I am trying to follow the example from the official MongoDB documentation.
It mentioned that the search vector was based on the string "historical heist" using the OpenAI text-embedding-ada-002 model. However, when I tried to validate this with my own OpenAI API key, I got different results.
Why?
curl https://api.openai.com/v1/embeddings \ -H "Content-Type: application/json" \ -H "Authorization: Bearer sk-xxxxx" \ -d '{"input": "historical heist","model": "text-embedding-ada-002","encoding_format": "float" }'My results are the following:
"data": [ {108 "object": "embedding","index": 0,19 "embedding": [8 -0.02273331,97 -0.027646419, 0.0030520316, 3 -0.033631727,1 -0.026207773,4 0:00:01 --:- 0.0016854883,-:-- 0:0 -0.017983066,0:0 0.006311038,1 -0.0013012275,201 -0.024823416,54 0.015227924, 0.005751188, 0.0004003777, -0.008095773,The example results are the following:
'queryVector': [-0.020156775, -0.024996493, 0.010778184, ....