Adapters are Lightweight 🤖

"Adapter" refers to a set of newly introduced weights, typically within the layers of a transformer model. Adapters provide an alternative to fully fine-tuning the model for each downstream task, while maintaining performance. They also have the added benefit of requiring as little as 1MB of storage space per task!

Modular, Composable, and Extensible 🔧

Adapters, being self-contained moduar units, allow for easy extension and composition. This opens up opportunities to compose adapters to solve new tasks.

Built on HuggingFace 🤗 Transformers 🚀

AdapterHub builds on the HuggingFace transformers framework requiring as little as two additional lines of code to train adapters for a downstream task.

Quickstart 🔥

Inference Training

Load an Adapter for Inference 🏄

Open In Colab

Loading existing adapters from our repository is as simple as adding one additional line of code:

model = AutoModelForSequenceClassification.from_pretrained("bert-base-uncased")
model.load_adapter("sentiment/sst-2@ukp")

The SST adapter is light-weight: it is only 3MB! At the same time, it achieves results that are on-par with fully fine-tuned BERT. We can now leverage SST adapter to predict the sentiment of sentences:

tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
tokens = tokenizer.tokenize("AdapterHub is awesome!")
input_tensor = torch.tensor([
    tokenizer.convert_tokens_to_ids(tokens)
])
outputs = model(
    input_tensor,
    adapter_names=['sst-2']
)

Training a new task adapter requires only few modifications compared to fully fine-tuning a model with Hugging Face's . We first load a pre-trained model, e.g., and add a new task adapter:

By calling we freeze all transformer parameters except for the parameters of sst-2 adapter. Before training we add a new classification head to our model:

The weights of this classification head can be stored together with the adapter weights to allow for a full reproducibility. The method call registers the sst-2 adapter as a default for training. This also supports adapter stacking and adapter fusion!

We can then train our adapter using the Hugging Face :

That's it! exports all adapters. Consider sharing them on AdapterHub!

Citation 📝

@article{pfeiffer2020AdapterHub,
    title={AdapterHub: A Framework for Adapting Transformers},
    author={Jonas Pfeiffer and
            Andreas R\"uckl\'{e} and
            Clifton Poth and
            Aishwarya Kamath and
            Ivan Vuli\'{c} and
            Sebastian Ruder and
            Kyunghyun Cho and
            Iryna Gurevych},
    journal={arXiv preprint},
    year={2020},
    url={https://arxiv.org/abs/2007.07779}
}