LiveTune: Dynamic Parameter Tuning for Feedback-Driven Optimization

Method and contribution

LiveVariables expose optimization parameters through designated system ports for real-time adjustment without restarting the optimization loop.

An earlier system for changing running optimization processes; compare its parameter mechanism with Interactive Training’s human/agent command interface and v2 protocol.

Read in context

Controlling a running training job with humans and AI agents →

Citation

Soheil Zibakhsh Shabgahi, Nojan Sheybani, Aiden Tabrizi, Farinaz Koushanfar. LiveTune: Dynamic Parameter Tuning for Feedback-Driven Optimization. arXiv.

@misc{shabgahi2023livetune,
  title = {{LiveTune: Dynamic Parameter Tuning for Feedback-Driven Optimization}},
  author = {Shabgahi, Soheil Zibakhsh and Sheybani, Nojan and Tabrizi, Aiden and Koushanfar, Farinaz},
  year = {2023},
  url = {https://arxiv.org/abs/2311.17279},
  eprint = {2311.17279},
  archivePrefix = {arXiv},
  doi = {10.48550/arXiv.2311.17279}
}