| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.992, | |
| "eval_steps": 500, | |
| "global_step": 498, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006, | |
| "grad_norm": 7.031697750091553, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 1.2696, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012, | |
| "grad_norm": 6.887118339538574, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 1.2574, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.018, | |
| "grad_norm": 6.771040439605713, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 1.2562, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 6.925063133239746, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.2938, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 6.506284713745117, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.2508, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.036, | |
| "grad_norm": 6.504903316497803, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.2622, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.042, | |
| "grad_norm": 6.177670478820801, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 1.2409, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 5.105483531951904, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 1.2019, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.054, | |
| "grad_norm": 4.707829475402832, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 1.1884, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.526137351989746, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.2281, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.066, | |
| "grad_norm": 2.691443681716919, | |
| "learning_rate": 2.2e-06, | |
| "loss": 1.1239, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 2.7380127906799316, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.1911, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.078, | |
| "grad_norm": 2.5737719535827637, | |
| "learning_rate": 2.6e-06, | |
| "loss": 1.1153, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.084, | |
| "grad_norm": 2.6669070720672607, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.1497, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.018786430358887, | |
| "learning_rate": 3e-06, | |
| "loss": 1.1334, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 4.247231483459473, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.1403, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.102, | |
| "grad_norm": 3.968531847000122, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 1.1283, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.108, | |
| "grad_norm": 3.5382187366485596, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.0746, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.114, | |
| "grad_norm": 2.893420457839966, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 1.1355, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.0976672172546387, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0489, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.126, | |
| "grad_norm": 2.1325271129608154, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 1.0639, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.132, | |
| "grad_norm": 1.6968716382980347, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.0202, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.138, | |
| "grad_norm": 1.4321019649505615, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 1.029, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.4450711011886597, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.0487, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.3008755445480347, | |
| "learning_rate": 5e-06, | |
| "loss": 0.996, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.156, | |
| "grad_norm": 1.0525767803192139, | |
| "learning_rate": 5.2e-06, | |
| "loss": 0.9986, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.162, | |
| "grad_norm": 1.0507460832595825, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 1.01, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 1.1463662385940552, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.022, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.174, | |
| "grad_norm": 1.0144261121749878, | |
| "learning_rate": 5.8e-06, | |
| "loss": 0.9666, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9777807593345642, | |
| "learning_rate": 6e-06, | |
| "loss": 1.0246, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.186, | |
| "grad_norm": 1.0017383098602295, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 0.9475, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 0.8886977434158325, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.9389, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.198, | |
| "grad_norm": 0.8426088690757751, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 0.9617, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.204, | |
| "grad_norm": 1.0878757238388062, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.958, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.0810081958770752, | |
| "learning_rate": 7e-06, | |
| "loss": 0.967, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 0.9367354512214661, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.9741, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.222, | |
| "grad_norm": 0.7725072503089905, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.9337, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.228, | |
| "grad_norm": 0.9318351149559021, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.9611, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.234, | |
| "grad_norm": 0.902850866317749, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.9655, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.8286356329917908, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9412, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.246, | |
| "grad_norm": 0.7070385217666626, | |
| "learning_rate": 8.2e-06, | |
| "loss": 0.9366, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.252, | |
| "grad_norm": 0.832440972328186, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.9556, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.258, | |
| "grad_norm": 0.7998343110084534, | |
| "learning_rate": 8.6e-06, | |
| "loss": 0.9534, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 0.8385363817214966, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.9778, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.8293184041976929, | |
| "learning_rate": 9e-06, | |
| "loss": 0.9147, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.276, | |
| "grad_norm": 0.9170862436294556, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 0.9265, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.282, | |
| "grad_norm": 0.7676896452903748, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.9426, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.9426606297492981, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.9417, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.294, | |
| "grad_norm": 0.8821513056755066, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.9448, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.6868188381195068, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9279, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.306, | |
| "grad_norm": 0.9008329510688782, | |
| "learning_rate": 9.999877063188124e-06, | |
| "loss": 0.9508, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 0.6640547513961792, | |
| "learning_rate": 9.999508258797876e-06, | |
| "loss": 0.9041, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.318, | |
| "grad_norm": 0.683678388595581, | |
| "learning_rate": 9.998893604965111e-06, | |
| "loss": 0.9314, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.324, | |
| "grad_norm": 0.6909109950065613, | |
| "learning_rate": 9.998033131915266e-06, | |
| "loss": 0.9362, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.6794642806053162, | |
| "learning_rate": 9.996926881961862e-06, | |
| "loss": 0.9242, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.745780348777771, | |
| "learning_rate": 9.995574909504434e-06, | |
| "loss": 0.9422, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.342, | |
| "grad_norm": 0.6801639199256897, | |
| "learning_rate": 9.993977281025862e-06, | |
| "loss": 0.9283, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.348, | |
| "grad_norm": 0.7167317271232605, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.9116, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.354, | |
| "grad_norm": 0.7170763611793518, | |
| "learning_rate": 9.990045382333243e-06, | |
| "loss": 0.8707, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.5977501273155212, | |
| "learning_rate": 9.987711305469232e-06, | |
| "loss": 0.9255, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.366, | |
| "grad_norm": 0.6289224028587341, | |
| "learning_rate": 9.985131959274637e-06, | |
| "loss": 0.9027, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.372, | |
| "grad_norm": 0.6277759671211243, | |
| "learning_rate": 9.982307470588097e-06, | |
| "loss": 0.9195, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.378, | |
| "grad_norm": 0.6727122664451599, | |
| "learning_rate": 9.979237978303066e-06, | |
| "loss": 0.9432, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6464661955833435, | |
| "learning_rate": 9.975923633360985e-06, | |
| "loss": 0.9259, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.8382951617240906, | |
| "learning_rate": 9.972364598743851e-06, | |
| "loss": 0.8883, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.396, | |
