boostr 0.2.0

ML framework built on numr - attention, quantization, model architectures
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14

pub mod device_map;
pub mod gguf;
pub mod gguf_tokenizer;
pub mod safetensors;
pub mod safetensors_loader;
pub mod safetensors_name_map;
pub mod torch_pt;

pub use device_map::{DevicePlacement, LayerDeviceMap};
pub use gguf::{GgmlType, Gguf, GgufMetadata, GgufTensorInfo, GgufValue, GgufValueType};
pub use gguf_tokenizer::GgufTokenizer;
pub use safetensors::{SafeTensors, TensorInfo};
pub use safetensors_loader::SafeTensorsLoader;
pub use torch_pt::{TorchStateDict, load_tensor_pt, load_voice_pt};