Create tomczak2019simulating.markdown

Author: Miltos

_publications/tomczak2019simulating.markdown

@@ -0,0 +1,12 @@
+---
+layout: publication
+title: "Simulating Execution Time of Tensor Programs using Graph Neural Networks"
+authors: J. M. Tomczak, R. Lepert, A. Wiggers
+conference: Representation Learning on Graphs and Manifolds at ICLR
+year: 2019
+bibkey: tomczak2019simulating
+additional_links:
+   - {name: "ArXiV", url: "https://arxiv.org/abs/1904.11876"}
+tags: ["GNN"]
+---
+Optimizing the execution time of tensor program, e.g., a convolution, involves finding its optimal configuration. Searching the configuration space exhaustively is typically infeasible in practice. In line with recent research using TVM, we propose to learn a surrogate model to overcome this issue. The model is trained on an acyclic graph called an abstract syntax tree, and utilizes a graph convolutional network to exploit structure in the graph. We claim that a learnable graph-based data processing is a strong competitor to heuristic-based feature extraction. We present a new dataset of graphs corresponding to configurations and their execution time for various tensor programs. We provide baselines for a runtime prediction task.