meganeura 0.2.0

E-graph optimized neural network training on Blade
Documentation 
1
2
3
4
5
6
7
8

//! Model loading from standard interchange formats.
//!
//! Each submodule imports a model from a specific format into Meganeura's
//! `Graph` IR, which then flows through the normal pipeline:
//! `Graph -> optimize (e-graph) -> compile -> ExecutionPlan -> Session`.

pub mod nnef;
pub mod onnx;