Greg Tatum

Is it appropriate to do data augmentation at training time to swap between `.` and `。`?

Is it possible to dynamically determine this value? Like run N translations, measure and adjust?

Training uses dynamic batch sizes, so it changes the batch size over time to find the best value, so there's not really a need to adjust it. It starts somewhat...

I think the only thing left to do here is change the config generation to use CTranslate2 by default.

Ok, en-lt sounds like a great choice. I read a bit more on it and it's got a lot of qualitative feedback in #756.

Lithuanian has a similar use of declensions: https://en.wikipedia.org/wiki/Lithuanian_declension

I've got the first one started, and will wait until it gets to the student step before kicking off the rest: https://firefox-ci-tc.services.mozilla.com/tasks/groups/Wxvkl1ruQkCj6URG6oIuuQ The configs are each done on a commit-by-commit...

They are all in student training now on [the dashboard](https://gregtatum.github.io/taskcluster-tools/src/training/?taskGroupIds=O3KtCsLKTHSRyvarD0OmkA%2CMb6nq4IiRx2ZQALGOl506w%2CF4lNZ2k6T7mr6wyNhdNpqg%2CCP2Xt9qhScW0ByzsFMTGng%2CGoDGckYuSuS9-eJbwx9a9g%2CUBduTJojQDWmoedTxg74ww%2CAuR23z6EQMeZrNM398e10A%2CV4rRhXt6Te-g_004KEiEUQ%2CFiSNJUysQKqf0HDZZ7uBAw%2CQb2MenUpRE-dyR8owan07Q%2CPy4okXqaTzKG4sFYqi9Nqg%2CI5x8d1_RS5avEFpYbEPdmw%2CL4LaCVDQT_6nwt7tMaCSOQ%2CTaeCdUs5Rqq7w1Tbf1PShQ&showAll=false&hidden=CGdli4TrRCu_FyLy3bsmcg%2CK_6r19C1Q-aAjFVS_VP21Q%2Ccfo9K-R8QtmuE27UKQ4BbQ&dashboardName=Experiment+Dashboard). They are all the ones named `decoder-*`.

So I misread the paper a bit when it was talking about decoders, and the ffn and embedding size affect both decoder and encoder equally. The decoder depth is the...

The [model inspector](https://gregtatum.github.io/taskcluster-tools/src/model-inspector) on `decoder-base` shows that the `dim-rnn` is not used: ``` bert-train-type-embeddings: true bert-type-vocab-size: 2 dec-cell: ssru dec-cell-base-depth: 2 dec-cell-high-depth: 1 dec-depth: 2 dim-emb: 512 dim-rnn: 1024 dim-vocabs:...