rinnaJapanese GPT
Japanese GPT-2 XSmall
A compact Japanese text-completion model trained on Japanese CC-100 and Wikipedia.
Research release
How this checkpoint was trained
The card records about four days of training on eight V100 GPUs. Its SentencePiece vocabulary was trained on Japanese Wikipedia.Japanese GPT-2 XSmall official model card
Use it to inspect tokenization and short continuations
Practical guidance. A small completion model is useful for controlled language experiments. Check unfamiliar names and mixed scripts separately; low memory use does not establish reliable instruction following.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | japanese-gpt2-xsmall |
| Architecture | 6 layers with hidden width 512Japanese GPT-2 XSmall official model card |
| License | MIT.Japanese GPT-2 XSmall official model card |
Related models
- Japanese GPT-2 Medium
rinna · Japanese GPT
A Japanese GPT-2 language model trained to predict continuations of text.
Launch date not confirmedResearch releaseDownloadable - 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