| "learning_rate": 9.72073547163753e-05, | |
| "loss": 0.7694, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.16693833470344543, | |
| "learning_rate": 9.708309243671165e-05, | |
| "loss": 0.7513, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.2053443044424057, | |
| "learning_rate": 9.695620836333219e-05, | |
| "loss": 0.7768, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.20644915103912354, | |
| "learning_rate": 9.68267095617003e-05, | |
| "loss": 0.7645, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.179234117269516, | |
| "learning_rate": 9.669460324287898e-05, | |
| "loss": 0.8299, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 0.18114909529685974, | |
| "learning_rate": 9.655989676312918e-05, | |
| "loss": 0.7326, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.19695445895195007, | |
| "learning_rate": 9.642259762350033e-05, | |
| "loss": 0.7831, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.17625164985656738, | |
| "learning_rate": 9.628271346941252e-05, | |
| "loss": 0.7511, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.2164350301027298, | |
| "learning_rate": 9.614025209023084e-05, | |
| "loss": 0.7882, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 0.18725153803825378, | |
| "learning_rate": 9.59952214188316e-05, | |
| "loss": 0.7671, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.18197496235370636, | |
| "learning_rate": 9.58476295311606e-05, | |
| "loss": 0.7698, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 0.19369439780712128, | |
| "learning_rate": 9.569748464578343e-05, | |
| "loss": 0.7534, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.2009640485048294, | |
| "learning_rate": 9.554479512342784e-05, | |
| "loss": 0.7415, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.19398680329322815, | |
| "learning_rate": 9.538956946651815e-05, | |
| "loss": 0.7429, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.20824606716632843, | |
| "learning_rate": 9.52318163187018e-05, | |
| "loss": 0.815, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 0.21556821465492249, | |
| "learning_rate": 9.507154446436805e-05, | |
| "loss": 0.7476, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.17891979217529297, | |
| "learning_rate": 9.490876282815884e-05, | |
| "loss": 0.7209, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 0.21326586604118347, | |
| "learning_rate": 9.474348047447177e-05, | |
| "loss": 0.7386, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.17727015912532806, | |
| "learning_rate": 9.457570660695541e-05, | |
| "loss": 0.7015, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.20510976016521454, | |
| "learning_rate": 9.440545056799677e-05, | |
| "loss": 0.7266, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.17708668112754822, | |
| "learning_rate": 9.423272183820108e-05, | |
| "loss": 0.7472, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 0.18563494086265564, | |
| "learning_rate": 9.405753003586395e-05, | |
| "loss": 0.7652, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.17155848443508148, | |
| "learning_rate": 9.387988491643558e-05, | |
| "loss": 0.7821, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.22160275280475616, | |
| "learning_rate": 9.369979637197775e-05, | |
| "loss": 0.7631, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.20093092322349548, | |
| "learning_rate": 9.351727443061283e-05, | |
| "loss": 0.7896, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.19560478627681732, | |
| "learning_rate": 9.333232925596552e-05, | |
| "loss": 0.7933, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.1992943435907364, | |
| "learning_rate": 9.314497114659671e-05, | |
| "loss": 0.7548, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 0.19478629529476166, | |
| "learning_rate": 9.295521053543019e-05, | |
| "loss": 0.7775, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.18368546664714813, | |
| "learning_rate": 9.276305798917159e-05, | |
| "loss": 0.7601, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 0.2044031322002411, | |
| "learning_rate": 9.256852420771998e-05, | |
| "loss": 0.7497, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.1834166944026947, | |
| "learning_rate": 9.237162002357214e-05, | |
| "loss": 0.7278, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.19447724521160126, | |
| "learning_rate": 9.217235640121926e-05, | |
| "loss": 0.72, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.1712876558303833, | |
| "learning_rate": 9.197074443653642e-05, | |
| "loss": 0.7404, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.20997127890586853, | |
| "learning_rate": 9.176679535616477e-05, | |
| "loss": 0.7825, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.1855696141719818, | |
| "learning_rate": 9.156052051688632e-05, | |
| "loss": 0.7547, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 0.20914626121520996, | |
| "learning_rate": 9.135193140499156e-05, | |
| "loss": 0.7972, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.21978524327278137, | |
| "learning_rate": 9.114103963563986e-05, | |
| "loss": 0.7325, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.2079222947359085, | |
| "learning_rate": 9.092785695221271e-05, | |
| "loss": 0.8084, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.1860748827457428, | |
| "learning_rate": 9.071239522565977e-05, | |
| "loss": 0.7359, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 0.1871631145477295, | |
| "learning_rate": 9.049466645383784e-05, | |
| "loss": 0.7138, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.19846683740615845, | |
| "learning_rate": 9.027468276084275e-05, | |
| "loss": 0.7176, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 0.2049422711133957, | |
| "learning_rate": 9.00524563963343e-05, | |
| "loss": 0.7595, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.17989754676818848, | |
| "learning_rate": 8.982799973485407e-05, | |
| "loss": 0.7445, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.32045578956604004, | |
| "learning_rate": 8.960132527513643e-05, | |
| "loss": 0.7214, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.2243393063545227, | |
| "learning_rate": 8.937244563941247e-05, | |
| "loss": 0.8011, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 0.20574022829532623, | |
| "learning_rate": 8.914137357270723e-05, | |
| "loss": 0.768, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.2292594015598297, | |
| "learning_rate": 8.890812194212988e-05, | |
| "loss": 0.7083, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 0.2046431005001068, | |
| "learning_rate": 8.867270373615734e-05, | |
| "loss": 0.77, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.22153380513191223, | |
| "learning_rate": 8.843513206391101e-05, | |
| "loss": 0.7262, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.214860737323761, | |
| "learning_rate": 8.81954201544267e-05, | |
| "loss": 0.8284, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.21554942429065704, | |
| "learning_rate": 8.795358135591811e-05, | |
| "loss": 0.7712, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 0.21957093477249146, | |
| "learning_rate": 8.77096291350334e-05, | |
| "loss": 0.7235, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.20161031186580658, | |
| "learning_rate": 8.746357707610545e-05, | |
| "loss": 0.7071, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.20437265932559967, | |
