berhaan/LLaVA-595K-Translation-TR-EN
收藏官方服务：
资源简介：
Turkish-LLaVA-Pretrain数据集是一个多语言视觉-语言数据集，专为预训练支持土耳其语的多模态模型设计。该数据集在LLaVA-CC3M-Pretrain-595K的基础上增加了高质量的土耳其语翻译，适用于图像字幕、视觉问答和多模态对话生成等任务。
The Turkish-LLaVA-Pretrain dataset is a multilingual vision-language dataset designed for pretraining multimodal models with Turkish support. It extends the LLaVA-CC3M-Pretrain-595K with high-quality Turkish translations, suitable for tasks such as image captioning, visual question answering, and multimodal dialog generation.
提供机构：
berhaan
