2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Model: "SequenceTagger( (embeddings): StackedEmbeddings( (list_embedding_0): WordEmbeddings( 'de' (embedding): Embedding(1000000, 300) ) (list_embedding_1): FlairEmbeddings( (lm): LanguageModel( (drop): Dropout(p=0.25, inplace=False) (encoder): Embedding(275, 100) (rnn): LSTM(100, 2048) ) ) (list_embedding_2): FlairEmbeddings( (lm): LanguageModel( (drop): Dropout(p=0.25, inplace=False) (encoder): Embedding(275, 100) (rnn): LSTM(100, 2048) ) ) ) (word_dropout): WordDropout(p=0.05) (locked_dropout): LockedDropout(p=0.5) (embedding2nn): Linear(in_features=4396, out_features=4396, bias=True) (rnn): LSTM(4396, 256, batch_first=True, bidirectional=True) (linear): Linear(in_features=512, out_features=71, bias=True) (loss_function): ViterbiLoss() (crf): CRF() )" 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Corpus: 420 train + 500 dev + 506 test sentences 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Train: 420 sentences 2023-05-15 21:27:39,581 (train_with_dev=False, train_with_test=False) 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Training Params: 2023-05-15 21:27:39,581 - learning_rate: "0.1" 2023-05-15 21:27:39,581 - mini_batch_size: "4" 2023-05-15 21:27:39,581 - max_epochs: "150" 2023-05-15 21:27:39,581 - shuffle: "True" 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Plugins: 2023-05-15 21:27:39,581 - AnnealOnPlateau | patience: '3', anneal_factor: '0.5', min_learning_rate: '0.0001' 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Final evaluation on model from best epoch (best-model.pt) 2023-05-15 21:27:39,581 - metric: "('micro avg', 'accuracy')" 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Computation: 2023-05-15 21:27:39,581 - compute on device: cuda:0 2023-05-15 21:27:39,581 - embedding storage: cpu 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 Model training base path: "pos-twitter-german-bs4-4" 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:39,581 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:40,533 epoch 1 - iter 10/105 - loss 4.13732332 - time (sec): 0.95 - samples/sec: 579.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:41,490 epoch 1 - iter 20/105 - loss 3.80546391 - time (sec): 1.91 - samples/sec: 605.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:42,512 epoch 1 - iter 30/105 - loss 3.50840566 - time (sec): 2.93 - samples/sec: 597.21 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:43,510 epoch 1 - iter 40/105 - loss 3.26294657 - time (sec): 3.93 - samples/sec: 616.23 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:44,492 epoch 1 - iter 50/105 - loss 3.05152618 - time (sec): 4.91 - samples/sec: 620.65 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:45,477 epoch 1 - iter 60/105 - loss 2.92978663 - time (sec): 5.90 - samples/sec: 610.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:46,428 epoch 1 - iter 70/105 - loss 2.80236063 - time (sec): 6.85 - samples/sec: 608.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:47,452 epoch 1 - iter 80/105 - loss 2.66861768 - time (sec): 7.87 - samples/sec: 609.51 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:48,403 epoch 1 - iter 90/105 - loss 2.56291144 - time (sec): 8.82 - samples/sec: 611.03 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:49,333 epoch 1 - iter 100/105 - loss 2.48654514 - time (sec): 9.75 - samples/sec: 607.29 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:49,822 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:49,822 EPOCH 1 done: loss 2.4446 - lr: 0.100000 2023-05-15 21:27:51,262 DEV : loss 1.257121205329895 - accuracy (micro avg) 0.6784 2023-05-15 21:27:51,274 - 0 epochs without improvement 2023-05-15 21:27:51,274 saving best model 2023-05-15 21:27:52,427 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:52,588 epoch 2 - iter 10/105 - loss 1.50705724 - time (sec): 0.16 - samples/sec: 3730.85 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:52,740 epoch 2 - iter 20/105 - loss 1.42392433 - time (sec): 0.31 - samples/sec: 3666.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:52,895 epoch 2 - iter 30/105 - loss 1.32810853 - time (sec): 0.47 - samples/sec: 3740.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,018 epoch 2 - iter 40/105 - loss 1.26945619 - time (sec): 0.59 - samples/sec: 3883.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,142 epoch 2 - iter 50/105 - loss 1.25974974 - time (sec): 0.71 - samples/sec: 4067.80 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,272 epoch 2 - iter 60/105 - loss 1.24602583 - time (sec): 0.84 - samples/sec: 4120.93 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,394 epoch 2 - iter 70/105 - loss 1.26179294 - time (sec): 0.97 - samples/sec: 4184.40 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,517 epoch 2 - iter 80/105 - loss 1.22697354 - time (sec): 1.09 - samples/sec: 4240.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,646 epoch 2 - iter 90/105 - loss 1.20106764 - time (sec): 1.22 - samples/sec: 4334.29 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,772 epoch 2 - iter 100/105 - loss 1.17538832 - time (sec): 1.34 - samples/sec: 4373.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:53,839 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:53,839 EPOCH 2 done: loss 1.1693 - lr: 0.100000 2023-05-15 21:27:54,478 DEV : loss 0.7305335402488708 - accuracy (micro avg) 0.8135 2023-05-15 21:27:54,490 - 0 epochs without improvement 2023-05-15 21:27:54,490 saving best model 2023-05-15 21:27:56,027 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:56,188 epoch 3 - iter 10/105 - loss 1.03272509 - time (sec): 0.16 - samples/sec: 3494.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:56,358 epoch 3 - iter 20/105 - loss 0.97097376 - time (sec): 0.33 - samples/sec: 3768.79 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:56,509 epoch 3 - iter 30/105 - loss 0.95689355 - time (sec): 0.48 - samples/sec: 3821.29 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:56,667 epoch 3 - iter 40/105 - loss 0.93238795 - time (sec): 0.64 - samples/sec: 3918.58 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:56,810 epoch 3 - iter 50/105 - loss 0.91059667 - time (sec): 0.78 - samples/sec: 3897.35 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:56,947 epoch 3 - iter 60/105 - loss 0.92685783 - time (sec): 0.92 - samples/sec: 3814.35 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:57,084 epoch 3 - iter 70/105 - loss 0.91279862 - time (sec): 1.06 - samples/sec: 3916.54 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:57,212 epoch 3 - iter 80/105 - loss 0.89905473 - time (sec): 1.18 - samples/sec: 4005.24 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:57,333 epoch 3 - iter 90/105 - loss 0.87855579 - time (sec): 1.31 - samples/sec: 4077.26 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:57,458 epoch 3 - iter 100/105 - loss 0.86400929 - time (sec): 1.43 - samples/sec: 4140.61 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:27:57,525 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:57,526 EPOCH 3 done: loss 0.8662 - lr: 0.100000 2023-05-15 21:27:58,166 DEV : loss 0.6227904558181763 - accuracy (micro avg) 0.8302 2023-05-15 21:27:58,178 - 0 epochs without improvement 2023-05-15 21:27:58,178 saving best model 2023-05-15 21:27:59,690 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:27:59,852 epoch 4 - iter 10/105 - loss 0.70690933 - time (sec): 0.16 - samples/sec: 3773.47 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,017 epoch 4 - iter 20/105 - loss 0.67778427 - time (sec): 0.33 - samples/sec: 3685.11 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,162 epoch 4 - iter 30/105 - loss 0.72460377 - time (sec): 0.47 - samples/sec: 3659.52 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,291 epoch 4 - iter 40/105 - loss 0.74667061 - time (sec): 0.60 - samples/sec: 3901.05 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,413 epoch 4 - iter 50/105 - loss 0.72456310 - time (sec): 0.72 - samples/sec: 4004.33 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,544 epoch 4 - iter 60/105 - loss 0.72599638 - time (sec): 0.85 - samples/sec: 4146.27 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,677 epoch 4 - iter 70/105 - loss 0.70078356 - time (sec): 0.99 - samples/sec: 4201.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,799 epoch 4 - iter 80/105 - loss 0.70384380 - time (sec): 1.11 - samples/sec: 4253.46 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:00,927 epoch 4 - iter 90/105 - loss 0.69472975 - time (sec): 1.24 - samples/sec: 4321.83 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:01,053 epoch 4 - iter 100/105 - loss 0.69011472 - time (sec): 1.36 - samples/sec: 4356.35 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:01,117 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:01,117 EPOCH 4 done: loss 0.6919 - lr: 0.100000 2023-05-15 21:28:02,126 DEV : loss 0.4872177243232727 - accuracy (micro avg) 0.8758 2023-05-15 21:28:02,138 - 0 epochs without improvement 2023-05-15 21:28:02,138 saving best model 2023-05-15 21:28:03,629 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:03,791 epoch 5 - iter 10/105 - loss 0.61365618 - time (sec): 0.16 - samples/sec: 3683.44 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:03,957 epoch 5 - iter 20/105 - loss 0.59538471 - time (sec): 0.33 - samples/sec: 3731.82 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,125 epoch 5 - iter 30/105 - loss 0.57676758 - time (sec): 0.50 - samples/sec: 3761.28 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,258 epoch 5 - iter 40/105 - loss 0.57542241 - time (sec): 0.63 - samples/sec: 3923.23 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,384 epoch 5 - iter 50/105 - loss 0.56359304 - time (sec): 0.76 - samples/sec: 4025.06 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,518 epoch 5 - iter 60/105 - loss 0.56716628 - time (sec): 0.89 - samples/sec: 4121.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,641 epoch 5 - iter 70/105 - loss 0.57142354 - time (sec): 1.01 - samples/sec: 4161.49 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,771 epoch 5 - iter 80/105 - loss 0.57661931 - time (sec): 1.14 - samples/sec: 4218.39 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:04,899 epoch 5 - iter 90/105 - loss 0.58855084 - time (sec): 1.27 - samples/sec: 4241.27 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:05,023 epoch 5 - iter 100/105 - loss 0.58498679 - time (sec): 1.39 - samples/sec: 4268.28 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:05,085 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:05,086 EPOCH 5 done: loss 0.5872 - lr: 0.100000 2023-05-15 21:28:05,762 DEV : loss 0.4932672083377838 - accuracy (micro avg) 0.8797 2023-05-15 21:28:05,774 - 0 epochs without improvement 2023-05-15 21:28:05,774 saving best model 2023-05-15 21:28:07,277 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:07,441 epoch 6 - iter 10/105 - loss 0.53339850 - time (sec): 0.16 - samples/sec: 3641.02 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:07,607 epoch 6 - iter 20/105 - loss 0.50043912 - time (sec): 0.33 - samples/sec: 3706.63 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:07,730 epoch 6 - iter 30/105 - loss 0.53429793 - time (sec): 0.45 - samples/sec: 4028.08 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:07,864 epoch 6 - iter 40/105 - loss 0.52955419 - time (sec): 0.59 - samples/sec: 4164.53 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:07,990 epoch 6 - iter 50/105 - loss 0.50279857 - time (sec): 0.71 - samples/sec: 4250.14 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:08,121 epoch 6 - iter 60/105 - loss 0.49586015 - time (sec): 0.84 - samples/sec: 4329.78 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:08,255 epoch 6 - iter 70/105 - loss 0.50257764 - time (sec): 0.98 - samples/sec: 4351.11 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:08,378 epoch 6 - iter 80/105 - loss 0.51000387 - time (sec): 1.10 - samples/sec: 4362.95 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:08,512 epoch 6 - iter 90/105 - loss 0.51874780 - time (sec): 1.23 - samples/sec: 4366.77 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:08,635 epoch 6 - iter 100/105 - loss 0.52402548 - time (sec): 1.36 - samples/sec: 4361.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:08,703 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:08,703 EPOCH 6 done: loss 0.5268 - lr: 0.100000 2023-05-15 21:28:09,375 DEV : loss 0.4175605773925781 - accuracy (micro avg) 0.8935 2023-05-15 21:28:09,387 - 0 epochs without improvement 2023-05-15 21:28:09,387 saving best model 2023-05-15 21:28:10,889 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:11,057 epoch 7 - iter 10/105 - loss 0.42810660 - time (sec): 0.17 - samples/sec: 3461.94 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:11,219 epoch 7 - iter 20/105 - loss 0.50242569 - time (sec): 0.33 - samples/sec: 3551.85 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:11,385 epoch 7 - iter 30/105 - loss 0.47000735 - time (sec): 0.50 - samples/sec: 3508.06 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:11,547 epoch 7 - iter 40/105 - loss 0.46886899 - time (sec): 0.66 - samples/sec: 3537.36 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:11,706 epoch 7 - iter 50/105 - loss 0.46468136 - time (sec): 0.82 - samples/sec: 3480.86 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:11,868 epoch 7 - iter 60/105 - loss 0.44673748 - time (sec): 0.98 - samples/sec: 3517.24 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:12,042 epoch 7 - iter 70/105 - loss 0.46160434 - time (sec): 1.15 - samples/sec: 3541.95 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:12,201 epoch 7 - iter 80/105 - loss 0.46348997 - time (sec): 1.31 - samples/sec: 3550.57 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:12,364 epoch 7 - iter 90/105 - loss 0.46535512 - time (sec): 1.47 - samples/sec: 3597.04 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:12,522 epoch 7 - iter 100/105 - loss 0.46283468 - time (sec): 1.63 - samples/sec: 3614.16 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:12,602 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:12,602 EPOCH 7 done: loss 0.4622 - lr: 0.100000 2023-05-15 21:28:13,404 DEV : loss 0.3824714422225952 - accuracy (micro avg) 0.9053 2023-05-15 21:28:13,416 - 0 epochs without improvement 2023-05-15 21:28:13,416 saving best model 2023-05-15 21:28:14,931 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:15,098 epoch 8 - iter 10/105 - loss 0.52925490 - time (sec): 0.17 - samples/sec: 3717.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:15,255 epoch 8 - iter 20/105 - loss 0.45622410 - time (sec): 0.32 - samples/sec: 3607.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:15,408 epoch 8 - iter 30/105 - loss 0.43188253 - time (sec): 0.48 - samples/sec: 3586.79 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:15,550 epoch 8 - iter 40/105 - loss 0.45116733 - time (sec): 0.62 - samples/sec: 3625.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:15,679 epoch 8 - iter 50/105 - loss 0.47200060 - time (sec): 0.75 - samples/sec: 3805.42 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:15,808 epoch 8 - iter 60/105 - loss 0.47657265 - time (sec): 0.88 - samples/sec: 3985.65 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:15,936 epoch 8 - iter 70/105 - loss 0.46800503 - time (sec): 1.00 - samples/sec: 4075.81 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:16,063 epoch 8 - iter 80/105 - loss 0.45493913 - time (sec): 1.13 - samples/sec: 4110.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:16,196 epoch 8 - iter 90/105 - loss 0.44503478 - time (sec): 1.26 - samples/sec: 4186.41 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:16,330 epoch 8 - iter 100/105 - loss 0.42980643 - time (sec): 1.40 - samples/sec: 4242.37 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:16,395 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:16,395 EPOCH 8 done: loss 0.4262 - lr: 0.100000 2023-05-15 21:28:17,065 DEV : loss 0.4457044303417206 - accuracy (micro avg) 0.8953 2023-05-15 21:28:17,078 - 1 epochs without improvement 2023-05-15 21:28:17,078 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:17,232 epoch 9 - iter 10/105 - loss 0.43019313 - time (sec): 0.15 - samples/sec: 3666.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:17,391 epoch 9 - iter 20/105 - loss 0.42917599 - time (sec): 0.31 - samples/sec: 3602.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:17,560 epoch 9 - iter 30/105 - loss 0.43572778 - time (sec): 0.48 - samples/sec: 3684.15 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:17,724 epoch 9 - iter 40/105 - loss 0.42423583 - time (sec): 0.65 - samples/sec: 3740.41 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:17,885 epoch 9 - iter 50/105 - loss 0.40547926 - time (sec): 0.81 - samples/sec: 3749.20 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:18,043 epoch 9 - iter 60/105 - loss 0.40662170 - time (sec): 0.96 - samples/sec: 3707.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:18,201 epoch 9 - iter 70/105 - loss 0.41125568 - time (sec): 1.12 - samples/sec: 3713.64 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:18,355 epoch 9 - iter 80/105 - loss 0.40591552 - time (sec): 1.28 - samples/sec: 3672.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:18,520 epoch 9 - iter 90/105 - loss 0.40433275 - time (sec): 1.44 - samples/sec: 3672.57 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:18,677 epoch 9 - iter 100/105 - loss 0.39568312 - time (sec): 1.60 - samples/sec: 3692.30 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:18,760 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:18,760 EPOCH 9 done: loss 0.3993 - lr: 0.100000 2023-05-15 21:28:19,433 DEV : loss 0.4055745005607605 - accuracy (micro avg) 0.9067 2023-05-15 21:28:19,445 - 0 epochs without improvement 2023-05-15 21:28:19,445 saving best model 2023-05-15 21:28:20,938 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:21,109 epoch 10 - iter 10/105 - loss 0.29233026 - time (sec): 0.17 - samples/sec: 3313.66 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:21,290 epoch 10 - iter 20/105 - loss 0.32887121 - time (sec): 0.35 - samples/sec: 3504.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:21,455 epoch 10 - iter 30/105 - loss 0.31719053 - time (sec): 0.52 - samples/sec: 3589.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:21,617 epoch 10 - iter 40/105 - loss 0.32200724 - time (sec): 0.68 - samples/sec: 3565.44 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:21,773 epoch 10 - iter 50/105 - loss 0.33617648 - time (sec): 0.83 - samples/sec: 3622.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:21,940 epoch 10 - iter 60/105 - loss 0.34266434 - time (sec): 1.00 - samples/sec: 3641.11 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:22,103 epoch 10 - iter 70/105 - loss 0.34127674 - time (sec): 1.16 - samples/sec: 3617.73 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:22,249 epoch 10 - iter 80/105 - loss 0.34237331 - time (sec): 1.31 - samples/sec: 3665.35 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:22,404 epoch 10 - iter 90/105 - loss 0.34708513 - time (sec): 1.46 - samples/sec: 3646.44 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:22,554 epoch 10 - iter 100/105 - loss 0.34402692 - time (sec): 1.62 - samples/sec: 3635.62 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:22,644 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:22,644 EPOCH 10 done: loss 0.3448 - lr: 0.100000 2023-05-15 21:28:23,321 DEV : loss 0.3692632019519806 - accuracy (micro avg) 0.9107 2023-05-15 21:28:23,333 - 0 epochs without improvement 2023-05-15 21:28:23,333 saving best model 2023-05-15 21:28:24,834 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:25,004 epoch 11 - iter 10/105 - loss 0.34743618 - time (sec): 0.17 - samples/sec: 3556.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,163 epoch 11 - iter 20/105 - loss 0.34974343 - time (sec): 0.33 - samples/sec: 3539.54 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,326 epoch 11 - iter 30/105 - loss 0.32850942 - time (sec): 0.49 - samples/sec: 3729.38 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,480 epoch 11 - iter 40/105 - loss 0.31337183 - time (sec): 0.65 - samples/sec: 3812.20 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,608 epoch 11 - iter 50/105 - loss 0.30750245 - time (sec): 0.77 - samples/sec: 3931.40 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,726 epoch 11 - iter 60/105 - loss 0.31700868 - time (sec): 0.89 - samples/sec: 3986.15 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,855 epoch 11 - iter 70/105 - loss 0.32409246 - time (sec): 1.02 - samples/sec: 4094.85 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:25,976 epoch 11 - iter 80/105 - loss 0.32874190 - time (sec): 1.14 - samples/sec: 4126.07 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:26,105 epoch 11 - iter 90/105 - loss 0.34002788 - time (sec): 1.27 - samples/sec: 4201.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:26,233 epoch 11 - iter 100/105 - loss 0.33955505 - time (sec): 1.40 - samples/sec: 4249.07 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:26,299 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:26,299 EPOCH 11 done: loss 0.3348 - lr: 0.100000 2023-05-15 21:28:27,102 DEV : loss 0.37659206986427307 - accuracy (micro avg) 0.9107 2023-05-15 21:28:27,114 - 1 epochs without improvement 2023-05-15 21:28:27,114 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:27,266 epoch 12 - iter 10/105 - loss 0.28334459 - time (sec): 0.15 - samples/sec: 3790.16 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:27,430 epoch 12 - iter 20/105 - loss 0.26780325 - time (sec): 0.32 - samples/sec: 3825.09 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:27,584 epoch 12 - iter 30/105 - loss 0.29387770 - time (sec): 0.47 - samples/sec: 3834.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:27,735 epoch 12 - iter 40/105 - loss 0.30666119 - time (sec): 0.62 - samples/sec: 3791.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:27,889 epoch 12 - iter 50/105 - loss 0.30770345 - time (sec): 0.77 - samples/sec: 3792.88 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:28,058 epoch 12 - iter 60/105 - loss 0.31630196 - time (sec): 0.94 - samples/sec: 3808.64 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:28,212 epoch 12 - iter 70/105 - loss 0.32114589 - time (sec): 1.10 - samples/sec: 3793.90 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:28,373 epoch 12 - iter 80/105 - loss 0.32696461 - time (sec): 1.26 - samples/sec: 3737.04 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:28,541 epoch 12 - iter 90/105 - loss 0.32321903 - time (sec): 1.43 - samples/sec: 3732.76 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:28,704 epoch 12 - iter 100/105 - loss 0.32507218 - time (sec): 1.59 - samples/sec: 3733.42 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:28,789 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:28,789 EPOCH 12 done: loss 0.3271 - lr: 0.100000 2023-05-15 21:28:29,460 DEV : loss 0.3771950602531433 - accuracy (micro avg) 0.9105 2023-05-15 21:28:29,472 - 2 epochs without improvement 2023-05-15 21:28:29,472 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:29,633 epoch 13 - iter 10/105 - loss 0.25558085 - time (sec): 0.16 - samples/sec: 3839.70 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:29,804 epoch 13 - iter 20/105 - loss 0.28810978 - time (sec): 0.33 - samples/sec: 3857.76 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:29,949 epoch 13 - iter 30/105 - loss 0.28257029 - time (sec): 0.48 - samples/sec: 3848.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,114 epoch 13 - iter 40/105 - loss 0.26605552 - time (sec): 0.64 - samples/sec: 3808.50 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,272 epoch 13 - iter 50/105 - loss 0.26701675 - time (sec): 0.80 - samples/sec: 3777.93 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,428 epoch 13 - iter 60/105 - loss 0.27406176 - time (sec): 0.96 - samples/sec: 3764.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,578 epoch 13 - iter 70/105 - loss 0.28434073 - time (sec): 1.11 - samples/sec: 3737.53 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,721 epoch 13 - iter 80/105 - loss 0.28311288 - time (sec): 1.25 - samples/sec: 3787.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,855 epoch 13 - iter 90/105 - loss 0.28752102 - time (sec): 1.38 - samples/sec: 3884.81 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:30,979 epoch 13 - iter 100/105 - loss 0.28659536 - time (sec): 1.51 - samples/sec: 3909.75 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:31,048 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:31,048 EPOCH 13 done: loss 0.2864 - lr: 0.100000 2023-05-15 21:28:31,726 DEV : loss 0.3924044966697693 - accuracy (micro avg) 0.9078 2023-05-15 21:28:31,738 - 3 epochs without improvement 2023-05-15 21:28:31,738 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:31,896 epoch 14 - iter 10/105 - loss 0.26458731 - time (sec): 0.16 - samples/sec: 4180.82 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,059 epoch 14 - iter 20/105 - loss 0.25633094 - time (sec): 0.32 - samples/sec: 3916.83 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,213 epoch 14 - iter 30/105 - loss 0.29500461 - time (sec): 0.47 - samples/sec: 3861.29 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,365 epoch 14 - iter 40/105 - loss 0.29691722 - time (sec): 0.63 - samples/sec: 3870.16 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,521 epoch 14 - iter 50/105 - loss 0.29505478 - time (sec): 0.78 - samples/sec: 3847.72 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,652 epoch 14 - iter 60/105 - loss 0.29782093 - time (sec): 0.91 - samples/sec: 3959.37 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,783 epoch 14 - iter 70/105 - loss 0.30191361 - time (sec): 1.04 - samples/sec: 4012.08 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:32,914 epoch 14 - iter 80/105 - loss 0.29182818 - time (sec): 1.17 - samples/sec: 4042.65 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:33,043 epoch 14 - iter 90/105 - loss 0.29334758 - time (sec): 1.30 - samples/sec: 4090.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:33,173 epoch 14 - iter 100/105 - loss 0.29052290 - time (sec): 1.43 - samples/sec: 4130.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:33,245 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:33,246 EPOCH 14 done: loss 0.2933 - lr: 0.100000 2023-05-15 21:28:34,072 DEV : loss 0.3501710295677185 - accuracy (micro avg) 0.9157 2023-05-15 21:28:34,085 - 0 epochs without improvement 2023-05-15 21:28:34,085 saving best model 2023-05-15 21:28:35,588 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:35,755 epoch 15 - iter 10/105 - loss 0.26030079 - time (sec): 0.17 - samples/sec: 3653.21 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:35,920 epoch 15 - iter 20/105 - loss 0.24296167 - time (sec): 0.33 - samples/sec: 3614.18 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,077 epoch 15 - iter 30/105 - loss 0.22735185 - time (sec): 0.49 - samples/sec: 3687.43 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,240 epoch 15 - iter 40/105 - loss 0.24004719 - time (sec): 0.65 - samples/sec: 3695.51 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,395 epoch 15 - iter 50/105 - loss 0.24872295 - time (sec): 0.81 - samples/sec: 3757.80 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,533 epoch 15 - iter 60/105 - loss 0.24078150 - time (sec): 0.94 - samples/sec: 3875.50 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,660 epoch 15 - iter 70/105 - loss 0.23893907 - time (sec): 1.07 - samples/sec: 3967.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,783 epoch 15 - iter 80/105 - loss 0.23832398 - time (sec): 1.19 - samples/sec: 4028.86 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:36,907 epoch 15 - iter 90/105 - loss 0.24174430 - time (sec): 1.32 - samples/sec: 4089.50 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:37,031 epoch 15 - iter 100/105 - loss 0.24719004 - time (sec): 1.44 - samples/sec: 4123.74 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:37,094 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:37,094 EPOCH 15 done: loss 0.2450 - lr: 0.100000 2023-05-15 21:28:37,764 DEV : loss 0.36203423142433167 - accuracy (micro avg) 0.9168 2023-05-15 21:28:37,776 - 0 epochs without improvement 2023-05-15 21:28:37,777 saving best model 2023-05-15 21:28:39,292 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:39,460 epoch 16 - iter 10/105 - loss 0.28751220 - time (sec): 0.17 - samples/sec: 3624.96 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:39,636 epoch 16 - iter 20/105 - loss 0.26365962 - time (sec): 0.34 - samples/sec: 3574.88 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:39,800 epoch 16 - iter 30/105 - loss 0.26388788 - time (sec): 0.51 - samples/sec: 3657.97 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:39,958 epoch 16 - iter 40/105 - loss 0.25961111 - time (sec): 0.67 - samples/sec: 3606.68 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,117 epoch 16 - iter 50/105 - loss 0.25009359 - time (sec): 0.82 - samples/sec: 3602.32 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,268 epoch 16 - iter 60/105 - loss 0.24935196 - time (sec): 0.98 - samples/sec: 3631.04 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,433 epoch 16 - iter 70/105 - loss 0.25583697 - time (sec): 1.14 - samples/sec: 3660.28 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,592 epoch 16 - iter 80/105 - loss 0.25461234 - time (sec): 1.30 - samples/sec: 3649.21 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,749 epoch 16 - iter 90/105 - loss 0.25699326 - time (sec): 1.46 - samples/sec: 3649.05 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,915 epoch 16 - iter 100/105 - loss 0.25579934 - time (sec): 1.62 - samples/sec: 3648.15 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:40,993 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:40,993 EPOCH 16 done: loss 0.2513 - lr: 0.100000 2023-05-15 21:28:41,665 DEV : loss 0.37685805559158325 - accuracy (micro avg) 0.9161 2023-05-15 21:28:41,677 - 1 epochs without improvement 2023-05-15 21:28:41,678 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:41,839 epoch 17 - iter 10/105 - loss 0.25334569 - time (sec): 0.16 - samples/sec: 3850.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:41,981 epoch 17 - iter 20/105 - loss 0.24333967 - time (sec): 0.30 - samples/sec: 3621.39 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,132 epoch 17 - iter 30/105 - loss 0.23545959 - time (sec): 0.45 - samples/sec: 3627.00 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,285 epoch 17 - iter 40/105 - loss 0.23830042 - time (sec): 0.61 - samples/sec: 3707.61 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,413 epoch 17 - iter 50/105 - loss 0.24765555 - time (sec): 0.74 - samples/sec: 3912.11 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,538 epoch 17 - iter 60/105 - loss 0.24520233 - time (sec): 0.86 - samples/sec: 4023.16 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,668 epoch 17 - iter 70/105 - loss 0.24207840 - time (sec): 0.99 - samples/sec: 4189.92 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,797 epoch 17 - iter 80/105 - loss 0.23634265 - time (sec): 1.12 - samples/sec: 4230.99 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:42,925 epoch 17 - iter 90/105 - loss 0.23920227 - time (sec): 1.25 - samples/sec: 4271.72 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:43,059 epoch 17 - iter 100/105 - loss 0.23783792 - time (sec): 1.38 - samples/sec: 4332.38 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:43,119 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:43,119 EPOCH 17 done: loss 0.2414 - lr: 0.100000 2023-05-15 21:28:43,790 DEV : loss 0.37420299649238586 - accuracy (micro avg) 0.9168 2023-05-15 21:28:43,802 - 2 epochs without improvement 2023-05-15 21:28:43,802 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:43,959 epoch 18 - iter 10/105 - loss 0.24061614 - time (sec): 0.16 - samples/sec: 4009.23 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:44,125 epoch 18 - iter 20/105 - loss 0.21762997 - time (sec): 0.32 - samples/sec: 3788.05 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:44,286 epoch 18 - iter 30/105 - loss 0.21340251 - time (sec): 0.48 - samples/sec: 3721.72 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:44,458 epoch 18 - iter 40/105 - loss 0.20776087 - time (sec): 0.66 - samples/sec: 3772.39 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:44,611 epoch 18 - iter 50/105 - loss 0.21267624 - time (sec): 0.81 - samples/sec: 3704.43 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:44,774 epoch 18 - iter 60/105 - loss 0.22464270 - time (sec): 0.97 - samples/sec: 3733.26 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:44,937 epoch 18 - iter 70/105 - loss 0.23068581 - time (sec): 1.13 - samples/sec: 3712.31 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:45,074 epoch 18 - iter 80/105 - loss 0.23007149 - time (sec): 1.27 - samples/sec: 3804.07 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:45,198 epoch 18 - iter 90/105 - loss 0.23322089 - time (sec): 1.40 - samples/sec: 3876.73 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:45,319 epoch 18 - iter 100/105 - loss 0.23723561 - time (sec): 1.52 - samples/sec: 3933.25 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:45,382 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:45,382 EPOCH 18 done: loss 0.2391 - lr: 0.100000 2023-05-15 21:28:46,199 DEV : loss 0.3612535893917084 - accuracy (micro avg) 0.9186 2023-05-15 21:28:46,211 - 0 epochs without improvement 2023-05-15 21:28:46,211 saving best model 2023-05-15 21:28:47,711 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:47,875 epoch 19 - iter 10/105 - loss 0.22788334 - time (sec): 0.16 - samples/sec: 3354.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,045 epoch 19 - iter 20/105 - loss 0.21366377 - time (sec): 0.33 - samples/sec: 3320.64 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,208 epoch 19 - iter 30/105 - loss 0.20840273 - time (sec): 0.50 - samples/sec: 3421.34 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,372 epoch 19 - iter 40/105 - loss 0.20833268 - time (sec): 0.66 - samples/sec: 3520.90 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,527 epoch 19 - iter 50/105 - loss 0.20487635 - time (sec): 0.82 - samples/sec: 3632.63 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,684 epoch 19 - iter 60/105 - loss 0.21416832 - time (sec): 0.97 - samples/sec: 3691.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,810 epoch 19 - iter 70/105 - loss 0.20718833 - time (sec): 1.10 - samples/sec: 3836.58 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:48,936 epoch 19 - iter 80/105 - loss 0.20743279 - time (sec): 1.22 - samples/sec: 3893.28 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:49,067 epoch 19 - iter 90/105 - loss 0.21436857 - time (sec): 1.36 - samples/sec: 3980.06 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:49,192 epoch 19 - iter 100/105 - loss 0.21507106 - time (sec): 1.48 - samples/sec: 4031.59 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:49,260 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:49,260 EPOCH 19 done: loss 0.2162 - lr: 0.100000 2023-05-15 21:28:49,929 DEV : loss 0.38264331221580505 - accuracy (micro avg) 0.9151 2023-05-15 21:28:49,941 - 1 epochs without improvement 2023-05-15 21:28:49,941 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:50,098 epoch 20 - iter 10/105 - loss 0.18692402 - time (sec): 0.16 - samples/sec: 3641.69 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:50,243 epoch 20 - iter 20/105 - loss 0.18505662 - time (sec): 0.30 - samples/sec: 3892.05 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:50,374 epoch 20 - iter 30/105 - loss 0.18862972 - time (sec): 0.43 - samples/sec: 4193.31 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:50,498 epoch 20 - iter 40/105 - loss 0.19222678 - time (sec): 0.56 - samples/sec: 4259.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:50,621 epoch 20 - iter 50/105 - loss 0.19376221 - time (sec): 0.68 - samples/sec: 4310.40 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:50,755 epoch 20 - iter 60/105 - loss 0.20313456 - time (sec): 0.81 - samples/sec: 4423.94 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:50,892 epoch 20 - iter 70/105 - loss 0.21258358 - time (sec): 0.95 - samples/sec: 4528.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:51,018 epoch 20 - iter 80/105 - loss 0.21183394 - time (sec): 1.08 - samples/sec: 4531.58 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:51,138 epoch 20 - iter 90/105 - loss 0.20992423 - time (sec): 1.20 - samples/sec: 4477.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:51,269 epoch 20 - iter 100/105 - loss 0.21406864 - time (sec): 1.33 - samples/sec: 4483.06 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:51,331 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:51,331 EPOCH 20 done: loss 0.2133 - lr: 0.100000 2023-05-15 21:28:52,001 DEV : loss 0.37900087237358093 - accuracy (micro avg) 0.9165 2023-05-15 21:28:52,014 - 2 epochs without improvement 2023-05-15 21:28:52,014 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:52,173 epoch 21 - iter 10/105 - loss 0.21559606 - time (sec): 0.16 - samples/sec: 3872.09 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:52,337 epoch 21 - iter 20/105 - loss 0.20114522 - time (sec): 0.32 - samples/sec: 4092.61 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:52,468 epoch 21 - iter 30/105 - loss 0.18528392 - time (sec): 0.45 - samples/sec: 4232.20 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:52,593 epoch 21 - iter 40/105 - loss 0.18452745 - time (sec): 0.58 - samples/sec: 4270.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:52,719 epoch 21 - iter 50/105 - loss 0.18896888 - time (sec): 0.71 - samples/sec: 4306.03 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:52,844 epoch 21 - iter 60/105 - loss 0.19499164 - time (sec): 0.83 - samples/sec: 4292.80 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:52,973 epoch 21 - iter 70/105 - loss 0.18755836 - time (sec): 0.96 - samples/sec: 4357.24 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:53,100 epoch 21 - iter 80/105 - loss 0.19307926 - time (sec): 1.09 - samples/sec: 4405.82 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:53,226 epoch 21 - iter 90/105 - loss 0.19528471 - time (sec): 1.21 - samples/sec: 4401.00 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:53,353 epoch 21 - iter 100/105 - loss 0.19506996 - time (sec): 1.34 - samples/sec: 4414.11 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:53,419 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:53,419 EPOCH 21 done: loss 0.1992 - lr: 0.100000 2023-05-15 21:28:54,225 DEV : loss 0.38944578170776367 - accuracy (micro avg) 0.9169 2023-05-15 21:28:54,237 - 3 epochs without improvement 2023-05-15 21:28:54,237 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:54,390 epoch 22 - iter 10/105 - loss 0.20993071 - time (sec): 0.15 - samples/sec: 4066.32 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:54,543 epoch 22 - iter 20/105 - loss 0.18478512 - time (sec): 0.31 - samples/sec: 3890.44 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:54,700 epoch 22 - iter 30/105 - loss 0.19947340 - time (sec): 0.46 - samples/sec: 3819.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:54,859 epoch 22 - iter 40/105 - loss 0.19935648 - time (sec): 0.62 - samples/sec: 3886.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:54,988 epoch 22 - iter 50/105 - loss 0.20181282 - time (sec): 0.75 - samples/sec: 3961.08 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:55,122 epoch 22 - iter 60/105 - loss 0.20980325 - time (sec): 0.88 - samples/sec: 4090.00 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:55,252 epoch 22 - iter 70/105 - loss 0.20766586 - time (sec): 1.01 - samples/sec: 4117.60 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:55,377 epoch 22 - iter 80/105 - loss 0.20463919 - time (sec): 1.14 - samples/sec: 4147.86 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:55,503 epoch 22 - iter 90/105 - loss 0.20755944 - time (sec): 1.27 - samples/sec: 4188.49 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:55,637 epoch 22 - iter 100/105 - loss 0.20476708 - time (sec): 1.40 - samples/sec: 4243.65 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:55,703 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:55,703 EPOCH 22 done: loss 0.2058 - lr: 0.100000 2023-05-15 21:28:56,376 DEV : loss 0.3886520266532898 - accuracy (micro avg) 0.9206 2023-05-15 21:28:56,388 - 0 epochs without improvement 2023-05-15 21:28:56,388 saving best model 2023-05-15 21:28:57,893 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:58,068 epoch 23 - iter 10/105 - loss 0.19153291 - time (sec): 0.17 - samples/sec: 3812.49 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:58,231 epoch 23 - iter 20/105 - loss 0.16076549 - time (sec): 0.34 - samples/sec: 3733.08 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:58,391 epoch 23 - iter 30/105 - loss 0.17099648 - time (sec): 0.50 - samples/sec: 3819.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:58,545 epoch 23 - iter 40/105 - loss 0.17438607 - time (sec): 0.65 - samples/sec: 3778.95 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:58,700 epoch 23 - iter 50/105 - loss 0.17411032 - time (sec): 0.81 - samples/sec: 3782.28 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:58,828 epoch 23 - iter 60/105 - loss 0.17241215 - time (sec): 0.93 - samples/sec: 3890.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:58,960 epoch 23 - iter 70/105 - loss 0.18217880 - time (sec): 1.07 - samples/sec: 3984.89 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:59,088 epoch 23 - iter 80/105 - loss 0.18175217 - time (sec): 1.19 - samples/sec: 4030.69 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:59,212 epoch 23 - iter 90/105 - loss 0.18610610 - time (sec): 1.32 - samples/sec: 4069.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:59,344 epoch 23 - iter 100/105 - loss 0.18463815 - time (sec): 1.45 - samples/sec: 4099.