Paddle Inference 3.0.0 beta.1 CPU-MKL C++ 推理速度比旧版本明显下降 #13549
Unanswered
hiroi-sora
asked this question in
Q&A
Replies: 1 comment 1 reply
-
这个是历史悠久的问题,应该是跟onednn相关,相关issue有: |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
我对
Paddle Inference 3.0.0 beta.1
C++ 推理后端进行了一些测试,不一定严谨。发现推理速度比旧版本有明显下降。测试结果:
测试环境:
PaddleOCR/deploy/cpp_infer
,使用单个 PPOCR 实例对所有图片进行推理。结论:
询问:
Beta Was this translation helpful? Give feedback.
All reactions