├── .env
├── README.md
├── chain.py
├── multiple_messages.py
├── output_parser.py
├── prompt_template.py
└── simple_example.py


/.env:
--------------------------------------------------------------------------------
1 | OPENAI_API_KEY=""
2 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # LangChain-Quick-Start


--------------------------------------------------------------------------------
/chain.py:
--------------------------------------------------------------------------------
 1 | from langchain.chat_models import ChatOpenAI
 2 | from langchain.prompts.chat import ChatPromptTemplate
 3 | from langchain.schema import BaseOutputParser
 4 | from dotenv import load_dotenv
 5 | import os
 6 | 
 7 | load_dotenv()
 8 | api_key = os.getenv("OPENAI_API_KEY")
 9 | 
10 | chat_model = ChatOpenAI(openai_api_key=api_key)
11 | 
12 | class CommaSeparatedListOutputParser(BaseOutputParser):
13 |     def parse(self, text: str):
14 |         return text.strip().split(", ")
15 | 
16 | template = """You are a helpful assistant who generates comma separated lists.
17 | A user will pass in a category, and you should generate 5 objects in that category in a comma separated list.
18 | ONLY return a comma separated list, and nothing more."""
19 | human_template = "{text}"
20 | 
21 | chat_prompt = ChatPromptTemplate.from_messages([
22 |     ("system", template),
23 |     ("human", human_template),
24 | ])
25 | 
26 | chain = chat_prompt | chat_model | CommaSeparatedListOutputParser()
27 | result = chain.invoke({"text": "colors"})
28 | print(result)


--------------------------------------------------------------------------------
/multiple_messages.py:
--------------------------------------------------------------------------------
 1 | from langchain.chat_models import ChatOpenAI
 2 | from langchain.schema import HumanMessage
 3 | from dotenv import load_dotenv
 4 | import os
 5 | 
 6 | load_dotenv()
 7 | 
 8 | api_key = os.getenv("OPENAI_API_KEY")
 9 | 
10 | chat_model = ChatOpenAI(openai_api_key=api_key)
11 | 
12 | messages = [HumanMessage(content="from now on 1 + 1 = 3, use this in your replies"), 
13 |             HumanMessage(content="what is 1 + 1?"), 
14 |             HumanMessage(content="what is 1 + 1 + 1?")]
15 | 
16 | result = chat_model.predict_messages(messages)
17 | print(result.content)


--------------------------------------------------------------------------------
/output_parser.py:
--------------------------------------------------------------------------------
 1 | from langchain.chat_models import ChatOpenAI
 2 | from langchain.prompts.chat import ChatPromptTemplate
 3 | from langchain.schema import BaseOutputParser
 4 | from dotenv import load_dotenv
 5 | import os
 6 | 
 7 | load_dotenv()
 8 | api_key = os.getenv("OPENAI_API_KEY")
 9 | 
10 | class AnswerOutputParser(BaseOutputParser):
11 |     def parse(self, text: str):
12 |         """Parse the output of an LLM call."""
13 |         return text.strip().split("answer =")
14 | 
15 | chat_model = ChatOpenAI(openai_api_key=api_key)
16 | 
17 | template = """You are a helpful assistant that solves math problems and shows your work. 
18 |             Output each step then return the answer in the following format: answer = <answer here>. 
19 |             Make sure to output answer in all lowercases and to have exactly one space and one equal sign following it.
20 |             """
21 | human_template = "{problem}"
22 | 
23 | chat_prompt = ChatPromptTemplate.from_messages([
24 |     ("system", template),
25 |     ("human", human_template),
26 | ])
27 | 
28 | messages = chat_prompt.format_messages(problem="2x^2 - 5x + 3 = 0")
29 | result = chat_model.predict_messages(messages)
30 | parsed = AnswerOutputParser().parse(result.content)
31 | steps, answer = parsed
32 | 
33 | print(steps, answer)


--------------------------------------------------------------------------------
/prompt_template.py:
--------------------------------------------------------------------------------
 1 | from langchain.chat_models import ChatOpenAI
 2 | from langchain.prompts.chat import ChatPromptTemplate
 3 | from dotenv import load_dotenv
 4 | import os
 5 | 
 6 | load_dotenv()
 7 | api_key = os.getenv("OPENAI_API_KEY")
 8 | 
 9 | chat_model = ChatOpenAI(openai_api_key=api_key)
10 | 
11 | template = "You are a helpful assistant that translates {input_language} to {output_language}."
12 | human_template = "{text}"
13 | 
14 | chat_prompt = ChatPromptTemplate.from_messages([
15 |     ("system", template),
16 |     ("human", human_template),
17 | ])
18 | 
19 | messages = chat_prompt.format_messages(input_language="English", 
20 |                                         output_language="French", 
21 |                                         text="I love programming.")
22 | result = chat_model.predict_messages(messages)
23 | print(result.content)


--------------------------------------------------------------------------------
/simple_example.py:
--------------------------------------------------------------------------------
 1 | from langchain.chat_models import ChatOpenAI
 2 | from dotenv import load_dotenv
 3 | import os
 4 | 
 5 | load_dotenv()
 6 | 
 7 | api_key = os.getenv("OPENAI_API_KEY")
 8 | 
 9 | chat_model = ChatOpenAI(openai_api_key=api_key)
10 | 
11 | result = chat_model.predict("hi!")
12 | print(result)


--------------------------------------------------------------------------------