infrawhispers / anansi Goto Github PK

add support for https://huggingface.co/BAAI/AltCLIP-m18

I created a pull request for export AltCLIP-m18 for onnx

https://github.com/FlagAI-Open/FlagAI/pull/322/files#diff-e0fa65fd7d57ecb0570cc6e05c2db4e3f26fcae4177f7eb7401dbc478b082077

Support for multilingual text search I think would be very useful

add support for https://huggingface.co/BAAI/AltCLIP-m18

I created a pull request for export AltCLIP-m18 for onnx

https://github.com/FlagAI-Open/FlagAI/pull/322/files#diff-e0fa65fd7d57ecb0570cc6e05c2db4e3f26fcae4177f7eb7401dbc478b082077

Support for multilingual text & image search I think would be very useful

there are a bunch of things that we need to do to get to a 0.1.0 release. here is an evolving list:

• rename embeddings -> embedds
• document the HTTP apis in a README.md
• document the list of available models
• validate that we can load and run all CLIP models exported in .onnx by JINA
• add usage of setence-t5-* models
• add usage of gtr-t5-xxl
• add usage of msmarco-bert-condensor
• add usage of bert-base-uncased
• create and publish a python client

we will likely want to exclusively this from a configuration file on startup (likely using the same config.yaml) - which allows us to:

1. validate that the custom models exist and can operate on a piece of content in an expected manner (simply a warmup eval)
2. continue to expand upon the existing config.yaml file and keep the general configuration of the binary efficient.