training_log.txt

Epoch [1/3], Step [1/3236], Loss: 9.2074, Perplexity: 9971.0991
Epoch [1/3], Step [2/3236], Loss: 9.0559, Perplexity: 8569.1387
Epoch [1/3], Step [3/3236], Loss: 8.9044, Perplexity: 7364.1709
Epoch [1/3], Step [4/3236], Loss: 8.7128, Perplexity: 6080.0058
Epoch [1/3], Step [5/3236], Loss: 8.2568, Perplexity: 3853.7669
Epoch [1/3], Step [6/3236], Loss: 7.8992, Perplexity: 2695.2242
Epoch [1/3], Step [7/3236], Loss: 7.0748, Perplexity: 1181.7519
Epoch [1/3], Step [8/3236], Loss: 6.2328, Perplexity: 509.1559
Epoch [1/3], Step [9/3236], Loss: 5.6064, Perplexity: 272.1659
Epoch [1/3], Step [10/3236], Loss: 5.3177, Perplexity: 203.9087
Epoch [1/3], Step [11/3236], Loss: 5.2345, Perplexity: 187.6309
Epoch [1/3], Step [12/3236], Loss: 5.1802, Perplexity: 177.7202
Epoch [1/3], Step [13/3236], Loss: 5.0887, Perplexity: 162.1813
Epoch [1/3], Step [14/3236], Loss: 5.2100, Perplexity: 183.0936
Epoch [1/3], Step [15/3236], Loss: 5.4153, Perplexity: 224.8217
Epoch [1/3], Step [16/3236], Loss: 4.7276, Perplexity: 113.0278
Epoch [1/3], Step [17/3236], Loss: 4.8178, Perplexity: 123.6982
Epoch [1/3], Step [18/3236], Loss: 4.8233, Perplexity: 124.3742
Epoch [1/3], Step [19/3236], Loss: 4.8232, Perplexity: 124.3645
Epoch [1/3], Step [20/3236], Loss: 4.8637, Perplexity: 129.4985
Epoch [1/3], Step [21/3236], Loss: 4.7208, Perplexity: 112.2584
Epoch [1/3], Step [22/3236], Loss: 4.6124, Perplexity: 100.7260
Epoch [1/3], Step [23/3236], Loss: 4.6654, Perplexity: 106.2063
Epoch [1/3], Step [24/3236], Loss: 4.6078, Perplexity: 100.2662
Epoch [1/3], Step [25/3236], Loss: 4.5653, Perplexity: 96.0932
Epoch [1/3], Step [26/3236], Loss: 4.7316, Perplexity: 113.4755
Epoch [1/3], Step [27/3236], Loss: 4.4951, Perplexity: 89.5772
Epoch [1/3], Step [28/3236], Loss: 4.5164, Perplexity: 91.5035
Epoch [1/3], Step [29/3236], Loss: 4.4536, Perplexity: 85.9335
Epoch [1/3], Step [30/3236], Loss: 4.5179, Perplexity: 91.6433
Epoch [1/3], Step [31/3236], Loss: 4.5320, Perplexity: 92.9416
Epoch [1/3], Step [32/3236], Loss: 4.4096, Perplexity: 82.2367
Epoch [1/3], Step [33/3236], Loss: 4.4790, Perplexity: 88.1504
Epoch [1/3], Step [34/3236], Loss: 4.6706, Perplexity: 106.7610
Epoch [1/3], Step [35/3236], Loss: 4.5446, Perplexity: 94.1210
Epoch [1/3], Step [36/3236], Loss: 4.3708, Perplexity: 79.1070
Epoch [1/3], Step [37/3236], Loss: 4.3754, Perplexity: 79.4742
Epoch [1/3], Step [38/3236], Loss: 4.2960, Perplexity: 73.4088
Epoch [1/3], Step [39/3236], Loss: 4.2328, Perplexity: 68.9099
Epoch [1/3], Step [40/3236], Loss: 4.3319, Perplexity: 76.0906
Epoch [1/3], Step [41/3236], Loss: 4.3038, Perplexity: 73.9774
Epoch [1/3], Step [42/3236], Loss: 4.2249, Perplexity: 68.3664
Epoch [1/3], Step [43/3236], Loss: 4.3781, Perplexity: 79.6839
Epoch [1/3], Step [44/3236], Loss: 4.5484, Perplexity: 94.4810
Epoch [1/3], Step [45/3236], Loss: 4.2087, Perplexity: 67.2711
Epoch [1/3], Step [46/3236], Loss: 4.1453, Perplexity: 63.1353
Epoch [1/3], Step [47/3236], Loss: 4.3169, Perplexity: 74.9541
Epoch [1/3], Step [48/3236], Loss: 4.1681, Perplexity: 64.5936
Epoch [1/3], Step [49/3236], Loss: 4.1522, Perplexity: 63.5760
Epoch [1/3], Step [50/3236], Loss: 4.0536, Perplexity: 57.6050
Epoch [1/3], Step [51/3236], Loss: 4.8521, Perplexity: 128.0112
Epoch [1/3], Step [52/3236], Loss: 4.1742, Perplexity: 64.9856
Epoch [1/3], Step [53/3236], Loss: 4.5702, Perplexity: 96.5590
Epoch [1/3], Step [54/3236], Loss: 4.1209, Perplexity: 61.6176
Epoch [1/3], Step [55/3236], Loss: 4.2278, Perplexity: 68.5692
Epoch [1/3], Step [56/3236], Loss: 4.0813, Perplexity: 59.2205
Epoch [1/3], Step [57/3236], Loss: 4.1281, Perplexity: 62.0580
Epoch [1/3], Step [58/3236], Loss: 4.0760, Perplexity: 58.9111
Epoch [1/3], Step [59/3236], Loss: 4.2013, Perplexity: 66.7716
Epoch [1/3], Step [60/3236], Loss: 4.0416, Perplexity: 56.9175
Epoch [1/3], Step [61/3236], Loss: 4.1470, Perplexity: 63.2451
Epoch [1/3], Step [62/3236], Loss: 4.0052, Perplexity: 54.8817
Epoch [1/3], Step [63/3236], Loss: 4.0595, Perplexity: 57.9436
Epoch [1/3], Step [64/3236], Loss: 3.9794, Perplexity: 53.4850
Epoch [1/3], Step [65/3236], Loss: 4.0664, Perplexity: 58.3451
Epoch [1/3], Step [66/3236], Loss: 4.8803, Perplexity: 131.6647
Epoch [1/3], Step [67/3236], Loss: 4.0321, Perplexity: 56.3818
Epoch [1/3], Step [68/3236], Loss: 4.3702, Perplexity: 79.0599
Epoch [1/3], Step [69/3236], Loss: 4.6115, Perplexity: 100.6348
Epoch [1/3], Step [70/3236], Loss: 4.0102, Perplexity: 55.1573
Epoch [1/3], Step [71/3236], Loss: 3.8959, Perplexity: 49.2004
Epoch [1/3], Step [72/3236], Loss: 4.0810, Perplexity: 59.2049
Epoch [1/3], Step [73/3236], Loss: 3.9111, Perplexity: 49.9520
Epoch [1/3], Step [74/3236], Loss: 4.0747, Perplexity: 58.8354
Epoch [1/3], Step [75/3236], Loss: 4.0023, Perplexity: 54.7221
Epoch [1/3], Step [76/3236], Loss: 4.3395, Perplexity: 76.6664
Epoch [1/3], Step [77/3236], Loss: 3.8838, Perplexity: 48.6062
Epoch [1/3], Step [78/3236], Loss: 4.4794, Perplexity: 88.1852
Epoch [1/3], Step [79/3236], Loss: 4.0486, Perplexity: 57.3186
Epoch [1/3], Step [80/3236], Loss: 3.9362, Perplexity: 51.2239
Epoch [1/3], Step [81/3236], Loss: 3.9327, Perplexity: 51.0431
Epoch [1/3], Step [82/3236], Loss: 3.9042, Perplexity: 49.6090
Epoch [1/3], Step [83/3236], Loss: 4.2714, Perplexity: 71.6198
Epoch [1/3], Step [84/3236], Loss: 3.8088, Perplexity: 45.0974
Epoch [1/3], Step [85/3236], Loss: 3.8764, Perplexity: 48.2480
Epoch [1/3], Step [86/3236], Loss: 3.9619, Perplexity: 52.5575
Epoch [1/3], Step [87/3236], Loss: 3.9181, Perplexity: 50.3047
Epoch [1/3], Step [88/3236], Loss: 4.2296, Perplexity: 68.6917
Epoch [1/3], Step [89/3236], Loss: 4.4092, Perplexity: 82.2065
Epoch [1/3], Step [90/3236], Loss: 3.9192, Perplexity: 50.3623
Epoch [1/3], Step [91/3236], Loss: 3.7880, Perplexity: 44.1689
Epoch [1/3], Step [92/3236], Loss: 3.7743, Perplexity: 43.5686
Epoch [1/3], Step [93/3236], Loss: 3.7990, Perplexity: 44.6581
Epoch [1/3], Step [94/3236], Loss: 3.8203, Perplexity: 45.6166
Epoch [1/3], Step [95/3236], Loss: 3.9275, Perplexity: 50.7799
Epoch [1/3], Step [96/3236], Loss: 3.8648, Perplexity: 47.6922
Epoch [1/3], Step [97/3236], Loss: 3.7868, Perplexity: 44.1139
Epoch [1/3], Step [98/3236], Loss: 3.9210, Perplexity: 50.4516
Epoch [1/3], Step [99/3236], Loss: 3.8002, Perplexity: 44.7098
Epoch [1/3], Step [100/3236], Loss: 4.0337, Perplexity: 56.4693
Epoch [1/3], Step [101/3236], Loss: 3.8043, Perplexity: 44.8953
Epoch [1/3], Step [102/3236], Loss: 3.6890, Perplexity: 40.0046
Epoch [1/3], Step [103/3236], Loss: 3.7702, Perplexity: 43.3881
Epoch [1/3], Step [104/3236], Loss: 3.7135, Perplexity: 40.9967
Epoch [1/3], Step [105/3236], Loss: 3.7574, Perplexity: 42.8364
Epoch [1/3], Step [106/3236], Loss: 4.0484, Perplexity: 57.3077
Epoch [1/3], Step [107/3236], Loss: 3.7564, Perplexity: 42.7926
Epoch [1/3], Step [108/3236], Loss: 3.8896, Perplexity: 48.8926
Epoch [1/3], Step [109/3236], Loss: 3.9323, Perplexity: 51.0242
Epoch [1/3], Step [110/3236], Loss: 3.7873, Perplexity: 44.1370
Epoch [1/3], Step [111/3236], Loss: 3.7921, Perplexity: 44.3498
Epoch [1/3], Step [112/3236], Loss: 3.7328, Perplexity: 41.7980
Epoch [1/3], Step [113/3236], Loss: 3.7268, Perplexity: 41.5475
Epoch [1/3], Step [114/3236], Loss: 4.2500, Perplexity: 70.1081
Epoch [1/3], Step [115/3236], Loss: 3.7192, Perplexity: 41.2322
Epoch [1/3], Step [116/3236], Loss: 3.7828, Perplexity: 43.9405
Epoch [1/3], Step [117/3236], Loss: 3.6932, Perplexity: 40.1743
Epoch [1/3], Step [118/3236], Loss: 3.6115, Perplexity: 37.0201
Epoch [1/3], Step [119/3236], Loss: 3.6681, Perplexity: 39.1763
Epoch [1/3], Step [120/3236], Loss: 3.7813, Perplexity: 43.8745
Epoch [1/3], Step [121/3236], Loss: 3.8499, Perplexity: 46.9881
Epoch [1/3], Step [122/3236], Loss: 3.7034, Perplexity: 40.5834
Epoch [1/3], Step [123/3236], Loss: 3.7147, Perplexity: 41.0443
Epoch [1/3], Step [124/3236], Loss: 3.7257, Perplexity: 41.5023
Epoch [1/3], Step [125/3236], Loss: 3.8021, Perplexity: 44.7961
Epoch [1/3], Step [126/3236], Loss: 3.5834, Perplexity: 35.9962
Epoch [1/3], Step [127/3236], Loss: 3.8172, Perplexity: 45.4779
Epoch [1/3], Step [128/3236], Loss: 3.6394, Perplexity: 38.0701
Epoch [1/3], Step [129/3236], Loss: 3.6363, Perplexity: 37.9525
Epoch [1/3], Step [130/3236], Loss: 3.6813, Perplexity: 39.6997
Epoch [1/3], Step [131/3236], Loss: 3.6842, Perplexity: 39.8136
Epoch [1/3], Step [132/3236], Loss: 3.6948, Perplexity: 40.2390
Epoch [1/3], Step [133/3236], Loss: 3.4712, Perplexity: 32.1765
Epoch [1/3], Step [134/3236], Loss: 3.5564, Perplexity: 35.0356
Epoch [1/3], Step [135/3236], Loss: 3.6076, Perplexity: 36.8790
Epoch [1/3], Step [136/3236], Loss: 3.7536, Perplexity: 42.6762
Epoch [1/3], Step [137/3236], Loss: 3.5915, Perplexity: 36.2890
Epoch [1/3], Step [138/3236], Loss: 4.0280, Perplexity: 56.1484
Epoch [1/3], Step [139/3236], Loss: 3.5818, Perplexity: 35.9386
Epoch [1/3], Step [140/3236], Loss: 3.7307, Perplexity: 41.7074
Epoch [1/3], Step [141/3236], Loss: 3.6326, Perplexity: 37.8126
Epoch [1/3], Step [142/3236], Loss: 3.6423, Perplexity: 38.1792
Epoch [1/3], Step [143/3236], Loss: 3.4708, Perplexity: 32.1623
Epoch [1/3], Step [144/3236], Loss: 4.0719, Perplexity: 58.6701
Epoch [1/3], Step [145/3236], Loss: 3.6128, Perplexity: 37.0694
Epoch [1/3], Step [146/3236], Loss: 4.4026, Perplexity: 81.6594
Epoch [1/3], Step [147/3236], Loss: 3.8537, Perplexity: 47.1671
Epoch [1/3], Step [148/3236], Loss: 3.6843, Perplexity: 39.8166
Epoch [1/3], Step [149/3236], Loss: 3.6270, Perplexity: 37.6010
Epoch [1/3], Step [150/3236], Loss: 3.7239, Perplexity: 41.4274
Epoch [1/3], Step [151/3236], Loss: 3.6712, Perplexity: 39.2996
Epoch [1/3], Step [152/3236], Loss: 3.7269, Perplexity: 41.5506
Epoch [1/3], Step [153/3236], Loss: 3.5361, Perplexity: 34.3337
Epoch [1/3], Step [154/3236], Loss: 3.6141, Perplexity: 37.1186
Epoch [1/3], Step [155/3236], Loss: 3.5294, Perplexity: 34.1020
Epoch [1/3], Step [156/3236], Loss: 3.7214, Perplexity: 41.3221
Epoch [1/3], Step [157/3236], Loss: 3.6243, Perplexity: 37.4969
Epoch [1/3], Step [158/3236], Loss: 3.7389, Perplexity: 42.0529
Epoch [1/3], Step [159/3236], Loss: 3.7025, Perplexity: 40.5481
Epoch [1/3], Step [160/3236], Loss: 3.6144, Perplexity: 37.1286
Epoch [1/3], Step [161/3236], Loss: 3.6746, Perplexity: 39.4318
Epoch [1/3], Step [162/3236], Loss: 3.8175, Perplexity: 45.4911
Epoch [1/3], Step [163/3236], Loss: 3.7539, Perplexity: 42.6851
Epoch [1/3], Step [164/3236], Loss: 3.6697, Perplexity: 39.2389
Epoch [1/3], Step [165/3236], Loss: 3.5008, Perplexity: 33.1416
Epoch [1/3], Step [166/3236], Loss: 3.8001, Perplexity: 44.7056
Epoch [1/3], Step [167/3236], Loss: 3.5243, Perplexity: 33.9284
Epoch [1/3], Step [168/3236], Loss: 3.5682, Perplexity: 35.4539
Epoch [1/3], Step [169/3236], Loss: 3.5825, Perplexity: 35.9643
Epoch [1/3], Step [170/3236], Loss: 3.6222, Perplexity: 37.4181
Epoch [1/3], Step [171/3236], Loss: 3.5330, Perplexity: 34.2252
Epoch [1/3], Step [172/3236], Loss: 3.6451, Perplexity: 38.2865
Epoch [1/3], Step [173/3236], Loss: 3.4728, Perplexity: 32.2282
Epoch [1/3], Step [174/3236], Loss: 3.6134, Perplexity: 37.0927
Epoch [1/3], Step [175/3236], Loss: 3.8127, Perplexity: 45.2733
Epoch [1/3], Step [176/3236], Loss: 3.4370, Perplexity: 31.0927
Epoch [1/3], Step [177/3236], Loss: 3.5371, Perplexity: 34.3675
Epoch [1/3], Step [178/3236], Loss: 3.5884, Perplexity: 36.1777
Epoch [1/3], Step [179/3236], Loss: 3.4052, Perplexity: 30.1194
Epoch [1/3], Step [180/3236], Loss: 3.4205, Perplexity: 30.5848
Epoch [1/3], Step [181/3236], Loss: 3.3794, Perplexity: 29.3530
Epoch [1/3], Step [182/3236], Loss: 4.1145, Perplexity: 61.2239
Epoch [1/3], Step [183/3236], Loss: 3.3077, Perplexity: 27.3215
Epoch [1/3], Step [184/3236], Loss: 4.6241, Perplexity: 101.9134
Epoch [1/3], Step [185/3236], Loss: 3.5696, Perplexity: 35.5016
Epoch [1/3], Step [186/3236], Loss: 3.3692, Perplexity: 29.0555
Epoch [1/3], Step [187/3236], Loss: 3.5375, Perplexity: 34.3821
Epoch [1/3], Step [188/3236], Loss: 3.4795, Perplexity: 32.4448
Epoch [1/3], Step [189/3236], Loss: 3.4843, Perplexity: 32.5993
Epoch [1/3], Step [190/3236], Loss: 3.3526, Perplexity: 28.5781
Epoch [1/3], Step [191/3236], Loss: 3.3476, Perplexity: 28.4333
Epoch [1/3], Step [192/3236], Loss: 3.6415, Perplexity: 38.1502
Epoch [1/3], Step [193/3236], Loss: 3.4159, Perplexity: 30.4443
Epoch [1/3], Step [194/3236], Loss: 3.5194, Perplexity: 33.7640
Epoch [1/3], Step [195/3236], Loss: 3.4656, Perplexity: 31.9967
Epoch [1/3], Step [196/3236], Loss: 3.4788, Perplexity: 32.4194
Epoch [1/3], Step [197/3236], Loss: 3.6849, Perplexity: 39.8408
Epoch [1/3], Step [198/3236], Loss: 3.4118, Perplexity: 30.3186
Epoch [1/3], Step [199/3236], Loss: 4.0704, Perplexity: 58.5797
Epoch [1/3], Step [200/3236], Loss: 3.3766, Perplexity: 29.2698
Epoch [1/3], Step [201/3236], Loss: 3.6498, Perplexity: 38.4662
Epoch [1/3], Step [202/3236], Loss: 3.2835, Perplexity: 26.6682
Epoch [1/3], Step [203/3236], Loss: 3.3052, Perplexity: 27.2547
Epoch [1/3], Step [204/3236], Loss: 3.5920, Perplexity: 36.3079
Epoch [1/3], Step [205/3236], Loss: 3.2511, Perplexity: 25.8176
Epoch [1/3], Step [206/3236], Loss: 3.2647, Perplexity: 26.1712
Epoch [1/3], Step [207/3236], Loss: 3.3743, Perplexity: 29.2051
Epoch [1/3], Step [208/3236], Loss: 3.4941, Perplexity: 32.9213
Epoch [1/3], Step [209/3236], Loss: 3.6360, Perplexity: 37.9398
Epoch [1/3], Step [210/3236], Loss: 3.4018, Perplexity: 30.0193
Epoch [1/3], Step [211/3236], Loss: 3.5336, Perplexity: 34.2486
Epoch [1/3], Step [212/3236], Loss: 3.5306, Perplexity: 34.1458
Epoch [1/3], Step [213/3236], Loss: 3.2928, Perplexity: 26.9177
Epoch [1/3], Step [214/3236], Loss: 3.4710, Perplexity: 32.1690
Epoch [1/3], Step [215/3236], Loss: 3.4725, Perplexity: 32.2178
Epoch [1/3], Step [216/3236], Loss: 3.5442, Perplexity: 34.6118
Epoch [1/3], Step [217/3236], Loss: 3.4789, Perplexity: 32.4231
Epoch [1/3], Step [218/3236], Loss: 3.4513, Perplexity: 31.5428
Epoch [1/3], Step [219/3236], Loss: 3.4146, Perplexity: 30.4058
Epoch [1/3], Step [220/3236], Loss: 3.8760, Perplexity: 48.2322
Epoch [1/3], Step [221/3236], Loss: 3.6699, Perplexity: 39.2476
Epoch [1/3], Step [222/3236], Loss: 3.3711, Perplexity: 29.1091
Epoch [1/3], Step [223/3236], Loss: 3.5976, Perplexity: 36.5103
Epoch [1/3], Step [224/3236], Loss: 3.6598, Perplexity: 38.8552
Epoch [1/3], Step [225/3236], Loss: 3.3319, Perplexity: 27.9911
Epoch [1/3], Step [226/3236], Loss: 4.1944, Perplexity: 66.3110
Epoch [1/3], Step [227/3236], Loss: 3.3369, Perplexity: 28.1323
Epoch [1/3], Step [228/3236], Loss: 3.4713, Perplexity: 32.1791
Epoch [1/3], Step [229/3236], Loss: 3.5034, Perplexity: 33.2297
Epoch [1/3], Step [230/3236], Loss: 3.8595, Perplexity: 47.4418
Epoch [1/3], Step [231/3236], Loss: 3.3217, Perplexity: 27.7083
Epoch [1/3], Step [232/3236], Loss: 3.4625, Perplexity: 31.8956
Epoch [1/3], Step [233/3236], Loss: 3.3391, Perplexity: 28.1944
Epoch [1/3], Step [234/3236], Loss: 3.5731, Perplexity: 35.6273
Epoch [1/3], Step [235/3236], Loss: 3.2887, Perplexity: 26.8076
Epoch [1/3], Step [236/3236], Loss: 3.3397, Perplexity: 28.2098
Epoch [1/3], Step [237/3236], Loss: 3.6820, Perplexity: 39.7268
Epoch [1/3], Step [238/3236], Loss: 3.4409, Perplexity: 31.2137
Epoch [1/3], Step [239/3236], Loss: 3.5566, Perplexity: 35.0435
Epoch [1/3], Step [240/3236], Loss: 3.3506, Perplexity: 28.5199
Epoch [1/3], Step [241/3236], Loss: 3.4602, Perplexity: 31.8248
Epoch [1/3], Step [242/3236], Loss: 3.2585, Perplexity: 26.0095
Epoch [1/3], Step [243/3236], Loss: 3.3850, Perplexity: 29.5175
Epoch [1/3], Step [244/3236], Loss: 3.3700, Perplexity: 29.0787
Epoch [1/3], Step [245/3236], Loss: 3.8614, Perplexity: 47.5317
Epoch [1/3], Step [246/3236], Loss: 3.3061, Perplexity: 27.2794
Epoch [1/3], Step [247/3236], Loss: 3.3614, Perplexity: 28.8298
Epoch [1/3], Step [248/3236], Loss: 3.7443, Perplexity: 42.2789
Epoch [1/3], Step [249/3236], Loss: 3.3536, Perplexity: 28.6051
Epoch [1/3], Step [250/3236], Loss: 3.5930, Perplexity: 36.3411
Epoch [1/3], Step [251/3236], Loss: 3.4791, Perplexity: 32.4312
Epoch [1/3], Step [252/3236], Loss: 3.2242, Perplexity: 25.1347
Epoch [1/3], Step [253/3236], Loss: 3.4059, Perplexity: 30.1414
Epoch [1/3], Step [254/3236], Loss: 3.5432, Perplexity: 34.5763
Epoch [1/3], Step [255/3236], Loss: 3.2910, Perplexity: 26.8685
Epoch [1/3], Step [256/3236], Loss: 3.2928, Perplexity: 26.9179
Epoch [1/3], Step [257/3236], Loss: 3.4497, Perplexity: 31.4896
Epoch [1/3], Step [258/3236], Loss: 3.4113, Perplexity: 30.3036
Epoch [1/3], Step [259/3236], Loss: 3.5723, Perplexity: 35.5990
Epoch [1/3], Step [260/3236], Loss: 3.3434, Perplexity: 28.3149
Epoch [1/3], Step [261/3236], Loss: 3.4666, Perplexity: 32.0282
Epoch [1/3], Step [262/3236], Loss: 3.3809, Perplexity: 29.3972
Epoch [1/3], Step [263/3236], Loss: 3.2823, Perplexity: 26.6379
Epoch [1/3], Step [264/3236], Loss: 3.2957, Perplexity: 26.9958
Epoch [1/3], Step [265/3236], Loss: 4.2631, Perplexity: 71.0300
Epoch [1/3], Step [266/3236], Loss: 3.1797, Perplexity: 24.0397
Epoch [1/3], Step [267/3236], Loss: 3.6141, Perplexity: 37.1170
Epoch [1/3], Step [268/3236], Loss: 3.3794, Perplexity: 29.3540
Epoch [1/3], Step [269/3236], Loss: 3.3168, Perplexity: 27.5708
Epoch [1/3], Step [270/3236], Loss: 3.3267, Perplexity: 27.8452
Epoch [1/3], Step [271/3236], Loss: 3.4546, Perplexity: 31.6453
Epoch [1/3], Step [272/3236], Loss: 3.2927, Perplexity: 26.9160
Epoch [1/3], Step [273/3236], Loss: 3.4761, Perplexity: 32.3340
Epoch [1/3], Step [274/3236], Loss: 3.2616, Perplexity: 26.0924
Epoch [1/3], Step [275/3236], Loss: 3.3694, Perplexity: 29.0608
Epoch [1/3], Step [276/3236], Loss: 3.3331, Perplexity: 28.0254
Epoch [1/3], Step [277/3236], Loss: 3.3428, Perplexity: 28.2988
Epoch [1/3], Step [278/3236], Loss: 3.2306, Perplexity: 25.2953
Epoch [1/3], Step [279/3236], Loss: 3.1335, Perplexity: 22.9546
Epoch [1/3], Step [280/3236], Loss: 3.9199, Perplexity: 50.3949
Epoch [1/3], Step [281/3236], Loss: 3.2978, Perplexity: 27.0534
Epoch [1/3], Step [282/3236], Loss: 4.1304, Perplexity: 62.2000
Epoch [1/3], Step [283/3236], Loss: 4.0054, Perplexity: 54.8951
Epoch [1/3], Step [284/3236], Loss: 3.4448, Perplexity: 31.3369
Epoch [1/3], Step [285/3236], Loss: 3.2363, Perplexity: 25.4387
Epoch [1/3], Step [286/3236], Loss: 3.2084, Perplexity: 24.7391
Epoch [1/3], Step [287/3236], Loss: 3.4126, Perplexity: 30.3433
Epoch [1/3], Step [288/3236], Loss: 3.3320, Perplexity: 27.9937
Epoch [1/3], Step [289/3236], Loss: 3.2265, Perplexity: 25.1907
Epoch [1/3], Step [290/3236], Loss: 3.4724, Perplexity: 32.2147
Epoch [1/3], Step [291/3236], Loss: 3.2784, Perplexity: 26.5330
Epoch [1/3], Step [292/3236], Loss: 3.7265, Perplexity: 41.5321
Epoch [1/3], Step [293/3236], Loss: 3.4153, Perplexity: 30.4248
Epoch [1/3], Step [294/3236], Loss: 3.3980, Perplexity: 29.9038
Epoch [1/3], Step [295/3236], Loss: 3.3365, Perplexity: 28.1213
Epoch [1/3], Step [296/3236], Loss: 3.6670, Perplexity: 39.1344
Epoch [1/3], Step [297/3236], Loss: 3.1524, Perplexity: 23.3921
Epoch [1/3], Step [298/3236], Loss: 3.3173, Perplexity: 27.5865
Epoch [1/3], Step [299/3236], Loss: 3.1456, Perplexity: 23.2328
Epoch [1/3], Step [300/3236], Loss: 3.1631, Perplexity: 23.6430
Epoch [1/3], Step [301/3236], Loss: 3.4341, Perplexity: 31.0043
Epoch [1/3], Step [302/3236], Loss: 3.1690, Perplexity: 23.7846
Epoch [1/3], Step [303/3236], Loss: 3.2009, Perplexity: 24.5542
Epoch [1/3], Step [304/3236], Loss: 3.2384, Perplexity: 25.4927
Epoch [1/3], Step [305/3236], Loss: 3.4216, Perplexity: 30.6171
Epoch [1/3], Step [306/3236], Loss: 3.7130, Perplexity: 40.9747
Epoch [1/3], Step [307/3236], Loss: 3.2238, Perplexity: 25.1224
Epoch [1/3], Step [308/3236], Loss: 3.3303, Perplexity: 27.9473
Epoch [1/3], Step [309/3236], Loss: 3.8373, Perplexity: 46.4014
Epoch [1/3], Step [310/3236], Loss: 3.1579, Perplexity: 23.5208
Epoch [1/3], Step [311/3236], Loss: 3.4803, Perplexity: 32.4696
Epoch [1/3], Step [312/3236], Loss: 3.4005, Perplexity: 29.9789
Epoch [1/3], Step [313/3236], Loss: 3.3987, Perplexity: 29.9240
Epoch [1/3], Step [314/3236], Loss: 3.1196, Perplexity: 22.6373
Epoch [1/3], Step [315/3236], Loss: 3.2362, Perplexity: 25.4373
Epoch [1/3], Step [316/3236], Loss: 3.1371, Perplexity: 23.0364
Epoch [1/3], Step [317/3236], Loss: 3.2052, Perplexity: 24.6615
Epoch [1/3], Step [318/3236], Loss: 3.2235, Perplexity: 25.1154
Epoch [1/3], Step [319/3236], Loss: 3.2452, Perplexity: 25.6677
Epoch [1/3], Step [320/3236], Loss: 3.6777, Perplexity: 39.5544
Epoch [1/3], Step [321/3236], Loss: 3.3078, Perplexity: 27.3255
Epoch [1/3], Step [322/3236], Loss: 3.8821, Perplexity: 48.5278
Epoch [1/3], Step [323/3236], Loss: 3.2390, Perplexity: 25.5084
Epoch [1/3], Step [324/3236], Loss: 3.1882, Perplexity: 24.2442
Epoch [1/3], Step [325/3236], Loss: 3.5282, Perplexity: 34.0627
Epoch [1/3], Step [326/3236], Loss: 3.3296, Perplexity: 27.9258
Epoch [1/3], Step [327/3236], Loss: 3.5578, Perplexity: 35.0876
Epoch [1/3], Step [328/3236], Loss: 3.1090, Perplexity: 22.3997
Epoch [1/3], Step [329/3236], Loss: 3.2927, Perplexity: 26.9153
Epoch [1/3], Step [330/3236], Loss: 3.3347, Perplexity: 28.0696
Epoch [1/3], Step [331/3236], Loss: 3.0945, Perplexity: 22.0757
Epoch [1/3], Step [332/3236], Loss: 3.7099, Perplexity: 40.8500
Epoch [1/3], Step [333/3236], Loss: 3.1199, Perplexity: 22.6432
Epoch [1/3], Step [334/3236], Loss: 3.3782, Perplexity: 29.3165
Epoch [1/3], Step [335/3236], Loss: 3.2099, Perplexity: 24.7769
Epoch [1/3], Step [336/3236], Loss: 3.4216, Perplexity: 30.6186
Epoch [1/3], Step [337/3236], Loss: 3.2510, Perplexity: 25.8173
Epoch [1/3], Step [338/3236], Loss: 3.2412, Perplexity: 25.5633
Epoch [1/3], Step [339/3236], Loss: 3.3506, Perplexity: 28.5189
Epoch [1/3], Step [340/3236], Loss: 3.2224, Perplexity: 25.0875
Epoch [1/3], Step [341/3236], Loss: 3.1393, Perplexity: 23.0877
Epoch [1/3], Step [342/3236], Loss: 3.5323, Perplexity: 34.2040
Epoch [1/3], Step [343/3236], Loss: 3.1978, Perplexity: 24.4780
Epoch [1/3], Step [344/3236], Loss: 3.4200, Perplexity: 30.5681
Epoch [1/3], Step [345/3236], Loss: 3.2121, Perplexity: 24.8317
Epoch [1/3], Step [346/3236], Loss: 3.5228, Perplexity: 33.8791
Epoch [1/3], Step [347/3236], Loss: 3.4949, Perplexity: 32.9457
Epoch [1/3], Step [348/3236], Loss: 3.6132, Perplexity: 37.0838
Epoch [1/3], Step [349/3236], Loss: 3.0852, Perplexity: 21.8723
Epoch [1/3], Step [350/3236], Loss: 3.2841, Perplexity: 26.6850
Epoch [1/3], Step [351/3236], Loss: 3.2916, Perplexity: 26.8857
Epoch [1/3], Step [352/3236], Loss: 3.1794, Perplexity: 24.0317
Epoch [1/3], Step [353/3236], Loss: 3.1455, Perplexity: 23.2316
Epoch [1/3], Step [354/3236], Loss: 3.1721, Perplexity: 23.8577
Epoch [1/3], Step [355/3236], Loss: 3.2100, Perplexity: 24.7785
Epoch [1/3], Step [356/3236], Loss: 3.1337, Perplexity: 22.9595
Epoch [1/3], Step [357/3236], Loss: 3.2997, Perplexity: 27.1033
Epoch [1/3], Step [358/3236], Loss: 3.1215, Perplexity: 22.6812
Epoch [1/3], Step [359/3236], Loss: 3.2993, Perplexity: 27.0928
Epoch [1/3], Step [360/3236], Loss: 4.0909, Perplexity: 59.7943
Epoch [1/3], Step [361/3236], Loss: 3.1566, Perplexity: 23.4896
Epoch [1/3], Step [362/3236], Loss: 3.5650, Perplexity: 35.3398
Epoch [1/3], Step [363/3236], Loss: 3.2972, Perplexity: 27.0377
Epoch [1/3], Step [364/3236], Loss: 3.0871, Perplexity: 21.9131
Epoch [1/3], Step [365/3236], Loss: 3.2033, Perplexity: 24.6127
Epoch [1/3], Step [366/3236], Loss: 3.1515, Perplexity: 23.3712
Epoch [1/3], Step [367/3236], Loss: 3.2562, Perplexity: 25.9508
Epoch [1/3], Step [368/3236], Loss: 3.0977, Perplexity: 22.1471
Epoch [1/3], Step [369/3236], Loss: 3.2012, Perplexity: 24.5619
Epoch [1/3], Step [370/3236], Loss: 3.4185, Perplexity: 30.5238
Epoch [1/3], Step [371/3236], Loss: 3.5416, Perplexity: 34.5233
Epoch [1/3], Step [372/3236], Loss: 3.2485, Perplexity: 25.7512
Epoch [1/3], Step [373/3236], Loss: 3.5711, Perplexity: 35.5552
Epoch [1/3], Step [374/3236], Loss: 3.2591, Perplexity: 26.0253
Epoch [1/3], Step [375/3236], Loss: 3.1464, Perplexity: 23.2524
Epoch [1/3], Step [376/3236], Loss: 3.1437, Perplexity: 23.1889
Epoch [1/3], Step [377/3236], Loss: 3.5383, Perplexity: 34.4079
Epoch [1/3], Step [378/3236], Loss: 3.4070, Perplexity: 30.1747
Epoch [1/3], Step [379/3236], Loss: 4.2136, Perplexity: 67.6020
Epoch [1/3], Step [380/3236], Loss: 3.3290, Perplexity: 27.9096
Epoch [1/3], Step [381/3236], Loss: 3.9271, Perplexity: 50.7605
Epoch [1/3], Step [382/3236], Loss: 3.2456, Perplexity: 25.6773
Epoch [1/3], Step [383/3236], Loss: 3.1288, Perplexity: 22.8460
Epoch [1/3], Step [384/3236], Loss: 3.2715, Perplexity: 26.3498
Epoch [1/3], Step [385/3236], Loss: 3.2851, Perplexity: 26.7121
Epoch [1/3], Step [386/3236], Loss: 3.1629, Perplexity: 23.6394
Epoch [1/3], Step [387/3236], Loss: 3.1661, Perplexity: 23.7155
Epoch [1/3], Step [388/3236], Loss: 3.0834, Perplexity: 21.8328
Epoch [1/3], Step [389/3236], Loss: 3.0974, Perplexity: 22.1408
Epoch [1/3], Step [390/3236], Loss: 3.7581, Perplexity: 42.8671
Epoch [1/3], Step [391/3236], Loss: 3.1623, Perplexity: 23.6243
Epoch [1/3], Step [392/3236], Loss: 3.1598, Perplexity: 23.5662
Epoch [1/3], Step [393/3236], Loss: 3.1898, Perplexity: 24.2827
Epoch [1/3], Step [394/3236], Loss: 3.3020, Perplexity: 27.1657
Epoch [1/3], Step [395/3236], Loss: 3.1496, Perplexity: 23.3277
Epoch [1/3], Step [396/3236], Loss: 4.3625, Perplexity: 78.4542
Epoch [1/3], Step [397/3236], Loss: 3.2371, Perplexity: 25.4599
Epoch [1/3], Step [398/3236], Loss: 2.8794, Perplexity: 17.8030
Epoch [1/3], Step [399/3236], Loss: 3.1983, Perplexity: 24.4900
Epoch [1/3], Step [400/3236], Loss: 3.1680, Perplexity: 23.7606
Epoch [1/3], Step [401/3236], Loss: 3.3688, Perplexity: 29.0446
Epoch [1/3], Step [402/3236], Loss: 3.3054, Perplexity: 27.2595
Epoch [1/3], Step [403/3236], Loss: 3.0837, Perplexity: 21.8389
Epoch [1/3], Step [404/3236], Loss: 3.1750, Perplexity: 23.9269
Epoch [1/3], Step [405/3236], Loss: 3.2043, Perplexity: 24.6389
Epoch [1/3], Step [406/3236], Loss: 3.1351, Perplexity: 22.9917
Epoch [1/3], Step [407/3236], Loss: 3.0860, Perplexity: 21.8884
Epoch [1/3], Step [408/3236], Loss: 3.1067, Perplexity: 22.3470
Epoch [1/3], Step [409/3236], Loss: 3.0868, Perplexity: 21.9067
Epoch [1/3], Step [410/3236], Loss: 3.1468, Perplexity: 23.2613
Epoch [1/3], Step [411/3236], Loss: 3.1052, Perplexity: 22.3148
Epoch [1/3], Step [412/3236], Loss: 3.2867, Perplexity: 26.7555
Epoch [1/3], Step [413/3236], Loss: 3.1007, Perplexity: 22.2141
Epoch [1/3], Step [414/3236], Loss: 3.2181, Perplexity: 24.9802
Epoch [1/3], Step [415/3236], Loss: 2.9817, Perplexity: 19.7208
Epoch [1/3], Step [416/3236], Loss: 3.3853, Perplexity: 29.5276
Epoch [1/3], Step [417/3236], Loss: 3.0649, Perplexity: 21.4319
Epoch [1/3], Step [418/3236], Loss: 3.1691, Perplexity: 23.7860
Epoch [1/3], Step [419/3236], Loss: 3.0897, Perplexity: 21.9713
Epoch [1/3], Step [420/3236], Loss: 3.3743, Perplexity: 29.2045
Epoch [1/3], Step [421/3236], Loss: 3.3470, Perplexity: 28.4161
Epoch [1/3], Step [422/3236], Loss: 3.3306, Perplexity: 27.9549
Epoch [1/3], Step [423/3236], Loss: 3.0222, Perplexity: 20.5357
Epoch [1/3], Step [424/3236], Loss: 3.0854, Perplexity: 21.8758
Epoch [1/3], Step [425/3236], Loss: 3.0674, Perplexity: 21.4854
Epoch [1/3], Step [426/3236], Loss: 3.1354, Perplexity: 22.9969
Epoch [1/3], Step [427/3236], Loss: 3.5919, Perplexity: 36.3013
Epoch [1/3], Step [428/3236], Loss: 3.1457, Perplexity: 23.2359
Epoch [1/3], Step [429/3236], Loss: 3.1645, Perplexity: 23.6777
Epoch [1/3], Step [430/3236], Loss: 3.1993, Perplexity: 24.5163
Epoch [1/3], Step [431/3236], Loss: 3.1545, Perplexity: 23.4409
Epoch [1/3], Step [432/3236], Loss: 3.0273, Perplexity: 20.6412
Epoch [1/3], Step [433/3236], Loss: 3.1909, Perplexity: 24.3106
Epoch [1/3], Step [434/3236], Loss: 3.3600, Perplexity: 28.7899
Epoch [1/3], Step [435/3236], Loss: 3.3390, Perplexity: 28.1919
Epoch [1/3], Step [436/3236], Loss: 3.2188, Perplexity: 24.9974
Epoch [1/3], Step [437/3236], Loss: 3.0250, Perplexity: 20.5947
Epoch [1/3], Step [438/3236], Loss: 3.0671, Perplexity: 21.4787
Epoch [1/3], Step [439/3236], Loss: 3.1214, Perplexity: 22.6782
Epoch [1/3], Step [440/3236], Loss: 3.2172, Perplexity: 24.9593
Epoch [1/3], Step [441/3236], Loss: 3.2774, Perplexity: 26.5066
Epoch [1/3], Step [442/3236], Loss: 3.1121, Perplexity: 22.4686
Epoch [1/3], Step [443/3236], Loss: 3.1786, Perplexity: 24.0138
Epoch [1/3], Step [444/3236], Loss: 3.5328, Perplexity: 34.2181
Epoch [1/3], Step [445/3236], Loss: 3.6481, Perplexity: 38.4010
Epoch [1/3], Step [446/3236], Loss: 2.9653, Perplexity: 19.3996
Epoch [1/3], Step [447/3236], Loss: 3.1495, Perplexity: 23.3246
Epoch [1/3], Step [448/3236], Loss: 3.2048, Perplexity: 24.6499
Epoch [1/3], Step [449/3236], Loss: 3.0597, Perplexity: 21.3221
Epoch [1/3], Step [450/3236], Loss: 3.2402, Perplexity: 25.5391
Epoch [1/3], Step [451/3236], Loss: 3.2487, Perplexity: 25.7573
Epoch [1/3], Step [452/3236], Loss: 3.5940, Perplexity: 36.3796
Epoch [1/3], Step [453/3236], Loss: 3.3191, Perplexity: 27.6347
Epoch [1/3], Step [454/3236], Loss: 3.0537, Perplexity: 21.1931
Epoch [1/3], Step [455/3236], Loss: 3.1197, Perplexity: 22.6405
Epoch [1/3], Step [456/3236], Loss: 3.5730, Perplexity: 35.6228
Epoch [1/3], Step [457/3236], Loss: 2.9531, Perplexity: 19.1645
Epoch [1/3], Step [458/3236], Loss: 3.0264, Perplexity: 20.6234
Epoch [1/3], Step [459/3236], Loss: 3.2685, Perplexity: 26.2710
Epoch [1/3], Step [460/3236], Loss: 3.1043, Perplexity: 22.2932
Epoch [1/3], Step [461/3236], Loss: 3.4249, Perplexity: 30.7181
Epoch [1/3], Step [462/3236], Loss: 3.2145, Perplexity: 24.8913
Epoch [1/3], Step [463/3236], Loss: 3.7458, Perplexity: 42.3423
Epoch [1/3], Step [464/3236], Loss: 3.2292, Perplexity: 25.2601
Epoch [1/3], Step [465/3236], Loss: 3.0337, Perplexity: 20.7747
Epoch [1/3], Step [466/3236], Loss: 3.2926, Perplexity: 26.9121
Epoch [1/3], Step [467/3236], Loss: 3.1942, Perplexity: 24.3912
Epoch [1/3], Step [468/3236], Loss: 3.0430, Perplexity: 20.9671
Epoch [1/3], Step [469/3236], Loss: 2.9626, Perplexity: 19.3478
Epoch [1/3], Step [470/3236], Loss: 3.1765, Perplexity: 23.9635
Epoch [1/3], Step [471/3236], Loss: 3.0605, Perplexity: 21.3381
Epoch [1/3], Step [472/3236], Loss: 3.0815, Perplexity: 21.7916
Epoch [1/3], Step [473/3236], Loss: 3.2175, Perplexity: 24.9653
Epoch [1/3], Step [474/3236], Loss: 3.2421, Perplexity: 25.5875
Epoch [1/3], Step [475/3236], Loss: 3.1007, Perplexity: 22.2140
Epoch [1/3], Step [476/3236], Loss: 3.3879, Perplexity: 29.6044
Epoch [1/3], Step [477/3236], Loss: 2.9501, Perplexity: 19.1076
Epoch [1/3], Step [478/3236], Loss: 3.1079, Perplexity: 22.3749
Epoch [1/3], Step [479/3236], Loss: 3.1732, Perplexity: 23.8836
Epoch [1/3], Step [480/3236], Loss: 3.1350, Perplexity: 22.9894
Epoch [1/3], Step [481/3236], Loss: 3.2022, Perplexity: 24.5860
Epoch [1/3], Step [482/3236], Loss: 3.3291, Perplexity: 27.9133
Epoch [1/3], Step [483/3236], Loss: 2.8677, Perplexity: 17.5971
Epoch [1/3], Step [484/3236], Loss: 3.3279, Perplexity: 27.8808
Epoch [1/3], Step [485/3236], Loss: 2.9924, Perplexity: 19.9328
Epoch [1/3], Step [486/3236], Loss: 3.0859, Perplexity: 21.8870
Epoch [1/3], Step [487/3236], Loss: 3.0589, Perplexity: 21.3048
Epoch [1/3], Step [488/3236], Loss: 2.9630, Perplexity: 19.3569
Epoch [1/3], Step [489/3236], Loss: 2.9618, Perplexity: 19.3334
Epoch [1/3], Step [490/3236], Loss: 3.1235, Perplexity: 22.7253
Epoch [1/3], Step [491/3236], Loss: 3.5390, Perplexity: 34.4331
Epoch [1/3], Step [492/3236], Loss: 3.1626, Perplexity: 23.6322
Epoch [1/3], Step [493/3236], Loss: 3.2692, Perplexity: 26.2893
Epoch [1/3], Step [494/3236], Loss: 2.9986, Perplexity: 20.0571
Epoch [1/3], Step [495/3236], Loss: 3.2133, Perplexity: 24.8621
Epoch [1/3], Step [496/3236], Loss: 3.3515, Perplexity: 28.5460
Epoch [1/3], Step [497/3236], Loss: 3.0910, Perplexity: 21.9994
Epoch [1/3], Step [498/3236], Loss: 3.2294, Perplexity: 25.2644
Epoch [1/3], Step [499/3236], Loss: 3.1151, Perplexity: 22.5359
Epoch [1/3], Step [500/3236], Loss: 2.9885, Perplexity: 19.8551
Epoch [1/3], Step [501/3236], Loss: 2.9890, Perplexity: 19.8655
Epoch [1/3], Step [502/3236], Loss: 3.4381, Perplexity: 31.1290
Epoch [1/3], Step [503/3236], Loss: 3.3279, Perplexity: 27.8792
Epoch [1/3], Step [504/3236], Loss: 3.1604, Perplexity: 23.5797
Epoch [1/3], Step [505/3236], Loss: 3.0053, Perplexity: 20.1932
Epoch [1/3], Step [506/3236], Loss: 3.0020, Perplexity: 20.1249
Epoch [1/3], Step [507/3236], Loss: 3.0261, Perplexity: 20.6159
Epoch [1/3], Step [508/3236], Loss: 3.1275, Perplexity: 22.8162
Epoch [1/3], Step [509/3236], Loss: 2.9913, Perplexity: 19.9121
Epoch [1/3], Step [510/3236], Loss: 2.9760, Perplexity: 19.6085
Epoch [1/3], Step [511/3236], Loss: 3.2600, Perplexity: 26.0483
Epoch [1/3], Step [512/3236], Loss: 3.1213, Perplexity: 22.6764
Epoch [1/3], Step [513/3236], Loss: 2.9448, Perplexity: 19.0066
Epoch [1/3], Step [514/3236], Loss: 3.0341, Perplexity: 20.7817
Epoch [1/3], Step [515/3236], Loss: 3.2899, Perplexity: 26.8401
Epoch [1/3], Step [516/3236], Loss: 3.2357, Perplexity: 25.4253
Epoch [1/3], Step [517/3236], Loss: 3.9203, Perplexity: 50.4148
Epoch [1/3], Step [518/3236], Loss: 3.1222, Perplexity: 22.6968
Epoch [1/3], Step [519/3236], Loss: 3.0649, Perplexity: 21.4320
Epoch [1/3], Step [520/3236], Loss: 2.9382, Perplexity: 18.8810
Epoch [1/3], Step [521/3236], Loss: 2.9048, Perplexity: 18.2618
Epoch [1/3], Step [522/3236], Loss: 3.1556, Perplexity: 23.4661
Epoch [1/3], Step [523/3236], Loss: 3.0275, Perplexity: 20.6450
Epoch [1/3], Step [524/3236], Loss: 3.9120, Perplexity: 49.9966
Epoch [1/3], Step [525/3236], Loss: 3.0742, Perplexity: 21.6331
Epoch [1/3], Step [526/3236], Loss: 3.1275, Perplexity: 22.8172
Epoch [1/3], Step [527/3236], Loss: 3.0786, Perplexity: 21.7284
Epoch [1/3], Step [528/3236], Loss: 3.1094, Perplexity: 22.4069
Epoch [1/3], Step [529/3236], Loss: 2.8601, Perplexity: 17.4631
Epoch [1/3], Step [530/3236], Loss: 3.1640, Perplexity: 23.6644
Epoch [1/3], Step [531/3236], Loss: 3.0955, Perplexity: 22.0993
Epoch [1/3], Step [532/3236], Loss: 3.0381, Perplexity: 20.8658
Epoch [1/3], Step [533/3236], Loss: 2.8767, Perplexity: 17.7561
Epoch [1/3], Step [534/3236], Loss: 3.0381, Perplexity: 20.8647
Epoch [1/3], Step [535/3236], Loss: 2.8869, Perplexity: 17.9371
Epoch [1/3], Step [536/3236], Loss: 3.1455, Perplexity: 23.2316
Epoch [1/3], Step [537/3236], Loss: 3.0091, Perplexity: 20.2685
Epoch [1/3], Step [538/3236], Loss: 3.0695, Perplexity: 21.5320
Epoch [1/3], Step [539/3236], Loss: 3.0997, Perplexity: 22.1912
Epoch [1/3], Step [540/3236], Loss: 3.0416, Perplexity: 20.9384
Epoch [1/3], Step [541/3236], Loss: 3.0235, Perplexity: 20.5635
Epoch [1/3], Step [542/3236], Loss: 2.9669, Perplexity: 19.4310
Epoch [1/3], Step [543/3236], Loss: 2.9589, Perplexity: 19.2769
Epoch [1/3], Step [544/3236], Loss: 3.5408, Perplexity: 34.4958
Epoch [1/3], Step [545/3236], Loss: 2.8109, Perplexity: 16.6254
Epoch [1/3], Step [546/3236], Loss: 3.4533, Perplexity: 31.6060
Epoch [1/3], Step [547/3236], Loss: 2.9369, Perplexity: 18.8579
Epoch [1/3], Step [548/3236], Loss: 2.8957, Perplexity: 18.0953
Epoch [1/3], Step [549/3236], Loss: 3.1468, Perplexity: 23.2616
Epoch [1/3], Step [550/3236], Loss: 3.2036, Perplexity: 24.6208
Epoch [1/3], Step [551/3236], Loss: 3.0223, Perplexity: 20.5383
Epoch [1/3], Step [552/3236], Loss: 3.1834, Perplexity: 24.1278
Epoch [1/3], Step [553/3236], Loss: 3.0233, Perplexity: 20.5599
Epoch [1/3], Step [554/3236], Loss: 2.9568, Perplexity: 19.2370
Epoch [1/3], Step [555/3236], Loss: 3.0089, Perplexity: 20.2644
Epoch [1/3], Step [556/3236], Loss: 3.2934, Perplexity: 26.9330
Epoch [1/3], Step [557/3236], Loss: 2.8735, Perplexity: 17.6981
Epoch [1/3], Step [558/3236], Loss: 3.2545, Perplexity: 25.9068
Epoch [1/3], Step [559/3236], Loss: 3.0050, Perplexity: 20.1862
Epoch [1/3], Step [560/3236], Loss: 4.1281, Perplexity: 62.0592
Epoch [1/3], Step [561/3236], Loss: 2.7944, Perplexity: 16.3523
Epoch [1/3], Step [562/3236], Loss: 3.5659, Perplexity: 35.3721
Epoch [1/3], Step [563/3236], Loss: 2.8977, Perplexity: 18.1321
Epoch [1/3], Step [564/3236], Loss: 2.8131, Perplexity: 16.6610
Epoch [1/3], Step [565/3236], Loss: 3.2383, Perplexity: 25.4892
Epoch [1/3], Step [566/3236], Loss: 3.0320, Perplexity: 20.7377
Epoch [1/3], Step [567/3236], Loss: 3.0220, Perplexity: 20.5327
Epoch [1/3], Step [568/3236], Loss: 2.8569, Perplexity: 17.4076
Epoch [1/3], Step [569/3236], Loss: 3.1077, Perplexity: 22.3698
Epoch [1/3], Step [570/3236], Loss: 2.9023, Perplexity: 18.2163
Epoch [1/3], Step [571/3236], Loss: 3.1340, Perplexity: 22.9656
Epoch [1/3], Step [572/3236], Loss: 2.8252, Perplexity: 16.8640
Epoch [1/3], Step [573/3236], Loss: 3.1192, Perplexity: 22.6292
Epoch [1/3], Step [574/3236], Loss: 3.0837, Perplexity: 21.8395
Epoch [1/3], Step [575/3236], Loss: 3.5916, Perplexity: 36.2903
Epoch [1/3], Step [576/3236], Loss: 3.1552, Perplexity: 23.4583
Epoch [1/3], Step [577/3236], Loss: 3.0023, Perplexity: 20.1322
Epoch [1/3], Step [578/3236], Loss: 3.0774, Perplexity: 21.7027
Epoch [1/3], Step [579/3236], Loss: 2.9280, Perplexity: 18.6896
Epoch [1/3], Step [580/3236], Loss: 3.0049, Perplexity: 20.1838
Epoch [1/3], Step [581/3236], Loss: 3.2909, Perplexity: 26.8683
Epoch [1/3], Step [582/3236], Loss: 3.0178, Perplexity: 20.4467
Epoch [1/3], Step [583/3236], Loss: 2.9868, Perplexity: 19.8216
Epoch [1/3], Step [584/3236], Loss: 3.1249, Perplexity: 22.7577
Epoch [1/3], Step [585/3236], Loss: 3.0071, Perplexity: 20.2281
Epoch [1/3], Step [586/3236], Loss: 2.9826, Perplexity: 19.7388
Epoch [1/3], Step [587/3236], Loss: 3.0678, Perplexity: 21.4954
Epoch [1/3], Step [588/3236], Loss: 3.0665, Perplexity: 21.4658
Epoch [1/3], Step [589/3236], Loss: 2.7551, Perplexity: 15.7225
Epoch [1/3], Step [590/3236], Loss: 2.9974, Perplexity: 20.0334
Epoch [1/3], Step [591/3236], Loss: 2.8889, Perplexity: 17.9729
Epoch [1/3], Step [592/3236], Loss: 3.0494, Perplexity: 21.1027
Epoch [1/3], Step [593/3236], Loss: 2.8924, Perplexity: 18.0369
Epoch [1/3], Step [594/3236], Loss: 3.0008, Perplexity: 20.1009
Epoch [1/3], Step [595/3236], Loss: 2.9333, Perplexity: 18.7888
Epoch [1/3], Step [596/3236], Loss: 2.9468, Perplexity: 19.0440
Epoch [1/3], Step [597/3236], Loss: 3.1158, Perplexity: 22.5521
Epoch [1/3], Step [598/3236], Loss: 3.0798, Perplexity: 21.7536
Epoch [1/3], Step [599/3236], Loss: 3.0428, Perplexity: 20.9641
Epoch [1/3], Step [600/3236], Loss: 3.4334, Perplexity: 30.9808
Epoch [1/3], Step [601/3236], Loss: 2.9844, Perplexity: 19.7756
Epoch [1/3], Step [602/3236], Loss: 3.3333, Perplexity: 28.0301
Epoch [1/3], Step [603/3236], Loss: 2.9104, Perplexity: 18.3639
Epoch [1/3], Step [604/3236], Loss: 2.8089, Perplexity: 16.5908
Epoch [1/3], Step [605/3236], Loss: 2.9325, Perplexity: 18.7747
Epoch [1/3], Step [606/3236], Loss: 2.6438, Perplexity: 14.0661
Epoch [1/3], Step [607/3236], Loss: 2.8879, Perplexity: 17.9547
Epoch [1/3], Step [608/3236], Loss: 3.6907, Perplexity: 40.0718
Epoch [1/3], Step [609/3236], Loss: 2.9514, Perplexity: 19.1320
Epoch [1/3], Step [610/3236], Loss: 3.2335, Perplexity: 25.3683
Epoch [1/3], Step [611/3236], Loss: 2.9268, Perplexity: 18.6683
Epoch [1/3], Step [612/3236], Loss: 2.9810, Perplexity: 19.7072
Epoch [1/3], Step [613/3236], Loss: 2.8173, Perplexity: 16.7320
Epoch [1/3], Step [614/3236], Loss: 3.0647, Perplexity: 21.4285
Epoch [1/3], Step [615/3236], Loss: 3.3390, Perplexity: 28.1917
Epoch [1/3], Step [616/3236], Loss: 3.0478, Perplexity: 21.0689
Epoch [1/3], Step [617/3236], Loss: 3.5557, Perplexity: 35.0121
Epoch [1/3], Step [618/3236], Loss: 2.9994, Perplexity: 20.0734
Epoch [1/3], Step [619/3236], Loss: 3.4499, Perplexity: 31.4962
Epoch [1/3], Step [620/3236], Loss: 3.0856, Perplexity: 21.8808
Epoch [1/3], Step [621/3236], Loss: 3.1264, Perplexity: 22.7926
Epoch [1/3], Step [622/3236], Loss: 3.1785, Perplexity: 24.0097
Epoch [1/3], Step [623/3236], Loss: 3.2422, Perplexity: 25.5888
Epoch [1/3], Step [624/3236], Loss: 2.8827, Perplexity: 17.8618
Epoch [1/3], Step [625/3236], Loss: 3.0634, Perplexity: 21.3998
Epoch [1/3], Step [626/3236], Loss: 3.2150, Perplexity: 24.9030
Epoch [1/3], Step [627/3236], Loss: 3.2357, Perplexity: 25.4236
Epoch [1/3], Step [628/3236], Loss: 3.0666, Perplexity: 21.4688
Epoch [1/3], Step [629/3236], Loss: 3.4814, Perplexity: 32.5061
Epoch [1/3], Step [630/3236], Loss: 2.9719, Perplexity: 19.5285
Epoch [1/3], Step [631/3236], Loss: 3.0062, Perplexity: 20.2102
Epoch [1/3], Step [632/3236], Loss: 3.1060, Perplexity: 22.3323
Epoch [1/3], Step [633/3236], Loss: 3.2881, Perplexity: 26.7921
Epoch [1/3], Step [634/3236], Loss: 3.0194, Perplexity: 20.4795
Epoch [1/3], Step [635/3236], Loss: 3.0102, Perplexity: 20.2910
Epoch [1/3], Step [636/3236], Loss: 2.9658, Perplexity: 19.4111
Epoch [1/3], Step [637/3236], Loss: 2.8641, Perplexity: 17.5326
Epoch [1/3], Step [638/3236], Loss: 2.9813, Perplexity: 19.7138
Epoch [1/3], Step [639/3236], Loss: 3.0147, Perplexity: 20.3821
Epoch [1/3], Step [640/3236], Loss: 3.0401, Perplexity: 20.9081
Epoch [1/3], Step [641/3236], Loss: 3.3188, Perplexity: 27.6259
Epoch [1/3], Step [642/3236], Loss: 2.8626, Perplexity: 17.5078
Epoch [1/3], Step [643/3236], Loss: 2.9795, Perplexity: 19.6778
Epoch [1/3], Step [644/3236], Loss: 3.0151, Perplexity: 20.3910
Epoch [1/3], Step [645/3236], Loss: 3.0133, Perplexity: 20.3551
Epoch [1/3], Step [646/3236], Loss: 3.0329, Perplexity: 20.7575
Epoch [1/3], Step [647/3236], Loss: 2.8096, Perplexity: 16.6036
Epoch [1/3], Step [648/3236], Loss: 2.9128, Perplexity: 18.4079
Epoch [1/3], Step [649/3236], Loss: 3.0306, Perplexity: 20.7094
Epoch [1/3], Step [650/3236], Loss: 2.8431, Perplexity: 17.1685
Epoch [1/3], Step [651/3236], Loss: 2.9709, Perplexity: 19.5090
Epoch [1/3], Step [652/3236], Loss: 2.9102, Perplexity: 18.3604
Epoch [1/3], Step [653/3236], Loss: 3.4424, Perplexity: 31.2633
Epoch [1/3], Step [654/3236], Loss: 2.9805, Perplexity: 19.6976
Epoch [1/3], Step [655/3236], Loss: 2.9707, Perplexity: 19.5057
Epoch [1/3], Step [656/3236], Loss: 3.2333, Perplexity: 25.3635
Epoch [1/3], Step [657/3236], Loss: 3.0122, Perplexity: 20.3330
Epoch [1/3], Step [658/3236], Loss: 3.1044, Perplexity: 22.2968
Epoch [1/3], Step [659/3236], Loss: 3.1941, Perplexity: 24.3873
Epoch [1/3], Step [660/3236], Loss: 2.8938, Perplexity: 18.0621
Epoch [1/3], Step [661/3236], Loss: 3.2486, Perplexity: 25.7543
Epoch [1/3], Step [662/3236], Loss: 3.6519, Perplexity: 38.5489
Epoch [1/3], Step [663/3236], Loss: 2.9427, Perplexity: 18.9665
Epoch [1/3], Step [664/3236], Loss: 2.8854, Perplexity: 17.9114
Epoch [1/3], Step [665/3236], Loss: 3.7665, Perplexity: 43.2283
Epoch [1/3], Step [666/3236], Loss: 2.9400, Perplexity: 18.9160
Epoch [1/3], Step [667/3236], Loss: 2.8787, Perplexity: 17.7910
Epoch [1/3], Step [668/3236], Loss: 2.9994, Perplexity: 20.0727
Epoch [1/3], Step [669/3236], Loss: 2.8281, Perplexity: 16.9137
Epoch [1/3], Step [670/3236], Loss: 2.8538, Perplexity: 17.3540
Epoch [1/3], Step [671/3236], Loss: 3.1478, Perplexity: 23.2859
Epoch [1/3], Step [672/3236], Loss: 3.0667, Perplexity: 21.4710
Epoch [1/3], Step [673/3236], Loss: 3.0821, Perplexity: 21.8034
Epoch [1/3], Step [674/3236], Loss: 3.0650, Perplexity: 21.4355
Epoch [1/3], Step [675/3236], Loss: 2.9674, Perplexity: 19.4407
Epoch [1/3], Step [676/3236], Loss: 2.9773, Perplexity: 19.6354
Epoch [1/3], Step [677/3236], Loss: 2.8304, Perplexity: 16.9524
Epoch [1/3], Step [678/3236], Loss: 2.9158, Perplexity: 18.4644
Epoch [1/3], Step [679/3236], Loss: 2.6886, Perplexity: 14.7107
Epoch [1/3], Step [680/3236], Loss: 2.8394, Perplexity: 17.1051
Epoch [1/3], Step [681/3236], Loss: 2.9582, Perplexity: 19.2632
Epoch [1/3], Step [682/3236], Loss: 2.8847, Perplexity: 17.8988
Epoch [1/3], Step [683/3236], Loss: 2.8860, Perplexity: 17.9223
Epoch [1/3], Step [684/3236], Loss: 3.1201, Perplexity: 22.6493
Epoch [1/3], Step [685/3236], Loss: 2.9200, Perplexity: 18.5417
Epoch [1/3], Step [686/3236], Loss: 2.9877, Perplexity: 19.8395
Epoch [1/3], Step [687/3236], Loss: 2.8934, Perplexity: 18.0549
Epoch [1/3], Step [688/3236], Loss: 3.1053, Perplexity: 22.3151
Epoch [1/3], Step [689/3236], Loss: 2.7365, Perplexity: 15.4332
Epoch [1/3], Step [690/3236], Loss: 3.7845, Perplexity: 44.0115
Epoch [1/3], Step [691/3236], Loss: 2.9026, Perplexity: 18.2217
Epoch [1/3], Step [692/3236], Loss: 2.8648, Perplexity: 17.5450
Epoch [1/3], Step [693/3236], Loss: 2.8528, Perplexity: 17.3366
Epoch [1/3], Step [694/3236], Loss: 3.0784, Perplexity: 21.7236
Epoch [1/3], Step [695/3236], Loss: 2.8322, Perplexity: 16.9824
Epoch [1/3], Step [696/3236], Loss: 2.9726, Perplexity: 19.5427
Epoch [1/3], Step [697/3236], Loss: 2.9949, Perplexity: 19.9832
Epoch [1/3], Step [698/3236], Loss: 3.0182, Perplexity: 20.4551
Epoch [1/3], Step [699/3236], Loss: 2.9871, Perplexity: 19.8272
Epoch [1/3], Step [700/3236], Loss: 3.0217, Perplexity: 20.5255
Epoch [1/3], Step [701/3236], Loss: 2.8337, Perplexity: 17.0084
Epoch [1/3], Step [702/3236], Loss: 3.1129, Perplexity: 22.4858
Epoch [1/3], Step [703/3236], Loss: 2.8002, Perplexity: 16.4479
Epoch [1/3], Step [704/3236], Loss: 2.9070, Perplexity: 18.3020
Epoch [1/3], Step [705/3236], Loss: 2.7436, Perplexity: 15.5427
Epoch [1/3], Step [706/3236], Loss: 3.3622, Perplexity: 28.8534
Epoch [1/3], Step [707/3236], Loss: 2.7377, Perplexity: 15.4516
Epoch [1/3], Step [708/3236], Loss: 2.8651, Perplexity: 17.5508
Epoch [1/3], Step [709/3236], Loss: 2.8832, Perplexity: 17.8716
Epoch [1/3], Step [710/3236], Loss: 3.1702, Perplexity: 23.8128
Epoch [1/3], Step [711/3236], Loss: 2.9184, Perplexity: 18.5111
Epoch [1/3], Step [712/3236], Loss: 3.2056, Perplexity: 24.6714
Epoch [1/3], Step [713/3236], Loss: 2.8647, Perplexity: 17.5444
Epoch [1/3], Step [714/3236], Loss: 3.0800, Perplexity: 21.7575
Epoch [1/3], Step [715/3236], Loss: 2.9860, Perplexity: 19.8062
Epoch [1/3], Step [716/3236], Loss: 3.2609, Perplexity: 26.0718
Epoch [1/3], Step [717/3236], Loss: 2.8915, Perplexity: 18.0211
Epoch [1/3], Step [718/3236], Loss: 2.7804, Perplexity: 16.1249
Epoch [1/3], Step [719/3236], Loss: 2.8816, Perplexity: 17.8422
Epoch [1/3], Step [720/3236], Loss: 2.5759, Perplexity: 13.1435
Epoch [1/3], Step [721/3236], Loss: 2.8522, Perplexity: 17.3257
Epoch [1/3], Step [722/3236], Loss: 2.6952, Perplexity: 14.8081
Epoch [1/3], Step [723/3236], Loss: 2.8874, Perplexity: 17.9466
Epoch [1/3], Step [724/3236], Loss: 3.1312, Perplexity: 22.9020
Epoch [1/3], Step [725/3236], Loss: 2.7569, Perplexity: 15.7512
Epoch [1/3], Step [726/3236], Loss: 2.8532, Perplexity: 17.3425
Epoch [1/3], Step [727/3236], Loss: 2.8754, Perplexity: 17.7323
Epoch [1/3], Step [728/3236], Loss: 2.8815, Perplexity: 17.8414
Epoch [1/3], Step [729/3236], Loss: 2.6185, Perplexity: 13.7156
Epoch [1/3], Step [730/3236], Loss: 2.7917, Perplexity: 16.3092
Epoch [1/3], Step [731/3236], Loss: 2.9114, Perplexity: 18.3823
Epoch [1/3], Step [732/3236], Loss: 2.9020, Perplexity: 18.2110
Epoch [1/3], Step [733/3236], Loss: 3.0667, Perplexity: 21.4716
Epoch [1/3], Step [734/3236], Loss: 3.0283, Perplexity: 20.6616
Epoch [1/3], Step [735/3236], Loss: 3.3465, Perplexity: 28.4037
Epoch [1/3], Step [736/3236], Loss: 2.7320, Perplexity: 15.3629
Epoch [1/3], Step [737/3236], Loss: 2.9496, Perplexity: 19.0985
Epoch [1/3], Step [738/3236], Loss: 2.8309, Perplexity: 16.9602
Epoch [1/3], Step [739/3236], Loss: 2.9157, Perplexity: 18.4622
Epoch [1/3], Step [740/3236], Loss: 3.2892, Perplexity: 26.8209
Epoch [1/3], Step [741/3236], Loss: 2.8166, Perplexity: 16.7197
Epoch [1/3], Step [742/3236], Loss: 3.0774, Perplexity: 21.7020
Epoch [1/3], Step [743/3236], Loss: 3.2096, Perplexity: 24.7699
Epoch [1/3], Step [744/3236], Loss: 3.2951, Perplexity: 26.9812
Epoch [1/3], Step [745/3236], Loss: 2.8587, Perplexity: 17.4384
Epoch [1/3], Step [746/3236], Loss: 3.1125, Perplexity: 22.4761
Epoch [1/3], Step [747/3236], Loss: 2.8654, Perplexity: 17.5554
Epoch [1/3], Step [748/3236], Loss: 2.8876, Perplexity: 17.9499
Epoch [1/3], Step [749/3236], Loss: 2.8558, Perplexity: 17.3878
Epoch [1/3], Step [750/3236], Loss: 2.7952, Perplexity: 16.3659
Epoch [1/3], Step [751/3236], Loss: 2.9779, Perplexity: 19.6462
Epoch [1/3], Step [752/3236], Loss: 2.7929, Perplexity: 16.3291
Epoch [1/3], Step [753/3236], Loss: 3.2362, Perplexity: 25.4372
Epoch [1/3], Step [754/3236], Loss: 2.6972, Perplexity: 14.8386
Epoch [1/3], Step [755/3236], Loss: 2.8370, Perplexity: 17.0645
Epoch [1/3], Step [756/3236], Loss: 3.0466, Perplexity: 21.0433
Epoch [1/3], Step [757/3236], Loss: 3.0188, Perplexity: 20.4662
Epoch [1/3], Step [758/3236], Loss: 2.8614, Perplexity: 17.4864
Epoch [1/3], Step [759/3236], Loss: 2.9053, Perplexity: 18.2711
Epoch [1/3], Step [760/3236], Loss: 2.7732, Perplexity: 16.0106
Epoch [1/3], Step [761/3236], Loss: 2.8456, Perplexity: 17.2114
Epoch [1/3], Step [762/3236], Loss: 2.9437, Perplexity: 18.9868
Epoch [1/3], Step [763/3236], Loss: 2.8160, Perplexity: 16.7091
Epoch [1/3], Step [764/3236], Loss: 2.7582, Perplexity: 15.7711
Epoch [1/3], Step [765/3236], Loss: 2.8130, Perplexity: 16.6595
Epoch [1/3], Step [766/3236], Loss: 2.7884, Perplexity: 16.2547
Epoch [1/3], Step [767/3236], Loss: 2.7506, Perplexity: 15.6517
Epoch [1/3], Step [768/3236], Loss: 2.7973, Perplexity: 16.4002
Epoch [1/3], Step [769/3236], Loss: 2.8793, Perplexity: 17.8013
Epoch [1/3], Step [770/3236], Loss: 2.9800, Perplexity: 19.6884
Epoch [1/3], Step [771/3236], Loss: 2.7052, Perplexity: 14.9571
Epoch [1/3], Step [772/3236], Loss: 2.8774, Perplexity: 17.7679
Epoch [1/3], Step [773/3236], Loss: 2.7005, Perplexity: 14.8866
Epoch [1/3], Step [774/3236], Loss: 2.7952, Perplexity: 16.3665
Epoch [1/3], Step [775/3236], Loss: 2.8126, Perplexity: 16.6523
Epoch [1/3], Step [776/3236], Loss: 2.8077, Perplexity: 16.5724
Epoch [1/3], Step [777/3236], Loss: 2.8769, Perplexity: 17.7586
Epoch [1/3], Step [778/3236], Loss: 2.8200, Perplexity: 16.7770
Epoch [1/3], Step [779/3236], Loss: 2.6814, Perplexity: 14.6050
Epoch [1/3], Step [780/3236], Loss: 3.1490, Perplexity: 23.3136
Epoch [1/3], Step [781/3236], Loss: 2.5938, Perplexity: 13.3799
Epoch [1/3], Step [782/3236], Loss: 2.7908, Perplexity: 16.2942
Epoch [1/3], Step [783/3236], Loss: 2.8472, Perplexity: 17.2393
Epoch [1/3], Step [784/3236], Loss: 2.9147, Perplexity: 18.4429
Epoch [1/3], Step [785/3236], Loss: 2.7867, Perplexity: 16.2281
Epoch [1/3], Step [786/3236], Loss: 2.8382, Perplexity: 17.0847
Epoch [1/3], Step [787/3236], Loss: 2.7561, Perplexity: 15.7388
Epoch [1/3], Step [788/3236], Loss: 3.0030, Perplexity: 20.1453
Epoch [1/3], Step [789/3236], Loss: 2.7982, Perplexity: 16.4155
Epoch [1/3], Step [790/3236], Loss: 3.7346, Perplexity: 41.8719
Epoch [1/3], Step [791/3236], Loss: 2.8019, Perplexity: 16.4756
Epoch [1/3], Step [792/3236], Loss: 2.6922, Perplexity: 14.7641
Epoch [1/3], Step [793/3236], Loss: 2.9362, Perplexity: 18.8449
Epoch [1/3], Step [794/3236], Loss: 2.8707, Perplexity: 17.6502
Epoch [1/3], Step [795/3236], Loss: 2.9171, Perplexity: 18.4882
Epoch [1/3], Step [796/3236], Loss: 2.7108, Perplexity: 15.0409
Epoch [1/3], Step [797/3236], Loss: 2.9720, Perplexity: 19.5309
Epoch [1/3], Step [798/3236], Loss: 2.8097, Perplexity: 16.6056
Epoch [1/3], Step [799/3236], Loss: 2.9914, Perplexity: 19.9139
Epoch [1/3], Step [800/3236], Loss: 3.4460, Perplexity: 31.3745
Epoch [1/3], Step [801/3236], Loss: 2.8354, Perplexity: 17.0371
Epoch [1/3], Step [802/3236], Loss: 3.8516, Perplexity: 47.0683
Epoch [1/3], Step [803/3236], Loss: 2.8393, Perplexity: 17.1033
Epoch [1/3], Step [804/3236], Loss: 2.9585, Perplexity: 19.2694
Epoch [1/3], Step [805/3236], Loss: 2.8007, Perplexity: 16.4555
Epoch [1/3], Step [806/3236], Loss: 2.9421, Perplexity: 18.9550
Epoch [1/3], Step [807/3236], Loss: 2.7587, Perplexity: 15.7792
Epoch [1/3], Step [808/3236], Loss: 2.8495, Perplexity: 17.2789
Epoch [1/3], Step [809/3236], Loss: 2.9405, Perplexity: 18.9257
Epoch [1/3], Step [810/3236], Loss: 2.7043, Perplexity: 14.9435
Epoch [1/3], Step [811/3236], Loss: 2.9957, Perplexity: 19.9993
Epoch [1/3], Step [812/3236], Loss: 2.7584, Perplexity: 15.7743
Epoch [1/3], Step [813/3236], Loss: 2.6503, Perplexity: 14.1582
Epoch [1/3], Step [814/3236], Loss: 2.8129, Perplexity: 16.6590
Epoch [1/3], Step [815/3236], Loss: 2.9199, Perplexity: 18.5394
Epoch [1/3], Step [816/3236], Loss: 2.6886, Perplexity: 14.7115
Epoch [1/3], Step [817/3236], Loss: 2.8255, Perplexity: 16.8691
Epoch [1/3], Step [818/3236], Loss: 3.2042, Perplexity: 24.6368
Epoch [1/3], Step [819/3236], Loss: 2.7061, Perplexity: 14.9715
Epoch [1/3], Step [820/3236], Loss: 2.9081, Perplexity: 18.3215
Epoch [1/3], Step [821/3236], Loss: 2.7864, Perplexity: 16.2232
Epoch [1/3], Step [822/3236], Loss: 2.9376, Perplexity: 18.8706
Epoch [1/3], Step [823/3236], Loss: 2.7820, Perplexity: 16.1508
Epoch [1/3], Step [824/3236], Loss: 3.2008, Perplexity: 24.5525
Epoch [1/3], Step [825/3236], Loss: 2.8232, Perplexity: 16.8301
Epoch [1/3], Step [826/3236], Loss: 2.6852, Perplexity: 14.6613
Epoch [1/3], Step [827/3236], Loss: 2.7927, Perplexity: 16.3244
Epoch [1/3], Step [828/3236], Loss: 2.7304, Perplexity: 15.3386
Epoch [1/3], Step [829/3236], Loss: 3.1283, Perplexity: 22.8345
Epoch [1/3], Step [830/3236], Loss: 2.8350, Perplexity: 17.0299
Epoch [1/3], Step [831/3236], Loss: 2.9350, Perplexity: 18.8208
Epoch [1/3], Step [832/3236], Loss: 3.4094, Perplexity: 30.2475
Epoch [1/3], Step [833/3236], Loss: 2.6525, Perplexity: 14.1891
Epoch [1/3], Step [834/3236], Loss: 2.9953, Perplexity: 19.9909
Epoch [1/3], Step [835/3236], Loss: 3.0369, Perplexity: 20.8407
Epoch [1/3], Step [836/3236], Loss: 2.8906, Perplexity: 18.0038
Epoch [1/3], Step [837/3236], Loss: 2.9593, Perplexity: 19.2848
Epoch [1/3], Step [838/3236], Loss: 3.1723, Perplexity: 23.8614
Epoch [1/3], Step [839/3236], Loss: 2.6558, Perplexity: 14.2370
Epoch [1/3], Step [840/3236], Loss: 2.9573, Perplexity: 19.2459
Epoch [1/3], Step [841/3236], Loss: 2.7959, Perplexity: 16.3778
Epoch [1/3], Step [842/3236], Loss: 2.8474, Perplexity: 17.2429
Epoch [1/3], Step [843/3236], Loss: 2.9036, Perplexity: 18.2401
Epoch [1/3], Step [844/3236], Loss: 2.6988, Perplexity: 14.8617
Epoch [1/3], Step [845/3236], Loss: 2.7209, Perplexity: 15.1933
Epoch [1/3], Step [846/3236], Loss: 2.7797, Perplexity: 16.1143
Epoch [1/3], Step [847/3236], Loss: 2.5914, Perplexity: 13.3485
Epoch [1/3], Step [848/3236], Loss: 2.7366, Perplexity: 15.4350
Epoch [1/3], Step [849/3236], Loss: 2.9229, Perplexity: 18.5951
Epoch [1/3], Step [850/3236], Loss: 2.7428, Perplexity: 15.5299
Epoch [1/3], Step [851/3236], Loss: 2.6923, Perplexity: 14.7656
Epoch [1/3], Step [852/3236], Loss: 2.6822, Perplexity: 14.6179
Epoch [1/3], Step [853/3236], Loss: 2.8745, Perplexity: 17.7170
Epoch [1/3], Step [854/3236], Loss: 2.8339, Perplexity: 17.0124
Epoch [1/3], Step [855/3236], Loss: 2.8170, Perplexity: 16.7273
Epoch [1/3], Step [856/3236], Loss: 2.6425, Perplexity: 14.0480
Epoch [1/3], Step [857/3236], Loss: 2.6589, Perplexity: 14.2806
Epoch [1/3], Step [858/3236], Loss: 2.6965, Perplexity: 14.8273
Epoch [1/3], Step [859/3236], Loss: 2.7729, Perplexity: 16.0046
Epoch [1/3], Step [860/3236], Loss: 2.7460, Perplexity: 15.5800
Epoch [1/3], Step [861/3236], Loss: 2.9927, Perplexity: 19.9395
Epoch [1/3], Step [862/3236], Loss: 2.7531, Perplexity: 15.6913
Epoch [1/3], Step [863/3236], Loss: 3.6542, Perplexity: 38.6354
Epoch [1/3], Step [864/3236], Loss: 2.7441, Perplexity: 15.5511
Epoch [1/3], Step [865/3236], Loss: 2.8720, Perplexity: 17.6723
Epoch [1/3], Step [866/3236], Loss: 2.9451, Perplexity: 19.0121
Epoch [1/3], Step [867/3236], Loss: 2.8268, Perplexity: 16.8920
Epoch [1/3], Step [868/3236], Loss: 2.8857, Perplexity: 17.9153
Epoch [1/3], Step [869/3236], Loss: 2.7509, Perplexity: 15.6569
Epoch [1/3], Step [870/3236], Loss: 2.7079, Perplexity: 14.9982
Epoch [1/3], Step [871/3236], Loss: 2.6553, Perplexity: 14.2290
Epoch [1/3], Step [872/3236], Loss: 2.6158, Perplexity: 13.6779
Epoch [1/3], Step [873/3236], Loss: 3.0600, Perplexity: 21.3269
Epoch [1/3], Step [874/3236], Loss: 2.9437, Perplexity: 18.9865
Epoch [1/3], Step [875/3236], Loss: 2.5929, Perplexity: 13.3688
Epoch [1/3], Step [876/3236], Loss: 2.5483, Perplexity: 12.7847
Epoch [1/3], Step [877/3236], Loss: 2.7057, Perplexity: 14.9645
Epoch [1/3], Step [878/3236], Loss: 2.9063, Perplexity: 18.2888
Epoch [1/3], Step [879/3236], Loss: 2.6345, Perplexity: 13.9364
Epoch [1/3], Step [880/3236], Loss: 3.0275, Perplexity: 20.6450
Epoch [1/3], Step [881/3236], Loss: 2.6879, Perplexity: 14.7001
Epoch [1/3], Step [882/3236], Loss: 2.6297, Perplexity: 13.8694
Epoch [1/3], Step [883/3236], Loss: 2.9339, Perplexity: 18.8011
Epoch [1/3], Step [884/3236], Loss: 2.7206, Perplexity: 15.1902
Epoch [1/3], Step [885/3236], Loss: 2.9351, Perplexity: 18.8235
Epoch [1/3], Step [886/3236], Loss: 3.1631, Perplexity: 23.6426
Epoch [1/3], Step [887/3236], Loss: 2.7594, Perplexity: 15.7907
Epoch [1/3], Step [888/3236], Loss: 2.6755, Perplexity: 14.5196
Epoch [1/3], Step [889/3236], Loss: 2.7852, Perplexity: 16.2027
Epoch [1/3], Step [890/3236], Loss: 2.8985, Perplexity: 18.1478
Epoch [1/3], Step [891/3236], Loss: 2.5813, Perplexity: 13.2142
Epoch [1/3], Step [892/3236], Loss: 2.7128, Perplexity: 15.0712
Epoch [1/3], Step [893/3236], Loss: 2.7591, Perplexity: 15.7852
Epoch [1/3], Step [894/3236], Loss: 2.7094, Perplexity: 15.0200
Epoch [1/3], Step [895/3236], Loss: 2.8054, Perplexity: 16.5329
Epoch [1/3], Step [896/3236], Loss: 2.6428, Perplexity: 14.0531
Epoch [1/3], Step [897/3236], Loss: 2.8414, Perplexity: 17.1399
Epoch [1/3], Step [898/3236], Loss: 2.7542, Perplexity: 15.7078
Epoch [1/3], Step [899/3236], Loss: 2.8806, Perplexity: 17.8248
Epoch [1/3], Step [900/3236], Loss: 3.1789, Perplexity: 24.0203
Epoch [1/3], Step [901/3236], Loss: 2.8711, Perplexity: 17.6565
Epoch [1/3], Step [902/3236], Loss: 2.7891, Perplexity: 16.2665
Epoch [1/3], Step [903/3236], Loss: 2.6611, Perplexity: 14.3126
Epoch [1/3], Step [904/3236], Loss: 2.6236, Perplexity: 13.7850
Epoch [1/3], Step [905/3236], Loss: 3.0395, Perplexity: 20.8950
Epoch [1/3], Step [906/3236], Loss: 2.7066, Perplexity: 14.9781
Epoch [1/3], Step [907/3236], Loss: 3.2984, Perplexity: 27.0706
Epoch [1/3], Step [908/3236], Loss: 2.7804, Perplexity: 16.1257
Epoch [1/3], Step [909/3236], Loss: 2.8034, Perplexity: 16.5005
Epoch [1/3], Step [910/3236], Loss: 2.7387, Perplexity: 15.4666
Epoch [1/3], Step [911/3236], Loss: 3.1512, Perplexity: 23.3634
Epoch [1/3], Step [912/3236], Loss: 2.6574, Perplexity: 14.2597
Epoch [1/3], Step [913/3236], Loss: 2.8272, Perplexity: 16.8988
Epoch [1/3], Step [914/3236], Loss: 2.8784, Perplexity: 17.7863
Epoch [1/3], Step [915/3236], Loss: 2.9021, Perplexity: 18.2124
Epoch [1/3], Step [916/3236], Loss: 2.6719, Perplexity: 14.4680
Epoch [1/3], Step [917/3236], Loss: 2.6910, Perplexity: 14.7466
Epoch [1/3], Step [918/3236], Loss: 2.6315, Perplexity: 13.8939
Epoch [1/3], Step [919/3236], Loss: 2.7486, Perplexity: 15.6205
Epoch [1/3], Step [920/3236], Loss: 2.8519, Perplexity: 17.3203
Epoch [1/3], Step [921/3236], Loss: 2.8599, Perplexity: 17.4591
Epoch [1/3], Step [922/3236], Loss: 2.8785, Perplexity: 17.7877
Epoch [1/3], Step [923/3236], Loss: 2.9287, Perplexity: 18.7025
Epoch [1/3], Step [924/3236], Loss: 2.6594, Perplexity: 14.2876
Epoch [1/3], Step [925/3236], Loss: 3.7789, Perplexity: 43.7696
Epoch [1/3], Step [926/3236], Loss: 2.5910, Perplexity: 13.3435
Epoch [1/3], Step [927/3236], Loss: 2.6072, Perplexity: 13.5607
Epoch [1/3], Step [928/3236], Loss: 2.6752, Perplexity: 14.5147
Epoch [1/3], Step [929/3236], Loss: 2.5252, Perplexity: 12.4934
Epoch [1/3], Step [930/3236], Loss: 2.6529, Perplexity: 14.1949
Epoch [1/3], Step [931/3236], Loss: 2.8349, Perplexity: 17.0282
Epoch [1/3], Step [932/3236], Loss: 2.7502, Perplexity: 15.6465
Epoch [1/3], Step [933/3236], Loss: 3.0746, Perplexity: 21.6414
Epoch [1/3], Step [934/3236], Loss: 2.7649, Perplexity: 15.8770
Epoch [1/3], Step [935/3236], Loss: 3.4656, Perplexity: 31.9955
Epoch [1/3], Step [936/3236], Loss: 2.5006, Perplexity: 12.1902
Epoch [1/3], Step [937/3236], Loss: 2.4600, Perplexity: 11.7051
Epoch [1/3], Step [938/3236], Loss: 2.5935, Perplexity: 13.3763
Epoch [1/3], Step [939/3236], Loss: 2.6653, Perplexity: 14.3729
Epoch [1/3], Step [940/3236], Loss: 2.7608, Perplexity: 15.8125
Epoch [1/3], Step [941/3236], Loss: 2.6175, Perplexity: 13.7008
Epoch [1/3], Step [942/3236], Loss: 2.8000, Perplexity: 16.4447
Epoch [1/3], Step [943/3236], Loss: 2.7286, Perplexity: 15.3110
Epoch [1/3], Step [944/3236], Loss: 2.7130, Perplexity: 15.0740
Epoch [1/3], Step [945/3236], Loss: 2.8561, Perplexity: 17.3937
Epoch [1/3], Step [946/3236], Loss: 2.9260, Perplexity: 18.6525
Epoch [1/3], Step [947/3236], Loss: 2.6846, Perplexity: 14.6517
Epoch [1/3], Step [948/3236], Loss: 2.6614, Perplexity: 14.3161
Epoch [1/3], Step [949/3236], Loss: 2.9372, Perplexity: 18.8637
Epoch [1/3], Step [950/3236], Loss: 2.7220, Perplexity: 15.2105
Epoch [1/3], Step [951/3236], Loss: 3.3506, Perplexity: 28.5194
Epoch [1/3], Step [952/3236], Loss: 2.7961, Perplexity: 16.3808
Epoch [1/3], Step [953/3236], Loss: 2.6636, Perplexity: 14.3485
Epoch [1/3], Step [954/3236], Loss: 3.0255, Perplexity: 20.6039
Epoch [1/3], Step [955/3236], Loss: 2.6536, Perplexity: 14.2046
Epoch [1/3], Step [956/3236], Loss: 2.5110, Perplexity: 12.3178
Epoch [1/3], Step [957/3236], Loss: 2.6999, Perplexity: 14.8775
Epoch [1/3], Step [958/3236], Loss: 2.9999, Perplexity: 20.0834
Epoch [1/3], Step [959/3236], Loss: 2.8498, Perplexity: 17.2837
Epoch [1/3], Step [960/3236], Loss: 2.8087, Perplexity: 16.5876
Epoch [1/3], Step [961/3236], Loss: 2.7229, Perplexity: 15.2241
Epoch [1/3], Step [962/3236], Loss: 2.7603, Perplexity: 15.8045
Epoch [1/3], Step [963/3236], Loss: 3.0488, Perplexity: 21.0900
Epoch [1/3], Step [964/3236], Loss: 2.6631, Perplexity: 14.3411
Epoch [1/3], Step [965/3236], Loss: 2.7867, Perplexity: 16.2273
Epoch [1/3], Step [966/3236], Loss: 2.7191, Perplexity: 15.1663
Epoch [1/3], Step [967/3236], Loss: 2.6535, Perplexity: 14.2044
Epoch [1/3], Step [968/3236], Loss: 2.9303, Perplexity: 18.7330
Epoch [1/3], Step [969/3236], Loss: 2.8735, Perplexity: 17.6984
Epoch [1/3], Step [970/3236], Loss: 2.6653, Perplexity: 14.3719
Epoch [1/3], Step [971/3236], Loss: 2.7148, Perplexity: 15.1010
Epoch [1/3], Step [972/3236], Loss: 2.7747, Perplexity: 16.0340
Epoch [1/3], Step [973/3236], Loss: 2.9097, Perplexity: 18.3505
Epoch [1/3], Step [974/3236], Loss: 2.7824, Perplexity: 16.1583
Epoch [1/3], Step [975/3236], Loss: 3.0247, Perplexity: 20.5875
Epoch [1/3], Step [976/3236], Loss: 2.6954, Perplexity: 14.8114
Epoch [1/3], Step [977/3236], Loss: 2.6822, Perplexity: 14.6172
Epoch [1/3], Step [978/3236], Loss: 2.9095, Perplexity: 18.3485
Epoch [1/3], Step [979/3236], Loss: 2.5461, Perplexity: 12.7569
Epoch [1/3], Step [980/3236], Loss: 2.7461, Perplexity: 15.5812
Epoch [1/3], Step [981/3236], Loss: 2.5722, Perplexity: 13.0952
Epoch [1/3], Step [982/3236], Loss: 2.6912, Perplexity: 14.7490
Epoch [1/3], Step [983/3236], Loss: 2.6845, Perplexity: 14.6504
Epoch [1/3], Step [984/3236], Loss: 2.7498, Perplexity: 15.6390
Epoch [1/3], Step [985/3236], Loss: 2.7639, Perplexity: 15.8615
Epoch [1/3], Step [986/3236], Loss: 2.8038, Perplexity: 16.5073
Epoch [1/3], Step [987/3236], Loss: 2.6141, Perplexity: 13.6556
Epoch [1/3], Step [988/3236], Loss: 2.5119, Perplexity: 12.3280
Epoch [1/3], Step [989/3236], Loss: 2.8617, Perplexity: 17.4912
Epoch [1/3], Step [990/3236], Loss: 2.8344, Perplexity: 17.0205
Epoch [1/3], Step [991/3236], Loss: 2.9171, Perplexity: 18.4872
Epoch [1/3], Step [992/3236], Loss: 2.6937, Perplexity: 14.7856
Epoch [1/3], Step [993/3236], Loss: 2.6077, Perplexity: 13.5673
Epoch [1/3], Step [994/3236], Loss: 3.0408, Perplexity: 20.9223
Epoch [1/3], Step [995/3236], Loss: 2.4954, Perplexity: 12.1270
Epoch [1/3], Step [996/3236], Loss: 2.5250, Perplexity: 12.4914
Epoch [1/3], Step [997/3236], Loss: 2.8896, Perplexity: 17.9861
Epoch [1/3], Step [998/3236], Loss: 2.7712, Perplexity: 15.9779
Epoch [1/3], Step [999/3236], Loss: 2.8441, Perplexity: 17.1861
Epoch [1/3], Step [1000/3236], Loss: 2.7010, Perplexity: 14.8951
Epoch [1/3], Step [1001/3236], Loss: 2.5857, Perplexity: 13.2724
Epoch [1/3], Step [1002/3236], Loss: 2.8806, Perplexity: 17.8244
Epoch [1/3], Step [1003/3236], Loss: 3.2066, Perplexity: 24.6942
Epoch [1/3], Step [1004/3236], Loss: 2.4951, Perplexity: 12.1235
Epoch [1/3], Step [1005/3236], Loss: 2.5126, Perplexity: 12.3365
Epoch [1/3], Step [1006/3236], Loss: 2.6428, Perplexity: 14.0518
Epoch [1/3], Step [1007/3236], Loss: 2.7709, Perplexity: 15.9735
Epoch [1/3], Step [1008/3236], Loss: 2.5954, Perplexity: 13.4019
Epoch [1/3], Step [1009/3236], Loss: 2.7003, Perplexity: 14.8844
Epoch [1/3], Step [1010/3236], Loss: 2.5345, Perplexity: 12.6098
Epoch [1/3], Step [1011/3236], Loss: 5.0858, Perplexity: 161.7016
Epoch [1/3], Step [1012/3236], Loss: 2.7556, Perplexity: 15.7306
Epoch [1/3], Step [1013/3236], Loss: 2.6853, Perplexity: 14.6625
Epoch [1/3], Step [1014/3236], Loss: 2.9704, Perplexity: 19.4989
Epoch [1/3], Step [1015/3236], Loss: 2.7691, Perplexity: 15.9440
Epoch [1/3], Step [1016/3236], Loss: 2.7958, Perplexity: 16.3759
Epoch [1/3], Step [1017/3236], Loss: 2.6749, Perplexity: 14.5112
Epoch [1/3], Step [1018/3236], Loss: 2.8519, Perplexity: 17.3214
Epoch [1/3], Step [1019/3236], Loss: 3.0124, Perplexity: 20.3364
Epoch [1/3], Step [1020/3236], Loss: 2.5960, Perplexity: 13.4096
Epoch [1/3], Step [1021/3236], Loss: 2.7261, Perplexity: 15.2726
Epoch [1/3], Step [1022/3236], Loss: 2.7586, Perplexity: 15.7778
Epoch [1/3], Step [1023/3236], Loss: 2.5733, Perplexity: 13.1096
Epoch [1/3], Step [1024/3236], Loss: 2.6830, Perplexity: 14.6285
Epoch [1/3], Step [1025/3236], Loss: 3.7203, Perplexity: 41.2761
Epoch [1/3], Step [1026/3236], Loss: 2.6380, Perplexity: 13.9854
Epoch [1/3], Step [1027/3236], Loss: 2.7048, Perplexity: 14.9511
Epoch [1/3], Step [1028/3236], Loss: 2.6231, Perplexity: 13.7789
Epoch [1/3], Step [1029/3236], Loss: 2.5480, Perplexity: 12.7818
Epoch [1/3], Step [1030/3236], Loss: 2.7677, Perplexity: 15.9226
Epoch [1/3], Step [1031/3236], Loss: 2.6942, Perplexity: 14.7943
Epoch [1/3], Step [1032/3236], Loss: 2.5981, Perplexity: 13.4382
Epoch [1/3], Step [1033/3236], Loss: 2.7282, Perplexity: 15.3048
Epoch [1/3], Step [1034/3236], Loss: 2.7357, Perplexity: 15.4201
Epoch [1/3], Step [1035/3236], Loss: 2.6944, Perplexity: 14.7961
Epoch [1/3], Step [1036/3236], Loss: 2.6021, Perplexity: 13.4921
Epoch [1/3], Step [1037/3236], Loss: 2.7962, Perplexity: 16.3824
Epoch [1/3], Step [1038/3236], Loss: 2.8697, Perplexity: 17.6313
Epoch [1/3], Step [1039/3236], Loss: 2.5159, Perplexity: 12.3772
Epoch [1/3], Step [1040/3236], Loss: 2.6320, Perplexity: 13.9018
Epoch [1/3], Step [1041/3236], Loss: 2.5582, Perplexity: 12.9130
Epoch [1/3], Step [1042/3236], Loss: 2.7945, Perplexity: 16.3553
Epoch [1/3], Step [1043/3236], Loss: 2.8068, Perplexity: 16.5568
Epoch [1/3], Step [1044/3236], Loss: 3.1028, Perplexity: 22.2605
Epoch [1/3], Step [1045/3236], Loss: 2.8697, Perplexity: 17.6323
Epoch [1/3], Step [1046/3236], Loss: 2.6301, Perplexity: 13.8750
Epoch [1/3], Step [1047/3236], Loss: 2.7594, Perplexity: 15.7911
Epoch [1/3], Step [1048/3236], Loss: 2.6785, Perplexity: 14.5638
Epoch [1/3], Step [1049/3236], Loss: 2.7817, Perplexity: 16.1471
Epoch [1/3], Step [1050/3236], Loss: 2.8074, Perplexity: 16.5666
Epoch [1/3], Step [1051/3236], Loss: 2.6362, Perplexity: 13.9597
Epoch [1/3], Step [1052/3236], Loss: 2.6182, Perplexity: 13.7106
Epoch [1/3], Step [1053/3236], Loss: 3.5090, Perplexity: 33.4140
Epoch [1/3], Step [1054/3236], Loss: 2.4996, Perplexity: 12.1778
Epoch [1/3], Step [1055/3236], Loss: 2.7851, Perplexity: 16.2013
Epoch [1/3], Step [1056/3236], Loss: 2.9456, Perplexity: 19.0216
Epoch [1/3], Step [1057/3236], Loss: 2.6364, Perplexity: 13.9625
Epoch [1/3], Step [1058/3236], Loss: 2.6314, Perplexity: 13.8931
Epoch [1/3], Step [1059/3236], Loss: 2.7456, Perplexity: 15.5734
Epoch [1/3], Step [1060/3236], Loss: 2.4948, Perplexity: 12.1199
Epoch [1/3], Step [1061/3236], Loss: 2.6174, Perplexity: 13.6999
Epoch [1/3], Step [1062/3236], Loss: 2.5118, Perplexity: 12.3268
Epoch [1/3], Step [1063/3236], Loss: 2.7159, Perplexity: 15.1187
Epoch [1/3], Step [1064/3236], Loss: 2.6459, Perplexity: 14.0959
Epoch [1/3], Step [1065/3236], Loss: 2.6249, Perplexity: 13.8026
Epoch [1/3], Step [1066/3236], Loss: 2.5777, Perplexity: 13.1674
Epoch [1/3], Step [1067/3236], Loss: 2.6218, Perplexity: 13.7598
Epoch [1/3], Step [1068/3236], Loss: 2.5693, Perplexity: 13.0563
Epoch [1/3], Step [1069/3236], Loss: 2.5106, Perplexity: 12.3120
Epoch [1/3], Step [1070/3236], Loss: 2.7896, Perplexity: 16.2747
Epoch [1/3], Step [1071/3236], Loss: 2.5032, Perplexity: 12.2219
Epoch [1/3], Step [1072/3236], Loss: 2.7155, Perplexity: 15.1123
Epoch [1/3], Step [1073/3236], Loss: 3.4127, Perplexity: 30.3480
Epoch [1/3], Step [1074/3236], Loss: 2.8962, Perplexity: 18.1045
Epoch [1/3], Step [1075/3236], Loss: 2.6561, Perplexity: 14.2408
Epoch [1/3], Step [1076/3236], Loss: 2.7375, Perplexity: 15.4490
Epoch [1/3], Step [1077/3236], Loss: 2.7289, Perplexity: 15.3157
Epoch [1/3], Step [1078/3236], Loss: 2.7001, Perplexity: 14.8813
Epoch [1/3], Step [1079/3236], Loss: 2.6983, Perplexity: 14.8551
Epoch [1/3], Step [1080/3236], Loss: 2.8723, Perplexity: 17.6776
Epoch [1/3], Step [1081/3236], Loss: 2.5110, Perplexity: 12.3176
Epoch [1/3], Step [1082/3236], Loss: 2.6788, Perplexity: 14.5675
Epoch [1/3], Step [1083/3236], Loss: 2.8221, Perplexity: 16.8118
Epoch [1/3], Step [1084/3236], Loss: 2.9438, Perplexity: 18.9877
Epoch [1/3], Step [1085/3236], Loss: 2.5845, Perplexity: 13.2564
Epoch [1/3], Step [1086/3236], Loss: 2.6615, Perplexity: 14.3174
Epoch [1/3], Step [1087/3236], Loss: 2.5323, Perplexity: 12.5823
Epoch [1/3], Step [1088/3236], Loss: 2.6148, Perplexity: 13.6642
Epoch [1/3], Step [1089/3236], Loss: 3.0367, Perplexity: 20.8362
Epoch [1/3], Step [1090/3236], Loss: 2.7001, Perplexity: 14.8816
Epoch [1/3], Step [1091/3236], Loss: 2.6504, Perplexity: 14.1603
Epoch [1/3], Step [1092/3236], Loss: 2.7446, Perplexity: 15.5580
Epoch [1/3], Step [1093/3236], Loss: 2.6241, Perplexity: 13.7919
Epoch [1/3], Step [1094/3236], Loss: 2.8551, Perplexity: 17.3761
Epoch [1/3], Step [1095/3236], Loss: 2.4744, Perplexity: 11.8749
Epoch [1/3], Step [1096/3236], Loss: 2.6142, Perplexity: 13.6569
Epoch [1/3], Step [1097/3236], Loss: 2.5801, Perplexity: 13.1988
Epoch [1/3], Step [1098/3236], Loss: 2.7860, Perplexity: 16.2158
Epoch [1/3], Step [1099/3236], Loss: 2.6444, Perplexity: 14.0748
Epoch [1/3], Step [1100/3236], Loss: 2.7518, Perplexity: 15.6703
Epoch [1/3], Step [1101/3236], Loss: 2.5652, Perplexity: 13.0034
Epoch [1/3], Step [1102/3236], Loss: 2.7746, Perplexity: 16.0321
Epoch [1/3], Step [1103/3236], Loss: 2.8865, Perplexity: 17.9309
Epoch [1/3], Step [1104/3236], Loss: 2.5788, Perplexity: 13.1811
Epoch [1/3], Step [1105/3236], Loss: 2.8690, Perplexity: 17.6202
Epoch [1/3], Step [1106/3236], Loss: 2.7745, Perplexity: 16.0309
Epoch [1/3], Step [1107/3236], Loss: 2.5970, Perplexity: 13.4241
Epoch [1/3], Step [1108/3236], Loss: 2.8331, Perplexity: 16.9975
Epoch [1/3], Step [1109/3236], Loss: 2.6587, Perplexity: 14.2779
Epoch [1/3], Step [1110/3236], Loss: 2.5114, Perplexity: 12.3217
Epoch [1/3], Step [1111/3236], Loss: 2.9442, Perplexity: 18.9961
Epoch [1/3], Step [1112/3236], Loss: 2.6632, Perplexity: 14.3423
Epoch [1/3], Step [1113/3236], Loss: 2.6323, Perplexity: 13.9062
Epoch [1/3], Step [1114/3236], Loss: 3.1682, Perplexity: 23.7637
Epoch [1/3], Step [1115/3236], Loss: 2.5960, Perplexity: 13.4104
Epoch [1/3], Step [1116/3236], Loss: 2.6393, Perplexity: 14.0040
Epoch [1/3], Step [1117/3236], Loss: 2.5722, Perplexity: 13.0940
Epoch [1/3], Step [1118/3236], Loss: 2.5915, Perplexity: 13.3499
Epoch [1/3], Step [1119/3236], Loss: 2.8306, Perplexity: 16.9560
Epoch [1/3], Step [1120/3236], Loss: 2.5969, Perplexity: 13.4226
Epoch [1/3], Step [1121/3236], Loss: 2.8512, Perplexity: 17.3080
Epoch [1/3], Step [1122/3236], Loss: 2.8812, Perplexity: 17.8352
Epoch [1/3], Step [1123/3236], Loss: 2.8231, Perplexity: 16.8283
Epoch [1/3], Step [1124/3236], Loss: 3.0552, Perplexity: 21.2263
Epoch [1/3], Step [1125/3236], Loss: 2.6098, Perplexity: 13.5968
Epoch [1/3], Step [1126/3236], Loss: 2.5942, Perplexity: 13.3859
Epoch [1/3], Step [1127/3236], Loss: 2.6282, Perplexity: 13.8485
Epoch [1/3], Step [1128/3236], Loss: 2.8527, Perplexity: 17.3351
Epoch [1/3], Step [1129/3236], Loss: 2.7576, Perplexity: 15.7616
Epoch [1/3], Step [1130/3236], Loss: 3.0596, Perplexity: 21.3182
Epoch [1/3], Step [1131/3236], Loss: 2.5008, Perplexity: 12.1927
Epoch [1/3], Step [1132/3236], Loss: 2.5731, Perplexity: 13.1058
Epoch [1/3], Step [1133/3236], Loss: 2.7313, Perplexity: 15.3536
Epoch [1/3], Step [1134/3236], Loss: 2.5476, Perplexity: 12.7768
Epoch [1/3], Step [1135/3236], Loss: 2.3742, Perplexity: 10.7428
Epoch [1/3], Step [1136/3236], Loss: 2.5182, Perplexity: 12.4059
Epoch [1/3], Step [1137/3236], Loss: 2.5883, Perplexity: 13.3065
Epoch [1/3], Step [1138/3236], Loss: 2.6086, Perplexity: 13.5794
Epoch [1/3], Step [1139/3236], Loss: 2.7426, Perplexity: 15.5270
Epoch [1/3], Step [1140/3236], Loss: 2.6323, Perplexity: 13.9054
Epoch [1/3], Step [1141/3236], Loss: 2.7981, Perplexity: 16.4139
Epoch [1/3], Step [1142/3236], Loss: 2.7567, Perplexity: 15.7472
Epoch [1/3], Step [1143/3236], Loss: 2.4929, Perplexity: 12.0961
Epoch [1/3], Step [1144/3236], Loss: 2.8526, Perplexity: 17.3326
Epoch [1/3], Step [1145/3236], Loss: 2.6664, Perplexity: 14.3886
Epoch [1/3], Step [1146/3236], Loss: 2.6583, Perplexity: 14.2723
Epoch [1/3], Step [1147/3236], Loss: 2.8006, Perplexity: 16.4537
Epoch [1/3], Step [1148/3236], Loss: 2.5300, Perplexity: 12.5541
Epoch [1/3], Step [1149/3236], Loss: 2.6946, Perplexity: 14.8003
Epoch [1/3], Step [1150/3236], Loss: 2.6897, Perplexity: 14.7271
Epoch [1/3], Step [1151/3236], Loss: 2.4668, Perplexity: 11.7842
Epoch [1/3], Step [1152/3236], Loss: 2.6414, Perplexity: 14.0331
Epoch [1/3], Step [1153/3236], Loss: 2.7385, Perplexity: 15.4644
Epoch [1/3], Step [1154/3236], Loss: 2.8597, Perplexity: 17.4562
Epoch [1/3], Step [1155/3236], Loss: 2.6268, Perplexity: 13.8296
Epoch [1/3], Step [1156/3236], Loss: 2.9014, Perplexity: 18.1987
Epoch [1/3], Step [1157/3236], Loss: 2.5574, Perplexity: 12.9019
Epoch [1/3], Step [1158/3236], Loss: 2.6377, Perplexity: 13.9817
Epoch [1/3], Step [1159/3236], Loss: 2.5790, Perplexity: 13.1840
Epoch [1/3], Step [1160/3236], Loss: 2.6394, Perplexity: 14.0054
Epoch [1/3], Step [1161/3236], Loss: 2.7714, Perplexity: 15.9807
Epoch [1/3], Step [1162/3236], Loss: 2.6716, Perplexity: 14.4637
Epoch [1/3], Step [1163/3236], Loss: 2.7014, Perplexity: 14.9003
Epoch [1/3], Step [1164/3236], Loss: 2.9250, Perplexity: 18.6339
Epoch [1/3], Step [1165/3236], Loss: 2.7414, Perplexity: 15.5085
Epoch [1/3], Step [1166/3236], Loss: 2.6543, Perplexity: 14.2154
Epoch [1/3], Step [1167/3236], Loss: 2.6770, Perplexity: 14.5419
Epoch [1/3], Step [1168/3236], Loss: 2.6231, Perplexity: 13.7786
Epoch [1/3], Step [1169/3236], Loss: 2.6112, Perplexity: 13.6157
Epoch [1/3], Step [1170/3236], Loss: 2.6494, Perplexity: 14.1455
Epoch [1/3], Step [1171/3236], Loss: 2.6469, Perplexity: 14.1105
Epoch [1/3], Step [1172/3236], Loss: 2.9575, Perplexity: 19.2500
Epoch [1/3], Step [1173/3236], Loss: 2.6460, Perplexity: 14.0976
Epoch [1/3], Step [1174/3236], Loss: 2.6478, Perplexity: 14.1229
Epoch [1/3], Step [1175/3236], Loss: 2.9600, Perplexity: 19.2977
Epoch [1/3], Step [1176/3236], Loss: 2.7334, Perplexity: 15.3845
Epoch [1/3], Step [1177/3236], Loss: 2.6072, Perplexity: 13.5607
Epoch [1/3], Step [1178/3236], Loss: 2.5013, Perplexity: 12.1980
Epoch [1/3], Step [1179/3236], Loss: 2.5661, Perplexity: 13.0154
Epoch [1/3], Step [1180/3236], Loss: 2.7018, Perplexity: 14.9067
Epoch [1/3], Step [1181/3236], Loss: 2.9159, Perplexity: 18.4659
Epoch [1/3], Step [1182/3236], Loss: 2.5145, Perplexity: 12.3603
Epoch [1/3], Step [1183/3236], Loss: 2.6162, Perplexity: 13.6843
Epoch [1/3], Step [1184/3236], Loss: 2.7024, Perplexity: 14.9151
Epoch [1/3], Step [1185/3236], Loss: 2.8193, Perplexity: 16.7650
Epoch [1/3], Step [1186/3236], Loss: 2.3899, Perplexity: 10.9122
Epoch [1/3], Step [1187/3236], Loss: 3.1470, Perplexity: 23.2664
Epoch [1/3], Step [1188/3236], Loss: 2.7003, Perplexity: 14.8849
Epoch [1/3], Step [1189/3236], Loss: 2.5650, Perplexity: 13.0003
Epoch [1/3], Step [1190/3236], Loss: 2.4256, Perplexity: 11.3091
Epoch [1/3], Step [1191/3236], Loss: 2.4342, Perplexity: 11.4070
Epoch [1/3], Step [1192/3236], Loss: 2.6177, Perplexity: 13.7042
Epoch [1/3], Step [1193/3236], Loss: 2.6528, Perplexity: 14.1935
Epoch [1/3], Step [1194/3236], Loss: 2.4632, Perplexity: 11.7427
Epoch [1/3], Step [1195/3236], Loss: 2.5951, Perplexity: 13.3975
Epoch [1/3], Step [1196/3236], Loss: 2.5688, Perplexity: 13.0505
Epoch [1/3], Step [1197/3236], Loss: 2.5311, Perplexity: 12.5678
Epoch [1/3], Step [1198/3236], Loss: 2.7940, Perplexity: 16.3462
Epoch [1/3], Step [1199/3236], Loss: 2.9381, Perplexity: 18.8803
Epoch [1/3], Step [1200/3236], Loss: 3.2331, Perplexity: 25.3585
Epoch [1/3], Step [1201/3236], Loss: 2.4720, Perplexity: 11.8459
Epoch [1/3], Step [1202/3236], Loss: 2.6313, Perplexity: 13.8921
Epoch [1/3], Step [1203/3236], Loss: 2.5861, Perplexity: 13.2784
Epoch [1/3], Step [1204/3236], Loss: 2.6835, Perplexity: 14.6367
Epoch [1/3], Step [1205/3236], Loss: 2.6441, Perplexity: 14.0713
Epoch [1/3], Step [1206/3236], Loss: 2.6980, Perplexity: 14.8497
Epoch [1/3], Step [1207/3236], Loss: 2.5447, Perplexity: 12.7389
Epoch [1/3], Step [1208/3236], Loss: 2.5793, Perplexity: 13.1879
Epoch [1/3], Step [1209/3236], Loss: 2.4431, Perplexity: 11.5083
Epoch [1/3], Step [1210/3236], Loss: 2.5625, Perplexity: 12.9682
Epoch [1/3], Step [1211/3236], Loss: 2.7260, Perplexity: 15.2714
Epoch [1/3], Step [1212/3236], Loss: 2.4963, Perplexity: 12.1372
