Bilingual GPT-NeoX 4B MiniGPT4
This visual-language release joins rinna's 3.8B bilingual GPT-NeoX model with BLIP-2 components and a projection layer.
Research release
The training stages differ from original MiniGPT-4
Training uses CC12M, COCO, Visual Genome, STAIR captions and Japanese Visual Genome VQA. rinna runs MiniGPT-4's first pretraining stage, but not the second stage with ChatGPT-created aligned image-text data.Bilingual GPT-NeoX 4B MiniGPT4 official model card
Check image grounding before fluent dialogue
Practical guidance. Ask questions whose answers are visible in the image and include images with missing requested details. A fluent bilingual response does not establish that the visual components supplied the answer.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | bilingual-gpt-neox-4b-minigpt4 |
| Scale | 3.8B language-model parameters; the 4B name is rounded.Bilingual GPT-NeoX 4B MiniGPT4 official model card |
| License | MIT for this published release.Bilingual GPT-NeoX 4B MiniGPT4 official model card |
Related models
- Bilingual GPT-NeoX 4B
rinna · Bilingual GPT-NeoX
An English-Japanese GPT-NeoX language model with 3.8B parameters.
Launch date not confirmedResearch releaseDownloadable - Bilingual GPT-NeoX 4B SFT
rinna · Bilingual GPT-NeoX
The supervised instruction release adapts Bilingual GPT-NeoX 4B using English and Japanese instruction data.
Launch date not confirmedDownloadable