附表2
TensorFlow 安装包下载
| 深度学习框架 | Python 版本 | TensorFlow 版本 | 下载地址 |
|---|---|---|---|
CUDA 9.1 |
2.7 |
1.6.0 |
|
CUDA 9.1 |
3.6 |
1.6.0 |
|
CUDA 9.1 |
2.7 |
1.7.0 |
|
CUDA 9.1 |
3.6 |
1.7.0 |
|
CUDA 9.1 |
2.7 |
1.8.0 |
|
CUDA 9.1 |
3.6 |
1.8.0 |
|
CUDA 9.1 |
2.7 |
1.9.0 |
|
CUDA 9.1 |
3.6 |
1.9.0 |
|
CUDA 9.1 |
2.7 |
1.10.0 |
|
CUDA 9.1 |
3.6 |
1.10.0 |
|
CUDA 9.1 |
2.7 |
1.11.0 |
|
CUDA 9.1 |
3.6 |
1.11.0 |
|
CUDA 9.1 |
2.7 |
1.12.0 |
|
CUDA 9.1 |
3.6 |
1.12.0 |
|
CUDA 8.0 |
2.7 |
1.6.0 |
|
CUDA 8.0 |
3.6 |
1.6.0 |
|
CPU |
2.7 |
1.6.0 |
|
CPU |
3.6 |
1.6.0 |
|
CPU |
2.7 |
1.7.0 |
|
CPU |
3.6 |
1.7.0 |
|
CPU |
2.7 |
1.8.0 |
|
CPU |
3.6 |
1.8.0 |
|
CPU |
2.7 |
1.9.0 |
|
CPU |
3.6 |
1.9.0 |
|
CPU |
2.7 |
1.10.0 |
|
CPU |
3.6 |
1.10.0 |
|
CPU |
2.7 |
1.11.0 |
|
CPU |
3.6 |
1.11.0 |
|
CPU |
2.7 |
1.12.0 |
|
CPU |
3.6 |
1.12.0 |
docker 镜像获取命令
| 框架版本 | Python 版本 | CUDA 版本 | image 获取命令 |
|---|---|---|---|
deeplearning1.0 |
2.7 |
9.1 |
docker pull qingcloud/deeplearning:1.0-py27-cu91-cudnn7.1 |
deeplearning1.0 |
2.7 |
- |
docker pull qingcloud/deeplearning:1.0-py27-cpu |
deeplearning1.0 |
2.7 |
- |
docker pull qingcloud/deeplearning:1.0-py27-cpu-avx |
deeplearning1.0 |
2.7 |
- |
docker pull qingcloud/deeplearning:1.0-py27-cpu-avx2-mkldnn |
caffe1.0 |
3.6 |
- |
docker pull qingcloud/caffe:1.0-bvlc-py36-cpu |
caffe1.0 |
2.7 |
- |
docker pull qingcloud/caffe:1.0-bvlc-py27-cpu |
caffe1.0 |
3.6 |
9.1 |
docker pull qingcloud/caffe:1.0-bvlc-py36-cu91-cudnn7.1 |
caffe1.0 |
3.6 |
8.0 |
docker pull /caffe:1.0-bvlc-py36-cu80-cudnn7.1 |
caffe1.0 |
2.7 |
9.1 |
docker pull /caffe:1.0-bvlc-py27-cu91-cudnn7.1 |
caffe1.0 |
2.7 |
8.0 |
docker pull /caffe:1.0-bvlc-py27-cu80-cudnn7.1 |
pytorch0.4.0 |
3.6 |
9.1 |
docker pull /pytorch:0.4.0-py36-cpu-cu91-cudnn7.1 |
pytorch0.4.1 |
3.6 |
8.0 |
docker pull /pytorch:0.4.1-py36-cpu-cu80-cudnn7.1 |
pytorch0.4.1 |
2.7 |
9.1 |
docker pull /pytorch:0.4.1-py27-cpu-cu91-cudnn7.1 |
pytorch0.4.1 |
2.7 |
8.0 |
docker pull /pytorch:0.4.1-py27-cpu-cu80-cudnn7.1 |
keras2.2 |
3.6 |
9.1 |
docker pull /keras:2.2-tf1.8-py36-cu91-cudnn7.1 |
keras2.2 |
2.7 |
9.1 |
docker pull qingcloud/keras:2.2-tf1.8-py27-cu91-cudnn7.1 |
keras2.2 |
3.6 |
- |
docker pull qingcloud/keras:2.2-tf1.8-py36-cpu |
keras2.2 |
2.7 |
- |
docker pull qingcloud/keras:2.2-tf1.8-py27-cpu |
tensorflow1.9 |
3.6 |
9.1 |
docker pull qingcloud/tensorflow:1.9-py36-cu91-cudnn7.1 |
tensorflow1.9 |
2.7 |
9.1 |
docker pull qingcloud/tensorflow:1.9-py27-cu91-cudnn7.1 |
tensorflow1.9 |
3.6 |
- |
docker pull qingcloud/tensorflow:1.9-py36-cpu |
tensorflow1.9 |
2.7 |
- |
docker pull qingcloud/tensorflow:1.9-py27-cpu |
tensorflow1.8 |
3.6 |
9.1 |
docker pull qingcloud/tensorflow:1.8-py36-cu91-cudnn7.1 |
tensorflow1.8 |
2.7 |
9.1 |
docker pull qingcloud/tensorflow:1.8-py27-cu91-cudnn7.1 |
tensorflow1.8 |
3.6 |
- |
docker pull qingcloud/tensorflow:1.8-py36-cpu |
tensorflow1.8 |
2.7 |
- |
docker pull qingcloud/tensorflow:1.8-py27-cpu |
tensorflow1.7 |
3.6 |
9.1 |
docker pull qingcloud/tensorflow:1.7-py36-cu91-cudnn7.1 |
tensorflow1.7 |
2.7 |
9.1 |
docker pull qingcloud/tensorflow:1.7-py27-cu91-cudnn7.1 |
tensorflow1.7 |
3.6 |
- |
docker pull qingcloud/tensorflow:1.7-py36-cpu |
tensorflow1.7 |
2.7 |
- |
docker pull qingcloud/tensorflow:1.7-py27-cpu |
tensorflow1.6 |
3.6 |
9.1 |
docker pull qingcloud/tensorflow:1.6-py36-cu91-cudnn7.1 |
tensorflow1.6 |
3.6 |
8.0 |
docker pull qingcloud/tensorflow:1.6-py36-cu80-cudnn7.1 |
tensorflow1.6 |
2.7 |
9.1 |
docker pull qingcloud/tensorflow:1.6-py27-cu91-cudnn7.1 |
tensorflow1.6 |
2.7 |
8.0 |
docker pull qingcloud/tensorflow:1.6-py27-cu80-cudnn7.1 |
tensorflow1.6 |
3.6 |
- |
docker pull qingcloud/tensorflow:1.6-py36-cpu |
tensorflow1.6 |
2.7 |
- |
docker pull qingcloud/tensorflow:1.6-py27-cpu |
CPU 容器版镜像与 CPU 微架构适配表
| Intel CPU 微架构 | AVX2 和 MKLDNN 优化版 | AVX 优化版 | 无优化版 |
|---|---|---|---|
平台默认 |
不支持 |
不支持 |
支持 |
Westmere |
不支持 |
不支持 |
支持 |
SandyBridge |
不支持 |
支持 |
支持 |
IvyBridge |
不支持 |
支持 |
支持 |
Haswell |
支持 |
支持 |
支持 |
Broadwell |
支持 |
支持 |
支持 |