Doesn't work on the new version of ComfyUI

[olesyavlasova909-droid]

Error message occurred while importing the 'ComfyUI-SeedVR2_VideoUpscaler' module.

Traceback (most recent call last):
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\utils\import_utils.py", line 1010, in get_module
return importlib.import_module("." + module_name, self.name)
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "importlib_init.py", line 88, in import_module
File "", line 1387, in _gcd_import
File "", line 1360, in _find_and_load
File "", line 1331, in _find_and_load_unlocked
File "", line 935, in _load_unlocked
File "", line 1023, in exec_module
File "", line 488, in _call_with_frames_removed
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\loaders\single_file_model.py", line 28, in
from .single_file_utils import (
...<33 lines>...
)
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\loaders\single_file_utils.py", line 53, in
from transformers import AutoImageProcessor
File "", line 1412, in _handle_fromlist
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\utils\import_utils.py", line 2207, in getattr
module = self._get_module(self._class_to_module[name])
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\utils\import_utils.py", line 2441, in _get_module
raise e
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\utils\import_utils.py", line 2439, in get_module
return importlib.import_module("." + module_name, self.name)
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "importlib_init.py", line 88, in import_module
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\models\auto\image_processing_auto.py", line 24, in
from ...image_processing_utils import ImageProcessingMixin
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\image_processing_utils.py", line 34, in
from .processing_utils import ImagesKwargs, Unpack
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\processing_utils.py", line 62, in
from .utils.type_validators import (
...<9 lines>...
)
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\utils\type_validators.py", line 19, in
from ..activations import ACT2FN
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\activations.py", line 22, in
from .integrations.hub_kernels import use_kernel_forward_from_hub
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\integrations\hub_kernels.py", line 25, in
from .flash_attention import flash_attention_forward
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\integrations\flash_attention.py", line 9, in
use_top_left_mask = flash_attn_supports_top_left_mask()
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\modeling_flash_attention_utils.py", line 42, in flash_attn_supports_top_left_mask
if is_flash_attn_2_available() or is_flash_attn_3_available() or is_flash_attn_4_available():
~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "Q:\Upscaler\python_embeded\Lib\site-packages\transformers\utils\import_utils.py", line 951, in is_flash_attn_2_available
pkg.replace("", "-") for pkg in PACKAGE_DISTRIBUTION_MAPPING["flash_attn"]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^
KeyError: 'flash_attn'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "Q:\Upscaler\ComfyUI\nodes.py", line 2227, in load_custom_node
module_spec.loader.exec_module(module)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^
File "", line 1023, in exec_module
File "", line 488, in call_with_frames_removed
File "Q:\Upscaler\ComfyUI\custom_nodes\seedvr2_videoupscaler_init.py", line 7, in
from .src.interfaces import comfy_entrypoint, SeedVR2Extension
File "Q:\Upscaler\ComfyUI\custom_nodes\seedvr2_videoupscaler\src\interfaces_init_.py", line 8, in
from .video_upscaler import SeedVR2VideoUpscaler
File "Q:\Upscaler\ComfyUI\custom_nodes\seedvr2_videoupscaler\src\interfaces\video_upscaler.py", line 10, in
from ..utils.downloads import download_weight
File "Q:\Upscaler\ComfyUI\custom_nodes\seedvr2_videoupscaler\src\utils\downloads.py", line 14, in
from .model_registry import MODEL_REGISTRY, DEFAULT_VAE
File "Q:\Upscaler\ComfyUI\custom_nodes\seedvr2_videoupscaler\src\utils\model_registry.py", line 14, in
from ..models.video_vae_v3.modules.attn_video_vae import VideoAutoencoderKLWrapper
File "Q:\Upscaler\ComfyUI\custom_nodes\seedvr2_videoupscaler\src\models\video_vae_v3\modules\attn_video_vae.py", line 20, in
from diffusers.models.autoencoders.vae import DecoderOutput, DiagonalGaussianDistribution
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\autoencoders_init_.py", line 1, in
from .autoencoder_asym_kl import AsymmetricAutoencoderKL
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\autoencoders\autoencoder_asym_kl.py", line 21, in
from .vae import AutoencoderMixin, DecoderOutput, DiagonalGaussianDistribution, Encoder, MaskConditionDecoder
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\autoencoders\vae.py", line 24, in
from ..unets.unet_2d_blocks import (
...<4 lines>...
)
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\unets_init_.py", line 6, in
from .unet_2d import UNet2DModel
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\unets\unet_2d.py", line 23, in
from .unet_2d_blocks import UNetMidBlock2D, get_down_block, get_up_block
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\unets\unet_2d_blocks.py", line 36, in
from ..transformers.dual_transformer_2d import DualTransformer2DModel
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\transformers_init_.py", line 5, in
from .auraflow_transformer_2d import AuraFlowTransformer2DModel
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\models\transformers\auraflow_transformer_2d.py", line 23, in
from ...loaders import FromOriginalModelMixin, PeftAdapterMixin
File "", line 1412, in _handle_fromlist
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\utils\import_utils.py", line 1000, in getattr
module = self._get_module(self._class_to_module[name])
File "Q:\Upscaler\python_embeded\Lib\site-packages\diffusers\utils\import_utils.py", line 1012, in _get_module
raise RuntimeError(
...<2 lines>...
) from e
RuntimeError: Failed to import diffusers.loaders.single_file_model because of the following error (look up to see its traceback):
'flash_attn'
ComfyUI Manager V3.39.2
DB
Channel
Preview method
Share
Component
Update
EXPERIMENTAL

