metadata
tags:
- text2text-generation
- adapter-transformers
- adapterhub:mt/wmt16_en_ro
- mbart
license: apache-2.0
Adapter facebook-mbart-large-cc25_translation_wmt16_ro_en_pfeiffer for facebook/mbart-large-cc25
Adapter for mbart-large-cc25 in Pfeiffer architecture with reduction factor 2 trained on the WMT16 Romanian-English translation task. Training for 10 epochs with early stopping and a learning rate of 1e-4. After post-processing following https://github.com/huggingface/transformers/blob/master/examples/legacy/seq2seq/romanian_postprocessing.md, it achieves a BLEU score of 36.3.
This adapter was created for usage with the Adapters library.
Usage
First, install adapters:
pip install -U adapters
Now, the adapter can be loaded and activated like this:
from adapters import AutoAdapterModel
model = AutoAdapterModel.from_pretrained("facebook/mbart-large-cc25")
adapter_name = model.load_adapter("AdapterHub/facebook-mbart-large-cc25_translation_wmt16_ro_en_pfeiffer")
model.set_active_adapters(adapter_name)
Architecture & Training
- Adapter architecture: pfeiffer
- Prediction head: seq2seq lm
- Dataset: WMT16 English-Romanian
Author Information
- Author name(s): Clifton Poth
- Author email: calpt@mail.de
- Author links: Website, GitHub, Twitter
Citation
This adapter has been auto-imported from https://github.com/Adapter-Hub/Hub/blob/master/adapters/ukp/facebook-mbart-large-cc25_translation_wmt16_ro_en_pfeiffer.yaml.