Closed nabenabe0928 closed 1 year ago
NASOA: Towards Faster Task-oriented Online Fine-tuning with a Zoo of Models
This was not really about multi-fidelity optimization. They use an MLP predictor with learnable layers. The input is the architecture and dataset statistics.
NASOA: Towards Faster Task-oriented Online Fine-tuning with a Zoo of Models
This was not really about multi-fidelity optimization. They use an MLP predictor with learnable layers. The input is the architecture and dataset statistics.