issues
search
intel
/
auto-round
Advanced Quantization Algorithm for LLMs. This is official implementation of "Optimize Weight Rounding via Signed Gradient Descent for the Quantization of LLMs"
https://arxiv.org/abs/2309.05516
Apache License 2.0
247
stars
20
forks
source link
eval for MLLMs
#296
Closed
n1ck-guo
closed
1 week ago