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:28:59,410 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:28:59,410 EPOCH 23 done: loss 0.1822 - lr: 0.100000 2023-05-15 21:29:00,084 DEV : loss 0.36885181069374084 - accuracy (micro avg) 0.9197 2023-05-15 21:29:00,097 - 1 epochs without improvement 2023-05-15 21:29:00,097 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:00,254 epoch 24 - iter 10/105 - loss 0.16730327 - time (sec): 0.16 - samples/sec: 3740.77 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:00,427 epoch 24 - iter 20/105 - loss 0.16385379 - time (sec): 0.33 - samples/sec: 3664.66 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:00,584 epoch 24 - iter 30/105 - loss 0.16301537 - time (sec): 0.49 - samples/sec: 3636.25 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:00,747 epoch 24 - iter 40/105 - loss 0.17770530 - time (sec): 0.65 - samples/sec: 3696.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:00,874 epoch 24 - iter 50/105 - loss 0.16391138 - time (sec): 0.78 - samples/sec: 3792.58 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:01,005 epoch 24 - iter 60/105 - loss 0.17546450 - time (sec): 0.91 - samples/sec: 3931.20 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:01,128 epoch 24 - iter 70/105 - loss 0.17562979 - time (sec): 1.03 - samples/sec: 4042.74 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:01,255 epoch 24 - iter 80/105 - loss 0.17835859 - time (sec): 1.16 - samples/sec: 4076.50 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:01,383 epoch 24 - iter 90/105 - loss 0.17677116 - time (sec): 1.29 - samples/sec: 4133.64 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:01,513 epoch 24 - iter 100/105 - loss 0.17858467 - time (sec): 1.42 - samples/sec: 4189.34 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:01,579 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:01,579 EPOCH 24 done: loss 0.1767 - lr: 0.100000 2023-05-15 21:29:02,389 DEV : loss 0.37519118189811707 - accuracy (micro avg) 0.9212 2023-05-15 21:29:02,401 - 0 epochs without improvement 2023-05-15 21:29:02,401 saving best model 2023-05-15 21:29:03,937 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:04,110 epoch 25 - iter 10/105 - loss 0.11805303 - time (sec): 0.17 - samples/sec: 3739.77 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:04,278 epoch 25 - iter 20/105 - loss 0.14216948 - time (sec): 0.34 - samples/sec: 3772.14 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:04,439 epoch 25 - iter 30/105 - loss 0.15601505 - time (sec): 0.50 - samples/sec: 3784.47 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:04,602 epoch 25 - iter 40/105 - loss 0.17032954 - time (sec): 0.67 - samples/sec: 3779.74 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:04,765 epoch 25 - iter 50/105 - loss 0.17585301 - time (sec): 0.83 - samples/sec: 3753.25 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:04,916 epoch 25 - iter 60/105 - loss 0.18142867 - time (sec): 0.98 - samples/sec: 3716.42 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:05,071 epoch 25 - iter 70/105 - loss 0.18596428 - time (sec): 1.13 - samples/sec: 3687.15 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:05,233 epoch 25 - iter 80/105 - loss 0.19091255 - time (sec): 1.30 - samples/sec: 3662.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:05,391 epoch 25 - iter 90/105 - loss 0.19683841 - time (sec): 1.45 - samples/sec: 3663.45 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:05,556 epoch 25 - iter 100/105 - loss 0.19210034 - time (sec): 1.62 - samples/sec: 3665.75 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:05,630 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:05,630 EPOCH 25 done: loss 0.1916 - lr: 0.100000 2023-05-15 21:29:06,302 DEV : loss 0.4038028419017792 - accuracy (micro avg) 0.919 2023-05-15 21:29:06,314 - 1 epochs without improvement 2023-05-15 21:29:06,314 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:06,463 epoch 26 - iter 10/105 - loss 0.23133541 - time (sec): 0.15 - samples/sec: 3868.62 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:06,622 epoch 26 - iter 20/105 - loss 0.22254807 - time (sec): 0.31 - samples/sec: 3844.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:06,788 epoch 26 - iter 30/105 - loss 0.20389609 - time (sec): 0.47 - samples/sec: 3715.95 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:06,949 epoch 26 - iter 40/105 - loss 0.19783433 - time (sec): 0.63 - samples/sec: 3742.26 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,112 epoch 26 - iter 50/105 - loss 0.19137730 - time (sec): 0.80 - samples/sec: 3719.80 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,266 epoch 26 - iter 60/105 - loss 0.18776910 - time (sec): 0.95 - samples/sec: 3715.59 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,425 epoch 26 - iter 70/105 - loss 0.18005685 - time (sec): 1.11 - samples/sec: 3739.70 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,584 epoch 26 - iter 80/105 - loss 0.17298422 - time (sec): 1.27 - samples/sec: 3768.56 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,708 epoch 26 - iter 90/105 - loss 0.18011843 - time (sec): 1.39 - samples/sec: 3809.45 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,840 epoch 26 - iter 100/105 - loss 0.18028208 - time (sec): 1.53 - samples/sec: 3873.21 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:07,909 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:07,909 EPOCH 26 done: loss 0.1815 - lr: 0.100000 2023-05-15 21:29:08,581 DEV : loss 0.3833042085170746 - accuracy (micro avg) 0.9165 2023-05-15 21:29:08,593 - 2 epochs without improvement 2023-05-15 21:29:08,593 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:08,749 epoch 27 - iter 10/105 - loss 0.14937348 - time (sec): 0.16 - samples/sec: 3789.39 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:08,918 epoch 27 - iter 20/105 - loss 0.16552775 - time (sec): 0.32 - samples/sec: 3704.49 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,091 epoch 27 - iter 30/105 - loss 0.16062877 - time (sec): 0.50 - samples/sec: 3704.97 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,252 epoch 27 - iter 40/105 - loss 0.17210393 - time (sec): 0.66 - samples/sec: 3731.70 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,382 epoch 27 - iter 50/105 - loss 0.17608282 - time (sec): 0.79 - samples/sec: 3950.03 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,499 epoch 27 - iter 60/105 - loss 0.17527874 - time (sec): 0.91 - samples/sec: 3986.92 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,624 epoch 27 - iter 70/105 - loss 0.16902907 - time (sec): 1.03 - samples/sec: 4045.54 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,746 epoch 27 - iter 80/105 - loss 0.17736055 - time (sec): 1.15 - samples/sec: 4100.51 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:09,874 epoch 27 - iter 90/105 - loss 0.17949764 - time (sec): 1.28 - samples/sec: 4159.39 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:10,005 epoch 27 - iter 100/105 - loss 0.17488097 - time (sec): 1.41 - samples/sec: 4200.05 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:10,070 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:10,070 EPOCH 27 done: loss 0.1752 - lr: 0.100000 2023-05-15 21:29:10,742 DEV : loss 0.40154093503952026 - accuracy (micro avg) 0.9191 2023-05-15 21:29:10,755 - 3 epochs without improvement 2023-05-15 21:29:10,755 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:10,916 epoch 28 - iter 10/105 - loss 0.14716557 - time (sec): 0.16 - samples/sec: 4006.33 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:11,074 epoch 28 - iter 20/105 - loss 0.14562331 - time (sec): 0.32 - samples/sec: 3926.32 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:11,231 epoch 28 - iter 30/105 - loss 0.13775937 - time (sec): 0.48 - samples/sec: 3861.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:11,388 epoch 28 - iter 40/105 - loss 0.14650236 - time (sec): 0.63 - samples/sec: 3868.72 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:11,552 epoch 28 - iter 50/105 - loss 0.15787049 - time (sec): 0.80 - samples/sec: 3829.39 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:11,706 epoch 28 - iter 60/105 - loss 0.16049621 - time (sec): 0.95 - samples/sec: 3837.66 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:11,861 epoch 28 - iter 70/105 - loss 0.16097042 - time (sec): 1.11 - samples/sec: 3780.49 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:12,020 epoch 28 - iter 80/105 - loss 0.15843466 - time (sec): 1.26 - samples/sec: 3762.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:12,174 epoch 28 - iter 90/105 - loss 0.16246614 - time (sec): 1.42 - samples/sec: 3759.14 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:12,330 epoch 28 - iter 100/105 - loss 0.16244747 - time (sec): 1.57 - samples/sec: 3744.50 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:12,421 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:12,421 EPOCH 28 done: loss 0.1658 - lr: 0.100000 2023-05-15 21:29:13,226 DEV : loss 0.3700896203517914 - accuracy (micro avg) 0.9233 2023-05-15 21:29:13,238 - 0 epochs without improvement 2023-05-15 21:29:13,238 saving best model 2023-05-15 21:29:14,742 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:14,915 epoch 29 - iter 10/105 - loss 0.19017553 - time (sec): 0.17 - samples/sec: 3425.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,078 epoch 29 - iter 20/105 - loss 0.17749714 - time (sec): 0.34 - samples/sec: 3402.37 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,229 epoch 29 - iter 30/105 - loss 0.16616767 - time (sec): 0.49 - samples/sec: 3573.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,365 epoch 29 - iter 40/105 - loss 0.16236268 - time (sec): 0.62 - samples/sec: 3704.61 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,485 epoch 29 - iter 50/105 - loss 0.16865431 - time (sec): 0.74 - samples/sec: 3801.47 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,618 epoch 29 - iter 60/105 - loss 0.17369814 - time (sec): 0.88 - samples/sec: 3945.68 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,747 epoch 29 - iter 70/105 - loss 0.17680836 - time (sec): 1.00 - samples/sec: 4061.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:15,878 epoch 29 - iter 80/105 - loss 0.17509046 - time (sec): 1.14 - samples/sec: 4114.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:16,007 epoch 29 - iter 90/105 - loss 0.17525190 - time (sec): 1.26 - samples/sec: 4165.68 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:16,140 epoch 29 - iter 100/105 - loss 0.17246774 - time (sec): 1.40 - samples/sec: 4236.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:16,207 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:16,207 EPOCH 29 done: loss 0.1727 - lr: 0.100000 2023-05-15 21:29:16,879 DEV : loss 0.4101061522960663 - accuracy (micro avg) 0.9205 2023-05-15 21:29:16,891 - 1 epochs without improvement 2023-05-15 21:29:16,891 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:17,040 epoch 30 - iter 10/105 - loss 0.12426246 - time (sec): 0.15 - samples/sec: 3641.65 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:17,194 epoch 30 - iter 20/105 - loss 0.17071040 - time (sec): 0.30 - samples/sec: 3819.79 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:17,358 epoch 30 - iter 30/105 - loss 0.16425334 - time (sec): 0.47 - samples/sec: 3694.92 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:17,520 epoch 30 - iter 40/105 - loss 0.16927551 - time (sec): 0.63 - samples/sec: 3729.25 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:17,650 epoch 30 - iter 50/105 - loss 0.16612028 - time (sec): 0.76 - samples/sec: 3912.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:17,782 epoch 30 - iter 60/105 - loss 0.16233095 - time (sec): 0.89 - samples/sec: 3977.99 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:17,907 epoch 30 - iter 70/105 - loss 0.16040288 - time (sec): 1.02 - samples/sec: 4047.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:18,035 epoch 30 - iter 80/105 - loss 0.16245743 - time (sec): 1.14 - samples/sec: 4098.75 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:18,159 epoch 30 - iter 90/105 - loss 0.16439274 - time (sec): 1.27 - samples/sec: 4129.95 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:18,300 epoch 30 - iter 100/105 - loss 0.16738507 - time (sec): 1.41 - samples/sec: 4203.40 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:18,367 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:18,367 EPOCH 30 done: loss 0.1668 - lr: 0.100000 2023-05-15 21:29:19,046 DEV : loss 0.40395256876945496 - accuracy (micro avg) 0.9234 2023-05-15 21:29:19,058 - 0 epochs without improvement 2023-05-15 21:29:19,059 saving best model 2023-05-15 21:29:20,549 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:20,721 epoch 31 - iter 10/105 - loss 0.17948015 - time (sec): 0.17 - samples/sec: 3359.44 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:20,888 epoch 31 - iter 20/105 - loss 0.16896526 - time (sec): 0.34 - samples/sec: 3340.57 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,043 epoch 31 - iter 30/105 - loss 0.15090088 - time (sec): 0.49 - samples/sec: 3470.71 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,195 epoch 31 - iter 40/105 - loss 0.14980059 - time (sec): 0.65 - samples/sec: 3531.56 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,349 epoch 31 - iter 50/105 - loss 0.15409275 - time (sec): 0.80 - samples/sec: 3568.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,510 epoch 31 - iter 60/105 - loss 0.16090804 - time (sec): 0.96 - samples/sec: 3606.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,674 epoch 31 - iter 70/105 - loss 0.17302166 - time (sec): 1.12 - samples/sec: 3636.50 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,827 epoch 31 - iter 80/105 - loss 0.17423631 - time (sec): 1.28 - samples/sec: 3660.33 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:21,986 epoch 31 - iter 90/105 - loss 0.16943651 - time (sec): 1.44 - samples/sec: 3668.25 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:22,152 epoch 31 - iter 100/105 - loss 0.17237235 - time (sec): 1.60 - samples/sec: 3678.88 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:22,241 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:22,241 EPOCH 31 done: loss 0.1689 - lr: 0.100000 2023-05-15 21:29:23,046 DEV : loss 0.4038907289505005 - accuracy (micro avg) 0.9215 2023-05-15 21:29:23,058 - 1 epochs without improvement 2023-05-15 21:29:23,058 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:23,221 epoch 32 - iter 10/105 - loss 0.13153856 - time (sec): 0.16 - samples/sec: 3930.46 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:23,386 epoch 32 - iter 20/105 - loss 0.12559417 - time (sec): 0.33 - samples/sec: 3898.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:23,540 epoch 32 - iter 30/105 - loss 0.13186378 - time (sec): 0.48 - samples/sec: 3882.91 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:23,706 epoch 32 - iter 40/105 - loss 0.14738308 - time (sec): 0.65 - samples/sec: 3803.48 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:23,860 epoch 32 - iter 50/105 - loss 0.14542930 - time (sec): 0.80 - samples/sec: 3822.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:24,013 epoch 32 - iter 60/105 - loss 0.14034148 - time (sec): 0.95 - samples/sec: 3772.52 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:24,162 epoch 32 - iter 70/105 - loss 0.14608369 - time (sec): 1.10 - samples/sec: 3736.89 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:24,322 epoch 32 - iter 80/105 - loss 0.14451093 - time (sec): 1.26 - samples/sec: 3765.69 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:24,452 epoch 32 - iter 90/105 - loss 0.14694589 - time (sec): 1.39 - samples/sec: 3835.94 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:24,578 epoch 32 - iter 100/105 - loss 0.14257599 - time (sec): 1.52 - samples/sec: 3899.13 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:24,645 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:24,645 EPOCH 32 done: loss 0.1465 - lr: 0.100000 2023-05-15 21:29:25,317 DEV : loss 0.4025160074234009 - accuracy (micro avg) 0.9226 2023-05-15 21:29:25,329 - 2 epochs without improvement 2023-05-15 21:29:25,330 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:25,489 epoch 33 - iter 10/105 - loss 0.11692645 - time (sec): 0.16 - samples/sec: 3837.74 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:25,661 epoch 33 - iter 20/105 - loss 0.11261481 - time (sec): 0.33 - samples/sec: 3799.27 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:25,818 epoch 33 - iter 30/105 - loss 0.12293677 - time (sec): 0.49 - samples/sec: 3747.77 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:25,963 epoch 33 - iter 40/105 - loss 0.12200511 - time (sec): 0.63 - samples/sec: 3718.00 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,124 epoch 33 - iter 50/105 - loss 0.13014301 - time (sec): 0.79 - samples/sec: 3746.61 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,275 epoch 33 - iter 60/105 - loss 0.13305045 - time (sec): 0.95 - samples/sec: 3739.00 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,432 epoch 33 - iter 70/105 - loss 0.14157664 - time (sec): 1.10 - samples/sec: 3697.61 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,561 epoch 33 - iter 80/105 - loss 0.14296348 - time (sec): 1.23 - samples/sec: 3733.29 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,693 epoch 33 - iter 90/105 - loss 0.14671581 - time (sec): 1.36 - samples/sec: 3850.84 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,823 epoch 33 - iter 100/105 - loss 0.14655293 - time (sec): 1.49 - samples/sec: 3942.58 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:26,890 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:26,890 EPOCH 33 done: loss 0.1479 - lr: 0.100000 2023-05-15 21:29:27,561 DEV : loss 0.3954732418060303 - accuracy (micro avg) 0.9216 2023-05-15 21:29:27,574 - 3 epochs without improvement 2023-05-15 21:29:27,574 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:27,737 epoch 34 - iter 10/105 - loss 0.18187757 - time (sec): 0.16 - samples/sec: 3964.56 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:27,885 epoch 34 - iter 20/105 - loss 0.15943059 - time (sec): 0.31 - samples/sec: 3973.64 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,012 epoch 34 - iter 30/105 - loss 0.18407409 - time (sec): 0.44 - samples/sec: 4058.33 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,142 epoch 34 - iter 40/105 - loss 0.17524787 - time (sec): 0.57 - samples/sec: 4222.45 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,274 epoch 34 - iter 50/105 - loss 0.17493872 - time (sec): 0.70 - samples/sec: 4284.73 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,402 epoch 34 - iter 60/105 - loss 0.16409113 - time (sec): 0.83 - samples/sec: 4326.04 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,533 epoch 34 - iter 70/105 - loss 0.16252810 - time (sec): 0.96 - samples/sec: 4330.