Epoch [1/3], Step [1213/3236], Loss: 2.4898, Perplexity: 12.0583
Epoch [1/3], Step [1214/3236], Loss: 2.6546, Perplexity: 14.2190
Epoch [1/3], Step [1215/3236], Loss: 2.4023, Perplexity: 11.0489
Epoch [1/3], Step [1216/3236], Loss: 2.4034, Perplexity: 11.0609
Epoch [1/3], Step [1217/3236], Loss: 2.5973, Perplexity: 13.4275
Epoch [1/3], Step [1218/3236], Loss: 2.4681, Perplexity: 11.7998
Epoch [1/3], Step [1219/3236], Loss: 2.3909, Perplexity: 10.9231
Epoch [1/3], Step [1220/3236], Loss: 2.5218, Perplexity: 12.4506
Epoch [1/3], Step [1221/3236], Loss: 3.0876, Perplexity: 21.9244
Epoch [1/3], Step [1222/3236], Loss: 2.4999, Perplexity: 12.1812
Epoch [1/3], Step [1223/3236], Loss: 2.6269, Perplexity: 13.8311
Epoch [1/3], Step [1224/3236], Loss: 2.2741, Perplexity: 9.7187
Epoch [1/3], Step [1225/3236], Loss: 2.6400, Perplexity: 14.0127
Epoch [1/3], Step [1226/3236], Loss: 2.8212, Perplexity: 16.7970
Epoch [1/3], Step [1227/3236], Loss: 2.6271, Perplexity: 13.8335
Epoch [1/3], Step [1228/3236], Loss: 2.6112, Perplexity: 13.6160
Epoch [1/3], Step [1229/3236], Loss: 2.5353, Perplexity: 12.6203
Epoch [1/3], Step [1230/3236], Loss: 2.5638, Perplexity: 12.9852
Epoch [1/3], Step [1231/3236], Loss: 2.4193, Perplexity: 11.2383
Epoch [1/3], Step [1232/3236], Loss: 2.6168, Perplexity: 13.6920
Epoch [1/3], Step [1233/3236], Loss: 2.7730, Perplexity: 16.0071
Epoch [1/3], Step [1234/3236], Loss: 2.5141, Perplexity: 12.3560
Epoch [1/3], Step [1235/3236], Loss: 2.5224, Perplexity: 12.4587
Epoch [1/3], Step [1236/3236], Loss: 2.4992, Perplexity: 12.1728
Epoch [1/3], Step [1237/3236], Loss: 3.2242, Perplexity: 25.1341
Epoch [1/3], Step [1238/3236], Loss: 2.5417, Perplexity: 12.7010
Epoch [1/3], Step [1239/3236], Loss: 2.4582, Perplexity: 11.6833
Epoch [1/3], Step [1240/3236], Loss: 3.1097, Perplexity: 22.4139
Epoch [1/3], Step [1241/3236], Loss: 2.6207, Perplexity: 13.7447
Epoch [1/3], Step [1242/3236], Loss: 3.0446, Perplexity: 21.0017
Epoch [1/3], Step [1243/3236], Loss: 2.3902, Perplexity: 10.9161
Epoch [1/3], Step [1244/3236], Loss: 2.6531, Perplexity: 14.1976
Epoch [1/3], Step [1245/3236], Loss: 2.6218, Perplexity: 13.7608
Epoch [1/3], Step [1246/3236], Loss: 2.4908, Perplexity: 12.0706
Epoch [1/3], Step [1247/3236], Loss: 2.6817, Perplexity: 14.6093
Epoch [1/3], Step [1248/3236], Loss: 2.9245, Perplexity: 18.6255
Epoch [1/3], Step [1249/3236], Loss: 2.5334, Perplexity: 12.5964
Epoch [1/3], Step [1250/3236], Loss: 2.4491, Perplexity: 11.5781
Epoch [1/3], Step [1251/3236], Loss: 2.6332, Perplexity: 13.9188
Epoch [1/3], Step [1252/3236], Loss: 2.7334, Perplexity: 15.3852
Epoch [1/3], Step [1253/3236], Loss: 2.5737, Perplexity: 13.1143
Epoch [1/3], Step [1254/3236], Loss: 2.5036, Perplexity: 12.2262
Epoch [1/3], Step [1255/3236], Loss: 2.5525, Perplexity: 12.8388
Epoch [1/3], Step [1256/3236], Loss: 2.5349, Perplexity: 12.6146
Epoch [1/3], Step [1257/3236], Loss: 2.7413, Perplexity: 15.5078
Epoch [1/3], Step [1258/3236], Loss: 2.5006, Perplexity: 12.1900
Epoch [1/3], Step [1259/3236], Loss: 2.6327, Perplexity: 13.9120
Epoch [1/3], Step [1260/3236], Loss: 2.5755, Perplexity: 13.1377
Epoch [1/3], Step [1261/3236], Loss: 2.5879, Perplexity: 13.3021
Epoch [1/3], Step [1262/3236], Loss: 2.5141, Perplexity: 12.3557
Epoch [1/3], Step [1263/3236], Loss: 2.4977, Perplexity: 12.1541
Epoch [1/3], Step [1264/3236], Loss: 2.5590, Perplexity: 12.9234
Epoch [1/3], Step [1265/3236], Loss: 2.6530, Perplexity: 14.1968
Epoch [1/3], Step [1266/3236], Loss: 2.5631, Perplexity: 12.9754
Epoch [1/3], Step [1267/3236], Loss: 2.5777, Perplexity: 13.1670
Epoch [1/3], Step [1268/3236], Loss: 2.6320, Perplexity: 13.9009
Epoch [1/3], Step [1269/3236], Loss: 2.4171, Perplexity: 11.2135
Epoch [1/3], Step [1270/3236], Loss: 2.3686, Perplexity: 10.6821
Epoch [1/3], Step [1271/3236], Loss: 2.5404, Perplexity: 12.6850
Epoch [1/3], Step [1272/3236], Loss: 2.5988, Perplexity: 13.4473
Epoch [1/3], Step [1273/3236], Loss: 2.6556, Perplexity: 14.2336
Epoch [1/3], Step [1274/3236], Loss: 2.5055, Perplexity: 12.2498
Epoch [1/3], Step [1275/3236], Loss: 2.7314, Perplexity: 15.3538
Epoch [1/3], Step [1276/3236], Loss: 2.6610, Perplexity: 14.3111
Epoch [1/3], Step [1277/3236], Loss: 2.6071, Perplexity: 13.5595
Epoch [1/3], Step [1278/3236], Loss: 2.4749, Perplexity: 11.8803
Epoch [1/3], Step [1279/3236], Loss: 2.5599, Perplexity: 12.9352
Epoch [1/3], Step [1280/3236], Loss: 2.4701, Perplexity: 11.8233
Epoch [1/3], Step [1281/3236], Loss: 2.6169, Perplexity: 13.6930
Epoch [1/3], Step [1282/3236], Loss: 2.3457, Perplexity: 10.4405
Epoch [1/3], Step [1283/3236], Loss: 2.9963, Perplexity: 20.0111
Epoch [1/3], Step [1284/3236], Loss: 3.5569, Perplexity: 35.0549
Epoch [1/3], Step [1285/3236], Loss: 2.6261, Perplexity: 13.8196
Epoch [1/3], Step [1286/3236], Loss: 3.4697, Perplexity: 32.1274
Epoch [1/3], Step [1287/3236], Loss: 2.5553, Perplexity: 12.8750
Epoch [1/3], Step [1288/3236], Loss: 3.1746, Perplexity: 23.9168
Epoch [1/3], Step [1289/3236], Loss: 2.6253, Perplexity: 13.8093
Epoch [1/3], Step [1290/3236], Loss: 2.4979, Perplexity: 12.1575
Epoch [1/3], Step [1291/3236], Loss: 2.6215, Perplexity: 13.7565
Epoch [1/3], Step [1292/3236], Loss: 2.6344, Perplexity: 13.9355
Epoch [1/3], Step [1293/3236], Loss: 2.4880, Perplexity: 12.0375
Epoch [1/3], Step [1294/3236], Loss: 2.5705, Perplexity: 13.0718
Epoch [1/3], Step [1295/3236], Loss: 2.6452, Perplexity: 14.0861
Epoch [1/3], Step [1296/3236], Loss: 3.0394, Perplexity: 20.8935
Epoch [1/3], Step [1297/3236], Loss: 2.5419, Perplexity: 12.7037
Epoch [1/3], Step [1298/3236], Loss: 2.4318, Perplexity: 11.3791
Epoch [1/3], Step [1299/3236], Loss: 2.5348, Perplexity: 12.6142
Epoch [1/3], Step [1300/3236], Loss: 2.8728, Perplexity: 17.6863
Epoch [1/3], Step [1301/3236], Loss: 2.4841, Perplexity: 11.9899
Epoch [1/3], Step [1302/3236], Loss: 2.5040, Perplexity: 12.2310
Epoch [1/3], Step [1303/3236], Loss: 2.6059, Perplexity: 13.5429
Epoch [1/3], Step [1304/3236], Loss: 2.6422, Perplexity: 14.0436
Epoch [1/3], Step [1305/3236], Loss: 2.7004, Perplexity: 14.8863
Epoch [1/3], Step [1306/3236], Loss: 2.4082, Perplexity: 11.1138
Epoch [1/3], Step [1307/3236], Loss: 2.5723, Perplexity: 13.0960
Epoch [1/3], Step [1308/3236], Loss: 2.6180, Perplexity: 13.7087
Epoch [1/3], Step [1309/3236], Loss: 2.6002, Perplexity: 13.4659
Epoch [1/3], Step [1310/3236], Loss: 2.5355, Perplexity: 12.6233
Epoch [1/3], Step [1311/3236], Loss: 2.8573, Perplexity: 17.4147
Epoch [1/3], Step [1312/3236], Loss: 2.5100, Perplexity: 12.3048
Epoch [1/3], Step [1313/3236], Loss: 2.8776, Perplexity: 17.7712
Epoch [1/3], Step [1314/3236], Loss: 2.4878, Perplexity: 12.0344
Epoch [1/3], Step [1315/3236], Loss: 2.5129, Perplexity: 12.3403
Epoch [1/3], Step [1316/3236], Loss: 2.6045, Perplexity: 13.5242
Epoch [1/3], Step [1317/3236], Loss: 2.5114, Perplexity: 12.3224
Epoch [1/3], Step [1318/3236], Loss: 2.4590, Perplexity: 11.6930
Epoch [1/3], Step [1319/3236], Loss: 2.6628, Perplexity: 14.3357
Epoch [1/3], Step [1320/3236], Loss: 2.6374, Perplexity: 13.9771
Epoch [1/3], Step [1321/3236], Loss: 3.3221, Perplexity: 27.7197
Epoch [1/3], Step [1322/3236], Loss: 2.4437, Perplexity: 11.5161
Epoch [1/3], Step [1323/3236], Loss: 2.5542, Perplexity: 12.8608
Epoch [1/3], Step [1324/3236], Loss: 2.9430, Perplexity: 18.9735
Epoch [1/3], Step [1325/3236], Loss: 2.8021, Perplexity: 16.4797
Epoch [1/3], Step [1326/3236], Loss: 2.4586, Perplexity: 11.6879
Epoch [1/3], Step [1327/3236], Loss: 2.5685, Perplexity: 13.0464
Epoch [1/3], Step [1328/3236], Loss: 2.5611, Perplexity: 12.9496
Epoch [1/3], Step [1329/3236], Loss: 2.6965, Perplexity: 14.8271
Epoch [1/3], Step [1330/3236], Loss: 2.5493, Perplexity: 12.7980
Epoch [1/3], Step [1331/3236], Loss: 2.5303, Perplexity: 12.5568
Epoch [1/3], Step [1332/3236], Loss: 2.5203, Perplexity: 12.4321
Epoch [1/3], Step [1333/3236], Loss: 2.5140, Perplexity: 12.3540
Epoch [1/3], Step [1334/3236], Loss: 2.8004, Perplexity: 16.4516
Epoch [1/3], Step [1335/3236], Loss: 2.7315, Perplexity: 15.3556
Epoch [1/3], Step [1336/3236], Loss: 2.4266, Perplexity: 11.3204
Epoch [1/3], Step [1337/3236], Loss: 2.4421, Perplexity: 11.4975
Epoch [1/3], Step [1338/3236], Loss: 2.4236, Perplexity: 11.2868
Epoch [1/3], Step [1339/3236], Loss: 2.5226, Perplexity: 12.4611
Epoch [1/3], Step [1340/3236], Loss: 2.4709, Perplexity: 11.8336
Epoch [1/3], Step [1341/3236], Loss: 2.3357, Perplexity: 10.3367
Epoch [1/3], Step [1342/3236], Loss: 2.5793, Perplexity: 13.1882
Epoch [1/3], Step [1343/3236], Loss: 2.6698, Perplexity: 14.4369
Epoch [1/3], Step [1344/3236], Loss: 2.6440, Perplexity: 14.0690
Epoch [1/3], Step [1345/3236], Loss: 2.3899, Perplexity: 10.9122
Epoch [1/3], Step [1346/3236], Loss: 2.5137, Perplexity: 12.3508
Epoch [1/3], Step [1347/3236], Loss: 2.6879, Perplexity: 14.7004
Epoch [1/3], Step [1348/3236], Loss: 3.2717, Perplexity: 26.3569
Epoch [1/3], Step [1349/3236], Loss: 2.9155, Perplexity: 18.4574
Epoch [1/3], Step [1350/3236], Loss: 2.3324, Perplexity: 10.3022
Epoch [1/3], Step [1351/3236], Loss: 2.4732, Perplexity: 11.8599
Epoch [1/3], Step [1352/3236], Loss: 2.5955, Perplexity: 13.4034
Epoch [1/3], Step [1353/3236], Loss: 2.5731, Perplexity: 13.1064
Epoch [1/3], Step [1354/3236], Loss: 2.6094, Perplexity: 13.5911
Epoch [1/3], Step [1355/3236], Loss: 2.5423, Perplexity: 12.7090
Epoch [1/3], Step [1356/3236], Loss: 2.8269, Perplexity: 16.8928
Epoch [1/3], Step [1357/3236], Loss: 2.7571, Perplexity: 15.7541
Epoch [1/3], Step [1358/3236], Loss: 2.4423, Perplexity: 11.4993
Epoch [1/3], Step [1359/3236], Loss: 2.6338, Perplexity: 13.9269
Epoch [1/3], Step [1360/3236], Loss: 2.5433, Perplexity: 12.7212
Epoch [1/3], Step [1361/3236], Loss: 2.5462, Perplexity: 12.7590
Epoch [1/3], Step [1362/3236], Loss: 2.5691, Perplexity: 13.0539
Epoch [1/3], Step [1363/3236], Loss: 2.5092, Perplexity: 12.2947
Epoch [1/3], Step [1364/3236], Loss: 2.5988, Perplexity: 13.4472
Epoch [1/3], Step [1365/3236], Loss: 2.3628, Perplexity: 10.6201
Epoch [1/3], Step [1366/3236], Loss: 2.4962, Perplexity: 12.1362
Epoch [1/3], Step [1367/3236], Loss: 2.6853, Perplexity: 14.6633
Epoch [1/3], Step [1368/3236], Loss: 2.8501, Perplexity: 17.2898
Epoch [1/3], Step [1369/3236], Loss: 3.3455, Perplexity: 28.3759
Epoch [1/3], Step [1370/3236], Loss: 2.7806, Perplexity: 16.1286
Epoch [1/3], Step [1371/3236], Loss: 2.5057, Perplexity: 12.2518
Epoch [1/3], Step [1372/3236], Loss: 2.8479, Perplexity: 17.2513
Epoch [1/3], Step [1373/3236], Loss: 2.3951, Perplexity: 10.9697
Epoch [1/3], Step [1374/3236], Loss: 2.6064, Perplexity: 13.5498
Epoch [1/3], Step [1375/3236], Loss: 2.5970, Perplexity: 13.4235
Epoch [1/3], Step [1376/3236], Loss: 2.2982, Perplexity: 9.9558
Epoch [1/3], Step [1377/3236], Loss: 2.5643, Perplexity: 12.9916
Epoch [1/3], Step [1378/3236], Loss: 2.7811, Perplexity: 16.1368
Epoch [1/3], Step [1379/3236], Loss: 2.4894, Perplexity: 12.0537
Epoch [1/3], Step [1380/3236], Loss: 2.5218, Perplexity: 12.4510
Epoch [1/3], Step [1381/3236], Loss: 2.4677, Perplexity: 11.7949
Epoch [1/3], Step [1382/3236], Loss: 2.3579, Perplexity: 10.5692
Epoch [1/3], Step [1383/3236], Loss: 2.5307, Perplexity: 12.5629
Epoch [1/3], Step [1384/3236], Loss: 2.3396, Perplexity: 10.3768
Epoch [1/3], Step [1385/3236], Loss: 2.5492, Perplexity: 12.7973
Epoch [1/3], Step [1386/3236], Loss: 2.5570, Perplexity: 12.8972
Epoch [1/3], Step [1387/3236], Loss: 2.5073, Perplexity: 12.2716
Epoch [1/3], Step [1388/3236], Loss: 2.6230, Perplexity: 13.7770
Epoch [1/3], Step [1389/3236], Loss: 3.3695, Perplexity: 29.0634
Epoch [1/3], Step [1390/3236], Loss: 2.6257, Perplexity: 13.8136
Epoch [1/3], Step [1391/3236], Loss: 2.7483, Perplexity: 15.6164
Epoch [1/3], Step [1392/3236], Loss: 2.7264, Perplexity: 15.2776
Epoch [1/3], Step [1393/3236], Loss: 2.5501, Perplexity: 12.8087
Epoch [1/3], Step [1394/3236], Loss: 2.6220, Perplexity: 13.7635
Epoch [1/3], Step [1395/3236], Loss: 2.5607, Perplexity: 12.9452
Epoch [1/3], Step [1396/3236], Loss: 2.4576, Perplexity: 11.6773
Epoch [1/3], Step [1397/3236], Loss: 2.4760, Perplexity: 11.8939
Epoch [1/3], Step [1398/3236], Loss: 2.6889, Perplexity: 14.7158
Epoch [1/3], Step [1399/3236], Loss: 2.5918, Perplexity: 13.3541
Epoch [1/3], Step [1400/3236], Loss: 2.3799, Perplexity: 10.8042
Epoch [1/3], Step [1401/3236], Loss: 2.8440, Perplexity: 17.1836
Epoch [1/3], Step [1402/3236], Loss: 2.3994, Perplexity: 11.0167
Epoch [1/3], Step [1403/3236], Loss: 2.6489, Perplexity: 14.1391
Epoch [1/3], Step [1404/3236], Loss: 2.4489, Perplexity: 11.5751
Epoch [1/3], Step [1405/3236], Loss: 2.8687, Perplexity: 17.6139
Epoch [1/3], Step [1406/3236], Loss: 2.6073, Perplexity: 13.5627
Epoch [1/3], Step [1407/3236], Loss: 2.4674, Perplexity: 11.7922
Epoch [1/3], Step [1408/3236], Loss: 2.3140, Perplexity: 10.1148
Epoch [1/3], Step [1409/3236], Loss: 2.4958, Perplexity: 12.1311
Epoch [1/3], Step [1410/3236], Loss: 2.4528, Perplexity: 11.6204
Epoch [1/3], Step [1411/3236], Loss: 2.5192, Perplexity: 12.4189
Epoch [1/3], Step [1412/3236], Loss: 2.9211, Perplexity: 18.5620
Epoch [1/3], Step [1413/3236], Loss: 2.4790, Perplexity: 11.9296
Epoch [1/3], Step [1414/3236], Loss: 2.4941, Perplexity: 12.1110
Epoch [1/3], Step [1415/3236], Loss: 2.4579, Perplexity: 11.6806
Epoch [1/3], Step [1416/3236], Loss: 2.4692, Perplexity: 11.8130
Epoch [1/3], Step [1417/3236], Loss: 2.3939, Perplexity: 10.9557
Epoch [1/3], Step [1418/3236], Loss: 2.5076, Perplexity: 12.2756
Epoch [1/3], Step [1419/3236], Loss: 2.9853, Perplexity: 19.7921
Epoch [1/3], Step [1420/3236], Loss: 2.3949, Perplexity: 10.9670
Epoch [1/3], Step [1421/3236], Loss: 2.4085, Perplexity: 11.1175
Epoch [1/3], Step [1422/3236], Loss: 2.5825, Perplexity: 13.2296
Epoch [1/3], Step [1423/3236], Loss: 2.3861, Perplexity: 10.8715
Epoch [1/3], Step [1424/3236], Loss: 2.4468, Perplexity: 11.5509
Epoch [1/3], Step [1425/3236], Loss: 2.5894, Perplexity: 13.3213
Epoch [1/3], Step [1426/3236], Loss: 2.7563, Perplexity: 15.7410
Epoch [1/3], Step [1427/3236], Loss: 2.6000, Perplexity: 13.4638
Epoch [1/3], Step [1428/3236], Loss: 2.4241, Perplexity: 11.2922
Epoch [1/3], Step [1429/3236], Loss: 2.3818, Perplexity: 10.8239
Epoch [1/3], Step [1430/3236], Loss: 2.5264, Perplexity: 12.5088
Epoch [1/3], Step [1431/3236], Loss: 2.4912, Perplexity: 12.0763
Epoch [1/3], Step [1432/3236], Loss: 2.8700, Perplexity: 17.6377
Epoch [1/3], Step [1433/3236], Loss: 2.5153, Perplexity: 12.3704
Epoch [1/3], Step [1434/3236], Loss: 2.6537, Perplexity: 14.2069
Epoch [1/3], Step [1435/3236], Loss: 2.3524, Perplexity: 10.5107
Epoch [1/3], Step [1436/3236], Loss: 2.7695, Perplexity: 15.9514
Epoch [1/3], Step [1437/3236], Loss: 2.4816, Perplexity: 11.9598
Epoch [1/3], Step [1438/3236], Loss: 2.3629, Perplexity: 10.6214
Epoch [1/3], Step [1439/3236], Loss: 3.6998, Perplexity: 40.4396
Epoch [1/3], Step [1440/3236], Loss: 2.5171, Perplexity: 12.3922
Epoch [1/3], Step [1441/3236], Loss: 2.4732, Perplexity: 11.8604
Epoch [1/3], Step [1442/3236], Loss: 2.5100, Perplexity: 12.3053
Epoch [1/3], Step [1443/3236], Loss: 2.3292, Perplexity: 10.2693
Epoch [1/3], Step [1444/3236], Loss: 2.5476, Perplexity: 12.7759
Epoch [1/3], Step [1445/3236], Loss: 2.4783, Perplexity: 11.9210
Epoch [1/3], Step [1446/3236], Loss: 3.7880, Perplexity: 44.1673
Epoch [1/3], Step [1447/3236], Loss: 2.6458, Perplexity: 14.0946
Epoch [1/3], Step [1448/3236], Loss: 2.4225, Perplexity: 11.2735
Epoch [1/3], Step [1449/3236], Loss: 2.7503, Perplexity: 15.6472
Epoch [1/3], Step [1450/3236], Loss: 4.0729, Perplexity: 58.7273
Epoch [1/3], Step [1451/3236], Loss: 2.6216, Perplexity: 13.7582
Epoch [1/3], Step [1452/3236], Loss: 2.9906, Perplexity: 19.8968
Epoch [1/3], Step [1453/3236], Loss: 2.4104, Perplexity: 11.1386
Epoch [1/3], Step [1454/3236], Loss: 2.7712, Perplexity: 15.9772
Epoch [1/3], Step [1455/3236], Loss: 2.6993, Perplexity: 14.8687
Epoch [1/3], Step [1456/3236], Loss: 2.5901, Perplexity: 13.3310
Epoch [1/3], Step [1457/3236], Loss: 2.4560, Perplexity: 11.6584
Epoch [1/3], Step [1458/3236], Loss: 2.8185, Perplexity: 16.7514
Epoch [1/3], Step [1459/3236], Loss: 2.5294, Perplexity: 12.5460
Epoch [1/3], Step [1460/3236], Loss: 2.8725, Perplexity: 17.6812
Epoch [1/3], Step [1461/3236], Loss: 2.4027, Perplexity: 11.0535
Epoch [1/3], Step [1462/3236], Loss: 2.4911, Perplexity: 12.0751
Epoch [1/3], Step [1463/3236], Loss: 2.3993, Perplexity: 11.0156
Epoch [1/3], Step [1464/3236], Loss: 2.6575, Perplexity: 14.2602
Epoch [1/3], Step [1465/3236], Loss: 2.4657, Perplexity: 11.7716
Epoch [1/3], Step [1466/3236], Loss: 2.5913, Perplexity: 13.3475
Epoch [1/3], Step [1467/3236], Loss: 2.3592, Perplexity: 10.5822
Epoch [1/3], Step [1468/3236], Loss: 2.4192, Perplexity: 11.2367
Epoch [1/3], Step [1469/3236], Loss: 2.4512, Perplexity: 11.6021
Epoch [1/3], Step [1470/3236], Loss: 2.2779, Perplexity: 9.7559
Epoch [1/3], Step [1471/3236], Loss: 2.3983, Perplexity: 11.0043
Epoch [1/3], Step [1472/3236], Loss: 2.3697, Perplexity: 10.6946
Epoch [1/3], Step [1473/3236], Loss: 2.5031, Perplexity: 12.2209
Epoch [1/3], Step [1474/3236], Loss: 2.5080, Perplexity: 12.2803
Epoch [1/3], Step [1475/3236], Loss: 2.4357, Perplexity: 11.4237
Epoch [1/3], Step [1476/3236], Loss: 2.3696, Perplexity: 10.6927
Epoch [1/3], Step [1477/3236], Loss: 2.6145, Perplexity: 13.6599
Epoch [1/3], Step [1478/3236], Loss: 2.4718, Perplexity: 11.8440
Epoch [1/3], Step [1479/3236], Loss: 2.4484, Perplexity: 11.5698
Epoch [1/3], Step [1480/3236], Loss: 2.9296, Perplexity: 18.7203
Epoch [1/3], Step [1481/3236], Loss: 3.2856, Perplexity: 26.7253
Epoch [1/3], Step [1482/3236], Loss: 2.4102, Perplexity: 11.1366
Epoch [1/3], Step [1483/3236], Loss: 2.6637, Perplexity: 14.3489
Epoch [1/3], Step [1484/3236], Loss: 2.7575, Perplexity: 15.7607
Epoch [1/3], Step [1485/3236], Loss: 2.2415, Perplexity: 9.4076
Epoch [1/3], Step [1486/3236], Loss: 2.5117, Perplexity: 12.3265
Epoch [1/3], Step [1487/3236], Loss: 2.5361, Perplexity: 12.6304
Epoch [1/3], Step [1488/3236], Loss: 2.6367, Perplexity: 13.9677
Epoch [1/3], Step [1489/3236], Loss: 2.4800, Perplexity: 11.9417
Epoch [1/3], Step [1490/3236], Loss: 2.4741, Perplexity: 11.8714
Epoch [1/3], Step [1491/3236], Loss: 2.5378, Perplexity: 12.6516
Epoch [1/3], Step [1492/3236], Loss: 2.3574, Perplexity: 10.5636
Epoch [1/3], Step [1493/3236], Loss: 2.4325, Perplexity: 11.3868
Epoch [1/3], Step [1494/3236], Loss: 2.8400, Perplexity: 17.1150
Epoch [1/3], Step [1495/3236], Loss: 2.5851, Perplexity: 13.2648
Epoch [1/3], Step [1496/3236], Loss: 2.4524, Perplexity: 11.6168
Epoch [1/3], Step [1497/3236], Loss: 2.5898, Perplexity: 13.3274
Epoch [1/3], Step [1498/3236], Loss: 2.6449, Perplexity: 14.0815
Epoch [1/3], Step [1499/3236], Loss: 2.3108, Perplexity: 10.0824
Epoch [1/3], Step [1500/3236], Loss: 2.5878, Perplexity: 13.3004
Epoch [1/3], Step [1501/3236], Loss: 2.3652, Perplexity: 10.6460
Epoch [1/3], Step [1502/3236], Loss: 2.3317, Perplexity: 10.2952
Epoch [1/3], Step [1503/3236], Loss: 2.3945, Perplexity: 10.9632
Epoch [1/3], Step [1504/3236], Loss: 2.4658, Perplexity: 11.7727
Epoch [1/3], Step [1505/3236], Loss: 2.7589, Perplexity: 15.7826
Epoch [1/3], Step [1506/3236], Loss: 3.0845, Perplexity: 21.8556
Epoch [1/3], Step [1507/3236], Loss: 2.5515, Perplexity: 12.8260
Epoch [1/3], Step [1508/3236], Loss: 2.5671, Perplexity: 13.0273
Epoch [1/3], Step [1509/3236], Loss: 2.6605, Perplexity: 14.3041
Epoch [1/3], Step [1510/3236], Loss: 2.4149, Perplexity: 11.1884
Epoch [1/3], Step [1511/3236], Loss: 2.3885, Perplexity: 10.8972
Epoch [1/3], Step [1512/3236], Loss: 2.3373, Perplexity: 10.3529
Epoch [1/3], Step [1513/3236], Loss: 3.1596, Perplexity: 23.5608
Epoch [1/3], Step [1514/3236], Loss: 2.3536, Perplexity: 10.5229
Epoch [1/3], Step [1515/3236], Loss: 2.4006, Perplexity: 11.0294
Epoch [1/3], Step [1516/3236], Loss: 2.3361, Perplexity: 10.3406
Epoch [1/3], Step [1517/3236], Loss: 2.5908, Perplexity: 13.3410
Epoch [1/3], Step [1518/3236], Loss: 2.9037, Perplexity: 18.2418
Epoch [1/3], Step [1519/3236], Loss: 2.4889, Perplexity: 12.0483
Epoch [1/3], Step [1520/3236], Loss: 2.3175, Perplexity: 10.1501
Epoch [1/3], Step [1521/3236], Loss: 2.7125, Perplexity: 15.0670
Epoch [1/3], Step [1522/3236], Loss: 3.0696, Perplexity: 21.5339
Epoch [1/3], Step [1523/3236], Loss: 2.3710, Perplexity: 10.7078
Epoch [1/3], Step [1524/3236], Loss: 2.4759, Perplexity: 11.8922
Epoch [1/3], Step [1525/3236], Loss: 2.4060, Perplexity: 11.0893
Epoch [1/3], Step [1526/3236], Loss: 2.9144, Perplexity: 18.4372
Epoch [1/3], Step [1527/3236], Loss: 2.8023, Perplexity: 16.4832
Epoch [1/3], Step [1528/3236], Loss: 2.4931, Perplexity: 12.0993
Epoch [1/3], Step [1529/3236], Loss: 2.4417, Perplexity: 11.4927
Epoch [1/3], Step [1530/3236], Loss: 2.3330, Perplexity: 10.3088
Epoch [1/3], Step [1531/3236], Loss: 2.4617, Perplexity: 11.7247
Epoch [1/3], Step [1532/3236], Loss: 2.5051, Perplexity: 12.2453
Epoch [1/3], Step [1533/3236], Loss: 2.4850, Perplexity: 12.0011
Epoch [1/3], Step [1534/3236], Loss: 2.4058, Perplexity: 11.0868
Epoch [1/3], Step [1535/3236], Loss: 2.4369, Perplexity: 11.4371
Epoch [1/3], Step [1536/3236], Loss: 2.5381, Perplexity: 12.6556
Epoch [1/3], Step [1537/3236], Loss: 2.3752, Perplexity: 10.7534
Epoch [1/3], Step [1538/3236], Loss: 2.5639, Perplexity: 12.9869
Epoch [1/3], Step [1539/3236], Loss: 2.5082, Perplexity: 12.2825
Epoch [1/3], Step [1540/3236], Loss: 2.5931, Perplexity: 13.3707
Epoch [1/3], Step [1541/3236], Loss: 2.2694, Perplexity: 9.6733
Epoch [1/3], Step [1542/3236], Loss: 2.5701, Perplexity: 13.0672
Epoch [1/3], Step [1543/3236], Loss: 2.4837, Perplexity: 11.9855
Epoch [1/3], Step [1544/3236], Loss: 2.4862, Perplexity: 12.0159
Epoch [1/3], Step [1545/3236], Loss: 2.5370, Perplexity: 12.6418
Epoch [1/3], Step [1546/3236], Loss: 2.4557, Perplexity: 11.6549
Epoch [1/3], Step [1547/3236], Loss: 2.5845, Perplexity: 13.2561
Epoch [1/3], Step [1548/3236], Loss: 2.8136, Perplexity: 16.6702
Epoch [1/3], Step [1549/3236], Loss: 2.3758, Perplexity: 10.7595
Epoch [1/3], Step [1550/3236], Loss: 2.4356, Perplexity: 11.4228
Epoch [1/3], Step [1551/3236], Loss: 2.4691, Perplexity: 11.8119
Epoch [1/3], Step [1552/3236], Loss: 2.3999, Perplexity: 11.0218
Epoch [1/3], Step [1553/3236], Loss: 2.3037, Perplexity: 10.0116
Epoch [1/3], Step [1554/3236], Loss: 2.5634, Perplexity: 12.9802
Epoch [1/3], Step [1555/3236], Loss: 2.5960, Perplexity: 13.4099
Epoch [1/3], Step [1556/3236], Loss: 2.4184, Perplexity: 11.2276
Epoch [1/3], Step [1557/3236], Loss: 2.4854, Perplexity: 12.0059
Epoch [1/3], Step [1558/3236], Loss: 2.2053, Perplexity: 9.0730
Epoch [1/3], Step [1559/3236], Loss: 2.5077, Perplexity: 12.2768
Epoch [1/3], Step [1560/3236], Loss: 2.9356, Perplexity: 18.8325
Epoch [1/3], Step [1561/3236], Loss: 2.5513, Perplexity: 12.8242
Epoch [1/3], Step [1562/3236], Loss: 2.5333, Perplexity: 12.5947
Epoch [1/3], Step [1563/3236], Loss: 2.6976, Perplexity: 14.8437
Epoch [1/3], Step [1564/3236], Loss: 2.3908, Perplexity: 10.9223
Epoch [1/3], Step [1565/3236], Loss: 2.4553, Perplexity: 11.6497
Epoch [1/3], Step [1566/3236], Loss: 2.4178, Perplexity: 11.2215
Epoch [1/3], Step [1567/3236], Loss: 2.4313, Perplexity: 11.3733
Epoch [1/3], Step [1568/3236], Loss: 2.4345, Perplexity: 11.4099
Epoch [1/3], Step [1569/3236], Loss: 2.5745, Perplexity: 13.1242
Epoch [1/3], Step [1570/3236], Loss: 2.7713, Perplexity: 15.9801
Epoch [1/3], Step [1571/3236], Loss: 2.4367, Perplexity: 11.4358
Epoch [1/3], Step [1572/3236], Loss: 2.3497, Perplexity: 10.4822
Epoch [1/3], Step [1573/3236], Loss: 2.4928, Perplexity: 12.0956
Epoch [1/3], Step [1574/3236], Loss: 2.2529, Perplexity: 9.5150
Epoch [1/3], Step [1575/3236], Loss: 2.4906, Perplexity: 12.0686
Epoch [1/3], Step [1576/3236], Loss: 2.3098, Perplexity: 10.0727
Epoch [1/3], Step [1577/3236], Loss: 2.4407, Perplexity: 11.4805
Epoch [1/3], Step [1578/3236], Loss: 3.2081, Perplexity: 24.7314
Epoch [1/3], Step [1579/3236], Loss: 2.3510, Perplexity: 10.4958
Epoch [1/3], Step [1580/3236], Loss: 2.5134, Perplexity: 12.3463
Epoch [1/3], Step [1581/3236], Loss: 2.4868, Perplexity: 12.0228
Epoch [1/3], Step [1582/3236], Loss: 2.4387, Perplexity: 11.4576
Epoch [1/3], Step [1583/3236], Loss: 2.5862, Perplexity: 13.2795
Epoch [1/3], Step [1584/3236], Loss: 2.5030, Perplexity: 12.2188
Epoch [1/3], Step [1585/3236], Loss: 2.4677, Perplexity: 11.7956
Epoch [1/3], Step [1586/3236], Loss: 2.5551, Perplexity: 12.8729
Epoch [1/3], Step [1587/3236], Loss: 3.3379, Perplexity: 28.1606
Epoch [1/3], Step [1588/3236], Loss: 2.2237, Perplexity: 9.2417
Epoch [1/3], Step [1589/3236], Loss: 2.5054, Perplexity: 12.2489
Epoch [1/3], Step [1590/3236], Loss: 2.5191, Perplexity: 12.4172
Epoch [1/3], Step [1591/3236], Loss: 2.2738, Perplexity: 9.7166
Epoch [1/3], Step [1592/3236], Loss: 2.4226, Perplexity: 11.2752
Epoch [1/3], Step [1593/3236], Loss: 2.2631, Perplexity: 9.6128
Epoch [1/3], Step [1594/3236], Loss: 2.4893, Perplexity: 12.0527
Epoch [1/3], Step [1595/3236], Loss: 2.2463, Perplexity: 9.4532
Epoch [1/3], Step [1596/3236], Loss: 2.3665, Perplexity: 10.6596
Epoch [1/3], Step [1597/3236], Loss: 2.3800, Perplexity: 10.8046
Epoch [1/3], Step [1598/3236], Loss: 2.7940, Perplexity: 16.3466
Epoch [1/3], Step [1599/3236], Loss: 2.4514, Perplexity: 11.6042
Epoch [1/3], Step [1600/3236], Loss: 2.4797, Perplexity: 11.9376
Epoch [1/3], Step [1601/3236], Loss: 2.4982, Perplexity: 12.1609
Epoch [1/3], Step [1602/3236], Loss: 2.6525, Perplexity: 14.1898
Epoch [1/3], Step [1603/3236], Loss: 2.4032, Perplexity: 11.0581
Epoch [1/3], Step [1604/3236], Loss: 2.2997, Perplexity: 9.9715
Epoch [1/3], Step [1605/3236], Loss: 2.7630, Perplexity: 15.8476
Epoch [1/3], Step [1606/3236], Loss: 2.4830, Perplexity: 11.9775
Epoch [1/3], Step [1607/3236], Loss: 2.4012, Perplexity: 11.0363
Epoch [1/3], Step [1608/3236], Loss: 2.5243, Perplexity: 12.4825
Epoch [1/3], Step [1609/3236], Loss: 2.3653, Perplexity: 10.6475
Epoch [1/3], Step [1610/3236], Loss: 2.3549, Perplexity: 10.5375
Epoch [1/3], Step [1611/3236], Loss: 2.6231, Perplexity: 13.7787
Epoch [1/3], Step [1612/3236], Loss: 2.4935, Perplexity: 12.1033
Epoch [1/3], Step [1613/3236], Loss: 2.4370, Perplexity: 11.4384
Epoch [1/3], Step [1614/3236], Loss: 3.2181, Perplexity: 24.9804
Epoch [1/3], Step [1615/3236], Loss: 2.6946, Perplexity: 14.7994
Epoch [1/3], Step [1616/3236], Loss: 2.6250, Perplexity: 13.8044
Epoch [1/3], Step [1617/3236], Loss: 2.5832, Perplexity: 13.2395
Epoch [1/3], Step [1618/3236], Loss: 2.4670, Perplexity: 11.7873
Epoch [1/3], Step [1619/3236], Loss: 2.4753, Perplexity: 11.8856
Epoch [1/3], Step [1620/3236], Loss: 2.4736, Perplexity: 11.8650
Epoch [1/3], Step [1621/3236], Loss: 2.7014, Perplexity: 14.9006
Epoch [1/3], Step [1622/3236], Loss: 2.4030, Perplexity: 11.0561
Epoch [1/3], Step [1623/3236], Loss: 2.5459, Perplexity: 12.7541
Epoch [1/3], Step [1624/3236], Loss: 2.4673, Perplexity: 11.7910
Epoch [1/3], Step [1625/3236], Loss: 2.4379, Perplexity: 11.4488
Epoch [1/3], Step [1626/3236], Loss: 2.3413, Perplexity: 10.3949
Epoch [1/3], Step [1627/3236], Loss: 2.5281, Perplexity: 12.5300
Epoch [1/3], Step [1628/3236], Loss: 2.5111, Perplexity: 12.3184
Epoch [1/3], Step [1629/3236], Loss: 2.5976, Perplexity: 13.4309
Epoch [1/3], Step [1630/3236], Loss: 3.0389, Perplexity: 20.8830
Epoch [1/3], Step [1631/3236], Loss: 2.5045, Perplexity: 12.2371
Epoch [1/3], Step [1632/3236], Loss: 2.4209, Perplexity: 11.2559
Epoch [1/3], Step [1633/3236], Loss: 2.4134, Perplexity: 11.1715
Epoch [1/3], Step [1634/3236], Loss: 2.6860, Perplexity: 14.6736
Epoch [1/3], Step [1635/3236], Loss: 2.4714, Perplexity: 11.8387
Epoch [1/3], Step [1636/3236], Loss: 2.6147, Perplexity: 13.6636
Epoch [1/3], Step [1637/3236], Loss: 2.6238, Perplexity: 13.7886
Epoch [1/3], Step [1638/3236], Loss: 2.8694, Perplexity: 17.6267
Epoch [1/3], Step [1639/3236], Loss: 2.3523, Perplexity: 10.5102
Epoch [1/3], Step [1640/3236], Loss: 2.5488, Perplexity: 12.7923
Epoch [1/3], Step [1641/3236], Loss: 2.2956, Perplexity: 9.9302
Epoch [1/3], Step [1642/3236], Loss: 2.4289, Perplexity: 11.3462
Epoch [1/3], Step [1643/3236], Loss: 2.5117, Perplexity: 12.3263
Epoch [1/3], Step [1644/3236], Loss: 2.3597, Perplexity: 10.5876
Epoch [1/3], Step [1645/3236], Loss: 2.3688, Perplexity: 10.6843
Epoch [1/3], Step [1646/3236], Loss: 2.4177, Perplexity: 11.2198
Epoch [1/3], Step [1647/3236], Loss: 2.4389, Perplexity: 11.4609
Epoch [1/3], Step [1648/3236], Loss: 2.3788, Perplexity: 10.7919
Epoch [1/3], Step [1649/3236], Loss: 2.3374, Perplexity: 10.3545
Epoch [1/3], Step [1650/3236], Loss: 2.3085, Perplexity: 10.0597
Epoch [1/3], Step [1651/3236], Loss: 2.4458, Perplexity: 11.5392
Epoch [1/3], Step [1652/3236], Loss: 2.5341, Perplexity: 12.6048
Epoch [1/3], Step [1653/3236], Loss: 2.3870, Perplexity: 10.8808
Epoch [1/3], Step [1654/3236], Loss: 2.4974, Perplexity: 12.1512
Epoch [1/3], Step [1655/3236], Loss: 2.5369, Perplexity: 12.6401
Epoch [1/3], Step [1656/3236], Loss: 2.5697, Perplexity: 13.0614
Epoch [1/3], Step [1657/3236], Loss: 2.4704, Perplexity: 11.8267
Epoch [1/3], Step [1658/3236], Loss: 2.3780, Perplexity: 10.7837
Epoch [1/3], Step [1659/3236], Loss: 2.6432, Perplexity: 14.0577
Epoch [1/3], Step [1660/3236], Loss: 2.4373, Perplexity: 11.4419
Epoch [1/3], Step [1661/3236], Loss: 2.4342, Perplexity: 11.4071
Epoch [1/3], Step [1662/3236], Loss: 2.6558, Perplexity: 14.2365
Epoch [1/3], Step [1663/3236], Loss: 2.4853, Perplexity: 12.0052
Epoch [1/3], Step [1664/3236], Loss: 2.5498, Perplexity: 12.8045
Epoch [1/3], Step [1665/3236], Loss: 2.3641, Perplexity: 10.6347
Epoch [1/3], Step [1666/3236], Loss: 2.4715, Perplexity: 11.8404
Epoch [1/3], Step [1667/3236], Loss: 2.3541, Perplexity: 10.5281
Epoch [1/3], Step [1668/3236], Loss: 2.3477, Perplexity: 10.4611
Epoch [1/3], Step [1669/3236], Loss: 2.3396, Perplexity: 10.3771
Epoch [1/3], Step [1670/3236], Loss: 2.4461, Perplexity: 11.5430
Epoch [1/3], Step [1671/3236], Loss: 2.4201, Perplexity: 11.2465
Epoch [1/3], Step [1672/3236], Loss: 2.3851, Perplexity: 10.8598
Epoch [1/3], Step [1673/3236], Loss: 2.9021, Perplexity: 18.2130
Epoch [1/3], Step [1674/3236], Loss: 2.6356, Perplexity: 13.9522
Epoch [1/3], Step [1675/3236], Loss: 2.5286, Perplexity: 12.5354
Epoch [1/3], Step [1676/3236], Loss: 2.5824, Perplexity: 13.2287
Epoch [1/3], Step [1677/3236], Loss: 2.4202, Perplexity: 11.2477
Epoch [1/3], Step [1678/3236], Loss: 2.7064, Perplexity: 14.9750
Epoch [1/3], Step [1679/3236], Loss: 2.4425, Perplexity: 11.5013
Epoch [1/3], Step [1680/3236], Loss: 2.3946, Perplexity: 10.9639
Epoch [1/3], Step [1681/3236], Loss: 2.6021, Perplexity: 13.4925
Epoch [1/3], Step [1682/3236], Loss: 2.2860, Perplexity: 9.8358
Epoch [1/3], Step [1683/3236], Loss: 2.2797, Perplexity: 9.7741
Epoch [1/3], Step [1684/3236], Loss: 2.5717, Perplexity: 13.0881
Epoch [1/3], Step [1685/3236], Loss: 2.4080, Perplexity: 11.1112
Epoch [1/3], Step [1686/3236], Loss: 2.4290, Perplexity: 11.3477
Epoch [1/3], Step [1687/3236], Loss: 2.5781, Perplexity: 13.1718
Epoch [1/3], Step [1688/3236], Loss: 2.3186, Perplexity: 10.1618
Epoch [1/3], Step [1689/3236], Loss: 2.2865, Perplexity: 9.8406
Epoch [1/3], Step [1690/3236], Loss: 2.4252, Perplexity: 11.3048
Epoch [1/3], Step [1691/3236], Loss: 2.3812, Perplexity: 10.8175
Epoch [1/3], Step [1692/3236], Loss: 2.5444, Perplexity: 12.7352
Epoch [1/3], Step [1693/3236], Loss: 2.4642, Perplexity: 11.7536
Epoch [1/3], Step [1694/3236], Loss: 2.4431, Perplexity: 11.5086
Epoch [1/3], Step [1695/3236], Loss: 2.3644, Perplexity: 10.6372
Epoch [1/3], Step [1696/3236], Loss: 2.3861, Perplexity: 10.8710
Epoch [1/3], Step [1697/3236], Loss: 2.3375, Perplexity: 10.3550
Epoch [1/3], Step [1698/3236], Loss: 2.2813, Perplexity: 9.7898
Epoch [1/3], Step [1699/3236], Loss: 2.2400, Perplexity: 9.3930
Epoch [1/3], Step [1700/3236], Loss: 2.3419, Perplexity: 10.4011
Epoch [1/3], Step [1701/3236], Loss: 2.4882, Perplexity: 12.0398
Epoch [1/3], Step [1702/3236], Loss: 2.5628, Perplexity: 12.9723
Epoch [1/3], Step [1703/3236], Loss: 2.3855, Perplexity: 10.8640
Epoch [1/3], Step [1704/3236], Loss: 2.3095, Perplexity: 10.0695
Epoch [1/3], Step [1705/3236], Loss: 2.4453, Perplexity: 11.5335
Epoch [1/3], Step [1706/3236], Loss: 2.7607, Perplexity: 15.8115
Epoch [1/3], Step [1707/3236], Loss: 2.2093, Perplexity: 9.1098
Epoch [1/3], Step [1708/3236], Loss: 2.3856, Perplexity: 10.8654
Epoch [1/3], Step [1709/3236], Loss: 2.3844, Perplexity: 10.8520
Epoch [1/3], Step [1710/3236], Loss: 2.4230, Perplexity: 11.2800
Epoch [1/3], Step [1711/3236], Loss: 2.4877, Perplexity: 12.0330
Epoch [1/3], Step [1712/3236], Loss: 2.2972, Perplexity: 9.9459
Epoch [1/3], Step [1713/3236], Loss: 2.3422, Perplexity: 10.4038
Epoch [1/3], Step [1714/3236], Loss: 2.3061, Perplexity: 10.0357
Epoch [1/3], Step [1715/3236], Loss: 2.7154, Perplexity: 15.1103
Epoch [1/3], Step [1716/3236], Loss: 2.4847, Perplexity: 11.9974
Epoch [1/3], Step [1717/3236], Loss: 2.2476, Perplexity: 9.4652
Epoch [1/3], Step [1718/3236], Loss: 2.4495, Perplexity: 11.5829
Epoch [1/3], Step [1719/3236], Loss: 2.3454, Perplexity: 10.4374
Epoch [1/3], Step [1720/3236], Loss: 2.3500, Perplexity: 10.4855
Epoch [1/3], Step [1721/3236], Loss: 2.3832, Perplexity: 10.8393
Epoch [1/3], Step [1722/3236], Loss: 2.6406, Perplexity: 14.0219
Epoch [1/3], Step [1723/3236], Loss: 2.4641, Perplexity: 11.7533
Epoch [1/3], Step [1724/3236], Loss: 2.3326, Perplexity: 10.3052
Epoch [1/3], Step [1725/3236], Loss: 2.3495, Perplexity: 10.4803
Epoch [1/3], Step [1726/3236], Loss: 2.4166, Perplexity: 11.2081
Epoch [1/3], Step [1727/3236], Loss: 2.4403, Perplexity: 11.4760
Epoch [1/3], Step [1728/3236], Loss: 2.7186, Perplexity: 15.1587
Epoch [1/3], Step [1729/3236], Loss: 2.4484, Perplexity: 11.5698
Epoch [1/3], Step [1730/3236], Loss: 2.3405, Perplexity: 10.3863
Epoch [1/3], Step [1731/3236], Loss: 2.4801, Perplexity: 11.9421
Epoch [1/3], Step [1732/3236], Loss: 2.5166, Perplexity: 12.3864
Epoch [1/3], Step [1733/3236], Loss: 2.3928, Perplexity: 10.9438
Epoch [1/3], Step [1734/3236], Loss: 2.1689, Perplexity: 8.7489
Epoch [1/3], Step [1735/3236], Loss: 2.6648, Perplexity: 14.3651
Epoch [1/3], Step [1736/3236], Loss: 2.4037, Perplexity: 11.0640
Epoch [1/3], Step [1737/3236], Loss: 3.5847, Perplexity: 36.0417
Epoch [1/3], Step [1738/3236], Loss: 2.2949, Perplexity: 9.9230
Epoch [1/3], Step [1739/3236], Loss: 2.5380, Perplexity: 12.6538
Epoch [1/3], Step [1740/3236], Loss: 2.3831, Perplexity: 10.8382
Epoch [1/3], Step [1741/3236], Loss: 2.3150, Perplexity: 10.1253
Epoch [1/3], Step [1742/3236], Loss: 2.3296, Perplexity: 10.2740
Epoch [1/3], Step [1743/3236], Loss: 2.5032, Perplexity: 12.2220
Epoch [1/3], Step [1744/3236], Loss: 2.5247, Perplexity: 12.4871
Epoch [1/3], Step [1745/3236], Loss: 2.4353, Perplexity: 11.4195
Epoch [1/3], Step [1746/3236], Loss: 2.3113, Perplexity: 10.0874
Epoch [1/3], Step [1747/3236], Loss: 2.3758, Perplexity: 10.7599
Epoch [1/3], Step [1748/3236], Loss: 2.6746, Perplexity: 14.5061
Epoch [1/3], Step [1749/3236], Loss: 2.2539, Perplexity: 9.5249
Epoch [1/3], Step [1750/3236], Loss: 2.7110, Perplexity: 15.0440
Epoch [1/3], Step [1751/3236], Loss: 2.6522, Perplexity: 14.1847
Epoch [1/3], Step [1752/3236], Loss: 2.4019, Perplexity: 11.0440
Epoch [1/3], Step [1753/3236], Loss: 2.5550, Perplexity: 12.8718
Epoch [1/3], Step [1754/3236], Loss: 2.7465, Perplexity: 15.5881
Epoch [1/3], Step [1755/3236], Loss: 2.4600, Perplexity: 11.7046
Epoch [1/3], Step [1756/3236], Loss: 3.0974, Perplexity: 22.1394
Epoch [1/3], Step [1757/3236], Loss: 2.5755, Perplexity: 13.1377
Epoch [1/3], Step [1758/3236], Loss: 2.2811, Perplexity: 9.7875
Epoch [1/3], Step [1759/3236], Loss: 2.3578, Perplexity: 10.5675
Epoch [1/3], Step [1760/3236], Loss: 2.7540, Perplexity: 15.7052
Epoch [1/3], Step [1761/3236], Loss: 2.3568, Perplexity: 10.5567
Epoch [1/3], Step [1762/3236], Loss: 2.4952, Perplexity: 12.1242
Epoch [1/3], Step [1763/3236], Loss: 2.3138, Perplexity: 10.1129
Epoch [1/3], Step [1764/3236], Loss: 2.3811, Perplexity: 10.8170
Epoch [1/3], Step [1765/3236], Loss: 2.5042, Perplexity: 12.2338
Epoch [1/3], Step [1766/3236], Loss: 2.3449, Perplexity: 10.4325
Epoch [1/3], Step [1767/3236], Loss: 2.7145, Perplexity: 15.0976
Epoch [1/3], Step [1768/3236], Loss: 2.4790, Perplexity: 11.9295
Epoch [1/3], Step [1769/3236], Loss: 2.4343, Perplexity: 11.4073
Epoch [1/3], Step [1770/3236], Loss: 2.3234, Perplexity: 10.2104
Epoch [1/3], Step [1771/3236], Loss: 2.3575, Perplexity: 10.5649
Epoch [1/3], Step [1772/3236], Loss: 2.4073, Perplexity: 11.1035
Epoch [1/3], Step [1773/3236], Loss: 2.6637, Perplexity: 14.3491
Epoch [1/3], Step [1774/3236], Loss: 2.4680, Perplexity: 11.7983
Epoch [1/3], Step [1775/3236], Loss: 2.3570, Perplexity: 10.5588
Epoch [1/3], Step [1776/3236], Loss: 2.3777, Perplexity: 10.7799
Epoch [1/3], Step [1777/3236], Loss: 2.4386, Perplexity: 11.4572
Epoch [1/3], Step [1778/3236], Loss: 2.3117, Perplexity: 10.0914
Epoch [1/3], Step [1779/3236], Loss: 2.6756, Perplexity: 14.5216
Epoch [1/3], Step [1780/3236], Loss: 2.4806, Perplexity: 11.9487
Epoch [1/3], Step [1781/3236], Loss: 2.6431, Perplexity: 14.0560
Epoch [1/3], Step [1782/3236], Loss: 2.9908, Perplexity: 19.9013
Epoch [1/3], Step [1783/3236], Loss: 2.2539, Perplexity: 9.5250
Epoch [1/3], Step [1784/3236], Loss: 2.4210, Perplexity: 11.2576
Epoch [1/3], Step [1785/3236], Loss: 2.3282, Perplexity: 10.2593
Epoch [1/3], Step [1786/3236], Loss: 2.3297, Perplexity: 10.2753
Epoch [1/3], Step [1787/3236], Loss: 2.4915, Perplexity: 12.0790
Epoch [1/3], Step [1788/3236], Loss: 2.2767, Perplexity: 9.7442
Epoch [1/3], Step [1789/3236], Loss: 2.2412, Perplexity: 9.4050
Epoch [1/3], Step [1790/3236], Loss: 2.3535, Perplexity: 10.5221
Epoch [1/3], Step [1791/3236], Loss: 2.7676, Perplexity: 15.9200
Epoch [1/3], Step [1792/3236], Loss: 2.5276, Perplexity: 12.5231
Epoch [1/3], Step [1793/3236], Loss: 2.3202, Perplexity: 10.1780
Epoch [1/3], Step [1794/3236], Loss: 2.4412, Perplexity: 11.4869
Epoch [1/3], Step [1795/3236], Loss: 2.4721, Perplexity: 11.8472
Epoch [1/3], Step [1796/3236], Loss: 2.8425, Perplexity: 17.1584
Epoch [1/3], Step [1797/3236], Loss: 2.5580, Perplexity: 12.9099
Epoch [1/3], Step [1798/3236], Loss: 2.5973, Perplexity: 13.4275
Epoch [1/3], Step [1799/3236], Loss: 2.3760, Perplexity: 10.7618
Epoch [1/3], Step [1800/3236], Loss: 2.4197, Perplexity: 11.2423
Epoch [1/3], Step [1801/3236], Loss: 2.4390, Perplexity: 11.4611
Epoch [1/3], Step [1802/3236], Loss: 2.4274, Perplexity: 11.3291
Epoch [1/3], Step [1803/3236], Loss: 2.5229, Perplexity: 12.4653
Epoch [1/3], Step [1804/3236], Loss: 2.3876, Perplexity: 10.8874
Epoch [1/3], Step [1805/3236], Loss: 2.3066, Perplexity: 10.0397
Epoch [1/3], Step [1806/3236], Loss: 2.5929, Perplexity: 13.3690
Epoch [1/3], Step [1807/3236], Loss: 2.3655, Perplexity: 10.6493
Epoch [1/3], Step [1808/3236], Loss: 2.6267, Perplexity: 13.8287
Epoch [1/3], Step [1809/3236], Loss: 2.3874, Perplexity: 10.8854
Epoch [1/3], Step [1810/3236], Loss: 2.3332, Perplexity: 10.3112
Epoch [1/3], Step [1811/3236], Loss: 2.3225, Perplexity: 10.2012
Epoch [1/3], Step [1812/3236], Loss: 2.5068, Perplexity: 12.2652
Epoch [1/3], Step [1813/3236], Loss: 2.3602, Perplexity: 10.5926
Epoch [1/3], Step [1814/3236], Loss: 2.4538, Perplexity: 11.6322
Epoch [1/3], Step [1815/3236], Loss: 2.7293, Perplexity: 15.3214
Epoch [1/3], Step [1816/3236], Loss: 2.4860, Perplexity: 12.0128
Epoch [1/3], Step [1817/3236], Loss: 2.4043, Perplexity: 11.0708
Epoch [1/3], Step [1818/3236], Loss: 2.4282, Perplexity: 11.3388
Epoch [1/3], Step [1819/3236], Loss: 2.4254, Perplexity: 11.3073
Epoch [1/3], Step [1820/3236], Loss: 2.7635, Perplexity: 15.8552
Epoch [1/3], Step [1821/3236], Loss: 2.4356, Perplexity: 11.4226
Epoch [1/3], Step [1822/3236], Loss: 2.4918, Perplexity: 12.0833
Epoch [1/3], Step [1823/3236], Loss: 2.2919, Perplexity: 9.8936
Epoch [1/3], Step [1824/3236], Loss: 2.4254, Perplexity: 11.3065
Epoch [1/3], Step [1825/3236], Loss: 2.5080, Perplexity: 12.2807
Epoch [1/3], Step [1826/3236], Loss: 2.4324, Perplexity: 11.3861
Epoch [1/3], Step [1827/3236], Loss: 2.4955, Perplexity: 12.1275
Epoch [1/3], Step [1828/3236], Loss: 2.4519, Perplexity: 11.6103
Epoch [1/3], Step [1829/3236], Loss: 2.3344, Perplexity: 10.3236
Epoch [1/3], Step [1830/3236], Loss: 2.7450, Perplexity: 15.5648
Epoch [1/3], Step [1831/3236], Loss: 2.4003, Perplexity: 11.0264
Epoch [1/3], Step [1832/3236], Loss: 2.3244, Perplexity: 10.2201
Epoch [1/3], Step [1833/3236], Loss: 2.4860, Perplexity: 12.0134
Epoch [1/3], Step [1834/3236], Loss: 2.4801, Perplexity: 11.9423
Epoch [1/3], Step [1835/3236], Loss: 2.4314, Perplexity: 11.3743
Epoch [1/3], Step [1836/3236], Loss: 2.3959, Perplexity: 10.9782
Epoch [1/3], Step [1837/3236], Loss: 2.4820, Perplexity: 11.9657
Epoch [1/3], Step [1838/3236], Loss: 2.5138, Perplexity: 12.3521
Epoch [1/3], Step [1839/3236], Loss: 2.3277, Perplexity: 10.2544
Epoch [1/3], Step [1840/3236], Loss: 2.4308, Perplexity: 11.3676
Epoch [1/3], Step [1841/3236], Loss: 2.2504, Perplexity: 9.4914
Epoch [1/3], Step [1842/3236], Loss: 2.3752, Perplexity: 10.7527
Epoch [1/3], Step [1843/3236], Loss: 2.4187, Perplexity: 11.2315
Epoch [1/3], Step [1844/3236], Loss: 2.9508, Perplexity: 19.1218
Epoch [1/3], Step [1845/3236], Loss: 2.2448, Perplexity: 9.4382
Epoch [1/3], Step [1846/3236], Loss: 2.4262, Perplexity: 11.3161
Epoch [1/3], Step [1847/3236], Loss: 2.2426, Perplexity: 9.4181
Epoch [1/3], Step [1848/3236], Loss: 2.2488, Perplexity: 9.4761
Epoch [1/3], Step [1849/3236], Loss: 2.3220, Perplexity: 10.1960
Epoch [1/3], Step [1850/3236], Loss: 2.4360, Perplexity: 11.4272
Epoch [1/3], Step [1851/3236], Loss: 2.2749, Perplexity: 9.7268
Epoch [1/3], Step [1852/3236], Loss: 2.3961, Perplexity: 10.9799
Epoch [1/3], Step [1853/3236], Loss: 2.4203, Perplexity: 11.2489
Epoch [1/3], Step [1854/3236], Loss: 2.6288, Perplexity: 13.8578
Epoch [1/3], Step [1855/3236], Loss: 2.6180, Perplexity: 13.7089
Epoch [1/3], Step [1856/3236], Loss: 3.5185, Perplexity: 33.7351
Epoch [1/3], Step [1857/3236], Loss: 2.3682, Perplexity: 10.6782
Epoch [1/3], Step [1858/3236], Loss: 2.3981, Perplexity: 11.0021
Epoch [1/3], Step [1859/3236], Loss: 2.4895, Perplexity: 12.0551
Epoch [1/3], Step [1860/3236], Loss: 2.4498, Perplexity: 11.5855
Epoch [1/3], Step [1861/3236], Loss: 2.3926, Perplexity: 10.9425
Epoch [1/3], Step [1862/3236], Loss: 2.3909, Perplexity: 10.9229
Epoch [1/3], Step [1863/3236], Loss: 2.6243, Perplexity: 13.7953
Epoch [1/3], Step [1864/3236], Loss: 2.3815, Perplexity: 10.8208
Epoch [1/3], Step [1865/3236], Loss: 2.4525, Perplexity: 11.6172
Epoch [1/3], Step [1866/3236], Loss: 2.3994, Perplexity: 11.0167
Epoch [1/3], Step [1867/3236], Loss: 2.3792, Perplexity: 10.7961
Epoch [1/3], Step [1868/3236], Loss: 2.3176, Perplexity: 10.1514
Epoch [1/3], Step [1869/3236], Loss: 2.4435, Perplexity: 11.5129
Epoch [1/3], Step [1870/3236], Loss: 2.4188, Perplexity: 11.2325
Epoch [1/3], Step [1871/3236], Loss: 2.7104, Perplexity: 15.0347
Epoch [1/3], Step [1872/3236], Loss: 2.6056, Perplexity: 13.5394
Epoch [1/3], Step [1873/3236], Loss: 2.3074, Perplexity: 10.0485
Epoch [1/3], Step [1874/3236], Loss: 2.3706, Perplexity: 10.7036
Epoch [1/3], Step [1875/3236], Loss: 2.2470, Perplexity: 9.4592
Epoch [1/3], Step [1876/3236], Loss: 2.4006, Perplexity: 11.0300
Epoch [1/3], Step [1877/3236], Loss: 2.2695, Perplexity: 9.6746
Epoch [1/3], Step [1878/3236], Loss: 2.4625, Perplexity: 11.7338
Epoch [1/3], Step [1879/3236], Loss: 2.3675, Perplexity: 10.6704
Epoch [1/3], Step [1880/3236], Loss: 2.5845, Perplexity: 13.2563
Epoch [1/3], Step [1881/3236], Loss: 2.2521, Perplexity: 9.5073
Epoch [1/3], Step [1882/3236], Loss: 2.4509, Perplexity: 11.5983
Epoch [1/3], Step [1883/3236], Loss: 2.3875, Perplexity: 10.8861
Epoch [1/3], Step [1884/3236], Loss: 2.3644, Perplexity: 10.6375
Epoch [1/3], Step [1885/3236], Loss: 2.2842, Perplexity: 9.8177
Epoch [1/3], Step [1886/3236], Loss: 2.5171, Perplexity: 12.3925
Epoch [1/3], Step [1887/3236], Loss: 2.5065, Perplexity: 12.2622
Epoch [1/3], Step [1888/3236], Loss: 2.6303, Perplexity: 13.8772
Epoch [1/3], Step [1889/3236], Loss: 2.4669, Perplexity: 11.7857
Epoch [1/3], Step [1890/3236], Loss: 2.2625, Perplexity: 9.6075
Epoch [1/3], Step [1891/3236], Loss: 2.8266, Perplexity: 16.8878
Epoch [1/3], Step [1892/3236], Loss: 2.3308, Perplexity: 10.2865
Epoch [1/3], Step [1893/3236], Loss: 2.4232, Perplexity: 11.2817
Epoch [1/3], Step [1894/3236], Loss: 2.3818, Perplexity: 10.8247
Epoch [1/3], Step [1895/3236], Loss: 2.5406, Perplexity: 12.6874
Epoch [1/3], Step [1896/3236], Loss: 2.4652, Perplexity: 11.7654
Epoch [1/3], Step [1897/3236], Loss: 2.3473, Perplexity: 10.4572
Epoch [1/3], Step [1898/3236], Loss: 2.4581, Perplexity: 11.6830
Epoch [1/3], Step [1899/3236], Loss: 3.8101, Perplexity: 45.1554
Epoch [1/3], Step [1900/3236], Loss: 2.4797, Perplexity: 11.9382
Epoch [1/3], Step [1901/3236], Loss: 2.3641, Perplexity: 10.6345
Epoch [1/3], Step [1902/3236], Loss: 2.3550, Perplexity: 10.5379
Epoch [1/3], Step [1903/3236], Loss: 2.4716, Perplexity: 11.8408
Epoch [1/3], Step [1904/3236], Loss: 2.3994, Perplexity: 11.0167
Epoch [1/3], Step [1905/3236], Loss: 2.4902, Perplexity: 12.0636
Epoch [1/3], Step [1906/3236], Loss: 2.5002, Perplexity: 12.1844
Epoch [1/3], Step [1907/3236], Loss: 2.5012, Perplexity: 12.1975
Epoch [1/3], Step [1908/3236], Loss: 2.4226, Perplexity: 11.2747
Epoch [1/3], Step [1909/3236], Loss: 2.4547, Perplexity: 11.6431
Epoch [1/3], Step [1910/3236], Loss: 2.3550, Perplexity: 10.5380
Epoch [1/3], Step [1911/3236], Loss: 2.5373, Perplexity: 12.6454
Epoch [1/3], Step [1912/3236], Loss: 3.4545, Perplexity: 31.6412
Epoch [1/3], Step [1913/3236], Loss: 2.2372, Perplexity: 9.3673
Epoch [1/3], Step [1914/3236], Loss: 2.2525, Perplexity: 9.5113
Epoch [1/3], Step [1915/3236], Loss: 2.4668, Perplexity: 11.7850
Epoch [1/3], Step [1916/3236], Loss: 2.4355, Perplexity: 11.4220
Epoch [1/3], Step [1917/3236], Loss: 2.3975, Perplexity: 10.9954
Epoch [1/3], Step [1918/3236], Loss: 2.4203, Perplexity: 11.2488
Epoch [1/3], Step [1919/3236], Loss: 3.6261, Perplexity: 37.5675
Epoch [1/3], Step [1920/3236], Loss: 2.4696, Perplexity: 11.8180
Epoch [1/3], Step [1921/3236], Loss: 2.5034, Perplexity: 12.2245
Epoch [1/3], Step [1922/3236], Loss: 2.2675, Perplexity: 9.6557
Epoch [1/3], Step [1923/3236], Loss: 2.5315, Perplexity: 12.5725
Epoch [1/3], Step [1924/3236], Loss: 2.3960, Perplexity: 10.9797
Epoch [1/3], Step [1925/3236], Loss: 2.2982, Perplexity: 9.9558
Epoch [1/3], Step [1926/3236], Loss: 2.4765, Perplexity: 11.8991
Epoch [1/3], Step [1927/3236], Loss: 2.3033, Perplexity: 10.0074
Epoch [1/3], Step [1928/3236], Loss: 2.2737, Perplexity: 9.7148
Epoch [1/3], Step [1929/3236], Loss: 2.9554, Perplexity: 19.2086
Epoch [1/3], Step [1930/3236], Loss: 2.5261, Perplexity: 12.5052
Epoch [1/3], Step [1931/3236], Loss: 2.4521, Perplexity: 11.6123
Epoch [1/3], Step [1932/3236], Loss: 2.4227, Perplexity: 11.2767
Epoch [1/3], Step [1933/3236], Loss: 2.4200, Perplexity: 11.2461
Epoch [1/3], Step [1934/3236], Loss: 2.2071, Perplexity: 9.0898
Epoch [1/3], Step [1935/3236], Loss: 2.4406, Perplexity: 11.4801
Epoch [1/3], Step [1936/3236], Loss: 2.4616, Perplexity: 11.7239
Epoch [1/3], Step [1937/3236], Loss: 2.3402, Perplexity: 10.3837
Epoch [1/3], Step [1938/3236], Loss: 2.6402, Perplexity: 14.0167
Epoch [1/3], Step [1939/3236], Loss: 2.4009, Perplexity: 11.0335
Epoch [1/3], Step [1940/3236], Loss: 2.3182, Perplexity: 10.1572
Epoch [1/3], Step [1941/3236], Loss: 2.4835, Perplexity: 11.9828
Epoch [1/3], Step [1942/3236], Loss: 2.4046, Perplexity: 11.0736
Epoch [1/3], Step [1943/3236], Loss: 2.4730, Perplexity: 11.8582
Epoch [1/3], Step [1944/3236], Loss: 2.7476, Perplexity: 15.6050
Epoch [1/3], Step [1945/3236], Loss: 2.3687, Perplexity: 10.6839
Epoch [1/3], Step [1946/3236], Loss: 2.4933, Perplexity: 12.1006
Epoch [1/3], Step [1947/3236], Loss: 2.5312, Perplexity: 12.5684
Epoch [1/3], Step [1948/3236], Loss: 2.2805, Perplexity: 9.7819
Epoch [1/3], Step [1949/3236], Loss: 2.3120, Perplexity: 10.0943
Epoch [1/3], Step [1950/3236], Loss: 2.2888, Perplexity: 9.8633
Epoch [1/3], Step [1951/3236], Loss: 2.3227, Perplexity: 10.2031
Epoch [1/3], Step [1952/3236], Loss: 2.6477, Perplexity: 14.1214
Epoch [1/3], Step [1953/3236], Loss: 2.3995, Perplexity: 11.0182
Epoch [1/3], Step [1954/3236], Loss: 2.2271, Perplexity: 9.2725
Epoch [1/3], Step [1955/3236], Loss: 2.3117, Perplexity: 10.0912
Epoch [1/3], Step [1956/3236], Loss: 2.3813, Perplexity: 10.8187
Epoch [1/3], Step [1957/3236], Loss: 2.6079, Perplexity: 13.5701
Epoch [1/3], Step [1958/3236], Loss: 3.6070, Perplexity: 36.8540
Epoch [1/3], Step [1959/3236], Loss: 2.4930, Perplexity: 12.0969
Epoch [1/3], Step [1960/3236], Loss: 2.4045, Perplexity: 11.0733
Epoch [1/3], Step [1961/3236], Loss: 2.4593, Perplexity: 11.6962
Epoch [1/3], Step [1962/3236], Loss: 2.3893, Perplexity: 10.9057
Epoch [1/3], Step [1963/3236], Loss: 2.2882, Perplexity: 9.8573
Epoch [1/3], Step [1964/3236], Loss: 2.4437, Perplexity: 11.5157
Epoch [1/3], Step [1965/3236], Loss: 2.4529, Perplexity: 11.6217
Epoch [1/3], Step [1966/3236], Loss: 2.5723, Perplexity: 13.0961
Epoch [1/3], Step [1967/3236], Loss: 2.4825, Perplexity: 11.9707
Epoch [1/3], Step [1968/3236], Loss: 2.4139, Perplexity: 11.1780
Epoch [1/3], Step [1969/3236], Loss: 2.3251, Perplexity: 10.2275
Epoch [1/3], Step [1970/3236], Loss: 2.3455, Perplexity: 10.4389
Epoch [1/3], Step [1971/3236], Loss: 2.4375, Perplexity: 11.4441
Epoch [1/3], Step [1972/3236], Loss: 2.2778, Perplexity: 9.7549
Epoch [1/3], Step [1973/3236], Loss: 2.3478, Perplexity: 10.4630
Epoch [1/3], Step [1974/3236], Loss: 2.2180, Perplexity: 9.1888
Epoch [1/3], Step [1975/3236], Loss: 2.3188, Perplexity: 10.1640
Epoch [1/3], Step [1976/3236], Loss: 2.3640, Perplexity: 10.6338
Epoch [1/3], Step [1977/3236], Loss: 2.3504, Perplexity: 10.4893
Epoch [1/3], Step [1978/3236], Loss: 2.4544, Perplexity: 11.6391
Epoch [1/3], Step [1979/3236], Loss: 2.3364, Perplexity: 10.3440
Epoch [1/3], Step [1980/3236], Loss: 2.3276, Perplexity: 10.2530
Epoch [1/3], Step [1981/3236], Loss: 2.4034, Perplexity: 11.0611
Epoch [1/3], Step [1982/3236], Loss: 2.1818, Perplexity: 8.8624
Epoch [1/3], Step [1983/3236], Loss: 2.3228, Perplexity: 10.2038
Epoch [1/3], Step [1984/3236], Loss: 2.8607, Perplexity: 17.4731
Epoch [1/3], Step [1985/3236], Loss: 2.6277, Perplexity: 13.8418
Epoch [1/3], Step [1986/3236], Loss: 2.4132, Perplexity: 11.1698
Epoch [1/3], Step [1987/3236], Loss: 2.3575, Perplexity: 10.5645
Epoch [1/3], Step [1988/3236], Loss: 2.4827, Perplexity: 11.9738
Epoch [1/3], Step [1989/3236], Loss: 2.2753, Perplexity: 9.7311
Epoch [1/3], Step [1990/3236], Loss: 2.2605, Perplexity: 9.5881
Epoch [1/3], Step [1991/3236], Loss: 2.3276, Perplexity: 10.2532
Epoch [1/3], Step [1992/3236], Loss: 2.4788, Perplexity: 11.9274
Epoch [1/3], Step [1993/3236], Loss: 2.3800, Perplexity: 10.8051
Epoch [1/3], Step [1994/3236], Loss: 2.3801, Perplexity: 10.8058
Epoch [1/3], Step [1995/3236], Loss: 2.2075, Perplexity: 9.0931
Epoch [1/3], Step [1996/3236], Loss: 2.3153, Perplexity: 10.1281
Epoch [1/3], Step [1997/3236], Loss: 2.2916, Perplexity: 9.8912
Epoch [1/3], Step [1998/3236], Loss: 3.0577, Perplexity: 21.2777
Epoch [1/3], Step [1999/3236], Loss: 2.5616, Perplexity: 12.9572
Epoch [1/3], Step [2000/3236], Loss: 3.3091, Perplexity: 27.3605
Epoch [1/3], Step [2001/3236], Loss: 2.1826, Perplexity: 8.8691
Epoch [1/3], Step [2002/3236], Loss: 2.6346, Perplexity: 13.9382
Epoch [1/3], Step [2003/3236], Loss: 2.2890, Perplexity: 9.8650
Epoch [1/3], Step [2004/3236], Loss: 2.4529, Perplexity: 11.6217
Epoch [1/3], Step [2005/3236], Loss: 2.4204, Perplexity: 11.2501
Epoch [1/3], Step [2006/3236], Loss: 3.7621, Perplexity: 43.0385
Epoch [1/3], Step [2007/3236], Loss: 2.2787, Perplexity: 9.7643
Epoch [1/3], Step [2008/3236], Loss: 2.4956, Perplexity: 12.1294
Epoch [1/3], Step [2009/3236], Loss: 2.5189, Perplexity: 12.4155
Epoch [1/3], Step [2010/3236], Loss: 2.3119, Perplexity: 10.0936
Epoch [1/3], Step [2011/3236], Loss: 2.3894, Perplexity: 10.9069
Epoch [1/3], Step [2012/3236], Loss: 2.5953, Perplexity: 13.4006
Epoch [1/3], Step [2013/3236], Loss: 2.3768, Perplexity: 10.7705
Epoch [1/3], Step [2014/3236], Loss: 2.4346, Perplexity: 11.4110
Epoch [1/3], Step [2015/3236], Loss: 2.3010, Perplexity: 9.9841
Epoch [1/3], Step [2016/3236], Loss: 2.2952, Perplexity: 9.9265
Epoch [1/3], Step [2017/3236], Loss: 2.3182, Perplexity: 10.1570
Epoch [1/3], Step [2018/3236], Loss: 2.5295, Perplexity: 12.5468
Epoch [1/3], Step [2019/3236], Loss: 2.5877, Perplexity: 13.2990
Epoch [1/3], Step [2020/3236], Loss: 2.3861, Perplexity: 10.8713
Epoch [1/3], Step [2021/3236], Loss: 2.2266, Perplexity: 9.2679
Epoch [1/3], Step [2022/3236], Loss: 2.3176, Perplexity: 10.1513
Epoch [1/3], Step [2023/3236], Loss: 3.0958, Perplexity: 22.1045
Epoch [1/3], Step [2024/3236], Loss: 2.3158, Perplexity: 10.1334
Epoch [1/3], Step [2025/3236], Loss: 2.4474, Perplexity: 11.5586
Epoch [1/3], Step [2026/3236], Loss: 2.3665, Perplexity: 10.6601
Epoch [1/3], Step [2027/3236], Loss: 2.3807, Perplexity: 10.8125
Epoch [1/3], Step [2028/3236], Loss: 2.4612, Perplexity: 11.7184
Epoch [1/3], Step [2029/3236], Loss: 2.3525, Perplexity: 10.5123
Epoch [1/3], Step [2030/3236], Loss: 2.3576, Perplexity: 10.5655
Epoch [1/3], Step [2031/3236], Loss: 2.3634, Perplexity: 10.6268
Epoch [1/3], Step [2032/3236], Loss: 2.5683, Perplexity: 13.0437
Epoch [1/3], Step [2033/3236], Loss: 2.4052, Perplexity: 11.0812
Epoch [1/3], Step [2034/3236], Loss: 2.6370, Perplexity: 13.9712
Epoch [1/3], Step [2035/3236], Loss: 2.3735, Perplexity: 10.7354
Epoch [1/3], Step [2036/3236], Loss: 2.3669, Perplexity: 10.6640
Epoch [1/3], Step [2037/3236], Loss: 2.2713, Perplexity: 9.6916
Epoch [1/3], Step [2038/3236], Loss: 2.3088, Perplexity: 10.0627
Epoch [1/3], Step [2039/3236], Loss: 2.2883, Perplexity: 9.8584
Epoch [1/3], Step [2040/3236], Loss: 2.3574, Perplexity: 10.5639
Epoch [1/3], Step [2041/3236], Loss: 2.2570, Perplexity: 9.5544
Epoch [1/3], Step [2042/3236], Loss: 2.5014, Perplexity: 12.1997
Epoch [1/3], Step [2043/3236], Loss: 2.3608, Perplexity: 10.5999
Epoch [1/3], Step [2044/3236], Loss: 2.4178, Perplexity: 11.2214
Epoch [1/3], Step [2045/3236], Loss: 2.2685, Perplexity: 9.6647
Epoch [1/3], Step [2046/3236], Loss: 2.2441, Perplexity: 9.4323
Epoch [1/3], Step [2047/3236], Loss: 2.3895, Perplexity: 10.9078
Epoch [1/3], Step [2048/3236], Loss: 2.4335, Perplexity: 11.3985
Epoch [1/3], Step [2049/3236], Loss: 2.3768, Perplexity: 10.7700
Epoch [1/3], Step [2050/3236], Loss: 2.3060, Perplexity: 10.0340
Epoch [1/3], Step [2051/3236], Loss: 2.7596, Perplexity: 15.7934
Epoch [1/3], Step [2052/3236], Loss: 3.1618, Perplexity: 23.6119
Epoch [1/3], Step [2053/3236], Loss: 2.5100, Perplexity: 12.3054
Epoch [1/3], Step [2054/3236], Loss: 2.3053, Perplexity: 10.0270
Epoch [1/3], Step [2055/3236], Loss: 2.5027, Perplexity: 12.2150
Epoch [1/3], Step [2056/3236], Loss: 2.6464, Perplexity: 14.1027
Epoch [1/3], Step [2057/3236], Loss: 2.3416, Perplexity: 10.3977
Epoch [1/3], Step [2058/3236], Loss: 2.3817, Perplexity: 10.8235
Epoch [1/3], Step [2059/3236], Loss: 2.3003, Perplexity: 9.9777
Epoch [1/3], Step [2060/3236], Loss: 2.3976, Perplexity: 10.9965
Epoch [1/3], Step [2061/3236], Loss: 2.5278, Perplexity: 12.5255
Epoch [1/3], Step [2062/3236], Loss: 2.3567, Perplexity: 10.5561
Epoch [1/3], Step [2063/3236], Loss: 2.3789, Perplexity: 10.7934
Epoch [1/3], Step [2064/3236], Loss: 2.2798, Perplexity: 9.7744
Epoch [1/3], Step [2065/3236], Loss: 2.4714, Perplexity: 11.8387
Epoch [1/3], Step [2066/3236], Loss: 3.1379, Perplexity: 23.0552
Epoch [1/3], Step [2067/3236], Loss: 2.4896, Perplexity: 12.0569
Epoch [1/3], Step [2068/3236], Loss: 2.4057, Perplexity: 11.0863
Epoch [1/3], Step [2069/3236], Loss: 2.2460, Perplexity: 9.4497
Epoch [1/3], Step [2070/3236], Loss: 2.5012, Perplexity: 12.1975
Epoch [1/3], Step [2071/3236], Loss: 2.3739, Perplexity: 10.7388
Epoch [1/3], Step [2072/3236], Loss: 2.2624, Perplexity: 9.6057
Epoch [1/3], Step [2073/3236], Loss: 2.4562, Perplexity: 11.6604
Epoch [1/3], Step [2074/3236], Loss: 2.3717, Perplexity: 10.7154
Epoch [1/3], Step [2075/3236], Loss: 2.4049, Perplexity: 11.0772
Epoch [1/3], Step [2076/3236], Loss: 2.7301, Perplexity: 15.3342
Epoch [1/3], Step [2077/3236], Loss: 2.2755, Perplexity: 9.7327
Epoch [1/3], Step [2078/3236], Loss: 2.2443, Perplexity: 9.4335
Epoch [1/3], Step [2079/3236], Loss: 2.4335, Perplexity: 11.3983
Epoch [1/3], Step [2080/3236], Loss: 2.3416, Perplexity: 10.3975
Epoch [1/3], Step [2081/3236], Loss: 2.2933, Perplexity: 9.9074
Epoch [1/3], Step [2082/3236], Loss: 2.4523, Perplexity: 11.6145
Epoch [1/3], Step [2083/3236], Loss: 2.4611, Perplexity: 11.7174
Epoch [1/3], Step [2084/3236], Loss: 2.3268, Perplexity: 10.2449
Epoch [1/3], Step [2085/3236], Loss: 2.3691, Perplexity: 10.6881
Epoch [1/3], Step [2086/3236], Loss: 2.3964, Perplexity: 10.9835
Epoch [1/3], Step [2087/3236], Loss: 2.4472, Perplexity: 11.5559
Epoch [1/3], Step [2088/3236], Loss: 2.3238, Perplexity: 10.2145
Epoch [1/3], Step [2089/3236], Loss: 2.4690, Perplexity: 11.8104
Epoch [1/3], Step [2090/3236], Loss: 2.3553, Perplexity: 10.5408
Epoch [1/3], Step [2091/3236], Loss: 2.2756, Perplexity: 9.7340
Epoch [1/3], Step [2092/3236], Loss: 2.4304, Perplexity: 11.3635
Epoch [1/3], Step [2093/3236], Loss: 2.3864, Perplexity: 10.8744
Epoch [1/3], Step [2094/3236], Loss: 2.4989, Perplexity: 12.1693
Epoch [1/3], Step [2095/3236], Loss: 2.1066, Perplexity: 8.2199
Epoch [1/3], Step [2096/3236], Loss: 2.3959, Perplexity: 10.9780
Epoch [1/3], Step [2097/3236], Loss: 2.3793, Perplexity: 10.7972
Epoch [1/3], Step [2098/3236], Loss: 2.3659, Perplexity: 10.6535
Epoch [1/3], Step [2099/3236], Loss: 2.1697, Perplexity: 8.7553
Epoch [1/3], Step [2100/3236], Loss: 2.3513, Perplexity: 10.4992
Epoch [1/3], Step [2101/3236], Loss: 2.3978, Perplexity: 10.9988
Epoch [1/3], Step [2102/3236], Loss: 3.1092, Perplexity: 22.4032
Epoch [1/3], Step [2103/3236], Loss: 2.3182, Perplexity: 10.1574
Epoch [1/3], Step [2104/3236], Loss: 2.2492, Perplexity: 9.4799
Epoch [1/3], Step [2105/3236], Loss: 4.0883, Perplexity: 59.6402
Epoch [1/3], Step [2106/3236], Loss: 2.2857, Perplexity: 9.8321
Epoch [1/3], Step [2107/3236], Loss: 2.2788, Perplexity: 9.7647
Epoch [1/3], Step [2108/3236], Loss: 2.2927, Perplexity: 9.9013
Epoch [1/3], Step [2109/3236], Loss: 2.4873, Perplexity: 12.0291
Epoch [1/3], Step [2110/3236], Loss: 2.3287, Perplexity: 10.2645
Epoch [1/3], Step [2111/3236], Loss: 2.4429, Perplexity: 11.5060
Epoch [1/3], Step [2112/3236], Loss: 2.3529, Perplexity: 10.5165
Epoch [1/3], Step [2113/3236], Loss: 2.2533, Perplexity: 9.5189
Epoch [1/3], Step [2114/3236], Loss: 2.3462, Perplexity: 10.4454
Epoch [1/3], Step [2115/3236], Loss: 2.6617, Perplexity: 14.3213
Epoch [1/3], Step [2116/3236], Loss: 2.8859, Perplexity: 17.9197
Epoch [1/3], Step [2117/3236], Loss: 2.4592, Perplexity: 11.6959
Epoch [1/3], Step [2118/3236], Loss: 2.6888, Perplexity: 14.7144
Epoch [1/3], Step [2119/3236], Loss: 2.2708, Perplexity: 9.6868
Epoch [1/3], Step [2120/3236], Loss: 2.2656, Perplexity: 9.6366
Epoch [1/3], Step [2121/3236], Loss: 2.1564, Perplexity: 8.6396
Epoch [1/3], Step [2122/3236], Loss: 2.2901, Perplexity: 9.8758
Epoch [1/3], Step [2123/3236], Loss: 2.5219, Perplexity: 12.4518
Epoch [1/3], Step [2124/3236], Loss: 2.2948, Perplexity: 9.9220
Epoch [1/3], Step [2125/3236], Loss: 2.3655, Perplexity: 10.6495
Epoch [1/3], Step [2126/3236], Loss: 2.2622, Perplexity: 9.6046
Epoch [1/3], Step [2127/3236], Loss: 2.6576, Perplexity: 14.2624
Epoch [1/3], Step [2128/3236], Loss: 2.3438, Perplexity: 10.4204
Epoch [1/3], Step [2129/3236], Loss: 2.3543, Perplexity: 10.5311
Epoch [1/3], Step [2130/3236], Loss: 2.2352, Perplexity: 9.3484
Epoch [1/3], Step [2131/3236], Loss: 2.6659, Perplexity: 14.3815
Epoch [1/3], Step [2132/3236], Loss: 2.3146, Perplexity: 10.1210
Epoch [1/3], Step [2133/3236], Loss: 2.4299, Perplexity: 11.3575
Epoch [1/3], Step [2134/3236], Loss: 2.4176, Perplexity: 11.2190
Epoch [1/3], Step [2135/3236], Loss: 2.5938, Perplexity: 13.3811
Epoch [1/3], Step [2136/3236], Loss: 2.3744, Perplexity: 10.7441
Epoch [1/3], Step [2137/3236], Loss: 2.2527, Perplexity: 9.5134
Epoch [1/3], Step [2138/3236], Loss: 2.3471, Perplexity: 10.4553
Epoch [1/3], Step [2139/3236], Loss: 2.2795, Perplexity: 9.7723
Epoch [1/3], Step [2140/3236], Loss: 2.2981, Perplexity: 9.9552
Epoch [1/3], Step [2141/3236], Loss: 2.5860, Perplexity: 13.2769
Epoch [1/3], Step [2142/3236], Loss: 2.2946, Perplexity: 9.9202
Epoch [1/3], Step [2143/3236], Loss: 2.0274, Perplexity: 7.5941
Epoch [1/3], Step [2144/3236], Loss: 2.4859, Perplexity: 12.0124
Epoch [1/3], Step [2145/3236], Loss: 2.4815, Perplexity: 11.9597
Epoch [1/3], Step [2146/3236], Loss: 2.3707, Perplexity: 10.7049
Epoch [1/3], Step [2147/3236], Loss: 3.0508, Perplexity: 21.1320
Epoch [1/3], Step [2148/3236], Loss: 2.3499, Perplexity: 10.4846
Epoch [1/3], Step [2149/3236], Loss: 2.4411, Perplexity: 11.4851
Epoch [1/3], Step [2150/3236], Loss: 2.2604, Perplexity: 9.5867
Epoch [1/3], Step [2151/3236], Loss: 2.2606, Perplexity: 9.5888
Epoch [1/3], Step [2152/3236], Loss: 2.3406, Perplexity: 10.3877
Epoch [1/3], Step [2153/3236], Loss: 2.1907, Perplexity: 8.9414
Epoch [1/3], Step [2154/3236], Loss: 2.4446, Perplexity: 11.5254
Epoch [1/3], Step [2155/3236], Loss: 2.3023, Perplexity: 9.9974
Epoch [1/3], Step [2156/3236], Loss: 2.3309, Perplexity: 10.2867
Epoch [1/3], Step [2157/3236], Loss: 2.5279, Perplexity: 12.5278
Epoch [1/3], Step [2158/3236], Loss: 2.5826, Perplexity: 13.2316
Epoch [1/3], Step [2159/3236], Loss: 2.3971, Perplexity: 10.9917
Epoch [1/3], Step [2160/3236], Loss: 2.3270, Perplexity: 10.2469
Epoch [1/3], Step [2161/3236], Loss: 2.3450, Perplexity: 10.4330
Epoch [1/3], Step [2162/3236], Loss: 2.3270, Perplexity: 10.2470
Epoch [1/3], Step [2163/3236], Loss: 2.3791, Perplexity: 10.7953
Epoch [1/3], Step [2164/3236], Loss: 2.5110, Perplexity: 12.3170
Epoch [1/3], Step [2165/3236], Loss: 2.2589, Perplexity: 9.5723
Epoch [1/3], Step [2166/3236], Loss: 2.2958, Perplexity: 9.9319
Epoch [1/3], Step [2167/3236], Loss: 2.3136, Perplexity: 10.1103
Epoch [1/3], Step [2168/3236], Loss: 2.4876, Perplexity: 12.0322
Epoch [1/3], Step [2169/3236], Loss: 2.3113, Perplexity: 10.0871
Epoch [1/3], Step [2170/3236], Loss: 2.3048, Perplexity: 10.0221
Epoch [1/3], Step [2171/3236], Loss: 2.2859, Perplexity: 9.8349
Epoch [1/3], Step [2172/3236], Loss: 2.2745, Perplexity: 9.7234
Epoch [1/3], Step [2173/3236], Loss: 3.1927, Perplexity: 24.3541
Epoch [1/3], Step [2174/3236], Loss: 2.4289, Perplexity: 11.3464
Epoch [1/3], Step [2175/3236], Loss: 2.3808, Perplexity: 10.8136
Epoch [1/3], Step [2176/3236], Loss: 2.2409, Perplexity: 9.4021
Epoch [1/3], Step [2177/3236], Loss: 2.4022, Perplexity: 11.0472
Epoch [1/3], Step [2178/3236], Loss: 2.4059, Perplexity: 11.0886
Epoch [1/3], Step [2179/3236], Loss: 2.2941, Perplexity: 9.9156
Epoch [1/3], Step [2180/3236], Loss: 2.1781, Perplexity: 8.8291
Epoch [1/3], Step [2181/3236], Loss: 2.6716, Perplexity: 14.4637
Epoch [1/3], Step [2182/3236], Loss: 2.3905, Perplexity: 10.9190
Epoch [1/3], Step [2183/3236], Loss: 2.7238, Perplexity: 15.2376
Epoch [1/3], Step [2184/3236], Loss: 2.3982, Perplexity: 11.0030
Epoch [1/3], Step [2185/3236], Loss: 2.1443, Perplexity: 8.5357
Epoch [1/3], Step [2186/3236], Loss: 2.3731, Perplexity: 10.7301
Epoch [1/3], Step [2187/3236], Loss: 2.4454, Perplexity: 11.5346
Epoch [1/3], Step [2188/3236], Loss: 3.2104, Perplexity: 24.7890
Epoch [1/3], Step [2189/3236], Loss: 2.2507, Perplexity: 9.4945
Epoch [1/3], Step [2190/3236], Loss: 2.3865, Perplexity: 10.8751
Epoch [1/3], Step [2191/3236], Loss: 2.4197, Perplexity: 11.2419
Epoch [1/3], Step [2192/3236], Loss: 2.8057, Perplexity: 16.5393
Epoch [1/3], Step [2193/3236], Loss: 2.5990, Perplexity: 13.4503
Epoch [1/3], Step [2194/3236], Loss: 2.9060, Perplexity: 18.2836
Epoch [1/3], Step [2195/3236], Loss: 2.2624, Perplexity: 9.6057
Epoch [1/3], Step [2196/3236], Loss: 2.3673, Perplexity: 10.6685
Epoch [1/3], Step [2197/3236], Loss: 2.2597, Perplexity: 9.5798
Epoch [1/3], Step [2198/3236], Loss: 2.7109, Perplexity: 15.0426
Epoch [1/3], Step [2199/3236], Loss: 2.1081, Perplexity: 8.2322
Epoch [1/3], Step [2200/3236], Loss: 2.2070, Perplexity: 9.0884
Epoch [1/3], Step [2201/3236], Loss: 2.5003, Perplexity: 12.1859
Epoch [1/3], Step [2202/3236], Loss: 2.3340, Perplexity: 10.3195
Epoch [1/3], Step [2203/3236], Loss: 2.7460, Perplexity: 15.5796
Epoch [1/3], Step [2204/3236], Loss: 2.3374, Perplexity: 10.3542
Epoch [1/3], Step [2205/3236], Loss: 2.2479, Perplexity: 9.4682
Epoch [1/3], Step [2206/3236], Loss: 2.5543, Perplexity: 12.8624
Epoch [1/3], Step [2207/3236], Loss: 2.4912, Perplexity: 12.0754
Epoch [1/3], Step [2208/3236], Loss: 2.5903, Perplexity: 13.3336
Epoch [1/3], Step [2209/3236], Loss: 2.3315, Perplexity: 10.2939
Epoch [1/3], Step [2210/3236], Loss: 2.3579, Perplexity: 10.5692
Epoch [1/3], Step [2211/3236], Loss: 2.3798, Perplexity: 10.8031
Epoch [1/3], Step [2212/3236], Loss: 2.1401, Perplexity: 8.5001
Epoch [1/3], Step [2213/3236], Loss: 2.2882, Perplexity: 9.8572
Epoch [1/3], Step [2214/3236], Loss: 2.4149, Perplexity: 11.1888
Epoch [1/3], Step [2215/3236], Loss: 2.4145, Perplexity: 11.1841
Epoch [1/3], Step [2216/3236], Loss: 2.4700, Perplexity: 11.8223
Epoch [1/3], Step [2217/3236], Loss: 2.5735, Perplexity: 13.1115
Epoch [1/3], Step [2218/3236], Loss: 2.3470, Perplexity: 10.4545
Epoch [1/3], Step [2219/3236], Loss: 2.2662, Perplexity: 9.6424
Epoch [1/3], Step [2220/3236], Loss: 2.5327, Perplexity: 12.5875
Epoch [1/3], Step [2221/3236], Loss: 2.6211, Perplexity: 13.7509
Epoch [1/3], Step [2222/3236], Loss: 2.6032, Perplexity: 13.5073
Epoch [1/3], Step [2223/3236], Loss: 2.3021, Perplexity: 9.9951
Epoch [1/3], Step [2224/3236], Loss: 2.1839, Perplexity: 8.8810
Epoch [1/3], Step [2225/3236], Loss: 2.3733, Perplexity: 10.7330
Epoch [1/3], Step [2226/3236], Loss: 2.2136, Perplexity: 9.1482
Epoch [1/3], Step [2227/3236], Loss: 2.9263, Perplexity: 18.6584
Epoch [1/3], Step [2228/3236], Loss: 2.2059, Perplexity: 9.0782
Epoch [1/3], Step [2229/3236], Loss: 2.2321, Perplexity: 9.3191
Epoch [1/3], Step [2230/3236], Loss: 2.2352, Perplexity: 9.3484
Epoch [1/3], Step [2231/3236], Loss: 2.3242, Perplexity: 10.2186
Epoch [1/3], Step [2232/3236], Loss: 2.4224, Perplexity: 11.2732
Epoch [1/3], Step [2233/3236], Loss: 2.2706, Perplexity: 9.6855
Epoch [1/3], Step [2234/3236], Loss: 2.3286, Perplexity: 10.2631
Epoch [1/3], Step [2235/3236], Loss: 2.4232, Perplexity: 11.2822
Epoch [1/3], Step [2236/3236], Loss: 2.3502, Perplexity: 10.4880
Epoch [1/3], Step [2237/3236], Loss: 2.3043, Perplexity: 10.0173
Epoch [1/3], Step [2238/3236], Loss: 2.4218, Perplexity: 11.2659
Epoch [1/3], Step [2239/3236], Loss: 2.5039, Perplexity: 12.2300
Epoch [1/3], Step [2240/3236], Loss: 3.1503, Perplexity: 23.3436
Epoch [1/3], Step [2241/3236], Loss: 2.2961, Perplexity: 9.9357
Epoch [1/3], Step [2242/3236], Loss: 2.9110, Perplexity: 18.3747
Epoch [1/3], Step [2243/3236], Loss: 2.3960, Perplexity: 10.9789
Epoch [1/3], Step [2244/3236], Loss: 2.2576, Perplexity: 9.5598
Epoch [1/3], Step [2245/3236], Loss: 2.2072, Perplexity: 9.0903
Epoch [1/3], Step [2246/3236], Loss: 2.3345, Perplexity: 10.3241
Epoch [1/3], Step [2247/3236], Loss: 2.2421, Perplexity: 9.4129
Epoch [1/3], Step [2248/3236], Loss: 2.3705, Perplexity: 10.7032
Epoch [1/3], Step [2249/3236], Loss: 2.1400, Perplexity: 8.4998
Epoch [1/3], Step [2250/3236], Loss: 2.2353, Perplexity: 9.3494
Epoch [1/3], Step [2251/3236], Loss: 2.2936, Perplexity: 9.9103
Epoch [1/3], Step [2252/3236], Loss: 2.3465, Perplexity: 10.4491
Epoch [1/3], Step [2253/3236], Loss: 2.3249, Perplexity: 10.2253
Epoch [1/3], Step [2254/3236], Loss: 2.4017, Perplexity: 11.0416
Epoch [1/3], Step [2255/3236], Loss: 2.3515, Perplexity: 10.5009
Epoch [1/3], Step [2256/3236], Loss: 2.2677, Perplexity: 9.6568
Epoch [1/3], Step [2257/3236], Loss: 2.2287, Perplexity: 9.2882
Epoch [1/3], Step [2258/3236], Loss: 2.4138, Perplexity: 11.1764
Epoch [1/3], Step [2259/3236], Loss: 2.3944, Perplexity: 10.9613
Epoch [1/3], Step [2260/3236], Loss: 2.3300, Perplexity: 10.2781
Epoch [1/3], Step [2261/3236], Loss: 2.3919, Perplexity: 10.9346
Epoch [1/3], Step [2262/3236], Loss: 2.3821, Perplexity: 10.8277
Epoch [1/3], Step [2263/3236], Loss: 2.2708, Perplexity: 9.6873
Epoch [1/3], Step [2264/3236], Loss: 2.4751, Perplexity: 11.8833
Epoch [1/3], Step [2265/3236], Loss: 2.5746, Perplexity: 13.1260
Epoch [1/3], Step [2266/3236], Loss: 3.0218, Perplexity: 20.5290
Epoch [1/3], Step [2267/3236], Loss: 2.3414, Perplexity: 10.3956
Epoch [1/3], Step [2268/3236], Loss: 2.8061, Perplexity: 16.5459
Epoch [1/3], Step [2269/3236], Loss: 2.2862, Perplexity: 9.8376
Epoch [1/3], Step [2270/3236], Loss: 2.4026, Perplexity: 11.0522
Epoch [1/3], Step [2271/3236], Loss: 2.7797, Perplexity: 16.1150
Epoch [1/3], Step [2272/3236], Loss: 2.4447, Perplexity: 11.5274
Epoch [1/3], Step [2273/3236], Loss: 2.3049, Perplexity: 10.0234
Epoch [1/3], Step [2274/3236], Loss: 2.3259, Perplexity: 10.2362
Epoch [1/3], Step [2275/3236], Loss: 2.6809, Perplexity: 14.5981
Epoch [1/3], Step [2276/3236], Loss: 2.3294, Perplexity: 10.2721
Epoch [1/3], Step [2277/3236], Loss: 2.3423, Perplexity: 10.4052
Epoch [1/3], Step [2278/3236], Loss: 2.2229, Perplexity: 9.2342
Epoch [1/3], Step [2279/3236], Loss: 2.3074, Perplexity: 10.0488
Epoch [1/3], Step [2280/3236], Loss: 2.2650, Perplexity: 9.6311
Epoch [1/3], Step [2281/3236], Loss: 2.3598, Perplexity: 10.5889
Epoch [1/3], Step [2282/3236], Loss: 2.3564, Perplexity: 10.5532
Epoch [1/3], Step [2283/3236], Loss: 2.3480, Perplexity: 10.4645
Epoch [1/3], Step [2284/3236], Loss: 2.3947, Perplexity: 10.9644
Epoch [1/3], Step [2285/3236], Loss: 2.2268, Perplexity: 9.2704
Epoch [1/3], Step [2286/3236], Loss: 2.3653, Perplexity: 10.6470
Epoch [1/3], Step [2287/3236], Loss: 2.2059, Perplexity: 9.0788
Epoch [1/3], Step [2288/3236], Loss: 2.6819, Perplexity: 14.6132
Epoch [1/3], Step [2289/3236], Loss: 2.5624, Perplexity: 12.9669
Epoch [1/3], Step [2290/3236], Loss: 2.2375, Perplexity: 9.3700
Epoch [1/3], Step [2291/3236], Loss: 2.3357, Perplexity: 10.3366
Epoch [1/3], Step [2292/3236], Loss: 2.2191, Perplexity: 9.1987
Epoch [1/3], Step [2293/3236], Loss: 2.5085, Perplexity: 12.2866
Epoch [1/3], Step [2294/3236], Loss: 2.2876, Perplexity: 9.8517
Epoch [1/3], Step [2295/3236], Loss: 2.1863, Perplexity: 8.9022
Epoch [1/3], Step [2296/3236], Loss: 2.3516, Perplexity: 10.5021
Epoch [1/3], Step [2297/3236], Loss: 2.3498, Perplexity: 10.4832
Epoch [1/3], Step [2298/3236], Loss: 2.3830, Perplexity: 10.8375
Epoch [1/3], Step [2299/3236], Loss: 2.3811, Perplexity: 10.8172
Epoch [1/3], Step [2300/3236], Loss: 2.2896, Perplexity: 9.8712
Epoch [1/3], Step [2301/3236], Loss: 2.3699, Perplexity: 10.6966
Epoch [1/3], Step [2302/3236], Loss: 2.3999, Perplexity: 11.0221
Epoch [1/3], Step [2303/3236], Loss: 2.4850, Perplexity: 12.0009
Epoch [1/3], Step [2304/3236], Loss: 2.6873, Perplexity: 14.6920
Epoch [1/3], Step [2305/3236], Loss: 2.6243, Perplexity: 13.7955
Epoch [1/3], Step [2306/3236], Loss: 2.0691, Perplexity: 7.9180
Epoch [1/3], Step [2307/3236], Loss: 2.2488, Perplexity: 9.4760
Epoch [1/3], Step [2308/3236], Loss: 2.2522, Perplexity: 9.5088
Epoch [1/3], Step [2309/3236], Loss: 2.5091, Perplexity: 12.2944
Epoch [1/3], Step [2310/3236], Loss: 2.4037, Perplexity: 11.0646
Epoch [1/3], Step [2311/3236], Loss: 2.3358, Perplexity: 10.3376
Epoch [1/3], Step [2312/3236], Loss: 2.1700, Perplexity: 8.7579
Epoch [1/3], Step [2313/3236], Loss: 2.2483, Perplexity: 9.4720
Epoch [1/3], Step [2314/3236], Loss: 2.1782, Perplexity: 8.8308
Epoch [1/3], Step [2315/3236], Loss: 2.2588, Perplexity: 9.5711
Epoch [1/3], Step [2316/3236], Loss: 2.5940, Perplexity: 13.3831
Epoch [1/3], Step [2317/3236], Loss: 2.1946, Perplexity: 8.9768
Epoch [1/3], Step [2318/3236], Loss: 2.2917, Perplexity: 9.8920
Epoch [1/3], Step [2319/3236], Loss: 2.7593, Perplexity: 15.7887
Epoch [1/3], Step [2320/3236], Loss: 2.5327, Perplexity: 12.5877
Epoch [1/3], Step [2321/3236], Loss: 2.5128, Perplexity: 12.3395
Epoch [1/3], Step [2322/3236], Loss: 2.3082, Perplexity: 10.0568
Epoch [1/3], Step [2323/3236], Loss: 2.3745, Perplexity: 10.7455
Epoch [1/3], Step [2324/3236], Loss: 2.4935, Perplexity: 12.1038
Epoch [1/3], Step [2325/3236], Loss: 2.5486, Perplexity: 12.7889
Epoch [1/3], Step [2326/3236], Loss: 2.2771, Perplexity: 9.7483
Epoch [1/3], Step [2327/3236], Loss: 2.4132, Perplexity: 11.1691
Epoch [1/3], Step [2328/3236], Loss: 2.3874, Perplexity: 10.8855
Epoch [1/3], Step [2329/3236], Loss: 2.2154, Perplexity: 9.1652
Epoch [1/3], Step [2330/3236], Loss: 2.2217, Perplexity: 9.2233
Epoch [1/3], Step [2331/3236], Loss: 2.2760, Perplexity: 9.7380
Epoch [1/3], Step [2332/3236], Loss: 2.4662, Perplexity: 11.7781
Epoch [1/3], Step [2333/3236], Loss: 2.3766, Perplexity: 10.7684
Epoch [1/3], Step [2334/3236], Loss: 2.2829, Perplexity: 9.8050
Epoch [1/3], Step [2335/3236], Loss: 2.9579, Perplexity: 19.2566
Epoch [1/3], Step [2336/3236], Loss: 3.4126, Perplexity: 30.3425
Epoch [1/3], Step [2337/3236], Loss: 2.2172, Perplexity: 9.1813
Epoch [1/3], Step [2338/3236], Loss: 2.2973, Perplexity: 9.9476
Epoch [1/3], Step [2339/3236], Loss: 2.3782, Perplexity: 10.7858
Epoch [1/3], Step [2340/3236], Loss: 2.6778, Perplexity: 14.5532
Epoch [1/3], Step [2341/3236], Loss: 2.2161, Perplexity: 9.1715
Epoch [1/3], Step [2342/3236], Loss: 2.2919, Perplexity: 9.8940
Epoch [1/3], Step [2343/3236], Loss: 2.4841, Perplexity: 11.9898
Epoch [1/3], Step [2344/3236], Loss: 2.2329, Perplexity: 9.3273
Epoch [1/3], Step [2345/3236], Loss: 2.4942, Perplexity: 12.1121
Epoch [1/3], Step [2346/3236], Loss: 3.1410, Perplexity: 23.1259
Epoch [1/3], Step [2347/3236], Loss: 2.4420, Perplexity: 11.4954
Epoch [1/3], Step [2348/3236], Loss: 2.4232, Perplexity: 11.2821
Epoch [1/3], Step [2349/3236], Loss: 2.5422, Perplexity: 12.7070
Epoch [1/3], Step [2350/3236], Loss: 2.3965, Perplexity: 10.9846
Epoch [1/3], Step [2351/3236], Loss: 2.3961, Perplexity: 10.9805
Epoch [1/3], Step [2352/3236], Loss: 2.4126, Perplexity: 11.1628
Epoch [1/3], Step [2353/3236], Loss: 2.4110, Perplexity: 11.1454
Epoch [1/3], Step [2354/3236], Loss: 2.2658, Perplexity: 9.6392
Epoch [1/3], Step [2355/3236], Loss: 2.4530, Perplexity: 11.6234
Epoch [1/3], Step [2356/3236], Loss: 2.6714, Perplexity: 14.4609
Epoch [1/3], Step [2357/3236], Loss: 3.0696, Perplexity: 21.5337
Epoch [1/3], Step [2358/3236], Loss: 2.7190, Perplexity: 15.1659
Epoch [1/3], Step [2359/3236], Loss: 2.1217, Perplexity: 8.3454
Epoch [1/3], Step [2360/3236], Loss: 2.1640, Perplexity: 8.7055
Epoch [1/3], Step [2361/3236], Loss: 2.4309, Perplexity: 11.3687
Epoch [1/3], Step [2362/3236], Loss: 2.4405, Perplexity: 11.4790
Epoch [1/3], Step [2363/3236], Loss: 2.2235, Perplexity: 9.2395
Epoch [1/3], Step [2364/3236], Loss: 2.4358, Perplexity: 11.4253
Epoch [1/3], Step [2365/3236], Loss: 2.1651, Perplexity: 8.7156
Epoch [1/3], Step [2366/3236], Loss: 2.2752, Perplexity: 9.7301
Epoch [1/3], Step [2367/3236], Loss: 2.2105, Perplexity: 9.1202
Epoch [1/3], Step [2368/3236], Loss: 2.3071, Perplexity: 10.0457
Epoch [1/3], Step [2369/3236], Loss: 2.4657, Perplexity: 11.7712
Epoch [1/3], Step [2370/3236], Loss: 2.4014, Perplexity: 11.0384
Epoch [1/3], Step [2371/3236], Loss: 2.2416, Perplexity: 9.4084
Epoch [1/3], Step [2372/3236], Loss: 2.1757, Perplexity: 8.8082
Epoch [1/3], Step [2373/3236], Loss: 2.2644, Perplexity: 9.6258
Epoch [1/3], Step [2374/3236], Loss: 2.4920, Perplexity: 12.0852
Epoch [1/3], Step [2375/3236], Loss: 2.4080, Perplexity: 11.1116
Epoch [1/3], Step [2376/3236], Loss: 2.2980, Perplexity: 9.9545
Epoch [1/3], Step [2377/3236], Loss: 2.5743, Perplexity: 13.1218
Epoch [1/3], Step [2378/3236], Loss: 2.4619, Perplexity: 11.7273
Epoch [1/3], Step [2379/3236], Loss: 2.2899, Perplexity: 9.8739
Epoch [1/3], Step [2380/3236], Loss: 2.4127, Perplexity: 11.1645
Epoch [1/3], Step [2381/3236], Loss: 2.2790, Perplexity: 9.7674
Epoch [1/3], Step [2382/3236], Loss: 2.3339, Perplexity: 10.3186
Epoch [1/3], Step [2383/3236], Loss: 2.6787, Perplexity: 14.5660
Epoch [1/3], Step [2384/3236], Loss: 2.6811, Perplexity: 14.6012
Epoch [1/3], Step [2385/3236], Loss: 2.6338, Perplexity: 13.9261
Epoch [1/3], Step [2386/3236], Loss: 2.3739, Perplexity: 10.7397
Epoch [1/3], Step [2387/3236], Loss: 2.2051, Perplexity: 9.0711
Epoch [1/3], Step [2388/3236], Loss: 2.5333, Perplexity: 12.5947
Epoch [1/3], Step [2389/3236], Loss: 2.3138, Perplexity: 10.1125
Epoch [1/3], Step [2390/3236], Loss: 2.2403, Perplexity: 9.3966
Epoch [1/3], Step [2391/3236], Loss: 2.4108, Perplexity: 11.1433
Epoch [1/3], Step [2392/3236], Loss: 2.3303, Perplexity: 10.2807
Epoch [1/3], Step [2393/3236], Loss: 2.2744, Perplexity: 9.7223
Epoch [1/3], Step [2394/3236], Loss: 2.5156, Perplexity: 12.3737
Epoch [1/3], Step [2395/3236], Loss: 2.3301, Perplexity: 10.2787
Epoch [1/3], Step [2396/3236], Loss: 2.5524, Perplexity: 12.8378
Epoch [1/3], Step [2397/3236], Loss: 2.5232, Perplexity: 12.4685
Epoch [1/3], Step [2398/3236], Loss: 2.3500, Perplexity: 10.4858
Epoch [1/3], Step [2399/3236], Loss: 2.3204, Perplexity: 10.1802
Epoch [1/3], Step [2400/3236], Loss: 2.3041, Perplexity: 10.0154
Epoch [1/3], Step [2401/3236], Loss: 2.5617, Perplexity: 12.9576
Epoch [1/3], Step [2402/3236], Loss: 2.3661, Perplexity: 10.6559
Epoch [1/3], Step [2403/3236], Loss: 2.3257, Perplexity: 10.2339
Epoch [1/3], Step [2404/3236], Loss: 2.6174, Perplexity: 13.6998
Epoch [1/3], Step [2405/3236], Loss: 2.5317, Perplexity: 12.5743
Epoch [1/3], Step [2406/3236], Loss: 2.2878, Perplexity: 9.8528
Epoch [1/3], Step [2407/3236], Loss: 2.3136, Perplexity: 10.1104
Epoch [1/3], Step [2408/3236], Loss: 2.1843, Perplexity: 8.8845
Epoch [1/3], Step [2409/3236], Loss: 2.3099, Perplexity: 10.0730
Epoch [1/3], Step [2410/3236], Loss: 2.1865, Perplexity: 8.9042
