| abstractive-text-summarization-on-cnn-daily | Transformer | #39 | ROUGE-1: 39.50ROUGE-2: 16.06ROUGE-L: 36.63 |
| constituency-parsing-on-penn-treebank | Transformer | #22 | F1 score: 92.7 |
| coreference-resolution-on-winograd-schema | Subword-level Transformer LM | #56 | Accuracy: 54.1 |
| image-guided-story-ending-generation-on-lsmdc | Transformer | #2 | BLEU-1: 15.35BLEU-2: 4.49BLEU-3: 1.82BLEU-4: 0.76CIDEr: 9.32… |
| image-guided-story-ending-generation-on-vist | Transformer | #3 | BLEU-1: 17.18BLEU-2: 6.29BLEU-3: 3.07BLEU-4: 2.01CIDEr: 12.75… |
| machine-translation-on-iwslt2014-german | Transformer | #24 | BLEU score: 34.44 |
| machine-translation-on-iwslt2015-english | Transformer | #2 | BLEU score: 28.50 |
| machine-translation-on-wmt2014-english-french | Transformer Big | #26 | BLEU score: 41.0Hardware Burden: 23G… |
| machine-translation-on-wmt2014-english-french | Transformer Base | #38 | BLEU score: 38.1Hardware Burden: 23G… |
| machine-translation-on-wmt2014-english-german | Transformer Big | #37 | BLEU score: 28.4Hardware Burden: 871G… |
| machine-translation-on-wmt2014-english-german | Transformer Base | #43 | BLEU score: 27.3Operations per network pass: 330000000.0G |
| multimodal-machine-translation-on-multi30k | Transformer | #8 | BLUE (DE-EN): 29.0 |
| natural-language-understanding-on-pdp60 | Subword-level Transformer LM | #6 | Accuracy: 58.3 |
| supervised-only-3d-point-cloud-classification | Transformer | #11 | Overall Accuracy (PB_T50_RS): 77.24GFLOPs: 4.8… |
| text-summarization-on-gigaword | Transformer | #20 | ROUGE-1: 37.57ROUGE-2: 18.90ROUGE-L: 34.69 |