maximal update parametrization (µP)
☆1,756Jul 17, 2024Updated 2 years ago
Alternatives and similar repositories for mup
Users that are interested in mup are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- some common Huggingface transformers in maximal update parametrization (µP)☆88Mar 14, 2022Updated 4 years ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.☆206Jan 19, 2026Updated 9 months ago
- A PyTorch native platform for training generative AI models☆5,858Updated this week
- WIP☆97Aug 13, 2024Updated 2 years ago
- PyTorch extensions for high performance and large scale training.☆3,404Apr 26, 2025Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Schedule-Free Optimization in PyTorch☆2,330Jul 28, 2026Updated 3 months ago
- Helpful tools and examples for working with flex-attention☆1,262Updated this week
- ☆1,600Mar 25, 2026Updated 7 months ago
- Hackable and optimized Transformers building blocks, supporting a composable construction.☆10,572Updated this week
- Accessible large language models via k-bit quantization for PyTorch.☆8,535Updated this week
- Supercharge Your Model Training☆5,508Apr 29, 2026Updated 6 months ago
- Flexible and powerful tensor operations for readable and reliable code (for pytorch, jax, TF and others)☆9,626Aug 26, 2026Updated 2 months ago
- 🚀 Efficient implementations for emerging model architectures☆6,003Updated this week
- Minimalistic large language model 3D-parallelism training☆2,850Oct 8, 2026Updated 3 weeks ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Fast and memory-efficient exact attention☆25,221Updated this week
- A repo for distributed training of language models with Reinforcement Learning via Human Feedback (RLHF)☆4,756Jan 8, 2024Updated 2 years ago
- A library for unit scaling in PyTorch☆136Jul 11, 2025Updated last year
- Ongoing research training transformer models at scale☆18,293Updated this week
- 🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (i…☆9,930Updated this week
- functorch is JAX-like composable function transforms for PyTorch.☆1,434Aug 21, 2025Updated last year
- A high-performance Python-based I/O system for large (and small) deep learning problems, with strong support for PyTorch.☆3,212Feb 9, 2026Updated 8 months ago
- A concise but complete full-attention transformer with a set of promising experimental features from various papers☆5,961Updated this week
- The official implementation of “Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training”☆1,000Jan 30, 2024Updated 2 years ago
- Deploy open-source AI quickly and easily - Special Bonus Offer • AdRunpod Hub is built for open source. One-click deployment and autoscaling endpoints without provisioning your own infrastructure.
- Foundation Architecture for (M)LLMs☆3,143Apr 11, 2024Updated 2 years ago
- Simple implementation of muP, based on Spectral Condition for Feature Learning. The implementation is SGD only, dont use it for Adam☆89Jul 28, 2024Updated 2 years ago
- ☆3,005Aug 3, 2026Updated 2 months ago
- Flax is a neural network library for JAX that is designed for flexibility.☆7,361Updated this week
- Efficient Triton Kernels for LLM Training☆6,678Updated this week
- Development repository for the Triton language and compiler☆20,481Updated this week
- A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on H…☆3,598Updated this week
- Repo for external large-scale work☆6,544Apr 27, 2024Updated 2 years ago
- Structured state space sequence models☆2,930Jul 17, 2024Updated 2 years ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- An implementation of model parallel autoregressive transformers on GPUs, based on the Megatron and DeepSpeed libraries☆7,470Oct 11, 2026Updated 2 weeks ago
- ☆1,273Jul 30, 2024Updated 2 years ago
- Dion optimizer algorithm☆558Sep 30, 2026Updated 3 weeks ago
- Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.☆6,267Aug 22, 2025Updated last year
- DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.☆43,269Updated this week
- Scaling Data-Constrained Language Models☆351Jun 28, 2025Updated last year
- Freeing data processing from scripting madness by providing a set of platform-agnostic customizable pipeline processing blocks.☆3,409Updated this week