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,657 epoch 34 - iter 80/105 - loss 0.16419135 - time (sec): 1.08 - samples/sec: 4339.85 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,784 epoch 34 - iter 90/105 - loss 0.15812127 - time (sec): 1.21 - samples/sec: 4404.06 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,909 epoch 34 - iter 100/105 - loss 0.15864944 - time (sec): 1.33 - samples/sec: 4432.72 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:28,976 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:28,976 EPOCH 34 done: loss 0.1557 - lr: 0.100000 2023-05-15 21:29:29,783 DEV : loss 0.40327951312065125 - accuracy (micro avg) 0.9237 2023-05-15 21:29:29,795 - 0 epochs without improvement 2023-05-15 21:29:29,795 saving best model 2023-05-15 21:29:31,337 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:31,498 epoch 35 - iter 10/105 - loss 0.11365990 - time (sec): 0.16 - samples/sec: 3319.82 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:31,666 epoch 35 - iter 20/105 - loss 0.15195260 - time (sec): 0.33 - samples/sec: 3488.78 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:31,838 epoch 35 - iter 30/105 - loss 0.15559549 - time (sec): 0.50 - samples/sec: 3580.59 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:31,993 epoch 35 - iter 40/105 - loss 0.14416862 - time (sec): 0.66 - samples/sec: 3531.45 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:32,155 epoch 35 - iter 50/105 - loss 0.15059920 - time (sec): 0.82 - samples/sec: 3596.12 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:32,320 epoch 35 - iter 60/105 - loss 0.15134831 - time (sec): 0.98 - samples/sec: 3625.10 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:32,479 epoch 35 - iter 70/105 - loss 0.16104896 - time (sec): 1.14 - samples/sec: 3645.90 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:32,641 epoch 35 - iter 80/105 - loss 0.16009531 - time (sec): 1.30 - samples/sec: 3666.99 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:32,792 epoch 35 - iter 90/105 - loss 0.15922036 - time (sec): 1.45 - samples/sec: 3666.51 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:32,948 epoch 35 - iter 100/105 - loss 0.16487008 - time (sec): 1.61 - samples/sec: 3675.78 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:33,017 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:33,017 EPOCH 35 done: loss 0.1689 - lr: 0.100000 2023-05-15 21:29:33,692 DEV : loss 0.39580753445625305 - accuracy (micro avg) 0.9242 2023-05-15 21:29:33,704 - 0 epochs without improvement 2023-05-15 21:29:33,704 saving best model 2023-05-15 21:29:35,238 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:35,407 epoch 36 - iter 10/105 - loss 0.12681475 - time (sec): 0.17 - samples/sec: 3601.05 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:35,573 epoch 36 - iter 20/105 - loss 0.13123437 - time (sec): 0.34 - samples/sec: 3565.71 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:35,738 epoch 36 - iter 30/105 - loss 0.12768796 - time (sec): 0.50 - samples/sec: 3642.53 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:35,899 epoch 36 - iter 40/105 - loss 0.12212410 - time (sec): 0.66 - samples/sec: 3679.92 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,048 epoch 36 - iter 50/105 - loss 0.12578326 - time (sec): 0.81 - samples/sec: 3698.25 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,209 epoch 36 - iter 60/105 - loss 0.12588040 - time (sec): 0.97 - samples/sec: 3697.77 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,364 epoch 36 - iter 70/105 - loss 0.13398106 - time (sec): 1.13 - samples/sec: 3700.07 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,525 epoch 36 - iter 80/105 - loss 0.13156100 - time (sec): 1.29 - samples/sec: 3718.51 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,671 epoch 36 - iter 90/105 - loss 0.13473477 - time (sec): 1.43 - samples/sec: 3739.14 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,800 epoch 36 - iter 100/105 - loss 0.13714964 - time (sec): 1.56 - samples/sec: 3799.85 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:36,866 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:36,866 EPOCH 36 done: loss 0.1372 - lr: 0.100000 2023-05-15 21:29:37,544 DEV : loss 0.4013025164604187 - accuracy (micro avg) 0.9229 2023-05-15 21:29:37,556 - 1 epochs without improvement 2023-05-15 21:29:37,556 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:37,717 epoch 37 - iter 10/105 - loss 0.17773883 - time (sec): 0.16 - samples/sec: 3833.71 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:37,876 epoch 37 - iter 20/105 - loss 0.13500703 - time (sec): 0.32 - samples/sec: 3767.94 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,032 epoch 37 - iter 30/105 - loss 0.13968275 - time (sec): 0.48 - samples/sec: 3886.64 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,193 epoch 37 - iter 40/105 - loss 0.13029297 - time (sec): 0.64 - samples/sec: 3799.45 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,344 epoch 37 - iter 50/105 - loss 0.13293094 - time (sec): 0.79 - samples/sec: 3762.98 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,502 epoch 37 - iter 60/105 - loss 0.13691739 - time (sec): 0.95 - samples/sec: 3754.09 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,646 epoch 37 - iter 70/105 - loss 0.13779774 - time (sec): 1.09 - samples/sec: 3831.76 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,779 epoch 37 - iter 80/105 - loss 0.13771613 - time (sec): 1.22 - samples/sec: 3888.86 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:38,908 epoch 37 - iter 90/105 - loss 0.13710010 - time (sec): 1.35 - samples/sec: 3953.89 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:39,036 epoch 37 - iter 100/105 - loss 0.14205918 - time (sec): 1.48 - samples/sec: 3993.58 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:39,103 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:39,103 EPOCH 37 done: loss 0.1438 - lr: 0.100000 2023-05-15 21:29:39,810 DEV : loss 0.415070116519928 - accuracy (micro avg) 0.9219 2023-05-15 21:29:39,822 - 2 epochs without improvement 2023-05-15 21:29:39,822 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:39,994 epoch 38 - iter 10/105 - loss 0.20516581 - time (sec): 0.17 - samples/sec: 3954.67 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:40,149 epoch 38 - iter 20/105 - loss 0.18937984 - time (sec): 0.33 - samples/sec: 3794.82 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:40,309 epoch 38 - iter 30/105 - loss 0.15848597 - time (sec): 0.49 - samples/sec: 3844.54 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:40,456 epoch 38 - iter 40/105 - loss 0.16197244 - time (sec): 0.63 - samples/sec: 3741.04 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:40,622 epoch 38 - iter 50/105 - loss 0.16457943 - time (sec): 0.80 - samples/sec: 3794.73 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:40,781 epoch 38 - iter 60/105 - loss 0.16114683 - time (sec): 0.96 - samples/sec: 3756.04 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:40,938 epoch 38 - iter 70/105 - loss 0.16400070 - time (sec): 1.12 - samples/sec: 3776.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:41,085 epoch 38 - iter 80/105 - loss 0.16291289 - time (sec): 1.26 - samples/sec: 3786.55 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:41,217 epoch 38 - iter 90/105 - loss 0.15458326 - time (sec): 1.40 - samples/sec: 3846.21 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:41,342 epoch 38 - iter 100/105 - loss 0.15470577 - time (sec): 1.52 - samples/sec: 3906.01 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:41,404 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:41,404 EPOCH 38 done: loss 0.1544 - lr: 0.100000 2023-05-15 21:29:42,219 DEV : loss 0.41072356700897217 - accuracy (micro avg) 0.9237 2023-05-15 21:29:42,231 - 3 epochs without improvement 2023-05-15 21:29:42,231 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:42,389 epoch 39 - iter 10/105 - loss 0.12626244 - time (sec): 0.16 - samples/sec: 3574.30 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:42,519 epoch 39 - iter 20/105 - loss 0.10437065 - time (sec): 0.29 - samples/sec: 3972.66 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:42,644 epoch 39 - iter 30/105 - loss 0.11959350 - time (sec): 0.41 - samples/sec: 4001.79 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:42,780 epoch 39 - iter 40/105 - loss 0.11522957 - time (sec): 0.55 - samples/sec: 4191.23 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:42,913 epoch 39 - iter 50/105 - loss 0.11583460 - time (sec): 0.68 - samples/sec: 4244.33 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:43,039 epoch 39 - iter 60/105 - loss 0.12341357 - time (sec): 0.81 - samples/sec: 4253.17 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:43,166 epoch 39 - iter 70/105 - loss 0.12221358 - time (sec): 0.93 - samples/sec: 4329.07 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:43,297 epoch 39 - iter 80/105 - loss 0.12490269 - time (sec): 1.07 - samples/sec: 4363.97 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:43,434 epoch 39 - iter 90/105 - loss 0.12776413 - time (sec): 1.20 - samples/sec: 4432.09 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:43,562 epoch 39 - iter 100/105 - loss 0.12640114 - time (sec): 1.33 - samples/sec: 4439.21 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:43,632 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:43,632 EPOCH 39 done: loss 0.1244 - lr: 0.100000 2023-05-15 21:29:44,318 DEV : loss 0.41140687465667725 - accuracy (micro avg) 0.9249 2023-05-15 21:29:44,330 - 0 epochs without improvement 2023-05-15 21:29:44,330 saving best model 2023-05-15 21:29:45,827 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:45,986 epoch 40 - iter 10/105 - loss 0.14795394 - time (sec): 0.16 - samples/sec: 3761.38 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,147 epoch 40 - iter 20/105 - loss 0.11385624 - time (sec): 0.32 - samples/sec: 3628.90 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,280 epoch 40 - iter 30/105 - loss 0.10842964 - time (sec): 0.45 - samples/sec: 3770.91 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,419 epoch 40 - iter 40/105 - loss 0.11264527 - time (sec): 0.59 - samples/sec: 4036.28 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,543 epoch 40 - iter 50/105 - loss 0.11175295 - time (sec): 0.72 - samples/sec: 4150.69 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,666 epoch 40 - iter 60/105 - loss 0.11581112 - time (sec): 0.84 - samples/sec: 4231.95 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,796 epoch 40 - iter 70/105 - loss 0.11738182 - time (sec): 0.97 - samples/sec: 4282.03 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:46,925 epoch 40 - iter 80/105 - loss 0.12190069 - time (sec): 1.10 - samples/sec: 4315.15 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:47,050 epoch 40 - iter 90/105 - loss 0.12535250 - time (sec): 1.22 - samples/sec: 4350.73 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:47,180 epoch 40 - iter 100/105 - loss 0.12320186 - time (sec): 1.35 - samples/sec: 4363.43 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:47,251 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:47,251 EPOCH 40 done: loss 0.1229 - lr: 0.100000 2023-05-15 21:29:47,931 DEV : loss 0.4322431683540344 - accuracy (micro avg) 0.9233 2023-05-15 21:29:47,945 - 1 epochs without improvement 2023-05-15 21:29:47,945 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:48,108 epoch 41 - iter 10/105 - loss 0.13569982 - time (sec): 0.16 - samples/sec: 3739.71 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:48,255 epoch 41 - iter 20/105 - loss 0.12269332 - time (sec): 0.31 - samples/sec: 3765.33 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:48,414 epoch 41 - iter 30/105 - loss 0.12385531 - time (sec): 0.47 - samples/sec: 3683.30 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:48,576 epoch 41 - iter 40/105 - loss 0.12539752 - time (sec): 0.63 - samples/sec: 3765.40 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:48,742 epoch 41 - iter 50/105 - loss 0.13098185 - time (sec): 0.80 - samples/sec: 3827.90 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:48,907 epoch 41 - iter 60/105 - loss 0.13122920 - time (sec): 0.96 - samples/sec: 3754.83 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:49,070 epoch 41 - iter 70/105 - loss 0.13069313 - time (sec): 1.12 - samples/sec: 3760.26 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:49,202 epoch 41 - iter 80/105 - loss 0.12619702 - time (sec): 1.26 - samples/sec: 3872.70 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:49,332 epoch 41 - iter 90/105 - loss 0.12526959 - time (sec): 1.39 - samples/sec: 3931.93 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:49,457 epoch 41 - iter 100/105 - loss 0.12524086 - time (sec): 1.51 - samples/sec: 3954.19 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:49,519 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:49,519 EPOCH 41 done: loss 0.1237 - lr: 0.100000 2023-05-15 21:29:50,333 DEV : loss 0.42178237438201904 - accuracy (micro avg) 0.9234 2023-05-15 21:29:50,345 - 2 epochs without improvement 2023-05-15 21:29:50,345 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:50,508 epoch 42 - iter 10/105 - loss 0.13757911 - time (sec): 0.16 - samples/sec: 3642.86 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:50,675 epoch 42 - iter 20/105 - loss 0.12492803 - time (sec): 0.33 - samples/sec: 3727.46 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:50,835 epoch 42 - iter 30/105 - loss 0.12348396 - time (sec): 0.49 - samples/sec: 3722.72 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:50,989 epoch 42 - iter 40/105 - loss 0.12797827 - time (sec): 0.64 - samples/sec: 3679.45 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:51,152 epoch 42 - iter 50/105 - loss 0.12548280 - time (sec): 0.81 - samples/sec: 3696.99 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:51,309 epoch 42 - iter 60/105 - loss 0.11989623 - time (sec): 0.96 - samples/sec: 3749.20 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:51,473 epoch 42 - iter 70/105 - loss 0.12276712 - time (sec): 1.13 - samples/sec: 3762.06 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:51,627 epoch 42 - iter 80/105 - loss 0.12305123 - time (sec): 1.28 - samples/sec: 3763.51 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:51,780 epoch 42 - iter 90/105 - loss 0.11996811 - time (sec): 1.43 - samples/sec: 3766.57 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:51,934 epoch 42 - iter 100/105 - loss 0.12027167 - time (sec): 1.59 - samples/sec: 3740.47 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:52,014 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:52,014 EPOCH 42 done: loss 0.1211 - lr: 0.100000 2023-05-15 21:29:52,690 DEV : loss 0.42137962579727173 - accuracy (micro avg) 0.9237 2023-05-15 21:29:52,704 - 3 epochs without improvement 2023-05-15 21:29:52,704 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:52,871 epoch 43 - iter 10/105 - loss 0.10634964 - time (sec): 0.17 - samples/sec: 3955.16 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,030 epoch 43 - iter 20/105 - loss 0.13513222 - time (sec): 0.33 - samples/sec: 3909.74 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,182 epoch 43 - iter 30/105 - loss 0.13494545 - time (sec): 0.48 - samples/sec: 3942.48 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,341 epoch 43 - iter 40/105 - loss 0.12992074 - time (sec): 0.64 - samples/sec: 3855.31 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,492 epoch 43 - iter 50/105 - loss 0.13240206 - time (sec): 0.79 - samples/sec: 3849.44 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,658 epoch 43 - iter 60/105 - loss 0.13833264 - time (sec): 0.95 - samples/sec: 3793.30 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,816 epoch 43 - iter 70/105 - loss 0.13098679 - time (sec): 1.11 - samples/sec: 3794.22 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:53,977 epoch 43 - iter 80/105 - loss 0.12497234 - time (sec): 1.27 - samples/sec: 3781.76 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:54,136 epoch 43 - iter 90/105 - loss 0.12201816 - time (sec): 1.43 - samples/sec: 3748.20 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:54,300 epoch 43 - iter 100/105 - loss 0.12714844 - time (sec): 1.60 - samples/sec: 3725.40 - lr: 0.100000 - momentum: 0.000000 2023-05-15 21:29:54,377 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:54,377 EPOCH 43 done: loss 0.1301 - lr: 0.100000 2023-05-15 21:29:55,050 DEV : loss 0.4403139650821686 - accuracy (micro avg) 0.9215 2023-05-15 21:29:55,061 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.05] 2023-05-15 21:29:55,062 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:55,215 epoch 44 - iter 10/105 - loss 0.11070459 - time (sec): 0.15 - samples/sec: 3645.21 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:55,376 epoch 44 - iter 20/105 - loss 0.12333639 - time (sec): 0.31 - samples/sec: 3957.75 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:55,540 epoch 44 - iter 30/105 - loss 0.12356758 - time (sec): 0.48 - samples/sec: 3879.69 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:55,688 epoch 44 - iter 40/105 - loss 0.12342539 - time (sec): 0.63 - samples/sec: 3764.11 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:55,844 epoch 44 - iter 50/105 - loss 0.12214429 - time (sec): 0.78 - samples/sec: 3710.79 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:56,007 epoch 44 - iter 60/105 - loss 0.12046220 - time (sec): 0.95 - samples/sec: 3746.26 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:56,169 epoch 44 - iter 70/105 - loss 0.12088169 - time (sec): 1.11 - samples/sec: 3733.92 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:56,304 epoch 44 - iter 80/105 - loss 0.11876220 - time (sec): 1.24 - samples/sec: 3789.32 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:56,435 epoch 44 - iter 90/105 - loss 0.12117199 - time (sec): 1.37 - samples/sec: 3869.54 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:56,560 epoch 44 - iter 100/105 - loss 0.11797416 - time (sec): 1.50 - samples/sec: 3946.47 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:56,632 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:56,632 EPOCH 44 done: loss 0.1203 - lr: 0.050000 2023-05-15 21:29:57,437 DEV : loss 0.4078696370124817 - accuracy (micro avg) 0.9234 2023-05-15 21:29:57,450 - 1 epochs without improvement 2023-05-15 21:29:57,450 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:57,598 epoch 45 - iter 10/105 - loss 0.12240043 - time (sec): 0.15 - samples/sec: 3917.08 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:57,760 epoch 45 - iter 20/105 - loss 0.12615430 - time (sec): 0.31 - samples/sec: 3888.29 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:57,924 epoch 45 - iter 30/105 - loss 0.11850485 - time (sec): 0.47 - samples/sec: 3763.81 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,082 epoch 45 - iter 40/105 - loss 0.11245046 - time (sec): 0.63 - samples/sec: 3766.75 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,241 epoch 45 - iter 50/105 - loss 0.11174340 - time (sec): 0.79 - samples/sec: 3729.85 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,403 epoch 45 - iter 60/105 - loss 0.11224976 - time (sec): 0.95 - samples/sec: 3692.49 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,563 epoch 45 - iter 70/105 - loss 0.11477408 - time (sec): 1.11 - samples/sec: 3693.73 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,732 epoch 45 - iter 80/105 - loss 0.11335217 - time (sec): 1.28 - samples/sec: 3710.14 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,870 epoch 45 - iter 90/105 - loss 0.11603745 - time (sec): 1.42 - samples/sec: 3801.20 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:58,997 epoch 45 - iter 100/105 - loss 0.11783449 - time (sec): 1.55 - samples/sec: 3827.74 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:29:59,067 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:29:59,067 EPOCH 45 done: loss 0.1205 - lr: 0.050000 2023-05-15 21:29:59,738 DEV : loss 0.42651140689849854 - accuracy (micro avg) 0.9262 2023-05-15 21:29:59,750 - 0 epochs without improvement 2023-05-15 21:29:59,750 saving best model 2023-05-15 21:30:01,303 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:01,462 epoch 46 - iter 10/105 - loss 0.07757579 - time (sec): 0.16 - samples/sec: 3575.33 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:01,635 epoch 46 - iter 20/105 - loss 0.10217992 - time (sec): 0.33 - samples/sec: 3588.55 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:01,786 epoch 46 - iter 30/105 - loss 0.08863273 - time (sec): 0.48 - samples/sec: 3626.34 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:01,938 epoch 46 - iter 40/105 - loss 0.09988256 - time (sec): 0.64 - samples/sec: 3916.20 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,068 epoch 46 - iter 50/105 - loss 0.10114883 - time (sec): 0.77 - samples/sec: 3999.98 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,194 epoch 46 - iter 60/105 - loss 0.09889308 - time (sec): 0.89 - samples/sec: 4105.93 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,314 epoch 46 - iter 70/105 - loss 0.10302309 - time (sec): 1.01 - samples/sec: 4154.63 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,447 epoch 46 - iter 80/105 - loss 0.10177534 - time (sec): 1.14 - samples/sec: 4239.99 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,574 epoch 46 - iter 90/105 - loss 0.10155713 - time (sec): 1.27 - samples/sec: 4239.72 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,706 epoch 46 - iter 100/105 - loss 0.10475806 - time (sec): 1.40 - samples/sec: 4257.72 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:02,771 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:02,771 EPOCH 46 done: loss 0.1047 - lr: 0.050000 2023-05-15 21:30:03,442 DEV : loss 0.42641904950141907 - accuracy (micro avg) 0.9244 2023-05-15 21:30:03,455 - 1 epochs without improvement 2023-05-15 21:30:03,455 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:03,604 epoch 47 - iter 10/105 - loss 0.12040184 - time (sec): 0.15 - samples/sec: 3787.05 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:03,767 epoch 47 - iter 20/105 - loss 0.10311603 - time (sec): 0.31 - samples/sec: 3748.26 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:03,924 epoch 47 - iter 30/105 - loss 0.09732495 - time (sec): 0.47 - samples/sec: 3790.45 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,059 epoch 47 - iter 40/105 - loss 0.10650195 - time (sec): 0.60 - samples/sec: 3867.01 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,187 epoch 47 - iter 50/105 - loss 0.10210912 - time (sec): 0.73 - samples/sec: 3995.66 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,316 epoch 47 - iter 60/105 - loss 0.10302949 - time (sec): 0.86 - samples/sec: 4087.62 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,443 epoch 47 - iter 70/105 - loss 0.09700798 - time (sec): 0.99 - samples/sec: 4161.11 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,568 epoch 47 - iter 80/105 - loss 0.09692983 - time (sec): 1.11 - samples/sec: 4210.37 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,691 epoch 47 - iter 90/105 - loss 0.09707526 - time (sec): 1.24 - samples/sec: 4230.86 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,826 epoch 47 - iter 100/105 - loss 0.09855310 - time (sec): 1.37 - samples/sec: 4304.02 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:04,895 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:04,895 EPOCH 47 done: loss 0.0997 - lr: 0.050000 2023-05-15 21:30:05,699 DEV : loss 0.4337489902973175 - accuracy (micro avg) 0.9248 2023-05-15 21:30:05,711 - 2 epochs without improvement 2023-05-15 21:30:05,711 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:05,868 epoch 48 - iter 10/105 - loss 0.09515060 - time (sec): 0.16 - samples/sec: 3527.84 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,012 epoch 48 - iter 20/105 - loss 0.09041357 - time (sec): 0.30 - samples/sec: 3751.47 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,172 epoch 48 - iter 30/105 - loss 0.08203553 - time (sec): 0.46 - samples/sec: 3713.72 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,342 epoch 48 - iter 40/105 - loss 0.08376034 - time (sec): 0.63 - samples/sec: 3837.15 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,504 epoch 48 - iter 50/105 - loss 0.08316523 - time (sec): 0.79 - samples/sec: 3744.48 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,639 epoch 48 - iter 60/105 - loss 0.08656137 - time (sec): 0.93 - samples/sec: 3854.43 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,771 epoch 48 - iter 70/105 - loss 0.09194897 - time (sec): 1.06 - samples/sec: 3914.28 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:06,899 epoch 48 - iter 80/105 - loss 0.09788336 - time (sec): 1.19 - samples/sec: 3988.64 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:07,031 epoch 48 - iter 90/105 - loss 0.10283371 - time (sec): 1.32 - samples/sec: 4077.49 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:07,151 epoch 48 - iter 100/105 - loss 0.10118860 - time (sec): 1.44 - samples/sec: 4109.14 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:07,217 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:07,217 EPOCH 48 done: loss 0.1001 - lr: 0.050000 2023-05-15 21:30:07,889 DEV : loss 0.43896815180778503 - accuracy (micro avg) 0.9234 2023-05-15 21:30:07,901 - 3 epochs without improvement 2023-05-15 21:30:07,901 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:08,064 epoch 49 - iter 10/105 - loss 0.10410754 - time (sec): 0.16 - samples/sec: 3487.74 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:08,225 epoch 49 - iter 20/105 - loss 0.08994063 - time (sec): 0.32 - samples/sec: 3541.51 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:08,386 epoch 49 - iter 30/105 - loss 0.09238845 - time (sec): 0.49 - samples/sec: 3686.53 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:08,541 epoch 49 - iter 40/105 - loss 0.09027927 - time (sec): 0.64 - samples/sec: 3669.78 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:08,707 epoch 49 - iter 50/105 - loss 0.09236148 - time (sec): 0.81 - samples/sec: 3593.40 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:08,872 epoch 49 - iter 60/105 - loss 0.09347906 - time (sec): 0.97 - samples/sec: 3589.49 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:09,042 epoch 49 - iter 70/105 - loss 0.09866856 - time (sec): 1.14 - samples/sec: 3585.54 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:09,198 epoch 49 - iter 80/105 - loss 0.10140349 - time (sec): 1.30 - samples/sec: 3582.01 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:09,334 epoch 49 - iter 90/105 - loss 0.10317846 - time (sec): 1.43 - samples/sec: 3683.47 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:09,469 epoch 49 - iter 100/105 - loss 0.10346212 - time (sec): 1.57 - samples/sec: 3800.14 - lr: 0.050000 - momentum: 0.000000 2023-05-15 21:30:09,529 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:09,529 EPOCH 49 done: loss 0.1022 - lr: 0.050000 2023-05-15 21:30:10,202 DEV : loss 0.4417934715747833 - accuracy (micro avg) 0.9247 2023-05-15 21:30:10,214 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.025] 2023-05-15 21:30:10,214 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:10,374 epoch 50 - iter 10/105 - loss 0.10155014 - time (sec): 0.16 - samples/sec: 3410.51 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:10,528 epoch 50 - iter 20/105 - loss 0.09421830 - time (sec): 0.31 - samples/sec: 3582.91 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:10,683 epoch 50 - iter 30/105 - loss 0.08456918 - time (sec): 0.47 - samples/sec: 3597.25 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:10,838 epoch 50 - iter 40/105 - loss 0.08748446 - time (sec): 0.62 - samples/sec: 3703.35 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,003 epoch 50 - iter 50/105 - loss 0.08878558 - time (sec): 0.79 - samples/sec: 3695.58 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,158 epoch 50 - iter 60/105 - loss 0.08631747 - time (sec): 0.94 - samples/sec: 3650.32 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,322 epoch 50 - iter 70/105 - loss 0.08712266 - time (sec): 1.11 - samples/sec: 3653.78 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,489 epoch 50 - iter 80/105 - loss 0.08719100 - time (sec): 1.27 - samples/sec: 3674.25 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,646 epoch 50 - iter 90/105 - loss 0.08340004 - time (sec): 1.43 - samples/sec: 3690.59 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,800 epoch 50 - iter 100/105 - loss 0.08787648 - time (sec): 1.59 - samples/sec: 3705.52 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:11,888 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:11,888 EPOCH 50 done: loss 0.0871 - lr: 0.025000 2023-05-15 21:30:12,560 DEV : loss 0.43998345732688904 - accuracy (micro avg) 0.9248 2023-05-15 21:30:12,572 - 1 epochs without improvement 2023-05-15 21:30:12,572 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:12,714 epoch 51 - iter 10/105 - loss 0.08250246 - time (sec): 0.14 - samples/sec: 3380.14 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:12,885 epoch 51 - iter 20/105 - loss 0.07379753 - time (sec): 0.31 - samples/sec: 3591.82 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,052 epoch 51 - iter 30/105 - loss 0.08802722 - time (sec): 0.48 - samples/sec: 3742.11 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,207 epoch 51 - iter 40/105 - loss 0.09066879 - time (sec): 0.63 - samples/sec: 3831.15 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,357 epoch 51 - iter 50/105 - loss 0.08766548 - time (sec): 0.78 - samples/sec: 3935.49 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,486 epoch 51 - iter 60/105 - loss 0.08422949 - time (sec): 0.91 - samples/sec: 4058.85 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,616 epoch 51 - iter 70/105 - loss 0.08933460 - time (sec): 1.04 - samples/sec: 4091.37 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,740 epoch 51 - iter 80/105 - loss 0.09256634 - time (sec): 1.17 - samples/sec: 4130.07 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,871 epoch 51 - iter 90/105 - loss 0.09040581 - time (sec): 1.30 - samples/sec: 4153.59 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:13,996 epoch 51 - iter 100/105 - loss 0.08850502 - time (sec): 1.42 - samples/sec: 4178.50 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:14,060 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:14,060 EPOCH 51 done: loss 0.0889 - lr: 0.025000 2023-05-15 21:30:14,868 DEV : loss 0.4459802806377411 - accuracy (micro avg) 0.9266 2023-05-15 21:30:14,880 - 0 epochs without improvement 2023-05-15 21:30:14,880 saving best model 2023-05-15 21:30:16,376 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:16,538 epoch 52 - iter 10/105 - loss 0.08790156 - time (sec): 0.16 - samples/sec: 3551.43 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:16,709 epoch 52 - iter 20/105 - loss 0.09278011 - time (sec): 0.33 - samples/sec: 3493.27 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:16,874 epoch 52 - iter 30/105 - loss 0.09091522 - time (sec): 0.50 - samples/sec: 3689.32 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,041 epoch 52 - iter 40/105 - loss 0.08797693 - time (sec): 0.66 - samples/sec: 3717.58 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,199 epoch 52 - iter 50/105 - loss 0.09140983 - time (sec): 0.82 - samples/sec: 3722.50 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,362 epoch 52 - iter 60/105 - loss 0.08783638 - time (sec): 0.99 - samples/sec: 3691.63 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,521 epoch 52 - iter 70/105 - loss 0.09044234 - time (sec): 1.14 - samples/sec: 3683.92 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,683 epoch 52 - iter 80/105 - loss 0.09242811 - time (sec): 1.31 - samples/sec: 3702.29 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,837 epoch 52 - iter 90/105 - loss 0.09060020 - time (sec): 1.46 - samples/sec: 3683.11 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:17,993 epoch 52 - iter 100/105 - loss 0.09166655 - time (sec): 1.62 - samples/sec: 3671.75 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:18,075 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:18,075 EPOCH 52 done: loss 0.0923 - lr: 0.025000 2023-05-15 21:30:18,746 DEV : loss 0.44730234146118164 - accuracy (micro avg) 0.9249 2023-05-15 21:30:18,758 - 1 epochs without improvement 2023-05-15 21:30:18,758 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:18,905 epoch 53 - iter 10/105 - loss 0.05499141 - time (sec): 0.15 - samples/sec: 3610.68 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:19,062 epoch 53 - iter 20/105 - loss 0.07059527 - time (sec): 0.30 - samples/sec: 3575.99 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:19,223 epoch 53 - iter 30/105 - loss 0.08327849 - time (sec): 0.46 - samples/sec: 3646.81 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:19,371 epoch 53 - iter 40/105 - loss 0.08524775 - time (sec): 0.61 - samples/sec: 3626.90 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:19,531 epoch 53 - iter 50/105 - loss 0.08712003 - time (sec): 0.77 - samples/sec: 3684.46 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:19,703 epoch 53 - iter 60/105 - loss 0.09188637 - time (sec): 0.94 - samples/sec: 3692.12 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:19,858 epoch 53 - iter 70/105 - loss 0.09597465 - time (sec): 1.10 - samples/sec: 3664.92 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:20,023 epoch 53 - iter 80/105 - loss 0.09462198 - time (sec): 1.26 - samples/sec: 3699.23 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:20,155 epoch 53 - iter 90/105 - loss 0.09487981 - time (sec): 1.40 - samples/sec: 3804.17 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:20,285 epoch 53 - iter 100/105 - loss 0.09549078 - time (sec): 1.53 - samples/sec: 3868.82 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:20,356 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:20,356 EPOCH 53 done: loss 0.0979 - lr: 0.025000 2023-05-15 21:30:21,036 DEV : loss 0.4442404806613922 - accuracy (micro avg) 0.9267 2023-05-15 21:30:21,048 - 0 epochs without improvement 2023-05-15 21:30:21,048 saving best model 2023-05-15 21:30:22,517 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:22,691 epoch 54 - iter 10/105 - loss 0.08495571 - time (sec): 0.17 - samples/sec: 3491.90 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:22,854 epoch 54 - iter 20/105 - loss 0.08501973 - time (sec): 0.34 - samples/sec: 3554.35 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,021 epoch 54 - iter 30/105 - loss 0.09743852 - time (sec): 0.50 - samples/sec: 3672.88 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,172 epoch 54 - iter 40/105 - loss 0.09451822 - time (sec): 0.65 - samples/sec: 3701.29 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,309 epoch 54 - iter 50/105 - loss 0.08993559 - time (sec): 0.79 - samples/sec: 3787.10 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,443 epoch 54 - iter 60/105 - loss 0.08867751 - time (sec): 0.93 - samples/sec: 3908.89 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,567 epoch 54 - iter 70/105 - loss 0.08874145 - time (sec): 1.05 - samples/sec: 3967.35 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,694 epoch 54 - iter 80/105 - loss 0.08686806 - time (sec): 1.18 - samples/sec: 4038.98 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,825 epoch 54 - iter 90/105 - loss 0.09092987 - time (sec): 1.31 - samples/sec: 4116.77 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:23,946 epoch 54 - iter 100/105 - loss 0.08850743 - time (sec): 1.43 - samples/sec: 4145.41 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:24,013 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:24,013 EPOCH 54 done: loss 0.0894 - lr: 0.025000 2023-05-15 21:30:24,822 DEV : loss 0.44815778732299805 - accuracy (micro avg) 0.9267 2023-05-15 21:30:24,834 - 1 epochs without improvement 2023-05-15 21:30:24,835 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:24,998 epoch 55 - iter 10/105 - loss 0.11340921 - time (sec): 0.16 - samples/sec: 3674.32 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:25,161 epoch 55 - iter 20/105 - loss 0.10073815 - time (sec): 0.33 - samples/sec: 3971.60 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:25,306 epoch 55 - iter 30/105 - loss 0.10311474 - time (sec): 0.47 - samples/sec: 3810.26 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:25,472 epoch 55 - iter 40/105 - loss 0.09085139 - time (sec): 0.64 - samples/sec: 3794.66 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:25,628 epoch 55 - iter 50/105 - loss 0.10506698 - time (sec): 0.79 - samples/sec: 3785.43 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:25,784 epoch 55 - iter 60/105 - loss 0.10309018 - time (sec): 0.95 - samples/sec: 3817.32 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:25,935 epoch 55 - iter 70/105 - loss 0.10555289 - time (sec): 1.10 - samples/sec: 3784.28 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:26,092 epoch 55 - iter 80/105 - loss 0.10515112 - time (sec): 1.26 - samples/sec: 3806.57 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:26,237 epoch 55 - iter 90/105 - loss 0.10222695 - time (sec): 1.40 - samples/sec: 3807.05 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:26,390 epoch 55 - iter 100/105 - loss 0.09704806 - time (sec): 1.56 - samples/sec: 3812.01 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:26,471 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:26,472 EPOCH 55 done: loss 0.0994 - lr: 0.025000 2023-05-15 21:30:27,160 DEV : loss 0.4603947103023529 - accuracy (micro avg) 0.9273 2023-05-15 21:30:27,172 - 0 epochs without improvement 2023-05-15 21:30:27,172 saving best model 2023-05-15 21:30:28,679 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:28,850 epoch 56 - iter 10/105 - loss 0.08967255 - time (sec): 0.17 - samples/sec: 3319.85 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,019 epoch 56 - iter 20/105 - loss 0.06883085 - time (sec): 0.34 - samples/sec: 3569.10 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,172 epoch 56 - iter 30/105 - loss 0.07577760 - time (sec): 0.49 - samples/sec: 3659.54 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,318 epoch 56 - iter 40/105 - loss 0.08986062 - time (sec): 0.64 - samples/sec: 3724.69 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,476 epoch 56 - iter 50/105 - loss 0.08512969 - time (sec): 0.80 - samples/sec: 3717.46 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,634 epoch 56 - iter 60/105 - loss 0.09044687 - time (sec): 0.95 - samples/sec: 3722.60 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,786 epoch 56 - iter 70/105 - loss 0.09543241 - time (sec): 1.11 - samples/sec: 3771.58 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:29,950 epoch 56 - iter 80/105 - loss 0.09562096 - time (sec): 1.27 - samples/sec: 3734.26 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:30,107 epoch 56 - iter 90/105 - loss 0.09317804 - time (sec): 1.43 - samples/sec: 3722.42 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:30,255 epoch 56 - iter 100/105 - loss 0.09849251 - time (sec): 1.58 - samples/sec: 3748.79 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:30,337 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:30,337 EPOCH 56 done: loss 0.0992 - lr: 0.025000 2023-05-15 21:30:31,005 DEV : loss 0.45674797892570496 - accuracy (micro avg) 0.9253 2023-05-15 21:30:31,018 - 1 epochs without improvement 2023-05-15 21:30:31,018 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:31,172 epoch 57 - iter 10/105 - loss 0.02979771 - time (sec): 0.15 - samples/sec: 3689.67 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:31,326 epoch 57 - iter 20/105 - loss 0.06583308 - time (sec): 0.31 - samples/sec: 3619.17 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:31,490 epoch 57 - iter 30/105 - loss 0.07248027 - time (sec): 0.47 - samples/sec: 3718.69 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:31,635 epoch 57 - iter 40/105 - loss 0.06894876 - time (sec): 0.62 - samples/sec: 3819.61 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:31,759 epoch 57 - iter 50/105 - loss 0.06672600 - time (sec): 0.74 - samples/sec: 3965.13 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:31,887 epoch 57 - iter 60/105 - loss 0.07004378 - time (sec): 0.87 - samples/sec: 4066.85 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:32,012 epoch 57 - iter 70/105 - loss 0.07138190 - time (sec): 0.99 - samples/sec: 4149.22 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:32,140 epoch 57 - iter 80/105 - loss 0.07725337 - time (sec): 1.12 - samples/sec: 4193.19 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:32,274 epoch 57 - iter 90/105 - loss 0.07986006 - time (sec): 1.26 - samples/sec: 4260.96 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:32,401 epoch 57 - iter 100/105 - loss 0.08470070 - time (sec): 1.38 - samples/sec: 4325.36 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:32,469 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:32,469 EPOCH 57 done: loss 0.0849 - lr: 0.025000 2023-05-15 21:30:33,267 DEV : loss 0.4544132947921753 - accuracy (micro avg) 0.9276 2023-05-15 21:30:33,279 - 0 epochs without improvement 2023-05-15 21:30:33,280 saving best model 2023-05-15 21:30:34,740 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:34,910 epoch 58 - iter 10/105 - loss 0.09034536 - time (sec): 0.17 - samples/sec: 3468.88 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,063 epoch 58 - iter 20/105 - loss 0.11065231 - time (sec): 0.32 - samples/sec: 3467.45 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,224 epoch 58 - iter 30/105 - loss 0.10656638 - time (sec): 0.48 - samples/sec: 3512.14 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,385 epoch 58 - iter 40/105 - loss 0.09482706 - time (sec): 0.64 - samples/sec: 3652.50 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,550 epoch 58 - iter 50/105 - loss 0.08608497 - time (sec): 0.81 - samples/sec: 3591.58 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,701 epoch 58 - iter 60/105 - loss 0.08756389 - time (sec): 0.96 - samples/sec: 3609.97 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,829 epoch 58 - iter 70/105 - loss 0.09090279 - time (sec): 1.09 - samples/sec: 3715.99 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:35,955 epoch 58 - iter 80/105 - loss 0.08859676 - time (sec): 1.21 - samples/sec: 3827.45 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:36,086 epoch 58 - iter 90/105 - loss 0.08564067 - time (sec): 1.35 - samples/sec: 3907.50 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:36,220 epoch 58 - iter 100/105 - loss 0.08297622 - time (sec): 1.48 - samples/sec: 3985.55 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:36,290 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:36,291 EPOCH 58 done: loss 0.0848 - lr: 0.025000 2023-05-15 21:30:36,963 DEV : loss 0.4505839943885803 - accuracy (micro avg) 0.9269 2023-05-15 21:30:36,976 - 1 epochs without improvement 2023-05-15 21:30:36,976 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:37,141 epoch 59 - iter 10/105 - loss 0.09168406 - time (sec): 0.16 - samples/sec: 3648.56 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:37,304 epoch 59 - iter 20/105 - loss 0.09114142 - time (sec): 0.33 - samples/sec: 3729.72 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:37,467 epoch 59 - iter 30/105 - loss 0.08044970 - time (sec): 0.49 - samples/sec: 3824.77 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:37,612 epoch 59 - iter 40/105 - loss 0.08763567 - time (sec): 0.64 - samples/sec: 3884.63 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:37,740 epoch 59 - iter 50/105 - loss 0.08274090 - time (sec): 0.76 - samples/sec: 3976.87 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:37,869 epoch 59 - iter 60/105 - loss 0.08555697 - time (sec): 0.89 - samples/sec: 4118.47 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:37,989 epoch 59 - iter 70/105 - loss 0.08258492 - time (sec): 1.01 - samples/sec: 4109.92 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:38,112 epoch 59 - iter 80/105 - loss 0.08036995 - time (sec): 1.14 - samples/sec: 4127.41 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:38,242 epoch 59 - iter 90/105 - loss 0.08190745 - time (sec): 1.27 - samples/sec: 4182.97 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:38,371 epoch 59 - iter 100/105 - loss 0.08489307 - time (sec): 1.40 - samples/sec: 4245.97 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:38,437 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:38,437 EPOCH 59 done: loss 0.0831 - lr: 0.025000 2023-05-15 21:30:39,126 DEV : loss 0.449642151594162 - accuracy (micro avg) 0.9287 2023-05-15 21:30:39,139 - 0 epochs without improvement 2023-05-15 21:30:39,139 saving best model 2023-05-15 21:30:40,634 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:40,810 epoch 60 - iter 10/105 - loss 0.07294341 - time (sec): 0.18 - samples/sec: 3747.11 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:40,975 epoch 60 - iter 20/105 - loss 0.07297189 - time (sec): 0.34 - samples/sec: 3538.07 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,132 epoch 60 - iter 30/105 - loss 0.08514902 - time (sec): 0.50 - samples/sec: 3549.28 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,259 epoch 60 - iter 40/105 - loss 0.09686086 - time (sec): 0.62 - samples/sec: 3697.61 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,391 epoch 60 - iter 50/105 - loss 0.09001737 - time (sec): 0.76 - samples/sec: 3913.06 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,509 epoch 60 - iter 60/105 - loss 0.08812331 - time (sec): 0.87 - samples/sec: 3958.10 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,640 epoch 60 - iter 70/105 - loss 0.08630214 - time (sec): 1.01 - samples/sec: 4044.86 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,771 epoch 60 - iter 80/105 - loss 0.08846802 - time (sec): 1.14 - samples/sec: 4130.01 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:41,904 epoch 60 - iter 90/105 - loss 0.08861147 - time (sec): 1.27 - samples/sec: 4195.05 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:42,036 epoch 60 - iter 100/105 - loss 0.08885317 - time (sec): 1.40 - samples/sec: 4207.76 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:42,108 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:42,108 EPOCH 60 done: loss 0.0874 - lr: 0.025000 2023-05-15 21:30:42,795 DEV : loss 0.4487907290458679 - accuracy (micro avg) 0.9277 2023-05-15 21:30:42,808 - 1 epochs without improvement 2023-05-15 21:30:42,808 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:42,973 epoch 61 - iter 10/105 - loss 0.08015160 - time (sec): 0.16 - samples/sec: 3824.57 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,123 epoch 61 - iter 20/105 - loss 0.06057929 - time (sec): 0.31 - samples/sec: 3784.72 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,275 epoch 61 - iter 30/105 - loss 0.06926931 - time (sec): 0.47 - samples/sec: 3624.66 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,427 epoch 61 - iter 40/105 - loss 0.07085607 - time (sec): 0.62 - samples/sec: 3670.65 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,588 epoch 61 - iter 50/105 - loss 0.07059675 - time (sec): 0.78 - samples/sec: 3724.09 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,723 epoch 61 - iter 60/105 - loss 0.07058916 - time (sec): 0.91 - samples/sec: 3895.12 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,845 epoch 61 - iter 70/105 - loss 0.07125948 - time (sec): 1.04 - samples/sec: 3951.28 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:43,978 epoch 61 - iter 80/105 - loss 0.07479688 - time (sec): 1.17 - samples/sec: 4037.05 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:44,111 epoch 61 - iter 90/105 - loss 0.07313734 - time (sec): 1.30 - samples/sec: 4092.27 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:44,241 epoch 61 - iter 100/105 - loss 0.07799588 - time (sec): 1.43 - samples/sec: 4126.79 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:44,312 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:44,312 EPOCH 61 done: loss 0.0760 - lr: 0.025000 2023-05-15 21:30:45,140 DEV : loss 0.4585801362991333 - accuracy (micro avg) 0.927 2023-05-15 21:30:45,153 - 2 epochs without improvement 2023-05-15 21:30:45,153 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:45,308 epoch 62 - iter 10/105 - loss 0.09004574 - time (sec): 0.15 - samples/sec: 3944.17 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:45,466 epoch 62 - iter 20/105 - loss 0.07939139 - time (sec): 0.31 - samples/sec: 3844.98 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:45,595 epoch 62 - iter 30/105 - loss 0.06903933 - time (sec): 0.44 - samples/sec: 4024.22 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:45,726 epoch 62 - iter 40/105 - loss 0.07174227 - time (sec): 0.57 - samples/sec: 4081.11 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:45,861 epoch 62 - iter 50/105 - loss 0.07977434 - time (sec): 0.71 - samples/sec: 4136.79 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:45,992 epoch 62 - iter 60/105 - loss 0.08341777 - time (sec): 0.84 - samples/sec: 4264.15 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:46,125 epoch 62 - iter 70/105 - loss 0.08811842 - time (sec): 0.97 - samples/sec: 4249.91 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:46,252 epoch 62 - iter 80/105 - loss 0.08805974 - time (sec): 1.10 - samples/sec: 4258.69 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:46,386 epoch 62 - iter 90/105 - loss 0.08797355 - time (sec): 1.23 - samples/sec: 4307.63 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:46,521 epoch 62 - iter 100/105 - loss 0.08905213 - time (sec): 1.37 - samples/sec: 4343.29 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:46,587 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:46,587 EPOCH 62 done: loss 0.0876 - lr: 0.025000 2023-05-15 21:30:47,277 DEV : loss 0.45762529969215393 - accuracy (micro avg) 0.9273 2023-05-15 21:30:47,289 - 3 epochs without improvement 2023-05-15 21:30:47,290 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:47,455 epoch 63 - iter 10/105 - loss 0.13866874 - time (sec): 0.17 - samples/sec: 3764.67 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:47,616 epoch 63 - iter 20/105 - loss 0.11774307 - time (sec): 0.33 - samples/sec: 3679.75 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:47,773 epoch 63 - iter 30/105 - loss 0.10617288 - time (sec): 0.48 - samples/sec: 3634.33 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:47,903 epoch 63 - iter 40/105 - loss 0.10151569 - time (sec): 0.61 - samples/sec: 3826.73 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,036 epoch 63 - iter 50/105 - loss 0.10331057 - time (sec): 0.75 - samples/sec: 3977.39 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,162 epoch 63 - iter 60/105 - loss 0.09765511 - time (sec): 0.87 - samples/sec: 4060.06 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,297 epoch 63 - iter 70/105 - loss 0.09812690 - time (sec): 1.01 - samples/sec: 4107.99 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,422 epoch 63 - iter 80/105 - loss 0.10264448 - time (sec): 1.13 - samples/sec: 4126.25 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,555 epoch 63 - iter 90/105 - loss 0.09810914 - time (sec): 1.27 - samples/sec: 4178.06 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,688 epoch 63 - iter 100/105 - loss 0.09693463 - time (sec): 1.40 - samples/sec: 4235.89 - lr: 0.025000 - momentum: 0.000000 2023-05-15 21:30:48,757 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:48,757 EPOCH 63 done: loss 0.0968 - lr: 0.025000 2023-05-15 21:30:49,446 DEV : loss 0.4547846019268036 - accuracy (micro avg) 0.9287 2023-05-15 21:30:49,459 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.0125] 2023-05-15 21:30:49,459 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:49,621 epoch 64 - iter 10/105 - loss 0.10850674 - time (sec): 0.16 - samples/sec: 3610.44 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:49,789 epoch 64 - iter 20/105 - loss 0.09252579 - time (sec): 0.33 - samples/sec: 3587.76 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:49,937 epoch 64 - iter 30/105 - loss 0.08403065 - time (sec): 0.48 - samples/sec: 3635.53 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,095 epoch 64 - iter 40/105 - loss 0.08705889 - time (sec): 0.64 - samples/sec: 3693.26 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,249 epoch 64 - iter 50/105 - loss 0.08250025 - time (sec): 0.79 - samples/sec: 3689.38 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,380 epoch 64 - iter 60/105 - loss 0.07915800 - time (sec): 0.92 - samples/sec: 3822.31 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,505 epoch 64 - iter 70/105 - loss 0.08280372 - time (sec): 1.05 - samples/sec: 3914.92 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,632 epoch 64 - iter 80/105 - loss 0.08043726 - time (sec): 1.17 - samples/sec: 4009.72 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,760 epoch 64 - iter 90/105 - loss 0.07739270 - time (sec): 1.30 - samples/sec: 4072.08 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,892 epoch 64 - iter 100/105 - loss 0.07765963 - time (sec): 1.43 - samples/sec: 4135.64 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:50,958 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:50,958 EPOCH 64 done: loss 0.0806 - lr: 0.012500 2023-05-15 21:30:51,770 DEV : loss 0.4491409659385681 - accuracy (micro avg) 0.927 2023-05-15 21:30:51,782 - 1 epochs without improvement 2023-05-15 21:30:51,782 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:51,936 epoch 65 - iter 10/105 - loss 0.09353496 - time (sec): 0.15 - samples/sec: 3789.70 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,097 epoch 65 - iter 20/105 - loss 0.08407092 - time (sec): 0.31 - samples/sec: 3859.28 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,230 epoch 65 - iter 30/105 - loss 0.07108789 - time (sec): 0.45 - samples/sec: 3987.09 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,358 epoch 65 - iter 40/105 - loss 0.07327740 - time (sec): 0.58 - samples/sec: 4156.74 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,485 epoch 65 - iter 50/105 - loss 0.08257736 - time (sec): 0.70 - samples/sec: 4220.56 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,609 epoch 65 - iter 60/105 - loss 0.08262387 - time (sec): 0.83 - samples/sec: 4243.88 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,737 epoch 65 - iter 70/105 - loss 0.07754773 - time (sec): 0.95 - samples/sec: 4326.77 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:52,864 epoch 65 - iter 80/105 - loss 0.07909466 - time (sec): 1.08 - samples/sec: 4357.30 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:53,001 epoch 65 - iter 90/105 - loss 0.07937441 - time (sec): 1.22 - samples/sec: 4386.93 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:53,128 epoch 65 - iter 100/105 - loss 0.07883681 - time (sec): 1.35 - samples/sec: 4390.93 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:53,196 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:53,196 EPOCH 65 done: loss 0.0776 - lr: 0.012500 2023-05-15 21:30:53,867 DEV : loss 0.45759317278862 - accuracy (micro avg) 0.9269 2023-05-15 21:30:53,879 - 2 epochs without improvement 2023-05-15 21:30:53,880 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:54,039 epoch 66 - iter 10/105 - loss 0.08663228 - time (sec): 0.16 - samples/sec: 3779.37 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:54,208 epoch 66 - iter 20/105 - loss 0.08645371 - time (sec): 0.33 - samples/sec: 3807.32 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:54,361 epoch 66 - iter 30/105 - loss 0.08114543 - time (sec): 0.48 - samples/sec: 3776.59 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:54,513 epoch 66 - iter 40/105 - loss 0.08233481 - time (sec): 0.63 - samples/sec: 3703.76 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:54,669 epoch 66 - iter 50/105 - loss 0.08185727 - time (sec): 0.79 - samples/sec: 3738.31 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:54,800 epoch 66 - iter 60/105 - loss 0.08319666 - time (sec): 0.92 - samples/sec: 3843.32 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:54,930 epoch 66 - iter 70/105 - loss 0.08390522 - time (sec): 1.05 - samples/sec: 3922.32 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:55,059 epoch 66 - iter 80/105 - loss 0.08176695 - time (sec): 1.18 - samples/sec: 4003.55 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:55,193 epoch 66 - iter 90/105 - loss 0.07874792 - time (sec): 1.31 - samples/sec: 4082.74 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:55,320 epoch 66 - iter 100/105 - loss 0.07989898 - time (sec): 1.44 - samples/sec: 4126.29 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:55,385 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:55,385 EPOCH 66 done: loss 0.0814 - lr: 0.012500 2023-05-15 21:30:56,058 DEV : loss 0.4588263928890228 - accuracy (micro avg) 0.9269 2023-05-15 21:30:56,071 - 3 epochs without improvement 2023-05-15 21:30:56,071 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:56,230 epoch 67 - iter 10/105 - loss 0.06845870 - time (sec): 0.16 - samples/sec: 3905.46 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:56,386 epoch 67 - iter 20/105 - loss 0.08727215 - time (sec): 0.32 - samples/sec: 3764.82 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:56,539 epoch 67 - iter 30/105 - loss 0.07657684 - time (sec): 0.47 - samples/sec: 3771.44 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:56,679 epoch 67 - iter 40/105 - loss 0.08364560 - time (sec): 0.61 - samples/sec: 3889.19 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:56,810 epoch 67 - iter 50/105 - loss 0.07990151 - time (sec): 0.74 - samples/sec: 4035.15 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:56,939 epoch 67 - iter 60/105 - loss 0.07751244 - time (sec): 0.87 - samples/sec: 4144.07 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:57,061 epoch 67 - iter 70/105 - loss 0.07772931 - time (sec): 0.99 - samples/sec: 4169.81 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:57,188 epoch 67 - iter 80/105 - loss 0.07758064 - time (sec): 1.12 - samples/sec: 4225.12 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:57,317 epoch 67 - iter 90/105 - loss 0.07968912 - time (sec): 1.25 - samples/sec: 4258.92 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:57,456 epoch 67 - iter 100/105 - loss 0.08729335 - time (sec): 1.38 - samples/sec: 4313.57 - lr: 0.012500 - momentum: 0.000000 2023-05-15 21:30:57,519 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:57,519 EPOCH 67 done: loss 0.0866 - lr: 0.012500 2023-05-15 21:30:58,192 DEV : loss 0.45442140102386475 - accuracy (micro avg) 0.9276 2023-05-15 21:30:58,204 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.00625] 2023-05-15 21:30:58,204 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:58,360 epoch 68 - iter 10/105 - loss 0.10963916 - time (sec): 0.16 - samples/sec: 3625.19 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:58,512 epoch 68 - iter 20/105 - loss 0.07913487 - time (sec): 0.31 - samples/sec: 3774.24 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:58,671 epoch 68 - iter 30/105 - loss 0.07409651 - time (sec): 0.47 - samples/sec: 3819.83 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:58,818 epoch 68 - iter 40/105 - loss 0.07495369 - time (sec): 0.61 - samples/sec: 3801.11 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:58,964 epoch 68 - iter 50/105 - loss 0.07448298 - time (sec): 0.76 - samples/sec: 3813.46 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:59,116 epoch 68 - iter 60/105 - loss 0.07435783 - time (sec): 0.91 - samples/sec: 3849.08 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:59,248 epoch 68 - iter 70/105 - loss 0.07258293 - time (sec): 1.04 - samples/sec: 3964.16 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:59,374 epoch 68 - iter 80/105 - loss 0.07064422 - time (sec): 1.17 - samples/sec: 4006.50 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:59,507 epoch 68 - iter 90/105 - loss 0.06948889 - time (sec): 1.30 - samples/sec: 4037.94 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:59,644 epoch 68 - iter 100/105 - loss 0.06781102 - time (sec): 1.44 - samples/sec: 4113.60 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:30:59,713 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:30:59,713 EPOCH 68 done: loss 0.0686 - lr: 0.006250 2023-05-15 21:31:00,522 DEV : loss 0.4559873044490814 - accuracy (micro avg) 0.9287 2023-05-15 21:31:00,535 - 1 epochs without improvement 2023-05-15 21:31:00,535 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:00,692 epoch 69 - iter 10/105 - loss 0.08944258 - time (sec): 0.16 - samples/sec: 4206.96 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:00,846 epoch 69 - iter 20/105 - loss 0.10336843 - time (sec): 0.31 - samples/sec: 3939.59 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:00,983 epoch 69 - iter 30/105 - loss 0.09478803 - time (sec): 0.45 - samples/sec: 4098.96 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,111 epoch 69 - iter 40/105 - loss 0.09967452 - time (sec): 0.58 - samples/sec: 4125.35 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,240 epoch 69 - iter 50/105 - loss 0.09868047 - time (sec): 0.71 - samples/sec: 4244.02 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,372 epoch 69 - iter 60/105 - loss 0.09570703 - time (sec): 0.84 - samples/sec: 4329.31 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,501 epoch 69 - iter 70/105 - loss 0.08997756 - time (sec): 0.97 - samples/sec: 4394.20 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,630 epoch 69 - iter 80/105 - loss 0.08595509 - time (sec): 1.09 - samples/sec: 4411.81 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,750 epoch 69 - iter 90/105 - loss 0.08389853 - time (sec): 1.21 - samples/sec: 4403.53 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,879 epoch 69 - iter 100/105 - loss 0.08742698 - time (sec): 1.34 - samples/sec: 4413.41 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:01,943 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:01,943 EPOCH 69 done: loss 0.0885 - lr: 0.006250 2023-05-15 21:31:02,617 DEV : loss 0.45404523611068726 - accuracy (micro avg) 0.9284 2023-05-15 21:31:02,630 - 2 epochs without improvement 2023-05-15 21:31:02,630 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:02,786 epoch 70 - iter 10/105 - loss 0.05790097 - time (sec): 0.16 - samples/sec: 3862.98 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:02,951 epoch 70 - iter 20/105 - loss 0.05650145 - time (sec): 0.32 - samples/sec: 3842.87 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,114 epoch 70 - iter 30/105 - loss 0.06989721 - time (sec): 0.48 - samples/sec: 3892.36 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,266 epoch 70 - iter 40/105 - loss 0.07572799 - time (sec): 0.64 - samples/sec: 3841.64 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,411 epoch 70 - iter 50/105 - loss 0.07726074 - time (sec): 0.78 - samples/sec: 3841.64 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,540 epoch 70 - iter 60/105 - loss 0.07660886 - time (sec): 0.91 - samples/sec: 3979.72 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,666 epoch 70 - iter 70/105 - loss 0.07545937 - time (sec): 1.04 - samples/sec: 4038.69 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,792 epoch 70 - iter 80/105 - loss 0.08047299 - time (sec): 1.16 - samples/sec: 4091.16 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:03,921 epoch 70 - iter 90/105 - loss 0.08212963 - time (sec): 1.29 - samples/sec: 4152.37 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:04,047 epoch 70 - iter 100/105 - loss 0.08173741 - time (sec): 1.42 - samples/sec: 4160.89 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:04,118 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:04,118 EPOCH 70 done: loss 0.0821 - lr: 0.006250 2023-05-15 21:31:04,789 DEV : loss 0.4548334777355194 - accuracy (micro avg) 0.9288 2023-05-15 21:31:04,802 - 0 epochs without improvement 2023-05-15 21:31:04,802 saving best model 2023-05-15 21:31:06,278 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:06,446 epoch 71 - iter 10/105 - loss 0.05925519 - time (sec): 0.17 - samples/sec: 3561.96 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:06,627 epoch 71 - iter 20/105 - loss 0.08341835 - time (sec): 0.35 - samples/sec: 3688.93 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:06,786 epoch 71 - iter 30/105 - loss 0.08201720 - time (sec): 0.51 - samples/sec: 3799.69 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:06,940 epoch 71 - iter 40/105 - loss 0.08307218 - time (sec): 0.66 - samples/sec: 3764.18 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,107 epoch 71 - iter 50/105 - loss 0.07868981 - time (sec): 0.83 - samples/sec: 3729.79 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,253 epoch 71 - iter 60/105 - loss 0.07536770 - time (sec): 0.97 - samples/sec: 3724.63 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,415 epoch 71 - iter 70/105 - loss 0.07608214 - time (sec): 1.14 - samples/sec: 3728.62 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,563 epoch 71 - iter 80/105 - loss 0.07578846 - time (sec): 1.28 - samples/sec: 3705.30 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,717 epoch 71 - iter 90/105 - loss 0.07440881 - time (sec): 1.44 - samples/sec: 3724.39 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,841 epoch 71 - iter 100/105 - loss 0.07231427 - time (sec): 1.56 - samples/sec: 3793.05 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:07,908 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:07,908 EPOCH 71 done: loss 0.0727 - lr: 0.006250 2023-05-15 21:31:08,710 DEV : loss 0.4581780433654785 - accuracy (micro avg) 0.9285 2023-05-15 21:31:08,722 - 1 epochs without improvement 2023-05-15 21:31:08,722 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:08,883 epoch 72 - iter 10/105 - loss 0.05354000 - time (sec): 0.16 - samples/sec: 3907.77 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,033 epoch 72 - iter 20/105 - loss 0.06611641 - time (sec): 0.31 - samples/sec: 3858.84 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,189 epoch 72 - iter 30/105 - loss 0.09281519 - time (sec): 0.47 - samples/sec: 3720.04 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,316 epoch 72 - iter 40/105 - loss 0.09454611 - time (sec): 0.59 - samples/sec: 3910.86 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,445 epoch 72 - iter 50/105 - loss 0.09008855 - time (sec): 0.72 - samples/sec: 4117.39 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,577 epoch 72 - iter 60/105 - loss 0.09084776 - time (sec): 0.86 - samples/sec: 4184.40 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,707 