e-lab / torch-PredNetNew

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

I got following error while I was trying to train with your script.

spk921 opened this issue · comments

Loaded 2000 image sequences
Image resolution: 64 x 64

of parameters 1011768

TRAINING
@codeAC29

I was trying to run your code but encounter errors. Thank you.

th main.lua --datapath ./dataset/data-big-test.t7 --saveGraph --layers 3 --seq 20 --save /ly3 --devID 1

Epoch: 1
/home/elab/torch/install/bin/luajit: /home/elab/torch/install/share/lua/5.1/nn/CSubTable.lua:11: bad argument #2 to 'add' (sizes do not match at /tmp/luarock
s_cutorch-scm-1-8073/cutorch/lib/THC/generated/../generic/THCTensorMathPointwise.cu:111)
stack traceback:
[C]: in function 'add'
/home/elab/torch/install/share/lua/5.1/nn/CSubTable.lua:11: in function 'func'
/home/elab/torch/install/share/lua/5.1/nngraph/gmodule.lua:345: in function 'neteval'
/home/elab/torch/install/share/lua/5.1/nngraph/gmodule.lua:380: in function 'func'
/home/elab/torch/install/share/lua/5.1/nngraph/gmodule.lua:345: in function 'neteval'
/home/elab/torch/install/share/lua/5.1/nngraph/gmodule.lua:380: in function 'func'
/home/elab/torch/install/share/lua/5.1/nngraph/gmodule.lua:345: in function 'neteval'
/home/elab/torch/install/share/lua/5.1/nngraph/gmodule.lua:380: in function 'forward'
./train.lua:126: in function 'opfunc'
/home/elab/torch/install/share/lua/5.1/optim/adam.lua:35: in function 'adam'
./train.lua:165: in function 'updateModel'
main.lua:53: in main chunk
[C]: in function 'dofile'
...elab/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x00406670

Just E-mail Abhi when he is back. Closing without solve.

In case of MNIST your input channel for A of first layer will be 1 instead of 3. You will need to set value of opt.channel[1]=1 in main.lua

@spk921 did it work?