Created: 2021-12-08 18:37:30

Modified: 2021-12-08 18:37:30

Google Colab is a free platform to use GPUs and TPUs, different from Google Cloud.

It is based on driver.google to store jupyter notebook and your own data, though some public ram and rom delivered by colab.

use with vscode

with colab_ssh, Cloudflare and Remote-SSH

refer to Colab + Vs Code + GitHub + Jupyter (Perfect for Deep Learning) and How to Connect to VSCode to Colab.

1. Push Colab to Cloudflare, then there will be tips for following steps in your screen.

   !pip install colab_ssh --upgrade from colab_ssh import launch_ssh_cloudflared, init_git_cloudflared launch_ssh_cloudflared(password="yourpassword")

   

2. Following the tips in the screen, download binary file of Cloudflare as /to/path/cloudflared-linux-amd64, and

   chmod +x /to/path/cloudflared-linux-amd64

3. Remote Configuration modifying ~/.ssh/config.

   # Read more about SSH config files: https://linux.die.net/man/5/ssh_config Host *.trycloudflare.com HostName %h User root Port 22 ProxyCommand /to/path/cloudflared-linux-amd64 access ssh --hostname %h

4. Test the connection in local terminal.

   ssh channels-enhancing-aw-specifies.trycloudflare.com nvidia-smi exit

5. Connect with VSCode Remote SSH. PressCtrl+Shift+P and type “Connect to Host…” and then paste the following hostname in the opened command palette.

   channels-enhancing-aw-specifies.trycloudflare.com

6. Test in the remote terminal.

   nvidia-smi

   runtime

   restart

   import os os.kill(os.getpid(), 9)