# Train Big, Then Compress: Rethinking Model Size for Efficient Training and Inference of Transformers

> Research article (International Conference on Machine Learning, 2020) · cited 111× · AI/ML

**Wikidata**: [openalex:W3034340181](https://www.wikidata.org/wiki/openalex:W3034340181)  
**Source**: https://4ort.xyz/entity/train-big-then-compress-rethinking-model-size-for-efficient-training-and-inference-of-transformers