pytorch

https://github.com/Lightning-AI/lightning-thunder Make PyTorch models up to 40% faster! Thunder is a source to source compiler for PyTorch. It enables using different hardware executors at once; across one or thousands of GPUs.

https://github.com/roboflow/notebooks Examples and tutorials on using SOTA computer vision models and techniques. Learn everything from old-school ResNet, through YOLO and object-detection transformers like DETR, to the latest models like Grounding D

🔧🔗https://github.com/modelscope/FunASR A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models, Supporting Speech Recognition, Voice Activity Detection, Text Post-processing

🔧🔗https://github.com/yandexdataschool/Practical_RL A course in reinforcement learning in the wild

🔧🔗https://github.com/lucidrains/alphafold3-pytorch Implementation of Alphafold 3 from Google Deepmind in Pytorch

🔧🔗https://github.com/MinishLab/tokenlearn

Pre-train Static Word Embeddings

Allosaurus

🔧🔗https://github.com/xinjli/allosaurus Allosaurus is a pretrained universal phone recognizer for more than 2000 languages

🔧🔗https://github.com/huggingface/lerobot

🤗 LeRobot: Making AI for Robotics more accessible with end-to-end learning

🔧🔗https://github.com/pytorch/serve

Serve, optimize and scale PyTorch models in production

Faceformer

🔧🔗 github.com/manycore-research/faceformer [CVPR 2022] Neural Face Identification in a 2D Wireframe Projection of a Manifold Object

🔧🔗https://github.com/manycore-research/PlankAssembly [ICCV 2023] PlankAssembly: Robust 3D Reconstruction from Three Orthographic Views with Learnt Shape Programs

🔧🔗https://github.com/lucidrains/rectified-flow-pytorch

Implementation of rectified flow and some of its followup research / improvements in Pytorch

https://github.com/Lightning-AI/lightning-Habana Lightning support for Intel Habana accelerators.

🔧🔗https://github.com/blaisewf/rvc-cli 🚀 RVC + UVR = A perfect set of tools for voice cloning, easily and free!

🔧🔗https://github.com/cognitivecomputations/dolphinflow-optimizer DolphinFlow is a pragmatic, hardware agnostic, memory-efficient PyTorch optimizer that uses orthogonalization techniques to improve neural network training stability and performance, particularly for models with large, dense layers.

DolphinFlow is inspired by the optimizers I've worked with, including Muon and AdamW, with a focus on fine tuning pretrained and instruct tuned large language models such as Llama, Mistral, Qwen, DeepSeek, Gemma, Phi, and others.

DolphinFlow is about results. I'm a practitioner, not a theoretician. I build things that do things.

https://github.com/homebrewltd/WhisperSpeech An Open Source text-to-speech system built by inverting Whisper. 🔗 https://collabora.github.io/WhisperSpeech/

https://github.com/lucidrains/genetic-algorithm-pytorch Toy genetic algorithm in Pytorch

🔧🔗https://github.com/lucidrains/e2-tts-pytorch Implementation of E2-TTS, "Embarrassingly Easy Fully Non-Autoregressive Zero-Shot TTS", in Pytorch

🔧🔗 meta-llama/llama-recipes Scripts for fine-tuning Meta Llama with composable FSDP & PEFT methods to cover single/multi-node GPUs.

https://github.com/THUDM/SwissArmyTransformer SwissArmyTransformer is a flexible and powerful library to develop your own Transformer variants.