refiner

Attributes

log

Classes

JanusRefiner	Base class to parse the output of an LLM call.
SimpleRetry	Base class to parse the output of an LLM call.
FixParserExceptions	Base class to parse the output of an LLM call.
ReflectionRefiner	Base class to parse the output of an LLM call.
RequirementsReflectionRefiner	This requirements-specific refiner is intended to address a common issue with
HallucinationRefiner	Base class to parse the output of an LLM call.

Module Contents

refiner.log

class refiner.JanusRefiner

Bases: janus.parsers.parser.JanusParser

Base class to parse the output of an LLM call.

Output parsers help structure language model responses.

Example

class BooleanOutputParser(BaseOutputParser[bool]):
    true_val: str = "YES"
    false_val: str = "NO"

    def parse(self, text: str) -> bool:
        cleaned_text = text.strip().upper()
        if cleaned_text not in (self.true_val.upper(), self.false_val.upper()):
            raise OutputParserException(
                f"BooleanOutputParser expected output value to either be "
                f"{self.true_val} or {self.false_val} (case-insensitive). "
                f"Received {cleaned_text}."
            )
        return cleaned_text == self.true_val.upper()

    @property
    def _type(self) -> str:
        return "boolean_output_parser"

parser: janus.parsers.parser.JanusParser

parse_runnable(input)

Parameters:

input (dict[str, Any]) –

Return type:

Any

parse_completion(completion, **kwargs)

Parameters:

completion (str) –

Return type:

Any

abstract parse(text)

Parse a single string model output into some structure.

Parameters:

text (str) – String output of a language model.

Returns:

Structured output.

Return type:

str

class refiner.SimpleRetry(llm, parser, max_retries)

Bases: JanusRefiner

Base class to parse the output of an LLM call.

Output parsers help structure language model responses.

Example

class BooleanOutputParser(BaseOutputParser[bool]):
    true_val: str = "YES"
    false_val: str = "NO"

    def parse(self, text: str) -> bool:
        cleaned_text = text.strip().upper()
        if cleaned_text not in (self.true_val.upper(), self.false_val.upper()):
            raise OutputParserException(
                f"BooleanOutputParser expected output value to either be "
                f"{self.true_val} or {self.false_val} (case-insensitive). "
                f"Received {cleaned_text}."
            )
        return cleaned_text == self.true_val.upper()

    @property
    def _type(self) -> str:
        return "boolean_output_parser"

Parameters:

• llm (janus.llm.models_info.JanusModel) –

• parser (janus.parsers.parser.JanusParser) –

• max_retries (int) –

max_retries: int

retry_chain: langchain_core.runnables.RunnableSerializable

parse_completion(completion, prompt_value, **kwargs)

Parameters:

• completion (str) –

• prompt_value (langchain_core.prompt_values.PromptValue) –

Return type:

Any

class refiner.FixParserExceptions(llm, parser, max_retries)

Bases: JanusRefiner, langchain.output_parsers.RetryWithErrorOutputParser

Base class to parse the output of an LLM call.

Output parsers help structure language model responses.

Example

class BooleanOutputParser(BaseOutputParser[bool]):
    true_val: str = "YES"
    false_val: str = "NO"

    def parse(self, text: str) -> bool:
        cleaned_text = text.strip().upper()
        if cleaned_text not in (self.true_val.upper(), self.false_val.upper()):
            raise OutputParserException(
                f"BooleanOutputParser expected output value to either be "
                f"{self.true_val} or {self.false_val} (case-insensitive). "
                f"Received {cleaned_text}."
            )
        return cleaned_text == self.true_val.upper()

    @property
    def _type(self) -> str:
        return "boolean_output_parser"

Parameters:

• llm (janus.llm.models_info.JanusModel) –

• parser (janus.parsers.parser.JanusParser) –

• max_retries (int) –

parse_completion(completion, prompt_value, **kwargs)

Parameters:

• completion (str) –

• prompt_value (langchain_core.prompt_values.PromptValue) –

Return type:

Any

class refiner.ReflectionRefiner(llm, parser, max_retries, prompt_template_name='refinement/reflection')

Bases: JanusRefiner

Base class to parse the output of an LLM call.

Output parsers help structure language model responses.

Example

class BooleanOutputParser(BaseOutputParser[bool]):
    true_val: str = "YES"
    false_val: str = "NO"

    def parse(self, text: str) -> bool:
        cleaned_text = text.strip().upper()
        if cleaned_text not in (self.true_val.upper(), self.false_val.upper()):
            raise OutputParserException(
                f"BooleanOutputParser expected output value to either be "
                f"{self.true_val} or {self.false_val} (case-insensitive). "
                f"Received {cleaned_text}."
            )
        return cleaned_text == self.true_val.upper()

    @property
    def _type(self) -> str:
        return "boolean_output_parser"

Parameters:

• llm (janus.llm.models_info.JanusModel) –

• parser (janus.parsers.parser.JanusParser) –

• max_retries (int) –

• prompt_template_name (str) –

max_retries: int

reflection_chain: langchain_core.runnables.RunnableSerializable

revision_chain: langchain_core.runnables.RunnableSerializable

reflection_prompt_name: str

parse_completion(completion, prompt_value, **kwargs)

Parameters:

• completion (str) –

• prompt_value (langchain_core.prompt_values.PromptValue) –

Return type:

Any

class refiner.RequirementsReflectionRefiner(llm, parser, max_retries, prompt_template_name='refinement/reflection/incose')

Bases: JanusRefiner

This requirements-specific refiner is intended to address a common issue with requirements reflection, where over the course of several reflection loops, requirements lists grow increasingly verbose, eventually becoming too wordy to be useful. To reduce this, this refiner interlaces an additional reflection -> revision loop which de-duplicates requirements.

Parameters:

• llm (janus.llm.models_info.JanusModel) –

• parser (janus.parsers.parser.JanusParser) –

• max_retries (int) –

• prompt_template_name (str) –

max_retries: int

reflection_chain: langchain_core.runnables.RunnableSerializable

revision_chain: langchain_core.runnables.RunnableSerializable

reflect_duplication_chain: langchain_core.runnables.RunnableSerializable

revise_duplication_chain: langchain_core.runnables.RunnableSerializable

reflection_prompt_name: str

parse_completion(completion, prompt_value, **kwargs)

Parameters:

• completion (str) –

• prompt_value (langchain_core.prompt_values.PromptValue) –

Return type:

Any

class refiner.HallucinationRefiner(**kwargs)

Bases: ReflectionRefiner

Base class to parse the output of an LLM call.

Output parsers help structure language model responses.

Example

class BooleanOutputParser(BaseOutputParser[bool]):
    true_val: str = "YES"
    false_val: str = "NO"

    def parse(self, text: str) -> bool:
        cleaned_text = text.strip().upper()
        if cleaned_text not in (self.true_val.upper(), self.false_val.upper()):
            raise OutputParserException(
                f"BooleanOutputParser expected output value to either be "
                f"{self.true_val} or {self.false_val} (case-insensitive). "
                f"Received {cleaned_text}."
            )
        return cleaned_text == self.true_val.upper()

    @property
    def _type(self) -> str:
        return "boolean_output_parser"