A large language model optimized for Japanese by NII, available in 8 sizes from 150M to 172B parameters. Trained on llm-jp-corpus v3 with SFT/DPO fine-tuning via Instruct2/Instruct3. Licensed under Apache 2.0 for models 13B and under.
Parameters
Undisclosed
Context Window
License
Proprietary
Release Date
2025-03-18
API Pricing
API pricing for this model is not yet available
Strengths
- •Mixture-of-Experts instruction model
- •8 experts with 1.8B each
- •Efficient parameter scaling
Weaknesses
- •Complex routing inference
- •Specialized instruction tuning
- •High total but sparse active parameters
Use Cases
- •Efficient instruction-tuned generation
- •Scalable conversational AI
- •MoE architecture research