Source of problem:
Recently, in the field of deep learning in learning machine learning, certain services are required to provide computing power to run models. Renting machines from server manufacturers such as Amazon can be said to be an ideal training environment, but Amazon’s servers are expensive, so you can choose to download and install CUDA and cudaa in the local window system. There are nanny-level tutorials on how to download. Teacher Mu's deep learning teaching how to install the GPU version locally: Install CUDA and Pytorch under Windows to run deep learning-hands-on deep learning v2_哔哩哔哩_bilibili
Solution:
When I followed Mr. Li Mu's video to download cuda and check whether the download was successful, I was stuck in the error report shown in the title, and it was useless to consult a large amount of information. In the end, I removed the habitual space between the command and the parameter, and found that it ran magically. The code is as follows:
Before the modification, an error was reported:
nvidia-smi
After modification, success:
nvidia-smi
Hope this blog is helpful to you, thanks.