怀疑是因为安装了和cuda不匹配的torch 导致的，本地的ubuntu是cuda 11.0,安装了cuda 11.3.
只能删除虚拟环境，重新安装，根据提示逐步安装 pip install torch
免责声明！
本站转载的文章为个人学习借鉴使用，本站对版权不负任何法律责任。如果侵犯了您的隐私权益，请联系本站邮箱yoyou2525@163.com删除。
猜您在找
import pandas出现‘ValueError: numpy.ufunc size changed, may indicate binary incompatibility. Expected 216 from C header, got 192 from PyObject’问题
安装Python3.6.4后，在使用numpy时报错RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88
numpy.ufunc size changed, may indicate binary incompatibility错误解决
Python：Expected 88 from C header, got 80 from PyObjec
解决Tensorflow ValueError: Failed to convert a NumPy array to a Tensor (Unsupported object type numpy.ndarray)
pytorch报错：ValueError: Expected more than 1 value per channel when training, got input size torch.Size([1,512,1,1])
将numpy.ndarray写入excel
numpy.ndarray的赋值操作
TypeError: unhashable type: 'numpy.ndarray'
ValueError:Expected more than 1 value per channel when training, got input size torch.Size([1, 512, 1, 1])