Enable kwargs for service post("/model/train")

[lsgome]

Description of Problem:
Currently the http / model / train service does not allow you to define extra parameters that exist on the command line, for example the augmentation_factor.

Overview of the Solution:
Change the / model / train service to include in the request body a dictionary where the kwargs parameter can be set. Pass the parameter to train_async.

Examples (if relevant):
Request:
{
"domain": ...
"config": ...
"nlu": ...
"kwargs": ...
}

Definition of Done:
Test training by changing the augmentation_factor parameter

[sara-tagger]

Thanks for submitting this feature request 🚀@msamogh will get back to you about it soon!✨

[wochinge]

Hi @lsgome , sounds like a great idea! Do you want to work on it?

[lsgome]

Yes, I can do.

[wochinge]

Amazing @lsgome ! Let me know if you need any help! I assign you for now so that nobody starts working on it at the same time, but feel free to un-assign you in case you change your mind.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[theanht1]

@wochinge does anyone take this issue? I could work for this enhancement

[lsgome]

no, you can work on this. Tks a lot!

…

On Wed, Mar 3, 2021 at 4:37 AM Anh Mai ***@***.***> wrote: @wochinge <https://github.com/wochinge> does anyone take this issue? I could work for this enhancement — You are receiving this because you were assigned. Reply to this email directly, view it on GitHub <#4596 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIVQDP6KFSKYAZVDIVF5KYTTBXRNFANCNFSM4I7SDRVA> .

[wochinge]

Feel free to start working on it @theanht1 You can also share an early draft in case you need feedback / help. Just tag me in the PR then 🚀

[m-vdb]

Fixed in #8132