DiamondH
DiamondH
One correction, `labels = torch.cat([torch.arange(features.shape[0]//2) for i in range(self.args.n_views)], dim=0)`--> `labels = torch.cat([torch.arange(features.shape[0]//self.args.n_views) for i in range(self.args.n_views)], dim=0)`
`assert args.n_views == 2, "Only two view training is supported. Please use --n-views 2."` can be found in run.py, this works for 2_views, if we wanna try more, we can...
I'm also look forward to that day, can't wait to see your great work!!👍🏻👍🏻👍🏻
> 本来想测试一下baseline，但是发现Unet量化结果奇高啊 是的，Unet就是最高的