(1) Generate the data-set of 40k triplets of images,
(2) Train the deep network, and output validation results every 100
- epochs. This takes 15h on a GTX 1080 with cuda 8.0, cudnn 5.1,
+ epochs. This takes ~30h on a GTX 1080 with cuda 8.0, cudnn 5.1,
and recent torch.
(3) Generate two pictures of the internal activations.
--
Francois Fleuret
-Nov 6, 2016
+Nov 24, 2016
Martigny
cmd:text('')
cmd:text('Training')
-cmd:option('-nbEpochs', 2000, 'nb of epochs for the heavy setting')
+cmd:option('-nbEpochs', 1000, 'nb of epochs for the heavy setting')
cmd:option('-learningRate', 0.1, 'learning rate')
cmd:option('-batchSize', 128, 'size of the mini-batches')
cmd:option('-nbTrainSamples', 32768)
--dir "${DYNCNN_DATA_DIR}"
fi
-# Train the model (takes 30h on a GTX 1080 with cuda 8.0, cudnn 5.1,
-# and recent torch)
+# Train the model (2000 epochs takes 30h on a GTX 1080 with cuda 8.0,
+# cudnn 5.1, and recent torch)
if [[ ! -f "${DYNCNN_RUNDIR}"/scheme_02000.t7 ]]; then
./dyncnn.lua -rundir "${DYNCNN_RUNDIR}"