-
### 🐛 Describe the bug
Consider a simple example:
```python
import torch
a = torch.arange(16).view(4, 4)
a[:3, :3] = a[:3, :3].T
b = torch.arange(16, device='cuda').view(4, 4)
b[:3, :3] =…
-
### 🐛 Describe the bug
JIT tracing a quantized model that has forward_pre_hooks throws the following error:
`RuntimeError: Couldn't find method: 'forward' on class: '__torch__.torch.ao.nn.intri…
-
# 🐛 Bug
Reading Cuda tensors from multiprocessing queue causes child (reader) process to hang.
I discovered that process hangs on any operation (sum, min, max, mean, etc.), in my example code on "…
-
### 🐛 Describe the bug
Hi, I think I found a bug in torchvision's write_video.
When reading a video (with audio) and writing it again without any modifications, the number of audio frames differs…
-
### 🐛 Describe the bug
When trying to compile a model with the QNN partitioner with the GPU or DSP i get the following error:
```
[ERROR] [Qnn ExecuTorch]: Cannot Open QNN library libQnnDsp.so, w…
-
I am trying to use zenpower. When I install it, and use `sensors` or `psensor`, I do not get any data about the CPU temperature. In my attempts to use zenmonitor, I get an error saying Zen CPU Not Det…
-
### 🐛 Describe the bug
We recently try to generate high resolution contents using diffusion models but notice unusual memory usage when resolution becoms large
Shown in above, the memory usa…
-
### 🐛 Describe the bug
The delay between inference request will impact the performance of latency.
If the delay is 0 second, the performance of bs=1 is best.
If the delay is 2s, the performance…
-
### 🐛 Describe the bug
Hello, I am trying to create a straight through estimator for fp8 type casting function, but it fails with message:
"RuntimeError: "fill_out" not implemented for 'Float8_e5m2'…
-
### 🐛 Describe the bug
The following code:
```python
import torch
from torch.profiler import ProfilerActivity, profile, record_function, tensorboard_trace_handler
DEVICE = "cuda:1"
def…