MagellaX / StreamAttnLinks
A high-performance attention mechanism that computes softmax normalization in a single streaming pass using running accumulators (online softmax).
☆28Updated 2 months ago
Alternatives and similar repositories for StreamAttn
Users that are interested in StreamAttn are comparing it to the libraries listed below
Sorting:
- Learning about CUDA by writing PTX code.☆150Updated last year
- PTX-Tutorial Written Purely By AIs (Deep Research of Openai and Claude 3.7)☆66Updated 8 months ago
- Quantized LLM training in pure CUDA/C++.☆221Updated this week
- SIMD quantization kernels☆93Updated 3 months ago
- PCCL (Prime Collective Communications Library) implements fault tolerant collective communications over IP☆141Updated 3 months ago
- in this repository, i'm going to implement increasingly complex llm inference optimizations☆75Updated 6 months ago
- NanoGPT-speedrunning for the poor T4 enjoyers☆73Updated 7 months ago
- NSA Triton Kernels written with GPT5 and Opus 4.1☆66Updated 4 months ago
- My submission for the GPUMODE/AMD fp8 mm challenge☆29Updated 6 months ago
- Tensor library with autograd using only Rust's standard library☆70Updated last year
- pytorch from scratch in pure C/CUDA and python☆41Updated last year
- an open source reproduction of NVIDIA's nGPT (Normalized Transformer with Representation Learning on the Hypersphere)☆108Updated 9 months ago
- Experimental GPU language with meta-programming☆24Updated last year
- ☆86Updated last month
- High-Performance SGEMM on CUDA devices☆113Updated 10 months ago
- A simple MLX implementation for pretraining LLMs on Apple Silicon.☆84Updated 3 months ago
- coding CUDA everyday!☆71Updated last week
- Write a fast kernel and run it on Discord. See how you compare against the best!☆64Updated this week
- look how they massacred my boy☆63Updated last year
- Rust Implementation of micrograd☆53Updated last year
- Samples of good AI generated CUDA kernels☆94Updated 6 months ago
- ☆28Updated last year
- LLM training in simple, raw C/CUDA☆108Updated last year
- Because it's there.☆16Updated last year
- ☆70Updated last month
- A really tiny autograd engine☆96Updated 6 months ago
- ☆81Updated last week
- Official Problem Sets / Reference Kernels for the GPU MODE Leaderboard!☆174Updated 2 weeks ago
- Simple Transformer in Jax☆140Updated last year
- NanoGPT (124M) quality in 2.67B tokens☆28Updated 3 months ago