Crate meganeura 

Meganeura: E-graph optimized neural network framework on blade-graphics.

Models are defined as declarative computation graphs, optimized via equality saturation (egglog), and compiled to static GPU dispatch sequences — no manual CUDA-graphing needed.

Re-exports

pub use data::DataLoader;

pub use data::MnistDataset;

pub use graph::DType;

pub use graph::Graph;

pub use graph::NodeId;

pub use graph::TensorType;

pub use load::nnef::NnefError;

pub use load::nnef::NnefModel;

pub use load::nnef::load_nnef;

pub use load::onnx::OnnxError;

pub use load::onnx::OnnxModel;

pub use load::onnx::load_onnx;

pub use load::onnx::load_onnx_bytes;

pub use optimize::OptimizeReport;

pub use runtime::MemorySummary;

pub use runtime::Session;

pub use train::EpochStats;

pub use train::LossHistory;

pub use train::MetricCallback;

pub use train::Optimizer;

pub use train::StepMetrics;

pub use train::TrainConfig;

pub use train::TrainHistory;

pub use train::Trainer;

pub use train::build_inference_session;

pub use train::build_session;

pub use train::build_session_cached;

pub use train::build_session_unoptimized;

pub use train::build_session_with_report;

pub use train::compile_training_graph;

Modules

autodiff

cache

codegen

Shader codegen via WGSL templates.

compile

data

Data loading utilities for training.

graph

load

Model loading from standard interchange formats.

models

nn

High-level neural network building blocks.

optimize

profiler

Profiling infrastructure producing Perfetto binary traces (.pftrace).

runtime

train