| { | |
| "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": 5.443678855895996, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.9061, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 5.82219934463501, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.9105, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 5.593341827392578, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.9029, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.499074459075928, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9133, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 5.513024806976318, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.8816, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 4.841058254241943, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.8616, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.233567714691162, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.8717, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.8331143856048584, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.8967, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.1296498775482178, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.8247, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.0691421031951904, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8194, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 1.935678482055664, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8173, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.1325912475585938, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8072, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 3.440077543258667, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.7925, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.7032241821289062, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.7964, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.4100658893585205, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.7774, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.6250312328338623, | |
| "learning_rate": 5e-06, | |
| "loss": 0.757, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.285292387008667, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.7718, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.9034889936447144, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7582, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.5559135675430298, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7192, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.4654297828674316, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7072, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.4360804557800293, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7075, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.46616530418396, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7336, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.2139849662780762, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.6851, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.1717193126678467, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.6955, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0088032484054565, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.6746, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 0.9678511023521423, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.6963, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 0.8797613382339478, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.6886, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.7748528718948364, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.6746, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.9056878089904785, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.6852, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.7696974277496338, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.6524, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.7831515073776245, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6993, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8536770343780518, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6839, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.734736442565918, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6732, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.7764032483100891, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.6645, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.6298977732658386, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6251, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6373680830001831, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6305, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.7237633466720581, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.6501, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.7214115262031555, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6473, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.5279441475868225, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6267, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.5429782271385193, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6353, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.6703981161117554, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6697, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.576714813709259, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6229, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.539608895778656, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6199, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.5225086212158203, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6642, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.5761452913284302, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6415, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.5128948092460632, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6498, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.4713912606239319, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6146, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.5027300715446472, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6323, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.4705498516559601, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6343, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.45382463932037354, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6362, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.5028925538063049, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6336, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.48574110865592957, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6389, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.5203267931938171, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6277, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.5508487224578857, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6166, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.46763473749160767, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6193, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.524610698223114, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6277, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.50461345911026, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.5966, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.499636173248291, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6435, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.4767281711101532, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.653, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.5102624297142029, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.621, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.44457393884658813, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6077, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.48511043190956116, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6279, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.5544952154159546, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6158, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.48609480261802673, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.5928, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.4879539906978607, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6023, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.5057247281074524, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6342, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.4872565269470215, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.5781, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.519126296043396, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.5852, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.43009886145591736, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.5965, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.640880286693573, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.5982, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.4887712001800537, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6152, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.5316668152809143, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.5856, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.4966796338558197, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6151, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.5114216804504395, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6251, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5646020770072937, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6015, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.46756139397621155, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.5881, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.47471338510513306, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.5908, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.5130011439323425, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6226, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.4613938629627228, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6182, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5221911072731018, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6044, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.4437088370323181, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6141, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.5470189452171326, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6298, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.4281877279281616, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.5925, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.45089638233184814, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.5883, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.5195220112800598, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6336, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.5485154390335083, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6144, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.5088512897491455, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.5838, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6507291197776794, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6099, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.5225642323493958, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6068, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.4308846592903137, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.5969, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.6871822476387024, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6134, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.435263067483902, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6013, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.4600350856781006, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.5696, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.492987722158432, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.5992, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.4614384174346924, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.5747, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.487991601228714, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.5639, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.45146673917770386, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.5801, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.5142231583595276, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.3947697579860687, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.5812, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.4276128113269806, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.5749, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.5193365216255188, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.583, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.4325678050518036, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6007, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.4741419851779938, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.5838, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.43793222308158875, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6134, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 0.9734936952590942, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.0593, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.44788506627082825, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5288, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.4729693531990051, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5346, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5325874090194702, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.6252, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.5393266081809998, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5493, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5109856128692627, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5092, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.4956968426704407, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5695, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.4617892801761627, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5158, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.470152884721756, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.56, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.40656527876853943, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.516, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.5289188027381897, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.6162, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.4350636601448059, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5425, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.4797382056713104, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5956, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.44591331481933594, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5188, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.3946237564086914, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5597, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.4344748258590698, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5617, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.39545729756355286, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5186, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.4443671405315399, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5828, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.3955932855606079, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5833, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.3952138423919678, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5227, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.3857017457485199, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5335, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.3564269244670868, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5236, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.39288026094436646, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5508, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.4181830883026123, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5738, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.38415423035621643, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5319, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.39088174700737, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5163, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.4108724892139435, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5885, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.3562670350074768, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5172, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.4148976802825928, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.5517, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.4055224061012268, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5682, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.41268885135650635, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5578, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.3958832621574402, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5493, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.3835631012916565, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5248, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.4725865125656128, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5647, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.41323113441467285, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5551, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.41332513093948364, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5003, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.4922088384628296, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5717, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.3990492820739746, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5622, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.4874456822872162, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5472, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.44028475880622864, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5737, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.4180607497692108, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5355, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.36899641156196594, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.4625, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.5210156440734863, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.6048, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.46662116050720215, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5512, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.447454035282135, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.598, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.5043928027153015, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5772, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.39012083411216736, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5113, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.4258650243282318, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5898, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.4456270635128021, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.5674, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.3997485041618347, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5014, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.4093655049800873, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5621, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.3858829736709595, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5435, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.44281816482543945, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.561, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.4816691279411316, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