moca-embed/MAmmoTH-VL-Instruct-12M
收藏Hugging Face2025-07-01 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/moca-embed/MAmmoTH-VL-Instruct-12M
下载链接
链接失效反馈官方服务：
资源简介：
MAmmoTH-VL-Instruct-12M是一个视觉问答风格的持续预训练数据集，用于MoCa模型。它改编自MAmmoTH-VL-Instruct-12M数据集，通过拼接提示和响应。该数据集包含文本和图像二进制序列的交错多模态示例。
MAmmoTH-VL-Instruct-12M is a VQA-style dataset used for modality-aware continual pre-training of MoCa models. It is adapted from the MAmmoTH-VL-Instruct-12M dataset by concatenating prompts and responses. The dataset consists of interleaved multimodal examples with text strings and image binary sequences.
提供机构：
moca-embed
