I wanted to deploy a Tensorflow model which uses mixed precision feature. Wanted to ask does streamlit (or streamlit-cloud) support mixed precision
Mixed precision really isn’t a function of Streamlit or Streamlit Cloud, but rather the underlying Python libraries you are using. If it’s already possible in Python, then it’s possible within Streamlit, but it’s important to recognize that Streamlit is only defining what a user interface looks like. Model training is done by individual Python libraries irrespective of Streamlit.