The CoRa Tensor Compiler: Compilation for Ragged Tensors with Minimal Padding

2021/10/19 by Pratik Fegade, Fegade, Pratik, Tianqi Chen +5 · 1 citation
Computer Science · Mathematics · #Advanced Neural Network Applications #Computational Physics and Python Applications #FOS: Computer and information sciences #Machine Learning (cs.LG) #Parallel Computing and Optimization Techniques #Tensor decomposition and applications

paper · pdf · doi:10.48550/arxiv.2110.10221

openalex publication_date 2021/10/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

The CoRa tensor compiler and benchmarks to evaluate it are being released here as part of the artifact evaluation process for MLSys 2022.

Cited by

• Operon: Incremental Construction of Ragged Data via Named Dimensions

Related

• TensorIR: An Abstraction for Automatic Tensorized Program Optimization
• FTuner: A Fast Dynamic Shape Tensors Program Auto-Tuner for Deep Learning Compilers
• Tensor Comprehensions: Framework-Agnostic High-Performance Machine Learning Abstractions
• Learning to Optimize Tensor Programs
• QiMeng-TensorOp: Automatically Generating High-Performance Tensor Operators with Hardware Primitives