rumus 0.3.1

A native-Rust deep learning framework with explicit memory safety and hardware acceleration
Documentation 
1
2
3
4
5
6
7
8
9
10

// SPDX-License-Identifier: Apache-2.0 OR MIT
//! WGPU-based GPU backend.
//!
//! Provides the [`GpuContext`] device singleton, [`BufferPool`] for
//! efficient GPU memory management, and [`compute`] dispatch functions
//! for all tensor operations.

pub mod compute;
pub mod context;
pub mod pool;