mirror of https://github.com/hwchase17/langchain
You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
390 lines
14 KiB
Python
390 lines
14 KiB
Python
"""
|
|
UpTrain Callback Handler
|
|
|
|
UpTrain is an open-source platform to evaluate and improve LLM applications. It provides
|
|
grades for 20+ preconfigured checks (covering language, code, embedding use cases),
|
|
performs root cause analyses on instances of failure cases and provides guidance for
|
|
resolving them.
|
|
|
|
This module contains a callback handler for integrating UpTrain seamlessly into your
|
|
pipeline and facilitating diverse evaluations. The callback handler automates various
|
|
evaluations to assess the performance and effectiveness of the components within the
|
|
pipeline.
|
|
|
|
The evaluations conducted include:
|
|
|
|
1. RAG:
|
|
- Context Relevance: Determines the relevance of the context extracted from the query
|
|
to the response.
|
|
- Factual Accuracy: Assesses if the Language Model (LLM) is providing accurate
|
|
information or hallucinating.
|
|
- Response Completeness: Checks if the response contains all the information
|
|
requested by the query.
|
|
|
|
2. Multi Query Generation:
|
|
MultiQueryRetriever generates multiple variants of a question with similar meanings
|
|
to the original question. This evaluation includes previous assessments and adds:
|
|
- Multi Query Accuracy: Ensures that the multi-queries generated convey the same
|
|
meaning as the original query.
|
|
|
|
3. Context Compression and Reranking:
|
|
Re-ranking involves reordering nodes based on relevance to the query and selecting
|
|
top n nodes.
|
|
Due to the potential reduction in the number of nodes after re-ranking, the following
|
|
evaluations
|
|
are performed in addition to the RAG evaluations:
|
|
- Context Reranking: Determines if the order of re-ranked nodes is more relevant to
|
|
the query than the original order.
|
|
- Context Conciseness: Examines whether the reduced number of nodes still provides
|
|
all the required information.
|
|
|
|
These evaluations collectively ensure the robustness and effectiveness of the RAG query
|
|
engine, MultiQueryRetriever, and the re-ranking process within the pipeline.
|
|
|
|
Useful links:
|
|
Github: https://github.com/uptrain-ai/uptrain
|
|
Website: https://uptrain.ai/
|
|
Docs: https://docs.uptrain.ai/getting-started/introduction
|
|
|
|
"""
|
|
|
|
import logging
|
|
import sys
|
|
from collections import defaultdict
|
|
from typing import (
|
|
Any,
|
|
DefaultDict,
|
|
Dict,
|
|
List,
|
|
Optional,
|
|
Sequence,
|
|
Set,
|
|
)
|
|
from uuid import UUID
|
|
|
|
from langchain_core.callbacks.base import BaseCallbackHandler
|
|
from langchain_core.documents import Document
|
|
from langchain_core.outputs import LLMResult
|
|
|
|
logger = logging.getLogger(__name__)
|
|
handler = logging.StreamHandler(sys.stdout)
|
|
formatter = logging.Formatter("%(message)s")
|
|
handler.setFormatter(formatter)
|
|
logger.addHandler(handler)
|
|
|
|
|
|
def import_uptrain() -> Any:
|
|
try:
|
|
import uptrain
|
|
except ImportError as e:
|
|
raise ImportError(
|
|
"To use the UpTrainCallbackHandler, you need the"
|
|
"`uptrain` package. Please install it with"
|
|
"`pip install uptrain`.",
|
|
e,
|
|
)
|
|
|
|
return uptrain
|
|
|
|
|
|
class UpTrainDataSchema:
|
|
"""The UpTrain data schema for tracking evaluation results.
|
|
|
|
Args:
|
|
project_name_prefix (str): Prefix for the project name.
|
|
|
|
Attributes:
|
|
project_name_prefix (str): Prefix for the project name.
|
|
uptrain_results (DefaultDict[str, Any]): Dictionary to store evaluation results.
|
|
eval_types (Set[str]): Set to store the types of evaluations.
|
|
query (str): Query for the RAG evaluation.
|
|
context (str): Context for the RAG evaluation.
|
|
response (str): Response for the RAG evaluation.
|
|
old_context (List[str]): Old context nodes for Context Conciseness evaluation.
|
|
new_context (List[str]): New context nodes for Context Conciseness evaluation.
|
|
context_conciseness_run_id (str): Run ID for Context Conciseness evaluation.
|
|
multi_queries (List[str]): List of multi queries for Multi Query evaluation.
|
|
multi_query_run_id (str): Run ID for Multi Query evaluation.
|
|
multi_query_daugher_run_id (str): Run ID for Multi Query daughter evaluation.
|
|
|
|
"""
|
|
|
|
def __init__(self, project_name_prefix: str) -> None:
|
|
"""Initialize the UpTrain data schema."""
|
|
# For tracking project name and results
|
|
self.project_name_prefix: str = project_name_prefix
|
|
self.uptrain_results: DefaultDict[str, Any] = defaultdict(list)
|
|
|
|
# For tracking event types
|
|
self.eval_types: Set[str] = set()
|
|
|
|
## RAG
|
|
self.query: str = ""
|
|
self.context: str = ""
|
|
self.response: str = ""
|
|
|
|
## CONTEXT CONCISENESS
|
|
self.old_context: List[str] = []
|
|
self.new_context: List[str] = []
|
|
self.context_conciseness_run_id: UUID = UUID(int=0)
|
|
|
|
# MULTI QUERY
|
|
self.multi_queries: List[str] = []
|
|
self.multi_query_run_id: UUID = UUID(int=0)
|
|
self.multi_query_daugher_run_id: UUID = UUID(int=0)
|
|
|
|
|
|
class UpTrainCallbackHandler(BaseCallbackHandler):
|
|
"""Callback Handler that logs evaluation results to uptrain and the console.
|
|
|
|
Args:
|
|
project_name_prefix (str): Prefix for the project name.
|
|
key_type (str): Type of key to use. Must be 'uptrain' or 'openai'.
|
|
api_key (str): API key for the UpTrain or OpenAI API.
|
|
(This key is required to perform evaluations using GPT.)
|
|
|
|
Raises:
|
|
ValueError: If the key type is invalid.
|
|
ImportError: If the `uptrain` package is not installed.
|
|
|
|
"""
|
|
|
|
def __init__(
|
|
self,
|
|
*,
|
|
project_name_prefix: str = "langchain",
|
|
key_type: str = "openai",
|
|
api_key: str = "sk-****************", # The API key to use for evaluation
|
|
model: str = "gpt-3.5-turbo", # The model to use for evaluation
|
|
log_results: bool = True,
|
|
) -> None:
|
|
"""Initializes the `UpTrainCallbackHandler`."""
|
|
super().__init__()
|
|
|
|
uptrain = import_uptrain()
|
|
|
|
self.log_results = log_results
|
|
|
|
# Set uptrain variables
|
|
self.schema = UpTrainDataSchema(project_name_prefix=project_name_prefix)
|
|
self.first_score_printed_flag = False
|
|
|
|
if key_type == "uptrain":
|
|
settings = uptrain.Settings(uptrain_access_token=api_key, model=model)
|
|
self.uptrain_client = uptrain.APIClient(settings=settings)
|
|
elif key_type == "openai":
|
|
settings = uptrain.Settings(
|
|
openai_api_key=api_key, evaluate_locally=False, model=model
|
|
)
|
|
self.uptrain_client = uptrain.EvalLLM(settings=settings)
|
|
else:
|
|
raise ValueError("Invalid key type: Must be 'uptrain' or 'openai'")
|
|
|
|
def uptrain_evaluate(
|
|
self,
|
|
project_name: str,
|
|
data: List[Dict[str, Any]],
|
|
checks: List[str],
|
|
) -> None:
|
|
"""Run an evaluation on the UpTrain server using UpTrain client."""
|
|
if self.uptrain_client.__class__.__name__ == "APIClient":
|
|
uptrain_result = self.uptrain_client.log_and_evaluate(
|
|
project_name=project_name,
|
|
data=data,
|
|
checks=checks,
|
|
)
|
|
else:
|
|
uptrain_result = self.uptrain_client.evaluate(
|
|
data=data,
|
|
checks=checks,
|
|
)
|
|
self.schema.uptrain_results[project_name].append(uptrain_result)
|
|
|
|
score_name_map = {
|
|
"score_context_relevance": "Context Relevance Score",
|
|
"score_factual_accuracy": "Factual Accuracy Score",
|
|
"score_response_completeness": "Response Completeness Score",
|
|
"score_sub_query_completeness": "Sub Query Completeness Score",
|
|
"score_context_reranking": "Context Reranking Score",
|
|
"score_context_conciseness": "Context Conciseness Score",
|
|
"score_multi_query_accuracy": "Multi Query Accuracy Score",
|
|
}
|
|
|
|
if self.log_results:
|
|
# Set logger level to INFO to print the evaluation results
|
|
logger.setLevel(logging.INFO)
|
|
|
|
for row in uptrain_result:
|
|
columns = list(row.keys())
|
|
for column in columns:
|
|
if column == "question":
|
|
logger.info(f"\nQuestion: {row[column]}")
|
|
self.first_score_printed_flag = False
|
|
elif column == "response":
|
|
logger.info(f"Response: {row[column]}")
|
|
self.first_score_printed_flag = False
|
|
elif column == "variants":
|
|
logger.info("Multi Queries:")
|
|
for variant in row[column]:
|
|
logger.info(f" - {variant}")
|
|
self.first_score_printed_flag = False
|
|
elif column.startswith("score"):
|
|
if not self.first_score_printed_flag:
|
|
logger.info("")
|
|
self.first_score_printed_flag = True
|
|
if column in score_name_map:
|
|
logger.info(f"{score_name_map[column]}: {row[column]}")
|
|
else:
|
|
logger.info(f"{column}: {row[column]}")
|
|
|
|
if self.log_results:
|
|
# Set logger level back to WARNING
|
|
# (We are doing this to avoid printing the logs from HTTP requests)
|
|
logger.setLevel(logging.WARNING)
|
|
|
|
def on_llm_end(
|
|
self,
|
|
response: LLMResult,
|
|
*,
|
|
run_id: UUID,
|
|
parent_run_id: Optional[UUID] = None,
|
|
**kwargs: Any,
|
|
) -> None:
|
|
"""Log records to uptrain when an LLM ends."""
|
|
uptrain = import_uptrain()
|
|
self.schema.response = response.generations[0][0].text
|
|
if (
|
|
"qa_rag" in self.schema.eval_types
|
|
and parent_run_id != self.schema.multi_query_daugher_run_id
|
|
):
|
|
data = [
|
|
{
|
|
"question": self.schema.query,
|
|
"context": self.schema.context,
|
|
"response": self.schema.response,
|
|
}
|
|
]
|
|
|
|
self.uptrain_evaluate(
|
|
project_name=f"{self.schema.project_name_prefix}_rag",
|
|
data=data,
|
|
checks=[
|
|
uptrain.Evals.CONTEXT_RELEVANCE,
|
|
uptrain.Evals.FACTUAL_ACCURACY,
|
|
uptrain.Evals.RESPONSE_COMPLETENESS,
|
|
],
|
|
)
|
|
|
|
def on_chain_start(
|
|
self,
|
|
serialized: Dict[str, Any],
|
|
inputs: Dict[str, Any],
|
|
*,
|
|
run_id: UUID,
|
|
tags: Optional[List[str]] = None,
|
|
parent_run_id: Optional[UUID] = None,
|
|
metadata: Optional[Dict[str, Any]] = None,
|
|
run_type: Optional[str] = None,
|
|
name: Optional[str] = None,
|
|
**kwargs: Any,
|
|
) -> None:
|
|
"""Do nothing when chain starts"""
|
|
if parent_run_id == self.schema.multi_query_run_id:
|
|
self.schema.multi_query_daugher_run_id = run_id
|
|
if isinstance(inputs, dict) and set(inputs.keys()) == {"context", "question"}:
|
|
self.schema.eval_types.add("qa_rag")
|
|
|
|
context = ""
|
|
if isinstance(inputs["context"], Document):
|
|
context = inputs["context"].page_content
|
|
elif isinstance(inputs["context"], list):
|
|
for doc in inputs["context"]:
|
|
context += doc.page_content + "\n"
|
|
elif isinstance(inputs["context"], str):
|
|
context = inputs["context"]
|
|
self.schema.context = context
|
|
self.schema.query = inputs["question"]
|
|
pass
|
|
|
|
def on_retriever_start(
|
|
self,
|
|
serialized: Dict[str, Any],
|
|
query: str,
|
|
*,
|
|
run_id: UUID,
|
|
parent_run_id: Optional[UUID] = None,
|
|
tags: Optional[List[str]] = None,
|
|
metadata: Optional[Dict[str, Any]] = None,
|
|
**kwargs: Any,
|
|
) -> None:
|
|
if "contextual_compression" in serialized["id"]:
|
|
self.schema.eval_types.add("contextual_compression")
|
|
self.schema.query = query
|
|
self.schema.context_conciseness_run_id = run_id
|
|
|
|
if "multi_query" in serialized["id"]:
|
|
self.schema.eval_types.add("multi_query")
|
|
self.schema.multi_query_run_id = run_id
|
|
self.schema.query = query
|
|
elif "multi_query" in self.schema.eval_types:
|
|
self.schema.multi_queries.append(query)
|
|
|
|
def on_retriever_end(
|
|
self,
|
|
documents: Sequence[Document],
|
|
*,
|
|
run_id: UUID,
|
|
parent_run_id: Optional[UUID] = None,
|
|
**kwargs: Any,
|
|
) -> Any:
|
|
"""Run when Retriever ends running."""
|
|
uptrain = import_uptrain()
|
|
if run_id == self.schema.multi_query_run_id:
|
|
data = [
|
|
{
|
|
"question": self.schema.query,
|
|
"variants": self.schema.multi_queries,
|
|
}
|
|
]
|
|
|
|
self.uptrain_evaluate(
|
|
project_name=f"{self.schema.project_name_prefix}_multi_query",
|
|
data=data,
|
|
checks=[uptrain.Evals.MULTI_QUERY_ACCURACY],
|
|
)
|
|
if "contextual_compression" in self.schema.eval_types:
|
|
if parent_run_id == self.schema.context_conciseness_run_id:
|
|
for doc in documents:
|
|
self.schema.old_context.append(doc.page_content)
|
|
elif run_id == self.schema.context_conciseness_run_id:
|
|
for doc in documents:
|
|
self.schema.new_context.append(doc.page_content)
|
|
context = "\n".join(
|
|
[
|
|
f"{index}. {string}"
|
|
for index, string in enumerate(self.schema.old_context, start=1)
|
|
]
|
|
)
|
|
reranked_context = "\n".join(
|
|
[
|
|
f"{index}. {string}"
|
|
for index, string in enumerate(self.schema.new_context, start=1)
|
|
]
|
|
)
|
|
data = [
|
|
{
|
|
"question": self.schema.query,
|
|
"context": context,
|
|
"concise_context": reranked_context,
|
|
"reranked_context": reranked_context,
|
|
}
|
|
]
|
|
self.uptrain_evaluate(
|
|
project_name=f"{self.schema.project_name_prefix}_context_reranking",
|
|
data=data,
|
|
checks=[
|
|
uptrain.Evals.CONTEXT_CONCISENESS,
|
|
uptrain.Evals.CONTEXT_RERANKING,
|
|
],
|
|
)
|