SthPhoenix

Hi! That's strange I have just build image from scratch without cache and pulled latest `python:3.8-slim` base image - everything works as expected.

Have you managed to figure out the issue? I still can't reproduce it.

I wasn't able to reproduce this bug, @jinzaz could you share more info about your environment? Are you running service in docker? Could you compute md5 sums of onnx files?...

Proper md5 sum for `glintr100` model should be `3b366b98f786426f79629ddb2e56629c`, in case you got different check sum you just need to re-download the model, with latest commit it'll be downloaded again...

It might be google drive is inaccessible in your region, you can try manually downloading models from google drive using proxy: [srcfd_2.5g_gnkps](https://drive.google.com/file/d/1F__ILEeCTzeR71BAV-vInuyBezYmNMsB/view?usp=sharing) [glintr100](https://drive.google.com/file/d/1TR_ImGvuY7Dt22a9BOAUAlHasFfkrJp-/view?usp=sharing) Models should be placed under following path:...

Hi! It's not possible in original plugin at the moment, but you can try my [fork](https://github.com/SthPhoenix/elastik-nearest-neighbors-extended). It has `_aknn_search_vec` endpoint, which can be used like this: ``` POST :9200/_aknn_search_vec {...

Hi! Great work @linghu8812! I've got a bit overlapping project, though more focused on deployment of TensorRT face recognition pipeline over REST API. I hope it won't be too rude...

Have you tried Chinese whispers with original insightface models? I have tried it some time ago and I recall it was working pretty well.

Are you using normalized embeddings for clustering?

Link is broken