The official code of Yume
-
Updated
Jan 14, 2026 - Python
The official code of Yume
official code for "magicworld: towards long-horizon stability for interactive video world exploration"
Inject a message into a running LLM generation — no restart, no KV cache recompute. The model finishes its current sentence, then changes course.
Instead of writing prompts, you simply toggle the exact attributes you want and generate a face image.
To associate your repository with the interactive-generation topic, visit your repo's landing page and select "manage topics."