Byt5 Xxl

🚀 ByT5 - xxl

ByT5 is a tokenizer-free version of Google's T5, generally following the architecture of MT5. It offers a unique approach to language processing, eliminating the need for a tokenizer and showing great potential in handling various language tasks.

📦 Information Table

🚀 Quick Start

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

✨ Features

• ByT5 is a tokenizer - free version, following the architecture of MT5.
• It works especially well on noisy text data. For example, google/byt5 - xxl significantly outperforms mt5 - xxl on TweetQA.

💻 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-xxl')

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 using a tokenizer class for padding:

from transformers import T5ForConditionalGeneration, AutoTokenizer

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

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

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, the authors show that a standard Transformer architecture can be used with minimal modifications to process byte sequences. They 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. They 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 their contribution, they 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 their experiments.

📄 License

This model is licensed under the apache - 2.0 license.

📖 Paper and Authors

• Paper: ByT5: Towards a token - free future with pre - trained byte - to - byte models
• Authors: Linting Xue, Aditya Barua, Noah Constant, Rami Al - Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel