We ran the benchmark of DSSTNE using the Movielens data set. Training Parameter Specific
- 27278 input/output dimensions
- 3 Hidden Sigmoid layers with 1024 each
- RMSProp learner
- 256 Batchsize
Time taken to run one epoch is considered for performance comparison
Use the Config at config.json. Follow the example and change the training command
train -i gl_input.nc -o gl_output.nc -d gl -c config.json -b 256 -e 20 -n gl_network.nc
autoencoder.py -u 1024 -b 256 -i 1082 -v54 --vocab_size 27278 -l 3 -f /input/data/ml20m-all.remotcc