You can deploy this model and start earning money today!

gtz1/
LLaVA-Instruct-150K

visual-question-answering

question-answering

cc-by-4.0

100K<n<1M

Views

No views yet

how I filtered llava_v1_5_mix665k.json into llava_v1_5_mix665k_vqa_clean_st_single_shuf.jsonl https://colab.research.google.com/drive/1um_wlBFacXCHSabmJCg9Ll3DvP5rGxuE?usp=sharing

	LLaVA Visual Instruct 150K Dataset Card




	
	Dataset details

Dataset type: LLaVA Visual Instruct 150K is a set of GPT-generated multimodal instruction-following data. It is constructed for visual instruction tuning and for building large multimodal towards GPT-4 vision/language capability. Dataset… See the full description on the dataset page: https://huggingface.co/datasets/gtz1/LLaVA-Instruct-150K.