Björn Hartmann on Debugging Deep Learning Programs
Training deep neural networks can generate non-descriptive error messages or produce unusual output without any explicit errors at all. While experts rely on tacit knowledge to apply debugging strategies, non-experts lack the experience required to interpret model output and correct Deep Learning (DL) programs.
With Eldon Schoop and Forrest Huang, Björn Hartmann published "UMLAUT: Debugging Deep Learning Programs using Program Structure and Model Behavior" to identify DL debugging heuristics and strategies used by experts, and use them to guide the design of Umlaut. Umlaut checks DL program structure and model behavior against these heuristics; provides human-readable error messages to users; and annotates erroneous model output to facilitate error correction. Umlaut links code, model output, and tutorial-driven error messages in a single interface. They evaluated Umlaut in a study with 15 participants to determine its effectiveness in helping developers find and fix errors in their DL programs. Participants using Umlaut found and fixed significantly more bugs compared to a baseline condition.
To read the paper, please visit this document.