Do you have a reference to the difference between search embeddings and similarity embeddings? Can you describe the difference, ex. Is a Dual-Encoding model generating the search embedings using query-retrieval pairs?
Do you have a reference to the difference between search embeddings and similarity embeddings? Can you describe the difference, ex. Is a Dual-Encoding model generating the search embedings using query-retrieval pairs?
this should help: [2201.10005] Text and Code Embeddings by Contrastive Pre-Training