A Japanese-optimized large language model developed by NII. It is one of eight model sizes ranging from 150M to 172B parameters. Trained on the llm-jp-corpus v3 dataset, with SFT and DPO applied via Instruct2/Instruct3. Models under 13B are released under Apache 2.0 license.
Parameters
Undisclosed
Context Window
License
Proprietary
Release Date
2024-12-16
API Pricing
API pricing for this model is not yet available
Strengths
- •172B parameters represent the maximum capacity in the llm-jp-3 family.
- •Achieves top-tier results on Japanese language benchmarks.
- •Trained on an extensive and refined Japanese corpus.
Weaknesses
- •Impractical for most commercial deployment due to extreme resource needs.
- •Very slow inference unless using optimized, distributed hardware.
- •Potential for high computational cost per query.