I have the same issue. I'm hoping for a solution.

Hey, I solved the problem by removing the useless dependencies and reducing the total amount of them. Also, the modules have to be compatible with each other (versions).

Thanks. My solution came late last night. I built my initial requirements.txt with an older version of pip and had no deployment issues. I had updated pip without thinking and the newest version of pip adds paths to requirements.txt file now.

example:
binaryornot==0.4.4
bleach @ file:///opt/conda/conda-bld/bleach_1641577558959/work
boto3==1.34.7

The @ and path was crashing the build. I also ran into issues with dependency versions conflicting. I was too timid to start guessing which dependencies were useless in the file. I ended up reverting to my initial requirements text version and adding the new dependencies by hand. I'm probably going to switch to poetry given that my solution won't work long term. Is there a way you removed the unneeded dependencies via pip? or was it just a matter of editing the requirements.txt file by hand?

Thanks. My solution came late last night. I built my initial requirements.txt with an older version of pip and had no deployment issues. I had updated pip without thinking and the newest version of pip adds paths to requirements.txt file now.

example:

binaryornot==0.4.4

bleach @ file:///opt/conda/conda-bld/bleach_1641577558959/work

boto3==1.34.7

The @ and path was crashing the build. I also ran into issues with dependency versions conflicting. I was too timid to start guessing which dependencies were useless in the file. I ended up reverting to my initial requirements text version and adding the new dependencies by hand. I'm probably going to switch to poetry given that my solution won't work long term. Is there a way you removed the unneeded dependencies via pip? or was it just a matter of editing the requirements.txt file by hand?