Chat Prompt Templates¶

The Cassandra-specific approach can be seamlessly integrated with LangChain's "chat prompt templates".

In [1]:

from langchain.prompts import createCassandraPromptTemplate

from langchain.prompts import createCassandraPromptTemplate

In [2]:

from cqlsession import getCQLSession, getCQLKeyspace
cqlMode = 'astra_db' # 'astra_db'/'local'
session = getCQLSession(mode=cqlMode)
keyspace = getCQLKeyspace(mode=cqlMode)

from cqlsession import getCQLSession, getCQLKeyspace cqlMode = 'astra_db' # 'astra_db'/'local' session = getCQLSession(mode=cqlMode) keyspace = getCQLKeyspace(mode=cqlMode)

This is the prompt for a single message in the chat sequence.

We create it similarly as for a "stand-alone Cassandra prompt template".

In [3]:

systemTemplate = """
You are a chat assistant, helping a user of age {user_age} from a city
they refer to as {city_nickname}.
"""

systemTemplate = """ You are a chat assistant, helping a user of age {user_age} from a city they refer to as {city_nickname}. """

In [4]:

cassSystemPrompt = createCassandraPromptTemplate(
    session=session,
    keyspace=keyspace,
    template=systemTemplate,
    input_variables=['city', 'name'],
    field_mapper={
        'user_age': ('people', 'age'),
        'city_nickname': ('nickname_by_city', 'nickname'),
    },
)

cassSystemPrompt = createCassandraPromptTemplate( session=session, keyspace=keyspace, template=systemTemplate, input_variables=['city', 'name'], field_mapper={ 'user_age': ('people', 'age'), 'city_nickname': ('nickname_by_city', 'nickname'), }, )

In [5]:

from langchain.prompts import (
    ChatPromptTemplate,
    SystemMessagePromptTemplate,
    HumanMessagePromptTemplate,
)
systemMessagePrompt = SystemMessagePromptTemplate(prompt=cassSystemPrompt)

from langchain.prompts import ( ChatPromptTemplate, SystemMessagePromptTemplate, HumanMessagePromptTemplate, ) systemMessagePrompt = SystemMessagePromptTemplate(prompt=cassSystemPrompt)

A sequence of messages¶

Once we wrapped a single prompt template as a "system message prompt", let's make it part of a longer chat conversation:

In [6]:

humanTemplate = "{text}"
humanMessagePrompt = HumanMessagePromptTemplate.from_template(humanTemplate)

humanTemplate = "{text}" humanMessagePrompt = HumanMessagePromptTemplate.from_template(humanTemplate)

In [7]:

cassChatPrompt = ChatPromptTemplate.from_messages(
    [systemMessagePrompt, humanMessagePrompt]
)

cassChatPrompt = ChatPromptTemplate.from_messages( [systemMessagePrompt, humanMessagePrompt] )

Rendering¶

LangChain takes care of correctly propagating the rendering steps throughout the sequence of messages, including the Cassandra-backed template:

In [8]:

print(cassChatPrompt.format_prompt(
    city='turin',
    name='beppe',
    text='Assistant, please help me!'
).to_string())

print(cassChatPrompt.format_prompt( city='turin', name='beppe', text='Assistant, please help me!' ).to_string())

System: 
You are a chat assistant, helping a user of age 2 from a city
they refer to as CereaNeh.

Human: Assistant, please help me!