-
## Current issue
- [ ] Infeasible to merge multiple views
- [ ] Cannot support multiple views (e.g. [A, B] -> View -> Shape cannot be inferenced)
- [ ] Parallelize Attention QKV Projection
…
-
### Qubes OS release
4.1
### Brief summary
On recent linux kernel (5.6+), the driver "amdgpu" start using MSI to do power management at runtime. In later kernel version (5.10+) also seems MSI see…
-
Hi @Dad0u,
Try to use GPUCorrel and get the error bellow, any idea ?
```shell
nvcc compilation of /tmp/tmpfapg_rx8/kernel.cu failed
[command: nvcc --cubin -arch sm_86 -I/home/jeff/.local/lib/py…
-
Since kernel 6.6 (?), the gpio pins have an offset of 512, so the correct number should be **537**, now.
Cf. this bug report concerning the [GPIO library](https://github.com/raspberrypi/linux/issue…
-
I use `pip install autoawq-kernels`to download the library.
![image](https://github.com/casper-hansen/AutoAWQ_kernels/assets/125335633/a951144a-858a-4d43-810c-f84e2c872049)
However, the error still …
-
With the release of Ubuntu 24.04 LTS, which comes with linux 6.5, and with more modern hardware getting on the market there is a need for datto to have support for the 6.x kernels.
I am currently in …
-
### Describe the bug
I'm integrating RPi Linux into NixOS. I've seen the announcement that RPi OS is receiving Linux 6.6, and I've seen that the default branch of this repository is now `rpi-6.6.y`. …
-
### Motivation.
Currently vLLM generally has a tight coupling between the checkpoint format and the kernel used during model execution. This model causes issues as the diversity of hardware and ker…
-
Hey!
I'm a big fan of the flash attention varlen kernels, and they are fantastic for saving the memory & compute of pad tokens.
When training with fixed batches of N tokens, I've noticed that th…
-
Since the introduction of mixed-precision fp16-int4 [MARLIN](https://github.com/IST-DASLab/marlin) (Mixed Auto-Regressive Linear) kernels by IST-DASLab, new mixed-precision MARLIN kernels have been in…