# Contra-OOD

**Repository Path**: bobtuan/Contra-OOD

## Basic Information

- **Project Name**: Contra-OOD
- **Description**: Source code for paper "Contrastive Out-of-Distribution Detection for Pretrained Transformers", EMNLP 2021
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-11-25
- **Last Updated**: 2021-11-25

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Contra-OOD

Code for EMNLP 2021 paper [Contrastive Out-of-Distribution Detection for Pretrained Transformers](https://arxiv.org/abs/2104.08812).

## Requirements
* [PyTorch](http://pytorch.org/)
* [Transformers](https://github.com/huggingface/transformers)
* datasets
* wandb
* tqdm
* scikit-learn

## Dataset
Most datasets used in paper are automatically downloaded by the datasets package. Instructions on downloading sst2 and multi30k are provided in ``readme.txt`` under the ``data`` folder.

## Training and Evaluation

Finetune the PLM with the following command:

```bash
>> python run.py --task_name task
```

The ``task_name`` parameter can take ``sst2``, ``imdb``, ``trec``, or ``20ng``. The training loss and evaluation results on the test set are synced to the wandb dashboard.