dotnet / dotnet/machinelearning

Request: expose Experiment._modelDirectory

Open
#4,952 1 comment 0 reactions 0 assignees View on GitHub
area-AutoML enhancement Priority:2
Dominant language
C#
Stars
9.4k
Forks
2k
Avg merge
2d 20h
Merged PRs (30d)
11

Description

Hi all,

In AutoML, when configuring the `ExperimentSettings`, one can set a `CacheDirectory` where the intermediate models will be stored. However, the models are actually stored in a subfolder, called "ExperimentXX", where XX in an auto-increment integer in v.0.14.0, and some random characters in the recent preview releases. Navigating the source code, I found that this path is stored inside the private readonly field `_modelDirectory` belonging to the internal class `Experiment`. When running many experiments, there is no way to associate each experiment to each subfolder.
In my scenario, caching is necessary due to dataset size, and I would like to delete the temporary models at the end of the experiment without bothering other experiments that may be running simultaneously.

Thanks

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.