| "learning_rate": 8.721543888039533e-05, | |
| "loss": 0.7134, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.19950765371322632, | |
| "learning_rate": 8.69652283653294e-05, | |
| "loss": 0.6978, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.19119159877300262, | |
| "learning_rate": 8.671295946372988e-05, | |
| "loss": 0.7583, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.20924100279808044, | |
| "learning_rate": 8.645864622303898e-05, | |
| "loss": 0.7563, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 0.18858124315738678, | |
| "learning_rate": 8.620230280453673e-05, | |
| "loss": 0.7121, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.17950883507728577, | |
| "learning_rate": 8.594394348255238e-05, | |
| "loss": 0.6578, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 0.20606084167957306, | |
| "learning_rate": 8.568358264366957e-05, | |
| "loss": 0.71, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.18285425007343292, | |
| "learning_rate": 8.542123478592518e-05, | |
| "loss": 0.6941, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.20143665373325348, | |
| "learning_rate": 8.515691451800205e-05, | |
| "loss": 0.636, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.20506098866462708, | |
| "learning_rate": 8.489063655841551e-05, | |
| "loss": 0.7111, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.18759530782699585, | |
| "learning_rate": 8.462241573469379e-05, | |
| "loss": 0.733, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.22259001433849335, | |
| "learning_rate": 8.435226698255227e-05, | |
| "loss": 0.7405, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.3408335745334625, | |
| "learning_rate": 8.408020534506195e-05, | |
| "loss": 0.6648, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0064, | |
| "grad_norm": 0.18838848173618317, | |
| "learning_rate": 8.380624597181165e-05, | |
| "loss": 0.6529, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0128, | |
| "grad_norm": 0.20577505230903625, | |
| "learning_rate": 8.353040411806447e-05, | |
| "loss": 0.6837, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0192, | |
| "grad_norm": 0.20288875699043274, | |
| "learning_rate": 8.325269514390835e-05, | |
| "loss": 0.7313, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0256, | |
| "grad_norm": 0.2250303328037262, | |
| "learning_rate": 8.297313451340064e-05, | |
| "loss": 0.7675, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 0.21464917063713074, | |
| "learning_rate": 8.269173779370711e-05, | |
| "loss": 0.778, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0384, | |
| "grad_norm": 0.19386309385299683, | |
| "learning_rate": 8.240852065423506e-05, | |
| "loss": 0.7163, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0448, | |
| "grad_norm": 0.1796194165945053, | |
| "learning_rate": 8.21234988657607e-05, | |
| "loss": 0.6736, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0512, | |
| "grad_norm": 0.21359069645404816, | |
| "learning_rate": 8.183668829955111e-05, | |
| "loss": 0.7748, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0576, | |
| "grad_norm": 0.19759395718574524, | |
| "learning_rate": 8.154810492648037e-05, | |
| "loss": 0.746, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.19113849103450775, | |
| "learning_rate": 8.125776481614024e-05, | |
| "loss": 0.6629, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.0704, | |
| "grad_norm": 0.18363501131534576, | |
| "learning_rate": 8.096568413594533e-05, | |
| "loss": 0.674, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0768, | |
| "grad_norm": 0.20529989898204803, | |
| "learning_rate": 8.067187915023282e-05, | |
| "loss": 0.7051, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0832, | |
| "grad_norm": 0.176677405834198, | |
| "learning_rate": 8.037636621935685e-05, | |
| "loss": 0.7434, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0896, | |
| "grad_norm": 0.2223319411277771, | |
| "learning_rate": 8.007916179877741e-05, | |
| "loss": 0.6931, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 0.18287041783332825, | |
| "learning_rate": 7.978028243814415e-05, | |
| "loss": 0.7043, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1024, | |
| "grad_norm": 0.2017446905374527, | |
| "learning_rate": 7.947974478037468e-05, | |
| "loss": 0.7687, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1088, | |
| "grad_norm": 0.17362761497497559, | |
| "learning_rate": 7.91775655607279e-05, | |
| "loss": 0.7475, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1152, | |
| "grad_norm": 0.23194265365600586, | |
| "learning_rate": 7.887376160587215e-05, | |
| "loss": 0.7347, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1216, | |
| "grad_norm": 0.19501113891601562, | |
| "learning_rate": 7.85683498329481e-05, | |
| "loss": 0.7044, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 0.20629528164863586, | |
| "learning_rate": 7.826134724862687e-05, | |
| "loss": 0.7334, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1344, | |
| "grad_norm": 0.21323485672473907, | |
| "learning_rate": 7.795277094816291e-05, | |
| "loss": 0.7211, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1408, | |
| "grad_norm": 0.23606210947036743, | |
| "learning_rate": 7.764263811444215e-05, | |
| "loss": 0.6876, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1472, | |
| "grad_norm": 0.1839892566204071, | |
| "learning_rate": 7.733096601702507e-05, | |
| "loss": 0.7583, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1536, | |
| "grad_norm": 0.21783791482448578, | |
| "learning_rate": 7.70177720111852e-05, | |
| "loss": 0.6689, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.1926935911178589, | |
| "learning_rate": 7.67030735369426e-05, | |
| "loss": 0.76, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1663999999999999, | |
| "grad_norm": 0.18627937138080597, | |
| "learning_rate": 7.638688811809274e-05, | |
| "loss": 0.6958, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1728, | |
| "grad_norm": 0.20846357941627502, | |
| "learning_rate": 7.60692333612307e-05, | |
| "loss": 0.6813, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.1792, | |
| "grad_norm": 0.1936255395412445, | |
| "learning_rate": 7.575012695477076e-05, | |
| "loss": 0.6926, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1856, | |
| "grad_norm": 0.18793527781963348, | |
| "learning_rate": 7.54295866679615e-05, | |
| "loss": 0.6974, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 0.19964028894901276, | |
| "learning_rate": 7.510763034989617e-05, | |
| "loss": 0.7603, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.1984, | |
| "grad_norm": 0.20926351845264435, | |
| "learning_rate": 7.478427592851893e-05, | |
| "loss": 0.6682, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2048, | |
| "grad_norm": 0.22575032711029053, | |
| "learning_rate": 7.44595414096265e-05, | |
| "loss": 0.7939, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2112, | |
| "grad_norm": 0.19395670294761658, | |
| "learning_rate": 7.413344487586542e-05, | |
| "loss": 0.6991, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2176, | |
| "grad_norm": 0.19221967458724976, | |
| "learning_rate": 7.380600448572531e-05, | |
| "loss": 0.7237, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 0.2095392495393753, | |
| "learning_rate": 7.347723847252756e-05, | |
