Closed hwanheelee1993 closed 3 years ago
In your code, you did not wrap the eval part with "torch.no_grad()", which means that your code computes gradient for validation data. That leads to unnecessary computation... You'd better edit that part.
Thanks! I updated the code and will close this
In your code, you did not wrap the eval part with "torch.no_grad()", which means that your code computes gradient for validation data. That leads to unnecessary computation... You'd better edit that part.