Crate burn_fusion

Expand description

§Burn Fusion

This library is a part of the Burn project. It is a standalone crate that can be used to perform automatic operation fusion on backends that support it.

Modules§

client: Client module exposing types to communicate with the fusion server.
stream: Stream module exposing all tensor operations that can be optimized.

Macros§

binary_float_cmp_ops
binary_float_ops
binary_int_cmp_ops
binary_int_ops
memory_checksNon-memory-checks: Export memory checks tests.
reduce_float2int_ops
reduce_float_ops
reduce_int_ops
scalar_float2int_ops
scalar_float_cmp_ops
scalar_float_ops
scalar_int_cmp_ops
scalar_int_ops
unary_float_ops
unary_int_ops

Structs§

Fusion: Enable dynamic operation fusion on a backend that implements fusion backend.
FusionTensor: Tensor primitive for the fusion backend for all kind.
OptimizationProperties: The properties of a builder.

Enums§

OptimizationStatus: The status of a builder.

Traits§

FusionBackend: Trait that allows an existing backend to specify graph optimizations using operation builder.
FusionRuntime: Trait that defines a runtime that will benefits from fused operations.
NumOperations: The number of operations contained in the data strusture.
Optimization: The operation created from the builder.
OptimizationBuilder: The fusion operation abstraction allows implementations to fuse many tensor operations into one, improving the performance of the backend.

Type Aliases§

Client: Type alias for <R as FusionRuntime>::FusionClient.
FusionDevice: Type alias for <R as FusionRuntime>::FusionDevice.
FusionHandle: Type alias for <R as FusionRuntime>::FusionHandle.
Handle: Type alias for representation backend handle.