| { | |
| "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": 7.511359905912763, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.1926, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 7.28262435081321, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.1781, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 7.457651222427291, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.1735, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 6.861950997514667, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.127, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 5.015833529845837, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 1.0954, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.0315795853784198, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.0403, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.373470688343003, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 1.0216, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 3.867248078079527, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 1.0108, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.7854002771061452, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0268, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 3.6023698978814034, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9537, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.4294152602833585, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.9501, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.0371640781234364, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.942, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.4146993273227397, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8908, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.9909841147739862, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8717, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.320385240709792, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8521, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 2.1242317195998304, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.8543, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4809652658218375, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.8451, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.294043890814804, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.8273, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.3569542854579595, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7948, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.0568847200469609, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.8236, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.1472933431486327, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7837, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.9816002705686646, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7843, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.00482997620653, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7548, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8691457531114991, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7813, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.0047344766786888, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7946, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.8754870205684157, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7581, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.9452018409115547, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7749, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.8223965474970097, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.8086, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.9144628612846286, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7511, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.8928961205828579, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.765, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.7584371273838272, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.7564, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.8942296049296377, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.7491, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.8627289582557884, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.7646, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.9869187027630572, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.7555, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7247315245850969, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.7443, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.1883748183923173, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.7631, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.9250970210131236, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.7561, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.8118657195876166, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.7504, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.7880519981498711, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7467, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.7466274201390836, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.7685, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.8124659466824775, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.7321, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.7355951940953518, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.7534, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.8896957560011128, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.7507, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.8771070685697864, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.7405, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.8419310540856586, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.788, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7779348563101531, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.7341, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.9036635505985329, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.7522, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.175352616172817, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.7362, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.721612411957382, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.763, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.1163554642177993, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.7596, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.0728158964937573, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.7303, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.710654386180622, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.76, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.9933870258540802, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.7483, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.9912456612159388, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.7412, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.7659934869807872, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.7421, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.8860610152269798, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.7062, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.032756261646297, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.7769, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.8083649124559691, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.741, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.8268800650406273, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.7318, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.6779236481233356, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.7273, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.8350494452645356, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.7217, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.7429158377538772, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.7219, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.9676173355627138, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.7186, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.8976458517522923, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.732, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7902905188479334, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.7448, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.8175451429579409, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.7223, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.924004494956127, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.7293, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6844398628724695, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.7508, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.7875102086141258, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.7199, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6156421168279543, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.7101, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.6785037490419621, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.715, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6906120279127116, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.7319, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.7945205223793728, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.7206, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.7078831584936823, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.7169, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.8759629031643444, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.7181, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7949597403256198, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.7147, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.8578764602490665, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.7237, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.9121215466460237, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.7404, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.8286012742877019, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.6044, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.857156702901733, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.559, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.863990123852963, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5802, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.8108250146911395, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.548, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7945029299398011, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5617, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8770636966737697, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.569, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.9020139029372768, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5579, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.7688602186580494, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5638, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.9616184912950552, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5663, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.8561360976859292, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5776, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.9572760221561845, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5607, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.7915050561494705, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5345, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.8494273575243793, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5608, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.7184926499002829, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5508, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.8578429399631107, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5585, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.713985140098294, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.551, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.7312489232084477, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5908, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.9680097022393541, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5826, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.7635181239808696, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5524, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.8955330415394881, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5673, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.825016031167963, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5417, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.8851484177724676, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5444, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.9504008665165626, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5687, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.7376749398622102, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5515, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.7434147529578286, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5447, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.7123791554300906, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5491, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.723159636246862, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5655, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.7346052751848078, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5344, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.6770195867050284, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.541, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.836925564357309, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5465, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.7745553349971575, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5621, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.6613253410706297, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5532, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.852624548927096, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5424, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.7544998597715881, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5415, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.836252122521665, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5483, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.8349756975894402, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5348, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.7035166790179582, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5422, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.734808884191683, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5336, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5741373010783888, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5766, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.6719211603856002, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.543, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.6512853360546929, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5616, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.6733165535088171, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5303, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.670465510019021, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5396, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.6392214761021744, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5602, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.6356188581391619, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5504, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5731682601422339, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5476, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.6289952487122805, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5496, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.6349204454340878, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5329, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.6801347942930306, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5401, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.7012846715222437, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5695, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.6017136521163042, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5479, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.6891024794532379, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5644, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.586119819092562, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