/MF_NAS_KD

Multi-fidelity Neural Architecture Search with Knowledge Distillation

Primary LanguageJupyter Notebook

Multi-fidelity Neural Architecture Search with Knowledge Distillation

  1. Run NAS.ipynb to study the benchmark and NAS algorithms.
  2. mobilenetv2.py, ShuffleNetv2.py define search spaces.
  3. Training with various KD methods were done via https://github.com/HobbitLong/RepDistiller.