Language resources

On this page you can browse and search our corpora and lexicons. Click on a resource name to see what files are available for download. You can go directly to the search interface by clicking on the Korp or Karp logo.

All (1169) Collections (28) Corpora (1095) Lexicons (59) Training and evaluation data (45) Models (15)

Resource	Type	Language	Access
Dependency parsing model: Stanza Pretrained models for dependency parsing.	Model	Swedish	Dataset: synt_stanza_eval.zip 2020-12-09 – 99.05 MB – CC BY 4.0 Dataset: synt_stanza_full2.zip 2020-12-09 – 99.17 MB – CC BY 4.0 Dataset: stanza_pretrain.zip 2020-12-09 – 105.77 MB – CC BY 4.0
Collection Kubord-fasttext A collection of fasttext models trained on modern newspaper texts from the National Library of Sweden	Model	Swedish	See 6 collected resources
Kubord-fasttext - Aftonbladet 2010–2022 - lemma Fasttext model trained on Aftonbladet 2010–2022	Model	Swedish	Dataset: kubord-fasttext-afb-2010-2022-lemma.zip 2024-06-11 – 3.03 GB – CC BY 4.0
Kubord-fasttext - Aftonbladet 2010–2022 - token Fasttext model trained on Aftonbladet 2010–2022	Model	Swedish	Dataset: kubord-fasttext-afb-2010-2022-token.zip 2024-06-11 – 3.18 GB – CC BY 4.0
Kubord-fasttext - Dagens Nyheter 2010–2022 - lemma Fasttext model trained on Dagens Nyheter 2010–2022	Model	Swedish	Dataset: kubord-fasttext-dn-2010-2022-lemma.zip 2024-06-11 – 2.93 GB – CC BY 4.0
Kubord-fasttext - Dagens Nyheter 2010–2022 - token Fasttext model trained on Dagens Nyheter 2010–2022	Model	Swedish	Dataset: kubord-fasttext-dn-2010-2022-token.zip 2024-06-11 – 3.1 GB – CC BY 4.0
Kubord-fasttext - Göteborgsposten 2013–2022 - lemma Fasttext model trained on Göteborgsposten 2013–2022	Model	Swedish	Dataset: kubord-fasttext-gp-2013-2022-lemma.zip 2024-06-11 – 2.71 GB – CC BY 4.0
Kubord-fasttext - Göteborgsposten 2013–2022 - token Fasttext model trained on Göteborgsposten 2013–2022	Model	Swedish	Dataset: kubord-fasttext-gp-2013-2022-token.zip 2024-06-11 – 2.84 GB – CC BY 4.0
Lemmatization model: Stanza Pretrained model for lemmatization.	Model	Swedish	Dataset: lem_stanza.zip 2020-11-19 – 3.74 MB – CC BY 4.0
POS-tagging model: Flair Pretrained models for POS-tagging.	Model	Swedish	Dataset: flair_eval.zip 2020-06-18 – 1.37 GB – CC BY 4.0 Dataset: flair_full.zip 2020-06-18 – 1.37 GB – CC BY 4.0
POS-tagging model: Marmot Pretrained models for POS-tagging.	Model	Swedish	Dataset: marmot_eval.marmot 2020-06-29 – 108.59 MB – CC BY 4.0 Dataset: marmot_full.marmot 2020-06-29 – 113.41 MB – CC BY 4.0 Dataset: saldo_marmot.txt 2020-06-29 – 46.33 MB – CC BY 4.0
POS-tagging model: Stanza Pretrained models for POS-tagging.	Model	Swedish	Dataset: morph_stanza_eval.zip 2020-12-09 – 19.94 MB – CC BY 4.0 Dataset: morph_stanza_full2.zip 2020-12-09 – 20.19 MB – CC BY 4.0 Dataset: stanza_pretrain.zip 2020-12-09 – 105.77 MB – CC BY 4.0
Pretrained embeddings A list of pretrained embeddings for Swedish	Model	Swedish
Swedish Diachronic Word Embeddings Swedish Diachronic Word Embedding Models Trained on Historical Newspaper Data	Model	Swedish	Dataset: HENGCHEN-TAHMASEBI_-_2020_-_Kubhist2_diachronic_embeddings.zip 2024-01-25 – 15.13 GB – CC BY 4.0
Word Embeddings trained on English Wikipedia Word Embeddings trained on English Wikipedia	Model	English	Dataset: wiki_300_5_word2vec.model 2024-01-25 – 112.01 MB – CC BY 4.0 Dataset: wiki_300_5_word2vec.model.syn1neg.npy 2024-01-25 – 3.75 GB – CC BY 4.0 Dataset: wiki_300_5_word2vec.model.wv.vectors.npy 2024-01-25 – 3.75 GB – CC BY 4.0 Dataset: wiki_300_50_word2vec.model 2024-01-25 – 28.04 MB – CC BY 4.0 Dataset: wiki_300_50_word2vec.model.syn1neg.npy 2024-01-25 – 949.26 MB – CC BY 4.0 Dataset: wiki_300_50_word2vec.model.wv.vectors.npy 2024-01-25 – 949.26 MB – CC BY 4.0

Page manager: sb-webb