---
language: de
datasets:
- deepset/germanquad
license: mit
thumbnail: https://thumb.tildacdn.com/tild3433-3637-4830-a533-353833613061/-/resize/720x/-/format/webp/germanquad.jpg
tags:
- exbert
model-index:
- name: deepset/gelectra-large-germanquad
  results:
  - task:
      type: question-answering
      name: Question Answering
    dataset:
      name: deepset/germanquad
      type: deepset/germanquad
      config: plain_text
      split: test
    metrics:
    - name: Exact Match
      type: exact_match
      value: 70.2359
      verified: true
    - name: F1
      type: f1
      value: 86.27
      verified: true
---

![bert_image](https://thumb.tildacdn.com/tild3433-3637-4830-a533-353833613061/-/resize/720x/-/format/webp/germanquad.jpg)

## Overview
**Language model:** gelectra-large-germanquad   
**Language:** German  
**Training data:** GermanQuAD train set (~ 12MB)  
**Eval data:** GermanQuAD test set (~ 5MB)   
**Infrastructure**: 1x V100 GPU  
**Published**: Apr 21st, 2021

## Details
- We trained a German question answering model with a gelectra-large model as its basis.
- The dataset is GermanQuAD, a new, German language dataset, which we hand-annotated and published [online](https://deepset.ai/germanquad).
- The training dataset is one-way annotated and contains 11518 questions and 11518 answers, while the test dataset is three-way annotated so that there are 2204 questions and with 2204·3−76 = 6536 answers, because we removed 76 wrong answers.

See https://deepset.ai/germanquad for more details and dataset download in SQuAD format.

## Hyperparameters
```
batch_size = 24
n_epochs = 2
max_seq_len = 384
learning_rate = 3e-5
lr_schedule = LinearWarmup
embeds_dropout_prob = 0.1
```
## Performance
We evaluated the extractive question answering performance on our GermanQuAD test set.
Model types and training data are included in the model name. 
For finetuning XLM-Roberta, we use the English SQuAD v2.0 dataset.
The GELECTRA models are warm started on the German translation of SQuAD v1.1 and finetuned on [GermanQuAD](https://deepset.ai/germanquad). 
The human baseline was computed for the 3-way test set by taking one answer as prediction and the other two as ground truth.
![performancetable](https://images.prismic.io/deepset/1c63afd8-40e6-4fd9-85c4-0dbb81996183_german-qa-vs-xlm-r.png) 

## Authors
 **Timo Möller:** timo.moeller@deepset.ai     
 **Julian Risch:** julian.risch@deepset.ai   
 **Malte Pietsch:** malte.pietsch@deepset.ai   
 
## About us
<div class="grid lg:grid-cols-2 gap-x-4 gap-y-3">
    <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center">
         <img alt="" src="https://huggingface.co/spaces/deepset/README/resolve/main/haystack-logo-colored.svg" class="w-40"/>
     </div>
    <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center">
         <img alt="" src="https://huggingface.co/spaces/deepset/README/resolve/main/deepset-logo-colored.svg" class="w-40"/>
     </div>
</div>

[deepset](http://deepset.ai/) is the company behind the open-source NLP framework [Haystack](https://haystack.deepset.ai/) which is designed to help you build production ready NLP systems that use: Question answering, summarization, ranking etc.


Some of our other work: 
- [Distilled roberta-base-squad2 (aka "tinyroberta-squad2")]([https://huggingface.co/deepset/tinyroberta-squad2)
- [German BERT (aka "bert-base-german-cased")](https://deepset.ai/german-bert)
- [GermanQuAD and GermanDPR datasets and models (aka "gelectra-base-germanquad", "gbert-base-germandpr")](https://deepset.ai/germanquad)

## Get in touch and join the Haystack community

<p>For more info on Haystack, visit our <strong><a href="https://github.com/deepset-ai/haystack">GitHub</a></strong> repo and <strong><a href="https://haystack.deepset.ai">Documentation</a></strong>. 

We also have a <strong><a class="h-7" href="https://haystack.deepset.ai/community/join"><img alt="slack" class="h-7 inline-block m-0" style="margin: 0" src="https://huggingface.co/spaces/deepset/README/resolve/main/Slack_RGB.png"/>community open to everyone!</a></strong></p>

[Twitter](https://twitter.com/deepset_ai) | [LinkedIn](https://www.linkedin.com/company/deepset-ai/) | [Slack](https://haystack.deepset.ai/community/join) | [GitHub Discussions](https://github.com/deepset-ai/haystack/discussions) | [Website](https://deepset.ai)

By the way: [we're hiring!](http://www.deepset.ai/jobs)