Merge pull request #32 from Vort3xed/vulkan-q2_0-kernel

vulkan: Q2_0

release-prism: install spirv-headers for ubuntu-arm64 vulkan build

Remove Windows CUDA 12.8 (not supported by setup action)

fix windows-hip artifact path

Merge pull request #8 from PrismML-Eng/cpu-fixes

some cpu fixes; getting ready for upstream PR; e.g. id 40 is taken by…

add slim release workflow for prism

add slim release workflow for prism

Merge pull request #1 from PrismML-Eng/mmq

[cuda] Fix mmq/mma path

ggml: add Q1_0 and Q1_0_g128 1-bit quantization support (CPU, Metal, …

…CUDA)

Adds two 1-bit quantization types:
- Q1_0: block size 32, ~1.5 bpw
- Q1_0_g128: block size 128, ~1.125 bpw

Backend support: CPU (x86 SSE/AVX + ARM NEON), Metal, CUDA.
Kernel implementations follow Q4_0 as boilerplate, adapted for
1-bit sign-based dequantization.

CUDA MMQ kernels included but disabled (cuBLAS fallback used for
prompt processing) pending accuracy debugging.

Made-with: Cursor

ggml: add Q1_0 and Q1_0_g128 1-bit quantization support (CPU, Metal, …

…CUDA)

Adds two 1-bit quantization types:
- Q1_0: block size 32, ~1.5 bpw
- Q1_0_g128: block size 128, ~1.125 bpw

Backend support: CPU (x86 SSE/AVX + ARM NEON), Metal, CUDA.
Kernel implementations follow Q4_0 as boilerplate, adapted for
1-bit sign-based dequantization.

CUDA MMQ kernels included but disabled (cuBLAS fallback used for
prompt processing) pending accuracy debugging.

Made-with: Cursor