Docker Images¶
We provide docker utility scripts to help developers to setup development environment. They are also helpful run through TVM demo and tutorials. We need docker and nvidia-docker if we want to use cuda.
Get a tvm source distribution or clone the github repo to get the auxiliary scripts
git clone --recursive https://github.com/apache/tvm tvm
We can then use the following command to launch a docker image.
/path/to/tvm/docker/bash.sh <image-name>
Here the image-name can be a local docker image name, e.g. tvm.ci_cpu
after you have done the local build.
This auxiliary script does the following things:
Mount current directory to
/workspace
Switch user to be the same user that calls the
bash.sh
(so you can read/write host system)Use the host-side network on Linux. Use the bridge network and expose port 8888 on macOS, because host networking driver isn’t supported. (so you can use
jupyter notebook
)
Then you can start a Jupyter notebook by typing
jupyter notebook
You might see an error OSError: [Errno 99] Cannot assign requested address
when starting
a Jupyter notebook on macOS. You can change the binding IP address by
jupyter notebook --ip=0.0.0.0
Note that on macOS, because bash.sh
uses the Docker bridge network, Jupyter will be reportedly running
at an URL like http://{container_hostname}:8888/?token=...
. You should replace the container_hostname
with localhost
when pasting it into browser.
Docker Source¶
Check out the docker source if you are interested in building your own docker images.
Run the following command to build the docker image.
/path/to/tvm/docker/build.sh <image-name>
You can find some un-official third party pre-built images at https://hub.docker.com/r/tlcpack/. These images are used for test purposes and are NOT of the ASF release.