using GPT-4's pre-trained embeddings?

[chonchonj23]

Dear Maarten,

The code below has generated some excellent results, and I want to confirm if the code uses GPT-4's pre-trained embeddings.

from bertopic.representation import OpenAI
representation_model = OpenAI(model="gpt-4", chat=False)
model = BERTopic(embedding_model=representation_model)
topics, probs = model.fit_transform(docs)

Best regards,
Chon

[MaartenGr]

It does not. What you are using is a representation model, not an embedding model. I would advise reading the documentation instead.

[chonchonj23]

Thanks Maarten!

I can now see the magic a representation model can do on topic keywords!