bugbug
bugbug copied to clipboard
mozilla
Platform for Machine Learning projects on Software Engineering
Suggested by @marco-c Depends on: https://github.com/mozilla/bugbug/issues/3876 See also: https://github.com/mozilla/bugbug/issues/3901
> Turns out that there is an issue; I tried to use the model to classify a certain bug but it didn't work. However, it seems like that there are...
> > I think using the `n_features_in_` attribute from the `base_clf` (after fitting) should solve this. > > Yes. Let's solve it that way for now. But at some point,...
See also https://github.com/mozilla/bugbug/issues/3372 Depends on https://github.com/mozilla/bugbug/issues/3903
Depends on https://github.com/mozilla/bugbug/issues/3903
The logic responsible for aggregating small components should live in a separate class for reusability, rather than being part of the `component` model. https://github.com/mozilla/bugbug/blob/c9fdb64529f6838aaead42b7157c206e13191612/bugbug/models/component.py#L130-L143
> I plan to simplify the train() method and remove unnecessary data duplicates where possible. _Originally posted by @suhaibmujahid in https://github.com/mozilla/bugbug/pull/3883#discussion_r1413236467_
Instead of having specialized feature extractor classes (e.g., `CommitExtractor`, `IssueExtractor` and `BugExtractor`), create a generic one and refactor the specialized logic to be as feature classes and/or pipeline steps.
``` 2023-12-01T14:26:38.680345+00:00 app[web.1]: 2023-12-01 14:26:38,679:ERROR:bugbug_http.app:Exception on /swagger [GET] 2023-12-01T14:26:38.680359+00:00 app[web.1]: Traceback (most recent call last): 2023-12-01T14:26:38.680368+00:00 app[web.1]: File "/usr/local/lib/python3.10/site-packages/flask/app.py", line 1455, in wsgi_app 2023-12-01T14:26:38.680368+00:00 app[web.1]: response = self.full_dispatch_request() 2023-12-01T14:26:38.680369+00:00 app[web.1]:...
