TypeError

At a glance

A community member is following the llama-index documentation and encountering a TypeError: 'RetrieverQueryEngine' object is not callable error. The community members discuss the issue, with some suggesting the community member is using an outdated version of the library and should follow the latest documentation. They provide links to the official llama-index documentation and suggest the community member install the latest version of the library and set up a new environment. The community members also discuss how to correctly import and use the LLM object directly, without the deprecated LLMPredictor. The community members suggest the community member is trying to use the llama-index to generate text in the style of the documents they have, but it's unclear if they have found a solution to their original issue.

Useful resources

SSlurm

people i am following the llama-index documentation to the word, copy/pasting the code

Plain Text

documents = SimpleDirectoryReader('Knowlege').load_data()
index = VectorStoreIndex.from_documents(documents)
query_engine = index.as_query_engine()
response = query_engine('what are these documents about?')
print(response)

but i am getting this error: TypeError: 'RetrieverQueryEngine' object is not callable
how do i fix this error?

14 comments

WWhiteFang_Jr

Can you share the doc link that you are using to follow.

what version are you using?

SSlurm

https://pypi.org/project/gpt-index/

WWhiteFang_Jr

Where did you get this link 😅
This is a very old version. Like wayyy old

Use this: https://docs.llamaindex.ai/en/stable/index.html
This is the latest one

WWhiteFang_Jr

Make sure you spin up a new env and install llama-index

SSlurm

i got it from a tutorial that is just a few months old

SSlurm

ok let me see

SSlurm

ok i have reached this error, i cannot find a solution for it in the documents

Plain Text

from llama_index import LLMPredictor

how to correctly import it?

WWhiteFang_Jr

You don't have to use it. This has been deprecated as well 😅

You can directly use llm object.

WWhiteFang_Jr

Are you using openai?

WWhiteFang_Jr

Follow this tutorial. This is the latest and official documentation

SSlurm

yes, i am confused how can i use it directly? basically i have 5 documents and i want the LLM model to read them and emulate their content, like 5 documents about james bond and i want the chat bot to talk in the style of james bond, am I implementing the correct technique here?

SSlurm

This is my code so far, it is broken i am not sure how to stick the LLM to the context

Plain Text

def train(directory):
    documents = SimpleDirectoryReader(directory).load_data()
    index = VectorStoreIndex.from_documents(documents)
    index.storage_context.persist()

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

service_context = ServiceContext.from_defaults(llm_predictor=OpenAI())
storage_context = StorageContext.from_defaults(persist_dir='./storage')
index = load_index_from_storage(storage_context, service_context=service_context)
query_engine = index.as_query_engine()

response = query_engine.query('Write me a dialogue like the style from the documents I have presented you?')
print(response)

WWhiteFang_Jr

If you check the 5 line of code ( starter tutorial here: https://docs.llamaindex.ai/en/stable/getting_started/starter_example.html#starter-tutorial

It can get you started with openAI

Just install pip install llama-index

Plain Text

from llama_index.core import VectorStoreIndex, SimpleDirectoryReader

documents = SimpleDirectoryReader("data").load_data()
index = VectorStoreIndex.from_documents(documents)
query_engine = index.as_query_engine()
response = query_engine.query("What did the author do growing up?")
print(response)

SSlurm

nice excellent,

Add a reply

Find answers from the community

TypeError