aprender-gpu 0.31.1

Pure Rust PTX generation for NVIDIA CUDA - no LLVM, no nvcc
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12

//! GPU-Resident Tensor Operations Tests (PMAT-018: Coverage Killer Remediation)
//!
//! Comprehensive tests for all f32-specialized operations on `GpuResidentTensor`.
//! These tests exercise the actual CUDA kernel paths to achieve coverage.

#![cfg(all(test, feature = "cuda"))]

mod attention;
mod elementwise;
mod linear_conv;
mod matmul;
mod weights_and_tensor;