pl752

Hello, I was reinstalling windows on that tablet and forgot to backup SileadTouch.fw, so can you, please, upload the original file from windows driver to flex 8 page, cause I...

AMDGPU is broken on 1.8

6

### Describe the bug llamacpp doesn't see radeon rx6900xt, previous version worked fine, it seems it has missing dependencies (rocm 5.7.1 is installed) in particular llama_cpp_cuda can not be imported...

Feature Request: Add kv-quant fa kernel variants for head sizes other than 128

1

### Prerequisites - [x] I am running the latest code. Mention the version if possible as well. - [x] I carefully followed the [README.md](https://github.com/ggml-org/llama.cpp/blob/master/README.md). - [x] I searched using keywords...

Eval: HIP: Llama-server multi-instance lockup

3

**Follow-up on the https://github.com/ggml-org/llama.cpp/issues/12991** According to rocgdb backtrace, threads that are working with gpus are stuck somewhere in the libhsa-runtime ``` Thread 2 (Thread 0x7fffd7fff6c0 (LWP 11602) "llama-server"): #0 __GI___ioctl...

(Improvements, need help testing) Rune related and span-ish optimizations

8

I want to propose set of changes aimed at improving performance, which I have implemented and used for some time in my (private) projects. The main goal of these changes...

I don't think that was how you intended to calculate offset there (the i starts with offset and is then added with offset again, also it performs byte writing count-offset...