I stored sentence transformers in elasticsearch and used its cosineSimilarity function to find similar documents. I'm wondering if maybe if makes sense to use

I stored sentence transformers in elasticsearch and used its cosineSimilarity function to find similar documents. I'm wondering if maybe if makes sense to use pgvector? or sqlite? instead