| "loss": 0.7686, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2304, | |
| "grad_norm": 0.19817109405994415, | |
| "learning_rate": 7.314716514341006e-05, | |
| "loss": 0.711, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2368000000000001, | |
| "grad_norm": 0.19115367531776428, | |
| "learning_rate": 7.28158028783079e-05, | |
| "loss": 0.6759, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2432, | |
| "grad_norm": 0.1996709704399109, | |
| "learning_rate": 7.248317012892969e-05, | |
| "loss": 0.7026, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2496, | |
| "grad_norm": 0.1884952187538147, | |
| "learning_rate": 7.214928541773027e-05, | |
| "loss": 0.7281, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 0.1934693157672882, | |
| "learning_rate": 7.181416733687919e-05, | |
| "loss": 0.6471, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2624, | |
| "grad_norm": 0.19305215775966644, | |
| "learning_rate": 7.147783454722545e-05, | |
| "loss": 0.7218, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.2688, | |
| "grad_norm": 0.17377148568630219, | |
| "learning_rate": 7.114030577725836e-05, | |
| "loss": 0.6985, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2752, | |
| "grad_norm": 0.19253912568092346, | |
| "learning_rate": 7.080159982206471e-05, | |
| "loss": 0.723, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2816, | |
| "grad_norm": 0.2046697437763214, | |
| "learning_rate": 7.046173554228213e-05, | |
| "loss": 0.6556, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.288, | |
| "grad_norm": 0.18565595149993896, | |
| "learning_rate": 7.012073186304886e-05, | |
| "loss": 0.7111, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.2944, | |
| "grad_norm": 0.1991768181324005, | |
| "learning_rate": 6.977860777294988e-05, | |
| "loss": 0.7236, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.3008, | |
| "grad_norm": 0.20029094815254211, | |
| "learning_rate": 6.943538232295964e-05, | |
| "loss": 0.6902, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.3072, | |
| "grad_norm": 0.20728713274002075, | |
| "learning_rate": 6.909107462538113e-05, | |
| "loss": 0.723, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3136, | |
| "grad_norm": 0.20966534316539764, | |
| "learning_rate": 6.874570385278161e-05, | |
| "loss": 0.7214, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.19986563920974731, | |
| "learning_rate": 6.839928923692504e-05, | |
| "loss": 0.7464, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3264, | |
| "grad_norm": 0.22603270411491394, | |
| "learning_rate": 6.805185006770124e-05, | |
| "loss": 0.7287, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3328, | |
| "grad_norm": 0.1987605094909668, | |
| "learning_rate": 6.770340569205158e-05, | |
| "loss": 0.7483, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3392, | |
| "grad_norm": 0.23768657445907593, | |
| "learning_rate": 6.735397551289178e-05, | |
| "loss": 0.685, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3456000000000001, | |
| "grad_norm": 0.19051390886306763, | |
| "learning_rate": 6.700357898803145e-05, | |
| "loss": 0.7484, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 0.1968153864145279, | |
| "learning_rate": 6.665223562909058e-05, | |
| "loss": 0.7095, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.3584, | |
| "grad_norm": 0.21399809420108795, | |
| "learning_rate": 6.629996500041299e-05, | |
| "loss": 0.6991, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3648, | |
| "grad_norm": 0.21878115832805634, | |
| "learning_rate": 6.594678671797703e-05, | |
| "loss": 0.7192, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.3712, | |
| "grad_norm": 0.19469957053661346, | |
| "learning_rate": 6.559272044830317e-05, | |
| "loss": 0.6861, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3776, | |
| "grad_norm": 0.23191341757774353, | |
| "learning_rate": 6.523778590735891e-05, | |
| "loss": 0.7194, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.384, | |
| "grad_norm": 0.21132110059261322, | |
| "learning_rate": 6.488200285946094e-05, | |
| "loss": 0.685, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.3904, | |
| "grad_norm": 0.22124455869197845, | |
| "learning_rate": 6.452539111617453e-05, | |
| "loss": 0.6685, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.3968, | |
| "grad_norm": 0.19542793929576874, | |
| "learning_rate": 6.416797053521038e-05, | |
| "loss": 0.6813, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.4032, | |
| "grad_norm": 0.2289544641971588, | |
| "learning_rate": 6.38097610193188e-05, | |
| "loss": 0.6714, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4096, | |
| "grad_norm": 0.2579493224620819, | |
| "learning_rate": 6.345078251518143e-05, | |
| "loss": 0.7159, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 0.19421054422855377, | |
| "learning_rate": 6.309105501230067e-05, | |
| "loss": 0.7322, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4224, | |
| "grad_norm": 0.20962068438529968, | |
| "learning_rate": 6.273059854188636e-05, | |
| "loss": 0.7126, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.4288, | |
| "grad_norm": 0.19890280067920685, | |
| "learning_rate": 6.236943317574054e-05, | |
| "loss": 0.7368, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4352, | |
| "grad_norm": 0.20143720507621765, | |
| "learning_rate": 6.200757902513962e-05, | |
| "loss": 0.7455, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4416, | |
| "grad_norm": 0.23619429767131805, | |
| "learning_rate": 6.164505623971457e-05, | |
| "loss": 0.7332, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 0.18844173848628998, | |
| "learning_rate": 6.128188500632892e-05, | |
| "loss": 0.726, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4544000000000001, | |
| "grad_norm": 0.2009935975074768, | |
| "learning_rate": 6.091808554795462e-05, | |
| "loss": 0.7047, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4607999999999999, | |
| "grad_norm": 0.21597343683242798, | |
| "learning_rate": 6.055367812254592e-05, | |
| "loss": 0.6953, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4672, | |
| "grad_norm": 0.21373218297958374, | |
| "learning_rate": 6.0188683021911396e-05, | |
| "loss": 0.7517, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4736, | |
| "grad_norm": 0.2007780373096466, | |
| "learning_rate": 5.9823120570583926e-05, | |
| "loss": 0.6925, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.24474382400512695, | |
| "learning_rate": 5.9457011124689023e-05, | |
| "loss": 0.6962, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.4864, | |
| "grad_norm": 0.20391108095645905, | |
| "learning_rate": 5.909037507081121e-05, | |
| "loss": 0.75, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4928, | |
| "grad_norm": 0.20164735615253448, | |
| "learning_rate": 5.8723232824858886e-05, | |
| "loss": 0.7141, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.4992, | |
| "grad_norm": 0.19647496938705444, | |
| "learning_rate": 5.835560483092743e-05, | |
| "loss": 0.7297, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5056, | |
| "grad_norm": 0.1943722367286682, | |
| "learning_rate": 5.798751156016085e-05, | |
| "loss": 0.7056, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 0.1920538693666458, | |
| "learning_rate": 5.761897350961175e-05, | |
| "loss": 0.7046, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.5184, | |
