From a091d03fb7630154dfeee41f99a425760eef7d37 Mon Sep 17 00:00:00 2001 From: SHYuanBest Date: Fri, 27 Dec 2024 15:27:06 +0800 Subject: [PATCH] update consisid readme --- TeaCache4ConsisID/README.md | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/TeaCache4ConsisID/README.md b/TeaCache4ConsisID/README.md index 857b53a..8b66fa1 100644 --- a/TeaCache4ConsisID/README.md +++ b/TeaCache4ConsisID/README.md @@ -1,7 +1,9 @@ # TeaCache4ConsisID -[TeaCache](https://github.com/LiewFeng/TeaCache) can speedup [ConsisID](https://github.com/PKU-YuanGroup/ConsisID) 2x without much visual quality degradation, in a training-free manner. +[TeaCache](https://github.com/LiewFeng/TeaCache) can speedup [ConsisID](https://github.com/PKU-YuanGroup/ConsisID) 2x without much visual quality degradation, in a training-free manner. The following video shows the videos generated by TeaCache-ConsisID with various `rel_l1_thresh` values: 0 (1.6x speedup), 0.1 (1.6x speedup), 0.15 (1.6x speedup), and 2.0 (1.6x speedup). + + ## 📈 Inference Latency Comparisons on a Single H100 GPU @@ -53,4 +55,4 @@ If you find TeaCache is useful in your research or applications, please consider ## Acknowledgements -We would like to thank the contributors to the [ConsisID](https://github.com/PKU-YuanGroup/ConsisID). +We would like to thank the contributors to the [ConsisID](https://github.com/PKU-YuanGroup/ConsisID). \ No newline at end of file