Sakana AIEvoVLM-JP
EvoVLM-JP-v1-7B
EvoVLM-JP-v1-7B combines a Japanese language checkpoint with LLaVA's visual capabilities through evolutionary model merging.
Research release
The release joins language and image capabilities
The card gives an image-and-text inference example and identifies Japanese Visual Genome visual question answering in its development account.EvoVLM-JP-v1-7B official model card
Working with this release
Practical guidance. Use images whose correct answer depends on a visible detail. Separate failures to read the image from failures to express the answer in Japanese.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Research release |
| Access | Downloadable |
| Model identifiers | SakanaAI/EvoVLM-JP-v1-7B |
| Ancestry | Shisa Gamma 7B v1 and LLaVA-1.6-Mistral-7B.EvoVLM-JP-v1-7B official model card |
| Terms | Apache 2.0; research prototype with commercial and critical deployment discouraged in the card.EvoVLM-JP-v1-7B official model card |
Related models
- Llama-3-EvoVLM-JP-v2
Sakana AI ยท EvoVLM-JP
A merged Llama 3 vision-language model for interleaved images and Japanese text.
Launch date not confirmedResearch releaseDownloadable