TypeError: 'CreateEmbeddingResponse' object is not subscriptable

Question

I am trying to run sample code from the openai website for getting embeddings for dataset: https://platform.openai.com/docs/guides/embeddings/use-cases. However, the code returns an error that I can't resolve looking at historical posts

I tried running this code where df is a dataframe I have created with my own data that loaded succesfully.

from openai import OpenAI
client = OpenAI()

def get_embedding(text, model="text-embedding-ada-002"):
   text = text.replace("\n", " ")
   return client.embeddings.create(input = [text], model=model)['data'][0]['embedding']

df['embedding'] = df.ITEM_DESCRIPTION.apply(lambda x: get_embedding(x, model='text-embedding-ada-002'))
df.to_csv('embedded_output.csv', index=False)

Neomineomycin

Can you please edit to provide a minimal reproducible example including the full traceback and whatever input data you used. – Transformism

Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. – Wheedle

Answer 1

They didn't update their documentation. It is now:

def get_embedding(text, model="text-embedding-ada-002"):
    text = text.replace("\n", " ")
    return client.embeddings.create(input = [text], model=model).data[0].embedding

Minette

Thank you Florian, your answer works for me. There is a similar case with the ChatCompletion, TypeError: 'ChatCompletion' object is not subscriptable But why this error occurs in the first place, can anyone explain it please? – Fishmonger

It seems create() was returning dict like object before and returning class instance currently. – Tokenism

This is the best answer – Strega