Data-Juicer Sandbox: A Feedback-Driven Suite for Multimodal Data-Model Co-development

Daoyuan Chen, Haibin Wang, Yilun Huang, Ce Ge, Yaliang Li, Bolin Ding, Jingren Zhou
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:9198-9230, 2025.

Abstract

The emergence of multimodal large models has advanced artificial intelligence, introducing unprecedented levels of performance and functionality. However, optimizing these models remains challenging due to historically isolated paths of model-centric and data-centric developments, leading to suboptimal outcomes and inefficient resource utilization. In response, we present a new sandbox suite tailored for integrated data-model co-development. This sandbox provides a feedback-driven experimental platform, enabling cost-effective iteration and guided refinement of both data and models. Our proposed “Probe-Analyze-Refine” workflow, validated through practical use cases on multimodal tasks such as image-text pre-training with CLIP, image-to-text generation with LLaVA-like models, and text-to-video generation with DiT-based models, yields transferable and notable performance boosts, such as topping the VBench leaderboard. A comprehensive set of over 100 experiments demonstrated the suite’s usability and extensibility, while also uncovering insights into the interplay between data quality, diversity, model behavior, and computational costs. All codes, datasets, and models are open-sourced to foster future research and applications that would otherwise be infeasible due to the lack of a dedicated co-development infrastructure.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-chen25bm, title = {Data-Juicer Sandbox: A Feedback-Driven Suite for Multimodal Data-Model Co-development}, author = {Chen, Daoyuan and Wang, Haibin and Huang, Yilun and Ge, Ce and Li, Yaliang and Ding, Bolin and Zhou, Jingren}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {9198--9230}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/chen25bm/chen25bm.pdf}, url = {https://proceedings.mlr.press/v267/chen25bm.html}, abstract = {The emergence of multimodal large models has advanced artificial intelligence, introducing unprecedented levels of performance and functionality. However, optimizing these models remains challenging due to historically isolated paths of model-centric and data-centric developments, leading to suboptimal outcomes and inefficient resource utilization. In response, we present a new sandbox suite tailored for integrated data-model co-development. This sandbox provides a feedback-driven experimental platform, enabling cost-effective iteration and guided refinement of both data and models. Our proposed “Probe-Analyze-Refine” workflow, validated through practical use cases on multimodal tasks such as image-text pre-training with CLIP, image-to-text generation with LLaVA-like models, and text-to-video generation with DiT-based models, yields transferable and notable performance boosts, such as topping the VBench leaderboard. A comprehensive set of over 100 experiments demonstrated the suite’s usability and extensibility, while also uncovering insights into the interplay between data quality, diversity, model behavior, and computational costs. All codes, datasets, and models are open-sourced to foster future research and applications that would otherwise be infeasible due to the lack of a dedicated co-development infrastructure.} }
Endnote
%0 Conference Paper %T Data-Juicer Sandbox: A Feedback-Driven Suite for Multimodal Data-Model Co-development %A Daoyuan Chen %A Haibin Wang %A Yilun Huang %A Ce Ge %A Yaliang Li %A Bolin Ding %A Jingren Zhou %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-chen25bm %I PMLR %P 9198--9230 %U https://proceedings.mlr.press/v267/chen25bm.html %V 267 %X The emergence of multimodal large models has advanced artificial intelligence, introducing unprecedented levels of performance and functionality. However, optimizing these models remains challenging due to historically isolated paths of model-centric and data-centric developments, leading to suboptimal outcomes and inefficient resource utilization. In response, we present a new sandbox suite tailored for integrated data-model co-development. This sandbox provides a feedback-driven experimental platform, enabling cost-effective iteration and guided refinement of both data and models. Our proposed “Probe-Analyze-Refine” workflow, validated through practical use cases on multimodal tasks such as image-text pre-training with CLIP, image-to-text generation with LLaVA-like models, and text-to-video generation with DiT-based models, yields transferable and notable performance boosts, such as topping the VBench leaderboard. A comprehensive set of over 100 experiments demonstrated the suite’s usability and extensibility, while also uncovering insights into the interplay between data quality, diversity, model behavior, and computational costs. All codes, datasets, and models are open-sourced to foster future research and applications that would otherwise be infeasible due to the lack of a dedicated co-development infrastructure.
APA
Chen, D., Wang, H., Huang, Y., Ge, C., Li, Y., Ding, B. & Zhou, J.. (2025). Data-Juicer Sandbox: A Feedback-Driven Suite for Multimodal Data-Model Co-development. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:9198-9230 Available from https://proceedings.mlr.press/v267/chen25bm.html.

Related Material