CODEWITHSUNDEEP
nlpembeddinglangchain
mechanicalsloth
mechanicalsloth

Reputation: 11

How to speed up the embedding process in LangChain

I have approximately 1600 short text files to embed using Sentence Transformers and store in a chroma vector in LangChain.

I want to create a Retrieval Question/Answering (QA) capability to retrieve those text files. I've done the processing and started the embedding process, but it's been 3-4 hours and it's still running. Is there any way to speed up this process?

Upvotes: 1

Views: 2141

Answers (1)

tomasantunes
tomasantunes

Reputation: 945

You should try a smaller sample and do the math to see how long it really takes. It's a computationally intensive task so if you want to speed it up you have to rent compute time with a faster GPU.

It helps to do these tasks on a pre-processing stage and then keep a memory cache for fast delivery.

Upvotes: 0

Related Questions