## Update for GLM-4.7-Flash-GGUF An update has been released for the GLM-4.7-Flash-GGUF model that fixes a critical issue. Previously, a bug caused unexpected behavior, such as infinite loops and lower quality outputs. The engineers at llama.cpp have identified and corrected the root cause. To benefit from the improvements, you must redownload the updated model. This will ensure that your applications take advantage of the latest fixes and optimizations. ## Recommended Parameters Z.ai has provided recommended parameters for optimal performance with the model: * **General use:** `--temp 1.0 --top-p 0.95` * **Tool-calling:** `--temp 0.7 --top-p 1.0` * **Llama.cpp:** If using llama.cpp, set `--min-p 0.01` (llama.cpp's default is 0.1). The updated model is available for download on Hugging Face.