| "grad_norm": 0.736031174659729, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.8914, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.402, | |
| "grad_norm": 0.7064986824989319, | |
| "learning_rate": 9.96451317256656e-06, | |
| "loss": 0.8648, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 0.6144806742668152, | |
| "learning_rate": 9.960221167098124e-06, | |
| "loss": 0.9126, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.414, | |
| "grad_norm": 0.7800092101097107, | |
| "learning_rate": 9.955685244119092e-06, | |
| "loss": 0.8957, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.8123449683189392, | |
| "learning_rate": 9.950905626682229e-06, | |
| "loss": 0.8948, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.426, | |
| "grad_norm": 0.7551453113555908, | |
| "learning_rate": 9.945882549823906e-06, | |
| "loss": 0.911, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.7735251784324646, | |
| "learning_rate": 9.940616260552545e-06, | |
| "loss": 0.8891, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.438, | |
| "grad_norm": 0.7624951601028442, | |
| "learning_rate": 9.935107017836472e-06, | |
| "loss": 0.9046, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.444, | |
| "grad_norm": 0.7585752010345459, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.8906, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.8881312012672424, | |
| "learning_rate": 9.92336076766601e-06, | |
| "loss": 0.9212, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 0.6675351858139038, | |
| "learning_rate": 9.917124337830242e-06, | |
| "loss": 0.9124, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.462, | |
| "grad_norm": 0.7288476228713989, | |
| "learning_rate": 9.910646109758596e-06, | |
| "loss": 0.9014, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.468, | |
| "grad_norm": 0.8076652884483337, | |
| "learning_rate": 9.903926402016153e-06, | |
| "loss": 0.8854, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.474, | |
| "grad_norm": 0.6705262660980225, | |
| "learning_rate": 9.896965545042692e-06, | |
| "loss": 0.9108, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.957453727722168, | |
| "learning_rate": 9.889763881136439e-06, | |
| "loss": 0.904, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.486, | |
| "grad_norm": 0.687865138053894, | |
| "learning_rate": 9.882321764437234e-06, | |
| "loss": 0.8938, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.492, | |
| "grad_norm": 0.8026799559593201, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.8863, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.498, | |
| "grad_norm": 0.8794582486152649, | |
| "learning_rate": 9.866717648322335e-06, | |
| "loss": 0.9067, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 0.6872241497039795, | |
| "learning_rate": 9.858556416234755e-06, | |
| "loss": 0.8882, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.8959775567054749, | |
| "learning_rate": 9.850156265972722e-06, | |
| "loss": 0.913, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.516, | |
| "grad_norm": 0.5751799941062927, | |
| "learning_rate": 9.841517610611309e-06, | |
| "loss": 0.9085, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.522, | |
| "grad_norm": 0.9262533187866211, | |
| "learning_rate": 9.832640874954017e-06, | |
| "loss": 0.9203, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.659100353717804, | |
| "learning_rate": 9.82352649551188e-06, | |
| "loss": 0.8833, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.534, | |
| "grad_norm": 0.7242835760116577, | |
| "learning_rate": 9.814174920481999e-06, | |
| "loss": 0.8802, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.8866828083992004, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.8726, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.546, | |
| "grad_norm": 0.7923456430435181, | |
| "learning_rate": 9.794762034744925e-06, | |
| "loss": 0.9173, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 0.7427409291267395, | |
| "learning_rate": 9.784701678661045e-06, | |
| "loss": 0.9242, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.558, | |
| "grad_norm": 0.6884157061576843, | |
| "learning_rate": 9.774406036189104e-06, | |
| "loss": 0.9162, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.564, | |
| "grad_norm": 0.6347200274467468, | |
| "learning_rate": 9.763875613614482e-06, | |
| "loss": 0.8712, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.703681230545044, | |
| "learning_rate": 9.753110928767816e-06, | |
| "loss": 0.91, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.6913872957229614, | |
| "learning_rate": 9.742112510999516e-06, | |
| "loss": 0.9004, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.582, | |
| "grad_norm": 0.5897263884544373, | |
| "learning_rate": 9.730880901153747e-06, | |
| "loss": 0.9157, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.588, | |
| "grad_norm": 0.5855635404586792, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.9058, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.594, | |
| "grad_norm": 0.724475622177124, | |
| "learning_rate": 9.707720325915105e-06, | |
| "loss": 0.9156, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.6259617209434509, | |
| "learning_rate": 9.69579249943714e-06, | |
| "loss": 0.9283, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.606, | |
| "grad_norm": 0.676932156085968, | |
| "learning_rate": 9.683633758655529e-06, | |
| "loss": 0.9148, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.612, | |
| "grad_norm": 0.6588666439056396, | |
| "learning_rate": 9.671244701472999e-06, | |
| "loss": 0.8542, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.618, | |
| "grad_norm": 0.7273979187011719, | |
| "learning_rate": 9.658625937118033e-06, | |
| "loss": 0.9185, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.605010986328125, | |
| "learning_rate": 9.645778086114892e-06, | |
| "loss": 0.9427, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.7519787549972534, | |
| "learning_rate": 9.632701780253111e-06, | |
| "loss": 0.9034, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.636, | |
| "grad_norm": 0.6903522610664368, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.8935, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.642, | |
| "grad_norm": 0.6054957509040833, | |
| "learning_rate": 9.605866387251186e-06, | |
| "loss": 0.8894, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 0.6449054479598999, | |
| "learning_rate": 9.592108619734107e-06, | |
| "loss": 0.8794, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.654, | |
| "grad_norm": 0.6131969094276428, | |
| "learning_rate": 9.578125036539625e-06, | |
| "loss": 0.8985, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.6487088203430176, | |
| "learning_rate": 9.563916325306595e-06, | |
| "loss": 0.9301, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.666, | |
| "grad_norm": 0.5402405261993408, | |
| "learning_rate": 9.549483184744483e-06, | |
| "loss": 0.9235, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.5446547269821167, | |
| "learning_rate": 9.534826324599002e-06, | |
| "loss": 0.8869, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.678, | |
| "grad_norm": 0.5393290519714355, | |
| "learning_rate": 9.519946465617217e-06, | |
| "loss": 0.891, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.684, | |
| "grad_norm": 0.6562361717224121, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.9527, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.5437909960746765, | |
| "learning_rate": 9.489520688926534e-06, | |
| "loss": 0.9027, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 0.6590492129325867, | |
| "learning_rate": 9.473976267396831e-06, | |
| "loss": 0.8903, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.702, | |
| "grad_norm": 0.6078689694404602, | |
| "learning_rate": 9.458211839315636e-06, | |
| "loss": 0.8749, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.708, | |
| "grad_norm": 0.6383506655693054, | |
| "learning_rate": 9.442228179894362e-06, | |
| "loss": 0.8689, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.714, | |
| "grad_norm": 0.5695559978485107, | |
| "learning_rate": 9.426026075125062e-06, | |
| "loss": 0.864, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5810907483100891, | |
| "learning_rate": 9.409606321741776e-06, | |
| "loss": 0.8627, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.726, | |
| "grad_norm": 0.6629945635795593, | |
| "learning_rate": 9.39296972718136e-06, | |
| "loss": 0.902, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.732, | |
| "grad_norm": 0.5847179889678955, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.8713, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.738, | |
| "grad_norm": 0.5658993124961853, | |
| "learning_rate": 9.35904929755184e-06, | |
| "loss": 0.9075, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 0.5689570903778076, | |
| "learning_rate": 9.341767130510529e-06, | |
| "loss": 0.9153, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.6111438870429993, | |
| "learning_rate": 9.324271458265642e-06, | |
| "loss": 0.9107, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.756, | |
| "grad_norm": 0.6272052526473999, | |
| "learning_rate": 9.306563141162046e-06, | |
| "loss": 0.9211, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.762, | |
| "grad_norm": 0.6451931595802307, | |
| "learning_rate": 9.288643050001362e-06, | |
| "loss": 0.8603, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5918119549751282, | |
| "learning_rate": 9.270512065999139e-06, | |
| "loss": 0.8865, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.774, | |
| "grad_norm": 0.5935030579566956, | |
| "learning_rate": 9.252171080741525e-06, | |
