felladrin / MiniSearch

Self-hosted web-searching platform with an AI assistant that runs directly from your browser. Uses Web-LLM, Wllama and SearXNG. Demo: https://felladrin-minisearch.hf.space
https://felladrin-minisearch.hf.space
Apache License 2.0
127 stars 20 forks source link

Error since the update- Failed to generate response Could not generate response. It's possible that your browser or your system is out of memory. #654

Closed comeback01 closed 1 day ago

comeback01 commented 3 days ago

Bug description

Since the update, I have this error for Internal API Failed to generate response:

Failed to generate response
Could not generate response. It's possible that your browser or your system is out of memory.

Steps to reproduce

To be sure, I deleted the container:

docker stop xxxx
docker rm xxxx
docker system prune -a

I deleted the directory:

rm -rf MiniSearch

Then I cloned the directory again:

git clone https://github.com/felladrin/MiniSearch

I created the .env file

Then I ran this command:

docker compose -f docker-compose.production.yml up --build

.env:


ACCESS_KEYS="XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"
INTERNAL_OPENAI_COMPATIBLE_API_BASE_URL="https://openrouter.ai/api/v1"
INTERNAL_OPENAI_COMPATIBLE_API_KEY="XXXXXX"
INTERNAL_OPENAI_COMPATIBLE_API_MODEL="meta-llama/llama-3.2-11b-vision-instruct:free"
INTERNAL_OPENAI_COMPATIBLE_API_NAME="Internal API"
DEFAULT_INFERENCE_TYPE="internal"

image

image

Expected behavior

No response

Device info

No response

Additional context

No response

felladrin commented 3 days ago

Hi, @comeback01.

I couldn't reproduce it.

Here's the .env I used to test it:

Screenshot 2024-10-14 at 13 42 28

And it responded fine using meta-llama/llama-3.2-11b-vision-instruct:free from OpenRouter:

image

Could you please let me know if, in the server terminal, you see any message starting with Error in internal API endpoint:?

felladrin commented 1 day ago

@comeback01, I'm closing this issue for now, but I can re-open it if you're still facing it, ok?