diff --git "a/exp/log/log-train-2022-12-09-10-39-23-2" "b/exp/log/log-train-2022-12-09-10-39-23-2"
new file mode 100644--- /dev/null
+++ "b/exp/log/log-train-2022-12-09-10-39-23-2"
@@ -0,0 +1,6042 @@
+2022-12-09 10:39:23,973 INFO [train.py:493] (2/8) Training started
+2022-12-09 10:39:23,974 INFO [train.py:494] (2/8) {'max_sent_len': 200, 'sos_id': 1, 'eos_id': 1, 'blank_id': 0, 'lr': 0.001, 'weight_decay': 1e-06, 'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 200, 'reset_interval': 2000, 'valid_interval': 1000, 'nhead': 8, 'embedding_dim': 768, 'encoder_dim': 768, 'dim_feedforward': 2048, 'dropout': 0.1, 'env_info': {'k2-version': '1.22', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': '96c9a2aece2a3a7633da07740e24fa3d96f5498c', 'k2-git-date': 'Thu Nov 10 08:14:02 2022', 'lhotse-version': '1.10.0', 'torch-version': '1.12.1', 'torch-cuda-available': True, 'torch-cuda-version': '11.6', 'python-version': '3.8', 'icefall-git-branch': 'transformer_lm', 'icefall-git-sha1': '5b028fe-dirty', 'icefall-git-date': 'Thu Dec 8 23:29:10 2022', 'icefall-path': '/ceph-data4/yangxiaoyu/softwares/icefall_development/icefall_zipformer_mvq', 'k2-path': '/ceph-data4/yangxiaoyu/softwares/anaconda3/envs/k2_latest/lib/python3.8/site-packages/k2/__init__.py', 'lhotse-path': '/ceph-data4/yangxiaoyu/softwares/anaconda3/envs/k2_latest/lib/python3.8/site-packages/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-1-0307195509-567fcb96d6-kdztg', 'IP address': '10.177.22.10'}, 'world_size': 8, 'master_port': 12354, 'tensorboard': True, 'num_epochs': 20, 'start_epoch': 0, 'exp_dir': PosixPath('transformer_lm/exp_full_libri_16layer_maxlen200_8gpu'), 'use_fp16': False, 'batch_size': 70, 'lm_data': './transformer_lm/libri_lm_training_bpe500/sorted-lm-data-libri-lm_maxlen200.pt', 'lm_data_valid': './transformer_lm/libri_lm_training_bpe500/sorted_lm_data-test.pt', 'vocab_size': 500, 'num_layers': 16, 'tie_weights': True, 'seed': 42}
+2022-12-09 10:39:23,974 INFO [train.py:505] (2/8) Device: cuda:2
+2022-12-09 10:39:23,974 INFO [train.py:507] (2/8) About to create model
+2022-12-09 10:39:24,476 INFO [model.py:64] (2/8) Tying weights
+2022-12-09 10:39:24,477 INFO [train.py:520] (2/8) Number of model parameters: 98611638
+2022-12-09 10:39:27,367 INFO [train.py:539] (2/8) Loading LM training data from ./transformer_lm/libri_lm_training_bpe500/sorted-lm-data-libri-lm_maxlen200.pt
+2022-12-09 10:39:41,359 INFO [train.py:546] (2/8) Loading LM validation data from ./transformer_lm/libri_lm_training_bpe500/sorted_lm_data-test.pt
+2022-12-09 10:39:44,759 INFO [train.py:421] (2/8) Epoch 0, batch 0, loss[loss=79.79, over 2450.00 frames. , ppl: 4.49693072748638e+34] tot_loss[loss=79.79, over 2450.00 frames. , ppl: 4.49693072748638e+34], batch size: 70
+2022-12-09 10:39:44,761 INFO [distributed.py:995] (2/8) Reducer buckets have been rebuilt in this iteration.
+2022-12-09 10:41:21,499 INFO [train.py:421] (2/8) Epoch 0, batch 200, loss[loss=8.194, over 2590.00 frames. , ppl: 3619.6735340733685] tot_loss[loss=17.58, over 501285.51 frames. , ppl: 43071984.177458726], batch size: 70
+2022-12-09 10:43:01,687 INFO [train.py:421] (2/8) Epoch 0, batch 400, loss[loss=6.573, over 4900.00 frames. , ppl: 715.6602309707547] tot_loss[loss=12.13, over 984727.61 frames. , ppl: 185364.9066801909], batch size: 70
+2022-12-09 10:44:42,142 INFO [train.py:421] (2/8) Epoch 0, batch 600, loss[loss=6.418, over 4060.00 frames. , ppl: 612.9517693059069] tot_loss[loss=10, over 1432237.02 frames. , ppl: 22133.24061194677], batch size: 70
+2022-12-09 10:46:22,383 INFO [train.py:421] (2/8) Epoch 0, batch 800, loss[loss=5.799, over 630.00 frames. , ppl: 330.12962060913645] tot_loss[loss=8.93, over 1788683.70 frames. , ppl: 7556.819372178538], batch size: 70
+2022-12-09 10:48:02,280 INFO [train.py:421] (2/8) Epoch 0, batch 1000, loss[loss=5.464, over 2100.00 frames. , ppl: 235.95128683975] tot_loss[loss=8.131, over 2101123.55 frames. , ppl: 3398.9755677606295], batch size: 70
+2022-12-09 10:48:02,281 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 10:48:03,028 INFO [train.py:452] (2/8) Epoch 0, validation: loss=5.053, over 211138.00 frames. , ppl: 156.50471436114304
+2022-12-09 10:49:43,730 INFO [train.py:421] (2/8) Epoch 0, batch 1200, loss[loss=5.243, over 6300.00 frames. , ppl: 189.235976971383] tot_loss[loss=7.601, over 2414589.32 frames. , ppl: 1999.7061705357546], batch size: 70
+2022-12-09 10:51:23,207 INFO [train.py:421] (2/8) Epoch 0, batch 1400, loss[loss=5.178, over 3360.00 frames. , ppl: 177.3476998673368] tot_loss[loss=7.191, over 2691219.58 frames. , ppl: 1327.6435237501441], batch size: 70
+2022-12-09 10:53:02,875 INFO [train.py:421] (2/8) Epoch 0, batch 1600, loss[loss=4.793, over 2660.00 frames. , ppl: 120.60412773169607] tot_loss[loss=6.813, over 2949554.84 frames. , ppl: 909.4089720596776], batch size: 70
+2022-12-09 10:54:41,870 INFO [train.py:421] (2/8) Epoch 0, batch 1800, loss[loss=4.467, over 2450.00 frames. , ppl: 87.12915705526105] tot_loss[loss=6.435, over 3246211.69 frames. , ppl: 623.3146641166776], batch size: 70
+2022-12-09 10:56:19,726 INFO [train.py:421] (2/8) Epoch 0, batch 2000, loss[loss=4.419, over 2730.00 frames. , ppl: 83.03616195031789] tot_loss[loss=6.128, over 3488801.60 frames. , ppl: 458.32818527572744], batch size: 70
+2022-12-09 10:56:19,726 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 10:56:20,475 INFO [train.py:452] (2/8) Epoch 0, validation: loss=4.224, over 211138.00 frames. , ppl: 68.31401714385908
+2022-12-09 10:58:02,726 INFO [train.py:421] (2/8) Epoch 0, batch 2200, loss[loss=4.437, over 1890.00 frames. , ppl: 84.5199358590285] tot_loss[loss=5.88, over 3720270.14 frames. , ppl: 357.6622779174864], batch size: 70
+2022-12-09 10:59:45,477 INFO [train.py:421] (2/8) Epoch 0, batch 2400, loss[loss=4.223, over 1820.00 frames. , ppl: 68.2686126332504] tot_loss[loss=5.654, over 3977718.67 frames. , ppl: 285.54702791042223], batch size: 70
+2022-12-09 11:01:26,791 INFO [train.py:421] (2/8) Epoch 0, batch 2600, loss[loss=4.29, over 770.00 frames. , ppl: 72.95590112025636] tot_loss[loss=5.48, over 4103421.70 frames. , ppl: 239.87165915577927], batch size: 70
+2022-12-09 11:03:09,434 INFO [train.py:421] (2/8) Epoch 0, batch 2800, loss[loss=4.008, over 7350.00 frames. , ppl: 55.01499371357638] tot_loss[loss=5.307, over 4243157.50 frames. , ppl: 201.7851413353427], batch size: 70
+2022-12-09 11:04:49,031 INFO [train.py:421] (2/8) Epoch 0, batch 3000, loss[loss=3.741, over 1260.00 frames. , ppl: 42.12656243469766] tot_loss[loss=5.142, over 4337171.34 frames. , ppl: 171.02603206221232], batch size: 70
+2022-12-09 11:04:49,031 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:04:49,779 INFO [train.py:452] (2/8) Epoch 0, validation: loss=3.608, over 211138.00 frames. , ppl: 36.89311779083857
+2022-12-09 11:06:26,626 INFO [train.py:421] (2/8) Epoch 0, batch 3200, loss[loss=3.68, over 1120.00 frames. , ppl: 39.65827373027577] tot_loss[loss=4.972, over 4425552.91 frames. , ppl: 144.2893015012269], batch size: 70
+2022-12-09 11:08:06,512 INFO [train.py:421] (2/8) Epoch 0, batch 3400, loss[loss=3.46, over 5040.00 frames. , ppl: 31.805908685086347] tot_loss[loss=4.81, over 4508403.76 frames. , ppl: 122.69904458012519], batch size: 70
+2022-12-09 11:09:49,633 INFO [train.py:421] (2/8) Epoch 0, batch 3600, loss[loss=3.846, over 560.00 frames. , ppl: 46.82063102873267] tot_loss[loss=4.646, over 4612049.50 frames. , ppl: 104.142654058839], batch size: 70
+2022-12-09 11:11:30,340 INFO [train.py:421] (2/8) Epoch 0, batch 3800, loss[loss=3.21, over 5600.00 frames. , ppl: 24.768562188095803] tot_loss[loss=4.492, over 4707549.98 frames. , ppl: 89.26988526407024], batch size: 70
+2022-12-09 11:13:10,810 INFO [train.py:421] (2/8) Epoch 0, batch 4000, loss[loss=3.146, over 1190.00 frames. , ppl: 23.24329282498255] tot_loss[loss=4.345, over 4808005.67 frames. , ppl: 77.1200272169229], batch size: 70
+2022-12-09 11:13:10,811 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:13:11,561 INFO [train.py:452] (2/8) Epoch 0, validation: loss=3.125, over 211138.00 frames. , ppl: 22.762158928396982
+2022-12-09 11:14:48,443 INFO [train.py:421] (2/8) Epoch 0, batch 4200, loss[loss=3.18, over 2310.00 frames. , ppl: 24.04648720776747] tot_loss[loss=4.221, over 4861926.66 frames. , ppl: 68.0878110084979], batch size: 70
+2022-12-09 11:16:24,327 INFO [train.py:421] (2/8) Epoch 0, batch 4400, loss[loss=3.045, over 4620.00 frames. , ppl: 21.009233607087715] tot_loss[loss=4.11, over 4887822.36 frames. , ppl: 60.972451860691436], batch size: 70
+2022-12-09 11:18:03,018 INFO [train.py:421] (2/8) Epoch 0, batch 4600, loss[loss=3.138, over 1750.00 frames. , ppl: 23.04915031907031] tot_loss[loss=4.008, over 4911182.18 frames. , ppl: 55.062202565647134], batch size: 70
+2022-12-09 11:19:43,496 INFO [train.py:421] (2/8) Epoch 0, batch 4800, loss[loss=3.214, over 1050.00 frames. , ppl: 24.886517025000263] tot_loss[loss=3.899, over 5007583.44 frames. , ppl: 49.3756832094654], batch size: 70
+2022-12-09 11:21:23,999 INFO [train.py:421] (2/8) Epoch 0, batch 5000, loss[loss=2.939, over 4620.00 frames. , ppl: 18.904997027325326] tot_loss[loss=3.811, over 5038418.85 frames. , ppl: 45.209635737135514], batch size: 70
+2022-12-09 11:21:24,000 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:21:24,746 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.954, over 211138.00 frames. , ppl: 19.189241084043
+2022-12-09 11:23:04,400 INFO [train.py:421] (2/8) Epoch 0, batch 5200, loss[loss=2.894, over 5880.00 frames. , ppl: 18.072901539459895] tot_loss[loss=3.724, over 5110832.46 frames. , ppl: 41.439280906747726], batch size: 70
+2022-12-09 11:24:42,543 INFO [train.py:421] (2/8) Epoch 0, batch 5400, loss[loss=2.968, over 4410.00 frames. , ppl: 19.450781691871182] tot_loss[loss=3.65, over 5138499.98 frames. , ppl: 38.4627032577034], batch size: 70
+2022-12-09 11:26:20,377 INFO [train.py:421] (2/8) Epoch 0, batch 5600, loss[loss=2.739, over 910.00 frames. , ppl: 15.478351860663224] tot_loss[loss=3.582, over 5156770.44 frames. , ppl: 35.95409774630693], batch size: 70
+2022-12-09 11:28:00,840 INFO [train.py:421] (2/8) Epoch 0, batch 5800, loss[loss=2.823, over 1260.00 frames. , ppl: 16.82886077205874] tot_loss[loss=3.519, over 5172579.09 frames. , ppl: 33.76511684281297], batch size: 70
+2022-12-09 11:29:42,214 INFO [train.py:421] (2/8) Epoch 0, batch 6000, loss[loss=2.864, over 3850.00 frames. , ppl: 17.533129005937518] tot_loss[loss=3.453, over 5256033.24 frames. , ppl: 31.601625547714416], batch size: 70
+2022-12-09 11:29:42,214 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:29:42,974 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.875, over 211138.00 frames. , ppl: 17.725366404663895
+2022-12-09 11:31:26,510 INFO [train.py:421] (2/8) Epoch 0, batch 6200, loss[loss=2.86, over 4760.00 frames. , ppl: 17.46727263462596] tot_loss[loss=3.399, over 5286368.13 frames. , ppl: 29.930396769720048], batch size: 70
+2022-12-09 11:33:07,647 INFO [train.py:421] (2/8) Epoch 0, batch 6400, loss[loss=2.928, over 1260.00 frames. , ppl: 18.694037754185057] tot_loss[loss=3.352, over 5284668.11 frames. , ppl: 28.562005773089194], batch size: 70
+2022-12-09 11:34:47,512 INFO [train.py:421] (2/8) Epoch 0, batch 6600, loss[loss=2.886, over 3850.00 frames. , ppl: 17.914061041109495] tot_loss[loss=3.307, over 5304681.12 frames. , ppl: 27.292603302003908], batch size: 70
+2022-12-09 11:36:26,665 INFO [train.py:421] (2/8) Epoch 0, batch 6800, loss[loss=2.763, over 4060.00 frames. , ppl: 15.843198509438] tot_loss[loss=3.26, over 5368471.09 frames. , ppl: 26.05586953138987], batch size: 70
+2022-12-09 11:38:05,181 INFO [train.py:421] (2/8) Epoch 0, batch 7000, loss[loss=2.936, over 1120.00 frames. , ppl: 18.839735609068732] tot_loss[loss=3.222, over 5366584.77 frames. , ppl: 25.09054248947788], batch size: 70
+2022-12-09 11:38:05,181 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:38:05,970 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.816, over 211138.00 frames. , ppl: 16.709705528719027
+2022-12-09 11:39:47,610 INFO [train.py:421] (2/8) Epoch 0, batch 7200, loss[loss=2.817, over 2310.00 frames. , ppl: 16.71824309717566] tot_loss[loss=3.188, over 5363924.15 frames. , ppl: 24.230443914828054], batch size: 70
+2022-12-09 11:41:23,910 INFO [train.py:421] (2/8) Epoch 0, batch 7400, loss[loss=2.883, over 770.00 frames. , ppl: 17.87176599660814] tot_loss[loss=3.156, over 5354342.00 frames. , ppl: 23.482566416492396], batch size: 70
+2022-12-09 11:43:02,498 INFO [train.py:421] (2/8) Epoch 0, batch 7600, loss[loss=2.933, over 1680.00 frames. , ppl: 18.782095569669526] tot_loss[loss=3.123, over 5393192.02 frames. , ppl: 22.706908269560895], batch size: 70
+2022-12-09 11:44:44,319 INFO [train.py:421] (2/8) Epoch 0, batch 7800, loss[loss=2.844, over 3570.00 frames. , ppl: 17.177222459261763] tot_loss[loss=3.095, over 5389106.14 frames. , ppl: 22.080061459059433], batch size: 70
+2022-12-09 11:46:23,635 INFO [train.py:421] (2/8) Epoch 0, batch 8000, loss[loss=2.723, over 3220.00 frames. , ppl: 15.22534312002679] tot_loss[loss=3.068, over 5393676.64 frames. , ppl: 21.49955578094218], batch size: 70
+2022-12-09 11:46:23,636 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:46:24,382 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.773, over 211138.00 frames. , ppl: 16.01179093613074
+2022-12-09 11:47:58,403 INFO [train.py:421] (2/8) Epoch 0, batch 8200, loss[loss=2.771, over 1050.00 frames. , ppl: 15.975251390066099] tot_loss[loss=3.045, over 5380385.92 frames. , ppl: 21.004131349262927], batch size: 70
+2022-12-09 11:49:42,155 INFO [train.py:421] (2/8) Epoch 0, batch 8400, loss[loss=2.894, over 3570.00 frames. , ppl: 18.05744328365137] tot_loss[loss=3.022, over 5389613.46 frames. , ppl: 20.526953944675117], batch size: 70
+2022-12-09 11:51:19,107 INFO [train.py:421] (2/8) Epoch 0, batch 8600, loss[loss=2.704, over 3080.00 frames. , ppl: 14.94493809760303] tot_loss[loss=2.999, over 5395053.91 frames. , ppl: 20.074758890289832], batch size: 70
+2022-12-09 11:53:00,420 INFO [train.py:421] (2/8) Epoch 0, batch 8800, loss[loss=2.947, over 1750.00 frames. , ppl: 19.045426784148873] tot_loss[loss=2.978, over 5410723.90 frames. , ppl: 19.653436748045756], batch size: 70
+2022-12-09 11:54:40,037 INFO [train.py:421] (2/8) Epoch 0, batch 9000, loss[loss=2.799, over 2240.00 frames. , ppl: 16.422716189194226] tot_loss[loss=2.958, over 5413368.00 frames. , ppl: 19.268861570537762], batch size: 70
+2022-12-09 11:54:40,038 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 11:54:40,785 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.745, over 211138.00 frames. , ppl: 15.565240729798518
+2022-12-09 11:56:21,426 INFO [train.py:421] (2/8) Epoch 0, batch 9200, loss[loss=2.724, over 6160.00 frames. , ppl: 15.247269360198016] tot_loss[loss=2.94, over 5446618.85 frames. , ppl: 18.910471928169784], batch size: 70
+2022-12-09 11:58:04,161 INFO [train.py:421] (2/8) Epoch 0, batch 9400, loss[loss=2.946, over 1120.00 frames. , ppl: 19.03489781771371] tot_loss[loss=2.923, over 5460384.96 frames. , ppl: 18.603527888428953], batch size: 70
+2022-12-09 11:59:43,941 INFO [train.py:421] (2/8) Epoch 0, batch 9600, loss[loss=2.654, over 3990.00 frames. , ppl: 14.215119839952841] tot_loss[loss=2.908, over 5457479.21 frames. , ppl: 18.32049873701085], batch size: 70
+2022-12-09 12:01:21,922 INFO [train.py:421] (2/8) Epoch 0, batch 9800, loss[loss=2.727, over 3850.00 frames. , ppl: 15.282817372462752] tot_loss[loss=2.894, over 5444199.96 frames. , ppl: 18.06883334100013], batch size: 70
+2022-12-09 12:03:00,591 INFO [train.py:421] (2/8) Epoch 0, batch 10000, loss[loss=2.921, over 1120.00 frames. , ppl: 18.56213645792199] tot_loss[loss=2.881, over 5439306.64 frames. , ppl: 17.82380960168632], batch size: 70
+2022-12-09 12:03:00,592 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:03:01,337 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.716, over 211138.00 frames. , ppl: 15.11237530515154
+2022-12-09 12:04:40,082 INFO [train.py:421] (2/8) Epoch 0, batch 10200, loss[loss=2.639, over 10780.00 frames. , ppl: 13.99479737828827] tot_loss[loss=2.868, over 5429307.94 frames. , ppl: 17.603951367833883], batch size: 70
+2022-12-09 12:06:22,432 INFO [train.py:421] (2/8) Epoch 0, batch 10400, loss[loss=2.729, over 3360.00 frames. , ppl: 15.320346478303057] tot_loss[loss=2.856, over 5433344.59 frames. , ppl: 17.395588135014723], batch size: 70
+2022-12-09 12:08:01,855 INFO [train.py:421] (2/8) Epoch 0, batch 10600, loss[loss=2.66, over 2660.00 frames. , ppl: 14.296903720475981] tot_loss[loss=2.844, over 5450078.83 frames. , ppl: 17.178308697868637], batch size: 70
+2022-12-09 12:09:45,720 INFO [train.py:421] (2/8) Epoch 0, batch 10800, loss[loss=2.733, over 2800.00 frames. , ppl: 15.372993554979294] tot_loss[loss=2.834, over 5418427.61 frames. , ppl: 17.01696278910551], batch size: 70
+2022-12-09 12:11:29,859 INFO [train.py:421] (2/8) Epoch 0, batch 11000, loss[loss=2.678, over 4270.00 frames. , ppl: 14.561887598324724] tot_loss[loss=2.823, over 5447799.26 frames. , ppl: 16.83294448100107], batch size: 70
+2022-12-09 12:11:29,859 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:11:30,607 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.692, over 211138.00 frames. , ppl: 14.756337129748474
+2022-12-09 12:13:13,813 INFO [train.py:421] (2/8) Epoch 0, batch 11200, loss[loss=2.767, over 1190.00 frames. , ppl: 15.906857469478222] tot_loss[loss=2.813, over 5467291.90 frames. , ppl: 16.66186535129564], batch size: 70
+2022-12-09 12:14:55,284 INFO [train.py:421] (2/8) Epoch 0, batch 11400, loss[loss=2.777, over 2520.00 frames. , ppl: 16.075218130743178] tot_loss[loss=2.804, over 5442389.70 frames. , ppl: 16.5172785808231], batch size: 70
+2022-12-09 12:16:35,960 INFO [train.py:421] (2/8) Epoch 0, batch 11600, loss[loss=2.706, over 5390.00 frames. , ppl: 14.962930279030093] tot_loss[loss=2.797, over 5412369.95 frames. , ppl: 16.399820996531282], batch size: 70
+2022-12-09 12:18:18,166 INFO [train.py:421] (2/8) Epoch 0, batch 11800, loss[loss=2.789, over 1820.00 frames. , ppl: 16.25893981669551] tot_loss[loss=2.788, over 5431893.78 frames. , ppl: 16.25070393861208], batch size: 70
+2022-12-09 12:19:58,700 INFO [train.py:421] (2/8) Epoch 0, batch 12000, loss[loss=2.739, over 1120.00 frames. , ppl: 15.471312675379801] tot_loss[loss=2.779, over 5462874.67 frames. , ppl: 16.107737871462117], batch size: 70
+2022-12-09 12:19:58,701 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:19:59,461 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.67, over 211138.00 frames. , ppl: 14.443332034874045
+2022-12-09 12:21:38,234 INFO [train.py:421] (2/8) Epoch 0, batch 12200, loss[loss=2.719, over 3220.00 frames. , ppl: 15.166670480083285] tot_loss[loss=2.771, over 5473612.11 frames. , ppl: 15.982452582776572], batch size: 70
+2022-12-09 12:23:19,260 INFO [train.py:421] (2/8) Epoch 0, batch 12400, loss[loss=2.868, over 770.00 frames. , ppl: 17.60695614474787] tot_loss[loss=2.765, over 5434820.32 frames. , ppl: 15.876323944897527], batch size: 70
+2022-12-09 12:25:01,000 INFO [train.py:421] (2/8) Epoch 0, batch 12600, loss[loss=2.554, over 2170.00 frames. , ppl: 12.856760702149385] tot_loss[loss=2.758, over 5449248.73 frames. , ppl: 15.767479842195252], batch size: 70
+2022-12-09 12:26:41,330 INFO [train.py:421] (2/8) Epoch 0, batch 12800, loss[loss=2.68, over 2660.00 frames. , ppl: 14.583957092489754] tot_loss[loss=2.751, over 5450598.65 frames. , ppl: 15.664883911082038], batch size: 70
+2022-12-09 12:28:17,832 INFO [train.py:421] (2/8) Epoch 0, batch 13000, loss[loss=2.771, over 1680.00 frames. , ppl: 15.97393001714953] tot_loss[loss=2.745, over 5460871.24 frames. , ppl: 15.56008230953126], batch size: 70
+2022-12-09 12:28:17,833 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:28:18,579 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.655, over 211138.00 frames. , ppl: 14.220370751795132
+2022-12-09 12:29:56,774 INFO [train.py:421] (2/8) Epoch 0, batch 13200, loss[loss=2.791, over 980.00 frames. , ppl: 16.30171356206948] tot_loss[loss=2.741, over 5437452.97 frames. , ppl: 15.4953109517092], batch size: 70
+2022-12-09 12:31:35,006 INFO [train.py:421] (2/8) Epoch 0, batch 13400, loss[loss=2.625, over 3500.00 frames. , ppl: 13.808052729869155] tot_loss[loss=2.734, over 5468479.54 frames. , ppl: 15.391162119779654], batch size: 70
+2022-12-09 12:33:14,488 INFO [train.py:421] (2/8) Epoch 0, batch 13600, loss[loss=2.627, over 3710.00 frames. , ppl: 13.830943669363535] tot_loss[loss=2.728, over 5481191.04 frames. , ppl: 15.304163205463], batch size: 70
+2022-12-09 12:34:55,726 INFO [train.py:421] (2/8) Epoch 0, batch 13800, loss[loss=2.651, over 2170.00 frames. , ppl: 14.166069796711948] tot_loss[loss=2.722, over 5518521.61 frames. , ppl: 15.207203776010324], batch size: 70
+2022-12-09 12:36:40,249 INFO [train.py:421] (2/8) Epoch 0, batch 14000, loss[loss=2.744, over 1470.00 frames. , ppl: 15.555562849986526] tot_loss[loss=2.716, over 5554421.63 frames. , ppl: 15.120802024196136], batch size: 70
+2022-12-09 12:36:40,250 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:36:41,012 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.637, over 211138.00 frames. , ppl: 13.971456542635757
+2022-12-09 12:38:24,303 INFO [train.py:421] (2/8) Epoch 0, batch 14200, loss[loss=2.675, over 2520.00 frames. , ppl: 14.517257518041866] tot_loss[loss=2.711, over 5539713.58 frames. , ppl: 15.048195316311817], batch size: 70
+2022-12-09 12:40:02,584 INFO [train.py:421] (2/8) Epoch 0, batch 14400, loss[loss=2.545, over 4830.00 frames. , ppl: 12.737657851723464] tot_loss[loss=2.706, over 5573342.57 frames. , ppl: 14.970896984824616], batch size: 70
+2022-12-09 12:41:41,085 INFO [train.py:421] (2/8) Epoch 0, batch 14600, loss[loss=2.592, over 6090.00 frames. , ppl: 13.359615774517602] tot_loss[loss=2.703, over 5551089.57 frames. , ppl: 14.929196632828297], batch size: 70
+2022-12-09 12:43:22,516 INFO [train.py:421] (2/8) Epoch 0, batch 14800, loss[loss=2.681, over 3990.00 frames. , ppl: 14.602367484935854] tot_loss[loss=2.699, over 5554019.38 frames. , ppl: 14.862162921833226], batch size: 70
+2022-12-09 12:45:00,508 INFO [train.py:421] (2/8) Epoch 0, batch 15000, loss[loss=3.133, over 840.00 frames. , ppl: 22.94684501169565] tot_loss[loss=2.695, over 5538764.18 frames. , ppl: 14.809466574351065], batch size: 70
+2022-12-09 12:45:00,509 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:45:01,254 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.624, over 211138.00 frames. , ppl: 13.790181361567887
+2022-12-09 12:46:44,176 INFO [train.py:421] (2/8) Epoch 0, batch 15200, loss[loss=5.147, over 280.00 frames. , ppl: 171.8599133164025] tot_loss[loss=2.691, over 5551481.94 frames. , ppl: 14.743443399874653], batch size: 70
+2022-12-09 12:48:23,231 INFO [train.py:421] (2/8) Epoch 0, batch 15400, loss[loss=2.639, over 5460.00 frames. , ppl: 13.994684772834097] tot_loss[loss=2.687, over 5516252.77 frames. , ppl: 14.693152227951414], batch size: 70
+2022-12-09 12:50:02,366 INFO [train.py:421] (2/8) Epoch 0, batch 15600, loss[loss=2.628, over 6650.00 frames. , ppl: 13.848220213351834] tot_loss[loss=2.683, over 5518324.76 frames. , ppl: 14.6340844467068], batch size: 70
+2022-12-09 12:51:41,372 INFO [train.py:421] (2/8) Epoch 0, batch 15800, loss[loss=3.095, over 700.00 frames. , ppl: 22.076825972305844] tot_loss[loss=2.68, over 5496159.06 frames. , ppl: 14.589593718653639], batch size: 70
+2022-12-09 12:53:18,247 INFO [train.py:421] (2/8) Epoch 0, batch 16000, loss[loss=2.627, over 3570.00 frames. , ppl: 13.831169105449304] tot_loss[loss=2.678, over 5466595.57 frames. , ppl: 14.55189656371591], batch size: 70
+2022-12-09 12:53:18,247 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 12:53:18,996 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.614, over 211138.00 frames. , ppl: 13.652930714419472
+2022-12-09 12:54:59,738 INFO [train.py:421] (2/8) Epoch 0, batch 16200, loss[loss=2.784, over 770.00 frames. , ppl: 16.181315811184398] tot_loss[loss=2.675, over 5436225.36 frames. , ppl: 14.51199678040019], batch size: 70
+2022-12-09 12:56:38,667 INFO [train.py:421] (2/8) Epoch 0, batch 16400, loss[loss=2.615, over 10430.00 frames. , ppl: 13.661304587436813] tot_loss[loss=2.672, over 5420810.97 frames. , ppl: 14.471036295334383], batch size: 70
+2022-12-09 12:58:19,164 INFO [train.py:421] (2/8) Epoch 0, batch 16600, loss[loss=2.995, over 700.00 frames. , ppl: 19.98874776239348] tot_loss[loss=2.67, over 5411527.86 frames. , ppl: 14.434477649229185], batch size: 70
+2022-12-09 13:00:02,403 INFO [train.py:421] (2/8) Epoch 0, batch 16800, loss[loss=2.601, over 4830.00 frames. , ppl: 13.480811543484853] tot_loss[loss=2.665, over 5456002.84 frames. , ppl: 14.361273119133203], batch size: 70
+2022-12-09 13:01:41,561 INFO [train.py:421] (2/8) Epoch 0, batch 17000, loss[loss=2.672, over 1540.00 frames. , ppl: 14.469510029986893] tot_loss[loss=2.662, over 5439135.02 frames. , ppl: 14.32350766864772], batch size: 70
+2022-12-09 13:01:41,562 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:01:42,309 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.601, over 211138.00 frames. , ppl: 13.482567378311968
+2022-12-09 13:03:23,533 INFO [train.py:421] (2/8) Epoch 0, batch 17200, loss[loss=2.7, over 1330.00 frames. , ppl: 14.879027043268854] tot_loss[loss=2.66, over 5443087.56 frames. , ppl: 14.289492990855951], batch size: 70
+2022-12-09 13:05:01,411 INFO [train.py:421] (2/8) Epoch 0, batch 17400, loss[loss=2.744, over 840.00 frames. , ppl: 15.545351044780798] tot_loss[loss=2.655, over 5507579.88 frames. , ppl: 14.22910684633654], batch size: 70
+2022-12-09 13:06:45,177 INFO [train.py:421] (2/8) Epoch 0, batch 17600, loss[loss=2.69, over 2800.00 frames. , ppl: 14.7357714762852] tot_loss[loss=2.652, over 5555115.93 frames. , ppl: 14.179012690834005], batch size: 70
+2022-12-09 13:08:28,800 INFO [train.py:421] (2/8) Epoch 0, batch 17800, loss[loss=2.74, over 1680.00 frames. , ppl: 15.493695127770719] tot_loss[loss=2.65, over 5546736.66 frames. , ppl: 14.147061701930308], batch size: 70
+2022-12-09 13:10:04,650 INFO [train.py:421] (2/8) Epoch 0, batch 18000, loss[loss=2.638, over 2310.00 frames. , ppl: 13.979588345051008] tot_loss[loss=2.645, over 5579187.08 frames. , ppl: 14.09002226461178], batch size: 70
+2022-12-09 13:10:04,651 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:10:05,409 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.595, over 211138.00 frames. , ppl: 13.393325029986839
+2022-12-09 13:11:44,552 INFO [train.py:421] (2/8) Epoch 0, batch 18200, loss[loss=2.664, over 1820.00 frames. , ppl: 14.350935617961943] tot_loss[loss=2.642, over 5632754.56 frames. , ppl: 14.039473035742867], batch size: 70
+2022-12-09 13:13:26,207 INFO [train.py:421] (2/8) Epoch 0, batch 18400, loss[loss=2.576, over 5110.00 frames. , ppl: 13.144661934441698] tot_loss[loss=2.64, over 5632323.28 frames. , ppl: 14.007855065097539], batch size: 70
+2022-12-09 13:15:03,885 INFO [train.py:421] (2/8) Epoch 0, batch 18600, loss[loss=2.98, over 630.00 frames. , ppl: 19.691434131187478] tot_loss[loss=2.638, over 5584466.14 frames. , ppl: 13.99076253743009], batch size: 70
+2022-12-09 13:16:42,944 INFO [train.py:421] (2/8) Epoch 0, batch 18800, loss[loss=2.77, over 1610.00 frames. , ppl: 15.956119375637273] tot_loss[loss=2.636, over 5595671.39 frames. , ppl: 13.953979674208595], batch size: 70
+2022-12-09 13:18:25,094 INFO [train.py:421] (2/8) Epoch 0, batch 19000, loss[loss=3.396, over 560.00 frames. , ppl: 29.838616103824045] tot_loss[loss=2.633, over 5579415.76 frames. , ppl: 13.917852281777316], batch size: 70
+2022-12-09 13:18:25,094 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:18:25,840 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.582, over 211138.00 frames. , ppl: 13.223891354355008
+2022-12-09 13:20:07,272 INFO [train.py:421] (2/8) Epoch 0, batch 19200, loss[loss=2.524, over 5040.00 frames. , ppl: 12.478069603256724] tot_loss[loss=2.631, over 5588576.49 frames. , ppl: 13.888335363785137], batch size: 70
+2022-12-09 13:21:45,761 INFO [train.py:421] (2/8) Epoch 0, batch 19400, loss[loss=2.593, over 6230.00 frames. , ppl: 13.363331273536136] tot_loss[loss=2.63, over 5554682.42 frames. , ppl: 13.869082917780913], batch size: 70
+2022-12-09 13:23:23,703 INFO [train.py:421] (2/8) Epoch 0, batch 19600, loss[loss=2.563, over 4550.00 frames. , ppl: 12.970613492247859] tot_loss[loss=2.628, over 5553839.67 frames. , ppl: 13.847444288055296], batch size: 70
+2022-12-09 13:25:02,643 INFO [train.py:421] (2/8) Epoch 0, batch 19800, loss[loss=3.222, over 560.00 frames. , ppl: 25.068828748301048] tot_loss[loss=2.626, over 5528224.52 frames. , ppl: 13.82094638810833], batch size: 70
+2022-12-09 13:26:43,265 INFO [train.py:421] (2/8) Epoch 0, batch 20000, loss[loss=2.68, over 2170.00 frames. , ppl: 14.579282942882665] tot_loss[loss=2.623, over 5538398.34 frames. , ppl: 13.781979992040219], batch size: 70
+2022-12-09 13:26:43,266 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:26:44,026 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.574, over 211138.00 frames. , ppl: 13.115298232215096
+2022-12-09 13:28:25,520 INFO [train.py:421] (2/8) Epoch 0, batch 20200, loss[loss=2.734, over 3430.00 frames. , ppl: 15.397309467548142] tot_loss[loss=2.621, over 5539622.83 frames. , ppl: 13.74990201443613], batch size: 70
+2022-12-09 13:30:02,452 INFO [train.py:421] (2/8) Epoch 0, batch 20400, loss[loss=2.508, over 4340.00 frames. , ppl: 12.281197895388432] tot_loss[loss=2.62, over 5509036.60 frames. , ppl: 13.730243712603194], batch size: 70
+2022-12-09 13:31:42,412 INFO [train.py:421] (2/8) Epoch 0, batch 20600, loss[loss=2.547, over 2310.00 frames. , ppl: 12.770561042427188] tot_loss[loss=2.618, over 5522291.33 frames. , ppl: 13.70552030255501], batch size: 70
+2022-12-09 13:33:22,027 INFO [train.py:421] (2/8) Epoch 0, batch 20800, loss[loss=2.628, over 1190.00 frames. , ppl: 13.850883689950178] tot_loss[loss=2.616, over 5520582.62 frames. , ppl: 13.681813371999743], batch size: 70
+2022-12-09 13:35:01,471 INFO [train.py:421] (2/8) Epoch 0, batch 21000, loss[loss=2.572, over 2940.00 frames. , ppl: 13.092602724217098] tot_loss[loss=2.614, over 5518790.30 frames. , ppl: 13.65300748989969], batch size: 70
+2022-12-09 13:35:01,472 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:35:02,241 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.567, over 211138.00 frames. , ppl: 13.024711781587353
+2022-12-09 13:36:41,726 INFO [train.py:421] (2/8) Epoch 0, batch 21200, loss[loss=2.683, over 1050.00 frames. , ppl: 14.632079997161878] tot_loss[loss=2.612, over 5526610.48 frames. , ppl: 13.626774517748514], batch size: 70
+2022-12-09 13:38:26,894 INFO [train.py:421] (2/8) Epoch 0, batch 21400, loss[loss=2.542, over 1330.00 frames. , ppl: 12.70379972381762] tot_loss[loss=2.61, over 5538554.64 frames. , ppl: 13.598419246709966], batch size: 70
+2022-12-09 13:40:04,866 INFO [train.py:421] (2/8) Epoch 0, batch 21600, loss[loss=2.547, over 2660.00 frames. , ppl: 12.76879639283336] tot_loss[loss=2.608, over 5528150.06 frames. , ppl: 13.570164341788546], batch size: 70
+2022-12-09 13:41:41,323 INFO [train.py:421] (2/8) Epoch 0, batch 21800, loss[loss=2.544, over 4270.00 frames. , ppl: 12.733714320507342] tot_loss[loss=2.606, over 5542117.96 frames. , ppl: 13.540374615771952], batch size: 70
+2022-12-09 13:43:22,212 INFO [train.py:421] (2/8) Epoch 0, batch 22000, loss[loss=2.592, over 3500.00 frames. , ppl: 13.362332379067844] tot_loss[loss=2.603, over 5574165.32 frames. , ppl: 13.506182702645773], batch size: 70
+2022-12-09 13:43:22,213 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:43:22,962 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.558, over 211138.00 frames. , ppl: 12.907422572347382
+2022-12-09 13:45:08,009 INFO [train.py:421] (2/8) Epoch 0, batch 22200, loss[loss=2.566, over 910.00 frames. , ppl: 13.010576855363869] tot_loss[loss=2.601, over 5588548.11 frames. , ppl: 13.47844726412569], batch size: 70
+2022-12-09 13:46:46,059 INFO [train.py:421] (2/8) Epoch 0, batch 22400, loss[loss=2.658, over 2520.00 frames. , ppl: 14.273964164187122] tot_loss[loss=2.601, over 5504771.92 frames. , ppl: 13.477821101237607], batch size: 70
+2022-12-09 13:48:21,209 INFO [train.py:421] (2/8) Epoch 0, batch 22600, loss[loss=2.568, over 1610.00 frames. , ppl: 13.034796089033131] tot_loss[loss=2.598, over 5517020.65 frames. , ppl: 13.440419856732866], batch size: 70
+2022-12-09 13:50:00,273 INFO [train.py:421] (2/8) Epoch 0, batch 22800, loss[loss=2.545, over 2310.00 frames. , ppl: 12.746375709015458] tot_loss[loss=2.597, over 5533273.36 frames. , ppl: 13.42356538971654], batch size: 70
+2022-12-09 13:51:39,299 INFO [train.py:421] (2/8) Epoch 0, batch 23000, loss[loss=2.53, over 5390.00 frames. , ppl: 12.555355855343494] tot_loss[loss=2.596, over 5517422.37 frames. , ppl: 13.412141833915603], batch size: 70
+2022-12-09 13:51:39,299 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 13:51:40,059 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.55, over 211138.00 frames. , ppl: 12.811174547900729
+2022-12-09 13:53:19,944 INFO [train.py:421] (2/8) Epoch 0, batch 23200, loss[loss=2.643, over 1120.00 frames. , ppl: 14.050811342421204] tot_loss[loss=2.594, over 5537819.36 frames. , ppl: 13.381071404354584], batch size: 70
+2022-12-09 13:55:01,929 INFO [train.py:421] (2/8) Epoch 0, batch 23400, loss[loss=3.397, over 490.00 frames. , ppl: 29.876011378759237] tot_loss[loss=2.593, over 5511508.54 frames. , ppl: 13.3686477516057], batch size: 70
+2022-12-09 13:56:43,808 INFO [train.py:421] (2/8) Epoch 0, batch 23600, loss[loss=2.715, over 1120.00 frames. , ppl: 15.104265017257031] tot_loss[loss=2.589, over 5571126.24 frames. , ppl: 13.31862081785601], batch size: 70
+2022-12-09 13:58:26,276 INFO [train.py:421] (2/8) Epoch 0, batch 23800, loss[loss=2.559, over 2310.00 frames. , ppl: 12.917529887696327] tot_loss[loss=2.587, over 5604089.27 frames. , ppl: 13.285149365243882], batch size: 70
+2022-12-09 14:00:07,995 INFO [train.py:421] (2/8) Epoch 0, batch 24000, loss[loss=2.619, over 1260.00 frames. , ppl: 13.725791389979074] tot_loss[loss=2.585, over 5576567.60 frames. , ppl: 13.265972264873136], batch size: 70
+2022-12-09 14:00:07,996 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:00:08,746 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.543, over 211138.00 frames. , ppl: 12.713002157513388
+2022-12-09 14:01:49,803 INFO [train.py:421] (2/8) Epoch 0, batch 24200, loss[loss=2.571, over 2590.00 frames. , ppl: 13.076821530573861] tot_loss[loss=2.585, over 5508334.65 frames. , ppl: 13.266676262777477], batch size: 70
+2022-12-09 14:03:30,660 INFO [train.py:421] (2/8) Epoch 0, batch 24400, loss[loss=2.655, over 2450.00 frames. , ppl: 14.229664608156474] tot_loss[loss=2.584, over 5514477.45 frames. , ppl: 13.245142833128778], batch size: 70
+2022-12-09 14:05:11,533 INFO [train.py:421] (2/8) Epoch 0, batch 24600, loss[loss=2.399, over 3220.00 frames. , ppl: 11.009792006172267] tot_loss[loss=2.582, over 5491331.75 frames. , ppl: 13.225005388401293], batch size: 70
+2022-12-09 14:06:52,051 INFO [train.py:421] (2/8) Epoch 0, batch 24800, loss[loss=2.523, over 1470.00 frames. , ppl: 12.459944588770412] tot_loss[loss=2.581, over 5472535.09 frames. , ppl: 13.215031575232388], batch size: 70
+2022-12-09 14:08:35,039 INFO [train.py:421] (2/8) Epoch 0, batch 25000, loss[loss=2.455, over 5180.00 frames. , ppl: 11.640920258134479] tot_loss[loss=2.58, over 5504005.14 frames. , ppl: 13.192289201090146], batch size: 70
+2022-12-09 14:08:35,040 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:08:35,787 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.538, over 211138.00 frames. , ppl: 12.65462202047239
+2022-12-09 14:10:12,803 INFO [train.py:421] (2/8) Epoch 0, batch 25200, loss[loss=2.711, over 1330.00 frames. , ppl: 15.04797883329658] tot_loss[loss=2.58, over 5442848.60 frames. , ppl: 13.191866522882277], batch size: 70
+2022-12-09 14:11:51,083 INFO [train.py:421] (2/8) Epoch 0, batch 25400, loss[loss=2.463, over 5460.00 frames. , ppl: 11.743663306826507] tot_loss[loss=2.577, over 5467026.82 frames. , ppl: 13.158515420157448], batch size: 70
+2022-12-09 14:13:29,055 INFO [train.py:421] (2/8) Epoch 0, batch 25600, loss[loss=2.56, over 3360.00 frames. , ppl: 12.93146279763361] tot_loss[loss=2.576, over 5457611.51 frames. , ppl: 13.137895920025878], batch size: 70
+2022-12-09 14:15:06,161 INFO [train.py:421] (2/8) Epoch 0, batch 25800, loss[loss=2.438, over 4480.00 frames. , ppl: 11.445536584764687] tot_loss[loss=2.573, over 5496343.56 frames. , ppl: 13.101230125193469], batch size: 70
+2022-12-09 14:16:45,358 INFO [train.py:421] (2/8) Epoch 0, batch 26000, loss[loss=2.673, over 1750.00 frames. , ppl: 14.479507515229617] tot_loss[loss=2.571, over 5492169.68 frames. , ppl: 13.081690149383915], batch size: 70
+2022-12-09 14:16:45,359 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:16:46,124 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.534, over 211138.00 frames. , ppl: 12.606376393324691
+2022-12-09 14:18:23,649 INFO [train.py:421] (2/8) Epoch 0, batch 26200, loss[loss=2.669, over 2170.00 frames. , ppl: 14.419931638867585] tot_loss[loss=2.571, over 5437777.79 frames. , ppl: 13.07323984211717], batch size: 70
+2022-12-09 14:20:04,884 INFO [train.py:421] (2/8) Epoch 0, batch 26400, loss[loss=2.589, over 3220.00 frames. , ppl: 13.318746988320234] tot_loss[loss=2.57, over 5408307.83 frames. , ppl: 13.059335923300779], batch size: 70
+2022-12-09 14:21:42,869 INFO [train.py:421] (2/8) Epoch 0, batch 26600, loss[loss=2.714, over 980.00 frames. , ppl: 15.082642406166778] tot_loss[loss=2.57, over 5343524.55 frames. , ppl: 13.06650252893489], batch size: 70
+2022-12-09 14:23:20,634 INFO [train.py:421] (2/8) Epoch 0, batch 26800, loss[loss=2.685, over 1050.00 frames. , ppl: 14.657584500002805] tot_loss[loss=2.569, over 5344916.73 frames. , ppl: 13.046575109899806], batch size: 70
+2022-12-09 14:25:00,562 INFO [train.py:421] (2/8) Epoch 0, batch 27000, loss[loss=2.661, over 1540.00 frames. , ppl: 14.305334115051629] tot_loss[loss=2.567, over 5356609.61 frames. , ppl: 13.025017057604865], batch size: 70
+2022-12-09 14:25:00,562 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:25:01,326 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.527, over 211138.00 frames. , ppl: 12.51153246345996
+2022-12-09 14:26:38,312 INFO [train.py:421] (2/8) Epoch 0, batch 27200, loss[loss=2.417, over 4900.00 frames. , ppl: 11.212102574406616] tot_loss[loss=2.565, over 5408015.39 frames. , ppl: 12.99707094840033], batch size: 70
+2022-12-09 14:28:19,674 INFO [train.py:421] (2/8) Epoch 0, batch 27400, loss[loss=2.557, over 1890.00 frames. , ppl: 12.901056851151253] tot_loss[loss=2.563, over 5400848.66 frames. , ppl: 12.97312902395537], batch size: 70
+2022-12-09 14:30:01,584 INFO [train.py:421] (2/8) Epoch 0, batch 27600, loss[loss=2.705, over 1680.00 frames. , ppl: 14.95195592519589] tot_loss[loss=2.562, over 5395997.26 frames. , ppl: 12.959965258752424], batch size: 70
+2022-12-09 14:31:41,951 INFO [train.py:421] (2/8) Epoch 0, batch 27800, loss[loss=2.519, over 4130.00 frames. , ppl: 12.41772369658789] tot_loss[loss=2.56, over 5431701.22 frames. , ppl: 12.93474184620587], batch size: 70
+2022-12-09 14:33:19,871 INFO [train.py:421] (2/8) Epoch 0, batch 28000, loss[loss=2.466, over 3640.00 frames. , ppl: 11.78092113740556] tot_loss[loss=2.559, over 5442483.38 frames. , ppl: 12.919330454809826], batch size: 70
+2022-12-09 14:33:19,871 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:33:20,620 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.521, over 211138.00 frames. , ppl: 12.4409233226362
+2022-12-09 14:35:00,358 INFO [train.py:421] (2/8) Epoch 0, batch 28200, loss[loss=2.457, over 4760.00 frames. , ppl: 11.670282007962712] tot_loss[loss=2.557, over 5467494.68 frames. , ppl: 12.902626492836452], batch size: 70
+2022-12-09 14:36:41,343 INFO [train.py:421] (2/8) Epoch 0, batch 28400, loss[loss=2.498, over 3010.00 frames. , ppl: 12.1602933675607] tot_loss[loss=2.556, over 5492080.55 frames. , ppl: 12.885000983620417], batch size: 70
+2022-12-09 14:38:25,476 INFO [train.py:421] (2/8) Epoch 0, batch 28600, loss[loss=2.598, over 2240.00 frames. , ppl: 13.433572149538927] tot_loss[loss=2.554, over 5519739.12 frames. , ppl: 12.856261661317474], batch size: 70
+2022-12-09 14:40:02,955 INFO [train.py:421] (2/8) Epoch 0, batch 28800, loss[loss=2.612, over 1750.00 frames. , ppl: 13.627759020462934] tot_loss[loss=2.553, over 5505054.93 frames. , ppl: 12.845599643798279], batch size: 70
+2022-12-09 14:41:43,067 INFO [train.py:421] (2/8) Epoch 0, batch 29000, loss[loss=2.909, over 700.00 frames. , ppl: 18.3451723937443] tot_loss[loss=2.552, over 5509815.30 frames. , ppl: 12.830560160550068], batch size: 70
+2022-12-09 14:41:43,067 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:41:43,812 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.513, over 211138.00 frames. , ppl: 12.34438956706025
+2022-12-09 14:43:21,607 INFO [train.py:421] (2/8) Epoch 0, batch 29200, loss[loss=2.418, over 2030.00 frames. , ppl: 11.219055582061646] tot_loss[loss=2.551, over 5520813.07 frames. , ppl: 12.815706522711832], batch size: 70
+2022-12-09 14:45:00,657 INFO [train.py:421] (2/8) Epoch 0, batch 29400, loss[loss=2.471, over 5110.00 frames. , ppl: 11.838343581084912] tot_loss[loss=2.55, over 5508307.43 frames. , ppl: 12.800819119996586], batch size: 70
+2022-12-09 14:46:41,805 INFO [train.py:421] (2/8) Epoch 0, batch 29600, loss[loss=2.607, over 2660.00 frames. , ppl: 13.563560397071134] tot_loss[loss=2.547, over 5542261.98 frames. , ppl: 12.774086737330897], batch size: 70
+2022-12-09 14:48:22,705 INFO [train.py:421] (2/8) Epoch 0, batch 29800, loss[loss=2.64, over 2520.00 frames. , ppl: 14.017911524549131] tot_loss[loss=2.546, over 5571603.85 frames. , ppl: 12.749870201061228], batch size: 70
+2022-12-09 14:50:00,643 INFO [train.py:421] (2/8) Epoch 0, batch 30000, loss[loss=2.655, over 1400.00 frames. , ppl: 14.222847907751149] tot_loss[loss=2.545, over 5541355.06 frames. , ppl: 12.748514031297136], batch size: 70
+2022-12-09 14:50:00,643 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:50:01,403 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.512, over 211138.00 frames. , ppl: 12.331413290958714
+2022-12-09 14:51:44,211 INFO [train.py:421] (2/8) Epoch 0, batch 30200, loss[loss=3.332, over 490.00 frames. , ppl: 27.99555364406169] tot_loss[loss=2.545, over 5509230.03 frames. , ppl: 12.739554433594746], batch size: 70
+2022-12-09 14:53:26,934 INFO [train.py:421] (2/8) Epoch 0, batch 30400, loss[loss=2.423, over 5740.00 frames. , ppl: 11.285101854324337] tot_loss[loss=2.543, over 5582100.00 frames. , ppl: 12.712040104430313], batch size: 70
+2022-12-09 14:55:12,236 INFO [train.py:421] (2/8) Epoch 0, batch 30600, loss[loss=2.485, over 4690.00 frames. , ppl: 11.999552047259053] tot_loss[loss=2.54, over 5631200.58 frames. , ppl: 12.684498875251828], batch size: 70
+2022-12-09 14:56:52,333 INFO [train.py:421] (2/8) Epoch 0, batch 30800, loss[loss=2.488, over 3080.00 frames. , ppl: 12.03380063074638] tot_loss[loss=2.54, over 5568942.93 frames. , ppl: 12.684585419802024], batch size: 70
+2022-12-09 14:58:33,195 INFO [train.py:421] (2/8) Epoch 0, batch 31000, loss[loss=2.506, over 3570.00 frames. , ppl: 12.25762223798952] tot_loss[loss=2.539, over 5586140.11 frames. , ppl: 12.671257010604634], batch size: 70
+2022-12-09 14:58:33,196 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 14:58:33,957 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.506, over 211138.00 frames. , ppl: 12.259548686709403
+2022-12-09 15:00:11,902 INFO [train.py:421] (2/8) Epoch 0, batch 31200, loss[loss=2.416, over 3220.00 frames. , ppl: 11.201626534820308] tot_loss[loss=2.539, over 5590288.94 frames. , ppl: 12.660868441133507], batch size: 70
+2022-12-09 15:01:49,406 INFO [train.py:421] (2/8) Epoch 0, batch 31400, loss[loss=2.556, over 1400.00 frames. , ppl: 12.8831474928301] tot_loss[loss=2.537, over 5563305.24 frames. , ppl: 12.643483530794056], batch size: 70
+2022-12-09 15:03:31,993 INFO [train.py:421] (2/8) Epoch 0, batch 31600, loss[loss=2.584, over 1400.00 frames. , ppl: 13.244960199119605] tot_loss[loss=2.537, over 5547879.31 frames. , ppl: 12.642867835312009], batch size: 70
+2022-12-09 15:05:12,045 INFO [train.py:421] (2/8) Epoch 0, batch 31800, loss[loss=2.412, over 5530.00 frames. , ppl: 11.157704134145238] tot_loss[loss=2.535, over 5594061.65 frames. , ppl: 12.621571726248876], batch size: 70
+2022-12-09 15:06:52,419 INFO [train.py:421] (2/8) Epoch 0, batch 32000, loss[loss=2.65, over 980.00 frames. , ppl: 14.151796224828653] tot_loss[loss=2.535, over 5569305.67 frames. , ppl: 12.622242677799502], batch size: 70
+2022-12-09 15:06:52,420 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:06:53,178 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.502, over 211138.00 frames. , ppl: 12.211804295938897
+2022-12-09 15:08:31,772 INFO [train.py:421] (2/8) Epoch 0, batch 32200, loss[loss=2.581, over 910.00 frames. , ppl: 13.216311036867623] tot_loss[loss=2.536, over 5529071.49 frames. , ppl: 12.622802996808828], batch size: 70
+2022-12-09 15:10:13,701 INFO [train.py:421] (2/8) Epoch 0, batch 32400, loss[loss=3.203, over 490.00 frames. , ppl: 24.606788527905366] tot_loss[loss=2.534, over 5544433.95 frames. , ppl: 12.608706482576565], batch size: 70
+2022-12-09 15:11:52,318 INFO [train.py:421] (2/8) Epoch 0, batch 32600, loss[loss=3.098, over 630.00 frames. , ppl: 22.14826871526334] tot_loss[loss=2.535, over 5491168.88 frames. , ppl: 12.61169743367509], batch size: 70
+2022-12-09 15:13:32,200 INFO [train.py:421] (2/8) Epoch 0, batch 32800, loss[loss=2.479, over 2030.00 frames. , ppl: 11.923621084760216] tot_loss[loss=2.532, over 5536261.81 frames. , ppl: 12.580173848420827], batch size: 70
+2022-12-09 15:15:11,967 INFO [train.py:421] (2/8) Epoch 0, batch 33000, loss[loss=2.493, over 2940.00 frames. , ppl: 12.097801273549889] tot_loss[loss=2.531, over 5534964.51 frames. , ppl: 12.571400138517856], batch size: 70
+2022-12-09 15:15:11,967 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:15:12,723 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.498, over 211138.00 frames. , ppl: 12.152444554503326
+2022-12-09 15:16:47,646 INFO [train.py:421] (2/8) Epoch 0, batch 33200, loss[loss=2.569, over 1190.00 frames. , ppl: 13.056036460753441] tot_loss[loss=2.529, over 5558701.01 frames. , ppl: 12.543578851716164], batch size: 70
+2022-12-09 15:18:24,715 INFO [train.py:421] (2/8) Epoch 0, batch 33400, loss[loss=2.651, over 980.00 frames. , ppl: 14.170585708851] tot_loss[loss=2.53, over 5499847.12 frames. , ppl: 12.548951002305524], batch size: 70
+2022-12-09 15:20:01,825 INFO [train.py:421] (2/8) Epoch 0, batch 33600, loss[loss=2.448, over 3990.00 frames. , ppl: 11.560476993926608] tot_loss[loss=2.528, over 5498764.30 frames. , ppl: 12.534341664052493], batch size: 70
+2022-12-09 15:21:40,604 INFO [train.py:421] (2/8) Epoch 0, batch 33800, loss[loss=2.553, over 840.00 frames. , ppl: 12.848174383314072] tot_loss[loss=2.529, over 5458713.08 frames. , ppl: 12.54053415004675], batch size: 70
+2022-12-09 15:23:22,961 INFO [train.py:421] (2/8) Epoch 0, batch 34000, loss[loss=2.479, over 3080.00 frames. , ppl: 11.926255652149488] tot_loss[loss=2.527, over 5488728.96 frames. , ppl: 12.516754278998487], batch size: 70
+2022-12-09 15:23:22,962 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:23:23,718 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.494, over 211138.00 frames. , ppl: 12.106278320487071
+2022-12-09 15:25:05,072 INFO [train.py:421] (2/8) Epoch 0, batch 34200, loss[loss=2.623, over 1330.00 frames. , ppl: 13.773607670501159] tot_loss[loss=2.525, over 5533887.99 frames. , ppl: 12.489150311243613], batch size: 70
+2022-12-09 15:26:48,264 INFO [train.py:421] (2/8) Epoch 0, batch 34400, loss[loss=2.593, over 1680.00 frames. , ppl: 13.372061266169574] tot_loss[loss=2.524, over 5538517.77 frames. , ppl: 12.479622368474104], batch size: 70
+2022-12-09 15:28:27,299 INFO [train.py:421] (2/8) Epoch 0, batch 34600, loss[loss=2.616, over 840.00 frames. , ppl: 13.677510069771476] tot_loss[loss=2.522, over 5544915.33 frames. , ppl: 12.455857829918417], batch size: 70
+2022-12-09 15:30:05,481 INFO [train.py:421] (2/8) Epoch 0, batch 34800, loss[loss=2.484, over 4830.00 frames. , ppl: 11.992445853771414] tot_loss[loss=2.521, over 5552092.96 frames. , ppl: 12.443357917358076], batch size: 70
+2022-12-09 15:31:45,503 INFO [train.py:421] (2/8) Epoch 0, batch 35000, loss[loss=2.437, over 4270.00 frames. , ppl: 11.439296464946457] tot_loss[loss=2.52, over 5560437.10 frames. , ppl: 12.423798608469845], batch size: 70
+2022-12-09 15:31:45,504 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:31:46,264 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.491, over 211138.00 frames. , ppl: 12.072202954749265
+2022-12-09 15:33:26,039 INFO [train.py:421] (2/8) Epoch 0, batch 35200, loss[loss=2.468, over 6860.00 frames. , ppl: 11.799353810268318] tot_loss[loss=2.519, over 5548979.36 frames. , ppl: 12.416738725958618], batch size: 70
+2022-12-09 15:35:05,216 INFO [train.py:421] (2/8) Epoch 0, batch 35400, loss[loss=2.566, over 1820.00 frames. , ppl: 13.020085125214473] tot_loss[loss=2.52, over 5505458.19 frames. , ppl: 12.43067406280058], batch size: 70
+2022-12-09 15:36:46,623 INFO [train.py:421] (2/8) Epoch 0, batch 35600, loss[loss=2.481, over 2520.00 frames. , ppl: 11.957956490191696] tot_loss[loss=2.519, over 5486230.19 frames. , ppl: 12.422258021730723], batch size: 70
+2022-12-09 15:38:27,733 INFO [train.py:421] (2/8) Epoch 0, batch 35800, loss[loss=2.485, over 3010.00 frames. , ppl: 11.999592489363854] tot_loss[loss=2.518, over 5520464.60 frames. , ppl: 12.40493677121553], batch size: 70
+2022-12-09 15:40:08,921 INFO [train.py:421] (2/8) Epoch 0, batch 36000, loss[loss=3.188, over 560.00 frames. , ppl: 24.24112629174183] tot_loss[loss=2.517, over 5550062.90 frames. , ppl: 12.391769067953474], batch size: 70
+2022-12-09 15:40:08,922 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:40:09,682 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.487, over 211138.00 frames. , ppl: 12.026263233378076
+2022-12-09 15:41:50,311 INFO [train.py:421] (2/8) Epoch 0, batch 36200, loss[loss=2.619, over 1470.00 frames. , ppl: 13.727330211416083] tot_loss[loss=2.515, over 5559833.89 frames. , ppl: 12.365532991931042], batch size: 70
+2022-12-09 15:43:29,538 INFO [train.py:421] (2/8) Epoch 0, batch 36400, loss[loss=2.521, over 3990.00 frames. , ppl: 12.436918519044154] tot_loss[loss=2.515, over 5539152.29 frames. , ppl: 12.36131723204119], batch size: 70
+2022-12-09 15:45:09,050 INFO [train.py:421] (2/8) Epoch 0, batch 36600, loss[loss=2.548, over 3220.00 frames. , ppl: 12.783413034449836] tot_loss[loss=2.513, over 5542488.95 frames. , ppl: 12.34394012896217], batch size: 70
+2022-12-09 15:46:47,616 INFO [train.py:421] (2/8) Epoch 0, batch 36800, loss[loss=2.438, over 3010.00 frames. , ppl: 11.454601870981351] tot_loss[loss=2.513, over 5508821.76 frames. , ppl: 12.344042895283577], batch size: 70
+2022-12-09 15:48:30,558 INFO [train.py:421] (2/8) Epoch 0, batch 37000, loss[loss=2.743, over 1120.00 frames. , ppl: 15.536237206527348] tot_loss[loss=2.514, over 5425020.02 frames. , ppl: 12.3557292739585], batch size: 70
+2022-12-09 15:48:30,558 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:48:31,316 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.483, over 211138.00 frames. , ppl: 11.975344738756288
+2022-12-09 15:50:06,980 INFO [train.py:421] (2/8) Epoch 0, batch 37200, loss[loss=2.64, over 1330.00 frames. , ppl: 14.008877110599986] tot_loss[loss=2.514, over 5367021.81 frames. , ppl: 12.35030401880852], batch size: 70
+2022-12-09 15:51:48,848 INFO [train.py:421] (2/8) Epoch 0, batch 37400, loss[loss=2.49, over 1750.00 frames. , ppl: 12.058146786854046] tot_loss[loss=2.513, over 5359479.83 frames. , ppl: 12.339587115006706], batch size: 70
+2022-12-09 15:53:29,771 INFO [train.py:421] (2/8) Epoch 0, batch 37600, loss[loss=2.41, over 5740.00 frames. , ppl: 11.136337548880334] tot_loss[loss=2.512, over 5359034.32 frames. , ppl: 12.325117347274054], batch size: 70
+2022-12-09 15:55:11,150 INFO [train.py:421] (2/8) Epoch 0, batch 37800, loss[loss=3.064, over 630.00 frames. , ppl: 21.41290779338029] tot_loss[loss=2.511, over 5341088.32 frames. , ppl: 12.321059365907225], batch size: 70
+2022-12-09 15:56:53,633 INFO [train.py:421] (2/8) Epoch 0, batch 38000, loss[loss=2.462, over 4410.00 frames. , ppl: 11.723918352294534] tot_loss[loss=2.51, over 5329541.59 frames. , ppl: 12.310523204249153], batch size: 70
+2022-12-09 15:56:53,633 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 15:56:54,399 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.477, over 211138.00 frames. , ppl: 11.90581072793103
+2022-12-09 15:58:34,659 INFO [train.py:421] (2/8) Epoch 0, batch 38200, loss[loss=2.504, over 2940.00 frames. , ppl: 12.232823882015959] tot_loss[loss=2.51, over 5303494.72 frames. , ppl: 12.308181603701522], batch size: 70
+2022-12-09 16:00:16,324 INFO [train.py:421] (2/8) Epoch 0, batch 38400, loss[loss=2.454, over 4900.00 frames. , ppl: 11.633579341856347] tot_loss[loss=2.509, over 5289126.97 frames. , ppl: 12.296858199046708], batch size: 70
+2022-12-09 16:01:53,958 INFO [train.py:421] (2/8) Epoch 0, batch 38600, loss[loss=2.531, over 1960.00 frames. , ppl: 12.56447422728518] tot_loss[loss=2.507, over 5339550.89 frames. , ppl: 12.272049681063553], batch size: 70
+2022-12-09 16:03:33,396 INFO [train.py:421] (2/8) Epoch 0, batch 38800, loss[loss=2.66, over 1120.00 frames. , ppl: 14.290434064866547] tot_loss[loss=2.507, over 5350345.35 frames. , ppl: 12.264846563354805], batch size: 70
+2022-12-09 16:05:12,805 INFO [train.py:421] (2/8) Epoch 0, batch 39000, loss[loss=2.554, over 2170.00 frames. , ppl: 12.855435796432003] tot_loss[loss=2.506, over 5392634.92 frames. , ppl: 12.25247258487315], batch size: 70
+2022-12-09 16:05:12,806 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:05:13,565 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.476, over 211138.00 frames. , ppl: 11.897084242143393
+2022-12-09 16:06:53,459 INFO [train.py:421] (2/8) Epoch 0, batch 39200, loss[loss=2.543, over 3920.00 frames. , ppl: 12.721862962921026] tot_loss[loss=2.505, over 5439822.18 frames. , ppl: 12.238476407012856], batch size: 70
+2022-12-09 16:08:35,763 INFO [train.py:421] (2/8) Epoch 0, batch 39400, loss[loss=2.446, over 1400.00 frames. , ppl: 11.542765450182268] tot_loss[loss=2.504, over 5492702.86 frames. , ppl: 12.227455084682672], batch size: 70
+2022-12-09 16:10:13,121 INFO [train.py:421] (2/8) Epoch 0, batch 39600, loss[loss=2.43, over 4060.00 frames. , ppl: 11.359195738807529] tot_loss[loss=2.502, over 5520578.34 frames. , ppl: 12.208146985694505], batch size: 70
+2022-12-09 16:11:57,760 INFO [train.py:421] (2/8) Epoch 0, batch 39800, loss[loss=2.378, over 5880.00 frames. , ppl: 10.786113572343663] tot_loss[loss=2.501, over 5515181.30 frames. , ppl: 12.195539422653036], batch size: 70
+2022-12-09 16:13:37,671 INFO [train.py:421] (2/8) Epoch 0, batch 40000, loss[loss=2.464, over 2800.00 frames. , ppl: 11.750094209296995] tot_loss[loss=2.5, over 5522732.86 frames. , ppl: 12.184233951273095], batch size: 70
+2022-12-09 16:13:37,671 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:13:38,430 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.473, over 211138.00 frames. , ppl: 11.854175758684555
+2022-12-09 16:15:17,653 INFO [train.py:421] (2/8) Epoch 0, batch 40200, loss[loss=2.523, over 2940.00 frames. , ppl: 12.472064590550715] tot_loss[loss=2.5, over 5493361.87 frames. , ppl: 12.185220054333199], batch size: 70
+2022-12-09 16:16:59,417 INFO [train.py:421] (2/8) Epoch 0, batch 40400, loss[loss=2.537, over 1890.00 frames. , ppl: 12.641804444958275] tot_loss[loss=2.5, over 5502776.97 frames. , ppl: 12.179836192270049], batch size: 70
+2022-12-09 16:18:39,707 INFO [train.py:421] (2/8) Epoch 0, batch 40600, loss[loss=2.495, over 2240.00 frames. , ppl: 12.118076031310238] tot_loss[loss=2.5, over 5466524.72 frames. , ppl: 12.186863126013737], batch size: 70
+2022-12-09 16:20:18,313 INFO [train.py:421] (2/8) Epoch 0, batch 40800, loss[loss=2.602, over 1400.00 frames. , ppl: 13.49321278637831] tot_loss[loss=2.5, over 5436573.19 frames. , ppl: 12.181362723440312], batch size: 70
+2022-12-09 16:21:58,921 INFO [train.py:421] (2/8) Epoch 0, batch 41000, loss[loss=2.703, over 1190.00 frames. , ppl: 14.920400702568006] tot_loss[loss=2.498, over 5473803.88 frames. , ppl: 12.162858138818551], batch size: 70
+2022-12-09 16:21:58,922 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:21:59,668 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.47, over 211138.00 frames. , ppl: 11.826051100421541
+2022-12-09 16:23:39,477 INFO [train.py:421] (2/8) Epoch 0, batch 41200, loss[loss=2.502, over 1610.00 frames. , ppl: 12.209956243254734] tot_loss[loss=2.497, over 5509728.04 frames. , ppl: 12.150322897312012], batch size: 70
+2022-12-09 16:25:18,888 INFO [train.py:421] (2/8) Epoch 0, batch 41400, loss[loss=2.82, over 700.00 frames. , ppl: 16.769941708474732] tot_loss[loss=2.496, over 5537950.33 frames. , ppl: 12.135692546694408], batch size: 70
+2022-12-09 16:26:57,211 INFO [train.py:421] (2/8) Epoch 0, batch 41600, loss[loss=2.548, over 2730.00 frames. , ppl: 12.781387335778959] tot_loss[loss=2.494, over 5574997.05 frames. , ppl: 12.106146883756123], batch size: 70
+2022-12-09 16:28:34,642 INFO [train.py:421] (2/8) Epoch 0, batch 41800, loss[loss=2.987, over 560.00 frames. , ppl: 19.83086829783716] tot_loss[loss=2.492, over 5595967.44 frames. , ppl: 12.089434278761283], batch size: 70
+2022-12-09 16:30:16,900 INFO [train.py:421] (2/8) Epoch 0, batch 42000, loss[loss=2.594, over 1820.00 frames. , ppl: 13.378994194300008] tot_loss[loss=2.492, over 5592838.44 frames. , ppl: 12.089198416288939], batch size: 70
+2022-12-09 16:30:16,901 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:30:17,662 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.468, over 211138.00 frames. , ppl: 11.797221378196006
+2022-12-09 16:31:58,193 INFO [train.py:421] (2/8) Epoch 0, batch 42200, loss[loss=2.494, over 1890.00 frames. , ppl: 12.10994734393403] tot_loss[loss=2.492, over 5562422.89 frames. , ppl: 12.089233801545008], batch size: 70
+2022-12-09 16:33:38,765 INFO [train.py:421] (2/8) Epoch 0, batch 42400, loss[loss=2.445, over 7350.00 frames. , ppl: 11.534490611995537] tot_loss[loss=2.492, over 5574742.37 frames. , ppl: 12.082642205512863], batch size: 70
+2022-12-09 16:35:21,172 INFO [train.py:421] (2/8) Epoch 0, batch 42600, loss[loss=2.449, over 840.00 frames. , ppl: 11.573751927783198] tot_loss[loss=2.491, over 5584075.20 frames. , ppl: 12.07535186618562], batch size: 70
+2022-12-09 16:37:02,071 INFO [train.py:421] (2/8) Epoch 0, batch 42800, loss[loss=2.563, over 1330.00 frames. , ppl: 12.972428153862358] tot_loss[loss=2.491, over 5579646.29 frames. , ppl: 12.078502815125862], batch size: 70
+2022-12-09 16:38:43,436 INFO [train.py:421] (2/8) Epoch 0, batch 43000, loss[loss=2.836, over 1120.00 frames. , ppl: 17.052907440956506] tot_loss[loss=2.491, over 5571946.96 frames. , ppl: 12.07216499361305], batch size: 70
+2022-12-09 16:38:43,437 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:38:44,197 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.462, over 211138.00 frames. , ppl: 11.728416107172176
+2022-12-09 16:40:25,945 INFO [train.py:421] (2/8) Epoch 0, batch 43200, loss[loss=2.562, over 1120.00 frames. , ppl: 12.961903466118358] tot_loss[loss=2.49, over 5589688.34 frames. , ppl: 12.06513524233437], batch size: 70
+2022-12-09 16:42:04,146 INFO [train.py:421] (2/8) Epoch 0, batch 43400, loss[loss=2.628, over 1680.00 frames. , ppl: 13.847990245380199] tot_loss[loss=2.49, over 5622261.04 frames. , ppl: 12.058919289715469], batch size: 70
+2022-12-09 16:43:44,124 INFO [train.py:421] (2/8) Epoch 0, batch 43600, loss[loss=2.518, over 1120.00 frames. , ppl: 12.404782714194996] tot_loss[loss=2.49, over 5593052.80 frames. , ppl: 12.055592527342592], batch size: 70
+2022-12-09 16:45:29,464 INFO [train.py:421] (2/8) Epoch 0, batch 43800, loss[loss=2.704, over 980.00 frames. , ppl: 14.945875678614833] tot_loss[loss=2.488, over 5636595.55 frames. , ppl: 12.032994028536455], batch size: 70
+2022-12-09 16:47:12,070 INFO [train.py:421] (2/8) Epoch 0, batch 44000, loss[loss=2.379, over 7140.00 frames. , ppl: 10.790900990921497] tot_loss[loss=2.487, over 5638086.96 frames. , ppl: 12.030744253197279], batch size: 70
+2022-12-09 16:47:12,071 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:47:12,815 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.46, over 211138.00 frames. , ppl: 11.703982858703393
+2022-12-09 16:48:52,628 INFO [train.py:421] (2/8) Epoch 0, batch 44200, loss[loss=2.581, over 980.00 frames. , ppl: 13.215369925287941] tot_loss[loss=2.486, over 5632522.21 frames. , ppl: 12.015915744399104], batch size: 70
+2022-12-09 16:50:32,161 INFO [train.py:421] (2/8) Epoch 0, batch 44400, loss[loss=2.373, over 6160.00 frames. , ppl: 10.729788954679305] tot_loss[loss=2.485, over 5622191.48 frames. , ppl: 12.005022583931664], batch size: 70
+2022-12-09 16:52:13,144 INFO [train.py:421] (2/8) Epoch 0, batch 44600, loss[loss=2.493, over 2450.00 frames. , ppl: 12.098934932948616] tot_loss[loss=2.485, over 5578511.01 frames. , ppl: 12.00524555787096], batch size: 70
+2022-12-09 16:53:54,414 INFO [train.py:421] (2/8) Epoch 0, batch 44800, loss[loss=2.453, over 2800.00 frames. , ppl: 11.618994929346968] tot_loss[loss=2.485, over 5564703.33 frames. , ppl: 11.99802673500258], batch size: 70
+2022-12-09 16:55:31,011 INFO [train.py:421] (2/8) Epoch 0, batch 45000, loss[loss=2.623, over 1750.00 frames. , ppl: 13.782006151606693] tot_loss[loss=2.484, over 5572515.22 frames. , ppl: 11.992219786414323], batch size: 70
+2022-12-09 16:55:31,012 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 16:55:31,771 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.459, over 211138.00 frames. , ppl: 11.693083254495477
+2022-12-09 16:57:12,294 INFO [train.py:421] (2/8) Epoch 0, batch 45200, loss[loss=2.5, over 2800.00 frames. , ppl: 12.188099571107626] tot_loss[loss=2.484, over 5564441.70 frames. , ppl: 11.987654850859991], batch size: 70
+2022-12-09 16:58:52,875 INFO [train.py:421] (2/8) Epoch 0, batch 45400, loss[loss=2.468, over 3570.00 frames. , ppl: 11.79698848958135] tot_loss[loss=2.483, over 5552169.45 frames. , ppl: 11.975407719841948], batch size: 70
+2022-12-09 17:00:33,542 INFO [train.py:421] (2/8) Epoch 0, batch 45600, loss[loss=2.514, over 2030.00 frames. , ppl: 12.357457620793898] tot_loss[loss=2.483, over 5558579.48 frames. , ppl: 11.972611612305474], batch size: 70
+2022-12-09 17:02:15,320 INFO [train.py:421] (2/8) Epoch 0, batch 45800, loss[loss=2.443, over 4620.00 frames. , ppl: 11.503749587888883] tot_loss[loss=2.482, over 5591981.02 frames. , ppl: 11.96238293564803], batch size: 70
+2022-12-09 17:03:53,494 INFO [train.py:421] (2/8) Epoch 0, batch 46000, loss[loss=2.429, over 2730.00 frames. , ppl: 11.342753926167966] tot_loss[loss=2.481, over 5582718.46 frames. , ppl: 11.958585987382998], batch size: 70
+2022-12-09 17:03:53,494 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:03:54,253 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.455, over 211138.00 frames. , ppl: 11.65209699143015
+2022-12-09 17:05:29,358 INFO [train.py:421] (2/8) Epoch 0, batch 46200, loss[loss=2.498, over 1540.00 frames. , ppl: 12.160411905806392] tot_loss[loss=2.481, over 5548881.25 frames. , ppl: 11.95874520208113], batch size: 70
+2022-12-09 17:07:09,588 INFO [train.py:421] (2/8) Epoch 0, batch 46400, loss[loss=2.743, over 980.00 frames. , ppl: 15.53402879631778] tot_loss[loss=2.481, over 5577799.11 frames. , ppl: 11.947565272204825], batch size: 70
+2022-12-09 17:08:51,898 INFO [train.py:421] (2/8) Epoch 0, batch 46600, loss[loss=2.456, over 3780.00 frames. , ppl: 11.654967980579807] tot_loss[loss=2.481, over 5560220.66 frames. , ppl: 11.950159902518477], batch size: 70
+2022-12-09 17:10:31,042 INFO [train.py:421] (2/8) Epoch 0, batch 46800, loss[loss=2.496, over 1890.00 frames. , ppl: 12.129507508657415] tot_loss[loss=2.48, over 5587223.28 frames. , ppl: 11.936578402842413], batch size: 70
+2022-12-09 17:12:09,903 INFO [train.py:421] (2/8) Epoch 0, batch 47000, loss[loss=2.698, over 840.00 frames. , ppl: 14.847702928596672] tot_loss[loss=2.479, over 5583956.87 frames. , ppl: 11.929957211041962], batch size: 70
+2022-12-09 17:12:09,904 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:12:10,666 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.453, over 211138.00 frames. , ppl: 11.621817980016578
+2022-12-09 17:13:49,886 INFO [train.py:421] (2/8) Epoch 0, batch 47200, loss[loss=2.572, over 1610.00 frames. , ppl: 13.098359843784419] tot_loss[loss=2.478, over 5604136.51 frames. , ppl: 11.919301319985443], batch size: 70
+2022-12-09 17:15:28,650 INFO [train.py:421] (2/8) Epoch 0, batch 47400, loss[loss=2.591, over 1960.00 frames. , ppl: 13.341298936332594] tot_loss[loss=2.478, over 5586320.00 frames. , ppl: 11.914671534750804], batch size: 70
+2022-12-09 17:17:08,452 INFO [train.py:421] (2/8) Epoch 0, batch 47600, loss[loss=2.946, over 770.00 frames. , ppl: 19.0280154093149] tot_loss[loss=2.477, over 5566509.17 frames. , ppl: 11.907228077923644], batch size: 70
+2022-12-09 17:18:46,707 INFO [train.py:421] (2/8) Epoch 0, batch 47800, loss[loss=2.62, over 1050.00 frames. , ppl: 13.733165621421069] tot_loss[loss=2.476, over 5576195.30 frames. , ppl: 11.896404145307324], batch size: 70
+2022-12-09 17:20:33,545 INFO [train.py:421] (2/8) Epoch 0, batch 48000, loss[loss=2.467, over 1190.00 frames. , ppl: 11.792324740042396] tot_loss[loss=2.475, over 5598880.51 frames. , ppl: 11.880246713911495], batch size: 70
+2022-12-09 17:20:33,545 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:20:34,309 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.453, over 211138.00 frames. , ppl: 11.621976231748523
+2022-12-09 17:22:13,219 INFO [train.py:421] (2/8) Epoch 0, batch 48200, loss[loss=2.578, over 1470.00 frames. , ppl: 13.17258735107842] tot_loss[loss=2.475, over 5570294.61 frames. , ppl: 11.880129097209357], batch size: 70
+2022-12-09 17:23:50,812 INFO [train.py:421] (2/8) Epoch 0, batch 48400, loss[loss=2.618, over 910.00 frames. , ppl: 13.705063853843605] tot_loss[loss=2.475, over 5567860.13 frames. , ppl: 11.88662210348006], batch size: 70
+2022-12-09 17:25:30,262 INFO [train.py:421] (2/8) Epoch 0, batch 48600, loss[loss=2.457, over 1610.00 frames. , ppl: 11.672420501664599] tot_loss[loss=2.476, over 5525224.39 frames. , ppl: 11.890182462116085], batch size: 70
+2022-12-09 17:27:09,468 INFO [train.py:421] (2/8) Epoch 0, batch 48800, loss[loss=2.53, over 2380.00 frames. , ppl: 12.54836086787289] tot_loss[loss=2.475, over 5531319.77 frames. , ppl: 11.880711794318088], batch size: 70
+2022-12-09 17:28:45,143 INFO [train.py:421] (2/8) Epoch 0, batch 49000, loss[loss=2.769, over 1120.00 frames. , ppl: 15.939803064697387] tot_loss[loss=2.474, over 5555242.97 frames. , ppl: 11.86452155689741], batch size: 70
+2022-12-09 17:28:45,143 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:28:45,901 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.449, over 211138.00 frames. , ppl: 11.574441333093757
+2022-12-09 17:30:31,205 INFO [train.py:421] (2/8) Epoch 0, batch 49200, loss[loss=2.433, over 1680.00 frames. , ppl: 11.39162214256082] tot_loss[loss=2.472, over 5596512.79 frames. , ppl: 11.841041273048758], batch size: 70
+2022-12-09 17:32:12,942 INFO [train.py:421] (2/8) Epoch 0, batch 49400, loss[loss=3.122, over 560.00 frames. , ppl: 22.68339228223606] tot_loss[loss=2.472, over 5578066.37 frames. , ppl: 11.843401519075128], batch size: 70
+2022-12-09 17:33:55,712 INFO [train.py:421] (2/8) Epoch 0, batch 49600, loss[loss=2.529, over 1890.00 frames. , ppl: 12.53780073551526] tot_loss[loss=2.471, over 5564218.60 frames. , ppl: 11.836696162982479], batch size: 70
+2022-12-09 17:35:36,196 INFO [train.py:421] (2/8) Epoch 0, batch 49800, loss[loss=2.565, over 2310.00 frames. , ppl: 13.006174310789472] tot_loss[loss=2.471, over 5567897.73 frames. , ppl: 11.835799446662636], batch size: 70
+2022-12-09 17:37:18,308 INFO [train.py:421] (2/8) Epoch 0, batch 50000, loss[loss=2.447, over 6510.00 frames. , ppl: 11.548332781528545] tot_loss[loss=2.47, over 5576575.45 frames. , ppl: 11.821036313578801], batch size: 70
+2022-12-09 17:37:18,309 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:37:19,072 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.445, over 211138.00 frames. , ppl: 11.528131384960666
+2022-12-09 17:38:55,888 INFO [train.py:421] (2/8) Epoch 0, batch 50200, loss[loss=2.504, over 4270.00 frames. , ppl: 12.234001897550867] tot_loss[loss=2.469, over 5587577.92 frames. , ppl: 11.809506754878571], batch size: 70
+2022-12-09 17:40:41,959 INFO [train.py:421] (2/8) Epoch 0, batch 50400, loss[loss=2.459, over 6720.00 frames. , ppl: 11.696277852771628] tot_loss[loss=2.467, over 5648161.39 frames. , ppl: 11.792330240603912], batch size: 70
+2022-12-09 17:42:23,292 INFO [train.py:421] (2/8) Epoch 0, batch 50600, loss[loss=2.493, over 2100.00 frames. , ppl: 12.091720066452043] tot_loss[loss=2.465, over 5683044.06 frames. , ppl: 11.763577535853036], batch size: 70
+2022-12-09 17:44:03,398 INFO [train.py:421] (2/8) Epoch 0, batch 50800, loss[loss=2.427, over 2380.00 frames. , ppl: 11.319494222738076] tot_loss[loss=2.466, over 5618567.46 frames. , ppl: 11.77136983168327], batch size: 70
+2022-12-09 17:45:43,892 INFO [train.py:421] (2/8) Epoch 0, batch 51000, loss[loss=2.554, over 1820.00 frames. , ppl: 12.86142109781232] tot_loss[loss=2.466, over 5648967.54 frames. , ppl: 11.769903161871422], batch size: 70
+2022-12-09 17:45:43,892 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:45:44,651 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.442, over 211138.00 frames. , ppl: 11.496775701837747
+2022-12-09 17:47:23,507 INFO [train.py:421] (2/8) Epoch 0, batch 51200, loss[loss=2.444, over 1750.00 frames. , ppl: 11.517302230105827] tot_loss[loss=2.465, over 5650603.74 frames. , ppl: 11.762988729348924], batch size: 70
+2022-12-09 17:49:00,218 INFO [train.py:421] (2/8) Epoch 0, batch 51400, loss[loss=2.366, over 5880.00 frames. , ppl: 10.651738885872302] tot_loss[loss=2.466, over 5602106.56 frames. , ppl: 11.77789730154617], batch size: 70
+2022-12-09 17:50:40,892 INFO [train.py:421] (2/8) Epoch 0, batch 51600, loss[loss=2.526, over 3430.00 frames. , ppl: 12.509440943966355] tot_loss[loss=2.466, over 5571930.99 frames. , ppl: 11.776463203799977], batch size: 70
+2022-12-09 17:52:21,992 INFO [train.py:421] (2/8) Epoch 0, batch 51800, loss[loss=2.511, over 1820.00 frames. , ppl: 12.320688403450932] tot_loss[loss=2.466, over 5552857.77 frames. , ppl: 11.770873021627377], batch size: 70
+2022-12-09 17:54:03,075 INFO [train.py:421] (2/8) Epoch 0, batch 52000, loss[loss=2.458, over 4970.00 frames. , ppl: 11.680928985758307] tot_loss[loss=2.466, over 5508799.66 frames. , ppl: 11.778162992062105], batch size: 70
+2022-12-09 17:54:03,075 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 17:54:03,822 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.441, over 211138.00 frames. , ppl: 11.488961168872661
+2022-12-09 17:55:46,383 INFO [train.py:421] (2/8) Epoch 0, batch 52200, loss[loss=2.417, over 3010.00 frames. , ppl: 11.212366618756471] tot_loss[loss=2.467, over 5468675.29 frames. , ppl: 11.785259879435335], batch size: 70
+2022-12-09 17:57:27,048 INFO [train.py:421] (2/8) Epoch 0, batch 52400, loss[loss=3.596, over 420.00 frames. , ppl: 36.45233053771302] tot_loss[loss=2.466, over 5474913.18 frames. , ppl: 11.78022159945444], batch size: 70
+2022-12-09 17:59:04,842 INFO [train.py:421] (2/8) Epoch 0, batch 52600, loss[loss=2.482, over 1750.00 frames. , ppl: 11.965250969323469] tot_loss[loss=2.465, over 5529822.38 frames. , ppl: 11.761182870499693], batch size: 70
+2022-12-09 18:00:44,978 INFO [train.py:421] (2/8) Epoch 0, batch 52800, loss[loss=2.632, over 840.00 frames. , ppl: 13.903199929082174] tot_loss[loss=2.464, over 5545699.39 frames. , ppl: 11.74755404278991], batch size: 70
+2022-12-09 18:02:25,843 INFO [train.py:421] (2/8) Epoch 0, batch 53000, loss[loss=2.584, over 1680.00 frames. , ppl: 13.2487856771876] tot_loss[loss=2.463, over 5519507.92 frames. , ppl: 11.739836181267698], batch size: 70
+2022-12-09 18:02:25,844 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:02:26,589 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.439, over 211138.00 frames. , ppl: 11.45966239282214
+2022-12-09 18:04:04,857 INFO [train.py:421] (2/8) Epoch 0, batch 53200, loss[loss=2.828, over 630.00 frames. , ppl: 16.914337264018595] tot_loss[loss=2.462, over 5544609.50 frames. , ppl: 11.724901485226598], batch size: 70
+2022-12-09 18:05:43,608 INFO [train.py:421] (2/8) Epoch 0, batch 53400, loss[loss=2.507, over 1120.00 frames. , ppl: 12.264427119424294] tot_loss[loss=2.461, over 5568026.65 frames. , ppl: 11.719170860984812], batch size: 70
+2022-12-09 18:07:22,503 INFO [train.py:421] (2/8) Epoch 0, batch 53600, loss[loss=2.528, over 2240.00 frames. , ppl: 12.529687192607799] tot_loss[loss=2.461, over 5565027.87 frames. , ppl: 11.721328776333259], batch size: 70
+2022-12-09 18:09:01,077 INFO [train.py:421] (2/8) Epoch 0, batch 53800, loss[loss=2.565, over 1540.00 frames. , ppl: 12.997928613669247] tot_loss[loss=2.46, over 5571086.16 frames. , ppl: 11.710637533842432], batch size: 70
+2022-12-09 18:10:37,935 INFO [train.py:421] (2/8) Epoch 0, batch 54000, loss[loss=2.592, over 2310.00 frames. , ppl: 13.354131781071892] tot_loss[loss=2.46, over 5572584.04 frames. , ppl: 11.708121358425732], batch size: 70
+2022-12-09 18:10:37,936 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:10:38,696 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.437, over 211138.00 frames. , ppl: 11.437813533531314
+2022-12-09 18:12:17,944 INFO [train.py:421] (2/8) Epoch 0, batch 54200, loss[loss=2.369, over 6510.00 frames. , ppl: 10.68349360691527] tot_loss[loss=2.46, over 5524407.74 frames. , ppl: 11.707836848715626], batch size: 70
+2022-12-09 18:13:57,230 INFO [train.py:421] (2/8) Epoch 0, batch 54400, loss[loss=2.467, over 1190.00 frames. , ppl: 11.781288151633046] tot_loss[loss=2.46, over 5530149.05 frames. , ppl: 11.70629545667379], batch size: 70
+2022-12-09 18:15:38,992 INFO [train.py:421] (2/8) Epoch 0, batch 54600, loss[loss=3.073, over 560.00 frames. , ppl: 21.60435085924366] tot_loss[loss=2.46, over 5539413.06 frames. , ppl: 11.70974016351168], batch size: 70
+2022-12-09 18:17:17,687 INFO [train.py:421] (2/8) Epoch 0, batch 54800, loss[loss=3.465, over 490.00 frames. , ppl: 31.966167668530414] tot_loss[loss=2.46, over 5565713.20 frames. , ppl: 11.701573954354426], batch size: 70
+2022-12-09 18:18:54,917 INFO [train.py:421] (2/8) Epoch 0, batch 55000, loss[loss=2.396, over 2870.00 frames. , ppl: 10.984550870037816] tot_loss[loss=2.459, over 5558534.35 frames. , ppl: 11.688822846604625], batch size: 70
+2022-12-09 18:18:54,917 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:18:55,677 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.435, over 211138.00 frames. , ppl: 11.413029548583495
+2022-12-09 18:20:34,862 INFO [train.py:421] (2/8) Epoch 0, batch 55200, loss[loss=2.514, over 1400.00 frames. , ppl: 12.351522031223443] tot_loss[loss=2.458, over 5543566.67 frames. , ppl: 11.687165096142287], batch size: 70
+2022-12-09 18:22:16,307 INFO [train.py:421] (2/8) Epoch 0, batch 55400, loss[loss=2.459, over 3220.00 frames. , ppl: 11.694086005214942] tot_loss[loss=2.459, over 5502994.45 frames. , ppl: 11.691375451996311], batch size: 70
+2022-12-09 18:23:55,792 INFO [train.py:421] (2/8) Epoch 0, batch 55600, loss[loss=2.581, over 1750.00 frames. , ppl: 13.209059260250664] tot_loss[loss=2.458, over 5558154.56 frames. , ppl: 11.678129696505243], batch size: 70
+2022-12-09 18:25:34,331 INFO [train.py:421] (2/8) Epoch 0, batch 55800, loss[loss=3.505, over 420.00 frames. , ppl: 33.28702372882368] tot_loss[loss=2.458, over 5552202.61 frames. , ppl: 11.678627278112277], batch size: 70
+2022-12-09 18:27:11,392 INFO [train.py:421] (2/8) Epoch 0, batch 56000, loss[loss=2.488, over 2940.00 frames. , ppl: 12.033082973640234] tot_loss[loss=2.458, over 5542196.43 frames. , ppl: 11.677767391942414], batch size: 70
+2022-12-09 18:27:11,393 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:27:12,150 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.435, over 211138.00 frames. , ppl: 11.411955259320733
+2022-12-09 18:28:51,285 INFO [train.py:421] (2/8) Epoch 0, batch 56200, loss[loss=2.41, over 3010.00 frames. , ppl: 11.13164084835434] tot_loss[loss=2.458, over 5522768.14 frames. , ppl: 11.676917302911464], batch size: 70
+2022-12-09 18:30:31,855 INFO [train.py:421] (2/8) Epoch 0, batch 56400, loss[loss=2.571, over 1330.00 frames. , ppl: 13.08288689889971] tot_loss[loss=2.457, over 5517641.50 frames. , ppl: 11.672168126989952], batch size: 70
+2022-12-09 18:32:12,541 INFO [train.py:421] (2/8) Epoch 0, batch 56600, loss[loss=2.764, over 840.00 frames. , ppl: 15.857448437690174] tot_loss[loss=2.457, over 5521755.86 frames. , ppl: 11.66650609393284], batch size: 70
+2022-12-09 18:33:53,494 INFO [train.py:421] (2/8) Epoch 0, batch 56800, loss[loss=2.81, over 840.00 frames. , ppl: 16.602734422832913] tot_loss[loss=2.455, over 5549173.78 frames. , ppl: 11.65006592719798], batch size: 70
+2022-12-09 18:35:37,612 INFO [train.py:421] (2/8) Epoch 0, batch 57000, loss[loss=2.421, over 3920.00 frames. , ppl: 11.260971221108132] tot_loss[loss=2.455, over 5586559.04 frames. , ppl: 11.645157351991188], batch size: 70
+2022-12-09 18:35:37,613 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:35:38,358 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.428, over 211138.00 frames. , ppl: 11.336884940977864
+2022-12-09 18:37:20,806 INFO [train.py:421] (2/8) Epoch 0, batch 57200, loss[loss=2.7, over 770.00 frames. , ppl: 14.878255110077326] tot_loss[loss=2.455, over 5601238.92 frames. , ppl: 11.641226819610203], batch size: 70
+2022-12-09 18:39:00,405 INFO [train.py:421] (2/8) Epoch 0, batch 57400, loss[loss=2.551, over 1260.00 frames. , ppl: 12.81729950388002] tot_loss[loss=2.454, over 5608804.32 frames. , ppl: 11.634814774541612], batch size: 70
+2022-12-09 18:40:42,488 INFO [train.py:421] (2/8) Epoch 0, batch 57600, loss[loss=2.507, over 1960.00 frames. , ppl: 12.268576225193684] tot_loss[loss=2.453, over 5644766.69 frames. , ppl: 11.623473982294046], batch size: 70
+2022-12-09 18:42:20,988 INFO [train.py:421] (2/8) Epoch 0, batch 57800, loss[loss=2.762, over 630.00 frames. , ppl: 15.826666547779125] tot_loss[loss=2.453, over 5623414.94 frames. , ppl: 11.621895426296097], batch size: 70
+2022-12-09 18:44:01,539 INFO [train.py:421] (2/8) Epoch 0, batch 58000, loss[loss=2.473, over 1890.00 frames. , ppl: 11.862292187442687] tot_loss[loss=2.453, over 5591921.78 frames. , ppl: 11.626213728280781], batch size: 70
+2022-12-09 18:44:01,540 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:44:02,301 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.427, over 211138.00 frames. , ppl: 11.325093455985686
+2022-12-09 18:45:45,018 INFO [train.py:421] (2/8) Epoch 0, batch 58200, loss[loss=2.56, over 1540.00 frames. , ppl: 12.939707357120257] tot_loss[loss=2.452, over 5599206.36 frames. , ppl: 11.608819981009221], batch size: 70
+2022-12-09 18:47:28,520 INFO [train.py:421] (2/8) Epoch 0, batch 58400, loss[loss=2.813, over 840.00 frames. , ppl: 16.660005243106017] tot_loss[loss=2.451, over 5621436.20 frames. , ppl: 11.602389975182215], batch size: 70
+2022-12-09 18:49:11,522 INFO [train.py:421] (2/8) Epoch 0, batch 58600, loss[loss=2.413, over 4620.00 frames. , ppl: 11.169388086179344] tot_loss[loss=2.451, over 5647194.13 frames. , ppl: 11.595570074905897], batch size: 70
+2022-12-09 18:50:48,038 INFO [train.py:421] (2/8) Epoch 0, batch 58800, loss[loss=2.438, over 2590.00 frames. , ppl: 11.449892750360936] tot_loss[loss=2.45, over 5608117.16 frames. , ppl: 11.592257663272866], batch size: 70
+2022-12-09 18:52:28,262 INFO [train.py:421] (2/8) Epoch 0, batch 59000, loss[loss=2.515, over 2170.00 frames. , ppl: 12.367951766190451] tot_loss[loss=2.451, over 5573073.52 frames. , ppl: 11.597467522870344], batch size: 70
+2022-12-09 18:52:28,262 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 18:52:29,026 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.429, over 211138.00 frames. , ppl: 11.349824230730555
+2022-12-09 18:54:09,183 INFO [train.py:421] (2/8) Epoch 0, batch 59200, loss[loss=2.529, over 1960.00 frames. , ppl: 12.54522009583324] tot_loss[loss=2.451, over 5553771.56 frames. , ppl: 11.601297299584692], batch size: 70
+2022-12-09 18:55:49,439 INFO [train.py:421] (2/8) Epoch 0, batch 59400, loss[loss=2.413, over 2520.00 frames. , ppl: 11.172848993228785] tot_loss[loss=2.451, over 5517744.99 frames. , ppl: 11.600053828428496], batch size: 70
+2022-12-09 18:57:31,356 INFO [train.py:421] (2/8) Epoch 0, batch 59600, loss[loss=2.642, over 840.00 frames. , ppl: 14.044176769371859] tot_loss[loss=2.451, over 5510430.74 frames. , ppl: 11.604031221503597], batch size: 70
+2022-12-09 18:59:09,088 INFO [train.py:421] (2/8) Epoch 0, batch 59800, loss[loss=2.432, over 2520.00 frames. , ppl: 11.384607398073324] tot_loss[loss=2.451, over 5496724.31 frames. , ppl: 11.602851310215412], batch size: 70
+2022-12-09 19:00:50,598 INFO [train.py:421] (2/8) Epoch 0, batch 60000, loss[loss=2.665, over 980.00 frames. , ppl: 14.3726063683329] tot_loss[loss=2.452, over 5480540.40 frames. , ppl: 11.606651945566847], batch size: 70
+2022-12-09 19:00:50,598 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:00:51,357 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.426, over 211138.00 frames. , ppl: 11.316449249268606
+2022-12-09 19:02:33,547 INFO [train.py:421] (2/8) Epoch 0, batch 60200, loss[loss=2.494, over 1750.00 frames. , ppl: 12.111091090162558] tot_loss[loss=2.451, over 5477472.53 frames. , ppl: 11.600026528533574], batch size: 70
+2022-12-09 19:04:11,496 INFO [train.py:421] (2/8) Epoch 0, batch 60400, loss[loss=2.418, over 3640.00 frames. , ppl: 11.223722862802994] tot_loss[loss=2.45, over 5512702.90 frames. , ppl: 11.589174046457149], batch size: 70
+2022-12-09 19:05:47,668 INFO [train.py:421] (2/8) Epoch 0, batch 60600, loss[loss=2.491, over 3150.00 frames. , ppl: 12.073515233035332] tot_loss[loss=2.449, over 5532599.83 frames. , ppl: 11.579625814173111], batch size: 70
+2022-12-09 19:07:33,494 INFO [train.py:421] (2/8) Epoch 0, batch 60800, loss[loss=2.601, over 980.00 frames. , ppl: 13.473263772523582] tot_loss[loss=2.449, over 5527724.71 frames. , ppl: 11.575597053991515], batch size: 70
+2022-12-09 19:09:14,068 INFO [train.py:421] (2/8) Epoch 0, batch 61000, loss[loss=2.48, over 3710.00 frames. , ppl: 11.94201504526207] tot_loss[loss=2.448, over 5536165.60 frames. , ppl: 11.563225568683285], batch size: 70
+2022-12-09 19:09:14,068 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:09:14,814 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.424, over 211138.00 frames. , ppl: 11.292349414932128
+2022-12-09 19:10:56,124 INFO [train.py:421] (2/8) Epoch 0, batch 61200, loss[loss=2.415, over 4200.00 frames. , ppl: 11.19038179446612] tot_loss[loss=2.448, over 5514426.18 frames. , ppl: 11.559970170924872], batch size: 70
+2022-12-09 19:12:36,985 INFO [train.py:421] (2/8) Epoch 0, batch 61400, loss[loss=2.363, over 7840.00 frames. , ppl: 10.61820788706812] tot_loss[loss=2.449, over 5451850.43 frames. , ppl: 11.574759992353885], batch size: 70
+2022-12-09 19:14:17,498 INFO [train.py:421] (2/8) Epoch 0, batch 61600, loss[loss=3.33, over 490.00 frames. , ppl: 27.93407967173942] tot_loss[loss=2.449, over 5441883.77 frames. , ppl: 11.578858024635435], batch size: 70
+2022-12-09 19:16:00,806 INFO [train.py:421] (2/8) Epoch 0, batch 61800, loss[loss=2.808, over 840.00 frames. , ppl: 16.577733159716654] tot_loss[loss=2.449, over 5416612.60 frames. , ppl: 11.57728029455418], batch size: 70
+2022-12-09 19:17:37,907 INFO [train.py:421] (2/8) Epoch 0, batch 62000, loss[loss=2.397, over 6370.00 frames. , ppl: 10.991732944114178] tot_loss[loss=2.447, over 5443717.99 frames. , ppl: 11.556588998728161], batch size: 70
+2022-12-09 19:17:37,907 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:17:38,663 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.422, over 211138.00 frames. , ppl: 11.26730360842701
+2022-12-09 19:19:16,418 INFO [train.py:421] (2/8) Epoch 0, batch 62200, loss[loss=2.412, over 6510.00 frames. , ppl: 11.159892104852819] tot_loss[loss=2.447, over 5450684.08 frames. , ppl: 11.552143482044286], batch size: 70
+2022-12-09 19:20:55,497 INFO [train.py:421] (2/8) Epoch 0, batch 62400, loss[loss=2.634, over 840.00 frames. , ppl: 13.933211533990887] tot_loss[loss=2.447, over 5441815.80 frames. , ppl: 11.548651706480046], batch size: 70
+2022-12-09 19:22:38,366 INFO [train.py:421] (2/8) Epoch 0, batch 62600, loss[loss=2.506, over 2100.00 frames. , ppl: 12.253786128319886] tot_loss[loss=2.444, over 5507249.66 frames. , ppl: 11.520438719961762], batch size: 70
+2022-12-09 19:24:18,377 INFO [train.py:421] (2/8) Epoch 0, batch 62800, loss[loss=2.633, over 1400.00 frames. , ppl: 13.919270884586215] tot_loss[loss=2.444, over 5493416.60 frames. , ppl: 11.517550433631325], batch size: 70
+2022-12-09 19:25:57,775 INFO [train.py:421] (2/8) Epoch 0, batch 63000, loss[loss=2.798, over 840.00 frames. , ppl: 16.416089997298076] tot_loss[loss=2.443, over 5479665.90 frames. , ppl: 11.511178817979967], batch size: 70
+2022-12-09 19:25:57,775 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:25:58,533 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.419, over 211138.00 frames. , ppl: 11.238974307884806
+2022-12-09 19:27:42,798 INFO [train.py:421] (2/8) Epoch 0, batch 63200, loss[loss=3.537, over 420.00 frames. , ppl: 34.35426956036431] tot_loss[loss=2.443, over 5497969.86 frames. , ppl: 11.503095089240702], batch size: 70
+2022-12-09 19:29:19,322 INFO [train.py:421] (2/8) Epoch 0, batch 63400, loss[loss=2.424, over 2870.00 frames. , ppl: 11.29035994697309] tot_loss[loss=2.442, over 5508399.69 frames. , ppl: 11.495349922236548], batch size: 70
+2022-12-09 19:30:58,088 INFO [train.py:421] (2/8) Epoch 0, batch 63600, loss[loss=2.345, over 3010.00 frames. , ppl: 10.435351979854524] tot_loss[loss=2.442, over 5491541.22 frames. , ppl: 11.499530642486597], batch size: 70
+2022-12-09 19:32:37,331 INFO [train.py:421] (2/8) Epoch 0, batch 63800, loss[loss=2.557, over 1960.00 frames. , ppl: 12.89856675552951] tot_loss[loss=2.441, over 5514583.80 frames. , ppl: 11.489922416321596], batch size: 70
+2022-12-09 19:34:18,429 INFO [train.py:421] (2/8) Epoch 0, batch 64000, loss[loss=2.379, over 6020.00 frames. , ppl: 10.796730345475957] tot_loss[loss=2.441, over 5515309.68 frames. , ppl: 11.482313666259511], batch size: 70
+2022-12-09 19:34:18,429 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:34:19,175 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.418, over 211138.00 frames. , ppl: 11.224827241902219
+2022-12-09 19:35:57,968 INFO [train.py:421] (2/8) Epoch 0, batch 64200, loss[loss=2.546, over 2170.00 frames. , ppl: 12.75620053827317] tot_loss[loss=2.44, over 5519179.63 frames. , ppl: 11.475191266277921], batch size: 70
+2022-12-09 19:37:37,732 INFO [train.py:421] (2/8) Epoch 0, batch 64400, loss[loss=2.296, over 5110.00 frames. , ppl: 9.932012567009044] tot_loss[loss=2.44, over 5545888.87 frames. , ppl: 11.471549089424657], batch size: 70
+2022-12-09 19:39:14,488 INFO [train.py:421] (2/8) Epoch 0, batch 64600, loss[loss=2.77, over 700.00 frames. , ppl: 15.958439203438973] tot_loss[loss=2.44, over 5529883.54 frames. , ppl: 11.476324929582194], batch size: 70
+2022-12-09 19:40:56,610 INFO [train.py:421] (2/8) Epoch 0, batch 64800, loss[loss=2.534, over 1330.00 frames. , ppl: 12.608048124556289] tot_loss[loss=2.44, over 5543291.72 frames. , ppl: 11.468840560816567], batch size: 70
+2022-12-09 19:42:37,261 INFO [train.py:421] (2/8) Epoch 0, batch 65000, loss[loss=2.65, over 1050.00 frames. , ppl: 14.15735638696597] tot_loss[loss=2.44, over 5540399.63 frames. , ppl: 11.468144092715267], batch size: 70
+2022-12-09 19:42:37,261 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:42:38,022 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.417, over 211138.00 frames. , ppl: 11.211114467296362
+2022-12-09 19:44:16,899 INFO [train.py:421] (2/8) Epoch 0, batch 65200, loss[loss=2.372, over 4060.00 frames. , ppl: 10.71481380127597] tot_loss[loss=2.439, over 5572074.30 frames. , ppl: 11.457128625774304], batch size: 70
+2022-12-09 19:45:56,839 INFO [train.py:421] (2/8) Epoch 0, batch 65400, loss[loss=2.468, over 1540.00 frames. , ppl: 11.795311213380039] tot_loss[loss=2.438, over 5573203.52 frames. , ppl: 11.445248280681174], batch size: 70
+2022-12-09 19:47:35,890 INFO [train.py:421] (2/8) Epoch 0, batch 65600, loss[loss=2.375, over 2800.00 frames. , ppl: 10.75532427957971] tot_loss[loss=2.44, over 5475604.49 frames. , ppl: 11.467685643698648], batch size: 70
+2022-12-09 19:49:13,440 INFO [train.py:421] (2/8) Epoch 0, batch 65800, loss[loss=2.358, over 1400.00 frames. , ppl: 10.570592182028841] tot_loss[loss=2.438, over 5480738.91 frames. , ppl: 11.455592156834449], batch size: 70
+2022-12-09 19:50:53,677 INFO [train.py:421] (2/8) Epoch 0, batch 66000, loss[loss=2.491, over 1890.00 frames. , ppl: 12.067981185262118] tot_loss[loss=2.439, over 5441634.12 frames. , ppl: 11.456563668782277], batch size: 70
+2022-12-09 19:50:53,678 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:50:54,438 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.415, over 211138.00 frames. , ppl: 11.192550351531178
+2022-12-09 19:52:35,266 INFO [train.py:421] (2/8) Epoch 0, batch 66200, loss[loss=2.702, over 910.00 frames. , ppl: 14.913902755187657] tot_loss[loss=2.439, over 5425201.56 frames. , ppl: 11.457225230138203], batch size: 70
+2022-12-09 19:54:17,399 INFO [train.py:421] (2/8) Epoch 0, batch 66400, loss[loss=2.591, over 2310.00 frames. , ppl: 13.347660764146125] tot_loss[loss=2.438, over 5447009.73 frames. , ppl: 11.447845634524867], batch size: 70
+2022-12-09 19:56:01,926 INFO [train.py:421] (2/8) Epoch 0, batch 66600, loss[loss=2.57, over 1400.00 frames. , ppl: 13.066503451267824] tot_loss[loss=2.437, over 5447474.12 frames. , ppl: 11.43966629955807], batch size: 70
+2022-12-09 19:57:44,195 INFO [train.py:421] (2/8) Epoch 0, batch 66800, loss[loss=2.377, over 7000.00 frames. , ppl: 10.769387184159934] tot_loss[loss=2.436, over 5497583.76 frames. , ppl: 11.42164160499524], batch size: 70
+2022-12-09 19:59:25,927 INFO [train.py:421] (2/8) Epoch 0, batch 67000, loss[loss=2.383, over 5950.00 frames. , ppl: 10.840597942089632] tot_loss[loss=2.435, over 5517615.00 frames. , ppl: 11.41068448135001], batch size: 70
+2022-12-09 19:59:25,927 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 19:59:26,673 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.412, over 211138.00 frames. , ppl: 11.1568600153867
+2022-12-09 20:01:08,429 INFO [train.py:421] (2/8) Epoch 0, batch 67200, loss[loss=2.425, over 7210.00 frames. , ppl: 11.298623343147225] tot_loss[loss=2.432, over 5593554.37 frames. , ppl: 11.384489702042368], batch size: 70
+2022-12-09 20:02:47,748 INFO [train.py:421] (2/8) Epoch 0, batch 67400, loss[loss=2.375, over 8400.00 frames. , ppl: 10.746839379652704] tot_loss[loss=2.432, over 5586667.62 frames. , ppl: 11.386167520928755], batch size: 70
+2022-12-09 20:04:25,980 INFO [train.py:421] (2/8) Epoch 0, batch 67600, loss[loss=2.508, over 3990.00 frames. , ppl: 12.274800390055272] tot_loss[loss=2.432, over 5568338.94 frames. , ppl: 11.3870496970513], batch size: 70
+2022-12-09 20:06:07,197 INFO [train.py:421] (2/8) Epoch 0, batch 67800, loss[loss=2.352, over 4620.00 frames. , ppl: 10.505580813626299] tot_loss[loss=2.432, over 5592605.35 frames. , ppl: 11.382239338043721], batch size: 70
+2022-12-09 20:07:47,450 INFO [train.py:421] (2/8) Epoch 0, batch 68000, loss[loss=2.386, over 4760.00 frames. , ppl: 10.871966442771866] tot_loss[loss=2.431, over 5621624.26 frames. , ppl: 11.368321864530301], batch size: 70
+2022-12-09 20:07:47,450 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 20:07:48,196 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.413, over 211138.00 frames. , ppl: 11.167173836606553
+2022-12-09 20:09:29,914 INFO [train.py:421] (2/8) Epoch 0, batch 68200, loss[loss=2.377, over 3570.00 frames. , ppl: 10.772677115008568] tot_loss[loss=2.431, over 5589422.72 frames. , ppl: 11.371561639112437], batch size: 70
+2022-12-09 20:11:13,456 INFO [train.py:421] (2/8) Epoch 0, batch 68400, loss[loss=3.526, over 420.00 frames. , ppl: 33.9999103179307] tot_loss[loss=2.434, over 5504263.34 frames. , ppl: 11.399131843061433], batch size: 70
+2022-12-09 20:12:55,465 INFO [train.py:421] (2/8) Epoch 0, batch 68600, loss[loss=2.466, over 3220.00 frames. , ppl: 11.772725235181928] tot_loss[loss=2.434, over 5493980.20 frames. , ppl: 11.404304413867106], batch size: 70
+2022-12-09 20:14:35,016 INFO [train.py:421] (2/8) Epoch 0, batch 68800, loss[loss=2.601, over 1260.00 frames. , ppl: 13.482390578151595] tot_loss[loss=2.435, over 5453922.51 frames. , ppl: 11.410227037727214], batch size: 70
+2022-12-09 20:16:15,816 INFO [train.py:421] (2/8) Epoch 0, batch 69000, loss[loss=2.329, over 6510.00 frames. , ppl: 10.26272424212102] tot_loss[loss=2.434, over 5450975.98 frames. , ppl: 11.4075532868935], batch size: 70
+2022-12-09 20:16:15,817 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 20:16:16,582 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.409, over 211138.00 frames. , ppl: 11.12763404873676
+2022-12-09 20:17:58,587 INFO [train.py:421] (2/8) Epoch 0, batch 69200, loss[loss=2.595, over 1540.00 frames. , ppl: 13.395099085401904] tot_loss[loss=2.433, over 5487330.35 frames. , ppl: 11.397006103399452], batch size: 70
+2022-12-09 20:19:40,453 INFO [train.py:421] (2/8) Epoch 0, batch 69400, loss[loss=2.386, over 3920.00 frames. , ppl: 10.868576300873409] tot_loss[loss=2.433, over 5495414.38 frames. , ppl: 11.394114195771264], batch size: 70
+2022-12-09 20:21:21,256 INFO [train.py:421] (2/8) Epoch 0, batch 69600, loss[loss=2.439, over 1680.00 frames. , ppl: 11.46650152263681] tot_loss[loss=2.432, over 5518439.26 frames. , ppl: 11.379035069085914], batch size: 70
+2022-12-09 20:23:02,999 INFO [train.py:421] (2/8) Epoch 0, batch 69800, loss[loss=2.444, over 3780.00 frames. , ppl: 11.515922278893328] tot_loss[loss=2.431, over 5500712.95 frames. , ppl: 11.37499681701668], batch size: 70
+2022-12-09 20:24:45,759 INFO [train.py:421] (2/8) Epoch 0, batch 70000, loss[loss=2.668, over 1190.00 frames. , ppl: 14.404915583138198] tot_loss[loss=2.432, over 5484581.40 frames. , ppl: 11.382927477639045], batch size: 70
+2022-12-09 20:24:45,759 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 20:24:46,509 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.41, over 211138.00 frames. , ppl: 11.135786255375706
+2022-12-09 20:26:29,616 INFO [train.py:421] (2/8) Epoch 0, batch 70200, loss[loss=2.868, over 630.00 frames. , ppl: 17.602733442516186] tot_loss[loss=2.431, over 5510378.13 frames. , ppl: 11.369388957848178], batch size: 70
+2022-12-09 20:28:07,706 INFO [train.py:421] (2/8) Epoch 0, batch 70400, loss[loss=2.355, over 2450.00 frames. , ppl: 10.537679421428027] tot_loss[loss=2.43, over 5544968.67 frames. , ppl: 11.361887010127628], batch size: 70
+2022-12-09 20:29:47,709 INFO [train.py:421] (2/8) Epoch 0, batch 70600, loss[loss=2.529, over 1610.00 frames. , ppl: 12.541618360447739] tot_loss[loss=2.431, over 5538707.91 frames. , ppl: 11.365537725007323], batch size: 70
+2022-12-09 20:31:33,136 INFO [train.py:421] (2/8) Epoch 0, batch 70800, loss[loss=2.356, over 3990.00 frames. , ppl: 10.551493115335408] tot_loss[loss=2.429, over 5591083.38 frames. , ppl: 11.344895650308894], batch size: 70
+2022-12-09 20:33:13,790 INFO [train.py:421] (2/8) Epoch 0, batch 71000, loss[loss=2.452, over 2800.00 frames. , ppl: 11.611455889960638] tot_loss[loss=2.427, over 5650860.36 frames. , ppl: 11.330009312709114], batch size: 70
+2022-12-09 20:33:13,790 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 20:33:14,538 INFO [train.py:452] (2/8) Epoch 0, validation: loss=2.406, over 211138.00 frames. , ppl: 11.087587669142305
+2022-12-09 20:34:53,367 INFO [train.py:421] (2/8) Epoch 0, batch 71200, loss[loss=2.535, over 1400.00 frames. , ppl: 12.618281293844072] tot_loss[loss=2.426, over 5666824.65 frames. , ppl: 11.319172460307856], batch size: 70
+2022-12-09 20:36:32,752 INFO [train.py:421] (2/8) Epoch 0, batch 71400, loss[loss=2.382, over 3920.00 frames. , ppl: 10.823207721469688] tot_loss[loss=2.427, over 5643059.45 frames. , ppl: 11.325278090540936], batch size: 70
+2022-12-09 20:38:12,563 INFO [train.py:421] (2/8) Epoch 0, batch 71600, loss[loss=2.391, over 4200.00 frames. , ppl: 10.921236214734469] tot_loss[loss=2.429, over 5562283.68 frames. , ppl: 11.343217980655211], batch size: 70
+2022-12-09 20:39:51,524 INFO [train.py:421] (2/8) Epoch 0, batch 71800, loss[loss=2.376, over 5250.00 frames. , ppl: 10.76087880474916] tot_loss[loss=2.428, over 5538558.24 frames. , ppl: 11.340294882641889], batch size: 70
+2022-12-09 20:41:07,494 INFO [train.py:421] (2/8) Epoch 1, batch 0, loss[loss=2.543, over 1190.00 frames. , ppl: 12.716417390252655] tot_loss[loss=2.543, over 1190.00 frames. , ppl: 12.716417390252655], batch size: 70
+2022-12-09 20:42:47,122 INFO [train.py:421] (2/8) Epoch 1, batch 200, loss[loss=2.811, over 770.00 frames. , ppl: 16.633005452547437] tot_loss[loss=2.424, over 492022.26 frames. , ppl: 11.28569641110615], batch size: 70
+2022-12-09 20:44:26,253 INFO [train.py:421] (2/8) Epoch 1, batch 400, loss[loss=2.67, over 770.00 frames. , ppl: 14.435160821587573] tot_loss[loss=2.419, over 972143.27 frames. , ppl: 11.23074860850753], batch size: 70
+2022-12-09 20:46:09,618 INFO [train.py:421] (2/8) Epoch 1, batch 600, loss[loss=2.361, over 12180.00 frames. , ppl: 10.598005875601094] tot_loss[loss=2.417, over 1442907.02 frames. , ppl: 11.217744078093217], batch size: 70
+2022-12-09 20:47:54,141 INFO [train.py:421] (2/8) Epoch 1, batch 800, loss[loss=2.47, over 4410.00 frames. , ppl: 11.820470954103685] tot_loss[loss=2.416, over 1843689.91 frames. , ppl: 11.206177359962766], batch size: 70
+2022-12-09 20:49:35,988 INFO [train.py:421] (2/8) Epoch 1, batch 1000, loss[loss=2.533, over 1330.00 frames. , ppl: 12.585821836004461] tot_loss[loss=2.418, over 2201145.48 frames. , ppl: 11.219455250241825], batch size: 70
+2022-12-09 20:49:35,988 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 20:49:36,749 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.403, over 211138.00 frames. , ppl: 11.059317373352997
+2022-12-09 20:51:14,564 INFO [train.py:421] (2/8) Epoch 1, batch 1200, loss[loss=2.35, over 5460.00 frames. , ppl: 10.4899234870907] tot_loss[loss=2.418, over 2535381.96 frames. , ppl: 11.224172461713408], batch size: 70
+2022-12-09 20:52:52,422 INFO [train.py:421] (2/8) Epoch 1, batch 1400, loss[loss=2.452, over 3850.00 frames. , ppl: 11.617336149183412] tot_loss[loss=2.419, over 2790730.59 frames. , ppl: 11.235967975364815], batch size: 70
+2022-12-09 20:54:34,022 INFO [train.py:421] (2/8) Epoch 1, batch 1600, loss[loss=2.494, over 2520.00 frames. , ppl: 12.11391662757801] tot_loss[loss=2.419, over 3057508.59 frames. , ppl: 11.239888032520513], batch size: 70
+2022-12-09 20:56:14,665 INFO [train.py:421] (2/8) Epoch 1, batch 1800, loss[loss=2.45, over 980.00 frames. , ppl: 11.591231115914258] tot_loss[loss=2.421, over 3247955.64 frames. , ppl: 11.256247416478512], batch size: 70
+2022-12-09 20:57:57,542 INFO [train.py:421] (2/8) Epoch 1, batch 2000, loss[loss=2.46, over 3220.00 frames. , ppl: 11.709758219987448] tot_loss[loss=2.418, over 3543889.28 frames. , ppl: 11.221736193771129], batch size: 70
+2022-12-09 20:57:57,543 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 20:57:58,304 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.404, over 211138.00 frames. , ppl: 11.068564594231407
+2022-12-09 20:59:42,326 INFO [train.py:421] (2/8) Epoch 1, batch 2200, loss[loss=2.696, over 770.00 frames. , ppl: 14.81851646321762] tot_loss[loss=2.419, over 3697861.99 frames. , ppl: 11.232762472991524], batch size: 70
+2022-12-09 21:01:20,668 INFO [train.py:421] (2/8) Epoch 1, batch 2400, loss[loss=2.334, over 7140.00 frames. , ppl: 10.322446931461734] tot_loss[loss=2.419, over 3890498.29 frames. , ppl: 11.236007485237128], batch size: 70
+2022-12-09 21:03:02,870 INFO [train.py:421] (2/8) Epoch 1, batch 2600, loss[loss=2.343, over 6440.00 frames. , ppl: 10.412310067376907] tot_loss[loss=2.42, over 4011786.48 frames. , ppl: 11.246860862167637], batch size: 70
+2022-12-09 21:04:41,939 INFO [train.py:421] (2/8) Epoch 1, batch 2800, loss[loss=2.371, over 3990.00 frames. , ppl: 10.711502756852402] tot_loss[loss=2.42, over 4114229.18 frames. , ppl: 11.249462478367974], batch size: 70
+2022-12-09 21:06:24,186 INFO [train.py:421] (2/8) Epoch 1, batch 3000, loss[loss=2.397, over 5810.00 frames. , ppl: 10.991213680332597] tot_loss[loss=2.419, over 4251996.24 frames. , ppl: 11.235745009152433], batch size: 70
+2022-12-09 21:06:24,187 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:06:24,931 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.404, over 211138.00 frames. , ppl: 11.068386028582267
+2022-12-09 21:08:06,436 INFO [train.py:421] (2/8) Epoch 1, batch 3200, loss[loss=2.371, over 5040.00 frames. , ppl: 10.705565780514618] tot_loss[loss=2.417, over 4397439.55 frames. , ppl: 11.217648889919317], batch size: 70
+2022-12-09 21:09:45,268 INFO [train.py:421] (2/8) Epoch 1, batch 3400, loss[loss=2.32, over 8050.00 frames. , ppl: 10.178825065607354] tot_loss[loss=2.416, over 4544305.34 frames. , ppl: 11.19983058580011], batch size: 70
+2022-12-09 21:11:24,245 INFO [train.py:421] (2/8) Epoch 1, batch 3600, loss[loss=2.398, over 3640.00 frames. , ppl: 11.000209757914254] tot_loss[loss=2.416, over 4628620.11 frames. , ppl: 11.2025585141374], batch size: 70
+2022-12-09 21:13:06,615 INFO [train.py:421] (2/8) Epoch 1, batch 3800, loss[loss=2.444, over 2240.00 frames. , ppl: 11.514463624421667] tot_loss[loss=2.415, over 4763260.62 frames. , ppl: 11.19156106619156], batch size: 70
+2022-12-09 21:14:40,540 INFO [train.py:421] (2/8) Epoch 1, batch 4000, loss[loss=2.384, over 7490.00 frames. , ppl: 10.844990529599206] tot_loss[loss=2.416, over 4805279.14 frames. , ppl: 11.199553153616364], batch size: 70
+2022-12-09 21:14:40,541 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:14:41,285 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.405, over 211138.00 frames. , ppl: 11.079054234738054
+2022-12-09 21:16:21,687 INFO [train.py:421] (2/8) Epoch 1, batch 4200, loss[loss=2.408, over 2940.00 frames. , ppl: 11.114445315215997] tot_loss[loss=2.416, over 4875825.85 frames. , ppl: 11.197529612481485], batch size: 70
+2022-12-09 21:18:00,238 INFO [train.py:421] (2/8) Epoch 1, batch 4400, loss[loss=3.275, over 490.00 frames. , ppl: 26.448733302701736] tot_loss[loss=2.416, over 4916945.79 frames. , ppl: 11.204769412954617], batch size: 70
+2022-12-09 21:19:45,666 INFO [train.py:421] (2/8) Epoch 1, batch 4600, loss[loss=2.345, over 6370.00 frames. , ppl: 10.437474470467711] tot_loss[loss=2.415, over 5024969.80 frames. , ppl: 11.18655005457308], batch size: 70
+2022-12-09 21:21:25,275 INFO [train.py:421] (2/8) Epoch 1, batch 4800, loss[loss=2.341, over 5810.00 frames. , ppl: 10.387117527192382] tot_loss[loss=2.413, over 5107725.61 frames. , ppl: 11.172861309864654], batch size: 70
+2022-12-09 21:23:05,028 INFO [train.py:421] (2/8) Epoch 1, batch 5000, loss[loss=2.255, over 7630.00 frames. , ppl: 9.534417576475537] tot_loss[loss=2.414, over 5115195.47 frames. , ppl: 11.179016248107176], batch size: 70
+2022-12-09 21:23:05,028 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:23:05,773 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.399, over 211138.00 frames. , ppl: 11.009013517670198
+2022-12-09 21:24:42,080 INFO [train.py:421] (2/8) Epoch 1, batch 5200, loss[loss=2.639, over 1050.00 frames. , ppl: 13.998409064330476] tot_loss[loss=2.415, over 5123393.99 frames. , ppl: 11.19371127831864], batch size: 70
+2022-12-09 21:26:20,267 INFO [train.py:421] (2/8) Epoch 1, batch 5400, loss[loss=2.461, over 1680.00 frames. , ppl: 11.717765081837095] tot_loss[loss=2.416, over 5139660.68 frames. , ppl: 11.19774921880975], batch size: 70
+2022-12-09 21:28:00,127 INFO [train.py:421] (2/8) Epoch 1, batch 5600, loss[loss=2.417, over 2450.00 frames. , ppl: 11.207958241225484] tot_loss[loss=2.417, over 5151389.88 frames. , ppl: 11.207009015947639], batch size: 70
+2022-12-09 21:29:38,416 INFO [train.py:421] (2/8) Epoch 1, batch 5800, loss[loss=2.367, over 2590.00 frames. , ppl: 10.663159905493906] tot_loss[loss=2.416, over 5176415.22 frames. , ppl: 11.206426636834928], batch size: 70
+2022-12-09 21:31:19,571 INFO [train.py:421] (2/8) Epoch 1, batch 6000, loss[loss=2.476, over 1330.00 frames. , ppl: 11.889945643741441] tot_loss[loss=2.416, over 5234658.80 frames. , ppl: 11.198858270600871], batch size: 70
+2022-12-09 21:31:19,571 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:31:20,318 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.4, over 211138.00 frames. , ppl: 11.020588453227733
+2022-12-09 21:33:01,351 INFO [train.py:421] (2/8) Epoch 1, batch 6200, loss[loss=2.366, over 5180.00 frames. , ppl: 10.657036346188233] tot_loss[loss=2.416, over 5244939.01 frames. , ppl: 11.199322288097543], batch size: 70
+2022-12-09 21:34:44,217 INFO [train.py:421] (2/8) Epoch 1, batch 6400, loss[loss=2.635, over 1330.00 frames. , ppl: 13.948563488854244] tot_loss[loss=2.416, over 5242141.42 frames. , ppl: 11.205312708083584], batch size: 70
+2022-12-09 21:36:22,197 INFO [train.py:421] (2/8) Epoch 1, batch 6600, loss[loss=2.571, over 980.00 frames. , ppl: 13.085103032931151] tot_loss[loss=2.416, over 5261416.81 frames. , ppl: 11.205565128478213], batch size: 70
+2022-12-09 21:38:05,940 INFO [train.py:421] (2/8) Epoch 1, batch 6800, loss[loss=2.361, over 3990.00 frames. , ppl: 10.605239746645376] tot_loss[loss=2.415, over 5320478.92 frames. , ppl: 11.192404860672367], batch size: 70
+2022-12-09 21:39:45,054 INFO [train.py:421] (2/8) Epoch 1, batch 7000, loss[loss=2.601, over 1470.00 frames. , ppl: 13.474999195327204] tot_loss[loss=2.415, over 5337821.70 frames. , ppl: 11.190356814282053], batch size: 70
+2022-12-09 21:39:45,054 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:39:45,806 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.398, over 211138.00 frames. , ppl: 11.000016301250483
+2022-12-09 21:41:28,147 INFO [train.py:421] (2/8) Epoch 1, batch 7200, loss[loss=2.736, over 700.00 frames. , ppl: 15.419329135326494] tot_loss[loss=2.415, over 5348987.73 frames. , ppl: 11.184500986914983], batch size: 70
+2022-12-09 21:43:10,157 INFO [train.py:421] (2/8) Epoch 1, batch 7400, loss[loss=2.926, over 630.00 frames. , ppl: 18.645042972292124] tot_loss[loss=2.414, over 5377478.74 frames. , ppl: 11.181693081100923], batch size: 70
+2022-12-09 21:44:48,881 INFO [train.py:421] (2/8) Epoch 1, batch 7600, loss[loss=2.339, over 3290.00 frames. , ppl: 10.367304268488557] tot_loss[loss=2.415, over 5350180.26 frames. , ppl: 11.185730620083438], batch size: 70
+2022-12-09 21:46:31,987 INFO [train.py:421] (2/8) Epoch 1, batch 7800, loss[loss=2.641, over 1050.00 frames. , ppl: 14.030995317764853] tot_loss[loss=2.414, over 5363975.35 frames. , ppl: 11.18151925464811], batch size: 70
+2022-12-09 21:48:11,092 INFO [train.py:421] (2/8) Epoch 1, batch 8000, loss[loss=2.413, over 3710.00 frames. , ppl: 11.1673411042867] tot_loss[loss=2.414, over 5374203.39 frames. , ppl: 11.182830440335726], batch size: 70
+2022-12-09 21:48:11,093 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:48:11,857 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.397, over 211138.00 frames. , ppl: 10.99293966703717
+2022-12-09 21:49:52,602 INFO [train.py:421] (2/8) Epoch 1, batch 8200, loss[loss=2.404, over 5040.00 frames. , ppl: 11.06934365994679] tot_loss[loss=2.415, over 5343109.97 frames. , ppl: 11.193088664626218], batch size: 70
+2022-12-09 21:51:31,078 INFO [train.py:421] (2/8) Epoch 1, batch 8400, loss[loss=2.543, over 1120.00 frames. , ppl: 12.714273542222767] tot_loss[loss=2.414, over 5388133.15 frames. , ppl: 11.183435374942277], batch size: 70
+2022-12-09 21:53:08,570 INFO [train.py:421] (2/8) Epoch 1, batch 8600, loss[loss=2.3, over 7350.00 frames. , ppl: 9.973318444758227] tot_loss[loss=2.415, over 5372581.09 frames. , ppl: 11.190648308581972], batch size: 70
+2022-12-09 21:54:48,827 INFO [train.py:421] (2/8) Epoch 1, batch 8800, loss[loss=2.46, over 1470.00 frames. , ppl: 11.70849164433871] tot_loss[loss=2.414, over 5410647.16 frames. , ppl: 11.175889094086754], batch size: 70
+2022-12-09 21:56:32,515 INFO [train.py:421] (2/8) Epoch 1, batch 9000, loss[loss=2.569, over 2520.00 frames. , ppl: 13.057449338760742] tot_loss[loss=2.412, over 5478479.66 frames. , ppl: 11.153006024259447], batch size: 70
+2022-12-09 21:56:32,516 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 21:56:33,273 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.396, over 211138.00 frames. , ppl: 10.981821240677286
+2022-12-09 21:58:14,759 INFO [train.py:421] (2/8) Epoch 1, batch 9200, loss[loss=2.44, over 5250.00 frames. , ppl: 11.472532832963159] tot_loss[loss=2.411, over 5487343.88 frames. , ppl: 11.147466717546262], batch size: 70
+2022-12-09 21:59:56,480 INFO [train.py:421] (2/8) Epoch 1, batch 9400, loss[loss=2.372, over 4410.00 frames. , ppl: 10.71376699297206] tot_loss[loss=2.411, over 5536246.42 frames. , ppl: 11.145433230018076], batch size: 70
+2022-12-09 22:01:34,865 INFO [train.py:421] (2/8) Epoch 1, batch 9600, loss[loss=2.387, over 3850.00 frames. , ppl: 10.877615806482437] tot_loss[loss=2.412, over 5498642.38 frames. , ppl: 11.15623204241553], batch size: 70
+2022-12-09 22:03:13,432 INFO [train.py:421] (2/8) Epoch 1, batch 9800, loss[loss=2.334, over 4270.00 frames. , ppl: 10.324203081752453] tot_loss[loss=2.412, over 5483612.65 frames. , ppl: 11.157813464238805], batch size: 70
+2022-12-09 22:04:51,393 INFO [train.py:421] (2/8) Epoch 1, batch 10000, loss[loss=2.608, over 1750.00 frames. , ppl: 13.572114713541769] tot_loss[loss=2.413, over 5500368.91 frames. , ppl: 11.163437832399689], batch size: 70
+2022-12-09 22:04:51,393 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:04:52,141 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.395, over 211138.00 frames. , ppl: 10.972836503018678
+2022-12-09 22:06:33,532 INFO [train.py:421] (2/8) Epoch 1, batch 10200, loss[loss=2.405, over 1120.00 frames. , ppl: 11.07721903147965] tot_loss[loss=2.413, over 5470071.27 frames. , ppl: 11.166693614508105], batch size: 70
+2022-12-09 22:08:13,457 INFO [train.py:421] (2/8) Epoch 1, batch 10400, loss[loss=2.528, over 1190.00 frames. , ppl: 12.5223527070692] tot_loss[loss=2.413, over 5469302.21 frames. , ppl: 11.171086281007064], batch size: 70
+2022-12-09 22:09:54,328 INFO [train.py:421] (2/8) Epoch 1, batch 10600, loss[loss=2.429, over 1820.00 frames. , ppl: 11.35006988411155] tot_loss[loss=2.413, over 5480015.53 frames. , ppl: 11.164023386854318], batch size: 70
+2022-12-09 22:11:32,870 INFO [train.py:421] (2/8) Epoch 1, batch 10800, loss[loss=2.448, over 1610.00 frames. , ppl: 11.566272002517513] tot_loss[loss=2.413, over 5460976.95 frames. , ppl: 11.171395628416061], batch size: 70
+2022-12-09 22:13:12,635 INFO [train.py:421] (2/8) Epoch 1, batch 11000, loss[loss=3.09, over 560.00 frames. , ppl: 21.967198100453594] tot_loss[loss=2.412, over 5483889.27 frames. , ppl: 11.160270964503248], batch size: 70
+2022-12-09 22:13:12,635 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:13:13,393 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.394, over 211138.00 frames. , ppl: 10.959451037049323
+2022-12-09 22:14:51,704 INFO [train.py:421] (2/8) Epoch 1, batch 11200, loss[loss=2.389, over 2590.00 frames. , ppl: 10.904352718609134] tot_loss[loss=2.413, over 5496774.49 frames. , ppl: 11.163376783171218], batch size: 70
+2022-12-09 22:16:32,958 INFO [train.py:421] (2/8) Epoch 1, batch 11400, loss[loss=3.108, over 560.00 frames. , ppl: 22.36958936291382] tot_loss[loss=2.412, over 5466328.72 frames. , ppl: 11.160189922246875], batch size: 70
+2022-12-09 22:18:09,021 INFO [train.py:421] (2/8) Epoch 1, batch 11600, loss[loss=3.823, over 420.00 frames. , ppl: 45.76382284881111] tot_loss[loss=2.41, over 5500035.66 frames. , ppl: 11.139009143617997], batch size: 70
+2022-12-09 22:19:53,253 INFO [train.py:421] (2/8) Epoch 1, batch 11800, loss[loss=2.368, over 4550.00 frames. , ppl: 10.67633372095037] tot_loss[loss=2.412, over 5437733.67 frames. , ppl: 11.154737752784294], batch size: 70
+2022-12-09 22:21:30,851 INFO [train.py:421] (2/8) Epoch 1, batch 12000, loss[loss=2.418, over 2030.00 frames. , ppl: 11.224753665534132] tot_loss[loss=2.412, over 5446901.10 frames. , ppl: 11.151186129863714], batch size: 70
+2022-12-09 22:21:30,852 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:21:31,611 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.393, over 211138.00 frames. , ppl: 10.943533776107909
+2022-12-09 22:23:19,195 INFO [train.py:421] (2/8) Epoch 1, batch 12200, loss[loss=2.439, over 1890.00 frames. , ppl: 11.456866024124375] tot_loss[loss=2.412, over 5427239.80 frames. , ppl: 11.153400032124493], batch size: 70
+2022-12-09 22:24:55,362 INFO [train.py:421] (2/8) Epoch 1, batch 12400, loss[loss=2.479, over 1890.00 frames. , ppl: 11.928722365039429] tot_loss[loss=2.411, over 5431642.48 frames. , ppl: 11.148640945011348], batch size: 70
+2022-12-09 22:26:36,489 INFO [train.py:421] (2/8) Epoch 1, batch 12600, loss[loss=2.495, over 2310.00 frames. , ppl: 12.126162933810518] tot_loss[loss=2.41, over 5459026.45 frames. , ppl: 11.137487546062593], batch size: 70
+2022-12-09 22:28:14,892 INFO [train.py:421] (2/8) Epoch 1, batch 12800, loss[loss=2.469, over 3570.00 frames. , ppl: 11.808672893244626] tot_loss[loss=2.41, over 5455509.87 frames. , ppl: 11.132793558403144], batch size: 70
+2022-12-09 22:29:53,721 INFO [train.py:421] (2/8) Epoch 1, batch 13000, loss[loss=2.417, over 3360.00 frames. , ppl: 11.207194425034057] tot_loss[loss=2.412, over 5393216.79 frames. , ppl: 11.15073906366618], batch size: 70
+2022-12-09 22:29:53,722 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:29:54,481 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.391, over 211138.00 frames. , ppl: 10.927798136667734
+2022-12-09 22:31:36,569 INFO [train.py:421] (2/8) Epoch 1, batch 13200, loss[loss=2.385, over 3220.00 frames. , ppl: 10.86023252214268] tot_loss[loss=2.409, over 5455624.05 frames. , ppl: 11.127817041934144], batch size: 70
+2022-12-09 22:33:15,742 INFO [train.py:421] (2/8) Epoch 1, batch 13400, loss[loss=2.31, over 8260.00 frames. , ppl: 10.074576637728955] tot_loss[loss=2.409, over 5457095.04 frames. , ppl: 11.127892279095867], batch size: 70
+2022-12-09 22:34:54,052 INFO [train.py:421] (2/8) Epoch 1, batch 13600, loss[loss=2.68, over 840.00 frames. , ppl: 14.585668972138041] tot_loss[loss=2.409, over 5474964.90 frames. , ppl: 11.1220065051188], batch size: 70
+2022-12-09 22:36:30,646 INFO [train.py:421] (2/8) Epoch 1, batch 13800, loss[loss=2.294, over 2800.00 frames. , ppl: 9.917036635316546] tot_loss[loss=2.41, over 5450536.33 frames. , ppl: 11.129289949468887], batch size: 70
+2022-12-09 22:38:12,924 INFO [train.py:421] (2/8) Epoch 1, batch 14000, loss[loss=3.246, over 490.00 frames. , ppl: 25.683873666184585] tot_loss[loss=2.41, over 5429996.96 frames. , ppl: 11.13436961785497], batch size: 70
+2022-12-09 22:38:12,925 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:38:13,682 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.39, over 211138.00 frames. , ppl: 10.914509612851669
+2022-12-09 22:39:52,639 INFO [train.py:421] (2/8) Epoch 1, batch 14200, loss[loss=2.348, over 3990.00 frames. , ppl: 10.464505619025978] tot_loss[loss=2.409, over 5446967.43 frames. , ppl: 11.122783380168553], batch size: 70
+2022-12-09 22:41:35,426 INFO [train.py:421] (2/8) Epoch 1, batch 14400, loss[loss=2.761, over 840.00 frames. , ppl: 15.812630754744063] tot_loss[loss=2.408, over 5480807.55 frames. , ppl: 11.11493016750142], batch size: 70
+2022-12-09 22:43:15,976 INFO [train.py:421] (2/8) Epoch 1, batch 14600, loss[loss=2.352, over 7000.00 frames. , ppl: 10.507130579261005] tot_loss[loss=2.408, over 5478937.91 frames. , ppl: 11.11589873997829], batch size: 70
+2022-12-09 22:44:53,638 INFO [train.py:421] (2/8) Epoch 1, batch 14800, loss[loss=2.436, over 1260.00 frames. , ppl: 11.431240706331806] tot_loss[loss=2.408, over 5486018.92 frames. , ppl: 11.114697796092338], batch size: 70
+2022-12-09 22:46:32,344 INFO [train.py:421] (2/8) Epoch 1, batch 15000, loss[loss=2.445, over 1890.00 frames. , ppl: 11.533801842073448] tot_loss[loss=2.408, over 5447106.76 frames. , ppl: 11.11199177445997], batch size: 70
+2022-12-09 22:46:32,345 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:46:33,092 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.389, over 211138.00 frames. , ppl: 10.905875331584172
+2022-12-09 22:48:15,322 INFO [train.py:421] (2/8) Epoch 1, batch 15200, loss[loss=2.333, over 2100.00 frames. , ppl: 10.303817070184465] tot_loss[loss=2.409, over 5399971.37 frames. , ppl: 11.118426652647026], batch size: 70
+2022-12-09 22:49:57,178 INFO [train.py:421] (2/8) Epoch 1, batch 15400, loss[loss=2.443, over 2590.00 frames. , ppl: 11.511945127253494] tot_loss[loss=2.407, over 5430113.82 frames. , ppl: 11.105239089470034], batch size: 70
+2022-12-09 22:51:38,845 INFO [train.py:421] (2/8) Epoch 1, batch 15600, loss[loss=2.368, over 4130.00 frames. , ppl: 10.675008761274213] tot_loss[loss=2.407, over 5455309.44 frames. , ppl: 11.103469468608898], batch size: 70
+2022-12-09 22:53:18,733 INFO [train.py:421] (2/8) Epoch 1, batch 15800, loss[loss=2.425, over 1680.00 frames. , ppl: 11.307825023866311] tot_loss[loss=2.407, over 5473285.58 frames. , ppl: 11.096872731829707], batch size: 70
+2022-12-09 22:55:01,078 INFO [train.py:421] (2/8) Epoch 1, batch 16000, loss[loss=2.3, over 6930.00 frames. , ppl: 9.978475906751479] tot_loss[loss=2.406, over 5529961.65 frames. , ppl: 11.08756546480651], batch size: 70
+2022-12-09 22:55:01,079 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 22:55:01,823 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.387, over 211138.00 frames. , ppl: 10.886231072245774
+2022-12-09 22:56:40,411 INFO [train.py:421] (2/8) Epoch 1, batch 16200, loss[loss=2.328, over 3150.00 frames. , ppl: 10.252308965967295] tot_loss[loss=2.406, over 5518305.88 frames. , ppl: 11.093732226457975], batch size: 70
+2022-12-09 22:58:20,171 INFO [train.py:421] (2/8) Epoch 1, batch 16400, loss[loss=2.325, over 1960.00 frames. , ppl: 10.224542787091437] tot_loss[loss=2.406, over 5511939.63 frames. , ppl: 11.089530768721295], batch size: 70
+2022-12-09 23:00:00,604 INFO [train.py:421] (2/8) Epoch 1, batch 16600, loss[loss=2.51, over 1260.00 frames. , ppl: 12.304928629970822] tot_loss[loss=2.406, over 5486979.54 frames. , ppl: 11.08857796852457], batch size: 70
+2022-12-09 23:01:39,179 INFO [train.py:421] (2/8) Epoch 1, batch 16800, loss[loss=2.565, over 1470.00 frames. , ppl: 13.004923873894672] tot_loss[loss=2.406, over 5493130.78 frames. , ppl: 11.088385383765582], batch size: 70
+2022-12-09 23:03:20,352 INFO [train.py:421] (2/8) Epoch 1, batch 17000, loss[loss=2.69, over 1050.00 frames. , ppl: 14.72459060907327] tot_loss[loss=2.406, over 5456853.20 frames. , ppl: 11.093342686039978], batch size: 70
+2022-12-09 23:03:20,352 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:03:21,099 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.388, over 211138.00 frames. , ppl: 10.88762972140702
+2022-12-09 23:04:58,074 INFO [train.py:421] (2/8) Epoch 1, batch 17200, loss[loss=2.57, over 980.00 frames. , ppl: 13.069032310198903] tot_loss[loss=2.407, over 5453845.62 frames. , ppl: 11.098792960343816], batch size: 70
+2022-12-09 23:06:36,567 INFO [train.py:421] (2/8) Epoch 1, batch 17400, loss[loss=2.285, over 7140.00 frames. , ppl: 9.825807712704277] tot_loss[loss=2.408, over 5410617.47 frames. , ppl: 11.115243443480448], batch size: 70
+2022-12-09 23:08:16,406 INFO [train.py:421] (2/8) Epoch 1, batch 17600, loss[loss=2.57, over 1610.00 frames. , ppl: 13.060208256779468] tot_loss[loss=2.408, over 5439327.35 frames. , ppl: 11.107890602282255], batch size: 70
+2022-12-09 23:09:57,151 INFO [train.py:421] (2/8) Epoch 1, batch 17800, loss[loss=2.718, over 770.00 frames. , ppl: 15.145402562113036] tot_loss[loss=2.408, over 5417237.79 frames. , ppl: 11.108218320800345], batch size: 70
+2022-12-09 23:11:36,545 INFO [train.py:421] (2/8) Epoch 1, batch 18000, loss[loss=2.604, over 980.00 frames. , ppl: 13.520249254887563] tot_loss[loss=2.408, over 5418544.51 frames. , ppl: 11.10935065081393], batch size: 70
+2022-12-09 23:11:36,545 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:11:37,293 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.386, over 211138.00 frames. , ppl: 10.869570686577362
+2022-12-09 23:13:24,071 INFO [train.py:421] (2/8) Epoch 1, batch 18200, loss[loss=2.646, over 770.00 frames. , ppl: 14.101226717569244] tot_loss[loss=2.406, over 5469364.34 frames. , ppl: 11.087347783793751], batch size: 70
+2022-12-09 23:15:06,572 INFO [train.py:421] (2/8) Epoch 1, batch 18400, loss[loss=2.55, over 2100.00 frames. , ppl: 12.813379633374286] tot_loss[loss=2.405, over 5510719.91 frames. , ppl: 11.076880412003836], batch size: 70
+2022-12-09 23:16:40,257 INFO [train.py:421] (2/8) Epoch 1, batch 18600, loss[loss=2.315, over 5670.00 frames. , ppl: 10.123361946058202] tot_loss[loss=2.405, over 5477344.08 frames. , ppl: 11.083208918943065], batch size: 70
+2022-12-09 23:18:19,348 INFO [train.py:421] (2/8) Epoch 1, batch 18800, loss[loss=2.52, over 2380.00 frames. , ppl: 12.434738737325908] tot_loss[loss=2.405, over 5466133.04 frames. , ppl: 11.0751086383386], batch size: 70
+2022-12-09 23:19:58,913 INFO [train.py:421] (2/8) Epoch 1, batch 19000, loss[loss=2.604, over 1260.00 frames. , ppl: 13.517711736771522] tot_loss[loss=2.406, over 5414272.67 frames. , ppl: 11.09231392920473], batch size: 70
+2022-12-09 23:19:58,913 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:19:59,658 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.385, over 211138.00 frames. , ppl: 10.86154424255148
+2022-12-09 23:21:39,312 INFO [train.py:421] (2/8) Epoch 1, batch 19200, loss[loss=2.5, over 910.00 frames. , ppl: 12.186851513659335] tot_loss[loss=2.405, over 5436585.87 frames. , ppl: 11.079824605206614], batch size: 70
+2022-12-09 23:23:22,546 INFO [train.py:421] (2/8) Epoch 1, batch 19400, loss[loss=2.39, over 2100.00 frames. , ppl: 10.918141185801126] tot_loss[loss=2.404, over 5446849.01 frames. , ppl: 11.071960844238395], batch size: 70
+2022-12-09 23:25:00,328 INFO [train.py:421] (2/8) Epoch 1, batch 19600, loss[loss=2.42, over 2800.00 frames. , ppl: 11.245863237126992] tot_loss[loss=2.405, over 5415765.31 frames. , ppl: 11.078757940914628], batch size: 70
+2022-12-09 23:26:38,753 INFO [train.py:421] (2/8) Epoch 1, batch 19800, loss[loss=2.415, over 2030.00 frames. , ppl: 11.189163170509394] tot_loss[loss=2.405, over 5388883.13 frames. , ppl: 11.077034489521761], batch size: 70
+2022-12-09 23:28:20,837 INFO [train.py:421] (2/8) Epoch 1, batch 20000, loss[loss=2.616, over 1050.00 frames. , ppl: 13.68611846546223] tot_loss[loss=2.404, over 5375587.39 frames. , ppl: 11.07213813749936], batch size: 70
+2022-12-09 23:28:20,837 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:28:21,595 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.383, over 211138.00 frames. , ppl: 10.840206792981473
+2022-12-09 23:30:07,792 INFO [train.py:421] (2/8) Epoch 1, batch 20200, loss[loss=2.397, over 2380.00 frames. , ppl: 10.990599968134736] tot_loss[loss=2.403, over 5414134.46 frames. , ppl: 11.061477260046406], batch size: 70
+2022-12-09 23:31:46,093 INFO [train.py:421] (2/8) Epoch 1, batch 20400, loss[loss=2.434, over 3430.00 frames. , ppl: 11.39933834571622] tot_loss[loss=2.403, over 5420902.13 frames. , ppl: 11.05403135799571], batch size: 70
+2022-12-09 23:33:26,949 INFO [train.py:421] (2/8) Epoch 1, batch 20600, loss[loss=2.421, over 2030.00 frames. , ppl: 11.259536511929939] tot_loss[loss=2.403, over 5408280.75 frames. , ppl: 11.054407505644429], batch size: 70
+2022-12-09 23:35:04,515 INFO [train.py:421] (2/8) Epoch 1, batch 20800, loss[loss=2.351, over 4480.00 frames. , ppl: 10.495350105806603] tot_loss[loss=2.401, over 5462754.81 frames. , ppl: 11.033498798858577], batch size: 70
+2022-12-09 23:36:44,985 INFO [train.py:421] (2/8) Epoch 1, batch 21000, loss[loss=2.937, over 770.00 frames. , ppl: 18.86355516539902] tot_loss[loss=2.401, over 5474788.80 frames. , ppl: 11.033441890214517], batch size: 70
+2022-12-09 23:36:44,985 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:36:45,739 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.381, over 211138.00 frames. , ppl: 10.817335716309938
+2022-12-09 23:38:28,715 INFO [train.py:421] (2/8) Epoch 1, batch 21200, loss[loss=2.343, over 8400.00 frames. , ppl: 10.414576175358293] tot_loss[loss=2.401, over 5524201.30 frames. , ppl: 11.030936241294489], batch size: 70
+2022-12-09 23:40:05,767 INFO [train.py:421] (2/8) Epoch 1, batch 21400, loss[loss=2.395, over 2310.00 frames. , ppl: 10.967357429808867] tot_loss[loss=2.401, over 5520366.30 frames. , ppl: 11.034819206687951], batch size: 70
+2022-12-09 23:41:45,963 INFO [train.py:421] (2/8) Epoch 1, batch 21600, loss[loss=2.671, over 910.00 frames. , ppl: 14.454127249314297] tot_loss[loss=2.402, over 5508952.67 frames. , ppl: 11.042243729171942], batch size: 70
+2022-12-09 23:43:25,697 INFO [train.py:421] (2/8) Epoch 1, batch 21800, loss[loss=2.542, over 1470.00 frames. , ppl: 12.702530087783723] tot_loss[loss=2.401, over 5541739.08 frames. , ppl: 11.03481190328151], batch size: 70
+2022-12-09 23:45:06,450 INFO [train.py:421] (2/8) Epoch 1, batch 22000, loss[loss=2.371, over 1820.00 frames. , ppl: 10.703271012916069] tot_loss[loss=2.4, over 5543801.26 frames. , ppl: 11.028551372474515], batch size: 70
+2022-12-09 23:45:06,450 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:45:07,207 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.382, over 211138.00 frames. , ppl: 10.821137268370757
+2022-12-09 23:46:46,113 INFO [train.py:421] (2/8) Epoch 1, batch 22200, loss[loss=2.507, over 1190.00 frames. , ppl: 12.263612221231087] tot_loss[loss=2.401, over 5469322.74 frames. , ppl: 11.038834104702538], batch size: 70
+2022-12-09 23:48:27,624 INFO [train.py:421] (2/8) Epoch 1, batch 22400, loss[loss=2.47, over 1260.00 frames. , ppl: 11.821138450117324] tot_loss[loss=2.401, over 5493416.41 frames. , ppl: 11.036517377134645], batch size: 70
+2022-12-09 23:50:05,038 INFO [train.py:421] (2/8) Epoch 1, batch 22600, loss[loss=2.352, over 2450.00 frames. , ppl: 10.5048005731933] tot_loss[loss=2.401, over 5456462.50 frames. , ppl: 11.035578072349411], batch size: 70
+2022-12-09 23:51:43,332 INFO [train.py:421] (2/8) Epoch 1, batch 22800, loss[loss=2.3, over 3080.00 frames. , ppl: 9.976863009636876] tot_loss[loss=2.401, over 5450082.03 frames. , ppl: 11.038401307939651], batch size: 70
+2022-12-09 23:53:27,617 INFO [train.py:421] (2/8) Epoch 1, batch 23000, loss[loss=2.566, over 1190.00 frames. , ppl: 13.007243796766518] tot_loss[loss=2.4, over 5486895.70 frames. , ppl: 11.017830766971231], batch size: 70
+2022-12-09 23:53:27,617 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-09 23:53:28,375 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.38, over 211138.00 frames. , ppl: 10.800644888892075
+2022-12-09 23:55:09,816 INFO [train.py:421] (2/8) Epoch 1, batch 23200, loss[loss=2.437, over 1400.00 frames. , ppl: 11.434511318394204] tot_loss[loss=2.399, over 5491283.42 frames. , ppl: 11.013012515770422], batch size: 70
+2022-12-09 23:56:52,352 INFO [train.py:421] (2/8) Epoch 1, batch 23400, loss[loss=2.449, over 1260.00 frames. , ppl: 11.572774213812183] tot_loss[loss=2.399, over 5461279.95 frames. , ppl: 11.013958045162003], batch size: 70
+2022-12-09 23:58:28,672 INFO [train.py:421] (2/8) Epoch 1, batch 23600, loss[loss=2.471, over 1540.00 frames. , ppl: 11.8359382477211] tot_loss[loss=2.4, over 5401642.71 frames. , ppl: 11.027820211358021], batch size: 70
+2022-12-10 00:00:06,981 INFO [train.py:421] (2/8) Epoch 1, batch 23800, loss[loss=2.725, over 840.00 frames. , ppl: 15.256209955984755] tot_loss[loss=2.401, over 5403501.97 frames. , ppl: 11.032425396262578], batch size: 70
+2022-12-10 00:01:47,275 INFO [train.py:421] (2/8) Epoch 1, batch 24000, loss[loss=2.288, over 4340.00 frames. , ppl: 9.854422378877848] tot_loss[loss=2.399, over 5440422.21 frames. , ppl: 11.015184811809059], batch size: 70
+2022-12-10 00:01:47,276 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:01:48,035 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.379, over 211138.00 frames. , ppl: 10.78979920368168
+2022-12-10 00:03:26,226 INFO [train.py:421] (2/8) Epoch 1, batch 24200, loss[loss=2.519, over 1610.00 frames. , ppl: 12.416741689821936] tot_loss[loss=2.4, over 5424782.32 frames. , ppl: 11.021746516384614], batch size: 70
+2022-12-10 00:05:05,803 INFO [train.py:421] (2/8) Epoch 1, batch 24400, loss[loss=2.357, over 2450.00 frames. , ppl: 10.55930786222987] tot_loss[loss=2.399, over 5428203.39 frames. , ppl: 11.016988653113966], batch size: 70
+2022-12-10 00:06:43,340 INFO [train.py:421] (2/8) Epoch 1, batch 24600, loss[loss=2.318, over 3360.00 frames. , ppl: 10.157977867998582] tot_loss[loss=2.4, over 5392796.91 frames. , ppl: 11.026144270410539], batch size: 70
+2022-12-10 00:08:25,287 INFO [train.py:421] (2/8) Epoch 1, batch 24800, loss[loss=2.427, over 3290.00 frames. , ppl: 11.319199046216545] tot_loss[loss=2.4, over 5427358.54 frames. , ppl: 11.023406171144016], batch size: 70
+2022-12-10 00:10:06,641 INFO [train.py:421] (2/8) Epoch 1, batch 25000, loss[loss=2.325, over 5530.00 frames. , ppl: 10.223444306641984] tot_loss[loss=2.399, over 5461541.97 frames. , ppl: 11.01075807339044], batch size: 70
+2022-12-10 00:10:06,641 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:10:07,386 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.38, over 211138.00 frames. , ppl: 10.800197296869092
+2022-12-10 00:11:43,308 INFO [train.py:421] (2/8) Epoch 1, batch 25200, loss[loss=2.357, over 12320.00 frames. , ppl: 10.556365958608561] tot_loss[loss=2.398, over 5441933.18 frames. , ppl: 11.006015829785463], batch size: 70
+2022-12-10 00:13:24,086 INFO [train.py:421] (2/8) Epoch 1, batch 25400, loss[loss=2.408, over 2870.00 frames. , ppl: 11.108057871719497] tot_loss[loss=2.398, over 5434724.45 frames. , ppl: 11.00593584500469], batch size: 70
+2022-12-10 00:15:07,554 INFO [train.py:421] (2/8) Epoch 1, batch 25600, loss[loss=2.375, over 2660.00 frames. , ppl: 10.755026066540436] tot_loss[loss=2.397, over 5453767.98 frames. , ppl: 10.995301684453295], batch size: 70
+2022-12-10 00:16:48,181 INFO [train.py:421] (2/8) Epoch 1, batch 25800, loss[loss=2.525, over 1120.00 frames. , ppl: 12.491292798599222] tot_loss[loss=2.397, over 5456629.26 frames. , ppl: 10.990693436877843], batch size: 70
+2022-12-10 00:18:27,901 INFO [train.py:421] (2/8) Epoch 1, batch 26000, loss[loss=2.379, over 1330.00 frames. , ppl: 10.79352932204098] tot_loss[loss=2.399, over 5410700.32 frames. , ppl: 11.007072409439168], batch size: 70
+2022-12-10 00:18:27,901 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:18:28,648 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.379, over 211138.00 frames. , ppl: 10.799350119895008
+2022-12-10 00:20:08,713 INFO [train.py:421] (2/8) Epoch 1, batch 26200, loss[loss=2.3, over 3360.00 frames. , ppl: 9.970826056057865] tot_loss[loss=2.398, over 5397479.82 frames. , ppl: 11.002235137374054], batch size: 70
+2022-12-10 00:21:44,686 INFO [train.py:421] (2/8) Epoch 1, batch 26400, loss[loss=2.483, over 1750.00 frames. , ppl: 11.9791673247059] tot_loss[loss=2.398, over 5361690.24 frames. , ppl: 11.006079013571346], batch size: 70
+2022-12-10 00:23:22,432 INFO [train.py:421] (2/8) Epoch 1, batch 26600, loss[loss=2.296, over 8470.00 frames. , ppl: 9.933661154255587] tot_loss[loss=2.397, over 5403893.81 frames. , ppl: 10.991572539898502], batch size: 70
+2022-12-10 00:25:04,486 INFO [train.py:421] (2/8) Epoch 1, batch 26800, loss[loss=2.331, over 5250.00 frames. , ppl: 10.290475410567026] tot_loss[loss=2.397, over 5411031.87 frames. , ppl: 10.98887530696989], batch size: 70
+2022-12-10 00:26:48,138 INFO [train.py:421] (2/8) Epoch 1, batch 27000, loss[loss=3.098, over 560.00 frames. , ppl: 22.1446485930108] tot_loss[loss=2.397, over 5408238.88 frames. , ppl: 10.98717674251201], batch size: 70
+2022-12-10 00:26:48,139 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:26:48,899 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.379, over 211138.00 frames. , ppl: 10.791004984020308
+2022-12-10 00:28:29,579 INFO [train.py:421] (2/8) Epoch 1, batch 27200, loss[loss=2.413, over 2940.00 frames. , ppl: 11.171177832371384] tot_loss[loss=2.397, over 5413798.08 frames. , ppl: 10.988356296945097], batch size: 70
+2022-12-10 00:30:09,144 INFO [train.py:421] (2/8) Epoch 1, batch 27400, loss[loss=2.377, over 6230.00 frames. , ppl: 10.76847293786593] tot_loss[loss=2.397, over 5420921.81 frames. , ppl: 10.991139230567313], batch size: 70
+2022-12-10 00:31:48,115 INFO [train.py:421] (2/8) Epoch 1, batch 27600, loss[loss=3.223, over 490.00 frames. , ppl: 25.097534168284795] tot_loss[loss=2.398, over 5392604.94 frames. , ppl: 10.997128471981423], batch size: 70
+2022-12-10 00:33:29,296 INFO [train.py:421] (2/8) Epoch 1, batch 27800, loss[loss=2.567, over 1610.00 frames. , ppl: 13.022088806791134] tot_loss[loss=2.396, over 5419746.64 frames. , ppl: 10.982522871222024], batch size: 70
+2022-12-10 00:35:07,716 INFO [train.py:421] (2/8) Epoch 1, batch 28000, loss[loss=2.521, over 1610.00 frames. , ppl: 12.44642111113349] tot_loss[loss=2.396, over 5420756.27 frames. , ppl: 10.981597158594658], batch size: 70
+2022-12-10 00:35:07,717 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:35:08,463 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.377, over 211138.00 frames. , ppl: 10.77500384045564
+2022-12-10 00:36:52,790 INFO [train.py:421] (2/8) Epoch 1, batch 28200, loss[loss=2.67, over 980.00 frames. , ppl: 14.440046175863616] tot_loss[loss=2.394, over 5490011.44 frames. , ppl: 10.957363590439847], batch size: 70
+2022-12-10 00:38:30,674 INFO [train.py:421] (2/8) Epoch 1, batch 28400, loss[loss=2.534, over 1400.00 frames. , ppl: 12.608653583133652] tot_loss[loss=2.394, over 5484189.59 frames. , ppl: 10.95624265083344], batch size: 70
+2022-12-10 00:40:09,191 INFO [train.py:421] (2/8) Epoch 1, batch 28600, loss[loss=2.303, over 5950.00 frames. , ppl: 10.009086231254933] tot_loss[loss=2.394, over 5458203.69 frames. , ppl: 10.961021932342625], batch size: 70
+2022-12-10 00:41:48,525 INFO [train.py:421] (2/8) Epoch 1, batch 28800, loss[loss=2.356, over 5460.00 frames. , ppl: 10.551068202592946] tot_loss[loss=2.395, over 5443708.39 frames. , ppl: 10.965008955190966], batch size: 70
+2022-12-10 00:43:30,618 INFO [train.py:421] (2/8) Epoch 1, batch 29000, loss[loss=2.281, over 3500.00 frames. , ppl: 9.783147595729133] tot_loss[loss=2.394, over 5491585.22 frames. , ppl: 10.957473007768609], batch size: 70
+2022-12-10 00:43:30,618 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:43:31,364 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.375, over 211138.00 frames. , ppl: 10.749719593753966
+2022-12-10 00:45:08,473 INFO [train.py:421] (2/8) Epoch 1, batch 29200, loss[loss=2.516, over 1260.00 frames. , ppl: 12.381631425555968] tot_loss[loss=2.394, over 5501388.73 frames. , ppl: 10.953500314560747], batch size: 70
+2022-12-10 00:46:48,118 INFO [train.py:421] (2/8) Epoch 1, batch 29400, loss[loss=2.462, over 2030.00 frames. , ppl: 11.727256481084337] tot_loss[loss=2.393, over 5517372.86 frames. , ppl: 10.948599021871413], batch size: 70
+2022-12-10 00:48:31,855 INFO [train.py:421] (2/8) Epoch 1, batch 29600, loss[loss=2.349, over 1400.00 frames. , ppl: 10.477309815431095] tot_loss[loss=2.393, over 5512512.20 frames. , ppl: 10.94617255499015], batch size: 70
+2022-12-10 00:50:13,022 INFO [train.py:421] (2/8) Epoch 1, batch 29800, loss[loss=2.42, over 2310.00 frames. , ppl: 11.243098393748454] tot_loss[loss=2.393, over 5501886.66 frames. , ppl: 10.943050137633788], batch size: 70
+2022-12-10 00:51:49,940 INFO [train.py:421] (2/8) Epoch 1, batch 30000, loss[loss=2.63, over 980.00 frames. , ppl: 13.870402515148252] tot_loss[loss=2.392, over 5505724.75 frames. , ppl: 10.932922588133431], batch size: 70
+2022-12-10 00:51:49,940 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 00:51:50,686 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.373, over 211138.00 frames. , ppl: 10.73370182407569
+2022-12-10 00:53:27,749 INFO [train.py:421] (2/8) Epoch 1, batch 30200, loss[loss=2.554, over 2030.00 frames. , ppl: 12.85964490402234] tot_loss[loss=2.392, over 5500847.07 frames. , ppl: 10.937140268967388], batch size: 70
+2022-12-10 00:55:09,215 INFO [train.py:421] (2/8) Epoch 1, batch 30400, loss[loss=2.612, over 1400.00 frames. , ppl: 13.629667952985018] tot_loss[loss=2.393, over 5467236.09 frames. , ppl: 10.944859951008473], batch size: 70
+2022-12-10 00:56:50,132 INFO [train.py:421] (2/8) Epoch 1, batch 30600, loss[loss=3.466, over 490.00 frames. , ppl: 32.01772515032718] tot_loss[loss=2.392, over 5488706.96 frames. , ppl: 10.936782975598943], batch size: 70
+2022-12-10 00:58:29,426 INFO [train.py:421] (2/8) Epoch 1, batch 30800, loss[loss=2.742, over 1260.00 frames. , ppl: 15.51068128258092] tot_loss[loss=2.392, over 5496515.34 frames. , ppl: 10.93366628450317], batch size: 70
+2022-12-10 01:00:09,322 INFO [train.py:421] (2/8) Epoch 1, batch 31000, loss[loss=2.351, over 3500.00 frames. , ppl: 10.495515835692112] tot_loss[loss=2.39, over 5551562.43 frames. , ppl: 10.917349438738327], batch size: 70
+2022-12-10 01:00:09,322 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:00:10,085 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.373, over 211138.00 frames. , ppl: 10.729127435277624
+2022-12-10 01:01:48,639 INFO [train.py:421] (2/8) Epoch 1, batch 31200, loss[loss=2.381, over 4200.00 frames. , ppl: 10.815082474021027] tot_loss[loss=2.39, over 5568377.56 frames. , ppl: 10.908563568167587], batch size: 70
+2022-12-10 01:03:31,207 INFO [train.py:421] (2/8) Epoch 1, batch 31400, loss[loss=2.323, over 3500.00 frames. , ppl: 10.205697569182746] tot_loss[loss=2.389, over 5588964.50 frames. , ppl: 10.907708325632742], batch size: 70
+2022-12-10 01:05:10,005 INFO [train.py:421] (2/8) Epoch 1, batch 31600, loss[loss=2.303, over 7490.00 frames. , ppl: 10.005716228066866] tot_loss[loss=2.388, over 5619760.64 frames. , ppl: 10.894992536201373], batch size: 70
+2022-12-10 01:06:49,277 INFO [train.py:421] (2/8) Epoch 1, batch 31800, loss[loss=2.265, over 3990.00 frames. , ppl: 9.633606153560773] tot_loss[loss=2.388, over 5609789.33 frames. , ppl: 10.896663693755192], batch size: 70
+2022-12-10 01:08:31,499 INFO [train.py:421] (2/8) Epoch 1, batch 32000, loss[loss=2.377, over 1610.00 frames. , ppl: 10.772302937824389] tot_loss[loss=2.389, over 5587496.10 frames. , ppl: 10.904057613980449], batch size: 70
+2022-12-10 01:08:31,499 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:08:32,284 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.372, over 211138.00 frames. , ppl: 10.718152095318677
+2022-12-10 01:10:13,323 INFO [train.py:421] (2/8) Epoch 1, batch 32200, loss[loss=2.384, over 2030.00 frames. , ppl: 10.845887062891409] tot_loss[loss=2.39, over 5568298.08 frames. , ppl: 10.916680807218867], batch size: 70
+2022-12-10 01:11:53,782 INFO [train.py:421] (2/8) Epoch 1, batch 32400, loss[loss=2.672, over 910.00 frames. , ppl: 14.462273019907837] tot_loss[loss=2.391, over 5525761.02 frames. , ppl: 10.923108619840274], batch size: 70
+2022-12-10 01:13:33,493 INFO [train.py:421] (2/8) Epoch 1, batch 32600, loss[loss=2.676, over 840.00 frames. , ppl: 14.526073005057315] tot_loss[loss=2.391, over 5501543.27 frames. , ppl: 10.928540546120711], batch size: 70
+2022-12-10 01:15:16,739 INFO [train.py:421] (2/8) Epoch 1, batch 32800, loss[loss=2.548, over 1750.00 frames. , ppl: 12.778651777430923] tot_loss[loss=2.391, over 5527379.31 frames. , ppl: 10.9204133308793], batch size: 70
+2022-12-10 01:16:56,205 INFO [train.py:421] (2/8) Epoch 1, batch 33000, loss[loss=2.259, over 4830.00 frames. , ppl: 9.570311250695877] tot_loss[loss=2.389, over 5561285.53 frames. , ppl: 10.906591690205884], batch size: 70
+2022-12-10 01:16:56,205 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:16:56,953 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.371, over 211138.00 frames. , ppl: 10.707864689524108
+2022-12-10 01:18:40,594 INFO [train.py:421] (2/8) Epoch 1, batch 33200, loss[loss=2.318, over 4760.00 frames. , ppl: 10.158097352093801] tot_loss[loss=2.39, over 5552265.16 frames. , ppl: 10.913789187688883], batch size: 70
+2022-12-10 01:20:20,452 INFO [train.py:421] (2/8) Epoch 1, batch 33400, loss[loss=2.369, over 3360.00 frames. , ppl: 10.683215354493873] tot_loss[loss=2.388, over 5599240.47 frames. , ppl: 10.896251489384982], batch size: 70
+2022-12-10 01:21:57,275 INFO [train.py:421] (2/8) Epoch 1, batch 33600, loss[loss=2.459, over 1890.00 frames. , ppl: 11.691830331593929] tot_loss[loss=2.389, over 5578770.87 frames. , ppl: 10.898774656372911], batch size: 70
+2022-12-10 01:23:38,628 INFO [train.py:421] (2/8) Epoch 1, batch 33800, loss[loss=2.574, over 1260.00 frames. , ppl: 13.119773405739501] tot_loss[loss=2.388, over 5609129.91 frames. , ppl: 10.895687549808445], batch size: 70
+2022-12-10 01:25:17,021 INFO [train.py:421] (2/8) Epoch 1, batch 34000, loss[loss=2.282, over 3990.00 frames. , ppl: 9.795192862264576] tot_loss[loss=2.388, over 5609953.14 frames. , ppl: 10.88778217317525], batch size: 70
+2022-12-10 01:25:17,021 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:25:17,782 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.373, over 211138.00 frames. , ppl: 10.73305843282348
+2022-12-10 01:27:03,119 INFO [train.py:421] (2/8) Epoch 1, batch 34200, loss[loss=2.791, over 630.00 frames. , ppl: 16.30088601437611] tot_loss[loss=2.387, over 5626605.13 frames. , ppl: 10.880848641963022], batch size: 70
+2022-12-10 01:28:41,223 INFO [train.py:421] (2/8) Epoch 1, batch 34400, loss[loss=2.285, over 2660.00 frames. , ppl: 9.822846252806773] tot_loss[loss=2.387, over 5588632.56 frames. , ppl: 10.885816034843584], batch size: 70
+2022-12-10 01:30:22,448 INFO [train.py:421] (2/8) Epoch 1, batch 34600, loss[loss=2.493, over 2170.00 frames. , ppl: 12.095263625857958] tot_loss[loss=2.387, over 5607549.19 frames. , ppl: 10.87776792327517], batch size: 70
+2022-12-10 01:32:03,744 INFO [train.py:421] (2/8) Epoch 1, batch 34800, loss[loss=2.487, over 1050.00 frames. , ppl: 12.027555251178164] tot_loss[loss=2.386, over 5601384.46 frames. , ppl: 10.867751862619592], batch size: 70
+2022-12-10 01:33:43,070 INFO [train.py:421] (2/8) Epoch 1, batch 35000, loss[loss=2.401, over 2590.00 frames. , ppl: 11.03357203304402] tot_loss[loss=2.386, over 5608808.39 frames. , ppl: 10.867889339541717], batch size: 70
+2022-12-10 01:33:43,070 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:33:43,828 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.372, over 211138.00 frames. , ppl: 10.721890226540799
+2022-12-10 01:35:22,933 INFO [train.py:421] (2/8) Epoch 1, batch 35200, loss[loss=2.424, over 3080.00 frames. , ppl: 11.289096178481678] tot_loss[loss=2.386, over 5582990.68 frames. , ppl: 10.872805828459501], batch size: 70
+2022-12-10 01:37:03,046 INFO [train.py:421] (2/8) Epoch 1, batch 35400, loss[loss=2.358, over 2450.00 frames. , ppl: 10.565404119823315] tot_loss[loss=2.387, over 5553455.11 frames. , ppl: 10.878143310843809], batch size: 70
+2022-12-10 01:38:46,643 INFO [train.py:421] (2/8) Epoch 1, batch 35600, loss[loss=2.346, over 2100.00 frames. , ppl: 10.440603919275711] tot_loss[loss=2.387, over 5546221.01 frames. , ppl: 10.877623844982606], batch size: 70
+2022-12-10 01:40:24,015 INFO [train.py:421] (2/8) Epoch 1, batch 35800, loss[loss=2.711, over 910.00 frames. , ppl: 15.051276498487363] tot_loss[loss=2.387, over 5531408.64 frames. , ppl: 10.87859174790672], batch size: 70
+2022-12-10 01:42:02,432 INFO [train.py:421] (2/8) Epoch 1, batch 36000, loss[loss=2.514, over 1470.00 frames. , ppl: 12.35673606320898] tot_loss[loss=2.388, over 5496379.75 frames. , ppl: 10.89655418082219], batch size: 70
+2022-12-10 01:42:02,433 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:42:03,181 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.371, over 211138.00 frames. , ppl: 10.709730213282327
+2022-12-10 01:43:40,117 INFO [train.py:421] (2/8) Epoch 1, batch 36200, loss[loss=2.48, over 2030.00 frames. , ppl: 11.940071913467019] tot_loss[loss=2.388, over 5526928.56 frames. , ppl: 10.887811874705706], batch size: 70
+2022-12-10 01:45:21,067 INFO [train.py:421] (2/8) Epoch 1, batch 36400, loss[loss=2.364, over 2240.00 frames. , ppl: 10.631438689754399] tot_loss[loss=2.388, over 5510800.18 frames. , ppl: 10.895293823303396], batch size: 70
+2022-12-10 01:47:02,480 INFO [train.py:421] (2/8) Epoch 1, batch 36600, loss[loss=2.405, over 1540.00 frames. , ppl: 11.081927266846977] tot_loss[loss=2.387, over 5533124.28 frames. , ppl: 10.876042891173809], batch size: 70
+2022-12-10 01:48:44,159 INFO [train.py:421] (2/8) Epoch 1, batch 36800, loss[loss=3.059, over 560.00 frames. , ppl: 21.305842337316435] tot_loss[loss=2.385, over 5558396.54 frames. , ppl: 10.861029626900821], batch size: 70
+2022-12-10 01:50:20,607 INFO [train.py:421] (2/8) Epoch 1, batch 37000, loss[loss=2.318, over 6230.00 frames. , ppl: 10.159929924623466] tot_loss[loss=2.386, over 5525165.68 frames. , ppl: 10.874949553444106], batch size: 70
+2022-12-10 01:50:20,607 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:50:21,353 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.371, over 211138.00 frames. , ppl: 10.706359193662431
+2022-12-10 01:52:04,192 INFO [train.py:421] (2/8) Epoch 1, batch 37200, loss[loss=2.47, over 840.00 frames. , ppl: 11.817710080588949] tot_loss[loss=2.386, over 5493022.68 frames. , ppl: 10.872611436996758], batch size: 70
+2022-12-10 01:53:44,571 INFO [train.py:421] (2/8) Epoch 1, batch 37400, loss[loss=2.565, over 1330.00 frames. , ppl: 12.994690731517359] tot_loss[loss=2.387, over 5483333.46 frames. , ppl: 10.886212579823747], batch size: 70
+2022-12-10 01:55:24,444 INFO [train.py:421] (2/8) Epoch 1, batch 37600, loss[loss=2.466, over 2100.00 frames. , ppl: 11.76996085264386] tot_loss[loss=2.387, over 5489643.66 frames. , ppl: 10.881268942565148], batch size: 70
+2022-12-10 01:57:05,222 INFO [train.py:421] (2/8) Epoch 1, batch 37800, loss[loss=2.252, over 3850.00 frames. , ppl: 9.50795585210101] tot_loss[loss=2.387, over 5475744.86 frames. , ppl: 10.883263214034816], batch size: 70
+2022-12-10 01:58:47,483 INFO [train.py:421] (2/8) Epoch 1, batch 38000, loss[loss=2.442, over 2450.00 frames. , ppl: 11.493493934175762] tot_loss[loss=2.386, over 5506691.19 frames. , ppl: 10.873601701407557], batch size: 70
+2022-12-10 01:58:47,483 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 01:58:48,230 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.367, over 211138.00 frames. , ppl: 10.660825546799813
+2022-12-10 02:00:27,318 INFO [train.py:421] (2/8) Epoch 1, batch 38200, loss[loss=2.484, over 1330.00 frames. , ppl: 11.986530434578647] tot_loss[loss=2.386, over 5509066.58 frames. , ppl: 10.871234463282372], batch size: 70
+2022-12-10 02:02:09,571 INFO [train.py:421] (2/8) Epoch 1, batch 38400, loss[loss=2.589, over 1050.00 frames. , ppl: 13.31354698387978] tot_loss[loss=2.385, over 5577350.15 frames. , ppl: 10.857393823802596], batch size: 70
+2022-12-10 02:03:49,946 INFO [train.py:421] (2/8) Epoch 1, batch 38600, loss[loss=2.398, over 2730.00 frames. , ppl: 10.997965102409328] tot_loss[loss=2.385, over 5557087.40 frames. , ppl: 10.862806292403453], batch size: 70
+2022-12-10 02:05:31,070 INFO [train.py:421] (2/8) Epoch 1, batch 38800, loss[loss=2.28, over 5390.00 frames. , ppl: 9.777274181775779] tot_loss[loss=2.385, over 5566507.57 frames. , ppl: 10.858471569911371], batch size: 70
+2022-12-10 02:07:09,883 INFO [train.py:421] (2/8) Epoch 1, batch 39000, loss[loss=2.42, over 2030.00 frames. , ppl: 11.249038675731617] tot_loss[loss=2.385, over 5586252.84 frames. , ppl: 10.86054563029782], batch size: 70
+2022-12-10 02:07:09,883 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:07:10,643 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.368, over 211138.00 frames. , ppl: 10.674933246579991
+2022-12-10 02:08:52,861 INFO [train.py:421] (2/8) Epoch 1, batch 39200, loss[loss=2.626, over 770.00 frames. , ppl: 13.816949011575351] tot_loss[loss=2.386, over 5543849.65 frames. , ppl: 10.868072088119382], batch size: 70
+2022-12-10 02:10:35,768 INFO [train.py:421] (2/8) Epoch 1, batch 39400, loss[loss=2.516, over 1120.00 frames. , ppl: 12.373978880189698] tot_loss[loss=2.386, over 5533471.03 frames. , ppl: 10.866491244867053], batch size: 70
+2022-12-10 02:12:14,141 INFO [train.py:421] (2/8) Epoch 1, batch 39600, loss[loss=2.348, over 1890.00 frames. , ppl: 10.467308389260142] tot_loss[loss=2.386, over 5514708.55 frames. , ppl: 10.871285997845613], batch size: 70
+2022-12-10 02:13:56,724 INFO [train.py:421] (2/8) Epoch 1, batch 39800, loss[loss=2.578, over 1470.00 frames. , ppl: 13.175484224635639] tot_loss[loss=2.385, over 5567348.66 frames. , ppl: 10.857472112118495], batch size: 70
+2022-12-10 02:15:38,369 INFO [train.py:421] (2/8) Epoch 1, batch 40000, loss[loss=2.61, over 1190.00 frames. , ppl: 13.60364840620564] tot_loss[loss=2.386, over 5527304.01 frames. , ppl: 10.87053197821333], batch size: 70
+2022-12-10 02:15:38,369 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:15:39,131 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.368, over 211138.00 frames. , ppl: 10.68034759574071
+2022-12-10 02:17:21,026 INFO [train.py:421] (2/8) Epoch 1, batch 40200, loss[loss=2.359, over 3780.00 frames. , ppl: 10.581261087561941] tot_loss[loss=2.385, over 5540658.62 frames. , ppl: 10.863928512593414], batch size: 70
+2022-12-10 02:19:03,094 INFO [train.py:421] (2/8) Epoch 1, batch 40400, loss[loss=2.908, over 700.00 frames. , ppl: 18.32737140015592] tot_loss[loss=2.385, over 5564091.83 frames. , ppl: 10.857212471569241], batch size: 70
+2022-12-10 02:20:42,126 INFO [train.py:421] (2/8) Epoch 1, batch 40600, loss[loss=2.48, over 1680.00 frames. , ppl: 11.938199531309957] tot_loss[loss=2.384, over 5577061.65 frames. , ppl: 10.845629137033407], batch size: 70
+2022-12-10 02:22:24,716 INFO [train.py:421] (2/8) Epoch 1, batch 40800, loss[loss=2.481, over 2030.00 frames. , ppl: 11.950792069083999] tot_loss[loss=2.384, over 5569925.85 frames. , ppl: 10.844613043985216], batch size: 70
+2022-12-10 02:24:04,620 INFO [train.py:421] (2/8) Epoch 1, batch 41000, loss[loss=2.388, over 5740.00 frames. , ppl: 10.89399626161463] tot_loss[loss=2.385, over 5509112.35 frames. , ppl: 10.864481785740974], batch size: 70
+2022-12-10 02:24:04,621 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:24:05,368 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.367, over 211138.00 frames. , ppl: 10.668684715535875
+2022-12-10 02:25:40,536 INFO [train.py:421] (2/8) Epoch 1, batch 41200, loss[loss=2.383, over 3080.00 frames. , ppl: 10.837167428988229] tot_loss[loss=2.386, over 5490639.73 frames. , ppl: 10.870377690802526], batch size: 70
+2022-12-10 02:27:22,732 INFO [train.py:421] (2/8) Epoch 1, batch 41400, loss[loss=2.411, over 1890.00 frames. , ppl: 11.148833148783275] tot_loss[loss=2.386, over 5490299.70 frames. , ppl: 10.86639473652908], batch size: 70
+2022-12-10 02:29:04,770 INFO [train.py:421] (2/8) Epoch 1, batch 41600, loss[loss=2.488, over 2030.00 frames. , ppl: 12.037468022284578] tot_loss[loss=2.386, over 5483768.98 frames. , ppl: 10.869306635952723], batch size: 70
+2022-12-10 02:30:42,768 INFO [train.py:421] (2/8) Epoch 1, batch 41800, loss[loss=2.501, over 1120.00 frames. , ppl: 12.191039934925826] tot_loss[loss=2.386, over 5477205.14 frames. , ppl: 10.864530973966383], batch size: 70
+2022-12-10 02:32:23,879 INFO [train.py:421] (2/8) Epoch 1, batch 42000, loss[loss=2.342, over 3850.00 frames. , ppl: 10.405728092293382] tot_loss[loss=2.383, over 5551509.08 frames. , ppl: 10.83952226386799], batch size: 70
+2022-12-10 02:32:23,880 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:32:24,639 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.364, over 211138.00 frames. , ppl: 10.63275418399596
+2022-12-10 02:34:04,507 INFO [train.py:421] (2/8) Epoch 1, batch 42200, loss[loss=2.425, over 1330.00 frames. , ppl: 11.30323153739545] tot_loss[loss=2.384, over 5536198.68 frames. , ppl: 10.842887806099665], batch size: 70
+2022-12-10 02:35:44,518 INFO [train.py:421] (2/8) Epoch 1, batch 42400, loss[loss=2.768, over 770.00 frames. , ppl: 15.930220468822617] tot_loss[loss=2.382, over 5585360.56 frames. , ppl: 10.824910017718596], batch size: 70
+2022-12-10 02:37:26,135 INFO [train.py:421] (2/8) Epoch 1, batch 42600, loss[loss=2.372, over 2380.00 frames. , ppl: 10.719562170451956] tot_loss[loss=2.381, over 5589759.28 frames. , ppl: 10.820211024103678], batch size: 70
+2022-12-10 02:39:06,197 INFO [train.py:421] (2/8) Epoch 1, batch 42800, loss[loss=2.329, over 8260.00 frames. , ppl: 10.266075017539382] tot_loss[loss=2.381, over 5557093.55 frames. , ppl: 10.819453997807134], batch size: 70
+2022-12-10 02:40:45,344 INFO [train.py:421] (2/8) Epoch 1, batch 43000, loss[loss=2.38, over 3570.00 frames. , ppl: 10.810075303132615] tot_loss[loss=2.383, over 5518623.98 frames. , ppl: 10.832086585886252], batch size: 70
+2022-12-10 02:40:45,345 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:40:46,105 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.364, over 211138.00 frames. , ppl: 10.630343507571958
+2022-12-10 02:42:23,862 INFO [train.py:421] (2/8) Epoch 1, batch 43200, loss[loss=2.527, over 1400.00 frames. , ppl: 12.514368215226725] tot_loss[loss=2.383, over 5499798.72 frames. , ppl: 10.842285968960619], batch size: 70
+2022-12-10 02:44:06,618 INFO [train.py:421] (2/8) Epoch 1, batch 43400, loss[loss=2.487, over 910.00 frames. , ppl: 12.024138954241495] tot_loss[loss=2.383, over 5503445.97 frames. , ppl: 10.84235837053094], batch size: 70
+2022-12-10 02:45:46,628 INFO [train.py:421] (2/8) Epoch 1, batch 43600, loss[loss=2.35, over 4060.00 frames. , ppl: 10.487572480613458] tot_loss[loss=2.384, over 5495085.04 frames. , ppl: 10.843157731220419], batch size: 70
+2022-12-10 02:47:26,671 INFO [train.py:421] (2/8) Epoch 1, batch 43800, loss[loss=2.601, over 1260.00 frames. , ppl: 13.471034121496027] tot_loss[loss=2.382, over 5567093.12 frames. , ppl: 10.828907561820161], batch size: 70
+2022-12-10 02:49:05,960 INFO [train.py:421] (2/8) Epoch 1, batch 44000, loss[loss=2.336, over 1680.00 frames. , ppl: 10.338694592639255] tot_loss[loss=2.381, over 5606152.04 frames. , ppl: 10.817169931321029], batch size: 70
+2022-12-10 02:49:05,961 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:49:06,722 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.365, over 211138.00 frames. , ppl: 10.648347077171886
+2022-12-10 02:50:46,031 INFO [train.py:421] (2/8) Epoch 1, batch 44200, loss[loss=2.327, over 4760.00 frames. , ppl: 10.247375750910145] tot_loss[loss=2.381, over 5606701.97 frames. , ppl: 10.810397931394535], batch size: 70
+2022-12-10 02:52:25,678 INFO [train.py:421] (2/8) Epoch 1, batch 44400, loss[loss=2.282, over 4900.00 frames. , ppl: 9.79633885129347] tot_loss[loss=2.379, over 5641970.57 frames. , ppl: 10.7983214240511], batch size: 70
+2022-12-10 02:54:07,176 INFO [train.py:421] (2/8) Epoch 1, batch 44600, loss[loss=2.304, over 3080.00 frames. , ppl: 10.01818382023282] tot_loss[loss=2.379, over 5631749.30 frames. , ppl: 10.798657342741157], batch size: 70
+2022-12-10 02:55:46,318 INFO [train.py:421] (2/8) Epoch 1, batch 44800, loss[loss=2.37, over 2100.00 frames. , ppl: 10.70212418046199] tot_loss[loss=2.379, over 5629764.19 frames. , ppl: 10.789365296633976], batch size: 70
+2022-12-10 02:57:25,620 INFO [train.py:421] (2/8) Epoch 1, batch 45000, loss[loss=2.456, over 2170.00 frames. , ppl: 11.65445329504277] tot_loss[loss=2.38, over 5587378.94 frames. , ppl: 10.808297707392768], batch size: 70
+2022-12-10 02:57:25,621 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 02:57:26,378 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.364, over 211138.00 frames. , ppl: 10.628203939461411
+2022-12-10 02:59:10,718 INFO [train.py:421] (2/8) Epoch 1, batch 45200, loss[loss=2.361, over 3430.00 frames. , ppl: 10.599919345015593] tot_loss[loss=2.379, over 5651139.65 frames. , ppl: 10.78929978300868], batch size: 70
+2022-12-10 03:00:53,879 INFO [train.py:421] (2/8) Epoch 1, batch 45400, loss[loss=2.628, over 1330.00 frames. , ppl: 13.84713531319868] tot_loss[loss=2.379, over 5612796.03 frames. , ppl: 10.79393387833019], batch size: 70
+2022-12-10 03:02:33,640 INFO [train.py:421] (2/8) Epoch 1, batch 45600, loss[loss=2.518, over 1120.00 frames. , ppl: 12.409710431776348] tot_loss[loss=2.378, over 5657096.40 frames. , ppl: 10.779844765030129], batch size: 70
+2022-12-10 03:04:17,136 INFO [train.py:421] (2/8) Epoch 1, batch 45800, loss[loss=2.541, over 1050.00 frames. , ppl: 12.697192498013926] tot_loss[loss=2.378, over 5653250.17 frames. , ppl: 10.778091772332047], batch size: 70
+2022-12-10 03:05:58,452 INFO [train.py:421] (2/8) Epoch 1, batch 46000, loss[loss=2.369, over 1820.00 frames. , ppl: 10.684989177951733] tot_loss[loss=2.377, over 5676392.72 frames. , ppl: 10.770420880514878], batch size: 70
+2022-12-10 03:05:58,453 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:05:59,200 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.363, over 211138.00 frames. , ppl: 10.624262598332244
+2022-12-10 03:07:41,815 INFO [train.py:421] (2/8) Epoch 1, batch 46200, loss[loss=2.415, over 2940.00 frames. , ppl: 11.190359863932303] tot_loss[loss=2.377, over 5678817.96 frames. , ppl: 10.771884644038824], batch size: 70
+2022-12-10 03:09:19,254 INFO [train.py:421] (2/8) Epoch 1, batch 46400, loss[loss=2.561, over 1260.00 frames. , ppl: 12.945376531167144] tot_loss[loss=2.376, over 5661387.10 frames. , ppl: 10.766118737484526], batch size: 70
+2022-12-10 03:11:00,973 INFO [train.py:421] (2/8) Epoch 1, batch 46600, loss[loss=2.499, over 1680.00 frames. , ppl: 12.171734792264973] tot_loss[loss=2.376, over 5654666.42 frames. , ppl: 10.760247541647528], batch size: 70
+2022-12-10 03:12:37,298 INFO [train.py:421] (2/8) Epoch 1, batch 46800, loss[loss=2.373, over 2590.00 frames. , ppl: 10.727102821951604] tot_loss[loss=2.377, over 5617662.16 frames. , ppl: 10.774816332097068], batch size: 70
+2022-12-10 03:14:18,442 INFO [train.py:421] (2/8) Epoch 1, batch 47000, loss[loss=2.298, over 6370.00 frames. , ppl: 9.95902829391357] tot_loss[loss=2.377, over 5590114.03 frames. , ppl: 10.776765004488462], batch size: 70
+2022-12-10 03:14:18,443 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:14:19,232 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.363, over 211138.00 frames. , ppl: 10.618849962774785
+2022-12-10 03:15:58,272 INFO [train.py:421] (2/8) Epoch 1, batch 47200, loss[loss=2.549, over 1680.00 frames. , ppl: 12.795064073615341] tot_loss[loss=2.379, over 5520903.66 frames. , ppl: 10.790813068432449], batch size: 70
+2022-12-10 03:17:35,093 INFO [train.py:421] (2/8) Epoch 1, batch 47400, loss[loss=2.464, over 1610.00 frames. , ppl: 11.75432332637622] tot_loss[loss=2.379, over 5522563.21 frames. , ppl: 10.795536787814683], batch size: 70
+2022-12-10 03:19:18,529 INFO [train.py:421] (2/8) Epoch 1, batch 47600, loss[loss=2.624, over 700.00 frames. , ppl: 13.788156359352755] tot_loss[loss=2.379, over 5542123.18 frames. , ppl: 10.795952421139797], batch size: 70
+2022-12-10 03:20:57,292 INFO [train.py:421] (2/8) Epoch 1, batch 47800, loss[loss=2.573, over 1470.00 frames. , ppl: 13.107090372573987] tot_loss[loss=2.379, over 5543859.15 frames. , ppl: 10.794610731041361], batch size: 70
+2022-12-10 03:22:37,620 INFO [train.py:421] (2/8) Epoch 1, batch 48000, loss[loss=4.323, over 350.00 frames. , ppl: 75.40710014365162] tot_loss[loss=2.379, over 5549941.50 frames. , ppl: 10.798835048475953], batch size: 70
+2022-12-10 03:22:37,621 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:22:38,366 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.361, over 211138.00 frames. , ppl: 10.600952749654093
+2022-12-10 03:24:15,507 INFO [train.py:421] (2/8) Epoch 1, batch 48200, loss[loss=2.555, over 1260.00 frames. , ppl: 12.865298560490924] tot_loss[loss=2.379, over 5542925.49 frames. , ppl: 10.799149724713896], batch size: 70
+2022-12-10 03:25:55,536 INFO [train.py:421] (2/8) Epoch 1, batch 48400, loss[loss=3.622, over 420.00 frames. , ppl: 37.39724896828318] tot_loss[loss=2.38, over 5531305.05 frames. , ppl: 10.799850033321837], batch size: 70
+2022-12-10 03:27:34,538 INFO [train.py:421] (2/8) Epoch 1, batch 48600, loss[loss=2.383, over 4620.00 frames. , ppl: 10.834821403040147] tot_loss[loss=2.379, over 5506996.11 frames. , ppl: 10.7976053013021], batch size: 70
+2022-12-10 03:29:11,885 INFO [train.py:421] (2/8) Epoch 1, batch 48800, loss[loss=2.342, over 3500.00 frames. , ppl: 10.406632667682475] tot_loss[loss=2.379, over 5517488.64 frames. , ppl: 10.789117075697346], batch size: 70
+2022-12-10 03:30:50,523 INFO [train.py:421] (2/8) Epoch 1, batch 49000, loss[loss=5.1, over 280.00 frames. , ppl: 163.9799375467727] tot_loss[loss=2.379, over 5492637.41 frames. , ppl: 10.79531495935104], batch size: 70
+2022-12-10 03:30:50,524 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:30:51,283 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.36, over 211138.00 frames. , ppl: 10.585974491786418
+2022-12-10 03:32:29,119 INFO [train.py:421] (2/8) Epoch 1, batch 49200, loss[loss=2.352, over 5530.00 frames. , ppl: 10.50928138859311] tot_loss[loss=2.38, over 5467499.10 frames. , ppl: 10.800762588626212], batch size: 70
+2022-12-10 03:34:05,518 INFO [train.py:421] (2/8) Epoch 1, batch 49400, loss[loss=2.726, over 770.00 frames. , ppl: 15.277054064796294] tot_loss[loss=2.38, over 5451626.39 frames. , ppl: 10.80761251519937], batch size: 70
+2022-12-10 03:35:47,724 INFO [train.py:421] (2/8) Epoch 1, batch 49600, loss[loss=2.434, over 1820.00 frames. , ppl: 11.409751262356233] tot_loss[loss=2.38, over 5449874.85 frames. , ppl: 10.809789107391309], batch size: 70
+2022-12-10 03:37:28,143 INFO [train.py:421] (2/8) Epoch 1, batch 49800, loss[loss=2.49, over 2660.00 frames. , ppl: 12.061618855713863] tot_loss[loss=2.38, over 5442356.95 frames. , ppl: 10.807909695281646], batch size: 70
+2022-12-10 03:39:06,578 INFO [train.py:421] (2/8) Epoch 1, batch 50000, loss[loss=2.488, over 910.00 frames. , ppl: 12.039617160620542] tot_loss[loss=2.379, over 5460953.77 frames. , ppl: 10.793933132514132], batch size: 70
+2022-12-10 03:39:06,578 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:39:07,338 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.359, over 211138.00 frames. , ppl: 10.58257820728752
+2022-12-10 03:40:43,902 INFO [train.py:421] (2/8) Epoch 1, batch 50200, loss[loss=2.404, over 2870.00 frames. , ppl: 11.071195372155833] tot_loss[loss=2.379, over 5461034.47 frames. , ppl: 10.792579476298373], batch size: 70
+2022-12-10 03:42:22,998 INFO [train.py:421] (2/8) Epoch 1, batch 50400, loss[loss=2.391, over 2310.00 frames. , ppl: 10.92959109999966] tot_loss[loss=2.378, over 5446831.25 frames. , ppl: 10.787576419310243], batch size: 70
+2022-12-10 03:44:03,134 INFO [train.py:421] (2/8) Epoch 1, batch 50600, loss[loss=2.326, over 5390.00 frames. , ppl: 10.233178326636128] tot_loss[loss=2.378, over 5453578.72 frames. , ppl: 10.78525263772274], batch size: 70
+2022-12-10 03:45:44,369 INFO [train.py:421] (2/8) Epoch 1, batch 50800, loss[loss=2.299, over 4410.00 frames. , ppl: 9.959997170258038] tot_loss[loss=2.377, over 5486959.50 frames. , ppl: 10.774259815125383], batch size: 70
+2022-12-10 03:47:26,021 INFO [train.py:421] (2/8) Epoch 1, batch 51000, loss[loss=2.376, over 1750.00 frames. , ppl: 10.759355655716694] tot_loss[loss=2.378, over 5465527.69 frames. , ppl: 10.778544870992295], batch size: 70
+2022-12-10 03:47:26,021 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:47:26,769 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.361, over 211138.00 frames. , ppl: 10.598020656585552
+2022-12-10 03:49:07,094 INFO [train.py:421] (2/8) Epoch 1, batch 51200, loss[loss=2.708, over 1050.00 frames. , ppl: 14.992303506667259] tot_loss[loss=2.377, over 5464288.19 frames. , ppl: 10.773946948356908], batch size: 70
+2022-12-10 03:50:49,015 INFO [train.py:421] (2/8) Epoch 1, batch 51400, loss[loss=2.731, over 840.00 frames. , ppl: 15.353814308095918] tot_loss[loss=2.377, over 5470150.92 frames. , ppl: 10.773794293932301], batch size: 70
+2022-12-10 03:52:29,274 INFO [train.py:421] (2/8) Epoch 1, batch 51600, loss[loss=2.255, over 11690.00 frames. , ppl: 9.538795002298547] tot_loss[loss=2.378, over 5443197.06 frames. , ppl: 10.781575333738045], batch size: 70
+2022-12-10 03:54:10,138 INFO [train.py:421] (2/8) Epoch 1, batch 51800, loss[loss=2.532, over 1050.00 frames. , ppl: 12.579277047893774] tot_loss[loss=2.378, over 5455651.02 frames. , ppl: 10.783214694617897], batch size: 70
+2022-12-10 03:55:47,567 INFO [train.py:421] (2/8) Epoch 1, batch 52000, loss[loss=2.357, over 3780.00 frames. , ppl: 10.557226684506162] tot_loss[loss=2.378, over 5481099.37 frames. , ppl: 10.779227080600094], batch size: 70
+2022-12-10 03:55:47,568 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 03:55:48,316 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.359, over 211138.00 frames. , ppl: 10.580972870980991
+2022-12-10 03:57:25,692 INFO [train.py:421] (2/8) Epoch 1, batch 52200, loss[loss=2.513, over 1120.00 frames. , ppl: 12.344456993471494] tot_loss[loss=2.376, over 5543366.73 frames. , ppl: 10.766075741092063], batch size: 70
+2022-12-10 03:59:05,787 INFO [train.py:421] (2/8) Epoch 1, batch 52400, loss[loss=2.432, over 2240.00 frames. , ppl: 11.380906484681367] tot_loss[loss=2.377, over 5563746.85 frames. , ppl: 10.768376691281032], batch size: 70
+2022-12-10 04:00:46,728 INFO [train.py:421] (2/8) Epoch 1, batch 52600, loss[loss=2.328, over 4270.00 frames. , ppl: 10.254588119777537] tot_loss[loss=2.374, over 5642574.80 frames. , ppl: 10.74033524253337], batch size: 70
+2022-12-10 04:02:26,552 INFO [train.py:421] (2/8) Epoch 1, batch 52800, loss[loss=2.253, over 7000.00 frames. , ppl: 9.514010346299687] tot_loss[loss=2.374, over 5627393.06 frames. , ppl: 10.741430663551666], batch size: 70
+2022-12-10 04:04:06,246 INFO [train.py:421] (2/8) Epoch 1, batch 53000, loss[loss=2.352, over 2730.00 frames. , ppl: 10.508578250908151] tot_loss[loss=2.374, over 5616461.48 frames. , ppl: 10.742679626228718], batch size: 70
+2022-12-10 04:04:06,247 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:04:07,006 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.358, over 211138.00 frames. , ppl: 10.567887116951528
+2022-12-10 04:05:52,856 INFO [train.py:421] (2/8) Epoch 1, batch 53200, loss[loss=2.258, over 8120.00 frames. , ppl: 9.561592484833621] tot_loss[loss=2.374, over 5636925.35 frames. , ppl: 10.736039138426639], batch size: 70
+2022-12-10 04:07:33,548 INFO [train.py:421] (2/8) Epoch 1, batch 53400, loss[loss=2.335, over 3080.00 frames. , ppl: 10.331209973109146] tot_loss[loss=2.372, over 5657637.82 frames. , ppl: 10.722508254184152], batch size: 70
+2022-12-10 04:09:19,575 INFO [train.py:421] (2/8) Epoch 1, batch 53600, loss[loss=2.408, over 3710.00 frames. , ppl: 11.107790051759103] tot_loss[loss=2.372, over 5675409.19 frames. , ppl: 10.71668034107373], batch size: 70
+2022-12-10 04:10:59,120 INFO [train.py:421] (2/8) Epoch 1, batch 53800, loss[loss=2.3, over 2870.00 frames. , ppl: 9.974997017347158] tot_loss[loss=2.373, over 5627852.63 frames. , ppl: 10.725201721852109], batch size: 70
+2022-12-10 04:12:36,584 INFO [train.py:421] (2/8) Epoch 1, batch 54000, loss[loss=2.222, over 8750.00 frames. , ppl: 9.221833545730743] tot_loss[loss=2.373, over 5606476.31 frames. , ppl: 10.73037104740901], batch size: 70
+2022-12-10 04:12:36,585 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:12:37,329 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.356, over 211138.00 frames. , ppl: 10.55320704955568
+2022-12-10 04:14:15,873 INFO [train.py:421] (2/8) Epoch 1, batch 54200, loss[loss=4.133, over 350.00 frames. , ppl: 62.36322292645905] tot_loss[loss=2.373, over 5580898.56 frames. , ppl: 10.73078575559422], batch size: 70
+2022-12-10 04:15:51,797 INFO [train.py:421] (2/8) Epoch 1, batch 54400, loss[loss=2.553, over 1400.00 frames. , ppl: 12.840245305920455] tot_loss[loss=2.374, over 5545802.17 frames. , ppl: 10.735675003098264], batch size: 70
+2022-12-10 04:17:31,784 INFO [train.py:421] (2/8) Epoch 1, batch 54600, loss[loss=2.365, over 2520.00 frames. , ppl: 10.646967023861329] tot_loss[loss=2.374, over 5549769.70 frames. , ppl: 10.738973020419294], batch size: 70
+2022-12-10 04:19:11,953 INFO [train.py:421] (2/8) Epoch 1, batch 54800, loss[loss=2.339, over 3850.00 frames. , ppl: 10.367300351924277] tot_loss[loss=2.374, over 5556737.74 frames. , ppl: 10.736564175424215], batch size: 70
+2022-12-10 04:20:51,927 INFO [train.py:421] (2/8) Epoch 1, batch 55000, loss[loss=2.347, over 2660.00 frames. , ppl: 10.45387813935252] tot_loss[loss=2.374, over 5512279.76 frames. , ppl: 10.742770710414069], batch size: 70
+2022-12-10 04:20:51,927 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:20:52,675 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.355, over 211138.00 frames. , ppl: 10.542236908630983
+2022-12-10 04:22:35,385 INFO [train.py:421] (2/8) Epoch 1, batch 55200, loss[loss=2.517, over 1190.00 frames. , ppl: 12.387886739231194] tot_loss[loss=2.373, over 5535303.58 frames. , ppl: 10.732388171894108], batch size: 70
+2022-12-10 04:24:13,632 INFO [train.py:421] (2/8) Epoch 1, batch 55400, loss[loss=2.549, over 1470.00 frames. , ppl: 12.798779166294867] tot_loss[loss=2.373, over 5566634.50 frames. , ppl: 10.726179649868635], batch size: 70
+2022-12-10 04:25:54,098 INFO [train.py:421] (2/8) Epoch 1, batch 55600, loss[loss=2.336, over 2730.00 frames. , ppl: 10.341874470904605] tot_loss[loss=2.372, over 5586790.04 frames. , ppl: 10.723466738928531], batch size: 70
+2022-12-10 04:27:33,001 INFO [train.py:421] (2/8) Epoch 1, batch 55800, loss[loss=2.331, over 3850.00 frames. , ppl: 10.287868667218438] tot_loss[loss=2.373, over 5554793.35 frames. , ppl: 10.727900783559909], batch size: 70
+2022-12-10 04:29:13,169 INFO [train.py:421] (2/8) Epoch 1, batch 56000, loss[loss=2.415, over 1680.00 frames. , ppl: 11.194655955776607] tot_loss[loss=2.374, over 5485442.85 frames. , ppl: 10.739399266091143], batch size: 70
+2022-12-10 04:29:13,169 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:29:13,929 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.355, over 211138.00 frames. , ppl: 10.539062115723747
+2022-12-10 04:30:54,312 INFO [train.py:421] (2/8) Epoch 1, batch 56200, loss[loss=2.462, over 2100.00 frames. , ppl: 11.722645775573017] tot_loss[loss=2.374, over 5458556.02 frames. , ppl: 10.745520357248772], batch size: 70
+2022-12-10 04:32:32,240 INFO [train.py:421] (2/8) Epoch 1, batch 56400, loss[loss=2.372, over 3640.00 frames. , ppl: 10.720261040032343] tot_loss[loss=2.374, over 5471645.02 frames. , ppl: 10.74267013615018], batch size: 70
+2022-12-10 04:34:13,605 INFO [train.py:421] (2/8) Epoch 1, batch 56600, loss[loss=2.518, over 1050.00 frames. , ppl: 12.402883550294876] tot_loss[loss=2.375, over 5419199.56 frames. , ppl: 10.752419590728696], batch size: 70
+2022-12-10 04:35:52,337 INFO [train.py:421] (2/8) Epoch 1, batch 56800, loss[loss=2.808, over 840.00 frames. , ppl: 16.584373985701866] tot_loss[loss=2.376, over 5397964.77 frames. , ppl: 10.758059474917744], batch size: 70
+2022-12-10 04:37:33,495 INFO [train.py:421] (2/8) Epoch 1, batch 57000, loss[loss=2.619, over 980.00 frames. , ppl: 13.727514881581476] tot_loss[loss=2.375, over 5414151.17 frames. , ppl: 10.75180927987586], batch size: 70
+2022-12-10 04:37:33,496 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:37:34,257 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.354, over 211138.00 frames. , ppl: 10.531887643629647
+2022-12-10 04:39:15,037 INFO [train.py:421] (2/8) Epoch 1, batch 57200, loss[loss=2.387, over 1610.00 frames. , ppl: 10.876208798362537] tot_loss[loss=2.375, over 5417765.43 frames. , ppl: 10.753547569715298], batch size: 70
+2022-12-10 04:40:55,985 INFO [train.py:421] (2/8) Epoch 1, batch 57400, loss[loss=2.655, over 1120.00 frames. , ppl: 14.226212168837664] tot_loss[loss=2.375, over 5437211.50 frames. , ppl: 10.755929281341166], batch size: 70
+2022-12-10 04:42:35,563 INFO [train.py:421] (2/8) Epoch 1, batch 57600, loss[loss=2.429, over 840.00 frames. , ppl: 11.351732867576677] tot_loss[loss=2.375, over 5450373.36 frames. , ppl: 10.74793613700056], batch size: 70
+2022-12-10 04:44:14,430 INFO [train.py:421] (2/8) Epoch 1, batch 57800, loss[loss=2.431, over 2380.00 frames. , ppl: 11.369396538009203] tot_loss[loss=2.374, over 5464518.62 frames. , ppl: 10.741212424485779], batch size: 70
+2022-12-10 04:45:53,663 INFO [train.py:421] (2/8) Epoch 1, batch 58000, loss[loss=2.298, over 6580.00 frames. , ppl: 9.95401889301435] tot_loss[loss=2.373, over 5498298.42 frames. , ppl: 10.7329337303719], batch size: 70
+2022-12-10 04:45:53,664 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:45:54,422 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.353, over 211138.00 frames. , ppl: 10.515346212674748
+2022-12-10 04:47:35,224 INFO [train.py:421] (2/8) Epoch 1, batch 58200, loss[loss=3.225, over 490.00 frames. , ppl: 25.146907657849166] tot_loss[loss=2.373, over 5515481.15 frames. , ppl: 10.728791845624752], batch size: 70
+2022-12-10 04:49:15,814 INFO [train.py:421] (2/8) Epoch 1, batch 58400, loss[loss=2.458, over 1680.00 frames. , ppl: 11.685978662081359] tot_loss[loss=2.372, over 5524313.21 frames. , ppl: 10.719014984782884], batch size: 70
+2022-12-10 04:50:58,482 INFO [train.py:421] (2/8) Epoch 1, batch 58600, loss[loss=2.305, over 7770.00 frames. , ppl: 10.0244675241299] tot_loss[loss=2.372, over 5549822.89 frames. , ppl: 10.720219056488645], batch size: 70
+2022-12-10 04:52:37,253 INFO [train.py:421] (2/8) Epoch 1, batch 58800, loss[loss=2.328, over 1750.00 frames. , ppl: 10.257174372823831] tot_loss[loss=2.374, over 5494722.44 frames. , ppl: 10.736739360446363], batch size: 70
+2022-12-10 04:54:20,777 INFO [train.py:421] (2/8) Epoch 1, batch 59000, loss[loss=2.29, over 5040.00 frames. , ppl: 9.873767902819939] tot_loss[loss=2.374, over 5461060.56 frames. , ppl: 10.737380527743378], batch size: 70
+2022-12-10 04:54:20,777 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 04:54:21,536 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.352, over 211138.00 frames. , ppl: 10.503966120477648
+2022-12-10 04:55:59,455 INFO [train.py:421] (2/8) Epoch 1, batch 59200, loss[loss=2.395, over 2940.00 frames. , ppl: 10.971108666759687] tot_loss[loss=2.374, over 5463222.98 frames. , ppl: 10.735230611172154], batch size: 70
+2022-12-10 04:57:42,858 INFO [train.py:421] (2/8) Epoch 1, batch 59400, loss[loss=2.41, over 1610.00 frames. , ppl: 11.130335823304133] tot_loss[loss=2.371, over 5528000.99 frames. , ppl: 10.713179572157152], batch size: 70
+2022-12-10 04:59:24,822 INFO [train.py:421] (2/8) Epoch 1, batch 59600, loss[loss=3.283, over 560.00 frames. , ppl: 26.662981673273737] tot_loss[loss=2.37, over 5579370.81 frames. , ppl: 10.693891036246509], batch size: 70
+2022-12-10 05:01:09,076 INFO [train.py:421] (2/8) Epoch 1, batch 59800, loss[loss=2.494, over 1190.00 frames. , ppl: 12.10589289284915] tot_loss[loss=2.368, over 5599248.46 frames. , ppl: 10.67987381689782], batch size: 70
+2022-12-10 05:02:50,323 INFO [train.py:421] (2/8) Epoch 1, batch 60000, loss[loss=2.309, over 3570.00 frames. , ppl: 10.063779559655163] tot_loss[loss=2.368, over 5566481.50 frames. , ppl: 10.680535724244699], batch size: 70
+2022-12-10 05:02:50,324 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:02:51,100 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.353, over 211138.00 frames. , ppl: 10.521841308858221
+2022-12-10 05:04:30,566 INFO [train.py:421] (2/8) Epoch 1, batch 60200, loss[loss=2.291, over 3360.00 frames. , ppl: 9.889089881456805] tot_loss[loss=2.369, over 5569601.66 frames. , ppl: 10.68185744930842], batch size: 70
+2022-12-10 05:06:11,840 INFO [train.py:421] (2/8) Epoch 1, batch 60400, loss[loss=2.265, over 3430.00 frames. , ppl: 9.630343683345899] tot_loss[loss=2.368, over 5583871.44 frames. , ppl: 10.676672414192216], batch size: 70
+2022-12-10 05:07:53,808 INFO [train.py:421] (2/8) Epoch 1, batch 60600, loss[loss=2.365, over 3290.00 frames. , ppl: 10.643141889288973] tot_loss[loss=2.369, over 5555024.58 frames. , ppl: 10.681965119055485], batch size: 70
+2022-12-10 05:09:33,763 INFO [train.py:421] (2/8) Epoch 1, batch 60800, loss[loss=2.312, over 5250.00 frames. , ppl: 10.093042793501787] tot_loss[loss=2.368, over 5559765.17 frames. , ppl: 10.677131039338066], batch size: 70
+2022-12-10 05:11:18,589 INFO [train.py:421] (2/8) Epoch 1, batch 61000, loss[loss=2.525, over 1540.00 frames. , ppl: 12.49212701963166] tot_loss[loss=2.368, over 5574246.22 frames. , ppl: 10.678312831483515], batch size: 70
+2022-12-10 05:11:18,590 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:11:19,337 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.352, over 211138.00 frames. , ppl: 10.508142781873898
+2022-12-10 05:12:58,120 INFO [train.py:421] (2/8) Epoch 1, batch 61200, loss[loss=2.817, over 700.00 frames. , ppl: 16.72821507445118] tot_loss[loss=2.369, over 5559330.59 frames. , ppl: 10.683348587788537], batch size: 70
+2022-12-10 05:14:39,423 INFO [train.py:421] (2/8) Epoch 1, batch 61400, loss[loss=2.216, over 8890.00 frames. , ppl: 9.169092595983736] tot_loss[loss=2.37, over 5529179.26 frames. , ppl: 10.692760964557577], batch size: 70
+2022-12-10 05:16:19,246 INFO [train.py:421] (2/8) Epoch 1, batch 61600, loss[loss=2.389, over 2100.00 frames. , ppl: 10.902022638586464] tot_loss[loss=2.369, over 5534193.14 frames. , ppl: 10.68992964397853], batch size: 70
+2022-12-10 05:17:59,811 INFO [train.py:421] (2/8) Epoch 1, batch 61800, loss[loss=2.547, over 980.00 frames. , ppl: 12.7636903990452] tot_loss[loss=2.37, over 5532401.80 frames. , ppl: 10.695651882181142], batch size: 70
+2022-12-10 05:19:41,041 INFO [train.py:421] (2/8) Epoch 1, batch 62000, loss[loss=2.393, over 3220.00 frames. , ppl: 10.943088618561578] tot_loss[loss=2.37, over 5515266.93 frames. , ppl: 10.697828567228086], batch size: 70
+2022-12-10 05:19:41,042 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:19:41,789 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.351, over 211138.00 frames. , ppl: 10.500484247218756
+2022-12-10 05:21:23,028 INFO [train.py:421] (2/8) Epoch 1, batch 62200, loss[loss=2.694, over 700.00 frames. , ppl: 14.793048375575648] tot_loss[loss=2.369, over 5538582.55 frames. , ppl: 10.690673921342698], batch size: 70
+2022-12-10 05:23:03,309 INFO [train.py:421] (2/8) Epoch 1, batch 62400, loss[loss=2.475, over 1470.00 frames. , ppl: 11.878589647522698] tot_loss[loss=2.37, over 5514116.08 frames. , ppl: 10.692599950338824], batch size: 70
+2022-12-10 05:24:44,527 INFO [train.py:421] (2/8) Epoch 1, batch 62600, loss[loss=2.469, over 1680.00 frames. , ppl: 11.806082334232535] tot_loss[loss=2.371, over 5488570.98 frames. , ppl: 10.704278298661377], batch size: 70
+2022-12-10 05:26:22,732 INFO [train.py:421] (2/8) Epoch 1, batch 62800, loss[loss=2.344, over 2030.00 frames. , ppl: 10.427743663167307] tot_loss[loss=2.37, over 5502011.78 frames. , ppl: 10.697741369231883], batch size: 70
+2022-12-10 05:28:05,200 INFO [train.py:421] (2/8) Epoch 1, batch 63000, loss[loss=2.376, over 2380.00 frames. , ppl: 10.763237391401146] tot_loss[loss=2.371, over 5488066.09 frames. , ppl: 10.706285007205596], batch size: 70
+2022-12-10 05:28:05,201 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:28:05,959 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.352, over 211138.00 frames. , ppl: 10.50832786224836
+2022-12-10 05:29:47,018 INFO [train.py:421] (2/8) Epoch 1, batch 63200, loss[loss=2.352, over 3220.00 frames. , ppl: 10.506665217699704] tot_loss[loss=2.37, over 5521491.38 frames. , ppl: 10.69984574528347], batch size: 70
+2022-12-10 05:31:24,706 INFO [train.py:421] (2/8) Epoch 1, batch 63400, loss[loss=2.329, over 2730.00 frames. , ppl: 10.262986553809414] tot_loss[loss=2.369, over 5541725.48 frames. , ppl: 10.690956675080495], batch size: 70
+2022-12-10 05:33:03,873 INFO [train.py:421] (2/8) Epoch 1, batch 63600, loss[loss=2.649, over 1050.00 frames. , ppl: 14.145524035767266] tot_loss[loss=2.37, over 5515252.62 frames. , ppl: 10.695959447937613], batch size: 70
+2022-12-10 05:34:41,284 INFO [train.py:421] (2/8) Epoch 1, batch 63800, loss[loss=2.388, over 1750.00 frames. , ppl: 10.886298958805382] tot_loss[loss=2.37, over 5489464.00 frames. , ppl: 10.69886552641615], batch size: 70
+2022-12-10 05:36:22,761 INFO [train.py:421] (2/8) Epoch 1, batch 64000, loss[loss=2.365, over 2590.00 frames. , ppl: 10.647999122391553] tot_loss[loss=2.37, over 5506507.94 frames. , ppl: 10.699229131982491], batch size: 70
+2022-12-10 05:36:22,762 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:36:23,521 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.351, over 211138.00 frames. , ppl: 10.500185854649434
+2022-12-10 05:38:06,566 INFO [train.py:421] (2/8) Epoch 1, batch 64200, loss[loss=2.388, over 2310.00 frames. , ppl: 10.896873664987178] tot_loss[loss=2.37, over 5469762.32 frames. , ppl: 10.700856268519317], batch size: 70
+2022-12-10 05:39:47,998 INFO [train.py:421] (2/8) Epoch 1, batch 64400, loss[loss=2.64, over 980.00 frames. , ppl: 14.018140070544849] tot_loss[loss=2.371, over 5462210.32 frames. , ppl: 10.706204189216486], batch size: 70
+2022-12-10 05:41:29,580 INFO [train.py:421] (2/8) Epoch 1, batch 64600, loss[loss=2.34, over 5880.00 frames. , ppl: 10.382604238388504] tot_loss[loss=2.371, over 5437806.63 frames. , ppl: 10.709315974973256], batch size: 70
+2022-12-10 05:43:06,385 INFO [train.py:421] (2/8) Epoch 1, batch 64800, loss[loss=2.554, over 980.00 frames. , ppl: 12.860227757757107] tot_loss[loss=2.372, over 5396743.28 frames. , ppl: 10.714002154853059], batch size: 70
+2022-12-10 05:44:45,913 INFO [train.py:421] (2/8) Epoch 1, batch 65000, loss[loss=2.254, over 5810.00 frames. , ppl: 9.521495539381139] tot_loss[loss=2.372, over 5366878.03 frames. , ppl: 10.721589675828598], batch size: 70
+2022-12-10 05:44:45,914 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:44:46,659 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.351, over 211138.00 frames. , ppl: 10.499536258484806
+2022-12-10 05:46:27,197 INFO [train.py:421] (2/8) Epoch 1, batch 65200, loss[loss=2.421, over 2310.00 frames. , ppl: 11.252028200787624] tot_loss[loss=2.372, over 5366653.34 frames. , ppl: 10.717538970701845], batch size: 70
+2022-12-10 05:48:06,797 INFO [train.py:421] (2/8) Epoch 1, batch 65400, loss[loss=2.472, over 3150.00 frames. , ppl: 11.850235967569402] tot_loss[loss=2.372, over 5399062.62 frames. , ppl: 10.714308188096606], batch size: 70
+2022-12-10 05:49:47,266 INFO [train.py:421] (2/8) Epoch 1, batch 65600, loss[loss=2.459, over 910.00 frames. , ppl: 11.688338130086374] tot_loss[loss=2.37, over 5432791.69 frames. , ppl: 10.700323987714974], batch size: 70
+2022-12-10 05:51:25,736 INFO [train.py:421] (2/8) Epoch 1, batch 65800, loss[loss=2.376, over 9100.00 frames. , ppl: 10.76096764765068] tot_loss[loss=2.371, over 5412555.45 frames. , ppl: 10.705983490125414], batch size: 70
+2022-12-10 05:53:05,387 INFO [train.py:421] (2/8) Epoch 1, batch 66000, loss[loss=2.361, over 3290.00 frames. , ppl: 10.605615571592617] tot_loss[loss=2.372, over 5397048.22 frames. , ppl: 10.714830606009599], batch size: 70
+2022-12-10 05:53:05,388 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 05:53:06,133 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.349, over 211138.00 frames. , ppl: 10.479744419827652
+2022-12-10 05:54:45,805 INFO [train.py:421] (2/8) Epoch 1, batch 66200, loss[loss=2.366, over 3850.00 frames. , ppl: 10.654865471456544] tot_loss[loss=2.371, over 5438432.84 frames. , ppl: 10.710181216197874], batch size: 70
+2022-12-10 05:56:26,060 INFO [train.py:421] (2/8) Epoch 1, batch 66400, loss[loss=2.399, over 1540.00 frames. , ppl: 11.0163176896375] tot_loss[loss=2.372, over 5410932.14 frames. , ppl: 10.714596734620914], batch size: 70
+2022-12-10 05:58:06,910 INFO [train.py:421] (2/8) Epoch 1, batch 66600, loss[loss=2.678, over 980.00 frames. , ppl: 14.558765611225557] tot_loss[loss=2.37, over 5447734.35 frames. , ppl: 10.695225000442345], batch size: 70
+2022-12-10 05:59:44,019 INFO [train.py:421] (2/8) Epoch 1, batch 66800, loss[loss=2.335, over 3570.00 frames. , ppl: 10.328793676278657] tot_loss[loss=2.369, over 5454882.34 frames. , ppl: 10.69127730591622], batch size: 70
+2022-12-10 06:01:21,583 INFO [train.py:421] (2/8) Epoch 1, batch 67000, loss[loss=2.581, over 840.00 frames. , ppl: 13.215354012737743] tot_loss[loss=2.369, over 5449744.22 frames. , ppl: 10.691334281200449], batch size: 70
+2022-12-10 06:01:21,584 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:01:22,329 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.348, over 211138.00 frames. , ppl: 10.464386029268178
+2022-12-10 06:02:59,567 INFO [train.py:421] (2/8) Epoch 1, batch 67200, loss[loss=2.419, over 2310.00 frames. , ppl: 11.233059644345282] tot_loss[loss=2.369, over 5451404.69 frames. , ppl: 10.692004447427042], batch size: 70
+2022-12-10 06:04:38,154 INFO [train.py:421] (2/8) Epoch 1, batch 67400, loss[loss=2.544, over 1540.00 frames. , ppl: 12.726282036188111] tot_loss[loss=2.369, over 5476146.63 frames. , ppl: 10.68418520799201], batch size: 70
+2022-12-10 06:06:18,494 INFO [train.py:421] (2/8) Epoch 1, batch 67600, loss[loss=2.343, over 2520.00 frames. , ppl: 10.41259796186093] tot_loss[loss=2.368, over 5500527.52 frames. , ppl: 10.680000162252984], batch size: 70
+2022-12-10 06:08:00,629 INFO [train.py:421] (2/8) Epoch 1, batch 67800, loss[loss=2.316, over 5670.00 frames. , ppl: 10.131328006140237] tot_loss[loss=2.368, over 5504176.28 frames. , ppl: 10.67903190955711], batch size: 70
+2022-12-10 06:09:41,543 INFO [train.py:421] (2/8) Epoch 1, batch 68000, loss[loss=2.425, over 2030.00 frames. , ppl: 11.301511743538622] tot_loss[loss=2.368, over 5490593.23 frames. , ppl: 10.67804777554391], batch size: 70
+2022-12-10 06:09:41,544 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:09:42,303 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.35, over 211138.00 frames. , ppl: 10.481405758605055
+2022-12-10 06:11:26,714 INFO [train.py:421] (2/8) Epoch 1, batch 68200, loss[loss=2.34, over 3780.00 frames. , ppl: 10.37794949856016] tot_loss[loss=2.367, over 5475239.33 frames. , ppl: 10.668991790966714], batch size: 70
+2022-12-10 06:13:04,726 INFO [train.py:421] (2/8) Epoch 1, batch 68400, loss[loss=2.385, over 3220.00 frames. , ppl: 10.859899864107746] tot_loss[loss=2.368, over 5431433.00 frames. , ppl: 10.68000128703256], batch size: 70
+2022-12-10 06:14:45,539 INFO [train.py:421] (2/8) Epoch 1, batch 68600, loss[loss=2.739, over 840.00 frames. , ppl: 15.477753201226092] tot_loss[loss=2.369, over 5422089.07 frames. , ppl: 10.683063180004703], batch size: 70
+2022-12-10 06:16:28,483 INFO [train.py:421] (2/8) Epoch 1, batch 68800, loss[loss=2.478, over 1680.00 frames. , ppl: 11.921632517729718] tot_loss[loss=2.369, over 5422257.49 frames. , ppl: 10.683131467928993], batch size: 70
+2022-12-10 06:18:08,236 INFO [train.py:421] (2/8) Epoch 1, batch 69000, loss[loss=2.403, over 1680.00 frames. , ppl: 11.055368781749882] tot_loss[loss=2.368, over 5418065.39 frames. , ppl: 10.675134612546385], batch size: 70
+2022-12-10 06:18:08,237 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:18:08,996 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.348, over 211138.00 frames. , ppl: 10.467406639264686
+2022-12-10 06:19:44,538 INFO [train.py:421] (2/8) Epoch 1, batch 69200, loss[loss=2.361, over 2100.00 frames. , ppl: 10.598065696205383] tot_loss[loss=2.369, over 5376999.29 frames. , ppl: 10.682905273314525], batch size: 70
+2022-12-10 06:21:26,218 INFO [train.py:421] (2/8) Epoch 1, batch 69400, loss[loss=2.297, over 4480.00 frames. , ppl: 9.941674736873118] tot_loss[loss=2.369, over 5367336.43 frames. , ppl: 10.68214053844763], batch size: 70
+2022-12-10 06:23:06,573 INFO [train.py:421] (2/8) Epoch 1, batch 69600, loss[loss=2.333, over 5390.00 frames. , ppl: 10.30792539453986] tot_loss[loss=2.37, over 5345612.54 frames. , ppl: 10.695568056887002], batch size: 70
+2022-12-10 06:24:43,354 INFO [train.py:421] (2/8) Epoch 1, batch 69800, loss[loss=2.49, over 1400.00 frames. , ppl: 12.05964405483825] tot_loss[loss=2.37, over 5333250.89 frames. , ppl: 10.699157527761], batch size: 70
+2022-12-10 06:26:25,253 INFO [train.py:421] (2/8) Epoch 1, batch 70000, loss[loss=2.465, over 1960.00 frames. , ppl: 11.767535221488481] tot_loss[loss=2.37, over 5332768.19 frames. , ppl: 10.692942946319931], batch size: 70
+2022-12-10 06:26:25,254 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:26:26,013 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.348, over 211138.00 frames. , ppl: 10.463410326196138
+2022-12-10 06:28:04,613 INFO [train.py:421] (2/8) Epoch 1, batch 70200, loss[loss=2.402, over 1750.00 frames. , ppl: 11.049923988895797] tot_loss[loss=2.37, over 5314440.15 frames. , ppl: 10.698270359946076], batch size: 70
+2022-12-10 06:29:39,960 INFO [train.py:421] (2/8) Epoch 1, batch 70400, loss[loss=2.523, over 1260.00 frames. , ppl: 12.467498710508735] tot_loss[loss=2.37, over 5332381.18 frames. , ppl: 10.692172057380025], batch size: 70
+2022-12-10 06:31:16,084 INFO [train.py:421] (2/8) Epoch 1, batch 70600, loss[loss=2.371, over 3080.00 frames. , ppl: 10.712053440274632] tot_loss[loss=2.371, over 5286089.58 frames. , ppl: 10.707675967544414], batch size: 70
+2022-12-10 06:32:53,012 INFO [train.py:421] (2/8) Epoch 1, batch 70800, loss[loss=2.343, over 8680.00 frames. , ppl: 10.415443124170718] tot_loss[loss=2.37, over 5311797.77 frames. , ppl: 10.700088949079882], batch size: 70
+2022-12-10 06:34:35,543 INFO [train.py:421] (2/8) Epoch 1, batch 71000, loss[loss=2.38, over 2310.00 frames. , ppl: 10.804175691543726] tot_loss[loss=2.369, over 5365866.99 frames. , ppl: 10.689585434606443], batch size: 70
+2022-12-10 06:34:35,544 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:34:36,305 INFO [train.py:452] (2/8) Epoch 1, validation: loss=2.347, over 211138.00 frames. , ppl: 10.45096802929164
+2022-12-10 06:36:12,870 INFO [train.py:421] (2/8) Epoch 1, batch 71200, loss[loss=2.378, over 1890.00 frames. , ppl: 10.788276980071954] tot_loss[loss=2.368, over 5384291.51 frames. , ppl: 10.677895671220748], batch size: 70
+2022-12-10 06:37:52,675 INFO [train.py:421] (2/8) Epoch 1, batch 71400, loss[loss=2.354, over 2870.00 frames. , ppl: 10.522968769657243] tot_loss[loss=2.368, over 5372301.44 frames. , ppl: 10.677533632772343], batch size: 70
+2022-12-10 06:39:34,126 INFO [train.py:421] (2/8) Epoch 1, batch 71600, loss[loss=2.359, over 1680.00 frames. , ppl: 10.575392165023692] tot_loss[loss=2.37, over 5355571.21 frames. , ppl: 10.692460718309366], batch size: 70
+2022-12-10 06:41:16,766 INFO [train.py:421] (2/8) Epoch 1, batch 71800, loss[loss=2.421, over 2240.00 frames. , ppl: 11.253779152767459] tot_loss[loss=2.368, over 5396112.14 frames. , ppl: 10.680100952633547], batch size: 70
+2022-12-10 06:42:32,557 INFO [train.py:421] (2/8) Epoch 2, batch 0, loss[loss=2.274, over 2240.00 frames. , ppl: 9.717644171083355] tot_loss[loss=2.274, over 2240.00 frames. , ppl: 9.717644171083355], batch size: 70
+2022-12-10 06:44:11,579 INFO [train.py:421] (2/8) Epoch 2, batch 200, loss[loss=2.527, over 910.00 frames. , ppl: 12.510065059548339] tot_loss[loss=2.362, over 501229.23 frames. , ppl: 10.608554657441443], batch size: 70
+2022-12-10 06:45:50,262 INFO [train.py:421] (2/8) Epoch 2, batch 400, loss[loss=2.681, over 840.00 frames. , ppl: 14.600157377583098] tot_loss[loss=2.362, over 954440.65 frames. , ppl: 10.614775495287558], batch size: 70
+2022-12-10 06:47:30,605 INFO [train.py:421] (2/8) Epoch 2, batch 600, loss[loss=2.308, over 7630.00 frames. , ppl: 10.058588545954624] tot_loss[loss=2.361, over 1392752.68 frames. , ppl: 10.600790762166838], batch size: 70
+2022-12-10 06:49:11,810 INFO [train.py:421] (2/8) Epoch 2, batch 800, loss[loss=2.524, over 770.00 frames. , ppl: 12.473764278694745] tot_loss[loss=2.358, over 1793100.67 frames. , ppl: 10.568411766287715], batch size: 70
+2022-12-10 06:50:52,992 INFO [train.py:421] (2/8) Epoch 2, batch 1000, loss[loss=2.332, over 3080.00 frames. , ppl: 10.298967171170625] tot_loss[loss=2.36, over 2127106.39 frames. , ppl: 10.589573259562528], batch size: 70
+2022-12-10 06:50:52,993 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:50:53,753 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.346, over 211138.00 frames. , ppl: 10.441904495381296
+2022-12-10 06:52:33,130 INFO [train.py:421] (2/8) Epoch 2, batch 1200, loss[loss=2.37, over 2170.00 frames. , ppl: 10.694847835199733] tot_loss[loss=2.358, over 2446348.62 frames. , ppl: 10.57380191703776], batch size: 70
+2022-12-10 06:54:14,266 INFO [train.py:421] (2/8) Epoch 2, batch 1400, loss[loss=2.665, over 770.00 frames. , ppl: 14.36465714669262] tot_loss[loss=2.357, over 2744296.36 frames. , ppl: 10.555695565493998], batch size: 70
+2022-12-10 06:55:55,764 INFO [train.py:421] (2/8) Epoch 2, batch 1600, loss[loss=2.289, over 11130.00 frames. , ppl: 9.86942449393351] tot_loss[loss=2.357, over 3001106.17 frames. , ppl: 10.554437139597859], batch size: 70
+2022-12-10 06:57:35,263 INFO [train.py:421] (2/8) Epoch 2, batch 1800, loss[loss=2.382, over 1750.00 frames. , ppl: 10.82490620704712] tot_loss[loss=2.359, over 3219192.81 frames. , ppl: 10.584725911976058], batch size: 70
+2022-12-10 06:59:14,992 INFO [train.py:421] (2/8) Epoch 2, batch 2000, loss[loss=2.468, over 980.00 frames. , ppl: 11.800323462357102] tot_loss[loss=2.356, over 3478031.18 frames. , ppl: 10.553707698960748], batch size: 70
+2022-12-10 06:59:14,993 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 06:59:15,754 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.345, over 211138.00 frames. , ppl: 10.430619322580018
+2022-12-10 07:00:58,120 INFO [train.py:421] (2/8) Epoch 2, batch 2200, loss[loss=2.51, over 1470.00 frames. , ppl: 12.303961010222448] tot_loss[loss=2.355, over 3719655.77 frames. , ppl: 10.537118467861832], batch size: 70
+2022-12-10 07:02:37,033 INFO [train.py:421] (2/8) Epoch 2, batch 2400, loss[loss=2.295, over 10920.00 frames. , ppl: 9.925883147357737] tot_loss[loss=2.355, over 3912697.47 frames. , ppl: 10.54050654724156], batch size: 70
+2022-12-10 07:04:16,524 INFO [train.py:421] (2/8) Epoch 2, batch 2600, loss[loss=2.545, over 1190.00 frames. , ppl: 12.741746328552358] tot_loss[loss=2.358, over 4006651.07 frames. , ppl: 10.568317810105835], batch size: 70
+2022-12-10 07:05:57,393 INFO [train.py:421] (2/8) Epoch 2, batch 2800, loss[loss=2.365, over 4760.00 frames. , ppl: 10.647503655380204] tot_loss[loss=2.358, over 4132543.04 frames. , ppl: 10.568362642274682], batch size: 70
+2022-12-10 07:07:34,871 INFO [train.py:421] (2/8) Epoch 2, batch 3000, loss[loss=2.715, over 770.00 frames. , ppl: 15.10154149345072] tot_loss[loss=2.358, over 4266919.83 frames. , ppl: 10.568853049205636], batch size: 70
+2022-12-10 07:07:34,871 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:07:35,617 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.346, over 211138.00 frames. , ppl: 10.439906381190209
+2022-12-10 07:09:13,504 INFO [train.py:421] (2/8) Epoch 2, batch 3200, loss[loss=2.404, over 2380.00 frames. , ppl: 11.067119071643482] tot_loss[loss=2.357, over 4410005.68 frames. , ppl: 10.562806082375925], batch size: 70
+2022-12-10 07:10:54,074 INFO [train.py:421] (2/8) Epoch 2, batch 3400, loss[loss=2.426, over 1610.00 frames. , ppl: 11.313846866338833] tot_loss[loss=2.356, over 4553468.25 frames. , ppl: 10.54377722171455], batch size: 70
+2022-12-10 07:12:38,066 INFO [train.py:421] (2/8) Epoch 2, batch 3600, loss[loss=2.233, over 7840.00 frames. , ppl: 9.326816207507589] tot_loss[loss=2.356, over 4652817.69 frames. , ppl: 10.551519952393617], batch size: 70
+2022-12-10 07:14:17,881 INFO [train.py:421] (2/8) Epoch 2, batch 3800, loss[loss=2.358, over 2380.00 frames. , ppl: 10.567089284746706] tot_loss[loss=2.357, over 4714686.02 frames. , ppl: 10.554585943691649], batch size: 70
+2022-12-10 07:15:59,649 INFO [train.py:421] (2/8) Epoch 2, batch 4000, loss[loss=2.538, over 770.00 frames. , ppl: 12.656684519779374] tot_loss[loss=2.355, over 4797805.54 frames. , ppl: 10.541684648538936], batch size: 70
+2022-12-10 07:15:59,649 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:16:00,395 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.345, over 211138.00 frames. , ppl: 10.432709850132296
+2022-12-10 07:17:44,062 INFO [train.py:421] (2/8) Epoch 2, batch 4200, loss[loss=2.654, over 980.00 frames. , ppl: 14.210076722408406] tot_loss[loss=2.354, over 4888968.75 frames. , ppl: 10.528755554888395], batch size: 70
+2022-12-10 07:19:26,573 INFO [train.py:421] (2/8) Epoch 2, batch 4400, loss[loss=2.303, over 5600.00 frames. , ppl: 10.00431182974334] tot_loss[loss=2.353, over 4971419.59 frames. , ppl: 10.522177362097754], batch size: 70
+2022-12-10 07:21:09,714 INFO [train.py:421] (2/8) Epoch 2, batch 4600, loss[loss=2.587, over 1050.00 frames. , ppl: 13.292099402954777] tot_loss[loss=2.355, over 4998153.43 frames. , ppl: 10.537014538009903], batch size: 70
+2022-12-10 07:22:49,446 INFO [train.py:421] (2/8) Epoch 2, batch 4800, loss[loss=2.284, over 2660.00 frames. , ppl: 9.811759690919079] tot_loss[loss=2.354, over 5062080.72 frames. , ppl: 10.530760507641848], batch size: 70
+2022-12-10 07:24:27,421 INFO [train.py:421] (2/8) Epoch 2, batch 5000, loss[loss=2.354, over 2870.00 frames. , ppl: 10.523529149278653] tot_loss[loss=2.356, over 5067216.20 frames. , ppl: 10.546583655885591], batch size: 70
+2022-12-10 07:24:27,422 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:24:28,168 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.417051893133054
+2022-12-10 07:26:10,665 INFO [train.py:421] (2/8) Epoch 2, batch 5200, loss[loss=2.368, over 3850.00 frames. , ppl: 10.680747553243773] tot_loss[loss=2.356, over 5088108.20 frames. , ppl: 10.547902602238828], batch size: 70
+2022-12-10 07:27:50,415 INFO [train.py:421] (2/8) Epoch 2, batch 5400, loss[loss=2.582, over 1260.00 frames. , ppl: 13.22185066920533] tot_loss[loss=2.356, over 5122808.41 frames. , ppl: 10.54925968273731], batch size: 70
+2022-12-10 07:29:28,145 INFO [train.py:421] (2/8) Epoch 2, batch 5600, loss[loss=2.389, over 1470.00 frames. , ppl: 10.901197746606009] tot_loss[loss=2.357, over 5131523.24 frames. , ppl: 10.556094105009098], batch size: 70
+2022-12-10 07:31:06,307 INFO [train.py:421] (2/8) Epoch 2, batch 5800, loss[loss=2.377, over 3570.00 frames. , ppl: 10.767318159821496] tot_loss[loss=2.357, over 5160123.72 frames. , ppl: 10.555895975799865], batch size: 70
+2022-12-10 07:32:43,819 INFO [train.py:421] (2/8) Epoch 2, batch 6000, loss[loss=3.635, over 420.00 frames. , ppl: 37.899887705997386] tot_loss[loss=2.356, over 5207433.20 frames. , ppl: 10.551015867669326], batch size: 70
+2022-12-10 07:32:43,820 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:32:44,567 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.413985701421621
+2022-12-10 07:34:23,914 INFO [train.py:421] (2/8) Epoch 2, batch 6200, loss[loss=2.445, over 1050.00 frames. , ppl: 11.534311689624372] tot_loss[loss=2.356, over 5239640.79 frames. , ppl: 10.551709688141722], batch size: 70
+2022-12-10 07:36:04,956 INFO [train.py:421] (2/8) Epoch 2, batch 6400, loss[loss=2.885, over 700.00 frames. , ppl: 17.91023147023445] tot_loss[loss=2.356, over 5252986.34 frames. , ppl: 10.553354773986701], batch size: 70
+2022-12-10 07:37:46,139 INFO [train.py:421] (2/8) Epoch 2, batch 6600, loss[loss=2.419, over 1750.00 frames. , ppl: 11.23639970422334] tot_loss[loss=2.357, over 5272857.60 frames. , ppl: 10.55793428818705], batch size: 70
+2022-12-10 07:39:25,805 INFO [train.py:421] (2/8) Epoch 2, batch 6800, loss[loss=2.515, over 1470.00 frames. , ppl: 12.362966157830279] tot_loss[loss=2.356, over 5326841.62 frames. , ppl: 10.547640814557695], batch size: 70
+2022-12-10 07:41:05,167 INFO [train.py:421] (2/8) Epoch 2, batch 7000, loss[loss=2.31, over 2870.00 frames. , ppl: 10.078947862592461] tot_loss[loss=2.355, over 5339953.84 frames. , ppl: 10.54238893342639], batch size: 70
+2022-12-10 07:41:05,168 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:41:05,947 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.402274966255137
+2022-12-10 07:42:45,099 INFO [train.py:421] (2/8) Epoch 2, batch 7200, loss[loss=2.342, over 2450.00 frames. , ppl: 10.400101950751178] tot_loss[loss=2.356, over 5328362.46 frames. , ppl: 10.546035528085127], batch size: 70
+2022-12-10 07:44:25,815 INFO [train.py:421] (2/8) Epoch 2, batch 7400, loss[loss=2.592, over 1820.00 frames. , ppl: 13.361521635428636] tot_loss[loss=2.355, over 5363232.54 frames. , ppl: 10.53511894935256], batch size: 70
+2022-12-10 07:46:04,860 INFO [train.py:421] (2/8) Epoch 2, batch 7600, loss[loss=3.064, over 560.00 frames. , ppl: 21.405392650582698] tot_loss[loss=2.354, over 5388395.23 frames. , ppl: 10.525258941883322], batch size: 70
+2022-12-10 07:47:41,328 INFO [train.py:421] (2/8) Epoch 2, batch 7800, loss[loss=2.223, over 3500.00 frames. , ppl: 9.239192777448043] tot_loss[loss=2.354, over 5421014.00 frames. , ppl: 10.525084819560695], batch size: 70
+2022-12-10 07:49:26,400 INFO [train.py:421] (2/8) Epoch 2, batch 8000, loss[loss=2.248, over 2660.00 frames. , ppl: 9.472096887480669] tot_loss[loss=2.354, over 5453321.74 frames. , ppl: 10.522755640964323], batch size: 70
+2022-12-10 07:49:26,400 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:49:27,161 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.408854278509716
+2022-12-10 07:51:05,853 INFO [train.py:421] (2/8) Epoch 2, batch 8200, loss[loss=2.327, over 2800.00 frames. , ppl: 10.24798345422148] tot_loss[loss=2.354, over 5445422.13 frames. , ppl: 10.527569161323607], batch size: 70
+2022-12-10 07:52:44,252 INFO [train.py:421] (2/8) Epoch 2, batch 8400, loss[loss=2.351, over 5950.00 frames. , ppl: 10.500320259630724] tot_loss[loss=2.353, over 5474756.25 frames. , ppl: 10.52175482090245], batch size: 70
+2022-12-10 07:54:24,068 INFO [train.py:421] (2/8) Epoch 2, batch 8600, loss[loss=2.438, over 1400.00 frames. , ppl: 11.452423662008114] tot_loss[loss=2.354, over 5460476.80 frames. , ppl: 10.530639927921955], batch size: 70
+2022-12-10 07:56:01,956 INFO [train.py:421] (2/8) Epoch 2, batch 8800, loss[loss=2.413, over 2590.00 frames. , ppl: 11.164299476421794] tot_loss[loss=2.354, over 5462098.48 frames. , ppl: 10.53203922236953], batch size: 70
+2022-12-10 07:57:42,022 INFO [train.py:421] (2/8) Epoch 2, batch 9000, loss[loss=2.509, over 840.00 frames. , ppl: 12.291924749260156] tot_loss[loss=2.354, over 5483235.23 frames. , ppl: 10.52662521277193], batch size: 70
+2022-12-10 07:57:42,022 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 07:57:42,768 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.399154636076556
+2022-12-10 07:59:22,462 INFO [train.py:421] (2/8) Epoch 2, batch 9200, loss[loss=2.384, over 2030.00 frames. , ppl: 10.848016044293264] tot_loss[loss=2.354, over 5523415.69 frames. , ppl: 10.52606835892949], batch size: 70
+2022-12-10 08:00:56,951 INFO [train.py:421] (2/8) Epoch 2, batch 9400, loss[loss=2.364, over 2240.00 frames. , ppl: 10.635103240803709] tot_loss[loss=2.356, over 5441873.95 frames. , ppl: 10.547662962602317], batch size: 70
+2022-12-10 08:02:37,352 INFO [train.py:421] (2/8) Epoch 2, batch 9600, loss[loss=2.319, over 4340.00 frames. , ppl: 10.169747610709432] tot_loss[loss=2.357, over 5439498.93 frames. , ppl: 10.555743818652234], batch size: 70
+2022-12-10 08:04:17,626 INFO [train.py:421] (2/8) Epoch 2, batch 9800, loss[loss=2.252, over 2450.00 frames. , ppl: 9.505614018132503] tot_loss[loss=2.356, over 5443977.84 frames. , ppl: 10.551839272252419], batch size: 70
+2022-12-10 08:06:01,213 INFO [train.py:421] (2/8) Epoch 2, batch 10000, loss[loss=2.816, over 700.00 frames. , ppl: 16.704963906712365] tot_loss[loss=2.356, over 5465173.91 frames. , ppl: 10.549587037254737], batch size: 70
+2022-12-10 08:06:01,214 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:06:01,975 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.415013831030345
+2022-12-10 08:07:39,459 INFO [train.py:421] (2/8) Epoch 2, batch 10200, loss[loss=2.521, over 1890.00 frames. , ppl: 12.444210280262231] tot_loss[loss=2.356, over 5470783.62 frames. , ppl: 10.550262161986101], batch size: 70
+2022-12-10 08:09:21,336 INFO [train.py:421] (2/8) Epoch 2, batch 10400, loss[loss=2.351, over 2100.00 frames. , ppl: 10.493218240834468] tot_loss[loss=2.357, over 5452156.09 frames. , ppl: 10.554404970861887], batch size: 70
+2022-12-10 08:11:01,466 INFO [train.py:421] (2/8) Epoch 2, batch 10600, loss[loss=2.477, over 2450.00 frames. , ppl: 11.904096366259159] tot_loss[loss=2.355, over 5498305.77 frames. , ppl: 10.540699119882301], batch size: 70
+2022-12-10 08:12:42,386 INFO [train.py:421] (2/8) Epoch 2, batch 10800, loss[loss=2.232, over 5320.00 frames. , ppl: 9.319276848861389] tot_loss[loss=2.357, over 5450211.00 frames. , ppl: 10.55455365865452], batch size: 70
+2022-12-10 08:14:21,836 INFO [train.py:421] (2/8) Epoch 2, batch 11000, loss[loss=2.375, over 1890.00 frames. , ppl: 10.755591518275192] tot_loss[loss=2.356, over 5474250.90 frames. , ppl: 10.547960847140347], batch size: 70
+2022-12-10 08:14:21,837 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:14:22,596 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.400310603854379
+2022-12-10 08:16:01,433 INFO [train.py:421] (2/8) Epoch 2, batch 11200, loss[loss=2.332, over 3570.00 frames. , ppl: 10.297735529994213] tot_loss[loss=2.355, over 5500909.44 frames. , ppl: 10.541215584723918], batch size: 70
+2022-12-10 08:17:42,428 INFO [train.py:421] (2/8) Epoch 2, batch 11400, loss[loss=3.22, over 490.00 frames. , ppl: 25.032824369652204] tot_loss[loss=2.356, over 5502280.35 frames. , ppl: 10.54568743083762], batch size: 70
+2022-12-10 08:19:25,307 INFO [train.py:421] (2/8) Epoch 2, batch 11600, loss[loss=2.748, over 630.00 frames. , ppl: 15.604700096555725] tot_loss[loss=2.357, over 5480483.46 frames. , ppl: 10.554287883121397], batch size: 70
+2022-12-10 08:21:08,341 INFO [train.py:421] (2/8) Epoch 2, batch 11800, loss[loss=2.295, over 2030.00 frames. , ppl: 9.92558477570617] tot_loss[loss=2.356, over 5471871.72 frames. , ppl: 10.548583058417446], batch size: 70
+2022-12-10 08:22:51,113 INFO [train.py:421] (2/8) Epoch 2, batch 12000, loss[loss=2.284, over 4270.00 frames. , ppl: 9.81797742701235] tot_loss[loss=2.355, over 5492718.27 frames. , ppl: 10.541964336889873], batch size: 70
+2022-12-10 08:22:51,113 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:22:51,875 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.405858265858988
+2022-12-10 08:24:33,733 INFO [train.py:421] (2/8) Epoch 2, batch 12200, loss[loss=2.246, over 4900.00 frames. , ppl: 9.449616615727281] tot_loss[loss=2.355, over 5497807.71 frames. , ppl: 10.540041416316717], batch size: 70
+2022-12-10 08:26:16,320 INFO [train.py:421] (2/8) Epoch 2, batch 12400, loss[loss=2.934, over 840.00 frames. , ppl: 18.81004405467282] tot_loss[loss=2.356, over 5465347.05 frames. , ppl: 10.551962712300748], batch size: 70
+2022-12-10 08:27:55,049 INFO [train.py:421] (2/8) Epoch 2, batch 12600, loss[loss=2.525, over 1260.00 frames. , ppl: 12.496741573388347] tot_loss[loss=2.356, over 5479113.40 frames. , ppl: 10.548552704011328], batch size: 70
+2022-12-10 08:29:33,545 INFO [train.py:421] (2/8) Epoch 2, batch 12800, loss[loss=2.288, over 1610.00 frames. , ppl: 9.857537239230991] tot_loss[loss=2.356, over 5480799.51 frames. , ppl: 10.548263253267624], batch size: 70
+2022-12-10 08:31:14,850 INFO [train.py:421] (2/8) Epoch 2, batch 13000, loss[loss=2.51, over 770.00 frames. , ppl: 12.304660081621005] tot_loss[loss=2.357, over 5434698.61 frames. , ppl: 10.559271943660429], batch size: 70
+2022-12-10 08:31:14,850 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:31:15,609 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.345, over 211138.00 frames. , ppl: 10.435761475361208
+2022-12-10 08:32:58,530 INFO [train.py:421] (2/8) Epoch 2, batch 13200, loss[loss=2.282, over 3150.00 frames. , ppl: 9.79369132716836] tot_loss[loss=2.356, over 5468499.20 frames. , ppl: 10.551029726056141], batch size: 70
+2022-12-10 08:34:37,211 INFO [train.py:421] (2/8) Epoch 2, batch 13400, loss[loss=3.151, over 560.00 frames. , ppl: 23.35216495242751] tot_loss[loss=2.354, over 5531706.98 frames. , ppl: 10.529142342204015], batch size: 70
+2022-12-10 08:36:17,992 INFO [train.py:421] (2/8) Epoch 2, batch 13600, loss[loss=2.523, over 1120.00 frames. , ppl: 12.46144756172456] tot_loss[loss=2.352, over 5576226.67 frames. , ppl: 10.510982013654209], batch size: 70
+2022-12-10 08:38:00,730 INFO [train.py:421] (2/8) Epoch 2, batch 13800, loss[loss=2.313, over 2940.00 frames. , ppl: 10.100745994831582] tot_loss[loss=2.353, over 5550471.26 frames. , ppl: 10.518007066682818], batch size: 70
+2022-12-10 08:39:39,900 INFO [train.py:421] (2/8) Epoch 2, batch 14000, loss[loss=2.955, over 560.00 frames. , ppl: 19.20711207329518] tot_loss[loss=2.353, over 5526772.98 frames. , ppl: 10.520501779474598], batch size: 70
+2022-12-10 08:39:39,901 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:39:40,646 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.34, over 211138.00 frames. , ppl: 10.38519312924439
+2022-12-10 08:41:22,077 INFO [train.py:421] (2/8) Epoch 2, batch 14200, loss[loss=2.393, over 2590.00 frames. , ppl: 10.951716966568714] tot_loss[loss=2.352, over 5581471.95 frames. , ppl: 10.50565958361184], batch size: 70
+2022-12-10 08:43:05,914 INFO [train.py:421] (2/8) Epoch 2, batch 14400, loss[loss=2.281, over 2240.00 frames. , ppl: 9.782891265743539] tot_loss[loss=2.353, over 5575641.99 frames. , ppl: 10.511896032706238], batch size: 70
+2022-12-10 08:44:46,234 INFO [train.py:421] (2/8) Epoch 2, batch 14600, loss[loss=2.428, over 1330.00 frames. , ppl: 11.331690958642424] tot_loss[loss=2.352, over 5581612.82 frames. , ppl: 10.511568656692944], batch size: 70
+2022-12-10 08:46:23,518 INFO [train.py:421] (2/8) Epoch 2, batch 14800, loss[loss=3.256, over 490.00 frames. , ppl: 25.94307501940731] tot_loss[loss=2.353, over 5547810.73 frames. , ppl: 10.519377215706598], batch size: 70
+2022-12-10 08:48:03,521 INFO [train.py:421] (2/8) Epoch 2, batch 15000, loss[loss=2.469, over 1400.00 frames. , ppl: 11.808495514786971] tot_loss[loss=2.353, over 5529492.60 frames. , ppl: 10.519787822868553], batch size: 70
+2022-12-10 08:48:03,521 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:48:04,267 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.339, over 211138.00 frames. , ppl: 10.36868412657422
+2022-12-10 08:49:43,381 INFO [train.py:421] (2/8) Epoch 2, batch 15200, loss[loss=2.227, over 6510.00 frames. , ppl: 9.273226121036265] tot_loss[loss=2.351, over 5593736.88 frames. , ppl: 10.500348480389036], batch size: 70
+2022-12-10 08:51:24,495 INFO [train.py:421] (2/8) Epoch 2, batch 15400, loss[loss=2.34, over 3430.00 frames. , ppl: 10.383426053658491] tot_loss[loss=2.351, over 5573815.15 frames. , ppl: 10.495013367101828], batch size: 70
+2022-12-10 08:53:07,440 INFO [train.py:421] (2/8) Epoch 2, batch 15600, loss[loss=2.301, over 3780.00 frames. , ppl: 9.979922703523203] tot_loss[loss=2.35, over 5596458.91 frames. , ppl: 10.480829139554494], batch size: 70
+2022-12-10 08:54:45,435 INFO [train.py:421] (2/8) Epoch 2, batch 15800, loss[loss=2.353, over 2100.00 frames. , ppl: 10.519777621572857] tot_loss[loss=2.349, over 5583439.92 frames. , ppl: 10.479513871981249], batch size: 70
+2022-12-10 08:56:24,652 INFO [train.py:421] (2/8) Epoch 2, batch 16000, loss[loss=2.444, over 1260.00 frames. , ppl: 11.523940087816518] tot_loss[loss=2.35, over 5582630.85 frames. , ppl: 10.481283072181494], batch size: 70
+2022-12-10 08:56:24,653 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 08:56:25,397 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.34, over 211138.00 frames. , ppl: 10.380303272822726
+2022-12-10 08:58:07,170 INFO [train.py:421] (2/8) Epoch 2, batch 16200, loss[loss=2.326, over 3920.00 frames. , ppl: 10.236480693549261] tot_loss[loss=2.349, over 5630627.74 frames. , ppl: 10.475230853891475], batch size: 70
+2022-12-10 08:59:47,484 INFO [train.py:421] (2/8) Epoch 2, batch 16400, loss[loss=2.43, over 2310.00 frames. , ppl: 11.358537780631385] tot_loss[loss=2.349, over 5619560.69 frames. , ppl: 10.478661282441589], batch size: 70
+2022-12-10 09:01:27,414 INFO [train.py:421] (2/8) Epoch 2, batch 16600, loss[loss=2.581, over 1610.00 frames. , ppl: 13.210639660280046] tot_loss[loss=2.349, over 5609728.83 frames. , ppl: 10.476363240726839], batch size: 70
+2022-12-10 09:03:04,463 INFO [train.py:421] (2/8) Epoch 2, batch 16800, loss[loss=2.349, over 2310.00 frames. , ppl: 10.472082685539329] tot_loss[loss=2.351, over 5545834.50 frames. , ppl: 10.491665311179062], batch size: 70
+2022-12-10 09:04:43,291 INFO [train.py:421] (2/8) Epoch 2, batch 17000, loss[loss=2.471, over 1820.00 frames. , ppl: 11.835333922745805] tot_loss[loss=2.35, over 5561701.47 frames. , ppl: 10.48838581314238], batch size: 70
+2022-12-10 09:04:43,291 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:04:44,079 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.34, over 211138.00 frames. , ppl: 10.380625913933697
+2022-12-10 09:06:22,870 INFO [train.py:421] (2/8) Epoch 2, batch 17200, loss[loss=2.368, over 3710.00 frames. , ppl: 10.679372748616176] tot_loss[loss=2.351, over 5559161.54 frames. , ppl: 10.492890492929654], batch size: 70
+2022-12-10 09:08:07,716 INFO [train.py:421] (2/8) Epoch 2, batch 17400, loss[loss=2.273, over 7840.00 frames. , ppl: 9.712982656296457] tot_loss[loss=2.351, over 5565697.82 frames. , ppl: 10.497490107797363], batch size: 70
+2022-12-10 09:09:46,649 INFO [train.py:421] (2/8) Epoch 2, batch 17600, loss[loss=2.586, over 1050.00 frames. , ppl: 13.281774593334854] tot_loss[loss=2.352, over 5558005.51 frames. , ppl: 10.504012030187104], batch size: 70
+2022-12-10 09:11:23,684 INFO [train.py:421] (2/8) Epoch 2, batch 17800, loss[loss=2.42, over 2800.00 frames. , ppl: 11.2419514774047] tot_loss[loss=2.353, over 5527213.85 frames. , ppl: 10.513633483561398], batch size: 70
+2022-12-10 09:13:01,348 INFO [train.py:421] (2/8) Epoch 2, batch 18000, loss[loss=2.398, over 2520.00 frames. , ppl: 10.999173933138831] tot_loss[loss=2.352, over 5544248.20 frames. , ppl: 10.506535925935326], batch size: 70
+2022-12-10 09:13:01,348 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:13:02,108 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.339, over 211138.00 frames. , ppl: 10.371116820501767
+2022-12-10 09:14:39,901 INFO [train.py:421] (2/8) Epoch 2, batch 18200, loss[loss=2.407, over 3290.00 frames. , ppl: 11.104004571296551] tot_loss[loss=2.352, over 5521130.48 frames. , ppl: 10.511174743148308], batch size: 70
+2022-12-10 09:16:18,845 INFO [train.py:421] (2/8) Epoch 2, batch 18400, loss[loss=2.354, over 3570.00 frames. , ppl: 10.529978663599364] tot_loss[loss=2.353, over 5496766.66 frames. , ppl: 10.512513217219713], batch size: 70
+2022-12-10 09:17:58,062 INFO [train.py:421] (2/8) Epoch 2, batch 18600, loss[loss=2.323, over 5740.00 frames. , ppl: 10.210504916206377] tot_loss[loss=2.352, over 5499951.67 frames. , ppl: 10.508776504091268], batch size: 70
+2022-12-10 09:19:35,458 INFO [train.py:421] (2/8) Epoch 2, batch 18800, loss[loss=3.048, over 700.00 frames. , ppl: 21.08133336602138] tot_loss[loss=2.353, over 5461024.11 frames. , ppl: 10.518213114161332], batch size: 70
+2022-12-10 09:21:19,032 INFO [train.py:421] (2/8) Epoch 2, batch 19000, loss[loss=2.37, over 5950.00 frames. , ppl: 10.694025298604812] tot_loss[loss=2.353, over 5488219.13 frames. , ppl: 10.512099879013814], batch size: 70
+2022-12-10 09:21:19,032 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:21:19,781 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.337, over 211138.00 frames. , ppl: 10.346198405029323
+2022-12-10 09:23:00,697 INFO [train.py:421] (2/8) Epoch 2, batch 19200, loss[loss=2.311, over 6370.00 frames. , ppl: 10.08341627375813] tot_loss[loss=2.353, over 5500368.96 frames. , ppl: 10.516965118933431], batch size: 70
+2022-12-10 09:24:40,865 INFO [train.py:421] (2/8) Epoch 2, batch 19400, loss[loss=2.491, over 1680.00 frames. , ppl: 12.078354535838253] tot_loss[loss=2.352, over 5529305.39 frames. , ppl: 10.50866756489429], batch size: 70
+2022-12-10 09:26:20,689 INFO [train.py:421] (2/8) Epoch 2, batch 19600, loss[loss=2.272, over 1610.00 frames. , ppl: 9.697707888713337] tot_loss[loss=2.352, over 5540191.61 frames. , ppl: 10.502350229554509], batch size: 70
+2022-12-10 09:28:00,785 INFO [train.py:421] (2/8) Epoch 2, batch 19800, loss[loss=2.481, over 1610.00 frames. , ppl: 11.947892244407331] tot_loss[loss=2.352, over 5514411.00 frames. , ppl: 10.50832673726987], batch size: 70
+2022-12-10 09:29:37,936 INFO [train.py:421] (2/8) Epoch 2, batch 20000, loss[loss=2.199, over 3010.00 frames. , ppl: 9.014095480819876] tot_loss[loss=2.352, over 5489779.64 frames. , ppl: 10.511804118400946], batch size: 70
+2022-12-10 09:29:37,936 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:29:38,694 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.339, over 211138.00 frames. , ppl: 10.367246266120773
+2022-12-10 09:31:19,562 INFO [train.py:421] (2/8) Epoch 2, batch 20200, loss[loss=2.282, over 3500.00 frames. , ppl: 9.792820876237524] tot_loss[loss=2.352, over 5463734.54 frames. , ppl: 10.511281914681645], batch size: 70
+2022-12-10 09:32:58,803 INFO [train.py:421] (2/8) Epoch 2, batch 20400, loss[loss=2.313, over 4200.00 frames. , ppl: 10.102300870319668] tot_loss[loss=2.351, over 5486431.79 frames. , ppl: 10.499273833745274], batch size: 70
+2022-12-10 09:34:39,600 INFO [train.py:421] (2/8) Epoch 2, batch 20600, loss[loss=2.623, over 770.00 frames. , ppl: 13.773253247237639] tot_loss[loss=2.351, over 5474251.25 frames. , ppl: 10.500989805198163], batch size: 70
+2022-12-10 09:36:18,783 INFO [train.py:421] (2/8) Epoch 2, batch 20800, loss[loss=2.292, over 1820.00 frames. , ppl: 9.891933316841309] tot_loss[loss=2.352, over 5461766.20 frames. , ppl: 10.505717275354156], batch size: 70
+2022-12-10 09:37:55,971 INFO [train.py:421] (2/8) Epoch 2, batch 21000, loss[loss=2.555, over 770.00 frames. , ppl: 12.86828166812146] tot_loss[loss=2.352, over 5462672.90 frames. , ppl: 10.504016990451728], batch size: 70
+2022-12-10 09:37:55,972 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:37:56,718 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.338, over 211138.00 frames. , ppl: 10.362088796314158
+2022-12-10 09:39:34,399 INFO [train.py:421] (2/8) Epoch 2, batch 21200, loss[loss=2.418, over 2380.00 frames. , ppl: 11.221086979659791] tot_loss[loss=2.352, over 5433203.10 frames. , ppl: 10.509083749615948], batch size: 70
+2022-12-10 09:41:12,871 INFO [train.py:421] (2/8) Epoch 2, batch 21400, loss[loss=2.246, over 6230.00 frames. , ppl: 9.45042075305107] tot_loss[loss=2.352, over 5453715.68 frames. , ppl: 10.506033691755142], batch size: 70
+2022-12-10 09:42:51,082 INFO [train.py:421] (2/8) Epoch 2, batch 21600, loss[loss=2.504, over 1330.00 frames. , ppl: 12.22792109705709] tot_loss[loss=2.351, over 5489781.14 frames. , ppl: 10.494897688563126], batch size: 70
+2022-12-10 09:44:31,993 INFO [train.py:421] (2/8) Epoch 2, batch 21800, loss[loss=2.394, over 1820.00 frames. , ppl: 10.960972780670565] tot_loss[loss=2.351, over 5514369.90 frames. , ppl: 10.491315456046896], batch size: 70
+2022-12-10 09:46:13,455 INFO [train.py:421] (2/8) Epoch 2, batch 22000, loss[loss=2.249, over 3500.00 frames. , ppl: 9.482143850502316] tot_loss[loss=2.351, over 5473345.61 frames. , ppl: 10.499748535749532], batch size: 70
+2022-12-10 09:46:13,456 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:46:14,216 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.337, over 211138.00 frames. , ppl: 10.353811838414043
+2022-12-10 09:47:52,635 INFO [train.py:421] (2/8) Epoch 2, batch 22200, loss[loss=2.241, over 5040.00 frames. , ppl: 9.404474198759303] tot_loss[loss=2.351, over 5509750.19 frames. , ppl: 10.49260477076819], batch size: 70
+2022-12-10 09:49:28,392 INFO [train.py:421] (2/8) Epoch 2, batch 22400, loss[loss=2.891, over 840.00 frames. , ppl: 18.007432031822045] tot_loss[loss=2.35, over 5512887.51 frames. , ppl: 10.49070387954277], batch size: 70
+2022-12-10 09:51:04,810 INFO [train.py:421] (2/8) Epoch 2, batch 22600, loss[loss=2.364, over 2380.00 frames. , ppl: 10.629401283832447] tot_loss[loss=2.351, over 5484770.76 frames. , ppl: 10.495233070927329], batch size: 70
+2022-12-10 09:52:46,947 INFO [train.py:421] (2/8) Epoch 2, batch 22800, loss[loss=2.233, over 1890.00 frames. , ppl: 9.328723930753535] tot_loss[loss=2.351, over 5472020.11 frames. , ppl: 10.496514361189764], batch size: 70
+2022-12-10 09:54:25,159 INFO [train.py:421] (2/8) Epoch 2, batch 23000, loss[loss=3.339, over 490.00 frames. , ppl: 28.199005483455608] tot_loss[loss=2.351, over 5466553.28 frames. , ppl: 10.494507277669124], batch size: 70
+2022-12-10 09:54:25,159 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 09:54:25,907 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.337, over 211138.00 frames. , ppl: 10.35141844156331
+2022-12-10 09:56:05,774 INFO [train.py:421] (2/8) Epoch 2, batch 23200, loss[loss=2.733, over 700.00 frames. , ppl: 15.386363339478319] tot_loss[loss=2.35, over 5474368.21 frames. , ppl: 10.488717763522576], batch size: 70
+2022-12-10 09:57:48,629 INFO [train.py:421] (2/8) Epoch 2, batch 23400, loss[loss=2.576, over 770.00 frames. , ppl: 13.140171724117792] tot_loss[loss=2.352, over 5451996.43 frames. , ppl: 10.503700877607924], batch size: 70
+2022-12-10 09:59:32,018 INFO [train.py:421] (2/8) Epoch 2, batch 23600, loss[loss=2.458, over 2170.00 frames. , ppl: 11.681046078700167] tot_loss[loss=2.352, over 5435892.87 frames. , ppl: 10.506229860958202], batch size: 70
+2022-12-10 10:01:14,603 INFO [train.py:421] (2/8) Epoch 2, batch 23800, loss[loss=2.648, over 840.00 frames. , ppl: 14.128408342418364] tot_loss[loss=2.352, over 5404233.84 frames. , ppl: 10.511040975152714], batch size: 70
+2022-12-10 10:02:54,344 INFO [train.py:421] (2/8) Epoch 2, batch 24000, loss[loss=2.344, over 1960.00 frames. , ppl: 10.418269165573028] tot_loss[loss=2.353, over 5404931.76 frames. , ppl: 10.514502050225053], batch size: 70
+2022-12-10 10:02:54,345 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:02:55,105 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.32757755196688
+2022-12-10 10:04:39,373 INFO [train.py:421] (2/8) Epoch 2, batch 24200, loss[loss=2.405, over 2100.00 frames. , ppl: 11.081037399605334] tot_loss[loss=2.352, over 5415620.99 frames. , ppl: 10.511441081253022], batch size: 70
+2022-12-10 10:06:16,094 INFO [train.py:421] (2/8) Epoch 2, batch 24400, loss[loss=2.342, over 3500.00 frames. , ppl: 10.405294177411498] tot_loss[loss=2.352, over 5439514.12 frames. , ppl: 10.504789497484653], batch size: 70
+2022-12-10 10:07:52,407 INFO [train.py:421] (2/8) Epoch 2, batch 24600, loss[loss=2.343, over 1680.00 frames. , ppl: 10.407502827085327] tot_loss[loss=2.352, over 5398974.38 frames. , ppl: 10.507755838074939], batch size: 70
+2022-12-10 10:09:36,242 INFO [train.py:421] (2/8) Epoch 2, batch 24800, loss[loss=2.512, over 980.00 frames. , ppl: 12.330907022951186] tot_loss[loss=2.351, over 5426692.76 frames. , ppl: 10.499853166664035], batch size: 70
+2022-12-10 10:11:14,032 INFO [train.py:421] (2/8) Epoch 2, batch 25000, loss[loss=2.654, over 1190.00 frames. , ppl: 14.205632145757473] tot_loss[loss=2.352, over 5408385.39 frames. , ppl: 10.506803118455162], batch size: 70
+2022-12-10 10:11:14,032 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:11:14,793 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.336, over 211138.00 frames. , ppl: 10.337237971792588
+2022-12-10 10:12:54,450 INFO [train.py:421] (2/8) Epoch 2, batch 25200, loss[loss=2.648, over 1330.00 frames. , ppl: 14.12930225422263] tot_loss[loss=2.352, over 5414452.83 frames. , ppl: 10.509284763174673], batch size: 70
+2022-12-10 10:14:32,509 INFO [train.py:421] (2/8) Epoch 2, batch 25400, loss[loss=2.824, over 700.00 frames. , ppl: 16.84547426740087] tot_loss[loss=2.352, over 5445735.93 frames. , ppl: 10.501389629155105], batch size: 70
+2022-12-10 10:16:11,533 INFO [train.py:421] (2/8) Epoch 2, batch 25600, loss[loss=2.594, over 980.00 frames. , ppl: 13.387260576407492] tot_loss[loss=2.351, over 5465539.96 frames. , ppl: 10.499464006692683], batch size: 70
+2022-12-10 10:17:49,158 INFO [train.py:421] (2/8) Epoch 2, batch 25800, loss[loss=2.482, over 3220.00 frames. , ppl: 11.964307366968415] tot_loss[loss=2.353, over 5416915.69 frames. , ppl: 10.520051448271584], batch size: 70
+2022-12-10 10:19:31,250 INFO [train.py:421] (2/8) Epoch 2, batch 26000, loss[loss=2.326, over 3850.00 frames. , ppl: 10.234034192650796] tot_loss[loss=2.353, over 5406504.83 frames. , ppl: 10.517851621834652], batch size: 70
+2022-12-10 10:19:31,251 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:19:31,994 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.33015502328756
+2022-12-10 10:21:14,018 INFO [train.py:421] (2/8) Epoch 2, batch 26200, loss[loss=2.263, over 5600.00 frames. , ppl: 9.609358947591359] tot_loss[loss=2.353, over 5417152.92 frames. , ppl: 10.515964842864241], batch size: 70
+2022-12-10 10:22:55,811 INFO [train.py:421] (2/8) Epoch 2, batch 26400, loss[loss=2.276, over 4830.00 frames. , ppl: 9.73775433410851] tot_loss[loss=2.351, over 5470295.89 frames. , ppl: 10.500512486449981], batch size: 70
+2022-12-10 10:24:38,308 INFO [train.py:421] (2/8) Epoch 2, batch 26600, loss[loss=2.324, over 2940.00 frames. , ppl: 10.21430744250387] tot_loss[loss=2.351, over 5489757.63 frames. , ppl: 10.497896642751321], batch size: 70
+2022-12-10 10:26:19,546 INFO [train.py:421] (2/8) Epoch 2, batch 26800, loss[loss=2.5, over 840.00 frames. , ppl: 12.188532459237999] tot_loss[loss=2.35, over 5487273.12 frames. , ppl: 10.486894910022349], batch size: 70
+2022-12-10 10:27:59,904 INFO [train.py:421] (2/8) Epoch 2, batch 27000, loss[loss=2.495, over 980.00 frames. , ppl: 12.124824379660806] tot_loss[loss=2.351, over 5456705.25 frames. , ppl: 10.491735027639944], batch size: 70
+2022-12-10 10:27:59,904 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:28:00,666 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.332974773515254
+2022-12-10 10:29:43,671 INFO [train.py:421] (2/8) Epoch 2, batch 27200, loss[loss=2.413, over 1540.00 frames. , ppl: 11.171477135198447] tot_loss[loss=2.351, over 5452459.39 frames. , ppl: 10.49534415552974], batch size: 70
+2022-12-10 10:31:24,575 INFO [train.py:421] (2/8) Epoch 2, batch 27400, loss[loss=2.343, over 4200.00 frames. , ppl: 10.413936907378837] tot_loss[loss=2.35, over 5489456.65 frames. , ppl: 10.486115820977739], batch size: 70
+2022-12-10 10:33:03,927 INFO [train.py:421] (2/8) Epoch 2, batch 27600, loss[loss=2.448, over 2310.00 frames. , ppl: 11.570867949887486] tot_loss[loss=2.35, over 5502508.96 frames. , ppl: 10.482223639405225], batch size: 70
+2022-12-10 10:34:44,430 INFO [train.py:421] (2/8) Epoch 2, batch 27800, loss[loss=4.234, over 350.00 frames. , ppl: 68.9847161430649] tot_loss[loss=2.35, over 5496823.04 frames. , ppl: 10.480524176866844], batch size: 70
+2022-12-10 10:36:25,384 INFO [train.py:421] (2/8) Epoch 2, batch 28000, loss[loss=2.456, over 1540.00 frames. , ppl: 11.657453669872272] tot_loss[loss=2.35, over 5459830.16 frames. , ppl: 10.48596780989223], batch size: 70
+2022-12-10 10:36:25,385 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:36:26,133 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.333, over 211138.00 frames. , ppl: 10.313298466854016
+2022-12-10 10:38:05,410 INFO [train.py:421] (2/8) Epoch 2, batch 28200, loss[loss=2.274, over 3850.00 frames. , ppl: 9.719984264516711] tot_loss[loss=2.348, over 5513136.15 frames. , ppl: 10.468561347195573], batch size: 70
+2022-12-10 10:39:46,466 INFO [train.py:421] (2/8) Epoch 2, batch 28400, loss[loss=2.344, over 1680.00 frames. , ppl: 10.422168665665637] tot_loss[loss=2.347, over 5547837.59 frames. , ppl: 10.456839357011601], batch size: 70
+2022-12-10 10:41:25,663 INFO [train.py:421] (2/8) Epoch 2, batch 28600, loss[loss=2.321, over 2730.00 frames. , ppl: 10.190370826117483] tot_loss[loss=2.348, over 5531718.43 frames. , ppl: 10.459770476952798], batch size: 70
+2022-12-10 10:43:03,861 INFO [train.py:421] (2/8) Epoch 2, batch 28800, loss[loss=2.208, over 7630.00 frames. , ppl: 9.096875642626134] tot_loss[loss=2.349, over 5475067.32 frames. , ppl: 10.479628584421636], batch size: 70
+2022-12-10 10:44:44,078 INFO [train.py:421] (2/8) Epoch 2, batch 29000, loss[loss=2.566, over 910.00 frames. , ppl: 13.013205513188293] tot_loss[loss=2.349, over 5470319.96 frames. , ppl: 10.480129225842987], batch size: 70
+2022-12-10 10:44:44,079 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:44:44,826 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.328367846981726
+2022-12-10 10:46:24,218 INFO [train.py:421] (2/8) Epoch 2, batch 29200, loss[loss=2.25, over 2030.00 frames. , ppl: 9.488817618215597] tot_loss[loss=2.35, over 5450901.80 frames. , ppl: 10.486009234119892], batch size: 70
+2022-12-10 10:48:02,866 INFO [train.py:421] (2/8) Epoch 2, batch 29400, loss[loss=2.302, over 7350.00 frames. , ppl: 9.992928144925035] tot_loss[loss=2.351, over 5412891.85 frames. , ppl: 10.500439809852189], batch size: 70
+2022-12-10 10:49:40,055 INFO [train.py:421] (2/8) Epoch 2, batch 29600, loss[loss=2.413, over 1610.00 frames. , ppl: 11.169524611720908] tot_loss[loss=2.351, over 5412586.03 frames. , ppl: 10.492624305307942], batch size: 70
+2022-12-10 10:51:22,076 INFO [train.py:421] (2/8) Epoch 2, batch 29800, loss[loss=2.303, over 3570.00 frames. , ppl: 10.000327515548797] tot_loss[loss=2.35, over 5438191.65 frames. , ppl: 10.483213906406668], batch size: 70
+2022-12-10 10:53:03,452 INFO [train.py:421] (2/8) Epoch 2, batch 30000, loss[loss=2.44, over 2030.00 frames. , ppl: 11.470850900926619] tot_loss[loss=2.35, over 5419117.47 frames. , ppl: 10.485984345792238], batch size: 70
+2022-12-10 10:53:03,452 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 10:53:04,215 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.333, over 211138.00 frames. , ppl: 10.311935441408778
+2022-12-10 10:54:45,305 INFO [train.py:421] (2/8) Epoch 2, batch 30200, loss[loss=2.734, over 770.00 frames. , ppl: 15.394340028112383] tot_loss[loss=2.349, over 5447735.05 frames. , ppl: 10.474437387657357], batch size: 70
+2022-12-10 10:56:29,740 INFO [train.py:421] (2/8) Epoch 2, batch 30400, loss[loss=2.328, over 2590.00 frames. , ppl: 10.258541463215943] tot_loss[loss=2.349, over 5447925.08 frames. , ppl: 10.4754669546392], batch size: 70
+2022-12-10 10:58:09,968 INFO [train.py:421] (2/8) Epoch 2, batch 30600, loss[loss=2.338, over 2660.00 frames. , ppl: 10.365178476320022] tot_loss[loss=2.349, over 5430118.09 frames. , ppl: 10.477368095522356], batch size: 70
+2022-12-10 10:59:47,294 INFO [train.py:421] (2/8) Epoch 2, batch 30800, loss[loss=2.34, over 4480.00 frames. , ppl: 10.381755239241159] tot_loss[loss=2.35, over 5417971.04 frames. , ppl: 10.486045157827036], batch size: 70
+2022-12-10 11:01:28,587 INFO [train.py:421] (2/8) Epoch 2, batch 31000, loss[loss=2.367, over 1400.00 frames. , ppl: 10.663994656013132] tot_loss[loss=2.351, over 5369859.88 frames. , ppl: 10.499587912515615], batch size: 70
+2022-12-10 11:01:28,588 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:01:29,350 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.334, over 211138.00 frames. , ppl: 10.318448451765484
+2022-12-10 11:03:12,446 INFO [train.py:421] (2/8) Epoch 2, batch 31200, loss[loss=2.33, over 6230.00 frames. , ppl: 10.27851412818698] tot_loss[loss=2.35, over 5409148.62 frames. , ppl: 10.489277092586972], batch size: 70
+2022-12-10 11:04:53,797 INFO [train.py:421] (2/8) Epoch 2, batch 31400, loss[loss=2.352, over 3850.00 frames. , ppl: 10.510920588497312] tot_loss[loss=2.349, over 5444606.89 frames. , ppl: 10.47427789949935], batch size: 70
+2022-12-10 11:06:34,443 INFO [train.py:421] (2/8) Epoch 2, batch 31600, loss[loss=2.435, over 1190.00 frames. , ppl: 11.419954623884617] tot_loss[loss=2.349, over 5435294.83 frames. , ppl: 10.4789501884006], batch size: 70
+2022-12-10 11:08:17,173 INFO [train.py:421] (2/8) Epoch 2, batch 31800, loss[loss=2.396, over 3220.00 frames. , ppl: 10.975561604781348] tot_loss[loss=2.349, over 5436290.98 frames. , ppl: 10.477482348690694], batch size: 70
+2022-12-10 11:09:56,529 INFO [train.py:421] (2/8) Epoch 2, batch 32000, loss[loss=2.268, over 5600.00 frames. , ppl: 9.664376196766682] tot_loss[loss=2.349, over 5432669.57 frames. , ppl: 10.478102413712374], batch size: 70
+2022-12-10 11:09:56,529 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:09:57,291 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.334, over 211138.00 frames. , ppl: 10.31484792506888
+2022-12-10 11:11:36,906 INFO [train.py:421] (2/8) Epoch 2, batch 32200, loss[loss=2.443, over 1470.00 frames. , ppl: 11.506683723571866] tot_loss[loss=2.349, over 5442977.68 frames. , ppl: 10.476675995719187], batch size: 70
+2022-12-10 11:13:14,855 INFO [train.py:421] (2/8) Epoch 2, batch 32400, loss[loss=2.603, over 1260.00 frames. , ppl: 13.500726270779323] tot_loss[loss=2.349, over 5458257.51 frames. , ppl: 10.478718538672853], batch size: 70
+2022-12-10 11:14:55,253 INFO [train.py:421] (2/8) Epoch 2, batch 32600, loss[loss=2.325, over 3500.00 frames. , ppl: 10.223558913735786] tot_loss[loss=2.349, over 5475067.01 frames. , ppl: 10.479400532042636], batch size: 70
+2022-12-10 11:16:38,284 INFO [train.py:421] (2/8) Epoch 2, batch 32800, loss[loss=2.719, over 630.00 frames. , ppl: 15.165783520496682] tot_loss[loss=2.349, over 5486149.24 frames. , ppl: 10.47506659114931], batch size: 70
+2022-12-10 11:18:17,211 INFO [train.py:421] (2/8) Epoch 2, batch 33000, loss[loss=3.347, over 490.00 frames. , ppl: 28.40612953492327] tot_loss[loss=2.347, over 5550420.79 frames. , ppl: 10.450030521715755], batch size: 70
+2022-12-10 11:18:17,211 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:18:17,973 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.332, over 211138.00 frames. , ppl: 10.303007731813349
+2022-12-10 11:20:00,455 INFO [train.py:421] (2/8) Epoch 2, batch 33200, loss[loss=2.274, over 5600.00 frames. , ppl: 9.721240825789394] tot_loss[loss=2.346, over 5565096.74 frames. , ppl: 10.447754646452177], batch size: 70
+2022-12-10 11:21:39,767 INFO [train.py:421] (2/8) Epoch 2, batch 33400, loss[loss=2.257, over 3990.00 frames. , ppl: 9.556197048285288] tot_loss[loss=2.347, over 5532626.15 frames. , ppl: 10.458981861261037], batch size: 70
+2022-12-10 11:23:22,970 INFO [train.py:421] (2/8) Epoch 2, batch 33600, loss[loss=2.326, over 2660.00 frames. , ppl: 10.236989526966093] tot_loss[loss=2.346, over 5585082.75 frames. , ppl: 10.446668390828917], batch size: 70
+2022-12-10 11:25:04,054 INFO [train.py:421] (2/8) Epoch 2, batch 33800, loss[loss=2.474, over 1890.00 frames. , ppl: 11.872247188687302] tot_loss[loss=2.346, over 5572444.31 frames. , ppl: 10.443462452222612], batch size: 70
+2022-12-10 11:26:42,437 INFO [train.py:421] (2/8) Epoch 2, batch 34000, loss[loss=2.386, over 1680.00 frames. , ppl: 10.868140609503403] tot_loss[loss=2.345, over 5603366.93 frames. , ppl: 10.433315240892224], batch size: 70
+2022-12-10 11:26:42,437 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:26:43,186 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.291255765342557
+2022-12-10 11:28:20,402 INFO [train.py:421] (2/8) Epoch 2, batch 34200, loss[loss=2.286, over 4970.00 frames. , ppl: 9.835883870003926] tot_loss[loss=2.346, over 5564725.09 frames. , ppl: 10.442830299964953], batch size: 70
+2022-12-10 11:30:00,241 INFO [train.py:421] (2/8) Epoch 2, batch 34400, loss[loss=2.249, over 8540.00 frames. , ppl: 9.480213692687393] tot_loss[loss=2.347, over 5551704.77 frames. , ppl: 10.450727905381482], batch size: 70
+2022-12-10 11:31:39,812 INFO [train.py:421] (2/8) Epoch 2, batch 34600, loss[loss=2.193, over 9170.00 frames. , ppl: 8.96448408634109] tot_loss[loss=2.346, over 5572987.95 frames. , ppl: 10.447226230923002], batch size: 70
+2022-12-10 11:33:17,825 INFO [train.py:421] (2/8) Epoch 2, batch 34800, loss[loss=2.267, over 5950.00 frames. , ppl: 9.647861761815882] tot_loss[loss=2.346, over 5567117.69 frames. , ppl: 10.439148644573805], batch size: 70
+2022-12-10 11:35:00,073 INFO [train.py:421] (2/8) Epoch 2, batch 35000, loss[loss=2.235, over 7490.00 frames. , ppl: 9.343531320789934] tot_loss[loss=2.345, over 5553119.50 frames. , ppl: 10.43834271852632], batch size: 70
+2022-12-10 11:35:00,073 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:35:00,819 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.332, over 211138.00 frames. , ppl: 10.297171983552662
+2022-12-10 11:36:40,695 INFO [train.py:421] (2/8) Epoch 2, batch 35200, loss[loss=2.212, over 7560.00 frames. , ppl: 9.1305293968674] tot_loss[loss=2.345, over 5565648.62 frames. , ppl: 10.43173522631984], batch size: 70
+2022-12-10 11:38:22,509 INFO [train.py:421] (2/8) Epoch 2, batch 35400, loss[loss=2.334, over 2590.00 frames. , ppl: 10.317641890678427] tot_loss[loss=2.345, over 5541008.84 frames. , ppl: 10.430877104223468], batch size: 70
+2022-12-10 11:40:03,278 INFO [train.py:421] (2/8) Epoch 2, batch 35600, loss[loss=2.3, over 6510.00 frames. , ppl: 9.973337124571916] tot_loss[loss=2.345, over 5540770.20 frames. , ppl: 10.43353849278998], batch size: 70
+2022-12-10 11:41:42,466 INFO [train.py:421] (2/8) Epoch 2, batch 35800, loss[loss=2.226, over 5740.00 frames. , ppl: 9.26137351606144] tot_loss[loss=2.345, over 5547759.45 frames. , ppl: 10.437093705710424], batch size: 70
+2022-12-10 11:43:21,048 INFO [train.py:421] (2/8) Epoch 2, batch 36000, loss[loss=2.364, over 3710.00 frames. , ppl: 10.635845112607026] tot_loss[loss=2.347, over 5496651.43 frames. , ppl: 10.451531748434926], batch size: 70
+2022-12-10 11:43:21,049 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:43:21,794 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.288832667020564
+2022-12-10 11:45:02,829 INFO [train.py:421] (2/8) Epoch 2, batch 36200, loss[loss=2.389, over 3570.00 frames. , ppl: 10.898555917970173] tot_loss[loss=2.347, over 5494917.58 frames. , ppl: 10.454844259995674], batch size: 70
+2022-12-10 11:46:46,962 INFO [train.py:421] (2/8) Epoch 2, batch 36400, loss[loss=2.246, over 5670.00 frames. , ppl: 9.450264863149965] tot_loss[loss=2.348, over 5466140.64 frames. , ppl: 10.461842184620311], batch size: 70
+2022-12-10 11:48:25,554 INFO [train.py:421] (2/8) Epoch 2, batch 36600, loss[loss=2.441, over 1330.00 frames. , ppl: 11.482477160480345] tot_loss[loss=2.349, over 5431339.81 frames. , ppl: 10.47075704155563], batch size: 70
+2022-12-10 11:50:04,140 INFO [train.py:421] (2/8) Epoch 2, batch 36800, loss[loss=2.397, over 2240.00 frames. , ppl: 10.988888119077604] tot_loss[loss=2.349, over 5421840.83 frames. , ppl: 10.48018249782364], batch size: 70
+2022-12-10 11:51:43,181 INFO [train.py:421] (2/8) Epoch 2, batch 37000, loss[loss=2.517, over 1050.00 frames. , ppl: 12.39117774260452] tot_loss[loss=2.35, over 5390535.48 frames. , ppl: 10.487688398358117], batch size: 70
+2022-12-10 11:51:43,181 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 11:51:43,927 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.333, over 211138.00 frames. , ppl: 10.313814418221853
+2022-12-10 11:53:27,750 INFO [train.py:421] (2/8) Epoch 2, batch 37200, loss[loss=2.473, over 2100.00 frames. , ppl: 11.860266032688923] tot_loss[loss=2.35, over 5382661.32 frames. , ppl: 10.490179436712276], batch size: 70
+2022-12-10 11:55:05,711 INFO [train.py:421] (2/8) Epoch 2, batch 37400, loss[loss=2.452, over 1540.00 frames. , ppl: 11.615107869996162] tot_loss[loss=2.35, over 5405147.38 frames. , ppl: 10.480884269663422], batch size: 70
+2022-12-10 11:56:46,031 INFO [train.py:421] (2/8) Epoch 2, batch 37600, loss[loss=2.377, over 1470.00 frames. , ppl: 10.769772455918844] tot_loss[loss=2.35, over 5406436.94 frames. , ppl: 10.48557260542505], batch size: 70
+2022-12-10 11:58:28,997 INFO [train.py:421] (2/8) Epoch 2, batch 37800, loss[loss=2.292, over 4760.00 frames. , ppl: 9.89069466099406] tot_loss[loss=2.348, over 5465828.14 frames. , ppl: 10.46156166200898], batch size: 70
+2022-12-10 12:00:10,577 INFO [train.py:421] (2/8) Epoch 2, batch 38000, loss[loss=2.43, over 1120.00 frames. , ppl: 11.356823191925503] tot_loss[loss=2.349, over 5458072.34 frames. , ppl: 10.470396949033482], batch size: 70
+2022-12-10 12:00:10,577 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:00:11,327 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.269646244641864
+2022-12-10 12:01:54,474 INFO [train.py:421] (2/8) Epoch 2, batch 38200, loss[loss=2.299, over 4550.00 frames. , ppl: 9.962257049107633] tot_loss[loss=2.348, over 5459528.00 frames. , ppl: 10.469645313656073], batch size: 70
+2022-12-10 12:03:35,677 INFO [train.py:421] (2/8) Epoch 2, batch 38400, loss[loss=2.318, over 2170.00 frames. , ppl: 10.152626950706738] tot_loss[loss=2.348, over 5444828.43 frames. , ppl: 10.463922252820419], batch size: 70
+2022-12-10 12:05:13,565 INFO [train.py:421] (2/8) Epoch 2, batch 38600, loss[loss=2.222, over 8330.00 frames. , ppl: 9.227063841913427] tot_loss[loss=2.348, over 5466353.14 frames. , ppl: 10.464588601693992], batch size: 70
+2022-12-10 12:06:50,147 INFO [train.py:421] (2/8) Epoch 2, batch 38800, loss[loss=2.418, over 1680.00 frames. , ppl: 11.222037987495508] tot_loss[loss=2.348, over 5470180.97 frames. , ppl: 10.461157336480545], batch size: 70
+2022-12-10 12:08:30,334 INFO [train.py:421] (2/8) Epoch 2, batch 39000, loss[loss=2.365, over 3430.00 frames. , ppl: 10.647385606905026] tot_loss[loss=2.347, over 5470312.22 frames. , ppl: 10.455923127918458], batch size: 70
+2022-12-10 12:08:30,334 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:08:31,096 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.258752219498415
+2022-12-10 12:10:14,097 INFO [train.py:421] (2/8) Epoch 2, batch 39200, loss[loss=2.308, over 3010.00 frames. , ppl: 10.052662068134675] tot_loss[loss=2.347, over 5449372.31 frames. , ppl: 10.457808142916951], batch size: 70
+2022-12-10 12:11:55,246 INFO [train.py:421] (2/8) Epoch 2, batch 39400, loss[loss=2.399, over 2380.00 frames. , ppl: 11.011788744372485] tot_loss[loss=2.348, over 5424149.92 frames. , ppl: 10.469213713226406], batch size: 70
+2022-12-10 12:13:32,407 INFO [train.py:421] (2/8) Epoch 2, batch 39600, loss[loss=2.232, over 5530.00 frames. , ppl: 9.322149795390644] tot_loss[loss=2.349, over 5413446.66 frames. , ppl: 10.473900693592414], batch size: 70
+2022-12-10 12:15:17,688 INFO [train.py:421] (2/8) Epoch 2, batch 39800, loss[loss=2.419, over 1610.00 frames. , ppl: 11.238434741148826] tot_loss[loss=2.348, over 5446567.58 frames. , ppl: 10.463163446309558], batch size: 70
+2022-12-10 12:16:56,217 INFO [train.py:421] (2/8) Epoch 2, batch 40000, loss[loss=2.323, over 4410.00 frames. , ppl: 10.210412655071] tot_loss[loss=2.348, over 5449595.67 frames. , ppl: 10.4603985165209], batch size: 70
+2022-12-10 12:16:56,217 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:16:56,977 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.292756209643162
+2022-12-10 12:18:37,688 INFO [train.py:421] (2/8) Epoch 2, batch 40200, loss[loss=2.538, over 1260.00 frames. , ppl: 12.656719299170271] tot_loss[loss=2.346, over 5504963.04 frames. , ppl: 10.440709058727922], batch size: 70
+2022-12-10 12:20:17,609 INFO [train.py:421] (2/8) Epoch 2, batch 40400, loss[loss=2.303, over 3150.00 frames. , ppl: 10.007615510054489] tot_loss[loss=2.345, over 5499540.00 frames. , ppl: 10.435617177143122], batch size: 70
+2022-12-10 12:22:00,252 INFO [train.py:421] (2/8) Epoch 2, batch 40600, loss[loss=2.301, over 3780.00 frames. , ppl: 9.979476665565114] tot_loss[loss=2.346, over 5487094.75 frames. , ppl: 10.444769672132201], batch size: 70
+2022-12-10 12:23:41,676 INFO [train.py:421] (2/8) Epoch 2, batch 40800, loss[loss=2.344, over 1820.00 frames. , ppl: 10.423471508853844] tot_loss[loss=2.346, over 5485172.52 frames. , ppl: 10.444631258368208], batch size: 70
+2022-12-10 12:25:27,338 INFO [train.py:421] (2/8) Epoch 2, batch 41000, loss[loss=2.464, over 1820.00 frames. , ppl: 11.75384331481308] tot_loss[loss=2.346, over 5459997.49 frames. , ppl: 10.447286356578411], batch size: 70
+2022-12-10 12:25:27,338 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:25:28,086 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.27130012072034
+2022-12-10 12:27:08,687 INFO [train.py:421] (2/8) Epoch 2, batch 41200, loss[loss=2.359, over 2870.00 frames. , ppl: 10.580445285979625] tot_loss[loss=2.346, over 5477134.76 frames. , ppl: 10.446185571784731], batch size: 70
+2022-12-10 12:28:47,117 INFO [train.py:421] (2/8) Epoch 2, batch 41400, loss[loss=2.325, over 2590.00 frames. , ppl: 10.224628909325604] tot_loss[loss=2.346, over 5467867.68 frames. , ppl: 10.445726988793178], batch size: 70
+2022-12-10 12:30:25,485 INFO [train.py:421] (2/8) Epoch 2, batch 41600, loss[loss=2.353, over 3010.00 frames. , ppl: 10.51204119463401] tot_loss[loss=2.347, over 5468564.05 frames. , ppl: 10.44920034504873], batch size: 70
+2022-12-10 12:32:02,273 INFO [train.py:421] (2/8) Epoch 2, batch 41800, loss[loss=2.425, over 3220.00 frames. , ppl: 11.306981270131512] tot_loss[loss=2.347, over 5447780.81 frames. , ppl: 10.450802821920565], batch size: 70
+2022-12-10 12:33:43,237 INFO [train.py:421] (2/8) Epoch 2, batch 42000, loss[loss=2.395, over 2030.00 frames. , ppl: 10.967269976215697] tot_loss[loss=2.346, over 5446876.84 frames. , ppl: 10.447824899584424], batch size: 70
+2022-12-10 12:33:43,237 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:33:43,996 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.283807052673856
+2022-12-10 12:35:25,662 INFO [train.py:421] (2/8) Epoch 2, batch 42200, loss[loss=2.251, over 5600.00 frames. , ppl: 9.492578225949535] tot_loss[loss=2.345, over 5474997.82 frames. , ppl: 10.435970793705263], batch size: 70
+2022-12-10 12:37:03,123 INFO [train.py:421] (2/8) Epoch 2, batch 42400, loss[loss=2.911, over 560.00 frames. , ppl: 18.3710565176829] tot_loss[loss=2.346, over 5427416.44 frames. , ppl: 10.446530673887752], batch size: 70
+2022-12-10 12:38:43,256 INFO [train.py:421] (2/8) Epoch 2, batch 42600, loss[loss=2.287, over 11340.00 frames. , ppl: 9.845069064360185] tot_loss[loss=2.347, over 5417458.43 frames. , ppl: 10.449408825854574], batch size: 70
+2022-12-10 12:40:22,507 INFO [train.py:421] (2/8) Epoch 2, batch 42800, loss[loss=2.223, over 8260.00 frames. , ppl: 9.239053003905902] tot_loss[loss=2.345, over 5458264.04 frames. , ppl: 10.430862270524031], batch size: 70
+2022-12-10 12:42:04,985 INFO [train.py:421] (2/8) Epoch 2, batch 43000, loss[loss=2.29, over 5950.00 frames. , ppl: 9.875485511648291] tot_loss[loss=2.344, over 5480765.40 frames. , ppl: 10.425149065946666], batch size: 70
+2022-12-10 12:42:04,985 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:42:05,730 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.261654238592559
+2022-12-10 12:43:44,225 INFO [train.py:421] (2/8) Epoch 2, batch 43200, loss[loss=2.476, over 1190.00 frames. , ppl: 11.895133975161745] tot_loss[loss=2.345, over 5458446.39 frames. , ppl: 10.430886083059956], batch size: 70
+2022-12-10 12:45:21,821 INFO [train.py:421] (2/8) Epoch 2, batch 43400, loss[loss=2.327, over 1890.00 frames. , ppl: 10.249262903762972] tot_loss[loss=2.343, over 5512148.09 frames. , ppl: 10.415828172197536], batch size: 70
+2022-12-10 12:47:02,721 INFO [train.py:421] (2/8) Epoch 2, batch 43600, loss[loss=2.265, over 8680.00 frames. , ppl: 9.631986295208526] tot_loss[loss=2.343, over 5533008.05 frames. , ppl: 10.411151923011307], batch size: 70
+2022-12-10 12:48:41,286 INFO [train.py:421] (2/8) Epoch 2, batch 43800, loss[loss=2.652, over 700.00 frames. , ppl: 14.178868966028052] tot_loss[loss=2.343, over 5523146.69 frames. , ppl: 10.412139658209137], batch size: 70
+2022-12-10 12:50:21,826 INFO [train.py:421] (2/8) Epoch 2, batch 44000, loss[loss=2.303, over 4200.00 frames. , ppl: 9.999851266106118] tot_loss[loss=2.343, over 5487985.10 frames. , ppl: 10.41645657872725], batch size: 70
+2022-12-10 12:50:21,826 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:50:22,572 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.270837981475065
+2022-12-10 12:52:01,332 INFO [train.py:421] (2/8) Epoch 2, batch 44200, loss[loss=2.192, over 6020.00 frames. , ppl: 8.954135337521516] tot_loss[loss=2.344, over 5430655.57 frames. , ppl: 10.423690636768885], batch size: 70
+2022-12-10 12:53:43,812 INFO [train.py:421] (2/8) Epoch 2, batch 44400, loss[loss=2.646, over 770.00 frames. , ppl: 14.093283959961472] tot_loss[loss=2.345, over 5425567.04 frames. , ppl: 10.428070350301729], batch size: 70
+2022-12-10 12:55:21,482 INFO [train.py:421] (2/8) Epoch 2, batch 44600, loss[loss=2.5, over 1680.00 frames. , ppl: 12.18365892772911] tot_loss[loss=2.345, over 5438458.32 frames. , ppl: 10.432248376675723], batch size: 70
+2022-12-10 12:57:01,753 INFO [train.py:421] (2/8) Epoch 2, batch 44800, loss[loss=2.347, over 2240.00 frames. , ppl: 10.453448181296702] tot_loss[loss=2.345, over 5451574.43 frames. , ppl: 10.43185749832749], batch size: 70
+2022-12-10 12:58:44,440 INFO [train.py:421] (2/8) Epoch 2, batch 45000, loss[loss=2.302, over 2870.00 frames. , ppl: 9.993927156067299] tot_loss[loss=2.344, over 5477971.33 frames. , ppl: 10.423244782225275], batch size: 70
+2022-12-10 12:58:44,440 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 12:58:45,186 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.267509369063996
+2022-12-10 13:00:27,490 INFO [train.py:421] (2/8) Epoch 2, batch 45200, loss[loss=2.312, over 2310.00 frames. , ppl: 10.094041974999845] tot_loss[loss=2.344, over 5517995.56 frames. , ppl: 10.417904894540342], batch size: 70
+2022-12-10 13:02:10,351 INFO [train.py:421] (2/8) Epoch 2, batch 45400, loss[loss=2.505, over 1260.00 frames. , ppl: 12.246719806923593] tot_loss[loss=2.342, over 5555084.81 frames. , ppl: 10.40249705038589], batch size: 70
+2022-12-10 13:03:49,609 INFO [train.py:421] (2/8) Epoch 2, batch 45600, loss[loss=2.279, over 4550.00 frames. , ppl: 9.767617597711823] tot_loss[loss=2.344, over 5472858.02 frames. , ppl: 10.422894655990255], batch size: 70
+2022-12-10 13:05:28,544 INFO [train.py:421] (2/8) Epoch 2, batch 45800, loss[loss=2.795, over 700.00 frames. , ppl: 16.354586950247533] tot_loss[loss=2.344, over 5466118.04 frames. , ppl: 10.426274636173872], batch size: 70
+2022-12-10 13:07:05,405 INFO [train.py:421] (2/8) Epoch 2, batch 46000, loss[loss=2.463, over 1120.00 frames. , ppl: 11.742239226884957] tot_loss[loss=2.344, over 5455756.66 frames. , ppl: 10.424530563676914], batch size: 70
+2022-12-10 13:07:05,405 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:07:06,169 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.269302733935248
+2022-12-10 13:08:48,309 INFO [train.py:421] (2/8) Epoch 2, batch 46200, loss[loss=2.442, over 1050.00 frames. , ppl: 11.496294195677118] tot_loss[loss=2.345, over 5463634.73 frames. , ppl: 10.428530562770653], batch size: 70
+2022-12-10 13:10:28,306 INFO [train.py:421] (2/8) Epoch 2, batch 46400, loss[loss=2.431, over 1610.00 frames. , ppl: 11.366753359030014] tot_loss[loss=2.344, over 5493775.23 frames. , ppl: 10.421226867505897], batch size: 70
+2022-12-10 13:12:09,472 INFO [train.py:421] (2/8) Epoch 2, batch 46600, loss[loss=2.463, over 1050.00 frames. , ppl: 11.740288799222887] tot_loss[loss=2.343, over 5523314.85 frames. , ppl: 10.414855880870364], batch size: 70
+2022-12-10 13:13:46,920 INFO [train.py:421] (2/8) Epoch 2, batch 46800, loss[loss=2.655, over 910.00 frames. , ppl: 14.224151062013332] tot_loss[loss=2.343, over 5532372.83 frames. , ppl: 10.409543269983356], batch size: 70
+2022-12-10 13:15:22,164 INFO [train.py:421] (2/8) Epoch 2, batch 47000, loss[loss=2.509, over 980.00 frames. , ppl: 12.293762207924013] tot_loss[loss=2.344, over 5495456.37 frames. , ppl: 10.42057856603507], batch size: 70
+2022-12-10 13:15:22,164 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:15:22,923 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.265565897614955
+2022-12-10 13:17:01,353 INFO [train.py:421] (2/8) Epoch 2, batch 47200, loss[loss=2.372, over 3360.00 frames. , ppl: 10.717705450267628] tot_loss[loss=2.344, over 5475204.93 frames. , ppl: 10.420380350484496], batch size: 70
+2022-12-10 13:18:45,164 INFO [train.py:421] (2/8) Epoch 2, batch 47400, loss[loss=2.382, over 2940.00 frames. , ppl: 10.828253695089664] tot_loss[loss=2.344, over 5494067.17 frames. , ppl: 10.417920914902798], batch size: 70
+2022-12-10 13:20:26,212 INFO [train.py:421] (2/8) Epoch 2, batch 47600, loss[loss=2.204, over 4060.00 frames. , ppl: 9.064624113854421] tot_loss[loss=2.342, over 5540056.71 frames. , ppl: 10.405233917178938], batch size: 70
+2022-12-10 13:22:08,963 INFO [train.py:421] (2/8) Epoch 2, batch 47800, loss[loss=2.327, over 5320.00 frames. , ppl: 10.24253639090687] tot_loss[loss=2.341, over 5589756.79 frames. , ppl: 10.396649211379248], batch size: 70
+2022-12-10 13:23:44,809 INFO [train.py:421] (2/8) Epoch 2, batch 48000, loss[loss=2.245, over 5530.00 frames. , ppl: 9.439184637682487] tot_loss[loss=2.342, over 5577393.74 frames. , ppl: 10.397812084667194], batch size: 70
+2022-12-10 13:23:44,809 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:23:45,556 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.26050912621634
+2022-12-10 13:25:25,350 INFO [train.py:421] (2/8) Epoch 2, batch 48200, loss[loss=2.35, over 8750.00 frames. , ppl: 10.480964575303588] tot_loss[loss=2.343, over 5541495.18 frames. , ppl: 10.409949324657607], batch size: 70
+2022-12-10 13:27:08,696 INFO [train.py:421] (2/8) Epoch 2, batch 48400, loss[loss=2.449, over 2520.00 frames. , ppl: 11.571034266656323] tot_loss[loss=2.343, over 5555177.38 frames. , ppl: 10.412280075500517], batch size: 70
+2022-12-10 13:28:49,067 INFO [train.py:421] (2/8) Epoch 2, batch 48600, loss[loss=2.353, over 2660.00 frames. , ppl: 10.521435628551208] tot_loss[loss=2.343, over 5547622.76 frames. , ppl: 10.411915663477544], batch size: 70
+2022-12-10 13:30:27,730 INFO [train.py:421] (2/8) Epoch 2, batch 48800, loss[loss=2.874, over 630.00 frames. , ppl: 17.705638594711985] tot_loss[loss=2.343, over 5529805.63 frames. , ppl: 10.411236030695605], batch size: 70
+2022-12-10 13:32:08,063 INFO [train.py:421] (2/8) Epoch 2, batch 49000, loss[loss=2.34, over 1960.00 frames. , ppl: 10.377264393669444] tot_loss[loss=2.344, over 5491316.58 frames. , ppl: 10.418405980525772], batch size: 70
+2022-12-10 13:32:08,064 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:32:08,810 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.253641134938565
+2022-12-10 13:33:44,602 INFO [train.py:421] (2/8) Epoch 2, batch 49200, loss[loss=2.387, over 2520.00 frames. , ppl: 10.879555498377606] tot_loss[loss=2.344, over 5430710.29 frames. , ppl: 10.427829201028608], batch size: 70
+2022-12-10 13:35:23,413 INFO [train.py:421] (2/8) Epoch 2, batch 49400, loss[loss=2.525, over 1120.00 frames. , ppl: 12.49148067912834] tot_loss[loss=2.344, over 5467092.39 frames. , ppl: 10.419929921324934], batch size: 70
+2022-12-10 13:37:03,013 INFO [train.py:421] (2/8) Epoch 2, batch 49600, loss[loss=2.282, over 5460.00 frames. , ppl: 9.791414646416342] tot_loss[loss=2.343, over 5487554.05 frames. , ppl: 10.4117357967455], batch size: 70
+2022-12-10 13:38:43,471 INFO [train.py:421] (2/8) Epoch 2, batch 49800, loss[loss=2.395, over 2730.00 frames. , ppl: 10.963334401039782] tot_loss[loss=2.342, over 5553883.28 frames. , ppl: 10.39992597519875], batch size: 70
+2022-12-10 13:40:22,316 INFO [train.py:421] (2/8) Epoch 2, batch 50000, loss[loss=2.421, over 980.00 frames. , ppl: 11.259613504894931] tot_loss[loss=2.342, over 5530940.26 frames. , ppl: 10.39816199040216], batch size: 70
+2022-12-10 13:40:22,316 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:40:23,076 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.327, over 211138.00 frames. , ppl: 10.24865242939489
+2022-12-10 13:42:06,023 INFO [train.py:421] (2/8) Epoch 2, batch 50200, loss[loss=2.547, over 1470.00 frames. , ppl: 12.76311170131353] tot_loss[loss=2.341, over 5544008.33 frames. , ppl: 10.390622801705552], batch size: 70
+2022-12-10 13:43:50,046 INFO [train.py:421] (2/8) Epoch 2, batch 50400, loss[loss=2.232, over 6230.00 frames. , ppl: 9.322492476863264] tot_loss[loss=2.34, over 5599466.49 frames. , ppl: 10.382099958285627], batch size: 70
+2022-12-10 13:45:26,740 INFO [train.py:421] (2/8) Epoch 2, batch 50600, loss[loss=2.482, over 1750.00 frames. , ppl: 11.966452896413932] tot_loss[loss=2.34, over 5618323.06 frames. , ppl: 10.38228098822767], batch size: 70
+2022-12-10 13:47:06,683 INFO [train.py:421] (2/8) Epoch 2, batch 50800, loss[loss=2.265, over 5460.00 frames. , ppl: 9.63302378239938] tot_loss[loss=2.34, over 5600056.26 frames. , ppl: 10.383381238713179], batch size: 70
+2022-12-10 13:48:47,679 INFO [train.py:421] (2/8) Epoch 2, batch 51000, loss[loss=2.369, over 2240.00 frames. , ppl: 10.691679853307548] tot_loss[loss=2.34, over 5631323.57 frames. , ppl: 10.377018489033453], batch size: 70
+2022-12-10 13:48:47,680 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:48:48,425 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.265193656171405
+2022-12-10 13:50:27,915 INFO [train.py:421] (2/8) Epoch 2, batch 51200, loss[loss=2.354, over 2800.00 frames. , ppl: 10.52248325080423] tot_loss[loss=2.342, over 5574058.25 frames. , ppl: 10.397007001858867], batch size: 70
+2022-12-10 13:52:05,819 INFO [train.py:421] (2/8) Epoch 2, batch 51400, loss[loss=2.84, over 700.00 frames. , ppl: 17.11389121512274] tot_loss[loss=2.342, over 5518760.48 frames. , ppl: 10.405506249251316], batch size: 70
+2022-12-10 13:53:44,960 INFO [train.py:421] (2/8) Epoch 2, batch 51600, loss[loss=2.307, over 2590.00 frames. , ppl: 10.049254135573175] tot_loss[loss=2.341, over 5552452.45 frames. , ppl: 10.392326974583247], batch size: 70
+2022-12-10 13:55:22,666 INFO [train.py:421] (2/8) Epoch 2, batch 51800, loss[loss=2.432, over 1960.00 frames. , ppl: 11.378210484584638] tot_loss[loss=2.342, over 5533079.73 frames. , ppl: 10.400191158569614], batch size: 70
+2022-12-10 13:57:00,788 INFO [train.py:421] (2/8) Epoch 2, batch 52000, loss[loss=2.265, over 3430.00 frames. , ppl: 9.627421287706897] tot_loss[loss=2.343, over 5510727.36 frames. , ppl: 10.41086327017319], batch size: 70
+2022-12-10 13:57:00,788 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 13:57:01,533 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.267428826974491
+2022-12-10 13:58:39,504 INFO [train.py:421] (2/8) Epoch 2, batch 52200, loss[loss=2.39, over 2450.00 frames. , ppl: 10.909820914258699] tot_loss[loss=2.343, over 5496973.72 frames. , ppl: 10.415951233684089], batch size: 70
+2022-12-10 14:00:16,926 INFO [train.py:421] (2/8) Epoch 2, batch 52400, loss[loss=2.311, over 2870.00 frames. , ppl: 10.080391793950819] tot_loss[loss=2.343, over 5499346.20 frames. , ppl: 10.417502395771347], batch size: 70
+2022-12-10 14:01:58,866 INFO [train.py:421] (2/8) Epoch 2, batch 52600, loss[loss=2.274, over 3920.00 frames. , ppl: 9.718869218387098] tot_loss[loss=2.344, over 5470111.17 frames. , ppl: 10.419436846421643], batch size: 70
+2022-12-10 14:03:38,030 INFO [train.py:421] (2/8) Epoch 2, batch 52800, loss[loss=3.087, over 560.00 frames. , ppl: 21.91362310623958] tot_loss[loss=2.343, over 5472517.07 frames. , ppl: 10.407521613157899], batch size: 70
+2022-12-10 14:05:17,808 INFO [train.py:421] (2/8) Epoch 2, batch 53000, loss[loss=2.273, over 6790.00 frames. , ppl: 9.706614423648897] tot_loss[loss=2.342, over 5473930.09 frames. , ppl: 10.403261442516483], batch size: 70
+2022-12-10 14:05:17,808 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:05:18,555 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.326, over 211138.00 frames. , ppl: 10.233852338507358
+2022-12-10 14:06:58,251 INFO [train.py:421] (2/8) Epoch 2, batch 53200, loss[loss=2.311, over 4830.00 frames. , ppl: 10.08460990005661] tot_loss[loss=2.342, over 5479782.47 frames. , ppl: 10.397554687304755], batch size: 70
+2022-12-10 14:08:37,567 INFO [train.py:421] (2/8) Epoch 2, batch 53400, loss[loss=2.455, over 1190.00 frames. , ppl: 11.646134401780921] tot_loss[loss=2.341, over 5513559.81 frames. , ppl: 10.391942164179714], batch size: 70
+2022-12-10 14:10:19,462 INFO [train.py:421] (2/8) Epoch 2, batch 53600, loss[loss=2.462, over 1890.00 frames. , ppl: 11.726879232887905] tot_loss[loss=2.34, over 5528850.79 frames. , ppl: 10.381128358641723], batch size: 70
+2022-12-10 14:11:57,897 INFO [train.py:421] (2/8) Epoch 2, batch 53800, loss[loss=2.338, over 2450.00 frames. , ppl: 10.356281788464308] tot_loss[loss=2.341, over 5522560.77 frames. , ppl: 10.38729631926569], batch size: 70
+2022-12-10 14:13:35,086 INFO [train.py:421] (2/8) Epoch 2, batch 54000, loss[loss=2.228, over 9450.00 frames. , ppl: 9.282450923279185] tot_loss[loss=2.342, over 5459363.85 frames. , ppl: 10.405317783919449], batch size: 70
+2022-12-10 14:13:35,086 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:13:35,833 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.325, over 211138.00 frames. , ppl: 10.222168662360373
+2022-12-10 14:15:13,891 INFO [train.py:421] (2/8) Epoch 2, batch 54200, loss[loss=2.572, over 1260.00 frames. , ppl: 13.089610303174736] tot_loss[loss=2.342, over 5454958.75 frames. , ppl: 10.403089696637013], batch size: 70
+2022-12-10 14:16:54,954 INFO [train.py:421] (2/8) Epoch 2, batch 54400, loss[loss=2.289, over 4410.00 frames. , ppl: 9.86543958371817] tot_loss[loss=2.341, over 5498535.54 frames. , ppl: 10.390754827528097], batch size: 70
+2022-12-10 14:18:34,362 INFO [train.py:421] (2/8) Epoch 2, batch 54600, loss[loss=3.173, over 560.00 frames. , ppl: 23.888255297593357] tot_loss[loss=2.34, over 5519460.05 frames. , ppl: 10.378284276062825], batch size: 70
+2022-12-10 14:20:12,829 INFO [train.py:421] (2/8) Epoch 2, batch 54800, loss[loss=2.773, over 770.00 frames. , ppl: 15.999538243034529] tot_loss[loss=2.339, over 5537022.07 frames. , ppl: 10.375362350407633], batch size: 70
+2022-12-10 14:21:50,925 INFO [train.py:421] (2/8) Epoch 2, batch 55000, loss[loss=2.346, over 5040.00 frames. , ppl: 10.438711779843175] tot_loss[loss=2.341, over 5500268.68 frames. , ppl: 10.390108528687714], batch size: 70
+2022-12-10 14:21:50,925 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:21:51,690 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.326, over 211138.00 frames. , ppl: 10.232852694220295
+2022-12-10 14:23:32,569 INFO [train.py:421] (2/8) Epoch 2, batch 55200, loss[loss=2.222, over 3500.00 frames. , ppl: 9.229971079352278] tot_loss[loss=2.34, over 5517941.81 frames. , ppl: 10.378859377839394], batch size: 70
+2022-12-10 14:25:08,970 INFO [train.py:421] (2/8) Epoch 2, batch 55400, loss[loss=2.948, over 560.00 frames. , ppl: 19.06228614322677] tot_loss[loss=2.34, over 5525028.30 frames. , ppl: 10.378065941430615], batch size: 70
+2022-12-10 14:26:48,412 INFO [train.py:421] (2/8) Epoch 2, batch 55600, loss[loss=2.247, over 5530.00 frames. , ppl: 9.45537409726143] tot_loss[loss=2.34, over 5504369.42 frames. , ppl: 10.381853380583948], batch size: 70
+2022-12-10 14:28:34,501 INFO [train.py:421] (2/8) Epoch 2, batch 55800, loss[loss=3.133, over 560.00 frames. , ppl: 22.94301378787986] tot_loss[loss=2.339, over 5542439.15 frames. , ppl: 10.374500956195734], batch size: 70
+2022-12-10 14:30:15,209 INFO [train.py:421] (2/8) Epoch 2, batch 56000, loss[loss=2.286, over 5810.00 frames. , ppl: 9.839581454242053] tot_loss[loss=2.339, over 5566841.01 frames. , ppl: 10.368487318116829], batch size: 70
+2022-12-10 14:30:15,210 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:30:15,958 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.327, over 211138.00 frames. , ppl: 10.2458754038067
+2022-12-10 14:31:57,749 INFO [train.py:421] (2/8) Epoch 2, batch 56200, loss[loss=2.673, over 910.00 frames. , ppl: 14.488731846159801] tot_loss[loss=2.339, over 5543955.61 frames. , ppl: 10.373434234034079], batch size: 70
+2022-12-10 14:33:41,513 INFO [train.py:421] (2/8) Epoch 2, batch 56400, loss[loss=2.441, over 3290.00 frames. , ppl: 11.478857472668984] tot_loss[loss=2.34, over 5556093.42 frames. , ppl: 10.381582900917683], batch size: 70
+2022-12-10 14:35:23,938 INFO [train.py:421] (2/8) Epoch 2, batch 56600, loss[loss=2.591, over 770.00 frames. , ppl: 13.345537836497924] tot_loss[loss=2.34, over 5538958.29 frames. , ppl: 10.379514253214513], batch size: 70
+2022-12-10 14:37:04,930 INFO [train.py:421] (2/8) Epoch 2, batch 56800, loss[loss=2.299, over 2660.00 frames. , ppl: 9.965145544155746] tot_loss[loss=2.34, over 5510965.40 frames. , ppl: 10.380695409222582], batch size: 70
+2022-12-10 14:38:48,284 INFO [train.py:421] (2/8) Epoch 2, batch 57000, loss[loss=2.303, over 5460.00 frames. , ppl: 9.999346550201514] tot_loss[loss=2.339, over 5525163.11 frames. , ppl: 10.37372735090927], batch size: 70
+2022-12-10 14:38:48,285 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:38:49,030 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.216746149309023
+2022-12-10 14:40:28,426 INFO [train.py:421] (2/8) Epoch 2, batch 57200, loss[loss=2.253, over 10010.00 frames. , ppl: 9.520900924676965] tot_loss[loss=2.34, over 5503665.54 frames. , ppl: 10.380994129772777], batch size: 70
+2022-12-10 14:42:06,873 INFO [train.py:421] (2/8) Epoch 2, batch 57400, loss[loss=2.513, over 1190.00 frames. , ppl: 12.344528539378272] tot_loss[loss=2.339, over 5529898.76 frames. , ppl: 10.372227811227656], batch size: 70
+2022-12-10 14:43:46,025 INFO [train.py:421] (2/8) Epoch 2, batch 57600, loss[loss=2.378, over 2380.00 frames. , ppl: 10.788238779790538] tot_loss[loss=2.339, over 5547395.00 frames. , ppl: 10.369685777453514], batch size: 70
+2022-12-10 14:45:24,392 INFO [train.py:421] (2/8) Epoch 2, batch 57800, loss[loss=2.715, over 910.00 frames. , ppl: 15.106221376756297] tot_loss[loss=2.339, over 5530867.54 frames. , ppl: 10.37550169013769], batch size: 70
+2022-12-10 14:47:02,164 INFO [train.py:421] (2/8) Epoch 2, batch 58000, loss[loss=2.498, over 700.00 frames. , ppl: 12.158135935795043] tot_loss[loss=2.341, over 5496862.85 frames. , ppl: 10.387525273330507], batch size: 70
+2022-12-10 14:47:02,165 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:47:02,926 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.325, over 211138.00 frames. , ppl: 10.22471528659933
+2022-12-10 14:48:40,205 INFO [train.py:421] (2/8) Epoch 2, batch 58200, loss[loss=3.395, over 490.00 frames. , ppl: 29.814908982669422] tot_loss[loss=2.342, over 5481903.46 frames. , ppl: 10.397149357597106], batch size: 70
+2022-12-10 14:50:20,524 INFO [train.py:421] (2/8) Epoch 2, batch 58400, loss[loss=2.668, over 910.00 frames. , ppl: 14.404019468509698] tot_loss[loss=2.342, over 5463946.01 frames. , ppl: 10.398462452279151], batch size: 70
+2022-12-10 14:51:59,910 INFO [train.py:421] (2/8) Epoch 2, batch 58600, loss[loss=2.309, over 4270.00 frames. , ppl: 10.065192401098276] tot_loss[loss=2.342, over 5441037.18 frames. , ppl: 10.402765527337953], batch size: 70
+2022-12-10 14:53:41,810 INFO [train.py:421] (2/8) Epoch 2, batch 58800, loss[loss=2.736, over 910.00 frames. , ppl: 15.42111840354121] tot_loss[loss=2.341, over 5453779.84 frames. , ppl: 10.394720369699602], batch size: 70
+2022-12-10 14:55:21,055 INFO [train.py:421] (2/8) Epoch 2, batch 59000, loss[loss=2.786, over 630.00 frames. , ppl: 16.22244767160564] tot_loss[loss=2.341, over 5469677.69 frames. , ppl: 10.38916975749762], batch size: 70
+2022-12-10 14:55:21,055 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 14:55:21,801 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.325, over 211138.00 frames. , ppl: 10.225733811230693
+2022-12-10 14:57:02,577 INFO [train.py:421] (2/8) Epoch 2, batch 59200, loss[loss=2.574, over 840.00 frames. , ppl: 13.121005120109155] tot_loss[loss=2.34, over 5457906.42 frames. , ppl: 10.38628181588914], batch size: 70
+2022-12-10 14:58:41,089 INFO [train.py:421] (2/8) Epoch 2, batch 59400, loss[loss=2.463, over 1540.00 frames. , ppl: 11.735395674638804] tot_loss[loss=2.342, over 5444874.62 frames. , ppl: 10.398136990466291], batch size: 70
+2022-12-10 15:00:19,436 INFO [train.py:421] (2/8) Epoch 2, batch 59600, loss[loss=2.481, over 1750.00 frames. , ppl: 11.947523226745592] tot_loss[loss=2.342, over 5417134.15 frames. , ppl: 10.401582361382639], batch size: 70
+2022-12-10 15:02:02,511 INFO [train.py:421] (2/8) Epoch 2, batch 59800, loss[loss=2.395, over 2100.00 frames. , ppl: 10.963996020459604] tot_loss[loss=2.341, over 5455807.64 frames. , ppl: 10.390748368027428], batch size: 70
+2022-12-10 15:03:44,826 INFO [train.py:421] (2/8) Epoch 2, batch 60000, loss[loss=2.394, over 1540.00 frames. , ppl: 10.959334008791533] tot_loss[loss=2.339, over 5525100.36 frames. , ppl: 10.373461485733523], batch size: 70
+2022-12-10 15:03:44,827 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:03:45,573 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.21343657885031
+2022-12-10 15:05:24,002 INFO [train.py:421] (2/8) Epoch 2, batch 60200, loss[loss=2.402, over 2870.00 frames. , ppl: 11.047959535418839] tot_loss[loss=2.339, over 5527299.04 frames. , ppl: 10.369122946953853], batch size: 70
+2022-12-10 15:07:02,397 INFO [train.py:421] (2/8) Epoch 2, batch 60400, loss[loss=2.302, over 1960.00 frames. , ppl: 9.992514735640649] tot_loss[loss=2.34, over 5478714.92 frames. , ppl: 10.384582475813714], batch size: 70
+2022-12-10 15:08:40,165 INFO [train.py:421] (2/8) Epoch 2, batch 60600, loss[loss=2.547, over 980.00 frames. , ppl: 12.763143497210969] tot_loss[loss=2.34, over 5484221.69 frames. , ppl: 10.382453476408491], batch size: 70
+2022-12-10 15:10:20,017 INFO [train.py:421] (2/8) Epoch 2, batch 60800, loss[loss=2.572, over 1050.00 frames. , ppl: 13.095131414656915] tot_loss[loss=2.341, over 5455898.04 frames. , ppl: 10.386453846516547], batch size: 70
+2022-12-10 15:11:55,175 INFO [train.py:421] (2/8) Epoch 2, batch 61000, loss[loss=2.261, over 2730.00 frames. , ppl: 9.590550748799616] tot_loss[loss=2.342, over 5413244.60 frames. , ppl: 10.400007409284965], batch size: 70
+2022-12-10 15:11:55,175 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:11:55,926 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.211778417126597
+2022-12-10 15:13:35,397 INFO [train.py:421] (2/8) Epoch 2, batch 61200, loss[loss=2.359, over 3010.00 frames. , ppl: 10.581539289557549] tot_loss[loss=2.341, over 5428383.53 frames. , ppl: 10.395479028176672], batch size: 70
+2022-12-10 15:15:14,571 INFO [train.py:421] (2/8) Epoch 2, batch 61400, loss[loss=2.565, over 1470.00 frames. , ppl: 12.994603722877455] tot_loss[loss=2.341, over 5452745.45 frames. , ppl: 10.388239454431412], batch size: 70
+2022-12-10 15:16:52,161 INFO [train.py:421] (2/8) Epoch 2, batch 61600, loss[loss=2.237, over 8330.00 frames. , ppl: 9.364500713765885] tot_loss[loss=2.339, over 5501318.46 frames. , ppl: 10.374449548629368], batch size: 70
+2022-12-10 15:18:32,329 INFO [train.py:421] (2/8) Epoch 2, batch 61800, loss[loss=2.179, over 5670.00 frames. , ppl: 8.834158493121794] tot_loss[loss=2.34, over 5469765.98 frames. , ppl: 10.380476413984047], batch size: 70
+2022-12-10 15:20:14,992 INFO [train.py:421] (2/8) Epoch 2, batch 62000, loss[loss=2.411, over 3010.00 frames. , ppl: 11.139664974978698] tot_loss[loss=2.34, over 5454116.53 frames. , ppl: 10.386367163709416], batch size: 70
+2022-12-10 15:20:14,993 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:20:15,753 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.219063032959546
+2022-12-10 15:21:54,084 INFO [train.py:421] (2/8) Epoch 2, batch 62200, loss[loss=2.281, over 2380.00 frames. , ppl: 9.790427289571875] tot_loss[loss=2.339, over 5503167.68 frames. , ppl: 10.371160607194092], batch size: 70
+2022-12-10 15:23:35,865 INFO [train.py:421] (2/8) Epoch 2, batch 62400, loss[loss=2.341, over 1470.00 frames. , ppl: 10.395859258628255] tot_loss[loss=2.338, over 5509894.01 frames. , ppl: 10.363363287921954], batch size: 70
+2022-12-10 15:25:13,326 INFO [train.py:421] (2/8) Epoch 2, batch 62600, loss[loss=2.677, over 1050.00 frames. , ppl: 14.5389301862712] tot_loss[loss=2.339, over 5491581.58 frames. , ppl: 10.367415084882193], batch size: 70
+2022-12-10 15:26:54,725 INFO [train.py:421] (2/8) Epoch 2, batch 62800, loss[loss=2.246, over 8540.00 frames. , ppl: 9.446204455974591] tot_loss[loss=2.338, over 5530625.00 frames. , ppl: 10.35582860919285], batch size: 70
+2022-12-10 15:28:32,241 INFO [train.py:421] (2/8) Epoch 2, batch 63000, loss[loss=2.573, over 910.00 frames. , ppl: 13.099339553683174] tot_loss[loss=2.339, over 5512590.26 frames. , ppl: 10.370644333225174], batch size: 70
+2022-12-10 15:28:32,241 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:28:33,000 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.20283624301792
+2022-12-10 15:30:14,780 INFO [train.py:421] (2/8) Epoch 2, batch 63200, loss[loss=2.379, over 3570.00 frames. , ppl: 10.795412311147192] tot_loss[loss=2.338, over 5511038.66 frames. , ppl: 10.364054044181941], batch size: 70
+2022-12-10 15:31:59,064 INFO [train.py:421] (2/8) Epoch 2, batch 63400, loss[loss=2.448, over 1050.00 frames. , ppl: 11.568847147569903] tot_loss[loss=2.338, over 5497953.90 frames. , ppl: 10.364740645047277], batch size: 70
+2022-12-10 15:33:37,718 INFO [train.py:421] (2/8) Epoch 2, batch 63600, loss[loss=2.626, over 1120.00 frames. , ppl: 13.823933627296968] tot_loss[loss=2.338, over 5486062.18 frames. , ppl: 10.365575202969463], batch size: 70
+2022-12-10 15:35:17,819 INFO [train.py:421] (2/8) Epoch 2, batch 63800, loss[loss=2.26, over 5040.00 frames. , ppl: 9.580964060656772] tot_loss[loss=2.338, over 5516440.22 frames. , ppl: 10.359988172026382], batch size: 70
+2022-12-10 15:36:57,203 INFO [train.py:421] (2/8) Epoch 2, batch 64000, loss[loss=3.084, over 630.00 frames. , ppl: 21.843512432371284] tot_loss[loss=2.337, over 5529114.40 frames. , ppl: 10.351134815925025], batch size: 70
+2022-12-10 15:36:57,203 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:36:57,966 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.220371427142487
+2022-12-10 15:38:37,720 INFO [train.py:421] (2/8) Epoch 2, batch 64200, loss[loss=2.431, over 3500.00 frames. , ppl: 11.370963651441361] tot_loss[loss=2.339, over 5476340.93 frames. , ppl: 10.370485381856147], batch size: 70
+2022-12-10 15:40:17,493 INFO [train.py:421] (2/8) Epoch 2, batch 64400, loss[loss=2.444, over 2100.00 frames. , ppl: 11.520486745884055] tot_loss[loss=2.338, over 5516869.29 frames. , ppl: 10.361817602745488], batch size: 70
+2022-12-10 15:41:55,772 INFO [train.py:421] (2/8) Epoch 2, batch 64600, loss[loss=2.342, over 1330.00 frames. , ppl: 10.401734240205627] tot_loss[loss=2.338, over 5509703.69 frames. , ppl: 10.363529145224717], batch size: 70
+2022-12-10 15:43:33,788 INFO [train.py:421] (2/8) Epoch 2, batch 64800, loss[loss=2.332, over 2800.00 frames. , ppl: 10.297297192321189] tot_loss[loss=2.339, over 5492344.26 frames. , ppl: 10.370714222232754], batch size: 70
+2022-12-10 15:45:14,653 INFO [train.py:421] (2/8) Epoch 2, batch 65000, loss[loss=2.573, over 1050.00 frames. , ppl: 13.105216671845097] tot_loss[loss=2.339, over 5441331.66 frames. , ppl: 10.372461208198034], batch size: 70
+2022-12-10 15:45:14,654 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:45:15,399 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.211031803223284
+2022-12-10 15:46:58,768 INFO [train.py:421] (2/8) Epoch 2, batch 65200, loss[loss=2.483, over 1750.00 frames. , ppl: 11.979264254744376] tot_loss[loss=2.338, over 5453082.76 frames. , ppl: 10.36531926019529], batch size: 70
+2022-12-10 15:48:38,204 INFO [train.py:421] (2/8) Epoch 2, batch 65400, loss[loss=2.607, over 700.00 frames. , ppl: 13.553190294223691] tot_loss[loss=2.338, over 5441605.24 frames. , ppl: 10.36399160988631], batch size: 70
+2022-12-10 15:50:19,932 INFO [train.py:421] (2/8) Epoch 2, batch 65600, loss[loss=2.333, over 2100.00 frames. , ppl: 10.306869759780511] tot_loss[loss=2.337, over 5462992.05 frames. , ppl: 10.355244558767666], batch size: 70
+2022-12-10 15:51:59,509 INFO [train.py:421] (2/8) Epoch 2, batch 65800, loss[loss=2.252, over 4970.00 frames. , ppl: 9.509524175390247] tot_loss[loss=2.337, over 5486433.61 frames. , ppl: 10.353588154081981], batch size: 70
+2022-12-10 15:53:41,829 INFO [train.py:421] (2/8) Epoch 2, batch 66000, loss[loss=2.307, over 3850.00 frames. , ppl: 10.041597870304386] tot_loss[loss=2.337, over 5476037.02 frames. , ppl: 10.34906554440258], batch size: 70
+2022-12-10 15:53:41,830 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 15:53:42,595 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.321, over 211138.00 frames. , ppl: 10.187981682837393
+2022-12-10 15:55:24,763 INFO [train.py:421] (2/8) Epoch 2, batch 66200, loss[loss=2.407, over 3150.00 frames. , ppl: 11.098743539171606] tot_loss[loss=2.337, over 5469868.33 frames. , ppl: 10.354653120514204], batch size: 70
+2022-12-10 15:57:04,875 INFO [train.py:421] (2/8) Epoch 2, batch 66400, loss[loss=2.512, over 1470.00 frames. , ppl: 12.326684888695029] tot_loss[loss=2.336, over 5474107.31 frames. , ppl: 10.34149544267806], batch size: 70
+2022-12-10 15:58:43,335 INFO [train.py:421] (2/8) Epoch 2, batch 66600, loss[loss=2.259, over 2170.00 frames. , ppl: 9.577319263379495] tot_loss[loss=2.336, over 5488734.34 frames. , ppl: 10.341571328927406], batch size: 70
+2022-12-10 16:00:20,766 INFO [train.py:421] (2/8) Epoch 2, batch 66800, loss[loss=2.293, over 3570.00 frames. , ppl: 9.905020868636656] tot_loss[loss=2.335, over 5537016.13 frames. , ppl: 10.325158886356261], batch size: 70
+2022-12-10 16:02:02,722 INFO [train.py:421] (2/8) Epoch 2, batch 67000, loss[loss=3.084, over 560.00 frames. , ppl: 21.85455663105156] tot_loss[loss=2.336, over 5495502.64 frames. , ppl: 10.337812314487271], batch size: 70
+2022-12-10 16:02:02,722 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 16:02:03,483 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.203262098954827
+2022-12-10 16:03:41,290 INFO [train.py:421] (2/8) Epoch 2, batch 67200, loss[loss=2.235, over 3080.00 frames. , ppl: 9.345298918121966] tot_loss[loss=2.337, over 5459937.44 frames. , ppl: 10.354634693962385], batch size: 70
+2022-12-10 16:05:20,731 INFO [train.py:421] (2/8) Epoch 2, batch 67400, loss[loss=2.456, over 840.00 frames. , ppl: 11.66047199091334] tot_loss[loss=2.338, over 5440594.89 frames. , ppl: 10.358396320918978], batch size: 70
+2022-12-10 16:07:01,225 INFO [train.py:421] (2/8) Epoch 2, batch 67600, loss[loss=2.418, over 1540.00 frames. , ppl: 11.225233341041402] tot_loss[loss=2.337, over 5470492.27 frames. , ppl: 10.351971006240982], batch size: 70
+2022-12-10 16:08:44,331 INFO [train.py:421] (2/8) Epoch 2, batch 67800, loss[loss=2.498, over 1400.00 frames. , ppl: 12.155047181294917] tot_loss[loss=2.338, over 5483278.38 frames. , ppl: 10.360140416368811], batch size: 70
+2022-12-10 16:10:23,873 INFO [train.py:421] (2/8) Epoch 2, batch 68000, loss[loss=2.347, over 1330.00 frames. , ppl: 10.454098822042257] tot_loss[loss=2.339, over 5444770.02 frames. , ppl: 10.367768306809445], batch size: 70
+2022-12-10 16:10:23,874 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 16:10:24,633 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.201776210751579
+2022-12-10 16:12:06,495 INFO [train.py:421] (2/8) Epoch 2, batch 68200, loss[loss=2.243, over 5670.00 frames. , ppl: 9.426095924827559] tot_loss[loss=2.337, over 5498454.80 frames. , ppl: 10.351314350728464], batch size: 70
+2022-12-10 16:13:47,728 INFO [train.py:421] (2/8) Epoch 2, batch 68400, loss[loss=2.426, over 1470.00 frames. , ppl: 11.318534375478302] tot_loss[loss=2.337, over 5475251.44 frames. , ppl: 10.35398641880803], batch size: 70
+2022-12-10 16:15:28,292 INFO [train.py:421] (2/8) Epoch 2, batch 68600, loss[loss=2.222, over 5390.00 frames. , ppl: 9.230151220199591] tot_loss[loss=2.337, over 5503492.84 frames. , ppl: 10.347282256624258], batch size: 70
+2022-12-10 16:17:07,753 INFO [train.py:421] (2/8) Epoch 2, batch 68800, loss[loss=2.378, over 2310.00 frames. , ppl: 10.785962120375848] tot_loss[loss=2.337, over 5500930.63 frames. , ppl: 10.346594205096041], batch size: 70
+2022-12-10 16:18:50,958 INFO [train.py:421] (2/8) Epoch 2, batch 69000, loss[loss=2.569, over 1610.00 frames. , ppl: 13.050626386722811] tot_loss[loss=2.337, over 5523396.68 frames. , ppl: 10.351021311807374], batch size: 70
+2022-12-10 16:18:50,959 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 16:18:51,729 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.32, over 211138.00 frames. , ppl: 10.176567267443412
+2022-12-10 16:20:33,667 INFO [train.py:421] (2/8) Epoch 2, batch 69200, loss[loss=2.344, over 4060.00 frames. , ppl: 10.426662679542087] tot_loss[loss=2.337, over 5515422.75 frames. , ppl: 10.350597201149712], batch size: 70
+2022-12-10 16:22:12,409 INFO [train.py:421] (2/8) Epoch 2, batch 69400, loss[loss=2.279, over 2870.00 frames. , ppl: 9.771058625523798] tot_loss[loss=2.336, over 5527497.58 frames. , ppl: 10.340139551676472], batch size: 70
+2022-12-10 16:23:54,608 INFO [train.py:421] (2/8) Epoch 2, batch 69600, loss[loss=2.317, over 4060.00 frames. , ppl: 10.144142603490016] tot_loss[loss=2.335, over 5542739.35 frames. , ppl: 10.332016104911943], batch size: 70
+2022-12-10 16:25:37,452 INFO [train.py:421] (2/8) Epoch 2, batch 69800, loss[loss=2.322, over 3920.00 frames. , ppl: 10.19947630124195] tot_loss[loss=2.336, over 5505602.78 frames. , ppl: 10.336002838780841], batch size: 70
+2022-12-10 16:27:18,878 INFO [train.py:421] (2/8) Epoch 2, batch 70000, loss[loss=2.247, over 3990.00 frames. , ppl: 9.456542736685133] tot_loss[loss=2.335, over 5522119.84 frames. , ppl: 10.331288587324273], batch size: 70
+2022-12-10 16:27:18,878 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 16:27:19,638 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.322, over 211138.00 frames. , ppl: 10.191266412638488
+2022-12-10 16:29:02,824 INFO [train.py:421] (2/8) Epoch 2, batch 70200, loss[loss=2.228, over 3780.00 frames. , ppl: 9.27787308221377] tot_loss[loss=2.334, over 5565199.14 frames. , ppl: 10.320330436360898], batch size: 70
+2022-12-10 16:30:41,310 INFO [train.py:421] (2/8) Epoch 2, batch 70400, loss[loss=2.242, over 4620.00 frames. , ppl: 9.411665326543398] tot_loss[loss=2.335, over 5550914.77 frames. , ppl: 10.326965866057172], batch size: 70
+2022-12-10 16:32:24,246 INFO [train.py:421] (2/8) Epoch 2, batch 70600, loss[loss=2.628, over 980.00 frames. , ppl: 13.850271936828188] tot_loss[loss=2.335, over 5530227.69 frames. , ppl: 10.32682344446933], batch size: 70
+2022-12-10 16:34:01,869 INFO [train.py:421] (2/8) Epoch 2, batch 70800, loss[loss=2.544, over 1260.00 frames. , ppl: 12.725920678551569] tot_loss[loss=2.336, over 5485980.45 frames. , ppl: 10.336751755272344], batch size: 70
+2022-12-10 16:35:42,854 INFO [train.py:421] (2/8) Epoch 2, batch 71000, loss[loss=2.389, over 3010.00 frames. , ppl: 10.907410476899734] tot_loss[loss=2.336, over 5462050.99 frames. , ppl: 10.339576052373806], batch size: 70
+2022-12-10 16:35:42,854 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 16:35:43,613 INFO [train.py:452] (2/8) Epoch 2, validation: loss=2.319, over 211138.00 frames. , ppl: 10.169496572570102
+2022-12-10 16:37:26,616 INFO [train.py:421] (2/8) Epoch 2, batch 71200, loss[loss=2.261, over 2590.00 frames. , ppl: 9.589770345895488] tot_loss[loss=2.335, over 5520572.79 frames. , ppl: 10.324859045177542], batch size: 70
+2022-12-10 16:39:05,532 INFO [train.py:421] (2/8) Epoch 2, batch 71400, loss[loss=2.254, over 4130.00 frames. , ppl: 9.526559092389737] tot_loss[loss=2.336, over 5465249.70 frames. , ppl: 10.338602447505387], batch size: 70
+2022-12-10 16:40:47,631 INFO [train.py:421] (2/8) Epoch 2, batch 71600, loss[loss=2.383, over 2730.00 frames. , ppl: 10.833218357961323] tot_loss[loss=2.335, over 5519822.74 frames. , ppl: 10.325084891135694], batch size: 70
+2022-12-10 16:42:29,323 INFO [train.py:421] (2/8) Epoch 2, batch 71800, loss[loss=2.26, over 4200.00 frames. , ppl: 9.584221164722557] tot_loss[loss=2.335, over 5531535.85 frames. , ppl: 10.324324042878366], batch size: 70
+2022-12-10 16:43:45,587 INFO [train.py:421] (2/8) Epoch 3, batch 0, loss[loss=2.609, over 840.00 frames. , ppl: 13.583430154655446] tot_loss[loss=2.609, over 840.00 frames. , ppl: 13.583430154655446], batch size: 70
+2022-12-10 16:45:28,209 INFO [train.py:421] (2/8) Epoch 3, batch 200, loss[loss=2.21, over 7490.00 frames. , ppl: 9.116061547798305] tot_loss[loss=2.316, over 573161.92 frames. , ppl: 10.131972028326242], batch size: 70
+2022-12-10 16:47:09,531 INFO [train.py:421] (2/8) Epoch 3, batch 400, loss[loss=2.359, over 1680.00 frames. , ppl: 10.5835497563637] tot_loss[loss=2.325, over 1022734.22 frames. , ppl: 10.225256823660398], batch size: 70
+2022-12-10 16:48:48,730 INFO [train.py:421] (2/8) Epoch 3, batch 600, loss[loss=2.728, over 700.00 frames. , ppl: 15.299572417165871] tot_loss[loss=2.323, over 1483837.58 frames. , ppl: 10.206947237708956], batch size: 70
+2022-12-10 16:50:28,168 INFO [train.py:421] (2/8) Epoch 3, batch 800, loss[loss=3.02, over 560.00 frames. , ppl: 20.492442350553475] tot_loss[loss=2.321, over 1890757.68 frames. , ppl: 10.189808080272808], batch size: 70
+2022-12-10 16:52:09,310 INFO [train.py:421] (2/8) Epoch 3, batch 1000, loss[loss=2.412, over 1680.00 frames. , ppl: 11.15905911874444] tot_loss[loss=2.321, over 2256304.16 frames. , ppl: 10.185689544334645], batch size: 70
+2022-12-10 16:52:09,311 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 16:52:10,090 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.164404381725806
+2022-12-10 16:53:53,057 INFO [train.py:421] (2/8) Epoch 3, batch 1200, loss[loss=2.296, over 1960.00 frames. , ppl: 9.936039363875917] tot_loss[loss=2.325, over 2526248.13 frames. , ppl: 10.225942553428075], batch size: 70
+2022-12-10 16:55:35,512 INFO [train.py:421] (2/8) Epoch 3, batch 1400, loss[loss=2.533, over 910.00 frames. , ppl: 12.5865424951964] tot_loss[loss=2.327, over 2784579.37 frames. , ppl: 10.244164045447224], batch size: 70
+2022-12-10 16:57:16,482 INFO [train.py:421] (2/8) Epoch 3, batch 1600, loss[loss=2.272, over 4970.00 frames. , ppl: 9.702615669884063] tot_loss[loss=2.326, over 3031609.81 frames. , ppl: 10.236132655321821], batch size: 70
+2022-12-10 16:58:58,235 INFO [train.py:421] (2/8) Epoch 3, batch 1800, loss[loss=2.226, over 3010.00 frames. , ppl: 9.25893339088627] tot_loss[loss=2.324, over 3279998.16 frames. , ppl: 10.218541431282715], batch size: 70
+2022-12-10 17:00:33,685 INFO [train.py:421] (2/8) Epoch 3, batch 2000, loss[loss=2.636, over 840.00 frames. , ppl: 13.961242668588298] tot_loss[loss=2.326, over 3457769.33 frames. , ppl: 10.236367038602976], batch size: 70
+2022-12-10 17:00:33,686 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:00:34,445 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.321, over 211138.00 frames. , ppl: 10.184881917574328
+2022-12-10 17:02:13,490 INFO [train.py:421] (2/8) Epoch 3, batch 2200, loss[loss=2.227, over 6580.00 frames. , ppl: 9.26741624621157] tot_loss[loss=2.327, over 3635266.90 frames. , ppl: 10.251741923437697], batch size: 70
+2022-12-10 17:03:54,052 INFO [train.py:421] (2/8) Epoch 3, batch 2400, loss[loss=2.469, over 1610.00 frames. , ppl: 11.81054678370656] tot_loss[loss=2.328, over 3821228.12 frames. , ppl: 10.25298663167366], batch size: 70
+2022-12-10 17:05:32,316 INFO [train.py:421] (2/8) Epoch 3, batch 2600, loss[loss=2.251, over 3710.00 frames. , ppl: 9.494891096761368] tot_loss[loss=2.327, over 3994463.93 frames. , ppl: 10.245093499056152], batch size: 70
+2022-12-10 17:07:12,597 INFO [train.py:421] (2/8) Epoch 3, batch 2800, loss[loss=2.423, over 2450.00 frames. , ppl: 11.28283838607687] tot_loss[loss=2.327, over 4129951.53 frames. , ppl: 10.244124065706353], batch size: 70
+2022-12-10 17:08:55,682 INFO [train.py:421] (2/8) Epoch 3, batch 3000, loss[loss=2.431, over 1120.00 frames. , ppl: 11.37270777977079] tot_loss[loss=2.326, over 4258048.27 frames. , ppl: 10.23913915218522], batch size: 70
+2022-12-10 17:08:55,682 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:08:56,443 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.32, over 211138.00 frames. , ppl: 10.171098190408838
+2022-12-10 17:10:38,822 INFO [train.py:421] (2/8) Epoch 3, batch 3200, loss[loss=2.514, over 980.00 frames. , ppl: 12.36024793137721] tot_loss[loss=2.326, over 4402528.85 frames. , ppl: 10.233507376074694], batch size: 70
+2022-12-10 17:12:18,604 INFO [train.py:421] (2/8) Epoch 3, batch 3400, loss[loss=2.263, over 5180.00 frames. , ppl: 9.609813944204578] tot_loss[loss=2.327, over 4471715.60 frames. , ppl: 10.246111979385358], batch size: 70
+2022-12-10 17:13:58,760 INFO [train.py:421] (2/8) Epoch 3, batch 3600, loss[loss=2.425, over 1050.00 frames. , ppl: 11.307312991438888] tot_loss[loss=2.328, over 4522101.61 frames. , ppl: 10.259386486488303], batch size: 70
+2022-12-10 17:15:38,952 INFO [train.py:421] (2/8) Epoch 3, batch 3800, loss[loss=2.369, over 2730.00 frames. , ppl: 10.687740548382918] tot_loss[loss=2.326, over 4669876.13 frames. , ppl: 10.241614638716054], batch size: 70
+2022-12-10 17:17:17,264 INFO [train.py:421] (2/8) Epoch 3, batch 4000, loss[loss=2.541, over 1400.00 frames. , ppl: 12.69251103533866] tot_loss[loss=2.326, over 4752089.87 frames. , ppl: 10.238916998668017], batch size: 70
+2022-12-10 17:17:17,265 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:17:18,024 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.163163321129922
+2022-12-10 17:18:57,020 INFO [train.py:421] (2/8) Epoch 3, batch 4200, loss[loss=2.551, over 700.00 frames. , ppl: 12.818034147404871] tot_loss[loss=2.327, over 4801025.54 frames. , ppl: 10.250319703024571], batch size: 70
+2022-12-10 17:20:36,279 INFO [train.py:421] (2/8) Epoch 3, batch 4400, loss[loss=2.47, over 1260.00 frames. , ppl: 11.818637497267638] tot_loss[loss=2.326, over 4900196.41 frames. , ppl: 10.240218269678554], batch size: 70
+2022-12-10 17:22:15,217 INFO [train.py:421] (2/8) Epoch 3, batch 4600, loss[loss=2.471, over 1120.00 frames. , ppl: 11.830086779780986] tot_loss[loss=2.328, over 4900816.14 frames. , ppl: 10.261376092161994], batch size: 70
+2022-12-10 17:23:55,351 INFO [train.py:421] (2/8) Epoch 3, batch 4800, loss[loss=2.683, over 630.00 frames. , ppl: 14.622162046752813] tot_loss[loss=2.328, over 4990228.03 frames. , ppl: 10.252606245215432], batch size: 70
+2022-12-10 17:25:31,817 INFO [train.py:421] (2/8) Epoch 3, batch 5000, loss[loss=2.877, over 630.00 frames. , ppl: 17.766197491526448] tot_loss[loss=2.328, over 5036266.71 frames. , ppl: 10.255117272417783], batch size: 70
+2022-12-10 17:25:31,818 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:25:32,565 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.167579177724292
+2022-12-10 17:27:12,749 INFO [train.py:421] (2/8) Epoch 3, batch 5200, loss[loss=2.356, over 2660.00 frames. , ppl: 10.543725865516477] tot_loss[loss=2.327, over 5075980.81 frames. , ppl: 10.249509167716697], batch size: 70
+2022-12-10 17:28:53,461 INFO [train.py:421] (2/8) Epoch 3, batch 5400, loss[loss=3.028, over 560.00 frames. , ppl: 20.662118549099638] tot_loss[loss=2.327, over 5138523.16 frames. , ppl: 10.246362610541171], batch size: 70
+2022-12-10 17:30:38,125 INFO [train.py:421] (2/8) Epoch 3, batch 5600, loss[loss=2.244, over 4690.00 frames. , ppl: 9.429162318274063] tot_loss[loss=2.327, over 5156411.92 frames. , ppl: 10.249209765629256], batch size: 70
+2022-12-10 17:32:21,540 INFO [train.py:421] (2/8) Epoch 3, batch 5800, loss[loss=2.268, over 5740.00 frames. , ppl: 9.663633856441908] tot_loss[loss=2.328, over 5157720.07 frames. , ppl: 10.253265111662111], batch size: 70
+2022-12-10 17:34:02,533 INFO [train.py:421] (2/8) Epoch 3, batch 6000, loss[loss=2.292, over 2800.00 frames. , ppl: 9.89104165631127] tot_loss[loss=2.327, over 5229578.96 frames. , ppl: 10.249159289524737], batch size: 70
+2022-12-10 17:34:02,534 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:34:03,279 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.146697899157441
+2022-12-10 17:35:42,385 INFO [train.py:421] (2/8) Epoch 3, batch 6200, loss[loss=2.43, over 2310.00 frames. , ppl: 11.357001285837566] tot_loss[loss=2.328, over 5244597.07 frames. , ppl: 10.253977671776495], batch size: 70
+2022-12-10 17:37:19,186 INFO [train.py:421] (2/8) Epoch 3, batch 6400, loss[loss=2.447, over 1120.00 frames. , ppl: 11.552640423121858] tot_loss[loss=2.329, over 5205208.84 frames. , ppl: 10.268617025778923], batch size: 70
+2022-12-10 17:38:59,188 INFO [train.py:421] (2/8) Epoch 3, batch 6600, loss[loss=2.5, over 1050.00 frames. , ppl: 12.18411998799907] tot_loss[loss=2.328, over 5280140.51 frames. , ppl: 10.25480985909732], batch size: 70
+2022-12-10 17:40:37,955 INFO [train.py:421] (2/8) Epoch 3, batch 6800, loss[loss=2.313, over 10010.00 frames. , ppl: 10.106171014183618] tot_loss[loss=2.329, over 5280946.48 frames. , ppl: 10.265872581430118], batch size: 70
+2022-12-10 17:42:18,035 INFO [train.py:421] (2/8) Epoch 3, batch 7000, loss[loss=2.494, over 1540.00 frames. , ppl: 12.113098968533171] tot_loss[loss=2.329, over 5317642.08 frames. , ppl: 10.26721557786411], batch size: 70
+2022-12-10 17:42:18,035 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:42:18,803 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.159209485439936
+2022-12-10 17:44:00,386 INFO [train.py:421] (2/8) Epoch 3, batch 7200, loss[loss=2.473, over 1120.00 frames. , ppl: 11.853576733754274] tot_loss[loss=2.328, over 5379246.79 frames. , ppl: 10.254459763902743], batch size: 70
+2022-12-10 17:45:38,827 INFO [train.py:421] (2/8) Epoch 3, batch 7400, loss[loss=2.404, over 1120.00 frames. , ppl: 11.06372940239562] tot_loss[loss=2.328, over 5395156.18 frames. , ppl: 10.260729915636754], batch size: 70
+2022-12-10 17:47:16,693 INFO [train.py:421] (2/8) Epoch 3, batch 7600, loss[loss=2.518, over 1540.00 frames. , ppl: 12.403785705378825] tot_loss[loss=2.327, over 5436057.38 frames. , ppl: 10.250592394765592], batch size: 70
+2022-12-10 17:48:56,787 INFO [train.py:421] (2/8) Epoch 3, batch 7800, loss[loss=2.401, over 1260.00 frames. , ppl: 11.032212706754535] tot_loss[loss=2.327, over 5458771.42 frames. , ppl: 10.242992958203725], batch size: 70
+2022-12-10 17:50:40,362 INFO [train.py:421] (2/8) Epoch 3, batch 8000, loss[loss=2.417, over 2590.00 frames. , ppl: 11.209523107755139] tot_loss[loss=2.325, over 5511059.44 frames. , ppl: 10.229623496572062], batch size: 70
+2022-12-10 17:50:40,363 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:50:41,126 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.153032911836572
+2022-12-10 17:52:20,457 INFO [train.py:421] (2/8) Epoch 3, batch 8200, loss[loss=2.426, over 910.00 frames. , ppl: 11.316113444963584] tot_loss[loss=2.325, over 5523241.89 frames. , ppl: 10.229853240466412], batch size: 70
+2022-12-10 17:53:58,404 INFO [train.py:421] (2/8) Epoch 3, batch 8400, loss[loss=2.309, over 3360.00 frames. , ppl: 10.065596303990274] tot_loss[loss=2.325, over 5528463.88 frames. , ppl: 10.22873141465446], batch size: 70
+2022-12-10 17:55:40,041 INFO [train.py:421] (2/8) Epoch 3, batch 8600, loss[loss=2.246, over 7700.00 frames. , ppl: 9.445699491209247] tot_loss[loss=2.326, over 5526509.62 frames. , ppl: 10.236151487236421], batch size: 70
+2022-12-10 17:57:23,065 INFO [train.py:421] (2/8) Epoch 3, batch 8800, loss[loss=2.316, over 2590.00 frames. , ppl: 10.136894774785823] tot_loss[loss=2.327, over 5492006.94 frames. , ppl: 10.245233984672034], batch size: 70
+2022-12-10 17:59:05,032 INFO [train.py:421] (2/8) Epoch 3, batch 9000, loss[loss=2.282, over 3640.00 frames. , ppl: 9.792381394588944] tot_loss[loss=2.328, over 5487351.13 frames. , ppl: 10.25317093651913], batch size: 70
+2022-12-10 17:59:05,032 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 17:59:05,778 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.158504303301722
+2022-12-10 18:00:47,946 INFO [train.py:421] (2/8) Epoch 3, batch 9200, loss[loss=2.387, over 1260.00 frames. , ppl: 10.8786406436684] tot_loss[loss=2.329, over 5481982.05 frames. , ppl: 10.264403122939852], batch size: 70
+2022-12-10 18:02:28,034 INFO [train.py:421] (2/8) Epoch 3, batch 9400, loss[loss=2.203, over 6230.00 frames. , ppl: 9.05286627514558] tot_loss[loss=2.33, over 5419119.06 frames. , ppl: 10.281130114772806], batch size: 70
+2022-12-10 18:04:09,125 INFO [train.py:421] (2/8) Epoch 3, batch 9600, loss[loss=2.525, over 1540.00 frames. , ppl: 12.490039838932113] tot_loss[loss=2.33, over 5440948.52 frames. , ppl: 10.272873814156128], batch size: 70
+2022-12-10 18:05:47,176 INFO [train.py:421] (2/8) Epoch 3, batch 9800, loss[loss=2.408, over 2730.00 frames. , ppl: 11.115342463658035] tot_loss[loss=2.33, over 5451777.12 frames. , ppl: 10.273363450985455], batch size: 70
+2022-12-10 18:07:29,496 INFO [train.py:421] (2/8) Epoch 3, batch 10000, loss[loss=2.231, over 4760.00 frames. , ppl: 9.307059588416616] tot_loss[loss=2.329, over 5483724.78 frames. , ppl: 10.262787071334836], batch size: 70
+2022-12-10 18:07:29,497 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:07:30,247 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.15942601173964
+2022-12-10 18:09:07,694 INFO [train.py:421] (2/8) Epoch 3, batch 10200, loss[loss=2.369, over 1610.00 frames. , ppl: 10.688849864685317] tot_loss[loss=2.33, over 5462000.77 frames. , ppl: 10.27583785136508], batch size: 70
+2022-12-10 18:10:48,659 INFO [train.py:421] (2/8) Epoch 3, batch 10400, loss[loss=2.526, over 1190.00 frames. , ppl: 12.50734795216072] tot_loss[loss=2.33, over 5436340.18 frames. , ppl: 10.277671745203206], batch size: 70
+2022-12-10 18:12:27,603 INFO [train.py:421] (2/8) Epoch 3, batch 10600, loss[loss=2.215, over 3360.00 frames. , ppl: 9.162726979060155] tot_loss[loss=2.331, over 5408077.58 frames. , ppl: 10.287809105656187], batch size: 70
+2022-12-10 18:14:05,263 INFO [train.py:421] (2/8) Epoch 3, batch 10800, loss[loss=2.265, over 6720.00 frames. , ppl: 9.629437935990541] tot_loss[loss=2.332, over 5376672.17 frames. , ppl: 10.298719845189659], batch size: 70
+2022-12-10 18:15:47,986 INFO [train.py:421] (2/8) Epoch 3, batch 11000, loss[loss=2.529, over 1400.00 frames. , ppl: 12.54236170599362] tot_loss[loss=2.333, over 5343647.18 frames. , ppl: 10.30382802574605], batch size: 70
+2022-12-10 18:15:47,987 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:15:48,745 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.163211456412954
+2022-12-10 18:17:26,886 INFO [train.py:421] (2/8) Epoch 3, batch 11200, loss[loss=2.526, over 1330.00 frames. , ppl: 12.50888441117256] tot_loss[loss=2.331, over 5376025.23 frames. , ppl: 10.286568080377641], batch size: 70
+2022-12-10 18:19:04,381 INFO [train.py:421] (2/8) Epoch 3, batch 11400, loss[loss=2.267, over 4830.00 frames. , ppl: 9.65401329215104] tot_loss[loss=2.33, over 5418044.96 frames. , ppl: 10.276387794479403], batch size: 70
+2022-12-10 18:20:41,826 INFO [train.py:421] (2/8) Epoch 3, batch 11600, loss[loss=2.32, over 3920.00 frames. , ppl: 10.180320492026318] tot_loss[loss=2.33, over 5425753.12 frames. , ppl: 10.274471054328467], batch size: 70
+2022-12-10 18:22:26,235 INFO [train.py:421] (2/8) Epoch 3, batch 11800, loss[loss=2.371, over 1540.00 frames. , ppl: 10.708594739676379] tot_loss[loss=2.33, over 5412631.31 frames. , ppl: 10.278612240225325], batch size: 70
+2022-12-10 18:24:10,856 INFO [train.py:421] (2/8) Epoch 3, batch 12000, loss[loss=2.193, over 8750.00 frames. , ppl: 8.958566880821225] tot_loss[loss=2.328, over 5482877.77 frames. , ppl: 10.256755578804334], batch size: 70
+2022-12-10 18:24:10,857 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:24:11,627 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.149768016214665
+2022-12-10 18:25:51,546 INFO [train.py:421] (2/8) Epoch 3, batch 12200, loss[loss=2.338, over 2310.00 frames. , ppl: 10.36152540273136] tot_loss[loss=2.328, over 5500110.87 frames. , ppl: 10.256990669869213], batch size: 70
+2022-12-10 18:27:29,158 INFO [train.py:421] (2/8) Epoch 3, batch 12400, loss[loss=2.288, over 3920.00 frames. , ppl: 9.856385991438687] tot_loss[loss=2.328, over 5510839.79 frames. , ppl: 10.252935546466814], batch size: 70
+2022-12-10 18:29:06,470 INFO [train.py:421] (2/8) Epoch 3, batch 12600, loss[loss=2.394, over 1960.00 frames. , ppl: 10.955937393091574] tot_loss[loss=2.328, over 5513100.47 frames. , ppl: 10.25532862852183], batch size: 70
+2022-12-10 18:30:46,101 INFO [train.py:421] (2/8) Epoch 3, batch 12800, loss[loss=2.247, over 4270.00 frames. , ppl: 9.462345443550799] tot_loss[loss=2.328, over 5490104.54 frames. , ppl: 10.261720070691975], batch size: 70
+2022-12-10 18:32:27,360 INFO [train.py:421] (2/8) Epoch 3, batch 13000, loss[loss=2.228, over 5390.00 frames. , ppl: 9.278939282498497] tot_loss[loss=2.327, over 5524338.75 frames. , ppl: 10.250684839316836], batch size: 70
+2022-12-10 18:32:27,360 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:32:28,107 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.154893455412658
+2022-12-10 18:34:13,397 INFO [train.py:421] (2/8) Epoch 3, batch 13200, loss[loss=2.421, over 1750.00 frames. , ppl: 11.253282650375294] tot_loss[loss=2.326, over 5587605.98 frames. , ppl: 10.236080149496418], batch size: 70
+2022-12-10 18:35:54,363 INFO [train.py:421] (2/8) Epoch 3, batch 13400, loss[loss=2.37, over 2030.00 frames. , ppl: 10.693217517770977] tot_loss[loss=2.326, over 5559036.83 frames. , ppl: 10.237501231731478], batch size: 70
+2022-12-10 18:37:33,223 INFO [train.py:421] (2/8) Epoch 3, batch 13600, loss[loss=2.633, over 1120.00 frames. , ppl: 13.911192047557558] tot_loss[loss=2.326, over 5567042.87 frames. , ppl: 10.234683773716872], batch size: 70
+2022-12-10 18:39:15,490 INFO [train.py:421] (2/8) Epoch 3, batch 13800, loss[loss=2.37, over 1750.00 frames. , ppl: 10.700723798809095] tot_loss[loss=2.327, over 5555091.87 frames. , ppl: 10.245194407519673], batch size: 70
+2022-12-10 18:40:59,171 INFO [train.py:421] (2/8) Epoch 3, batch 14000, loss[loss=2.399, over 910.00 frames. , ppl: 11.017310523173483] tot_loss[loss=2.327, over 5517501.15 frames. , ppl: 10.252054770549456], batch size: 70
+2022-12-10 18:40:59,171 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:40:59,950 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.146131741243801
+2022-12-10 18:42:39,823 INFO [train.py:421] (2/8) Epoch 3, batch 14200, loss[loss=2.389, over 1960.00 frames. , ppl: 10.901300945215688] tot_loss[loss=2.327, over 5530734.29 frames. , ppl: 10.24485255193298], batch size: 70
+2022-12-10 18:44:22,939 INFO [train.py:421] (2/8) Epoch 3, batch 14400, loss[loss=2.385, over 910.00 frames. , ppl: 10.862662400242549] tot_loss[loss=2.327, over 5522416.13 frames. , ppl: 10.24806552198118], batch size: 70
+2022-12-10 18:46:05,764 INFO [train.py:421] (2/8) Epoch 3, batch 14600, loss[loss=2.3, over 2870.00 frames. , ppl: 9.97888407526753] tot_loss[loss=2.326, over 5535078.73 frames. , ppl: 10.237253214404797], batch size: 70
+2022-12-10 18:47:46,146 INFO [train.py:421] (2/8) Epoch 3, batch 14800, loss[loss=2.377, over 1680.00 frames. , ppl: 10.77737115253391] tot_loss[loss=2.327, over 5500475.37 frames. , ppl: 10.24945744455342], batch size: 70
+2022-12-10 18:49:30,929 INFO [train.py:421] (2/8) Epoch 3, batch 15000, loss[loss=2.926, over 560.00 frames. , ppl: 18.647871034162655] tot_loss[loss=2.328, over 5495441.29 frames. , ppl: 10.254025914978591], batch size: 70
+2022-12-10 18:49:30,929 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:49:31,678 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.126931133917406
+2022-12-10 18:51:11,705 INFO [train.py:421] (2/8) Epoch 3, batch 15200, loss[loss=2.508, over 1610.00 frames. , ppl: 12.283141232535371] tot_loss[loss=2.329, over 5490784.05 frames. , ppl: 10.264985083810556], batch size: 70
+2022-12-10 18:52:50,289 INFO [train.py:421] (2/8) Epoch 3, batch 15400, loss[loss=2.188, over 11550.00 frames. , ppl: 8.914757014420088] tot_loss[loss=2.33, over 5466640.53 frames. , ppl: 10.277767910761067], batch size: 70
+2022-12-10 18:54:28,578 INFO [train.py:421] (2/8) Epoch 3, batch 15600, loss[loss=2.223, over 5040.00 frames. , ppl: 9.231162101738708] tot_loss[loss=2.328, over 5501791.36 frames. , ppl: 10.262016116715229], batch size: 70
+2022-12-10 18:56:05,396 INFO [train.py:421] (2/8) Epoch 3, batch 15800, loss[loss=2.316, over 840.00 frames. , ppl: 10.13095999164702] tot_loss[loss=2.328, over 5500627.83 frames. , ppl: 10.260358247699674], batch size: 70
+2022-12-10 18:57:45,622 INFO [train.py:421] (2/8) Epoch 3, batch 16000, loss[loss=2.22, over 3430.00 frames. , ppl: 9.209023422126396] tot_loss[loss=2.328, over 5503183.45 frames. , ppl: 10.25859869801809], batch size: 70
+2022-12-10 18:57:45,623 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 18:57:46,369 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.316, over 211138.00 frames. , ppl: 10.134522713241985
+2022-12-10 18:59:25,158 INFO [train.py:421] (2/8) Epoch 3, batch 16200, loss[loss=2.34, over 4620.00 frames. , ppl: 10.380198245236691] tot_loss[loss=2.328, over 5533797.51 frames. , ppl: 10.25274052462091], batch size: 70
+2022-12-10 19:01:05,960 INFO [train.py:421] (2/8) Epoch 3, batch 16400, loss[loss=3.129, over 560.00 frames. , ppl: 22.849796631459164] tot_loss[loss=2.327, over 5547423.61 frames. , ppl: 10.245933152586836], batch size: 70
+2022-12-10 19:02:45,401 INFO [train.py:421] (2/8) Epoch 3, batch 16600, loss[loss=2.488, over 1330.00 frames. , ppl: 12.032669085201478] tot_loss[loss=2.326, over 5585824.77 frames. , ppl: 10.232497810847049], batch size: 70
+2022-12-10 19:04:28,700 INFO [train.py:421] (2/8) Epoch 3, batch 16800, loss[loss=2.41, over 1330.00 frames. , ppl: 11.135696874178583] tot_loss[loss=2.325, over 5599009.31 frames. , ppl: 10.231000644726771], batch size: 70
+2022-12-10 19:06:10,450 INFO [train.py:421] (2/8) Epoch 3, batch 17000, loss[loss=2.344, over 2660.00 frames. , ppl: 10.426561041956248] tot_loss[loss=2.324, over 5629695.30 frames. , ppl: 10.21905317997958], batch size: 70
+2022-12-10 19:06:10,450 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:06:11,210 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.149083017239782
+2022-12-10 19:07:50,801 INFO [train.py:421] (2/8) Epoch 3, batch 17200, loss[loss=2.442, over 1120.00 frames. , ppl: 11.492803041934119] tot_loss[loss=2.323, over 5663019.07 frames. , ppl: 10.206157622010954], batch size: 70
+2022-12-10 19:09:33,686 INFO [train.py:421] (2/8) Epoch 3, batch 17400, loss[loss=2.36, over 2870.00 frames. , ppl: 10.58735985622875] tot_loss[loss=2.324, over 5630077.58 frames. , ppl: 10.21206033030677], batch size: 70
+2022-12-10 19:11:12,100 INFO [train.py:421] (2/8) Epoch 3, batch 17600, loss[loss=2.313, over 10990.00 frames. , ppl: 10.107521283589012] tot_loss[loss=2.324, over 5612120.72 frames. , ppl: 10.220782309830957], batch size: 70
+2022-12-10 19:12:53,522 INFO [train.py:421] (2/8) Epoch 3, batch 17800, loss[loss=2.374, over 1610.00 frames. , ppl: 10.743001690825604] tot_loss[loss=2.325, over 5605914.40 frames. , ppl: 10.226168898917987], batch size: 70
+2022-12-10 19:14:33,706 INFO [train.py:421] (2/8) Epoch 3, batch 18000, loss[loss=2.594, over 1260.00 frames. , ppl: 13.386666946098902] tot_loss[loss=2.325, over 5561018.37 frames. , ppl: 10.230431712683776], batch size: 70
+2022-12-10 19:14:33,707 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:14:34,466 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.129355082846555
+2022-12-10 19:16:15,834 INFO [train.py:421] (2/8) Epoch 3, batch 18200, loss[loss=3.557, over 490.00 frames. , ppl: 35.053756129796] tot_loss[loss=2.325, over 5585118.85 frames. , ppl: 10.221931111902904], batch size: 70
+2022-12-10 19:17:56,506 INFO [train.py:421] (2/8) Epoch 3, batch 18400, loss[loss=2.43, over 1190.00 frames. , ppl: 11.3545455853831] tot_loss[loss=2.325, over 5582184.41 frames. , ppl: 10.225175731138151], batch size: 70
+2022-12-10 19:19:35,454 INFO [train.py:421] (2/8) Epoch 3, batch 18600, loss[loss=2.22, over 4690.00 frames. , ppl: 9.205883131842235] tot_loss[loss=2.326, over 5554535.91 frames. , ppl: 10.233888500757075], batch size: 70
+2022-12-10 19:21:19,540 INFO [train.py:421] (2/8) Epoch 3, batch 18800, loss[loss=2.406, over 2030.00 frames. , ppl: 11.08632014055243] tot_loss[loss=2.324, over 5617888.28 frames. , ppl: 10.212936940653266], batch size: 70
+2022-12-10 19:22:57,474 INFO [train.py:421] (2/8) Epoch 3, batch 19000, loss[loss=2.407, over 2380.00 frames. , ppl: 11.103514112115048] tot_loss[loss=2.323, over 5656168.57 frames. , ppl: 10.201651297006567], batch size: 70
+2022-12-10 19:22:57,474 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:22:58,219 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.14855577978015
+2022-12-10 19:24:36,563 INFO [train.py:421] (2/8) Epoch 3, batch 19200, loss[loss=2.233, over 3220.00 frames. , ppl: 9.328852241911592] tot_loss[loss=2.322, over 5678479.59 frames. , ppl: 10.200840685117115], batch size: 70
+2022-12-10 19:26:16,122 INFO [train.py:421] (2/8) Epoch 3, batch 19400, loss[loss=2.909, over 630.00 frames. , ppl: 18.33342498910041] tot_loss[loss=2.323, over 5666793.60 frames. , ppl: 10.20425446994917], batch size: 70
+2022-12-10 19:27:55,649 INFO [train.py:421] (2/8) Epoch 3, batch 19600, loss[loss=2.357, over 1540.00 frames. , ppl: 10.555581101734678] tot_loss[loss=2.324, over 5634742.82 frames. , ppl: 10.216271775666275], batch size: 70
+2022-12-10 19:29:36,902 INFO [train.py:421] (2/8) Epoch 3, batch 19800, loss[loss=2.223, over 3430.00 frames. , ppl: 9.2366013520767] tot_loss[loss=2.323, over 5647738.47 frames. , ppl: 10.204877804964628], batch size: 70
+2022-12-10 19:31:19,065 INFO [train.py:421] (2/8) Epoch 3, batch 20000, loss[loss=2.467, over 1050.00 frames. , ppl: 11.784553718167528] tot_loss[loss=2.323, over 5644182.65 frames. , ppl: 10.204380236319844], batch size: 70
+2022-12-10 19:31:19,065 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:31:19,811 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.12899827473271
+2022-12-10 19:33:00,567 INFO [train.py:421] (2/8) Epoch 3, batch 20200, loss[loss=2.289, over 2940.00 frames. , ppl: 9.869065097080385] tot_loss[loss=2.322, over 5664766.44 frames. , ppl: 10.199121525583063], batch size: 70
+2022-12-10 19:34:42,566 INFO [train.py:421] (2/8) Epoch 3, batch 20400, loss[loss=2.863, over 630.00 frames. , ppl: 17.51790514140213] tot_loss[loss=2.323, over 5646897.89 frames. , ppl: 10.203229178887675], batch size: 70
+2022-12-10 19:36:23,578 INFO [train.py:421] (2/8) Epoch 3, batch 20600, loss[loss=2.657, over 770.00 frames. , ppl: 14.252088715562316] tot_loss[loss=2.324, over 5584424.05 frames. , ppl: 10.219945768111872], batch size: 70
+2022-12-10 19:38:05,418 INFO [train.py:421] (2/8) Epoch 3, batch 20800, loss[loss=2.267, over 3780.00 frames. , ppl: 9.654939194691845] tot_loss[loss=2.324, over 5583365.67 frames. , ppl: 10.220417120687598], batch size: 70
+2022-12-10 19:39:45,171 INFO [train.py:421] (2/8) Epoch 3, batch 21000, loss[loss=2.516, over 1890.00 frames. , ppl: 12.381845746531202] tot_loss[loss=2.324, over 5603305.97 frames. , ppl: 10.21563066862656], batch size: 70
+2022-12-10 19:39:45,172 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:39:45,934 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.115766279020953
+2022-12-10 19:41:26,042 INFO [train.py:421] (2/8) Epoch 3, batch 21200, loss[loss=2.394, over 1610.00 frames. , ppl: 10.954606012720527] tot_loss[loss=2.325, over 5569021.67 frames. , ppl: 10.225734989331684], batch size: 70
+2022-12-10 19:43:03,101 INFO [train.py:421] (2/8) Epoch 3, batch 21400, loss[loss=4.961, over 280.00 frames. , ppl: 142.76749383553462] tot_loss[loss=2.326, over 5546624.79 frames. , ppl: 10.241483036453591], batch size: 70
+2022-12-10 19:44:44,717 INFO [train.py:421] (2/8) Epoch 3, batch 21600, loss[loss=2.42, over 2030.00 frames. , ppl: 11.244612921717298] tot_loss[loss=2.327, over 5530800.33 frames. , ppl: 10.252231950090131], batch size: 70
+2022-12-10 19:46:24,994 INFO [train.py:421] (2/8) Epoch 3, batch 21800, loss[loss=2.311, over 3080.00 frames. , ppl: 10.086558306211918] tot_loss[loss=2.327, over 5525771.70 frames. , ppl: 10.245446997770593], batch size: 70
+2022-12-10 19:48:03,093 INFO [train.py:421] (2/8) Epoch 3, batch 22000, loss[loss=2.202, over 4620.00 frames. , ppl: 9.039649642957905] tot_loss[loss=2.326, over 5515851.57 frames. , ppl: 10.239508786881014], batch size: 70
+2022-12-10 19:48:03,093 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:48:03,838 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.119331761662655
+2022-12-10 19:49:42,909 INFO [train.py:421] (2/8) Epoch 3, batch 22200, loss[loss=2.207, over 4480.00 frames. , ppl: 9.08970073506241] tot_loss[loss=2.328, over 5484137.41 frames. , ppl: 10.254027508120227], batch size: 70
+2022-12-10 19:51:27,707 INFO [train.py:421] (2/8) Epoch 3, batch 22400, loss[loss=2.331, over 2660.00 frames. , ppl: 10.285467548373994] tot_loss[loss=2.327, over 5499297.56 frames. , ppl: 10.246063636646086], batch size: 70
+2022-12-10 19:53:09,801 INFO [train.py:421] (2/8) Epoch 3, batch 22600, loss[loss=2.727, over 700.00 frames. , ppl: 15.293839886973338] tot_loss[loss=2.328, over 5488577.59 frames. , ppl: 10.257141467794096], batch size: 70
+2022-12-10 19:54:48,467 INFO [train.py:421] (2/8) Epoch 3, batch 22800, loss[loss=2.252, over 6790.00 frames. , ppl: 9.507024377986218] tot_loss[loss=2.327, over 5498868.98 frames. , ppl: 10.247922145149314], batch size: 70
+2022-12-10 19:56:25,441 INFO [train.py:421] (2/8) Epoch 3, batch 23000, loss[loss=2.313, over 2030.00 frames. , ppl: 10.100963847642742] tot_loss[loss=2.327, over 5523516.97 frames. , ppl: 10.247912597604223], batch size: 70
+2022-12-10 19:56:25,442 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 19:56:26,223 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.111284628584807
+2022-12-10 19:58:04,891 INFO [train.py:421] (2/8) Epoch 3, batch 23200, loss[loss=2.302, over 3640.00 frames. , ppl: 9.99362869331929] tot_loss[loss=2.328, over 5490480.83 frames. , ppl: 10.25371816203903], batch size: 70
+2022-12-10 19:59:44,527 INFO [train.py:421] (2/8) Epoch 3, batch 23400, loss[loss=2.254, over 3290.00 frames. , ppl: 9.526351375231831] tot_loss[loss=2.328, over 5460754.53 frames. , ppl: 10.259554725551311], batch size: 70
+2022-12-10 20:01:22,837 INFO [train.py:421] (2/8) Epoch 3, batch 23600, loss[loss=2.346, over 1400.00 frames. , ppl: 10.443581183018297] tot_loss[loss=2.327, over 5504832.71 frames. , ppl: 10.252048615179117], batch size: 70
+2022-12-10 20:03:03,118 INFO [train.py:421] (2/8) Epoch 3, batch 23800, loss[loss=2.399, over 980.00 frames. , ppl: 11.00688467138036] tot_loss[loss=2.327, over 5524343.92 frames. , ppl: 10.242340904533167], batch size: 70
+2022-12-10 20:04:45,087 INFO [train.py:421] (2/8) Epoch 3, batch 24000, loss[loss=2.552, over 1190.00 frames. , ppl: 12.832908223238336] tot_loss[loss=2.326, over 5575337.19 frames. , ppl: 10.234810243278394], batch size: 70
+2022-12-10 20:04:45,087 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:04:45,874 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.115662972133224
+2022-12-10 20:06:27,302 INFO [train.py:421] (2/8) Epoch 3, batch 24200, loss[loss=2.475, over 1190.00 frames. , ppl: 11.882131272340493] tot_loss[loss=2.326, over 5559728.30 frames. , ppl: 10.240574675199941], batch size: 70
+2022-12-10 20:08:06,569 INFO [train.py:421] (2/8) Epoch 3, batch 24400, loss[loss=2.354, over 2240.00 frames. , ppl: 10.528068562926336] tot_loss[loss=2.327, over 5533649.53 frames. , ppl: 10.2481282230541], batch size: 70
+2022-12-10 20:09:46,778 INFO [train.py:421] (2/8) Epoch 3, batch 24600, loss[loss=2.357, over 2940.00 frames. , ppl: 10.560690929750825] tot_loss[loss=2.327, over 5530788.83 frames. , ppl: 10.242574943322607], batch size: 70
+2022-12-10 20:11:29,174 INFO [train.py:421] (2/8) Epoch 3, batch 24800, loss[loss=2.586, over 980.00 frames. , ppl: 13.274280739630132] tot_loss[loss=2.327, over 5508490.27 frames. , ppl: 10.245946916217033], batch size: 70
+2022-12-10 20:13:09,465 INFO [train.py:421] (2/8) Epoch 3, batch 25000, loss[loss=2.445, over 1050.00 frames. , ppl: 11.535633941668614] tot_loss[loss=2.326, over 5521307.39 frames. , ppl: 10.241638492378558], batch size: 70
+2022-12-10 20:13:09,466 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:13:10,215 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.122748675677348
+2022-12-10 20:14:53,081 INFO [train.py:421] (2/8) Epoch 3, batch 25200, loss[loss=2.498, over 2100.00 frames. , ppl: 12.152386577594143] tot_loss[loss=2.327, over 5524532.32 frames. , ppl: 10.243951637546926], batch size: 70
+2022-12-10 20:16:34,327 INFO [train.py:421] (2/8) Epoch 3, batch 25400, loss[loss=2.259, over 4690.00 frames. , ppl: 9.574381951221651] tot_loss[loss=2.326, over 5538977.82 frames. , ppl: 10.23739480834063], batch size: 70
+2022-12-10 20:18:12,387 INFO [train.py:421] (2/8) Epoch 3, batch 25600, loss[loss=2.242, over 7840.00 frames. , ppl: 9.411083157592387] tot_loss[loss=2.326, over 5537362.63 frames. , ppl: 10.233555330498348], batch size: 70
+2022-12-10 20:19:49,966 INFO [train.py:421] (2/8) Epoch 3, batch 25800, loss[loss=3.63, over 420.00 frames. , ppl: 37.707717909876045] tot_loss[loss=2.326, over 5531655.27 frames. , ppl: 10.232644651660223], batch size: 70
+2022-12-10 20:21:34,321 INFO [train.py:421] (2/8) Epoch 3, batch 26000, loss[loss=2.43, over 1470.00 frames. , ppl: 11.35395823360794] tot_loss[loss=2.326, over 5530442.68 frames. , ppl: 10.235041401810518], batch size: 70
+2022-12-10 20:21:34,321 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:21:35,067 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.316, over 211138.00 frames. , ppl: 10.132153014018082
+2022-12-10 20:23:08,525 INFO [train.py:421] (2/8) Epoch 3, batch 26200, loss[loss=2.325, over 2520.00 frames. , ppl: 10.227012990951724] tot_loss[loss=2.327, over 5483642.90 frames. , ppl: 10.248600604281144], batch size: 70
+2022-12-10 20:24:45,539 INFO [train.py:421] (2/8) Epoch 3, batch 26400, loss[loss=2.683, over 630.00 frames. , ppl: 14.628336947969128] tot_loss[loss=2.327, over 5457027.34 frames. , ppl: 10.250880620496849], batch size: 70
+2022-12-10 20:26:27,607 INFO [train.py:421] (2/8) Epoch 3, batch 26600, loss[loss=3.192, over 490.00 frames. , ppl: 24.33551685386046] tot_loss[loss=2.328, over 5410981.33 frames. , ppl: 10.25952325558923], batch size: 70
+2022-12-10 20:28:07,421 INFO [train.py:421] (2/8) Epoch 3, batch 26800, loss[loss=2.42, over 1400.00 frames. , ppl: 11.241537832015144] tot_loss[loss=2.328, over 5423173.59 frames. , ppl: 10.258319094471174], batch size: 70
+2022-12-10 20:29:49,943 INFO [train.py:421] (2/8) Epoch 3, batch 27000, loss[loss=2.245, over 3990.00 frames. , ppl: 9.43643021054755] tot_loss[loss=2.327, over 5444514.64 frames. , ppl: 10.249930740025547], batch size: 70
+2022-12-10 20:29:49,943 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:29:50,703 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.101093802370023
+2022-12-10 20:31:29,520 INFO [train.py:421] (2/8) Epoch 3, batch 27200, loss[loss=2.384, over 2660.00 frames. , ppl: 10.84870265574738] tot_loss[loss=2.327, over 5452080.28 frames. , ppl: 10.246990342330133], batch size: 70
+2022-12-10 20:33:11,530 INFO [train.py:421] (2/8) Epoch 3, batch 27400, loss[loss=2.234, over 5600.00 frames. , ppl: 9.34143053671509] tot_loss[loss=2.327, over 5473395.85 frames. , ppl: 10.24229230476096], batch size: 70
+2022-12-10 20:34:50,136 INFO [train.py:421] (2/8) Epoch 3, batch 27600, loss[loss=2.315, over 2870.00 frames. , ppl: 10.128950426987002] tot_loss[loss=2.327, over 5432372.04 frames. , ppl: 10.251255820889906], batch size: 70
+2022-12-10 20:36:32,924 INFO [train.py:421] (2/8) Epoch 3, batch 27800, loss[loss=2.512, over 1050.00 frames. , ppl: 12.326495725060296] tot_loss[loss=2.327, over 5472824.68 frames. , ppl: 10.243089943130142], batch size: 70
+2022-12-10 20:38:09,407 INFO [train.py:421] (2/8) Epoch 3, batch 28000, loss[loss=2.634, over 840.00 frames. , ppl: 13.929117991082128] tot_loss[loss=2.327, over 5462575.56 frames. , ppl: 10.250339260117752], batch size: 70
+2022-12-10 20:38:09,408 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:38:10,154 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.106116881121231
+2022-12-10 20:39:49,676 INFO [train.py:421] (2/8) Epoch 3, batch 28200, loss[loss=2.898, over 840.00 frames. , ppl: 18.134993684384003] tot_loss[loss=2.326, over 5490117.60 frames. , ppl: 10.241295322229975], batch size: 70
+2022-12-10 20:41:31,327 INFO [train.py:421] (2/8) Epoch 3, batch 28400, loss[loss=2.402, over 1400.00 frames. , ppl: 11.04790164325207] tot_loss[loss=2.326, over 5504520.95 frames. , ppl: 10.233350613376702], batch size: 70
+2022-12-10 20:43:09,535 INFO [train.py:421] (2/8) Epoch 3, batch 28600, loss[loss=2.381, over 2870.00 frames. , ppl: 10.816991091133975] tot_loss[loss=2.327, over 5439915.25 frames. , ppl: 10.247744396407196], batch size: 70
+2022-12-10 20:44:46,043 INFO [train.py:421] (2/8) Epoch 3, batch 28800, loss[loss=2.571, over 980.00 frames. , ppl: 13.075809381700006] tot_loss[loss=2.327, over 5398535.09 frames. , ppl: 10.251840481251289], batch size: 70
+2022-12-10 20:46:28,958 INFO [train.py:421] (2/8) Epoch 3, batch 29000, loss[loss=2.222, over 3150.00 frames. , ppl: 9.227957393525198] tot_loss[loss=2.325, over 5452615.42 frames. , ppl: 10.229803452549067], batch size: 70
+2022-12-10 20:46:28,959 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:46:29,722 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.10355792431578
+2022-12-10 20:48:07,969 INFO [train.py:421] (2/8) Epoch 3, batch 29200, loss[loss=2.25, over 3500.00 frames. , ppl: 9.484538499977202] tot_loss[loss=2.325, over 5458202.30 frames. , ppl: 10.22768530027584], batch size: 70
+2022-12-10 20:49:48,612 INFO [train.py:421] (2/8) Epoch 3, batch 29400, loss[loss=2.548, over 1050.00 frames. , ppl: 12.776688541655068] tot_loss[loss=2.325, over 5457806.75 frames. , ppl: 10.227794056572678], batch size: 70
+2022-12-10 20:51:26,256 INFO [train.py:421] (2/8) Epoch 3, batch 29600, loss[loss=2.349, over 1680.00 frames. , ppl: 10.472059443194105] tot_loss[loss=2.326, over 5470093.53 frames. , ppl: 10.23489197790462], batch size: 70
+2022-12-10 20:53:01,786 INFO [train.py:421] (2/8) Epoch 3, batch 29800, loss[loss=2.412, over 1260.00 frames. , ppl: 11.158025091594846] tot_loss[loss=2.327, over 5444823.03 frames. , ppl: 10.249024784483241], batch size: 70
+2022-12-10 20:54:41,558 INFO [train.py:421] (2/8) Epoch 3, batch 30000, loss[loss=2.984, over 560.00 frames. , ppl: 19.768416126451324] tot_loss[loss=2.328, over 5430990.00 frames. , ppl: 10.255418941700032], batch size: 70
+2022-12-10 20:54:41,559 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 20:54:42,324 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.100455441623417
+2022-12-10 20:56:19,483 INFO [train.py:421] (2/8) Epoch 3, batch 30200, loss[loss=2.264, over 3570.00 frames. , ppl: 9.618999543667085] tot_loss[loss=2.328, over 5446505.64 frames. , ppl: 10.25529678814455], batch size: 70
+2022-12-10 20:58:04,768 INFO [train.py:421] (2/8) Epoch 3, batch 30400, loss[loss=2.995, over 560.00 frames. , ppl: 19.975462715410853] tot_loss[loss=2.328, over 5423104.45 frames. , ppl: 10.261207340636094], batch size: 70
+2022-12-10 20:59:43,368 INFO [train.py:421] (2/8) Epoch 3, batch 30600, loss[loss=2.292, over 3640.00 frames. , ppl: 9.892262402342702] tot_loss[loss=2.328, over 5443427.69 frames. , ppl: 10.254411158420664], batch size: 70
+2022-12-10 21:01:20,859 INFO [train.py:421] (2/8) Epoch 3, batch 30800, loss[loss=2.583, over 1120.00 frames. , ppl: 13.232518869233473] tot_loss[loss=2.327, over 5465109.01 frames. , ppl: 10.25095850093752], batch size: 70
+2022-12-10 21:02:59,411 INFO [train.py:421] (2/8) Epoch 3, batch 31000, loss[loss=2.543, over 910.00 frames. , ppl: 12.72320535832427] tot_loss[loss=2.327, over 5470403.70 frames. , ppl: 10.245430013280107], batch size: 70
+2022-12-10 21:02:59,411 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:03:00,157 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.091295598195819
+2022-12-10 21:04:37,247 INFO [train.py:421] (2/8) Epoch 3, batch 31200, loss[loss=2.323, over 2590.00 frames. , ppl: 10.208268942612907] tot_loss[loss=2.327, over 5440606.77 frames. , ppl: 10.248962607940381], batch size: 70
+2022-12-10 21:06:17,343 INFO [train.py:421] (2/8) Epoch 3, batch 31400, loss[loss=2.239, over 2730.00 frames. , ppl: 9.38543895404589] tot_loss[loss=2.327, over 5426410.84 frames. , ppl: 10.249211037455265], batch size: 70
+2022-12-10 21:07:59,102 INFO [train.py:421] (2/8) Epoch 3, batch 31600, loss[loss=2.398, over 1680.00 frames. , ppl: 11.003995823755433] tot_loss[loss=2.326, over 5469440.45 frames. , ppl: 10.239154734451693], batch size: 70
+2022-12-10 21:09:38,631 INFO [train.py:421] (2/8) Epoch 3, batch 31800, loss[loss=2.243, over 2660.00 frames. , ppl: 9.424600982669377] tot_loss[loss=2.326, over 5466745.06 frames. , ppl: 10.238810397683094], batch size: 70
+2022-12-10 21:11:16,581 INFO [train.py:421] (2/8) Epoch 3, batch 32000, loss[loss=2.484, over 1750.00 frames. , ppl: 11.98376066491997] tot_loss[loss=2.325, over 5507636.52 frames. , ppl: 10.2226718680682], batch size: 70
+2022-12-10 21:11:16,581 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:11:17,340 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.101464578420366
+2022-12-10 21:12:58,747 INFO [train.py:421] (2/8) Epoch 3, batch 32200, loss[loss=2.86, over 630.00 frames. , ppl: 17.454411093188202] tot_loss[loss=2.325, over 5505328.46 frames. , ppl: 10.222140284889283], batch size: 70
+2022-12-10 21:14:34,252 INFO [train.py:421] (2/8) Epoch 3, batch 32400, loss[loss=2.399, over 2380.00 frames. , ppl: 11.010340704785126] tot_loss[loss=2.326, over 5447161.10 frames. , ppl: 10.241888111288883], batch size: 70
+2022-12-10 21:16:14,624 INFO [train.py:421] (2/8) Epoch 3, batch 32600, loss[loss=2.367, over 1960.00 frames. , ppl: 10.6613625091317] tot_loss[loss=2.327, over 5442967.49 frames. , ppl: 10.242262355968448], batch size: 70
+2022-12-10 21:17:51,467 INFO [train.py:421] (2/8) Epoch 3, batch 32800, loss[loss=2.564, over 840.00 frames. , ppl: 12.989276292195266] tot_loss[loss=2.327, over 5416018.24 frames. , ppl: 10.243977530984326], batch size: 70
+2022-12-10 21:19:29,969 INFO [train.py:421] (2/8) Epoch 3, batch 33000, loss[loss=2.395, over 2380.00 frames. , ppl: 10.963847412725595] tot_loss[loss=2.326, over 5418578.52 frames. , ppl: 10.235793416482199], batch size: 70
+2022-12-10 21:19:29,970 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:19:30,729 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.111637819514526
+2022-12-10 21:21:05,637 INFO [train.py:421] (2/8) Epoch 3, batch 33200, loss[loss=2.694, over 910.00 frames. , ppl: 14.790514927402354] tot_loss[loss=2.327, over 5381662.90 frames. , ppl: 10.24451031259014], batch size: 70
+2022-12-10 21:22:50,813 INFO [train.py:421] (2/8) Epoch 3, batch 33400, loss[loss=2.347, over 2380.00 frames. , ppl: 10.457782709578607] tot_loss[loss=2.327, over 5384049.13 frames. , ppl: 10.24899031085805], batch size: 70
+2022-12-10 21:24:31,309 INFO [train.py:421] (2/8) Epoch 3, batch 33600, loss[loss=2.389, over 1890.00 frames. , ppl: 10.898848582883641] tot_loss[loss=2.326, over 5389042.12 frames. , ppl: 10.241561292960778], batch size: 70
+2022-12-10 21:26:13,417 INFO [train.py:421] (2/8) Epoch 3, batch 33800, loss[loss=2.259, over 4550.00 frames. , ppl: 9.574653790061072] tot_loss[loss=2.327, over 5385032.64 frames. , ppl: 10.247788090490964], batch size: 70
+2022-12-10 21:27:54,536 INFO [train.py:421] (2/8) Epoch 3, batch 34000, loss[loss=2.797, over 700.00 frames. , ppl: 16.4002056692269] tot_loss[loss=2.326, over 5412451.72 frames. , ppl: 10.237390776143075], batch size: 70
+2022-12-10 21:27:54,537 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:27:55,295 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.071958753555176
+2022-12-10 21:29:32,478 INFO [train.py:421] (2/8) Epoch 3, batch 34200, loss[loss=2.299, over 3780.00 frames. , ppl: 9.963834128994183] tot_loss[loss=2.327, over 5398817.04 frames. , ppl: 10.24228612689309], batch size: 70
+2022-12-10 21:31:11,666 INFO [train.py:421] (2/8) Epoch 3, batch 34400, loss[loss=2.479, over 1190.00 frames. , ppl: 11.92969389163677] tot_loss[loss=2.326, over 5434602.92 frames. , ppl: 10.233504000526295], batch size: 70
+2022-12-10 21:32:50,593 INFO [train.py:421] (2/8) Epoch 3, batch 34600, loss[loss=2.865, over 630.00 frames. , ppl: 17.5497897351332] tot_loss[loss=2.328, over 5356747.52 frames. , ppl: 10.261257775049959], batch size: 70
+2022-12-10 21:34:29,939 INFO [train.py:421] (2/8) Epoch 3, batch 34800, loss[loss=2.429, over 1540.00 frames. , ppl: 11.350660642949073] tot_loss[loss=2.33, over 5333446.40 frames. , ppl: 10.273029343997726], batch size: 70
+2022-12-10 21:36:14,331 INFO [train.py:421] (2/8) Epoch 3, batch 35000, loss[loss=2.437, over 1330.00 frames. , ppl: 11.43887443783593] tot_loss[loss=2.328, over 5366140.48 frames. , ppl: 10.25689299292322], batch size: 70
+2022-12-10 21:36:14,332 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:36:15,077 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.083123013689438
+2022-12-10 21:37:54,319 INFO [train.py:421] (2/8) Epoch 3, batch 35200, loss[loss=2.494, over 1260.00 frames. , ppl: 12.112963693104081] tot_loss[loss=2.327, over 5418200.55 frames. , ppl: 10.246604595998466], batch size: 70
+2022-12-10 21:39:36,771 INFO [train.py:421] (2/8) Epoch 3, batch 35400, loss[loss=2.459, over 1260.00 frames. , ppl: 11.692544721255063] tot_loss[loss=2.326, over 5445756.78 frames. , ppl: 10.231819297484394], batch size: 70
+2022-12-10 21:41:19,923 INFO [train.py:421] (2/8) Epoch 3, batch 35600, loss[loss=2.229, over 3990.00 frames. , ppl: 9.286817336055156] tot_loss[loss=2.325, over 5460090.96 frames. , ppl: 10.222683048129259], batch size: 70
+2022-12-10 21:42:54,211 INFO [train.py:421] (2/8) Epoch 3, batch 35800, loss[loss=2.398, over 1890.00 frames. , ppl: 10.997629612869265] tot_loss[loss=2.326, over 5447744.06 frames. , ppl: 10.232792392592694], batch size: 70
+2022-12-10 21:44:36,435 INFO [train.py:421] (2/8) Epoch 3, batch 36000, loss[loss=2.522, over 2170.00 frames. , ppl: 12.447299929259408] tot_loss[loss=2.325, over 5449013.40 frames. , ppl: 10.231162182797776], batch size: 70
+2022-12-10 21:44:36,435 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:44:37,182 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.086282871808761
+2022-12-10 21:46:17,615 INFO [train.py:421] (2/8) Epoch 3, batch 36200, loss[loss=2.657, over 700.00 frames. , ppl: 14.257338099408457] tot_loss[loss=2.324, over 5488325.33 frames. , ppl: 10.21906833883563], batch size: 70
+2022-12-10 21:47:56,619 INFO [train.py:421] (2/8) Epoch 3, batch 36400, loss[loss=2.31, over 2800.00 frames. , ppl: 10.071327821749707] tot_loss[loss=2.325, over 5453408.96 frames. , ppl: 10.231768140979165], batch size: 70
+2022-12-10 21:49:39,024 INFO [train.py:421] (2/8) Epoch 3, batch 36600, loss[loss=2.227, over 7910.00 frames. , ppl: 9.276322944811099] tot_loss[loss=2.325, over 5464776.73 frames. , ppl: 10.222567202098492], batch size: 70
+2022-12-10 21:51:17,838 INFO [train.py:421] (2/8) Epoch 3, batch 36800, loss[loss=2.245, over 4200.00 frames. , ppl: 9.438385365474321] tot_loss[loss=2.324, over 5504246.40 frames. , ppl: 10.21375775641569], batch size: 70
+2022-12-10 21:53:00,121 INFO [train.py:421] (2/8) Epoch 3, batch 37000, loss[loss=2.566, over 840.00 frames. , ppl: 13.01872259182609] tot_loss[loss=2.324, over 5524613.04 frames. , ppl: 10.215508308358435], batch size: 70
+2022-12-10 21:53:00,121 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 21:53:00,870 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.091421060289873
+2022-12-10 21:54:40,044 INFO [train.py:421] (2/8) Epoch 3, batch 37200, loss[loss=3.71, over 420.00 frames. , ppl: 40.83407212774567] tot_loss[loss=2.322, over 5574098.57 frames. , ppl: 10.19915996355107], batch size: 70
+2022-12-10 21:56:18,550 INFO [train.py:421] (2/8) Epoch 3, batch 37400, loss[loss=2.473, over 1190.00 frames. , ppl: 11.86222674893806] tot_loss[loss=2.322, over 5551927.18 frames. , ppl: 10.200924482967464], batch size: 70
+2022-12-10 21:57:58,918 INFO [train.py:421] (2/8) Epoch 3, batch 37600, loss[loss=2.308, over 1470.00 frames. , ppl: 10.052868397647217] tot_loss[loss=2.322, over 5544006.66 frames. , ppl: 10.199481920947338], batch size: 70
+2022-12-10 21:59:41,112 INFO [train.py:421] (2/8) Epoch 3, batch 37800, loss[loss=2.341, over 2940.00 frames. , ppl: 10.390656691849438] tot_loss[loss=2.322, over 5561740.40 frames. , ppl: 10.199047765209125], batch size: 70
+2022-12-10 22:01:19,439 INFO [train.py:421] (2/8) Epoch 3, batch 38000, loss[loss=2.311, over 5320.00 frames. , ppl: 10.081998927506394] tot_loss[loss=2.322, over 5566251.18 frames. , ppl: 10.199261850472444], batch size: 70
+2022-12-10 22:01:19,439 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:01:20,188 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.088428318438387
+2022-12-10 22:02:59,176 INFO [train.py:421] (2/8) Epoch 3, batch 38200, loss[loss=2.287, over 3430.00 frames. , ppl: 9.847154655673785] tot_loss[loss=2.321, over 5590576.22 frames. , ppl: 10.182640679706584], batch size: 70
+2022-12-10 22:04:43,417 INFO [train.py:421] (2/8) Epoch 3, batch 38400, loss[loss=2.959, over 560.00 frames. , ppl: 19.276230717348625] tot_loss[loss=2.321, over 5595509.40 frames. , ppl: 10.190545724625663], batch size: 70
+2022-12-10 22:06:21,541 INFO [train.py:421] (2/8) Epoch 3, batch 38600, loss[loss=2.793, over 700.00 frames. , ppl: 16.321807427437044] tot_loss[loss=2.321, over 5599109.47 frames. , ppl: 10.183782137890095], batch size: 70
+2022-12-10 22:08:03,079 INFO [train.py:421] (2/8) Epoch 3, batch 38800, loss[loss=2.264, over 4550.00 frames. , ppl: 9.624445986579477] tot_loss[loss=2.32, over 5603648.70 frames. , ppl: 10.180297094710117], batch size: 70
+2022-12-10 22:09:44,350 INFO [train.py:421] (2/8) Epoch 3, batch 39000, loss[loss=2.164, over 7560.00 frames. , ppl: 8.704112582381551] tot_loss[loss=2.32, over 5604736.57 frames. , ppl: 10.175327187872627], batch size: 70
+2022-12-10 22:09:44,350 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:09:45,098 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.087220422135697
+2022-12-10 22:11:24,489 INFO [train.py:421] (2/8) Epoch 3, batch 39200, loss[loss=2.26, over 3010.00 frames. , ppl: 9.57875477444971] tot_loss[loss=2.321, over 5561665.90 frames. , ppl: 10.18501376216129], batch size: 70
+2022-12-10 22:13:02,104 INFO [train.py:421] (2/8) Epoch 3, batch 39400, loss[loss=2.345, over 1960.00 frames. , ppl: 10.438072937305087] tot_loss[loss=2.322, over 5518836.71 frames. , ppl: 10.196896181170823], batch size: 70
+2022-12-10 22:14:41,111 INFO [train.py:421] (2/8) Epoch 3, batch 39600, loss[loss=2.216, over 5110.00 frames. , ppl: 9.170772725030307] tot_loss[loss=2.323, over 5499500.60 frames. , ppl: 10.208867357452771], batch size: 70
+2022-12-10 22:16:21,148 INFO [train.py:421] (2/8) Epoch 3, batch 39800, loss[loss=2.288, over 2590.00 frames. , ppl: 9.850414877515338] tot_loss[loss=2.323, over 5506918.14 frames. , ppl: 10.210523815133298], batch size: 70
+2022-12-10 22:17:59,566 INFO [train.py:421] (2/8) Epoch 3, batch 40000, loss[loss=2.319, over 4130.00 frames. , ppl: 10.169881110994664] tot_loss[loss=2.324, over 5500213.44 frames. , ppl: 10.213574799335934], batch size: 70
+2022-12-10 22:17:59,566 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:18:00,311 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.0863605000554
+2022-12-10 22:19:42,202 INFO [train.py:421] (2/8) Epoch 3, batch 40200, loss[loss=2.298, over 4060.00 frames. , ppl: 9.949973608787989] tot_loss[loss=2.324, over 5487061.73 frames. , ppl: 10.211866469415023], batch size: 70
+2022-12-10 22:21:21,716 INFO [train.py:421] (2/8) Epoch 3, batch 40400, loss[loss=2.395, over 1400.00 frames. , ppl: 10.966180344136387] tot_loss[loss=2.324, over 5451367.35 frames. , ppl: 10.218104268678243], batch size: 70
+2022-12-10 22:23:02,836 INFO [train.py:421] (2/8) Epoch 3, batch 40600, loss[loss=2.468, over 1820.00 frames. , ppl: 11.793083312922795] tot_loss[loss=2.325, over 5443641.15 frames. , ppl: 10.22482263749659], batch size: 70
+2022-12-10 22:24:42,985 INFO [train.py:421] (2/8) Epoch 3, batch 40800, loss[loss=2.348, over 1540.00 frames. , ppl: 10.459553694265049] tot_loss[loss=2.325, over 5434908.22 frames. , ppl: 10.227711824287283], batch size: 70
+2022-12-10 22:26:20,554 INFO [train.py:421] (2/8) Epoch 3, batch 41000, loss[loss=2.339, over 4410.00 frames. , ppl: 10.372622671157206] tot_loss[loss=2.326, over 5401024.55 frames. , ppl: 10.238243228581176], batch size: 70
+2022-12-10 22:26:20,554 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:26:21,301 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.09838814732227
+2022-12-10 22:28:02,987 INFO [train.py:421] (2/8) Epoch 3, batch 41200, loss[loss=2.398, over 2030.00 frames. , ppl: 10.997495389773405] tot_loss[loss=2.326, over 5387553.56 frames. , ppl: 10.235339841924153], batch size: 70
+2022-12-10 22:29:45,178 INFO [train.py:421] (2/8) Epoch 3, batch 41400, loss[loss=2.405, over 2310.00 frames. , ppl: 11.07466214744063] tot_loss[loss=2.325, over 5391787.78 frames. , ppl: 10.23117587577415], batch size: 70
+2022-12-10 22:31:24,818 INFO [train.py:421] (2/8) Epoch 3, batch 41600, loss[loss=2.441, over 1400.00 frames. , ppl: 11.480706134751527] tot_loss[loss=2.325, over 5409274.94 frames. , ppl: 10.228718813637299], batch size: 70
+2022-12-10 22:32:59,993 INFO [train.py:421] (2/8) Epoch 3, batch 41800, loss[loss=2.254, over 6930.00 frames. , ppl: 9.522819082969546] tot_loss[loss=2.324, over 5456793.52 frames. , ppl: 10.220069264745048], batch size: 70
+2022-12-10 22:34:40,203 INFO [train.py:421] (2/8) Epoch 3, batch 42000, loss[loss=2.4, over 1890.00 frames. , ppl: 11.023791530415746] tot_loss[loss=2.323, over 5502621.40 frames. , ppl: 10.208968099361156], batch size: 70
+2022-12-10 22:34:40,204 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:34:40,954 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.07229715383208
+2022-12-10 22:36:24,003 INFO [train.py:421] (2/8) Epoch 3, batch 42200, loss[loss=2.473, over 1470.00 frames. , ppl: 11.854368410173066] tot_loss[loss=2.323, over 5519225.07 frames. , ppl: 10.208799579761077], batch size: 70
+2022-12-10 22:38:04,422 INFO [train.py:421] (2/8) Epoch 3, batch 42400, loss[loss=2.258, over 4270.00 frames. , ppl: 9.560318815928548] tot_loss[loss=2.323, over 5512008.70 frames. , ppl: 10.209198196118445], batch size: 70
+2022-12-10 22:39:45,512 INFO [train.py:421] (2/8) Epoch 3, batch 42600, loss[loss=2.317, over 3780.00 frames. , ppl: 10.149908690201482] tot_loss[loss=2.324, over 5506886.47 frames. , ppl: 10.213817580937445], batch size: 70
+2022-12-10 22:41:24,894 INFO [train.py:421] (2/8) Epoch 3, batch 42800, loss[loss=2.451, over 1890.00 frames. , ppl: 11.598656126586382] tot_loss[loss=2.324, over 5509559.14 frames. , ppl: 10.211773702701887], batch size: 70
+2022-12-10 22:43:08,205 INFO [train.py:421] (2/8) Epoch 3, batch 43000, loss[loss=2.329, over 2520.00 frames. , ppl: 10.270709435124171] tot_loss[loss=2.323, over 5528088.87 frames. , ppl: 10.209162223596175], batch size: 70
+2022-12-10 22:43:08,206 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:43:08,978 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.073643413765264
+2022-12-10 22:44:46,748 INFO [train.py:421] (2/8) Epoch 3, batch 43200, loss[loss=2.384, over 3360.00 frames. , ppl: 10.851109323374036] tot_loss[loss=2.323, over 5505147.05 frames. , ppl: 10.209631271446648], batch size: 70
+2022-12-10 22:46:25,132 INFO [train.py:421] (2/8) Epoch 3, batch 43400, loss[loss=2.409, over 980.00 frames. , ppl: 11.120359337658973] tot_loss[loss=2.323, over 5506493.80 frames. , ppl: 10.209462368955545], batch size: 70
+2022-12-10 22:48:04,841 INFO [train.py:421] (2/8) Epoch 3, batch 43600, loss[loss=2.435, over 1820.00 frames. , ppl: 11.41780474656659] tot_loss[loss=2.324, over 5476484.99 frames. , ppl: 10.217884658920402], batch size: 70
+2022-12-10 22:49:41,094 INFO [train.py:421] (2/8) Epoch 3, batch 43800, loss[loss=2.293, over 2660.00 frames. , ppl: 9.902770395485598] tot_loss[loss=2.324, over 5463320.42 frames. , ppl: 10.219413596588115], batch size: 70
+2022-12-10 22:51:19,560 INFO [train.py:421] (2/8) Epoch 3, batch 44000, loss[loss=2.258, over 3360.00 frames. , ppl: 9.564631198344687] tot_loss[loss=2.323, over 5505377.34 frames. , ppl: 10.203143612649658], batch size: 70
+2022-12-10 22:51:19,560 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:51:20,323 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.093368910010755
+2022-12-10 22:53:02,986 INFO [train.py:421] (2/8) Epoch 3, batch 44200, loss[loss=2.458, over 1190.00 frames. , ppl: 11.682388960893054] tot_loss[loss=2.323, over 5507702.84 frames. , ppl: 10.20373112798944], batch size: 70
+2022-12-10 22:54:43,985 INFO [train.py:421] (2/8) Epoch 3, batch 44400, loss[loss=2.596, over 910.00 frames. , ppl: 13.414407359621332] tot_loss[loss=2.323, over 5495361.99 frames. , ppl: 10.202927728231701], batch size: 70
+2022-12-10 22:56:24,869 INFO [train.py:421] (2/8) Epoch 3, batch 44600, loss[loss=2.267, over 3150.00 frames. , ppl: 9.654295673254738] tot_loss[loss=2.323, over 5458532.71 frames. , ppl: 10.20946505727901], batch size: 70
+2022-12-10 22:58:04,158 INFO [train.py:421] (2/8) Epoch 3, batch 44800, loss[loss=2.641, over 770.00 frames. , ppl: 14.029551328470053] tot_loss[loss=2.323, over 5480543.10 frames. , ppl: 10.205744539749258], batch size: 70
+2022-12-10 22:59:45,431 INFO [train.py:421] (2/8) Epoch 3, batch 45000, loss[loss=2.52, over 1260.00 frames. , ppl: 12.433329185367871] tot_loss[loss=2.323, over 5478650.44 frames. , ppl: 10.203294363368974], batch size: 70
+2022-12-10 22:59:45,431 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 22:59:46,181 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.10122686156436
+2022-12-10 23:01:30,008 INFO [train.py:421] (2/8) Epoch 3, batch 45200, loss[loss=4.108, over 350.00 frames. , ppl: 60.800860033006224] tot_loss[loss=2.323, over 5472019.93 frames. , ppl: 10.204081276810749], batch size: 70
+2022-12-10 23:03:11,281 INFO [train.py:421] (2/8) Epoch 3, batch 45400, loss[loss=2.148, over 10990.00 frames. , ppl: 8.566676231264623] tot_loss[loss=2.322, over 5472498.54 frames. , ppl: 10.197644102053527], batch size: 70
+2022-12-10 23:04:51,282 INFO [train.py:421] (2/8) Epoch 3, batch 45600, loss[loss=2.628, over 700.00 frames. , ppl: 13.848865271013114] tot_loss[loss=2.322, over 5476679.85 frames. , ppl: 10.191223715076216], batch size: 70
+2022-12-10 23:06:30,767 INFO [train.py:421] (2/8) Epoch 3, batch 45800, loss[loss=2.364, over 2660.00 frames. , ppl: 10.637700018532428] tot_loss[loss=2.322, over 5470589.65 frames. , ppl: 10.193504593040183], batch size: 70
+2022-12-10 23:08:09,340 INFO [train.py:421] (2/8) Epoch 3, batch 46000, loss[loss=2.291, over 2870.00 frames. , ppl: 9.884211883899587] tot_loss[loss=2.321, over 5490362.46 frames. , ppl: 10.182315301266552], batch size: 70
+2022-12-10 23:08:09,340 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:08:10,128 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.07448734063421
+2022-12-10 23:09:50,388 INFO [train.py:421] (2/8) Epoch 3, batch 46200, loss[loss=2.892, over 630.00 frames. , ppl: 18.030868277862236] tot_loss[loss=2.321, over 5506903.64 frames. , ppl: 10.182658552866485], batch size: 70
+2022-12-10 23:11:28,888 INFO [train.py:421] (2/8) Epoch 3, batch 46400, loss[loss=2.541, over 840.00 frames. , ppl: 12.697768944979705] tot_loss[loss=2.321, over 5485225.81 frames. , ppl: 10.187278772819527], batch size: 70
+2022-12-10 23:13:11,853 INFO [train.py:421] (2/8) Epoch 3, batch 46600, loss[loss=2.2, over 5110.00 frames. , ppl: 9.028311830453138] tot_loss[loss=2.322, over 5463238.52 frames. , ppl: 10.196770107814714], batch size: 70
+2022-12-10 23:14:52,161 INFO [train.py:421] (2/8) Epoch 3, batch 46800, loss[loss=2.235, over 4690.00 frames. , ppl: 9.34889810983377] tot_loss[loss=2.322, over 5472751.15 frames. , ppl: 10.200116576158091], batch size: 70
+2022-12-10 23:16:28,906 INFO [train.py:421] (2/8) Epoch 3, batch 47000, loss[loss=2.6, over 840.00 frames. , ppl: 13.467522582185557] tot_loss[loss=2.322, over 5501639.54 frames. , ppl: 10.200745793902923], batch size: 70
+2022-12-10 23:16:28,907 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:16:29,653 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.048629133994561
+2022-12-10 23:18:11,341 INFO [train.py:421] (2/8) Epoch 3, batch 47200, loss[loss=2.454, over 1260.00 frames. , ppl: 11.629816678806877] tot_loss[loss=2.322, over 5515592.45 frames. , ppl: 10.194121398227475], batch size: 70
+2022-12-10 23:19:51,596 INFO [train.py:421] (2/8) Epoch 3, batch 47400, loss[loss=2.535, over 1050.00 frames. , ppl: 12.61852554656094] tot_loss[loss=2.321, over 5525001.98 frames. , ppl: 10.189569350671226], batch size: 70
+2022-12-10 23:21:30,718 INFO [train.py:421] (2/8) Epoch 3, batch 47600, loss[loss=2.918, over 630.00 frames. , ppl: 18.500723969986172] tot_loss[loss=2.32, over 5523573.04 frames. , ppl: 10.180048685550107], batch size: 70
+2022-12-10 23:23:11,823 INFO [train.py:421] (2/8) Epoch 3, batch 47800, loss[loss=2.257, over 4130.00 frames. , ppl: 9.559151420661195] tot_loss[loss=2.32, over 5528152.73 frames. , ppl: 10.178022630035077], batch size: 70
+2022-12-10 23:24:53,411 INFO [train.py:421] (2/8) Epoch 3, batch 48000, loss[loss=2.194, over 8050.00 frames. , ppl: 8.975509565392406] tot_loss[loss=2.319, over 5542174.71 frames. , ppl: 10.167689393872102], batch size: 70
+2022-12-10 23:24:53,412 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:24:54,156 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.0626251790151
+2022-12-10 23:26:31,205 INFO [train.py:421] (2/8) Epoch 3, batch 48200, loss[loss=2.412, over 1680.00 frames. , ppl: 11.153958559517585] tot_loss[loss=2.319, over 5540120.28 frames. , ppl: 10.16994737223703], batch size: 70
+2022-12-10 23:28:10,222 INFO [train.py:421] (2/8) Epoch 3, batch 48400, loss[loss=2.578, over 1680.00 frames. , ppl: 13.177034663724283] tot_loss[loss=2.32, over 5534794.48 frames. , ppl: 10.176444758558329], batch size: 70
+2022-12-10 23:29:50,970 INFO [train.py:421] (2/8) Epoch 3, batch 48600, loss[loss=2.421, over 1540.00 frames. , ppl: 11.25989044122999] tot_loss[loss=2.32, over 5530256.19 frames. , ppl: 10.178166625138653], batch size: 70
+2022-12-10 23:31:26,389 INFO [train.py:421] (2/8) Epoch 3, batch 48800, loss[loss=2.371, over 1890.00 frames. , ppl: 10.705713059392737] tot_loss[loss=2.321, over 5490738.79 frames. , ppl: 10.183809083748185], batch size: 70
+2022-12-10 23:33:09,328 INFO [train.py:421] (2/8) Epoch 3, batch 49000, loss[loss=2.331, over 3360.00 frames. , ppl: 10.283580855993803] tot_loss[loss=2.319, over 5553009.88 frames. , ppl: 10.166617742418897], batch size: 70
+2022-12-10 23:33:09,328 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:33:10,087 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.064253163288134
+2022-12-10 23:34:51,166 INFO [train.py:421] (2/8) Epoch 3, batch 49200, loss[loss=2.435, over 840.00 frames. , ppl: 11.416218863300013] tot_loss[loss=2.319, over 5580501.12 frames. , ppl: 10.16751971056616], batch size: 70
+2022-12-10 23:36:30,854 INFO [train.py:421] (2/8) Epoch 3, batch 49400, loss[loss=2.342, over 2870.00 frames. , ppl: 10.400218529513513] tot_loss[loss=2.32, over 5547044.60 frames. , ppl: 10.174020052462378], batch size: 70
+2022-12-10 23:38:10,928 INFO [train.py:421] (2/8) Epoch 3, batch 49600, loss[loss=2.272, over 3780.00 frames. , ppl: 9.70069391553753] tot_loss[loss=2.32, over 5553398.38 frames. , ppl: 10.17089503236166], batch size: 70
+2022-12-10 23:39:52,726 INFO [train.py:421] (2/8) Epoch 3, batch 49800, loss[loss=2.515, over 1120.00 frames. , ppl: 12.369135453591461] tot_loss[loss=2.32, over 5533705.21 frames. , ppl: 10.180725299000088], batch size: 70
+2022-12-10 23:41:29,023 INFO [train.py:421] (2/8) Epoch 3, batch 50000, loss[loss=2.423, over 2590.00 frames. , ppl: 11.278322051384187] tot_loss[loss=2.321, over 5480257.62 frames. , ppl: 10.188115064182673], batch size: 70
+2022-12-10 23:41:29,023 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:41:29,779 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.074415768120824
+2022-12-10 23:43:13,634 INFO [train.py:421] (2/8) Epoch 3, batch 50200, loss[loss=2.374, over 1540.00 frames. , ppl: 10.742716014936196] tot_loss[loss=2.321, over 5500153.99 frames. , ppl: 10.182061740149686], batch size: 70
+2022-12-10 23:44:54,564 INFO [train.py:421] (2/8) Epoch 3, batch 50400, loss[loss=2.488, over 1610.00 frames. , ppl: 12.036978205441505] tot_loss[loss=2.321, over 5468167.85 frames. , ppl: 10.189908002100978], batch size: 70
+2022-12-10 23:46:36,439 INFO [train.py:421] (2/8) Epoch 3, batch 50600, loss[loss=2.45, over 1470.00 frames. , ppl: 11.588223544423943] tot_loss[loss=2.32, over 5498599.30 frames. , ppl: 10.17616295262813], batch size: 70
+2022-12-10 23:48:17,653 INFO [train.py:421] (2/8) Epoch 3, batch 50800, loss[loss=2.435, over 1540.00 frames. , ppl: 11.41795735272057] tot_loss[loss=2.321, over 5474125.25 frames. , ppl: 10.181423014468457], batch size: 70
+2022-12-10 23:49:58,970 INFO [train.py:421] (2/8) Epoch 3, batch 51000, loss[loss=2.298, over 4830.00 frames. , ppl: 9.954573714560047] tot_loss[loss=2.321, over 5460344.92 frames. , ppl: 10.186048367450825], batch size: 70
+2022-12-10 23:49:58,971 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:49:59,716 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.069029904707316
+2022-12-10 23:51:40,230 INFO [train.py:421] (2/8) Epoch 3, batch 51200, loss[loss=2.374, over 1120.00 frames. , ppl: 10.739236347627816] tot_loss[loss=2.322, over 5434922.65 frames. , ppl: 10.19329258784412], batch size: 70
+2022-12-10 23:53:19,770 INFO [train.py:421] (2/8) Epoch 3, batch 51400, loss[loss=2.324, over 2310.00 frames. , ppl: 10.219207704666182] tot_loss[loss=2.321, over 5466819.79 frames. , ppl: 10.187450379895646], batch size: 70
+2022-12-10 23:55:01,898 INFO [train.py:421] (2/8) Epoch 3, batch 51600, loss[loss=2.304, over 3220.00 frames. , ppl: 10.018308253921408] tot_loss[loss=2.32, over 5502531.05 frames. , ppl: 10.17316045694847], batch size: 70
+2022-12-10 23:56:41,075 INFO [train.py:421] (2/8) Epoch 3, batch 51800, loss[loss=2.175, over 8540.00 frames. , ppl: 8.803153470568224] tot_loss[loss=2.32, over 5495510.29 frames. , ppl: 10.177439563424382], batch size: 70
+2022-12-10 23:58:20,748 INFO [train.py:421] (2/8) Epoch 3, batch 52000, loss[loss=2.237, over 4970.00 frames. , ppl: 9.364480736163243] tot_loss[loss=2.321, over 5465227.57 frames. , ppl: 10.182199549912704], batch size: 70
+2022-12-10 23:58:20,749 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-10 23:58:21,496 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.064753676155952
+2022-12-11 00:00:01,575 INFO [train.py:421] (2/8) Epoch 3, batch 52200, loss[loss=2.241, over 5880.00 frames. , ppl: 9.406737424729638] tot_loss[loss=2.32, over 5473186.82 frames. , ppl: 10.176873320128756], batch size: 70
+2022-12-11 00:01:45,811 INFO [train.py:421] (2/8) Epoch 3, batch 52400, loss[loss=2.444, over 1400.00 frames. , ppl: 11.516022913341972] tot_loss[loss=2.32, over 5498514.68 frames. , ppl: 10.17093493409833], batch size: 70
+2022-12-11 00:03:24,670 INFO [train.py:421] (2/8) Epoch 3, batch 52600, loss[loss=2.496, over 1470.00 frames. , ppl: 12.13284867057936] tot_loss[loss=2.32, over 5488045.76 frames. , ppl: 10.17422252809664], batch size: 70
+2022-12-11 00:05:03,992 INFO [train.py:421] (2/8) Epoch 3, batch 52800, loss[loss=2.577, over 1190.00 frames. , ppl: 13.15236895089844] tot_loss[loss=2.319, over 5522082.38 frames. , ppl: 10.164169555417153], batch size: 70
+2022-12-11 00:06:42,493 INFO [train.py:421] (2/8) Epoch 3, batch 53000, loss[loss=2.266, over 6440.00 frames. , ppl: 9.639159924970965] tot_loss[loss=2.318, over 5525413.04 frames. , ppl: 10.155479142580058], batch size: 70
+2022-12-11 00:06:42,494 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:06:43,239 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.088516415422001
+2022-12-11 00:08:25,757 INFO [train.py:421] (2/8) Epoch 3, batch 53200, loss[loss=2.232, over 7630.00 frames. , ppl: 9.31770702531334] tot_loss[loss=2.319, over 5501250.15 frames. , ppl: 10.16348804439321], batch size: 70
+2022-12-11 00:10:06,549 INFO [train.py:421] (2/8) Epoch 3, batch 53400, loss[loss=2.493, over 1960.00 frames. , ppl: 12.102109824853196] tot_loss[loss=2.32, over 5492590.02 frames. , ppl: 10.17298623008694], batch size: 70
+2022-12-11 00:11:48,855 INFO [train.py:421] (2/8) Epoch 3, batch 53600, loss[loss=2.301, over 1960.00 frames. , ppl: 9.986449991596583] tot_loss[loss=2.319, over 5497654.44 frames. , ppl: 10.170534496486784], batch size: 70
+2022-12-11 00:13:30,371 INFO [train.py:421] (2/8) Epoch 3, batch 53800, loss[loss=2.504, over 1260.00 frames. , ppl: 12.229836024654386] tot_loss[loss=2.321, over 5456375.74 frames. , ppl: 10.185522536510431], batch size: 70
+2022-12-11 00:15:11,539 INFO [train.py:421] (2/8) Epoch 3, batch 54000, loss[loss=2.272, over 3010.00 frames. , ppl: 9.699857472653067] tot_loss[loss=2.321, over 5441954.04 frames. , ppl: 10.187150264664064], batch size: 70
+2022-12-11 00:15:11,539 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:15:12,298 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.032199299773053
+2022-12-11 00:16:54,776 INFO [train.py:421] (2/8) Epoch 3, batch 54200, loss[loss=2.268, over 4830.00 frames. , ppl: 9.66027901843153] tot_loss[loss=2.321, over 5460616.40 frames. , ppl: 10.183239000076505], batch size: 70
+2022-12-11 00:18:34,608 INFO [train.py:421] (2/8) Epoch 3, batch 54400, loss[loss=2.314, over 2660.00 frames. , ppl: 10.115756488858901] tot_loss[loss=2.32, over 5504035.68 frames. , ppl: 10.179836237766606], batch size: 70
+2022-12-11 00:20:14,978 INFO [train.py:421] (2/8) Epoch 3, batch 54600, loss[loss=2.265, over 4620.00 frames. , ppl: 9.631346098676493] tot_loss[loss=2.32, over 5518196.13 frames. , ppl: 10.176785826103414], batch size: 70
+2022-12-11 00:21:53,921 INFO [train.py:421] (2/8) Epoch 3, batch 54800, loss[loss=2.356, over 1680.00 frames. , ppl: 10.546177401066183] tot_loss[loss=2.322, over 5443957.54 frames. , ppl: 10.193130398195626], batch size: 70
+2022-12-11 00:23:33,768 INFO [train.py:421] (2/8) Epoch 3, batch 55000, loss[loss=2.373, over 4060.00 frames. , ppl: 10.72474601659482] tot_loss[loss=2.322, over 5392540.45 frames. , ppl: 10.200887484291057], batch size: 70
+2022-12-11 00:23:33,768 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:23:34,513 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.042009979542168
+2022-12-11 00:25:14,234 INFO [train.py:421] (2/8) Epoch 3, batch 55200, loss[loss=2.371, over 3640.00 frames. , ppl: 10.711420678818543] tot_loss[loss=2.323, over 5359989.40 frames. , ppl: 10.208214876380248], batch size: 70
+2022-12-11 00:26:51,652 INFO [train.py:421] (2/8) Epoch 3, batch 55400, loss[loss=2.283, over 3710.00 frames. , ppl: 9.80506769789096] tot_loss[loss=2.323, over 5365981.31 frames. , ppl: 10.201384643876171], batch size: 70
+2022-12-11 00:28:28,803 INFO [train.py:421] (2/8) Epoch 3, batch 55600, loss[loss=3.215, over 490.00 frames. , ppl: 24.898166706266032] tot_loss[loss=2.322, over 5364310.85 frames. , ppl: 10.199788530289405], batch size: 70
+2022-12-11 00:30:05,107 INFO [train.py:421] (2/8) Epoch 3, batch 55800, loss[loss=2.155, over 5740.00 frames. , ppl: 8.625075523612725] tot_loss[loss=2.323, over 5355654.35 frames. , ppl: 10.204107339950832], batch size: 70
+2022-12-11 00:31:49,168 INFO [train.py:421] (2/8) Epoch 3, batch 56000, loss[loss=2.204, over 6230.00 frames. , ppl: 9.065200549834755] tot_loss[loss=2.321, over 5432500.32 frames. , ppl: 10.189565586574588], batch size: 70
+2022-12-11 00:31:49,168 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:31:49,915 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.029381468540643
+2022-12-11 00:33:27,282 INFO [train.py:421] (2/8) Epoch 3, batch 56200, loss[loss=2.434, over 2310.00 frames. , ppl: 11.399945177543966] tot_loss[loss=2.321, over 5439835.34 frames. , ppl: 10.18524458527557], batch size: 70
+2022-12-11 00:35:04,950 INFO [train.py:421] (2/8) Epoch 3, batch 56400, loss[loss=2.224, over 5740.00 frames. , ppl: 9.242895257428202] tot_loss[loss=2.321, over 5436776.50 frames. , ppl: 10.183490501231613], batch size: 70
+2022-12-11 00:36:45,142 INFO [train.py:421] (2/8) Epoch 3, batch 56600, loss[loss=2.286, over 6020.00 frames. , ppl: 9.836133861419116] tot_loss[loss=2.322, over 5414367.60 frames. , ppl: 10.19240311845507], batch size: 70
+2022-12-11 00:38:27,987 INFO [train.py:421] (2/8) Epoch 3, batch 56800, loss[loss=2.245, over 4130.00 frames. , ppl: 9.441615685563105] tot_loss[loss=2.321, over 5426850.46 frames. , ppl: 10.18732933590061], batch size: 70
+2022-12-11 00:40:12,326 INFO [train.py:421] (2/8) Epoch 3, batch 57000, loss[loss=2.264, over 3150.00 frames. , ppl: 9.625662064116835] tot_loss[loss=2.319, over 5472632.28 frames. , ppl: 10.169766286281046], batch size: 70
+2022-12-11 00:40:12,326 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:40:13,085 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.04296571105421
+2022-12-11 00:41:53,207 INFO [train.py:421] (2/8) Epoch 3, batch 57200, loss[loss=4.793, over 292.00 frames. , ppl: 120.70868855551744] tot_loss[loss=2.318, over 5487435.49 frames. , ppl: 10.159389703762836], batch size: 70
+2022-12-11 00:43:33,443 INFO [train.py:421] (2/8) Epoch 3, batch 57400, loss[loss=2.304, over 2870.00 frames. , ppl: 10.011215391087411] tot_loss[loss=2.319, over 5472760.37 frames. , ppl: 10.169195059243531], batch size: 70
+2022-12-11 00:45:12,036 INFO [train.py:421] (2/8) Epoch 3, batch 57600, loss[loss=2.36, over 2730.00 frames. , ppl: 10.590882501151379] tot_loss[loss=2.319, over 5475594.91 frames. , ppl: 10.167110540510029], batch size: 70
+2022-12-11 00:46:53,805 INFO [train.py:421] (2/8) Epoch 3, batch 57800, loss[loss=2.351, over 2870.00 frames. , ppl: 10.494907452819993] tot_loss[loss=2.32, over 5447861.06 frames. , ppl: 10.175463894651648], batch size: 70
+2022-12-11 00:48:33,991 INFO [train.py:421] (2/8) Epoch 3, batch 58000, loss[loss=2.395, over 1680.00 frames. , ppl: 10.966541146788297] tot_loss[loss=2.319, over 5450844.94 frames. , ppl: 10.16956660768477], batch size: 70
+2022-12-11 00:48:33,992 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:48:34,750 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.037739283472558
+2022-12-11 00:50:13,898 INFO [train.py:421] (2/8) Epoch 3, batch 58200, loss[loss=2.218, over 1610.00 frames. , ppl: 9.189221542282423] tot_loss[loss=2.319, over 5468703.42 frames. , ppl: 10.161121592582191], batch size: 70
+2022-12-11 00:51:51,052 INFO [train.py:421] (2/8) Epoch 3, batch 58400, loss[loss=2.269, over 1890.00 frames. , ppl: 9.666764671798035] tot_loss[loss=2.319, over 5426975.36 frames. , ppl: 10.16900266357281], batch size: 70
+2022-12-11 00:53:28,852 INFO [train.py:421] (2/8) Epoch 3, batch 58600, loss[loss=2.309, over 6090.00 frames. , ppl: 10.068816282499723] tot_loss[loss=2.319, over 5442880.47 frames. , ppl: 10.169057039840277], batch size: 70
+2022-12-11 00:55:08,239 INFO [train.py:421] (2/8) Epoch 3, batch 58800, loss[loss=2.581, over 770.00 frames. , ppl: 13.212059230669105] tot_loss[loss=2.319, over 5429485.04 frames. , ppl: 10.169459339305003], batch size: 70
+2022-12-11 00:56:50,214 INFO [train.py:421] (2/8) Epoch 3, batch 59000, loss[loss=2.385, over 1960.00 frames. , ppl: 10.861381847020134] tot_loss[loss=2.32, over 5417538.48 frames. , ppl: 10.17100996142633], batch size: 70
+2022-12-11 00:56:50,214 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 00:56:50,962 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.063399667509817
+2022-12-11 00:58:29,069 INFO [train.py:421] (2/8) Epoch 3, batch 59200, loss[loss=2.623, over 770.00 frames. , ppl: 13.778488123460788] tot_loss[loss=2.321, over 5365097.90 frames. , ppl: 10.186473875037398], batch size: 70
+2022-12-11 01:00:05,132 INFO [train.py:421] (2/8) Epoch 3, batch 59400, loss[loss=2.415, over 1260.00 frames. , ppl: 11.194913539770424] tot_loss[loss=2.321, over 5363737.99 frames. , ppl: 10.187108557475495], batch size: 70
+2022-12-11 01:01:44,064 INFO [train.py:421] (2/8) Epoch 3, batch 59600, loss[loss=2.546, over 980.00 frames. , ppl: 12.75573401585891] tot_loss[loss=2.321, over 5366374.11 frames. , ppl: 10.182225755251753], batch size: 70
+2022-12-11 01:03:24,733 INFO [train.py:421] (2/8) Epoch 3, batch 59800, loss[loss=2.298, over 3290.00 frames. , ppl: 9.951583101440319] tot_loss[loss=2.32, over 5398085.00 frames. , ppl: 10.176775746208087], batch size: 70
+2022-12-11 01:05:04,163 INFO [train.py:421] (2/8) Epoch 3, batch 60000, loss[loss=2.824, over 630.00 frames. , ppl: 16.84321441186948] tot_loss[loss=2.32, over 5378062.26 frames. , ppl: 10.179585095215607], batch size: 70
+2022-12-11 01:05:04,164 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:05:04,908 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.045027606490608
+2022-12-11 01:06:48,968 INFO [train.py:421] (2/8) Epoch 3, batch 60200, loss[loss=2.468, over 1680.00 frames. , ppl: 11.793811301543984] tot_loss[loss=2.319, over 5443177.70 frames. , ppl: 10.161346017100513], batch size: 70
+2022-12-11 01:08:28,494 INFO [train.py:421] (2/8) Epoch 3, batch 60400, loss[loss=2.35, over 3220.00 frames. , ppl: 10.483159511356105] tot_loss[loss=2.318, over 5465495.39 frames. , ppl: 10.155868236112122], batch size: 70
+2022-12-11 01:10:08,957 INFO [train.py:421] (2/8) Epoch 3, batch 60600, loss[loss=2.468, over 1540.00 frames. , ppl: 11.79331428162777] tot_loss[loss=2.319, over 5443886.33 frames. , ppl: 10.16254370310335], batch size: 70
+2022-12-11 01:11:50,384 INFO [train.py:421] (2/8) Epoch 3, batch 60800, loss[loss=2.331, over 1540.00 frames. , ppl: 10.290792098660617] tot_loss[loss=2.318, over 5477323.65 frames. , ppl: 10.155576404291793], batch size: 70
+2022-12-11 01:13:29,059 INFO [train.py:421] (2/8) Epoch 3, batch 61000, loss[loss=2.529, over 1190.00 frames. , ppl: 12.54119314469717] tot_loss[loss=2.318, over 5513583.59 frames. , ppl: 10.15250060027236], batch size: 70
+2022-12-11 01:13:29,060 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:13:29,818 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.0369311167839
+2022-12-11 01:15:11,766 INFO [train.py:421] (2/8) Epoch 3, batch 61200, loss[loss=2.359, over 2450.00 frames. , ppl: 10.576232803285722] tot_loss[loss=2.318, over 5509739.46 frames. , ppl: 10.155275843718918], batch size: 70
+2022-12-11 01:16:52,649 INFO [train.py:421] (2/8) Epoch 3, batch 61400, loss[loss=2.536, over 1190.00 frames. , ppl: 12.633991302016389] tot_loss[loss=2.319, over 5479255.78 frames. , ppl: 10.168461797925248], batch size: 70
+2022-12-11 01:18:30,897 INFO [train.py:421] (2/8) Epoch 3, batch 61600, loss[loss=2.715, over 560.00 frames. , ppl: 15.106727984471917] tot_loss[loss=2.319, over 5478724.73 frames. , ppl: 10.16960157356837], batch size: 70
+2022-12-11 01:20:08,760 INFO [train.py:421] (2/8) Epoch 3, batch 61800, loss[loss=2.524, over 1120.00 frames. , ppl: 12.478803723523267] tot_loss[loss=2.32, over 5457386.56 frames. , ppl: 10.179504624145798], batch size: 70
+2022-12-11 01:21:46,221 INFO [train.py:421] (2/8) Epoch 3, batch 62000, loss[loss=2.217, over 4200.00 frames. , ppl: 9.178965682470833] tot_loss[loss=2.32, over 5452829.67 frames. , ppl: 10.171880613345454], batch size: 70
+2022-12-11 01:21:46,221 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:21:46,967 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.04664679581944
+2022-12-11 01:23:28,132 INFO [train.py:421] (2/8) Epoch 3, batch 62200, loss[loss=2.535, over 1400.00 frames. , ppl: 12.616912685912988] tot_loss[loss=2.32, over 5475174.70 frames. , ppl: 10.171976417384874], batch size: 70
+2022-12-11 01:25:03,203 INFO [train.py:421] (2/8) Epoch 3, batch 62400, loss[loss=2.214, over 5740.00 frames. , ppl: 9.153990729433088] tot_loss[loss=2.321, over 5432944.75 frames. , ppl: 10.18457446664341], batch size: 70
+2022-12-11 01:26:46,722 INFO [train.py:421] (2/8) Epoch 3, batch 62600, loss[loss=2.445, over 2450.00 frames. , ppl: 11.525352647109234] tot_loss[loss=2.32, over 5446107.74 frames. , ppl: 10.177695934783179], batch size: 70
+2022-12-11 01:28:27,835 INFO [train.py:421] (2/8) Epoch 3, batch 62800, loss[loss=2.243, over 3920.00 frames. , ppl: 9.42208537358746] tot_loss[loss=2.319, over 5474018.93 frames. , ppl: 10.170176445718091], batch size: 70
+2022-12-11 01:30:02,598 INFO [train.py:421] (2/8) Epoch 3, batch 63000, loss[loss=2.474, over 1330.00 frames. , ppl: 11.86901359645342] tot_loss[loss=2.321, over 5436633.93 frames. , ppl: 10.180797993530424], batch size: 70
+2022-12-11 01:30:02,599 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:30:03,344 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.036726114257576
+2022-12-11 01:31:44,167 INFO [train.py:421] (2/8) Epoch 3, batch 63200, loss[loss=2.262, over 5950.00 frames. , ppl: 9.60567677275079] tot_loss[loss=2.321, over 5426532.70 frames. , ppl: 10.185661004153147], batch size: 70
+2022-12-11 01:33:23,261 INFO [train.py:421] (2/8) Epoch 3, batch 63400, loss[loss=2.463, over 1050.00 frames. , ppl: 11.739805636007384] tot_loss[loss=2.321, over 5415715.03 frames. , ppl: 10.187120764066922], batch size: 70
+2022-12-11 01:35:03,944 INFO [train.py:421] (2/8) Epoch 3, batch 63600, loss[loss=2.717, over 770.00 frames. , ppl: 15.128940352234292] tot_loss[loss=2.32, over 5448528.52 frames. , ppl: 10.178930863992528], batch size: 70
+2022-12-11 01:36:42,933 INFO [train.py:421] (2/8) Epoch 3, batch 63800, loss[loss=2.412, over 1120.00 frames. , ppl: 11.151769730843311] tot_loss[loss=2.319, over 5458393.91 frames. , ppl: 10.169311441686952], batch size: 70
+2022-12-11 01:38:28,651 INFO [train.py:421] (2/8) Epoch 3, batch 64000, loss[loss=2.506, over 980.00 frames. , ppl: 12.257758336032362] tot_loss[loss=2.32, over 5432563.76 frames. , ppl: 10.174325074953861], batch size: 70
+2022-12-11 01:38:28,652 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:38:29,412 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.036485464460732
+2022-12-11 01:40:09,936 INFO [train.py:421] (2/8) Epoch 3, batch 64200, loss[loss=2.744, over 770.00 frames. , ppl: 15.552772450684106] tot_loss[loss=2.318, over 5488958.24 frames. , ppl: 10.157906910787508], batch size: 70
+2022-12-11 01:41:46,111 INFO [train.py:421] (2/8) Epoch 3, batch 64400, loss[loss=2.297, over 7490.00 frames. , ppl: 9.941236625117387] tot_loss[loss=2.319, over 5474711.95 frames. , ppl: 10.164050704476654], batch size: 70
+2022-12-11 01:43:25,726 INFO [train.py:421] (2/8) Epoch 3, batch 64600, loss[loss=2.473, over 1470.00 frames. , ppl: 11.859730659278346] tot_loss[loss=2.319, over 5479233.08 frames. , ppl: 10.168227735085305], batch size: 70
+2022-12-11 01:45:05,832 INFO [train.py:421] (2/8) Epoch 3, batch 64800, loss[loss=2.26, over 9520.00 frames. , ppl: 9.581657189551525] tot_loss[loss=2.32, over 5472210.93 frames. , ppl: 10.17359985838921], batch size: 70
+2022-12-11 01:46:48,564 INFO [train.py:421] (2/8) Epoch 3, batch 65000, loss[loss=2.602, over 700.00 frames. , ppl: 13.492233961580407] tot_loss[loss=2.319, over 5484222.32 frames. , ppl: 10.167952670462308], batch size: 70
+2022-12-11 01:46:48,565 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:46:49,313 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.022263217124767
+2022-12-11 01:48:29,590 INFO [train.py:421] (2/8) Epoch 3, batch 65200, loss[loss=2.256, over 5250.00 frames. , ppl: 9.54213329121046] tot_loss[loss=2.319, over 5481341.23 frames. , ppl: 10.163803242064178], batch size: 70
+2022-12-11 01:50:10,536 INFO [train.py:421] (2/8) Epoch 3, batch 65400, loss[loss=2.308, over 3920.00 frames. , ppl: 10.049726699452798] tot_loss[loss=2.319, over 5469349.44 frames. , ppl: 10.170434660562458], batch size: 70
+2022-12-11 01:51:49,275 INFO [train.py:421] (2/8) Epoch 3, batch 65600, loss[loss=2.349, over 1680.00 frames. , ppl: 10.472563177244021] tot_loss[loss=2.32, over 5455746.18 frames. , ppl: 10.170980735023669], batch size: 70
+2022-12-11 01:53:27,733 INFO [train.py:421] (2/8) Epoch 3, batch 65800, loss[loss=2.556, over 1400.00 frames. , ppl: 12.887165116170955] tot_loss[loss=2.319, over 5456746.24 frames. , ppl: 10.16929122878556], batch size: 70
+2022-12-11 01:55:11,465 INFO [train.py:421] (2/8) Epoch 3, batch 66000, loss[loss=2.582, over 980.00 frames. , ppl: 13.228664143702998] tot_loss[loss=2.317, over 5551845.25 frames. , ppl: 10.145721071550538], batch size: 70
+2022-12-11 01:55:11,466 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 01:55:12,223 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.026417513222102
+2022-12-11 01:56:53,388 INFO [train.py:421] (2/8) Epoch 3, batch 66200, loss[loss=2.326, over 4130.00 frames. , ppl: 10.236124933078454] tot_loss[loss=2.317, over 5582965.17 frames. , ppl: 10.14393971243112], batch size: 70
+2022-12-11 01:58:33,170 INFO [train.py:421] (2/8) Epoch 3, batch 66400, loss[loss=2.301, over 5040.00 frames. , ppl: 9.98671951288534] tot_loss[loss=2.318, over 5515302.07 frames. , ppl: 10.160076653908254], batch size: 70
+2022-12-11 02:00:07,972 INFO [train.py:421] (2/8) Epoch 3, batch 66600, loss[loss=2.308, over 2800.00 frames. , ppl: 10.056229699463378] tot_loss[loss=2.319, over 5481783.99 frames. , ppl: 10.164400420051019], batch size: 70
+2022-12-11 02:01:49,360 INFO [train.py:421] (2/8) Epoch 3, batch 66800, loss[loss=2.681, over 770.00 frames. , ppl: 14.592685816546208] tot_loss[loss=2.318, over 5502210.76 frames. , ppl: 10.159548157400183], batch size: 70
+2022-12-11 02:03:29,963 INFO [train.py:421] (2/8) Epoch 3, batch 67000, loss[loss=2.265, over 3710.00 frames. , ppl: 9.631391809303606] tot_loss[loss=2.318, over 5510394.11 frames. , ppl: 10.159579293439563], batch size: 70
+2022-12-11 02:03:29,963 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 02:03:30,724 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.023177014556284
+2022-12-11 02:05:08,154 INFO [train.py:421] (2/8) Epoch 3, batch 67200, loss[loss=2.389, over 1680.00 frames. , ppl: 10.904677617512522] tot_loss[loss=2.319, over 5479587.19 frames. , ppl: 10.1639765886865], batch size: 70
+2022-12-11 02:06:47,972 INFO [train.py:421] (2/8) Epoch 3, batch 67400, loss[loss=2.351, over 2730.00 frames. , ppl: 10.491931215926815] tot_loss[loss=2.318, over 5527601.77 frames. , ppl: 10.156842103673595], batch size: 70
+2022-12-11 02:08:29,679 INFO [train.py:421] (2/8) Epoch 3, batch 67600, loss[loss=2.475, over 1470.00 frames. , ppl: 11.879398531513523] tot_loss[loss=2.318, over 5521427.21 frames. , ppl: 10.160259437445148], batch size: 70
+2022-12-11 02:10:11,841 INFO [train.py:421] (2/8) Epoch 3, batch 67800, loss[loss=2.448, over 1400.00 frames. , ppl: 11.563234627374014] tot_loss[loss=2.317, over 5566875.76 frames. , ppl: 10.143728411284007], batch size: 70
+2022-12-11 02:11:53,373 INFO [train.py:421] (2/8) Epoch 3, batch 68000, loss[loss=2.348, over 2310.00 frames. , ppl: 10.467136549289895] tot_loss[loss=2.317, over 5568806.04 frames. , ppl: 10.145023517740242], batch size: 70
+2022-12-11 02:11:53,374 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 02:11:54,123 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.025138400067002
+2022-12-11 02:13:31,881 INFO [train.py:421] (2/8) Epoch 3, batch 68200, loss[loss=2.406, over 1680.00 frames. , ppl: 11.08429619271269] tot_loss[loss=2.317, over 5544529.80 frames. , ppl: 10.14704617858798], batch size: 70
+2022-12-11 02:15:11,916 INFO [train.py:421] (2/8) Epoch 3, batch 68400, loss[loss=2.395, over 2520.00 frames. , ppl: 10.970614280264709] tot_loss[loss=2.318, over 5547320.06 frames. , ppl: 10.15781865375574], batch size: 70
+2022-12-11 02:16:58,787 INFO [train.py:421] (2/8) Epoch 3, batch 68600, loss[loss=2.644, over 770.00 frames. , ppl: 14.064063508845958] tot_loss[loss=2.318, over 5566174.55 frames. , ppl: 10.155753813249389], batch size: 70
+2022-12-11 02:18:39,312 INFO [train.py:421] (2/8) Epoch 3, batch 68800, loss[loss=2.408, over 2240.00 frames. , ppl: 11.115533070878861] tot_loss[loss=2.319, over 5522375.37 frames. , ppl: 10.167798946870445], batch size: 70
+2022-12-11 02:20:20,721 INFO [train.py:421] (2/8) Epoch 3, batch 69000, loss[loss=2.208, over 3150.00 frames. , ppl: 9.094141030303861] tot_loss[loss=2.318, over 5534461.11 frames. , ppl: 10.15383739215458], batch size: 70
+2022-12-11 02:20:20,722 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 02:20:21,480 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.019322126504301
+2022-12-11 02:22:00,195 INFO [train.py:421] (2/8) Epoch 3, batch 69200, loss[loss=2.204, over 6160.00 frames. , ppl: 9.064879316090597] tot_loss[loss=2.319, over 5516541.23 frames. , ppl: 10.162666344306414], batch size: 70
+2022-12-11 02:23:41,999 INFO [train.py:421] (2/8) Epoch 3, batch 69400, loss[loss=2.387, over 1820.00 frames. , ppl: 10.880702337362186] tot_loss[loss=2.318, over 5551071.79 frames. , ppl: 10.151737413069677], batch size: 70
+2022-12-11 02:25:20,933 INFO [train.py:421] (2/8) Epoch 3, batch 69600, loss[loss=2.473, over 1120.00 frames. , ppl: 11.860312224058644] tot_loss[loss=2.318, over 5526298.01 frames. , ppl: 10.154487634907529], batch size: 70
+2022-12-11 02:27:00,942 INFO [train.py:421] (2/8) Epoch 3, batch 69800, loss[loss=2.399, over 2380.00 frames. , ppl: 11.017596369605476] tot_loss[loss=2.318, over 5513499.17 frames. , ppl: 10.158892986608526], batch size: 70
+2022-12-11 02:28:40,245 INFO [train.py:421] (2/8) Epoch 3, batch 70000, loss[loss=2.3, over 3920.00 frames. , ppl: 9.978489541374072] tot_loss[loss=2.318, over 5524292.55 frames. , ppl: 10.154428171155038], batch size: 70
+2022-12-11 02:28:40,246 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 02:28:40,991 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.023684386112267
+2022-12-11 02:30:18,974 INFO [train.py:421] (2/8) Epoch 3, batch 70200, loss[loss=2.639, over 1120.00 frames. , ppl: 13.99962988539597] tot_loss[loss=2.317, over 5544163.17 frames. , ppl: 10.149574460971184], batch size: 70
+2022-12-11 02:31:59,181 INFO [train.py:421] (2/8) Epoch 3, batch 70400, loss[loss=2.264, over 3780.00 frames. , ppl: 9.62231841178812] tot_loss[loss=2.318, over 5536451.51 frames. , ppl: 10.1509769187085], batch size: 70
+2022-12-11 02:33:37,898 INFO [train.py:421] (2/8) Epoch 3, batch 70600, loss[loss=2.315, over 3150.00 frames. , ppl: 10.122145349406608] tot_loss[loss=2.318, over 5552194.58 frames. , ppl: 10.150320778582092], batch size: 70
+2022-12-11 02:35:16,197 INFO [train.py:421] (2/8) Epoch 3, batch 70800, loss[loss=2.434, over 1820.00 frames. , ppl: 11.400088462205295] tot_loss[loss=2.316, over 5562815.24 frames. , ppl: 10.136260944455989], batch size: 70
+2022-12-11 02:36:54,122 INFO [train.py:421] (2/8) Epoch 3, batch 71000, loss[loss=2.416, over 1610.00 frames. , ppl: 11.205552724796517] tot_loss[loss=2.316, over 5544161.82 frames. , ppl: 10.134750448932357], batch size: 70
+2022-12-11 02:36:54,123 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 02:36:54,881 INFO [train.py:452] (2/8) Epoch 3, validation: loss=2.303, over 211138.00 frames. , ppl: 10.006615510581163
+2022-12-11 02:38:34,986 INFO [train.py:421] (2/8) Epoch 3, batch 71200, loss[loss=2.334, over 1820.00 frames. , ppl: 10.31412522918662] tot_loss[loss=2.314, over 5586821.37 frames. , ppl: 10.119103763165953], batch size: 70
+2022-12-11 02:40:17,300 INFO [train.py:421] (2/8) Epoch 3, batch 71400, loss[loss=2.253, over 4410.00 frames. , ppl: 9.512378784056397] tot_loss[loss=2.315, over 5541719.32 frames. , ppl: 10.126681612661146], batch size: 70
+2022-12-11 02:41:59,128 INFO [train.py:421] (2/8) Epoch 3, batch 71600, loss[loss=2.587, over 770.00 frames. , ppl: 13.289778837714167] tot_loss[loss=2.314, over 5583466.15 frames. , ppl: 10.11977394814855], batch size: 70
+2022-12-11 02:43:42,978 INFO [train.py:421] (2/8) Epoch 3, batch 71800, loss[loss=2.265, over 4410.00 frames. , ppl: 9.62911448767633] tot_loss[loss=2.315, over 5600660.69 frames. , ppl: 10.12635113691184], batch size: 70
+2022-12-11 02:44:57,811 INFO [train.py:421] (2/8) Epoch 4, batch 0, loss[loss=3.29, over 490.00 frames. , ppl: 26.844955486382663] tot_loss[loss=3.29, over 490.00 frames. , ppl: 26.844955486382663], batch size: 70
+2022-12-11 02:46:37,238 INFO [train.py:421] (2/8) Epoch 4, batch 200, loss[loss=2.533, over 910.00 frames. , ppl: 12.589298264155873] tot_loss[loss=2.322, over 486366.21 frames. , ppl: 10.195216305316478], batch size: 70
+2022-12-11 02:48:15,754 INFO [train.py:421] (2/8) Epoch 4, batch 400, loss[loss=2.802, over 630.00 frames. , ppl: 16.47013822410627] tot_loss[loss=2.313, over 976275.45 frames. , ppl: 10.106361444839484], batch size: 70
+2022-12-11 02:49:54,209 INFO [train.py:421] (2/8) Epoch 4, batch 600, loss[loss=2.361, over 1610.00 frames. , ppl: 10.60493524071632] tot_loss[loss=2.313, over 1420468.01 frames. , ppl: 10.102536376656941], batch size: 70
+2022-12-11 02:51:35,082 INFO [train.py:421] (2/8) Epoch 4, batch 800, loss[loss=2.396, over 2100.00 frames. , ppl: 10.976663117415166] tot_loss[loss=2.313, over 1775335.46 frames. , ppl: 10.106958106345237], batch size: 70
+2022-12-11 02:53:15,879 INFO [train.py:421] (2/8) Epoch 4, batch 1000, loss[loss=2.411, over 2030.00 frames. , ppl: 11.146572938683335] tot_loss[loss=2.309, over 2141586.95 frames. , ppl: 10.06930812170635], batch size: 70
+2022-12-11 02:53:15,880 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 02:53:16,642 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.006100117136109
+2022-12-11 02:54:58,857 INFO [train.py:421] (2/8) Epoch 4, batch 1200, loss[loss=2.367, over 1750.00 frames. , ppl: 10.670062942217445] tot_loss[loss=2.309, over 2479441.37 frames. , ppl: 10.063404020301075], batch size: 70
+2022-12-11 02:56:36,318 INFO [train.py:421] (2/8) Epoch 4, batch 1400, loss[loss=2.323, over 3010.00 frames. , ppl: 10.20174636805427] tot_loss[loss=2.308, over 2782467.70 frames. , ppl: 10.057811198343797], batch size: 70
+2022-12-11 02:58:17,517 INFO [train.py:421] (2/8) Epoch 4, batch 1600, loss[loss=2.257, over 7910.00 frames. , ppl: 9.55285097425726] tot_loss[loss=2.307, over 3065472.17 frames. , ppl: 10.044712811910465], batch size: 70
+2022-12-11 02:59:58,142 INFO [train.py:421] (2/8) Epoch 4, batch 1800, loss[loss=2.352, over 1820.00 frames. , ppl: 10.506385057683625] tot_loss[loss=2.307, over 3295442.23 frames. , ppl: 10.04613016246219], batch size: 70
+2022-12-11 03:01:36,803 INFO [train.py:421] (2/8) Epoch 4, batch 2000, loss[loss=2.252, over 5530.00 frames. , ppl: 9.50541176929968] tot_loss[loss=2.309, over 3488653.27 frames. , ppl: 10.062332366780696], batch size: 70
+2022-12-11 03:01:36,803 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:01:37,550 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.020806653868545
+2022-12-11 03:03:19,539 INFO [train.py:421] (2/8) Epoch 4, batch 2200, loss[loss=2.287, over 1610.00 frames. , ppl: 9.84400814339423] tot_loss[loss=2.308, over 3681981.97 frames. , ppl: 10.053431797528573], batch size: 70
+2022-12-11 03:04:59,303 INFO [train.py:421] (2/8) Epoch 4, batch 2400, loss[loss=2.434, over 1750.00 frames. , ppl: 11.405980633621693] tot_loss[loss=2.306, over 3878835.18 frames. , ppl: 10.038881200242328], batch size: 70
+2022-12-11 03:06:39,920 INFO [train.py:421] (2/8) Epoch 4, batch 2600, loss[loss=2.226, over 3150.00 frames. , ppl: 9.266629635608076] tot_loss[loss=2.307, over 4039187.94 frames. , ppl: 10.039472723692457], batch size: 70
+2022-12-11 03:08:21,017 INFO [train.py:421] (2/8) Epoch 4, batch 2800, loss[loss=2.206, over 4200.00 frames. , ppl: 9.079552666496372] tot_loss[loss=2.307, over 4169948.37 frames. , ppl: 10.047528701672995], batch size: 70
+2022-12-11 03:09:57,239 INFO [train.py:421] (2/8) Epoch 4, batch 3000, loss[loss=2.658, over 770.00 frames. , ppl: 14.265823593253472] tot_loss[loss=2.308, over 4245661.34 frames. , ppl: 10.057038413214931], batch size: 70
+2022-12-11 03:09:57,239 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:09:58,000 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.00204158109846
+2022-12-11 03:11:38,666 INFO [train.py:421] (2/8) Epoch 4, batch 3200, loss[loss=2.327, over 4480.00 frames. , ppl: 10.244154400535688] tot_loss[loss=2.308, over 4378364.60 frames. , ppl: 10.055576218733655], batch size: 70
+2022-12-11 03:13:21,691 INFO [train.py:421] (2/8) Epoch 4, batch 3400, loss[loss=2.239, over 13440.00 frames. , ppl: 9.384924669569857] tot_loss[loss=2.308, over 4507133.64 frames. , ppl: 10.058848331281279], batch size: 70
+2022-12-11 03:15:01,795 INFO [train.py:421] (2/8) Epoch 4, batch 3600, loss[loss=3.372, over 490.00 frames. , ppl: 29.13765139340937] tot_loss[loss=2.308, over 4614455.81 frames. , ppl: 10.052759828653091], batch size: 70
+2022-12-11 03:16:43,478 INFO [train.py:421] (2/8) Epoch 4, batch 3800, loss[loss=2.374, over 2310.00 frames. , ppl: 10.742787544267115] tot_loss[loss=2.309, over 4649265.23 frames. , ppl: 10.065409368708334], batch size: 70
+2022-12-11 03:18:23,730 INFO [train.py:421] (2/8) Epoch 4, batch 4000, loss[loss=2.261, over 3150.00 frames. , ppl: 9.595370014366146] tot_loss[loss=2.308, over 4766256.04 frames. , ppl: 10.055644314504693], batch size: 70
+2022-12-11 03:18:23,731 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:18:24,476 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.024831259142056
+2022-12-11 03:20:05,988 INFO [train.py:421] (2/8) Epoch 4, batch 4200, loss[loss=2.395, over 3500.00 frames. , ppl: 10.96345252482752] tot_loss[loss=2.308, over 4878250.58 frames. , ppl: 10.051193040354567], batch size: 70
+2022-12-11 03:21:49,026 INFO [train.py:421] (2/8) Epoch 4, batch 4400, loss[loss=2.342, over 2030.00 frames. , ppl: 10.402726708368094] tot_loss[loss=2.307, over 4993121.24 frames. , ppl: 10.04087711158418], batch size: 70
+2022-12-11 03:23:27,772 INFO [train.py:421] (2/8) Epoch 4, batch 4600, loss[loss=2.203, over 4200.00 frames. , ppl: 9.052608247991195] tot_loss[loss=2.307, over 5039624.34 frames. , ppl: 10.0429895487929], batch size: 70
+2022-12-11 03:25:07,823 INFO [train.py:421] (2/8) Epoch 4, batch 4800, loss[loss=2.466, over 1190.00 frames. , ppl: 11.771309957661087] tot_loss[loss=2.308, over 5072187.71 frames. , ppl: 10.058680308887675], batch size: 70
+2022-12-11 03:26:49,290 INFO [train.py:421] (2/8) Epoch 4, batch 5000, loss[loss=2.38, over 2730.00 frames. , ppl: 10.80623794435023] tot_loss[loss=2.308, over 5138185.94 frames. , ppl: 10.05310956379257], batch size: 70
+2022-12-11 03:26:49,290 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:26:50,036 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.007444935075092
+2022-12-11 03:28:31,298 INFO [train.py:421] (2/8) Epoch 4, batch 5200, loss[loss=2.231, over 10780.00 frames. , ppl: 9.311606153175237] tot_loss[loss=2.308, over 5188898.04 frames. , ppl: 10.051656415887512], batch size: 70
+2022-12-11 03:30:12,448 INFO [train.py:421] (2/8) Epoch 4, batch 5400, loss[loss=2.252, over 6720.00 frames. , ppl: 9.508931944450666] tot_loss[loss=2.307, over 5237153.73 frames. , ppl: 10.0491168318482], batch size: 70
+2022-12-11 03:31:52,290 INFO [train.py:421] (2/8) Epoch 4, batch 5600, loss[loss=2.214, over 4830.00 frames. , ppl: 9.15059463714673] tot_loss[loss=2.309, over 5235685.78 frames. , ppl: 10.059908446378778], batch size: 70
+2022-12-11 03:33:32,828 INFO [train.py:421] (2/8) Epoch 4, batch 5800, loss[loss=2.193, over 2870.00 frames. , ppl: 8.961807911984549] tot_loss[loss=2.31, over 5231634.43 frames. , ppl: 10.069648388353697], batch size: 70
+2022-12-11 03:35:14,355 INFO [train.py:421] (2/8) Epoch 4, batch 6000, loss[loss=2.446, over 1330.00 frames. , ppl: 11.53955797127059] tot_loss[loss=2.309, over 5263768.10 frames. , ppl: 10.06645562040811], batch size: 70
+2022-12-11 03:35:14,355 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:35:15,103 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.002629307959912
+2022-12-11 03:36:56,273 INFO [train.py:421] (2/8) Epoch 4, batch 6200, loss[loss=3.248, over 490.00 frames. , ppl: 25.746097022665154] tot_loss[loss=2.309, over 5269537.83 frames. , ppl: 10.069208672381409], batch size: 70
+2022-12-11 03:38:34,613 INFO [train.py:421] (2/8) Epoch 4, batch 6400, loss[loss=2.382, over 2100.00 frames. , ppl: 10.824551826037679] tot_loss[loss=2.309, over 5313024.44 frames. , ppl: 10.060705218061793], batch size: 70
+2022-12-11 03:40:16,694 INFO [train.py:421] (2/8) Epoch 4, batch 6600, loss[loss=2.353, over 2800.00 frames. , ppl: 10.522191493751695] tot_loss[loss=2.31, over 5288394.60 frames. , ppl: 10.073352218660398], batch size: 70
+2022-12-11 03:41:57,721 INFO [train.py:421] (2/8) Epoch 4, batch 6800, loss[loss=2.729, over 700.00 frames. , ppl: 15.316782903176572] tot_loss[loss=2.31, over 5329057.82 frames. , ppl: 10.072513110391576], batch size: 70
+2022-12-11 03:43:38,826 INFO [train.py:421] (2/8) Epoch 4, batch 7000, loss[loss=2.244, over 3500.00 frames. , ppl: 9.433432654232266] tot_loss[loss=2.31, over 5369098.86 frames. , ppl: 10.071898041621509], batch size: 70
+2022-12-11 03:43:38,826 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:43:39,587 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.304, over 211138.00 frames. , ppl: 10.018172918207874
+2022-12-11 03:45:21,082 INFO [train.py:421] (2/8) Epoch 4, batch 7200, loss[loss=2.294, over 4130.00 frames. , ppl: 9.919254933705252] tot_loss[loss=2.309, over 5417171.90 frames. , ppl: 10.065110372722208], batch size: 70
+2022-12-11 03:47:01,073 INFO [train.py:421] (2/8) Epoch 4, batch 7400, loss[loss=2.286, over 2310.00 frames. , ppl: 9.836640216129693] tot_loss[loss=2.31, over 5417095.63 frames. , ppl: 10.070668614866191], batch size: 70
+2022-12-11 03:48:40,870 INFO [train.py:421] (2/8) Epoch 4, batch 7600, loss[loss=2.404, over 1330.00 frames. , ppl: 11.065596724738327] tot_loss[loss=2.311, over 5379973.55 frames. , ppl: 10.08370429585872], batch size: 70
+2022-12-11 03:50:23,906 INFO [train.py:421] (2/8) Epoch 4, batch 7800, loss[loss=2.181, over 5180.00 frames. , ppl: 8.85737439528425] tot_loss[loss=2.312, over 5355719.69 frames. , ppl: 10.092058008336528], batch size: 70
+2022-12-11 03:52:01,577 INFO [train.py:421] (2/8) Epoch 4, batch 8000, loss[loss=2.358, over 1820.00 frames. , ppl: 10.564648292722854] tot_loss[loss=2.312, over 5344143.77 frames. , ppl: 10.099174254124769], batch size: 70
+2022-12-11 03:52:01,577 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 03:52:02,323 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.304, over 211138.00 frames. , ppl: 10.015572487058913
+2022-12-11 03:53:42,801 INFO [train.py:421] (2/8) Epoch 4, batch 8200, loss[loss=2.36, over 3290.00 frames. , ppl: 10.590423012047617] tot_loss[loss=2.312, over 5388557.78 frames. , ppl: 10.093809521979594], batch size: 70
+2022-12-11 03:55:20,997 INFO [train.py:421] (2/8) Epoch 4, batch 8400, loss[loss=2.282, over 13440.00 frames. , ppl: 9.79913662650255] tot_loss[loss=2.313, over 5412935.46 frames. , ppl: 10.107783403170828], batch size: 70
+2022-12-11 03:56:59,259 INFO [train.py:421] (2/8) Epoch 4, batch 8600, loss[loss=2.397, over 1190.00 frames. , ppl: 10.993030947531576] tot_loss[loss=2.315, over 5362414.11 frames. , ppl: 10.125977218128455], batch size: 70
+2022-12-11 03:58:39,457 INFO [train.py:421] (2/8) Epoch 4, batch 8800, loss[loss=2.539, over 1050.00 frames. , ppl: 12.668407909182141] tot_loss[loss=2.316, over 5334334.21 frames. , ppl: 10.135821463611956], batch size: 70
+2022-12-11 04:00:15,936 INFO [train.py:421] (2/8) Epoch 4, batch 9000, loss[loss=2.269, over 1960.00 frames. , ppl: 9.668255649777926] tot_loss[loss=2.317, over 5321162.25 frames. , ppl: 10.141232287682254], batch size: 70
+2022-12-11 04:00:15,936 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:00:16,697 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.02282839694073
+2022-12-11 04:01:59,796 INFO [train.py:421] (2/8) Epoch 4, batch 9200, loss[loss=2.51, over 1190.00 frames. , ppl: 12.299772615584436] tot_loss[loss=2.317, over 5306607.78 frames. , ppl: 10.144335134724972], batch size: 70
+2022-12-11 04:03:46,666 INFO [train.py:421] (2/8) Epoch 4, batch 9400, loss[loss=2.285, over 3570.00 frames. , ppl: 9.827248489537249] tot_loss[loss=2.316, over 5314089.08 frames. , ppl: 10.134506688722317], batch size: 70
+2022-12-11 04:05:23,936 INFO [train.py:421] (2/8) Epoch 4, batch 9600, loss[loss=2.266, over 5460.00 frames. , ppl: 9.641663026960254] tot_loss[loss=2.316, over 5298362.94 frames. , ppl: 10.138984744044503], batch size: 70
+2022-12-11 04:07:04,977 INFO [train.py:421] (2/8) Epoch 4, batch 9800, loss[loss=2.483, over 1470.00 frames. , ppl: 11.974325720534425] tot_loss[loss=2.314, over 5378319.88 frames. , ppl: 10.116731591838443], batch size: 70
+2022-12-11 04:08:44,054 INFO [train.py:421] (2/8) Epoch 4, batch 10000, loss[loss=2.247, over 5600.00 frames. , ppl: 9.459099519221976] tot_loss[loss=2.313, over 5428477.27 frames. , ppl: 10.102160373272026], batch size: 70
+2022-12-11 04:08:44,055 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:08:44,815 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.02091344153873
+2022-12-11 04:10:25,960 INFO [train.py:421] (2/8) Epoch 4, batch 10200, loss[loss=2.325, over 1890.00 frames. , ppl: 10.22810689849685] tot_loss[loss=2.313, over 5421035.28 frames. , ppl: 10.102481282749581], batch size: 70
+2022-12-11 04:12:05,325 INFO [train.py:421] (2/8) Epoch 4, batch 10400, loss[loss=2.218, over 4970.00 frames. , ppl: 9.185197811892465] tot_loss[loss=2.313, over 5431519.05 frames. , ppl: 10.100226988282916], batch size: 70
+2022-12-11 04:13:45,272 INFO [train.py:421] (2/8) Epoch 4, batch 10600, loss[loss=2.953, over 630.00 frames. , ppl: 19.16143107595757] tot_loss[loss=2.313, over 5436749.26 frames. , ppl: 10.100661353564353], batch size: 70
+2022-12-11 04:15:25,274 INFO [train.py:421] (2/8) Epoch 4, batch 10800, loss[loss=2.328, over 1960.00 frames. , ppl: 10.252669041779516] tot_loss[loss=2.312, over 5454436.69 frames. , ppl: 10.097622112642998], batch size: 70
+2022-12-11 04:17:08,163 INFO [train.py:421] (2/8) Epoch 4, batch 11000, loss[loss=2.519, over 1050.00 frames. , ppl: 12.41958942761692] tot_loss[loss=2.313, over 5437608.83 frames. , ppl: 10.107994746759728], batch size: 70
+2022-12-11 04:17:08,163 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:17:08,910 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.005364098386623
+2022-12-11 04:18:47,976 INFO [train.py:421] (2/8) Epoch 4, batch 11200, loss[loss=4.07, over 350.00 frames. , ppl: 58.55361330784481] tot_loss[loss=2.314, over 5411671.69 frames. , ppl: 10.117529050999616], batch size: 70
+2022-12-11 04:20:23,515 INFO [train.py:421] (2/8) Epoch 4, batch 11400, loss[loss=2.349, over 2240.00 frames. , ppl: 10.47151084198284] tot_loss[loss=2.314, over 5419941.09 frames. , ppl: 10.114498166442846], batch size: 70
+2022-12-11 04:22:05,494 INFO [train.py:421] (2/8) Epoch 4, batch 11600, loss[loss=2.53, over 1330.00 frames. , ppl: 12.550150491210431] tot_loss[loss=2.316, over 5390546.23 frames. , ppl: 10.133415757420716], batch size: 70
+2022-12-11 04:23:44,973 INFO [train.py:421] (2/8) Epoch 4, batch 11800, loss[loss=2.49, over 1190.00 frames. , ppl: 12.055423878413396] tot_loss[loss=2.316, over 5344186.40 frames. , ppl: 10.139756096314883], batch size: 70
+2022-12-11 04:25:25,251 INFO [train.py:421] (2/8) Epoch 4, batch 12000, loss[loss=3.634, over 420.00 frames. , ppl: 37.87462689259497] tot_loss[loss=2.316, over 5335576.95 frames. , ppl: 10.138385436509413], batch size: 70
+2022-12-11 04:25:25,252 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:25:25,998 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.98219907371343
+2022-12-11 04:27:07,827 INFO [train.py:421] (2/8) Epoch 4, batch 12200, loss[loss=2.212, over 3780.00 frames. , ppl: 9.132039199161914] tot_loss[loss=2.318, over 5285768.61 frames. , ppl: 10.158134829766633], batch size: 70
+2022-12-11 04:28:46,443 INFO [train.py:421] (2/8) Epoch 4, batch 12400, loss[loss=2.243, over 4620.00 frames. , ppl: 9.417934058351335] tot_loss[loss=2.318, over 5286337.22 frames. , ppl: 10.156973456492379], batch size: 70
+2022-12-11 04:30:27,175 INFO [train.py:421] (2/8) Epoch 4, batch 12600, loss[loss=2.245, over 6020.00 frames. , ppl: 9.445083563238326] tot_loss[loss=2.316, over 5320402.20 frames. , ppl: 10.139911548945147], batch size: 70
+2022-12-11 04:32:07,000 INFO [train.py:421] (2/8) Epoch 4, batch 12800, loss[loss=2.408, over 1820.00 frames. , ppl: 11.112953068203309] tot_loss[loss=2.315, over 5361154.30 frames. , ppl: 10.123134506982458], batch size: 70
+2022-12-11 04:33:50,382 INFO [train.py:421] (2/8) Epoch 4, batch 13000, loss[loss=2.5, over 1190.00 frames. , ppl: 12.185613564787975] tot_loss[loss=2.314, over 5402997.14 frames. , ppl: 10.11377966737028], batch size: 70
+2022-12-11 04:33:50,382 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:33:51,146 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.97099750372906
+2022-12-11 04:35:32,578 INFO [train.py:421] (2/8) Epoch 4, batch 13200, loss[loss=2.296, over 2590.00 frames. , ppl: 9.935802541004787] tot_loss[loss=2.314, over 5399916.54 frames. , ppl: 10.111252807693965], batch size: 70
+2022-12-11 04:37:11,847 INFO [train.py:421] (2/8) Epoch 4, batch 13400, loss[loss=2.343, over 1750.00 frames. , ppl: 10.40865379482148] tot_loss[loss=2.314, over 5397604.08 frames. , ppl: 10.118678140881453], batch size: 70
+2022-12-11 04:38:49,040 INFO [train.py:421] (2/8) Epoch 4, batch 13600, loss[loss=2.144, over 6090.00 frames. , ppl: 8.535779348118613] tot_loss[loss=2.315, over 5387086.18 frames. , ppl: 10.124203902018156], batch size: 70
+2022-12-11 04:40:27,700 INFO [train.py:421] (2/8) Epoch 4, batch 13800, loss[loss=2.566, over 910.00 frames. , ppl: 13.010824687543952] tot_loss[loss=2.314, over 5414926.69 frames. , ppl: 10.11653012838351], batch size: 70
+2022-12-11 04:42:07,519 INFO [train.py:421] (2/8) Epoch 4, batch 14000, loss[loss=2.257, over 3500.00 frames. , ppl: 9.554772201147427] tot_loss[loss=2.315, over 5386498.89 frames. , ppl: 10.12932061635242], batch size: 70
+2022-12-11 04:42:07,520 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:42:08,266 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.986692979928648
+2022-12-11 04:43:46,528 INFO [train.py:421] (2/8) Epoch 4, batch 14200, loss[loss=4.095, over 350.00 frames. , ppl: 60.04673145357088] tot_loss[loss=2.316, over 5348006.13 frames. , ppl: 10.132624841405764], batch size: 70
+2022-12-11 04:45:27,608 INFO [train.py:421] (2/8) Epoch 4, batch 14400, loss[loss=2.286, over 2730.00 frames. , ppl: 9.83711173533918] tot_loss[loss=2.316, over 5360630.33 frames. , ppl: 10.134715202088206], batch size: 70
+2022-12-11 04:47:07,723 INFO [train.py:421] (2/8) Epoch 4, batch 14600, loss[loss=2.239, over 4340.00 frames. , ppl: 9.380450336724161] tot_loss[loss=2.317, over 5302454.25 frames. , ppl: 10.14700189875959], batch size: 70
+2022-12-11 04:48:49,492 INFO [train.py:421] (2/8) Epoch 4, batch 14800, loss[loss=2.181, over 4690.00 frames. , ppl: 8.856257167234519] tot_loss[loss=2.317, over 5339276.55 frames. , ppl: 10.144051678761338], batch size: 70
+2022-12-11 04:50:23,944 INFO [train.py:421] (2/8) Epoch 4, batch 15000, loss[loss=2.39, over 2030.00 frames. , ppl: 10.917063021474032] tot_loss[loss=2.317, over 5319283.68 frames. , ppl: 10.140746100452759], batch size: 70
+2022-12-11 04:50:23,944 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:50:24,694 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.975186652961758
+2022-12-11 04:52:04,459 INFO [train.py:421] (2/8) Epoch 4, batch 15200, loss[loss=2.238, over 4270.00 frames. , ppl: 9.374528580642195] tot_loss[loss=2.317, over 5308611.41 frames. , ppl: 10.146358742328427], batch size: 70
+2022-12-11 04:53:41,643 INFO [train.py:421] (2/8) Epoch 4, batch 15400, loss[loss=2.208, over 3220.00 frames. , ppl: 9.09867796695057] tot_loss[loss=2.317, over 5304488.24 frames. , ppl: 10.143163712772], batch size: 70
+2022-12-11 04:55:25,248 INFO [train.py:421] (2/8) Epoch 4, batch 15600, loss[loss=2.211, over 8050.00 frames. , ppl: 9.123560219661758] tot_loss[loss=2.316, over 5353017.86 frames. , ppl: 10.13210585072025], batch size: 70
+2022-12-11 04:57:01,735 INFO [train.py:421] (2/8) Epoch 4, batch 15800, loss[loss=2.374, over 1540.00 frames. , ppl: 10.74149327765565] tot_loss[loss=2.315, over 5365631.17 frames. , ppl: 10.124728108405549], batch size: 70
+2022-12-11 04:58:46,153 INFO [train.py:421] (2/8) Epoch 4, batch 16000, loss[loss=2.153, over 7350.00 frames. , ppl: 8.609309536546423] tot_loss[loss=2.314, over 5383877.03 frames. , ppl: 10.116011067170772], batch size: 70
+2022-12-11 04:58:46,154 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 04:58:46,898 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.980866511809946
+2022-12-11 05:00:26,722 INFO [train.py:421] (2/8) Epoch 4, batch 16200, loss[loss=2.138, over 6160.00 frames. , ppl: 8.4830125927396] tot_loss[loss=2.313, over 5401173.12 frames. , ppl: 10.105982541869665], batch size: 70
+2022-12-11 05:02:10,985 INFO [train.py:421] (2/8) Epoch 4, batch 16400, loss[loss=2.264, over 2310.00 frames. , ppl: 9.624381950514701] tot_loss[loss=2.314, over 5379827.68 frames. , ppl: 10.119161613750348], batch size: 70
+2022-12-11 05:03:48,548 INFO [train.py:421] (2/8) Epoch 4, batch 16600, loss[loss=2.74, over 700.00 frames. , ppl: 15.490912435099824] tot_loss[loss=2.314, over 5404601.45 frames. , ppl: 10.113952984367119], batch size: 70
+2022-12-11 05:05:30,767 INFO [train.py:421] (2/8) Epoch 4, batch 16800, loss[loss=2.244, over 4060.00 frames. , ppl: 9.42857931942926] tot_loss[loss=2.315, over 5409405.39 frames. , ppl: 10.12190278069514], batch size: 70
+2022-12-11 05:07:12,257 INFO [train.py:421] (2/8) Epoch 4, batch 17000, loss[loss=2.439, over 2170.00 frames. , ppl: 11.46093839659551] tot_loss[loss=2.314, over 5429252.28 frames. , ppl: 10.117254735406181], batch size: 70
+2022-12-11 05:07:12,258 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:07:13,003 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.993913175932885
+2022-12-11 05:08:54,249 INFO [train.py:421] (2/8) Epoch 4, batch 17200, loss[loss=2.268, over 5320.00 frames. , ppl: 9.664463124883493] tot_loss[loss=2.313, over 5467984.49 frames. , ppl: 10.103036241947942], batch size: 70
+2022-12-11 05:10:34,554 INFO [train.py:421] (2/8) Epoch 4, batch 17400, loss[loss=2.393, over 1540.00 frames. , ppl: 10.943907945940275] tot_loss[loss=2.31, over 5558464.43 frames. , ppl: 10.075630356683995], batch size: 70
+2022-12-11 05:12:12,659 INFO [train.py:421] (2/8) Epoch 4, batch 17600, loss[loss=2.46, over 1330.00 frames. , ppl: 11.701926773617158] tot_loss[loss=2.309, over 5555219.82 frames. , ppl: 10.06848895208552], batch size: 70
+2022-12-11 05:13:51,762 INFO [train.py:421] (2/8) Epoch 4, batch 17800, loss[loss=2.318, over 2520.00 frames. , ppl: 10.155061363711146] tot_loss[loss=2.311, over 5489799.93 frames. , ppl: 10.088681892765717], batch size: 70
+2022-12-11 05:15:30,814 INFO [train.py:421] (2/8) Epoch 4, batch 18000, loss[loss=2.499, over 1050.00 frames. , ppl: 12.165165051744012] tot_loss[loss=2.312, over 5478973.30 frames. , ppl: 10.09544553867245], batch size: 70
+2022-12-11 05:15:30,815 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:15:31,578 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.995456073238609
+2022-12-11 05:17:10,754 INFO [train.py:421] (2/8) Epoch 4, batch 18200, loss[loss=2.371, over 2240.00 frames. , ppl: 10.711207126997849] tot_loss[loss=2.311, over 5507248.75 frames. , ppl: 10.082921512370232], batch size: 70
+2022-12-11 05:18:48,983 INFO [train.py:421] (2/8) Epoch 4, batch 18400, loss[loss=2.648, over 910.00 frames. , ppl: 14.123518232229634] tot_loss[loss=2.312, over 5460157.58 frames. , ppl: 10.094563487249367], batch size: 70
+2022-12-11 05:20:26,264 INFO [train.py:421] (2/8) Epoch 4, batch 18600, loss[loss=2.801, over 630.00 frames. , ppl: 16.462592543951462] tot_loss[loss=2.313, over 5439519.28 frames. , ppl: 10.104785147670231], batch size: 70
+2022-12-11 05:22:06,444 INFO [train.py:421] (2/8) Epoch 4, batch 18800, loss[loss=2.374, over 1470.00 frames. , ppl: 10.742718934491402] tot_loss[loss=2.313, over 5437874.20 frames. , ppl: 10.104331910607742], batch size: 70
+2022-12-11 05:23:48,711 INFO [train.py:421] (2/8) Epoch 4, batch 19000, loss[loss=2.591, over 1190.00 frames. , ppl: 13.345232929344228] tot_loss[loss=2.312, over 5440645.56 frames. , ppl: 10.097703731538406], batch size: 70
+2022-12-11 05:23:48,712 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:23:49,478 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.977298131379836
+2022-12-11 05:25:29,358 INFO [train.py:421] (2/8) Epoch 4, batch 19200, loss[loss=2.22, over 7840.00 frames. , ppl: 9.210633094168687] tot_loss[loss=2.311, over 5469050.64 frames. , ppl: 10.087689121151119], batch size: 70
+2022-12-11 05:27:11,528 INFO [train.py:421] (2/8) Epoch 4, batch 19400, loss[loss=2.281, over 3150.00 frames. , ppl: 9.789670654645347] tot_loss[loss=2.311, over 5470999.39 frames. , ppl: 10.084504514829206], batch size: 70
+2022-12-11 05:28:50,724 INFO [train.py:421] (2/8) Epoch 4, batch 19600, loss[loss=2.15, over 4830.00 frames. , ppl: 8.582687254145323] tot_loss[loss=2.312, over 5454578.41 frames. , ppl: 10.094130236758124], batch size: 70
+2022-12-11 05:30:32,725 INFO [train.py:421] (2/8) Epoch 4, batch 19800, loss[loss=2.236, over 4270.00 frames. , ppl: 9.352244806686457] tot_loss[loss=2.31, over 5511320.22 frames. , ppl: 10.073923753440756], batch size: 70
+2022-12-11 05:32:10,679 INFO [train.py:421] (2/8) Epoch 4, batch 20000, loss[loss=2.238, over 12740.00 frames. , ppl: 9.37694879161519] tot_loss[loss=2.309, over 5561580.62 frames. , ppl: 10.06625232399244], batch size: 70
+2022-12-11 05:32:10,679 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:32:11,424 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.990693533416067
+2022-12-11 05:33:50,978 INFO [train.py:421] (2/8) Epoch 4, batch 20200, loss[loss=2.612, over 700.00 frames. , ppl: 13.627063201201569] tot_loss[loss=2.308, over 5588882.62 frames. , ppl: 10.058008614767619], batch size: 70
+2022-12-11 05:35:32,012 INFO [train.py:421] (2/8) Epoch 4, batch 20400, loss[loss=2.243, over 5600.00 frames. , ppl: 9.417008212457644] tot_loss[loss=2.309, over 5576979.77 frames. , ppl: 10.061486204233816], batch size: 70
+2022-12-11 05:37:16,459 INFO [train.py:421] (2/8) Epoch 4, batch 20600, loss[loss=2.424, over 840.00 frames. , ppl: 11.289097371610312] tot_loss[loss=2.309, over 5582545.10 frames. , ppl: 10.060823173803199], batch size: 70
+2022-12-11 05:38:53,403 INFO [train.py:421] (2/8) Epoch 4, batch 20800, loss[loss=2.358, over 1820.00 frames. , ppl: 10.569575544231041] tot_loss[loss=2.311, over 5507937.01 frames. , ppl: 10.086813804881945], batch size: 70
+2022-12-11 05:40:32,952 INFO [train.py:421] (2/8) Epoch 4, batch 21000, loss[loss=2.243, over 5460.00 frames. , ppl: 9.422325880424442] tot_loss[loss=2.312, over 5478346.82 frames. , ppl: 10.097159883793056], batch size: 70
+2022-12-11 05:40:32,952 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:40:33,720 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.982080879215324
+2022-12-11 05:42:16,404 INFO [train.py:421] (2/8) Epoch 4, batch 21200, loss[loss=2.246, over 3570.00 frames. , ppl: 9.450565695916] tot_loss[loss=2.312, over 5495886.93 frames. , ppl: 10.093123539089147], batch size: 70
+2022-12-11 05:43:57,340 INFO [train.py:421] (2/8) Epoch 4, batch 21400, loss[loss=2.48, over 1190.00 frames. , ppl: 11.945750486074301] tot_loss[loss=2.311, over 5519012.59 frames. , ppl: 10.080880910404527], batch size: 70
+2022-12-11 05:45:35,240 INFO [train.py:421] (2/8) Epoch 4, batch 21600, loss[loss=2.213, over 5040.00 frames. , ppl: 9.147292751083729] tot_loss[loss=2.31, over 5533937.03 frames. , ppl: 10.077633806616753], batch size: 70
+2022-12-11 05:47:14,448 INFO [train.py:421] (2/8) Epoch 4, batch 21800, loss[loss=2.225, over 11970.00 frames. , ppl: 9.256885178544236] tot_loss[loss=2.31, over 5523281.30 frames. , ppl: 10.07460388687934], batch size: 70
+2022-12-11 05:48:55,726 INFO [train.py:421] (2/8) Epoch 4, batch 22000, loss[loss=2.169, over 7070.00 frames. , ppl: 8.749093954647961] tot_loss[loss=2.31, over 5539480.86 frames. , ppl: 10.07256056789395], batch size: 70
+2022-12-11 05:48:55,726 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:48:56,472 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.988403293971805
+2022-12-11 05:50:35,529 INFO [train.py:421] (2/8) Epoch 4, batch 22200, loss[loss=2.769, over 630.00 frames. , ppl: 15.947489963582541] tot_loss[loss=2.312, over 5463220.31 frames. , ppl: 10.090923612665797], batch size: 70
+2022-12-11 05:52:14,294 INFO [train.py:421] (2/8) Epoch 4, batch 22400, loss[loss=2.239, over 3360.00 frames. , ppl: 9.382085596383208] tot_loss[loss=2.311, over 5463095.95 frames. , ppl: 10.086917397774585], batch size: 70
+2022-12-11 05:53:54,669 INFO [train.py:421] (2/8) Epoch 4, batch 22600, loss[loss=2.45, over 1540.00 frames. , ppl: 11.592080376660126] tot_loss[loss=2.312, over 5427697.65 frames. , ppl: 10.09635107782772], batch size: 70
+2022-12-11 05:55:36,838 INFO [train.py:421] (2/8) Epoch 4, batch 22800, loss[loss=2.358, over 2870.00 frames. , ppl: 10.570651979159756] tot_loss[loss=2.31, over 5486793.09 frames. , ppl: 10.077652149906127], batch size: 70
+2022-12-11 05:57:12,517 INFO [train.py:421] (2/8) Epoch 4, batch 23000, loss[loss=2.157, over 5670.00 frames. , ppl: 8.647722335864986] tot_loss[loss=2.31, over 5492419.51 frames. , ppl: 10.079144643711215], batch size: 70
+2022-12-11 05:57:12,518 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 05:57:13,263 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.298, over 211138.00 frames. , ppl: 9.957022795141143
+2022-12-11 05:58:56,278 INFO [train.py:421] (2/8) Epoch 4, batch 23200, loss[loss=2.416, over 1050.00 frames. , ppl: 11.196547465280315] tot_loss[loss=2.311, over 5482368.42 frames. , ppl: 10.087091609318975], batch size: 70
+2022-12-11 06:00:34,256 INFO [train.py:421] (2/8) Epoch 4, batch 23400, loss[loss=2.363, over 2170.00 frames. , ppl: 10.623680166501522] tot_loss[loss=2.311, over 5484945.58 frames. , ppl: 10.087686080662362], batch size: 70
+2022-12-11 06:02:14,832 INFO [train.py:421] (2/8) Epoch 4, batch 23600, loss[loss=2.612, over 1680.00 frames. , ppl: 13.630499867285915] tot_loss[loss=2.311, over 5476185.55 frames. , ppl: 10.088627191162882], batch size: 70
+2022-12-11 06:03:55,549 INFO [train.py:421] (2/8) Epoch 4, batch 23800, loss[loss=2.365, over 3360.00 frames. , ppl: 10.64550343881422] tot_loss[loss=2.311, over 5491535.69 frames. , ppl: 10.083694947390383], batch size: 70
+2022-12-11 06:05:34,129 INFO [train.py:421] (2/8) Epoch 4, batch 24000, loss[loss=2.364, over 1960.00 frames. , ppl: 10.631331093490443] tot_loss[loss=2.31, over 5538352.91 frames. , ppl: 10.076957359972054], batch size: 70
+2022-12-11 06:05:34,129 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:05:34,888 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.975346105661082
+2022-12-11 06:07:17,288 INFO [train.py:421] (2/8) Epoch 4, batch 24200, loss[loss=2.208, over 13370.00 frames. , ppl: 9.0956367816996] tot_loss[loss=2.309, over 5552381.17 frames. , ppl: 10.067239095465132], batch size: 70
+2022-12-11 06:08:57,901 INFO [train.py:421] (2/8) Epoch 4, batch 24400, loss[loss=2.196, over 7350.00 frames. , ppl: 8.992369237788466] tot_loss[loss=2.31, over 5552810.48 frames. , ppl: 10.074627564776614], batch size: 70
+2022-12-11 06:10:39,362 INFO [train.py:421] (2/8) Epoch 4, batch 24600, loss[loss=2.327, over 1400.00 frames. , ppl: 10.244956542940203] tot_loss[loss=2.31, over 5550515.03 frames. , ppl: 10.076206088911713], batch size: 70
+2022-12-11 06:12:21,102 INFO [train.py:421] (2/8) Epoch 4, batch 24800, loss[loss=2.404, over 1890.00 frames. , ppl: 11.063169723132578] tot_loss[loss=2.311, over 5498988.67 frames. , ppl: 10.08304020658716], batch size: 70
+2022-12-11 06:14:01,869 INFO [train.py:421] (2/8) Epoch 4, batch 25000, loss[loss=2.376, over 2590.00 frames. , ppl: 10.761710090291267] tot_loss[loss=2.31, over 5534709.14 frames. , ppl: 10.069473233394746], batch size: 70
+2022-12-11 06:14:01,869 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:14:02,628 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.974898758710204
+2022-12-11 06:15:42,489 INFO [train.py:421] (2/8) Epoch 4, batch 25200, loss[loss=2.422, over 2800.00 frames. , ppl: 11.273769282532651] tot_loss[loss=2.31, over 5540167.37 frames. , ppl: 10.07000686337678], batch size: 70
+2022-12-11 06:17:23,860 INFO [train.py:421] (2/8) Epoch 4, batch 25400, loss[loss=2.55, over 1260.00 frames. , ppl: 12.804982245826448] tot_loss[loss=2.31, over 5524952.94 frames. , ppl: 10.071174064805465], batch size: 70
+2022-12-11 06:19:05,299 INFO [train.py:421] (2/8) Epoch 4, batch 25600, loss[loss=2.249, over 3850.00 frames. , ppl: 9.479482421974271] tot_loss[loss=2.309, over 5538164.70 frames. , ppl: 10.06600271804959], batch size: 70
+2022-12-11 06:20:48,760 INFO [train.py:421] (2/8) Epoch 4, batch 25800, loss[loss=2.266, over 3500.00 frames. , ppl: 9.638522768933496] tot_loss[loss=2.309, over 5542753.86 frames. , ppl: 10.061378943313517], batch size: 70
+2022-12-11 06:22:27,270 INFO [train.py:421] (2/8) Epoch 4, batch 26000, loss[loss=2.241, over 5460.00 frames. , ppl: 9.399249344839165] tot_loss[loss=2.31, over 5518987.62 frames. , ppl: 10.070734343500268], batch size: 70
+2022-12-11 06:22:27,271 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:22:28,032 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.983234812864332
+2022-12-11 06:24:08,765 INFO [train.py:421] (2/8) Epoch 4, batch 26200, loss[loss=3.062, over 560.00 frames. , ppl: 21.368586219239777] tot_loss[loss=2.31, over 5499876.87 frames. , ppl: 10.07137082459992], batch size: 70
+2022-12-11 06:25:49,600 INFO [train.py:421] (2/8) Epoch 4, batch 26400, loss[loss=2.431, over 1050.00 frames. , ppl: 11.366158023815093] tot_loss[loss=2.31, over 5478335.76 frames. , ppl: 10.073145554548113], batch size: 70
+2022-12-11 06:27:29,119 INFO [train.py:421] (2/8) Epoch 4, batch 26600, loss[loss=2.336, over 1680.00 frames. , ppl: 10.336547832057011] tot_loss[loss=2.31, over 5505532.32 frames. , ppl: 10.077222409600587], batch size: 70
+2022-12-11 06:29:04,074 INFO [train.py:421] (2/8) Epoch 4, batch 26800, loss[loss=2.253, over 1680.00 frames. , ppl: 9.520896657788695] tot_loss[loss=2.311, over 5491529.06 frames. , ppl: 10.081901543963594], batch size: 70
+2022-12-11 06:30:43,400 INFO [train.py:421] (2/8) Epoch 4, batch 27000, loss[loss=2.34, over 1540.00 frames. , ppl: 10.379249324415657] tot_loss[loss=2.312, over 5441752.87 frames. , ppl: 10.095464266023502], batch size: 70
+2022-12-11 06:30:43,400 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:30:44,146 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.968636532001739
+2022-12-11 06:32:23,959 INFO [train.py:421] (2/8) Epoch 4, batch 27200, loss[loss=2.211, over 7210.00 frames. , ppl: 9.121992719524233] tot_loss[loss=2.311, over 5462797.12 frames. , ppl: 10.089527241252739], batch size: 70
+2022-12-11 06:34:06,619 INFO [train.py:421] (2/8) Epoch 4, batch 27400, loss[loss=2.373, over 2450.00 frames. , ppl: 10.73226627216695] tot_loss[loss=2.311, over 5465097.58 frames. , ppl: 10.087056392570666], batch size: 70
+2022-12-11 06:35:42,978 INFO [train.py:421] (2/8) Epoch 4, batch 27600, loss[loss=2.408, over 1260.00 frames. , ppl: 11.116746787870266] tot_loss[loss=2.313, over 5418836.43 frames. , ppl: 10.101572510755819], batch size: 70
+2022-12-11 06:37:23,603 INFO [train.py:421] (2/8) Epoch 4, batch 27800, loss[loss=2.398, over 840.00 frames. , ppl: 11.00409656880728] tot_loss[loss=2.311, over 5473660.76 frames. , ppl: 10.087667988473187], batch size: 70
+2022-12-11 06:39:01,492 INFO [train.py:421] (2/8) Epoch 4, batch 28000, loss[loss=2.279, over 3920.00 frames. , ppl: 9.76426821296031] tot_loss[loss=2.311, over 5451943.25 frames. , ppl: 10.086326325674984], batch size: 70
+2022-12-11 06:39:01,492 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:39:02,238 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.969008347954837
+2022-12-11 06:40:41,648 INFO [train.py:421] (2/8) Epoch 4, batch 28200, loss[loss=2.368, over 1540.00 frames. , ppl: 10.673157043200781] tot_loss[loss=2.311, over 5483731.82 frames. , ppl: 10.081728776399677], batch size: 70
+2022-12-11 06:42:23,114 INFO [train.py:421] (2/8) Epoch 4, batch 28400, loss[loss=2.205, over 3710.00 frames. , ppl: 9.066660983984688] tot_loss[loss=2.311, over 5493089.67 frames. , ppl: 10.08125671677948], batch size: 70
+2022-12-11 06:44:01,460 INFO [train.py:421] (2/8) Epoch 4, batch 28600, loss[loss=2.215, over 8610.00 frames. , ppl: 9.157077486392428] tot_loss[loss=2.31, over 5502529.47 frames. , ppl: 10.078898788066782], batch size: 70
+2022-12-11 06:45:43,568 INFO [train.py:421] (2/8) Epoch 4, batch 28800, loss[loss=2.247, over 5460.00 frames. , ppl: 9.462985862977517] tot_loss[loss=2.31, over 5498892.63 frames. , ppl: 10.078664181042049], batch size: 70
+2022-12-11 06:47:27,934 INFO [train.py:421] (2/8) Epoch 4, batch 29000, loss[loss=2.41, over 910.00 frames. , ppl: 11.135133341025483] tot_loss[loss=2.311, over 5497449.73 frames. , ppl: 10.08162446924917], batch size: 70
+2022-12-11 06:47:27,935 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:47:28,683 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.996510942600489
+2022-12-11 06:49:06,727 INFO [train.py:421] (2/8) Epoch 4, batch 29200, loss[loss=2.512, over 1610.00 frames. , ppl: 12.329723992276955] tot_loss[loss=2.313, over 5399574.49 frames. , ppl: 10.101458186583484], batch size: 70
+2022-12-11 06:50:43,153 INFO [train.py:421] (2/8) Epoch 4, batch 29400, loss[loss=2.222, over 4270.00 frames. , ppl: 9.225015876980676] tot_loss[loss=2.313, over 5381266.38 frames. , ppl: 10.103888649207004], batch size: 70
+2022-12-11 06:52:19,515 INFO [train.py:421] (2/8) Epoch 4, batch 29600, loss[loss=2.227, over 3430.00 frames. , ppl: 9.271420570592655] tot_loss[loss=2.312, over 5404451.77 frames. , ppl: 10.09420503947847], batch size: 70
+2022-12-11 06:53:56,647 INFO [train.py:421] (2/8) Epoch 4, batch 29800, loss[loss=2.303, over 4620.00 frames. , ppl: 10.004624384783027] tot_loss[loss=2.313, over 5387989.16 frames. , ppl: 10.103016395471155], batch size: 70
+2022-12-11 06:55:37,408 INFO [train.py:421] (2/8) Epoch 4, batch 30000, loss[loss=2.415, over 2660.00 frames. , ppl: 11.187701716423941] tot_loss[loss=2.313, over 5393456.22 frames. , ppl: 10.108456502760793], batch size: 70
+2022-12-11 06:55:37,409 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 06:55:38,160 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.970644798055698
+2022-12-11 06:57:17,485 INFO [train.py:421] (2/8) Epoch 4, batch 30200, loss[loss=2.27, over 2240.00 frames. , ppl: 9.680974114626851] tot_loss[loss=2.314, over 5385737.41 frames. , ppl: 10.118347245597347], batch size: 70
+2022-12-11 06:58:56,929 INFO [train.py:421] (2/8) Epoch 4, batch 30400, loss[loss=2.347, over 980.00 frames. , ppl: 10.454155569772603] tot_loss[loss=2.314, over 5404672.89 frames. , ppl: 10.115177988449686], batch size: 70
+2022-12-11 07:00:39,366 INFO [train.py:421] (2/8) Epoch 4, batch 30600, loss[loss=2.322, over 1890.00 frames. , ppl: 10.196186051832026] tot_loss[loss=2.312, over 5456245.04 frames. , ppl: 10.094653259142898], batch size: 70
+2022-12-11 07:02:17,072 INFO [train.py:421] (2/8) Epoch 4, batch 30800, loss[loss=2.24, over 4340.00 frames. , ppl: 9.395826976179864] tot_loss[loss=2.311, over 5478014.39 frames. , ppl: 10.089364979404172], batch size: 70
+2022-12-11 07:04:02,218 INFO [train.py:421] (2/8) Epoch 4, batch 31000, loss[loss=2.424, over 980.00 frames. , ppl: 11.290505521734453] tot_loss[loss=2.311, over 5486516.89 frames. , ppl: 10.084039322777596], batch size: 70
+2022-12-11 07:04:02,218 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:04:02,984 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.976530269338536
+2022-12-11 07:05:48,983 INFO [train.py:421] (2/8) Epoch 4, batch 31200, loss[loss=2.273, over 3710.00 frames. , ppl: 9.711624051099324] tot_loss[loss=2.31, over 5506481.53 frames. , ppl: 10.077212233094185], batch size: 70
+2022-12-11 07:07:34,367 INFO [train.py:421] (2/8) Epoch 4, batch 31400, loss[loss=2.368, over 1470.00 frames. , ppl: 10.67444952249524] tot_loss[loss=2.31, over 5477435.31 frames. , ppl: 10.079248966427311], batch size: 70
+2022-12-11 07:09:13,736 INFO [train.py:421] (2/8) Epoch 4, batch 31600, loss[loss=2.697, over 770.00 frames. , ppl: 14.837510339393974] tot_loss[loss=2.309, over 5509003.28 frames. , ppl: 10.069204345447542], batch size: 70
+2022-12-11 07:10:54,254 INFO [train.py:421] (2/8) Epoch 4, batch 31800, loss[loss=2.343, over 1470.00 frames. , ppl: 10.412170822235002] tot_loss[loss=2.308, over 5544412.55 frames. , ppl: 10.058843076461708], batch size: 70
+2022-12-11 07:12:34,096 INFO [train.py:421] (2/8) Epoch 4, batch 32000, loss[loss=2.456, over 1470.00 frames. , ppl: 11.652782371459715] tot_loss[loss=2.308, over 5573684.18 frames. , ppl: 10.056476873190654], batch size: 70
+2022-12-11 07:12:34,097 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:12:34,843 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.965888179968418
+2022-12-11 07:14:15,364 INFO [train.py:421] (2/8) Epoch 4, batch 32200, loss[loss=2.444, over 1960.00 frames. , ppl: 11.52183879560312] tot_loss[loss=2.309, over 5547582.21 frames. , ppl: 10.064024826936741], batch size: 70
+2022-12-11 07:15:52,080 INFO [train.py:421] (2/8) Epoch 4, batch 32400, loss[loss=2.469, over 1890.00 frames. , ppl: 11.811388030906251] tot_loss[loss=2.309, over 5526433.96 frames. , ppl: 10.06428406185985], batch size: 70
+2022-12-11 07:17:32,125 INFO [train.py:421] (2/8) Epoch 4, batch 32600, loss[loss=2.782, over 630.00 frames. , ppl: 16.144643279222176] tot_loss[loss=2.31, over 5495843.88 frames. , ppl: 10.071716735320823], batch size: 70
+2022-12-11 07:19:13,614 INFO [train.py:421] (2/8) Epoch 4, batch 32800, loss[loss=2.54, over 840.00 frames. , ppl: 12.675320893072058] tot_loss[loss=2.31, over 5492123.96 frames. , ppl: 10.0701356240134], batch size: 70
+2022-12-11 07:20:53,530 INFO [train.py:421] (2/8) Epoch 4, batch 33000, loss[loss=2.526, over 1750.00 frames. , ppl: 12.505339235673274] tot_loss[loss=2.31, over 5486579.14 frames. , ppl: 10.069616815673706], batch size: 70
+2022-12-11 07:20:53,530 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:20:54,297 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.98494896820417
+2022-12-11 07:22:34,966 INFO [train.py:421] (2/8) Epoch 4, batch 33200, loss[loss=2.298, over 2170.00 frames. , ppl: 9.958874385433528] tot_loss[loss=2.31, over 5448663.08 frames. , ppl: 10.077146030293202], batch size: 70
+2022-12-11 07:24:11,942 INFO [train.py:421] (2/8) Epoch 4, batch 33400, loss[loss=2.515, over 770.00 frames. , ppl: 12.365468596720785] tot_loss[loss=2.311, over 5434401.41 frames. , ppl: 10.080658113768443], batch size: 70
+2022-12-11 07:25:47,473 INFO [train.py:421] (2/8) Epoch 4, batch 33600, loss[loss=2.232, over 3430.00 frames. , ppl: 9.319817057664626] tot_loss[loss=2.31, over 5442415.13 frames. , ppl: 10.078487716544245], batch size: 70
+2022-12-11 07:27:27,222 INFO [train.py:421] (2/8) Epoch 4, batch 33800, loss[loss=2.14, over 4480.00 frames. , ppl: 8.50186912941104] tot_loss[loss=2.31, over 5470412.03 frames. , ppl: 10.073029059034498], batch size: 70
+2022-12-11 07:29:07,735 INFO [train.py:421] (2/8) Epoch 4, batch 34000, loss[loss=2.233, over 4410.00 frames. , ppl: 9.328533192514252] tot_loss[loss=2.309, over 5486681.47 frames. , ppl: 10.065786815886048], batch size: 70
+2022-12-11 07:29:07,735 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:29:08,495 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.963913316067512
+2022-12-11 07:30:52,351 INFO [train.py:421] (2/8) Epoch 4, batch 34200, loss[loss=2.251, over 10080.00 frames. , ppl: 9.493554211218001] tot_loss[loss=2.309, over 5484216.58 frames. , ppl: 10.068766117807687], batch size: 70
+2022-12-11 07:32:31,250 INFO [train.py:421] (2/8) Epoch 4, batch 34400, loss[loss=2.568, over 840.00 frames. , ppl: 13.038428429702904] tot_loss[loss=2.309, over 5478712.60 frames. , ppl: 10.06872952885477], batch size: 70
+2022-12-11 07:34:12,812 INFO [train.py:421] (2/8) Epoch 4, batch 34600, loss[loss=2.478, over 1120.00 frames. , ppl: 11.919651599330441] tot_loss[loss=2.31, over 5457461.63 frames. , ppl: 10.0756125442982], batch size: 70
+2022-12-11 07:35:50,515 INFO [train.py:421] (2/8) Epoch 4, batch 34800, loss[loss=2.291, over 2870.00 frames. , ppl: 9.88360483445957] tot_loss[loss=2.31, over 5472176.12 frames. , ppl: 10.078069085987934], batch size: 70
+2022-12-11 07:37:31,269 INFO [train.py:421] (2/8) Epoch 4, batch 35000, loss[loss=2.264, over 3640.00 frames. , ppl: 9.61677811635243] tot_loss[loss=2.31, over 5468297.40 frames. , ppl: 10.077788715591435], batch size: 70
+2022-12-11 07:37:31,269 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:37:32,028 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.97713864747801
+2022-12-11 07:39:12,241 INFO [train.py:421] (2/8) Epoch 4, batch 35200, loss[loss=2.65, over 840.00 frames. , ppl: 14.15717369346482] tot_loss[loss=2.311, over 5445960.06 frames. , ppl: 10.086721534900695], batch size: 70
+2022-12-11 07:40:54,303 INFO [train.py:421] (2/8) Epoch 4, batch 35400, loss[loss=2.161, over 5950.00 frames. , ppl: 8.683636466186123] tot_loss[loss=2.312, over 5387792.58 frames. , ppl: 10.098234419541463], batch size: 70
+2022-12-11 07:42:36,894 INFO [train.py:421] (2/8) Epoch 4, batch 35600, loss[loss=2.398, over 1260.00 frames. , ppl: 10.999497884314193] tot_loss[loss=2.313, over 5367729.34 frames. , ppl: 10.10643642194012], batch size: 70
+2022-12-11 07:44:17,399 INFO [train.py:421] (2/8) Epoch 4, batch 35800, loss[loss=2.47, over 1820.00 frames. , ppl: 11.824688893667348] tot_loss[loss=2.311, over 5424839.21 frames. , ppl: 10.085298390912088], batch size: 70
+2022-12-11 07:46:03,625 INFO [train.py:421] (2/8) Epoch 4, batch 36000, loss[loss=2.994, over 560.00 frames. , ppl: 19.97356432850703] tot_loss[loss=2.309, over 5503855.97 frames. , ppl: 10.05956763718359], batch size: 70
+2022-12-11 07:46:03,625 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:46:04,385 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.962801429018057
+2022-12-11 07:47:46,400 INFO [train.py:421] (2/8) Epoch 4, batch 36200, loss[loss=2.887, over 630.00 frames. , ppl: 17.939627742478333] tot_loss[loss=2.308, over 5498529.19 frames. , ppl: 10.05885102409319], batch size: 70
+2022-12-11 07:49:21,365 INFO [train.py:421] (2/8) Epoch 4, batch 36400, loss[loss=3.032, over 560.00 frames. , ppl: 20.744615638968806] tot_loss[loss=2.309, over 5506264.55 frames. , ppl: 10.059474814407169], batch size: 70
+2022-12-11 07:51:00,542 INFO [train.py:421] (2/8) Epoch 4, batch 36600, loss[loss=2.319, over 2660.00 frames. , ppl: 10.166313754045705] tot_loss[loss=2.308, over 5509625.34 frames. , ppl: 10.058605259500382], batch size: 70
+2022-12-11 07:52:42,134 INFO [train.py:421] (2/8) Epoch 4, batch 36800, loss[loss=2.448, over 1260.00 frames. , ppl: 11.569964867108666] tot_loss[loss=2.309, over 5483969.65 frames. , ppl: 10.060871151813803], batch size: 70
+2022-12-11 07:54:20,219 INFO [train.py:421] (2/8) Epoch 4, batch 37000, loss[loss=2.498, over 1330.00 frames. , ppl: 12.16046426158863] tot_loss[loss=2.309, over 5483018.56 frames. , ppl: 10.06811466966973], batch size: 70
+2022-12-11 07:54:20,219 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 07:54:20,987 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.297, over 211138.00 frames. , ppl: 9.946271967014194
+2022-12-11 07:56:00,677 INFO [train.py:421] (2/8) Epoch 4, batch 37200, loss[loss=3.408, over 490.00 frames. , ppl: 30.20848687843285] tot_loss[loss=2.31, over 5469534.10 frames. , ppl: 10.070052330282794], batch size: 70
+2022-12-11 07:57:39,426 INFO [train.py:421] (2/8) Epoch 4, batch 37400, loss[loss=2.394, over 1190.00 frames. , ppl: 10.953977233611193] tot_loss[loss=2.31, over 5435256.30 frames. , ppl: 10.075572589884894], batch size: 70
+2022-12-11 07:59:17,910 INFO [train.py:421] (2/8) Epoch 4, batch 37600, loss[loss=2.981, over 560.00 frames. , ppl: 19.704055899334612] tot_loss[loss=2.311, over 5407138.77 frames. , ppl: 10.080229708871162], batch size: 70
+2022-12-11 08:00:55,523 INFO [train.py:421] (2/8) Epoch 4, batch 37800, loss[loss=6.38, over 210.00 frames. , ppl: 589.9641952010261] tot_loss[loss=2.31, over 5418048.35 frames. , ppl: 10.077737927100438], batch size: 70
+2022-12-11 08:02:37,238 INFO [train.py:421] (2/8) Epoch 4, batch 38000, loss[loss=2.476, over 1050.00 frames. , ppl: 11.893266233577176] tot_loss[loss=2.308, over 5464607.22 frames. , ppl: 10.05834803849088], batch size: 70
+2022-12-11 08:02:37,239 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:02:37,986 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.297, over 211138.00 frames. , ppl: 9.945532988975891
+2022-12-11 08:04:17,113 INFO [train.py:421] (2/8) Epoch 4, batch 38200, loss[loss=2.744, over 700.00 frames. , ppl: 15.543383963771936] tot_loss[loss=2.309, over 5459268.99 frames. , ppl: 10.062786497124842], batch size: 70
+2022-12-11 08:05:57,529 INFO [train.py:421] (2/8) Epoch 4, batch 38400, loss[loss=2.566, over 770.00 frames. , ppl: 13.015366786343604] tot_loss[loss=2.308, over 5500259.34 frames. , ppl: 10.055039055740304], batch size: 70
+2022-12-11 08:07:37,177 INFO [train.py:421] (2/8) Epoch 4, batch 38600, loss[loss=2.33, over 2450.00 frames. , ppl: 10.276601981122287] tot_loss[loss=2.308, over 5517407.65 frames. , ppl: 10.055862075890246], batch size: 70
+2022-12-11 08:09:17,864 INFO [train.py:421] (2/8) Epoch 4, batch 38800, loss[loss=2.289, over 4480.00 frames. , ppl: 9.86112185666521] tot_loss[loss=2.307, over 5564089.01 frames. , ppl: 10.046779116360042], batch size: 70
+2022-12-11 08:10:56,256 INFO [train.py:421] (2/8) Epoch 4, batch 39000, loss[loss=2.282, over 2870.00 frames. , ppl: 9.79628947010168] tot_loss[loss=2.309, over 5525816.37 frames. , ppl: 10.059809816507853], batch size: 70
+2022-12-11 08:10:56,256 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:10:57,005 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.298, over 211138.00 frames. , ppl: 9.949938227560008
+2022-12-11 08:12:35,868 INFO [train.py:421] (2/8) Epoch 4, batch 39200, loss[loss=2.988, over 560.00 frames. , ppl: 19.841937718871538] tot_loss[loss=2.309, over 5496457.33 frames. , ppl: 10.066782962598518], batch size: 70
+2022-12-11 08:14:16,246 INFO [train.py:421] (2/8) Epoch 4, batch 39400, loss[loss=2.462, over 1260.00 frames. , ppl: 11.732445998197802] tot_loss[loss=2.308, over 5541602.42 frames. , ppl: 10.052931851649939], batch size: 70
+2022-12-11 08:15:56,340 INFO [train.py:421] (2/8) Epoch 4, batch 39600, loss[loss=2.361, over 2100.00 frames. , ppl: 10.605983688608934] tot_loss[loss=2.308, over 5505920.93 frames. , ppl: 10.056146547811402], batch size: 70
+2022-12-11 08:17:35,780 INFO [train.py:421] (2/8) Epoch 4, batch 39800, loss[loss=2.246, over 2940.00 frames. , ppl: 9.445151125254606] tot_loss[loss=2.307, over 5530115.14 frames. , ppl: 10.046697201929964], batch size: 70
+2022-12-11 08:19:17,000 INFO [train.py:421] (2/8) Epoch 4, batch 40000, loss[loss=2.304, over 2660.00 frames. , ppl: 10.016137744112553] tot_loss[loss=2.308, over 5515058.28 frames. , ppl: 10.052474530290553], batch size: 70
+2022-12-11 08:19:17,000 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:19:17,760 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.93333160275137
+2022-12-11 08:20:58,993 INFO [train.py:421] (2/8) Epoch 4, batch 40200, loss[loss=2.389, over 3640.00 frames. , ppl: 10.906517716702442] tot_loss[loss=2.308, over 5509314.07 frames. , ppl: 10.052711466525732], batch size: 70
+2022-12-11 08:22:39,534 INFO [train.py:421] (2/8) Epoch 4, batch 40400, loss[loss=2.471, over 1260.00 frames. , ppl: 11.832259436473095] tot_loss[loss=2.307, over 5538904.93 frames. , ppl: 10.041772316450952], batch size: 70
+2022-12-11 08:24:20,706 INFO [train.py:421] (2/8) Epoch 4, batch 40600, loss[loss=2.632, over 770.00 frames. , ppl: 13.897872136651795] tot_loss[loss=2.307, over 5540555.64 frames. , ppl: 10.040917324990035], batch size: 70
+2022-12-11 08:26:03,834 INFO [train.py:421] (2/8) Epoch 4, batch 40800, loss[loss=2.366, over 1610.00 frames. , ppl: 10.65597925000948] tot_loss[loss=2.305, over 5577108.59 frames. , ppl: 10.026448222803483], batch size: 70
+2022-12-11 08:27:44,150 INFO [train.py:421] (2/8) Epoch 4, batch 41000, loss[loss=2.371, over 2870.00 frames. , ppl: 10.712502937589662] tot_loss[loss=2.306, over 5546986.26 frames. , ppl: 10.037245382390562], batch size: 70
+2022-12-11 08:27:44,150 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:27:44,909 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.931032464771791
+2022-12-11 08:29:23,713 INFO [train.py:421] (2/8) Epoch 4, batch 41200, loss[loss=2.439, over 2590.00 frames. , ppl: 11.456569080029395] tot_loss[loss=2.307, over 5516156.88 frames. , ppl: 10.046256484637619], batch size: 70
+2022-12-11 08:31:02,540 INFO [train.py:421] (2/8) Epoch 4, batch 41400, loss[loss=2.638, over 1190.00 frames. , ppl: 13.982843705701587] tot_loss[loss=2.308, over 5485101.31 frames. , ppl: 10.057119933284351], batch size: 70
+2022-12-11 08:32:46,787 INFO [train.py:421] (2/8) Epoch 4, batch 41600, loss[loss=5.975, over 210.00 frames. , ppl: 393.4905160907653] tot_loss[loss=2.308, over 5492954.67 frames. , ppl: 10.055826788011041], batch size: 70
+2022-12-11 08:34:26,738 INFO [train.py:421] (2/8) Epoch 4, batch 41800, loss[loss=2.609, over 840.00 frames. , ppl: 13.591932723126233] tot_loss[loss=2.308, over 5510027.02 frames. , ppl: 10.053944828645578], batch size: 70
+2022-12-11 08:36:06,525 INFO [train.py:421] (2/8) Epoch 4, batch 42000, loss[loss=2.248, over 4060.00 frames. , ppl: 9.465183485000182] tot_loss[loss=2.309, over 5499662.71 frames. , ppl: 10.060889112607068], batch size: 70
+2022-12-11 08:36:06,525 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:36:07,272 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.298, over 211138.00 frames. , ppl: 9.952867789688982
+2022-12-11 08:37:48,692 INFO [train.py:421] (2/8) Epoch 4, batch 42200, loss[loss=2.279, over 3710.00 frames. , ppl: 9.762086408304453] tot_loss[loss=2.308, over 5510170.49 frames. , ppl: 10.057262189387101], batch size: 70
+2022-12-11 08:39:25,695 INFO [train.py:421] (2/8) Epoch 4, batch 42400, loss[loss=2.264, over 4970.00 frames. , ppl: 9.62478287392058] tot_loss[loss=2.308, over 5546622.09 frames. , ppl: 10.052504340852327], batch size: 70
+2022-12-11 08:41:06,782 INFO [train.py:421] (2/8) Epoch 4, batch 42600, loss[loss=2.274, over 8050.00 frames. , ppl: 9.717230691951631] tot_loss[loss=2.306, over 5600054.66 frames. , ppl: 10.038733550950608], batch size: 70
+2022-12-11 08:42:47,989 INFO [train.py:421] (2/8) Epoch 4, batch 42800, loss[loss=2.366, over 3080.00 frames. , ppl: 10.65636317394154] tot_loss[loss=2.306, over 5630522.99 frames. , ppl: 10.029899091243585], batch size: 70
+2022-12-11 08:44:22,384 INFO [train.py:421] (2/8) Epoch 4, batch 43000, loss[loss=2.357, over 1890.00 frames. , ppl: 10.556490296497916] tot_loss[loss=2.306, over 5607960.27 frames. , ppl: 10.036039134725543], batch size: 70
+2022-12-11 08:44:22,384 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:44:23,129 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.932262819602796
+2022-12-11 08:46:01,985 INFO [train.py:421] (2/8) Epoch 4, batch 43200, loss[loss=2.28, over 2870.00 frames. , ppl: 9.773366278578775] tot_loss[loss=2.307, over 5567334.76 frames. , ppl: 10.04243313489477], batch size: 70
+2022-12-11 08:47:41,482 INFO [train.py:421] (2/8) Epoch 4, batch 43400, loss[loss=2.419, over 1540.00 frames. , ppl: 11.230489637074854] tot_loss[loss=2.308, over 5533413.20 frames. , ppl: 10.052666789433616], batch size: 70
+2022-12-11 08:49:23,891 INFO [train.py:421] (2/8) Epoch 4, batch 43600, loss[loss=2.39, over 1470.00 frames. , ppl: 10.909036378723465] tot_loss[loss=2.307, over 5560356.93 frames. , ppl: 10.043521545740418], batch size: 70
+2022-12-11 08:51:04,780 INFO [train.py:421] (2/8) Epoch 4, batch 43800, loss[loss=2.187, over 7840.00 frames. , ppl: 8.907320399344872] tot_loss[loss=2.306, over 5608999.68 frames. , ppl: 10.033809613914217], batch size: 70
+2022-12-11 08:52:41,191 INFO [train.py:421] (2/8) Epoch 4, batch 44000, loss[loss=2.27, over 3990.00 frames. , ppl: 9.683525748240505] tot_loss[loss=2.305, over 5622180.46 frames. , ppl: 10.028115249594086], batch size: 70
+2022-12-11 08:52:41,191 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 08:52:41,937 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.938447770769514
+2022-12-11 08:54:20,560 INFO [train.py:421] (2/8) Epoch 4, batch 44200, loss[loss=2.494, over 840.00 frames. , ppl: 12.108899307183364] tot_loss[loss=2.307, over 5564048.38 frames. , ppl: 10.041535818451443], batch size: 70
+2022-12-11 08:55:59,750 INFO [train.py:421] (2/8) Epoch 4, batch 44400, loss[loss=2.511, over 1120.00 frames. , ppl: 12.311129792599411] tot_loss[loss=2.306, over 5565571.12 frames. , ppl: 10.037934927537675], batch size: 70
+2022-12-11 08:57:40,777 INFO [train.py:421] (2/8) Epoch 4, batch 44600, loss[loss=2.413, over 1190.00 frames. , ppl: 11.16849665345704] tot_loss[loss=2.306, over 5576023.69 frames. , ppl: 10.033706885917004], batch size: 70
+2022-12-11 08:59:24,320 INFO [train.py:421] (2/8) Epoch 4, batch 44800, loss[loss=2.447, over 2170.00 frames. , ppl: 11.553109888003265] tot_loss[loss=2.307, over 5541015.20 frames. , ppl: 10.03935081378956], batch size: 70
+2022-12-11 09:01:09,940 INFO [train.py:421] (2/8) Epoch 4, batch 45000, loss[loss=2.192, over 5390.00 frames. , ppl: 8.956584612110236] tot_loss[loss=2.306, over 5544718.57 frames. , ppl: 10.034519117472563], batch size: 70
+2022-12-11 09:01:09,940 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:01:10,691 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.909560376874897
+2022-12-11 09:02:55,711 INFO [train.py:421] (2/8) Epoch 4, batch 45200, loss[loss=2.318, over 4130.00 frames. , ppl: 10.160069749178442] tot_loss[loss=2.307, over 5525334.98 frames. , ppl: 10.041727414831545], batch size: 70
+2022-12-11 09:04:33,450 INFO [train.py:421] (2/8) Epoch 4, batch 45400, loss[loss=2.35, over 2870.00 frames. , ppl: 10.481503130862624] tot_loss[loss=2.306, over 5544011.33 frames. , ppl: 10.03773440380295], batch size: 70
+2022-12-11 09:06:16,158 INFO [train.py:421] (2/8) Epoch 4, batch 45600, loss[loss=2.208, over 4690.00 frames. , ppl: 9.09960897873921] tot_loss[loss=2.307, over 5529685.86 frames. , ppl: 10.042847524978791], batch size: 70
+2022-12-11 09:07:57,246 INFO [train.py:421] (2/8) Epoch 4, batch 45800, loss[loss=2.223, over 4970.00 frames. , ppl: 9.238354780177534] tot_loss[loss=2.307, over 5518747.30 frames. , ppl: 10.043025127912664], batch size: 70
+2022-12-11 09:09:38,445 INFO [train.py:421] (2/8) Epoch 4, batch 46000, loss[loss=2.298, over 3150.00 frames. , ppl: 9.959071535920906] tot_loss[loss=2.307, over 5532874.99 frames. , ppl: 10.039363489098177], batch size: 70
+2022-12-11 09:09:38,445 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:09:39,192 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.295, over 211138.00 frames. , ppl: 9.921327269150712
+2022-12-11 09:11:18,479 INFO [train.py:421] (2/8) Epoch 4, batch 46200, loss[loss=2.49, over 1260.00 frames. , ppl: 12.06722203425377] tot_loss[loss=2.309, over 5488394.91 frames. , ppl: 10.059453191200031], batch size: 70
+2022-12-11 09:12:54,785 INFO [train.py:421] (2/8) Epoch 4, batch 46400, loss[loss=2.46, over 1540.00 frames. , ppl: 11.701256013246349] tot_loss[loss=2.308, over 5518077.10 frames. , ppl: 10.051751375985356], batch size: 70
+2022-12-11 09:14:38,860 INFO [train.py:421] (2/8) Epoch 4, batch 46600, loss[loss=2.296, over 2170.00 frames. , ppl: 9.933097230602668] tot_loss[loss=2.308, over 5488077.71 frames. , ppl: 10.056881531519855], batch size: 70
+2022-12-11 09:16:20,806 INFO [train.py:421] (2/8) Epoch 4, batch 46800, loss[loss=2.326, over 2030.00 frames. , ppl: 10.233117079183282] tot_loss[loss=2.308, over 5510747.60 frames. , ppl: 10.049658424741287], batch size: 70
+2022-12-11 09:18:01,883 INFO [train.py:421] (2/8) Epoch 4, batch 47000, loss[loss=2.33, over 2310.00 frames. , ppl: 10.273844104244525] tot_loss[loss=2.307, over 5504866.34 frames. , ppl: 10.048202854000943], batch size: 70
+2022-12-11 09:18:01,883 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:18:02,643 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.912724529894838
+2022-12-11 09:19:43,173 INFO [train.py:421] (2/8) Epoch 4, batch 47200, loss[loss=2.374, over 2030.00 frames. , ppl: 10.74252421532079] tot_loss[loss=2.307, over 5496532.99 frames. , ppl: 10.047951527561667], batch size: 70
+2022-12-11 09:21:18,453 INFO [train.py:421] (2/8) Epoch 4, batch 47400, loss[loss=2.449, over 1890.00 frames. , ppl: 11.580206313606856] tot_loss[loss=2.307, over 5503463.49 frames. , ppl: 10.040006783553187], batch size: 70
+2022-12-11 09:22:56,915 INFO [train.py:421] (2/8) Epoch 4, batch 47600, loss[loss=2.231, over 3990.00 frames. , ppl: 9.308959706766473] tot_loss[loss=2.307, over 5499497.45 frames. , ppl: 10.047915683745922], batch size: 70
+2022-12-11 09:24:37,051 INFO [train.py:421] (2/8) Epoch 4, batch 47800, loss[loss=2.27, over 1890.00 frames. , ppl: 9.68412723730666] tot_loss[loss=2.308, over 5450773.08 frames. , ppl: 10.059035526679459], batch size: 70
+2022-12-11 09:26:19,548 INFO [train.py:421] (2/8) Epoch 4, batch 48000, loss[loss=2.17, over 11900.00 frames. , ppl: 8.757895929479083] tot_loss[loss=2.308, over 5471487.13 frames. , ppl: 10.050213118308552], batch size: 70
+2022-12-11 09:26:19,548 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:26:20,321 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.295, over 211138.00 frames. , ppl: 9.92101743515498
+2022-12-11 09:28:03,643 INFO [train.py:421] (2/8) Epoch 4, batch 48200, loss[loss=2.371, over 1750.00 frames. , ppl: 10.706592294214936] tot_loss[loss=2.308, over 5438848.26 frames. , ppl: 10.057147419217019], batch size: 70
+2022-12-11 09:29:44,757 INFO [train.py:421] (2/8) Epoch 4, batch 48400, loss[loss=2.388, over 3010.00 frames. , ppl: 10.888213511406521] tot_loss[loss=2.307, over 5481475.94 frames. , ppl: 10.048365660064762], batch size: 70
+2022-12-11 09:31:30,698 INFO [train.py:421] (2/8) Epoch 4, batch 48600, loss[loss=2.428, over 2100.00 frames. , ppl: 11.34110813064026] tot_loss[loss=2.308, over 5464165.61 frames. , ppl: 10.051223316152761], batch size: 70
+2022-12-11 09:33:10,103 INFO [train.py:421] (2/8) Epoch 4, batch 48800, loss[loss=2.283, over 2520.00 frames. , ppl: 9.808633797626673] tot_loss[loss=2.308, over 5467801.40 frames. , ppl: 10.049928473983845], batch size: 70
+2022-12-11 09:34:51,408 INFO [train.py:421] (2/8) Epoch 4, batch 49000, loss[loss=2.272, over 1750.00 frames. , ppl: 9.703599807158525] tot_loss[loss=2.306, over 5496411.88 frames. , ppl: 10.03608300059489], batch size: 70
+2022-12-11 09:34:51,409 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:34:52,168 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.912736267159561
+2022-12-11 09:36:32,382 INFO [train.py:421] (2/8) Epoch 4, batch 49200, loss[loss=2.543, over 980.00 frames. , ppl: 12.723435037691859] tot_loss[loss=2.307, over 5473415.55 frames. , ppl: 10.048225782337253], batch size: 70
+2022-12-11 09:38:12,918 INFO [train.py:421] (2/8) Epoch 4, batch 49400, loss[loss=2.491, over 1470.00 frames. , ppl: 12.067742456183444] tot_loss[loss=2.308, over 5494743.41 frames. , ppl: 10.050580635655896], batch size: 70
+2022-12-11 09:39:55,520 INFO [train.py:421] (2/8) Epoch 4, batch 49600, loss[loss=2.37, over 1120.00 frames. , ppl: 10.697932353384317] tot_loss[loss=2.309, over 5454988.78 frames. , ppl: 10.061928501038333], batch size: 70
+2022-12-11 09:41:34,985 INFO [train.py:421] (2/8) Epoch 4, batch 49800, loss[loss=2.306, over 4130.00 frames. , ppl: 10.033783413986166] tot_loss[loss=2.309, over 5444721.27 frames. , ppl: 10.063433218414513], batch size: 70
+2022-12-11 09:43:18,302 INFO [train.py:421] (2/8) Epoch 4, batch 50000, loss[loss=2.44, over 1680.00 frames. , ppl: 11.471151533328825] tot_loss[loss=2.308, over 5466870.93 frames. , ppl: 10.054543411236779], batch size: 70
+2022-12-11 09:43:18,303 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:43:19,051 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.911439383482614
+2022-12-11 09:44:57,840 INFO [train.py:421] (2/8) Epoch 4, batch 50200, loss[loss=2.427, over 1470.00 frames. , ppl: 11.320673796285622] tot_loss[loss=2.309, over 5459665.49 frames. , ppl: 10.063764450896675], batch size: 70
+2022-12-11 09:46:40,421 INFO [train.py:421] (2/8) Epoch 4, batch 50400, loss[loss=2.266, over 3500.00 frames. , ppl: 9.644028033881288] tot_loss[loss=2.308, over 5484204.00 frames. , ppl: 10.05450058816935], batch size: 70
+2022-12-11 09:48:16,348 INFO [train.py:421] (2/8) Epoch 4, batch 50600, loss[loss=2.371, over 3150.00 frames. , ppl: 10.708255702621067] tot_loss[loss=2.309, over 5436035.79 frames. , ppl: 10.064786604606304], batch size: 70
+2022-12-11 09:49:58,423 INFO [train.py:421] (2/8) Epoch 4, batch 50800, loss[loss=2.364, over 2590.00 frames. , ppl: 10.630420617684196] tot_loss[loss=2.308, over 5456280.26 frames. , ppl: 10.054627679723405], batch size: 70
+2022-12-11 09:51:36,249 INFO [train.py:421] (2/8) Epoch 4, batch 51000, loss[loss=2.235, over 3080.00 frames. , ppl: 9.349482233354262] tot_loss[loss=2.307, over 5495698.79 frames. , ppl: 10.0458231740511], batch size: 70
+2022-12-11 09:51:36,250 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:51:36,999 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.909437175772835
+2022-12-11 09:53:14,567 INFO [train.py:421] (2/8) Epoch 4, batch 51200, loss[loss=2.306, over 1540.00 frames. , ppl: 10.031654549075629] tot_loss[loss=2.308, over 5453397.43 frames. , ppl: 10.051285936318605], batch size: 70
+2022-12-11 09:54:54,304 INFO [train.py:421] (2/8) Epoch 4, batch 51400, loss[loss=2.234, over 4690.00 frames. , ppl: 9.341645763843177] tot_loss[loss=2.308, over 5465120.95 frames. , ppl: 10.051186521433117], batch size: 70
+2022-12-11 09:56:32,679 INFO [train.py:421] (2/8) Epoch 4, batch 51600, loss[loss=2.263, over 3920.00 frames. , ppl: 9.613610520060046] tot_loss[loss=2.307, over 5473248.94 frames. , ppl: 10.046017653723469], batch size: 70
+2022-12-11 09:58:13,022 INFO [train.py:421] (2/8) Epoch 4, batch 51800, loss[loss=2.226, over 2870.00 frames. , ppl: 9.259041827240933] tot_loss[loss=2.307, over 5485000.07 frames. , ppl: 10.048899582492927], batch size: 70
+2022-12-11 09:59:53,302 INFO [train.py:421] (2/8) Epoch 4, batch 52000, loss[loss=2.315, over 3640.00 frames. , ppl: 10.123074863061472] tot_loss[loss=2.307, over 5509149.70 frames. , ppl: 10.041873328433075], batch size: 70
+2022-12-11 09:59:53,303 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 09:59:54,056 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.914090547381605
+2022-12-11 10:01:33,675 INFO [train.py:421] (2/8) Epoch 4, batch 52200, loss[loss=2.355, over 1540.00 frames. , ppl: 10.542541611049561] tot_loss[loss=2.307, over 5478018.02 frames. , ppl: 10.048974040878084], batch size: 70
+2022-12-11 10:03:14,831 INFO [train.py:421] (2/8) Epoch 4, batch 52400, loss[loss=2.292, over 2730.00 frames. , ppl: 9.898007238137886] tot_loss[loss=2.307, over 5487961.08 frames. , ppl: 10.043369765606641], batch size: 70
+2022-12-11 10:04:57,997 INFO [train.py:421] (2/8) Epoch 4, batch 52600, loss[loss=2.332, over 1610.00 frames. , ppl: 10.30124202106155] tot_loss[loss=2.307, over 5486841.06 frames. , ppl: 10.045292766828377], batch size: 70
+2022-12-11 10:06:38,641 INFO [train.py:421] (2/8) Epoch 4, batch 52800, loss[loss=2.341, over 3080.00 frames. , ppl: 10.38940343739473] tot_loss[loss=2.307, over 5483384.72 frames. , ppl: 10.04334570655798], batch size: 70
+2022-12-11 10:08:18,149 INFO [train.py:421] (2/8) Epoch 4, batch 53000, loss[loss=2.14, over 3500.00 frames. , ppl: 8.50269905992133] tot_loss[loss=2.306, over 5527152.72 frames. , ppl: 10.033026339832283], batch size: 70
+2022-12-11 10:08:18,150 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:08:18,896 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.929699384943623
+2022-12-11 10:09:58,059 INFO [train.py:421] (2/8) Epoch 4, batch 53200, loss[loss=2.182, over 9800.00 frames. , ppl: 8.862154437437006] tot_loss[loss=2.307, over 5510341.54 frames. , ppl: 10.04155945991138], batch size: 70
+2022-12-11 10:11:38,307 INFO [train.py:421] (2/8) Epoch 4, batch 53400, loss[loss=2.154, over 4060.00 frames. , ppl: 8.622578659268655] tot_loss[loss=2.307, over 5474281.91 frames. , ppl: 10.04831854595674], batch size: 70
+2022-12-11 10:13:16,451 INFO [train.py:421] (2/8) Epoch 4, batch 53600, loss[loss=2.148, over 5460.00 frames. , ppl: 8.563456139868231] tot_loss[loss=2.307, over 5482760.20 frames. , ppl: 10.047650837570332], batch size: 70
+2022-12-11 10:14:54,533 INFO [train.py:421] (2/8) Epoch 4, batch 53800, loss[loss=2.461, over 980.00 frames. , ppl: 11.716344507261343] tot_loss[loss=2.306, over 5512895.58 frames. , ppl: 10.036371286071438], batch size: 70
+2022-12-11 10:16:36,543 INFO [train.py:421] (2/8) Epoch 4, batch 54000, loss[loss=2.352, over 1400.00 frames. , ppl: 10.505699285432101] tot_loss[loss=2.306, over 5528361.74 frames. , ppl: 10.030153675149363], batch size: 70
+2022-12-11 10:16:36,543 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:16:37,305 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.903191128032574
+2022-12-11 10:18:18,565 INFO [train.py:421] (2/8) Epoch 4, batch 54200, loss[loss=2.366, over 1750.00 frames. , ppl: 10.658514922880697] tot_loss[loss=2.306, over 5539220.29 frames. , ppl: 10.029281664698683], batch size: 70
+2022-12-11 10:19:56,516 INFO [train.py:421] (2/8) Epoch 4, batch 54400, loss[loss=3.18, over 490.00 frames. , ppl: 24.042779715432314] tot_loss[loss=2.305, over 5536901.78 frames. , ppl: 10.026816255824688], batch size: 70
+2022-12-11 10:21:35,656 INFO [train.py:421] (2/8) Epoch 4, batch 54600, loss[loss=2.36, over 1470.00 frames. , ppl: 10.588311220304242] tot_loss[loss=2.305, over 5543601.78 frames. , ppl: 10.023484528444984], batch size: 70
+2022-12-11 10:23:17,361 INFO [train.py:421] (2/8) Epoch 4, batch 54800, loss[loss=3.072, over 560.00 frames. , ppl: 21.592486493813922] tot_loss[loss=2.305, over 5563477.82 frames. , ppl: 10.020017021976358], batch size: 70
+2022-12-11 10:24:58,184 INFO [train.py:421] (2/8) Epoch 4, batch 55000, loss[loss=2.332, over 3780.00 frames. , ppl: 10.296333741299424] tot_loss[loss=2.306, over 5543969.31 frames. , ppl: 10.029707882057027], batch size: 70
+2022-12-11 10:24:58,185 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:24:58,943 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.91692148777633
+2022-12-11 10:26:38,690 INFO [train.py:421] (2/8) Epoch 4, batch 55200, loss[loss=2.45, over 1540.00 frames. , ppl: 11.59368849827807] tot_loss[loss=2.305, over 5587303.09 frames. , ppl: 10.020847734143766], batch size: 70
+2022-12-11 10:28:18,394 INFO [train.py:421] (2/8) Epoch 4, batch 55400, loss[loss=2.293, over 3570.00 frames. , ppl: 9.906946146905026] tot_loss[loss=2.305, over 5579418.03 frames. , ppl: 10.021733739687967], batch size: 70
+2022-12-11 10:29:59,489 INFO [train.py:421] (2/8) Epoch 4, batch 55600, loss[loss=2.657, over 840.00 frames. , ppl: 14.251465881130677] tot_loss[loss=2.304, over 5605035.82 frames. , ppl: 10.0158601199075], batch size: 70
+2022-12-11 10:31:38,242 INFO [train.py:421] (2/8) Epoch 4, batch 55800, loss[loss=2.229, over 7700.00 frames. , ppl: 9.286184892851082] tot_loss[loss=2.304, over 5604149.52 frames. , ppl: 10.016623637390385], batch size: 70
+2022-12-11 10:33:18,377 INFO [train.py:421] (2/8) Epoch 4, batch 56000, loss[loss=2.752, over 910.00 frames. , ppl: 15.680969527530127] tot_loss[loss=2.306, over 5558787.49 frames. , ppl: 10.031312778888086], batch size: 70
+2022-12-11 10:33:18,378 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:33:19,123 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.91445299177917
+2022-12-11 10:34:53,562 INFO [train.py:421] (2/8) Epoch 4, batch 56200, loss[loss=2.299, over 3850.00 frames. , ppl: 9.959286962228438] tot_loss[loss=2.307, over 5506165.90 frames. , ppl: 10.04630695200559], batch size: 70
+2022-12-11 10:36:32,168 INFO [train.py:421] (2/8) Epoch 4, batch 56400, loss[loss=2.42, over 1330.00 frames. , ppl: 11.244906456282061] tot_loss[loss=2.308, over 5514615.44 frames. , ppl: 10.050638089002247], batch size: 70
+2022-12-11 10:38:13,264 INFO [train.py:421] (2/8) Epoch 4, batch 56600, loss[loss=2.057, over 3290.00 frames. , ppl: 7.819788781608339] tot_loss[loss=2.308, over 5499608.40 frames. , ppl: 10.054363717600982], batch size: 70
+2022-12-11 10:39:51,813 INFO [train.py:421] (2/8) Epoch 4, batch 56800, loss[loss=2.348, over 3920.00 frames. , ppl: 10.465480297804849] tot_loss[loss=2.308, over 5490680.56 frames. , ppl: 10.05259199895096], batch size: 70
+2022-12-11 10:41:31,980 INFO [train.py:421] (2/8) Epoch 4, batch 57000, loss[loss=2.252, over 2030.00 frames. , ppl: 9.50833138059489] tot_loss[loss=2.31, over 5431944.52 frames. , ppl: 10.07041915562931], batch size: 70
+2022-12-11 10:41:31,981 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:41:32,726 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.90575504999001
+2022-12-11 10:43:12,819 INFO [train.py:421] (2/8) Epoch 4, batch 57200, loss[loss=2.315, over 2380.00 frames. , ppl: 10.128779824980889] tot_loss[loss=2.309, over 5454344.23 frames. , ppl: 10.064844969448986], batch size: 70
+2022-12-11 10:44:55,967 INFO [train.py:421] (2/8) Epoch 4, batch 57400, loss[loss=2.299, over 1610.00 frames. , ppl: 9.9606866022774] tot_loss[loss=2.308, over 5470064.87 frames. , ppl: 10.057759919556057], batch size: 70
+2022-12-11 10:46:33,832 INFO [train.py:421] (2/8) Epoch 4, batch 57600, loss[loss=2.296, over 3640.00 frames. , ppl: 9.932999402917792] tot_loss[loss=2.309, over 5457224.74 frames. , ppl: 10.059523534814062], batch size: 70
+2022-12-11 10:48:10,496 INFO [train.py:421] (2/8) Epoch 4, batch 57800, loss[loss=2.3, over 2940.00 frames. , ppl: 9.976561519524555] tot_loss[loss=2.309, over 5423864.00 frames. , ppl: 10.065876160044331], batch size: 70
+2022-12-11 10:49:55,349 INFO [train.py:421] (2/8) Epoch 4, batch 58000, loss[loss=4.288, over 350.00 frames. , ppl: 72.85150565703849] tot_loss[loss=2.309, over 5417842.39 frames. , ppl: 10.069320418940526], batch size: 70
+2022-12-11 10:49:55,349 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:49:56,110 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.915101609561932
+2022-12-11 10:51:34,570 INFO [train.py:421] (2/8) Epoch 4, batch 58200, loss[loss=2.484, over 1190.00 frames. , ppl: 11.987007132586129] tot_loss[loss=2.308, over 5448259.83 frames. , ppl: 10.056661804448606], batch size: 70
+2022-12-11 10:53:13,755 INFO [train.py:421] (2/8) Epoch 4, batch 58400, loss[loss=2.178, over 5810.00 frames. , ppl: 8.828161938576399] tot_loss[loss=2.308, over 5448938.43 frames. , ppl: 10.05761051794182], batch size: 70
+2022-12-11 10:54:51,409 INFO [train.py:421] (2/8) Epoch 4, batch 58600, loss[loss=2.307, over 2240.00 frames. , ppl: 10.03997451829192] tot_loss[loss=2.308, over 5459377.38 frames. , ppl: 10.055463922268071], batch size: 70
+2022-12-11 10:56:31,627 INFO [train.py:421] (2/8) Epoch 4, batch 58800, loss[loss=2.291, over 1890.00 frames. , ppl: 9.885656446509172] tot_loss[loss=2.308, over 5454020.51 frames. , ppl: 10.049517918981412], batch size: 70
+2022-12-11 10:58:12,932 INFO [train.py:421] (2/8) Epoch 4, batch 59000, loss[loss=2.415, over 1400.00 frames. , ppl: 11.18492037496605] tot_loss[loss=2.307, over 5467466.65 frames. , ppl: 10.040109548130525], batch size: 70
+2022-12-11 10:58:12,932 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 10:58:13,677 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.911222274742915
+2022-12-11 10:59:52,975 INFO [train.py:421] (2/8) Epoch 4, batch 59200, loss[loss=2.629, over 770.00 frames. , ppl: 13.859337594010368] tot_loss[loss=2.306, over 5498407.27 frames. , ppl: 10.033184692620734], batch size: 70
+2022-12-11 11:01:31,036 INFO [train.py:421] (2/8) Epoch 4, batch 59400, loss[loss=2.218, over 4620.00 frames. , ppl: 9.188503729948149] tot_loss[loss=2.305, over 5537350.58 frames. , ppl: 10.020000006365708], batch size: 70
+2022-12-11 11:03:11,599 INFO [train.py:421] (2/8) Epoch 4, batch 59600, loss[loss=2.252, over 5670.00 frames. , ppl: 9.50477848355732] tot_loss[loss=2.305, over 5515093.34 frames. , ppl: 10.028643379885962], batch size: 70
+2022-12-11 11:04:50,759 INFO [train.py:421] (2/8) Epoch 4, batch 59800, loss[loss=2.369, over 3360.00 frames. , ppl: 10.68659879657486] tot_loss[loss=2.305, over 5543831.13 frames. , ppl: 10.02141015708587], batch size: 70
+2022-12-11 11:06:31,797 INFO [train.py:421] (2/8) Epoch 4, batch 60000, loss[loss=2.421, over 1610.00 frames. , ppl: 11.253347836929406] tot_loss[loss=2.306, over 5536723.42 frames. , ppl: 10.032206605796215], batch size: 70
+2022-12-11 11:06:31,798 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:06:32,558 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.908699467886942
+2022-12-11 11:08:12,353 INFO [train.py:421] (2/8) Epoch 4, batch 60200, loss[loss=2.374, over 1330.00 frames. , ppl: 10.73930485195504] tot_loss[loss=2.306, over 5549612.58 frames. , ppl: 10.031781307084469], batch size: 70
+2022-12-11 11:09:54,641 INFO [train.py:421] (2/8) Epoch 4, batch 60400, loss[loss=2.496, over 1540.00 frames. , ppl: 12.129405807681254] tot_loss[loss=2.307, over 5512861.27 frames. , ppl: 10.040320525054433], batch size: 70
+2022-12-11 11:11:33,931 INFO [train.py:421] (2/8) Epoch 4, batch 60600, loss[loss=2.475, over 1050.00 frames. , ppl: 11.883392463582734] tot_loss[loss=2.307, over 5502062.73 frames. , ppl: 10.039413055767051], batch size: 70
+2022-12-11 11:13:09,650 INFO [train.py:421] (2/8) Epoch 4, batch 60800, loss[loss=2.225, over 2940.00 frames. , ppl: 9.250317460235589] tot_loss[loss=2.307, over 5487754.99 frames. , ppl: 10.0418115019311], batch size: 70
+2022-12-11 11:14:48,151 INFO [train.py:421] (2/8) Epoch 4, batch 61000, loss[loss=2.338, over 2170.00 frames. , ppl: 10.36204738757842] tot_loss[loss=2.307, over 5481714.86 frames. , ppl: 10.045332359120147], batch size: 70
+2022-12-11 11:14:48,152 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:14:48,898 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.292, over 211138.00 frames. , ppl: 9.890057963547763
+2022-12-11 11:16:30,418 INFO [train.py:421] (2/8) Epoch 4, batch 61200, loss[loss=2.258, over 3010.00 frames. , ppl: 9.56416114839344] tot_loss[loss=2.308, over 5470217.99 frames. , ppl: 10.050174940325801], batch size: 70
+2022-12-11 11:18:08,905 INFO [train.py:421] (2/8) Epoch 4, batch 61400, loss[loss=2.346, over 2450.00 frames. , ppl: 10.44584780220845] tot_loss[loss=2.308, over 5479768.13 frames. , ppl: 10.059322131846821], batch size: 70
+2022-12-11 11:19:47,063 INFO [train.py:421] (2/8) Epoch 4, batch 61600, loss[loss=2.43, over 2380.00 frames. , ppl: 11.362589870062815] tot_loss[loss=2.308, over 5492713.01 frames. , ppl: 10.054654337528598], batch size: 70
+2022-12-11 11:21:28,011 INFO [train.py:421] (2/8) Epoch 4, batch 61800, loss[loss=2.723, over 630.00 frames. , ppl: 15.221801617422164] tot_loss[loss=2.307, over 5495358.24 frames. , ppl: 10.046313741674604], batch size: 70
+2022-12-11 11:23:08,281 INFO [train.py:421] (2/8) Epoch 4, batch 62000, loss[loss=2.293, over 2730.00 frames. , ppl: 9.909294235845927] tot_loss[loss=2.307, over 5493835.67 frames. , ppl: 10.041937934243052], batch size: 70
+2022-12-11 11:23:08,282 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:23:09,027 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.916335861485702
+2022-12-11 11:24:49,035 INFO [train.py:421] (2/8) Epoch 4, batch 62200, loss[loss=2.314, over 3850.00 frames. , ppl: 10.116351285914403] tot_loss[loss=2.307, over 5488397.92 frames. , ppl: 10.048160424620452], batch size: 70
+2022-12-11 11:26:31,324 INFO [train.py:421] (2/8) Epoch 4, batch 62400, loss[loss=3.296, over 490.00 frames. , ppl: 26.99107667231945] tot_loss[loss=2.308, over 5506915.29 frames. , ppl: 10.049440786220673], batch size: 70
+2022-12-11 11:28:12,734 INFO [train.py:421] (2/8) Epoch 4, batch 62600, loss[loss=2.342, over 3430.00 frames. , ppl: 10.403275775991586] tot_loss[loss=2.307, over 5527242.47 frames. , ppl: 10.04592350142537], batch size: 70
+2022-12-11 11:29:56,768 INFO [train.py:421] (2/8) Epoch 4, batch 62800, loss[loss=2.469, over 1470.00 frames. , ppl: 11.809201854612867] tot_loss[loss=2.307, over 5532976.23 frames. , ppl: 10.041360384278006], batch size: 70
+2022-12-11 11:31:39,087 INFO [train.py:421] (2/8) Epoch 4, batch 63000, loss[loss=2.56, over 840.00 frames. , ppl: 12.93707161644944] tot_loss[loss=2.306, over 5528935.99 frames. , ppl: 10.038251062604338], batch size: 70
+2022-12-11 11:31:39,087 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:31:39,834 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.912125947821703
+2022-12-11 11:33:18,807 INFO [train.py:421] (2/8) Epoch 4, batch 63200, loss[loss=2.421, over 1330.00 frames. , ppl: 11.258720064576622] tot_loss[loss=2.308, over 5494046.48 frames. , ppl: 10.051197049478526], batch size: 70
+2022-12-11 11:34:57,557 INFO [train.py:421] (2/8) Epoch 4, batch 63400, loss[loss=2.22, over 2660.00 frames. , ppl: 9.202965260690581] tot_loss[loss=2.308, over 5446130.27 frames. , ppl: 10.05531942376601], batch size: 70
+2022-12-11 11:36:40,497 INFO [train.py:421] (2/8) Epoch 4, batch 63600, loss[loss=2.732, over 700.00 frames. , ppl: 15.364149336861983] tot_loss[loss=2.307, over 5466404.78 frames. , ppl: 10.048424824685625], batch size: 70
+2022-12-11 11:38:20,357 INFO [train.py:421] (2/8) Epoch 4, batch 63800, loss[loss=2.254, over 4690.00 frames. , ppl: 9.525024482378885] tot_loss[loss=2.307, over 5484811.54 frames. , ppl: 10.043263749885174], batch size: 70
+2022-12-11 11:39:59,621 INFO [train.py:421] (2/8) Epoch 4, batch 64000, loss[loss=2.249, over 2730.00 frames. , ppl: 9.476356107214391] tot_loss[loss=2.306, over 5497522.22 frames. , ppl: 10.034877107580655], batch size: 70
+2022-12-11 11:39:59,621 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:40:00,404 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.292, over 211138.00 frames. , ppl: 9.899215356849181
+2022-12-11 11:41:39,829 INFO [train.py:421] (2/8) Epoch 4, batch 64200, loss[loss=2.356, over 1820.00 frames. , ppl: 10.54880765257204] tot_loss[loss=2.307, over 5481568.13 frames. , ppl: 10.040830287638794], batch size: 70
+2022-12-11 11:43:17,850 INFO [train.py:421] (2/8) Epoch 4, batch 64400, loss[loss=2.323, over 3080.00 frames. , ppl: 10.206864752080541] tot_loss[loss=2.306, over 5486642.55 frames. , ppl: 10.03736326245192], batch size: 70
+2022-12-11 11:44:57,183 INFO [train.py:421] (2/8) Epoch 4, batch 64600, loss[loss=2.385, over 2030.00 frames. , ppl: 10.85463786340012] tot_loss[loss=2.305, over 5484942.39 frames. , ppl: 10.027597174948522], batch size: 70
+2022-12-11 11:46:37,645 INFO [train.py:421] (2/8) Epoch 4, batch 64800, loss[loss=2.469, over 1330.00 frames. , ppl: 11.80985219716362] tot_loss[loss=2.306, over 5477930.10 frames. , ppl: 10.03404400652406], batch size: 70
+2022-12-11 11:48:19,715 INFO [train.py:421] (2/8) Epoch 4, batch 65000, loss[loss=2.325, over 2520.00 frames. , ppl: 10.228362556800695] tot_loss[loss=2.306, over 5473377.98 frames. , ppl: 10.031568483071391], batch size: 70
+2022-12-11 11:48:19,715 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:48:20,473 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.916086357031197
+2022-12-11 11:50:01,971 INFO [train.py:421] (2/8) Epoch 4, batch 65200, loss[loss=4.209, over 350.00 frames. , ppl: 67.25935271151657] tot_loss[loss=2.305, over 5500003.74 frames. , ppl: 10.021764753754896], batch size: 70
+2022-12-11 11:51:41,797 INFO [train.py:421] (2/8) Epoch 4, batch 65400, loss[loss=2.321, over 2380.00 frames. , ppl: 10.18905904920925] tot_loss[loss=2.306, over 5457262.84 frames. , ppl: 10.035134212001395], batch size: 70
+2022-12-11 11:53:19,725 INFO [train.py:421] (2/8) Epoch 4, batch 65600, loss[loss=2.367, over 1540.00 frames. , ppl: 10.667682985772316] tot_loss[loss=2.306, over 5497603.14 frames. , ppl: 10.02947529217873], batch size: 70
+2022-12-11 11:55:00,754 INFO [train.py:421] (2/8) Epoch 4, batch 65800, loss[loss=2.881, over 630.00 frames. , ppl: 17.83281934146816] tot_loss[loss=2.304, over 5552700.64 frames. , ppl: 10.015471683400932], batch size: 70
+2022-12-11 11:56:40,731 INFO [train.py:421] (2/8) Epoch 4, batch 66000, loss[loss=2.596, over 1050.00 frames. , ppl: 13.408198529793237] tot_loss[loss=2.305, over 5494641.97 frames. , ppl: 10.024275559765378], batch size: 70
+2022-12-11 11:56:40,731 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 11:56:41,491 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.919058803417657
+2022-12-11 11:58:22,192 INFO [train.py:421] (2/8) Epoch 4, batch 66200, loss[loss=4.043, over 350.00 frames. , ppl: 56.97518330476399] tot_loss[loss=2.305, over 5503817.57 frames. , ppl: 10.022817829757301], batch size: 70
+2022-12-11 12:00:04,131 INFO [train.py:421] (2/8) Epoch 4, batch 66400, loss[loss=2.992, over 560.00 frames. , ppl: 19.924686490931826] tot_loss[loss=2.305, over 5497415.99 frames. , ppl: 10.02213803421136], batch size: 70
+2022-12-11 12:01:41,417 INFO [train.py:421] (2/8) Epoch 4, batch 66600, loss[loss=2.34, over 2450.00 frames. , ppl: 10.382271096554893] tot_loss[loss=2.305, over 5496675.65 frames. , ppl: 10.022083783180369], batch size: 70
+2022-12-11 12:03:22,240 INFO [train.py:421] (2/8) Epoch 4, batch 66800, loss[loss=2.349, over 980.00 frames. , ppl: 10.4797513635856] tot_loss[loss=2.305, over 5491069.86 frames. , ppl: 10.026676829311153], batch size: 70
+2022-12-11 12:05:01,127 INFO [train.py:421] (2/8) Epoch 4, batch 67000, loss[loss=2.521, over 840.00 frames. , ppl: 12.435574680231207] tot_loss[loss=2.305, over 5484475.97 frames. , ppl: 10.027054466285865], batch size: 70
+2022-12-11 12:05:01,128 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 12:05:01,878 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.901741611125681
+2022-12-11 12:06:40,972 INFO [train.py:421] (2/8) Epoch 4, batch 67200, loss[loss=2.29, over 3360.00 frames. , ppl: 9.879873742786062] tot_loss[loss=2.306, over 5471054.65 frames. , ppl: 10.029973925904018], batch size: 70
+2022-12-11 12:08:24,020 INFO [train.py:421] (2/8) Epoch 4, batch 67400, loss[loss=2.682, over 770.00 frames. , ppl: 14.617455786929975] tot_loss[loss=2.305, over 5513996.71 frames. , ppl: 10.024239605494808], batch size: 70
+2022-12-11 12:10:06,522 INFO [train.py:421] (2/8) Epoch 4, batch 67600, loss[loss=3.223, over 560.00 frames. , ppl: 25.10303842503206] tot_loss[loss=2.305, over 5504470.47 frames. , ppl: 10.027125561489711], batch size: 70
+2022-12-11 12:11:44,348 INFO [train.py:421] (2/8) Epoch 4, batch 67800, loss[loss=2.254, over 5530.00 frames. , ppl: 9.521736676418534] tot_loss[loss=2.305, over 5534514.39 frames. , ppl: 10.028202893546593], batch size: 70
+2022-12-11 12:13:27,419 INFO [train.py:421] (2/8) Epoch 4, batch 68000, loss[loss=2.215, over 3220.00 frames. , ppl: 9.162028176690672] tot_loss[loss=2.306, over 5496704.78 frames. , ppl: 10.03613412326291], batch size: 70
+2022-12-11 12:13:27,419 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 12:13:28,169 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.910991968643046
+2022-12-11 12:15:09,924 INFO [train.py:421] (2/8) Epoch 4, batch 68200, loss[loss=2.435, over 1400.00 frames. , ppl: 11.41776982566411] tot_loss[loss=2.305, over 5553877.13 frames. , ppl: 10.0253775589097], batch size: 70
+2022-12-11 12:16:49,680 INFO [train.py:421] (2/8) Epoch 4, batch 68400, loss[loss=2.239, over 5740.00 frames. , ppl: 9.386694427858211] tot_loss[loss=2.306, over 5516445.60 frames. , ppl: 10.031352834384151], batch size: 70
+2022-12-11 12:18:30,823 INFO [train.py:421] (2/8) Epoch 4, batch 68600, loss[loss=2.283, over 3290.00 frames. , ppl: 9.80939457058427] tot_loss[loss=2.304, over 5565084.84 frames. , ppl: 10.013576607853864], batch size: 70
+2022-12-11 12:20:12,549 INFO [train.py:421] (2/8) Epoch 4, batch 68800, loss[loss=2.33, over 3010.00 frames. , ppl: 10.272930231226628] tot_loss[loss=2.304, over 5562230.25 frames. , ppl: 10.011613837166054], batch size: 70
+2022-12-11 12:21:53,589 INFO [train.py:421] (2/8) Epoch 4, batch 69000, loss[loss=2.146, over 6020.00 frames. , ppl: 8.547643189783988] tot_loss[loss=2.305, over 5523879.05 frames. , ppl: 10.022369662850567], batch size: 70
+2022-12-11 12:21:53,589 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 12:21:54,350 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.910044396033411
+2022-12-11 12:23:33,504 INFO [train.py:421] (2/8) Epoch 4, batch 69200, loss[loss=2.26, over 7000.00 frames. , ppl: 9.586270242770006] tot_loss[loss=2.304, over 5521400.46 frames. , ppl: 10.012542130316032], batch size: 70
+2022-12-11 12:25:12,763 INFO [train.py:421] (2/8) Epoch 4, batch 69400, loss[loss=2.589, over 840.00 frames. , ppl: 13.310269149390534] tot_loss[loss=2.306, over 5460416.67 frames. , ppl: 10.031598188568317], batch size: 70
+2022-12-11 12:26:51,729 INFO [train.py:421] (2/8) Epoch 4, batch 69600, loss[loss=2.365, over 2380.00 frames. , ppl: 10.642921678327317] tot_loss[loss=2.305, over 5478629.84 frames. , ppl: 10.019712001846736], batch size: 70
+2022-12-11 12:28:30,365 INFO [train.py:421] (2/8) Epoch 4, batch 69800, loss[loss=2.345, over 2590.00 frames. , ppl: 10.428131629873068] tot_loss[loss=2.305, over 5485577.01 frames. , ppl: 10.024850623427545], batch size: 70
+2022-12-11 12:30:08,269 INFO [train.py:421] (2/8) Epoch 4, batch 70000, loss[loss=2.2, over 4270.00 frames. , ppl: 9.026491479611972] tot_loss[loss=2.306, over 5434209.82 frames. , ppl: 10.031122512479762], batch size: 70
+2022-12-11 12:30:08,270 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 12:30:09,018 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.292, over 211138.00 frames. , ppl: 9.896056980662939
+2022-12-11 12:31:48,774 INFO [train.py:421] (2/8) Epoch 4, batch 70200, loss[loss=2.44, over 1050.00 frames. , ppl: 11.476578595636363] tot_loss[loss=2.306, over 5404468.66 frames. , ppl: 10.03774781570945], batch size: 70
+2022-12-11 12:33:26,587 INFO [train.py:421] (2/8) Epoch 4, batch 70400, loss[loss=2.429, over 1750.00 frames. , ppl: 11.342961021608328] tot_loss[loss=2.308, over 5354232.65 frames. , ppl: 10.056357549084778], batch size: 70
+2022-12-11 12:35:03,418 INFO [train.py:421] (2/8) Epoch 4, batch 70600, loss[loss=2.36, over 2450.00 frames. , ppl: 10.593556451100591] tot_loss[loss=2.308, over 5362243.35 frames. , ppl: 10.058297471700884], batch size: 70
+2022-12-11 12:36:42,005 INFO [train.py:421] (2/8) Epoch 4, batch 70800, loss[loss=2.395, over 3150.00 frames. , ppl: 10.96956508859027] tot_loss[loss=2.308, over 5391073.78 frames. , ppl: 10.051233299482046], batch size: 70
+2022-12-11 12:38:21,442 INFO [train.py:421] (2/8) Epoch 4, batch 71000, loss[loss=2.424, over 1260.00 frames. , ppl: 11.286085720005504] tot_loss[loss=2.308, over 5373233.98 frames. , ppl: 10.05581496899964], batch size: 70
+2022-12-11 12:38:21,443 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 12:38:22,205 INFO [train.py:452] (2/8) Epoch 4, validation: loss=2.295, over 211138.00 frames. , ppl: 9.920506450462009
+2022-12-11 12:40:01,637 INFO [train.py:421] (2/8) Epoch 4, batch 71200, loss[loss=2.364, over 1680.00 frames. , ppl: 10.637717023332947] tot_loss[loss=2.307, over 5415496.75 frames. , ppl: 10.04896564321948], batch size: 70
+2022-12-11 12:41:40,328 INFO [train.py:421] (2/8) Epoch 4, batch 71400, loss[loss=2.381, over 1470.00 frames. , ppl: 10.819991500507253] tot_loss[loss=2.308, over 5386719.63 frames. , ppl: 10.055926836318095], batch size: 70
+2022-12-11 12:43:20,043 INFO [train.py:421] (2/8) Epoch 4, batch 71600, loss[loss=2.199, over 4550.00 frames. , ppl: 9.019024251632521] tot_loss[loss=2.307, over 5414444.60 frames. , ppl: 10.047696354739834], batch size: 70
+2022-12-11 12:45:06,136 INFO [train.py:421] (2/8) Epoch 4, batch 71800, loss[loss=2.395, over 1260.00 frames. , ppl: 10.965008136187373] tot_loss[loss=2.307, over 5441415.42 frames. , ppl: 10.040630206577918], batch size: 70
+2022-12-11 12:46:20,322 INFO [train.py:421] (2/8) Epoch 5, batch 0, loss[loss=2.293, over 3710.00 frames. , ppl: 9.908404421035582] tot_loss[loss=2.293, over 3710.00 frames. , ppl: 9.908404421035582], batch size: 70
+2022-12-11 12:48:02,165 INFO [train.py:421] (2/8) Epoch 5, batch 200, loss[loss=2.297, over 3010.00 frames. , ppl: 9.943868298894488] tot_loss[loss=2.291, over 530038.23 frames. , ppl: 9.883439711453322], batch size: 70
+2022-12-11 12:49:43,355 INFO [train.py:421] (2/8) Epoch 5, batch 400, loss[loss=2.442, over 1750.00 frames. , ppl: 11.4931361424409] tot_loss[loss=2.281, over 1096755.13 frames. , ppl: 9.787157389987653], batch size: 70
+2022-12-11 12:51:24,412 INFO [train.py:421] (2/8) Epoch 5, batch 600, loss[loss=2.57, over 1050.00 frames. , ppl: 13.068540693997537] tot_loss[loss=2.285, over 1512710.25 frames. , ppl: 9.822198613089732], batch size: 70
+2022-12-11 12:53:04,445 INFO [train.py:421] (2/8) Epoch 5, batch 800, loss[loss=2.384, over 2520.00 frames. , ppl: 10.853253597309152] tot_loss[loss=2.287, over 1908959.61 frames. , ppl: 9.84468776282165], batch size: 70
+2022-12-11 12:54:44,645 INFO [train.py:421] (2/8) Epoch 5, batch 1000, loss[loss=2.225, over 10500.00 frames. , ppl: 9.25032999654804] tot_loss[loss=2.288, over 2270555.92 frames. , ppl: 9.853738309179157], batch size: 70
+2022-12-11 12:54:44,646 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 12:54:45,418 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.90861440768603
+2022-12-11 12:56:25,125 INFO [train.py:421] (2/8) Epoch 5, batch 1200, loss[loss=2.413, over 3080.00 frames. , ppl: 11.166083830249478] tot_loss[loss=2.288, over 2597226.44 frames. , ppl: 9.856025765753673], batch size: 70
+2022-12-11 12:58:03,058 INFO [train.py:421] (2/8) Epoch 5, batch 1400, loss[loss=2.583, over 980.00 frames. , ppl: 13.23524372844277] tot_loss[loss=2.292, over 2833798.55 frames. , ppl: 9.893688642785627], batch size: 70
+2022-12-11 12:59:45,897 INFO [train.py:421] (2/8) Epoch 5, batch 1600, loss[loss=2.618, over 980.00 frames. , ppl: 13.711703810100703] tot_loss[loss=2.293, over 3087346.78 frames. , ppl: 9.908194964055939], batch size: 70
+2022-12-11 13:01:25,813 INFO [train.py:421] (2/8) Epoch 5, batch 1800, loss[loss=2.581, over 1050.00 frames. , ppl: 13.215952447555946] tot_loss[loss=2.296, over 3286475.79 frames. , ppl: 9.937865842164085], batch size: 70
+2022-12-11 13:03:08,689 INFO [train.py:421] (2/8) Epoch 5, batch 2000, loss[loss=2.411, over 1540.00 frames. , ppl: 11.149702116657245] tot_loss[loss=2.297, over 3482219.26 frames. , ppl: 9.940593540201192], batch size: 70
+2022-12-11 13:03:08,690 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:03:09,438 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.893647858407924
+2022-12-11 13:04:49,704 INFO [train.py:421] (2/8) Epoch 5, batch 2200, loss[loss=2.212, over 4550.00 frames. , ppl: 9.132116801034291] tot_loss[loss=2.297, over 3664896.16 frames. , ppl: 9.942553468537731], batch size: 70
+2022-12-11 13:06:31,502 INFO [train.py:421] (2/8) Epoch 5, batch 2400, loss[loss=2.25, over 4340.00 frames. , ppl: 9.489960639205064] tot_loss[loss=2.295, over 3865809.10 frames. , ppl: 9.925944919008364], batch size: 70
+2022-12-11 13:08:14,048 INFO [train.py:421] (2/8) Epoch 5, batch 2600, loss[loss=2.255, over 2870.00 frames. , ppl: 9.539301475106287] tot_loss[loss=2.297, over 3981422.21 frames. , ppl: 9.944334655205974], batch size: 70
+2022-12-11 13:09:54,954 INFO [train.py:421] (2/8) Epoch 5, batch 2800, loss[loss=2.328, over 3010.00 frames. , ppl: 10.259167737698894] tot_loss[loss=2.297, over 4143112.75 frames. , ppl: 9.940156819869669], batch size: 70
+2022-12-11 13:11:38,193 INFO [train.py:421] (2/8) Epoch 5, batch 3000, loss[loss=2.445, over 3430.00 frames. , ppl: 11.530499042490186] tot_loss[loss=2.295, over 4299865.48 frames. , ppl: 9.926264770811331], batch size: 70
+2022-12-11 13:11:38,194 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:11:38,953 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.878481675447484
+2022-12-11 13:13:17,481 INFO [train.py:421] (2/8) Epoch 5, batch 3200, loss[loss=2.271, over 3150.00 frames. , ppl: 9.692391011070995] tot_loss[loss=2.295, over 4438440.81 frames. , ppl: 9.924350762263613], batch size: 70
+2022-12-11 13:14:57,406 INFO [train.py:421] (2/8) Epoch 5, batch 3400, loss[loss=2.37, over 1820.00 frames. , ppl: 10.696315527483824] tot_loss[loss=2.295, over 4530078.73 frames. , ppl: 9.925344352654363], batch size: 70
+2022-12-11 13:16:37,748 INFO [train.py:421] (2/8) Epoch 5, batch 3600, loss[loss=2.258, over 3920.00 frames. , ppl: 9.566564217612127] tot_loss[loss=2.295, over 4625314.85 frames. , ppl: 9.92878095829202], batch size: 70
+2022-12-11 13:18:14,022 INFO [train.py:421] (2/8) Epoch 5, batch 3800, loss[loss=2.618, over 770.00 frames. , ppl: 13.714902709586696] tot_loss[loss=2.297, over 4674002.26 frames. , ppl: 9.9474057460938], batch size: 70
+2022-12-11 13:19:56,337 INFO [train.py:421] (2/8) Epoch 5, batch 4000, loss[loss=2.96, over 630.00 frames. , ppl: 19.301540798343275] tot_loss[loss=2.298, over 4733912.45 frames. , ppl: 9.95097011760639], batch size: 70
+2022-12-11 13:19:56,337 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:19:57,099 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.905227258803862
+2022-12-11 13:21:39,463 INFO [train.py:421] (2/8) Epoch 5, batch 4200, loss[loss=2.232, over 4410.00 frames. , ppl: 9.320191351114723] tot_loss[loss=2.297, over 4844563.31 frames. , ppl: 9.942268610959221], batch size: 70
+2022-12-11 13:23:20,083 INFO [train.py:421] (2/8) Epoch 5, batch 4400, loss[loss=2.191, over 5740.00 frames. , ppl: 8.942646203168424] tot_loss[loss=2.297, over 4902703.73 frames. , ppl: 9.943163699077468], batch size: 70
+2022-12-11 13:25:01,536 INFO [train.py:421] (2/8) Epoch 5, batch 4600, loss[loss=2.377, over 3010.00 frames. , ppl: 10.776555090745282] tot_loss[loss=2.296, over 5017440.83 frames. , ppl: 9.931171701566335], batch size: 70
+2022-12-11 13:26:39,712 INFO [train.py:421] (2/8) Epoch 5, batch 4800, loss[loss=2.372, over 1400.00 frames. , ppl: 10.721962675978933] tot_loss[loss=2.295, over 5048790.96 frames. , ppl: 9.928408687680848], batch size: 70
+2022-12-11 13:28:19,897 INFO [train.py:421] (2/8) Epoch 5, batch 5000, loss[loss=2.154, over 7140.00 frames. , ppl: 8.616368865769356] tot_loss[loss=2.295, over 5116663.65 frames. , ppl: 9.920470824469417], batch size: 70
+2022-12-11 13:28:19,898 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:28:20,645 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.89573182475821
+2022-12-11 13:30:01,901 INFO [train.py:421] (2/8) Epoch 5, batch 5200, loss[loss=2.692, over 770.00 frames. , ppl: 14.754928291502775] tot_loss[loss=2.295, over 5142374.46 frames. , ppl: 9.923524369830025], batch size: 70
+2022-12-11 13:31:43,064 INFO [train.py:421] (2/8) Epoch 5, batch 5400, loss[loss=4.118, over 350.00 frames. , ppl: 61.425928313381064] tot_loss[loss=2.294, over 5231830.49 frames. , ppl: 9.911892880028901], batch size: 70
+2022-12-11 13:33:23,011 INFO [train.py:421] (2/8) Epoch 5, batch 5600, loss[loss=2.515, over 980.00 frames. , ppl: 12.372046984004406] tot_loss[loss=2.294, over 5249532.50 frames. , ppl: 9.917002812327956], batch size: 70
+2022-12-11 13:35:01,458 INFO [train.py:421] (2/8) Epoch 5, batch 5800, loss[loss=2.205, over 12741.00 frames. , ppl: 9.074656985850737] tot_loss[loss=2.295, over 5255574.14 frames. , ppl: 9.924878068629665], batch size: 70
+2022-12-11 13:36:38,803 INFO [train.py:421] (2/8) Epoch 5, batch 6000, loss[loss=2.599, over 910.00 frames. , ppl: 13.452508971625164] tot_loss[loss=2.296, over 5235778.92 frames. , ppl: 9.936195112185352], batch size: 70
+2022-12-11 13:36:38,803 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:36:39,553 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.896351387792528
+2022-12-11 13:38:21,622 INFO [train.py:421] (2/8) Epoch 5, batch 6200, loss[loss=2.219, over 6510.00 frames. , ppl: 9.198590991430045] tot_loss[loss=2.297, over 5218347.89 frames. , ppl: 9.94873324670464], batch size: 70
+2022-12-11 13:39:57,535 INFO [train.py:421] (2/8) Epoch 5, batch 6400, loss[loss=2.524, over 840.00 frames. , ppl: 12.474783063252152] tot_loss[loss=2.298, over 5239355.74 frames. , ppl: 9.954339800407523], batch size: 70
+2022-12-11 13:41:36,935 INFO [train.py:421] (2/8) Epoch 5, batch 6600, loss[loss=2.245, over 3430.00 frames. , ppl: 9.444979469482506] tot_loss[loss=2.298, over 5272112.05 frames. , ppl: 9.957213968329823], batch size: 70
+2022-12-11 13:43:15,795 INFO [train.py:421] (2/8) Epoch 5, batch 6800, loss[loss=2.244, over 4690.00 frames. , ppl: 9.431874110644056] tot_loss[loss=2.301, over 5231383.71 frames. , ppl: 9.981500214599507], batch size: 70
+2022-12-11 13:44:54,863 INFO [train.py:421] (2/8) Epoch 5, batch 7000, loss[loss=2.826, over 560.00 frames. , ppl: 16.877326820837187] tot_loss[loss=2.301, over 5241727.40 frames. , ppl: 9.981921221805512], batch size: 70
+2022-12-11 13:44:54,864 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:44:55,610 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.88711469496257
+2022-12-11 13:46:38,710 INFO [train.py:421] (2/8) Epoch 5, batch 7200, loss[loss=2.27, over 2660.00 frames. , ppl: 9.681770280103157] tot_loss[loss=2.3, over 5271724.62 frames. , ppl: 9.977474001975207], batch size: 70
+2022-12-11 13:48:18,508 INFO [train.py:421] (2/8) Epoch 5, batch 7400, loss[loss=2.238, over 2730.00 frames. , ppl: 9.378723372112736] tot_loss[loss=2.301, over 5241650.19 frames. , ppl: 9.98736515924531], batch size: 70
+2022-12-11 13:49:58,377 INFO [train.py:421] (2/8) Epoch 5, batch 7600, loss[loss=2.283, over 2450.00 frames. , ppl: 9.806373441959797] tot_loss[loss=2.301, over 5272113.57 frames. , ppl: 9.983959825068473], batch size: 70
+2022-12-11 13:51:37,159 INFO [train.py:421] (2/8) Epoch 5, batch 7800, loss[loss=2.308, over 3850.00 frames. , ppl: 10.051411457556938] tot_loss[loss=2.302, over 5263326.36 frames. , ppl: 9.991686682398853], batch size: 70
+2022-12-11 13:53:18,458 INFO [train.py:421] (2/8) Epoch 5, batch 8000, loss[loss=4.146, over 350.00 frames. , ppl: 63.19030886559822] tot_loss[loss=2.302, over 5280776.76 frames. , ppl: 9.993055919543462], batch size: 70
+2022-12-11 13:53:18,459 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 13:53:19,219 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.903350895657939
+2022-12-11 13:54:59,219 INFO [train.py:421] (2/8) Epoch 5, batch 8200, loss[loss=2.536, over 770.00 frames. , ppl: 12.628244278085525] tot_loss[loss=2.301, over 5291978.83 frames. , ppl: 9.983629325673679], batch size: 70
+2022-12-11 13:56:40,090 INFO [train.py:421] (2/8) Epoch 5, batch 8400, loss[loss=2.273, over 3920.00 frames. , ppl: 9.706161419046554] tot_loss[loss=2.299, over 5336340.66 frames. , ppl: 9.966675044326028], batch size: 70
+2022-12-11 13:58:20,775 INFO [train.py:421] (2/8) Epoch 5, batch 8600, loss[loss=2.736, over 630.00 frames. , ppl: 15.428677450528195] tot_loss[loss=2.3, over 5313272.00 frames. , ppl: 9.978535310390226], batch size: 70
+2022-12-11 14:00:00,769 INFO [train.py:421] (2/8) Epoch 5, batch 8800, loss[loss=2.383, over 1960.00 frames. , ppl: 10.84184447180123] tot_loss[loss=2.299, over 5365800.24 frames. , ppl: 9.964552530302734], batch size: 70
+2022-12-11 14:01:38,184 INFO [train.py:421] (2/8) Epoch 5, batch 9000, loss[loss=2.334, over 3360.00 frames. , ppl: 10.318122675504851] tot_loss[loss=2.299, over 5348659.49 frames. , ppl: 9.966086502627116], batch size: 70
+2022-12-11 14:01:38,184 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:01:38,944 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.898942841256256
+2022-12-11 14:03:18,627 INFO [train.py:421] (2/8) Epoch 5, batch 9200, loss[loss=2.279, over 3010.00 frames. , ppl: 9.771512465243934] tot_loss[loss=2.299, over 5338354.77 frames. , ppl: 9.968283766486415], batch size: 70
+2022-12-11 14:04:53,956 INFO [train.py:421] (2/8) Epoch 5, batch 9400, loss[loss=2.261, over 2170.00 frames. , ppl: 9.593819476597206] tot_loss[loss=2.299, over 5316309.54 frames. , ppl: 9.967521041545314], batch size: 70
+2022-12-11 14:06:31,800 INFO [train.py:421] (2/8) Epoch 5, batch 9600, loss[loss=2.239, over 4060.00 frames. , ppl: 9.388622517132122] tot_loss[loss=2.3, over 5310601.37 frames. , ppl: 9.9695957983056], batch size: 70
+2022-12-11 14:08:10,021 INFO [train.py:421] (2/8) Epoch 5, batch 9800, loss[loss=2.553, over 770.00 frames. , ppl: 12.841890176862151] tot_loss[loss=2.3, over 5301985.78 frames. , ppl: 9.973502595482802], batch size: 70
+2022-12-11 14:09:51,392 INFO [train.py:421] (2/8) Epoch 5, batch 10000, loss[loss=2.256, over 3150.00 frames. , ppl: 9.549577673107901] tot_loss[loss=2.299, over 5341902.97 frames. , ppl: 9.963517519519444], batch size: 70
+2022-12-11 14:09:51,392 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:09:52,185 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.904837297826665
+2022-12-11 14:11:34,876 INFO [train.py:421] (2/8) Epoch 5, batch 10200, loss[loss=2.248, over 6230.00 frames. , ppl: 9.469822869572699] tot_loss[loss=2.3, over 5355972.88 frames. , ppl: 9.971469144974069], batch size: 70
+2022-12-11 14:13:12,042 INFO [train.py:421] (2/8) Epoch 5, batch 10400, loss[loss=2.34, over 1610.00 frames. , ppl: 10.385893496843801] tot_loss[loss=2.3, over 5333755.42 frames. , ppl: 9.97838929745139], batch size: 70
+2022-12-11 14:14:53,715 INFO [train.py:421] (2/8) Epoch 5, batch 10600, loss[loss=2.687, over 700.00 frames. , ppl: 14.68432330216069] tot_loss[loss=2.3, over 5395831.03 frames. , ppl: 9.970664585533417], batch size: 70
+2022-12-11 14:16:33,811 INFO [train.py:421] (2/8) Epoch 5, batch 10800, loss[loss=2.375, over 3290.00 frames. , ppl: 10.754030885757631] tot_loss[loss=2.3, over 5395969.63 frames. , ppl: 9.974894271672007], batch size: 70
+2022-12-11 14:18:14,047 INFO [train.py:421] (2/8) Epoch 5, batch 11000, loss[loss=2.292, over 3430.00 frames. , ppl: 9.896991046430758] tot_loss[loss=2.3, over 5397153.93 frames. , ppl: 9.972312252876614], batch size: 70
+2022-12-11 14:18:14,047 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:18:14,795 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.8904824753545
+2022-12-11 14:19:51,221 INFO [train.py:421] (2/8) Epoch 5, batch 11200, loss[loss=2.238, over 5040.00 frames. , ppl: 9.371800132393366] tot_loss[loss=2.3, over 5403076.45 frames. , ppl: 9.974285728787347], batch size: 70
+2022-12-11 14:21:32,934 INFO [train.py:421] (2/8) Epoch 5, batch 11400, loss[loss=2.338, over 2030.00 frames. , ppl: 10.359865069762929] tot_loss[loss=2.3, over 5410938.89 frames. , ppl: 9.976793517889657], batch size: 70
+2022-12-11 14:23:12,643 INFO [train.py:421] (2/8) Epoch 5, batch 11600, loss[loss=2.177, over 11130.00 frames. , ppl: 8.818010196486002] tot_loss[loss=2.299, over 5451709.07 frames. , ppl: 9.959576012535324], batch size: 70
+2022-12-11 14:24:53,833 INFO [train.py:421] (2/8) Epoch 5, batch 11800, loss[loss=2.452, over 1680.00 frames. , ppl: 11.616457063477776] tot_loss[loss=2.298, over 5475505.21 frames. , ppl: 9.955256570669485], batch size: 70
+2022-12-11 14:26:33,013 INFO [train.py:421] (2/8) Epoch 5, batch 12000, loss[loss=2.465, over 1330.00 frames. , ppl: 11.764490182690606] tot_loss[loss=2.299, over 5447609.87 frames. , ppl: 9.965105230619477], batch size: 70
+2022-12-11 14:26:33,014 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:26:33,774 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.883088329655019
+2022-12-11 14:28:11,331 INFO [train.py:421] (2/8) Epoch 5, batch 12200, loss[loss=2.195, over 7000.00 frames. , ppl: 8.97797176581429] tot_loss[loss=2.299, over 5423086.35 frames. , ppl: 9.968905796856019], batch size: 70
+2022-12-11 14:29:52,550 INFO [train.py:421] (2/8) Epoch 5, batch 12400, loss[loss=2.259, over 5530.00 frames. , ppl: 9.577292721026176] tot_loss[loss=2.299, over 5450452.46 frames. , ppl: 9.969064577804371], batch size: 70
+2022-12-11 14:31:29,829 INFO [train.py:421] (2/8) Epoch 5, batch 12600, loss[loss=2.724, over 770.00 frames. , ppl: 15.235946598094728] tot_loss[loss=2.3, over 5430399.65 frames. , ppl: 9.972455138681662], batch size: 70
+2022-12-11 14:33:09,806 INFO [train.py:421] (2/8) Epoch 5, batch 12800, loss[loss=2.386, over 1890.00 frames. , ppl: 10.868006714191226] tot_loss[loss=2.302, over 5355640.22 frames. , ppl: 9.992730795228708], batch size: 70
+2022-12-11 14:34:47,875 INFO [train.py:421] (2/8) Epoch 5, batch 13000, loss[loss=2.427, over 1400.00 frames. , ppl: 11.319306813906788] tot_loss[loss=2.303, over 5319064.65 frames. , ppl: 10.002670684346375], batch size: 70
+2022-12-11 14:34:47,876 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:34:48,636 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.882500313269402
+2022-12-11 14:36:27,618 INFO [train.py:421] (2/8) Epoch 5, batch 13200, loss[loss=2.617, over 770.00 frames. , ppl: 13.699934560398395] tot_loss[loss=2.304, over 5280784.81 frames. , ppl: 10.017136699116692], batch size: 70
+2022-12-11 14:38:12,162 INFO [train.py:421] (2/8) Epoch 5, batch 13400, loss[loss=2.252, over 4200.00 frames. , ppl: 9.508596988671155] tot_loss[loss=2.303, over 5313928.41 frames. , ppl: 10.008701806136632], batch size: 70
+2022-12-11 14:39:51,871 INFO [train.py:421] (2/8) Epoch 5, batch 13600, loss[loss=2.227, over 3920.00 frames. , ppl: 9.268719245974175] tot_loss[loss=2.303, over 5338833.34 frames. , ppl: 10.00603124157102], batch size: 70
+2022-12-11 14:41:35,231 INFO [train.py:421] (2/8) Epoch 5, batch 13800, loss[loss=2.409, over 1750.00 frames. , ppl: 11.12598632047743] tot_loss[loss=2.303, over 5340949.48 frames. , ppl: 10.006916808977415], batch size: 70
+2022-12-11 14:43:18,791 INFO [train.py:421] (2/8) Epoch 5, batch 14000, loss[loss=2.353, over 1820.00 frames. , ppl: 10.520104262367306] tot_loss[loss=2.303, over 5375471.53 frames. , ppl: 10.002709081356192], batch size: 70
+2022-12-11 14:43:18,791 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:43:19,550 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.88920899459524
+2022-12-11 14:45:00,093 INFO [train.py:421] (2/8) Epoch 5, batch 14200, loss[loss=2.386, over 2310.00 frames. , ppl: 10.870608021196078] tot_loss[loss=2.302, over 5376994.34 frames. , ppl: 9.992273618309513], batch size: 70
+2022-12-11 14:46:37,397 INFO [train.py:421] (2/8) Epoch 5, batch 14400, loss[loss=2.574, over 980.00 frames. , ppl: 13.111886856134394] tot_loss[loss=2.302, over 5365449.68 frames. , ppl: 9.992165893522062], batch size: 70
+2022-12-11 14:48:13,476 INFO [train.py:421] (2/8) Epoch 5, batch 14600, loss[loss=2.204, over 7140.00 frames. , ppl: 9.063467197780643] tot_loss[loss=2.301, over 5356205.98 frames. , ppl: 9.988349557864183], batch size: 70
+2022-12-11 14:49:54,282 INFO [train.py:421] (2/8) Epoch 5, batch 14800, loss[loss=2.241, over 5110.00 frames. , ppl: 9.399295564435862] tot_loss[loss=2.303, over 5322727.64 frames. , ppl: 10.004293602575892], batch size: 70
+2022-12-11 14:51:36,210 INFO [train.py:421] (2/8) Epoch 5, batch 15000, loss[loss=2.31, over 2380.00 frames. , ppl: 10.07328504687798] tot_loss[loss=2.302, over 5354382.56 frames. , ppl: 9.997101998197262], batch size: 70
+2022-12-11 14:51:36,211 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 14:51:36,973 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.882491537170422
+2022-12-11 14:53:18,611 INFO [train.py:421] (2/8) Epoch 5, batch 15200, loss[loss=2.455, over 1260.00 frames. , ppl: 11.64146819230091] tot_loss[loss=2.303, over 5353751.91 frames. , ppl: 10.000421451704536], batch size: 70
+2022-12-11 14:55:01,738 INFO [train.py:421] (2/8) Epoch 5, batch 15400, loss[loss=2.197, over 7910.00 frames. , ppl: 9.002125270100242] tot_loss[loss=2.301, over 5409378.62 frames. , ppl: 9.981828654278273], batch size: 70
+2022-12-11 14:56:39,993 INFO [train.py:421] (2/8) Epoch 5, batch 15600, loss[loss=2.222, over 7210.00 frames. , ppl: 9.223516368115435] tot_loss[loss=2.301, over 5404755.77 frames. , ppl: 9.986703655914672], batch size: 70
+2022-12-11 14:58:22,472 INFO [train.py:421] (2/8) Epoch 5, batch 15800, loss[loss=3.62, over 420.00 frames. , ppl: 37.3310485546201] tot_loss[loss=2.301, over 5390341.20 frames. , ppl: 9.983856385653127], batch size: 70
+2022-12-11 15:00:04,356 INFO [train.py:421] (2/8) Epoch 5, batch 16000, loss[loss=2.545, over 910.00 frames. , ppl: 12.741856523359555] tot_loss[loss=2.301, over 5403266.75 frames. , ppl: 9.98282610675353], batch size: 70
+2022-12-11 15:00:04,357 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:00:05,117 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.876905669210421
+2022-12-11 15:01:47,827 INFO [train.py:421] (2/8) Epoch 5, batch 16200, loss[loss=2.149, over 4900.00 frames. , ppl: 8.573908468119294] tot_loss[loss=2.3, over 5428226.84 frames. , ppl: 9.975190567673657], batch size: 70
+2022-12-11 15:03:30,619 INFO [train.py:421] (2/8) Epoch 5, batch 16400, loss[loss=2.206, over 4410.00 frames. , ppl: 9.076863684786003] tot_loss[loss=2.3, over 5440132.45 frames. , ppl: 9.974190553032622], batch size: 70
+2022-12-11 15:05:08,591 INFO [train.py:421] (2/8) Epoch 5, batch 16600, loss[loss=2.361, over 1470.00 frames. , ppl: 10.603667264685901] tot_loss[loss=2.299, over 5460521.05 frames. , ppl: 9.963647876415138], batch size: 70
+2022-12-11 15:06:50,353 INFO [train.py:421] (2/8) Epoch 5, batch 16800, loss[loss=2.227, over 3360.00 frames. , ppl: 9.27370679266764] tot_loss[loss=2.298, over 5494912.94 frames. , ppl: 9.952882805117907], batch size: 70
+2022-12-11 15:08:31,261 INFO [train.py:421] (2/8) Epoch 5, batch 17000, loss[loss=2.643, over 840.00 frames. , ppl: 14.0507745884636] tot_loss[loss=2.296, over 5563288.89 frames. , ppl: 9.935826387625516], batch size: 70
+2022-12-11 15:08:31,262 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:08:32,009 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.891302273858114
+2022-12-11 15:10:09,655 INFO [train.py:421] (2/8) Epoch 5, batch 17200, loss[loss=2.179, over 6720.00 frames. , ppl: 8.837280570437867] tot_loss[loss=2.296, over 5558049.91 frames. , ppl: 9.938109882649943], batch size: 70
+2022-12-11 15:11:48,015 INFO [train.py:421] (2/8) Epoch 5, batch 17400, loss[loss=2.204, over 10710.00 frames. , ppl: 9.056705409462829] tot_loss[loss=2.298, over 5526054.24 frames. , ppl: 9.94979979813178], batch size: 70
+2022-12-11 15:13:29,110 INFO [train.py:421] (2/8) Epoch 5, batch 17600, loss[loss=2.487, over 1400.00 frames. , ppl: 12.028941039159317] tot_loss[loss=2.298, over 5498826.75 frames. , ppl: 9.957559575208245], batch size: 70
+2022-12-11 15:15:10,239 INFO [train.py:421] (2/8) Epoch 5, batch 17800, loss[loss=2.939, over 560.00 frames. , ppl: 18.896650425133743] tot_loss[loss=2.299, over 5447360.75 frames. , ppl: 9.961587017175773], batch size: 70
+2022-12-11 15:16:49,287 INFO [train.py:421] (2/8) Epoch 5, batch 18000, loss[loss=2.516, over 1050.00 frames. , ppl: 12.384120976395904] tot_loss[loss=2.299, over 5469275.30 frames. , ppl: 9.960360238451859], batch size: 70
+2022-12-11 15:16:49,288 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:16:50,060 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.884169376428208
+2022-12-11 15:18:27,178 INFO [train.py:421] (2/8) Epoch 5, batch 18200, loss[loss=2.276, over 2170.00 frames. , ppl: 9.73738413460448] tot_loss[loss=2.299, over 5487340.74 frames. , ppl: 9.962496053041978], batch size: 70
+2022-12-11 15:20:07,209 INFO [train.py:421] (2/8) Epoch 5, batch 18400, loss[loss=2.343, over 1680.00 frames. , ppl: 10.413250161781468] tot_loss[loss=2.299, over 5476785.41 frames. , ppl: 9.964403479298728], batch size: 70
+2022-12-11 15:21:46,277 INFO [train.py:421] (2/8) Epoch 5, batch 18600, loss[loss=2.212, over 5810.00 frames. , ppl: 9.13644082166634] tot_loss[loss=2.298, over 5511169.44 frames. , ppl: 9.956115658888942], batch size: 70
+2022-12-11 15:23:25,436 INFO [train.py:421] (2/8) Epoch 5, batch 18800, loss[loss=2.242, over 4130.00 frames. , ppl: 9.414244496202011] tot_loss[loss=2.297, over 5538518.18 frames. , ppl: 9.943015330774118], batch size: 70
+2022-12-11 15:25:02,768 INFO [train.py:421] (2/8) Epoch 5, batch 19000, loss[loss=2.567, over 700.00 frames. , ppl: 13.025811594355568] tot_loss[loss=2.299, over 5487514.56 frames. , ppl: 9.959848066499918], batch size: 70
+2022-12-11 15:25:02,769 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:25:03,528 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.892699015651312
+2022-12-11 15:26:43,245 INFO [train.py:421] (2/8) Epoch 5, batch 19200, loss[loss=2.307, over 3850.00 frames. , ppl: 10.041271849692073] tot_loss[loss=2.296, over 5575098.98 frames. , ppl: 9.934949282322725], batch size: 70
+2022-12-11 15:28:24,621 INFO [train.py:421] (2/8) Epoch 5, batch 19400, loss[loss=2.537, over 840.00 frames. , ppl: 12.6395445690862] tot_loss[loss=2.295, over 5598269.56 frames. , ppl: 9.928685911904802], batch size: 70
+2022-12-11 15:30:03,434 INFO [train.py:421] (2/8) Epoch 5, batch 19600, loss[loss=2.532, over 1610.00 frames. , ppl: 12.576506103235282] tot_loss[loss=2.295, over 5596608.07 frames. , ppl: 9.929283338741879], batch size: 70
+2022-12-11 15:31:37,912 INFO [train.py:421] (2/8) Epoch 5, batch 19800, loss[loss=2.458, over 1470.00 frames. , ppl: 11.677412062584787] tot_loss[loss=2.298, over 5519849.61 frames. , ppl: 9.95046493686272], batch size: 70
+2022-12-11 15:33:17,500 INFO [train.py:421] (2/8) Epoch 5, batch 20000, loss[loss=2.318, over 3220.00 frames. , ppl: 10.15864999941874] tot_loss[loss=2.297, over 5530725.29 frames. , ppl: 9.941318062276455], batch size: 70
+2022-12-11 15:33:17,500 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:33:18,262 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.897619928623687
+2022-12-11 15:34:59,004 INFO [train.py:421] (2/8) Epoch 5, batch 20200, loss[loss=2.171, over 3220.00 frames. , ppl: 8.767449852882173] tot_loss[loss=2.297, over 5539375.02 frames. , ppl: 9.941591595399462], batch size: 70
+2022-12-11 15:36:41,035 INFO [train.py:421] (2/8) Epoch 5, batch 20400, loss[loss=2.321, over 3850.00 frames. , ppl: 10.188730065254722] tot_loss[loss=2.297, over 5523372.51 frames. , ppl: 9.946126153407189], batch size: 70
+2022-12-11 15:38:20,075 INFO [train.py:421] (2/8) Epoch 5, batch 20600, loss[loss=2.329, over 2450.00 frames. , ppl: 10.269550706899603] tot_loss[loss=2.296, over 5565146.49 frames. , ppl: 9.934026407926767], batch size: 70
+2022-12-11 15:40:02,095 INFO [train.py:421] (2/8) Epoch 5, batch 20800, loss[loss=2.338, over 2590.00 frames. , ppl: 10.356198336619487] tot_loss[loss=2.296, over 5582324.95 frames. , ppl: 9.937296522209854], batch size: 70
+2022-12-11 15:41:40,027 INFO [train.py:421] (2/8) Epoch 5, batch 21000, loss[loss=2.455, over 1400.00 frames. , ppl: 11.643385451596394] tot_loss[loss=2.297, over 5547716.41 frames. , ppl: 9.94180784144071], batch size: 70
+2022-12-11 15:41:40,027 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:41:40,786 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.8614396619666
+2022-12-11 15:43:22,020 INFO [train.py:421] (2/8) Epoch 5, batch 21200, loss[loss=2.319, over 1820.00 frames. , ppl: 10.163513444654027] tot_loss[loss=2.298, over 5505566.26 frames. , ppl: 9.955641644469571], batch size: 70
+2022-12-11 15:45:08,222 INFO [train.py:421] (2/8) Epoch 5, batch 21400, loss[loss=2.308, over 3500.00 frames. , ppl: 10.057292477995414] tot_loss[loss=2.297, over 5528341.91 frames. , ppl: 9.943711153594483], batch size: 70
+2022-12-11 15:46:45,918 INFO [train.py:421] (2/8) Epoch 5, batch 21600, loss[loss=2.463, over 1120.00 frames. , ppl: 11.739869271833427] tot_loss[loss=2.298, over 5494461.45 frames. , ppl: 9.953848540540768], batch size: 70
+2022-12-11 15:48:20,255 INFO [train.py:421] (2/8) Epoch 5, batch 21800, loss[loss=2.225, over 3640.00 frames. , ppl: 9.251139541731936] tot_loss[loss=2.299, over 5487414.60 frames. , ppl: 9.961099243585222], batch size: 70
+2022-12-11 15:49:59,709 INFO [train.py:421] (2/8) Epoch 5, batch 22000, loss[loss=2.349, over 1960.00 frames. , ppl: 10.472381204055822] tot_loss[loss=2.299, over 5442870.45 frames. , ppl: 9.967144882090093], batch size: 70
+2022-12-11 15:49:59,709 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:50:00,468 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.884660933415132
+2022-12-11 15:51:45,352 INFO [train.py:421] (2/8) Epoch 5, batch 22200, loss[loss=2.466, over 1680.00 frames. , ppl: 11.776643696162669] tot_loss[loss=2.3, over 5420554.17 frames. , ppl: 9.972256923300984], batch size: 70
+2022-12-11 15:53:28,191 INFO [train.py:421] (2/8) Epoch 5, batch 22400, loss[loss=2.274, over 2940.00 frames. , ppl: 9.719836130231924] tot_loss[loss=2.298, over 5477131.43 frames. , ppl: 9.956843894035677], batch size: 70
+2022-12-11 15:55:07,005 INFO [train.py:421] (2/8) Epoch 5, batch 22600, loss[loss=2.214, over 1610.00 frames. , ppl: 9.148614287375887] tot_loss[loss=2.298, over 5485494.48 frames. , ppl: 9.954376274623753], batch size: 70
+2022-12-11 15:56:53,257 INFO [train.py:421] (2/8) Epoch 5, batch 22800, loss[loss=2.436, over 1050.00 frames. , ppl: 11.426688129923523] tot_loss[loss=2.297, over 5501901.96 frames. , ppl: 9.948949093530262], batch size: 70
+2022-12-11 15:58:36,997 INFO [train.py:421] (2/8) Epoch 5, batch 23000, loss[loss=2.296, over 1400.00 frames. , ppl: 9.932857631852718] tot_loss[loss=2.299, over 5434469.61 frames. , ppl: 9.962375366120888], batch size: 70
+2022-12-11 15:58:36,997 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 15:58:37,746 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.87439159731824
+2022-12-11 16:00:16,678 INFO [train.py:421] (2/8) Epoch 5, batch 23200, loss[loss=2.222, over 5180.00 frames. , ppl: 9.221310632079522] tot_loss[loss=2.297, over 5515560.89 frames. , ppl: 9.94243726612554], batch size: 70
+2022-12-11 16:01:54,225 INFO [train.py:421] (2/8) Epoch 5, batch 23400, loss[loss=2.35, over 1050.00 frames. , ppl: 10.488515303985865] tot_loss[loss=2.298, over 5480374.49 frames. , ppl: 9.951267362304128], batch size: 70
+2022-12-11 16:03:33,363 INFO [train.py:421] (2/8) Epoch 5, batch 23600, loss[loss=2.327, over 3150.00 frames. , ppl: 10.247073857387921] tot_loss[loss=2.298, over 5496821.73 frames. , ppl: 9.954695578048181], batch size: 70
+2022-12-11 16:05:15,464 INFO [train.py:421] (2/8) Epoch 5, batch 23800, loss[loss=2.459, over 1330.00 frames. , ppl: 11.690607695395265] tot_loss[loss=2.297, over 5540076.69 frames. , ppl: 9.949089185829084], batch size: 70
+2022-12-11 16:06:54,589 INFO [train.py:421] (2/8) Epoch 5, batch 24000, loss[loss=2.174, over 4200.00 frames. , ppl: 8.796624114710644] tot_loss[loss=2.299, over 5498643.35 frames. , ppl: 9.960716182151554], batch size: 70
+2022-12-11 16:06:54,589 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:06:55,332 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.869746633606848
+2022-12-11 16:08:33,924 INFO [train.py:421] (2/8) Epoch 5, batch 24200, loss[loss=2.309, over 4970.00 frames. , ppl: 10.061920156153073] tot_loss[loss=2.299, over 5491144.60 frames. , ppl: 9.960721861340984], batch size: 70
+2022-12-11 16:10:16,216 INFO [train.py:421] (2/8) Epoch 5, batch 24400, loss[loss=2.294, over 2450.00 frames. , ppl: 9.915337774824417] tot_loss[loss=2.299, over 5454326.41 frames. , ppl: 9.969039956978847], batch size: 70
+2022-12-11 16:11:57,882 INFO [train.py:421] (2/8) Epoch 5, batch 24600, loss[loss=2.342, over 3150.00 frames. , ppl: 10.403655898374787] tot_loss[loss=2.299, over 5483116.69 frames. , ppl: 9.965793140769625], batch size: 70
+2022-12-11 16:13:31,672 INFO [train.py:421] (2/8) Epoch 5, batch 24800, loss[loss=3.129, over 490.00 frames. , ppl: 22.854252089623554] tot_loss[loss=2.3, over 5435602.02 frames. , ppl: 9.977574663940427], batch size: 70
+2022-12-11 16:15:09,415 INFO [train.py:421] (2/8) Epoch 5, batch 25000, loss[loss=2.437, over 1400.00 frames. , ppl: 11.444080636824134] tot_loss[loss=2.301, over 5420509.29 frames. , ppl: 9.987138860469699], batch size: 70
+2022-12-11 16:15:09,416 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:15:10,164 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.867343916412231
+2022-12-11 16:16:49,431 INFO [train.py:421] (2/8) Epoch 5, batch 25200, loss[loss=2.124, over 6720.00 frames. , ppl: 8.361931892354802] tot_loss[loss=2.302, over 5394422.25 frames. , ppl: 9.989917846087687], batch size: 70
+2022-12-11 16:18:26,295 INFO [train.py:421] (2/8) Epoch 5, batch 25400, loss[loss=2.306, over 2800.00 frames. , ppl: 10.034605712946442] tot_loss[loss=2.302, over 5393067.46 frames. , ppl: 9.992587874973081], batch size: 70
+2022-12-11 16:20:05,519 INFO [train.py:421] (2/8) Epoch 5, batch 25600, loss[loss=2.411, over 1750.00 frames. , ppl: 11.143141747495443] tot_loss[loss=2.301, over 5420910.38 frames. , ppl: 9.987918501230538], batch size: 70
+2022-12-11 16:21:46,552 INFO [train.py:421] (2/8) Epoch 5, batch 25800, loss[loss=2.282, over 2870.00 frames. , ppl: 9.795641156712646] tot_loss[loss=2.301, over 5451008.84 frames. , ppl: 9.98548294302587], batch size: 70
+2022-12-11 16:23:25,042 INFO [train.py:421] (2/8) Epoch 5, batch 26000, loss[loss=2.331, over 2520.00 frames. , ppl: 10.287264285445524] tot_loss[loss=2.302, over 5429030.75 frames. , ppl: 9.990974122349671], batch size: 70
+2022-12-11 16:23:25,042 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:23:25,801 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.864416167537827
+2022-12-11 16:25:05,333 INFO [train.py:421] (2/8) Epoch 5, batch 26200, loss[loss=2.615, over 910.00 frames. , ppl: 13.670265969323866] tot_loss[loss=2.3, over 5472583.00 frames. , ppl: 9.97491955248408], batch size: 70
+2022-12-11 16:26:45,397 INFO [train.py:421] (2/8) Epoch 5, batch 26400, loss[loss=2.547, over 1470.00 frames. , ppl: 12.774088109197718] tot_loss[loss=2.3, over 5472142.58 frames. , ppl: 9.976656815925656], batch size: 70
+2022-12-11 16:28:28,145 INFO [train.py:421] (2/8) Epoch 5, batch 26600, loss[loss=3.298, over 490.00 frames. , ppl: 27.06827868221877] tot_loss[loss=2.3, over 5483886.85 frames. , ppl: 9.971732952532834], batch size: 70
+2022-12-11 16:30:12,513 INFO [train.py:421] (2/8) Epoch 5, batch 26800, loss[loss=2.25, over 4060.00 frames. , ppl: 9.485955957447976] tot_loss[loss=2.298, over 5521199.51 frames. , ppl: 9.957493117931458], batch size: 70
+2022-12-11 16:31:49,995 INFO [train.py:421] (2/8) Epoch 5, batch 27000, loss[loss=2.454, over 1260.00 frames. , ppl: 11.636666842475856] tot_loss[loss=2.299, over 5506717.17 frames. , ppl: 9.9642687933642], batch size: 70
+2022-12-11 16:31:49,995 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:31:50,754 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.875375224811005
+2022-12-11 16:33:32,081 INFO [train.py:421] (2/8) Epoch 5, batch 27200, loss[loss=2.288, over 1890.00 frames. , ppl: 9.85778360721871] tot_loss[loss=2.298, over 5518975.42 frames. , ppl: 9.958681742936841], batch size: 70
+2022-12-11 16:35:10,657 INFO [train.py:421] (2/8) Epoch 5, batch 27400, loss[loss=2.283, over 3220.00 frames. , ppl: 9.805961583060284] tot_loss[loss=2.298, over 5544202.27 frames. , ppl: 9.953254013250705], batch size: 70
+2022-12-11 16:36:49,937 INFO [train.py:421] (2/8) Epoch 5, batch 27600, loss[loss=2.603, over 840.00 frames. , ppl: 13.507303020093481] tot_loss[loss=2.299, over 5511658.02 frames. , ppl: 9.961619494502111], batch size: 70
+2022-12-11 16:38:28,348 INFO [train.py:421] (2/8) Epoch 5, batch 27800, loss[loss=2.324, over 1890.00 frames. , ppl: 10.211364966490587] tot_loss[loss=2.298, over 5514941.86 frames. , ppl: 9.954232104149911], batch size: 70
+2022-12-11 16:40:08,929 INFO [train.py:421] (2/8) Epoch 5, batch 28000, loss[loss=2.233, over 9450.00 frames. , ppl: 9.328244820227988] tot_loss[loss=2.298, over 5513329.50 frames. , ppl: 9.949843594167447], batch size: 70
+2022-12-11 16:40:08,929 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:40:09,677 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.858846353025392
+2022-12-11 16:41:47,857 INFO [train.py:421] (2/8) Epoch 5, batch 28200, loss[loss=2.275, over 6580.00 frames. , ppl: 9.727726994777031] tot_loss[loss=2.298, over 5509173.89 frames. , ppl: 9.949589898734834], batch size: 70
+2022-12-11 16:43:28,937 INFO [train.py:421] (2/8) Epoch 5, batch 28400, loss[loss=2.542, over 1120.00 frames. , ppl: 12.7093218510701] tot_loss[loss=2.297, over 5524612.37 frames. , ppl: 9.948028258059635], batch size: 70
+2022-12-11 16:45:09,673 INFO [train.py:421] (2/8) Epoch 5, batch 28600, loss[loss=2.275, over 2380.00 frames. , ppl: 9.728102626209472] tot_loss[loss=2.298, over 5514725.81 frames. , ppl: 9.955306830662042], batch size: 70
+2022-12-11 16:46:50,655 INFO [train.py:421] (2/8) Epoch 5, batch 28800, loss[loss=2.33, over 2940.00 frames. , ppl: 10.282907317784082] tot_loss[loss=2.297, over 5563323.61 frames. , ppl: 9.943529791613493], batch size: 70
+2022-12-11 16:48:28,317 INFO [train.py:421] (2/8) Epoch 5, batch 29000, loss[loss=2.239, over 4200.00 frames. , ppl: 9.385989066096638] tot_loss[loss=2.296, over 5586222.14 frames. , ppl: 9.935135066243715], batch size: 70
+2022-12-11 16:48:28,318 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:48:29,078 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.86421176864417
+2022-12-11 16:50:05,458 INFO [train.py:421] (2/8) Epoch 5, batch 29200, loss[loss=2.378, over 840.00 frames. , ppl: 10.778231023487303] tot_loss[loss=2.295, over 5607964.67 frames. , ppl: 9.92350884507282], batch size: 70
+2022-12-11 16:51:46,725 INFO [train.py:421] (2/8) Epoch 5, batch 29400, loss[loss=2.335, over 2240.00 frames. , ppl: 10.333866885338983] tot_loss[loss=2.294, over 5614521.58 frames. , ppl: 9.912799680304012], batch size: 70
+2022-12-11 16:53:26,109 INFO [train.py:421] (2/8) Epoch 5, batch 29600, loss[loss=2.28, over 2380.00 frames. , ppl: 9.780973331789973] tot_loss[loss=2.293, over 5631332.41 frames. , ppl: 9.899690249608813], batch size: 70
+2022-12-11 16:55:06,784 INFO [train.py:421] (2/8) Epoch 5, batch 29800, loss[loss=4.165, over 350.00 frames. , ppl: 64.4078436381303] tot_loss[loss=2.293, over 5614203.97 frames. , ppl: 9.90488222132788], batch size: 70
+2022-12-11 16:56:46,812 INFO [train.py:421] (2/8) Epoch 5, batch 30000, loss[loss=2.148, over 4900.00 frames. , ppl: 8.567045391257187] tot_loss[loss=2.293, over 5640214.28 frames. , ppl: 9.905761703864032], batch size: 70
+2022-12-11 16:56:46,812 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 16:56:47,572 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.86519146183926
+2022-12-11 16:58:29,824 INFO [train.py:421] (2/8) Epoch 5, batch 30200, loss[loss=2.957, over 560.00 frames. , ppl: 19.23135647855246] tot_loss[loss=2.294, over 5617694.28 frames. , ppl: 9.916623374039325], batch size: 70
+2022-12-11 17:00:10,447 INFO [train.py:421] (2/8) Epoch 5, batch 30400, loss[loss=3.16, over 490.00 frames. , ppl: 23.567872662253766] tot_loss[loss=2.296, over 5573240.18 frames. , ppl: 9.929517062443715], batch size: 70
+2022-12-11 17:01:51,074 INFO [train.py:421] (2/8) Epoch 5, batch 30600, loss[loss=2.331, over 1610.00 frames. , ppl: 10.289806124492374] tot_loss[loss=2.297, over 5546717.63 frames. , ppl: 9.942618732046402], batch size: 70
+2022-12-11 17:03:31,209 INFO [train.py:421] (2/8) Epoch 5, batch 30800, loss[loss=2.246, over 6370.00 frames. , ppl: 9.450643791193894] tot_loss[loss=2.295, over 5592267.71 frames. , ppl: 9.927183035737485], batch size: 70
+2022-12-11 17:05:13,223 INFO [train.py:421] (2/8) Epoch 5, batch 31000, loss[loss=2.613, over 700.00 frames. , ppl: 13.64426709773295] tot_loss[loss=2.295, over 5603765.63 frames. , ppl: 9.92322918117559], batch size: 70
+2022-12-11 17:05:13,224 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:05:13,971 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.862906635580453
+2022-12-11 17:06:53,704 INFO [train.py:421] (2/8) Epoch 5, batch 31200, loss[loss=2.32, over 2380.00 frames. , ppl: 10.176002488886645] tot_loss[loss=2.293, over 5663642.23 frames. , ppl: 9.90600581714101], batch size: 70
+2022-12-11 17:08:34,095 INFO [train.py:421] (2/8) Epoch 5, batch 31400, loss[loss=2.252, over 8470.00 frames. , ppl: 9.508884847290593] tot_loss[loss=2.294, over 5628307.86 frames. , ppl: 9.916734986542897], batch size: 70
+2022-12-11 17:10:13,743 INFO [train.py:421] (2/8) Epoch 5, batch 31600, loss[loss=2.309, over 2870.00 frames. , ppl: 10.06538638363277] tot_loss[loss=2.294, over 5598902.79 frames. , ppl: 9.918879943309335], batch size: 70
+2022-12-11 17:11:57,220 INFO [train.py:421] (2/8) Epoch 5, batch 31800, loss[loss=2.195, over 7420.00 frames. , ppl: 8.981327694295786] tot_loss[loss=2.294, over 5620998.45 frames. , ppl: 9.917503585712948], batch size: 70
+2022-12-11 17:13:39,204 INFO [train.py:421] (2/8) Epoch 5, batch 32000, loss[loss=2.192, over 7840.00 frames. , ppl: 8.956721330500196] tot_loss[loss=2.294, over 5637163.26 frames. , ppl: 9.915305039509354], batch size: 70
+2022-12-11 17:13:39,205 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:13:39,950 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.866756836775595
+2022-12-11 17:15:20,454 INFO [train.py:421] (2/8) Epoch 5, batch 32200, loss[loss=2.42, over 1330.00 frames. , ppl: 11.249021088299392] tot_loss[loss=2.295, over 5601204.16 frames. , ppl: 9.92671758853836], batch size: 70
+2022-12-11 17:17:01,704 INFO [train.py:421] (2/8) Epoch 5, batch 32400, loss[loss=2.605, over 980.00 frames. , ppl: 13.531962210431216] tot_loss[loss=2.295, over 5605336.62 frames. , ppl: 9.925984135190904], batch size: 70
+2022-12-11 17:18:38,646 INFO [train.py:421] (2/8) Epoch 5, batch 32600, loss[loss=2.259, over 3640.00 frames. , ppl: 9.57007429164072] tot_loss[loss=2.296, over 5570692.72 frames. , ppl: 9.936563975016364], batch size: 70
+2022-12-11 17:20:17,523 INFO [train.py:421] (2/8) Epoch 5, batch 32800, loss[loss=2.278, over 5390.00 frames. , ppl: 9.754618815412645] tot_loss[loss=2.297, over 5564351.45 frames. , ppl: 9.944703921390117], batch size: 70
+2022-12-11 17:21:54,553 INFO [train.py:421] (2/8) Epoch 5, batch 33000, loss[loss=2.331, over 2590.00 frames. , ppl: 10.283596002239616] tot_loss[loss=2.297, over 5547262.30 frames. , ppl: 9.946445136477855], batch size: 70
+2022-12-11 17:21:54,553 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:21:55,298 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.874875360260138
+2022-12-11 17:23:34,425 INFO [train.py:421] (2/8) Epoch 5, batch 33200, loss[loss=3.079, over 560.00 frames. , ppl: 21.72649339559756] tot_loss[loss=2.299, over 5488985.63 frames. , ppl: 9.960013170037401], batch size: 70
+2022-12-11 17:25:20,345 INFO [train.py:421] (2/8) Epoch 5, batch 33400, loss[loss=2.201, over 5390.00 frames. , ppl: 9.036217995735548] tot_loss[loss=2.3, over 5459259.87 frames. , ppl: 9.973375564423096], batch size: 70
+2022-12-11 17:27:01,179 INFO [train.py:421] (2/8) Epoch 5, batch 33600, loss[loss=2.276, over 2030.00 frames. , ppl: 9.738838345947485] tot_loss[loss=2.299, over 5501124.98 frames. , ppl: 9.961639980513727], batch size: 70
+2022-12-11 17:28:43,674 INFO [train.py:421] (2/8) Epoch 5, batch 33800, loss[loss=2.308, over 2940.00 frames. , ppl: 10.052634656214774] tot_loss[loss=2.298, over 5503669.64 frames. , ppl: 9.959105169344314], batch size: 70
+2022-12-11 17:30:21,134 INFO [train.py:421] (2/8) Epoch 5, batch 34000, loss[loss=2.254, over 5740.00 frames. , ppl: 9.527161437437645] tot_loss[loss=2.299, over 5496198.21 frames. , ppl: 9.966165417048666], batch size: 70
+2022-12-11 17:30:21,135 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:30:21,882 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.858973302750663
+2022-12-11 17:32:01,801 INFO [train.py:421] (2/8) Epoch 5, batch 34200, loss[loss=2.433, over 1330.00 frames. , ppl: 11.392916953113952] tot_loss[loss=2.3, over 5465450.63 frames. , ppl: 9.975364556266095], batch size: 70
+2022-12-11 17:33:39,848 INFO [train.py:421] (2/8) Epoch 5, batch 34400, loss[loss=3.84, over 420.00 frames. , ppl: 46.51990898107072] tot_loss[loss=2.302, over 5422439.10 frames. , ppl: 9.991903419796145], batch size: 70
+2022-12-11 17:35:26,817 INFO [train.py:421] (2/8) Epoch 5, batch 34600, loss[loss=2.509, over 1470.00 frames. , ppl: 12.296833418565056] tot_loss[loss=2.301, over 5426765.70 frames. , ppl: 9.98815303980221], batch size: 70
+2022-12-11 17:37:05,657 INFO [train.py:421] (2/8) Epoch 5, batch 34800, loss[loss=2.468, over 1540.00 frames. , ppl: 11.795363571944463] tot_loss[loss=2.302, over 5390390.75 frames. , ppl: 9.996514127530112], batch size: 70
+2022-12-11 17:38:42,833 INFO [train.py:421] (2/8) Epoch 5, batch 35000, loss[loss=3.625, over 420.00 frames. , ppl: 37.532969243938325] tot_loss[loss=2.302, over 5389742.89 frames. , ppl: 9.992643696863189], batch size: 70
+2022-12-11 17:38:42,833 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:38:43,595 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.860366938807065
+2022-12-11 17:40:23,338 INFO [train.py:421] (2/8) Epoch 5, batch 35200, loss[loss=2.34, over 1470.00 frames. , ppl: 10.385356688551038] tot_loss[loss=2.302, over 5392180.90 frames. , ppl: 9.991111714017208], batch size: 70
+2022-12-11 17:42:03,891 INFO [train.py:421] (2/8) Epoch 5, batch 35400, loss[loss=2.197, over 3080.00 frames. , ppl: 8.996292410293801] tot_loss[loss=2.303, over 5374396.74 frames. , ppl: 10.003447620315754], batch size: 70
+2022-12-11 17:44:04,291 INFO [train.py:421] (2/8) Epoch 5, batch 35600, loss[loss=2.753, over 770.00 frames. , ppl: 15.681853020502816] tot_loss[loss=2.301, over 5440116.03 frames. , ppl: 9.982473021128293], batch size: 70
+2022-12-11 17:45:45,024 INFO [train.py:421] (2/8) Epoch 5, batch 35800, loss[loss=2.411, over 910.00 frames. , ppl: 11.141641810142863] tot_loss[loss=2.3, over 5460795.44 frames. , ppl: 9.97666036515308], batch size: 70
+2022-12-11 17:47:22,412 INFO [train.py:421] (2/8) Epoch 5, batch 36000, loss[loss=2.244, over 4270.00 frames. , ppl: 9.434295330831391] tot_loss[loss=2.3, over 5459576.51 frames. , ppl: 9.977142465190406], batch size: 70
+2022-12-11 17:47:22,413 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:47:23,172 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.84657788756117
+2022-12-11 17:49:05,397 INFO [train.py:421] (2/8) Epoch 5, batch 36200, loss[loss=2.256, over 7070.00 frames. , ppl: 9.54333172158078] tot_loss[loss=2.299, over 5477415.94 frames. , ppl: 9.963770998841158], batch size: 70
+2022-12-11 17:50:49,468 INFO [train.py:421] (2/8) Epoch 5, batch 36400, loss[loss=2.208, over 3990.00 frames. , ppl: 9.097298153141427] tot_loss[loss=2.298, over 5486948.57 frames. , ppl: 9.956711383302997], batch size: 70
+2022-12-11 17:52:28,479 INFO [train.py:421] (2/8) Epoch 5, batch 36600, loss[loss=2.721, over 770.00 frames. , ppl: 15.20064062824651] tot_loss[loss=2.298, over 5503202.89 frames. , ppl: 9.95019738391085], batch size: 70
+2022-12-11 17:54:09,080 INFO [train.py:421] (2/8) Epoch 5, batch 36800, loss[loss=2.264, over 7210.00 frames. , ppl: 9.625527865389074] tot_loss[loss=2.297, over 5497333.01 frames. , ppl: 9.949144571045663], batch size: 70
+2022-12-11 17:55:52,423 INFO [train.py:421] (2/8) Epoch 5, batch 37000, loss[loss=2.366, over 1540.00 frames. , ppl: 10.65575670238919] tot_loss[loss=2.297, over 5499624.25 frames. , ppl: 9.947094199012513], batch size: 70
+2022-12-11 17:55:52,423 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 17:55:53,182 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.841108451353056
+2022-12-11 17:57:35,866 INFO [train.py:421] (2/8) Epoch 5, batch 37200, loss[loss=2.383, over 3080.00 frames. , ppl: 10.83466108785837] tot_loss[loss=2.299, over 5481848.62 frames. , ppl: 9.960361468719825], batch size: 70
+2022-12-11 17:59:14,078 INFO [train.py:421] (2/8) Epoch 5, batch 37400, loss[loss=2.477, over 1190.00 frames. , ppl: 11.908986469092739] tot_loss[loss=2.3, over 5480731.33 frames. , ppl: 9.970244657739176], batch size: 70
+2022-12-11 18:00:54,461 INFO [train.py:421] (2/8) Epoch 5, batch 37600, loss[loss=2.408, over 3150.00 frames. , ppl: 11.116946681567748] tot_loss[loss=2.3, over 5477814.54 frames. , ppl: 9.977495247413621], batch size: 70
+2022-12-11 18:02:36,711 INFO [train.py:421] (2/8) Epoch 5, batch 37800, loss[loss=2.27, over 2170.00 frames. , ppl: 9.676190532208436] tot_loss[loss=2.3, over 5450475.54 frames. , ppl: 9.976318392682147], batch size: 70
+2022-12-11 18:04:14,228 INFO [train.py:421] (2/8) Epoch 5, batch 38000, loss[loss=2.256, over 2800.00 frames. , ppl: 9.548919896659008] tot_loss[loss=2.299, over 5472467.78 frames. , ppl: 9.966944759944992], batch size: 70
+2022-12-11 18:04:14,229 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:04:14,993 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.84971901441481
+2022-12-11 18:05:55,141 INFO [train.py:421] (2/8) Epoch 5, batch 38200, loss[loss=2.44, over 1260.00 frames. , ppl: 11.478160665782687] tot_loss[loss=2.3, over 5469307.31 frames. , ppl: 9.977167145377798], batch size: 70
+2022-12-11 18:07:31,753 INFO [train.py:421] (2/8) Epoch 5, batch 38400, loss[loss=2.342, over 2800.00 frames. , ppl: 10.406908153827194] tot_loss[loss=2.3, over 5447984.54 frames. , ppl: 9.973607433506006], batch size: 70
+2022-12-11 18:09:11,040 INFO [train.py:421] (2/8) Epoch 5, batch 38600, loss[loss=2.366, over 1960.00 frames. , ppl: 10.652763222433538] tot_loss[loss=2.3, over 5435130.02 frames. , ppl: 9.975939693768158], batch size: 70
+2022-12-11 18:10:51,066 INFO [train.py:421] (2/8) Epoch 5, batch 38800, loss[loss=2.349, over 1400.00 frames. , ppl: 10.478954332529174] tot_loss[loss=2.3, over 5450496.47 frames. , ppl: 9.969593370353966], batch size: 70
+2022-12-11 18:12:31,801 INFO [train.py:421] (2/8) Epoch 5, batch 39000, loss[loss=2.164, over 5670.00 frames. , ppl: 8.702344529229293] tot_loss[loss=2.299, over 5467259.72 frames. , ppl: 9.963785553007693], batch size: 70
+2022-12-11 18:12:31,802 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:12:32,547 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.839695691985794
+2022-12-11 18:14:11,127 INFO [train.py:421] (2/8) Epoch 5, batch 39200, loss[loss=2.489, over 1470.00 frames. , ppl: 12.04332353907669] tot_loss[loss=2.3, over 5445846.35 frames. , ppl: 9.974231773550649], batch size: 70
+2022-12-11 18:15:49,010 INFO [train.py:421] (2/8) Epoch 5, batch 39400, loss[loss=2.533, over 980.00 frames. , ppl: 12.596159429348889] tot_loss[loss=2.301, over 5392249.36 frames. , ppl: 9.986484844088102], batch size: 70
+2022-12-11 18:17:29,939 INFO [train.py:421] (2/8) Epoch 5, batch 39600, loss[loss=2.256, over 1960.00 frames. , ppl: 9.544323387548266] tot_loss[loss=2.3, over 5465195.78 frames. , ppl: 9.969792235234605], batch size: 70
+2022-12-11 18:19:09,419 INFO [train.py:421] (2/8) Epoch 5, batch 39800, loss[loss=2.197, over 3570.00 frames. , ppl: 8.99394215863086] tot_loss[loss=2.3, over 5453391.58 frames. , ppl: 9.975032324609872], batch size: 70
+2022-12-11 18:20:46,152 INFO [train.py:421] (2/8) Epoch 5, batch 40000, loss[loss=2.319, over 3640.00 frames. , ppl: 10.16116327305099] tot_loss[loss=2.301, over 5446363.02 frames. , ppl: 9.980194505264327], batch size: 70
+2022-12-11 18:20:46,153 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:20:46,898 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.863040936604593
+2022-12-11 18:22:26,580 INFO [train.py:421] (2/8) Epoch 5, batch 40200, loss[loss=2.395, over 1260.00 frames. , ppl: 10.97235323458615] tot_loss[loss=2.3, over 5463690.01 frames. , ppl: 9.974856987436002], batch size: 70
+2022-12-11 18:24:04,874 INFO [train.py:421] (2/8) Epoch 5, batch 40400, loss[loss=2.826, over 700.00 frames. , ppl: 16.87382260234604] tot_loss[loss=2.301, over 5451519.95 frames. , ppl: 9.980962850667664], batch size: 70
+2022-12-11 18:25:44,588 INFO [train.py:421] (2/8) Epoch 5, batch 40600, loss[loss=2.214, over 5530.00 frames. , ppl: 9.15487405126113] tot_loss[loss=2.3, over 5463882.29 frames. , ppl: 9.976734096168151], batch size: 70
+2022-12-11 18:27:25,884 INFO [train.py:421] (2/8) Epoch 5, batch 40800, loss[loss=2.216, over 1890.00 frames. , ppl: 9.171237275586972] tot_loss[loss=2.299, over 5518055.03 frames. , ppl: 9.96496626795192], batch size: 70
+2022-12-11 18:29:07,897 INFO [train.py:421] (2/8) Epoch 5, batch 41000, loss[loss=2.18, over 4480.00 frames. , ppl: 8.846657995245138] tot_loss[loss=2.299, over 5474895.90 frames. , ppl: 9.967010614529732], batch size: 70
+2022-12-11 18:29:07,897 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:29:08,657 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.844880201345818
+2022-12-11 18:30:49,865 INFO [train.py:421] (2/8) Epoch 5, batch 41200, loss[loss=2.342, over 1820.00 frames. , ppl: 10.403738703099824] tot_loss[loss=2.3, over 5440649.24 frames. , ppl: 9.972550787107282], batch size: 70
+2022-12-11 18:32:28,300 INFO [train.py:421] (2/8) Epoch 5, batch 41400, loss[loss=2.38, over 1750.00 frames. , ppl: 10.80720638672771] tot_loss[loss=2.3, over 5448435.58 frames. , ppl: 9.972122443946398], batch size: 70
+2022-12-11 18:34:08,974 INFO [train.py:421] (2/8) Epoch 5, batch 41600, loss[loss=2.287, over 1470.00 frames. , ppl: 9.844985372365786] tot_loss[loss=2.3, over 5444075.89 frames. , ppl: 9.971731411857732], batch size: 70
+2022-12-11 18:35:50,122 INFO [train.py:421] (2/8) Epoch 5, batch 41800, loss[loss=2.338, over 2520.00 frames. , ppl: 10.355356015166786] tot_loss[loss=2.3, over 5435572.56 frames. , ppl: 9.97351570590567], batch size: 70
+2022-12-11 18:37:33,725 INFO [train.py:421] (2/8) Epoch 5, batch 42000, loss[loss=2.348, over 1330.00 frames. , ppl: 10.465110082981804] tot_loss[loss=2.299, over 5468464.13 frames. , ppl: 9.965263443873804], batch size: 70
+2022-12-11 18:37:33,725 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:37:34,472 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.285, over 211138.00 frames. , ppl: 9.829064270772456
+2022-12-11 18:39:16,852 INFO [train.py:421] (2/8) Epoch 5, batch 42200, loss[loss=2.218, over 5810.00 frames. , ppl: 9.186936359071048] tot_loss[loss=2.298, over 5471272.89 frames. , ppl: 9.957982703777297], batch size: 70
+2022-12-11 18:40:58,489 INFO [train.py:421] (2/8) Epoch 5, batch 42400, loss[loss=2.274, over 4410.00 frames. , ppl: 9.713827984804237] tot_loss[loss=2.298, over 5479810.81 frames. , ppl: 9.95689377374292], batch size: 70
+2022-12-11 18:42:36,705 INFO [train.py:421] (2/8) Epoch 5, batch 42600, loss[loss=2.295, over 2590.00 frames. , ppl: 9.924536298930562] tot_loss[loss=2.298, over 5483065.05 frames. , ppl: 9.953324520615542], batch size: 70
+2022-12-11 18:44:18,977 INFO [train.py:421] (2/8) Epoch 5, batch 42800, loss[loss=3.148, over 560.00 frames. , ppl: 23.281272326118053] tot_loss[loss=2.298, over 5485450.17 frames. , ppl: 9.95778142015021], batch size: 70
+2022-12-11 18:45:56,834 INFO [train.py:421] (2/8) Epoch 5, batch 43000, loss[loss=2.411, over 3290.00 frames. , ppl: 11.148400867096804] tot_loss[loss=2.3, over 5427508.26 frames. , ppl: 9.973723406821419], batch size: 70
+2022-12-11 18:45:56,835 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:45:57,597 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.842544721854633
+2022-12-11 18:47:37,836 INFO [train.py:421] (2/8) Epoch 5, batch 43200, loss[loss=2.286, over 2310.00 frames. , ppl: 9.838197691014036] tot_loss[loss=2.3, over 5431837.79 frames. , ppl: 9.975822212986074], batch size: 70
+2022-12-11 18:49:19,142 INFO [train.py:421] (2/8) Epoch 5, batch 43400, loss[loss=2.264, over 4900.00 frames. , ppl: 9.625110865460746] tot_loss[loss=2.301, over 5415539.10 frames. , ppl: 9.984394099492974], batch size: 70
+2022-12-11 18:51:01,641 INFO [train.py:421] (2/8) Epoch 5, batch 43600, loss[loss=2.415, over 1260.00 frames. , ppl: 11.187644898544] tot_loss[loss=2.302, over 5382000.13 frames. , ppl: 9.989415168331227], batch size: 70
+2022-12-11 18:52:42,485 INFO [train.py:421] (2/8) Epoch 5, batch 43800, loss[loss=2.291, over 2100.00 frames. , ppl: 9.887367235616786] tot_loss[loss=2.302, over 5373434.47 frames. , ppl: 9.99234340453092], batch size: 70
+2022-12-11 18:54:22,169 INFO [train.py:421] (2/8) Epoch 5, batch 44000, loss[loss=2.2, over 5950.00 frames. , ppl: 9.027013420344213] tot_loss[loss=2.301, over 5382699.43 frames. , ppl: 9.98724282205154], batch size: 70
+2022-12-11 18:54:22,169 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 18:54:22,915 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.849284590078504
+2022-12-11 18:56:01,438 INFO [train.py:421] (2/8) Epoch 5, batch 44200, loss[loss=2.176, over 10850.00 frames. , ppl: 8.814175230231825] tot_loss[loss=2.3, over 5420454.14 frames. , ppl: 9.974935151606115], batch size: 70
+2022-12-11 18:57:41,930 INFO [train.py:421] (2/8) Epoch 5, batch 44400, loss[loss=2.295, over 2310.00 frames. , ppl: 9.926306166625126] tot_loss[loss=2.299, over 5440572.03 frames. , ppl: 9.964387975951158], batch size: 70
+2022-12-11 18:59:21,017 INFO [train.py:421] (2/8) Epoch 5, batch 44600, loss[loss=2.231, over 3990.00 frames. , ppl: 9.312644594493861] tot_loss[loss=2.299, over 5429056.80 frames. , ppl: 9.96534995885989], batch size: 70
+2022-12-11 19:01:02,197 INFO [train.py:421] (2/8) Epoch 5, batch 44800, loss[loss=2.535, over 770.00 frames. , ppl: 12.612862340396786] tot_loss[loss=2.298, over 5460493.16 frames. , ppl: 9.954346125592584], batch size: 70
+2022-12-11 19:02:40,877 INFO [train.py:421] (2/8) Epoch 5, batch 45000, loss[loss=2.23, over 5530.00 frames. , ppl: 9.304413725066016] tot_loss[loss=2.298, over 5476049.04 frames. , ppl: 9.951554599328412], batch size: 70
+2022-12-11 19:02:40,877 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:02:41,620 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.835643508977666
+2022-12-11 19:04:22,576 INFO [train.py:421] (2/8) Epoch 5, batch 45200, loss[loss=2.338, over 2100.00 frames. , ppl: 10.359217199752983] tot_loss[loss=2.298, over 5467806.55 frames. , ppl: 9.954484251564109], batch size: 70
+2022-12-11 19:06:01,351 INFO [train.py:421] (2/8) Epoch 5, batch 45400, loss[loss=2.34, over 3500.00 frames. , ppl: 10.37727194250331] tot_loss[loss=2.299, over 5457515.90 frames. , ppl: 9.962356456512898], batch size: 70
+2022-12-11 19:07:43,218 INFO [train.py:421] (2/8) Epoch 5, batch 45600, loss[loss=2.26, over 2030.00 frames. , ppl: 9.587468832056121] tot_loss[loss=2.298, over 5492102.43 frames. , ppl: 9.95839400093763], batch size: 70
+2022-12-11 19:09:21,551 INFO [train.py:421] (2/8) Epoch 5, batch 45800, loss[loss=2.404, over 1330.00 frames. , ppl: 11.07135884775085] tot_loss[loss=2.3, over 5450821.97 frames. , ppl: 9.970025587364258], batch size: 70
+2022-12-11 19:11:03,159 INFO [train.py:421] (2/8) Epoch 5, batch 46000, loss[loss=2.766, over 630.00 frames. , ppl: 15.896130126694448] tot_loss[loss=2.298, over 5478379.80 frames. , ppl: 9.958171540858121], batch size: 70
+2022-12-11 19:11:03,159 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:11:03,906 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.285, over 211138.00 frames. , ppl: 9.824885591047718
+2022-12-11 19:12:44,281 INFO [train.py:421] (2/8) Epoch 5, batch 46200, loss[loss=2.263, over 2590.00 frames. , ppl: 9.613738878487894] tot_loss[loss=2.298, over 5470590.97 frames. , ppl: 9.956617380214531], batch size: 70
+2022-12-11 19:14:23,630 INFO [train.py:421] (2/8) Epoch 5, batch 46400, loss[loss=2.42, over 1820.00 frames. , ppl: 11.242549427737295] tot_loss[loss=2.299, over 5474312.70 frames. , ppl: 9.960301065804067], batch size: 70
+2022-12-11 19:15:58,261 INFO [train.py:421] (2/8) Epoch 5, batch 46600, loss[loss=2.388, over 1330.00 frames. , ppl: 10.887227505574858] tot_loss[loss=2.3, over 5454562.87 frames. , ppl: 9.972102677639038], batch size: 70
+2022-12-11 19:17:38,452 INFO [train.py:421] (2/8) Epoch 5, batch 46800, loss[loss=2.493, over 980.00 frames. , ppl: 12.100264837550418] tot_loss[loss=2.3, over 5457135.60 frames. , ppl: 9.969426165977737], batch size: 70
+2022-12-11 19:19:20,571 INFO [train.py:421] (2/8) Epoch 5, batch 47000, loss[loss=2.519, over 770.00 frames. , ppl: 12.411541675146234] tot_loss[loss=2.298, over 5501450.44 frames. , ppl: 9.955522738356043], batch size: 70
+2022-12-11 19:19:20,572 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:19:21,331 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.837504131725721
+2022-12-11 19:21:03,345 INFO [train.py:421] (2/8) Epoch 5, batch 47200, loss[loss=2.238, over 6300.00 frames. , ppl: 9.370475706707612] tot_loss[loss=2.298, over 5515866.64 frames. , ppl: 9.954278715407337], batch size: 70
+2022-12-11 19:22:43,309 INFO [train.py:421] (2/8) Epoch 5, batch 47400, loss[loss=2.332, over 2520.00 frames. , ppl: 10.293768436466255] tot_loss[loss=2.299, over 5465890.78 frames. , ppl: 9.96440932522435], batch size: 70
+2022-12-11 19:24:25,849 INFO [train.py:421] (2/8) Epoch 5, batch 47600, loss[loss=2.284, over 3500.00 frames. , ppl: 9.816421441498933] tot_loss[loss=2.298, over 5514868.53 frames. , ppl: 9.949789612226233], batch size: 70
+2022-12-11 19:26:04,799 INFO [train.py:421] (2/8) Epoch 5, batch 47800, loss[loss=2.658, over 840.00 frames. , ppl: 14.264154628607592] tot_loss[loss=2.297, over 5511003.92 frames. , ppl: 9.949269063607392], batch size: 70
+2022-12-11 19:27:44,266 INFO [train.py:421] (2/8) Epoch 5, batch 48000, loss[loss=2.363, over 1330.00 frames. , ppl: 10.625219810747339] tot_loss[loss=2.297, over 5501351.65 frames. , ppl: 9.94774727413184], batch size: 70
+2022-12-11 19:27:44,267 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:27:45,017 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.847490241584369
+2022-12-11 19:29:31,062 INFO [train.py:421] (2/8) Epoch 5, batch 48200, loss[loss=2.332, over 1960.00 frames. , ppl: 10.296220594519545] tot_loss[loss=2.296, over 5533247.75 frames. , ppl: 9.937660371649324], batch size: 70
+2022-12-11 19:31:11,712 INFO [train.py:421] (2/8) Epoch 5, batch 48400, loss[loss=2.403, over 1540.00 frames. , ppl: 11.057775855784225] tot_loss[loss=2.296, over 5539185.87 frames. , ppl: 9.935175849467095], batch size: 70
+2022-12-11 19:32:50,749 INFO [train.py:421] (2/8) Epoch 5, batch 48600, loss[loss=2.218, over 3850.00 frames. , ppl: 9.192920667780678] tot_loss[loss=2.296, over 5534376.99 frames. , ppl: 9.93595030854555], batch size: 70
+2022-12-11 19:34:30,308 INFO [train.py:421] (2/8) Epoch 5, batch 48800, loss[loss=2.295, over 1540.00 frames. , ppl: 9.929253518214315] tot_loss[loss=2.295, over 5547200.46 frames. , ppl: 9.927876844180556], batch size: 70
+2022-12-11 19:36:12,848 INFO [train.py:421] (2/8) Epoch 5, batch 49000, loss[loss=2.172, over 6090.00 frames. , ppl: 8.774726856651705] tot_loss[loss=2.294, over 5567658.60 frames. , ppl: 9.916295305177139], batch size: 70
+2022-12-11 19:36:12,849 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:36:13,593 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.8508430669708
+2022-12-11 19:37:58,713 INFO [train.py:421] (2/8) Epoch 5, batch 49200, loss[loss=2.283, over 2590.00 frames. , ppl: 9.808678948502394] tot_loss[loss=2.294, over 5596506.48 frames. , ppl: 9.911815312693355], batch size: 70
+2022-12-11 19:39:40,032 INFO [train.py:421] (2/8) Epoch 5, batch 49400, loss[loss=2.182, over 4480.00 frames. , ppl: 8.861760746083748] tot_loss[loss=2.294, over 5609655.14 frames. , ppl: 9.91462256506342], batch size: 70
+2022-12-11 19:41:17,702 INFO [train.py:421] (2/8) Epoch 5, batch 49600, loss[loss=2.274, over 4760.00 frames. , ppl: 9.717159720446466] tot_loss[loss=2.294, over 5593398.13 frames. , ppl: 9.917441615698037], batch size: 70
+2022-12-11 19:42:57,244 INFO [train.py:421] (2/8) Epoch 5, batch 49800, loss[loss=2.239, over 5740.00 frames. , ppl: 9.385452055768932] tot_loss[loss=2.294, over 5606487.91 frames. , ppl: 9.91346315491181], batch size: 70
+2022-12-11 19:44:39,917 INFO [train.py:421] (2/8) Epoch 5, batch 50000, loss[loss=2.495, over 1190.00 frames. , ppl: 12.127527383624676] tot_loss[loss=2.294, over 5607794.41 frames. , ppl: 9.911195677915465], batch size: 70
+2022-12-11 19:44:39,918 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:44:40,663 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.837191091537502
+2022-12-11 19:46:20,685 INFO [train.py:421] (2/8) Epoch 5, batch 50200, loss[loss=3.038, over 560.00 frames. , ppl: 20.85621847387642] tot_loss[loss=2.294, over 5571209.63 frames. , ppl: 9.913812683752704], batch size: 70
+2022-12-11 19:48:00,103 INFO [train.py:421] (2/8) Epoch 5, batch 50400, loss[loss=2.203, over 2730.00 frames. , ppl: 9.051505689113426] tot_loss[loss=2.295, over 5548313.81 frames. , ppl: 9.922185105446463], batch size: 70
+2022-12-11 19:49:38,962 INFO [train.py:421] (2/8) Epoch 5, batch 50600, loss[loss=4.847, over 280.00 frames. , ppl: 127.31035017219152] tot_loss[loss=2.294, over 5550962.91 frames. , ppl: 9.918298605833256], batch size: 70
+2022-12-11 19:51:17,597 INFO [train.py:421] (2/8) Epoch 5, batch 50800, loss[loss=2.255, over 3080.00 frames. , ppl: 9.530878478759604] tot_loss[loss=2.295, over 5544658.68 frames. , ppl: 9.92178272416548], batch size: 70
+2022-12-11 19:52:59,276 INFO [train.py:421] (2/8) Epoch 5, batch 51000, loss[loss=2.573, over 980.00 frames. , ppl: 13.110201464020312] tot_loss[loss=2.295, over 5537796.75 frames. , ppl: 9.920463675405523], batch size: 70
+2022-12-11 19:52:59,276 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 19:53:00,024 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.852051821719042
+2022-12-11 19:54:41,942 INFO [train.py:421] (2/8) Epoch 5, batch 51200, loss[loss=2.426, over 1750.00 frames. , ppl: 11.311940134905909] tot_loss[loss=2.293, over 5571120.86 frames. , ppl: 9.907922344237829], batch size: 70
+2022-12-11 19:56:21,646 INFO [train.py:421] (2/8) Epoch 5, batch 51400, loss[loss=2.454, over 1330.00 frames. , ppl: 11.636175436348937] tot_loss[loss=2.294, over 5550344.78 frames. , ppl: 9.916497906281808], batch size: 70
+2022-12-11 19:58:02,211 INFO [train.py:421] (2/8) Epoch 5, batch 51600, loss[loss=2.596, over 1190.00 frames. , ppl: 13.408639769952] tot_loss[loss=2.297, over 5473239.97 frames. , ppl: 9.94014092817279], batch size: 70
+2022-12-11 19:59:40,708 INFO [train.py:421] (2/8) Epoch 5, batch 51800, loss[loss=2.682, over 630.00 frames. , ppl: 14.618816390578017] tot_loss[loss=2.296, over 5493875.89 frames. , ppl: 9.932956809835837], batch size: 70
+2022-12-11 20:01:19,087 INFO [train.py:421] (2/8) Epoch 5, batch 52000, loss[loss=2.16, over 6440.00 frames. , ppl: 8.672021837274084] tot_loss[loss=2.296, over 5465032.43 frames. , ppl: 9.937318981447797], batch size: 70
+2022-12-11 20:01:19,088 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:01:19,832 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.840256402098536
+2022-12-11 20:03:00,929 INFO [train.py:421] (2/8) Epoch 5, batch 52200, loss[loss=2.222, over 3010.00 frames. , ppl: 9.221474856699798] tot_loss[loss=2.297, over 5469343.84 frames. , ppl: 9.939759200072508], batch size: 70
+2022-12-11 20:04:40,923 INFO [train.py:421] (2/8) Epoch 5, batch 52400, loss[loss=2.296, over 3710.00 frames. , ppl: 9.93424554279179] tot_loss[loss=2.297, over 5473174.41 frames. , ppl: 9.9409088418711], batch size: 70
+2022-12-11 20:06:20,201 INFO [train.py:421] (2/8) Epoch 5, batch 52600, loss[loss=2.26, over 3920.00 frames. , ppl: 9.578375750470315] tot_loss[loss=2.297, over 5450520.45 frames. , ppl: 9.94738020617009], batch size: 70
+2022-12-11 20:07:55,988 INFO [train.py:421] (2/8) Epoch 5, batch 52800, loss[loss=2.323, over 2520.00 frames. , ppl: 10.21006446365191] tot_loss[loss=2.298, over 5416004.53 frames. , ppl: 9.95598882031081], batch size: 70
+2022-12-11 20:09:33,171 INFO [train.py:421] (2/8) Epoch 5, batch 53000, loss[loss=2.295, over 3290.00 frames. , ppl: 9.926860150060982] tot_loss[loss=2.296, over 5463242.02 frames. , ppl: 9.936569953698122], batch size: 70
+2022-12-11 20:09:33,171 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:09:33,930 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.842675832035669
+2022-12-11 20:11:15,343 INFO [train.py:421] (2/8) Epoch 5, batch 53200, loss[loss=2.317, over 2660.00 frames. , ppl: 10.142873691603585] tot_loss[loss=2.297, over 5455033.02 frames. , ppl: 9.945531655130193], batch size: 70
+2022-12-11 20:12:50,608 INFO [train.py:421] (2/8) Epoch 5, batch 53400, loss[loss=2.398, over 1190.00 frames. , ppl: 10.996479891551285] tot_loss[loss=2.298, over 5421289.48 frames. , ppl: 9.95680036138342], batch size: 70
+2022-12-11 20:14:31,569 INFO [train.py:421] (2/8) Epoch 5, batch 53600, loss[loss=2.741, over 630.00 frames. , ppl: 15.505137334871607] tot_loss[loss=2.298, over 5424621.10 frames. , ppl: 9.951526762038174], batch size: 70
+2022-12-11 20:16:11,456 INFO [train.py:421] (2/8) Epoch 5, batch 53800, loss[loss=2.22, over 4900.00 frames. , ppl: 9.204088979469372] tot_loss[loss=2.298, over 5438444.26 frames. , ppl: 9.951994448084193], batch size: 70
+2022-12-11 20:17:55,393 INFO [train.py:421] (2/8) Epoch 5, batch 54000, loss[loss=2.358, over 1610.00 frames. , ppl: 10.571296691515185] tot_loss[loss=2.297, over 5452337.49 frames. , ppl: 9.947485243553347], batch size: 70
+2022-12-11 20:17:55,394 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:17:56,155 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.832751361696745
+2022-12-11 20:19:35,176 INFO [train.py:421] (2/8) Epoch 5, batch 54200, loss[loss=2.231, over 4550.00 frames. , ppl: 9.309395976329329] tot_loss[loss=2.297, over 5450690.07 frames. , ppl: 9.94508914121644], batch size: 70
+2022-12-11 20:21:11,752 INFO [train.py:421] (2/8) Epoch 5, batch 54400, loss[loss=2.434, over 1050.00 frames. , ppl: 11.402114585152747] tot_loss[loss=2.298, over 5409349.04 frames. , ppl: 9.957483146385432], batch size: 70
+2022-12-11 20:22:53,112 INFO [train.py:421] (2/8) Epoch 5, batch 54600, loss[loss=2.381, over 2870.00 frames. , ppl: 10.81847081396585] tot_loss[loss=2.298, over 5450657.79 frames. , ppl: 9.950920551685599], batch size: 70
+2022-12-11 20:24:36,518 INFO [train.py:421] (2/8) Epoch 5, batch 54800, loss[loss=2.362, over 2730.00 frames. , ppl: 10.608280397987052] tot_loss[loss=2.297, over 5486163.53 frames. , ppl: 9.94546569319174], batch size: 70
+2022-12-11 20:26:21,050 INFO [train.py:421] (2/8) Epoch 5, batch 55000, loss[loss=2.688, over 910.00 frames. , ppl: 14.703900020885657] tot_loss[loss=2.296, over 5543163.11 frames. , ppl: 9.932785974391559], batch size: 70
+2022-12-11 20:26:21,051 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:26:21,809 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.833813803126963
+2022-12-11 20:27:59,953 INFO [train.py:421] (2/8) Epoch 5, batch 55200, loss[loss=2.551, over 980.00 frames. , ppl: 12.81545935453945] tot_loss[loss=2.298, over 5488297.26 frames. , ppl: 9.9495094314003], batch size: 70
+2022-12-11 20:29:36,219 INFO [train.py:421] (2/8) Epoch 5, batch 55400, loss[loss=2.789, over 630.00 frames. , ppl: 16.261143771890655] tot_loss[loss=2.298, over 5470646.84 frames. , ppl: 9.952550315947494], batch size: 70
+2022-12-11 20:31:18,017 INFO [train.py:421] (2/8) Epoch 5, batch 55600, loss[loss=2.217, over 5670.00 frames. , ppl: 9.175492893561954] tot_loss[loss=2.298, over 5491540.29 frames. , ppl: 9.949982596540332], batch size: 70
+2022-12-11 20:32:57,832 INFO [train.py:421] (2/8) Epoch 5, batch 55800, loss[loss=2.175, over 4060.00 frames. , ppl: 8.804427536113526] tot_loss[loss=2.296, over 5533904.58 frames. , ppl: 9.931473019911888], batch size: 70
+2022-12-11 20:34:40,844 INFO [train.py:421] (2/8) Epoch 5, batch 56000, loss[loss=2.242, over 5600.00 frames. , ppl: 9.41651392446222] tot_loss[loss=2.295, over 5548611.04 frames. , ppl: 9.91997152886551], batch size: 70
+2022-12-11 20:34:40,845 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:34:41,592 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.835016101318848
+2022-12-11 20:36:19,908 INFO [train.py:421] (2/8) Epoch 5, batch 56200, loss[loss=2.391, over 1960.00 frames. , ppl: 10.921481446138895] tot_loss[loss=2.296, over 5504425.14 frames. , ppl: 9.934942167430922], batch size: 70
+2022-12-11 20:38:00,524 INFO [train.py:421] (2/8) Epoch 5, batch 56400, loss[loss=2.259, over 2800.00 frames. , ppl: 9.575231262341518] tot_loss[loss=2.296, over 5525750.73 frames. , ppl: 9.932627534209375], batch size: 70
+2022-12-11 20:39:43,937 INFO [train.py:421] (2/8) Epoch 5, batch 56600, loss[loss=2.446, over 1120.00 frames. , ppl: 11.54552848408935] tot_loss[loss=2.298, over 5465225.95 frames. , ppl: 9.950585343981762], batch size: 70
+2022-12-11 20:41:26,632 INFO [train.py:421] (2/8) Epoch 5, batch 56800, loss[loss=2.284, over 3850.00 frames. , ppl: 9.81811750030068] tot_loss[loss=2.297, over 5479921.90 frames. , ppl: 9.942709231052167], batch size: 70
+2022-12-11 20:43:05,319 INFO [train.py:421] (2/8) Epoch 5, batch 57000, loss[loss=2.161, over 11550.00 frames. , ppl: 8.676362208031925] tot_loss[loss=2.297, over 5460913.10 frames. , ppl: 9.947740912214911], batch size: 70
+2022-12-11 20:43:05,320 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:43:06,065 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.842181992787078
+2022-12-11 20:44:43,547 INFO [train.py:421] (2/8) Epoch 5, batch 57200, loss[loss=2.272, over 7630.00 frames. , ppl: 9.7004268393476] tot_loss[loss=2.298, over 5468853.17 frames. , ppl: 9.950314684780334], batch size: 70
+2022-12-11 20:46:22,290 INFO [train.py:421] (2/8) Epoch 5, batch 57400, loss[loss=2.197, over 6860.00 frames. , ppl: 8.996434275457425] tot_loss[loss=2.298, over 5439494.84 frames. , ppl: 9.957143757232926], batch size: 70
+2022-12-11 20:48:03,932 INFO [train.py:421] (2/8) Epoch 5, batch 57600, loss[loss=2.71, over 840.00 frames. , ppl: 15.023050425243479] tot_loss[loss=2.296, over 5482697.67 frames. , ppl: 9.93927850460295], batch size: 70
+2022-12-11 20:49:40,763 INFO [train.py:421] (2/8) Epoch 5, batch 57800, loss[loss=2.306, over 2730.00 frames. , ppl: 10.036701948760127] tot_loss[loss=2.298, over 5409244.08 frames. , ppl: 9.958500489350708], batch size: 70
+2022-12-11 20:51:21,984 INFO [train.py:421] (2/8) Epoch 5, batch 58000, loss[loss=2.381, over 2310.00 frames. , ppl: 10.820699870965615] tot_loss[loss=2.299, over 5417167.29 frames. , ppl: 9.960978333053214], batch size: 70
+2022-12-11 20:51:21,985 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:51:22,729 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.846384059654563
+2022-12-11 20:53:06,173 INFO [train.py:421] (2/8) Epoch 5, batch 58200, loss[loss=2.341, over 3150.00 frames. , ppl: 10.389221237336061] tot_loss[loss=2.298, over 5467529.64 frames. , ppl: 9.949513547011442], batch size: 70
+2022-12-11 20:54:48,411 INFO [train.py:421] (2/8) Epoch 5, batch 58400, loss[loss=2.432, over 1190.00 frames. , ppl: 11.382739535791638] tot_loss[loss=2.299, over 5435219.96 frames. , ppl: 9.96060219164402], batch size: 70
+2022-12-11 20:56:30,697 INFO [train.py:421] (2/8) Epoch 5, batch 58600, loss[loss=2.606, over 840.00 frames. , ppl: 13.540111546801006] tot_loss[loss=2.298, over 5466808.54 frames. , ppl: 9.954577361285926], batch size: 70
+2022-12-11 20:58:11,550 INFO [train.py:421] (2/8) Epoch 5, batch 58800, loss[loss=2.332, over 2030.00 frames. , ppl: 10.301213257251005] tot_loss[loss=2.298, over 5475114.61 frames. , ppl: 9.954510069363524], batch size: 70
+2022-12-11 20:59:49,673 INFO [train.py:421] (2/8) Epoch 5, batch 59000, loss[loss=2.581, over 770.00 frames. , ppl: 13.206124621375169] tot_loss[loss=2.297, over 5507807.71 frames. , ppl: 9.946313809034017], batch size: 70
+2022-12-11 20:59:49,673 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 20:59:50,433 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.833901132185455
+2022-12-11 21:01:27,627 INFO [train.py:421] (2/8) Epoch 5, batch 59200, loss[loss=2.431, over 1190.00 frames. , ppl: 11.371181076915756] tot_loss[loss=2.296, over 5546163.33 frames. , ppl: 9.933809265174808], batch size: 70
+2022-12-11 21:03:06,545 INFO [train.py:421] (2/8) Epoch 5, batch 59400, loss[loss=2.38, over 3080.00 frames. , ppl: 10.808842983857424] tot_loss[loss=2.297, over 5525597.01 frames. , ppl: 9.943094117765394], batch size: 70
+2022-12-11 21:04:48,214 INFO [train.py:421] (2/8) Epoch 5, batch 59600, loss[loss=2.263, over 2660.00 frames. , ppl: 9.608681796415208] tot_loss[loss=2.297, over 5530313.16 frames. , ppl: 9.941421665501506], batch size: 70
+2022-12-11 21:06:24,260 INFO [train.py:421] (2/8) Epoch 5, batch 59800, loss[loss=2.364, over 1330.00 frames. , ppl: 10.629033550254324] tot_loss[loss=2.296, over 5525090.58 frames. , ppl: 9.935515534832879], batch size: 70
+2022-12-11 21:08:10,094 INFO [train.py:421] (2/8) Epoch 5, batch 60000, loss[loss=2.277, over 2870.00 frames. , ppl: 9.748851257228624] tot_loss[loss=2.297, over 5497930.08 frames. , ppl: 9.943540173920388], batch size: 70
+2022-12-11 21:08:10,094 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:08:10,853 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.831021137755421
+2022-12-11 21:09:53,985 INFO [train.py:421] (2/8) Epoch 5, batch 60200, loss[loss=2.247, over 10080.00 frames. , ppl: 9.458585611062523] tot_loss[loss=2.295, over 5573162.92 frames. , ppl: 9.925967201051371], batch size: 70
+2022-12-11 21:11:35,809 INFO [train.py:421] (2/8) Epoch 5, batch 60400, loss[loss=2.605, over 1050.00 frames. , ppl: 13.535769194545326] tot_loss[loss=2.295, over 5586497.18 frames. , ppl: 9.922658673001065], batch size: 70
+2022-12-11 21:13:18,754 INFO [train.py:421] (2/8) Epoch 5, batch 60600, loss[loss=2.27, over 6230.00 frames. , ppl: 9.676830307371757] tot_loss[loss=2.295, over 5567942.58 frames. , ppl: 9.927592384259949], batch size: 70
+2022-12-11 21:14:58,852 INFO [train.py:421] (2/8) Epoch 5, batch 60800, loss[loss=2.173, over 3640.00 frames. , ppl: 8.787904880768673] tot_loss[loss=2.295, over 5573305.71 frames. , ppl: 9.928153118369465], batch size: 70
+2022-12-11 21:16:39,557 INFO [train.py:421] (2/8) Epoch 5, batch 61000, loss[loss=2.806, over 700.00 frames. , ppl: 16.537235524315104] tot_loss[loss=2.296, over 5561810.53 frames. , ppl: 9.934194070044482], batch size: 70
+2022-12-11 21:16:39,558 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:16:40,311 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816234263345795
+2022-12-11 21:18:20,759 INFO [train.py:421] (2/8) Epoch 5, batch 61200, loss[loss=2.328, over 1960.00 frames. , ppl: 10.256393707600804] tot_loss[loss=2.297, over 5520707.16 frames. , ppl: 9.946058886808116], batch size: 70
+2022-12-11 21:20:01,590 INFO [train.py:421] (2/8) Epoch 5, batch 61400, loss[loss=2.266, over 5670.00 frames. , ppl: 9.640492747694708] tot_loss[loss=2.297, over 5517340.58 frames. , ppl: 9.947580279541326], batch size: 70
+2022-12-11 21:21:39,915 INFO [train.py:421] (2/8) Epoch 5, batch 61600, loss[loss=2.481, over 910.00 frames. , ppl: 11.948517185500151] tot_loss[loss=2.298, over 5446663.77 frames. , ppl: 9.959015586624792], batch size: 70
+2022-12-11 21:23:20,559 INFO [train.py:421] (2/8) Epoch 5, batch 61800, loss[loss=2.199, over 8050.00 frames. , ppl: 9.018168945280605] tot_loss[loss=2.298, over 5469432.20 frames. , ppl: 9.953926682120064], batch size: 70
+2022-12-11 21:24:59,391 INFO [train.py:421] (2/8) Epoch 5, batch 62000, loss[loss=2.278, over 1330.00 frames. , ppl: 9.7592960315926] tot_loss[loss=2.298, over 5436160.77 frames. , ppl: 9.95908138801635], batch size: 70
+2022-12-11 21:24:59,391 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:25:00,150 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.841092429233514
+2022-12-11 21:26:42,941 INFO [train.py:421] (2/8) Epoch 5, batch 62200, loss[loss=2.273, over 3640.00 frames. , ppl: 9.708575656090451] tot_loss[loss=2.3, over 5400862.60 frames. , ppl: 9.973728674036371], batch size: 70
+2022-12-11 21:28:21,541 INFO [train.py:421] (2/8) Epoch 5, batch 62400, loss[loss=2.257, over 4130.00 frames. , ppl: 9.555686982793311] tot_loss[loss=2.299, over 5444859.15 frames. , ppl: 9.959718911089055], batch size: 70
+2022-12-11 21:30:03,693 INFO [train.py:421] (2/8) Epoch 5, batch 62600, loss[loss=2.192, over 5180.00 frames. , ppl: 8.954955862235142] tot_loss[loss=2.297, over 5479269.32 frames. , ppl: 9.945555696647682], batch size: 70
+2022-12-11 21:31:48,069 INFO [train.py:421] (2/8) Epoch 5, batch 62800, loss[loss=2.509, over 910.00 frames. , ppl: 12.293474792423316] tot_loss[loss=2.297, over 5459617.03 frames. , ppl: 9.9468724669891], batch size: 70
+2022-12-11 21:33:23,566 INFO [train.py:421] (2/8) Epoch 5, batch 63000, loss[loss=2.24, over 3430.00 frames. , ppl: 9.390774111575817] tot_loss[loss=2.297, over 5456058.97 frames. , ppl: 9.94077095888672], batch size: 70
+2022-12-11 21:33:23,567 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:33:24,313 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.831210297961553
+2022-12-11 21:35:04,616 INFO [train.py:421] (2/8) Epoch 5, batch 63200, loss[loss=2.461, over 1820.00 frames. , ppl: 11.711893042331187] tot_loss[loss=2.295, over 5506831.55 frames. , ppl: 9.928447304475512], batch size: 70
+2022-12-11 21:36:45,647 INFO [train.py:421] (2/8) Epoch 5, batch 63400, loss[loss=2.355, over 1890.00 frames. , ppl: 10.540479751271702] tot_loss[loss=2.294, over 5559141.54 frames. , ppl: 9.919188093190998], batch size: 70
+2022-12-11 21:38:24,035 INFO [train.py:421] (2/8) Epoch 5, batch 63600, loss[loss=2.717, over 910.00 frames. , ppl: 15.141209072664203] tot_loss[loss=2.295, over 5523644.95 frames. , ppl: 9.924296892611641], batch size: 70
+2022-12-11 21:40:04,571 INFO [train.py:421] (2/8) Epoch 5, batch 63800, loss[loss=2.274, over 2450.00 frames. , ppl: 9.720701751279602] tot_loss[loss=2.294, over 5540455.52 frames. , ppl: 9.916528523418753], batch size: 70
+2022-12-11 21:41:46,258 INFO [train.py:421] (2/8) Epoch 5, batch 64000, loss[loss=2.536, over 1050.00 frames. , ppl: 12.63454651891178] tot_loss[loss=2.295, over 5493741.97 frames. , ppl: 9.925125640540108], batch size: 70
+2022-12-11 21:41:46,258 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:41:47,016 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.285, over 211138.00 frames. , ppl: 9.828019797907846
+2022-12-11 21:43:27,812 INFO [train.py:421] (2/8) Epoch 5, batch 64200, loss[loss=3.309, over 490.00 frames. , ppl: 27.36038001630865] tot_loss[loss=2.295, over 5496020.17 frames. , ppl: 9.924149261614101], batch size: 70
+2022-12-11 21:45:08,808 INFO [train.py:421] (2/8) Epoch 5, batch 64400, loss[loss=2.212, over 3990.00 frames. , ppl: 9.13513507440185] tot_loss[loss=2.295, over 5506458.65 frames. , ppl: 9.928038747730854], batch size: 70
+2022-12-11 21:46:48,251 INFO [train.py:421] (2/8) Epoch 5, batch 64600, loss[loss=2.437, over 1610.00 frames. , ppl: 11.434544436522245] tot_loss[loss=2.294, over 5561799.55 frames. , ppl: 9.916677694847488], batch size: 70
+2022-12-11 21:48:30,043 INFO [train.py:421] (2/8) Epoch 5, batch 64800, loss[loss=2.258, over 4550.00 frames. , ppl: 9.567880853281636] tot_loss[loss=2.294, over 5555218.40 frames. , ppl: 9.915927743527012], batch size: 70
+2022-12-11 21:50:08,871 INFO [train.py:421] (2/8) Epoch 5, batch 65000, loss[loss=2.572, over 770.00 frames. , ppl: 13.094440952330094] tot_loss[loss=2.295, over 5511604.05 frames. , ppl: 9.923970429568248], batch size: 70
+2022-12-11 21:50:08,872 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:50:09,631 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.835764336858247
+2022-12-11 21:51:51,527 INFO [train.py:421] (2/8) Epoch 5, batch 65200, loss[loss=2.18, over 4340.00 frames. , ppl: 8.848799171990628] tot_loss[loss=2.294, over 5549639.21 frames. , ppl: 9.916567204799597], batch size: 70
+2022-12-11 21:53:29,532 INFO [train.py:421] (2/8) Epoch 5, batch 65400, loss[loss=2.41, over 1260.00 frames. , ppl: 11.128401766049885] tot_loss[loss=2.295, over 5535978.38 frames. , ppl: 9.920877076588035], batch size: 70
+2022-12-11 21:55:05,986 INFO [train.py:421] (2/8) Epoch 5, batch 65600, loss[loss=2.505, over 1120.00 frames. , ppl: 12.24145446472596] tot_loss[loss=2.294, over 5519692.68 frames. , ppl: 9.91634228956622], batch size: 70
+2022-12-11 21:56:43,518 INFO [train.py:421] (2/8) Epoch 5, batch 65800, loss[loss=2.357, over 1330.00 frames. , ppl: 10.56000202254871] tot_loss[loss=2.294, over 5506758.05 frames. , ppl: 9.918813138920687], batch size: 70
+2022-12-11 21:58:21,357 INFO [train.py:421] (2/8) Epoch 5, batch 66000, loss[loss=2.299, over 2730.00 frames. , ppl: 9.96345236928615] tot_loss[loss=2.295, over 5502665.46 frames. , ppl: 9.92197617188107], batch size: 70
+2022-12-11 21:58:21,357 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 21:58:22,104 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816399892600549
+2022-12-11 22:00:01,629 INFO [train.py:421] (2/8) Epoch 5, batch 66200, loss[loss=2.525, over 980.00 frames. , ppl: 12.492787363291932] tot_loss[loss=2.296, over 5483964.15 frames. , ppl: 9.93241769102722], batch size: 70
+2022-12-11 22:01:40,816 INFO [train.py:421] (2/8) Epoch 5, batch 66400, loss[loss=2.319, over 3920.00 frames. , ppl: 10.170403895010203] tot_loss[loss=2.296, over 5459457.13 frames. , ppl: 9.937912991072563], batch size: 70
+2022-12-11 22:03:20,545 INFO [train.py:421] (2/8) Epoch 5, batch 66600, loss[loss=2.283, over 2590.00 frames. , ppl: 9.807922660226614] tot_loss[loss=2.295, over 5494163.73 frames. , ppl: 9.925833602086827], batch size: 70
+2022-12-11 22:05:00,223 INFO [train.py:421] (2/8) Epoch 5, batch 66800, loss[loss=2.247, over 6020.00 frames. , ppl: 9.455156875842611] tot_loss[loss=2.296, over 5476697.56 frames. , ppl: 9.93245269537887], batch size: 70
+2022-12-11 22:06:37,492 INFO [train.py:421] (2/8) Epoch 5, batch 67000, loss[loss=2.431, over 1050.00 frames. , ppl: 11.36621880839786] tot_loss[loss=2.296, over 5440993.63 frames. , ppl: 9.937417622161474], batch size: 70
+2022-12-11 22:06:37,493 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 22:06:38,242 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.812880145960932
+2022-12-11 22:08:16,524 INFO [train.py:421] (2/8) Epoch 5, batch 67200, loss[loss=2.269, over 2450.00 frames. , ppl: 9.667962288491566] tot_loss[loss=2.296, over 5454897.84 frames. , ppl: 9.929528940347549], batch size: 70
+2022-12-11 22:09:55,880 INFO [train.py:421] (2/8) Epoch 5, batch 67400, loss[loss=2.242, over 4480.00 frames. , ppl: 9.412368920521535] tot_loss[loss=2.295, over 5469250.54 frames. , ppl: 9.926982314992703], batch size: 70
+2022-12-11 22:11:35,230 INFO [train.py:421] (2/8) Epoch 5, batch 67600, loss[loss=2.416, over 1750.00 frames. , ppl: 11.199315703933618] tot_loss[loss=2.295, over 5466503.88 frames. , ppl: 9.92577397867612], batch size: 70
+2022-12-11 22:13:14,516 INFO [train.py:421] (2/8) Epoch 5, batch 67800, loss[loss=2.418, over 1540.00 frames. , ppl: 11.222436207922275] tot_loss[loss=2.295, over 5466303.07 frames. , ppl: 9.921816021738566], batch size: 70
+2022-12-11 22:14:52,339 INFO [train.py:421] (2/8) Epoch 5, batch 68000, loss[loss=2.255, over 4690.00 frames. , ppl: 9.534912506421755] tot_loss[loss=2.295, over 5452761.58 frames. , ppl: 9.924844650573952], batch size: 70
+2022-12-11 22:14:52,339 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 22:14:53,100 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.814543261416496
+2022-12-11 22:16:33,228 INFO [train.py:421] (2/8) Epoch 5, batch 68200, loss[loss=2.178, over 5670.00 frames. , ppl: 8.829617876104738] tot_loss[loss=2.294, over 5496748.26 frames. , ppl: 9.914919487329739], batch size: 70
+2022-12-11 22:18:14,671 INFO [train.py:421] (2/8) Epoch 5, batch 68400, loss[loss=2.22, over 2940.00 frames. , ppl: 9.202821829085707] tot_loss[loss=2.293, over 5526067.95 frames. , ppl: 9.902928252373671], batch size: 70
+2022-12-11 22:19:55,894 INFO [train.py:421] (2/8) Epoch 5, batch 68600, loss[loss=2.468, over 1330.00 frames. , ppl: 11.803144550700454] tot_loss[loss=2.293, over 5538356.23 frames. , ppl: 9.90688508553848], batch size: 70
+2022-12-11 22:21:39,030 INFO [train.py:421] (2/8) Epoch 5, batch 68800, loss[loss=2.386, over 1610.00 frames. , ppl: 10.868288591340344] tot_loss[loss=2.294, over 5538230.71 frames. , ppl: 9.913163288469432], batch size: 70
+2022-12-11 22:23:22,189 INFO [train.py:421] (2/8) Epoch 5, batch 69000, loss[loss=2.269, over 2450.00 frames. , ppl: 9.673721311619083] tot_loss[loss=2.294, over 5545813.22 frames. , ppl: 9.910138246934913], batch size: 70
+2022-12-11 22:23:22,190 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 22:23:22,952 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816581506837656
+2022-12-11 22:25:00,962 INFO [train.py:421] (2/8) Epoch 5, batch 69200, loss[loss=2.28, over 3360.00 frames. , ppl: 9.778308173176699] tot_loss[loss=2.294, over 5557467.22 frames. , ppl: 9.912112410079954], batch size: 70
+2022-12-11 22:26:43,990 INFO [train.py:421] (2/8) Epoch 5, batch 69400, loss[loss=2.316, over 3850.00 frames. , ppl: 10.13684644235621] tot_loss[loss=2.296, over 5490126.61 frames. , ppl: 9.929481758118936], batch size: 70
+2022-12-11 22:28:25,073 INFO [train.py:421] (2/8) Epoch 5, batch 69600, loss[loss=3.559, over 420.00 frames. , ppl: 35.121857205999895] tot_loss[loss=2.295, over 5511221.52 frames. , ppl: 9.926579704455882], batch size: 70
+2022-12-11 22:30:05,465 INFO [train.py:421] (2/8) Epoch 5, batch 69800, loss[loss=2.463, over 1470.00 frames. , ppl: 11.739942510735002] tot_loss[loss=2.295, over 5507445.47 frames. , ppl: 9.924780663333669], batch size: 70
+2022-12-11 22:31:48,584 INFO [train.py:421] (2/8) Epoch 5, batch 70000, loss[loss=2.219, over 6720.00 frames. , ppl: 9.199250596265388] tot_loss[loss=2.294, over 5531535.62 frames. , ppl: 9.918204002827132], batch size: 70
+2022-12-11 22:31:48,584 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 22:31:49,328 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.283, over 211138.00 frames. , ppl: 9.80977109406468
+2022-12-11 22:33:34,153 INFO [train.py:421] (2/8) Epoch 5, batch 70200, loss[loss=2.223, over 4340.00 frames. , ppl: 9.233696334923422] tot_loss[loss=2.294, over 5525950.18 frames. , ppl: 9.91817723523485], batch size: 70
+2022-12-11 22:35:14,438 INFO [train.py:421] (2/8) Epoch 5, batch 70400, loss[loss=2.447, over 1820.00 frames. , ppl: 11.553621623767196] tot_loss[loss=2.294, over 5544974.56 frames. , ppl: 9.909602424392347], batch size: 70
+2022-12-11 22:36:54,989 INFO [train.py:421] (2/8) Epoch 5, batch 70600, loss[loss=2.5, over 1120.00 frames. , ppl: 12.184119279751785] tot_loss[loss=2.293, over 5538201.01 frames. , ppl: 9.90939007869066], batch size: 70
+2022-12-11 22:38:32,952 INFO [train.py:421] (2/8) Epoch 5, batch 70800, loss[loss=2.296, over 2520.00 frames. , ppl: 9.934472899530165] tot_loss[loss=2.295, over 5516053.75 frames. , ppl: 9.919814094889047], batch size: 70
+2022-12-11 22:40:16,508 INFO [train.py:421] (2/8) Epoch 5, batch 71000, loss[loss=2.311, over 4620.00 frames. , ppl: 10.079554230367828] tot_loss[loss=2.294, over 5503668.62 frames. , ppl: 9.9181097725587], batch size: 70
+2022-12-11 22:40:16,509 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 22:40:17,255 INFO [train.py:452] (2/8) Epoch 5, validation: loss=2.283, over 211138.00 frames. , ppl: 9.803261396653303
+2022-12-11 22:41:56,395 INFO [train.py:421] (2/8) Epoch 5, batch 71200, loss[loss=2.205, over 4270.00 frames. , ppl: 9.06933390184022] tot_loss[loss=2.295, over 5503314.75 frames. , ppl: 9.92114875212958], batch size: 70
+2022-12-11 22:43:36,034 INFO [train.py:421] (2/8) Epoch 5, batch 71400, loss[loss=2.267, over 3220.00 frames. , ppl: 9.651963999392569] tot_loss[loss=2.295, over 5514294.12 frames. , ppl: 9.91997211372207], batch size: 70
+2022-12-11 22:45:15,393 INFO [train.py:421] (2/8) Epoch 5, batch 71600, loss[loss=2.196, over 8750.00 frames. , ppl: 8.986832864879755] tot_loss[loss=2.295, over 5508693.73 frames. , ppl: 9.921513454649446], batch size: 70
+2022-12-11 22:46:51,070 INFO [train.py:421] (2/8) Epoch 5, batch 71800, loss[loss=2.481, over 1120.00 frames. , ppl: 11.950968136906091] tot_loss[loss=2.295, over 5487070.50 frames. , ppl: 9.923409142072343], batch size: 70
+2022-12-11 22:48:06,679 INFO [train.py:421] (2/8) Epoch 6, batch 0, loss[loss=2.371, over 1960.00 frames. , ppl: 10.710120806081568] tot_loss[loss=2.371, over 1960.00 frames. , ppl: 10.710120806081568], batch size: 70
+2022-12-11 22:49:46,706 INFO [train.py:421] (2/8) Epoch 6, batch 200, loss[loss=2.291, over 2590.00 frames. , ppl: 9.88477447299645] tot_loss[loss=2.279, over 548393.67 frames. , ppl: 9.76651002882609], batch size: 70
+2022-12-11 22:51:26,927 INFO [train.py:421] (2/8) Epoch 6, batch 400, loss[loss=2.444, over 1610.00 frames. , ppl: 11.523602826817891] tot_loss[loss=2.285, over 1009711.56 frames. , ppl: 9.82085157986299], batch size: 70
+2022-12-11 22:53:08,650 INFO [train.py:421] (2/8) Epoch 6, batch 600, loss[loss=2.139, over 5600.00 frames. , ppl: 8.487327925025104] tot_loss[loss=2.277, over 1490471.30 frames. , ppl: 9.749154373585302], batch size: 70
+2022-12-11 22:54:51,848 INFO [train.py:421] (2/8) Epoch 6, batch 800, loss[loss=2.31, over 1890.00 frames. , ppl: 10.075426233869154] tot_loss[loss=2.279, over 1885825.77 frames. , ppl: 9.770236460946062], batch size: 70
+2022-12-11 22:56:35,370 INFO [train.py:421] (2/8) Epoch 6, batch 1000, loss[loss=2.303, over 1470.00 frames. , ppl: 10.004074565116738] tot_loss[loss=2.281, over 2264646.36 frames. , ppl: 9.787958131864082], batch size: 70
+2022-12-11 22:56:35,370 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 22:56:36,121 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816725347697853
+2022-12-11 22:58:21,089 INFO [train.py:421] (2/8) Epoch 6, batch 1200, loss[loss=2.341, over 2450.00 frames. , ppl: 10.391864062862108] tot_loss[loss=2.281, over 2590640.51 frames. , ppl: 9.783055487678782], batch size: 70
+2022-12-11 22:59:58,474 INFO [train.py:421] (2/8) Epoch 6, batch 1400, loss[loss=2.434, over 980.00 frames. , ppl: 11.400191140502814] tot_loss[loss=2.282, over 2848132.88 frames. , ppl: 9.791858585238417], batch size: 70
+2022-12-11 23:01:38,963 INFO [train.py:421] (2/8) Epoch 6, batch 1600, loss[loss=2.17, over 13230.00 frames. , ppl: 8.762596906313437] tot_loss[loss=2.284, over 3073181.64 frames. , ppl: 9.814092310920577], batch size: 70
+2022-12-11 23:03:19,723 INFO [train.py:421] (2/8) Epoch 6, batch 1800, loss[loss=2.368, over 3710.00 frames. , ppl: 10.676188279583787] tot_loss[loss=2.284, over 3323236.68 frames. , ppl: 9.817078100139668], batch size: 70
+2022-12-11 23:04:57,205 INFO [train.py:421] (2/8) Epoch 6, batch 2000, loss[loss=2.463, over 1330.00 frames. , ppl: 11.74435890585506] tot_loss[loss=2.282, over 3578096.33 frames. , ppl: 9.796821573764912], batch size: 70
+2022-12-11 23:04:57,206 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:04:57,952 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.286, over 211138.00 frames. , ppl: 9.840409328583648
+2022-12-11 23:06:35,159 INFO [train.py:421] (2/8) Epoch 6, batch 2200, loss[loss=2.174, over 7840.00 frames. , ppl: 8.79147310464111] tot_loss[loss=2.283, over 3731163.32 frames. , ppl: 9.808583746537728], batch size: 70
+2022-12-11 23:08:19,874 INFO [train.py:421] (2/8) Epoch 6, batch 2400, loss[loss=2.219, over 4060.00 frames. , ppl: 9.201860500863132] tot_loss[loss=2.282, over 3966615.23 frames. , ppl: 9.795039014469168], batch size: 70
+2022-12-11 23:10:02,603 INFO [train.py:421] (2/8) Epoch 6, batch 2600, loss[loss=2.428, over 1750.00 frames. , ppl: 11.337085345664999] tot_loss[loss=2.283, over 4106023.21 frames. , ppl: 9.802748883622263], batch size: 70
+2022-12-11 23:11:44,606 INFO [train.py:421] (2/8) Epoch 6, batch 2800, loss[loss=2.212, over 5320.00 frames. , ppl: 9.136917218437624] tot_loss[loss=2.282, over 4231294.51 frames. , ppl: 9.798144609034598], batch size: 70
+2022-12-11 23:13:24,432 INFO [train.py:421] (2/8) Epoch 6, batch 3000, loss[loss=2.246, over 3150.00 frames. , ppl: 9.454028738796712] tot_loss[loss=2.283, over 4359810.89 frames. , ppl: 9.809811519260531], batch size: 70
+2022-12-11 23:13:24,433 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:13:25,195 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.824575860616017
+2022-12-11 23:15:01,075 INFO [train.py:421] (2/8) Epoch 6, batch 3200, loss[loss=2.417, over 1470.00 frames. , ppl: 11.209749834999283] tot_loss[loss=2.283, over 4483721.59 frames. , ppl: 9.807581517378098], batch size: 70
+2022-12-11 23:16:39,607 INFO [train.py:421] (2/8) Epoch 6, batch 3400, loss[loss=2.306, over 1890.00 frames. , ppl: 10.035946223092244] tot_loss[loss=2.285, over 4546511.94 frames. , ppl: 9.823190526262335], batch size: 70
+2022-12-11 23:18:19,908 INFO [train.py:421] (2/8) Epoch 6, batch 3600, loss[loss=2.36, over 2730.00 frames. , ppl: 10.589238411240107] tot_loss[loss=2.286, over 4589703.10 frames. , ppl: 9.838774866445029], batch size: 70
+2022-12-11 23:19:59,208 INFO [train.py:421] (2/8) Epoch 6, batch 3800, loss[loss=2.151, over 4410.00 frames. , ppl: 8.592119692799633] tot_loss[loss=2.285, over 4736398.29 frames. , ppl: 9.823942101903818], batch size: 70
+2022-12-11 23:21:41,562 INFO [train.py:421] (2/8) Epoch 6, batch 4000, loss[loss=2.227, over 3990.00 frames. , ppl: 9.268161363405738] tot_loss[loss=2.285, over 4836043.40 frames. , ppl: 9.821491919701646], batch size: 70
+2022-12-11 23:21:41,563 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:21:42,322 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.808357026550453
+2022-12-11 23:23:26,050 INFO [train.py:421] (2/8) Epoch 6, batch 4200, loss[loss=2.352, over 2590.00 frames. , ppl: 10.50925871688848] tot_loss[loss=2.285, over 4889897.69 frames. , ppl: 9.824348783199987], batch size: 70
+2022-12-11 23:25:03,905 INFO [train.py:421] (2/8) Epoch 6, batch 4400, loss[loss=2.406, over 1400.00 frames. , ppl: 11.091130302860847] tot_loss[loss=2.287, over 4926794.19 frames. , ppl: 9.84126547915421], batch size: 70
+2022-12-11 23:26:43,142 INFO [train.py:421] (2/8) Epoch 6, batch 4600, loss[loss=2.258, over 3570.00 frames. , ppl: 9.559933832567467] tot_loss[loss=2.287, over 4950133.35 frames. , ppl: 9.850144959968608], batch size: 70
+2022-12-11 23:28:17,780 INFO [train.py:421] (2/8) Epoch 6, batch 4800, loss[loss=2.425, over 1820.00 frames. , ppl: 11.30324223096481] tot_loss[loss=2.289, over 4947462.29 frames. , ppl: 9.864488392217007], batch size: 70
+2022-12-11 23:29:55,650 INFO [train.py:421] (2/8) Epoch 6, batch 5000, loss[loss=2.42, over 1540.00 frames. , ppl: 11.24139811206521] tot_loss[loss=2.289, over 4983750.72 frames. , ppl: 9.864751208975957], batch size: 70
+2022-12-11 23:29:55,651 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:29:56,411 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.82324543954202
+2022-12-11 23:31:38,998 INFO [train.py:421] (2/8) Epoch 6, batch 5200, loss[loss=2.276, over 2100.00 frames. , ppl: 9.739293896412361] tot_loss[loss=2.288, over 5037324.84 frames. , ppl: 9.856026816104865], batch size: 70
+2022-12-11 23:33:16,575 INFO [train.py:421] (2/8) Epoch 6, batch 5400, loss[loss=2.248, over 3920.00 frames. , ppl: 9.470742506178672] tot_loss[loss=2.289, over 5033286.95 frames. , ppl: 9.868759160851592], batch size: 70
+2022-12-11 23:34:57,714 INFO [train.py:421] (2/8) Epoch 6, batch 5600, loss[loss=2.323, over 2590.00 frames. , ppl: 10.208880958259261] tot_loss[loss=2.29, over 5056708.15 frames. , ppl: 9.878865535825039], batch size: 70
+2022-12-11 23:36:40,248 INFO [train.py:421] (2/8) Epoch 6, batch 5800, loss[loss=2.33, over 1400.00 frames. , ppl: 10.281749148705835] tot_loss[loss=2.291, over 5073523.26 frames. , ppl: 9.882668152540578], batch size: 70
+2022-12-11 23:38:22,691 INFO [train.py:421] (2/8) Epoch 6, batch 6000, loss[loss=2.179, over 3290.00 frames. , ppl: 8.838142917754121] tot_loss[loss=2.29, over 5093992.30 frames. , ppl: 9.879169538031855], batch size: 70
+2022-12-11 23:38:22,691 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:38:23,436 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.813196769801683
+2022-12-11 23:40:02,703 INFO [train.py:421] (2/8) Epoch 6, batch 6200, loss[loss=2.202, over 5810.00 frames. , ppl: 9.04669702656977] tot_loss[loss=2.292, over 5058516.43 frames. , ppl: 9.89943760384488], batch size: 70
+2022-12-11 23:41:42,417 INFO [train.py:421] (2/8) Epoch 6, batch 6400, loss[loss=2.418, over 1750.00 frames. , ppl: 11.225629337607797] tot_loss[loss=2.293, over 5065502.76 frames. , ppl: 9.906835035912923], batch size: 70
+2022-12-11 23:43:22,509 INFO [train.py:421] (2/8) Epoch 6, batch 6600, loss[loss=2.308, over 5670.00 frames. , ppl: 10.053796120276168] tot_loss[loss=2.292, over 5128234.33 frames. , ppl: 9.896644904783642], batch size: 70
+2022-12-11 23:45:02,938 INFO [train.py:421] (2/8) Epoch 6, batch 6800, loss[loss=2.225, over 4690.00 frames. , ppl: 9.257084658974971] tot_loss[loss=2.29, over 5225181.55 frames. , ppl: 9.876388370999905], batch size: 70
+2022-12-11 23:46:38,670 INFO [train.py:421] (2/8) Epoch 6, batch 7000, loss[loss=2.108, over 3640.00 frames. , ppl: 8.229533508235896] tot_loss[loss=2.291, over 5239247.34 frames. , ppl: 9.880447725339923], batch size: 70
+2022-12-11 23:46:38,671 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:46:39,429 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.822367314949352
+2022-12-11 23:48:19,180 INFO [train.py:421] (2/8) Epoch 6, batch 7200, loss[loss=2.349, over 2310.00 frames. , ppl: 10.474657371281028] tot_loss[loss=2.292, over 5243896.89 frames. , ppl: 9.88978261168555], batch size: 70
+2022-12-11 23:50:03,974 INFO [train.py:421] (2/8) Epoch 6, batch 7400, loss[loss=2.204, over 5740.00 frames. , ppl: 9.065131615008026] tot_loss[loss=2.292, over 5265840.73 frames. , ppl: 9.896712532176211], batch size: 70
+2022-12-11 23:51:43,015 INFO [train.py:421] (2/8) Epoch 6, batch 7600, loss[loss=2.286, over 2310.00 frames. , ppl: 9.839547424738006] tot_loss[loss=2.292, over 5283161.06 frames. , ppl: 9.891506784136308], batch size: 70
+2022-12-11 23:53:23,567 INFO [train.py:421] (2/8) Epoch 6, batch 7800, loss[loss=2.615, over 770.00 frames. , ppl: 13.67120672735957] tot_loss[loss=2.293, over 5266089.28 frames. , ppl: 9.900808594933311], batch size: 70
+2022-12-11 23:55:07,597 INFO [train.py:421] (2/8) Epoch 6, batch 8000, loss[loss=2.22, over 3430.00 frames. , ppl: 9.209372144140621] tot_loss[loss=2.291, over 5323987.42 frames. , ppl: 9.888178922476158], batch size: 70
+2022-12-11 23:55:07,598 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-11 23:55:08,354 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.793563654653626
+2022-12-11 23:56:47,538 INFO [train.py:421] (2/8) Epoch 6, batch 8200, loss[loss=2.312, over 1680.00 frames. , ppl: 10.092106939973615] tot_loss[loss=2.291, over 5329706.96 frames. , ppl: 9.884004694329219], batch size: 70
+2022-12-11 23:58:28,578 INFO [train.py:421] (2/8) Epoch 6, batch 8400, loss[loss=2.235, over 1400.00 frames. , ppl: 9.343330162735288] tot_loss[loss=2.289, over 5401781.70 frames. , ppl: 9.866123895279314], batch size: 70
+2022-12-12 00:00:05,797 INFO [train.py:421] (2/8) Epoch 6, batch 8600, loss[loss=2.28, over 3500.00 frames. , ppl: 9.78022318841341] tot_loss[loss=2.29, over 5421329.73 frames. , ppl: 9.87096713603385], batch size: 70
+2022-12-12 00:01:46,887 INFO [train.py:421] (2/8) Epoch 6, batch 8800, loss[loss=2.239, over 4410.00 frames. , ppl: 9.38737040777569] tot_loss[loss=2.29, over 5413917.03 frames. , ppl: 9.87297658387095], batch size: 70
+2022-12-12 00:03:27,014 INFO [train.py:421] (2/8) Epoch 6, batch 9000, loss[loss=2.122, over 7070.00 frames. , ppl: 8.346962105259733] tot_loss[loss=2.29, over 5405390.19 frames. , ppl: 9.87430143811922], batch size: 70
+2022-12-12 00:03:27,015 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:03:27,760 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.82595299982713
+2022-12-12 00:05:06,482 INFO [train.py:421] (2/8) Epoch 6, batch 9200, loss[loss=2.354, over 1890.00 frames. , ppl: 10.532800595813828] tot_loss[loss=2.291, over 5404084.75 frames. , ppl: 9.884183366837416], batch size: 70
+2022-12-12 00:06:43,068 INFO [train.py:421] (2/8) Epoch 6, batch 9400, loss[loss=2.246, over 2100.00 frames. , ppl: 9.452082797264033] tot_loss[loss=2.291, over 5410572.31 frames. , ppl: 9.881126677404483], batch size: 70
+2022-12-12 00:08:21,540 INFO [train.py:421] (2/8) Epoch 6, batch 9600, loss[loss=2.332, over 1540.00 frames. , ppl: 10.298725530726376] tot_loss[loss=2.292, over 5361748.15 frames. , ppl: 9.891176886808909], batch size: 70
+2022-12-12 00:10:01,615 INFO [train.py:421] (2/8) Epoch 6, batch 9800, loss[loss=2.343, over 1820.00 frames. , ppl: 10.414766920258435] tot_loss[loss=2.293, over 5335376.09 frames. , ppl: 9.903693313657843], batch size: 70
+2022-12-12 00:11:46,128 INFO [train.py:421] (2/8) Epoch 6, batch 10000, loss[loss=3.212, over 490.00 frames. , ppl: 24.828273631978824] tot_loss[loss=2.292, over 5381685.99 frames. , ppl: 9.88981343074749], batch size: 70
+2022-12-12 00:11:46,129 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:11:46,874 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.804028982352971
+2022-12-12 00:13:23,522 INFO [train.py:421] (2/8) Epoch 6, batch 10200, loss[loss=2.608, over 770.00 frames. , ppl: 13.574232423323885] tot_loss[loss=2.293, over 5324451.93 frames. , ppl: 9.906523575545659], batch size: 70
+2022-12-12 00:15:02,166 INFO [train.py:421] (2/8) Epoch 6, batch 10400, loss[loss=2.241, over 12530.00 frames. , ppl: 9.407352027121751] tot_loss[loss=2.293, over 5338374.98 frames. , ppl: 9.901309162181736], batch size: 70
+2022-12-12 00:16:43,003 INFO [train.py:421] (2/8) Epoch 6, batch 10600, loss[loss=2.989, over 560.00 frames. , ppl: 19.86416438154406] tot_loss[loss=2.292, over 5362367.01 frames. , ppl: 9.894102020797115], batch size: 70
+2022-12-12 00:18:24,430 INFO [train.py:421] (2/8) Epoch 6, batch 10800, loss[loss=2.239, over 3640.00 frames. , ppl: 9.3867513359762] tot_loss[loss=2.291, over 5390859.29 frames. , ppl: 9.888473092838412], batch size: 70
+2022-12-12 00:20:06,164 INFO [train.py:421] (2/8) Epoch 6, batch 11000, loss[loss=2.236, over 4900.00 frames. , ppl: 9.359507744473134] tot_loss[loss=2.291, over 5406995.80 frames. , ppl: 9.884359210149206], batch size: 70
+2022-12-12 00:20:06,164 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:20:06,914 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.804163932742426
+2022-12-12 00:21:45,894 INFO [train.py:421] (2/8) Epoch 6, batch 11200, loss[loss=2.248, over 2590.00 frames. , ppl: 9.471137104863656] tot_loss[loss=2.291, over 5402867.39 frames. , ppl: 9.888497184131076], batch size: 70
+2022-12-12 00:23:23,950 INFO [train.py:421] (2/8) Epoch 6, batch 11400, loss[loss=2.129, over 8470.00 frames. , ppl: 8.40798853628819] tot_loss[loss=2.29, over 5430298.31 frames. , ppl: 9.879011719599958], batch size: 70
+2022-12-12 00:25:02,701 INFO [train.py:421] (2/8) Epoch 6, batch 11600, loss[loss=2.347, over 2240.00 frames. , ppl: 10.449529602046233] tot_loss[loss=2.29, over 5451959.38 frames. , ppl: 9.87580299672426], batch size: 70
+2022-12-12 00:26:41,475 INFO [train.py:421] (2/8) Epoch 6, batch 11800, loss[loss=2.29, over 2730.00 frames. , ppl: 9.876874341135538] tot_loss[loss=2.29, over 5473503.89 frames. , ppl: 9.87100819815142], batch size: 70
+2022-12-12 00:28:20,379 INFO [train.py:421] (2/8) Epoch 6, batch 12000, loss[loss=4.077, over 350.00 frames. , ppl: 58.94781041810843] tot_loss[loss=2.291, over 5454062.44 frames. , ppl: 9.880318628876697], batch size: 70
+2022-12-12 00:28:20,379 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:28:21,131 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.80987272894533
+2022-12-12 00:30:01,071 INFO [train.py:421] (2/8) Epoch 6, batch 12200, loss[loss=2.32, over 1470.00 frames. , ppl: 10.17794523931397] tot_loss[loss=2.291, over 5435474.18 frames. , ppl: 9.886075720856637], batch size: 70
+2022-12-12 00:31:40,361 INFO [train.py:421] (2/8) Epoch 6, batch 12400, loss[loss=2.237, over 5600.00 frames. , ppl: 9.36199993136714] tot_loss[loss=2.291, over 5458350.43 frames. , ppl: 9.882654307074636], batch size: 70
+2022-12-12 00:33:20,724 INFO [train.py:421] (2/8) Epoch 6, batch 12600, loss[loss=2.501, over 1890.00 frames. , ppl: 12.197629498573988] tot_loss[loss=2.291, over 5462798.44 frames. , ppl: 9.882691357391046], batch size: 70
+2022-12-12 00:35:01,350 INFO [train.py:421] (2/8) Epoch 6, batch 12800, loss[loss=2.157, over 4340.00 frames. , ppl: 8.649286037447327] tot_loss[loss=2.29, over 5464919.08 frames. , ppl: 9.877954266065826], batch size: 70
+2022-12-12 00:36:42,468 INFO [train.py:421] (2/8) Epoch 6, batch 13000, loss[loss=2.232, over 5530.00 frames. , ppl: 9.319223254699] tot_loss[loss=2.29, over 5472349.04 frames. , ppl: 9.875712590111052], batch size: 70
+2022-12-12 00:36:42,469 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:36:43,214 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.805265371495329
+2022-12-12 00:38:26,661 INFO [train.py:421] (2/8) Epoch 6, batch 13200, loss[loss=2.685, over 910.00 frames. , ppl: 14.66096389604185] tot_loss[loss=2.289, over 5520248.58 frames. , ppl: 9.861502124047504], batch size: 70
+2022-12-12 00:40:05,802 INFO [train.py:421] (2/8) Epoch 6, batch 13400, loss[loss=2.325, over 1470.00 frames. , ppl: 10.22743932920205] tot_loss[loss=2.288, over 5541722.68 frames. , ppl: 9.853196251130848], batch size: 70
+2022-12-12 00:41:44,769 INFO [train.py:421] (2/8) Epoch 6, batch 13600, loss[loss=2.284, over 2660.00 frames. , ppl: 9.816249032885457] tot_loss[loss=2.288, over 5551982.33 frames. , ppl: 9.85089310507244], batch size: 70
+2022-12-12 00:43:25,849 INFO [train.py:421] (2/8) Epoch 6, batch 13800, loss[loss=2.258, over 4270.00 frames. , ppl: 9.56314634633173] tot_loss[loss=2.287, over 5591751.34 frames. , ppl: 9.844383660916272], batch size: 70
+2022-12-12 00:45:05,054 INFO [train.py:421] (2/8) Epoch 6, batch 14000, loss[loss=2.639, over 910.00 frames. , ppl: 13.999170967625897] tot_loss[loss=2.287, over 5571616.32 frames. , ppl: 9.848909640378967], batch size: 70
+2022-12-12 00:45:05,054 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:45:05,803 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.822797644344924
+2022-12-12 00:46:45,819 INFO [train.py:421] (2/8) Epoch 6, batch 14200, loss[loss=2.196, over 10920.00 frames. , ppl: 8.991638983007123] tot_loss[loss=2.288, over 5551737.12 frames. , ppl: 9.85298567915496], batch size: 70
+2022-12-12 00:48:26,187 INFO [train.py:421] (2/8) Epoch 6, batch 14400, loss[loss=2.383, over 3290.00 frames. , ppl: 10.840153404373986] tot_loss[loss=2.29, over 5517300.54 frames. , ppl: 9.870227321340979], batch size: 70
+2022-12-12 00:50:08,143 INFO [train.py:421] (2/8) Epoch 6, batch 14600, loss[loss=2.262, over 2800.00 frames. , ppl: 9.600685886295386] tot_loss[loss=2.29, over 5511894.93 frames. , ppl: 9.877494907412817], batch size: 70
+2022-12-12 00:51:47,346 INFO [train.py:421] (2/8) Epoch 6, batch 14800, loss[loss=2.453, over 1050.00 frames. , ppl: 11.626098227042714] tot_loss[loss=2.292, over 5437164.84 frames. , ppl: 9.891850435632195], batch size: 70
+2022-12-12 00:53:30,891 INFO [train.py:421] (2/8) Epoch 6, batch 15000, loss[loss=2.571, over 1120.00 frames. , ppl: 13.085096513330258] tot_loss[loss=2.293, over 5403039.88 frames. , ppl: 9.901620082879942], batch size: 70
+2022-12-12 00:53:30,892 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 00:53:31,639 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.800679035400885
+2022-12-12 00:55:13,336 INFO [train.py:421] (2/8) Epoch 6, batch 15200, loss[loss=2.514, over 980.00 frames. , ppl: 12.354380325279159] tot_loss[loss=2.293, over 5388404.96 frames. , ppl: 9.903107557753039], batch size: 70
+2022-12-12 00:56:51,707 INFO [train.py:421] (2/8) Epoch 6, batch 15400, loss[loss=2.451, over 770.00 frames. , ppl: 11.604848571347024] tot_loss[loss=2.292, over 5391559.09 frames. , ppl: 9.899336288606891], batch size: 70
+2022-12-12 00:58:34,882 INFO [train.py:421] (2/8) Epoch 6, batch 15600, loss[loss=2.285, over 3080.00 frames. , ppl: 9.827464648721786] tot_loss[loss=2.293, over 5385217.29 frames. , ppl: 9.903507811371341], batch size: 70
+2022-12-12 01:00:14,970 INFO [train.py:421] (2/8) Epoch 6, batch 15800, loss[loss=2.484, over 1120.00 frames. , ppl: 11.986204519568505] tot_loss[loss=2.293, over 5409030.87 frames. , ppl: 9.899939759043356], batch size: 70
+2022-12-12 01:01:54,258 INFO [train.py:421] (2/8) Epoch 6, batch 16000, loss[loss=2.494, over 980.00 frames. , ppl: 12.111513981625427] tot_loss[loss=2.292, over 5454234.01 frames. , ppl: 9.894789529417944], batch size: 70
+2022-12-12 01:01:54,259 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:01:55,007 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.805294396587579
+2022-12-12 01:03:38,907 INFO [train.py:421] (2/8) Epoch 6, batch 16200, loss[loss=2.186, over 7560.00 frames. , ppl: 8.903874271917136] tot_loss[loss=2.291, over 5507761.56 frames. , ppl: 9.880879376749926], batch size: 70
+2022-12-12 01:05:19,348 INFO [train.py:421] (2/8) Epoch 6, batch 16400, loss[loss=2.344, over 1820.00 frames. , ppl: 10.426654384009687] tot_loss[loss=2.293, over 5424326.00 frames. , ppl: 9.907332829098795], batch size: 70
+2022-12-12 01:07:02,718 INFO [train.py:421] (2/8) Epoch 6, batch 16600, loss[loss=2.351, over 1820.00 frames. , ppl: 10.496991646091821] tot_loss[loss=2.293, over 5422793.98 frames. , ppl: 9.907874458033229], batch size: 70
+2022-12-12 01:08:42,226 INFO [train.py:421] (2/8) Epoch 6, batch 16800, loss[loss=2.184, over 5670.00 frames. , ppl: 8.880624727193547] tot_loss[loss=2.292, over 5433883.04 frames. , ppl: 9.898560523124829], batch size: 70
+2022-12-12 01:10:19,697 INFO [train.py:421] (2/8) Epoch 6, batch 17000, loss[loss=2.337, over 2100.00 frames. , ppl: 10.346601037649085] tot_loss[loss=2.294, over 5388128.47 frames. , ppl: 9.911863322006015], batch size: 70
+2022-12-12 01:10:19,698 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:10:20,459 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.80433371167732
+2022-12-12 01:12:00,280 INFO [train.py:421] (2/8) Epoch 6, batch 17200, loss[loss=2.384, over 2170.00 frames. , ppl: 10.851826125647744] tot_loss[loss=2.292, over 5436038.08 frames. , ppl: 9.895396491750931], batch size: 70
+2022-12-12 01:13:40,302 INFO [train.py:421] (2/8) Epoch 6, batch 17400, loss[loss=2.586, over 980.00 frames. , ppl: 13.27976476172428] tot_loss[loss=2.29, over 5489681.15 frames. , ppl: 9.87191962509011], batch size: 70
+2022-12-12 01:15:20,953 INFO [train.py:421] (2/8) Epoch 6, batch 17600, loss[loss=3.007, over 560.00 frames. , ppl: 20.232729157858003] tot_loss[loss=2.289, over 5499236.67 frames. , ppl: 9.868747104747106], batch size: 70
+2022-12-12 01:17:03,064 INFO [train.py:421] (2/8) Epoch 6, batch 17800, loss[loss=2.407, over 1680.00 frames. , ppl: 11.102488870042936] tot_loss[loss=2.289, over 5507154.97 frames. , ppl: 9.866198198369894], batch size: 70
+2022-12-12 01:18:44,984 INFO [train.py:421] (2/8) Epoch 6, batch 18000, loss[loss=2.775, over 630.00 frames. , ppl: 16.03467451013855] tot_loss[loss=2.289, over 5499628.98 frames. , ppl: 9.86812476778987], batch size: 70
+2022-12-12 01:18:44,985 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:18:45,748 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.802139871930427
+2022-12-12 01:20:22,555 INFO [train.py:421] (2/8) Epoch 6, batch 18200, loss[loss=2.311, over 2170.00 frames. , ppl: 10.089257401244142] tot_loss[loss=2.29, over 5487062.22 frames. , ppl: 9.876120743654251], batch size: 70
+2022-12-12 01:22:05,358 INFO [train.py:421] (2/8) Epoch 6, batch 18400, loss[loss=2.186, over 3360.00 frames. , ppl: 8.899600450547748] tot_loss[loss=2.291, over 5453184.06 frames. , ppl: 9.883658135368169], batch size: 70
+2022-12-12 01:23:45,396 INFO [train.py:421] (2/8) Epoch 6, batch 18600, loss[loss=2.25, over 1750.00 frames. , ppl: 9.486327605090066] tot_loss[loss=2.291, over 5452046.82 frames. , ppl: 9.886620002346927], batch size: 70
+2022-12-12 01:25:24,439 INFO [train.py:421] (2/8) Epoch 6, batch 18800, loss[loss=2.359, over 2030.00 frames. , ppl: 10.575539278255446] tot_loss[loss=2.292, over 5430189.44 frames. , ppl: 9.890215887889195], batch size: 70
+2022-12-12 01:27:04,203 INFO [train.py:421] (2/8) Epoch 6, batch 19000, loss[loss=2.312, over 3080.00 frames. , ppl: 10.094179062941377] tot_loss[loss=2.292, over 5406112.85 frames. , ppl: 9.899072219353029], batch size: 70
+2022-12-12 01:27:04,203 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:27:04,968 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.798580280075496
+2022-12-12 01:28:43,265 INFO [train.py:421] (2/8) Epoch 6, batch 19200, loss[loss=2.22, over 3920.00 frames. , ppl: 9.202807309073085] tot_loss[loss=2.293, over 5371693.88 frames. , ppl: 9.906820639917461], batch size: 70
+2022-12-12 01:30:21,610 INFO [train.py:421] (2/8) Epoch 6, batch 19400, loss[loss=2.455, over 1120.00 frames. , ppl: 11.648742490641137] tot_loss[loss=2.292, over 5401967.65 frames. , ppl: 9.896253760107577], batch size: 70
+2022-12-12 01:32:02,249 INFO [train.py:421] (2/8) Epoch 6, batch 19600, loss[loss=2.153, over 9450.00 frames. , ppl: 8.607101620073184] tot_loss[loss=2.292, over 5418509.31 frames. , ppl: 9.896639953500497], batch size: 70
+2022-12-12 01:33:40,021 INFO [train.py:421] (2/8) Epoch 6, batch 19800, loss[loss=2.231, over 9100.00 frames. , ppl: 9.310423039662577] tot_loss[loss=2.291, over 5427780.38 frames. , ppl: 9.886954832141111], batch size: 70
+2022-12-12 01:35:20,119 INFO [train.py:421] (2/8) Epoch 6, batch 20000, loss[loss=2.329, over 2520.00 frames. , ppl: 10.264929887263266] tot_loss[loss=2.292, over 5394166.39 frames. , ppl: 9.898829157103746], batch size: 70
+2022-12-12 01:35:20,120 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:35:20,866 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.811063387394697
+2022-12-12 01:36:57,457 INFO [train.py:421] (2/8) Epoch 6, batch 20200, loss[loss=2.294, over 2100.00 frames. , ppl: 9.9114528419099] tot_loss[loss=2.292, over 5412284.39 frames. , ppl: 9.894729034953556], batch size: 70
+2022-12-12 01:38:35,707 INFO [train.py:421] (2/8) Epoch 6, batch 20400, loss[loss=2.372, over 910.00 frames. , ppl: 10.714326412390886] tot_loss[loss=2.293, over 5403679.38 frames. , ppl: 9.900326002634072], batch size: 70
+2022-12-12 01:40:16,777 INFO [train.py:421] (2/8) Epoch 6, batch 20600, loss[loss=2.586, over 910.00 frames. , ppl: 13.278728542057136] tot_loss[loss=2.294, over 5377822.36 frames. , ppl: 9.910698485693372], batch size: 70
+2022-12-12 01:41:54,631 INFO [train.py:421] (2/8) Epoch 6, batch 20800, loss[loss=2.758, over 700.00 frames. , ppl: 15.765950907578452] tot_loss[loss=2.293, over 5405567.71 frames. , ppl: 9.901656985852366], batch size: 70
+2022-12-12 01:43:37,344 INFO [train.py:421] (2/8) Epoch 6, batch 21000, loss[loss=2.234, over 5810.00 frames. , ppl: 9.340273182587719] tot_loss[loss=2.291, over 5445726.98 frames. , ppl: 9.886204567278813], batch size: 70
+2022-12-12 01:43:37,345 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:43:38,089 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.790676635383708
+2022-12-12 01:45:21,894 INFO [train.py:421] (2/8) Epoch 6, batch 21200, loss[loss=2.773, over 700.00 frames. , ppl: 16.00868045425863] tot_loss[loss=2.292, over 5432257.00 frames. , ppl: 9.890118942036173], batch size: 70
+2022-12-12 01:47:02,204 INFO [train.py:421] (2/8) Epoch 6, batch 21400, loss[loss=2.503, over 1820.00 frames. , ppl: 12.21667041639867] tot_loss[loss=2.291, over 5415258.97 frames. , ppl: 9.886437503151566], batch size: 70
+2022-12-12 01:48:39,309 INFO [train.py:421] (2/8) Epoch 6, batch 21600, loss[loss=2.395, over 1400.00 frames. , ppl: 10.969582942714645] tot_loss[loss=2.291, over 5409084.14 frames. , ppl: 9.88143298422772], batch size: 70
+2022-12-12 01:50:16,059 INFO [train.py:421] (2/8) Epoch 6, batch 21800, loss[loss=2.458, over 1120.00 frames. , ppl: 11.680531169359366] tot_loss[loss=2.291, over 5411578.63 frames. , ppl: 9.888200990451855], batch size: 70
+2022-12-12 01:51:58,032 INFO [train.py:421] (2/8) Epoch 6, batch 22000, loss[loss=2.261, over 4550.00 frames. , ppl: 9.59037304092196] tot_loss[loss=2.291, over 5435272.50 frames. , ppl: 9.883132314098265], batch size: 70
+2022-12-12 01:51:58,033 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 01:51:58,792 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.776022789235363
+2022-12-12 01:53:38,879 INFO [train.py:421] (2/8) Epoch 6, batch 22200, loss[loss=2.503, over 1260.00 frames. , ppl: 12.220027105597612] tot_loss[loss=2.291, over 5417673.55 frames. , ppl: 9.888928552407043], batch size: 70
+2022-12-12 01:55:20,557 INFO [train.py:421] (2/8) Epoch 6, batch 22400, loss[loss=2.207, over 5530.00 frames. , ppl: 9.085823340323639] tot_loss[loss=2.29, over 5443457.99 frames. , ppl: 9.879659632248156], batch size: 70
+2022-12-12 01:57:04,222 INFO [train.py:421] (2/8) Epoch 6, batch 22600, loss[loss=2.472, over 2660.00 frames. , ppl: 11.847094069591138] tot_loss[loss=2.292, over 5388737.22 frames. , ppl: 9.893188774423887], batch size: 70
+2022-12-12 01:58:41,918 INFO [train.py:421] (2/8) Epoch 6, batch 22800, loss[loss=2.282, over 2730.00 frames. , ppl: 9.798231205494204] tot_loss[loss=2.292, over 5381838.76 frames. , ppl: 9.896690762774721], batch size: 70
+2022-12-12 02:00:22,256 INFO [train.py:421] (2/8) Epoch 6, batch 23000, loss[loss=2.252, over 3990.00 frames. , ppl: 9.506847753550586] tot_loss[loss=2.291, over 5413253.14 frames. , ppl: 9.883504914951509], batch size: 70
+2022-12-12 02:00:22,256 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:00:23,004 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.793408557185776
+2022-12-12 02:01:58,429 INFO [train.py:421] (2/8) Epoch 6, batch 23200, loss[loss=2.172, over 8050.00 frames. , ppl: 8.775887975966675] tot_loss[loss=2.291, over 5400216.24 frames. , ppl: 9.883334821486093], batch size: 70
+2022-12-12 02:03:43,836 INFO [train.py:421] (2/8) Epoch 6, batch 23400, loss[loss=2.884, over 630.00 frames. , ppl: 17.894426932510903] tot_loss[loss=2.291, over 5432303.06 frames. , ppl: 9.886792499252847], batch size: 70
+2022-12-12 02:05:24,922 INFO [train.py:421] (2/8) Epoch 6, batch 23600, loss[loss=2.204, over 4620.00 frames. , ppl: 9.060930118947667] tot_loss[loss=2.291, over 5448075.91 frames. , ppl: 9.880362521354227], batch size: 70
+2022-12-12 02:07:03,708 INFO [train.py:421] (2/8) Epoch 6, batch 23800, loss[loss=2.172, over 6650.00 frames. , ppl: 8.774968639406394] tot_loss[loss=2.29, over 5443418.59 frames. , ppl: 9.877843419356271], batch size: 70
+2022-12-12 02:08:44,636 INFO [train.py:421] (2/8) Epoch 6, batch 24000, loss[loss=2.292, over 3360.00 frames. , ppl: 9.89863395615282] tot_loss[loss=2.291, over 5404500.20 frames. , ppl: 9.886385325357832], batch size: 70
