I have gotten this up and running. It doesn't seem to return high quality results. I have a powerful machine and a 30,000 document corpus to play with. What are the current best practices for models to use, transformers, etc to get the most out of this?
I have gotten this up and running. It doesn't seem to return high quality results. I have a powerful machine and a 30,000 document corpus to play with. What are the current best practices for models to use, transformers, etc to get the most out of this?