rinnaJapanese GPT
Japanese GPT-NeoX Small
A Japanese GPT-NeoX model trained with EleutherAI's training code.
Research release
How this checkpoint was trained
Its training mix adds Japanese C4 to CC-100 and Wikipedia. The repository also contains a demonstration prefix that encourages a smiling emoji at the end of each generated sentence.Japanese GPT-NeoX Small official model card
Separate the base model from the learned prefix
Practical guidance. The emoji behavior comes from a small task-specific prefix. Keep that attachment separate from the base weights when comparing normal text completion.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | japanese-gpt-neox-small |
| Architecture | 12 layers with hidden width 768Japanese GPT-NeoX Small official model card |
| License | MIT.Japanese GPT-NeoX Small official model card |
Related models
- Japanese GPT-2 Small
rinna · Japanese GPT
A Japanese GPT-2 completion model with a larger decoder than the XSmall release.
Launch date not confirmedResearch releaseDownloadable - Japanese GPT-2 XSmall
rinna · Japanese GPT
A compact Japanese text-completion model trained on Japanese CC-100 and Wikipedia.
Launch date not confirmedResearch releaseDownloadable