InternVL-Chat

This folder contains the implementation of the training code for InternVL3_5-GPT-OSS-20B-A4B. We provide a conda environment package along with corresponding training scripts. The efficient implementation of GptOssAttention is provided in this folder, with credit to Wenhao Li. We also provide examples scrits about how to finetune Qwen3-based InternVL3.5. Please refer to the model card for more details.

📖 Documents

🌟 Get Started

Installation: 🌱 conda environment archive | 📄 requirements.txt
Tutorials: 🚀 Enhancing InternVL on COCO Caption Using LoRA Fine-Tuning

🏆 InternVL Family

InternVL 3.5: 📖 Introduction | ⚡ Quick Start | ✨ Finetune | 📊 Evaluation | 📦 Deployment | 🎯 Preference Optimization
InternVL 3.0: 📖 Introduction | ⚡ Quick Start | ✨ Finetune | 📊 Evaluation | 📦 Deployment | 🎯 Preference Optimization
InternVL 2.5: 📖 Introduction | ⚡ Quick Start | ✨ Finetune | 📊 Evaluation | 📦 Deployment | 🎯 Preference Optimization
InternVL 2.0: 📖 Introduction | ⚡ Quick Start | ✨ Finetune | 📊 Evaluation | 📦 Deployment | 🎯 Preference Optimization
InternVL 1.5: 📖 Introduction | ⚡ Quick Start | ✨ Finetune | 📊 Evaluation | 📦 Deployment
InternVL 1.2: 📖 Introduction | ⚡ Quick Start | ✨ Finetune | 📊 Evaluation
InternVL 1.1: 📖 Introduction | ⚡ Quick Start | 📊 Evaluation

Citation

If you find this project useful in your research, please consider citing:

@article{wang2025internvl3_5,
  title={InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency},
  author={Wang, Weiyun and Gao, Zhangwei and Gu, Lixin and Pu, Hengjun and Cui, Long and Wei, Xingguang and Liu, Zhaoyang and Jing, Linglin and Ye, Shenglong and Shao, Jie and others},
  journal={arXiv preprint arXiv:2508.18265},
  year={2025}
}