| "loss": 0.8983, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.7970097064971924, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.8849, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.786, | |
| "grad_norm": 0.5275349020957947, | |
| "learning_rate": 9.214862724394133e-06, | |
| "loss": 0.885, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 0.7283827066421509, | |
| "learning_rate": 9.195897187932513e-06, | |
| "loss": 0.8919, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.798, | |
| "grad_norm": 0.6388846635818481, | |
| "learning_rate": 9.176725319381589e-06, | |
| "loss": 0.8673, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.804, | |
| "grad_norm": 0.5710157155990601, | |
| "learning_rate": 9.157348061512728e-06, | |
| "loss": 0.8724, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.7034839391708374, | |
| "learning_rate": 9.137766367197246e-06, | |
| "loss": 0.8901, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.6445233225822449, | |
| "learning_rate": 9.117981199359575e-06, | |
| "loss": 0.881, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.822, | |
| "grad_norm": 0.5970780253410339, | |
| "learning_rate": 9.097993530929895e-06, | |
| "loss": 0.8504, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.828, | |
| "grad_norm": 0.7257516384124756, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.8811, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.834, | |
| "grad_norm": 0.6562220454216003, | |
| "learning_rate": 9.057414633756466e-06, | |
| "loss": 0.8868, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.7522321939468384, | |
| "learning_rate": 9.036825400468814e-06, | |
| "loss": 0.8844, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.846, | |
| "grad_norm": 0.6761206388473511, | |
| "learning_rate": 9.016037657403225e-06, | |
| "loss": 0.8805, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.852, | |
| "grad_norm": 0.6708723306655884, | |
| "learning_rate": 8.995052426791247e-06, | |
| "loss": 0.8913, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.858, | |
| "grad_norm": 0.7201263904571533, | |
| "learning_rate": 8.973870740575814e-06, | |
| "loss": 0.8479, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.7299242615699768, | |
| "learning_rate": 8.952493640360518e-06, | |
| "loss": 0.9066, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.683100163936615, | |
| "learning_rate": 8.930922177358379e-06, | |
| "loss": 0.8713, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.876, | |
| "grad_norm": 0.699967622756958, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.8871, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.882, | |
| "grad_norm": 0.599724292755127, | |
| "learning_rate": 8.88720041558216e-06, | |
| "loss": 0.8972, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 0.7004441022872925, | |
| "learning_rate": 8.865052266813686e-06, | |
| "loss": 0.9008, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.894, | |
| "grad_norm": 0.6089579463005066, | |
| "learning_rate": 8.842714055163841e-06, | |
| "loss": 0.866, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.6796953082084656, | |
| "learning_rate": 8.820186879108038e-06, | |
| "loss": 0.8995, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.906, | |
| "grad_norm": 0.627301812171936, | |
| "learning_rate": 8.797471846413957e-06, | |
| "loss": 0.881, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.6668423414230347, | |
| "learning_rate": 8.77457007408708e-06, | |
| "loss": 0.8812, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.918, | |
| "grad_norm": 0.6487154364585876, | |
| "learning_rate": 8.751482688315758e-06, | |
| "loss": 0.8645, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.924, | |
| "grad_norm": 0.7148600816726685, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.9134, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.7381123304367065, | |
| "learning_rate": 8.704755626774796e-06, | |
| "loss": 0.9009, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 0.6762275099754333, | |
| "learning_rate": 8.681118248795548e-06, | |
| "loss": 0.8654, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.942, | |
| "grad_norm": 0.7460249662399292, | |
| "learning_rate": 8.65729985283964e-06, | |
| "loss": 0.917, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.948, | |
| "grad_norm": 0.695384681224823, | |
| "learning_rate": 8.633301610170136e-06, | |
| "loss": 0.899, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.954, | |
| "grad_norm": 0.6210854649543762, | |
| "learning_rate": 8.609124700894017e-06, | |
| "loss": 0.8325, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.719469428062439, | |
| "learning_rate": 8.584770313904138e-06, | |
| "loss": 0.8266, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.966, | |
| "grad_norm": 0.5835751891136169, | |
| "learning_rate": 8.560239646820779e-06, | |
| "loss": 0.895, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.972, | |
| "grad_norm": 0.7177435159683228, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.9012, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.978, | |
| "grad_norm": 0.6134434342384338, | |
| "learning_rate": 8.510654306138028e-06, | |
| "loss": 0.8691, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 0.6205304265022278, | |
| "learning_rate": 8.485602070884118e-06, | |
| "loss": 0.8629, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.5721279978752136, | |
| "learning_rate": 8.460378432107779e-06, | |
| "loss": 0.8684, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.996, | |
| "grad_norm": 0.6909639239311218, | |
| "learning_rate": 8.43498463017451e-06, | |
| "loss": 0.9096, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.004, | |
| "grad_norm": 1.0973442792892456, | |
| "learning_rate": 8.409421913817526e-06, | |
| "loss": 1.4196, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.5294898152351379, | |
| "learning_rate": 8.383691540076372e-06, | |
| "loss": 0.7903, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 0.6551557183265686, | |
| "learning_rate": 8.357794774235094e-06, | |
| "loss": 0.8664, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.022, | |
| "grad_norm": 0.5237061381340027, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.7368, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.028, | |
| "grad_norm": 0.7365502715110779, | |
| "learning_rate": 8.305507168237152e-06, | |
| "loss": 0.9978, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.034, | |
| "grad_norm": 0.5492303371429443, | |
| "learning_rate": 8.279118899309121e-06, | |
| "loss": 0.7603, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.6328789591789246, | |
| "learning_rate": 8.252569380611793e-06, | |
| "loss": 0.8794, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.046, | |
| "grad_norm": 0.6356245279312134, | |
| "learning_rate": 8.22585991771044e-06, | |
| "loss": 0.7344, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.052, | |
| "grad_norm": 0.6173516511917114, | |
| "learning_rate": 8.198991824035546e-06, | |
| "loss": 0.7881, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.058, | |
| "grad_norm": 0.6230927109718323, | |
| "learning_rate": 8.171966420818227e-06, | |
| "loss": 0.8771, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.7857466340065002, | |
| "learning_rate": 8.144785037025246e-06, | |
| "loss": 0.9245, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.6522354483604431, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.7209, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.076, | |
| "grad_norm": 0.6525973677635193, | |
| "learning_rate": 8.089959681865134e-06, | |
| "loss": 0.9277, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.082, | |
| "grad_norm": 0.6244474649429321, | |
| "learning_rate": 8.062318406519751e-06, | |
| "loss": 0.7997, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6072016954421997, | |
| "learning_rate": 8.034526542509629e-06, | |
| "loss": 0.792, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.094, | |
| "grad_norm": 0.6310375332832336, | |
| "learning_rate": 8.00658545649203e-06, | |
| "loss": 0.9122, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.5891605019569397, | |
| "learning_rate": 7.978496522462167e-06, | |
| "loss": 0.8256, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.106, | |
| "grad_norm": 0.5977054238319397, | |
| "learning_rate": 7.950261121685642e-06, | |
| "loss": 0.8305, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.112, | |
| "grad_norm": 0.5768773555755615, | |
| "learning_rate": 7.921880642630514e-06, | |
| "loss": 0.8153, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1179999999999999, | |
| "grad_norm": 0.6263236999511719, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.803, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.124, | |
| "grad_norm": 0.5831404328346252, | |
| "learning_rate": 7.864690039158991e-06, | |
| "loss": 0.8451, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.6220493316650391, | |
| "learning_rate": 7.835882727074779e-06, | |
| "loss": 0.8614, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 0.5806083679199219, | |
| "learning_rate": 7.806935961238041e-06, | |
| "loss": 0.8056, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.142, | |
| "grad_norm": 0.6333805918693542, | |
| "learning_rate": 7.777851165098012e-06, | |
| "loss": 0.7189, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.148, | |
| "grad_norm": 0.7612046003341675, | |
| "learning_rate": 7.748629768891542e-06, | |
| "loss": 0.9553, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.154, | |
| "grad_norm": 0.6898669600486755, | |
| "learning_rate": 7.719273209572745e-06, | |
| "loss": 0.8351, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.7063294649124146, | |
