A fine-tuned Flan-T5-XL model trained on Midjourney image prompts.
Use an input in the format: “Image: [your thing]”
Such as: “Image: a photo of a tiger”
Documentation on how to fine tune: https://replicate.com/docs/guides/fine-tune-a-language-model
fofr
/
image-prompts
Generate image prompts for Midjourney. Prefix inputs with "Image: "
- Public
- 53.3K runs