Evaluating language models for efficient code generation

J Liu, S **e, J Wang, Y Wei, Y Ding, L Zhang - arxiv preprint arxiv …, 2024 - arxiv.org
We introduce Differential Performance Evaluation (DPE), a framework designed to reliably
evaluate Large Language Models (LLMs) for efficient code generation. Traditional coding …

Tpugraphs: A performance prediction dataset on large tensor computational graphs

M Phothilimthana, S Abu-El-Haija… - Advances in …, 2024 - proceedings.neurips.cc
Precise hardware performance models play a crucial role in code optimizations. They can
assist compilers in making heuristic decisions or aid autotuners in identifying the optimal …

Language models for code optimization: Survey, challenges and future directions

J Gong, V Voskanyan, P Brookes, F Wu, W Jie… - arxiv preprint arxiv …, 2025 - arxiv.org
Language models (LMs) built upon deep neural networks (DNNs) have recently
demonstrated breakthrough effectiveness in software engineering tasks like code …

WACO: learning workload-aware co-optimization of the format and schedule of a sparse tensor program

J Won, C Mendis, JS Emer… - Proceedings of the 28th …, 2023 - dl.acm.org
In this paper, we present WACO, a novel method of co-optimizing the format and the
schedule of a given sparsity pattern in a sparse tensor program. A core challenge in this …

Supersonic: Learning to generate source code optimizations in C/C++

Z Chen, S Fang, M Monperrus - IEEE Transactions on Software …, 2024 - ieeexplore.ieee.org
Software optimization refines programs for resource efficiency while preserving functionality.
Traditionally, it is a process done by developers and compilers. This paper introduces a third …

Tenset: A large-scale program performance dataset for learned tensor compilers

L Zheng, R Liu, J Shao, T Chen… - Thirty-fifth Conference …, 2021 - openreview.net
Search-based tensor compilers can greatly accelerate the execution of machine learning
models by generating high-performance tensor programs, such as matrix multiplications and …

A flexible approach to autotuning multi-pass machine learning compilers

PM Phothilimthana, A Sabne, N Sarda… - 2021 30th …, 2021 - ieeexplore.ieee.org
Search-based techniques have been demonstrated effective in solving complex optimization
problems that arise in domain-specific compilers for machine learning (ML). Unfortunately …

Tensor program optimization with probabilistic programs

J Shao, X Zhou, S Feng, B Hou, R Lai… - Advances in …, 2022 - proceedings.neurips.cc
Automatic optimization for tensor programs becomes increasingly important as we deploy
deep learning in various environments, and efficient optimization relies on a rich search …

Tlp: A deep learning-based cost model for tensor program tuning

Y Zhai, Y Zhang, S Liu, X Chu, J Peng, J Ji… - Proceedings of the 28th …, 2023 - dl.acm.org
Tensor program tuning is a non-convex objective optimization problem, to which search-
based approaches have proven to be effective. At the core of the search-based approaches …

PyGim: An Efficient Graph Neural Network Library for Real Processing-In-Memory Architectures

C Giannoula, P Yang, I Fernandez, J Yang… - Proceedings of the …, 2024 - dl.acm.org
Graph Neural Networks (GNNs) are emerging models to analyze graph-structure data. GNN
execution involves both compute-intensive and memory-intensive kernels. The latter kernels …