matbee/kimodo-llm2vec-nf4
Adapter vs McGill-NLP/LLM2Vec-Meta-Llama-3-8B-Instruct-mntp-supervised. McGill-NLP/LLM2Vec-Meta-Llama-3-8B-Instruct-mntp-supervised is a LoRA adapter, not full weights: rank 16, on down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj in 32 layers (224 adapted matrices). It modifies matbee/kimodo-llm2vec-nf4 at load time.