9060人阅读
神经网络（28）
机器学习（10）
Python（11）
系统设计（17）
C/C++（26）
TensorFlow 支持以下几种安装方式：
源码编译安装
Docker 镜像安装
源码编译安装
优点：可以定制
缺点：耗时较长
$ git clone --recurse-submodules /tensorflow/tensorflow
安装 Bazel
$ ./configure
仅 CPU 支持，无 GPU 支持：
$ bazel build -c opt //tensorflow/tools/pip_package:build_pip_package
有 GPU 支持：
$ bazel build -c opt --config=cuda //tensorflow/tools/pip_package:build_pip_package
生成 pip 安装包
$ bazel-bin/tensorflow/tools/pip_package/build_pip_package /tmp/tensorflow_pkg
使用 PIP 工具安装
$ pip install /tmp/tensorflow_pkg/tensorflow-x.x.x-py2-none-linux_x86_64.whl
优点：安装快捷
缺点：不能灵活定制，不支持某些系统发行版
PIP 是一种包管理系统，用于安装和管理用 Python 写的软件包。
# Ubuntu/Linux 64-bit
$ sudo apt-get install python-pip python-dev
# CentOS, Fedora, RHEL
$ sudo yum install python-pip python-devel
# Mac OS X
$ sudo easy_install pip
注意 TensorFlow 从 0.8.0rc 开始支持多机多卡分布式计算，而更早的版本只支持单计算节点。
安装 TensorFlow
# Python 2
$ sudo pip install --upgrade $TF_BINARY_URL
# Python 3
$ sudo pip3 install --upgrade $TF_BINARY_URL
其中环境变量 TF_BINARY_URL 根据你的系统进行设置，典型选项如下：
# Ubuntu/Linux 64-bit, CPU only, Python 2.7
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.11.0rc0-cp27-none-linux_x86_64.whl
# Ubuntu/Linux 64-bit, GPU enabled, Python 2.7
# Requires CUDA toolkit 7.5 and CuDNN v5. For other versions, see "Install from sources" below.
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.11.0rc0-cp27-none-linux_x86_64.whl
# Mac OS X, CPU only, Python 2.7:
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-0.11.0rc0-py2-none-any.whl
# Mac OS X, GPU enabled, Python 2.7:
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/gpu/tensorflow-0.11.0rc0-py2-none-any.whl
# Ubuntu/Linux 64-bit, CPU only, Python 3.4
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.11.0rc0-cp34-cp34m-linux_x86_64.whl
# Ubuntu/Linux 64-bit, GPU enabled, Python 3.4
# Requires CUDA toolkit 7.5 and CuDNN v5. For other versions, see "Install from sources" below.
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.11.0rc0-cp34-cp34m-linux_x86_64.whl
# Ubuntu/Linux 64-bit, CPU only, Python 3.5
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.11.0rc0-cp35-cp35m-linux_x86_64.whl
# Ubuntu/Linux 64-bit, GPU enabled, Python 3.5
# Requires CUDA toolkit 7.5 and CuDNN v5. For other versions, see "Install from sources" below.
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.11.0rc0-cp35-cp35m-linux_x86_64.whl
# Mac OS X, CPU only, Python 3.4 or 3.5:
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-0.11.0rc0-py3-none-any.whl
# Mac OS X, GPU enabled, Python 3.4 or 3.5:
$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/gpu/tensorflow-0.11.0rc0-py3-none-any.whl
运行记录如下：
Processing tensorflow-0.11.0rc0-cp27-none-linux_x86_64.whl
Collecting six&=1.10.0 (from tensorflow==0.11.0rc0)
Downloading six-1.10.0-py2.py3-none-any.whl
Collecting wheel (from tensorflow==0.11.0rc0)
Downloading wheel-0.29.0-py2.py3-none-any.whl (66kB)
100% |████████████████████████████████| 71kB 23kB/s
Collecting mock&=2.0.0 (from tensorflow==0.11.0rc0)
Downloading mock-2.0.0-py2.py3-none-any.whl (56kB)
100% |████████████████████████████████| 61kB 10kB/s
Collecting numpy&=1.11.0 (from tensorflow==0.11.0rc0)
Downloading numpy-1.11.2-cp27-cp27mu-manylinux1_x86_64.whl (15.3MB)
100% |████████████████████████████████| 15.3MB 78kB/s
Collecting protobuf==3.0.0 (from tensorflow==0.11.0rc0)
Downloading protobuf-3.0.0-py2.py3-none-any.whl (342kB)
100% |████████████████████████████████| 348kB 1.4MB/s
Collecting funcsigs&=1; python_version & "3.3" (from mock&=2.0.0-&tensorflow==0.11.0rc0)
Downloading funcsigs-1.0.2-py2.py3-none-any.whl
