- KeLP – Kernel-based Learning Platform for scalable ML.
- GAN-BERT – Few-shot adversarial learning on Transformers.
- GAN-BERT-PyTorch – PyTorch/HF port of GAN-BERT.
- MT-GANBERT – Multi-task + GAN-BERT for sustainable NLP.
- EthicalNN – Ethics by Design framework in PyTorch.
- GrUT – Semantic parsing for Human-Robot Interaction.
- LU4R – Adaptive spoken language understanding for robots.
- ACLPUB2 – ACL proceedings generation tool.
- BacKGen – Background Knowledge Generator.
- dats – Data augmentation for NLP.
📚 Datasets & Benchmarks
- ExtremITA – Instruction-tuned LLM for Italian.
- U-DepPLLaMA – Universal dependency parsing with LLMs.
- MM-IGLU, MM-IGLU-IT & MM-IGLU-Dialogues – Multimodal grounded understanding benchmarks.
- FEVER-it – Italian fact-checking dataset & pipeline.
- HuRIC – Human-Robot Interaction Corpus 2.0.
- GQA-it – 1M+ Visual Question Answering pairs in Italian.
- mscoco-it – 600K captions for Italian Image Captioning.
- msr-vtt-it – 200K Italian video caption pairs.
- SQuAD-it – 60K Q/A triples for reading comprehension.
- ABSITA – Tourism opinion mining dataset.
- SENTIPOLC – 10K annotated Italian tweets.
🎓 Teaching Material
