Skup podataka sa Hugging Face · lightonai

lightonai/trec-covid-decontaminated

Licenca
MITdozvoljena komercijalna upotrebaTekst licence
Licenca dozvoljava komercijalnu upotrebu. Obično treba navesti autora — tačni uslovi su u tekstu licence.
Formati
parquet, PARQUET
Veličina
54 MB
Ažurirano
25. 3. 2026.

Opis

trec-covid (Decontaminated) A decontaminated version of the trec-covid dataset from the BEIR benchmark, with samples found in the mgte-en pre-training dataset removed. Decontamination methodology Contamination was detected using a two-pass approach against the full mgte-en dataset (484 GB, 1,235 parquet files): Pass 1: Exact hash matching All texts (queries and corpus documents) were normalized (lowercased, unicode NFKD, whitespace collapsed) and…

Opis sa Hugging Face, skraćen.

Fajlovi (3)

  • PARQUETcorpus.parquet54 MB
  • PARQUETqrels_test.parquet299 kB
  • PARQUETqueries.parquet5,1 kB

DataHunter ne čuva i ne deli ove fajlove: svaki link vodi do izvornog sajta. Pre upotrebe tamo proverite licencu.

Pronađi druge skupove