b7956

vulkan: For coopmat2 FA, use fp16 accumulators for the final result (#19376)

The cpu and cuda backends use fp16 for the VKQ accumulator type, this change does the same for vulkan. This helps particularly with large head sizes which are very register-limited.

I tried this for the coopmat1 path and it slowed down a bit. I didn't try for scalar.

I applied the softmax bias that the cuda backend uses to avoid overflow, although I was not able to reproduce the original bug without it.

macOS/iOS:

• macOS Apple Silicon (arm64)
• macOS Intel (x64)
• iOS XCFramework

Linux:

• Ubuntu x64 (CPU)
• Ubuntu x64 (Vulkan)
• Ubuntu s390x (CPU)

Windows:

• Windows x64 (CPU)
• Windows arm64 (CPU)
• Windows x64 (CUDA 12) - CUDA 12.4 DLLs
• Windows x64 (CUDA 13) - CUDA 13.1 DLLs
• Windows x64 (Vulkan)
• Windows x64 (SYCL)
• Windows x64 (HIP)

openEuler:

• openEuler x86 (310p)
• openEuler x86 (910b, ACL Graph)
• openEuler aarch64 (310p)
• openEuler aarch64 (910b, ACL Graph)