| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 6.455806688672532, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0316, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.3584318209329265, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0063, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.27601957590265, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.019, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.633561959867806, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9994, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.168257171686445, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9575, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.7379628063731998, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9572, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.5053177560943887, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9424, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.253893464411118, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.9186, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.349687589283012, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9434, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.086114442629868, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9009, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.5542371555285537, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8844, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.3660399950585864, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8672, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.463822846330684, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8248, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.6085823891800346, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7911, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.251506150307515, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8038, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.2138517226169034, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7948, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.0115516940069706, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7719, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.9967586365102533, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7766, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.9748420422414402, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.757, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9745033899149671, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7388, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.7953897403498978, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7582, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.67039020165351, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.721, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7476588826201536, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7254, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.851933767925763, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.72, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.050103153251014, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7221, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.9569659066942513, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7271, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 1.0180903946753956, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6983, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.9834048884701009, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7135, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 1.0073676867959942, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7113, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7994115171324645, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.7019, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.8419690903250425, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6884, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9649544417885602, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6883, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.9482176195571902, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6922, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.2054232359185417, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6842, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.8304160439310835, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6849, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.7788944301557317, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6808, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.9211808519940944, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6816, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.141755499672058, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6771, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.2582244397210618, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6694, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.213751424428597, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6798, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.072160256932152, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6817, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.0276409716576014, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6856, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9611453357798541, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6655, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.4420963670568259, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6697, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7531021939810414, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.668, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 1.3611499420758055, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6726, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.7676770155990247, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6728, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.0672519923351085, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6539, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.9360474930296524, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6615, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.9115905560265815, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6737, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.9376645251323517, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6509, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.9909901658839197, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6613, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.9219417875130282, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6518, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.8877846456269147, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6687, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.8164231309596293, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6441, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.7957660461983604, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6528, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.1236119183964566, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6545, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.9479944170480047, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6562, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.626548034871867, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6558, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.0298462713659975, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6539, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.8077874189369452, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6793, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.740398264505655, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6522, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6750195443469825, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.665, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.7647643575597695, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6413, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.0572104788097436, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6669, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 1.1325165717999284, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6227, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.7775720377493975, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6501, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.44099236187160734, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6446, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.6272417464679025, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6545, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.8503735452053673, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6513, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.333730016691775, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6548, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6253427770084057, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6416, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.7920186893055566, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6457, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.6339368533950605, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6484, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5544575336714153, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6438, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 1.974943623020223, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6444, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.9821828355476883, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6424, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 2.663484130789532, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6477, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 4.941382266435169, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.183, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.7995378733417778, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.59, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.7720817417042583, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.6096, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.26620382611777, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5998, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 1.4754633366791623, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5792, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 1.2481412136313652, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5849, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.1376920215674418, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5971, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 1.1320442541159759, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.6035, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 1.0232067338865352, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.6091, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 1.0202253176884335, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.58, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6918802575336643, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.568, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.9802445576062785, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5868, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6331372584047888, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.583, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.8631768926664705, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5861, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6615914830584512, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.576, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.800863810967628, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5875, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.6580379221698681, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5921, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.785879346898882, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5748, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.6096781399430284, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.59, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.6853611392115616, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5897, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5845789885416388, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5869, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6403201084831845, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5817, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.8028194459196691, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.588, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.7762362375714167, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5813, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5011274095834896, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5811, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.6540273787471301, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5645, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.48135368825756675, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.6093, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5457309582236752, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5732, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.43409957473161764, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5858, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.6416881133504726, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5818, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.6620870091612782, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5754, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.41657718891348167, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5933, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.6336040794251142, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5724, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.4627677166330687, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5716, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5680979429728227, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5781, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.6426557810505764, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5893, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.6851542652536005, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5967, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.7823470379961811, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5705, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.45695030313416946, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5788, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.8111261846226543, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5911, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.6168312996627079, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5807, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.49043925134838406, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5831, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.6270161616203219, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5751, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.39170508583909547, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5613, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.522661971617142, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5717, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5257015468025916, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5798, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.47688461562895773, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5757, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.4351215404541433, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5776, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.6051250233610711, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.6019, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.6232000529993532, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5904, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.40200857504745496, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5782, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.49383447796538155, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5649, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.4397035126378412, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.579, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.47236407026339106, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.591, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.42274762850501924, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.574, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5543584627415455, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5895, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.42076772007078395, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5691, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5069394232619601, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5884, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.4038923709639309, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5411, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5133335475939002, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.6003, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.39682004790635145, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5524, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.5296349936034231, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.581, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5996083460308522, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5717, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.5413798728390828, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5788, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.5175466976647898, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5904, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4682093401048202, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5642, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.47551566805255907, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5824, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4318024287468201, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5857, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.420772862898559, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5474, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.48689725875900847, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5799, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.43171984330104424, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5766, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5738204026591428, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5673, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.5224543547495724, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.581, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.48031966263105436, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5571, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.5100900921302143, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5794, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.4832266559009672, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5719, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.49697395866229604, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.595, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5025796919566643, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5512, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.0942861291092771, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 1.0569, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 1.1203214893028626, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4969, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.7092678871521405, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.5063, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7519363641433118, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.5289, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.753356480873723, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4947, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5426401035171584, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4962, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.707369552029798, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.5074, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5497395140252154, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.5197, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6185549300177703, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.5023, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.580497252773744, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.5072, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5409738558428386, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.4982, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.43534172535797905, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.4771, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.4601930566955383, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.4929, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4320507312702627, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.5226, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.48502043578579096, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.499, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4859522614450258, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.5045, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.49391904855284163, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4968, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.4325613973134699, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.5024, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.3873002923244623, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.5131, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.41706430806430034, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.4776, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3502080718383108, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.5073, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.42777947299044994, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.4961, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.38837734226091947, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.5049, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3670279921533217, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.4865, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.39350246399828265, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.5026, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.3650725578145916, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.5152, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.3585014755905389, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.477, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.4033941898356016, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.5262, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4119980141819767, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.4792, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.38231858641717104, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.4831, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.39842931794253045, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.5192, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.35082108346987945, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.4981, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.37643446548871795, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.5059, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.46134742515544525, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.5068, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.31260136088701385, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.4882, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.4080202121089193, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.4889, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.3951121417796483, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.5041, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.4414864009705837, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.5, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.36475828248511283, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.4911, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.3681976840329648, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.4974, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.4408783566439563, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.5057, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.2998572928867727, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.5128, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.4093518980642346, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.5151, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.38801526961333443, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.4786, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3069570187353235, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.4941, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.3712540930355693, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.5228, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.35360170691882187, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4948, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.2977082345525892, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.4809, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.32965077669737364, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.5145, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.3343456972774645, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.505, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3159281951261117, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.4877, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.3724089659208682, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.5022, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3938799705682781, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.4937, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.32624238141968814, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.5074, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.32607377269848736, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.5063, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.33016534479694615, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.4901, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3220830396065913, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4913, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.33650418658391196, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.5169, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.31364559213449145, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.4862, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.35159923424406647, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.5215, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.350127339673872, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.4974, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3309668325479677, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.4947, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3890538321973108, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.5292, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.304625863899522, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.4897, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.36889928691140406, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.5319, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.32072648363389383, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.4795, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 1.5334556064043223, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.512, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.32011564920171937, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.472, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3654491843646982, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.5271, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.2947574558063537, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.4804, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3638823371729275, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.5218, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3208139595881512, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.5259, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.31957527909719796, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.49, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.31359088379829064, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.5195, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 1.1655499021539601, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.4975, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.39715740402723715, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.4893, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.35280843733532735, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.5173, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3559930169592642, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.4812, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.7231980902652241, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.9067, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.5034560975349548, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.4311, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.4368081123116843, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.4352, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.6460103786782603, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.4447, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5604762648113057, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.4459, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4659688108810321, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.4242, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5562504118798672, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.44, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.3783187935203311, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.4269, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.5092961964016923, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.4262, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.44649585580545287, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.4497, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.41274305264763705, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.4172, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.5213506905074292, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.4531, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3666854421866173, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.4049, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.43103085237184946, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.4483, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.352188864622588, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.4269, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3934477385172981, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.4291, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.37685856219419706, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.4484, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.37721604742051845, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.4273, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.32492704599266914, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.4257, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3693100483401799, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.4354, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3787891582540323, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.4389, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.34605983599487977, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.4342, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.42900570051436154, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.4297, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.29955937843112695, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.4192, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3868524850348689, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.4211, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3558484048797036, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.4379, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3807498715101144, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.4284, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.30985472150031523, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.4311, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.36061598013651475, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.4457, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.28201169754914196, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.4256, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3311788452757313, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.4217, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.5540609151275562, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.4397, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 6.832443750094971, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.4437, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.4466069899969703, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.4332, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.2768477569237576, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.4285, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.35203192444621906, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.4213, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3597046552499742, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.4374, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3325530857201198, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.4219, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.3853480285752419, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.4229, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.30159057588033156, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.4324, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3847830760236813, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.4325, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.31767472597893276, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.4228, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.2977997235495541, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.438, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.30218898899305674, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.4203, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.27407360658766305, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.4457, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.28664639374216394, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.4288, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.29135021472808337, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.4316, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.23942524410753038, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.4228, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.2610404113372267, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.4356, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.26152990417585215, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.43, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.25792082318056, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.4137, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2814226776112004, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.4345, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.24079584409939367, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.4295, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.24764066775640362, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.439, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.25293795245606016, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.4176, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.25567663000353846, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.4448, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2557748628132262, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.4143, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.26348617282981335, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.4341, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.26100409390539614, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.4364, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.2691727892889049, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.4171, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.257825726884648, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.4407, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.2316927863965696, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.4372, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.25125850891129237, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.4151, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.23884683068419738, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.4134, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.22498688360615401, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.4137, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.23545994713685783, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.4279, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.24709602404173453, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.4324, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.22722612778685342, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.4148, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.2276563993650903, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.428, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.24042359633073718, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.4307, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.2184993596582375, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.412, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.23158938650439423, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.4254, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.2400100801267434, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.4328, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.22379826927352578, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.4498, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.2175943729139426, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.4241, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.24142393569400988, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.4397, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.2277954625839659, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.4156, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.21963572957000518, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.4332, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.7453095594065533, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.7858, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3623525450413221, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.4057, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.29588229171753117, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.3755, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3163251340108107, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.3829, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.3065162031868669, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.3984, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3400715690966586, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.3754, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.32985853310217333, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.3703, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.3393092530537702, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.3987, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.2859983479785996, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.381, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.2603753990950839, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.3842, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.27353545934037826, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.3951, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.2976771126638214, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.3638, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.2979409629661625, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.4115, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.24959324888076445, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.3815, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.25345576845513385, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.3773, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.2556273873961127, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.3958, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.26751038831185286, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.373, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2642931307211567, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.3754, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2556489340286757, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.3946, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.22787664846961111, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.3816, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.23950933032081287, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.3879, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.24496434324273658, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.3913, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.24368645441312148, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.3767, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2327625767748398, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.3781, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.22176499332523203, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.3843, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.22707367923274444, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.3896, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.23024969123081462, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3713, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.22965164941152233, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.387, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.22295693061640967, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.3816, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.21915300930601672, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.3935, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.20845120406783407, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.3684, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.21830097067368961, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.3949, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2319295084970909, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.3861, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2252688795708634, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.3822, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.21942449224589794, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.4034, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.22032239874191956, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.379, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.21751705116878547, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.3882, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.20693329192460425, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.3717, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.20856743603119338, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.3733, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2250156081858486, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.3992, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.20967206248158893, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.3932, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.21304814301631292, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.384, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.21347126689517681, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.3774, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.2144707901587489, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3902, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.21283483621641808, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.38, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.20874683219822784, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.3766, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2106890163018537, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.3992, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.20561612581088098, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.3875, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.20439781434323565, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.3829, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2681421275066011, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.4002, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.21204428712133977, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.3782, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.20647473756251203, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.3687, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.21537782712985679, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.3877, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.20929612677212645, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.3881, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.20469289983672867, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.3782, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.20501899025609002, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.363, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.20999132281758767, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.392, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.2074595523353208, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.3824, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.20946142382537822, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3833, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.20538052481360528, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.3728, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.21061747295792488, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.3963, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.20832891708472415, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.3678, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.21826074453447222, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.3896, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.20430725766989333, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.3842, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.20976252169539303, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.3753, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.20149655479697867, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.3836, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2094211498322376, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.386, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.2010445971986087, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.3792, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2055018594838335, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.3959, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.20749079229469425, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.3768, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.2002473418624163, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.3918, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.1966493907159764, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.3755, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.20815227965413274, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.3894, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.20441975721928723, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.3768, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.21525302379268882, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.389, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.19748050149463348, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.3797, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.20299006619692414, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.3827, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.20924765505845955, | |
| "learning_rate": 0.0, | |
| "loss": 0.3776, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.7434270791410647e+18, | |
| "train_loss": 0.529817401445829, | |
| "train_runtime": 20813.8025, | |
| "train_samples_per_second": 2.402, | |
| "train_steps_per_second": 0.019 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 1.7434270791410647e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |