experimental: docstrings update (#18048)

Added missed docstrings. Formatted docsctrings to the consistent format.

libs/experimental/langchain_experimental/autonomous_agents/autogpt/agent.py

@@ -26,7 +26,7 @@ from langchain_experimental.pydantic_v1 import ValidationError
 
 
 class AutoGPT:
-    """Agent class for interacting with Auto-GPT."""
+    """Agent for interacting with AutoGPT."""
 
     def __init__(
         self,

libs/experimental/langchain_experimental/autonomous_agents/autogpt/prompt_generator.py

@@ -7,7 +7,7 @@ FINISH_NAME = "finish"
 
 
 class PromptGenerator:
-    """A class for generating custom prompt strings.
+    """Generator of custom prompt strings.
 
     Does this based on constraints, commands, resources, and performance evaluations.
     """

libs/experimental/langchain_experimental/autonomous_agents/hugginggpt/hugginggpt.py

@@ -15,6 +15,8 @@ from langchain_experimental.autonomous_agents.hugginggpt.task_planner import (
 
 
 class HuggingGPT:
+    """Agent for interacting with HuggingGPT."""
+
     def __init__(self, llm: BaseLanguageModel, tools: List[BaseTool]):
         self.llm = llm
         self.tools = tools

libs/experimental/langchain_experimental/autonomous_agents/hugginggpt/repsonse_generator.py

@@ -25,6 +25,8 @@ class ResponseGenerationChain(LLMChain):
 
 
 class ResponseGenerator:
+    """Generates a response based on the input."""
+
     def __init__(self, llm_chain: LLMChain, stop: Optional[List] = None):
         self.llm_chain = llm_chain
         self.stop = stop
@@ -36,6 +38,8 @@ class ResponseGenerator:
 
 
 def load_response_generator(llm: BaseLanguageModel) -> ResponseGenerator:
+    """Load the ResponseGenerator."""
+
     llm_chain = ResponseGenerationChain.from_llm(llm)
     return ResponseGenerator(
         llm_chain=llm_chain,

libs/experimental/langchain_experimental/autonomous_agents/hugginggpt/task_executor.py

@@ -9,6 +9,8 @@ from langchain_experimental.autonomous_agents.hugginggpt.task_planner import Pla
 
 
 class Task:
+    """Task to be executed."""
+
     def __init__(self, task: str, id: int, dep: List[int], args: Dict, tool: BaseTool):
         self.task = task
         self.id = id
@@ -74,7 +76,7 @@ class Task:
 
 
 class TaskExecutor:
-    """Load tools to execute tasks."""
+    """Load tools and execute tasks."""
 
     def __init__(self, plan: Plan):
         self.plan = plan

libs/experimental/langchain_experimental/autonomous_agents/hugginggpt/task_planner.py

@@ -76,6 +76,8 @@ class TaskPlaningChain(LLMChain):
 
 
 class Step:
+    """A step in the plan."""
+
     def __init__(
         self, task: str, id: int, dep: List[int], args: Dict[str, str], tool: BaseTool
     ):
@@ -87,6 +89,8 @@ class Step:
 
 
 class Plan:
+    """A plan to execute."""
+
     def __init__(self, steps: List[Step]):
         self.steps = steps
 
@@ -98,6 +102,8 @@ class Plan:
 
 
 class BasePlanner(BaseModel):
+    """Base class for a planner."""
+
     @abstractmethod
     def plan(self, inputs: dict, callbacks: Callbacks = None, **kwargs: Any) -> Plan:
         """Given input, decide what to do."""
@@ -106,11 +112,22 @@ class BasePlanner(BaseModel):
     async def aplan(
         self, inputs: dict, callbacks: Callbacks = None, **kwargs: Any
     ) -> Plan:
-        """Given input, decide what to do."""
+        """Asynchronous Given input, decide what to do."""
 
 
 class PlanningOutputParser(BaseModel):
+    """Parses the output of the planning stage."""
+
     def parse(self, text: str, hf_tools: List[BaseTool]) -> Plan:
+        """Parse the output of the planning stage.
+
+        Args:
+            text: The output of the planning stage.
+            hf_tools: The tools available.
+
+        Returns:
+            The plan.
+        """
         steps = []
         for v in json.loads(re.findall(r"\[.*\]", text)[0]):
             choose_tool = None
@@ -124,6 +141,8 @@ class PlanningOutputParser(BaseModel):
 
 
 class TaskPlanner(BasePlanner):
+    """Planner for tasks."""
+
     llm_chain: LLMChain
     output_parser: PlanningOutputParser
     stop: Optional[List] = None
@@ -139,7 +158,7 @@ class TaskPlanner(BasePlanner):
     async def aplan(
         self, inputs: dict, callbacks: Callbacks = None, **kwargs: Any
     ) -> Plan:
-        """Given input, decided what to do."""
+        """Asynchronous Given input, decided what to do."""
         inputs["hf_tools"] = [
             f"{tool.name}: {tool.description}" for tool in inputs["hf_tools"]
         ]
@@ -150,5 +169,7 @@ class TaskPlanner(BasePlanner):
 
 
 def load_chat_planner(llm: BaseLanguageModel) -> TaskPlanner:
+    """Load the chat planner."""
+
     llm_chain = TaskPlaningChain.from_llm(llm)
     return TaskPlanner(llm_chain=llm_chain, output_parser=PlanningOutputParser())

libs/experimental/langchain_experimental/chat_models/llm_wrapper.py

@@ -24,6 +24,8 @@ If a question does not make any sense, or is not factually coherent, explain why
 
 
 class ChatWrapper(BaseChatModel):
+    """Wrapper for chat LLMs."""
+
     llm: LLM
     sys_beg: str
     sys_end: str
@@ -130,6 +132,8 @@ class ChatWrapper(BaseChatModel):
 
 
 class Llama2Chat(ChatWrapper):
+    """Wrapper for Llama-2-chat model."""
+
     @property
     def _llm_type(self) -> str:
         return "llama-2-chat"
@@ -145,6 +149,8 @@ class Llama2Chat(ChatWrapper):
 
 
 class Orca(ChatWrapper):
+    """Wrapper for Orca-style models."""
+
     @property
     def _llm_type(self) -> str:
         return "orca-style"
@@ -158,6 +164,8 @@ class Orca(ChatWrapper):
 
 
 class Vicuna(ChatWrapper):
+    """Wrapper for Vicuna-style models."""
+
     @property
     def _llm_type(self) -> str:
         return "vicuna-style"

libs/experimental/langchain_experimental/comprehend_moderation/amazon_comprehend_moderation.py

@@ -14,7 +14,10 @@ from langchain_experimental.pydantic_v1 import root_validator
 
 
 class AmazonComprehendModerationChain(Chain):
-    """A subclass of Chain, designed to apply moderation to LLMs."""
+    """Moderation Chain, based on `Amazon Comprehend` service.
+
+    See more at https://aws.amazon.com/comprehend/
+    """
 
     output_key: str = "output"  #: :meta private:
     """Key used to fetch/store the output in data containers. Defaults to `output`"""
@@ -54,7 +57,7 @@ class AmazonComprehendModerationChain(Chain):
     @root_validator(pre=True)
     def create_client(cls, values: Dict[str, Any]) -> Dict[str, Any]:
         """
-        Creates an Amazon Comprehend client
+        Creates an Amazon Comprehend client.
 
         Args:
             values (Dict[str, Any]): A dictionary containing configuration values.

libs/experimental/langchain_experimental/comprehend_moderation/base_moderation.py

@@ -13,6 +13,8 @@ from langchain_experimental.comprehend_moderation.toxicity import ComprehendToxi
 
 
 class BaseModeration:
+    """Base class for moderation."""
+
     def __init__(
         self,
         client: Any,
@@ -109,6 +111,8 @@ class BaseModeration:
             self.run_manager.on_text(message)
 
     def moderate(self, prompt: Any) -> str:
+        """Moderate the input prompt."""
+
         from langchain_experimental.comprehend_moderation.base_moderation_config import (  # noqa: E501
             ModerationPiiConfig,
             ModerationPromptSafetyConfig,

libs/experimental/langchain_experimental/comprehend_moderation/base_moderation_callbacks.py

@@ -2,6 +2,8 @@ from typing import Any, Callable, Dict
 
 
 class BaseModerationCallbackHandler:
+    """Base class for moderation callback handlers."""
+
     def __init__(self) -> None:
         if (
             self._is_method_unchanged(

libs/experimental/langchain_experimental/comprehend_moderation/base_moderation_config.py

@@ -4,6 +4,8 @@ from pydantic import BaseModel
 
 
 class ModerationPiiConfig(BaseModel):
+    """Configuration for PII moderation filter."""
+
     threshold: float = 0.5
     """Threshold for PII confidence score, defaults to 0.5 i.e. 50%"""
 
@@ -21,6 +23,8 @@ class ModerationPiiConfig(BaseModel):
 
 
 class ModerationToxicityConfig(BaseModel):
+    """Configuration for Toxicity moderation filter."""
+
     threshold: float = 0.5
     """Threshold for Toxic label confidence score, defaults to 0.5 i.e. 50%"""
 
@@ -29,6 +33,8 @@ class ModerationToxicityConfig(BaseModel):
 
 
 class ModerationPromptSafetyConfig(BaseModel):
+    """Configuration for Prompt Safety moderation filter."""
+
     threshold: float = 0.5
     """
     Threshold for Prompt Safety classification
@@ -37,6 +43,8 @@ class ModerationPromptSafetyConfig(BaseModel):
 
 
 class BaseModerationConfig(BaseModel):
+    """Base configuration settings for moderation."""
+
     filters: List[
         Union[
             ModerationPiiConfig, ModerationToxicityConfig, ModerationPromptSafetyConfig

libs/experimental/langchain_experimental/comprehend_moderation/base_moderation_exceptions.py

@@ -27,7 +27,7 @@ class ModerationToxicityError(Exception):
 
 
 class ModerationPromptSafetyError(Exception):
-    """Exception raised if Intention entities are detected.
+    """Exception raised if Unsafe prompts are detected.
 
     Attributes:
         message -- explanation of the error

libs/experimental/langchain_experimental/comprehend_moderation/pii.py

@@ -7,6 +7,8 @@ from langchain_experimental.comprehend_moderation.base_moderation_exceptions imp
 
 
 class ComprehendPII:
+    """Class to handle Personally Identifiable Information (PII) moderation."""
+
     def __init__(
         self,
         client: Any,

libs/experimental/langchain_experimental/comprehend_moderation/prompt_safety.py

@@ -7,6 +7,8 @@ from langchain_experimental.comprehend_moderation.base_moderation_exceptions imp
 
 
 class ComprehendPromptSafety:
+    """Class to handle prompt safety moderation."""
+
     def __init__(
         self,
         client: Any,

libs/experimental/langchain_experimental/comprehend_moderation/toxicity.py

@@ -8,6 +8,8 @@ from langchain_experimental.comprehend_moderation.base_moderation_exceptions imp
 
 
 class ComprehendToxicity:
+    """Class to handle toxicity moderation."""
+
     def __init__(
         self,
         client: Any,

libs/experimental/langchain_experimental/cpal/base.py

@@ -105,7 +105,7 @@ class _BaseStoryElementChain(Chain):
 
 
 class NarrativeChain(_BaseStoryElementChain):
-    """Decompose the narrative into its story elements
+    """Decompose the narrative into its story elements.
 
     - causal model
     - query

libs/experimental/langchain_experimental/cpal/models.py

@@ -17,7 +17,7 @@ from langchain_experimental.pydantic_v1 import (
 
 class NarrativeModel(BaseModel):
     """
-    Represent the narrative input as three story elements.
+    Narrative input as three story elements.
     """
 
     story_outcome_question: str
@@ -33,6 +33,8 @@ class NarrativeModel(BaseModel):
 
 
 class EntityModel(BaseModel):
+    """Entity in the story."""
+
     name: str = Field(description="entity name")
     code: str = Field(description="entity actions")
     value: float = Field(description="entity initial value")
@@ -51,6 +53,8 @@ class EntityModel(BaseModel):
 
 
 class CausalModel(BaseModel):
+    """Casual data."""
+
     attribute: str = Field(description="name of the attribute to be calculated")
     entities: List[EntityModel] = Field(description="entities in the story")
 
@@ -58,7 +62,8 @@ class CausalModel(BaseModel):
 
 
 class EntitySettingModel(BaseModel):
-    """
+    """Entity initial conditions.
+
     Initial conditions for an entity
 
     {"name": "bud", "attribute": "pet_count", "value": 12}
@@ -75,7 +80,8 @@ class EntitySettingModel(BaseModel):
 
 
 class SystemSettingModel(BaseModel):
-    """
+    """System initial conditions.
+
     Initial global conditions for the system.
 
     {"parameter": "interest_rate", "value": .05}
@@ -86,8 +92,7 @@ class SystemSettingModel(BaseModel):
 
 
 class InterventionModel(BaseModel):
-    """
-    aka initial conditions
+    """Intervention data of the story aka initial conditions.
 
     >>> intervention.dict()
     {
@@ -110,7 +115,9 @@ class InterventionModel(BaseModel):
 
 
 class QueryModel(BaseModel):
-    """translate a question about the story outcome into a programmatic expression"""
+    """Query data of the story.
+
+    translate a question about the story outcome into a programmatic expression"""
 
     question: str = Field(alias=Constant.narrative_input.value)  # input
     expression: str  # output, part of llm completion
@@ -119,11 +126,15 @@ class QueryModel(BaseModel):
 
 
 class ResultModel(BaseModel):
+    """Result of the story query."""
+
     question: str = Field(alias=Constant.narrative_input.value)  # input
     _result_table: str = PrivateAttr()  # result of the executed query
 
 
 class StoryModel(BaseModel):
+    """Story data."""
+
     causal_operations: Any = Field(required=True)
     intervention: Any = Field(required=True)
     query: Any = Field(required=True)

libs/experimental/langchain_experimental/data_anonymizer/base.py

@@ -10,8 +10,8 @@ DEFAULT_DEANONYMIZER_MATCHING_STRATEGY = exact_matching_strategy
 
 
 class AnonymizerBase(ABC):
-    """
-    Base abstract class for anonymizers.
+    """Base abstract class for anonymizers.
+
     It is public and non-virtual because it allows
         wrapping the behavior for all methods in a base class.
     """
@@ -22,7 +22,8 @@ class AnonymizerBase(ABC):
         language: Optional[str] = None,
         allow_list: Optional[List[str]] = None,
     ) -> str:
-        """Anonymize text"""
+        """Anonymize text."""
+
         return self._anonymize(text, language, allow_list)
 
     @abstractmethod

libs/experimental/langchain_experimental/data_anonymizer/deanonymizer_mapping.py

@@ -11,7 +11,7 @@ MappingDataType = Dict[str, Dict[str, str]]
 
 
 def format_duplicated_operator(operator_name: str, count: int) -> str:
-    """Format the operator name with the count"""
+    """Format the operator name with the count."""
 
     clean_operator_name = re.sub(r"[<>]", "", operator_name)
     clean_operator_name = re.sub(r"_\d+$", "", clean_operator_name)
@@ -24,17 +24,20 @@ def format_duplicated_operator(operator_name: str, count: int) -> str:
 
 @dataclass
 class DeanonymizerMapping:
+    """Deanonymizer mapping."""
+
     mapping: MappingDataType = field(
         default_factory=lambda: defaultdict(lambda: defaultdict(str))
     )
 
     @property
     def data(self) -> MappingDataType:
-        """Return the deanonymizer mapping"""
+        """Return the deanonymizer mapping."""
         return {k: dict(v) for k, v in self.mapping.items()}
 
     def update(self, new_mapping: MappingDataType) -> None:
-        """Update the deanonymizer mapping with new values
+        """Update the deanonymizer mapping with new values.
+
         Duplicated values will not be added
         If there are multiple entities of the same type, the mapping will
         include a count to differentiate them. For example, if there are
@@ -67,7 +70,8 @@ def create_anonymizer_mapping(
     anonymizer_results: "EngineResult",
     is_reversed: bool = False,
 ) -> MappingDataType:
-    """Creates or updates the mapping used to anonymize and/or deanonymize text.
+    """Create or update the mapping used to anonymize and/or
+     deanonymize a text.
 
     This method exploits the results returned by the
     analysis and anonymization processes.

libs/experimental/langchain_experimental/data_anonymizer/deanonymizer_matching_strategies.py

@@ -5,8 +5,8 @@ from langchain_experimental.data_anonymizer.deanonymizer_mapping import MappingD
 
 
 def exact_matching_strategy(text: str, deanonymizer_mapping: MappingDataType) -> str:
-    """
-    Exact matching strategy for deanonymization.
+    """Exact matching strategy for deanonymization.
+
     It replaces all the anonymized entities with the original ones.
 
     Args:
@@ -23,8 +23,8 @@ def exact_matching_strategy(text: str, deanonymizer_mapping: MappingDataType) ->
 def case_insensitive_matching_strategy(
     text: str, deanonymizer_mapping: MappingDataType
 ) -> str:
-    """
-    Case insensitive matching strategy for deanonymization.
+    """Case insensitive matching strategy for deanonymization.
+
     It replaces all the anonymized entities with the original ones
         irrespective of their letter case.
 
@@ -48,8 +48,8 @@ def case_insensitive_matching_strategy(
 def fuzzy_matching_strategy(
     text: str, deanonymizer_mapping: MappingDataType, max_l_dist: int = 3
 ) -> str:
-    """
-    Fuzzy matching strategy for deanonymization.
+    """Fuzzy matching strategy for deanonymization.
+
     It uses fuzzy matching to find the position of the anonymized entity in the text.
     It replaces all the anonymized entities with the original ones.
 
@@ -93,9 +93,9 @@ def fuzzy_matching_strategy(
 def combined_exact_fuzzy_matching_strategy(
     text: str, deanonymizer_mapping: MappingDataType, max_l_dist: int = 3
 ) -> str:
-    """
-    RECOMMENDED STRATEGY.
-    Combined exact and fuzzy matching strategy for deanonymization.
+    """Combined exact and fuzzy matching strategy for deanonymization.
+
+    It is a RECOMMENDED STRATEGY.
 
     Args:
         text: text to deanonymize
@@ -118,8 +118,8 @@ def ngram_fuzzy_matching_strategy(
     fuzzy_threshold: int = 85,
     use_variable_length: bool = True,
 ) -> str:
-    """
-    N-gram fuzzy matching strategy for deanonymization.
+    """N-gram fuzzy matching strategy for deanonymization.
+
     It replaces all the anonymized entities with the original ones.
     It uses fuzzy matching to find the position of the anonymized entity in the text.
     It generates n-grams of the same length as the anonymized entity from the text and

libs/experimental/langchain_experimental/data_anonymizer/faker_presidio_mapping.py

@@ -3,6 +3,8 @@ from typing import Callable, Dict, Optional
 
 
 def get_pseudoanonymizer_mapping(seed: Optional[int] = None) -> Dict[str, Callable]:
+    """Get a mapping of entities to pseudo anonymize them."""
+
     try:
         from faker import Faker
     except ImportError as e:

libs/experimental/langchain_experimental/data_anonymizer/presidio.py

@@ -98,6 +98,11 @@ DEFAULT_LANGUAGES_CONFIG = {
 
 
 class PresidioAnonymizerBase(AnonymizerBase):
+    """Base Anonymizer using Microsoft Presidio.
+
+    See more: https://microsoft.github.io/presidio/
+    """
+
     def __init__(
         self,
         analyzed_fields: Optional[List[str]] = None,
@@ -180,6 +185,8 @@ class PresidioAnonymizerBase(AnonymizerBase):
 
 
 class PresidioAnonymizer(PresidioAnonymizerBase):
+    """Anonymizer using Microsoft Presidio."""
+
     def _anonymize(
         self,
         text: str,
@@ -258,6 +265,8 @@ class PresidioAnonymizer(PresidioAnonymizerBase):
 
 
 class PresidioReversibleAnonymizer(PresidioAnonymizerBase, ReversibleAnonymizerBase):
+    """Reversible Anonymizer using Microsoft Presidio."""
+
     def __init__(
         self,
         analyzed_fields: Optional[List[str]] = None,

libs/experimental/langchain_experimental/fallacy_removal/base.py

@@ -18,9 +18,10 @@ from langchain_experimental.fallacy_removal.prompts import (
 
 
 class FallacyChain(Chain):
-    """Chain for applying logical fallacy evaluations, modeled after Constitutional AI \
-    and in same format, but applying logical fallacies as generalized rules to remove \
-    in output
+    """Chain for applying logical fallacy evaluations.
+
+    It is modeled after Constitutional AI and in same format, but
+    applying logical fallacies as generalized rules to remove in output.
 
