| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9952, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 6.269186019897461, | |
| "learning_rate": 3.125e-07, | |
| "loss": 1.0433, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 5.88113260269165, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.0077, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 6.405559062957764, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.0115, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.136922836303711, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9998, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.04248046875, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 1.0162, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 5.5942230224609375, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 1.0018, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.604358196258545, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.9726, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 4.445125102996826, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9681, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.557436943054199, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.9057, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.421494960784912, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8879, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 2.1727352142333984, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8817, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.646085739135742, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.9022, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 4.079718589782715, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.9261, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.6011457443237305, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8719, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.366440773010254, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.855, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.50834059715271, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8304, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.122934103012085, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.7902, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.9464235305786133, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.861, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.5832701921463013, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.799, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.474358320236206, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7878, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.1883777379989624, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7172, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.2394137382507324, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7376, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.3630831241607666, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7649, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.2758921384811401, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7702, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0363675355911255, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7414, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.1511645317077637, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7156, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 1.3097772598266602, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7275, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.0546869039535522, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7264, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.7884509563446045, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7222, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.0245709419250488, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.7395, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.9652952551841736, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.7196, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8711326718330383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.714, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.9263397455215454, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.7318, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.935722291469574, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.6992, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.8034794926643372, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6659, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7833803296089172, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6875, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.8410481214523315, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.698, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.7238081097602844, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6955, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.7183172106742859, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6812, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.8978593945503235, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.702, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.8105177879333496, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.7173, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.7390109896659851, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6812, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.7539540529251099, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.687, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.902419924736023, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6998, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.8423213958740234, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6714, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.8182217478752136, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6751, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.7603351473808289, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6631, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.7307762503623962, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6804, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.76885986328125, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.69, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6869661808013916, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6906, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.6438692808151245, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6733, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.7209842801094055, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6967, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.7346424460411072, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6443, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.7073663473129272, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6859, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.5940643548965454, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.666, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.7106372117996216, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6793, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.7265703678131104, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6658, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.5696526169776917, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6394, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.6545392870903015, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6516, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7480618357658386, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6525, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.5585401654243469, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6514, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.538166344165802, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6435, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.6424883008003235, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6284, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.6507071852684021, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6461, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.5412493944168091, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6377, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.5966684818267822, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.671, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.6684296131134033, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6321, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.5935611128807068, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6631, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.6499678492546082, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6322, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.621015191078186, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6306, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.657986581325531, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6515, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6854347586631775, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6523, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.6243841052055359, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6377, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.6308167576789856, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6265, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.657517671585083, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6339, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.6046967506408691, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.633, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.5470812916755676, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.653, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.5507336854934692, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6315, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.6676499843597412, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6411, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5771962404251099, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6365, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.5356599688529968, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6323, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.6016016602516174, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6308, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.592422366142273, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6506, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.5272514224052429, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.637, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.649434506893158, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6459, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.5559666156768799, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6292, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.6839029788970947, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6732, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6171189546585083, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6458, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.541252076625824, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6382, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.5734552145004272, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6416, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.618876576423645, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6572, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5409361124038696, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.652, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.6081405282020569, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6423, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.5875936150550842, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6345, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.6266653537750244, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6438, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5741010904312134, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.618, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.5930486917495728, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.6686, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.6157588362693787, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6275, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.5080038905143738, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.6106, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6011462211608887, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6525, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.5805977582931519, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6289, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.7061988711357117, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6406, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.6194775700569153, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.6076, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5543082356452942, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6568, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.4446027278900146, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.1839, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.5336834788322449, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5265, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.6875979900360107, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.6195, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.6482763290405273, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5916, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.5494636297225952, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.6107, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.6209378242492676, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.574, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.6424750089645386, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5733, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5096311569213867, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5392, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.7727171182632446, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.6129, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.5170197486877441, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5385, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.6770474910736084, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5992, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5207650065422058, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5639, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.6146526336669922, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5616, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.6538062691688538, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.6215, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.5413742065429688, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5937, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6126624941825867, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5782, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.5789539813995361, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5605, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.544375479221344, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.6192, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.5714592337608337, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.572, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.5741793513298035, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.6344, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.46821609139442444, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5583, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.6880884766578674, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.6031, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.5428798198699951, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5596, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.5233510732650757, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5966, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.6209893822669983, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5785, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.6243672966957092, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.6048, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.6112650036811829, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.6424, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5332583785057068, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.4882, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.6027964353561401, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.6386, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.5317227244377136, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.55, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.5819597244262695, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5872, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6841590404510498, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5644, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.5862336754798889, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5939, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.5149160027503967, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5159, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.7106786966323853, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.6378, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5135704874992371, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5809, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.5317544341087341, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5644, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.556183934211731, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5674, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.5038918852806091, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5975, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.4887760877609253, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5691, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.5479956865310669, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5846, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.6294195055961609, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5742, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.5385369062423706, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.6475, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4947960674762726, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5551, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.506729006767273, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5611, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.564635157585144, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.6391, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.4597212076187134, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5015, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5252997279167175, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5833, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.5613046884536743, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.6352, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.49604713916778564, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5687, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.496060311794281, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5447, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4726210832595825, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5974, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.5450307726860046, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.6202, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