metadata
tags:
- roberta
- adapterhub:dp/ud_ewt
- adapter-transformers
datasets:
- universal_dependencies
language:
- en
Adapter AdapterHub/roberta-base-pf-ud_en_ewt for roberta-base
An adapter for the roberta-base model that was trained on the dp/ud_ewt dataset and includes a prediction head for dependency parsing.
This adapter was created for usage with the adapter-transformers library.
Usage
First, install adapter-transformers:
pip install -U adapter-transformers
Note: adapter-transformers is a fork of transformers that acts as a drop-in replacement with adapter support. More
Now, the adapter can be loaded and activated like this:
from transformers import AutoModelWithHeads
model = AutoModelWithHeads.from_pretrained("roberta-base")
adapter_name = model.load_adapter("AdapterHub/roberta-base-pf-ud_en_ewt", source="hf", set_active=True)
Architecture & Training
This adapter was trained using adapter-transformer's example script for dependency parsing. See https://github.com/Adapter-Hub/adapter-transformers/tree/master/examples/dependency-parsing.
Evaluation results
Scores achieved by dependency parsing adapters on the test set of UD English EWT after training:
| UAS | LAS | |
|---|---|---|
bert-base-uncased |
91.74 | 89.15 |
roberta-base |
91.43 | 88.43 |