Pinned Loading

1. Tencent/WeDLM Tencent/WeDLM Public

   WeDLM: The fastest diffusion language model with standard causal attention and native KV cache compatibility, delivering real speedups over vLLM-optimized baselines.

   Python 639 43

2. THU-BPM/MarkLLM THU-BPM/MarkLLM Public

   [EMNLP 2024 Demo] MarkLLM: An Open-Source Toolkit for LLM Watermarking

   Python 936 83

3. THU-BPM/chatgpt-sql THU-BPM/chatgpt-sql Public

   The prediction results of ChatGPT on various datasets of Text-to-SQL.

   Python 105 11

4. THU-BPM/Robust_Watermark THU-BPM/Robust_Watermark Public

   [ICLR 2024] Code and data for paper "A Semantic Invariant Robust Watermark for Large Language Models".

   Python 36 8

5. THU-BPM/unforgeable_watermark THU-BPM/unforgeable_watermark Public

   [ICLR 2024] Source code of paper "An Unforgeable Publicly Verifiable Watermark for Large Language Models"

   Python 34 2

6. THU-BPM/Watermarked_LLM_Identification THU-BPM/Watermarked_LLM_Identification Public

   Code and data for paper "Can Watermarked LLMs be Identified by Users via Crafted Prompts?" Accepted by ICLR 2025 (Spotlight)

   Python 28 5