-
**Overview**
I am exploring some techniques for saving GPU memory during training, including [Fairscale's `OffloadModel`](https://github.com/facebookresearch/fairscale/blob/main/fairscale/experimenta…
-
**Describe the bug**
`torchinfo.summary` does not report the correct number of trainable parameters.
**To Reproduce**
```python
import torch
from torch import nn, Tensor
from torchinfo imp…
-
### 🐛 Describe the bug
Nvidia driver crash if number of elements argument of torch.topk > 60000000
```
x = torch.randn((50, 1000,1000),device='cuda')
torch.topk (x,10,-1) # Works
x = torch.…
-
### 🐛 Describe the bug
```python
import torch
A = torch.randn(5, 5, device=device, requires_grad=True) # 5×5
t = torch.rand(7, 3, device=device) # 7×3
At = torch.einsum("ij…
-
### 🐛 Describe the bug
Hi, torch seems to run into an exception when I run `torchvision.transforms.functional.rotate` on a 5D tensor.
```
>>> import numpy as np
>>> import torch
>>> from torchvis…
-
**Describe the bug**
`nn.Parameter` is omitted in `summary` when there are other pytorch predefined layers in the networks.
Details are as follows:
**To Reproduce**
```python
import torch
imp…
-
**Describe the bug**
If using both `nn.Parameter` and modules, the `nn.Parameter` doesn't show up.
However, if not using the module, the `nn.Parameter` appears.
**To Reproduce**
Example 1:
```p…
-
**Describe the bug**
`summary` wrongly reports modules as recursive and several entries are duplicated and not in any logical order.
**To Reproduce**
Steps to reproduce the behavior:
1. `conda c…
-
Simply rerunning `pytest --cov=mypkg tests/`, changing nothing else, fixes it, and it's now the second time this randomly popped up. Affected 200+ lines out of 800, within just one file.
Python…
-
![image](https://user-images.githubusercontent.com/4758917/168480439-eda81859-41cf-4ae9-a231-561bd9266763.png)
Maybe I supply the wrong input_size? My model:
![image](https://user-images.githubu…