Sharpness-Aware Minimization, or SAM, is a procedure that improves model generalization by simultaneously minimizing loss value and loss sharpness. SAM functions by seeking parameters that lie in neighborhoods having uniformly low loss value (rather than parameters that only themselves have low loss value).
Source: Sharpness-Aware Minimization for Efficiently Improving GeneralizationPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Image Classification | 13 | 18.06% |
Domain Generalization | 4 | 5.56% |
Adversarial Robustness | 3 | 4.17% |
Language Modelling | 2 | 2.78% |
Federated Learning | 2 | 2.78% |
Node Classification | 2 | 2.78% |
Long-tail Learning | 2 | 2.78% |
Model Compression | 2 | 2.78% |
Quantization | 2 | 2.78% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |