LLM-jp-4 8B Base
LLM-jp-4 8B Base is the pretrained and mid-trained checkpoint for this size. It is the starting point for adaptation before instruction alignment.
Research release
Pretraining before instruction alignment
The training pipeline used 11.7 trillion tokens across pretraining and mid-training. The project releases much of the corpus, but excludes some material because of license restrictions. Use the supplied tokenizer; the format resembles Harmony, but its token IDs differ.LLM-jp-4 8B Base official model card
Choose the checkpoint for adaptation
Practical guidance. This is the smallest dense model in the generation. Compare answer quality and output length on the same Japanese tasks before choosing it for a local application.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | llm-jp-4-8b-base |
| Context window | 65,536 tokens Maximum sequence length stated in the model card. LLM-jp-4 8B Base official model card |
| Architecture | Dense; 32 layers, hidden size 4,096.LLM-jp-4 8B Base official model card |
| Parameters | 8,590,200,832 total parameters.LLM-jp-4 8B Base official model card |
| License | Apache 2.0; downloadable research weights.LLM-jp-4 8B Base official model card |
Related models
- LLM-jp-4.1 8B Thinking
LLM-jp / NII LLMC · LLM-jp-4.1
LLM-jp-4.1 8B Thinking is a Japanese-English reasoning release with new SFT and preference training.
Launch date Research releaseDownloadable - LLM-jp-4 8B Thinking
LLM-jp / NII LLMC · LLM-jp-4
LLM-jp-4 8B Thinking provides Japanese and English text reasoning weights with a 65,536-token context.
Launch date not confirmedResearch releaseDownloadable