Summary of A Two-phase Recall-and-select Framework For Fast Model Selection, by Jianwei Cui et al.
A Two-Phase Recall-and-Select Framework for Fast Model Selection
by Jianwei Cui, Wenhang Shi, Honglin Tao, Wei Lu, Xiaoyong Du
First submitted to arxiv on: 28 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a two-phase model selection framework for efficiently choosing a robust deep learning model from public repositories. The framework, comprising coarse-recall and fine-selection phases, leverages models’ training performances on benchmark datasets to recall a smaller subset of candidate models. A light-weight proxy score is computed between the model cluster and target dataset in the coarse-recall phase, while the fine-selection phase involves fine-tuning the recalled models on the target dataset with successive halving. The proposed methodology is demonstrated to facilitate selecting high-performing models at a rate about 3x times faster than conventional baseline methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us find the best deep learning model for a task by looking at how well models performed in the past. It’s like searching through a big library of books (models) and finding the ones that are most relevant to what you’re trying to do (your target dataset). The authors suggest a two-step process: first, they group similar models together based on their performance on some test datasets. Then, they pick one model from this group by fine-tuning it on your actual task. This way, you can find the best model quickly and efficiently. |
Keywords
» Artificial intelligence » Deep learning » Fine tuning » Recall
