| { | |
| "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": 6.2944841384887695, | |
| "learning_rate": 3.125e-07, | |
| "loss": 1.0265, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 6.261817932128906, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.0057, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 6.048126220703125, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.0008, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.2389984130859375, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9984, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 5.99343204498291, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.9975, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 5.678822994232178, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.9607, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.678098678588867, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.9757, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 4.243505954742432, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9083, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.43308424949646, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.9022, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.4154136180877686, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.9135, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 2.2478554248809814, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8983, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.213982343673706, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8729, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 3.9511802196502686, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.9055, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.8154666423797607, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8707, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.459975481033325, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8631, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.4299709796905518, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8467, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.4652292728424072, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.8559, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 2.0541021823883057, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.826, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.7855989933013916, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7943, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.5372544527053833, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7867, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.342797040939331, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7834, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.3770509958267212, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7868, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.3067494630813599, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7418, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.169503092765808, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7465, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.993989109992981, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7539, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.0650089979171753, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.745, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 1.1361355781555176, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7436, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9510095119476318, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7235, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.928550124168396, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7152, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.9140353798866272, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.7189, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.9672402739524841, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6841, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8036115169525146, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7011, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.822393536567688, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.7023, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.7571423053741455, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.6653, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.8241838812828064, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.685, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7730281949043274, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6926, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.713372528553009, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.7199, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.7029000520706177, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.7199, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.7670215964317322, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6906, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6846963167190552, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6521, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.6741933226585388, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6861, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.6933429837226868, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6681, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.7184209227561951, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.676, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.6944095492362976, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6803, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.664179801940918, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6705, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.6152117252349854, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.7002, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.6948118805885315, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6463, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.5532994270324707, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6628, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.635856568813324, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6417, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.766160249710083, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6774, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.5958895683288574, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6737, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.7605820298194885, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.682, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.6079903841018677, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6683, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.561535120010376, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.636, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.7909425497055054, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6805, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.5663748383522034, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6687, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.6579152941703796, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6715, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.6015044450759888, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6472, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.6228909492492676, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6436, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.5659533739089966, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6464, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.5795664191246033, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6675, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.58271723985672, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6503, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.5349075198173523, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6445, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.5833298563957214, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6297, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.5726664066314697, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6752, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.5394296646118164, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6437, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.5713369846343994, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6894, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.5965147614479065, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6406, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.5532094240188599, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6221, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.5668460726737976, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6503, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.5831488966941833, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6564, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6880363821983337, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6607, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.5172518491744995, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6542, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.5872878432273865, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6654, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6247705817222595, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6524, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5679057836532593, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6422, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.5778680443763733, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6439, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.5653820633888245, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6406, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.5885083079338074, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6251, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.567931592464447, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6287, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.5375138521194458, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.632, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.6063922643661499, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6545, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.5123437643051147, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6306, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.5764344334602356, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.6223, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.6248758435249329, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6647, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.6314254999160767, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6681, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.6592370867729187, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6452, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.599755585193634, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6397, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.6832008957862854, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6696, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.6271616220474243, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6499, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.5687711238861084, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6374, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.6185305714607239, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6274, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.6857321262359619, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6345, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.5555843710899353, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6189, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.5853341221809387, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6121, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6267306804656982, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6638, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.5644423961639404, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.6404, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.6241409778594971, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6537, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.5932403206825256, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.655, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5342922210693359, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6456, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.6301948428153992, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6164, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.5069014430046082, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6365, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.7131555080413818, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.6633, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5053462982177734, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6223, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.445302128791809, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.0996, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.605025589466095, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5613, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.5477588176727295, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.6005, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.6136167049407959, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5986, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.5717005133628845, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5701, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5678885579109192, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.6099, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.515852689743042, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5727, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5165748596191406, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.557, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.6090772151947021, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5926, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.5668388605117798, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5784, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.571506679058075, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.6301, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5189209580421448, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5324, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.5037117004394531, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5836, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.5364065170288086, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5625, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.6278241872787476, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5764, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5745953917503357, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.6015, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.5853832960128784, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.591, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.5691282153129578, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5772, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.4796733260154724, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5646, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.5633418560028076, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.6268, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5098775029182434, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5387, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.4989281892776489, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.599, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.5478797554969788, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5541, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.588407039642334, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.6224, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.5483384728431702, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5626, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.6058452725410461, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.6336, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.5169090628623962, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5884, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5238736867904663, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5409, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.6074072122573853, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.6037, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.446067750453949, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5229, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.6598799824714661, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.6079, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6147785782814026, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.6333, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.506599485874176, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5664, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.5260503888130188, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5631, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.5902073979377747, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.6112, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.47030794620513916, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5739, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.5535953640937805, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5579, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.5441429018974304, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.6087, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.5218523740768433, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5783, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.49848422408103943, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.583, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.5394670367240906, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.6202, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.4939243793487549, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5437, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.49003490805625916, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.6121, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.48296260833740234, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5696, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.5095104575157166, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.6269, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.4607343375682831, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5727, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.45730894804000854, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5589, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.4680177569389343, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5476, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.5357135534286499, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.6474, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.5005162358283997, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5268, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.4949341118335724, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5528, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4929472804069519, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5785, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.5368454456329346, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.6097, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.580466628074646, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5839, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.5195851922035217, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5234, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.6399260759353638, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.6137, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.4763248860836029, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5787, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.5006374716758728, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5274, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.5833279490470886, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.6125, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.48551830649375916, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.6128, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.5455009341239929, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5653, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.5103030800819397, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.6492, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.557127833366394, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5357, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.49870356917381287, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5971, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.43568363785743713, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5439, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.5349757075309753, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5737, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.5567549467086792, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.6446, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4268059730529785, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5168, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.5026167631149292, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.6265, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.41909241676330566, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5317, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.47246718406677246, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5649, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.4850209951400757, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5611, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.4260648190975189, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5731, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.46243417263031006, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5747, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.4310469627380371, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5509, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5433486104011536, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.6075, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.4814201891422272, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.6185, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.4939287602901459, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5572, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.48393356800079346, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5778, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.44161245226860046, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5546, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.5093246698379517, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5712, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.46280810236930847, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5479, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.654803454875946, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.6766, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.48899897933006287, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5681, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.48702487349510193, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.6134, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.44153550267219543, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5429, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.5268108248710632, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5738, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4569060802459717, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5194, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.46703416109085083, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.6334, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.43373048305511475, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.566, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.4687662720680237, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5649, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.501291036605835, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.6314, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.45364248752593994, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5467, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.4657815098762512, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.545, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.4723939299583435, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.584, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5213126540184021, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.641, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.42041918635368347, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5352, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.48415398597717285, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5842, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.4329238831996918, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5228, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.44082117080688477, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5502, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.474412739276886, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.6056, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.519351065158844, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5689, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.43768221139907837, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5394, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.43219730257987976, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5516, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.2905503511428833, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0586, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.46712183952331543, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5416, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.4269493818283081, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5414, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.38151392340660095, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4622, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.5331805944442749, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.628, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.44074368476867676, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5521, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.44412657618522644, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5426, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.41349828243255615, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4324, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.4773734509944916, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5879, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.46014338731765747, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5222, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.46258142590522766, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5448, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.43709850311279297, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.4832, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.5018278360366821, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5765, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.4334341883659363, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.5388, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.42537960410118103, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.4986, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.45592159032821655, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5617, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.41241732239723206, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5039, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.4029995799064636, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.4829, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.4592655599117279, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.5629, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.42388197779655457, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.4749, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.47804251313209534, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5853, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.462676465511322, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.6035, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.37558698654174805, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.4216, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.4392584264278412, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5112, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.4294244050979614, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.5268, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.4404008388519287, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.4891, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.4791399836540222, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5967, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.46237269043922424, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.5589, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.39166179299354553, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.4898, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.44791609048843384, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5668, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.41991719603538513, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5073, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3884223997592926, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.475, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.4285506308078766, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5496, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.4607200622558594, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5542, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.4037615954875946, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5352, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4245019853115082, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.5025, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.4666139781475067, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5653, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.4049258530139923, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5139, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.4625247120857239, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.5451, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.40945371985435486, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5544, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.3963390588760376, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.5222, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3810086250305176, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5443, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.3690873980522156, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5427, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.40942856669425964, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5673, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.3911842107772827, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5246, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.38057664036750793, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.523, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.4790250062942505, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5832, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.37658169865608215, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4281, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.3943864107131958, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5563, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.3827630877494812, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.5046, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.4126964509487152, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5456, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.3857049345970154, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5446, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.3478200435638428, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.4994, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.38740620017051697, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5476, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.3769795000553131, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.4792, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.42110389471054077, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5477, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.38507241010665894, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.4999, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.4063776135444641, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.5754, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.38943275809288025, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5088, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.41452136635780334, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5459, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.39360129833221436, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5027, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.3715846538543701, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5465, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.36886662244796753, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.5161, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3784402310848236, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5559, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.37773990631103516, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5534, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.35538250207901, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.523, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3803144097328186, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5093, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3719671666622162, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5254, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.4069719612598419, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5537, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.37681666016578674, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4967, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.3786385953426361, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5403, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.37524834275245667, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.4861, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.41508930921554565, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5344, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.4199521541595459, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.5255, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.37261250615119934, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.5422, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3805762529373169, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5761, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.38191166520118713, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.5083, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.3519704341888428, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.5136, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.38643679022789, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5539, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.36911097168922424, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5487, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.3468243479728699, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.477, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.38326430320739746, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.5226, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.3730164170265198, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5597, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3744042217731476, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5092, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.3927542567253113, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5508, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.3964044451713562, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5709, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.35273244976997375, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.4695, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3758268654346466, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.4991, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.39026081562042236, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.5671, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.3942229449748993, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5563, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.3643067181110382, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5059, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.38008803129196167, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5394, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.3755616843700409, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.5302, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.38908350467681885, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5444, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.4015740752220154, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.5632, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.36633241176605225, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.4896, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.3795105814933777, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5353, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.4037023186683655, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5591, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.3747823238372803, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5343, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3887001574039459, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5405, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.35715678334236145, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5039, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.38841620087623596, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.6049, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.3829285800457001, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5292, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.37096184492111206, | |
| "learning_rate": 0.0, | |
| "loss": 0.5472, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 312961936211968.0, | |
| "train_loss": 0.6116033505934936, | |
| "train_runtime": 29493.0805, | |
| "train_samples_per_second": 1.017, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "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": 312961936211968.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |