-
The kernel size can not be larger than 13x13? I set the kernel size to 15x15 and encounter the error "CUDA error: an illegal memory access was encountered". But everything is ok when the kernel size i…
-
- It's hard to assemble `methods` and `results` when we are developing incrementally
---
### (x) Chapter `Methods` then `Results`?
- If we use the traditional way of **thoroughly dividing meth…
-
OpenCL transpose breaks with matrices of rank 1296 or greater with the NVIDIA OpenCL implementation. This is NVIDIA-specific, because the Intel OpenCL is fine for much larger matrices.
It is possi…
-
## Detailed Description
https://github.com/facebookincubator/AITemplate
This is a new thing for PyTorch that compiles the models down to CUDA kernels for fast inference. There are still some c…
-
There's couple issues with exporting weights from darknet:
- the darknet fork is broken, there are unmerged changes in [convolution_kernels.cu](https://git.hipert.unimore.it/fgatti/darknet/-/blob/m…
-
### Description
AMReX is a C++ library and uses (for historic reasons) F-based indexing. While writing CUDA Array Interface bindings for AMReX, we noticed the following interesting behavior.
We …
ax3l updated
5 months ago
-
### Why/User Benefit/User Problem
Very difficult to switch to, and *use*, a System76 machine after using MacBooks for a decade.
There are those that think a trackpad just moves the mouse. Then …
-
inspired by GitHub issue https://github.com/apache/arrow/issues/5874
**Reporter**: [Wes McKinney](https://issues.apache.org/jira/browse/ARROW-7245) / @wesm
#### Related issues:
- [[C++] C++ array ker…
-
If you getting `EOVERFLOW` out of nowhere - be aware of this line:
https://github.com/bm16ton/ft2232-mpsse-i2c-spi-kern-drivers/blob/923b7059795bcba5d82a3ae0a9fc7de93c20df03/i2c-ftdi/i2c-ftdi.c#L45…
-
### Issue type
Bug
### Have you reproduced the bug with TensorFlow Nightly?
Yes
### Source
source
### TensorFlow version
tf 2.16.1
### Custom code
Yes
### OS platform and distribution
Ubunt…