automl / automl/ConfigurableOptimizer
Handle configuring nested configs in the Profiles
Open
- Dominant language
- Python
- Stars
- 12
- Forks
- 2
- PR merge metrics
- No merged PRs in 30d
Description
Currently, if we use `discrete_profile.configure_trainer(...)` to configure the `optim_config`, the entire dictionary is replaced with the new dictionary, instead of merging them.
This applies for supernet profiles, too.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.