Pytorch

An open-source educational book teaching practical debugging methodologies and tools across Unix, compiled programs, Python, and PyTorch...

InferenceX is an open-source continuous inference benchmarking research platform that tracks real-time performance of LLM inference frameworks across...

LMCache is a KV cache management layer that optimizes Large Language Model inference by providing efficient memory management across multiple...

Supervision is a Python toolkit for building computer vision applications, providing reusable components for model integration, real-time annotation,...

A hands-on curriculum and code repository that teaches modern reinforcement learning from foundational control algorithms through LLM alignment,...

This is an educational resource portal for fine-tuning large language models, covering SFT and reinforcement learning workflows (GRPO, GSPO), data...

VoxCPM2 is a tokenizer-free text-to-speech system generating multilingual speech with 48kHz audio quality, supporting 30 languages, voice design from...

vLLM-Omni is a framework for efficient serving of multimodal and omni-modality models (supporting text, audio, image, video, and world models) built...

FunASR is an industrial-grade speech-to-text toolkit supporting 50+ languages with specialized features including speaker diarization, emotion...

This repository is the official code companion to a Manning book that teaches how to build a reasoning LLM from scratch using PyTorch, starting with...

PhysicsNeMo is NVIDIA's open-source deep-learning framework for building Physics-informed AI models, combining neural operators, GNNs, and...

Applio is a voice conversion tool that uses RVC (Retrieval-based Voice Conversion) and related models to transform speech in real-time with high...

Chatterbox TTS Server is a self-hosted text-to-speech API server that wraps Resemble AI's open-source Chatterbox TTS model, offering...

FlashInfer is a specialized GPU kernel library designed for high-performance LLM inference, providing optimized implementations for attention, GEMM,...

YomiToku is a specialized Python package for AI-powered optical character recognition and document layout analysis designed specifically for Japanese...

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

Qualcomm AI Hub Models is a curated collection of pre-optimized machine learning models designed specifically for deployment on Qualcomm devices...

Machine Learning Zoomcamp is a free 4-month educational course teaching ML engineering end-to-end, from algorithm fundamentals (regression,...

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

SenseVoice is a multilingual speech foundation model supporting ASR, emotion recognition, language identification, and audio event detection across...

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

stable-worldmodel is a unified research platform for world model development, spanning data collection, training, and model-predictive control...

PaddleOCR2Pytorch converts PaddleOCR models to PyTorch format, enabling OCR inference across detection, recognition, and document structure parsing....

TensorRT LLM is an NVIDIA-developed framework that optimizes inference for large language models and visual generation models on NVIDIA GPUs through...

Silero VAD is a pre-trained, enterprise-grade Voice Activity Detector (VAD) available via PyPI, PyTorch Hub, and ONNX Runtime. It identifies speech...

Braindecode is a Python toolbox for decoding raw electrophysiological brain signals (EEG, ECoG, MEG) using deep learning models. It provides dataset...

Nano-vLLM is a lightweight, from-scratch implementation of vLLM optimized for fast LLM inference on consumer GPUs, achieving comparable or better...

ComfyUI is a node-graph-based GUI, API, and backend engine for running diffusion models (image, video, 3D, audio generation) with granular control...

vLLM is a high-throughput, memory-efficient inference and serving engine for large language models, using PagedAttention and continuous batching to...

TorchRL is a PyTorch-native reinforcement learning library providing modular, composable components for building RL systems including environments,...