does anyone know how to download or where to find this great ‘pool’ of models/ ? It is used in assignments and seems to be very useful for local future experimental implementations..
Hi!
The models we use are essentially the following three, all available on Hugging Face:
-
Embedding model: BAAI/bge-base-en-v1.5
-
Reranker model: BAAI/bge-reranker-base
-
LLMs:
The embedding and reranker models are typically used locally in our courses. However, the Llama models (8B and 70B) require significant GPU VRAM, which is why we use together.ai for them—this also reflects how LLMs are typically deployed in production environments, since running them locally is quite resource-intensive.
If you still have any questions or need clarification, just let me know.
1 Like