How do you specify in the app.py to not use flash_attn ??

How do you specify in the app.py to not use flash_attn ??

I did not install flash_attn (because it's giving me an error) but it still tries to use it. How do I prevent this?

Error trying to install flash_attn:

(trellis) C:\inetpub\wwwroot\TRELLIS>pip install flash-attn
Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com
Collecting flash-attn
Downloading flash_attn-2.7.0.post2.tar.gz (2.7 MB)
---------------------------------------- 2.7/2.7 MB 8.5 MB/s eta 0:00:00
Preparing metadata (setup.py) ... error
error: subprocess-exited-with-error

× python setup.py egg_info did not run successfully.
│ exit code: 1
╰─> [22 lines of output]
fatal: not a git repository (or any of the parent directories): .git
Traceback (most recent call last):
File "", line 2, in
File "", line 34, in
File "C:\Users\jeffc\AppData\Local\Temp\pip-install-34xtqzwr\flash-attn_72812057375a47f494a81dc7ec330502\setup.py", line 162, in
_, bare_metal_version = get_cuda_bare_metal_version(CUDA_HOME)
File "C:\Users\jeffc\AppData\Local\Temp\pip-install-34xtqzwr\flash-attn_72812057375a47f494a81dc7ec330502\setup.py", line 82, in get_cuda_bare_metal_version
raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True)
File "C:\Users\jeffc\anaconda3\envs\trellis\lib\subprocess.py", line 421, in check_output
return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
File "C:\Users\jeffc\anaconda3\envs\trellis\lib\subprocess.py", line 503, in run
with Popen(*popenargs, **kwargs) as process:
File "C:\Users\jeffc\anaconda3\envs\trellis\lib\subprocess.py", line 971, in init
self._execute_child(args, executable, preexec_fn, close_fds,
File "C:\Users\jeffc\anaconda3\envs\trellis\lib\subprocess.py", line 1456, in _execute_child
hp, ht, pid, tid = _winapi.CreateProcess(executable, args,
FileNotFoundError: [WinError 2] The system cannot find the file specified

  torch.__version__  = 2.4.0


  [end of output]

Start by inspecting app.py and the dependency or import path that triggers flash_attn, then compare it with the reported Windows installation error from pip install flash-attn. Done should mean the application has a documented way to run without flash_attn, or the issue clearly identifies the supported installation requirements.