mikeee/baai-m3 · Apply for community grant: Personal project (gpu)

BAAI/bge-m3 is a popular model for embedding and reranking.

However, the Huggingface's Inference Endpoint often returns 500 Server Error: Internal Server Error for url: https://api-inference.huggingface.co/pipeline/feature-extraction/BAAI/bge-m3. This tiny project provides an embed API (more of less equivalent to huggingface_hub.InferenceClient.feature_extraction('BAAI/bge-m3') and a rerank API at the same time and would be useful to toy with ideas for hobby projects.

It would be greatly appreciated if Huggignface can kindly sponsor a GPU. (The project will automatically make use of a GPU if one is available.)