|
--- |
|
license: apache-2.0 |
|
language: en |
|
--- |
|
|
|
# LongT5 (transient-global attention, large-sized model) |
|
|
|
LongT5 model pre-trained on English language. The model was introduced in the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/pdf/2112.07916.pdf) by Guo et al. and first released in [the LongT5 repository](https://github.com/google-research/longt5). All the model architecture and configuration can be found in [Flaxformer repository](https://github.com/google/flaxformer) which uses another Google research project repository [T5x](https://github.com/google-research/t5x). |
|
|
|
Disclaimer: The team releasing LongT5 did not write a model card for this model so this model card has been written by the Hugging Face team. |
|
|
|
## Model description |
|
LongT5 model is an encoder-decoder transformer pre-trained in a text-to-text denoising generative setting ([Pegasus-like generation pre-training](https://arxiv.org/pdf/1912.08777.pdf)). LongT5 model is an extension of [T5 model](https://arxiv.org/pdf/1910.10683.pdf), and it enables using one of the two different efficient attention mechanisms - (1) Local attention, or (2) Transient-Global attention. The usage of attention sparsity patterns allows the model to efficiently handle input sequence. |
|
|
|
LongT5 is particularly effective when fine-tuned for text generation (summarization, question answering) which requires handling long input sequences (up to 16,384 tokens). |
|
|
|
Results of LongT5 (transient-global attention, large-sized model) fine-tuned on multiple (summarization, QA) tasks. |
|
|
|
| Dataset | Rouge-1 | Rouge-2 | Rouge-Lsum | |
|
| --- | --- | --- | --- | |
|
| arXiv (16k input) | 48.28 | 21.63 | 44.11 | |
|
| PubMed (16k input) | 49.98 | 24.69 | 46.46 | |
|
| BigPatent (16k input) | 70.38 | 56.81 | 62.73 | |
|
| MultiNews (8k input) | 47.18 | 18.44 | 24.18 | |
|
| MediaSum (4k input) | 35.54 | 19.04 | 32.20 | |
|
| CNN / DailyMail (4k input) | 42.49 | 20.51 | 40.18 | |
|
|
|
| Dataset | EM | F1 | |
|
| --- | --- | --- | |
|
| Natural Questions (4k input) | 60.77 | 65.38 | |
|
| Trivia QA (16k input) | 78.38 | 82.45 | |
|
|
|
## Intended uses & limitations |
|
|
|
The model is mostly meant to be fine-tuned on a supervised dataset. See the [model hub](https://huggingface.co/models?search=longt5) to look for fine-tuned versions on a task that interests you. |
|
|
|
### How to use |
|
|
|
```python |
|
from transformers import AutoTokenizer, LongT5Model |
|
|
|
tokenizer = AutoTokenizer.from_pretrained("google/long-t5-tglobal-large") |
|
model = LongT5Model.from_pretrained("google/long-t5-tglobal-large") |
|
|
|
inputs = tokenizer("Hello, my dog is cute", return_tensors="pt") |
|
outputs = model(**inputs) |
|
|
|
last_hidden_states = outputs.last_hidden_state |
|
``` |
|
|
|
### BibTeX entry and citation info |
|
|
|
```bibtex |
|
@article{guo2021longt5, |
|
title={LongT5: Efficient Text-To-Text Transformer for Long Sequences}, |
|
author={Guo, Mandy and Ainslie, Joshua and Uthus, David and Ontanon, Santiago and Ni, Jianmo and Sung, Yun-Hsuan and Yang, Yinfei}, |
|
journal={arXiv preprint arXiv:2112.07916}, |
|
year={2021} |
|
} |
|
``` |