| "learning_rate": 7.689782930742345e-06, | |
| "loss": 0.7966, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.166, | |
| "grad_norm": 0.5425288677215576, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.8904, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.172, | |
| "grad_norm": 0.6887763738632202, | |
| "learning_rate": 7.630407021756419e-06, | |
| "loss": 0.7649, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.178, | |
| "grad_norm": 0.6101463437080383, | |
| "learning_rate": 7.600524311394873e-06, | |
| "loss": 0.8274, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 0.5494688749313354, | |
| "learning_rate": 7.570513720966108e-06, | |
| "loss": 0.7915, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.6239190101623535, | |
| "learning_rate": 7.540376726232648e-06, | |
| "loss": 0.8471, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.196, | |
| "grad_norm": 0.6039848327636719, | |
| "learning_rate": 7.51011480917291e-06, | |
| "loss": 0.8698, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.202, | |
| "grad_norm": 0.6674699187278748, | |
| "learning_rate": 7.4797294579083405e-06, | |
| "loss": 0.9333, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.208, | |
| "grad_norm": 0.6486606597900391, | |
| "learning_rate": 7.449222166630218e-06, | |
| "loss": 0.8904, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.214, | |
| "grad_norm": 0.4372139275074005, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.7593, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.536348283290863, | |
| "learning_rate": 7.3878477707065314e-06, | |
| "loss": 0.7174, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.226, | |
| "grad_norm": 0.6209816336631775, | |
| "learning_rate": 7.3569836841299905e-06, | |
| "loss": 0.9422, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.232, | |
| "grad_norm": 0.5448585748672485, | |
| "learning_rate": 7.326003693529538e-06, | |
| "loss": 0.7474, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.238, | |
| "grad_norm": 0.6505892872810364, | |
| "learning_rate": 7.294909322337689e-06, | |
| "loss": 0.8327, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.244, | |
| "grad_norm": 0.568920373916626, | |
| "learning_rate": 7.263702099611586e-06, | |
| "loss": 0.8315, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.614250659942627, | |
| "learning_rate": 7.232383559957815e-06, | |
| "loss": 0.8018, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 0.6246547698974609, | |
| "learning_rate": 7.200955243456946e-06, | |
| "loss": 0.8009, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.262, | |
| "grad_norm": 0.5993421673774719, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.8355, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.268, | |
| "grad_norm": 0.5505126714706421, | |
| "learning_rate": 7.137775467151411e-06, | |
| "loss": 0.8522, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.274, | |
| "grad_norm": 0.637431800365448, | |
| "learning_rate": 7.106027114194856e-06, | |
| "loss": 0.8511, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5857645273208618, | |
| "learning_rate": 7.0741751979346395e-06, | |
| "loss": 0.8573, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.286, | |
| "grad_norm": 0.6337495446205139, | |
| "learning_rate": 7.042221284679982e-06, | |
| "loss": 0.8298, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.292, | |
| "grad_norm": 0.5764760971069336, | |
| "learning_rate": 7.010166945755768e-06, | |
| "loss": 0.8252, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.298, | |
| "grad_norm": 0.632314145565033, | |
| "learning_rate": 6.978013757425295e-06, | |
| "loss": 0.8492, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.304, | |
| "grad_norm": 0.5820547342300415, | |
| "learning_rate": 6.945763300812746e-06, | |
| "loss": 0.8621, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.5542883276939392, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.7941, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.316, | |
| "grad_norm": 0.6561512351036072, | |
| "learning_rate": 6.880976931075887e-06, | |
| "loss": 0.9207, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.322, | |
| "grad_norm": 0.5089630484580994, | |
| "learning_rate": 6.848444203803476e-06, | |
| "loss": 0.7331, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.328, | |
| "grad_norm": 0.6927565932273865, | |
| "learning_rate": 6.8158205797961265e-06, | |
| "loss": 0.9655, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.334, | |
| "grad_norm": 0.4950551390647888, | |
| "learning_rate": 6.783107663311566e-06, | |
| "loss": 0.6941, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.6248024702072144, | |
| "learning_rate": 6.750307062998462e-06, | |
| "loss": 0.8132, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.346, | |
| "grad_norm": 0.5446671843528748, | |
| "learning_rate": 6.717420391817306e-06, | |
| "loss": 0.8235, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 0.6277297735214233, | |
| "learning_rate": 6.684449266961101e-06, | |
| "loss": 0.816, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.358, | |
| "grad_norm": 0.5996410250663757, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.9069, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.3639999999999999, | |
| "grad_norm": 0.5332024097442627, | |
| "learning_rate": 6.618260145680758e-06, | |
| "loss": 0.7992, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.6639161109924316, | |
| "learning_rate": 6.585045404088442e-06, | |
| "loss": 0.7207, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 0.5572250485420227, | |
| "learning_rate": 6.55175271832466e-06, | |
| "loss": 0.8031, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.3820000000000001, | |
| "grad_norm": 0.6577343940734863, | |
| "learning_rate": 6.518383725548074e-06, | |
| "loss": 0.8563, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.388, | |
| "grad_norm": 0.56269770860672, | |
| "learning_rate": 6.484940066669718e-06, | |
| "loss": 0.8494, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.3940000000000001, | |
| "grad_norm": 0.5737781524658203, | |
| "learning_rate": 6.451423386272312e-06, | |
| "loss": 0.8715, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.5935752987861633, | |
| "learning_rate": 6.417835332529389e-06, | |
| "loss": 0.8714, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4060000000000001, | |
| "grad_norm": 0.5302621126174927, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.9023, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.412, | |
| "grad_norm": 0.47902560234069824, | |
| "learning_rate": 6.350451715168728e-06, | |
| "loss": 0.7481, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.418, | |
| "grad_norm": 0.6068887114524841, | |
| "learning_rate": 6.3166594651218235e-06, | |
| "loss": 0.9063, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.424, | |
| "grad_norm": 0.5201268792152405, | |
| "learning_rate": 6.282802468708133e-06, | |
| "loss": 0.7979, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.5422048568725586, | |
| "learning_rate": 6.248882390836135e-06, | |
| "loss": 0.9127, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.436, | |
| "grad_norm": 0.49023619294166565, | |
| "learning_rate": 6.21490089951632e-06, | |
| "loss": 0.7837, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.442, | |
| "grad_norm": 0.5470706820487976, | |
| "learning_rate": 6.180859665779173e-06, | |
| "loss": 0.8533, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 0.5331608653068542, | |
| "learning_rate": 6.14676036359299e-06, | |
| "loss": 0.8272, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.454, | |
| "grad_norm": 0.5213180780410767, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.8428, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.5773241519927979, | |
| "learning_rate": 6.078394263941762e-06, | |
| "loss": 0.8013, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.466, | |
| "grad_norm": 0.5300610661506653, | |
| "learning_rate": 6.04413082836085e-06, | |
| "loss": 0.8215, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5806800723075867, | |
| "learning_rate": 6.009816047933849e-06, | |
| "loss": 0.87, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.478, | |
| "grad_norm": 0.5193670988082886, | |
| "learning_rate": 5.975451610080643e-06, | |
| "loss": 0.7694, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.484, | |
| "grad_norm": 0.6043456792831421, | |
| "learning_rate": 5.941039204663001e-06, | |
| "loss": 0.9321, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.5155271291732788, | |
| "learning_rate": 5.906580523901493e-06, | |
| "loss": 0.7809, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.496, | |
| "grad_norm": 0.5329350829124451, | |
| "learning_rate": 5.872077262292255e-06, | |
| "loss": 0.8733, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.502, | |
| "grad_norm": 0.49170058965682983, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.7802, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.508, | |
| "grad_norm": 0.5601338744163513, | |
| "learning_rate": 5.802943785392986e-06, | |
| "loss": 0.8852, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.514, | |
| "grad_norm": 0.4916960895061493, | |
| "learning_rate": 5.768316969722651e-06, | |
| "loss": 0.795, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.612114429473877, | |
| "learning_rate": 5.733652372276809e-06, | |
| "loss": 0.8377, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.526, | |
| "grad_norm": 0.6219832301139832, | |
| "learning_rate": 5.698951697677498e-06, | |
| "loss": 0.8324, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.532, | |
| "grad_norm": 0.5465147495269775, | |
| "learning_rate": 5.66421665232084e-06, | |
