Can you support ollama's new deepseek-ocr feature?

[qianjun2659]

The latest version of ollama supports the new features of deepseek-ocr. There is a slight change in the way to use it. Please help me see if it can be supported. Thank you.
Reference link：
https://github.com/ollama/ollama/releases/tag/v0.13.0
https://ollama.com/library/deepseek-ocr

[n4ze3m]

Hey, Seems like they're using the same Ollama API. Is the current version of Page Assist not working with deepseek OCR?

[qianjun2659]

Hey, Seems like they're using the same Ollama API. Is the current version of Page Assist not working with deepseek OCR?嘿，看起来他们用的是同一个 Ollama API。当前版本的 Page Assist 不能支持 deepseek OCR 吗？

When I was using it, I noticed that it often runs into the problem of infinite loop outputs during recognition. I'm not sure if it's an issue with the model or the plugin. I tested it today and found that the built-in chat feature in ollama seems to have the same problem. It appears to be unrelated to the plugin. Thanks, expert.

[amonpaike]

I'm not sure if it's a page assist problem. I'm using the LLM as Assistant and as executor tools in home assistant, the domotics manager, And often the models managed by ollama go into infinite loops, which doesn't happen with llama cpp.