Collecting pbr&=0.11 (from mock&=2.0.0-&tensorflow==0.11.0rc0)
Downloading pbr-1.10.0-py2.py3-none-any.whl (96kB)
100% |████████████████████████████████| 102kB 2.6MB/s
Collecting setuptools (from protobuf==3.0.0-&tensorflow==0.11.0rc0)
Downloading setuptools-28.3.0-py2.py3-none-any.whl (467kB)
100% |████████████████████████████████| 471kB 1.1MB/s
Installing collected packages: six, wheel, funcsigs, pbr, mock, numpy, setuptools, protobuf, tensorflow
Successfully installed funcsigs-1.0.2 mock-2.0.0 numpy-1.11.1 pbr-1.10.0 protobuf-3.0.0 setuptools-0.9.8 six-1.3.0 tensorflow-0.11.0rc0 wheel-0.29.0
Docker 镜像安装
优点：适合批量部署
缺点：有墙的孩子像根草
Docker 是一个开源的应用容器引擎，让开发者可以打包他们的应用以及依赖包到一个可移植的容器中，然后发布到任何流行的 Linux 机器上，也可以实现虚拟化。
当你通过 Docker 安装和运行 TensorFlow 时，它与你机器上之前已安装的软件包完全隔离。
官方提供了 4 个 Docker 镜像可供使用：
仅 CPU 版，无开发环境：gcr.io/tensorflow/tensorflow
仅 CPU 版，有开发环境：gcr.io/tensorflow/tensorflow:latest-devel
支持 GPU，无开发环境：gcr.io/tensorflow/tensorflow:latest-gpu
支持 GPU，有开发环境：gcr.io/tensorflow/tensorflow:latest-devel-gpu
另外提供了对应某个发布版本的镜像，只需将上面 tag 中 latest 替换为发布版本号。安装详细步骤如下：
创建 Docker 用户组，允许普通用户无需 sudo 即可启动容器。
$ usermod -a -G docker YOURNAME
选择上述 4 个镜像中的一个，创建容器。第一次执行该命令时会自动下载镜像，以后不需要再次下载。
$ docker run -it gcr.io/tensorflow/tensorflow
如果你使用了支持 GPU 的容器，在运行该命令时需要增加额外参数，目的是将宿主机上的 GPU 设备暴露给容器。使用 TensorFlow 源码中提供的脚本可以实现该功能
$ cd $TENSORFLOW_ROOT/tensorflow/tools/docker/
$ ./docker_run_gpu.sh gcr.io/tensorflow/tensorflow:gpu
好奇心驱使我们查看该脚本的具体细节：
#!/usr/bin/env bash
export CUDA_HOME=${CUDA_HOME:-/usr/local/cuda}
if [ ! -d ${CUDA_HOME}/lib64 ]; then
echo "Failed to locate CUDA libs at ${CUDA_HOME}/lib64."
export CUDA_SO=$(\ls /usr/lib/x86_64-linux-gnu/libcuda.* | \
xargs -I{} echo '-v {}:{}')
export DEVICES=$(\ls /dev/nvidia* | \
xargs -I{} echo '--device {}:{}')
if [[ "${DEVICES}" = "" ]]; then
echo "Failed to locate NVidia device(s). Did you want the non-GPU container?"
docker run -it $CUDA_SO $DEVICES "$@"
主要做了几件事：
暴露宿主机的 CUDA_HOME 环境变量给容器使用；
暴露宿主机的 libcuda.* 动态链接库给容器访问；
暴露宿主机的 /dev/nvidia* 设备给容器访问；
假设读者已经按照上述步骤安装了 GPU 版本 TensorFlow 0.11.0rc，接下来可以运行经典例程(MNIST）：
# python -m tensorflow.models.image.mnist.convolutional
K40 上运行输出
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcurand.so locally
Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.
Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.
Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.
Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.
Extracting data/train-images-idx3-ubyte.gz
Extracting data/train-labels-idx1-ubyte.gz
Extracting data/t10k-images-idx3-ubyte.gz
Extracting data/t10k-labels-idx1-ubyte.gz
I tensorflow/core/common_runtime/gpu/gpu_device.cc:951] Found device 0 with properties:
name: Tesla K40m
major: 3 minor: 5 memoryClockRate (GHz) 0.745
pciBusID 0000:02:00.0
Total memory: 11.25GiB
Free memory: 11.12GiB
W tensorflow/stream_executor/cuda/cuda_driver.cc:572] creating context when one is existing: 0x4c81230
I tensorflow/core/common_runtime/gpu/gpu_device.cc:951] Found device 1 with properties:
name: Tesla K40m
major: 3 minor: 5 memoryClockRate (GHz) 0.745
pciBusID 0000:03:00.0
Total memory: 11.25GiB
Free memory: 11.12GiB
I tensorflow/core/common_runtime/gpu/gpu_device.cc:972] DMA: 0 1
I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] 0:
I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] 1:
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1041] Creating TensorFlow device (/gpu:0) -& (device: 0, name: Tesla K40m, pci bus id: 0000:02:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1041] Creating TensorFlow device (/gpu:1) -& (device: 1, name: Tesla K40m, pci bus id: 0000:03:00.0)
Initialized!
