-
Notifications
You must be signed in to change notification settings - Fork 2.2k
Open
Description
Describe the bug(问题描述)
A clear and concise description of what the bug is.Better with standalone code to reproduce the issue.
在使用 deepctr 库中的 DeepFM 模型时,调用 predict 方法会出现内存持续增长的问题,导致内存泄漏。该问题在重复预测(如批量处理或 API 服务)时尤为明显。
To Reproduce(复现步骤)
Steps to reproduce the behavior:
-
安装对应版本依赖
-
model训练,重复调用预测
-
观察内存使用:使用 tracemalloc 或 psutil 监控内存,也可以from memory_profiler import profile看每一个语句运行后内存增加状况
-
See error
Operating environment(运行环境):
- python version [3.8.16]
- tensorflow version [2.10.0]
- deepctr version [0.9.3]
Additional context
Add any other context about the problem here.
Metadata
Metadata
Assignees
Labels
No labels