.4789316952228546, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5773, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.48762327432632446, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5287, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.4888775646686554, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5825, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.4770006835460663, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5705, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.46331989765167236, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5636, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.5133760571479797, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.6215, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.48929649591445923, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5536, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.46117785573005676, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5631, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.5345510244369507, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.6346, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.4624938368797302, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5556, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.4785770773887634, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5818, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.4525819718837738, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.525, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.4890022277832031, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5892, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.5301545858383179, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5804, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.47773250937461853, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5514, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.4779603183269501, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5864, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.41888129711151123, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5485, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.5274120569229126, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.6079, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.4637974202632904, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5801, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.47610342502593994, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.6149, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.4770704507827759, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5747, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.45518961548805237, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5494, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.4997996985912323, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.6107, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.46732598543167114, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.6397, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.4510185718536377, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5693, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.5135465264320374, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5641, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.48011690378189087, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5919, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.46342650055885315, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5703, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.47104477882385254, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5699, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.4818515479564667, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.6142, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.42683175206184387, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5429, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.44734159111976624, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5619, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.47292158007621765, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5374, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.4907752275466919, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.602, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.43983301520347595, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5084, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.4868846535682678, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.6236, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.5178371071815491, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5434, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.4733627736568451, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5635, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4717698395252228, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.6102, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.4741760194301605, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5521, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.5023679137229919, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5657, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.528271496295929, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.6052, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4362201392650604, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5717, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.4392755329608917, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5916, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.4845729470252991, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.575, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.44493991136550903, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5483, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4418197274208069, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5633, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.4906143546104431, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.6315, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.4616820514202118, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5895, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.471195250749588, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5481, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4188118875026703, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5961, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.4321852922439575, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0621, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.44777438044548035, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.4849, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.56546950340271, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5795, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.45097553730010986, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4568, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.4803258180618286, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5951, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.44294473528862, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5379, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.5006292462348938, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5837, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.44567424058914185, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4792, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.4596433937549591, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5511, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.4522572457790375, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5351, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.44120803475379944, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5411, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.48423171043395996, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.5397, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.4955926537513733, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5167, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.4581254720687866, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.57, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.41280773282051086, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5103, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.4327993392944336, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5307, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.42601215839385986, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5087, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.4611905515193939, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.538, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.43144893646240234, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.575, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.6415021419525146, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.555, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.42779651284217834, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5422, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.46901237964630127, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.5854, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.382344514131546, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.4453, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.40246301889419556, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5092, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.4600931406021118, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.5872, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.4324595034122467, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.4654, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.4273560047149658, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5499, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.42452070116996765, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.5656, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.3744601011276245, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.4674, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.40509456396102905, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.523, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.4461797773838043, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5752, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4019639492034912, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.493, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.3972235321998596, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5161, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.4078989028930664, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5526, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.3724250793457031, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5245, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.38890838623046875, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.5309, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.4284942150115967, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5639, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.4045864939689636, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5412, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.4144849181175232, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.552, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3892313241958618, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5759, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.3391750454902649, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.4385, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.4045711159706116, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5846, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.3971869647502899, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5431, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.38077765703201294, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5354, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.394400030374527, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5272, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.3723466396331787, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.496, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.4069453775882721, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.565, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.3614712059497833, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4724, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.39318445324897766, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5631, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.4134117662906647, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.5371, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.39242568612098694, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5362, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4147869050502777, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5081, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.39689701795578003, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.5168, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.39359962940216064, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5103, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.4367649257183075, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.5391, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.39729082584381104, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5303, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.4398985803127289, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.5244, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.4133533537387848, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.535, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.38750264048576355, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.4918, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.40344518423080444, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5765, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.37539300322532654, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5492, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.3672214150428772, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5143, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.36059069633483887, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4981, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3750021755695343, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5736, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.3868529796600342, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5255, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.3691962659358978, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5642, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.36560919880867004, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5252, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3951523005962372, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5152, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.38338950276374817, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5598, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.36706873774528503, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4719, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.4016658067703247, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5953, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.38280174136161804, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.4747, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.37302348017692566, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5089, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.42362266778945923, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.5667, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.38069814443588257, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.548, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.4120860993862152, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5604, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.385165810585022, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.5363, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.39202815294265747, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.5391, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.37232810258865356, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5389, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3825536072254181, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5847, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.34925132989883423, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.4945, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.3492256700992584, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.5212, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.406555712223053, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5622, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.41345784068107605, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.57, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.38691234588623047, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5209, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.3592824339866638, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.4982, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.36266496777534485, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.5115, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.4156397581100464, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.477, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.3903021812438965, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.5434, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.34341850876808167, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.4994, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.35946303606033325, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5125, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3683798611164093, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5463, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.4044525921344757, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.5705, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.3820551037788391, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.569, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.3507567048072815, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.51, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3491106927394867, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.4716, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.38632506132125854, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5607, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.388088583946228, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5586, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.43631306290626526, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5705, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3891598582267761, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.4909, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.3835952579975128, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5154, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.3801961839199066, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5555, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.38301390409469604, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5316, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3846304416656494, | |
| "learning_rate": 0.0, | |
| "loss": 0.5283, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 314195249823744.0, | |
| "train_loss": 0.6126962422560422, | |
| "train_runtime": 29740.425, | |
| "train_samples_per_second": 1.009, | |
| "train_steps_per_second": 0.01 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 314195249823744.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |