leafby autumnai
Open Machine Intelligence framework.. Abandoned project. The most updated fork is juice.
leaf0.2.1231/90d updated 2y ago
Machine learning
30 projects
Rust's ML ecosystem is smaller than Python's but strong in inference and tooling. Candle, from Hugging Face, is a minimal tensor framework focused on running models with CPU, CUDA and Metal backends and small binaries. Burn is a full deep learning framework that supports training as well as inference, with swappable backends including WGPU, CUDA and LibTorch. Hugging Face's tokenizers library is written in Rust and powers its Python package.
5.5k
tictonixby ave-sergeev
A library that provides the ability to convert tokens into embeddings, as well as to encode their positions.
tictonix1.0.111/90d updated 1y ago
41485
burnby tracel-ai
A Flexible and Comprehensive Deep Learning Framework.
burn0.22.1724.3k/90d updated today
dfdxby chelsea0x3b
CUDA accelerated machine learning framework that leverages many of Rust's unique features.
v0.13.0updated 2y ago
rust-automlby cmccomb
Automated classification, regression, and clustering workflows built on SmartCore
automl0.3.1228/90d updated 1mo ago
mistral.rsby EricLBuehler
Fast, flexible LLM inference engine supporting multimodal models, quantization (GGUF/GPTQ/ISQ), and OpenAI-compatible API
mistralrs0.8.1136.9k/90d updated today
furnaceby Gilfeather
Pure Rust inference server for ONNX models, built on Burn
furnace-server0.10.062/90d updated today
3.1k
candleby huggingface
A minimalist ML framework with a focus on easiness of use and on performance (including GPU support)
candle-core0.11.03.9m/90d updated 6d ago
tokenizersby huggingface
Hugging Face's tokenizers for modern NLP pipelines (original implementation) with bindings for Python.
tokenizers0.23.217.1m/90d updated 1d ago
7.1k
5.5k
luminalby luminal-ai
High-performance general-purpose inference compiler with RISC-style architecture, search-based optimization, and native CUDA/Metal backends. Supports transformers, convnets, and autograd.
luminal0.2.0305/90d updated today
647
shimmyby Michael-A-Kuykendall
Pure-Rust WebGPU inference engine with an OpenAI-compatible API and native GGUF support.
shimmy2.6.51k/90d updated 2d ago
shimmytokby Michael-A-Kuykendall
Pure-Rust tokenizer for GGUF models, compatible with llama.cpp tokenization.
shimmytok0.8.43.3k/90d updated 2d ago
62
nobodywhoby nobodywho-ooo
On-device LLM inference engine that embeds directly in games and apps with no server or API key. Supports streaming generation, embeddings, GBNF grammar-constrained structured output, and Whisper speech-to-text, with bindings for Godot, Flutter, React Native, and Swift.
updated 1d ago
openinferby openinfer-project
LLM inference engine in pure Rust + CUDA with no PyTorch and no Python runtime — OpenAI-compatible API, paged KV cache, CUDA Graph, serving models from Qwen3 to the trillion-parameter Kimi-K2.
v0.1.1updated today
perpetualby perpetual-ml
A self-generalizing gradient boosting machine which doesn't need hyperparameter optimization.
perpetual2.2.04.5k/90d updated 16d ago
RustTensorby ramsyana
A learning-focused, high-performance tensor computation library built from scratch in Rust with automatic differentiation and CPU/CUDA backends.
updated 1y ago
edgequakeby raphaelmansuy
A high-performance Graph-RAG framework that transforms documents into intelligent knowledge graphs.
v0.36.0updated today
4.8k
anofox-regressionby sipemu
Statistical regression models (OLS, Elastic Net, GLM, Quantile & Isotonic) with R-like inference (p-values, confidence & prediction intervals) and Wasm support.
anofox-regression0.5.2910.9k/90d updated 5d ago
967
rust-lstmby SyntaxSpirits
LSTM, BiLSTM, GRU and peephole LSTM networks with backpropagation through time, SGD/Adam/RMSprop optimizers, learning rate schedulers, dropout and zoneout.
rust-lstm0.10.029/90d updated 9d ago
festeby tag1consulting
A GPT-2 style transformer language model implemented from scratch in Rust for educational purposes.
updated 6mo ago
5.5k