A mixed precision NVFP4 quantized version of the GLM-4.7-FLASH model is now available on Hugging Face. The announcement was made by user DataGOGO, who invites the community to test the model and provide feedback on its performance. ## Model Details * **Model:** GLM-4.7-FLASH NVFP4 * **Size:** 20.5 GB * **Availability:** Hugging Face ([https://huggingface.co/GadflyII/GLM-4.7-Flash-NVFP4](https://huggingface.co/GadflyII/GLM-4.7-Flash-NVFP4)) NVFP4 quantization is a technique that reduces the model size and potentially increases its inference speed, while minimizing the loss of accuracy. This initiative aims to make large language models more accessible and usable on hardware with limited resources. User feedback is crucial to assess the effectiveness of this implementation.