libmir-cuda 0.1.0

CUDA inference backend for libmir
1
2
3
4
5
6
7
8
9
10
11
12
13
14
mod batch;
mod config;
mod graph;
mod layer;
mod prefill;
mod session;
mod template;
#[cfg(test)]
mod tests;

pub use batch::CudaDecodeBatch;
pub use config::CudaModelSessionConfig;
pub use session::CudaMoeModelSession;
pub use template::CudaMoeModelTemplate;