| { | |
| "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": 6.131838798522949, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.9833, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 5.853671073913574, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.9681, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 6.1819634437561035, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.976, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.768021106719971, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9338, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 5.764317035675049, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.9222, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 4.990548133850098, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.9031, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.159934043884277, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.9052, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.8818681240081787, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.921, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.117541790008545, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.838, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.0639071464538574, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8353, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 1.9629915952682495, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8412, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.111118793487549, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.806, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 3.886829137802124, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.8445, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.862375259399414, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8326, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.491795539855957, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8159, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.792435884475708, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7965, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.2920074462890625, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.779, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 2.0210556983947754, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7785, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.6636329889297485, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7553, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.6136822700500488, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7712, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.368077278137207, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7064, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.1975337266921997, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7226, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.339500904083252, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7725, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.2119865417480469, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7173, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9528706073760986, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.6951, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.0267130136489868, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7063, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 0.808690071105957, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7245, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.8393774628639221, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7201, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.8567103743553162, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7011, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.7556109428405762, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.6766, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.6630239486694336, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6664, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8722522854804993, | |
| "learning_rate": 1e-05, | |
| "loss": 0.675, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.7277824878692627, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6737, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.659622311592102, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.6708, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.72553950548172, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6797, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6049491167068481, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.67, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.6887332797050476, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.6685, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.6767848134040833, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6634, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.6132428050041199, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6826, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.742757260799408, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6597, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.6259972453117371, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6406, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.5580882430076599, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6579, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.5948535799980164, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6373, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7255340218544006, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6355, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.49734628200531006, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6663, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.5538824200630188, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6638, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.7608892321586609, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.648, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6479229927062988, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6549, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.6154243350028992, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6575, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.5901063084602356, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6448, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.6039822697639465, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6708, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.5183172225952148, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6261, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.5812297463417053, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.65, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.5127038359642029, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6749, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.48981958627700806, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6335, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.5423652529716492, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6275, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.5755580067634583, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.677, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.524773895740509, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6527, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.48854127526283264, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6288, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.6286006569862366, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6747, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.4807024300098419, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6411, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.4745778441429138, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.632, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.620383620262146, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6433, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.563988208770752, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6609, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.5120728015899658, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.657, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.4716770350933075, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6279, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.5320466160774231, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6161, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.643553614616394, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6514, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.4202897846698761, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6218, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.723480224609375, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6807, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.5218155980110168, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6305, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.4688321053981781, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6234, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.5682986974716187, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6399, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.5219847559928894, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6155, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.4187687933444977, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6225, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.48372235894203186, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6201, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.5802701115608215, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.645, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.40434369444847107, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6171, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.4700346887111664, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6386, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5105234384536743, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6152, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.5602794885635376, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6634, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.5358520150184631, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6164, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.5795727968215942, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6433, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.5619863867759705, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.6257, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.4195852279663086, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6172, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.5690880417823792, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.5954, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.4654615521430969, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6478, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.4986039996147156, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6117, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.5075365900993347, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6673, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.5060116052627563, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.666, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.521640419960022, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6208, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.4387621283531189, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6364, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.5321857929229736, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6209, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.49556243419647217, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6148, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.4558052718639374, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.63, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.51129150390625, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.639, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.47831830382347107, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.6025, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.53434157371521, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6446, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.5062114000320435, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.6368, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.4139803349971771, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6095, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.5093127489089966, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.5982, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.5289537310600281, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6499, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.49252796173095703, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.6085, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.4343496859073639, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6182, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.1592433452606201, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.1477, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.4558708965778351, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5865, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.5685394406318665, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5767, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5245770812034607, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5751, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.49450862407684326, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5354, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5683928728103638, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5974, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.4599214494228363, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5465, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5915585160255432, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5876, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.5455578565597534, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.6087, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.5110136866569519, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.6009, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.5638250112533569, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.6016, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.476537823677063, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5742, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.4985634684562683, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5472, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.5420597791671753, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.6081, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.39116644859313965, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5605, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5126307010650635, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5806, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.45386430621147156, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5724, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.4348129630088806, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5648, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.451242595911026, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.579, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.3860410153865814, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5446, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.46640846133232117, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.6146, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.4218214154243469, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5192, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.6103668212890625, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.6102, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.47190529108047485, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5858, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.46076807379722595, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5825, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.5487145781517029, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5827, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.4499678909778595, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5646, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.4505446255207062, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5866, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.40683647990226746, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.523, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.4905054271221161, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.587, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.45940226316452026, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5842, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.4480723738670349, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5384, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.4832042455673218, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.6043, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.4512084126472473, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5981, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.46236684918403625, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5586, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.41917967796325684, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5108, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.46481794118881226, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.6092, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.43681877851486206, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5353, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.4951077401638031, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.6309, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.41214001178741455, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5337, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.4620398283004761, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.565, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.5365075469017029, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.626, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.45632481575012207, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.5707, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4703211784362793, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5905, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.36476150155067444, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5169, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.45678749680519104, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5861, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.4231281876564026, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5575, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.456800252199173, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5864, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.5350721478462219, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.6067, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.5283902883529663, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.6154, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.42444902658462524, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5159, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.48835116624832153, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5731, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.41756582260131836, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.5676, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