File size: 2,850 Bytes
9544b88 990f01b b93142f 990f01b 9544b88 13954e4 9544b88 13954e4 9544b88 13954e4 9544b88 13954e4 9544b88 13954e4 990f01b 055199f 13954e4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 |
# MAmmoTH-VL-8B
[🏠 Homepage](https://mammoth-vl.github.io/) | [🤖 MAmmoTH-VL-8B](https://huggingface.co/MAmmoTH-VL/MAmmoTH-VL-8B) | [💻 Code](https://github.com/MAmmoTH-VL/MAmmoTH-VL) | [📄 Arxiv](https://arxiv.org/abs/2412.05237) | [📕 PDF](https://arxiv.org/pdf/2412.05237) | [🖥️ Demo](https://huggingface.co/spaces/paralym/MAmmoTH-VL-8B)
# Abstract
Open-source multimodal large language models (MLLMs) have shown significant potential in a broad range of multimodal tasks. However, their reasoning capabilities remain constrained by existing instruction-tuning datasets, which were predominately repurposed from academic datasets such as VQA, AI2D, and ChartQA. These datasets target simplistic tasks, and only provide phrase-level answers without any intermediate rationales.
To address these challenges, we introduce a scalable and cost-effective method to construct a large-scale multimodal instruction-tuning dataset with rich intermediate rationales designed to elicit CoT reasoning. Using only open models, we create a dataset containing 12M instruction-response pairs to cover diverse, reasoning-intensive tasks with detailed and faithful rationales. Experiments demonstrate that training MLLMs on this dataset significantly improves reasoning capabilities, achieving state-of-the-art performance on benchmarks such as MathVerse (+8.1%), MMMU-Pro (+7%), and MuirBench (+13.3%). Additionally, the model demonstrates notable improvements of up to 4% on non-reasoning-based benchmarks. Ablation studies further highlight the importance of key components, such as rewriting and self-filtering, in the dataset construction process.
# Performance
We highlight different groups of models with different colors: <span style="background-color: #f2f2f2">closed-source models</span>, <span style="background-color: #cce0ff">open weights</span> but closed training details, and <span style="background-color: #e0f7e0">fully open-source</span> models. Results are from official sources or running with lmms-eval package if unavailable.
## Multi-Discipline Knowledge and Mathematical Reasoning
## Chart & Doc Understanding and Multimodal Interactions & Preferences
## Multi-Image and Video
## Citing the Model
```
@article{guo2024mammothvlelicitingmultimodalreasoning,
title={MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale},
author={Jarvis Guo and Tuney Zheng and Yuelin Bai and Bo Li and Yubo Wang and King Zhu and Yizhi Li and Graham Neubig and Wenhu Chen and Xiang Yue},
year={2024},
eprint={2412.05237},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2412.05237},
}
``` |