Dataset · MohamedRashad
Midjourney prompts of varying detail
MohamedRashad/midjourney-detailed-prompts
1-10k English Midjourney images with short and detailed prompts for text-to-image.
What's inside
An English dataset of images and text prompts for text-to-image tasks. Prompts feature multiple detail levels ranging from short to long. Images are sourced from the midjourney-images dataset, descriptions were generated using Qwen-VL-Max, and the short and long prompts were created by Command R.
What you can do with it
- Fine-tune a text-to-image model on detailed prompts
- Train a LoRA for prompt generation
- Evaluate image description quality
- Tasks
- Text → image
- Languages
- English
- Size
- 1K–10K rows
- Downloads this month
- 308
- Likes
- 75
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("MohamedRashad/midjourney-detailed-prompts")