Epoch [1/3], Step [2411/3236], Loss: 2.1804, Perplexity: 8.8497
Epoch [1/3], Step [2412/3236], Loss: 2.4875, Perplexity: 12.0308
Epoch [1/3], Step [2413/3236], Loss: 2.3937, Perplexity: 10.9541
Epoch [1/3], Step [2414/3236], Loss: 2.1756, Perplexity: 8.8076
Epoch [1/3], Step [2415/3236], Loss: 2.3043, Perplexity: 10.0168
Epoch [1/3], Step [2416/3236], Loss: 2.3208, Perplexity: 10.1835
Epoch [1/3], Step [2417/3236], Loss: 2.1149, Perplexity: 8.2889
Epoch [1/3], Step [2418/3236], Loss: 2.4560, Perplexity: 11.6578
Epoch [1/3], Step [2419/3236], Loss: 2.2421, Perplexity: 9.4130
Epoch [1/3], Step [2420/3236], Loss: 2.2594, Perplexity: 9.5778
Epoch [1/3], Step [2421/3236], Loss: 2.1617, Perplexity: 8.6857
Epoch [1/3], Step [2422/3236], Loss: 2.2497, Perplexity: 9.4853
Epoch [1/3], Step [2423/3236], Loss: 2.6463, Perplexity: 14.1014
Epoch [1/3], Step [2424/3236], Loss: 2.1341, Perplexity: 8.4492
Epoch [1/3], Step [2425/3236], Loss: 2.3734, Perplexity: 10.7340
Epoch [1/3], Step [2426/3236], Loss: 2.5350, Perplexity: 12.6166
Epoch [1/3], Step [2427/3236], Loss: 2.1598, Perplexity: 8.6694
Epoch [1/3], Step [2428/3236], Loss: 2.4601, Perplexity: 11.7061
Epoch [1/3], Step [2429/3236], Loss: 2.5146, Perplexity: 12.3615
Epoch [1/3], Step [2430/3236], Loss: 2.1448, Perplexity: 8.5403
Epoch [1/3], Step [2431/3236], Loss: 2.1948, Perplexity: 8.9782
Epoch [1/3], Step [2432/3236], Loss: 2.3164, Perplexity: 10.1388
Epoch [1/3], Step [2433/3236], Loss: 2.3654, Perplexity: 10.6485
Epoch [1/3], Step [2434/3236], Loss: 2.3001, Perplexity: 9.9748
Epoch [1/3], Step [2435/3236], Loss: 2.7478, Perplexity: 15.6084
Epoch [1/3], Step [2436/3236], Loss: 3.3294, Perplexity: 27.9227
Epoch [1/3], Step [2437/3236], Loss: 2.2074, Perplexity: 9.0919
Epoch [1/3], Step [2438/3236], Loss: 2.5303, Perplexity: 12.5571
Epoch [1/3], Step [2439/3236], Loss: 2.3068, Perplexity: 10.0424
Epoch [1/3], Step [2440/3236], Loss: 2.6006, Perplexity: 13.4719
Epoch [1/3], Step [2441/3236], Loss: 2.3585, Perplexity: 10.5756
Epoch [1/3], Step [2442/3236], Loss: 2.1037, Perplexity: 8.1968
Epoch [1/3], Step [2443/3236], Loss: 2.3089, Perplexity: 10.0633
Epoch [1/3], Step [2444/3236], Loss: 2.1354, Perplexity: 8.4606
Epoch [1/3], Step [2445/3236], Loss: 2.2820, Perplexity: 9.7962
Epoch [1/3], Step [2446/3236], Loss: 2.4966, Perplexity: 12.1417
Epoch [1/3], Step [2447/3236], Loss: 2.2908, Perplexity: 9.8828
Epoch [1/3], Step [2448/3236], Loss: 2.2867, Perplexity: 9.8419
Epoch [1/3], Step [2449/3236], Loss: 2.4414, Perplexity: 11.4896
Epoch [1/3], Step [2450/3236], Loss: 2.3107, Perplexity: 10.0817
Epoch [1/3], Step [2451/3236], Loss: 2.1978, Perplexity: 9.0054
Epoch [1/3], Step [2452/3236], Loss: 2.2698, Perplexity: 9.6773
Epoch [1/3], Step [2453/3236], Loss: 2.2031, Perplexity: 9.0526
Epoch [1/3], Step [2454/3236], Loss: 2.2421, Perplexity: 9.4129
Epoch [1/3], Step [2455/3236], Loss: 2.5418, Perplexity: 12.7024
Epoch [1/3], Step [2456/3236], Loss: 2.3024, Perplexity: 9.9978
Epoch [1/3], Step [2457/3236], Loss: 2.3599, Perplexity: 10.5900
Epoch [1/3], Step [2458/3236], Loss: 2.3147, Perplexity: 10.1215
Epoch [1/3], Step [2459/3236], Loss: 2.4618, Perplexity: 11.7261
Epoch [1/3], Step [2460/3236], Loss: 2.3166, Perplexity: 10.1410
Epoch [1/3], Step [2461/3236], Loss: 2.2780, Perplexity: 9.7567
Epoch [1/3], Step [2462/3236], Loss: 2.2529, Perplexity: 9.5149
Epoch [1/3], Step [2463/3236], Loss: 2.2934, Perplexity: 9.9084
Epoch [1/3], Step [2464/3236], Loss: 2.3426, Perplexity: 10.4083
Epoch [1/3], Step [2465/3236], Loss: 2.4494, Perplexity: 11.5811
Epoch [1/3], Step [2466/3236], Loss: 2.6692, Perplexity: 14.4279
Epoch [1/3], Step [2467/3236], Loss: 2.5429, Perplexity: 12.7167
Epoch [1/3], Step [2468/3236], Loss: 2.2510, Perplexity: 9.4971
Epoch [1/3], Step [2469/3236], Loss: 2.2548, Perplexity: 9.5338
Epoch [1/3], Step [2470/3236], Loss: 2.5841, Perplexity: 13.2520
Epoch [1/3], Step [2471/3236], Loss: 2.2050, Perplexity: 9.0703
Epoch [1/3], Step [2472/3236], Loss: 2.1331, Perplexity: 8.4410
Epoch [1/3], Step [2473/3236], Loss: 2.2926, Perplexity: 9.9005
Epoch [1/3], Step [2474/3236], Loss: 2.2055, Perplexity: 9.0746
Epoch [1/3], Step [2475/3236], Loss: 2.2683, Perplexity: 9.6629
Epoch [1/3], Step [2476/3236], Loss: 2.3379, Perplexity: 10.3597
Epoch [1/3], Step [2477/3236], Loss: 2.3365, Perplexity: 10.3447
Epoch [1/3], Step [2478/3236], Loss: 2.2854, Perplexity: 9.8296
Epoch [1/3], Step [2479/3236], Loss: 2.2216, Perplexity: 9.2223
Epoch [1/3], Step [2480/3236], Loss: 2.2363, Perplexity: 9.3589
Epoch [1/3], Step [2481/3236], Loss: 2.3058, Perplexity: 10.0318
Epoch [1/3], Step [2482/3236], Loss: 2.3538, Perplexity: 10.5256
Epoch [1/3], Step [2483/3236], Loss: 2.1588, Perplexity: 8.6604
Epoch [1/3], Step [2484/3236], Loss: 2.3996, Perplexity: 11.0191
Epoch [1/3], Step [2485/3236], Loss: 2.2447, Perplexity: 9.4379
Epoch [1/3], Step [2486/3236], Loss: 2.3486, Perplexity: 10.4711
Epoch [1/3], Step [2487/3236], Loss: 2.4210, Perplexity: 11.2567
Epoch [1/3], Step [2488/3236], Loss: 2.2730, Perplexity: 9.7089
Epoch [1/3], Step [2489/3236], Loss: 2.3712, Perplexity: 10.7102
Epoch [1/3], Step [2490/3236], Loss: 2.2895, Perplexity: 9.8701
Epoch [1/3], Step [2491/3236], Loss: 2.2669, Perplexity: 9.6491
Epoch [1/3], Step [2492/3236], Loss: 2.1933, Perplexity: 8.9645
Epoch [1/3], Step [2493/3236], Loss: 2.2416, Perplexity: 9.4083
Epoch [1/3], Step [2494/3236], Loss: 2.3940, Perplexity: 10.9577
Epoch [1/3], Step [2495/3236], Loss: 2.5154, Perplexity: 12.3716
Epoch [1/3], Step [2496/3236], Loss: 2.9753, Perplexity: 19.5962
Epoch [1/3], Step [2497/3236], Loss: 2.2731, Perplexity: 9.7097
Epoch [1/3], Step [2498/3236], Loss: 2.1233, Perplexity: 8.3589
Epoch [1/3], Step [2499/3236], Loss: 2.4568, Perplexity: 11.6672
Epoch [1/3], Step [2500/3236], Loss: 2.4354, Perplexity: 11.4208
Epoch [1/3], Step [2501/3236], Loss: 2.3063, Perplexity: 10.0374
Epoch [1/3], Step [2502/3236], Loss: 2.4382, Perplexity: 11.4520
Epoch [1/3], Step [2503/3236], Loss: 2.2769, Perplexity: 9.7461
Epoch [1/3], Step [2504/3236], Loss: 2.4771, Perplexity: 11.9069
Epoch [1/3], Step [2505/3236], Loss: 3.0377, Perplexity: 20.8574
Epoch [1/3], Step [2506/3236], Loss: 2.6593, Perplexity: 14.2869
Epoch [1/3], Step [2507/3236], Loss: 2.3617, Perplexity: 10.6091
Epoch [1/3], Step [2508/3236], Loss: 2.2768, Perplexity: 9.7455
Epoch [1/3], Step [2509/3236], Loss: 2.2944, Perplexity: 9.9185
Epoch [1/3], Step [2510/3236], Loss: 2.7209, Perplexity: 15.1945
Epoch [1/3], Step [2511/3236], Loss: 2.3634, Perplexity: 10.6272
Epoch [1/3], Step [2512/3236], Loss: 2.3858, Perplexity: 10.8679
Epoch [1/3], Step [2513/3236], Loss: 2.4368, Perplexity: 11.4367
Epoch [1/3], Step [2514/3236], Loss: 2.1286, Perplexity: 8.4030
Epoch [1/3], Step [2515/3236], Loss: 2.3743, Perplexity: 10.7436
Epoch [1/3], Step [2516/3236], Loss: 2.1258, Perplexity: 8.3797
Epoch [1/3], Step [2517/3236], Loss: 2.3677, Perplexity: 10.6725
Epoch [1/3], Step [2518/3236], Loss: 2.2549, Perplexity: 9.5344
Epoch [1/3], Step [2519/3236], Loss: 2.5935, Perplexity: 13.3767
Epoch [1/3], Step [2520/3236], Loss: 2.6939, Perplexity: 14.7890
Epoch [1/3], Step [2521/3236], Loss: 2.2812, Perplexity: 9.7884
Epoch [1/3], Step [2522/3236], Loss: 2.3153, Perplexity: 10.1280
Epoch [1/3], Step [2523/3236], Loss: 2.4327, Perplexity: 11.3892
Epoch [1/3], Step [2524/3236], Loss: 2.1028, Perplexity: 8.1892
Epoch [1/3], Step [2525/3236], Loss: 2.2009, Perplexity: 9.0329
Epoch [1/3], Step [2526/3236], Loss: 2.1302, Perplexity: 8.4167
Epoch [1/3], Step [2527/3236], Loss: 2.3124, Perplexity: 10.0988
Epoch [1/3], Step [2528/3236], Loss: 2.2560, Perplexity: 9.5447
Epoch [1/3], Step [2529/3236], Loss: 2.5120, Perplexity: 12.3290
Epoch [1/3], Step [2530/3236], Loss: 2.2793, Perplexity: 9.7701
Epoch [1/3], Step [2531/3236], Loss: 2.2601, Perplexity: 9.5845
Epoch [1/3], Step [2532/3236], Loss: 2.2351, Perplexity: 9.3476
Epoch [1/3], Step [2533/3236], Loss: 2.4593, Perplexity: 11.6972
Epoch [1/3], Step [2534/3236], Loss: 2.5540, Perplexity: 12.8590
Epoch [1/3], Step [2535/3236], Loss: 2.9401, Perplexity: 18.9184
Epoch [1/3], Step [2536/3236], Loss: 2.2672, Perplexity: 9.6525
Epoch [1/3], Step [2537/3236], Loss: 3.3551, Perplexity: 28.6493
Epoch [1/3], Step [2538/3236], Loss: 2.3910, Perplexity: 10.9241
Epoch [1/3], Step [2539/3236], Loss: 2.4549, Perplexity: 11.6448
Epoch [1/3], Step [2540/3236], Loss: 2.1521, Perplexity: 8.6030
Epoch [1/3], Step [2541/3236], Loss: 2.3854, Perplexity: 10.8639
Epoch [1/3], Step [2542/3236], Loss: 2.2803, Perplexity: 9.7795
Epoch [1/3], Step [2543/3236], Loss: 2.1911, Perplexity: 8.9451
Epoch [1/3], Step [2544/3236], Loss: 2.3459, Perplexity: 10.4425
Epoch [1/3], Step [2545/3236], Loss: 2.4589, Perplexity: 11.6924
Epoch [1/3], Step [2546/3236], Loss: 2.3127, Perplexity: 10.1018
Epoch [1/3], Step [2547/3236], Loss: 2.2269, Perplexity: 9.2710
Epoch [1/3], Step [2548/3236], Loss: 2.1654, Perplexity: 8.7181
Epoch [1/3], Step [2549/3236], Loss: 2.3590, Perplexity: 10.5805
Epoch [1/3], Step [2550/3236], Loss: 2.3601, Perplexity: 10.5922
Epoch [1/3], Step [2551/3236], Loss: 2.2358, Perplexity: 9.3538
Epoch [1/3], Step [2552/3236], Loss: 2.4219, Perplexity: 11.2670
Epoch [1/3], Step [2553/3236], Loss: 2.7407, Perplexity: 15.4977
Epoch [1/3], Step [2554/3236], Loss: 2.3588, Perplexity: 10.5780
Epoch [1/3], Step [2555/3236], Loss: 2.2147, Perplexity: 9.1588
Epoch [1/3], Step [2556/3236], Loss: 2.1898, Perplexity: 8.9335
Epoch [1/3], Step [2557/3236], Loss: 2.2939, Perplexity: 9.9133
Epoch [1/3], Step [2558/3236], Loss: 2.2408, Perplexity: 9.4010
Epoch [1/3], Step [2559/3236], Loss: 2.4268, Perplexity: 11.3222
Epoch [1/3], Step [2560/3236], Loss: 2.7444, Perplexity: 15.5557
Epoch [1/3], Step [2561/3236], Loss: 2.3634, Perplexity: 10.6268
Epoch [1/3], Step [2562/3236], Loss: 2.3286, Perplexity: 10.2633
Epoch [1/3], Step [2563/3236], Loss: 2.2385, Perplexity: 9.3788
Epoch [1/3], Step [2564/3236], Loss: 2.1099, Perplexity: 8.2473
Epoch [1/3], Step [2565/3236], Loss: 2.1617, Perplexity: 8.6856
Epoch [1/3], Step [2566/3236], Loss: 2.1276, Perplexity: 8.3946
Epoch [1/3], Step [2567/3236], Loss: 2.3769, Perplexity: 10.7710
Epoch [1/3], Step [2568/3236], Loss: 2.2702, Perplexity: 9.6817
Epoch [1/3], Step [2569/3236], Loss: 2.6522, Perplexity: 14.1852
Epoch [1/3], Step [2570/3236], Loss: 2.2350, Perplexity: 9.3466
Epoch [1/3], Step [2571/3236], Loss: 2.3631, Perplexity: 10.6237
Epoch [1/3], Step [2572/3236], Loss: 2.2416, Perplexity: 9.4087
Epoch [1/3], Step [2573/3236], Loss: 2.2140, Perplexity: 9.1526
Epoch [1/3], Step [2574/3236], Loss: 2.1717, Perplexity: 8.7736
Epoch [1/3], Step [2575/3236], Loss: 2.3676, Perplexity: 10.6719
Epoch [1/3], Step [2576/3236], Loss: 2.2405, Perplexity: 9.3979
Epoch [1/3], Step [2577/3236], Loss: 2.2406, Perplexity: 9.3994
Epoch [1/3], Step [2578/3236], Loss: 2.4188, Perplexity: 11.2328
Epoch [1/3], Step [2579/3236], Loss: 2.2565, Perplexity: 9.5500
Epoch [1/3], Step [2580/3236], Loss: 2.9102, Perplexity: 18.3598
Epoch [1/3], Step [2581/3236], Loss: 3.0939, Perplexity: 22.0631
Epoch [1/3], Step [2582/3236], Loss: 2.3003, Perplexity: 9.9776
Epoch [1/3], Step [2583/3236], Loss: 2.0637, Perplexity: 7.8747
Epoch [1/3], Step [2584/3236], Loss: 2.3602, Perplexity: 10.5926
Epoch [1/3], Step [2585/3236], Loss: 2.3640, Perplexity: 10.6329
Epoch [1/3], Step [2586/3236], Loss: 2.4642, Perplexity: 11.7540
Epoch [1/3], Step [2587/3236], Loss: 2.3984, Perplexity: 11.0052
Epoch [1/3], Step [2588/3236], Loss: 2.4220, Perplexity: 11.2681
Epoch [1/3], Step [2589/3236], Loss: 2.4950, Perplexity: 12.1213
Epoch [1/3], Step [2590/3236], Loss: 2.1798, Perplexity: 8.8447
Epoch [1/3], Step [2591/3236], Loss: 2.5644, Perplexity: 12.9932
Epoch [1/3], Step [2592/3236], Loss: 2.3003, Perplexity: 9.9767
Epoch [1/3], Step [2593/3236], Loss: 2.3301, Perplexity: 10.2794
Epoch [1/3], Step [2594/3236], Loss: 2.4734, Perplexity: 11.8630
Epoch [1/3], Step [2595/3236], Loss: 2.6449, Perplexity: 14.0814
Epoch [1/3], Step [2596/3236], Loss: 2.5692, Perplexity: 13.0551
Epoch [1/3], Step [2597/3236], Loss: 2.3041, Perplexity: 10.0152
Epoch [1/3], Step [2598/3236], Loss: 2.2781, Perplexity: 9.7581
Epoch [1/3], Step [2599/3236], Loss: 2.3243, Perplexity: 10.2193
Epoch [1/3], Step [2600/3236], Loss: 2.7327, Perplexity: 15.3747
Epoch [1/3], Step [2601/3236], Loss: 2.2522, Perplexity: 9.5090
Epoch [1/3], Step [2602/3236], Loss: 2.5709, Perplexity: 13.0781
Epoch [1/3], Step [2603/3236], Loss: 2.4748, Perplexity: 11.8789
Epoch [1/3], Step [2604/3236], Loss: 2.2550, Perplexity: 9.5352
Epoch [1/3], Step [2605/3236], Loss: 2.3063, Perplexity: 10.0370
Epoch [1/3], Step [2606/3236], Loss: 2.1290, Perplexity: 8.4068
Epoch [1/3], Step [2607/3236], Loss: 2.1723, Perplexity: 8.7785
Epoch [1/3], Step [2608/3236], Loss: 2.2772, Perplexity: 9.7498
Epoch [1/3], Step [2609/3236], Loss: 2.2462, Perplexity: 9.4519
Epoch [1/3], Step [2610/3236], Loss: 2.5384, Perplexity: 12.6600
Epoch [1/3], Step [2611/3236], Loss: 2.2968, Perplexity: 9.9427
Epoch [1/3], Step [2612/3236], Loss: 2.2660, Perplexity: 9.6403
Epoch [1/3], Step [2613/3236], Loss: 2.2138, Perplexity: 9.1507
Epoch [1/3], Step [2614/3236], Loss: 2.2074, Perplexity: 9.0922
Epoch [1/3], Step [2615/3236], Loss: 2.2537, Perplexity: 9.5233
Epoch [1/3], Step [2616/3236], Loss: 2.4140, Perplexity: 11.1781
Epoch [1/3], Step [2617/3236], Loss: 2.2280, Perplexity: 9.2816
Epoch [1/3], Step [2618/3236], Loss: 2.3788, Perplexity: 10.7921
Epoch [1/3], Step [2619/3236], Loss: 2.3290, Perplexity: 10.2677
Epoch [1/3], Step [2620/3236], Loss: 2.7883, Perplexity: 16.2535
Epoch [1/3], Step [2621/3236], Loss: 2.2390, Perplexity: 9.3837
Epoch [1/3], Step [2622/3236], Loss: 2.9263, Perplexity: 18.6583
Epoch [1/3], Step [2623/3236], Loss: 2.1353, Perplexity: 8.4597
Epoch [1/3], Step [2624/3236], Loss: 2.2322, Perplexity: 9.3202
Epoch [1/3], Step [2625/3236], Loss: 2.6962, Perplexity: 14.8232
Epoch [1/3], Step [2626/3236], Loss: 2.6015, Perplexity: 13.4840
Epoch [1/3], Step [2627/3236], Loss: 2.2498, Perplexity: 9.4860
Epoch [1/3], Step [2628/3236], Loss: 2.4376, Perplexity: 11.4451
Epoch [1/3], Step [2629/3236], Loss: 2.2707, Perplexity: 9.6862
Epoch [1/3], Step [2630/3236], Loss: 2.3526, Perplexity: 10.5129
Epoch [1/3], Step [2631/3236], Loss: 2.2083, Perplexity: 9.1000
Epoch [1/3], Step [2632/3236], Loss: 2.4203, Perplexity: 11.2498
Epoch [1/3], Step [2633/3236], Loss: 2.2680, Perplexity: 9.6601
Epoch [1/3], Step [2634/3236], Loss: 2.2502, Perplexity: 9.4892
Epoch [1/3], Step [2635/3236], Loss: 2.2264, Perplexity: 9.2660
Epoch [1/3], Step [2636/3236], Loss: 2.4083, Perplexity: 11.1151
Epoch [1/3], Step [2637/3236], Loss: 2.5538, Perplexity: 12.8553
Epoch [1/3], Step [2638/3236], Loss: 2.9156, Perplexity: 18.4594
Epoch [1/3], Step [2639/3236], Loss: 2.3425, Perplexity: 10.4072
Epoch [1/3], Step [2640/3236], Loss: 2.6028, Perplexity: 13.5013
Epoch [1/3], Step [2641/3236], Loss: 2.4241, Perplexity: 11.2919
Epoch [1/3], Step [2642/3236], Loss: 2.3049, Perplexity: 10.0234
Epoch [1/3], Step [2643/3236], Loss: 2.3809, Perplexity: 10.8150
Epoch [1/3], Step [2644/3236], Loss: 2.1859, Perplexity: 8.8985
Epoch [1/3], Step [2645/3236], Loss: 2.1698, Perplexity: 8.7563
Epoch [1/3], Step [2646/3236], Loss: 2.2726, Perplexity: 9.7042
Epoch [1/3], Step [2647/3236], Loss: 2.3861, Perplexity: 10.8708
Epoch [1/3], Step [2648/3236], Loss: 2.0883, Perplexity: 8.0713
Epoch [1/3], Step [2649/3236], Loss: 2.4306, Perplexity: 11.3656
Epoch [1/3], Step [2650/3236], Loss: 2.3345, Perplexity: 10.3246
Epoch [1/3], Step [2651/3236], Loss: 2.2631, Perplexity: 9.6126
Epoch [1/3], Step [2652/3236], Loss: 2.2253, Perplexity: 9.2565
Epoch [1/3], Step [2653/3236], Loss: 2.3099, Perplexity: 10.0734
Epoch [1/3], Step [2654/3236], Loss: 2.3329, Perplexity: 10.3080
Epoch [1/3], Step [2655/3236], Loss: 2.3404, Perplexity: 10.3852
Epoch [1/3], Step [2656/3236], Loss: 2.2466, Perplexity: 9.4557
Epoch [1/3], Step [2657/3236], Loss: 2.2090, Perplexity: 9.1062
Epoch [1/3], Step [2658/3236], Loss: 2.2152, Perplexity: 9.1633
Epoch [1/3], Step [2659/3236], Loss: 2.3116, Perplexity: 10.0909
Epoch [1/3], Step [2660/3236], Loss: 2.2638, Perplexity: 9.6195
Epoch [1/3], Step [2661/3236], Loss: 2.3812, Perplexity: 10.8174
Epoch [1/3], Step [2662/3236], Loss: 2.3086, Perplexity: 10.0604
Epoch [1/3], Step [2663/3236], Loss: 2.3374, Perplexity: 10.3546
Epoch [1/3], Step [2664/3236], Loss: 2.3194, Perplexity: 10.1694
Epoch [1/3], Step [2665/3236], Loss: 2.3476, Perplexity: 10.4605
Epoch [1/3], Step [2666/3236], Loss: 2.2684, Perplexity: 9.6642
Epoch [1/3], Step [2667/3236], Loss: 2.3047, Perplexity: 10.0208
Epoch [1/3], Step [2668/3236], Loss: 2.1437, Perplexity: 8.5308
Epoch [1/3], Step [2669/3236], Loss: 2.7004, Perplexity: 14.8863
Epoch [1/3], Step [2670/3236], Loss: 2.3991, Perplexity: 11.0128
Epoch [1/3], Step [2671/3236], Loss: 2.3058, Perplexity: 10.0320
Epoch [1/3], Step [2672/3236], Loss: 2.2144, Perplexity: 9.1563
Epoch [1/3], Step [2673/3236], Loss: 2.3093, Perplexity: 10.0676
Epoch [1/3], Step [2674/3236], Loss: 2.2695, Perplexity: 9.6742
Epoch [1/3], Step [2675/3236], Loss: 2.2067, Perplexity: 9.0855
Epoch [1/3], Step [2676/3236], Loss: 2.1868, Perplexity: 8.9071
Epoch [1/3], Step [2677/3236], Loss: 2.1664, Perplexity: 8.7271
Epoch [1/3], Step [2678/3236], Loss: 2.2642, Perplexity: 9.6231
Epoch [1/3], Step [2679/3236], Loss: 2.6576, Perplexity: 14.2627
Epoch [1/3], Step [2680/3236], Loss: 2.2406, Perplexity: 9.3986
Epoch [1/3], Step [2681/3236], Loss: 2.2429, Perplexity: 9.4204
Epoch [1/3], Step [2682/3236], Loss: 2.1775, Perplexity: 8.8244
Epoch [1/3], Step [2683/3236], Loss: 2.1017, Perplexity: 8.1802
Epoch [1/3], Step [2684/3236], Loss: 2.2456, Perplexity: 9.4461
Epoch [1/3], Step [2685/3236], Loss: 2.3329, Perplexity: 10.3075
Epoch [1/3], Step [2686/3236], Loss: 2.3132, Perplexity: 10.1064
Epoch [1/3], Step [2687/3236], Loss: 3.0760, Perplexity: 21.6720
Epoch [1/3], Step [2688/3236], Loss: 2.3340, Perplexity: 10.3196
Epoch [1/3], Step [2689/3236], Loss: 2.3562, Perplexity: 10.5512
Epoch [1/3], Step [2690/3236], Loss: 2.3492, Perplexity: 10.4768
Epoch [1/3], Step [2691/3236], Loss: 2.2043, Perplexity: 9.0636
Epoch [1/3], Step [2692/3236], Loss: 2.1183, Perplexity: 8.3174
Epoch [1/3], Step [2693/3236], Loss: 2.3754, Perplexity: 10.7556
Epoch [1/3], Step [2694/3236], Loss: 2.6320, Perplexity: 13.9012
Epoch [1/3], Step [2695/3236], Loss: 2.1177, Perplexity: 8.3118
Epoch [1/3], Step [2696/3236], Loss: 2.1983, Perplexity: 9.0100
Epoch [1/3], Step [2697/3236], Loss: 2.2182, Perplexity: 9.1907
Epoch [1/3], Step [2698/3236], Loss: 2.2343, Perplexity: 9.3402
Epoch [1/3], Step [2699/3236], Loss: 2.1282, Perplexity: 8.3996
Epoch [1/3], Step [2700/3236], Loss: 2.2433, Perplexity: 9.4248
Epoch [1/3], Step [2701/3236], Loss: 2.2745, Perplexity: 9.7233
Epoch [1/3], Step [2702/3236], Loss: 2.4339, Perplexity: 11.4034
Epoch [1/3], Step [2703/3236], Loss: 2.2139, Perplexity: 9.1516
Epoch [1/3], Step [2704/3236], Loss: 2.3241, Perplexity: 10.2179
Epoch [1/3], Step [2705/3236], Loss: 2.5224, Perplexity: 12.4579
Epoch [1/3], Step [2706/3236], Loss: 2.3037, Perplexity: 10.0109
Epoch [1/3], Step [2707/3236], Loss: 2.2316, Perplexity: 9.3150
Epoch [1/3], Step [2708/3236], Loss: 2.3183, Perplexity: 10.1585
Epoch [1/3], Step [2709/3236], Loss: 2.2280, Perplexity: 9.2814
Epoch [1/3], Step [2710/3236], Loss: 2.1159, Perplexity: 8.2968
Epoch [1/3], Step [2711/3236], Loss: 2.3213, Perplexity: 10.1886
Epoch [1/3], Step [2712/3236], Loss: 2.1903, Perplexity: 8.9378
Epoch [1/3], Step [2713/3236], Loss: 2.3863, Perplexity: 10.8735
Epoch [1/3], Step [2714/3236], Loss: 2.1657, Perplexity: 8.7209
Epoch [1/3], Step [2715/3236], Loss: 2.3127, Perplexity: 10.1021
Epoch [1/3], Step [2716/3236], Loss: 2.3240, Perplexity: 10.2165
Epoch [1/3], Step [2717/3236], Loss: 2.2398, Perplexity: 9.3912
Epoch [1/3], Step [2718/3236], Loss: 2.2353, Perplexity: 9.3491
Epoch [1/3], Step [2719/3236], Loss: 2.3358, Perplexity: 10.3377
Epoch [1/3], Step [2720/3236], Loss: 2.4827, Perplexity: 11.9731
Epoch [1/3], Step [2721/3236], Loss: 2.7355, Perplexity: 15.4172
Epoch [1/3], Step [2722/3236], Loss: 2.5936, Perplexity: 13.3772
Epoch [1/3], Step [2723/3236], Loss: 2.1380, Perplexity: 8.4827
Epoch [1/3], Step [2724/3236], Loss: 2.3137, Perplexity: 10.1121
Epoch [1/3], Step [2725/3236], Loss: 2.6434, Perplexity: 14.0611
Epoch [1/3], Step [2726/3236], Loss: 2.2622, Perplexity: 9.6043
Epoch [1/3], Step [2727/3236], Loss: 2.1052, Perplexity: 8.2091
Epoch [1/3], Step [2728/3236], Loss: 2.2140, Perplexity: 9.1518
Epoch [1/3], Step [2729/3236], Loss: 2.5772, Perplexity: 13.1605
Epoch [1/3], Step [2730/3236], Loss: 2.2958, Perplexity: 9.9323
Epoch [1/3], Step [2731/3236], Loss: 2.3095, Perplexity: 10.0692
Epoch [1/3], Step [2732/3236], Loss: 2.2251, Perplexity: 9.2548
Epoch [1/3], Step [2733/3236], Loss: 2.2047, Perplexity: 9.0673
Epoch [1/3], Step [2734/3236], Loss: 2.5259, Perplexity: 12.5020
Epoch [1/3], Step [2735/3236], Loss: 2.3783, Perplexity: 10.7861
Epoch [1/3], Step [2736/3236], Loss: 2.4252, Perplexity: 11.3047
Epoch [1/3], Step [2737/3236], Loss: 2.1332, Perplexity: 8.4416
Epoch [1/3], Step [2738/3236], Loss: 2.4873, Perplexity: 12.0293
Epoch [1/3], Step [2739/3236], Loss: 2.9668, Perplexity: 19.4303
Epoch [1/3], Step [2740/3236], Loss: 2.1085, Perplexity: 8.2356
Epoch [1/3], Step [2741/3236], Loss: 2.1114, Perplexity: 8.2600
Epoch [1/3], Step [2742/3236], Loss: 2.4958, Perplexity: 12.1318
Epoch [1/3], Step [2743/3236], Loss: 2.5586, Perplexity: 12.9179
Epoch [1/3], Step [2744/3236], Loss: 2.6577, Perplexity: 14.2637
Epoch [1/3], Step [2745/3236], Loss: 2.1851, Perplexity: 8.8911
Epoch [1/3], Step [2746/3236], Loss: 2.5819, Perplexity: 13.2227
Epoch [1/3], Step [2747/3236], Loss: 2.7221, Perplexity: 15.2118
Epoch [1/3], Step [2748/3236], Loss: 2.3049, Perplexity: 10.0234
Epoch [1/3], Step [2749/3236], Loss: 2.3684, Perplexity: 10.6803
Epoch [1/3], Step [2750/3236], Loss: 2.4651, Perplexity: 11.7643
Epoch [1/3], Step [2751/3236], Loss: 2.1681, Perplexity: 8.7416
Epoch [1/3], Step [2752/3236], Loss: 2.3256, Perplexity: 10.2324
Epoch [1/3], Step [2753/3236], Loss: 2.2655, Perplexity: 9.6362
Epoch [1/3], Step [2754/3236], Loss: 2.6433, Perplexity: 14.0601
Epoch [1/3], Step [2755/3236], Loss: 2.6665, Perplexity: 14.3890
Epoch [1/3], Step [2756/3236], Loss: 2.1825, Perplexity: 8.8682
Epoch [1/3], Step [2757/3236], Loss: 2.1573, Perplexity: 8.6479
Epoch [1/3], Step [2758/3236], Loss: 2.4750, Perplexity: 11.8820
Epoch [1/3], Step [2759/3236], Loss: 2.1841, Perplexity: 8.8830
Epoch [1/3], Step [2760/3236], Loss: 2.1100, Perplexity: 8.2481
Epoch [1/3], Step [2761/3236], Loss: 2.3014, Perplexity: 9.9882
Epoch [1/3], Step [2762/3236], Loss: 2.4416, Perplexity: 11.4917
Epoch [1/3], Step [2763/3236], Loss: 2.4927, Perplexity: 12.0938
Epoch [1/3], Step [2764/3236], Loss: 2.5657, Perplexity: 13.0098
Epoch [1/3], Step [2765/3236], Loss: 2.2985, Perplexity: 9.9596
Epoch [1/3], Step [2766/3236], Loss: 2.2946, Perplexity: 9.9207
Epoch [1/3], Step [2767/3236], Loss: 2.2379, Perplexity: 9.3734
Epoch [1/3], Step [2768/3236], Loss: 2.3848, Perplexity: 10.8566
Epoch [1/3], Step [2769/3236], Loss: 3.2690, Perplexity: 26.2839
Epoch [1/3], Step [2770/3236], Loss: 2.1842, Perplexity: 8.8836
Epoch [1/3], Step [2771/3236], Loss: 2.2193, Perplexity: 9.2013
Epoch [1/3], Step [2772/3236], Loss: 2.3692, Perplexity: 10.6889
Epoch [1/3], Step [2773/3236], Loss: 2.2272, Perplexity: 9.2737
Epoch [1/3], Step [2774/3236], Loss: 2.2790, Perplexity: 9.7668
Epoch [1/3], Step [2775/3236], Loss: 2.4623, Perplexity: 11.7312
Epoch [1/3], Step [2776/3236], Loss: 2.2691, Perplexity: 9.6702
Epoch [1/3], Step [2777/3236], Loss: 2.8978, Perplexity: 18.1338
Epoch [1/3], Step [2778/3236], Loss: 2.1878, Perplexity: 8.9154
Epoch [1/3], Step [2779/3236], Loss: 2.8255, Perplexity: 16.8695
Epoch [1/3], Step [2780/3236], Loss: 2.3519, Perplexity: 10.5054
Epoch [1/3], Step [2781/3236], Loss: 2.5174, Perplexity: 12.3965
Epoch [1/3], Step [2782/3236], Loss: 2.6748, Perplexity: 14.5095
Epoch [1/3], Step [2783/3236], Loss: 2.3435, Perplexity: 10.4180
Epoch [1/3], Step [2784/3236], Loss: 2.3312, Perplexity: 10.2900
Epoch [1/3], Step [2785/3236], Loss: 2.6015, Perplexity: 13.4844
Epoch [1/3], Step [2786/3236], Loss: 2.3745, Perplexity: 10.7456
Epoch [1/3], Step [2787/3236], Loss: 2.2389, Perplexity: 9.3832
Epoch [1/3], Step [2788/3236], Loss: 2.2667, Perplexity: 9.6475
Epoch [1/3], Step [2789/3236], Loss: 2.4668, Perplexity: 11.7848
Epoch [1/3], Step [2790/3236], Loss: 2.3081, Perplexity: 10.0556
Epoch [1/3], Step [2791/3236], Loss: 2.3685, Perplexity: 10.6814
Epoch [1/3], Step [2792/3236], Loss: 2.3133, Perplexity: 10.1080
Epoch [1/3], Step [2793/3236], Loss: 2.2678, Perplexity: 9.6582
Epoch [1/3], Step [2794/3236], Loss: 2.2657, Perplexity: 9.6378
Epoch [1/3], Step [2795/3236], Loss: 2.2963, Perplexity: 9.9377
Epoch [1/3], Step [2796/3236], Loss: 2.2291, Perplexity: 9.2913
Epoch [1/3], Step [2797/3236], Loss: 2.3439, Perplexity: 10.4214
Epoch [1/3], Step [2798/3236], Loss: 2.2419, Perplexity: 9.4116
Epoch [1/3], Step [2799/3236], Loss: 2.2935, Perplexity: 9.9097
Epoch [1/3], Step [2800/3236], Loss: 2.4214, Perplexity: 11.2616
Epoch [1/3], Step [2801/3236], Loss: 2.5117, Perplexity: 12.3256
Epoch [1/3], Step [2802/3236], Loss: 2.2364, Perplexity: 9.3598
Epoch [1/3], Step [2803/3236], Loss: 2.2815, Perplexity: 9.7912
Epoch [1/3], Step [2804/3236], Loss: 2.6861, Perplexity: 14.6737
Epoch [1/3], Step [2805/3236], Loss: 2.3990, Perplexity: 11.0127
Epoch [1/3], Step [2806/3236], Loss: 2.5175, Perplexity: 12.3979
Epoch [1/3], Step [2807/3236], Loss: 2.1854, Perplexity: 8.8944
Epoch [1/3], Step [2808/3236], Loss: 2.3484, Perplexity: 10.4688
Epoch [1/3], Step [2809/3236], Loss: 2.1922, Perplexity: 8.9548
Epoch [1/3], Step [2810/3236], Loss: 2.2754, Perplexity: 9.7322
Epoch [1/3], Step [2811/3236], Loss: 2.3047, Perplexity: 10.0217
Epoch [1/3], Step [2812/3236], Loss: 2.7449, Perplexity: 15.5626
Epoch [1/3], Step [2813/3236], Loss: 2.2957, Perplexity: 9.9317
Epoch [1/3], Step [2814/3236], Loss: 2.3064, Perplexity: 10.0386
Epoch [1/3], Step [2815/3236], Loss: 2.7574, Perplexity: 15.7591
Epoch [1/3], Step [2816/3236], Loss: 2.1373, Perplexity: 8.4761
Epoch [1/3], Step [2817/3236], Loss: 2.2853, Perplexity: 9.8284
Epoch [1/3], Step [2818/3236], Loss: 2.2640, Perplexity: 9.6218
Epoch [1/3], Step [2819/3236], Loss: 2.1970, Perplexity: 8.9980
Epoch [1/3], Step [2820/3236], Loss: 2.3275, Perplexity: 10.2524
Epoch [1/3], Step [2821/3236], Loss: 2.2698, Perplexity: 9.6778
Epoch [1/3], Step [2822/3236], Loss: 2.2544, Perplexity: 9.5294
Epoch [1/3], Step [2823/3236], Loss: 2.1482, Perplexity: 8.5691
Epoch [1/3], Step [2824/3236], Loss: 2.1652, Perplexity: 8.7164
Epoch [1/3], Step [2825/3236], Loss: 2.2164, Perplexity: 9.1740
Epoch [1/3], Step [2826/3236], Loss: 2.2978, Perplexity: 9.9522
Epoch [1/3], Step [2827/3236], Loss: 2.3828, Perplexity: 10.8349
Epoch [1/3], Step [2828/3236], Loss: 2.3224, Perplexity: 10.2004
Epoch [1/3], Step [2829/3236], Loss: 2.1661, Perplexity: 8.7238
Epoch [1/3], Step [2830/3236], Loss: 2.2844, Perplexity: 9.8198
Epoch [1/3], Step [2831/3236], Loss: 2.4928, Perplexity: 12.0951
Epoch [1/3], Step [2832/3236], Loss: 2.3108, Perplexity: 10.0828
Epoch [1/3], Step [2833/3236], Loss: 2.2536, Perplexity: 9.5220
Epoch [1/3], Step [2834/3236], Loss: 2.2631, Perplexity: 9.6131
Epoch [1/3], Step [2835/3236], Loss: 2.1632, Perplexity: 8.6991
Epoch [1/3], Step [2836/3236], Loss: 2.2804, Perplexity: 9.7803
Epoch [1/3], Step [2837/3236], Loss: 2.3634, Perplexity: 10.6275
Epoch [1/3], Step [2838/3236], Loss: 2.3047, Perplexity: 10.0212
Epoch [1/3], Step [2839/3236], Loss: 2.2345, Perplexity: 9.3422
Epoch [1/3], Step [2840/3236], Loss: 2.2073, Perplexity: 9.0910
Epoch [1/3], Step [2841/3236], Loss: 2.2991, Perplexity: 9.9653
Epoch [1/3], Step [2842/3236], Loss: 2.2430, Perplexity: 9.4212
Epoch [1/3], Step [2843/3236], Loss: 2.3037, Perplexity: 10.0110
Epoch [1/3], Step [2844/3236], Loss: 2.4569, Perplexity: 11.6680
Epoch [1/3], Step [2845/3236], Loss: 2.4494, Perplexity: 11.5811
Epoch [1/3], Step [2846/3236], Loss: 2.2778, Perplexity: 9.7549
Epoch [1/3], Step [2847/3236], Loss: 2.2824, Perplexity: 9.7997
Epoch [1/3], Step [2848/3236], Loss: 3.4695, Perplexity: 32.1192
Epoch [1/3], Step [2849/3236], Loss: 2.2604, Perplexity: 9.5866
Epoch [1/3], Step [2850/3236], Loss: 2.2069, Perplexity: 9.0879
Epoch [1/3], Step [2851/3236], Loss: 2.1789, Perplexity: 8.8364
Epoch [1/3], Step [2852/3236], Loss: 2.6191, Perplexity: 13.7234
Epoch [1/3], Step [2853/3236], Loss: 2.2046, Perplexity: 9.0665
Epoch [1/3], Step [2854/3236], Loss: 2.4448, Perplexity: 11.5285
Epoch [1/3], Step [2855/3236], Loss: 2.1649, Perplexity: 8.7135
Epoch [1/3], Step [2856/3236], Loss: 2.3002, Perplexity: 9.9761
Epoch [1/3], Step [2857/3236], Loss: 2.4854, Perplexity: 12.0054
Epoch [1/3], Step [2858/3236], Loss: 2.5862, Perplexity: 13.2791
Epoch [1/3], Step [2859/3236], Loss: 2.4173, Perplexity: 11.2160
Epoch [1/3], Step [2860/3236], Loss: 2.1666, Perplexity: 8.7290
Epoch [1/3], Step [2861/3236], Loss: 2.5472, Perplexity: 12.7715
Epoch [1/3], Step [2862/3236], Loss: 2.3236, Perplexity: 10.2126
Epoch [1/3], Step [2863/3236], Loss: 2.6199, Perplexity: 13.7349
Epoch [1/3], Step [2864/3236], Loss: 2.2066, Perplexity: 9.0852
Epoch [1/3], Step [2865/3236], Loss: 2.3315, Perplexity: 10.2935
Epoch [1/3], Step [2866/3236], Loss: 2.1493, Perplexity: 8.5792
Epoch [1/3], Step [2867/3236], Loss: 2.2570, Perplexity: 9.5541
Epoch [1/3], Step [2868/3236], Loss: 2.3093, Perplexity: 10.0676
Epoch [1/3], Step [2869/3236], Loss: 2.1524, Perplexity: 8.6059
Epoch [1/3], Step [2870/3236], Loss: 2.2226, Perplexity: 9.2310
Epoch [1/3], Step [2871/3236], Loss: 2.6233, Perplexity: 13.7806
Epoch [1/3], Step [2872/3236], Loss: 2.3525, Perplexity: 10.5115
Epoch [1/3], Step [2873/3236], Loss: 2.1896, Perplexity: 8.9315
Epoch [1/3], Step [2874/3236], Loss: 2.3671, Perplexity: 10.6661
Epoch [1/3], Step [2875/3236], Loss: 2.1576, Perplexity: 8.6504
Epoch [1/3], Step [2876/3236], Loss: 2.2726, Perplexity: 9.7042
Epoch [1/3], Step [2877/3236], Loss: 2.0890, Perplexity: 8.0772
Epoch [1/3], Step [2878/3236], Loss: 2.3098, Perplexity: 10.0721
Epoch [1/3], Step [2879/3236], Loss: 2.3032, Perplexity: 10.0062
Epoch [1/3], Step [2880/3236], Loss: 2.2839, Perplexity: 9.8152
Epoch [1/3], Step [2881/3236], Loss: 2.2078, Perplexity: 9.0959
Epoch [1/3], Step [2882/3236], Loss: 2.1866, Perplexity: 8.9053
Epoch [1/3], Step [2883/3236], Loss: 2.2925, Perplexity: 9.8999
Epoch [1/3], Step [2884/3236], Loss: 2.1668, Perplexity: 8.7307
Epoch [1/3], Step [2885/3236], Loss: 2.2321, Perplexity: 9.3190
Epoch [1/3], Step [2886/3236], Loss: 2.5076, Perplexity: 12.2752
Epoch [1/3], Step [2887/3236], Loss: 2.3061, Perplexity: 10.0356
Epoch [1/3], Step [2888/3236], Loss: 2.1526, Perplexity: 8.6071
Epoch [1/3], Step [2889/3236], Loss: 2.1745, Perplexity: 8.7981
Epoch [1/3], Step [2890/3236], Loss: 2.1639, Perplexity: 8.7051
Epoch [1/3], Step [2891/3236], Loss: 2.3603, Perplexity: 10.5941
Epoch [1/3], Step [2892/3236], Loss: 2.2945, Perplexity: 9.9191
Epoch [1/3], Step [2893/3236], Loss: 2.1939, Perplexity: 8.9702
Epoch [1/3], Step [2894/3236], Loss: 2.3257, Perplexity: 10.2337
Epoch [1/3], Step [2895/3236], Loss: 2.5945, Perplexity: 13.3892
Epoch [1/3], Step [2896/3236], Loss: 2.3292, Perplexity: 10.2695
Epoch [1/3], Step [2897/3236], Loss: 2.5210, Perplexity: 12.4407
Epoch [1/3], Step [2898/3236], Loss: 2.2997, Perplexity: 9.9712
Epoch [1/3], Step [2899/3236], Loss: 2.2996, Perplexity: 9.9704
Epoch [1/3], Step [2900/3236], Loss: 2.2113, Perplexity: 9.1275
Epoch [1/3], Step [2901/3236], Loss: 2.6211, Perplexity: 13.7515
Epoch [1/3], Step [2902/3236], Loss: 2.2018, Perplexity: 9.0410
Epoch [1/3], Step [2903/3236], Loss: 2.3763, Perplexity: 10.7647
Epoch [1/3], Step [2904/3236], Loss: 2.2613, Perplexity: 9.5957
Epoch [1/3], Step [2905/3236], Loss: 2.1636, Perplexity: 8.7024
Epoch [1/3], Step [2906/3236], Loss: 2.2633, Perplexity: 9.6149
Epoch [1/3], Step [2907/3236], Loss: 2.4202, Perplexity: 11.2480
Epoch [1/3], Step [2908/3236], Loss: 2.3236, Perplexity: 10.2122
Epoch [1/3], Step [2909/3236], Loss: 2.2956, Perplexity: 9.9299
Epoch [1/3], Step [2910/3236], Loss: 2.2835, Perplexity: 9.8114
Epoch [1/3], Step [2911/3236], Loss: 2.2840, Perplexity: 9.8156
Epoch [1/3], Step [2912/3236], Loss: 2.1845, Perplexity: 8.8865
Epoch [1/3], Step [2913/3236], Loss: 2.5070, Perplexity: 12.2684
Epoch [1/3], Step [2914/3236], Loss: 2.1421, Perplexity: 8.5172
Epoch [1/3], Step [2915/3236], Loss: 2.1706, Perplexity: 8.7636
Epoch [1/3], Step [2916/3236], Loss: 2.3238, Perplexity: 10.2148
Epoch [1/3], Step [2917/3236], Loss: 2.1871, Perplexity: 8.9092
Epoch [1/3], Step [2918/3236], Loss: 2.3120, Perplexity: 10.0947
Epoch [1/3], Step [2919/3236], Loss: 2.3814, Perplexity: 10.8197
Epoch [1/3], Step [2920/3236], Loss: 2.2326, Perplexity: 9.3240
Epoch [1/3], Step [2921/3236], Loss: 2.4482, Perplexity: 11.5671
Epoch [1/3], Step [2922/3236], Loss: 2.2296, Perplexity: 9.2963
Epoch [1/3], Step [2923/3236], Loss: 2.1595, Perplexity: 8.6671
Epoch [1/3], Step [2924/3236], Loss: 2.1373, Perplexity: 8.4766
Epoch [1/3], Step [2925/3236], Loss: 2.1665, Perplexity: 8.7274
Epoch [1/3], Step [2926/3236], Loss: 2.5834, Perplexity: 13.2426
Epoch [1/3], Step [2927/3236], Loss: 2.3001, Perplexity: 9.9750
Epoch [1/3], Step [2928/3236], Loss: 2.3961, Perplexity: 10.9803
Epoch [1/3], Step [2929/3236], Loss: 2.2197, Perplexity: 9.2043
Epoch [1/3], Step [2930/3236], Loss: 2.1657, Perplexity: 8.7208
Epoch [1/3], Step [2931/3236], Loss: 2.1196, Perplexity: 8.3281
Epoch [1/3], Step [2932/3236], Loss: 2.2452, Perplexity: 9.4424
Epoch [1/3], Step [2933/3236], Loss: 2.2470, Perplexity: 9.4595
Epoch [1/3], Step [2934/3236], Loss: 2.2948, Perplexity: 9.9227
Epoch [1/3], Step [2935/3236], Loss: 2.2339, Perplexity: 9.3365
Epoch [1/3], Step [2936/3236], Loss: 2.2988, Perplexity: 9.9620
Epoch [1/3], Step [2937/3236], Loss: 2.3225, Perplexity: 10.2014
Epoch [1/3], Step [2938/3236], Loss: 2.1947, Perplexity: 8.9770
Epoch [1/3], Step [2939/3236], Loss: 2.2185, Perplexity: 9.1933
Epoch [1/3], Step [2940/3236], Loss: 2.2855, Perplexity: 9.8303
Epoch [1/3], Step [2941/3236], Loss: 2.1560, Perplexity: 8.6364
Epoch [1/3], Step [2942/3236], Loss: 2.1672, Perplexity: 8.7334
Epoch [1/3], Step [2943/3236], Loss: 2.1120, Perplexity: 8.2651
Epoch [1/3], Step [2944/3236], Loss: 2.1797, Perplexity: 8.8432
Epoch [1/3], Step [2945/3236], Loss: 2.5849, Perplexity: 13.2624
Epoch [1/3], Step [2946/3236], Loss: 2.3763, Perplexity: 10.7655
Epoch [1/3], Step [2947/3236], Loss: 2.2311, Perplexity: 9.3101
Epoch [1/3], Step [2948/3236], Loss: 2.3832, Perplexity: 10.8400
Epoch [1/3], Step [2949/3236], Loss: 2.6023, Perplexity: 13.4946
Epoch [1/3], Step [2950/3236], Loss: 3.2409, Perplexity: 25.5569
Epoch [1/3], Step [2951/3236], Loss: 2.1647, Perplexity: 8.7121
Epoch [1/3], Step [2952/3236], Loss: 2.2491, Perplexity: 9.4796
Epoch [1/3], Step [2953/3236], Loss: 2.2759, Perplexity: 9.7366
Epoch [1/3], Step [2954/3236], Loss: 2.3538, Perplexity: 10.5260
Epoch [1/3], Step [2955/3236], Loss: 2.2651, Perplexity: 9.6324
Epoch [1/3], Step [2956/3236], Loss: 2.2177, Perplexity: 9.1860
Epoch [1/3], Step [2957/3236], Loss: 2.1418, Perplexity: 8.5151
Epoch [1/3], Step [2958/3236], Loss: 2.4639, Perplexity: 11.7500
Epoch [1/3], Step [2959/3236], Loss: 2.8755, Perplexity: 17.7343
Epoch [1/3], Step [2960/3236], Loss: 2.2705, Perplexity: 9.6843
Epoch [1/3], Step [2961/3236], Loss: 2.4920, Perplexity: 12.0858
Epoch [1/3], Step [2962/3236], Loss: 2.1548, Perplexity: 8.6263
Epoch [1/3], Step [2963/3236], Loss: 2.1708, Perplexity: 8.7654
Epoch [1/3], Step [2964/3236], Loss: 3.2040, Perplexity: 24.6310
Epoch [1/3], Step [2965/3236], Loss: 2.4454, Perplexity: 11.5353
Epoch [1/3], Step [2966/3236], Loss: 2.7007, Perplexity: 14.8895
Epoch [1/3], Step [2967/3236], Loss: 2.0648, Perplexity: 7.8839
Epoch [1/3], Step [2968/3236], Loss: 2.2875, Perplexity: 9.8505
Epoch [1/3], Step [2969/3236], Loss: 2.2021, Perplexity: 9.0443
Epoch [1/3], Step [2970/3236], Loss: 2.0950, Perplexity: 8.1253
Epoch [1/3], Step [2971/3236], Loss: 2.5154, Perplexity: 12.3719
Epoch [1/3], Step [2972/3236], Loss: 2.2285, Perplexity: 9.2856
Epoch [1/3], Step [2973/3236], Loss: 2.2264, Perplexity: 9.2663
Epoch [1/3], Step [2974/3236], Loss: 2.2196, Perplexity: 9.2035
Epoch [1/3], Step [2975/3236], Loss: 2.4253, Perplexity: 11.3057
Epoch [1/3], Step [2976/3236], Loss: 2.2893, Perplexity: 9.8684
Epoch [1/3], Step [2977/3236], Loss: 2.2341, Perplexity: 9.3379
Epoch [1/3], Step [2978/3236], Loss: 2.2060, Perplexity: 9.0796
Epoch [1/3], Step [2979/3236], Loss: 2.1635, Perplexity: 8.7017
Epoch [1/3], Step [2980/3236], Loss: 2.6115, Perplexity: 13.6198
Epoch [1/3], Step [2981/3236], Loss: 2.1782, Perplexity: 8.8301
Epoch [1/3], Step [2982/3236], Loss: 2.4552, Perplexity: 11.6483
Epoch [1/3], Step [2983/3236], Loss: 2.3497, Perplexity: 10.4826
Epoch [1/3], Step [2984/3236], Loss: 2.1801, Perplexity: 8.8470
Epoch [1/3], Step [2985/3236], Loss: 2.3503, Perplexity: 10.4888
Epoch [1/3], Step [2986/3236], Loss: 2.2238, Perplexity: 9.2422
Epoch [1/3], Step [2987/3236], Loss: 2.6816, Perplexity: 14.6091
Epoch [1/3], Step [2988/3236], Loss: 2.1502, Perplexity: 8.5870
Epoch [1/3], Step [2989/3236], Loss: 2.5313, Perplexity: 12.5704
Epoch [1/3], Step [2990/3236], Loss: 2.0330, Perplexity: 7.6369
Epoch [1/3], Step [2991/3236], Loss: 2.9944, Perplexity: 19.9724
Epoch [1/3], Step [2992/3236], Loss: 2.1982, Perplexity: 9.0083
Epoch [1/3], Step [2993/3236], Loss: 2.2753, Perplexity: 9.7309
Epoch [1/3], Step [2994/3236], Loss: 2.3668, Perplexity: 10.6629
Epoch [1/3], Step [2995/3236], Loss: 2.2167, Perplexity: 9.1769
Epoch [1/3], Step [2996/3236], Loss: 2.2612, Perplexity: 9.5944
Epoch [1/3], Step [2997/3236], Loss: 2.3083, Perplexity: 10.0571
Epoch [1/3], Step [2998/3236], Loss: 2.2698, Perplexity: 9.6771
Epoch [1/3], Step [2999/3236], Loss: 2.3367, Perplexity: 10.3471
Epoch [1/3], Step [3000/3236], Loss: 2.3797, Perplexity: 10.8013
Epoch [1/3], Step [3001/3236], Loss: 2.5036, Perplexity: 12.2266
Epoch [1/3], Step [3002/3236], Loss: 2.1513, Perplexity: 8.5962
Epoch [1/3], Step [3003/3236], Loss: 2.3763, Perplexity: 10.7655
Epoch [1/3], Step [3004/3236], Loss: 2.0529, Perplexity: 7.7904
Epoch [1/3], Step [3005/3236], Loss: 2.2060, Perplexity: 9.0796
Epoch [1/3], Step [3006/3236], Loss: 2.1190, Perplexity: 8.3226
Epoch [1/3], Step [3007/3236], Loss: 2.1785, Perplexity: 8.8334
Epoch [1/3], Step [3008/3236], Loss: 2.2419, Perplexity: 9.4115
Epoch [1/3], Step [3009/3236], Loss: 2.2411, Perplexity: 9.4033
Epoch [1/3], Step [3010/3236], Loss: 2.0845, Perplexity: 8.0409
Epoch [1/3], Step [3011/3236], Loss: 2.4258, Perplexity: 11.3111
Epoch [1/3], Step [3012/3236], Loss: 2.1841, Perplexity: 8.8830
Epoch [1/3], Step [3013/3236], Loss: 3.1326, Perplexity: 22.9338
Epoch [1/3], Step [3014/3236], Loss: 2.1311, Perplexity: 8.4242
Epoch [1/3], Step [3015/3236], Loss: 2.2749, Perplexity: 9.7268
Epoch [1/3], Step [3016/3236], Loss: 2.1987, Perplexity: 9.0133
Epoch [1/3], Step [3017/3236], Loss: 2.7699, Perplexity: 15.9577
Epoch [1/3], Step [3018/3236], Loss: 2.3287, Perplexity: 10.2650
Epoch [1/3], Step [3019/3236], Loss: 2.2351, Perplexity: 9.3470
Epoch [1/3], Step [3020/3236], Loss: 2.1915, Perplexity: 8.9490
Epoch [1/3], Step [3021/3236], Loss: 2.7526, Perplexity: 15.6841
Epoch [1/3], Step [3022/3236], Loss: 2.2814, Perplexity: 9.7908
Epoch [1/3], Step [3023/3236], Loss: 2.1798, Perplexity: 8.8445
Epoch [1/3], Step [3024/3236], Loss: 2.2261, Perplexity: 9.2634
Epoch [1/3], Step [3025/3236], Loss: 2.1899, Perplexity: 8.9345
Epoch [1/3], Step [3026/3236], Loss: 2.1736, Perplexity: 8.7895
Epoch [1/3], Step [3027/3236], Loss: 2.1933, Perplexity: 8.9651
Epoch [1/3], Step [3028/3236], Loss: 2.2128, Perplexity: 9.1409
Epoch [1/3], Step [3029/3236], Loss: 2.1905, Perplexity: 8.9393
Epoch [1/3], Step [3030/3236], Loss: 2.1461, Perplexity: 8.5518
Epoch [1/3], Step [3031/3236], Loss: 2.2103, Perplexity: 9.1188
Epoch [1/3], Step [3032/3236], Loss: 2.2754, Perplexity: 9.7321
Epoch [1/3], Step [3033/3236], Loss: 2.1822, Perplexity: 8.8655
Epoch [1/3], Step [3034/3236], Loss: 2.2745, Perplexity: 9.7226
Epoch [1/3], Step [3035/3236], Loss: 2.3290, Perplexity: 10.2681
Epoch [1/3], Step [3036/3236], Loss: 2.3343, Perplexity: 10.3220
Epoch [1/3], Step [3037/3236], Loss: 2.1772, Perplexity: 8.8219
Epoch [1/3], Step [3038/3236], Loss: 2.2551, Perplexity: 9.5366
Epoch [1/3], Step [3039/3236], Loss: 2.1415, Perplexity: 8.5123
Epoch [1/3], Step [3040/3236], Loss: 2.1701, Perplexity: 8.7595
Epoch [1/3], Step [3041/3236], Loss: 2.3452, Perplexity: 10.4349
Epoch [1/3], Step [3042/3236], Loss: 2.5909, Perplexity: 13.3419
Epoch [1/3], Step [3043/3236], Loss: 2.1527, Perplexity: 8.6083
Epoch [1/3], Step [3044/3236], Loss: 2.1469, Perplexity: 8.5584
Epoch [1/3], Step [3045/3236], Loss: 2.2850, Perplexity: 9.8258
Epoch [1/3], Step [3046/3236], Loss: 2.0684, Perplexity: 7.9119
Epoch [1/3], Step [3047/3236], Loss: 2.2508, Perplexity: 9.4955
Epoch [1/3], Step [3048/3236], Loss: 2.1277, Perplexity: 8.3952
Epoch [1/3], Step [3049/3236], Loss: 2.2068, Perplexity: 9.0867
Epoch [1/3], Step [3050/3236], Loss: 2.1212, Perplexity: 8.3408
Epoch [1/3], Step [3051/3236], Loss: 2.1537, Perplexity: 8.6171
Epoch [1/3], Step [3052/3236], Loss: 2.2725, Perplexity: 9.7038
Epoch [1/3], Step [3053/3236], Loss: 2.1641, Perplexity: 8.7070
Epoch [1/3], Step [3054/3236], Loss: 3.1554, Perplexity: 23.4629
Epoch [1/3], Step [3055/3236], Loss: 2.2261, Perplexity: 9.2633
Epoch [1/3], Step [3056/3236], Loss: 2.3327, Perplexity: 10.3061
Epoch [1/3], Step [3057/3236], Loss: 2.4183, Perplexity: 11.2264
Epoch [1/3], Step [3058/3236], Loss: 2.2584, Perplexity: 9.5676
Epoch [1/3], Step [3059/3236], Loss: 2.2141, Perplexity: 9.1531
Epoch [1/3], Step [3060/3236], Loss: 2.2655, Perplexity: 9.6355
Epoch [1/3], Step [3061/3236], Loss: 2.3481, Perplexity: 10.4655
Epoch [1/3], Step [3062/3236], Loss: 2.2451, Perplexity: 9.4410
Epoch [1/3], Step [3063/3236], Loss: 2.2227, Perplexity: 9.2326
Epoch [1/3], Step [3064/3236], Loss: 2.5934, Perplexity: 13.3756
Epoch [1/3], Step [3065/3236], Loss: 2.1842, Perplexity: 8.8832
Epoch [1/3], Step [3066/3236], Loss: 2.2315, Perplexity: 9.3140
Epoch [1/3], Step [3067/3236], Loss: 2.5275, Perplexity: 12.5220
Epoch [1/3], Step [3068/3236], Loss: 2.2690, Perplexity: 9.6693
Epoch [1/3], Step [3069/3236], Loss: 2.1920, Perplexity: 8.9532
Epoch [1/3], Step [3070/3236], Loss: 2.4042, Perplexity: 11.0694
Epoch [1/3], Step [3071/3236], Loss: 2.3914, Perplexity: 10.9283
Epoch [1/3], Step [3072/3236], Loss: 2.2364, Perplexity: 9.3592
Epoch [1/3], Step [3073/3236], Loss: 2.7197, Perplexity: 15.1757
Epoch [1/3], Step [3074/3236], Loss: 2.2641, Perplexity: 9.6227
Epoch [1/3], Step [3075/3236], Loss: 2.2415, Perplexity: 9.4070
Epoch [1/3], Step [3076/3236], Loss: 2.1909, Perplexity: 8.9430
Epoch [1/3], Step [3077/3236], Loss: 2.1273, Perplexity: 8.3924
Epoch [1/3], Step [3078/3236], Loss: 2.1480, Perplexity: 8.5679
Epoch [1/3], Step [3079/3236], Loss: 2.0959, Perplexity: 8.1326
Epoch [1/3], Step [3080/3236], Loss: 2.2140, Perplexity: 9.1526
Epoch [1/3], Step [3081/3236], Loss: 2.0266, Perplexity: 7.5885
Epoch [1/3], Step [3082/3236], Loss: 2.4447, Perplexity: 11.5268
Epoch [1/3], Step [3083/3236], Loss: 2.4508, Perplexity: 11.5981
Epoch [1/3], Step [3084/3236], Loss: 2.1655, Perplexity: 8.7186
Epoch [1/3], Step [3085/3236], Loss: 2.2785, Perplexity: 9.7621
Epoch [1/3], Step [3086/3236], Loss: 2.2175, Perplexity: 9.1847
Epoch [1/3], Step [3087/3236], Loss: 2.3466, Perplexity: 10.4497
Epoch [1/3], Step [3088/3236], Loss: 2.2244, Perplexity: 9.2479
Epoch [1/3], Step [3089/3236], Loss: 2.4994, Perplexity: 12.1755
Epoch [1/3], Step [3090/3236], Loss: 2.4594, Perplexity: 11.6978
Epoch [1/3], Step [3091/3236], Loss: 2.4115, Perplexity: 11.1502
Epoch [1/3], Step [3092/3236], Loss: 2.5197, Perplexity: 12.4251
Epoch [1/3], Step [3093/3236], Loss: 2.2037, Perplexity: 9.0582
Epoch [1/3], Step [3094/3236], Loss: 2.2589, Perplexity: 9.5723
Epoch [1/3], Step [3095/3236], Loss: 1.9705, Perplexity: 7.1745
Epoch [1/3], Step [3096/3236], Loss: 2.3068, Perplexity: 10.0419
Epoch [1/3], Step [3097/3236], Loss: 2.2440, Perplexity: 9.4306
Epoch [1/3], Step [3098/3236], Loss: 2.2477, Perplexity: 9.4659
Epoch [1/3], Step [3099/3236], Loss: 2.3166, Perplexity: 10.1415
Epoch [1/3], Step [3100/3236], Loss: 3.0284, Perplexity: 20.6643
Epoch [1/3], Step [3101/3236], Loss: 2.2328, Perplexity: 9.3255
Epoch [1/3], Step [3102/3236], Loss: 2.1879, Perplexity: 8.9163
Epoch [1/3], Step [3103/3236], Loss: 2.1023, Perplexity: 8.1851
Epoch [1/3], Step [3104/3236], Loss: 2.0955, Perplexity: 8.1294
Epoch [1/3], Step [3105/3236], Loss: 2.3182, Perplexity: 10.1573
Epoch [1/3], Step [3106/3236], Loss: 2.4177, Perplexity: 11.2196
Epoch [1/3], Step [3107/3236], Loss: 2.1171, Perplexity: 8.3071
Epoch [1/3], Step [3108/3236], Loss: 2.2472, Perplexity: 9.4613
Epoch [1/3], Step [3109/3236], Loss: 2.2862, Perplexity: 9.8379
Epoch [1/3], Step [3110/3236], Loss: 2.1118, Perplexity: 8.2632
Epoch [1/3], Step [3111/3236], Loss: 2.5801, Perplexity: 13.1987
Epoch [1/3], Step [3112/3236], Loss: 2.1357, Perplexity: 8.4630
Epoch [1/3], Step [3113/3236], Loss: 2.2915, Perplexity: 9.8899
Epoch [1/3], Step [3114/3236], Loss: 2.1070, Perplexity: 8.2236
Epoch [1/3], Step [3115/3236], Loss: 2.5953, Perplexity: 13.4008
Epoch [1/3], Step [3116/3236], Loss: 3.4162, Perplexity: 30.4538
Epoch [1/3], Step [3117/3236], Loss: 2.2029, Perplexity: 9.0515
Epoch [1/3], Step [3118/3236], Loss: 2.0998, Perplexity: 8.1649
Epoch [1/3], Step [3119/3236], Loss: 2.1272, Perplexity: 8.3915
Epoch [1/3], Step [3120/3236], Loss: 2.3166, Perplexity: 10.1407
Epoch [1/3], Step [3121/3236], Loss: 2.1699, Perplexity: 8.7575
Epoch [1/3], Step [3122/3236], Loss: 2.4522, Perplexity: 11.6139
Epoch [1/3], Step [3123/3236], Loss: 2.1617, Perplexity: 8.6862
Epoch [1/3], Step [3124/3236], Loss: 2.4340, Perplexity: 11.4044
Epoch [1/3], Step [3125/3236], Loss: 2.2411, Perplexity: 9.4037
Epoch [1/3], Step [3126/3236], Loss: 2.2010, Perplexity: 9.0336
Epoch [1/3], Step [3127/3236], Loss: 2.3653, Perplexity: 10.6474
Epoch [1/3], Step [3128/3236], Loss: 2.9769, Perplexity: 19.6274
Epoch [1/3], Step [3129/3236], Loss: 2.1950, Perplexity: 8.9802
Epoch [1/3], Step [3130/3236], Loss: 2.5286, Perplexity: 12.5364
Epoch [1/3], Step [3131/3236], Loss: 2.2059, Perplexity: 9.0785
Epoch [1/3], Step [3132/3236], Loss: 2.1967, Perplexity: 8.9954
Epoch [1/3], Step [3133/3236], Loss: 2.1298, Perplexity: 8.4136
Epoch [1/3], Step [3134/3236], Loss: 2.2575, Perplexity: 9.5596
Epoch [1/3], Step [3135/3236], Loss: 2.2914, Perplexity: 9.8887
Epoch [1/3], Step [3136/3236], Loss: 2.2560, Perplexity: 9.5449
Epoch [1/3], Step [3137/3236], Loss: 2.7105, Perplexity: 15.0366
Epoch [1/3], Step [3138/3236], Loss: 2.3616, Perplexity: 10.6080
Epoch [1/3], Step [3139/3236], Loss: 2.4160, Perplexity: 11.2010
Epoch [1/3], Step [3140/3236], Loss: 2.2652, Perplexity: 9.6328
Epoch [1/3], Step [3141/3236], Loss: 2.0998, Perplexity: 8.1642
Epoch [1/3], Step [3142/3236], Loss: 2.5398, Perplexity: 12.6775
Epoch [1/3], Step [3143/3236], Loss: 2.1890, Perplexity: 8.9263
Epoch [1/3], Step [3144/3236], Loss: 2.2549, Perplexity: 9.5341
Epoch [1/3], Step [3145/3236], Loss: 2.1745, Perplexity: 8.7975
Epoch [1/3], Step [3146/3236], Loss: 2.4911, Perplexity: 12.0747
Epoch [1/3], Step [3147/3236], Loss: 2.2722, Perplexity: 9.7003
Epoch [1/3], Step [3148/3236], Loss: 2.4461, Perplexity: 11.5428
Epoch [1/3], Step [3149/3236], Loss: 2.2866, Perplexity: 9.8417
Epoch [1/3], Step [3150/3236], Loss: 2.2435, Perplexity: 9.4265
Epoch [1/3], Step [3151/3236], Loss: 2.2320, Perplexity: 9.3184
Epoch [1/3], Step [3152/3236], Loss: 2.4792, Perplexity: 11.9316
Epoch [1/3], Step [3153/3236], Loss: 2.3018, Perplexity: 9.9924
Epoch [1/3], Step [3154/3236], Loss: 2.3413, Perplexity: 10.3951
Epoch [1/3], Step [3155/3236], Loss: 2.6453, Perplexity: 14.0882
Epoch [1/3], Step [3156/3236], Loss: 2.3386, Perplexity: 10.3672
Epoch [1/3], Step [3157/3236], Loss: 2.1126, Perplexity: 8.2694
Epoch [1/3], Step [3158/3236], Loss: 2.1602, Perplexity: 8.6731
Epoch [1/3], Step [3159/3236], Loss: 2.3600, Perplexity: 10.5914
Epoch [1/3], Step [3160/3236], Loss: 2.0513, Perplexity: 7.7781
Epoch [1/3], Step [3161/3236], Loss: 2.3345, Perplexity: 10.3243
Epoch [1/3], Step [3162/3236], Loss: 2.2730, Perplexity: 9.7082
Epoch [1/3], Step [3163/3236], Loss: 3.1083, Perplexity: 22.3826
Epoch [1/3], Step [3164/3236], Loss: 2.3285, Perplexity: 10.2629
Epoch [1/3], Step [3165/3236], Loss: 2.5269, Perplexity: 12.5148
Epoch [1/3], Step [3166/3236], Loss: 2.1324, Perplexity: 8.4352
Epoch [1/3], Step [3167/3236], Loss: 3.0754, Perplexity: 21.6596
Epoch [1/3], Step [3168/3236], Loss: 2.1686, Perplexity: 8.7461
Epoch [1/3], Step [3169/3236], Loss: 2.2822, Perplexity: 9.7985
Epoch [1/3], Step [3170/3236], Loss: 1.9976, Perplexity: 7.3714
Epoch [1/3], Step [3171/3236], Loss: 2.2661, Perplexity: 9.6415
Epoch [1/3], Step [3172/3236], Loss: 2.0665, Perplexity: 7.8973
Epoch [1/3], Step [3173/3236], Loss: 3.0832, Perplexity: 21.8291
Epoch [1/3], Step [3174/3236], Loss: 2.3519, Perplexity: 10.5055
Epoch [1/3], Step [3175/3236], Loss: 2.1717, Perplexity: 8.7730
Epoch [1/3], Step [3176/3236], Loss: 2.3010, Perplexity: 9.9837
Epoch [1/3], Step [3177/3236], Loss: 2.3914, Perplexity: 10.9292
Epoch [1/3], Step [3178/3236], Loss: 2.2984, Perplexity: 9.9586
Epoch [1/3], Step [3179/3236], Loss: 2.1921, Perplexity: 8.9536
Epoch [1/3], Step [3180/3236], Loss: 2.2133, Perplexity: 9.1460
Epoch [1/3], Step [3181/3236], Loss: 2.7213, Perplexity: 15.1998
Epoch [1/3], Step [3182/3236], Loss: 2.3918, Perplexity: 10.9332
Epoch [1/3], Step [3183/3236], Loss: 2.2212, Perplexity: 9.2187
Epoch [1/3], Step [3184/3236], Loss: 2.3457, Perplexity: 10.4405
Epoch [1/3], Step [3185/3236], Loss: 2.3371, Perplexity: 10.3513
Epoch [1/3], Step [3186/3236], Loss: 2.3705, Perplexity: 10.7024
Epoch [1/3], Step [3187/3236], Loss: 2.3026, Perplexity: 10.0002
Epoch [1/3], Step [3188/3236], Loss: 2.4220, Perplexity: 11.2688
Epoch [1/3], Step [3189/3236], Loss: 2.2781, Perplexity: 9.7578
Epoch [1/3], Step [3190/3236], Loss: 2.0774, Perplexity: 7.9834
Epoch [1/3], Step [3191/3236], Loss: 2.7932, Perplexity: 16.3325
Epoch [1/3], Step [3192/3236], Loss: 2.2446, Perplexity: 9.4368
Epoch [1/3], Step [3193/3236], Loss: 2.1812, Perplexity: 8.8570
Epoch [1/3], Step [3194/3236], Loss: 2.2422, Perplexity: 9.4137
Epoch [1/3], Step [3195/3236], Loss: 2.3136, Perplexity: 10.1105
Epoch [1/3], Step [3196/3236], Loss: 2.1862, Perplexity: 8.9009
Epoch [1/3], Step [3197/3236], Loss: 2.3854, Perplexity: 10.8639
Epoch [1/3], Step [3198/3236], Loss: 2.3373, Perplexity: 10.3532
Epoch [1/3], Step [3199/3236], Loss: 2.2172, Perplexity: 9.1818
Epoch [1/3], Step [3200/3236], Loss: 2.2423, Perplexity: 9.4153
Epoch [1/3], Step [3201/3236], Loss: 2.3600, Perplexity: 10.5911
Epoch [1/3], Step [3202/3236], Loss: 2.7233, Perplexity: 15.2302
Epoch [1/3], Step [3203/3236], Loss: 2.2072, Perplexity: 9.0904
Epoch [1/3], Step [3204/3236], Loss: 2.2251, Perplexity: 9.2546
Epoch [1/3], Step [3205/3236], Loss: 2.2148, Perplexity: 9.1593
Epoch [1/3], Step [3206/3236], Loss: 2.3528, Perplexity: 10.5150
Epoch [1/3], Step [3207/3236], Loss: 2.4967, Perplexity: 12.1423
Epoch [1/3], Step [3208/3236], Loss: 2.2416, Perplexity: 9.4083
Epoch [1/3], Step [3209/3236], Loss: 2.3732, Perplexity: 10.7320
Epoch [1/3], Step [3210/3236], Loss: 2.3611, Perplexity: 10.6023
Epoch [1/3], Step [3211/3236], Loss: 2.2218, Perplexity: 9.2238
Epoch [1/3], Step [3212/3236], Loss: 2.4045, Perplexity: 11.0724
Epoch [1/3], Step [3213/3236], Loss: 2.4078, Perplexity: 11.1097
Epoch [1/3], Step [3214/3236], Loss: 2.1871, Perplexity: 8.9095
Epoch [1/3], Step [3215/3236], Loss: 2.2678, Perplexity: 9.6581
Epoch [1/3], Step [3216/3236], Loss: 2.2568, Perplexity: 9.5526
Epoch [1/3], Step [3217/3236], Loss: 2.1360, Perplexity: 8.4656
Epoch [1/3], Step [3218/3236], Loss: 2.2067, Perplexity: 9.0861
Epoch [1/3], Step [3219/3236], Loss: 2.1261, Perplexity: 8.3825
Epoch [1/3], Step [3220/3236], Loss: 3.3941, Perplexity: 29.7883
Epoch [1/3], Step [3221/3236], Loss: 2.3156, Perplexity: 10.1313
Epoch [1/3], Step [3222/3236], Loss: 2.5045, Perplexity: 12.2376
Epoch [1/3], Step [3223/3236], Loss: 2.1927, Perplexity: 8.9590
Epoch [1/3], Step [3224/3236], Loss: 2.3161, Perplexity: 10.1362
Epoch [1/3], Step [3225/3236], Loss: 2.3526, Perplexity: 10.5133
Epoch [1/3], Step [3226/3236], Loss: 2.2637, Perplexity: 9.6188
Epoch [1/3], Step [3227/3236], Loss: 2.2476, Perplexity: 9.4649
Epoch [1/3], Step [3228/3236], Loss: 2.1654, Perplexity: 8.7178
Epoch [1/3], Step [3229/3236], Loss: 2.4138, Perplexity: 11.1761
Epoch [1/3], Step [3230/3236], Loss: 2.2626, Perplexity: 9.6081
Epoch [1/3], Step [3231/3236], Loss: 2.2638, Perplexity: 9.6193
Epoch [1/3], Step [3232/3236], Loss: 2.1338, Perplexity: 8.4467
Epoch [1/3], Step [3233/3236], Loss: 2.3302, Perplexity: 10.2798
Epoch [1/3], Step [3234/3236], Loss: 2.0347, Perplexity: 7.6500
Epoch [1/3], Step [3235/3236], Loss: 2.3006, Perplexity: 9.9805
Epoch [1/3], Step [3236/3236], Loss: 2.3706, Perplexity: 10.7036
Epoch [2/3], Step [1/3236], Loss: 2.3297, Perplexity: 10.2752
Epoch [2/3], Step [2/3236], Loss: 2.2440, Perplexity: 9.4306
Epoch [2/3], Step [3/3236], Loss: 2.7388, Perplexity: 15.4682
Epoch [2/3], Step [4/3236], Loss: 2.1888, Perplexity: 8.9244
Epoch [2/3], Step [5/3236], Loss: 2.4249, Perplexity: 11.3016
Epoch [2/3], Step [6/3236], Loss: 2.1375, Perplexity: 8.4786
Epoch [2/3], Step [7/3236], Loss: 2.3907, Perplexity: 10.9215
Epoch [2/3], Step [8/3236], Loss: 2.4370, Perplexity: 11.4389
Epoch [2/3], Step [9/3236], Loss: 2.7704, Perplexity: 15.9654
Epoch [2/3], Step [10/3236], Loss: 2.2323, Perplexity: 9.3211
Epoch [2/3], Step [11/3236], Loss: 2.2361, Perplexity: 9.3567
Epoch [2/3], Step [12/3236], Loss: 2.1499, Perplexity: 8.5843
Epoch [2/3], Step [13/3236], Loss: 2.3726, Perplexity: 10.7257
Epoch [2/3], Step [14/3236], Loss: 2.4376, Perplexity: 11.4455
Epoch [2/3], Step [15/3236], Loss: 2.2058, Perplexity: 9.0779
Epoch [2/3], Step [16/3236], Loss: 2.0862, Perplexity: 8.0546
Epoch [2/3], Step [17/3236], Loss: 2.3097, Perplexity: 10.0711
Epoch [2/3], Step [18/3236], Loss: 2.1401, Perplexity: 8.5002
Epoch [2/3], Step [19/3236], Loss: 2.1339, Perplexity: 8.4477
Epoch [2/3], Step [20/3236], Loss: 2.3597, Perplexity: 10.5874
Epoch [2/3], Step [21/3236], Loss: 2.1002, Perplexity: 8.1677
Epoch [2/3], Step [22/3236], Loss: 2.7455, Perplexity: 15.5724
Epoch [2/3], Step [23/3236], Loss: 2.1521, Perplexity: 8.6026
Epoch [2/3], Step [24/3236], Loss: 2.2192, Perplexity: 9.1999
Epoch [2/3], Step [25/3236], Loss: 2.4359, Perplexity: 11.4263
Epoch [2/3], Step [26/3236], Loss: 2.2003, Perplexity: 9.0277
Epoch [2/3], Step [27/3236], Loss: 2.5895, Perplexity: 13.3229
Epoch [2/3], Step [28/3236], Loss: 2.0871, Perplexity: 8.0617
Epoch [2/3], Step [29/3236], Loss: 2.3187, Perplexity: 10.1622
Epoch [2/3], Step [30/3236], Loss: 2.6931, Perplexity: 14.7767
Epoch [2/3], Step [31/3236], Loss: 2.1672, Perplexity: 8.7342
Epoch [2/3], Step [32/3236], Loss: 2.4934, Perplexity: 12.1020
Epoch [2/3], Step [33/3236], Loss: 2.2020, Perplexity: 9.0430
Epoch [2/3], Step [34/3236], Loss: 2.3666, Perplexity: 10.6609
Epoch [2/3], Step [35/3236], Loss: 2.2546, Perplexity: 9.5318
Epoch [2/3], Step [36/3236], Loss: 2.0521, Perplexity: 7.7846
Epoch [2/3], Step [37/3236], Loss: 2.0782, Perplexity: 7.9898
Epoch [2/3], Step [38/3236], Loss: 2.1826, Perplexity: 8.8689
Epoch [2/3], Step [39/3236], Loss: 2.2958, Perplexity: 9.9326
Epoch [2/3], Step [40/3236], Loss: 2.1769, Perplexity: 8.8191
Epoch [2/3], Step [41/3236], Loss: 2.0948, Perplexity: 8.1236
Epoch [2/3], Step [42/3236], Loss: 2.4942, Perplexity: 12.1116
Epoch [2/3], Step [43/3236], Loss: 2.3378, Perplexity: 10.3586
Epoch [2/3], Step [44/3236], Loss: 2.0876, Perplexity: 8.0654
Epoch [2/3], Step [45/3236], Loss: 2.3444, Perplexity: 10.4271
Epoch [2/3], Step [46/3236], Loss: 2.2549, Perplexity: 9.5341
Epoch [2/3], Step [47/3236], Loss: 2.1807, Perplexity: 8.8524
Epoch [2/3], Step [48/3236], Loss: 2.2399, Perplexity: 9.3925
Epoch [2/3], Step [49/3236], Loss: 2.2261, Perplexity: 9.2639
Epoch [2/3], Step [50/3236], Loss: 2.3357, Perplexity: 10.3362
Epoch [2/3], Step [51/3236], Loss: 2.1818, Perplexity: 8.8618
Epoch [2/3], Step [52/3236], Loss: 2.3032, Perplexity: 10.0057
Epoch [2/3], Step [53/3236], Loss: 2.2485, Perplexity: 9.4734
Epoch [2/3], Step [54/3236], Loss: 2.1140, Perplexity: 8.2813
Epoch [2/3], Step [55/3236], Loss: 2.1409, Perplexity: 8.5067
Epoch [2/3], Step [56/3236], Loss: 2.6976, Perplexity: 14.8447
Epoch [2/3], Step [57/3236], Loss: 2.2145, Perplexity: 9.1564
Epoch [2/3], Step [58/3236], Loss: 2.4292, Perplexity: 11.3493
Epoch [2/3], Step [59/3236], Loss: 2.5546, Perplexity: 12.8660
Epoch [2/3], Step [60/3236], Loss: 2.8645, Perplexity: 17.5403
Epoch [2/3], Step [61/3236], Loss: 2.2634, Perplexity: 9.6162
Epoch [2/3], Step [62/3236], Loss: 2.3087, Perplexity: 10.0610
Epoch [2/3], Step [63/3236], Loss: 2.3613, Perplexity: 10.6045
Epoch [2/3], Step [64/3236], Loss: 2.2111, Perplexity: 9.1255
Epoch [2/3], Step [65/3236], Loss: 2.6621, Perplexity: 14.3266
Epoch [2/3], Step [66/3236], Loss: 2.3674, Perplexity: 10.6698
Epoch [2/3], Step [67/3236], Loss: 2.2106, Perplexity: 9.1216
Epoch [2/3], Step [68/3236], Loss: 2.9285, Perplexity: 18.7004
Epoch [2/3], Step [69/3236], Loss: 2.3461, Perplexity: 10.4452
Epoch [2/3], Step [70/3236], Loss: 2.4511, Perplexity: 11.6014
Epoch [2/3], Step [71/3236], Loss: 2.6719, Perplexity: 14.4672
Epoch [2/3], Step [72/3236], Loss: 2.2708, Perplexity: 9.6876
Epoch [2/3], Step [73/3236], Loss: 2.2003, Perplexity: 9.0274
Epoch [2/3], Step [74/3236], Loss: 2.3411, Perplexity: 10.3926
Epoch [2/3], Step [75/3236], Loss: 2.3537, Perplexity: 10.5241
Epoch [2/3], Step [76/3236], Loss: 3.0495, Perplexity: 21.1054
Epoch [2/3], Step [77/3236], Loss: 2.1592, Perplexity: 8.6644
Epoch [2/3], Step [78/3236], Loss: 2.4721, Perplexity: 11.8475
Epoch [2/3], Step [79/3236], Loss: 2.2734, Perplexity: 9.7120
Epoch [2/3], Step [80/3236], Loss: 2.3471, Perplexity: 10.4547
Epoch [2/3], Step [81/3236], Loss: 2.3656, Perplexity: 10.6505
Epoch [2/3], Step [82/3236], Loss: 2.5535, Perplexity: 12.8523
Epoch [2/3], Step [83/3236], Loss: 2.1924, Perplexity: 8.9570
Epoch [2/3], Step [84/3236], Loss: 2.4052, Perplexity: 11.0802
Epoch [2/3], Step [85/3236], Loss: 2.2699, Perplexity: 9.6785
Epoch [2/3], Step [86/3236], Loss: 2.4511, Perplexity: 11.6015
Epoch [2/3], Step [87/3236], Loss: 2.2286, Perplexity: 9.2868
Epoch [2/3], Step [88/3236], Loss: 2.1918, Perplexity: 8.9515
Epoch [2/3], Step [89/3236], Loss: 2.3860, Perplexity: 10.8695
Epoch [2/3], Step [90/3236], Loss: 2.2248, Perplexity: 9.2514
Epoch [2/3], Step [91/3236], Loss: 2.2970, Perplexity: 9.9440
Epoch [2/3], Step [92/3236], Loss: 2.2371, Perplexity: 9.3659
Epoch [2/3], Step [93/3236], Loss: 2.4764, Perplexity: 11.8986
Epoch [2/3], Step [94/3236], Loss: 2.2143, Perplexity: 9.1548
Epoch [2/3], Step [95/3236], Loss: 2.2164, Perplexity: 9.1745
Epoch [2/3], Step [96/3236], Loss: 2.1459, Perplexity: 8.5496
Epoch [2/3], Step [97/3236], Loss: 2.1494, Perplexity: 8.5801
Epoch [2/3], Step [98/3236], Loss: 2.4359, Perplexity: 11.4264
Epoch [2/3], Step [99/3236], Loss: 2.5688, Perplexity: 13.0503
Epoch [2/3], Step [100/3236], Loss: 2.3018, Perplexity: 9.9924
Epoch [2/3], Step [101/3236], Loss: 2.2608, Perplexity: 9.5907
Epoch [2/3], Step [102/3236], Loss: 2.1472, Perplexity: 8.5608
Epoch [2/3], Step [103/3236], Loss: 2.2816, Perplexity: 9.7922
Epoch [2/3], Step [104/3236], Loss: 2.6163, Perplexity: 13.6844
Epoch [2/3], Step [105/3236], Loss: 2.5606, Perplexity: 12.9432
Epoch [2/3], Step [106/3236], Loss: 2.4057, Perplexity: 11.0862
Epoch [2/3], Step [107/3236], Loss: 2.1283, Perplexity: 8.4007
Epoch [2/3], Step [108/3236], Loss: 2.1127, Perplexity: 8.2705
Epoch [2/3], Step [109/3236], Loss: 2.3920, Perplexity: 10.9357
Epoch [2/3], Step [110/3236], Loss: 2.2044, Perplexity: 9.0651
Epoch [2/3], Step [111/3236], Loss: 2.4699, Perplexity: 11.8217
Epoch [2/3], Step [112/3236], Loss: 2.2837, Perplexity: 9.8127
Epoch [2/3], Step [113/3236], Loss: 2.2127, Perplexity: 9.1405
Epoch [2/3], Step [114/3236], Loss: 2.0803, Perplexity: 8.0070
Epoch [2/3], Step [115/3236], Loss: 2.4588, Perplexity: 11.6913
Epoch [2/3], Step [116/3236], Loss: 2.1144, Perplexity: 8.2845
Epoch [2/3], Step [117/3236], Loss: 2.3749, Perplexity: 10.7501
Epoch [2/3], Step [118/3236], Loss: 2.2600, Perplexity: 9.5834
Epoch [2/3], Step [119/3236], Loss: 2.1601, Perplexity: 8.6723
Epoch [2/3], Step [120/3236], Loss: 2.2870, Perplexity: 9.8458
Epoch [2/3], Step [121/3236], Loss: 2.3031, Perplexity: 10.0056
Epoch [2/3], Step [122/3236], Loss: 2.0032, Perplexity: 7.4124
Epoch [2/3], Step [123/3236], Loss: 2.1710, Perplexity: 8.7667
Epoch [2/3], Step [124/3236], Loss: 2.2344, Perplexity: 9.3412
Epoch [2/3], Step [125/3236], Loss: 2.2177, Perplexity: 9.1859
Epoch [2/3], Step [126/3236], Loss: 2.1483, Perplexity: 8.5703
Epoch [2/3], Step [127/3236], Loss: 2.8839, Perplexity: 17.8843
Epoch [2/3], Step [128/3236], Loss: 2.1259, Perplexity: 8.3806
Epoch [2/3], Step [129/3236], Loss: 2.1128, Perplexity: 8.2716
Epoch [2/3], Step [130/3236], Loss: 2.1808, Perplexity: 8.8530
Epoch [2/3], Step [131/3236], Loss: 2.3210, Perplexity: 10.1863
Epoch [2/3], Step [132/3236], Loss: 2.1440, Perplexity: 8.5332
Epoch [2/3], Step [133/3236], Loss: 2.2676, Perplexity: 9.6564
Epoch [2/3], Step [134/3236], Loss: 2.2599, Perplexity: 9.5825
Epoch [2/3], Step [135/3236], Loss: 2.6537, Perplexity: 14.2064
Epoch [2/3], Step [136/3236], Loss: 2.3469, Perplexity: 10.4528
Epoch [2/3], Step [137/3236], Loss: 2.5686, Perplexity: 13.0470
Epoch [2/3], Step [138/3236], Loss: 2.5117, Perplexity: 12.3263
Epoch [2/3], Step [139/3236], Loss: 2.3256, Perplexity: 10.2330
Epoch [2/3], Step [140/3236], Loss: 2.2413, Perplexity: 9.4057
Epoch [2/3], Step [141/3236], Loss: 2.2094, Perplexity: 9.1099
Epoch [2/3], Step [142/3236], Loss: 2.0365, Perplexity: 7.6635
Epoch [2/3], Step [143/3236], Loss: 2.4480, Perplexity: 11.5654
Epoch [2/3], Step [144/3236], Loss: 2.2480, Perplexity: 9.4683
Epoch [2/3], Step [145/3236], Loss: 2.3340, Perplexity: 10.3188
Epoch [2/3], Step [146/3236], Loss: 2.1721, Perplexity: 8.7770
Epoch [2/3], Step [147/3236], Loss: 2.2848, Perplexity: 9.8234
Epoch [2/3], Step [148/3236], Loss: 2.2201, Perplexity: 9.2082
Epoch [2/3], Step [149/3236], Loss: 2.2415, Perplexity: 9.4071
Epoch [2/3], Step [150/3236], Loss: 2.2709, Perplexity: 9.6882
Epoch [2/3], Step [151/3236], Loss: 2.1543, Perplexity: 8.6219
Epoch [2/3], Step [152/3236], Loss: 2.1370, Perplexity: 8.4736
Epoch [2/3], Step [153/3236], Loss: 2.1787, Perplexity: 8.8351
Epoch [2/3], Step [154/3236], Loss: 2.2204, Perplexity: 9.2114
Epoch [2/3], Step [155/3236], Loss: 2.2553, Perplexity: 9.5378
Epoch [2/3], Step [156/3236], Loss: 2.1490, Perplexity: 8.5764
Epoch [2/3], Step [157/3236], Loss: 2.1617, Perplexity: 8.6861
Epoch [2/3], Step [158/3236], Loss: 2.3208, Perplexity: 10.1843
Epoch [2/3], Step [159/3236], Loss: 2.0334, Perplexity: 7.6397
Epoch [2/3], Step [160/3236], Loss: 2.3314, Perplexity: 10.2919
Epoch [2/3], Step [161/3236], Loss: 2.3761, Perplexity: 10.7627
Epoch [2/3], Step [162/3236], Loss: 2.1880, Perplexity: 8.9174
Epoch [2/3], Step [163/3236], Loss: 2.4809, Perplexity: 11.9516
Epoch [2/3], Step [164/3236], Loss: 2.0910, Perplexity: 8.0932
Epoch [2/3], Step [165/3236], Loss: 2.4116, Perplexity: 11.1516
Epoch [2/3], Step [166/3236], Loss: 2.2788, Perplexity: 9.7650
Epoch [2/3], Step [167/3236], Loss: 2.2482, Perplexity: 9.4705
Epoch [2/3], Step [168/3236], Loss: 2.2104, Perplexity: 9.1195
Epoch [2/3], Step [169/3236], Loss: 2.0341, Perplexity: 7.6452
Epoch [2/3], Step [170/3236], Loss: 2.1528, Perplexity: 8.6088
Epoch [2/3], Step [171/3236], Loss: 2.2366, Perplexity: 9.3616
Epoch [2/3], Step [172/3236], Loss: 2.1429, Perplexity: 8.5245
Epoch [2/3], Step [173/3236], Loss: 2.6308, Perplexity: 13.8853
Epoch [2/3], Step [174/3236], Loss: 2.0751, Perplexity: 7.9657
Epoch [2/3], Step [175/3236], Loss: 2.2639, Perplexity: 9.6203
Epoch [2/3], Step [176/3236], Loss: 2.2349, Perplexity: 9.3455
Epoch [2/3], Step [177/3236], Loss: 2.1836, Perplexity: 8.8785
Epoch [2/3], Step [178/3236], Loss: 2.2868, Perplexity: 9.8437
Epoch [2/3], Step [179/3236], Loss: 2.4888, Perplexity: 12.0470
Epoch [2/3], Step [180/3236], Loss: 2.0933, Perplexity: 8.1120
Epoch [2/3], Step [181/3236], Loss: 2.0752, Perplexity: 7.9662
Epoch [2/3], Step [182/3236], Loss: 2.2660, Perplexity: 9.6412
Epoch [2/3], Step [183/3236], Loss: 2.3140, Perplexity: 10.1148
Epoch [2/3], Step [184/3236], Loss: 2.5953, Perplexity: 13.3999
Epoch [2/3], Step [185/3236], Loss: 2.1856, Perplexity: 8.8961
Epoch [2/3], Step [186/3236], Loss: 2.2073, Perplexity: 9.0907
Epoch [2/3], Step [187/3236], Loss: 2.4003, Perplexity: 11.0270
Epoch [2/3], Step [188/3236], Loss: 2.5231, Perplexity: 12.4671
Epoch [2/3], Step [189/3236], Loss: 2.2310, Perplexity: 9.3088
Epoch [2/3], Step [190/3236], Loss: 2.6563, Perplexity: 14.2435
Epoch [2/3], Step [191/3236], Loss: 2.1824, Perplexity: 8.8677
Epoch [2/3], Step [192/3236], Loss: 2.2328, Perplexity: 9.3260
Epoch [2/3], Step [193/3236], Loss: 2.2157, Perplexity: 9.1676
Epoch [2/3], Step [194/3236], Loss: 2.2590, Perplexity: 9.5739
Epoch [2/3], Step [195/3236], Loss: 2.5389, Perplexity: 12.6660
Epoch [2/3], Step [196/3236], Loss: 2.4595, Perplexity: 11.6986
Epoch [2/3], Step [197/3236], Loss: 2.3318, Perplexity: 10.2969
Epoch [2/3], Step [198/3236], Loss: 2.3104, Perplexity: 10.0780
Epoch [2/3], Step [199/3236], Loss: 2.1704, Perplexity: 8.7614
Epoch [2/3], Step [200/3236], Loss: 2.3570, Perplexity: 10.5596
Epoch [2/3], Step [201/3236], Loss: 2.1989, Perplexity: 9.0147
Epoch [2/3], Step [202/3236], Loss: 2.3265, Perplexity: 10.2422
Epoch [2/3], Step [203/3236], Loss: 2.3517, Perplexity: 10.5038
Epoch [2/3], Step [204/3236], Loss: 2.3666, Perplexity: 10.6607
Epoch [2/3], Step [205/3236], Loss: 2.1762, Perplexity: 8.8124
Epoch [2/3], Step [206/3236], Loss: 2.2405, Perplexity: 9.3984
Epoch [2/3], Step [207/3236], Loss: 2.6946, Perplexity: 14.8001
Epoch [2/3], Step [208/3236], Loss: 2.1665, Perplexity: 8.7275
Epoch [2/3], Step [209/3236], Loss: 2.2604, Perplexity: 9.5868
Epoch [2/3], Step [210/3236], Loss: 2.4087, Perplexity: 11.1191
Epoch [2/3], Step [211/3236], Loss: 2.2833, Perplexity: 9.8089
Epoch [2/3], Step [212/3236], Loss: 2.3477, Perplexity: 10.4615
Epoch [2/3], Step [213/3236], Loss: 2.4843, Perplexity: 11.9928
Epoch [2/3], Step [214/3236], Loss: 2.1647, Perplexity: 8.7123
Epoch [2/3], Step [215/3236], Loss: 2.3914, Perplexity: 10.9284
Epoch [2/3], Step [216/3236], Loss: 2.2804, Perplexity: 9.7806
Epoch [2/3], Step [217/3236], Loss: 2.0992, Perplexity: 8.1595
Epoch [2/3], Step [218/3236], Loss: 2.3634, Perplexity: 10.6270
Epoch [2/3], Step [219/3236], Loss: 2.0901, Perplexity: 8.0858
Epoch [2/3], Step [220/3236], Loss: 2.1051, Perplexity: 8.2080
Epoch [2/3], Step [221/3236], Loss: 2.0120, Perplexity: 7.4783
Epoch [2/3], Step [222/3236], Loss: 2.2613, Perplexity: 9.5957
Epoch [2/3], Step [223/3236], Loss: 2.2038, Perplexity: 9.0590
Epoch [2/3], Step [224/3236], Loss: 2.4375, Perplexity: 11.4441
Epoch [2/3], Step [225/3236], Loss: 2.0707, Perplexity: 7.9303
Epoch [2/3], Step [226/3236], Loss: 2.2290, Perplexity: 9.2903
Epoch [2/3], Step [227/3236], Loss: 2.2032, Perplexity: 9.0540
Epoch [2/3], Step [228/3236], Loss: 2.1657, Perplexity: 8.7206
Epoch [2/3], Step [229/3236], Loss: 2.3042, Perplexity: 10.0160
Epoch [2/3], Step [230/3236], Loss: 2.2661, Perplexity: 9.6416
Epoch [2/3], Step [231/3236], Loss: 3.3307, Perplexity: 27.9588
Epoch [2/3], Step [232/3236], Loss: 2.3431, Perplexity: 10.4134
Epoch [2/3], Step [233/3236], Loss: 2.4119, Perplexity: 11.1549
Epoch [2/3], Step [234/3236], Loss: 2.1775, Perplexity: 8.8239
Epoch [2/3], Step [235/3236], Loss: 2.0720, Perplexity: 7.9404
Epoch [2/3], Step [236/3236], Loss: 2.2173, Perplexity: 9.1828
Epoch [2/3], Step [237/3236], Loss: 2.4443, Perplexity: 11.5230
Epoch [2/3], Step [238/3236], Loss: 2.5365, Perplexity: 12.6358
Epoch [2/3], Step [239/3236], Loss: 2.0792, Perplexity: 7.9984
Epoch [2/3], Step [240/3236], Loss: 2.2021, Perplexity: 9.0443
Epoch [2/3], Step [241/3236], Loss: 2.1619, Perplexity: 8.6874
Epoch [2/3], Step [242/3236], Loss: 2.1968, Perplexity: 8.9963
Epoch [2/3], Step [243/3236], Loss: 2.2965, Perplexity: 9.9395
Epoch [2/3], Step [244/3236], Loss: 2.2514, Perplexity: 9.5010
Epoch [2/3], Step [245/3236], Loss: 2.3797, Perplexity: 10.8012
Epoch [2/3], Step [246/3236], Loss: 2.7030, Perplexity: 14.9242
Epoch [2/3], Step [247/3236], Loss: 3.0295, Perplexity: 20.6877
Epoch [2/3], Step [248/3236], Loss: 2.1505, Perplexity: 8.5889
Epoch [2/3], Step [249/3236], Loss: 2.3180, Perplexity: 10.1550
Epoch [2/3], Step [250/3236], Loss: 2.2500, Perplexity: 9.4881
Epoch [2/3], Step [251/3236], Loss: 2.5267, Perplexity: 12.5127
Epoch [2/3], Step [252/3236], Loss: 2.3453, Perplexity: 10.4365
Epoch [2/3], Step [253/3236], Loss: 2.2037, Perplexity: 9.0587
Epoch [2/3], Step [254/3236], Loss: 2.1921, Perplexity: 8.9538
Epoch [2/3], Step [255/3236], Loss: 2.1541, Perplexity: 8.6203
Epoch [2/3], Step [256/3236], Loss: 2.2078, Perplexity: 9.0953
Epoch [2/3], Step [257/3236], Loss: 2.2526, Perplexity: 9.5127
Epoch [2/3], Step [258/3236], Loss: 2.1284, Perplexity: 8.4014
Epoch [2/3], Step [259/3236], Loss: 2.3403, Perplexity: 10.3841
Epoch [2/3], Step [260/3236], Loss: 2.1891, Perplexity: 8.9270
Epoch [2/3], Step [261/3236], Loss: 2.1180, Perplexity: 8.3148
Epoch [2/3], Step [262/3236], Loss: 2.4041, Perplexity: 11.0683
Epoch [2/3], Step [263/3236], Loss: 2.3253, Perplexity: 10.2296
Epoch [2/3], Step [264/3236], Loss: 2.2850, Perplexity: 9.8254
Epoch [2/3], Step [265/3236], Loss: 2.3428, Perplexity: 10.4105
Epoch [2/3], Step [266/3236], Loss: 2.2801, Perplexity: 9.7777
Epoch [2/3], Step [267/3236], Loss: 2.3761, Perplexity: 10.7629
Epoch [2/3], Step [268/3236], Loss: 2.5320, Perplexity: 12.5791
Epoch [2/3], Step [269/3236], Loss: 2.1459, Perplexity: 8.5500
Epoch [2/3], Step [270/3236], Loss: 2.4030, Perplexity: 11.0563
Epoch [2/3], Step [271/3236], Loss: 2.1211, Perplexity: 8.3402
Epoch [2/3], Step [272/3236], Loss: 2.6694, Perplexity: 14.4316
Epoch [2/3], Step [273/3236], Loss: 2.1341, Perplexity: 8.4493
Epoch [2/3], Step [274/3236], Loss: 2.1770, Perplexity: 8.8200
Epoch [2/3], Step [275/3236], Loss: 2.2514, Perplexity: 9.5011
Epoch [2/3], Step [276/3236], Loss: 2.4476, Perplexity: 11.5605
Epoch [2/3], Step [277/3236], Loss: 2.4485, Perplexity: 11.5715
Epoch [2/3], Step [278/3236], Loss: 2.2147, Perplexity: 9.1583
Epoch [2/3], Step [279/3236], Loss: 2.1295, Perplexity: 8.4107
Epoch [2/3], Step [280/3236], Loss: 2.6076, Perplexity: 13.5671
Epoch [2/3], Step [281/3236], Loss: 2.5016, Perplexity: 12.2016
Epoch [2/3], Step [282/3236], Loss: 2.3195, Perplexity: 10.1707
Epoch [2/3], Step [283/3236], Loss: 3.0197, Perplexity: 20.4845
Epoch [2/3], Step [284/3236], Loss: 2.1398, Perplexity: 8.4974
Epoch [2/3], Step [285/3236], Loss: 2.4177, Perplexity: 11.2203
Epoch [2/3], Step [286/3236], Loss: 2.9364, Perplexity: 18.8484
Epoch [2/3], Step [287/3236], Loss: 2.4117, Perplexity: 11.1531
Epoch [2/3], Step [288/3236], Loss: 2.4055, Perplexity: 11.0842
Epoch [2/3], Step [289/3236], Loss: 2.2426, Perplexity: 9.4182
Epoch [2/3], Step [290/3236], Loss: 2.3700, Perplexity: 10.6979
Epoch [2/3], Step [291/3236], Loss: 2.0747, Perplexity: 7.9620
Epoch [2/3], Step [292/3236], Loss: 2.3733, Perplexity: 10.7328
Epoch [2/3], Step [293/3236], Loss: 2.5175, Perplexity: 12.3981
Epoch [2/3], Step [294/3236], Loss: 2.3142, Perplexity: 10.1170
Epoch [2/3], Step [295/3236], Loss: 2.1394, Perplexity: 8.4947
Epoch [2/3], Step [296/3236], Loss: 2.3056, Perplexity: 10.0301
Epoch [2/3], Step [297/3236], Loss: 2.3946, Perplexity: 10.9643
Epoch [2/3], Step [298/3236], Loss: 2.2322, Perplexity: 9.3208
Epoch [2/3], Step [299/3236], Loss: 2.2086, Perplexity: 9.1026
Epoch [2/3], Step [300/3236], Loss: 2.4123, Perplexity: 11.1596
Epoch [2/3], Step [301/3236], Loss: 2.2381, Perplexity: 9.3758
Epoch [2/3], Step [302/3236], Loss: 2.3341, Perplexity: 10.3201
Epoch [2/3], Step [303/3236], Loss: 2.2440, Perplexity: 9.4309
Epoch [2/3], Step [304/3236], Loss: 2.2626, Perplexity: 9.6082
Epoch [2/3], Step [305/3236], Loss: 2.0650, Perplexity: 7.8855
Epoch [2/3], Step [306/3236], Loss: 2.3091, Perplexity: 10.0657
Epoch [2/3], Step [307/3236], Loss: 2.2104, Perplexity: 9.1197
Epoch [2/3], Step [308/3236], Loss: 2.2623, Perplexity: 9.6050
Epoch [2/3], Step [309/3236], Loss: 2.2663, Perplexity: 9.6432
Epoch [2/3], Step [310/3236], Loss: 2.1650, Perplexity: 8.7143
Epoch [2/3], Step [311/3236], Loss: 2.5847, Perplexity: 13.2593
Epoch [2/3], Step [312/3236], Loss: 2.1692, Perplexity: 8.7515
Epoch [2/3], Step [313/3236], Loss: 2.1615, Perplexity: 8.6839
Epoch [2/3], Step [314/3236], Loss: 2.1150, Perplexity: 8.2899
Epoch [2/3], Step [315/3236], Loss: 2.0739, Perplexity: 7.9556
Epoch [2/3], Step [316/3236], Loss: 2.5511, Perplexity: 12.8211
Epoch [2/3], Step [317/3236], Loss: 2.2481, Perplexity: 9.4700
Epoch [2/3], Step [318/3236], Loss: 2.1426, Perplexity: 8.5217
Epoch [2/3], Step [319/3236], Loss: 2.1902, Perplexity: 8.9372
Epoch [2/3], Step [320/3236], Loss: 2.4283, Perplexity: 11.3399
Epoch [2/3], Step [321/3236], Loss: 2.2298, Perplexity: 9.2978
Epoch [2/3], Step [322/3236], Loss: 2.4074, Perplexity: 11.1052
Epoch [2/3], Step [323/3236], Loss: 2.2962, Perplexity: 9.9366
Epoch [2/3], Step [324/3236], Loss: 2.1579, Perplexity: 8.6531
Epoch [2/3], Step [325/3236], Loss: 2.2228, Perplexity: 9.2329
Epoch [2/3], Step [326/3236], Loss: 2.1814, Perplexity: 8.8584
Epoch [2/3], Step [327/3236], Loss: 2.5506, Perplexity: 12.8153
Epoch [2/3], Step [328/3236], Loss: 2.0577, Perplexity: 7.8277
Epoch [2/3], Step [329/3236], Loss: 2.2296, Perplexity: 9.2957
Epoch [2/3], Step [330/3236], Loss: 2.2131, Perplexity: 9.1437
Epoch [2/3], Step [331/3236], Loss: 2.5342, Perplexity: 12.6067
Epoch [2/3], Step [332/3236], Loss: 2.2153, Perplexity: 9.1638
Epoch [2/3], Step [333/3236], Loss: 2.0894, Perplexity: 8.0799
Epoch [2/3], Step [334/3236], Loss: 2.1446, Perplexity: 8.5385
Epoch [2/3], Step [335/3236], Loss: 2.1504, Perplexity: 8.5884
Epoch [2/3], Step [336/3236], Loss: 2.1817, Perplexity: 8.8613
Epoch [2/3], Step [337/3236], Loss: 2.3178, Perplexity: 10.1529
Epoch [2/3], Step [338/3236], Loss: 2.5769, Perplexity: 13.1566
Epoch [2/3], Step [339/3236], Loss: 2.5026, Perplexity: 12.2145
Epoch [2/3], Step [340/3236], Loss: 2.0519, Perplexity: 7.7826
Epoch [2/3], Step [341/3236], Loss: 2.3508, Perplexity: 10.4943
Epoch [2/3], Step [342/3236], Loss: 2.2975, Perplexity: 9.9495
Epoch [2/3], Step [343/3236], Loss: 2.1452, Perplexity: 8.5438
Epoch [2/3], Step [344/3236], Loss: 2.1800, Perplexity: 8.8461
Epoch [2/3], Step [345/3236], Loss: 2.2060, Perplexity: 9.0797
Epoch [2/3], Step [346/3236], Loss: 2.3070, Perplexity: 10.0441
Epoch [2/3], Step [347/3236], Loss: 2.1681, Perplexity: 8.7416
Epoch [2/3], Step [348/3236], Loss: 2.1322, Perplexity: 8.4333
Epoch [2/3], Step [349/3236], Loss: 2.1492, Perplexity: 8.5776
Epoch [2/3], Step [350/3236], Loss: 2.2095, Perplexity: 9.1115
Epoch [2/3], Step [351/3236], Loss: 2.1226, Perplexity: 8.3527
Epoch [2/3], Step [352/3236], Loss: 2.1014, Perplexity: 8.1774
Epoch [2/3], Step [353/3236], Loss: 2.1137, Perplexity: 8.2784
Epoch [2/3], Step [354/3236], Loss: 2.5816, Perplexity: 13.2186
Epoch [2/3], Step [355/3236], Loss: 2.1279, Perplexity: 8.3968
Epoch [2/3], Step [356/3236], Loss: 2.2060, Perplexity: 9.0790
Epoch [2/3], Step [357/3236], Loss: 2.2068, Perplexity: 9.0862
Epoch [2/3], Step [358/3236], Loss: 2.3040, Perplexity: 10.0137
Epoch [2/3], Step [359/3236], Loss: 2.0077, Perplexity: 7.4464
Epoch [2/3], Step [360/3236], Loss: 2.2735, Perplexity: 9.7138
Epoch [2/3], Step [361/3236], Loss: 2.0354, Perplexity: 7.6552
Epoch [2/3], Step [362/3236], Loss: 2.5555, Perplexity: 12.8772
Epoch [2/3], Step [363/3236], Loss: 2.4189, Perplexity: 11.2340
Epoch [2/3], Step [364/3236], Loss: 2.7775, Perplexity: 16.0785
Epoch [2/3], Step [365/3236], Loss: 2.3275, Perplexity: 10.2527
Epoch [2/3], Step [366/3236], Loss: 2.3245, Perplexity: 10.2211
Epoch [2/3], Step [367/3236], Loss: 2.1473, Perplexity: 8.5621
Epoch [2/3], Step [368/3236], Loss: 2.3174, Perplexity: 10.1492
Epoch [2/3], Step [369/3236], Loss: 2.3139, Perplexity: 10.1143
Epoch [2/3], Step [370/3236], Loss: 2.6008, Perplexity: 13.4746
Epoch [2/3], Step [371/3236], Loss: 2.2541, Perplexity: 9.5272
Epoch [2/3], Step [372/3236], Loss: 2.4658, Perplexity: 11.7733
Epoch [2/3], Step [373/3236], Loss: 2.1544, Perplexity: 8.6226
Epoch [2/3], Step [374/3236], Loss: 2.2406, Perplexity: 9.3994
Epoch [2/3], Step [375/3236], Loss: 2.0879, Perplexity: 8.0676
Epoch [2/3], Step [376/3236], Loss: 2.6807, Perplexity: 14.5949
Epoch [2/3], Step [377/3236], Loss: 2.2793, Perplexity: 9.7700
Epoch [2/3], Step [378/3236], Loss: 2.2711, Perplexity: 9.6900
Epoch [2/3], Step [379/3236], Loss: 3.1879, Perplexity: 24.2371
Epoch [2/3], Step [380/3236], Loss: 2.4168, Perplexity: 11.2095
Epoch [2/3], Step [381/3236], Loss: 2.2209, Perplexity: 9.2154
Epoch [2/3], Step [382/3236], Loss: 2.1869, Perplexity: 8.9077
Epoch [2/3], Step [383/3236], Loss: 2.1152, Perplexity: 8.2910
Epoch [2/3], Step [384/3236], Loss: 2.3144, Perplexity: 10.1184
Epoch [2/3], Step [385/3236], Loss: 2.2777, Perplexity: 9.7545
Epoch [2/3], Step [386/3236], Loss: 2.0864, Perplexity: 8.0559
Epoch [2/3], Step [387/3236], Loss: 2.0540, Perplexity: 7.7987
Epoch [2/3], Step [388/3236], Loss: 2.6283, Perplexity: 13.8498
Epoch [2/3], Step [389/3236], Loss: 2.5608, Perplexity: 12.9464
Epoch [2/3], Step [390/3236], Loss: 2.3391, Perplexity: 10.3714
Epoch [2/3], Step [391/3236], Loss: 2.2098, Perplexity: 9.1143
Epoch [2/3], Step [392/3236], Loss: 2.0665, Perplexity: 7.8971
Epoch [2/3], Step [393/3236], Loss: 2.2861, Perplexity: 9.8363
Epoch [2/3], Step [394/3236], Loss: 2.4242, Perplexity: 11.2934
Epoch [2/3], Step [395/3236], Loss: 2.1262, Perplexity: 8.3825
Epoch [2/3], Step [396/3236], Loss: 2.1350, Perplexity: 8.4573
Epoch [2/3], Step [397/3236], Loss: 2.2029, Perplexity: 9.0510
Epoch [2/3], Step [398/3236], Loss: 2.3251, Perplexity: 10.2276
Epoch [2/3], Step [399/3236], Loss: 2.2867, Perplexity: 9.8427
Epoch [2/3], Step [400/3236], Loss: 2.0517, Perplexity: 7.7812
Epoch [2/3], Step [401/3236], Loss: 2.1322, Perplexity: 8.4330
Epoch [2/3], Step [402/3236], Loss: 2.0645, Perplexity: 7.8814
Epoch [2/3], Step [403/3236], Loss: 2.3892, Perplexity: 10.9050
Epoch [2/3], Step [404/3236], Loss: 2.2479, Perplexity: 9.4674
Epoch [2/3], Step [405/3236], Loss: 2.2630, Perplexity: 9.6122
Epoch [2/3], Step [406/3236], Loss: 2.8060, Perplexity: 16.5439
Epoch [2/3], Step [407/3236], Loss: 2.2047, Perplexity: 9.0680
Epoch [2/3], Step [408/3236], Loss: 2.1309, Perplexity: 8.4221
Epoch [2/3], Step [409/3236], Loss: 2.2898, Perplexity: 9.8727
Epoch [2/3], Step [410/3236], Loss: 2.2847, Perplexity: 9.8228
Epoch [2/3], Step [411/3236], Loss: 2.0253, Perplexity: 7.5781
Epoch [2/3], Step [412/3236], Loss: 2.1828, Perplexity: 8.8708
Epoch [2/3], Step [413/3236], Loss: 2.2023, Perplexity: 9.0462
Epoch [2/3], Step [414/3236], Loss: 2.2990, Perplexity: 9.9640
Epoch [2/3], Step [415/3236], Loss: 2.1457, Perplexity: 8.5483
Epoch [2/3], Step [416/3236], Loss: 2.3160, Perplexity: 10.1351
Epoch [2/3], Step [417/3236], Loss: 2.3927, Perplexity: 10.9427
Epoch [2/3], Step [418/3236], Loss: 2.1954, Perplexity: 8.9832
Epoch [2/3], Step [419/3236], Loss: 2.1994, Perplexity: 9.0196
Epoch [2/3], Step [420/3236], Loss: 2.0846, Perplexity: 8.0411
Epoch [2/3], Step [421/3236], Loss: 2.0900, Perplexity: 8.0852
Epoch [2/3], Step [422/3236], Loss: 2.0664, Perplexity: 7.8967
Epoch [2/3], Step [423/3236], Loss: 2.3302, Perplexity: 10.2799
Epoch [2/3], Step [424/3236], Loss: 2.2745, Perplexity: 9.7226
Epoch [2/3], Step [425/3236], Loss: 2.3943, Perplexity: 10.9606
Epoch [2/3], Step [426/3236], Loss: 2.1254, Perplexity: 8.3761
Epoch [2/3], Step [427/3236], Loss: 2.3849, Perplexity: 10.8575
Epoch [2/3], Step [428/3236], Loss: 2.2020, Perplexity: 9.0426
Epoch [2/3], Step [429/3236], Loss: 2.2301, Perplexity: 9.3012
Epoch [2/3], Step [430/3236], Loss: 2.2395, Perplexity: 9.3887
Epoch [2/3], Step [431/3236], Loss: 2.2562, Perplexity: 9.5471
Epoch [2/3], Step [432/3236], Loss: 2.2035, Perplexity: 9.0562
Epoch [2/3], Step [433/3236], Loss: 2.1694, Perplexity: 8.7534
Epoch [2/3], Step [434/3236], Loss: 2.3051, Perplexity: 10.0250
Epoch [2/3], Step [435/3236], Loss: 2.2727, Perplexity: 9.7054
Epoch [2/3], Step [436/3236], Loss: 2.0782, Perplexity: 7.9898
Epoch [2/3], Step [437/3236], Loss: 2.1381, Perplexity: 8.4836
Epoch [2/3], Step [438/3236], Loss: 2.2486, Perplexity: 9.4741
Epoch [2/3], Step [439/3236], Loss: 2.2633, Perplexity: 9.6143
Epoch [2/3], Step [440/3236], Loss: 2.1243, Perplexity: 8.3672
Epoch [2/3], Step [441/3236], Loss: 2.3581, Perplexity: 10.5714
Epoch [2/3], Step [442/3236], Loss: 2.0969, Perplexity: 8.1408
Epoch [2/3], Step [443/3236], Loss: 2.5316, Perplexity: 12.5736
Epoch [2/3], Step [444/3236], Loss: 2.3776, Perplexity: 10.7790
Epoch [2/3], Step [445/3236], Loss: 2.1542, Perplexity: 8.6210
Epoch [2/3], Step [446/3236], Loss: 2.1684, Perplexity: 8.7445
Epoch [2/3], Step [447/3236], Loss: 2.8787, Perplexity: 17.7913
Epoch [2/3], Step [448/3236], Loss: 2.1713, Perplexity: 8.7694
Epoch [2/3], Step [449/3236], Loss: 2.1707, Perplexity: 8.7642
Epoch [2/3], Step [450/3236], Loss: 2.2109, Perplexity: 9.1240
Epoch [2/3], Step [451/3236], Loss: 2.1518, Perplexity: 8.6001
Epoch [2/3], Step [452/3236], Loss: 2.2866, Perplexity: 9.8412
Epoch [2/3], Step [453/3236], Loss: 2.2505, Perplexity: 9.4928
Epoch [2/3], Step [454/3236], Loss: 2.2743, Perplexity: 9.7216
Epoch [2/3], Step [455/3236], Loss: 2.6297, Perplexity: 13.8689
Epoch [2/3], Step [456/3236], Loss: 2.1631, Perplexity: 8.6980
Epoch [2/3], Step [457/3236], Loss: 2.1798, Perplexity: 8.8450
Epoch [2/3], Step [458/3236], Loss: 2.1142, Perplexity: 8.2831
Epoch [2/3], Step [459/3236], Loss: 2.3109, Perplexity: 10.0834
Epoch [2/3], Step [460/3236], Loss: 2.7285, Perplexity: 15.3106
Epoch [2/3], Step [461/3236], Loss: 2.2207, Perplexity: 9.2139
Epoch [2/3], Step [462/3236], Loss: 2.1554, Perplexity: 8.6310
Epoch [2/3], Step [463/3236], Loss: 4.6363, Perplexity: 103.1620
Epoch [2/3], Step [464/3236], Loss: 2.7487, Perplexity: 15.6221
Epoch [2/3], Step [465/3236], Loss: 2.1220, Perplexity: 8.3474
Epoch [2/3], Step [466/3236], Loss: 2.3067, Perplexity: 10.0411
Epoch [2/3], Step [467/3236], Loss: 2.1302, Perplexity: 8.4164
Epoch [2/3], Step [468/3236], Loss: 2.2915, Perplexity: 9.8894
Epoch [2/3], Step [469/3236], Loss: 2.3070, Perplexity: 10.0439
Epoch [2/3], Step [470/3236], Loss: 2.1452, Perplexity: 8.5439
Epoch [2/3], Step [471/3236], Loss: 2.1944, Perplexity: 8.9745
Epoch [2/3], Step [472/3236], Loss: 2.1231, Perplexity: 8.3571
Epoch [2/3], Step [473/3236], Loss: 2.2305, Perplexity: 9.3048
Epoch [2/3], Step [474/3236], Loss: 2.5186, Perplexity: 12.4112
Epoch [2/3], Step [475/3236], Loss: 2.1270, Perplexity: 8.3899
Epoch [2/3], Step [476/3236], Loss: 2.3405, Perplexity: 10.3864
Epoch [2/3], Step [477/3236], Loss: 2.1924, Perplexity: 8.9563
Epoch [2/3], Step [478/3236], Loss: 2.8747, Perplexity: 17.7200
Epoch [2/3], Step [479/3236], Loss: 2.7374, Perplexity: 15.4468
Epoch [2/3], Step [480/3236], Loss: 2.2329, Perplexity: 9.3273
Epoch [2/3], Step [481/3236], Loss: 2.3115, Perplexity: 10.0891
Epoch [2/3], Step [482/3236], Loss: 2.6229, Perplexity: 13.7754
Epoch [2/3], Step [483/3236], Loss: 2.2104, Perplexity: 9.1191
Epoch [2/3], Step [484/3236], Loss: 2.2212, Perplexity: 9.2180
Epoch [2/3], Step [485/3236], Loss: 2.2129, Perplexity: 9.1423
Epoch [2/3], Step [486/3236], Loss: 2.2035, Perplexity: 9.0571
Epoch [2/3], Step [487/3236], Loss: 2.4811, Perplexity: 11.9546
Epoch [2/3], Step [488/3236], Loss: 2.1611, Perplexity: 8.6810
Epoch [2/3], Step [489/3236], Loss: 2.3152, Perplexity: 10.1274
Epoch [2/3], Step [490/3236], Loss: 2.0959, Perplexity: 8.1327
Epoch [2/3], Step [491/3236], Loss: 2.6454, Perplexity: 14.0891
Epoch [2/3], Step [492/3236], Loss: 2.2015, Perplexity: 9.0383
Epoch [2/3], Step [493/3236], Loss: 2.1157, Perplexity: 8.2956
Epoch [2/3], Step [494/3236], Loss: 2.4577, Perplexity: 11.6780
Epoch [2/3], Step [495/3236], Loss: 2.2454, Perplexity: 9.4446
Epoch [2/3], Step [496/3236], Loss: 2.4064, Perplexity: 11.0941
Epoch [2/3], Step [497/3236], Loss: 2.1869, Perplexity: 8.9072
Epoch [2/3], Step [498/3236], Loss: 2.2611, Perplexity: 9.5935
Epoch [2/3], Step [499/3236], Loss: 2.7747, Perplexity: 16.0341
Epoch [2/3], Step [500/3236], Loss: 2.1903, Perplexity: 8.9376
Epoch [2/3], Step [501/3236], Loss: 2.1887, Perplexity: 8.9234
Epoch [2/3], Step [502/3236], Loss: 2.2861, Perplexity: 9.8366
Epoch [2/3], Step [503/3236], Loss: 2.2989, Perplexity: 9.9633
Epoch [2/3], Step [504/3236], Loss: 1.9797, Perplexity: 7.2406
Epoch [2/3], Step [505/3236], Loss: 2.5881, Perplexity: 13.3042
Epoch [2/3], Step [506/3236], Loss: 2.1481, Perplexity: 8.5689
Epoch [2/3], Step [507/3236], Loss: 2.2841, Perplexity: 9.8168
Epoch [2/3], Step [508/3236], Loss: 2.1086, Perplexity: 8.2371
Epoch [2/3], Step [509/3236], Loss: 2.3482, Perplexity: 10.4671
Epoch [2/3], Step [510/3236], Loss: 2.1214, Perplexity: 8.3428
Epoch [2/3], Step [511/3236], Loss: 2.1669, Perplexity: 8.7307
Epoch [2/3], Step [512/3236], Loss: 2.1288, Perplexity: 8.4050
Epoch [2/3], Step [513/3236], Loss: 2.1535, Perplexity: 8.6150
Epoch [2/3], Step [514/3236], Loss: 2.5536, Perplexity: 12.8530
Epoch [2/3], Step [515/3236], Loss: 2.2601, Perplexity: 9.5839
Epoch [2/3], Step [516/3236], Loss: 2.2431, Perplexity: 9.4225
Epoch [2/3], Step [517/3236], Loss: 2.5007, Perplexity: 12.1910
Epoch [2/3], Step [518/3236], Loss: 3.2480, Perplexity: 25.7400
Epoch [2/3], Step [519/3236], Loss: 2.5391, Perplexity: 12.6685
Epoch [2/3], Step [520/3236], Loss: 2.1266, Perplexity: 8.3865
Epoch [2/3], Step [521/3236], Loss: 2.4343, Perplexity: 11.4075
Epoch [2/3], Step [522/3236], Loss: 2.0951, Perplexity: 8.1263
Epoch [2/3], Step [523/3236], Loss: 3.0822, Perplexity: 21.8060
Epoch [2/3], Step [524/3236], Loss: 2.0545, Perplexity: 7.8026
Epoch [2/3], Step [525/3236], Loss: 2.0945, Perplexity: 8.1214
Epoch [2/3], Step [526/3236], Loss: 2.0856, Perplexity: 8.0497
Epoch [2/3], Step [527/3236], Loss: 2.2909, Perplexity: 9.8840
Epoch [2/3], Step [528/3236], Loss: 2.4840, Perplexity: 11.9897
Epoch [2/3], Step [529/3236], Loss: 2.1481, Perplexity: 8.5683
Epoch [2/3], Step [530/3236], Loss: 2.1804, Perplexity: 8.8495
Epoch [2/3], Step [531/3236], Loss: 2.3702, Perplexity: 10.6997
Epoch [2/3], Step [532/3236], Loss: 2.1632, Perplexity: 8.6993
Epoch [2/3], Step [533/3236], Loss: 2.0877, Perplexity: 8.0663
Epoch [2/3], Step [534/3236], Loss: 2.1515, Perplexity: 8.5980
Epoch [2/3], Step [535/3236], Loss: 2.1092, Perplexity: 8.2420
Epoch [2/3], Step [536/3236], Loss: 2.5998, Perplexity: 13.4609
Epoch [2/3], Step [537/3236], Loss: 1.9664, Perplexity: 7.1447
Epoch [2/3], Step [538/3236], Loss: 2.9803, Perplexity: 19.6933
Epoch [2/3], Step [539/3236], Loss: 2.2275, Perplexity: 9.2770
Epoch [2/3], Step [540/3236], Loss: 2.1870, Perplexity: 8.9085
Epoch [2/3], Step [541/3236], Loss: 2.1536, Perplexity: 8.6160
Epoch [2/3], Step [542/3236], Loss: 2.1361, Perplexity: 8.4667
Epoch [2/3], Step [543/3236], Loss: 2.1551, Perplexity: 8.6287
Epoch [2/3], Step [544/3236], Loss: 2.2308, Perplexity: 9.3069
Epoch [2/3], Step [545/3236], Loss: 2.1745, Perplexity: 8.7980
Epoch [2/3], Step [546/3236], Loss: 2.0920, Perplexity: 8.1013
Epoch [2/3], Step [547/3236], Loss: 2.3379, Perplexity: 10.3591
Epoch [2/3], Step [548/3236], Loss: 2.7633, Perplexity: 15.8524
Epoch [2/3], Step [549/3236], Loss: 2.2837, Perplexity: 9.8126
Epoch [2/3], Step [550/3236], Loss: 2.4436, Perplexity: 11.5141
Epoch [2/3], Step [551/3236], Loss: 2.1600, Perplexity: 8.6709
Epoch [2/3], Step [552/3236], Loss: 2.1252, Perplexity: 8.3748
Epoch [2/3], Step [553/3236], Loss: 2.1880, Perplexity: 8.9171
Epoch [2/3], Step [554/3236], Loss: 2.3261, Perplexity: 10.2381
Epoch [2/3], Step [555/3236], Loss: 2.1784, Perplexity: 8.8321
Epoch [2/3], Step [556/3236], Loss: 2.1278, Perplexity: 8.3960
Epoch [2/3], Step [557/3236], Loss: 2.2363, Perplexity: 9.3584
Epoch [2/3], Step [558/3236], Loss: 2.3361, Perplexity: 10.3411
Epoch [2/3], Step [559/3236], Loss: 2.1622, Perplexity: 8.6903
Epoch [2/3], Step [560/3236], Loss: 2.0974, Perplexity: 8.1453
Epoch [2/3], Step [561/3236], Loss: 2.2619, Perplexity: 9.6011
Epoch [2/3], Step [562/3236], Loss: 2.3408, Perplexity: 10.3896
Epoch [2/3], Step [563/3236], Loss: 2.4632, Perplexity: 11.7421
Epoch [2/3], Step [564/3236], Loss: 2.3294, Perplexity: 10.2716
Epoch [2/3], Step [565/3236], Loss: 2.1486, Perplexity: 8.5729
Epoch [2/3], Step [566/3236], Loss: 2.2279, Perplexity: 9.2801
Epoch [2/3], Step [567/3236], Loss: 2.3652, Perplexity: 10.6464
Epoch [2/3], Step [568/3236], Loss: 2.3066, Perplexity: 10.0398
Epoch [2/3], Step [569/3236], Loss: 2.4790, Perplexity: 11.9294
Epoch [2/3], Step [570/3236], Loss: 2.6904, Perplexity: 14.7374
Epoch [2/3], Step [571/3236], Loss: 2.3429, Perplexity: 10.4109
Epoch [2/3], Step [572/3236], Loss: 2.0931, Perplexity: 8.1098
Epoch [2/3], Step [573/3236], Loss: 3.0330, Perplexity: 20.7597
Epoch [2/3], Step [574/3236], Loss: 2.2812, Perplexity: 9.7886
Epoch [2/3], Step [575/3236], Loss: 2.1212, Perplexity: 8.3415
Epoch [2/3], Step [576/3236], Loss: 2.0944, Perplexity: 8.1210
Epoch [2/3], Step [577/3236], Loss: 2.4258, Perplexity: 11.3109
Epoch [2/3], Step [578/3236], Loss: 2.1907, Perplexity: 8.9416
Epoch [2/3], Step [579/3236], Loss: 2.7788, Perplexity: 16.0994
Epoch [2/3], Step [580/3236], Loss: 2.1925, Perplexity: 8.9573
Epoch [2/3], Step [581/3236], Loss: 2.1486, Perplexity: 8.5728
Epoch [2/3], Step [582/3236], Loss: 2.9934, Perplexity: 19.9543
Epoch [2/3], Step [583/3236], Loss: 2.2509, Perplexity: 9.4966
Epoch [2/3], Step [584/3236], Loss: 2.2405, Perplexity: 9.3978
Epoch [2/3], Step [585/3236], Loss: 2.1896, Perplexity: 8.9321
Epoch [2/3], Step [586/3236], Loss: 2.1823, Perplexity: 8.8667
Epoch [2/3], Step [587/3236], Loss: 2.1021, Perplexity: 8.1830
Epoch [2/3], Step [588/3236], Loss: 2.1626, Perplexity: 8.6939
Epoch [2/3], Step [589/3236], Loss: 2.0804, Perplexity: 8.0074
Epoch [2/3], Step [590/3236], Loss: 2.3282, Perplexity: 10.2598
Epoch [2/3], Step [591/3236], Loss: 2.2492, Perplexity: 9.4798
Epoch [2/3], Step [592/3236], Loss: 2.2656, Perplexity: 9.6370
Epoch [2/3], Step [593/3236], Loss: 2.3360, Perplexity: 10.3396
Epoch [2/3], Step [594/3236], Loss: 2.3119, Perplexity: 10.0941
Epoch [2/3], Step [595/3236], Loss: 2.1227, Perplexity: 8.3541
Epoch [2/3], Step [596/3236], Loss: 2.5666, Perplexity: 13.0217
Epoch [2/3], Step [597/3236], Loss: 2.1852, Perplexity: 8.8924
Epoch [2/3], Step [598/3236], Loss: 2.1620, Perplexity: 8.6889
Epoch [2/3], Step [599/3236], Loss: 2.1510, Perplexity: 8.5938
Epoch [2/3], Step [600/3236], Loss: 2.2419, Perplexity: 9.4116
Epoch [2/3], Step [601/3236], Loss: 2.0640, Perplexity: 7.8774
Epoch [2/3], Step [602/3236], Loss: 2.2751, Perplexity: 9.7292
Epoch [2/3], Step [603/3236], Loss: 2.1448, Perplexity: 8.5403
Epoch [2/3], Step [604/3236], Loss: 2.1841, Perplexity: 8.8826
Epoch [2/3], Step [605/3236], Loss: 2.4287, Perplexity: 11.3438
Epoch [2/3], Step [606/3236], Loss: 3.3727, Perplexity: 29.1582
Epoch [2/3], Step [607/3236], Loss: 2.4512, Perplexity: 11.6028
Epoch [2/3], Step [608/3236], Loss: 2.2433, Perplexity: 9.4243
Epoch [2/3], Step [609/3236], Loss: 2.2251, Perplexity: 9.2540
Epoch [2/3], Step [610/3236], Loss: 2.1017, Perplexity: 8.1802
Epoch [2/3], Step [611/3236], Loss: 2.5115, Perplexity: 12.3235
Epoch [2/3], Step [612/3236], Loss: 2.2184, Perplexity: 9.1928
Epoch [2/3], Step [613/3236], Loss: 2.1691, Perplexity: 8.7501
Epoch [2/3], Step [614/3236], Loss: 2.1603, Perplexity: 8.6740
Epoch [2/3], Step [615/3236], Loss: 2.1797, Perplexity: 8.8441
Epoch [2/3], Step [616/3236], Loss: 2.2006, Perplexity: 9.0303
Epoch [2/3], Step [617/3236], Loss: 2.1668, Perplexity: 8.7305
Epoch [2/3], Step [618/3236], Loss: 2.2749, Perplexity: 9.7272
Epoch [2/3], Step [619/3236], Loss: 1.9784, Perplexity: 7.2312
Epoch [2/3], Step [620/3236], Loss: 2.3655, Perplexity: 10.6496
Epoch [2/3], Step [621/3236], Loss: 2.0808, Perplexity: 8.0105
Epoch [2/3], Step [622/3236], Loss: 2.2092, Perplexity: 9.1085
Epoch [2/3], Step [623/3236], Loss: 2.2319, Perplexity: 9.3174
Epoch [2/3], Step [624/3236], Loss: 2.5643, Perplexity: 12.9916
Epoch [2/3], Step [625/3236], Loss: 2.2756, Perplexity: 9.7338
Epoch [2/3], Step [626/3236], Loss: 2.1817, Perplexity: 8.8612
Epoch [2/3], Step [627/3236], Loss: 2.6448, Perplexity: 14.0807
Epoch [2/3], Step [628/3236], Loss: 2.1242, Perplexity: 8.3664
Epoch [2/3], Step [629/3236], Loss: 2.2770, Perplexity: 9.7469
Epoch [2/3], Step [630/3236], Loss: 2.1006, Perplexity: 8.1708
Epoch [2/3], Step [631/3236], Loss: 2.1309, Perplexity: 8.4223
Epoch [2/3], Step [632/3236], Loss: 2.2541, Perplexity: 9.5266
Epoch [2/3], Step [633/3236], Loss: 2.5614, Perplexity: 12.9534
Epoch [2/3], Step [634/3236], Loss: 1.9730, Perplexity: 7.1919
Epoch [2/3], Step [635/3236], Loss: 2.2747, Perplexity: 9.7245
Epoch [2/3], Step [636/3236], Loss: 2.2365, Perplexity: 9.3605
Epoch [2/3], Step [637/3236], Loss: 2.0496, Perplexity: 7.7647
Epoch [2/3], Step [638/3236], Loss: 2.4633, Perplexity: 11.7432
Epoch [2/3], Step [639/3236], Loss: 2.1107, Perplexity: 8.2543
Epoch [2/3], Step [640/3236], Loss: 2.3424, Perplexity: 10.4061
Epoch [2/3], Step [641/3236], Loss: 2.1904, Perplexity: 8.9389
Epoch [2/3], Step [642/3236], Loss: 2.1517, Perplexity: 8.5998
Epoch [2/3], Step [643/3236], Loss: 2.2678, Perplexity: 9.6584
Epoch [2/3], Step [644/3236], Loss: 2.1224, Perplexity: 8.3515
Epoch [2/3], Step [645/3236], Loss: 2.3449, Perplexity: 10.4322
Epoch [2/3], Step [646/3236], Loss: 2.2104, Perplexity: 9.1196
Epoch [2/3], Step [647/3236], Loss: 2.2656, Perplexity: 9.6369
Epoch [2/3], Step [648/3236], Loss: 2.1303, Perplexity: 8.4176
Epoch [2/3], Step [649/3236], Loss: 2.3488, Perplexity: 10.4728
Epoch [2/3], Step [650/3236], Loss: 2.1556, Perplexity: 8.6333
Epoch [2/3], Step [651/3236], Loss: 2.1536, Perplexity: 8.6161
Epoch [2/3], Step [652/3236], Loss: 2.3487, Perplexity: 10.4718
Epoch [2/3], Step [653/3236], Loss: 2.3912, Perplexity: 10.9270
Epoch [2/3], Step [654/3236], Loss: 2.2029, Perplexity: 9.0509
Epoch [2/3], Step [655/3236], Loss: 2.6754, Perplexity: 14.5186
Epoch [2/3], Step [656/3236], Loss: 2.2292, Perplexity: 9.2921
Epoch [2/3], Step [657/3236], Loss: 2.1463, Perplexity: 8.5530
Epoch [2/3], Step [658/3236], Loss: 2.1687, Perplexity: 8.7465
Epoch [2/3], Step [659/3236], Loss: 2.1647, Perplexity: 8.7120
Epoch [2/3], Step [660/3236], Loss: 2.7153, Perplexity: 15.1092
Epoch [2/3], Step [661/3236], Loss: 2.4804, Perplexity: 11.9464
Epoch [2/3], Step [662/3236], Loss: 2.2816, Perplexity: 9.7925
Epoch [2/3], Step [663/3236], Loss: 2.1913, Perplexity: 8.9470
Epoch [2/3], Step [664/3236], Loss: 2.2739, Perplexity: 9.7177
Epoch [2/3], Step [665/3236], Loss: 2.2873, Perplexity: 9.8484
Epoch [2/3], Step [666/3236], Loss: 2.1701, Perplexity: 8.7593
Epoch [2/3], Step [667/3236], Loss: 2.1395, Perplexity: 8.4950
Epoch [2/3], Step [668/3236], Loss: 2.6083, Perplexity: 13.5757
Epoch [2/3], Step [669/3236], Loss: 2.3422, Perplexity: 10.4045
Epoch [2/3], Step [670/3236], Loss: 2.1132, Perplexity: 8.2743
Epoch [2/3], Step [671/3236], Loss: 2.2015, Perplexity: 9.0384
Epoch [2/3], Step [672/3236], Loss: 2.0692, Perplexity: 7.9185
Epoch [2/3], Step [673/3236], Loss: 2.9740, Perplexity: 19.5696
Epoch [2/3], Step [674/3236], Loss: 2.3740, Perplexity: 10.7404
Epoch [2/3], Step [675/3236], Loss: 2.2535, Perplexity: 9.5213
Epoch [2/3], Step [676/3236], Loss: 2.2232, Perplexity: 9.2370
Epoch [2/3], Step [677/3236], Loss: 2.0709, Perplexity: 7.9322
Epoch [2/3], Step [678/3236], Loss: 2.0455, Perplexity: 7.7332
Epoch [2/3], Step [679/3236], Loss: 2.0526, Perplexity: 7.7882
Epoch [2/3], Step [680/3236], Loss: 2.1168, Perplexity: 8.3046
Epoch [2/3], Step [681/3236], Loss: 2.1834, Perplexity: 8.8769
Epoch [2/3], Step [682/3236], Loss: 2.1421, Perplexity: 8.5171
Epoch [2/3], Step [683/3236], Loss: 2.2179, Perplexity: 9.1876
Epoch [2/3], Step [684/3236], Loss: 2.1955, Perplexity: 8.9841
Epoch [2/3], Step [685/3236], Loss: 2.6921, Perplexity: 14.7629
Epoch [2/3], Step [686/3236], Loss: 2.3052, Perplexity: 10.0265
Epoch [2/3], Step [687/3236], Loss: 2.1459, Perplexity: 8.5500
Epoch [2/3], Step [688/3236], Loss: 2.0945, Perplexity: 8.1217
Epoch [2/3], Step [689/3236], Loss: 1.9232, Perplexity: 6.8426
Epoch [2/3], Step [690/3236], Loss: 2.0504, Perplexity: 7.7709
Epoch [2/3], Step [691/3236], Loss: 2.2621, Perplexity: 9.6034
Epoch [2/3], Step [692/3236], Loss: 2.1404, Perplexity: 8.5025
Epoch [2/3], Step [693/3236], Loss: 2.0969, Perplexity: 8.1411
Epoch [2/3], Step [694/3236], Loss: 2.2281, Perplexity: 9.2822
Epoch [2/3], Step [695/3236], Loss: 2.1055, Perplexity: 8.2112
Epoch [2/3], Step [696/3236], Loss: 2.2851, Perplexity: 9.8270
Epoch [2/3], Step [697/3236], Loss: 2.1441, Perplexity: 8.5347
Epoch [2/3], Step [698/3236], Loss: 2.3298, Perplexity: 10.2760
Epoch [2/3], Step [699/3236], Loss: 2.2220, Perplexity: 9.2258
Epoch [2/3], Step [700/3236], Loss: 2.2138, Perplexity: 9.1507
Epoch [2/3], Step [701/3236], Loss: 2.9523, Perplexity: 19.1507
Epoch [2/3], Step [702/3236], Loss: 2.2137, Perplexity: 9.1497
Epoch [2/3], Step [703/3236], Loss: 2.1871, Perplexity: 8.9098
Epoch [2/3], Step [704/3236], Loss: 2.2292, Perplexity: 9.2921
Epoch [2/3], Step [705/3236], Loss: 1.9815, Perplexity: 7.2535
Epoch [2/3], Step [706/3236], Loss: 2.1359, Perplexity: 8.4650
Epoch [2/3], Step [707/3236], Loss: 2.2814, Perplexity: 9.7902
Epoch [2/3], Step [708/3236], Loss: 2.2321, Perplexity: 9.3194
Epoch [2/3], Step [709/3236], Loss: 2.1844, Perplexity: 8.8849
Epoch [2/3], Step [710/3236], Loss: 2.2175, Perplexity: 9.1845
Epoch [2/3], Step [711/3236], Loss: 2.1099, Perplexity: 8.2475
Epoch [2/3], Step [712/3236], Loss: 2.2009, Perplexity: 9.0333
Epoch [2/3], Step [713/3236], Loss: 2.6371, Perplexity: 13.9731
Epoch [2/3], Step [714/3236], Loss: 2.6232, Perplexity: 13.7793
Epoch [2/3], Step [715/3236], Loss: 2.4713, Perplexity: 11.8375
Epoch [2/3], Step [716/3236], Loss: 2.1969, Perplexity: 8.9968
Epoch [2/3], Step [717/3236], Loss: 2.0778, Perplexity: 7.9866
Epoch [2/3], Step [718/3236], Loss: 2.2227, Perplexity: 9.2319
Epoch [2/3], Step [719/3236], Loss: 2.2210, Perplexity: 9.2162
Epoch [2/3], Step [720/3236], Loss: 2.5547, Perplexity: 12.8677
Epoch [2/3], Step [721/3236], Loss: 2.2201, Perplexity: 9.2083
Epoch [2/3], Step [722/3236], Loss: 2.0310, Perplexity: 7.6220
Epoch [2/3], Step [723/3236], Loss: 2.1784, Perplexity: 8.8319
Epoch [2/3], Step [724/3236], Loss: 2.3547, Perplexity: 10.5351
Epoch [2/3], Step [725/3236], Loss: 2.2022, Perplexity: 9.0453
Epoch [2/3], Step [726/3236], Loss: 2.3167, Perplexity: 10.1424
Epoch [2/3], Step [727/3236], Loss: 2.2753, Perplexity: 9.7308
Epoch [2/3], Step [728/3236], Loss: 2.1611, Perplexity: 8.6805
Epoch [2/3], Step [729/3236], Loss: 2.2621, Perplexity: 9.6033
Epoch [2/3], Step [730/3236], Loss: 2.1901, Perplexity: 8.9358
Epoch [2/3], Step [731/3236], Loss: 2.3037, Perplexity: 10.0107
Epoch [2/3], Step [732/3236], Loss: 2.2061, Perplexity: 9.0804
Epoch [2/3], Step [733/3236], Loss: 2.5167, Perplexity: 12.3875
Epoch [2/3], Step [734/3236], Loss: 2.1298, Perplexity: 8.4128
Epoch [2/3], Step [735/3236], Loss: 2.0718, Perplexity: 7.9389
Epoch [2/3], Step [736/3236], Loss: 2.0794, Perplexity: 7.9998
Epoch [2/3], Step [737/3236], Loss: 2.1912, Perplexity: 8.9461
Epoch [2/3], Step [738/3236], Loss: 2.0220, Perplexity: 7.5531
Epoch [2/3], Step [739/3236], Loss: 1.9760, Perplexity: 7.2138
Epoch [2/3], Step [740/3236], Loss: 2.3016, Perplexity: 9.9902
Epoch [2/3], Step [741/3236], Loss: 2.0923, Perplexity: 8.1034
Epoch [2/3], Step [742/3236], Loss: 2.3160, Perplexity: 10.1350
Epoch [2/3], Step [743/3236], Loss: 2.7625, Perplexity: 15.8392
Epoch [2/3], Step [744/3236], Loss: 2.0344, Perplexity: 7.6477
Epoch [2/3], Step [745/3236], Loss: 3.3397, Perplexity: 28.2120
Epoch [2/3], Step [746/3236], Loss: 2.2283, Perplexity: 9.2839
Epoch [2/3], Step [747/3236], Loss: 2.1934, Perplexity: 8.9660
Epoch [2/3], Step [748/3236], Loss: 2.5214, Perplexity: 12.4459
Epoch [2/3], Step [749/3236], Loss: 2.1312, Perplexity: 8.4250
Epoch [2/3], Step [750/3236], Loss: 2.1600, Perplexity: 8.6711
Epoch [2/3], Step [751/3236], Loss: 2.3344, Perplexity: 10.3233
Epoch [2/3], Step [752/3236], Loss: 2.4303, Perplexity: 11.3628
Epoch [2/3], Step [753/3236], Loss: 2.2016, Perplexity: 9.0394
Epoch [2/3], Step [754/3236], Loss: 2.4985, Perplexity: 12.1641
Epoch [2/3], Step [755/3236], Loss: 2.7457, Perplexity: 15.5758
Epoch [2/3], Step [756/3236], Loss: 2.1125, Perplexity: 8.2692
Epoch [2/3], Step [757/3236], Loss: 2.0964, Perplexity: 8.1365
Epoch [2/3], Step [758/3236], Loss: 2.1041, Perplexity: 8.1999
Epoch [2/3], Step [759/3236], Loss: 2.2532, Perplexity: 9.5182
Epoch [2/3], Step [760/3236], Loss: 1.9940, Perplexity: 7.3446
Epoch [2/3], Step [761/3236], Loss: 2.1690, Perplexity: 8.7495
Epoch [2/3], Step [762/3236], Loss: 2.2612, Perplexity: 9.5947
Epoch [2/3], Step [763/3236], Loss: 2.3262, Perplexity: 10.2394
Epoch [2/3], Step [764/3236], Loss: 2.9564, Perplexity: 19.2285
Epoch [2/3], Step [765/3236], Loss: 2.3406, Perplexity: 10.3873
Epoch [2/3], Step [766/3236], Loss: 2.3690, Perplexity: 10.6864
Epoch [2/3], Step [767/3236], Loss: 2.2119, Perplexity: 9.1334
Epoch [2/3], Step [768/3236], Loss: 2.5798, Perplexity: 13.1944
Epoch [2/3], Step [769/3236], Loss: 2.3113, Perplexity: 10.0879
Epoch [2/3], Step [770/3236], Loss: 2.1622, Perplexity: 8.6900
Epoch [2/3], Step [771/3236], Loss: 2.1260, Perplexity: 8.3812
Epoch [2/3], Step [772/3236], Loss: 2.1991, Perplexity: 9.0168
Epoch [2/3], Step [773/3236], Loss: 2.2351, Perplexity: 9.3471
Epoch [2/3], Step [774/3236], Loss: 2.2162, Perplexity: 9.1724
Epoch [2/3], Step [775/3236], Loss: 2.1159, Perplexity: 8.2971
Epoch [2/3], Step [776/3236], Loss: 2.2491, Perplexity: 9.4793
Epoch [2/3], Step [777/3236], Loss: 2.1914, Perplexity: 8.9478
Epoch [2/3], Step [778/3236], Loss: 2.1219, Perplexity: 8.3469
Epoch [2/3], Step [779/3236], Loss: 2.6163, Perplexity: 13.6848
Epoch [2/3], Step [780/3236], Loss: 2.1724, Perplexity: 8.7794
Epoch [2/3], Step [781/3236], Loss: 2.0520, Perplexity: 7.7831
Epoch [2/3], Step [782/3236], Loss: 2.2305, Perplexity: 9.3043
Epoch [2/3], Step [783/3236], Loss: 2.5054, Perplexity: 12.2491
Epoch [2/3], Step [784/3236], Loss: 2.1815, Perplexity: 8.8593
Epoch [2/3], Step [785/3236], Loss: 2.2269, Perplexity: 9.2714
Epoch [2/3], Step [786/3236], Loss: 2.6962, Perplexity: 14.8226
Epoch [2/3], Step [787/3236], Loss: 2.1547, Perplexity: 8.6256
Epoch [2/3], Step [788/3236], Loss: 2.2667, Perplexity: 9.6473
Epoch [2/3], Step [789/3236], Loss: 2.2300, Perplexity: 9.2994
Epoch [2/3], Step [790/3236], Loss: 2.3172, Perplexity: 10.1473
Epoch [2/3], Step [791/3236], Loss: 2.1232, Perplexity: 8.3580
Epoch [2/3], Step [792/3236], Loss: 2.3870, Perplexity: 10.8806
Epoch [2/3], Step [793/3236], Loss: 2.0817, Perplexity: 8.0183
Epoch [2/3], Step [794/3236], Loss: 2.4132, Perplexity: 11.1691
Epoch [2/3], Step [795/3236], Loss: 2.0926, Perplexity: 8.1060
Epoch [2/3], Step [796/3236], Loss: 1.9818, Perplexity: 7.2556
Epoch [2/3], Step [797/3236], Loss: 2.1565, Perplexity: 8.6410
Epoch [2/3], Step [798/3236], Loss: 2.0773, Perplexity: 7.9825
Epoch [2/3], Step [799/3236], Loss: 2.4208, Perplexity: 11.2548
Epoch [2/3], Step [800/3236], Loss: 2.1187, Perplexity: 8.3201
Epoch [2/3], Step [801/3236], Loss: 2.0188, Perplexity: 7.5296
Epoch [2/3], Step [802/3236], Loss: 2.4628, Perplexity: 11.7372
Epoch [2/3], Step [803/3236], Loss: 2.3220, Perplexity: 10.1959
Epoch [2/3], Step [804/3236], Loss: 2.1285, Perplexity: 8.4022
Epoch [2/3], Step [805/3236], Loss: 2.4263, Perplexity: 11.3173
Epoch [2/3], Step [806/3236], Loss: 2.1263, Perplexity: 8.3839
Epoch [2/3], Step [807/3236], Loss: 2.1472, Perplexity: 8.5613
Epoch [2/3], Step [808/3236], Loss: 2.2897, Perplexity: 9.8720
Epoch [2/3], Step [809/3236], Loss: 2.3155, Perplexity: 10.1295
Epoch [2/3], Step [810/3236], Loss: 2.4166, Perplexity: 11.2077
Epoch [2/3], Step [811/3236], Loss: 2.2362, Perplexity: 9.3573
Epoch [2/3], Step [812/3236], Loss: 2.2013, Perplexity: 9.0368
Epoch [2/3], Step [813/3236], Loss: 2.1470, Perplexity: 8.5592
Epoch [2/3], Step [814/3236], Loss: 2.1267, Perplexity: 8.3868
Epoch [2/3], Step [815/3236], Loss: 2.3379, Perplexity: 10.3598
Epoch [2/3], Step [816/3236], Loss: 2.2388, Perplexity: 9.3817
Epoch [2/3], Step [817/3236], Loss: 2.1808, Perplexity: 8.8530
Epoch [2/3], Step [818/3236], Loss: 2.0728, Perplexity: 7.9470
Epoch [2/3], Step [819/3236], Loss: 2.2211, Perplexity: 9.2171
Epoch [2/3], Step [820/3236], Loss: 2.2664, Perplexity: 9.6444
Epoch [2/3], Step [821/3236], Loss: 2.4524, Perplexity: 11.6157
Epoch [2/3], Step [822/3236], Loss: 2.1767, Perplexity: 8.8171
Epoch [2/3], Step [823/3236], Loss: 2.1546, Perplexity: 8.6242
Epoch [2/3], Step [824/3236], Loss: 2.0933, Perplexity: 8.1119
Epoch [2/3], Step [825/3236], Loss: 2.0351, Perplexity: 7.6529
Epoch [2/3], Step [826/3236], Loss: 2.2505, Perplexity: 9.4928
Epoch [2/3], Step [827/3236], Loss: 2.0882, Perplexity: 8.0707
Epoch [2/3], Step [828/3236], Loss: 2.0117, Perplexity: 7.4759
Epoch [2/3], Step [829/3236], Loss: 2.5640, Perplexity: 12.9874
Epoch [2/3], Step [830/3236], Loss: 2.4514, Perplexity: 11.6047
Epoch [2/3], Step [831/3236], Loss: 2.1850, Perplexity: 8.8907
Epoch [2/3], Step [832/3236], Loss: 2.0674, Perplexity: 7.9045
Epoch [2/3], Step [833/3236], Loss: 2.1497, Perplexity: 8.5825
Epoch [2/3], Step [834/3236], Loss: 2.0534, Perplexity: 7.7941
Epoch [2/3], Step [835/3236], Loss: 2.2121, Perplexity: 9.1351
Epoch [2/3], Step [836/3236], Loss: 2.3526, Perplexity: 10.5124
Epoch [2/3], Step [837/3236], Loss: 2.1018, Perplexity: 8.1806
Epoch [2/3], Step [838/3236], Loss: 2.2544, Perplexity: 9.5298
Epoch [2/3], Step [839/3236], Loss: 2.1008, Perplexity: 8.1723
Epoch [2/3], Step [840/3236], Loss: 2.1935, Perplexity: 8.9667
Epoch [2/3], Step [841/3236], Loss: 2.4536, Perplexity: 11.6299
Epoch [2/3], Step [842/3236], Loss: 2.3771, Perplexity: 10.7737
Epoch [2/3], Step [843/3236], Loss: 2.1109, Perplexity: 8.2554
Epoch [2/3], Step [844/3236], Loss: 2.0069, Perplexity: 7.4405
Epoch [2/3], Step [845/3236], Loss: 2.1661, Perplexity: 8.7246
Epoch [2/3], Step [846/3236], Loss: 2.7787, Perplexity: 16.0983
Epoch [2/3], Step [847/3236], Loss: 3.2793, Perplexity: 26.5578
Epoch [2/3], Step [848/3236], Loss: 2.2234, Perplexity: 9.2391
Epoch [2/3], Step [849/3236], Loss: 2.2177, Perplexity: 9.1862
Epoch [2/3], Step [850/3236], Loss: 2.0756, Perplexity: 7.9696
Epoch [2/3], Step [851/3236], Loss: 2.1658, Perplexity: 8.7217
Epoch [2/3], Step [852/3236], Loss: 2.5071, Perplexity: 12.2691
Epoch [2/3], Step [853/3236], Loss: 2.2824, Perplexity: 9.8004
Epoch [2/3], Step [854/3236], Loss: 2.9461, Perplexity: 19.0318
Epoch [2/3], Step [855/3236], Loss: 2.5416, Perplexity: 12.7006
Epoch [2/3], Step [856/3236], Loss: 2.3024, Perplexity: 9.9978
Epoch [2/3], Step [857/3236], Loss: 2.4860, Perplexity: 12.0127
Epoch [2/3], Step [858/3236], Loss: 2.0670, Perplexity: 7.9012
Epoch [2/3], Step [859/3236], Loss: 2.0848, Perplexity: 8.0428
Epoch [2/3], Step [860/3236], Loss: 2.1144, Perplexity: 8.2844
Epoch [2/3], Step [861/3236], Loss: 2.2355, Perplexity: 9.3516
Epoch [2/3], Step [862/3236], Loss: 2.1430, Perplexity: 8.5249
Epoch [2/3], Step [863/3236], Loss: 2.2512, Perplexity: 9.4991
Epoch [2/3], Step [864/3236], Loss: 2.2477, Perplexity: 9.4655
Epoch [2/3], Step [865/3236], Loss: 2.2816, Perplexity: 9.7925
Epoch [2/3], Step [866/3236], Loss: 2.1946, Perplexity: 8.9763
Epoch [2/3], Step [867/3236], Loss: 2.0020, Perplexity: 7.4041
Epoch [2/3], Step [868/3236], Loss: 2.1293, Perplexity: 8.4091
Epoch [2/3], Step [869/3236], Loss: 2.1175, Perplexity: 8.3101
Epoch [2/3], Step [870/3236], Loss: 2.5547, Perplexity: 12.8678
Epoch [2/3], Step [871/3236], Loss: 2.2083, Perplexity: 9.1004
Epoch [2/3], Step [872/3236], Loss: 2.1969, Perplexity: 8.9969
Epoch [2/3], Step [873/3236], Loss: 2.0643, Perplexity: 7.8795
Epoch [2/3], Step [874/3236], Loss: 2.2008, Perplexity: 9.0326
Epoch [2/3], Step [875/3236], Loss: 2.0111, Perplexity: 7.4717
Epoch [2/3], Step [876/3236], Loss: 2.5497, Perplexity: 12.8033
Epoch [2/3], Step [877/3236], Loss: 2.0517, Perplexity: 7.7810
Epoch [2/3], Step [878/3236], Loss: 2.0057, Perplexity: 7.4316
Epoch [2/3], Step [879/3236], Loss: 2.3428, Perplexity: 10.4098
Epoch [2/3], Step [880/3236], Loss: 2.1787, Perplexity: 8.8350
Epoch [2/3], Step [881/3236], Loss: 2.2074, Perplexity: 9.0921
Epoch [2/3], Step [882/3236], Loss: 2.1690, Perplexity: 8.7497
Epoch [2/3], Step [883/3236], Loss: 2.8190, Perplexity: 16.7598
Epoch [2/3], Step [884/3236], Loss: 2.3219, Perplexity: 10.1955
Epoch [2/3], Step [885/3236], Loss: 2.2217, Perplexity: 9.2228
Epoch [2/3], Step [886/3236], Loss: 2.3033, Perplexity: 10.0071
Epoch [2/3], Step [887/3236], Loss: 2.2375, Perplexity: 9.3699
Epoch [2/3], Step [888/3236], Loss: 2.1966, Perplexity: 8.9942
Epoch [2/3], Step [889/3236], Loss: 2.0706, Perplexity: 7.9293
Epoch [2/3], Step [890/3236], Loss: 3.2032, Perplexity: 24.6104
Epoch [2/3], Step [891/3236], Loss: 2.0421, Perplexity: 7.7069
Epoch [2/3], Step [892/3236], Loss: 2.2781, Perplexity: 9.7581
Epoch [2/3], Step [893/3236], Loss: 2.0960, Perplexity: 8.1332
Epoch [2/3], Step [894/3236], Loss: 2.1547, Perplexity: 8.6253
Epoch [2/3], Step [895/3236], Loss: 2.0853, Perplexity: 8.0467
Epoch [2/3], Step [896/3236], Loss: 2.0614, Perplexity: 7.8572
Epoch [2/3], Step [897/3236], Loss: 2.8699, Perplexity: 17.6348
Epoch [2/3], Step [898/3236], Loss: 2.0860, Perplexity: 8.0526
Epoch [2/3], Step [899/3236], Loss: 2.2258, Perplexity: 9.2610
Epoch [2/3], Step [900/3236], Loss: 2.0482, Perplexity: 7.7537
Epoch [2/3], Step [901/3236], Loss: 2.4223, Perplexity: 11.2714
Epoch [2/3], Step [902/3236], Loss: 2.1018, Perplexity: 8.1810
Epoch [2/3], Step [903/3236], Loss: 2.0942, Perplexity: 8.1188
Epoch [2/3], Step [904/3236], Loss: 2.1330, Perplexity: 8.4400
Epoch [2/3], Step [905/3236], Loss: 2.3670, Perplexity: 10.6651
Epoch [2/3], Step [906/3236], Loss: 2.2849, Perplexity: 9.8251
Epoch [2/3], Step [907/3236], Loss: 3.0062, Perplexity: 20.2113
Epoch [2/3], Step [908/3236], Loss: 2.2853, Perplexity: 9.8287
Epoch [2/3], Step [909/3236], Loss: 2.0204, Perplexity: 7.5416
Epoch [2/3], Step [910/3236], Loss: 1.9570, Perplexity: 7.0781
Epoch [2/3], Step [911/3236], Loss: 2.2842, Perplexity: 9.8176
Epoch [2/3], Step [912/3236], Loss: 2.0526, Perplexity: 7.7884
Epoch [2/3], Step [913/3236], Loss: 2.2905, Perplexity: 9.8802
Epoch [2/3], Step [914/3236], Loss: 2.3962, Perplexity: 10.9815
Epoch [2/3], Step [915/3236], Loss: 2.2115, Perplexity: 9.1294
Epoch [2/3], Step [916/3236], Loss: 2.1518, Perplexity: 8.6002
Epoch [2/3], Step [917/3236], Loss: 2.1791, Perplexity: 8.8387
Epoch [2/3], Step [918/3236], Loss: 2.2617, Perplexity: 9.5995
Epoch [2/3], Step [919/3236], Loss: 2.1921, Perplexity: 8.9542
Epoch [2/3], Step [920/3236], Loss: 2.3326, Perplexity: 10.3044
Epoch [2/3], Step [921/3236], Loss: 2.2387, Perplexity: 9.3810
Epoch [2/3], Step [922/3236], Loss: 2.4023, Perplexity: 11.0483
Epoch [2/3], Step [923/3236], Loss: 2.2079, Perplexity: 9.0970
Epoch [2/3], Step [924/3236], Loss: 2.3007, Perplexity: 9.9811
Epoch [2/3], Step [925/3236], Loss: 2.2314, Perplexity: 9.3131
Epoch [2/3], Step [926/3236], Loss: 2.3518, Perplexity: 10.5039
Epoch [2/3], Step [927/3236], Loss: 2.2676, Perplexity: 9.6558
Epoch [2/3], Step [928/3236], Loss: 2.2891, Perplexity: 9.8656
Epoch [2/3], Step [929/3236], Loss: 2.2126, Perplexity: 9.1391
Epoch [2/3], Step [930/3236], Loss: 2.2827, Perplexity: 9.8035
Epoch [2/3], Step [931/3236], Loss: 2.3679, Perplexity: 10.6746
Epoch [2/3], Step [932/3236], Loss: 2.1746, Perplexity: 8.7983
Epoch [2/3], Step [933/3236], Loss: 2.3059, Perplexity: 10.0328
Epoch [2/3], Step [934/3236], Loss: 2.2082, Perplexity: 9.0989
Epoch [2/3], Step [935/3236], Loss: 2.4975, Perplexity: 12.1521
Epoch [2/3], Step [936/3236], Loss: 2.1810, Perplexity: 8.8549
Epoch [2/3], Step [937/3236], Loss: 2.3262, Perplexity: 10.2390
Epoch [2/3], Step [938/3236], Loss: 2.1302, Perplexity: 8.4167
Epoch [2/3], Step [939/3236], Loss: 2.2282, Perplexity: 9.2835
Epoch [2/3], Step [940/3236], Loss: 2.2694, Perplexity: 9.6736
Epoch [2/3], Step [941/3236], Loss: 2.1766, Perplexity: 8.8163
Epoch [2/3], Step [942/3236], Loss: 2.1254, Perplexity: 8.3764
Epoch [2/3], Step [943/3236], Loss: 2.5638, Perplexity: 12.9849
Epoch [2/3], Step [944/3236], Loss: 2.7628, Perplexity: 15.8447
Epoch [2/3], Step [945/3236], Loss: 2.2450, Perplexity: 9.4404
Epoch [2/3], Step [946/3236], Loss: 2.2346, Perplexity: 9.3426
Epoch [2/3], Step [947/3236], Loss: 2.1974, Perplexity: 9.0019
Epoch [2/3], Step [948/3236], Loss: 2.3448, Perplexity: 10.4308
Epoch [2/3], Step [949/3236], Loss: 2.5442, Perplexity: 12.7333
Epoch [2/3], Step [950/3236], Loss: 2.2269, Perplexity: 9.2714
Epoch [2/3], Step [951/3236], Loss: 2.4302, Perplexity: 11.3606
Epoch [2/3], Step [952/3236], Loss: 2.6167, Perplexity: 13.6904
Epoch [2/3], Step [953/3236], Loss: 2.0604, Perplexity: 7.8494
Epoch [2/3], Step [954/3236], Loss: 2.4777, Perplexity: 11.9141
Epoch [2/3], Step [955/3236], Loss: 2.5261, Perplexity: 12.5052
Epoch [2/3], Step [956/3236], Loss: 2.1829, Perplexity: 8.8722
Epoch [2/3], Step [957/3236], Loss: 2.1715, Perplexity: 8.7711
Epoch [2/3], Step [958/3236], Loss: 2.1646, Perplexity: 8.7113
Epoch [2/3], Step [959/3236], Loss: 2.1506, Perplexity: 8.5901
Epoch [2/3], Step [960/3236], Loss: 2.2065, Perplexity: 9.0835
Epoch [2/3], Step [961/3236], Loss: 2.1971, Perplexity: 8.9986
Epoch [2/3], Step [962/3236], Loss: 2.2525, Perplexity: 9.5112
Epoch [2/3], Step [963/3236], Loss: 2.2770, Perplexity: 9.7478
Epoch [2/3], Step [964/3236], Loss: 2.0325, Perplexity: 7.6331
Epoch [2/3], Step [965/3236], Loss: 2.3452, Perplexity: 10.4351
Epoch [2/3], Step [966/3236], Loss: 2.8064, Perplexity: 16.5509
Epoch [2/3], Step [967/3236], Loss: 2.0343, Perplexity: 7.6470
Epoch [2/3], Step [968/3236], Loss: 2.7001, Perplexity: 14.8813
Epoch [2/3], Step [969/3236], Loss: 3.4030, Perplexity: 30.0544
Epoch [2/3], Step [970/3236], Loss: 2.1147, Perplexity: 8.2868
Epoch [2/3], Step [971/3236], Loss: 2.4136, Perplexity: 11.1746
Epoch [2/3], Step [972/3236], Loss: 2.2212, Perplexity: 9.2183
Epoch [2/3], Step [973/3236], Loss: 2.3027, Perplexity: 10.0010
Epoch [2/3], Step [974/3236], Loss: 2.1633, Perplexity: 8.7000
Epoch [2/3], Step [975/3236], Loss: 2.1671, Perplexity: 8.7329
Epoch [2/3], Step [976/3236], Loss: 2.1051, Perplexity: 8.2082
Epoch [2/3], Step [977/3236], Loss: 2.2757, Perplexity: 9.7348
Epoch [2/3], Step [978/3236], Loss: 2.1547, Perplexity: 8.6253
Epoch [2/3], Step [979/3236], Loss: 2.1598, Perplexity: 8.6693
Epoch [2/3], Step [980/3236], Loss: 2.8333, Perplexity: 17.0007
Epoch [2/3], Step [981/3236], Loss: 2.2244, Perplexity: 9.2479
Epoch [2/3], Step [982/3236], Loss: 2.1351, Perplexity: 8.4576
Epoch [2/3], Step [983/3236], Loss: 2.4058, Perplexity: 11.0870
Epoch [2/3], Step [984/3236], Loss: 2.1673, Perplexity: 8.7349
Epoch [2/3], Step [985/3236], Loss: 2.1481, Perplexity: 8.5683
Epoch [2/3], Step [986/3236], Loss: 2.4741, Perplexity: 11.8713
Epoch [2/3], Step [987/3236], Loss: 2.0781, Perplexity: 7.9892
Epoch [2/3], Step [988/3236], Loss: 2.1962, Perplexity: 8.9906
Epoch [2/3], Step [989/3236], Loss: 2.1445, Perplexity: 8.5375
Epoch [2/3], Step [990/3236], Loss: 2.6080, Perplexity: 13.5717
Epoch [2/3], Step [991/3236], Loss: 2.2056, Perplexity: 9.0756
Epoch [2/3], Step [992/3236], Loss: 2.2317, Perplexity: 9.3161
Epoch [2/3], Step [993/3236], Loss: 2.4428, Perplexity: 11.5052
Epoch [2/3], Step [994/3236], Loss: 2.1630, Perplexity: 8.6970
Epoch [2/3], Step [995/3236], Loss: 2.3832, Perplexity: 10.8398
Epoch [2/3], Step [996/3236], Loss: 2.1545, Perplexity: 8.6239
Epoch [2/3], Step [997/3236], Loss: 2.4006, Perplexity: 11.0301
Epoch [2/3], Step [998/3236], Loss: 1.9694, Perplexity: 7.1664
Epoch [2/3], Step [999/3236], Loss: 2.2137, Perplexity: 9.1493
Epoch [2/3], Step [1000/3236], Loss: 2.2054, Perplexity: 9.0741
Epoch [2/3], Step [1001/3236], Loss: 2.2164, Perplexity: 9.1738
Epoch [2/3], Step [1002/3236], Loss: 2.4059, Perplexity: 11.0880
Epoch [2/3], Step [1003/3236], Loss: 2.5288, Perplexity: 12.5389
Epoch [2/3], Step [1004/3236], Loss: 2.2000, Perplexity: 9.0249
Epoch [2/3], Step [1005/3236], Loss: 2.0418, Perplexity: 7.7041
Epoch [2/3], Step [1006/3236], Loss: 2.1252, Perplexity: 8.3748
Epoch [2/3], Step [1007/3236], Loss: 2.1851, Perplexity: 8.8914
Epoch [2/3], Step [1008/3236], Loss: 2.1599, Perplexity: 8.6701
Epoch [2/3], Step [1009/3236], Loss: 2.2120, Perplexity: 9.1341
Epoch [2/3], Step [1010/3236], Loss: 2.1103, Perplexity: 8.2509
Epoch [2/3], Step [1011/3236], Loss: 2.2147, Perplexity: 9.1588
Epoch [2/3], Step [1012/3236], Loss: 2.1739, Perplexity: 8.7921
Epoch [2/3], Step [1013/3236], Loss: 2.2199, Perplexity: 9.2063
Epoch [2/3], Step [1014/3236], Loss: 2.2832, Perplexity: 9.8076
Epoch [2/3], Step [1015/3236], Loss: 2.2152, Perplexity: 9.1633
Epoch [2/3], Step [1016/3236], Loss: 2.1949, Perplexity: 8.9792
Epoch [2/3], Step [1017/3236], Loss: 2.0969, Perplexity: 8.1406
Epoch [2/3], Step [1018/3236], Loss: 2.2311, Perplexity: 9.3096
Epoch [2/3], Step [1019/3236], Loss: 2.1424, Perplexity: 8.5196
Epoch [2/3], Step [1020/3236], Loss: 2.1546, Perplexity: 8.6246
Epoch [2/3], Step [1021/3236], Loss: 2.0630, Perplexity: 7.8694
Epoch [2/3], Step [1022/3236], Loss: 2.2176, Perplexity: 9.1848
Epoch [2/3], Step [1023/3236], Loss: 2.0847, Perplexity: 8.0419
Epoch [2/3], Step [1024/3236], Loss: 2.0905, Perplexity: 8.0891
Epoch [2/3], Step [1025/3236], Loss: 2.0917, Perplexity: 8.0987
Epoch [2/3], Step [1026/3236], Loss: 3.0747, Perplexity: 21.6430
Epoch [2/3], Step [1027/3236], Loss: 2.2145, Perplexity: 9.1564
Epoch [2/3], Step [1028/3236], Loss: 2.0885, Perplexity: 8.0726
Epoch [2/3], Step [1029/3236], Loss: 2.3734, Perplexity: 10.7340
Epoch [2/3], Step [1030/3236], Loss: 2.2661, Perplexity: 9.6419
Epoch [2/3], Step [1031/3236], Loss: 2.0044, Perplexity: 7.4218
Epoch [2/3], Step [1032/3236], Loss: 2.0490, Perplexity: 7.7605
Epoch [2/3], Step [1033/3236], Loss: 2.1682, Perplexity: 8.7428
Epoch [2/3], Step [1034/3236], Loss: 2.2912, Perplexity: 9.8871
Epoch [2/3], Step [1035/3236], Loss: 2.1879, Perplexity: 8.9162
Epoch [2/3], Step [1036/3236], Loss: 1.9398, Perplexity: 6.9576
Epoch [2/3], Step [1037/3236], Loss: 2.2280, Perplexity: 9.2812
Epoch [2/3], Step [1038/3236], Loss: 2.1492, Perplexity: 8.5783
Epoch [2/3], Step [1039/3236], Loss: 1.9873, Perplexity: 7.2959
Epoch [2/3], Step [1040/3236], Loss: 2.3931, Perplexity: 10.9476
Epoch [2/3], Step [1041/3236], Loss: 2.0720, Perplexity: 7.9407
Epoch [2/3], Step [1042/3236], Loss: 2.2488, Perplexity: 9.4763
Epoch [2/3], Step [1043/3236], Loss: 2.9216, Perplexity: 18.5712
Epoch [2/3], Step [1044/3236], Loss: 2.2608, Perplexity: 9.5910
Epoch [2/3], Step [1045/3236], Loss: 2.1401, Perplexity: 8.5001
Epoch [2/3], Step [1046/3236], Loss: 2.9990, Perplexity: 20.0652
Epoch [2/3], Step [1047/3236], Loss: 2.0495, Perplexity: 7.7637
Epoch [2/3], Step [1048/3236], Loss: 2.0851, Perplexity: 8.0455
Epoch [2/3], Step [1049/3236], Loss: 2.2091, Perplexity: 9.1076
Epoch [2/3], Step [1050/3236], Loss: 2.0926, Perplexity: 8.1057
Epoch [2/3], Step [1051/3236], Loss: 2.4352, Perplexity: 11.4181
Epoch [2/3], Step [1052/3236], Loss: 2.3225, Perplexity: 10.2013
Epoch [2/3], Step [1053/3236], Loss: 2.0046, Perplexity: 7.4234
Epoch [2/3], Step [1054/3236], Loss: 2.1149, Perplexity: 8.2890
Epoch [2/3], Step [1055/3236], Loss: 2.3106, Perplexity: 10.0809
Epoch [2/3], Step [1056/3236], Loss: 2.0942, Perplexity: 8.1186
Epoch [2/3], Step [1057/3236], Loss: 2.2113, Perplexity: 9.1273
Epoch [2/3], Step [1058/3236], Loss: 2.5718, Perplexity: 13.0888
Epoch [2/3], Step [1059/3236], Loss: 2.1152, Perplexity: 8.2916
Epoch [2/3], Step [1060/3236], Loss: 2.3933, Perplexity: 10.9500
Epoch [2/3], Step [1061/3236], Loss: 2.6602, Perplexity: 14.2989
Epoch [2/3], Step [1062/3236], Loss: 2.1322, Perplexity: 8.4333
Epoch [2/3], Step [1063/3236], Loss: 2.0899, Perplexity: 8.0838
Epoch [2/3], Step [1064/3236], Loss: 2.1712, Perplexity: 8.7692
Epoch [2/3], Step [1065/3236], Loss: 2.1063, Perplexity: 8.2177
Epoch [2/3], Step [1066/3236], Loss: 2.1872, Perplexity: 8.9103
Epoch [2/3], Step [1067/3236], Loss: 2.2093, Perplexity: 9.1096
Epoch [2/3], Step [1068/3236], Loss: 2.2659, Perplexity: 9.6403
Epoch [2/3], Step [1069/3236], Loss: 2.1175, Perplexity: 8.3105
Epoch [2/3], Step [1070/3236], Loss: 2.1946, Perplexity: 8.9767
Epoch [2/3], Step [1071/3236], Loss: 2.0444, Perplexity: 7.7248
Epoch [2/3], Step [1072/3236], Loss: 2.0391, Perplexity: 7.6836
Epoch [2/3], Step [1073/3236], Loss: 2.1912, Perplexity: 8.9461
Epoch [2/3], Step [1074/3236], Loss: 2.5147, Perplexity: 12.3627
Epoch [2/3], Step [1075/3236], Loss: 2.1360, Perplexity: 8.4654
Epoch [2/3], Step [1076/3236], Loss: 2.3410, Perplexity: 10.3912
Epoch [2/3], Step [1077/3236], Loss: 2.1848, Perplexity: 8.8892
Epoch [2/3], Step [1078/3236], Loss: 2.0155, Perplexity: 7.5045
Epoch [2/3], Step [1079/3236], Loss: 2.1925, Perplexity: 8.9575
Epoch [2/3], Step [1080/3236], Loss: 2.0665, Perplexity: 7.8970
Epoch [2/3], Step [1081/3236], Loss: 2.1892, Perplexity: 8.9278
Epoch [2/3], Step [1082/3236], Loss: 2.1917, Perplexity: 8.9508
Epoch [2/3], Step [1083/3236], Loss: 1.9312, Perplexity: 6.8976
Epoch [2/3], Step [1084/3236], Loss: 2.2147, Perplexity: 9.1588
Epoch [2/3], Step [1085/3236], Loss: 2.0862, Perplexity: 8.0543
Epoch [2/3], Step [1086/3236], Loss: 2.1845, Perplexity: 8.8859
Epoch [2/3], Step [1087/3236], Loss: 2.0219, Perplexity: 7.5529
Epoch [2/3], Step [1088/3236], Loss: 2.2066, Perplexity: 9.0844
Epoch [2/3], Step [1089/3236], Loss: 2.3052, Perplexity: 10.0261
Epoch [2/3], Step [1090/3236], Loss: 2.1633, Perplexity: 8.7001
Epoch [2/3], Step [1091/3236], Loss: 2.1607, Perplexity: 8.6770
Epoch [2/3], Step [1092/3236], Loss: 2.1624, Perplexity: 8.6924
Epoch [2/3], Step [1093/3236], Loss: 2.2197, Perplexity: 9.2049
Epoch [2/3], Step [1094/3236], Loss: 2.2388, Perplexity: 9.3821
Epoch [2/3], Step [1095/3236], Loss: 2.5440, Perplexity: 12.7299
Epoch [2/3], Step [1096/3236], Loss: 2.1887, Perplexity: 8.9237
Epoch [2/3], Step [1097/3236], Loss: 2.1499, Perplexity: 8.5844
Epoch [2/3], Step [1098/3236], Loss: 2.3713, Perplexity: 10.7115
Epoch [2/3], Step [1099/3236], Loss: 2.2185, Perplexity: 9.1932
Epoch [2/3], Step [1100/3236], Loss: 2.6538, Perplexity: 14.2074
Epoch [2/3], Step [1101/3236], Loss: 2.2530, Perplexity: 9.5158
Epoch [2/3], Step [1102/3236], Loss: 2.4619, Perplexity: 11.7276
Epoch [2/3], Step [1103/3236], Loss: 2.1203, Perplexity: 8.3337
Epoch [2/3], Step [1104/3236], Loss: 2.0952, Perplexity: 8.1274
Epoch [2/3], Step [1105/3236], Loss: 2.1949, Perplexity: 8.9789
Epoch [2/3], Step [1106/3236], Loss: 2.1687, Perplexity: 8.7467
Epoch [2/3], Step [1107/3236], Loss: 2.0512, Perplexity: 7.7771
Epoch [2/3], Step [1108/3236], Loss: 2.2026, Perplexity: 9.0486
Epoch [2/3], Step [1109/3236], Loss: 2.1074, Perplexity: 8.2271
Epoch [2/3], Step [1110/3236], Loss: 2.4649, Perplexity: 11.7620
Epoch [2/3], Step [1111/3236], Loss: 2.1284, Perplexity: 8.4011
Epoch [2/3], Step [1112/3236], Loss: 2.2365, Perplexity: 9.3602
Epoch [2/3], Step [1113/3236], Loss: 2.1082, Perplexity: 8.2336
Epoch [2/3], Step [1114/3236], Loss: 2.1175, Perplexity: 8.3105
Epoch [2/3], Step [1115/3236], Loss: 2.5130, Perplexity: 12.3419
Epoch [2/3], Step [1116/3236], Loss: 2.1270, Perplexity: 8.3895
Epoch [2/3], Step [1117/3236], Loss: 2.0755, Perplexity: 7.9688
Epoch [2/3], Step [1118/3236], Loss: 2.2046, Perplexity: 9.0670
Epoch [2/3], Step [1119/3236], Loss: 2.0840, Perplexity: 8.0364
Epoch [2/3], Step [1120/3236], Loss: 2.5987, Perplexity: 13.4465
Epoch [2/3], Step [1121/3236], Loss: 2.1267, Perplexity: 8.3873
Epoch [2/3], Step [1122/3236], Loss: 2.2385, Perplexity: 9.3790
Epoch [2/3], Step [1123/3236], Loss: 2.2718, Perplexity: 9.6968
Epoch [2/3], Step [1124/3236], Loss: 2.3359, Perplexity: 10.3386
Epoch [2/3], Step [1125/3236], Loss: 2.0931, Perplexity: 8.1104
Epoch [2/3], Step [1126/3236], Loss: 2.1505, Perplexity: 8.5890
Epoch [2/3], Step [1127/3236], Loss: 2.3990, Perplexity: 11.0127
Epoch [2/3], Step [1128/3236], Loss: 2.3563, Perplexity: 10.5513
Epoch [2/3], Step [1129/3236], Loss: 2.0500, Perplexity: 7.7680
Epoch [2/3], Step [1130/3236], Loss: 2.0096, Perplexity: 7.4604
Epoch [2/3], Step [1131/3236], Loss: 2.0797, Perplexity: 8.0022
Epoch [2/3], Step [1132/3236], Loss: 2.2093, Perplexity: 9.1090
Epoch [2/3], Step [1133/3236], Loss: 2.0722, Perplexity: 7.9425
Epoch [2/3], Step [1134/3236], Loss: 2.1119, Perplexity: 8.2640
Epoch [2/3], Step [1135/3236], Loss: 2.1462, Perplexity: 8.5526
Epoch [2/3], Step [1136/3236], Loss: 2.1588, Perplexity: 8.6605
Epoch [2/3], Step [1137/3236], Loss: 2.3176, Perplexity: 10.1512
Epoch [2/3], Step [1138/3236], Loss: 2.2662, Perplexity: 9.6425
Epoch [2/3], Step [1139/3236], Loss: 2.0098, Perplexity: 7.4617
Epoch [2/3], Step [1140/3236], Loss: 2.1312, Perplexity: 8.4252
Epoch [2/3], Step [1141/3236], Loss: 2.2618, Perplexity: 9.6007
Epoch [2/3], Step [1142/3236], Loss: 2.5515, Perplexity: 12.8259
Epoch [2/3], Step [1143/3236], Loss: 2.1564, Perplexity: 8.6401
Epoch [2/3], Step [1144/3236], Loss: 2.2074, Perplexity: 9.0917
Epoch [2/3], Step [1145/3236], Loss: 2.1428, Perplexity: 8.5235
Epoch [2/3], Step [1146/3236], Loss: 2.1325, Perplexity: 8.4358
Epoch [2/3], Step [1147/3236], Loss: 2.1502, Perplexity: 8.5865
Epoch [2/3], Step [1148/3236], Loss: 2.1591, Perplexity: 8.6633
Epoch [2/3], Step [1149/3236], Loss: 2.7121, Perplexity: 15.0610
Epoch [2/3], Step [1150/3236], Loss: 2.0680, Perplexity: 7.9090
Epoch [2/3], Step [1151/3236], Loss: 2.0393, Perplexity: 7.6852
Epoch [2/3], Step [1152/3236], Loss: 2.1158, Perplexity: 8.2966
Epoch [2/3], Step [1153/3236], Loss: 2.2243, Perplexity: 9.2469
Epoch [2/3], Step [1154/3236], Loss: 2.1491, Perplexity: 8.5775
Epoch [2/3], Step [1155/3236], Loss: 2.0682, Perplexity: 7.9105
Epoch [2/3], Step [1156/3236], Loss: 2.2632, Perplexity: 9.6140
Epoch [2/3], Step [1157/3236], Loss: 2.0302, Perplexity: 7.6156
Epoch [2/3], Step [1158/3236], Loss: 2.1199, Perplexity: 8.3305
Epoch [2/3], Step [1159/3236], Loss: 2.2437, Perplexity: 9.4284
Epoch [2/3], Step [1160/3236], Loss: 2.1924, Perplexity: 8.9571
Epoch [2/3], Step [1161/3236], Loss: 2.1571, Perplexity: 8.6460
Epoch [2/3], Step [1162/3236], Loss: 2.0076, Perplexity: 7.4456
Epoch [2/3], Step [1163/3236], Loss: 2.1956, Perplexity: 8.9851
Epoch [2/3], Step [1164/3236], Loss: 2.0613, Perplexity: 7.8559
Epoch [2/3], Step [1165/3236], Loss: 2.4224, Perplexity: 11.2734
Epoch [2/3], Step [1166/3236], Loss: 2.2276, Perplexity: 9.2778
Epoch [2/3], Step [1167/3236], Loss: 2.2089, Perplexity: 9.1060
Epoch [2/3], Step [1168/3236], Loss: 2.2628, Perplexity: 9.6097
Epoch [2/3], Step [1169/3236], Loss: 2.0229, Perplexity: 7.5602
Epoch [2/3], Step [1170/3236], Loss: 2.0236, Perplexity: 7.5654
Epoch [2/3], Step [1171/3236], Loss: 2.2075, Perplexity: 9.0932
Epoch [2/3], Step [1172/3236], Loss: 2.0245, Perplexity: 7.5720
Epoch [2/3], Step [1173/3236], Loss: 2.0497, Perplexity: 7.7652
Epoch [2/3], Step [1174/3236], Loss: 2.0709, Perplexity: 7.9317
Epoch [2/3], Step [1175/3236], Loss: 2.1295, Perplexity: 8.4111
Epoch [2/3], Step [1176/3236], Loss: 2.3714, Perplexity: 10.7123
Epoch [2/3], Step [1177/3236], Loss: 2.1535, Perplexity: 8.6152
Epoch [2/3], Step [1178/3236], Loss: 2.4385, Perplexity: 11.4562
Epoch [2/3], Step [1179/3236], Loss: 2.0843, Perplexity: 8.0388
Epoch [2/3], Step [1180/3236], Loss: 2.1621, Perplexity: 8.6892
Epoch [2/3], Step [1181/3236], Loss: 2.1314, Perplexity: 8.4266
Epoch [2/3], Step [1182/3236], Loss: 2.1038, Perplexity: 8.1974
Epoch [2/3], Step [1183/3236], Loss: 2.0583, Perplexity: 7.8326
Epoch [2/3], Step [1184/3236], Loss: 2.4508, Perplexity: 11.5978
Epoch [2/3], Step [1185/3236], Loss: 2.2067, Perplexity: 9.0855
Epoch [2/3], Step [1186/3236], Loss: 2.0329, Perplexity: 7.6362
Epoch [2/3], Step [1187/3236], Loss: 2.1083, Perplexity: 8.2344
Epoch [2/3], Step [1188/3236], Loss: 2.5109, Perplexity: 12.3158
Epoch [2/3], Step [1189/3236], Loss: 2.0907, Perplexity: 8.0904
Epoch [2/3], Step [1190/3236], Loss: 2.1748, Perplexity: 8.8007
Epoch [2/3], Step [1191/3236], Loss: 2.3758, Perplexity: 10.7599
Epoch [2/3], Step [1192/3236], Loss: 2.2133, Perplexity: 9.1460
Epoch [2/3], Step [1193/3236], Loss: 2.3070, Perplexity: 10.0447
Epoch [2/3], Step [1194/3236], Loss: 1.8934, Perplexity: 6.6419
Epoch [2/3], Step [1195/3236], Loss: 2.1384, Perplexity: 8.4861
Epoch [2/3], Step [1196/3236], Loss: 2.0244, Perplexity: 7.5712
Epoch [2/3], Step [1197/3236], Loss: 2.1860, Perplexity: 8.8992
Epoch [2/3], Step [1198/3236], Loss: 2.2345, Perplexity: 9.3422
Epoch [2/3], Step [1199/3236], Loss: 2.2207, Perplexity: 9.2134
Epoch [2/3], Step [1200/3236], Loss: 2.1813, Perplexity: 8.8575
Epoch [2/3], Step [1201/3236], Loss: 2.0838, Perplexity: 8.0350
Epoch [2/3], Step [1202/3236], Loss: 2.0071, Perplexity: 7.4420
Epoch [2/3], Step [1203/3236], Loss: 2.2490, Perplexity: 9.4782
Epoch [2/3], Step [1204/3236], Loss: 2.6248, Perplexity: 13.8013
Epoch [2/3], Step [1205/3236], Loss: 2.0176, Perplexity: 7.5199
Epoch [2/3], Step [1206/3236], Loss: 2.3489, Perplexity: 10.4736
Epoch [2/3], Step [1207/3236], Loss: 2.1389, Perplexity: 8.4901
Epoch [2/3], Step [1208/3236], Loss: 2.0769, Perplexity: 7.9795
Epoch [2/3], Step [1209/3236], Loss: 2.7953, Perplexity: 16.3682
Epoch [2/3], Step [1210/3236], Loss: 2.3657, Perplexity: 10.6518
Epoch [2/3], Step [1211/3236], Loss: 2.0948, Perplexity: 8.1236
Epoch [2/3], Step [1212/3236], Loss: 2.1772, Perplexity: 8.8220
Epoch [2/3], Step [1213/3236], Loss: 2.2487, Perplexity: 9.4758
Epoch [2/3], Step [1214/3236], Loss: 2.1507, Perplexity: 8.5907
Epoch [2/3], Step [1215/3236], Loss: 2.0887, Perplexity: 8.0745
Epoch [2/3], Step [1216/3236], Loss: 2.1118, Perplexity: 8.2635
Epoch [2/3], Step [1217/3236], Loss: 2.1533, Perplexity: 8.6130
Epoch [2/3], Step [1218/3236], Loss: 2.1580, Perplexity: 8.6540
Epoch [2/3], Step [1219/3236], Loss: 2.0809, Perplexity: 8.0120
Epoch [2/3], Step [1220/3236], Loss: 2.1248, Perplexity: 8.3711
Epoch [2/3], Step [1221/3236], Loss: 2.4742, Perplexity: 11.8718
Epoch [2/3], Step [1222/3236], Loss: 2.2040, Perplexity: 9.0611
Epoch [2/3], Step [1223/3236], Loss: 2.0786, Perplexity: 7.9936
Epoch [2/3], Step [1224/3236], Loss: 2.1255, Perplexity: 8.3774
Epoch [2/3], Step [1225/3236], Loss: 2.6241, Perplexity: 13.7918
Epoch [2/3], Step [1226/3236], Loss: 2.3739, Perplexity: 10.7387
Epoch [2/3], Step [1227/3236], Loss: 2.0139, Perplexity: 7.4924
Epoch [2/3], Step [1228/3236], Loss: 2.1731, Perplexity: 8.7853
Epoch [2/3], Step [1229/3236], Loss: 2.0723, Perplexity: 7.9428
Epoch [2/3], Step [1230/3236], Loss: 2.1037, Perplexity: 8.1961
Epoch [2/3], Step [1231/3236], Loss: 2.2500, Perplexity: 9.4878
Epoch [2/3], Step [1232/3236], Loss: 2.0360, Perplexity: 7.6599
Epoch [2/3], Step [1233/3236], Loss: 2.4202, Perplexity: 11.2484
Epoch [2/3], Step [1234/3236], Loss: 2.0854, Perplexity: 8.0477
Epoch [2/3], Step [1235/3236], Loss: 2.0479, Perplexity: 7.7516
Epoch [2/3], Step [1236/3236], Loss: 2.1137, Perplexity: 8.2787
Epoch [2/3], Step [1237/3236], Loss: 2.1054, Perplexity: 8.2105
Epoch [2/3], Step [1238/3236], Loss: 2.1999, Perplexity: 9.0237
Epoch [2/3], Step [1239/3236], Loss: 2.1834, Perplexity: 8.8766
Epoch [2/3], Step [1240/3236], Loss: 2.2511, Perplexity: 9.4982
Epoch [2/3], Step [1241/3236], Loss: 2.2168, Perplexity: 9.1778
Epoch [2/3], Step [1242/3236], Loss: 2.5472, Perplexity: 12.7707
Epoch [2/3], Step [1243/3236], Loss: 2.1322, Perplexity: 8.4335
Epoch [2/3], Step [1244/3236], Loss: 2.3042, Perplexity: 10.0163
Epoch [2/3], Step [1245/3236], Loss: 2.1732, Perplexity: 8.7866
Epoch [2/3], Step [1246/3236], Loss: 2.1144, Perplexity: 8.2843
Epoch [2/3], Step [1247/3236], Loss: 2.2049, Perplexity: 9.0693
Epoch [2/3], Step [1248/3236], Loss: 2.0849, Perplexity: 8.0439
Epoch [2/3], Step [1249/3236], Loss: 2.2255, Perplexity: 9.2578
Epoch [2/3], Step [1250/3236], Loss: 2.0104, Perplexity: 7.4663
Epoch [2/3], Step [1251/3236], Loss: 2.1904, Perplexity: 8.9384
Epoch [2/3], Step [1252/3236], Loss: 2.0849, Perplexity: 8.0435
Epoch [2/3], Step [1253/3236], Loss: 2.1552, Perplexity: 8.6293
Epoch [2/3], Step [1254/3236], Loss: 2.2149, Perplexity: 9.1607
Epoch [2/3], Step [1255/3236], Loss: 2.2357, Perplexity: 9.3533
Epoch [2/3], Step [1256/3236], Loss: 2.0829, Perplexity: 8.0278
Epoch [2/3], Step [1257/3236], Loss: 2.0378, Perplexity: 7.6735
Epoch [2/3], Step [1258/3236], Loss: 2.1963, Perplexity: 8.9917
Epoch [2/3], Step [1259/3236], Loss: 2.1179, Perplexity: 8.3137
Epoch [2/3], Step [1260/3236], Loss: 2.1592, Perplexity: 8.6644
Epoch [2/3], Step [1261/3236], Loss: 2.1989, Perplexity: 9.0148
Epoch [2/3], Step [1262/3236], Loss: 2.0983, Perplexity: 8.1522
Epoch [2/3], Step [1263/3236], Loss: 2.1477, Perplexity: 8.5648
Epoch [2/3], Step [1264/3236], Loss: 2.7597, Perplexity: 15.7946
Epoch [2/3], Step [1265/3236], Loss: 2.1342, Perplexity: 8.4502
Epoch [2/3], Step [1266/3236], Loss: 2.1434, Perplexity: 8.5283
Epoch [2/3], Step [1267/3236], Loss: 2.0718, Perplexity: 7.9389
Epoch [2/3], Step [1268/3236], Loss: 2.0715, Perplexity: 7.9370
Epoch [2/3], Step [1269/3236], Loss: 2.0895, Perplexity: 8.0806
Epoch [2/3], Step [1270/3236], Loss: 2.4068, Perplexity: 11.0989
Epoch [2/3], Step [1271/3236], Loss: 2.2748, Perplexity: 9.7257
Epoch [2/3], Step [1272/3236], Loss: 2.4300, Perplexity: 11.3594
Epoch [2/3], Step [1273/3236], Loss: 2.1746, Perplexity: 8.7986
Epoch [2/3], Step [1274/3236], Loss: 2.3583, Perplexity: 10.5733
Epoch [2/3], Step [1275/3236], Loss: 2.0836, Perplexity: 8.0330
Epoch [2/3], Step [1276/3236], Loss: 2.2924, Perplexity: 9.8983
Epoch [2/3], Step [1277/3236], Loss: 2.0879, Perplexity: 8.0682
Epoch [2/3], Step [1278/3236], Loss: 2.0284, Perplexity: 7.6022
Epoch [2/3], Step [1279/3236], Loss: 2.4457, Perplexity: 11.5388
Epoch [2/3], Step [1280/3236], Loss: 2.0712, Perplexity: 7.9347
Epoch [2/3], Step [1281/3236], Loss: 2.9392, Perplexity: 18.9001
Epoch [2/3], Step [1282/3236], Loss: 2.1646, Perplexity: 8.7108
Epoch [2/3], Step [1283/3236], Loss: 1.9246, Perplexity: 6.8526
Epoch [2/3], Step [1284/3236], Loss: 2.1335, Perplexity: 8.4441
Epoch [2/3], Step [1285/3236], Loss: 2.0491, Perplexity: 7.7608
Epoch [2/3], Step [1286/3236], Loss: 2.0269, Perplexity: 7.5907
Epoch [2/3], Step [1287/3236], Loss: 2.0993, Perplexity: 8.1601
Epoch [2/3], Step [1288/3236], Loss: 2.2244, Perplexity: 9.2475
Epoch [2/3], Step [1289/3236], Loss: 2.2475, Perplexity: 9.4643
Epoch [2/3], Step [1290/3236], Loss: 2.9000, Perplexity: 18.1742
Epoch [2/3], Step [1291/3236], Loss: 2.2756, Perplexity: 9.7342
Epoch [2/3], Step [1292/3236], Loss: 2.1781, Perplexity: 8.8291
Epoch [2/3], Step [1293/3236], Loss: 2.2379, Perplexity: 9.3739
Epoch [2/3], Step [1294/3236], Loss: 2.0925, Perplexity: 8.1051
Epoch [2/3], Step [1295/3236], Loss: 2.0844, Perplexity: 8.0396
Epoch [2/3], Step [1296/3236], Loss: 2.1701, Perplexity: 8.7590
Epoch [2/3], Step [1297/3236], Loss: 2.0468, Perplexity: 7.7428
Epoch [2/3], Step [1298/3236], Loss: 2.4397, Perplexity: 11.4698
Epoch [2/3], Step [1299/3236], Loss: 2.0686, Perplexity: 7.9141
Epoch [2/3], Step [1300/3236], Loss: 2.5956, Perplexity: 13.4042
Epoch [2/3], Step [1301/3236], Loss: 2.2543, Perplexity: 9.5290
Epoch [2/3], Step [1302/3236], Loss: 2.2019, Perplexity: 9.0426
Epoch [2/3], Step [1303/3236], Loss: 2.1722, Perplexity: 8.7776
Epoch [2/3], Step [1304/3236], Loss: 2.2306, Perplexity: 9.3050
Epoch [2/3], Step [1305/3236], Loss: 2.5519, Perplexity: 12.8316
Epoch [2/3], Step [1306/3236], Loss: 2.0632, Perplexity: 7.8713
Epoch [2/3], Step [1307/3236], Loss: 2.0670, Perplexity: 7.9007
Epoch [2/3], Step [1308/3236], Loss: 2.2783, Perplexity: 9.7602
Epoch [2/3], Step [1309/3236], Loss: 2.2209, Perplexity: 9.2159
Epoch [2/3], Step [1310/3236], Loss: 2.1507, Perplexity: 8.5906
Epoch [2/3], Step [1311/3236], Loss: 2.6814, Perplexity: 14.6054
Epoch [2/3], Step [1312/3236], Loss: 2.2524, Perplexity: 9.5104
Epoch [2/3], Step [1313/3236], Loss: 2.0566, Perplexity: 7.8196
Epoch [2/3], Step [1314/3236], Loss: 2.3860, Perplexity: 10.8697
Epoch [2/3], Step [1315/3236], Loss: 2.1440, Perplexity: 8.5334
Epoch [2/3], Step [1316/3236], Loss: 2.1859, Perplexity: 8.8982
Epoch [2/3], Step [1317/3236], Loss: 2.0536, Perplexity: 7.7959
Epoch [2/3], Step [1318/3236], Loss: 2.0567, Perplexity: 7.8200
Epoch [2/3], Step [1319/3236], Loss: 2.4504, Perplexity: 11.5932
Epoch [2/3], Step [1320/3236], Loss: 2.3457, Perplexity: 10.4402
Epoch [2/3], Step [1321/3236], Loss: 2.1133, Perplexity: 8.2757
Epoch [2/3], Step [1322/3236], Loss: 2.3407, Perplexity: 10.3889
Epoch [2/3], Step [1323/3236], Loss: 2.2099, Perplexity: 9.1144
Epoch [2/3], Step [1324/3236], Loss: 2.0812, Perplexity: 8.0142
Epoch [2/3], Step [1325/3236], Loss: 2.3827, Perplexity: 10.8337
Epoch [2/3], Step [1326/3236], Loss: 2.0984, Perplexity: 8.1530
Epoch [2/3], Step [1327/3236], Loss: 2.0695, Perplexity: 7.9209
Epoch [2/3], Step [1328/3236], Loss: 2.1905, Perplexity: 8.9398
Epoch [2/3], Step [1329/3236], Loss: 2.0711, Perplexity: 7.9332
Epoch [2/3], Step [1330/3236], Loss: 2.2912, Perplexity: 9.8866
Epoch [2/3], Step [1331/3236], Loss: 2.0476, Perplexity: 7.7489
Epoch [2/3], Step [1332/3236], Loss: 2.0485, Perplexity: 7.7562
Epoch [2/3], Step [1333/3236], Loss: 2.1331, Perplexity: 8.4412
Epoch [2/3], Step [1334/3236], Loss: 2.1646, Perplexity: 8.7110
Epoch [2/3], Step [1335/3236], Loss: 2.6319, Perplexity: 13.9000
Epoch [2/3], Step [1336/3236], Loss: 2.5509, Perplexity: 12.8180
Epoch [2/3], Step [1337/3236], Loss: 2.1770, Perplexity: 8.8199
Epoch [2/3], Step [1338/3236], Loss: 2.0644, Perplexity: 7.8808
Epoch [2/3], Step [1339/3236], Loss: 2.1899, Perplexity: 8.9342
Epoch [2/3], Step [1340/3236], Loss: 3.0296, Perplexity: 20.6892
Epoch [2/3], Step [1341/3236], Loss: 2.1697, Perplexity: 8.7555
Epoch [2/3], Step [1342/3236], Loss: 2.1425, Perplexity: 8.5206
Epoch [2/3], Step [1343/3236], Loss: 2.7606, Perplexity: 15.8086
Epoch [2/3], Step [1344/3236], Loss: 2.0506, Perplexity: 7.7729
Epoch [2/3], Step [1345/3236], Loss: 2.5486, Perplexity: 12.7895
Epoch [2/3], Step [1346/3236], Loss: 2.1331, Perplexity: 8.4414
Epoch [2/3], Step [1347/3236], Loss: 2.4753, Perplexity: 11.8855
Epoch [2/3], Step [1348/3236], Loss: 2.0554, Perplexity: 7.8099
Epoch [2/3], Step [1349/3236], Loss: 2.6368, Perplexity: 13.9683
Epoch [2/3], Step [1350/3236], Loss: 2.1009, Perplexity: 8.1732
Epoch [2/3], Step [1351/3236], Loss: 2.5735, Perplexity: 13.1114
Epoch [2/3], Step [1352/3236], Loss: 2.3174, Perplexity: 10.1496
Epoch [2/3], Step [1353/3236], Loss: 2.1671, Perplexity: 8.7332
Epoch [2/3], Step [1354/3236], Loss: 2.1285, Perplexity: 8.4024
Epoch [2/3], Step [1355/3236], Loss: 2.1889, Perplexity: 8.9250
Epoch [2/3], Step [1356/3236], Loss: 2.2702, Perplexity: 9.6818
Epoch [2/3], Step [1357/3236], Loss: 2.1785, Perplexity: 8.8330
Epoch [2/3], Step [1358/3236], Loss: 2.1337, Perplexity: 8.4461
Epoch [2/3], Step [1359/3236], Loss: 2.2875, Perplexity: 9.8504
Epoch [2/3], Step [1360/3236], Loss: 2.4429, Perplexity: 11.5059
Epoch [2/3], Step [1361/3236], Loss: 2.2182, Perplexity: 9.1908
Epoch [2/3], Step [1362/3236], Loss: 2.1146, Perplexity: 8.2865
Epoch [2/3], Step [1363/3236], Loss: 2.0943, Perplexity: 8.1199
Epoch [2/3], Step [1364/3236], Loss: 2.1782, Perplexity: 8.8305
Epoch [2/3], Step [1365/3236], Loss: 2.0432, Perplexity: 7.7156
Epoch [2/3], Step [1366/3236], Loss: 2.1079, Perplexity: 8.2309
Epoch [2/3], Step [1367/3236], Loss: 2.1204, Perplexity: 8.3345
Epoch [2/3], Step [1368/3236], Loss: 2.1857, Perplexity: 8.8965
Epoch [2/3], Step [1369/3236], Loss: 2.0774, Perplexity: 7.9839
Epoch [2/3], Step [1370/3236], Loss: 2.1011, Perplexity: 8.1753
Epoch [2/3], Step [1371/3236], Loss: 2.0102, Perplexity: 7.4648
Epoch [2/3], Step [1372/3236], Loss: 2.8368, Perplexity: 17.0611
Epoch [2/3], Step [1373/3236], Loss: 2.1703, Perplexity: 8.7609
Epoch [2/3], Step [1374/3236], Loss: 2.1858, Perplexity: 8.8979
Epoch [2/3], Step [1375/3236], Loss: 2.1695, Perplexity: 8.7537
Epoch [2/3], Step [1376/3236], Loss: 2.1565, Perplexity: 8.6408
Epoch [2/3], Step [1377/3236], Loss: 2.2278, Perplexity: 9.2790
Epoch [2/3], Step [1378/3236], Loss: 2.0871, Perplexity: 8.0617
Epoch [2/3], Step [1379/3236], Loss: 2.0801, Perplexity: 8.0053
Epoch [2/3], Step [1380/3236], Loss: 2.1395, Perplexity: 8.4951
Epoch [2/3], Step [1381/3236], Loss: 2.2709, Perplexity: 9.6885
Epoch [2/3], Step [1382/3236], Loss: 2.2350, Perplexity: 9.3463
Epoch [2/3], Step [1383/3236], Loss: 2.5201, Perplexity: 12.4297
Epoch [2/3], Step [1384/3236], Loss: 2.0368, Perplexity: 7.6661
Epoch [2/3], Step [1385/3236], Loss: 2.0996, Perplexity: 8.1628
Epoch [2/3], Step [1386/3236], Loss: 2.1777, Perplexity: 8.8256
Epoch [2/3], Step [1387/3236], Loss: 2.2355, Perplexity: 9.3508
Epoch [2/3], Step [1388/3236], Loss: 2.0979, Perplexity: 8.1494
Epoch [2/3], Step [1389/3236], Loss: 2.1455, Perplexity: 8.5464
Epoch [2/3], Step [1390/3236], Loss: 2.2414, Perplexity: 9.4063
Epoch [2/3], Step [1391/3236], Loss: 2.3795, Perplexity: 10.7991
Epoch [2/3], Step [1392/3236], Loss: 2.3125, Perplexity: 10.0999
Epoch [2/3], Step [1393/3236], Loss: 2.1152, Perplexity: 8.2910
Epoch [2/3], Step [1394/3236], Loss: 2.1292, Perplexity: 8.4082
Epoch [2/3], Step [1395/3236], Loss: 2.1032, Perplexity: 8.1927
Epoch [2/3], Step [1396/3236], Loss: 2.0546, Perplexity: 7.8038
Epoch [2/3], Step [1397/3236], Loss: 2.4809, Perplexity: 11.9518
Epoch [2/3], Step [1398/3236], Loss: 2.1138, Perplexity: 8.2794
Epoch [2/3], Step [1399/3236], Loss: 1.9764, Perplexity: 7.2167
Epoch [2/3], Step [1400/3236], Loss: 2.1604, Perplexity: 8.6748
Epoch [2/3], Step [1401/3236], Loss: 2.1995, Perplexity: 9.0202
Epoch [2/3], Step [1402/3236], Loss: 2.1582, Perplexity: 8.6557
Epoch [2/3], Step [1403/3236], Loss: 2.2222, Perplexity: 9.2273
Epoch [2/3], Step [1404/3236], Loss: 2.1710, Perplexity: 8.7670
Epoch [2/3], Step [1405/3236], Loss: 2.0882, Perplexity: 8.0701
Epoch [2/3], Step [1406/3236], Loss: 2.0972, Perplexity: 8.1433
Epoch [2/3], Step [1407/3236], Loss: 2.0581, Perplexity: 7.8308
Epoch [2/3], Step [1408/3236], Loss: 2.0536, Perplexity: 7.7959
Epoch [2/3], Step [1409/3236], Loss: 2.2392, Perplexity: 9.3856
Epoch [2/3], Step [1410/3236], Loss: 2.1366, Perplexity: 8.4709
Epoch [2/3], Step [1411/3236], Loss: 1.9597, Perplexity: 7.0972
Epoch [2/3], Step [1412/3236], Loss: 2.1363, Perplexity: 8.4681
Epoch [2/3], Step [1413/3236], Loss: 2.1300, Perplexity: 8.4145
Epoch [2/3], Step [1414/3236], Loss: 2.0540, Perplexity: 7.7992
Epoch [2/3], Step [1415/3236], Loss: 2.0734, Perplexity: 7.9518
Epoch [2/3], Step [1416/3236], Loss: 2.0867, Perplexity: 8.0585
Epoch [2/3], Step [1417/3236], Loss: 2.2247, Perplexity: 9.2505
Epoch [2/3], Step [1418/3236], Loss: 2.0835, Perplexity: 8.0324
Epoch [2/3], Step [1419/3236], Loss: 2.3154, Perplexity: 10.1290
Epoch [2/3], Step [1420/3236], Loss: 2.1475, Perplexity: 8.5637
Epoch [2/3], Step [1421/3236], Loss: 2.4202, Perplexity: 11.2484
Epoch [2/3], Step [1422/3236], Loss: 2.5492, Perplexity: 12.7966
Epoch [2/3], Step [1423/3236], Loss: 2.0338, Perplexity: 7.6432
Epoch [2/3], Step [1424/3236], Loss: 2.0762, Perplexity: 7.9744
Epoch [2/3], Step [1425/3236], Loss: 2.3534, Perplexity: 10.5212
Epoch [2/3], Step [1426/3236], Loss: 2.1731, Perplexity: 8.7855
Epoch [2/3], Step [1427/3236], Loss: 2.1640, Perplexity: 8.7059
Epoch [2/3], Step [1428/3236], Loss: 2.3721, Perplexity: 10.7198
Epoch [2/3], Step [1429/3236], Loss: 2.1546, Perplexity: 8.6240
Epoch [2/3], Step [1430/3236], Loss: 2.1918, Perplexity: 8.9513
Epoch [2/3], Step [1431/3236], Loss: 2.1410, Perplexity: 8.5077
Epoch [2/3], Step [1432/3236], Loss: 2.3399, Perplexity: 10.3799
Epoch [2/3], Step [1433/3236], Loss: 2.8911, Perplexity: 18.0130
Epoch [2/3], Step [1434/3236], Loss: 2.2652, Perplexity: 9.6330
Epoch [2/3], Step [1435/3236], Loss: 2.4209, Perplexity: 11.2554
Epoch [2/3], Step [1436/3236], Loss: 2.0503, Perplexity: 7.7705
Epoch [2/3], Step [1437/3236], Loss: 2.2197, Perplexity: 9.2050
Epoch [2/3], Step [1438/3236], Loss: 2.1480, Perplexity: 8.5673
Epoch [2/3], Step [1439/3236], Loss: 2.2201, Perplexity: 9.2081
Epoch [2/3], Step [1440/3236], Loss: 2.8303, Perplexity: 16.9500
Epoch [2/3], Step [1441/3236], Loss: 1.9898, Perplexity: 7.3141
Epoch [2/3], Step [1442/3236], Loss: 2.1691, Perplexity: 8.7504
Epoch [2/3], Step [1443/3236], Loss: 2.0212, Perplexity: 7.5472
Epoch [2/3], Step [1444/3236], Loss: 2.0751, Perplexity: 7.9657
Epoch [2/3], Step [1445/3236], Loss: 2.4393, Perplexity: 11.4655
Epoch [2/3], Step [1446/3236], Loss: 1.9932, Perplexity: 7.3391
Epoch [2/3], Step [1447/3236], Loss: 2.0886, Perplexity: 8.0734
Epoch [2/3], Step [1448/3236], Loss: 2.1386, Perplexity: 8.4871
Epoch [2/3], Step [1449/3236], Loss: 2.7869, Perplexity: 16.2313
Epoch [2/3], Step [1450/3236], Loss: 2.1261, Perplexity: 8.3823
Epoch [2/3], Step [1451/3236], Loss: 2.2455, Perplexity: 9.4450
Epoch [2/3], Step [1452/3236], Loss: 2.2568, Perplexity: 9.5528
Epoch [2/3], Step [1453/3236], Loss: 2.8549, Perplexity: 17.3733
Epoch [2/3], Step [1454/3236], Loss: 2.3413, Perplexity: 10.3950
Epoch [2/3], Step [1455/3236], Loss: 2.1590, Perplexity: 8.6622
Epoch [2/3], Step [1456/3236], Loss: 3.6746, Perplexity: 39.4325
Epoch [2/3], Step [1457/3236], Loss: 2.0711, Perplexity: 7.9338
Epoch [2/3], Step [1458/3236], Loss: 2.0847, Perplexity: 8.0419
Epoch [2/3], Step [1459/3236], Loss: 2.0799, Perplexity: 8.0035
Epoch [2/3], Step [1460/3236], Loss: 2.2374, Perplexity: 9.3686
Epoch [2/3], Step [1461/3236], Loss: 2.4482, Perplexity: 11.5678
Epoch [2/3], Step [1462/3236], Loss: 2.1599, Perplexity: 8.6707
Epoch [2/3], Step [1463/3236], Loss: 2.0484, Perplexity: 7.7557
Epoch [2/3], Step [1464/3236], Loss: 2.3083, Perplexity: 10.0572
Epoch [2/3], Step [1465/3236], Loss: 2.0906, Perplexity: 8.0894
Epoch [2/3], Step [1466/3236], Loss: 2.4475, Perplexity: 11.5597
Epoch [2/3], Step [1467/3236], Loss: 3.1238, Perplexity: 22.7321
Epoch [2/3], Step [1468/3236], Loss: 2.1656, Perplexity: 8.7201
Epoch [2/3], Step [1469/3236], Loss: 2.1311, Perplexity: 8.4241
Epoch [2/3], Step [1470/3236], Loss: 2.0357, Perplexity: 7.6575
Epoch [2/3], Step [1471/3236], Loss: 2.1783, Perplexity: 8.8313
Epoch [2/3], Step [1472/3236], Loss: 2.2059, Perplexity: 9.0786
Epoch [2/3], Step [1473/3236], Loss: 2.0282, Perplexity: 7.6004
Epoch [2/3], Step [1474/3236], Loss: 2.0638, Perplexity: 7.8757
Epoch [2/3], Step [1475/3236], Loss: 2.2046, Perplexity: 9.0669
Epoch [2/3], Step [1476/3236], Loss: 2.1495, Perplexity: 8.5808
Epoch [2/3], Step [1477/3236], Loss: 2.0910, Perplexity: 8.0931
Epoch [2/3], Step [1478/3236], Loss: 2.0854, Perplexity: 8.0478
Epoch [2/3], Step [1479/3236], Loss: 2.1087, Perplexity: 8.2378
Epoch [2/3], Step [1480/3236], Loss: 1.9971, Perplexity: 7.3679
Epoch [2/3], Step [1481/3236], Loss: 2.0603, Perplexity: 7.8480
Epoch [2/3], Step [1482/3236], Loss: 2.2134, Perplexity: 9.1463
Epoch [2/3], Step [1483/3236], Loss: 2.1687, Perplexity: 8.7465
Epoch [2/3], Step [1484/3236], Loss: 2.2546, Perplexity: 9.5313
Epoch [2/3], Step [1485/3236], Loss: 2.2697, Perplexity: 9.6768
Epoch [2/3], Step [1486/3236], Loss: 2.1766, Perplexity: 8.8165
Epoch [2/3], Step [1487/3236], Loss: 2.0029, Perplexity: 7.4107
Epoch [2/3], Step [1488/3236], Loss: 2.2146, Perplexity: 9.1577
Epoch [2/3], Step [1489/3236], Loss: 2.1560, Perplexity: 8.6364
Epoch [2/3], Step [1490/3236], Loss: 2.0907, Perplexity: 8.0906
Epoch [2/3], Step [1491/3236], Loss: 2.1720, Perplexity: 8.7758
Epoch [2/3], Step [1492/3236], Loss: 2.0332, Perplexity: 7.6381
Epoch [2/3], Step [1493/3236], Loss: 2.0221, Perplexity: 7.5538
Epoch [2/3], Step [1494/3236], Loss: 2.2708, Perplexity: 9.6868
Epoch [2/3], Step [1495/3236], Loss: 2.0798, Perplexity: 8.0030
Epoch [2/3], Step [1496/3236], Loss: 2.0352, Perplexity: 7.6538
Epoch [2/3], Step [1497/3236], Loss: 2.2348, Perplexity: 9.3448
Epoch [2/3], Step [1498/3236], Loss: 2.1432, Perplexity: 8.5267
Epoch [2/3], Step [1499/3236], Loss: 2.4455, Perplexity: 11.5365
Epoch [2/3], Step [1500/3236], Loss: 2.2215, Perplexity: 9.2214
Epoch [2/3], Step [1501/3236], Loss: 2.1497, Perplexity: 8.5821
Epoch [2/3], Step [1502/3236], Loss: 2.1307, Perplexity: 8.4207
Epoch [2/3], Step [1503/3236], Loss: 2.1123, Perplexity: 8.2676
Epoch [2/3], Step [1504/3236], Loss: 2.3817, Perplexity: 10.8235
Epoch [2/3], Step [1505/3236], Loss: 2.1384, Perplexity: 8.4854
Epoch [2/3], Step [1506/3236], Loss: 2.1740, Perplexity: 8.7930
Epoch [2/3], Step [1507/3236], Loss: 2.0536, Perplexity: 7.7955
Epoch [2/3], Step [1508/3236], Loss: 2.1411, Perplexity: 8.5084
Epoch [2/3], Step [1509/3236], Loss: 2.0378, Perplexity: 7.6739
Epoch [2/3], Step [1510/3236], Loss: 2.1376, Perplexity: 8.4791
Epoch [2/3], Step [1511/3236], Loss: 2.0569, Perplexity: 7.8221
Epoch [2/3], Step [1512/3236], Loss: 2.4592, Perplexity: 11.6960
Epoch [2/3], Step [1513/3236], Loss: 2.0288, Perplexity: 7.6049
Epoch [2/3], Step [1514/3236], Loss: 2.2613, Perplexity: 9.5958
Epoch [2/3], Step [1515/3236], Loss: 2.4582, Perplexity: 11.6835
Epoch [2/3], Step [1516/3236], Loss: 2.4802, Perplexity: 11.9442
Epoch [2/3], Step [1517/3236], Loss: 2.2524, Perplexity: 9.5108
Epoch [2/3], Step [1518/3236], Loss: 1.9908, Perplexity: 7.3217
Epoch [2/3], Step [1519/3236], Loss: 2.1727, Perplexity: 8.7818
Epoch [2/3], Step [1520/3236], Loss: 2.3188, Perplexity: 10.1633
Epoch [2/3], Step [1521/3236], Loss: 2.3869, Perplexity: 10.8797
Epoch [2/3], Step [1522/3236], Loss: 2.0958, Perplexity: 8.1316
Epoch [2/3], Step [1523/3236], Loss: 2.7485, Perplexity: 15.6185
Epoch [2/3], Step [1524/3236], Loss: 2.2908, Perplexity: 9.8831
Epoch [2/3], Step [1525/3236], Loss: 2.2636, Perplexity: 9.6172
Epoch [2/3], Step [1526/3236], Loss: 1.9782, Perplexity: 7.2297
Epoch [2/3], Step [1527/3236], Loss: 2.0670, Perplexity: 7.9007
Epoch [2/3], Step [1528/3236], Loss: 2.2501, Perplexity: 9.4883
Epoch [2/3], Step [1529/3236], Loss: 2.1139, Perplexity: 8.2804
Epoch [2/3], Step [1530/3236], Loss: 2.2400, Perplexity: 9.3934
Epoch [2/3], Step [1531/3236], Loss: 2.0428, Perplexity: 7.7124
Epoch [2/3], Step [1532/3236], Loss: 2.2385, Perplexity: 9.3797
Epoch [2/3], Step [1533/3236], Loss: 2.1029, Perplexity: 8.1897
Epoch [2/3], Step [1534/3236], Loss: 2.4149, Perplexity: 11.1888
Epoch [2/3], Step [1535/3236], Loss: 2.1206, Perplexity: 8.3357
Epoch [2/3], Step [1536/3236], Loss: 2.0428, Perplexity: 7.7125
Epoch [2/3], Step [1537/3236], Loss: 2.0327, Perplexity: 7.6350
Epoch [2/3], Step [1538/3236], Loss: 2.3296, Perplexity: 10.2742
Epoch [2/3], Step [1539/3236], Loss: 2.0379, Perplexity: 7.6742
Epoch [2/3], Step [1540/3236], Loss: 2.4822, Perplexity: 11.9670
Epoch [2/3], Step [1541/3236], Loss: 2.1607, Perplexity: 8.6772
Epoch [2/3], Step [1542/3236], Loss: 2.0919, Perplexity: 8.1000
Epoch [2/3], Step [1543/3236], Loss: 2.0548, Perplexity: 7.8054
Epoch [2/3], Step [1544/3236], Loss: 2.0361, Perplexity: 7.6609
Epoch [2/3], Step [1545/3236], Loss: 2.0329, Perplexity: 7.6361
Epoch [2/3], Step [1546/3236], Loss: 2.0733, Perplexity: 7.9513
Epoch [2/3], Step [1547/3236], Loss: 2.2142, Perplexity: 9.1539
Epoch [2/3], Step [1548/3236], Loss: 2.3141, Perplexity: 10.1154
Epoch [2/3], Step [1549/3236], Loss: 2.0243, Perplexity: 7.5711
Epoch [2/3], Step [1550/3236], Loss: 2.1293, Perplexity: 8.4087
Epoch [2/3], Step [1551/3236], Loss: 2.1638, Perplexity: 8.7045
Epoch [2/3], Step [1552/3236], Loss: 2.0942, Perplexity: 8.1186
Epoch [2/3], Step [1553/3236], Loss: 2.1743, Perplexity: 8.7959
Epoch [2/3], Step [1554/3236], Loss: 2.0919, Perplexity: 8.0999
Epoch [2/3], Step [1555/3236], Loss: 2.1067, Perplexity: 8.2213
Epoch [2/3], Step [1556/3236], Loss: 2.0423, Perplexity: 7.7081
Epoch [2/3], Step [1557/3236], Loss: 2.1302, Perplexity: 8.4164
Epoch [2/3], Step [1558/3236], Loss: 2.3429, Perplexity: 10.4116
Epoch [2/3], Step [1559/3236], Loss: 2.0994, Perplexity: 8.1614
Epoch [2/3], Step [1560/3236], Loss: 2.1853, Perplexity: 8.8937
Epoch [2/3], Step [1561/3236], Loss: 2.2337, Perplexity: 9.3345
Epoch [2/3], Step [1562/3236], Loss: 2.1387, Perplexity: 8.4887
Epoch [2/3], Step [1563/3236], Loss: 2.2497, Perplexity: 9.4854
Epoch [2/3], Step [1564/3236], Loss: 2.0680, Perplexity: 7.9092
Epoch [2/3], Step [1565/3236], Loss: 2.2412, Perplexity: 9.4050
Epoch [2/3], Step [1566/3236], Loss: 2.1320, Perplexity: 8.4313
Epoch [2/3], Step [1567/3236], Loss: 2.2252, Perplexity: 9.2554
Epoch [2/3], Step [1568/3236], Loss: 2.0726, Perplexity: 7.9458
Epoch [2/3], Step [1569/3236], Loss: 2.2076, Perplexity: 9.0941
Epoch [2/3], Step [1570/3236], Loss: 1.9424, Perplexity: 6.9755
Epoch [2/3], Step [1571/3236], Loss: 2.0901, Perplexity: 8.0859
Epoch [2/3], Step [1572/3236], Loss: 2.0868, Perplexity: 8.0588
Epoch [2/3], Step [1573/3236], Loss: 2.0969, Perplexity: 8.1406
Epoch [2/3], Step [1574/3236], Loss: 2.3155, Perplexity: 10.1296
Epoch [2/3], Step [1575/3236], Loss: 2.0312, Perplexity: 7.6235
Epoch [2/3], Step [1576/3236], Loss: 2.0797, Perplexity: 8.0024
Epoch [2/3], Step [1577/3236], Loss: 2.3450, Perplexity: 10.4334
Epoch [2/3], Step [1578/3236], Loss: 2.2388, Perplexity: 9.3820
Epoch [2/3], Step [1579/3236], Loss: 2.0862, Perplexity: 8.0543
Epoch [2/3], Step [1580/3236], Loss: 2.1599, Perplexity: 8.6699
Epoch [2/3], Step [1581/3236], Loss: 2.4735, Perplexity: 11.8633
Epoch [2/3], Step [1582/3236], Loss: 2.0159, Perplexity: 7.5073
Epoch [2/3], Step [1583/3236], Loss: 1.9744, Perplexity: 7.2021
Epoch [2/3], Step [1584/3236], Loss: 2.2165, Perplexity: 9.1755
Epoch [2/3], Step [1585/3236], Loss: 2.0380, Perplexity: 7.6754
Epoch [2/3], Step [1586/3236], Loss: 2.0442, Perplexity: 7.7230
Epoch [2/3], Step [1587/3236], Loss: 2.0120, Perplexity: 7.4781
Epoch [2/3], Step [1588/3236], Loss: 2.2242, Perplexity: 9.2458
Epoch [2/3], Step [1589/3236], Loss: 2.2096, Perplexity: 9.1122
Epoch [2/3], Step [1590/3236], Loss: 2.2833, Perplexity: 9.8088
Epoch [2/3], Step [1591/3236], Loss: 2.1548, Perplexity: 8.6265
Epoch [2/3], Step [1592/3236], Loss: 2.5530, Perplexity: 12.8460
Epoch [2/3], Step [1593/3236], Loss: 2.1235, Perplexity: 8.3604
Epoch [2/3], Step [1594/3236], Loss: 2.0792, Perplexity: 7.9985
Epoch [2/3], Step [1595/3236], Loss: 2.1179, Perplexity: 8.3138
Epoch [2/3], Step [1596/3236], Loss: 2.1533, Perplexity: 8.6131
Epoch [2/3], Step [1597/3236], Loss: 2.1768, Perplexity: 8.8183
Epoch [2/3], Step [1598/3236], Loss: 1.9882, Perplexity: 7.3022
Epoch [2/3], Step [1599/3236], Loss: 2.7217, Perplexity: 15.2056
Epoch [2/3], Step [1600/3236], Loss: 2.1515, Perplexity: 8.5979
Epoch [2/3], Step [1601/3236], Loss: 2.1471, Perplexity: 8.5603
Epoch [2/3], Step [1602/3236], Loss: 1.8405, Perplexity: 6.2996
Epoch [2/3], Step [1603/3236], Loss: 2.1401, Perplexity: 8.5003
Epoch [2/3], Step [1604/3236], Loss: 2.3036, Perplexity: 10.0105
Epoch [2/3], Step [1605/3236], Loss: 2.2061, Perplexity: 9.0803
Epoch [2/3], Step [1606/3236], Loss: 2.1579, Perplexity: 8.6530
Epoch [2/3], Step [1607/3236], Loss: 2.2042, Perplexity: 9.0627
Epoch [2/3], Step [1608/3236], Loss: 2.2743, Perplexity: 9.7212
Epoch [2/3], Step [1609/3236], Loss: 2.0022, Perplexity: 7.4057
Epoch [2/3], Step [1610/3236], Loss: 2.5053, Perplexity: 12.2467
Epoch [2/3], Step [1611/3236], Loss: 1.9652, Perplexity: 7.1364
Epoch [2/3], Step [1612/3236], Loss: 2.2321, Perplexity: 9.3190
Epoch [2/3], Step [1613/3236], Loss: 1.9751, Perplexity: 7.2072
Epoch [2/3], Step [1614/3236], Loss: 2.0996, Perplexity: 8.1633
Epoch [2/3], Step [1615/3236], Loss: 2.1365, Perplexity: 8.4699
Epoch [2/3], Step [1616/3236], Loss: 2.1748, Perplexity: 8.8007
Epoch [2/3], Step [1617/3236], Loss: 2.0639, Perplexity: 7.8769
Epoch [2/3], Step [1618/3236], Loss: 2.0913, Perplexity: 8.0951
Epoch [2/3], Step [1619/3236], Loss: 2.1312, Perplexity: 8.4252
Epoch [2/3], Step [1620/3236], Loss: 2.1925, Perplexity: 8.9576
Epoch [2/3], Step [1621/3236], Loss: 2.0534, Perplexity: 7.7947
Epoch [2/3], Step [1622/3236], Loss: 2.1125, Perplexity: 8.2692
Epoch [2/3], Step [1623/3236], Loss: 2.2676, Perplexity: 9.6564
Epoch [2/3], Step [1624/3236], Loss: 2.0152, Perplexity: 7.5023
Epoch [2/3], Step [1625/3236], Loss: 2.0607, Perplexity: 7.8513
Epoch [2/3], Step [1626/3236], Loss: 2.0989, Perplexity: 8.1576
Epoch [2/3], Step [1627/3236], Loss: 2.1989, Perplexity: 9.0153
Epoch [2/3], Step [1628/3236], Loss: 2.0905, Perplexity: 8.0886
Epoch [2/3], Step [1629/3236], Loss: 2.4800, Perplexity: 11.9415
Epoch [2/3], Step [1630/3236], Loss: 2.0271, Perplexity: 7.5923
Epoch [2/3], Step [1631/3236], Loss: 2.0560, Perplexity: 7.8145
Epoch [2/3], Step [1632/3236], Loss: 2.0608, Perplexity: 7.8526
Epoch [2/3], Step [1633/3236], Loss: 2.1121, Perplexity: 8.2654
Epoch [2/3], Step [1634/3236], Loss: 2.0578, Perplexity: 7.8287
Epoch [2/3], Step [1635/3236], Loss: 2.2122, Perplexity: 9.1355
Epoch [2/3], Step [1636/3236], Loss: 2.0414, Perplexity: 7.7012
Epoch [2/3], Step [1637/3236], Loss: 2.2271, Perplexity: 9.2725
Epoch [2/3], Step [1638/3236], Loss: 2.0369, Perplexity: 7.6665
Epoch [2/3], Step [1639/3236], Loss: 1.9705, Perplexity: 7.1742
Epoch [2/3], Step [1640/3236], Loss: 2.0813, Perplexity: 8.0150
Epoch [2/3], Step [1641/3236], Loss: 2.1666, Perplexity: 8.7282
Epoch [2/3], Step [1642/3236], Loss: 2.1380, Perplexity: 8.4826
Epoch [2/3], Step [1643/3236], Loss: 2.0522, Perplexity: 7.7853
Epoch [2/3], Step [1644/3236], Loss: 2.5104, Perplexity: 12.3099
Epoch [2/3], Step [1645/3236], Loss: 2.3929, Perplexity: 10.9454
Epoch [2/3], Step [1646/3236], Loss: 2.0340, Perplexity: 7.6445
Epoch [2/3], Step [1647/3236], Loss: 2.5871, Perplexity: 13.2912
Epoch [2/3], Step [1648/3236], Loss: 1.9952, Perplexity: 7.3537
Epoch [2/3], Step [1649/3236], Loss: 2.0612, Perplexity: 7.8557
Epoch [2/3], Step [1650/3236], Loss: 2.5442, Perplexity: 12.7324
Epoch [2/3], Step [1651/3236], Loss: 2.6835, Perplexity: 14.6357
Epoch [2/3], Step [1652/3236], Loss: 2.1945, Perplexity: 8.9759
Epoch [2/3], Step [1653/3236], Loss: 2.1622, Perplexity: 8.6905
Epoch [2/3], Step [1654/3236], Loss: 2.3011, Perplexity: 9.9856
Epoch [2/3], Step [1655/3236], Loss: 2.0507, Perplexity: 7.7735
Epoch [2/3], Step [1656/3236], Loss: 1.9959, Perplexity: 7.3589
Epoch [2/3], Step [1657/3236], Loss: 2.1676, Perplexity: 8.7372
Epoch [2/3], Step [1658/3236], Loss: 2.0724, Perplexity: 7.9441
Epoch [2/3], Step [1659/3236], Loss: 2.3943, Perplexity: 10.9600
Epoch [2/3], Step [1660/3236], Loss: 2.0768, Perplexity: 7.9791
Epoch [2/3], Step [1661/3236], Loss: 2.1753, Perplexity: 8.8052
Epoch [2/3], Step [1662/3236], Loss: 2.2248, Perplexity: 9.2512
Epoch [2/3], Step [1663/3236], Loss: 2.1169, Perplexity: 8.3050
Epoch [2/3], Step [1664/3236], Loss: 2.0914, Perplexity: 8.0965
Epoch [2/3], Step [1665/3236], Loss: 2.1554, Perplexity: 8.6312
Epoch [2/3], Step [1666/3236], Loss: 2.0638, Perplexity: 7.8760
Epoch [2/3], Step [1667/3236], Loss: 2.0631, Perplexity: 7.8705
Epoch [2/3], Step [1668/3236], Loss: 2.0469, Perplexity: 7.7438
Epoch [2/3], Step [1669/3236], Loss: 1.9890, Perplexity: 7.3084
Epoch [2/3], Step [1670/3236], Loss: 2.0885, Perplexity: 8.0725
Epoch [2/3], Step [1671/3236], Loss: 2.2508, Perplexity: 9.4953
Epoch [2/3], Step [1672/3236], Loss: 2.2379, Perplexity: 9.3740
Epoch [2/3], Step [1673/3236], Loss: 2.4426, Perplexity: 11.5027
Epoch [2/3], Step [1674/3236], Loss: 2.1472, Perplexity: 8.5605
Epoch [2/3], Step [1675/3236], Loss: 2.1740, Perplexity: 8.7934
Epoch [2/3], Step [1676/3236], Loss: 2.2110, Perplexity: 9.1248
Epoch [2/3], Step [1677/3236], Loss: 2.3214, Perplexity: 10.1902
Epoch [2/3], Step [1678/3236], Loss: 2.1766, Perplexity: 8.8161
Epoch [2/3], Step [1679/3236], Loss: 2.1978, Perplexity: 9.0049
Epoch [2/3], Step [1680/3236], Loss: 2.0377, Perplexity: 7.6727
Epoch [2/3], Step [1681/3236], Loss: 2.0559, Perplexity: 7.8138
Epoch [2/3], Step [1682/3236], Loss: 2.1954, Perplexity: 8.9840
Epoch [2/3], Step [1683/3236], Loss: 2.3216, Perplexity: 10.1923
Epoch [2/3], Step [1684/3236], Loss: 2.0700, Perplexity: 7.9245
Epoch [2/3], Step [1685/3236], Loss: 2.1883, Perplexity: 8.9199
Epoch [2/3], Step [1686/3236], Loss: 2.4190, Perplexity: 11.2349
Epoch [2/3], Step [1687/3236], Loss: 3.2193, Perplexity: 25.0103
Epoch [2/3], Step [1688/3236], Loss: 2.1160, Perplexity: 8.2975
Epoch [2/3], Step [1689/3236], Loss: 2.3699, Perplexity: 10.6968
Epoch [2/3], Step [1690/3236], Loss: 2.2716, Perplexity: 9.6948
Epoch [2/3], Step [1691/3236], Loss: 2.5636, Perplexity: 12.9829
Epoch [2/3], Step [1692/3236], Loss: 2.0865, Perplexity: 8.0565
Epoch [2/3], Step [1693/3236], Loss: 2.0679, Perplexity: 7.9085
Epoch [2/3], Step [1694/3236], Loss: 2.1580, Perplexity: 8.6539
Epoch [2/3], Step [1695/3236], Loss: 2.8285, Perplexity: 16.9198
Epoch [2/3], Step [1696/3236], Loss: 2.0730, Perplexity: 7.9487
Epoch [2/3], Step [1697/3236], Loss: 2.2752, Perplexity: 9.7294
Epoch [2/3], Step [1698/3236], Loss: 2.2228, Perplexity: 9.2330
Epoch [2/3], Step [1699/3236], Loss: 2.0282, Perplexity: 7.6002
Epoch [2/3], Step [1700/3236], Loss: 2.1572, Perplexity: 8.6466
Epoch [2/3], Step [1701/3236], Loss: 2.4251, Perplexity: 11.3034
Epoch [2/3], Step [1702/3236], Loss: 2.2356, Perplexity: 9.3521
Epoch [2/3], Step [1703/3236], Loss: 2.2092, Perplexity: 9.1081
Epoch [2/3], Step [1704/3236], Loss: 2.0423, Perplexity: 7.7080
Epoch [2/3], Step [1705/3236], Loss: 1.9963, Perplexity: 7.3616
Epoch [2/3], Step [1706/3236], Loss: 2.4213, Perplexity: 11.2607
Epoch [2/3], Step [1707/3236], Loss: 2.1940, Perplexity: 8.9707
Epoch [2/3], Step [1708/3236], Loss: 2.2188, Perplexity: 9.1965
Epoch [2/3], Step [1709/3236], Loss: 2.1959, Perplexity: 8.9881
Epoch [2/3], Step [1710/3236], Loss: 2.5388, Perplexity: 12.6640
Epoch [2/3], Step [1711/3236], Loss: 2.2525, Perplexity: 9.5118
Epoch [2/3], Step [1712/3236], Loss: 2.1473, Perplexity: 8.5621
Epoch [2/3], Step [1713/3236], Loss: 2.0634, Perplexity: 7.8729
Epoch [2/3], Step [1714/3236], Loss: 2.0828, Perplexity: 8.0273
Epoch [2/3], Step [1715/3236], Loss: 2.2619, Perplexity: 9.6009
Epoch [2/3], Step [1716/3236], Loss: 2.1168, Perplexity: 8.3043
Epoch [2/3], Step [1717/3236], Loss: 2.1792, Perplexity: 8.8389
Epoch [2/3], Step [1718/3236], Loss: 2.1753, Perplexity: 8.8046
Epoch [2/3], Step [1719/3236], Loss: 2.0483, Perplexity: 7.7546
Epoch [2/3], Step [1720/3236], Loss: 2.0953, Perplexity: 8.1275
Epoch [2/3], Step [1721/3236], Loss: 2.4104, Perplexity: 11.1386
Epoch [2/3], Step [1722/3236], Loss: 2.0277, Perplexity: 7.5967
Epoch [2/3], Step [1723/3236], Loss: 2.3381, Perplexity: 10.3616
Epoch [2/3], Step [1724/3236], Loss: 2.0552, Perplexity: 7.8082
Epoch [2/3], Step [1725/3236], Loss: 2.0385, Perplexity: 7.6789
Epoch [2/3], Step [1726/3236], Loss: 2.3903, Perplexity: 10.9168
Epoch [2/3], Step [1727/3236], Loss: 2.1619, Perplexity: 8.6878
Epoch [2/3], Step [1728/3236], Loss: 1.9922, Perplexity: 7.3316
Epoch [2/3], Step [1729/3236], Loss: 2.0931, Perplexity: 8.1101
Epoch [2/3], Step [1730/3236], Loss: 2.3438, Perplexity: 10.4207
Epoch [2/3], Step [1731/3236], Loss: 2.3641, Perplexity: 10.6347
Epoch [2/3], Step [1732/3236], Loss: 2.2034, Perplexity: 9.0560
Epoch [2/3], Step [1733/3236], Loss: 2.0434, Perplexity: 7.7168
Epoch [2/3], Step [1734/3236], Loss: 2.4099, Perplexity: 11.1330
Epoch [2/3], Step [1735/3236], Loss: 2.2527, Perplexity: 9.5136
Epoch [2/3], Step [1736/3236], Loss: 2.1874, Perplexity: 8.9117
Epoch [2/3], Step [1737/3236], Loss: 2.0719, Perplexity: 7.9399
Epoch [2/3], Step [1738/3236], Loss: 2.9426, Perplexity: 18.9647
Epoch [2/3], Step [1739/3236], Loss: 1.9980, Perplexity: 7.3742
Epoch [2/3], Step [1740/3236], Loss: 2.1723, Perplexity: 8.7787
Epoch [2/3], Step [1741/3236], Loss: 2.1863, Perplexity: 8.9021
Epoch [2/3], Step [1742/3236], Loss: 2.0383, Perplexity: 7.6774
Epoch [2/3], Step [1743/3236], Loss: 2.1734, Perplexity: 8.7880
Epoch [2/3], Step [1744/3236], Loss: 2.0841, Perplexity: 8.0377
Epoch [2/3], Step [1745/3236], Loss: 1.9749, Perplexity: 7.2059
Epoch [2/3], Step [1746/3236], Loss: 2.1009, Perplexity: 8.1735
Epoch [2/3], Step [1747/3236], Loss: 2.2061, Perplexity: 9.0801
Epoch [2/3], Step [1748/3236], Loss: 2.1293, Perplexity: 8.4092
Epoch [2/3], Step [1749/3236], Loss: 2.1105, Perplexity: 8.2523
Epoch [2/3], Step [1750/3236], Loss: 2.1327, Perplexity: 8.4372
Epoch [2/3], Step [1751/3236], Loss: 2.1162, Perplexity: 8.2999
Epoch [2/3], Step [1752/3236], Loss: 2.2034, Perplexity: 9.0557
Epoch [2/3], Step [1753/3236], Loss: 2.3706, Perplexity: 10.7034
Epoch [2/3], Step [1754/3236], Loss: 2.2836, Perplexity: 9.8118
Epoch [2/3], Step [1755/3236], Loss: 2.2201, Perplexity: 9.2082
Epoch [2/3], Step [1756/3236], Loss: 2.1533, Perplexity: 8.6131
Epoch [2/3], Step [1757/3236], Loss: 2.0165, Perplexity: 7.5117
Epoch [2/3], Step [1758/3236], Loss: 2.5256, Perplexity: 12.4982
Epoch [2/3], Step [1759/3236], Loss: 2.1858, Perplexity: 8.8979
Epoch [2/3], Step [1760/3236], Loss: 1.9239, Perplexity: 6.8476
Epoch [2/3], Step [1761/3236], Loss: 2.1189, Perplexity: 8.3222
Epoch [2/3], Step [1762/3236], Loss: 2.1645, Perplexity: 8.7101
Epoch [2/3], Step [1763/3236], Loss: 2.0932, Perplexity: 8.1107
Epoch [2/3], Step [1764/3236], Loss: 1.9686, Perplexity: 7.1604
Epoch [2/3], Step [1765/3236], Loss: 2.1798, Perplexity: 8.8447
Epoch [2/3], Step [1766/3236], Loss: 2.3427, Perplexity: 10.4094
Epoch [2/3], Step [1767/3236], Loss: 2.3299, Perplexity: 10.2773
Epoch [2/3], Step [1768/3236], Loss: 2.0960, Perplexity: 8.1338
Epoch [2/3], Step [1769/3236], Loss: 2.0998, Perplexity: 8.1642
Epoch [2/3], Step [1770/3236], Loss: 2.1553, Perplexity: 8.6305
Epoch [2/3], Step [1771/3236], Loss: 2.1952, Perplexity: 8.9815
Epoch [2/3], Step [1772/3236], Loss: 2.3846, Perplexity: 10.8552
Epoch [2/3], Step [1773/3236], Loss: 2.2287, Perplexity: 9.2882
Epoch [2/3], Step [1774/3236], Loss: 2.1054, Perplexity: 8.2103
Epoch [2/3], Step [1775/3236], Loss: 2.1461, Perplexity: 8.5515
Epoch [2/3], Step [1776/3236], Loss: 2.5242, Perplexity: 12.4806
Epoch [2/3], Step [1777/3236], Loss: 2.1653, Perplexity: 8.7168
Epoch [2/3], Step [1778/3236], Loss: 1.9273, Perplexity: 6.8712
Epoch [2/3], Step [1779/3236], Loss: 2.1987, Perplexity: 9.0133
Epoch [2/3], Step [1780/3236], Loss: 2.2480, Perplexity: 9.4692
Epoch [2/3], Step [1781/3236], Loss: 2.3018, Perplexity: 9.9921
Epoch [2/3], Step [1782/3236], Loss: 2.1305, Perplexity: 8.4188
Epoch [2/3], Step [1783/3236], Loss: 1.9906, Perplexity: 7.3199
Epoch [2/3], Step [1784/3236], Loss: 2.1243, Perplexity: 8.3672
Epoch [2/3], Step [1785/3236], Loss: 2.1231, Perplexity: 8.3566
Epoch [2/3], Step [1786/3236], Loss: 1.8961, Perplexity: 6.6595
Epoch [2/3], Step [1787/3236], Loss: 2.2185, Perplexity: 9.1938
Epoch [2/3], Step [1788/3236], Loss: 2.4012, Perplexity: 11.0364
Epoch [2/3], Step [1789/3236], Loss: 2.6715, Perplexity: 14.4621
Epoch [2/3], Step [1790/3236], Loss: 2.3603, Perplexity: 10.5943
Epoch [2/3], Step [1791/3236], Loss: 2.0115, Perplexity: 7.4742
Epoch [2/3], Step [1792/3236], Loss: 2.1940, Perplexity: 8.9707
Epoch [2/3], Step [1793/3236], Loss: 2.1658, Perplexity: 8.7219
Epoch [2/3], Step [1794/3236], Loss: 2.0456, Perplexity: 7.7338
Epoch [2/3], Step [1795/3236], Loss: 2.0721, Perplexity: 7.9414
Epoch [2/3], Step [1796/3236], Loss: 2.1025, Perplexity: 8.1863
Epoch [2/3], Step [1797/3236], Loss: 2.1260, Perplexity: 8.3810
Epoch [2/3], Step [1798/3236], Loss: 2.0124, Perplexity: 7.4812
Epoch [2/3], Step [1799/3236], Loss: 2.1986, Perplexity: 9.0121
Epoch [2/3], Step [1800/3236], Loss: 2.7443, Perplexity: 15.5534
Epoch [2/3], Step [1801/3236], Loss: 2.1328, Perplexity: 8.4383
Epoch [2/3], Step [1802/3236], Loss: 2.2680, Perplexity: 9.6601
Epoch [2/3], Step [1803/3236], Loss: 2.0821, Perplexity: 8.0211
Epoch [2/3], Step [1804/3236], Loss: 2.4042, Perplexity: 11.0694
Epoch [2/3], Step [1805/3236], Loss: 2.2157, Perplexity: 9.1677
Epoch [2/3], Step [1806/3236], Loss: 2.2337, Perplexity: 9.3340
Epoch [2/3], Step [1807/3236], Loss: 2.2358, Perplexity: 9.3541
Epoch [2/3], Step [1808/3236], Loss: 2.0282, Perplexity: 7.6007
Epoch [2/3], Step [1809/3236], Loss: 2.2514, Perplexity: 9.5013
Epoch [2/3], Step [1810/3236], Loss: 2.1788, Perplexity: 8.8356
Epoch [2/3], Step [1811/3236], Loss: 2.1674, Perplexity: 8.7359
Epoch [2/3], Step [1812/3236], Loss: 2.1851, Perplexity: 8.8911
Epoch [2/3], Step [1813/3236], Loss: 2.2083, Perplexity: 9.1001
Epoch [2/3], Step [1814/3236], Loss: 2.1498, Perplexity: 8.5830
Epoch [2/3], Step [1815/3236], Loss: 2.2154, Perplexity: 9.1648
Epoch [2/3], Step [1816/3236], Loss: 2.3153, Perplexity: 10.1278
Epoch [2/3], Step [1817/3236], Loss: 2.0116, Perplexity: 7.4755
Epoch [2/3], Step [1818/3236], Loss: 2.1866, Perplexity: 8.9049
Epoch [2/3], Step [1819/3236], Loss: 2.3048, Perplexity: 10.0220
Epoch [2/3], Step [1820/3236], Loss: 2.0666, Perplexity: 7.8977
Epoch [2/3], Step [1821/3236], Loss: 2.0018, Perplexity: 7.4025
Epoch [2/3], Step [1822/3236], Loss: 2.0053, Perplexity: 7.4286
Epoch [2/3], Step [1823/3236], Loss: 2.4364, Perplexity: 11.4322
Epoch [2/3], Step [1824/3236], Loss: 2.0532, Perplexity: 7.7930
Epoch [2/3], Step [1825/3236], Loss: 2.0430, Perplexity: 7.7136
Epoch [2/3], Step [1826/3236], Loss: 2.0074, Perplexity: 7.4439
Epoch [2/3], Step [1827/3236], Loss: 2.1186, Perplexity: 8.3192
Epoch [2/3], Step [1828/3236], Loss: 2.1740, Perplexity: 8.7930
Epoch [2/3], Step [1829/3236], Loss: 2.0135, Perplexity: 7.4892
Epoch [2/3], Step [1830/3236], Loss: 2.6948, Perplexity: 14.8028
Epoch [2/3], Step [1831/3236], Loss: 2.2230, Perplexity: 9.2349
Epoch [2/3], Step [1832/3236], Loss: 2.2304, Perplexity: 9.3040
Epoch [2/3], Step [1833/3236], Loss: 2.1331, Perplexity: 8.4412
Epoch [2/3], Step [1834/3236], Loss: 2.3521, Perplexity: 10.5071
Epoch [2/3], Step [1835/3236], Loss: 2.3625, Perplexity: 10.6173
Epoch [2/3], Step [1836/3236], Loss: 2.1143, Perplexity: 8.2837
Epoch [2/3], Step [1837/3236], Loss: 2.2756, Perplexity: 9.7336
Epoch [2/3], Step [1838/3236], Loss: 2.2849, Perplexity: 9.8247
Epoch [2/3], Step [1839/3236], Loss: 2.1218, Perplexity: 8.3458
Epoch [2/3], Step [1840/3236], Loss: 2.0757, Perplexity: 7.9698
Epoch [2/3], Step [1841/3236], Loss: 2.2095, Perplexity: 9.1110
Epoch [2/3], Step [1842/3236], Loss: 2.2252, Perplexity: 9.2556
Epoch [2/3], Step [1843/3236], Loss: 2.2756, Perplexity: 9.7335
Epoch [2/3], Step [1844/3236], Loss: 2.0399, Perplexity: 7.6898
Epoch [2/3], Step [1845/3236], Loss: 2.2939, Perplexity: 9.9135
Epoch [2/3], Step [1846/3236], Loss: 2.3996, Perplexity: 11.0190
Epoch [2/3], Step [1847/3236], Loss: 2.0474, Perplexity: 7.7479
Epoch [2/3], Step [1848/3236], Loss: 2.1081, Perplexity: 8.2328
Epoch [2/3], Step [1849/3236], Loss: 1.9996, Perplexity: 7.3859
Epoch [2/3], Step [1850/3236], Loss: 1.9413, Perplexity: 6.9680
Epoch [2/3], Step [1851/3236], Loss: 2.0797, Perplexity: 8.0022
Epoch [2/3], Step [1852/3236], Loss: 2.4829, Perplexity: 11.9757
Epoch [2/3], Step [1853/3236], Loss: 2.9348, Perplexity: 18.8168
Epoch [2/3], Step [1854/3236], Loss: 2.1469, Perplexity: 8.5583
Epoch [2/3], Step [1855/3236], Loss: 2.4410, Perplexity: 11.4848
Epoch [2/3], Step [1856/3236], Loss: 2.1916, Perplexity: 8.9493
Epoch [2/3], Step [1857/3236], Loss: 2.1493, Perplexity: 8.5793
Epoch [2/3], Step [1858/3236], Loss: 2.0604, Perplexity: 7.8492
Epoch [2/3], Step [1859/3236], Loss: 2.1525, Perplexity: 8.6060
Epoch [2/3], Step [1860/3236], Loss: 2.2139, Perplexity: 9.1512
Epoch [2/3], Step [1861/3236], Loss: 2.5468, Perplexity: 12.7668
Epoch [2/3], Step [1862/3236], Loss: 2.1725, Perplexity: 8.7803
Epoch [2/3], Step [1863/3236], Loss: 2.1646, Perplexity: 8.7112
Epoch [2/3], Step [1864/3236], Loss: 2.0525, Perplexity: 7.7876
Epoch [2/3], Step [1865/3236], Loss: 2.2576, Perplexity: 9.5603
Epoch [2/3], Step [1866/3236], Loss: 2.1641, Perplexity: 8.7067
Epoch [2/3], Step [1867/3236], Loss: 2.1463, Perplexity: 8.5534
Epoch [2/3], Step [1868/3236], Loss: 1.9566, Perplexity: 7.0750
Epoch [2/3], Step [1869/3236], Loss: 2.1713, Perplexity: 8.7693
Epoch [2/3], Step [1870/3236], Loss: 2.2096, Perplexity: 9.1119
Epoch [2/3], Step [1871/3236], Loss: 2.1533, Perplexity: 8.6135
Epoch [2/3], Step [1872/3236], Loss: 3.0109, Perplexity: 20.3066
Epoch [2/3], Step [1873/3236], Loss: 2.1331, Perplexity: 8.4414
Epoch [2/3], Step [1874/3236], Loss: 2.1729, Perplexity: 8.7835
Epoch [2/3], Step [1875/3236], Loss: 1.9938, Perplexity: 7.3437
Epoch [2/3], Step [1876/3236], Loss: 2.1198, Perplexity: 8.3297
Epoch [2/3], Step [1877/3236], Loss: 2.2377, Perplexity: 9.3722
Epoch [2/3], Step [1878/3236], Loss: 2.0796, Perplexity: 8.0013
Epoch [2/3], Step [1879/3236], Loss: 2.2301, Perplexity: 9.3008
Epoch [2/3], Step [1880/3236], Loss: 2.4837, Perplexity: 11.9861
Epoch [2/3], Step [1881/3236], Loss: 2.1173, Perplexity: 8.3088
Epoch [2/3], Step [1882/3236], Loss: 1.9395, Perplexity: 6.9551
Epoch [2/3], Step [1883/3236], Loss: 2.1854, Perplexity: 8.8943
Epoch [2/3], Step [1884/3236], Loss: 2.1228, Perplexity: 8.3541
Epoch [2/3], Step [1885/3236], Loss: 2.2938, Perplexity: 9.9125
Epoch [2/3], Step [1886/3236], Loss: 2.1883, Perplexity: 8.9203
Epoch [2/3], Step [1887/3236], Loss: 2.2032, Perplexity: 9.0537
Epoch [2/3], Step [1888/3236], Loss: 2.2634, Perplexity: 9.6155
Epoch [2/3], Step [1889/3236], Loss: 2.1090, Perplexity: 8.2397
Epoch [2/3], Step [1890/3236], Loss: 2.2239, Perplexity: 9.2431
Epoch [2/3], Step [1891/3236], Loss: 2.0146, Perplexity: 7.4979
Epoch [2/3], Step [1892/3236], Loss: 2.0873, Perplexity: 8.0627
Epoch [2/3], Step [1893/3236], Loss: 2.1175, Perplexity: 8.3106
Epoch [2/3], Step [1894/3236], Loss: 2.2459, Perplexity: 9.4493
Epoch [2/3], Step [1895/3236], Loss: 2.5565, Perplexity: 12.8901
Epoch [2/3], Step [1896/3236], Loss: 2.1946, Perplexity: 8.9768
Epoch [2/3], Step [1897/3236], Loss: 2.0267, Perplexity: 7.5892
Epoch [2/3], Step [1898/3236], Loss: 2.0347, Perplexity: 7.6496
Epoch [2/3], Step [1899/3236], Loss: 1.9533, Perplexity: 7.0518
Epoch [2/3], Step [1900/3236], Loss: 2.1761, Perplexity: 8.8121
Epoch [2/3], Step [1901/3236], Loss: 2.0417, Perplexity: 7.7034
Epoch [2/3], Step [1902/3236], Loss: 2.0826, Perplexity: 8.0257
Epoch [2/3], Step [1903/3236], Loss: 2.0847, Perplexity: 8.0424
Epoch [2/3], Step [1904/3236], Loss: 2.1076, Perplexity: 8.2284
Epoch [2/3], Step [1905/3236], Loss: 2.0832, Perplexity: 8.0304
Epoch [2/3], Step [1906/3236], Loss: 2.0887, Perplexity: 8.0742
Epoch [2/3], Step [1907/3236], Loss: 2.1204, Perplexity: 8.3347
Epoch [2/3], Step [1908/3236], Loss: 2.1464, Perplexity: 8.5544
Epoch [2/3], Step [1909/3236], Loss: 2.4212, Perplexity: 11.2598
Epoch [2/3], Step [1910/3236], Loss: 2.0827, Perplexity: 8.0261
Epoch [2/3], Step [1911/3236], Loss: 2.2843, Perplexity: 9.8186
Epoch [2/3], Step [1912/3236], Loss: 2.1493, Perplexity: 8.5785
Epoch [2/3], Step [1913/3236], Loss: 2.1437, Perplexity: 8.5308
Epoch [2/3], Step [1914/3236], Loss: 2.2043, Perplexity: 9.0638
Epoch [2/3], Step [1915/3236], Loss: 2.2399, Perplexity: 9.3925
Epoch [2/3], Step [1916/3236], Loss: 2.0630, Perplexity: 7.8699
Epoch [2/3], Step [1917/3236], Loss: 2.4621, Perplexity: 11.7299
Epoch [2/3], Step [1918/3236], Loss: 2.1002, Perplexity: 8.1678
Epoch [2/3], Step [1919/3236], Loss: 2.0340, Perplexity: 7.6445
Epoch [2/3], Step [1920/3236], Loss: 2.1751, Perplexity: 8.8031
Epoch [2/3], Step [1921/3236], Loss: 2.0762, Perplexity: 7.9740
Epoch [2/3], Step [1922/3236], Loss: 2.0790, Perplexity: 7.9961
Epoch [2/3], Step [1923/3236], Loss: 2.0379, Perplexity: 7.6746
Epoch [2/3], Step [1924/3236], Loss: 2.1925, Perplexity: 8.9573
Epoch [2/3], Step [1925/3236], Loss: 2.0994, Perplexity: 8.1615
Epoch [2/3], Step [1926/3236], Loss: 2.2087, Perplexity: 9.1036
Epoch [2/3], Step [1927/3236], Loss: 2.0180, Perplexity: 7.5232
Epoch [2/3], Step [1928/3236], Loss: 2.0867, Perplexity: 8.0584
Epoch [2/3], Step [1929/3236], Loss: 2.0033, Perplexity: 7.4136
Epoch [2/3], Step [1930/3236], Loss: 2.2334, Perplexity: 9.3313
Epoch [2/3], Step [1931/3236], Loss: 2.2467, Perplexity: 9.4565
Epoch [2/3], Step [1932/3236], Loss: 2.4153, Perplexity: 11.1935
Epoch [2/3], Step [1933/3236], Loss: 2.0853, Perplexity: 8.0470
Epoch [2/3], Step [1934/3236], Loss: 2.1782, Perplexity: 8.8306
Epoch [2/3], Step [1935/3236], Loss: 2.0318, Perplexity: 7.6276
Epoch [2/3], Step [1936/3236], Loss: 2.1231, Perplexity: 8.3568
Epoch [2/3], Step [1937/3236], Loss: 1.9604, Perplexity: 7.1020
Epoch [2/3], Step [1938/3236], Loss: 2.2131, Perplexity: 9.1436
Epoch [2/3], Step [1939/3236], Loss: 2.2082, Perplexity: 9.0992
Epoch [2/3], Step [1940/3236], Loss: 1.9875, Perplexity: 7.2975
Epoch [2/3], Step [1941/3236], Loss: 2.1111, Perplexity: 8.2576
Epoch [2/3], Step [1942/3236], Loss: 2.0815, Perplexity: 8.0167
Epoch [2/3], Step [1943/3236], Loss: 2.0798, Perplexity: 8.0029
Epoch [2/3], Step [1944/3236], Loss: 2.1360, Perplexity: 8.4653
Epoch [2/3], Step [1945/3236], Loss: 2.1170, Perplexity: 8.3061
Epoch [2/3], Step [1946/3236], Loss: 2.0968, Perplexity: 8.1398
Epoch [2/3], Step [1947/3236], Loss: 1.9534, Perplexity: 7.0526
Epoch [2/3], Step [1948/3236], Loss: 2.2160, Perplexity: 9.1703
Epoch [2/3], Step [1949/3236], Loss: 2.5696, Perplexity: 13.0606
Epoch [2/3], Step [1950/3236], Loss: 2.0671, Perplexity: 7.9016
Epoch [2/3], Step [1951/3236], Loss: 1.9272, Perplexity: 6.8705
Epoch [2/3], Step [1952/3236], Loss: 2.0906, Perplexity: 8.0894
Epoch [2/3], Step [1953/3236], Loss: 2.0852, Perplexity: 8.0462
Epoch [2/3], Step [1954/3236], Loss: 2.8280, Perplexity: 16.9120
Epoch [2/3], Step [1955/3236], Loss: 2.1293, Perplexity: 8.4092
Epoch [2/3], Step [1956/3236], Loss: 2.0684, Perplexity: 7.9121
Epoch [2/3], Step [1957/3236], Loss: 2.7127, Perplexity: 15.0694
Epoch [2/3], Step [1958/3236], Loss: 2.0074, Perplexity: 7.4438
Epoch [2/3], Step [1959/3236], Loss: 2.2505, Perplexity: 9.4925
Epoch [2/3], Step [1960/3236], Loss: 2.0635, Perplexity: 7.8735
Epoch [2/3], Step [1961/3236], Loss: 2.0276, Perplexity: 7.5959
Epoch [2/3], Step [1962/3236], Loss: 2.1351, Perplexity: 8.4582
Epoch [2/3], Step [1963/3236], Loss: 2.0840, Perplexity: 8.0368
Epoch [2/3], Step [1964/3236], Loss: 1.9577, Perplexity: 7.0831
Epoch [2/3], Step [1965/3236], Loss: 2.2005, Perplexity: 9.0298
Epoch [2/3], Step [1966/3236], Loss: 2.1449, Perplexity: 8.5414
Epoch [2/3], Step [1967/3236], Loss: 2.4635, Perplexity: 11.7458
Epoch [2/3], Step [1968/3236], Loss: 2.0824, Perplexity: 8.0236
Epoch [2/3], Step [1969/3236], Loss: 2.0441, Perplexity: 7.7222
Epoch [2/3], Step [1970/3236], Loss: 2.1316, Perplexity: 8.4281
Epoch [2/3], Step [1971/3236], Loss: 2.2651, Perplexity: 9.6321
Epoch [2/3], Step [1972/3236], Loss: 2.0742, Perplexity: 7.9583
Epoch [2/3], Step [1973/3236], Loss: 1.9044, Perplexity: 6.7155
Epoch [2/3], Step [1974/3236], Loss: 2.0634, Perplexity: 7.8723
Epoch [2/3], Step [1975/3236], Loss: 2.1349, Perplexity: 8.4559
Epoch [2/3], Step [1976/3236], Loss: 2.1672, Perplexity: 8.7334
Epoch [2/3], Step [1977/3236], Loss: 2.0118, Perplexity: 7.4771
Epoch [2/3], Step [1978/3236], Loss: 2.1598, Perplexity: 8.6691
Epoch [2/3], Step [1979/3236], Loss: 2.0994, Perplexity: 8.1615
Epoch [2/3], Step [1980/3236], Loss: 2.4293, Perplexity: 11.3513
Epoch [2/3], Step [1981/3236], Loss: 2.0448, Perplexity: 7.7277
Epoch [2/3], Step [1982/3236], Loss: 2.0064, Perplexity: 7.4365
Epoch [2/3], Step [1983/3236], Loss: 2.0941, Perplexity: 8.1179
Epoch [2/3], Step [1984/3236], Loss: 2.0081, Perplexity: 7.4488
Epoch [2/3], Step [1985/3236], Loss: 2.0022, Perplexity: 7.4054
Epoch [2/3], Step [1986/3236], Loss: 2.0067, Perplexity: 7.4388
Epoch [2/3], Step [1987/3236], Loss: 2.1914, Perplexity: 8.9478
Epoch [2/3], Step [1988/3236], Loss: 2.0762, Perplexity: 7.9739
Epoch [2/3], Step [1989/3236], Loss: 2.1245, Perplexity: 8.3684
Epoch [2/3], Step [1990/3236], Loss: 2.0301, Perplexity: 7.6146
Epoch [2/3], Step [1991/3236], Loss: 2.0466, Perplexity: 7.7412
Epoch [2/3], Step [1992/3236], Loss: 2.1496, Perplexity: 8.5818
Epoch [2/3], Step [1993/3236], Loss: 2.1767, Perplexity: 8.8168
Epoch [2/3], Step [1994/3236], Loss: 2.1801, Perplexity: 8.8471
Epoch [2/3], Step [1995/3236], Loss: 2.1016, Perplexity: 8.1791
Epoch [2/3], Step [1996/3236], Loss: 2.2175, Perplexity: 9.1840
Epoch [2/3], Step [1997/3236], Loss: 2.6106, Perplexity: 13.6072
Epoch [2/3], Step [1998/3236], Loss: 2.1747, Perplexity: 8.7992
Epoch [2/3], Step [1999/3236], Loss: 2.1310, Perplexity: 8.4232
Epoch [2/3], Step [2000/3236], Loss: 2.0298, Perplexity: 7.6129
Epoch [2/3], Step [2001/3236], Loss: 2.0402, Perplexity: 7.6918
Epoch [2/3], Step [2002/3236], Loss: 2.2936, Perplexity: 9.9106
Epoch [2/3], Step [2003/3236], Loss: 2.1606, Perplexity: 8.6762
Epoch [2/3], Step [2004/3236], Loss: 2.0817, Perplexity: 8.0182
Epoch [2/3], Step [2005/3236], Loss: 2.1429, Perplexity: 8.5237
Epoch [2/3], Step [2006/3236], Loss: 2.0483, Perplexity: 7.7549
Epoch [2/3], Step [2007/3236], Loss: 2.1722, Perplexity: 8.7773
Epoch [2/3], Step [2008/3236], Loss: 2.1143, Perplexity: 8.2837
Epoch [2/3], Step [2009/3236], Loss: 2.2067, Perplexity: 9.0853
Epoch [2/3], Step [2010/3236], Loss: 2.1889, Perplexity: 8.9252
Epoch [2/3], Step [2011/3236], Loss: 2.1235, Perplexity: 8.3607
Epoch [2/3], Step [2012/3236], Loss: 2.1154, Perplexity: 8.2929
Epoch [2/3], Step [2013/3236], Loss: 2.0950, Perplexity: 8.1256
Epoch [2/3], Step [2014/3236], Loss: 1.9598, Perplexity: 7.0979
Epoch [2/3], Step [2015/3236], Loss: 2.0291, Perplexity: 7.6072
Epoch [2/3], Step [2016/3236], Loss: 2.2458, Perplexity: 9.4478
Epoch [2/3], Step [2017/3236], Loss: 1.9739, Perplexity: 7.1984
Epoch [2/3], Step [2018/3236], Loss: 2.0721, Perplexity: 7.9417
Epoch [2/3], Step [2019/3236], Loss: 2.9224, Perplexity: 18.5864
Epoch [2/3], Step [2020/3236], Loss: 2.0784, Perplexity: 7.9913
Epoch [2/3], Step [2021/3236], Loss: 2.1348, Perplexity: 8.4556
Epoch [2/3], Step [2022/3236], Loss: 2.0725, Perplexity: 7.9447
Epoch [2/3], Step [2023/3236], Loss: 2.3603, Perplexity: 10.5943
Epoch [2/3], Step [2024/3236], Loss: 2.1128, Perplexity: 8.2713
Epoch [2/3], Step [2025/3236], Loss: 2.8411, Perplexity: 17.1354
Epoch [2/3], Step [2026/3236], Loss: 2.0649, Perplexity: 7.8843
Epoch [2/3], Step [2027/3236], Loss: 2.0191, Perplexity: 7.5317
Epoch [2/3], Step [2028/3236], Loss: 2.1079, Perplexity: 8.2309
Epoch [2/3], Step [2029/3236], Loss: 2.3234, Perplexity: 10.2101
Epoch [2/3], Step [2030/3236], Loss: 2.6377, Perplexity: 13.9805
Epoch [2/3], Step [2031/3236], Loss: 2.3777, Perplexity: 10.7798
Epoch [2/3], Step [2032/3236], Loss: 2.0110, Perplexity: 7.4711
Epoch [2/3], Step [2033/3236], Loss: 2.2022, Perplexity: 9.0453
Epoch [2/3], Step [2034/3236], Loss: 2.1107, Perplexity: 8.2544
Epoch [2/3], Step [2035/3236], Loss: 2.3964, Perplexity: 10.9839
Epoch [2/3], Step [2036/3236], Loss: 2.1048, Perplexity: 8.2058
Epoch [2/3], Step [2037/3236], Loss: 2.5139, Perplexity: 12.3531
Epoch [2/3], Step [2038/3236], Loss: 2.4779, Perplexity: 11.9158
Epoch [2/3], Step [2039/3236], Loss: 2.1772, Perplexity: 8.8218
Epoch [2/3], Step [2040/3236], Loss: 2.2158, Perplexity: 9.1688
Epoch [2/3], Step [2041/3236], Loss: 2.2654, Perplexity: 9.6349
Epoch [2/3], Step [2042/3236], Loss: 2.2360, Perplexity: 9.3556
Epoch [2/3], Step [2043/3236], Loss: 2.1082, Perplexity: 8.2337
Epoch [2/3], Step [2044/3236], Loss: 1.8997, Perplexity: 6.6842
Epoch [2/3], Step [2045/3236], Loss: 2.2661, Perplexity: 9.6418
Epoch [2/3], Step [2046/3236], Loss: 2.1440, Perplexity: 8.5337
Epoch [2/3], Step [2047/3236], Loss: 2.1223, Perplexity: 8.3507
Epoch [2/3], Step [2048/3236], Loss: 2.0674, Perplexity: 7.9043
Epoch [2/3], Step [2049/3236], Loss: 2.0952, Perplexity: 8.1270
Epoch [2/3], Step [2050/3236], Loss: 1.9530, Perplexity: 7.0501
Epoch [2/3], Step [2051/3236], Loss: 2.0689, Perplexity: 7.9164
Epoch [2/3], Step [2052/3236], Loss: 2.8627, Perplexity: 17.5082
Epoch [2/3], Step [2053/3236], Loss: 2.0139, Perplexity: 7.4928
Epoch [2/3], Step [2054/3236], Loss: 2.3655, Perplexity: 10.6495
Epoch [2/3], Step [2055/3236], Loss: 2.0103, Perplexity: 7.4655
Epoch [2/3], Step [2056/3236], Loss: 2.5663, Perplexity: 13.0175
Epoch [2/3], Step [2057/3236], Loss: 2.2905, Perplexity: 9.8804
Epoch [2/3], Step [2058/3236], Loss: 2.1643, Perplexity: 8.7084
Epoch [2/3], Step [2059/3236], Loss: 2.3746, Perplexity: 10.7469
Epoch [2/3], Step [2060/3236], Loss: 2.1289, Perplexity: 8.4053
Epoch [2/3], Step [2061/3236], Loss: 2.0657, Perplexity: 7.8909
Epoch [2/3], Step [2062/3236], Loss: 2.1176, Perplexity: 8.3116
Epoch [2/3], Step [2063/3236], Loss: 2.1166, Perplexity: 8.3031
Epoch [2/3], Step [2064/3236], Loss: 2.2444, Perplexity: 9.4346
Epoch [2/3], Step [2065/3236], Loss: 2.0075, Perplexity: 7.4447
Epoch [2/3], Step [2066/3236], Loss: 1.9710, Perplexity: 7.1781
Epoch [2/3], Step [2067/3236], Loss: 2.1615, Perplexity: 8.6843
Epoch [2/3], Step [2068/3236], Loss: 2.1644, Perplexity: 8.7098
Epoch [2/3], Step [2069/3236], Loss: 2.0411, Perplexity: 7.6992
Epoch [2/3], Step [2070/3236], Loss: 2.2349, Perplexity: 9.3451
Epoch [2/3], Step [2071/3236], Loss: 2.1496, Perplexity: 8.5817
Epoch [2/3], Step [2072/3236], Loss: 2.0075, Perplexity: 7.4446
Epoch [2/3], Step [2073/3236], Loss: 2.1659, Perplexity: 8.7222
Epoch [2/3], Step [2074/3236], Loss: 2.2061, Perplexity: 9.0798
Epoch [2/3], Step [2075/3236], Loss: 1.9910, Perplexity: 7.3228
Epoch [2/3], Step [2076/3236], Loss: 2.2491, Perplexity: 9.4788
Epoch [2/3], Step [2077/3236], Loss: 2.2139, Perplexity: 9.1517
Epoch [2/3], Step [2078/3236], Loss: 2.1668, Perplexity: 8.7301
Epoch [2/3], Step [2079/3236], Loss: 1.9343, Perplexity: 6.9193
Epoch [2/3], Step [2080/3236], Loss: 2.2030, Perplexity: 9.0521
Epoch [2/3], Step [2081/3236], Loss: 2.4506, Perplexity: 11.5953
Epoch [2/3], Step [2082/3236], Loss: 2.1092, Perplexity: 8.2417
Epoch [2/3], Step [2083/3236], Loss: 2.1211, Perplexity: 8.3400
Epoch [2/3], Step [2084/3236], Loss: 2.6223, Perplexity: 13.7667
Epoch [2/3], Step [2085/3236], Loss: 2.2721, Perplexity: 9.6995
Epoch [2/3], Step [2086/3236], Loss: 2.1754, Perplexity: 8.8057
Epoch [2/3], Step [2087/3236], Loss: 2.0926, Perplexity: 8.1058
Epoch [2/3], Step [2088/3236], Loss: 2.0668, Perplexity: 7.8996
Epoch [2/3], Step [2089/3236], Loss: 2.1551, Perplexity: 8.6283
Epoch [2/3], Step [2090/3236], Loss: 2.0619, Perplexity: 7.8612
Epoch [2/3], Step [2091/3236], Loss: 2.0573, Perplexity: 7.8246
Epoch [2/3], Step [2092/3236], Loss: 2.0969, Perplexity: 8.1412
Epoch [2/3], Step [2093/3236], Loss: 2.6638, Perplexity: 14.3501
Epoch [2/3], Step [2094/3236], Loss: 2.0804, Perplexity: 8.0074
Epoch [2/3], Step [2095/3236], Loss: 2.2094, Perplexity: 9.1106
Epoch [2/3], Step [2096/3236], Loss: 2.1651, Perplexity: 8.7152
Epoch [2/3], Step [2097/3236], Loss: 2.1247, Perplexity: 8.3704
Epoch [2/3], Step [2098/3236], Loss: 2.0829, Perplexity: 8.0281
Epoch [2/3], Step [2099/3236], Loss: 2.1524, Perplexity: 8.6053
Epoch [2/3], Step [2100/3236], Loss: 2.3892, Perplexity: 10.9050
Epoch [2/3], Step [2101/3236], Loss: 1.9850, Perplexity: 7.2787
Epoch [2/3], Step [2102/3236], Loss: 2.1517, Perplexity: 8.5996
Epoch [2/3], Step [2103/3236], Loss: 2.1422, Perplexity: 8.5182
Epoch [2/3], Step [2104/3236], Loss: 2.0181, Perplexity: 7.5237
Epoch [2/3], Step [2105/3236], Loss: 2.2436, Perplexity: 9.4276
Epoch [2/3], Step [2106/3236], Loss: 2.0723, Perplexity: 7.9430
Epoch [2/3], Step [2107/3236], Loss: 2.5915, Perplexity: 13.3503
Epoch [2/3], Step [2108/3236], Loss: 2.9203, Perplexity: 18.5464
Epoch [2/3], Step [2109/3236], Loss: 2.0953, Perplexity: 8.1275
Epoch [2/3], Step [2110/3236], Loss: 2.7284, Perplexity: 15.3081
Epoch [2/3], Step [2111/3236], Loss: 2.0334, Perplexity: 7.6398
Epoch [2/3], Step [2112/3236], Loss: 2.2249, Perplexity: 9.2530
Epoch [2/3], Step [2113/3236], Loss: 2.1469, Perplexity: 8.5584
Epoch [2/3], Step [2114/3236], Loss: 2.1432, Perplexity: 8.5263
Epoch [2/3], Step [2115/3236], Loss: 1.9783, Perplexity: 7.2302
Epoch [2/3], Step [2116/3236], Loss: 1.9550, Perplexity: 7.0638
Epoch [2/3], Step [2117/3236], Loss: 2.2395, Perplexity: 9.3888
Epoch [2/3], Step [2118/3236], Loss: 2.0297, Perplexity: 7.6119
Epoch [2/3], Step [2119/3236], Loss: 1.9847, Perplexity: 7.2765
Epoch [2/3], Step [2120/3236], Loss: 2.0503, Perplexity: 7.7706
Epoch [2/3], Step [2121/3236], Loss: 2.2486, Perplexity: 9.4740
Epoch [2/3], Step [2122/3236], Loss: 2.0721, Perplexity: 7.9415
Epoch [2/3], Step [2123/3236], Loss: 2.2154, Perplexity: 9.1646
Epoch [2/3], Step [2124/3236], Loss: 2.0841, Perplexity: 8.0371
Epoch [2/3], Step [2125/3236], Loss: 2.3740, Perplexity: 10.7402
Epoch [2/3], Step [2126/3236], Loss: 2.1877, Perplexity: 8.9143
Epoch [2/3], Step [2127/3236], Loss: 2.1165, Perplexity: 8.3023
Epoch [2/3], Step [2128/3236], Loss: 2.0231, Perplexity: 7.5619
Epoch [2/3], Step [2129/3236], Loss: 1.9872, Perplexity: 7.2951
Epoch [2/3], Step [2130/3236], Loss: 2.3370, Perplexity: 10.3506
Epoch [2/3], Step [2131/3236], Loss: 1.8925, Perplexity: 6.6360
Epoch [2/3], Step [2132/3236], Loss: 2.2420, Perplexity: 9.4122
Epoch [2/3], Step [2133/3236], Loss: 2.4130, Perplexity: 11.1678
Epoch [2/3], Step [2134/3236], Loss: 2.0420, Perplexity: 7.7062
Epoch [2/3], Step [2135/3236], Loss: 2.3046, Perplexity: 10.0204
Epoch [2/3], Step [2136/3236], Loss: 2.1175, Perplexity: 8.3106
Epoch [2/3], Step [2137/3236], Loss: 1.9974, Perplexity: 7.3697
Epoch [2/3], Step [2138/3236], Loss: 2.1846, Perplexity: 8.8872
Epoch [2/3], Step [2139/3236], Loss: 2.0138, Perplexity: 7.4917
Epoch [2/3], Step [2140/3236], Loss: 2.0837, Perplexity: 8.0338
Epoch [2/3], Step [2141/3236], Loss: 2.4910, Perplexity: 12.0738
Epoch [2/3], Step [2142/3236], Loss: 2.1729, Perplexity: 8.7838
Epoch [2/3], Step [2143/3236], Loss: 2.4585, Perplexity: 11.6870
Epoch [2/3], Step [2144/3236], Loss: 2.1077, Perplexity: 8.2297
Epoch [2/3], Step [2145/3236], Loss: 2.1035, Perplexity: 8.1947
Epoch [2/3], Step [2146/3236], Loss: 1.9361, Perplexity: 6.9314
Epoch [2/3], Step [2147/3236], Loss: 2.3549, Perplexity: 10.5367
Epoch [2/3], Step [2148/3236], Loss: 2.4218, Perplexity: 11.2660
Epoch [2/3], Step [2149/3236], Loss: 1.9674, Perplexity: 7.1520
Epoch [2/3], Step [2150/3236], Loss: 2.1225, Perplexity: 8.3520
Epoch [2/3], Step [2151/3236], Loss: 2.0670, Perplexity: 7.9009
Epoch [2/3], Step [2152/3236], Loss: 2.1973, Perplexity: 9.0005
Epoch [2/3], Step [2153/3236], Loss: 2.1624, Perplexity: 8.6924
Epoch [2/3], Step [2154/3236], Loss: 2.1232, Perplexity: 8.3582
Epoch [2/3], Step [2155/3236], Loss: 2.3907, Perplexity: 10.9215
Epoch [2/3], Step [2156/3236], Loss: 1.9122, Perplexity: 6.7676
Epoch [2/3], Step [2157/3236], Loss: 2.1335, Perplexity: 8.4440
Epoch [2/3], Step [2158/3236], Loss: 2.6555, Perplexity: 14.2320
Epoch [2/3], Step [2159/3236], Loss: 2.1354, Perplexity: 8.4603
Epoch [2/3], Step [2160/3236], Loss: 2.1875, Perplexity: 8.9128
Epoch [2/3], Step [2161/3236], Loss: 2.1997, Perplexity: 9.0226
Epoch [2/3], Step [2162/3236], Loss: 2.1422, Perplexity: 8.5179
Epoch [2/3], Step [2163/3236], Loss: 2.1394, Perplexity: 8.4940
Epoch [2/3], Step [2164/3236], Loss: 2.0428, Perplexity: 7.7121
Epoch [2/3], Step [2165/3236], Loss: 2.0492, Perplexity: 7.7618
Epoch [2/3], Step [2166/3236], Loss: 2.3443, Perplexity: 10.4263
Epoch [2/3], Step [2167/3236], Loss: 2.1197, Perplexity: 8.3286
Epoch [2/3], Step [2168/3236], Loss: 2.0756, Perplexity: 7.9695
Epoch [2/3], Step [2169/3236], Loss: 3.0647, Perplexity: 21.4274
Epoch [2/3], Step [2170/3236], Loss: 2.0668, Perplexity: 7.8993
Epoch [2/3], Step [2171/3236], Loss: 2.5229, Perplexity: 12.4651
Epoch [2/3], Step [2172/3236], Loss: 2.0575, Perplexity: 7.8264
Epoch [2/3], Step [2173/3236], Loss: 2.1041, Perplexity: 8.1996
Epoch [2/3], Step [2174/3236], Loss: 2.1020, Perplexity: 8.1827
Epoch [2/3], Step [2175/3236], Loss: 2.1208, Perplexity: 8.3375
Epoch [2/3], Step [2176/3236], Loss: 2.2155, Perplexity: 9.1662
Epoch [2/3], Step [2177/3236], Loss: 2.1486, Perplexity: 8.5732
Epoch [2/3], Step [2178/3236], Loss: 2.0337, Perplexity: 7.6424
Epoch [2/3], Step [2179/3236], Loss: 2.0711, Perplexity: 7.9333
Epoch [2/3], Step [2180/3236], Loss: 2.1064, Perplexity: 8.2189
Epoch [2/3], Step [2181/3236], Loss: 1.9357, Perplexity: 6.9286
Epoch [2/3], Step [2182/3236], Loss: 2.0897, Perplexity: 8.0822
Epoch [2/3], Step [2183/3236], Loss: 2.2792, Perplexity: 9.7684
Epoch [2/3], Step [2184/3236], Loss: 2.3356, Perplexity: 10.3354
Epoch [2/3], Step [2185/3236], Loss: 2.3053, Perplexity: 10.0269
Epoch [2/3], Step [2186/3236], Loss: 2.1234, Perplexity: 8.3598
Epoch [2/3], Step [2187/3236], Loss: 2.2455, Perplexity: 9.4452
Epoch [2/3], Step [2188/3236], Loss: 2.3319, Perplexity: 10.2976
Epoch [2/3], Step [2189/3236], Loss: 2.1810, Perplexity: 8.8551
Epoch [2/3], Step [2190/3236], Loss: 2.0232, Perplexity: 7.5623
Epoch [2/3], Step [2191/3236], Loss: 2.2397, Perplexity: 9.3909
Epoch [2/3], Step [2192/3236], Loss: 2.3509, Perplexity: 10.4951
Epoch [2/3], Step [2193/3236], Loss: 2.4075, Perplexity: 11.1066
Epoch [2/3], Step [2194/3236], Loss: 2.2471, Perplexity: 9.4604
Epoch [2/3], Step [2195/3236], Loss: 2.2511, Perplexity: 9.4977
Epoch [2/3], Step [2196/3236], Loss: 2.0488, Perplexity: 7.7586
Epoch [2/3], Step [2197/3236], Loss: 1.9148, Perplexity: 6.7858
Epoch [2/3], Step [2198/3236], Loss: 2.0218, Perplexity: 7.5517
Epoch [2/3], Step [2199/3236], Loss: 2.1292, Perplexity: 8.4083
Epoch [2/3], Step [2200/3236], Loss: 2.2292, Perplexity: 9.2924
Epoch [2/3], Step [2201/3236], Loss: 2.1975, Perplexity: 9.0029
Epoch [2/3], Step [2202/3236], Loss: 1.9885, Perplexity: 7.3044
Epoch [2/3], Step [2203/3236], Loss: 2.2460, Perplexity: 9.4503
Epoch [2/3], Step [2204/3236], Loss: 2.1508, Perplexity: 8.5915
Epoch [2/3], Step [2205/3236], Loss: 2.3475, Perplexity: 10.4598
Epoch [2/3], Step [2206/3236], Loss: 2.3158, Perplexity: 10.1334
Epoch [2/3], Step [2207/3236], Loss: 2.1768, Perplexity: 8.8183
Epoch [2/3], Step [2208/3236], Loss: 1.8616, Perplexity: 6.4337
Epoch [2/3], Step [2209/3236], Loss: 2.0190, Perplexity: 7.5305
Epoch [2/3], Step [2210/3236], Loss: 2.4003, Perplexity: 11.0266
Epoch [2/3], Step [2211/3236], Loss: 1.9728, Perplexity: 7.1906
Epoch [2/3], Step [2212/3236], Loss: 1.9988, Perplexity: 7.3804
Epoch [2/3], Step [2213/3236], Loss: 2.1036, Perplexity: 8.1956
Epoch [2/3], Step [2214/3236], Loss: 2.2614, Perplexity: 9.5965
Epoch [2/3], Step [2215/3236], Loss: 2.4861, Perplexity: 12.0146
Epoch [2/3], Step [2216/3236], Loss: 2.0019, Perplexity: 7.4030
Epoch [2/3], Step [2217/3236], Loss: 2.3100, Perplexity: 10.0748
Epoch [2/3], Step [2218/3236], Loss: 2.0031, Perplexity: 7.4122
Epoch [2/3], Step [2219/3236], Loss: 2.0361, Perplexity: 7.6609
Epoch [2/3], Step [2220/3236], Loss: 2.0745, Perplexity: 7.9608
Epoch [2/3], Step [2221/3236], Loss: 2.0425, Perplexity: 7.7099
Epoch [2/3], Step [2222/3236], Loss: 2.1490, Perplexity: 8.5765
Epoch [2/3], Step [2223/3236], Loss: 2.0445, Perplexity: 7.7256
Epoch [2/3], Step [2224/3236], Loss: 1.9436, Perplexity: 6.9837
Epoch [2/3], Step [2225/3236], Loss: 2.0179, Perplexity: 7.5225
Epoch [2/3], Step [2226/3236], Loss: 2.4466, Perplexity: 11.5488
Epoch [2/3], Step [2227/3236], Loss: 2.1476, Perplexity: 8.5646
Epoch [2/3], Step [2228/3236], Loss: 2.3747, Perplexity: 10.7481
Epoch [2/3], Step [2229/3236], Loss: 2.2195, Perplexity: 9.2029
Epoch [2/3], Step [2230/3236], Loss: 2.3684, Perplexity: 10.6806
Epoch [2/3], Step [2231/3236], Loss: 2.1641, Perplexity: 8.7066
Epoch [2/3], Step [2232/3236], Loss: 2.0990, Perplexity: 8.1582
Epoch [2/3], Step [2233/3236], Loss: 2.1000, Perplexity: 8.1659
Epoch [2/3], Step [2234/3236], Loss: 2.0431, Perplexity: 7.7144
Epoch [2/3], Step [2235/3236], Loss: 2.1397, Perplexity: 8.4965
Epoch [2/3], Step [2236/3236], Loss: 2.2580, Perplexity: 9.5638
Epoch [2/3], Step [2237/3236], Loss: 2.1621, Perplexity: 8.6895
Epoch [2/3], Step [2238/3236], Loss: 2.3645, Perplexity: 10.6387
Epoch [2/3], Step [2239/3236], Loss: 2.2550, Perplexity: 9.5353
Epoch [2/3], Step [2240/3236], Loss: 1.9051, Perplexity: 6.7199
Epoch [2/3], Step [2241/3236], Loss: 2.0562, Perplexity: 7.8159
Epoch [2/3], Step [2242/3236], Loss: 2.0456, Perplexity: 7.7338
Epoch [2/3], Step [2243/3236], Loss: 1.9989, Perplexity: 7.3807
Epoch [2/3], Step [2244/3236], Loss: 2.1323, Perplexity: 8.4343
Epoch [2/3], Step [2245/3236], Loss: 2.0667, Perplexity: 7.8991
Epoch [2/3], Step [2246/3236], Loss: 2.2770, Perplexity: 9.7471
Epoch [2/3], Step [2247/3236], Loss: 2.0421, Perplexity: 7.7070
Epoch [2/3], Step [2248/3236], Loss: 2.5005, Perplexity: 12.1886
Epoch [2/3], Step [2249/3236], Loss: 2.2652, Perplexity: 9.6330
Epoch [2/3], Step [2250/3236], Loss: 2.1168, Perplexity: 8.3045
Epoch [2/3], Step [2251/3236], Loss: 2.2442, Perplexity: 9.4327
Epoch [2/3], Step [2252/3236], Loss: 2.2562, Perplexity: 9.5469
Epoch [2/3], Step [2253/3236], Loss: 1.9641, Perplexity: 7.1283
Epoch [2/3], Step [2254/3236], Loss: 2.0957, Perplexity: 8.1312
Epoch [2/3], Step [2255/3236], Loss: 2.4101, Perplexity: 11.1350
Epoch [2/3], Step [2256/3236], Loss: 2.1131, Perplexity: 8.2735
Epoch [2/3], Step [2257/3236], Loss: 1.9615, Perplexity: 7.1101
Epoch [2/3], Step [2258/3236], Loss: 2.1493, Perplexity: 8.5788
Epoch [2/3], Step [2259/3236], Loss: 2.2237, Perplexity: 9.2411
Epoch [2/3], Step [2260/3236], Loss: 2.3966, Perplexity: 10.9860
Epoch [2/3], Step [2261/3236], Loss: 2.1101, Perplexity: 8.2488
Epoch [2/3], Step [2262/3236], Loss: 2.3457, Perplexity: 10.4406
Epoch [2/3], Step [2263/3236], Loss: 1.9267, Perplexity: 6.8668
Epoch [2/3], Step [2264/3236], Loss: 2.0058, Perplexity: 7.4318
Epoch [2/3], Step [2265/3236], Loss: 2.8754, Perplexity: 17.7321
Epoch [2/3], Step [2266/3236], Loss: 2.2234, Perplexity: 9.2385
Epoch [2/3], Step [2267/3236], Loss: 2.1431, Perplexity: 8.5256
Epoch [2/3], Step [2268/3236], Loss: 2.1401, Perplexity: 8.5005
Epoch [2/3], Step [2269/3236], Loss: 2.3995, Perplexity: 11.0175
Epoch [2/3], Step [2270/3236], Loss: 2.0555, Perplexity: 7.8111
Epoch [2/3], Step [2271/3236], Loss: 2.2462, Perplexity: 9.4516
Epoch [2/3], Step [2272/3236], Loss: 2.3954, Perplexity: 10.9726
Epoch [2/3], Step [2273/3236], Loss: 2.2313, Perplexity: 9.3124
Epoch [2/3], Step [2274/3236], Loss: 2.0641, Perplexity: 7.8781
Epoch [2/3], Step [2275/3236], Loss: 2.4168, Perplexity: 11.2101
Epoch [2/3], Step [2276/3236], Loss: 2.0258, Perplexity: 7.5825
Epoch [2/3], Step [2277/3236], Loss: 2.1755, Perplexity: 8.8065
Epoch [2/3], Step [2278/3236], Loss: 2.1619, Perplexity: 8.6877
Epoch [2/3], Step [2279/3236], Loss: 2.1810, Perplexity: 8.8551
Epoch [2/3], Step [2280/3236], Loss: 2.1826, Perplexity: 8.8696
Epoch [2/3], Step [2281/3236], Loss: 2.0000, Perplexity: 7.3893
Epoch [2/3], Step [2282/3236], Loss: 2.0182, Perplexity: 7.5250
Epoch [2/3], Step [2283/3236], Loss: 2.1726, Perplexity: 8.7807
Epoch [2/3], Step [2284/3236], Loss: 2.0568, Perplexity: 7.8206
Epoch [2/3], Step [2285/3236], Loss: 2.0799, Perplexity: 8.0036
Epoch [2/3], Step [2286/3236], Loss: 2.1262, Perplexity: 8.3825
Epoch [2/3], Step [2287/3236], Loss: 2.0188, Perplexity: 7.5296
Epoch [2/3], Step [2288/3236], Loss: 2.2020, Perplexity: 9.0431
Epoch [2/3], Step [2289/3236], Loss: 2.0845, Perplexity: 8.0405
Epoch [2/3], Step [2290/3236], Loss: 2.1511, Perplexity: 8.5941
Epoch [2/3], Step [2291/3236], Loss: 2.0981, Perplexity: 8.1509
Epoch [2/3], Step [2292/3236], Loss: 2.1796, Perplexity: 8.8430
Epoch [2/3], Step [2293/3236], Loss: 3.6294, Perplexity: 37.6892
Epoch [2/3], Step [2294/3236], Loss: 2.2149, Perplexity: 9.1607
Epoch [2/3], Step [2295/3236], Loss: 2.4906, Perplexity: 12.0684
Epoch [2/3], Step [2296/3236], Loss: 2.4111, Perplexity: 11.1460
Epoch [2/3], Step [2297/3236], Loss: 3.2426, Perplexity: 25.6003
Epoch [2/3], Step [2298/3236], Loss: 2.0451, Perplexity: 7.7303
Epoch [2/3], Step [2299/3236], Loss: 2.0274, Perplexity: 7.5940
Epoch [2/3], Step [2300/3236], Loss: 2.1927, Perplexity: 8.9593
Epoch [2/3], Step [2301/3236], Loss: 1.9535, Perplexity: 7.0537
Epoch [2/3], Step [2302/3236], Loss: 1.9621, Perplexity: 7.1146
Epoch [2/3], Step [2303/3236], Loss: 2.0721, Perplexity: 7.9414
Epoch [2/3], Step [2304/3236], Loss: 2.2322, Perplexity: 9.3208
Epoch [2/3], Step [2305/3236], Loss: 2.3695, Perplexity: 10.6920
Epoch [2/3], Step [2306/3236], Loss: 2.0694, Perplexity: 7.9199
Epoch [2/3], Step [2307/3236], Loss: 2.0916, Perplexity: 8.0976
Epoch [2/3], Step [2308/3236], Loss: 2.4499, Perplexity: 11.5870
Epoch [2/3], Step [2309/3236], Loss: 2.2388, Perplexity: 9.3824
Epoch [2/3], Step [2310/3236], Loss: 2.1096, Perplexity: 8.2446
Epoch [2/3], Step [2311/3236], Loss: 2.2463, Perplexity: 9.4528
Epoch [2/3], Step [2312/3236], Loss: 2.2471, Perplexity: 9.4606
Epoch [2/3], Step [2313/3236], Loss: 2.1107, Perplexity: 8.2544
Epoch [2/3], Step [2314/3236], Loss: 2.1427, Perplexity: 8.5223
Epoch [2/3], Step [2315/3236], Loss: 2.1740, Perplexity: 8.7930
Epoch [2/3], Step [2316/3236], Loss: 2.0939, Perplexity: 8.1165
Epoch [2/3], Step [2317/3236], Loss: 2.1498, Perplexity: 8.5833
Epoch [2/3], Step [2318/3236], Loss: 2.3605, Perplexity: 10.5959
Epoch [2/3], Step [2319/3236], Loss: 2.0199, Perplexity: 7.5373
Epoch [2/3], Step [2320/3236], Loss: 2.1358, Perplexity: 8.4634
Epoch [2/3], Step [2321/3236], Loss: 2.0470, Perplexity: 7.7449
Epoch [2/3], Step [2322/3236], Loss: 2.7745, Perplexity: 16.0298
Epoch [2/3], Step [2323/3236], Loss: 2.1178, Perplexity: 8.3131
Epoch [2/3], Step [2324/3236], Loss: 2.1347, Perplexity: 8.4548
Epoch [2/3], Step [2325/3236], Loss: 2.1633, Perplexity: 8.7002
Epoch [2/3], Step [2326/3236], Loss: 2.0673, Perplexity: 7.9035
Epoch [2/3], Step [2327/3236], Loss: 2.0761, Perplexity: 7.9729
Epoch [2/3], Step [2328/3236], Loss: 2.0925, Perplexity: 8.1052
Epoch [2/3], Step [2329/3236], Loss: 2.0990, Perplexity: 8.1580
Epoch [2/3], Step [2330/3236], Loss: 2.0963, Perplexity: 8.1360
Epoch [2/3], Step [2331/3236], Loss: 2.2406, Perplexity: 9.3991
Epoch [2/3], Step [2332/3236], Loss: 2.1042, Perplexity: 8.2002
Epoch [2/3], Step [2333/3236], Loss: 2.3873, Perplexity: 10.8841
Epoch [2/3], Step [2334/3236], Loss: 2.3955, Perplexity: 10.9737
Epoch [2/3], Step [2335/3236], Loss: 2.4097, Perplexity: 11.1302
Epoch [2/3], Step [2336/3236], Loss: 1.9687, Perplexity: 7.1616
Epoch [2/3], Step [2337/3236], Loss: 2.1825, Perplexity: 8.8688
Epoch [2/3], Step [2338/3236], Loss: 2.0312, Perplexity: 7.6235
Epoch [2/3], Step [2339/3236], Loss: 2.0472, Perplexity: 7.7463
Epoch [2/3], Step [2340/3236], Loss: 1.9716, Perplexity: 7.1820
Epoch [2/3], Step [2341/3236], Loss: 2.0321, Perplexity: 7.6300
Epoch [2/3], Step [2342/3236], Loss: 2.0874, Perplexity: 8.0636
Epoch [2/3], Step [2343/3236], Loss: 2.1135, Perplexity: 8.2769
Epoch [2/3], Step [2344/3236], Loss: 2.5007, Perplexity: 12.1910
Epoch [2/3], Step [2345/3236], Loss: 1.9704, Perplexity: 7.1737
Epoch [2/3], Step [2346/3236], Loss: 2.4340, Perplexity: 11.4049
Epoch [2/3], Step [2347/3236], Loss: 1.9167, Perplexity: 6.7984
Epoch [2/3], Step [2348/3236], Loss: 2.1117, Perplexity: 8.2625
Epoch [2/3], Step [2349/3236], Loss: 2.0505, Perplexity: 7.7717
Epoch [2/3], Step [2350/3236], Loss: 1.8759, Perplexity: 6.5265
Epoch [2/3], Step [2351/3236], Loss: 2.5164, Perplexity: 12.3837
Epoch [2/3], Step [2352/3236], Loss: 2.1072, Perplexity: 8.2250
Epoch [2/3], Step [2353/3236], Loss: 2.4562, Perplexity: 11.6602
Epoch [2/3], Step [2354/3236], Loss: 2.1875, Perplexity: 8.9133
Epoch [2/3], Step [2355/3236], Loss: 2.0537, Perplexity: 7.7968
Epoch [2/3], Step [2356/3236], Loss: 2.0891, Perplexity: 8.0775
Epoch [2/3], Step [2357/3236], Loss: 2.1810, Perplexity: 8.8556
Epoch [2/3], Step [2358/3236], Loss: 1.9276, Perplexity: 6.8730
Epoch [2/3], Step [2359/3236], Loss: 2.2181, Perplexity: 9.1895
Epoch [2/3], Step [2360/3236], Loss: 2.8728, Perplexity: 17.6864
Epoch [2/3], Step [2361/3236], Loss: 2.0166, Perplexity: 7.5128
Epoch [2/3], Step [2362/3236], Loss: 2.1026, Perplexity: 8.1871
Epoch [2/3], Step [2363/3236], Loss: 2.3863, Perplexity: 10.8729
Epoch [2/3], Step [2364/3236], Loss: 2.2099, Perplexity: 9.1145
Epoch [2/3], Step [2365/3236], Loss: 2.2142, Perplexity: 9.1539
Epoch [2/3], Step [2366/3236], Loss: 2.0182, Perplexity: 7.5248
Epoch [2/3], Step [2367/3236], Loss: 2.0540, Perplexity: 7.7992
Epoch [2/3], Step [2368/3236], Loss: 2.0974, Perplexity: 8.1451
Epoch [2/3], Step [2369/3236], Loss: 2.2757, Perplexity: 9.7350
Epoch [2/3], Step [2370/3236], Loss: 2.3674, Perplexity: 10.6692
Epoch [2/3], Step [2371/3236], Loss: 2.3438, Perplexity: 10.4211
Epoch [2/3], Step [2372/3236], Loss: 2.3897, Perplexity: 10.9102
Epoch [2/3], Step [2373/3236], Loss: 2.1821, Perplexity: 8.8648
Epoch [2/3], Step [2374/3236], Loss: 2.0163, Perplexity: 7.5102
Epoch [2/3], Step [2375/3236], Loss: 2.0903, Perplexity: 8.0871
Epoch [2/3], Step [2376/3236], Loss: 2.1974, Perplexity: 9.0019
Epoch [2/3], Step [2377/3236], Loss: 1.9826, Perplexity: 7.2613
Epoch [2/3], Step [2378/3236], Loss: 2.0375, Perplexity: 7.6711
Epoch [2/3], Step [2379/3236], Loss: 1.9935, Perplexity: 7.3415
Epoch [2/3], Step [2380/3236], Loss: 2.4094, Perplexity: 11.1278
Epoch [2/3], Step [2381/3236], Loss: 2.0491, Perplexity: 7.7610
Epoch [2/3], Step [2382/3236], Loss: 2.1356, Perplexity: 8.4623
Epoch [2/3], Step [2383/3236], Loss: 2.1548, Perplexity: 8.6263
Epoch [2/3], Step [2384/3236], Loss: 1.9938, Perplexity: 7.3437
Epoch [2/3], Step [2385/3236], Loss: 2.3330, Perplexity: 10.3091
Epoch [2/3], Step [2386/3236], Loss: 2.0215, Perplexity: 7.5496
Epoch [2/3], Step [2387/3236], Loss: 2.2521, Perplexity: 9.5073
Epoch [2/3], Step [2388/3236], Loss: 2.2365, Perplexity: 9.3606
Epoch [2/3], Step [2389/3236], Loss: 2.0506, Perplexity: 7.7728
Epoch [2/3], Step [2390/3236], Loss: 1.9789, Perplexity: 7.2348
Epoch [2/3], Step [2391/3236], Loss: 2.0001, Perplexity: 7.3895
Epoch [2/3], Step [2392/3236], Loss: 2.2888, Perplexity: 9.8628
Epoch [2/3], Step [2393/3236], Loss: 2.1186, Perplexity: 8.3192
Epoch [2/3], Step [2394/3236], Loss: 2.0081, Perplexity: 7.4489
Epoch [2/3], Step [2395/3236], Loss: 2.2150, Perplexity: 9.1613
Epoch [2/3], Step [2396/3236], Loss: 2.1349, Perplexity: 8.4560
Epoch [2/3], Step [2397/3236], Loss: 2.2985, Perplexity: 9.9596
Epoch [2/3], Step [2398/3236], Loss: 2.3022, Perplexity: 9.9960
Epoch [2/3], Step [2399/3236], Loss: 2.7685, Perplexity: 15.9342
Epoch [2/3], Step [2400/3236], Loss: 2.5796, Perplexity: 13.1913
Epoch [2/3], Step [2401/3236], Loss: 2.1513, Perplexity: 8.5957
Epoch [2/3], Step [2402/3236], Loss: 2.2572, Perplexity: 9.5566
Epoch [2/3], Step [2403/3236], Loss: 2.0391, Perplexity: 7.6833
Epoch [2/3], Step [2404/3236], Loss: 1.9800, Perplexity: 7.2427
Epoch [2/3], Step [2405/3236], Loss: 2.0417, Perplexity: 7.7035
Epoch [2/3], Step [2406/3236], Loss: 1.9640, Perplexity: 7.1280
Epoch [2/3], Step [2407/3236], Loss: 2.1091, Perplexity: 8.2406
Epoch [2/3], Step [2408/3236], Loss: 2.9327, Perplexity: 18.7784
Epoch [2/3], Step [2409/3236], Loss: 2.1339, Perplexity: 8.4477
Epoch [2/3], Step [2410/3236], Loss: 2.0535, Perplexity: 7.7951
Epoch [2/3], Step [2411/3236], Loss: 2.0219, Perplexity: 7.5528
Epoch [2/3], Step [2412/3236], Loss: 2.5367, Perplexity: 12.6382
Epoch [2/3], Step [2413/3236], Loss: 2.0102, Perplexity: 7.4646
Epoch [2/3], Step [2414/3236], Loss: 2.0802, Perplexity: 8.0063
Epoch [2/3], Step [2415/3236], Loss: 2.1179, Perplexity: 8.3133
Epoch [2/3], Step [2416/3236], Loss: 2.1169, Perplexity: 8.3054
Epoch [2/3], Step [2417/3236], Loss: 3.0157, Perplexity: 20.4032
Epoch [2/3], Step [2418/3236], Loss: 1.7716, Perplexity: 5.8802
Epoch [2/3], Step [2419/3236], Loss: 1.9965, Perplexity: 7.3631
Epoch [2/3], Step [2420/3236], Loss: 2.3084, Perplexity: 10.0579
Epoch [2/3], Step [2421/3236], Loss: 2.0362, Perplexity: 7.6618
Epoch [2/3], Step [2422/3236], Loss: 2.1022, Perplexity: 8.1842
Epoch [2/3], Step [2423/3236], Loss: 2.1395, Perplexity: 8.4952
Epoch [2/3], Step [2424/3236], Loss: 2.0193, Perplexity: 7.5331
Epoch [2/3], Step [2425/3236], Loss: 2.5469, Perplexity: 12.7678
Epoch [2/3], Step [2426/3236], Loss: 2.0802, Perplexity: 8.0057
Epoch [2/3], Step [2427/3236], Loss: 2.2251, Perplexity: 9.2544
Epoch [2/3], Step [2428/3236], Loss: 3.1076, Perplexity: 22.3681
Epoch [2/3], Step [2429/3236], Loss: 2.1522, Perplexity: 8.6034
Epoch [2/3], Step [2430/3236], Loss: 2.1441, Perplexity: 8.5346
Epoch [2/3], Step [2431/3236], Loss: 2.0342, Perplexity: 7.6461
Epoch [2/3], Step [2432/3236], Loss: 2.0238, Perplexity: 7.5668
Epoch [2/3], Step [2433/3236], Loss: 2.1544, Perplexity: 8.6226
Epoch [2/3], Step [2434/3236], Loss: 2.0476, Perplexity: 7.7491
Epoch [2/3], Step [2435/3236], Loss: 2.1228, Perplexity: 8.3544
Epoch [2/3], Step [2436/3236], Loss: 2.1956, Perplexity: 8.9852
Epoch [2/3], Step [2437/3236], Loss: 2.1369, Perplexity: 8.4727
Epoch [2/3], Step [2438/3236], Loss: 1.9685, Perplexity: 7.1597
Epoch [2/3], Step [2439/3236], Loss: 2.0584, Perplexity: 7.8335
Epoch [2/3], Step [2440/3236], Loss: 1.9026, Perplexity: 6.7033
Epoch [2/3], Step [2441/3236], Loss: 2.5490, Perplexity: 12.7947
Epoch [2/3], Step [2442/3236], Loss: 1.9565, Perplexity: 7.0747
Epoch [2/3], Step [2443/3236], Loss: 2.2221, Perplexity: 9.2267
Epoch [2/3], Step [2444/3236], Loss: 2.0850, Perplexity: 8.0448
Epoch [2/3], Step [2445/3236], Loss: 2.2831, Perplexity: 9.8070
Epoch [2/3], Step [2446/3236], Loss: 2.1334, Perplexity: 8.4434
Epoch [2/3], Step [2447/3236], Loss: 2.1794, Perplexity: 8.8408
Epoch [2/3], Step [2448/3236], Loss: 2.1900, Perplexity: 8.9356
Epoch [2/3], Step [2449/3236], Loss: 2.0042, Perplexity: 7.4202
Epoch [2/3], Step [2450/3236], Loss: 2.0874, Perplexity: 8.0642
Epoch [2/3], Step [2451/3236], Loss: 2.1605, Perplexity: 8.6757
Epoch [2/3], Step [2452/3236], Loss: 2.1586, Perplexity: 8.6592
Epoch [2/3], Step [2453/3236], Loss: 1.9635, Perplexity: 7.1239
Epoch [2/3], Step [2454/3236], Loss: 1.9838, Perplexity: 7.2704
Epoch [2/3], Step [2455/3236], Loss: 2.1209, Perplexity: 8.3385
Epoch [2/3], Step [2456/3236], Loss: 2.3142, Perplexity: 10.1164
Epoch [2/3], Step [2457/3236], Loss: 2.0954, Perplexity: 8.1288
Epoch [2/3], Step [2458/3236], Loss: 2.0888, Perplexity: 8.0751
Epoch [2/3], Step [2459/3236], Loss: 2.1263, Perplexity: 8.3836
Epoch [2/3], Step [2460/3236], Loss: 1.9875, Perplexity: 7.2972
Epoch [2/3], Step [2461/3236], Loss: 2.3332, Perplexity: 10.3111
Epoch [2/3], Step [2462/3236], Loss: 2.4259, Perplexity: 11.3129
Epoch [2/3], Step [2463/3236], Loss: 2.0913, Perplexity: 8.0951
Epoch [2/3], Step [2464/3236], Loss: 2.0253, Perplexity: 7.5781
Epoch [2/3], Step [2465/3236], Loss: 2.1072, Perplexity: 8.2249
Epoch [2/3], Step [2466/3236], Loss: 2.1193, Perplexity: 8.3257
Epoch [2/3], Step [2467/3236], Loss: 2.0100, Perplexity: 7.4635
Epoch [2/3], Step [2468/3236], Loss: 2.1941, Perplexity: 8.9716
Epoch [2/3], Step [2469/3236], Loss: 2.0555, Perplexity: 7.8106
Epoch [2/3], Step [2470/3236], Loss: 2.6210, Perplexity: 13.7490
Epoch [2/3], Step [2471/3236], Loss: 2.2733, Perplexity: 9.7112
Epoch [2/3], Step [2472/3236], Loss: 2.1318, Perplexity: 8.4302
Epoch [2/3], Step [2473/3236], Loss: 2.1210, Perplexity: 8.3391
Epoch [2/3], Step [2474/3236], Loss: 2.2710, Perplexity: 9.6890
Epoch [2/3], Step [2475/3236], Loss: 2.0007, Perplexity: 7.3941
Epoch [2/3], Step [2476/3236], Loss: 2.1303, Perplexity: 8.4170
Epoch [2/3], Step [2477/3236], Loss: 2.0637, Perplexity: 7.8747
Epoch [2/3], Step [2478/3236], Loss: 2.0845, Perplexity: 8.0408
Epoch [2/3], Step [2479/3236], Loss: 2.1625, Perplexity: 8.6925
Epoch [2/3], Step [2480/3236], Loss: 2.0627, Perplexity: 7.8670
Epoch [2/3], Step [2481/3236], Loss: 2.0170, Perplexity: 7.5154
Epoch [2/3], Step [2482/3236], Loss: 2.4376, Perplexity: 11.4460
Epoch [2/3], Step [2483/3236], Loss: 2.2590, Perplexity: 9.5739
Epoch [2/3], Step [2484/3236], Loss: 2.0324, Perplexity: 7.6322
Epoch [2/3], Step [2485/3236], Loss: 2.2427, Perplexity: 9.4186
Epoch [2/3], Step [2486/3236], Loss: 2.1232, Perplexity: 8.3577
Epoch [2/3], Step [2487/3236], Loss: 2.2286, Perplexity: 9.2869
Epoch [2/3], Step [2488/3236], Loss: 2.8515, Perplexity: 17.3136
Epoch [2/3], Step [2489/3236], Loss: 2.2277, Perplexity: 9.2787
Epoch [2/3], Step [2490/3236], Loss: 2.2596, Perplexity: 9.5790
Epoch [2/3], Step [2491/3236], Loss: 2.0875, Perplexity: 8.0651
Epoch [2/3], Step [2492/3236], Loss: 2.1920, Perplexity: 8.9530
Epoch [2/3], Step [2493/3236], Loss: 2.0897, Perplexity: 8.0822
Epoch [2/3], Step [2494/3236], Loss: 2.0715, Perplexity: 7.9364
Epoch [2/3], Step [2495/3236], Loss: 2.1073, Perplexity: 8.2263
Epoch [2/3], Step [2496/3236], Loss: 2.5547, Perplexity: 12.8676
Epoch [2/3], Step [2497/3236], Loss: 2.1334, Perplexity: 8.4438
Epoch [2/3], Step [2498/3236], Loss: 1.9876, Perplexity: 7.2978
Epoch [2/3], Step [2499/3236], Loss: 2.0834, Perplexity: 8.0317
Epoch [2/3], Step [2500/3236], Loss: 2.1082, Perplexity: 8.2331
Epoch [2/3], Step [2501/3236], Loss: 2.1786, Perplexity: 8.8338
Epoch [2/3], Step [2502/3236], Loss: 2.1077, Perplexity: 8.2291
Epoch [2/3], Step [2503/3236], Loss: 2.0759, Perplexity: 7.9714
Epoch [2/3], Step [2504/3236], Loss: 2.3456, Perplexity: 10.4392
Epoch [2/3], Step [2505/3236], Loss: 2.1935, Perplexity: 8.9664
Epoch [2/3], Step [2506/3236], Loss: 2.3710, Perplexity: 10.7078
Epoch [2/3], Step [2507/3236], Loss: 2.2181, Perplexity: 9.1898
Epoch [2/3], Step [2508/3236], Loss: 2.0368, Perplexity: 7.6663
Epoch [2/3], Step [2509/3236], Loss: 2.1710, Perplexity: 8.7666
Epoch [2/3], Step [2510/3236], Loss: 1.9616, Perplexity: 7.1107
Epoch [2/3], Step [2511/3236], Loss: 2.1922, Perplexity: 8.9553
Epoch [2/3], Step [2512/3236], Loss: 2.3552, Perplexity: 10.5405
Epoch [2/3], Step [2513/3236], Loss: 2.0195, Perplexity: 7.5346
Epoch [2/3], Step [2514/3236], Loss: 2.1249, Perplexity: 8.3724
Epoch [2/3], Step [2515/3236], Loss: 2.2791, Perplexity: 9.7676
Epoch [2/3], Step [2516/3236], Loss: 1.9202, Perplexity: 6.8220
Epoch [2/3], Step [2517/3236], Loss: 2.1184, Perplexity: 8.3174
Epoch [2/3], Step [2518/3236], Loss: 2.5207, Perplexity: 12.4372
Epoch [2/3], Step [2519/3236], Loss: 2.1974, Perplexity: 9.0015
Epoch [2/3], Step [2520/3236], Loss: 2.0710, Perplexity: 7.9328
Epoch [2/3], Step [2521/3236], Loss: 2.1072, Perplexity: 8.2254
Epoch [2/3], Step [2522/3236], Loss: 2.1152, Perplexity: 8.2915
Epoch [2/3], Step [2523/3236], Loss: 2.1462, Perplexity: 8.5523
Epoch [2/3], Step [2524/3236], Loss: 1.9496, Perplexity: 7.0261
Epoch [2/3], Step [2525/3236], Loss: 2.0244, Perplexity: 7.5719
Epoch [2/3], Step [2526/3236], Loss: 1.9701, Perplexity: 7.1716
Epoch [2/3], Step [2527/3236], Loss: 2.1885, Perplexity: 8.9216
Epoch [2/3], Step [2528/3236], Loss: 2.1902, Perplexity: 8.9366
Epoch [2/3], Step [2529/3236], Loss: 2.2508, Perplexity: 9.4957
Epoch [2/3], Step [2530/3236], Loss: 2.1594, Perplexity: 8.6659
Epoch [2/3], Step [2531/3236], Loss: 2.2275, Perplexity: 9.2763
Epoch [2/3], Step [2532/3236], Loss: 2.2035, Perplexity: 9.0562
Epoch [2/3], Step [2533/3236], Loss: 2.1069, Perplexity: 8.2227
Epoch [2/3], Step [2534/3236], Loss: 2.4400, Perplexity: 11.4732
Epoch [2/3], Step [2535/3236], Loss: 2.0354, Perplexity: 7.6557
Epoch [2/3], Step [2536/3236], Loss: 2.2351, Perplexity: 9.3471
Epoch [2/3], Step [2537/3236], Loss: 2.0276, Perplexity: 7.5958
Epoch [2/3], Step [2538/3236], Loss: 2.1819, Perplexity: 8.8627
Epoch [2/3], Step [2539/3236], Loss: 2.0167, Perplexity: 7.5133
Epoch [2/3], Step [2540/3236], Loss: 2.0741, Perplexity: 7.9577
Epoch [2/3], Step [2541/3236], Loss: 2.0158, Perplexity: 7.5071
Epoch [2/3], Step [2542/3236], Loss: 2.1859, Perplexity: 8.8990
Epoch [2/3], Step [2543/3236], Loss: 2.1216, Perplexity: 8.3448
Epoch [2/3], Step [2544/3236], Loss: 1.9462, Perplexity: 7.0021
Epoch [2/3], Step [2545/3236], Loss: 2.0823, Perplexity: 8.0233
Epoch [2/3], Step [2546/3236], Loss: 1.9696, Perplexity: 7.1680
Epoch [2/3], Step [2547/3236], Loss: 2.0134, Perplexity: 7.4889
Epoch [2/3], Step [2548/3236], Loss: 2.4462, Perplexity: 11.5445
Epoch [2/3], Step [2549/3236], Loss: 2.2446, Perplexity: 9.4369
Epoch [2/3], Step [2550/3236], Loss: 2.1079, Perplexity: 8.2310
Epoch [2/3], Step [2551/3236], Loss: 2.5335, Perplexity: 12.5969
Epoch [2/3], Step [2552/3236], Loss: 2.3572, Perplexity: 10.5609
Epoch [2/3], Step [2553/3236], Loss: 2.1829, Perplexity: 8.8719
Epoch [2/3], Step [2554/3236], Loss: 2.0335, Perplexity: 7.6405
Epoch [2/3], Step [2555/3236], Loss: 2.1017, Perplexity: 8.1801
Epoch [2/3], Step [2556/3236], Loss: 2.0505, Perplexity: 7.7722
Epoch [2/3], Step [2557/3236], Loss: 2.1258, Perplexity: 8.3794
Epoch [2/3], Step [2558/3236], Loss: 2.1800, Perplexity: 8.8459
Epoch [2/3], Step [2559/3236], Loss: 2.0745, Perplexity: 7.9603
Epoch [2/3], Step [2560/3236], Loss: 2.0812, Perplexity: 8.0144
Epoch [2/3], Step [2561/3236], Loss: 2.6120, Perplexity: 13.6262
Epoch [2/3], Step [2562/3236], Loss: 2.0825, Perplexity: 8.0248
Epoch [2/3], Step [2563/3236], Loss: 2.0931, Perplexity: 8.1097
Epoch [2/3], Step [2564/3236], Loss: 2.0677, Perplexity: 7.9067
Epoch [2/3], Step [2565/3236], Loss: 2.3383, Perplexity: 10.3633
Epoch [2/3], Step [2566/3236], Loss: 2.0534, Perplexity: 7.7943
Epoch [2/3], Step [2567/3236], Loss: 2.1720, Perplexity: 8.7761
Epoch [2/3], Step [2568/3236], Loss: 2.4465, Perplexity: 11.5480
Epoch [2/3], Step [2569/3236], Loss: 2.1186, Perplexity: 8.3193
Epoch [2/3], Step [2570/3236], Loss: 2.1351, Perplexity: 8.4581
Epoch [2/3], Step [2571/3236], Loss: 2.1512, Perplexity: 8.5948
Epoch [2/3], Step [2572/3236], Loss: 2.6522, Perplexity: 14.1855
Epoch [2/3], Step [2573/3236], Loss: 2.1427, Perplexity: 8.5224
Epoch [2/3], Step [2574/3236], Loss: 2.0477, Perplexity: 7.7498
Epoch [2/3], Step [2575/3236], Loss: 2.1058, Perplexity: 8.2133
Epoch [2/3], Step [2576/3236], Loss: 2.0058, Perplexity: 7.4320
Epoch [2/3], Step [2577/3236], Loss: 2.2673, Perplexity: 9.6537
Epoch [2/3], Step [2578/3236], Loss: 2.1433, Perplexity: 8.5279
Epoch [2/3], Step [2579/3236], Loss: 2.1032, Perplexity: 8.1922
Epoch [2/3], Step [2580/3236], Loss: 2.0317, Perplexity: 7.6270
Epoch [2/3], Step [2581/3236], Loss: 2.1018, Perplexity: 8.1807
Epoch [2/3], Step [2582/3236], Loss: 2.1966, Perplexity: 8.9943
Epoch [2/3], Step [2583/3236], Loss: 2.1550, Perplexity: 8.6279
Epoch [2/3], Step [2584/3236], Loss: 2.1119, Perplexity: 8.2636
Epoch [2/3], Step [2585/3236], Loss: 1.9692, Perplexity: 7.1650
Epoch [2/3], Step [2586/3236], Loss: 2.3735, Perplexity: 10.7346
Epoch [2/3], Step [2587/3236], Loss: 2.1978, Perplexity: 9.0054
Epoch [2/3], Step [2588/3236], Loss: 2.0656, Perplexity: 7.8902
Epoch [2/3], Step [2589/3236], Loss: 2.1695, Perplexity: 8.7542
Epoch [2/3], Step [2590/3236], Loss: 2.0324, Perplexity: 7.6321
Epoch [2/3], Step [2591/3236], Loss: 2.0337, Perplexity: 7.6422
Epoch [2/3], Step [2592/3236], Loss: 2.8238, Perplexity: 16.8406
Epoch [2/3], Step [2593/3236], Loss: 2.2301, Perplexity: 9.3004
Epoch [2/3], Step [2594/3236], Loss: 2.2203, Perplexity: 9.2103
Epoch [2/3], Step [2595/3236], Loss: 2.3230, Perplexity: 10.2063
Epoch [2/3], Step [2596/3236], Loss: 2.1130, Perplexity: 8.2727
Epoch [2/3], Step [2597/3236], Loss: 2.0664, Perplexity: 7.8965
Epoch [2/3], Step [2598/3236], Loss: 2.0488, Perplexity: 7.7589
Epoch [2/3], Step [2599/3236], Loss: 2.0701, Perplexity: 7.9257
Epoch [2/3], Step [2600/3236], Loss: 2.0069, Perplexity: 7.4403
Epoch [2/3], Step [2601/3236], Loss: 2.0567, Perplexity: 7.8201
Epoch [2/3], Step [2602/3236], Loss: 2.2050, Perplexity: 9.0698
Epoch [2/3], Step [2603/3236], Loss: 2.0104, Perplexity: 7.4663
Epoch [2/3], Step [2604/3236], Loss: 2.0676, Perplexity: 7.9057
Epoch [2/3], Step [2605/3236], Loss: 2.3010, Perplexity: 9.9841
Epoch [2/3], Step [2606/3236], Loss: 2.0062, Perplexity: 7.4351
Epoch [2/3], Step [2607/3236], Loss: 2.0957, Perplexity: 8.1310
Epoch [2/3], Step [2608/3236], Loss: 2.1708, Perplexity: 8.7650
Epoch [2/3], Step [2609/3236], Loss: 2.8509, Perplexity: 17.3033
Epoch [2/3], Step [2610/3236], Loss: 1.9872, Perplexity: 7.2952
Epoch [2/3], Step [2611/3236], Loss: 2.1227, Perplexity: 8.3536
Epoch [2/3], Step [2612/3236], Loss: 2.1328, Perplexity: 8.4389
Epoch [2/3], Step [2613/3236], Loss: 2.1776, Perplexity: 8.8254
Epoch [2/3], Step [2614/3236], Loss: 2.1419, Perplexity: 8.5158
Epoch [2/3], Step [2615/3236], Loss: 1.9736, Perplexity: 7.1967
Epoch [2/3], Step [2616/3236], Loss: 2.0501, Perplexity: 7.7684
Epoch [2/3], Step [2617/3236], Loss: 1.9495, Perplexity: 7.0255
Epoch [2/3], Step [2618/3236], Loss: 2.2297, Perplexity: 9.2973
Epoch [2/3], Step [2619/3236], Loss: 2.2355, Perplexity: 9.3516
Epoch [2/3], Step [2620/3236], Loss: 2.1036, Perplexity: 8.1954
Epoch [2/3], Step [2621/3236], Loss: 2.1761, Perplexity: 8.8115
Epoch [2/3], Step [2622/3236], Loss: 2.1539, Perplexity: 8.6187
Epoch [2/3], Step [2623/3236], Loss: 2.0779, Perplexity: 7.9879
Epoch [2/3], Step [2624/3236], Loss: 1.9535, Perplexity: 7.0530
Epoch [2/3], Step [2625/3236], Loss: 2.1009, Perplexity: 8.1739
Epoch [2/3], Step [2626/3236], Loss: 2.0316, Perplexity: 7.6259
Epoch [2/3], Step [2627/3236], Loss: 2.1051, Perplexity: 8.2078
Epoch [2/3], Step [2628/3236], Loss: 2.0115, Perplexity: 7.4742
Epoch [2/3], Step [2629/3236], Loss: 2.2307, Perplexity: 9.3060
Epoch [2/3], Step [2630/3236], Loss: 2.0003, Perplexity: 7.3915
Epoch [2/3], Step [2631/3236], Loss: 1.9422, Perplexity: 6.9738
Epoch [2/3], Step [2632/3236], Loss: 2.1910, Perplexity: 8.9445
Epoch [2/3], Step [2633/3236], Loss: 1.9546, Perplexity: 7.0613
Epoch [2/3], Step [2634/3236], Loss: 1.9593, Perplexity: 7.0942
Epoch [2/3], Step [2635/3236], Loss: 2.0794, Perplexity: 8.0000
Epoch [2/3], Step [2636/3236], Loss: 2.1258, Perplexity: 8.3792
Epoch [2/3], Step [2637/3236], Loss: 2.2528, Perplexity: 9.5139
Epoch [2/3], Step [2638/3236], Loss: 2.0962, Perplexity: 8.1352
Epoch [2/3], Step [2639/3236], Loss: 2.0585, Perplexity: 7.8343
Epoch [2/3], Step [2640/3236], Loss: 2.0662, Perplexity: 7.8950
Epoch [2/3], Step [2641/3236], Loss: 2.2024, Perplexity: 9.0464
Epoch [2/3], Step [2642/3236], Loss: 2.2884, Perplexity: 9.8596
Epoch [2/3], Step [2643/3236], Loss: 2.1298, Perplexity: 8.4130
Epoch [2/3], Step [2644/3236], Loss: 2.1895, Perplexity: 8.9311
Epoch [2/3], Step [2645/3236], Loss: 2.0993, Perplexity: 8.1603
Epoch [2/3], Step [2646/3236], Loss: 2.0171, Perplexity: 7.5165
Epoch [2/3], Step [2647/3236], Loss: 2.4612, Perplexity: 11.7190
Epoch [2/3], Step [2648/3236], Loss: 2.0985, Perplexity: 8.1539
Epoch [2/3], Step [2649/3236], Loss: 1.9271, Perplexity: 6.8693
Epoch [2/3], Step [2650/3236], Loss: 2.1058, Perplexity: 8.2136
Epoch [2/3], Step [2651/3236], Loss: 2.0793, Perplexity: 7.9990
Epoch [2/3], Step [2652/3236], Loss: 1.8954, Perplexity: 6.6549
Epoch [2/3], Step [2653/3236], Loss: 2.0296, Perplexity: 7.6107
Epoch [2/3], Step [2654/3236], Loss: 2.1500, Perplexity: 8.5852
Epoch [2/3], Step [2655/3236], Loss: 2.0797, Perplexity: 8.0020
Epoch [2/3], Step [2656/3236], Loss: 2.0786, Perplexity: 7.9931
Epoch [2/3], Step [2657/3236], Loss: 2.0944, Perplexity: 8.1205
Epoch [2/3], Step [2658/3236], Loss: 2.7503, Perplexity: 15.6469
Epoch [2/3], Step [2659/3236], Loss: 1.9509, Perplexity: 7.0348
Epoch [2/3], Step [2660/3236], Loss: 2.1438, Perplexity: 8.5318
Epoch [2/3], Step [2661/3236], Loss: 2.1156, Perplexity: 8.2946
Epoch [2/3], Step [2662/3236], Loss: 2.0372, Perplexity: 7.6688
Epoch [2/3], Step [2663/3236], Loss: 1.9637, Perplexity: 7.1254
Epoch [2/3], Step [2664/3236], Loss: 2.2084, Perplexity: 9.1009
Epoch [2/3], Step [2665/3236], Loss: 2.0895, Perplexity: 8.0808
Epoch [2/3], Step [2666/3236], Loss: 1.9796, Perplexity: 7.2400
Epoch [2/3], Step [2667/3236], Loss: 2.1256, Perplexity: 8.3782
Epoch [2/3], Step [2668/3236], Loss: 2.0565, Perplexity: 7.8186
Epoch [2/3], Step [2669/3236], Loss: 2.0461, Perplexity: 7.7379
Epoch [2/3], Step [2670/3236], Loss: 2.0460, Perplexity: 7.7366
Epoch [2/3], Step [2671/3236], Loss: 2.0697, Perplexity: 7.9223
Epoch [2/3], Step [2672/3236], Loss: 2.0592, Perplexity: 7.8400
Epoch [2/3], Step [2673/3236], Loss: 2.4053, Perplexity: 11.0819
Epoch [2/3], Step [2674/3236], Loss: 1.9970, Perplexity: 7.3672
Epoch [2/3], Step [2675/3236], Loss: 2.3404, Perplexity: 10.3849
Epoch [2/3], Step [2676/3236], Loss: 2.2038, Perplexity: 9.0595
Epoch [2/3], Step [2677/3236], Loss: 2.0002, Perplexity: 7.3907
Epoch [2/3], Step [2678/3236], Loss: 2.2215, Perplexity: 9.2211
Epoch [2/3], Step [2679/3236], Loss: 2.2967, Perplexity: 9.9416
Epoch [2/3], Step [2680/3236], Loss: 2.0665, Perplexity: 7.8974
Epoch [2/3], Step [2681/3236], Loss: 1.9700, Perplexity: 7.1708
Epoch [2/3], Step [2682/3236], Loss: 2.0824, Perplexity: 8.0233
Epoch [2/3], Step [2683/3236], Loss: 2.0793, Perplexity: 7.9988
Epoch [2/3], Step [2684/3236], Loss: 2.1329, Perplexity: 8.4395
Epoch [2/3], Step [2685/3236], Loss: 2.0768, Perplexity: 7.9791
Epoch [2/3], Step [2686/3236], Loss: 2.0413, Perplexity: 7.7009
Epoch [2/3], Step [2687/3236], Loss: 2.1376, Perplexity: 8.4793
Epoch [2/3], Step [2688/3236], Loss: 2.0509, Perplexity: 7.7752
Epoch [2/3], Step [2689/3236], Loss: 2.0375, Perplexity: 7.6714
Epoch [2/3], Step [2690/3236], Loss: 2.3575, Perplexity: 10.5642
Epoch [2/3], Step [2691/3236], Loss: 1.9940, Perplexity: 7.3447
Epoch [2/3], Step [2692/3236], Loss: 2.1761, Perplexity: 8.8116
Epoch [2/3], Step [2693/3236], Loss: 2.0701, Perplexity: 7.9256
Epoch [2/3], Step [2694/3236], Loss: 2.0863, Perplexity: 8.0548
Epoch [2/3], Step [2695/3236], Loss: 2.1156, Perplexity: 8.2949
Epoch [2/3], Step [2696/3236], Loss: 2.1443, Perplexity: 8.5362
Epoch [2/3], Step [2697/3236], Loss: 2.1745, Perplexity: 8.7978
Epoch [2/3], Step [2698/3236], Loss: 2.7545, Perplexity: 15.7136
Epoch [2/3], Step [2699/3236], Loss: 2.0835, Perplexity: 8.0322
Epoch [2/3], Step [2700/3236], Loss: 2.2576, Perplexity: 9.5598
Epoch [2/3], Step [2701/3236], Loss: 2.0863, Perplexity: 8.0552
Epoch [2/3], Step [2702/3236], Loss: 2.4957, Perplexity: 12.1302
Epoch [2/3], Step [2703/3236], Loss: 2.0891, Perplexity: 8.0780
Epoch [2/3], Step [2704/3236], Loss: 2.1336, Perplexity: 8.4456
Epoch [2/3], Step [2705/3236], Loss: 2.0530, Perplexity: 7.7911
Epoch [2/3], Step [2706/3236], Loss: 2.3040, Perplexity: 10.0146
Epoch [2/3], Step [2707/3236], Loss: 2.1194, Perplexity: 8.3257
Epoch [2/3], Step [2708/3236], Loss: 2.2227, Perplexity: 9.2327
Epoch [2/3], Step [2709/3236], Loss: 2.0449, Perplexity: 7.7286
Epoch [2/3], Step [2710/3236], Loss: 2.2037, Perplexity: 9.0586
Epoch [2/3], Step [2711/3236], Loss: 2.1810, Perplexity: 8.8552
Epoch [2/3], Step [2712/3236], Loss: 1.9528, Perplexity: 7.0486
Epoch [2/3], Step [2713/3236], Loss: 2.1194, Perplexity: 8.3260
Epoch [2/3], Step [2714/3236], Loss: 2.1060, Perplexity: 8.2156
Epoch [2/3], Step [2715/3236], Loss: 2.2137, Perplexity: 9.1494
Epoch [2/3], Step [2716/3236], Loss: 2.1111, Perplexity: 8.2575
Epoch [2/3], Step [2717/3236], Loss: 1.9221, Perplexity: 6.8350
Epoch [2/3], Step [2718/3236], Loss: 2.0561, Perplexity: 7.8153
Epoch [2/3], Step [2719/3236], Loss: 2.0668, Perplexity: 7.8991
Epoch [2/3], Step [2720/3236], Loss: 2.0906, Perplexity: 8.0894
Epoch [2/3], Step [2721/3236], Loss: 2.1020, Perplexity: 8.1823
Epoch [2/3], Step [2722/3236], Loss: 1.9820, Perplexity: 7.2573
Epoch [2/3], Step [2723/3236], Loss: 2.3439, Perplexity: 10.4217
Epoch [2/3], Step [2724/3236], Loss: 2.0859, Perplexity: 8.0517
Epoch [2/3], Step [2725/3236], Loss: 1.9366, Perplexity: 6.9354
Epoch [2/3], Step [2726/3236], Loss: 2.0325, Perplexity: 7.6330
Epoch [2/3], Step [2727/3236], Loss: 2.0745, Perplexity: 7.9607
Epoch [2/3], Step [2728/3236], Loss: 2.4103, Perplexity: 11.1378
Epoch [2/3], Step [2729/3236], Loss: 1.9848, Perplexity: 7.2775
Epoch [2/3], Step [2730/3236], Loss: 2.0810, Perplexity: 8.0128
Epoch [2/3], Step [2731/3236], Loss: 2.1823, Perplexity: 8.8665
Epoch [2/3], Step [2732/3236], Loss: 2.0867, Perplexity: 8.0584
Epoch [2/3], Step [2733/3236], Loss: 2.1653, Perplexity: 8.7172
Epoch [2/3], Step [2734/3236], Loss: 2.0742, Perplexity: 7.9581
Epoch [2/3], Step [2735/3236], Loss: 2.3670, Perplexity: 10.6659
Epoch [2/3], Step [2736/3236], Loss: 1.9050, Perplexity: 6.7193
Epoch [2/3], Step [2737/3236], Loss: 3.0703, Perplexity: 21.5474
Epoch [2/3], Step [2738/3236], Loss: 2.1049, Perplexity: 8.2066
Epoch [2/3], Step [2739/3236], Loss: 1.9450, Perplexity: 6.9938
Epoch [2/3], Step [2740/3236], Loss: 2.1110, Perplexity: 8.2562
Epoch [2/3], Step [2741/3236], Loss: 1.9803, Perplexity: 7.2450
Epoch [2/3], Step [2742/3236], Loss: 2.1608, Perplexity: 8.6777
Epoch [2/3], Step [2743/3236], Loss: 2.1365, Perplexity: 8.4694
Epoch [2/3], Step [2744/3236], Loss: 2.1350, Perplexity: 8.4571
Epoch [2/3], Step [2745/3236], Loss: 2.9774, Perplexity: 19.6374
Epoch [2/3], Step [2746/3236], Loss: 2.5459, Perplexity: 12.7545
Epoch [2/3], Step [2747/3236], Loss: 2.1257, Perplexity: 8.3791
Epoch [2/3], Step [2748/3236], Loss: 2.0342, Perplexity: 7.6461
Epoch [2/3], Step [2749/3236], Loss: 2.2614, Perplexity: 9.5964
Epoch [2/3], Step [2750/3236], Loss: 2.1193, Perplexity: 8.3254
Epoch [2/3], Step [2751/3236], Loss: 2.0392, Perplexity: 7.6842
Epoch [2/3], Step [2752/3236], Loss: 2.1194, Perplexity: 8.3258
Epoch [2/3], Step [2753/3236], Loss: 2.1052, Perplexity: 8.2091
Epoch [2/3], Step [2754/3236], Loss: 2.1707, Perplexity: 8.7644
Epoch [2/3], Step [2755/3236], Loss: 2.0479, Perplexity: 7.7513
Epoch [2/3], Step [2756/3236], Loss: 2.1392, Perplexity: 8.4926
Epoch [2/3], Step [2757/3236], Loss: 2.4012, Perplexity: 11.0360
Epoch [2/3], Step [2758/3236], Loss: 2.2206, Perplexity: 9.2133
Epoch [2/3], Step [2759/3236], Loss: 2.0794, Perplexity: 7.9997
Epoch [2/3], Step [2760/3236], Loss: 1.9124, Perplexity: 6.7692
Epoch [2/3], Step [2761/3236], Loss: 2.1509, Perplexity: 8.5927
Epoch [2/3], Step [2762/3236], Loss: 2.2164, Perplexity: 9.1744
Epoch [2/3], Step [2763/3236], Loss: 2.1898, Perplexity: 8.9336
Epoch [2/3], Step [2764/3236], Loss: 2.1005, Perplexity: 8.1705
Epoch [2/3], Step [2765/3236], Loss: 2.1432, Perplexity: 8.5270
Epoch [2/3], Step [2766/3236], Loss: 1.9611, Perplexity: 7.1069
Epoch [2/3], Step [2767/3236], Loss: 2.0679, Perplexity: 7.9080
Epoch [2/3], Step [2768/3236], Loss: 2.0636, Perplexity: 7.8743
Epoch [2/3], Step [2769/3236], Loss: 2.3163, Perplexity: 10.1376
Epoch [2/3], Step [2770/3236], Loss: 1.9179, Perplexity: 6.8068
Epoch [2/3], Step [2771/3236], Loss: 2.0524, Perplexity: 7.7866
Epoch [2/3], Step [2772/3236], Loss: 2.1511, Perplexity: 8.5941
Epoch [2/3], Step [2773/3236], Loss: 2.4732, Perplexity: 11.8606
Epoch [2/3], Step [2774/3236], Loss: 2.6554, Perplexity: 14.2303
Epoch [2/3], Step [2775/3236], Loss: 2.0250, Perplexity: 7.5761
Epoch [2/3], Step [2776/3236], Loss: 1.9594, Perplexity: 7.0952
Epoch [2/3], Step [2777/3236], Loss: 2.1322, Perplexity: 8.4334
Epoch [2/3], Step [2778/3236], Loss: 2.0738, Perplexity: 7.9551
Epoch [2/3], Step [2779/3236], Loss: 2.0200, Perplexity: 7.5384
Epoch [2/3], Step [2780/3236], Loss: 2.1792, Perplexity: 8.8390
Epoch [2/3], Step [2781/3236], Loss: 2.4459, Perplexity: 11.5413
Epoch [2/3], Step [2782/3236], Loss: 2.1994, Perplexity: 9.0199
Epoch [2/3], Step [2783/3236], Loss: 2.1787, Perplexity: 8.8346
Epoch [2/3], Step [2784/3236], Loss: 2.0101, Perplexity: 7.4640
Epoch [2/3], Step [2785/3236], Loss: 2.3791, Perplexity: 10.7950
Epoch [2/3], Step [2786/3236], Loss: 1.9581, Perplexity: 7.0856
Epoch [2/3], Step [2787/3236], Loss: 2.0081, Perplexity: 7.4489
Epoch [2/3], Step [2788/3236], Loss: 1.8745, Perplexity: 6.5173
Epoch [2/3], Step [2789/3236], Loss: 2.1113, Perplexity: 8.2586
Epoch [2/3], Step [2790/3236], Loss: 2.0542, Perplexity: 7.8009
Epoch [2/3], Step [2791/3236], Loss: 1.9914, Perplexity: 7.3259
Epoch [2/3], Step [2792/3236], Loss: 2.3134, Perplexity: 10.1091
Epoch [2/3], Step [2793/3236], Loss: 1.9840, Perplexity: 7.2718
Epoch [2/3], Step [2794/3236], Loss: 2.8510, Perplexity: 17.3051
Epoch [2/3], Step [2795/3236], Loss: 2.0283, Perplexity: 7.6010
Epoch [2/3], Step [2796/3236], Loss: 1.9934, Perplexity: 7.3407
Epoch [2/3], Step [2797/3236], Loss: 2.1623, Perplexity: 8.6911
Epoch [2/3], Step [2798/3236], Loss: 2.2016, Perplexity: 9.0392
Epoch [2/3], Step [2799/3236], Loss: 2.2862, Perplexity: 9.8377
Epoch [2/3], Step [2800/3236], Loss: 2.0544, Perplexity: 7.8023
Epoch [2/3], Step [2801/3236], Loss: 2.1629, Perplexity: 8.6965
Epoch [2/3], Step [2802/3236], Loss: 1.9668, Perplexity: 7.1477
Epoch [2/3], Step [2803/3236], Loss: 2.2340, Perplexity: 9.3367
Epoch [2/3], Step [2804/3236], Loss: 2.0703, Perplexity: 7.9269
Epoch [2/3], Step [2805/3236], Loss: 1.9714, Perplexity: 7.1805
Epoch [2/3], Step [2806/3236], Loss: 2.1225, Perplexity: 8.3517
Epoch [2/3], Step [2807/3236], Loss: 2.0441, Perplexity: 7.7223
Epoch [2/3], Step [2808/3236], Loss: 2.1440, Perplexity: 8.5339
Epoch [2/3], Step [2809/3236], Loss: 1.9612, Perplexity: 7.1079
Epoch [2/3], Step [2810/3236], Loss: 1.9144, Perplexity: 6.7831
Epoch [2/3], Step [2811/3236], Loss: 2.8969, Perplexity: 18.1184
Epoch [2/3], Step [2812/3236], Loss: 2.0252, Perplexity: 7.5775
Epoch [2/3], Step [2813/3236], Loss: 2.6333, Perplexity: 13.9196
Epoch [2/3], Step [2814/3236], Loss: 2.0295, Perplexity: 7.6107
Epoch [2/3], Step [2815/3236], Loss: 1.9344, Perplexity: 6.9196
Epoch [2/3], Step [2816/3236], Loss: 2.4163, Perplexity: 11.2044
Epoch [2/3], Step [2817/3236], Loss: 2.1038, Perplexity: 8.1972
Epoch [2/3], Step [2818/3236], Loss: 2.2496, Perplexity: 9.4835
Epoch [2/3], Step [2819/3236], Loss: 1.9707, Perplexity: 7.1755
Epoch [2/3], Step [2820/3236], Loss: 2.0320, Perplexity: 7.6295
Epoch [2/3], Step [2821/3236], Loss: 2.0288, Perplexity: 7.6046
Epoch [2/3], Step [2822/3236], Loss: 2.2088, Perplexity: 9.1051
Epoch [2/3], Step [2823/3236], Loss: 2.1854, Perplexity: 8.8939
Epoch [2/3], Step [2824/3236], Loss: 1.9970, Perplexity: 7.3666
Epoch [2/3], Step [2825/3236], Loss: 2.0769, Perplexity: 7.9796
Epoch [2/3], Step [2826/3236], Loss: 1.9895, Perplexity: 7.3122
Epoch [2/3], Step [2827/3236], Loss: 2.9128, Perplexity: 18.4078
Epoch [2/3], Step [2828/3236], Loss: 2.0404, Perplexity: 7.6933
Epoch [2/3], Step [2829/3236], Loss: 2.1009, Perplexity: 8.1735
Epoch [2/3], Step [2830/3236], Loss: 2.1874, Perplexity: 8.9122
Epoch [2/3], Step [2831/3236], Loss: 2.4385, Perplexity: 11.4561
Epoch [2/3], Step [2832/3236], Loss: 1.9912, Perplexity: 7.3241
Epoch [2/3], Step [2833/3236], Loss: 2.0331, Perplexity: 7.6379
Epoch [2/3], Step [2834/3236], Loss: 2.4159, Perplexity: 11.2001
Epoch [2/3], Step [2835/3236], Loss: 2.1798, Perplexity: 8.8444
Epoch [2/3], Step [2836/3236], Loss: 2.0990, Perplexity: 8.1582
Epoch [2/3], Step [2837/3236], Loss: 2.3367, Perplexity: 10.3474
Epoch [2/3], Step [2838/3236], Loss: 2.9508, Perplexity: 19.1221
Epoch [2/3], Step [2839/3236], Loss: 2.3225, Perplexity: 10.2013
Epoch [2/3], Step [2840/3236], Loss: 2.0560, Perplexity: 7.8143
Epoch [2/3], Step [2841/3236], Loss: 1.9601, Perplexity: 7.0999
Epoch [2/3], Step [2842/3236], Loss: 2.1387, Perplexity: 8.4886
Epoch [2/3], Step [2843/3236], Loss: 1.9997, Perplexity: 7.3870
Epoch [2/3], Step [2844/3236], Loss: 2.5422, Perplexity: 12.7070
Epoch [2/3], Step [2845/3236], Loss: 2.0603, Perplexity: 7.8487
Epoch [2/3], Step [2846/3236], Loss: 2.2740, Perplexity: 9.7185
Epoch [2/3], Step [2847/3236], Loss: 2.0508, Perplexity: 7.7742
Epoch [2/3], Step [2848/3236], Loss: 2.0013, Perplexity: 7.3987
Epoch [2/3], Step [2849/3236], Loss: 1.9767, Perplexity: 7.2187
Epoch [2/3], Step [2850/3236], Loss: 2.1731, Perplexity: 8.7853
Epoch [2/3], Step [2851/3236], Loss: 2.2504, Perplexity: 9.4911
Epoch [2/3], Step [2852/3236], Loss: 1.9872, Perplexity: 7.2951
Epoch [2/3], Step [2853/3236], Loss: 2.0972, Perplexity: 8.1435
Epoch [2/3], Step [2854/3236], Loss: 2.1506, Perplexity: 8.5900
Epoch [2/3], Step [2855/3236], Loss: 2.1210, Perplexity: 8.3399
Epoch [2/3], Step [2856/3236], Loss: 2.1208, Perplexity: 8.3380
Epoch [2/3], Step [2857/3236], Loss: 2.0173, Perplexity: 7.5180
Epoch [2/3], Step [2858/3236], Loss: 2.0655, Perplexity: 7.8892
Epoch [2/3], Step [2859/3236], Loss: 2.0149, Perplexity: 7.4997
Epoch [2/3], Step [2860/3236], Loss: 2.1325, Perplexity: 8.4358
Epoch [2/3], Step [2861/3236], Loss: 2.2762, Perplexity: 9.7399
Epoch [2/3], Step [2862/3236], Loss: 2.1064, Perplexity: 8.2189
Epoch [2/3], Step [2863/3236], Loss: 2.4635, Perplexity: 11.7454
Epoch [2/3], Step [2864/3236], Loss: 2.4806, Perplexity: 11.9490
Epoch [2/3], Step [2865/3236], Loss: 2.0098, Perplexity: 7.4619
Epoch [2/3], Step [2866/3236], Loss: 2.0331, Perplexity: 7.6378
Epoch [2/3], Step [2867/3236], Loss: 1.9144, Perplexity: 6.7832
Epoch [2/3], Step [2868/3236], Loss: 1.8304, Perplexity: 6.2365
Epoch [2/3], Step [2869/3236], Loss: 2.4975, Perplexity: 12.1526
Epoch [2/3], Step [2870/3236], Loss: 2.0529, Perplexity: 7.7905
Epoch [2/3], Step [2871/3236], Loss: 2.0704, Perplexity: 7.9277
Epoch [2/3], Step [2872/3236], Loss: 2.0788, Perplexity: 7.9952
Epoch [2/3], Step [2873/3236], Loss: 2.0209, Perplexity: 7.5452
Epoch [2/3], Step [2874/3236], Loss: 2.0138, Perplexity: 7.4921
Epoch [2/3], Step [2875/3236], Loss: 3.1482, Perplexity: 23.2932
Epoch [2/3], Step [2876/3236], Loss: 2.1410, Perplexity: 8.5081
Epoch [2/3], Step [2877/3236], Loss: 2.0720, Perplexity: 7.9409
Epoch [2/3], Step [2878/3236], Loss: 2.2215, Perplexity: 9.2213
Epoch [2/3], Step [2879/3236], Loss: 2.0690, Perplexity: 7.9170
Epoch [2/3], Step [2880/3236], Loss: 2.1089, Perplexity: 8.2395
Epoch [2/3], Step [2881/3236], Loss: 2.0930, Perplexity: 8.1091
Epoch [2/3], Step [2882/3236], Loss: 2.1749, Perplexity: 8.8016
Epoch [2/3], Step [2883/3236], Loss: 2.1364, Perplexity: 8.4691
Epoch [2/3], Step [2884/3236], Loss: 1.9724, Perplexity: 7.1876
Epoch [2/3], Step [2885/3236], Loss: 2.1035, Perplexity: 8.1947
Epoch [2/3], Step [2886/3236], Loss: 2.2028, Perplexity: 9.0506
Epoch [2/3], Step [2887/3236], Loss: 2.1213, Perplexity: 8.3418
Epoch [2/3], Step [2888/3236], Loss: 2.1344, Perplexity: 8.4520
Epoch [2/3], Step [2889/3236], Loss: 2.3412, Perplexity: 10.3939
Epoch [2/3], Step [2890/3236], Loss: 1.9463, Perplexity: 7.0027
Epoch [2/3], Step [2891/3236], Loss: 2.0733, Perplexity: 7.9513
Epoch [2/3], Step [2892/3236], Loss: 2.6205, Perplexity: 13.7427
Epoch [2/3], Step [2893/3236], Loss: 2.1585, Perplexity: 8.6586
Epoch [2/3], Step [2894/3236], Loss: 2.5966, Perplexity: 13.4177
Epoch [2/3], Step [2895/3236], Loss: 2.4912, Perplexity: 12.0756
Epoch [2/3], Step [2896/3236], Loss: 2.0570, Perplexity: 7.8227
Epoch [2/3], Step [2897/3236], Loss: 2.2485, Perplexity: 9.4739
Epoch [2/3], Step [2898/3236], Loss: 2.2407, Perplexity: 9.4001
Epoch [2/3], Step [2899/3236], Loss: 2.0556, Perplexity: 7.8114
Epoch [2/3], Step [2900/3236], Loss: 1.9703, Perplexity: 7.1730
Epoch [2/3], Step [2901/3236], Loss: 1.9932, Perplexity: 7.3389
Epoch [2/3], Step [2902/3236], Loss: 2.4575, Perplexity: 11.6758
Epoch [2/3], Step [2903/3236], Loss: 1.9761, Perplexity: 7.2142
Epoch [2/3], Step [2904/3236], Loss: 2.0188, Perplexity: 7.5292
Epoch [2/3], Step [2905/3236], Loss: 2.0759, Perplexity: 7.9720
Epoch [2/3], Step [2906/3236], Loss: 2.0818, Perplexity: 8.0186
Epoch [2/3], Step [2907/3236], Loss: 2.1231, Perplexity: 8.3572
Epoch [2/3], Step [2908/3236], Loss: 2.9411, Perplexity: 18.9373
Epoch [2/3], Step [2909/3236], Loss: 2.0694, Perplexity: 7.9202
Epoch [2/3], Step [2910/3236], Loss: 2.0521, Perplexity: 7.7839
Epoch [2/3], Step [2911/3236], Loss: 2.3499, Perplexity: 10.4843
Epoch [2/3], Step [2912/3236], Loss: 2.0698, Perplexity: 7.9236
Epoch [2/3], Step [2913/3236], Loss: 2.4760, Perplexity: 11.8939
Epoch [2/3], Step [2914/3236], Loss: 2.3231, Perplexity: 10.2074
Epoch [2/3], Step [2915/3236], Loss: 1.9847, Perplexity: 7.2771
Epoch [2/3], Step [2916/3236], Loss: 2.1145, Perplexity: 8.2851
Epoch [2/3], Step [2917/3236], Loss: 2.1081, Perplexity: 8.2328
Epoch [2/3], Step [2918/3236], Loss: 1.9828, Perplexity: 7.2633
Epoch [2/3], Step [2919/3236], Loss: 2.0729, Perplexity: 7.9479
Epoch [2/3], Step [2920/3236], Loss: 2.1265, Perplexity: 8.3855
Epoch [2/3], Step [2921/3236], Loss: 2.2399, Perplexity: 9.3925
Epoch [2/3], Step [2922/3236], Loss: 2.0930, Perplexity: 8.1091
Epoch [2/3], Step [2923/3236], Loss: 2.0475, Perplexity: 7.7488
Epoch [2/3], Step [2924/3236], Loss: 2.1149, Perplexity: 8.2884
Epoch [2/3], Step [2925/3236], Loss: 2.3535, Perplexity: 10.5223
Epoch [2/3], Step [2926/3236], Loss: 2.3351, Perplexity: 10.3308
Epoch [2/3], Step [2927/3236], Loss: 1.9596, Perplexity: 7.0963
Epoch [2/3], Step [2928/3236], Loss: 2.1233, Perplexity: 8.3589
Epoch [2/3], Step [2929/3236], Loss: 2.0416, Perplexity: 7.7029
Epoch [2/3], Step [2930/3236], Loss: 2.1926, Perplexity: 8.9582
Epoch [2/3], Step [2931/3236], Loss: 2.2032, Perplexity: 9.0536
Epoch [2/3], Step [2932/3236], Loss: 2.0452, Perplexity: 7.7306
Epoch [2/3], Step [2933/3236], Loss: 2.2963, Perplexity: 9.9370
Epoch [2/3], Step [2934/3236], Loss: 2.0864, Perplexity: 8.0556
Epoch [2/3], Step [2935/3236], Loss: 2.1471, Perplexity: 8.5603
Epoch [2/3], Step [2936/3236], Loss: 2.0380, Perplexity: 7.6752
Epoch [2/3], Step [2937/3236], Loss: 2.0881, Perplexity: 8.0699
Epoch [2/3], Step [2938/3236], Loss: 2.1999, Perplexity: 9.0241
Epoch [2/3], Step [2939/3236], Loss: 2.0848, Perplexity: 8.0430
Epoch [2/3], Step [2940/3236], Loss: 2.0905, Perplexity: 8.0890
Epoch [2/3], Step [2941/3236], Loss: 2.1518, Perplexity: 8.5999
Epoch [2/3], Step [2942/3236], Loss: 2.3280, Perplexity: 10.2576
Epoch [2/3], Step [2943/3236], Loss: 2.0080, Perplexity: 7.4485
Epoch [2/3], Step [2944/3236], Loss: 2.1739, Perplexity: 8.7925
Epoch [2/3], Step [2945/3236], Loss: 2.0942, Perplexity: 8.1191
Epoch [2/3], Step [2946/3236], Loss: 1.9180, Perplexity: 6.8071
Epoch [2/3], Step [2947/3236], Loss: 2.0595, Perplexity: 7.8423
Epoch [2/3], Step [2948/3236], Loss: 2.0641, Perplexity: 7.8781
Epoch [2/3], Step [2949/3236], Loss: 2.1858, Perplexity: 8.8976
Epoch [2/3], Step [2950/3236], Loss: 2.1185, Perplexity: 8.3184
Epoch [2/3], Step [2951/3236], Loss: 2.1126, Perplexity: 8.2699
Epoch [2/3], Step [2952/3236], Loss: 2.1005, Perplexity: 8.1702
Epoch [2/3], Step [2953/3236], Loss: 2.2961, Perplexity: 9.9357
Epoch [2/3], Step [2954/3236], Loss: 1.9562, Perplexity: 7.0723
Epoch [2/3], Step [2955/3236], Loss: 2.2991, Perplexity: 9.9648
Epoch [2/3], Step [2956/3236], Loss: 2.1734, Perplexity: 8.7879
Epoch [2/3], Step [2957/3236], Loss: 1.9961, Perplexity: 7.3603
Epoch [2/3], Step [2958/3236], Loss: 2.0716, Perplexity: 7.9372
Epoch [2/3], Step [2959/3236], Loss: 2.0038, Perplexity: 7.4170
Epoch [2/3], Step [2960/3236], Loss: 1.9715, Perplexity: 7.1813
Epoch [2/3], Step [2961/3236], Loss: 2.1596, Perplexity: 8.6679
Epoch [2/3], Step [2962/3236], Loss: 2.1819, Perplexity: 8.8632
Epoch [2/3], Step [2963/3236], Loss: 1.9483, Perplexity: 7.0169
Epoch [2/3], Step [2964/3236], Loss: 1.9708, Perplexity: 7.1763
Epoch [2/3], Step [2965/3236], Loss: 2.0025, Perplexity: 7.4074
Epoch [2/3], Step [2966/3236], Loss: 2.4261, Perplexity: 11.3148
Epoch [2/3], Step [2967/3236], Loss: 2.0036, Perplexity: 7.4157
Epoch [2/3], Step [2968/3236], Loss: 2.2493, Perplexity: 9.4814
Epoch [2/3], Step [2969/3236], Loss: 2.1354, Perplexity: 8.4607
Epoch [2/3], Step [2970/3236], Loss: 2.0052, Perplexity: 7.4275
Epoch [2/3], Step [2971/3236], Loss: 2.1111, Perplexity: 8.2574
Epoch [2/3], Step [2972/3236], Loss: 2.3272, Perplexity: 10.2488
Epoch [2/3], Step [2973/3236], Loss: 1.9665, Perplexity: 7.1458
Epoch [2/3], Step [2974/3236], Loss: 2.1573, Perplexity: 8.6478
Epoch [2/3], Step [2975/3236], Loss: 2.0218, Perplexity: 7.5517
Epoch [2/3], Step [2976/3236], Loss: 2.0259, Perplexity: 7.5830
Epoch [2/3], Step [2977/3236], Loss: 2.2414, Perplexity: 9.4069
Epoch [2/3], Step [2978/3236], Loss: 3.8314, Perplexity: 46.1281
Epoch [2/3], Step [2979/3236], Loss: 2.0732, Perplexity: 7.9502
Epoch [2/3], Step [2980/3236], Loss: 1.9608, Perplexity: 7.1048
Epoch [2/3], Step [2981/3236], Loss: 2.1292, Perplexity: 8.4083
Epoch [2/3], Step [2982/3236], Loss: 2.0990, Perplexity: 8.1584
Epoch [2/3], Step [2983/3236], Loss: 2.0704, Perplexity: 7.9280
Epoch [2/3], Step [2984/3236], Loss: 2.0005, Perplexity: 7.3926
Epoch [2/3], Step [2985/3236], Loss: 2.1844, Perplexity: 8.8855
Epoch [2/3], Step [2986/3236], Loss: 2.1286, Perplexity: 8.4031
Epoch [2/3], Step [2987/3236], Loss: 3.0602, Perplexity: 21.3312
Epoch [2/3], Step [2988/3236], Loss: 2.1854, Perplexity: 8.8942
Epoch [2/3], Step [2989/3236], Loss: 2.0866, Perplexity: 8.0578
Epoch [2/3], Step [2990/3236], Loss: 2.1912, Perplexity: 8.9463
Epoch [2/3], Step [2991/3236], Loss: 2.1667, Perplexity: 8.7293
Epoch [2/3], Step [2992/3236], Loss: 2.0476, Perplexity: 7.7494
Epoch [2/3], Step [2993/3236], Loss: 2.0298, Perplexity: 7.6124
Epoch [2/3], Step [2994/3236], Loss: 2.0607, Perplexity: 7.8517
Epoch [2/3], Step [2995/3236], Loss: 1.9807, Perplexity: 7.2476
Epoch [2/3], Step [2996/3236], Loss: 2.0238, Perplexity: 7.5671
Epoch [2/3], Step [2997/3236], Loss: 2.2271, Perplexity: 9.2729
Epoch [2/3], Step [2998/3236], Loss: 1.9694, Perplexity: 7.1664
Epoch [2/3], Step [2999/3236], Loss: 2.0676, Perplexity: 7.9055
Epoch [2/3], Step [3000/3236], Loss: 2.1532, Perplexity: 8.6121
Epoch [2/3], Step [3001/3236], Loss: 2.3744, Perplexity: 10.7449
Epoch [2/3], Step [3002/3236], Loss: 2.1132, Perplexity: 8.2751
Epoch [2/3], Step [3003/3236], Loss: 2.0954, Perplexity: 8.1290
Epoch [2/3], Step [3004/3236], Loss: 2.8482, Perplexity: 17.2563
Epoch [2/3], Step [3005/3236], Loss: 1.9708, Perplexity: 7.1762
Epoch [2/3], Step [3006/3236], Loss: 2.3259, Perplexity: 10.2360
Epoch [2/3], Step [3007/3236], Loss: 2.1491, Perplexity: 8.5773
Epoch [2/3], Step [3008/3236], Loss: 2.0787, Perplexity: 7.9937
Epoch [2/3], Step [3009/3236], Loss: 1.9866, Perplexity: 7.2908
Epoch [2/3], Step [3010/3236], Loss: 2.3380, Perplexity: 10.3601
Epoch [2/3], Step [3011/3236], Loss: 2.0539, Perplexity: 7.7980
Epoch [2/3], Step [3012/3236], Loss: 2.0702, Perplexity: 7.9266
Epoch [2/3], Step [3013/3236], Loss: 1.9892, Perplexity: 7.3097
Epoch [2/3], Step [3014/3236], Loss: 2.1249, Perplexity: 8.3719
Epoch [2/3], Step [3015/3236], Loss: 2.0034, Perplexity: 7.4139
Epoch [2/3], Step [3016/3236], Loss: 2.1498, Perplexity: 8.5832
Epoch [2/3], Step [3017/3236], Loss: 2.1477, Perplexity: 8.5654
Epoch [2/3], Step [3018/3236], Loss: 2.1224, Perplexity: 8.3508
Epoch [2/3], Step [3019/3236], Loss: 2.0016, Perplexity: 7.4011
Epoch [2/3], Step [3020/3236], Loss: 2.4975, Perplexity: 12.1526
Epoch [2/3], Step [3021/3236], Loss: 2.2468, Perplexity: 9.4575
Epoch [2/3], Step [3022/3236], Loss: 2.1987, Perplexity: 9.0130
Epoch [2/3], Step [3023/3236], Loss: 2.0401, Perplexity: 7.6916
Epoch [2/3], Step [3024/3236], Loss: 2.0690, Perplexity: 7.9167
Epoch [2/3], Step [3025/3236], Loss: 1.9482, Perplexity: 7.0158
Epoch [2/3], Step [3026/3236], Loss: 2.5217, Perplexity: 12.4496
Epoch [2/3], Step [3027/3236], Loss: 1.9745, Perplexity: 7.2028
Epoch [2/3], Step [3028/3236], Loss: 2.1244, Perplexity: 8.3681
Epoch [2/3], Step [3029/3236], Loss: 1.9560, Perplexity: 7.0710
Epoch [2/3], Step [3030/3236], Loss: 1.9892, Perplexity: 7.3099
Epoch [2/3], Step [3031/3236], Loss: 2.0921, Perplexity: 8.1022
Epoch [2/3], Step [3032/3236], Loss: 2.1810, Perplexity: 8.8548
Epoch [2/3], Step [3033/3236], Loss: 2.1027, Perplexity: 8.1881
Epoch [2/3], Step [3034/3236], Loss: 2.1590, Perplexity: 8.6625
Epoch [2/3], Step [3035/3236], Loss: 2.1318, Perplexity: 8.4303
Epoch [2/3], Step [3036/3236], Loss: 2.2334, Perplexity: 9.3317
Epoch [2/3], Step [3037/3236], Loss: 2.0756, Perplexity: 7.9690
Epoch [2/3], Step [3038/3236], Loss: 2.1456, Perplexity: 8.5469
Epoch [2/3], Step [3039/3236], Loss: 2.0229, Perplexity: 7.5599
Epoch [2/3], Step [3040/3236], Loss: 2.1419, Perplexity: 8.5157
Epoch [2/3], Step [3041/3236], Loss: 2.2196, Perplexity: 9.2037
Epoch [2/3], Step [3042/3236], Loss: 2.0116, Perplexity: 7.4750
Epoch [2/3], Step [3043/3236], Loss: 2.0510, Perplexity: 7.7760
Epoch [2/3], Step [3044/3236], Loss: 2.1999, Perplexity: 9.0239
Epoch [2/3], Step [3045/3236], Loss: 2.0668, Perplexity: 7.8999
Epoch [2/3], Step [3046/3236], Loss: 2.1460, Perplexity: 8.5508
Epoch [2/3], Step [3047/3236], Loss: 2.2074, Perplexity: 9.0922
Epoch [2/3], Step [3048/3236], Loss: 2.1168, Perplexity: 8.3045
Epoch [2/3], Step [3049/3236], Loss: 1.9188, Perplexity: 6.8129
Epoch [2/3], Step [3050/3236], Loss: 2.1028, Perplexity: 8.1895
Epoch [2/3], Step [3051/3236], Loss: 2.2851, Perplexity: 9.8262
Epoch [2/3], Step [3052/3236], Loss: 2.0557, Perplexity: 7.8123
Epoch [2/3], Step [3053/3236], Loss: 2.1617, Perplexity: 8.6859
Epoch [2/3], Step [3054/3236], Loss: 2.0467, Perplexity: 7.7423
Epoch [2/3], Step [3055/3236], Loss: 2.6084, Perplexity: 13.5775
Epoch [2/3], Step [3056/3236], Loss: 2.0250, Perplexity: 7.5764
Epoch [2/3], Step [3057/3236], Loss: 2.0158, Perplexity: 7.5070
Epoch [2/3], Step [3058/3236], Loss: 2.1167, Perplexity: 8.3038
Epoch [2/3], Step [3059/3236], Loss: 2.0616, Perplexity: 7.8586
Epoch [2/3], Step [3060/3236], Loss: 1.9827, Perplexity: 7.2624
Epoch [2/3], Step [3061/3236], Loss: 2.1065, Perplexity: 8.2191
Epoch [2/3], Step [3062/3236], Loss: 2.1762, Perplexity: 8.8128
Epoch [2/3], Step [3063/3236], Loss: 2.0144, Perplexity: 7.4959
Epoch [2/3], Step [3064/3236], Loss: 1.9844, Perplexity: 7.2743
Epoch [2/3], Step [3065/3236], Loss: 2.0146, Perplexity: 7.4974
Epoch [2/3], Step [3066/3236], Loss: 2.1038, Perplexity: 8.1972
Epoch [2/3], Step [3067/3236], Loss: 2.0201, Perplexity: 7.5389
Epoch [2/3], Step [3068/3236], Loss: 2.0845, Perplexity: 8.0402
Epoch [2/3], Step [3069/3236], Loss: 2.1396, Perplexity: 8.4961
Epoch [2/3], Step [3070/3236], Loss: 2.1090, Perplexity: 8.2403
Epoch [2/3], Step [3071/3236], Loss: 2.5014, Perplexity: 12.1991
Epoch [2/3], Step [3072/3236], Loss: 1.8621, Perplexity: 6.4370
Epoch [2/3], Step [3073/3236], Loss: 1.9996, Perplexity: 7.3861
Epoch [2/3], Step [3074/3236], Loss: 1.9720, Perplexity: 7.1853
Epoch [2/3], Step [3075/3236], Loss: 2.3270, Perplexity: 10.2473
Epoch [2/3], Step [3076/3236], Loss: 2.3773, Perplexity: 10.7760
Epoch [2/3], Step [3077/3236], Loss: 2.0184, Perplexity: 7.5262
Epoch [2/3], Step [3078/3236], Loss: 2.0104, Perplexity: 7.4665
Epoch [2/3], Step [3079/3236], Loss: 2.0443, Perplexity: 7.7235
Epoch [2/3], Step [3080/3236], Loss: 2.1135, Perplexity: 8.2775
Epoch [2/3], Step [3081/3236], Loss: 2.0522, Perplexity: 7.7852
Epoch [2/3], Step [3082/3236], Loss: 2.1379, Perplexity: 8.4819
Epoch [2/3], Step [3083/3236], Loss: 2.1537, Perplexity: 8.6169
Epoch [2/3], Step [3084/3236], Loss: 2.1855, Perplexity: 8.8950
Epoch [2/3], Step [3085/3236], Loss: 2.1316, Perplexity: 8.4284
Epoch [2/3], Step [3086/3236], Loss: 2.1580, Perplexity: 8.6536
Epoch [2/3], Step [3087/3236], Loss: 2.2274, Perplexity: 9.2759
Epoch [2/3], Step [3088/3236], Loss: 2.1175, Perplexity: 8.3106
Epoch [2/3], Step [3089/3236], Loss: 2.0935, Perplexity: 8.1130
Epoch [2/3], Step [3090/3236], Loss: 2.0691, Perplexity: 7.9179
Epoch [2/3], Step [3091/3236], Loss: 2.2713, Perplexity: 9.6924
Epoch [2/3], Step [3092/3236], Loss: 2.0170, Perplexity: 7.5159
Epoch [2/3], Step [3093/3236], Loss: 2.4327, Perplexity: 11.3899
Epoch [2/3], Step [3094/3236], Loss: 2.3569, Perplexity: 10.5578
Epoch [2/3], Step [3095/3236], Loss: 2.1217, Perplexity: 8.3453
Epoch [2/3], Step [3096/3236], Loss: 2.0535, Perplexity: 7.7952
Epoch [2/3], Step [3097/3236], Loss: 2.1071, Perplexity: 8.2245
Epoch [2/3], Step [3098/3236], Loss: 2.2051, Perplexity: 9.0712
Epoch [2/3], Step [3099/3236], Loss: 2.0529, Perplexity: 7.7905
Epoch [2/3], Step [3100/3236], Loss: 2.3302, Perplexity: 10.2803
Epoch [2/3], Step [3101/3236], Loss: 2.1135, Perplexity: 8.2776
Epoch [2/3], Step [3102/3236], Loss: 1.9454, Perplexity: 6.9967
Epoch [2/3], Step [3103/3236], Loss: 2.0936, Perplexity: 8.1141
Epoch [2/3], Step [3104/3236], Loss: 2.0546, Perplexity: 7.8037
Epoch [2/3], Step [3105/3236], Loss: 2.1011, Perplexity: 8.1755
Epoch [2/3], Step [3106/3236], Loss: 2.1258, Perplexity: 8.3794
Epoch [2/3], Step [3107/3236], Loss: 2.0601, Perplexity: 7.8470
Epoch [2/3], Step [3108/3236], Loss: 2.0078, Perplexity: 7.4469
Epoch [2/3], Step [3109/3236], Loss: 2.2089, Perplexity: 9.1060
Epoch [2/3], Step [3110/3236], Loss: 2.0708, Perplexity: 7.9309
Epoch [2/3], Step [3111/3236], Loss: 2.1229, Perplexity: 8.3555
Epoch [2/3], Step [3112/3236], Loss: 2.1772, Perplexity: 8.8218
Epoch [2/3], Step [3113/3236], Loss: 1.9613, Perplexity: 7.1085
Epoch [2/3], Step [3114/3236], Loss: 2.1203, Perplexity: 8.3334
Epoch [2/3], Step [3115/3236], Loss: 2.1336, Perplexity: 8.4450
Epoch [2/3], Step [3116/3236], Loss: 2.1284, Perplexity: 8.4015
Epoch [2/3], Step [3117/3236], Loss: 1.9369, Perplexity: 6.9375
Epoch [2/3], Step [3118/3236], Loss: 2.1401, Perplexity: 8.5004
Epoch [2/3], Step [3119/3236], Loss: 1.9712, Perplexity: 7.1791
Epoch [2/3], Step [3120/3236], Loss: 2.1983, Perplexity: 9.0094
Epoch [2/3], Step [3121/3236], Loss: 2.1863, Perplexity: 8.9021
Epoch [2/3], Step [3122/3236], Loss: 2.5244, Perplexity: 12.4840
Epoch [2/3], Step [3123/3236], Loss: 2.0796, Perplexity: 8.0014
Epoch [2/3], Step [3124/3236], Loss: 1.9495, Perplexity: 7.0252
Epoch [2/3], Step [3125/3236], Loss: 2.0423, Perplexity: 7.7083
Epoch [2/3], Step [3126/3236], Loss: 2.0611, Perplexity: 7.8542
Epoch [2/3], Step [3127/3236], Loss: 2.0766, Perplexity: 7.9770
Epoch [2/3], Step [3128/3236], Loss: 2.3225, Perplexity: 10.2009
Epoch [2/3], Step [3129/3236], Loss: 2.3802, Perplexity: 10.8068
Epoch [2/3], Step [3130/3236], Loss: 2.3540, Perplexity: 10.5272
Epoch [2/3], Step [3131/3236], Loss: 2.2591, Perplexity: 9.5743
Epoch [2/3], Step [3132/3236], Loss: 2.0024, Perplexity: 7.4066
Epoch [2/3], Step [3133/3236], Loss: 2.0245, Perplexity: 7.5723
Epoch [2/3], Step [3134/3236], Loss: 2.2070, Perplexity: 9.0886
Epoch [2/3], Step [3135/3236], Loss: 2.3986, Perplexity: 11.0075
Epoch [2/3], Step [3136/3236], Loss: 2.0477, Perplexity: 7.7500
Epoch [2/3], Step [3137/3236], Loss: 2.5584, Perplexity: 12.9152
Epoch [2/3], Step [3138/3236], Loss: 2.0000, Perplexity: 7.3892
Epoch [2/3], Step [3139/3236], Loss: 2.1090, Perplexity: 8.2398
Epoch [2/3], Step [3140/3236], Loss: 2.0166, Perplexity: 7.5128
Epoch [2/3], Step [3141/3236], Loss: 1.9991, Perplexity: 7.3823
Epoch [2/3], Step [3142/3236], Loss: 2.1860, Perplexity: 8.8991
Epoch [2/3], Step [3143/3236], Loss: 2.2066, Perplexity: 9.0849
Epoch [2/3], Step [3144/3236], Loss: 2.1482, Perplexity: 8.5691
Epoch [2/3], Step [3145/3236], Loss: 2.4697, Perplexity: 11.8188
Epoch [2/3], Step [3146/3236], Loss: 2.1228, Perplexity: 8.3541
Epoch [2/3], Step [3147/3236], Loss: 2.2864, Perplexity: 9.8393
Epoch [2/3], Step [3148/3236], Loss: 1.9422, Perplexity: 6.9742
Epoch [2/3], Step [3149/3236], Loss: 2.0416, Perplexity: 7.7027
Epoch [2/3], Step [3150/3236], Loss: 2.0842, Perplexity: 8.0385
Epoch [2/3], Step [3151/3236], Loss: 2.0770, Perplexity: 7.9802
Epoch [2/3], Step [3152/3236], Loss: 1.9329, Perplexity: 6.9093
Epoch [2/3], Step [3153/3236], Loss: 2.0456, Perplexity: 7.7339
Epoch [2/3], Step [3154/3236], Loss: 2.0592, Perplexity: 7.8399
Epoch [2/3], Step [3155/3236], Loss: 2.0769, Perplexity: 7.9799
Epoch [2/3], Step [3156/3236], Loss: 1.9636, Perplexity: 7.1246
Epoch [2/3], Step [3157/3236], Loss: 2.0537, Perplexity: 7.7963
Epoch [2/3], Step [3158/3236], Loss: 2.0810, Perplexity: 8.0128
Epoch [2/3], Step [3159/3236], Loss: 1.9842, Perplexity: 7.2734
Epoch [2/3], Step [3160/3236], Loss: 2.0231, Perplexity: 7.5618
Epoch [2/3], Step [3161/3236], Loss: 2.1415, Perplexity: 8.5121
Epoch [2/3], Step [3162/3236], Loss: 2.3734, Perplexity: 10.7339
Epoch [2/3], Step [3163/3236], Loss: 2.7632, Perplexity: 15.8504
Epoch [2/3], Step [3164/3236], Loss: 2.0159, Perplexity: 7.5073
Epoch [2/3], Step [3165/3236], Loss: 1.9948, Perplexity: 7.3506
Epoch [2/3], Step [3166/3236], Loss: 1.9847, Perplexity: 7.2770
Epoch [2/3], Step [3167/3236], Loss: 2.0562, Perplexity: 7.8161
Epoch [2/3], Step [3168/3236], Loss: 2.1187, Perplexity: 8.3207
Epoch [2/3], Step [3169/3236], Loss: 2.5122, Perplexity: 12.3318
Epoch [2/3], Step [3170/3236], Loss: 1.9905, Perplexity: 7.3192
Epoch [2/3], Step [3171/3236], Loss: 2.1373, Perplexity: 8.4765
Epoch [2/3], Step [3172/3236], Loss: 2.1598, Perplexity: 8.6696
Epoch [2/3], Step [3173/3236], Loss: 2.0661, Perplexity: 7.8937
Epoch [2/3], Step [3174/3236], Loss: 2.1541, Perplexity: 8.6199
Epoch [2/3], Step [3175/3236], Loss: 2.0138, Perplexity: 7.4916
Epoch [2/3], Step [3176/3236], Loss: 2.3414, Perplexity: 10.3953
Epoch [2/3], Step [3177/3236], Loss: 1.9221, Perplexity: 6.8356
Epoch [2/3], Step [3178/3236], Loss: 2.3513, Perplexity: 10.4995
Epoch [2/3], Step [3179/3236], Loss: 2.8493, Perplexity: 17.2760
Epoch [2/3], Step [3180/3236], Loss: 2.0649, Perplexity: 7.8844
Epoch [2/3], Step [3181/3236], Loss: 2.0242, Perplexity: 7.5704
Epoch [2/3], Step [3182/3236], Loss: 1.9961, Perplexity: 7.3602
Epoch [2/3], Step [3183/3236], Loss: 2.1786, Perplexity: 8.8338
Epoch [2/3], Step [3184/3236], Loss: 2.1202, Perplexity: 8.3331
Epoch [2/3], Step [3185/3236], Loss: 2.2659, Perplexity: 9.6395
Epoch [2/3], Step [3186/3236], Loss: 2.1034, Perplexity: 8.1936
Epoch [2/3], Step [3187/3236], Loss: 1.9505, Perplexity: 7.0321
Epoch [2/3], Step [3188/3236], Loss: 2.2424, Perplexity: 9.4157
Epoch [2/3], Step [3189/3236], Loss: 2.1072, Perplexity: 8.2253
Epoch [2/3], Step [3190/3236], Loss: 2.9055, Perplexity: 18.2740
Epoch [2/3], Step [3191/3236], Loss: 2.0330, Perplexity: 7.6370
Epoch [2/3], Step [3192/3236], Loss: 2.0469, Perplexity: 7.7440
Epoch [2/3], Step [3193/3236], Loss: 2.1653, Perplexity: 8.7170
Epoch [2/3], Step [3194/3236], Loss: 1.9532, Perplexity: 7.0511
Epoch [2/3], Step [3195/3236], Loss: 2.0471, Perplexity: 7.7453
Epoch [2/3], Step [3196/3236], Loss: 1.9518, Perplexity: 7.0413
Epoch [2/3], Step [3197/3236], Loss: 2.3376, Perplexity: 10.3568
Epoch [2/3], Step [3198/3236], Loss: 2.1020, Perplexity: 8.1827
Epoch [2/3], Step [3199/3236], Loss: 2.0966, Perplexity: 8.1388
Epoch [2/3], Step [3200/3236], Loss: 2.1242, Perplexity: 8.3662
Epoch [2/3], Step [3201/3236], Loss: 2.1141, Perplexity: 8.2824
Epoch [2/3], Step [3202/3236], Loss: 1.9858, Perplexity: 7.2852
Epoch [2/3], Step [3203/3236], Loss: 1.9710, Perplexity: 7.1781
Epoch [2/3], Step [3204/3236], Loss: 2.0174, Perplexity: 7.5190
Epoch [2/3], Step [3205/3236], Loss: 2.1680, Perplexity: 8.7411
Epoch [2/3], Step [3206/3236], Loss: 2.4028, Perplexity: 11.0541
Epoch [2/3], Step [3207/3236], Loss: 2.1630, Perplexity: 8.6974
Epoch [2/3], Step [3208/3236], Loss: 2.0314, Perplexity: 7.6245
Epoch [2/3], Step [3209/3236], Loss: 2.0582, Perplexity: 7.8320
Epoch [2/3], Step [3210/3236], Loss: 2.0502, Perplexity: 7.7691
Epoch [2/3], Step [3211/3236], Loss: 2.1382, Perplexity: 8.4840
Epoch [2/3], Step [3212/3236], Loss: 2.0783, Perplexity: 7.9909
Epoch [2/3], Step [3213/3236], Loss: 2.2422, Perplexity: 9.4139
Epoch [2/3], Step [3214/3236], Loss: 1.9599, Perplexity: 7.0985
Epoch [2/3], Step [3215/3236], Loss: 2.0467, Perplexity: 7.7423
Epoch [2/3], Step [3216/3236], Loss: 2.1539, Perplexity: 8.6183
Epoch [2/3], Step [3217/3236], Loss: 2.0972, Perplexity: 8.1429
Epoch [2/3], Step [3218/3236], Loss: 2.0601, Perplexity: 7.8467
Epoch [2/3], Step [3219/3236], Loss: 1.9859, Perplexity: 7.2855
Epoch [2/3], Step [3220/3236], Loss: 1.9457, Perplexity: 6.9985
Epoch [2/3], Step [3221/3236], Loss: 2.0655, Perplexity: 7.8895
Epoch [2/3], Step [3222/3236], Loss: 1.9041, Perplexity: 6.7137
Epoch [2/3], Step [3223/3236], Loss: 1.9312, Perplexity: 6.8979
Epoch [2/3], Step [3224/3236], Loss: 2.4164, Perplexity: 11.2057
Epoch [2/3], Step [3225/3236], Loss: 2.1056, Perplexity: 8.2124
Epoch [2/3], Step [3226/3236], Loss: 2.0652, Perplexity: 7.8870
Epoch [2/3], Step [3227/3236], Loss: 2.0854, Perplexity: 8.0482
Epoch [2/3], Step [3228/3236], Loss: 2.0110, Perplexity: 7.4711
Epoch [2/3], Step [3229/3236], Loss: 2.0425, Perplexity: 7.7098
Epoch [2/3], Step [3230/3236], Loss: 2.0943, Perplexity: 8.1196
Epoch [2/3], Step [3231/3236], Loss: 2.1633, Perplexity: 8.7001
Epoch [2/3], Step [3232/3236], Loss: 2.0309, Perplexity: 7.6211
Epoch [2/3], Step [3233/3236], Loss: 1.9133, Perplexity: 6.7757
Epoch [2/3], Step [3234/3236], Loss: 2.1955, Perplexity: 8.9841
Epoch [2/3], Step [3235/3236], Loss: 2.7626, Perplexity: 15.8407
Epoch [2/3], Step [3236/3236], Loss: 2.0134, Perplexity: 7.4885
Epoch [3/3], Step [1/3236], Loss: 1.9977, Perplexity: 7.3721
Epoch [3/3], Step [2/3236], Loss: 2.1061, Perplexity: 8.2160
Epoch [3/3], Step [3/3236], Loss: 2.0688, Perplexity: 7.9155
Epoch [3/3], Step [4/3236], Loss: 1.9464, Perplexity: 7.0032
Epoch [3/3], Step [5/3236], Loss: 2.4163, Perplexity: 11.2047
Epoch [3/3], Step [6/3236], Loss: 1.9837, Perplexity: 7.2696
Epoch [3/3], Step [7/3236], Loss: 1.9524, Perplexity: 7.0457
Epoch [3/3], Step [8/3236], Loss: 2.0357, Perplexity: 7.6577
Epoch [3/3], Step [9/3236], Loss: 2.9388, Perplexity: 18.8924
Epoch [3/3], Step [10/3236], Loss: 1.9215, Perplexity: 6.8309
Epoch [3/3], Step [11/3236], Loss: 2.0296, Perplexity: 7.6108
Epoch [3/3], Step [12/3236], Loss: 2.0987, Perplexity: 8.1552
Epoch [3/3], Step [13/3236], Loss: 2.0778, Perplexity: 7.9872
Epoch [3/3], Step [14/3236], Loss: 2.1839, Perplexity: 8.8811
Epoch [3/3], Step [15/3236], Loss: 2.0537, Perplexity: 7.7970
Epoch [3/3], Step [16/3236], Loss: 2.2914, Perplexity: 9.8890
Epoch [3/3], Step [17/3236], Loss: 2.3808, Perplexity: 10.8137
Epoch [3/3], Step [18/3236], Loss: 2.2615, Perplexity: 9.5976
Epoch [3/3], Step [19/3236], Loss: 2.1894, Perplexity: 8.9294
Epoch [3/3], Step [20/3236], Loss: 2.1714, Perplexity: 8.7707
Epoch [3/3], Step [21/3236], Loss: 2.2150, Perplexity: 9.1618
Epoch [3/3], Step [22/3236], Loss: 2.0853, Perplexity: 8.0472
Epoch [3/3], Step [23/3236], Loss: 2.0300, Perplexity: 7.6138
Epoch [3/3], Step [24/3236], Loss: 2.2243, Perplexity: 9.2471
Epoch [3/3], Step [25/3236], Loss: 2.3244, Perplexity: 10.2205
Epoch [3/3], Step [26/3236], Loss: 1.9659, Perplexity: 7.1413
Epoch [3/3], Step [27/3236], Loss: 2.0386, Perplexity: 7.6798
Epoch [3/3], Step [28/3236], Loss: 2.0567, Perplexity: 7.8199
Epoch [3/3], Step [29/3236], Loss: 2.4135, Perplexity: 11.1731
Epoch [3/3], Step [30/3236], Loss: 2.0821, Perplexity: 8.0211
Epoch [3/3], Step [31/3236], Loss: 2.0967, Perplexity: 8.1392
Epoch [3/3], Step [32/3236], Loss: 2.1633, Perplexity: 8.6999
Epoch [3/3], Step [33/3236], Loss: 2.0407, Perplexity: 7.6962
Epoch [3/3], Step [34/3236], Loss: 2.0538, Perplexity: 7.7976
Epoch [3/3], Step [35/3236], Loss: 2.1084, Perplexity: 8.2348
Epoch [3/3], Step [36/3236], Loss: 2.0575, Perplexity: 7.8264
Epoch [3/3], Step [37/3236], Loss: 2.1713, Perplexity: 8.7694
Epoch [3/3], Step [38/3236], Loss: 2.3611, Perplexity: 10.6031
Epoch [3/3], Step [39/3236], Loss: 2.0936, Perplexity: 8.1143
Epoch [3/3], Step [40/3236], Loss: 2.2298, Perplexity: 9.2976
Epoch [3/3], Step [41/3236], Loss: 2.0260, Perplexity: 7.5840
Epoch [3/3], Step [42/3236], Loss: 2.0941, Perplexity: 8.1185
Epoch [3/3], Step [43/3236], Loss: 2.1081, Perplexity: 8.2327
Epoch [3/3], Step [44/3236], Loss: 2.0242, Perplexity: 7.5703
Epoch [3/3], Step [45/3236], Loss: 2.1091, Perplexity: 8.2409
Epoch [3/3], Step [46/3236], Loss: 1.9793, Perplexity: 7.2378
Epoch [3/3], Step [47/3236], Loss: 2.1855, Perplexity: 8.8953
Epoch [3/3], Step [48/3236], Loss: 2.1941, Perplexity: 8.9720
Epoch [3/3], Step [49/3236], Loss: 2.0853, Perplexity: 8.0473
Epoch [3/3], Step [50/3236], Loss: 2.0367, Perplexity: 7.6654
Epoch [3/3], Step [51/3236], Loss: 2.0947, Perplexity: 8.1227
Epoch [3/3], Step [52/3236], Loss: 2.0812, Perplexity: 8.0140
Epoch [3/3], Step [53/3236], Loss: 2.1170, Perplexity: 8.3058
Epoch [3/3], Step [54/3236], Loss: 2.1689, Perplexity: 8.7491
Epoch [3/3], Step [55/3236], Loss: 2.1278, Perplexity: 8.3966
Epoch [3/3], Step [56/3236], Loss: 1.9937, Perplexity: 7.3428
Epoch [3/3], Step [57/3236], Loss: 2.0521, Perplexity: 7.7840
Epoch [3/3], Step [58/3236], Loss: 2.1779, Perplexity: 8.8282
Epoch [3/3], Step [59/3236], Loss: 2.0861, Perplexity: 8.0536
Epoch [3/3], Step [60/3236], Loss: 2.0099, Perplexity: 7.4629
Epoch [3/3], Step [61/3236], Loss: 1.9854, Perplexity: 7.2820
Epoch [3/3], Step [62/3236], Loss: 1.9847, Perplexity: 7.2769
Epoch [3/3], Step [63/3236], Loss: 2.4542, Perplexity: 11.6370
Epoch [3/3], Step [64/3236], Loss: 2.2309, Perplexity: 9.3079
Epoch [3/3], Step [65/3236], Loss: 2.1060, Perplexity: 8.2156
Epoch [3/3], Step [66/3236], Loss: 2.0769, Perplexity: 7.9797
Epoch [3/3], Step [67/3236], Loss: 1.9369, Perplexity: 6.9373
Epoch [3/3], Step [68/3236], Loss: 1.9766, Perplexity: 7.2178
Epoch [3/3], Step [69/3236], Loss: 2.3944, Perplexity: 10.9614
Epoch [3/3], Step [70/3236], Loss: 2.0512, Perplexity: 7.7773
Epoch [3/3], Step [71/3236], Loss: 2.1166, Perplexity: 8.3028
Epoch [3/3], Step [72/3236], Loss: 2.1026, Perplexity: 8.1873
Epoch [3/3], Step [73/3236], Loss: 2.0876, Perplexity: 8.0659
Epoch [3/3], Step [74/3236], Loss: 2.0423, Perplexity: 7.7080
Epoch [3/3], Step [75/3236], Loss: 2.0878, Perplexity: 8.0674
Epoch [3/3], Step [76/3236], Loss: 2.1231, Perplexity: 8.3567
Epoch [3/3], Step [77/3236], Loss: 2.2560, Perplexity: 9.5448
Epoch [3/3], Step [78/3236], Loss: 2.1072, Perplexity: 8.2250
Epoch [3/3], Step [79/3236], Loss: 1.9374, Perplexity: 6.9408
Epoch [3/3], Step [80/3236], Loss: 2.0818, Perplexity: 8.0191
Epoch [3/3], Step [81/3236], Loss: 3.0106, Perplexity: 20.3004
Epoch [3/3], Step [82/3236], Loss: 2.0644, Perplexity: 7.8802
Epoch [3/3], Step [83/3236], Loss: 2.2065, Perplexity: 9.0843
Epoch [3/3], Step [84/3236], Loss: 2.0447, Perplexity: 7.7266
Epoch [3/3], Step [85/3236], Loss: 2.2085, Perplexity: 9.1021
Epoch [3/3], Step [86/3236], Loss: 2.1276, Perplexity: 8.3950
Epoch [3/3], Step [87/3236], Loss: 2.1260, Perplexity: 8.3814
Epoch [3/3], Step [88/3236], Loss: 2.1964, Perplexity: 8.9923
Epoch [3/3], Step [89/3236], Loss: 2.3813, Perplexity: 10.8186
Epoch [3/3], Step [90/3236], Loss: 2.0709, Perplexity: 7.9321
Epoch [3/3], Step [91/3236], Loss: 2.7899, Perplexity: 16.2795
Epoch [3/3], Step [92/3236], Loss: 2.1211, Perplexity: 8.3405
Epoch [3/3], Step [93/3236], Loss: 2.0285, Perplexity: 7.6024
Epoch [3/3], Step [94/3236], Loss: 2.0996, Perplexity: 8.1625
Epoch [3/3], Step [95/3236], Loss: 2.0838, Perplexity: 8.0352
Epoch [3/3], Step [96/3236], Loss: 1.8884, Perplexity: 6.6088
Epoch [3/3], Step [97/3236], Loss: 2.1459, Perplexity: 8.5496
Epoch [3/3], Step [98/3236], Loss: 1.9496, Perplexity: 7.0255
Epoch [3/3], Step [99/3236], Loss: 2.1978, Perplexity: 9.0056
Epoch [3/3], Step [100/3236], Loss: 2.1301, Perplexity: 8.4159
Epoch [3/3], Step [101/3236], Loss: 2.1937, Perplexity: 8.9682
Epoch [3/3], Step [102/3236], Loss: 2.0001, Perplexity: 7.3897
Epoch [3/3], Step [103/3236], Loss: 2.5840, Perplexity: 13.2505
Epoch [3/3], Step [104/3236], Loss: 2.5213, Perplexity: 12.4452
Epoch [3/3], Step [105/3236], Loss: 1.9628, Perplexity: 7.1192
Epoch [3/3], Step [106/3236], Loss: 2.1299, Perplexity: 8.4142
Epoch [3/3], Step [107/3236], Loss: 2.0918, Perplexity: 8.0993
Epoch [3/3], Step [108/3236], Loss: 2.0664, Perplexity: 7.8965
Epoch [3/3], Step [109/3236], Loss: 2.0705, Perplexity: 7.9284
Epoch [3/3], Step [110/3236], Loss: 1.9938, Perplexity: 7.3432
Epoch [3/3], Step [111/3236], Loss: 2.0484, Perplexity: 7.7552
Epoch [3/3], Step [112/3236], Loss: 1.9732, Perplexity: 7.1939
Epoch [3/3], Step [113/3236], Loss: 1.9502, Perplexity: 7.0299
Epoch [3/3], Step [114/3236], Loss: 2.0766, Perplexity: 7.9770
Epoch [3/3], Step [115/3236], Loss: 2.3041, Perplexity: 10.0154
Epoch [3/3], Step [116/3236], Loss: 2.5210, Perplexity: 12.4414
Epoch [3/3], Step [117/3236], Loss: 2.0187, Perplexity: 7.5288
Epoch [3/3], Step [118/3236], Loss: 2.0832, Perplexity: 8.0304
Epoch [3/3], Step [119/3236], Loss: 2.0631, Perplexity: 7.8706
Epoch [3/3], Step [120/3236], Loss: 2.0541, Perplexity: 7.8002
Epoch [3/3], Step [121/3236], Loss: 2.0060, Perplexity: 7.4335
Epoch [3/3], Step [122/3236], Loss: 2.1223, Perplexity: 8.3502
Epoch [3/3], Step [123/3236], Loss: 2.0079, Perplexity: 7.4478
Epoch [3/3], Step [124/3236], Loss: 2.1539, Perplexity: 8.6180
Epoch [3/3], Step [125/3236], Loss: 2.0675, Perplexity: 7.9049
Epoch [3/3], Step [126/3236], Loss: 2.1026, Perplexity: 8.1870
Epoch [3/3], Step [127/3236], Loss: 2.0290, Perplexity: 7.6067
Epoch [3/3], Step [128/3236], Loss: 2.1761, Perplexity: 8.8121
Epoch [3/3], Step [129/3236], Loss: 2.2224, Perplexity: 9.2293
Epoch [3/3], Step [130/3236], Loss: 2.0496, Perplexity: 7.7650
Epoch [3/3], Step [131/3236], Loss: 2.1649, Perplexity: 8.7140
Epoch [3/3], Step [132/3236], Loss: 2.0711, Perplexity: 7.9332
Epoch [3/3], Step [133/3236], Loss: 2.0788, Perplexity: 7.9948
Epoch [3/3], Step [134/3236], Loss: 1.8894, Perplexity: 6.6152
Epoch [3/3], Step [135/3236], Loss: 2.0281, Perplexity: 7.5994
Epoch [3/3], Step [136/3236], Loss: 2.0214, Perplexity: 7.5487
Epoch [3/3], Step [137/3236], Loss: 1.9855, Perplexity: 7.2825
Epoch [3/3], Step [138/3236], Loss: 2.0855, Perplexity: 8.0490
Epoch [3/3], Step [139/3236], Loss: 2.3630, Perplexity: 10.6228
Epoch [3/3], Step [140/3236], Loss: 1.9721, Perplexity: 7.1857
Epoch [3/3], Step [141/3236], Loss: 2.1668, Perplexity: 8.7307
Epoch [3/3], Step [142/3236], Loss: 2.1101, Perplexity: 8.2492
Epoch [3/3], Step [143/3236], Loss: 2.0504, Perplexity: 7.7710
Epoch [3/3], Step [144/3236], Loss: 2.0319, Perplexity: 7.6288
Epoch [3/3], Step [145/3236], Loss: 2.0095, Perplexity: 7.4599
Epoch [3/3], Step [146/3236], Loss: 2.2518, Perplexity: 9.5051
Epoch [3/3], Step [147/3236], Loss: 2.2207, Perplexity: 9.2139
Epoch [3/3], Step [148/3236], Loss: 1.9798, Perplexity: 7.2415
Epoch [3/3], Step [149/3236], Loss: 1.9186, Perplexity: 6.8115
Epoch [3/3], Step [150/3236], Loss: 2.2035, Perplexity: 9.0564
Epoch [3/3], Step [151/3236], Loss: 2.1642, Perplexity: 8.7079
Epoch [3/3], Step [152/3236], Loss: 2.5934, Perplexity: 13.3757
Epoch [3/3], Step [153/3236], Loss: 2.0101, Perplexity: 7.4643
Epoch [3/3], Step [154/3236], Loss: 2.1371, Perplexity: 8.4751
Epoch [3/3], Step [155/3236], Loss: 2.0759, Perplexity: 7.9717
Epoch [3/3], Step [156/3236], Loss: 2.1204, Perplexity: 8.3349
Epoch [3/3], Step [157/3236], Loss: 1.9238, Perplexity: 6.8467
Epoch [3/3], Step [158/3236], Loss: 2.1290, Perplexity: 8.4062
Epoch [3/3], Step [159/3236], Loss: 2.0098, Perplexity: 7.4619
Epoch [3/3], Step [160/3236], Loss: 2.4621, Perplexity: 11.7291
Epoch [3/3], Step [161/3236], Loss: 2.0249, Perplexity: 7.5755
Epoch [3/3], Step [162/3236], Loss: 2.0026, Perplexity: 7.4083
Epoch [3/3], Step [163/3236], Loss: 2.9121, Perplexity: 18.3950
Epoch [3/3], Step [164/3236], Loss: 2.1735, Perplexity: 8.7894
Epoch [3/3], Step [165/3236], Loss: 2.2026, Perplexity: 9.0483
Epoch [3/3], Step [166/3236], Loss: 2.0401, Perplexity: 7.6914
Epoch [3/3], Step [167/3236], Loss: 2.0677, Perplexity: 7.9069
Epoch [3/3], Step [168/3236], Loss: 2.1002, Perplexity: 8.1677
Epoch [3/3], Step [169/3236], Loss: 1.9551, Perplexity: 7.0646
Epoch [3/3], Step [170/3236], Loss: 2.3610, Perplexity: 10.6014
Epoch [3/3], Step [171/3236], Loss: 2.2709, Perplexity: 9.6882
Epoch [3/3], Step [172/3236], Loss: 2.1074, Perplexity: 8.2270
Epoch [3/3], Step [173/3236], Loss: 2.1428, Perplexity: 8.5233
Epoch [3/3], Step [174/3236], Loss: 2.6019, Perplexity: 13.4890
Epoch [3/3], Step [175/3236], Loss: 1.9716, Perplexity: 7.1818
Epoch [3/3], Step [176/3236], Loss: 2.1518, Perplexity: 8.6001
Epoch [3/3], Step [177/3236], Loss: 2.1950, Perplexity: 8.9798
Epoch [3/3], Step [178/3236], Loss: 2.0778, Perplexity: 7.9869
Epoch [3/3], Step [179/3236], Loss: 2.4355, Perplexity: 11.4213
Epoch [3/3], Step [180/3236], Loss: 2.2626, Perplexity: 9.6083
Epoch [3/3], Step [181/3236], Loss: 2.0317, Perplexity: 7.6273
Epoch [3/3], Step [182/3236], Loss: 1.8616, Perplexity: 6.4340
Epoch [3/3], Step [183/3236], Loss: 2.2693, Perplexity: 9.6727
Epoch [3/3], Step [184/3236], Loss: 2.1552, Perplexity: 8.6293
Epoch [3/3], Step [185/3236], Loss: 2.3704, Perplexity: 10.7015
Epoch [3/3], Step [186/3236], Loss: 2.5347, Perplexity: 12.6125
Epoch [3/3], Step [187/3236], Loss: 2.7034, Perplexity: 14.9311
Epoch [3/3], Step [188/3236], Loss: 2.0819, Perplexity: 8.0198
Epoch [3/3], Step [189/3236], Loss: 2.0345, Perplexity: 7.6482
Epoch [3/3], Step [190/3236], Loss: 2.0467, Perplexity: 7.7420
Epoch [3/3], Step [191/3236], Loss: 2.0104, Perplexity: 7.4664
Epoch [3/3], Step [192/3236], Loss: 2.4145, Perplexity: 11.1847
Epoch [3/3], Step [193/3236], Loss: 2.0600, Perplexity: 7.8459
Epoch [3/3], Step [194/3236], Loss: 2.1300, Perplexity: 8.4146
Epoch [3/3], Step [195/3236], Loss: 2.0541, Perplexity: 7.7999
Epoch [3/3], Step [196/3236], Loss: 2.1666, Perplexity: 8.7287
Epoch [3/3], Step [197/3236], Loss: 2.2979, Perplexity: 9.9537
Epoch [3/3], Step [198/3236], Loss: 2.0746, Perplexity: 7.9615
Epoch [3/3], Step [199/3236], Loss: 1.9935, Perplexity: 7.3414
Epoch [3/3], Step [200/3236], Loss: 2.0850, Perplexity: 8.0445
Epoch [3/3], Step [201/3236], Loss: 1.9597, Perplexity: 7.0974
Epoch [3/3], Step [202/3236], Loss: 2.2326, Perplexity: 9.3241
Epoch [3/3], Step [203/3236], Loss: 2.0782, Perplexity: 7.9901
Epoch [3/3], Step [204/3236], Loss: 2.6389, Perplexity: 13.9982
Epoch [3/3], Step [205/3236], Loss: 1.9644, Perplexity: 7.1308
Epoch [3/3], Step [206/3236], Loss: 2.2883, Perplexity: 9.8578
Epoch [3/3], Step [207/3236], Loss: 2.2682, Perplexity: 9.6617
Epoch [3/3], Step [208/3236], Loss: 2.0714, Perplexity: 7.9358
Epoch [3/3], Step [209/3236], Loss: 2.1611, Perplexity: 8.6805
Epoch [3/3], Step [210/3236], Loss: 2.1685, Perplexity: 8.7451
Epoch [3/3], Step [211/3236], Loss: 2.0429, Perplexity: 7.7129
Epoch [3/3], Step [212/3236], Loss: 1.9502, Perplexity: 7.0299
Epoch [3/3], Step [213/3236], Loss: 2.4189, Perplexity: 11.2338
Epoch [3/3], Step [214/3236], Loss: 2.0527, Perplexity: 7.7891
Epoch [3/3], Step [215/3236], Loss: 2.1680, Perplexity: 8.7412
Epoch [3/3], Step [216/3236], Loss: 2.0133, Perplexity: 7.4884
Epoch [3/3], Step [217/3236], Loss: 2.4061, Perplexity: 11.0903
Epoch [3/3], Step [218/3236], Loss: 2.1444, Perplexity: 8.5365
Epoch [3/3], Step [219/3236], Loss: 2.9338, Perplexity: 18.7987
Epoch [3/3], Step [220/3236], Loss: 2.0718, Perplexity: 7.9391
Epoch [3/3], Step [221/3236], Loss: 2.0796, Perplexity: 8.0010
Epoch [3/3], Step [222/3236], Loss: 2.0613, Perplexity: 7.8560
Epoch [3/3], Step [223/3236], Loss: 2.1802, Perplexity: 8.8482
Epoch [3/3], Step [224/3236], Loss: 2.0755, Perplexity: 7.9683
Epoch [3/3], Step [225/3236], Loss: 2.0421, Perplexity: 7.7071
Epoch [3/3], Step [226/3236], Loss: 2.0949, Perplexity: 8.1246
Epoch [3/3], Step [227/3236], Loss: 2.0297, Perplexity: 7.6118
Epoch [3/3], Step [228/3236], Loss: 2.3707, Perplexity: 10.7050
Epoch [3/3], Step [229/3236], Loss: 2.0272, Perplexity: 7.5931
Epoch [3/3], Step [230/3236], Loss: 2.0705, Perplexity: 7.9288
Epoch [3/3], Step [231/3236], Loss: 2.0015, Perplexity: 7.4000
Epoch [3/3], Step [232/3236], Loss: 2.0884, Perplexity: 8.0723
Epoch [3/3], Step [233/3236], Loss: 2.0048, Perplexity: 7.4246
Epoch [3/3], Step [234/3236], Loss: 2.0217, Perplexity: 7.5512
Epoch [3/3], Step [235/3236], Loss: 2.0204, Perplexity: 7.5411
Epoch [3/3], Step [236/3236], Loss: 2.1375, Perplexity: 8.4786
Epoch [3/3], Step [237/3236], Loss: 2.0676, Perplexity: 7.9061
Epoch [3/3], Step [238/3236], Loss: 2.1930, Perplexity: 8.9620
Epoch [3/3], Step [239/3236], Loss: 2.0062, Perplexity: 7.4351
Epoch [3/3], Step [240/3236], Loss: 2.2315, Perplexity: 9.3134
Epoch [3/3], Step [241/3236], Loss: 2.1222, Perplexity: 8.3496
Epoch [3/3], Step [242/3236], Loss: 2.0129, Perplexity: 7.4850
Epoch [3/3], Step [243/3236], Loss: 2.0510, Perplexity: 7.7759
Epoch [3/3], Step [244/3236], Loss: 2.0412, Perplexity: 7.6996
Epoch [3/3], Step [245/3236], Loss: 2.0222, Perplexity: 7.5550
Epoch [3/3], Step [246/3236], Loss: 2.0609, Perplexity: 7.8533
Epoch [3/3], Step [247/3236], Loss: 2.2132, Perplexity: 9.1453
Epoch [3/3], Step [248/3236], Loss: 2.0596, Perplexity: 7.8428
Epoch [3/3], Step [249/3236], Loss: 2.0455, Perplexity: 7.7330
Epoch [3/3], Step [250/3236], Loss: 2.0170, Perplexity: 7.5155
Epoch [3/3], Step [251/3236], Loss: 2.6107, Perplexity: 13.6085
Epoch [3/3], Step [252/3236], Loss: 1.8558, Perplexity: 6.3971
Epoch [3/3], Step [253/3236], Loss: 2.0673, Perplexity: 7.9034
Epoch [3/3], Step [254/3236], Loss: 2.0456, Perplexity: 7.7338
Epoch [3/3], Step [255/3236], Loss: 2.0759, Perplexity: 7.9717
Epoch [3/3], Step [256/3236], Loss: 2.1812, Perplexity: 8.8572
Epoch [3/3], Step [257/3236], Loss: 2.1203, Perplexity: 8.3335
Epoch [3/3], Step [258/3236], Loss: 2.0491, Perplexity: 7.7610
Epoch [3/3], Step [259/3236], Loss: 2.0129, Perplexity: 7.4850
Epoch [3/3], Step [260/3236], Loss: 1.8673, Perplexity: 6.4711
Epoch [3/3], Step [261/3236], Loss: 2.0001, Perplexity: 7.3899
Epoch [3/3], Step [262/3236], Loss: 1.9712, Perplexity: 7.1793
Epoch [3/3], Step [263/3236], Loss: 2.1789, Perplexity: 8.8369
Epoch [3/3], Step [264/3236], Loss: 2.0572, Perplexity: 7.8238
Epoch [3/3], Step [265/3236], Loss: 1.9254, Perplexity: 6.8581
Epoch [3/3], Step [266/3236], Loss: 2.1583, Perplexity: 8.6568
Epoch [3/3], Step [267/3236], Loss: 2.0345, Perplexity: 7.6485
Epoch [3/3], Step [268/3236], Loss: 2.1255, Perplexity: 8.3767
Epoch [3/3], Step [269/3236], Loss: 2.0247, Perplexity: 7.5735
Epoch [3/3], Step [270/3236], Loss: 2.3589, Perplexity: 10.5788
Epoch [3/3], Step [271/3236], Loss: 2.0581, Perplexity: 7.8310
Epoch [3/3], Step [272/3236], Loss: 1.9243, Perplexity: 6.8505
Epoch [3/3], Step [273/3236], Loss: 2.1584, Perplexity: 8.6576
Epoch [3/3], Step [274/3236], Loss: 1.9916, Perplexity: 7.3270
Epoch [3/3], Step [275/3236], Loss: 2.0892, Perplexity: 8.0784
Epoch [3/3], Step [276/3236], Loss: 1.9124, Perplexity: 6.7694
Epoch [3/3], Step [277/3236], Loss: 2.1565, Perplexity: 8.6406
Epoch [3/3], Step [278/3236], Loss: 2.0808, Perplexity: 8.0109
Epoch [3/3], Step [279/3236], Loss: 2.5691, Perplexity: 13.0542
Epoch [3/3], Step [280/3236], Loss: 2.4469, Perplexity: 11.5528
Epoch [3/3], Step [281/3236], Loss: 2.0219, Perplexity: 7.5530
Epoch [3/3], Step [282/3236], Loss: 2.0320, Perplexity: 7.6295
Epoch [3/3], Step [283/3236], Loss: 1.9821, Perplexity: 7.2581
Epoch [3/3], Step [284/3236], Loss: 2.0820, Perplexity: 8.0204
Epoch [3/3], Step [285/3236], Loss: 2.3197, Perplexity: 10.1729
Epoch [3/3], Step [286/3236], Loss: 2.0751, Perplexity: 7.9651
Epoch [3/3], Step [287/3236], Loss: 2.1112, Perplexity: 8.2581
Epoch [3/3], Step [288/3236], Loss: 2.1238, Perplexity: 8.3628
Epoch [3/3], Step [289/3236], Loss: 2.4773, Perplexity: 11.9087
Epoch [3/3], Step [290/3236], Loss: 1.9640, Perplexity: 7.1276
Epoch [3/3], Step [291/3236], Loss: 2.1208, Perplexity: 8.3377
Epoch [3/3], Step [292/3236], Loss: 2.0785, Perplexity: 7.9924
Epoch [3/3], Step [293/3236], Loss: 2.1017, Perplexity: 8.1803
Epoch [3/3], Step [294/3236], Loss: 2.3737, Perplexity: 10.7365
Epoch [3/3], Step [295/3236], Loss: 1.9951, Perplexity: 7.3532
Epoch [3/3], Step [296/3236], Loss: 2.3251, Perplexity: 10.2281
Epoch [3/3], Step [297/3236], Loss: 2.7684, Perplexity: 15.9336
Epoch [3/3], Step [298/3236], Loss: 2.0620, Perplexity: 7.8621
Epoch [3/3], Step [299/3236], Loss: 2.0479, Perplexity: 7.7516
Epoch [3/3], Step [300/3236], Loss: 2.1348, Perplexity: 8.4555
Epoch [3/3], Step [301/3236], Loss: 2.0990, Perplexity: 8.1583
Epoch [3/3], Step [302/3236], Loss: 1.9316, Perplexity: 6.9008
Epoch [3/3], Step [303/3236], Loss: 2.1704, Perplexity: 8.7619
Epoch [3/3], Step [304/3236], Loss: 2.0662, Perplexity: 7.8948
Epoch [3/3], Step [305/3236], Loss: 2.0659, Perplexity: 7.8920
Epoch [3/3], Step [306/3236], Loss: 2.6834, Perplexity: 14.6354
Epoch [3/3], Step [307/3236], Loss: 2.1738, Perplexity: 8.7913
Epoch [3/3], Step [308/3236], Loss: 2.0368, Perplexity: 7.6658
Epoch [3/3], Step [309/3236], Loss: 2.2608, Perplexity: 9.5908
Epoch [3/3], Step [310/3236], Loss: 1.9870, Perplexity: 7.2933
Epoch [3/3], Step [311/3236], Loss: 2.1010, Perplexity: 8.1747
Epoch [3/3], Step [312/3236], Loss: 2.1001, Perplexity: 8.1668
Epoch [3/3], Step [313/3236], Loss: 2.4878, Perplexity: 12.0344
Epoch [3/3], Step [314/3236], Loss: 2.1142, Perplexity: 8.2826
Epoch [3/3], Step [315/3236], Loss: 1.9686, Perplexity: 7.1603
Epoch [3/3], Step [316/3236], Loss: 2.1660, Perplexity: 8.7231
Epoch [3/3], Step [317/3236], Loss: 2.1314, Perplexity: 8.4268
Epoch [3/3], Step [318/3236], Loss: 2.0977, Perplexity: 8.1471
Epoch [3/3], Step [319/3236], Loss: 1.9943, Perplexity: 7.3473
Epoch [3/3], Step [320/3236], Loss: 2.0811, Perplexity: 8.0133
Epoch [3/3], Step [321/3236], Loss: 1.9977, Perplexity: 7.3722
Epoch [3/3], Step [322/3236], Loss: 1.9202, Perplexity: 6.8221
Epoch [3/3], Step [323/3236], Loss: 2.0626, Perplexity: 7.8663
Epoch [3/3], Step [324/3236], Loss: 1.8959, Perplexity: 6.6584
Epoch [3/3], Step [325/3236], Loss: 2.0821, Perplexity: 8.0214
Epoch [3/3], Step [326/3236], Loss: 2.0757, Perplexity: 7.9697
Epoch [3/3], Step [327/3236], Loss: 2.2991, Perplexity: 9.9657
Epoch [3/3], Step [328/3236], Loss: 2.2057, Perplexity: 9.0764
Epoch [3/3], Step [329/3236], Loss: 1.9650, Perplexity: 7.1350
Epoch [3/3], Step [330/3236], Loss: 2.6703, Perplexity: 14.4440
Epoch [3/3], Step [331/3236], Loss: 1.9377, Perplexity: 6.9426
Epoch [3/3], Step [332/3236], Loss: 2.1116, Perplexity: 8.2614
Epoch [3/3], Step [333/3236], Loss: 2.3171, Perplexity: 10.1463
Epoch [3/3], Step [334/3236], Loss: 2.0621, Perplexity: 7.8621
Epoch [3/3], Step [335/3236], Loss: 2.0948, Perplexity: 8.1235
Epoch [3/3], Step [336/3236], Loss: 2.2549, Perplexity: 9.5346
Epoch [3/3], Step [337/3236], Loss: 2.1231, Perplexity: 8.3573
Epoch [3/3], Step [338/3236], Loss: 2.0850, Perplexity: 8.0445
Epoch [3/3], Step [339/3236], Loss: 1.9956, Perplexity: 7.3566
Epoch [3/3], Step [340/3236], Loss: 2.3844, Perplexity: 10.8523
Epoch [3/3], Step [341/3236], Loss: 2.1362, Perplexity: 8.4676
Epoch [3/3], Step [342/3236], Loss: 2.1389, Perplexity: 8.4901
Epoch [3/3], Step [343/3236], Loss: 1.9843, Perplexity: 7.2740
Epoch [3/3], Step [344/3236], Loss: 2.1306, Perplexity: 8.4196
Epoch [3/3], Step [345/3236], Loss: 2.0348, Perplexity: 7.6511
Epoch [3/3], Step [346/3236], Loss: 2.2458, Perplexity: 9.4477
Epoch [3/3], Step [347/3236], Loss: 2.0634, Perplexity: 7.8726
Epoch [3/3], Step [348/3236], Loss: 1.9858, Perplexity: 7.2846
Epoch [3/3], Step [349/3236], Loss: 2.1253, Perplexity: 8.3751
Epoch [3/3], Step [350/3236], Loss: 2.1011, Perplexity: 8.1754
Epoch [3/3], Step [351/3236], Loss: 2.5327, Perplexity: 12.5878
Epoch [3/3], Step [352/3236], Loss: 1.9938, Perplexity: 7.3433
Epoch [3/3], Step [353/3236], Loss: 3.5749, Perplexity: 35.6918
Epoch [3/3], Step [354/3236], Loss: 2.0475, Perplexity: 7.7487
Epoch [3/3], Step [355/3236], Loss: 2.0214, Perplexity: 7.5491
Epoch [3/3], Step [356/3236], Loss: 1.9480, Perplexity: 7.0150
Epoch [3/3], Step [357/3236], Loss: 1.9471, Perplexity: 7.0085
Epoch [3/3], Step [358/3236], Loss: 1.9419, Perplexity: 6.9718
Epoch [3/3], Step [359/3236], Loss: 1.9427, Perplexity: 6.9775
Epoch [3/3], Step [360/3236], Loss: 2.6126, Perplexity: 13.6338
Epoch [3/3], Step [361/3236], Loss: 1.9790, Perplexity: 7.2355
Epoch [3/3], Step [362/3236], Loss: 2.0773, Perplexity: 7.9832
Epoch [3/3], Step [363/3236], Loss: 2.0009, Perplexity: 7.3955
Epoch [3/3], Step [364/3236], Loss: 1.9776, Perplexity: 7.2252
Epoch [3/3], Step [365/3236], Loss: 2.1725, Perplexity: 8.7802
Epoch [3/3], Step [366/3236], Loss: 1.9866, Perplexity: 7.2909
Epoch [3/3], Step [367/3236], Loss: 2.1215, Perplexity: 8.3439
Epoch [3/3], Step [368/3236], Loss: 2.1494, Perplexity: 8.5795
Epoch [3/3], Step [369/3236], Loss: 2.0657, Perplexity: 7.8910
Epoch [3/3], Step [370/3236], Loss: 1.9618, Perplexity: 7.1123
Epoch [3/3], Step [371/3236], Loss: 2.1137, Perplexity: 8.2787
Epoch [3/3], Step [372/3236], Loss: 2.8308, Perplexity: 16.9591
Epoch [3/3], Step [373/3236], Loss: 2.1597, Perplexity: 8.6690
Epoch [3/3], Step [374/3236], Loss: 2.1291, Perplexity: 8.4070
Epoch [3/3], Step [375/3236], Loss: 2.0126, Perplexity: 7.4828
Epoch [3/3], Step [376/3236], Loss: 2.0030, Perplexity: 7.4115
Epoch [3/3], Step [377/3236], Loss: 2.9964, Perplexity: 20.0124
Epoch [3/3], Step [378/3236], Loss: 2.0886, Perplexity: 8.0733
Epoch [3/3], Step [379/3236], Loss: 1.9846, Perplexity: 7.2761
Epoch [3/3], Step [380/3236], Loss: 2.3340, Perplexity: 10.3196
Epoch [3/3], Step [381/3236], Loss: 2.0335, Perplexity: 7.6407
Epoch [3/3], Step [382/3236], Loss: 2.2571, Perplexity: 9.5557
Epoch [3/3], Step [383/3236], Loss: 1.9348, Perplexity: 6.9225
Epoch [3/3], Step [384/3236], Loss: 2.1859, Perplexity: 8.8990
Epoch [3/3], Step [385/3236], Loss: 1.8679, Perplexity: 6.4747
Epoch [3/3], Step [386/3236], Loss: 2.1376, Perplexity: 8.4788
Epoch [3/3], Step [387/3236], Loss: 2.2228, Perplexity: 9.2330
Epoch [3/3], Step [388/3236], Loss: 2.0590, Perplexity: 7.8378
Epoch [3/3], Step [389/3236], Loss: 2.7439, Perplexity: 15.5476
Epoch [3/3], Step [390/3236], Loss: 2.1131, Perplexity: 8.2739
Epoch [3/3], Step [391/3236], Loss: 2.1540, Perplexity: 8.6196
Epoch [3/3], Step [392/3236], Loss: 2.0004, Perplexity: 7.3923
Epoch [3/3], Step [393/3236], Loss: 2.1784, Perplexity: 8.8325
Epoch [3/3], Step [394/3236], Loss: 2.1159, Perplexity: 8.2968
Epoch [3/3], Step [395/3236], Loss: 2.0184, Perplexity: 7.5260
Epoch [3/3], Step [396/3236], Loss: 2.2724, Perplexity: 9.7025
Epoch [3/3], Step [397/3236], Loss: 2.1146, Perplexity: 8.2860
Epoch [3/3], Step [398/3236], Loss: 2.0833, Perplexity: 8.0310
Epoch [3/3], Step [399/3236], Loss: 2.0618, Perplexity: 7.8602
Epoch [3/3], Step [400/3236], Loss: 2.1822, Perplexity: 8.8660
Epoch [3/3], Step [401/3236], Loss: 2.0880, Perplexity: 8.0684
Epoch [3/3], Step [402/3236], Loss: 2.1470, Perplexity: 8.5592
Epoch [3/3], Step [403/3236], Loss: 2.0648, Perplexity: 7.8840
Epoch [3/3], Step [404/3236], Loss: 2.2884, Perplexity: 9.8587
Epoch [3/3], Step [405/3236], Loss: 1.9022, Perplexity: 6.7007
Epoch [3/3], Step [406/3236], Loss: 1.9072, Perplexity: 6.7341
Epoch [3/3], Step [407/3236], Loss: 2.5930, Perplexity: 13.3696
Epoch [3/3], Step [408/3236], Loss: 2.0149, Perplexity: 7.5000
Epoch [3/3], Step [409/3236], Loss: 2.3306, Perplexity: 10.2839
Epoch [3/3], Step [410/3236], Loss: 1.8800, Perplexity: 6.5532
Epoch [3/3], Step [411/3236], Loss: 2.0432, Perplexity: 7.7155
Epoch [3/3], Step [412/3236], Loss: 2.0839, Perplexity: 8.0355
Epoch [3/3], Step [413/3236], Loss: 1.9869, Perplexity: 7.2929
Epoch [3/3], Step [414/3236], Loss: 2.2551, Perplexity: 9.5358
Epoch [3/3], Step [415/3236], Loss: 2.2224, Perplexity: 9.2296
Epoch [3/3], Step [416/3236], Loss: 1.9656, Perplexity: 7.1393
Epoch [3/3], Step [417/3236], Loss: 2.0782, Perplexity: 7.9902
Epoch [3/3], Step [418/3236], Loss: 2.1534, Perplexity: 8.6142
Epoch [3/3], Step [419/3236], Loss: 2.3202, Perplexity: 10.1779
Epoch [3/3], Step [420/3236], Loss: 2.0597, Perplexity: 7.8433
Epoch [3/3], Step [421/3236], Loss: 2.1160, Perplexity: 8.2978
Epoch [3/3], Step [422/3236], Loss: 2.0651, Perplexity: 7.8864
Epoch [3/3], Step [423/3236], Loss: 2.1365, Perplexity: 8.4701
Epoch [3/3], Step [424/3236], Loss: 2.0906, Perplexity: 8.0896
Epoch [3/3], Step [425/3236], Loss: 2.0917, Perplexity: 8.0983
Epoch [3/3], Step [426/3236], Loss: 2.0962, Perplexity: 8.1349
Epoch [3/3], Step [427/3236], Loss: 1.9987, Perplexity: 7.3793
Epoch [3/3], Step [428/3236], Loss: 2.0148, Perplexity: 7.4994
Epoch [3/3], Step [429/3236], Loss: 1.9573, Perplexity: 7.0800
Epoch [3/3], Step [430/3236], Loss: 2.0964, Perplexity: 8.1369
Epoch [3/3], Step [431/3236], Loss: 2.4642, Perplexity: 11.7540
Epoch [3/3], Step [432/3236], Loss: 1.9308, Perplexity: 6.8953
Epoch [3/3], Step [433/3236], Loss: 2.0811, Perplexity: 8.0133
Epoch [3/3], Step [434/3236], Loss: 2.3554, Perplexity: 10.5421
Epoch [3/3], Step [435/3236], Loss: 2.0435, Perplexity: 7.7177
Epoch [3/3], Step [436/3236], Loss: 2.1462, Perplexity: 8.5520
Epoch [3/3], Step [437/3236], Loss: 1.9766, Perplexity: 7.2183
Epoch [3/3], Step [438/3236], Loss: 1.9894, Perplexity: 7.3109
Epoch [3/3], Step [439/3236], Loss: 2.2605, Perplexity: 9.5880
Epoch [3/3], Step [440/3236], Loss: 2.2481, Perplexity: 9.4702
Epoch [3/3], Step [441/3236], Loss: 2.0499, Perplexity: 7.7672
Epoch [3/3], Step [442/3236], Loss: 2.1777, Perplexity: 8.8259
Epoch [3/3], Step [443/3236], Loss: 2.1072, Perplexity: 8.2254
Epoch [3/3], Step [444/3236], Loss: 2.8704, Perplexity: 17.6444
Epoch [3/3], Step [445/3236], Loss: 2.1993, Perplexity: 9.0184
Epoch [3/3], Step [446/3236], Loss: 2.5918, Perplexity: 13.3535
Epoch [3/3], Step [447/3236], Loss: 2.5009, Perplexity: 12.1938
Epoch [3/3], Step [448/3236], Loss: 1.9413, Perplexity: 6.9680
Epoch [3/3], Step [449/3236], Loss: 1.9531, Perplexity: 7.0504
Epoch [3/3], Step [450/3236], Loss: 2.1756, Perplexity: 8.8071
Epoch [3/3], Step [451/3236], Loss: 2.0014, Perplexity: 7.3994
Epoch [3/3], Step [452/3236], Loss: 2.8695, Perplexity: 17.6286
Epoch [3/3], Step [453/3236], Loss: 2.4230, Perplexity: 11.2797
Epoch [3/3], Step [454/3236], Loss: 2.6389, Perplexity: 13.9981
Epoch [3/3], Step [455/3236], Loss: 2.0416, Perplexity: 7.7027
Epoch [3/3], Step [456/3236], Loss: 2.1394, Perplexity: 8.4946
Epoch [3/3], Step [457/3236], Loss: 1.9549, Perplexity: 7.0634
Epoch [3/3], Step [458/3236], Loss: 2.1942, Perplexity: 8.9728
Epoch [3/3], Step [459/3236], Loss: 2.1101, Perplexity: 8.2490
Epoch [3/3], Step [460/3236], Loss: 2.3716, Perplexity: 10.7148
Epoch [3/3], Step [461/3236], Loss: 2.0632, Perplexity: 7.8709
Epoch [3/3], Step [462/3236], Loss: 2.0883, Perplexity: 8.0709
Epoch [3/3], Step [463/3236], Loss: 2.1142, Perplexity: 8.2826
Epoch [3/3], Step [464/3236], Loss: 2.0646, Perplexity: 7.8823
Epoch [3/3], Step [465/3236], Loss: 2.0131, Perplexity: 7.4867
Epoch [3/3], Step [466/3236], Loss: 2.0783, Perplexity: 7.9912
Epoch [3/3], Step [467/3236], Loss: 2.0995, Perplexity: 8.1621
Epoch [3/3], Step [468/3236], Loss: 2.3629, Perplexity: 10.6218
Epoch [3/3], Step [469/3236], Loss: 2.1231, Perplexity: 8.3572
Epoch [3/3], Step [470/3236], Loss: 2.1913, Perplexity: 8.9469
Epoch [3/3], Step [471/3236], Loss: 2.0300, Perplexity: 7.6144
Epoch [3/3], Step [472/3236], Loss: 2.0499, Perplexity: 7.7668
Epoch [3/3], Step [473/3236], Loss: 2.0543, Perplexity: 7.8013
Epoch [3/3], Step [474/3236], Loss: 2.1929, Perplexity: 8.9615
Epoch [3/3], Step [475/3236], Loss: 2.0160, Perplexity: 7.5081
Epoch [3/3], Step [476/3236], Loss: 2.1317, Perplexity: 8.4289
Epoch [3/3], Step [477/3236], Loss: 2.1237, Perplexity: 8.3616
Epoch [3/3], Step [478/3236], Loss: 2.0124, Perplexity: 7.4815
Epoch [3/3], Step [479/3236], Loss: 2.0060, Perplexity: 7.4334
Epoch [3/3], Step [480/3236], Loss: 2.0225, Perplexity: 7.5572
Epoch [3/3], Step [481/3236], Loss: 2.7098, Perplexity: 15.0259
Epoch [3/3], Step [482/3236], Loss: 2.0068, Perplexity: 7.4393
Epoch [3/3], Step [483/3236], Loss: 2.1498, Perplexity: 8.5831
Epoch [3/3], Step [484/3236], Loss: 2.6593, Perplexity: 14.2869
Epoch [3/3], Step [485/3236], Loss: 2.1666, Perplexity: 8.7285
Epoch [3/3], Step [486/3236], Loss: 1.9821, Perplexity: 7.2577
Epoch [3/3], Step [487/3236], Loss: 2.5310, Perplexity: 12.5665
Epoch [3/3], Step [488/3236], Loss: 1.9222, Perplexity: 6.8360
Epoch [3/3], Step [489/3236], Loss: 2.0063, Perplexity: 7.4360
Epoch [3/3], Step [490/3236], Loss: 2.0061, Perplexity: 7.4346
Epoch [3/3], Step [491/3236], Loss: 2.0266, Perplexity: 7.5880
Epoch [3/3], Step [492/3236], Loss: 2.1874, Perplexity: 8.9120
Epoch [3/3], Step [493/3236], Loss: 2.3720, Perplexity: 10.7187
Epoch [3/3], Step [494/3236], Loss: 2.0652, Perplexity: 7.8868
Epoch [3/3], Step [495/3236], Loss: 2.1461, Perplexity: 8.5516
Epoch [3/3], Step [496/3236], Loss: 1.9531, Perplexity: 7.0502
Epoch [3/3], Step [497/3236], Loss: 1.9053, Perplexity: 6.7216
Epoch [3/3], Step [498/3236], Loss: 2.1179, Perplexity: 8.3141
Epoch [3/3], Step [499/3236], Loss: 2.0434, Perplexity: 7.7167
Epoch [3/3], Step [500/3236], Loss: 1.9743, Perplexity: 7.2019
Epoch [3/3], Step [501/3236], Loss: 1.9290, Perplexity: 6.8825
Epoch [3/3], Step [502/3236], Loss: 2.1425, Perplexity: 8.5211
Epoch [3/3], Step [503/3236], Loss: 2.0056, Perplexity: 7.4308
Epoch [3/3], Step [504/3236], Loss: 2.1404, Perplexity: 8.5031
Epoch [3/3], Step [505/3236], Loss: 2.0683, Perplexity: 7.9113
Epoch [3/3], Step [506/3236], Loss: 2.0338, Perplexity: 7.6428
Epoch [3/3], Step [507/3236], Loss: 2.1044, Perplexity: 8.2020
Epoch [3/3], Step [508/3236], Loss: 2.0218, Perplexity: 7.5515
Epoch [3/3], Step [509/3236], Loss: 2.0186, Perplexity: 7.5276
Epoch [3/3], Step [510/3236], Loss: 2.9744, Perplexity: 19.5777
Epoch [3/3], Step [511/3236], Loss: 2.2518, Perplexity: 9.5050
Epoch [3/3], Step [512/3236], Loss: 1.9324, Perplexity: 6.9058
Epoch [3/3], Step [513/3236], Loss: 2.1051, Perplexity: 8.2083
Epoch [3/3], Step [514/3236], Loss: 2.4244, Perplexity: 11.2959
Epoch [3/3], Step [515/3236], Loss: 2.1110, Perplexity: 8.2564
Epoch [3/3], Step [516/3236], Loss: 1.9876, Perplexity: 7.2982
Epoch [3/3], Step [517/3236], Loss: 2.6954, Perplexity: 14.8118
Epoch [3/3], Step [518/3236], Loss: 2.3441, Perplexity: 10.4240
Epoch [3/3], Step [519/3236], Loss: 2.0549, Perplexity: 7.8057
Epoch [3/3], Step [520/3236], Loss: 2.0585, Perplexity: 7.8341
Epoch [3/3], Step [521/3236], Loss: 2.0098, Perplexity: 7.4615
Epoch [3/3], Step [522/3236], Loss: 1.9547, Perplexity: 7.0621
Epoch [3/3], Step [523/3236], Loss: 2.2899, Perplexity: 9.8736
Epoch [3/3], Step [524/3236], Loss: 2.0872, Perplexity: 8.0619
Epoch [3/3], Step [525/3236], Loss: 2.0262, Perplexity: 7.5855
Epoch [3/3], Step [526/3236], Loss: 2.1225, Perplexity: 8.3523
Epoch [3/3], Step [527/3236], Loss: 2.0701, Perplexity: 7.9258
Epoch [3/3], Step [528/3236], Loss: 1.9181, Perplexity: 6.8080
Epoch [3/3], Step [529/3236], Loss: 2.2925, Perplexity: 9.8999
Epoch [3/3], Step [530/3236], Loss: 2.0597, Perplexity: 7.8438
Epoch [3/3], Step [531/3236], Loss: 2.1051, Perplexity: 8.2076
Epoch [3/3], Step [532/3236], Loss: 2.2922, Perplexity: 9.8972
Epoch [3/3], Step [533/3236], Loss: 2.0461, Perplexity: 7.7380
Epoch [3/3], Step [534/3236], Loss: 2.0374, Perplexity: 7.6709
Epoch [3/3], Step [535/3236], Loss: 2.2563, Perplexity: 9.5477
Epoch [3/3], Step [536/3236], Loss: 2.4537, Perplexity: 11.6313
Epoch [3/3], Step [537/3236], Loss: 2.2531, Perplexity: 9.5168
Epoch [3/3], Step [538/3236], Loss: 2.1594, Perplexity: 8.6661
Epoch [3/3], Step [539/3236], Loss: 2.1656, Perplexity: 8.7200
Epoch [3/3], Step [540/3236], Loss: 1.9821, Perplexity: 7.2582
Epoch [3/3], Step [541/3236], Loss: 2.1401, Perplexity: 8.5001
Epoch [3/3], Step [542/3236], Loss: 2.0563, Perplexity: 7.8170
Epoch [3/3], Step [543/3236], Loss: 2.3211, Perplexity: 10.1866
Epoch [3/3], Step [544/3236], Loss: 2.0063, Perplexity: 7.4360
Epoch [3/3], Step [545/3236], Loss: 2.2069, Perplexity: 9.0876
Epoch [3/3], Step [546/3236], Loss: 1.9336, Perplexity: 6.9147
Epoch [3/3], Step [547/3236], Loss: 1.9672, Perplexity: 7.1504
Epoch [3/3], Step [548/3236], Loss: 2.0855, Perplexity: 8.0489
Epoch [3/3], Step [549/3236], Loss: 2.0538, Perplexity: 7.7978
Epoch [3/3], Step [550/3236], Loss: 2.0127, Perplexity: 7.4837
Epoch [3/3], Step [551/3236], Loss: 2.2332, Perplexity: 9.3296
Epoch [3/3], Step [552/3236], Loss: 2.1146, Perplexity: 8.2865
Epoch [3/3], Step [553/3236], Loss: 2.0640, Perplexity: 7.8778
Epoch [3/3], Step [554/3236], Loss: 2.2795, Perplexity: 9.7722
Epoch [3/3], Step [555/3236], Loss: 2.4548, Perplexity: 11.6436
Epoch [3/3], Step [556/3236], Loss: 2.0672, Perplexity: 7.9023
Epoch [3/3], Step [557/3236], Loss: 2.0433, Perplexity: 7.7162
Epoch [3/3], Step [558/3236], Loss: 2.0780, Perplexity: 7.9887
Epoch [3/3], Step [559/3236], Loss: 2.0529, Perplexity: 7.7908
Epoch [3/3], Step [560/3236], Loss: 2.0763, Perplexity: 7.9745
Epoch [3/3], Step [561/3236], Loss: 2.1100, Perplexity: 8.2483
Epoch [3/3], Step [562/3236], Loss: 1.9573, Perplexity: 7.0803
Epoch [3/3], Step [563/3236], Loss: 2.0592, Perplexity: 7.8394
Epoch [3/3], Step [564/3236], Loss: 2.8170, Perplexity: 16.7272
Epoch [3/3], Step [565/3236], Loss: 2.1328, Perplexity: 8.4387
Epoch [3/3], Step [566/3236], Loss: 2.0656, Perplexity: 7.8903
Epoch [3/3], Step [567/3236], Loss: 2.1525, Perplexity: 8.6060
Epoch [3/3], Step [568/3236], Loss: 2.0207, Perplexity: 7.5432
Epoch [3/3], Step [569/3236], Loss: 2.4020, Perplexity: 11.0455
Epoch [3/3], Step [570/3236], Loss: 2.0116, Perplexity: 7.4752
Epoch [3/3], Step [571/3236], Loss: 1.9819, Perplexity: 7.2566
Epoch [3/3], Step [572/3236], Loss: 1.9284, Perplexity: 6.8787
Epoch [3/3], Step [573/3236], Loss: 2.1996, Perplexity: 9.0215
Epoch [3/3], Step [574/3236], Loss: 1.9194, Perplexity: 6.8166
Epoch [3/3], Step [575/3236], Loss: 2.0645, Perplexity: 7.8811
Epoch [3/3], Step [576/3236], Loss: 2.5433, Perplexity: 12.7212
Epoch [3/3], Step [577/3236], Loss: 2.5392, Perplexity: 12.6697
Epoch [3/3], Step [578/3236], Loss: 1.9950, Perplexity: 7.3519
Epoch [3/3], Step [579/3236], Loss: 2.0723, Perplexity: 7.9428
Epoch [3/3], Step [580/3236], Loss: 1.9961, Perplexity: 7.3604
Epoch [3/3], Step [581/3236], Loss: 2.0085, Perplexity: 7.4523
Epoch [3/3], Step [582/3236], Loss: 2.0382, Perplexity: 7.6768
Epoch [3/3], Step [583/3236], Loss: 2.1292, Perplexity: 8.4082
Epoch [3/3], Step [584/3236], Loss: 2.1923, Perplexity: 8.9554
Epoch [3/3], Step [585/3236], Loss: 2.0789, Perplexity: 7.9959
Epoch [3/3], Step [586/3236], Loss: 2.4429, Perplexity: 11.5068
Epoch [3/3], Step [587/3236], Loss: 2.1110, Perplexity: 8.2567
Epoch [3/3], Step [588/3236], Loss: 2.3548, Perplexity: 10.5362
Epoch [3/3], Step [589/3236], Loss: 1.9811, Perplexity: 7.2504
Epoch [3/3], Step [590/3236], Loss: 2.2967, Perplexity: 9.9415
Epoch [3/3], Step [591/3236], Loss: 1.9748, Perplexity: 7.2049
Epoch [3/3], Step [592/3236], Loss: 2.0836, Perplexity: 8.0331
Epoch [3/3], Step [593/3236], Loss: 2.0899, Perplexity: 8.0841
Epoch [3/3], Step [594/3236], Loss: 2.0292, Perplexity: 7.6082
Epoch [3/3], Step [595/3236], Loss: 2.0767, Perplexity: 7.9780
Epoch [3/3], Step [596/3236], Loss: 2.1341, Perplexity: 8.4490
Epoch [3/3], Step [597/3236], Loss: 2.0542, Perplexity: 7.8005
Epoch [3/3], Step [598/3236], Loss: 2.0566, Perplexity: 7.8191
Epoch [3/3], Step [599/3236], Loss: 2.6030, Perplexity: 13.5039
Epoch [3/3], Step [600/3236], Loss: 2.0442, Perplexity: 7.7231
Epoch [3/3], Step [601/3236], Loss: 2.7424, Perplexity: 15.5244
Epoch [3/3], Step [602/3236], Loss: 2.4543, Perplexity: 11.6382
Epoch [3/3], Step [603/3236], Loss: 2.0288, Perplexity: 7.6051
Epoch [3/3], Step [604/3236], Loss: 2.1590, Perplexity: 8.6625
Epoch [3/3], Step [605/3236], Loss: 2.1021, Perplexity: 8.1836
Epoch [3/3], Step [606/3236], Loss: 2.2443, Perplexity: 9.4339
Epoch [3/3], Step [607/3236], Loss: 1.9111, Perplexity: 6.7608
Epoch [3/3], Step [608/3236], Loss: 2.6559, Perplexity: 14.2380
Epoch [3/3], Step [609/3236], Loss: 2.0319, Perplexity: 7.6289
Epoch [3/3], Step [610/3236], Loss: 1.9800, Perplexity: 7.2429
Epoch [3/3], Step [611/3236], Loss: 1.8994, Perplexity: 6.6821
Epoch [3/3], Step [612/3236], Loss: 2.1649, Perplexity: 8.7137
Epoch [3/3], Step [613/3236], Loss: 3.4203, Perplexity: 30.5780
Epoch [3/3], Step [614/3236], Loss: 2.1067, Perplexity: 8.2210
Epoch [3/3], Step [615/3236], Loss: 2.0155, Perplexity: 7.5043
Epoch [3/3], Step [616/3236], Loss: 1.9292, Perplexity: 6.8840
Epoch [3/3], Step [617/3236], Loss: 2.0722, Perplexity: 7.9424
Epoch [3/3], Step [618/3236], Loss: 2.1523, Perplexity: 8.6044
Epoch [3/3], Step [619/3236], Loss: 2.2702, Perplexity: 9.6812
Epoch [3/3], Step [620/3236], Loss: 3.2203, Perplexity: 25.0368
Epoch [3/3], Step [621/3236], Loss: 1.9280, Perplexity: 6.8759
Epoch [3/3], Step [622/3236], Loss: 2.1065, Perplexity: 8.2194
Epoch [3/3], Step [623/3236], Loss: 2.2231, Perplexity: 9.2355
Epoch [3/3], Step [624/3236], Loss: 2.0855, Perplexity: 8.0485
Epoch [3/3], Step [625/3236], Loss: 2.0002, Perplexity: 7.3905
Epoch [3/3], Step [626/3236], Loss: 2.2432, Perplexity: 9.4234
Epoch [3/3], Step [627/3236], Loss: 2.3222, Perplexity: 10.1985
Epoch [3/3], Step [628/3236], Loss: 2.0190, Perplexity: 7.5309
Epoch [3/3], Step [629/3236], Loss: 2.0797, Perplexity: 8.0021
Epoch [3/3], Step [630/3236], Loss: 2.0030, Perplexity: 7.4113
Epoch [3/3], Step [631/3236], Loss: 1.9674, Perplexity: 7.1519
Epoch [3/3], Step [632/3236], Loss: 1.9850, Perplexity: 7.2794
Epoch [3/3], Step [633/3236], Loss: 2.0917, Perplexity: 8.0985
Epoch [3/3], Step [634/3236], Loss: 2.1046, Perplexity: 8.2035
Epoch [3/3], Step [635/3236], Loss: 2.0012, Perplexity: 7.3976
Epoch [3/3], Step [636/3236], Loss: 2.0190, Perplexity: 7.5309
Epoch [3/3], Step [637/3236], Loss: 2.0871, Perplexity: 8.0611
Epoch [3/3], Step [638/3236], Loss: 2.0117, Perplexity: 7.4759
Epoch [3/3], Step [639/3236], Loss: 2.1071, Perplexity: 8.2241
Epoch [3/3], Step [640/3236], Loss: 2.8327, Perplexity: 16.9919
Epoch [3/3], Step [641/3236], Loss: 2.1536, Perplexity: 8.6156
Epoch [3/3], Step [642/3236], Loss: 2.0015, Perplexity: 7.3998
Epoch [3/3], Step [643/3236], Loss: 2.1220, Perplexity: 8.3476
Epoch [3/3], Step [644/3236], Loss: 2.0684, Perplexity: 7.9120
Epoch [3/3], Step [645/3236], Loss: 2.0032, Perplexity: 7.4126
Epoch [3/3], Step [646/3236], Loss: 2.0845, Perplexity: 8.0406
Epoch [3/3], Step [647/3236], Loss: 2.1327, Perplexity: 8.4380
Epoch [3/3], Step [648/3236], Loss: 2.0700, Perplexity: 7.9249
Epoch [3/3], Step [649/3236], Loss: 2.1396, Perplexity: 8.4961
Epoch [3/3], Step [650/3236], Loss: 1.9596, Perplexity: 7.0962
Epoch [3/3], Step [651/3236], Loss: 1.9098, Perplexity: 6.7515
Epoch [3/3], Step [652/3236], Loss: 2.2847, Perplexity: 9.8227
Epoch [3/3], Step [653/3236], Loss: 1.9347, Perplexity: 6.9222
Epoch [3/3], Step [654/3236], Loss: 2.1032, Perplexity: 8.1927
Epoch [3/3], Step [655/3236], Loss: 2.1044, Perplexity: 8.2025
Epoch [3/3], Step [656/3236], Loss: 2.4180, Perplexity: 11.2229
Epoch [3/3], Step [657/3236], Loss: 2.1761, Perplexity: 8.8120
Epoch [3/3], Step [658/3236], Loss: 2.2746, Perplexity: 9.7242
Epoch [3/3], Step [659/3236], Loss: 2.0998, Perplexity: 8.1643
Epoch [3/3], Step [660/3236], Loss: 2.4294, Perplexity: 11.3525
Epoch [3/3], Step [661/3236], Loss: 2.0300, Perplexity: 7.6140
Epoch [3/3], Step [662/3236], Loss: 2.2448, Perplexity: 9.4389
Epoch [3/3], Step [663/3236], Loss: 1.9226, Perplexity: 6.8390
Epoch [3/3], Step [664/3236], Loss: 2.6081, Perplexity: 13.5727
Epoch [3/3], Step [665/3236], Loss: 2.0211, Perplexity: 7.5463
Epoch [3/3], Step [666/3236], Loss: 2.4351, Perplexity: 11.4165
Epoch [3/3], Step [667/3236], Loss: 2.0580, Perplexity: 7.8305
Epoch [3/3], Step [668/3236], Loss: 2.1416, Perplexity: 8.5131
Epoch [3/3], Step [669/3236], Loss: 2.2035, Perplexity: 9.0571
Epoch [3/3], Step [670/3236], Loss: 2.0781, Perplexity: 7.9892
Epoch [3/3], Step [671/3236], Loss: 2.2311, Perplexity: 9.3099
Epoch [3/3], Step [672/3236], Loss: 1.9067, Perplexity: 6.7309
Epoch [3/3], Step [673/3236], Loss: 2.1041, Perplexity: 8.1994
Epoch [3/3], Step [674/3236], Loss: 1.9479, Perplexity: 7.0138
Epoch [3/3], Step [675/3236], Loss: 2.0319, Perplexity: 7.6286
Epoch [3/3], Step [676/3236], Loss: 2.0273, Perplexity: 7.5938
Epoch [3/3], Step [677/3236], Loss: 2.0789, Perplexity: 7.9958
Epoch [3/3], Step [678/3236], Loss: 2.3133, Perplexity: 10.1074
Epoch [3/3], Step [679/3236], Loss: 2.0864, Perplexity: 8.0557
Epoch [3/3], Step [680/3236], Loss: 2.3023, Perplexity: 9.9976
Epoch [3/3], Step [681/3236], Loss: 2.0593, Perplexity: 7.8405
Epoch [3/3], Step [682/3236], Loss: 1.9999, Perplexity: 7.3884
Epoch [3/3], Step [683/3236], Loss: 1.9489, Perplexity: 7.0206
Epoch [3/3], Step [684/3236], Loss: 1.7908, Perplexity: 5.9942
Epoch [3/3], Step [685/3236], Loss: 2.0419, Perplexity: 7.7050
Epoch [3/3], Step [686/3236], Loss: 2.0662, Perplexity: 7.8944
Epoch [3/3], Step [687/3236], Loss: 2.0234, Perplexity: 7.5642
Epoch [3/3], Step [688/3236], Loss: 2.0192, Perplexity: 7.5323
Epoch [3/3], Step [689/3236], Loss: 1.9324, Perplexity: 6.9061
Epoch [3/3], Step [690/3236], Loss: 2.0571, Perplexity: 7.8233
Epoch [3/3], Step [691/3236], Loss: 2.2062, Perplexity: 9.0814
Epoch [3/3], Step [692/3236], Loss: 2.0517, Perplexity: 7.7811
Epoch [3/3], Step [693/3236], Loss: 2.1166, Perplexity: 8.3030
Epoch [3/3], Step [694/3236], Loss: 2.1023, Perplexity: 8.1851
Epoch [3/3], Step [695/3236], Loss: 2.4182, Perplexity: 11.2253
Epoch [3/3], Step [696/3236], Loss: 2.5295, Perplexity: 12.5468
Epoch [3/3], Step [697/3236], Loss: 2.2362, Perplexity: 9.3573
Epoch [3/3], Step [698/3236], Loss: 1.9146, Perplexity: 6.7842
Epoch [3/3], Step [699/3236], Loss: 1.9518, Perplexity: 7.0411
Epoch [3/3], Step [700/3236], Loss: 2.2991, Perplexity: 9.9653
Epoch [3/3], Step [701/3236], Loss: 2.0155, Perplexity: 7.5042
Epoch [3/3], Step [702/3236], Loss: 1.9936, Perplexity: 7.3418
Epoch [3/3], Step [703/3236], Loss: 2.2652, Perplexity: 9.6334
Epoch [3/3], Step [704/3236], Loss: 2.2178, Perplexity: 9.1875
Epoch [3/3], Step [705/3236], Loss: 2.0950, Perplexity: 8.1256
Epoch [3/3], Step [706/3236], Loss: 2.5475, Perplexity: 12.7755
Epoch [3/3], Step [707/3236], Loss: 2.0411, Perplexity: 7.6992
Epoch [3/3], Step [708/3236], Loss: 2.0861, Perplexity: 8.0533
Epoch [3/3], Step [709/3236], Loss: 2.2826, Perplexity: 9.8018
Epoch [3/3], Step [710/3236], Loss: 2.1175, Perplexity: 8.3105
Epoch [3/3], Step [711/3236], Loss: 2.2212, Perplexity: 9.2183
Epoch [3/3], Step [712/3236], Loss: 2.4278, Perplexity: 11.3334
Epoch [3/3], Step [713/3236], Loss: 2.0725, Perplexity: 7.9446
Epoch [3/3], Step [714/3236], Loss: 2.2155, Perplexity: 9.1662
Epoch [3/3], Step [715/3236], Loss: 2.8450, Perplexity: 17.2009
Epoch [3/3], Step [716/3236], Loss: 2.1131, Perplexity: 8.2738
Epoch [3/3], Step [717/3236], Loss: 2.0826, Perplexity: 8.0255
Epoch [3/3], Step [718/3236], Loss: 2.0232, Perplexity: 7.5627
Epoch [3/3], Step [719/3236], Loss: 2.0860, Perplexity: 8.0530
Epoch [3/3], Step [720/3236], Loss: 2.0908, Perplexity: 8.0911
Epoch [3/3], Step [721/3236], Loss: 2.1820, Perplexity: 8.8643
Epoch [3/3], Step [722/3236], Loss: 2.2996, Perplexity: 9.9700
Epoch [3/3], Step [723/3236], Loss: 2.0323, Perplexity: 7.6316
Epoch [3/3], Step [724/3236], Loss: 2.0235, Perplexity: 7.5650
Epoch [3/3], Step [725/3236], Loss: 1.9432, Perplexity: 6.9811
Epoch [3/3], Step [726/3236], Loss: 2.0204, Perplexity: 7.5411
Epoch [3/3], Step [727/3236], Loss: 2.1682, Perplexity: 8.7429
Epoch [3/3], Step [728/3236], Loss: 2.1161, Perplexity: 8.2987
Epoch [3/3], Step [729/3236], Loss: 2.2182, Perplexity: 9.1911
Epoch [3/3], Step [730/3236], Loss: 2.3074, Perplexity: 10.0481
Epoch [3/3], Step [731/3236], Loss: 2.2236, Perplexity: 9.2407
Epoch [3/3], Step [732/3236], Loss: 2.0457, Perplexity: 7.7343
Epoch [3/3], Step [733/3236], Loss: 3.1256, Perplexity: 22.7737
Epoch [3/3], Step [734/3236], Loss: 2.0013, Perplexity: 7.3984
Epoch [3/3], Step [735/3236], Loss: 2.0303, Perplexity: 7.6164
Epoch [3/3], Step [736/3236], Loss: 2.1542, Perplexity: 8.6208
Epoch [3/3], Step [737/3236], Loss: 2.3704, Perplexity: 10.7016
Epoch [3/3], Step [738/3236], Loss: 1.9402, Perplexity: 6.9599
Epoch [3/3], Step [739/3236], Loss: 2.0979, Perplexity: 8.1489
Epoch [3/3], Step [740/3236], Loss: 2.0399, Perplexity: 7.6897
Epoch [3/3], Step [741/3236], Loss: 2.0320, Perplexity: 7.6293
Epoch [3/3], Step [742/3236], Loss: 2.0775, Perplexity: 7.9841
Epoch [3/3], Step [743/3236], Loss: 2.0697, Perplexity: 7.9223
Epoch [3/3], Step [744/3236], Loss: 2.1226, Perplexity: 8.3529
Epoch [3/3], Step [745/3236], Loss: 1.9832, Perplexity: 7.2658
Epoch [3/3], Step [746/3236], Loss: 2.1825, Perplexity: 8.8684
Epoch [3/3], Step [747/3236], Loss: 2.3302, Perplexity: 10.2805
Epoch [3/3], Step [748/3236], Loss: 2.0883, Perplexity: 8.0715
Epoch [3/3], Step [749/3236], Loss: 2.3546, Perplexity: 10.5343
Epoch [3/3], Step [750/3236], Loss: 2.1085, Perplexity: 8.2358
Epoch [3/3], Step [751/3236], Loss: 2.0393, Perplexity: 7.6850
Epoch [3/3], Step [752/3236], Loss: 2.0908, Perplexity: 8.0917
Epoch [3/3], Step [753/3236], Loss: 2.3379, Perplexity: 10.3593
Epoch [3/3], Step [754/3236], Loss: 1.9584, Perplexity: 7.0879
Epoch [3/3], Step [755/3236], Loss: 2.0592, Perplexity: 7.8400
Epoch [3/3], Step [756/3236], Loss: 2.0235, Perplexity: 7.5648
Epoch [3/3], Step [757/3236], Loss: 1.9201, Perplexity: 6.8215
Epoch [3/3], Step [758/3236], Loss: 2.1330, Perplexity: 8.4400
Epoch [3/3], Step [759/3236], Loss: 2.0515, Perplexity: 7.7794
Epoch [3/3], Step [760/3236], Loss: 2.0768, Perplexity: 7.9793
Epoch [3/3], Step [761/3236], Loss: 1.9261, Perplexity: 6.8628
Epoch [3/3], Step [762/3236], Loss: 2.0282, Perplexity: 7.6007
Epoch [3/3], Step [763/3236], Loss: 2.0772, Perplexity: 7.9824
Epoch [3/3], Step [764/3236], Loss: 2.0472, Perplexity: 7.7464
Epoch [3/3], Step [765/3236], Loss: 2.3905, Perplexity: 10.9189
Epoch [3/3], Step [766/3236], Loss: 2.1172, Perplexity: 8.3082
Epoch [3/3], Step [767/3236], Loss: 2.2258, Perplexity: 9.2610
Epoch [3/3], Step [768/3236], Loss: 2.6381, Perplexity: 13.9873
Epoch [3/3], Step [769/3236], Loss: 2.0212, Perplexity: 7.5477
Epoch [3/3], Step [770/3236], Loss: 1.8651, Perplexity: 6.4564
Epoch [3/3], Step [771/3236], Loss: 1.9689, Perplexity: 7.1627
Epoch [3/3], Step [772/3236], Loss: 2.0817, Perplexity: 8.0181
Epoch [3/3], Step [773/3236], Loss: 1.9276, Perplexity: 6.8728
Epoch [3/3], Step [774/3236], Loss: 1.9630, Perplexity: 7.1209
Epoch [3/3], Step [775/3236], Loss: 2.5989, Perplexity: 13.4491
Epoch [3/3], Step [776/3236], Loss: 2.0699, Perplexity: 7.9237
Epoch [3/3], Step [777/3236], Loss: 2.0452, Perplexity: 7.7307
Epoch [3/3], Step [778/3236], Loss: 2.0624, Perplexity: 7.8645
Epoch [3/3], Step [779/3236], Loss: 1.9993, Perplexity: 7.3839
Epoch [3/3], Step [780/3236], Loss: 2.1700, Perplexity: 8.7586
Epoch [3/3], Step [781/3236], Loss: 2.0322, Perplexity: 7.6305
Epoch [3/3], Step [782/3236], Loss: 2.0967, Perplexity: 8.1389
Epoch [3/3], Step [783/3236], Loss: 2.0006, Perplexity: 7.3935
Epoch [3/3], Step [784/3236], Loss: 2.1787, Perplexity: 8.8346
Epoch [3/3], Step [785/3236], Loss: 2.0513, Perplexity: 7.7783
Epoch [3/3], Step [786/3236], Loss: 2.0569, Perplexity: 7.8218
Epoch [3/3], Step [787/3236], Loss: 2.0835, Perplexity: 8.0322
Epoch [3/3], Step [788/3236], Loss: 2.0102, Perplexity: 7.4651
Epoch [3/3], Step [789/3236], Loss: 2.1441, Perplexity: 8.5344
Epoch [3/3], Step [790/3236], Loss: 1.9366, Perplexity: 6.9352
Epoch [3/3], Step [791/3236], Loss: 2.0071, Perplexity: 7.4416
Epoch [3/3], Step [792/3236], Loss: 2.1125, Perplexity: 8.2690
Epoch [3/3], Step [793/3236], Loss: 2.8119, Perplexity: 16.6412
Epoch [3/3], Step [794/3236], Loss: 1.9300, Perplexity: 6.8895
Epoch [3/3], Step [795/3236], Loss: 1.9393, Perplexity: 6.9540
Epoch [3/3], Step [796/3236], Loss: 2.1515, Perplexity: 8.5974
Epoch [3/3], Step [797/3236], Loss: 2.0501, Perplexity: 7.7690
Epoch [3/3], Step [798/3236], Loss: 2.0798, Perplexity: 8.0027
Epoch [3/3], Step [799/3236], Loss: 1.8947, Perplexity: 6.6502
Epoch [3/3], Step [800/3236], Loss: 2.0241, Perplexity: 7.5690
Epoch [3/3], Step [801/3236], Loss: 2.0953, Perplexity: 8.1282
Epoch [3/3], Step [802/3236], Loss: 2.4275, Perplexity: 11.3303
Epoch [3/3], Step [803/3236], Loss: 1.9510, Perplexity: 7.0357
Epoch [3/3], Step [804/3236], Loss: 2.0505, Perplexity: 7.7719
Epoch [3/3], Step [805/3236], Loss: 2.6090, Perplexity: 13.5855
Epoch [3/3], Step [806/3236], Loss: 2.0888, Perplexity: 8.0752
Epoch [3/3], Step [807/3236], Loss: 1.9941, Perplexity: 7.3455
Epoch [3/3], Step [808/3236], Loss: 2.3321, Perplexity: 10.2999
Epoch [3/3], Step [809/3236], Loss: 2.1493, Perplexity: 8.5792
Epoch [3/3], Step [810/3236], Loss: 1.9381, Perplexity: 6.9453
Epoch [3/3], Step [811/3236], Loss: 1.8766, Perplexity: 6.5310
Epoch [3/3], Step [812/3236], Loss: 2.3480, Perplexity: 10.4643
Epoch [3/3], Step [813/3236], Loss: 2.2748, Perplexity: 9.7257
Epoch [3/3], Step [814/3236], Loss: 1.9964, Perplexity: 7.3624
Epoch [3/3], Step [815/3236], Loss: 1.9894, Perplexity: 7.3115
Epoch [3/3], Step [816/3236], Loss: 1.9665, Perplexity: 7.1453
Epoch [3/3], Step [817/3236], Loss: 1.9967, Perplexity: 7.3650
Epoch [3/3], Step [818/3236], Loss: 2.3463, Perplexity: 10.4470
Epoch [3/3], Step [819/3236], Loss: 1.9006, Perplexity: 6.6899
Epoch [3/3], Step [820/3236], Loss: 1.9947, Perplexity: 7.3499
Epoch [3/3], Step [821/3236], Loss: 2.1829, Perplexity: 8.8717
Epoch [3/3], Step [822/3236], Loss: 2.7133, Perplexity: 15.0784
Epoch [3/3], Step [823/3236], Loss: 2.0774, Perplexity: 7.9835
Epoch [3/3], Step [824/3236], Loss: 2.1694, Perplexity: 8.7531
Epoch [3/3], Step [825/3236], Loss: 1.9693, Perplexity: 7.1656
Epoch [3/3], Step [826/3236], Loss: 1.9756, Perplexity: 7.2109
Epoch [3/3], Step [827/3236], Loss: 2.2568, Perplexity: 9.5528
Epoch [3/3], Step [828/3236], Loss: 1.9621, Perplexity: 7.1144
Epoch [3/3], Step [829/3236], Loss: 2.2765, Perplexity: 9.7428
Epoch [3/3], Step [830/3236], Loss: 2.2169, Perplexity: 9.1792
Epoch [3/3], Step [831/3236], Loss: 2.0204, Perplexity: 7.5416
Epoch [3/3], Step [832/3236], Loss: 1.9580, Perplexity: 7.0851
Epoch [3/3], Step [833/3236], Loss: 2.2141, Perplexity: 9.1530
Epoch [3/3], Step [834/3236], Loss: 2.1602, Perplexity: 8.6727
Epoch [3/3], Step [835/3236], Loss: 3.1656, Perplexity: 23.7023
Epoch [3/3], Step [836/3236], Loss: 2.0181, Perplexity: 7.5242
Epoch [3/3], Step [837/3236], Loss: 2.3335, Perplexity: 10.3144
Epoch [3/3], Step [838/3236], Loss: 1.9996, Perplexity: 7.3863
Epoch [3/3], Step [839/3236], Loss: 2.0662, Perplexity: 7.8947
Epoch [3/3], Step [840/3236], Loss: 1.9585, Perplexity: 7.0886
Epoch [3/3], Step [841/3236], Loss: 2.1087, Perplexity: 8.2373
Epoch [3/3], Step [842/3236], Loss: 2.1610, Perplexity: 8.6796
Epoch [3/3], Step [843/3236], Loss: 2.1038, Perplexity: 8.1969
Epoch [3/3], Step [844/3236], Loss: 2.0117, Perplexity: 7.4757
Epoch [3/3], Step [845/3236], Loss: 1.8006, Perplexity: 6.0534
Epoch [3/3], Step [846/3236], Loss: 2.5077, Perplexity: 12.2764
Epoch [3/3], Step [847/3236], Loss: 1.9361, Perplexity: 6.9318
Epoch [3/3], Step [848/3236], Loss: 2.4096, Perplexity: 11.1295
Epoch [3/3], Step [849/3236], Loss: 2.1160, Perplexity: 8.2983
Epoch [3/3], Step [850/3236], Loss: 1.9465, Perplexity: 7.0041
Epoch [3/3], Step [851/3236], Loss: 2.1126, Perplexity: 8.2695
Epoch [3/3], Step [852/3236], Loss: 1.9712, Perplexity: 7.1791
Epoch [3/3], Step [853/3236], Loss: 2.0810, Perplexity: 8.0127
Epoch [3/3], Step [854/3236], Loss: 2.6405, Perplexity: 14.0197
Epoch [3/3], Step [855/3236], Loss: 2.5371, Perplexity: 12.6427
Epoch [3/3], Step [856/3236], Loss: 2.6856, Perplexity: 14.6672
Epoch [3/3], Step [857/3236], Loss: 1.8823, Perplexity: 6.5689
Epoch [3/3], Step [858/3236], Loss: 2.1518, Perplexity: 8.6006
Epoch [3/3], Step [859/3236], Loss: 2.2460, Perplexity: 9.4503
Epoch [3/3], Step [860/3236], Loss: 1.9903, Perplexity: 7.3178
Epoch [3/3], Step [861/3236], Loss: 1.9636, Perplexity: 7.1246
Epoch [3/3], Step [862/3236], Loss: 1.9834, Perplexity: 7.2671
Epoch [3/3], Step [863/3236], Loss: 2.0761, Perplexity: 7.9730
Epoch [3/3], Step [864/3236], Loss: 2.0781, Perplexity: 7.9892
Epoch [3/3], Step [865/3236], Loss: 2.0710, Perplexity: 7.9325
Epoch [3/3], Step [866/3236], Loss: 1.9638, Perplexity: 7.1262
Epoch [3/3], Step [867/3236], Loss: 1.9808, Perplexity: 7.2488
Epoch [3/3], Step [868/3236], Loss: 2.1102, Perplexity: 8.2502
Epoch [3/3], Step [869/3236], Loss: 2.0650, Perplexity: 7.8853
Epoch [3/3], Step [870/3236], Loss: 2.0610, Perplexity: 7.8535
Epoch [3/3], Step [871/3236], Loss: 2.1040, Perplexity: 8.1990
Epoch [3/3], Step [872/3236], Loss: 2.5900, Perplexity: 13.3302
Epoch [3/3], Step [873/3236], Loss: 1.9918, Perplexity: 7.3289
Epoch [3/3], Step [874/3236], Loss: 2.1042, Perplexity: 8.2004
Epoch [3/3], Step [875/3236], Loss: 2.0151, Perplexity: 7.5012
Epoch [3/3], Step [876/3236], Loss: 2.1376, Perplexity: 8.4791
Epoch [3/3], Step [877/3236], Loss: 2.9706, Perplexity: 19.5031
Epoch [3/3], Step [878/3236], Loss: 2.4297, Perplexity: 11.3556
Epoch [3/3], Step [879/3236], Loss: 2.0003, Perplexity: 7.3914
Epoch [3/3], Step [880/3236], Loss: 2.0696, Perplexity: 7.9216
Epoch [3/3], Step [881/3236], Loss: 2.2148, Perplexity: 9.1598
Epoch [3/3], Step [882/3236], Loss: 2.0744, Perplexity: 7.9596
Epoch [3/3], Step [883/3236], Loss: 2.3350, Perplexity: 10.3295
Epoch [3/3], Step [884/3236], Loss: 2.1674, Perplexity: 8.7355
Epoch [3/3], Step [885/3236], Loss: 1.9410, Perplexity: 6.9656
Epoch [3/3], Step [886/3236], Loss: 2.1471, Perplexity: 8.5597
Epoch [3/3], Step [887/3236], Loss: 2.0686, Perplexity: 7.9135
Epoch [3/3], Step [888/3236], Loss: 1.9446, Perplexity: 6.9909
Epoch [3/3], Step [889/3236], Loss: 1.9596, Perplexity: 7.0963
Epoch [3/3], Step [890/3236], Loss: 1.8732, Perplexity: 6.5089
Epoch [3/3], Step [891/3236], Loss: 2.0334, Perplexity: 7.6397
Epoch [3/3], Step [892/3236], Loss: 2.1056, Perplexity: 8.2119
Epoch [3/3], Step [893/3236], Loss: 2.0332, Perplexity: 7.6382
Epoch [3/3], Step [894/3236], Loss: 2.2030, Perplexity: 9.0521
Epoch [3/3], Step [895/3236], Loss: 1.9477, Perplexity: 7.0123
Epoch [3/3], Step [896/3236], Loss: 2.5639, Perplexity: 12.9863
Epoch [3/3], Step [897/3236], Loss: 2.1442, Perplexity: 8.5349
Epoch [3/3], Step [898/3236], Loss: 2.0645, Perplexity: 7.8813
Epoch [3/3], Step [899/3236], Loss: 2.3551, Perplexity: 10.5389
Epoch [3/3], Step [900/3236], Loss: 2.0550, Perplexity: 7.8070
Epoch [3/3], Step [901/3236], Loss: 2.3941, Perplexity: 10.9583
Epoch [3/3], Step [902/3236], Loss: 2.0369, Perplexity: 7.6667
Epoch [3/3], Step [903/3236], Loss: 2.0179, Perplexity: 7.5228
Epoch [3/3], Step [904/3236], Loss: 1.9389, Perplexity: 6.9513
Epoch [3/3], Step [905/3236], Loss: 2.0583, Perplexity: 7.8325
Epoch [3/3], Step [906/3236], Loss: 2.0222, Perplexity: 7.5547
Epoch [3/3], Step [907/3236], Loss: 2.8373, Perplexity: 17.0693
Epoch [3/3], Step [908/3236], Loss: 1.9063, Perplexity: 6.7284
Epoch [3/3], Step [909/3236], Loss: 1.9073, Perplexity: 6.7348
Epoch [3/3], Step [910/3236], Loss: 2.6277, Perplexity: 13.8417
Epoch [3/3], Step [911/3236], Loss: 2.0378, Perplexity: 7.6733
Epoch [3/3], Step [912/3236], Loss: 1.8578, Perplexity: 6.4095
Epoch [3/3], Step [913/3236], Loss: 2.5652, Perplexity: 13.0029
Epoch [3/3], Step [914/3236], Loss: 2.0998, Perplexity: 8.1645
Epoch [3/3], Step [915/3236], Loss: 2.1836, Perplexity: 8.8782
Epoch [3/3], Step [916/3236], Loss: 2.0254, Perplexity: 7.5794
Epoch [3/3], Step [917/3236], Loss: 2.2719, Perplexity: 9.6978
Epoch [3/3], Step [918/3236], Loss: 1.9324, Perplexity: 6.9062
Epoch [3/3], Step [919/3236], Loss: 2.2314, Perplexity: 9.3125
Epoch [3/3], Step [920/3236], Loss: 2.1710, Perplexity: 8.7671
Epoch [3/3], Step [921/3236], Loss: 2.2054, Perplexity: 9.0742
Epoch [3/3], Step [922/3236], Loss: 2.0436, Perplexity: 7.7186
Epoch [3/3], Step [923/3236], Loss: 2.0181, Perplexity: 7.5241
Epoch [3/3], Step [924/3236], Loss: 1.9674, Perplexity: 7.1521
Epoch [3/3], Step [925/3236], Loss: 2.3157, Perplexity: 10.1319
Epoch [3/3], Step [926/3236], Loss: 2.3710, Perplexity: 10.7083
Epoch [3/3], Step [927/3236], Loss: 2.5410, Perplexity: 12.6919
Epoch [3/3], Step [928/3236], Loss: 1.9445, Perplexity: 6.9903
Epoch [3/3], Step [929/3236], Loss: 2.1775, Perplexity: 8.8246
Epoch [3/3], Step [930/3236], Loss: 1.9817, Perplexity: 7.2553
Epoch [3/3], Step [931/3236], Loss: 2.0508, Perplexity: 7.7740
Epoch [3/3], Step [932/3236], Loss: 2.0814, Perplexity: 8.0158
Epoch [3/3], Step [933/3236], Loss: 2.3931, Perplexity: 10.9476
Epoch [3/3], Step [934/3236], Loss: 1.9542, Perplexity: 7.0583
Epoch [3/3], Step [935/3236], Loss: 2.1028, Perplexity: 8.1889
Epoch [3/3], Step [936/3236], Loss: 2.1900, Perplexity: 8.9350
Epoch [3/3], Step [937/3236], Loss: 2.1066, Perplexity: 8.2203
Epoch [3/3], Step [938/3236], Loss: 2.0097, Perplexity: 7.4612
Epoch [3/3], Step [939/3236], Loss: 2.0738, Perplexity: 7.9551
Epoch [3/3], Step [940/3236], Loss: 2.0970, Perplexity: 8.1413
Epoch [3/3], Step [941/3236], Loss: 2.3543, Perplexity: 10.5313
Epoch [3/3], Step [942/3236], Loss: 2.0125, Perplexity: 7.4824
Epoch [3/3], Step [943/3236], Loss: 2.0259, Perplexity: 7.5829
Epoch [3/3], Step [944/3236], Loss: 1.8830, Perplexity: 6.5731
Epoch [3/3], Step [945/3236], Loss: 1.9878, Perplexity: 7.2994
Epoch [3/3], Step [946/3236], Loss: 1.9817, Perplexity: 7.2551
Epoch [3/3], Step [947/3236], Loss: 2.0657, Perplexity: 7.8910
Epoch [3/3], Step [948/3236], Loss: 1.9839, Perplexity: 7.2707
Epoch [3/3], Step [949/3236], Loss: 1.9831, Perplexity: 7.2655
Epoch [3/3], Step [950/3236], Loss: 1.9498, Perplexity: 7.0276
Epoch [3/3], Step [951/3236], Loss: 2.0667, Perplexity: 7.8988
Epoch [3/3], Step [952/3236], Loss: 2.4644, Perplexity: 11.7562
Epoch [3/3], Step [953/3236], Loss: 2.1332, Perplexity: 8.4417
Epoch [3/3], Step [954/3236], Loss: 1.9476, Perplexity: 7.0115
Epoch [3/3], Step [955/3236], Loss: 2.0625, Perplexity: 7.8658
Epoch [3/3], Step [956/3236], Loss: 1.9890, Perplexity: 7.3082
Epoch [3/3], Step [957/3236], Loss: 1.9987, Perplexity: 7.3798
Epoch [3/3], Step [958/3236], Loss: 2.0571, Perplexity: 7.8230
Epoch [3/3], Step [959/3236], Loss: 2.0381, Perplexity: 7.6762
Epoch [3/3], Step [960/3236], Loss: 1.9411, Perplexity: 6.9665
Epoch [3/3], Step [961/3236], Loss: 2.1220, Perplexity: 8.3479
Epoch [3/3], Step [962/3236], Loss: 2.2560, Perplexity: 9.5452
Epoch [3/3], Step [963/3236], Loss: 1.8497, Perplexity: 6.3579
Epoch [3/3], Step [964/3236], Loss: 1.9711, Perplexity: 7.1785
Epoch [3/3], Step [965/3236], Loss: 2.3471, Perplexity: 10.4556
Epoch [3/3], Step [966/3236], Loss: 2.1119, Perplexity: 8.2641
Epoch [3/3], Step [967/3236], Loss: 2.2667, Perplexity: 9.6478
Epoch [3/3], Step [968/3236], Loss: 1.9478, Perplexity: 7.0134
Epoch [3/3], Step [969/3236], Loss: 2.0168, Perplexity: 7.5140
Epoch [3/3], Step [970/3236], Loss: 2.0097, Perplexity: 7.4611
Epoch [3/3], Step [971/3236], Loss: 2.0123, Perplexity: 7.4809
Epoch [3/3], Step [972/3236], Loss: 1.9732, Perplexity: 7.1935
Epoch [3/3], Step [973/3236], Loss: 2.0786, Perplexity: 7.9934
Epoch [3/3], Step [974/3236], Loss: 2.0356, Perplexity: 7.6570
Epoch [3/3], Step [975/3236], Loss: 2.2292, Perplexity: 9.2923
Epoch [3/3], Step [976/3236], Loss: 2.0955, Perplexity: 8.1293
Epoch [3/3], Step [977/3236], Loss: 2.0564, Perplexity: 7.8176
Epoch [3/3], Step [978/3236], Loss: 2.1730, Perplexity: 8.7845
Epoch [3/3], Step [979/3236], Loss: 2.3427, Perplexity: 10.4095
Epoch [3/3], Step [980/3236], Loss: 2.0665, Perplexity: 7.8971
Epoch [3/3], Step [981/3236], Loss: 2.0060, Perplexity: 7.4337
Epoch [3/3], Step [982/3236], Loss: 1.9233, Perplexity: 6.8438
Epoch [3/3], Step [983/3236], Loss: 1.9090, Perplexity: 6.7462
Epoch [3/3], Step [984/3236], Loss: 2.3760, Perplexity: 10.7619
Epoch [3/3], Step [985/3236], Loss: 2.0456, Perplexity: 7.7342
Epoch [3/3], Step [986/3236], Loss: 2.2388, Perplexity: 9.3823
Epoch [3/3], Step [987/3236], Loss: 2.2835, Perplexity: 9.8109
Epoch [3/3], Step [988/3236], Loss: 2.1307, Perplexity: 8.4209
Epoch [3/3], Step [989/3236], Loss: 2.0460, Perplexity: 7.7371
Epoch [3/3], Step [990/3236], Loss: 2.0037, Perplexity: 7.4167
Epoch [3/3], Step [991/3236], Loss: 1.9776, Perplexity: 7.2254
Epoch [3/3], Step [992/3236], Loss: 1.9972, Perplexity: 7.3687
Epoch [3/3], Step [993/3236], Loss: 2.0032, Perplexity: 7.4125
Epoch [3/3], Step [994/3236], Loss: 1.9752, Perplexity: 7.2078
Epoch [3/3], Step [995/3236], Loss: 2.2327, Perplexity: 9.3254
Epoch [3/3], Step [996/3236], Loss: 1.9938, Perplexity: 7.3431
Epoch [3/3], Step [997/3236], Loss: 1.8831, Perplexity: 6.5740
Epoch [3/3], Step [998/3236], Loss: 2.0970, Perplexity: 8.1413
Epoch [3/3], Step [999/3236], Loss: 2.1399, Perplexity: 8.4983
Epoch [3/3], Step [1000/3236], Loss: 2.1597, Perplexity: 8.6686
Epoch [3/3], Step [1001/3236], Loss: 1.9722, Perplexity: 7.1863
Epoch [3/3], Step [1002/3236], Loss: 2.0807, Perplexity: 8.0103
Epoch [3/3], Step [1003/3236], Loss: 2.5845, Perplexity: 13.2565
Epoch [3/3], Step [1004/3236], Loss: 2.1425, Perplexity: 8.5209
Epoch [3/3], Step [1005/3236], Loss: 2.1929, Perplexity: 8.9616
Epoch [3/3], Step [1006/3236], Loss: 2.1240, Perplexity: 8.3648
Epoch [3/3], Step [1007/3236], Loss: 1.9944, Perplexity: 7.3481
Epoch [3/3], Step [1008/3236], Loss: 1.9179, Perplexity: 6.8066
Epoch [3/3], Step [1009/3236], Loss: 2.1126, Perplexity: 8.2694
Epoch [3/3], Step [1010/3236], Loss: 2.5079, Perplexity: 12.2787
Epoch [3/3], Step [1011/3236], Loss: 2.1287, Perplexity: 8.4035
Epoch [3/3], Step [1012/3236], Loss: 1.8782, Perplexity: 6.5415
Epoch [3/3], Step [1013/3236], Loss: 2.0086, Perplexity: 7.4531
Epoch [3/3], Step [1014/3236], Loss: 2.0029, Perplexity: 7.4103
Epoch [3/3], Step [1015/3236], Loss: 2.1877, Perplexity: 8.9145
Epoch [3/3], Step [1016/3236], Loss: 2.0704, Perplexity: 7.9279
Epoch [3/3], Step [1017/3236], Loss: 2.0698, Perplexity: 7.9230
Epoch [3/3], Step [1018/3236], Loss: 2.0437, Perplexity: 7.7193
Epoch [3/3], Step [1019/3236], Loss: 1.9428, Perplexity: 6.9786
Epoch [3/3], Step [1020/3236], Loss: 2.0845, Perplexity: 8.0410
Epoch [3/3], Step [1021/3236], Loss: 2.1044, Perplexity: 8.2021
Epoch [3/3], Step [1022/3236], Loss: 2.1335, Perplexity: 8.4442
Epoch [3/3], Step [1023/3236], Loss: 2.4357, Perplexity: 11.4242
Epoch [3/3], Step [1024/3236], Loss: 1.9912, Perplexity: 7.3244
Epoch [3/3], Step [1025/3236], Loss: 2.0556, Perplexity: 7.8115
Epoch [3/3], Step [1026/3236], Loss: 1.9841, Perplexity: 7.2723
Epoch [3/3], Step [1027/3236], Loss: 1.9731, Perplexity: 7.1927
Epoch [3/3], Step [1028/3236], Loss: 3.4645, Perplexity: 31.9614
Epoch [3/3], Step [1029/3236], Loss: 2.0254, Perplexity: 7.5790
Epoch [3/3], Step [1030/3236], Loss: 2.4325, Perplexity: 11.3877
Epoch [3/3], Step [1031/3236], Loss: 2.0949, Perplexity: 8.1245
Epoch [3/3], Step [1032/3236], Loss: 2.1140, Perplexity: 8.2812
Epoch [3/3], Step [1033/3236], Loss: 2.0698, Perplexity: 7.9233
Epoch [3/3], Step [1034/3236], Loss: 1.9978, Perplexity: 7.3730
Epoch [3/3], Step [1035/3236], Loss: 2.1718, Perplexity: 8.7740
Epoch [3/3], Step [1036/3236], Loss: 2.0010, Perplexity: 7.3965
Epoch [3/3], Step [1037/3236], Loss: 1.9814, Perplexity: 7.2531
Epoch [3/3], Step [1038/3236], Loss: 2.1256, Perplexity: 8.3783
Epoch [3/3], Step [1039/3236], Loss: 1.8894, Perplexity: 6.6152
Epoch [3/3], Step [1040/3236], Loss: 2.0416, Perplexity: 7.7028
Epoch [3/3], Step [1041/3236], Loss: 1.9333, Perplexity: 6.9124
Epoch [3/3], Step [1042/3236], Loss: 2.0233, Perplexity: 7.5635
Epoch [3/3], Step [1043/3236], Loss: 2.5029, Perplexity: 12.2181
Epoch [3/3], Step [1044/3236], Loss: 2.0519, Perplexity: 7.7824
Epoch [3/3], Step [1045/3236], Loss: 2.2073, Perplexity: 9.0907
Epoch [3/3], Step [1046/3236], Loss: 1.8879, Perplexity: 6.6052
Epoch [3/3], Step [1047/3236], Loss: 1.9417, Perplexity: 6.9705
Epoch [3/3], Step [1048/3236], Loss: 1.9741, Perplexity: 7.2000
Epoch [3/3], Step [1049/3236], Loss: 1.9470, Perplexity: 7.0078
Epoch [3/3], Step [1050/3236], Loss: 2.0653, Perplexity: 7.8874
Epoch [3/3], Step [1051/3236], Loss: 1.9946, Perplexity: 7.3489
Epoch [3/3], Step [1052/3236], Loss: 2.2828, Perplexity: 9.8038
Epoch [3/3], Step [1053/3236], Loss: 2.0453, Perplexity: 7.7317
Epoch [3/3], Step [1054/3236], Loss: 1.9525, Perplexity: 7.0466
Epoch [3/3], Step [1055/3236], Loss: 2.2083, Perplexity: 9.1006
Epoch [3/3], Step [1056/3236], Loss: 1.8768, Perplexity: 6.5324
Epoch [3/3], Step [1057/3236], Loss: 2.0617, Perplexity: 7.8590
Epoch [3/3], Step [1058/3236], Loss: 2.3226, Perplexity: 10.2023
Epoch [3/3], Step [1059/3236], Loss: 2.2442, Perplexity: 9.4324
Epoch [3/3], Step [1060/3236], Loss: 1.9676, Perplexity: 7.1535
Epoch [3/3], Step [1061/3236], Loss: 2.0479, Perplexity: 7.7516
Epoch [3/3], Step [1062/3236], Loss: 2.1477, Perplexity: 8.5654
Epoch [3/3], Step [1063/3236], Loss: 2.7920, Perplexity: 16.3136
Epoch [3/3], Step [1064/3236], Loss: 1.8503, Perplexity: 6.3620
Epoch [3/3], Step [1065/3236], Loss: 2.8475, Perplexity: 17.2441
Epoch [3/3], Step [1066/3236], Loss: 2.0089, Perplexity: 7.4549
Epoch [3/3], Step [1067/3236], Loss: 1.9237, Perplexity: 6.8461
Epoch [3/3], Step [1068/3236], Loss: 1.9660, Perplexity: 7.1423
Epoch [3/3], Step [1069/3236], Loss: 2.0941, Perplexity: 8.1183
Epoch [3/3], Step [1070/3236], Loss: 2.0949, Perplexity: 8.1243
Epoch [3/3], Step [1071/3236], Loss: 2.2319, Perplexity: 9.3178
Epoch [3/3], Step [1072/3236], Loss: 2.0235, Perplexity: 7.5644
Epoch [3/3], Step [1073/3236], Loss: 2.1907, Perplexity: 8.9416
Epoch [3/3], Step [1074/3236], Loss: 2.1317, Perplexity: 8.4292
Epoch [3/3], Step [1075/3236], Loss: 2.0927, Perplexity: 8.1066
Epoch [3/3], Step [1076/3236], Loss: 2.0958, Perplexity: 8.1320
Epoch [3/3], Step [1077/3236], Loss: 2.0493, Perplexity: 7.7621
Epoch [3/3], Step [1078/3236], Loss: 2.1392, Perplexity: 8.4925
Epoch [3/3], Step [1079/3236], Loss: 2.0437, Perplexity: 7.7193
Epoch [3/3], Step [1080/3236], Loss: 2.3113, Perplexity: 10.0880
Epoch [3/3], Step [1081/3236], Loss: 2.4587, Perplexity: 11.6900
Epoch [3/3], Step [1082/3236], Loss: 2.0804, Perplexity: 8.0073
Epoch [3/3], Step [1083/3236], Loss: 2.0988, Perplexity: 8.1566
Epoch [3/3], Step [1084/3236], Loss: 1.9371, Perplexity: 6.9388
Epoch [3/3], Step [1085/3236], Loss: 2.3757, Perplexity: 10.7587
Epoch [3/3], Step [1086/3236], Loss: 2.1899, Perplexity: 8.9339
Epoch [3/3], Step [1087/3236], Loss: 2.1211, Perplexity: 8.3403
Epoch [3/3], Step [1088/3236], Loss: 2.0785, Perplexity: 7.9927
Epoch [3/3], Step [1089/3236], Loss: 2.0121, Perplexity: 7.4790
Epoch [3/3], Step [1090/3236], Loss: 2.0040, Perplexity: 7.4184
Epoch [3/3], Step [1091/3236], Loss: 1.9994, Perplexity: 7.3843
Epoch [3/3], Step [1092/3236], Loss: 2.3426, Perplexity: 10.4087
Epoch [3/3], Step [1093/3236], Loss: 1.9282, Perplexity: 6.8770
Epoch [3/3], Step [1094/3236], Loss: 1.9071, Perplexity: 6.7335
Epoch [3/3], Step [1095/3236], Loss: 2.1140, Perplexity: 8.2816
Epoch [3/3], Step [1096/3236], Loss: 2.0180, Perplexity: 7.5235
Epoch [3/3], Step [1097/3236], Loss: 2.0824, Perplexity: 8.0239
Epoch [3/3], Step [1098/3236], Loss: 1.9343, Perplexity: 6.9191
Epoch [3/3], Step [1099/3236], Loss: 1.9866, Perplexity: 7.2905
Epoch [3/3], Step [1100/3236], Loss: 2.5204, Perplexity: 12.4330
Epoch [3/3], Step [1101/3236], Loss: 1.9927, Perplexity: 7.3355
Epoch [3/3], Step [1102/3236], Loss: 2.1343, Perplexity: 8.4510
Epoch [3/3], Step [1103/3236], Loss: 2.0500, Perplexity: 7.7682
Epoch [3/3], Step [1104/3236], Loss: 1.9261, Perplexity: 6.8626
Epoch [3/3], Step [1105/3236], Loss: 2.2922, Perplexity: 9.8962
Epoch [3/3], Step [1106/3236], Loss: 2.2039, Perplexity: 9.0600
Epoch [3/3], Step [1107/3236], Loss: 2.5688, Perplexity: 13.0507
Epoch [3/3], Step [1108/3236], Loss: 2.0972, Perplexity: 8.1434
Epoch [3/3], Step [1109/3236], Loss: 2.1954, Perplexity: 8.9836
Epoch [3/3], Step [1110/3236], Loss: 2.2628, Perplexity: 9.6097
Epoch [3/3], Step [1111/3236], Loss: 2.1589, Perplexity: 8.6620
Epoch [3/3], Step [1112/3236], Loss: 2.7065, Perplexity: 14.9769
Epoch [3/3], Step [1113/3236], Loss: 2.0336, Perplexity: 7.6414
Epoch [3/3], Step [1114/3236], Loss: 2.0536, Perplexity: 7.7961
Epoch [3/3], Step [1115/3236], Loss: 2.0582, Perplexity: 7.8316
Epoch [3/3], Step [1116/3236], Loss: 2.0758, Perplexity: 7.9707
Epoch [3/3], Step [1117/3236], Loss: 2.5063, Perplexity: 12.2592
Epoch [3/3], Step [1118/3236], Loss: 2.2527, Perplexity: 9.5134
Epoch [3/3], Step [1119/3236], Loss: 2.0263, Perplexity: 7.5858
Epoch [3/3], Step [1120/3236], Loss: 2.4953, Perplexity: 12.1258
Epoch [3/3], Step [1121/3236], Loss: 1.9423, Perplexity: 6.9747
Epoch [3/3], Step [1122/3236], Loss: 1.9921, Perplexity: 7.3312
Epoch [3/3], Step [1123/3236], Loss: 2.0578, Perplexity: 7.8286
Epoch [3/3], Step [1124/3236], Loss: 2.4195, Perplexity: 11.2405
Epoch [3/3], Step [1125/3236], Loss: 2.0670, Perplexity: 7.9014
Epoch [3/3], Step [1126/3236], Loss: 2.1858, Perplexity: 8.8981
Epoch [3/3], Step [1127/3236], Loss: 2.0955, Perplexity: 8.1294
Epoch [3/3], Step [1128/3236], Loss: 2.1031, Perplexity: 8.1918
Epoch [3/3], Step [1129/3236], Loss: 2.0667, Perplexity: 7.8983
Epoch [3/3], Step [1130/3236], Loss: 2.1025, Perplexity: 8.1867
Epoch [3/3], Step [1131/3236], Loss: 1.9414, Perplexity: 6.9682
Epoch [3/3], Step [1132/3236], Loss: 2.1068, Perplexity: 8.2216
Epoch [3/3], Step [1133/3236], Loss: 2.0345, Perplexity: 7.6484
Epoch [3/3], Step [1134/3236], Loss: 2.3010, Perplexity: 9.9842
Epoch [3/3], Step [1135/3236], Loss: 2.0848, Perplexity: 8.0426
Epoch [3/3], Step [1136/3236], Loss: 2.0475, Perplexity: 7.7488
Epoch [3/3], Step [1137/3236], Loss: 2.0322, Perplexity: 7.6309
Epoch [3/3], Step [1138/3236], Loss: 2.0095, Perplexity: 7.4593
Epoch [3/3], Step [1139/3236], Loss: 1.9652, Perplexity: 7.1363
Epoch [3/3], Step [1140/3236], Loss: 1.9928, Perplexity: 7.3360
Epoch [3/3], Step [1141/3236], Loss: 2.3571, Perplexity: 10.5606
Epoch [3/3], Step [1142/3236], Loss: 2.0893, Perplexity: 8.0789
Epoch [3/3], Step [1143/3236], Loss: 2.2803, Perplexity: 9.7800
Epoch [3/3], Step [1144/3236], Loss: 1.9758, Perplexity: 7.2122
Epoch [3/3], Step [1145/3236], Loss: 2.2150, Perplexity: 9.1613
Epoch [3/3], Step [1146/3236], Loss: 2.0218, Perplexity: 7.5517
Epoch [3/3], Step [1147/3236], Loss: 1.9729, Perplexity: 7.1915
Epoch [3/3], Step [1148/3236], Loss: 1.9799, Perplexity: 7.2420
Epoch [3/3], Step [1149/3236], Loss: 1.8930, Perplexity: 6.6392
Epoch [3/3], Step [1150/3236], Loss: 2.4142, Perplexity: 11.1811
Epoch [3/3], Step [1151/3236], Loss: 2.0878, Perplexity: 8.0672
Epoch [3/3], Step [1152/3236], Loss: 2.1972, Perplexity: 8.9998
Epoch [3/3], Step [1153/3236], Loss: 1.9842, Perplexity: 7.2734
Epoch [3/3], Step [1154/3236], Loss: 2.0158, Perplexity: 7.5065
Epoch [3/3], Step [1155/3236], Loss: 2.1922, Perplexity: 8.9546
Epoch [3/3], Step [1156/3236], Loss: 2.1970, Perplexity: 8.9976
Epoch [3/3], Step [1157/3236], Loss: 1.9280, Perplexity: 6.8760
Epoch [3/3], Step [1158/3236], Loss: 2.2807, Perplexity: 9.7834
Epoch [3/3], Step [1159/3236], Loss: 2.2044, Perplexity: 9.0649
Epoch [3/3], Step [1160/3236], Loss: 2.0593, Perplexity: 7.8409
Epoch [3/3], Step [1161/3236], Loss: 2.4004, Perplexity: 11.0280
Epoch [3/3], Step [1162/3236], Loss: 2.1411, Perplexity: 8.5091
Epoch [3/3], Step [1163/3236], Loss: 2.2830, Perplexity: 9.8064
Epoch [3/3], Step [1164/3236], Loss: 1.9952, Perplexity: 7.3540
Epoch [3/3], Step [1165/3236], Loss: 2.6174, Perplexity: 13.6998
Epoch [3/3], Step [1166/3236], Loss: 2.0563, Perplexity: 7.8173
Epoch [3/3], Step [1167/3236], Loss: 1.9666, Perplexity: 7.1460
Epoch [3/3], Step [1168/3236], Loss: 1.9773, Perplexity: 7.2234
Epoch [3/3], Step [1169/3236], Loss: 1.9832, Perplexity: 7.2657
Epoch [3/3], Step [1170/3236], Loss: 1.9741, Perplexity: 7.2002
Epoch [3/3], Step [1171/3236], Loss: 2.0372, Perplexity: 7.6688
Epoch [3/3], Step [1172/3236], Loss: 1.9857, Perplexity: 7.2840
Epoch [3/3], Step [1173/3236], Loss: 2.0100, Perplexity: 7.4635
Epoch [3/3], Step [1174/3236], Loss: 1.8959, Perplexity: 6.6588
Epoch [3/3], Step [1175/3236], Loss: 2.1622, Perplexity: 8.6906
Epoch [3/3], Step [1176/3236], Loss: 2.2010, Perplexity: 9.0343
Epoch [3/3], Step [1177/3236], Loss: 2.0949, Perplexity: 8.1246
Epoch [3/3], Step [1178/3236], Loss: 2.0810, Perplexity: 8.0124
Epoch [3/3], Step [1179/3236], Loss: 2.2041, Perplexity: 9.0624
Epoch [3/3], Step [1180/3236], Loss: 1.9116, Perplexity: 6.7640
Epoch [3/3], Step [1181/3236], Loss: 2.2886, Perplexity: 9.8610
Epoch [3/3], Step [1182/3236], Loss: 2.0812, Perplexity: 8.0137
Epoch [3/3], Step [1183/3236], Loss: 1.9772, Perplexity: 7.2228
Epoch [3/3], Step [1184/3236], Loss: 2.2575, Perplexity: 9.5589
Epoch [3/3], Step [1185/3236], Loss: 2.0093, Perplexity: 7.4578
Epoch [3/3], Step [1186/3236], Loss: 2.8748, Perplexity: 17.7212
Epoch [3/3], Step [1187/3236], Loss: 2.0605, Perplexity: 7.8497
Epoch [3/3], Step [1188/3236], Loss: 2.1123, Perplexity: 8.2671
Epoch [3/3], Step [1189/3236], Loss: 2.0315, Perplexity: 7.6252
Epoch [3/3], Step [1190/3236], Loss: 1.9814, Perplexity: 7.2527
Epoch [3/3], Step [1191/3236], Loss: 2.1575, Perplexity: 8.6495
Epoch [3/3], Step [1192/3236], Loss: 2.1463, Perplexity: 8.5534
Epoch [3/3], Step [1193/3236], Loss: 2.1305, Perplexity: 8.4192
Epoch [3/3], Step [1194/3236], Loss: 2.1084, Perplexity: 8.2350
Epoch [3/3], Step [1195/3236], Loss: 2.1102, Perplexity: 8.2499
Epoch [3/3], Step [1196/3236], Loss: 2.4254, Perplexity: 11.3069
Epoch [3/3], Step [1197/3236], Loss: 2.3195, Perplexity: 10.1709
Epoch [3/3], Step [1198/3236], Loss: 2.0530, Perplexity: 7.7915
Epoch [3/3], Step [1199/3236], Loss: 2.0552, Perplexity: 7.8082
Epoch [3/3], Step [1200/3236], Loss: 2.1104, Perplexity: 8.2515
Epoch [3/3], Step [1201/3236], Loss: 2.3094, Perplexity: 10.0680
Epoch [3/3], Step [1202/3236], Loss: 2.2966, Perplexity: 9.9402
Epoch [3/3], Step [1203/3236], Loss: 2.2702, Perplexity: 9.6811
Epoch [3/3], Step [1204/3236], Loss: 2.0565, Perplexity: 7.8182
Epoch [3/3], Step [1205/3236], Loss: 2.2464, Perplexity: 9.4535
Epoch [3/3], Step [1206/3236], Loss: 1.9436, Perplexity: 6.9841
Epoch [3/3], Step [1207/3236], Loss: 2.0366, Perplexity: 7.6644
Epoch [3/3], Step [1208/3236], Loss: 1.8989, Perplexity: 6.6784
Epoch [3/3], Step [1209/3236], Loss: 2.0771, Perplexity: 7.9815
Epoch [3/3], Step [1210/3236], Loss: 1.9555, Perplexity: 7.0677
Epoch [3/3], Step [1211/3236], Loss: 2.1024, Perplexity: 8.1860
Epoch [3/3], Step [1212/3236], Loss: 2.0781, Perplexity: 7.9894
Epoch [3/3], Step [1213/3236], Loss: 2.4891, Perplexity: 12.0506
Epoch [3/3], Step [1214/3236], Loss: 2.0890, Perplexity: 8.0770
Epoch [3/3], Step [1215/3236], Loss: 2.0210, Perplexity: 7.5462
Epoch [3/3], Step [1216/3236], Loss: 2.1507, Perplexity: 8.5909
Epoch [3/3], Step [1217/3236], Loss: 2.7843, Perplexity: 16.1883
Epoch [3/3], Step [1218/3236], Loss: 2.2456, Perplexity: 9.4457
Epoch [3/3], Step [1219/3236], Loss: 1.9090, Perplexity: 6.7462
Epoch [3/3], Step [1220/3236], Loss: 1.9508, Perplexity: 7.0342
Epoch [3/3], Step [1221/3236], Loss: 1.9425, Perplexity: 6.9765
Epoch [3/3], Step [1222/3236], Loss: 2.0780, Perplexity: 7.9888
Epoch [3/3], Step [1223/3236], Loss: 2.0353, Perplexity: 7.6543
Epoch [3/3], Step [1224/3236], Loss: 2.0495, Perplexity: 7.7641
Epoch [3/3], Step [1225/3236], Loss: 1.9865, Perplexity: 7.2900
Epoch [3/3], Step [1226/3236], Loss: 2.0287, Perplexity: 7.6042
Epoch [3/3], Step [1227/3236], Loss: 2.5231, Perplexity: 12.4672
Epoch [3/3], Step [1228/3236], Loss: 2.0551, Perplexity: 7.8072
Epoch [3/3], Step [1229/3236], Loss: 1.9293, Perplexity: 6.8847
Epoch [3/3], Step [1230/3236], Loss: 2.2093, Perplexity: 9.1098
Epoch [3/3], Step [1231/3236], Loss: 2.0092, Perplexity: 7.4576
Epoch [3/3], Step [1232/3236], Loss: 1.9784, Perplexity: 7.2313
Epoch [3/3], Step [1233/3236], Loss: 1.9566, Perplexity: 7.0750
Epoch [3/3], Step [1234/3236], Loss: 2.1499, Perplexity: 8.5844
Epoch [3/3], Step [1235/3236], Loss: 2.3561, Perplexity: 10.5493
Epoch [3/3], Step [1236/3236], Loss: 2.1331, Perplexity: 8.4407
Epoch [3/3], Step [1237/3236], Loss: 1.9917, Perplexity: 7.3282
Epoch [3/3], Step [1238/3236], Loss: 2.1972, Perplexity: 8.9996
Epoch [3/3], Step [1239/3236], Loss: 1.9886, Perplexity: 7.3056
Epoch [3/3], Step [1240/3236], Loss: 2.0696, Perplexity: 7.9219
Epoch [3/3], Step [1241/3236], Loss: 2.0610, Perplexity: 7.8541
Epoch [3/3], Step [1242/3236], Loss: 2.0775, Perplexity: 7.9841
Epoch [3/3], Step [1243/3236], Loss: 2.0416, Perplexity: 7.7028
Epoch [3/3], Step [1244/3236], Loss: 2.0776, Perplexity: 7.9856
Epoch [3/3], Step [1245/3236], Loss: 2.3031, Perplexity: 10.0049
Epoch [3/3], Step [1246/3236], Loss: 1.9870, Perplexity: 7.2935
Epoch [3/3], Step [1247/3236], Loss: 1.9255, Perplexity: 6.8584
Epoch [3/3], Step [1248/3236], Loss: 2.2275, Perplexity: 9.2770
Epoch [3/3], Step [1249/3236], Loss: 2.0322, Perplexity: 7.6310
Epoch [3/3], Step [1250/3236], Loss: 1.8209, Perplexity: 6.1773
Epoch [3/3], Step [1251/3236], Loss: 2.1363, Perplexity: 8.4676
Epoch [3/3], Step [1252/3236], Loss: 2.0959, Perplexity: 8.1328
Epoch [3/3], Step [1253/3236], Loss: 2.0831, Perplexity: 8.0295
Epoch [3/3], Step [1254/3236], Loss: 1.9700, Perplexity: 7.1708
Epoch [3/3], Step [1255/3236], Loss: 2.3567, Perplexity: 10.5560
Epoch [3/3], Step [1256/3236], Loss: 2.8817, Perplexity: 17.8450
Epoch [3/3], Step [1257/3236], Loss: 2.2602, Perplexity: 9.5850
Epoch [3/3], Step [1258/3236], Loss: 2.1109, Perplexity: 8.2555
Epoch [3/3], Step [1259/3236], Loss: 2.0342, Perplexity: 7.6464
Epoch [3/3], Step [1260/3236], Loss: 1.9007, Perplexity: 6.6907
Epoch [3/3], Step [1261/3236], Loss: 2.1418, Perplexity: 8.5148
Epoch [3/3], Step [1262/3236], Loss: 1.9434, Perplexity: 6.9822
Epoch [3/3], Step [1263/3236], Loss: 2.0798, Perplexity: 8.0030
Epoch [3/3], Step [1264/3236], Loss: 2.8242, Perplexity: 16.8469
Epoch [3/3], Step [1265/3236], Loss: 1.9758, Perplexity: 7.2124
Epoch [3/3], Step [1266/3236], Loss: 1.9492, Perplexity: 7.0228
Epoch [3/3], Step [1267/3236], Loss: 2.1334, Perplexity: 8.4434
Epoch [3/3], Step [1268/3236], Loss: 1.9788, Perplexity: 7.2341
Epoch [3/3], Step [1269/3236], Loss: 2.0104, Perplexity: 7.4666
Epoch [3/3], Step [1270/3236], Loss: 1.9417, Perplexity: 6.9703
Epoch [3/3], Step [1271/3236], Loss: 2.0800, Perplexity: 8.0047
Epoch [3/3], Step [1272/3236], Loss: 2.0793, Perplexity: 7.9989
Epoch [3/3], Step [1273/3236], Loss: 2.1719, Perplexity: 8.7747
Epoch [3/3], Step [1274/3236], Loss: 2.0567, Perplexity: 7.8204
Epoch [3/3], Step [1275/3236], Loss: 2.0457, Perplexity: 7.7344
Epoch [3/3], Step [1276/3236], Loss: 2.0388, Perplexity: 7.6816
Epoch [3/3], Step [1277/3236], Loss: 2.0242, Perplexity: 7.5702
Epoch [3/3], Step [1278/3236], Loss: 2.3153, Perplexity: 10.1281
Epoch [3/3], Step [1279/3236], Loss: 2.1027, Perplexity: 8.1879
Epoch [3/3], Step [1280/3236], Loss: 2.0471, Perplexity: 7.7454
Epoch [3/3], Step [1281/3236], Loss: 2.3598, Perplexity: 10.5889
Epoch [3/3], Step [1282/3236], Loss: 2.3470, Perplexity: 10.4540
Epoch [3/3], Step [1283/3236], Loss: 2.0638, Perplexity: 7.8757
Epoch [3/3], Step [1284/3236], Loss: 2.2029, Perplexity: 9.0515
Epoch [3/3], Step [1285/3236], Loss: 2.0294, Perplexity: 7.6098
Epoch [3/3], Step [1286/3236], Loss: 1.8772, Perplexity: 6.5350
Epoch [3/3], Step [1287/3236], Loss: 1.9568, Perplexity: 7.0763
Epoch [3/3], Step [1288/3236], Loss: 2.1040, Perplexity: 8.1988
Epoch [3/3], Step [1289/3236], Loss: 1.9354, Perplexity: 6.9270
Epoch [3/3], Step [1290/3236], Loss: 2.0839, Perplexity: 8.0360
Epoch [3/3], Step [1291/3236], Loss: 1.9586, Perplexity: 7.0891
Epoch [3/3], Step [1292/3236], Loss: 2.1578, Perplexity: 8.6523
Epoch [3/3], Step [1293/3236], Loss: 2.1711, Perplexity: 8.7680
Epoch [3/3], Step [1294/3236], Loss: 2.8284, Perplexity: 16.9192
Epoch [3/3], Step [1295/3236], Loss: 2.0718, Perplexity: 7.9388
Epoch [3/3], Step [1296/3236], Loss: 1.9442, Perplexity: 6.9878
Epoch [3/3], Step [1297/3236], Loss: 2.1770, Perplexity: 8.8195
Epoch [3/3], Step [1298/3236], Loss: 2.0591, Perplexity: 7.8389
Epoch [3/3], Step [1299/3236], Loss: 2.0454, Perplexity: 7.7322
Epoch [3/3], Step [1300/3236], Loss: 1.9676, Perplexity: 7.1536
Epoch [3/3], Step [1301/3236], Loss: 2.0282, Perplexity: 7.6001
Epoch [3/3], Step [1302/3236], Loss: 1.9478, Perplexity: 7.0132
Epoch [3/3], Step [1303/3236], Loss: 2.1355, Perplexity: 8.4610
Epoch [3/3], Step [1304/3236], Loss: 2.0104, Perplexity: 7.4663
Epoch [3/3], Step [1305/3236], Loss: 2.2105, Perplexity: 9.1207
Epoch [3/3], Step [1306/3236], Loss: 1.9472, Perplexity: 7.0090
Epoch [3/3], Step [1307/3236], Loss: 2.2022, Perplexity: 9.0448
Epoch [3/3], Step [1308/3236], Loss: 2.0402, Perplexity: 7.6920
Epoch [3/3], Step [1309/3236], Loss: 2.6536, Perplexity: 14.2044
Epoch [3/3], Step [1310/3236], Loss: 2.0424, Perplexity: 7.7091
Epoch [3/3], Step [1311/3236], Loss: 1.9423, Perplexity: 6.9748
Epoch [3/3], Step [1312/3236], Loss: 2.0816, Perplexity: 8.0174
Epoch [3/3], Step [1313/3236], Loss: 2.2925, Perplexity: 9.8997
Epoch [3/3], Step [1314/3236], Loss: 1.9665, Perplexity: 7.1459
Epoch [3/3], Step [1315/3236], Loss: 1.9639, Perplexity: 7.1268
Epoch [3/3], Step [1316/3236], Loss: 2.4409, Perplexity: 11.4834
Epoch [3/3], Step [1317/3236], Loss: 2.1309, Perplexity: 8.4221
Epoch [3/3], Step [1318/3236], Loss: 2.1492, Perplexity: 8.5778
Epoch [3/3], Step [1319/3236], Loss: 1.8747, Perplexity: 6.5188
Epoch [3/3], Step [1320/3236], Loss: 1.8920, Perplexity: 6.6327
Epoch [3/3], Step [1321/3236], Loss: 2.1038, Perplexity: 8.1969
Epoch [3/3], Step [1322/3236], Loss: 2.1910, Perplexity: 8.9444
Epoch [3/3], Step [1323/3236], Loss: 2.0982, Perplexity: 8.1517
Epoch [3/3], Step [1324/3236], Loss: 2.2086, Perplexity: 9.1027
Epoch [3/3], Step [1325/3236], Loss: 2.1194, Perplexity: 8.3265
Epoch [3/3], Step [1326/3236], Loss: 2.3186, Perplexity: 10.1614
Epoch [3/3], Step [1327/3236], Loss: 2.6875, Perplexity: 14.6948
Epoch [3/3], Step [1328/3236], Loss: 2.3207, Perplexity: 10.1826
Epoch [3/3], Step [1329/3236], Loss: 1.8717, Perplexity: 6.4991
Epoch [3/3], Step [1330/3236], Loss: 2.0530, Perplexity: 7.7914
Epoch [3/3], Step [1331/3236], Loss: 2.1419, Perplexity: 8.5156
Epoch [3/3], Step [1332/3236], Loss: 2.0563, Perplexity: 7.8170
Epoch [3/3], Step [1333/3236], Loss: 2.1064, Perplexity: 8.2186
Epoch [3/3], Step [1334/3236], Loss: 2.0430, Perplexity: 7.7138
Epoch [3/3], Step [1335/3236], Loss: 2.0523, Perplexity: 7.7858
Epoch [3/3], Step [1336/3236], Loss: 2.0937, Perplexity: 8.1148
Epoch [3/3], Step [1337/3236], Loss: 2.3378, Perplexity: 10.3586
Epoch [3/3], Step [1338/3236], Loss: 2.0522, Perplexity: 7.7854
Epoch [3/3], Step [1339/3236], Loss: 1.9268, Perplexity: 6.8678
Epoch [3/3], Step [1340/3236], Loss: 1.8372, Perplexity: 6.2792
Epoch [3/3], Step [1341/3236], Loss: 2.2244, Perplexity: 9.2477
Epoch [3/3], Step [1342/3236], Loss: 2.0697, Perplexity: 7.9221
Epoch [3/3], Step [1343/3236], Loss: 2.1173, Perplexity: 8.3087
Epoch [3/3], Step [1344/3236], Loss: 2.2485, Perplexity: 9.4738
Epoch [3/3], Step [1345/3236], Loss: 1.9727, Perplexity: 7.1902
Epoch [3/3], Step [1346/3236], Loss: 2.1079, Perplexity: 8.2309
Epoch [3/3], Step [1347/3236], Loss: 2.1287, Perplexity: 8.4039
Epoch [3/3], Step [1348/3236], Loss: 2.5719, Perplexity: 13.0909
Epoch [3/3], Step [1349/3236], Loss: 2.0743, Perplexity: 7.9587
Epoch [3/3], Step [1350/3236], Loss: 2.6470, Perplexity: 14.1121
Epoch [3/3], Step [1351/3236], Loss: 1.9661, Perplexity: 7.1426
Epoch [3/3], Step [1352/3236], Loss: 1.9890, Perplexity: 7.3083
Epoch [3/3], Step [1353/3236], Loss: 1.9961, Perplexity: 7.3600
Epoch [3/3], Step [1354/3236], Loss: 1.9674, Perplexity: 7.1522
Epoch [3/3], Step [1355/3236], Loss: 1.9816, Perplexity: 7.2546
Epoch [3/3], Step [1356/3236], Loss: 2.2496, Perplexity: 9.4835
Epoch [3/3], Step [1357/3236], Loss: 2.0547, Perplexity: 7.8042
Epoch [3/3], Step [1358/3236], Loss: 2.3287, Perplexity: 10.2648
Epoch [3/3], Step [1359/3236], Loss: 2.1011, Perplexity: 8.1749
Epoch [3/3], Step [1360/3236], Loss: 2.1395, Perplexity: 8.4952
Epoch [3/3], Step [1361/3236], Loss: 2.0297, Perplexity: 7.6121
Epoch [3/3], Step [1362/3236], Loss: 2.2334, Perplexity: 9.3319
Epoch [3/3], Step [1363/3236], Loss: 2.2733, Perplexity: 9.7113
Epoch [3/3], Step [1364/3236], Loss: 2.0683, Perplexity: 7.9116
Epoch [3/3], Step [1365/3236], Loss: 2.0410, Perplexity: 7.6984
Epoch [3/3], Step [1366/3236], Loss: 2.0349, Perplexity: 7.6516
Epoch [3/3], Step [1367/3236], Loss: 2.1343, Perplexity: 8.4507
Epoch [3/3], Step [1368/3236], Loss: 2.2637, Perplexity: 9.6186
Epoch [3/3], Step [1369/3236], Loss: 2.1734, Perplexity: 8.7882
Epoch [3/3], Step [1370/3236], Loss: 2.0860, Perplexity: 8.0523
Epoch [3/3], Step [1371/3236], Loss: 2.1989, Perplexity: 9.0148
Epoch [3/3], Step [1372/3236], Loss: 2.0757, Perplexity: 7.9698
Epoch [3/3], Step [1373/3236], Loss: 2.0940, Perplexity: 8.1170
Epoch [3/3], Step [1374/3236], Loss: 2.6745, Perplexity: 14.5048
Epoch [3/3], Step [1375/3236], Loss: 2.0485, Perplexity: 7.7563
Epoch [3/3], Step [1376/3236], Loss: 2.0437, Perplexity: 7.7191
Epoch [3/3], Step [1377/3236], Loss: 1.9931, Perplexity: 7.3383
Epoch [3/3], Step [1378/3236], Loss: 1.9942, Perplexity: 7.3465
Epoch [3/3], Step [1379/3236], Loss: 2.0062, Perplexity: 7.4350
Epoch [3/3], Step [1380/3236], Loss: 2.1911, Perplexity: 8.9454
Epoch [3/3], Step [1381/3236], Loss: 2.8384, Perplexity: 17.0892
Epoch [3/3], Step [1382/3236], Loss: 2.0713, Perplexity: 7.9352
Epoch [3/3], Step [1383/3236], Loss: 2.1242, Perplexity: 8.3666
Epoch [3/3], Step [1384/3236], Loss: 1.9689, Perplexity: 7.1625
Epoch [3/3], Step [1385/3236], Loss: 1.9684, Perplexity: 7.1592
Epoch [3/3], Step [1386/3236], Loss: 1.9887, Perplexity: 7.3063
Epoch [3/3], Step [1387/3236], Loss: 1.9852, Perplexity: 7.2808
Epoch [3/3], Step [1388/3236], Loss: 2.3160, Perplexity: 10.1354
Epoch [3/3], Step [1389/3236], Loss: 1.9972, Perplexity: 7.3686
Epoch [3/3], Step [1390/3236], Loss: 2.6409, Perplexity: 14.0259
Epoch [3/3], Step [1391/3236], Loss: 2.1103, Perplexity: 8.2511
Epoch [3/3], Step [1392/3236], Loss: 2.0621, Perplexity: 7.8624
Epoch [3/3], Step [1393/3236], Loss: 1.9789, Perplexity: 7.2346
Epoch [3/3], Step [1394/3236], Loss: 2.3909, Perplexity: 10.9239
Epoch [3/3], Step [1395/3236], Loss: 2.0369, Perplexity: 7.6667
Epoch [3/3], Step [1396/3236], Loss: 1.9848, Perplexity: 7.2774
Epoch [3/3], Step [1397/3236], Loss: 1.9083, Perplexity: 6.7418
Epoch [3/3], Step [1398/3236], Loss: 1.9918, Perplexity: 7.3289
Epoch [3/3], Step [1399/3236], Loss: 1.9148, Perplexity: 6.7858
Epoch [3/3], Step [1400/3236], Loss: 1.9429, Perplexity: 6.9791
Epoch [3/3], Step [1401/3236], Loss: 2.2030, Perplexity: 9.0518
Epoch [3/3], Step [1402/3236], Loss: 2.1077, Perplexity: 8.2297
Epoch [3/3], Step [1403/3236], Loss: 1.9292, Perplexity: 6.8839
Epoch [3/3], Step [1404/3236], Loss: 2.0685, Perplexity: 7.9127
Epoch [3/3], Step [1405/3236], Loss: 2.2353, Perplexity: 9.3490
Epoch [3/3], Step [1406/3236], Loss: 2.8407, Perplexity: 17.1285
Epoch [3/3], Step [1407/3236], Loss: 1.9894, Perplexity: 7.3114
Epoch [3/3], Step [1408/3236], Loss: 1.9363, Perplexity: 6.9332
Epoch [3/3], Step [1409/3236], Loss: 2.0215, Perplexity: 7.5494
Epoch [3/3], Step [1410/3236], Loss: 1.9694, Perplexity: 7.1665
Epoch [3/3], Step [1411/3236], Loss: 1.9758, Perplexity: 7.2123
Epoch [3/3], Step [1412/3236], Loss: 1.8743, Perplexity: 6.5166
Epoch [3/3], Step [1413/3236], Loss: 2.0269, Perplexity: 7.5908
Epoch [3/3], Step [1414/3236], Loss: 2.1333, Perplexity: 8.4425
Epoch [3/3], Step [1415/3236], Loss: 2.5271, Perplexity: 12.5171
Epoch [3/3], Step [1416/3236], Loss: 1.9903, Perplexity: 7.3174
Epoch [3/3], Step [1417/3236], Loss: 1.9293, Perplexity: 6.8847
Epoch [3/3], Step [1418/3236], Loss: 1.9584, Perplexity: 7.0882
Epoch [3/3], Step [1419/3236], Loss: 1.9510, Perplexity: 7.0360
Epoch [3/3], Step [1420/3236], Loss: 1.9272, Perplexity: 6.8701
Epoch [3/3], Step [1421/3236], Loss: 2.8012, Perplexity: 16.4648
Epoch [3/3], Step [1422/3236], Loss: 2.0837, Perplexity: 8.0342
Epoch [3/3], Step [1423/3236], Loss: 2.0414, Perplexity: 7.7010
Epoch [3/3], Step [1424/3236], Loss: 1.7961, Perplexity: 6.0262
Epoch [3/3], Step [1425/3236], Loss: 2.2433, Perplexity: 9.4240
Epoch [3/3], Step [1426/3236], Loss: 1.9705, Perplexity: 7.1741
Epoch [3/3], Step [1427/3236], Loss: 1.9146, Perplexity: 6.7844
Epoch [3/3], Step [1428/3236], Loss: 2.1156, Perplexity: 8.2944
Epoch [3/3], Step [1429/3236], Loss: 1.9167, Perplexity: 6.7987
Epoch [3/3], Step [1430/3236], Loss: 2.0974, Perplexity: 8.1446
Epoch [3/3], Step [1431/3236], Loss: 2.1245, Perplexity: 8.3685
Epoch [3/3], Step [1432/3236], Loss: 2.0043, Perplexity: 7.4211
Epoch [3/3], Step [1433/3236], Loss: 2.0544, Perplexity: 7.8022
Epoch [3/3], Step [1434/3236], Loss: 2.8908, Perplexity: 18.0069
Epoch [3/3], Step [1435/3236], Loss: 2.1418, Perplexity: 8.5151
Epoch [3/3], Step [1436/3236], Loss: 2.0324, Perplexity: 7.6324
Epoch [3/3], Step [1437/3236], Loss: 2.0339, Perplexity: 7.6441
Epoch [3/3], Step [1438/3236], Loss: 2.2643, Perplexity: 9.6246
Epoch [3/3], Step [1439/3236], Loss: 1.9570, Perplexity: 7.0780
Epoch [3/3], Step [1440/3236], Loss: 2.2471, Perplexity: 9.4606
Epoch [3/3], Step [1441/3236], Loss: 1.7638, Perplexity: 5.8346
Epoch [3/3], Step [1442/3236], Loss: 2.1052, Perplexity: 8.2087
Epoch [3/3], Step [1443/3236], Loss: 2.0154, Perplexity: 7.5038
Epoch [3/3], Step [1444/3236], Loss: 2.1719, Perplexity: 8.7750
Epoch [3/3], Step [1445/3236], Loss: 1.9467, Perplexity: 7.0054
Epoch [3/3], Step [1446/3236], Loss: 2.0744, Perplexity: 7.9598
Epoch [3/3], Step [1447/3236], Loss: 1.9737, Perplexity: 7.1973
Epoch [3/3], Step [1448/3236], Loss: 1.9333, Perplexity: 6.9121
Epoch [3/3], Step [1449/3236], Loss: 2.0668, Perplexity: 7.8996
Epoch [3/3], Step [1450/3236], Loss: 1.9746, Perplexity: 7.2038
Epoch [3/3], Step [1451/3236], Loss: 2.0431, Perplexity: 7.7144
Epoch [3/3], Step [1452/3236], Loss: 2.0377, Perplexity: 7.6732
Epoch [3/3], Step [1453/3236], Loss: 2.6499, Perplexity: 14.1526
Epoch [3/3], Step [1454/3236], Loss: 1.9038, Perplexity: 6.7111
Epoch [3/3], Step [1455/3236], Loss: 2.0671, Perplexity: 7.9017
Epoch [3/3], Step [1456/3236], Loss: 2.0578, Perplexity: 7.8290
Epoch [3/3], Step [1457/3236], Loss: 2.0490, Perplexity: 7.7600
Epoch [3/3], Step [1458/3236], Loss: 2.1574, Perplexity: 8.6488
Epoch [3/3], Step [1459/3236], Loss: 1.8933, Perplexity: 6.6414
Epoch [3/3], Step [1460/3236], Loss: 1.8806, Perplexity: 6.5577
Epoch [3/3], Step [1461/3236], Loss: 1.9767, Perplexity: 7.2185
Epoch [3/3], Step [1462/3236], Loss: 2.4495, Perplexity: 11.5822
Epoch [3/3], Step [1463/3236], Loss: 1.9762, Perplexity: 7.2149
Epoch [3/3], Step [1464/3236], Loss: 1.9016, Perplexity: 6.6968
Epoch [3/3], Step [1465/3236], Loss: 1.9763, Perplexity: 7.2161
Epoch [3/3], Step [1466/3236], Loss: 2.0408, Perplexity: 7.6970
Epoch [3/3], Step [1467/3236], Loss: 2.0569, Perplexity: 7.8218
Epoch [3/3], Step [1468/3236], Loss: 2.0266, Perplexity: 7.5879
Epoch [3/3], Step [1469/3236], Loss: 2.0105, Perplexity: 7.4672
Epoch [3/3], Step [1470/3236], Loss: 2.4830, Perplexity: 11.9767
Epoch [3/3], Step [1471/3236], Loss: 2.0592, Perplexity: 7.8399
Epoch [3/3], Step [1472/3236], Loss: 2.1975, Perplexity: 9.0028
Epoch [3/3], Step [1473/3236], Loss: 2.1018, Perplexity: 8.1805
Epoch [3/3], Step [1474/3236], Loss: 2.0259, Perplexity: 7.5828
Epoch [3/3], Step [1475/3236], Loss: 1.9441, Perplexity: 6.9874
Epoch [3/3], Step [1476/3236], Loss: 2.3662, Perplexity: 10.6564
Epoch [3/3], Step [1477/3236], Loss: 1.9971, Perplexity: 7.3677
Epoch [3/3], Step [1478/3236], Loss: 2.0864, Perplexity: 8.0560
Epoch [3/3], Step [1479/3236], Loss: 1.8709, Perplexity: 6.4942
Epoch [3/3], Step [1480/3236], Loss: 2.0503, Perplexity: 7.7700
Epoch [3/3], Step [1481/3236], Loss: 2.3588, Perplexity: 10.5779
Epoch [3/3], Step [1482/3236], Loss: 2.0151, Perplexity: 7.5012
Epoch [3/3], Step [1483/3236], Loss: 2.0272, Perplexity: 7.5928
Epoch [3/3], Step [1484/3236], Loss: 2.1782, Perplexity: 8.8307
Epoch [3/3], Step [1485/3236], Loss: 1.8880, Perplexity: 6.6061
Epoch [3/3], Step [1486/3236], Loss: 2.1501, Perplexity: 8.5859
Epoch [3/3], Step [1487/3236], Loss: 2.0146, Perplexity: 7.4975
Epoch [3/3], Step [1488/3236], Loss: 2.5307, Perplexity: 12.5622
Epoch [3/3], Step [1489/3236], Loss: 1.9129, Perplexity: 6.7730
Epoch [3/3], Step [1490/3236], Loss: 2.6153, Perplexity: 13.6714
Epoch [3/3], Step [1491/3236], Loss: 2.2505, Perplexity: 9.4929
Epoch [3/3], Step [1492/3236], Loss: 1.9335, Perplexity: 6.9135
Epoch [3/3], Step [1493/3236], Loss: 2.0931, Perplexity: 8.1099
Epoch [3/3], Step [1494/3236], Loss: 2.0108, Perplexity: 7.4695
Epoch [3/3], Step [1495/3236], Loss: 2.6041, Perplexity: 13.5188
Epoch [3/3], Step [1496/3236], Loss: 2.0520, Perplexity: 7.7838
Epoch [3/3], Step [1497/3236], Loss: 2.1833, Perplexity: 8.8753
Epoch [3/3], Step [1498/3236], Loss: 2.0035, Perplexity: 7.4151
Epoch [3/3], Step [1499/3236], Loss: 1.9501, Perplexity: 7.0297
Epoch [3/3], Step [1500/3236], Loss: 2.0859, Perplexity: 8.0517
Epoch [3/3], Step [1501/3236], Loss: 4.3132, Perplexity: 74.6779
Epoch [3/3], Step [1502/3236], Loss: 2.0863, Perplexity: 8.0553
Epoch [3/3], Step [1503/3236], Loss: 2.0340, Perplexity: 7.6443
Epoch [3/3], Step [1504/3236], Loss: 2.0047, Perplexity: 7.4242
Epoch [3/3], Step [1505/3236], Loss: 2.2164, Perplexity: 9.1746
Epoch [3/3], Step [1506/3236], Loss: 2.4872, Perplexity: 12.0277
Epoch [3/3], Step [1507/3236], Loss: 1.8865, Perplexity: 6.5962
Epoch [3/3], Step [1508/3236], Loss: 2.1047, Perplexity: 8.2043
Epoch [3/3], Step [1509/3236], Loss: 2.0251, Perplexity: 7.5772
Epoch [3/3], Step [1510/3236], Loss: 1.9650, Perplexity: 7.1347
Epoch [3/3], Step [1511/3236], Loss: 2.2586, Perplexity: 9.5693
Epoch [3/3], Step [1512/3236], Loss: 2.0303, Perplexity: 7.6165
Epoch [3/3], Step [1513/3236], Loss: 2.2847, Perplexity: 9.8229
Epoch [3/3], Step [1514/3236], Loss: 2.3290, Perplexity: 10.2677
Epoch [3/3], Step [1515/3236], Loss: 2.1716, Perplexity: 8.7721
Epoch [3/3], Step [1516/3236], Loss: 2.1174, Perplexity: 8.3094
Epoch [3/3], Step [1517/3236], Loss: 2.0014, Perplexity: 7.3992
Epoch [3/3], Step [1518/3236], Loss: 2.6575, Perplexity: 14.2601
Epoch [3/3], Step [1519/3236], Loss: 2.0213, Perplexity: 7.5485
Epoch [3/3], Step [1520/3236], Loss: 1.9625, Perplexity: 7.1171
Epoch [3/3], Step [1521/3236], Loss: 2.0590, Perplexity: 7.8384
Epoch [3/3], Step [1522/3236], Loss: 2.0458, Perplexity: 7.7357
Epoch [3/3], Step [1523/3236], Loss: 2.1302, Perplexity: 8.4168
Epoch [3/3], Step [1524/3236], Loss: 2.0191, Perplexity: 7.5314
Epoch [3/3], Step [1525/3236], Loss: 2.0636, Perplexity: 7.8746
Epoch [3/3], Step [1526/3236], Loss: 2.3205, Perplexity: 10.1807
Epoch [3/3], Step [1527/3236], Loss: 1.9185, Perplexity: 6.8108
Epoch [3/3], Step [1528/3236], Loss: 2.0797, Perplexity: 8.0021
Epoch [3/3], Step [1529/3236], Loss: 2.3827, Perplexity: 10.8345
Epoch [3/3], Step [1530/3236], Loss: 2.0755, Perplexity: 7.9685
Epoch [3/3], Step [1531/3236], Loss: 1.9058, Perplexity: 6.7245
Epoch [3/3], Step [1532/3236], Loss: 1.9829, Perplexity: 7.2635
Epoch [3/3], Step [1533/3236], Loss: 1.9006, Perplexity: 6.6897
Epoch [3/3], Step [1534/3236], Loss: 2.2561, Perplexity: 9.5461
Epoch [3/3], Step [1535/3236], Loss: 2.7332, Perplexity: 15.3822
Epoch [3/3], Step [1536/3236], Loss: 1.9878, Perplexity: 7.2992
Epoch [3/3], Step [1537/3236], Loss: 2.9645, Perplexity: 19.3844
Epoch [3/3], Step [1538/3236], Loss: 2.0014, Perplexity: 7.3995
Epoch [3/3], Step [1539/3236], Loss: 2.0018, Perplexity: 7.4021
Epoch [3/3], Step [1540/3236], Loss: 1.9865, Perplexity: 7.2898
Epoch [3/3], Step [1541/3236], Loss: 2.0145, Perplexity: 7.4973
Epoch [3/3], Step [1542/3236], Loss: 2.2816, Perplexity: 9.7923
Epoch [3/3], Step [1543/3236], Loss: 1.9788, Perplexity: 7.2337
Epoch [3/3], Step [1544/3236], Loss: 2.0889, Perplexity: 8.0760
Epoch [3/3], Step [1545/3236], Loss: 2.2329, Perplexity: 9.3266
Epoch [3/3], Step [1546/3236], Loss: 2.4167, Perplexity: 11.2094
Epoch [3/3], Step [1547/3236], Loss: 2.0125, Perplexity: 7.4821
Epoch [3/3], Step [1548/3236], Loss: 1.9129, Perplexity: 6.7728
Epoch [3/3], Step [1549/3236], Loss: 1.9350, Perplexity: 6.9241
Epoch [3/3], Step [1550/3236], Loss: 2.0493, Perplexity: 7.7627
Epoch [3/3], Step [1551/3236], Loss: 2.2332, Perplexity: 9.3294
Epoch [3/3], Step [1552/3236], Loss: 2.1201, Perplexity: 8.3318
Epoch [3/3], Step [1553/3236], Loss: 2.0895, Perplexity: 8.0810
Epoch [3/3], Step [1554/3236], Loss: 1.9924, Perplexity: 7.3329
Epoch [3/3], Step [1555/3236], Loss: 2.1218, Perplexity: 8.3462
Epoch [3/3], Step [1556/3236], Loss: 2.0821, Perplexity: 8.0213
Epoch [3/3], Step [1557/3236], Loss: 2.1306, Perplexity: 8.4197
Epoch [3/3], Step [1558/3236], Loss: 1.9544, Perplexity: 7.0599
Epoch [3/3], Step [1559/3236], Loss: 2.1141, Perplexity: 8.2820
Epoch [3/3], Step [1560/3236], Loss: 2.3494, Perplexity: 10.4797
Epoch [3/3], Step [1561/3236], Loss: 1.9265, Perplexity: 6.8655
Epoch [3/3], Step [1562/3236], Loss: 2.2575, Perplexity: 9.5587
Epoch [3/3], Step [1563/3236], Loss: 1.9492, Perplexity: 7.0230
Epoch [3/3], Step [1564/3236], Loss: 2.1117, Perplexity: 8.2619
Epoch [3/3], Step [1565/3236], Loss: 2.1019, Perplexity: 8.1818
Epoch [3/3], Step [1566/3236], Loss: 2.3015, Perplexity: 9.9894
Epoch [3/3], Step [1567/3236], Loss: 2.0303, Perplexity: 7.6161
Epoch [3/3], Step [1568/3236], Loss: 2.0523, Perplexity: 7.7861
Epoch [3/3], Step [1569/3236], Loss: 2.2832, Perplexity: 9.8082
Epoch [3/3], Step [1570/3236], Loss: 2.1303, Perplexity: 8.4176
Epoch [3/3], Step [1571/3236], Loss: 2.1879, Perplexity: 8.9165
Epoch [3/3], Step [1572/3236], Loss: 2.0095, Perplexity: 7.4597
Epoch [3/3], Step [1573/3236], Loss: 1.9897, Perplexity: 7.3134
Epoch [3/3], Step [1574/3236], Loss: 2.0053, Perplexity: 7.4285
Epoch [3/3], Step [1575/3236], Loss: 2.5732, Perplexity: 13.1072
Epoch [3/3], Step [1576/3236], Loss: 1.9088, Perplexity: 6.7449
Epoch [3/3], Step [1577/3236], Loss: 2.0022, Perplexity: 7.4050
Epoch [3/3], Step [1578/3236], Loss: 2.2231, Perplexity: 9.2355
Epoch [3/3], Step [1579/3236], Loss: 2.0504, Perplexity: 7.7706
Epoch [3/3], Step [1580/3236], Loss: 1.9276, Perplexity: 6.8731
Epoch [3/3], Step [1581/3236], Loss: 2.1869, Perplexity: 8.9078
Epoch [3/3], Step [1582/3236], Loss: 2.0836, Perplexity: 8.0329
Epoch [3/3], Step [1583/3236], Loss: 1.8546, Perplexity: 6.3894
Epoch [3/3], Step [1584/3236], Loss: 2.0539, Perplexity: 7.7985
Epoch [3/3], Step [1585/3236], Loss: 2.3957, Perplexity: 10.9754
Epoch [3/3], Step [1586/3236], Loss: 2.0204, Perplexity: 7.5416
Epoch [3/3], Step [1587/3236], Loss: 1.9401, Perplexity: 6.9598
Epoch [3/3], Step [1588/3236], Loss: 2.0922, Perplexity: 8.1026
Epoch [3/3], Step [1589/3236], Loss: 2.1521, Perplexity: 8.6031
Epoch [3/3], Step [1590/3236], Loss: 2.1963, Perplexity: 8.9914
Epoch [3/3], Step [1591/3236], Loss: 2.0118, Perplexity: 7.4771
Epoch [3/3], Step [1592/3236], Loss: 2.5298, Perplexity: 12.5514
Epoch [3/3], Step [1593/3236], Loss: 2.0279, Perplexity: 7.5984
Epoch [3/3], Step [1594/3236], Loss: 2.1188, Perplexity: 8.3209
Epoch [3/3], Step [1595/3236], Loss: 1.9557, Perplexity: 7.0690
Epoch [3/3], Step [1596/3236], Loss: 1.9978, Perplexity: 7.3725
Epoch [3/3], Step [1597/3236], Loss: 2.1877, Perplexity: 8.9148
Epoch [3/3], Step [1598/3236], Loss: 2.1180, Perplexity: 8.3145
Epoch [3/3], Step [1599/3236], Loss: 2.2166, Perplexity: 9.1757
Epoch [3/3], Step [1600/3236], Loss: 2.1853, Perplexity: 8.8937
Epoch [3/3], Step [1601/3236], Loss: 1.9493, Perplexity: 7.0238
Epoch [3/3], Step [1602/3236], Loss: 2.1808, Perplexity: 8.8531
Epoch [3/3], Step [1603/3236], Loss: 2.0367, Perplexity: 7.6654
Epoch [3/3], Step [1604/3236], Loss: 2.0802, Perplexity: 8.0064
Epoch [3/3], Step [1605/3236], Loss: 2.0211, Perplexity: 7.5463
Epoch [3/3], Step [1606/3236], Loss: 2.0836, Perplexity: 8.0330
Epoch [3/3], Step [1607/3236], Loss: 2.0507, Perplexity: 7.7733
Epoch [3/3], Step [1608/3236], Loss: 2.1143, Perplexity: 8.2837
Epoch [3/3], Step [1609/3236], Loss: 1.9995, Perplexity: 7.3855
Epoch [3/3], Step [1610/3236], Loss: 2.6410, Perplexity: 14.0273
Epoch [3/3], Step [1611/3236], Loss: 1.8812, Perplexity: 6.5611
Epoch [3/3], Step [1612/3236], Loss: 2.0563, Perplexity: 7.8168
Epoch [3/3], Step [1613/3236], Loss: 1.9971, Perplexity: 7.3674
Epoch [3/3], Step [1614/3236], Loss: 2.0156, Perplexity: 7.5049
Epoch [3/3], Step [1615/3236], Loss: 1.9101, Perplexity: 6.7537
Epoch [3/3], Step [1616/3236], Loss: 1.9632, Perplexity: 7.1221
Epoch [3/3], Step [1617/3236], Loss: 2.5034, Perplexity: 12.2236
Epoch [3/3], Step [1618/3236], Loss: 1.9445, Perplexity: 6.9898
Epoch [3/3], Step [1619/3236], Loss: 2.0045, Perplexity: 7.4227
Epoch [3/3], Step [1620/3236], Loss: 1.9267, Perplexity: 6.8670
Epoch [3/3], Step [1621/3236], Loss: 2.0432, Perplexity: 7.7155
Epoch [3/3], Step [1622/3236], Loss: 2.0064, Perplexity: 7.4363
Epoch [3/3], Step [1623/3236], Loss: 2.0042, Perplexity: 7.4201
Epoch [3/3], Step [1624/3236], Loss: 2.1494, Perplexity: 8.5796
Epoch [3/3], Step [1625/3236], Loss: 2.1340, Perplexity: 8.4489
Epoch [3/3], Step [1626/3236], Loss: 2.3715, Perplexity: 10.7130
Epoch [3/3], Step [1627/3236], Loss: 1.9283, Perplexity: 6.8775
Epoch [3/3], Step [1628/3236], Loss: 2.0211, Perplexity: 7.5463
Epoch [3/3], Step [1629/3236], Loss: 1.9938, Perplexity: 7.3437
Epoch [3/3], Step [1630/3236], Loss: 2.0521, Perplexity: 7.7842
Epoch [3/3], Step [1631/3236], Loss: 1.9601, Perplexity: 7.1003
Epoch [3/3], Step [1632/3236], Loss: 2.2888, Perplexity: 9.8627
Epoch [3/3], Step [1633/3236], Loss: 1.9535, Perplexity: 7.0534
Epoch [3/3], Step [1634/3236], Loss: 2.1335, Perplexity: 8.4442
Epoch [3/3], Step [1635/3236], Loss: 2.1388, Perplexity: 8.4888
Epoch [3/3], Step [1636/3236], Loss: 2.5203, Perplexity: 12.4329
Epoch [3/3], Step [1637/3236], Loss: 2.0351, Perplexity: 7.6527
Epoch [3/3], Step [1638/3236], Loss: 1.9226, Perplexity: 6.8388
Epoch [3/3], Step [1639/3236], Loss: 2.2181, Perplexity: 9.1901
Epoch [3/3], Step [1640/3236], Loss: 1.9913, Perplexity: 7.3247
Epoch [3/3], Step [1641/3236], Loss: 1.9221, Perplexity: 6.8356
Epoch [3/3], Step [1642/3236], Loss: 2.1156, Perplexity: 8.2948
Epoch [3/3], Step [1643/3236], Loss: 2.0594, Perplexity: 7.8416
Epoch [3/3], Step [1644/3236], Loss: 2.3948, Perplexity: 10.9655
Epoch [3/3], Step [1645/3236], Loss: 2.2618, Perplexity: 9.6000
Epoch [3/3], Step [1646/3236], Loss: 2.1869, Perplexity: 8.9076
Epoch [3/3], Step [1647/3236], Loss: 2.2127, Perplexity: 9.1402
Epoch [3/3], Step [1648/3236], Loss: 1.9797, Perplexity: 7.2408
Epoch [3/3], Step [1649/3236], Loss: 2.3070, Perplexity: 10.0444
Epoch [3/3], Step [1650/3236], Loss: 2.3390, Perplexity: 10.3705
Epoch [3/3], Step [1651/3236], Loss: 1.9862, Perplexity: 7.2878
Epoch [3/3], Step [1652/3236], Loss: 1.8963, Perplexity: 6.6613
Epoch [3/3], Step [1653/3236], Loss: 1.9837, Perplexity: 7.2694
Epoch [3/3], Step [1654/3236], Loss: 2.1012, Perplexity: 8.1758
Epoch [3/3], Step [1655/3236], Loss: 2.0370, Perplexity: 7.6672
Epoch [3/3], Step [1656/3236], Loss: 2.1682, Perplexity: 8.7421
Epoch [3/3], Step [1657/3236], Loss: 2.0657, Perplexity: 7.8908
Epoch [3/3], Step [1658/3236], Loss: 1.9757, Perplexity: 7.2117
Epoch [3/3], Step [1659/3236], Loss: 2.0914, Perplexity: 8.0962
Epoch [3/3], Step [1660/3236], Loss: 2.2014, Perplexity: 9.0373
Epoch [3/3], Step [1661/3236], Loss: 2.1198, Perplexity: 8.3297
Epoch [3/3], Step [1662/3236], Loss: 2.4564, Perplexity: 11.6629
Epoch [3/3], Step [1663/3236], Loss: 2.1535, Perplexity: 8.6152
Epoch [3/3], Step [1664/3236], Loss: 2.0131, Perplexity: 7.4866
Epoch [3/3], Step [1665/3236], Loss: 1.9835, Perplexity: 7.2685
Epoch [3/3], Step [1666/3236], Loss: 2.1005, Perplexity: 8.1701
Epoch [3/3], Step [1667/3236], Loss: 1.9669, Perplexity: 7.1486
Epoch [3/3], Step [1668/3236], Loss: 2.1333, Perplexity: 8.4431
Epoch [3/3], Step [1669/3236], Loss: 2.0475, Perplexity: 7.7487
Epoch [3/3], Step [1670/3236], Loss: 2.1036, Perplexity: 8.1958
Epoch [3/3], Step [1671/3236], Loss: 2.0080, Perplexity: 7.4485
Epoch [3/3], Step [1672/3236], Loss: 1.9733, Perplexity: 7.1947
Epoch [3/3], Step [1673/3236], Loss: 2.0059, Perplexity: 7.4331
Epoch [3/3], Step [1674/3236], Loss: 2.1793, Perplexity: 8.8401
Epoch [3/3], Step [1675/3236], Loss: 1.9660, Perplexity: 7.1420
Epoch [3/3], Step [1676/3236], Loss: 2.2252, Perplexity: 9.2553
Epoch [3/3], Step [1677/3236], Loss: 2.0837, Perplexity: 8.0342
Epoch [3/3], Step [1678/3236], Loss: 2.0558, Perplexity: 7.8134
Epoch [3/3], Step [1679/3236], Loss: 2.0964, Perplexity: 8.1370
Epoch [3/3], Step [1680/3236], Loss: 1.9259, Perplexity: 6.8615
Epoch [3/3], Step [1681/3236], Loss: 2.0460, Perplexity: 7.7365
Epoch [3/3], Step [1682/3236], Loss: 1.9601, Perplexity: 7.0999
Epoch [3/3], Step [1683/3236], Loss: 1.9823, Perplexity: 7.2598
Epoch [3/3], Step [1684/3236], Loss: 2.0815, Perplexity: 8.0164
Epoch [3/3], Step [1685/3236], Loss: 1.9365, Perplexity: 6.9345
Epoch [3/3], Step [1686/3236], Loss: 2.3644, Perplexity: 10.6377
Epoch [3/3], Step [1687/3236], Loss: 1.8059, Perplexity: 6.0856
Epoch [3/3], Step [1688/3236], Loss: 1.9402, Perplexity: 6.9599
Epoch [3/3], Step [1689/3236], Loss: 2.6435, Perplexity: 14.0628
Epoch [3/3], Step [1690/3236], Loss: 2.0295, Perplexity: 7.6104
Epoch [3/3], Step [1691/3236], Loss: 2.0166, Perplexity: 7.5127
Epoch [3/3], Step [1692/3236], Loss: 2.0413, Perplexity: 7.7004
Epoch [3/3], Step [1693/3236], Loss: 2.3401, Perplexity: 10.3826
Epoch [3/3], Step [1694/3236], Loss: 2.1203, Perplexity: 8.3337
Epoch [3/3], Step [1695/3236], Loss: 2.0483, Perplexity: 7.7547
Epoch [3/3], Step [1696/3236], Loss: 2.0534, Perplexity: 7.7945
Epoch [3/3], Step [1697/3236], Loss: 2.1984, Perplexity: 9.0108
Epoch [3/3], Step [1698/3236], Loss: 1.9834, Perplexity: 7.2675
Epoch [3/3], Step [1699/3236], Loss: 2.0030, Perplexity: 7.4111
Epoch [3/3], Step [1700/3236], Loss: 2.2788, Perplexity: 9.7652
Epoch [3/3], Step [1701/3236], Loss: 2.0169, Perplexity: 7.5151
Epoch [3/3], Step [1702/3236], Loss: 2.0536, Perplexity: 7.7955
Epoch [3/3], Step [1703/3236], Loss: 2.2386, Perplexity: 9.3806
Epoch [3/3], Step [1704/3236], Loss: 2.2053, Perplexity: 9.0729
Epoch [3/3], Step [1705/3236], Loss: 1.9414, Perplexity: 6.9685
Epoch [3/3], Step [1706/3236], Loss: 2.0672, Perplexity: 7.9029
Epoch [3/3], Step [1707/3236], Loss: 2.1506, Perplexity: 8.5897
Epoch [3/3], Step [1708/3236], Loss: 3.0430, Perplexity: 20.9671
Epoch [3/3], Step [1709/3236], Loss: 2.0099, Perplexity: 7.4625
Epoch [3/3], Step [1710/3236], Loss: 2.0020, Perplexity: 7.4037
Epoch [3/3], Step [1711/3236], Loss: 1.9940, Perplexity: 7.3447
Epoch [3/3], Step [1712/3236], Loss: 1.9045, Perplexity: 6.7157
Epoch [3/3], Step [1713/3236], Loss: 2.3121, Perplexity: 10.0957
Epoch [3/3], Step [1714/3236], Loss: 1.9632, Perplexity: 7.1221
Epoch [3/3], Step [1715/3236], Loss: 1.9168, Perplexity: 6.7990
Epoch [3/3], Step [1716/3236], Loss: 2.0865, Perplexity: 8.0565
Epoch [3/3], Step [1717/3236], Loss: 2.0643, Perplexity: 7.8794
Epoch [3/3], Step [1718/3236], Loss: 2.8904, Perplexity: 18.0000
Epoch [3/3], Step [1719/3236], Loss: 1.9628, Perplexity: 7.1190
Epoch [3/3], Step [1720/3236], Loss: 2.2963, Perplexity: 9.9371
Epoch [3/3], Step [1721/3236], Loss: 1.9209, Perplexity: 6.8268
Epoch [3/3], Step [1722/3236], Loss: 2.0991, Perplexity: 8.1584
Epoch [3/3], Step [1723/3236], Loss: 2.0678, Perplexity: 7.9075
Epoch [3/3], Step [1724/3236], Loss: 2.0122, Perplexity: 7.4795
Epoch [3/3], Step [1725/3236], Loss: 1.8598, Perplexity: 6.4226
Epoch [3/3], Step [1726/3236], Loss: 1.9010, Perplexity: 6.6928
Epoch [3/3], Step [1727/3236], Loss: 2.0233, Perplexity: 7.5633
Epoch [3/3], Step [1728/3236], Loss: 2.1236, Perplexity: 8.3615
Epoch [3/3], Step [1729/3236], Loss: 2.4619, Perplexity: 11.7270
Epoch [3/3], Step [1730/3236], Loss: 2.2629, Perplexity: 9.6107
Epoch [3/3], Step [1731/3236], Loss: 2.1261, Perplexity: 8.3819
Epoch [3/3], Step [1732/3236], Loss: 2.0240, Perplexity: 7.5686
Epoch [3/3], Step [1733/3236], Loss: 2.0871, Perplexity: 8.0618
Epoch [3/3], Step [1734/3236], Loss: 1.9851, Perplexity: 7.2797
Epoch [3/3], Step [1735/3236], Loss: 2.1309, Perplexity: 8.4220
Epoch [3/3], Step [1736/3236], Loss: 2.2689, Perplexity: 9.6691
Epoch [3/3], Step [1737/3236], Loss: 1.9667, Perplexity: 7.1472
Epoch [3/3], Step [1738/3236], Loss: 2.0578, Perplexity: 7.8290
Epoch [3/3], Step [1739/3236], Loss: 1.9849, Perplexity: 7.2787
Epoch [3/3], Step [1740/3236], Loss: 2.0061, Perplexity: 7.4346
Epoch [3/3], Step [1741/3236], Loss: 2.1029, Perplexity: 8.1900
Epoch [3/3], Step [1742/3236], Loss: 2.0085, Perplexity: 7.4522
Epoch [3/3], Step [1743/3236], Loss: 2.0488, Perplexity: 7.7585
Epoch [3/3], Step [1744/3236], Loss: 1.9754, Perplexity: 7.2092
Epoch [3/3], Step [1745/3236], Loss: 2.0538, Perplexity: 7.7976
Epoch [3/3], Step [1746/3236], Loss: 1.9726, Perplexity: 7.1893
Epoch [3/3], Step [1747/3236], Loss: 2.0100, Perplexity: 7.4635
Epoch [3/3], Step [1748/3236], Loss: 1.9089, Perplexity: 6.7459
Epoch [3/3], Step [1749/3236], Loss: 2.0839, Perplexity: 8.0360
Epoch [3/3], Step [1750/3236], Loss: 2.5119, Perplexity: 12.3283
Epoch [3/3], Step [1751/3236], Loss: 2.0364, Perplexity: 7.6626
Epoch [3/3], Step [1752/3236], Loss: 2.6053, Perplexity: 13.5356
Epoch [3/3], Step [1753/3236], Loss: 2.1125, Perplexity: 8.2693
Epoch [3/3], Step [1754/3236], Loss: 2.0208, Perplexity: 7.5446
Epoch [3/3], Step [1755/3236], Loss: 1.9281, Perplexity: 6.8767
Epoch [3/3], Step [1756/3236], Loss: 2.0206, Perplexity: 7.5427
Epoch [3/3], Step [1757/3236], Loss: 1.9495, Perplexity: 7.0248
Epoch [3/3], Step [1758/3236], Loss: 2.3175, Perplexity: 10.1500
Epoch [3/3], Step [1759/3236], Loss: 1.9209, Perplexity: 6.8271
Epoch [3/3], Step [1760/3236], Loss: 2.0135, Perplexity: 7.4896
Epoch [3/3], Step [1761/3236], Loss: 1.9801, Perplexity: 7.2435
Epoch [3/3], Step [1762/3236], Loss: 2.0148, Perplexity: 7.4989
Epoch [3/3], Step [1763/3236], Loss: 2.0956, Perplexity: 8.1306
Epoch [3/3], Step [1764/3236], Loss: 2.2100, Perplexity: 9.1153
Epoch [3/3], Step [1765/3236], Loss: 2.0248, Perplexity: 7.5747
Epoch [3/3], Step [1766/3236], Loss: 1.9954, Perplexity: 7.3553
Epoch [3/3], Step [1767/3236], Loss: 2.1533, Perplexity: 8.6133
Epoch [3/3], Step [1768/3236], Loss: 2.1595, Perplexity: 8.6672
Epoch [3/3], Step [1769/3236], Loss: 2.0912, Perplexity: 8.0944
Epoch [3/3], Step [1770/3236], Loss: 2.0829, Perplexity: 8.0280
Epoch [3/3], Step [1771/3236], Loss: 2.0810, Perplexity: 8.0123
Epoch [3/3], Step [1772/3236], Loss: 2.0549, Perplexity: 7.8057
Epoch [3/3], Step [1773/3236], Loss: 2.0470, Perplexity: 7.7449
Epoch [3/3], Step [1774/3236], Loss: 2.2506, Perplexity: 9.4937
Epoch [3/3], Step [1775/3236], Loss: 2.5139, Perplexity: 12.3528
Epoch [3/3], Step [1776/3236], Loss: 1.8585, Perplexity: 6.4143
Epoch [3/3], Step [1777/3236], Loss: 1.9100, Perplexity: 6.7533
Epoch [3/3], Step [1778/3236], Loss: 2.0107, Perplexity: 7.4685
Epoch [3/3], Step [1779/3236], Loss: 2.4511, Perplexity: 11.6010
Epoch [3/3], Step [1780/3236], Loss: 1.9138, Perplexity: 6.7786
Epoch [3/3], Step [1781/3236], Loss: 2.0243, Perplexity: 7.5710
Epoch [3/3], Step [1782/3236], Loss: 1.8639, Perplexity: 6.4491
Epoch [3/3], Step [1783/3236], Loss: 2.2864, Perplexity: 9.8391
Epoch [3/3], Step [1784/3236], Loss: 2.0318, Perplexity: 7.6281
Epoch [3/3], Step [1785/3236], Loss: 2.0824, Perplexity: 8.0234
Epoch [3/3], Step [1786/3236], Loss: 2.2898, Perplexity: 9.8733
Epoch [3/3], Step [1787/3236], Loss: 1.9321, Perplexity: 6.9038
Epoch [3/3], Step [1788/3236], Loss: 2.3119, Perplexity: 10.0938
Epoch [3/3], Step [1789/3236], Loss: 2.0243, Perplexity: 7.5709
Epoch [3/3], Step [1790/3236], Loss: 2.1249, Perplexity: 8.3723
Epoch [3/3], Step [1791/3236], Loss: 2.0871, Perplexity: 8.0612
Epoch [3/3], Step [1792/3236], Loss: 2.1384, Perplexity: 8.4855
Epoch [3/3], Step [1793/3236], Loss: 2.1966, Perplexity: 8.9945
Epoch [3/3], Step [1794/3236], Loss: 2.1039, Perplexity: 8.1980
Epoch [3/3], Step [1795/3236], Loss: 2.0512, Perplexity: 7.7776
Epoch [3/3], Step [1796/3236], Loss: 2.3560, Perplexity: 10.5491
Epoch [3/3], Step [1797/3236], Loss: 1.9722, Perplexity: 7.1861
Epoch [3/3], Step [1798/3236], Loss: 2.0357, Perplexity: 7.6573
Epoch [3/3], Step [1799/3236], Loss: 2.4484, Perplexity: 11.5700
Epoch [3/3], Step [1800/3236], Loss: 1.9868, Perplexity: 7.2922
Epoch [3/3], Step [1801/3236], Loss: 1.9439, Perplexity: 6.9860
Epoch [3/3], Step [1802/3236], Loss: 1.9343, Perplexity: 6.9190
Epoch [3/3], Step [1803/3236], Loss: 2.0083, Perplexity: 7.4504
Epoch [3/3], Step [1804/3236], Loss: 2.0902, Perplexity: 8.0863
Epoch [3/3], Step [1805/3236], Loss: 2.3622, Perplexity: 10.6146
Epoch [3/3], Step [1806/3236], Loss: 1.9108, Perplexity: 6.7586
Epoch [3/3], Step [1807/3236], Loss: 2.1319, Perplexity: 8.4305
Epoch [3/3], Step [1808/3236], Loss: 1.9528, Perplexity: 7.0485
Epoch [3/3], Step [1809/3236], Loss: 2.2130, Perplexity: 9.1434
Epoch [3/3], Step [1810/3236], Loss: 2.3722, Perplexity: 10.7205
Epoch [3/3], Step [1811/3236], Loss: 1.8987, Perplexity: 6.6771
Epoch [3/3], Step [1812/3236], Loss: 2.1568, Perplexity: 8.6434
Epoch [3/3], Step [1813/3236], Loss: 2.0777, Perplexity: 7.9861
Epoch [3/3], Step [1814/3236], Loss: 2.1612, Perplexity: 8.6814
Epoch [3/3], Step [1815/3236], Loss: 1.9794, Perplexity: 7.2381
Epoch [3/3], Step [1816/3236], Loss: 1.9047, Perplexity: 6.7173
Epoch [3/3], Step [1817/3236], Loss: 1.9978, Perplexity: 7.3728
Epoch [3/3], Step [1818/3236], Loss: 2.1486, Perplexity: 8.5727
Epoch [3/3], Step [1819/3236], Loss: 2.0597, Perplexity: 7.8439
Epoch [3/3], Step [1820/3236], Loss: 1.8971, Perplexity: 6.6663
Epoch [3/3], Step [1821/3236], Loss: 2.1191, Perplexity: 8.3239
Epoch [3/3], Step [1822/3236], Loss: 2.0945, Perplexity: 8.1217
Epoch [3/3], Step [1823/3236], Loss: 1.9969, Perplexity: 7.3661
Epoch [3/3], Step [1824/3236], Loss: 2.0610, Perplexity: 7.8535
Epoch [3/3], Step [1825/3236], Loss: 3.1525, Perplexity: 23.3953
Epoch [3/3], Step [1826/3236], Loss: 2.3192, Perplexity: 10.1674
Epoch [3/3], Step [1827/3236], Loss: 2.0598, Perplexity: 7.8446
Epoch [3/3], Step [1828/3236], Loss: 1.9444, Perplexity: 6.9894
Epoch [3/3], Step [1829/3236], Loss: 2.1987, Perplexity: 9.0132
Epoch [3/3], Step [1830/3236], Loss: 2.0320, Perplexity: 7.6291
Epoch [3/3], Step [1831/3236], Loss: 1.9630, Perplexity: 7.1208
Epoch [3/3], Step [1832/3236], Loss: 1.8802, Perplexity: 6.5549
Epoch [3/3], Step [1833/3236], Loss: 2.2552, Perplexity: 9.5372
Epoch [3/3], Step [1834/3236], Loss: 1.9781, Perplexity: 7.2291
Epoch [3/3], Step [1835/3236], Loss: 2.0425, Perplexity: 7.7095
Epoch [3/3], Step [1836/3236], Loss: 2.7340, Perplexity: 15.3938
Epoch [3/3], Step [1837/3236], Loss: 2.0078, Perplexity: 7.4468
Epoch [3/3], Step [1838/3236], Loss: 2.0628, Perplexity: 7.8676
Epoch [3/3], Step [1839/3236], Loss: 2.0482, Perplexity: 7.7539
Epoch [3/3], Step [1840/3236], Loss: 1.9092, Perplexity: 6.7474
Epoch [3/3], Step [1841/3236], Loss: 1.9822, Perplexity: 7.2590
Epoch [3/3], Step [1842/3236], Loss: 1.8432, Perplexity: 6.3170
Epoch [3/3], Step [1843/3236], Loss: 2.5054, Perplexity: 12.2488
Epoch [3/3], Step [1844/3236], Loss: 2.1792, Perplexity: 8.8395
Epoch [3/3], Step [1845/3236], Loss: 2.0363, Perplexity: 7.6624
Epoch [3/3], Step [1846/3236], Loss: 1.9743, Perplexity: 7.2017
Epoch [3/3], Step [1847/3236], Loss: 2.0176, Perplexity: 7.5203
Epoch [3/3], Step [1848/3236], Loss: 2.0849, Perplexity: 8.0436
Epoch [3/3], Step [1849/3236], Loss: 2.1958, Perplexity: 8.9873
Epoch [3/3], Step [1850/3236], Loss: 2.1056, Perplexity: 8.2119
Epoch [3/3], Step [1851/3236], Loss: 2.0555, Perplexity: 7.8104
Epoch [3/3], Step [1852/3236], Loss: 1.9223, Perplexity: 6.8368
Epoch [3/3], Step [1853/3236], Loss: 1.9274, Perplexity: 6.8717
Epoch [3/3], Step [1854/3236], Loss: 2.1169, Perplexity: 8.3053
Epoch [3/3], Step [1855/3236], Loss: 2.0028, Perplexity: 7.4096
Epoch [3/3], Step [1856/3236], Loss: 1.8914, Perplexity: 6.6286
Epoch [3/3], Step [1857/3236], Loss: 1.8993, Perplexity: 6.6811
Epoch [3/3], Step [1858/3236], Loss: 1.9352, Perplexity: 6.9256
Epoch [3/3], Step [1859/3236], Loss: 2.0343, Perplexity: 7.6473
Epoch [3/3], Step [1860/3236], Loss: 2.0849, Perplexity: 8.0439
Epoch [3/3], Step [1861/3236], Loss: 1.9922, Perplexity: 7.3320
Epoch [3/3], Step [1862/3236], Loss: 2.0249, Perplexity: 7.5751
Epoch [3/3], Step [1863/3236], Loss: 2.0329, Perplexity: 7.6359
Epoch [3/3], Step [1864/3236], Loss: 2.1679, Perplexity: 8.7402
Epoch [3/3], Step [1865/3236], Loss: 2.4507, Perplexity: 11.5963
Epoch [3/3], Step [1866/3236], Loss: 1.9940, Perplexity: 7.3449
Epoch [3/3], Step [1867/3236], Loss: 1.9071, Perplexity: 6.7332
Epoch [3/3], Step [1868/3236], Loss: 2.1872, Perplexity: 8.9106
Epoch [3/3], Step [1869/3236], Loss: 2.5614, Perplexity: 12.9542
Epoch [3/3], Step [1870/3236], Loss: 1.9637, Perplexity: 7.1260
Epoch [3/3], Step [1871/3236], Loss: 1.9829, Perplexity: 7.2638
Epoch [3/3], Step [1872/3236], Loss: 1.9910, Perplexity: 7.3229
Epoch [3/3], Step [1873/3236], Loss: 2.0542, Perplexity: 7.8008
Epoch [3/3], Step [1874/3236], Loss: 2.0760, Perplexity: 7.9725
Epoch [3/3], Step [1875/3236], Loss: 1.9815, Perplexity: 7.2539
Epoch [3/3], Step [1876/3236], Loss: 2.0665, Perplexity: 7.8975
Epoch [3/3], Step [1877/3236], Loss: 2.0593, Perplexity: 7.8406
Epoch [3/3], Step [1878/3236], Loss: 2.0049, Perplexity: 7.4253
Epoch [3/3], Step [1879/3236], Loss: 2.0166, Perplexity: 7.5128
Epoch [3/3], Step [1880/3236], Loss: 1.9888, Perplexity: 7.3071
Epoch [3/3], Step [1881/3236], Loss: 2.0811, Perplexity: 8.0135
Epoch [3/3], Step [1882/3236], Loss: 2.2016, Perplexity: 9.0394
Epoch [3/3], Step [1883/3236], Loss: 2.1236, Perplexity: 8.3614
Epoch [3/3], Step [1884/3236], Loss: 2.1369, Perplexity: 8.4728
Epoch [3/3], Step [1885/3236], Loss: 2.1705, Perplexity: 8.7627
Epoch [3/3], Step [1886/3236], Loss: 1.9488, Perplexity: 7.0202
Epoch [3/3], Step [1887/3236], Loss: 2.0410, Perplexity: 7.6983
Epoch [3/3], Step [1888/3236], Loss: 1.9342, Perplexity: 6.9187
Epoch [3/3], Step [1889/3236], Loss: 1.9874, Perplexity: 7.2967
Epoch [3/3], Step [1890/3236], Loss: 1.8989, Perplexity: 6.6788
Epoch [3/3], Step [1891/3236], Loss: 2.0409, Perplexity: 7.6975
Epoch [3/3], Step [1892/3236], Loss: 2.1080, Perplexity: 8.2315
Epoch [3/3], Step [1893/3236], Loss: 2.3734, Perplexity: 10.7336
Epoch [3/3], Step [1894/3236], Loss: 2.0602, Perplexity: 7.8471
Epoch [3/3], Step [1895/3236], Loss: 1.9520, Perplexity: 7.0428
Epoch [3/3], Step [1896/3236], Loss: 2.0894, Perplexity: 8.0800
Epoch [3/3], Step [1897/3236], Loss: 2.1459, Perplexity: 8.5496
Epoch [3/3], Step [1898/3236], Loss: 1.9643, Perplexity: 7.1298
Epoch [3/3], Step [1899/3236], Loss: 2.0367, Perplexity: 7.6650
Epoch [3/3], Step [1900/3236], Loss: 1.9791, Perplexity: 7.2366
Epoch [3/3], Step [1901/3236], Loss: 2.0530, Perplexity: 7.7912
Epoch [3/3], Step [1902/3236], Loss: 1.9449, Perplexity: 6.9928
Epoch [3/3], Step [1903/3236], Loss: 2.0347, Perplexity: 7.6499
Epoch [3/3], Step [1904/3236], Loss: 1.9664, Perplexity: 7.1448
Epoch [3/3], Step [1905/3236], Loss: 2.0027, Perplexity: 7.4089
Epoch [3/3], Step [1906/3236], Loss: 2.0478, Perplexity: 7.7512
Epoch [3/3], Step [1907/3236], Loss: 2.0201, Perplexity: 7.5394
Epoch [3/3], Step [1908/3236], Loss: 2.6859, Perplexity: 14.6715
Epoch [3/3], Step [1909/3236], Loss: 2.0005, Perplexity: 7.3928
Epoch [3/3], Step [1910/3236], Loss: 1.9683, Perplexity: 7.1586
Epoch [3/3], Step [1911/3236], Loss: 2.0946, Perplexity: 8.1225
Epoch [3/3], Step [1912/3236], Loss: 2.2933, Perplexity: 9.9072
Epoch [3/3], Step [1913/3236], Loss: 2.3066, Perplexity: 10.0405
Epoch [3/3], Step [1914/3236], Loss: 1.9519, Perplexity: 7.0417
Epoch [3/3], Step [1915/3236], Loss: 2.0269, Perplexity: 7.5903
Epoch [3/3], Step [1916/3236], Loss: 2.0457, Perplexity: 7.7343
Epoch [3/3], Step [1917/3236], Loss: 2.1888, Perplexity: 8.9245
Epoch [3/3], Step [1918/3236], Loss: 2.0611, Perplexity: 7.8545
Epoch [3/3], Step [1919/3236], Loss: 1.9784, Perplexity: 7.2313
Epoch [3/3], Step [1920/3236], Loss: 2.0819, Perplexity: 8.0201
Epoch [3/3], Step [1921/3236], Loss: 2.5517, Perplexity: 12.8285
Epoch [3/3], Step [1922/3236], Loss: 2.0120, Perplexity: 7.4780
Epoch [3/3], Step [1923/3236], Loss: 2.0801, Perplexity: 8.0049
Epoch [3/3], Step [1924/3236], Loss: 2.2181, Perplexity: 9.1902
Epoch [3/3], Step [1925/3236], Loss: 2.0021, Perplexity: 7.4048
Epoch [3/3], Step [1926/3236], Loss: 2.0448, Perplexity: 7.7278
Epoch [3/3], Step [1927/3236], Loss: 2.1589, Perplexity: 8.6617
Epoch [3/3], Step [1928/3236], Loss: 2.0302, Perplexity: 7.6156
Epoch [3/3], Step [1929/3236], Loss: 2.0544, Perplexity: 7.8020
Epoch [3/3], Step [1930/3236], Loss: 2.0202, Perplexity: 7.5399
Epoch [3/3], Step [1931/3236], Loss: 2.0349, Perplexity: 7.6513
Epoch [3/3], Step [1932/3236], Loss: 2.2362, Perplexity: 9.3576
Epoch [3/3], Step [1933/3236], Loss: 2.2113, Perplexity: 9.1279
Epoch [3/3], Step [1934/3236], Loss: 2.0467, Perplexity: 7.7423
Epoch [3/3], Step [1935/3236], Loss: 2.7246, Perplexity: 15.2496
Epoch [3/3], Step [1936/3236], Loss: 1.9462, Perplexity: 7.0017
Epoch [3/3], Step [1937/3236], Loss: 2.0884, Perplexity: 8.0721
Epoch [3/3], Step [1938/3236], Loss: 2.5520, Perplexity: 12.8330
Epoch [3/3], Step [1939/3236], Loss: 2.1078, Perplexity: 8.2302
Epoch [3/3], Step [1940/3236], Loss: 2.5587, Perplexity: 12.9187
Epoch [3/3], Step [1941/3236], Loss: 2.0654, Perplexity: 7.8883
Epoch [3/3], Step [1942/3236], Loss: 2.1595, Perplexity: 8.6664
Epoch [3/3], Step [1943/3236], Loss: 1.9168, Perplexity: 6.7990
Epoch [3/3], Step [1944/3236], Loss: 1.9762, Perplexity: 7.2153
Epoch [3/3], Step [1945/3236], Loss: 1.9722, Perplexity: 7.1863
Epoch [3/3], Step [1946/3236], Loss: 1.9961, Perplexity: 7.3602
Epoch [3/3], Step [1947/3236], Loss: 1.9743, Perplexity: 7.2015
Epoch [3/3], Step [1948/3236], Loss: 2.0115, Perplexity: 7.4743
Epoch [3/3], Step [1949/3236], Loss: 2.0560, Perplexity: 7.8150
Epoch [3/3], Step [1950/3236], Loss: 2.0344, Perplexity: 7.6476
Epoch [3/3], Step [1951/3236], Loss: 2.4799, Perplexity: 11.9397
Epoch [3/3], Step [1952/3236], Loss: 2.0258, Perplexity: 7.5823
Epoch [3/3], Step [1953/3236], Loss: 2.2962, Perplexity: 9.9361
Epoch [3/3], Step [1954/3236], Loss: 2.0855, Perplexity: 8.0487
Epoch [3/3], Step [1955/3236], Loss: 2.0911, Perplexity: 8.0938
Epoch [3/3], Step [1956/3236], Loss: 2.3653, Perplexity: 10.6469
Epoch [3/3], Step [1957/3236], Loss: 2.2960, Perplexity: 9.9346
Epoch [3/3], Step [1958/3236], Loss: 2.4337, Perplexity: 11.4008
Epoch [3/3], Step [1959/3236], Loss: 2.0442, Perplexity: 7.7229
Epoch [3/3], Step [1960/3236], Loss: 2.1463, Perplexity: 8.5536
Epoch [3/3], Step [1961/3236], Loss: 2.1305, Perplexity: 8.4187
Epoch [3/3], Step [1962/3236], Loss: 1.9987, Perplexity: 7.3797
Epoch [3/3], Step [1963/3236], Loss: 2.3138, Perplexity: 10.1129
Epoch [3/3], Step [1964/3236], Loss: 2.5194, Perplexity: 12.4217
Epoch [3/3], Step [1965/3236], Loss: 2.1966, Perplexity: 8.9943
Epoch [3/3], Step [1966/3236], Loss: 2.0272, Perplexity: 7.5926
Epoch [3/3], Step [1967/3236], Loss: 2.0323, Perplexity: 7.6317
Epoch [3/3], Step [1968/3236], Loss: 2.0645, Perplexity: 7.8817
Epoch [3/3], Step [1969/3236], Loss: 1.9978, Perplexity: 7.3725
Epoch [3/3], Step [1970/3236], Loss: 2.1042, Perplexity: 8.2004
Epoch [3/3], Step [1971/3236], Loss: 2.7472, Perplexity: 15.5984
Epoch [3/3], Step [1972/3236], Loss: 2.1033, Perplexity: 8.1934
Epoch [3/3], Step [1973/3236], Loss: 1.9901, Perplexity: 7.3161
Epoch [3/3], Step [1974/3236], Loss: 1.9476, Perplexity: 7.0117
Epoch [3/3], Step [1975/3236], Loss: 2.0065, Perplexity: 7.4372
Epoch [3/3], Step [1976/3236], Loss: 2.0094, Perplexity: 7.4587
Epoch [3/3], Step [1977/3236], Loss: 2.0001, Perplexity: 7.3900
Epoch [3/3], Step [1978/3236], Loss: 1.9166, Perplexity: 6.7976
Epoch [3/3], Step [1979/3236], Loss: 1.9889, Perplexity: 7.3078
Epoch [3/3], Step [1980/3236], Loss: 1.9596, Perplexity: 7.0961
Epoch [3/3], Step [1981/3236], Loss: 2.2723, Perplexity: 9.7013
Epoch [3/3], Step [1982/3236], Loss: 1.9575, Perplexity: 7.0813
Epoch [3/3], Step [1983/3236], Loss: 2.0061, Perplexity: 7.4343
Epoch [3/3], Step [1984/3236], Loss: 1.9685, Perplexity: 7.1603
Epoch [3/3], Step [1985/3236], Loss: 2.0083, Perplexity: 7.4504
Epoch [3/3], Step [1986/3236], Loss: 1.9954, Perplexity: 7.3552
Epoch [3/3], Step [1987/3236], Loss: 2.2024, Perplexity: 9.0463
Epoch [3/3], Step [1988/3236], Loss: 1.9941, Perplexity: 7.3456
Epoch [3/3], Step [1989/3236], Loss: 2.0026, Perplexity: 7.4080
Epoch [3/3], Step [1990/3236], Loss: 2.0449, Perplexity: 7.7287
Epoch [3/3], Step [1991/3236], Loss: 1.9804, Perplexity: 7.2456
Epoch [3/3], Step [1992/3236], Loss: 1.9433, Perplexity: 6.9818
Epoch [3/3], Step [1993/3236], Loss: 2.0706, Perplexity: 7.9292
Epoch [3/3], Step [1994/3236], Loss: 3.6682, Perplexity: 39.1833
Epoch [3/3], Step [1995/3236], Loss: 1.9754, Perplexity: 7.2095
Epoch [3/3], Step [1996/3236], Loss: 2.4786, Perplexity: 11.9245
Epoch [3/3], Step [1997/3236], Loss: 2.0813, Perplexity: 8.0150
Epoch [3/3], Step [1998/3236], Loss: 1.9417, Perplexity: 6.9709
Epoch [3/3], Step [1999/3236], Loss: 2.0834, Perplexity: 8.0317
Epoch [3/3], Step [2000/3236], Loss: 1.9623, Perplexity: 7.1156
Epoch [3/3], Step [2001/3236], Loss: 2.0294, Perplexity: 7.6098
Epoch [3/3], Step [2002/3236], Loss: 2.0952, Perplexity: 8.1273
Epoch [3/3], Step [2003/3236], Loss: 2.0086, Perplexity: 7.4527
Epoch [3/3], Step [2004/3236], Loss: 2.0298, Perplexity: 7.6122
Epoch [3/3], Step [2005/3236], Loss: 1.9134, Perplexity: 6.7759
Epoch [3/3], Step [2006/3236], Loss: 2.0840, Perplexity: 8.0367
Epoch [3/3], Step [2007/3236], Loss: 2.0572, Perplexity: 7.8241
Epoch [3/3], Step [2008/3236], Loss: 2.0585, Perplexity: 7.8344
Epoch [3/3], Step [2009/3236], Loss: 1.9236, Perplexity: 6.8457
Epoch [3/3], Step [2010/3236], Loss: 1.9176, Perplexity: 6.8046
Epoch [3/3], Step [2011/3236], Loss: 1.9695, Perplexity: 7.1674
Epoch [3/3], Step [2012/3236], Loss: 2.4697, Perplexity: 11.8185
Epoch [3/3], Step [2013/3236], Loss: 1.9960, Perplexity: 7.3597
Epoch [3/3], Step [2014/3236], Loss: 2.2993, Perplexity: 9.9670
Epoch [3/3], Step [2015/3236], Loss: 2.5519, Perplexity: 12.8320
Epoch [3/3], Step [2016/3236], Loss: 2.0345, Perplexity: 7.6480
Epoch [3/3], Step [2017/3236], Loss: 1.9370, Perplexity: 6.9376
Epoch [3/3], Step [2018/3236], Loss: 2.0093, Perplexity: 7.4582
Epoch [3/3], Step [2019/3236], Loss: 2.0477, Perplexity: 7.7501
Epoch [3/3], Step [2020/3236], Loss: 2.0636, Perplexity: 7.8739
Epoch [3/3], Step [2021/3236], Loss: 1.9827, Perplexity: 7.2625
Epoch [3/3], Step [2022/3236], Loss: 2.0048, Perplexity: 7.4244
Epoch [3/3], Step [2023/3236], Loss: 1.9081, Perplexity: 6.7404
Epoch [3/3], Step [2024/3236], Loss: 2.0442, Perplexity: 7.7226
Epoch [3/3], Step [2025/3236], Loss: 1.9395, Perplexity: 6.9550
Epoch [3/3], Step [2026/3236], Loss: 2.0333, Perplexity: 7.6391
Epoch [3/3], Step [2027/3236], Loss: 2.2060, Perplexity: 9.0790
Epoch [3/3], Step [2028/3236], Loss: 2.0656, Perplexity: 7.8900
Epoch [3/3], Step [2029/3236], Loss: 2.1216, Perplexity: 8.3442
Epoch [3/3], Step [2030/3236], Loss: 1.9588, Perplexity: 7.0906
Epoch [3/3], Step [2031/3236], Loss: 2.2955, Perplexity: 9.9293
Epoch [3/3], Step [2032/3236], Loss: 2.0932, Perplexity: 8.1106
Epoch [3/3], Step [2033/3236], Loss: 2.1173, Perplexity: 8.3083
Epoch [3/3], Step [2034/3236], Loss: 1.9868, Perplexity: 7.2923
Epoch [3/3], Step [2035/3236], Loss: 2.0227, Perplexity: 7.5589
Epoch [3/3], Step [2036/3236], Loss: 4.0627, Perplexity: 58.1310
Epoch [3/3], Step [2037/3236], Loss: 2.0490, Perplexity: 7.7598
Epoch [3/3], Step [2038/3236], Loss: 1.9907, Perplexity: 7.3208
Epoch [3/3], Step [2039/3236], Loss: 2.0142, Perplexity: 7.4948
Epoch [3/3], Step [2040/3236], Loss: 1.9350, Perplexity: 6.9238
Epoch [3/3], Step [2041/3236], Loss: 1.9691, Perplexity: 7.1643
Epoch [3/3], Step [2042/3236], Loss: 2.0012, Perplexity: 7.3981
Epoch [3/3], Step [2043/3236], Loss: 1.9296, Perplexity: 6.8868
Epoch [3/3], Step [2044/3236], Loss: 1.9558, Perplexity: 7.0699
Epoch [3/3], Step [2045/3236], Loss: 1.8596, Perplexity: 6.4214
Epoch [3/3], Step [2046/3236], Loss: 1.9583, Perplexity: 7.0872
Epoch [3/3], Step [2047/3236], Loss: 2.4285, Perplexity: 11.3418
Epoch [3/3], Step [2048/3236], Loss: 2.1686, Perplexity: 8.7462
Epoch [3/3], Step [2049/3236], Loss: 2.0276, Perplexity: 7.5962
Epoch [3/3], Step [2050/3236], Loss: 1.9733, Perplexity: 7.1946
Epoch [3/3], Step [2051/3236], Loss: 2.1283, Perplexity: 8.4004
Epoch [3/3], Step [2052/3236], Loss: 2.5122, Perplexity: 12.3324
Epoch [3/3], Step [2053/3236], Loss: 1.9055, Perplexity: 6.7228
Epoch [3/3], Step [2054/3236], Loss: 2.0313, Perplexity: 7.6242
Epoch [3/3], Step [2055/3236], Loss: 2.0083, Perplexity: 7.4505
Epoch [3/3], Step [2056/3236], Loss: 1.9904, Perplexity: 7.3188
Epoch [3/3], Step [2057/3236], Loss: 2.0712, Perplexity: 7.9343
Epoch [3/3], Step [2058/3236], Loss: 2.0093, Perplexity: 7.4580
Epoch [3/3], Step [2059/3236], Loss: 1.9162, Perplexity: 6.7948
Epoch [3/3], Step [2060/3236], Loss: 2.0553, Perplexity: 7.8094
Epoch [3/3], Step [2061/3236], Loss: 2.0204, Perplexity: 7.5414
Epoch [3/3], Step [2062/3236], Loss: 2.2000, Perplexity: 9.0248
Epoch [3/3], Step [2063/3236], Loss: 2.1221, Perplexity: 8.3483
Epoch [3/3], Step [2064/3236], Loss: 1.9417, Perplexity: 6.9707
Epoch [3/3], Step [2065/3236], Loss: 2.3713, Perplexity: 10.7118
Epoch [3/3], Step [2066/3236], Loss: 2.5798, Perplexity: 13.1946
Epoch [3/3], Step [2067/3236], Loss: 2.0582, Perplexity: 7.8320
Epoch [3/3], Step [2068/3236], Loss: 1.9297, Perplexity: 6.8874
Epoch [3/3], Step [2069/3236], Loss: 2.0526, Perplexity: 7.7880
Epoch [3/3], Step [2070/3236], Loss: 2.7560, Perplexity: 15.7373
Epoch [3/3], Step [2071/3236], Loss: 2.1567, Perplexity: 8.6422
Epoch [3/3], Step [2072/3236], Loss: 1.9606, Perplexity: 7.1037
Epoch [3/3], Step [2073/3236], Loss: 2.1045, Perplexity: 8.2028
Epoch [3/3], Step [2074/3236], Loss: 2.4196, Perplexity: 11.2416
Epoch [3/3], Step [2075/3236], Loss: 2.2108, Perplexity: 9.1227
Epoch [3/3], Step [2076/3236], Loss: 2.0225, Perplexity: 7.5574
Epoch [3/3], Step [2077/3236], Loss: 2.1274, Perplexity: 8.3928
Epoch [3/3], Step [2078/3236], Loss: 1.9284, Perplexity: 6.8783
Epoch [3/3], Step [2079/3236], Loss: 2.1249, Perplexity: 8.3721
Epoch [3/3], Step [2080/3236], Loss: 1.9282, Perplexity: 6.8768
Epoch [3/3], Step [2081/3236], Loss: 1.9688, Perplexity: 7.1623
Epoch [3/3], Step [2082/3236], Loss: 1.9414, Perplexity: 6.9685
Epoch [3/3], Step [2083/3236], Loss: 2.1371, Perplexity: 8.4749
Epoch [3/3], Step [2084/3236], Loss: 2.0239, Perplexity: 7.5676
Epoch [3/3], Step [2085/3236], Loss: 2.0065, Perplexity: 7.4372
Epoch [3/3], Step [2086/3236], Loss: 1.8785, Perplexity: 6.5437
Epoch [3/3], Step [2087/3236], Loss: 1.8907, Perplexity: 6.6240
Epoch [3/3], Step [2088/3236], Loss: 2.0543, Perplexity: 7.8011
Epoch [3/3], Step [2089/3236], Loss: 2.0769, Perplexity: 7.9794
Epoch [3/3], Step [2090/3236], Loss: 2.3511, Perplexity: 10.4974
Epoch [3/3], Step [2091/3236], Loss: 1.9525, Perplexity: 7.0460
Epoch [3/3], Step [2092/3236], Loss: 2.2103, Perplexity: 9.1182
Epoch [3/3], Step [2093/3236], Loss: 1.9854, Perplexity: 7.2816
Epoch [3/3], Step [2094/3236], Loss: 1.8339, Perplexity: 6.2584
Epoch [3/3], Step [2095/3236], Loss: 2.0887, Perplexity: 8.0744
Epoch [3/3], Step [2096/3236], Loss: 2.3318, Perplexity: 10.2965
Epoch [3/3], Step [2097/3236], Loss: 1.8369, Perplexity: 6.2770
Epoch [3/3], Step [2098/3236], Loss: 2.1247, Perplexity: 8.3702
Epoch [3/3], Step [2099/3236], Loss: 1.8469, Perplexity: 6.3400
Epoch [3/3], Step [2100/3236], Loss: 2.0252, Perplexity: 7.5779
Epoch [3/3], Step [2101/3236], Loss: 2.1963, Perplexity: 8.9914
Epoch [3/3], Step [2102/3236], Loss: 2.4972, Perplexity: 12.1479
Epoch [3/3], Step [2103/3236], Loss: 2.0879, Perplexity: 8.0684
Epoch [3/3], Step [2104/3236], Loss: 1.9907, Perplexity: 7.3210
Epoch [3/3], Step [2105/3236], Loss: 2.0057, Perplexity: 7.4314
Epoch [3/3], Step [2106/3236], Loss: 2.0086, Perplexity: 7.4530
Epoch [3/3], Step [2107/3236], Loss: 2.0832, Perplexity: 8.0301
Epoch [3/3], Step [2108/3236], Loss: 1.9310, Perplexity: 6.8963
Epoch [3/3], Step [2109/3236], Loss: 2.0324, Perplexity: 7.6323
Epoch [3/3], Step [2110/3236], Loss: 1.9332, Perplexity: 6.9118
Epoch [3/3], Step [2111/3236], Loss: 1.8625, Perplexity: 6.4395
Epoch [3/3], Step [2112/3236], Loss: 1.8942, Perplexity: 6.6470
Epoch [3/3], Step [2113/3236], Loss: 2.2213, Perplexity: 9.2193
Epoch [3/3], Step [2114/3236], Loss: 2.0670, Perplexity: 7.9013
Epoch [3/3], Step [2115/3236], Loss: 2.1427, Perplexity: 8.5222
Epoch [3/3], Step [2116/3236], Loss: 2.1105, Perplexity: 8.2525
Epoch [3/3], Step [2117/3236], Loss: 2.1006, Perplexity: 8.1713
Epoch [3/3], Step [2118/3236], Loss: 1.9841, Perplexity: 7.2728
Epoch [3/3], Step [2119/3236], Loss: 2.0425, Perplexity: 7.7101
Epoch [3/3], Step [2120/3236], Loss: 2.0505, Perplexity: 7.7716
Epoch [3/3], Step [2121/3236], Loss: 1.9290, Perplexity: 6.8826
Epoch [3/3], Step [2122/3236], Loss: 1.9945, Perplexity: 7.3488
Epoch [3/3], Step [2123/3236], Loss: 1.9238, Perplexity: 6.8467
Epoch [3/3], Step [2124/3236], Loss: 2.0772, Perplexity: 7.9819
Epoch [3/3], Step [2125/3236], Loss: 2.2930, Perplexity: 9.9045
Epoch [3/3], Step [2126/3236], Loss: 2.3302, Perplexity: 10.2802
Epoch [3/3], Step [2127/3236], Loss: 1.9551, Perplexity: 7.0648
Epoch [3/3], Step [2128/3236], Loss: 1.8981, Perplexity: 6.6729
Epoch [3/3], Step [2129/3236], Loss: 1.8896, Perplexity: 6.6165
Epoch [3/3], Step [2130/3236], Loss: 2.1689, Perplexity: 8.7482
Epoch [3/3], Step [2131/3236], Loss: 2.2013, Perplexity: 9.0365
Epoch [3/3], Step [2132/3236], Loss: 2.0781, Perplexity: 7.9893
Epoch [3/3], Step [2133/3236], Loss: 2.7113, Perplexity: 15.0490
Epoch [3/3], Step [2134/3236], Loss: 1.9846, Perplexity: 7.2762
Epoch [3/3], Step [2135/3236], Loss: 2.0668, Perplexity: 7.8996
Epoch [3/3], Step [2136/3236], Loss: 2.0275, Perplexity: 7.5948
Epoch [3/3], Step [2137/3236], Loss: 1.9683, Perplexity: 7.1588
Epoch [3/3], Step [2138/3236], Loss: 2.1297, Perplexity: 8.4124
Epoch [3/3], Step [2139/3236], Loss: 1.9238, Perplexity: 6.8472
Epoch [3/3], Step [2140/3236], Loss: 2.0518, Perplexity: 7.7816
Epoch [3/3], Step [2141/3236], Loss: 2.0447, Perplexity: 7.7271
Epoch [3/3], Step [2142/3236], Loss: 2.0578, Perplexity: 7.8284
Epoch [3/3], Step [2143/3236], Loss: 2.0197, Perplexity: 7.5363
Epoch [3/3], Step [2144/3236], Loss: 2.2192, Perplexity: 9.2001
Epoch [3/3], Step [2145/3236], Loss: 2.4835, Perplexity: 11.9826
Epoch [3/3], Step [2146/3236], Loss: 2.2246, Perplexity: 9.2495
Epoch [3/3], Step [2147/3236], Loss: 2.4289, Perplexity: 11.3468
Epoch [3/3], Step [2148/3236], Loss: 2.0623, Perplexity: 7.8642
Epoch [3/3], Step [2149/3236], Loss: 1.9843, Perplexity: 7.2743
Epoch [3/3], Step [2150/3236], Loss: 3.0120, Perplexity: 20.3284
Epoch [3/3], Step [2151/3236], Loss: 2.1471, Perplexity: 8.5604
Epoch [3/3], Step [2152/3236], Loss: 1.8918, Perplexity: 6.6315
Epoch [3/3], Step [2153/3236], Loss: 2.1697, Perplexity: 8.7558
Epoch [3/3], Step [2154/3236], Loss: 1.9024, Perplexity: 6.7021
Epoch [3/3], Step [2155/3236], Loss: 2.1996, Perplexity: 9.0212
Epoch [3/3], Step [2156/3236], Loss: 2.0719, Perplexity: 7.9401
Epoch [3/3], Step [2157/3236], Loss: 2.0951, Perplexity: 8.1259
Epoch [3/3], Step [2158/3236], Loss: 1.8701, Perplexity: 6.4889
Epoch [3/3], Step [2159/3236], Loss: 1.9701, Perplexity: 7.1714
Epoch [3/3], Step [2160/3236], Loss: 2.0041, Perplexity: 7.4197
Epoch [3/3], Step [2161/3236], Loss: 1.9563, Perplexity: 7.0734
Epoch [3/3], Step [2162/3236], Loss: 1.9726, Perplexity: 7.1891
Epoch [3/3], Step [2163/3236], Loss: 1.9820, Perplexity: 7.2573
Epoch [3/3], Step [2164/3236], Loss: 1.9762, Perplexity: 7.2153
Epoch [3/3], Step [2165/3236], Loss: 2.2754, Perplexity: 9.7318
Epoch [3/3], Step [2166/3236], Loss: 1.9775, Perplexity: 7.2247
Epoch [3/3], Step [2167/3236], Loss: 2.0436, Perplexity: 7.7187
Epoch [3/3], Step [2168/3236], Loss: 2.0241, Perplexity: 7.5693
Epoch [3/3], Step [2169/3236], Loss: 2.1047, Perplexity: 8.2050
Epoch [3/3], Step [2170/3236], Loss: 1.9205, Perplexity: 6.8244
Epoch [3/3], Step [2171/3236], Loss: 2.0383, Perplexity: 7.6776
Epoch [3/3], Step [2172/3236], Loss: 2.0549, Perplexity: 7.8058
Epoch [3/3], Step [2173/3236], Loss: 2.1522, Perplexity: 8.6041
Epoch [3/3], Step [2174/3236], Loss: 2.0415, Perplexity: 7.7023
Epoch [3/3], Step [2175/3236], Loss: 2.2244, Perplexity: 9.2477
Epoch [3/3], Step [2176/3236], Loss: 2.0428, Perplexity: 7.7119
Epoch [3/3], Step [2177/3236], Loss: 2.2845, Perplexity: 9.8209
Epoch [3/3], Step [2178/3236], Loss: 2.0394, Perplexity: 7.6857
Epoch [3/3], Step [2179/3236], Loss: 1.9877, Perplexity: 7.2984
Epoch [3/3], Step [2180/3236], Loss: 2.0239, Perplexity: 7.5680
Epoch [3/3], Step [2181/3236], Loss: 2.2550, Perplexity: 9.5350
Epoch [3/3], Step [2182/3236], Loss: 1.9731, Perplexity: 7.1928
Epoch [3/3], Step [2183/3236], Loss: 2.0523, Perplexity: 7.7858
Epoch [3/3], Step [2184/3236], Loss: 2.1054, Perplexity: 8.2102
Epoch [3/3], Step [2185/3236], Loss: 2.0927, Perplexity: 8.1065
Epoch [3/3], Step [2186/3236], Loss: 1.9317, Perplexity: 6.9016
Epoch [3/3], Step [2187/3236], Loss: 2.3596, Perplexity: 10.5864
Epoch [3/3], Step [2188/3236], Loss: 1.9198, Perplexity: 6.8196
Epoch [3/3], Step [2189/3236], Loss: 2.0932, Perplexity: 8.1107
Epoch [3/3], Step [2190/3236], Loss: 1.9931, Perplexity: 7.3385
Epoch [3/3], Step [2191/3236], Loss: 1.9234, Perplexity: 6.8445
Epoch [3/3], Step [2192/3236], Loss: 1.9927, Perplexity: 7.3353
Epoch [3/3], Step [2193/3236], Loss: 2.1093, Perplexity: 8.2422
Epoch [3/3], Step [2194/3236], Loss: 2.3967, Perplexity: 10.9873
Epoch [3/3], Step [2195/3236], Loss: 1.8832, Perplexity: 6.5748
Epoch [3/3], Step [2196/3236], Loss: 2.0280, Perplexity: 7.5991
Epoch [3/3], Step [2197/3236], Loss: 2.1830, Perplexity: 8.8729
Epoch [3/3], Step [2198/3236], Loss: 1.9642, Perplexity: 7.1290
Epoch [3/3], Step [2199/3236], Loss: 2.4274, Perplexity: 11.3298
Epoch [3/3], Step [2200/3236], Loss: 1.9417, Perplexity: 6.9705
Epoch [3/3], Step [2201/3236], Loss: 2.7054, Perplexity: 14.9610
Epoch [3/3], Step [2202/3236], Loss: 2.0415, Perplexity: 7.7021
Epoch [3/3], Step [2203/3236], Loss: 1.9621, Perplexity: 7.1145
Epoch [3/3], Step [2204/3236], Loss: 2.0826, Perplexity: 8.0254
Epoch [3/3], Step [2205/3236], Loss: 2.0248, Perplexity: 7.5744
Epoch [3/3], Step [2206/3236], Loss: 2.2278, Perplexity: 9.2790
Epoch [3/3], Step [2207/3236], Loss: 1.9708, Perplexity: 7.1762
Epoch [3/3], Step [2208/3236], Loss: 1.9993, Perplexity: 7.3840
Epoch [3/3], Step [2209/3236], Loss: 2.0192, Perplexity: 7.5319
Epoch [3/3], Step [2210/3236], Loss: 2.2613, Perplexity: 9.5954
Epoch [3/3], Step [2211/3236], Loss: 1.9825, Perplexity: 7.2612
Epoch [3/3], Step [2212/3236], Loss: 1.9112, Perplexity: 6.7609
Epoch [3/3], Step [2213/3236], Loss: 2.0056, Perplexity: 7.4305
Epoch [3/3], Step [2214/3236], Loss: 2.0264, Perplexity: 7.5870
Epoch [3/3], Step [2215/3236], Loss: 2.0433, Perplexity: 7.7157
Epoch [3/3], Step [2216/3236], Loss: 2.0134, Perplexity: 7.4889
Epoch [3/3], Step [2217/3236], Loss: 2.2119, Perplexity: 9.1334
Epoch [3/3], Step [2218/3236], Loss: 1.9489, Perplexity: 7.0212
Epoch [3/3], Step [2219/3236], Loss: 2.1342, Perplexity: 8.4503
Epoch [3/3], Step [2220/3236], Loss: 2.1059, Perplexity: 8.2142
Epoch [3/3], Step [2221/3236], Loss: 1.9783, Perplexity: 7.2301
Epoch [3/3], Step [2222/3236], Loss: 1.9598, Perplexity: 7.0982
Epoch [3/3], Step [2223/3236], Loss: 2.0472, Perplexity: 7.7459
Epoch [3/3], Step [2224/3236], Loss: 1.8924, Perplexity: 6.6355
Epoch [3/3], Step [2225/3236], Loss: 2.0355, Perplexity: 7.6561
Epoch [3/3], Step [2226/3236], Loss: 2.0210, Perplexity: 7.5457
Epoch [3/3], Step [2227/3236], Loss: 2.2256, Perplexity: 9.2590
Epoch [3/3], Step [2228/3236], Loss: 1.9946, Perplexity: 7.3490
Epoch [3/3], Step [2229/3236], Loss: 2.0106, Perplexity: 7.4679
Epoch [3/3], Step [2230/3236], Loss: 1.9183, Perplexity: 6.8093
Epoch [3/3], Step [2231/3236], Loss: 2.2879, Perplexity: 9.8544
Epoch [3/3], Step [2232/3236], Loss: 2.0575, Perplexity: 7.8261
Epoch [3/3], Step [2233/3236], Loss: 2.4038, Perplexity: 11.0655
Epoch [3/3], Step [2234/3236], Loss: 1.9885, Perplexity: 7.3045
Epoch [3/3], Step [2235/3236], Loss: 1.8943, Perplexity: 6.6477
Epoch [3/3], Step [2236/3236], Loss: 3.6233, Perplexity: 37.4625
Epoch [3/3], Step [2237/3236], Loss: 2.0715, Perplexity: 7.9371
Epoch [3/3], Step [2238/3236], Loss: 1.9388, Perplexity: 6.9502
Epoch [3/3], Step [2239/3236], Loss: 2.1531, Perplexity: 8.6119
Epoch [3/3], Step [2240/3236], Loss: 1.9472, Perplexity: 7.0089
Epoch [3/3], Step [2241/3236], Loss: 2.0398, Perplexity: 7.6890
Epoch [3/3], Step [2242/3236], Loss: 2.3161, Perplexity: 10.1361
Epoch [3/3], Step [2243/3236], Loss: 2.2065, Perplexity: 9.0841
Epoch [3/3], Step [2244/3236], Loss: 1.9796, Perplexity: 7.2401
Epoch [3/3], Step [2245/3236], Loss: 2.0058, Perplexity: 7.4319
Epoch [3/3], Step [2246/3236], Loss: 1.9980, Perplexity: 7.3741
Epoch [3/3], Step [2247/3236], Loss: 2.0801, Perplexity: 8.0051
Epoch [3/3], Step [2248/3236], Loss: 2.1962, Perplexity: 8.9911
Epoch [3/3], Step [2249/3236], Loss: 2.0835, Perplexity: 8.0324
Epoch [3/3], Step [2250/3236], Loss: 2.1042, Perplexity: 8.2006
Epoch [3/3], Step [2251/3236], Loss: 1.9665, Perplexity: 7.1457
Epoch [3/3], Step [2252/3236], Loss: 2.0579, Perplexity: 7.8294
Epoch [3/3], Step [2253/3236], Loss: 2.1955, Perplexity: 8.9841
Epoch [3/3], Step [2254/3236], Loss: 2.2717, Perplexity: 9.6960
Epoch [3/3], Step [2255/3236], Loss: 2.0382, Perplexity: 7.6768
Epoch [3/3], Step [2256/3236], Loss: 1.9186, Perplexity: 6.8116
Epoch [3/3], Step [2257/3236], Loss: 2.8085, Perplexity: 16.5853
Epoch [3/3], Step [2258/3236], Loss: 2.1840, Perplexity: 8.8815
Epoch [3/3], Step [2259/3236], Loss: 2.0407, Perplexity: 7.6957
Epoch [3/3], Step [2260/3236], Loss: 1.9781, Perplexity: 7.2287
Epoch [3/3], Step [2261/3236], Loss: 2.0936, Perplexity: 8.1141
Epoch [3/3], Step [2262/3236], Loss: 1.9407, Perplexity: 6.9637
Epoch [3/3], Step [2263/3236], Loss: 1.9979, Perplexity: 7.3736
Epoch [3/3], Step [2264/3236], Loss: 2.2922, Perplexity: 9.8968
Epoch [3/3], Step [2265/3236], Loss: 1.9549, Perplexity: 7.0629
Epoch [3/3], Step [2266/3236], Loss: 2.0137, Perplexity: 7.4911
Epoch [3/3], Step [2267/3236], Loss: 1.9419, Perplexity: 6.9718
Epoch [3/3], Step [2268/3236], Loss: 2.1330, Perplexity: 8.4398
Epoch [3/3], Step [2269/3236], Loss: 1.9653, Perplexity: 7.1367
Epoch [3/3], Step [2270/3236], Loss: 2.0140, Perplexity: 7.4931
Epoch [3/3], Step [2271/3236], Loss: 1.9321, Perplexity: 6.9039
Epoch [3/3], Step [2272/3236], Loss: 1.9316, Perplexity: 6.9004
Epoch [3/3], Step [2273/3236], Loss: 2.0528, Perplexity: 7.7895
Epoch [3/3], Step [2274/3236], Loss: 2.0390, Perplexity: 7.6826
Epoch [3/3], Step [2275/3236], Loss: 2.0136, Perplexity: 7.4900
Epoch [3/3], Step [2276/3236], Loss: 2.0605, Perplexity: 7.8501
Epoch [3/3], Step [2277/3236], Loss: 2.2585, Perplexity: 9.5683
Epoch [3/3], Step [2278/3236], Loss: 2.0642, Perplexity: 7.8789
Epoch [3/3], Step [2279/3236], Loss: 2.3340, Perplexity: 10.3194
Epoch [3/3], Step [2280/3236], Loss: 2.7253, Perplexity: 15.2614
Epoch [3/3], Step [2281/3236], Loss: 2.0435, Perplexity: 7.7175
Epoch [3/3], Step [2282/3236], Loss: 2.0091, Perplexity: 7.4569
Epoch [3/3], Step [2283/3236], Loss: 2.0545, Perplexity: 7.8030
Epoch [3/3], Step [2284/3236], Loss: 2.1132, Perplexity: 8.2746
Epoch [3/3], Step [2285/3236], Loss: 2.2238, Perplexity: 9.2425
Epoch [3/3], Step [2286/3236], Loss: 2.1761, Perplexity: 8.8122
Epoch [3/3], Step [2287/3236], Loss: 2.0855, Perplexity: 8.0488
Epoch [3/3], Step [2288/3236], Loss: 1.9669, Perplexity: 7.1484
Epoch [3/3], Step [2289/3236], Loss: 2.0169, Perplexity: 7.5151
Epoch [3/3], Step [2290/3236], Loss: 1.8794, Perplexity: 6.5498
Epoch [3/3], Step [2291/3236], Loss: 1.9507, Perplexity: 7.0336
Epoch [3/3], Step [2292/3236], Loss: 2.1987, Perplexity: 9.0135
Epoch [3/3], Step [2293/3236], Loss: 2.0474, Perplexity: 7.7476
Epoch [3/3], Step [2294/3236], Loss: 2.3521, Perplexity: 10.5073
Epoch [3/3], Step [2295/3236], Loss: 1.9721, Perplexity: 7.1859
Epoch [3/3], Step [2296/3236], Loss: 2.6380, Perplexity: 13.9855
Epoch [3/3], Step [2297/3236], Loss: 2.2595, Perplexity: 9.5780
Epoch [3/3], Step [2298/3236], Loss: 1.9115, Perplexity: 6.7634
Epoch [3/3], Step [2299/3236], Loss: 1.9530, Perplexity: 7.0496
Epoch [3/3], Step [2300/3236], Loss: 2.3944, Perplexity: 10.9618
Epoch [3/3], Step [2301/3236], Loss: 1.9828, Perplexity: 7.2628
Epoch [3/3], Step [2302/3236], Loss: 2.1240, Perplexity: 8.3642
Epoch [3/3], Step [2303/3236], Loss: 2.1596, Perplexity: 8.6674
Epoch [3/3], Step [2304/3236], Loss: 2.0196, Perplexity: 7.5354
Epoch [3/3], Step [2305/3236], Loss: 1.9260, Perplexity: 6.8617
Epoch [3/3], Step [2306/3236], Loss: 2.0993, Perplexity: 8.1604
Epoch [3/3], Step [2307/3236], Loss: 2.0884, Perplexity: 8.0722
Epoch [3/3], Step [2308/3236], Loss: 1.9185, Perplexity: 6.8106
Epoch [3/3], Step [2309/3236], Loss: 2.0340, Perplexity: 7.6443
Epoch [3/3], Step [2310/3236], Loss: 2.0870, Perplexity: 8.0608
Epoch [3/3], Step [2311/3236], Loss: 1.8924, Perplexity: 6.6350
Epoch [3/3], Step [2312/3236], Loss: 2.0131, Perplexity: 7.4868
Epoch [3/3], Step [2313/3236], Loss: 2.1414, Perplexity: 8.5117
Epoch [3/3], Step [2314/3236], Loss: 1.9343, Perplexity: 6.9195
Epoch [3/3], Step [2315/3236], Loss: 2.0583, Perplexity: 7.8328
Epoch [3/3], Step [2316/3236], Loss: 1.9537, Perplexity: 7.0551
Epoch [3/3], Step [2317/3236], Loss: 1.9722, Perplexity: 7.1866
Epoch [3/3], Step [2318/3236], Loss: 2.0170, Perplexity: 7.5158
Epoch [3/3], Step [2319/3236], Loss: 1.9448, Perplexity: 6.9920
Epoch [3/3], Step [2320/3236], Loss: 2.2394, Perplexity: 9.3880
Epoch [3/3], Step [2321/3236], Loss: 2.2178, Perplexity: 9.1872
Epoch [3/3], Step [2322/3236], Loss: 1.9801, Perplexity: 7.2437
Epoch [3/3], Step [2323/3236], Loss: 2.0662, Perplexity: 7.8948
Epoch [3/3], Step [2324/3236], Loss: 2.0066, Perplexity: 7.4383
Epoch [3/3], Step [2325/3236], Loss: 1.9931, Perplexity: 7.3382
Epoch [3/3], Step [2326/3236], Loss: 1.9322, Perplexity: 6.9046
Epoch [3/3], Step [2327/3236], Loss: 1.9572, Perplexity: 7.0798
Epoch [3/3], Step [2328/3236], Loss: 2.0135, Perplexity: 7.4893
Epoch [3/3], Step [2329/3236], Loss: 2.1833, Perplexity: 8.8755
Epoch [3/3], Step [2330/3236], Loss: 2.0726, Perplexity: 7.9453
Epoch [3/3], Step [2331/3236], Loss: 2.0599, Perplexity: 7.8450
Epoch [3/3], Step [2332/3236], Loss: 1.9689, Perplexity: 7.1628
Epoch [3/3], Step [2333/3236], Loss: 2.0246, Perplexity: 7.5734
Epoch [3/3], Step [2334/3236], Loss: 1.9934, Perplexity: 7.3404
Epoch [3/3], Step [2335/3236], Loss: 2.2095, Perplexity: 9.1110
Epoch [3/3], Step [2336/3236], Loss: 2.3754, Perplexity: 10.7549
Epoch [3/3], Step [2337/3236], Loss: 2.0179, Perplexity: 7.5223
Epoch [3/3], Step [2338/3236], Loss: 1.9695, Perplexity: 7.1673
Epoch [3/3], Step [2339/3236], Loss: 2.1716, Perplexity: 8.7723
Epoch [3/3], Step [2340/3236], Loss: 2.0086, Perplexity: 7.4529
Epoch [3/3], Step [2341/3236], Loss: 2.0878, Perplexity: 8.0668
Epoch [3/3], Step [2342/3236], Loss: 2.0461, Perplexity: 7.7377
Epoch [3/3], Step [2343/3236], Loss: 1.9725, Perplexity: 7.1883
Epoch [3/3], Step [2344/3236], Loss: 2.2283, Perplexity: 9.2843
Epoch [3/3], Step [2345/3236], Loss: 2.1594, Perplexity: 8.6662
Epoch [3/3], Step [2346/3236], Loss: 2.1964, Perplexity: 8.9923
Epoch [3/3], Step [2347/3236], Loss: 1.8944, Perplexity: 6.6486
Epoch [3/3], Step [2348/3236], Loss: 1.9643, Perplexity: 7.1298
Epoch [3/3], Step [2349/3236], Loss: 2.0686, Perplexity: 7.9140
Epoch [3/3], Step [2350/3236], Loss: 1.9594, Perplexity: 7.0948
Epoch [3/3], Step [2351/3236], Loss: 2.1152, Perplexity: 8.2915
Epoch [3/3], Step [2352/3236], Loss: 2.1146, Perplexity: 8.2863
Epoch [3/3], Step [2353/3236], Loss: 2.0802, Perplexity: 8.0065
Epoch [3/3], Step [2354/3236], Loss: 1.9634, Perplexity: 7.1233
Epoch [3/3], Step [2355/3236], Loss: 1.8655, Perplexity: 6.4589
Epoch [3/3], Step [2356/3236], Loss: 2.1667, Perplexity: 8.7293
Epoch [3/3], Step [2357/3236], Loss: 1.9621, Perplexity: 7.1141
Epoch [3/3], Step [2358/3236], Loss: 1.9924, Perplexity: 7.3330
Epoch [3/3], Step [2359/3236], Loss: 2.2588, Perplexity: 9.5712
Epoch [3/3], Step [2360/3236], Loss: 2.0090, Perplexity: 7.4562
Epoch [3/3], Step [2361/3236], Loss: 2.1557, Perplexity: 8.6337
Epoch [3/3], Step [2362/3236], Loss: 1.9639, Perplexity: 7.1272
Epoch [3/3], Step [2363/3236], Loss: 1.9345, Perplexity: 6.9206
Epoch [3/3], Step [2364/3236], Loss: 1.8488, Perplexity: 6.3524
Epoch [3/3], Step [2365/3236], Loss: 2.5705, Perplexity: 13.0729
Epoch [3/3], Step [2366/3236], Loss: 1.9919, Perplexity: 7.3296
Epoch [3/3], Step [2367/3236], Loss: 2.2321, Perplexity: 9.3196
Epoch [3/3], Step [2368/3236], Loss: 2.2771, Perplexity: 9.7485
Epoch [3/3], Step [2369/3236], Loss: 2.0570, Perplexity: 7.8223
Epoch [3/3], Step [2370/3236], Loss: 1.9571, Perplexity: 7.0789
Epoch [3/3], Step [2371/3236], Loss: 2.2379, Perplexity: 9.3733
Epoch [3/3], Step [2372/3236], Loss: 2.2323, Perplexity: 9.3217
Epoch [3/3], Step [2373/3236], Loss: 1.9032, Perplexity: 6.7075
Epoch [3/3], Step [2374/3236], Loss: 2.0101, Perplexity: 7.4638
Epoch [3/3], Step [2375/3236], Loss: 1.9695, Perplexity: 7.1669
Epoch [3/3], Step [2376/3236], Loss: 2.0377, Perplexity: 7.6731
Epoch [3/3], Step [2377/3236], Loss: 2.4450, Perplexity: 11.5309
Epoch [3/3], Step [2378/3236], Loss: 2.0030, Perplexity: 7.4115
Epoch [3/3], Step [2379/3236], Loss: 2.0014, Perplexity: 7.3995
Epoch [3/3], Step [2380/3236], Loss: 1.9671, Perplexity: 7.1502
Epoch [3/3], Step [2381/3236], Loss: 2.0032, Perplexity: 7.4126
Epoch [3/3], Step [2382/3236], Loss: 2.1600, Perplexity: 8.6708
Epoch [3/3], Step [2383/3236], Loss: 1.9026, Perplexity: 6.7035
Epoch [3/3], Step [2384/3236], Loss: 1.8912, Perplexity: 6.6275
Epoch [3/3], Step [2385/3236], Loss: 2.4841, Perplexity: 11.9906
Epoch [3/3], Step [2386/3236], Loss: 2.1961, Perplexity: 8.9903
Epoch [3/3], Step [2387/3236], Loss: 2.0558, Perplexity: 7.8131
Epoch [3/3], Step [2388/3236], Loss: 1.9117, Perplexity: 6.7645
Epoch [3/3], Step [2389/3236], Loss: 2.0325, Perplexity: 7.6330
Epoch [3/3], Step [2390/3236], Loss: 2.2581, Perplexity: 9.5650
Epoch [3/3], Step [2391/3236], Loss: 1.9630, Perplexity: 7.1209
Epoch [3/3], Step [2392/3236], Loss: 1.9577, Perplexity: 7.0832
Epoch [3/3], Step [2393/3236], Loss: 1.8754, Perplexity: 6.5232
Epoch [3/3], Step [2394/3236], Loss: 2.0105, Perplexity: 7.4669
Epoch [3/3], Step [2395/3236], Loss: 2.2770, Perplexity: 9.7471
Epoch [3/3], Step [2396/3236], Loss: 2.0548, Perplexity: 7.8054
Epoch [3/3], Step [2397/3236], Loss: 2.0584, Perplexity: 7.8337
Epoch [3/3], Step [2398/3236], Loss: 2.0498, Perplexity: 7.7662
Epoch [3/3], Step [2399/3236], Loss: 1.9947, Perplexity: 7.3499
Epoch [3/3], Step [2400/3236], Loss: 2.0985, Perplexity: 8.1541
Epoch [3/3], Step [2401/3236], Loss: 1.9766, Perplexity: 7.2180
Epoch [3/3], Step [2402/3236], Loss: 2.0076, Perplexity: 7.4454
Epoch [3/3], Step [2403/3236], Loss: 1.9809, Perplexity: 7.2495
Epoch [3/3], Step [2404/3236], Loss: 1.8955, Perplexity: 6.6556
Epoch [3/3], Step [2405/3236], Loss: 2.0821, Perplexity: 8.0212
Epoch [3/3], Step [2406/3236], Loss: 2.0048, Perplexity: 7.4246
Epoch [3/3], Step [2407/3236], Loss: 1.9953, Perplexity: 7.3544
Epoch [3/3], Step [2408/3236], Loss: 2.0344, Perplexity: 7.6479
Epoch [3/3], Step [2409/3236], Loss: 1.9993, Perplexity: 7.3837
Epoch [3/3], Step [2410/3236], Loss: 1.9969, Perplexity: 7.3665
Epoch [3/3], Step [2411/3236], Loss: 1.9396, Perplexity: 6.9562
Epoch [3/3], Step [2412/3236], Loss: 2.1755, Perplexity: 8.8062
Epoch [3/3], Step [2413/3236], Loss: 1.9793, Perplexity: 7.2377
Epoch [3/3], Step [2414/3236], Loss: 1.9526, Perplexity: 7.0467
Epoch [3/3], Step [2415/3236], Loss: 1.9858, Perplexity: 7.2850
Epoch [3/3], Step [2416/3236], Loss: 1.9811, Perplexity: 7.2509
Epoch [3/3], Step [2417/3236], Loss: 2.0092, Perplexity: 7.4573
Epoch [3/3], Step [2418/3236], Loss: 2.0582, Perplexity: 7.8318
Epoch [3/3], Step [2419/3236], Loss: 2.0655, Perplexity: 7.8889
Epoch [3/3], Step [2420/3236], Loss: 2.1576, Perplexity: 8.6500
Epoch [3/3], Step [2421/3236], Loss: 2.0684, Perplexity: 7.9120
Epoch [3/3], Step [2422/3236], Loss: 2.0365, Perplexity: 7.6634
Epoch [3/3], Step [2423/3236], Loss: 2.0808, Perplexity: 8.0106
Epoch [3/3], Step [2424/3236], Loss: 2.0671, Perplexity: 7.9019
Epoch [3/3], Step [2425/3236], Loss: 1.9082, Perplexity: 6.7407
Epoch [3/3], Step [2426/3236], Loss: 1.9029, Perplexity: 6.7056
Epoch [3/3], Step [2427/3236], Loss: 2.0444, Perplexity: 7.7245
Epoch [3/3], Step [2428/3236], Loss: 1.9976, Perplexity: 7.3711
Epoch [3/3], Step [2429/3236], Loss: 2.0170, Perplexity: 7.5155
Epoch [3/3], Step [2430/3236], Loss: 1.9056, Perplexity: 6.7233
Epoch [3/3], Step [2431/3236], Loss: 2.1720, Perplexity: 8.7760
Epoch [3/3], Step [2432/3236], Loss: 2.1241, Perplexity: 8.3653
Epoch [3/3], Step [2433/3236], Loss: 1.9971, Perplexity: 7.3680
Epoch [3/3], Step [2434/3236], Loss: 2.2038, Perplexity: 9.0590
Epoch [3/3], Step [2435/3236], Loss: 1.8162, Perplexity: 6.1484
Epoch [3/3], Step [2436/3236], Loss: 2.4687, Perplexity: 11.8066
Epoch [3/3], Step [2437/3236], Loss: 1.9527, Perplexity: 7.0474
Epoch [3/3], Step [2438/3236], Loss: 2.0008, Perplexity: 7.3949
Epoch [3/3], Step [2439/3236], Loss: 2.1032, Perplexity: 8.1921
Epoch [3/3], Step [2440/3236], Loss: 2.2477, Perplexity: 9.4660
Epoch [3/3], Step [2441/3236], Loss: 2.8973, Perplexity: 18.1259
Epoch [3/3], Step [2442/3236], Loss: 2.4938, Perplexity: 12.1071
Epoch [3/3], Step [2443/3236], Loss: 2.3944, Perplexity: 10.9614
Epoch [3/3], Step [2444/3236], Loss: 1.9689, Perplexity: 7.1625
Epoch [3/3], Step [2445/3236], Loss: 1.9845, Perplexity: 7.2757
Epoch [3/3], Step [2446/3236], Loss: 2.3146, Perplexity: 10.1212
Epoch [3/3], Step [2447/3236], Loss: 2.0890, Perplexity: 8.0772
Epoch [3/3], Step [2448/3236], Loss: 2.2571, Perplexity: 9.5551
Epoch [3/3], Step [2449/3236], Loss: 2.0779, Perplexity: 7.9876
Epoch [3/3], Step [2450/3236], Loss: 2.1338, Perplexity: 8.4468
Epoch [3/3], Step [2451/3236], Loss: 1.9776, Perplexity: 7.2257
Epoch [3/3], Step [2452/3236], Loss: 1.9512, Perplexity: 7.0372
Epoch [3/3], Step [2453/3236], Loss: 2.2112, Perplexity: 9.1266
Epoch [3/3], Step [2454/3236], Loss: 2.0464, Perplexity: 7.7399
Epoch [3/3], Step [2455/3236], Loss: 2.0224, Perplexity: 7.5562
Epoch [3/3], Step [2456/3236], Loss: 2.0778, Perplexity: 7.9865
Epoch [3/3], Step [2457/3236], Loss: 2.3644, Perplexity: 10.6382
Epoch [3/3], Step [2458/3236], Loss: 2.5503, Perplexity: 12.8115
Epoch [3/3], Step [2459/3236], Loss: 2.1570, Perplexity: 8.6448
Epoch [3/3], Step [2460/3236], Loss: 2.2863, Perplexity: 9.8383
Epoch [3/3], Step [2461/3236], Loss: 2.3654, Perplexity: 10.6485
Epoch [3/3], Step [2462/3236], Loss: 2.0767, Perplexity: 7.9783
Epoch [3/3], Step [2463/3236], Loss: 2.1108, Perplexity: 8.2546
Epoch [3/3], Step [2464/3236], Loss: 2.4256, Perplexity: 11.3085
Epoch [3/3], Step [2465/3236], Loss: 2.0508, Perplexity: 7.7744
Epoch [3/3], Step [2466/3236], Loss: 2.1698, Perplexity: 8.7567
Epoch [3/3], Step [2467/3236], Loss: 2.0720, Perplexity: 7.9406
Epoch [3/3], Step [2468/3236], Loss: 1.9651, Perplexity: 7.1360
Epoch [3/3], Step [2469/3236], Loss: 1.9536, Perplexity: 7.0541
Epoch [3/3], Step [2470/3236], Loss: 2.2702, Perplexity: 9.6816
Epoch [3/3], Step [2471/3236], Loss: 1.9398, Perplexity: 6.9571
Epoch [3/3], Step [2472/3236], Loss: 2.3779, Perplexity: 10.7818
Epoch [3/3], Step [2473/3236], Loss: 1.9013, Perplexity: 6.6947
Epoch [3/3], Step [2474/3236], Loss: 2.0496, Perplexity: 7.7649
Epoch [3/3], Step [2475/3236], Loss: 2.0998, Perplexity: 8.1642
Epoch [3/3], Step [2476/3236], Loss: 2.2402, Perplexity: 9.3955
Epoch [3/3], Step [2477/3236], Loss: 2.1093, Perplexity: 8.2427
Epoch [3/3], Step [2478/3236], Loss: 1.9525, Perplexity: 7.0463
Epoch [3/3], Step [2479/3236], Loss: 2.3058, Perplexity: 10.0319
Epoch [3/3], Step [2480/3236], Loss: 2.0379, Perplexity: 7.6744
Epoch [3/3], Step [2481/3236], Loss: 2.0396, Perplexity: 7.6874
Epoch [3/3], Step [2482/3236], Loss: 2.3192, Perplexity: 10.1679
Epoch [3/3], Step [2483/3236], Loss: 2.1333, Perplexity: 8.4430
Epoch [3/3], Step [2484/3236], Loss: 1.9446, Perplexity: 6.9908
Epoch [3/3], Step [2485/3236], Loss: 2.3204, Perplexity: 10.1794
Epoch [3/3], Step [2486/3236], Loss: 1.9728, Perplexity: 7.1908
Epoch [3/3], Step [2487/3236], Loss: 2.0342, Perplexity: 7.6463
Epoch [3/3], Step [2488/3236], Loss: 2.1345, Perplexity: 8.4529
Epoch [3/3], Step [2489/3236], Loss: 1.9559, Perplexity: 7.0702
Epoch [3/3], Step [2490/3236], Loss: 2.1895, Perplexity: 8.9305
Epoch [3/3], Step [2491/3236], Loss: 2.1204, Perplexity: 8.3344
Epoch [3/3], Step [2492/3236], Loss: 1.9361, Perplexity: 6.9320
Epoch [3/3], Step [2493/3236], Loss: 2.1773, Perplexity: 8.8220
Epoch [3/3], Step [2494/3236], Loss: 2.0269, Perplexity: 7.5905
Epoch [3/3], Step [2495/3236], Loss: 1.8338, Perplexity: 6.2579
Epoch [3/3], Step [2496/3236], Loss: 1.9653, Perplexity: 7.1374
Epoch [3/3], Step [2497/3236], Loss: 2.0955, Perplexity: 8.1292
Epoch [3/3], Step [2498/3236], Loss: 1.7837, Perplexity: 5.9521
Epoch [3/3], Step [2499/3236], Loss: 1.9539, Perplexity: 7.0564
Epoch [3/3], Step [2500/3236], Loss: 1.8562, Perplexity: 6.3992
Epoch [3/3], Step [2501/3236], Loss: 2.3157, Perplexity: 10.1324
Epoch [3/3], Step [2502/3236], Loss: 1.9571, Perplexity: 7.0786
Epoch [3/3], Step [2503/3236], Loss: 2.6889, Perplexity: 14.7151
Epoch [3/3], Step [2504/3236], Loss: 2.0260, Perplexity: 7.5836
Epoch [3/3], Step [2505/3236], Loss: 2.3201, Perplexity: 10.1772
Epoch [3/3], Step [2506/3236], Loss: 2.0136, Perplexity: 7.4905
Epoch [3/3], Step [2507/3236], Loss: 2.1423, Perplexity: 8.5194
Epoch [3/3], Step [2508/3236], Loss: 1.9875, Perplexity: 7.2971
Epoch [3/3], Step [2509/3236], Loss: 2.0033, Perplexity: 7.4133
Epoch [3/3], Step [2510/3236], Loss: 2.0085, Perplexity: 7.4519
Epoch [3/3], Step [2511/3236], Loss: 1.8657, Perplexity: 6.4603
Epoch [3/3], Step [2512/3236], Loss: 1.8195, Perplexity: 6.1688
Epoch [3/3], Step [2513/3236], Loss: 1.9226, Perplexity: 6.8388
Epoch [3/3], Step [2514/3236], Loss: 1.9329, Perplexity: 6.9097
Epoch [3/3], Step [2515/3236], Loss: 1.9526, Perplexity: 7.0468
Epoch [3/3], Step [2516/3236], Loss: 2.0648, Perplexity: 7.8836
Epoch [3/3], Step [2517/3236], Loss: 2.2615, Perplexity: 9.5976
Epoch [3/3], Step [2518/3236], Loss: 1.9825, Perplexity: 7.2608
Epoch [3/3], Step [2519/3236], Loss: 2.0869, Perplexity: 8.0598
Epoch [3/3], Step [2520/3236], Loss: 2.0585, Perplexity: 7.8341
Epoch [3/3], Step [2521/3236], Loss: 2.2592, Perplexity: 9.5750
Epoch [3/3], Step [2522/3236], Loss: 1.9862, Perplexity: 7.2881
Epoch [3/3], Step [2523/3236], Loss: 1.9411, Perplexity: 6.9666
Epoch [3/3], Step [2524/3236], Loss: 1.9708, Perplexity: 7.1763
Epoch [3/3], Step [2525/3236], Loss: 2.2305, Perplexity: 9.3048
Epoch [3/3], Step [2526/3236], Loss: 1.9329, Perplexity: 6.9094
Epoch [3/3], Step [2527/3236], Loss: 1.9752, Perplexity: 7.2083
Epoch [3/3], Step [2528/3236], Loss: 2.0962, Perplexity: 8.1354
Epoch [3/3], Step [2529/3236], Loss: 2.0269, Perplexity: 7.5903
Epoch [3/3], Step [2530/3236], Loss: 2.2937, Perplexity: 9.9114
Epoch [3/3], Step [2531/3236], Loss: 2.0768, Perplexity: 7.9788
Epoch [3/3], Step [2532/3236], Loss: 2.3024, Perplexity: 9.9984
Epoch [3/3], Step [2533/3236], Loss: 2.0902, Perplexity: 8.0868
Epoch [3/3], Step [2534/3236], Loss: 2.0342, Perplexity: 7.6464
Epoch [3/3], Step [2535/3236], Loss: 2.1343, Perplexity: 8.4513
Epoch [3/3], Step [2536/3236], Loss: 2.0850, Perplexity: 8.0446
Epoch [3/3], Step [2537/3236], Loss: 2.0475, Perplexity: 7.7484
Epoch [3/3], Step [2538/3236], Loss: 1.8840, Perplexity: 6.5796
Epoch [3/3], Step [2539/3236], Loss: 2.2157, Perplexity: 9.1674
Epoch [3/3], Step [2540/3236], Loss: 1.8982, Perplexity: 6.6739
Epoch [3/3], Step [2541/3236], Loss: 1.9191, Perplexity: 6.8151
Epoch [3/3], Step [2542/3236], Loss: 1.9839, Perplexity: 7.2710
Epoch [3/3], Step [2543/3236], Loss: 2.0971, Perplexity: 8.1429
Epoch [3/3], Step [2544/3236], Loss: 2.3076, Perplexity: 10.0504
Epoch [3/3], Step [2545/3236], Loss: 1.8250, Perplexity: 6.2028
Epoch [3/3], Step [2546/3236], Loss: 1.9710, Perplexity: 7.1775
Epoch [3/3], Step [2547/3236], Loss: 2.0036, Perplexity: 7.4160
Epoch [3/3], Step [2548/3236], Loss: 1.9560, Perplexity: 7.0713
Epoch [3/3], Step [2549/3236], Loss: 2.3137, Perplexity: 10.1121
Epoch [3/3], Step [2550/3236], Loss: 2.1882, Perplexity: 8.9188
Epoch [3/3], Step [2551/3236], Loss: 1.9590, Perplexity: 7.0919
Epoch [3/3], Step [2552/3236], Loss: 2.0100, Perplexity: 7.4633
Epoch [3/3], Step [2553/3236], Loss: 1.9420, Perplexity: 6.9729
Epoch [3/3], Step [2554/3236], Loss: 1.9880, Perplexity: 7.3008
Epoch [3/3], Step [2555/3236], Loss: 2.0040, Perplexity: 7.4189
Epoch [3/3], Step [2556/3236], Loss: 1.9930, Perplexity: 7.3378
Epoch [3/3], Step [2557/3236], Loss: 2.0774, Perplexity: 7.9835
Epoch [3/3], Step [2558/3236], Loss: 1.9455, Perplexity: 6.9972
Epoch [3/3], Step [2559/3236], Loss: 1.9135, Perplexity: 6.7769
Epoch [3/3], Step [2560/3236], Loss: 1.9304, Perplexity: 6.8925
Epoch [3/3], Step [2561/3236], Loss: 2.1154, Perplexity: 8.2928
Epoch [3/3], Step [2562/3236], Loss: 2.0263, Perplexity: 7.5861
Epoch [3/3], Step [2563/3236], Loss: 1.9457, Perplexity: 6.9989
Epoch [3/3], Step [2564/3236], Loss: 2.0401, Perplexity: 7.6915
Epoch [3/3], Step [2565/3236], Loss: 1.9012, Perplexity: 6.6938
Epoch [3/3], Step [2566/3236], Loss: 2.0151, Perplexity: 7.5017
Epoch [3/3], Step [2567/3236], Loss: 2.1256, Perplexity: 8.3776
Epoch [3/3], Step [2568/3236], Loss: 2.3400, Perplexity: 10.3813
Epoch [3/3], Step [2569/3236], Loss: 1.8296, Perplexity: 6.2316
Epoch [3/3], Step [2570/3236], Loss: 2.1446, Perplexity: 8.5383
Epoch [3/3], Step [2571/3236], Loss: 2.1055, Perplexity: 8.2110
Epoch [3/3], Step [2572/3236], Loss: 1.8723, Perplexity: 6.5032
Epoch [3/3], Step [2573/3236], Loss: 2.0058, Perplexity: 7.4317
Epoch [3/3], Step [2574/3236], Loss: 1.8913, Perplexity: 6.6280
Epoch [3/3], Step [2575/3236], Loss: 1.9285, Perplexity: 6.8792
Epoch [3/3], Step [2576/3236], Loss: 2.1563, Perplexity: 8.6388
Epoch [3/3], Step [2577/3236], Loss: 2.0159, Perplexity: 7.5077
Epoch [3/3], Step [2578/3236], Loss: 1.9906, Perplexity: 7.3199
Epoch [3/3], Step [2579/3236], Loss: 2.0818, Perplexity: 8.0192
Epoch [3/3], Step [2580/3236], Loss: 1.9323, Perplexity: 6.9051
Epoch [3/3], Step [2581/3236], Loss: 1.9287, Perplexity: 6.8808
Epoch [3/3], Step [2582/3236], Loss: 2.3026, Perplexity: 10.0006
Epoch [3/3], Step [2583/3236], Loss: 2.1108, Perplexity: 8.2544
Epoch [3/3], Step [2584/3236], Loss: 2.1310, Perplexity: 8.4230
Epoch [3/3], Step [2585/3236], Loss: 2.0808, Perplexity: 8.0112
Epoch [3/3], Step [2586/3236], Loss: 2.0870, Perplexity: 8.0610
Epoch [3/3], Step [2587/3236], Loss: 2.0534, Perplexity: 7.7940
Epoch [3/3], Step [2588/3236], Loss: 1.9840, Perplexity: 7.2720
Epoch [3/3], Step [2589/3236], Loss: 2.2072, Perplexity: 9.0904
Epoch [3/3], Step [2590/3236], Loss: 1.9423, Perplexity: 6.9746
Epoch [3/3], Step [2591/3236], Loss: 2.0994, Perplexity: 8.1616
Epoch [3/3], Step [2592/3236], Loss: 1.9118, Perplexity: 6.7654
Epoch [3/3], Step [2593/3236], Loss: 2.0363, Perplexity: 7.6619
Epoch [3/3], Step [2594/3236], Loss: 2.0947, Perplexity: 8.1228
Epoch [3/3], Step [2595/3236], Loss: 1.8850, Perplexity: 6.5865
Epoch [3/3], Step [2596/3236], Loss: 2.2509, Perplexity: 9.4961
Epoch [3/3], Step [2597/3236], Loss: 2.0198, Perplexity: 7.5370
Epoch [3/3], Step [2598/3236], Loss: 1.9130, Perplexity: 6.7733
Epoch [3/3], Step [2599/3236], Loss: 2.0154, Perplexity: 7.5037
Epoch [3/3], Step [2600/3236], Loss: 1.9882, Perplexity: 7.3022
Epoch [3/3], Step [2601/3236], Loss: 1.9269, Perplexity: 6.8679
Epoch [3/3], Step [2602/3236], Loss: 1.9241, Perplexity: 6.8489
Epoch [3/3], Step [2603/3236], Loss: 2.1812, Perplexity: 8.8572
Epoch [3/3], Step [2604/3236], Loss: 1.9727, Perplexity: 7.1903
Epoch [3/3], Step [2605/3236], Loss: 2.1597, Perplexity: 8.6685
Epoch [3/3], Step [2606/3236], Loss: 1.9842, Perplexity: 7.2730
Epoch [3/3], Step [2607/3236], Loss: 2.5548, Perplexity: 12.8689
Epoch [3/3], Step [2608/3236], Loss: 1.9715, Perplexity: 7.1815
Epoch [3/3], Step [2609/3236], Loss: 2.5031, Perplexity: 12.2201
Epoch [3/3], Step [2610/3236], Loss: 1.8550, Perplexity: 6.3920
Epoch [3/3], Step [2611/3236], Loss: 1.8925, Perplexity: 6.6362
Epoch [3/3], Step [2612/3236], Loss: 2.2904, Perplexity: 9.8792
Epoch [3/3], Step [2613/3236], Loss: 2.1473, Perplexity: 8.5619
Epoch [3/3], Step [2614/3236], Loss: 2.0406, Perplexity: 7.6950
Epoch [3/3], Step [2615/3236], Loss: 1.8506, Perplexity: 6.3636
Epoch [3/3], Step [2616/3236], Loss: 1.9682, Perplexity: 7.1578
Epoch [3/3], Step [2617/3236], Loss: 1.9271, Perplexity: 6.8693
Epoch [3/3], Step [2618/3236], Loss: 2.0276, Perplexity: 7.5955
Epoch [3/3], Step [2619/3236], Loss: 2.1892, Perplexity: 8.9277
Epoch [3/3], Step [2620/3236], Loss: 2.0799, Perplexity: 8.0035
Epoch [3/3], Step [2621/3236], Loss: 1.9815, Perplexity: 7.2537
Epoch [3/3], Step [2622/3236], Loss: 2.0804, Perplexity: 8.0075
Epoch [3/3], Step [2623/3236], Loss: 1.9792, Perplexity: 7.2371
Epoch [3/3], Step [2624/3236], Loss: 2.0346, Perplexity: 7.6489
Epoch [3/3], Step [2625/3236], Loss: 2.0270, Perplexity: 7.5910
Epoch [3/3], Step [2626/3236], Loss: 2.0161, Perplexity: 7.5088
Epoch [3/3], Step [2627/3236], Loss: 1.8707, Perplexity: 6.4929
Epoch [3/3], Step [2628/3236], Loss: 2.2408, Perplexity: 9.4009
Epoch [3/3], Step [2629/3236], Loss: 1.9620, Perplexity: 7.1134
Epoch [3/3], Step [2630/3236], Loss: 1.9999, Perplexity: 7.3883
Epoch [3/3], Step [2631/3236], Loss: 2.1682, Perplexity: 8.7424
Epoch [3/3], Step [2632/3236], Loss: 1.9652, Perplexity: 7.1367
Epoch [3/3], Step [2633/3236], Loss: 2.0003, Perplexity: 7.3912
Epoch [3/3], Step [2634/3236], Loss: 2.0821, Perplexity: 8.0212
Epoch [3/3], Step [2635/3236], Loss: 1.9177, Perplexity: 6.8053
Epoch [3/3], Step [2636/3236], Loss: 1.9188, Perplexity: 6.8127
Epoch [3/3], Step [2637/3236], Loss: 1.9474, Perplexity: 7.0104
Epoch [3/3], Step [2638/3236], Loss: 2.0122, Perplexity: 7.4796
Epoch [3/3], Step [2639/3236], Loss: 1.9838, Perplexity: 7.2707
Epoch [3/3], Step [2640/3236], Loss: 1.9635, Perplexity: 7.1242
Epoch [3/3], Step [2641/3236], Loss: 1.9549, Perplexity: 7.0631
Epoch [3/3], Step [2642/3236], Loss: 2.0388, Perplexity: 7.6815
Epoch [3/3], Step [2643/3236], Loss: 2.1489, Perplexity: 8.5756
Epoch [3/3], Step [2644/3236], Loss: 1.9777, Perplexity: 7.2262
Epoch [3/3], Step [2645/3236], Loss: 2.2161, Perplexity: 9.1714
Epoch [3/3], Step [2646/3236], Loss: 1.9680, Perplexity: 7.1567
Epoch [3/3], Step [2647/3236], Loss: 1.9731, Perplexity: 7.1926
Epoch [3/3], Step [2648/3236], Loss: 2.1742, Perplexity: 8.7955
Epoch [3/3], Step [2649/3236], Loss: 2.0179, Perplexity: 7.5228
Epoch [3/3], Step [2650/3236], Loss: 1.9320, Perplexity: 6.9032
Epoch [3/3], Step [2651/3236], Loss: 2.0503, Perplexity: 7.7705
Epoch [3/3], Step [2652/3236], Loss: 2.0272, Perplexity: 7.5927
Epoch [3/3], Step [2653/3236], Loss: 1.8766, Perplexity: 6.5315
Epoch [3/3], Step [2654/3236], Loss: 1.9864, Perplexity: 7.2890
Epoch [3/3], Step [2655/3236], Loss: 2.2494, Perplexity: 9.4819
Epoch [3/3], Step [2656/3236], Loss: 2.0308, Perplexity: 7.6201
Epoch [3/3], Step [2657/3236], Loss: 2.0126, Perplexity: 7.4827
Epoch [3/3], Step [2658/3236], Loss: 1.8596, Perplexity: 6.4212
Epoch [3/3], Step [2659/3236], Loss: 2.1082, Perplexity: 8.2335
Epoch [3/3], Step [2660/3236], Loss: 2.4459, Perplexity: 11.5406
Epoch [3/3], Step [2661/3236], Loss: 2.0021, Perplexity: 7.4048
Epoch [3/3], Step [2662/3236], Loss: 1.9485, Perplexity: 7.0182
Epoch [3/3], Step [2663/3236], Loss: 2.0854, Perplexity: 8.0474
Epoch [3/3], Step [2664/3236], Loss: 2.5127, Perplexity: 12.3381
Epoch [3/3], Step [2665/3236], Loss: 1.9719, Perplexity: 7.1840
Epoch [3/3], Step [2666/3236], Loss: 1.9819, Perplexity: 7.2567
Epoch [3/3], Step [2667/3236], Loss: 1.9842, Perplexity: 7.2735
Epoch [3/3], Step [2668/3236], Loss: 2.0409, Perplexity: 7.6975
Epoch [3/3], Step [2669/3236], Loss: 2.0266, Perplexity: 7.5886
Epoch [3/3], Step [2670/3236], Loss: 2.0122, Perplexity: 7.4799
Epoch [3/3], Step [2671/3236], Loss: 2.0608, Perplexity: 7.8521
Epoch [3/3], Step [2672/3236], Loss: 2.2421, Perplexity: 9.4132
Epoch [3/3], Step [2673/3236], Loss: 2.0032, Perplexity: 7.4128
Epoch [3/3], Step [2674/3236], Loss: 1.9283, Perplexity: 6.8781
Epoch [3/3], Step [2675/3236], Loss: 1.9273, Perplexity: 6.8709
Epoch [3/3], Step [2676/3236], Loss: 1.9395, Perplexity: 6.9554
Epoch [3/3], Step [2677/3236], Loss: 1.9790, Perplexity: 7.2357
Epoch [3/3], Step [2678/3236], Loss: 1.9704, Perplexity: 7.1733
Epoch [3/3], Step [2679/3236], Loss: 2.3041, Perplexity: 10.0156
Epoch [3/3], Step [2680/3236], Loss: 1.9805, Perplexity: 7.2466
Epoch [3/3], Step [2681/3236], Loss: 2.1661, Perplexity: 8.7240
Epoch [3/3], Step [2682/3236], Loss: 2.3553, Perplexity: 10.5415
Epoch [3/3], Step [2683/3236], Loss: 2.0115, Perplexity: 7.4748
Epoch [3/3], Step [2684/3236], Loss: 1.8885, Perplexity: 6.6095
Epoch [3/3], Step [2685/3236], Loss: 1.9263, Perplexity: 6.8640
Epoch [3/3], Step [2686/3236], Loss: 2.0426, Perplexity: 7.7109
Epoch [3/3], Step [2687/3236], Loss: 2.2635, Perplexity: 9.6168
Epoch [3/3], Step [2688/3236], Loss: 1.8574, Perplexity: 6.4070
Epoch [3/3], Step [2689/3236], Loss: 1.9849, Perplexity: 7.2785
Epoch [3/3], Step [2690/3236], Loss: 2.4087, Perplexity: 11.1197
Epoch [3/3], Step [2691/3236], Loss: 2.0060, Perplexity: 7.4336
Epoch [3/3], Step [2692/3236], Loss: 2.0484, Perplexity: 7.7558
Epoch [3/3], Step [2693/3236], Loss: 1.9914, Perplexity: 7.3256
Epoch [3/3], Step [2694/3236], Loss: 2.1160, Perplexity: 8.2976
Epoch [3/3], Step [2695/3236], Loss: 1.9887, Perplexity: 7.3059
Epoch [3/3], Step [2696/3236], Loss: 2.0461, Perplexity: 7.7375
Epoch [3/3], Step [2697/3236], Loss: 2.0343, Perplexity: 7.6468
Epoch [3/3], Step [2698/3236], Loss: 1.9663, Perplexity: 7.1443
Epoch [3/3], Step [2699/3236], Loss: 1.8925, Perplexity: 6.6356
Epoch [3/3], Step [2700/3236], Loss: 1.9946, Perplexity: 7.3493
Epoch [3/3], Step [2701/3236], Loss: 1.8532, Perplexity: 6.3804
Epoch [3/3], Step [2702/3236], Loss: 2.0523, Perplexity: 7.7860
Epoch [3/3], Step [2703/3236], Loss: 1.8824, Perplexity: 6.5691
Epoch [3/3], Step [2704/3236], Loss: 1.9882, Perplexity: 7.3022
Epoch [3/3], Step [2705/3236], Loss: 2.0628, Perplexity: 7.8683
Epoch [3/3], Step [2706/3236], Loss: 2.0781, Perplexity: 7.9892
Epoch [3/3], Step [2707/3236], Loss: 2.1252, Perplexity: 8.3748
Epoch [3/3], Step [2708/3236], Loss: 2.0998, Perplexity: 8.1649
Epoch [3/3], Step [2709/3236], Loss: 1.9814, Perplexity: 7.2528
Epoch [3/3], Step [2710/3236], Loss: 2.0974, Perplexity: 8.1447
Epoch [3/3], Step [2711/3236], Loss: 2.0331, Perplexity: 7.6378
Epoch [3/3], Step [2712/3236], Loss: 2.0647, Perplexity: 7.8829
Epoch [3/3], Step [2713/3236], Loss: 1.9743, Perplexity: 7.2018
Epoch [3/3], Step [2714/3236], Loss: 1.9141, Perplexity: 6.7807
Epoch [3/3], Step [2715/3236], Loss: 2.0524, Perplexity: 7.7866
Epoch [3/3], Step [2716/3236], Loss: 2.2986, Perplexity: 9.9603
Epoch [3/3], Step [2717/3236], Loss: 2.0852, Perplexity: 8.0462
Epoch [3/3], Step [2718/3236], Loss: 1.9936, Perplexity: 7.3418
Epoch [3/3], Step [2719/3236], Loss: 2.1795, Perplexity: 8.8418
Epoch [3/3], Step [2720/3236], Loss: 1.9927, Perplexity: 7.3355
Epoch [3/3], Step [2721/3236], Loss: 2.0172, Perplexity: 7.5172
Epoch [3/3], Step [2722/3236], Loss: 2.0399, Perplexity: 7.6899
Epoch [3/3], Step [2723/3236], Loss: 2.1506, Perplexity: 8.5898
Epoch [3/3], Step [2724/3236], Loss: 2.0480, Perplexity: 7.7526
Epoch [3/3], Step [2725/3236], Loss: 2.2142, Perplexity: 9.1544
Epoch [3/3], Step [2726/3236], Loss: 1.9707, Perplexity: 7.1759
Epoch [3/3], Step [2727/3236], Loss: 1.9375, Perplexity: 6.9411
Epoch [3/3], Step [2728/3236], Loss: 1.8428, Perplexity: 6.3139
Epoch [3/3], Step [2729/3236], Loss: 1.9909, Perplexity: 7.3224
Epoch [3/3], Step [2730/3236], Loss: 1.9686, Perplexity: 7.1604
Epoch [3/3], Step [2731/3236], Loss: 1.9741, Perplexity: 7.2002
Epoch [3/3], Step [2732/3236], Loss: 1.9648, Perplexity: 7.1336
Epoch [3/3], Step [2733/3236], Loss: 2.4433, Perplexity: 11.5115
Epoch [3/3], Step [2734/3236], Loss: 2.4453, Perplexity: 11.5344
Epoch [3/3], Step [2735/3236], Loss: 2.1209, Perplexity: 8.3389
Epoch [3/3], Step [2736/3236], Loss: 1.9278, Perplexity: 6.8746
Epoch [3/3], Step [2737/3236], Loss: 1.9353, Perplexity: 6.9258
Epoch [3/3], Step [2738/3236], Loss: 1.9466, Perplexity: 7.0046
Epoch [3/3], Step [2739/3236], Loss: 1.9555, Perplexity: 7.0678
Epoch [3/3], Step [2740/3236], Loss: 1.9157, Perplexity: 6.7917
Epoch [3/3], Step [2741/3236], Loss: 1.9732, Perplexity: 7.1935
Epoch [3/3], Step [2742/3236], Loss: 2.2697, Perplexity: 9.6761
Epoch [3/3], Step [2743/3236], Loss: 1.8664, Perplexity: 6.4651
Epoch [3/3], Step [2744/3236], Loss: 1.9254, Perplexity: 6.8581
Epoch [3/3], Step [2745/3236], Loss: 1.9761, Perplexity: 7.2144
Epoch [3/3], Step [2746/3236], Loss: 2.0301, Perplexity: 7.6146
Epoch [3/3], Step [2747/3236], Loss: 2.0809, Perplexity: 8.0116
Epoch [3/3], Step [2748/3236], Loss: 2.0735, Perplexity: 7.9524
Epoch [3/3], Step [2749/3236], Loss: 2.0999, Perplexity: 8.1650
Epoch [3/3], Step [2750/3236], Loss: 2.3392, Perplexity: 10.3733
Epoch [3/3], Step [2751/3236], Loss: 2.7131, Perplexity: 15.0754
Epoch [3/3], Step [2752/3236], Loss: 2.5123, Perplexity: 12.3332
Epoch [3/3], Step [2753/3236], Loss: 1.9338, Perplexity: 6.9154
Epoch [3/3], Step [2754/3236], Loss: 1.9841, Perplexity: 7.2722
Epoch [3/3], Step [2755/3236], Loss: 1.9503, Perplexity: 7.0307
Epoch [3/3], Step [2756/3236], Loss: 2.0348, Perplexity: 7.6510
Epoch [3/3], Step [2757/3236], Loss: 2.0460, Perplexity: 7.7373
Epoch [3/3], Step [2758/3236], Loss: 2.1024, Perplexity: 8.1855
Epoch [3/3], Step [2759/3236], Loss: 2.5276, Perplexity: 12.5235
Epoch [3/3], Step [2760/3236], Loss: 1.9513, Perplexity: 7.0382
Epoch [3/3], Step [2761/3236], Loss: 2.0606, Perplexity: 7.8504
Epoch [3/3], Step [2762/3236], Loss: 1.9080, Perplexity: 6.7396
Epoch [3/3], Step [2763/3236], Loss: 1.9815, Perplexity: 7.2537
Epoch [3/3], Step [2764/3236], Loss: 2.0850, Perplexity: 8.0448
Epoch [3/3], Step [2765/3236], Loss: 2.0399, Perplexity: 7.6897
Epoch [3/3], Step [2766/3236], Loss: 2.0709, Perplexity: 7.9317
Epoch [3/3], Step [2767/3236], Loss: 2.2126, Perplexity: 9.1395
Epoch [3/3], Step [2768/3236], Loss: 2.0162, Perplexity: 7.5098
Epoch [3/3], Step [2769/3236], Loss: 2.2629, Perplexity: 9.6114
Epoch [3/3], Step [2770/3236], Loss: 1.9615, Perplexity: 7.1100
Epoch [3/3], Step [2771/3236], Loss: 2.3228, Perplexity: 10.2038
Epoch [3/3], Step [2772/3236], Loss: 2.0151, Perplexity: 7.5017
Epoch [3/3], Step [2773/3236], Loss: 2.1025, Perplexity: 8.1865
Epoch [3/3], Step [2774/3236], Loss: 2.5813, Perplexity: 13.2150
Epoch [3/3], Step [2775/3236], Loss: 2.0119, Perplexity: 7.4774
Epoch [3/3], Step [2776/3236], Loss: 1.9991, Perplexity: 7.3823
Epoch [3/3], Step [2777/3236], Loss: 1.8408, Perplexity: 6.3017
Epoch [3/3], Step [2778/3236], Loss: 2.0084, Perplexity: 7.4514
Epoch [3/3], Step [2779/3236], Loss: 2.0541, Perplexity: 7.7998
Epoch [3/3], Step [2780/3236], Loss: 2.2229, Perplexity: 9.2343
Epoch [3/3], Step [2781/3236], Loss: 3.0101, Perplexity: 20.2890
Epoch [3/3], Step [2782/3236], Loss: 1.9139, Perplexity: 6.7794
Epoch [3/3], Step [2783/3236], Loss: 2.1162, Perplexity: 8.2993
Epoch [3/3], Step [2784/3236], Loss: 2.0470, Perplexity: 7.7449
Epoch [3/3], Step [2785/3236], Loss: 2.2039, Perplexity: 9.0602
Epoch [3/3], Step [2786/3236], Loss: 1.9533, Perplexity: 7.0521
Epoch [3/3], Step [2787/3236], Loss: 2.0050, Perplexity: 7.4261
Epoch [3/3], Step [2788/3236], Loss: 2.0762, Perplexity: 7.9741
Epoch [3/3], Step [2789/3236], Loss: 1.9730, Perplexity: 7.1923
Epoch [3/3], Step [2790/3236], Loss: 2.1932, Perplexity: 8.9640
Epoch [3/3], Step [2791/3236], Loss: 2.0331, Perplexity: 7.6380
Epoch [3/3], Step [2792/3236], Loss: 2.0095, Perplexity: 7.4594
Epoch [3/3], Step [2793/3236], Loss: 2.0444, Perplexity: 7.7245
Epoch [3/3], Step [2794/3236], Loss: 2.1552, Perplexity: 8.6294
Epoch [3/3], Step [2795/3236], Loss: 1.8999, Perplexity: 6.6853
Epoch [3/3], Step [2796/3236], Loss: 1.7972, Perplexity: 6.0328
Epoch [3/3], Step [2797/3236], Loss: 2.1285, Perplexity: 8.4019
Epoch [3/3], Step [2798/3236], Loss: 1.9029, Perplexity: 6.7054
Epoch [3/3], Step [2799/3236], Loss: 2.0249, Perplexity: 7.5754
Epoch [3/3], Step [2800/3236], Loss: 2.1285, Perplexity: 8.4022
Epoch [3/3], Step [2801/3236], Loss: 1.9108, Perplexity: 6.7583
Epoch [3/3], Step [2802/3236], Loss: 1.9458, Perplexity: 6.9995
Epoch [3/3], Step [2803/3236], Loss: 2.0685, Perplexity: 7.9128
Epoch [3/3], Step [2804/3236], Loss: 1.9261, Perplexity: 6.8629
Epoch [3/3], Step [2805/3236], Loss: 2.0651, Perplexity: 7.8863
Epoch [3/3], Step [2806/3236], Loss: 2.0337, Perplexity: 7.6420
Epoch [3/3], Step [2807/3236], Loss: 2.0379, Perplexity: 7.6745
Epoch [3/3], Step [2808/3236], Loss: 2.5242, Perplexity: 12.4805
Epoch [3/3], Step [2809/3236], Loss: 1.9174, Perplexity: 6.8035
Epoch [3/3], Step [2810/3236], Loss: 1.9400, Perplexity: 6.9588
Epoch [3/3], Step [2811/3236], Loss: 1.8952, Perplexity: 6.6537
Epoch [3/3], Step [2812/3236], Loss: 1.8788, Perplexity: 6.5456
Epoch [3/3], Step [2813/3236], Loss: 2.0231, Perplexity: 7.5614
Epoch [3/3], Step [2814/3236], Loss: 2.0619, Perplexity: 7.8610
Epoch [3/3], Step [2815/3236], Loss: 2.0169, Perplexity: 7.5148
Epoch [3/3], Step [2816/3236], Loss: 2.0514, Perplexity: 7.7785
Epoch [3/3], Step [2817/3236], Loss: 2.0883, Perplexity: 8.0710
Epoch [3/3], Step [2818/3236], Loss: 2.1924, Perplexity: 8.9567
Epoch [3/3], Step [2819/3236], Loss: 1.8870, Perplexity: 6.5994
Epoch [3/3], Step [2820/3236], Loss: 2.4963, Perplexity: 12.1373
Epoch [3/3], Step [2821/3236], Loss: 2.0131, Perplexity: 7.4862
Epoch [3/3], Step [2822/3236], Loss: 2.0339, Perplexity: 7.6440
Epoch [3/3], Step [2823/3236], Loss: 2.1320, Perplexity: 8.4315
Epoch [3/3], Step [2824/3236], Loss: 1.9271, Perplexity: 6.8692
Epoch [3/3], Step [2825/3236], Loss: 2.5403, Perplexity: 12.6829
Epoch [3/3], Step [2826/3236], Loss: 1.9343, Perplexity: 6.9189
Epoch [3/3], Step [2827/3236], Loss: 2.1084, Perplexity: 8.2347
Epoch [3/3], Step [2828/3236], Loss: 1.9566, Perplexity: 7.0752
Epoch [3/3], Step [2829/3236], Loss: 2.0407, Perplexity: 7.6958
Epoch [3/3], Step [2830/3236], Loss: 1.9636, Perplexity: 7.1249
Epoch [3/3], Step [2831/3236], Loss: 2.5815, Perplexity: 13.2173
Epoch [3/3], Step [2832/3236], Loss: 1.9950, Perplexity: 7.3522
Epoch [3/3], Step [2833/3236], Loss: 2.0005, Perplexity: 7.3925
Epoch [3/3], Step [2834/3236], Loss: 2.0725, Perplexity: 7.9446
Epoch [3/3], Step [2835/3236], Loss: 2.0308, Perplexity: 7.6204
Epoch [3/3], Step [2836/3236], Loss: 2.0867, Perplexity: 8.0583
Epoch [3/3], Step [2837/3236], Loss: 2.1999, Perplexity: 9.0244
Epoch [3/3], Step [2838/3236], Loss: 2.2446, Perplexity: 9.4366
Epoch [3/3], Step [2839/3236], Loss: 1.9630, Perplexity: 7.1205
Epoch [3/3], Step [2840/3236], Loss: 1.9993, Perplexity: 7.3837
Epoch [3/3], Step [2841/3236], Loss: 2.0230, Perplexity: 7.5613
Epoch [3/3], Step [2842/3236], Loss: 1.9083, Perplexity: 6.7413
Epoch [3/3], Step [2843/3236], Loss: 2.0846, Perplexity: 8.0411
Epoch [3/3], Step [2844/3236], Loss: 1.9351, Perplexity: 6.9249
Epoch [3/3], Step [2845/3236], Loss: 2.0157, Perplexity: 7.5058
Epoch [3/3], Step [2846/3236], Loss: 2.3082, Perplexity: 10.0563
Epoch [3/3], Step [2847/3236], Loss: 2.1595, Perplexity: 8.6667
Epoch [3/3], Step [2848/3236], Loss: 2.2367, Perplexity: 9.3627
Epoch [3/3], Step [2849/3236], Loss: 2.1862, Perplexity: 8.9009
Epoch [3/3], Step [2850/3236], Loss: 2.2862, Perplexity: 9.8379
Epoch [3/3], Step [2851/3236], Loss: 2.0748, Perplexity: 7.9632
Epoch [3/3], Step [2852/3236], Loss: 1.9375, Perplexity: 6.9416
Epoch [3/3], Step [2853/3236], Loss: 1.9604, Perplexity: 7.1024
Epoch [3/3], Step [2854/3236], Loss: 2.0371, Perplexity: 7.6682
Epoch [3/3], Step [2855/3236], Loss: 2.0178, Perplexity: 7.5221
Epoch [3/3], Step [2856/3236], Loss: 1.9396, Perplexity: 6.9557
Epoch [3/3], Step [2857/3236], Loss: 2.0353, Perplexity: 7.6546
Epoch [3/3], Step [2858/3236], Loss: 1.9521, Perplexity: 7.0432
Epoch [3/3], Step [2859/3236], Loss: 1.9466, Perplexity: 7.0050
Epoch [3/3], Step [2860/3236], Loss: 2.6269, Perplexity: 13.8314
Epoch [3/3], Step [2861/3236], Loss: 2.0177, Perplexity: 7.5208
Epoch [3/3], Step [2862/3236], Loss: 2.0750, Perplexity: 7.9642
Epoch [3/3], Step [2863/3236], Loss: 2.2831, Perplexity: 9.8068
Epoch [3/3], Step [2864/3236], Loss: 2.0356, Perplexity: 7.6571
Epoch [3/3], Step [2865/3236], Loss: 1.9363, Perplexity: 6.9329
Epoch [3/3], Step [2866/3236], Loss: 2.6407, Perplexity: 14.0225
Epoch [3/3], Step [2867/3236], Loss: 2.0391, Perplexity: 7.6839
Epoch [3/3], Step [2868/3236], Loss: 1.9958, Perplexity: 7.3579
Epoch [3/3], Step [2869/3236], Loss: 1.9304, Perplexity: 6.8920
Epoch [3/3], Step [2870/3236], Loss: 1.9303, Perplexity: 6.8916
Epoch [3/3], Step [2871/3236], Loss: 1.9311, Perplexity: 6.8968
Epoch [3/3], Step [2872/3236], Loss: 2.4554, Perplexity: 11.6509
Epoch [3/3], Step [2873/3236], Loss: 1.9904, Perplexity: 7.3185
Epoch [3/3], Step [2874/3236], Loss: 1.9085, Perplexity: 6.7430
Epoch [3/3], Step [2875/3236], Loss: 1.9456, Perplexity: 6.9981
Epoch [3/3], Step [2876/3236], Loss: 1.9351, Perplexity: 6.9244
Epoch [3/3], Step [2877/3236], Loss: 1.9789, Perplexity: 7.2345
Epoch [3/3], Step [2878/3236], Loss: 1.8476, Perplexity: 6.3445
Epoch [3/3], Step [2879/3236], Loss: 2.0084, Perplexity: 7.4513
Epoch [3/3], Step [2880/3236], Loss: 1.9844, Perplexity: 7.2747
Epoch [3/3], Step [2881/3236], Loss: 2.0841, Perplexity: 8.0374
Epoch [3/3], Step [2882/3236], Loss: 1.9504, Perplexity: 7.0316
Epoch [3/3], Step [2883/3236], Loss: 1.9780, Perplexity: 7.2284
Epoch [3/3], Step [2884/3236], Loss: 1.9892, Perplexity: 7.3099
Epoch [3/3], Step [2885/3236], Loss: 2.2085, Perplexity: 9.1020
Epoch [3/3], Step [2886/3236], Loss: 2.0454, Perplexity: 7.7325
Epoch [3/3], Step [2887/3236], Loss: 2.3847, Perplexity: 10.8553
Epoch [3/3], Step [2888/3236], Loss: 2.0097, Perplexity: 7.4610
Epoch [3/3], Step [2889/3236], Loss: 2.4225, Perplexity: 11.2736
Epoch [3/3], Step [2890/3236], Loss: 2.1261, Perplexity: 8.3820
Epoch [3/3], Step [2891/3236], Loss: 2.0057, Perplexity: 7.4314
Epoch [3/3], Step [2892/3236], Loss: 2.0380, Perplexity: 7.6750
Epoch [3/3], Step [2893/3236], Loss: 1.9153, Perplexity: 6.7892
Epoch [3/3], Step [2894/3236], Loss: 1.9462, Perplexity: 7.0020
Epoch [3/3], Step [2895/3236], Loss: 1.9117, Perplexity: 6.7647
Epoch [3/3], Step [2896/3236], Loss: 2.3484, Perplexity: 10.4692
Epoch [3/3], Step [2897/3236], Loss: 2.0498, Perplexity: 7.7665
Epoch [3/3], Step [2898/3236], Loss: 1.9049, Perplexity: 6.7186
Epoch [3/3], Step [2899/3236], Loss: 2.5413, Perplexity: 12.6965
Epoch [3/3], Step [2900/3236], Loss: 2.1325, Perplexity: 8.4358
Epoch [3/3], Step [2901/3236], Loss: 1.9790, Perplexity: 7.2357
Epoch [3/3], Step [2902/3236], Loss: 2.0334, Perplexity: 7.6398
Epoch [3/3], Step [2903/3236], Loss: 2.1017, Perplexity: 8.1802
Epoch [3/3], Step [2904/3236], Loss: 1.9826, Perplexity: 7.2618
Epoch [3/3], Step [2905/3236], Loss: 1.8938, Perplexity: 6.6446
Epoch [3/3], Step [2906/3236], Loss: 1.9476, Perplexity: 7.0121
Epoch [3/3], Step [2907/3236], Loss: 1.9746, Perplexity: 7.2040
Epoch [3/3], Step [2908/3236], Loss: 1.8453, Perplexity: 6.3297
Epoch [3/3], Step [2909/3236], Loss: 1.9991, Perplexity: 7.3825
Epoch [3/3], Step [2910/3236], Loss: 2.1616, Perplexity: 8.6846
Epoch [3/3], Step [2911/3236], Loss: 1.9368, Perplexity: 6.9365
Epoch [3/3], Step [2912/3236], Loss: 2.1331, Perplexity: 8.4406
Epoch [3/3], Step [2913/3236], Loss: 1.9133, Perplexity: 6.7754
Epoch [3/3], Step [2914/3236], Loss: 1.9608, Perplexity: 7.1047
Epoch [3/3], Step [2915/3236], Loss: 2.2370, Perplexity: 9.3651
Epoch [3/3], Step [2916/3236], Loss: 2.1352, Perplexity: 8.4588
Epoch [3/3], Step [2917/3236], Loss: 2.4369, Perplexity: 11.4373
Epoch [3/3], Step [2918/3236], Loss: 1.9462, Perplexity: 7.0022
Epoch [3/3], Step [2919/3236], Loss: 1.9024, Perplexity: 6.7017
Epoch [3/3], Step [2920/3236], Loss: 1.9347, Perplexity: 6.9218
Epoch [3/3], Step [2921/3236], Loss: 2.0364, Perplexity: 7.6632
Epoch [3/3], Step [2922/3236], Loss: 2.2991, Perplexity: 9.9653
Epoch [3/3], Step [2923/3236], Loss: 2.5490, Perplexity: 12.7949
Epoch [3/3], Step [2924/3236], Loss: 2.0027, Perplexity: 7.4091
Epoch [3/3], Step [2925/3236], Loss: 2.3361, Perplexity: 10.3413
Epoch [3/3], Step [2926/3236], Loss: 2.1269, Perplexity: 8.3887
Epoch [3/3], Step [2927/3236], Loss: 2.2832, Perplexity: 9.8082
Epoch [3/3], Step [2928/3236], Loss: 1.9418, Perplexity: 6.9712
Epoch [3/3], Step [2929/3236], Loss: 2.3017, Perplexity: 9.9910
Epoch [3/3], Step [2930/3236], Loss: 2.1322, Perplexity: 8.4336
Epoch [3/3], Step [2931/3236], Loss: 1.9055, Perplexity: 6.7226
Epoch [3/3], Step [2932/3236], Loss: 2.1256, Perplexity: 8.3782
Epoch [3/3], Step [2933/3236], Loss: 2.0139, Perplexity: 7.4924
Epoch [3/3], Step [2934/3236], Loss: 2.0774, Perplexity: 7.9840
Epoch [3/3], Step [2935/3236], Loss: 2.0326, Perplexity: 7.6337
Epoch [3/3], Step [2936/3236], Loss: 1.8936, Perplexity: 6.6431
Epoch [3/3], Step [2937/3236], Loss: 1.9312, Perplexity: 6.8981
Epoch [3/3], Step [2938/3236], Loss: 2.5300, Perplexity: 12.5540
Epoch [3/3], Step [2939/3236], Loss: 2.0304, Perplexity: 7.6172
Epoch [3/3], Step [2940/3236], Loss: 2.1071, Perplexity: 8.2240
Epoch [3/3], Step [2941/3236], Loss: 2.1537, Perplexity: 8.6163
Epoch [3/3], Step [2942/3236], Loss: 1.9807, Perplexity: 7.2476
Epoch [3/3], Step [2943/3236], Loss: 2.0070, Perplexity: 7.4411
Epoch [3/3], Step [2944/3236], Loss: 2.0165, Perplexity: 7.5123
Epoch [3/3], Step [2945/3236], Loss: 2.0234, Perplexity: 7.5640
Epoch [3/3], Step [2946/3236], Loss: 2.0643, Perplexity: 7.8799
Epoch [3/3], Step [2947/3236], Loss: 2.0503, Perplexity: 7.7703
Epoch [3/3], Step [2948/3236], Loss: 2.0788, Perplexity: 7.9949
Epoch [3/3], Step [2949/3236], Loss: 1.9118, Perplexity: 6.7653
Epoch [3/3], Step [2950/3236], Loss: 2.3228, Perplexity: 10.2044
Epoch [3/3], Step [2951/3236], Loss: 2.0661, Perplexity: 7.8943
Epoch [3/3], Step [2952/3236], Loss: 1.9519, Perplexity: 7.0419
Epoch [3/3], Step [2953/3236], Loss: 1.9139, Perplexity: 6.7796
Epoch [3/3], Step [2954/3236], Loss: 1.9542, Perplexity: 7.0582
Epoch [3/3], Step [2955/3236], Loss: 1.8386, Perplexity: 6.2877
Epoch [3/3], Step [2956/3236], Loss: 2.0427, Perplexity: 7.7113
Epoch [3/3], Step [2957/3236], Loss: 2.3279, Perplexity: 10.2561
Epoch [3/3], Step [2958/3236], Loss: 1.9053, Perplexity: 6.7217
Epoch [3/3], Step [2959/3236], Loss: 2.0204, Perplexity: 7.5413
Epoch [3/3], Step [2960/3236], Loss: 2.4151, Perplexity: 11.1908
Epoch [3/3], Step [2961/3236], Loss: 2.2213, Perplexity: 9.2191
Epoch [3/3], Step [2962/3236], Loss: 1.8898, Perplexity: 6.6183
Epoch [3/3], Step [2963/3236], Loss: 2.1074, Perplexity: 8.2272
Epoch [3/3], Step [2964/3236], Loss: 1.9319, Perplexity: 6.9027
Epoch [3/3], Step [2965/3236], Loss: 2.0384, Perplexity: 7.6786
Epoch [3/3], Step [2966/3236], Loss: 1.8813, Perplexity: 6.5619
Epoch [3/3], Step [2967/3236], Loss: 1.9694, Perplexity: 7.1663
Epoch [3/3], Step [2968/3236], Loss: 1.9780, Perplexity: 7.2283
Epoch [3/3], Step [2969/3236], Loss: 2.0703, Perplexity: 7.9272
Epoch [3/3], Step [2970/3236], Loss: 1.9447, Perplexity: 6.9913
Epoch [3/3], Step [2971/3236], Loss: 2.0804, Perplexity: 8.0074
Epoch [3/3], Step [2972/3236], Loss: 1.9972, Perplexity: 7.3680
Epoch [3/3], Step [2973/3236], Loss: 2.1160, Perplexity: 8.2976
Epoch [3/3], Step [2974/3236], Loss: 2.1605, Perplexity: 8.6753
Epoch [3/3], Step [2975/3236], Loss: 1.9878, Perplexity: 7.2998
Epoch [3/3], Step [2976/3236], Loss: 1.8122, Perplexity: 6.1241
Epoch [3/3], Step [2977/3236], Loss: 2.1942, Perplexity: 8.9727
Epoch [3/3], Step [2978/3236], Loss: 2.5765, Perplexity: 13.1512
Epoch [3/3], Step [2979/3236], Loss: 1.9537, Perplexity: 7.0544
Epoch [3/3], Step [2980/3236], Loss: 2.1667, Perplexity: 8.7297
Epoch [3/3], Step [2981/3236], Loss: 1.9305, Perplexity: 6.8929
Epoch [3/3], Step [2982/3236], Loss: 1.9158, Perplexity: 6.7924
Epoch [3/3], Step [2983/3236], Loss: 2.2242, Perplexity: 9.2463
Epoch [3/3], Step [2984/3236], Loss: 1.9514, Perplexity: 7.0389
Epoch [3/3], Step [2985/3236], Loss: 1.8893, Perplexity: 6.6149
Epoch [3/3], Step [2986/3236], Loss: 2.1688, Perplexity: 8.7477
Epoch [3/3], Step [2987/3236], Loss: 1.9946, Perplexity: 7.3491
Epoch [3/3], Step [2988/3236], Loss: 1.9437, Perplexity: 6.9846
Epoch [3/3], Step [2989/3236], Loss: 1.9987, Perplexity: 7.3795
Epoch [3/3], Step [2990/3236], Loss: 1.9438, Perplexity: 6.9851
Epoch [3/3], Step [2991/3236], Loss: 2.6094, Perplexity: 13.5906
Epoch [3/3], Step [2992/3236], Loss: 1.9520, Perplexity: 7.0425
Epoch [3/3], Step [2993/3236], Loss: 2.1410, Perplexity: 8.5078
Epoch [3/3], Step [2994/3236], Loss: 2.5629, Perplexity: 12.9733
Epoch [3/3], Step [2995/3236], Loss: 2.0016, Perplexity: 7.4007
Epoch [3/3], Step [2996/3236], Loss: 1.9172, Perplexity: 6.8017
Epoch [3/3], Step [2997/3236], Loss: 2.0851, Perplexity: 8.0456
Epoch [3/3], Step [2998/3236], Loss: 2.6525, Perplexity: 14.1902
Epoch [3/3], Step [2999/3236], Loss: 2.3510, Perplexity: 10.4959
Epoch [3/3], Step [3000/3236], Loss: 2.0486, Perplexity: 7.7570
Epoch [3/3], Step [3001/3236], Loss: 1.9956, Perplexity: 7.3568
Epoch [3/3], Step [3002/3236], Loss: 2.0686, Perplexity: 7.9135
Epoch [3/3], Step [3003/3236], Loss: 2.0637, Perplexity: 7.8749
Epoch [3/3], Step [3004/3236], Loss: 2.0614, Perplexity: 7.8568
Epoch [3/3], Step [3005/3236], Loss: 2.0335, Perplexity: 7.6410
Epoch [3/3], Step [3006/3236], Loss: 2.0669, Perplexity: 7.9003
Epoch [3/3], Step [3007/3236], Loss: 2.1309, Perplexity: 8.4227
Epoch [3/3], Step [3008/3236], Loss: 1.9766, Perplexity: 7.2185
Epoch [3/3], Step [3009/3236], Loss: 2.0636, Perplexity: 7.8741
Epoch [3/3], Step [3010/3236], Loss: 2.6458, Perplexity: 14.0945
Epoch [3/3], Step [3011/3236], Loss: 2.0375, Perplexity: 7.6711
Epoch [3/3], Step [3012/3236], Loss: 2.1844, Perplexity: 8.8857
Epoch [3/3], Step [3013/3236], Loss: 1.9954, Perplexity: 7.3548
Epoch [3/3], Step [3014/3236], Loss: 1.8776, Perplexity: 6.5376
Epoch [3/3], Step [3015/3236], Loss: 2.0250, Perplexity: 7.5760
Epoch [3/3], Step [3016/3236], Loss: 1.8906, Perplexity: 6.6233
Epoch [3/3], Step [3017/3236], Loss: 2.2394, Perplexity: 9.3877
Epoch [3/3], Step [3018/3236], Loss: 1.8950, Perplexity: 6.6523
Epoch [3/3], Step [3019/3236], Loss: 2.1678, Perplexity: 8.7391
Epoch [3/3], Step [3020/3236], Loss: 2.0334, Perplexity: 7.6399
Epoch [3/3], Step [3021/3236], Loss: 1.9973, Perplexity: 7.3690
Epoch [3/3], Step [3022/3236], Loss: 2.0101, Perplexity: 7.4637
Epoch [3/3], Step [3023/3236], Loss: 1.8965, Perplexity: 6.6626
Epoch [3/3], Step [3024/3236], Loss: 1.8719, Perplexity: 6.5009
Epoch [3/3], Step [3025/3236], Loss: 1.9581, Perplexity: 7.0860
Epoch [3/3], Step [3026/3236], Loss: 2.0578, Perplexity: 7.8284
Epoch [3/3], Step [3027/3236], Loss: 1.9192, Perplexity: 6.8153
Epoch [3/3], Step [3028/3236], Loss: 1.9549, Perplexity: 7.0631
Epoch [3/3], Step [3029/3236], Loss: 2.0164, Perplexity: 7.5114
Epoch [3/3], Step [3030/3236], Loss: 2.0253, Perplexity: 7.5787
Epoch [3/3], Step [3031/3236], Loss: 2.7408, Perplexity: 15.4992
Epoch [3/3], Step [3032/3236], Loss: 2.1627, Perplexity: 8.6950
Epoch [3/3], Step [3033/3236], Loss: 2.0328, Perplexity: 7.6355
Epoch [3/3], Step [3034/3236], Loss: 1.9325, Perplexity: 6.9070
Epoch [3/3], Step [3035/3236], Loss: 2.0773, Perplexity: 7.9833
Epoch [3/3], Step [3036/3236], Loss: 1.9980, Perplexity: 7.3745
Epoch [3/3], Step [3037/3236], Loss: 2.1899, Perplexity: 8.9342
Epoch [3/3], Step [3038/3236], Loss: 2.1327, Perplexity: 8.4376
Epoch [3/3], Step [3039/3236], Loss: 2.0387, Perplexity: 7.6805
Epoch [3/3], Step [3040/3236], Loss: 1.9714, Perplexity: 7.1808
Epoch [3/3], Step [3041/3236], Loss: 1.9117, Perplexity: 6.7648
Epoch [3/3], Step [3042/3236], Loss: 1.9455, Perplexity: 6.9970
Epoch [3/3], Step [3043/3236], Loss: 2.0588, Perplexity: 7.8364
Epoch [3/3], Step [3044/3236], Loss: 2.0258, Perplexity: 7.5825
Epoch [3/3], Step [3045/3236], Loss: 2.1101, Perplexity: 8.2491
Epoch [3/3], Step [3046/3236], Loss: 2.1103, Perplexity: 8.2507
Epoch [3/3], Step [3047/3236], Loss: 1.8773, Perplexity: 6.5356
Epoch [3/3], Step [3048/3236], Loss: 1.8766, Perplexity: 6.5313
Epoch [3/3], Step [3049/3236], Loss: 2.0515, Perplexity: 7.7795
Epoch [3/3], Step [3050/3236], Loss: 1.8046, Perplexity: 6.0774
Epoch [3/3], Step [3051/3236], Loss: 1.8520, Perplexity: 6.3725
Epoch [3/3], Step [3052/3236], Loss: 2.0325, Perplexity: 7.6334
Epoch [3/3], Step [3053/3236], Loss: 1.9353, Perplexity: 6.9264
Epoch [3/3], Step [3054/3236], Loss: 1.8481, Perplexity: 6.3476
Epoch [3/3], Step [3055/3236], Loss: 2.0850, Perplexity: 8.0449
Epoch [3/3], Step [3056/3236], Loss: 2.1337, Perplexity: 8.4458
Epoch [3/3], Step [3057/3236], Loss: 1.9765, Perplexity: 7.2172
Epoch [3/3], Step [3058/3236], Loss: 2.0457, Perplexity: 7.7348
Epoch [3/3], Step [3059/3236], Loss: 1.9648, Perplexity: 7.1337
Epoch [3/3], Step [3060/3236], Loss: 2.0559, Perplexity: 7.8140
Epoch [3/3], Step [3061/3236], Loss: 2.5818, Perplexity: 13.2215
Epoch [3/3], Step [3062/3236], Loss: 2.2805, Perplexity: 9.7811
Epoch [3/3], Step [3063/3236], Loss: 2.0868, Perplexity: 8.0587
Epoch [3/3], Step [3064/3236], Loss: 2.2751, Perplexity: 9.7285
Epoch [3/3], Step [3065/3236], Loss: 1.9294, Perplexity: 6.8854
Epoch [3/3], Step [3066/3236], Loss: 2.1721, Perplexity: 8.7764
Epoch [3/3], Step [3067/3236], Loss: 1.9277, Perplexity: 6.8737
Epoch [3/3], Step [3068/3236], Loss: 2.0013, Perplexity: 7.3990
Epoch [3/3], Step [3069/3236], Loss: 2.1111, Perplexity: 8.2575
Epoch [3/3], Step [3070/3236], Loss: 1.8886, Perplexity: 6.6099
Epoch [3/3], Step [3071/3236], Loss: 2.1512, Perplexity: 8.5950
Epoch [3/3], Step [3072/3236], Loss: 2.0840, Perplexity: 8.0362
Epoch [3/3], Step [3073/3236], Loss: 1.9183, Perplexity: 6.8095
Epoch [3/3], Step [3074/3236], Loss: 2.0760, Perplexity: 7.9729
Epoch [3/3], Step [3075/3236], Loss: 1.8186, Perplexity: 6.1630
Epoch [3/3], Step [3076/3236], Loss: 1.9626, Perplexity: 7.1179
Epoch [3/3], Step [3077/3236], Loss: 2.5567, Perplexity: 12.8926
Epoch [3/3], Step [3078/3236], Loss: 2.1836, Perplexity: 8.8779
Epoch [3/3], Step [3079/3236], Loss: 2.0711, Perplexity: 7.9333
Epoch [3/3], Step [3080/3236], Loss: 1.9942, Perplexity: 7.3464
Epoch [3/3], Step [3081/3236], Loss: 2.1189, Perplexity: 8.3218
Epoch [3/3], Step [3082/3236], Loss: 2.0300, Perplexity: 7.6144
Epoch [3/3], Step [3083/3236], Loss: 2.0662, Perplexity: 7.8951
Epoch [3/3], Step [3084/3236], Loss: 1.8682, Perplexity: 6.4766
Epoch [3/3], Step [3085/3236], Loss: 1.8997, Perplexity: 6.6839
Epoch [3/3], Step [3086/3236], Loss: 2.3041, Perplexity: 10.0149
Epoch [3/3], Step [3087/3236], Loss: 1.9963, Perplexity: 7.3620
Epoch [3/3], Step [3088/3236], Loss: 1.9869, Perplexity: 7.2932
Epoch [3/3], Step [3089/3236], Loss: 2.1066, Perplexity: 8.2202
Epoch [3/3], Step [3090/3236], Loss: 2.0793, Perplexity: 7.9988
Epoch [3/3], Step [3091/3236], Loss: 2.0357, Perplexity: 7.6575
Epoch [3/3], Step [3092/3236], Loss: 1.9919, Perplexity: 7.3296
Epoch [3/3], Step [3093/3236], Loss: 2.1079, Perplexity: 8.2309
Epoch [3/3], Step [3094/3236], Loss: 1.8835, Perplexity: 6.5766
Epoch [3/3], Step [3095/3236], Loss: 2.2509, Perplexity: 9.4964
Epoch [3/3], Step [3096/3236], Loss: 2.3879, Perplexity: 10.8911
Epoch [3/3], Step [3097/3236], Loss: 1.9139, Perplexity: 6.7798
Epoch [3/3], Step [3098/3236], Loss: 1.9147, Perplexity: 6.7852
Epoch [3/3], Step [3099/3236], Loss: 2.6221, Perplexity: 13.7651
Epoch [3/3], Step [3100/3236], Loss: 1.9795, Perplexity: 7.2393
Epoch [3/3], Step [3101/3236], Loss: 1.9334, Perplexity: 6.9128
Epoch [3/3], Step [3102/3236], Loss: 1.9932, Perplexity: 7.3387
Epoch [3/3], Step [3103/3236], Loss: 1.9837, Perplexity: 7.2698
Epoch [3/3], Step [3104/3236], Loss: 2.0421, Perplexity: 7.7064
Epoch [3/3], Step [3105/3236], Loss: 2.1290, Perplexity: 8.4067
Epoch [3/3], Step [3106/3236], Loss: 2.1410, Perplexity: 8.5076
Epoch [3/3], Step [3107/3236], Loss: 1.9903, Perplexity: 7.3177
Epoch [3/3], Step [3108/3236], Loss: 2.0613, Perplexity: 7.8565
Epoch [3/3], Step [3109/3236], Loss: 1.9827, Perplexity: 7.2625
Epoch [3/3], Step [3110/3236], Loss: 2.1172, Perplexity: 8.3075
Epoch [3/3], Step [3111/3236], Loss: 1.9036, Perplexity: 6.7102
Epoch [3/3], Step [3112/3236], Loss: 1.8905, Perplexity: 6.6226
Epoch [3/3], Step [3113/3236], Loss: 2.2166, Perplexity: 9.1759
Epoch [3/3], Step [3114/3236], Loss: 2.0235, Perplexity: 7.5644
Epoch [3/3], Step [3115/3236], Loss: 1.8908, Perplexity: 6.6243
Epoch [3/3], Step [3116/3236], Loss: 2.2904, Perplexity: 9.8790
Epoch [3/3], Step [3117/3236], Loss: 1.9004, Perplexity: 6.6885
Epoch [3/3], Step [3118/3236], Loss: 2.3618, Perplexity: 10.6103
Epoch [3/3], Step [3119/3236], Loss: 1.9967, Perplexity: 7.3644
Epoch [3/3], Step [3120/3236], Loss: 2.0507, Perplexity: 7.7730
Epoch [3/3], Step [3121/3236], Loss: 2.0084, Perplexity: 7.4513
Epoch [3/3], Step [3122/3236], Loss: 2.0625, Perplexity: 7.8653
Epoch [3/3], Step [3123/3236], Loss: 1.8516, Perplexity: 6.3703
Epoch [3/3], Step [3124/3236], Loss: 2.0783, Perplexity: 7.9913
Epoch [3/3], Step [3125/3236], Loss: 1.8806, Perplexity: 6.5571
Epoch [3/3], Step [3126/3236], Loss: 2.0772, Perplexity: 7.9822
Epoch [3/3], Step [3127/3236], Loss: 2.0487, Perplexity: 7.7579
Epoch [3/3], Step [3128/3236], Loss: 2.0660, Perplexity: 7.8934
Epoch [3/3], Step [3129/3236], Loss: 1.9519, Perplexity: 7.0418
Epoch [3/3], Step [3130/3236], Loss: 2.0917, Perplexity: 8.0985
Epoch [3/3], Step [3131/3236], Loss: 2.0014, Perplexity: 7.3997
Epoch [3/3], Step [3132/3236], Loss: 2.0055, Perplexity: 7.4299
Epoch [3/3], Step [3133/3236], Loss: 2.0154, Perplexity: 7.5040
Epoch [3/3], Step [3134/3236], Loss: 2.0640, Perplexity: 7.8778
Epoch [3/3], Step [3135/3236], Loss: 1.9181, Perplexity: 6.8082
Epoch [3/3], Step [3136/3236], Loss: 1.9624, Perplexity: 7.1165
Epoch [3/3], Step [3137/3236], Loss: 2.0453, Perplexity: 7.7311
Epoch [3/3], Step [3138/3236], Loss: 1.8158, Perplexity: 6.1459
Epoch [3/3], Step [3139/3236], Loss: 2.2832, Perplexity: 9.8080
Epoch [3/3], Step [3140/3236], Loss: 1.8409, Perplexity: 6.3021
Epoch [3/3], Step [3141/3236], Loss: 1.9936, Perplexity: 7.3421
Epoch [3/3], Step [3142/3236], Loss: 2.1544, Perplexity: 8.6225
Epoch [3/3], Step [3143/3236], Loss: 2.0156, Perplexity: 7.5051
Epoch [3/3], Step [3144/3236], Loss: 1.9783, Perplexity: 7.2308
Epoch [3/3], Step [3145/3236], Loss: 1.8822, Perplexity: 6.5679
Epoch [3/3], Step [3146/3236], Loss: 2.5312, Perplexity: 12.5690
Epoch [3/3], Step [3147/3236], Loss: 1.8867, Perplexity: 6.5973
Epoch [3/3], Step [3148/3236], Loss: 2.1226, Perplexity: 8.3528
Epoch [3/3], Step [3149/3236], Loss: 1.9833, Perplexity: 7.2664
Epoch [3/3], Step [3150/3236], Loss: 2.1049, Perplexity: 8.2059
Epoch [3/3], Step [3151/3236], Loss: 1.9287, Perplexity: 6.8805
Epoch [3/3], Step [3152/3236], Loss: 1.9975, Perplexity: 7.3708
Epoch [3/3], Step [3153/3236], Loss: 1.9264, Perplexity: 6.8646
Epoch [3/3], Step [3154/3236], Loss: 2.1313, Perplexity: 8.4262
Epoch [3/3], Step [3155/3236], Loss: 1.8587, Perplexity: 6.4154
Epoch [3/3], Step [3156/3236], Loss: 2.0260, Perplexity: 7.5835
Epoch [3/3], Step [3157/3236], Loss: 2.0195, Perplexity: 7.5346
Epoch [3/3], Step [3158/3236], Loss: 1.9325, Perplexity: 6.9065
Epoch [3/3], Step [3159/3236], Loss: 1.8887, Perplexity: 6.6109
Epoch [3/3], Step [3160/3236], Loss: 1.8538, Perplexity: 6.3841
Epoch [3/3], Step [3161/3236], Loss: 2.0593, Perplexity: 7.8401
Epoch [3/3], Step [3162/3236], Loss: 2.1280, Perplexity: 8.3977
Epoch [3/3], Step [3163/3236], Loss: 1.9343, Perplexity: 6.9193
Epoch [3/3], Step [3164/3236], Loss: 2.1106, Perplexity: 8.2531
Epoch [3/3], Step [3165/3236], Loss: 1.9637, Perplexity: 7.1254
Epoch [3/3], Step [3166/3236], Loss: 2.0559, Perplexity: 7.8137
Epoch [3/3], Step [3167/3236], Loss: 2.0978, Perplexity: 8.1485
Epoch [3/3], Step [3168/3236], Loss: 1.8591, Perplexity: 6.4181
Epoch [3/3], Step [3169/3236], Loss: 2.1532, Perplexity: 8.6120
Epoch [3/3], Step [3170/3236], Loss: 1.8129, Perplexity: 6.1281
Epoch [3/3], Step [3171/3236], Loss: 2.3159, Perplexity: 10.1341
Epoch [3/3], Step [3172/3236], Loss: 2.0432, Perplexity: 7.7156
Epoch [3/3], Step [3173/3236], Loss: 1.8767, Perplexity: 6.5317
Epoch [3/3], Step [3174/3236], Loss: 2.0690, Perplexity: 7.9172
Epoch [3/3], Step [3175/3236], Loss: 2.3248, Perplexity: 10.2242
Epoch [3/3], Step [3176/3236], Loss: 2.0963, Perplexity: 8.1357
Epoch [3/3], Step [3177/3236], Loss: 2.0864, Perplexity: 8.0559
Epoch [3/3], Step [3178/3236], Loss: 2.1183, Perplexity: 8.3171
Epoch [3/3], Step [3179/3236], Loss: 2.2281, Perplexity: 9.2818
Epoch [3/3], Step [3180/3236], Loss: 1.9455, Perplexity: 6.9969
Epoch [3/3], Step [3181/3236], Loss: 2.0451, Perplexity: 7.7301
Epoch [3/3], Step [3182/3236], Loss: 1.9043, Perplexity: 6.7148
Epoch [3/3], Step [3183/3236], Loss: 1.8690, Perplexity: 6.4817
Epoch [3/3], Step [3184/3236], Loss: 2.0560, Perplexity: 7.8149
Epoch [3/3], Step [3185/3236], Loss: 2.0361, Perplexity: 7.6604
Epoch [3/3], Step [3186/3236], Loss: 1.9629, Perplexity: 7.1196
Epoch [3/3], Step [3187/3236], Loss: 1.9195, Perplexity: 6.8175
Epoch [3/3], Step [3188/3236], Loss: 1.9850, Perplexity: 7.2791
Epoch [3/3], Step [3189/3236], Loss: 1.9531, Perplexity: 7.0506
Epoch [3/3], Step [3190/3236], Loss: 2.0320, Perplexity: 7.6292
Epoch [3/3], Step [3191/3236], Loss: 2.0585, Perplexity: 7.8341
Epoch [3/3], Step [3192/3236], Loss: 2.1398, Perplexity: 8.4980
Epoch [3/3], Step [3193/3236], Loss: 1.9352, Perplexity: 6.9257
Epoch [3/3], Step [3194/3236], Loss: 2.0588, Perplexity: 7.8368
Epoch [3/3], Step [3195/3236], Loss: 1.9707, Perplexity: 7.1754
Epoch [3/3], Step [3196/3236], Loss: 2.1694, Perplexity: 8.7532
Epoch [3/3], Step [3197/3236], Loss: 1.9835, Perplexity: 7.2681
Epoch [3/3], Step [3198/3236], Loss: 1.9795, Perplexity: 7.2389
Epoch [3/3], Step [3199/3236], Loss: 1.9855, Perplexity: 7.2830
Epoch [3/3], Step [3200/3236], Loss: 2.2395, Perplexity: 9.3887
Epoch [3/3], Step [3201/3236], Loss: 1.9869, Perplexity: 7.2933
Epoch [3/3], Step [3202/3236], Loss: 1.9066, Perplexity: 6.7303
Epoch [3/3], Step [3203/3236], Loss: 2.0221, Perplexity: 7.5541
Epoch [3/3], Step [3204/3236], Loss: 2.0796, Perplexity: 8.0012
Epoch [3/3], Step [3205/3236], Loss: 2.3915, Perplexity: 10.9300
Epoch [3/3], Step [3206/3236], Loss: 2.0021, Perplexity: 7.4044
Epoch [3/3], Step [3207/3236], Loss: 2.1501, Perplexity: 8.5858
Epoch [3/3], Step [3208/3236], Loss: 2.0284, Perplexity: 7.6016
Epoch [3/3], Step [3209/3236], Loss: 1.9537, Perplexity: 7.0548
Epoch [3/3], Step [3210/3236], Loss: 1.9569, Perplexity: 7.0774
Epoch [3/3], Step [3211/3236], Loss: 2.0760, Perplexity: 7.9729
Epoch [3/3], Step [3212/3236], Loss: 2.0927, Perplexity: 8.1069
Epoch [3/3], Step [3213/3236], Loss: 2.1024, Perplexity: 8.1854
Epoch [3/3], Step [3214/3236], Loss: 2.0120, Perplexity: 7.4780
Epoch [3/3], Step [3215/3236], Loss: 2.0482, Perplexity: 7.7536
Epoch [3/3], Step [3216/3236], Loss: 2.0757, Perplexity: 7.9698
Epoch [3/3], Step [3217/3236], Loss: 2.0222, Perplexity: 7.5546
Epoch [3/3], Step [3218/3236], Loss: 1.9744, Perplexity: 7.2021
Epoch [3/3], Step [3219/3236], Loss: 2.2189, Perplexity: 9.1975
Epoch [3/3], Step [3220/3236], Loss: 2.1049, Perplexity: 8.2062
Epoch [3/3], Step [3221/3236], Loss: 1.9504, Perplexity: 7.0316
Epoch [3/3], Step [3222/3236], Loss: 2.1155, Perplexity: 8.2940
Epoch [3/3], Step [3223/3236], Loss: 1.9995, Perplexity: 7.3854
Epoch [3/3], Step [3224/3236], Loss: 2.9216, Perplexity: 18.5719
Epoch [3/3], Step [3225/3236], Loss: 1.9016, Perplexity: 6.6965
Epoch [3/3], Step [3226/3236], Loss: 2.2681, Perplexity: 9.6605
Epoch [3/3], Step [3227/3236], Loss: 2.2583, Perplexity: 9.5669
Epoch [3/3], Step [3228/3236], Loss: 1.9060, Perplexity: 6.7261
Epoch [3/3], Step [3229/3236], Loss: 2.0741, Perplexity: 7.9575
Epoch [3/3], Step [3230/3236], Loss: 1.9533, Perplexity: 7.0521
Epoch [3/3], Step [3231/3236], Loss: 1.9933, Perplexity: 7.3397
Epoch [3/3], Step [3232/3236], Loss: 2.9029, Perplexity: 18.2276
Epoch [3/3], Step [3233/3236], Loss: 2.3291, Perplexity: 10.2686
Epoch [3/3], Step [3234/3236], Loss: 1.9689, Perplexity: 7.1626
Epoch [3/3], Step [3235/3236], Loss: 1.9525, Perplexity: 7.0463
Epoch [3/3], Step [3236/3236], Loss: 2.2994, Perplexity: 9.9681