OpenNLP Research Group

Advancing Open Research in Language Models and Dialogue Systems

Our Mission

We are an open research collective dedicated to advancing NLP technologies through transparent collaboration. Our focus areas include:


Recent Demo Pages

Demos for F5R-TTS

Demonstrate the F5R-TTS arch and results.

View

Open Projects

InternVL-Llama.cpp: High-Performance Adapter for InternVL

A specialized adaptation of llama.cpp for InternVL series multimodal models, enabling efficient CPU/GPU inference for vision-language tasks with optimized memory usage and performance.

View on GitHub →

Optimized Sign Language Processing Implementations (CSLR and Two-Stream)

Optimizing the Comprehensive CSLR Workflow Implementation, Traning Efficiency, Fixing Errors and Providing more dataset support for CSLR and Two-Stream methods.

View on GitHub →

Recent Publications

RAIDEN Benchmark: Evaluating Role-playing Conversational Agents with Measurement-Driven Custom Dialogues

Wu et al. | COLING 2025 | GitHub Logo HF Logo
More →

Huggingface Hub

HF Logo Our Data & Models Hub