Adapter lenglaender/xlm-roberta-base-lora-cls-yelp-polarity for google-t5/t5-base

An adapter for the google-t5/t5-base model that was trained on the yelp_polarity dataset and includes a prediction head for classification.

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("google-t5/t5-base")
adapter_name = model.load_adapter("AdapterHub/xlm-roberta-base-lora-cls-yelp-polarity", source="hf", set_active=True)

Architecture & Training

LoRA has r=8 and alpha=8 and was trained with dropout=0.1

Evaluation results

Accuracy on Amazon polarity dataset: 98.15%

Author Information

• Author name: Leon Engländer
• Author links: GitHub, Twitter