KellerJordan / Muon
Muon optimizer: +>30% sample efficiency with <3% wallclock overhead
β597Updated last month
Alternatives and similar repositories for Muon:
Users that are interested in Muon are comparing it to the libraries listed below
- Helpful tools and examples for working with flex-attentionβ757Updated this week
- Implementation of π Ring Attention, from Liu et al. at Berkeley AI, in Pytorchβ511Updated 6 months ago
- [ICLR2025 Spotlightπ₯] Official Implementation of TokenFormer: Rethinking Transformer Scaling with Tokenized Model Parametersβ555Updated 2 months ago
- Code for Adam-mini: Use Fewer Learning Rates To Gain More https://arxiv.org/abs/2406.16793β409Updated 3 weeks ago
- Implementation of the sparse attention pattern proposed by the Deepseek team in their "Native Sparse Attention" paperβ607Updated last month
- When it comes to optimizers, it's always better to be safe than sorryβ222Updated last month
- Muon is Scalable for LLM Trainingβ1,039Updated last month
- β436Updated 2 weeks ago
- β280Updated 2 weeks ago
- π³ Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"β647Updated last month
- Normalized Transformer (nGPT)β174Updated 5 months ago
- Quick implementation of nGPT, learning entirely on the hypersphere, from NvidiaAIβ281Updated last month
- Memory layers use a trainable key-value lookup mechanism to add extra parameters to a model without increasing FLOPs. Conceptually, sparsβ¦β322Updated 4 months ago
- A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.β536Updated last week
- Understanding R1-Zero-Like Training: A Critical Perspectiveβ908Updated 3 weeks ago
- Large Context Attentionβ707Updated 3 months ago
- Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"β232Updated 2 months ago
- Annotated version of the Mamba paperβ483Updated last year
- CIFAR-10 speedruns: 94% in 2.6 seconds and 96% in 27 secondsβ233Updated 2 months ago
- π Efficiently (pre)training foundation models with native PyTorch features, including FSDP for training and SDPA implementation of Flashβ¦β244Updated this week
- [ICLR 2025] Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modelingβ867Updated last week
- Ring attention implementation with flash attentionβ757Updated last month
- Scalable and Performant Data Loadingβ252Updated this week
- [ICML2024 (Oral)] Official PyTorch implementation of DoRA: Weight-Decomposed Low-Rank Adaptationβ775Updated 7 months ago
- Repo for "LoLCATs: On Low-Rank Linearizing of Large Language Models"β231Updated 3 months ago
- Simple and Effective Masked Diffusion Language Modelβ376Updated 3 weeks ago
- The official implementation of Tensor ProducT ATTenTion Transformer (T6)β367Updated 3 weeks ago
- Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Modelsβ569Updated 3 weeks ago
- πΎ OAT: A research-friendly framework for LLM online alignment, including preference learning, reinforcement learning, etc.β338Updated this week
- Code for BLT research paperβ1,546Updated this week