-
```python
import torch
class TestModule(torch.nn.Module):
def forward(self, x):
return x + torch.rand_like(x)
trace_input = torch.zeros([1], device='cuda', dtype=torch.float32…
-
## 🐛 Bug
Torch.jit.trace fails if model slicing is used anywhere within the model functions.
Steps to reproduce the behavior:
Say you have VGG16 model like below -
```python3
class VGG1…
-
## 🐛 Bug
When tracing BCELoss with the optional weight parameter (see https://pytorch.org/docs/master/generated/torch.nn.BCELoss.html) it throws:
```Traceback (most recent call last):
File "…
-
## 🐛 Bug
There seems to be a type error when tracing a Graph-UNet with Torch JIT.
```
[omitted]/torch_sparse/matmul.py in spspmm(src, other, reduce)
94 if reduce == 'sum' or reduce == '…
-
Today I want to do a simple experiment to improve my mental model of how [Snabb Switch](https://github.com/SnabbCo/snabbswitch) code is just-in-time compiled into traces by [LuaJIT](http://luajit.org/…
-
#### Issue description
After updating jax and mlir dependency chain to v0.4.28 [(PR#931)](https://github.com/PennyLaneAI/catalyst/pull/931), jax introduces new `_sin_lowering` and `_cos_lowering` wit…
-
This is the reason we are so slow on the "Acid" benchmark presented here: http://www.chrisseaton.com/rubytruffle/pushing-pixels/
-
In Heptapod by @belm0 on Oct 25, 2022, 15:19
any "import all" seems to cause an abort. Perhaps this is just at module load time, so won't happen repeatedly and hurt performance significantly. In any…
-
While preparing the benchmark for eager and dynamo using the code from the fork: https://github.com/tfogal/NeMo I get errors for dynamo case.
## 🐛 Bug
After fixing [1187](https://github.com/Ligh…
-
## 🐛 Bug
I am calling `torch.jit.trace` on a model and I am getting the corresponding error:
```TracerWarning: Output nr 1. of the traced function does not match the corresponding output of the …