Open INF800 opened 1 year ago
Note that colab uses Python 3.10.12
Trying to manually comment out the packages and see which is causing the issue.
Hey @mmirman @INF800, would like to work on this could you assign me this issue.
Trying to manually comment out the packages and see which is causing the issue.
Found out the issue. It is due to flask package https://github.com/anarchy-ai/LLM-VM/blob/d188e720ce65dcfddf5178da0ba1dd14bf612bd4/pyproject.toml#L20
It is causing
ERROR: Cannot uninstall 'blinker'. It is a distutils installed project and thus we cannot accurately determine which files belong to it which would lead to only a partial uninstall.
Replacing it with "flask>=2.0"
does the job but we need to restart the session to be able to run import llm_vm
, need to figure out the reason behind it.
To be able to properly run LLM-VM on jupyter notebooks (also colab) it would be necessary to refactor the way the arguments are parsed.
I created a story for that #176
I wanted to get this done but wasnt able to finish. So if anyone wants to finish trying to make anarchy usable in google colab maybe this helps
https://colab.research.google.com/drive/1LGBspWILLrpedJ3GRBSnr4G1_pFV88om
@dafinley will take a shot at it
Trying to manually comment out the packages and see which is causing the issue.
Found out the issue. It is due to flask package
It is causing
ERROR: Cannot uninstall 'blinker'. It is a distutils installed project and thus we cannot accurately determine which files belong to it which would lead to only a partial uninstall.
Replacing it with
"flask>=2.0"
does the job but we need to restart the session to be able to runimport llm_vm
, need to figure out the reason behind it.
This was mentioned by @INF800 last month the task is to get it done without the need for restarting the session.
Ahh. At least we have a path forward now.
With the downgrade of flask + #176 it is working properly in colab, but the proper solution would be #177
To be able to properly run LLM-VM on jupyter notebooks (also colab) it would be necessary to refactor the way the arguments are parsed.
I created a story for that #176
There is a link to a colab notebook to test it
causes