Production¤

Module: generative_models.inference.optimization.production

Source: generative_models/inference/optimization/production.py

Overview¤

Note: This module is experimental and under active development as part of ongoing work towards production-ready inference capabilities.

Inference optimization infrastructure for scaled models.

This module provides optimization infrastructure including:

All implementations follow JAX/Flax NNX best practices and prioritize performance through hardware-aware optimization.

class CompiledModel

class MonitoringMetrics

class OptimizationResult

class OptimizationTarget

class ProductionMonitor

class ProductionOptimizer

class ProductionPipeline

def __call__()

def __init__()

def __init__()

def __init__()

def __init__()

def compiled_forward()

def create_production_optimizer()

def create_production_pipeline()

def create_production_pipeline()

def get_metrics()

def get_monitoring_metrics()

def optimize_for_production()

def predict()

def predict_batch()

def record_request()

def reset()

def reset_monitoring()