Using Langchain to access an LLM on a vLLM server

langchain_chatopenai.py

from langchain import PromptTemplate
from langchain.prompts.chat import ChatPromptTemplate, HumanMessagePromptTemplate
from langchain.chains import LLMChain
from langchain.chat_models import ChatOpenAI

import requests

base_url = "http://localhost:9999/v1"
x = requests.get(base_url + "/models")
model = str(x.json()["data"][0]["id"])
print(f"model: {model}")

llm = ChatOpenAI(temperature=0.7, 
	openai_api_key="***", 
	verbose=True,
	openai_api_base = base_url,
	model_name = model,
	max_tokens = 1024)

prompt = PromptTemplate(
    input_variables=["country"],
    template="What is the capital of {country}?"
)

message_prompt = HumanMessagePromptTemplate(prompt=prompt)
chat_prompt = ChatPromptTemplate.from_messages([message_prompt])

chain = LLMChain(llm=llm, prompt=chat_prompt)

print(chain.run("France"))