keras实验1Keras框架搭建实验介绍

53 阅读 0 评论 35 点赞

我是靠谱客的博主大气小蝴蝶，最近开发中收集的这篇文章主要介绍keras实验1Keras框架搭建实验介绍，觉得挺不错的，现在分享给大家，希望可以做个参考。

概述

实验介绍

本文参考keras中文官方文档操作。传送门

Keras框架搭建

Keras中mnist数据集测试

3行代码这么方便

>>> git clone https://github.com/fchollet/keras.git
>>> cd keras/examples/
>>> python mnist_mlp.py

执行日志

root@astron:/asdata# git clone https://github.com/fchollet/keras.git
Cloning into 'keras'...
remote: Counting objects: 23274, done.
remote: Compressing objects: 100% (7/7), done.
remote: Total 23274 (delta 1), reused 0 (delta 0), pack-reused 23267
Receiving objects: 100% (23274/23274), 8.84 MiB | 1.12 MiB/s, done.
Resolving deltas: 100% (16738/16738), done.
Checking connectivity... done.
root@astron:/asdata# cd keras/
root@astron:/asdata/keras# ls
CONTRIBUTING.md
docs
ISSUE_TEMPLATE.md
LICENSE
pytest.ini
setup.cfg
tests
docker
examples
keras
MANIFEST.in
README.md
setup.py
root@astron:/asdata/keras# cd examples/
root@astron:/asdata/keras/examples# ls
addition_rnn.py
imdb_bidirectional_lstm.py
mnist_hierarchical_rnn.py
neural_doodle.py
antirectifier.py
imdb_cnn_lstm.py
mnist_irnn.py
neural_style_transfer.py
babi_memnn.py
imdb_cnn.py
mnist_mlp.py
pretrained_word_embeddings.py
babi_rnn.py
imdb_fasttext.py
mnist_net2net.py
README.md
cifar10_cnn.py
imdb_lstm.py
mnist_siamese_graph.py
reuters_mlp.py
conv_filter_visualization.py
lstm_benchmark.py
mnist_sklearn_wrapper.py
reuters_mlp_relu_vs_selu.py
conv_lstm.py
lstm_text_generation.py
mnist_swwae.py
stateful_lstm.py
deep_dream.py
mnist_acgan.py
mnist_tfrecord.py
variational_autoencoder_deconv.py
image_ocr.py
mnist_cnn.py
mnist_transfer_cnn.py
variational_autoencoder.py
root@astron:/asdata/keras/examples# python mnist_mlp.py
Using TensorFlow backend.
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcublas.so.8.0 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcudnn.so.5 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcufft.so.8.0 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcurand.so.8.0 locally
60000 train samples
10000 test samples
_________________________________________________________________
Layer (type)
Output Shape
Param #
=================================================================
dense_1 (Dense)
(None, 512)
401920
_________________________________________________________________
dropout_1 (Dropout)
(None, 512)
0
_________________________________________________________________
dense_2 (Dense)
(None, 512)
262656
_________________________________________________________________
dropout_2 (Dropout)
(None, 512)
0
_________________________________________________________________
dense_3 (Dense)
(None, 10)
5130
=================================================================
Total params: 669,706.0
Trainable params: 669,706.0
Non-trainable params: 0.0
_________________________________________________________________
Train on 60000 samples, validate on 10000 samples
Epoch 1/20
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.
I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:910] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
I tensorflow/core/common_runtime/gpu/gpu_device.cc:885] Found device 0 with properties:
name: Tesla M60
major: 5 minor: 2 memoryClockRate (GHz) 1.1775
pciBusID 0000:00:15.0
Total memory: 7.93GiB
Free memory: 7.86GiB
I tensorflow/core/common_runtime/gpu/gpu_device.cc:906] DMA: 0
I tensorflow/core/common_runtime/gpu/gpu_device.cc:916] 0:
Y
I tensorflow/core/common_runtime/gpu/gpu_device.cc:975] Creating TensorFlow device (/gpu:0) -> (device: 0, name: Tesla M60, pci bus id: 0000:00:15.0)
60000/60000 [==============================] - 3s - loss: 0.2442 - acc: 0.9246 - val_loss: 0.0994 - val_acc: 0.9692
Epoch 2/20
60000/60000 [==============================] - 1s - loss: 0.1041 - acc: 0.9684 - val_loss: 0.0818 - val_acc: 0.9751
Epoch 3/20
60000/60000 [==============================] - 1s - loss: 0.0751 - acc: 0.9766 - val_loss: 0.0821 - val_acc: 0.9762
Epoch 4/20
60000/60000 [==============================] - 1s - loss: 0.0596 - acc: 0.9821 - val_loss: 0.0688 - val_acc: 0.9809
Epoch 5/20
60000/60000 [==============================] - 1s - loss: 0.0501 - acc: 0.9845 - val_loss: 0.0789 - val_acc: 0.9801
Epoch 6/20
60000/60000 [==============================] - 1s - loss: 0.0429 - acc: 0.9874 - val_loss: 0.0918 - val_acc: 0.9796
Epoch 7/20
60000/60000 [==============================] - 1s - loss: 0.0367 - acc: 0.9889 - val_loss: 0.0879 - val_acc: 0.9803
Epoch 8/20
60000/60000 [==============================] - 1s - loss: 0.0336 - acc: 0.9900 - val_loss: 0.0799 - val_acc: 0.9828
Epoch 9/20
60000/60000 [==============================] - 1s - loss: 0.0324 - acc: 0.9907 - val_loss: 0.0896 - val_acc: 0.9826
Epoch 10/20
60000/60000 [==============================] - 1s - loss: 0.0285 - acc: 0.9913 - val_loss: 0.0860 - val_acc: 0.9829
Epoch 11/20
60000/60000 [==============================] - 1s - loss: 0.0265 - acc: 0.9923 - val_loss: 0.0994 - val_acc: 0.9822
Epoch 12/20
60000/60000 [==============================] - 1s - loss: 0.0237 - acc: 0.9933 - val_loss: 0.1013 - val_acc: 0.9844
Epoch 13/20
60000/60000 [==============================] - 1s - loss: 0.0226 - acc: 0.9933 - val_loss: 0.1026 - val_acc: 0.9818
Epoch 14/20
60000/60000 [==============================] - 1s - loss: 0.0229 - acc: 0.9938 - val_loss: 0.1056 - val_acc: 0.9830
Epoch 15/20
60000/60000 [==============================] - 1s - loss: 0.0219 - acc: 0.9942 - val_loss: 0.0991 - val_acc: 0.9825
Epoch 16/20
60000/60000 [==============================] - 1s - loss: 0.0210 - acc: 0.9943 - val_loss: 0.1119 - val_acc: 0.9827
Epoch 17/20
60000/60000 [==============================] - 1s - loss: 0.0206 - acc: 0.9948 - val_loss: 0.1041 - val_acc: 0.9837
Epoch 18/20
60000/60000 [==============================] - 1s - loss: 0.0206 - acc: 0.9947 - val_loss: 0.1147 - val_acc: 0.9836
Epoch 19/20
60000/60000 [==============================] - 1s - loss: 0.0179 - acc: 0.9953 - val_loss: 0.1231 - val_acc: 0.9807
Epoch 20/20
60000/60000 [==============================] - 1s - loss: 0.0174 - acc: 0.9955 - val_loss: 0.1126 - val_acc: 0.9823
Test loss: 0.112580380508
Test accuracy: 0.9823