求救！ 大神帮忙 用不了！

windy

The above exception was the direct cause of the following exception:Traceback (most recent call last):  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\queueing.py", line 407, in call_prediction    output = await route_utils.call_process_api(  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\route_utils.py", line 226, in call_process_api    output = await app.get_blocks().process_api(  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\blocks.py", line 1550, in process_api    result = await self.call_function(  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\blocks.py", line 1199, in call_function    prediction = await utils.async_iteration(iterator)  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\utils.py", line 519, in async_iteration    return await iterator.__anext__()  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\utils.py", line 512, in __anext__    return await anyio.to_thread.run_sync(  File "D:\facechain2.0.0\python310\lib\site-packages\anyio\to_thread.py", line 56, in run_sync    return await get_async_backend().run_sync_in_worker_thread(  File "D:\facechain2.0.0\python310\lib\site-packages\anyio\_backends\_asyncio.py", line 2144, in run_sync_in_worker_thread    return await future  File "D:\facechain2.0.0\python310\lib\site-packages\anyio\_backends\_asyncio.py", line 851, in run    result = context.run(func, *args)  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\utils.py", line 495, in run_sync_iterator_async    return next(iterator)  File "D:\facechain2.0.0\python310\lib\site-packages\gradio\utils.py", line 649, in gen_wrapper    yield from f(*args, **kwargs)  File "D:\facechain2.0.0\app.py", line 351, in launch_pipeline    outputs = future.result()  File "D:\facechain2.0.0\python310\lib\concurrent\futures\_base.py", line 451, in result    return self.__get_result()  File "D:\facechain2.0.0\python310\lib\concurrent\futures\_base.py", line 403, in __get_result    raise self._exceptionRuntimeError: CUDA error: no kernel image is available for execution on the deviceCUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.For debugging consider passing CUDA_LAUNCH_BLOCKING=1.Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

windy

这个错误信息是来自Python的并行计算库PyTorch，具体错误为RuntimeError: CUDA error: no kernel image is available for execution on the device，表明CUDA在尝试执行内核时遇到了问题，因为没有合适的内核映像可用于当前的设备。
以下列出了可能的解决办法：
检查CUDA版本与PyTorch版本的兼容性：确保你安装的PyTorch版本与CUDA版本相兼容。可以去PyTorch官网查看支持的CUDA版本，并执行适当的更新。
更新显卡驱动：有可能是因为显卡驱动版本过低造成的，可以尝试更新显卡驱动到最新版本。
设置环境变量：错误提示中提到“For debugging consider passing CUDA_LAUNCH_BLOCKING=1”。你可以尝试设置环境变量CUDA_LAUNCH_BLOCKING=1，这将使得CUDA调用变为阻塞操作，便于调试。
检查GPU兼容性：确保你的GPU硬件支持你所使用的CUDA版本。一些较旧的GPU可能不支持最新的CUDA版本。
**设置TORCH_USE_CUDA_DSA**：另外，可以设置环境变量TORCH_USE_CUDA_DSA来开启device-side断言，以便获得更多的调试信息。
你可以根据这些建议逐一排查问题。在实践中，编译相匹配的PyTorch和CUDA版本常常是必须的，确保它们之间相互兼容并使用相应驱动版本。如果你不确定如何操作，可以参考PyTorch和CUDA的官方文档进行配置。

maomao5441

支持一下