auto-round
5c0e8480
- Enhance E8M0 block scale expansion for DeepSeek variants and improve logging for fp32 to UE8M0 conversion
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
10 days ago
Enhance E8M0 block scale expansion for DeepSeek variants and improve logging for fp32 to UE8M0 conversion Signed-off-by: Xin He <xin3.he@intel.com>
References
xinhe/8-6
#2123 - [Experimental] Add NVFP4 E5M3 support and related components for model-free quantization
Author
xin3he
Parents
52adf875
Loading