Step 0 (epoch 0.00), 67.1 ms
Minibatch loss: 12.054, learning rate: 0.010000
Minibatch error: 90.6%
Validation error: 84.6%
Step 100 (epoch 0.12), 14.2 ms
Minibatch loss: 3.303, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 7.0%
Step 200 (epoch 0.23), 13.8 ms
Minibatch loss: 3.481, learning rate: 0.010000
Minibatch error: 12.5%
Validation error: 4.0%
Step 300 (epoch 0.35), 13.8 ms
Minibatch loss: 3.170, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 3.5%
Step 400 (epoch 0.47), 13.8 ms
Minibatch loss: 3.219, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 2.9%
Step 500 (epoch 0.58), 13.8 ms
Minibatch loss: 3.268, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 2.7%
Step 600 (epoch 0.70), 13.7 ms
Minibatch loss: 3.220, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 2.4%
Step 700 (epoch 0.81), 13.9 ms
Minibatch loss: 3.035, learning rate: 0.010000
Minibatch error: 1.6%
Validation error: 2.7%
Step 800 (epoch 0.93), 13.7 ms
Minibatch loss: 3.073, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 2.1%
Step 900 (epoch 1.05), 13.8 ms
Minibatch loss: 2.933, learning rate: 0.009500
Minibatch error: 3.1%
Validation error: 1.7%
Step 1000 (epoch 1.16), 13.9 ms
Minibatch loss: 2.874, learning rate: 0.009500
Minibatch error: 1.6%
Validation error: 1.6%
Step 1100 (epoch 1.28), 13.9 ms
Minibatch loss: 2.821, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.5%
Step 1200 (epoch 1.40), 13.9 ms
Minibatch loss: 2.923, learning rate: 0.009500
Minibatch error: 7.8%
Validation error: 1.6%
Step 1300 (epoch 1.51), 13.9 ms
Minibatch loss: 2.765, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.7%
Step 1400 (epoch 1.63), 14.1 ms
Minibatch loss: 2.789, learning rate: 0.009500
Minibatch error: 3.1%
Validation error: 1.5%
Step 1500 (epoch 1.75), 13.9 ms
Minibatch loss: 2.886, learning rate: 0.009500
Minibatch error: 6.2%
^^Validation error: 1.2%
Step 1600 (epoch 1.86), 13.9 ms
Minibatch loss: 2.691, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.3%
Step 1700 (epoch 1.98), 13.8 ms
Minibatch loss: 2.649, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.4%
Step 1800 (epoch 2.09), 13.9 ms
Minibatch loss: 2.667, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.4%
Step 1900 (epoch 2.21), 13.9 ms
Minibatch loss: 2.639, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.2%
Step 2000 (epoch 2.33), 13.9 ms
Minibatch loss: 2.636, learning rate: 0.009025
Minibatch error: 3.1%
Validation error: 1.2%
Step 2100 (epoch 2.44), 13.9 ms
Minibatch loss: 2.585, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.1%
Step 2200 (epoch 2.56), 13.9 ms
Minibatch loss: 2.573, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.2%
Step 2300 (epoch 2.68), 13.8 ms
Minibatch loss: 2.553, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.2%
Step 2400 (epoch 2.79), 13.9 ms
Minibatch loss: 2.499, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.2%
Step 2500 (epoch 2.91), 13.9 ms
Minibatch loss: 2.471, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.1%
Step 2600 (epoch 3.03), 13.9 ms
Minibatch loss: 2.451, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.3%
Step 2700 (epoch 3.14), 13.9 ms
Minibatch loss: 2.488, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.2%
Step 2800 (epoch 3.26), 13.9 ms
Minibatch loss: 2.433, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.1%
Step 2900 (epoch 3.37), 13.9 ms
Minibatch loss: 2.472, learning rate: 0.008574
Minibatch error: 3.1%
Validation error: 1.1%
Step 3000 (epoch 3.49), 13.9 ms
Minibatch loss: 2.405, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.2%
Step 3100 (epoch 3.61), 13.9 ms
Minibatch loss: 2.396, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.1%
Step 3200 (epoch 3.72), 13.9 ms
Minibatch loss: 2.332, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.1%
Step 3300 (epoch 3.84), 13.9 ms
Minibatch loss: 2.319, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.0%
Step 3400 (epoch 3.96), 13.9 ms
Minibatch loss: 2.297, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.1%
Step 3500 (epoch 4.07), 13.8 ms
Minibatch loss: 2.273, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.1%
Step 3600 (epoch 4.19), 13.9 ms
Minibatch loss: 2.261, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.0%
Step 3700 (epoch 4.31), 13.8 ms
Minibatch loss: 2.239, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.0%
Step 3800 (epoch 4.42), 13.9 ms
Minibatch loss: 2.220, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 3900 (epoch 4.54), 13.9 ms
Minibatch loss: 2.319, learning rate: 0.008145
Minibatch error: 3.1%
Validation error: 1.0%
Step 4000 (epoch 4.65), 13.9 ms
Minibatch loss: 2.213, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.1%
Step 4100 (epoch 4.77), 14.0 ms
Minibatch loss: 2.191, learning rate: 0.008145
Minibatch error: 1.6%
Validation error: 1.0%
Step 4200 (epoch 4.89), 13.9 ms
Minibatch loss: 2.231, learning rate: 0.008145
Minibatch error: 1.6%
Validation error: 1.0%
Step 4300 (epoch 5.00), 13.9 ms
Minibatch loss: 2.208, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 0.9%
Step 4400 (epoch 5.12), 13.8 ms
Minibatch loss: 2.167, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 1.0%
Step 4500 (epoch 5.24), 14.0 ms
Minibatch loss: 2.125, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.9%
Step 4600 (epoch 5.35), 13.9 ms
Minibatch loss: 2.096, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.9%
Step 4700 (epoch 5.47), 13.9 ms
Minibatch loss: 2.091, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 0.9%
Step 4800 (epoch 5.59), 13.8 ms
Minibatch loss: 2.053, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.9%
Step 4900 (epoch 5.70), 13.9 ms
Minibatch loss: 2.049, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 1.0%
Step 5000 (epoch 5.82), 13.9 ms
Minibatch loss: 2.170, learning rate: 0.007738
