task_path stringlengths 3 199 ⌀ | dataset stringlengths 1 128 ⌀ | model_name stringlengths 1 223 ⌀ | paper_url stringlengths 21 601 ⌀ | metric_name stringlengths 1 50 ⌀ | metric_value stringlengths 1 9.22k ⌀ |
|---|---|---|---|---|---|
Machine Translation | WMT2016 German-English | Unsupervised S2S with attention | http://arxiv.org/abs/1711.00043v2 | BLEU score | 13.33 |
Machine Translation | WMT2016 German-English | Exploiting Mono at Scale (single) | https://aclanthology.org/D19-1430 | SacreBLEU | 47.5 |
Machine Translation | WMT2015 English-Russian | C2-50k Segmentation | http://arxiv.org/abs/1508.07909v5 | BLEU score | 20.9 |
Machine Translation | 20NEWS | tensorflow/tensor2tensor | http://arxiv.org/abs/1709.07809v1 | 1-of-100 Accuracy | 5 |
Machine Translation | 20NEWS | 12 | http://arxiv.org/abs/1508.04025v5 | Accuracy | 1.0 |
Machine Translation | flores95-devtest X-eng | SeamlessM4T Large | https://arxiv.org/abs/2308.11596v3 | ChrF++ | 60.8 |
Machine Translation | flores95-devtest X-eng | SeamlessM4T-NLLB-1.3B | https://arxiv.org/abs/2308.11596v3 | ChrF++ | 60.7 |
Machine Translation | flores95-devtest X-eng | SeamlessM4T Medium | https://arxiv.org/abs/2308.11596v3 | ChrF++ | 55.4 |
Machine Translation | IWSLT2017 German-English | Adaptively Sparse Transformer (alpha-entmax) | https://arxiv.org/abs/1909.00015v2 | BLEU score | 29.9 |
Machine Translation | IWSLT2017 German-English | Adaptively Sparse Transformer (1.5-entmax) | https://arxiv.org/abs/1909.00015v2 | BLEU score | 29.83 |
Machine Translation | FRMT (Chinese - Mainland) | PaLM 2 | https://arxiv.org/abs/2305.10403v3 | BLEURT | 74.4 |
Machine Translation | FRMT (Chinese - Mainland) | Google Translate | https://arxiv.org/abs/2305.10403v3 | BLEURT | 72.3 |
Machine Translation | FRMT (Chinese - Mainland) | PaLM | https://arxiv.org/abs/2305.10403v3 | BLEURT | 70.3 |
Machine Translation | IWSLT 2017 | GPT-4o (HPT) | https://arxiv.org/abs/2406.12644v4 | BLEU score | 32 |
Machine Translation | Tatoeba (EL-to-EN) | PENELOPIE (Transformers-based Greek-to-English NMT) | https://arxiv.org/abs/2103.15075v1 | BLEU | 79.3 |
Machine Translation | WMT 2022 Japanese-English | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 25.6 |
Machine Translation | Multi Lingual Bug Reports | ChatGPT | https://arxiv.org/abs/2502.14338v4 | BERTScore | 79 |
Machine Translation | IWSLT2017 Arabic-English | Transformer base + BPE-Dropout | https://arxiv.org/abs/1910.13267v2 | Cased sacreBLEU | 33.0 |
Machine Translation | IWSLT2017 Arabic-English | NLLB-200 | https://arxiv.org/abs/2207.04672v3 | SacreBLEU | 44.7 |
Machine Translation | IWSLT2015 Thai-English | Seq-KD + Seq-Inter + Word-KD | http://arxiv.org/abs/1606.07947v4 | BLEU score | 14.2 |
Machine Translation | ACCURAT balanced test corpus for under resourced languages Estonian-Russian | Multilingual Transformer | https://aclanthology.org/L18-1595 | BLEU | 19.18 |
Machine Translation | Itihasa | Baseline (en->sn) | https://arxiv.org/abs/2106.03269v3 | SacreBLEU | 7.59 |
Machine Translation | Itihasa | Baseline (sn->en) | https://arxiv.org/abs/2106.03269v3 | SacreBLEU | 7.49 |
Machine Translation | slone/myv_ru_2022 myv-ru | slone/mbart-large-51-myv-mul-v1 | https://arxiv.org/abs/2209.09368v1 | ChrF++ | 38.63 |
Machine Translation | WMT 2018 Estonian-English | Multi-pass backtranslated adapted transformer | https://aclanthology.org/W18-6423 | BLEU | 29.00 |
Machine Translation | WMT 2022 English-Chinese | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 49.7 |
Machine Translation | WMT2019 German-English | Exploiting Mono at Scale (single) | https://aclanthology.org/D19-1430 | SacreBLEU | 41.9 |
Machine Translation | IWSLT2015 Chinese-English | BP-Transformer | https://arxiv.org/abs/1911.04070v1 | BLEU | 19.84 |
Machine Translation | WMT 2022 German-English | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 33.7 |
Machine Translation | WMT 2022 English-Japanese | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 41.5 |
Machine Translation | WMT 2022 English-Russian | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 32.7 |
Machine Translation | WMT2016 Finnish-English | CT+B/S construction | https://arxiv.org/abs/1907.00494v1 | BLEU | 32.4 |
Machine Translation | WMT2017 Finnish-English | CT+B/S construction | https://arxiv.org/abs/1907.00494v1 | BLEU | 35.5 |
Machine Translation | Tatoeba (EN-to-EL) | PENELOPIE Transformers-based NMT (EN2EL) | https://arxiv.org/abs/2103.15075v1 | BLEU | 76.9 |
Machine Translation | WMT2016 Russian-English | Attentional encoder-decoder + BPE | http://arxiv.org/abs/1606.02891v2 | BLEU score | 28.0 |
Machine Translation | WMT 2018 English-Finnish | Transformer trained on highly filtered data | http://arxiv.org/abs/1810.08392v1 | BLEU | 17.40 |
Machine Translation | IWSLT2014 English-German | PiNMT | https://arxiv.org/abs/2310.19680v4 | BLEU score | 32.2 |
Machine Translation | IWSLT2014 English-German | Bi-SimCut | https://arxiv.org/abs/2206.02368v2 | BLEU score | 31.16 |
Machine Translation | IWSLT2014 English-German | SimCut | https://arxiv.org/abs/2206.02368v2 | BLEU score | 30.98 |
Machine Translation | IWSLT2014 English-German | DRDA | https://arxiv.org/abs/2406.02517v1 | BLEU score | 30.92 |
Machine Translation | IWSLT2014 English-German | Unidrop | https://arxiv.org/abs/2104.04946v1 | BLEU score | 29.99 |
Machine Translation | IWSLT2014 English-German | MixedRepresentations | https://icml.cc/Conferences/2020/ScheduleMultitrack?event=6391 | BLEU score | 29.93 |
Machine Translation | flores95-devtest eng-X | SeamlessM4T Large | https://arxiv.org/abs/2308.11596v3 | ChrF++ | 50.9 |
Machine Translation | flores95-devtest eng-X | SeamlessM4T-NLLB-1.3B | https://arxiv.org/abs/2308.11596v3 | ChrF++ | 49.6 |
Machine Translation | flores95-devtest eng-X | SeamlessM4T Medium | https://arxiv.org/abs/2308.11596v3 | ChrF++ | 48.4 |
Machine Translation | OpenSubtitles | Fine tuned MarianMT | https://arxiv.org/abs/2501.01629v1 | BLEU score | 27 |
Machine Translation | OpenSubtitles | Fine tuned MarianMT | https://arxiv.org/abs/2501.01629v1 | METEOR | 61 |
Machine Translation | FRMT (Portuguese - Portugal) | PaLM 2 | https://arxiv.org/abs/2305.10403v3 | BLEURT | 78.3 |
Machine Translation | FRMT (Portuguese - Portugal) | PaLM | https://arxiv.org/abs/2305.10403v3 | BLEURT | 76.1 |
Machine Translation | FRMT (Portuguese - Portugal) | Google Translate | https://arxiv.org/abs/2305.10403v3 | BLEURT | 75.3 |
Machine Translation | WMT 2022 English-Czech | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 41.4 |
Machine Translation | WMT2016 Czech-English | Attentional encoder-decoder + BPE | http://arxiv.org/abs/1606.02891v2 | BLEU score | 31.4 |
Machine Translation | WMT2019 English-German | Facebook FAIR (ensemble) | https://arxiv.org/abs/1907.06616v1 | BLEU score | 43.1 |
Machine Translation | WMT2019 English-German | Facebook FAIR (ensemble) | https://arxiv.org/abs/1907.06616v1 | SacreBLEU | 42.7 |
Machine Translation | WMT2019 English-German | Exploiting Mono at Scale (single) | https://aclanthology.org/D19-1430 | SacreBLEU | 43.8 |
Machine Translation | Business Scene Dialogue EN-JA | Transformer-base | https://arxiv.org/abs/2008.01940v1 | BLEU | 13.53 |
Machine Translation | FRMT (Portuguese - Brazil) | PaLM 2 | https://arxiv.org/abs/2305.10403v3 | BLEURT | 81.1 |
Machine Translation | FRMT (Portuguese - Brazil) | Google Translate | https://arxiv.org/abs/2305.10403v3 | BLEURT | 80.2 |
Machine Translation | FRMT (Portuguese - Brazil) | PaLM | https://arxiv.org/abs/2305.10403v3 | BLEURT | 78.5 |
Machine Translation | WMT2017 English-Finnish | OmniNetP | https://arxiv.org/abs/2103.01075v1 | BLEU | 20.9 |
Machine Translation | WMT2016 English-French | DeLighT | https://arxiv.org/abs/2008.00623v2 | BLEU score | 40.5 |
Machine Translation | WMT 2022 Russian-English | Vega-MT | https://arxiv.org/abs/2209.09444v4 | SacreBLEU | 45.1 |
Machine Translation | IWSLT2017 English-Arabic | Transformer base + BPE-Dropout | https://arxiv.org/abs/1910.13267v2 | Cased sacreBLEU | 15.2 |
Machine Translation | IWSLT2017 English-Arabic | NLLB-200 | https://arxiv.org/abs/2207.04672v3 | SacreBLEU | 25.2 |
Machine Translation | WMT 2018 Finnish-English | CT+B/S construction | https://arxiv.org/abs/1907.00494v1 | BLEU | 26.5 |
Machine Translation | WMT 2018 Finnish-English | Transformer trained on highly filtered data | http://arxiv.org/abs/1810.08392v1 | BLEU | 24.00 |
Machine Translation | WMT2014 German-English | Bi-SimCut | https://arxiv.org/abs/2206.02368v2 | BLEU score | 35.15 |
Machine Translation | WMT2014 German-English | BiBERT | https://arxiv.org/abs/2109.04588v1 | BLEU score | 34.94 |
Machine Translation | WMT2014 German-English | SimCut | https://arxiv.org/abs/2206.02368v2 | BLEU score | 34.86 |
Machine Translation | WMT2014 German-English | Mega | https://arxiv.org/abs/2209.10655v3 | BLEU score | 33.12 |
Machine Translation | WMT2014 German-English | CMLM+LAT+4 iterations | https://arxiv.org/abs/2011.06132v1 | BLEU score | 32.04 |
Machine Translation | WMT2014 German-English | MAT+Knee | https://arxiv.org/abs/2003.03977v5 | BLEU score | 31.9 |
Machine Translation | WMT2014 German-English | CNAT | https://arxiv.org/abs/2103.11405v1 | BLEU score | 30.75 |
Machine Translation | WMT2014 German-English | CMLM+LAT+1 iterations | https://arxiv.org/abs/2011.06132v1 | BLEU score | 29.91 |
Machine Translation | WMT2014 German-English | FlowSeq-large (NPD n = 30) | https://arxiv.org/abs/1909.02480v3 | BLEU score | 28.29 |
Machine Translation | WMT2014 German-English | FlowSeq-large (NPD n = 15) | https://arxiv.org/abs/1909.02480v3 | BLEU score | 27.71 |
Machine Translation | WMT2014 German-English | FlowSeq-large (IWD n=15) | https://arxiv.org/abs/1909.02480v3 | BLEU score | 27.16 |
Machine Translation | WMT2014 German-English | Denoising autoencoders (non-autoregressive) | http://arxiv.org/abs/1802.06901v3 | BLEU score | 25.43 |
Machine Translation | WMT2014 German-English | FlowSeq-large | https://arxiv.org/abs/1909.02480v3 | BLEU score | 25.4 |
Machine Translation | WMT2014 German-English | FlowSeq-base | https://arxiv.org/abs/1909.02480v3 | BLEU score | 23.36 |
Machine Translation | WMT2014 German-English | NAT +FT + NPD | http://arxiv.org/abs/1711.02281v2 | BLEU score | 23.20 |
Machine Translation | WMT2014 German-English | SMT + iterative backtranslation (unsupervised) | http://arxiv.org/abs/1809.01272v1 | BLEU score | 17.43 |
Machine Translation | WMT2014 English-German | Transformer Cycle (Rev) | https://arxiv.org/abs/2104.06022v4 | BLEU score | 35.14 |
Machine Translation | WMT2014 English-German | Transformer Cycle (Rev) | https://arxiv.org/abs/2104.06022v4 | SacreBLEU | 33.54 |
Machine Translation | WMT2014 English-German | Noisy back-translation | http://arxiv.org/abs/1808.09381v2 | BLEU score | 35.0 |
Machine Translation | WMT2014 English-German | Noisy back-translation | http://arxiv.org/abs/1808.09381v2 | SacreBLEU | 33.8 |
Machine Translation | WMT2014 English-German | Noisy back-translation | http://arxiv.org/abs/1808.09381v2 | Hardware Burden | 146G |
Machine Translation | WMT2014 English-German | Noisy back-translation | http://arxiv.org/abs/1808.09381v2 | Operations per network pass | null |
Machine Translation | WMT2014 English-German | Transformer+Rep(Uni) | https://arxiv.org/abs/2104.01853v1 | BLEU score | 33.89 |
Machine Translation | WMT2014 English-German | Transformer+Rep(Uni) | https://arxiv.org/abs/2104.01853v1 | SacreBLEU | 32.35 |
Machine Translation | WMT2014 English-German | Transformer+Rep(Uni) | https://arxiv.org/abs/2104.01853v1 | Hardware Burden | null |
Machine Translation | WMT2014 English-German | Transformer+Rep(Uni) | https://arxiv.org/abs/2104.01853v1 | Operations per network pass | null |
Machine Translation | WMT2014 English-German | T5-11B | https://arxiv.org/abs/1910.10683v4 | BLEU score | 32.1 |
Machine Translation | WMT2014 English-German | T5-11B | https://arxiv.org/abs/1910.10683v4 | Number of Params | 11110M |
Machine Translation | WMT2014 English-German | BiBERT | https://arxiv.org/abs/2109.04588v1 | BLEU score | 31.26 |
Machine Translation | WMT2014 English-German | Transformer + R-Drop | https://arxiv.org/abs/2106.14448v2 | BLEU score | 30.91 |
Machine Translation | WMT2014 English-German | Transformer + R-Drop | https://arxiv.org/abs/2106.14448v2 | Hardware Burden | 49G |
Machine Translation | WMT2014 English-German | Transformer + R-Drop | https://arxiv.org/abs/2106.14448v2 | Operations per network pass | null |
Machine Translation | WMT2014 English-German | Bi-SimCut | https://arxiv.org/abs/2206.02368v2 | BLEU score | 30.78 |
Machine Translation | WMT2014 English-German | BERT-fused NMT | https://arxiv.org/abs/2002.06823v1 | BLEU score | 30.75 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.