GitHub · Python
speridlabs/iris-3b
Iris-3B: pixel-space generation & general vision learner. A 3B diffusion transformer that generates every pixel directly (no VAE); the same prior, fine-tuned, does monocular depth estimation and 4x image restoration. Paper, weights, training code and demo.
- Stars
- 237
- Forks
- 21
- Stars gained in 24 hours
- 0
- Forks gained in 24 hours
- 0
Topics: depth-estimation, diffusion, pixel-space, pytorch, super-resolution, text-to-image, flow-matching, generative-prior, image-restoration, vision-foundation-model
Latest data point: Nov 22, 2026 · 237 stars · 21 forks