T0: Multitask Prompted Training Enables Zero-Shot Task Generalization

About Share Download Add to

❤️ Become The AI Epiphany Patreon ❤️ ► 👨‍👩‍👧‍👦 JOIN OUR DISCORD COMMUNITY: Discord ► 📢 SUBSCRIBE TO MY MONTHLY AI NEWSLETTER: Substack ► In this video I cover the “Multitask Prompted Training Enables Zero-Shot Task Generalization“ paper that introduced the T0 transformer. T0 basically = Google's T5 LM pretraining additional training on prompted datasets. The paper came out of the BigScience workshop. ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ ✅ Paper: ✅ BigScience: ✅ Models on HF hub:

Share with your friends

Link:

Embed:

<iframe width="640" height="360" src="//myvideo.cc/embed/U3VqVHRaUWpBQ2JnVC8zQ29yM2VDMmEwbi94RjMwRzhCck5HRHh5Q1krRT0" frameborder="0" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>

Video Size:

Custom size:

Autoplay video

Hide player controls

Hide resume playing

Add to Playlist:

Favorites

My Playlist

Watch Later

T0: Multitask Prompted Training Enables Zero-Shot Task Generalization | Paper Explained