boostr 0.1.0

ML framework built on numr - attention, quantization, model architectures
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12

pub mod core;
pub mod gpu_cache;
pub mod gpu_radix;
pub mod types;

pub use core::PrefixCache;
#[cfg(feature = "cuda")]
pub use gpu_cache::GpuPrefixCache;
pub use gpu_radix::{GpuRadixStats, GpuRadixTree, PrefixLookup};
pub use types::{
    BlockHash, CacheResult, PrefixCacheConfig, PrefixCacheStats, SequenceId, SequencePrefixState,
};