Arguably most interesting facts about the new Wan 2.2 model:
- they are now using a 27B MoE architecture (with two 14B experts, for low level and high level detail), which were usually only used for autoregressive LLMs rather than diffusion models
- the smaller 5B model supports up to 720p24 video and runs on 24 GB of VRAM, e.g. an RTX 4090, a consumer graphics card
- if their benchmarks are reliable, the model performance is SOTA even compared to closed source models
esseph · 2h ago
Ugh hate they used this name
yorwba · 2h ago
You can call it Wanxiang (万相, ten thousand pictures) if you want. Similarly, Qwen is Qianwen (千问, one thousand questions).
CapsAdmin · 1h ago
Its original name was WanX, but the gen ai community found that to be too funny / unfortunate, so they changed it to just Wan.
latentsea · 2h ago
They should just pretend it's an acronym. Wide Art Network.
- they are now using a 27B MoE architecture (with two 14B experts, for low level and high level detail), which were usually only used for autoregressive LLMs rather than diffusion models
- the smaller 5B model supports up to 720p24 video and runs on 24 GB of VRAM, e.g. an RTX 4090, a consumer graphics card
- if their benchmarks are reliable, the model performance is SOTA even compared to closed source models