Ranking and performance of all 326 ranked bert-base-cased models (full table). The top 241 models were fully tested.
Notes:
- The baseline results can be found here
- While the average improvement is small, many datasets show large gains
model_name | avg | mnli_lp | 20_newsgroup | ag_news | amazon_reviews_multi | anli | boolq | cb | cola | copa | dbpedia | esnli | financial_phrasebank | imdb | isear | mnli | mrpc | multirc | poem_sentiment | qnli | qqp | rotten_tomatoes | rte | sst2 | sst_5bins | stsb | trec_coarse | trec_fine | tweet_ev_emoji | tweet_ev_emotion | tweet_ev_hate | tweet_ev_irony | tweet_ev_offensive | tweet_ev_sentiment | wic | wnli | wsc | yahoo_answers | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
baseline | bert-base-cased | 72.43 | nan | 81.74 | 89.06 | 65.71 | 46.57 | 68.27 | 63.48 | 81.85 | 52.15 | 78.77 | 89.64 | 68.36 | 91.15 | 68.39 | 83.39 | 82.93 | 60.47 | 67.69 | 90.00 | 89.95 | 84.55 | 62.64 | 91.49 | 51.41 | 84.52 | 96.63 | 72.98 | 44.24 | 78.84 | 52.78 | 65.20 | 84.25 | 68.23 | 64.78 | 52.32 | 61.92 | 71.03 |
1 | skim945/bert-finetuned-squad | 74.43 | 52.13 | 81.35 | 89.20 | 65.86 | 46.88 | 70.70 | 95.83 | 78.04 | 50.00 | 79.17 | 88.99 | 81.40 | 90.84 | 69.95 | 83.47 | 80.60 | 57.80 | 74.04 | 91.20 | 90.20 | 84.62 | 70.16 | 96.10 | 51.36 | 86.82 | 96.52 | 86.45 | 44.30 | 78.75 | 53.60 | 66.96 | 84.30 | 68.46 | 69.93 | 53.12 | 51.79 | 70.90 |
2 | ellabettison/finetuned_orgnames_bert | 74.26 | 52.01 | 82.81 | 89.07 | 65.76 | 46.75 | 70.03 | 69.64 | 82.84 | 55.00 | 79.70 | 89.88 | 83.40 | 91.44 | 69.95 | 83.13 | 85.29 | 59.84 | 79.81 | 89.38 | 90.29 | 85.27 | 64.26 | 92.20 | 51.18 | 85.69 | 97.00 | 78.60 | 44.92 | 80.72 | 54.21 | 66.45 | 85.47 | 69.20 | 63.17 | 56.34 | 63.46 | 71.10 |
3 | algoprivacy/bert-finetuned-squad | 74.10 | 55.66 | 82.32 | 89.37 | 83.60 | 46.25 | 71.22 | 71.43 | 81.88 | 55.00 | 78.63 | 46.25 | 44.11 | 91.20 | 70.53 | 91.10 | 85.29 | 61.96 | 75.00 | 67.15 | 90.25 | 85.46 | 54.93 | 66.10 | 92.55 | 85.91 | 97.20 | 81.20 | 79.59 | 53.77 | 68.11 | 84.19 | 66.70 | 83.19 | 62.54 | 89.61 | 63.46 | 70.50 |
4 | momtaz/bert-finetuned-squad | 74.08 | 54.84 | 81.56 | 89.37 | 65.72 | 48.22 | 72.78 | 73.21 | 82.65 | 53.00 | 78.87 | 89.48 | 81.30 | 91.29 | 69.62 | 82.92 | 85.54 | 60.95 | 68.27 | 90.59 | 90.15 | 84.33 | 64.98 | 92.09 | 52.22 | 86.23 | 96.80 | 80.00 | 44.98 | 79.03 | 54.95 | 66.96 | 83.84 | 69.11 | 65.36 | 56.34 | 63.46 | 70.87 |
5 | Dylan1999/bert-finetuned-squad-accelerate | 74.07 | 56.27 | 81.70 | 89.13 | 66.04 | 46.94 | 71.04 | 75.00 | 80.06 | 55.00 | 79.57 | 89.63 | 80.00 | 91.04 | 69.82 | 83.27 | 86.27 | 59.28 | 73.08 | 91.01 | 88.91 | 84.80 | 67.87 | 91.97 | 50.05 | 86.03 | 96.40 | 82.60 | 44.21 | 79.38 | 54.34 | 68.24 | 84.19 | 66.78 | 63.48 | 54.93 | 63.46 | 70.93 |
6 | jfarmerphd/bert-finetuned-squad-accelerate | 74.05 | 54.32 | 81.09 | 88.73 | 65.84 | 47.41 | 71.44 | 71.43 | 81.59 | 50.00 | 77.63 | 89.51 | 82.60 | 91.01 | 69.43 | 83.06 | 88.24 | 60.58 | 73.08 | 90.98 | 89.85 | 84.90 | 69.31 | 91.86 | 51.45 | 85.56 | 97.00 | 80.40 | 44.02 | 77.76 | 53.40 | 67.86 | 85.23 | 68.50 | 65.36 | 56.34 | 63.46 | 69.83 |
7 | Moussab/deepset_bert-base-cased-squad2-orkg-unchanged-5e-05 | 74.04 | 54.75 | 78.66 | 88.70 | 65.12 | 47.84 | 71.59 | 75.00 | 81.11 | 55.00 | 79.60 | 89.52 | 82.20 | 90.97 | 69.62 | 83.24 | 83.09 | 60.97 | 76.92 | 91.10 | 89.88 | 84.15 | 67.51 | 89.91 | 50.36 | 85.25 | 96.20 | 79.20 | 44.06 | 79.24 | 51.21 | 71.43 | 84.65 | 68.15 | 63.95 | 56.34 | 63.46 | 70.40 |
8 | relevanthint/bert-finetuned-ner | 74.04 | 49.62 | 82.43 | 89.23 | 66.00 | 47.16 | 69.36 | 66.07 | 83.03 | 58.00 | 79.23 | 89.59 | 81.50 | 91.07 | 70.60 | 83.39 | 85.54 | 61.70 | 81.73 | 90.33 | 89.88 | 83.96 | 63.54 | 91.40 | 50.59 | 85.08 | 97.00 | 79.20 | 44.52 | 80.30 | 52.49 | 67.22 | 84.30 | 68.60 | 64.11 | 52.11 | 63.46 | 71.57 |
9 | Hudayday/bert-finetuned-squad | 73.99 | 55.81 | 81.25 | 88.70 | 66.12 | 47.22 | 73.25 | 70.83 | 75.82 | 57.50 | 79.23 | 90.15 | 77.80 | 91.00 | 68.90 | 82.27 | 83.06 | 61.20 | 73.08 | 92.20 | 87.70 | 84.05 | 67.74 | 96.20 | 50.72 | 86.25 | 96.15 | 86.63 | 43.75 | 79.66 | 54.01 | 65.69 | 84.77 | 68.17 | 72.14 | 57.81 | 51.79 | 70.97 |
10 | chiranthans23/bert-base-cased | 73.98 | 54.98 | 81.43 | 89.13 | 65.60 | 47.44 | 71.77 | 69.64 | 80.92 | 60.00 | 78.90 | 89.60 | 75.30 | 90.99 | 70.53 | 83.44 | 84.56 | 60.48 | 79.81 | 91.31 | 89.92 | 84.90 | 65.70 | 91.63 | 50.90 | 84.98 | 97.20 | 78.80 | 44.39 | 79.52 | 53.54 | 66.07 | 84.30 | 68.59 | 65.05 | 53.52 | 63.46 | 70.10 |
Download full models ranking table: csv