metadata

pipeline_tag: zero-shot-classification
language:
  - da
  - 'no'
  - nb
  - sv
license: mit
datasets:
  - strombergnlp/danfever
  - mnli_da
  - mnli_sv
  - mnli_nb
  - cb_da
  - cb_sv
  - cb_nb
  - fever_sv
  - anli_sv
model-index:
  - name: nb-bert-large-ner-scandi
    results: []
widget:
  - example_title: Nyhetsartikkel om FHI
    text: >-
      Folkehelseinstituttets mest optimistiske anslag er at alle voksne er
      ferdigvaksinert innen midten av september.
    candidate_labels: helse, politikk, sport, religion

ScandiNLI - Natural Language Inference model for Scandinavian Languages

This model is a fine-tuned version of NbAiLab/nb-bert-large for Natural Language Inference in Danish, Norwegian Bokmål and Swedish.

It has been fine-tuned on a dataset composed of DanFEVER as well as machine translated versions of MultiNLI and CommitmentBank into all three languages, and machine translated versions of FEVER and Adversarial NLI into Swedish.

The three languages are sampled equally during training, and they're validated on validation splits of DanFEVER and machine translated versions of MultiNLI for Swedish and Norwegian Bokmål, sampled equally.

Quick start

You can use this model in your scripts as follows:

>>> from transformers import pipeline
>>> classifier = pipeline("zero-shot-classification", model="alexandrainst/nb-bert-large-nli-scandi")
>>> classifier(
...     'Folkehelseinstituttets mest optimistiske anslag er at alle over 18 år er ferdigvaksinert innen midten av september.',
...     candidate_labels=['helse', 'politikk', 'sport', 'religion'],
...     hypothesis_template="Dette eksempelet er {}",
)
{
    'labels': ['helse', 'politikk', 'sport', 'religion'], 
    'scores': [0.4210019111633301, 0.0674605593085289, 0.000840459018945694, 0.0007541406666859984],
    'sequence': 'Folkehelseinstituttets mest optimistiske anslag er at alle over 18 år er ferdigvaksinert innen midten av september.',
}

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 2
eval_batch_size: 2
seed: 4242
gradient_accumulation_steps: 16
total_train_batch_size: 32
optimizer: Adam with betas=(0.9, 0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
max_steps: 50,000