You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Dec 1, 2021. It is now read-only.
To generate config.py, it duplicated the input python config file, while config.yaml is generated from the config dictionary.
It's possible that when we overwrite some variables, the config.py will not get update, but config.yaml will give us the correct configuration variables.
In train.py, it happens immediately after loading the config file. So I think there is no problem to not save the config.yaml there.
In export.py, it loads the config file, then overrides some things like BATCH_SIZE and PRE_PROCESSOR, before it then saves as config.yaml. But, after that, is the config.yaml used for anything? I understand that meta.yaml could be used for application such as demo. But I'm not sure about config.yaml.
The problem with not removing the config.yaml is that currently it can cause error when loading #1110
Currently, we are saving
config.yaml
the contents is same as Python config.IMO, nobody are using
config.yaml
.https://github.com/blue-oil/blueoil/blob/master/blueoil/utils/config.py#L214-L236
The text was updated successfully, but these errors were encountered: