llm of DatasetGenerator

[xinshuaiiii]

from openai import OpenAI
client=OpenAI(
base_url="",
api_key='',
)
message_text = [{"role":"user","content":"what's 1+1? Answer in one word."}]

llm = client.chat.completions.create(
model="xxxxx",
messages = message_text,
temperature=0.7,
max_tokens=800,
top_p=0.95,
)
dataset_generator = DatasetGenerator.from_documents(
[doc],
question_gen_query=QUESTION_GEN_PROMPT,
llm=llm,
show_progress=True,
num_questions_per_chunk=1,
)

How to apply the llm in OpenAI to DatasetGenerator？

[dosubot[bot]]

Hey @xinshuaiiii! I'm here to help you with any bugs, questions, or contributions while you wait for a human maintainer. Let's tackle this together!

To apply the LLM from OpenAI to DatasetGenerator, you need to pass an instance of the LLM to the DatasetGenerator class during its initialization. Here's how you can do it:

1. First, ensure you have an instance of the LLM from OpenAI. You seem to have created it using the client.chat.completions.create method. However, you need to wrap this in a class or function that can be passed to DatasetGenerator.

2. Use this LLM instance when initializing DatasetGenerator. Here's an example:

from llama_index.core.llms.llm import LLM

# Assuming you have a function or class that wraps your OpenAI LLM
class OpenAILLM(LLM):
    def __init__(self, client):
        self.client = client

    def generate(self, prompt):
        response = self.client.chat.completions.create(
            model="xxxxx",
            messages=[{"role": "user", "content": prompt}],
            temperature=0.7,
            max_tokens=800,
            top_p=0.95,
        )
        return response['choices'][0]['message']['content']

# Initialize your OpenAI LLM
openai_llm = OpenAILLM(client)

# Create a DatasetGenerator instance with the OpenAI LLM
dataset_generator = DatasetGenerator.from_documents(
    [doc],
    question_gen_query=QUESTION_GEN_PROMPT,
    llm=openai_llm,  # Pass the OpenAI LLM instance
    show_progress=True,
    num_questions_per_chunk=1,
)

This example demonstrates how to wrap your OpenAI LLM in a class that can be used with DatasetGenerator ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}