How to call it through transformer
#4
by
awelker
- opened
Could you describe the usage prompts and how to inject image and text stop and start words, maybe even a e.g. hugging face transformer call?
Thanks in advance.
Hi @awelker ! To run this model, please follow these 2 steps.
Step 1:
Install Nexa-SDK
https://github.com/NexaAI/nexa-sdk?tab=readme-ov-file#install-option-1-executable-installer
Step 2:
To use CLI, type in terminal: nexa run omnivision
To use local UI, type in terminal: nexa run omnivision -st
Here is a quick tutorial video on how to inject image and input prompts. You can drag a photo into your terminal and write prompt for image captioning and question answering tasks.
