Instructions to use mattiaferrarini/BERToli with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use mattiaferrarini/BERToli with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("fill-mask", model="mattiaferrarini/BERToli")# Load model directly from transformers import AutoTokenizer, AutoModelForMaskedLM tokenizer = AutoTokenizer.from_pretrained("mattiaferrarini/BERToli") model = AutoModelForMaskedLM.from_pretrained("mattiaferrarini/BERToli") - Notebooks
- Google Colab
- Kaggle
BERToli ๐ถ๐ฎ๐น
About the model
BERToli is a BERT model for Italian song lyrics. It was obtained via continued pretraining of dbmdz/bert-base-italian-xxl-cased on ~106k Italian song lyrics from the Genius Song Lyrics Dataset.
The objective was Masked Language Modeling (MLM).
The training code is available on GitHub.
Evaluation
The base model and the adapted model were tested on a held-out set of ~6k songs with the following results:
| Model | MLM Loss | Perplexity |
|---|---|---|
| Base | 1.94 | 6.95 |
| BERToli | 1.45 | 4.26 |
Evaluation of the learned representations will be made available in the future, once a suitable dataset has been created / identified.
Why BERToli?
Pierangelo Bertoli (5 November 1942 โ 7 October 2002) was an Italian singer-songwriter and poet.
- Downloads last month
- 1
Model tree for mattiaferrarini/BERToli
Base model
dbmdz/bert-base-italian-xxl-cased