A large language model optimized for Japanese, developed by NII. It comes in 8 sizes ranging from 150M to 172B parameters and was trained on llm-jp-corpus v3. It supports SFT/DPO through Instruct2/Instruct3 versions and is available under the Apache 2.0 license (for 13B and smaller).
Parameters
Undisclosed
Context Window
License
Proprietary
Release Date
2024-09-23
API Pricing
API pricing for this model is not yet available
Strengths
- •Balanced size for general tasks.
- •Japanese language training data.
Weaknesses
- •May struggle with very complex reasoning.
- •Lower performance than larger models.
Use Cases
- •General Japanese text generation.
- •Supporting light commercial applications.