Byt5 Xl

🚀 ByT5 - xl

ByT5 is a tokenizer-free version of Google's T5. It generally follows the architecture of MT5. This model offers the ability to process text in various languages without the need for a tokenizer, and it performs well on noisy text data.

🚀 Quick Start

ByT5 was pre - trained on mC4 with an average span - mask of 20 UTF - 8 characters, excluding any supervised training. So, it needs to be fine - tuned before being used on downstream tasks.

✨ Features

• Tokenizer - free: ByT5 can work directly on raw UTF - 8 bytes, eliminating the need for a tokenizer.
• Multilingual support: It supports a wide range of languages, including af, am, ar, etc.
• Good performance on noisy data: It significantly outperforms mt5 - xl on tasks like TweetQA.

📦 Installation

The README doesn't provide specific installation steps, so this section is skipped.

💻 Usage Examples

Basic Usage

ByT5 works on raw UTF - 8 bytes and can be used without a tokenizer:

from transformers import T5ForConditionalGeneration
import torch

model = T5ForConditionalGeneration.from_pretrained('google/byt5-xl')

input_ids = torch.tensor([list("Life is like a box of chocolates.".encode("utf-8"))]) + 3  # add 3 for special tokens
labels = torch.tensor([list("La vie est comme une boîte de chocolat.".encode("utf-8"))]) + 3  # add 3 for special tokens

loss = model(input_ids, labels=labels).loss # forward pass

Advanced Usage

For batched inference & training, it is recommended to use a tokenizer class for padding:

from transformers import T5ForConditionalGeneration, AutoTokenizer

model = T5ForConditionalGeneration.from_pretrained('google/byt5-xl')
tokenizer = AutoTokenizer.from_pretrained('google/byt5-xl')

model_inputs = tokenizer(["Life is like a box of chocolates.", "Today is Monday."], padding="longest", return_tensors="pt")
labels = tokenizer(["La vie est comme une boîte de chocolat.", "Aujourd'hui c'est lundi."], padding="longest", return_tensors="pt").input_ids

loss = model(**model_inputs, labels=labels).loss # forward pass

📚 Documentation

Abstract

Most widely - used pre - trained language models operate on sequences of tokens corresponding to word or subword units. Encoding text as a sequence of tokens requires a tokenizer, which is typically created as an independent artifact from the model. Token - free models that instead operate directly on raw text (bytes or characters) have many benefits: they can process text in any language out of the box, they are more robust to noise, and they minimize technical debt by removing complex and error - prone text preprocessing pipelines. Since byte or character sequences are longer than token sequences, past work on token - free models has often introduced new model architectures designed to amortize the cost of operating directly on raw text. In this paper, we show that a standard Transformer architecture can be used with minimal modifications to process byte sequences. We carefully characterize the trade - offs in terms of parameter count, training FLOPs, and inference speed, and show that byte - level models are competitive with their token - level counterparts. We also demonstrate that byte - level models are significantly more robust to noise and perform better on tasks that are sensitive to spelling and pronunciation. As part of our contribution, we release a new set of pre - trained byte - level Transformer models based on the T5 architecture, as well as all code and data used in our experiments.

🔧 Technical Details

The README doesn't provide in - depth technical details (>50 words), so this section is skipped.

📄 License

This project is licensed under the Apache 2.0 license.