| "loss": 0.8394, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.538, | |
| "grad_norm": 0.7250228524208069, | |
| "learning_rate": 5.629448944293128e-06, | |
| "loss": 0.8914, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 0.640958309173584, | |
| "learning_rate": 5.594650283286835e-06, | |
| "loss": 0.8513, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.6299832463264465, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.8624, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.556, | |
| "grad_norm": 0.5859169363975525, | |
| "learning_rate": 5.524966948634774e-06, | |
| "loss": 0.7434, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.562, | |
| "grad_norm": 0.5168582797050476, | |
| "learning_rate": 5.490085701647805e-06, | |
| "loss": 0.8681, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 0.6453350782394409, | |
| "learning_rate": 5.4551803548313505e-06, | |
| "loss": 0.7842, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.5739999999999998, | |
| "grad_norm": 0.6329458355903625, | |
| "learning_rate": 5.420252624646238e-06, | |
| "loss": 0.942, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.5038691163063049, | |
| "learning_rate": 5.385304228653983e-06, | |
| "loss": 0.8119, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.5859999999999999, | |
| "grad_norm": 0.56325364112854, | |
| "learning_rate": 5.350336885432337e-06, | |
| "loss": 0.7853, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.592, | |
| "grad_norm": 0.520606279373169, | |
| "learning_rate": 5.315352314490781e-06, | |
| "loss": 0.7943, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.5979999999999999, | |
| "grad_norm": 0.5728681087493896, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.8226, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.604, | |
| "grad_norm": 0.567225992679596, | |
| "learning_rate": 5.245338371637091e-06, | |
| "loss": 0.9019, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.6099999999999999, | |
| "grad_norm": 0.6226865649223328, | |
| "learning_rate": 5.210312442641327e-06, | |
| "loss": 0.8341, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.616, | |
| "grad_norm": 0.5414499044418335, | |
| "learning_rate": 5.175276171589082e-06, | |
| "loss": 0.8727, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.6219999999999999, | |
| "grad_norm": 0.5580145120620728, | |
| "learning_rate": 5.140231281379345e-06, | |
| "loss": 0.8318, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.6280000000000001, | |
| "grad_norm": 0.4797461926937103, | |
| "learning_rate": 5.1051794953349445e-06, | |
| "loss": 0.8061, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.634, | |
| "grad_norm": 0.5352113246917725, | |
| "learning_rate": 5.070122537117812e-06, | |
| "loss": 0.8135, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.5719079971313477, | |
| "learning_rate": 5.0350621306442185e-06, | |
| "loss": 0.897, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.646, | |
| "grad_norm": 0.43101564049720764, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7674, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.6520000000000001, | |
| "grad_norm": 0.5771524310112, | |
| "learning_rate": 4.964937869355782e-06, | |
| "loss": 0.9136, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.658, | |
| "grad_norm": 0.4993324875831604, | |
| "learning_rate": 4.92987746288219e-06, | |
| "loss": 0.7627, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5497719645500183, | |
| "learning_rate": 4.894820504665056e-06, | |
| "loss": 0.8228, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.5680533051490784, | |
| "learning_rate": 4.859768718620656e-06, | |
| "loss": 0.8384, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.6760000000000002, | |
| "grad_norm": 0.6155795454978943, | |
| "learning_rate": 4.82472382841092e-06, | |
| "loss": 0.8781, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.682, | |
| "grad_norm": 0.5147539973258972, | |
| "learning_rate": 4.789687557358676e-06, | |
| "loss": 0.8741, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.688, | |
| "grad_norm": 0.45784464478492737, | |
| "learning_rate": 4.75466162836291e-06, | |
| "loss": 0.7455, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.694, | |
| "grad_norm": 0.5640389323234558, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.8629, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.5667287707328796, | |
| "learning_rate": 4.684647685509221e-06, | |
| "loss": 0.8333, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.706, | |
| "grad_norm": 0.5320158004760742, | |
| "learning_rate": 4.649663114567663e-06, | |
| "loss": 0.7671, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.712, | |
| "grad_norm": 0.4992581903934479, | |
| "learning_rate": 4.61469577134602e-06, | |
| "loss": 0.8266, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.718, | |
| "grad_norm": 0.6357414126396179, | |
| "learning_rate": 4.579747375353763e-06, | |
| "loss": 0.9051, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.724, | |
| "grad_norm": 0.4872035086154938, | |
| "learning_rate": 4.54481964516865e-06, | |
| "loss": 0.7786, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.5206552147865295, | |
| "learning_rate": 4.509914298352197e-06, | |
| "loss": 0.7692, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 0.5295038819313049, | |
| "learning_rate": 4.475033051365228e-06, | |
| "loss": 0.8352, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.742, | |
| "grad_norm": 0.5258405804634094, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.8252, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.748, | |
| "grad_norm": 0.5072481632232666, | |
| "learning_rate": 4.405349716713165e-06, | |
| "loss": 0.8048, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.754, | |
| "grad_norm": 0.5262322425842285, | |
| "learning_rate": 4.3705510557068746e-06, | |
| "loss": 0.8922, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.46330228447914124, | |
| "learning_rate": 4.335783347679162e-06, | |
| "loss": 0.7183, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.766, | |
| "grad_norm": 0.6510980129241943, | |
| "learning_rate": 4.3010483023225045e-06, | |
| "loss": 0.8822, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.772, | |
| "grad_norm": 0.47852858901023865, | |
| "learning_rate": 4.266347627723192e-06, | |
| "loss": 0.7692, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.778, | |
| "grad_norm": 0.5477906465530396, | |
| "learning_rate": 4.231683030277349e-06, | |
| "loss": 0.8841, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.784, | |
| "grad_norm": 0.5225306749343872, | |
| "learning_rate": 4.197056214607016e-06, | |
| "loss": 0.7784, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.6084046363830566, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.9245, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.796, | |
| "grad_norm": 0.4767346978187561, | |
| "learning_rate": 4.1279227377077465e-06, | |
| "loss": 0.799, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.802, | |
| "grad_norm": 0.4934868812561035, | |
| "learning_rate": 4.0934194760985095e-06, | |
| "loss": 0.7813, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.808, | |
| "grad_norm": 0.5745190382003784, | |
| "learning_rate": 4.058960795337001e-06, | |
| "loss": 0.8129, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.814, | |
| "grad_norm": 0.618204653263092, | |
| "learning_rate": 4.02454838991936e-06, | |
| "loss": 0.7941, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.8199999999999998, | |
| "grad_norm": 0.5389032363891602, | |
| "learning_rate": 3.990183952066151e-06, | |
| "loss": 0.8791, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.826, | |
| "grad_norm": 0.515533983707428, | |
| "learning_rate": 3.955869171639151e-06, | |
| "loss": 0.771, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.5190464854240417, | |
| "learning_rate": 3.921605736058238e-06, | |
| "loss": 0.8339, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.838, | |
| "grad_norm": 0.547839343547821, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.8521, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.8439999999999999, | |
| "grad_norm": 0.5869026780128479, | |
| "learning_rate": 3.853239636407012e-06, | |
| "loss": 0.8309, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.4525587856769562, | |
| "learning_rate": 3.81914033422083e-06, | |
| "loss": 0.7261, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.49729108810424805, | |
| "learning_rate": 3.7850991004836813e-06, | |
| "loss": 0.8609, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.862, | |
| "grad_norm": 0.5660340189933777, | |
| "learning_rate": 3.751117609163865e-06, | |
| "loss": 0.8973, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.8679999999999999, | |
| "grad_norm": 0.48715630173683167, | |
| "learning_rate": 3.7171975312918674e-06, | |
| "loss": 0.7725, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.874, | |
| "grad_norm": 0.5296438336372375, | |
| "learning_rate": 3.683340534878176e-06, | |
| "loss": 0.7935, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.4918104410171509, | |
| "learning_rate": 3.6495482848312745e-06, | |
| "loss": 0.8476, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.8860000000000001, | |
| "grad_norm": 0.5101162791252136, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.8902, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.892, | |
| "grad_norm": 0.5076822638511658, | |
| "learning_rate": 3.5821646674706124e-06, | |
| "loss": 0.851, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.8980000000000001, | |
| "grad_norm": 0.48573198914527893, | |
| "learning_rate": 3.5485766137276894e-06, | |
