yujunhuics / Reyes
从零到一实现了一个多模态大模型，并命名为Reyes（睿视），R：睿，eyes：眼。Reyes的参数量为8B，视觉编码器使用的是InternViT-300M-448px-V2_5,语言模型侧使用的是Qwen2.5-7B-Instruct，Reyes也通过一个两层MLP投影层连接视觉编码器与语言模型。
☆12Updated 2 months ago
Alternatives and similar repositories for Reyes:
Users that are interested in Reyes are comparing it to the libraries listed below
- Exploring Efficient Fine-Grained Perception of Multimodal Large Language Models☆60Updated 6 months ago
- Multimodal Open-O1 (MO1) is designed to enhance the accuracy of inference models by utilizing a novel prompt-based approach. This tool wo…☆29Updated 7 months ago
- ☆56Updated last year
- This project aims to collect and collate various datasets for multimodal large model training, including but not limited to pre-training …☆43Updated this week
- [NeurIPS 2024] Needle In A Multimodal Haystack (MM-NIAH): A comprehensive benchmark designed to systematically evaluate the capability of…☆116Updated 5 months ago
- Official PyTorch Implementation of MLLM Is a Strong Reranker: Advancing Multimodal Retrieval-augmented Generation via Knowledge-enhanced …☆70Updated 5 months ago
- 训练一个对中文支持更好的LLaVA模型，并开源训练代码和数据。☆54Updated 8 months ago
- ZO2 (Zeroth-Order Offloading): Full Parameter Fine-Tuning 175B LLMs with 18GB GPU Memory☆93Updated this week
- The huggingface implementation of Fine-grained Late-interaction Multi-modal Retriever.☆86Updated 3 months ago
- A Survey of Multimodal Retrieval-Augmented Generation☆17Updated 3 weeks ago
- The official repo for “TextCoT: Zoom In for Enhanced Multimodal Text-Rich Image Understanding”.☆39Updated 7 months ago
- ☆53Updated 3 weeks ago
- The proposed simulated dataset consisting of 9,536 charts and associated data annotations in CSV format.☆24Updated last year
- MMR1: Advancing the Frontiers of Multimodal Reasoning☆159Updated last month
- ☆29Updated 8 months ago
- Official code of *Virgo: A Preliminary Exploration on Reproducing o1-like MLLM*☆100Updated 2 months ago
- A Token-level Text Image Foundation Model for Document Understanding☆91Updated this week
- Rethinking RL Scaling for Vision Language Models: A Transparent, From-Scratch Framework and Comprehensive Evaluation Scheme☆120Updated 3 weeks ago
- Vary-tiny codebase upon LAVIS （for training from scratch）and a PDF image-text pairs data (about 600k including English/Chinese)☆81Updated 7 months ago
- ☆83Updated 11 months ago
- Building a VLM model starts from the basic module.☆15Updated last year
- ☆63Updated last year
- Pruning the VLLMs☆92Updated 4 months ago
- Our 2nd-gen LMM☆33Updated 11 months ago
- [ArXiv] V2PE: Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding☆46Updated 4 months ago
- Open-Qwen2VL: Compute-Efficient Pre-Training of Fully-Open Multimodal LLMs on Academic Resources☆190Updated last month
- ☆73Updated last year
- ☆25Updated 8 months ago
- GOT的vLLM加速实现 并结合 MinerU 实现RAG中的pdf 解析☆56Updated 6 months ago
- Official Repository of MMLONGBENCH-DOC: Benchmarking Long-context Document Understanding with Visualizations☆80Updated 9 months ago