PaddleOCR
PaddleOCR copied to clipboard
paddle 开启GPU以后 速度比 use-gpu=false更慢
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment:win11 CUDA:12.1 python:3.8.9、paddleocr:2.7.3
- 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components:
- 运行指令/Command Code:
- 完整报错/Complete Error Message:
-
程序执行时间:5.861234664916992秒 Running verify PaddlePaddle program ... PaddlePaddle works well on 1 GPU. PaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now.
`
# 记录开始时间 start_time = time.time() orc = PaddleOCR(lang='ch',use_angle_cls=True,use_gpu=True) result = orc.ocr('images/sfz/03-1.png') print(result) end_time = time.time() execution_time = end_time - start_time print(f"程序执行时间:{execution_time}秒")
# 记录开始时间 start_time = time.time() orc = PaddleOCR(lang='ch',use_angle_cls=True,use_gpu=True) result = orc.ocr('images/sfz/03-1.png') print(result) end_time = time.time() execution_time = end_time - start_time print(f"程序执行时间:{execution_time}秒")
测一张图没法说明问题,需要批量测试。
因为将数据移动到显存内会消耗一定时间,建议批量测试后计算一下平均时间