|
--- |
|
license: gpl-3.0 |
|
language: |
|
- it |
|
widget: |
|
- text: "Sentence: We tried 4 different styles of donuts. The distribution of numerals in the sentence is equal to" |
|
example_title: "Example 1" |
|
- text: "Sentence: No one's going to take you seriously if they're full of typos. The distribution of subordinates in the sentence is equal to" |
|
example_title: "Example 2" |
|
--- |
|
|
|
# Li-IT5 Base |
|
|
|
<p align="center"> |
|
<img src="lit5.png" alt="Linguistically-Informed T5" width="500"/> |
|
</p> |
|
|
|
|
|
This model is released as part of the paper "Linguistic Knowledge Can Enhance Encoder-Decoder Models (If You Let It)" (Miaschi et al., 2024). |
|
If you use this model in your work, we kindly ask you to cite our paper: |
|
|
|
```bibtex |
|
@inproceedings{miaschi_linguistic_knowledge, |
|
title = "Linguistic Knowledge Can Enhance Encoder-Deocer Models (If You Let It)", |
|
author = "Miaschi, Alessio and Dell'Orletta Felice and Venturi, Giulia", |
|
} |
|
``` |
|
|
|
Other information can be found in the original [GitHub repository](https://github.com/alemiaschi/linguistically_informed_t5/tree/main). |
|
|
|
## Model Description |
|
|
|
The model is based on a T5 model fine-tuned in a multitask fashion to solve a set of raw, morpho-syntactic and syntactic tasks (i.e. predictions of linguistic properties). |
|
The full list of the 10 linguistic properties used as intermediate tasks can be found in the original paper. |
|
|
|
This model is based on the Italian version of t5-base, [it5-base](https://huggingface.co/gsarti/it5-base). |
|
|
|
## Model variations |
|
|
|
The other fine-tuned models presented in the original study are the following: |
|
|
|
- [li-it5-small](https://huggingface.co/alemiaschi/li-it5-small) |
|
- [lit5-small](https://huggingface.co/alemiaschi/lit5-small) |