| { | |
| "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": 5.830983638763428, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.9551, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 5.870426654815674, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.9523, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 5.710000514984131, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.9228, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.736249923706055, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9472, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 5.407976150512695, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.9213, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 5.176140308380127, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.9449, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.032590866088867, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.8826, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.869014263153076, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.8546, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.0670173168182373, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.8725, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.0075252056121826, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8611, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 1.888549566268921, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8235, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.298048973083496, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8324, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 3.7331154346466064, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.8225, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.683380365371704, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8178, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.4067821502685547, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8126, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.6179301738739014, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8143, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.067991256713867, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.7502, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.9287723302841187, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7955, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.6081292629241943, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7664, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.3914157152175903, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7688, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.4573328495025635, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.754, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.2828798294067383, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7382, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.307642936706543, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7697, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.079110026359558, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7381, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0884710550308228, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7332, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.00990891456604, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.6895, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 0.9771729111671448, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.709, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.8624317646026611, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.6993, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.7236678004264832, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7039, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.838710367679596, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.6622, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.8697376847267151, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6993, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.6491643190383911, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6688, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.6650396585464478, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6634, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.8129804730415344, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.677, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.632296085357666, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6569, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6362555623054504, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6501, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.7144654989242554, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.7063, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.6496362686157227, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6827, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.6045290231704712, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6626, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6381681561470032, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6802, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.575342059135437, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6523, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.6357355117797852, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6949, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.6100252866744995, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.673, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.505431592464447, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6805, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.6085166931152344, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6514, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.5448803901672363, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6353, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.671501874923706, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6715, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.5686109662055969, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6496, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.48324066400527954, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6609, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.5451517701148987, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.667, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.5837141275405884, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6462, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.5317370891571045, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6506, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.517959475517273, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6667, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.5565412044525146, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6669, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.5011281371116638, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.636, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.5035920143127441, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6563, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.533829391002655, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6646, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.5612527132034302, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6368, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.5507520437240601, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6504, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.4911407232284546, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6519, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.5910381078720093, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6351, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.48854464292526245, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6233, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.5122814178466797, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6455, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.4857119023799896, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.639, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.49830198287963867, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6468, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.542622983455658, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6338, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.4335653483867645, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6488, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.48472529649734497, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6423, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.6232064962387085, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6462, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.4528062343597412, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6366, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.5472282767295837, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.621, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.5490003824234009, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6385, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.4690479338169098, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6335, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.544563353061676, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6629, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5248415470123291, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6536, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.4670054614543915, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6655, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.4673774242401123, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6267, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.5523642301559448, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6389, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.5403356552124023, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6126, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.4651912748813629, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6431, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.45636171102523804, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6372, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.5118988156318665, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6245, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.4754661023616791, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6361, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.5229830741882324, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.679, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.4797511100769043, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6443, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.5188183784484863, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6349, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.4860018193721771, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6564, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6160600781440735, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.5984, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.43828341364860535, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6337, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.5764517188072205, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6276, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.5239168405532837, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6288, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.49708092212677, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6337, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.4697052538394928, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.632, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.5427669882774353, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6329, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.4586198627948761, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6171, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.510936975479126, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6085, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.5019223690032959, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.5948, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.48964670300483704, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6276, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.511461615562439, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.6194, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.47996845841407776, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6418, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.5433138012886047, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6284, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.4707951545715332, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6208, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.6000167727470398, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.598, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.46327921748161316, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6096, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.4142744541168213, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.111, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.4859069585800171, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5687, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.5571460127830505, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.6305, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.4870958924293518, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5559, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.4879758059978485, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5767, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5278946757316589, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5591, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.4541994035243988, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5814, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5303021669387817, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5899, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.4370093047618866, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5491, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.5089861154556274, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5576, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.5528135895729065, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5802, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5348351001739502, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5793, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.5000813007354736, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5996, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.564142107963562, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5751, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.5259392857551575, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5712, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.44391924142837524, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5667, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.5595237612724304, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.588, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.4950568377971649, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5891, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.5585800409317017, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5716, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6349160671234131, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5954, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.4565514326095581, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5734, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.5305560827255249, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5629, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.5316576361656189, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5622, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.4759328365325928, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5932, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.5879533886909485, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.6242, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.4395292401313782, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5471, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.4417659044265747, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5659, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.509986937046051, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5983, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.4841684401035309, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.5948, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.47024214267730713, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5938, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.44979485869407654, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5707, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5013648271560669, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5754, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.5056862235069275, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5968, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.46595561504364014, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5772, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.490436851978302, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5456, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5751674175262451, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5749, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.47302526235580444, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5556, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.4148563742637634, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.6011, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.5700293779373169, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5436, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.564984142780304, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.6139, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.4018213450908661, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5559, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.6509582996368408, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5854, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.5213705897331238, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.5989, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.3932231068611145, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.597, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.47424086928367615, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5531, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.4648714065551758, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5716, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.45904025435447693, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5306, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5592586398124695, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.6029, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.43010833859443665, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.5649, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.4213835299015045, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.571, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.4220429062843323, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5938, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4041115641593933, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.577, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.39277032017707825, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.5794, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