A computational model for TensorFlow: an introduction
Martı́n Abadi, Michael Isard, Derek Gordon Murray · 2017
TensorFlow is a powerful, programmable system for machine learning. This paper aims to provide the basics of a conceptual framework for understanding the behavior of TensorFlow models during training and inference: it describes an operational semantics, of the kind common in the literature on programming languages. More broadly, the paper suggests that a programming-language perspective is fruitful in designing and in explaining systems such as TensorFlow.
Cited by 69 works. doi:10.1145/3088525.3088527 Open-access PDF
Authors
References
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- Theano: new features and speed improvements
- The Semantics of a Simple Language for Parallel Programming.
- HOGWILD!: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent
- Static Analysis of Shape in TensorFlow Programs
- Specifying Systems: The TLA+ Language and Tools for Hardware and Software Engineers
- Executing a program on the MIT tagged-token dataflow architecture
- Incremental, iterative data processing with timely dataflow
- Deep Learning with Dynamic Computation Graphs
- A fully abstract trace model for dataflow and asynchronous networks