I need help " TypeError: can’t convert CUDA tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first "

projects
1

i trying to run this plot but is said error: can’t convert CUDA tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first , anyone know why?

def plot_training_history(history):
  fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(18, 6))

  ax1.plot(history['train_loss'], label='train loss')
  ax1.plot(history['val_loss'], label='validation loss')

  ax1.xaxis.set_major_locator(MaxNLocator(integer=True))
  ax1.set_ylim([-0.05, 1.05])
  ax1.legend()
  ax1.set_ylabel('Loss')
  ax1.set_xlabel('Epoch')

  ax2.plot(history['train_acc'], label='train accuracy')
  ax2.plot(history['val_acc'], label='validation accuracy')

  ax2.xaxis.set_major_locator(MaxNLocator(integer=True))
  ax2.set_ylim([-0.05, 1.05])
  ax2.legend()

  ax2.set_ylabel('Accuracy')
  ax2.set_xlabel('Epoch')

  fig.suptitle('Training history')
2

Some aspects of matplotlib will try to convert the input you provide it into numpy. It can do it only if the provided input tensor is in cpu.

So, to begin with I would suggest you to try and do .cpu() to the inputs that you provide, that is, check if every value in the dict history that you are using is on cpu.