Optimisation in deployment #5520
-
|
I'm seeking optimization strategies for a Flask-based face recognition app using YOLO and DeepFace libraries. The app works fine on localhost but consistently hits memory limits on free-tier hosting services. What are the best practices to reduce memory usage in Flask apps with ML models? Looking for specific optimizations for resource-constrained platforms. |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments
-
|
Not using free hosting for ML and other compute-heavy deployment would probably be a good idea. Also, this sounds pretty much like a case on how to optimize your ML usage, and not how to optimize Flask (IMHO those two things should even run separately and not in the same process) |
Beta Was this translation helpful? Give feedback.
-
|
Consider using Streamlit instead of Flask. It offers better memory allocation for ML projects like your face recognition app. This can help address your deployment memory issues. For implementation details, check this |
Beta Was this translation helpful? Give feedback.
Consider using Streamlit instead of Flask. It offers better memory allocation for ML projects like your face recognition app. This can help address your deployment memory issues. For implementation details, check this