Hours debugging , sorted the conflicts, no errors but app won't deploy.

Community Cloud
,
1

The app works fine on my localhost before deploying. The machine model works perfectly and gives the necessary output, but I haven’t been able to deploy since last night

Streamlit

Everything is on the root folder, I know it’s not a good practice but I was trying to exhaust all of my options.

The app is streamlist_implement.py

ERROR:

Error installing requirements.

Requirements file

2

Hi @ryanjuguna2k2, welcome to the forum!

I tried to clone your repo and look at the errors, and saw that it had an issue with a particular version of numpy:

  × Failed to download and build `scikit-learn==1.4.2`

  ├─▶ Failed to resolve requirements from `build-system.requires`

  ├─▶ No solution found when resolving: `setuptools`, `wheel`,

  │   `cython>=3.0.8`, `numpy==2.0.0rc1`, `scipy>=1.6.0`

  ╰─▶ Because there is no version of numpy==2.0.0rc1 and you require

      numpy==2.0.0rc1, we can conclude that your requirements are

      unsatisfiable.

So, I did what I often try, and just removed all the specific versions, like this, genetic-makeup-healthrisk-predictor/requirements.txt at main · sfc-gh-zblackwood/genetic-makeup-healthrisk-predictor · GitHub, and then the app loaded. I don’t know if it’s actually working properly, but you might start by removing all the version numbers, and only add back in specific versions if you know that you need one of them.

1 Like
3

It sounds like you’ve already done a lot of work fixing the conflicts and making sure your app runs locally, so the deployment issue might be linked to the environment setup or dependencies. Double-check your requirements.txt file for any packages that might not be supported in the deployment environment.

In my work with https://www.aftermathinvestigations.com/, I’ve seen similar cases where a small dependency version mismatch caused major deployment failures. Cleaning up unnecessary packages and specifying exact versions often solves it. You might want to try that approach here as well.

4

Thanks for your input. I actually realized this. I tried doing so and nothing changed and then it clicked. “What if I changed the python version that better supports the dependencies?” So on my app I persisted on using python 3.11 and it fixed my problems.

1 Like
5

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.