Vision · Haotian Liu

LLaVA Instruct

Visual instruction tuning dataset. Contains 158k unique language-image instruction-following samples to make LLMs 'see'.

View on Hugging Face →Read Paper

Load it

from datasets import load_dataset
ds = load_dataset("liuhaotian/LLaVA-Instruct-150K")

Preview a sample row

{
  "image": "coco_train_000000581929.jpg",
  "conversations": [
    { "from": "human", "value": "What is unusual about this image?" },
    { "from": "gpt", "value": "The man is ironing clothes on the back of a taxi..." }
  ]
}

Frequently asked questions

Can I use LLaVA Instruct commercially?

Not in a product — LLaVA Instruct is released under CC-BY-NC 4.0, which restricts use to research and other non-commercial purposes. For commercial fine-tuning, pick a permissively licensed dataset from the same category instead.

How much data does LLaVA Instruct contain, and do I need all of it?

LLaVA Instruct contains 158k Samples. You rarely need all of it: for style and format fine-tuning, a few hundred to a few thousand examples are enough — load a slice (e.g. split="train[:1000]") and scale up only if quality plateaus.

What is LLaVA Instruct best used for?

Adding image understanding to an open LLM (the LLaVA recipe). It belongs to the Vision section of our dataset hub, where you'll find alternatives and complementary sets.