PolEmo 2.0

Polish four-class review sentiment from PolEmo 2.0, scored in-domain and out-of-domain as two lm-evaluation-harness tasks.

Also known as: PolEmo2, PolEmo 2.0, klej-polemo2, polemo2_in, polemo2_out

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categorydomain
Subcategoryfour-class Polish consumer-review sentiment, in-domain and out-of-domain
Page statusunknown
Metricmicro-F1 and accuracy (Hugging Face evaluate); KLEJ cards also quote accuracy
Directionhigher_is_better
Dataset size1216
Dataset licenceCC-BY-NC-SA-4.0
PublisherWrocław University of Science and Technology and University of Wrocław; KLEJ packaging by Allegro

What it measures

PolEmo 2.0 is a Polish consumer-review corpus. The model reads a review and must label it neutral, negative, positive, or ambiguous. KLEJ splits the same hotel and medicine training reviews into an in-domain test (hotels and medicine) and an out-of-domain test (products and university). lm-evaluation-harness follows that pair as polemo2_in and polemo2_out. It is review-level four-way sentiment, not sentence-level tagging, and not English IMDb-style polarity.

Task format

lm-eval generate_until. Polish prompt "Opinia:" plus four lettered options (A Neutralny, B Negatywny, C Pozytywny, D Niejednoznaczny). Decoding stops at "." or ",", temperature 0, max 50 tokens. A regex keeps the first A–D letter. Gold labels are __label__meta_zero, __label__meta_minus_m, __label__meta_plus_m, __label__meta_amb. Tag polemo2 runs both tasks.

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub