Before Running
Before running the caffe container, use docker pull command or click the Pull Button to ensure the desired image is installed. Once the pull is complete, you can go to next step.
| Tag | Pull command | Description |
| cpu | docker pull bvlc/caffe:cpu | CPU version |
| gpu | docker pull bvlc/caffe:gpu | GPU support
(CUDA 8.0+ required) |
| intel | docker pull bvlc/caffe:intel | Optimized for Intel CPU |
Run Container with Options
- Select GPU. To run the container with GPU support, you must specify the GPU.
NVIDIA GPU
GPU=nvidia0 gpu-docker run -it --rm bvlc/caffe:gpu bash
Intel CPU (launch Intel CPU optimized version caffe)
docker run -it --rm bvlc/caffe:intel bash
CPU
docker run -it --rm bvlc/caffe:cpu bash
- Shared folder. For leverage the NAS’s advantage, you can use shared folder to store the training data set, or the data to be inference.
Mount data folder in the volume 1 to the data folder in the container
GPU=nvidia0 gpu-docker run -it -v /share/CACHEDEV1_DATA/data:/data --rm bvlc/caffe:gpu bash
Mount videos folder in the volume 2 to the videos folder in the container
GPU=nvidia0 gpu-docker run -it -v /share/CACHEDEV2_DATA/videos:/videos --rm bvlc/caffe:gpu bash
Comments
Post a Comment