-
OS version: macOS 10.13.6
Cuda version: 10.1
GPU: GTX 1060
`[1/893] Linking CXX shared library lib/libtorch_cpu.dylib
FAILED: lib/libtorch_cpu.dylib
: && /Library/Developer/CommandLineTools…
-
First - this is great. Thank you. Opens up so many possibilities. I made it through Into the Storm so far. Below is general feedback.
If you would like me to create a separate issue for any of…
-
Platforms: rocm
This test was disabled because it is failing on main branch ([recent examples](http://torch-ci.com/failure/inductor%2Ftest_snode_runtime.py%3A%3AComputeBoundedTests%3A%3Atest_bmm)).…
-
### Description
…
-
### 🐛 Describe the bug
For the following model, the result optimized by `torch.compile` on CUDA is totally wrong, which may out-of-bound access some data.
```py
import torch
torch.manual_seed(…
-
### 🐛 Describe the bug
We found there are negative numbers in PyTorch profilings, which is inconvenient for users to get solid profiling for operators:
```python
from transformers import AutoMode…
-
-
(update) as I mentioned in this [comment](https://github.com/tenstorrent-metal/tt-metal/issues/5168#issuecomment-1940293930), this issue is related to the TT_METAL_DEVICE_PROFILER env variable.
*…
-
[24.08.2023 12:07:39] jasiu-rozowyczlowiek:
[24.08.2023 12:07:39] jasiu-rozowyczlowiek
-
When trying to convert an openai clip model to neuron everything works fine if using a single batch size, but when switching to dynamic batch size it crashes when doing inference on a batch size diff…