How to use CMB-AI-LAB/lagkv_cache with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("CMB-AI-LAB/lagkv_cache", dtype="auto")