-
### 🐛 Describe the bug
Trying to do a sparse-dense matrix multiplication in with bfloat16 dtype produces incorrect output.
More specifically, I would like to use torch.sparse_csr_tensor to impleme…
-
### 🐛 Describe the bug
I am trying to export a PyTorch `nn.Module` subclass (the `CPnet.net` object in Cellpose) to ONNX but I am not able to do so.
You can reproduce the error by issuing the fo…
-
### 🐛 Describe the bug
Thanks @jjsjann123 for reporting.
```C++
TEST_F(NVFuserTest, TMP) {
auto fusion = std::make_unique();
FusionGuard fg(fusion.get());
auto tv0 = makeSymbolicTens…
-
### 🐛 Describe the bug
`parallelize_module` failed with `nn.Transformer` and the `PairwiseParallel` plan, which is unexpected according to [the doc](https://pytorch.org/docs/stable/distributed.tens…
-
## 🚀 Feature
Implementation of Grad-CAM, probably as a vision callback.
### Motivation
[Grad-CAM](https://arxiv.org/pdf/1610.02391.pdf) is widely used localization method that uses the gradient…
-
### 🐛 Describe the bug
I'm testing sd2.1 that batchsize in [1, 2, 4, 8]. it's fine when batchsize=1, and then it fails when batchsize switch to 2
after @BoyuanFeng fix, there is a new error
### Err…
-
## 🐛 Bug
**Describe the bug**
Hi,
I've scripted a Roberta model and when I do two inference calls on it, the second call returns a result only after several minutes (up to 16 minutes).
I see t…
mreso updated
2 years ago
-
### 🐛 Describe the bug
We have recently found out at NVIDIA that Triton's libtorch backend can't run TorchScript modules exported via torch.jit.trace() with amp.autocast.
There is still outstanding…
-
## 🐛 Bug
When using `nn.RNN`, I forgot to cast my output (from `nn.functional.one_hot`) to a float before passing into the RNN. However, I was prompted with the error message `RuntimeError: expec…
-
Could you please provide a code for the metrics used in the paper?
Thank you