The reproduction of state-of-the-art multimodal LLM pre-training faces barriers at every stage of the pipeline, including high-quality data filtering, multimodal data mixture strategies, sequence packing techniques, and training frameworks. We introduce Open-Qwen2VL, a fully open-source 2B-parameter Multimodal Large Language Model pre-trained efficiently on 29M image-text pairs using only 220 A100-40G GPU hours. Our approach employs low-to-high dynamic image resolution and multimodal sequence packing to significantly enhance pre-training efficiency. The training dataset was carefully curated using both MLLM-based filtering techniques (e.g., MLM-Filter) and conventional CLIP-based filtering methods, substantially improving data quality and training efficiency. The Open-Qwen2VL pre-training is conducted on academic level 8xA100-40G GPUs at UCSB on 5B packed multimodal tokens, which is 0.36% of 1.4T multimodal pre-training tokens of Qwen2-VL. The final instruction-tuned Open-Qwen2VL outperforms partially-open state-of-the-art MLLM Qwen2-VL-2B on various multimodal benchmarks of MMBench, SEEDBench, MMstar, and MathVista, indicating the remarkable training efficiency of Open-Qwen2VL. We open-source all aspects of our work, including compute-efficient and data-efficient training details, data filtering methods, sequence packing scripts, pre-training data in WebDataset format, FSDP-based training codebase, and both base and instruction-tuned model checkpoints. We redefine "fully open" for multimodal LLMs as the complete release of: 1) the training codebase, 2) detailed data filtering techniques, and 3) all pre-training and supervised fine-tuning data used to develop the model.

翻译：复现前沿多模态大语言模型预训练在流程的每个阶段都面临障碍，包括高质量数据过滤、多模态数据混合策略、序列打包技术以及训练框架。我们推出了Open-Qwen2VL，这是一个完全开源的20亿参数多模态大语言模型，仅使用220个A100-40G GPU小时，在2900万图文对上实现了高效预训练。我们的方法采用从低到高的动态图像分辨率与多模态序列打包技术，显著提升了预训练效率。训练数据集通过结合基于MLLM的过滤技术（例如MLM-Filter）和传统的基于CLIP的过滤方法精心筛选，大幅提高了数据质量和训练效率。Open-Qwen2VL的预训练是在UCSB的学术级8xA100-40G GPU集群上完成的，使用了50亿个打包的多模态词元，这仅占Qwen2-VL所用1.4万亿多模态预训练词元的0.36%。最终经过指令微调的Open-Qwen2VL在MMBench、SEEDBench、MMstar和MathVista等多个多模态基准测试中，表现优于部分开源的前沿MLLM模型Qwen2-VL-2B，这证明了Open-Qwen2VL卓越的训练效率。我们开源了工作的所有方面，包括计算高效与数据高效的训练细节、数据过滤方法、序列打包脚本、WebDataset格式的预训练数据、基于FSDP的训练代码库，以及基础模型和指令微调模型的检查点。我们为多模态大语言模型重新定义了“完全开放”，即完整发布：1）训练代码库，2）详细的数据过滤技术，以及3）用于开发模型的所有预训练数据和监督微调数据。