hi everyone please someone can recommend me the best pipeline for french language and thank you very much
Unfortunately the “best” general pipeline doens’t really exist. A pipeline for a digital assistant that needs to mostly handle FAQ will be very different compared to one that needs to handle very long contextualized conversations. We have a guide in our documentation that suggests a place to get started though and it even lists an example in french that uses spaCy word vectors.
Other word vectors also exist though and you might like to learn that we’re working on an example packages that contains more embeddings for non-english languages here but we cannot (and should not) guarantee their effectiveness upfront. This always needs to be confirmed and benchmarked on your own dataset.
1 Like