README.md

---
license: mit
language:
- it
base_model:
- GroNLP/gpt2-small-italian
pipeline_tag: text-generation
tags:
- legal
library_name: transformers
---

## Usage
```
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("VerbACxSS/sempl-it-gpt2-small-italian", model_max_length=1024)
model = AutoModelForCausalLM.from_pretrained("VerbACxSS/sempl-it-gpt2-small-italian")

model.eval()

text_to_simplify = 'Nella fattispecie, questo documento è di natura prescrittiva'
prompt = f'### [Input]:\n{text_to_simplify}\n\n###[Output]:\n'

x = tokenizer(prompt, max_length=1024, truncation=True, padding=True, return_tensors='pt').input_ids
y = model.generate(x, max_length=1024)[0]
y_dec = tokenizer.decode(y, max_length=1024, truncation=True)
output = y_dec.split('###[Output]:\n')[1].split('<|endoftext|>')[0].strip()

print(output)
```

## Acknowledgements
This contribution is a result of the research conducted within the framework of the PRIN 2020 (Progetti di Rilevante Interesse Nazionale) "VerbACxSS: on analytic verbs, complexity, synthetic verbs, and simplification. For accessibility" (Prot. 2020BJKB9M), funded by the Italian Ministero dell'Università e della Ricerca.