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
17
18
19
20
21

pub mod amp;
pub mod async_checkpoint;
pub mod checkpoint;
pub mod config;
#[path = "distributed_checkpoint/mod.rs"]
pub mod distributed_checkpoint;
pub mod simple;
#[cfg(test)]
mod test_helpers;

pub use amp::AmpTrainer;
pub use async_checkpoint::AsyncCheckpointer;
pub use checkpoint::{
    CHECKPOINT_VERSION, CheckpointData, TrainingState, load_checkpoint, save_checkpoint,
};
pub use config::{LossScaleStrategy, MixedPrecisionConfig, TrainingConfig, TrainingMetrics};
pub use distributed_checkpoint::{
    ShardingConfig, ShardingMeta, ShardingStrategy, consolidate_checkpoint,
    load_distributed_checkpoint, save_distributed_checkpoint,
};
pub use simple::SimpleTrainer;