I am trying the following code for entity extraction using Azure OpenAI endpoint. The idea of using LLM is that I can leverage language understanding for more evolved use cases over time.

from langchain.chains import create_extraction_chain

from langchain_openai import AzureChatOpenAI

azure_api_key = "key comes here"

azure_endpoint = "endpoint comes here"

MODEL_NAME = "gpt-35-turbo"

api_version = "2023-09-15-preview"

import os
os.environ["AZURE_OPENAI_API_KEY"] = azure_api_key

os.environ["AZURE_OPENAI_ENDPOINT"] = azure_endpoint

llm = AzureChatOpenAI(
    azure_deployment=MODEL_NAME,
    openai_api_version=api_version) 

schema={
    "properties": {
     "name":{"type":"string"},
     "age":{"type":"integer"}, 
    },
    "required":["name","age"],
}
chain=create_extraction_chain(schema,llm)

inp = """Return the information in JSON format: Names, age. For example, '{"name": "John", "age": 7, }'. Now, Peter is 5 years old, Agatha is twice his age and Charles is 1 year younger than Agatha."""

result = chain.invoke(inp)

I get an error:
JSONDecodeError: Expecting property name enclosed in double quotes: line 2 column 3 (char 4)

During handling of the above exception, another exception occurred:

OutputParserException Traceback (most recent call last)
OutputParserException: Could not parse function call data: Expecting property name enclosed in double quotes: line 2 column 3 (char 4)

I have tried doing a try-catch, but looks like the issue arises while passing the schema.