TF-Coder: Program Synthesis for Tensor Manipulations

Abstract

Deep learning frameworks such as TensorFlow and PyTorch come with steep learning curves. We present a tool called TF-Coder for programming by example in TensorFlow. It uses a bottom-up weighted enumerative search with learned models that prioritize relevant operations. TF-Coder solves 63 of 70 real-world tasks within 5 minutes, often achieving superhuman performance -- finding solutions that are simpler than those written by TensorFlow experts, in less time.

PDF NeurIPSW OpenReview Semantic Scholar

Cite

Text

Shi et al. "TF-Coder: Program Synthesis for Tensor Manipulations." NeurIPS 2020 Workshops: CAP, 2020.

Markdown

[Shi et al. "TF-Coder: Program Synthesis for Tensor Manipulations." NeurIPS 2020 Workshops: CAP, 2020.](https://mlanthology.org/neuripsw/2020/shi2020neuripsw-tfcoder/)

BibTeX

@inproceedings{shi2020neuripsw-tfcoder,
  title     = {{TF-Coder: Program Synthesis for Tensor Manipulations}},
  author    = {Shi, Kensen and Bieber, David and Singh, Rishabh},
  booktitle = {NeurIPS 2020 Workshops: CAP},
  year      = {2020},
  url       = {https://mlanthology.org/neuripsw/2020/shi2020neuripsw-tfcoder/}
}