zjunlp/KnowLM
KnowLM is an open-source framework for building, fine-tuning, and deploying knowledgeable large language models with knowledge augmentation and editing capabilities.
Velocity · 7d
+1.2
★ / day
Trend
→steady
star history
KnowLM provides a standard framework for LLM pre-training and fine-tuning, along with a model zoo featuring models like ZhiXi and OneKE. It includes modules for instruction processing, knowledge augmentation via RAG, hallucination detection, and knowledge editing. All model weights and datasets are hosted on HuggingFace.