LLM-jp(NII)Proprietary

llm-jp-3-440m-instruct3

Compare this model

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

2025-01-31

API Pricing

API pricing for this model is not yet available

Strengths

  • Strong Japanese language proficiency due to dedicated corpus training.
  • Available in a compact 440M parameter size for efficient deployment.
  • Instruction-tuned for following commands and dialogue.

Weaknesses

  • Performance may lag behind much larger models on complex reasoning tasks.
  • Limited to Japanese and may not generalize well to other languages.
  • Context window may be smaller than larger-scale models.

Use Cases

  • General-purpose Japanese chatbots and virtual assistants.
  • Running lightweight language models on resource-constrained devices.
  • Serving as a base for further fine-tuning on specific Japanese tasks.