-
-
It's a very nieche problem, but tripped me over big time :')
### Issue
For `model.eval()` , `z_pred` will not have tracked gradients (`z_pred.requires_gradient==False`).
For custom torch.autog…
-
Is there a recommended way to use nsys / nsight? I know there's a profiling hook for using the Pytorch profiler, but I'm wondering how to use nsys instead.
Can I use these APIs:
```
with torch.au…
-
Inplace operations (+=, -=, x[:]=, etc) are not supported when recording with autograd. But in object detection, we need to use inplace operations to calcuate regression target, which does not need gr…
-
#### Problem description
It looks like lighkurve calls functions in [oktopus](https://github.com/KeplerGO/oktopus), which depends on [autograd](https://github.com/HIPS/autograd), which makes depr…
-
### 🐛 Describe the bug
Good morning,
I have the following simple model:
```
import torch
class Model1(torch.nn.Module):
def __init__(self):
super(Model1, self).__init__()
…
-
# Reference
- [Autograd:自动求导](https://pytorch.org/tutorials/beginner/blitz/autograd_tutorial.html#sphx-glr-beginner-blitz-autograd-tutorial-py)
# Brief
```
x = torch.ones(2, 2, requires_grad=Tru…
-
I am interested in some of the functionality of torch.autograd.functional (https://github.com/pytorch/pytorch/blob/master/torch/autograd/functional.py) in Pytorch for R (especially `jvp()` and `jacobi…
-
First of all thank you for "mergeing" my pull request in a slightly edited way.
If i want to use the vectorized implementation it seems not to work.
The error message is:
```
Using Materials Pr…
-
For most of our operators an adjoint/derivative method will be defined. pytorch.autograd should use that. If the method is not defined it should default to standard pytorch.autgrad. More details can b…