ruda-model 0.21.6

Ruda model configuration, module parameters, records and data loading.
Documentation

ruda-model

Model configuration, modules, parameters, records, and data loading for Ruda. Neural-network layer implementations are in ruda-nn; storage-format adapters are in ruda-store.

Interfaces

  • config, module, and record define model configuration, parameter ownership, and recording.
  • Tensor is re-exported from the model tensor interface.
  • data provides standard-library data-loading support; optional dataset integration comes from ruda-dataset.

Usage

Cargo package: ruda-model. Rust import: ruda_model.

[dependencies]
ruda-model = "0.21"

Features

Default features: std, ruda-core/std, ruda-core/id-tests-std, ruda-tensor-config/std, ruda-dataset?/default, ruda-tensor/api-std.

Feature Purpose
dataset Enable dataset integration.
network Enable model I/O dependency integration.
record-item-custom-serde Enable custom record-item serialization support.
distributed Enable distributed model and tensor integration.

Links