Fixes float32->float64 precision widening a.k.a. float bloat - #3743

Fixes float32->float64 precision widening a.k.a. float bloat#3743

binarymax wants to merge 5 commits into

openai:mainopenai/openai-python:mainfrom

omc:maxirwin/retain-float32-embeddingsomc/openai-python:maxirwin/retain-float32-embeddingsCopy head branch name to clipboard