Langchain Output Parsers

from langfuse.callback import CallbackHandler

handler = CallbackHandler()

from langchain.llms import OpenAI
from langchain.output_parsers import PydanticOutputParser
from langchain.prompts import PromptTemplate
from langchain_core.pydantic_v1 import BaseModel, Field, validator

model = OpenAI(model_name="text-davinci-003", temperature=0.0)


# Define your desired data structure.
class Joke(BaseModel):
    setup: str = Field(description="question to set up a joke")
    punchline: str = Field(description="answer to resolve the joke")

    # You can add custom validation logic easily with Pydantic.
    @validator("setup")
    def question_ends_with_question_mark(cls, field):
        if field[-1] != "?":
            raise ValueError("Badly formed question!")
        return field


# Set up a parser + inject instructions into the prompt template.
parser = PydanticOutputParser(pydantic_object=Joke)

prompt = PromptTemplate(
    template="Answer the user query.\n{format_instructions}\n{query}\n",
    input_variables=["query"],
    partial_variables={"format_instructions": parser.get_format_instructions()},
)

chain = prompt | model | parser
chain.invoke({"query": "Tell me a joke."}, {"callbacks":[handler]})

from langfuse.callback import CallbackHandler

handler = CallbackHandler()

from langchain.llms import OpenAI
from langchain.output_parsers import PydanticOutputParser
from langchain.prompts import PromptTemplate
from langchain_core.pydantic_v1 import BaseModel, Field, validator

model = OpenAI(model_name="text-davinci-003", temperature=0.0)


# Define your desired data structure.
class Joke(BaseModel):
    setup: str = Field(description="question to set up a joke")
    punchline: str = Field(description="answer to resolve the joke")

    # You can add custom validation logic easily with Pydantic.
    @validator("setup")
    def question_ends_with_question_mark(cls, field):
        if field[-1] != "?":
            raise ValueError("Badly formed question!")
        return field


# Set up a parser + inject instructions into the prompt template.
parser = PydanticOutputParser(pydantic_object=Joke)

prompt = PromptTemplate(
    template="Answer the user query.\n{format_instructions}\n{query}\n",
    input_variables=["query"],
    partial_variables={"format_instructions": parser.get_format_instructions()},
)

chain = prompt | model | parser
chain.invoke({"query": "Tell me a joke."}, {"callbacks":[handler]})

class WForceConversationSummarizationTask:
    def get_chain(self):
        template = """Read a support chat between a customer and a support agent. \
Based on the informations given in the chat, answer in a few words (max. 80) what is the main subject of the conversation,\
stay conscious and be precise, explain what was defined and the result of the service without inferring information. At the end, give a list of 5 keywords which address the main topics, separated by comma. \
You must answer in Portuguese, with only a json containing two attributes: text and keywords.\
The suport chat data is: {question} \
"""

        prompt = PromptTemplate(template=template, input_variables=["question"])

        return LLMChain(
            llm=AzureChatOpenAI(openai_api_version=os.environ["AZURE_OPENAI_API_VERSION"], azure_deployment=os.environ["AZURE_GPT35_DEPLOYMENT_NAME"], temperature=0.3),
            prompt=prompt,
            output_parser=WForceConversationSummarizationOutputParser()
        )

class WForceConversationSummarizationTask:
    def get_chain(self):
        template = """Read a support chat between a customer and a support agent. \
Based on the informations given in the chat, answer in a few words (max. 80) what is the main subject of the conversation,\
stay conscious and be precise, explain what was defined and the result of the service without inferring information. At the end, give a list of 5 keywords which address the main topics, separated by comma. \
You must answer in Portuguese, with only a json containing two attributes: text and keywords.\
The suport chat data is: {question} \
"""

        prompt = PromptTemplate(template=template, input_variables=["question"])

        return LLMChain(
            llm=AzureChatOpenAI(openai_api_version=os.environ["AZURE_OPENAI_API_VERSION"], azure_deployment=os.environ["AZURE_GPT35_DEPLOYMENT_NAME"], temperature=0.3),
            prompt=prompt,
            output_parser=WForceConversationSummarizationOutputParser()
        )

task = WForceConversationSummarizationTask()
response = task.get_chain().invoke(input={"question": conversation}, config={"callbacks": [CallbackHandler()]})

task = WForceConversationSummarizationTask()
response = task.get_chain().invoke(input={"question": conversation}, config={"callbacks": [CallbackHandler()]})