Search

Training-free language model modification: Remove censorship, suppress slop, reduce sycophancy, and more

Scale computer-use 2.0 with open-source drivers, cross-OS sandboxes, and benchmarks for training, evaluation, and data generation.

🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

An open-source document parsing model with SOTA accuracy and speed, using an RSI loop during post-training.

vibe coded supermimic within 1h, LAFAN passed within 1.5 gpuh training

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

Ongoing research training transformer models at scale

A high-performance distributed file system designed to address the challenges of AI training and inference workloads.

ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

🦆 Microduck 飞特 HD1910 机器人鸭复刻：单脚独立实机、现成动作模型、中文中控、自制 HAT/IMU 与三步教程。正常复刻无需训练。Feetech HD1910 Microduck replica: one-legged balance, ready-to-use models, Chinese console, custom HAT/IMU & 3-step build guides. No training required.

Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.

Open-source deep-learning framework for building, training, and fine-tuning deep learning models using state-of-the-art Physics-ML methods

A high-performance framework for training LLMs, VLMs, diffusion, and embodied models on NVIDIA GPUs and Kunlun XPUs. Supports Pi0.5, GR00T-N1.7, FastWAM, Cosmos3, DreamZero, DeepSeek-V4, GLM-5.3, Kimi-K3 and more.

Beginner, advanced, expert level Rust training material

A unified inference and post-training framework for accelerated video generation.

⚡️SwanLab - an open-source, modern-design AI training tracking and visualization tool. Supports Cloud / Self-hosted use. Integrated with PyTorch / Transformers / verl / LLaMA Factory / ms-swift / Ultralytics / MMEngine / Keras etc.

CPS-2 Recomp: unofficial native PC ports of Capcom CPS-2 arcade games (Street Fighter Alpha 2, Super Puzzle Fighter II Turbo, Vampire Savior) via static recompilation, with training modes, in one launcher: macOS, Linux, Steam Deck, Windows. Requires your own ROM sets. Beta.

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

AllenAI's post-training codebase

Popular Simple Open Source Online Judge based on PHP/C++/MySQL/Linux for ACM/ICPC and NOIP training, with easy installation. 简单实用的开源OJ系统

Ultralytics YOLOv5 in PyTorch for object detection, instance segmentation, classification, training, and export.

open source training courses about distributed database and distributed systems

Go Training Class Material :

Training library for Megatron-based models with bidirectional Hugging Face conversion capability

A nano replica of Jev: parallel decisions, dynamic candidates, and an end-to-end training pipeline.

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen3.6, GPT-OSS, Llama, and more!

Engineering toolkit for building scalable AI systems, including model training pipelines, LLM optimization utilities, and data engineering tools.

A SuperPod-Affinity Training Acceleration Framework.