paper-with-me

홈 › Papers

LazyTensor: combining eager execution with domain-specific compilers

2021-02-26 · Alex Suhan, Davide Libenzi, Ailing Zhang, Parker Schuh, Brennan Saeta, Jie Young Sohn, Denys Shabalin

Domain-specific optimizing compilers have demonstrated significant performance and portability benefits, but require programs to be represented in their specialized IRs. Existing frontends to these compilers suffer from the "language subset problem" where some host language features are unsupported in the subset of the user's program that interacts with the domain-specific compiler. By contrast, define-by-run ML frameworks-colloquially called "eager" mode-are popular due to their ease of use and expressivity, where the full power of the host programming language can be used. LazyTensor is a technique to target domain specific compilers without sacrificing define-by-run ergonomics. Initially developed to support PyTorch on Cloud TPUs, the technique, along with a substantially shared implementation, has been used by Swift for TensorFlow across CPUs, GPUs, and TPUs, demonstrating the generality of the approach across (1) Tensor implementations, (2) hardware accelerators, and (3) programming languages.

📄 PDF Abstract BibTeX arXiv:2102.13267

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

TensorFlow Eager: A Multi-Stage, Python-Embedded DSL for Machine Learning

2019-02-27 · Akshay Agrawal, Akshay Naresh Modi, Alexandre Passos, Allen Lavoie 외

TensorFlow Eager is a multi-stage, Python-embedded domain-specific language for hardware-accelerated machine learning, suitable for both interactive research and production. TensorFlow, which TensorFlow Eager extends, re…

BIG-bench Machine Learning

AscendCraft: Automatic Ascend NPU Kernel Generation via DSL-Guided Transcompilation

2026-01-30 · Zhongzhen Wen, Shudi Shao, Zhong Li, Yu Ge 외 arxiv

The performance of deep learning models critically depends on efficient kernel implementations, yet developing high-performance kernels for specialized accelerators remains time-consuming and expertise-intensive. While r…

Executing as You Generate: Hiding Execution Latency in LLM Code Interpreters

2026-04-01 · Zhensu Sun, Zhihao Lin, Zhi Chen, Chengran Yang 외 arxiv

Current LLM systems are increasingly equipped with a code interpreter that executes generated code to obtain results. This works serially: the model first generates the complete code, then an interpreter executes it. Thi…

veScale: Consistent and Efficient Tensor Programming with Eager-Mode SPMD

2025-09-05 · Youjie Li, Cheng Wan, Zhiqi Lin, Hongyu Zhu 외 arxiv

Large Language Models (LLMs) have scaled rapidly in size and complexity, requiring increasingly intricate parallelism for distributed training, such as 3D parallelism. This sophistication motivates a shift toward simpler…

TimeRL: Efficient Deep Reinforcement Learning with Polyhedral Dependence Graphs

2025-01-09 · Pedro F. Silvestre, Peter Pietzuch

Modern deep learning (DL) workloads increasingly use complex deep reinforcement learning (DRL) algorithms that generate training data within the learning loop. This results in programs with several nested loops and dynam…

CPUDeep Reinforcement LearningGPUreinforcement-learning+2