Jan-scraper

🎉**jan-scraper for conversation**: Now jan-scraper is optimized also to use Jan as an interface to hold a conversation with several text-generation and text2text-generation HuggingFace models, in 89 different languages, with *your own* pdfs.

Added `conversator`, `anylang` and `models_source` modules to manage this new conversation function

Added `scraper.convert_stream_to_jsonl(stream)` and `scraper.mine_content_from_jsonl(jsonlfile)` to adjust to new Jan v0.4.5 API "stream" parameter. See the README for more information.

Adjustments of the package on new Jan release (v0.4.5)

Added the model `jan_scraper.formatter` with the `convert_code_to_curl_json` function to format code in a JSON-compatible string that will be piped in the API

Added the possibility to activate more than one model at once within `activate_jan_api` function: this may be incompatible with RAM/CPU resources and/or may result in the activation of only one model, which may also happen not to be the desired one: set the parameter `auto` to False and activate the API manually to avoid this inconvenient.
Check out the README and the user_case_API.py for more detailed explanations.

Introduced the possibility to scrape Jan also via API.
May be unstable, may not work for all the machines (tested on Windows 11).