Skip to content
This repository has been archived by the owner on Sep 4, 2023. It is now read-only.

Add Ukrainian models #409

Merged
merged 4 commits into from
Jun 24, 2022
Merged

Add Ukrainian models #409

merged 4 commits into from
Jun 24, 2022

Conversation

eu9ene
Copy link
Collaborator

@eu9ene eu9ene commented Jun 16, 2022

The models were not quantized using browsermt marian (it produces so-called alpha models) and should be used with gemm-precision: int8shiftAll. See browsermt/marian-dev config docs for an explanation of the difference. This mode is ~15% slower mozilla/translation-service#20 (comment).

Also, we need to update Gecko to support non-alpha models, so they are slow for now ( I assume the same as usage on ARM).

We distinguish between different model types based on naming convention in the model registy.

The models include two different vocabularies for source and target language, unlike all the other models that currently reuse one vocabulary. I added two registry keys to support this.

Also, the model itself is a bit heavier and there's more data to download because of two vocabularies.

Despite all that it's good to have an extra language. It's marked as "Beta", so there shouldn't be expectations for speed or quality.

fixes #166

@eu9ene eu9ene requested review from andrenatal and abhi-agg June 16, 2022 23:13
@eu9ene eu9ene changed the title Add Ukranian models Add Ukrainian models Jun 16, 2022
@andrenatal
Copy link
Contributor

There's so many caveats that I don't know if we should implement it now.

@andrenatal
Copy link
Contributor

I will test it tomorrow morning.

@eu9ene
Copy link
Collaborator Author

eu9ene commented Jun 16, 2022

It looks pretty usable to me on a regular website like this one https://www.pravda.com.ua/

Of course, if you will try to translate huge texts on Wikipedia it will be noticeably slow but it's the same for our other models.

@andrenatal
Copy link
Contributor

andrenatal commented Jun 23, 2022

I've pulled and tested the xpi built from CI and Ukrainian is not being displayed in the translation bar at least in a pt-br browser.

Screen Shot 2022-06-23 at 3 58 40 PM

Copy link
Contributor

@andrenatal andrenatal left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Translation bar does not display Ukrainian in a pt-br browser

@andrenatal
Copy link
Contributor

Same on english browser
Screen Shot 2022-06-23 at 4 20 04 PM

@eu9ene
Copy link
Collaborator Author

eu9ene commented Jun 23, 2022

Translation bar does not display Ukrainian in a pt-br browser

did other languages work? I can see it working using npm run once but when I do npm build and then install it manually the translation bar does not appear at all

@andrenatal
Copy link
Contributor

Translation bar does not display Ukrainian in a pt-br browser

did other languages work? I can see it working using npm run once but when I do npm build and then install it manually the translation bar does not appear at all

yes
Screen Shot 2022-06-23 at 4 33 12 PM

@eu9ene
Copy link
Collaborator Author

eu9ene commented Jun 23, 2022

not sure what the issue was but I synced it with main and it looks working now

@eu9ene eu9ene requested a review from andrenatal June 23, 2022 23:44
@andrenatal andrenatal merged commit 7dfb026 into mozilla:main Jun 24, 2022
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Dump of Ukranian-centric models
2 participants