-
## š Bug
### To Reproduce
Steps to reproduce the behavior (**always include the command you ran**):
1. Run cmd python examples/speech_recognition/infer.py ~/fairseq/benchmark --task audiā¦
-
### š Describe the bug
Indicies with **only** null tensors causes an error or segfault.
This doesn't happen if `indices` contains other tensors that aren't null
```python
import torch
input =ā¦
-
https://arxiv.org/pdf/1711.07971.pdf
It improves >1% mAP for mask r-cnn, even with backbone X-152. It seems worth to implement.
-
```
# create a color pallette, selecting a color for each class
palette = torch.tensor([2 ** 25 - 1, 2 ** 15 - 1, 2 ** 21 - 1])
colors = torch.as_tensor([i for i in range(21)])[:, None] * palette
ā¦
-
### š Describe the bug
The codes below runs normally on the CPU, but there is a segmentation fault in running line `print(net(x).shape)`. Additionally, I tested a simple model without convolutional lā¦
-
Hi,
I'm using segnet to load a custom segmentation network and I was wondering if I can change precision to fp32?
Currently I am getting this when i run it
...
[TRT] desired precision specā¦
-
(yw_fast_bert_test)$ python Fast_Bert_model_fitting.py
09/13/2019 18:42:45 - INFO - pytorch_transformers.tokenization_utils - loading file https://s3.amazonaws.com/models.huggingface.co/bert/bert-bā¦
-
Based on the theoretical research check of "implementability" of methods (available code, built-in into pytorch, etc.).
Make sure to think ahead, i.e. how you can progress with training/pruning etā¦
-
A1111 v1.6
latest inpaint anything version: [4f0d8eba]
Model: sam_vit_h_4b8939.pth
nvidia driver version: 537.34
Windows 11
When just clicking on "Run Segment Anything" after adding a picture, ā¦
-
Background, dataloader slows down over time, especially when using a large number of slides; data that is persistent in memory loads quickly (case for very small number sslides), but not when trainingā¦