"Assertion Error" upon generation

[neggy555]

Looks like I'm the first one with this issue, I am getting this error code upon generating and not sure what it means:

Traceback (most recent call last): File "C:\Users\joeyb\Downloads\YUME\webapp_single_gpu.py", line 1444, in api_generate_long out_path, final_prompt = long_generate(g) ^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\webapp_single_gpu.py", line 806, in long_generate noise_pred = transformer(latent_model_input, t=tvec,latent_frame_zero = latent_frame_zero, **arg_c)[0] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\wan23\modules\model.py", line 852, in forward x = block(x, **kwargs) ^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\wan23\modules\model.py", line 300, in forward y = self.self_attn( ^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1736, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1747, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\wan23\modules\model.py", line 197, in forward x = flash_attention( ^^^^^^^^^^^^^^^^ File "C:\Users\joeyb\Downloads\YUME\wan23\modules\attention.py", line 112, in flash_attention assert FLASH_ATTN_2_AVAILABLE ^^^^^^^^^^^^^^^^^^^^^^ AssertionError

[stdstu12]

It seems that flash_attention failed to install on your computer. You can try installing it manually by finding the version that exactly matches your Python, PyTorch, and CUDA versions.