User-friendly implementation of the Mixture-of-Sparse-Attention (MoSA). MoSA selects distinct tokens for each head with expert choice routing providing a content-based sparse attention mechanism.
☆29May 3, 2025Updated last year
Alternatives and similar repositories for MoSA
Users that are interested in MoSA are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- The official implementation of HybridNorm: Towards Stable and Efficient Transformer Training via Hybrid Normalization☆19Mar 7, 2025Updated last year
- sigma-MoE layer☆21Jan 5, 2024Updated 2 years ago
- ☆12Apr 29, 2024Updated 2 years ago
- ☆14Oct 3, 2024Updated last year
- A collection of AI for Drug Design related papers and corresponding code sources (in progress).☆18Feb 25, 2022Updated 4 years ago
- 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.
- [ICML 2025] LaCache: Ladder-Shaped KV Caching for Efficient Long-Context Modeling of Large Language Models☆17Nov 4, 2025Updated 8 months ago
- Official code repo for paper "Great Memory, Shallow Reasoning: Limits of kNN-LMs"☆24Apr 30, 2025Updated last year
- MonarchAttention: Zero-Shot Conversion to Fast, Hardware-Aware Structured Attention (NeurIPS'25 Spotlight)☆26Feb 22, 2026Updated 5 months ago
- ☆18Dec 2, 2024Updated last year
- Activation-Steered Compression☆17Jan 30, 2026Updated 5 months ago
- ☆135May 29, 2025Updated last year
- Official repository for the paper "SwitchHead: Accelerating Transformers with Mixture-of-Experts Attention"☆101Sep 30, 2024Updated last year
- ☆13Aug 13, 2024Updated last year
- ☆20Jan 10, 2025Updated last year
- Virtual machines for every use case on DigitalOcean • AdGet dependable uptime with 99.99% SLA, simple security tools, and predictable monthly pricing with DigitalOcean's virtual machines, called Droplets.
- Template for projects in PyTorch powered with PyTorch Lightning, Telegrad and MLflow. Get updates on mobile and streamline PyTorch code f…☆10May 1, 2023Updated 3 years ago
- Clustered Compositional Embeddings☆13Oct 25, 2023Updated 2 years ago
- Revealing and unlocking the context boundary of reward models☆21May 10, 2026Updated 2 months ago
- [ICML 2024] SIRFShampoo: Structured inverse- and root-free Shampoo in PyTorch (https://arxiv.org/abs/2402.03496)☆15Nov 4, 2024Updated last year
- codes and plots for "Active-Dormant Attention Heads: Mechanistically Demystifying Extreme-Token Phenomena in LLMs"☆11Dec 30, 2024Updated last year
- Don't just regulate gradients like in Muon, regulate the weights too☆32Jul 30, 2025Updated 11 months ago
- ☆152Jun 2, 2023Updated 3 years ago
- [NeurIPS 2024] | An Efficient Recipe for Long Context Extension via Middle-Focused Positional Encoding☆22Oct 10, 2024Updated last year
- KnowLA: Enhancing Parameter-efficient Finetuning with Knowledgeable Adaptation, NAACL 2024☆16Jul 29, 2024Updated last year
- 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.
- ☆28Jul 2, 2026Updated 3 weeks ago
- ☆10Jun 11, 2024Updated 2 years ago
- Mixture of Attention Heads☆53Oct 10, 2022Updated 3 years ago
- Python script to transform the Mobile Detect JSON database into an UA-based mobile detection VCL subroutine easily integrable in any Varn…☆14Nov 13, 2023Updated 2 years ago
- Code and models for the paper: Hybrid Linear Attention Done Right: Efficient Distillation and Effective Architectures for Extremely Long …☆40Apr 9, 2026Updated 3 months ago
- The official implementation of paper: SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction.☆54Oct 18, 2024Updated last year
- RAG-RewardBench: Benchmarking Reward Models in Retrieval Augmented Generation for Preference Alignment☆18Dec 19, 2024Updated last year
- Suri: Multi-constraint instruction following for long-form text generation [EMNLP’24]☆27Oct 3, 2025Updated 9 months ago
- Codebase for Context-aware Meta-learned Loss Scaling (CaMeLS). https://arxiv.org/abs/2305.15076.☆26Jan 23, 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.
- ☆29Apr 7, 2024Updated 2 years ago
- iADMM for a low-rank representation optimization problem☆13Feb 5, 2021Updated 5 years ago
- ☆27May 12, 2026Updated 2 months ago
- ☆23Mar 31, 2023Updated 3 years ago
- The repository for papaer "Distance between Relevant Information Pieces Causes Bias in Long-Context LLMs"☆14Dec 16, 2024Updated last year
- Code for experiments on transformers using Markovian data.☆22Nov 22, 2024Updated last year
- ☆16May 18, 2026Updated 2 months ago