llama.cpp b10251 adds MTP support for GLM-4.7-Flash

The latest release of llama.cpp (b10251) introduces Multi-Tenant Plugin (MTP) support for the GLM-4.7-Flash model, enhancing compatibility and performance in multi-user environments. This update is crucial for operators running local AI services on consumer hardware such as Apple Silicon or gaming GPUs, as it allows more efficient management of resources across multiple users or applications. The MTP feature ensures that each user's session can be isolated and optimized without affecting others, making the deployment of GLM-4.7-Flash models in shared setups smoother.