AUTOSIGNAL 车智信号
Human-curated. Expert-annotated. Every signal traced to source.
人工精选 · 专家点评 · 每条信号可溯源
Research Frontier研究前沿
DataComp-VLM: Data Mixing Beats Filtering in New Vision-Language Model BenchmarkDataComp-VLM发布，数据混合策略优于过滤提升视觉语言模型训练
S 1.7
T1
1 sources1 个来源
R7-research
- Researchers introduced DataComp-VLM (DCVLM), a benchmark for data-centric vision-language model training, comprising 160 datasets with 6 trillion multimodal tokens across image-caption pairs, multimodal interleaved documents, text-only, and instruction-tuning data.
- The benchmark enables testing of curation strategies (filtering, mixing, formatting, sampling) on 1B-8B parameter models with token budgets from 6.25B to 200B, evaluated across up to 52 downstream tasks in 9 domains.
- Key findings show data mixing outperforms filtering, with instruction-heavy mixtures scaling better than caption-heavy ones, enabling an 8B VLM to achieve 63.6% accuracy with 200B tokens—a +5.4 percentage point improvement over FineVision.
- 研究团队推出DataComp-VLM（DCVLM）基准，这是用于视觉语言模型数据驱动训练的基准，包含160个数据集共6万亿个多模态令牌，涵盖图像-标题对、多模态交错文档、纯文本和指令调优四种数据类型。
- 该基准允许在1B至8B参数的模型上测试多种数据策略（过滤、混合、格式化、采样），令牌预算范围为6.25B至200B，并在9个领域的最多52个下游任务上进行评估。
- 研究发现数据混合策略优于过滤策略，指令密集型混合优于标题密集型，8B参数模型使用200B令牌可达到63.6%的准确率，相比FineVision提升5.4个百分点。