| "grad_norm": 0.2010536789894104, | |
| "learning_rate": 5.7250011201100095e-05, | |
| "loss": 0.7222, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5248, | |
| "grad_norm": 0.20179545879364014, | |
| "learning_rate": 5.688064518007036e-05, | |
| "loss": 0.6906, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5312000000000001, | |
| "grad_norm": 0.19752287864685059, | |
| "learning_rate": 5.6510896014447526e-05, | |
| "loss": 0.7016, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5375999999999999, | |
| "grad_norm": 0.20254461467266083, | |
| "learning_rate": 5.6140784293491725e-05, | |
| "loss": 0.6707, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 0.2074304223060608, | |
| "learning_rate": 5.577033062665179e-05, | |
| "loss": 0.7068, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5504, | |
| "grad_norm": 0.1754094362258911, | |
| "learning_rate": 5.53995556424176e-05, | |
| "loss": 0.6778, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5568, | |
| "grad_norm": 0.19732975959777832, | |
| "learning_rate": 5.50284799871714e-05, | |
| "loss": 0.7431, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5632000000000001, | |
| "grad_norm": 0.22642773389816284, | |
| "learning_rate": 5.465712432403812e-05, | |
| "loss": 0.7276, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5695999999999999, | |
| "grad_norm": 0.19384561479091644, | |
| "learning_rate": 5.428550933173476e-05, | |
| "loss": 0.6988, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 0.19731836020946503, | |
| "learning_rate": 5.391365570341893e-05, | |
| "loss": 0.6659, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.5824, | |
| "grad_norm": 0.1964057832956314, | |
| "learning_rate": 5.3541584145536475e-05, | |
| "loss": 0.6876, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.5888, | |
| "grad_norm": 0.21582961082458496, | |
| "learning_rate": 5.316931537666856e-05, | |
| "loss": 0.7048, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.5952, | |
| "grad_norm": 0.18675744533538818, | |
| "learning_rate": 5.279687012637799e-05, | |
| "loss": 0.657, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6016, | |
| "grad_norm": 0.21063970029354095, | |
| "learning_rate": 5.24242691340547e-05, | |
| "loss": 0.6786, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 0.20948725938796997, | |
| "learning_rate": 5.2051533147761155e-05, | |
| "loss": 0.712, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.6143999999999998, | |
| "grad_norm": 0.2167293280363083, | |
| "learning_rate": 5.167868292307678e-05, | |
| "loss": 0.668, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6208, | |
| "grad_norm": 0.19093649089336395, | |
| "learning_rate": 5.1305739221942364e-05, | |
| "loss": 0.7175, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6272, | |
| "grad_norm": 0.2434680163860321, | |
| "learning_rate": 5.093272281150383e-05, | |
| "loss": 0.6888, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6336, | |
| "grad_norm": 0.21365131437778473, | |
| "learning_rate": 5.05596544629559e-05, | |
| "loss": 0.6907, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.24530403316020966, | |
| "learning_rate": 5.018655495038541e-05, | |
| "loss": 0.6995, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.6463999999999999, | |
| "grad_norm": 0.2155607044696808, | |
| "learning_rate": 4.981344504961459e-05, | |
| "loss": 0.6894, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6528, | |
| "grad_norm": 0.193691223859787, | |
| "learning_rate": 4.944034553704412e-05, | |
| "loss": 0.6693, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6592, | |
| "grad_norm": 0.23936335742473602, | |
| "learning_rate": 4.9067277188496185e-05, | |
| "loss": 0.7161, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.6656, | |
| "grad_norm": 0.19866809248924255, | |
| "learning_rate": 4.869426077805765e-05, | |
| "loss": 0.6558, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6720000000000002, | |
| "grad_norm": 0.21519921720027924, | |
| "learning_rate": 4.8321317076923223e-05, | |
| "loss": 0.6388, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.6784, | |
| "grad_norm": 0.23403239250183105, | |
| "learning_rate": 4.794846685223884e-05, | |
| "loss": 0.723, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.6848, | |
| "grad_norm": 0.20778563618659973, | |
| "learning_rate": 4.757573086594529e-05, | |
| "loss": 0.7281, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.6912, | |
| "grad_norm": 0.2214445173740387, | |
| "learning_rate": 4.7203129873622045e-05, | |
| "loss": 0.6819, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.6976, | |
| "grad_norm": 0.23840227723121643, | |
| "learning_rate": 4.6830684623331446e-05, | |
| "loss": 0.6612, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 0.21336662769317627, | |
| "learning_rate": 4.645841585446356e-05, | |
| "loss": 0.6873, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.7104, | |
| "grad_norm": 0.20929306745529175, | |
| "learning_rate": 4.60863442965811e-05, | |
| "loss": 0.6937, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7168, | |
| "grad_norm": 0.2372131198644638, | |
| "learning_rate": 4.5714490668265245e-05, | |
| "loss": 0.6603, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.7231999999999998, | |
| "grad_norm": 0.19182223081588745, | |
| "learning_rate": 4.5342875675961885e-05, | |
| "loss": 0.7286, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7296, | |
| "grad_norm": 0.19139911234378815, | |
| "learning_rate": 4.497152001282861e-05, | |
| "loss": 0.6884, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 0.22471372783184052, | |
| "learning_rate": 4.460044435758241e-05, | |
| "loss": 0.6634, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.7424, | |
| "grad_norm": 0.1903676688671112, | |
| "learning_rate": 4.4229669373348226e-05, | |
| "loss": 0.6813, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7488000000000001, | |
| "grad_norm": 0.18955154716968536, | |
| "learning_rate": 4.3859215706508294e-05, | |
| "loss": 0.6851, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.7551999999999999, | |
| "grad_norm": 0.208612322807312, | |
| "learning_rate": 4.348910398555249e-05, | |
| "loss": 0.6685, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7616, | |
| "grad_norm": 0.19398434460163116, | |
| "learning_rate": 4.3119354819929655e-05, | |
| "loss": 0.6544, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.768, | |
| "grad_norm": 0.1975647658109665, | |
| "learning_rate": 4.27499887988999e-05, | |
| "loss": 0.6744, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7744, | |
| "grad_norm": 0.19207845628261566, | |
| "learning_rate": 4.2381026490388245e-05, | |
| "loss": 0.6928, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.7808000000000002, | |
| "grad_norm": 0.19270960986614227, | |
| "learning_rate": 4.201248843983918e-05, | |
| "loss": 0.6818, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.7872, | |
| "grad_norm": 0.20096321403980255, | |
| "learning_rate": 4.164439516907258e-05, | |
| "loss": 0.6712, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7936, | |
| "grad_norm": 0.20221710205078125, | |
| "learning_rate": 4.127676717514113e-05, | |
| "loss": 0.697, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.19503732025623322, | |
| "learning_rate": 4.0909624929188805e-05, | |
