Open WujiangXu opened 1 year ago
Hi, Sorry for the wait and thanks for the interest! I add to my backlog to release the pretrained weights, I hope to have them ready by the end of the week. I'll post new updates in this issue (:
Hi! I uploaded (some of) the models to this gdrive folder: https://drive.google.com/drive/folders/1KuuiyNJUa92rCUtv9JO6gyAJ0plpqxcS?usp=sharing
I hope to upload all the models to HF in the near future, I'll post updates in this issue.
Hi @dveni, though the ViT models are here, the outputs when passing images to this model are 1024-tensors. According to the code, I think there should also be an SGD regressor model that we pass the tensor through to get the aesthetic score. Can this be published? Thanks.
Hello, this work contributes a lot to the community and is impressive for me.
We utilize your provided code to extract the aesthetic feature combined with the language modality. But I have not found the pre-trained aesthetic model in this repository. Will you release the pre-trained model in the future?