41 10 36

wenhua cheng

wenhuach

wenhuach21

AI & ML interests

Model Compression, CV

Recent Activity

updated a model less than a minute ago

wenhuach/test_mlx

published a model 2 minutes ago

wenhuach/test_mlx

reacted to theirpost with 🔥 about 20 hours ago

We demonstrate that AutoRound achieves SOTA or near SOTA performance under INT4 (W4A4) quantization. Check out the accuracy data at https://github.com/intel/auto-round/blob/main/docs/int4_acc.md This capability is currently a research-only feature, with no production model export.

View all activity

Organizations

Posts 16

Post

110

We demonstrate that AutoRound achieves SOTA or near SOTA performance under INT4 (W4A4) quantization.
Check out the accuracy data at https://github.com/intel/auto-round/blob/main/docs/int4_acc.md
This capability is currently a research-only feature, with no production model export.