llama.cpp b9804 release: Mamba2 fixes and new binaries

The llama.cpp b9804 release introduces a fix for the Mamba2 architecture by removing a hardcoded 2x expansion factor and an invalid parameter check, allowing support for any expand value. This change updates the `convert_hf_to_gguf.py` script to make the expand parameter optional with a default of 2.

The release provides binaries for macOS (Apple Silicon and Intel), iOS, Linux (Ubuntu x64, arm64, s390x), Android (arm64), Windows (CPU, CUDA 12/13, Vulkan, OpenCL, ROCm, SYCL, OpenVINO), and openEuler.

This update ensures broader compatibility for Mamba2 models by removing restrictive configuration checks while maintaining extensive platform support across CPU and GPU backends.