49xxy

Results 20 comments of 49xxy

> Hi, thank you provide a code for Dual Contrastive Loss, and is it useful for performance on FFHQ dataset?? > > and did you add self-attention and reference-attention in...

> 大家好。我遇到了错误说 “RuntimeError:预计在开始新的迭代之前已经完成了先前迭代中的缩减。这个错误表明您的模块具有未用于产生损失的参数。您可以通过(1)传递启用未使用的参数检测; (2) 确保所有函数输出`find_unused_parameters=True`都参与计算损失。如果您已经完成上述两个步骤,那么分布式数据并行模块无法在模块的返回值中定位输出张量功能。请在报告此问题时包括您的模块的损失函数和返回值的结构(例如列表、字典、可迭代)。`torch.nn.parallel.DistributedDataParallel``forward``forward``forward` 当我尝试使用 DDP 应用对比学习损失时。 那么有什么想法吗?TAT Hi,may I ask if you have successfully used the loss?

> 默认增强模式应该是“ada”(我使用的是 stylegan2 pytorch 版本)但似乎总是设置为“noaug”,无论我通过哪个 --aug 值 如果我运行确实: `python train.py -outdir=clusterresults/ --data=/mydata/ --gpus 4 --workers 8 --augpipe bg` 我收到以下错误: `Error: --augpipe cannot be specified with --aug=noaug` > > 如果我尝试指定 --aug...

> 我认为您使用了 stylegan2-ada-pytorch 中的检查点。目前`convert_weight.py`不支持 pytorch 版本的检查点。 Can you support it now,please?

> > 我使用 google colab 从使用 stylegan2-ADA-PyTorch 和所有默认设置的自定义数据集训练的模型进行转换,并收到此错误消息: Traceback (last recent call last): File "/content/stylegan2-pytorch/convert_weight.py ",第 236 行,在生成器、鉴别器、g_ema = pickle.load(f) ModuleNotFoundError: No module named 'torch_utils' > > 在我 pip install...

> Is it possible to convert stylegan2-ada-PyTorch pkl model,please?

Hello, from the beginning of training to 1200kimg, the generation quality has hardly improved and the generator has been affected by data enhancement in the middle of the process. ![image](https://user-images.githubusercontent.com/102204219/195740583-9e3d235f-a672-4948-a596-b934adfcf794.png)...

Hello! 1)This is some data from my previous experiment: ![image](https://user-images.githubusercontent.com/102204219/196182967-3eee7a39-b48c-4165-b0ef-b253cd672c2d.png) 2)I am having this problem while running with diffaugment: ![image](https://user-images.githubusercontent.com/102204219/196179269-a57dd414-fbee-45ab-92d5-6ac44bb0c79e.png) Thanks!

> ![图片](https://github.com/assets/8830319/7ff98c27-7c4e-4f1d-bee1-792afc71cb35) 我有这个带有默认参数和 300 个 epoch 的“带有方形轮子的公共汽车”。没有什么真正改变。 May I ask whether your problem has been solved? I have also been studying the combination of Clip and non-face generation model recently.

> please add colab for inference Hi,have you found any other code for global directions using PyTorch?I have a Stylegan2-Ada-Pytorch model that I have trained with my own data,however, problems...