Javier Martinez
Javier Martinez
Could you share your stack trace? I just ran it using Docker+Ollama, and everything is working as expected (using macOS; I don't have access to a Windows machine).
Failure when starting the container https://github.com/abetlen/llama-cpp-python/issues/1169#issuecomment-2228511452
Can you explain which changes are there in this PR?
Are you trying to ingest local files using `ingest_folder`? If yes, can you try to run using mock profile `PGPT_PROFILES=mock make ingest`?
Could you give us more details about profile, environment, machine?
I will try to explain how it works by default. Once you enable `auth.enable=True`, the system validates if a basic authentication is included. By default the Authorization header is: `Authorization:...
I reopened the issue and convert it to improvement. Currently, we consider gradio as testing ui, not the final one. Please consider use other alternatives or define a custom version...
Can you check ollama server logs? The problem is ollama related to, since it's throwing 500. Something with context window, computer resources, etc
Have you tried latest changes? It's like an read/write problem in your host machine. We've updated our docker-compose compatibility to have different profiles.
Sorry for the delay :( I just pushed a generic fix in https://github.com/zylon-ai/private-gpt/pull/2059. If somebody can check if this issue is fixed, it would be nice! @salimfadhley @iguy0 @theodufort @vilaca