---
language:
- de
tags:
- deepset/gbert-large
---
# German Sentiment Analysis

This model predicts sentiment for German text. 

# Usage
First set up the model:

```python
# if necessary:
# !pip install transformers
from transformers import pipeline

sentiment_model = pipeline(model="aari1995/German_Sentiment")
```

to use it:

```python
sentence = ["Ich liebe die Bahn. Pünktlich wie immer ... -.-","Krasser Service"]
result = sentiment_model(sentence)
print(result)
#Output:
#[{'label': 'negative', 'score': 0.4935680031776428},{'label': 'positive', 'score': 0.5790663957595825}]
```


# Credits / Special Thanks:
This model was fine-tuned by Aaron Chibb. It is trained on [twitter dataset by tygiangz](https://huggingface.co/datasets/tyqiangz/multilingual-sentiments) and based on gBERT-large by [deepset](https://huggingface.co/deepset/gbert-large).