search

andreas128 / SRFlow

Official SRFlow training code: Super-Resolution using Normalizing Flow in PyTorch
Other
824 stars 112 forks source link

Is there something wrong in original code? #63

Open danchelz opened 10 months ago

danchelz commented 10 months ago

Skipping ERROR caught in nll = model.optimize_parameters(current_step): 23-09-04 05:05:13.371 - WARNING: Params [module.flowUpsamplerNet.layers.57.affine.fFeatures.4.weight] will not optimize. module.flowUpsamplerNet.layers.57.affine.fFeatures.4.bias True 23-09-04 05:05:13.371 - WARNING: Params [module.flowUpsamplerNet.layers.57.affine.fFeatures.4.bias] will not optimize. module.flowUpsamplerNet.layers.57.affine.fFeatures.4.logs True 23-09-04 05:05:13.371 - WARNING: Params [module.flowUpsamplerNet.layers.57.affine.fFeatures.4.logs] will not optimize.

And I don't know what caused this problem。