Buckets:

download
raw
2.7 kB
"""
ACGS-2 Online Learning Module (Facade)
Constitutional Hash: 608508a9bd224290
Incremental learning using River with sklearn compatibility.
Delegates to specialized modules in .online_learning_infra.
"""
from enhanced_agent_bus.online_learning_infra.adapter import (
RIVER_AVAILABLE,
RiverSklearnAdapter,
)
from enhanced_agent_bus.online_learning_infra.config import (
ENABLE_COLD_START_FALLBACK,
KAFKA_AUTO_OFFSET_RESET,
KAFKA_BOOTSTRAP,
KAFKA_CONSUMER_GROUP,
KAFKA_MAX_POLL_RECORDS,
KAFKA_TOPIC_FEEDBACK,
MIN_SAMPLES_FOR_PREDICTION,
RIVER_MODEL_TYPE,
RIVER_N_MODELS,
RIVER_SEED,
LearningStatus,
ModelType,
)
from enhanced_agent_bus.online_learning_infra.consumer import FeedbackKafkaConsumer
from enhanced_agent_bus.online_learning_infra.evaluator import OnlineLearningEvaluator
from enhanced_agent_bus.online_learning_infra.models import (
ConsumerStats,
LearningResult,
LearningStats,
PipelineStats,
PredictionResult,
)
from enhanced_agent_bus.online_learning_infra.registry import (
get_consumer_stats,
get_feedback_kafka_consumer,
get_online_learning_adapter,
get_online_learning_pipeline,
start_feedback_consumer,
stop_feedback_consumer,
)
from enhanced_agent_bus.online_learning_infra.trainer import OnlineLearningPipeline
# Optional availability flags (backward compatibility)
NUMPY_AVAILABLE = True
KAFKA_AVAILABLE = True
# Helper for direct feedback learning (delegates to singleton pipeline)
def learn_from_feedback_event(
features: dict, outcome: float | None = None, decision_id: str | None = None
) -> LearningResult:
"""Convenience function to learn from a feedback event using the global pipeline."""
pipeline = get_online_learning_pipeline()
return pipeline.learn_from_feedback(features, outcome, decision_id)
__all__ = [
"ENABLE_COLD_START_FALLBACK",
"KAFKA_AUTO_OFFSET_RESET",
"KAFKA_AVAILABLE",
"KAFKA_BOOTSTRAP",
"KAFKA_CONSUMER_GROUP",
"KAFKA_MAX_POLL_RECORDS",
"KAFKA_TOPIC_FEEDBACK",
"MIN_SAMPLES_FOR_PREDICTION",
"NUMPY_AVAILABLE",
"RIVER_AVAILABLE",
"RIVER_MODEL_TYPE",
"RIVER_N_MODELS",
"RIVER_SEED",
"ConsumerStats",
"FeedbackKafkaConsumer",
"LearningResult",
"LearningStats",
"LearningStatus",
"ModelType",
"OnlineLearningEvaluator",
"OnlineLearningPipeline",
"PipelineStats",
"PredictionResult",
"RiverSklearnAdapter",
"get_consumer_stats",
"get_feedback_kafka_consumer",
"get_online_learning_adapter",
"get_online_learning_pipeline",
"learn_from_feedback_event",
"start_feedback_consumer",
"stop_feedback_consumer",
]

Xet Storage Details

Size:
2.7 kB
·
Xet hash:
619613088cca95f63eeb2f3ef681ec92a4e6af65ab7b0f08d00242c804e6acd4

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.