Minibatch error: 3.1%
Validation error: 1.0%
Step 5100 (epoch 5.93), 13.7 ms
Minibatch loss: 2.005, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.9%
Step 5200 (epoch 6.05), 13.8 ms
Minibatch loss: 2.074, learning rate: 0.007351
Minibatch error: 3.1%
Validation error: 1.0%
Step 5300 (epoch 6.17), 13.8 ms
Minibatch loss: 1.988, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 1.0%
Step 5400 (epoch 6.28), 13.8 ms
Minibatch loss: 1.959, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.8%
Step 5500 (epoch 6.40), 13.8 ms
Minibatch loss: 1.969, learning rate: 0.007351
Minibatch error: 3.1%
Validation error: 0.9%
Step 5600 (epoch 6.52), 13.9 ms
Minibatch loss: 1.931, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5700 (epoch 6.63), 13.8 ms
Minibatch loss: 1.926, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 1.0%
Step 5800 (epoch 6.75), 13.9 ms
Minibatch loss: 1.899, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
^^Step 5900 (epoch 6.87), 13.9 ms
Minibatch loss: 1.897, learning rate: 0.007351
Minibatch error: 1.6%
Validation error: 0.9%
Step 6000 (epoch 6.98), 13.8 ms
Minibatch loss: 1.877, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 6100 (epoch 7.10), 13.9 ms
Minibatch loss: 1.863, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6200 (epoch 7.21), 13.9 ms
Minibatch loss: 1.844, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6300 (epoch 7.33), 13.8 ms
Minibatch loss: 1.854, learning rate: 0.006983
Minibatch error: 1.6%
Validation error: 0.9%
Step 6400 (epoch 7.45), 13.9 ms
Minibatch loss: 1.880, learning rate: 0.006983
Minibatch error: 3.1%
Validation error: 0.8%
Step 6500 (epoch 7.56), 13.9 ms
Minibatch loss: 1.809, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6600 (epoch 7.68), 13.8 ms
Minibatch loss: 1.824, learning rate: 0.006983
Minibatch error: 1.6%
Validation error: 1.0%
Step 6700 (epoch 7.80), 13.9 ms
Minibatch loss: 1.787, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6800 (epoch 7.91), 13.8 ms
Minibatch loss: 1.771, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 1.0%
Step 6900 (epoch 8.03), 13.8 ms
Minibatch loss: 1.758, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7000 (epoch 8.15), 13.8 ms
Minibatch loss: 1.784, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.9%
Step 7100 (epoch 8.26), 13.9 ms
Minibatch loss: 1.757, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.9%
Step 7200 (epoch 8.38), 13.8 ms
Minibatch loss: 1.756, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.8%
Step 7300 (epoch 8.49), 13.8 ms
Minibatch loss: 1.721, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.7%
Step 7400 (epoch 8.61), 13.8 ms
Minibatch loss: 1.702, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.8%
Step 7500 (epoch 8.73), 13.8 ms
Minibatch loss: 1.696, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.8%
Step 7600 (epoch 8.84), 13.8 ms
Minibatch loss: 1.798, learning rate: 0.006634
Minibatch error: 3.1%
Validation error: 0.8%
Step 7700 (epoch 8.96), 14.0 ms
Minibatch loss: 1.667, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7800 (epoch 9.08), 13.9 ms
Minibatch loss: 1.658, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 7900 (epoch 9.19), 13.9 ms
Minibatch loss: 1.650, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 1.0%
Step 8000 (epoch 9.31), 13.9 ms
Minibatch loss: 1.654, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 8100 (epoch 9.43), 13.9 ms
Minibatch loss: 1.630, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 8200 (epoch 9.54), 14.0 ms
Minibatch loss: 1.630, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 8300 (epoch 9.66), 13.9 ms
Minibatch loss: 1.608, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 8400 (epoch 9.77), 13.9 ms
Minibatch loss: 1.596, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8500 (epoch 9.89), 13.9 ms
Minibatch loss: 1.618, learning rate: 0.006302
Minibatch error: 1.6%
Validation error: 0.8%
Test error: 0.8%
M40 上运行输出
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcudnn.so.5 locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcurand.so locally
Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.
Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.
Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.
Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.
Extracting data/train-images-idx3-ubyte.gz
Extracting data/train-labels-idx1-ubyte.gz
Extracting data/t10k-images-idx3-ubyte.gz
Extracting data/t10k-labels-idx1-ubyte.gz
I tensorflow/core/common_runtime/gpu/gpu_device.cc:951] Found device 0 with properties:
name: Tesla M40
major: 5 minor: 2 memoryClockRate (GHz) 1.112
pciBusID 0000:06:00.0
Total memory: 11.18GiB
Free memory: 11.07GiB
W tensorflow/stream_executor/cuda/cuda_driver.cc:572] creating context when one is existing: 0x37b6440
I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] 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:951] Found device 1 with properties:
name: Tesla M40
major: 5 minor: 2 memoryClockRate (GHz) 1.112
pciBusID 0000:87:00.0
Total memory: 11.18GiB
Free memory: 11.07GiB
I tensorflow/core/common_runtime/gpu/gpu_device.cc:855] cannot enable peer access from device ordinal 0 to device ordinal 1
I tensorflow/core/common_runtime/gpu/gpu_device.cc:855] cannot enable peer access from device ordinal 1 to device ordinal 0
I tensorflow/core/common_runtime/gpu/gpu_device.cc:972] DMA: 0 1
I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] 0:
I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] 1:
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1041] Creating TensorFlow device (/gpu:0) -& (device: 0, name: Tesla M40, pci bus id: 0000:06:00.0)
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1041] Creating TensorFlow device (/gpu:1) -& (device: 1, name: Tesla M40, pci bus id: 0000:87:00.0)
Initialized!