5612, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.6004177061714289, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5476, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.6679095830207941, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5525, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5738863516141487, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5591, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.6402257856859631, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5659, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.587838134705723, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5373, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.5677386377615509, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5366, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5610177288697289, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5453, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.550323142541207, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5494, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.6378838173182162, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5389, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5604437488955527, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5797, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.63381863366934, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5353, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.6346777898181406, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5375, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.5837487186299176, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5468, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.6255787463575647, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5423, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.583750423987297, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5373, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.6586802656570658, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.563, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.6514315479393031, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5413, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.7357144845772984, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5497, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.6961689372034049, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5563, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.6387171773459921, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5374, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5757774690936177, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5483, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.6183311529307285, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5526, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.5754694357118991, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5523, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.7850641752514341, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5653, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.6103048207999278, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5518, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.8345155097205981, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.4758, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.5205065406260636, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4029, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.0696613412523857, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.403, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7981066273258726, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3994, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6639873372149179, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3927, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.7323974048851529, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4048, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6254738536750247, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3755, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.7165577457987822, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.4098, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5965717307567397, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.4102, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.6641111890663406, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3807, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.6072104456800083, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3691, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.6046279519430399, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.393, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.7200670036750239, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3854, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5581545293812528, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3817, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.7055010190634289, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3927, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5770902875283219, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3941, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5831365353628903, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3556, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5850198656743264, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3769, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4803424586912255, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4058, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.6373057616264917, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3943, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.5303174061456862, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3864, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.5659878034089584, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.365, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5131485065746315, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.39, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.5486897838488941, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3939, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5645917491254367, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.4068, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.5495711512543889, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3946, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.5092064000074611, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3731, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.5371517653905579, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3747, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.5182412650474031, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3801, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.557350987982833, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3908, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.490155854864711, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3598, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.5452219843473595, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3808, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5432083893584774, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3758, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.5671196866696925, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3798, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.4790374323005663, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3839, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.5195035816219602, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.375, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.545646192631747, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3798, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.5728899421128788, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.4013, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4918766606175862, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3806, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.5393410680879338, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.387, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.48824237902125384, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.369, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.4860113062664166, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.4041, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.5022741070656771, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.4187, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4740542493805564, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3704, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.5318635727037249, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.4099, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.5174406023922865, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3929, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.5639182715541546, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4032, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.5292109083435944, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3734, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.5428611384224525, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.4241, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.575219449586423, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3936, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.5092558749130109, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3822, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.48120017035694174, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3844, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.4504453422060441, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3685, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.5041740213571579, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3969, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.4833580780256183, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3752, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.4659911711293527, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3542, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.50747462828893, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4179, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.46573616967371484, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3763, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4592547351914383, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3917, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.517512402943037, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3566, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.5177790795333941, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3676, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.4576139040751015, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3701, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.48232412101183647, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.398, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.5201088963477126, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3797, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.49254627321580385, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3824, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.5030772574222968, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3775, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.48184491872276675, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.363, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.49619969853454565, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3983, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.49724141647367104, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3641, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.47376207103226875, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3949, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.5065029584090022, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3888, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.4588134303289056, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3895, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.537376081347621, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3684, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.4849300819024804, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3768, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.48416733957784724, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3935, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.522010253606809, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3664, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.4634102160100665, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3915, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.4817855075430708, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3614, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.5178454128099733, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.3297, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.5575195436903574, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2702, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.47790204293401833, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2713, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.7256552154221274, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2641, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5898949203734531, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.281, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4991251499683754, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2593, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5851412198606553, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2594, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5841561653070572, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2849, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.530064468861995, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.24, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4921647645941968, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2766, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.5357430647881002, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2726, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.47013652169964254, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2633, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.4422514503491578, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.268, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4595431040165629, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2756, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.4638833440176158, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2705, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.42903002169219484, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2584, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.4445379357533598, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.3012, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.4513959212284946, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2687, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.4453400354617768, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2912, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.4377934600354535, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2785, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.4345193057218717, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2786, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.397408915198694, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2843, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.39699994407473127, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2724, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.41168809425208336, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2829, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.39915929776759457, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2544, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.42594282685098744, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.268, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.36119318836337994, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.243, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.37376871402766004, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2596, