License Clarification
Hi, tested out your model, it's great! Given this, I'd like to clarify what the license is for this model. It seems you have listed the model license as apache-2.0, but both this repo and your blog post clearly state that this is a fine-tune/derivative model of Qwen/Qwen2.5-VL-3B-Instruct which has the following license: https://hf.co/Qwen/Qwen2.5-VL-3B-Instruct/blob/main/LICENSE
It appears that derivative works are restricted to non-commercial use only. Could you help me understand the correct licensing terms for this model? Thanks!
I am also interested in a response here -- afaik only Qwen 2.5 7b has an open license and the 3b model does not
@pszemraj You're right. The Apache-2.0 license was mistakenly listed, and I apologize for the confusion. Since it's a derivative of Qwen-2.5-VL-3B, it will have the same license as the base model (Qwen RESEARCH LICENSE AGREEMENT). Thanks for pointing this out.
@Souvik3333 Any chance you would run the exact same fine-tuning set on https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct which has an apache 2.0 license? I tested Nanonets-OCR this morning and it is awesome but theres no way I can use it based off the 3B qwen 2.5. Would absolutely a use a 7b fine tune! cc @pszemraj
Actually we are already doing some experimentations based on the 7b model. Will keep you posted as we make progress on that. Feel free to share any improvement suggestions for the 3b model, we will try to address them when we release the next model.
Great to hear that you’re already looking at the 7b! Looking forward to it.
Btw @Souvik3333 feel free to close this out or leave it up for other people to see why the license is what it is