Yaziwel
commented
1 month ago 感谢您的开源，我尝试运行代码，单卡可以训练，但是多卡会有如下问题: Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. 请问您那边可以多卡训练吗？
可以多卡训练，你试试把/root/.cache/torch_extensions/py310_cu118/wkv这个缓存删掉
感谢您的开源，我尝试运行代码，单卡可以训练，但是多卡会有如下问题: Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. Loading extension module wkv... Using /root/.cache/torch_extensions/py310_cu118 as PyTorch extensions root... Detected CUDA files, patching ldflags Emitting ninja build file /root/.cache/torch_extensions/py310_cu118/wkv/build.ninja... Building extension module wkv... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) ninja: no work to do. 请问您那边可以多卡训练吗？