| "loss": 0.8573, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.904, | |
| "grad_norm": 0.423879474401474, | |
| "learning_rate": 3.5150599333302826e-06, | |
| "loss": 0.7711, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.9100000000000001, | |
| "grad_norm": 0.43153077363967896, | |
| "learning_rate": 3.4816162744519266e-06, | |
| "loss": 0.7175, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.916, | |
| "grad_norm": 0.5034720301628113, | |
| "learning_rate": 3.4482472816753404e-06, | |
| "loss": 0.8976, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.9220000000000002, | |
| "grad_norm": 0.529129683971405, | |
| "learning_rate": 3.4149545959115604e-06, | |
| "loss": 0.8046, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 0.5979866981506348, | |
| "learning_rate": 3.3817398543192426e-06, | |
| "loss": 0.872, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.9340000000000002, | |
| "grad_norm": 0.4693806767463684, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.8611, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.4347771406173706, | |
| "learning_rate": 3.3155507330389004e-06, | |
| "loss": 0.7822, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.946, | |
| "grad_norm": 0.5176924467086792, | |
| "learning_rate": 3.2825796081826943e-06, | |
| "loss": 0.8046, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 0.5300335884094238, | |
| "learning_rate": 3.2496929370015383e-06, | |
| "loss": 0.8355, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.958, | |
| "grad_norm": 0.4729841649532318, | |
| "learning_rate": 3.216892336688435e-06, | |
| "loss": 0.8173, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.964, | |
| "grad_norm": 0.4969511330127716, | |
| "learning_rate": 3.184179420203877e-06, | |
| "loss": 0.8948, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.5224483013153076, | |
| "learning_rate": 3.1515557961965254e-06, | |
| "loss": 0.779, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.976, | |
| "grad_norm": 0.4701511561870575, | |
| "learning_rate": 3.119023068924115e-06, | |
| "loss": 0.8453, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.982, | |
| "grad_norm": 0.44776278734207153, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.8481, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.988, | |
| "grad_norm": 0.4345719814300537, | |
| "learning_rate": 3.0542366991872546e-06, | |
| "loss": 0.7723, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.994, | |
| "grad_norm": 0.5247130393981934, | |
| "learning_rate": 3.021986242574707e-06, | |
| "loss": 0.9373, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.002, | |
| "grad_norm": 0.9623157978057861, | |
| "learning_rate": 2.989833054244232e-06, | |
| "loss": 1.2604, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.008, | |
| "grad_norm": 0.454515278339386, | |
| "learning_rate": 2.95777871532002e-06, | |
| "loss": 0.7464, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.014, | |
| "grad_norm": 0.45794516801834106, | |
| "learning_rate": 2.9258248020653617e-06, | |
| "loss": 0.7502, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.4778282344341278, | |
| "learning_rate": 2.893972885805148e-06, | |
| "loss": 0.8084, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.026, | |
| "grad_norm": 0.5253900289535522, | |
| "learning_rate": 2.862224532848591e-06, | |
| "loss": 0.7463, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.46551772952079773, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.7751, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.038, | |
| "grad_norm": 0.5346385836601257, | |
| "learning_rate": 2.799044756543054e-06, | |
| "loss": 0.7935, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.044, | |
| "grad_norm": 0.45891788601875305, | |
| "learning_rate": 2.7676164400421864e-06, | |
| "loss": 0.7491, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.5041545629501343, | |
| "learning_rate": 2.7362979003884154e-06, | |
| "loss": 0.8321, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.056, | |
| "grad_norm": 0.452737033367157, | |
| "learning_rate": 2.705090677662311e-06, | |
| "loss": 0.7378, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.062, | |
| "grad_norm": 0.5380454659461975, | |
| "learning_rate": 2.6739963064704634e-06, | |
| "loss": 0.9362, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.068, | |
| "grad_norm": 0.47858721017837524, | |
| "learning_rate": 2.6430163158700116e-06, | |
| "loss": 0.7629, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.074, | |
| "grad_norm": 0.44490736722946167, | |
| "learning_rate": 2.61215222929347e-06, | |
| "loss": 0.6881, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.5407552719116211, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.8796, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.086, | |
| "grad_norm": 0.5123006701469421, | |
| "learning_rate": 2.5507778333697817e-06, | |
| "loss": 0.7191, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.092, | |
| "grad_norm": 0.45387887954711914, | |
| "learning_rate": 2.520270542091663e-06, | |
| "loss": 0.7626, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.098, | |
| "grad_norm": 0.48711925745010376, | |
| "learning_rate": 2.4898851908270903e-06, | |
| "loss": 0.804, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.104, | |
| "grad_norm": 0.4527442157268524, | |
| "learning_rate": 2.4596232737673544e-06, | |
| "loss": 0.7554, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.45920926332473755, | |
| "learning_rate": 2.429486279033892e-06, | |
| "loss": 0.7417, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.116, | |
| "grad_norm": 0.4968525767326355, | |
| "learning_rate": 2.3994756886051267e-06, | |
| "loss": 0.8689, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.122, | |
| "grad_norm": 0.4496273994445801, | |
| "learning_rate": 2.3695929782435845e-06, | |
| "loss": 0.7424, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.48898282647132874, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.8748, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.134, | |
| "grad_norm": 0.43496692180633545, | |
| "learning_rate": 2.3102170692576593e-06, | |
| "loss": 0.6957, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.5999041199684143, | |
| "learning_rate": 2.280726790427258e-06, | |
| "loss": 0.8587, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.146, | |
| "grad_norm": 0.44731348752975464, | |
| "learning_rate": 2.2513702311084595e-06, | |
| "loss": 0.7431, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.152, | |
| "grad_norm": 0.5178326368331909, | |
| "learning_rate": 2.2221488349019903e-06, | |
| "loss": 0.8245, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.158, | |
| "grad_norm": 0.4318412244319916, | |
| "learning_rate": 2.1930640387619616e-06, | |
| "loss": 0.7456, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.164, | |
| "grad_norm": 0.4501084089279175, | |
| "learning_rate": 2.1641172729252206e-06, | |
| "loss": 0.79, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.45479288697242737, | |
| "learning_rate": 2.1353099608410095e-06, | |
| "loss": 0.6949, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5368492007255554, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.8474, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.182, | |
| "grad_norm": 0.4430352449417114, | |
| "learning_rate": 2.078119357369487e-06, | |
| "loss": 0.7613, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.188, | |
| "grad_norm": 0.3970995545387268, | |
| "learning_rate": 2.04973887831436e-06, | |
| "loss": 0.7985, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.194, | |
| "grad_norm": 0.47133341431617737, | |
| "learning_rate": 2.0215034775378336e-06, | |
| "loss": 0.7873, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.503981351852417, | |
| "learning_rate": 1.9934145435079705e-06, | |
| "loss": 0.8465, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.206, | |
| "grad_norm": 0.41882044076919556, | |
| "learning_rate": 1.965473457490372e-06, | |
| "loss": 0.7078, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.212, | |
| "grad_norm": 0.4320144057273865, | |
| "learning_rate": 1.9376815934802496e-06, | |
| "loss": 0.7788, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.218, | |
| "grad_norm": 0.5286938548088074, | |
| "learning_rate": 1.9100403181348687e-06, | |
| "loss": 0.8697, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.44443029165267944, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.7119, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.5040541291236877, | |
| "learning_rate": 1.8552149629747557e-06, | |
| "loss": 0.7846, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.2359999999999998, | |
| "grad_norm": 0.4279659390449524, | |
| "learning_rate": 1.8280335791817733e-06, | |
| "loss": 0.8005, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.242, | |
| "grad_norm": 0.4774853587150574, | |
| "learning_rate": 1.8010081759644538e-06, | |
| "loss": 0.8352, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.248, | |
| "grad_norm": 0.4227093756198883, | |
| "learning_rate": 1.7741400822895633e-06, | |
| "loss": 0.8064, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.254, | |
| "grad_norm": 0.4951332211494446, | |
| "learning_rate": 1.747430619388209e-06, | |
| "loss": 0.8362, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.4470813274383545, | |
| "learning_rate": 1.7208811006908798e-06, | |
| "loss": 0.8408, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.266, | |
| "grad_norm": 0.3740614354610443, | |
| "learning_rate": 1.6944928317628512e-06, | |
| "loss": 0.7427, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 0.4321645200252533, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.8121, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.278, | |
| "grad_norm": 0.43954160809516907, | |
| "learning_rate": 1.642205225764908e-06, | |
| "loss": 0.7566, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.284, | |
| "grad_norm": 0.47293248772621155, | |
| "learning_rate": 1.6163084599236278e-06, | |
| "loss": 0.8505, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.44587525725364685, | |
| "learning_rate": 1.5905780861824748e-06, | |
| "loss": 0.7285, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.296, | |
| "grad_norm": 0.4518946707248688, | |
| "learning_rate": 1.5650153698254916e-06, | |
| "loss": 0.779, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.302, | |
| "grad_norm": 0.4540093243122101, | |
| "learning_rate": 1.5396215678922222e-06, | |
| "loss": 0.8963, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.308, | |
| "grad_norm": 0.42568331956863403, | |
| "learning_rate": 1.514397929115884e-06, | |
| "loss": 0.6954, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.314, | |
| "grad_norm": 0.44443783164024353, | |
| "learning_rate": 1.4893456938619743e-06, | |
| "loss": 0.8091, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.45444992184638977, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.7894, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.326, | |
| "grad_norm": 0.3983505666255951, | |
| "learning_rate": 1.439760353179223e-06, | |
| "loss": 0.7593, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.332, | |
| "grad_norm": 0.3989236056804657, | |
| "learning_rate": 1.4152296860958641e-06, | |
| "loss": 0.8326, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.338, | |
| "grad_norm": 0.4200502038002014, | |
| "learning_rate": 1.3908752991059854e-06, | |
| "loss": 0.8052, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.344, | |
| "grad_norm": 0.44380584359169006, | |
| "learning_rate": 1.3666983898298659e-06, | |
| "loss": 0.8043, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.3878370225429535, | |
| "learning_rate": 1.3427001471603623e-06, | |
| "loss": 0.7374, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.356, | |
| "grad_norm": 0.417605996131897, | |
| "learning_rate": 1.3188817512044544e-06, | |
| "loss": 0.871, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.362, | |
| "grad_norm": 0.3884265720844269, | |
| "learning_rate": 1.2952443732252058e-06, | |
| "loss": 0.7358, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.3953470289707184, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.7017, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.374, | |
| "grad_norm": 0.47011804580688477, | |
| "learning_rate": 1.2485173116842432e-06, | |
| "loss": 0.8144, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.43607431650161743, | |
| "learning_rate": 1.225429925912921e-06, | |
| "loss": 0.7471, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.386, | |
| "grad_norm": 0.4099930226802826, | |
| "learning_rate": 1.2025281535860433e-06, | |
| "loss": 0.8123, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.392, | |
| "grad_norm": 0.4191509485244751, | |
| "learning_rate": 1.1798131208919628e-06, | |
| "loss": 0.7882, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.398, | |
| "grad_norm": 0.43705156445503235, | |
| "learning_rate": 1.1572859448361602e-06, | |
| "loss": 0.8712, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.404, | |
| "grad_norm": 0.4271179139614105, | |
| "learning_rate": 1.134947733186315e-06, | |
| "loss": 0.7635, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.426925927400589, | |
| "learning_rate": 1.1127995844178385e-06, | |
| "loss": 0.7843, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.408716082572937, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.8381, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.422, | |
| "grad_norm": 0.3790528476238251, | |
| "learning_rate": 1.069077822641622e-06, | |
| "loss": 0.7518, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.428, | |
| "grad_norm": 0.4429261386394501, | |
| "learning_rate": 1.047506359639483e-06, | |
| "loss": 0.7886, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.434, | |
| "grad_norm": 0.4311138987541199, | |
| "learning_rate": 1.0261292594241873e-06, | |
| "loss": 0.7454, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.4488796591758728, | |
| "learning_rate": 1.004947573208756e-06, | |
| "loss": 0.7773, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.446, | |
| "grad_norm": 0.4116864800453186, | |
| "learning_rate": 9.83962342596776e-07, | |
| "loss": 0.7395, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.452, | |
| "grad_norm": 0.5070677399635315, | |
| "learning_rate": 9.631745995311881e-07, | |
| "loss": 0.7972, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.458, | |
| "grad_norm": 0.43356719613075256, | |
| "learning_rate": 9.42585366243537e-07, | |
| "loss": 0.7628, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 0.41195109486579895, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.8116, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.4699999999999998, | |
| "grad_norm": 0.3985452651977539, | |
| "learning_rate": 9.020064690701069e-07, | |
| "loss": 0.8279, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.476, | |
| "grad_norm": 0.47739341855049133, | |
| "learning_rate": 8.820188006404268e-07, | |
| "loss": 0.8178, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.482, | |
| "grad_norm": 0.4421622157096863, | |
| "learning_rate": 8.622336328027553e-07, | |
| "loss": 0.8169, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.488, | |
| "grad_norm": 0.4162546396255493, | |
| "learning_rate": 8.426519384872733e-07, | |
| "loss": 0.8104, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.4939999999999998, | |
| "grad_norm": 0.36846503615379333, | |
| "learning_rate": 8.232746806184116e-07, | |
| "loss": 0.7281, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.3913734257221222, | |
| "learning_rate": 8.041028120674894e-07, | |
| "loss": 0.7757, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.5060000000000002, | |
| "grad_norm": 0.4026607573032379, | |
| "learning_rate": 7.851372756058662e-07, | |
| "loss": 0.8185, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.40786227583885193, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.8264, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.518, | |
| "grad_norm": 0.3854236602783203, | |
| "learning_rate": 7.478289192584765e-07, | |
| "loss": 0.6955, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.524, | |
| "grad_norm": 0.4671294391155243, | |
| "learning_rate": 7.294879340008632e-07, | |
| "loss": 0.8344, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.5300000000000002, | |
| "grad_norm": 0.4368779957294464, | |
| "learning_rate": 7.113569499986401e-07, | |
| "loss": 0.7613, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.536, | |
| "grad_norm": 0.40010130405426025, | |
| "learning_rate": 6.934368588379553e-07, | |
| "loss": 0.6928, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.542, | |
| "grad_norm": 0.404140442609787, | |
| "learning_rate": 6.757285417343601e-07, | |
| "loss": 0.7997, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.548, | |
| "grad_norm": 0.39803242683410645, | |
| "learning_rate": 6.582328694894729e-07, | |
| "loss": 0.7461, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.5540000000000003, | |
| "grad_norm": 0.4443919360637665, | |
| "learning_rate": 6.409507024481615e-07, | |
| "loss": 0.8448, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.43407124280929565, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.8036, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.566, | |
| "grad_norm": 0.4071224331855774, | |
| "learning_rate": 6.070302728186428e-07, | |
| "loss": 0.7716, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.572, | |
| "grad_norm": 0.3952639698982239, | |
| "learning_rate": 5.903936782582253e-07, | |
| "loss": 0.7355, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.578, | |
| "grad_norm": 0.4198220372200012, | |
| "learning_rate": 5.739739248749398e-07, | |
| "loss": 0.818, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.584, | |
| "grad_norm": 0.43490225076675415, | |
| "learning_rate": 5.577718201056392e-07, | |
| "loss": 0.7708, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.4053814113140106, | |
| "learning_rate": 5.41788160684365e-07, | |
| "loss": 0.7872, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.596, | |
| "grad_norm": 0.39600592851638794, | |
| "learning_rate": 5.260237326031698e-07, | |
| "loss": 0.8138, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.602, | |
| "grad_norm": 0.40536797046661377, | |
| "learning_rate": 5.104793110734668e-07, | |
| "loss": 0.7579, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.3923257887363434, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.7697, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.614, | |
| "grad_norm": 0.44755300879478455, | |
| "learning_rate": 4.800535343827834e-07, | |
| "loss": 0.8134, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.41581737995147705, | |
| "learning_rate": 4.651736754009972e-07, | |
| "loss": 0.8371, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.626, | |
| "grad_norm": 0.344684898853302, | |
| "learning_rate": 4.5051681525551726e-07, | |
| "loss": 0.6842, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.632, | |
| "grad_norm": 0.44776731729507446, | |
| "learning_rate": 4.3608367469340553e-07, | |
| "loss": 0.8512, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.638, | |
| "grad_norm": 0.40358269214630127, | |
| "learning_rate": 4.218749634603769e-07, | |
