Skip to content

Dataset · jat-project

JAT tokenized time series

jat-project/jat-dataset-tokenized

A large corpus of tokenized time series (10 to 100 million records) without a detailed card description.

What's inside

The repository contains from 10 to 100 million records with time series modalities presented in tokenized form. A detailed description of the sources, language, and data structure is absent from the dataset card.

What you can do with it

  • Pre-train or fine-tune transformers on tokenized time series
  • Benchmark loading and processing speed for large sequential data arrays
Size
10M–100M rows
Downloads this month
237.3K
Likes
43

How to load it

from datasets import load_dataset

ds = load_dataset("jat-project/jat-dataset-tokenized")