Comments (4)
Why during embedding generation, is gpt-3.5-turbo used instead of text-embedding-ada-002?
I am seeing things like this in my billing panel:
Local time: 8 jun 2023, 14:15
gpt-3.5-turbo-0301, 63 requests
18,792 prompt + 4,115 completion = 22,907 tokens
Local time: 8 jun 2023, 14:15
text-embedding-ada-002-v2, 4 requests
493 prompt + 0 completion = 493 tokensThat is definitely not normal, doing the embedding from GPT instead of Ada is a very high cost if you have to process up to 1M tokens as I usually have to do.
Reviewing the code, it has indeed been established that the embeddings are delegated to text-embedding-ada-002, but when it comes down to it, it is gpt-3.5 that is generating them.
It functions correctly to my knowledge, I think you have just misunderstood the purpose of the program. Having everything done in embeddings removes the ability to dynamically search the multiple dbs and load them in varying correct orders. Aetherius is a simulacrum of the human thought process/memory retrieval and doesn't function like other chatbots. The high usage of gpt-3.5 is from loading the various types of memories, not from it being embedded. The embeddings are only used for the database upsert.
from aetherius_ai_assistant.
Local time: 8 jun 2023, 14:15
gpt-3.5-turbo-0301, 63 requests
18,792 prompt + 4,115 completion = 22,907 tokensLocal time: 8 jun 2023, 14:15
text-embedding-ada-002-v2, 4 requests
493 prompt + 0 completion = 493 tokensLocal time: 8 jun 2023, 14:20
gpt-3.5-turbo-0301, 32 requests
8,624 prompt + 2,030 completion = 10,654 tokensLocal time: 8 jun 2023, 14:25
gpt-3.5-turbo-0301, 100 requests
26,169 prompt + 5,541 completion = 31,710 tokensLocal time: 8 jun 2023, 14:25
text-embedding-ada-002-v2, 3 requests
243 prompt + 0 completion = 243 tokens
Very little is using Ada
from aetherius_ai_assistant.
This is how the models should really be used for embedding:
Local time: 8 jun 2023, 2:50
gpt-3.5-turbo-0301, 1 request
55 prompt + 9 completion = 64 tokensLocal time: 8 jun 2023, 2:50
text-embedding-ada-002-v2, 75 requests
21,991 prompt + 0 completion = 21,991 tokensLocal time: 8 jun 2023, 2:55
text-embedding-ada-002-v2, 521 requests
149,767 prompt + 0 completion = 149,767 tokensLocal time: 8 jun 2023, 3:00
text-embedding-ada-002-v2, 535 requests
154,043 prompt + 0 completion = 154,043 tokensLocal time: 8 jun 2023, 3:05
text-embedding-ada-002-v2, 254 requests
85,398 prompt + 0 completion = 85,398 tokens
from aetherius_ai_assistant.
I am now understanding you then, it is not like the typical ChatPDF that you pass documentation to and it remains in the DB
from aetherius_ai_assistant.
Related Issues (6)
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from aetherius_ai_assistant.