     Example:
         .. code-block:: python

libs/experimental/langchain_experimental/fallacy_removal/models.py

@@ -3,7 +3,7 @@ from langchain_experimental.pydantic_v1 import BaseModel
 
 
 class LogicalFallacy(BaseModel):
-    """Class for a logical fallacy."""
+    """Logical fallacy."""
 
     fallacy_critique_request: str
     fallacy_revision_request: str

libs/experimental/langchain_experimental/generative_agents/generative_agent.py

@@ -11,7 +11,7 @@ from langchain_experimental.pydantic_v1 import BaseModel, Field
 
 
 class GenerativeAgent(BaseModel):
-    """An Agent as a character with memory and innate characteristics."""
+    """Agent as a character with memory and innate characteristics."""
 
     name: str
     """The character's name."""
@@ -48,6 +48,8 @@ class GenerativeAgent(BaseModel):
         return [re.sub(r"^\s*\d+\.\s*", "", line).strip() for line in lines]
 
     def chain(self, prompt: PromptTemplate) -> LLMChain:
+        """Create a chain with the same settings as the agent."""
+
         return LLMChain(
             llm=self.llm, prompt=prompt, verbose=self.verbose, memory=self.memory
         )

libs/experimental/langchain_experimental/graph_transformers/diffbot.py

@@ -7,6 +7,8 @@ from langchain_core.documents import Document
 
 
 def format_property_key(s: str) -> str:
+    """Formats a string to be used as a property key."""
+
     words = s.split()
     if not words:
         return s
@@ -16,8 +18,7 @@ def format_property_key(s: str) -> str:
 
 
 class NodesList:
-    """
-    Manages a list of nodes with associated properties.
+    """List of nodes with associated properties.
 
     Attributes:
         nodes (Dict[Tuple, Any]): Stores nodes as keys and their properties as values.
@@ -85,8 +86,7 @@ schema_mapping = [
 
 
 class SimplifiedSchema:
-    """
-    Provides functionality for working with a simplified schema mapping.
+    """Simplified schema mapping.
 
     Attributes:
         schema (Dict): A dictionary containing the mapping to simplified schema types.
@@ -116,7 +116,7 @@ class SimplifiedSchema:
 
 
 class DiffbotGraphTransformer:
-    """Transforms documents into graph documents using Diffbot's NLP API.
+    """Transform documents into graph documents using Diffbot NLP API.
 
     A graph document transformation system takes a sequence of Documents and returns a
     sequence of Graph Documents.

libs/experimental/langchain_experimental/llm_bash/bash.py

@@ -12,8 +12,8 @@ if TYPE_CHECKING:
 
 
 class BashProcess:
-    """
-    Wrapper class for starting subprocesses.
+    """Wrapper for starting subprocesses.
+
     Uses the python built-in subprocesses.run()
     Persistent processes are **not** available
     on Windows systems, as pexpect makes use of

libs/experimental/langchain_experimental/llm_bash/prompt.py

@@ -31,6 +31,8 @@ class BashOutputParser(BaseOutputParser):
     """Parser for bash output."""
 
     def parse(self, text: str) -> List[str]:
+        """Parse the output of a bash command."""
+
         if "```bash" in text:
             return self.get_code_blocks(text)
         else:

libs/experimental/langchain_experimental/llm_symbolic_math/base.py

@@ -20,6 +20,10 @@ from langchain_experimental.pydantic_v1 import Extra
 class LLMSymbolicMathChain(Chain):
     """Chain that interprets a prompt and executes python code to do symbolic math.
 
+    It is based on the sympy library and can be used to evaluate
+    mathematical expressions.
+    See https://www.sympy.org/ for more information.
+
     Example:
         .. code-block:: python
 

libs/experimental/langchain_experimental/llms/anthropic_functions.py

@@ -41,6 +41,8 @@ for the weather in SF you would respond:
 
 
 class TagParser(HTMLParser):
+    """Parser for the tool tags."""
+
     def __init__(self) -> None:
         """A heavy-handed solution, but it's fast for prototyping.
 
@@ -122,6 +124,8 @@ def _destrip(tool_input: Any) -> Any:
 
 
 class AnthropicFunctions(BaseChatModel):
+    """Chat model for interacting with Anthropic functions."""
+
     llm: BaseChatModel
 
     @root_validator(pre=True)

libs/experimental/langchain_experimental/llms/jsonformer_decoder.py

@@ -14,7 +14,7 @@ if TYPE_CHECKING:
 
 
 def import_jsonformer() -> jsonformer:
-    """Lazily import jsonformer."""
+    """Lazily import of the jsonformer package."""
     try:
         import jsonformer
     except ImportError:

libs/experimental/langchain_experimental/llms/llamaapi.py

@@ -76,6 +76,8 @@ def _convert_message_to_dict(message: BaseMessage) -> dict:
 
 
 class ChatLlamaAPI(BaseChatModel):
+    """Chat model using the Llama API."""
+
     client: Any  #: :meta private:
 
     def _generate(

libs/experimental/langchain_experimental/llms/lmformatenforcer_decoder.py

@@ -14,7 +14,7 @@ if TYPE_CHECKING:
 
 
 def import_lmformatenforcer() -> lmformatenforcer:
-    """Lazily import lmformatenforcer."""
+    """Lazily import of the lmformatenforcer package."""
     try:
         import lmformatenforcer
     except ImportError:

libs/experimental/langchain_experimental/llms/ollama_functions.py

@@ -42,6 +42,8 @@ DEFAULT_RESPONSE_FUNCTION = {
 
 
 class OllamaFunctions(BaseChatModel):
+    """Function chat model that uses Ollama API."""
+
     llm: ChatOllama
 
     tool_system_prompt_template: str

libs/experimental/langchain_experimental/llms/rellm_decoder.py

@@ -20,7 +20,7 @@ else:
 
 
 def import_rellm() -> rellm:
-    """Lazily import rellm."""
+    """Lazily import of the rellm package."""
     try:
         import rellm
     except ImportError:

libs/experimental/langchain_experimental/open_clip/open_clip.py

@@ -5,6 +5,8 @@ from langchain_core.embeddings import Embeddings
 
 
 class OpenCLIPEmbeddings(BaseModel, Embeddings):
+    """OpenCLIP Embeddings model."""
+
     model: Any
     preprocess: Any
     tokenizer: Any

libs/experimental/langchain_experimental/pal_chain/base.py

@@ -34,6 +34,8 @@ COMMAND_EXECUTION_ATTRIBUTES = [
 
 
 class PALValidation:
+    """Validation for PAL generated code."""
+
     SOLUTION_EXPRESSION_TYPE_FUNCTION = ast.FunctionDef
     SOLUTION_EXPRESSION_TYPE_VARIABLE = ast.Name
 
@@ -95,7 +97,7 @@ class PALValidation:
 
 
 class PALChain(Chain):
-    """Implements Program-Aided Language Models (PAL).
+    """Chain that implements Program-Aided Language Models (PAL).
 