Step 0 (epoch 0.00), 23.6 ms
Minibatch loss: 12.054, learning rate: 0.010000
Minibatch error: 90.6%
Validation error: 84.6%
Step 100 (epoch 0.12), 6.7 ms
Minibatch loss: 3.297, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 7.3%
Step 200 (epoch 0.23), 6.5 ms
Minibatch loss: 3.448, learning rate: 0.010000
Minibatch error: 10.9%
Validation error: 3.8%
Step 300 (epoch 0.35), 6.4 ms
Minibatch loss: 3.175, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 3.2%
Step 400 (epoch 0.47), 6.4 ms
Minibatch loss: 3.215, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 2.6%
Step 500 (epoch 0.58), 6.4 ms
Minibatch loss: 3.278, learning rate: 0.010000
Minibatch error: 9.4%
Validation error: 2.5%
Step 600 (epoch 0.70), 6.4 ms
Minibatch loss: 3.202, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 2.7%
Step 700 (epoch 0.81), 6.4 ms
Minibatch loss: 3.017, learning rate: 0.010000
Minibatch error: 3.1%
Validation error: 2.6%
Step 800 (epoch 0.93), 6.4 ms
Minibatch loss: 3.090, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 2.2%
Step 900 (epoch 1.05), 6.4 ms
Minibatch loss: 2.926, learning rate: 0.009500
Minibatch error: 3.1%
Validation error: 1.6%
Step 1000 (epoch 1.16), 6.5 ms
Minibatch loss: 2.856, learning rate: 0.009500
Minibatch error: 1.6%
Validation error: 1.6%
Step 1100 (epoch 1.28), 6.4 ms
Minibatch loss: 2.820, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.5%
Step 1200 (epoch 1.40), 6.4 ms
Minibatch loss: 2.956, learning rate: 0.009500
Minibatch error: 6.2%
Validation error: 1.4%
Step 1300 (epoch 1.51), 6.4 ms
Minibatch loss: 2.767, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.5%
Step 1400 (epoch 1.63), 6.4 ms
Minibatch loss: 2.780, learning rate: 0.009500
Minibatch error: 3.1%
Validation error: 1.5%
Step 1500 (epoch 1.75), 6.4 ms
Minibatch loss: 2.880, learning rate: 0.009500
Minibatch error: 6.2%
Validation error: 1.2%
Step 1600 (epoch 1.86), 6.4 ms
Minibatch loss: 2.718, learning rate: 0.009500
Minibatch error: 1.6%
Validation error: 1.3%
Step 1700 (epoch 1.98), 6.4 ms
Minibatch loss: 2.650, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.5%
Step 1800 (epoch 2.09), 6.6 ms
Minibatch loss: 2.657, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.4%
Step 1900 (epoch 2.21), 6.4 ms
Minibatch loss: 2.671, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.2%
Step 2000 (epoch 2.33), 6.4 ms
Minibatch loss: 2.642, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.3%
Step 2100 (epoch 2.44), 6.4 ms
Minibatch loss: 2.580, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.1%
Step 2200 (epoch 2.56), 6.4 ms
Minibatch loss: 2.564, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.2%
Step 2300 (epoch 2.68), 6.4 ms
Minibatch loss: 2.578, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.2%
Step 2400 (epoch 2.79), 6.4 ms
Minibatch loss: 2.498, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.1%
Step 2500 (epoch 2.91), 6.4 ms
Minibatch loss: 2.476, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.2%
Step 2600 (epoch 3.03), 6.5 ms
Minibatch loss: 2.453, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.1%
Step 2700 (epoch 3.14), 6.4 ms
Minibatch loss: 2.483, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.0%
Step 2800 (epoch 3.26), 6.4 ms
Minibatch loss: 2.424, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.2%
Step 2900 (epoch 3.37), 6.4 ms
Minibatch loss: 2.453, learning rate: 0.008574
Minibatch error: 4.7%
Validation error: 1.1%
Step 3000 (epoch 3.49), 6.4 ms
Minibatch loss: 2.405, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.1%
Step 3100 (epoch 3.61), 6.4 ms
Minibatch loss: 2.400, learning rate: 0.008574
Minibatch error: 3.1%
Validation error: 1.1%
Step 3200 (epoch 3.72), 6.4 ms
Minibatch loss: 2.342, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.2%
Step 3300 (epoch 3.84), 6.4 ms
Minibatch loss: 2.320, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.2%
Step 3400 (epoch 3.96), 6.4 ms
Minibatch loss: 2.295, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.1%
Step 3500 (epoch 4.07), 6.4 ms
Minibatch loss: 2.272, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.1%
Step 3600 (epoch 4.19), 6.4 ms
Minibatch loss: 2.256, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 3700 (epoch 4.31), 6.4 ms
Minibatch loss: 2.235, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 3800 (epoch 4.42), 6.4 ms
Minibatch loss: 2.220, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 3900 (epoch 4.54), 6.4 ms
Minibatch loss: 2.323, learning rate: 0.008145
Minibatch error: 3.1%
Validation error: 1.0%
Step 4000 (epoch 4.65), 6.4 ms
Minibatch loss: 2.205, learning rate: 0.008145
Minibatch error: 1.6%
Validation error: 1.1%
Step 4100 (epoch 4.77), 6.4 ms
Minibatch loss: 2.171, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 4200 (epoch 4.89), 6.4 ms
Minibatch loss: 2.231, learning rate: 0.008145
Minibatch error: 1.6%
Validation error: 1.0%
Step 4300 (epoch 5.00), 6.4 ms
Minibatch loss: 2.209, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 0.9%
Step 4400 (epoch 5.12), 6.4 ms
Minibatch loss: 2.142, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 1.0%
Step 4500 (epoch 5.24), 6.4 ms
Minibatch loss: 2.167, learning rate: 0.007738
Minibatch error: 6.2%
Validation error: 0.9%
Step 4600 (epoch 5.35), 6.4 ms
Minibatch loss: 2.126, learning rate: 0.007738
