Find answers from the community

Updated 6 months ago

Latency

At a glance

The community members are discussing tools to track indexing latency and get maximum stats while indexing. One community member suggests using Arize and time.time(), while another agrees but wants to compare the latency of different embedding methods, such as TGI-based and OpenAI embeddings, especially when working with terabytes of data. The community members also note that embedding calls are typically batched, and Arize tracks each embedding call.

is there a tool which i can use to track indexing latency and get maximum stats while indexing

L
B
6 comments
Arize? Using time.time()
ageed but for something like Actually latency per embedding call comparing TGI Based Embedding to OpenAI one
working with TB of Data
Arize tracks each embedding call
Keep in mind embedding calls are typically batched though
Add a reply
Sign up and join the conversation on Discord