| "loss": 0.6939, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.8064, | |
| "grad_norm": 0.20334869623184204, | |
| "learning_rate": 4.0542988875310995e-05, | |
| "loss": 0.644, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.8128, | |
| "grad_norm": 0.20078979432582855, | |
| "learning_rate": 4.0176879429416086e-05, | |
| "loss": 0.7029, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8192, | |
| "grad_norm": 0.20540271699428558, | |
| "learning_rate": 3.981131697808862e-05, | |
| "loss": 0.6893, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.8256000000000001, | |
| "grad_norm": 0.21139805018901825, | |
| "learning_rate": 3.9446321877454094e-05, | |
| "loss": 0.7459, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.2058684229850769, | |
| "learning_rate": 3.90819144520454e-05, | |
| "loss": 0.6872, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8384, | |
| "grad_norm": 0.20217950642108917, | |
| "learning_rate": 3.8718114993671084e-05, | |
| "loss": 0.679, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.8448, | |
| "grad_norm": 0.1952074021100998, | |
| "learning_rate": 3.835494376028544e-05, | |
| "loss": 0.7296, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8512, | |
| "grad_norm": 0.20614856481552124, | |
| "learning_rate": 3.7992420974860384e-05, | |
| "loss": 0.6543, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8576000000000001, | |
| "grad_norm": 0.19841645658016205, | |
| "learning_rate": 3.7630566824259456e-05, | |
| "loss": 0.6866, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.8639999999999999, | |
| "grad_norm": 0.20447316765785217, | |
| "learning_rate": 3.726940145811363e-05, | |
| "loss": 0.6684, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.8704, | |
| "grad_norm": 0.2253882735967636, | |
| "learning_rate": 3.6908944987699345e-05, | |
| "loss": 0.6934, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.8768, | |
| "grad_norm": 0.20601052045822144, | |
| "learning_rate": 3.654921748481858e-05, | |
| "loss": 0.6936, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.8832, | |
| "grad_norm": 0.19950821995735168, | |
| "learning_rate": 3.6190238980681236e-05, | |
| "loss": 0.6804, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.8896, | |
| "grad_norm": 0.23119482398033142, | |
| "learning_rate": 3.583202946478963e-05, | |
| "loss": 0.682, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 0.19434580206871033, | |
| "learning_rate": 3.547460888382547e-05, | |
| "loss": 0.6698, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.9024, | |
| "grad_norm": 0.19283151626586914, | |
| "learning_rate": 3.511799714053907e-05, | |
| "loss": 0.6838, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9088, | |
| "grad_norm": 0.2131282240152359, | |
| "learning_rate": 3.47622140926411e-05, | |
| "loss": 0.7203, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.9152, | |
| "grad_norm": 0.20812910795211792, | |
| "learning_rate": 3.4407279551696846e-05, | |
| "loss": 0.697, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9216, | |
| "grad_norm": 0.19355922937393188, | |
| "learning_rate": 3.4053213282022984e-05, | |
| "loss": 0.7057, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 0.19805647432804108, | |
| "learning_rate": 3.370003499958703e-05, | |
| "loss": 0.7373, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.9344000000000001, | |
| "grad_norm": 0.21588417887687683, | |
| "learning_rate": 3.334776437090944e-05, | |
| "loss": 0.7187, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9407999999999999, | |
| "grad_norm": 0.17682795226573944, | |
| "learning_rate": 3.299642101196854e-05, | |
| "loss": 0.7133, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.9472, | |
| "grad_norm": 0.201612651348114, | |
| "learning_rate": 3.2646024487108215e-05, | |
| "loss": 0.6767, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9536, | |
| "grad_norm": 0.20109272003173828, | |
| "learning_rate": 3.2296594307948425e-05, | |
| "loss": 0.6907, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.1948445439338684, | |
| "learning_rate": 3.1948149932298774e-05, | |
| "loss": 0.7196, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.9664000000000001, | |
| "grad_norm": 0.20404204726219177, | |
| "learning_rate": 3.160071076307497e-05, | |
| "loss": 0.7123, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.9727999999999999, | |
| "grad_norm": 0.20908042788505554, | |
| "learning_rate": 3.125429614721842e-05, | |
| "loss": 0.6982, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.9792, | |
| "grad_norm": 0.21416354179382324, | |
| "learning_rate": 3.0908925374618895e-05, | |
| "loss": 0.6476, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.9856, | |
| "grad_norm": 0.20553742349147797, | |
| "learning_rate": 3.056461767704037e-05, | |
| "loss": 0.729, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 0.2184964120388031, | |
| "learning_rate": 3.0221392227050126e-05, | |
| "loss": 0.6833, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.9984, | |
| "grad_norm": 0.19415715336799622, | |
| "learning_rate": 2.987926813695116e-05, | |
| "loss": 0.75, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.4077971279621124, | |
| "learning_rate": 2.9538264457717878e-05, | |
| "loss": 0.7006, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 0.23124489188194275, | |
| "learning_rate": 2.9198400177935305e-05, | |
| "loss": 0.6592, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.0128, | |
| "grad_norm": 0.2515321969985962, | |
| "learning_rate": 2.885969422274165e-05, | |
| "loss": 0.6851, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.0192, | |
| "grad_norm": 0.21909447014331818, | |
| "learning_rate": 2.8522165452774557e-05, | |
| "loss": 0.7244, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.20850086212158203, | |
| "learning_rate": 2.8185832663120815e-05, | |
| "loss": 0.7144, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.2121347337961197, | |
| "learning_rate": 2.7850714582269722e-05, | |
| "loss": 0.6756, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.0384, | |
| "grad_norm": 0.22937630116939545, | |
| "learning_rate": 2.7516829871070292e-05, | |
| "loss": 0.6907, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.20349667966365814, | |
| "learning_rate": 2.7184197121692127e-05, | |
| "loss": 0.6523, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0512, | |
| "grad_norm": 0.19774508476257324, | |
| "learning_rate": 2.6852834856589947e-05, | |
| "loss": 0.6745, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.0576, | |
| "grad_norm": 0.1831101030111313, | |
| "learning_rate": 2.652276152747246e-05, | |
| "loss": 0.6973, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.20878373086452484, | |
| "learning_rate": 2.6193995514274705e-05, | |
| "loss": 0.67, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.0704, | |
| "grad_norm": 0.21983082592487335, | |
| "learning_rate": 2.5866555124134577e-05, | |
| "loss": 0.6811, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.0768, | |
| "grad_norm": 0.19198405742645264, | |
| "learning_rate": 2.5540458590373527e-05, | |
| "loss": 0.6282, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.19146208465099335, | |
| "learning_rate": 2.5215724071481072e-05, | |
| "loss": 0.7243, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.0896, | |
| "grad_norm": 0.22032876312732697, | |
