Loading config ...
Loading data ...  
Building vocab ...
Creating iterator ...
Building dataset ...
Building vocab from dataset ...
Load vocab from path successful
Building encoder and decoder ...
src vocab size = 29393
trg vocab size = 15202                            
Encoder: 34065920 parameters                      
Decoder: 40908642 parameters                      
Starting training on cuda                         
Performing training...                            
==================================================
Epoch: 01 - 15.0m40.41644310951233s
        Train Loss/PPL:   9.094 / 8899.793
        Val   Loss/PPL:   8.620 / 5540.010
--------------------------------------------------
Epoch: 02 - 15.0m38.90395212173462s
        Train Loss/PPL:   8.185 / 3586.749
        Val   Loss/PPL:   8.069 / 3193.848
--------------------------------------------------
Epoch: 03 - 15.0m39.11497640609741s
        Train Loss/PPL:   7.795 / 2427.533
        Val   Loss/PPL:   7.912 / 2729.002
--------------------------------------------------
Epoch: 04 - 15.0m42.52194285392761s
        Train Loss/PPL:   7.661 / 2123.347
        Val   Loss/PPL:   7.859 / 2589.700
--------------------------------------------------
Epoch: 05 - 15.0m42.61946368217468s
        Train Loss/PPL:   7.604 / 2005.850
        Val   Loss/PPL:   7.837 / 2532.609
--------------------------------------------------
Epoch: 06 - 15.0m40.5325984954834s
        Train Loss/PPL:   7.570 / 1938.907
        Val   Loss/PPL:   7.822 / 2493.998
--------------------------------------------------
Epoch: 07 - 15.0m44.441715240478516s
        Train Loss/PPL:   7.546 / 1893.262
        Val   Loss/PPL:   7.812 / 2469.149
--------------------------------------------------
Epoch: 08 - 15.0m43.27636504173279s
        Train Loss/PPL:   7.525 / 1854.688
        Val   Loss/PPL:   7.800 / 2441.054
--------------------------------------------------
Epoch: 09 - 17.0m49.64024472236633s
        Train Loss/PPL:   7.509 / 1823.568
        Val   Loss/PPL:   7.790 / 2415.858
--------------------------------------------------
Epoch: 10 - 15.0m41.81872010231018s
        Train Loss/PPL:   7.492 / 1793.774
        Val   Loss/PPL:   7.780 / 2391.125
--------------------------------------------------
Epoch: 11 - 28.0m3.3641841411590576s
        Train Loss/PPL:   7.477 / 1767.388
        Val   Loss/PPL:   7.772 / 2373.962
--------------------------------------------------
Epoch: 12 - 15.0m45.12012314796448s
        Train Loss/PPL:   7.463 / 1742.621
        Val   Loss/PPL:   7.763 / 2350.974
--------------------------------------------------
Epoch: 13 - 15.0m42.93015956878662s
        Train Loss/PPL:   7.449 / 1718.568
        Val   Loss/PPL:   7.756 / 2335.491
--------------------------------------------------
Epoch: 14 - 15.0m44.00054144859314s
        Train Loss/PPL:   7.438 / 1699.215
        Val   Loss/PPL:   7.748 / 2317.051
--------------------------------------------------
Epoch: 15 - 15.0m55.463807582855225s
        Train Loss/PPL:   7.426 / 1679.351
        Val   Loss/PPL:   7.741 / 2301.697
--------------------------------------------------
Epoch: 16 - 15.0m44.77303099632263s
        Train Loss/PPL:   7.415 / 1660.209
        Val   Loss/PPL:   7.733 / 2282.415
--------------------------------------------------
Epoch: 17 - 15.0m44.32082152366638s
        Train Loss/PPL:   7.405 / 1643.373
        Val   Loss/PPL:   7.726 / 2266.476
--------------------------------------------------
Epoch: 18 - 15.0m43.58943033218384s
        Train Loss/PPL:   7.395 / 1627.018
        Val   Loss/PPL:   7.719 / 2250.929
--------------------------------------------------
Epoch: 19 - 15.0m45.2637825012207s
        Train Loss/PPL:   7.386 / 1613.311
        Val   Loss/PPL:   7.712 / 2234.829
--------------------------------------------------
Epoch: 20 - 15.0m43.84825682640076s
        Train Loss/PPL:   7.376 / 1597.986
        Val   Loss/PPL:   7.706 / 2220.886
--------------------------------------------------