auto-round
aa967f87 - Refactor NVFP4 E5M3 quantization to use "nvfp4_v2" data type

Commit
21 days ago
Refactor NVFP4 E5M3 quantization to use "nvfp4_v2" data type - Updated references from "fp4_v2" to "nvfp4_v2" across multiple modules including model_free, nvfp, and export functionalities. - Modified documentation to reflect the new data type and its properties. - Adjusted unit tests to ensure compatibility with the new naming convention. - Ensured backward compatibility by updating relevant configurations and checks. Signed-off-by: Xin He <xin3.he@intel.com>
Author
Parents
Loading