---
language: en
dataset_info:
  features:
  - name: prompt
    dtype: string
  splits:
  - name: train
    num_bytes: 284636288
    num_examples: 1819808
  download_size: 101931289
  dataset_size: 284636288
---
Subset dataset of [diffusiondb](https://huggingface.co/datasets/poloclub/diffusiondb) consisting of just unique prompts. 

Created this subset dataset for the [Prompt Extend](https://github.com/daspartho/prompt-extend) project.