     This class implements the Program-Aided Language Models (PAL) for generating code
     solutions. PAL is a technique described in the paper "Program-Aided Language Models"

libs/experimental/langchain_experimental/plan_and_execute/schema.py

@@ -40,7 +40,7 @@ class BaseStepContainer(BaseModel):
 
 
 class ListStepContainer(BaseStepContainer):
-    """List step container."""
+    """Container for List of steps."""
 
     steps: List[Tuple[Step, StepResponse]] = Field(default_factory=list)
     """The steps."""

libs/experimental/langchain_experimental/prompt_injection_identifier/hugging_face_identifier.py

@@ -11,6 +11,8 @@ if TYPE_CHECKING:
 
 
 class PromptInjectionException(ValueError):
+    """Exception raised when prompt injection attack is detected."""
+
     def __init__(
         self, message: str = "Prompt injection attack detected", score: float = 1.0
     ):
@@ -48,7 +50,8 @@ def _model_default_factory(
 
 
 class HuggingFaceInjectionIdentifier(BaseTool):
-    """Tool that uses HF model to detect prompt injection attacks."""
+    """Tool that uses HuggingFace Prompt Injection to
+    detect prompt injection attacks."""
 
     name: str = "hugging_face_injection_identifier"
     description: str = (

libs/experimental/langchain_experimental/prompts/load.py

@@ -10,7 +10,8 @@ from langchain_core.prompts import BasePromptTemplate
 
 
 def load_prompt(path: Union[str, Path]) -> BasePromptTemplate:
-    """Unified method for loading a prompt from LangChainHub or local fs."""
+    """Unified method for loading a prompt from LangChainHub or local file system."""
+
     if hub_result := try_load_from_hub(
         path, _load_prompt_from_file, "prompts", {"py", "json", "yaml"}
     ):

libs/experimental/langchain_experimental/recommenders/amazon_personalize.py

@@ -84,7 +84,9 @@ class AmazonPersonalize:
         metadata_columns: Optional[Mapping[str, Sequence[str]]] = None,
         **kwargs: Any,
     ) -> Mapping[str, Any]:
-        """Get recommendations from Amazon Personalize:
+        """Get recommendations from Amazon Personalize service.
+
+        See more details at:
         https://docs.aws.amazon.com/personalize/latest/dg/API_RS_GetRecommendations.html
 
         Args:
@@ -151,6 +153,7 @@ class AmazonPersonalize:
         **kwargs: Any,
     ) -> Mapping[str, Any]:
         """Re-ranks a list of recommended items for the given user.
+
         https://docs.aws.amazon.com/personalize/latest/dg/API_RS_GetPersonalizedRanking.html
 
         Args:

libs/experimental/langchain_experimental/recommenders/amazon_personalize_chain.py

@@ -39,12 +39,12 @@ RESULT_OUTPUT_KEY = "result"
 
 
 class AmazonPersonalizeChain(Chain):
-    """Amazon Personalize Chain for retrieving recommendations
-                        from Amazon Personalize, and summarizing
-    the recommendations in natural language.
-                    It will only return recommendations if return_direct=True.
-    Can also be used in sequential chains for working with
-                                                the output of Amazon Personalize.
+    """Chain for retrieving recommendations from Amazon Personalize,
+     and summarizing them.
+
+    It only returns recommendations if return_direct=True.
+    It can also be used in sequential chains for working with
+    the output of Amazon Personalize.
 
     Example:
         .. code-block:: python

libs/experimental/langchain_experimental/retrievers/vector_sql_database.py

@@ -13,7 +13,7 @@ from langchain_experimental.sql.vector_sql import VectorSQLDatabaseChain
 
 
 class VectorSQLDatabaseChainRetriever(BaseRetriever):
-    """Retriever that uses SQLDatabase as Retriever"""
+    """Retriever that uses Vector SQL Database."""
 
     sql_db_chain: VectorSQLDatabaseChain
     """SQL Database Chain"""

libs/experimental/langchain_experimental/rl_chain/base.py

@@ -51,6 +51,8 @@ class _BasedOn:
 
 
 def BasedOn(anything: Any) -> _BasedOn:
+    """Wrap a value to indicate that it should be based on."""
+
     return _BasedOn(anything)
 
 
@@ -65,6 +67,8 @@ class _ToSelectFrom:
 
 
 def ToSelectFrom(anything: Any) -> _ToSelectFrom:
+    """Wrap a value to indicate that it should be selected from."""
+
     if not isinstance(anything, list):
         raise ValueError("ToSelectFrom must be a list to select from")
     return _ToSelectFrom(anything)
@@ -82,6 +86,8 @@ class _Embed:
 
 
 def Embed(anything: Any, keep: bool = False) -> Any:
+    """Wrap a value to indicate that it should be embedded."""
+
     if isinstance(anything, _ToSelectFrom):
         return ToSelectFrom(Embed(anything.value, keep=keep))
     elif isinstance(anything, _BasedOn):
@@ -96,6 +102,8 @@ def Embed(anything: Any, keep: bool = False) -> Any:
 
 
 def EmbedAndKeep(anything: Any) -> Any:
+    """Wrap a value to indicate that it should be embedded and kept."""
+
     return Embed(anything, keep=True)
 
 
@@ -103,14 +111,19 @@ def EmbedAndKeep(anything: Any) -> Any:
 
 
 def stringify_embedding(embedding: List) -> str:
+    """Convert an embedding to a string."""
+
     return " ".join([f"{i}:{e}" for i, e in enumerate(embedding)])
 
 
 def parse_lines(parser: "vw.TextFormatParser", input_str: str) -> List["vw.Example"]:
+    """Parse the input string into a list of examples."""
+
     return [parser.parse_line(line) for line in input_str.split("\n")]
 
 
 def get_based_on_and_to_select_from(inputs: Dict[str, Any]) -> Tuple[Dict, Dict]:
+    """Get the BasedOn and ToSelectFrom from the inputs."""
     to_select_from = {
         k: inputs[k].value
         for k in inputs.keys()
@@ -132,8 +145,9 @@ def get_based_on_and_to_select_from(inputs: Dict[str, Any]) -> Tuple[Dict, Dict]
 
 
 def prepare_inputs_for_autoembed(inputs: Dict[str, Any]) -> Dict[str, Any]:
-    """
-    go over all the inputs and if something is either wrapped in _ToSelectFrom or _BasedOn, and if their inner values are not already _Embed,
+    """Prepare the inputs for auto embedding.
+
+    Go over all the inputs and if something is either wrapped in _ToSelectFrom or _BasedOn, and if their inner values are not already _Embed,
     then wrap them in EmbedAndKeep while retaining their _ToSelectFrom or _BasedOn status
     """  # noqa: E501
 
