dartwin/tensorflow repository overview

It is forked version of official tensorflow/tensorflow image. You can use vscode remote-development extension with this image.

⁠Usage

⁠1. Run Docker image.

$ docker run -it -p 8888:22 dartwin/tensorflow:tag

You can mount the volume of local machine with -v tag.

$ docker run -it -p 8888:22 -v ~/"your-dir":/root dartwin/tensorflow:tag

If you use gpus, you can use --runtime option

$ docker run --runtime nvidia -it -p 8888:22 dartwin/tensorflow:tag

After running this command you can set root password of container.

⁠2. Then open vscode in local machine and add host to SSH config file.

Host "host-name"
    HostName "ip"
    User root
    Port 8888

⁠3. Connect to Host.

⁠4. Set the working directory of jupyter notebook embedded in the vscode.

Not "fileDirName"