Keywords: security

Issue News:

[security] Compromised litellm PyPI packages (v1.82.7, v1.82.8) have been added to ComfyUI-Manager's security scanner blacklist. These packages were part of a supply chain attack that harvests credentials and exfiltrates them to an attacker-controlled server. If you have litellm installed, ComfyUI will block startup and display remediation steps. Please update ComfyUI-Manager to the latest version. https://github.com/BerriAI/litellm/issues/24518

Features/Updates News:

ComfyUI v0.18.2 is released (https://github.com/comfyanonymous/ComfyUI/releases/tag/v0.18.2)
ComfyUI: v0.18.1-47-g8cbbea8f
(2026-04-05)
Manager: V3.39.2
Filter
1 custom nodes
ID
Title
Version
Action
Nodes
Description
Author
★
Last Update
20

ComfyUI-SeedVR2_VideoUpscaler
2.5.24
[↑2.5.22]
4 nodes
4 conflicts
SeedVR2 official ComfyUI integration: ByteDance-Seed's one-step diffusion-based video/image upscaling with memory-efficient inference
✅ AInVFX
2 283
2025-12-24

[akza07]

Switch to tagged release v2.5.18. It seems to work for me.
I guess some new code changes introduced some bug.

[rgj7]

Was able to bypass temporarily by adding the follow lines of code to the top of __init__.py in project root:

# Patch for transformers>=5.5.0 bug: 'flash_attn' missing from PACKAGE_DISTRIBUTION_MAPPING
# but accessed unconditionally in is_flash_attn_2_available(), causing KeyError on import.
try:
    from transformers.utils.import_utils import PACKAGE_DISTRIBUTION_MAPPING
    if "flash_attn" not in PACKAGE_DISTRIBUTION_MAPPING:
        PACKAGE_DISTRIBUTION_MAPPING["flash_attn"] = ["flash_attn", "flash-attn"]
except Exception:
    pass

[Winnougan]

Not working with latest comfyui

[naxci1]

use portable Comfyui

[RogueWild]

Was able to bypass temporarily by adding the follow lines of code to the top of __init__.py in project root:

 # Patch for transformers>=5.5.0 bug: 'flash_attn' missing from PACKAGE_DISTRIBUTION_MAPPING
 # but accessed unconditionally in is_flash_attn_2_available(), causing KeyError on import.
 try:
     from transformers.utils.import_utils import PACKAGE_DISTRIBUTION_MAPPING
     if "flash_attn" not in PACKAGE_DISTRIBUTION_MAPPING:
         PACKAGE_DISTRIBUTION_MAPPING["flash_attn"] = ["flash_attn", "flash-attn"]
 except Exception:
     pass

Thanks 🙏

[JanBluePix]

works in Linx too. Thank you rgj7

[odeguard]

Switch to tagged release v2.5.18. It seems to work for me. I guess some new code changes introduced some bug.

It works for me. In Extensions Manager i clicked on version and select 2.5.18 in pop-up menu. Reload. Works. <3