Resume from previous checkpoint

[aramic]

Is there a way to do this?

I tried the code suggested in #62, but torch.load([checkpoint file]) throws errors:

Initializing discriminator network from /datadisk/ml/dcgan.torch/checkpoints/test_4415_net_G.t7 /home/username/torch/install/bin/luajit: ...e/username/torch/install/share/lua/5.1/torch/File.lua:343: unknown Torch class <torch.CudaTensor> stack traceback: [C]: in function 'error' ...e/username/torch/install/share/lua/5.1/torch/File.lua:343: in function 'readObject' ...e/username/torch/install/share/lua/5.1/torch/File.lua:369: in function 'readObject' /home/username/torch/install/share/lua/5.1/nn/Module.lua:192: in function 'read' ...e/username/torch/install/share/lua/5.1/torch/File.lua:351: in function 'readObject' ...e/username/torch/install/share/lua/5.1/torch/File.lua:409: in function 'load' main.lua:68: in main chunk [C]: in function 'dofile' ...name/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk [C]: at 0x00405d50

[Quasimondo]

You could try to put

require 'cunn'
require 'cudnn'

before the torch.load() line.

[maryam089]

Did you solve that problem of resuming training from a specific epoch ? If yes then kindly help me how to do it. I couldn’t understand the solution #62