Follow up of #529
Finally, it would be good to document the memory requirements for running the API to the Markdown documentation. In particular, the following estimates would be useful:
- Memory usage when idle and classification models not loaded?
- Memory usage when idle and classification models are loaded?
- How much does the memory usage increase when processing large documents?
Follow up of #529
Finally, it would be good to document the memory requirements for running the API to the Markdown documentation. In particular, the following estimates would be useful: