gmakstutis/cognee
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix: create enrichments

Browse source
This commit is contained in:
lxobr 2025-10-21 00:57:42 +02:00
parent 834cf8b113
commit 8e580bd3d3
3 changed files with 16 additions and 12 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

13
cognee/tasks/feedback/create_enrichments.py
View file

@ -68,17 +68,13 @@ async def _generate_enrichment_report(
async def _create_enrichment_datapoint(
improved_answer_item: Dict,
report_text: str,
nodeset: NodeSet,
) -> Optional[FeedbackEnrichment]:
"""Create a single FeedbackEnrichment DataPoint with proper ID and nodeset assignment."""
try:
question = improved_answer_item["question"]
improved_answer = improved_answer_item["improved_answer"]
# Create nodeset following UserQAFeedback pattern
nodeset = NodeSet(
id=uuid5(NAMESPACE_OID, name="FeedbackEnrichment"), name="FeedbackEnrichment"
)
enrichment = FeedbackEnrichment(
id=str(uuid5(NAMESPACE_OID, f"{question}_{improved_answer}")),
text=report_text,
@ -87,7 +83,7 @@ async def _create_enrichment_datapoint(
improved_answer=improved_answer,
feedback_id=improved_answer_item["feedback_id"],
interaction_id=improved_answer_item["interaction_id"],
belongs_to_set=nodeset,
belongs_to_set=[nodeset],
)
return enrichment
@ -119,6 +115,9 @@ async def create_enrichments(
logger.info("Creating enrichments", count=len(improved_answers))
# Create nodeset once for all enrichments
nodeset = NodeSet(id=uuid5(NAMESPACE_OID, name="FeedbackEnrichment"), name="FeedbackEnrichment")
enrichments: List[FeedbackEnrichment] = []
for improved_answer_item in improved_answers:
@ -130,7 +129,7 @@ async def create_enrichments(
question, improved_answer, new_context, report_prompt_location
)
enrichment = await _create_enrichment_datapoint(improved_answer_item, report_text)
enrichment = await _create_enrichment_datapoint(improved_answer_item, report_text, nodeset)
if enrichment:
enrichments.append(enrichment)

2
cognee/tasks/feedback/models.py
View file

@ -18,4 +18,4 @@ class FeedbackEnrichment(DataPoint):
improved_answer: str
feedback_id: UUID
interaction_id: UUID
belongs_to_set: Optional[NodeSet] = None
belongs_to_set: Optional[List[NodeSet]] = None

13
examples/python/feedback_enrichment_minimal_example.py
View file

@ -3,6 +3,9 @@ import asyncio
import cognee
from cognee.api.v1.search import SearchType
from cognee.modules.pipelines.tasks.task import Task
from cognee.tasks.graph import extract_graph_from_data
from cognee.tasks.storage import add_data_points
from cognee.shared.data_models import KnowledgeGraph
from cognee.tasks.feedback.extract_feedback_interactions import extract_feedback_interactions
from cognee.tasks.feedback.generate_improved_answers import generate_improved_answers
@ -49,12 +52,14 @@ async def run_question_and_submit_feedback(question_text: str) -> bool:
async def run_feedback_enrichment_memify(last_n: int = 5):
"""Execute memify with extraction, answer improvement, and enrichment creation tasks."""
"""Execute memify with extraction, answer improvement, enrichment creation, and graph processing tasks."""
# Instantiate tasks with their own kwargs
extraction_tasks = [Task(extract_feedback_interactions, last_n=last_n)]
enrichment_tasks = [
Task(generate_improved_answers, retriever_name="graph_completion_cot", top_k=20),
Task(create_enrichments),
Task(extract_graph_from_data, graph_model=KnowledgeGraph, task_config={"batch_size": 10}),
Task(add_data_points, task_config={"batch_size": 10}),
]
await cognee.memify(
extraction_tasks=extraction_tasks,
@ -65,9 +70,9 @@ async def run_feedback_enrichment_memify(last_n: int = 5):
async def main():
await initialize_conversation_and_graph(CONVERSATION)
is_correct = await run_question_and_submit_feedback("Who told Bob to bring the donuts?")
# is_correct = False
# await initialize_conversation_and_graph(CONVERSATION)
# is_correct = await run_question_and_submit_feedback("Who told Bob to bring the donuts?")
is_correct = False
if not is_correct:
await run_feedback_enrichment_memify(last_n=5)