| "learning_rate": 2.4892369650103836e-05, | |
| "loss": 0.6853, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 0.19633299112319946, | |
| "learning_rate": 2.457041333203852e-05, | |
| "loss": 0.6914, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.17531536519527435, | |
| "learning_rate": 2.4249873045229244e-05, | |
| "loss": 0.7256, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.1088, | |
| "grad_norm": 0.2074475884437561, | |
| "learning_rate": 2.3930766638769326e-05, | |
| "loss": 0.7376, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.1152, | |
| "grad_norm": 0.20909680426120758, | |
| "learning_rate": 2.3613111881907275e-05, | |
| "loss": 0.6866, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.19580084085464478, | |
| "learning_rate": 2.3296926463057396e-05, | |
| "loss": 0.7069, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.18059836328029633, | |
| "learning_rate": 2.2982227988814796e-05, | |
| "loss": 0.6449, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.1344, | |
| "grad_norm": 0.1941729485988617, | |
| "learning_rate": 2.2669033982974945e-05, | |
| "loss": 0.6295, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.2014259248971939, | |
| "learning_rate": 2.235736188555787e-05, | |
| "loss": 0.6776, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.1471999999999998, | |
| "grad_norm": 0.21426711976528168, | |
| "learning_rate": 2.2047229051837102e-05, | |
| "loss": 0.6988, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.1536, | |
| "grad_norm": 0.21534332633018494, | |
| "learning_rate": 2.173865275137314e-05, | |
| "loss": 0.6451, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.21011656522750854, | |
| "learning_rate": 2.143165016705192e-05, | |
| "loss": 0.6699, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.1664, | |
| "grad_norm": 0.21305830776691437, | |
| "learning_rate": 2.1126238394127868e-05, | |
| "loss": 0.6535, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.1728, | |
| "grad_norm": 0.22721970081329346, | |
| "learning_rate": 2.0822434439272122e-05, | |
| "loss": 0.6317, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.20172137022018433, | |
| "learning_rate": 2.052025521962534e-05, | |
| "loss": 0.6911, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.1856, | |
| "grad_norm": 0.19113659858703613, | |
| "learning_rate": 2.0219717561855855e-05, | |
| "loss": 0.7039, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 0.2135518491268158, | |
| "learning_rate": 1.992083820122259e-05, | |
| "loss": 0.684, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.19133660197257996, | |
| "learning_rate": 1.962363378064316e-05, | |
| "loss": 0.6835, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.2048, | |
| "grad_norm": 0.19742318987846375, | |
| "learning_rate": 1.9328120849767194e-05, | |
| "loss": 0.6411, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.2112, | |
| "grad_norm": 0.2134242057800293, | |
| "learning_rate": 1.903431586405468e-05, | |
| "loss": 0.6987, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.19351407885551453, | |
| "learning_rate": 1.8742235183859747e-05, | |
| "loss": 0.6387, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.19710959494113922, | |
| "learning_rate": 1.8451895073519643e-05, | |
| "loss": 0.6638, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.2304, | |
| "grad_norm": 0.18867988884449005, | |
| "learning_rate": 1.8163311700448898e-05, | |
| "loss": 0.6697, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.18838374316692352, | |
| "learning_rate": 1.7876501134239316e-05, | |
| "loss": 0.7276, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.2432, | |
| "grad_norm": 0.1833486109972, | |
| "learning_rate": 1.7591479345764973e-05, | |
| "loss": 0.6753, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.2496, | |
| "grad_norm": 0.23941516876220703, | |
| "learning_rate": 1.7308262206292897e-05, | |
| "loss": 0.6962, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.19911228120326996, | |
| "learning_rate": 1.7026865486599377e-05, | |
| "loss": 0.6764, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.2624, | |
| "grad_norm": 0.21614587306976318, | |
| "learning_rate": 1.6747304856091662e-05, | |
| "loss": 0.6218, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.2688, | |
| "grad_norm": 0.1931372880935669, | |
| "learning_rate": 1.6469595881935525e-05, | |
| "loss": 0.7262, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.19275271892547607, | |
| "learning_rate": 1.6193754028188364e-05, | |
| "loss": 0.7365, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.2816, | |
| "grad_norm": 0.19287358224391937, | |
| "learning_rate": 1.591979465493806e-05, | |
| "loss": 0.6888, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 0.18417416512966156, | |
| "learning_rate": 1.564773301744774e-05, | |
| "loss": 0.6533, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.1791851967573166, | |
| "learning_rate": 1.5377584265306223e-05, | |
| "loss": 0.6738, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.3008, | |
| "grad_norm": 0.1805466115474701, | |
| "learning_rate": 1.510936344158448e-05, | |
| "loss": 0.664, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.3072, | |
| "grad_norm": 0.1946578025817871, | |
| "learning_rate": 1.4843085481997959e-05, | |
| "loss": 0.7261, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.1858345866203308, | |
| "learning_rate": 1.457876521407484e-05, | |
| "loss": 0.6374, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.19666247069835663, | |
| "learning_rate": 1.431641735633044e-05, | |
| "loss": 0.6469, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.3264, | |
| "grad_norm": 0.16774997115135193, | |
| "learning_rate": 1.4056056517447635e-05, | |
| "loss": 0.6785, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.19413676857948303, | |
| "learning_rate": 1.3797697195463278e-05, | |
| "loss": 0.6797, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.3392, | |
| "grad_norm": 0.1797027438879013, | |
| "learning_rate": 1.3541353776961036e-05, | |
| "loss": 0.6968, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.3456, | |
| "grad_norm": 0.17702320218086243, | |
| "learning_rate": 1.3287040536270135e-05, | |
| "loss": 0.6696, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.19950032234191895, | |
| "learning_rate": 1.3034771634670601e-05, | |
| "loss": 0.6559, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.3584, | |
| "grad_norm": 0.19765135645866394, | |
| "learning_rate": 1.2784561119604682e-05, | |
| "loss": 0.6154, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.3648, | |
| "grad_norm": 0.21111543476581573, | |
| "learning_rate": 1.2536422923894564e-05, | |
| "loss": 0.6153, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.20043286681175232, | |
| "learning_rate": 1.2290370864966622e-05, | |
| "loss": 0.6988, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.3776, | |
| "grad_norm": 0.19548365473747253, | |
| "learning_rate": 1.2046418644081903e-05, | |
| "loss": 0.7292, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 0.1912689507007599, | |
| "learning_rate": 1.1804579845573289e-05, | |
| "loss": 0.7139, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.18108686804771423, | |
| "learning_rate": 1.1564867936088992e-05, | |
| "loss": 0.6878, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.3968, | |
