FontDiffuser
FontDiffuser copied to clipboard
[AAAI2024] FontDiffuser: One-Shot Font Generation via Denoising Diffusion with Multi-Scale Content Aggregation and Style Contrastive Learning
when ”Combined with InstructPix2Pix “ releasing ?
Thank you very much for your work, I am very interested in this, but I would like to ask whether there are experiments to prove that this model can be...
训练出错
在训练的时候默认开启了数据并行导致保存DistributedDataParallel,模型model没有unt这个参数导致出错,如何从并行模型提取需要模型参数
用Git Bash 2.46.0.windows.1运行**D:\FontDiffuser-main\train_phase_1.sh**的时候报出如下错误: > l257737602 MINGW64 /d/FontDiffuser-main > $ sh train_phase_1.sh > D:\Users\limin\AppData\Local\Programs\Python\Python312\Lib\site-packages\kornia\feature\lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead. > @torch.cuda.amp.custom_fwd(cast_inputs=torch.float32) > D:\Users\limin\AppData\Local\Programs\Python\Python312\Lib\site-packages\accelerate\accelerator.py:406: UserWarning: `log_with=tensorboard` was passed...
https://github.com/yeungchenwa/FontDiffuser?tab=readme-ov-file#training---phase-2 > **After the phase 2 training**, you should put the trained checkpoint files (unet.pth, content_encoder.pth, and style_encoder.pth) to the directory phase_1_ckpt. During phase 2, these parameters will be resumed....
Hello @yeungchenwa , Thank you for your excellent work on the FontDiffuser project! I am very curious about how you combined FontDiffuser and InstructPix2Pix. I have tried different text prompts...
First, thanks for open soucing your amazing work. Two weeks ago, I commented below an [existing issue](https://github.com/yeungchenwa/FontDiffuser/issues/40#issuecomment-2235240970) requesting for the script, but you might have missed my comment. It's been...
InfoNCE
您好,我想询问一下关于第二阶段的对比损失InfoNCE具体代码是怎么样的,感谢
关于数据集的问题
你好,你们的工作很棒,但是我作为一个初学者对于这个领域的了解没有很深,尤其是这个领域现存有哪些开源数据集可以使用。请问你们的数据集是怎么、从何构建的呢,可以给一个可供参考的开源数据集的链接吗?非常感谢
作者大佬们好,非常感谢你们的工作! 关于训练我有两个问题。 1、假如说有草书、楷书两个主要风格的笔记训练,我将他们混合到一起进行训练效果好,还是说进行分类后再单独生产效果好呢? 2、训练时loss值有推荐的区间不