@@ -149,6 +163,8 @@ def prepare_inputs_for_autoembed(inputs: Dict[str, Any]) -> Dict[str, Any]:
 
 
 class Selected(ABC):
+    """Abstract class to represent the selected item."""
+
     pass
 
 
@@ -156,6 +172,8 @@ TSelected = TypeVar("TSelected", bound=Selected)
 
 
 class Event(Generic[TSelected], ABC):
+    """Abstract class to represent an event."""
+
     inputs: Dict[str, Any]
     selected: Optional[TSelected]
 
@@ -168,6 +186,8 @@ TEvent = TypeVar("TEvent", bound=Event)
 
 
 class Policy(Generic[TEvent], ABC):
+    """Abstract class to represent a policy."""
+
     def __init__(self, **kwargs: Any):
         pass
 
@@ -188,6 +208,8 @@ class Policy(Generic[TEvent], ABC):
 
 
 class VwPolicy(Policy):
+    """Vowpal Wabbit policy."""
+
     def __init__(
         self,
         model_repo: ModelRepository,
@@ -229,6 +251,8 @@ class VwPolicy(Policy):
 
 
 class Embedder(Generic[TEvent], ABC):
+    """Abstract class to represent an embedder."""
+
     def __init__(self, *args: Any, **kwargs: Any):
         pass
 
@@ -238,7 +262,7 @@ class Embedder(Generic[TEvent], ABC):
 
 
 class SelectionScorer(Generic[TEvent], ABC, BaseModel):
-    """Abstract method to grade the chosen selection or the response of the llm"""
+    """Abstract class to grade the chosen selection or the response of the llm."""
 
     @abstractmethod
     def score_response(
@@ -248,6 +272,8 @@ class SelectionScorer(Generic[TEvent], ABC, BaseModel):
 
 
 class AutoSelectionScorer(SelectionScorer[Event], BaseModel):
+    """Auto selection scorer."""
+
     llm_chain: LLMChain
     prompt: Union[BasePromptTemplate, None] = None
     scoring_criteria_template_str: Optional[str] = None
@@ -308,8 +334,8 @@ class AutoSelectionScorer(SelectionScorer[Event], BaseModel):
 
 
 class RLChain(Chain, Generic[TEvent]):
-    """
-    The `RLChain` class leverages the Vowpal Wabbit (VW) model as a learned policy for reinforcement learning.
+    """Chain that leverages the Vowpal Wabbit (VW) model as a learned policy
+    for reinforcement learning.
 
     Attributes:
         - llm_chain (Chain): Represents the underlying Language Model chain.
@@ -547,7 +573,8 @@ class RLChain(Chain, Generic[TEvent]):
 
 
 def is_stringtype_instance(item: Any) -> bool:
-    """Helper function to check if an item is a string."""
+    """Check if an item is a string."""
+
     return isinstance(item, str) or (
         isinstance(item, _Embed) and isinstance(item.value, str)
     )
@@ -556,7 +583,8 @@ def is_stringtype_instance(item: Any) -> bool:
 def embed_string_type(
     item: Union[str, _Embed], model: Any, namespace: Optional[str] = None
 ) -> Dict[str, Union[str, List[str]]]:
-    """Helper function to embed a string or an _Embed object."""
+    """Embed a string or an _Embed object."""
+
     keep_str = ""
     if isinstance(item, _Embed):
         encoded = stringify_embedding(model.encode(item.value))
@@ -576,7 +604,7 @@ def embed_string_type(
 
 
 def embed_dict_type(item: Dict, model: Any) -> Dict[str, Any]:
-    """Helper function to embed a dictionary item."""
+    """Embed a dictionary item."""
     inner_dict: Dict = {}
     for ns, embed_item in item.items():
         if isinstance(embed_item, list):
@@ -592,6 +620,8 @@ def embed_dict_type(item: Dict, model: Any) -> Dict[str, Any]:
 def embed_list_type(
     item: list, model: Any, namespace: Optional[str] = None
 ) -> List[Dict[str, Union[str, List[str]]]]:
+    """Embed a list item."""
+
     ret_list: List = []
     for embed_item in item:
         if isinstance(embed_item, dict):
@@ -614,7 +644,8 @@ def embed(
     namespace: Optional[str] = None,
 ) -> List[Dict[str, Union[str, List[str]]]]:
     """
-    Embeds the actions or context using the SentenceTransformer model (or a model that has an `encode` function)
+    Embed the actions or context using the SentenceTransformer model
+    (or a model that has an `encode` function).
 
     Attributes:
         to_embed: (Union[Union(str, _Embed(str)), Dict, List[Union(str, _Embed(str))], List[Dict]], required) The text to be embedded, either a string, a list of strings or a dictionary or a list of dictionaries.

libs/experimental/langchain_experimental/rl_chain/metrics.py

@@ -6,6 +6,8 @@ if TYPE_CHECKING:
 
 
 class MetricsTrackerAverage:
+    """Metrics Tracker Average."""
+
     def __init__(self, step: int):
         self.history: List[Dict[str, Union[int, float]]] = [{"step": 0, "score": 0}]
         self.step: int = step
@@ -33,6 +35,8 @@ class MetricsTrackerAverage:
 
 
 class MetricsTrackerRollingWindow:
+    """Metrics Tracker Rolling Window."""
+
     def __init__(self, window_size: int, step: int):
         self.history: List[Dict[str, Union[int, float]]] = [{"step": 0, "score": 0}]
         self.step: int = step

libs/experimental/langchain_experimental/rl_chain/model_repository.py

@@ -13,6 +13,8 @@ logger = logging.getLogger(__name__)
 
 
 class ModelRepository:
+    """Model Repository."""
+
     def __init__(
         self,
         folder: Union[str, os.PathLike],

libs/experimental/langchain_experimental/rl_chain/pick_best_chain.py

@@ -18,6 +18,8 @@ SENTINEL = object()
 
 
 class PickBestSelected(base.Selected):
+    """Selected class for PickBest chain."""
+
     index: Optional[int]
     probability: Optional[float]
     score: Optional[float]
@@ -34,6 +36,8 @@ class PickBestSelected(base.Selected):
 
 
 class PickBestEvent(base.Event[PickBestSelected]):
+    """Event class for PickBest chain."""
+
     def __init__(
         self,
         inputs: Dict[str, Any],
@@ -47,8 +51,8 @@ class PickBestEvent(base.Event[PickBestSelected]):
 
 
 class PickBestFeatureEmbedder(base.Embedder[PickBestEvent]):
-    """
-    Text Embedder class that embeds the `BasedOn` and `ToSelectFrom` inputs into a format that can be used by the learning policy
+    """Embed the `BasedOn` and `ToSelectFrom` inputs into a format that can be used
+    by the learning policy.
 
     Attributes:
         model name (Any, optional): The type of embeddings to be used for feature representation. Defaults to BERT SentenceTransformer.
@@ -225,6 +229,8 @@ class PickBestFeatureEmbedder(base.Embedder[PickBestEvent]):
 
 
 class PickBestRandomPolicy(base.Policy[PickBestEvent]):
+    """Random policy for PickBest chain."""
+
     def __init__(self, feature_embedder: base.Embedder, **kwargs: Any):
         self.feature_embedder = feature_embedder
 
@@ -240,8 +246,8 @@ class PickBestRandomPolicy(base.Policy[PickBestEvent]):
 
 
 class PickBest(base.RLChain[PickBestEvent]):
-    """
-    `PickBest` is a class designed to leverage the Vowpal Wabbit (VW) model for reinforcement learning with a context, with the goal of modifying the prompt before the LLM call.
+    """Chain that leverages the Vowpal Wabbit (VW) model for reinforcement learning
+    with a context, with the goal of modifying the prompt before the LLM call.
 
     Each invocation of the chain's `run()` method should be equipped with a set of potential actions (`ToSelectFrom`) and will result in the selection of a specific action based on the `BasedOn` input. This chosen action then informs the LLM (Language Model) prompt for the subsequent response generation.
 

libs/experimental/langchain_experimental/rl_chain/vw_logger.py

@@ -4,6 +4,8 @@ from typing import Optional, Union
 
 
 class VwLogger:
+    """Vowpal Wabbit custom logger."""
+
     def __init__(self, path: Optional[Union[str, PathLike]]):
         self.path = Path(path) if path else None
         if self.path:

libs/experimental/langchain_experimental/smart_llm/__init__.py

@@ -1,4 +1,23 @@
-"""Generalized implementation of SmartGPT (origin: https://youtu.be/wVzuvf9D9BU)"""
+"""Chain for applying self-critique using the SmartGPT workflow.
+
+See details at https://youtu.be/wVzuvf9D9BU
+
+The workflow performs these 3 steps:
+1. **Ideate**: Pass the user prompt to an ideation LLM n_ideas times,
+   each result is an "idea"
+2. **Critique**: Pass the ideas to a critique LLM which looks for flaws in the ideas
+   & picks the best one
+3. **Resolve**: Pass the critique to a resolver LLM which improves upon the best idea
+   & outputs only the (improved version of) the best output
+
+In total, the SmartGPT workflow will use n_ideas+2 LLM calls
+
+Note that SmartLLMChain will only improve results (compared to a basic LLMChain),
+when the underlying models have the capability for reflection, which smaller models
+often don't.
+
+Finally, a SmartLLMChain assumes that each underlying LLM outputs exactly 1 result.
+"""
 
 from langchain_experimental.smart_llm.base import SmartLLMChain
 

libs/experimental/langchain_experimental/smart_llm/base.py

@@ -18,8 +18,9 @@ from langchain_experimental.pydantic_v1 import Extra, root_validator
 
 
 class SmartLLMChain(Chain):
-    """
-    Generalized implementation of SmartGPT (origin: https://youtu.be/wVzuvf9D9BU)
+    """Chain for applying self-critique using the SmartGPT workflow.
+
+    See details at https://youtu.be/wVzuvf9D9BU
 
     A SmartLLMChain is an LLMChain that instead of simply passing the prompt to the LLM
     performs these 3 steps:

libs/experimental/langchain_experimental/sql/vector_sql.py

@@ -19,7 +19,8 @@ from langchain_experimental.sql.base import INTERMEDIATE_STEPS_KEY, SQLDatabaseC
 
 
 class VectorSQLOutputParser(BaseOutputParser[str]):
-    """Output Parser for Vector SQL
+    """Output Parser for Vector SQL.
+
     1. finds for `NeuralArray()` and replace it with the embedding
     2. finds for `DISTANCE()` and replace it with the distance name in backend SQL
     """
@@ -61,8 +62,8 @@ class VectorSQLOutputParser(BaseOutputParser[str]):
 
 
 class VectorSQLRetrieveAllOutputParser(VectorSQLOutputParser):
-    """Based on VectorSQLOutputParser
-    It also modify the SQL to get all columns
+    """Parser based on VectorSQLOutputParser.
+    It also modifies the SQL to get all columns.
     """
 
     @property
@@ -79,6 +80,8 @@ class VectorSQLRetrieveAllOutputParser(VectorSQLOutputParser):
 
 
 def get_result_from_sqldb(db: SQLDatabase, cmd: str) -> Sequence[Dict[str, Any]]:
+    """Get result from SQL Database."""
+
     result = db._execute(cmd, fetch="all")
     assert isinstance(result, Sequence)
     return result

libs/experimental/langchain_experimental/synthetic_data/__init__.py

@@ -12,7 +12,7 @@ def create_data_generation_chain(
     llm: BaseLanguageModel,
     prompt: Optional[PromptTemplate] = None,
 ) -> Chain:
-    """Creates a chain that generates synthetic sentences with
+    """Create a chain that generates synthetic sentences with
      provided fields.
 
     Args:
@@ -28,7 +28,7 @@ def create_data_generation_chain(
 
 
 class DatasetGenerator:
-    """Generates synthetic dataset with a given language model."""
+    """Generate synthetic dataset with a given language model."""
 
     def __init__(
         self,

libs/experimental/langchain_experimental/tabular_synthetic_data/base.py

@@ -9,7 +9,7 @@ from langchain_core.language_models import BaseLanguageModel
 
 
 class SyntheticDataGenerator(BaseModel):
-    """Generates synthetic data using the given LLM and few-shot template.
+    """Generate synthetic data using the given LLM and few-shot template.
 
