mirror of
https://github.com/likelovewant/ollama-for-amd.git
synced 2025-12-23 15:08:27 +00:00
- default to embeddings enabled - move embedding logic for loaded model to request - allow embedding full directory - close llm on reload