🚩 Report
not indicating that it's a qwen2-72b finetune in the frontpage.
The config shows that it is indeed qwen finetune but it's hidden from the FRONT page.
Please show in the medtada that it's a derivative of qwen2 as that greatly affects how developers work with the model/tokenizer/weights/formatting etc...
Hi @nisten ,
Thanks for the suggestion. We have made it more clear in the README here: https://huggingface.co/nvidia/NVLM-D-72B#how-to-use
Thanks,
Boxin
I noticed that this repository is licensed under CC-BY-NC, but I'm unsure if this is compatible with the Qwen research license used in Qwen 2.5. Qwen research license should require all derivatives to include a copy.
In addition, you have Apache 2.0 and MIT licenses distributed in a folder. I assume that's for the code but it's not clear and clarity would be appreciated.