     Utilizes the provided LLM to produce synthetic data based on the
     few-shot prompt template.

libs/experimental/langchain_experimental/text_splitter.py

@@ -11,6 +11,16 @@ from langchain_core.embeddings import Embeddings
 
 
 def combine_sentences(sentences: List[dict], buffer_size: int = 1) -> List[dict]:
+    """Combine sentences based on buffer size.
+
+    Args:
+        sentences: List of sentences to combine.
+        buffer_size: Number of sentences to combine. Defaults to 1.
+
+    Returns:
+        List of sentences with combined sentences.
+    """
+
     # Go through each sentence dict
     for i in range(len(sentences)):
         # Create a string that will hold the sentences which are joined
@@ -42,6 +52,14 @@ def combine_sentences(sentences: List[dict], buffer_size: int = 1) -> List[dict]
 
 
 def calculate_cosine_distances(sentences: List[dict]) -> Tuple[List[float], List[dict]]:
+    """Calculate cosine distances between sentences.
+
+    Args:
+        sentences: List of sentences to calculate distances for.
+
+    Returns:
+        Tuple of distances and sentences.
+    """
     distances = []
     for i in range(len(sentences) - 1):
         embedding_current = sentences[i]["combined_sentence_embedding"]
@@ -66,12 +84,12 @@ def calculate_cosine_distances(sentences: List[dict]) -> Tuple[List[float], List
 
 
 class SemanticChunker(BaseDocumentTransformer):
-    """Splits the text based on semantic similarity.
+    """Split the text based on semantic similarity.
 
     Taken from Greg Kamradt's wonderful notebook:
     https://github.com/FullStackRetrieval-com/RetrievalTutorials/blob/main/5_Levels_Of_Text_Splitting.ipynb
 
-    All credit to him.
+    All credits to him.
 
     At a high level, this splits into sentences, then groups into groups of 3
     sentences, and then merges one that are similar in the embedding space.

libs/experimental/langchain_experimental/tools/python/tool.py

@@ -24,6 +24,7 @@ def _get_default_python_repl() -> PythonREPL:
 
 def sanitize_input(query: str) -> str:
     """Sanitize input to the python REPL.
+
     Remove whitespace, backtick & python (if llm mistakes python console as terminal)
 
     Args:
@@ -41,7 +42,7 @@ def sanitize_input(query: str) -> str:
 
 
 class PythonREPLTool(BaseTool):
-    """A tool for running python code in a REPL."""
+    """Tool for running python code in a REPL."""
 
     name: str = "Python_REPL"
     description: str = (
@@ -76,11 +77,13 @@ class PythonREPLTool(BaseTool):
 
 
 class PythonInputs(BaseModel):
+    """Python inputs."""
+
     query: str = Field(description="code snippet to run")
 
 
 class PythonAstREPLTool(BaseTool):
-    """A tool for running python code in a REPL."""
+    """Tool for running python code in a REPL."""
 
     name: str = "python_repl_ast"
     description: str = (

libs/experimental/langchain_experimental/tot/__init__.py

@@ -1,3 +1,12 @@
+"""Implementation of a Tree of Thought (ToT) chain based on the paper
+"Large Language Model Guided Tree-of-Thought"
+
+https://arxiv.org/pdf/2305.08291.pdf
+
+The Tree of Thought (ToT) chain uses a tree structure to explore the space of
+possible solutions to a problem.
+
+"""
 from langchain_experimental.tot.base import ToTChain
 from langchain_experimental.tot.checker import ToTChecker
 

libs/experimental/langchain_experimental/tot/base.py

@@ -1,14 +1,3 @@
-"""
-This a Tree of Thought (ToT) chain based on the paper "Large Language Model
-Guided Tree-of-Thought"
-
-https://arxiv.org/pdf/2305.08291.pdf
-
-The Tree of Thought (ToT) chain uses a tree structure to explore the space of
-possible solutions to a problem.
-
-"""
-
 from __future__ import annotations
 
 from textwrap import indent
@@ -34,7 +23,7 @@ from langchain_experimental.tot.thought_generation import (
 
 class ToTChain(Chain):
     """
-    A Chain implementing the Tree of Thought (ToT).
+    Chain implementing the Tree of Thought (ToT).
     """
 
     llm: BaseLanguageModel

libs/experimental/langchain_experimental/tot/memory.py

@@ -7,7 +7,9 @@ from langchain_experimental.tot.thought import Thought
 
 class ToTDFSMemory:
     """
-    Memory for the Tree of Thought (ToT) chain. Implemented as a stack of
+    Memory for the Tree of Thought (ToT) chain.
+
+    It is implemented as a stack of
     thoughts. This allows for a depth first search (DFS) of the ToT.
     """
 

libs/experimental/langchain_experimental/tot/prompts.py

@@ -9,6 +9,8 @@ from langchain_experimental.tot.thought import ThoughtValidity
 
 
 def get_cot_prompt() -> PromptTemplate:
+    """Get the prompt for the Chain of Thought (CoT) chain."""
+
     return PromptTemplate(
         template_format="jinja2",
         input_variables=["problem_description", "thoughts"],
@@ -36,7 +38,7 @@ def get_cot_prompt() -> PromptTemplate:
 
 
 class JSONListOutputParser(BaseOutputParser):
-    """Class to parse the output of a PROPOSE_PROMPT response."""
+    """Parse the output of a PROPOSE_PROMPT response."""
 
     @property
     def _type(self) -> str:
@@ -53,6 +55,8 @@ class JSONListOutputParser(BaseOutputParser):
 
 
 def get_propose_prompt() -> PromptTemplate:
+    """Get the prompt for the PROPOSE_PROMPT chain."""
+
     return PromptTemplate(
         template_format="jinja2",
         input_variables=["problem_description", "thoughts", "n"],
@@ -95,6 +99,8 @@ def get_propose_prompt() -> PromptTemplate:
 
 
 class CheckerOutputParser(BaseOutputParser):
+    """Parse and check the output of the language model."""
+
     def parse(self, text: str) -> ThoughtValidity:
         """Parse the output of the language model."""
         text = text.upper()

libs/experimental/langchain_experimental/tot/thought.py

@@ -7,12 +7,16 @@ from langchain_experimental.pydantic_v1 import BaseModel, Field
 
 
 class ThoughtValidity(Enum):
+    """Enum for the validity of a thought."""
+
     VALID_INTERMEDIATE = 0
     VALID_FINAL = 1
     INVALID = 2
 
 
 class Thought(BaseModel):
+    """A thought in the ToT."""
+
     text: str
     validity: ThoughtValidity
     children: Set[Thought] = Field(default_factory=set)

libs/experimental/langchain_experimental/tot/thought_generation.py

@@ -39,7 +39,7 @@ class BaseThoughtGenerationStrategy(LLMChain):
 
 class SampleCoTStrategy(BaseThoughtGenerationStrategy):
     """
-    Sample thoughts from a Chain-of-Thought (CoT) prompt.
+    Sample strategy from a Chain-of-Thought (CoT) prompt.
 
     This strategy works better when the thought space is rich, such as when each
     thought is a paragraph. Independent and identically distributed samples
@@ -62,7 +62,7 @@ class SampleCoTStrategy(BaseThoughtGenerationStrategy):
 
 class ProposePromptStrategy(BaseThoughtGenerationStrategy):
     """
-    Propose thoughts sequentially using a "propose prompt".
+    Strategy that is sequentially using a "propose prompt".
 
     This strategy works better when the thought space is more constrained, such
     as when each thought is just a word or a line. Proposing different thoughts