Module mixed_precision 

Mixed-precision operations for the array protocol.

This module provides support for mixed-precision operations, allowing arrays to use different numeric types (e.g., f32, f64) for storage and computation to optimize performance and memory usage.

Modules

ops

Implementation of common array operations with mixed precision.

Structs

MixedPrecisionArray

Mixed-precision array that can automatically convert between precisions.

MixedPrecisionConfig

Configuration for mixed-precision operations.

Enums

Precision

Precision levels for array operations.

Statics

MIXED_PRECISION_CONFIG

Global mixed-precision configuration.

Traits

MixedPrecisionSupport

Trait for arrays that support mixed-precision operations.

Functions

determine_optimal_precision

Determine the optimal precision for an array based on its size.

execute_with_precision

Execute an operation with a specific precision.

get_mixed_precision_config

Get the current mixed-precision configuration.

set_mixed_precision_config

Set the global mixed-precision configuration.