main_test_rrdb.py

[House-Leo]

用main_train_rrdb_psnr.py训练出model之后，用test代码不能正常超分图片。 用原model = net(in_nc=3, out_nc=3, nc=64, nb=23, gc=32, upscale=4, act_mode='L', upsample_mode='upconv') 得出的输出会变黄色，后面按照option改成了 model = net(in_nc=3, out_nc=3, nc=64, nb=23, gc=32, upscale=4, act_mode='R', upsample_mode='upconv') 得出的输出会变黑色 训练过程中的验证是能正常超分出来的

[cszn]

model.load_state_dict(torch.load(model_path), strict=True)  # strict=False

再检查一下.

[House-Leo]

@cszn 改成True后，提示以下错误 Traceback (most recent call last): File "main_test_rrdb.py", line 162, in main() File "main_test_rrdb.py", line 62, in main model.load_state_dict(torch.load(model_path), strict=True) File "/public_files/software/envs/mmsr/lib/python3.7/site-packages/torch/nn/modules/module.py", line 847, in load_state_dict self.class.name, "\n\t".join(error_msgs))) RuntimeError: Error(s) in loading state_dict for RRDB: Missing key(s) in state_dict: "model.6.weight", "model.6.bias", "model.8.weight", "model.8.bias", "model.10.weight", "model.10.bias". Unexpected key(s) in state_dict: "model.5.weight", "model.5.bias", "model.7.weight", "model.7.bias".

[tongchangD]

请问你是如何看直观看了loss 损失的直观感受模型的 以及可视化调参的？ 难道是沿用作者大神的参数直接训练到底吗？