Skip to content

Averaging Optimization

Richard Townsend edited this page May 23, 2016 · 1 revision

Before

INFO:root:Epoch 0, Update 1880, Cost 0.7312
INFO:root:Epoch 0, Update 1890, Cost 0.7798
INFO:root:Epoch 0, Update 1900, Cost 0.8291
^CWARNING:root:Training interrupted
INFO:root:Train 21.1506, Valid 17.7939, Test 19.1229
INFO:root:Saving to lstm_model.npz...
INFO:root:The code run for 1 epochs, with 35.593483 sec/epochs
INFO:root:Training took 35.6s

After

INFO:root:Epoch 0, Update 1900, Cost 0.8291
INFO:root:Epoch 0, Update 1910, Cost 0.7017
INFO:root:Epoch 0, Update 1920, Cost 0.8256
INFO:root:Epoch 0, Update 1930, Cost 0.7227
INFO:root:Epoch 0, Update 1940, Cost 0.6970
^CWARNING:root:Training interrupted
INFO:root:Train 21.1546, Valid 18.1854, Test 18.8214
INFO:root:Train 21.1546, Valid 18.1854, Test 18.8214
INFO:root:Saving to lstm_model.npz...
INFO:root:The code run for 1 epochs, with 36.752027 sec/epochs
INFO:root:Training took 36.8s

Clone this wiki locally