python - ChatGPT API 定制训练的 AI 聊天机器人对 Python 查询应答 "None"

Question

我正在连接到我的第一个聊天机器人。基于此处概述的流程: https://beebom.com/how-train-ai-chatbot-custom-knowledge-base-chatgpt-api/

我创建了他建议的代码来让 ChatGPT 分析我的 PDF。不过代码有点过时了，我必须做一些调整。这就是我现在拥有的:

from llama_index import *
from langchain.chat_models import ChatOpenAI
import gradio as gr
import sys
import os
import openai

os.environ["OPENAI_API_KEY"] = 'XXXX'
openai.api_key = "XXXX"

documents = ""
service_context = ""

def construct_index(directory_path):
    max_input_size = 4096
    num_outputs = 512
    max_chunk_overlap = 20
    chunk_size_limit = 600

    prompt_helper = PromptHelper(max_input_size, num_outputs, chunk_overlap_ratio=0.1, chunk_size_limit=chunk_size_limit)
    llm_predictor = LLMPredictor(llm=ChatOpenAI(temperature=0.7, model_name="gpt-3.5-turbo", max_tokens=num_outputs))
    documents = SimpleDirectoryReader(directory_path).load_data()
    service_context = ServiceContext.from_defaults(llm_predictor=llm_predictor, prompt_helper=prompt_helper)
    
    index = GPTVectorStoreIndex.from_documents(documents, service_context=service_context)
    
    # apparently this saves it to disk?
    index.storage_context.persist(persist_dir='docs')
    storage_context = StorageContext.from_defaults(persist_dir='docs')
    index = load_index_from_storage(storage_context)
    return index

def chatbot(input_text):
    index = GPTVectorStoreIndex.from_documents(documents, service_context=service_context)
    index.storage_context.persist(persist_dir='docs')
    storage_context = StorageContext.from_defaults(persist_dir='docs')
    index = load_index_from_storage(storage_context)

    # tried this method as well with no success instead of above
    #index = GPTVectorStoreIndex.from_documents(documents, service_context=service_context)

    query_engine = index.as_query_engine()
    response = query_engine.query(input_text)

    # am I returning the correct object here? I believe its supposed to be JSON?
    return response

iface = gr.Interface(fn=chatbot,
                     inputs=gr.components.Textbox(lines=7, label="Enter your text"),
                     outputs="text",
                     title="Custom-trained AI Chatbot")
index = construct_index("docs")
iface.launch(share=True)

当我运行该程序时，没有错误，它说它在我的 Ip 上运行。当我到达聊天机器人时，一切看起来都很好，直到我提出问题。然后它一直说“无”

控制台中没有错误或警告，程序继续运行。每当我查询它时，它总是说“无”。我哪里出错了？顺便说一句，我并不是 100% 理解代码，这是对原始示例的大量修改，以使所有库都能正常工作。如果有人可以简单地解释发生了什么，我们将不胜感激。谢谢G

Answer 1

Chatbot 函数返回 response=None，它只是在输出文本框组件中收集。我建议在 iface.launch() 中使用 debug=True。更多信息可以在这里找到 - https://www.gradio.app/docs/blocks#blocks-launch-arguments