.48155003786087036, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5806, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.4307321608066559, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5927, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.4028647541999817, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5624, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.41640058159828186, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5698, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.43608710169792175, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5822, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.3996332287788391, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5454, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.45532336831092834, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.6008, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.39399054646492004, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.549, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.43555116653442383, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5762, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.464605450630188, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5604, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.48288488388061523, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5993, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.4093412458896637, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5632, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.4191835820674896, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5478, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.49523329734802246, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5972, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4115680456161499, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5108, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.4010234773159027, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.594, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.44894617795944214, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5414, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.4436964690685272, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5792, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.3987331688404083, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5922, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.4431919455528259, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5692, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.4404211938381195, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5673, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.38400033116340637, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5912, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.3576762080192566, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.5335, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.410598486661911, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5854, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.4571099877357483, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.599, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.3698180615901947, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5697, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.39968162775039673, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5747, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.3965860605239868, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5838, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.397352010011673, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5522, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.3896408975124359, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5511, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.421280175447464, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5758, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.40224403142929077, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5664, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.43550077080726624, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5971, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.3845508098602295, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5385, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.424711674451828, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5983, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.3863534927368164, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5589, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.3979099690914154, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5715, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.38444915413856506, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5667, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.44589975476264954, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.6137, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.3899131715297699, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5696, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.3851713240146637, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5883, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.35874688625335693, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5771, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4088473618030548, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5934, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.3660665452480316, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5542, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.33395296335220337, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5515, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.3701746463775635, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5831, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.3489643335342407, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5573, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.38076457381248474, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.6108, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.387829452753067, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.6088, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.37397393584251404, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5632, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.3869399130344391, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5873, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.0653833150863647, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0771, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.41893067955970764, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5478, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.3772825002670288, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.4907, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.4282759130001068, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.5529, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.42536887526512146, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5381, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.40407678484916687, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5325, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.3934045732021332, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5494, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.38797062635421753, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.5188, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.3943995535373688, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5328, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.3562810719013214, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5397, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.4104858636856079, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5628, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.38845738768577576, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.531, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.3576855957508087, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5061, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.35464754700660706, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.5213, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.37027040123939514, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5519, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.355179101228714, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5271, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.33750709891319275, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5069, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.3592349588871002, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.5511, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.36242836713790894, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.5296, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.35842618346214294, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5432, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.3345714211463928, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5343, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.3113364577293396, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.4744, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.35715803503990173, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.5248, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.3679574131965637, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5757, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.3594515323638916, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.559, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.32706817984580994, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.5094, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.3269524574279785, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5231, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.36103031039237976, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.552, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.3555162847042084, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5191, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.32897675037384033, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.508, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.36962077021598816, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5772, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.35815155506134033, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.5398, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.3262934684753418, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5291, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.354318231344223, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5138, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.3910127580165863, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5585, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.305263489484787, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4313, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.3480761647224426, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5968, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.3212254047393799, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5017, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.3434459865093231, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.5225, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.35372763872146606, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5729, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.378040611743927, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.5652, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3556387424468994, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5358, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.3666606843471527, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5636, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.3441127836704254, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5075, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.34117305278778076, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5601, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.30562204122543335, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.5041, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.3496702313423157, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5545, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.34862929582595825, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.559, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.30959630012512207, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.4906, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.3355902135372162, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.5815, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.3508121967315674, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5443, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.3274599611759186, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5584, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.329504132270813, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.499, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.3280683755874634, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.549, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.3194555342197418, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.5199, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3465876281261444, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.563, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.33275124430656433, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.5151, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.32962992787361145, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.5344, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.30311086773872375, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.4918, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3393261134624481, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5499, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.3435958921909332, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5643, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.32636529207229614, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5406, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.28953099250793457, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4724, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3425915539264679, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5489, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.3282570242881775, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5522, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.33357545733451843, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5187, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.33037564158439636, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5595, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.33127933740615845, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.531, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.332841157913208, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5292, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.3178723454475403, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.5378, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.32591140270233154, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5181, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3086685836315155, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.5415, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.3303026258945465, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5511, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.33906057476997375, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.5328, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.3472452163696289, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.5633, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3318159282207489, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5824, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.31639620661735535, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.4953, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.3287310004234314, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.523, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.3125673532485962, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.4959, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3077879846096039, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5024, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.32344353199005127, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.52, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.3055321276187897, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.5476, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.31949734687805176, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5148, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3481059670448303, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5497, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.2971678376197815, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.4786, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.3506447374820709, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5871, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.322232723236084, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.5505, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.32472485303878784, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5239, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.2901249825954437, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.4817, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.3060782253742218, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5165, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.332023561000824, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5635, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.31392067670822144, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5259, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.3084220886230469, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.5028, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.3326255977153778, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.556, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.31157639622688293, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.5158, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3132721781730652, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.5279, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.29832401871681213, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5072, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.33024075627326965, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5767, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.32524237036705017, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5515, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.30261901021003723, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.4752, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.32912176847457886, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.542, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.3169884979724884, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5135, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.31885501742362976, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.566, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.30927810072898865, | |
| "learning_rate": 0.0, | |
| "loss": 0.5411, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 432583207665664.0, | |
| "train_loss": 0.6040483728433267, | |
| "train_runtime": 34034.1274, | |
| "train_samples_per_second": 0.881, | |
| "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": 432583207665664.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |