pytorch

error:

only integer tensors of a single element can be converted to an index

userEmbeddings = self.userEmbeds[userIndex]
movieEmbeddings = self.movieEmbeds[movieIndex]

To -->

userEmbeddings = self.userEmbeds(userIndex)
movieEmbeddings = self.movieEmbeds(movieIndex)

 can't covert CUDA tensor to numpy. use Tensor.cpu() to copy the tensor to host memory first.

                           Change

        index = output.data.numpy().argmax()

                             to

        index = output.cpu().data.numpy().argmax()

指定GPU运行python程序

一、命令行运行python程序时

1、首先查看哪些GPU空闲，nvidia-smi显示当前GPU使用情况。

1	nvidia-smi

2、然后指定空闲的GPU运行python程序。

1	CUDA_VISIBLE_DEVICES=0,2,3 python test.py

二、在python程序中指定GPU

1 2	import os os.environ["CUDA_VISIBLE_DEVICES"] = "0,2,3"

周期性输出GPU使用情况

每 10s 显示一次GPU使用情况：

1	watch -n 10 nvidia-smi

convert list to tensor:     torch.Tensor(a)