PaddleX/latest/en/module_usage/tutorials/time_series_modules/time_series_classification

[giscus[bot]]

PaddleX/latest/en/module_usage/tutorials/time_series_modules/time_series_classification

https://paddlepaddle.github.io/PaddleX/latest/en/module_usage/tutorials/time_series_modules/time_series_classification.html

[songmingjun3]

文档好像没有提到export模式的使用方式，使用以下命令进行导出
python main.py -c paddlex/configs/modules/ts_classification/Timesnet_cls.yaml -o Global.mode=export
出现以下错误
[2025-08-06 15:17:18,153] [paddlets] [ERROR] ValueError: failed to save model, path must be a file, not directory: D:\data\PycharmWorkspace\PaddleX\output\inference
Traceback (most recent call last):
File "tools/export.py", line 153, in
export(args)
File "tools/export.py", line 142, in export
model.save(
File "D:\data\PycharmWorkspace\PaddleX\paddlex\repo_manager\repos\PaddleTS\paddlets\models\classify\dl\paddle_base.py", line 721, in save
raise_if(
File "D:\data\PycharmWorkspace\PaddleX\paddlex\repo_manager\repos\PaddleTS\paddlets\logger\logger.py", line 153, in raise_if
raise_if_not(not condition, message, logger)
File "D:\data\PycharmWorkspace\PaddleX\paddlex\repo_manager\repos\PaddleTS\paddlets\logger\logger.py", line 133, in raise_if_not
raise ValueError(message)
ValueError: failed to save model, path must be a file, not directory: D:\data\PycharmWorkspace\PaddleX\output\inference
是否是配置方式有误，还是代码的bug？能否给个使用建议，另外train结束后会保存模型权重，这块是否和export逻辑相同，产物相同？