-
## 🐛 Bug
The return of torch.inverse contains nan sometime.
```
AssertionError: tensor([[[ nan, -0.8982, -0.0000],
[ nan, -0.4397, 0.0000],
[ nan, 0.0000, 1.00…
-
## 🐛 Bug
## To Reproduce
Steps to reproduce the behavior:
1. Create tensor
2. Create new_full with requires_grad=True
3. New full tensor should require gradient as per [documentation](h…
-
Hello! Thank you for your awesome work!
May I have a look at your conda list?
-
Hello,
very interesting paper and promising results!
Could you please provide test data for `eval.py.`
I want to try out your matting network.
Kind regards,
Vladimir.
-
## 🐛 Bug
I was trying to run a retinaface model in libtorch which this error popped up :
```python
isTensor() INTERNAL ASSERT FAILED at "C:\\w\\b\\windows\\pytorch\\aten\\src\\ATen/core/ivalue_inl…
-
when i download the dataset and run the process
2021-11-26 12:45:58.018258: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0
E…
-
In my ubuntu sever, pip is ok, but torchnet is only in 'pip list' and not in 'conda list', so when i run xx.py, error is no module named 'torchnet'. So i want to ask if i can use conda to install tor…
-
### Checklist
- I have searched related issues but cannot get the expected help.
- I have read related documents and don't know what to do.
### Describe the question you meet
I need help to ensu…
-
## 🐛 Bug
I pruned a resnet18 model and saved it as jit so it can be used in libtorch. The model is pruned and trained using Pytorch 1.5.1 and Python 3.7 under linux. Everything seems fine, the mode…
-
## 🐛 Bug
This is related to https://github.com/pytorch/pytorch/issues/41641. BatchNorm layer does not get fused while based on the comment :
> We do support quantized batch_norm so nn.BatchNorm…