boostr 0.1.0

ML framework built on numr - attention, quantization, model architectures
Documentation 
1
2
3

pub mod fused_fp8;

pub use fused_fp8::dynamic_loss_scale_update_impl;