| "grad_norm": 0.2090510129928589, | |
| "learning_rate": 1.1327296263842652e-05, | |
| "loss": 0.6882, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.4032, | |
| "grad_norm": 0.19618839025497437, | |
| "learning_rate": 1.1091878057870136e-05, | |
| "loss": 0.6613, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.18068960309028625, | |
| "learning_rate": 1.0858626427292795e-05, | |
| "loss": 0.6645, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.1936527043581009, | |
| "learning_rate": 1.0627554360587534e-05, | |
| "loss": 0.6942, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.4224, | |
| "grad_norm": 0.19242267310619354, | |
| "learning_rate": 1.0398674724863583e-05, | |
| "loss": 0.707, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.17682623863220215, | |
| "learning_rate": 1.0172000265145936e-05, | |
| "loss": 0.6429, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.4352, | |
| "grad_norm": 0.18140776455402374, | |
| "learning_rate": 9.94754360366571e-06, | |
| "loss": 0.7112, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.4416, | |
| "grad_norm": 0.1756039410829544, | |
| "learning_rate": 9.72531723915726e-06, | |
| "loss": 0.6616, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.19845832884311676, | |
| "learning_rate": 9.505333546162171e-06, | |
| "loss": 0.6774, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.4544, | |
| "grad_norm": 0.19389702379703522, | |
| "learning_rate": 9.287604774340236e-06, | |
| "loss": 0.7189, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.4608, | |
| "grad_norm": 0.20373724400997162, | |
| "learning_rate": 9.07214304778729e-06, | |
| "loss": 0.7034, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.1839911788702011, | |
| "learning_rate": 8.858960364360141e-06, | |
| "loss": 0.6394, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.4736000000000002, | |
| "grad_norm": 0.18362964689731598, | |
| "learning_rate": 8.648068595008457e-06, | |
| "loss": 0.673, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.18894684314727783, | |
| "learning_rate": 8.439479483113683e-06, | |
| "loss": 0.6781, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.1816522479057312, | |
| "learning_rate": 8.233204643835236e-06, | |
| "loss": 0.6894, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.4928, | |
| "grad_norm": 0.1781761348247528, | |
| "learning_rate": 8.029255563463589e-06, | |
| "loss": 0.681, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.4992, | |
| "grad_norm": 0.18166184425354004, | |
| "learning_rate": 7.827643598780749e-06, | |
| "loss": 0.6752, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.19368097186088562, | |
| "learning_rate": 7.628379976427868e-06, | |
| "loss": 0.6476, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.1752738654613495, | |
| "learning_rate": 7.431475792280018e-06, | |
| "loss": 0.6838, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.5183999999999997, | |
| "grad_norm": 0.1779792755842209, | |
| "learning_rate": 7.236942010828429e-06, | |
| "loss": 0.7104, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.18969357013702393, | |
| "learning_rate": 7.0447894645698175e-06, | |
| "loss": 0.6944, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.5312, | |
| "grad_norm": 0.18680952489376068, | |
| "learning_rate": 6.855028853403294e-06, | |
| "loss": 0.6931, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.5376, | |
| "grad_norm": 0.19200138747692108, | |
| "learning_rate": 6.667670744034499e-06, | |
| "loss": 0.732, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.1732376366853714, | |
| "learning_rate": 6.482725569387172e-06, | |
| "loss": 0.6958, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.5504, | |
| "grad_norm": 0.1787935346364975, | |
| "learning_rate": 6.300203628022272e-06, | |
| "loss": 0.6913, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.5568, | |
| "grad_norm": 0.18525420129299164, | |
| "learning_rate": 6.120115083564431e-06, | |
| "loss": 0.6651, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.19965125620365143, | |
| "learning_rate": 5.942469964136055e-06, | |
| "loss": 0.6701, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.5696, | |
| "grad_norm": 0.20355017483234406, | |
| "learning_rate": 5.767278161798911e-06, | |
| "loss": 0.6546, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 0.18207506835460663, | |
| "learning_rate": 5.5945494320032434e-06, | |
| "loss": 0.651, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.16914696991443634, | |
| "learning_rate": 5.424293393044611e-06, | |
| "loss": 0.6754, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.5888, | |
| "grad_norm": 0.17765933275222778, | |
| "learning_rate": 5.256519525528253e-06, | |
| "loss": 0.7015, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.5952, | |
| "grad_norm": 0.17559665441513062, | |
| "learning_rate": 5.091237171841173e-06, | |
| "loss": 0.7134, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.17291496694087982, | |
| "learning_rate": 4.928455535631959e-06, | |
| "loss": 0.7031, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.18128253519535065, | |
| "learning_rate": 4.768183681298211e-06, | |
| "loss": 0.7073, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.6144, | |
| "grad_norm": 0.1818084418773651, | |
| "learning_rate": 4.610430533481857e-06, | |
| "loss": 0.7378, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.17898201942443848, | |
| "learning_rate": 4.455204876572172e-06, | |
| "loss": 0.7126, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.6272, | |
| "grad_norm": 0.18443121016025543, | |
| "learning_rate": 4.302515354216574e-06, | |
| "loss": 0.6359, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.6336, | |
| "grad_norm": 0.16818998754024506, | |
| "learning_rate": 4.1523704688394176e-06, | |
| "loss": 0.6955, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.17661628127098083, | |
| "learning_rate": 4.004778581168412e-06, | |
| "loss": 0.6518, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.6464, | |
| "grad_norm": 0.17434482276439667, | |
| "learning_rate": 3.859747909769162e-06, | |
| "loss": 0.6369, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.6528, | |
| "grad_norm": 0.1895863562822342, | |
| "learning_rate": 3.7172865305874826e-06, | |
| "loss": 0.6902, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.17753319442272186, | |
| "learning_rate": 3.5774023764996723e-06, | |
| "loss": 0.7559, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.6656, | |
| "grad_norm": 0.16980338096618652, | |
| "learning_rate": 3.440103236870823e-06, | |
| "loss": 0.6977, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 0.17221416532993317, | |
| "learning_rate": 3.3053967571210378e-06, | |
| "loss": 0.673, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.18261651694774628, | |
| "learning_rate": 3.1732904382996976e-06, | |
| "loss": 0.6809, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.6848, | |
| "grad_norm": 0.18153472244739532, | |
| "learning_rate": 3.04379163666782e-06, | |
| "loss": 0.6428, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.6912000000000003, | |
| "grad_norm": 0.1869465857744217, | |
| "learning_rate": 2.916907563288357e-06, | |
| "loss": 0.6529, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.18409712612628937, | |
| "learning_rate": 2.792645283624712e-06, | |
