mahimairaja/tweet-summarization-llama-2-finetuned
Adapter vs meta-llama/Llama-2-7b-hf. mahimairaja/tweet-summarization-llama-2-finetuned is a LoRA adapter, not full weights: rank 16, on q_proj, v_proj in 32 layers (64 adapted matrices). It modifies meta-llama/Llama-2-7b-hf at load time.