Skip to content

Fix Low VRAM Audio VAE loader returning an unusable model #562

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Henry Kroll III (themanyone) wants to merge 2 commits into
base: master
Choose a base branch
from
Open

Fix Low VRAM Audio VAE loader returning an unusable model #562

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
814d5f1
Fix Low VRAM Audio VAE loader returning an unusable model
Sep 19, 2026
f97fa5f
Merge upstream master (Lightricks) into lowvram_pr
Sep 29, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 8 additions & 2 deletions low_vram_loaders.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,10 @@
This reduces peak VRAM usage in memory-constrained environments.
"""

import comfy.sd
import comfy.utils
import folder_paths
import nodes
from comfy.ldm.lightricks.vae.audio_vae import AudioVAE
from comfy_api.latest import io
from comfy_extras.nodes_hunyuan import LatentUpscaleModelLoader

Expand Down Expand Up @@ -93,7 +93,13 @@ def load_audio_vae_sequentially(
) -> tuple:
ckpt_path = folder_paths.get_full_path_or_raise("checkpoints", ckpt_name)
sd, metadata = comfy.utils.load_torch_file(ckpt_path, return_metadata=True)
audio_vae = AudioVAE(sd, metadata)
sd = comfy.utils.state_dict_prefix_replace(
sd,
{"audio_vae.": "autoencoder.", "vocoder.": "vocoder."},
filter_keys=True,
)
audio_vae = comfy.sd.VAE(sd=sd, metadata=metadata)
audio_vae.throw_exception_if_invalid()
return (audio_vae,)


Expand Down