Salesforce/cota-llava
收藏Hugging Face2025-01-06 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/Salesforce/cota-llava
下载链接
链接失效反馈官方服务：
资源简介：
TACO是一个大规模的合成链式思考和动作（CoTA）数据集，由多模态大型语言模型生成，包含15种动作和一个终止动作。数据集来源于Cauldron和Mantis-Instruct，收集自多个现有数据集，旨在微调多模态语言模型，以产生思考和动作链来回答复杂的视觉问题。
TACO is a large-scale dataset of synthetic Chains-of-Thought-and-Action (CoTA) generated by multi-modal large language models, containing 15 actions and a terminate action. The dataset is sourced from Cauldron and Mantis-Instruct, collected from various existing datasets, aiming to fine-tune multi-modal language models to produce chains of thoughts and actions for answering complex visual questions.
提供机构：
Salesforce
