Changing the tokenizer to multilingual

[SD-fan]

I need to use a multilang tokenizer as my documents are neither EN nor CN. Unfortunately just downloading the model (intfloat/multilingual-e5-large) and changing the .env seems not to do the trick.

Any ideas?

Btw: I tried looking at the code, but comments in Chinese make it very hard to read 😢