upstream-archive/stable-diffusion.cpp
1
0
Fork 0
mirror of https://github.com/leejet/stable-diffusion.cpp.git synced 2026-09-25 07:27:53 -05:00
Code Issues Packages Projects Releases Wiki Activity

feat: generalize temporal tiling across video VAEs (#1926)

Browse Source
This commit is contained in:
leejet
2026-08-31 00:11:38 +08:00
committed by GitHub
parent 40e605f3f1
commit 6b3edaaf32
11 changed files with 516 additions and 207 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/server/api.md
+2 -1
View File
@@ -518,7 +518,8 @@ Shared default fields used by both `img_gen` and `vid_gen`:
| `output_format` | `string` |
| `output_compression` | `integer` |
`vae_tiling_params.extra_tiling_args` accepts a key=value list. For LTX video VAE temporal tiling, `temporal_tile_frames` defaults to `4` and `temporal_tile_overlap` defaults to `1`.
`vae_tiling_params.extra_tiling_args` accepts a key=value list. Supported video VAEs accept `temporal_tile_frames` (alias `temporal_tile_size`, default `4`) and `temporal_tile_overlap` (default `1`).
LTX and Wan preserve causal state between temporal tiles. Hunyuan Video and TAEHV use overlap blending. MiniMax H3 keeps its model-specific fixed temporal windows because its latent-to-frame mapping is non-linear.
`img_gen`-specific default fields: