π© Report: Spam
This is spam. Spam account too
@lunarflu The model's entire description is a contradiction, it's tagged for Text-to-Speech, but the model tree shows it's a finetune of an Image Editing model (Qwen/Qwen-Image-Edit), supposedly trained on FinePdfs. None of this makes any technical sense
hmm, it was uploaded recently and it's just one model, I think it's okay to allow repo authors to experiment/have placeholders/try out features if the raw amount isn't too high
@lunarflu
I understand allowing for experimentation, but the combination of elements here is technically impossible, because it claims to be a Text-to-Speech model, but its a finetune of an Image Editing model, supposedly trained on a PDF dataset. To make it even more nonsensical, a banner at the top claims a newer version is a DeepSeek text model.
This isn't a user experimenting with a model. At best, they are stress-testing the UI by creating an incoherent page, which still amounts to spam.