.48666542768478394, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.6042, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.47318583726882935, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5934, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.3969041109085083, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5579, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.4392009675502777, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5505, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.4054500460624695, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5459, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.43042227625846863, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.6069, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5059496164321899, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.6048, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.4146595895290375, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5603, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.3976272642612457, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5354, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.4466738700866699, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5732, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.42778560519218445, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5468, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.379707932472229, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5267, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.4900387227535248, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5988, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.4723159968852997, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5843, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4003152549266815, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5585, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.38993823528289795, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5727, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.4870893359184265, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.605, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.4945925772190094, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5261, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.45155471563339233, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.6114, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.42216935753822327, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5623, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.38710522651672363, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5507, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.482290118932724, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5692, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.44206753373146057, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.5395, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.3879704773426056, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5805, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.37226903438568115, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5291, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.472578227519989, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.6054, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4130408465862274, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.566, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.392385870218277, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5762, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.406933069229126, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5618, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.38907113671302795, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5608, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.42358124256134033, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.6109, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.4014008641242981, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5966, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.38505733013153076, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5499, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.36976855993270874, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5698, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.3806343078613281, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5749, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.3810393214225769, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5963, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.38945069909095764, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.582, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.3987008035182953, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5477, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.38081228733062744, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.5817, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.3651650846004486, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5571, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.41797518730163574, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.601, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.40681710839271545, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5456, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.39941880106925964, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5652, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.4326709806919098, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5798, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.36714649200439453, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.6079, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.3356380760669708, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5534, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.3741372525691986, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5784, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.3816111087799072, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5549, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.4057197868824005, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.568, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.3479039967060089, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5511, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4004793167114258, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.6061, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.0402474403381348, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0123, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.41367122530937195, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5117, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.4018729031085968, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5235, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.45626595616340637, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.5914, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.39450162649154663, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5212, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.377545565366745, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5818, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.40043240785598755, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5042, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.4060075283050537, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.5457, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.41231614351272583, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5213, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.34756186604499817, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.523, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.36289355158805847, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5049, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.3698144257068634, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.537, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.39785948395729065, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5769, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.3399403989315033, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.4696, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.40045249462127686, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5951, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.3397459089756012, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.504, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.35173940658569336, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5393, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.36121654510498047, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.5251, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.341952919960022, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.4888, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.37249886989593506, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5503, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.36582255363464355, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5438, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.3314046561717987, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.535, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.37044402956962585, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.5697, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.34775087237358093, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.4777, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.3687901794910431, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.5504, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.3211866617202759, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.513, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.35890623927116394, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5659, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3151806890964508, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.4887, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.33889836072921753, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5496, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.3221040368080139, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5466, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.34383803606033325, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5201, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3385244607925415, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.5298, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.3405994176864624, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5097, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.32173824310302734, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5137, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.3235268294811249, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5115, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.3248879313468933, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.523, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.36188867688179016, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5563, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.39015740156173706, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5494, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.31805044412612915, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.5143, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3444110155105591, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5621, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.3224121928215027, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.4919, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.35199204087257385, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5601, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.3403403162956238, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5454, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.37926703691482544, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5478, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.3948769271373749, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5291, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.29835790395736694, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.5027, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.3501754105091095, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5728, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.30872753262519836, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4733, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.33777865767478943, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5625, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.3258240222930908, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.5068, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.3382646143436432, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5169, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.303226500749588, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5221, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.3068852722644806, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.4866, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.3425780236721039, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5587, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.32032689452171326, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.5227, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3151043951511383, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5385, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.3404954671859741, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.5179, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.3317325711250305, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.5278, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.3379656672477722, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5236, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3135997951030731, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5345, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.3309415578842163, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5594, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.3306780755519867, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5496, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.33503568172454834, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.5066, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.33311519026756287, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5255, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.3359110355377197, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5903, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.3060072660446167, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.518, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3306880295276642, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5346, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3175564408302307, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5211, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.3217380940914154, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5688, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.3017677068710327, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.5008, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.32683315873146057, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5231, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3304096460342407, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.4972, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.3331338167190552, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5311, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.36494651436805725, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.555, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.3253585994243622, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.496, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3310704827308655, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5523, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.32599544525146484, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.5155, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.31990212202072144, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.5416, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.30479979515075684, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5063, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.32286277413368225, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5139, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.3293840289115906, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.541, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.31361597776412964, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.514, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.32490071654319763, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5423, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3236123323440552, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5327, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.33006975054740906, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5237, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.3304724395275116, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5319, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.32301440834999084, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.5057, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.31884756684303284, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5244, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.31779152154922485, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.5257, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.3324771225452423, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5708, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.3106856346130371, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5108, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.31627193093299866, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5273, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.3130391836166382, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.5499, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.3078140616416931, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.4995, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.3132491409778595, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.5279, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3202137351036072, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.502, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.32776516675949097, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5358, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.3089218735694885, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5492, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.30734020471572876, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5316, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.41417551040649414, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5547, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.32749590277671814, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5429, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.3043902516365051, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5131, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.32178109884262085, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5376, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3101270794868469, | |
| "learning_rate": 0.0, | |
| "loss": 0.5542, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 443026603933696.0, | |
| "train_loss": 0.6015871439415675, | |
| "train_runtime": 34348.8651, | |
| "train_samples_per_second": 0.873, | |
| "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": 443026603933696.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |