language: nl
widget:
- text: Ik weet niet zeker of ik me beter voel
- text: >-
Ik denk dat het over het algemeen mijn vermogen heeft beïnvloed om te
vertrouwen op mijn vermogen om me niet gemarginaliseerd te voelen en ik
sta zeer wantrouwend tegenover kliekjes en autoriteit
- text: Ik voel me de hele tijd raar tenzij ik lig
- text: ik blog omdat het iets is waar ik een passie voor heb
pipeline_tag: text-classification
tags:
- text-classification
- RobBERT
- pytorch
- emotion
- latest
datasets:
- botdevringring/NL_emotion_classification
- dair-ai/emotion
metrics:
- Accuracy, F1 Score
model-index:
- name: botdevringring/nl-naxai-ai-emotion-classification-101608122023
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: emotion
type: emotion
config: default
split: test
metrics:
- type: accuracy
value: 0.896
name: Accuracy
verified: true
- type: precision
value: 0.897
name: Precision Macro
verified: true
- type: precision
value: 0.896
name: Precision Micro
verified: true
- type: precision
value: 0.897
name: Precision Weighted
verified: true
- type: recall
value: 0.896
name: Recall Macro
verified: true
- type: recall
value: 0.896
name: Recall Micro
verified: true
- type: recall
value: 0.896
name: Recall Weighted
verified: true
- type: f1
value: 0.895
name: F1 Macro
verified: true
- type: f1
value: 0.895
name: F1 Micro
verified: true
- type: f1
value: 0.895
name: F1 Weighted
verified: true
botdevringring/nl-naxai-ai-emotion-classification-101608122023(latest)
The model is trained on the emotion classification task in Dutch. It uses 6 labels: anger, fear, joy love, sadness, surprise.
This model is finetuned from robbert-v2-dutch-base. RobBERT is the state-of-the-art Dutch BERT model. It is a large pre-trained general Dutch language model that can be fine-tuned on a given dataset to perform any text classification, regression or token-tagging task.
Model Details
- Language: nl
- Problem type: Multi-class Classification
- Model Architecture: RobBERT
- Model Name: nl-naxai-ai-emotion-classification-101608122023
- Creation Date: 10:16h 08/12/2023
Dataset Summary
The botdevringring/NL_emotion_classification dataset is a balanced dataset translated from the English Twitter messages Emotion dataset of with six basic emotions: anger, fear, joy, love, sadness, and surprise. For more detailed information please refer to the paper.
Classification Report:
Label | Precision | Recall | f1-Score | Support |
---|---|---|---|---|
anger | 0.91 | 0.91 | 0.91 | 2956 |
fear | 0.89 | 0.83 | 0.86 | 2956 |
joy | 0.89 | 0.83 | 0.86 | 2956 |
love | 0.88 | 0.91 | 0.89 | 2956 |
sadness | 0.91 | 0.89 | 0.90 | 2956 |
surprise | 0.87 | 0.97 | 0.92 | 2956 |
How to use this model
You can use Python to access this model:
from transformers import pipeline
analyzer = pipeline(
task='text-classification',
model='botdevringring/nl-naxai-ai-emotion-classification-101608122023',
tokenizer='botdevringring/nl-naxai-ai-emotion-classification-101608122023'
)
result = analyzer(
"ik blog omdat het iets is waar ik een passie voor heb."
)
result
[
{
'label': 'love',
'score': 0.9728550314903259
}
]
Or you can use cURL:
curl https://api-inference.huggingface.co/models/botdevringring/nl-naxai-ai-emotion-classification-101608122023 \
-X POST \
-d '{"inputs": "ik blog omdat het iets is waar ik een passie voor heb"}' \
-H 'Content-Type: application/json' \
-H "Authorization: Bearer <Your HF API token>"
Acknowledgements
Model trained by Eduardo Brigham for Naxai powered by The Ring Ring Company