LLaVA-Instruct-150K

LLaVA Team (Liu et al.)
View on Hugging Face ↗
MultimodalCC-BY-NC-4.0publicTrusted158K rowsnon-commercial

158K GPT-4-generated visual instructions over COCO images. Research-only.

158K GPT-4-generated visual instructions over COCO images. Research-only.

At a glance

  • Modality: Multimodal
  • Use case: Instruction tuning
  • Rows: 158K
  • License: CC-BY-NC-4.0
  • Format: jsonl