| Benchmark | Model | Rank | Results |
|---|---|---|---|
| image-classification-on-imagenet | Mega | #525 | Top 1 Accuracy: 82.4%Number of params: 90M |
| language-modelling-on-wikitext-103 | Mega | #30 | Test perplexity: 18.07Number of params: 252M |
| machine-translation-on-wmt2014-english-german | Mega | #30 | BLEU score: 29.01SacreBLEU: 27.96Number of Params: 67M |
| machine-translation-on-wmt2014-german-english | Mega | #4 | BLEU score: 33.12 |