Commonsense Reasoning on HellaSwag (ACC)
34.7AccuracyADAMW
Evaluation Results
| Method | Links | |
|---|---|---|
| ADAMWSynchronization=DENSE all-reduce, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 34.7 | |
| ADAMSSynchronization=DENSE all-reduce, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 34.57 | |
| SCAPESparsity (d)=0.1, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 34.38 | |
| SCAPESparsity (d)=0.01, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 33.93 |