Created by: lifubang

root@cb462c7b76b1:/opt# git clone https://github.com/PaddlePaddle/book 正克隆到 'book'... remote: Counting objects: 4782, done. remote: Total 4782 (delta 0), reused 0 (delta 0), pack-reused 4782 接收对象中: 100% (4782/4782), 33.64 MiB | 281.00 KiB/s, 完成. 处理 delta 中: 100% (3156/3156), 完成. 检查连接... 完成。 root@cb462c7b76b1:/opt# cd book/ root@cb462c7b76b1:/opt/book# ls 01.fit_a_line 05.recommender_system index.cn.html pending 02.recognize_digits 06.understand_sentiment index.html README.cn.md 03.image_classification 07.label_semantic_roles mnist-client README.md 04.word2vec 08.machine_translation paddle serve root@cb462c7b76b1:/opt/book# cd 02.recognize_digits/ root@cb462c7b76b1:/opt/book/02.recognize_digits# ls client image index.cn.html index.html README.cn.md README.md train.py root@cb462c7b76b1:/opt/book/02.recognize_digits# python2 train.py I0122 08:59:13.437571 121 Util.cpp:166] commandline: --use_gpu=False --trainer_count=1 [INFO 2018-01-22 08:59:13,443 layers.py:2689] output for __conv_pool_0___conv: c = 20, h = 24, w = 24, size = 11520 [INFO 2018-01-22 08:59:13,444 layers.py:2829] output for __conv_pool_0___pool: c = 20, h = 12, w = 12, size = 2880 [INFO 2018-01-22 08:59:13,445 layers.py:2689] output for __conv_pool_1___conv: c = 50, h = 8, w = 8, size = 3200 [INFO 2018-01-22 08:59:13,446 layers.py:2829] output for __conv_pool_1___pool: c = 50, h = 4, w = 4, size = 800 I0122 08:59:13.449390 121 GradientMachine.cpp:94] Initing parameters.. I0122 08:59:13.451611 121 GradientMachine.cpp:101] Init parameters done. Pass 0, Batch 0, Cost 2.836447, {'classification_error_evaluator': 0.875} Pass 0, Batch 100, Cost 2.305167, {'classification_error_evaluator': 0.875} Pass 0, Batch 200, Cost 2.305494, {'classification_error_evaluator': 0.890625} Pass 0, Batch 300, Cost 2.297118, {'classification_error_evaluator': 0.859375} Pass 0, Batch 400, Cost 2.316209, {'classification_error_evaluator': 0.9140625} Test with Pass 0, Cost 2.302148, {'classification_error_evaluator': 0.8865000009536743}

Pass 1, Batch 0, Cost 2.303937, {'classification_error_evaluator': 0.890625} Pass 1, Batch 100, Cost 2.302149, {'classification_error_evaluator': 0.90625} Pass 1, Batch 200, Cost 2.302109, {'classification_error_evaluator': 0.890625} Pass 1, Batch 300, Cost 2.300121, {'classification_error_evaluator': 0.9140625} Pass 1, Batch 400, Cost 2.293390, {'classification_error_evaluator': 0.859375} Test with Pass 1, Cost 2.302162, {'classification_error_evaluator': 0.8971999883651733}

Pass 2, Batch 0, Cost 2.299892, {'classification_error_evaluator': 0.8984375} Pass 2, Batch 100, Cost 2.308000, {'classification_error_evaluator': 0.8984375} Pass 2, Batch 200, Cost 2.301318, {'classification_error_evaluator': 0.9140625} Pass 2, Batch 300, Cost 2.300943, {'classification_error_evaluator': 0.859375} Pass 2, Batch 400, Cost 2.313893, {'classification_error_evaluator': 0.953125} Test with Pass 2, Cost 2.302042, {'classification_error_evaluator': 0.8971999883651733}

Pass 3, Batch 0, Cost 2.294420, {'classification_error_evaluator': 0.84375} Pass 3, Batch 100, Cost 2.325497, {'classification_error_evaluator': 0.9296875} Pass 3, Batch 200, Cost 2.296998, {'classification_error_evaluator': 0.8671875} Pass 3, Batch 300, Cost 2.294434, {'classification_error_evaluator': 0.8515625} Pass 3, Batch 400, Cost 2.301574, {'classification_error_evaluator': 0.859375} Test with Pass 3, Cost 2.301726, {'classification_error_evaluator': 0.8971999883651733}

Pass 4, Batch 0, Cost 2.316931, {'classification_error_evaluator': 0.921875} Pass 4, Batch 100, Cost 2.304905, {'classification_error_evaluator': 0.921875} Pass 4, Batch 200, Cost 2.294854, {'classification_error_evaluator': 0.859375} Pass 4, Batch 300, Cost 2.303220, {'classification_error_evaluator': 0.8671875} Pass 4, Batch 400, Cost 2.307637, {'classification_error_evaluator': 0.890625} Test with Pass 4, Cost 2.302750, {'classification_error_evaluator': 0.8971999883651733}

Best pass is 3, testing Avgcost is 2.30172612419 The classification accuracy is 10.28% Label of image/infer_3.png is: 7 root@cb462c7b76b1:/opt/book/02.recognize_digits#