Zj-BinXia
inference_diffir.py 参数model_path问题
您好，请问对于SRGAN 的inference_diffir --model_path 参数，我下载了三个权重文件：SISR-DiffIRS1.pth，SISR-DiffIRS2.pth，SISR-DiffIRS2-GAN.pth，无论选择哪个都会报错：
anaconda3/envs/dfrsc/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1667, in load_state_dict raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for DiffIRS2: size mismatch for G.patch_embed.proj.weight: copying a param with shape torch.Size([64, 3, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 12, 3, 3]). size mismatch for condition.E.0.weight: copying a param with shape torch.Size([64, 3, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 12, 3, 3]). size mismatch for diffusion.condition.E.0.weight: copying a param with shape torch.Size([64, 3, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 12, 3, 3]).
请问除了使用test.sh外，进行无gt的测试时，如何选择权重文件作为--model_path的参数。谢谢！
Yes, I check the scale arguments is 2, :
CUDA_VISIBLE_DEVICES=1 python inference_diffir.py --scale 2 --model_path ./experiments/SISR-DiffIRS2-GAN.pth,
but still not works:
Traceback (most recent call last): File "/home/ps/PycharmProjects/DiffIR/DiffIR-SRGAN/inference_diffir.py", line 46, in <module> model.load_state_dict(loadnet['params_ema'], strict=True) File "/home/ps/anaconda3/envs/dfrsc/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1667, in load_state_dict raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for DiffIRS2: size mismatch for G.patch_embed.proj.weight: copying a param with shape torch.Size([64, 3, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 12, 3, 3]). size mismatch for condition.E.0.weight: copying a param with shape torch.Size([64, 3, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 12, 3, 3]). size mismatch for diffusion.condition.E.0.weight: copying a param with shape torch.Size([64, 3, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 12, 3, 3])..
But I try to set scale to 4, it works. Thank you for your kindly reply!