Viking 7B
Viking 7B is a base model pretrained on Finnish, English, Swedish, Danish, Norwegian, Icelandic and code. It uses a LLaMA-like architecture, not a claimed adaptation of Meta's released weights.
Model design and training
32 decoder layers. The card reports completion of two trillion training tokens. Viking uses rotary position embeddings and FlashAttention. The downloadable base is intended for further tuning. Training checkpoints and later packaging are grouped with the same size-specific release.Viking-7B: official model card
Practical assessment
Practical guidance. The smallest Viking release can serve as a baseline for Nordic-language adaptation. Compare terminology and spelling conventions separately for each language. Shared training does not establish equal performance across all six languages.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | Viking-7B |
| Context window | 4,096 tokens Sequence length reported by this release's model card; this is not a guaranteed useful retrieval length. Viking-7B: official model card |
| Model size | 7.55 billion parameters.Viking-7B: official model card |
| License | Apache 2.0Viking-7B: official model card |
Related models
- Viking 13B
LumiOpen · Viking
A 13B base model pretrained on Finnish, English, Swedish, Danish, Norwegian, Icelandic and code.
Launch date not confirmedDownloadable - Viking 33B
LumiOpen · Viking
A 33B base model pretrained on Finnish, English, Swedish, Danish, Norwegian, Icelandic and code.
Launch date not confirmedDownloadable