|
--- |
|
language: |
|
- it |
|
pipeline_tag: fill-mask |
|
library_name: transformers |
|
widget: |
|
- example_title: Esempio 1 |
|
text: >- |
|
Più tardi, durante la sua carriera, la gravidanza la costrinse a cancellare |
|
le sue apparizioni programmate nell'autunno del 1935. Il 29 marzo 1936 entrò |
|
in una clinica di Londra in [MASK] della nascita del suo bimbo |
|
- example_title: Esempio 2 |
|
text: >- |
|
Nella stagione 2017-2018 il Karlsruhe, allenato da Alois Schwartz, [MASK] il |
|
campionato di 3. Liga al 3º posto. |
|
- example_title: Esempio 3 |
|
text: >- |
|
Conchita Supervía, nome alla nascita María de la Concepción Supervía Pascual |
|
(Barcellona, 8 dicembre 1895[1] – Londra, 30 marzo 1936), è [MASK] un |
|
mezzosoprano spagnolo molto famosa nel panorama operistico europeo e |
|
statunitense, attiva anche in campo concertistico |
|
license: apache-2.0 |
|
datasets: |
|
- wikipedia |
|
metrics: |
|
- perplexity |
|
--- |
|
|
|
**bertina-3M** is an italian foundational model based on bert, pretrained from scratch on 30GB of italian Wikipedia articles (10M sentences, 329M tokens). |
|
|
|
It has 3M parameters and uses a 512 context window size. |
|
|
|
The project is still a work in progress, new versions will come with time. |
|
|
|
Use it as a foundational model to be finetuned on specific italian tasks. |
|
|
|
# Training |
|
|
|
- epochs: 4 |
|
- lr: 4e-4 |
|
- optim: AdamW (beta_1=0.8) |
|
- weight_decay: 1e-2 |
|
|
|
- Dev set perplexity: 19 (it's a 12MB model!) |
|
# Evaluation (UINAUIL) |
|
|
|
Following the [UINAUIL setup](https://github.com/valeriobasile/uinauil/tree/main) we can summarise the following results on BERTINA-3M: |
|
|
|
**CLASSIFICATION TASKS** |
|
``` |
|
task,type,p,r,f1,acc |
|
haspeede,classification,0.699,0.687,0.680,0.685 |
|
ironita,classification,0.701,0.701,0.701,0.701 |
|
sentipolc,classification,0.649,0.588,0.587,0.560 |
|
``` |
|
|
|
**ENTAILMENT TASKS** |
|
|
|
``` |
|
task,type,p,r,f1,acc |
|
textualentailment,entailment,0.423,0.530,0.401,0.530 |
|
``` |
|
|
|
**SEQUENCE TASKS** |
|
|
|
``` |
|
task,type,acc |
|
eventi,NER,0.835 |
|
facta,NER,0.967 |
|
``` |
|
|
|
# License |
|
|
|
BERTINA-3M can be freely used for research and commercial purposes. |
|
|
|
# Citation |
|
|
|
If you're using BERTINA-3M in your scientific work, please cite with: |
|
``` |
|
@misc{ |
|
Sciancalepore, |
|
title={mascit/bertina-3M}, |
|
url={https://huggingface.co/mascIT/bertina-3M}, |
|
journal={mascIT/bertina-3M · Hugging Face}, |
|
publisher={mascIT}, |
|
author={Sciancalepore, Mauro} |
|
} |
|
``` |