LLM-jp-3 13B Instruct
LLM-jp-3 13B Instruct is the 13B original instruction-tuned release in LLM-jp's generation-3 size series. The underlying model was exposed to 2.1 trillion pretraining tokens.
Research release
The original instruction data mixture
The card lists Japanese ichikara and AnswerCarefully data alongside synthetic conversation, coding and English FLAN material. This original Instruct checkpoint remains distinct from the later Instruct2 and Instruct3 releases.LLM-jp-3 13B Instruct official model card
Record the exact Instruct revision
Practical guidance. Keep this original 13B revision separate in results. Reusing an Instruct2 prompt or template without checking it can obscure the cause of a response change.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | llm-jp-3-13b-instruct |
| Context window | 4,096 tokens Maximum sequence length stated in the model card. LLM-jp-3 13B Instruct official model card |
| Architecture | 40 layers; hidden size 5,120; 40 attention heads.LLM-jp-3 13B Instruct official model card |
| Parameters | 13,707,924,480 total, from the card's embedding and non-embedding counts.LLM-jp-3 13B Instruct official model card |
| License | Apache 2.0; research weights.LLM-jp-3 13B Instruct official model card |
Related models
- LLM-jp-3 13B Base
LLM-jp / NII LLMC ยท LLM-jp-3
LLM-jp-3 13B Base is a pretrained text model with 4,096-token context and Apache 2.0 weights.
Launch date not confirmedResearch releaseDownloadable