Deprecation of tq1_0 quant support in ggml-org/llama.cpp

The latest release of ggml-org/llama.cpp, titled b10926, introduces a change that affects handling of unsupported tq1_0 quantization formats. According to the project's GitHub release notes, the system now fails gracefully when encountering tq1_0 quant formats, signaling the deprecation of support for this quantization type. This update impacts compatibility with certain configurations and may require adjustments in workflows relying on tq1_0 quants.

Operators should verify whether their current deployment or model training pipelines depend on tq1_0 quantization formats. If so, alternative quantization methods or configurations should be explored prior to upgrading. This change aligns with a broader pattern seen across projects where older quantization formats are phased out in favor of newer, more optimized approaches. Ensuring compatibility with supported formats will be critical to avoiding runtime errors or degraded performance.