Closed SWHL closed 1 week ago
该问题查看源码,当前逻辑是:如果不显式指定infer_num_threads
,则不通过inference.set_cpu_math_library_num_threads
来设置,采用Config初始化的默认值。@vivienfanghuagood
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.
This issue was closed because it has been stalled for 5 days with no activity.
你好,我实际测试下来,发现CPU上rapidocr_onnxruntime性能比rapidocr_paddle好,但是这个差异主要是默认的cpu_math_library_num_threads没有设置导致的。具体原因是:以下这个分支里的设置线程数没有跑进去,导致paddle CPU推理的默认性能很差。
希望把默认的infer_num_threads这块改一下。
Originally posted by @vivienfanghuagood in https://github.com/RapidAI/RapidOCR/discussions/196#discussioncomment-10405131