TensorFI 2: A fault injector for TensorFlow 2 applications
TensorFI 2 is a fault injector for TensorFlow 2 applications written in Python 3. We make use of the Keras Model APIs to inject static faults in the layer states and dynamic faults in the layer outputs. Similar to TensorFI 1, TensorFI 2 supports both hardware and software faults with different injection modes. The fault injector is easily configurable with YAML.
Z. Chen, N. Narayanan, B. Fang, G. Li, K. Pattabiraman and N. DeBardeleben, “TensorFI: A Flexible Fault Injection Framework for TensorFlow Applications,” 2020 IEEE 31st International Symposium on Software Reliability Engineering (ISSRE), 2020, pp. 426-435, doi: 10.1109/ISSRE5003.2020.00047. https://ieeexplore.ieee.org/abstract/document/9251063