v0.3.26

What's Changed

• Support "fixed" HunyuanVideo i2v model (actually a model with a different architecture from the original released version).
• Use fp16 as the default compute dtype for the WAN 2.1 models.
• Support fp8_scaled model files that don't enable the fp8 matrix mult by default.
• fixed: Incorrect guide message for missing frontend. by @ltdrdata in #7105
• Typo in node_typing.py by @JettHu in #7092
• Update frontend to 1.11.8 by @huchenlei in #7119
• Weight Hooks Switching Optimization by @Kosinkadink in #7067
• Fix stable cascade VAE issues with lowvram.

Full Changelog: v0.3.24...v0.3.26

v0.3.25

What's Changed

• Support "fixed" HunyuanVideo i2v model (actually a model with a different architecture from the original released version).
• Use fp16 as the default compute dtype for the WAN 2.1 models.
• Support fp8_scaled model files that don't enable the fp8 matrix mult by default.
• fixed: Incorrect guide message for missing frontend. by @ltdrdata in #7105
• Typo in node_typing.py by @JettHu in #7092
• Update frontend to 1.11.8 by @huchenlei in #7119
• Fix stable cascade VAE issues with lowvram.

Full Changelog: v0.3.24...v0.3.25

v0.3.24

• Fix regression when using incompatible embeddings.

Full Changelog: v0.3.23...v0.3.24

v0.3.23

What's Changed

• Support HunyuanVideo Image to Video model.
• [NodeDef] Explicitly add control_after_generate to seed/noise_seed by @huchenlei in #7059
• Better argument handling of front-end-root by @silveroxides in #7043
• Add type hint for FileLocator by @huchenlei in #6968

Full Changelog: v0.3.22...v0.3.23

v0.3.22

What's Changed

• fix: ltxv crop guides works with 0 keyframes by @kvochko in #7085

Full Changelog: v0.3.21...v0.3.22

v0.3.21

• Fix Issue where the LTXV 0.9.5 img2vid workflow didn't work with some image resolutions.

Full Changelog: v0.3.20...v0.3.21

v0.3.20

What's Changed

• Support LTXV 0.9.5
• Improve: Provide a better guide message for the missing frontend message. by @ltdrdata in #7079

Full Changelog: v0.3.19...v0.3.20

v0.3.19

What's Changed

• Temporal Area Composition: ConditioningSetAreaPercentageVideo
• enabling fp16 accumulations now prioritizes loading models in fp16 when possible.
• Use enum list for --fast options by @huchenlei in #7024
• Use comfyui_frontend_package pypi package to manage frontend dependency (Frontend v1.10.17) by @huchenlei in #7021
• improved: better frontend package installation guide by @ltdrdata in #7047

Full Changelog: v0.3.18...v0.3.19

v0.3.18

What's Changed

• Improve Wan performance by allowing batching and fix issue with long prompts.
• Support Cambricon MLU by @BiologicalExplosion in #6964

New Contributors

• @BiologicalExplosion made their first contribution in #6964

Full Changelog: v0.3.17...v0.3.18

v0.3.17

• Fix artifacts when inferencing wan with fp8 weights.

Full Changelog: v0.3.16...v0.3.17