Meta is about to release a pixel space model (Tuna-2)

There's a catch, though, they break it on purpose and want you to fix it: "Due to organizational policy constraints, we are unable to release the full production-trained model weights. To support the research community, we plan to release a foundation checkpoint with a small number of layers removed from both the LLM backbone and the diffusion head (flow head). The remaining layers and all other components (vision encoder, projections, embeddings, etc.) are fully preserved. With a short fine-tun

More from StableDiffusion

Looneytunes background style for ZIT

StableDiffusion·5

Moss-Audio Captioning is a first of its kind! | Here's the repo: I modified the GUI to allow for batch captioning, youtube videos, and file chunking.

StableDiffusion·13

Got early access access to LingBot-World-Fast at 17 FPS! Here's what I found.

StableDiffusion·8

Got early access access to LingBot-World-Fast at 17 FPS! Here's what I found.

StableDiffusion·9

Ernie VS Qwen and ZiT - Big Test

StableDiffusion·8

LTX 2.3 Prompt Relay with a messy zombie chase scene(Prompt Relay test)

StableDiffusion·12