andreas/ComfyUI
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Make the ltx audio vae more native. (#13486)

Browse Source
This commit is contained in:
comfyanonymous
2026-04-21 08:02:42 -07:00
committed by GitHub
parent e75f775ae8
commit ad94d47221
4 changed files with 41 additions and 70 deletions
Download Patch File Download Diff File Expand all files Collapse all files
comfy/sd.py
+18
View File
@@ -12,6 +12,7 @@ from .ldm.cascade.stage_c_coder import StageC_coder
from .ldm.audio.autoencoder import AudioOobleckVAE
import comfy.ldm.genmo.vae.model
import comfy.ldm.lightricks.vae.causal_video_autoencoder
import comfy.ldm.lightricks.vae.audio_vae
import comfy.ldm.cosmos.vae
import comfy.ldm.wan.vae
import comfy.ldm.wan.vae2_2
@@ -805,6 +806,23 @@ class VAE:
self.downscale_index_formula = (4, 8, 8)
self.memory_used_encode = lambda shape, dtype: (700 * (max(1, (shape[-3] ** 0.66 * 0.11)) * shape[-2] * shape[-1]) * model_management.dtype_size(dtype))
self.memory_used_decode = lambda shape, dtype: (50 * (max(1, (shape[-3] ** 0.65 * 0.26)) * shape[-2] * shape[-1] * 32 * 32) * model_management.dtype_size(dtype))
elif "vocoder.resblocks.0.convs1.0.weight" in sd or "vocoder.vocoder.resblocks.0.convs1.0.weight" in sd: # LTX Audio
self.first_stage_model = comfy.ldm.lightricks.vae.audio_vae.AudioVAE(metadata=metadata)
self.memory_used_encode = lambda shape, dtype: (shape[2] * 330) * model_management.dtype_size(dtype)
self.memory_used_decode = lambda shape, dtype: (shape[2] * shape[3] * 87000) * model_management.dtype_size(dtype)
self.latent_channels = self.first_stage_model.latent_channels
self.audio_sample_rate_output = self.first_stage_model.output_sample_rate
self.autoencoder = self.first_stage_model.autoencoder # TODO: remove hack for ltxv custom nodes
self.output_channels = 2
self.pad_channel_value = "replicate"
self.upscale_ratio = 4096
self.downscale_ratio = 4096
self.latent_dim = 2
self.process_output = lambda audio: audio
self.process_input = lambda audio: audio
self.working_dtypes = [torch.float32]
self.disable_offload = True
self.extra_1d_channel = 16
else:
logging.warning("WARNING: No VAE weights detected, VAE not initalized.")
self.first_stage_model = None