gfodor / instructblip

Image captioning via vision-language models with instruction tuning

  • Public
  • 528.9K runs
  • GitHub
  • Paper
  • License

Input

Output

Run time and cost

This model runs on Nvidia A100 (40GB) GPU hardware. Predictions typically complete within 70 seconds.

Readme

InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning

project page paper

InstructBLIP is an instruction tuned image captioning model.

Comparison

From the project page:

“The response from InstructBLIP is more comprehensive than GPT-4, more visually-grounded than LLaVA, and more logical than MiniGPT-4. The responses of GPT-4 and LLaVA are obtained from their respective papers, while the official demo is used for MiniGPT-4.”