NLLB-inference icon indicating copy to clipboard operation
NLLB-inference copied to clipboard

Character translation limit?

Open 314esther opened this issue 1 year ago • 1 comments

The sentence buffer size is quoted at 100000, however I hit a limit at about 500 characters being translated, is this baked in somewhere?

314esther avatar Jul 11 '22 03:07 314esther

Hi, buffer-size is simply set large enough to be greater than batch-size, it's not related to the generation result. Fairseq has a max_position=512 limit, maybe one of your input sentence reached this length limitation.

pluiez avatar Jul 11 '22 05:07 pluiez