Cameron

@guillaume-be i think it was failing cos im just using SentenceEmbeddingsBuilder::local(PATH) and its always trying to find a rust_model.ot file.

@saied71 we use pos_patterns passed into our service as a sequence. So i guess you could do something like this. ``` def kph_extr(docs:list, pos_patterns: Sequence[str] = ("*+",)) -> list :...

@huemans you probably have two options cos of the way digitalocean works. You can build or use a packer image that has puppet on it as the base build or...

Could use a puppet script and pass it to cf-init/cloud-init, or build our own packer boxes

@cjrh @imaurer I was hoping tantivy would publish a new release so will could pull in some upstream fixes as well. I'll make a release within the next few days

> Hi @cclauss, > > As I mention in the PR's description, the publishing step required some configuration on the PyPI repo, to configure the workflow as a trusted publisher...

As @Dunedan mentioned above. A workaround for now is to consider making your own package index. Considering the package index that is used is https://github.com/nltk/nltk_data. You can create your own...

From ruff pre-commit readme ``` When running with --fix, Ruff's lint hook should be placed before Ruff's formatter hook, and before Black, isort, and other formatting tools, as Ruff's fix...

@NDNM1408 it was exposed here https://github.com/quickwit-oss/tantivy-py/pull/233. What version of tantivy-py are you using?

We experienced something similiar with extra_body input_type using bedrock/amazon.titan-embed-text-v2:0. Input type is supported on vertex embeddings, but we use both.