epoch 72 - iter 70/105 - loss 0.08655022 - time (sec): 0.98 - samples/sec: 4259.24 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,833 epoch 72 - iter 80/105 - loss 0.08556694 - time (sec): 1.11 - samples/sec: 4252.43 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:09,963 epoch 72 - iter 90/105 - loss 0.08736164 - time (sec): 1.24 - samples/sec: 4288.88 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:10,088 epoch 72 - iter 100/105 - loss 0.08868731 - time (sec): 1.37 - samples/sec: 4315.20 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:10,159 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:10,159 EPOCH 72 done: loss 0.0863 - lr: 0.006250 2023-05-15 21:31:10,829 DEV : loss 0.45512455701828003 - accuracy (micro avg) 0.9277 2023-05-15 21:31:10,842 - 2 epochs without improvement 2023-05-15 21:31:10,842 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:11,014 epoch 73 - iter 10/105 - loss 0.05618360 - time (sec): 0.17 - samples/sec: 4063.87 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,176 epoch 73 - iter 20/105 - loss 0.06240535 - time (sec): 0.33 - samples/sec: 4017.13 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,317 epoch 73 - iter 30/105 - loss 0.05767794 - time (sec): 0.47 - samples/sec: 4047.38 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,442 epoch 73 - iter 40/105 - loss 0.06854809 - time (sec): 0.60 - samples/sec: 4149.84 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,565 epoch 73 - iter 50/105 - loss 0.06718055 - time (sec): 0.72 - samples/sec: 4220.82 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,693 epoch 73 - iter 60/105 - loss 0.06832022 - time (sec): 0.85 - samples/sec: 4251.94 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,825 epoch 73 - iter 70/105 - loss 0.07204937 - time (sec): 0.98 - samples/sec: 4323.67 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:11,948 epoch 73 - iter 80/105 - loss 0.07038789 - time (sec): 1.11 - samples/sec: 4367.21 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:12,068 epoch 73 - iter 90/105 - loss 0.07173083 - time (sec): 1.23 - samples/sec: 4385.62 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:12,197 epoch 73 - iter 100/105 - loss 0.07347684 - time (sec): 1.36 - samples/sec: 4375.11 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:12,264 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:12,264 EPOCH 73 done: loss 0.0732 - lr: 0.006250 2023-05-15 21:31:12,939 DEV : loss 0.45252254605293274 - accuracy (micro avg) 0.9277 2023-05-15 21:31:12,951 - 3 epochs without improvement 2023-05-15 21:31:12,951 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:13,116 epoch 74 - iter 10/105 - loss 0.02994434 - time (sec): 0.16 - samples/sec: 3313.18 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:13,280 epoch 74 - iter 20/105 - loss 0.04619729 - time (sec): 0.33 - samples/sec: 3407.00 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:13,433 epoch 74 - iter 30/105 - loss 0.05692570 - time (sec): 0.48 - samples/sec: 3506.18 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:13,587 epoch 74 - iter 40/105 - loss 0.06489997 - time (sec): 0.64 - samples/sec: 3678.64 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:13,741 epoch 74 - iter 50/105 - loss 0.06906347 - time (sec): 0.79 - samples/sec: 3707.45 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:13,873 epoch 74 - iter 60/105 - loss 0.07814785 - time (sec): 0.92 - samples/sec: 3834.11 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:14,001 epoch 74 - iter 70/105 - loss 0.07762192 - time (sec): 1.05 - samples/sec: 3913.42 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:14,137 epoch 74 - iter 80/105 - loss 0.07806681 - time (sec): 1.19 - samples/sec: 4023.56 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:14,270 epoch 74 - iter 90/105 - loss 0.07972765 - time (sec): 1.32 - samples/sec: 4095.73 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:14,393 epoch 74 - iter 100/105 - loss 0.08040769 - time (sec): 1.44 - samples/sec: 4103.49 - lr: 0.006250 - momentum: 0.000000 2023-05-15 21:31:14,464 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:14,464 EPOCH 74 done: loss 0.0806 - lr: 0.006250 2023-05-15 21:31:15,266 DEV : loss 0.452541708946228 - accuracy (micro avg) 0.9277 2023-05-15 21:31:15,278 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.003125] 2023-05-15 21:31:15,278 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:15,438 epoch 75 - iter 10/105 - loss 0.05128156 - time (sec): 0.16 - samples/sec: 3811.40 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:15,588 epoch 75 - iter 20/105 - loss 0.05844909 - time (sec): 0.31 - samples/sec: 3738.54 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:15,719 epoch 75 - iter 30/105 - loss 0.06017984 - time (sec): 0.44 - samples/sec: 4060.59 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:15,842 epoch 75 - iter 40/105 - loss 0.05969354 - time (sec): 0.56 - samples/sec: 4217.26 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:15,971 epoch 75 - iter 50/105 - loss 0.07116576 - time (sec): 0.69 - samples/sec: 4355.69 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:16,100 epoch 75 - iter 60/105 - loss 0.07862299 - time (sec): 0.82 - samples/sec: 4368.47 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:16,222 epoch 75 - iter 70/105 - loss 0.07367731 - time (sec): 0.94 - samples/sec: 4354.25 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:16,352 epoch 75 - iter 80/105 - loss 0.07187146 - time (sec): 1.07 - samples/sec: 4375.41 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:16,485 epoch 75 - iter 90/105 - loss 0.06910262 - time (sec): 1.21 - samples/sec: 4380.93 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:16,617 epoch 75 - iter 100/105 - loss 0.06635376 - time (sec): 1.34 - samples/sec: 4421.90 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:16,687 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:16,687 EPOCH 75 done: loss 0.0684 - lr: 0.003125 2023-05-15 21:31:17,359 DEV : loss 0.45416298508644104 - accuracy (micro avg) 0.9281 2023-05-15 21:31:17,372 - 1 epochs without improvement 2023-05-15 21:31:17,372 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:17,523 epoch 76 - iter 10/105 - loss 0.05918599 - time (sec): 0.15 - samples/sec: 3442.17 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:17,671 epoch 76 - iter 20/105 - loss 0.07186231 - time (sec): 0.30 - samples/sec: 3511.07 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:17,834 epoch 76 - iter 30/105 - loss 0.07661525 - time (sec): 0.46 - samples/sec: 3563.78 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:17,995 epoch 76 - iter 40/105 - loss 0.09041049 - time (sec): 0.62 - samples/sec: 3605.29 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,152 epoch 76 - iter 50/105 - loss 0.08889893 - time (sec): 0.78 - samples/sec: 3700.76 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,281 epoch 76 - iter 60/105 - loss 0.08529008 - time (sec): 0.91 - samples/sec: 3901.02 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,408 epoch 76 - iter 70/105 - loss 0.08495219 - time (sec): 1.04 - samples/sec: 3976.59 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,540 epoch 76 - iter 80/105 - loss 0.08500861 - time (sec): 1.17 - samples/sec: 4053.75 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,675 epoch 76 - iter 90/105 - loss 0.08691115 - time (sec): 1.30 - samples/sec: 4117.47 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,805 epoch 76 - iter 100/105 - loss 0.08528640 - time (sec): 1.43 - samples/sec: 4157.67 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:18,870 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:18,870 EPOCH 76 done: loss 0.0831 - lr: 0.003125 2023-05-15 21:31:19,545 DEV : loss 0.4525325298309326 - accuracy (micro avg) 0.9284 2023-05-15 21:31:19,558 - 2 epochs without improvement 2023-05-15 21:31:19,558 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:19,708 epoch 77 - iter 10/105 - loss 0.05640535 - time (sec): 0.15 - samples/sec: 3978.59 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:19,860 epoch 77 - iter 20/105 - loss 0.07251769 - time (sec): 0.30 - samples/sec: 3821.23 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,007 epoch 77 - iter 30/105 - loss 0.05696852 - time (sec): 0.45 - samples/sec: 3803.28 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,167 epoch 77 - iter 40/105 - loss 0.06423022 - time (sec): 0.61 - samples/sec: 3810.51 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,329 epoch 77 - iter 50/105 - loss 0.06844753 - time (sec): 0.77 - samples/sec: 3821.47 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,487 epoch 77 - iter 60/105 - loss 0.07693215 - time (sec): 0.93 - samples/sec: 3874.91 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,652 epoch 77 - iter 70/105 - loss 0.07380444 - time (sec): 1.09 - samples/sec: 3829.16 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,804 epoch 77 - iter 80/105 - loss 0.07705450 - time (sec): 1.25 - samples/sec: 3827.57 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:20,970 epoch 77 - iter 90/105 - loss 0.07583402 - time (sec): 1.41 - samples/sec: 3800.11 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:21,125 epoch 77 - iter 100/105 - loss 0.07763649 - time (sec): 1.57 - samples/sec: 3779.50 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:21,203 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:21,203 EPOCH 77 done: loss 0.0758 - lr: 0.003125 2023-05-15 21:31:21,874 DEV : loss 0.4528382122516632 - accuracy (micro avg) 0.9281 2023-05-15 21:31:21,887 - 3 epochs without improvement 2023-05-15 21:31:21,887 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:22,038 epoch 78 - iter 10/105 - loss 0.04706085 - time (sec): 0.15 - samples/sec: 3537.22 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:22,197 epoch 78 - iter 20/105 - loss 0.06569906 - time (sec): 0.31 - samples/sec: 3743.21 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:22,343 epoch 78 - iter 30/105 - loss 0.07837444 - time (sec): 0.46 - samples/sec: 3690.38 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:22,496 epoch 78 - iter 40/105 - loss 0.07520407 - time (sec): 0.61 - samples/sec: 3718.03 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:22,654 epoch 78 - iter 50/105 - loss 0.07961023 - time (sec): 0.77 - samples/sec: 3764.60 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:22,780 epoch 78 - iter 60/105 - loss 0.08524695 - time (sec): 0.89 - samples/sec: 3911.94 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:22,912 epoch 78 - iter 70/105 - loss 0.08718972 - time (sec): 1.03 - samples/sec: 4034.00 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:23,037 epoch 78 - iter 80/105 - loss 0.08715969 - time (sec): 1.15 - samples/sec: 4124.13 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:23,158 epoch 78 - iter 90/105 - loss 0.08897209 - time (sec): 1.27 - samples/sec: 4149.10 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:23,297 epoch 78 - iter 100/105 - loss 0.09199450 - time (sec): 1.41 - samples/sec: 4206.78 - lr: 0.003125 - momentum: 0.000000 2023-05-15 21:31:23,363 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:23,363 EPOCH 78 done: loss 0.0927 - lr: 0.003125 2023-05-15 21:31:24,176 DEV : loss 0.45365533232688904 - accuracy (micro avg) 0.9284 2023-05-15 21:31:24,188 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.0015625] 2023-05-15 21:31:24,189 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:24,350 epoch 79 - iter 10/105 - loss 0.09398302 - time (sec): 0.16 - samples/sec: 3712.77 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:24,504 epoch 79 - iter 20/105 - loss 0.07434979 - time (sec): 0.32 - samples/sec: 3928.85 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:24,666 epoch 79 - iter 30/105 - loss 0.08088376 - time (sec): 0.48 - samples/sec: 3863.00 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:24,821 epoch 79 - iter 40/105 - loss 0.08310857 - time (sec): 0.63 - samples/sec: 3812.20 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:24,979 epoch 79 - iter 50/105 - loss 0.07883039 - time (sec): 0.79 - samples/sec: 3788.24 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:25,132 epoch 79 - iter 60/105 - loss 0.08249951 - time (sec): 0.94 - samples/sec: 3807.23 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:25,293 epoch 79 - iter 70/105 - loss 0.08213901 - time (sec): 1.10 - samples/sec: 3808.40 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:25,450 epoch 79 - iter 80/105 - loss 0.07941353 - time (sec): 1.26 - samples/sec: 3763.70 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:25,609 epoch 79 - iter 90/105 - loss 0.07842659 - time (sec): 1.42 - samples/sec: 3803.21 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:25,758 epoch 79 - iter 100/105 - loss 0.08078698 - time (sec): 1.57 - samples/sec: 3769.21 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:25,838 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:25,838 EPOCH 79 done: loss 0.0814 - lr: 0.001563 2023-05-15 21:31:26,512 DEV : loss 0.45469412207603455 - accuracy (micro avg) 0.9287 2023-05-15 21:31:26,525 - 1 epochs without improvement 2023-05-15 21:31:26,525 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:26,674 epoch 80 - iter 10/105 - loss 0.08687815 - time (sec): 0.15 - samples/sec: 3496.86 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:26,832 epoch 80 - iter 20/105 - loss 0.08161370 - time (sec): 0.31 - samples/sec: 3707.59 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:26,963 epoch 80 - iter 30/105 - loss 0.08213428 - time (sec): 0.44 - samples/sec: 4044.88 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,091 epoch 80 - iter 40/105 - loss 0.07656094 - time (sec): 0.57 - samples/sec: 4128.56 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,225 epoch 80 - iter 50/105 - loss 0.07999229 - time (sec): 0.70 - samples/sec: 4259.62 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,356 epoch 80 - iter 60/105 - loss 0.08157347 - time (sec): 0.83 - samples/sec: 4344.42 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,482 epoch 80 - iter 70/105 - loss 0.08251190 - time (sec): 0.96 - samples/sec: 4392.56 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,608 epoch 80 - iter 80/105 - loss 0.08859252 - time (sec): 1.08 - samples/sec: 4404.76 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,732 epoch 80 - iter 90/105 - loss 0.08405423 - time (sec): 1.21 - samples/sec: 4405.96 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,865 epoch 80 - iter 100/105 - loss 0.08321391 - time (sec): 1.34 - samples/sec: 4431.27 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:27,928 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:27,928 EPOCH 80 done: loss 0.0847 - lr: 0.001563 2023-05-15 21:31:28,600 DEV : loss 0.45543888211250305 - accuracy (micro avg) 0.9281 2023-05-15 21:31:28,612 - 2 epochs without improvement 2023-05-15 21:31:28,612 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:28,780 epoch 81 - iter 10/105 - loss 0.09104405 - time (sec): 0.17 - samples/sec: 3878.73 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:28,939 epoch 81 - iter 20/105 - loss 0.09779127 - time (sec): 0.33 - samples/sec: 3779.11 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,068 epoch 81 - iter 30/105 - loss 0.08391870 - time (sec): 0.46 - samples/sec: 3956.76 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,198 epoch 81 - iter 40/105 - loss 0.08020966 - time (sec): 0.59 - samples/sec: 4117.64 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,324 epoch 81 - iter 50/105 - loss 0.07895226 - time (sec): 0.71 - samples/sec: 4247.92 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,454 epoch 81 - iter 60/105 - loss 0.07539052 - time (sec): 0.84 - samples/sec: 4255.05 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,574 epoch 81 - iter 70/105 - loss 0.07256490 - time (sec): 0.96 - samples/sec: 4249.92 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,707 epoch 81 - iter 80/105 - loss 0.07126569 - time (sec): 1.10 - samples/sec: 4375.21 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,832 epoch 81 - iter 90/105 - loss 0.07536926 - time (sec): 1.22 - samples/sec: 4392.35 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:29,958 epoch 81 - iter 100/105 - loss 0.07462281 - time (sec): 1.35 - samples/sec: 4402.65 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:30,028 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:30,028 EPOCH 81 done: loss 0.0738 - lr: 0.001563 2023-05-15 21:31:30,834 DEV : loss 0.45619630813598633 - accuracy (micro avg) 0.9281 2023-05-15 21:31:30,846 - 3 epochs without improvement 2023-05-15 21:31:30,846 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:31,008 epoch 82 - iter 10/105 - loss 0.12525001 - time (sec): 0.16 - samples/sec: 3724.13 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,171 epoch 82 - iter 20/105 - loss 0.11689806 - time (sec): 0.32 - samples/sec: 3620.25 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,299 epoch 82 - iter 30/105 - loss 0.10251425 - time (sec): 0.45 - samples/sec: 3802.14 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,419 epoch 82 - iter 40/105 - loss 0.09642245 - time (sec): 0.57 - samples/sec: 3950.44 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,549 epoch 82 - iter 50/105 - loss 0.09190527 - time (sec): 0.70 - samples/sec: 4075.66 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,674 epoch 82 - iter 60/105 - loss 0.09405102 - time (sec): 0.83 - samples/sec: 4142.18 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,804 epoch 82 - iter 70/105 - loss 0.08565949 - time (sec): 0.96 - samples/sec: 4217.21 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:31,929 epoch 82 - iter 80/105 - loss 0.08256859 - time (sec): 1.08 - samples/sec: 4213.75 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:32,059 epoch 82 - iter 90/105 - loss 0.08123900 - time (sec): 1.21 - samples/sec: 4286.50 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:32,194 epoch 82 - iter 100/105 - loss 0.07714071 - time (sec): 1.35 - samples/sec: 4364.26 - lr: 0.001563 - momentum: 0.000000 2023-05-15 21:31:32,267 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:32,267 EPOCH 82 done: loss 0.0819 - lr: 0.001563 2023-05-15 21:31:32,940 DEV : loss 0.45649954676628113 - accuracy (micro avg) 0.9282 2023-05-15 21:31:32,953 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.00078125] 2023-05-15 21:31:32,953 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:33,111 epoch 83 - iter 10/105 - loss 0.06625568 - time (sec): 0.16 - samples/sec: 3815.44 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:33,267 epoch 83 - iter 20/105 - loss 0.06653766 - time (sec): 0.31 - samples/sec: 3872.72 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:33,428 epoch 83 - iter 30/105 - loss 0.07237229 - time (sec): 0.47 - samples/sec: 3838.24 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:33,580 epoch 83 - iter 40/105 - loss 0.06671547 - time (sec): 0.63 - samples/sec: 3708.51 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:33,735 epoch 83 - iter 50/105 - loss 0.06586352 - time (sec): 0.78 - samples/sec: 3696.60 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:33,890 epoch 83 - iter 60/105 - loss 0.06541506 - time (sec): 0.94 - samples/sec: 3706.22 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:34,059 epoch 83 - iter 70/105 - loss 0.06899117 - time (sec): 1.11 - samples/sec: 3742.73 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:34,217 epoch 83 - iter 80/105 - loss 0.06740104 - time (sec): 1.26 - samples/sec: 3707.03 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:34,373 epoch 83 - iter 90/105 - loss 0.06348514 - time (sec): 1.42 - samples/sec: 3712.48 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:34,514 epoch 83 - iter 100/105 - loss 0.06313693 - time (sec): 1.56 - samples/sec: 3794.15 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:34,582 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:34,582 EPOCH 83 done: loss 0.0646 - lr: 0.000781 2023-05-15 21:31:35,256 DEV : loss 0.4569226801395416 - accuracy (micro avg) 0.9282 2023-05-15 21:31:35,268 - 1 epochs without improvement 2023-05-15 21:31:35,268 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:35,425 epoch 84 - iter 10/105 - loss 0.05634396 - time (sec): 0.16 - samples/sec: 3795.45 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:35,589 epoch 84 - iter 20/105 - loss 0.05880624 - time (sec): 0.32 - samples/sec: 3615.23 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:35,746 epoch 84 - iter 30/105 - loss 0.05553765 - time (sec): 0.48 - samples/sec: 3651.73 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:35,898 epoch 84 - iter 40/105 - loss 0.05421690 - time (sec): 0.63 - samples/sec: 3608.96 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,051 epoch 84 - iter 50/105 - loss 0.06769443 - time (sec): 0.78 - samples/sec: 3651.20 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,183 epoch 84 - iter 60/105 - loss 0.06904524 - time (sec): 0.91 - samples/sec: 3815.75 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,307 epoch 84 - iter 70/105 - loss 0.06988047 - time (sec): 1.04 - samples/sec: 3923.91 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,438 epoch 84 - iter 80/105 - loss 0.06880722 - time (sec): 1.17 - samples/sec: 4015.36 