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.39129428343302936, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.292, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3775317504712681, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2605, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.4647923919562145, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2614, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.4019796497891929, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2816, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.35826815671651413, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2647, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.36239369494432616, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2555, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.42665721079509045, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2768, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3714679540593856, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.3036, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3713473261225915, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.281, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.41932128838063465, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2573, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.4100660037606104, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.31, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.36373907463021976, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2821, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.35646977396365903, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2734, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.359585740902966, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2627, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3480365214389296, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2428, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.35823902488400805, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2671, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.36386530769864606, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.288, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.36272193061996166, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2582, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.34999164197126587, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2843, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.42504123595377974, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2646, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3631369201277819, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2635, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.34564819696603877, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2539, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3319639926747861, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2883, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.35496864101383674, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2715, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.35825302222496747, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2673, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3432234362926089, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2844, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3632957597794797, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2522, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3663440079341007, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2576, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.33063789598953014, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2517, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.3530473980233189, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2811, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.35309089688903844, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.285, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.3541914773769567, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2698, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.35550936211843454, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2321, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.35522264755823996, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2605, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.3628443127873584, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2911, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.335750342148056, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2765, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3454944454030594, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2405, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.35005653196076403, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2668, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.3377614185036671, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2897, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.33461749854266604, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2296, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.33049134984801565, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2604, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.32257364579402886, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2668, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.3505257775702328, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2704, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.34941456797915443, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2486, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.3497997790474214, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2711, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.35189624751605414, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2828, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.3639251545539861, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2673, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.34077733369189517, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2661, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.32497574289628633, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2534, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.3253370420307855, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2668, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.3565295314357817, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.2516, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.4216092593966632, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2124, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.40470073127793627, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.2029, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.2860069241760031, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2125, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.2990753383402229, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2247, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.43968773553331003, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.236, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.3824809877851468, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.207, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.42972369363504553, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.227, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.3657359469789783, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.2301, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.31557224393201466, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2204, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3029116822429697, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2063, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.32137576259932, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.234, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3485619041982408, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.2216, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.31314179476097936, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.2048, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.3033891867936396, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.2568, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.27517449717479464, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2095, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2692405790931033, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.2047, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.29445031313218717, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1927, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2933155524765344, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2159, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.2880231883657972, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1948, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.27843202803180184, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.2677, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.7918234394823368, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.2285, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2856922622994964, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.2162, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.29270813496135917, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1977, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.2765352040319683, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1868, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2773420935290148, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.2504, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2951989763955348, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.202, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2694819507577761, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.2053, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.26880423039477896, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1948, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.27233541311582243, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.2245, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.30059778886082356, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.209, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2542852865046565, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1966, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.26142277431675487, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.2174, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.28085365331071566, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1929, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.2620057586613848, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.2133, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.28012967225287866, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.2208, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.25022658550404375, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.2352, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.25777005246742474, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.2124, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2776133916654331, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2124, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.27726738839209575, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.2271, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.27877609622454136, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.2268, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2665115712864471, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1867, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.27260219647757444, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.2055, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.25789180521366417, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2074, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.25093686649152924, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.204, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2682259475992922, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2247, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2632131820948456, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2117, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2510605449258651, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1779, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2708835691693638, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.2235, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2571475890919035, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.2344, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2542457515693508, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2108, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2686788559436714, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1922, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.25639444515712656, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2131, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.24758957375451615, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1946, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.2500813766369497, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2193, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2700601426436986, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.2367, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.26217775420557915, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.2015, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.255667212672387, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.2107, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.25482614331996517, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.2038, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.24600788381355781, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.2192, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.27008669578003647, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2205, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.26125735753861795, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.208, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.25290289513173314, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.2176, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.27722115330198216, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.2372, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2715757787558839, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.2041, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.2581407584718076, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.2071, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2550767792625748, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.2006, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.2810305452313246, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2044, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2748382570801458, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2278, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.2611304637437827, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.2109, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.26434707527197937, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.2135, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2565016779068201, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2139, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.26306248826189155, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.2119, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.27944779516296414, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.2132, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2563144728702243, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1981, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2609112371572079, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1905, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.25407745161319095, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2077, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2545177508593149, | |
| "learning_rate": 0.0, | |
| "loss": 0.2084, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 7.101172747442586e+17, | |
| "train_loss": 0.445627332230409, | |
| "train_runtime": 12672.6478, | |
| "train_samples_per_second": 3.946, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "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": 7.101172747442586e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |