WaveCut/Anima-Preview-3-SDNQ-uint4
Text-to-Image • Updated • 6
SDNQ Anima Preview 3 component and full Diffusers checkpoints with 1024 benchmark grids.
Note UINT4 SDNQ transformer component; loads as CosmosTransformer3DModel, not a standalone pipeline.
Note INT8 SDNQ transformer component; loads as CosmosTransformer3DModel, not a standalone pipeline.
Note Runnable full Diffusers pipeline with UINT4 transformer, Anima adapter, Qwen3 text encoder, VAE, grid and stats.
Note Runnable full Diffusers pipeline with INT8 transformer, Anima adapter, Qwen3 text encoder, VAE, grid and stats.