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
13
14
15
16

pub mod flash;
pub mod flash_bwd;
pub mod flash_bwd_fp8;
pub mod flash_decode;
pub mod flash_fwd;
pub mod flash_utils;
pub mod flash_v3;
pub mod fused_qkv;
pub mod kv_insert;
pub mod mla;
pub mod mqa_gqa;
pub mod paged_attention;
pub mod paged_attention_bwd;
pub mod paged_attention_fwd;
pub mod paged_decode;
pub mod varlen_attention;