Leonardo Di Perna

Biomedical Engineer and Data Scientist.
In love with travelling 🌍 and Artificial Intelligence 📖

[Leonardo Di Perna @Leodipe94]

Oltre la pasta e la pizza, in Italia c'è .. GilBERTo!

@GiulioRavasio and I are very happy to release GilBERTo 🇮🇹 an Italian uncased language model based on RoBERTa trained on ~71GB Ita Common Crawl!
Try GilBERTo with #fairseq or find it directly into @huggingface 🤗library
Ciao!

[Leonardo Di Perna @Leodipe94]

@cappiello_an @huggingface Congrats Antonio and thanks for your fantastic contribution! #nlp #italian #model

[Leonardo Di Perna @Leodipe94]

@TensorFlow @GiulioRavasio

[Yann LeCun @ylecun]

How to train a model with 10^11 parameters without running out of GPU memory?
Use DeepSpeed from Microsoft Research!
It's PyTorch compatible.
It partitions the network onto multiple processors automatically and efficiently.

[Sebastian Ruder @seb_ruder]

Transfer learning is increasingly going multilingual with language-specific BERT models:

- 🇩🇪 German BERT https://t.co/M6Mo1drsgo
- 🇫🇷 CamemBERT https://t.co/flzuE9zNJN, FlauBERT https://t.co/xXlSeytUjj
- 🇮🇹 AlBERTo https://t.co/1nhchd2PG1
- 🇳🇱 RobBERT https://t.co/wR2Bln1Fyx

[Leonardo Di Perna @Leodipe94]

@seb_ruder There is also 🇮🇹 GilBERTo🇮🇹 as an italian pretrained language model based on RoBERTa and trained on a large common crawl corpus ~71Gb! You can find it here: https://t.co/fI02DZszaB or through the beautiful #huggingface

[Leonardo Di Perna @Leodipe94]

@AlbertoBernacc1 Thanks Alberto for sharing it! We hope to contribuite even more in the future and increase italian nlp applications!!

[Leonardo Di Perna @Leodipe94]

You can find more details about the model on:
- huggingface models https://t.co/dXfrD1Yvee
- github https://t.co/fI02DZszaB

If you find it useful, please star the repo and we strongly recommend to contribute to it

[Leonardo Di Perna @Leodipe94]

Oltre la pasta e la pizza, in Italia c'è .. GilBERTo!

@GiulioRavasio and I are very happy to release GilBERTo 🇮🇹 an Italian uncased language model based on RoBERTa trained on ~71GB Ita Common Crawl!
Try GilBERTo with #fairseq or find it directly into @huggingface 🤗library
Ciao!

[Pau Rué @PauRueQ]

Fascinated by how easy it is to put together a demo with @streamlit and @huggingface's transformers 🤯

[UCBM @CampusBioMedico]

6 mesi alla @Philips per fare pratica con il #machinelearning. L'esperienza di Leonardo, studente UCBM: https://t.co/j14AsHSwHu

[Leonardo Di Perna @Leodipe94]

#EMBC16 Campus Biomedico in tour