Minibatch error: 3.1%
Validation error: 0.8%
Step 4700 (epoch 5.47), 6.4 ms
Minibatch loss: 2.074, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 0.9%
Step 4800 (epoch 5.59), 6.4 ms
Minibatch loss: 2.052, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 1.0%
Step 4900 (epoch 5.70), 6.4 ms
Minibatch loss: 2.073, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 0.9%
Step 5000 (epoch 5.82), 6.4 ms
Minibatch loss: 2.129, learning rate: 0.007738
Minibatch error: 3.1%
Validation error: 1.0%
Step 5100 (epoch 5.93), 6.4 ms
Minibatch loss: 2.003, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.9%
Step 5200 (epoch 6.05), 6.4 ms
Minibatch loss: 2.076, learning rate: 0.007351
Minibatch error: 4.7%
Validation error: 1.0%
Step 5300 (epoch 6.17), 6.4 ms
Minibatch loss: 1.995, learning rate: 0.007351
Minibatch error: 1.6%
Validation error: 0.9%
Step 5400 (epoch 6.28), 6.4 ms
Minibatch loss: 1.960, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5500 (epoch 6.40), 6.4 ms
Minibatch loss: 1.965, learning rate: 0.007351
Minibatch error: 1.6%
Validation error: 0.8%
Step 5600 (epoch 6.52), 6.4 ms
Minibatch loss: 1.927, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.8%
Step 5700 (epoch 6.63), 6.4 ms
Minibatch loss: 1.912, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5800 (epoch 6.75), 6.4 ms
Minibatch loss: 1.899, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5900 (epoch 6.87), 6.4 ms
Minibatch loss: 1.886, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.8%
Step 6000 (epoch 6.98), 6.4 ms
Minibatch loss: 1.887, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 6100 (epoch 7.10), 6.4 ms
Minibatch loss: 1.860, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6200 (epoch 7.21), 6.4 ms
Minibatch loss: 1.847, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6300 (epoch 7.33), 6.4 ms
Minibatch loss: 1.840, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6400 (epoch 7.45), 6.4 ms
Minibatch loss: 1.904, learning rate: 0.006983
Minibatch error: 3.1%
Validation error: 0.9%
Step 6500 (epoch 7.56), 6.4 ms
Minibatch loss: 1.811, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6600 (epoch 7.68), 6.4 ms
Minibatch loss: 1.825, learning rate: 0.006983
Minibatch error: 1.6%
Validation error: 0.9%
Step 6700 (epoch 7.80), 6.4 ms
Minibatch loss: 1.784, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6800 (epoch 7.91), 6.4 ms
Minibatch loss: 1.773, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6900 (epoch 8.03), 6.4 ms
Minibatch loss: 1.757, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 1.0%
Step 7000 (epoch 8.15), 6.4 ms
Minibatch loss: 1.774, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.8%
Step 7100 (epoch 8.26), 6.4 ms
Minibatch loss: 1.741, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.8%
Step 7200 (epoch 8.38), 6.4 ms
Minibatch loss: 1.736, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7300 (epoch 8.49), 6.4 ms
Minibatch loss: 1.728, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.8%
Step 7400 (epoch 8.61), 6.4 ms
Minibatch loss: 1.700, learning rate: 0.006634
Minibatch error: 0.0%
^^Validation error: 0.8%
Step 7500 (epoch 8.73), 6.4 ms
Minibatch loss: 1.706, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.8%
Step 7600 (epoch 8.84), 6.4 ms
Minibatch loss: 1.793, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.8%
Step 7700 (epoch 8.96), 6.4 ms
Minibatch loss: 1.666, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7800 (epoch 9.08), 6.4 ms
Minibatch loss: 1.658, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 7900 (epoch 9.19), 6.4 ms
Minibatch loss: 1.652, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 8000 (epoch 9.31), 6.4 ms
Minibatch loss: 1.670, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8100 (epoch 9.43), 6.4 ms
Minibatch loss: 1.626, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8200 (epoch 9.54), 6.4 ms
Minibatch loss: 1.627, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8300 (epoch 9.66), 6.4 ms
Minibatch loss: 1.610, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8400 (epoch 9.77), 6.4 ms
Minibatch loss: 1.596, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8500 (epoch 9.89), 6.4 ms
Minibatch loss: 1.605, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Test error: 0.8%
GTX1080 上运行输出
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcurand.so locally
Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.
Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.
Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.
Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.
Extracting data/train-images-idx3-ubyte.gz
Extracting data/train-labels-idx1-ubyte.gz
Extracting data/t10k-images-idx3-ubyte.gz
Extracting data/t10k-labels-idx1-ubyte.gz
I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] 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_init.cc:118] Found device 0 with properties:
name: GeForce GTX 1080
major: 6 minor: 1 memoryClockRate (GHz) 1.8475
pciBusID 0000:01:00.0
Total memory: 7.92GiB
Free memory: 7.48GiB
I tensorflow/core/common_runtime/gpu/gpu_init.cc:138] DMA: 0
I tensorflow/core/common_runtime/gpu/gpu_init.cc:148] 0:
I tensorflow/core/common_runtime/gpu/gpu_device.cc:868] Creating TensorFlow device (/gpu:0) -& (device: 0, name: GeForce GTX 1080, pci bus id: 0000:01:00.0)
Initialized!
Step 0 (epoch 0.00), 8.2 ms
Minibatch loss: 12.054, learning rate: 0.010000
Minibatch error: 90.6%
Validation error: 84.6%
Step 100 (epoch 0.12), 5.1 ms
Minibatch loss: 3.289, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 7.0%
Step 200 (epoch 0.23), 5.1 ms
Minibatch loss: 3.476, learning rate: 0.010000
Minibatch error: 12.5%
Validation error: 3.7%
Step 300 (epoch 0.35), 4.9 ms
Minibatch loss: 3.182, learning rate: 0.010000
Minibatch error: 7.8%
Validation error: 3.2%
Step 400 (epoch 0.47), 5.0 ms
Minibatch loss: 3.241, learning rate: 0.010000
Minibatch error: 9.4%
Validation error: 2.7%
Step 500 (epoch 0.58), 5.1 ms
Minibatch loss: 3.295, learning rate: 0.010000
Minibatch error: 9.4%
Validation error: 2.7%
Step 600 (epoch 0.70), 4.9 ms
Minibatch loss: 3.206, learning rate: 0.010000
Minibatch error: 6.2%
Validation error: 2.8%
Step 700 (epoch 0.81), 5.0 ms
Minibatch loss: 2.999, learning rate: 0.010000
Minibatch error: 3.1%
Validation error: 2.5%
Step 800 (epoch 0.93), 5.0 ms
Minibatch loss: 3.090, learning rate: 0.010000
Minibatch error: 4.7%
Validation error: 2.3%
Step 900 (epoch 1.05), 5.0 ms
Minibatch loss: 2.919, learning rate: 0.009500
Minibatch error: 1.6%
Validation error: 1.7%
Step 1000 (epoch 1.16), 5.0 ms
Minibatch loss: 2.845, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.7%
Step 1100 (epoch 1.28), 4.8 ms
Minibatch loss: 2.824, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.5%
Step 1200 (epoch 1.40), 4.9 ms
Minibatch loss: 2.964, learning rate: 0.009500
Minibatch error: 6.2%
Validation error: 1.4%
Step 1300 (epoch 1.51), 5.0 ms
Minibatch loss: 2.761, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.8%
Step 1400 (epoch 1.63), 5.0 ms
Minibatch loss: 2.779, learning rate: 0.009500
Minibatch error: 4.7%
Validation error: 1.5%
Step 1500 (epoch 1.75), 5.0 ms
Minibatch loss: 2.840, learning rate: 0.009500
Minibatch error: 4.7%
Validation error: 1.2%
Step 1600 (epoch 1.86), 4.9 ms
Minibatch loss: 2.724, learning rate: 0.009500
Minibatch error: 1.6%
Validation error: 1.3%
Step 1700 (epoch 1.98), 5.0 ms
Minibatch loss: 2.650, learning rate: 0.009500
Minibatch error: 0.0%
Validation error: 1.3%
Step 1800 (epoch 2.09), 5.0 ms
Minibatch loss: 2.658, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.4%
Step 1900 (epoch 2.21), 5.0 ms
Minibatch loss: 2.653, learning rate: 0.009025
Minibatch error: 1.6%
Validation error: 1.2%
Step 2000 (epoch 2.33), 4.9 ms
Minibatch loss: 2.649, learning rate: 0.009025
Minibatch error: 3.1%
Validation error: 1.2%
Step 2100 (epoch 2.44), 5.0 ms
Minibatch loss: 2.573, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.0%
Step 2200 (epoch 2.56), 4.9 ms
Minibatch loss: 2.579, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.2%
Step 2300 (epoch 2.68), 5.2 ms
Minibatch loss: 2.566, learning rate: 0.009025
Minibatch error: 3.1%
Validation error: 1.2%
Step 2400 (epoch 2.79), 5.0 ms
Minibatch loss: 2.498, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.2%
Step 2500 (epoch 2.91), 5.1 ms
Minibatch loss: 2.476, learning rate: 0.009025
Minibatch error: 0.0%
Validation error: 1.1%
Step 2600 (epoch 3.03), 4.9 ms
Minibatch loss: 2.454, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.1%
Step 2700 (epoch 3.14), 4.9 ms
Minibatch loss: 2.480, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.0%
Step 2800 (epoch 3.26), 5.2 ms
Minibatch loss: 2.414, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.1%
Step 2900 (epoch 3.37), 5.1 ms
Minibatch loss: 2.455, learning rate: 0.008574
Minibatch error: 3.1%
Validation error: 1.1%
Step 3000 (epoch 3.49), 5.1 ms
Minibatch loss: 2.398, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.1%
Step 3100 (epoch 3.61), 4.9 ms
Minibatch loss: 2.400, learning rate: 0.008574
Minibatch error: 4.7%
Validation error: 1.0%
Step 3200 (epoch 3.72), 5.0 ms
Minibatch loss: 2.340, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.1%
Step 3300 (epoch 3.84), 5.1 ms
Minibatch loss: 2.321, learning rate: 0.008574
Minibatch error: 1.6%
Validation error: 1.2%
Step 3400 (epoch 3.96), 5.0 ms
Minibatch loss: 2.295, learning rate: 0.008574
Minibatch error: 0.0%
Validation error: 1.0%
Step 3500 (epoch 4.07), 5.0 ms
Minibatch loss: 2.274, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.0%
Step 3600 (epoch 4.19), 5.1 ms
Minibatch loss: 2.256, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 3700 (epoch 4.31), 4.9 ms
Minibatch loss: 2.243, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 1.0%
Step 3800 (epoch 4.42), 5.1 ms
Minibatch loss: 2.252, learning rate: 0.008145
Minibatch error: 1.6%
Validation error: 0.9%
Step 3900 (epoch 4.54), 4.9 ms
Minibatch loss: 2.299, learning rate: 0.008145
Minibatch error: 3.1%
Validation error: 1.1%
Step 4000 (epoch 4.65), 5.0 ms
Minibatch loss: 2.236, learning rate: 0.008145
Minibatch error: 3.1%
Validation error: 1.0%
Step 4100 (epoch 4.77), 5.0 ms
Minibatch loss: 2.174, learning rate: 0.008145
Minibatch error: 0.0%
Validation error: 0.9%
Step 4200 (epoch 4.89), 5.2 ms
Minibatch loss: 2.209, learning rate: 0.008145
Minibatch error: 1.6%
Validation error: 1.0%
Step 4300 (epoch 5.00), 5.1 ms
