rinnaJapanese GPT
Japanese GPT-2 Medium
A Japanese GPT-2 language model trained to predict continuations of text.
Research release
How this checkpoint was trained
The documented run used Japanese CC-100 and Wikipedia for about 30 days on eight V100 GPUs. The example explicitly enables lowercasing in the tokenizer.Japanese GPT-2 Medium official model card
Measure a completion task rather than a chat task
Practical guidance. This checkpoint can support adaptation or completion research. If the intended workflow is a conversation, instruction tuning and a conversation format are separate requirements.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | japanese-gpt2-medium |
| Architecture | 24 layers with hidden width 1,024Japanese GPT-2 Medium official model card |
| License | MIT.Japanese GPT-2 Medium 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