hi i am new to python and i need your

kkaandikbas

hi i am new to python and i need your help on something.

I want to make a chatbot using my own data but it always uses the text-davinci model, how can I change this model?

I am sending the code. thanks

Code :

import openai
from llama_index import VectorStoreIndex, SimpleDirectoryReader, Document, StorageContext, load_index_from_storage
import os

os.environ['OPENAI_API_KEY'] = 'token'
openai.api_key = os.environ['OPENAI_API_KEY']

documents = SimpleDirectoryReader('textFiles').load_data()

text_list = [documents[0].text]
documents = [Document(t) for t in text_list]

index = VectorStoreIndex.from_documents(documents)

index.storage_context.persist()

storage_context = StorageContext.from_defaults(persist_dir="./storage")
index = load_index_from_storage(storage_context)
query_engine = index.as_query_engine()

while True:
user_input = input("Please enter a question: ")
if user_input.lower() == 'exit':
break
response = query_engine.query(user_input)
print(response)

6 comments

LLogan M

You can use the service context to change the model

Plain Text

from langchain.chat_models import ChatOpenAI
from llama_index import LLMPredictor, ServiceContext

service_context = ServiceContext.from_defaults(llm_predictor=LLMPredictor(llm=ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0)))

index = load_index_from_storage(storage_context, service_context=service_context)

kkaandikbas

Hello again,
Thank you very much for your help.
In the example you gave, https://api.openai.com/v1/chat/completions sends a request here, I want it to send a request here https://api.openai.com/v1/completions and I want it to use text-ada-001 as a model.

Can you help me? @Logan M

kkaandikbas

from llama_index.llms import OpenAI
llm = OpenAI(temperature=0, model="text-davinci-002")
service_context = ServiceContext.from_defaults(llm=llm)

When I use this example written in the document, I get this error:

File "main.py", line 5, in <module>
    from llama_index.llms import OpenAI
ModuleNotFoundError: No module named 'llama_index.llms'

@Logan M

LLogan M

The docs are updated for a big release, but the release is not out yet 😅 use a pinned version of the docs for now

https://gpt-index.readthedocs.io/en/v0.6.38/

LLogan M

Why do you want to use text Ada 001 for generating text? It will be quite bad...

In any case, you can use

Plain Text

from langchain.llms import OpenAI

And use that instead of ChatOpenAI

kkaandikbas

thank you

Add a reply

Find answers from the community

hi i am new to python and i need your