Minibatch loss: 2.187, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 1.0%
Step 4400 (epoch 5.12), 5.0 ms
Minibatch loss: 2.157, learning rate: 0.007738
Minibatch error: 3.1%
Validation error: 1.0%
Step 4500 (epoch 5.24), 5.2 ms
Minibatch loss: 2.162, learning rate: 0.007738
Minibatch error: 4.7%
Validation error: 0.8%
Step 4600 (epoch 5.35), 5.1 ms
Minibatch loss: 2.098, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.8%
Step 4700 (epoch 5.47), 5.0 ms
Minibatch loss: 2.079, learning rate: 0.007738
Minibatch error: 1.6%
Validation error: 0.9%
Step 4800 (epoch 5.59), 5.0 ms
Minibatch loss: 2.057, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 1.0%
Step 4900 (epoch 5.70), 4.9 ms
Minibatch loss: 2.046, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 0.9%
Step 5000 (epoch 5.82), 4.9 ms
Minibatch loss: 2.104, learning rate: 0.007738
Minibatch error: 3.1%
Validation error: 1.0%
Step 5100 (epoch 5.93), 5.1 ms
Minibatch loss: 2.002, learning rate: 0.007738
Minibatch error: 0.0%
Validation error: 1.0%
Step 5200 (epoch 6.05), 5.1 ms
Minibatch loss: 2.062, learning rate: 0.007351
Minibatch error: 3.1%
Validation error: 0.9%
Step 5300 (epoch 6.17), 5.1 ms
Minibatch loss: 1.980, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 1.0%
Step 5400 (epoch 6.28), 5.0 ms
Minibatch loss: 1.957, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5500 (epoch 6.40), 5.1 ms
Minibatch loss: 1.953, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5600 (epoch 6.52), 5.0 ms
Minibatch loss: 1.929, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 5700 (epoch 6.63), 5.1 ms
Minibatch loss: 1.913, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 1.0%
Step 5800 (epoch 6.75), 5.0 ms
Minibatch loss: 1.897, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.8%
Step 5900 (epoch 6.87), 5.1 ms
Minibatch loss: 1.890, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.8%
Step 6000 (epoch 6.98), 5.0 ms
Minibatch loss: 1.892, learning rate: 0.007351
Minibatch error: 0.0%
Validation error: 0.9%
Step 6100 (epoch 7.10), 5.2 ms
Minibatch loss: 1.859, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6200 (epoch 7.21), 5.0 ms
Minibatch loss: 1.843, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6300 (epoch 7.33), 5.0 ms
Minibatch loss: 1.843, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6400 (epoch 7.45), 5.0 ms
Minibatch loss: 1.872, learning rate: 0.006983
Minibatch error: 1.6%
Validation error: 0.9%
Step 6500 (epoch 7.56), 5.1 ms
Minibatch loss: 1.808, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6600 (epoch 7.68), 4.9 ms
Minibatch loss: 1.838, learning rate: 0.006983
Minibatch error: 1.6%
Validation error: 1.0%
Step 6700 (epoch 7.80), 5.1 ms
Minibatch loss: 1.782, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.8%
Step 6800 (epoch 7.91), 5.0 ms
Minibatch loss: 1.774, learning rate: 0.006983
Minibatch error: 0.0%
Validation error: 0.9%
Step 6900 (epoch 8.03), 4.9 ms
Minibatch loss: 1.760, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7000 (epoch 8.15), 5.0 ms
Minibatch loss: 1.783, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.9%
Step 7100 (epoch 8.26), 5.1 ms
Minibatch loss: 1.736, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7200 (epoch 8.38), 4.9 ms
Minibatch loss: 1.747, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.9%
Step 7300 (epoch 8.49), 5.0 ms
Minibatch loss: 1.722, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.7%
Step 7400 (epoch 8.61), 5.0 ms
Minibatch loss: 1.700, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.8%
Step 7500 (epoch 8.73), 5.1 ms
Minibatch loss: 1.698, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 0.9%
Step 7600 (epoch 8.84), 5.1 ms
Minibatch loss: 1.803, learning rate: 0.006634
Minibatch error: 1.6%
Validation error: 0.8%
Step 7700 (epoch 8.96), 5.0 ms
Minibatch loss: 1.668, learning rate: 0.006634
Minibatch error: 0.0%
Validation error: 1.0%
Step 7800 (epoch 9.08), 5.0 ms
Minibatch loss: 1.660, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 7900 (epoch 9.19), 5.1 ms
Minibatch loss: 1.657, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 1.0%
Step 8000 (epoch 9.31), 5.0 ms
Minibatch loss: 1.666, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8100 (epoch 9.43), 4.9 ms
Minibatch loss: 1.625, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8200 (epoch 9.54), 5.0 ms
Minibatch loss: 1.627, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.9%
Step 8300 (epoch 9.66), 5.1 ms
Minibatch loss: 1.607, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8400 (epoch 9.77), 4.9 ms
Minibatch loss: 1.595, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 0.8%
Step 8500 (epoch 9.89), 4.9 ms
Minibatch loss: 1.612, learning rate: 0.006302
Minibatch error: 0.0%
Validation error: 1.0%
Test error: 0.8%
可以看出计算速度上 GTX1080 & M40 & K40。
(1) 如果需要 GPU，那么首先安装 CUDA 和 cuDNN。
(2) GTX1080 需要安装 CUDA 8.0，只支持源码安装。
(3) TensorFlow 0.11.0rc 支持 cuDNN v5。
参考知识库
* 以上用户言论只代表其个人观点，不代表CSDN网站的观点或立场
访问：361721次
积分：4845
积分：4845
排名：第4499名
原创：80篇
译文：18篇
评论：379条
文章：12篇
阅读：64435
文章：17篇
阅读：80514
文章：11篇
阅读：60698
阅读：56638
(1)(2)(1)(1)(1)(2)(6)(3)(1)(1)(1)(2)(4)(7)(1)(1)(9)(19)(1)(1)(4)(1)(7)(19)(1)(1)(1)(1)(1)(1)