Doc: add a bit more details to help run RAG_API locally using ollama

[mjh-sakh]

What features would you like to see added?

The proposal is to update the doc to include one more env variable OLLAMA_BASE_URL to help run rag_api locally.

More details

Current description provides almost all references to do RAG embeddings locally using ollama. The guide is focused on running RAG_API inside docker. It makes OLLAMA_BASE_URL env variable essential and may be easily missed. Providing it right away helps to save users time.

Which components are impacted by your request?

Other

Pictures

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[mjh-sakh]

PR is in place (not sure if it is picked up automatically)

[fuegovic]

I'll review your PR soon, thanks!

[fuegovic]

PR is in place (not sure if it is picked up automatically)

Merged, Thanks again!
/rag_api#configuration

[hulkds]

If RAG API service cannot communicate with Ollama service and cause this error on Linux:

rag_api           | 2024-07-15 15:19:07,918 - root - ERROR - Error raised by inference endpoint: HTTPConnectionPool(host='host.docker.internal', port=11434): Max retries exceeded with url: /api/embeddings (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x70420aba7310>: Failed to establish a new connection: [Errno -2] Name or service not known'))
rag_api           | 2024-07-15 15:19:07,919 - root - INFO - Request POST http://rag_api:8000/embed - 200
LibreChat         | 2024-07-15 15:19:07 error: Error embedding file File embedding failed.
LibreChat         | 2024-07-15 15:19:07 error: [/files] Error processing file: File embedding failed.

Consider adding extra_hosts in the rag_api service in docker-compose.override.yml file:

extra_hosts:
  - "host.docker.internal:host-gateway"

PR for the librechat.ai docs: LibreChat-AI/librechat.ai#92