norbert3-base
| Property | Value |
|---|---|
| Parameter Count | 123M parameters |
| Model Type | BERT-based Masked Language Model |
| License | Apache 2.0 |
| Paper | NorBench — A Benchmark for Norwegian Language Models |
What is norbert3-base?
norbert3-base is a powerful Norwegian language model that's part of the NorBERT 3 family, specifically designed to handle both Norwegian Bokmål and Nynorsk. With 123M parameters, it represents a balanced compromise between model size and performance, positioning itself as the standard variant in the NorBERT 3 series.
Implementation Details
The model requires custom code implementation and must be loaded with trust_remote_code=True. It supports multiple NLP tasks through various interfaces including AutoModel, AutoModelMaskedLM, AutoModelForSequenceClassification, AutoModelForTokenClassification, AutoModelForQuestionAnswering, and AutoModelForMultipleChoice.
- Custom wrapper requirement from modeling_norbert.py
- Built on PyTorch framework
- Supports masked language modeling as primary task
- Implements BERT architecture with Norwegian language specificity
Core Capabilities
- Masked language modeling for Norwegian text
- Sequence classification tasks
- Token classification capabilities
- Question answering functionality
- Multiple choice task handling
- Supports both Bokmål and Nynorsk variants
Frequently Asked Questions
Q: What makes this model unique?
This model is specifically trained for Norwegian language understanding, supporting both major written forms (Bokmål and Nynorsk). It's part of a comprehensive benchmark study (NorBench) and comes with multiple size variants for different use cases.
Q: What are the recommended use cases?
The model is ideal for Norwegian language processing tasks including text classification, named entity recognition, question answering, and general language understanding tasks. It's particularly useful for applications requiring deep understanding of Norwegian language nuances.
