class

`SushiAI::Train::IObjective`

What a model is being trained to minimise.

Declared in: include/SushiAI/train/objective.hpp

Public member functions

virtual ~IObjective()=default

virtual void declare(const Graph::GraphBuilder &builder, ObjectiveLeaves &leaves) const =0

Declares whatever leaves the graph needs beyond the model's input.

Parameters

builder: Where the leaves go.
leaves: Where to record them, by role.

virtual Graph::ValueId loss(const Graph::GraphBuilder &builder, const Graph::Graph &graph, Graph::ValueId output, const ObjectiveLeaves &leaves) const =0

Computes scalar loss from model output.

Parameters

builder: Where the nodes go.
graph: The graph, for checking shapes.
output: The model output value.
leaves: Declared auxiliary leaves.

Returns

The scalar loss value identifier.

virtual std::optional< Metric > score(const Data::Batch &batch, const SushiBLAS::Tensor &output) const =0

An optional per-step metric: accuracy, mAP, an L2 residual, or none.

Parameters

batch: The batch that was just stepped.
output: The model's output tensor.

Returns

The metric, or std::nullopt when this objective offers none.

virtual std::string_view name() const noexcept=0

A short name for a run report.