Ranking and performance of all 278 ranked t5-base models (full table). The top 84 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 | t5-base | 75.45 | nan | 85.12 | 89.42 | 66.54 | 47.05 | 76.66 | 75.54 | 81.91 | 49.65 | 76.41 | 89.72 | 85.30 | 92.33 | 71.28 | 83.80 | 85.66 | 60.28 | 74.42 | 90.38 | 88.94 | 88.61 | 73.68 | 93.84 | 55.55 | 85.31 | 97.21 | 92.33 | 44.88 | 79.51 | 52.74 | 73.74 | 84.03 | 70.21 | 67.19 | 55.35 | 60.00 | 71.59 |
1 | adit94/nlpcharade | 78.23 | 82.76 | 56.11 | 91.80 | 70.95 | 48.62 | 87.50 | 66.61 | 79.29 | 89.47 | 89.21 | 90.32 | 86.62 | 81.49 | 97.60 | 92.44 | 88.73 | 72.36 | 45.38 | 56.34 | 90.68 | 51.89 | 90.32 | 83.95 | 74.23 | 79.33 | 66.44 | 92.32 | 92.44 | 90.32 | 74.23 | 83.95 | 70.95 | 86.62 | 71.80 | 56.34 | 77.17 | 92.60 |
2 | zeineb/LearningQ-t5-Answer-agnostic-QG | 78.02 | 82.78 | 92.83 | 72.03 | 92.46 | 48.28 | 85.71 | 59.01 | 87.01 | 68.65 | 86.54 | 90.55 | 74.01 | 66.64 | 76.07 | 86.68 | 90.81 | 58.65 | 93.69 | 92.46 | 87.88 | 74.49 | 74.01 | 86.68 | 85.00 | 77.95 | 55.48 | 89.12 | 90.55 | 81.49 | 51.48 | 74.49 | 85.00 | 70.42 | 89.60 | 54.93 | 72.49 | 85.53 |
3 | hadifar/tqa_qg_t5 | 77.68 | 82.36 | 92.20 | 71.27 | 85.60 | 48.53 | 97.80 | 88.93 | 83.03 | 55.88 | 85.83 | 48.53 | 46.01 | 88.56 | 92.62 | 92.31 | 69.75 | 94.15 | 92.82 | 79.42 | 89.43 | 56.33 | 49.30 | 66.48 | 93.69 | 92.00 | 82.69 | 98.00 | 82.05 | 52.56 | 75.89 | 84.07 | 70.64 | 85.79 | 51.95 | 90.71 | 83.60 | 78.00 |
4 | yacine-djm/t5-ALL-1-Epoch | 77.24 | 82.24 | 85.30 | 89.63 | 66.86 | 50.50 | 79.66 | 89.29 | 82.07 | 55.00 | 77.33 | 90.22 | 85.30 | 92.39 | 70.60 | 86.63 | 88.73 | 60.91 | 82.69 | 92.35 | 90.90 | 89.02 | 79.78 | 94.15 | 54.39 | 88.51 | 97.60 | 91.60 | 45.03 | 81.42 | 52.90 | 74.87 | 84.30 | 69.72 | 70.85 | 54.93 | 63.46 | 71.73 |
5 | Zekunli/t5-base-extraction-cnndm_fs0.01-h-ppo | 77.23 | 82.87 | 92.80 | 71.77 | 45.54 | 47.84 | 79.60 | 87.50 | 82.26 | 54.00 | 85.08 | 90.71 | 81.91 | 88.09 | 72.30 | 90.71 | 86.52 | 61.18 | 92.70 | 47.84 | 90.99 | 54.71 | 77.26 | 86.10 | 93.23 | 88.71 | 87.50 | 98.20 | 53.80 | 77.68 | 84.19 | 69.92 | 92.62 | 77.26 | 67.24 | 56.34 | 89.43 | 76.63 |
6 | ammarpl/t5-base-finetuned-elif-attempt1 | 77.07 | 82.15 | 85.34 | 89.20 | 66.28 | 48.50 | 78.65 | 92.86 | 81.88 | 55.00 | 76.93 | 90.14 | 85.40 | 92.65 | 72.75 | 87.10 | 86.27 | 61.10 | 79.81 | 92.93 | 90.72 | 87.34 | 74.01 | 93.12 | 55.02 | 87.75 | 97.80 | 92.60 | 45.96 | 81.91 | 53.77 | 77.93 | 83.37 | 70.35 | 69.75 | 56.34 | 63.46 | 70.63 |
7 | Zekunli/t5-base-extraction-cnndm_fs0.2-c | 77.02 | 82.82 | 85.81 | 89.37 | 66.76 | 48.41 | 78.99 | 85.71 | 81.21 | 55.00 | 77.67 | 90.37 | 86.20 | 92.31 | 71.19 | 86.23 | 88.97 | 61.16 | 85.58 | 92.81 | 90.78 | 88.37 | 77.98 | 93.23 | 55.11 | 88.76 | 97.40 | 92.20 | 45.61 | 82.34 | 51.82 | 75.13 | 85.47 | 70.50 | 69.12 | 56.34 | 56.73 | 72.13 |
8 | tzytzytzy/t5_4248 | 77.01 | 82.90 | 98.00 | 72.29 | 46.20 | 48.47 | 79.33 | 87.50 | 81.78 | 61.43 | 86.01 | 90.53 | 81.42 | 56.02 | 71.70 | 77.26 | 89.95 | 68.65 | 88.56 | 56.34 | 90.67 | 93.69 | 90.53 | 85.60 | 66.82 | 88.38 | 92.52 | 84.62 | 52.49 | 74.11 | 83.14 | 70.77 | 87.24 | 92.64 | 63.46 | 48.47 | 89.13 | 76.73 |
9 | AyanSau/results_T5_Base | 77.00 | 83.08 | 86.02 | 89.23 | 66.44 | 49.19 | 79.27 | 89.29 | 79.77 | 49.00 | 76.13 | 90.58 | 85.80 | 92.70 | 71.64 | 86.53 | 87.01 | 61.43 | 84.62 | 92.57 | 90.42 | 88.27 | 77.62 | 93.69 | 55.88 | 88.54 | 98.20 | 92.40 | 45.27 | 81.35 | 52.96 | 78.95 | 84.30 | 70.51 | 69.91 | 56.34 | 58.65 | 71.47 |
10 | Zekunli/t5-base-summarization-cnndm_fs0.01 | 76.95 | 82.81 | 86.46 | 89.63 | 66.48 | 48.97 | 79.17 | 87.50 | 82.36 | 49.00 | 76.53 | 90.41 | 83.50 | 92.58 | 71.12 | 86.60 | 85.78 | 60.13 | 85.58 | 92.90 | 90.82 | 87.90 | 76.53 | 93.92 | 55.29 | 87.98 | 97.40 | 92.80 | 45.38 | 80.79 | 54.81 | 77.04 | 83.60 | 70.86 | 68.34 | 56.34 | 63.46 | 72.07 |
Download full models ranking table: csv