search

chrischoy / 3D-R2N2

Single/multi view image(s) to voxel reconstruction using a recurrent neural network
MIT License
1.36k stars 294 forks source link

Compiling testing function doesn't work #11

Open kealennieh opened 7 years ago

kealennieh commented 7 years ago

Dear author: sorry to bother you. Here is my question. After I change the training dataset, the training process cannot operate successfully. It always loops in "Compiling testing function", as follows:

lib/data_io.py: model paths from ./experiments/dataset/shapenet_1000.json lib/data_io.py: model paths from ./experiments/dataset/shapenet_1000.json [INFO/ReconstructionDataProcess-1] child process calling self.run() Set the learning rate to 0.000100. [INFO/ReconstructionDataProcess-2] child process calling self.run() Compiling training function Exiting [INFO/ReconstructionDataProcess-2] process shutting down 2017-06-27 09:06:57.187543 Iter: 0 Loss: 0.502873 [INFO/ReconstructionDataProcess-2] process exiting with exitcode 0 Compiling testing function

I attempted a lot of times, but failed at the same place. I don't know the reason, since I just change the training dataset. Thanks in advance.

chrischoy commented 7 years ago

It is possible that the json file is empty or the process cannot load it properly and the network fails. Try to make an isolated test function like https://github.com/chrischoy/3D-R2N2/blob/master/lib/data_process.py#L216 and see if you can load the dataset properly first.