Step‑by‑Step Guide to Building a Name‑Generator with LangChain and OpenAI

1. Install LangChain

pip install --upgrade langchain==0.0.279 -i https://pypi.org/simple

LangChain provides a unified interface for large language models (LLMs) and supports prompt templating, output parsing, and chain composition.

2. Create an LLM

Self‑hosted platform + open‑source model (requires substantial GPU resources and your own training data).

Third‑party model APIs such as OpenAI, Baidu Wenxin, Alibaba Tongyi, etc. (no GPU needed).

Example goal: let the LLM generate Chinese‑style names. The basic LangChain call is llm.predict(prompt). The temperature parameter controls creativity.

3. Custom Prompt Templates

Template the context of a question.

Allow parameter injection.

Example: generate American‑style names.

Usage

Import the prompt utilities:

from langchain.prompts import PromptTemplate

4. Output Parser

Format LLM results into JSON or other structured data.

Useful for returning arrays instead of plain text.

Example: ask the model to return four Chinese names as an array.

from langchain.schema import BaseOutputParser

class CommaSeparatedListOutputParser(BaseOutputParser):
    """Parse the output of an LLM call to a comma‑separated list."""
    def parse(self, text: str):
        """Parse the output of an LLM call."""
        return text.strip().split(", ")

CommaSeparatedListOutputParser().parse("hi, bye")

['hi', 'bye']

5. Run the Full Example

pip install openai==v0.28.1 -i https://pypi.org/simple

Set OpenAI credentials

import os
os.environ["OPENAI_KEY"] = "sk-ss"
# For proxy access
os.environ["OPENAI_API_BASE"] = "https://ai-yyds.com/v1"

Verify installations

! pip show langchain
! pip show openai

Use OpenAI SDK directly

import openai, os
openai.api_base = os.getenv("OPENAI_API_BASE")
openai.api_key = os.getenv("OPENAI_KEY")
messages = [{"role": "user", "content": "介绍下你自己"}]
res = openai.ChatCompletion.create(model="gpt-3.5-turbo", messages=messages, stream=False)
print(res['choices'][0]['message']['content'])

Call the model through LangChain

from langchain.llms import OpenAI
import os
api_base = os.getenv("OPENAI_API_BASE")
api_key = os.getenv("OPENAI_KEY")
llm = OpenAI(model="gpt-3.5-turbo-instruct", temperature=0, openai_api_key=api_key, openai_api_base=api_base)
llm.predict("介绍下你自己")

Name‑generator with a custom prompt and parser

from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate
from langchain.schema import BaseOutputParser
import os

api_base = os.getenv("OPENAI_API_BASE")
api_key = os.getenv("OPENAI_KEY")
llm = OpenAI(model="gpt-3.5-turbo-instruct", temperature=0, openai_api_key=api_key, openai_api_base=api_base)

prompt = PromptTemplate.from_template(
    "你是一个起名大师,请模仿示例起3个{county}名字,示例：男孩常用名{boy},女孩常用名{girl}。请返回以逗号分隔的列表。"
)
message = prompt.format(county="美国男孩", boy="sam", girl="lucy")
print(message)
result = llm.predict(message)
class CommaSeparatedListOutputParser(BaseOutputParser):
    def parse(self, text: str):
        return text.strip().split(",")
print(CommaSeparatedListOutputParser().parse(result))

Typical output:

'

男孩: 龙飞、铁柱、小虎
女孩: 玉兰、梅香、小红梅'

Conclusion

The guide demonstrates how to set up LangChain, choose an LLM source, craft reusable prompt templates, and convert free‑form text into structured data with a custom output parser, enabling practical applications such as culturally aware name generation.