Skip to main content

Week 2

ToDo​

  1. See the Action Items for CCTV project here.
  2. See the Action Items for DeepVibe project here.

October 13, 2023​

  • CCTV GIS+Prediction: Porting the latest code to our git repo and testing - ongoing
    • First nb complete 2 and 3 are in progress
  • Discussing with Srujana on how to run batch prediction for set of images

October 12, 2023​

  • CCTV GIS+Prediction: Porting the latest code to our git repo and testing - ongoing
  • Integrating wandb with scikit learn in the random tree training

October 11, 2023​

  • Had a call with Srujana on the issues of CCTV-Apps and pip acting weird on MicroCenter.

    • CCTV-Apps issue was actually due to bad csv files used for the input. Updating the csv files fixes it.
    • MicroCenter pip issue: The cause was the same issue we found on gqc-utility-notebooks package trial. conda activate calls virtual environments on top of each other as it seems. And that causes issues for pip in resolving the correct python installation to use and it reverts back to global installation which is problematic.
    • Introduce a new alias, coda for conda deactivate, which we should be using to explicitly exit from an environment before switching to another.

  • CCTV-Apps documentation update

    • Docs should be refreshed to reflect the latest status of the apps
    • Draw a mermaid diagram for the dataflow for each app.

  • Documented Vannary's reply on GIS+Prediction project

October 10, 2023​

  • Setting up remote login to microcenter and MSI machines with SSH keys. Assuming your local machine is running on Windows. Similar commands are available for linux.
    1. In your local machine (Windows), open a terminal window and run ssh-keygen -t ed25519.
      1. Give a unique name to the file. For example, I'm using C:/Users/pavan/.ssh/id_ed25519_microcenter as the name.
      2. This will generate a private key (with no extension) and a public key (with .pub extension)
      3. We use the private key at our end and put the public key in the remote in the next steps.
    2. SSH to the remote machine using the password for this time. I used VSCode itself to connect. Copy the content of the generated .pub file to ~/<user>/.ssh/authorized_keys file in the remote machine (linux).
    3. In VSCode, Open SSH configuration file by pressing ctrl+shift+p and typing Remote-SSH: Open SSH Configuration Fileconfig file
    4. If you already have a configuration for the same IP address, replace it with the following fields with the corresponding values. If not add these at the end of the file. ssh config
      • Host: Put a unique name for the machine. Doesn't affect the connection
      • HostName: IP address of the machine
      • User: Username. Typically gqc
      • PreferredAuthentications: Keep this as publickey
      • IdentityFile: Path to the private key file. Follow the exact same syntax as shown in the image. Note that C:/ is omitted from the path.

October 09, 2023​

  • Checking out the test failure issue in gqc-utility-notebooks and fixing the different issues it had.
  • Identifying the correct dependencies of the gqc-utility-notebooks.
    • Recognized the modules used by the python files by pipreqs modules.
    • Added the dependencies to requirements.txt
    • Updated the README
  • Resolving the issue with nbdev_preview not working outside MSI machine.
    • If nbdev python package is installed globally, it takes precedence.
    • documented the details in the readme
Last updated on by Pavan