Edit model card

ul2-large-en-nl for English to Dutch translation

Fine-tuned T5 model on English to Dutch translation that was pretrained on Dutch using a UL2 (Mixture-of-Denoisers) objective. The T5 model was introduced in this paper and first released at this page. The UL2 objective was introduced in this paper and first released at this page.

Model description

T5 is an encoder-decoder model and treats all NLP problems in a text-to-text format.

ul2-large-en-nl-v3 T5 is a transformers model fine-tuned on parallel sentence and paragraph pairs sampled from books.

This model used the T5 v1.1 improvements compared to the original T5 model during the pretraining:

  • GEGLU activation in the feed-forward hidden layer, rather than ReLU - see here
  • Dropout was turned off during pre-training. Dropout should be re-enabled during fine-tuning
  • Pre-trained on self-supervised objective only without mixing in the downstream tasks
  • No parameter sharing between embedding and classifier layer

UL2 pretraining objective

This model was pretrained with the UL2's Mixture-of-Denoisers (MoD) objective, that combines diverse pre-training paradigms together. UL2 frames different objective functions for training language models as denoising tasks, where the model has to recover missing sub-sequences of a given input. During pre-training it uses a novel mixture-of-denoisers that samples from a varied set of such objectives, each with different configurations. UL2 is trained using a mixture of three denoising tasks:

  1. R-denoising (or regular span corruption), which emulates the standard T5 span corruption objective;
  2. X-denoising (or extreme span corruption); and
  3. S-denoising (or sequential PrefixLM).

During pre-training, we sample from the available denoising tasks based on user-specified ratios. UL2 introduces a notion of mode switching, wherein downstream fine-tuning is associated with specific pre-training denoising task. During the pre-training, a paradigm token is inserted to the input ([NLU] for R-denoising, [NLG] for X-denoising, or [S2S] for S-denoising) indicating the denoising task at hand. Then, during fine-tuning the same input token should be inserted to get the best performance for different downstream fine-tuning tasks.

Intended uses & limitations

This model was fine-tuned on parallel sentence and paragraph pairs and can be used for machine translation.

How to use

Here is how to use this model in PyTorch:

model_name = "yhavinga/ul2-large-en-nl-v3"
from transformers import AutoTokenizer
from transformers import AutoModelForSeq2SeqLM
from transformers import pipeline
import torch
device_num = 0 if torch.cuda.is_available() else -1
device = "cpu" if device_num < 0 else f"cuda:{device_num}"

tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name, use_auth_token=True).to(
params = {"max_length": 370, "num_beams": 4, "early_stopping": True}
translator = pipeline("translation", tokenizer=tokenizer, model=model, device=device_num)
print(translator("Young Wehling was hunched in his chair, his head in his hand. He was so rumpled, so still and colorless as to be virtually invisible.",

Limitations and bias

The training data used for this model contains a lot of unfiltered content from the internet, which is far from neutral. Therefore, the model can have biased predictions. This bias will also affect all fine-tuned versions of this model.

Training data

The ul2-large-en-nl T5 model was pre-trained simultaneously on a combination of several datasets, including the full config of the "mc4_nl_cleaned" dataset, which is a cleaned version of Common Crawl's web crawl corpus, Dutch books, the Dutch subset of Wikipedia (2022-03-20), and a subset of "mc4_nl_cleaned" containing only texts from Dutch newspapers.

After pre-training, the model was fine-tuned on a translation dataset containing 13 million sentence and paragraph pairs sampled from books.

Training procedure


The ul2-large-en-nl T5 model uses a SentencePiece unigram tokenizer with a vocabulary of 32,000 tokens. The tokenizer includes the special tokens <pad>, </s>, <unk>, known from the original T5 paper, [NLU], [NLG] and [S2S] for the MoD pre-training, and <n> for newline. During pre-training with the UL2 objective, input and output sequences consist of 512 consecutive tokens. The tokenizer does not lowercase texts and is therefore case-sensitive; it distinguises between dutch and Dutch. Additionally, 100+28 extra tokens were added for pre-training tasks, resulting in a total of 32,128 tokens.


This model was fine-tuned on a dataset containing 13M sentence and paragraph translation pairs sampled from books for three epochs.

Wandb run https://wandb.ai/yepster/ul2-large-de-neddx2-en-nl/runs/30arxggk?workspace=user-yepster

  • Pre-trained model used as starting point: yhavinga/ul2-large-dutch-english (3150k checkpoint)

For the concluding ~half epoch, a HuggingFace Flax based trainer was used with the following settings:

  • Batch Size: Total effective batch size of 512, achieved via per-device settings and gradient accumulation.
  • Learning Rate: Set at 0.0009, with linear schedule and 500 step warmup.
  • Optimizer: AdamW with beta1=0.9, beta2=0.997, epsilon=1e-8.
  • Weight Decay: Configured to 0.001 for regularization.
  • Additional Parameters: Dropout rate of 0.01, label smoothing factor of 0.11, and sequence length of 370 tokens. Model datatype is bfloat16, z_loss at 0.0001.

Evaluation results



This project would not have been possible without compute generously provided by Google through the TPU Research Cloud. Thanks to the Finnish-NLP authors for releasing their code for the UL2 objective and associated task definitions. Thanks to Stephenn Fernandes for helping me get started with the t5x framework.

Created by Yeb Havinga

Downloads last month
Model size
783M params
Tensor type

Dataset used to train yhavinga/ul2-large-en-nl-v3