T5 V1 1 Xxl

🚀 Google's T5 Version 1.1

Google's T5 Version 1.1 is an enhanced natural language processing model with several improvements over the original T5, offering better performance and flexibility for various NLP tasks.

🚀 Quick Start

This README provides an in - depth introduction to Google's T5 Version 1.1, including its improvements, pre - training details, and related research information.

✨ Features

Improvements in Version 1.1

• Activation Function: T5 Version 1.1 uses GEGLU activation in the feed - forward hidden layer instead of ReLU. For more details, refer to here.
• Dropout: Dropout was turned off during pre - training, which led to quality improvements. It should be re - enabled during fine - tuning.
• Pre - training Dataset: It was pre - trained only on C4 without mixing in downstream tasks.
• Parameter Sharing: There is no parameter sharing between the embedding and classifier layer.
• Model Naming and Shape: "xl" and "xxl" replace "3B" and "11B". The model shapes have larger d_model and smaller num_heads and d_ff.

Note

T5 Version 1.1 was only pre - trained on C4 without any supervised training. So, it must be fine - tuned before being used on a downstream task.

📚 Documentation

Pretraining Dataset

The model was pre - trained on the C4 dataset.

Other Community Checkpoints

You can find other community checkpoints [here](https://huggingface.co/models?search=t5 - v1_1).

Paper

The related research paper is Exploring the Limits of Transfer Learning with a Unified Text - to - Text Transformer.

Authors

Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu

Abstract

Transfer learning, a powerful technique in natural language processing (NLP), involves pre - training a model on a data - rich task and then fine - tuning it on a downstream task. The effectiveness of transfer learning has led to a variety of approaches, methodologies, and practices. In this paper, the authors introduce a unified framework that converts every language problem into a text - to - text format to explore the landscape of transfer learning techniques for NLP. Their systematic study compares pre - training objectives, architectures, unlabeled datasets, transfer approaches, and other factors on dozens of language understanding tasks. By combining insights from their exploration with scale and the new “Colossal Clean Crawled Corpus”, they achieve state - of - the - art results on many benchmarks, including summarization, question answering, and text classification. To facilitate future work on transfer learning for NLP, they release their dataset, pre - trained models, and code.

📄 License

This project is licensed under the Apache-2.0 license.