| "loss": 0.6518, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.19606178998947144, | |
| "learning_rate": 2.671011717147276e-06, | |
| "loss": 0.6725, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.7104, | |
| "grad_norm": 0.17141272127628326, | |
| "learning_rate": 2.5520136369481198e-06, | |
| "loss": 0.6554, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.16549499332904816, | |
| "learning_rate": 2.4356576693638556e-06, | |
| "loss": 0.6629, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.7232, | |
| "grad_norm": 0.18200421333312988, | |
| "learning_rate": 2.321950293606623e-06, | |
| "loss": 0.7158, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.7296, | |
| "grad_norm": 0.1918717324733734, | |
| "learning_rate": 2.210897841403331e-06, | |
| "loss": 0.7156, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.16785910725593567, | |
| "learning_rate": 2.1025064966430696e-06, | |
| "loss": 0.6626, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.7424, | |
| "grad_norm": 0.17111513018608093, | |
| "learning_rate": 1.9967822950327454e-06, | |
| "loss": 0.6273, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.7488, | |
| "grad_norm": 0.18554243445396423, | |
| "learning_rate": 1.8937311237610166e-06, | |
| "loss": 0.6827, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.1815970093011856, | |
| "learning_rate": 1.793358721170435e-06, | |
| "loss": 0.6945, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.7616, | |
| "grad_norm": 0.18188947439193726, | |
| "learning_rate": 1.6956706764379438e-06, | |
| "loss": 0.6751, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 0.17477133870124817, | |
| "learning_rate": 1.6006724292636166e-06, | |
| "loss": 0.6617, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.17050299048423767, | |
| "learning_rate": 1.5083692695677832e-06, | |
| "loss": 0.6813, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.7808, | |
| "grad_norm": 0.18110086023807526, | |
| "learning_rate": 1.418766337196431e-06, | |
| "loss": 0.654, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.7872, | |
| "grad_norm": 0.17326004803180695, | |
| "learning_rate": 1.331868621635024e-06, | |
| "loss": 0.6791, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.16975191235542297, | |
| "learning_rate": 1.2476809617306407e-06, | |
| "loss": 0.6558, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.18156126141548157, | |
| "learning_rate": 1.166208045422551e-06, | |
| "loss": 0.6519, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.8064, | |
| "grad_norm": 0.17490343749523163, | |
| "learning_rate": 1.0874544094811423e-06, | |
| "loss": 0.6796, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.2003561407327652, | |
| "learning_rate": 1.0114244392553317e-06, | |
| "loss": 0.6837, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.8192, | |
| "grad_norm": 0.17029455304145813, | |
| "learning_rate": 9.381223684283291e-07, | |
| "loss": 0.619, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.8256, | |
| "grad_norm": 0.18762946128845215, | |
| "learning_rate": 8.675522787819023e-07, | |
| "loss": 0.7032, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.1766004115343094, | |
| "learning_rate": 7.9971809996911e-07, | |
| "loss": 0.6922, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.8384, | |
| "grad_norm": 0.17784501612186432, | |
| "learning_rate": 7.346236092954318e-07, | |
| "loss": 0.735, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.8448, | |
| "grad_norm": 0.1895914524793625, | |
| "learning_rate": 6.722724315084805e-07, | |
| "loss": 0.7254, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.17988507449626923, | |
| "learning_rate": 6.12668038596137e-07, | |
| "loss": 0.6898, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.8576, | |
| "grad_norm": 0.18528032302856445, | |
| "learning_rate": 5.558137495932037e-07, | |
| "loss": 0.6719, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 0.1748681515455246, | |
| "learning_rate": 5.017127303966085e-07, | |
| "loss": 0.6558, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.1760055422782898, | |
| "learning_rate": 4.5036799358910697e-07, | |
| "loss": 0.7183, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.8768000000000002, | |
| "grad_norm": 0.17728447914123535, | |
| "learning_rate": 4.0178239827151075e-07, | |
| "loss": 0.6566, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.8832, | |
| "grad_norm": 0.1707494556903839, | |
| "learning_rate": 3.5595864990352056e-07, | |
| "loss": 0.7113, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.17298643290996552, | |
| "learning_rate": 3.128993001530245e-07, | |
| "loss": 0.636, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.18352839350700378, | |
| "learning_rate": 2.72606746754045e-07, | |
| "loss": 0.6611, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.9024, | |
| "grad_norm": 0.1850634068250656, | |
| "learning_rate": 2.3508323337321226e-07, | |
| "loss": 0.6805, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.176837757229805, | |
| "learning_rate": 2.0033084948483105e-07, | |
| "loss": 0.7048, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.9152, | |
| "grad_norm": 0.1697586327791214, | |
| "learning_rate": 1.6835153025451245e-07, | |
| "loss": 0.6847, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.9215999999999998, | |
| "grad_norm": 0.16893458366394043, | |
| "learning_rate": 1.3914705643143789e-07, | |
| "loss": 0.6545, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.17263327538967133, | |
| "learning_rate": 1.1271905424918295e-07, | |
| "loss": 0.7085, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.9344, | |
| "grad_norm": 0.17428065836429596, | |
| "learning_rate": 8.906899533517865e-08, | |
| "loss": 0.6696, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.9408, | |
| "grad_norm": 0.17830806970596313, | |
| "learning_rate": 6.819819662874371e-08, | |
| "loss": 0.6562, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.16810065507888794, | |
| "learning_rate": 5.0107820307770947e-08, | |
| "loss": 0.7167, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.9536, | |
| "grad_norm": 0.18890230357646942, | |
| "learning_rate": 3.4798873723984605e-08, | |
| "loss": 0.7027, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.181377574801445, | |
| "learning_rate": 2.227220934688523e-08, | |
| "loss": 0.6166, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.16959308087825775, | |
| "learning_rate": 1.2528524716259871e-08, | |
| "loss": 0.7082, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.9728, | |
| "grad_norm": 0.16611914336681366, | |
| "learning_rate": 5.568362403318705e-09, | |
| "loss": 0.6582, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.9792, | |
| "grad_norm": 0.16966696083545685, | |
| "learning_rate": 1.3921099805302984e-09, | |
| "loss": 0.676, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.16860099136829376, | |
| "learning_rate": 0.0, | |
| "loss": 0.6525, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "step": 468, | |
| "total_flos": 6.618679011990045e+18, | |
| "train_loss": 0.7691191567314996, | |
| "train_runtime": 29137.5675, | |
| "train_samples_per_second": 1.03, | |
| "train_steps_per_second": 0.016 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 468, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.618679011990045e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |