ASTROMER: Building light curves embeddings using transfomers
Abstract
ASTROMER is a Transformer-based model trained on millions of stars for the representation of light curves. Pretrained models can be directly used or finetuned on specific datasets. ASTROMER is useful in downstream tasks in which data are limited to train deep learning models.
- Publication:
-
Astrophysics Source Code Library
- Pub Date:
- May 2022
- Bibcode:
- 2022ascl.soft05020D
- Keywords:
-
- Software