| "loss": 0.8017, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.644, | |
| "grad_norm": 0.3789862394332886, | |
| "learning_rate": 4.078913802658946e-07, | |
| "loss": 0.7429, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.4059142470359802, | |
| "learning_rate": 3.941336127488149e-07, | |
| "loss": 0.9286, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 0.47864121198654175, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.7965, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.662, | |
| "grad_norm": 0.34777745604515076, | |
| "learning_rate": 3.672982197468894e-07, | |
| "loss": 0.6891, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.668, | |
| "grad_norm": 0.4154059588909149, | |
| "learning_rate": 3.542219138851094e-07, | |
| "loss": 0.8237, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.674, | |
| "grad_norm": 0.4375707805156708, | |
| "learning_rate": 3.413740628819673e-07, | |
| "loss": 0.8458, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.3523346483707428, | |
| "learning_rate": 3.287552985270015e-07, | |
| "loss": 0.6672, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.686, | |
| "grad_norm": 0.43038198351860046, | |
| "learning_rate": 3.1636624134447347e-07, | |
| "loss": 0.9044, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.692, | |
| "grad_norm": 0.4230746030807495, | |
| "learning_rate": 3.0420750056286195e-07, | |
| "loss": 0.7871, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.698, | |
| "grad_norm": 0.4082009792327881, | |
| "learning_rate": 2.9227967408489653e-07, | |
| "loss": 0.8072, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.41478103399276733, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.8227, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.37652283906936646, | |
| "learning_rate": 2.691190988462522e-07, | |
| "loss": 0.7586, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.716, | |
| "grad_norm": 0.46765658259391785, | |
| "learning_rate": 2.5788748900048676e-07, | |
| "loss": 0.8312, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.722, | |
| "grad_norm": 0.39574170112609863, | |
| "learning_rate": 2.468890712321864e-07, | |
| "loss": 0.8476, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.7279999999999998, | |
| "grad_norm": 0.39904430508613586, | |
| "learning_rate": 2.3612438638551837e-07, | |
| "loss": 0.7627, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.734, | |
| "grad_norm": 0.369881272315979, | |
| "learning_rate": 2.2559396381089836e-07, | |
| "loss": 0.7599, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.3744626045227051, | |
| "learning_rate": 2.152983213389559e-07, | |
| "loss": 0.7736, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.746, | |
| "grad_norm": 0.4368973672389984, | |
| "learning_rate": 2.0523796525507622e-07, | |
| "loss": 0.8125, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.43715760111808777, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.7794, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.758, | |
| "grad_norm": 0.41397351026535034, | |
| "learning_rate": 1.8582507951800277e-07, | |
| "loss": 0.793, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.7640000000000002, | |
| "grad_norm": 0.39155086874961853, | |
| "learning_rate": 1.7647350448812105e-07, | |
| "loss": 0.7349, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 0.4406411647796631, | |
| "learning_rate": 1.6735912504598384e-07, | |
| "loss": 0.8252, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.776, | |
| "grad_norm": 0.41271498799324036, | |
| "learning_rate": 1.5848238938869332e-07, | |
| "loss": 0.8621, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.782, | |
| "grad_norm": 0.3822709321975708, | |
| "learning_rate": 1.4984373402728014e-07, | |
| "loss": 0.7701, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.7880000000000003, | |
| "grad_norm": 0.42628178000450134, | |
| "learning_rate": 1.4144358376524504e-07, | |
| "loss": 0.7945, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.794, | |
| "grad_norm": 0.40755292773246765, | |
| "learning_rate": 1.3328235167766545e-07, | |
| "loss": 0.7498, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.4093887209892273, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.8173, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.806, | |
| "grad_norm": 0.3912396728992462, | |
| "learning_rate": 1.1767823556276648e-07, | |
| "loss": 0.7292, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.8120000000000003, | |
| "grad_norm": 0.39809033274650574, | |
| "learning_rate": 1.10236118863562e-07, | |
| "loss": 0.8378, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.818, | |
| "grad_norm": 0.39227575063705444, | |
| "learning_rate": 1.0303445495730868e-07, | |
| "loss": 0.8025, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.824, | |
| "grad_norm": 0.3954123854637146, | |
| "learning_rate": 9.607359798384785e-08, | |
| "loss": 0.8015, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.37289682030677795, | |
| "learning_rate": 8.935389024140485e-08, | |
| "loss": 0.6498, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 2.836, | |
| "grad_norm": 0.4187394976615906, | |
| "learning_rate": 8.287566216975795e-08, | |
| "loss": 0.8355, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.842, | |
| "grad_norm": 0.389783650636673, | |
| "learning_rate": 7.663923233398928e-08, | |
| "loss": 0.8568, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 0.40914517641067505, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.7667, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.854, | |
| "grad_norm": 0.4102424681186676, | |
| "learning_rate": 6.489298216352913e-08, | |
| "loss": 0.7791, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.42990586161613464, | |
| "learning_rate": 5.938373944745612e-08, | |
| "loss": 0.8003, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.866, | |
| "grad_norm": 0.3794080317020416, | |
| "learning_rate": 5.411745017609493e-08, | |
| "loss": 0.8502, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 2.872, | |
| "grad_norm": 0.3640872538089752, | |
| "learning_rate": 4.909437331777178e-08, | |
| "loss": 0.667, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.878, | |
| "grad_norm": 0.4412924647331238, | |
| "learning_rate": 4.431475588090872e-08, | |
| "loss": 0.8964, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 2.884, | |
| "grad_norm": 0.38228335976600647, | |
| "learning_rate": 3.977883290187667e-08, | |
| "loss": 0.7143, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.42708632349967957, | |
| "learning_rate": 3.548682743344079e-08, | |
| "loss": 0.8222, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.39925622940063477, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.7455, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.902, | |
| "grad_norm": 0.4083234667778015, | |
| "learning_rate": 2.7635401256149163e-08, | |
| "loss": 0.8795, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 2.908, | |
| "grad_norm": 0.3775049149990082, | |
| "learning_rate": 2.4076366639015914e-08, | |
| "loss": 0.6865, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.914, | |
| "grad_norm": 0.40156200528144836, | |
| "learning_rate": 2.0762021696933975e-08, | |
| "loss": 0.7257, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.4004535675048828, | |
| "learning_rate": 1.769252941190458e-08, | |
| "loss": 0.8059, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.926, | |
| "grad_norm": 0.38527584075927734, | |
| "learning_rate": 1.4868040725365407e-08, | |
| "loss": 0.8935, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 2.932, | |
| "grad_norm": 0.39457565546035767, | |
| "learning_rate": 1.2288694530769862e-08, | |
| "loss": 0.7953, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.9379999999999997, | |
| "grad_norm": 0.42659154534339905, | |
| "learning_rate": 9.954617666758648e-09, | |
| "loss": 0.7744, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.4101230502128601, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.7164, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.3969820737838745, | |
| "learning_rate": 6.022718974137976e-09, | |
| "loss": 0.8194, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.956, | |
| "grad_norm": 0.40510863065719604, | |
| "learning_rate": 4.4250904955656095e-09, | |
| "loss": 0.8323, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.9619999999999997, | |
| "grad_norm": 0.3878780007362366, | |
| "learning_rate": 3.0731180381399216e-09, | |
| "loss": 0.7787, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.968, | |
| "grad_norm": 0.3961254954338074, | |
| "learning_rate": 1.9668680847356735e-09, | |
| "loss": 0.8071, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.974, | |
| "grad_norm": 0.38420915603637695, | |
| "learning_rate": 1.1063950348888519e-09, | |
| "loss": 0.7125, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.44326284527778625, | |
| "learning_rate": 4.91741202124918e-10, | |
| "loss": 0.7699, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.9859999999999998, | |
| "grad_norm": 0.43767398595809937, | |
| "learning_rate": 1.22936811877139e-10, | |
| "loss": 0.8057, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.992, | |
| "grad_norm": 0.3621894121170044, | |
| "learning_rate": 0.0, | |
| "loss": 0.7188, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.992, | |
| "step": 498, | |
| "total_flos": 499815018037248.0, | |
| "train_loss": 0.8564689170165234, | |
| "train_runtime": 27758.1426, | |
| "train_samples_per_second": 1.729, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 498, | |
| "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": 499815018037248.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |