How can I save and load trained model?

Question

How can I save and load trained model?

Mekako opened this issue a year ago · 4 comments

I have trained a RollingStandardDeviationDetector and saved it.

rsd = RollingStandardDeviationDetector ()
rsd.fit(data)
rsd.save('./test.model')

How Can I load it from the 'test.model'?
There are only 4 functions for it : fit, detect, validateand save.
Looking forward to reply!

Answer 1 · 2023-09-15T01:32:45.000Z

Btw, https://ts-outlier-detection.streamlit.app/ is broken.

Answer 2 · 2023-09-15T06:08:04.000Z

I have trained a RollingStandardDeviationDetector and saved it.
rsd = RollingStandardDeviationDetector ()
rsd.fit(data)
rsd.save('./test.model')
How Can I load it from the 'test.model'? There are only 4 functions for it : fit, detect, validateand save. Looking forward to reply!

It should be as simple as tsod.load https://dhi.github.io/tsod/getting_started.html#saving-and-loading, let us know if that is not the case.

Answer 3 · 2023-09-15T06:18:35.000Z

Btw, https://ts-outlier-detection.streamlit.app/ is broken.

Thanks for letting us know, at the moment we are not actively developing this app, but you can try out this application either by running it locally using Docker or in GitHub Codespaces.

The Dockerfile should tell you how to run it.

Answer 4 · 2023-09-15T08:34:33.000Z

I have trained a RollingStandardDeviationDetector and saved it.
rsd = RollingStandardDeviationDetector ()
rsd.fit(data)
rsd.save('./test.model')
How Can I load it from the 'test.model'? There are only 4 functions for it : fit, detect, validateand save. Looking forward to reply!
It should be as simple as tsod.load https://dhi.github.io/tsod/getting_started.html#saving-and-loading, let us know if that is not the case.

I use joblib.load() and it works well.
Thanks for replying!