Polish review sentiment on PolEmo 2.0: fine-tuned HerBERT 0.986 vs TF-IDF 0.944 macro-F1, with a paired McNemar test.
python nlp github-pages machine-learning sentiment-analysis text-classification scikit-learn transformers tf-idf poland polish bert herbert fastapi polemo
-
Updated
Sep 26, 2026 - Python