File size: 1,904 Bytes
110925c 7287889 110925c d30d440 110925c b0a3d86 110925c ca85174 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 |
---
pipeline_tag: zero-shot-classification
tags:
- zero-shot-classification
- swedish
- megatron-bert
language:
- sv
datasets:
- KBLab/overlim
widget:
- example_title: Zero-shot
text: Många skjuter upp sina tandläkarbesök
candidate_labels: hälsa, politik, sport, religion
inference:
parameters:
hypothesis_template: Detta exempel handlar om {}.
---
# Megatron-BERT-large Swedish 165k for zero-shot classification
This model is based on Megatron-BERT-large-165k (https://huggingface.co/KBLab/megatron-bert-large-swedish-cased-165k). It was fine-tuned on the QNLI task and further fine-tuned on the MNLI task.
The model can be used with the Hugging Face zero-shot classification pipeline.
You can read more about the model on our [blog](https://kb-labb.github.io/posts/2023-02-12-zero-shot-text-classification/).
## Usage
```python
>>> from transformers import pipeline
>>> classifier = pipeline(
... "zero-shot-classification",
... model="KBlab/megatron-bert-large-swedish-cased-165-zero-shot"
... )
>>> classifier(
... "Ruben Östlunds ”Triangle of sadness” nomineras till en Golden Globe i kategorin bästa musikal eller komedi.",
... candidate_labels=["hälsa", "politik", "sport", "religion", "nöje"],
... hypothesis_template="Detta exempel handlar om {}.",
... )
{'sequence': 'Ruben Östlunds ”Triangle of sadness” nomineras till en Golden Globe i kategorin bästa musikal eller komedi.',
'labels': ['nöje', 'sport', 'religion', 'hälsa', 'politik'],
'scores': [0.9274595379829407,
0.025105971843004227,
0.018440095707774162,
0.017049923539161682,
0.011944468133151531]}
```
## Citation
```
@misc{sikora2023swedish,
author = {Sikora, Justyna},
title = {The KBLab Blog: Swedish zero-shot classification model},
url = {https://kb-labb.github.io/posts/2023-02-12-zero-shot-text-classification/},
year = {2023}
}
```
|