I have a FastAPI-based project where I train classification models using Scikit-learn. My code was working fine for months, but suddenly, without any changes, it started throwing the error: "model {name} not recognized" The issue seems to be that model names are not being recognized. However, when my teammates run the same code in their environments, it works fine. This led me to suspect a version mismatch with one of the key libraries: FastAPI, Scikit-learn, or regex. To troubleshoot: I installed the exact versions of these libraries that my teammates are using. I also tried upgrading/downgrading different versions, but the error persists. Questions: a) What could be causing this issue if the versions match but the error persists? b) How can I systematically identify the right version of the libraries that would make my code work again? c) Are there other factors (e.g., environment, dependencies, serialization issues) that could lead to this problem? Any insights or debugging strategies would be greatly appreciated! My Python version - 3.11.7 Continue reading...
