298977887/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2025-04-13 11:53:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

Disable partial offloading of audio VAE.

Browse Source
This commit is contained in:
comfyanonymous 2025-04-04 21:24:56 -04:00
parent 721253cb05
commit 3a100b9a55
1 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
comfy/sd.py
View File

@ -265,6 +265,7 @@ class VAE:
self.process_input = lambda image: image * 2.0 - 1.0
self.process_output = lambda image: torch.clamp((image + 1.0) / 2.0, min=0.0, max=1.0)
self.working_dtypes = [torch.bfloat16, torch.float32]
self.disable_offload = False
self.downscale_index_formula = None
self.upscale_index_formula = None
@ -337,6 +338,7 @@ class VAE:
self.process_output = lambda audio: audio
self.process_input = lambda audio: audio
self.working_dtypes = [torch.float16, torch.bfloat16, torch.float32]
self.disable_offload = True
elif "blocks.2.blocks.3.stack.5.weight" in sd or "decoder.blocks.2.blocks.3.stack.5.weight" in sd or "layers.4.layers.1.attn_block.attn.qkv.weight" in sd or "encoder.layers.4.layers.1.attn_block.attn.qkv.weight" in sd: #genmo mochi vae
if "blocks.2.blocks.3.stack.5.weight" in sd:
sd = comfy.utils.state_dict_prefix_replace(sd, {"": "decoder."})
@ -515,7 +517,7 @@ class VAE:
pixel_samples = None
try:
memory_used = self.memory_used_decode(samples_in.shape, self.vae_dtype)
model_management.load_models_gpu([self.patcher], memory_required=memory_used)
model_management.load_models_gpu([self.patcher], memory_required=memory_used, force_full_load=self.disable_offload)
free_memory = model_management.get_free_memory(self.device)
batch_number = int(free_memory / memory_used)
batch_number = max(1, batch_number)
@ -544,7 +546,7 @@ class VAE:
def decode_tiled(self, samples, tile_x=None, tile_y=None, overlap=None, tile_t=None, overlap_t=None):
self.throw_exception_if_invalid()
memory_used = self.memory_used_decode(samples.shape, self.vae_dtype) #TODO: calculate mem required for tile
model_management.load_models_gpu([self.patcher], memory_required=memory_used)
model_management.load_models_gpu([self.patcher], memory_required=memory_used, force_full_load=self.disable_offload)
dims = samples.ndim - 2
args = {}
if tile_x is not None:
@ -578,7 +580,7 @@ class VAE:
pixel_samples = pixel_samples.movedim(1, 0).unsqueeze(0)
try:
memory_used = self.memory_used_encode(pixel_samples.shape, self.vae_dtype)
model_management.load_models_gpu([self.patcher], memory_required=memory_used)
model_management.load_models_gpu([self.patcher], memory_required=memory_used, force_full_load=self.disable_offload)
free_memory = model_management.get_free_memory(self.device)
batch_number = int(free_memory / max(1, memory_used))
batch_number = max(1, batch_number)
@ -612,7 +614,7 @@ class VAE:
pixel_samples = pixel_samples.movedim(1, 0).unsqueeze(0)
memory_used = self.memory_used_encode(pixel_samples.shape, self.vae_dtype) # TODO: calculate mem required for tile
model_management.load_models_gpu([self.patcher], memory_required=memory_used)
model_management.load_models_gpu([self.patcher], memory_required=memory_used, force_full_load=self.disable_offload)
args = {}
if tile_x is not None: