Diachronic equivalents
For a given Polish word, as used in a given year, give a diachronic equivalent (a.k.a. temporal word analogy) for a given year. [ver. 1.0.0]
This is a long list of all submissions, if you want to see only the best, click leaderboard.
| # | submitter | when | ver. | description | dev-0 MAP | test-A MAP | |
|---|---|---|---|---|---|---|---|
| 7 | zrostek | 2023-05-29 08:58 | 1.0.0 | word2vec (with close matches for words with no embedding) word2vec | 0.0467 | 0.0543 | |
| 13 | zrostek | 2023-05-26 12:30 | 1.0.0 | word2vec word2vec | 0.0419 | 0.0465 | |
| 16 | zrostek | 2023-05-25 06:20 | 1.0.0 | nanoT5 (adafactor, legacy), 70000 out of 364210 epochs=20 t5 | 0.3374 | 0.0320 | |
| 15 | zrostek | 2023-05-25 06:15 | 1.0.0 | nanoT5 (adafactor, legacy), 70000 out of 364210 epochs=10 t5 | 0.1181 | 0.0367 | |
| 1 | zrostek | 2023-05-18 17:56 | 1.0.0 | gpt-3.5 | 0.0000 | 0.0923 | |
| 5 | zrostek | 2023-05-18 13:34 | 1.0.0 | allegro plT5-large with wikipedia + wiktionary data t5 | 0.2646 | 0.0670 | |
| 4 | zrostek | 2023-05-17 06:33 | 1.0.0 | allegro plT5-large with wikipedia data t5 | 0.2188 | 0.0680 | |
| 6 | zrostek | 2023-05-16 13:11 | 1.0.0 | allegro plT5-large with wikipedia data t5 | 0.2118 | 0.0606 | |
| 3 | zrostek | 2023-05-16 10:31 | 1.0.0 | allegro plT5-large t5 | 0.2054 | 0.0680 | |
| 18 | zrostek | 2023-05-16 08:47 | 1.0.0 | allegro plT5 t5 | 0.0596 | 0.0310 | |
| 2 | p/tlen | 2018-01-31 21:48 | 1.0.0 | with Word2Vec, method=most-similar, model=normalized-lemmatized-with-years word2vec | 0.0678 | 0.0904 | |
| 14 | p/tlen | 2018-01-30 20:35 | 1.0.0 | with Word2Vec, method=year-analogy, model=normalized-lemmatized-with-years word2vec | 0.0281 | 0.0427 | |
| 20 | p/tlen | 2018-01-26 19:04 | 1.0.0 | with Word2Vec, method=year-analogy, model=every-n-5 word2vec | 0.0128 | 0.0204 | |
| 9 | p/tlen | 2018-01-26 16:22 | 1.0.0 | with Word2Vec, method=most-similar, model=every-n-5 | 0.0434 | 0.0517 | |
| 8 | p/tlen | 2018-01-25 19:59 | 1.0.0 | with Word2Vec, method=most-similar, model=normalized-with-years word2vec | 0.0488 | 0.0541 | |
| 19 | p/tlen | 2018-01-25 17:16 | 1.0.0 | with Word2Vec, method=year-analogy, model=normalized-with-years word2vec | 0.0295 | 0.0306 | |
| 17 | p/tlen | 2018-01-25 16:12 | 1.0.0 | with Word2Vec, method=year-analogy, model=with-years | 0.0254 | 0.0310 | |
| 10 | p/tlen | 2018-01-25 15:17 | 1.0.0 | with Word2Vec, method=most-similar, model=with-years word2vec | 0.0447 | 0.0516 | |
| 11 | p/tlen | 2018-01-25 13:44 | 1.0.0 | with Word2Vec, method=most-similar, model=raw word2vec | 0.0467 | 0.0499 | |
| 12 | p/tlen | 2018-01-25 10:34 | 1.0.0 | with Word2Vec, method=year-analogy, model=raw word2vec | 0.0432 | 0.0492 |