adeebjamal/private-ai-backend

oops

renamed column

removed sources list from LLM response

increased n_ctx

minor change

Updated LLM to Gemma 4 E4B

minor change

minor change

minor change

enhanced web search

Optimised internet access logic

dependency update

minor change

minor fix

Build time reduction

Added log

Added web search feature

getMessages query now returns message ordered by created_at

Added async processing in /ask API

feat: add rename and delete conversation APIs

Add dummy JSON responses to requests.txt for frontend dev reference

Update requests.txt for public space access

Auto-clear model_cache when model changes to save disk space

Fix severe CPU thread thrashing by hardcoding n_threads to 2

Change default model to Llama-3.2-3B-Instruct for much faster CPU inference

Add detailed timing logs to /ask API

Fix HF Space timeout

Fix musl libc error and prevent OOM via CMAKE_BUILD_PARALLEL_LEVEL=1

Force prefer-binary for llama-cpp-python

Fix OOM by using pre-compiled llama-cpp-python

added requests.txt

Merge HF Space config

renamed README_GITHUB.md

initial commit

first commit