meta-pytorch / krakenLinks
Triton-based Symmetric Memory operators and examples
β70Updated 2 months ago
Alternatives and similar repositories for kraken
Users that are interested in kraken are comparing it to the libraries listed below
Sorting:
- extensible collectives library in tritonβ91Updated 8 months ago
- A bunch of kernels that might make stuff slower πβ69Updated last week
- Framework to reduce autotune overhead to zero for well known deployments.β91Updated 3 months ago
- β99Updated last year
- PyTorch bindings for CUTLASS grouped GEMM.β135Updated 7 months ago
- Boosting 4-bit inference kernels with 2:4 Sparsityβ90Updated last year
- DeeperGEMM: crazy optimized versionβ73Updated 7 months ago
- QuTLASS: CUTLASS-Powered Quantized BLAS for Deep Learningβ148Updated last month
- Autonomous GPU Kernel Generation via Deep Agentsβ192Updated last week
- β115Updated last year
- Tritonbench is a collection of PyTorch custom operators with example inputs to measure their performance.β306Updated this week
- Efficient Long-context Language Model Training by Core Attention Disaggregationβ73Updated this week
- Applied AI experiments and examples for PyTorchβ311Updated 4 months ago
- Triton-based implementation of Sparse Mixture of Experts.β259Updated 2 months ago
- β39Updated 2 weeks ago
- β65Updated 8 months ago
- Collection of kernels written in Triton languageβ173Updated 8 months ago
- Ship correct and fast LLM kernels to PyTorchβ127Updated last week
- Accelerating MoE with IO and Tile-aware Optimizationsβ469Updated this week
- ring-attention experimentsβ160Updated last year
- β268Updated last week
- π Collection of components for development, training, tuning, and inference of foundation models leveraging PyTorch native components.β217Updated 2 weeks ago
- β15Updated last month
- Utility scripts for PyTorch (e.g. Make Perfetto show some disappearing kernels, Memory profiler that understands more low-level allocatioβ¦β73Updated 3 months ago
- This repository contains the experimental PyTorch native float8 training UXβ227Updated last year
- Cataloging released Triton kernels.β280Updated 3 months ago
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.β104Updated 6 months ago
- β52Updated 7 months ago
- Distributed MoE in a Single Kernel [NeurIPS '25]β157Updated last week
- Odysseus: Playground of LLM Sequence Parallelismβ78Updated last year