无法在nvidia docker容器中使用GPU运行代码。(nightly-gpu-py3)
我在自学什么是docker以及如何使用它。我对docker很陌生,所以希望在这里学习一些基础知识无法在nvidia docker容器中使用GPU运行代码。(nightly-gpu-py3),docker,tensorflow,nvidia-docker,Docker,Tensorflow,Nvidia Docker,我在自学什么是docker以及如何使用它。我对docker很陌生,所以希望在这里学习一些基础知识 我在我的电脑上安装了nvidia docker(以下)和tensorflow/tensorflow:nightly-gpu-py3(,启动gpu(CUDA)容器) Docker:NVIDIA Docker 2.0.3,版本:17.12.1-ce 主机操作系统:Ubuntu 16.04桌面 主持人:amd64 我的问题 当(用纯cuda编写)或简单矩阵乘法(用python用tensorflow编写
我在我的电脑上安装了nvidia docker(以下)和tensorflow/tensorflow:nightly-gpu-py3(,启动gpu(CUDA)容器)
- Docker:NVIDIA Docker 2.0.3,版本:17.12.1-ce
- 主机操作系统:Ubuntu 16.04桌面
- 主持人:amd64
仅使用CPU(如)的代码在tensorflow中工作良好 某些使用GPU的代码返回错误消息。(当代码使用
时)
细节
在容器(tensorflow/tensorflow:nightly-gpu-py3)中,当我运行monte-carlo模拟时,我得到以下错误:
fatal error: curand.h: No such file or directory
locate curand.h
不返回任何内容,但当我尝试locate curand
时,我得到:
/usr/local/cuda-9.0/targets/x86_64-linux/lib/libcurand.so.9.0
/usr/local/cuda-9.0/targets/x86_64-linux/lib/libcurand.so.9.0.176
/usr/share/doc/cuda-curand-9-0
/usr/share/doc/cuda-curand-9-0/changelog.Debian.gz
/usr/share/doc/cuda-curand-9-0/copyright
/var/lib/dpkg/info/cuda-curand-9-0.list
/var/lib/dpkg/info/cuda-curand-9-0.md5sums
/var/lib/dpkg/info/cuda-curand-9-0.postinst
/var/lib/dpkg/info/cuda-curand-9-0.postrm
/var/lib/dpkg/info/cuda-curand-9-0.shlibs
对于查找cudnn.h
:
/usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/core/util/use_cudnn.h
/usr/include/linux/cuda.h
/usr/local/cuda-9.0/targets/x86_64-linux/include/cuda.h
/usr/local/cuda-9.0/targets/x86_64-linux/include/dynlink_cuda.h
/usr/local/cuda-9.0/targets/x86_64-linux/include/dynlink_cuda_cuda.h
/usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/core/platform/cuda.h
/usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/core/platform/stream_executor_no_cuda.h
对于查找cuda.h
:
/usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/core/util/use_cudnn.h
/usr/include/linux/cuda.h
/usr/local/cuda-9.0/targets/x86_64-linux/include/cuda.h
/usr/local/cuda-9.0/targets/x86_64-linux/include/dynlink_cuda.h
/usr/local/cuda-9.0/targets/x86_64-linux/include/dynlink_cuda_cuda.h
/usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/core/platform/cuda.h
/usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/core/platform/stream_executor_no_cuda.h
nvcc--version
返回:
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Sep__1_21:08:03_CDT_2017
Cuda compilation tools, release 9.0, V9.0.176
在主机中(容器外部),当我尝试运行nvidia/cuda nvidia smi时,我得到
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.30 Driver Version: 390.30 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 108... Off | 00000000:03:00.0 On | N/A |
| 0% 48C P8 22W / 250W | 301MiB / 11177MiB | 1% Default |
+-------------------------------+----------------------+----------------------+
| 1 GeForce GTX 108... Off | 00000000:81:00.0 Off | N/A |
| 0% 51C P8 22W / 250W | 2MiB / 11178MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
+-----------------------------------------------------------------------------+
我所做的
#include
-->失败#include
-->失败nvidia docker run
运行容器。docker映像不是意味着它可以保存依赖项(路径、包等)以运行特定代码(在我的例子中,是使用
的代码)吗?(集装箱是否有实际工作?)
是否意味着我未正确安装/加载nvidia docker/nightly-gpu-py3
sudo nvidia docker run-it--rm-p 8888:8888-p 6006:6006/bin/bash
是我用给定图像启动新容器的命令。这可能是个问题吗来自“Devel-docker映像包括从源代码构建所需的所有依赖项,而其他二进制文件只是安装了TensorFlow。”您使用的是非Devel映像。如果您希望完整的CUDA工具包可用,请尝试使用devel映像。在你的情况下,这将是
nightly-devel-gpu-py3
而不是nightly-gpu-py3
我这样做:nvidia docker run-it--rm-p 8888:8888 tensorflow/tensorflow:nightly-devel-gpu-py3/bin/bash
,然后查看该容器中的/usr/local/cuda/include
,而且它确实有curand.h
,看起来是一个完整的CUDA工具包安装。