Transformers

A comprehensive educational project that implements a transformer model from scratch in PyTorch and demonstrates the complete LLM training...

A Go utility for downloading HuggingFace models and datasets with parallel connections, interactive GGUF quantization selection, and proxy support....

A comprehensive, open-source AI engineering curriculum spanning 503 lessons across 20 phases that teaches AI fundamentals from first principles...

MFLUX is a specialized library for running state-of-the-art generative image models (Flux, Qwen, Z-Image) natively on Apple Silicon using MLX, with...

Chronos is a family of pretrained transformer models for time series forecasting, including univariate, multivariate, and covariate-informed tasks....

MLX-Audio is a specialized speech processing library for Apple Silicon Macs, providing TTS, STT, and STS capabilities optimized for M-series chips...

Presidio is an open-source framework for detecting, redacting, masking, and anonymizing sensitive personally identifiable information (PII) across...

KVPress is a specialized Python library for compressing key-value (KV) caches in transformer-based LLMs during inference, addressing the memory...

SwanLab is an open-source AI training tracking and visualization platform designed for model training teams, offering cloud and self-hosted...

An open technical reference and practical guide for machine learning engineers working on large language model and multi-modal model training,...

AutoRound is a specialized quantization toolkit for Large Language Models and Vision-Language Models, designed to achieve high accuracy at ultra-low...

MLX-Tune enables fine-tuning of large language models, vision models, and audio systems directly on Apple Silicon Macs using the native MLX...

LLM Guard is a Python security toolkit for protecting Large Language Model interactions through input sanitization, harmful content detection, data...

Lagent is a lightweight Python framework for building LLM-based agents, inspired by PyTorch's design philosophy. It serves developers building...

A curated collection of 100+ original architecture diagrams and visual maps explaining LLM, reinforcement learning, and vision-language model...

GPTQModel is a specialized LLM quantization and compression toolkit providing hardware-accelerated inference across NVIDIA, AMD, Intel, Huawei, and...

An interactive 12-chapter textbook that teaches how to build a modern LLM (GPT/LLaMA-style) from scratch using 7,500+ lines of fully commented code....

Pruna is a model optimization framework providing compression algorithms (quantization, pruning, distillation, caching, compilation) for Python-based...

SANA is a specialized research framework for efficient high-resolution image and video generation using linear diffusion transformers, developed by...

SpeechBrain is an open-source PyTorch toolkit for building conversational AI systems, covering speech recognition, speaker recognition, speech...

Megatron-LM is a GPU-optimized library and research framework for training transformer models at scale, specifically designed for distributed...

Segmentation Models PyTorch (SMP) is a specialized deep learning library providing 500+ pretrained encoder-decoder architectures for semantic image...

Shimmy is a pure-Rust, single-binary WebGPU inference engine that provides OpenAI-compatible API endpoints for running GGUF models locally without...

OpenRLHF is a production-ready framework for reinforcement learning from human feedback (RLHF) built on Ray and vLLM, enabling scalable training of...

Argos Translate is an open-source Python library for offline machine translation using OpenNMT models. It serves developers and end-users who need...

GROBID is a machine learning library for extracting, parsing, and re-structuring scholarly PDF documents into structured XML/TEI encoded data. It...

LLaMA Factory is a unified fine-tuning framework that supports 100+ large language models and vision-language models through a zero-code CLI and web...

OpenVINO is an open-source toolkit for optimizing and deploying deep learning inference across diverse hardware platforms (CPU, GPU, NPU) and...

Optimum is a hardware optimization toolkit that extends HuggingFace's Transformers, Diffusers, TIMM, and Sentence-Transformers libraries to...

RTranslator is an offline, real-time translation app for Android that enables live conversations between two or more people speaking different...