5567, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.47301939129829407, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5652, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.3875133693218231, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5504, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.46024438738822937, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5634, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.42805129289627075, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5231, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.3890428841114044, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5603, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.4287274181842804, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5752, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.39753690361976624, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.505, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.39962896704673767, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5811, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.35847893357276917, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.4964, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.46846652030944824, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5461, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.37737685441970825, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.499, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.3535378575325012, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5219, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.35122019052505493, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5257, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.3954353630542755, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5665, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.3725343346595764, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5784, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.3784255087375641, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5383, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.4283500015735626, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5505, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.375383198261261, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5287, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.36719298362731934, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5617, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.32004213333129883, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.543, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.3863466680049896, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5558, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.3306654989719391, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.52, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.3595239222049713, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5619, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.37461021542549133, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5339, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.3879006803035736, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5487, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.38633430004119873, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5803, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.3449064791202545, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5511, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.3583766222000122, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5569, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.36931082606315613, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5681, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.3926275372505188, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5502, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.3721172511577606, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5423, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.34798112511634827, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5215, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.4220713675022125, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5386, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.37677109241485596, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.555, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.3716229796409607, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5494, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.36740225553512573, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5344, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.35373231768608093, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5116, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4049929678440094, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.553, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.36175477504730225, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5179, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.35334157943725586, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5621, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.331590861082077, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5135, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.33945655822753906, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5618, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.37211182713508606, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5485, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.3631335496902466, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5486, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.3692123591899872, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.539, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.3526400625705719, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5622, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.3384586572647095, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5608, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.3885630667209625, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5608, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.3339373767375946, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5353, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.35258418321609497, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5135, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 0.9603828191757202, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0023, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.41805851459503174, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5273, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.4096415936946869, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5329, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.3491603136062622, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4646, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.37214240431785583, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5397, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.3790673315525055, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5099, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.36157622933387756, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.4986, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.36601653695106506, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4959, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.3199480473995209, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5041, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.38439592719078064, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5489, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.35765281319618225, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.4804, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.36830228567123413, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.5071, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.3654811680316925, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5151, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.3462638258934021, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.4932, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.3995060622692108, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5102, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.3391073942184448, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.4859, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.3244698941707611, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.4726, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.37653449177742004, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.5547, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.350153386592865, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.4398, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.4024869501590729, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5674, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.3475300669670105, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5286, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.3343484401702881, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.5165, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.3347722291946411, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.5018, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.32682865858078003, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.4595, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.39053240418434143, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.5387, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.36222565174102783, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.4928, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.35628780722618103, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.4951, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.32255733013153076, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.507, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.3425965905189514, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5368, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.3360115885734558, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5216, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.31881076097488403, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5189, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.31721314787864685, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.4975, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.3104105591773987, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.4797, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.33582985401153564, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5312, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.3309701085090637, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.4946, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.30556604266166687, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4906, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.34476539492607117, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5575, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.32860761880874634, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.4896, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.3144643008708954, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.4995, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.34862715005874634, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5136, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.33808404207229614, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.5136, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.31220996379852295, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.4955, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.3128294348716736, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.4899, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.32542213797569275, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5572, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.32575318217277527, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5349, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.31433993577957153, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.4838, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.33171674609184265, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5294, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.30407729744911194, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4757, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.36378738284111023, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5431, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.30705177783966064, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.4695, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.3292585015296936, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5697, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.28501492738723755, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.4864, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.3187987208366394, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.5343, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.3403649628162384, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5249, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.30438101291656494, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.4767, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3251408338546753, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5451, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.3420161008834839, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.5298, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.3219169080257416, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.5286, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.3166907727718353, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5298, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3281399607658386, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5225, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.29590266942977905, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.4878, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.2966673672199249, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5033, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.29974713921546936, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4901, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3066723942756653, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.4788, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.3056885004043579, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5197, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.323575884103775, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5127, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3323626220226288, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.4795, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3106381595134735, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5181, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.32390648126602173, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5662, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.27066949009895325, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4112, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.31843921542167664, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5574, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.30422160029411316, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.5385, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.2964281439781189, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.4983, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.31731534004211426, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.5389, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.29196590185165405, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.496, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3442181646823883, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5581, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.3149075508117676, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.512, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.3001596927642822, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.4941, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.28749582171440125, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.4924, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3162883520126343, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.4948, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.2985377311706543, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.5109, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.28927546739578247, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.4975, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.30359500646591187, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.4968, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.33133822679519653, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5401, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.30433112382888794, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5028, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.3015415072441101, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.4919, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.2975231111049652, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.4839, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.32519227266311646, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5256, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.2897155284881592, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.4884, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.3230243921279907, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5589, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.30294710397720337, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.4802, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.2914966642856598, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.497, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.27703312039375305, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.4772, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.29439136385917664, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5001, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.31278231739997864, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.523, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.2997344136238098, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.4714, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.3155505657196045, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5405, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.33159124851226807, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5023, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.2960556745529175, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.4542, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.2954247295856476, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.4865, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.2954687178134918, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5101, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.2963257133960724, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5357, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.2885425388813019, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.4818, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.29874712228775024, | |
| "learning_rate": 0.0, | |
| "loss": 0.5586, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 7.00883763227263e+17, | |
| "train_loss": 0.5766575982173284, | |
| "train_runtime": 33757.7333, | |
| "train_samples_per_second": 0.889, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "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": 7.00883763227263e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |