Error with LangChain ChatPromptTemplate.from_messages

As shown in LangChain Quickstart, I am trying the following Python code:

from langchain.prompts.chat import ChatPromptTemplate
template = "You are a helpful assistant that translates {input_language} to {output_language}."
human_template = "{text}"
chat_prompt = ChatPromptTemplate.from_messages([ ("system", template), ("human", human_template),
])
chat_prompt.format_messages(input_language="English", output_language="French", text="I love programming.")

But when I run the above code, I get the following error:

Traceback (most recent call last): File "/home/yser364/Projets/SinappsIrdOpenaiQA/promptWorkout.py", line 6, in <module> chat_prompt = ChatPromptTemplate.from_messages([ File "/home/yser364/.local/lib/python3.10/site-packages/langchain/prompts/chat.py", line 220, in from_messages return cls(input_variables=list(input_vars), messages=messages) File "/home/yser364/.local/lib/python3.10/site-packages/langchain/load/serializable.py", line 64, in __init__ super().__init__(**kwargs) File "pydantic/main.py", line 341, in pydantic.main.BaseModel.__init__ pydantic.error_wrappers.ValidationError: 4 validation errors for ChatPromptTemplate messages -> 0 value is not a valid dict (type=type_error.dict) messages -> 0 value is not a valid dict (type=type_error.dict) messages -> 1 value is not a valid dict (type=type_error.dict) messages -> 1 value is not a valid dict (type=type_error.dict)

I use Python 3.10.12.

1 Answer

Your example is from the Prompt templates section of the LangChain Quickstart tutorial. I did not spot any differences, so it should work as given.

I tried out the example myself, with an additional loop to output the messages created by chat_prompt.format_messages:

from langchain.prompts.chat import ChatPromptTemplate
template = "You are a helpful assistant that translates {input_language} to {output_language}."
human_template = "{text}"
chat_prompt = ChatPromptTemplate.from_messages([ ("system", template), ("human", human_template),
])
messages = chat_prompt.format_messages(input_language="English", output_language="French", text="I love programming.")
for message in messages: print(message.__repr__())

The example works without any errors. The result is very similar to what is shown in the tutorial, although not identical:

SystemMessage(content='You are a helpful assistant that translates English to French.', additional_kwargs={})
HumanMessage(content='I love programming.', additional_kwargs={}, example=False)

I ran the test with Python 3.9.5 and LangChain 0.0.300, which is the lastest version on PyPI. According to PyPI, it supports Python >=3.8.1 and <4.0.

Maybe your version of LangChain or one of its dependencies is outdated? Try to run it in a new venv with a fresh install of LangChain.