Comfyui cannot start with torch 2.9 +cu130

### Custom Node Testing

- [x] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)

### Expected Behavior

start the comfyui and show the webui normally, with torch 2.9, cu130, 5060Ti

### Actual Behavior

After starting to run the main.py:

### Steps to Reproduce

5060 Ti, upgrade CUDA to 13.0, upgrade torch and torchvision to the newest version with cu130, then got this error after env check

### Debug Logs

after closing the dialog box, i got the error log:
```powershell
Prestartup times for custom nodes:
0.0 seconds: C:\comfy\ComfyUI\custom_nodes\rgthree-comfy
0.0 seconds: C:\comfy\ComfyUI\custom_nodes\ComfyUI-Easy-Use
0.0 seconds: C:\comfy\ComfyUI\custom_nodes\comfyui-deploy
3.4 seconds: C:\comfy\ComfyUI\custom_nodes\ComfyUI-Manager

Checkpoint files will always be loaded safely.
Total VRAM 16311 MB, total RAM 63095 MB
pytorch version: 2.9.0+cu130
xformers version: 0.0.33+00a7a5f.d20251028
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 5060 Ti : cudaMallocAsync
working around nvidia conv3d memory bug.
Traceback (most recent call last):
File "C:\comfy\ComfyUI\main.py", line 149, in
import execution
File "C:\comfy\ComfyUI\execution.py", line 16, in
import nodes
File "C:\comfy\ComfyUI\nodes.py", line 24, in
import comfy.diffusers_load
File "C:\comfy\ComfyUI\comfy\diffusers_load.py", line 3, in
import comfy.sd
File "C:\comfy\ComfyUI\comfy\sd.py", line 11, in
from .ldm.cascade.stage_c_coder import StageC_coder
File "C:\comfy\ComfyUI\comfy\ldm\cascade\stage_c_coder.py", line 19, in
import torchvision
File "C:\comfy\python312\Lib\site-packages\torchvision\__init__.py", line 10, in
from torchvision import _meta_registrations, datasets, io, models, ops, transforms, utils # usort:skip
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\comfy\python312\Lib\site-packages\torchvision\_meta_registrations.py", line 163, in
@torch.library.register_fake("torchvision::nms")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\comfy\python312\Lib\site-packages\torch\library.py", line 1063, in register
use_lib._register_fake(
File "C:\comfy\python312\Lib\site-packages\torch\library.py", line 211, in _register_fake
handle = entry.fake_impl.register(
^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\comfy\python312\Lib\site-packages\torch\_library\fake_impl.py", line 50, in register
if torch._C._dispatch_has_kernel_for_dispatch_key(self.qualname, "Meta"):
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: operator torchvision::nms does not exist
[W1028 21:53:38.000000000 AllocatorConfig.cpp:28] Warning: PYTORCH_CUDA_ALLOC_CONF is deprecated, use PYTORCH_ALLOC_CONF instead (function operator ())
```

### Other

If you need sth else, please let me know