This tweet makes me sad. We should fix at some point.
I really donât see how this kind of toolkit could actually be used by a translator directly.
Most of them 1) wonât be Linux savvy enough 2) wonât have the GPU to train a model.
There are 2 directions I think:
-
sdl studio integration as discussed in the other thread
-
create kind of âscript recipesâ where you can easily see all the necessary steps from
getting the corpus, pre-procesing, training, evaluating, spitting the bleu scores.
These ârecipesâ could make it easier for âopenNMT newbiesâ to step in the project and replicates some results.
@srush if youâre familiar with Kaldi, this is the kind of ârecipesâ I have in mind.
Cheers.
Vincent
This is very helpful. I am pretty out of touch with this community, but I would like to keep these ideas on our road map so we donât lose touch.