This is a Spanish
T5 (small arch) trained from scratch on the
large_spanish_corpus aka BETO's corpus with
FlaxThis is part of the
Flax/Jax Community Week, organised by
HuggingFace and TPU usage sponsored by Google.
The dataset is about 20 GB. 95% of the data was used for training and the rest 5% for validation.
1@misc{mromero2021spanish-t5-small,
2 title={Spanish T5 (small) by Manuel Romero},
3 author={Romero, Manuel},
4 publisher={Hugging Face},
5 journal={Hugging Face Hub},
6 howpublished={\url{https://huggingface.co/flax-community/spanish-t5-small}},
7 year={2021}
8}
