fal-ai-community / NativeSparseAttentionLinks
research impl of Native Sparse Attention (2502.11089)
☆54Updated 3 months ago
Alternatives and similar repositories for NativeSparseAttention
Users that are interested in NativeSparseAttention are comparing it to the libraries listed below
Sorting:
- supporting pytorch FSDP for optimizers☆79Updated 5 months ago
- Tiny re-implementation of MDM in style of LLaDA and nano-gpt speedrun☆52Updated 2 months ago
- ☆78Updated 11 months ago
- ☆21Updated 6 months ago
- Flash-Muon: An Efficient Implementation of Muon Optimizer☆115Updated last week
- Focused on fast experimentation and simplicity☆73Updated 5 months ago
- ☆56Updated 2 months ago
- Collection of autoregressive model implementation☆85Updated last month
- an open source reproduction of NVIDIA's nGPT (Normalized Transformer with Representation Learning on the Hypersphere)☆100Updated 2 months ago
- Triton Implementation of HyperAttention Algorithm☆48Updated last year
- Official repository for the paper "Approximating Two-Layer Feedforward Networks for Efficient Transformers"☆37Updated last year
- NanoGPT-speedrunning for the poor T4 enjoyers☆66Updated last month
- Fast, Modern, Memory Efficient, and Low Precision PyTorch Optimizers☆93Updated 10 months ago
- https://x.com/BlinkDL_AI/status/1884768989743882276☆28Updated last month
- ☆31Updated last month
- ☆61Updated 6 months ago
- Combining SOAP and MUON☆16Updated 3 months ago
- ☆49Updated last year
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training☆127Updated last year
- ☆19Updated 2 weeks ago
- Here we will test various linear attention designs.☆58Updated last year
- ☆33Updated 8 months ago
- Implementation of GateLoop Transformer in Pytorch and Jax☆88Updated 11 months ago
- Experiment of using Tangent to autodiff triton☆79Updated last year
- DPO, but faster 🚀☆42Updated 5 months ago
- ☆95Updated 4 months ago
- WIP☆93Updated 9 months ago
- GoldFinch and other hybrid transformer components☆45Updated 10 months ago
- Pytorch implementation of the PEER block from the paper, Mixture of A Million Experts, by Xu Owen He at Deepmind☆124Updated 9 months ago
- ☆53Updated last year