Expand description
NanoChat - A GPT implementation in Rust using Burn
This library provides a complete implementation of the NanoChat model with modern features like rotary embeddings, Multi-Query Attention, and efficient KV caching.
Re-exports§
pub use checkpoint::load_checkpoint;pub use checkpoint::save_checkpoint;pub use checkpoint::load_weights;pub use checkpoint::save_weights;pub use backend::get_device;pub use backend::print_backend_info;pub use backend::AutoBackend;pub use config::NanoChatConfig;
Modules§
- backend
- Backend selection with automatic GPU detection for Burn 0.18
- checkpoint
- Checkpoint save/load for NanoChat models
- config
- Configuration structures for NanoChat model and training
- engine
- gpt
- NanoChat GPT with numerically stable attention and logits
- prelude
- Prelude module for convenient imports
- sampling
- Sampling strategies for text generation
- tokenizer
- Tokenizer for NanoChat with special token support and chat templating Mirrors the Python tokenizer.py implementation