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,567 epoch 84 - iter 90/105 - loss 0.06763182 - time (sec): 1.30 - samples/sec: 4079.92 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,694 epoch 84 - iter 100/105 - loss 0.06676447 - time (sec): 1.43 - samples/sec: 4148.14 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:36,762 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:36,762 EPOCH 84 done: loss 0.0678 - lr: 0.000781 2023-05-15 21:31:37,575 DEV : loss 0.4565540850162506 - accuracy (micro avg) 0.9281 2023-05-15 21:31:37,587 - 2 epochs without improvement 2023-05-15 21:31:37,587 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:37,757 epoch 85 - iter 10/105 - loss 0.06256486 - time (sec): 0.17 - samples/sec: 3677.85 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:37,914 epoch 85 - iter 20/105 - loss 0.07293128 - time (sec): 0.33 - samples/sec: 3592.56 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:38,066 epoch 85 - iter 30/105 - loss 0.06617007 - time (sec): 0.48 - samples/sec: 3619.65 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:38,222 epoch 85 - iter 40/105 - loss 0.07255578 - time (sec): 0.63 - samples/sec: 3677.53 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:38,384 epoch 85 - iter 50/105 - loss 0.06989969 - time (sec): 0.80 - samples/sec: 3776.48 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:38,530 epoch 85 - iter 60/105 - loss 0.06554395 - time (sec): 0.94 - samples/sec: 3740.88 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:38,688 epoch 85 - iter 70/105 - loss 0.06226723 - time (sec): 1.10 - samples/sec: 3780.25 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:38,847 epoch 85 - iter 80/105 - loss 0.06101403 - time (sec): 1.26 - samples/sec: 3820.26 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:39,005 epoch 85 - iter 90/105 - loss 0.06516340 - time (sec): 1.42 - samples/sec: 3797.73 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:39,159 epoch 85 - iter 100/105 - loss 0.06652169 - time (sec): 1.57 - samples/sec: 3761.44 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:39,231 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:39,231 EPOCH 85 done: loss 0.0670 - lr: 0.000781 2023-05-15 21:31:39,903 DEV : loss 0.45625191926956177 - accuracy (micro avg) 0.9284 2023-05-15 21:31:39,916 - 3 epochs without improvement 2023-05-15 21:31:39,916 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:40,078 epoch 86 - iter 10/105 - loss 0.04161077 - time (sec): 0.16 - samples/sec: 3363.93 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:40,251 epoch 86 - iter 20/105 - loss 0.05284345 - time (sec): 0.33 - samples/sec: 3447.78 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:40,405 epoch 86 - iter 30/105 - loss 0.05720211 - time (sec): 0.49 - samples/sec: 3544.07 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:40,556 epoch 86 - iter 40/105 - loss 0.05855859 - time (sec): 0.64 - samples/sec: 3620.52 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:40,723 epoch 86 - iter 50/105 - loss 0.06731156 - time (sec): 0.81 - samples/sec: 3633.73 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:40,892 epoch 86 - iter 60/105 - loss 0.06409584 - time (sec): 0.98 - samples/sec: 3624.43 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:41,050 epoch 86 - iter 70/105 - loss 0.06448159 - time (sec): 1.13 - samples/sec: 3610.54 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:41,201 epoch 86 - iter 80/105 - loss 0.06684483 - time (sec): 1.28 - samples/sec: 3617.48 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:41,373 epoch 86 - iter 90/105 - loss 0.07065451 - time (sec): 1.46 - samples/sec: 3609.49 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:41,539 epoch 86 - iter 100/105 - loss 0.07105019 - time (sec): 1.62 - samples/sec: 3651.39 - lr: 0.000781 - momentum: 0.000000 2023-05-15 21:31:41,623 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:41,623 EPOCH 86 done: loss 0.0707 - lr: 0.000781 2023-05-15 21:31:42,293 DEV : loss 0.4558914303779602 - accuracy (micro avg) 0.9285 2023-05-15 21:31:42,305 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.000390625] 2023-05-15 21:31:42,305 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:42,466 epoch 87 - iter 10/105 - loss 0.04982021 - time (sec): 0.16 - samples/sec: 3886.65 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:42,623 epoch 87 - iter 20/105 - loss 0.08063016 - time (sec): 0.32 - samples/sec: 3594.55 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:42,767 epoch 87 - iter 30/105 - loss 0.07532039 - time (sec): 0.46 - samples/sec: 3603.82 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:42,937 epoch 87 - iter 40/105 - loss 0.07826631 - time (sec): 0.63 - samples/sec: 3606.43 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,105 epoch 87 - iter 50/105 - loss 0.07813172 - time (sec): 0.80 - samples/sec: 3646.58 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,268 epoch 87 - iter 60/105 - loss 0.08340842 - time (sec): 0.96 - samples/sec: 3654.19 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,423 epoch 87 - iter 70/105 - loss 0.07727099 - time (sec): 1.12 - samples/sec: 3714.27 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,575 epoch 87 - iter 80/105 - loss 0.07791657 - time (sec): 1.27 - samples/sec: 3727.90 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,735 epoch 87 - iter 90/105 - loss 0.07602732 - time (sec): 1.43 - samples/sec: 3703.95 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,871 epoch 87 - iter 100/105 - loss 0.07647325 - time (sec): 1.57 - samples/sec: 3788.99 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:43,937 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:43,937 EPOCH 87 done: loss 0.0766 - lr: 0.000391 2023-05-15 21:31:44,741 DEV : loss 0.45574912428855896 - accuracy (micro avg) 0.9282 2023-05-15 21:31:44,753 - 1 epochs without improvement 2023-05-15 21:31:44,754 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:44,913 epoch 88 - iter 10/105 - loss 0.07622109 - time (sec): 0.16 - samples/sec: 3928.45 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,073 epoch 88 - iter 20/105 - loss 0.08673939 - time (sec): 0.32 - samples/sec: 3721.81 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,241 epoch 88 - iter 30/105 - loss 0.09478922 - time (sec): 0.49 - samples/sec: 3719.78 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,385 epoch 88 - iter 40/105 - loss 0.09433885 - time (sec): 0.63 - samples/sec: 3828.89 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,515 epoch 88 - iter 50/105 - loss 0.09348278 - time (sec): 0.76 - samples/sec: 3975.06 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,641 epoch 88 - iter 60/105 - loss 0.08843286 - time (sec): 0.89 - samples/sec: 4030.94 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,767 epoch 88 - iter 70/105 - loss 0.08531614 - time (sec): 1.01 - samples/sec: 4080.20 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:45,901 epoch 88 - iter 80/105 - loss 0.08199163 - time (sec): 1.15 - samples/sec: 4150.58 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:46,028 epoch 88 - iter 90/105 - loss 0.08255142 - time (sec): 1.27 - samples/sec: 4176.91 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:46,154 epoch 88 - iter 100/105 - loss 0.07994260 - time (sec): 1.40 - samples/sec: 4233.18 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:46,223 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:46,223 EPOCH 88 done: loss 0.0802 - lr: 0.000391 2023-05-15 21:31:46,894 DEV : loss 0.4561077356338501 - accuracy (micro avg) 0.9282 2023-05-15 21:31:46,906 - 2 epochs without improvement 2023-05-15 21:31:46,906 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:47,070 epoch 89 - iter 10/105 - loss 0.07080775 - time (sec): 0.16 - samples/sec: 3693.68 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:47,223 epoch 89 - iter 20/105 - loss 0.07455936 - time (sec): 0.32 - samples/sec: 3613.52 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:47,394 epoch 89 - iter 30/105 - loss 0.07381745 - time (sec): 0.49 - samples/sec: 3669.90 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:47,523 epoch 89 - iter 40/105 - loss 0.08314483 - time (sec): 0.62 - samples/sec: 3816.89 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:47,649 epoch 89 - iter 50/105 - loss 0.07774133 - time (sec): 0.74 - samples/sec: 4036.84 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:47,777 epoch 89 - iter 60/105 - loss 0.07910999 - time (sec): 0.87 - samples/sec: 4118.80 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:47,901 epoch 89 - iter 70/105 - loss 0.07932378 - time (sec): 0.99 - samples/sec: 4202.82 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:48,033 epoch 89 - iter 80/105 - loss 0.07679682 - time (sec): 1.13 - samples/sec: 4251.01 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:48,164 epoch 89 - iter 90/105 - loss 0.07742812 - time (sec): 1.26 - samples/sec: 4295.76 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:48,289 epoch 89 - iter 100/105 - loss 0.07785464 - time (sec): 1.38 - samples/sec: 4305.90 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:48,357 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:48,357 EPOCH 89 done: loss 0.0786 - lr: 0.000391 2023-05-15 21:31:49,028 DEV : loss 0.45609503984451294 - accuracy (micro avg) 0.9282 2023-05-15 21:31:49,040 - 3 epochs without improvement 2023-05-15 21:31:49,040 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:49,204 epoch 90 - iter 10/105 - loss 0.05868137 - time (sec): 0.16 - samples/sec: 3826.04 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:49,349 epoch 90 - iter 20/105 - loss 0.09696209 - time (sec): 0.31 - samples/sec: 3764.51 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:49,507 epoch 90 - iter 30/105 - loss 0.09615798 - time (sec): 0.47 - samples/sec: 3705.81 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:49,667 epoch 90 - iter 40/105 - loss 0.08374592 - time (sec): 0.63 - samples/sec: 3715.04 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:49,823 epoch 90 - iter 50/105 - loss 0.07703084 - time (sec): 0.78 - samples/sec: 3662.11 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:49,978 epoch 90 - iter 60/105 - loss 0.08614811 - time (sec): 0.94 - samples/sec: 3732.03 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:50,131 epoch 90 - iter 70/105 - loss 0.09328989 - time (sec): 1.09 - samples/sec: 3688.87 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:50,280 epoch 90 - iter 80/105 - loss 0.08829302 - time (sec): 1.24 - samples/sec: 3726.86 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:50,418 epoch 90 - iter 90/105 - loss 0.08347571 - time (sec): 1.38 - samples/sec: 3848.06 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:50,553 epoch 90 - iter 100/105 - loss 0.08183453 - time (sec): 1.51 - samples/sec: 3907.42 - lr: 0.000391 - momentum: 0.000000 2023-05-15 21:31:50,620 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:50,620 EPOCH 90 done: loss 0.0803 - lr: 0.000391 2023-05-15 21:31:51,291 DEV : loss 0.4561174511909485 - accuracy (micro avg) 0.9282 2023-05-15 21:31:51,304 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [0.0001953125] 2023-05-15 21:31:51,304 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:51,463 epoch 91 - iter 10/105 - loss 0.10163174 - time (sec): 0.16 - samples/sec: 3736.41 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:51,612 epoch 91 - iter 20/105 - loss 0.08917681 - time (sec): 0.31 - samples/sec: 3612.29 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:51,768 epoch 91 - iter 30/105 - loss 0.07233568 - time (sec): 0.46 - samples/sec: 3660.52 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:51,923 epoch 91 - iter 40/105 - loss 0.07080886 - time (sec): 0.62 - samples/sec: 3707.63 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,087 epoch 91 - iter 50/105 - loss 0.06833288 - time (sec): 0.78 - samples/sec: 3758.81 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,237 epoch 91 - iter 60/105 - loss 0.06462794 - time (sec): 0.93 - samples/sec: 3804.06 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,391 epoch 91 - iter 70/105 - loss 0.06432342 - time (sec): 1.09 - samples/sec: 3788.99 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,524 epoch 91 - iter 80/105 - loss 0.06133304 - time (sec): 1.22 - samples/sec: 3898.58 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,648 epoch 91 - iter 90/105 - loss 0.06374294 - time (sec): 1.34 - samples/sec: 3970.37 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,775 epoch 91 - iter 100/105 - loss 0.06553967 - time (sec): 1.47 - samples/sec: 4014.74 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:52,842 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:52,842 EPOCH 91 done: loss 0.0674 - lr: 0.000195 2023-05-15 21:31:53,652 DEV : loss 0.4560907185077667 - accuracy (micro avg) 0.9282 2023-05-15 21:31:53,664 - 1 epochs without improvement 2023-05-15 21:31:53,664 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:53,830 epoch 92 - iter 10/105 - loss 0.09787210 - time (sec): 0.17 - samples/sec: 3797.34 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:53,989 epoch 92 - iter 20/105 - loss 0.08095827 - time (sec): 0.33 - samples/sec: 3682.45 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,141 epoch 92 - iter 30/105 - loss 0.07818815 - time (sec): 0.48 - samples/sec: 3688.05 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,293 epoch 92 - iter 40/105 - loss 0.07613048 - time (sec): 0.63 - samples/sec: 3823.41 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,420 epoch 92 - iter 50/105 - loss 0.07859168 - time (sec): 0.76 - samples/sec: 3881.09 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,550 epoch 92 - iter 60/105 - loss 0.07787762 - time (sec): 0.89 - samples/sec: 3994.22 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,676 epoch 92 - iter 70/105 - loss 0.08210023 - time (sec): 1.01 - samples/sec: 4065.12 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,807 epoch 92 - iter 80/105 - loss 0.08262567 - time (sec): 1.14 - samples/sec: 4158.56 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:54,932 epoch 92 - iter 90/105 - loss 0.08175313 - time (sec): 1.27 - samples/sec: 4191.78 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:55,059 epoch 92 - iter 100/105 - loss 0.08208601 - time (sec): 1.39 - samples/sec: 4205.72 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:55,130 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:55,131 EPOCH 92 done: loss 0.0814 - lr: 0.000195 2023-05-15 21:31:55,801 DEV : loss 0.45620062947273254 - accuracy (micro avg) 0.9282 2023-05-15 21:31:55,813 - 2 epochs without improvement 2023-05-15 21:31:55,813 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:55,966 epoch 93 - iter 10/105 - loss 0.08338736 - time (sec): 0.15 - samples/sec: 3324.33 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,116 epoch 93 - iter 20/105 - loss 0.09952859 - time (sec): 0.30 - samples/sec: 3364.67 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,272 epoch 93 - iter 30/105 - loss 0.09269119 - time (sec): 0.46 - samples/sec: 3448.47 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,428 epoch 93 - iter 40/105 - loss 0.08263464 - time (sec): 0.61 - samples/sec: 3554.14 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,563 epoch 93 - iter 50/105 - loss 0.07815594 - time (sec): 0.75 - samples/sec: 3720.25 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,687 epoch 93 - iter 60/105 - loss 0.07279853 - time (sec): 0.87 - samples/sec: 3844.56 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,813 epoch 93 - iter 70/105 - loss 0.06974049 - time (sec): 1.00 - samples/sec: 3966.49 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:56,940 epoch 93 - iter 80/105 - loss 0.07093432 - time (sec): 1.13 - samples/sec: 4056.96 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:57,079 epoch 93 - iter 90/105 - loss 0.06798311 - time (sec): 1.27 - samples/sec: 4180.91 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:57,209 epoch 93 - iter 100/105 - loss 0.06784165 - time (sec): 1.40 - samples/sec: 4266.46 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:57,274 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:57,274 EPOCH 93 done: loss 0.0690 - lr: 0.000195 2023-05-15 21:31:57,945 DEV : loss 0.45623788237571716 - accuracy (micro avg) 0.9281 2023-05-15 21:31:57,957 - 3 epochs without improvement 2023-05-15 21:31:57,957 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:58,125 epoch 94 - iter 10/105 - loss 0.08754979 - time (sec): 0.17 - samples/sec: 3932.26 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:58,285 epoch 94 - iter 20/105 - loss 0.08029897 - time (sec): 0.33 - samples/sec: 3869.40 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:58,460 epoch 94 - iter 30/105 - loss 0.08021386 - time (sec): 0.50 - samples/sec: 3876.31 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:58,622 epoch 94 - iter 40/105 - loss 0.08244415 - time (sec): 0.66 - samples/sec: 3822.84 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:58,780 epoch 94 - iter 50/105 - loss 0.08622380 - time (sec): 0.82 - samples/sec: 3797.24 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:58,933 epoch 94 - iter 60/105 - loss 0.08088922 - time (sec): 0.98 - samples/sec: 3785.69 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:59,077 epoch 94 - iter 70/105 - loss 0.07714372 - time (sec): 1.12 - samples/sec: 3802.97 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:59,229 epoch 94 - iter 80/105 - loss 0.07837018 - time (sec): 1.27 - samples/sec: 3800.98 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:59,382 epoch 94 - iter 90/105 - loss 0.07799857 - time (sec): 1.42 - samples/sec: 3792.89 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:59,534 epoch 94 - iter 100/105 - loss 0.07581181 - time (sec): 1.58 - samples/sec: 3812.30 - lr: 0.000195 - momentum: 0.000000 2023-05-15 21:31:59,598 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:31:59,598 EPOCH 94 done: loss 0.0771 - lr: 0.000195 2023-05-15 21:32:00,407 DEV : loss 0.45624786615371704 - accuracy (micro avg) 0.9282 2023-05-15 21:32:00,419 - 4 epochs without improvement (above 'patience')-> annealing learning_rate to [9.765625e-05] 2023-05-15 21:32:00,419 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:32:00,419 learning rate too small - quitting training! 2023-05-15 21:32:00,419 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:32:00,419 Saving model ... 2023-05-15 21:32:01,555 Done. 2023-05-15 21:32:01,555 ---------------------------------------------------------------------------------------------------- 2023-05-15 21:32:01,555 Loading model from best epoch ... 2023-05-15 21:32:03,450 SequenceTagger predicts: Dictionary with 71 tags: NN, $., NE, ADV, APPR, ART, VVFIN, PPER, ADDRESS, $(, ADJA, URL, VAFIN, ADJD, $,, HASH, KON, CARD, VVINF, APPRART, VVPP, EMO, VMFIN, PIS, PTKNEG, PDS, KOUS, PPOSAT, PTKVZ, PIAT, PRF, XYB, ITJ, PWAV, FM, PROAV, PWS, XY, PRELS, VAINF, VVIMP, PDAT, KOKOM, PTKZU, PTKANT, PAUSE, VVFIN_ES, PTKA, VVIZU, NINFL 2023-05-15 21:32:05,113 Results: - F-score (micro) 0.9316 - F-score (macro) 0.6573 - Accuracy 0.9316 By class: precision recall f1-score support NN 0.9181 0.9339 0.9260 1165 $. 0.9812 0.9946 0.9879 736 ADV 0.9162 0.9307 0.9234 505 NE 0.8333 0.8385 0.8359 483 APPR 0.9582 0.9750 0.9665 400 ART 0.9667 0.9915 0.9789 351 VVFIN 0.9453 0.9333 0.9393 315 PPER 0.9542 0.9927 0.9731 273 $( 0.9808 0.9623 0.9714 265 ADDRESS 0.9274 0.9914 0.9583 232 VAFIN 0.9775 0.9864 0.9819 220 URL 0.9910 1.0000 0.9955 220 ADJA 0.9312 0.9355 0.9333 217 ADJD 0.9149 0.7818 0.8431 220 $, 1.0000 1.0000 1.0000 198 HASH 0.9184 0.9507 0.9343 142 KON 0.9568 0.9433 0.9500 141 VVINF 0.8125 0.9100 0.8585 100 CARD 0.9623 0.9808 0.9714 104 VVPP 0.8725 0.9368 0.9036 95 APPRART 1.0000 1.0000 1.0000 97 EMO 0.8812 0.9780 0.9271 91 VMFIN 0.8481 0.9710 0.9054 69 PIS 0.8714 0.9104 0.8905 67 PDS 0.9385 0.8971 0.9173 68 PTKNEG 1.0000 1.0000 1.0000 58 PPOSAT 1.0000 0.9636 0.9815 55 KOUS 0.9375 0.9184 0.9278 49 PTKVZ 0.7778 0.7143 0.7447 49 PIAT 0.9000 0.8780 0.8889 41 ITJ 0.7000 0.5526 0.6176 38 PWAV 0.9062 0.9667 0.9355 30 PROAV 0.7931 0.7667 0.7797 30 PRF 0.9545 0.6562 0.7778 32 XYB 0.9600 0.8889 0.9231 27 PWS 1.0000 0.8148 0.8980 27 VAINF 1.0000 0.9524 0.9756 21 PDAT 0.9048 0.9500 0.9268 20 FM 1.0000 0.6400 0.7805 25 XY 0.7778 0.2500 0.3784 28 PTKANT 0.8947 0.9444 0.9189 18 PTKZU 0.9286 0.9286 0.9286 14 PRELS 0.5714 0.7273 0.6400 11 KOKOM 0.8333 0.8333 0.8333 12 VVFIN_ES 1.0000 0.7000 0.8235 10 PAUSE 0.4000 0.2500 0.3077 8 VVIMP 0.7500 0.3333 0.4615 9 PTKA 0.5000 0.1429 0.2222 7 PWAT 1.0000 1.0000 1.0000 4 PTKREZ 0.6667 0.5000 0.5714 4 VVIZU 0.0000 0.0000 0.0000 4 PTKPAU 0.5000 1.0000 0.6667 1 NINFL 0.0000 0.0000 0.0000 3 VAFIN_ES 0.0000 0.0000 0.0000 2 VMINF 0.0000 0.0000 0.0000 2 KOUS_ES 0.0000 0.0000 0.0000 1 VAPP 0.0000 0.0000 0.0000 1 XYU 0.0000 0.0000 0.0000 1 PIS_PPER 0.0000 0.0000 0.0000 1 KOUI 0.0000 0.0000 0.0000 1 VAFIN_PPER 0.0000 0.0000 0.0000 1 TRUNC 0.0000 0.0000 0.0000 1 VVFIN_DU 0.0000 0.0000 0.0000 1 PTKONO 0.0000 0.0000 0.0000 1 PTKQU 0.0000 0.0000 0.0000 1 PTK 0.0000 0.0000 0.0000 0 accuracy 0.9316 7423 macro avg 0.6805 0.6515 0.6573 7423 weighted avg 0.9284 0.9316 0.9285 7423 2023-05-15 21:32:05,113 ----------------------------------------------------------------------------------------------------