+2022-12-12 02:08:44,637 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:08:45,409 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.790136138531322
+2022-12-12 02:10:29,006 INFO [train.py:421] (2/8) Epoch 6, batch 24200, loss[loss=2.391, over 3150.00 frames. , ppl: 10.925020501632464] tot_loss[loss=2.291, over 5432455.23 frames. , ppl: 9.884003026473897], batch size: 70
+2022-12-12 02:12:08,530 INFO [train.py:421] (2/8) Epoch 6, batch 24400, loss[loss=2.345, over 1540.00 frames. , ppl: 10.437581007128284] tot_loss[loss=2.291, over 5406216.51 frames. , ppl: 9.88690644131115], batch size: 70
+2022-12-12 02:13:52,337 INFO [train.py:421] (2/8) Epoch 6, batch 24600, loss[loss=2.179, over 3990.00 frames. , ppl: 8.836283373151794] tot_loss[loss=2.291, over 5430472.48 frames. , ppl: 9.88700714685687], batch size: 70
+2022-12-12 02:15:29,117 INFO [train.py:421] (2/8) Epoch 6, batch 24800, loss[loss=2.154, over 3360.00 frames. , ppl: 8.617270089910756] tot_loss[loss=2.292, over 5415451.59 frames. , ppl: 9.894966717986579], batch size: 70
+2022-12-12 02:17:09,494 INFO [train.py:421] (2/8) Epoch 6, batch 25000, loss[loss=2.36, over 2240.00 frames. , ppl: 10.59510967500157] tot_loss[loss=2.292, over 5431625.75 frames. , ppl: 9.89272449075631], batch size: 70
+2022-12-12 02:17:09,495 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:17:10,254 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.788068615004098
+2022-12-12 02:18:49,721 INFO [train.py:421] (2/8) Epoch 6, batch 25200, loss[loss=5, over 280.00 frames. , ppl: 148.45211542194698] tot_loss[loss=2.292, over 5424101.23 frames. , ppl: 9.896862840918574], batch size: 70
+2022-12-12 02:20:33,642 INFO [train.py:421] (2/8) Epoch 6, batch 25400, loss[loss=2.517, over 1050.00 frames. , ppl: 12.385543527345888] tot_loss[loss=2.292, over 5413598.09 frames. , ppl: 9.895576286395373], batch size: 70
+2022-12-12 02:22:12,282 INFO [train.py:421] (2/8) Epoch 6, batch 25600, loss[loss=2.191, over 8540.00 frames. , ppl: 8.940827995510867] tot_loss[loss=2.292, over 5447047.18 frames. , ppl: 9.89081961250564], batch size: 70
+2022-12-12 02:23:50,882 INFO [train.py:421] (2/8) Epoch 6, batch 25800, loss[loss=2.312, over 1750.00 frames. , ppl: 10.097319067763419] tot_loss[loss=2.293, over 5417171.77 frames. , ppl: 9.902862889087158], batch size: 70
+2022-12-12 02:25:30,634 INFO [train.py:421] (2/8) Epoch 6, batch 26000, loss[loss=2.251, over 5670.00 frames. , ppl: 9.495952396507576] tot_loss[loss=2.293, over 5436979.71 frames. , ppl: 9.901104173552444], batch size: 70
+2022-12-12 02:25:30,634 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:25:31,395 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.806832849432856
+2022-12-12 02:27:11,364 INFO [train.py:421] (2/8) Epoch 6, batch 26200, loss[loss=2.195, over 3430.00 frames. , ppl: 8.984255415264537] tot_loss[loss=2.292, over 5437802.81 frames. , ppl: 9.897413428558718], batch size: 70
+2022-12-12 02:28:48,580 INFO [train.py:421] (2/8) Epoch 6, batch 26400, loss[loss=3.19, over 490.00 frames. , ppl: 24.278873886455127] tot_loss[loss=2.293, over 5390956.96 frames. , ppl: 9.90692865231438], batch size: 70
+2022-12-12 02:30:30,699 INFO [train.py:421] (2/8) Epoch 6, batch 26600, loss[loss=2.304, over 1470.00 frames. , ppl: 10.012262449649448] tot_loss[loss=2.292, over 5418016.32 frames. , ppl: 9.899182717513153], batch size: 70
+2022-12-12 02:32:10,389 INFO [train.py:421] (2/8) Epoch 6, batch 26800, loss[loss=2.351, over 1890.00 frames. , ppl: 10.49402775323148] tot_loss[loss=2.291, over 5446991.86 frames. , ppl: 9.884281316693865], batch size: 70
+2022-12-12 02:33:50,898 INFO [train.py:421] (2/8) Epoch 6, batch 27000, loss[loss=2.294, over 2730.00 frames. , ppl: 9.91648195880728] tot_loss[loss=2.29, over 5500813.82 frames. , ppl: 9.870904658956936], batch size: 70
+2022-12-12 02:33:50,899 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:33:51,645 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.810623407744394
+2022-12-12 02:35:28,234 INFO [train.py:421] (2/8) Epoch 6, batch 27200, loss[loss=2.179, over 6860.00 frames. , ppl: 8.836200661202415] tot_loss[loss=2.29, over 5481822.12 frames. , ppl: 9.878677624691905], batch size: 70
+2022-12-12 02:37:07,150 INFO [train.py:421] (2/8) Epoch 6, batch 27400, loss[loss=2.19, over 5180.00 frames. , ppl: 8.935895706745505] tot_loss[loss=2.291, over 5475336.87 frames. , ppl: 9.884677002475144], batch size: 70
+2022-12-12 02:38:49,954 INFO [train.py:421] (2/8) Epoch 6, batch 27600, loss[loss=3.746, over 420.00 frames. , ppl: 42.34816863981927] tot_loss[loss=2.29, over 5506607.66 frames. , ppl: 9.878674908010591], batch size: 70
+2022-12-12 02:40:23,826 INFO [train.py:421] (2/8) Epoch 6, batch 27800, loss[loss=2.363, over 1540.00 frames. , ppl: 10.620598251319118] tot_loss[loss=2.291, over 5504136.68 frames. , ppl: 9.88875897675691], batch size: 70
+2022-12-12 02:42:02,154 INFO [train.py:421] (2/8) Epoch 6, batch 28000, loss[loss=2.126, over 6370.00 frames. , ppl: 8.38075564249991] tot_loss[loss=2.291, over 5504910.37 frames. , ppl: 9.883572463656463], batch size: 70
+2022-12-12 02:42:02,154 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:42:02,898 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.819234910695574
+2022-12-12 02:43:40,823 INFO [train.py:421] (2/8) Epoch 6, batch 28200, loss[loss=2.232, over 4410.00 frames. , ppl: 9.314913418689464] tot_loss[loss=2.292, over 5481054.69 frames. , ppl: 9.890288370005237], batch size: 70
+2022-12-12 02:45:19,070 INFO [train.py:421] (2/8) Epoch 6, batch 28400, loss[loss=2.419, over 980.00 frames. , ppl: 11.229013018043984] tot_loss[loss=2.293, over 5432225.29 frames. , ppl: 9.900175474401967], batch size: 70
+2022-12-12 02:46:59,635 INFO [train.py:421] (2/8) Epoch 6, batch 28600, loss[loss=2.526, over 1120.00 frames. , ppl: 12.499101751894717] tot_loss[loss=2.292, over 5442936.95 frames. , ppl: 9.89378685264495], batch size: 70
+2022-12-12 02:48:38,764 INFO [train.py:421] (2/8) Epoch 6, batch 28800, loss[loss=2.247, over 3080.00 frames. , ppl: 9.457344275823052] tot_loss[loss=2.293, over 5410984.36 frames. , ppl: 9.899923556392087], batch size: 70
+2022-12-12 02:50:20,987 INFO [train.py:421] (2/8) Epoch 6, batch 29000, loss[loss=2.438, over 2310.00 frames. , ppl: 11.44559389267074] tot_loss[loss=2.291, over 5452108.47 frames. , ppl: 9.884576738262354], batch size: 70
+2022-12-12 02:50:20,987 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:50:21,747 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.796046998105865
+2022-12-12 02:51:59,584 INFO [train.py:421] (2/8) Epoch 6, batch 29200, loss[loss=2.241, over 2170.00 frames. , ppl: 9.399440832599229] tot_loss[loss=2.29, over 5483779.59 frames. , ppl: 9.879444159821423], batch size: 70
+2022-12-12 02:53:39,007 INFO [train.py:421] (2/8) Epoch 6, batch 29400, loss[loss=2.197, over 6160.00 frames. , ppl: 8.998096744440229] tot_loss[loss=2.291, over 5488537.41 frames. , ppl: 9.880023600696006], batch size: 70
+2022-12-12 02:55:17,412 INFO [train.py:421] (2/8) Epoch 6, batch 29600, loss[loss=2.31, over 2310.00 frames. , ppl: 10.079236777815662] tot_loss[loss=2.292, over 5462797.75 frames. , ppl: 9.890020327759558], batch size: 70
+2022-12-12 02:57:01,348 INFO [train.py:421] (2/8) Epoch 6, batch 29800, loss[loss=2.255, over 2310.00 frames. , ppl: 9.53414371357433] tot_loss[loss=2.291, over 5479033.61 frames. , ppl: 9.880708555251086], batch size: 70
+2022-12-12 02:58:45,206 INFO [train.py:421] (2/8) Epoch 6, batch 30000, loss[loss=2.359, over 1750.00 frames. , ppl: 10.582653928093757] tot_loss[loss=2.29, over 5493857.56 frames. , ppl: 9.879269255135107], batch size: 70
+2022-12-12 02:58:45,207 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 02:58:45,956 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.793146201640242
+2022-12-12 03:00:24,700 INFO [train.py:421] (2/8) Epoch 6, batch 30200, loss[loss=2.251, over 1120.00 frames. , ppl: 9.498086425952414] tot_loss[loss=2.289, over 5523353.56 frames. , ppl: 9.868086382392825], batch size: 70
+2022-12-12 03:02:01,496 INFO [train.py:421] (2/8) Epoch 6, batch 30400, loss[loss=2.295, over 2100.00 frames. , ppl: 9.919775797079108] tot_loss[loss=2.291, over 5477756.93 frames. , ppl: 9.880689192567168], batch size: 70
+2022-12-12 03:03:40,077 INFO [train.py:421] (2/8) Epoch 6, batch 30600, loss[loss=2.409, over 1820.00 frames. , ppl: 11.119878161626819] tot_loss[loss=2.293, over 5410585.10 frames. , ppl: 9.900410447587886], batch size: 70
+2022-12-12 03:05:19,243 INFO [train.py:421] (2/8) Epoch 6, batch 30800, loss[loss=2.427, over 1820.00 frames. , ppl: 11.324678089790284] tot_loss[loss=2.294, over 5373905.66 frames. , ppl: 9.911638828757766], batch size: 70
+2022-12-12 03:07:01,536 INFO [train.py:421] (2/8) Epoch 6, batch 31000, loss[loss=2.285, over 2310.00 frames. , ppl: 9.823088748749258] tot_loss[loss=2.294, over 5358359.26 frames. , ppl: 9.919169623571689], batch size: 70
+2022-12-12 03:07:01,537 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:07:02,295 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.792647600527868
+2022-12-12 03:08:42,302 INFO [train.py:421] (2/8) Epoch 6, batch 31200, loss[loss=2.362, over 2520.00 frames. , ppl: 10.617223315337322] tot_loss[loss=2.294, over 5392522.44 frames. , ppl: 9.912351684261376], batch size: 70
+2022-12-12 03:10:23,241 INFO [train.py:421] (2/8) Epoch 6, batch 31400, loss[loss=2.256, over 2800.00 frames. , ppl: 9.548725070307423] tot_loss[loss=2.292, over 5456039.94 frames. , ppl: 9.89572708835498], batch size: 70
+2022-12-12 03:12:05,123 INFO [train.py:421] (2/8) Epoch 6, batch 31600, loss[loss=2.504, over 1330.00 frames. , ppl: 12.232072086741017] tot_loss[loss=2.29, over 5533984.42 frames. , ppl: 9.871949212868872], batch size: 70
+2022-12-12 03:13:47,714 INFO [train.py:421] (2/8) Epoch 6, batch 31800, loss[loss=2.215, over 3920.00 frames. , ppl: 9.165312251022078] tot_loss[loss=2.29, over 5519450.87 frames. , ppl: 9.873721435784892], batch size: 70
+2022-12-12 03:15:28,315 INFO [train.py:421] (2/8) Epoch 6, batch 32000, loss[loss=2.194, over 5950.00 frames. , ppl: 8.973376977251814] tot_loss[loss=2.289, over 5539190.86 frames. , ppl: 9.860387285169933], batch size: 70
+2022-12-12 03:15:28,315 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:15:29,060 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.786621463485243
+2022-12-12 03:17:08,909 INFO [train.py:421] (2/8) Epoch 6, batch 32200, loss[loss=2.452, over 1120.00 frames. , ppl: 11.616814951375577] tot_loss[loss=2.289, over 5538886.04 frames. , ppl: 9.862213443428905], batch size: 70
+2022-12-12 03:18:47,923 INFO [train.py:421] (2/8) Epoch 6, batch 32400, loss[loss=3.13, over 490.00 frames. , ppl: 22.878924040182977] tot_loss[loss=2.289, over 5558269.36 frames. , ppl: 9.863270637284309], batch size: 70
+2022-12-12 03:20:28,895 INFO [train.py:421] (2/8) Epoch 6, batch 32600, loss[loss=2.298, over 2800.00 frames. , ppl: 9.958622120419552] tot_loss[loss=2.289, over 5561874.33 frames. , ppl: 9.863336561805413], batch size: 70
+2022-12-12 03:22:12,183 INFO [train.py:421] (2/8) Epoch 6, batch 32800, loss[loss=2.573, over 1050.00 frames. , ppl: 13.104780935510286] tot_loss[loss=2.288, over 5581541.65 frames. , ppl: 9.858519938313488], batch size: 70
+2022-12-12 03:23:51,245 INFO [train.py:421] (2/8) Epoch 6, batch 33000, loss[loss=2.845, over 910.00 frames. , ppl: 17.19831836029165] tot_loss[loss=2.289, over 5544553.50 frames. , ppl: 9.863953052570292], batch size: 70
+2022-12-12 03:23:51,246 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:23:52,003 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.805240700234473
+2022-12-12 03:25:32,458 INFO [train.py:421] (2/8) Epoch 6, batch 33200, loss[loss=2.675, over 770.00 frames. , ppl: 14.51390518566646] tot_loss[loss=2.288, over 5565499.33 frames. , ppl: 9.85919468161035], batch size: 70
+2022-12-12 03:27:11,786 INFO [train.py:421] (2/8) Epoch 6, batch 33400, loss[loss=2.537, over 1190.00 frames. , ppl: 12.636158393203267] tot_loss[loss=2.288, over 5561607.62 frames. , ppl: 9.856391174788694], batch size: 70
+2022-12-12 03:28:50,124 INFO [train.py:421] (2/8) Epoch 6, batch 33600, loss[loss=2.275, over 2450.00 frames. , ppl: 9.72403063858822] tot_loss[loss=2.287, over 5588992.76 frames. , ppl: 9.84931635330045], batch size: 70
+2022-12-12 03:30:30,180 INFO [train.py:421] (2/8) Epoch 6, batch 33800, loss[loss=2.159, over 8190.00 frames. , ppl: 8.66523219311457] tot_loss[loss=2.286, over 5639702.70 frames. , ppl: 9.837437356901638], batch size: 70
+2022-12-12 03:32:08,720 INFO [train.py:421] (2/8) Epoch 6, batch 34000, loss[loss=3.511, over 420.00 frames. , ppl: 33.47025148453787] tot_loss[loss=2.287, over 5613799.64 frames. , ppl: 9.842434301650863], batch size: 70
+2022-12-12 03:32:08,720 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:32:09,480 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.78966811844687
+2022-12-12 03:33:48,835 INFO [train.py:421] (2/8) Epoch 6, batch 34200, loss[loss=2.184, over 7140.00 frames. , ppl: 8.884741220059727] tot_loss[loss=2.286, over 5628919.73 frames. , ppl: 9.834559890217667], batch size: 70
+2022-12-12 03:35:27,527 INFO [train.py:421] (2/8) Epoch 6, batch 34400, loss[loss=2.311, over 2520.00 frames. , ppl: 10.080408277074458] tot_loss[loss=2.286, over 5609585.14 frames. , ppl: 9.835784553888967], batch size: 70
+2022-12-12 03:37:07,526 INFO [train.py:421] (2/8) Epoch 6, batch 34600, loss[loss=2.368, over 1890.00 frames. , ppl: 10.673967125418487] tot_loss[loss=2.286, over 5592294.96 frames. , ppl: 9.840278817807143], batch size: 70
+2022-12-12 03:38:50,945 INFO [train.py:421] (2/8) Epoch 6, batch 34800, loss[loss=2.405, over 1400.00 frames. , ppl: 11.074808515098777] tot_loss[loss=2.285, over 5630694.33 frames. , ppl: 9.829389912039847], batch size: 70
+2022-12-12 03:40:30,319 INFO [train.py:421] (2/8) Epoch 6, batch 35000, loss[loss=2.375, over 1540.00 frames. , ppl: 10.754543555236891] tot_loss[loss=2.284, over 5648117.68 frames. , ppl: 9.818937140361005], batch size: 70
+2022-12-12 03:40:30,319 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:40:31,078 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.778649309689856
+2022-12-12 03:42:08,421 INFO [train.py:421] (2/8) Epoch 6, batch 35200, loss[loss=2.214, over 5460.00 frames. , ppl: 9.148801325400875] tot_loss[loss=2.285, over 5619310.49 frames. , ppl: 9.82894060225542], batch size: 70
+2022-12-12 03:43:48,389 INFO [train.py:421] (2/8) Epoch 6, batch 35400, loss[loss=2.204, over 3780.00 frames. , ppl: 9.059922102622714] tot_loss[loss=2.286, over 5598971.28 frames. , ppl: 9.832728267116616], batch size: 70
+2022-12-12 03:45:30,989 INFO [train.py:421] (2/8) Epoch 6, batch 35600, loss[loss=2.262, over 1610.00 frames. , ppl: 9.599878506196804] tot_loss[loss=2.287, over 5561503.78 frames. , ppl: 9.849159825822591], batch size: 70
+2022-12-12 03:47:10,357 INFO [train.py:421] (2/8) Epoch 6, batch 35800, loss[loss=2.265, over 3080.00 frames. , ppl: 9.629357281888078] tot_loss[loss=2.286, over 5596958.43 frames. , ppl: 9.840083049678329], batch size: 70
+2022-12-12 03:48:50,030 INFO [train.py:421] (2/8) Epoch 6, batch 36000, loss[loss=2.299, over 3220.00 frames. , ppl: 9.96696748040108] tot_loss[loss=2.287, over 5553411.80 frames. , ppl: 9.849199646866833], batch size: 70
+2022-12-12 03:48:50,030 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:48:50,790 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.788203345702467
+2022-12-12 03:50:31,022 INFO [train.py:421] (2/8) Epoch 6, batch 36200, loss[loss=2.213, over 4130.00 frames. , ppl: 9.14730284476509] tot_loss[loss=2.287, over 5562417.56 frames. , ppl: 9.845546034692878], batch size: 70
+2022-12-12 03:52:13,148 INFO [train.py:421] (2/8) Epoch 6, batch 36400, loss[loss=2.746, over 770.00 frames. , ppl: 15.581360501011297] tot_loss[loss=2.287, over 5567008.57 frames. , ppl: 9.845810044250006], batch size: 70
+2022-12-12 03:53:51,588 INFO [train.py:421] (2/8) Epoch 6, batch 36600, loss[loss=2.248, over 4410.00 frames. , ppl: 9.473451462468505] tot_loss[loss=2.288, over 5547110.10 frames. , ppl: 9.853486192880325], batch size: 70
+2022-12-12 03:55:34,328 INFO [train.py:421] (2/8) Epoch 6, batch 36800, loss[loss=2.339, over 1960.00 frames. , ppl: 10.370384829708525] tot_loss[loss=2.288, over 5558773.06 frames. , ppl: 9.85297824891939], batch size: 70
+2022-12-12 03:57:15,030 INFO [train.py:421] (2/8) Epoch 6, batch 37000, loss[loss=2.358, over 2100.00 frames. , ppl: 10.574497150176905] tot_loss[loss=2.288, over 5563215.95 frames. , ppl: 9.85588819533942], batch size: 70
+2022-12-12 03:57:15,031 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 03:57:15,781 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.776730360844793
+2022-12-12 03:58:58,748 INFO [train.py:421] (2/8) Epoch 6, batch 37200, loss[loss=2.246, over 5460.00 frames. , ppl: 9.445567194005132] tot_loss[loss=2.288, over 5564722.66 frames. , ppl: 9.854276453927953], batch size: 70
+2022-12-12 04:00:39,203 INFO [train.py:421] (2/8) Epoch 6, batch 37400, loss[loss=2.377, over 2380.00 frames. , ppl: 10.771709123981625] tot_loss[loss=2.287, over 5590237.37 frames. , ppl: 9.843137875907766], batch size: 70
+2022-12-12 04:02:19,599 INFO [train.py:421] (2/8) Epoch 6, batch 37600, loss[loss=2.588, over 980.00 frames. , ppl: 13.304116270217879] tot_loss[loss=2.287, over 5606566.55 frames. , ppl: 9.843096100005607], batch size: 70
+2022-12-12 04:04:01,675 INFO [train.py:421] (2/8) Epoch 6, batch 37800, loss[loss=2.218, over 4970.00 frames. , ppl: 9.191706467061325] tot_loss[loss=2.288, over 5575733.44 frames. , ppl: 9.852209887162775], batch size: 70
+2022-12-12 04:05:40,692 INFO [train.py:421] (2/8) Epoch 6, batch 38000, loss[loss=2.577, over 1050.00 frames. , ppl: 13.160039696398929] tot_loss[loss=2.288, over 5565237.55 frames. , ppl: 9.857122368090623], batch size: 70
+2022-12-12 04:05:40,693 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:05:41,464 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.785526464019718
+2022-12-12 04:07:23,630 INFO [train.py:421] (2/8) Epoch 6, batch 38200, loss[loss=2.205, over 2940.00 frames. , ppl: 9.06862991609116] tot_loss[loss=2.289, over 5553592.15 frames. , ppl: 9.863949734975744], batch size: 70
+2022-12-12 04:08:58,703 INFO [train.py:421] (2/8) Epoch 6, batch 38400, loss[loss=2.457, over 910.00 frames. , ppl: 11.672663267522184] tot_loss[loss=2.288, over 5553917.74 frames. , ppl: 9.858002025966456], batch size: 70
+2022-12-12 04:10:38,731 INFO [train.py:421] (2/8) Epoch 6, batch 38600, loss[loss=2.367, over 2660.00 frames. , ppl: 10.668487839049755] tot_loss[loss=2.288, over 5576753.51 frames. , ppl: 9.852616334341448], batch size: 70
+2022-12-12 04:12:18,446 INFO [train.py:421] (2/8) Epoch 6, batch 38800, loss[loss=2.192, over 5460.00 frames. , ppl: 8.955858249393984] tot_loss[loss=2.288, over 5574260.11 frames. , ppl: 9.853951703799549], batch size: 70
+2022-12-12 04:13:53,332 INFO [train.py:421] (2/8) Epoch 6, batch 39000, loss[loss=2.452, over 1750.00 frames. , ppl: 11.60861169084875] tot_loss[loss=2.289, over 5556842.38 frames. , ppl: 9.860996122843618], batch size: 70
+2022-12-12 04:13:53,332 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:13:54,080 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.790575199383582
+2022-12-12 04:15:37,674 INFO [train.py:421] (2/8) Epoch 6, batch 39200, loss[loss=2.473, over 1120.00 frames. , ppl: 11.857184371677134] tot_loss[loss=2.289, over 5549247.07 frames. , ppl: 9.861409112302562], batch size: 70
+2022-12-12 04:17:15,042 INFO [train.py:421] (2/8) Epoch 6, batch 39400, loss[loss=2.275, over 2240.00 frames. , ppl: 9.724740875253412] tot_loss[loss=2.288, over 5536910.03 frames. , ppl: 9.856812859177388], batch size: 70
+2022-12-12 04:18:56,787 INFO [train.py:421] (2/8) Epoch 6, batch 39600, loss[loss=2.418, over 2450.00 frames. , ppl: 11.221789082548831] tot_loss[loss=2.288, over 5563824.83 frames. , ppl: 9.85119135046772], batch size: 70
+2022-12-12 04:20:36,866 INFO [train.py:421] (2/8) Epoch 6, batch 39800, loss[loss=2.236, over 3010.00 frames. , ppl: 9.359932790354259] tot_loss[loss=2.288, over 5528243.02 frames. , ppl: 9.856369066054944], batch size: 70
+2022-12-12 04:22:14,667 INFO [train.py:421] (2/8) Epoch 6, batch 40000, loss[loss=2.407, over 1260.00 frames. , ppl: 11.10394213107575] tot_loss[loss=2.288, over 5543409.39 frames. , ppl: 9.850428800257166], batch size: 70
+2022-12-12 04:22:14,667 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:22:15,413 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.796226785877156
+2022-12-12 04:23:58,047 INFO [train.py:421] (2/8) Epoch 6, batch 40200, loss[loss=3.493, over 420.00 frames. , ppl: 32.88898846304666] tot_loss[loss=2.286, over 5583860.66 frames. , ppl: 9.836410592736659], batch size: 70
+2022-12-12 04:25:35,491 INFO [train.py:421] (2/8) Epoch 6, batch 40400, loss[loss=2.476, over 1260.00 frames. , ppl: 11.891654597989595] tot_loss[loss=2.287, over 5572558.08 frames. , ppl: 9.842462197779975], batch size: 70
+2022-12-12 04:27:15,613 INFO [train.py:421] (2/8) Epoch 6, batch 40600, loss[loss=2.525, over 980.00 frames. , ppl: 12.493111040626943] tot_loss[loss=2.289, over 5508890.76 frames. , ppl: 9.863751847905032], batch size: 70
+2022-12-12 04:28:54,808 INFO [train.py:421] (2/8) Epoch 6, batch 40800, loss[loss=2.28, over 2380.00 frames. , ppl: 9.7746884662208] tot_loss[loss=2.288, over 5528794.23 frames. , ppl: 9.857366324349474], batch size: 70
+2022-12-12 04:30:38,979 INFO [train.py:421] (2/8) Epoch 6, batch 41000, loss[loss=2.822, over 630.00 frames. , ppl: 16.81162974742163] tot_loss[loss=2.287, over 5576619.06 frames. , ppl: 9.842629687849858], batch size: 70
+2022-12-12 04:30:38,979 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:30:39,735 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.77442840863681
+2022-12-12 04:32:21,591 INFO [train.py:421] (2/8) Epoch 6, batch 41200, loss[loss=2.177, over 7910.00 frames. , ppl: 8.817460129837603] tot_loss[loss=2.286, over 5593214.74 frames. , ppl: 9.839411990827161], batch size: 70
+2022-12-12 04:34:01,967 INFO [train.py:421] (2/8) Epoch 6, batch 41400, loss[loss=2.424, over 1680.00 frames. , ppl: 11.293870746133383] tot_loss[loss=2.285, over 5666666.21 frames. , ppl: 9.826453059051827], batch size: 70
+2022-12-12 04:35:43,802 INFO [train.py:421] (2/8) Epoch 6, batch 41600, loss[loss=2.327, over 1890.00 frames. , ppl: 10.249247016400993] tot_loss[loss=2.287, over 5618644.57 frames. , ppl: 9.84094429298711], batch size: 70
+2022-12-12 04:37:22,140 INFO [train.py:421] (2/8) Epoch 6, batch 41800, loss[loss=2.386, over 2030.00 frames. , ppl: 10.866406318803685] tot_loss[loss=2.289, over 5567966.29 frames. , ppl: 9.8602070905214], batch size: 70
+2022-12-12 04:39:01,080 INFO [train.py:421] (2/8) Epoch 6, batch 42000, loss[loss=2.171, over 8960.00 frames. , ppl: 8.770760843151132] tot_loss[loss=2.288, over 5586166.33 frames. , ppl: 9.853251455024514], batch size: 70
+2022-12-12 04:39:01,080 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:39:01,825 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.774656988070976
+2022-12-12 04:40:40,482 INFO [train.py:421] (2/8) Epoch 6, batch 42200, loss[loss=3.331, over 490.00 frames. , ppl: 27.96539890657477] tot_loss[loss=2.289, over 5521971.30 frames. , ppl: 9.868680606396055], batch size: 70
+2022-12-12 04:42:19,271 INFO [train.py:421] (2/8) Epoch 6, batch 42400, loss[loss=2.506, over 980.00 frames. , ppl: 12.256631575993799] tot_loss[loss=2.29, over 5506245.04 frames. , ppl: 9.871993993245608], batch size: 70
+2022-12-12 04:43:58,795 INFO [train.py:421] (2/8) Epoch 6, batch 42600, loss[loss=2.201, over 6090.00 frames. , ppl: 9.031996077883878] tot_loss[loss=2.29, over 5489345.35 frames. , ppl: 9.877260265176922], batch size: 70
+2022-12-12 04:45:37,699 INFO [train.py:421] (2/8) Epoch 6, batch 42800, loss[loss=2.305, over 2170.00 frames. , ppl: 10.023783984014443] tot_loss[loss=2.29, over 5482897.06 frames. , ppl: 9.873768808919818], batch size: 70
+2022-12-12 04:47:16,068 INFO [train.py:421] (2/8) Epoch 6, batch 43000, loss[loss=2.17, over 5250.00 frames. , ppl: 8.762586040004562] tot_loss[loss=2.289, over 5493880.19 frames. , ppl: 9.869288023997795], batch size: 70
+2022-12-12 04:47:16,068 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:47:16,826 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.774992633382912
+2022-12-12 04:48:53,986 INFO [train.py:421] (2/8) Epoch 6, batch 43200, loss[loss=2.292, over 1750.00 frames. , ppl: 9.893513345612616] tot_loss[loss=2.289, over 5502965.41 frames. , ppl: 9.864059200828452], batch size: 70
+2022-12-12 04:50:34,705 INFO [train.py:421] (2/8) Epoch 6, batch 43400, loss[loss=2.493, over 1190.00 frames. , ppl: 12.101788119968587] tot_loss[loss=2.288, over 5506507.89 frames. , ppl: 9.859963526155068], batch size: 70
+2022-12-12 04:52:11,333 INFO [train.py:421] (2/8) Epoch 6, batch 43600, loss[loss=2.236, over 4760.00 frames. , ppl: 9.35231218189974] tot_loss[loss=2.288, over 5519639.59 frames. , ppl: 9.856483812621935], batch size: 70
+2022-12-12 04:53:48,620 INFO [train.py:421] (2/8) Epoch 6, batch 43800, loss[loss=2.592, over 1050.00 frames. , ppl: 13.353412843832487] tot_loss[loss=2.288, over 5526930.77 frames. , ppl: 9.851537317044388], batch size: 70
+2022-12-12 04:55:27,839 INFO [train.py:421] (2/8) Epoch 6, batch 44000, loss[loss=2.553, over 700.00 frames. , ppl: 12.846715872088984] tot_loss[loss=2.287, over 5550389.66 frames. , ppl: 9.841238640243525], batch size: 70
+2022-12-12 04:55:27,839 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 04:55:28,598 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.775564124978258
+2022-12-12 04:57:06,980 INFO [train.py:421] (2/8) Epoch 6, batch 44200, loss[loss=2.383, over 1680.00 frames. , ppl: 10.835621010428643] tot_loss[loss=2.286, over 5542551.55 frames. , ppl: 9.839800257308662], batch size: 70
+2022-12-12 04:58:47,015 INFO [train.py:421] (2/8) Epoch 6, batch 44400, loss[loss=2.282, over 4130.00 frames. , ppl: 9.796740159483777] tot_loss[loss=2.287, over 5524886.45 frames. , ppl: 9.844163903663944], batch size: 70
+2022-12-12 05:00:25,090 INFO [train.py:421] (2/8) Epoch 6, batch 44600, loss[loss=2.395, over 1050.00 frames. , ppl: 10.970845558053174] tot_loss[loss=2.286, over 5571080.99 frames. , ppl: 9.836477549326704], batch size: 70
+2022-12-12 05:02:08,399 INFO [train.py:421] (2/8) Epoch 6, batch 44800, loss[loss=2.625, over 980.00 frames. , ppl: 13.810291050873989] tot_loss[loss=2.285, over 5619048.99 frames. , ppl: 9.828612003712125], batch size: 70
+2022-12-12 05:03:45,269 INFO [train.py:421] (2/8) Epoch 6, batch 45000, loss[loss=2.38, over 2170.00 frames. , ppl: 10.801762620299618] tot_loss[loss=2.287, over 5571160.26 frames. , ppl: 9.843977527098273], batch size: 70
+2022-12-12 05:03:45,269 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:03:46,014 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.792734564010422
+2022-12-12 05:05:26,479 INFO [train.py:421] (2/8) Epoch 6, batch 45200, loss[loss=2.252, over 2450.00 frames. , ppl: 9.50979220790985] tot_loss[loss=2.288, over 5519043.25 frames. , ppl: 9.859858492443385], batch size: 70
+2022-12-12 05:07:06,357 INFO [train.py:421] (2/8) Epoch 6, batch 45400, loss[loss=2.25, over 2940.00 frames. , ppl: 9.489141502958402] tot_loss[loss=2.288, over 5534599.39 frames. , ppl: 9.859463425448675], batch size: 70
+2022-12-12 05:08:44,593 INFO [train.py:421] (2/8) Epoch 6, batch 45600, loss[loss=2.5, over 1610.00 frames. , ppl: 12.185039882001716] tot_loss[loss=2.289, over 5535272.09 frames. , ppl: 9.866950964592242], batch size: 70
+2022-12-12 05:10:22,563 INFO [train.py:421] (2/8) Epoch 6, batch 45800, loss[loss=2.398, over 1330.00 frames. , ppl: 10.995903547832002] tot_loss[loss=2.289, over 5539071.61 frames. , ppl: 9.864144500398151], batch size: 70
+2022-12-12 05:12:00,167 INFO [train.py:421] (2/8) Epoch 6, batch 46000, loss[loss=2.224, over 3150.00 frames. , ppl: 9.240482103720757] tot_loss[loss=2.289, over 5525811.88 frames. , ppl: 9.867384709034491], batch size: 70
+2022-12-12 05:12:00,167 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:12:00,928 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.7772122333999
+2022-12-12 05:13:38,801 INFO [train.py:421] (2/8) Epoch 6, batch 46200, loss[loss=2.277, over 3640.00 frames. , ppl: 9.748210994959184] tot_loss[loss=2.29, over 5519452.40 frames. , ppl: 9.870441082523659], batch size: 70
+2022-12-12 05:15:18,383 INFO [train.py:421] (2/8) Epoch 6, batch 46400, loss[loss=2.169, over 5250.00 frames. , ppl: 8.753072327689825] tot_loss[loss=2.289, over 5555428.31 frames. , ppl: 9.86353019331463], batch size: 70
+2022-12-12 05:16:59,461 INFO [train.py:421] (2/8) Epoch 6, batch 46600, loss[loss=2.625, over 700.00 frames. , ppl: 13.811465674665797] tot_loss[loss=2.29, over 5531762.19 frames. , ppl: 9.871487483623998], batch size: 70
+2022-12-12 05:18:44,137 INFO [train.py:421] (2/8) Epoch 6, batch 46800, loss[loss=2.495, over 1120.00 frames. , ppl: 12.116258794292332] tot_loss[loss=2.288, over 5574688.68 frames. , ppl: 9.855992940351006], batch size: 70
+2022-12-12 05:20:24,469 INFO [train.py:421] (2/8) Epoch 6, batch 47000, loss[loss=2.349, over 2240.00 frames. , ppl: 10.474688714728815] tot_loss[loss=2.288, over 5553202.22 frames. , ppl: 9.854721335759553], batch size: 70
+2022-12-12 05:20:24,470 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:20:25,228 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.782431870183903
+2022-12-12 05:22:04,498 INFO [train.py:421] (2/8) Epoch 6, batch 47200, loss[loss=2.344, over 2100.00 frames. , ppl: 10.422802986179052] tot_loss[loss=2.288, over 5555164.49 frames. , ppl: 9.85770659655354], batch size: 70
+2022-12-12 05:23:42,961 INFO [train.py:421] (2/8) Epoch 6, batch 47400, loss[loss=2.843, over 700.00 frames. , ppl: 17.1602512918918] tot_loss[loss=2.288, over 5581679.34 frames. , ppl: 9.851438123325297], batch size: 70
+2022-12-12 05:25:19,149 INFO [train.py:421] (2/8) Epoch 6, batch 47600, loss[loss=2.161, over 7280.00 frames. , ppl: 8.68092524022972] tot_loss[loss=2.288, over 5546044.83 frames. , ppl: 9.856807717237672], batch size: 70
+2022-12-12 05:26:59,790 INFO [train.py:421] (2/8) Epoch 6, batch 47800, loss[loss=2.26, over 3850.00 frames. , ppl: 9.579591919177053] tot_loss[loss=2.289, over 5546063.72 frames. , ppl: 9.861638469081061], batch size: 70
+2022-12-12 05:28:40,367 INFO [train.py:421] (2/8) Epoch 6, batch 48000, loss[loss=2.237, over 4480.00 frames. , ppl: 9.366170937021586] tot_loss[loss=2.289, over 5530618.16 frames. , ppl: 9.862081690293275], batch size: 70
+2022-12-12 05:28:40,368 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:28:41,113 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.767148491772168
+2022-12-12 05:30:22,489 INFO [train.py:421] (2/8) Epoch 6, batch 48200, loss[loss=2.371, over 1330.00 frames. , ppl: 10.713133557947188] tot_loss[loss=2.288, over 5551424.35 frames. , ppl: 9.856404738485484], batch size: 70
+2022-12-12 05:31:59,718 INFO [train.py:421] (2/8) Epoch 6, batch 48400, loss[loss=2.371, over 1260.00 frames. , ppl: 10.707939084072946] tot_loss[loss=2.288, over 5547998.30 frames. , ppl: 9.850701761529027], batch size: 70
+2022-12-12 05:33:38,676 INFO [train.py:421] (2/8) Epoch 6, batch 48600, loss[loss=2.221, over 4830.00 frames. , ppl: 9.218823367420558] tot_loss[loss=2.288, over 5542327.83 frames. , ppl: 9.854726542853863], batch size: 70
+2022-12-12 05:35:20,096 INFO [train.py:421] (2/8) Epoch 6, batch 48800, loss[loss=2.9, over 700.00 frames. , ppl: 18.172706555376617] tot_loss[loss=2.288, over 5550582.46 frames. , ppl: 9.856578612095712], batch size: 70
+2022-12-12 05:36:57,680 INFO [train.py:421] (2/8) Epoch 6, batch 49000, loss[loss=2.549, over 1190.00 frames. , ppl: 12.795495462093337] tot_loss[loss=2.289, over 5562876.77 frames. , ppl: 9.865601140720948], batch size: 70
+2022-12-12 05:36:57,681 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:36:58,441 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.767533031836354
+2022-12-12 05:38:37,512 INFO [train.py:421] (2/8) Epoch 6, batch 49200, loss[loss=2.594, over 980.00 frames. , ppl: 13.386210068788197] tot_loss[loss=2.288, over 5561911.10 frames. , ppl: 9.858813707274049], batch size: 70
+2022-12-12 05:40:16,674 INFO [train.py:421] (2/8) Epoch 6, batch 49400, loss[loss=4.043, over 350.00 frames. , ppl: 57.00460051952303] tot_loss[loss=2.288, over 5587574.27 frames. , ppl: 9.850778402928073], batch size: 70
+2022-12-12 05:41:57,409 INFO [train.py:421] (2/8) Epoch 6, batch 49600, loss[loss=2.284, over 2660.00 frames. , ppl: 9.811828132561882] tot_loss[loss=2.286, over 5635238.56 frames. , ppl: 9.838435542768316], batch size: 70
+2022-12-12 05:43:39,729 INFO [train.py:421] (2/8) Epoch 6, batch 49800, loss[loss=2.414, over 1330.00 frames. , ppl: 11.178665141079069] tot_loss[loss=2.287, over 5584892.41 frames. , ppl: 9.849328895760491], batch size: 70
+2022-12-12 05:45:15,853 INFO [train.py:421] (2/8) Epoch 6, batch 50000, loss[loss=2.42, over 1260.00 frames. , ppl: 11.25059489653609] tot_loss[loss=2.288, over 5571057.29 frames. , ppl: 9.850471863844902], batch size: 70
+2022-12-12 05:45:15,853 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:45:16,606 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.785662608076919
+2022-12-12 05:46:53,730 INFO [train.py:421] (2/8) Epoch 6, batch 50200, loss[loss=2.308, over 1960.00 frames. , ppl: 10.05111380006104] tot_loss[loss=2.288, over 5543164.65 frames. , ppl: 9.855017553460156], batch size: 70
+2022-12-12 05:48:35,502 INFO [train.py:421] (2/8) Epoch 6, batch 50400, loss[loss=2.265, over 3780.00 frames. , ppl: 9.63218761871062] tot_loss[loss=2.289, over 5523048.02 frames. , ppl: 9.867908850974041], batch size: 70
+2022-12-12 05:50:15,396 INFO [train.py:421] (2/8) Epoch 6, batch 50600, loss[loss=2.246, over 4690.00 frames. , ppl: 9.451945160077988] tot_loss[loss=2.289, over 5515807.33 frames. , ppl: 9.868143854517289], batch size: 70
+2022-12-12 05:51:57,533 INFO [train.py:421] (2/8) Epoch 6, batch 50800, loss[loss=2.378, over 1260.00 frames. , ppl: 10.783998160105211] tot_loss[loss=2.29, over 5492515.48 frames. , ppl: 9.877538151693548], batch size: 70
+2022-12-12 05:53:42,722 INFO [train.py:421] (2/8) Epoch 6, batch 51000, loss[loss=2.338, over 2730.00 frames. , ppl: 10.362556528029494] tot_loss[loss=2.29, over 5503984.63 frames. , ppl: 9.878926794623204], batch size: 70
+2022-12-12 05:53:42,723 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 05:53:43,469 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.76539078711788
+2022-12-12 05:55:23,427 INFO [train.py:421] (2/8) Epoch 6, batch 51200, loss[loss=2.251, over 3080.00 frames. , ppl: 9.497788666790857] tot_loss[loss=2.29, over 5499149.94 frames. , ppl: 9.87132142988135], batch size: 70
+2022-12-12 05:57:03,660 INFO [train.py:421] (2/8) Epoch 6, batch 51400, loss[loss=2.305, over 2940.00 frames. , ppl: 10.023975810393074] tot_loss[loss=2.29, over 5493361.99 frames. , ppl: 9.877602026548313], batch size: 70
+2022-12-12 05:58:44,079 INFO [train.py:421] (2/8) Epoch 6, batch 51600, loss[loss=2.856, over 700.00 frames. , ppl: 17.399196680865924] tot_loss[loss=2.29, over 5502443.92 frames. , ppl: 9.873508767637663], batch size: 70
+2022-12-12 06:00:26,236 INFO [train.py:421] (2/8) Epoch 6, batch 51800, loss[loss=2.208, over 4970.00 frames. , ppl: 9.093806782956275] tot_loss[loss=2.289, over 5519308.54 frames. , ppl: 9.86313875511843], batch size: 70
+2022-12-12 06:02:05,611 INFO [train.py:421] (2/8) Epoch 6, batch 52000, loss[loss=2.217, over 4620.00 frames. , ppl: 9.17676068905592] tot_loss[loss=2.288, over 5546971.49 frames. , ppl: 9.857890387385579], batch size: 70
+2022-12-12 06:02:05,611 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:02:06,356 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758165262568566
+2022-12-12 06:03:45,494 INFO [train.py:421] (2/8) Epoch 6, batch 52200, loss[loss=2.406, over 1540.00 frames. , ppl: 11.094917994872647] tot_loss[loss=2.288, over 5532115.83 frames. , ppl: 9.860047332694542], batch size: 70
+2022-12-12 06:05:23,973 INFO [train.py:421] (2/8) Epoch 6, batch 52400, loss[loss=2.285, over 2450.00 frames. , ppl: 9.821754856315401] tot_loss[loss=2.288, over 5515842.90 frames. , ppl: 9.853899952720282], batch size: 70
+2022-12-12 06:07:01,650 INFO [train.py:421] (2/8) Epoch 6, batch 52600, loss[loss=2.286, over 1540.00 frames. , ppl: 9.830866299368552] tot_loss[loss=2.29, over 5455599.18 frames. , ppl: 9.873953344009148], batch size: 70
+2022-12-12 06:08:40,079 INFO [train.py:421] (2/8) Epoch 6, batch 52800, loss[loss=2.24, over 2870.00 frames. , ppl: 9.393974710225587] tot_loss[loss=2.289, over 5496193.38 frames. , ppl: 9.86297659508991], batch size: 70
+2022-12-12 06:10:20,319 INFO [train.py:421] (2/8) Epoch 6, batch 53000, loss[loss=2.676, over 770.00 frames. , ppl: 14.519612616674719] tot_loss[loss=2.289, over 5496888.41 frames. , ppl: 9.864484091346315], batch size: 70
+2022-12-12 06:10:20,320 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:10:21,067 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764269259359716
+2022-12-12 06:11:59,976 INFO [train.py:421] (2/8) Epoch 6, batch 53200, loss[loss=2.3, over 2240.00 frames. , ppl: 9.969378631708901] tot_loss[loss=2.29, over 5469959.10 frames. , ppl: 9.871463926144353], batch size: 70
+2022-12-12 06:13:40,686 INFO [train.py:421] (2/8) Epoch 6, batch 53400, loss[loss=2.397, over 1750.00 frames. , ppl: 10.989454211635067] tot_loss[loss=2.29, over 5479909.18 frames. , ppl: 9.873339646313301], batch size: 70
+2022-12-12 06:15:21,184 INFO [train.py:421] (2/8) Epoch 6, batch 53600, loss[loss=2.674, over 700.00 frames. , ppl: 14.499630790582914] tot_loss[loss=2.291, over 5431732.82 frames. , ppl: 9.882878481864882], batch size: 70
+2022-12-12 06:17:02,491 INFO [train.py:421] (2/8) Epoch 6, batch 53800, loss[loss=2.342, over 1610.00 frames. , ppl: 10.400238300863132] tot_loss[loss=2.289, over 5471736.82 frames. , ppl: 9.868753825475697], batch size: 70
+2022-12-12 06:18:44,476 INFO [train.py:421] (2/8) Epoch 6, batch 54000, loss[loss=2.356, over 980.00 frames. , ppl: 10.546497936081796] tot_loss[loss=2.29, over 5449173.62 frames. , ppl: 9.871107412349536], batch size: 70
+2022-12-12 06:18:44,477 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:18:45,222 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.773898935036751
+2022-12-12 06:20:25,788 INFO [train.py:421] (2/8) Epoch 6, batch 54200, loss[loss=2.628, over 770.00 frames. , ppl: 13.842563643047194] tot_loss[loss=2.29, over 5455875.43 frames. , ppl: 9.873132052139058], batch size: 70
+2022-12-12 06:22:07,259 INFO [train.py:421] (2/8) Epoch 6, batch 54400, loss[loss=2.178, over 8610.00 frames. , ppl: 8.830215700928344] tot_loss[loss=2.29, over 5471608.57 frames. , ppl: 9.877613532066768], batch size: 70
+2022-12-12 06:23:46,090 INFO [train.py:421] (2/8) Epoch 6, batch 54600, loss[loss=2.409, over 1330.00 frames. , ppl: 11.119021198997068] tot_loss[loss=2.291, over 5468135.89 frames. , ppl: 9.881919269225332], batch size: 70
+2022-12-12 06:25:25,861 INFO [train.py:421] (2/8) Epoch 6, batch 54800, loss[loss=2.313, over 2100.00 frames. , ppl: 10.101760629650038] tot_loss[loss=2.29, over 5475188.50 frames. , ppl: 9.871874311312427], batch size: 70
+2022-12-12 06:27:05,727 INFO [train.py:421] (2/8) Epoch 6, batch 55000, loss[loss=2.337, over 3220.00 frames. , ppl: 10.348525075307897] tot_loss[loss=2.29, over 5470122.22 frames. , ppl: 9.878617475634393], batch size: 70
+2022-12-12 06:27:05,727 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:27:06,473 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758666441048378
+2022-12-12 06:28:46,934 INFO [train.py:421] (2/8) Epoch 6, batch 55200, loss[loss=2.144, over 5670.00 frames. , ppl: 8.537219045619537] tot_loss[loss=2.29, over 5478426.72 frames. , ppl: 9.877878925719434], batch size: 70
+2022-12-12 06:30:30,148 INFO [train.py:421] (2/8) Epoch 6, batch 55400, loss[loss=3.018, over 560.00 frames. , ppl: 20.45540870542286] tot_loss[loss=2.291, over 5456934.91 frames. , ppl: 9.888290098986765], batch size: 70
+2022-12-12 06:32:09,447 INFO [train.py:421] (2/8) Epoch 6, batch 55600, loss[loss=2.881, over 630.00 frames. , ppl: 17.82983073157378] tot_loss[loss=2.291, over 5458943.78 frames. , ppl: 9.88478112120656], batch size: 70
+2022-12-12 06:33:49,115 INFO [train.py:421] (2/8) Epoch 6, batch 55800, loss[loss=2.189, over 3780.00 frames. , ppl: 8.926819521576352] tot_loss[loss=2.29, over 5486988.18 frames. , ppl: 9.877275765671431], batch size: 70
+2022-12-12 06:35:30,414 INFO [train.py:421] (2/8) Epoch 6, batch 56000, loss[loss=2.39, over 2030.00 frames. , ppl: 10.909707700741972] tot_loss[loss=2.291, over 5462858.04 frames. , ppl: 9.88185534640927], batch size: 70
+2022-12-12 06:35:30,414 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:35:31,174 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.763342939887206
+2022-12-12 06:37:13,841 INFO [train.py:421] (2/8) Epoch 6, batch 56200, loss[loss=2.389, over 1680.00 frames. , ppl: 10.907107213896355] tot_loss[loss=2.29, over 5492204.30 frames. , ppl: 9.874882095419824], batch size: 70
+2022-12-12 06:38:54,618 INFO [train.py:421] (2/8) Epoch 6, batch 56400, loss[loss=2.457, over 1820.00 frames. , ppl: 11.66870557495504] tot_loss[loss=2.29, over 5508814.44 frames. , ppl: 9.875089986932801], batch size: 70
+2022-12-12 06:40:33,674 INFO [train.py:421] (2/8) Epoch 6, batch 56600, loss[loss=2.413, over 1330.00 frames. , ppl: 11.17155925612407] tot_loss[loss=2.291, over 5475510.02 frames. , ppl: 9.8812556713813], batch size: 70
+2022-12-12 06:42:12,600 INFO [train.py:421] (2/8) Epoch 6, batch 56800, loss[loss=2.21, over 4900.00 frames. , ppl: 9.11751878554252] tot_loss[loss=2.288, over 5533603.04 frames. , ppl: 9.859038260982208], batch size: 70
+2022-12-12 06:43:53,582 INFO [train.py:421] (2/8) Epoch 6, batch 57000, loss[loss=2.233, over 4760.00 frames. , ppl: 9.32800674847783] tot_loss[loss=2.288, over 5544768.54 frames. , ppl: 9.850664760504525], batch size: 70
+2022-12-12 06:43:53,583 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:43:54,329 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.763007694592458
+2022-12-12 06:45:35,859 INFO [train.py:421] (2/8) Epoch 6, batch 57200, loss[loss=2.24, over 2100.00 frames. , ppl: 9.397342131457053] tot_loss[loss=2.288, over 5539487.00 frames. , ppl: 9.853168558656732], batch size: 70
+2022-12-12 06:47:18,379 INFO [train.py:421] (2/8) Epoch 6, batch 57400, loss[loss=2.156, over 10360.00 frames. , ppl: 8.638160644765264] tot_loss[loss=2.288, over 5543595.23 frames. , ppl: 9.850571178661191], batch size: 70
+2022-12-12 06:48:59,494 INFO [train.py:421] (2/8) Epoch 6, batch 57600, loss[loss=2.338, over 1680.00 frames. , ppl: 10.363954334160399] tot_loss[loss=2.289, over 5505007.77 frames. , ppl: 9.860868665721593], batch size: 70
+2022-12-12 06:50:45,868 INFO [train.py:421] (2/8) Epoch 6, batch 57800, loss[loss=2.321, over 2590.00 frames. , ppl: 10.187743431840328] tot_loss[loss=2.288, over 5532901.28 frames. , ppl: 9.854617901898926], batch size: 70
+2022-12-12 06:52:25,594 INFO [train.py:421] (2/8) Epoch 6, batch 58000, loss[loss=2.126, over 7490.00 frames. , ppl: 8.377859536332993] tot_loss[loss=2.288, over 5543185.64 frames. , ppl: 9.855888576913033], batch size: 70
+2022-12-12 06:52:25,595 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 06:52:26,339 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758448345808779
+2022-12-12 06:54:02,208 INFO [train.py:421] (2/8) Epoch 6, batch 58200, loss[loss=2.2, over 10220.00 frames. , ppl: 9.0221887955449] tot_loss[loss=2.29, over 5504792.60 frames. , ppl: 9.87160657674692], batch size: 70
+2022-12-12 06:55:40,804 INFO [train.py:421] (2/8) Epoch 6, batch 58400, loss[loss=2.148, over 4270.00 frames. , ppl: 8.569832039694354] tot_loss[loss=2.289, over 5498575.71 frames. , ppl: 9.869437450126199], batch size: 70
+2022-12-12 06:57:22,165 INFO [train.py:421] (2/8) Epoch 6, batch 58600, loss[loss=2.189, over 3780.00 frames. , ppl: 8.930286477333826] tot_loss[loss=2.289, over 5539333.20 frames. , ppl: 9.860399455317634], batch size: 70
+2022-12-12 06:59:04,023 INFO [train.py:421] (2/8) Epoch 6, batch 58800, loss[loss=3.23, over 490.00 frames. , ppl: 25.276575041863765] tot_loss[loss=2.288, over 5538727.22 frames. , ppl: 9.855154942778569], batch size: 70
+2022-12-12 07:00:45,054 INFO [train.py:421] (2/8) Epoch 6, batch 59000, loss[loss=2.302, over 2660.00 frames. , ppl: 9.99073656933049] tot_loss[loss=2.29, over 5490873.44 frames. , ppl: 9.87095141092695], batch size: 70
+2022-12-12 07:00:45,054 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:00:45,814 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.762743263537212
+2022-12-12 07:02:25,671 INFO [train.py:421] (2/8) Epoch 6, batch 59200, loss[loss=2.253, over 2660.00 frames. , ppl: 9.518650912952511] tot_loss[loss=2.289, over 5498286.95 frames. , ppl: 9.860309682512417], batch size: 70
+2022-12-12 07:04:06,628 INFO [train.py:421] (2/8) Epoch 6, batch 59400, loss[loss=2.272, over 4200.00 frames. , ppl: 9.701777394623937] tot_loss[loss=2.289, over 5510151.37 frames. , ppl: 9.863160138668785], batch size: 70
+2022-12-12 07:05:46,487 INFO [train.py:421] (2/8) Epoch 6, batch 59600, loss[loss=2.39, over 1540.00 frames. , ppl: 10.908291591608627] tot_loss[loss=2.29, over 5462564.14 frames. , ppl: 9.874161492676963], batch size: 70
+2022-12-12 07:07:27,936 INFO [train.py:421] (2/8) Epoch 6, batch 59800, loss[loss=2.541, over 770.00 frames. , ppl: 12.69434627639011] tot_loss[loss=2.29, over 5442302.79 frames. , ppl: 9.871387873440224], batch size: 70
+2022-12-12 07:09:08,960 INFO [train.py:421] (2/8) Epoch 6, batch 60000, loss[loss=2.283, over 3640.00 frames. , ppl: 9.80315342809449] tot_loss[loss=2.289, over 5469311.51 frames. , ppl: 9.863918532338497], batch size: 70
+2022-12-12 07:09:08,960 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:09:09,719 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.763438313496728
+2022-12-12 07:10:50,271 INFO [train.py:421] (2/8) Epoch 6, batch 60200, loss[loss=3.584, over 420.00 frames. , ppl: 36.031124631339374] tot_loss[loss=2.289, over 5463226.14 frames. , ppl: 9.86979281932899], batch size: 70
+2022-12-12 07:12:28,642 INFO [train.py:421] (2/8) Epoch 6, batch 60400, loss[loss=2.445, over 980.00 frames. , ppl: 11.530025087501635] tot_loss[loss=2.289, over 5487292.55 frames. , ppl: 9.862822535955468], batch size: 70
+2022-12-12 07:14:10,681 INFO [train.py:421] (2/8) Epoch 6, batch 60600, loss[loss=2.264, over 3080.00 frames. , ppl: 9.624540633039038] tot_loss[loss=2.287, over 5561977.60 frames. , ppl: 9.84729387214445], batch size: 70
+2022-12-12 07:15:51,737 INFO [train.py:421] (2/8) Epoch 6, batch 60800, loss[loss=2.339, over 1890.00 frames. , ppl: 10.368540277457969] tot_loss[loss=2.287, over 5565758.95 frames. , ppl: 9.849076470440473], batch size: 70
+2022-12-12 07:17:31,925 INFO [train.py:421] (2/8) Epoch 6, batch 61000, loss[loss=2.26, over 6090.00 frames. , ppl: 9.581705317257446] tot_loss[loss=2.288, over 5525980.51 frames. , ppl: 9.857688404957994], batch size: 70
+2022-12-12 07:17:31,926 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:17:32,674 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.767613989568469
+2022-12-12 07:19:10,821 INFO [train.py:421] (2/8) Epoch 6, batch 61200, loss[loss=2.163, over 5600.00 frames. , ppl: 8.699334662435609] tot_loss[loss=2.289, over 5494170.19 frames. , ppl: 9.860708907036498], batch size: 70
+2022-12-12 07:20:50,819 INFO [train.py:421] (2/8) Epoch 6, batch 61400, loss[loss=2.246, over 3850.00 frames. , ppl: 9.451818271795057] tot_loss[loss=2.29, over 5464828.37 frames. , ppl: 9.871595194354445], batch size: 70
+2022-12-12 07:22:30,434 INFO [train.py:421] (2/8) Epoch 6, batch 61600, loss[loss=2.343, over 2450.00 frames. , ppl: 10.410636722697003] tot_loss[loss=2.29, over 5447711.29 frames. , ppl: 9.876358805329179], batch size: 70
+2022-12-12 07:24:09,143 INFO [train.py:421] (2/8) Epoch 6, batch 61800, loss[loss=2.186, over 5600.00 frames. , ppl: 8.900390436889454] tot_loss[loss=2.29, over 5444759.36 frames. , ppl: 9.87324455886828], batch size: 70
+2022-12-12 07:25:49,776 INFO [train.py:421] (2/8) Epoch 6, batch 62000, loss[loss=2.448, over 1050.00 frames. , ppl: 11.559503419211564] tot_loss[loss=2.29, over 5439719.88 frames. , ppl: 9.873338259824246], batch size: 70
+2022-12-12 07:25:49,777 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:25:50,536 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.776349799606294
+2022-12-12 07:27:29,140 INFO [train.py:421] (2/8) Epoch 6, batch 62200, loss[loss=2.444, over 980.00 frames. , ppl: 11.516122464851236] tot_loss[loss=2.291, over 5424235.86 frames. , ppl: 9.884074345276808], batch size: 70
+2022-12-12 07:29:06,005 INFO [train.py:421] (2/8) Epoch 6, batch 62400, loss[loss=2.478, over 1260.00 frames. , ppl: 11.919694615341342] tot_loss[loss=2.291, over 5411623.08 frames. , ppl: 9.883293136851647], batch size: 70
+2022-12-12 07:30:49,310 INFO [train.py:421] (2/8) Epoch 6, batch 62600, loss[loss=2.601, over 980.00 frames. , ppl: 13.47978031648352] tot_loss[loss=2.29, over 5448640.37 frames. , ppl: 9.875874606752141], batch size: 70
+2022-12-12 07:32:26,314 INFO [train.py:421] (2/8) Epoch 6, batch 62800, loss[loss=2.258, over 2730.00 frames. , ppl: 9.565693392522167] tot_loss[loss=2.291, over 5415511.42 frames. , ppl: 9.884681925772425], batch size: 70
+2022-12-12 07:34:06,411 INFO [train.py:421] (2/8) Epoch 6, batch 63000, loss[loss=2.386, over 1260.00 frames. , ppl: 10.866278682632036] tot_loss[loss=2.291, over 5399207.28 frames. , ppl: 9.882411133771663], batch size: 70
+2022-12-12 07:34:06,412 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:34:07,172 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.784635781017636
+2022-12-12 07:35:46,470 INFO [train.py:421] (2/8) Epoch 6, batch 63200, loss[loss=2.467, over 910.00 frames. , ppl: 11.786972195909264] tot_loss[loss=2.29, over 5425950.32 frames. , ppl: 9.873166296327284], batch size: 70
+2022-12-12 07:37:31,524 INFO [train.py:421] (2/8) Epoch 6, batch 63400, loss[loss=2.499, over 910.00 frames. , ppl: 12.16750124952303] tot_loss[loss=2.289, over 5457316.68 frames. , ppl: 9.863373037648262], batch size: 70
+2022-12-12 07:39:13,151 INFO [train.py:421] (2/8) Epoch 6, batch 63600, loss[loss=2.326, over 2870.00 frames. , ppl: 10.240420552735246] tot_loss[loss=2.288, over 5492568.66 frames. , ppl: 9.851315926761849], batch size: 70
+2022-12-12 07:40:49,344 INFO [train.py:421] (2/8) Epoch 6, batch 63800, loss[loss=2.374, over 2240.00 frames. , ppl: 10.735154475816355] tot_loss[loss=2.289, over 5476252.13 frames. , ppl: 9.861120596729274], batch size: 70
+2022-12-12 07:42:29,285 INFO [train.py:421] (2/8) Epoch 6, batch 64000, loss[loss=2.448, over 980.00 frames. , ppl: 11.562318611010863] tot_loss[loss=2.288, over 5471893.45 frames. , ppl: 9.859188716876721], batch size: 70
+2022-12-12 07:42:29,286 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:42:30,050 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.768704092947166
+2022-12-12 07:44:07,948 INFO [train.py:421] (2/8) Epoch 6, batch 64200, loss[loss=2.147, over 5250.00 frames. , ppl: 8.556388521248419] tot_loss[loss=2.288, over 5504677.63 frames. , ppl: 9.854567207035405], batch size: 70
+2022-12-12 07:45:49,265 INFO [train.py:421] (2/8) Epoch 6, batch 64400, loss[loss=2.319, over 2520.00 frames. , ppl: 10.170274959544772] tot_loss[loss=2.289, over 5492297.62 frames. , ppl: 9.860928844861316], batch size: 70
+2022-12-12 07:47:28,533 INFO [train.py:421] (2/8) Epoch 6, batch 64600, loss[loss=2.218, over 3290.00 frames. , ppl: 9.189146592506534] tot_loss[loss=2.289, over 5471343.53 frames. , ppl: 9.866404509609344], batch size: 70
+2022-12-12 07:49:11,547 INFO [train.py:421] (2/8) Epoch 6, batch 64800, loss[loss=2.371, over 3360.00 frames. , ppl: 10.708342639496237] tot_loss[loss=2.291, over 5413738.67 frames. , ppl: 9.884518090954845], batch size: 70
+2022-12-12 07:50:54,147 INFO [train.py:421] (2/8) Epoch 6, batch 65000, loss[loss=2.416, over 2170.00 frames. , ppl: 11.197184383776714] tot_loss[loss=2.29, over 5448652.11 frames. , ppl: 9.8745982708246], batch size: 70
+2022-12-12 07:50:54,148 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:50:54,908 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.786995181821556
+2022-12-12 07:52:32,871 INFO [train.py:421] (2/8) Epoch 6, batch 65200, loss[loss=2.296, over 2100.00 frames. , ppl: 9.936780000279647] tot_loss[loss=2.289, over 5482400.60 frames. , ppl: 9.864735064365556], batch size: 70
+2022-12-12 07:54:12,248 INFO [train.py:421] (2/8) Epoch 6, batch 65400, loss[loss=2.569, over 1120.00 frames. , ppl: 13.04676617843376] tot_loss[loss=2.291, over 5429626.49 frames. , ppl: 9.881739823296291], batch size: 70
+2022-12-12 07:55:56,355 INFO [train.py:421] (2/8) Epoch 6, batch 65600, loss[loss=2.39, over 2590.00 frames. , ppl: 10.914574581236842] tot_loss[loss=2.291, over 5421460.18 frames. , ppl: 9.885655080235948], batch size: 70
+2022-12-12 07:57:38,642 INFO [train.py:421] (2/8) Epoch 6, batch 65800, loss[loss=2.404, over 1890.00 frames. , ppl: 11.070417553218459] tot_loss[loss=2.291, over 5443661.57 frames. , ppl: 9.884251577161752], batch size: 70
+2022-12-12 07:59:17,514 INFO [train.py:421] (2/8) Epoch 6, batch 66000, loss[loss=2.235, over 1960.00 frames. , ppl: 9.350523007620485] tot_loss[loss=2.291, over 5424983.28 frames. , ppl: 9.882844328552869], batch size: 70
+2022-12-12 07:59:17,514 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 07:59:18,272 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.756986800195627
+2022-12-12 08:00:55,080 INFO [train.py:421] (2/8) Epoch 6, batch 66200, loss[loss=2.311, over 2240.00 frames. , ppl: 10.081660602842724] tot_loss[loss=2.29, over 5443272.21 frames. , ppl: 9.879342574134421], batch size: 70
+2022-12-12 08:02:34,946 INFO [train.py:421] (2/8) Epoch 6, batch 66400, loss[loss=2.239, over 2940.00 frames. , ppl: 9.384047570365695] tot_loss[loss=2.29, over 5466396.41 frames. , ppl: 9.872366996176133], batch size: 70
+2022-12-12 08:04:12,888 INFO [train.py:421] (2/8) Epoch 6, batch 66600, loss[loss=2.277, over 2310.00 frames. , ppl: 9.747798656084685] tot_loss[loss=2.29, over 5435979.11 frames. , ppl: 9.878716574779217], batch size: 70
+2022-12-12 08:05:56,265 INFO [train.py:421] (2/8) Epoch 6, batch 66800, loss[loss=2.489, over 1750.00 frames. , ppl: 12.054063050461382] tot_loss[loss=2.29, over 5450878.64 frames. , ppl: 9.87518161886003], batch size: 70
+2022-12-12 08:07:41,370 INFO [train.py:421] (2/8) Epoch 6, batch 67000, loss[loss=2.339, over 2170.00 frames. , ppl: 10.375785004953379] tot_loss[loss=2.29, over 5455398.42 frames. , ppl: 9.877133402194236], batch size: 70
+2022-12-12 08:07:41,371 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 08:07:42,129 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.766049889291654
+2022-12-12 08:09:21,229 INFO [train.py:421] (2/8) Epoch 6, batch 67200, loss[loss=2.27, over 4340.00 frames. , ppl: 9.679742331214985] tot_loss[loss=2.289, over 5467255.31 frames. , ppl: 9.866184712291119], batch size: 70
+2022-12-12 08:11:01,763 INFO [train.py:421] (2/8) Epoch 6, batch 67400, loss[loss=2.23, over 4270.00 frames. , ppl: 9.304430694800057] tot_loss[loss=2.289, over 5500463.57 frames. , ppl: 9.860186849251072], batch size: 70
+2022-12-12 08:12:41,350 INFO [train.py:421] (2/8) Epoch 6, batch 67600, loss[loss=2.39, over 1330.00 frames. , ppl: 10.916158294784527] tot_loss[loss=2.288, over 5516814.21 frames. , ppl: 9.855651212349246], batch size: 70
+2022-12-12 08:14:15,982 INFO [train.py:421] (2/8) Epoch 6, batch 67800, loss[loss=2.34, over 1260.00 frames. , ppl: 10.379859987980396] tot_loss[loss=2.289, over 5498075.07 frames. , ppl: 9.862483052011173], batch size: 70
+2022-12-12 08:15:54,492 INFO [train.py:421] (2/8) Epoch 6, batch 68000, loss[loss=2.353, over 2730.00 frames. , ppl: 10.518559581683128] tot_loss[loss=2.289, over 5481350.09 frames. , ppl: 9.860939557228551], batch size: 70
+2022-12-12 08:15:54,493 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 08:15:55,254 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.779122592522853
+2022-12-12 08:17:39,587 INFO [train.py:421] (2/8) Epoch 6, batch 68200, loss[loss=2.426, over 1260.00 frames. , ppl: 11.30850755072128] tot_loss[loss=2.289, over 5470591.72 frames. , ppl: 9.864317599699682], batch size: 70
+2022-12-12 08:19:22,896 INFO [train.py:421] (2/8) Epoch 6, batch 68400, loss[loss=2.286, over 2310.00 frames. , ppl: 9.833675665112308] tot_loss[loss=2.289, over 5460002.04 frames. , ppl: 9.866491461360893], batch size: 70
+2022-12-12 08:21:04,974 INFO [train.py:421] (2/8) Epoch 6, batch 68600, loss[loss=2.272, over 2730.00 frames. , ppl: 9.699310028420552] tot_loss[loss=2.29, over 5459168.79 frames. , ppl: 9.870400526122738], batch size: 70
+2022-12-12 08:22:43,487 INFO [train.py:421] (2/8) Epoch 6, batch 68800, loss[loss=2.465, over 910.00 frames. , ppl: 11.761360260883784] tot_loss[loss=2.291, over 5409438.11 frames. , ppl: 9.881097281637473], batch size: 70
+2022-12-12 08:24:23,648 INFO [train.py:421] (2/8) Epoch 6, batch 69000, loss[loss=2.532, over 980.00 frames. , ppl: 12.581143670700556] tot_loss[loss=2.29, over 5440633.69 frames. , ppl: 9.876902118218592], batch size: 70
+2022-12-12 08:24:23,649 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 08:24:24,395 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.762691245156288
+2022-12-12 08:26:05,592 INFO [train.py:421] (2/8) Epoch 6, batch 69200, loss[loss=2.343, over 2030.00 frames. , ppl: 10.410696749673715] tot_loss[loss=2.29, over 5428250.56 frames. , ppl: 9.87512016452395], batch size: 70
+2022-12-12 08:27:44,996 INFO [train.py:421] (2/8) Epoch 6, batch 69400, loss[loss=2.445, over 1330.00 frames. , ppl: 11.528632966833722] tot_loss[loss=2.289, over 5452221.75 frames. , ppl: 9.867627702175056], batch size: 70
+2022-12-12 08:29:30,567 INFO [train.py:421] (2/8) Epoch 6, batch 69600, loss[loss=2.379, over 1680.00 frames. , ppl: 10.797992479292416] tot_loss[loss=2.288, over 5487252.17 frames. , ppl: 9.859932857663258], batch size: 70
+2022-12-12 08:31:11,121 INFO [train.py:421] (2/8) Epoch 6, batch 69800, loss[loss=3.342, over 420.00 frames. , ppl: 28.278326314517006] tot_loss[loss=2.289, over 5488868.86 frames. , ppl: 9.865351464027974], batch size: 70
+2022-12-12 08:32:55,420 INFO [train.py:421] (2/8) Epoch 6, batch 70000, loss[loss=2.367, over 1820.00 frames. , ppl: 10.661677350931088] tot_loss[loss=2.29, over 5462625.19 frames. , ppl: 9.87599193368197], batch size: 70
+2022-12-12 08:32:55,420 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 08:32:56,166 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.761897999189102
+2022-12-12 08:34:38,123 INFO [train.py:421] (2/8) Epoch 6, batch 70200, loss[loss=2.282, over 2800.00 frames. , ppl: 9.794540372196495] tot_loss[loss=2.291, over 5450586.15 frames. , ppl: 9.880613405140362], batch size: 70
+2022-12-12 08:36:17,125 INFO [train.py:421] (2/8) Epoch 6, batch 70400, loss[loss=2.235, over 2800.00 frames. , ppl: 9.345588711269862] tot_loss[loss=2.291, over 5429609.51 frames. , ppl: 9.885121776705557], batch size: 70
+2022-12-12 08:37:57,454 INFO [train.py:421] (2/8) Epoch 6, batch 70600, loss[loss=2.391, over 1260.00 frames. , ppl: 10.924227978668965] tot_loss[loss=2.29, over 5472263.84 frames. , ppl: 9.875564039092783], batch size: 70
+2022-12-12 08:39:34,984 INFO [train.py:421] (2/8) Epoch 6, batch 70800, loss[loss=2.451, over 1610.00 frames. , ppl: 11.594505789220353] tot_loss[loss=2.29, over 5483266.34 frames. , ppl: 9.87418180593554], batch size: 70
+2022-12-12 08:41:17,309 INFO [train.py:421] (2/8) Epoch 6, batch 71000, loss[loss=2.25, over 4480.00 frames. , ppl: 9.48879065775462] tot_loss[loss=2.29, over 5477188.70 frames. , ppl: 9.872412686365566], batch size: 70
+2022-12-12 08:41:17,310 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 08:41:18,069 INFO [train.py:452] (2/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.760633851500042
+2022-12-12 08:42:55,816 INFO [train.py:421] (2/8) Epoch 6, batch 71200, loss[loss=2.29, over 2730.00 frames. , ppl: 9.870845166571904] tot_loss[loss=2.289, over 5522447.03 frames. , ppl: 9.860916091984821], batch size: 70
+2022-12-12 08:44:37,451 INFO [train.py:421] (2/8) Epoch 6, batch 71400, loss[loss=2.326, over 2870.00 frames. , ppl: 10.241267312410427] tot_loss[loss=2.289, over 5506689.54 frames. , ppl: 9.862478631155085], batch size: 70
+2022-12-12 08:46:15,621 INFO [train.py:421] (2/8) Epoch 6, batch 71600, loss[loss=2.436, over 1400.00 frames. , ppl: 11.42206543846439] tot_loss[loss=2.289, over 5482595.47 frames. , ppl: 9.868692065707071], batch size: 70
+2022-12-12 08:47:54,854 INFO [train.py:421] (2/8) Epoch 6, batch 71800, loss[loss=2.33, over 2030.00 frames. , ppl: 10.28211245295905] tot_loss[loss=2.29, over 5475836.86 frames. , ppl: 9.870323993648457], batch size: 70
+2022-12-12 08:49:08,129 INFO [train.py:421] (2/8) Epoch 7, batch 0, loss[loss=2.359, over 2100.00 frames. , ppl: 10.582210532844584] tot_loss[loss=2.359, over 2100.00 frames. , ppl: 10.582210532844584], batch size: 70
+2022-12-12 08:50:48,794 INFO [train.py:421] (2/8) Epoch 7, batch 200, loss[loss=2.283, over 3780.00 frames. , ppl: 9.809120982614962] tot_loss[loss=2.284, over 512588.69 frames. , ppl: 9.811126885527907], batch size: 70
+2022-12-12 08:52:29,311 INFO [train.py:421] (2/8) Epoch 7, batch 400, loss[loss=2.242, over 8120.00 frames. , ppl: 9.407796717198044] tot_loss[loss=2.276, over 1018025.46 frames. , ppl: 9.735560968074628], batch size: 70
+2022-12-12 08:54:08,469 INFO [train.py:421] (2/8) Epoch 7, batch 600, loss[loss=2.459, over 1260.00 frames. , ppl: 11.697437122093604] tot_loss[loss=2.276, over 1427083.53 frames. , ppl: 9.735735244120692], batch size: 70
+2022-12-12 08:55:44,620 INFO [train.py:421] (2/8) Epoch 7, batch 800, loss[loss=2.353, over 1540.00 frames. , ppl: 10.514790510738612] tot_loss[loss=2.281, over 1783391.46 frames. , ppl: 9.78794366305025], batch size: 70
+2022-12-12 08:57:20,592 INFO [train.py:421] (2/8) Epoch 7, batch 1000, loss[loss=2.469, over 1050.00 frames. , ppl: 11.809413284273123] tot_loss[loss=2.282, over 2134720.19 frames. , ppl: 9.800839952817629], batch size: 70
+2022-12-12 08:57:20,593 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 08:57:21,340 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.766622303585429
+2022-12-12 08:58:56,584 INFO [train.py:421] (2/8) Epoch 7, batch 1200, loss[loss=2.408, over 1960.00 frames. , ppl: 11.112483286138193] tot_loss[loss=2.281, over 2471348.51 frames. , ppl: 9.79067259583073], batch size: 70
+2022-12-12 09:00:32,854 INFO [train.py:421] (2/8) Epoch 7, batch 1400, loss[loss=2.375, over 1330.00 frames. , ppl: 10.749888348176247] tot_loss[loss=2.282, over 2729950.13 frames. , ppl: 9.79357182540091], batch size: 70
+2022-12-12 09:02:12,540 INFO [train.py:421] (2/8) Epoch 7, batch 1600, loss[loss=2.345, over 1820.00 frames. , ppl: 10.436841637116643] tot_loss[loss=2.281, over 2987279.99 frames. , ppl: 9.78425514160984], batch size: 70
+2022-12-12 09:03:50,387 INFO [train.py:421] (2/8) Epoch 7, batch 1800, loss[loss=2.675, over 770.00 frames. , ppl: 14.513440405027698] tot_loss[loss=2.279, over 3260033.17 frames. , ppl: 9.762570669912646], batch size: 70
+2022-12-12 09:05:32,155 INFO [train.py:421] (2/8) Epoch 7, batch 2000, loss[loss=2.451, over 1890.00 frames. , ppl: 11.604770620720398] tot_loss[loss=2.28, over 3444683.78 frames. , ppl: 9.775158382098617], batch size: 70
+2022-12-12 09:05:32,156 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:05:32,901 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764543848320063
+2022-12-12 09:07:08,130 INFO [train.py:421] (2/8) Epoch 7, batch 2200, loss[loss=2.401, over 2030.00 frames. , ppl: 11.035055897193908] tot_loss[loss=2.282, over 3613033.68 frames. , ppl: 9.794762086493783], batch size: 70
+2022-12-12 09:08:47,615 INFO [train.py:421] (2/8) Epoch 7, batch 2400, loss[loss=3.027, over 560.00 frames. , ppl: 20.62904976249097] tot_loss[loss=2.284, over 3732636.52 frames. , ppl: 9.812342316908405], batch size: 70
+2022-12-12 09:10:26,432 INFO [train.py:421] (2/8) Epoch 7, batch 2600, loss[loss=2.399, over 1680.00 frames. , ppl: 11.007064963899257] tot_loss[loss=2.282, over 3929833.34 frames. , ppl: 9.799632974251582], batch size: 70
+2022-12-12 09:12:07,859 INFO [train.py:421] (2/8) Epoch 7, batch 2800, loss[loss=2.225, over 4970.00 frames. , ppl: 9.255228469577661] tot_loss[loss=2.28, over 4112929.48 frames. , ppl: 9.77913534299229], batch size: 70
+2022-12-12 09:13:50,922 INFO [train.py:421] (2/8) Epoch 7, batch 3000, loss[loss=2.324, over 4900.00 frames. , ppl: 10.21849810324716] tot_loss[loss=2.28, over 4276475.93 frames. , ppl: 9.778950059713642], batch size: 70
+2022-12-12 09:13:50,922 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:13:51,668 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.743588919093323
+2022-12-12 09:15:28,748 INFO [train.py:421] (2/8) Epoch 7, batch 3200, loss[loss=2.171, over 6580.00 frames. , ppl: 8.765284708580461] tot_loss[loss=2.282, over 4356896.59 frames. , ppl: 9.793472216164961], batch size: 70
+2022-12-12 09:17:10,259 INFO [train.py:421] (2/8) Epoch 7, batch 3400, loss[loss=2.32, over 2450.00 frames. , ppl: 10.177986484525329] tot_loss[loss=2.281, over 4472425.04 frames. , ppl: 9.786746566319236], batch size: 70
+2022-12-12 09:18:51,745 INFO [train.py:421] (2/8) Epoch 7, batch 3600, loss[loss=2.448, over 2170.00 frames. , ppl: 11.562248551476712] tot_loss[loss=2.281, over 4576017.13 frames. , ppl: 9.785463173084754], batch size: 70
+2022-12-12 09:20:35,629 INFO [train.py:421] (2/8) Epoch 7, batch 3800, loss[loss=2.333, over 2800.00 frames. , ppl: 10.308760172966881] tot_loss[loss=2.281, over 4661741.53 frames. , ppl: 9.786403487503565], batch size: 70
+2022-12-12 09:22:13,512 INFO [train.py:421] (2/8) Epoch 7, batch 4000, loss[loss=2.215, over 6090.00 frames. , ppl: 9.160829725336253] tot_loss[loss=2.282, over 4705970.42 frames. , ppl: 9.795252566440048], batch size: 70
+2022-12-12 09:22:13,513 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:22:14,273 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764824226068129
+2022-12-12 09:23:52,017 INFO [train.py:421] (2/8) Epoch 7, batch 4200, loss[loss=2.588, over 1050.00 frames. , ppl: 13.304430475494229] tot_loss[loss=2.283, over 4773966.35 frames. , ppl: 9.804001861550292], batch size: 70
+2022-12-12 09:25:30,244 INFO [train.py:421] (2/8) Epoch 7, batch 4400, loss[loss=2.267, over 2170.00 frames. , ppl: 9.651833501566081] tot_loss[loss=2.283, over 4813443.24 frames. , ppl: 9.808749852792875], batch size: 70
+2022-12-12 09:27:09,447 INFO [train.py:421] (2/8) Epoch 7, batch 4600, loss[loss=2.293, over 2310.00 frames. , ppl: 9.908397466014707] tot_loss[loss=2.283, over 4872755.33 frames. , ppl: 9.804155946569045], batch size: 70
+2022-12-12 09:28:50,536 INFO [train.py:421] (2/8) Epoch 7, batch 4800, loss[loss=2.279, over 2940.00 frames. , ppl: 9.768171541215079] tot_loss[loss=2.283, over 4911359.01 frames. , ppl: 9.8078495795928], batch size: 70
+2022-12-12 09:30:30,393 INFO [train.py:421] (2/8) Epoch 7, batch 5000, loss[loss=2.379, over 1470.00 frames. , ppl: 10.792625411925707] tot_loss[loss=2.282, over 5017442.38 frames. , ppl: 9.794838302269158], batch size: 70
+2022-12-12 09:30:30,394 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:30:31,152 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.770388643082995
+2022-12-12 09:32:12,050 INFO [train.py:421] (2/8) Epoch 7, batch 5200, loss[loss=2.388, over 2100.00 frames. , ppl: 10.890472232288447] tot_loss[loss=2.282, over 5062802.46 frames. , ppl: 9.792525755953074], batch size: 70
+2022-12-12 09:33:52,019 INFO [train.py:421] (2/8) Epoch 7, batch 5400, loss[loss=2.509, over 1190.00 frames. , ppl: 12.28728296706767] tot_loss[loss=2.281, over 5135023.68 frames. , ppl: 9.785136728518253], batch size: 70
+2022-12-12 09:35:30,665 INFO [train.py:421] (2/8) Epoch 7, batch 5600, loss[loss=2.342, over 1260.00 frames. , ppl: 10.400003842258984] tot_loss[loss=2.281, over 5169299.92 frames. , ppl: 9.782080256330183], batch size: 70
+2022-12-12 09:37:13,202 INFO [train.py:421] (2/8) Epoch 7, batch 5800, loss[loss=2.253, over 3290.00 frames. , ppl: 9.515784532815772] tot_loss[loss=2.282, over 5135956.30 frames. , ppl: 9.79444968481237], batch size: 70
+2022-12-12 09:38:55,653 INFO [train.py:421] (2/8) Epoch 7, batch 6000, loss[loss=2.421, over 1820.00 frames. , ppl: 11.25461475732561] tot_loss[loss=2.282, over 5177705.93 frames. , ppl: 9.796613993417488], batch size: 70
+2022-12-12 09:38:55,653 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:38:56,400 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.766977911177703
+2022-12-12 09:40:36,648 INFO [train.py:421] (2/8) Epoch 7, batch 6200, loss[loss=2.258, over 5110.00 frames. , ppl: 9.55965498918196] tot_loss[loss=2.283, over 5194619.08 frames. , ppl: 9.805411788511977], batch size: 70
+2022-12-12 09:42:17,041 INFO [train.py:421] (2/8) Epoch 7, batch 6400, loss[loss=2.249, over 5530.00 frames. , ppl: 9.482184942471244] tot_loss[loss=2.282, over 5262862.83 frames. , ppl: 9.79820366974148], batch size: 70
+2022-12-12 09:43:57,321 INFO [train.py:421] (2/8) Epoch 7, batch 6600, loss[loss=2.234, over 2940.00 frames. , ppl: 9.333009683251838] tot_loss[loss=2.282, over 5277284.02 frames. , ppl: 9.79626965211653], batch size: 70
+2022-12-12 09:45:38,137 INFO [train.py:421] (2/8) Epoch 7, batch 6800, loss[loss=2.231, over 5040.00 frames. , ppl: 9.310495089322185] tot_loss[loss=2.282, over 5308060.87 frames. , ppl: 9.79487638007873], batch size: 70
+2022-12-12 09:47:22,573 INFO [train.py:421] (2/8) Epoch 7, batch 7000, loss[loss=2.471, over 1050.00 frames. , ppl: 11.829475115081953] tot_loss[loss=2.283, over 5302168.00 frames. , ppl: 9.80745315022856], batch size: 70
+2022-12-12 09:47:22,573 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:47:23,333 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.761162606501355
+2022-12-12 09:49:03,422 INFO [train.py:421] (2/8) Epoch 7, batch 7200, loss[loss=2.928, over 560.00 frames. , ppl: 18.69671384716954] tot_loss[loss=2.282, over 5348346.69 frames. , ppl: 9.794794497540336], batch size: 70
+2022-12-12 09:50:43,930 INFO [train.py:421] (2/8) Epoch 7, batch 7400, loss[loss=2.125, over 5250.00 frames. , ppl: 8.37482895954682] tot_loss[loss=2.282, over 5370402.94 frames. , ppl: 9.79277866451412], batch size: 70
+2022-12-12 09:52:24,697 INFO [train.py:421] (2/8) Epoch 7, batch 7600, loss[loss=2.237, over 3780.00 frames. , ppl: 9.369623118196747] tot_loss[loss=2.282, over 5371578.80 frames. , ppl: 9.792457240353459], batch size: 70
+2022-12-12 09:54:01,291 INFO [train.py:421] (2/8) Epoch 7, batch 7800, loss[loss=2.968, over 560.00 frames. , ppl: 19.452460105112355] tot_loss[loss=2.282, over 5396049.63 frames. , ppl: 9.793062020835908], batch size: 70
+2022-12-12 09:55:38,126 INFO [train.py:421] (2/8) Epoch 7, batch 8000, loss[loss=2.43, over 1400.00 frames. , ppl: 11.356315212014819] tot_loss[loss=2.281, over 5395086.76 frames. , ppl: 9.787181529829056], batch size: 70
+2022-12-12 09:55:38,127 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 09:55:38,887 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.762243320567698
+2022-12-12 09:57:20,362 INFO [train.py:421] (2/8) Epoch 7, batch 8200, loss[loss=3.031, over 560.00 frames. , ppl: 20.708399199324077] tot_loss[loss=2.282, over 5390708.13 frames. , ppl: 9.794690484295007], batch size: 70
+2022-12-12 09:59:00,364 INFO [train.py:421] (2/8) Epoch 7, batch 8400, loss[loss=2.429, over 1680.00 frames. , ppl: 11.352768897726497] tot_loss[loss=2.282, over 5395737.94 frames. , ppl: 9.792081189084811], batch size: 70
+2022-12-12 10:00:40,150 INFO [train.py:421] (2/8) Epoch 7, batch 8600, loss[loss=2.203, over 7420.00 frames. , ppl: 9.053105883793078] tot_loss[loss=2.281, over 5420914.80 frames. , ppl: 9.789870101950601], batch size: 70
+2022-12-12 10:02:20,359 INFO [train.py:421] (2/8) Epoch 7, batch 8800, loss[loss=2.42, over 2240.00 frames. , ppl: 11.25080635292416] tot_loss[loss=2.28, over 5477209.74 frames. , ppl: 9.77500135833493], batch size: 70
+2022-12-12 10:04:02,209 INFO [train.py:421] (2/8) Epoch 7, batch 9000, loss[loss=2.215, over 2100.00 frames. , ppl: 9.159025768217472] tot_loss[loss=2.281, over 5438976.52 frames. , ppl: 9.789245007091507], batch size: 70
+2022-12-12 10:04:02,209 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:04:02,970 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758306803162158
+2022-12-12 10:05:41,185 INFO [train.py:421] (2/8) Epoch 7, batch 9200, loss[loss=2.438, over 1190.00 frames. , ppl: 11.454865842316392] tot_loss[loss=2.281, over 5462745.19 frames. , ppl: 9.786951422209075], batch size: 70
+2022-12-12 10:07:18,496 INFO [train.py:421] (2/8) Epoch 7, batch 9400, loss[loss=2.53, over 770.00 frames. , ppl: 12.556939205345945] tot_loss[loss=2.281, over 5474972.00 frames. , ppl: 9.78399736317454], batch size: 70
+2022-12-12 10:08:57,692 INFO [train.py:421] (2/8) Epoch 7, batch 9600, loss[loss=2.249, over 2170.00 frames. , ppl: 9.476033258112073] tot_loss[loss=2.28, over 5486503.30 frames. , ppl: 9.777744668717315], batch size: 70
+2022-12-12 10:10:41,235 INFO [train.py:421] (2/8) Epoch 7, batch 9800, loss[loss=2.299, over 2310.00 frames. , ppl: 9.962622945572225] tot_loss[loss=2.28, over 5507029.87 frames. , ppl: 9.774379656407346], batch size: 70
+2022-12-12 10:12:17,286 INFO [train.py:421] (2/8) Epoch 7, batch 10000, loss[loss=2.344, over 2310.00 frames. , ppl: 10.4186579369865] tot_loss[loss=2.279, over 5533694.97 frames. , ppl: 9.770694620484429], batch size: 70
+2022-12-12 10:12:17,286 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:12:18,035 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.743937919324686
+2022-12-12 10:13:56,296 INFO [train.py:421] (2/8) Epoch 7, batch 10200, loss[loss=2.375, over 1540.00 frames. , ppl: 10.756301500999909] tot_loss[loss=2.279, over 5562384.67 frames. , ppl: 9.763872543196884], batch size: 70
+2022-12-12 10:15:40,296 INFO [train.py:421] (2/8) Epoch 7, batch 10400, loss[loss=2.363, over 1890.00 frames. , ppl: 10.623709649141784] tot_loss[loss=2.279, over 5588794.94 frames. , ppl: 9.767082682486032], batch size: 70
+2022-12-12 10:17:21,281 INFO [train.py:421] (2/8) Epoch 7, batch 10600, loss[loss=3.15, over 560.00 frames. , ppl: 23.344225328389243] tot_loss[loss=2.281, over 5564349.41 frames. , ppl: 9.786130611189193], batch size: 70
+2022-12-12 10:18:59,616 INFO [train.py:421] (2/8) Epoch 7, batch 10800, loss[loss=2.249, over 8890.00 frames. , ppl: 9.482726148775903] tot_loss[loss=2.281, over 5552629.02 frames. , ppl: 9.789862573407511], batch size: 70
+2022-12-12 10:20:44,651 INFO [train.py:421] (2/8) Epoch 7, batch 11000, loss[loss=2.2, over 4550.00 frames. , ppl: 9.02224009767549] tot_loss[loss=2.281, over 5577862.63 frames. , ppl: 9.790707665952615], batch size: 70
+2022-12-12 10:20:44,652 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:20:45,398 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764081387151732
+2022-12-12 10:22:26,516 INFO [train.py:421] (2/8) Epoch 7, batch 11200, loss[loss=2.26, over 4200.00 frames. , ppl: 9.582683640972608] tot_loss[loss=2.282, over 5562210.34 frames. , ppl: 9.792716343467676], batch size: 70
+2022-12-12 10:24:09,619 INFO [train.py:421] (2/8) Epoch 7, batch 11400, loss[loss=2.336, over 2100.00 frames. , ppl: 10.337552202121303] tot_loss[loss=2.282, over 5550343.76 frames. , ppl: 9.79260338490083], batch size: 70
+2022-12-12 10:25:51,779 INFO [train.py:421] (2/8) Epoch 7, batch 11600, loss[loss=2.399, over 1750.00 frames. , ppl: 11.010327603268152] tot_loss[loss=2.282, over 5546128.71 frames. , ppl: 9.797125660571371], batch size: 70
+2022-12-12 10:27:33,031 INFO [train.py:421] (2/8) Epoch 7, batch 11800, loss[loss=2.288, over 2870.00 frames. , ppl: 9.851766091053044] tot_loss[loss=2.281, over 5556908.38 frames. , ppl: 9.79027018172558], batch size: 70
+2022-12-12 10:29:14,367 INFO [train.py:421] (2/8) Epoch 7, batch 12000, loss[loss=2.464, over 1120.00 frames. , ppl: 11.747869147553025] tot_loss[loss=2.283, over 5510756.31 frames. , ppl: 9.804894130496486], batch size: 70
+2022-12-12 10:29:14,367 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:29:15,131 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.770513007653056
+2022-12-12 10:30:57,656 INFO [train.py:421] (2/8) Epoch 7, batch 12200, loss[loss=2.295, over 3430.00 frames. , ppl: 9.924502696834812] tot_loss[loss=2.283, over 5489168.19 frames. , ppl: 9.805692505654193], batch size: 70
+2022-12-12 10:32:39,434 INFO [train.py:421] (2/8) Epoch 7, batch 12400, loss[loss=2.142, over 7980.00 frames. , ppl: 8.520659001000588] tot_loss[loss=2.283, over 5492609.97 frames. , ppl: 9.80470862141449], batch size: 70
+2022-12-12 10:34:19,731 INFO [train.py:421] (2/8) Epoch 7, batch 12600, loss[loss=2.258, over 3710.00 frames. , ppl: 9.563332129690323] tot_loss[loss=2.283, over 5464797.18 frames. , ppl: 9.808465278584091], batch size: 70
+2022-12-12 10:35:58,824 INFO [train.py:421] (2/8) Epoch 7, batch 12800, loss[loss=2.244, over 4480.00 frames. , ppl: 9.429882377988719] tot_loss[loss=2.283, over 5479906.50 frames. , ppl: 9.803713887461193], batch size: 70
+2022-12-12 10:37:35,199 INFO [train.py:421] (2/8) Epoch 7, batch 13000, loss[loss=2.822, over 630.00 frames. , ppl: 16.80309085199172] tot_loss[loss=2.284, over 5471462.81 frames. , ppl: 9.811700224649128], batch size: 70
+2022-12-12 10:37:35,199 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:37:35,946 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.769163881244294
+2022-12-12 10:39:15,582 INFO [train.py:421] (2/8) Epoch 7, batch 13200, loss[loss=2.486, over 1260.00 frames. , ppl: 12.010996424613808] tot_loss[loss=2.283, over 5472510.73 frames. , ppl: 9.81093011070764], batch size: 70
+2022-12-12 10:40:55,457 INFO [train.py:421] (2/8) Epoch 7, batch 13400, loss[loss=2.75, over 630.00 frames. , ppl: 15.647081957392547] tot_loss[loss=2.284, over 5475708.07 frames. , ppl: 9.813230755323007], batch size: 70
+2022-12-12 10:42:34,882 INFO [train.py:421] (2/8) Epoch 7, batch 13600, loss[loss=2.366, over 1610.00 frames. , ppl: 10.651336239425953] tot_loss[loss=2.283, over 5499585.49 frames. , ppl: 9.807622338663476], batch size: 70
+2022-12-12 10:44:12,329 INFO [train.py:421] (2/8) Epoch 7, batch 13800, loss[loss=2.723, over 700.00 frames. , ppl: 15.223268136119247] tot_loss[loss=2.284, over 5458973.86 frames. , ppl: 9.815573783827297], batch size: 70
+2022-12-12 10:45:49,784 INFO [train.py:421] (2/8) Epoch 7, batch 14000, loss[loss=2.36, over 1120.00 frames. , ppl: 10.594333694409777] tot_loss[loss=2.284, over 5477878.71 frames. , ppl: 9.811524339571053], batch size: 70
+2022-12-12 10:45:49,785 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:45:50,534 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.282, over 211138.00 frames. , ppl: 9.796524022897163
+2022-12-12 10:47:32,823 INFO [train.py:421] (2/8) Epoch 7, batch 14200, loss[loss=2.459, over 1120.00 frames. , ppl: 11.694484781240567] tot_loss[loss=2.283, over 5501924.53 frames. , ppl: 9.806551955660666], batch size: 70
+2022-12-12 10:49:07,720 INFO [train.py:421] (2/8) Epoch 7, batch 14400, loss[loss=2.397, over 1470.00 frames. , ppl: 10.9921084479297] tot_loss[loss=2.283, over 5517084.58 frames. , ppl: 9.802922588074743], batch size: 70
+2022-12-12 10:50:46,721 INFO [train.py:421] (2/8) Epoch 7, batch 14600, loss[loss=2.275, over 2660.00 frames. , ppl: 9.726053135633146] tot_loss[loss=2.283, over 5527001.22 frames. , ppl: 9.806844100789764], batch size: 70
+2022-12-12 10:52:28,728 INFO [train.py:421] (2/8) Epoch 7, batch 14800, loss[loss=2.236, over 1680.00 frames. , ppl: 9.358076519754404] tot_loss[loss=2.283, over 5518605.17 frames. , ppl: 9.805963173778887], batch size: 70
+2022-12-12 10:54:11,273 INFO [train.py:421] (2/8) Epoch 7, batch 15000, loss[loss=2.17, over 7210.00 frames. , ppl: 8.756378148378445] tot_loss[loss=2.284, over 5485741.00 frames. , ppl: 9.815372198122645], batch size: 70
+2022-12-12 10:54:11,274 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 10:54:12,021 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.747404083511375
+2022-12-12 10:55:53,710 INFO [train.py:421] (2/8) Epoch 7, batch 15200, loss[loss=2.297, over 2940.00 frames. , ppl: 9.94175068931798] tot_loss[loss=2.283, over 5496626.42 frames. , ppl: 9.806589138023238], batch size: 70
+2022-12-12 10:57:32,864 INFO [train.py:421] (2/8) Epoch 7, batch 15400, loss[loss=2.353, over 1820.00 frames. , ppl: 10.514365064979593] tot_loss[loss=2.282, over 5503694.65 frames. , ppl: 9.79875712505778], batch size: 70
+2022-12-12 10:59:08,551 INFO [train.py:421] (2/8) Epoch 7, batch 15600, loss[loss=2.339, over 1610.00 frames. , ppl: 10.36632626805861] tot_loss[loss=2.283, over 5506362.52 frames. , ppl: 9.807096127799431], batch size: 70
+2022-12-12 11:00:48,237 INFO [train.py:421] (2/8) Epoch 7, batch 15800, loss[loss=2.459, over 1400.00 frames. , ppl: 11.68863227390717] tot_loss[loss=2.283, over 5495694.83 frames. , ppl: 9.80724817162474], batch size: 70
+2022-12-12 11:02:25,848 INFO [train.py:421] (2/8) Epoch 7, batch 16000, loss[loss=2.756, over 840.00 frames. , ppl: 15.73017651298116] tot_loss[loss=2.284, over 5461053.62 frames. , ppl: 9.817975398343593], batch size: 70
+2022-12-12 11:02:25,849 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:02:26,595 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.759092535215544
+2022-12-12 11:04:07,742 INFO [train.py:421] (2/8) Epoch 7, batch 16200, loss[loss=2.77, over 700.00 frames. , ppl: 15.955111165099966] tot_loss[loss=2.285, over 5462111.56 frames. , ppl: 9.821394965301009], batch size: 70
+2022-12-12 11:05:46,866 INFO [train.py:421] (2/8) Epoch 7, batch 16400, loss[loss=2.484, over 1190.00 frames. , ppl: 11.993501328685431] tot_loss[loss=2.285, over 5453997.97 frames. , ppl: 9.82394596243275], batch size: 70
+2022-12-12 11:07:27,128 INFO [train.py:421] (2/8) Epoch 7, batch 16600, loss[loss=2.449, over 1260.00 frames. , ppl: 11.57157236669885] tot_loss[loss=2.285, over 5464740.05 frames. , ppl: 9.828406118250435], batch size: 70
+2022-12-12 11:09:03,975 INFO [train.py:421] (2/8) Epoch 7, batch 16800, loss[loss=2.753, over 700.00 frames. , ppl: 15.68311103210442] tot_loss[loss=2.285, over 5458873.16 frames. , ppl: 9.82606547540816], batch size: 70
+2022-12-12 11:10:44,194 INFO [train.py:421] (2/8) Epoch 7, batch 17000, loss[loss=2.366, over 2380.00 frames. , ppl: 10.656717254808784] tot_loss[loss=2.286, over 5432813.21 frames. , ppl: 9.8345612572577], batch size: 70
+2022-12-12 11:10:44,194 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:10:44,952 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.755632321970248
+2022-12-12 11:12:25,660 INFO [train.py:421] (2/8) Epoch 7, batch 17200, loss[loss=2.22, over 6230.00 frames. , ppl: 9.210501685205806] tot_loss[loss=2.286, over 5424938.30 frames. , ppl: 9.83801232464667], batch size: 70
+2022-12-12 11:14:04,099 INFO [train.py:421] (2/8) Epoch 7, batch 17400, loss[loss=2.367, over 2520.00 frames. , ppl: 10.666945093769534] tot_loss[loss=2.286, over 5410221.13 frames. , ppl: 9.83163365594701], batch size: 70
+2022-12-12 11:15:49,430 INFO [train.py:421] (2/8) Epoch 7, batch 17600, loss[loss=2.238, over 2940.00 frames. , ppl: 9.37687332068357] tot_loss[loss=2.286, over 5408804.09 frames. , ppl: 9.835997894165176], batch size: 70
+2022-12-12 11:17:26,211 INFO [train.py:421] (2/8) Epoch 7, batch 17800, loss[loss=2.593, over 980.00 frames. , ppl: 13.366639027016312] tot_loss[loss=2.286, over 5405907.44 frames. , ppl: 9.830816857641937], batch size: 70
+2022-12-12 11:19:03,110 INFO [train.py:421] (2/8) Epoch 7, batch 18000, loss[loss=2.423, over 910.00 frames. , ppl: 11.274494302222111] tot_loss[loss=2.286, over 5402165.27 frames. , ppl: 9.84006536698565], batch size: 70
+2022-12-12 11:19:03,110 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:19:03,871 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.756920371505483
+2022-12-12 11:20:45,119 INFO [train.py:421] (2/8) Epoch 7, batch 18200, loss[loss=2.373, over 1610.00 frames. , ppl: 10.733210555761877] tot_loss[loss=2.285, over 5423557.27 frames. , ppl: 9.825461694736255], batch size: 70
+2022-12-12 11:22:26,377 INFO [train.py:421] (2/8) Epoch 7, batch 18400, loss[loss=2.244, over 3990.00 frames. , ppl: 9.433169632991124] tot_loss[loss=2.286, over 5412616.57 frames. , ppl: 9.831054513030647], batch size: 70
+2022-12-12 11:24:05,728 INFO [train.py:421] (2/8) Epoch 7, batch 18600, loss[loss=2.731, over 770.00 frames. , ppl: 15.34973856931291] tot_loss[loss=2.284, over 5460041.80 frames. , ppl: 9.820568038182996], batch size: 70
+2022-12-12 11:25:43,136 INFO [train.py:421] (2/8) Epoch 7, batch 18800, loss[loss=2.194, over 2660.00 frames. , ppl: 8.974580089697142] tot_loss[loss=2.285, over 5451530.59 frames. , ppl: 9.821905660679578], batch size: 70
+2022-12-12 11:27:23,641 INFO [train.py:421] (2/8) Epoch 7, batch 19000, loss[loss=2.409, over 1120.00 frames. , ppl: 11.127756522203205] tot_loss[loss=2.284, over 5461918.48 frames. , ppl: 9.819524365615333], batch size: 70
+2022-12-12 11:27:23,642 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:27:24,402 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.753232844662795
+2022-12-12 11:29:02,273 INFO [train.py:421] (2/8) Epoch 7, batch 19200, loss[loss=2.245, over 4900.00 frames. , ppl: 9.443509184090667] tot_loss[loss=2.284, over 5476331.20 frames. , ppl: 9.812627065518296], batch size: 70
+2022-12-12 11:30:43,913 INFO [train.py:421] (2/8) Epoch 7, batch 19400, loss[loss=2.518, over 1610.00 frames. , ppl: 12.399401220113933] tot_loss[loss=2.283, over 5517048.92 frames. , ppl: 9.801260133162893], batch size: 70
+2022-12-12 11:32:24,645 INFO [train.py:421] (2/8) Epoch 7, batch 19600, loss[loss=2.306, over 1680.00 frames. , ppl: 10.03569829105291] tot_loss[loss=2.284, over 5499841.79 frames. , ppl: 9.81207274677833], batch size: 70
+2022-12-12 11:34:05,671 INFO [train.py:421] (2/8) Epoch 7, batch 19800, loss[loss=2.23, over 3290.00 frames. , ppl: 9.30139052825647] tot_loss[loss=2.284, over 5495925.51 frames. , ppl: 9.813569571718931], batch size: 70
+2022-12-12 11:35:44,817 INFO [train.py:421] (2/8) Epoch 7, batch 20000, loss[loss=2.545, over 980.00 frames. , ppl: 12.73935371894883] tot_loss[loss=2.284, over 5470666.42 frames. , ppl: 9.816744741431554], batch size: 70
+2022-12-12 11:35:44,818 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:35:45,578 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.73576853706518
+2022-12-12 11:37:27,966 INFO [train.py:421] (2/8) Epoch 7, batch 20200, loss[loss=2.87, over 630.00 frames. , ppl: 17.643118614058498] tot_loss[loss=2.284, over 5462746.34 frames. , ppl: 9.813070095770458], batch size: 70
+2022-12-12 11:39:05,108 INFO [train.py:421] (2/8) Epoch 7, batch 20400, loss[loss=2.355, over 1890.00 frames. , ppl: 10.538641797737277] tot_loss[loss=2.283, over 5475094.46 frames. , ppl: 9.810113187128765], batch size: 70
+2022-12-12 11:40:47,485 INFO [train.py:421] (2/8) Epoch 7, batch 20600, loss[loss=2.167, over 4130.00 frames. , ppl: 8.732475891310806] tot_loss[loss=2.284, over 5487395.53 frames. , ppl: 9.817676863406959], batch size: 70
+2022-12-12 11:42:27,078 INFO [train.py:421] (2/8) Epoch 7, batch 20800, loss[loss=2.533, over 1050.00 frames. , ppl: 12.59699667400344] tot_loss[loss=2.286, over 5431414.18 frames. , ppl: 9.837062701808627], batch size: 70
+2022-12-12 11:44:06,627 INFO [train.py:421] (2/8) Epoch 7, batch 21000, loss[loss=2.327, over 2030.00 frames. , ppl: 10.247422872735909] tot_loss[loss=2.286, over 5438250.34 frames. , ppl: 9.834506219792862], batch size: 70
+2022-12-12 11:44:06,628 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:44:07,376 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.750929208941372
+2022-12-12 11:45:45,791 INFO [train.py:421] (2/8) Epoch 7, batch 21200, loss[loss=2.305, over 1540.00 frames. , ppl: 10.025932522683636] tot_loss[loss=2.287, over 5414087.19 frames. , ppl: 9.843763702698842], batch size: 70
+2022-12-12 11:47:26,537 INFO [train.py:421] (2/8) Epoch 7, batch 21400, loss[loss=2.269, over 2240.00 frames. , ppl: 9.6654190531073] tot_loss[loss=2.285, over 5476421.34 frames. , ppl: 9.825640367487471], batch size: 70
+2022-12-12 11:49:07,973 INFO [train.py:421] (2/8) Epoch 7, batch 21600, loss[loss=2.402, over 1400.00 frames. , ppl: 11.047572199367199] tot_loss[loss=2.284, over 5489750.24 frames. , ppl: 9.819686976540064], batch size: 70
+2022-12-12 11:50:45,599 INFO [train.py:421] (2/8) Epoch 7, batch 21800, loss[loss=2.284, over 2660.00 frames. , ppl: 9.81252518385071] tot_loss[loss=2.284, over 5503761.16 frames. , ppl: 9.81949483415364], batch size: 70
+2022-12-12 11:52:26,437 INFO [train.py:421] (2/8) Epoch 7, batch 22000, loss[loss=2.327, over 2170.00 frames. , ppl: 10.252132513955438] tot_loss[loss=2.284, over 5521734.78 frames. , ppl: 9.812541575512283], batch size: 70
+2022-12-12 11:52:26,437 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 11:52:27,187 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.739122243408765
+2022-12-12 11:54:04,986 INFO [train.py:421] (2/8) Epoch 7, batch 22200, loss[loss=2.356, over 3150.00 frames. , ppl: 10.54786583679189] tot_loss[loss=2.284, over 5513462.20 frames. , ppl: 9.813839648913865], batch size: 70
+2022-12-12 11:55:47,110 INFO [train.py:421] (2/8) Epoch 7, batch 22400, loss[loss=2.428, over 1400.00 frames. , ppl: 11.340649306027002] tot_loss[loss=2.284, over 5519885.30 frames. , ppl: 9.815451398359725], batch size: 70
+2022-12-12 11:57:29,542 INFO [train.py:421] (2/8) Epoch 7, batch 22600, loss[loss=2.619, over 840.00 frames. , ppl: 13.727945788295358] tot_loss[loss=2.286, over 5491937.76 frames. , ppl: 9.831256088589178], batch size: 70
+2022-12-12 11:59:10,044 INFO [train.py:421] (2/8) Epoch 7, batch 22800, loss[loss=2.314, over 1960.00 frames. , ppl: 10.112235372435578] tot_loss[loss=2.285, over 5490316.94 frames. , ppl: 9.82983387257714], batch size: 70
+2022-12-12 12:00:50,386 INFO [train.py:421] (2/8) Epoch 7, batch 23000, loss[loss=2.263, over 2800.00 frames. , ppl: 9.607556021352698] tot_loss[loss=2.284, over 5515074.04 frames. , ppl: 9.81548684779676], batch size: 70
+2022-12-12 12:00:50,387 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:00:51,133 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.741689827028287
+2022-12-12 12:02:30,417 INFO [train.py:421] (2/8) Epoch 7, batch 23200, loss[loss=2.189, over 5460.00 frames. , ppl: 8.927530049989238] tot_loss[loss=2.284, over 5521535.77 frames. , ppl: 9.814186256898848], batch size: 70
+2022-12-12 12:04:11,347 INFO [train.py:421] (2/8) Epoch 7, batch 23400, loss[loss=2.165, over 7140.00 frames. , ppl: 8.712111861955734] tot_loss[loss=2.283, over 5535206.74 frames. , ppl: 9.806421906686655], batch size: 70
+2022-12-12 12:05:50,823 INFO [train.py:421] (2/8) Epoch 7, batch 23600, loss[loss=2.164, over 7910.00 frames. , ppl: 8.708125696841543] tot_loss[loss=2.283, over 5528463.82 frames. , ppl: 9.807393277139607], batch size: 70
+2022-12-12 12:07:34,136 INFO [train.py:421] (2/8) Epoch 7, batch 23800, loss[loss=2.581, over 910.00 frames. , ppl: 13.215545175649455] tot_loss[loss=2.283, over 5530790.96 frames. , ppl: 9.807282776410775], batch size: 70
+2022-12-12 12:09:14,676 INFO [train.py:421] (2/8) Epoch 7, batch 24000, loss[loss=2.4, over 1540.00 frames. , ppl: 11.023522398165653] tot_loss[loss=2.283, over 5540390.94 frames. , ppl: 9.806587538902438], batch size: 70
+2022-12-12 12:09:14,676 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:09:15,425 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.74665824206101
+2022-12-12 12:10:54,504 INFO [train.py:421] (2/8) Epoch 7, batch 24200, loss[loss=2.152, over 2730.00 frames. , ppl: 8.598645668006688] tot_loss[loss=2.283, over 5513765.60 frames. , ppl: 9.810063545862624], batch size: 70
+2022-12-12 12:12:37,649 INFO [train.py:421] (2/8) Epoch 7, batch 24400, loss[loss=2.199, over 6580.00 frames. , ppl: 9.01953686801547] tot_loss[loss=2.284, over 5514803.78 frames. , ppl: 9.813168884580172], batch size: 70
+2022-12-12 12:14:18,797 INFO [train.py:421] (2/8) Epoch 7, batch 24600, loss[loss=2.406, over 1400.00 frames. , ppl: 11.085443456634822] tot_loss[loss=2.284, over 5508656.38 frames. , ppl: 9.811456345368141], batch size: 70
+2022-12-12 12:15:58,738 INFO [train.py:421] (2/8) Epoch 7, batch 24800, loss[loss=2.275, over 3360.00 frames. , ppl: 9.723377923117301] tot_loss[loss=2.282, over 5536682.79 frames. , ppl: 9.798120016283029], batch size: 70
+2022-12-12 12:17:36,489 INFO [train.py:421] (2/8) Epoch 7, batch 25000, loss[loss=2.209, over 2940.00 frames. , ppl: 9.107872628028066] tot_loss[loss=2.284, over 5478716.50 frames. , ppl: 9.814477430699611], batch size: 70
+2022-12-12 12:17:36,489 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:17:37,246 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.73495874762003
+2022-12-12 12:19:18,780 INFO [train.py:421] (2/8) Epoch 7, batch 25200, loss[loss=2.321, over 2310.00 frames. , ppl: 10.181832919999945] tot_loss[loss=2.284, over 5483494.30 frames. , ppl: 9.819935232849645], batch size: 70
+2022-12-12 12:20:57,119 INFO [train.py:421] (2/8) Epoch 7, batch 25400, loss[loss=2.44, over 1190.00 frames. , ppl: 11.473498332818682] tot_loss[loss=2.286, over 5422089.56 frames. , ppl: 9.83509723772351], batch size: 70
+2022-12-12 12:22:35,788 INFO [train.py:421] (2/8) Epoch 7, batch 25600, loss[loss=2.645, over 770.00 frames. , ppl: 14.089586763236499] tot_loss[loss=2.287, over 5385875.54 frames. , ppl: 9.843121086235046], batch size: 70
+2022-12-12 12:24:16,860 INFO [train.py:421] (2/8) Epoch 7, batch 25800, loss[loss=2.329, over 2870.00 frames. , ppl: 10.267701849432449] tot_loss[loss=2.287, over 5408375.63 frames. , ppl: 9.840771632972409], batch size: 70
+2022-12-12 12:25:56,646 INFO [train.py:421] (2/8) Epoch 7, batch 26000, loss[loss=2.317, over 2240.00 frames. , ppl: 10.141527415984479] tot_loss[loss=2.286, over 5411397.95 frames. , ppl: 9.83641281025702], batch size: 70
+2022-12-12 12:25:56,647 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:25:57,404 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.75006332120515
+2022-12-12 12:27:34,574 INFO [train.py:421] (2/8) Epoch 7, batch 26200, loss[loss=2.472, over 1190.00 frames. , ppl: 11.851941709838929] tot_loss[loss=2.285, over 5442192.37 frames. , ppl: 9.829277513285703], batch size: 70
+2022-12-12 12:29:16,203 INFO [train.py:421] (2/8) Epoch 7, batch 26400, loss[loss=2.39, over 1750.00 frames. , ppl: 10.913707099302147] tot_loss[loss=2.285, over 5430028.57 frames. , ppl: 9.830059318612365], batch size: 70
+2022-12-12 12:30:55,036 INFO [train.py:421] (2/8) Epoch 7, batch 26600, loss[loss=2.373, over 2380.00 frames. , ppl: 10.724194744079789] tot_loss[loss=2.286, over 5418600.59 frames. , ppl: 9.83327744939549], batch size: 70
+2022-12-12 12:32:37,070 INFO [train.py:421] (2/8) Epoch 7, batch 26800, loss[loss=2.345, over 2170.00 frames. , ppl: 10.436759080756351] tot_loss[loss=2.287, over 5429852.29 frames. , ppl: 9.841923484932307], batch size: 70
+2022-12-12 12:34:13,501 INFO [train.py:421] (2/8) Epoch 7, batch 27000, loss[loss=2.583, over 840.00 frames. , ppl: 13.231684325065636] tot_loss[loss=2.288, over 5412021.27 frames. , ppl: 9.85439412758391], batch size: 70
+2022-12-12 12:34:13,501 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:34:14,248 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.741972432318025
+2022-12-12 12:35:56,492 INFO [train.py:421] (2/8) Epoch 7, batch 27200, loss[loss=2.38, over 910.00 frames. , ppl: 10.809835442508135] tot_loss[loss=2.287, over 5436457.89 frames. , ppl: 9.84402464179575], batch size: 70
+2022-12-12 12:37:39,587 INFO [train.py:421] (2/8) Epoch 7, batch 27400, loss[loss=2.135, over 4200.00 frames. , ppl: 8.455701603624577] tot_loss[loss=2.285, over 5456509.25 frames. , ppl: 9.829205388611467], batch size: 70
+2022-12-12 12:39:19,441 INFO [train.py:421] (2/8) Epoch 7, batch 27600, loss[loss=2.458, over 1190.00 frames. , ppl: 11.683170331150404] tot_loss[loss=2.286, over 5424229.00 frames. , ppl: 9.83591994232469], batch size: 70
+2022-12-12 12:41:01,763 INFO [train.py:421] (2/8) Epoch 7, batch 27800, loss[loss=2.2, over 4200.00 frames. , ppl: 9.025999824798465] tot_loss[loss=2.286, over 5429405.79 frames. , ppl: 9.836116121544647], batch size: 70
+2022-12-12 12:42:41,587 INFO [train.py:421] (2/8) Epoch 7, batch 28000, loss[loss=2.563, over 980.00 frames. , ppl: 12.975804561801372] tot_loss[loss=2.285, over 5452144.11 frames. , ppl: 9.828292500976227], batch size: 70
+2022-12-12 12:42:41,588 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:42:42,348 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.745110478587499
+2022-12-12 12:44:19,017 INFO [train.py:421] (2/8) Epoch 7, batch 28200, loss[loss=2.314, over 1470.00 frames. , ppl: 10.116752451264253] tot_loss[loss=2.285, over 5454646.31 frames. , ppl: 9.82993575910985], batch size: 70
+2022-12-12 12:45:59,452 INFO [train.py:421] (2/8) Epoch 7, batch 28400, loss[loss=2.452, over 1120.00 frames. , ppl: 11.60795793534174] tot_loss[loss=2.285, over 5457647.37 frames. , ppl: 9.829649546770705], batch size: 70
+2022-12-12 12:47:41,287 INFO [train.py:421] (2/8) Epoch 7, batch 28600, loss[loss=2.488, over 910.00 frames. , ppl: 12.03596128158188] tot_loss[loss=2.286, over 5436837.38 frames. , ppl: 9.838160996453036], batch size: 70
+2022-12-12 12:49:21,379 INFO [train.py:421] (2/8) Epoch 7, batch 28800, loss[loss=2.194, over 5810.00 frames. , ppl: 8.96721841930245] tot_loss[loss=2.285, over 5463916.14 frames. , ppl: 9.829271338393111], batch size: 70
+2022-12-12 12:50:59,772 INFO [train.py:421] (2/8) Epoch 7, batch 29000, loss[loss=2.222, over 5810.00 frames. , ppl: 9.22835912763616] tot_loss[loss=2.285, over 5439895.78 frames. , ppl: 9.828399890482125], batch size: 70
+2022-12-12 12:50:59,772 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:51:00,533 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.757074891112701
+2022-12-12 12:52:40,866 INFO [train.py:421] (2/8) Epoch 7, batch 29200, loss[loss=2.359, over 2870.00 frames. , ppl: 10.576032407387116] tot_loss[loss=2.285, over 5456822.91 frames. , ppl: 9.824439026518423], batch size: 70
+2022-12-12 12:54:23,366 INFO [train.py:421] (2/8) Epoch 7, batch 29400, loss[loss=2.226, over 4620.00 frames. , ppl: 9.26195605250457] tot_loss[loss=2.284, over 5494320.97 frames. , ppl: 9.81235750815133], batch size: 70
+2022-12-12 12:56:02,431 INFO [train.py:421] (2/8) Epoch 7, batch 29600, loss[loss=2.27, over 2590.00 frames. , ppl: 9.681998968837778] tot_loss[loss=2.283, over 5512575.73 frames. , ppl: 9.8059666696492], batch size: 70
+2022-12-12 12:57:39,009 INFO [train.py:421] (2/8) Epoch 7, batch 29800, loss[loss=2.429, over 1190.00 frames. , ppl: 11.351638740635606] tot_loss[loss=2.285, over 5442959.47 frames. , ppl: 9.825578444735964], batch size: 70
+2022-12-12 12:59:14,357 INFO [train.py:421] (2/8) Epoch 7, batch 30000, loss[loss=2.412, over 1400.00 frames. , ppl: 11.154627368412958] tot_loss[loss=2.286, over 5404553.90 frames. , ppl: 9.834828105651503], batch size: 70
+2022-12-12 12:59:14,357 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 12:59:15,118 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.757277069995297
+2022-12-12 13:00:54,968 INFO [train.py:421] (2/8) Epoch 7, batch 30200, loss[loss=2.361, over 1540.00 frames. , ppl: 10.60456961585665] tot_loss[loss=2.285, over 5415345.22 frames. , ppl: 9.829825280285133], batch size: 70
+2022-12-12 13:02:35,774 INFO [train.py:421] (2/8) Epoch 7, batch 30400, loss[loss=2.801, over 700.00 frames. , ppl: 16.461620949962423] tot_loss[loss=2.285, over 5426922.01 frames. , ppl: 9.825334252088542], batch size: 70
+2022-12-12 13:04:16,900 INFO [train.py:421] (2/8) Epoch 7, batch 30600, loss[loss=2.319, over 2800.00 frames. , ppl: 10.161764668932085] tot_loss[loss=2.286, over 5404162.07 frames. , ppl: 9.834797228617026], batch size: 70
+2022-12-12 13:05:54,569 INFO [train.py:421] (2/8) Epoch 7, batch 30800, loss[loss=2.375, over 1820.00 frames. , ppl: 10.752259297761654] tot_loss[loss=2.285, over 5437686.73 frames. , ppl: 9.821625580054318], batch size: 70
+2022-12-12 13:07:37,485 INFO [train.py:421] (2/8) Epoch 7, batch 31000, loss[loss=2.258, over 3080.00 frames. , ppl: 9.5618034484259] tot_loss[loss=2.285, over 5414225.99 frames. , ppl: 9.829948598989917], batch size: 70
+2022-12-12 13:07:37,486 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:07:38,249 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.74524606034524
+2022-12-12 13:09:21,953 INFO [train.py:421] (2/8) Epoch 7, batch 31200, loss[loss=3.179, over 490.00 frames. , ppl: 24.031695500363405] tot_loss[loss=2.288, over 5349344.74 frames. , ppl: 9.85074478620987], batch size: 70
+2022-12-12 13:10:57,985 INFO [train.py:421] (2/8) Epoch 7, batch 31400, loss[loss=2.489, over 840.00 frames. , ppl: 12.05440162729975] tot_loss[loss=2.288, over 5351433.95 frames. , ppl: 9.851048830161549], batch size: 70
+2022-12-12 13:12:37,920 INFO [train.py:421] (2/8) Epoch 7, batch 31600, loss[loss=2.7, over 770.00 frames. , ppl: 14.875656057943493] tot_loss[loss=2.287, over 5383231.88 frames. , ppl: 9.843207278147148], batch size: 70
+2022-12-12 13:14:17,600 INFO [train.py:421] (2/8) Epoch 7, batch 31800, loss[loss=2.308, over 1540.00 frames. , ppl: 10.055472781379423] tot_loss[loss=2.286, over 5397779.12 frames. , ppl: 9.838695365619161], batch size: 70
+2022-12-12 13:15:58,458 INFO [train.py:421] (2/8) Epoch 7, batch 32000, loss[loss=3.25, over 490.00 frames. , ppl: 25.784551669683708] tot_loss[loss=2.286, over 5436626.19 frames. , ppl: 9.831727714290228], batch size: 70
+2022-12-12 13:15:58,458 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:15:59,205 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.747510843049605
+2022-12-12 13:17:36,125 INFO [train.py:421] (2/8) Epoch 7, batch 32200, loss[loss=2.323, over 1680.00 frames. , ppl: 10.203551912068034] tot_loss[loss=2.284, over 5470918.80 frames. , ppl: 9.818013389525776], batch size: 70
+2022-12-12 13:19:19,499 INFO [train.py:421] (2/8) Epoch 7, batch 32400, loss[loss=2.453, over 1050.00 frames. , ppl: 11.625646794586183] tot_loss[loss=2.283, over 5481602.28 frames. , ppl: 9.809910168358634], batch size: 70
+2022-12-12 13:20:59,435 INFO [train.py:421] (2/8) Epoch 7, batch 32600, loss[loss=2.476, over 1400.00 frames. , ppl: 11.889917161230573] tot_loss[loss=2.283, over 5479982.20 frames. , ppl: 9.80684698075678], batch size: 70
+2022-12-12 13:22:38,744 INFO [train.py:421] (2/8) Epoch 7, batch 32800, loss[loss=2.376, over 1680.00 frames. , ppl: 10.757427009954187] tot_loss[loss=2.283, over 5501248.86 frames. , ppl: 9.803552655143292], batch size: 70
+2022-12-12 13:24:19,352 INFO [train.py:421] (2/8) Epoch 7, batch 33000, loss[loss=2.376, over 1960.00 frames. , ppl: 10.76469205455338] tot_loss[loss=2.285, over 5431579.11 frames. , ppl: 9.821292748830041], batch size: 70
+2022-12-12 13:24:19,353 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:24:20,118 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.737587206525212
+2022-12-12 13:26:00,217 INFO [train.py:421] (2/8) Epoch 7, batch 33200, loss[loss=2.19, over 2310.00 frames. , ppl: 8.932028956227066] tot_loss[loss=2.284, over 5470001.27 frames. , ppl: 9.818176124283186], batch size: 70
+2022-12-12 13:27:39,637 INFO [train.py:421] (2/8) Epoch 7, batch 33400, loss[loss=3.186, over 490.00 frames. , ppl: 24.193451355317872] tot_loss[loss=2.284, over 5496954.60 frames. , ppl: 9.81263835942126], batch size: 70
+2022-12-12 13:29:21,157 INFO [train.py:421] (2/8) Epoch 7, batch 33600, loss[loss=2.322, over 1960.00 frames. , ppl: 10.192440366864815] tot_loss[loss=2.283, over 5497705.98 frames. , ppl: 9.807865340903422], batch size: 70
+2022-12-12 13:31:03,137 INFO [train.py:421] (2/8) Epoch 7, batch 33800, loss[loss=2.407, over 1610.00 frames. , ppl: 11.100441120533281] tot_loss[loss=2.282, over 5536455.99 frames. , ppl: 9.797952364531195], batch size: 70
+2022-12-12 13:32:47,145 INFO [train.py:421] (2/8) Epoch 7, batch 34000, loss[loss=2.385, over 1610.00 frames. , ppl: 10.857841563273201] tot_loss[loss=2.282, over 5523346.77 frames. , ppl: 9.796138750275603], batch size: 70
+2022-12-12 13:32:47,145 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:32:47,894 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.749546711482623
+2022-12-12 13:34:29,525 INFO [train.py:421] (2/8) Epoch 7, batch 34200, loss[loss=2.355, over 2030.00 frames. , ppl: 10.538062451090344] tot_loss[loss=2.281, over 5549963.11 frames. , ppl: 9.785807734649817], batch size: 70
+2022-12-12 13:36:09,713 INFO [train.py:421] (2/8) Epoch 7, batch 34400, loss[loss=2.234, over 3500.00 frames. , ppl: 9.338377607718662] tot_loss[loss=2.281, over 5543957.43 frames. , ppl: 9.784172973023779], batch size: 70
+2022-12-12 13:37:49,610 INFO [train.py:421] (2/8) Epoch 7, batch 34600, loss[loss=2.648, over 910.00 frames. , ppl: 14.125700950053597] tot_loss[loss=2.281, over 5557522.47 frames. , ppl: 9.787917233595469], batch size: 70
+2022-12-12 13:39:31,007 INFO [train.py:421] (2/8) Epoch 7, batch 34800, loss[loss=2.567, over 1400.00 frames. , ppl: 13.023964978465658] tot_loss[loss=2.281, over 5566773.70 frames. , ppl: 9.783426586479203], batch size: 70
+2022-12-12 13:41:13,650 INFO [train.py:421] (2/8) Epoch 7, batch 35000, loss[loss=2.227, over 3290.00 frames. , ppl: 9.271584681861253] tot_loss[loss=2.281, over 5559320.71 frames. , ppl: 9.784949033950971], batch size: 70
+2022-12-12 13:41:13,651 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:41:14,435 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.743404328984035
+2022-12-12 13:42:53,064 INFO [train.py:421] (2/8) Epoch 7, batch 35200, loss[loss=2.444, over 1680.00 frames. , ppl: 11.520674925014374] tot_loss[loss=2.28, over 5569323.30 frames. , ppl: 9.77812875391881], batch size: 70
+2022-12-12 13:44:36,901 INFO [train.py:421] (2/8) Epoch 7, batch 35400, loss[loss=2.42, over 1400.00 frames. , ppl: 11.246753622175888] tot_loss[loss=2.279, over 5636925.73 frames. , ppl: 9.765235023030597], batch size: 70
+2022-12-12 13:46:17,869 INFO [train.py:421] (2/8) Epoch 7, batch 35600, loss[loss=2.255, over 2450.00 frames. , ppl: 9.533269631549755] tot_loss[loss=2.279, over 5611251.55 frames. , ppl: 9.771687769226302], batch size: 70
+2022-12-12 13:48:00,243 INFO [train.py:421] (2/8) Epoch 7, batch 35800, loss[loss=2.287, over 2170.00 frames. , ppl: 9.846398706419436] tot_loss[loss=2.28, over 5584161.70 frames. , ppl: 9.776075999989924], batch size: 70
+2022-12-12 13:49:39,568 INFO [train.py:421] (2/8) Epoch 7, batch 36000, loss[loss=2.276, over 3150.00 frames. , ppl: 9.739216902670826] tot_loss[loss=2.281, over 5556868.05 frames. , ppl: 9.79047134595769], batch size: 70
+2022-12-12 13:49:39,569 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:49:40,314 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730916571608322
+2022-12-12 13:51:17,495 INFO [train.py:421] (2/8) Epoch 7, batch 36200, loss[loss=2.821, over 560.00 frames. , ppl: 16.789492780615788] tot_loss[loss=2.283, over 5508323.56 frames. , ppl: 9.80192943923743], batch size: 70
+2022-12-12 13:52:56,019 INFO [train.py:421] (2/8) Epoch 7, batch 36400, loss[loss=2.236, over 5810.00 frames. , ppl: 9.352801861854768] tot_loss[loss=2.282, over 5505453.98 frames. , ppl: 9.800812761516498], batch size: 70
+2022-12-12 13:54:35,736 INFO [train.py:421] (2/8) Epoch 7, batch 36600, loss[loss=2.314, over 2590.00 frames. , ppl: 10.11770104299676] tot_loss[loss=2.282, over 5484758.22 frames. , ppl: 9.800112941670628], batch size: 70
+2022-12-12 13:56:13,084 INFO [train.py:421] (2/8) Epoch 7, batch 36800, loss[loss=2.439, over 1190.00 frames. , ppl: 11.457159752434253] tot_loss[loss=2.282, over 5521364.04 frames. , ppl: 9.798159892361435], batch size: 70
+2022-12-12 13:57:51,835 INFO [train.py:421] (2/8) Epoch 7, batch 37000, loss[loss=2.26, over 1260.00 frames. , ppl: 9.587554798693015] tot_loss[loss=2.283, over 5521568.22 frames. , ppl: 9.806008373159582], batch size: 70
+2022-12-12 13:57:51,835 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 13:57:52,594 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730660210790171
+2022-12-12 13:59:32,625 INFO [train.py:421] (2/8) Epoch 7, batch 37200, loss[loss=2.438, over 2520.00 frames. , ppl: 11.447410152542114] tot_loss[loss=2.284, over 5485906.83 frames. , ppl: 9.81657043467916], batch size: 70
+2022-12-12 14:01:13,834 INFO [train.py:421] (2/8) Epoch 7, batch 37400, loss[loss=2.413, over 2380.00 frames. , ppl: 11.170031950666573] tot_loss[loss=2.284, over 5492468.18 frames. , ppl: 9.817836396911853], batch size: 70
+2022-12-12 14:03:00,115 INFO [train.py:421] (2/8) Epoch 7, batch 37600, loss[loss=2.268, over 3850.00 frames. , ppl: 9.656128937963981] tot_loss[loss=2.282, over 5549179.03 frames. , ppl: 9.800139787932522], batch size: 70
+2022-12-12 14:04:41,219 INFO [train.py:421] (2/8) Epoch 7, batch 37800, loss[loss=2.213, over 6020.00 frames. , ppl: 9.14428939482946] tot_loss[loss=2.281, over 5577382.23 frames. , ppl: 9.789165541094361], batch size: 70
+2022-12-12 14:06:18,543 INFO [train.py:421] (2/8) Epoch 7, batch 38000, loss[loss=2.14, over 11550.00 frames. , ppl: 8.49797174029765] tot_loss[loss=2.283, over 5514975.49 frames. , ppl: 9.805916249404948], batch size: 70
+2022-12-12 14:06:18,544 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:06:19,290 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.735293029761026
+2022-12-12 14:08:00,975 INFO [train.py:421] (2/8) Epoch 7, batch 38200, loss[loss=2.48, over 1470.00 frames. , ppl: 11.937382678310307] tot_loss[loss=2.282, over 5545253.28 frames. , ppl: 9.79701013820148], batch size: 70
+2022-12-12 14:09:41,637 INFO [train.py:421] (2/8) Epoch 7, batch 38400, loss[loss=2.329, over 2800.00 frames. , ppl: 10.26982189899046] tot_loss[loss=2.282, over 5557077.48 frames. , ppl: 9.794055849346035], batch size: 70
+2022-12-12 14:11:19,544 INFO [train.py:421] (2/8) Epoch 7, batch 38600, loss[loss=2.27, over 3640.00 frames. , ppl: 9.679356148328218] tot_loss[loss=2.282, over 5551105.10 frames. , ppl: 9.79933177409452], batch size: 70
+2022-12-12 14:12:58,079 INFO [train.py:421] (2/8) Epoch 7, batch 38800, loss[loss=2.321, over 1960.00 frames. , ppl: 10.18811961185195] tot_loss[loss=2.283, over 5499162.02 frames. , ppl: 9.81062967332715], batch size: 70
+2022-12-12 14:14:39,697 INFO [train.py:421] (2/8) Epoch 7, batch 39000, loss[loss=2.522, over 1050.00 frames. , ppl: 12.45816588250607] tot_loss[loss=2.283, over 5508276.51 frames. , ppl: 9.806093128856295], batch size: 70
+2022-12-12 14:14:39,697 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:14:40,458 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.733431570068024
+2022-12-12 14:16:18,497 INFO [train.py:421] (2/8) Epoch 7, batch 39200, loss[loss=2.336, over 1750.00 frames. , ppl: 10.339451247959245] tot_loss[loss=2.284, over 5512572.47 frames. , ppl: 9.811112908736941], batch size: 70
+2022-12-12 14:17:57,353 INFO [train.py:421] (2/8) Epoch 7, batch 39400, loss[loss=2.385, over 1820.00 frames. , ppl: 10.859413436582283] tot_loss[loss=2.284, over 5491827.30 frames. , ppl: 9.815847866639446], batch size: 70
+2022-12-12 14:19:39,321 INFO [train.py:421] (2/8) Epoch 7, batch 39600, loss[loss=2.437, over 1680.00 frames. , ppl: 11.434022461478266] tot_loss[loss=2.284, over 5472750.60 frames. , ppl: 9.819709971128368], batch size: 70
+2022-12-12 14:21:22,520 INFO [train.py:421] (2/8) Epoch 7, batch 39800, loss[loss=2.322, over 2590.00 frames. , ppl: 10.198566278642819] tot_loss[loss=2.285, over 5465827.99 frames. , ppl: 9.821220827229892], batch size: 70
+2022-12-12 14:23:03,815 INFO [train.py:421] (2/8) Epoch 7, batch 40000, loss[loss=2.656, over 910.00 frames. , ppl: 14.232907996065672] tot_loss[loss=2.285, over 5458034.03 frames. , ppl: 9.826569241824638], batch size: 70
+2022-12-12 14:23:03,816 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:23:04,577 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.732923044306935
+2022-12-12 14:24:40,622 INFO [train.py:421] (2/8) Epoch 7, batch 40200, loss[loss=2.334, over 2450.00 frames. , ppl: 10.317855691502755] tot_loss[loss=2.284, over 5531465.04 frames. , ppl: 9.811847545315633], batch size: 70
+2022-12-12 14:26:24,154 INFO [train.py:421] (2/8) Epoch 7, batch 40400, loss[loss=2.314, over 2240.00 frames. , ppl: 10.113073983007347] tot_loss[loss=2.283, over 5551476.03 frames. , ppl: 9.804906172062022], batch size: 70
+2022-12-12 14:28:05,233 INFO [train.py:421] (2/8) Epoch 7, batch 40600, loss[loss=2.366, over 2590.00 frames. , ppl: 10.651352041912864] tot_loss[loss=2.284, over 5523663.16 frames. , ppl: 9.813264743647942], batch size: 70
+2022-12-12 14:29:48,173 INFO [train.py:421] (2/8) Epoch 7, batch 40800, loss[loss=2.541, over 1540.00 frames. , ppl: 12.687332739997371] tot_loss[loss=2.284, over 5529611.49 frames. , ppl: 9.81918137061427], batch size: 70
+2022-12-12 14:31:28,831 INFO [train.py:421] (2/8) Epoch 7, batch 41000, loss[loss=3.393, over 490.00 frames. , ppl: 29.74081451278262] tot_loss[loss=2.283, over 5548017.26 frames. , ppl: 9.809734959371221], batch size: 70
+2022-12-12 14:31:28,832 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:31:29,592 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.72567260083667
+2022-12-12 14:33:15,369 INFO [train.py:421] (2/8) Epoch 7, batch 41200, loss[loss=2.35, over 2240.00 frames. , ppl: 10.486945790044107] tot_loss[loss=2.284, over 5539248.46 frames. , ppl: 9.816015981546542], batch size: 70
+2022-12-12 14:34:58,545 INFO [train.py:421] (2/8) Epoch 7, batch 41400, loss[loss=2.438, over 1470.00 frames. , ppl: 11.449534634330254] tot_loss[loss=2.283, over 5553438.20 frames. , ppl: 9.806702845426889], batch size: 70
+2022-12-12 14:36:43,116 INFO [train.py:421] (2/8) Epoch 7, batch 41600, loss[loss=2.791, over 770.00 frames. , ppl: 16.294308475894685] tot_loss[loss=2.283, over 5572057.72 frames. , ppl: 9.805235330769463], batch size: 70
+2022-12-12 14:38:19,985 INFO [train.py:421] (2/8) Epoch 7, batch 41800, loss[loss=2.239, over 1960.00 frames. , ppl: 9.382768890246206] tot_loss[loss=2.284, over 5523829.55 frames. , ppl: 9.814474635844132], batch size: 70
+2022-12-12 14:39:58,260 INFO [train.py:421] (2/8) Epoch 7, batch 42000, loss[loss=2.387, over 1330.00 frames. , ppl: 10.885205204169585] tot_loss[loss=2.284, over 5527247.83 frames. , ppl: 9.812547265449457], batch size: 70
+2022-12-12 14:39:58,260 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:39:59,022 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.732604689011115
+2022-12-12 14:41:41,114 INFO [train.py:421] (2/8) Epoch 7, batch 42200, loss[loss=2.177, over 4340.00 frames. , ppl: 8.81860389788757] tot_loss[loss=2.283, over 5549682.30 frames. , ppl: 9.804923998163734], batch size: 70
+2022-12-12 14:43:22,131 INFO [train.py:421] (2/8) Epoch 7, batch 42400, loss[loss=2.467, over 1610.00 frames. , ppl: 11.781372433462618] tot_loss[loss=2.283, over 5570871.80 frames. , ppl: 9.805939503557328], batch size: 70
+2022-12-12 14:45:04,978 INFO [train.py:421] (2/8) Epoch 7, batch 42600, loss[loss=2.425, over 1260.00 frames. , ppl: 11.300746922325397] tot_loss[loss=2.282, over 5578738.70 frames. , ppl: 9.797706768039163], batch size: 70
+2022-12-12 14:46:47,201 INFO [train.py:421] (2/8) Epoch 7, batch 42800, loss[loss=2.198, over 3220.00 frames. , ppl: 9.007532116595982] tot_loss[loss=2.283, over 5544632.75 frames. , ppl: 9.80431270366755], batch size: 70
+2022-12-12 14:48:26,742 INFO [train.py:421] (2/8) Epoch 7, batch 43000, loss[loss=2.476, over 1050.00 frames. , ppl: 11.895622554864085] tot_loss[loss=2.284, over 5487282.06 frames. , ppl: 9.817974942129904], batch size: 70
+2022-12-12 14:48:26,743 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:48:27,491 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.722768179823841
+2022-12-12 14:50:02,821 INFO [train.py:421] (2/8) Epoch 7, batch 43200, loss[loss=2.421, over 2590.00 frames. , ppl: 11.25281252878657] tot_loss[loss=2.285, over 5469970.37 frames. , ppl: 9.824680265156363], batch size: 70
+2022-12-12 14:51:39,671 INFO [train.py:421] (2/8) Epoch 7, batch 43400, loss[loss=2.236, over 3780.00 frames. , ppl: 9.354129317208296] tot_loss[loss=2.284, over 5476046.32 frames. , ppl: 9.819249244705539], batch size: 70
+2022-12-12 14:53:19,379 INFO [train.py:421] (2/8) Epoch 7, batch 43600, loss[loss=2.383, over 1680.00 frames. , ppl: 10.83564305557845] tot_loss[loss=2.285, over 5448036.34 frames. , ppl: 9.823321579168217], batch size: 70
+2022-12-12 14:54:58,929 INFO [train.py:421] (2/8) Epoch 7, batch 43800, loss[loss=2.275, over 2520.00 frames. , ppl: 9.72363368288151] tot_loss[loss=2.284, over 5425478.56 frames. , ppl: 9.820765003092768], batch size: 70
+2022-12-12 14:56:40,178 INFO [train.py:421] (2/8) Epoch 7, batch 44000, loss[loss=2.214, over 6440.00 frames. , ppl: 9.155020292763071] tot_loss[loss=2.285, over 5406044.76 frames. , ppl: 9.823897012401728], batch size: 70
+2022-12-12 14:56:40,178 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 14:56:40,926 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.738247314799137
+2022-12-12 14:58:24,214 INFO [train.py:421] (2/8) Epoch 7, batch 44200, loss[loss=2.169, over 5320.00 frames. , ppl: 8.751891551521426] tot_loss[loss=2.283, over 5473088.20 frames. , ppl: 9.810036271963654], batch size: 70
+2022-12-12 15:00:05,404 INFO [train.py:421] (2/8) Epoch 7, batch 44400, loss[loss=2.264, over 2100.00 frames. , ppl: 9.621453100313914] tot_loss[loss=2.283, over 5504969.18 frames. , ppl: 9.802621618073976], batch size: 70
+2022-12-12 15:01:46,541 INFO [train.py:421] (2/8) Epoch 7, batch 44600, loss[loss=2.208, over 4480.00 frames. , ppl: 9.101495714977025] tot_loss[loss=2.284, over 5447150.76 frames. , ppl: 9.819386007193375], batch size: 70
+2022-12-12 15:03:25,945 INFO [train.py:421] (2/8) Epoch 7, batch 44800, loss[loss=2.195, over 3640.00 frames. , ppl: 8.9833970014802] tot_loss[loss=2.283, over 5484669.61 frames. , ppl: 9.806362785740879], batch size: 70
+2022-12-12 15:05:07,366 INFO [train.py:421] (2/8) Epoch 7, batch 45000, loss[loss=2.494, over 1400.00 frames. , ppl: 12.110649266424941] tot_loss[loss=2.283, over 5459774.05 frames. , ppl: 9.807669284817276], batch size: 70
+2022-12-12 15:05:07,367 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:05:08,112 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.739091972737263
+2022-12-12 15:06:49,436 INFO [train.py:421] (2/8) Epoch 7, batch 45200, loss[loss=2.413, over 1330.00 frames. , ppl: 11.172121162149509] tot_loss[loss=2.285, over 5405805.22 frames. , ppl: 9.825096570780737], batch size: 70
+2022-12-12 15:08:26,833 INFO [train.py:421] (2/8) Epoch 7, batch 45400, loss[loss=2.229, over 4690.00 frames. , ppl: 9.28704253262028] tot_loss[loss=2.286, over 5389411.31 frames. , ppl: 9.83180751608815], batch size: 70
+2022-12-12 15:10:04,368 INFO [train.py:421] (2/8) Epoch 7, batch 45600, loss[loss=2.515, over 1050.00 frames. , ppl: 12.369467277471893] tot_loss[loss=2.284, over 5439189.96 frames. , ppl: 9.81952999275768], batch size: 70
+2022-12-12 15:11:43,469 INFO [train.py:421] (2/8) Epoch 7, batch 45800, loss[loss=2.25, over 3290.00 frames. , ppl: 9.4836686771462] tot_loss[loss=2.283, over 5479071.56 frames. , ppl: 9.804600098036033], batch size: 70
+2022-12-12 15:13:23,822 INFO [train.py:421] (2/8) Epoch 7, batch 46000, loss[loss=2.535, over 1120.00 frames. , ppl: 12.617071102384873] tot_loss[loss=2.284, over 5453926.11 frames. , ppl: 9.814171530545611], batch size: 70
+2022-12-12 15:13:23,822 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:13:24,592 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.725171677432357
+2022-12-12 15:15:04,841 INFO [train.py:421] (2/8) Epoch 7, batch 46200, loss[loss=2.442, over 1610.00 frames. , ppl: 11.496245964487741] tot_loss[loss=2.283, over 5468728.45 frames. , ppl: 9.8081284714185], batch size: 70
+2022-12-12 15:16:41,455 INFO [train.py:421] (2/8) Epoch 7, batch 46400, loss[loss=2.45, over 1050.00 frames. , ppl: 11.592208528115004] tot_loss[loss=2.283, over 5463845.45 frames. , ppl: 9.806656565404337], batch size: 70
+2022-12-12 15:18:27,479 INFO [train.py:421] (2/8) Epoch 7, batch 46600, loss[loss=2.235, over 3220.00 frames. , ppl: 9.351054335578226] tot_loss[loss=2.282, over 5498804.75 frames. , ppl: 9.79503725958049], batch size: 70
+2022-12-12 15:20:06,523 INFO [train.py:421] (2/8) Epoch 7, batch 46800, loss[loss=2.777, over 770.00 frames. , ppl: 16.06459318910138] tot_loss[loss=2.282, over 5487724.13 frames. , ppl: 9.794544581854176], batch size: 70
+2022-12-12 15:21:46,494 INFO [train.py:421] (2/8) Epoch 7, batch 47000, loss[loss=2.302, over 1680.00 frames. , ppl: 9.994733837006493] tot_loss[loss=2.281, over 5509862.83 frames. , ppl: 9.786688789732633], batch size: 70
+2022-12-12 15:21:46,494 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:21:47,256 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.746437530099742
+2022-12-12 15:23:25,417 INFO [train.py:421] (2/8) Epoch 7, batch 47200, loss[loss=2.245, over 3570.00 frames. , ppl: 9.440168682574624] tot_loss[loss=2.282, over 5501964.29 frames. , ppl: 9.794713994365638], batch size: 70
+2022-12-12 15:25:07,547 INFO [train.py:421] (2/8) Epoch 7, batch 47400, loss[loss=2.273, over 3500.00 frames. , ppl: 9.7073097605324] tot_loss[loss=2.283, over 5497198.55 frames. , ppl: 9.802077673241017], batch size: 70
+2022-12-12 15:26:45,395 INFO [train.py:421] (2/8) Epoch 7, batch 47600, loss[loss=2.46, over 1960.00 frames. , ppl: 11.703170564675567] tot_loss[loss=2.283, over 5470790.16 frames. , ppl: 9.807401699424902], batch size: 70
+2022-12-12 15:28:25,616 INFO [train.py:421] (2/8) Epoch 7, batch 47800, loss[loss=2.16, over 7700.00 frames. , ppl: 8.668511904935091] tot_loss[loss=2.284, over 5436601.03 frames. , ppl: 9.81421128382787], batch size: 70
+2022-12-12 15:30:08,467 INFO [train.py:421] (2/8) Epoch 7, batch 48000, loss[loss=2.468, over 1050.00 frames. , ppl: 11.799022016337384] tot_loss[loss=2.285, over 5430851.09 frames. , ppl: 9.822525348395704], batch size: 70
+2022-12-12 15:30:08,467 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:30:09,214 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730277122379839
+2022-12-12 15:31:50,593 INFO [train.py:421] (2/8) Epoch 7, batch 48200, loss[loss=2.595, over 700.00 frames. , ppl: 13.390706152913037] tot_loss[loss=2.284, over 5432955.28 frames. , ppl: 9.819489608238813], batch size: 70
+2022-12-12 15:33:31,840 INFO [train.py:421] (2/8) Epoch 7, batch 48400, loss[loss=2.464, over 1330.00 frames. , ppl: 11.750780770227875] tot_loss[loss=2.285, over 5434242.62 frames. , ppl: 9.821337080470288], batch size: 70
+2022-12-12 15:35:15,561 INFO [train.py:421] (2/8) Epoch 7, batch 48600, loss[loss=2.289, over 1750.00 frames. , ppl: 9.862923691422402] tot_loss[loss=2.284, over 5468588.63 frames. , ppl: 9.813786363725798], batch size: 70
+2022-12-12 15:37:00,084 INFO [train.py:421] (2/8) Epoch 7, batch 48800, loss[loss=2.386, over 1190.00 frames. , ppl: 10.872325558348022] tot_loss[loss=2.282, over 5538609.03 frames. , ppl: 9.791953083330366], batch size: 70
+2022-12-12 15:38:39,938 INFO [train.py:421] (2/8) Epoch 7, batch 49000, loss[loss=2.82, over 840.00 frames. , ppl: 16.768491246112927] tot_loss[loss=2.283, over 5502467.34 frames. , ppl: 9.804196167634771], batch size: 70
+2022-12-12 15:38:39,939 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:38:40,684 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71475889319782
+2022-12-12 15:40:19,125 INFO [train.py:421] (2/8) Epoch 7, batch 49200, loss[loss=2.262, over 2800.00 frames. , ppl: 9.601325462217225] tot_loss[loss=2.283, over 5518289.41 frames. , ppl: 9.803840061402699], batch size: 70
+2022-12-12 15:42:00,516 INFO [train.py:421] (2/8) Epoch 7, batch 49400, loss[loss=2.957, over 700.00 frames. , ppl: 19.24344359316758] tot_loss[loss=2.284, over 5486385.88 frames. , ppl: 9.812528728195174], batch size: 70
+2022-12-12 15:43:37,696 INFO [train.py:421] (2/8) Epoch 7, batch 49600, loss[loss=2.244, over 3920.00 frames. , ppl: 9.43189262768206] tot_loss[loss=2.283, over 5477486.59 frames. , ppl: 9.81022729456482], batch size: 70
+2022-12-12 15:45:24,897 INFO [train.py:421] (2/8) Epoch 7, batch 49800, loss[loss=2.051, over 10850.00 frames. , ppl: 7.778936022872502] tot_loss[loss=2.283, over 5525482.55 frames. , ppl: 9.802435872278696], batch size: 70
+2022-12-12 15:47:06,659 INFO [train.py:421] (2/8) Epoch 7, batch 50000, loss[loss=2.17, over 4760.00 frames. , ppl: 8.757975347143214] tot_loss[loss=2.283, over 5532195.30 frames. , ppl: 9.801474650637632], batch size: 70
+2022-12-12 15:47:06,659 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:47:07,406 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714458385768214
+2022-12-12 15:48:44,231 INFO [train.py:421] (2/8) Epoch 7, batch 50200, loss[loss=2.471, over 1890.00 frames. , ppl: 11.832932729629558] tot_loss[loss=2.283, over 5532161.94 frames. , ppl: 9.802383696941249], batch size: 70
+2022-12-12 15:50:22,977 INFO [train.py:421] (2/8) Epoch 7, batch 50400, loss[loss=2.529, over 1050.00 frames. , ppl: 12.543568377259888] tot_loss[loss=2.283, over 5503151.40 frames. , ppl: 9.801656207196102], batch size: 70
+2022-12-12 15:52:02,484 INFO [train.py:421] (2/8) Epoch 7, batch 50600, loss[loss=2.74, over 630.00 frames. , ppl: 15.489582802688226] tot_loss[loss=2.282, over 5522641.42 frames. , ppl: 9.792712206165112], batch size: 70
+2022-12-12 15:53:41,810 INFO [train.py:421] (2/8) Epoch 7, batch 50800, loss[loss=3.066, over 560.00 frames. , ppl: 21.461013583005258] tot_loss[loss=2.283, over 5467891.35 frames. , ppl: 9.808260515013327], batch size: 70
+2022-12-12 15:55:22,386 INFO [train.py:421] (2/8) Epoch 7, batch 51000, loss[loss=2.385, over 1330.00 frames. , ppl: 10.856093397333222] tot_loss[loss=2.284, over 5422228.85 frames. , ppl: 9.817069731215271], batch size: 70
+2022-12-12 15:55:22,387 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 15:55:23,149 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.712544846014454
+2022-12-12 15:57:03,533 INFO [train.py:421] (2/8) Epoch 7, batch 51200, loss[loss=2.24, over 6790.00 frames. , ppl: 9.39460075691492] tot_loss[loss=2.283, over 5453377.61 frames. , ppl: 9.807362909104755], batch size: 70
+2022-12-12 15:58:38,215 INFO [train.py:421] (2/8) Epoch 7, batch 51400, loss[loss=2.176, over 4900.00 frames. , ppl: 8.812590532935545] tot_loss[loss=2.283, over 5445724.78 frames. , ppl: 9.809302056621423], batch size: 70
+2022-12-12 16:00:15,721 INFO [train.py:421] (2/8) Epoch 7, batch 51600, loss[loss=2.236, over 2380.00 frames. , ppl: 9.359222133037932] tot_loss[loss=2.283, over 5443982.92 frames. , ppl: 9.80728422030877], batch size: 70
+2022-12-12 16:01:57,240 INFO [train.py:421] (2/8) Epoch 7, batch 51800, loss[loss=2.469, over 1190.00 frames. , ppl: 11.813890860878889] tot_loss[loss=2.282, over 5473543.06 frames. , ppl: 9.799300570757868], batch size: 70
+2022-12-12 16:03:36,230 INFO [train.py:421] (2/8) Epoch 7, batch 52000, loss[loss=2.169, over 3780.00 frames. , ppl: 8.750197712976071] tot_loss[loss=2.283, over 5465334.09 frames. , ppl: 9.804297908912064], batch size: 70
+2022-12-12 16:03:36,230 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:03:36,974 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714971698339436
+2022-12-12 16:05:14,234 INFO [train.py:421] (2/8) Epoch 7, batch 52200, loss[loss=2.562, over 1260.00 frames. , ppl: 12.956314948287897] tot_loss[loss=2.283, over 5454017.63 frames. , ppl: 9.80776892179932], batch size: 70
+2022-12-12 16:06:54,093 INFO [train.py:421] (2/8) Epoch 7, batch 52400, loss[loss=2.466, over 1260.00 frames. , ppl: 11.7733890573446] tot_loss[loss=2.282, over 5476496.25 frames. , ppl: 9.797238691444868], batch size: 70
+2022-12-12 16:08:37,695 INFO [train.py:421] (2/8) Epoch 7, batch 52600, loss[loss=2.189, over 8260.00 frames. , ppl: 8.929747327113054] tot_loss[loss=2.281, over 5529430.56 frames. , ppl: 9.785182958086889], batch size: 70
+2022-12-12 16:10:20,600 INFO [train.py:421] (2/8) Epoch 7, batch 52800, loss[loss=2.706, over 910.00 frames. , ppl: 14.972065245039257] tot_loss[loss=2.281, over 5535826.56 frames. , ppl: 9.782228563942656], batch size: 70
+2022-12-12 16:12:02,794 INFO [train.py:421] (2/8) Epoch 7, batch 53000, loss[loss=2.285, over 2660.00 frames. , ppl: 9.82296165346412] tot_loss[loss=2.28, over 5555990.94 frames. , ppl: 9.777726442244981], batch size: 70
+2022-12-12 16:12:02,795 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:12:03,557 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.710289624670212
+2022-12-12 16:13:44,881 INFO [train.py:421] (2/8) Epoch 7, batch 53200, loss[loss=2.352, over 1960.00 frames. , ppl: 10.505315609124803] tot_loss[loss=2.28, over 5571941.09 frames. , ppl: 9.777557248059923], batch size: 70
+2022-12-12 16:15:28,421 INFO [train.py:421] (2/8) Epoch 7, batch 53400, loss[loss=2.402, over 1120.00 frames. , ppl: 11.04645582370205] tot_loss[loss=2.281, over 5567558.73 frames. , ppl: 9.786224009656653], batch size: 70
+2022-12-12 16:17:11,372 INFO [train.py:421] (2/8) Epoch 7, batch 53600, loss[loss=2.517, over 1330.00 frames. , ppl: 12.395801762507707] tot_loss[loss=2.281, over 5597189.49 frames. , ppl: 9.7823383784102], batch size: 70
+2022-12-12 16:18:49,868 INFO [train.py:421] (2/8) Epoch 7, batch 53800, loss[loss=2.399, over 1190.00 frames. , ppl: 11.007065810728612] tot_loss[loss=2.282, over 5566922.75 frames. , ppl: 9.791746943531615], batch size: 70
+2022-12-12 16:20:30,712 INFO [train.py:421] (2/8) Epoch 7, batch 54000, loss[loss=2.355, over 1680.00 frames. , ppl: 10.536893997076259] tot_loss[loss=2.283, over 5512515.15 frames. , ppl: 9.802712527277029], batch size: 70
+2022-12-12 16:20:30,713 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:20:31,473 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730233917866439
+2022-12-12 16:22:14,289 INFO [train.py:421] (2/8) Epoch 7, batch 54200, loss[loss=2.153, over 7420.00 frames. , ppl: 8.610395256434238] tot_loss[loss=2.284, over 5472151.95 frames. , ppl: 9.81754049436142], batch size: 70
+2022-12-12 16:23:57,483 INFO [train.py:421] (2/8) Epoch 7, batch 54400, loss[loss=2.295, over 1610.00 frames. , ppl: 9.926399856010816] tot_loss[loss=2.283, over 5503271.42 frames. , ppl: 9.81053015867425], batch size: 70
+2022-12-12 16:25:40,588 INFO [train.py:421] (2/8) Epoch 7, batch 54600, loss[loss=2.463, over 910.00 frames. , ppl: 11.743381850777991] tot_loss[loss=2.284, over 5485243.54 frames. , ppl: 9.811427195668857], batch size: 70
+2022-12-12 16:27:23,732 INFO [train.py:421] (2/8) Epoch 7, batch 54800, loss[loss=2.274, over 3290.00 frames. , ppl: 9.715043453657058] tot_loss[loss=2.282, over 5511086.37 frames. , ppl: 9.800909250009981], batch size: 70
+2022-12-12 16:29:03,501 INFO [train.py:421] (2/8) Epoch 7, batch 55000, loss[loss=2.637, over 1400.00 frames. , ppl: 13.974441446931595] tot_loss[loss=2.282, over 5502804.70 frames. , ppl: 9.799872673766913], batch size: 70
+2022-12-12 16:29:03,501 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:29:04,267 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.738142098037239
+2022-12-12 16:30:45,241 INFO [train.py:421] (2/8) Epoch 7, batch 55200, loss[loss=2.08, over 5250.00 frames. , ppl: 8.000732581053834] tot_loss[loss=2.282, over 5530768.04 frames. , ppl: 9.791675573385701], batch size: 70
+2022-12-12 16:32:27,237 INFO [train.py:421] (2/8) Epoch 7, batch 55400, loss[loss=2.289, over 2310.00 frames. , ppl: 9.861326782940221] tot_loss[loss=2.281, over 5551401.52 frames. , ppl: 9.785873056301181], batch size: 70
+2022-12-12 16:34:09,959 INFO [train.py:421] (2/8) Epoch 7, batch 55600, loss[loss=2.285, over 2870.00 frames. , ppl: 9.82745066494528] tot_loss[loss=2.28, over 5568879.96 frames. , ppl: 9.771976301168493], batch size: 70
+2022-12-12 16:35:52,341 INFO [train.py:421] (2/8) Epoch 7, batch 55800, loss[loss=2.236, over 2590.00 frames. , ppl: 9.360346779768312] tot_loss[loss=2.278, over 5566948.72 frames. , ppl: 9.761991638042131], batch size: 70
+2022-12-12 16:37:34,622 INFO [train.py:421] (2/8) Epoch 7, batch 56000, loss[loss=2.148, over 2030.00 frames. , ppl: 8.569570829958389] tot_loss[loss=2.28, over 5545191.27 frames. , ppl: 9.774206457547129], batch size: 70
+2022-12-12 16:37:34,622 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:37:35,393 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.72342008782373
+2022-12-12 16:39:17,250 INFO [train.py:421] (2/8) Epoch 7, batch 56200, loss[loss=2.182, over 2450.00 frames. , ppl: 8.859590266007457] tot_loss[loss=2.281, over 5508328.18 frames. , ppl: 9.784800036505338], batch size: 70
+2022-12-12 16:40:58,456 INFO [train.py:421] (2/8) Epoch 7, batch 56400, loss[loss=2.38, over 1750.00 frames. , ppl: 10.804088909163402] tot_loss[loss=2.28, over 5513231.78 frames. , ppl: 9.780607091511808], batch size: 70
+2022-12-12 16:42:37,535 INFO [train.py:421] (2/8) Epoch 7, batch 56600, loss[loss=2.334, over 1750.00 frames. , ppl: 10.323687248106623] tot_loss[loss=2.281, over 5489084.48 frames. , ppl: 9.789760146338354], batch size: 70
+2022-12-12 16:44:14,636 INFO [train.py:421] (2/8) Epoch 7, batch 56800, loss[loss=2.54, over 910.00 frames. , ppl: 12.680652763500275] tot_loss[loss=2.281, over 5503137.73 frames. , ppl: 9.784426323625294], batch size: 70
+2022-12-12 16:45:52,794 INFO [train.py:421] (2/8) Epoch 7, batch 57000, loss[loss=2.5, over 980.00 frames. , ppl: 12.183999384198234] tot_loss[loss=2.281, over 5499977.86 frames. , ppl: 9.783537267113248], batch size: 70
+2022-12-12 16:45:52,795 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:45:53,554 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.720317799677936
+2022-12-12 16:47:33,744 INFO [train.py:421] (2/8) Epoch 7, batch 57200, loss[loss=2.321, over 2800.00 frames. , ppl: 10.184559539395702] tot_loss[loss=2.281, over 5484584.33 frames. , ppl: 9.786847580511532], batch size: 70
+2022-12-12 16:49:17,327 INFO [train.py:421] (2/8) Epoch 7, batch 57400, loss[loss=2.349, over 910.00 frames. , ppl: 10.476234272199685] tot_loss[loss=2.282, over 5465033.40 frames. , ppl: 9.795040628007177], batch size: 70
+2022-12-12 16:50:58,688 INFO [train.py:421] (2/8) Epoch 7, batch 57600, loss[loss=2.456, over 1750.00 frames. , ppl: 11.652651622104065] tot_loss[loss=2.282, over 5453312.43 frames. , ppl: 9.79463922454444], batch size: 70
+2022-12-12 16:52:38,236 INFO [train.py:421] (2/8) Epoch 7, batch 57800, loss[loss=2.181, over 5180.00 frames. , ppl: 8.85760483642787] tot_loss[loss=2.283, over 5427729.81 frames. , ppl: 9.80136433633516], batch size: 70
+2022-12-12 16:54:19,003 INFO [train.py:421] (2/8) Epoch 7, batch 58000, loss[loss=2.357, over 1750.00 frames. , ppl: 10.555570592746447] tot_loss[loss=2.282, over 5429428.80 frames. , ppl: 9.800561785035487], batch size: 70
+2022-12-12 16:54:19,003 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 16:54:19,765 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714061557664111
+2022-12-12 16:56:00,041 INFO [train.py:421] (2/8) Epoch 7, batch 58200, loss[loss=2.21, over 4690.00 frames. , ppl: 9.117915411790756] tot_loss[loss=2.284, over 5392206.71 frames. , ppl: 9.815268628310058], batch size: 70
+2022-12-12 16:57:41,618 INFO [train.py:421] (2/8) Epoch 7, batch 58400, loss[loss=2.554, over 910.00 frames. , ppl: 12.855711714481346] tot_loss[loss=2.284, over 5406448.64 frames. , ppl: 9.815304025676062], batch size: 70
+2022-12-12 16:59:24,618 INFO [train.py:421] (2/8) Epoch 7, batch 58600, loss[loss=2.416, over 1750.00 frames. , ppl: 11.201615799570984] tot_loss[loss=2.284, over 5428513.69 frames. , ppl: 9.814954257246162], batch size: 70
+2022-12-12 17:01:06,528 INFO [train.py:421] (2/8) Epoch 7, batch 58800, loss[loss=2.209, over 3220.00 frames. , ppl: 9.10862293542888] tot_loss[loss=2.282, over 5454050.07 frames. , ppl: 9.799603512296352], batch size: 70
+2022-12-12 17:02:51,781 INFO [train.py:421] (2/8) Epoch 7, batch 59000, loss[loss=2.156, over 6370.00 frames. , ppl: 8.638340105319326] tot_loss[loss=2.282, over 5479060.62 frames. , ppl: 9.79566656070567], batch size: 70
+2022-12-12 17:02:51,782 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:02:52,548 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.716621096251133
+2022-12-12 17:04:37,257 INFO [train.py:421] (2/8) Epoch 7, batch 59200, loss[loss=2.528, over 700.00 frames. , ppl: 12.532297998158368] tot_loss[loss=2.281, over 5509134.80 frames. , ppl: 9.787624108745486], batch size: 70
+2022-12-12 17:06:15,593 INFO [train.py:421] (2/8) Epoch 7, batch 59400, loss[loss=2.506, over 1400.00 frames. , ppl: 12.256346390827837] tot_loss[loss=2.282, over 5493241.60 frames. , ppl: 9.795364236163344], batch size: 70
+2022-12-12 17:07:59,263 INFO [train.py:421] (2/8) Epoch 7, batch 59600, loss[loss=2.29, over 2310.00 frames. , ppl: 9.875177310290907] tot_loss[loss=2.281, over 5531899.80 frames. , ppl: 9.78640871902835], batch size: 70
+2022-12-12 17:09:44,695 INFO [train.py:421] (2/8) Epoch 7, batch 59800, loss[loss=2.216, over 3430.00 frames. , ppl: 9.169389484229196] tot_loss[loss=2.28, over 5568447.39 frames. , ppl: 9.772362212406579], batch size: 70
+2022-12-12 17:11:28,832 INFO [train.py:421] (2/8) Epoch 7, batch 60000, loss[loss=2.454, over 1470.00 frames. , ppl: 11.639666366613534] tot_loss[loss=2.279, over 5600596.95 frames. , ppl: 9.770582965316084], batch size: 70
+2022-12-12 17:11:28,833 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:11:29,585 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71337577339601
+2022-12-12 17:13:08,942 INFO [train.py:421] (2/8) Epoch 7, batch 60200, loss[loss=2.327, over 2870.00 frames. , ppl: 10.245706105460739] tot_loss[loss=2.278, over 5625690.85 frames. , ppl: 9.760665408882375], batch size: 70
+2022-12-12 17:14:49,931 INFO [train.py:421] (2/8) Epoch 7, batch 60400, loss[loss=2.264, over 2730.00 frames. , ppl: 9.61687959888566] tot_loss[loss=2.278, over 5636070.37 frames. , ppl: 9.758579252703402], batch size: 70
+2022-12-12 17:16:33,754 INFO [train.py:421] (2/8) Epoch 7, batch 60600, loss[loss=2.339, over 2240.00 frames. , ppl: 10.37043262461083] tot_loss[loss=2.278, over 5641333.18 frames. , ppl: 9.75722167527448], batch size: 70
+2022-12-12 17:18:13,221 INFO [train.py:421] (2/8) Epoch 7, batch 60800, loss[loss=2.346, over 1400.00 frames. , ppl: 10.441958605598318] tot_loss[loss=2.278, over 5616582.05 frames. , ppl: 9.758162663028221], batch size: 70
+2022-12-12 17:19:56,253 INFO [train.py:421] (2/8) Epoch 7, batch 61000, loss[loss=2.3, over 1680.00 frames. , ppl: 9.976079986484475] tot_loss[loss=2.278, over 5606137.25 frames. , ppl: 9.758572164687061], batch size: 70
+2022-12-12 17:19:56,253 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:19:57,004 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.70608961125915
+2022-12-12 17:21:39,568 INFO [train.py:421] (2/8) Epoch 7, batch 61200, loss[loss=2.135, over 6860.00 frames. , ppl: 8.456535582156773] tot_loss[loss=2.277, over 5678634.50 frames. , ppl: 9.744760565068795], batch size: 70
+2022-12-12 17:23:24,485 INFO [train.py:421] (2/8) Epoch 7, batch 61400, loss[loss=2.487, over 840.00 frames. , ppl: 12.028964812899053] tot_loss[loss=2.278, over 5639111.77 frames. , ppl: 9.759281766589577], batch size: 70
+2022-12-12 17:25:04,986 INFO [train.py:421] (2/8) Epoch 7, batch 61600, loss[loss=2.808, over 840.00 frames. , ppl: 16.57776688723542] tot_loss[loss=2.278, over 5625303.79 frames. , ppl: 9.761423985002], batch size: 70
+2022-12-12 17:26:46,419 INFO [train.py:421] (2/8) Epoch 7, batch 61800, loss[loss=2.916, over 630.00 frames. , ppl: 18.462804022863132] tot_loss[loss=2.279, over 5619663.46 frames. , ppl: 9.76495165013156], batch size: 70
+2022-12-12 17:28:30,621 INFO [train.py:421] (2/8) Epoch 7, batch 62000, loss[loss=2.203, over 2870.00 frames. , ppl: 9.054129526616041] tot_loss[loss=2.28, over 5561548.10 frames. , ppl: 9.77895766535307], batch size: 70
+2022-12-12 17:28:30,621 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:28:31,386 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71544765103413
+2022-12-12 17:30:11,235 INFO [train.py:421] (2/8) Epoch 7, batch 62200, loss[loss=2.14, over 4270.00 frames. , ppl: 8.500850535224087] tot_loss[loss=2.28, over 5522877.36 frames. , ppl: 9.781459063509665], batch size: 70
+2022-12-12 17:31:56,818 INFO [train.py:421] (2/8) Epoch 7, batch 62400, loss[loss=2.16, over 4620.00 frames. , ppl: 8.67128832254874] tot_loss[loss=2.279, over 5569570.66 frames. , ppl: 9.76933164608777], batch size: 70
+2022-12-12 17:33:40,864 INFO [train.py:421] (2/8) Epoch 7, batch 62600, loss[loss=2.41, over 1890.00 frames. , ppl: 11.130241931695306] tot_loss[loss=2.279, over 5582018.55 frames. , ppl: 9.77023701222694], batch size: 70
+2022-12-12 17:35:24,955 INFO [train.py:421] (2/8) Epoch 7, batch 62800, loss[loss=2.491, over 1050.00 frames. , ppl: 12.072594482944304] tot_loss[loss=2.281, over 5527824.05 frames. , ppl: 9.790410640523705], batch size: 70
+2022-12-12 17:37:08,813 INFO [train.py:421] (2/8) Epoch 7, batch 63000, loss[loss=2.441, over 1540.00 frames. , ppl: 11.479441984197035] tot_loss[loss=2.28, over 5562862.16 frames. , ppl: 9.77478174610669], batch size: 70
+2022-12-12 17:37:08,813 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:37:09,562 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.722217042257858
+2022-12-12 17:38:51,316 INFO [train.py:421] (2/8) Epoch 7, batch 63200, loss[loss=2.207, over 7630.00 frames. , ppl: 9.09224986040401] tot_loss[loss=2.28, over 5552930.41 frames. , ppl: 9.775785697806667], batch size: 70
+2022-12-12 17:40:30,752 INFO [train.py:421] (2/8) Epoch 7, batch 63400, loss[loss=2.243, over 3430.00 frames. , ppl: 9.422758725654884] tot_loss[loss=2.28, over 5543469.62 frames. , ppl: 9.777354348845885], batch size: 70
+2022-12-12 17:42:12,638 INFO [train.py:421] (2/8) Epoch 7, batch 63600, loss[loss=2.493, over 1330.00 frames. , ppl: 12.093592673670786] tot_loss[loss=2.28, over 5556346.28 frames. , ppl: 9.772314400409378], batch size: 70
+2022-12-12 17:43:53,023 INFO [train.py:421] (2/8) Epoch 7, batch 63800, loss[loss=2.304, over 1330.00 frames. , ppl: 10.01447762038716] tot_loss[loss=2.279, over 5554460.72 frames. , ppl: 9.771409631644426], batch size: 70
+2022-12-12 17:45:29,461 INFO [train.py:421] (2/8) Epoch 7, batch 64000, loss[loss=2.24, over 6020.00 frames. , ppl: 9.397529943587157] tot_loss[loss=2.282, over 5479154.51 frames. , ppl: 9.796156778199036], batch size: 70
+2022-12-12 17:45:29,462 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:45:30,209 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.717427922051451
+2022-12-12 17:47:12,552 INFO [train.py:421] (2/8) Epoch 7, batch 64200, loss[loss=2.198, over 5390.00 frames. , ppl: 9.005535316186053] tot_loss[loss=2.284, over 5431164.79 frames. , ppl: 9.81158600963025], batch size: 70
+2022-12-12 17:48:56,533 INFO [train.py:421] (2/8) Epoch 7, batch 64400, loss[loss=2.263, over 3500.00 frames. , ppl: 9.612800191376452] tot_loss[loss=2.283, over 5444193.22 frames. , ppl: 9.80747751762533], batch size: 70
+2022-12-12 17:50:37,244 INFO [train.py:421] (2/8) Epoch 7, batch 64600, loss[loss=2.25, over 3010.00 frames. , ppl: 9.48300583004217] tot_loss[loss=2.283, over 5460062.75 frames. , ppl: 9.805500442004202], batch size: 70
+2022-12-12 17:52:15,806 INFO [train.py:421] (2/8) Epoch 7, batch 64800, loss[loss=2.17, over 3920.00 frames. , ppl: 8.76143176547564] tot_loss[loss=2.283, over 5455835.52 frames. , ppl: 9.805861917497792], batch size: 70
+2022-12-12 17:53:56,980 INFO [train.py:421] (2/8) Epoch 7, batch 65000, loss[loss=2.338, over 2590.00 frames. , ppl: 10.361060974501191] tot_loss[loss=2.283, over 5458535.67 frames. , ppl: 9.802840903530159], batch size: 70
+2022-12-12 17:53:56,980 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 17:53:57,744 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.728521733447238
+2022-12-12 17:55:38,071 INFO [train.py:421] (2/8) Epoch 7, batch 65200, loss[loss=2.353, over 3920.00 frames. , ppl: 10.518330734505914] tot_loss[loss=2.283, over 5467602.44 frames. , ppl: 9.805803274986966], batch size: 70
+2022-12-12 17:57:16,314 INFO [train.py:421] (2/8) Epoch 7, batch 65400, loss[loss=2.644, over 770.00 frames. , ppl: 14.071048401352137] tot_loss[loss=2.283, over 5467951.31 frames. , ppl: 9.807069040176875], batch size: 70
+2022-12-12 17:58:55,635 INFO [train.py:421] (2/8) Epoch 7, batch 65600, loss[loss=2.445, over 980.00 frames. , ppl: 11.536142039413651] tot_loss[loss=2.282, over 5496773.78 frames. , ppl: 9.797203407609999], batch size: 70
+2022-12-12 18:00:40,204 INFO [train.py:421] (2/8) Epoch 7, batch 65800, loss[loss=2.579, over 910.00 frames. , ppl: 13.178330385264825] tot_loss[loss=2.283, over 5463991.70 frames. , ppl: 9.803280885587265], batch size: 70
+2022-12-12 18:02:18,980 INFO [train.py:421] (2/8) Epoch 7, batch 66000, loss[loss=2.823, over 630.00 frames. , ppl: 16.820345716190143] tot_loss[loss=2.283, over 5482699.68 frames. , ppl: 9.802404855730204], batch size: 70
+2022-12-12 18:02:18,981 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 18:02:19,737 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.725315618297568
+2022-12-12 18:04:01,116 INFO [train.py:421] (2/8) Epoch 7, batch 66200, loss[loss=2.49, over 1820.00 frames. , ppl: 12.063407456908422] tot_loss[loss=2.282, over 5518857.51 frames. , ppl: 9.792723532071422], batch size: 70
+2022-12-12 18:05:46,685 INFO [train.py:421] (2/8) Epoch 7, batch 66400, loss[loss=2.115, over 6160.00 frames. , ppl: 8.292023124652857] tot_loss[loss=2.28, over 5538179.50 frames. , ppl: 9.779023797090314], batch size: 70
+2022-12-12 18:07:29,647 INFO [train.py:421] (2/8) Epoch 7, batch 66600, loss[loss=2.517, over 1120.00 frames. , ppl: 12.389395529747125] tot_loss[loss=2.28, over 5559116.07 frames. , ppl: 9.774533538585969], batch size: 70
+2022-12-12 18:09:07,070 INFO [train.py:421] (2/8) Epoch 7, batch 66800, loss[loss=2.381, over 910.00 frames. , ppl: 10.812599621512007] tot_loss[loss=2.28, over 5544129.55 frames. , ppl: 9.772037721827541], batch size: 70
+2022-12-12 18:10:47,177 INFO [train.py:421] (2/8) Epoch 7, batch 67000, loss[loss=2.256, over 2870.00 frames. , ppl: 9.548891425851812] tot_loss[loss=2.28, over 5540678.56 frames. , ppl: 9.775394458135533], batch size: 70
+2022-12-12 18:10:47,177 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 18:10:47,941 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71459066540528
+2022-12-12 18:12:28,622 INFO [train.py:421] (2/8) Epoch 7, batch 67200, loss[loss=2.609, over 910.00 frames. , ppl: 13.584105572371158] tot_loss[loss=2.28, over 5536533.93 frames. , ppl: 9.776959820000446], batch size: 70
+2022-12-12 18:14:14,269 INFO [train.py:421] (2/8) Epoch 7, batch 67400, loss[loss=2.344, over 2660.00 frames. , ppl: 10.421941801743792] tot_loss[loss=2.28, over 5547314.01 frames. , ppl: 9.776780549293198], batch size: 70
+2022-12-12 18:15:54,839 INFO [train.py:421] (2/8) Epoch 7, batch 67600, loss[loss=2.319, over 1680.00 frames. , ppl: 10.160918835518485] tot_loss[loss=2.279, over 5587829.08 frames. , ppl: 9.765065073470632], batch size: 70
+2022-12-12 18:17:35,752 INFO [train.py:421] (2/8) Epoch 7, batch 67800, loss[loss=2.385, over 2030.00 frames. , ppl: 10.861940380150859] tot_loss[loss=2.28, over 5559629.75 frames. , ppl: 9.777550624523187], batch size: 70
+2022-12-12 18:19:18,257 INFO [train.py:421] (2/8) Epoch 7, batch 68000, loss[loss=2.224, over 3430.00 frames. , ppl: 9.245654832175788] tot_loss[loss=2.281, over 5541929.88 frames. , ppl: 9.788251592758547], batch size: 70
+2022-12-12 18:19:18,258 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 18:19:19,022 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.710605812793908
+2022-12-12 18:20:59,031 INFO [train.py:421] (2/8) Epoch 7, batch 68200, loss[loss=2.303, over 2870.00 frames. , ppl: 10.00200341923634] tot_loss[loss=2.279, over 5583183.35 frames. , ppl: 9.771549639436373], batch size: 70
+2022-12-12 18:22:38,124 INFO [train.py:421] (2/8) Epoch 7, batch 68400, loss[loss=2.479, over 1050.00 frames. , ppl: 11.934710849188644] tot_loss[loss=2.28, over 5560331.11 frames. , ppl: 9.772850745722767], batch size: 70
+2022-12-12 18:24:19,221 INFO [train.py:421] (2/8) Epoch 7, batch 68600, loss[loss=2.149, over 4620.00 frames. , ppl: 8.572563929397697] tot_loss[loss=2.28, over 5525292.61 frames. , ppl: 9.780544507849122], batch size: 70
+2022-12-12 18:25:59,214 INFO [train.py:421] (2/8) Epoch 7, batch 68800, loss[loss=2.175, over 4970.00 frames. , ppl: 8.799809035607] tot_loss[loss=2.281, over 5544464.84 frames. , ppl: 9.785832323263815], batch size: 70
+2022-12-12 18:27:41,660 INFO [train.py:421] (2/8) Epoch 7, batch 69000, loss[loss=2.213, over 4340.00 frames. , ppl: 9.141641794662748] tot_loss[loss=2.28, over 5570153.19 frames. , ppl: 9.778262809494707], batch size: 70
+2022-12-12 18:27:41,660 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 18:27:42,411 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.71098525213279
+2022-12-12 18:29:24,643 INFO [train.py:421] (2/8) Epoch 7, batch 69200, loss[loss=2.398, over 1680.00 frames. , ppl: 11.003668008938616] tot_loss[loss=2.279, over 5574833.42 frames. , ppl: 9.770795562675628], batch size: 70
+2022-12-12 18:31:04,117 INFO [train.py:421] (2/8) Epoch 7, batch 69400, loss[loss=2.184, over 7350.00 frames. , ppl: 8.880838865209459] tot_loss[loss=2.28, over 5555450.62 frames. , ppl: 9.775095197615881], batch size: 70
+2022-12-12 18:32:46,705 INFO [train.py:421] (2/8) Epoch 7, batch 69600, loss[loss=2.446, over 1190.00 frames. , ppl: 11.544792735436594] tot_loss[loss=2.28, over 5545375.38 frames. , ppl: 9.774695314688701], batch size: 70
+2022-12-12 18:34:31,065 INFO [train.py:421] (2/8) Epoch 7, batch 69800, loss[loss=2.269, over 3780.00 frames. , ppl: 9.668647900374799] tot_loss[loss=2.279, over 5556732.25 frames. , ppl: 9.768303897241799], batch size: 70
+2022-12-12 18:36:12,732 INFO [train.py:421] (2/8) Epoch 7, batch 70000, loss[loss=2.29, over 9030.00 frames. , ppl: 9.871682283886141] tot_loss[loss=2.28, over 5526592.01 frames. , ppl: 9.777781563645856], batch size: 70
+2022-12-12 18:36:12,733 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 18:36:13,480 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.71248878254001
+2022-12-12 18:37:52,814 INFO [train.py:421] (2/8) Epoch 7, batch 70200, loss[loss=2.384, over 1330.00 frames. , ppl: 10.848127146189984] tot_loss[loss=2.281, over 5494728.53 frames. , ppl: 9.790779098995895], batch size: 70
+2022-12-12 18:39:34,533 INFO [train.py:421] (2/8) Epoch 7, batch 70400, loss[loss=2.427, over 1330.00 frames. , ppl: 11.321856380236445] tot_loss[loss=2.28, over 5538579.64 frames. , ppl: 9.780042560629477], batch size: 70
+2022-12-12 18:41:16,484 INFO [train.py:421] (2/8) Epoch 7, batch 70600, loss[loss=2.975, over 560.00 frames. , ppl: 19.593244713037866] tot_loss[loss=2.281, over 5560209.25 frames. , ppl: 9.781874944219147], batch size: 70
+2022-12-12 18:43:01,174 INFO [train.py:421] (2/8) Epoch 7, batch 70800, loss[loss=2.495, over 1190.00 frames. , ppl: 12.119150405106662] tot_loss[loss=2.279, over 5628354.20 frames. , ppl: 9.770077179503147], batch size: 70
+2022-12-12 18:44:42,633 INFO [train.py:421] (2/8) Epoch 7, batch 71000, loss[loss=2.625, over 770.00 frames. , ppl: 13.808999999540898] tot_loss[loss=2.279, over 5599419.49 frames. , ppl: 9.771781730189016], batch size: 70
+2022-12-12 18:44:42,634 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 18:44:43,417 INFO [train.py:452] (2/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.707774858509012
+2022-12-12 18:46:25,686 INFO [train.py:421] (2/8) Epoch 7, batch 71200, loss[loss=2.179, over 6510.00 frames. , ppl: 8.836116288522106] tot_loss[loss=2.28, over 5586763.64 frames. , ppl: 9.780976743319117], batch size: 70
+2022-12-12 18:48:08,695 INFO [train.py:421] (2/8) Epoch 7, batch 71400, loss[loss=2.38, over 1400.00 frames. , ppl: 10.8059203936249] tot_loss[loss=2.28, over 5580265.21 frames. , ppl: 9.781383775150974], batch size: 70
+2022-12-12 18:49:50,948 INFO [train.py:421] (2/8) Epoch 7, batch 71600, loss[loss=2.13, over 5600.00 frames. , ppl: 8.41160973332709] tot_loss[loss=2.279, over 5600877.38 frames. , ppl: 9.769702204864497], batch size: 70
+2022-12-12 18:51:32,658 INFO [train.py:421] (2/8) Epoch 7, batch 71800, loss[loss=2.315, over 2030.00 frames. , ppl: 10.12786868897663] tot_loss[loss=2.278, over 5610431.64 frames. , ppl: 9.761797413976833], batch size: 70
+2022-12-12 18:52:48,019 INFO [train.py:421] (2/8) Epoch 8, batch 0, loss[loss=2.162, over 4270.00 frames. , ppl: 8.691856590464074] tot_loss[loss=2.162, over 4270.00 frames. , ppl: 8.691856590464074], batch size: 70
+2022-12-12 18:54:30,706 INFO [train.py:421] (2/8) Epoch 8, batch 200, loss[loss=2.218, over 3990.00 frames. , ppl: 9.19260645176792] tot_loss[loss=2.266, over 531454.28 frames. , ppl: 9.638884360184], batch size: 70
+2022-12-12 18:56:11,879 INFO [train.py:421] (2/8) Epoch 8, batch 400, loss[loss=2.412, over 1190.00 frames. , ppl: 11.152253775019659] tot_loss[loss=2.265, over 1035165.38 frames. , ppl: 9.63043979072652], batch size: 70
+2022-12-12 18:57:54,120 INFO [train.py:421] (2/8) Epoch 8, batch 600, loss[loss=2.296, over 5110.00 frames. , ppl: 9.939123451879222] tot_loss[loss=2.264, over 1510226.53 frames. , ppl: 9.61907478769246], batch size: 70
+2022-12-12 18:59:34,600 INFO [train.py:421] (2/8) Epoch 8, batch 800, loss[loss=2.459, over 1050.00 frames. , ppl: 11.694670024238508] tot_loss[loss=2.267, over 1884035.33 frames. , ppl: 9.649942991562398], batch size: 70
+2022-12-12 19:01:16,046 INFO [train.py:421] (2/8) Epoch 8, batch 1000, loss[loss=2.215, over 10290.00 frames. , ppl: 9.15857304175091] tot_loss[loss=2.269, over 2199970.66 frames. , ppl: 9.670158071843542], batch size: 70
+2022-12-12 19:01:16,046 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:01:16,794 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714002609944576
+2022-12-12 19:02:59,995 INFO [train.py:421] (2/8) Epoch 8, batch 1200, loss[loss=2.495, over 1610.00 frames. , ppl: 12.12411745470048] tot_loss[loss=2.27, over 2515584.21 frames. , ppl: 9.67716764921034], batch size: 70
+2022-12-12 19:04:36,353 INFO [train.py:421] (2/8) Epoch 8, batch 1400, loss[loss=2.265, over 2870.00 frames. , ppl: 9.634437705113706] tot_loss[loss=2.27, over 2803845.65 frames. , ppl: 9.679199900959986], batch size: 70
+2022-12-12 19:06:17,635 INFO [train.py:421] (2/8) Epoch 8, batch 1600, loss[loss=2.337, over 2100.00 frames. , ppl: 10.347368496758838] tot_loss[loss=2.271, over 3053699.74 frames. , ppl: 9.690745597897461], batch size: 70
+2022-12-12 19:08:00,658 INFO [train.py:421] (2/8) Epoch 8, batch 1800, loss[loss=2.202, over 4200.00 frames. , ppl: 9.041127175320252] tot_loss[loss=2.27, over 3313944.81 frames. , ppl: 9.683902255806183], batch size: 70
+2022-12-12 19:09:38,929 INFO [train.py:421] (2/8) Epoch 8, batch 2000, loss[loss=2.354, over 1960.00 frames. , ppl: 10.522400939370034] tot_loss[loss=2.272, over 3500412.62 frames. , ppl: 9.700834488548686], batch size: 70
+2022-12-12 19:09:38,930 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:09:39,713 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.70567014079976
+2022-12-12 19:11:23,733 INFO [train.py:421] (2/8) Epoch 8, batch 2200, loss[loss=2.192, over 5180.00 frames. , ppl: 8.948826279725136] tot_loss[loss=2.273, over 3686804.36 frames. , ppl: 9.710671252788822], batch size: 70
+2022-12-12 19:13:02,998 INFO [train.py:421] (2/8) Epoch 8, batch 2400, loss[loss=2.399, over 1540.00 frames. , ppl: 11.015105721053036] tot_loss[loss=2.274, over 3830062.78 frames. , ppl: 9.71843140294795], batch size: 70
+2022-12-12 19:14:41,152 INFO [train.py:421] (2/8) Epoch 8, batch 2600, loss[loss=2.302, over 3850.00 frames. , ppl: 9.99348966422935] tot_loss[loss=2.275, over 3964249.55 frames. , ppl: 9.724900120511611], batch size: 70
+2022-12-12 19:16:22,764 INFO [train.py:421] (2/8) Epoch 8, batch 2800, loss[loss=2.25, over 3150.00 frames. , ppl: 9.488963944640318] tot_loss[loss=2.276, over 4087574.33 frames. , ppl: 9.735539956561489], batch size: 70
+2022-12-12 19:18:01,847 INFO [train.py:421] (2/8) Epoch 8, batch 3000, loss[loss=2.273, over 2730.00 frames. , ppl: 9.711678324238116] tot_loss[loss=2.276, over 4203714.69 frames. , ppl: 9.740889092481897], batch size: 70
+2022-12-12 19:18:01,848 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:18:02,632 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.276, over 211138.00 frames. , ppl: 9.739168370327212
+2022-12-12 19:19:44,665 INFO [train.py:421] (2/8) Epoch 8, batch 3200, loss[loss=2.704, over 910.00 frames. , ppl: 14.936815188455379] tot_loss[loss=2.278, over 4306941.05 frames. , ppl: 9.75446801898727], batch size: 70
+2022-12-12 19:21:28,589 INFO [train.py:421] (2/8) Epoch 8, batch 3400, loss[loss=2.38, over 3290.00 frames. , ppl: 10.8080069097858] tot_loss[loss=2.276, over 4437677.59 frames. , ppl: 9.736250614548158], batch size: 70
+2022-12-12 19:23:11,769 INFO [train.py:421] (2/8) Epoch 8, batch 3600, loss[loss=2.357, over 1960.00 frames. , ppl: 10.557167849502141] tot_loss[loss=2.278, over 4476906.43 frames. , ppl: 9.756427550171411], batch size: 70
+2022-12-12 19:24:50,895 INFO [train.py:421] (2/8) Epoch 8, batch 3800, loss[loss=2.402, over 1750.00 frames. , ppl: 11.047642711091681] tot_loss[loss=2.277, over 4598694.08 frames. , ppl: 9.752153914790474], batch size: 70
+2022-12-12 19:26:35,932 INFO [train.py:421] (2/8) Epoch 8, batch 4000, loss[loss=2.203, over 4900.00 frames. , ppl: 9.055265580381162] tot_loss[loss=2.277, over 4703030.23 frames. , ppl: 9.748039714077299], batch size: 70
+2022-12-12 19:26:35,933 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:26:36,699 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.709655842207155
+2022-12-12 19:28:16,986 INFO [train.py:421] (2/8) Epoch 8, batch 4200, loss[loss=2.387, over 1330.00 frames. , ppl: 10.87747915752454] tot_loss[loss=2.277, over 4767029.58 frames. , ppl: 9.748652475950799], batch size: 70
+2022-12-12 19:29:59,663 INFO [train.py:421] (2/8) Epoch 8, batch 4400, loss[loss=2.415, over 1610.00 frames. , ppl: 11.188911121735945] tot_loss[loss=2.277, over 4831167.99 frames. , ppl: 9.749999323853944], batch size: 70
+2022-12-12 19:31:43,100 INFO [train.py:421] (2/8) Epoch 8, batch 4600, loss[loss=2.433, over 1610.00 frames. , ppl: 11.393639668232113] tot_loss[loss=2.276, over 4918527.89 frames. , ppl: 9.740831182833679], batch size: 70
+2022-12-12 19:33:23,008 INFO [train.py:421] (2/8) Epoch 8, batch 4800, loss[loss=2.322, over 2100.00 frames. , ppl: 10.19395145560806] tot_loss[loss=2.278, over 4911848.77 frames. , ppl: 9.759109193195481], batch size: 70
+2022-12-12 19:35:03,746 INFO [train.py:421] (2/8) Epoch 8, batch 5000, loss[loss=2.397, over 2660.00 frames. , ppl: 10.995170872637088] tot_loss[loss=2.278, over 4986402.77 frames. , ppl: 9.753400379775753], batch size: 70
+2022-12-12 19:35:03,746 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:35:04,513 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.275, over 211138.00 frames. , ppl: 9.73047586561236
+2022-12-12 19:36:45,801 INFO [train.py:421] (2/8) Epoch 8, batch 5200, loss[loss=2.268, over 3990.00 frames. , ppl: 9.658263031602985] tot_loss[loss=2.277, over 5033636.65 frames. , ppl: 9.752233684730445], batch size: 70
+2022-12-12 19:38:27,894 INFO [train.py:421] (2/8) Epoch 8, batch 5400, loss[loss=2.461, over 1050.00 frames. , ppl: 11.716432850301508] tot_loss[loss=2.278, over 5062018.17 frames. , ppl: 9.755783651552035], batch size: 70
+2022-12-12 19:40:10,516 INFO [train.py:421] (2/8) Epoch 8, batch 5600, loss[loss=2.289, over 3150.00 frames. , ppl: 9.869365926497522] tot_loss[loss=2.279, over 5085012.51 frames. , ppl: 9.765296420903725], batch size: 70
+2022-12-12 19:41:53,330 INFO [train.py:421] (2/8) Epoch 8, batch 5800, loss[loss=2.203, over 5320.00 frames. , ppl: 9.051555876100867] tot_loss[loss=2.277, over 5187136.57 frames. , ppl: 9.748221935572145], batch size: 70
+2022-12-12 19:43:36,498 INFO [train.py:421] (2/8) Epoch 8, batch 6000, loss[loss=2.321, over 2100.00 frames. , ppl: 10.1905246568519] tot_loss[loss=2.276, over 5254577.39 frames. , ppl: 9.736867421066261], batch size: 70
+2022-12-12 19:43:36,498 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:43:37,249 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.709706140891356
+2022-12-12 19:45:21,823 INFO [train.py:421] (2/8) Epoch 8, batch 6200, loss[loss=2.186, over 4200.00 frames. , ppl: 8.895971139602203] tot_loss[loss=2.274, over 5350129.73 frames. , ppl: 9.720912715246124], batch size: 70
+2022-12-12 19:47:05,569 INFO [train.py:421] (2/8) Epoch 8, batch 6400, loss[loss=2.225, over 6860.00 frames. , ppl: 9.251615076257863] tot_loss[loss=2.274, over 5410548.20 frames. , ppl: 9.715202138153918], batch size: 70
+2022-12-12 19:48:44,524 INFO [train.py:421] (2/8) Epoch 8, batch 6600, loss[loss=2.204, over 5390.00 frames. , ppl: 9.059450006877567] tot_loss[loss=2.274, over 5417264.31 frames. , ppl: 9.721894893667885], batch size: 70
+2022-12-12 19:50:24,283 INFO [train.py:421] (2/8) Epoch 8, batch 6800, loss[loss=2.277, over 1820.00 frames. , ppl: 9.74643274470885] tot_loss[loss=2.275, over 5408333.74 frames. , ppl: 9.729629503468049], batch size: 70
+2022-12-12 19:52:06,686 INFO [train.py:421] (2/8) Epoch 8, batch 7000, loss[loss=2.167, over 6930.00 frames. , ppl: 8.734337056949355] tot_loss[loss=2.275, over 5424178.85 frames. , ppl: 9.730400593187193], batch size: 70
+2022-12-12 19:52:06,687 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 19:52:07,437 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.703498378235237
+2022-12-12 19:53:49,599 INFO [train.py:421] (2/8) Epoch 8, batch 7200, loss[loss=2.393, over 1960.00 frames. , ppl: 10.949061521727229] tot_loss[loss=2.275, over 5456237.60 frames. , ppl: 9.723392126101444], batch size: 70
+2022-12-12 19:55:33,510 INFO [train.py:421] (2/8) Epoch 8, batch 7400, loss[loss=2.278, over 2520.00 frames. , ppl: 9.759141569228417] tot_loss[loss=2.274, over 5468562.55 frames. , ppl: 9.718149697665847], batch size: 70
+2022-12-12 19:57:12,619 INFO [train.py:421] (2/8) Epoch 8, batch 7600, loss[loss=2.35, over 2520.00 frames. , ppl: 10.484846461210786] tot_loss[loss=2.275, over 5444555.41 frames. , ppl: 9.73035863784032], batch size: 70
+2022-12-12 19:58:56,550 INFO [train.py:421] (2/8) Epoch 8, batch 7800, loss[loss=2.278, over 3080.00 frames. , ppl: 9.75407272078545] tot_loss[loss=2.275, over 5488470.64 frames. , ppl: 9.725555458660198], batch size: 70
+2022-12-12 20:00:37,121 INFO [train.py:421] (2/8) Epoch 8, batch 8000, loss[loss=2.2, over 5880.00 frames. , ppl: 9.028505096037795] tot_loss[loss=2.273, over 5534454.22 frames. , ppl: 9.713074368557109], batch size: 70
+2022-12-12 20:00:37,122 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:00:37,873 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.274, over 211138.00 frames. , ppl: 9.715900618606277
+2022-12-12 20:02:19,422 INFO [train.py:421] (2/8) Epoch 8, batch 8200, loss[loss=2.92, over 560.00 frames. , ppl: 18.533487042130993] tot_loss[loss=2.274, over 5495942.03 frames. , ppl: 9.718222802595555], batch size: 70
+2022-12-12 20:04:00,775 INFO [train.py:421] (2/8) Epoch 8, batch 8400, loss[loss=2.239, over 2660.00 frames. , ppl: 9.380643637734655] tot_loss[loss=2.274, over 5491202.39 frames. , ppl: 9.716181693682879], batch size: 70
+2022-12-12 20:05:43,630 INFO [train.py:421] (2/8) Epoch 8, batch 8600, loss[loss=2.249, over 3990.00 frames. , ppl: 9.477208421835964] tot_loss[loss=2.272, over 5548234.19 frames. , ppl: 9.703098996800538], batch size: 70
+2022-12-12 20:07:22,554 INFO [train.py:421] (2/8) Epoch 8, batch 8800, loss[loss=2.345, over 2660.00 frames. , ppl: 10.428693394936026] tot_loss[loss=2.273, over 5511065.49 frames. , ppl: 9.71145541468007], batch size: 70
+2022-12-12 20:09:06,514 INFO [train.py:421] (2/8) Epoch 8, batch 9000, loss[loss=2.161, over 3220.00 frames. , ppl: 8.675687590975384] tot_loss[loss=2.274, over 5485107.67 frames. , ppl: 9.721945768337685], batch size: 70
+2022-12-12 20:09:06,515 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:09:07,262 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.70899192399486
+2022-12-12 20:10:45,553 INFO [train.py:421] (2/8) Epoch 8, batch 9200, loss[loss=2.289, over 2940.00 frames. , ppl: 9.86260761745933] tot_loss[loss=2.275, over 5466823.13 frames. , ppl: 9.730282995641693], batch size: 70
+2022-12-12 20:12:29,543 INFO [train.py:421] (2/8) Epoch 8, batch 9400, loss[loss=2.273, over 2170.00 frames. , ppl: 9.70898140631201] tot_loss[loss=2.275, over 5493891.17 frames. , ppl: 9.723072370980143], batch size: 70
+2022-12-12 20:14:10,748 INFO [train.py:421] (2/8) Epoch 8, batch 9600, loss[loss=2.315, over 2940.00 frames. , ppl: 10.12734668465505] tot_loss[loss=2.275, over 5486483.71 frames. , ppl: 9.727246579914437], batch size: 70
+2022-12-12 20:15:50,130 INFO [train.py:421] (2/8) Epoch 8, batch 9800, loss[loss=2.404, over 1680.00 frames. , ppl: 11.068619805986764] tot_loss[loss=2.276, over 5457957.33 frames. , ppl: 9.736927618838267], batch size: 70
+2022-12-12 20:17:30,548 INFO [train.py:421] (2/8) Epoch 8, batch 10000, loss[loss=2.933, over 560.00 frames. , ppl: 18.782914421903104] tot_loss[loss=2.276, over 5441203.75 frames. , ppl: 9.737198930370463], batch size: 70
+2022-12-12 20:17:30,549 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:17:31,332 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.276, over 211138.00 frames. , ppl: 9.734638884938837
+2022-12-12 20:19:15,341 INFO [train.py:421] (2/8) Epoch 8, batch 10200, loss[loss=2.361, over 3850.00 frames. , ppl: 10.60677554409433] tot_loss[loss=2.275, over 5472771.68 frames. , ppl: 9.727103563292232], batch size: 70
+2022-12-12 20:20:52,751 INFO [train.py:421] (2/8) Epoch 8, batch 10400, loss[loss=2.452, over 1050.00 frames. , ppl: 11.607952031230525] tot_loss[loss=2.275, over 5480231.32 frames. , ppl: 9.726777450900398], batch size: 70
+2022-12-12 20:22:35,015 INFO [train.py:421] (2/8) Epoch 8, batch 10600, loss[loss=2.469, over 1330.00 frames. , ppl: 11.81091016422709] tot_loss[loss=2.276, over 5483473.17 frames. , ppl: 9.73570886804504], batch size: 70
+2022-12-12 20:24:13,555 INFO [train.py:421] (2/8) Epoch 8, batch 10800, loss[loss=2.429, over 1190.00 frames. , ppl: 11.34379537279937] tot_loss[loss=2.276, over 5484339.08 frames. , ppl: 9.735700586808749], batch size: 70
+2022-12-12 20:25:55,032 INFO [train.py:421] (2/8) Epoch 8, batch 11000, loss[loss=2.393, over 1400.00 frames. , ppl: 10.949007514113166] tot_loss[loss=2.276, over 5462817.23 frames. , ppl: 9.740485035752618], batch size: 70
+2022-12-12 20:25:55,033 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:25:55,810 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.706705920947643
+2022-12-12 20:27:38,328 INFO [train.py:421] (2/8) Epoch 8, batch 11200, loss[loss=2.269, over 4480.00 frames. , ppl: 9.667848607364848] tot_loss[loss=2.277, over 5468660.39 frames. , ppl: 9.745614435752252], batch size: 70
+2022-12-12 20:29:17,857 INFO [train.py:421] (2/8) Epoch 8, batch 11400, loss[loss=2.556, over 1680.00 frames. , ppl: 12.880426347142484] tot_loss[loss=2.277, over 5478692.30 frames. , ppl: 9.746381223891389], batch size: 70
+2022-12-12 20:30:55,493 INFO [train.py:421] (2/8) Epoch 8, batch 11600, loss[loss=2.331, over 1890.00 frames. , ppl: 10.28535622422131] tot_loss[loss=2.276, over 5484787.59 frames. , ppl: 9.740770805220915], batch size: 70
+2022-12-12 20:32:40,055 INFO [train.py:421] (2/8) Epoch 8, batch 11800, loss[loss=2.397, over 1260.00 frames. , ppl: 10.986838923073686] tot_loss[loss=2.276, over 5498042.12 frames. , ppl: 9.741349167827577], batch size: 70
+2022-12-12 20:34:17,575 INFO [train.py:421] (2/8) Epoch 8, batch 12000, loss[loss=2.823, over 630.00 frames. , ppl: 16.826819634759783] tot_loss[loss=2.277, over 5492327.45 frames. , ppl: 9.742925748442207], batch size: 70
+2022-12-12 20:34:17,575 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:34:18,324 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.712665599212405
+2022-12-12 20:36:00,924 INFO [train.py:421] (2/8) Epoch 8, batch 12200, loss[loss=2.229, over 2870.00 frames. , ppl: 9.29330411188987] tot_loss[loss=2.275, over 5548839.68 frames. , ppl: 9.730224695275789], batch size: 70
+2022-12-12 20:37:42,034 INFO [train.py:421] (2/8) Epoch 8, batch 12400, loss[loss=2.325, over 2450.00 frames. , ppl: 10.228946772750263] tot_loss[loss=2.276, over 5523857.87 frames. , ppl: 9.736068488282186], batch size: 70
+2022-12-12 20:39:22,979 INFO [train.py:421] (2/8) Epoch 8, batch 12600, loss[loss=2.268, over 4900.00 frames. , ppl: 9.65971328205001] tot_loss[loss=2.276, over 5532466.75 frames. , ppl: 9.736286229112443], batch size: 70
+2022-12-12 20:41:03,255 INFO [train.py:421] (2/8) Epoch 8, batch 12800, loss[loss=2.172, over 3010.00 frames. , ppl: 8.773098665340921] tot_loss[loss=2.274, over 5577557.84 frames. , ppl: 9.722616585930666], batch size: 70
+2022-12-12 20:42:44,548 INFO [train.py:421] (2/8) Epoch 8, batch 13000, loss[loss=2.246, over 2660.00 frames. , ppl: 9.447376504681301] tot_loss[loss=2.275, over 5563399.44 frames. , ppl: 9.72928248940891], batch size: 70
+2022-12-12 20:42:44,549 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:42:45,301 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.274, over 211138.00 frames. , ppl: 9.722002639478895
+2022-12-12 20:44:28,071 INFO [train.py:421] (2/8) Epoch 8, batch 13200, loss[loss=2.33, over 2100.00 frames. , ppl: 10.282683939132243] tot_loss[loss=2.275, over 5575444.55 frames. , ppl: 9.726132855758552], batch size: 70
+2022-12-12 20:46:10,752 INFO [train.py:421] (2/8) Epoch 8, batch 13400, loss[loss=2.278, over 2660.00 frames. , ppl: 9.752634078505313] tot_loss[loss=2.275, over 5574635.32 frames. , ppl: 9.728214718335318], batch size: 70
+2022-12-12 20:47:53,569 INFO [train.py:421] (2/8) Epoch 8, batch 13600, loss[loss=2.859, over 840.00 frames. , ppl: 17.4453124803324] tot_loss[loss=2.275, over 5596817.01 frames. , ppl: 9.728781107378895], batch size: 70
+2022-12-12 20:49:37,915 INFO [train.py:421] (2/8) Epoch 8, batch 13800, loss[loss=2.158, over 5390.00 frames. , ppl: 8.653361358159039] tot_loss[loss=2.276, over 5571907.87 frames. , ppl: 9.735208408943937], batch size: 70
+2022-12-12 20:51:19,926 INFO [train.py:421] (2/8) Epoch 8, batch 14000, loss[loss=2.221, over 2590.00 frames. , ppl: 9.220187659871785] tot_loss[loss=2.275, over 5608117.91 frames. , ppl: 9.727373406471834], batch size: 70
+2022-12-12 20:51:19,926 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:51:20,692 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.699218039613152
+2022-12-12 20:53:03,140 INFO [train.py:421] (2/8) Epoch 8, batch 14200, loss[loss=2.173, over 3220.00 frames. , ppl: 8.783780634494791] tot_loss[loss=2.275, over 5622742.83 frames. , ppl: 9.725072169317786], batch size: 70
+2022-12-12 20:54:47,641 INFO [train.py:421] (2/8) Epoch 8, batch 14400, loss[loss=2.21, over 7560.00 frames. , ppl: 9.11518369762455] tot_loss[loss=2.276, over 5591681.98 frames. , ppl: 9.736106724664785], batch size: 70
+2022-12-12 20:56:30,676 INFO [train.py:421] (2/8) Epoch 8, batch 14600, loss[loss=2.217, over 4970.00 frames. , ppl: 9.182592031864992] tot_loss[loss=2.276, over 5618929.76 frames. , ppl: 9.73355794550798], batch size: 70
+2022-12-12 20:58:11,812 INFO [train.py:421] (2/8) Epoch 8, batch 14800, loss[loss=2.532, over 980.00 frames. , ppl: 12.583127809508696] tot_loss[loss=2.275, over 5636246.69 frames. , ppl: 9.732413205664374], batch size: 70
+2022-12-12 20:59:52,436 INFO [train.py:421] (2/8) Epoch 8, batch 15000, loss[loss=2.263, over 2800.00 frames. , ppl: 9.607599582496475] tot_loss[loss=2.275, over 5651690.31 frames. , ppl: 9.725646110023842], batch size: 70
+2022-12-12 20:59:52,437 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 20:59:53,203 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.698958207336654
+2022-12-12 21:01:33,685 INFO [train.py:421] (2/8) Epoch 8, batch 15200, loss[loss=2.199, over 6370.00 frames. , ppl: 9.01846870985729] tot_loss[loss=2.275, over 5635754.24 frames. , ppl: 9.72995969358553], batch size: 70
+2022-12-12 21:03:13,943 INFO [train.py:421] (2/8) Epoch 8, batch 15400, loss[loss=2.264, over 4340.00 frames. , ppl: 9.622168936703309] tot_loss[loss=2.275, over 5645496.19 frames. , ppl: 9.725226183698688], batch size: 70
+2022-12-12 21:04:56,825 INFO [train.py:421] (2/8) Epoch 8, batch 15600, loss[loss=2.245, over 3220.00 frames. , ppl: 9.436615280808759] tot_loss[loss=2.275, over 5637867.22 frames. , ppl: 9.72849351473297], batch size: 70
+2022-12-12 21:06:37,397 INFO [train.py:421] (2/8) Epoch 8, batch 15800, loss[loss=2.212, over 5600.00 frames. , ppl: 9.134789922010254] tot_loss[loss=2.276, over 5601221.78 frames. , ppl: 9.735396971254652], batch size: 70
+2022-12-12 21:08:18,882 INFO [train.py:421] (2/8) Epoch 8, batch 16000, loss[loss=2.212, over 4620.00 frames. , ppl: 9.13849169758533] tot_loss[loss=2.275, over 5611811.68 frames. , ppl: 9.730000205551203], batch size: 70
+2022-12-12 21:08:18,883 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:08:19,648 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.704328531673617
+2022-12-12 21:09:58,158 INFO [train.py:421] (2/8) Epoch 8, batch 16200, loss[loss=2.404, over 1680.00 frames. , ppl: 11.067419921254535] tot_loss[loss=2.276, over 5584349.26 frames. , ppl: 9.740045979555216], batch size: 70
+2022-12-12 21:11:38,668 INFO [train.py:421] (2/8) Epoch 8, batch 16400, loss[loss=3.244, over 490.00 frames. , ppl: 25.641474208566045] tot_loss[loss=2.277, over 5558283.45 frames. , ppl: 9.744970212338066], batch size: 70
+2022-12-12 21:13:18,887 INFO [train.py:421] (2/8) Epoch 8, batch 16600, loss[loss=2.211, over 2940.00 frames. , ppl: 9.121158875534801] tot_loss[loss=2.276, over 5583318.93 frames. , ppl: 9.736150751724864], batch size: 70
+2022-12-12 21:14:59,036 INFO [train.py:421] (2/8) Epoch 8, batch 16800, loss[loss=2.15, over 5250.00 frames. , ppl: 8.587244476189024] tot_loss[loss=2.277, over 5538224.71 frames. , ppl: 9.743376665859293], batch size: 70
+2022-12-12 21:16:43,635 INFO [train.py:421] (2/8) Epoch 8, batch 17000, loss[loss=2.329, over 3290.00 frames. , ppl: 10.263927290078815] tot_loss[loss=2.277, over 5516483.44 frames. , ppl: 9.747302256751587], batch size: 70
+2022-12-12 21:16:43,635 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:16:44,419 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.701667408509223
+2022-12-12 21:18:25,170 INFO [train.py:421] (2/8) Epoch 8, batch 17200, loss[loss=2.228, over 3990.00 frames. , ppl: 9.27951941956132] tot_loss[loss=2.277, over 5516172.17 frames. , ppl: 9.746895935182858], batch size: 70
+2022-12-12 21:20:08,355 INFO [train.py:421] (2/8) Epoch 8, batch 17400, loss[loss=2.302, over 1890.00 frames. , ppl: 9.995024977324851] tot_loss[loss=2.278, over 5488872.62 frames. , ppl: 9.757262519384224], batch size: 70
+2022-12-12 21:21:48,862 INFO [train.py:421] (2/8) Epoch 8, batch 17600, loss[loss=2.247, over 5110.00 frames. , ppl: 9.463962199980138] tot_loss[loss=2.279, over 5471100.16 frames. , ppl: 9.767599353374], batch size: 70
+2022-12-12 21:23:31,880 INFO [train.py:421] (2/8) Epoch 8, batch 17800, loss[loss=2.252, over 2450.00 frames. , ppl: 9.507927428327772] tot_loss[loss=2.28, over 5457824.39 frames. , ppl: 9.772338572888337], batch size: 70
+2022-12-12 21:25:12,513 INFO [train.py:421] (2/8) Epoch 8, batch 18000, loss[loss=2.629, over 910.00 frames. , ppl: 13.854321201308752] tot_loss[loss=2.278, over 5481943.12 frames. , ppl: 9.76047297968516], batch size: 70
+2022-12-12 21:25:12,513 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:25:13,280 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.693148970049764
+2022-12-12 21:26:52,987 INFO [train.py:421] (2/8) Epoch 8, batch 18200, loss[loss=2.355, over 1680.00 frames. , ppl: 10.543360877750315] tot_loss[loss=2.278, over 5496754.28 frames. , ppl: 9.752749606828118], batch size: 70
+2022-12-12 21:28:32,811 INFO [train.py:421] (2/8) Epoch 8, batch 18400, loss[loss=2.381, over 1540.00 frames. , ppl: 10.817062613108009] tot_loss[loss=2.278, over 5490201.31 frames. , ppl: 9.757821868559585], batch size: 70
+2022-12-12 21:30:13,679 INFO [train.py:421] (2/8) Epoch 8, batch 18600, loss[loss=2.575, over 980.00 frames. , ppl: 13.130401222441943] tot_loss[loss=2.278, over 5485360.25 frames. , ppl: 9.754903565779621], batch size: 70
+2022-12-12 21:31:51,675 INFO [train.py:421] (2/8) Epoch 8, batch 18800, loss[loss=2.316, over 1400.00 frames. , ppl: 10.133387432690254] tot_loss[loss=2.278, over 5482247.58 frames. , ppl: 9.75975230173189], batch size: 70
+2022-12-12 21:33:32,792 INFO [train.py:421] (2/8) Epoch 8, batch 19000, loss[loss=2.803, over 700.00 frames. , ppl: 16.49790607308537] tot_loss[loss=2.28, over 5419933.85 frames. , ppl: 9.773066452750827], batch size: 70
+2022-12-12 21:33:32,792 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:33:33,556 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.703235558989775
+2022-12-12 21:35:14,757 INFO [train.py:421] (2/8) Epoch 8, batch 19200, loss[loss=2.215, over 3920.00 frames. , ppl: 9.159427816946096] tot_loss[loss=2.28, over 5374168.94 frames. , ppl: 9.776933288809195], batch size: 70
+2022-12-12 21:36:52,519 INFO [train.py:421] (2/8) Epoch 8, batch 19400, loss[loss=2.378, over 1260.00 frames. , ppl: 10.780958317569405] tot_loss[loss=2.281, over 5354211.55 frames. , ppl: 9.784021814154563], batch size: 70
+2022-12-12 21:38:36,758 INFO [train.py:421] (2/8) Epoch 8, batch 19600, loss[loss=2.651, over 840.00 frames. , ppl: 14.17089872515275] tot_loss[loss=2.279, over 5421875.09 frames. , ppl: 9.766450814998704], batch size: 70
+2022-12-12 21:40:17,794 INFO [train.py:421] (2/8) Epoch 8, batch 19800, loss[loss=2.313, over 1330.00 frames. , ppl: 10.108656367706567] tot_loss[loss=2.278, over 5454132.45 frames. , ppl: 9.755400445240443], batch size: 70
+2022-12-12 21:42:00,376 INFO [train.py:421] (2/8) Epoch 8, batch 20000, loss[loss=2.398, over 1470.00 frames. , ppl: 10.996651459352071] tot_loss[loss=2.277, over 5477330.50 frames. , ppl: 9.748643673423135], batch size: 70
+2022-12-12 21:42:00,377 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:42:01,125 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.70147643313627
+2022-12-12 21:43:41,743 INFO [train.py:421] (2/8) Epoch 8, batch 20200, loss[loss=2.269, over 3430.00 frames. , ppl: 9.664972616847145] tot_loss[loss=2.277, over 5454055.00 frames. , ppl: 9.752061972146638], batch size: 70
+2022-12-12 21:45:25,099 INFO [train.py:421] (2/8) Epoch 8, batch 20400, loss[loss=2.254, over 2800.00 frames. , ppl: 9.52161643051294] tot_loss[loss=2.277, over 5480761.04 frames. , ppl: 9.751631283834582], batch size: 70
+2022-12-12 21:47:09,832 INFO [train.py:421] (2/8) Epoch 8, batch 20600, loss[loss=2.16, over 9240.00 frames. , ppl: 8.675448198133298] tot_loss[loss=2.278, over 5461613.52 frames. , ppl: 9.759439113823177], batch size: 70
+2022-12-12 21:48:48,198 INFO [train.py:421] (2/8) Epoch 8, batch 20800, loss[loss=2.362, over 1470.00 frames. , ppl: 10.615555923595778] tot_loss[loss=2.278, over 5471444.84 frames. , ppl: 9.75695104726982], batch size: 70
+2022-12-12 21:50:31,019 INFO [train.py:421] (2/8) Epoch 8, batch 21000, loss[loss=3.137, over 490.00 frames. , ppl: 23.036099751054888] tot_loss[loss=2.278, over 5477863.30 frames. , ppl: 9.754763025805504], batch size: 70
+2022-12-12 21:50:31,020 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:50:31,777 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.699197941840623
+2022-12-12 21:52:13,972 INFO [train.py:421] (2/8) Epoch 8, batch 21200, loss[loss=2.43, over 770.00 frames. , ppl: 11.357936362209077] tot_loss[loss=2.279, over 5451698.17 frames. , ppl: 9.763886769332302], batch size: 70
+2022-12-12 21:53:55,495 INFO [train.py:421] (2/8) Epoch 8, batch 21400, loss[loss=2.327, over 2730.00 frames. , ppl: 10.246600281327247] tot_loss[loss=2.28, over 5422011.23 frames. , ppl: 9.773844995811103], batch size: 70
+2022-12-12 21:55:37,162 INFO [train.py:421] (2/8) Epoch 8, batch 21600, loss[loss=2.133, over 6370.00 frames. , ppl: 8.442878604213181] tot_loss[loss=2.28, over 5423034.54 frames. , ppl: 9.773914027331028], batch size: 70
+2022-12-12 21:57:17,867 INFO [train.py:421] (2/8) Epoch 8, batch 21800, loss[loss=2.414, over 2240.00 frames. , ppl: 11.17971376467778] tot_loss[loss=2.279, over 5432033.40 frames. , ppl: 9.767794670558892], batch size: 70
+2022-12-12 21:59:00,465 INFO [train.py:421] (2/8) Epoch 8, batch 22000, loss[loss=2.341, over 1680.00 frames. , ppl: 10.394546288333904] tot_loss[loss=2.279, over 5419248.22 frames. , ppl: 9.77140734447313], batch size: 70
+2022-12-12 21:59:00,466 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 21:59:01,248 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.689339266199932
+2022-12-12 22:00:41,770 INFO [train.py:421] (2/8) Epoch 8, batch 22200, loss[loss=2.41, over 1050.00 frames. , ppl: 11.12918322422793] tot_loss[loss=2.28, over 5402087.45 frames. , ppl: 9.776418191095788], batch size: 70
+2022-12-12 22:02:23,779 INFO [train.py:421] (2/8) Epoch 8, batch 22400, loss[loss=2.378, over 2450.00 frames. , ppl: 10.778515066712293] tot_loss[loss=2.278, over 5434956.56 frames. , ppl: 9.761630133870554], batch size: 70
+2022-12-12 22:04:05,301 INFO [train.py:421] (2/8) Epoch 8, batch 22600, loss[loss=2.472, over 980.00 frames. , ppl: 11.846464354046264] tot_loss[loss=2.279, over 5408708.79 frames. , ppl: 9.769683175747472], batch size: 70
+2022-12-12 22:05:46,665 INFO [train.py:421] (2/8) Epoch 8, batch 22800, loss[loss=3.353, over 490.00 frames. , ppl: 28.592435490238966] tot_loss[loss=2.277, over 5464099.75 frames. , ppl: 9.751762055605143], batch size: 70
+2022-12-12 22:07:29,980 INFO [train.py:421] (2/8) Epoch 8, batch 23000, loss[loss=2.415, over 1540.00 frames. , ppl: 11.187034202344105] tot_loss[loss=2.278, over 5465779.34 frames. , ppl: 9.753074193138687], batch size: 70
+2022-12-12 22:07:29,981 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:07:30,755 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.695581021092998
+2022-12-12 22:09:13,828 INFO [train.py:421] (2/8) Epoch 8, batch 23200, loss[loss=2.195, over 7560.00 frames. , ppl: 8.980473417594903] tot_loss[loss=2.277, over 5500767.76 frames. , ppl: 9.748604164346316], batch size: 70
+2022-12-12 22:10:53,394 INFO [train.py:421] (2/8) Epoch 8, batch 23400, loss[loss=2.603, over 770.00 frames. , ppl: 13.50681587139859] tot_loss[loss=2.277, over 5487206.99 frames. , ppl: 9.748244526814227], batch size: 70
+2022-12-12 22:12:32,785 INFO [train.py:421] (2/8) Epoch 8, batch 23600, loss[loss=2.518, over 910.00 frames. , ppl: 12.401383592735849] tot_loss[loss=2.277, over 5484841.67 frames. , ppl: 9.748787042417689], batch size: 70
+2022-12-12 22:14:11,811 INFO [train.py:421] (2/8) Epoch 8, batch 23800, loss[loss=2.602, over 770.00 frames. , ppl: 13.48929021461403] tot_loss[loss=2.277, over 5462975.98 frames. , ppl: 9.74967735306486], batch size: 70
+2022-12-12 22:15:55,041 INFO [train.py:421] (2/8) Epoch 8, batch 24000, loss[loss=2.415, over 1610.00 frames. , ppl: 11.192753082932878] tot_loss[loss=2.277, over 5470916.64 frames. , ppl: 9.748193616806685], batch size: 70
+2022-12-12 22:15:55,041 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:15:55,805 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.708671477552679
+2022-12-12 22:17:36,359 INFO [train.py:421] (2/8) Epoch 8, batch 24200, loss[loss=2.394, over 1470.00 frames. , ppl: 10.956308594643867] tot_loss[loss=2.278, over 5457029.96 frames. , ppl: 9.756137220014741], batch size: 70
+2022-12-12 22:19:17,624 INFO [train.py:421] (2/8) Epoch 8, batch 24400, loss[loss=2.421, over 1330.00 frames. , ppl: 11.257042029267431] tot_loss[loss=2.277, over 5456127.55 frames. , ppl: 9.749223688621026], batch size: 70
+2022-12-12 22:20:57,978 INFO [train.py:421] (2/8) Epoch 8, batch 24600, loss[loss=2.404, over 1330.00 frames. , ppl: 11.072114891812534] tot_loss[loss=2.277, over 5478899.12 frames. , ppl: 9.742617474156713], batch size: 70
+2022-12-12 22:22:37,693 INFO [train.py:421] (2/8) Epoch 8, batch 24800, loss[loss=2.384, over 1260.00 frames. , ppl: 10.846410645148326] tot_loss[loss=2.276, over 5480199.29 frames. , ppl: 9.741187841630879], batch size: 70
+2022-12-12 22:24:20,891 INFO [train.py:421] (2/8) Epoch 8, batch 25000, loss[loss=2.178, over 2730.00 frames. , ppl: 8.82479641494226] tot_loss[loss=2.276, over 5484558.99 frames. , ppl: 9.740864638057795], batch size: 70
+2022-12-12 22:24:20,892 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:24:21,639 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.690235616742381
+2022-12-12 22:26:02,834 INFO [train.py:421] (2/8) Epoch 8, batch 25200, loss[loss=2.285, over 2450.00 frames. , ppl: 9.829272423483665] tot_loss[loss=2.277, over 5472002.71 frames. , ppl: 9.749934288597485], batch size: 70
+2022-12-12 22:27:41,826 INFO [train.py:421] (2/8) Epoch 8, batch 25400, loss[loss=2.495, over 1260.00 frames. , ppl: 12.124192424993556] tot_loss[loss=2.278, over 5447913.52 frames. , ppl: 9.758885959813933], batch size: 70
+2022-12-12 22:29:21,135 INFO [train.py:421] (2/8) Epoch 8, batch 25600, loss[loss=2.211, over 11690.00 frames. , ppl: 9.126937613521235] tot_loss[loss=2.278, over 5434928.06 frames. , ppl: 9.759315883607375], batch size: 70
+2022-12-12 22:31:04,404 INFO [train.py:421] (2/8) Epoch 8, batch 25800, loss[loss=2.398, over 1680.00 frames. , ppl: 10.996247615897078] tot_loss[loss=2.278, over 5456532.17 frames. , ppl: 9.759406681453076], batch size: 70
+2022-12-12 22:32:42,660 INFO [train.py:421] (2/8) Epoch 8, batch 26000, loss[loss=2.37, over 1820.00 frames. , ppl: 10.701726234775606] tot_loss[loss=2.28, over 5421289.71 frames. , ppl: 9.772977326672795], batch size: 70
+2022-12-12 22:32:42,660 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:32:43,419 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.708037800704728
+2022-12-12 22:34:22,025 INFO [train.py:421] (2/8) Epoch 8, batch 26200, loss[loss=2.548, over 1050.00 frames. , ppl: 12.78086612740012] tot_loss[loss=2.279, over 5430416.58 frames. , ppl: 9.770583324767829], batch size: 70
+2022-12-12 22:36:01,414 INFO [train.py:421] (2/8) Epoch 8, batch 26400, loss[loss=2.318, over 2940.00 frames. , ppl: 10.157805789553622] tot_loss[loss=2.278, over 5472583.99 frames. , ppl: 9.756552832928469], batch size: 70
+2022-12-12 22:37:43,186 INFO [train.py:421] (2/8) Epoch 8, batch 26600, loss[loss=2.382, over 1890.00 frames. , ppl: 10.826715990712573] tot_loss[loss=2.277, over 5507255.45 frames. , ppl: 9.744713961860878], batch size: 70
+2022-12-12 22:39:21,935 INFO [train.py:421] (2/8) Epoch 8, batch 26800, loss[loss=2.258, over 3710.00 frames. , ppl: 9.56352848128541] tot_loss[loss=2.276, over 5536849.21 frames. , ppl: 9.738095350454339], batch size: 70
+2022-12-12 22:40:59,518 INFO [train.py:421] (2/8) Epoch 8, batch 27000, loss[loss=2.68, over 770.00 frames. , ppl: 14.587282667005436] tot_loss[loss=2.277, over 5514243.68 frames. , ppl: 9.744389056716642], batch size: 70
+2022-12-12 22:40:59,519 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:41:00,266 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.710776846115149
+2022-12-12 22:42:40,998 INFO [train.py:421] (2/8) Epoch 8, batch 27200, loss[loss=2.417, over 1470.00 frames. , ppl: 11.21340874781153] tot_loss[loss=2.279, over 5456650.04 frames. , ppl: 9.762211679874467], batch size: 70
+2022-12-12 22:44:23,273 INFO [train.py:421] (2/8) Epoch 8, batch 27400, loss[loss=2.451, over 3150.00 frames. , ppl: 11.605366656508275] tot_loss[loss=2.279, over 5463987.90 frames. , ppl: 9.763151463205789], batch size: 70
+2022-12-12 22:46:01,804 INFO [train.py:421] (2/8) Epoch 8, batch 27600, loss[loss=2.5, over 1120.00 frames. , ppl: 12.179801507318636] tot_loss[loss=2.279, over 5462821.96 frames. , ppl: 9.764204867096979], batch size: 70
+2022-12-12 22:47:40,353 INFO [train.py:421] (2/8) Epoch 8, batch 27800, loss[loss=2.282, over 2520.00 frames. , ppl: 9.793731557412846] tot_loss[loss=2.277, over 5488080.91 frames. , ppl: 9.750498501009595], batch size: 70
+2022-12-12 22:49:20,551 INFO [train.py:421] (2/8) Epoch 8, batch 28000, loss[loss=2.948, over 560.00 frames. , ppl: 19.07630695062492] tot_loss[loss=2.278, over 5480932.63 frames. , ppl: 9.754553343249261], batch size: 70
+2022-12-12 22:49:20,552 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:49:21,297 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.709539437680133
+2022-12-12 22:51:00,491 INFO [train.py:421] (2/8) Epoch 8, batch 28200, loss[loss=2.295, over 1890.00 frames. , ppl: 9.922953118940942] tot_loss[loss=2.278, over 5488474.57 frames. , ppl: 9.760543287110576], batch size: 70
+2022-12-12 22:52:39,726 INFO [train.py:421] (2/8) Epoch 8, batch 28400, loss[loss=2.21, over 8820.00 frames. , ppl: 9.113909109498207] tot_loss[loss=2.278, over 5477951.38 frames. , ppl: 9.760989336684343], batch size: 70
+2022-12-12 22:54:17,923 INFO [train.py:421] (2/8) Epoch 8, batch 28600, loss[loss=2.566, over 700.00 frames. , ppl: 13.008659282563242] tot_loss[loss=2.28, over 5438114.86 frames. , ppl: 9.772286023378712], batch size: 70
+2022-12-12 22:55:59,834 INFO [train.py:421] (2/8) Epoch 8, batch 28800, loss[loss=2.246, over 5600.00 frames. , ppl: 9.445377094890654] tot_loss[loss=2.279, over 5454129.68 frames. , ppl: 9.766776327449872], batch size: 70
+2022-12-12 22:57:42,761 INFO [train.py:421] (2/8) Epoch 8, batch 29000, loss[loss=2.221, over 7840.00 frames. , ppl: 9.218679114076833] tot_loss[loss=2.28, over 5441102.86 frames. , ppl: 9.772996723097569], batch size: 70
+2022-12-12 22:57:42,762 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 22:57:43,506 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.696305732449282
+2022-12-12 22:59:22,325 INFO [train.py:421] (2/8) Epoch 8, batch 29200, loss[loss=2.318, over 2240.00 frames. , ppl: 10.153447757327081] tot_loss[loss=2.278, over 5494817.30 frames. , ppl: 9.756267522393097], batch size: 70
+2022-12-12 23:01:03,744 INFO [train.py:421] (2/8) Epoch 8, batch 29400, loss[loss=2.378, over 2170.00 frames. , ppl: 10.784701479530034] tot_loss[loss=2.277, over 5519029.91 frames. , ppl: 9.744558655370414], batch size: 70
+2022-12-12 23:02:44,759 INFO [train.py:421] (2/8) Epoch 8, batch 29600, loss[loss=2.428, over 1680.00 frames. , ppl: 11.334488944885303] tot_loss[loss=2.276, over 5542252.65 frames. , ppl: 9.739309430796622], batch size: 70
+2022-12-12 23:04:24,090 INFO [train.py:421] (2/8) Epoch 8, batch 29800, loss[loss=2.648, over 770.00 frames. , ppl: 14.12741539170013] tot_loss[loss=2.277, over 5536598.17 frames. , ppl: 9.746132716595467], batch size: 70
+2022-12-12 23:06:04,136 INFO [train.py:421] (2/8) Epoch 8, batch 30000, loss[loss=2.141, over 7140.00 frames. , ppl: 8.50715499872523] tot_loss[loss=2.277, over 5493352.19 frames. , ppl: 9.750945125846695], batch size: 70
+2022-12-12 23:06:04,137 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:06:04,900 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.705356986317165
+2022-12-12 23:07:46,652 INFO [train.py:421] (2/8) Epoch 8, batch 30200, loss[loss=2.345, over 2450.00 frames. , ppl: 10.436712515024862] tot_loss[loss=2.278, over 5472521.94 frames. , ppl: 9.75790107319879], batch size: 70
+2022-12-12 23:09:28,046 INFO [train.py:421] (2/8) Epoch 8, batch 30400, loss[loss=2.336, over 1400.00 frames. , ppl: 10.341277906247706] tot_loss[loss=2.277, over 5517021.18 frames. , ppl: 9.746906071435776], batch size: 70
+2022-12-12 23:11:07,277 INFO [train.py:421] (2/8) Epoch 8, batch 30600, loss[loss=2.257, over 2450.00 frames. , ppl: 9.554852941751111] tot_loss[loss=2.277, over 5519632.16 frames. , ppl: 9.749921038281915], batch size: 70
+2022-12-12 23:12:43,210 INFO [train.py:421] (2/8) Epoch 8, batch 30800, loss[loss=2.35, over 1260.00 frames. , ppl: 10.485297479233987] tot_loss[loss=2.277, over 5520833.81 frames. , ppl: 9.74750775719102], batch size: 70
+2022-12-12 23:14:23,185 INFO [train.py:421] (2/8) Epoch 8, batch 31000, loss[loss=2.259, over 4830.00 frames. , ppl: 9.573812289845193] tot_loss[loss=2.279, over 5501888.10 frames. , ppl: 9.763681748242151], batch size: 70
+2022-12-12 23:14:23,186 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:14:23,945 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.697824214248042
+2022-12-12 23:16:06,469 INFO [train.py:421] (2/8) Epoch 8, batch 31200, loss[loss=2.266, over 2520.00 frames. , ppl: 9.643747291987843] tot_loss[loss=2.277, over 5519767.31 frames. , ppl: 9.750962914673797], batch size: 70
+2022-12-12 23:17:46,226 INFO [train.py:421] (2/8) Epoch 8, batch 31400, loss[loss=2.342, over 1820.00 frames. , ppl: 10.40093115328595] tot_loss[loss=2.278, over 5523862.77 frames. , ppl: 9.75289626782093], batch size: 70
+2022-12-12 23:19:24,625 INFO [train.py:421] (2/8) Epoch 8, batch 31600, loss[loss=2.547, over 1190.00 frames. , ppl: 12.768597852582054] tot_loss[loss=2.277, over 5534176.46 frames. , ppl: 9.748358769764065], batch size: 70
+2022-12-12 23:21:02,585 INFO [train.py:421] (2/8) Epoch 8, batch 31800, loss[loss=2.274, over 3010.00 frames. , ppl: 9.722088251109364] tot_loss[loss=2.278, over 5536244.91 frames. , ppl: 9.752363400913307], batch size: 70
+2022-12-12 23:22:42,639 INFO [train.py:421] (2/8) Epoch 8, batch 32000, loss[loss=2.164, over 7420.00 frames. , ppl: 8.709308983963835] tot_loss[loss=2.278, over 5532098.93 frames. , ppl: 9.753139680456915], batch size: 70
+2022-12-12 23:22:42,639 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:22:43,385 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.681927833421199
+2022-12-12 23:24:20,873 INFO [train.py:421] (2/8) Epoch 8, batch 32200, loss[loss=2.311, over 2940.00 frames. , ppl: 10.08420043838879] tot_loss[loss=2.279, over 5495535.59 frames. , ppl: 9.764127986179048], batch size: 70
+2022-12-12 23:26:03,836 INFO [train.py:421] (2/8) Epoch 8, batch 32400, loss[loss=2.227, over 4200.00 frames. , ppl: 9.269538841713791] tot_loss[loss=2.277, over 5539479.07 frames. , ppl: 9.748351881591041], batch size: 70
+2022-12-12 23:27:44,838 INFO [train.py:421] (2/8) Epoch 8, batch 32600, loss[loss=2.148, over 7560.00 frames. , ppl: 8.569074843731329] tot_loss[loss=2.277, over 5563632.62 frames. , ppl: 9.744369241884476], batch size: 70
+2022-12-12 23:29:26,644 INFO [train.py:421] (2/8) Epoch 8, batch 32800, loss[loss=2.575, over 1400.00 frames. , ppl: 13.134708967962412] tot_loss[loss=2.277, over 5559326.21 frames. , ppl: 9.745434084090077], batch size: 70
+2022-12-12 23:31:04,174 INFO [train.py:421] (2/8) Epoch 8, batch 33000, loss[loss=2.431, over 1960.00 frames. , ppl: 11.365629098403334] tot_loss[loss=2.276, over 5562179.08 frames. , ppl: 9.739675561501565], batch size: 70
+2022-12-12 23:31:04,174 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:31:04,954 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.70102126639035
+2022-12-12 23:32:46,880 INFO [train.py:421] (2/8) Epoch 8, batch 33200, loss[loss=2.261, over 4130.00 frames. , ppl: 9.591762264430397] tot_loss[loss=2.277, over 5538019.46 frames. , ppl: 9.7487762978215], batch size: 70
+2022-12-12 23:34:29,597 INFO [train.py:421] (2/8) Epoch 8, batch 33400, loss[loss=2.326, over 2310.00 frames. , ppl: 10.235598768219027] tot_loss[loss=2.276, over 5536285.25 frames. , ppl: 9.740800226923701], batch size: 70
+2022-12-12 23:36:09,713 INFO [train.py:421] (2/8) Epoch 8, batch 33600, loss[loss=2.209, over 4690.00 frames. , ppl: 9.111014885051544] tot_loss[loss=2.276, over 5563842.21 frames. , ppl: 9.737600160777136], batch size: 70
+2022-12-12 23:37:51,012 INFO [train.py:421] (2/8) Epoch 8, batch 33800, loss[loss=2.373, over 1400.00 frames. , ppl: 10.726885018869366] tot_loss[loss=2.277, over 5546618.74 frames. , ppl: 9.74477886716879], batch size: 70
+2022-12-12 23:39:28,191 INFO [train.py:421] (2/8) Epoch 8, batch 34000, loss[loss=2.467, over 840.00 frames. , ppl: 11.783378280711805] tot_loss[loss=2.276, over 5560827.38 frames. , ppl: 9.738876518270054], batch size: 70
+2022-12-12 23:39:28,192 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:39:28,937 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.680815891149308
+2022-12-12 23:41:01,892 INFO [train.py:421] (2/8) Epoch 8, batch 34200, loss[loss=2.219, over 1750.00 frames. , ppl: 9.196979726048566] tot_loss[loss=2.277, over 5519277.78 frames. , ppl: 9.746921753480018], batch size: 70
+2022-12-12 23:42:44,182 INFO [train.py:421] (2/8) Epoch 8, batch 34400, loss[loss=4.082, over 350.00 frames. , ppl: 59.289134523422256] tot_loss[loss=2.277, over 5540959.70 frames. , ppl: 9.74406808649771], batch size: 70
+2022-12-12 23:44:28,311 INFO [train.py:421] (2/8) Epoch 8, batch 34600, loss[loss=2.309, over 3360.00 frames. , ppl: 10.060828866476575] tot_loss[loss=2.276, over 5577181.97 frames. , ppl: 9.736344594943896], batch size: 70
+2022-12-12 23:46:06,962 INFO [train.py:421] (2/8) Epoch 8, batch 34800, loss[loss=2.243, over 8470.00 frames. , ppl: 9.425752932522785] tot_loss[loss=2.277, over 5505775.96 frames. , ppl: 9.75096915492476], batch size: 70
+2022-12-12 23:47:51,640 INFO [train.py:421] (2/8) Epoch 8, batch 35000, loss[loss=2.107, over 8680.00 frames. , ppl: 8.225883779608997] tot_loss[loss=2.277, over 5505041.27 frames. , ppl: 9.748398562990769], batch size: 70
+2022-12-12 23:47:51,640 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:47:52,385 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.680259967903076
+2022-12-12 23:49:30,428 INFO [train.py:421] (2/8) Epoch 8, batch 35200, loss[loss=2.201, over 5180.00 frames. , ppl: 9.029918398879369] tot_loss[loss=2.276, over 5518472.63 frames. , ppl: 9.735102433735214], batch size: 70
+2022-12-12 23:51:13,504 INFO [train.py:421] (2/8) Epoch 8, batch 35400, loss[loss=2.216, over 5530.00 frames. , ppl: 9.173421721474819] tot_loss[loss=2.277, over 5480804.33 frames. , ppl: 9.750141936226228], batch size: 70
+2022-12-12 23:52:48,344 INFO [train.py:421] (2/8) Epoch 8, batch 35600, loss[loss=2.251, over 2660.00 frames. , ppl: 9.492984759725736] tot_loss[loss=2.278, over 5480184.13 frames. , ppl: 9.759905334719818], batch size: 70
+2022-12-12 23:54:26,509 INFO [train.py:421] (2/8) Epoch 8, batch 35800, loss[loss=2.134, over 7000.00 frames. , ppl: 8.445598079557119] tot_loss[loss=2.278, over 5493985.05 frames. , ppl: 9.759837648353075], batch size: 70
+2022-12-12 23:56:06,756 INFO [train.py:421] (2/8) Epoch 8, batch 36000, loss[loss=2.212, over 3920.00 frames. , ppl: 9.135490860516493] tot_loss[loss=2.279, over 5471490.97 frames. , ppl: 9.763544654097624], batch size: 70
+2022-12-12 23:56:06,756 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-12 23:56:07,522 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675932595464811
+2022-12-12 23:57:48,184 INFO [train.py:421] (2/8) Epoch 8, batch 36200, loss[loss=2.409, over 1540.00 frames. , ppl: 11.119074730464302] tot_loss[loss=2.278, over 5477854.68 frames. , ppl: 9.758722123484084], batch size: 70
+2022-12-12 23:59:26,453 INFO [train.py:421] (2/8) Epoch 8, batch 36400, loss[loss=2.318, over 1050.00 frames. , ppl: 10.152919978502684] tot_loss[loss=2.279, over 5449625.24 frames. , ppl: 9.768212019823824], batch size: 70
+2022-12-13 00:01:08,145 INFO [train.py:421] (2/8) Epoch 8, batch 36600, loss[loss=2.139, over 6300.00 frames. , ppl: 8.492505152615118] tot_loss[loss=2.278, over 5464454.64 frames. , ppl: 9.761789087717155], batch size: 70
+2022-12-13 00:02:44,810 INFO [train.py:421] (2/8) Epoch 8, batch 36800, loss[loss=2.277, over 2450.00 frames. , ppl: 9.744214346761256] tot_loss[loss=2.278, over 5466282.90 frames. , ppl: 9.76196018249617], batch size: 70
+2022-12-13 00:04:22,094 INFO [train.py:421] (2/8) Epoch 8, batch 37000, loss[loss=2.558, over 700.00 frames. , ppl: 12.916274068681641] tot_loss[loss=2.279, over 5447250.10 frames. , ppl: 9.762211284712672], batch size: 70
+2022-12-13 00:04:22,094 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:04:22,851 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.678712720778323
+2022-12-13 00:06:04,185 INFO [train.py:421] (2/8) Epoch 8, batch 37200, loss[loss=2.188, over 7840.00 frames. , ppl: 8.916751782321937] tot_loss[loss=2.278, over 5471398.36 frames. , ppl: 9.754419034662089], batch size: 70
+2022-12-13 00:07:45,618 INFO [train.py:421] (2/8) Epoch 8, batch 37400, loss[loss=2.48, over 1190.00 frames. , ppl: 11.935814206694896] tot_loss[loss=2.277, over 5497253.79 frames. , ppl: 9.748546386310325], batch size: 70
+2022-12-13 00:09:28,550 INFO [train.py:421] (2/8) Epoch 8, batch 37600, loss[loss=2.269, over 2660.00 frames. , ppl: 9.664988258298822] tot_loss[loss=2.277, over 5498236.52 frames. , ppl: 9.742737859238535], batch size: 70
+2022-12-13 00:11:11,010 INFO [train.py:421] (2/8) Epoch 8, batch 37800, loss[loss=2.193, over 2030.00 frames. , ppl: 8.963456156739234] tot_loss[loss=2.275, over 5534234.31 frames. , ppl: 9.731913149353451], batch size: 70
+2022-12-13 00:12:52,619 INFO [train.py:421] (2/8) Epoch 8, batch 38000, loss[loss=2.567, over 1120.00 frames. , ppl: 13.027441512837019] tot_loss[loss=2.276, over 5523221.39 frames. , ppl: 9.740353455600676], batch size: 70
+2022-12-13 00:12:52,620 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:12:53,370 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.666419470350528
+2022-12-13 00:14:30,994 INFO [train.py:421] (2/8) Epoch 8, batch 38200, loss[loss=2.46, over 1540.00 frames. , ppl: 11.70817176687979] tot_loss[loss=2.275, over 5532124.37 frames. , ppl: 9.732029359313687], batch size: 70
+2022-12-13 00:16:13,236 INFO [train.py:421] (2/8) Epoch 8, batch 38400, loss[loss=2.191, over 4480.00 frames. , ppl: 8.9432553695719] tot_loss[loss=2.274, over 5562953.19 frames. , ppl: 9.715302843401465], batch size: 70
+2022-12-13 00:17:51,819 INFO [train.py:421] (2/8) Epoch 8, batch 38600, loss[loss=2.191, over 2660.00 frames. , ppl: 8.946074138862633] tot_loss[loss=2.274, over 5551196.25 frames. , ppl: 9.716463299456544], batch size: 70
+2022-12-13 00:19:32,517 INFO [train.py:421] (2/8) Epoch 8, batch 38800, loss[loss=2.202, over 5600.00 frames. , ppl: 9.046096030807707] tot_loss[loss=2.274, over 5507379.81 frames. , ppl: 9.721528118132301], batch size: 70
+2022-12-13 00:21:14,874 INFO [train.py:421] (2/8) Epoch 8, batch 39000, loss[loss=2.196, over 3710.00 frames. , ppl: 8.984893471592473] tot_loss[loss=2.274, over 5500138.16 frames. , ppl: 9.722214734894946], batch size: 70
+2022-12-13 00:21:14,874 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:21:15,629 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.679752787430928
+2022-12-13 00:22:54,860 INFO [train.py:421] (2/8) Epoch 8, batch 39200, loss[loss=2.489, over 1470.00 frames. , ppl: 12.053346609021386] tot_loss[loss=2.276, over 5443208.41 frames. , ppl: 9.74093140767516], batch size: 70
+2022-12-13 00:24:36,587 INFO [train.py:421] (2/8) Epoch 8, batch 39400, loss[loss=2.382, over 1260.00 frames. , ppl: 10.825488141939728] tot_loss[loss=2.277, over 5409825.34 frames. , ppl: 9.747394812661389], batch size: 70
+2022-12-13 00:26:21,910 INFO [train.py:421] (2/8) Epoch 8, batch 39600, loss[loss=2.306, over 2240.00 frames. , ppl: 10.036854144163618] tot_loss[loss=2.278, over 5405776.37 frames. , ppl: 9.753630665263044], batch size: 70
+2022-12-13 00:28:01,276 INFO [train.py:421] (2/8) Epoch 8, batch 39800, loss[loss=2.329, over 2310.00 frames. , ppl: 10.272530336349375] tot_loss[loss=2.277, over 5434763.26 frames. , ppl: 9.744657288244849], batch size: 70
+2022-12-13 00:29:43,006 INFO [train.py:421] (2/8) Epoch 8, batch 40000, loss[loss=3.055, over 560.00 frames. , ppl: 21.22779412175634] tot_loss[loss=2.278, over 5427256.71 frames. , ppl: 9.754456765938322], batch size: 70
+2022-12-13 00:29:43,006 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:29:43,770 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.689682020856184
+2022-12-13 00:31:23,406 INFO [train.py:421] (2/8) Epoch 8, batch 40200, loss[loss=2.288, over 1820.00 frames. , ppl: 9.858619762622238] tot_loss[loss=2.277, over 5460553.38 frames. , ppl: 9.744812650247846], batch size: 70
+2022-12-13 00:33:03,355 INFO [train.py:421] (2/8) Epoch 8, batch 40400, loss[loss=2.528, over 770.00 frames. , ppl: 12.523957088673042] tot_loss[loss=2.277, over 5430163.38 frames. , ppl: 9.747821433552007], batch size: 70
+2022-12-13 00:34:47,610 INFO [train.py:421] (2/8) Epoch 8, batch 40600, loss[loss=2.262, over 2590.00 frames. , ppl: 9.600366974020705] tot_loss[loss=2.276, over 5491673.21 frames. , ppl: 9.733753770521757], batch size: 70
+2022-12-13 00:36:29,478 INFO [train.py:421] (2/8) Epoch 8, batch 40800, loss[loss=2.477, over 1190.00 frames. , ppl: 11.90350019388894] tot_loss[loss=2.276, over 5502416.75 frames. , ppl: 9.73784040031422], batch size: 70
+2022-12-13 00:38:06,221 INFO [train.py:421] (2/8) Epoch 8, batch 41000, loss[loss=2.246, over 2870.00 frames. , ppl: 9.449354336209405] tot_loss[loss=2.277, over 5470635.69 frames. , ppl: 9.750609807413419], batch size: 70
+2022-12-13 00:38:06,221 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:38:06,968 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.67065241409767
+2022-12-13 00:39:49,691 INFO [train.py:421] (2/8) Epoch 8, batch 41200, loss[loss=2.226, over 2240.00 frames. , ppl: 9.25980453405471] tot_loss[loss=2.276, over 5493209.72 frames. , ppl: 9.742006149758843], batch size: 70
+2022-12-13 00:41:29,186 INFO [train.py:421] (2/8) Epoch 8, batch 41400, loss[loss=2.467, over 1050.00 frames. , ppl: 11.78257828311671] tot_loss[loss=2.276, over 5528192.58 frames. , ppl: 9.735327949654414], batch size: 70
+2022-12-13 00:43:12,245 INFO [train.py:421] (2/8) Epoch 8, batch 41600, loss[loss=2.487, over 1330.00 frames. , ppl: 12.021409632324321] tot_loss[loss=2.276, over 5510243.34 frames. , ppl: 9.740781206481909], batch size: 70
+2022-12-13 00:44:46,362 INFO [train.py:421] (2/8) Epoch 8, batch 41800, loss[loss=2.14, over 4410.00 frames. , ppl: 8.50023079917896] tot_loss[loss=2.277, over 5492306.71 frames. , ppl: 9.745205870331652], batch size: 70
+2022-12-13 00:46:26,562 INFO [train.py:421] (2/8) Epoch 8, batch 42000, loss[loss=2.35, over 1960.00 frames. , ppl: 10.484906247555323] tot_loss[loss=2.278, over 5459900.12 frames. , ppl: 9.753020316676986], batch size: 70
+2022-12-13 00:46:26,563 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:46:27,308 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675724941707049
+2022-12-13 00:48:05,872 INFO [train.py:421] (2/8) Epoch 8, batch 42200, loss[loss=2.386, over 2100.00 frames. , ppl: 10.874601868569243] tot_loss[loss=2.278, over 5457922.44 frames. , ppl: 9.75939704171034], batch size: 70
+2022-12-13 00:49:47,952 INFO [train.py:421] (2/8) Epoch 8, batch 42400, loss[loss=2.397, over 2030.00 frames. , ppl: 10.989692007489467] tot_loss[loss=2.278, over 5469475.49 frames. , ppl: 9.758983034348148], batch size: 70
+2022-12-13 00:51:28,027 INFO [train.py:421] (2/8) Epoch 8, batch 42600, loss[loss=2.317, over 2100.00 frames. , ppl: 10.146625930046213] tot_loss[loss=2.276, over 5520471.20 frames. , ppl: 9.739778489868291], batch size: 70
+2022-12-13 00:53:10,397 INFO [train.py:421] (2/8) Epoch 8, batch 42800, loss[loss=2.221, over 3850.00 frames. , ppl: 9.21644835531159] tot_loss[loss=2.277, over 5507868.04 frames. , ppl: 9.74381972324032], batch size: 70
+2022-12-13 00:54:50,418 INFO [train.py:421] (2/8) Epoch 8, batch 43000, loss[loss=2.238, over 2030.00 frames. , ppl: 9.369951663355291] tot_loss[loss=2.276, over 5514062.79 frames. , ppl: 9.741069899967629], batch size: 70
+2022-12-13 00:54:50,419 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 00:54:51,176 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675021816214633
+2022-12-13 00:56:31,414 INFO [train.py:421] (2/8) Epoch 8, batch 43200, loss[loss=2.226, over 3080.00 frames. , ppl: 9.26114655163772] tot_loss[loss=2.275, over 5561126.54 frames. , ppl: 9.73215785267713], batch size: 70
+2022-12-13 00:58:13,097 INFO [train.py:421] (2/8) Epoch 8, batch 43400, loss[loss=2.164, over 4970.00 frames. , ppl: 8.70980148233243] tot_loss[loss=2.276, over 5554082.02 frames. , ppl: 9.733165866954874], batch size: 70
+2022-12-13 00:59:53,240 INFO [train.py:421] (2/8) Epoch 8, batch 43600, loss[loss=2.347, over 2310.00 frames. , ppl: 10.4580094870715] tot_loss[loss=2.275, over 5580583.21 frames. , ppl: 9.724889577586454], batch size: 70
+2022-12-13 01:01:34,004 INFO [train.py:421] (2/8) Epoch 8, batch 43800, loss[loss=2.177, over 4690.00 frames. , ppl: 8.815466517042957] tot_loss[loss=2.276, over 5540040.41 frames. , ppl: 9.73298583254642], batch size: 70
+2022-12-13 01:03:12,892 INFO [train.py:421] (2/8) Epoch 8, batch 44000, loss[loss=2.198, over 4340.00 frames. , ppl: 9.007811117707531] tot_loss[loss=2.276, over 5524154.33 frames. , ppl: 9.737427969132376], batch size: 70
+2022-12-13 01:03:12,893 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:03:13,638 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674165532792536
+2022-12-13 01:04:54,832 INFO [train.py:421] (2/8) Epoch 8, batch 44200, loss[loss=2.519, over 1120.00 frames. , ppl: 12.414964852321846] tot_loss[loss=2.275, over 5550852.23 frames. , ppl: 9.728402947742383], batch size: 70
+2022-12-13 01:06:33,909 INFO [train.py:421] (2/8) Epoch 8, batch 44400, loss[loss=4.153, over 350.00 frames. , ppl: 63.63795699639972] tot_loss[loss=2.275, over 5542014.47 frames. , ppl: 9.72824203787658], batch size: 70
+2022-12-13 01:08:12,907 INFO [train.py:421] (2/8) Epoch 8, batch 44600, loss[loss=2.198, over 5810.00 frames. , ppl: 9.006151738339613] tot_loss[loss=2.276, over 5510914.36 frames. , ppl: 9.737918881920027], batch size: 70
+2022-12-13 01:09:58,344 INFO [train.py:421] (2/8) Epoch 8, batch 44800, loss[loss=2.642, over 910.00 frames. , ppl: 14.043589623206445] tot_loss[loss=2.276, over 5526302.23 frames. , ppl: 9.73998760095391], batch size: 70
+2022-12-13 01:11:44,374 INFO [train.py:421] (2/8) Epoch 8, batch 45000, loss[loss=2.478, over 700.00 frames. , ppl: 11.917611086584145] tot_loss[loss=2.275, over 5553539.18 frames. , ppl: 9.731400254804557], batch size: 70
+2022-12-13 01:11:44,374 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:11:45,133 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674410382020383
+2022-12-13 01:13:24,885 INFO [train.py:421] (2/8) Epoch 8, batch 45200, loss[loss=2.344, over 1190.00 frames. , ppl: 10.426248285482759] tot_loss[loss=2.276, over 5517127.94 frames. , ppl: 9.739307302108177], batch size: 70
+2022-12-13 01:15:00,536 INFO [train.py:421] (2/8) Epoch 8, batch 45400, loss[loss=2.232, over 4410.00 frames. , ppl: 9.316072739458955] tot_loss[loss=2.275, over 5555885.26 frames. , ppl: 9.728359439386523], batch size: 70
+2022-12-13 01:16:39,298 INFO [train.py:421] (2/8) Epoch 8, batch 45600, loss[loss=2.203, over 5390.00 frames. , ppl: 9.05183057436572] tot_loss[loss=2.276, over 5530084.03 frames. , ppl: 9.732870398730558], batch size: 70
+2022-12-13 01:18:18,206 INFO [train.py:421] (2/8) Epoch 8, batch 45800, loss[loss=2.223, over 10430.00 frames. , ppl: 9.237118769828621] tot_loss[loss=2.275, over 5547344.75 frames. , ppl: 9.732406812705742], batch size: 70
+2022-12-13 01:19:57,235 INFO [train.py:421] (2/8) Epoch 8, batch 46000, loss[loss=2.259, over 4410.00 frames. , ppl: 9.576668978064243] tot_loss[loss=2.275, over 5606601.77 frames. , ppl: 9.727648038385349], batch size: 70
+2022-12-13 01:19:57,236 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:19:57,983 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.68747512264899
+2022-12-13 01:21:36,866 INFO [train.py:421] (2/8) Epoch 8, batch 46200, loss[loss=2.327, over 3570.00 frames. , ppl: 10.243206931248798] tot_loss[loss=2.277, over 5548313.21 frames. , ppl: 9.747796921680541], batch size: 70
+2022-12-13 01:23:16,619 INFO [train.py:421] (2/8) Epoch 8, batch 46400, loss[loss=2.367, over 1330.00 frames. , ppl: 10.664327147592857] tot_loss[loss=2.278, over 5542513.08 frames. , ppl: 9.753381716996293], batch size: 70
+2022-12-13 01:24:55,148 INFO [train.py:421] (2/8) Epoch 8, batch 46600, loss[loss=3.029, over 560.00 frames. , ppl: 20.68076444472833] tot_loss[loss=2.278, over 5550812.63 frames. , ppl: 9.756743546367648], batch size: 70
+2022-12-13 01:26:38,479 INFO [train.py:421] (2/8) Epoch 8, batch 46800, loss[loss=2.187, over 4410.00 frames. , ppl: 8.911377443454418] tot_loss[loss=2.278, over 5543101.84 frames. , ppl: 9.761535180396256], batch size: 70
+2022-12-13 01:28:17,318 INFO [train.py:421] (2/8) Epoch 8, batch 47000, loss[loss=2.148, over 9450.00 frames. , ppl: 8.563476153310747] tot_loss[loss=2.278, over 5569457.95 frames. , ppl: 9.757218489671677], batch size: 70
+2022-12-13 01:28:17,318 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:28:18,064 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.67573496626884
+2022-12-13 01:29:59,144 INFO [train.py:421] (2/8) Epoch 8, batch 47200, loss[loss=2.241, over 5040.00 frames. , ppl: 9.407227996678127] tot_loss[loss=2.278, over 5563358.45 frames. , ppl: 9.761068179889454], batch size: 70
+2022-12-13 01:31:41,124 INFO [train.py:421] (2/8) Epoch 8, batch 47400, loss[loss=2.167, over 10640.00 frames. , ppl: 8.733022277170932] tot_loss[loss=2.279, over 5553154.71 frames. , ppl: 9.762484310796529], batch size: 70
+2022-12-13 01:33:21,451 INFO [train.py:421] (2/8) Epoch 8, batch 47600, loss[loss=2.206, over 3710.00 frames. , ppl: 9.076289922171695] tot_loss[loss=2.278, over 5588429.73 frames. , ppl: 9.758145543231219], batch size: 70
+2022-12-13 01:35:04,446 INFO [train.py:421] (2/8) Epoch 8, batch 47800, loss[loss=2.186, over 3430.00 frames. , ppl: 8.895256161544337] tot_loss[loss=2.277, over 5619540.75 frames. , ppl: 9.745079871115829], batch size: 70
+2022-12-13 01:36:41,911 INFO [train.py:421] (2/8) Epoch 8, batch 48000, loss[loss=2.394, over 1050.00 frames. , ppl: 10.952586230832681] tot_loss[loss=2.277, over 5573121.77 frames. , ppl: 9.751028770467293], batch size: 70
+2022-12-13 01:36:41,911 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:36:42,686 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664405251487315
+2022-12-13 01:38:24,574 INFO [train.py:421] (2/8) Epoch 8, batch 48200, loss[loss=2.171, over 6020.00 frames. , ppl: 8.769963983534495] tot_loss[loss=2.278, over 5556606.93 frames. , ppl: 9.75348127252643], batch size: 70
+2022-12-13 01:40:06,223 INFO [train.py:421] (2/8) Epoch 8, batch 48400, loss[loss=2.574, over 630.00 frames. , ppl: 13.120731819529329] tot_loss[loss=2.278, over 5540903.92 frames. , ppl: 9.756848013160841], batch size: 70
+2022-12-13 01:41:49,536 INFO [train.py:421] (2/8) Epoch 8, batch 48600, loss[loss=2.924, over 630.00 frames. , ppl: 18.609022981596343] tot_loss[loss=2.28, over 5468518.96 frames. , ppl: 9.775300243352055], batch size: 70
+2022-12-13 01:43:30,069 INFO [train.py:421] (2/8) Epoch 8, batch 48800, loss[loss=2.304, over 2100.00 frames. , ppl: 10.012733714456644] tot_loss[loss=2.279, over 5503961.17 frames. , ppl: 9.770729241702085], batch size: 70
+2022-12-13 01:45:08,311 INFO [train.py:421] (2/8) Epoch 8, batch 49000, loss[loss=2.354, over 2380.00 frames. , ppl: 10.529287379919088] tot_loss[loss=2.281, over 5478648.80 frames. , ppl: 9.784108292185376], batch size: 70
+2022-12-13 01:45:08,311 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:45:09,059 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.66599885304498
+2022-12-13 01:46:48,494 INFO [train.py:421] (2/8) Epoch 8, batch 49200, loss[loss=2.199, over 3570.00 frames. , ppl: 9.014752978708863] tot_loss[loss=2.279, over 5496954.81 frames. , ppl: 9.767935822015499], batch size: 70
+2022-12-13 01:48:27,303 INFO [train.py:421] (2/8) Epoch 8, batch 49400, loss[loss=2.37, over 1820.00 frames. , ppl: 10.699489860065734] tot_loss[loss=2.28, over 5465171.43 frames. , ppl: 9.779424897298892], batch size: 70
+2022-12-13 01:50:08,163 INFO [train.py:421] (2/8) Epoch 8, batch 49600, loss[loss=2.225, over 4830.00 frames. , ppl: 9.250212991625869] tot_loss[loss=2.279, over 5470119.98 frames. , ppl: 9.771370118616689], batch size: 70
+2022-12-13 01:51:45,763 INFO [train.py:421] (2/8) Epoch 8, batch 49800, loss[loss=2.277, over 2450.00 frames. , ppl: 9.74294629870207] tot_loss[loss=2.279, over 5478064.62 frames. , ppl: 9.765253568053001], batch size: 70
+2022-12-13 01:53:28,037 INFO [train.py:421] (2/8) Epoch 8, batch 50000, loss[loss=2.359, over 980.00 frames. , ppl: 10.584233977888573] tot_loss[loss=2.28, over 5446552.40 frames. , ppl: 9.774349200309722], batch size: 70
+2022-12-13 01:53:28,038 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 01:53:28,798 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.6543676004614
+2022-12-13 01:55:10,390 INFO [train.py:421] (2/8) Epoch 8, batch 50200, loss[loss=2.691, over 700.00 frames. , ppl: 14.740618755117797] tot_loss[loss=2.279, over 5478319.44 frames. , ppl: 9.765245946154593], batch size: 70
+2022-12-13 01:56:54,467 INFO [train.py:421] (2/8) Epoch 8, batch 50400, loss[loss=2.125, over 10010.00 frames. , ppl: 8.376518545327084] tot_loss[loss=2.279, over 5470338.98 frames. , ppl: 9.766775440038476], batch size: 70
+2022-12-13 01:58:36,242 INFO [train.py:421] (2/8) Epoch 8, batch 50600, loss[loss=2.388, over 3220.00 frames. , ppl: 10.890933542268094] tot_loss[loss=2.279, over 5474259.60 frames. , ppl: 9.77043191284221], batch size: 70
+2022-12-13 02:00:17,742 INFO [train.py:421] (2/8) Epoch 8, batch 50800, loss[loss=2.275, over 2170.00 frames. , ppl: 9.728234222415697] tot_loss[loss=2.28, over 5464694.49 frames. , ppl: 9.77195558290888], batch size: 70
+2022-12-13 02:01:59,743 INFO [train.py:421] (2/8) Epoch 8, batch 51000, loss[loss=2.238, over 3920.00 frames. , ppl: 9.379001654203181] tot_loss[loss=2.279, over 5462622.36 frames. , ppl: 9.771052270121874], batch size: 70
+2022-12-13 02:01:59,744 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:02:00,491 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66433230115276
+2022-12-13 02:03:40,185 INFO [train.py:421] (2/8) Epoch 8, batch 51200, loss[loss=2.359, over 1680.00 frames. , ppl: 10.58036653272013] tot_loss[loss=2.278, over 5488902.13 frames. , ppl: 9.760301841163917], batch size: 70
+2022-12-13 02:05:20,066 INFO [train.py:421] (2/8) Epoch 8, batch 51400, loss[loss=2.245, over 4550.00 frames. , ppl: 9.443795843234723] tot_loss[loss=2.278, over 5492202.08 frames. , ppl: 9.752706673104308], batch size: 70
+2022-12-13 02:07:01,059 INFO [train.py:421] (2/8) Epoch 8, batch 51600, loss[loss=2.909, over 560.00 frames. , ppl: 18.336037912850912] tot_loss[loss=2.278, over 5497798.40 frames. , ppl: 9.754254405854438], batch size: 70
+2022-12-13 02:08:42,081 INFO [train.py:421] (2/8) Epoch 8, batch 51800, loss[loss=2.156, over 5110.00 frames. , ppl: 8.636334492565833] tot_loss[loss=2.278, over 5473442.16 frames. , ppl: 9.757054636612269], batch size: 70
+2022-12-13 02:10:19,818 INFO [train.py:421] (2/8) Epoch 8, batch 52000, loss[loss=2.226, over 7770.00 frames. , ppl: 9.261906843922118] tot_loss[loss=2.278, over 5452643.74 frames. , ppl: 9.75666563230285], batch size: 70
+2022-12-13 02:10:19,819 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:10:20,563 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.669551784931777
+2022-12-13 02:11:58,361 INFO [train.py:421] (2/8) Epoch 8, batch 52200, loss[loss=2.23, over 6440.00 frames. , ppl: 9.304220770006744] tot_loss[loss=2.279, over 5440595.35 frames. , ppl: 9.765715703191939], batch size: 70
+2022-12-13 02:13:38,890 INFO [train.py:421] (2/8) Epoch 8, batch 52400, loss[loss=2.244, over 2520.00 frames. , ppl: 9.427182912103596] tot_loss[loss=2.28, over 5396716.34 frames. , ppl: 9.780233853821585], batch size: 70
+2022-12-13 02:15:21,731 INFO [train.py:421] (2/8) Epoch 8, batch 52600, loss[loss=2.513, over 910.00 frames. , ppl: 12.344891681429358] tot_loss[loss=2.279, over 5439804.50 frames. , ppl: 9.770888903354095], batch size: 70
+2022-12-13 02:17:02,387 INFO [train.py:421] (2/8) Epoch 8, batch 52800, loss[loss=2.287, over 2310.00 frames. , ppl: 9.840529165570885] tot_loss[loss=2.279, over 5427871.85 frames. , ppl: 9.76971927476727], batch size: 70
+2022-12-13 02:18:40,326 INFO [train.py:421] (2/8) Epoch 8, batch 53000, loss[loss=2.106, over 4130.00 frames. , ppl: 8.219342673179698] tot_loss[loss=2.28, over 5411103.70 frames. , ppl: 9.772401532787086], batch size: 70
+2022-12-13 02:18:40,326 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:18:41,091 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.68564430928278
+2022-12-13 02:20:19,613 INFO [train.py:421] (2/8) Epoch 8, batch 53200, loss[loss=2.228, over 3010.00 frames. , ppl: 9.28281258912397] tot_loss[loss=2.28, over 5400151.50 frames. , ppl: 9.77451071215249], batch size: 70
+2022-12-13 02:21:58,982 INFO [train.py:421] (2/8) Epoch 8, batch 53400, loss[loss=2.368, over 1400.00 frames. , ppl: 10.671590323370703] tot_loss[loss=2.281, over 5356806.08 frames. , ppl: 9.789589398599908], batch size: 70
+2022-12-13 02:23:40,432 INFO [train.py:421] (2/8) Epoch 8, batch 53600, loss[loss=2.447, over 1190.00 frames. , ppl: 11.557039799528242] tot_loss[loss=2.282, over 5346761.77 frames. , ppl: 9.793008675981364], batch size: 70
+2022-12-13 02:25:20,304 INFO [train.py:421] (2/8) Epoch 8, batch 53800, loss[loss=2.469, over 700.00 frames. , ppl: 11.811873148396726] tot_loss[loss=2.282, over 5339941.28 frames. , ppl: 9.796187855999573], batch size: 70
+2022-12-13 02:26:59,831 INFO [train.py:421] (2/8) Epoch 8, batch 54000, loss[loss=2.276, over 2870.00 frames. , ppl: 9.739279204589728] tot_loss[loss=2.281, over 5358232.04 frames. , ppl: 9.787786444169688], batch size: 70
+2022-12-13 02:26:59,832 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:27:00,595 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66121883652427
+2022-12-13 02:28:44,769 INFO [train.py:421] (2/8) Epoch 8, batch 54200, loss[loss=2.164, over 5880.00 frames. , ppl: 8.704695016808028] tot_loss[loss=2.28, over 5400487.90 frames. , ppl: 9.775871251188974], batch size: 70
+2022-12-13 02:30:23,555 INFO [train.py:421] (2/8) Epoch 8, batch 54400, loss[loss=2.223, over 3150.00 frames. , ppl: 9.233864077276369] tot_loss[loss=2.281, over 5386180.13 frames. , ppl: 9.78235061413166], batch size: 70
+2022-12-13 02:32:02,940 INFO [train.py:421] (2/8) Epoch 8, batch 54600, loss[loss=2.243, over 8890.00 frames. , ppl: 9.423076671355009] tot_loss[loss=2.282, over 5349765.31 frames. , ppl: 9.79577453662024], batch size: 70
+2022-12-13 02:33:42,229 INFO [train.py:421] (2/8) Epoch 8, batch 54800, loss[loss=2.179, over 3570.00 frames. , ppl: 8.83369944875206] tot_loss[loss=2.282, over 5353320.49 frames. , ppl: 9.792365573220104], batch size: 70
+2022-12-13 02:35:25,413 INFO [train.py:421] (2/8) Epoch 8, batch 55000, loss[loss=2.247, over 3570.00 frames. , ppl: 9.459743636462917] tot_loss[loss=2.28, over 5408030.55 frames. , ppl: 9.772165758378202], batch size: 70
+2022-12-13 02:35:25,413 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:35:26,159 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.668597244433377
+2022-12-13 02:37:08,637 INFO [train.py:421] (2/8) Epoch 8, batch 55200, loss[loss=2.514, over 910.00 frames. , ppl: 12.357049269377736] tot_loss[loss=2.28, over 5409601.21 frames. , ppl: 9.773395232470259], batch size: 70
+2022-12-13 02:38:47,759 INFO [train.py:421] (2/8) Epoch 8, batch 55400, loss[loss=2.481, over 1050.00 frames. , ppl: 11.957017677690937] tot_loss[loss=2.28, over 5419392.06 frames. , ppl: 9.773071885392584], batch size: 70
+2022-12-13 02:40:28,428 INFO [train.py:421] (2/8) Epoch 8, batch 55600, loss[loss=2.581, over 980.00 frames. , ppl: 13.210656258757355] tot_loss[loss=2.28, over 5429526.29 frames. , ppl: 9.775009318581787], batch size: 70
+2022-12-13 02:42:10,518 INFO [train.py:421] (2/8) Epoch 8, batch 55800, loss[loss=2.292, over 2870.00 frames. , ppl: 9.890250776159794] tot_loss[loss=2.279, over 5444270.35 frames. , ppl: 9.767967140329153], batch size: 70
+2022-12-13 02:43:52,136 INFO [train.py:421] (2/8) Epoch 8, batch 56000, loss[loss=2.195, over 5250.00 frames. , ppl: 8.97920932823706] tot_loss[loss=2.279, over 5445336.88 frames. , ppl: 9.767310308594896], batch size: 70
+2022-12-13 02:43:52,136 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:43:52,898 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.676379424179105
+2022-12-13 02:45:35,148 INFO [train.py:421] (2/8) Epoch 8, batch 56200, loss[loss=2.774, over 700.00 frames. , ppl: 16.020412650662426] tot_loss[loss=2.277, over 5514468.85 frames. , ppl: 9.743559915774346], batch size: 70
+2022-12-13 02:47:16,351 INFO [train.py:421] (2/8) Epoch 8, batch 56400, loss[loss=2.318, over 2800.00 frames. , ppl: 10.158756137605579] tot_loss[loss=2.276, over 5516650.96 frames. , ppl: 9.740809327779317], batch size: 70
+2022-12-13 02:48:53,801 INFO [train.py:421] (2/8) Epoch 8, batch 56600, loss[loss=2.195, over 5040.00 frames. , ppl: 8.977281521702634] tot_loss[loss=2.277, over 5476260.27 frames. , ppl: 9.749126208896916], batch size: 70
+2022-12-13 02:50:34,355 INFO [train.py:421] (2/8) Epoch 8, batch 56800, loss[loss=2.211, over 6930.00 frames. , ppl: 9.124088949203058] tot_loss[loss=2.278, over 5441165.65 frames. , ppl: 9.759745771511147], batch size: 70
+2022-12-13 02:52:17,206 INFO [train.py:421] (2/8) Epoch 8, batch 57000, loss[loss=2.932, over 560.00 frames. , ppl: 18.774084983968663] tot_loss[loss=2.279, over 5443250.47 frames. , ppl: 9.76448775530213], batch size: 70
+2022-12-13 02:52:17,207 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 02:52:17,969 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66042525657999
+2022-12-13 02:54:00,656 INFO [train.py:421] (2/8) Epoch 8, batch 57200, loss[loss=2.273, over 2660.00 frames. , ppl: 9.70847791258246] tot_loss[loss=2.278, over 5454766.55 frames. , ppl: 9.762005725392967], batch size: 70
+2022-12-13 02:55:44,458 INFO [train.py:421] (2/8) Epoch 8, batch 57400, loss[loss=2.373, over 1750.00 frames. , ppl: 10.73256572573833] tot_loss[loss=2.278, over 5477051.83 frames. , ppl: 9.754583092311869], batch size: 70
+2022-12-13 02:57:27,591 INFO [train.py:421] (2/8) Epoch 8, batch 57600, loss[loss=2.158, over 5600.00 frames. , ppl: 8.652552401007013] tot_loss[loss=2.276, over 5535905.25 frames. , ppl: 9.735904234092223], batch size: 70
+2022-12-13 02:59:04,483 INFO [train.py:421] (2/8) Epoch 8, batch 57800, loss[loss=3.071, over 560.00 frames. , ppl: 21.56788871794072] tot_loss[loss=2.276, over 5524077.73 frames. , ppl: 9.733735484115257], batch size: 70
+2022-12-13 03:00:43,920 INFO [train.py:421] (2/8) Epoch 8, batch 58000, loss[loss=2.262, over 3150.00 frames. , ppl: 9.603408192830248] tot_loss[loss=2.277, over 5462500.50 frames. , ppl: 9.750891723017181], batch size: 70
+2022-12-13 03:00:43,920 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:00:44,670 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.679271420376821
+2022-12-13 03:02:24,459 INFO [train.py:421] (2/8) Epoch 8, batch 58200, loss[loss=4.125, over 350.00 frames. , ppl: 61.85313810015094] tot_loss[loss=2.279, over 5429214.98 frames. , ppl: 9.76302752433997], batch size: 70
+2022-12-13 03:04:02,670 INFO [train.py:421] (2/8) Epoch 8, batch 58400, loss[loss=2.194, over 5040.00 frames. , ppl: 8.974062359263138] tot_loss[loss=2.279, over 5419242.24 frames. , ppl: 9.767160279947046], batch size: 70
+2022-12-13 03:05:40,979 INFO [train.py:421] (2/8) Epoch 8, batch 58600, loss[loss=2.407, over 1190.00 frames. , ppl: 11.100997208463998] tot_loss[loss=2.278, over 5464912.03 frames. , ppl: 9.75378385708769], batch size: 70
+2022-12-13 03:07:18,316 INFO [train.py:421] (2/8) Epoch 8, batch 58800, loss[loss=2.437, over 1190.00 frames. , ppl: 11.43876500329334] tot_loss[loss=2.277, over 5450045.08 frames. , ppl: 9.748027019506875], batch size: 70
+2022-12-13 03:09:03,624 INFO [train.py:421] (2/8) Epoch 8, batch 59000, loss[loss=2.464, over 1400.00 frames. , ppl: 11.755521356767778] tot_loss[loss=2.277, over 5463999.41 frames. , ppl: 9.747027036052017], batch size: 70
+2022-12-13 03:09:03,624 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:09:04,370 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.666656969839083
+2022-12-13 03:10:48,154 INFO [train.py:421] (2/8) Epoch 8, batch 59200, loss[loss=2.619, over 840.00 frames. , ppl: 13.724849944295665] tot_loss[loss=2.277, over 5483838.92 frames. , ppl: 9.744505656560579], batch size: 70
+2022-12-13 03:12:28,716 INFO [train.py:421] (2/8) Epoch 8, batch 59400, loss[loss=2.589, over 700.00 frames. , ppl: 13.321036566895197] tot_loss[loss=2.277, over 5493407.96 frames. , ppl: 9.743338170249254], batch size: 70
+2022-12-13 03:14:12,152 INFO [train.py:421] (2/8) Epoch 8, batch 59600, loss[loss=2.234, over 4130.00 frames. , ppl: 9.332590654947088] tot_loss[loss=2.276, over 5513356.82 frames. , ppl: 9.735861658808252], batch size: 70
+2022-12-13 03:15:48,387 INFO [train.py:421] (2/8) Epoch 8, batch 59800, loss[loss=2.711, over 700.00 frames. , ppl: 15.048467487306509] tot_loss[loss=2.276, over 5492699.86 frames. , ppl: 9.738511266856454], batch size: 70
+2022-12-13 03:17:30,951 INFO [train.py:421] (2/8) Epoch 8, batch 60000, loss[loss=2.245, over 4200.00 frames. , ppl: 9.439493685892614] tot_loss[loss=2.276, over 5493627.78 frames. , ppl: 9.73734066796009], batch size: 70
+2022-12-13 03:17:30,952 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:17:31,713 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674444747320207
+2022-12-13 03:19:10,630 INFO [train.py:421] (2/8) Epoch 8, batch 60200, loss[loss=2.271, over 3500.00 frames. , ppl: 9.684701088484674] tot_loss[loss=2.276, over 5461408.45 frames. , ppl: 9.738515732505695], batch size: 70
+2022-12-13 03:20:51,023 INFO [train.py:421] (2/8) Epoch 8, batch 60400, loss[loss=2.266, over 3360.00 frames. , ppl: 9.638067500226228] tot_loss[loss=2.276, over 5467857.28 frames. , ppl: 9.73541364027397], batch size: 70
+2022-12-13 03:22:27,997 INFO [train.py:421] (2/8) Epoch 8, batch 60600, loss[loss=2.394, over 1260.00 frames. , ppl: 10.961995858290571] tot_loss[loss=2.275, over 5508961.28 frames. , ppl: 9.72619814464181], batch size: 70
+2022-12-13 03:24:04,404 INFO [train.py:421] (2/8) Epoch 8, batch 60800, loss[loss=2.403, over 2800.00 frames. , ppl: 11.058283191693759] tot_loss[loss=2.276, over 5495843.50 frames. , ppl: 9.736310276735258], batch size: 70
+2022-12-13 03:25:46,091 INFO [train.py:421] (2/8) Epoch 8, batch 61000, loss[loss=2.824, over 630.00 frames. , ppl: 16.838391528797334] tot_loss[loss=2.276, over 5480530.68 frames. , ppl: 9.741515731893209], batch size: 70
+2022-12-13 03:25:46,092 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:25:46,839 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66320950762563
+2022-12-13 03:27:25,695 INFO [train.py:421] (2/8) Epoch 8, batch 61200, loss[loss=2.366, over 2240.00 frames. , ppl: 10.649458365931496] tot_loss[loss=2.274, over 5522558.47 frames. , ppl: 9.722010014803741], batch size: 70
+2022-12-13 03:29:02,363 INFO [train.py:421] (2/8) Epoch 8, batch 61400, loss[loss=2.878, over 770.00 frames. , ppl: 17.785328869224355] tot_loss[loss=2.275, over 5525328.88 frames. , ppl: 9.724314574621776], batch size: 70
+2022-12-13 03:30:39,175 INFO [train.py:421] (2/8) Epoch 8, batch 61600, loss[loss=2.233, over 2940.00 frames. , ppl: 9.325849671090953] tot_loss[loss=2.276, over 5506256.76 frames. , ppl: 9.737609828321519], batch size: 70
+2022-12-13 03:32:18,644 INFO [train.py:421] (2/8) Epoch 8, batch 61800, loss[loss=2.266, over 4830.00 frames. , ppl: 9.64158370098964] tot_loss[loss=2.277, over 5479703.56 frames. , ppl: 9.745735578577303], batch size: 70
+2022-12-13 03:33:59,913 INFO [train.py:421] (2/8) Epoch 8, batch 62000, loss[loss=2.223, over 4620.00 frames. , ppl: 9.233820746796301] tot_loss[loss=2.277, over 5500668.26 frames. , ppl: 9.748119551888287], batch size: 70
+2022-12-13 03:33:59,913 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:34:00,657 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.655050647650823
+2022-12-13 03:35:43,437 INFO [train.py:421] (2/8) Epoch 8, batch 62200, loss[loss=2.292, over 1610.00 frames. , ppl: 9.894856583963874] tot_loss[loss=2.276, over 5507538.42 frames. , ppl: 9.738400975148272], batch size: 70
+2022-12-13 03:37:22,807 INFO [train.py:421] (2/8) Epoch 8, batch 62400, loss[loss=2.195, over 3150.00 frames. , ppl: 8.97655515275275] tot_loss[loss=2.275, over 5531064.67 frames. , ppl: 9.731123448239131], batch size: 70
+2022-12-13 03:39:04,712 INFO [train.py:421] (2/8) Epoch 8, batch 62600, loss[loss=2.394, over 2030.00 frames. , ppl: 10.96037908361025] tot_loss[loss=2.276, over 5502194.51 frames. , ppl: 9.73595005666719], batch size: 70
+2022-12-13 03:40:42,484 INFO [train.py:421] (2/8) Epoch 8, batch 62800, loss[loss=2.307, over 2520.00 frames. , ppl: 10.039598825823697] tot_loss[loss=2.276, over 5512534.87 frames. , ppl: 9.739044164511418], batch size: 70
+2022-12-13 03:42:19,089 INFO [train.py:421] (2/8) Epoch 8, batch 63000, loss[loss=2.322, over 1820.00 frames. , ppl: 10.192961105789898] tot_loss[loss=2.276, over 5527132.86 frames. , ppl: 9.735731960325532], batch size: 70
+2022-12-13 03:42:19,090 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:42:19,855 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.660179331522313
+2022-12-13 03:44:04,039 INFO [train.py:421] (2/8) Epoch 8, batch 63200, loss[loss=2.24, over 5180.00 frames. , ppl: 9.389144436172922] tot_loss[loss=2.276, over 5518984.14 frames. , ppl: 9.7384582754904], batch size: 70
+2022-12-13 03:45:42,069 INFO [train.py:421] (2/8) Epoch 8, batch 63400, loss[loss=2.914, over 560.00 frames. , ppl: 18.436881189626813] tot_loss[loss=2.277, over 5505650.83 frames. , ppl: 9.743369844106173], batch size: 70
+2022-12-13 03:47:24,376 INFO [train.py:421] (2/8) Epoch 8, batch 63600, loss[loss=2.238, over 3780.00 frames. , ppl: 9.378571656767459] tot_loss[loss=2.276, over 5535547.77 frames. , ppl: 9.736100816124077], batch size: 70
+2022-12-13 03:49:06,228 INFO [train.py:421] (2/8) Epoch 8, batch 63800, loss[loss=2.478, over 1540.00 frames. , ppl: 11.915508066519244] tot_loss[loss=2.276, over 5582595.35 frames. , ppl: 9.733429452366815], batch size: 70
+2022-12-13 03:50:46,485 INFO [train.py:421] (2/8) Epoch 8, batch 64000, loss[loss=2.461, over 910.00 frames. , ppl: 11.711233211175632] tot_loss[loss=2.276, over 5534253.67 frames. , ppl: 9.741751848265865], batch size: 70
+2022-12-13 03:50:46,485 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:50:47,232 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.672815393681503
+2022-12-13 03:52:29,192 INFO [train.py:421] (2/8) Epoch 8, batch 64200, loss[loss=2.247, over 4270.00 frames. , ppl: 9.457280715549702] tot_loss[loss=2.276, over 5561755.87 frames. , ppl: 9.735417934050504], batch size: 70
+2022-12-13 03:54:10,504 INFO [train.py:421] (2/8) Epoch 8, batch 64400, loss[loss=2.126, over 5180.00 frames. , ppl: 8.38371091239143] tot_loss[loss=2.275, over 5577560.78 frames. , ppl: 9.730324891028856], batch size: 70
+2022-12-13 03:55:52,716 INFO [train.py:421] (2/8) Epoch 8, batch 64600, loss[loss=2.393, over 1680.00 frames. , ppl: 10.945403880759791] tot_loss[loss=2.275, over 5566423.83 frames. , ppl: 9.727812804204241], batch size: 70
+2022-12-13 03:57:33,013 INFO [train.py:421] (2/8) Epoch 8, batch 64800, loss[loss=2.387, over 1610.00 frames. , ppl: 10.882094985311044] tot_loss[loss=2.274, over 5594057.84 frames. , ppl: 9.716691726848241], batch size: 70
+2022-12-13 03:59:15,150 INFO [train.py:421] (2/8) Epoch 8, batch 65000, loss[loss=2.448, over 980.00 frames. , ppl: 11.559899294135086] tot_loss[loss=2.273, over 5645181.72 frames. , ppl: 9.70469320761626], batch size: 70
+2022-12-13 03:59:15,150 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 03:59:15,897 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65988194214127
+2022-12-13 04:00:56,732 INFO [train.py:421] (2/8) Epoch 8, batch 65200, loss[loss=2.189, over 3500.00 frames. , ppl: 8.930116212148222] tot_loss[loss=2.273, over 5625393.47 frames. , ppl: 9.707739571022952], batch size: 70
+2022-12-13 04:02:38,086 INFO [train.py:421] (2/8) Epoch 8, batch 65400, loss[loss=2.242, over 3990.00 frames. , ppl: 9.416388116651662] tot_loss[loss=2.274, over 5565798.19 frames. , ppl: 9.71993192473943], batch size: 70
+2022-12-13 04:04:12,168 INFO [train.py:421] (2/8) Epoch 8, batch 65600, loss[loss=2.169, over 6160.00 frames. , ppl: 8.747789567385146] tot_loss[loss=2.273, over 5588726.96 frames. , ppl: 9.709437394476241], batch size: 70
+2022-12-13 04:05:51,082 INFO [train.py:421] (2/8) Epoch 8, batch 65800, loss[loss=3.036, over 560.00 frames. , ppl: 20.823224859276674] tot_loss[loss=2.275, over 5537707.80 frames. , ppl: 9.724763214415203], batch size: 70
+2022-12-13 04:07:30,778 INFO [train.py:421] (2/8) Epoch 8, batch 66000, loss[loss=2.393, over 1050.00 frames. , ppl: 10.951228956200083] tot_loss[loss=2.275, over 5508515.24 frames. , ppl: 9.728888032377741], batch size: 70
+2022-12-13 04:07:30,778 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 04:07:31,525 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.665475252994222
+2022-12-13 04:09:13,155 INFO [train.py:421] (2/8) Epoch 8, batch 66200, loss[loss=2.363, over 3150.00 frames. , ppl: 10.618335917491045] tot_loss[loss=2.275, over 5503358.00 frames. , ppl: 9.725649599511023], batch size: 70
+2022-12-13 04:10:50,316 INFO [train.py:421] (2/8) Epoch 8, batch 66400, loss[loss=2.368, over 1820.00 frames. , ppl: 10.680773186468464] tot_loss[loss=2.275, over 5516985.52 frames. , ppl: 9.729154018152553], batch size: 70
+2022-12-13 04:12:32,866 INFO [train.py:421] (2/8) Epoch 8, batch 66600, loss[loss=2.189, over 3850.00 frames. , ppl: 8.926881035569325] tot_loss[loss=2.276, over 5527039.19 frames. , ppl: 9.735614531315667], batch size: 70
+2022-12-13 04:14:11,395 INFO [train.py:421] (2/8) Epoch 8, batch 66800, loss[loss=2.159, over 6510.00 frames. , ppl: 8.66485937232284] tot_loss[loss=2.275, over 5559327.62 frames. , ppl: 9.728639872539924], batch size: 70
+2022-12-13 04:15:48,896 INFO [train.py:421] (2/8) Epoch 8, batch 67000, loss[loss=2.261, over 2450.00 frames. , ppl: 9.590009303846351] tot_loss[loss=2.276, over 5534188.31 frames. , ppl: 9.73516210000189], batch size: 70
+2022-12-13 04:15:48,896 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 04:15:49,642 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.651940173203029
+2022-12-13 04:17:32,980 INFO [train.py:421] (2/8) Epoch 8, batch 67200, loss[loss=2.198, over 7700.00 frames. , ppl: 9.007485974012877] tot_loss[loss=2.275, over 5555779.11 frames. , ppl: 9.731636923253419], batch size: 70
+2022-12-13 04:19:12,943 INFO [train.py:421] (2/8) Epoch 8, batch 67400, loss[loss=2.506, over 1540.00 frames. , ppl: 12.255367529484925] tot_loss[loss=2.276, over 5535338.46 frames. , ppl: 9.734702296400993], batch size: 70
+2022-12-13 04:20:51,764 INFO [train.py:421] (2/8) Epoch 8, batch 67600, loss[loss=2.357, over 1610.00 frames. , ppl: 10.564357837306426] tot_loss[loss=2.276, over 5522330.10 frames. , ppl: 9.734955984875395], batch size: 70
+2022-12-13 04:22:29,390 INFO [train.py:421] (2/8) Epoch 8, batch 67800, loss[loss=2.288, over 2660.00 frames. , ppl: 9.85449870396627] tot_loss[loss=2.275, over 5527549.09 frames. , ppl: 9.730549006442807], batch size: 70
+2022-12-13 04:24:11,409 INFO [train.py:421] (2/8) Epoch 8, batch 68000, loss[loss=2.382, over 2100.00 frames. , ppl: 10.830727467623817] tot_loss[loss=2.276, over 5511784.72 frames. , ppl: 9.733999854566568], batch size: 70
+2022-12-13 04:24:11,409 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 04:24:12,154 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674181283141165
+2022-12-13 04:25:51,982 INFO [train.py:421] (2/8) Epoch 8, batch 68200, loss[loss=2.482, over 1120.00 frames. , ppl: 11.961646016640186] tot_loss[loss=2.277, over 5488639.04 frames. , ppl: 9.743641973029174], batch size: 70
+2022-12-13 04:27:29,339 INFO [train.py:421] (2/8) Epoch 8, batch 68400, loss[loss=2.449, over 1190.00 frames. , ppl: 11.57276577195732] tot_loss[loss=2.278, over 5430760.73 frames. , ppl: 9.759622574075452], batch size: 70
+2022-12-13 04:29:12,424 INFO [train.py:421] (2/8) Epoch 8, batch 68600, loss[loss=2.788, over 630.00 frames. , ppl: 16.253400953749754] tot_loss[loss=2.278, over 5455649.71 frames. , ppl: 9.75383291294352], batch size: 70
+2022-12-13 04:30:48,901 INFO [train.py:421] (2/8) Epoch 8, batch 68800, loss[loss=2.235, over 2450.00 frames. , ppl: 9.349119863581942] tot_loss[loss=2.278, over 5446615.05 frames. , ppl: 9.75535059637173], batch size: 70
+2022-12-13 04:32:30,672 INFO [train.py:421] (2/8) Epoch 8, batch 69000, loss[loss=2.385, over 1120.00 frames. , ppl: 10.855192692693166] tot_loss[loss=2.279, over 5409131.77 frames. , ppl: 9.767071738389774], batch size: 70
+2022-12-13 04:32:30,673 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 04:32:31,419 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.265, over 211138.00 frames. , ppl: 9.629115969513046
+2022-12-13 04:34:10,754 INFO [train.py:421] (2/8) Epoch 8, batch 69200, loss[loss=2.309, over 1260.00 frames. , ppl: 10.067330300350159] tot_loss[loss=2.279, over 5389990.57 frames. , ppl: 9.771059965759772], batch size: 70
+2022-12-13 04:35:51,600 INFO [train.py:421] (2/8) Epoch 8, batch 69400, loss[loss=2.353, over 2170.00 frames. , ppl: 10.514602201982445] tot_loss[loss=2.279, over 5391433.60 frames. , ppl: 9.764236593522165], batch size: 70
+2022-12-13 04:37:25,625 INFO [train.py:421] (2/8) Epoch 8, batch 69600, loss[loss=2.18, over 4830.00 frames. , ppl: 8.850407411356898] tot_loss[loss=2.279, over 5357124.13 frames. , ppl: 9.766227877853563], batch size: 70
+2022-12-13 04:39:03,669 INFO [train.py:421] (2/8) Epoch 8, batch 69800, loss[loss=2.239, over 3920.00 frames. , ppl: 9.37980078008366] tot_loss[loss=2.28, over 5319772.26 frames. , ppl: 9.778507136584409], batch size: 70
+2022-12-13 04:40:44,763 INFO [train.py:421] (2/8) Epoch 8, batch 70000, loss[loss=2.349, over 2870.00 frames. , ppl: 10.479841275585622] tot_loss[loss=2.281, over 5327287.72 frames. , ppl: 9.788378179594897], batch size: 70
+2022-12-13 04:40:44,764 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 04:40:45,525 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.655207841082497
+2022-12-13 04:42:28,641 INFO [train.py:421] (2/8) Epoch 8, batch 70200, loss[loss=2.165, over 5600.00 frames. , ppl: 8.716764726788343] tot_loss[loss=2.281, over 5352753.26 frames. , ppl: 9.782044663391535], batch size: 70
+2022-12-13 04:44:05,957 INFO [train.py:421] (2/8) Epoch 8, batch 70400, loss[loss=2.313, over 4270.00 frames. , ppl: 10.109237153401535] tot_loss[loss=2.279, over 5385197.13 frames. , ppl: 9.770190840213315], batch size: 70
+2022-12-13 04:45:49,700 INFO [train.py:421] (2/8) Epoch 8, batch 70600, loss[loss=2.317, over 2730.00 frames. , ppl: 10.1457724649521] tot_loss[loss=2.28, over 5371634.62 frames. , ppl: 9.777545647584589], batch size: 70
+2022-12-13 04:47:27,337 INFO [train.py:421] (2/8) Epoch 8, batch 70800, loss[loss=2.351, over 2030.00 frames. , ppl: 10.500619927188831] tot_loss[loss=2.279, over 5391390.44 frames. , ppl: 9.770329001263185], batch size: 70
+2022-12-13 04:49:05,280 INFO [train.py:421] (2/8) Epoch 8, batch 71000, loss[loss=2.234, over 3850.00 frames. , ppl: 9.340703727578182] tot_loss[loss=2.279, over 5393980.25 frames. , ppl: 9.769100913348762], batch size: 70
+2022-12-13 04:49:05,281 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 04:49:06,043 INFO [train.py:452] (2/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664858700229336
+2022-12-13 04:50:47,497 INFO [train.py:421] (2/8) Epoch 8, batch 71200, loss[loss=2.312, over 3080.00 frames. , ppl: 10.093575783247001] tot_loss[loss=2.279, over 5427023.18 frames. , ppl: 9.768299572500487], batch size: 70
+2022-12-13 04:52:24,900 INFO [train.py:421] (2/8) Epoch 8, batch 71400, loss[loss=2.387, over 980.00 frames. , ppl: 10.881851706944532] tot_loss[loss=2.28, over 5410104.85 frames. , ppl: 9.772506256164101], batch size: 70
+2022-12-13 04:54:03,433 INFO [train.py:421] (2/8) Epoch 8, batch 71600, loss[loss=2.369, over 1820.00 frames. , ppl: 10.684157885475926] tot_loss[loss=2.28, over 5409011.82 frames. , ppl: 9.774067584461655], batch size: 70
+2022-12-13 04:55:49,181 INFO [train.py:421] (2/8) Epoch 8, batch 71800, loss[loss=2.264, over 3010.00 frames. , ppl: 9.626146340772127] tot_loss[loss=2.279, over 5448416.34 frames. , ppl: 9.767447732049153], batch size: 70
+2022-12-13 04:57:06,855 INFO [train.py:421] (2/8) Epoch 9, batch 0, loss[loss=2.392, over 2310.00 frames. , ppl: 10.940124054148965] tot_loss[loss=2.392, over 2310.00 frames. , ppl: 10.940124054148965], batch size: 70
+2022-12-13 04:58:47,671 INFO [train.py:421] (2/8) Epoch 9, batch 200, loss[loss=2.539, over 910.00 frames. , ppl: 12.66086490502031] tot_loss[loss=2.285, over 501187.66 frames. , ppl: 9.829534128245001], batch size: 70
+2022-12-13 05:00:29,924 INFO [train.py:421] (2/8) Epoch 9, batch 400, loss[loss=2.344, over 1330.00 frames. , ppl: 10.420392308143038] tot_loss[loss=2.279, over 954572.75 frames. , ppl: 9.764860238317398], batch size: 70
+2022-12-13 05:02:09,652 INFO [train.py:421] (2/8) Epoch 9, batch 600, loss[loss=2.316, over 1960.00 frames. , ppl: 10.138807384200714] tot_loss[loss=2.278, over 1354656.67 frames. , ppl: 9.758840087130327], batch size: 70
+2022-12-13 05:03:48,157 INFO [train.py:421] (2/8) Epoch 9, batch 800, loss[loss=2.353, over 1400.00 frames. , ppl: 10.513660719539866] tot_loss[loss=2.278, over 1743563.56 frames. , ppl: 9.753216425050335], batch size: 70
+2022-12-13 05:05:26,178 INFO [train.py:421] (2/8) Epoch 9, batch 1000, loss[loss=4.103, over 350.00 frames. , ppl: 60.494892653464866] tot_loss[loss=2.276, over 2085273.05 frames. , ppl: 9.735088495525241], batch size: 70
+2022-12-13 05:05:26,179 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:05:26,939 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664010468014997
+2022-12-13 05:07:07,627 INFO [train.py:421] (2/8) Epoch 9, batch 1200, loss[loss=2.42, over 1120.00 frames. , ppl: 11.249778810245672] tot_loss[loss=2.274, over 2412865.90 frames. , ppl: 9.714957885100937], batch size: 70
+2022-12-13 05:08:50,641 INFO [train.py:421] (2/8) Epoch 9, batch 1400, loss[loss=2.374, over 1540.00 frames. , ppl: 10.744674728163394] tot_loss[loss=2.272, over 2737017.53 frames. , ppl: 9.701141845503017], batch size: 70
+2022-12-13 05:10:29,015 INFO [train.py:421] (2/8) Epoch 9, batch 1600, loss[loss=2.286, over 3290.00 frames. , ppl: 9.835100636458955] tot_loss[loss=2.269, over 3027447.89 frames. , ppl: 9.671818404376214], batch size: 70
+2022-12-13 05:12:07,731 INFO [train.py:421] (2/8) Epoch 9, batch 1800, loss[loss=2.293, over 3010.00 frames. , ppl: 9.903005781715619] tot_loss[loss=2.271, over 3219429.88 frames. , ppl: 9.69175826291933], batch size: 70
+2022-12-13 05:13:46,774 INFO [train.py:421] (2/8) Epoch 9, batch 2000, loss[loss=2.149, over 7000.00 frames. , ppl: 8.577280528013798] tot_loss[loss=2.27, over 3444587.69 frames. , ppl: 9.67999048672692], batch size: 70
+2022-12-13 05:13:46,775 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:13:47,536 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.667164894955036
+2022-12-13 05:15:30,256 INFO [train.py:421] (2/8) Epoch 9, batch 2200, loss[loss=2.202, over 4480.00 frames. , ppl: 9.044163782605004] tot_loss[loss=2.269, over 3675722.56 frames. , ppl: 9.66756594377213], batch size: 70
+2022-12-13 05:17:11,207 INFO [train.py:421] (2/8) Epoch 9, batch 2400, loss[loss=2.248, over 3290.00 frames. , ppl: 9.465547471042502] tot_loss[loss=2.267, over 3902298.84 frames. , ppl: 9.653809567171503], batch size: 70
+2022-12-13 05:18:52,165 INFO [train.py:421] (2/8) Epoch 9, batch 2600, loss[loss=2.515, over 840.00 frames. , ppl: 12.364615554452577] tot_loss[loss=2.266, over 4081003.52 frames. , ppl: 9.645291433457299], batch size: 70
+2022-12-13 05:20:30,384 INFO [train.py:421] (2/8) Epoch 9, batch 2800, loss[loss=2.255, over 3010.00 frames. , ppl: 9.534673986626936] tot_loss[loss=2.267, over 4198642.79 frames. , ppl: 9.65017438565499], batch size: 70
+2022-12-13 05:22:11,118 INFO [train.py:421] (2/8) Epoch 9, batch 3000, loss[loss=2.199, over 2800.00 frames. , ppl: 9.01622160708023] tot_loss[loss=2.267, over 4331856.89 frames. , ppl: 9.64866263189467], batch size: 70
+2022-12-13 05:22:11,118 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:22:11,878 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65743883465414
+2022-12-13 05:23:52,011 INFO [train.py:421] (2/8) Epoch 9, batch 3200, loss[loss=2.177, over 3150.00 frames. , ppl: 8.821537552937015] tot_loss[loss=2.267, over 4455083.34 frames. , ppl: 9.649558626013864], batch size: 70
+2022-12-13 05:25:30,729 INFO [train.py:421] (2/8) Epoch 9, batch 3400, loss[loss=2.129, over 8820.00 frames. , ppl: 8.40718337750383] tot_loss[loss=2.267, over 4562674.51 frames. , ppl: 9.64854607440218], batch size: 70
+2022-12-13 05:27:10,001 INFO [train.py:421] (2/8) Epoch 9, batch 3600, loss[loss=2.866, over 700.00 frames. , ppl: 17.561445721131413] tot_loss[loss=2.267, over 4657346.51 frames. , ppl: 9.64913620205139], batch size: 70
+2022-12-13 05:28:50,484 INFO [train.py:421] (2/8) Epoch 9, batch 3800, loss[loss=2.223, over 6300.00 frames. , ppl: 9.234639896437757] tot_loss[loss=2.268, over 4730343.14 frames. , ppl: 9.657424238668295], batch size: 70
+2022-12-13 05:30:30,350 INFO [train.py:421] (2/8) Epoch 9, batch 4000, loss[loss=2.55, over 910.00 frames. , ppl: 12.808491990494074] tot_loss[loss=2.268, over 4800819.76 frames. , ppl: 9.657379766949925], batch size: 70
+2022-12-13 05:30:30,351 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:30:31,100 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.671714518344173
+2022-12-13 05:32:07,786 INFO [train.py:421] (2/8) Epoch 9, batch 4200, loss[loss=2.169, over 6230.00 frames. , ppl: 8.753479746967587] tot_loss[loss=2.268, over 4838571.46 frames. , ppl: 9.660162821651905], batch size: 70
+2022-12-13 05:33:45,392 INFO [train.py:421] (2/8) Epoch 9, batch 4400, loss[loss=2.461, over 1470.00 frames. , ppl: 11.712363924164721] tot_loss[loss=2.269, over 4904509.74 frames. , ppl: 9.670483473613803], batch size: 70
+2022-12-13 05:35:25,495 INFO [train.py:421] (2/8) Epoch 9, batch 4600, loss[loss=2.214, over 5880.00 frames. , ppl: 9.153535759079206] tot_loss[loss=2.269, over 4968992.44 frames. , ppl: 9.672556493727493], batch size: 70
+2022-12-13 05:37:03,948 INFO [train.py:421] (2/8) Epoch 9, batch 4800, loss[loss=2.393, over 1960.00 frames. , ppl: 10.94721232170748] tot_loss[loss=2.27, over 5022673.95 frames. , ppl: 9.675354398391814], batch size: 70
+2022-12-13 05:38:46,306 INFO [train.py:421] (2/8) Epoch 9, batch 5000, loss[loss=2.28, over 3780.00 frames. , ppl: 9.773014639723689] tot_loss[loss=2.269, over 5079825.80 frames. , ppl: 9.674366355217089], batch size: 70
+2022-12-13 05:38:46,307 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:38:47,053 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.655562250208732
+2022-12-13 05:40:30,491 INFO [train.py:421] (2/8) Epoch 9, batch 5200, loss[loss=2.348, over 1820.00 frames. , ppl: 10.468423630583281] tot_loss[loss=2.27, over 5108909.63 frames. , ppl: 9.676252503768795], batch size: 70
+2022-12-13 05:42:13,825 INFO [train.py:421] (2/8) Epoch 9, batch 5400, loss[loss=2.253, over 1400.00 frames. , ppl: 9.515516275283161] tot_loss[loss=2.268, over 5202083.20 frames. , ppl: 9.658761547833882], batch size: 70
+2022-12-13 05:43:55,332 INFO [train.py:421] (2/8) Epoch 9, batch 5600, loss[loss=2.142, over 5740.00 frames. , ppl: 8.517989632916397] tot_loss[loss=2.269, over 5206058.91 frames. , ppl: 9.666989568481796], batch size: 70
+2022-12-13 05:45:33,974 INFO [train.py:421] (2/8) Epoch 9, batch 5800, loss[loss=2.627, over 980.00 frames. , ppl: 13.82534783505947] tot_loss[loss=2.269, over 5227053.45 frames. , ppl: 9.6661713495004], batch size: 70
+2022-12-13 05:47:17,485 INFO [train.py:421] (2/8) Epoch 9, batch 6000, loss[loss=2.206, over 4830.00 frames. , ppl: 9.076353856814714] tot_loss[loss=2.27, over 5230094.60 frames. , ppl: 9.678270002297694], batch size: 70
+2022-12-13 05:47:17,486 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:47:18,250 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.66698032189641
+2022-12-13 05:48:59,167 INFO [train.py:421] (2/8) Epoch 9, batch 6200, loss[loss=2.249, over 2030.00 frames. , ppl: 9.48169473288509] tot_loss[loss=2.27, over 5274451.58 frames. , ppl: 9.679331954315874], batch size: 70
+2022-12-13 05:50:38,216 INFO [train.py:421] (2/8) Epoch 9, batch 6400, loss[loss=2.186, over 2730.00 frames. , ppl: 8.89708880691365] tot_loss[loss=2.271, over 5263680.16 frames. , ppl: 9.688064982832664], batch size: 70
+2022-12-13 05:52:22,371 INFO [train.py:421] (2/8) Epoch 9, batch 6600, loss[loss=2.208, over 4900.00 frames. , ppl: 9.0951939223796] tot_loss[loss=2.27, over 5302781.60 frames. , ppl: 9.678092473833908], batch size: 70
+2022-12-13 05:54:01,401 INFO [train.py:421] (2/8) Epoch 9, batch 6800, loss[loss=2.296, over 2520.00 frames. , ppl: 9.931655205598135] tot_loss[loss=2.27, over 5298723.86 frames. , ppl: 9.683282050486778], batch size: 70
+2022-12-13 05:55:37,120 INFO [train.py:421] (2/8) Epoch 9, batch 7000, loss[loss=2.548, over 980.00 frames. , ppl: 12.775954145925391] tot_loss[loss=2.272, over 5265086.72 frames. , ppl: 9.70045883933326], batch size: 70
+2022-12-13 05:55:37,120 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 05:55:37,868 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.271, over 211138.00 frames. , ppl: 9.689801055644402
+2022-12-13 05:57:20,761 INFO [train.py:421] (2/8) Epoch 9, batch 7200, loss[loss=2.295, over 2940.00 frames. , ppl: 9.92494303332073] tot_loss[loss=2.273, over 5259967.17 frames. , ppl: 9.708949508723666], batch size: 70
+2022-12-13 05:59:03,230 INFO [train.py:421] (2/8) Epoch 9, batch 7400, loss[loss=2.387, over 1750.00 frames. , ppl: 10.884479665227612] tot_loss[loss=2.272, over 5311806.54 frames. , ppl: 9.69585632700267], batch size: 70
+2022-12-13 06:00:38,563 INFO [train.py:421] (2/8) Epoch 9, batch 7600, loss[loss=5.09, over 280.00 frames. , ppl: 162.3104904173964] tot_loss[loss=2.273, over 5253678.93 frames. , ppl: 9.711705003348568], batch size: 70
+2022-12-13 06:02:16,074 INFO [train.py:421] (2/8) Epoch 9, batch 7800, loss[loss=2.184, over 3990.00 frames. , ppl: 8.878067947925649] tot_loss[loss=2.273, over 5275911.67 frames. , ppl: 9.703913249109771], batch size: 70
+2022-12-13 06:03:55,032 INFO [train.py:421] (2/8) Epoch 9, batch 8000, loss[loss=2.61, over 1120.00 frames. , ppl: 13.59838329483187] tot_loss[loss=2.272, over 5298000.55 frames. , ppl: 9.700893011787675], batch size: 70
+2022-12-13 06:03:55,033 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:03:55,778 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.27, over 211138.00 frames. , ppl: 9.677170017837494
+2022-12-13 06:05:34,733 INFO [train.py:421] (2/8) Epoch 9, batch 8200, loss[loss=2.398, over 1470.00 frames. , ppl: 10.999774955389054] tot_loss[loss=2.272, over 5320904.20 frames. , ppl: 9.695655925433572], batch size: 70
+2022-12-13 06:07:15,119 INFO [train.py:421] (2/8) Epoch 9, batch 8400, loss[loss=2.398, over 3290.00 frames. , ppl: 10.995750446846737] tot_loss[loss=2.271, over 5354330.50 frames. , ppl: 9.688858531321685], batch size: 70
+2022-12-13 06:08:56,773 INFO [train.py:421] (2/8) Epoch 9, batch 8600, loss[loss=2.311, over 1960.00 frames. , ppl: 10.087160252682917] tot_loss[loss=2.27, over 5421913.68 frames. , ppl: 9.675617465705791], batch size: 70
+2022-12-13 06:10:33,930 INFO [train.py:421] (2/8) Epoch 9, batch 8800, loss[loss=2.227, over 3990.00 frames. , ppl: 9.270824854368241] tot_loss[loss=2.271, over 5417647.24 frames. , ppl: 9.685392983471962], batch size: 70
+2022-12-13 06:12:16,352 INFO [train.py:421] (2/8) Epoch 9, batch 9000, loss[loss=2.35, over 1330.00 frames. , ppl: 10.482080684453459] tot_loss[loss=2.271, over 5412291.62 frames. , ppl: 9.692216926853892], batch size: 70
+2022-12-13 06:12:16,352 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:12:17,100 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.667984785319996
+2022-12-13 06:13:54,674 INFO [train.py:421] (2/8) Epoch 9, batch 9200, loss[loss=2.23, over 3150.00 frames. , ppl: 9.301836916529439] tot_loss[loss=2.272, over 5391784.33 frames. , ppl: 9.695015237189615], batch size: 70
+2022-12-13 06:15:37,227 INFO [train.py:421] (2/8) Epoch 9, batch 9400, loss[loss=3.139, over 490.00 frames. , ppl: 23.07274831031649] tot_loss[loss=2.272, over 5402146.66 frames. , ppl: 9.696982830564433], batch size: 70
+2022-12-13 06:17:15,205 INFO [train.py:421] (2/8) Epoch 9, batch 9600, loss[loss=2.411, over 2030.00 frames. , ppl: 11.14603672866796] tot_loss[loss=2.272, over 5418571.10 frames. , ppl: 9.695917602260456], batch size: 70
+2022-12-13 06:18:55,627 INFO [train.py:421] (2/8) Epoch 9, batch 9800, loss[loss=3.518, over 420.00 frames. , ppl: 33.71242431506843] tot_loss[loss=2.272, over 5416154.88 frames. , ppl: 9.69400392022297], batch size: 70
+2022-12-13 06:20:37,101 INFO [train.py:421] (2/8) Epoch 9, batch 10000, loss[loss=2.277, over 2100.00 frames. , ppl: 9.749317334898803] tot_loss[loss=2.271, over 5426921.33 frames. , ppl: 9.692183931304696], batch size: 70
+2022-12-13 06:20:37,101 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:20:37,859 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.658802552803898
+2022-12-13 06:22:19,988 INFO [train.py:421] (2/8) Epoch 9, batch 10200, loss[loss=2.51, over 840.00 frames. , ppl: 12.307010241006667] tot_loss[loss=2.272, over 5427415.33 frames. , ppl: 9.696145434168793], batch size: 70
+2022-12-13 06:23:58,842 INFO [train.py:421] (2/8) Epoch 9, batch 10400, loss[loss=2.216, over 4060.00 frames. , ppl: 9.173008359799905] tot_loss[loss=2.271, over 5447603.24 frames. , ppl: 9.687575292722592], batch size: 70
+2022-12-13 06:25:43,291 INFO [train.py:421] (2/8) Epoch 9, batch 10600, loss[loss=2.205, over 5810.00 frames. , ppl: 9.066091705147256] tot_loss[loss=2.269, over 5506745.71 frames. , ppl: 9.673536244909048], batch size: 70
+2022-12-13 06:27:26,721 INFO [train.py:421] (2/8) Epoch 9, batch 10800, loss[loss=2.388, over 1400.00 frames. , ppl: 10.895795577649574] tot_loss[loss=2.27, over 5505517.58 frames. , ppl: 9.677056868306078], batch size: 70
+2022-12-13 06:29:05,717 INFO [train.py:421] (2/8) Epoch 9, batch 11000, loss[loss=2.406, over 1190.00 frames. , ppl: 11.085700795724817] tot_loss[loss=2.271, over 5468352.96 frames. , ppl: 9.687926066107222], batch size: 70
+2022-12-13 06:29:05,717 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:29:06,467 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.671750305558916
+2022-12-13 06:30:49,635 INFO [train.py:421] (2/8) Epoch 9, batch 11200, loss[loss=2.326, over 2800.00 frames. , ppl: 10.237207510296264] tot_loss[loss=2.271, over 5455254.08 frames. , ppl: 9.687047154527805], batch size: 70
+2022-12-13 06:32:30,897 INFO [train.py:421] (2/8) Epoch 9, batch 11400, loss[loss=2.276, over 2870.00 frames. , ppl: 9.739816078064482] tot_loss[loss=2.271, over 5448242.80 frames. , ppl: 9.688282384094917], batch size: 70
+2022-12-13 06:34:12,383 INFO [train.py:421] (2/8) Epoch 9, batch 11600, loss[loss=2.269, over 4340.00 frames. , ppl: 9.667567984171699] tot_loss[loss=2.271, over 5454289.37 frames. , ppl: 9.688249964866701], batch size: 70
+2022-12-13 06:35:50,674 INFO [train.py:421] (2/8) Epoch 9, batch 11800, loss[loss=2.119, over 6020.00 frames. , ppl: 8.324213249080536] tot_loss[loss=2.272, over 5425426.77 frames. , ppl: 9.697971778957566], batch size: 70
+2022-12-13 06:37:30,262 INFO [train.py:421] (2/8) Epoch 9, batch 12000, loss[loss=2.371, over 2170.00 frames. , ppl: 10.705118870778753] tot_loss[loss=2.273, over 5417572.31 frames. , ppl: 9.705297635115068], batch size: 70
+2022-12-13 06:37:30,263 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:37:31,009 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.655009206172396
+2022-12-13 06:39:10,542 INFO [train.py:421] (2/8) Epoch 9, batch 12200, loss[loss=2.266, over 840.00 frames. , ppl: 9.64212875686181] tot_loss[loss=2.271, over 5468707.93 frames. , ppl: 9.688921906266756], batch size: 70
+2022-12-13 06:40:53,668 INFO [train.py:421] (2/8) Epoch 9, batch 12400, loss[loss=2.332, over 3290.00 frames. , ppl: 10.294942908701067] tot_loss[loss=2.27, over 5524183.50 frames. , ppl: 9.682927372635795], batch size: 70
+2022-12-13 06:42:37,744 INFO [train.py:421] (2/8) Epoch 9, batch 12600, loss[loss=2.299, over 2450.00 frames. , ppl: 9.961242071252457] tot_loss[loss=2.27, over 5527912.41 frames. , ppl: 9.677881746721921], batch size: 70
+2022-12-13 06:44:16,096 INFO [train.py:421] (2/8) Epoch 9, batch 12800, loss[loss=2.396, over 1750.00 frames. , ppl: 10.981267298348651] tot_loss[loss=2.271, over 5496494.82 frames. , ppl: 9.689232976786933], batch size: 70
+2022-12-13 06:45:55,153 INFO [train.py:421] (2/8) Epoch 9, batch 13000, loss[loss=2.445, over 1120.00 frames. , ppl: 11.529231615565138] tot_loss[loss=2.27, over 5549235.66 frames. , ppl: 9.68222170879108], batch size: 70
+2022-12-13 06:45:55,153 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:45:55,900 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66105725550274
+2022-12-13 06:47:36,599 INFO [train.py:421] (2/8) Epoch 9, batch 13200, loss[loss=2.388, over 1750.00 frames. , ppl: 10.889057332606559] tot_loss[loss=2.271, over 5530821.31 frames. , ppl: 9.684704401519083], batch size: 70
+2022-12-13 06:49:17,402 INFO [train.py:421] (2/8) Epoch 9, batch 13400, loss[loss=2.404, over 1470.00 frames. , ppl: 11.062453570965163] tot_loss[loss=2.27, over 5550355.71 frames. , ppl: 9.681685264632613], batch size: 70
+2022-12-13 06:50:56,823 INFO [train.py:421] (2/8) Epoch 9, batch 13600, loss[loss=2.199, over 5810.00 frames. , ppl: 9.019971044005944] tot_loss[loss=2.271, over 5521785.73 frames. , ppl: 9.69034263344333], batch size: 70
+2022-12-13 06:52:32,854 INFO [train.py:421] (2/8) Epoch 9, batch 13800, loss[loss=2.331, over 1190.00 frames. , ppl: 10.291168581743765] tot_loss[loss=2.27, over 5549250.81 frames. , ppl: 9.68414566024426], batch size: 70
+2022-12-13 06:54:17,171 INFO [train.py:421] (2/8) Epoch 9, batch 14000, loss[loss=2.41, over 980.00 frames. , ppl: 11.138666934584515] tot_loss[loss=2.271, over 5520179.54 frames. , ppl: 9.693631412147765], batch size: 70
+2022-12-13 06:54:17,171 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 06:54:17,916 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675671954910115
+2022-12-13 06:55:57,100 INFO [train.py:421] (2/8) Epoch 9, batch 14200, loss[loss=2.312, over 2380.00 frames. , ppl: 10.094765810326889] tot_loss[loss=2.273, over 5495975.78 frames. , ppl: 9.7046822810394], batch size: 70
+2022-12-13 06:57:38,458 INFO [train.py:421] (2/8) Epoch 9, batch 14400, loss[loss=2.504, over 1120.00 frames. , ppl: 12.234937237170048] tot_loss[loss=2.274, over 5461549.38 frames. , ppl: 9.713521242243292], batch size: 70
+2022-12-13 06:59:18,391 INFO [train.py:421] (2/8) Epoch 9, batch 14600, loss[loss=2.641, over 770.00 frames. , ppl: 14.031986978737752] tot_loss[loss=2.274, over 5457856.13 frames. , ppl: 9.717018236081469], batch size: 70
+2022-12-13 07:00:59,400 INFO [train.py:421] (2/8) Epoch 9, batch 14800, loss[loss=2.319, over 1890.00 frames. , ppl: 10.168409613303474] tot_loss[loss=2.275, over 5420704.14 frames. , ppl: 9.726579480444618], batch size: 70
+2022-12-13 07:02:36,591 INFO [train.py:421] (2/8) Epoch 9, batch 15000, loss[loss=2.174, over 3220.00 frames. , ppl: 8.795425801432714] tot_loss[loss=2.275, over 5430856.87 frames. , ppl: 9.724191677656602], batch size: 70
+2022-12-13 07:02:36,592 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:02:37,339 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.661317502370732
+2022-12-13 07:04:17,436 INFO [train.py:421] (2/8) Epoch 9, batch 15200, loss[loss=2.146, over 6300.00 frames. , ppl: 8.55375696646496] tot_loss[loss=2.274, over 5475035.90 frames. , ppl: 9.71403992030066], batch size: 70
+2022-12-13 07:05:53,691 INFO [train.py:421] (2/8) Epoch 9, batch 15400, loss[loss=2.261, over 3010.00 frames. , ppl: 9.59281822141314] tot_loss[loss=2.274, over 5435753.94 frames. , ppl: 9.71995345789807], batch size: 70
+2022-12-13 07:07:28,823 INFO [train.py:421] (2/8) Epoch 9, batch 15600, loss[loss=2.486, over 1540.00 frames. , ppl: 12.018125303928004] tot_loss[loss=2.275, over 5399304.06 frames. , ppl: 9.728757856993274], batch size: 70
+2022-12-13 07:09:10,482 INFO [train.py:421] (2/8) Epoch 9, batch 15800, loss[loss=2.841, over 630.00 frames. , ppl: 17.127262684396126] tot_loss[loss=2.274, over 5436334.89 frames. , ppl: 9.718363954950929], batch size: 70
+2022-12-13 07:10:50,253 INFO [train.py:421] (2/8) Epoch 9, batch 16000, loss[loss=2.273, over 1750.00 frames. , ppl: 9.708574718406119] tot_loss[loss=2.274, over 5433822.04 frames. , ppl: 9.718747048704165], batch size: 70
+2022-12-13 07:10:50,254 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:10:51,000 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65729732776054
+2022-12-13 07:12:29,286 INFO [train.py:421] (2/8) Epoch 9, batch 16200, loss[loss=2.379, over 1890.00 frames. , ppl: 10.78945043400719] tot_loss[loss=2.275, over 5418638.06 frames. , ppl: 9.729712638882901], batch size: 70
+2022-12-13 07:14:07,469 INFO [train.py:421] (2/8) Epoch 9, batch 16400, loss[loss=2.315, over 2520.00 frames. , ppl: 10.125302736033722] tot_loss[loss=2.276, over 5399538.56 frames. , ppl: 9.737280913085431], batch size: 70
+2022-12-13 07:15:49,881 INFO [train.py:421] (2/8) Epoch 9, batch 16600, loss[loss=2.305, over 2240.00 frames. , ppl: 10.019438585864174] tot_loss[loss=2.276, over 5410201.26 frames. , ppl: 9.739175480012609], batch size: 70
+2022-12-13 07:17:30,241 INFO [train.py:421] (2/8) Epoch 9, batch 16800, loss[loss=2.502, over 980.00 frames. , ppl: 12.202121540539816] tot_loss[loss=2.276, over 5425081.70 frames. , ppl: 9.736689696330405], batch size: 70
+2022-12-13 07:19:11,498 INFO [train.py:421] (2/8) Epoch 9, batch 17000, loss[loss=2.221, over 3290.00 frames. , ppl: 9.213422799876271] tot_loss[loss=2.275, over 5454612.33 frames. , ppl: 9.729436365223512], batch size: 70
+2022-12-13 07:19:11,498 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:19:12,256 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.27, over 211138.00 frames. , ppl: 9.679668259935742
+2022-12-13 07:20:56,016 INFO [train.py:421] (2/8) Epoch 9, batch 17200, loss[loss=2.483, over 1330.00 frames. , ppl: 11.980399766751768] tot_loss[loss=2.274, over 5492174.88 frames. , ppl: 9.719550311832947], batch size: 70
+2022-12-13 07:22:38,913 INFO [train.py:421] (2/8) Epoch 9, batch 17400, loss[loss=2.605, over 980.00 frames. , ppl: 13.527627639091225] tot_loss[loss=2.275, over 5489946.85 frames. , ppl: 9.725382793175788], batch size: 70
+2022-12-13 07:24:16,738 INFO [train.py:421] (2/8) Epoch 9, batch 17600, loss[loss=2.648, over 770.00 frames. , ppl: 14.1304974996788] tot_loss[loss=2.276, over 5438425.56 frames. , ppl: 9.739093554400165], batch size: 70
+2022-12-13 07:25:57,036 INFO [train.py:421] (2/8) Epoch 9, batch 17800, loss[loss=2.251, over 3920.00 frames. , ppl: 9.498754307296926] tot_loss[loss=2.277, over 5453928.22 frames. , ppl: 9.743839535987547], batch size: 70
+2022-12-13 07:27:36,347 INFO [train.py:421] (2/8) Epoch 9, batch 18000, loss[loss=2.25, over 3010.00 frames. , ppl: 9.487110983345296] tot_loss[loss=2.275, over 5498761.02 frames. , ppl: 9.728256833328023], batch size: 70
+2022-12-13 07:27:36,348 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:27:37,108 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635843748058024
+2022-12-13 07:29:17,364 INFO [train.py:421] (2/8) Epoch 9, batch 18200, loss[loss=2.7, over 630.00 frames. , ppl: 14.875326950439032] tot_loss[loss=2.276, over 5453674.58 frames. , ppl: 9.738130067738766], batch size: 70
+2022-12-13 07:30:55,936 INFO [train.py:421] (2/8) Epoch 9, batch 18400, loss[loss=2.152, over 6860.00 frames. , ppl: 8.60591173535763] tot_loss[loss=2.275, over 5498605.28 frames. , ppl: 9.732502315977557], batch size: 70
+2022-12-13 07:32:36,191 INFO [train.py:421] (2/8) Epoch 9, batch 18600, loss[loss=2.205, over 1610.00 frames. , ppl: 9.07169751852175] tot_loss[loss=2.275, over 5499243.54 frames. , ppl: 9.729849301779797], batch size: 70
+2022-12-13 07:34:20,273 INFO [train.py:421] (2/8) Epoch 9, batch 18800, loss[loss=2.469, over 1540.00 frames. , ppl: 11.808415579174802] tot_loss[loss=2.274, over 5498729.72 frames. , ppl: 9.722482810000075], batch size: 70
+2022-12-13 07:36:01,308 INFO [train.py:421] (2/8) Epoch 9, batch 19000, loss[loss=2.348, over 1750.00 frames. , ppl: 10.463396211167183] tot_loss[loss=2.275, over 5468078.73 frames. , ppl: 9.72928245032699], batch size: 70
+2022-12-13 07:36:01,309 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:36:02,070 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.654813433317269
+2022-12-13 07:37:39,895 INFO [train.py:421] (2/8) Epoch 9, batch 19200, loss[loss=2.28, over 2870.00 frames. , ppl: 9.779094538419868] tot_loss[loss=2.276, over 5452040.97 frames. , ppl: 9.738234189081927], batch size: 70
+2022-12-13 07:39:21,549 INFO [train.py:421] (2/8) Epoch 9, batch 19400, loss[loss=2.549, over 1890.00 frames. , ppl: 12.794996308931008] tot_loss[loss=2.277, over 5397724.87 frames. , ppl: 9.749126574010795], batch size: 70
+2022-12-13 07:41:01,749 INFO [train.py:421] (2/8) Epoch 9, batch 19600, loss[loss=2.396, over 1470.00 frames. , ppl: 10.979384139812904] tot_loss[loss=2.276, over 5421307.46 frames. , ppl: 9.737762883976597], batch size: 70
+2022-12-13 07:42:40,799 INFO [train.py:421] (2/8) Epoch 9, batch 19800, loss[loss=2.263, over 2730.00 frames. , ppl: 9.609044515311027] tot_loss[loss=2.275, over 5462307.39 frames. , ppl: 9.726555533306819], batch size: 70
+2022-12-13 07:44:22,224 INFO [train.py:421] (2/8) Epoch 9, batch 20000, loss[loss=2.295, over 1890.00 frames. , ppl: 9.92350174314256] tot_loss[loss=2.275, over 5447080.60 frames. , ppl: 9.72887478040517], batch size: 70
+2022-12-13 07:44:22,224 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:44:22,986 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.671224962542615
+2022-12-13 07:46:03,948 INFO [train.py:421] (2/8) Epoch 9, batch 20200, loss[loss=2.478, over 1120.00 frames. , ppl: 11.921374382032006] tot_loss[loss=2.275, over 5452505.11 frames. , ppl: 9.72909986886356], batch size: 70
+2022-12-13 07:47:42,244 INFO [train.py:421] (2/8) Epoch 9, batch 20400, loss[loss=2.267, over 2380.00 frames. , ppl: 9.647220867260733] tot_loss[loss=2.276, over 5432191.88 frames. , ppl: 9.734076155076089], batch size: 70
+2022-12-13 07:49:23,913 INFO [train.py:421] (2/8) Epoch 9, batch 20600, loss[loss=2.161, over 6790.00 frames. , ppl: 8.681571554626283] tot_loss[loss=2.276, over 5411758.07 frames. , ppl: 9.737801669901591], batch size: 70
+2022-12-13 07:51:06,183 INFO [train.py:421] (2/8) Epoch 9, batch 20800, loss[loss=2.245, over 2450.00 frames. , ppl: 9.441352570223843] tot_loss[loss=2.274, over 5479712.11 frames. , ppl: 9.718946604137741], batch size: 70
+2022-12-13 07:52:48,709 INFO [train.py:421] (2/8) Epoch 9, batch 21000, loss[loss=2.159, over 5460.00 frames. , ppl: 8.659920444635567] tot_loss[loss=2.273, over 5504912.08 frames. , ppl: 9.712387138605981], batch size: 70
+2022-12-13 07:52:48,709 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 07:52:49,455 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.652867352648064
+2022-12-13 07:54:33,301 INFO [train.py:421] (2/8) Epoch 9, batch 21200, loss[loss=3.046, over 560.00 frames. , ppl: 21.02474491327871] tot_loss[loss=2.273, over 5502236.58 frames. , ppl: 9.707969723780629], batch size: 70
+2022-12-13 07:56:15,661 INFO [train.py:421] (2/8) Epoch 9, batch 21400, loss[loss=2.305, over 1540.00 frames. , ppl: 10.025050422350773] tot_loss[loss=2.27, over 5567843.31 frames. , ppl: 9.682329597221674], batch size: 70
+2022-12-13 07:57:54,060 INFO [train.py:421] (2/8) Epoch 9, batch 21600, loss[loss=2.26, over 3780.00 frames. , ppl: 9.58328921266966] tot_loss[loss=2.271, over 5564115.36 frames. , ppl: 9.686995839802798], batch size: 70
+2022-12-13 07:59:33,285 INFO [train.py:421] (2/8) Epoch 9, batch 21800, loss[loss=2.376, over 1820.00 frames. , ppl: 10.765389740397337] tot_loss[loss=2.271, over 5554298.24 frames. , ppl: 9.689563418888254], batch size: 70
+2022-12-13 08:01:12,351 INFO [train.py:421] (2/8) Epoch 9, batch 22000, loss[loss=2.587, over 1050.00 frames. , ppl: 13.289231625436804] tot_loss[loss=2.271, over 5537655.15 frames. , ppl: 9.692469490202411], batch size: 70
+2022-12-13 08:01:12,352 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:01:13,114 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65536646613963
+2022-12-13 08:02:53,702 INFO [train.py:421] (2/8) Epoch 9, batch 22200, loss[loss=2.354, over 2520.00 frames. , ppl: 10.528877431730086] tot_loss[loss=2.272, over 5523650.59 frames. , ppl: 9.697779092046689], batch size: 70
+2022-12-13 08:04:32,686 INFO [train.py:421] (2/8) Epoch 9, batch 22400, loss[loss=2.23, over 3920.00 frames. , ppl: 9.301422180154974] tot_loss[loss=2.272, over 5506988.35 frames. , ppl: 9.701089546158592], batch size: 70
+2022-12-13 08:06:15,197 INFO [train.py:421] (2/8) Epoch 9, batch 22600, loss[loss=2.161, over 5670.00 frames. , ppl: 8.680496532671816] tot_loss[loss=2.272, over 5527821.32 frames. , ppl: 9.699425459239276], batch size: 70
+2022-12-13 08:07:55,314 INFO [train.py:421] (2/8) Epoch 9, batch 22800, loss[loss=2.337, over 1260.00 frames. , ppl: 10.353714845701345] tot_loss[loss=2.272, over 5540945.84 frames. , ppl: 9.69530065301081], batch size: 70
+2022-12-13 08:09:36,093 INFO [train.py:421] (2/8) Epoch 9, batch 23000, loss[loss=2.355, over 1750.00 frames. , ppl: 10.537490830240154] tot_loss[loss=2.271, over 5527436.26 frames. , ppl: 9.691223281858123], batch size: 70
+2022-12-13 08:09:36,094 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:09:36,839 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.639090273243687
+2022-12-13 08:11:16,442 INFO [train.py:421] (2/8) Epoch 9, batch 23200, loss[loss=2.323, over 840.00 frames. , ppl: 10.207231398504062] tot_loss[loss=2.27, over 5574412.87 frames. , ppl: 9.68052699804231], batch size: 70
+2022-12-13 08:12:57,318 INFO [train.py:421] (2/8) Epoch 9, batch 23400, loss[loss=2.358, over 2380.00 frames. , ppl: 10.571198341189003] tot_loss[loss=2.271, over 5567090.38 frames. , ppl: 9.686881275198118], batch size: 70
+2022-12-13 08:14:44,696 INFO [train.py:421] (2/8) Epoch 9, batch 23600, loss[loss=2.128, over 7490.00 frames. , ppl: 8.401782744362412] tot_loss[loss=2.27, over 5572500.52 frames. , ppl: 9.677294677652908], batch size: 70
+2022-12-13 08:16:23,459 INFO [train.py:421] (2/8) Epoch 9, batch 23800, loss[loss=2.353, over 1820.00 frames. , ppl: 10.51216502347446] tot_loss[loss=2.269, over 5584711.72 frames. , ppl: 9.667462871302972], batch size: 70
+2022-12-13 08:18:03,862 INFO [train.py:421] (2/8) Epoch 9, batch 24000, loss[loss=2.367, over 1050.00 frames. , ppl: 10.661781278688293] tot_loss[loss=2.268, over 5608730.29 frames. , ppl: 9.662672931202618], batch size: 70
+2022-12-13 08:18:03,863 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:18:04,611 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.650770254134942
+2022-12-13 08:19:39,226 INFO [train.py:421] (2/8) Epoch 9, batch 24200, loss[loss=2.269, over 6580.00 frames. , ppl: 9.667055172463257] tot_loss[loss=2.27, over 5576239.07 frames. , ppl: 9.675232956921157], batch size: 70
+2022-12-13 08:21:19,434 INFO [train.py:421] (2/8) Epoch 9, batch 24400, loss[loss=2.178, over 6720.00 frames. , ppl: 8.832189987811878] tot_loss[loss=2.268, over 5631670.84 frames. , ppl: 9.66033923149422], batch size: 70
+2022-12-13 08:23:01,199 INFO [train.py:421] (2/8) Epoch 9, batch 24600, loss[loss=2.3, over 2450.00 frames. , ppl: 9.971435639202506] tot_loss[loss=2.269, over 5619723.12 frames. , ppl: 9.666422804796731], batch size: 70
+2022-12-13 08:24:43,348 INFO [train.py:421] (2/8) Epoch 9, batch 24800, loss[loss=2.257, over 4200.00 frames. , ppl: 9.55784564943415] tot_loss[loss=2.27, over 5549126.12 frames. , ppl: 9.677111791076761], batch size: 70
+2022-12-13 08:26:25,431 INFO [train.py:421] (2/8) Epoch 9, batch 25000, loss[loss=2.331, over 1610.00 frames. , ppl: 10.287080558205584] tot_loss[loss=2.271, over 5519836.50 frames. , ppl: 9.685564229430154], batch size: 70
+2022-12-13 08:26:25,432 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:26:26,196 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.640315849779926
+2022-12-13 08:28:06,125 INFO [train.py:421] (2/8) Epoch 9, batch 25200, loss[loss=2.286, over 4200.00 frames. , ppl: 9.833811399574534] tot_loss[loss=2.27, over 5533180.06 frames. , ppl: 9.680958006169682], batch size: 70
+2022-12-13 08:29:44,793 INFO [train.py:421] (2/8) Epoch 9, batch 25400, loss[loss=2.278, over 3360.00 frames. , ppl: 9.757515483671986] tot_loss[loss=2.272, over 5496054.20 frames. , ppl: 9.695478680623905], batch size: 70
+2022-12-13 08:31:25,693 INFO [train.py:421] (2/8) Epoch 9, batch 25600, loss[loss=2.371, over 1540.00 frames. , ppl: 10.711430214376463] tot_loss[loss=2.271, over 5518894.68 frames. , ppl: 9.691303880160781], batch size: 70
+2022-12-13 08:33:08,256 INFO [train.py:421] (2/8) Epoch 9, batch 25800, loss[loss=2.186, over 7630.00 frames. , ppl: 8.895147632718984] tot_loss[loss=2.271, over 5544179.73 frames. , ppl: 9.688402852337626], batch size: 70
+2022-12-13 08:34:47,565 INFO [train.py:421] (2/8) Epoch 9, batch 26000, loss[loss=2.5, over 1750.00 frames. , ppl: 12.185213570907761] tot_loss[loss=2.271, over 5560261.95 frames. , ppl: 9.68855925389985], batch size: 70
+2022-12-13 08:34:47,565 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:34:48,310 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.645219714400634
+2022-12-13 08:36:28,500 INFO [train.py:421] (2/8) Epoch 9, batch 26200, loss[loss=2.205, over 13020.00 frames. , ppl: 9.069048583177638] tot_loss[loss=2.271, over 5571443.96 frames. , ppl: 9.684937536141897], batch size: 70
+2022-12-13 08:38:06,093 INFO [train.py:421] (2/8) Epoch 9, batch 26400, loss[loss=2.373, over 2170.00 frames. , ppl: 10.732769069136143] tot_loss[loss=2.271, over 5540084.87 frames. , ppl: 9.693649290326299], batch size: 70
+2022-12-13 08:39:46,906 INFO [train.py:421] (2/8) Epoch 9, batch 26600, loss[loss=2.177, over 4970.00 frames. , ppl: 8.819234261155998] tot_loss[loss=2.27, over 5569490.72 frames. , ppl: 9.681629605187597], batch size: 70
+2022-12-13 08:41:30,299 INFO [train.py:421] (2/8) Epoch 9, batch 26800, loss[loss=2.288, over 2310.00 frames. , ppl: 9.8568545477102] tot_loss[loss=2.27, over 5566838.46 frames. , ppl: 9.68075902127046], batch size: 70
+2022-12-13 08:43:11,277 INFO [train.py:421] (2/8) Epoch 9, batch 27000, loss[loss=2.262, over 3220.00 frames. , ppl: 9.598267151747573] tot_loss[loss=2.271, over 5511950.41 frames. , ppl: 9.687135928989544], batch size: 70
+2022-12-13 08:43:11,277 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:43:12,039 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.652885925719394
+2022-12-13 08:44:53,174 INFO [train.py:421] (2/8) Epoch 9, batch 27200, loss[loss=2.251, over 4340.00 frames. , ppl: 9.493570113952396] tot_loss[loss=2.27, over 5506516.12 frames. , ppl: 9.68317239309766], batch size: 70
+2022-12-13 08:46:33,230 INFO [train.py:421] (2/8) Epoch 9, batch 27400, loss[loss=2.176, over 7630.00 frames. , ppl: 8.807912316447256] tot_loss[loss=2.271, over 5522519.83 frames. , ppl: 9.686833256776444], batch size: 70
+2022-12-13 08:48:13,489 INFO [train.py:421] (2/8) Epoch 9, batch 27600, loss[loss=2.387, over 2030.00 frames. , ppl: 10.878862626610708] tot_loss[loss=2.27, over 5540235.68 frames. , ppl: 9.681629645181408], batch size: 70
+2022-12-13 08:49:50,163 INFO [train.py:421] (2/8) Epoch 9, batch 27800, loss[loss=2.313, over 1890.00 frames. , ppl: 10.101187014419413] tot_loss[loss=2.27, over 5518725.09 frames. , ppl: 9.684007899815743], batch size: 70
+2022-12-13 08:51:34,721 INFO [train.py:421] (2/8) Epoch 9, batch 28000, loss[loss=2.299, over 2660.00 frames. , ppl: 9.968074597001754] tot_loss[loss=2.268, over 5605435.95 frames. , ppl: 9.662676925267986], batch size: 70
+2022-12-13 08:51:34,722 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:51:35,469 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.631405082310689
+2022-12-13 08:53:13,329 INFO [train.py:421] (2/8) Epoch 9, batch 28200, loss[loss=2.152, over 6090.00 frames. , ppl: 8.602527650875082] tot_loss[loss=2.269, over 5599783.55 frames. , ppl: 9.671318483261272], batch size: 70
+2022-12-13 08:54:53,476 INFO [train.py:421] (2/8) Epoch 9, batch 28400, loss[loss=2.256, over 5530.00 frames. , ppl: 9.543696836039839] tot_loss[loss=2.271, over 5552128.63 frames. , ppl: 9.685950880674834], batch size: 70
+2022-12-13 08:56:32,697 INFO [train.py:421] (2/8) Epoch 9, batch 28600, loss[loss=3.076, over 560.00 frames. , ppl: 21.673902304635675] tot_loss[loss=2.273, over 5487339.58 frames. , ppl: 9.706128141405227], batch size: 70
+2022-12-13 08:58:14,672 INFO [train.py:421] (2/8) Epoch 9, batch 28800, loss[loss=2.494, over 1190.00 frames. , ppl: 12.110525781286755] tot_loss[loss=2.272, over 5534613.13 frames. , ppl: 9.701992180274669], batch size: 70
+2022-12-13 08:59:57,440 INFO [train.py:421] (2/8) Epoch 9, batch 29000, loss[loss=2.338, over 2030.00 frames. , ppl: 10.357876735847093] tot_loss[loss=2.271, over 5553134.15 frames. , ppl: 9.691275099181793], batch size: 70
+2022-12-13 08:59:57,440 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 08:59:58,199 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.650904523356159
+2022-12-13 09:01:42,269 INFO [train.py:421] (2/8) Epoch 9, batch 29200, loss[loss=2.566, over 770.00 frames. , ppl: 13.012657872020075] tot_loss[loss=2.272, over 5558888.72 frames. , ppl: 9.69629833695668], batch size: 70
+2022-12-13 09:03:21,199 INFO [train.py:421] (2/8) Epoch 9, batch 29400, loss[loss=2.382, over 1610.00 frames. , ppl: 10.824410225704] tot_loss[loss=2.271, over 5574721.33 frames. , ppl: 9.693374279943544], batch size: 70
+2022-12-13 09:05:00,146 INFO [train.py:421] (2/8) Epoch 9, batch 29600, loss[loss=2.289, over 2660.00 frames. , ppl: 9.868585895050304] tot_loss[loss=2.272, over 5559953.99 frames. , ppl: 9.700251139761404], batch size: 70
+2022-12-13 09:06:41,422 INFO [train.py:421] (2/8) Epoch 9, batch 29800, loss[loss=2.486, over 1050.00 frames. , ppl: 12.017058655256104] tot_loss[loss=2.274, over 5519647.42 frames. , ppl: 9.719222302017288], batch size: 70
+2022-12-13 09:08:24,042 INFO [train.py:421] (2/8) Epoch 9, batch 30000, loss[loss=2.398, over 1120.00 frames. , ppl: 11.00324106477111] tot_loss[loss=2.274, over 5523501.61 frames. , ppl: 9.719781521959469], batch size: 70
+2022-12-13 09:08:24,043 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:08:24,805 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637924764691014
+2022-12-13 09:10:04,512 INFO [train.py:421] (2/8) Epoch 9, batch 30200, loss[loss=2.215, over 3150.00 frames. , ppl: 9.162425966294933] tot_loss[loss=2.274, over 5528884.41 frames. , ppl: 9.71480749530359], batch size: 70
+2022-12-13 09:11:45,358 INFO [train.py:421] (2/8) Epoch 9, batch 30400, loss[loss=2.452, over 1400.00 frames. , ppl: 11.60710726821463] tot_loss[loss=2.273, over 5547338.45 frames. , ppl: 9.706745119569666], batch size: 70
+2022-12-13 09:13:24,121 INFO [train.py:421] (2/8) Epoch 9, batch 30600, loss[loss=2.257, over 3570.00 frames. , ppl: 9.557709956849466] tot_loss[loss=2.272, over 5550076.08 frames. , ppl: 9.701654422581335], batch size: 70
+2022-12-13 09:15:01,162 INFO [train.py:421] (2/8) Epoch 9, batch 30800, loss[loss=2.336, over 3010.00 frames. , ppl: 10.34080555840567] tot_loss[loss=2.273, over 5520256.67 frames. , ppl: 9.707613772237854], batch size: 70
+2022-12-13 09:16:34,899 INFO [train.py:421] (2/8) Epoch 9, batch 31000, loss[loss=2.205, over 6790.00 frames. , ppl: 9.068909580490212] tot_loss[loss=2.274, over 5460255.92 frames. , ppl: 9.718257939894732], batch size: 70
+2022-12-13 09:16:34,899 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:16:35,648 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66225130295525
+2022-12-13 09:18:16,658 INFO [train.py:421] (2/8) Epoch 9, batch 31200, loss[loss=2.287, over 3080.00 frames. , ppl: 9.846448204648864] tot_loss[loss=2.272, over 5516323.28 frames. , ppl: 9.701570444959524], batch size: 70
+2022-12-13 09:19:54,609 INFO [train.py:421] (2/8) Epoch 9, batch 31400, loss[loss=2.176, over 4690.00 frames. , ppl: 8.80851242623217] tot_loss[loss=2.273, over 5462428.93 frames. , ppl: 9.709241007119093], batch size: 70
+2022-12-13 09:21:34,366 INFO [train.py:421] (2/8) Epoch 9, batch 31600, loss[loss=2.153, over 5110.00 frames. , ppl: 8.61234029025904] tot_loss[loss=2.273, over 5489353.08 frames. , ppl: 9.708303997092074], batch size: 70
+2022-12-13 09:23:15,091 INFO [train.py:421] (2/8) Epoch 9, batch 31800, loss[loss=2.589, over 770.00 frames. , ppl: 13.311830726244526] tot_loss[loss=2.272, over 5502594.98 frames. , ppl: 9.701510207979075], batch size: 70
+2022-12-13 09:24:55,733 INFO [train.py:421] (2/8) Epoch 9, batch 32000, loss[loss=2.169, over 8890.00 frames. , ppl: 8.747277684857947] tot_loss[loss=2.272, over 5506745.89 frames. , ppl: 9.696731958244488], batch size: 70
+2022-12-13 09:24:55,733 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:24:56,478 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664727097772253
+2022-12-13 09:26:38,798 INFO [train.py:421] (2/8) Epoch 9, batch 32200, loss[loss=2.289, over 2800.00 frames. , ppl: 9.865708351486136] tot_loss[loss=2.272, over 5513344.68 frames. , ppl: 9.699869645575454], batch size: 70
+2022-12-13 09:28:15,342 INFO [train.py:421] (2/8) Epoch 9, batch 32400, loss[loss=2.318, over 840.00 frames. , ppl: 10.153293566938931] tot_loss[loss=2.273, over 5468646.48 frames. , ppl: 9.710127077105211], batch size: 70
+2022-12-13 09:29:54,355 INFO [train.py:421] (2/8) Epoch 9, batch 32600, loss[loss=2.293, over 2100.00 frames. , ppl: 9.904071780484834] tot_loss[loss=2.272, over 5495062.39 frames. , ppl: 9.695935947394487], batch size: 70
+2022-12-13 09:31:30,579 INFO [train.py:421] (2/8) Epoch 9, batch 32800, loss[loss=2.261, over 4130.00 frames. , ppl: 9.589133980223702] tot_loss[loss=2.273, over 5461008.12 frames. , ppl: 9.71232098765314], batch size: 70
+2022-12-13 09:33:11,502 INFO [train.py:421] (2/8) Epoch 9, batch 33000, loss[loss=2.457, over 980.00 frames. , ppl: 11.666056531700258] tot_loss[loss=2.274, over 5463907.13 frames. , ppl: 9.713583529699571], batch size: 70
+2022-12-13 09:33:11,503 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:33:12,263 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.644468844695679
+2022-12-13 09:34:49,445 INFO [train.py:421] (2/8) Epoch 9, batch 33200, loss[loss=3.653, over 420.00 frames. , ppl: 38.57555164949308] tot_loss[loss=2.273, over 5460415.17 frames. , ppl: 9.71217067567523], batch size: 70
+2022-12-13 09:36:29,258 INFO [train.py:421] (2/8) Epoch 9, batch 33400, loss[loss=2.254, over 3640.00 frames. , ppl: 9.528688306907153] tot_loss[loss=2.274, over 5457847.91 frames. , ppl: 9.714242255172982], batch size: 70
+2022-12-13 09:38:10,170 INFO [train.py:421] (2/8) Epoch 9, batch 33600, loss[loss=2.236, over 4970.00 frames. , ppl: 9.351467820132937] tot_loss[loss=2.274, over 5464493.71 frames. , ppl: 9.717839008288115], batch size: 70
+2022-12-13 09:39:46,992 INFO [train.py:421] (2/8) Epoch 9, batch 33800, loss[loss=2.326, over 1470.00 frames. , ppl: 10.23324847182934] tot_loss[loss=2.275, over 5424490.43 frames. , ppl: 9.726402830198293], batch size: 70
+2022-12-13 09:41:27,520 INFO [train.py:421] (2/8) Epoch 9, batch 34000, loss[loss=2.476, over 980.00 frames. , ppl: 11.898099091671526] tot_loss[loss=2.276, over 5393947.16 frames. , ppl: 9.736264358254614], batch size: 70
+2022-12-13 09:41:27,520 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:41:28,282 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.6631322756689
+2022-12-13 09:43:09,936 INFO [train.py:421] (2/8) Epoch 9, batch 34200, loss[loss=2.168, over 9030.00 frames. , ppl: 8.737937335866642] tot_loss[loss=2.276, over 5387152.43 frames. , ppl: 9.737167381486431], batch size: 70
+2022-12-13 09:44:49,447 INFO [train.py:421] (2/8) Epoch 9, batch 34400, loss[loss=2.603, over 770.00 frames. , ppl: 13.500175271380472] tot_loss[loss=2.276, over 5360511.47 frames. , ppl: 9.740812846662601], batch size: 70
+2022-12-13 09:46:28,568 INFO [train.py:421] (2/8) Epoch 9, batch 34600, loss[loss=2.473, over 1120.00 frames. , ppl: 11.860940479020528] tot_loss[loss=2.276, over 5381726.83 frames. , ppl: 9.735494213757958], batch size: 70
+2022-12-13 09:48:11,569 INFO [train.py:421] (2/8) Epoch 9, batch 34800, loss[loss=2.25, over 7490.00 frames. , ppl: 9.488210868402103] tot_loss[loss=2.276, over 5399708.88 frames. , ppl: 9.733292919047226], batch size: 70
+2022-12-13 09:49:50,487 INFO [train.py:421] (2/8) Epoch 9, batch 35000, loss[loss=2.549, over 1400.00 frames. , ppl: 12.799097379693603] tot_loss[loss=2.276, over 5368286.67 frames. , ppl: 9.739739076567634], batch size: 70
+2022-12-13 09:49:50,488 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:49:51,240 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.644501676179262
+2022-12-13 09:51:35,402 INFO [train.py:421] (2/8) Epoch 9, batch 35200, loss[loss=2.189, over 4690.00 frames. , ppl: 8.926883029528758] tot_loss[loss=2.277, over 5329495.15 frames. , ppl: 9.747021280562585], batch size: 70
+2022-12-13 09:53:15,453 INFO [train.py:421] (2/8) Epoch 9, batch 35400, loss[loss=2.199, over 2800.00 frames. , ppl: 9.012510160799653] tot_loss[loss=2.276, over 5364454.66 frames. , ppl: 9.737021777533718], batch size: 70
+2022-12-13 09:54:55,419 INFO [train.py:421] (2/8) Epoch 9, batch 35600, loss[loss=2.39, over 980.00 frames. , ppl: 10.911794279586188] tot_loss[loss=2.274, over 5418151.46 frames. , ppl: 9.7194004233398], batch size: 70
+2022-12-13 09:56:35,170 INFO [train.py:421] (2/8) Epoch 9, batch 35800, loss[loss=2.17, over 5040.00 frames. , ppl: 8.75822769965412] tot_loss[loss=2.275, over 5384386.18 frames. , ppl: 9.731164044011107], batch size: 70
+2022-12-13 09:58:16,611 INFO [train.py:421] (2/8) Epoch 9, batch 36000, loss[loss=2.17, over 10640.00 frames. , ppl: 8.756439228238161] tot_loss[loss=2.275, over 5405770.29 frames. , ppl: 9.727080414792118], batch size: 70
+2022-12-13 09:58:16,612 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 09:58:17,362 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.647108567176648
+2022-12-13 09:59:55,817 INFO [train.py:421] (2/8) Epoch 9, batch 36200, loss[loss=2.263, over 2940.00 frames. , ppl: 9.607877867870883] tot_loss[loss=2.275, over 5393575.25 frames. , ppl: 9.727262706406579], batch size: 70
+2022-12-13 10:01:36,713 INFO [train.py:421] (2/8) Epoch 9, batch 36400, loss[loss=2.484, over 770.00 frames. , ppl: 11.988512068549396] tot_loss[loss=2.275, over 5374071.06 frames. , ppl: 9.723989300322442], batch size: 70
+2022-12-13 10:03:20,896 INFO [train.py:421] (2/8) Epoch 9, batch 36600, loss[loss=2.345, over 1120.00 frames. , ppl: 10.431802741068136] tot_loss[loss=2.275, over 5385659.84 frames. , ppl: 9.72671294871583], batch size: 70
+2022-12-13 10:05:04,616 INFO [train.py:421] (2/8) Epoch 9, batch 36800, loss[loss=2.306, over 5110.00 frames. , ppl: 10.034838667607572] tot_loss[loss=2.276, over 5378268.22 frames. , ppl: 9.733064284466183], batch size: 70
+2022-12-13 10:06:43,622 INFO [train.py:421] (2/8) Epoch 9, batch 37000, loss[loss=2.241, over 4340.00 frames. , ppl: 9.405648319454805] tot_loss[loss=2.275, over 5416433.45 frames. , ppl: 9.724876810180078], batch size: 70
+2022-12-13 10:06:43,622 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:06:44,368 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.65039602425002
+2022-12-13 10:08:24,181 INFO [train.py:421] (2/8) Epoch 9, batch 37200, loss[loss=2.267, over 3780.00 frames. , ppl: 9.652899030121805] tot_loss[loss=2.276, over 5391443.64 frames. , ppl: 9.73748332373656], batch size: 70
+2022-12-13 10:10:01,428 INFO [train.py:421] (2/8) Epoch 9, batch 37400, loss[loss=2.261, over 3010.00 frames. , ppl: 9.591302656945686] tot_loss[loss=2.276, over 5400341.67 frames. , ppl: 9.735516761892708], batch size: 70
+2022-12-13 10:11:39,958 INFO [train.py:421] (2/8) Epoch 9, batch 37600, loss[loss=2.584, over 840.00 frames. , ppl: 13.251905090933741] tot_loss[loss=2.276, over 5378431.88 frames. , ppl: 9.740203298830831], batch size: 70
+2022-12-13 10:13:18,222 INFO [train.py:421] (2/8) Epoch 9, batch 37800, loss[loss=2.615, over 840.00 frames. , ppl: 13.663031814953317] tot_loss[loss=2.277, over 5366355.13 frames. , ppl: 9.752089395477388], batch size: 70
+2022-12-13 10:14:59,581 INFO [train.py:421] (2/8) Epoch 9, batch 38000, loss[loss=2.201, over 6790.00 frames. , ppl: 9.03835651471378] tot_loss[loss=2.276, over 5417812.38 frames. , ppl: 9.74069919868796], batch size: 70
+2022-12-13 10:14:59,581 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:15:00,328 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637123113571448
+2022-12-13 10:16:42,062 INFO [train.py:421] (2/8) Epoch 9, batch 38200, loss[loss=2.226, over 2450.00 frames. , ppl: 9.263664724522728] tot_loss[loss=2.274, over 5466275.02 frames. , ppl: 9.718706106965298], batch size: 70
+2022-12-13 10:18:21,502 INFO [train.py:421] (2/8) Epoch 9, batch 38400, loss[loss=2.585, over 1190.00 frames. , ppl: 13.268143330048167] tot_loss[loss=2.273, over 5473822.04 frames. , ppl: 9.713151380531961], batch size: 70
+2022-12-13 10:20:03,170 INFO [train.py:421] (2/8) Epoch 9, batch 38600, loss[loss=2.282, over 1960.00 frames. , ppl: 9.799368949282268] tot_loss[loss=2.274, over 5434979.52 frames. , ppl: 9.721094480506343], batch size: 70
+2022-12-13 10:21:45,083 INFO [train.py:421] (2/8) Epoch 9, batch 38800, loss[loss=2.379, over 2800.00 frames. , ppl: 10.792007388917721] tot_loss[loss=2.274, over 5438816.23 frames. , ppl: 9.721600222831475], batch size: 70
+2022-12-13 10:23:26,884 INFO [train.py:421] (2/8) Epoch 9, batch 39000, loss[loss=2.837, over 630.00 frames. , ppl: 17.063623889065653] tot_loss[loss=2.274, over 5430394.79 frames. , ppl: 9.718672780183319], batch size: 70
+2022-12-13 10:23:26,885 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:23:27,632 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.646513174570288
+2022-12-13 10:25:07,539 INFO [train.py:421] (2/8) Epoch 9, batch 39200, loss[loss=2.278, over 1890.00 frames. , ppl: 9.753990707776039] tot_loss[loss=2.275, over 5410610.29 frames. , ppl: 9.724011679379046], batch size: 70
+2022-12-13 10:26:46,236 INFO [train.py:421] (2/8) Epoch 9, batch 39400, loss[loss=2.206, over 4060.00 frames. , ppl: 9.075573838019155] tot_loss[loss=2.277, over 5357421.95 frames. , ppl: 9.74347880503391], batch size: 70
+2022-12-13 10:28:27,571 INFO [train.py:421] (2/8) Epoch 9, batch 39600, loss[loss=2.373, over 1750.00 frames. , ppl: 10.72952067248491] tot_loss[loss=2.277, over 5332087.32 frames. , ppl: 9.749544570711326], batch size: 70
+2022-12-13 10:30:04,664 INFO [train.py:421] (2/8) Epoch 9, batch 39800, loss[loss=2.382, over 2240.00 frames. , ppl: 10.826957591248853] tot_loss[loss=2.277, over 5347544.38 frames. , ppl: 9.743003552795154], batch size: 70
+2022-12-13 10:31:44,860 INFO [train.py:421] (2/8) Epoch 9, batch 40000, loss[loss=2.319, over 2380.00 frames. , ppl: 10.160456756139062] tot_loss[loss=2.275, over 5359641.40 frames. , ppl: 9.731380288247095], batch size: 70
+2022-12-13 10:31:44,861 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:31:45,622 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.648730734417011
+2022-12-13 10:33:26,258 INFO [train.py:421] (2/8) Epoch 9, batch 40200, loss[loss=2.166, over 3850.00 frames. , ppl: 8.720943884073204] tot_loss[loss=2.275, over 5374731.60 frames. , ppl: 9.727816029894443], batch size: 70
+2022-12-13 10:35:02,355 INFO [train.py:421] (2/8) Epoch 9, batch 40400, loss[loss=2.392, over 2310.00 frames. , ppl: 10.934660999071887] tot_loss[loss=2.275, over 5372751.24 frames. , ppl: 9.726586193148863], batch size: 70
+2022-12-13 10:36:42,721 INFO [train.py:421] (2/8) Epoch 9, batch 40600, loss[loss=2.358, over 2310.00 frames. , ppl: 10.573471016331021] tot_loss[loss=2.276, over 5348169.24 frames. , ppl: 9.736452411585129], batch size: 70
+2022-12-13 10:38:21,180 INFO [train.py:421] (2/8) Epoch 9, batch 40800, loss[loss=2.26, over 1540.00 frames. , ppl: 9.586638148238656] tot_loss[loss=2.276, over 5364436.80 frames. , ppl: 9.73785949536539], batch size: 70
+2022-12-13 10:39:58,028 INFO [train.py:421] (2/8) Epoch 9, batch 41000, loss[loss=2.25, over 3080.00 frames. , ppl: 9.490014848837431] tot_loss[loss=2.276, over 5374455.45 frames. , ppl: 9.733321346668133], batch size: 70
+2022-12-13 10:39:58,028 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:39:58,774 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.663169461348792
+2022-12-13 10:41:39,549 INFO [train.py:421] (2/8) Epoch 9, batch 41200, loss[loss=2.221, over 4060.00 frames. , ppl: 9.213536671882418] tot_loss[loss=2.276, over 5359762.53 frames. , ppl: 9.735827857776236], batch size: 70
+2022-12-13 10:43:16,338 INFO [train.py:421] (2/8) Epoch 9, batch 41400, loss[loss=2.457, over 1260.00 frames. , ppl: 11.66558152539499] tot_loss[loss=2.276, over 5339378.59 frames. , ppl: 9.742457941786565], batch size: 70
+2022-12-13 10:44:58,173 INFO [train.py:421] (2/8) Epoch 9, batch 41600, loss[loss=2.348, over 1820.00 frames. , ppl: 10.467890022400388] tot_loss[loss=2.276, over 5333974.93 frames. , ppl: 9.73982214793203], batch size: 70
+2022-12-13 10:46:39,492 INFO [train.py:421] (2/8) Epoch 9, batch 41800, loss[loss=2.602, over 1400.00 frames. , ppl: 13.493316320076756] tot_loss[loss=2.277, over 5318478.16 frames. , ppl: 9.751525732338335], batch size: 70
+2022-12-13 10:48:20,323 INFO [train.py:421] (2/8) Epoch 9, batch 42000, loss[loss=2.253, over 1680.00 frames. , ppl: 9.515517381532794] tot_loss[loss=2.279, over 5289876.61 frames. , ppl: 9.764716962375607], batch size: 70
+2022-12-13 10:48:20,323 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:48:21,082 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635082199742264
+2022-12-13 10:50:03,535 INFO [train.py:421] (2/8) Epoch 9, batch 42200, loss[loss=2.176, over 5880.00 frames. , ppl: 8.809981882272721] tot_loss[loss=2.276, over 5362517.04 frames. , ppl: 9.742387932403982], batch size: 70
+2022-12-13 10:51:44,293 INFO [train.py:421] (2/8) Epoch 9, batch 42400, loss[loss=2.175, over 7280.00 frames. , ppl: 8.804637888432115] tot_loss[loss=2.275, over 5395464.19 frames. , ppl: 9.72820634987336], batch size: 70
+2022-12-13 10:53:20,409 INFO [train.py:421] (2/8) Epoch 9, batch 42600, loss[loss=2.286, over 2730.00 frames. , ppl: 9.831523575291984] tot_loss[loss=2.274, over 5421076.10 frames. , ppl: 9.719508764746726], batch size: 70
+2022-12-13 10:54:59,730 INFO [train.py:421] (2/8) Epoch 9, batch 42800, loss[loss=2.355, over 1330.00 frames. , ppl: 10.540739885205973] tot_loss[loss=2.274, over 5454231.09 frames. , ppl: 9.714245953270266], batch size: 70
+2022-12-13 10:56:37,392 INFO [train.py:421] (2/8) Epoch 9, batch 43000, loss[loss=2.19, over 3990.00 frames. , ppl: 8.932686709413172] tot_loss[loss=2.275, over 5449877.25 frames. , ppl: 9.723064771962038], batch size: 70
+2022-12-13 10:56:37,393 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 10:56:38,138 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.668268114386317
+2022-12-13 10:58:19,720 INFO [train.py:421] (2/8) Epoch 9, batch 43200, loss[loss=2.22, over 3080.00 frames. , ppl: 9.20860115365255] tot_loss[loss=2.274, over 5452919.58 frames. , ppl: 9.718945719568758], batch size: 70
+2022-12-13 11:00:01,375 INFO [train.py:421] (2/8) Epoch 9, batch 43400, loss[loss=2.403, over 1610.00 frames. , ppl: 11.057818538116539] tot_loss[loss=2.274, over 5465225.92 frames. , ppl: 9.715643595170084], batch size: 70
+2022-12-13 11:01:45,159 INFO [train.py:421] (2/8) Epoch 9, batch 43600, loss[loss=2.29, over 3570.00 frames. , ppl: 9.877827910689906] tot_loss[loss=2.275, over 5434972.05 frames. , ppl: 9.727355647846423], batch size: 70
+2022-12-13 11:03:26,870 INFO [train.py:421] (2/8) Epoch 9, batch 43800, loss[loss=2.464, over 1330.00 frames. , ppl: 11.74796834652623] tot_loss[loss=2.274, over 5466094.54 frames. , ppl: 9.71405401558434], batch size: 70
+2022-12-13 11:05:05,108 INFO [train.py:421] (2/8) Epoch 9, batch 44000, loss[loss=2.265, over 4200.00 frames. , ppl: 9.629328785962906] tot_loss[loss=2.273, over 5497514.08 frames. , ppl: 9.703971257782154], batch size: 70
+2022-12-13 11:05:05,109 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:05:05,859 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.623468208814646
+2022-12-13 11:06:45,137 INFO [train.py:421] (2/8) Epoch 9, batch 44200, loss[loss=2.337, over 1330.00 frames. , ppl: 10.354177328852735] tot_loss[loss=2.273, over 5480079.09 frames. , ppl: 9.711337699557493], batch size: 70
+2022-12-13 11:08:25,117 INFO [train.py:421] (2/8) Epoch 9, batch 44400, loss[loss=2.373, over 2940.00 frames. , ppl: 10.724494810624376] tot_loss[loss=2.274, over 5474526.82 frames. , ppl: 9.713913414428262], batch size: 70
+2022-12-13 11:10:01,989 INFO [train.py:421] (2/8) Epoch 9, batch 44600, loss[loss=2.323, over 2730.00 frames. , ppl: 10.209004135654855] tot_loss[loss=2.274, over 5485616.73 frames. , ppl: 9.715023046236455], batch size: 70
+2022-12-13 11:11:38,149 INFO [train.py:421] (2/8) Epoch 9, batch 44800, loss[loss=2.309, over 2940.00 frames. , ppl: 10.064052593624057] tot_loss[loss=2.275, over 5471679.73 frames. , ppl: 9.724478288240666], batch size: 70
+2022-12-13 11:13:19,789 INFO [train.py:421] (2/8) Epoch 9, batch 45000, loss[loss=2.332, over 2870.00 frames. , ppl: 10.297469712947326] tot_loss[loss=2.275, over 5467817.88 frames. , ppl: 9.72804098889746], batch size: 70
+2022-12-13 11:13:19,790 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:13:20,550 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.648610776092447
+2022-12-13 11:15:01,466 INFO [train.py:421] (2/8) Epoch 9, batch 45200, loss[loss=2.31, over 2380.00 frames. , ppl: 10.078649397899785] tot_loss[loss=2.274, over 5464635.53 frames. , ppl: 9.721237760610336], batch size: 70
+2022-12-13 11:16:44,206 INFO [train.py:421] (2/8) Epoch 9, batch 45400, loss[loss=2.562, over 910.00 frames. , ppl: 12.96757803576046] tot_loss[loss=2.273, over 5511599.40 frames. , ppl: 9.706754783168806], batch size: 70
+2022-12-13 11:18:27,657 INFO [train.py:421] (2/8) Epoch 9, batch 45600, loss[loss=2.352, over 1960.00 frames. , ppl: 10.508954031192648] tot_loss[loss=2.273, over 5504966.79 frames. , ppl: 9.710686054504142], batch size: 70
+2022-12-13 11:20:08,207 INFO [train.py:421] (2/8) Epoch 9, batch 45800, loss[loss=2.26, over 2800.00 frames. , ppl: 9.58076987998398] tot_loss[loss=2.273, over 5509197.02 frames. , ppl: 9.71100831893077], batch size: 70
+2022-12-13 11:21:46,359 INFO [train.py:421] (2/8) Epoch 9, batch 46000, loss[loss=3.224, over 490.00 frames. , ppl: 25.130580951829582] tot_loss[loss=2.273, over 5512980.22 frames. , ppl: 9.707989433723945], batch size: 70
+2022-12-13 11:21:46,359 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:21:47,106 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.636034857644312
+2022-12-13 11:23:26,556 INFO [train.py:421] (2/8) Epoch 9, batch 46200, loss[loss=2.682, over 630.00 frames. , ppl: 14.61719058181522] tot_loss[loss=2.273, over 5522190.23 frames. , ppl: 9.705996755776257], batch size: 70
+2022-12-13 11:25:07,646 INFO [train.py:421] (2/8) Epoch 9, batch 46400, loss[loss=2.527, over 1120.00 frames. , ppl: 12.519217206343363] tot_loss[loss=2.271, over 5539227.63 frames. , ppl: 9.692192908462902], batch size: 70
+2022-12-13 11:26:44,518 INFO [train.py:421] (2/8) Epoch 9, batch 46600, loss[loss=2.556, over 980.00 frames. , ppl: 12.87950331208894] tot_loss[loss=2.272, over 5507317.75 frames. , ppl: 9.694501508776021], batch size: 70
+2022-12-13 11:28:29,186 INFO [train.py:421] (2/8) Epoch 9, batch 46800, loss[loss=2.187, over 4480.00 frames. , ppl: 8.910451430783034] tot_loss[loss=2.271, over 5499593.20 frames. , ppl: 9.693673593175173], batch size: 70
+2022-12-13 11:30:14,119 INFO [train.py:421] (2/8) Epoch 9, batch 47000, loss[loss=2.646, over 1120.00 frames. , ppl: 14.103265089290794] tot_loss[loss=2.272, over 5502986.94 frames. , ppl: 9.69536832726647], batch size: 70
+2022-12-13 11:30:14,119 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:30:14,868 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.63486544043961
+2022-12-13 11:31:58,222 INFO [train.py:421] (2/8) Epoch 9, batch 47200, loss[loss=2.33, over 2380.00 frames. , ppl: 10.279841157456167] tot_loss[loss=2.271, over 5527355.99 frames. , ppl: 9.686931941004998], batch size: 70
+2022-12-13 11:33:39,674 INFO [train.py:421] (2/8) Epoch 9, batch 47400, loss[loss=2.419, over 1190.00 frames. , ppl: 11.231234707574096] tot_loss[loss=2.27, over 5532653.95 frames. , ppl: 9.67961206421579], batch size: 70
+2022-12-13 11:35:21,971 INFO [train.py:421] (2/8) Epoch 9, batch 47600, loss[loss=2.181, over 3780.00 frames. , ppl: 8.854097102610208] tot_loss[loss=2.27, over 5573600.26 frames. , ppl: 9.676742620806989], batch size: 70
+2022-12-13 11:37:01,535 INFO [train.py:421] (2/8) Epoch 9, batch 47800, loss[loss=2.574, over 770.00 frames. , ppl: 13.119731114195014] tot_loss[loss=2.27, over 5568484.37 frames. , ppl: 9.678869122713248], batch size: 70
+2022-12-13 11:38:43,294 INFO [train.py:421] (2/8) Epoch 9, batch 48000, loss[loss=2.108, over 4270.00 frames. , ppl: 8.23033467898974] tot_loss[loss=2.268, over 5620376.50 frames. , ppl: 9.663316378985575], batch size: 70
+2022-12-13 11:38:43,295 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:38:44,056 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.627387380087804
+2022-12-13 11:40:21,734 INFO [train.py:421] (2/8) Epoch 9, batch 48200, loss[loss=2.414, over 1680.00 frames. , ppl: 11.182973299962931] tot_loss[loss=2.268, over 5624465.53 frames. , ppl: 9.663920794757738], batch size: 70
+2022-12-13 11:42:04,578 INFO [train.py:421] (2/8) Epoch 9, batch 48400, loss[loss=2.449, over 1330.00 frames. , ppl: 11.581452363606907] tot_loss[loss=2.269, over 5609935.11 frames. , ppl: 9.668294993811804], batch size: 70
+2022-12-13 11:43:45,248 INFO [train.py:421] (2/8) Epoch 9, batch 48600, loss[loss=2.406, over 1330.00 frames. , ppl: 11.08475171649529] tot_loss[loss=2.269, over 5601409.84 frames. , ppl: 9.673524144830067], batch size: 70
+2022-12-13 11:45:22,550 INFO [train.py:421] (2/8) Epoch 9, batch 48800, loss[loss=2.267, over 3430.00 frames. , ppl: 9.652820347713092] tot_loss[loss=2.269, over 5601224.10 frames. , ppl: 9.672961448479391], batch size: 70
+2022-12-13 11:47:02,920 INFO [train.py:421] (2/8) Epoch 9, batch 49000, loss[loss=2.36, over 1680.00 frames. , ppl: 10.591519701970808] tot_loss[loss=2.27, over 5619433.77 frames. , ppl: 9.67806782470996], batch size: 70
+2022-12-13 11:47:02,920 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:47:03,679 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.644904227803092
+2022-12-13 11:48:42,705 INFO [train.py:421] (2/8) Epoch 9, batch 49200, loss[loss=2.325, over 3220.00 frames. , ppl: 10.224041006981594] tot_loss[loss=2.271, over 5574428.56 frames. , ppl: 9.685735255318402], batch size: 70
+2022-12-13 11:50:20,797 INFO [train.py:421] (2/8) Epoch 9, batch 49400, loss[loss=2.328, over 2380.00 frames. , ppl: 10.260431266555361] tot_loss[loss=2.272, over 5525284.00 frames. , ppl: 9.702081367470422], batch size: 70
+2022-12-13 11:52:05,012 INFO [train.py:421] (2/8) Epoch 9, batch 49600, loss[loss=2.339, over 4340.00 frames. , ppl: 10.373996778066134] tot_loss[loss=2.272, over 5522754.68 frames. , ppl: 9.70347153855963], batch size: 70
+2022-12-13 11:53:47,506 INFO [train.py:421] (2/8) Epoch 9, batch 49800, loss[loss=4.066, over 350.00 frames. , ppl: 58.305585410961136] tot_loss[loss=2.272, over 5552807.81 frames. , ppl: 9.702392449910642], batch size: 70
+2022-12-13 11:55:29,008 INFO [train.py:421] (2/8) Epoch 9, batch 50000, loss[loss=2.192, over 6440.00 frames. , ppl: 8.957341660626843] tot_loss[loss=2.271, over 5584542.23 frames. , ppl: 9.689381448522898], batch size: 70
+2022-12-13 11:55:29,009 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 11:55:29,755 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.643979240614769
+2022-12-13 11:57:09,485 INFO [train.py:421] (2/8) Epoch 9, batch 50200, loss[loss=2.309, over 2380.00 frames. , ppl: 10.065185021347913] tot_loss[loss=2.272, over 5550595.34 frames. , ppl: 9.697054539732715], batch size: 70
+2022-12-13 11:58:51,076 INFO [train.py:421] (2/8) Epoch 9, batch 50400, loss[loss=2.355, over 1610.00 frames. , ppl: 10.537382940140056] tot_loss[loss=2.273, over 5497904.64 frames. , ppl: 9.711720000624656], batch size: 70
+2022-12-13 12:00:30,170 INFO [train.py:421] (2/8) Epoch 9, batch 50600, loss[loss=2.156, over 3010.00 frames. , ppl: 8.637899859189213] tot_loss[loss=2.273, over 5496882.26 frames. , ppl: 9.709039086390831], batch size: 70
+2022-12-13 12:02:07,593 INFO [train.py:421] (2/8) Epoch 9, batch 50800, loss[loss=2.539, over 980.00 frames. , ppl: 12.667751479670056] tot_loss[loss=2.272, over 5526711.62 frames. , ppl: 9.69633647035242], batch size: 70
+2022-12-13 12:03:46,555 INFO [train.py:421] (2/8) Epoch 9, batch 51000, loss[loss=2.363, over 1960.00 frames. , ppl: 10.62794683645459] tot_loss[loss=2.271, over 5536776.76 frames. , ppl: 9.690932535425757], batch size: 70
+2022-12-13 12:03:46,555 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:03:47,301 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.639113099787668
+2022-12-13 12:05:25,875 INFO [train.py:421] (2/8) Epoch 9, batch 51200, loss[loss=2.221, over 2660.00 frames. , ppl: 9.219589815762193] tot_loss[loss=2.271, over 5521104.60 frames. , ppl: 9.690999729176356], batch size: 70
+2022-12-13 12:07:04,769 INFO [train.py:421] (2/8) Epoch 9, batch 51400, loss[loss=2.242, over 2660.00 frames. , ppl: 9.407629205861612] tot_loss[loss=2.272, over 5485210.15 frames. , ppl: 9.703007772081564], batch size: 70
+2022-12-13 12:08:46,931 INFO [train.py:421] (2/8) Epoch 9, batch 51600, loss[loss=2.337, over 1470.00 frames. , ppl: 10.354758675040111] tot_loss[loss=2.272, over 5477692.05 frames. , ppl: 9.699986475067103], batch size: 70
+2022-12-13 12:10:24,705 INFO [train.py:421] (2/8) Epoch 9, batch 51800, loss[loss=2.634, over 840.00 frames. , ppl: 13.93633007399341] tot_loss[loss=2.273, over 5466168.43 frames. , ppl: 9.706848623275615], batch size: 70
+2022-12-13 12:12:05,979 INFO [train.py:421] (2/8) Epoch 9, batch 52000, loss[loss=2.308, over 2520.00 frames. , ppl: 10.055282374046898] tot_loss[loss=2.271, over 5544961.66 frames. , ppl: 9.69103054992496], batch size: 70
+2022-12-13 12:12:05,980 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:12:06,730 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.646610262463758
+2022-12-13 12:13:45,404 INFO [train.py:421] (2/8) Epoch 9, batch 52200, loss[loss=3.21, over 490.00 frames. , ppl: 24.77373600907813] tot_loss[loss=2.273, over 5505702.39 frames. , ppl: 9.706054302941434], batch size: 70
+2022-12-13 12:15:24,527 INFO [train.py:421] (2/8) Epoch 9, batch 52400, loss[loss=2.18, over 3430.00 frames. , ppl: 8.848283368343653] tot_loss[loss=2.272, over 5545019.92 frames. , ppl: 9.702861186482972], batch size: 70
+2022-12-13 12:17:07,156 INFO [train.py:421] (2/8) Epoch 9, batch 52600, loss[loss=2.248, over 3570.00 frames. , ppl: 9.466033798275909] tot_loss[loss=2.273, over 5528954.85 frames. , ppl: 9.70568583231873], batch size: 70
+2022-12-13 12:18:50,844 INFO [train.py:421] (2/8) Epoch 9, batch 52800, loss[loss=2.533, over 1680.00 frames. , ppl: 12.594970708671479] tot_loss[loss=2.273, over 5520003.50 frames. , ppl: 9.707681865603364], batch size: 70
+2022-12-13 12:20:28,986 INFO [train.py:421] (2/8) Epoch 9, batch 53000, loss[loss=2.329, over 3010.00 frames. , ppl: 10.269757759450963] tot_loss[loss=2.272, over 5529375.30 frames. , ppl: 9.702207630076673], batch size: 70
+2022-12-13 12:20:28,986 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:20:29,752 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.642737499169275
+2022-12-13 12:22:10,639 INFO [train.py:421] (2/8) Epoch 9, batch 53200, loss[loss=2.361, over 1680.00 frames. , ppl: 10.59771824857518] tot_loss[loss=2.273, over 5526746.42 frames. , ppl: 9.712285300229313], batch size: 70
+2022-12-13 12:23:51,912 INFO [train.py:421] (2/8) Epoch 9, batch 53400, loss[loss=2.201, over 3080.00 frames. , ppl: 9.0329104625901] tot_loss[loss=2.273, over 5539112.59 frames. , ppl: 9.70898246756472], batch size: 70
+2022-12-13 12:25:35,097 INFO [train.py:421] (2/8) Epoch 9, batch 53600, loss[loss=2.286, over 1470.00 frames. , ppl: 9.836089678192183] tot_loss[loss=2.274, over 5499153.01 frames. , ppl: 9.713952120416643], batch size: 70
+2022-12-13 12:27:15,424 INFO [train.py:421] (2/8) Epoch 9, batch 53800, loss[loss=2.387, over 1960.00 frames. , ppl: 10.87687017566943] tot_loss[loss=2.275, over 5445659.59 frames. , ppl: 9.72570079437894], batch size: 70
+2022-12-13 12:28:55,762 INFO [train.py:421] (2/8) Epoch 9, batch 54000, loss[loss=2.307, over 1820.00 frames. , ppl: 10.046889151762825] tot_loss[loss=2.276, over 5426484.46 frames. , ppl: 9.73742277661355], batch size: 70
+2022-12-13 12:28:55,763 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:28:56,510 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.633422404126998
+2022-12-13 12:30:35,814 INFO [train.py:421] (2/8) Epoch 9, batch 54200, loss[loss=2.5, over 980.00 frames. , ppl: 12.182648743652232] tot_loss[loss=2.275, over 5410625.86 frames. , ppl: 9.732291175361711], batch size: 70
+2022-12-13 12:32:16,566 INFO [train.py:421] (2/8) Epoch 9, batch 54400, loss[loss=2.208, over 4270.00 frames. , ppl: 9.10098715021428] tot_loss[loss=2.275, over 5419870.87 frames. , ppl: 9.72434990355146], batch size: 70
+2022-12-13 12:33:55,484 INFO [train.py:421] (2/8) Epoch 9, batch 54600, loss[loss=2.434, over 1890.00 frames. , ppl: 11.399052021483067] tot_loss[loss=2.275, over 5422597.57 frames. , ppl: 9.727579034601039], batch size: 70
+2022-12-13 12:35:38,800 INFO [train.py:421] (2/8) Epoch 9, batch 54800, loss[loss=2.616, over 1050.00 frames. , ppl: 13.679564649179405] tot_loss[loss=2.274, over 5453733.51 frames. , ppl: 9.715986689505925], batch size: 70
+2022-12-13 12:37:17,706 INFO [train.py:421] (2/8) Epoch 9, batch 55000, loss[loss=2.295, over 2870.00 frames. , ppl: 9.924614654281658] tot_loss[loss=2.273, over 5469038.73 frames. , ppl: 9.712427826344307], batch size: 70
+2022-12-13 12:37:17,707 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:37:18,465 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.64445171527035
+2022-12-13 12:39:02,260 INFO [train.py:421] (2/8) Epoch 9, batch 55200, loss[loss=2.16, over 5740.00 frames. , ppl: 8.67181098654742] tot_loss[loss=2.274, over 5470800.33 frames. , ppl: 9.717197817068627], batch size: 70
+2022-12-13 12:40:39,117 INFO [train.py:421] (2/8) Epoch 9, batch 55400, loss[loss=2.641, over 910.00 frames. , ppl: 14.020749555677885] tot_loss[loss=2.275, over 5452162.80 frames. , ppl: 9.72383604766873], batch size: 70
+2022-12-13 12:42:18,508 INFO [train.py:421] (2/8) Epoch 9, batch 55600, loss[loss=3.272, over 490.00 frames. , ppl: 26.352899171099867] tot_loss[loss=2.273, over 5531485.66 frames. , ppl: 9.70984241405165], batch size: 70
+2022-12-13 12:43:58,843 INFO [train.py:421] (2/8) Epoch 9, batch 55800, loss[loss=2.479, over 1260.00 frames. , ppl: 11.929602246776367] tot_loss[loss=2.272, over 5593790.39 frames. , ppl: 9.69608128093531], batch size: 70
+2022-12-13 12:45:38,587 INFO [train.py:421] (2/8) Epoch 9, batch 56000, loss[loss=2.473, over 840.00 frames. , ppl: 11.853010017457331] tot_loss[loss=2.272, over 5595460.77 frames. , ppl: 9.69793408937651], batch size: 70
+2022-12-13 12:45:38,588 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:45:39,335 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.625552251361995
+2022-12-13 12:47:19,624 INFO [train.py:421] (2/8) Epoch 9, batch 56200, loss[loss=2.166, over 6580.00 frames. , ppl: 8.72229592889363] tot_loss[loss=2.272, over 5582806.23 frames. , ppl: 9.697267645216312], batch size: 70
+2022-12-13 12:48:56,801 INFO [train.py:421] (2/8) Epoch 9, batch 56400, loss[loss=2.249, over 2520.00 frames. , ppl: 9.473719463669351] tot_loss[loss=2.272, over 5555526.61 frames. , ppl: 9.702863760260085], batch size: 70
+2022-12-13 12:50:35,595 INFO [train.py:421] (2/8) Epoch 9, batch 56600, loss[loss=2.452, over 1540.00 frames. , ppl: 11.612033178396167] tot_loss[loss=2.274, over 5523260.83 frames. , ppl: 9.713589748894004], batch size: 70
+2022-12-13 12:52:13,282 INFO [train.py:421] (2/8) Epoch 9, batch 56800, loss[loss=2.277, over 3500.00 frames. , ppl: 9.750447657491614] tot_loss[loss=2.273, over 5553067.89 frames. , ppl: 9.70679582818711], batch size: 70
+2022-12-13 12:53:52,633 INFO [train.py:421] (2/8) Epoch 9, batch 57000, loss[loss=2.549, over 1190.00 frames. , ppl: 12.800263579572364] tot_loss[loss=2.273, over 5542734.18 frames. , ppl: 9.709985029737362], batch size: 70
+2022-12-13 12:53:52,633 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 12:53:53,394 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.636780791774152
+2022-12-13 12:55:34,866 INFO [train.py:421] (2/8) Epoch 9, batch 57200, loss[loss=2.509, over 770.00 frames. , ppl: 12.294742855077692] tot_loss[loss=2.275, over 5492023.28 frames. , ppl: 9.727981801746623], batch size: 70
+2022-12-13 12:57:12,827 INFO [train.py:421] (2/8) Epoch 9, batch 57400, loss[loss=2.909, over 560.00 frames. , ppl: 18.333931646945064] tot_loss[loss=2.275, over 5471353.30 frames. , ppl: 9.731297688664755], batch size: 70
+2022-12-13 12:58:55,766 INFO [train.py:421] (2/8) Epoch 9, batch 57600, loss[loss=2.176, over 10290.00 frames. , ppl: 8.812573136864748] tot_loss[loss=2.276, over 5449430.54 frames. , ppl: 9.738459826471821], batch size: 70
+2022-12-13 13:00:36,168 INFO [train.py:421] (2/8) Epoch 9, batch 57800, loss[loss=2.406, over 1050.00 frames. , ppl: 11.088197245645985] tot_loss[loss=2.277, over 5425309.70 frames. , ppl: 9.748971716934962], batch size: 70
+2022-12-13 13:02:18,862 INFO [train.py:421] (2/8) Epoch 9, batch 58000, loss[loss=2.126, over 5110.00 frames. , ppl: 8.378807247954027] tot_loss[loss=2.276, over 5464311.99 frames. , ppl: 9.733521445642712], batch size: 70
+2022-12-13 13:02:18,863 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:02:19,612 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635025157347723
+2022-12-13 13:04:04,620 INFO [train.py:421] (2/8) Epoch 9, batch 58200, loss[loss=2.226, over 1820.00 frames. , ppl: 9.260404351744887] tot_loss[loss=2.276, over 5455669.48 frames. , ppl: 9.733856251042026], batch size: 70
+2022-12-13 13:05:43,765 INFO [train.py:421] (2/8) Epoch 9, batch 58400, loss[loss=2.649, over 770.00 frames. , ppl: 14.139725442519877] tot_loss[loss=2.276, over 5419141.73 frames. , ppl: 9.73925751963939], batch size: 70
+2022-12-13 13:07:23,478 INFO [train.py:421] (2/8) Epoch 9, batch 58600, loss[loss=2.357, over 2170.00 frames. , ppl: 10.562450618917817] tot_loss[loss=2.277, over 5372973.75 frames. , ppl: 9.747357764584516], batch size: 70
+2022-12-13 13:09:02,105 INFO [train.py:421] (2/8) Epoch 9, batch 58800, loss[loss=2.345, over 2310.00 frames. , ppl: 10.428396502953284] tot_loss[loss=2.277, over 5369971.99 frames. , ppl: 9.751868539592937], batch size: 70
+2022-12-13 13:10:46,532 INFO [train.py:421] (2/8) Epoch 9, batch 59000, loss[loss=2.483, over 1260.00 frames. , ppl: 11.978057141755572] tot_loss[loss=2.277, over 5386410.94 frames. , ppl: 9.746554796597257], batch size: 70
+2022-12-13 13:10:46,532 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:10:47,278 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634386305600566
+2022-12-13 13:12:26,044 INFO [train.py:421] (2/8) Epoch 9, batch 59200, loss[loss=2.192, over 4130.00 frames. , ppl: 8.953512048504306] tot_loss[loss=2.277, over 5405239.74 frames. , ppl: 9.746182757293374], batch size: 70
+2022-12-13 13:14:06,375 INFO [train.py:421] (2/8) Epoch 9, batch 59400, loss[loss=2.438, over 2030.00 frames. , ppl: 11.455162280689063] tot_loss[loss=2.277, over 5419877.81 frames. , ppl: 9.743249834118755], batch size: 70
+2022-12-13 13:15:48,493 INFO [train.py:421] (2/8) Epoch 9, batch 59600, loss[loss=2.562, over 910.00 frames. , ppl: 12.965869947067505] tot_loss[loss=2.276, over 5441174.47 frames. , ppl: 9.737541625912943], batch size: 70
+2022-12-13 13:17:26,949 INFO [train.py:421] (2/8) Epoch 9, batch 59800, loss[loss=2.281, over 1820.00 frames. , ppl: 9.782450504107564] tot_loss[loss=2.275, over 5472749.01 frames. , ppl: 9.727809646545014], batch size: 70
+2022-12-13 13:19:05,927 INFO [train.py:421] (2/8) Epoch 9, batch 60000, loss[loss=2.389, over 1820.00 frames. , ppl: 10.898675742604954] tot_loss[loss=2.275, over 5482182.42 frames. , ppl: 9.724770209989995], batch size: 70
+2022-12-13 13:19:05,928 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:19:06,687 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.63870508346662
+2022-12-13 13:20:45,779 INFO [train.py:421] (2/8) Epoch 9, batch 60200, loss[loss=2.448, over 910.00 frames. , ppl: 11.569689126021636] tot_loss[loss=2.273, over 5503805.09 frames. , ppl: 9.712114261937618], batch size: 70
+2022-12-13 13:22:24,271 INFO [train.py:421] (2/8) Epoch 9, batch 60400, loss[loss=2.464, over 1190.00 frames. , ppl: 11.748534380222317] tot_loss[loss=2.273, over 5502689.09 frames. , ppl: 9.712747158014835], batch size: 70
+2022-12-13 13:24:05,149 INFO [train.py:421] (2/8) Epoch 9, batch 60600, loss[loss=2.205, over 5950.00 frames. , ppl: 9.072602503913641] tot_loss[loss=2.273, over 5516816.19 frames. , ppl: 9.707581554363754], batch size: 70
+2022-12-13 13:25:47,347 INFO [train.py:421] (2/8) Epoch 9, batch 60800, loss[loss=2.541, over 980.00 frames. , ppl: 12.693336343524292] tot_loss[loss=2.272, over 5527216.47 frames. , ppl: 9.702465316044627], batch size: 70
+2022-12-13 13:27:25,583 INFO [train.py:421] (2/8) Epoch 9, batch 61000, loss[loss=2.509, over 770.00 frames. , ppl: 12.28839815504157] tot_loss[loss=2.274, over 5459207.86 frames. , ppl: 9.719440333641671], batch size: 70
+2022-12-13 13:27:25,583 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:27:26,343 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.648386572270214
+2022-12-13 13:29:06,955 INFO [train.py:421] (2/8) Epoch 9, batch 61200, loss[loss=2.213, over 3500.00 frames. , ppl: 9.141729670170038] tot_loss[loss=2.274, over 5458387.06 frames. , ppl: 9.715611311587256], batch size: 70
+2022-12-13 13:30:44,585 INFO [train.py:421] (2/8) Epoch 9, batch 61400, loss[loss=2.44, over 1680.00 frames. , ppl: 11.469778000649336] tot_loss[loss=2.274, over 5448712.47 frames. , ppl: 9.717651364315518], batch size: 70
+2022-12-13 13:32:23,424 INFO [train.py:421] (2/8) Epoch 9, batch 61600, loss[loss=2.355, over 1960.00 frames. , ppl: 10.541992948473569] tot_loss[loss=2.274, over 5471375.23 frames. , ppl: 9.71733384912211], batch size: 70
+2022-12-13 13:34:08,412 INFO [train.py:421] (2/8) Epoch 9, batch 61800, loss[loss=2.597, over 840.00 frames. , ppl: 13.429994260182742] tot_loss[loss=2.275, over 5419424.82 frames. , ppl: 9.729305424556433], batch size: 70
+2022-12-13 13:35:49,227 INFO [train.py:421] (2/8) Epoch 9, batch 62000, loss[loss=2.348, over 1610.00 frames. , ppl: 10.466333678120936] tot_loss[loss=2.276, over 5409166.68 frames. , ppl: 9.733551450305113], batch size: 70
+2022-12-13 13:35:49,228 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:35:49,994 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.263, over 211138.00 frames. , ppl: 9.615221936937656
+2022-12-13 13:37:32,295 INFO [train.py:421] (2/8) Epoch 9, batch 62200, loss[loss=2.149, over 4760.00 frames. , ppl: 8.577319945753857] tot_loss[loss=2.275, over 5432783.64 frames. , ppl: 9.727619402998117], batch size: 70
+2022-12-13 13:39:16,055 INFO [train.py:421] (2/8) Epoch 9, batch 62400, loss[loss=2.522, over 1330.00 frames. , ppl: 12.45910597625979] tot_loss[loss=2.275, over 5446678.86 frames. , ppl: 9.725118673652041], batch size: 70
+2022-12-13 13:40:53,821 INFO [train.py:421] (2/8) Epoch 9, batch 62600, loss[loss=2.283, over 2240.00 frames. , ppl: 9.804733230809951] tot_loss[loss=2.274, over 5450275.18 frames. , ppl: 9.721701501863546], batch size: 70
+2022-12-13 13:42:33,914 INFO [train.py:421] (2/8) Epoch 9, batch 62800, loss[loss=2.163, over 13160.00 frames. , ppl: 8.695153565723173] tot_loss[loss=2.274, over 5465795.66 frames. , ppl: 9.714933806615747], batch size: 70
+2022-12-13 13:44:11,462 INFO [train.py:421] (2/8) Epoch 9, batch 63000, loss[loss=2.416, over 1400.00 frames. , ppl: 11.204700262773393] tot_loss[loss=2.274, over 5452220.60 frames. , ppl: 9.714122429832093], batch size: 70
+2022-12-13 13:44:11,462 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:44:12,223 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.639786507154147
+2022-12-13 13:45:47,948 INFO [train.py:421] (2/8) Epoch 9, batch 63200, loss[loss=2.305, over 2240.00 frames. , ppl: 10.021928728536123] tot_loss[loss=2.275, over 5383461.15 frames. , ppl: 9.732619663457063], batch size: 70
+2022-12-13 13:47:29,989 INFO [train.py:421] (2/8) Epoch 9, batch 63400, loss[loss=2.244, over 2940.00 frames. , ppl: 9.434939900361227] tot_loss[loss=2.275, over 5377640.38 frames. , ppl: 9.729424568087355], batch size: 70
+2022-12-13 13:49:09,472 INFO [train.py:421] (2/8) Epoch 9, batch 63600, loss[loss=2.21, over 4620.00 frames. , ppl: 9.114982676458833] tot_loss[loss=2.276, over 5342127.47 frames. , ppl: 9.73781136767972], batch size: 70
+2022-12-13 13:50:48,800 INFO [train.py:421] (2/8) Epoch 9, batch 63800, loss[loss=2.154, over 3850.00 frames. , ppl: 8.619525463028715] tot_loss[loss=2.275, over 5359664.16 frames. , ppl: 9.732191082038511], batch size: 70
+2022-12-13 13:52:29,000 INFO [train.py:421] (2/8) Epoch 9, batch 64000, loss[loss=2.218, over 5460.00 frames. , ppl: 9.187572697209761] tot_loss[loss=2.276, over 5342945.48 frames. , ppl: 9.734583673282597], batch size: 70
+2022-12-13 13:52:29,000 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 13:52:29,747 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.627401629350286
+2022-12-13 13:54:05,564 INFO [train.py:421] (2/8) Epoch 9, batch 64200, loss[loss=2.677, over 910.00 frames. , ppl: 14.534567372864572] tot_loss[loss=2.277, over 5304579.26 frames. , ppl: 9.745945489152898], batch size: 70
+2022-12-13 13:55:48,757 INFO [train.py:421] (2/8) Epoch 9, batch 64400, loss[loss=2.962, over 560.00 frames. , ppl: 19.33557686535027] tot_loss[loss=2.275, over 5362370.99 frames. , ppl: 9.731199849814352], batch size: 70
+2022-12-13 13:57:26,852 INFO [train.py:421] (2/8) Epoch 9, batch 64600, loss[loss=2.389, over 980.00 frames. , ppl: 10.901276503351419] tot_loss[loss=2.275, over 5370054.64 frames. , ppl: 9.724460449637592], batch size: 70
+2022-12-13 13:59:05,983 INFO [train.py:421] (2/8) Epoch 9, batch 64800, loss[loss=2.293, over 2100.00 frames. , ppl: 9.900742427681262] tot_loss[loss=2.274, over 5378310.35 frames. , ppl: 9.718278902702997], batch size: 70
+2022-12-13 14:00:44,182 INFO [train.py:421] (2/8) Epoch 9, batch 65000, loss[loss=2.211, over 2870.00 frames. , ppl: 9.123771202163702] tot_loss[loss=2.272, over 5437008.09 frames. , ppl: 9.701626843609123], batch size: 70
+2022-12-13 14:00:44,182 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:00:44,934 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.633545025291745
+2022-12-13 14:02:29,168 INFO [train.py:421] (2/8) Epoch 9, batch 65200, loss[loss=2.383, over 1260.00 frames. , ppl: 10.84209266228197] tot_loss[loss=2.272, over 5484762.12 frames. , ppl: 9.694308521066915], batch size: 70
+2022-12-13 14:04:10,421 INFO [train.py:421] (2/8) Epoch 9, batch 65400, loss[loss=2.221, over 3430.00 frames. , ppl: 9.216102704903843] tot_loss[loss=2.271, over 5509541.91 frames. , ppl: 9.691202465976097], batch size: 70
+2022-12-13 14:05:48,252 INFO [train.py:421] (2/8) Epoch 9, batch 65600, loss[loss=2.358, over 3010.00 frames. , ppl: 10.565608126827998] tot_loss[loss=2.271, over 5514035.39 frames. , ppl: 9.690651618631946], batch size: 70
+2022-12-13 14:07:28,847 INFO [train.py:421] (2/8) Epoch 9, batch 65800, loss[loss=2.192, over 7700.00 frames. , ppl: 8.949103579061784] tot_loss[loss=2.271, over 5525465.18 frames. , ppl: 9.69219211043028], batch size: 70
+2022-12-13 14:09:11,363 INFO [train.py:421] (2/8) Epoch 9, batch 66000, loss[loss=2.252, over 2310.00 frames. , ppl: 9.502037731657104] tot_loss[loss=2.271, over 5526210.85 frames. , ppl: 9.692077993910143], batch size: 70
+2022-12-13 14:09:11,363 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:09:12,123 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.633080213783197
+2022-12-13 14:10:51,455 INFO [train.py:421] (2/8) Epoch 9, batch 66200, loss[loss=2.152, over 6790.00 frames. , ppl: 8.60593843507951] tot_loss[loss=2.272, over 5509637.68 frames. , ppl: 9.695108452000754], batch size: 70
+2022-12-13 14:12:30,193 INFO [train.py:421] (2/8) Epoch 9, batch 66400, loss[loss=2.567, over 840.00 frames. , ppl: 13.021753775192527] tot_loss[loss=2.272, over 5474148.15 frames. , ppl: 9.70160841616799], batch size: 70
+2022-12-13 14:14:07,585 INFO [train.py:421] (2/8) Epoch 9, batch 66600, loss[loss=2.683, over 840.00 frames. , ppl: 14.636156372132081] tot_loss[loss=2.273, over 5457996.56 frames. , ppl: 9.707190523051594], batch size: 70
+2022-12-13 14:15:45,367 INFO [train.py:421] (2/8) Epoch 9, batch 66800, loss[loss=2.774, over 560.00 frames. , ppl: 16.028512347852228] tot_loss[loss=2.274, over 5441785.49 frames. , ppl: 9.720920293521003], batch size: 70
+2022-12-13 14:17:23,668 INFO [train.py:421] (2/8) Epoch 9, batch 67000, loss[loss=2.426, over 2380.00 frames. , ppl: 11.310918861047812] tot_loss[loss=2.274, over 5439021.64 frames. , ppl: 9.71837814368123], batch size: 70
+2022-12-13 14:17:23,668 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:17:24,429 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635444426828531
+2022-12-13 14:19:04,964 INFO [train.py:421] (2/8) Epoch 9, batch 67200, loss[loss=2.28, over 1960.00 frames. , ppl: 9.780501633241673] tot_loss[loss=2.273, over 5488235.80 frames. , ppl: 9.707727944633069], batch size: 70
+2022-12-13 14:20:42,284 INFO [train.py:421] (2/8) Epoch 9, batch 67400, loss[loss=2.235, over 4690.00 frames. , ppl: 9.349986349897527] tot_loss[loss=2.274, over 5465759.76 frames. , ppl: 9.713386509768393], batch size: 70
+2022-12-13 14:22:19,562 INFO [train.py:421] (2/8) Epoch 9, batch 67600, loss[loss=2.326, over 1820.00 frames. , ppl: 10.234573952186206] tot_loss[loss=2.274, over 5451199.62 frames. , ppl: 9.719526694527197], batch size: 70
+2022-12-13 14:23:58,429 INFO [train.py:421] (2/8) Epoch 9, batch 67800, loss[loss=2.269, over 2660.00 frames. , ppl: 9.666613514112468] tot_loss[loss=2.273, over 5491521.62 frames. , ppl: 9.709559265797], batch size: 70
+2022-12-13 14:25:36,735 INFO [train.py:421] (2/8) Epoch 9, batch 68000, loss[loss=2.873, over 700.00 frames. , ppl: 17.684576421668773] tot_loss[loss=2.273, over 5499858.32 frames. , ppl: 9.7093352065698], batch size: 70
+2022-12-13 14:25:36,735 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:25:37,495 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621968490265408
+2022-12-13 14:27:17,453 INFO [train.py:421] (2/8) Epoch 9, batch 68200, loss[loss=2.322, over 2450.00 frames. , ppl: 10.199529152619284] tot_loss[loss=2.274, over 5483966.32 frames. , ppl: 9.71399120275655], batch size: 70
+2022-12-13 14:28:57,911 INFO [train.py:421] (2/8) Epoch 9, batch 68400, loss[loss=2.418, over 1400.00 frames. , ppl: 11.22389073690935] tot_loss[loss=2.272, over 5530096.39 frames. , ppl: 9.700791849423938], batch size: 70
+2022-12-13 14:30:42,245 INFO [train.py:421] (2/8) Epoch 9, batch 68600, loss[loss=2.462, over 1400.00 frames. , ppl: 11.728112057265175] tot_loss[loss=2.272, over 5526194.49 frames. , ppl: 9.701030506695291], batch size: 70
+2022-12-13 14:32:21,733 INFO [train.py:421] (2/8) Epoch 9, batch 68800, loss[loss=2.211, over 6440.00 frames. , ppl: 9.12760978650394] tot_loss[loss=2.272, over 5515299.32 frames. , ppl: 9.701164713580607], batch size: 70
+2022-12-13 14:34:01,334 INFO [train.py:421] (2/8) Epoch 9, batch 69000, loss[loss=2.38, over 1960.00 frames. , ppl: 10.809737252775824] tot_loss[loss=2.272, over 5549693.42 frames. , ppl: 9.698312483492392], batch size: 70
+2022-12-13 14:34:01,335 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:34:02,095 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634126784178106
+2022-12-13 14:35:35,880 INFO [train.py:421] (2/8) Epoch 9, batch 69200, loss[loss=2.26, over 3570.00 frames. , ppl: 9.581317771815666] tot_loss[loss=2.273, over 5525925.21 frames. , ppl: 9.709559383893904], batch size: 70
+2022-12-13 14:37:12,959 INFO [train.py:421] (2/8) Epoch 9, batch 69400, loss[loss=2.319, over 2310.00 frames. , ppl: 10.165325974420142] tot_loss[loss=2.275, over 5480814.00 frames. , ppl: 9.723746806667828], batch size: 70
+2022-12-13 14:38:55,433 INFO [train.py:421] (2/8) Epoch 9, batch 69600, loss[loss=2.177, over 3570.00 frames. , ppl: 8.823462133182991] tot_loss[loss=2.275, over 5465088.81 frames. , ppl: 9.727212020462252], batch size: 70
+2022-12-13 14:40:38,350 INFO [train.py:421] (2/8) Epoch 9, batch 69800, loss[loss=2.2, over 8680.00 frames. , ppl: 9.023640811424395] tot_loss[loss=2.276, over 5449376.73 frames. , ppl: 9.735066044924142], batch size: 70
+2022-12-13 14:42:20,435 INFO [train.py:421] (2/8) Epoch 9, batch 70000, loss[loss=3.2, over 490.00 frames. , ppl: 24.52044439142376] tot_loss[loss=2.276, over 5420185.52 frames. , ppl: 9.740484024864633], batch size: 70
+2022-12-13 14:42:20,435 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:42:21,216 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.632411551904289
+2022-12-13 14:44:01,630 INFO [train.py:421] (2/8) Epoch 9, batch 70200, loss[loss=2.312, over 3360.00 frames. , ppl: 10.091750561700065] tot_loss[loss=2.275, over 5460292.54 frames. , ppl: 9.724697193072746], batch size: 70
+2022-12-13 14:45:44,015 INFO [train.py:421] (2/8) Epoch 9, batch 70400, loss[loss=2.157, over 5670.00 frames. , ppl: 8.648686047716405] tot_loss[loss=2.274, over 5498041.90 frames. , ppl: 9.716917200207185], batch size: 70
+2022-12-13 14:47:24,087 INFO [train.py:421] (2/8) Epoch 9, batch 70600, loss[loss=2.355, over 1540.00 frames. , ppl: 10.53920781614063] tot_loss[loss=2.274, over 5486551.70 frames. , ppl: 9.718563770749634], batch size: 70
+2022-12-13 14:49:06,611 INFO [train.py:421] (2/8) Epoch 9, batch 70800, loss[loss=2.234, over 2940.00 frames. , ppl: 9.339844785261185] tot_loss[loss=2.274, over 5488832.16 frames. , ppl: 9.71913279094594], batch size: 70
+2022-12-13 14:50:45,766 INFO [train.py:421] (2/8) Epoch 9, batch 71000, loss[loss=2.36, over 1400.00 frames. , ppl: 10.589756565604022] tot_loss[loss=2.275, over 5450947.35 frames. , ppl: 9.72960005920367], batch size: 70
+2022-12-13 14:50:45,766 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 14:50:46,533 INFO [train.py:452] (2/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622304589200576
+2022-12-13 14:52:28,161 INFO [train.py:421] (2/8) Epoch 9, batch 71200, loss[loss=2.203, over 3640.00 frames. , ppl: 9.051109059808967] tot_loss[loss=2.276, over 5413364.96 frames. , ppl: 9.736250750792093], batch size: 70
+2022-12-13 14:54:10,852 INFO [train.py:421] (2/8) Epoch 9, batch 71400, loss[loss=2.326, over 2100.00 frames. , ppl: 10.239440478085625] tot_loss[loss=2.276, over 5420554.06 frames. , ppl: 9.734885462807656], batch size: 70
+2022-12-13 14:55:50,276 INFO [train.py:421] (2/8) Epoch 9, batch 71600, loss[loss=2.453, over 1400.00 frames. , ppl: 11.621138838165352] tot_loss[loss=2.275, over 5488582.93 frames. , ppl: 9.724100361370066], batch size: 70
+2022-12-13 14:57:34,057 INFO [train.py:421] (2/8) Epoch 9, batch 71800, loss[loss=2.228, over 4690.00 frames. , ppl: 9.277224178856745] tot_loss[loss=2.273, over 5525300.61 frames. , ppl: 9.711172123527042], batch size: 70
+2022-12-13 14:58:47,403 INFO [train.py:421] (2/8) Epoch 10, batch 0, loss[loss=2.658, over 840.00 frames. , ppl: 14.273344647936952] tot_loss[loss=2.658, over 840.00 frames. , ppl: 14.273344647936952], batch size: 70
+2022-12-13 15:00:28,718 INFO [train.py:421] (2/8) Epoch 10, batch 200, loss[loss=2.214, over 2100.00 frames. , ppl: 9.151104917209466] tot_loss[loss=2.251, over 553206.77 frames. , ppl: 9.500426861248828], batch size: 70
+2022-12-13 15:02:15,425 INFO [train.py:421] (2/8) Epoch 10, batch 400, loss[loss=2.297, over 3430.00 frames. , ppl: 9.94102657004924] tot_loss[loss=2.264, over 1004367.52 frames. , ppl: 9.623271292870783], batch size: 70
+2022-12-13 15:03:56,689 INFO [train.py:421] (2/8) Epoch 10, batch 600, loss[loss=2.354, over 2520.00 frames. , ppl: 10.529432353707811] tot_loss[loss=2.259, over 1458571.54 frames. , ppl: 9.572170895435729], batch size: 70
+2022-12-13 15:05:40,249 INFO [train.py:421] (2/8) Epoch 10, batch 800, loss[loss=2.314, over 2520.00 frames. , ppl: 10.114314198808238] tot_loss[loss=2.264, over 1787484.02 frames. , ppl: 9.626237088498286], batch size: 70
+2022-12-13 15:07:21,187 INFO [train.py:421] (2/8) Epoch 10, batch 1000, loss[loss=2.253, over 2520.00 frames. , ppl: 9.516133213790418] tot_loss[loss=2.268, over 2093515.35 frames. , ppl: 9.659274255302893], batch size: 70
+2022-12-13 15:07:21,188 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:07:21,948 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.636474138823111
+2022-12-13 15:09:07,556 INFO [train.py:421] (2/8) Epoch 10, batch 1200, loss[loss=2.2, over 3080.00 frames. , ppl: 9.022617292380783] tot_loss[loss=2.27, over 2384364.94 frames. , ppl: 9.676632239301641], batch size: 70
+2022-12-13 15:10:54,512 INFO [train.py:421] (2/8) Epoch 10, batch 1400, loss[loss=2.114, over 6440.00 frames. , ppl: 8.280137756299517] tot_loss[loss=2.265, over 2762308.67 frames. , ppl: 9.627159318186713], batch size: 70
+2022-12-13 15:12:36,516 INFO [train.py:421] (2/8) Epoch 10, batch 1600, loss[loss=2.633, over 840.00 frames. , ppl: 13.917528172035944] tot_loss[loss=2.267, over 2959111.39 frames. , ppl: 9.652174939319812], batch size: 70
+2022-12-13 15:14:19,425 INFO [train.py:421] (2/8) Epoch 10, batch 1800, loss[loss=2.625, over 840.00 frames. , ppl: 13.80048233724717] tot_loss[loss=2.269, over 3176329.10 frames. , ppl: 9.66624737174852], batch size: 70
+2022-12-13 15:16:01,288 INFO [train.py:421] (2/8) Epoch 10, batch 2000, loss[loss=2.609, over 840.00 frames. , ppl: 13.585143667406436] tot_loss[loss=2.266, over 3429558.04 frames. , ppl: 9.636860988636162], batch size: 70
+2022-12-13 15:16:01,289 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:16:02,057 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634249414308712
+2022-12-13 15:17:44,577 INFO [train.py:421] (2/8) Epoch 10, batch 2200, loss[loss=2.767, over 630.00 frames. , ppl: 15.91777619929037] tot_loss[loss=2.266, over 3629146.08 frames. , ppl: 9.639986229074573], batch size: 70
+2022-12-13 15:19:27,874 INFO [train.py:421] (2/8) Epoch 10, batch 2400, loss[loss=2.218, over 4060.00 frames. , ppl: 9.192980410509161] tot_loss[loss=2.264, over 3841353.71 frames. , ppl: 9.617743536593595], batch size: 70
+2022-12-13 15:21:09,674 INFO [train.py:421] (2/8) Epoch 10, batch 2600, loss[loss=2.167, over 5600.00 frames. , ppl: 8.732111298974496] tot_loss[loss=2.266, over 3989752.43 frames. , ppl: 9.636770214937131], batch size: 70
+2022-12-13 15:22:53,436 INFO [train.py:421] (2/8) Epoch 10, batch 2800, loss[loss=2.187, over 7000.00 frames. , ppl: 8.905964859319608] tot_loss[loss=2.265, over 4122582.17 frames. , ppl: 9.634934438096694], batch size: 70
+2022-12-13 15:24:36,433 INFO [train.py:421] (2/8) Epoch 10, batch 3000, loss[loss=2.242, over 3570.00 frames. , ppl: 9.41150453668791] tot_loss[loss=2.265, over 4280982.98 frames. , ppl: 9.634917615813345], batch size: 70
+2022-12-13 15:24:36,433 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:24:37,197 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.624782969170553
+2022-12-13 15:26:21,732 INFO [train.py:421] (2/8) Epoch 10, batch 3200, loss[loss=2.213, over 6370.00 frames. , ppl: 9.14543250309867] tot_loss[loss=2.265, over 4398107.12 frames. , ppl: 9.63205509605208], batch size: 70
+2022-12-13 15:28:06,713 INFO [train.py:421] (2/8) Epoch 10, batch 3400, loss[loss=2.348, over 1470.00 frames. , ppl: 10.466281605720383] tot_loss[loss=2.263, over 4587858.26 frames. , ppl: 9.607267289874777], batch size: 70
+2022-12-13 15:29:50,698 INFO [train.py:421] (2/8) Epoch 10, batch 3600, loss[loss=2.252, over 4760.00 frames. , ppl: 9.510619257941766] tot_loss[loss=2.263, over 4689713.32 frames. , ppl: 9.609475101897978], batch size: 70
+2022-12-13 15:31:33,968 INFO [train.py:421] (2/8) Epoch 10, batch 3800, loss[loss=2.344, over 2450.00 frames. , ppl: 10.42524542434363] tot_loss[loss=2.264, over 4733565.63 frames. , ppl: 9.62041553067845], batch size: 70
+2022-12-13 15:33:14,219 INFO [train.py:421] (2/8) Epoch 10, batch 4000, loss[loss=2.185, over 4060.00 frames. , ppl: 8.889888566108969] tot_loss[loss=2.264, over 4801724.68 frames. , ppl: 9.620579605440938], batch size: 70
+2022-12-13 15:33:14,219 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:33:14,973 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.6403101424268
+2022-12-13 15:34:56,293 INFO [train.py:421] (2/8) Epoch 10, batch 4200, loss[loss=2.424, over 1680.00 frames. , ppl: 11.288808638158287] tot_loss[loss=2.263, over 4898579.43 frames. , ppl: 9.613318039353858], batch size: 70
+2022-12-13 15:36:36,696 INFO [train.py:421] (2/8) Epoch 10, batch 4400, loss[loss=2.359, over 2100.00 frames. , ppl: 10.581839000632062] tot_loss[loss=2.264, over 4944858.41 frames. , ppl: 9.619597800546442], batch size: 70
+2022-12-13 15:38:18,534 INFO [train.py:421] (2/8) Epoch 10, batch 4600, loss[loss=3.003, over 560.00 frames. , ppl: 20.14565319581518] tot_loss[loss=2.265, over 4964286.53 frames. , ppl: 9.633505102491883], batch size: 70
+2022-12-13 15:39:59,373 INFO [train.py:421] (2/8) Epoch 10, batch 4800, loss[loss=2.129, over 8610.00 frames. , ppl: 8.402593019177658] tot_loss[loss=2.265, over 5008144.00 frames. , ppl: 9.629296078214304], batch size: 70
+2022-12-13 15:41:41,307 INFO [train.py:421] (2/8) Epoch 10, batch 5000, loss[loss=2.353, over 1190.00 frames. , ppl: 10.51876818942004] tot_loss[loss=2.264, over 5076800.56 frames. , ppl: 9.619613286450948], batch size: 70
+2022-12-13 15:41:41,308 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:41:42,117 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634593072176935
+2022-12-13 15:43:26,470 INFO [train.py:421] (2/8) Epoch 10, batch 5200, loss[loss=2.466, over 910.00 frames. , ppl: 11.774734188456776] tot_loss[loss=2.262, over 5183980.19 frames. , ppl: 9.602657565722529], batch size: 70
+2022-12-13 15:45:06,994 INFO [train.py:421] (2/8) Epoch 10, batch 5400, loss[loss=2.516, over 1470.00 frames. , ppl: 12.376350431175789] tot_loss[loss=2.264, over 5155983.18 frames. , ppl: 9.62304931459162], batch size: 70
+2022-12-13 15:46:49,903 INFO [train.py:421] (2/8) Epoch 10, batch 5600, loss[loss=2.176, over 11970.00 frames. , ppl: 8.812365313991728] tot_loss[loss=2.264, over 5199898.51 frames. , ppl: 9.619968579659306], batch size: 70
+2022-12-13 15:48:31,065 INFO [train.py:421] (2/8) Epoch 10, batch 5800, loss[loss=2.728, over 770.00 frames. , ppl: 15.304912084163796] tot_loss[loss=2.263, over 5239275.86 frames. , ppl: 9.612459407397848], batch size: 70
+2022-12-13 15:50:14,249 INFO [train.py:421] (2/8) Epoch 10, batch 6000, loss[loss=2.256, over 2170.00 frames. , ppl: 9.54252922787431] tot_loss[loss=2.262, over 5286225.38 frames. , ppl: 9.601342217309536], batch size: 70
+2022-12-13 15:50:14,249 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:50:14,999 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622741820215266
+2022-12-13 15:51:54,506 INFO [train.py:421] (2/8) Epoch 10, batch 6200, loss[loss=2.175, over 4130.00 frames. , ppl: 8.802501489701974] tot_loss[loss=2.261, over 5324158.55 frames. , ppl: 9.596407525513955], batch size: 70
+2022-12-13 15:53:34,733 INFO [train.py:421] (2/8) Epoch 10, batch 6400, loss[loss=2.551, over 980.00 frames. , ppl: 12.825222931991998] tot_loss[loss=2.261, over 5364191.65 frames. , ppl: 9.595516198624841], batch size: 70
+2022-12-13 15:55:17,561 INFO [train.py:421] (2/8) Epoch 10, batch 6600, loss[loss=2.583, over 1190.00 frames. , ppl: 13.235730172062224] tot_loss[loss=2.262, over 5383215.01 frames. , ppl: 9.597786322522115], batch size: 70
+2022-12-13 15:56:56,905 INFO [train.py:421] (2/8) Epoch 10, batch 6800, loss[loss=2.18, over 5740.00 frames. , ppl: 8.848319036789663] tot_loss[loss=2.261, over 5436165.37 frames. , ppl: 9.588805997903656], batch size: 70
+2022-12-13 15:58:42,415 INFO [train.py:421] (2/8) Epoch 10, batch 7000, loss[loss=2.298, over 1890.00 frames. , ppl: 9.958162305799023] tot_loss[loss=2.261, over 5465000.85 frames. , ppl: 9.588089623786193], batch size: 70
+2022-12-13 15:58:42,415 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 15:58:43,176 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637175889263702
+2022-12-13 16:00:25,416 INFO [train.py:421] (2/8) Epoch 10, batch 7200, loss[loss=2.441, over 1190.00 frames. , ppl: 11.483007339823313] tot_loss[loss=2.26, over 5501295.37 frames. , ppl: 9.584482272605419], batch size: 70
+2022-12-13 16:02:05,016 INFO [train.py:421] (2/8) Epoch 10, batch 7400, loss[loss=2.187, over 3430.00 frames. , ppl: 8.906458510607184] tot_loss[loss=2.26, over 5506198.23 frames. , ppl: 9.585901101278216], batch size: 70
+2022-12-13 16:03:46,152 INFO [train.py:421] (2/8) Epoch 10, batch 7600, loss[loss=2.221, over 4620.00 frames. , ppl: 9.213775885551078] tot_loss[loss=2.26, over 5520474.05 frames. , ppl: 9.585202661481828], batch size: 70
+2022-12-13 16:05:27,150 INFO [train.py:421] (2/8) Epoch 10, batch 7800, loss[loss=2.933, over 630.00 frames. , ppl: 18.77831657430046] tot_loss[loss=2.261, over 5522579.48 frames. , ppl: 9.595772724600272], batch size: 70
+2022-12-13 16:07:06,113 INFO [train.py:421] (2/8) Epoch 10, batch 8000, loss[loss=2.245, over 2940.00 frames. , ppl: 9.443651595651868] tot_loss[loss=2.262, over 5520859.31 frames. , ppl: 9.60348025124673], batch size: 70
+2022-12-13 16:07:06,114 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:07:06,898 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.623687560464123
+2022-12-13 16:08:45,826 INFO [train.py:421] (2/8) Epoch 10, batch 8200, loss[loss=2.392, over 1260.00 frames. , ppl: 10.936872017028234] tot_loss[loss=2.263, over 5501583.58 frames. , ppl: 9.610944194081048], batch size: 70
+2022-12-13 16:10:28,728 INFO [train.py:421] (2/8) Epoch 10, batch 8400, loss[loss=2.376, over 1260.00 frames. , ppl: 10.765152428616144] tot_loss[loss=2.263, over 5502178.08 frames. , ppl: 9.616001226064745], batch size: 70
+2022-12-13 16:12:10,317 INFO [train.py:421] (2/8) Epoch 10, batch 8600, loss[loss=2.271, over 4060.00 frames. , ppl: 9.690093673689073] tot_loss[loss=2.263, over 5508457.05 frames. , ppl: 9.612593029284968], batch size: 70
+2022-12-13 16:13:50,371 INFO [train.py:421] (2/8) Epoch 10, batch 8800, loss[loss=2.256, over 2310.00 frames. , ppl: 9.547708263380969] tot_loss[loss=2.265, over 5487130.19 frames. , ppl: 9.628461568779185], batch size: 70
+2022-12-13 16:15:29,982 INFO [train.py:421] (2/8) Epoch 10, batch 9000, loss[loss=2.268, over 4130.00 frames. , ppl: 9.659311811692248] tot_loss[loss=2.265, over 5495580.51 frames. , ppl: 9.631657688163259], batch size: 70
+2022-12-13 16:15:29,983 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:15:30,747 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634809825352011
+2022-12-13 16:17:11,733 INFO [train.py:421] (2/8) Epoch 10, batch 9200, loss[loss=2.242, over 2730.00 frames. , ppl: 9.414056596387992] tot_loss[loss=2.266, over 5457211.68 frames. , ppl: 9.64275276501117], batch size: 70
+2022-12-13 16:18:52,527 INFO [train.py:421] (2/8) Epoch 10, batch 9400, loss[loss=2.139, over 5810.00 frames. , ppl: 8.494482363925389] tot_loss[loss=2.265, over 5531115.70 frames. , ppl: 9.629007273759262], batch size: 70
+2022-12-13 16:20:35,332 INFO [train.py:421] (2/8) Epoch 10, batch 9600, loss[loss=2.355, over 1960.00 frames. , ppl: 10.538610876181385] tot_loss[loss=2.264, over 5554116.67 frames. , ppl: 9.62207287807186], batch size: 70
+2022-12-13 16:22:23,363 INFO [train.py:421] (2/8) Epoch 10, batch 9800, loss[loss=2.226, over 3290.00 frames. , ppl: 9.261877742195308] tot_loss[loss=2.264, over 5558580.94 frames. , ppl: 9.62409019273281], batch size: 70
+2022-12-13 16:24:04,254 INFO [train.py:421] (2/8) Epoch 10, batch 10000, loss[loss=2.95, over 560.00 frames. , ppl: 19.10970238659096] tot_loss[loss=2.265, over 5535923.21 frames. , ppl: 9.631745800610178], batch size: 70
+2022-12-13 16:24:04,254 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:24:05,013 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621310567860883
+2022-12-13 16:25:47,065 INFO [train.py:421] (2/8) Epoch 10, batch 10200, loss[loss=2.179, over 3990.00 frames. , ppl: 8.8362595834563] tot_loss[loss=2.265, over 5558277.95 frames. , ppl: 9.628610100377033], batch size: 70
+2022-12-13 16:27:29,192 INFO [train.py:421] (2/8) Epoch 10, batch 10400, loss[loss=2.169, over 4480.00 frames. , ppl: 8.746839652596258] tot_loss[loss=2.265, over 5555941.16 frames. , ppl: 9.6279005103483], batch size: 70
+2022-12-13 16:29:09,746 INFO [train.py:421] (2/8) Epoch 10, batch 10600, loss[loss=2.278, over 2940.00 frames. , ppl: 9.75379379427965] tot_loss[loss=2.265, over 5564832.84 frames. , ppl: 9.628089817983588], batch size: 70
+2022-12-13 16:30:51,417 INFO [train.py:421] (2/8) Epoch 10, batch 10800, loss[loss=2.742, over 700.00 frames. , ppl: 15.522729746927784] tot_loss[loss=2.266, over 5540725.85 frames. , ppl: 9.64138171801174], batch size: 70
+2022-12-13 16:32:33,733 INFO [train.py:421] (2/8) Epoch 10, batch 11000, loss[loss=2.217, over 3710.00 frames. , ppl: 9.177784352600252] tot_loss[loss=2.267, over 5504755.53 frames. , ppl: 9.65276139821122], batch size: 70
+2022-12-13 16:32:33,734 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:32:34,520 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60629313197587
+2022-12-13 16:34:15,547 INFO [train.py:421] (2/8) Epoch 10, batch 11200, loss[loss=2.335, over 1470.00 frames. , ppl: 10.326914041719673] tot_loss[loss=2.267, over 5501842.19 frames. , ppl: 9.651349864617297], batch size: 70
+2022-12-13 16:35:59,433 INFO [train.py:421] (2/8) Epoch 10, batch 11400, loss[loss=2.513, over 980.00 frames. , ppl: 12.3470974116216] tot_loss[loss=2.267, over 5514087.18 frames. , ppl: 9.650354048084886], batch size: 70
+2022-12-13 16:37:41,285 INFO [train.py:421] (2/8) Epoch 10, batch 11600, loss[loss=2.431, over 1190.00 frames. , ppl: 11.368713122119916] tot_loss[loss=2.268, over 5477597.02 frames. , ppl: 9.65704786064711], batch size: 70
+2022-12-13 16:39:23,747 INFO [train.py:421] (2/8) Epoch 10, batch 11800, loss[loss=2.137, over 6230.00 frames. , ppl: 8.473241354927913] tot_loss[loss=2.267, over 5500186.10 frames. , ppl: 9.64903001080742], batch size: 70
+2022-12-13 16:41:07,319 INFO [train.py:421] (2/8) Epoch 10, batch 12000, loss[loss=2.251, over 5250.00 frames. , ppl: 9.49945272438952] tot_loss[loss=2.265, over 5539437.30 frames. , ppl: 9.634628824874829], batch size: 70
+2022-12-13 16:41:07,320 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:41:08,102 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.625352801927015
+2022-12-13 16:42:53,939 INFO [train.py:421] (2/8) Epoch 10, batch 12200, loss[loss=2.174, over 8050.00 frames. , ppl: 8.793535828763849] tot_loss[loss=2.265, over 5571501.69 frames. , ppl: 9.62753751987394], batch size: 70
+2022-12-13 16:44:34,130 INFO [train.py:421] (2/8) Epoch 10, batch 12400, loss[loss=2.212, over 3080.00 frames. , ppl: 9.135547343440122] tot_loss[loss=2.265, over 5548966.79 frames. , ppl: 9.627873139870436], batch size: 70
+2022-12-13 16:46:16,870 INFO [train.py:421] (2/8) Epoch 10, batch 12600, loss[loss=2.286, over 2240.00 frames. , ppl: 9.837676642471108] tot_loss[loss=2.266, over 5503062.23 frames. , ppl: 9.641604499135818], batch size: 70
+2022-12-13 16:47:56,481 INFO [train.py:421] (2/8) Epoch 10, batch 12800, loss[loss=2.424, over 1190.00 frames. , ppl: 11.295703973186273] tot_loss[loss=2.267, over 5486220.81 frames. , ppl: 9.649662033769193], batch size: 70
+2022-12-13 16:49:37,760 INFO [train.py:421] (2/8) Epoch 10, batch 13000, loss[loss=2.202, over 5040.00 frames. , ppl: 9.040438031464372] tot_loss[loss=2.266, over 5520931.13 frames. , ppl: 9.638286393341916], batch size: 70
+2022-12-13 16:49:37,761 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:49:38,513 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634478993517094
+2022-12-13 16:51:20,401 INFO [train.py:421] (2/8) Epoch 10, batch 13200, loss[loss=2.184, over 4830.00 frames. , ppl: 8.882803456693036] tot_loss[loss=2.266, over 5529288.31 frames. , ppl: 9.642602871790375], batch size: 70
+2022-12-13 16:53:01,609 INFO [train.py:421] (2/8) Epoch 10, batch 13400, loss[loss=2.25, over 1610.00 frames. , ppl: 9.489478288185506] tot_loss[loss=2.267, over 5506154.84 frames. , ppl: 9.647405596766378], batch size: 70
+2022-12-13 16:54:46,215 INFO [train.py:421] (2/8) Epoch 10, batch 13600, loss[loss=2.487, over 1050.00 frames. , ppl: 12.022240116755487] tot_loss[loss=2.268, over 5466268.68 frames. , ppl: 9.661799975102355], batch size: 70
+2022-12-13 16:56:30,070 INFO [train.py:421] (2/8) Epoch 10, batch 13800, loss[loss=2.301, over 1050.00 frames. , ppl: 9.981198017973153] tot_loss[loss=2.268, over 5465021.65 frames. , ppl: 9.662769869228562], batch size: 70
+2022-12-13 16:58:09,138 INFO [train.py:421] (2/8) Epoch 10, batch 14000, loss[loss=2.247, over 4270.00 frames. , ppl: 9.454882352523452] tot_loss[loss=2.269, over 5439797.04 frames. , ppl: 9.669557403698523], batch size: 70
+2022-12-13 16:58:09,139 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 16:58:09,934 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.641104923874973
+2022-12-13 16:59:47,673 INFO [train.py:421] (2/8) Epoch 10, batch 14200, loss[loss=2.244, over 4830.00 frames. , ppl: 9.42949555501388] tot_loss[loss=2.271, over 5376384.30 frames. , ppl: 9.692979276900067], batch size: 70
+2022-12-13 17:01:29,029 INFO [train.py:421] (2/8) Epoch 10, batch 14400, loss[loss=2.295, over 1820.00 frames. , ppl: 9.924257087768227] tot_loss[loss=2.271, over 5368379.31 frames. , ppl: 9.693226918213895], batch size: 70
+2022-12-13 17:03:09,990 INFO [train.py:421] (2/8) Epoch 10, batch 14600, loss[loss=2.238, over 2730.00 frames. , ppl: 9.374034377795398] tot_loss[loss=2.271, over 5391552.39 frames. , ppl: 9.689531150549863], batch size: 70
+2022-12-13 17:04:50,614 INFO [train.py:421] (2/8) Epoch 10, batch 14800, loss[loss=3.154, over 490.00 frames. , ppl: 23.42244485544623] tot_loss[loss=2.271, over 5394470.65 frames. , ppl: 9.689474714062012], batch size: 70
+2022-12-13 17:06:32,745 INFO [train.py:421] (2/8) Epoch 10, batch 15000, loss[loss=2.174, over 3710.00 frames. , ppl: 8.791394755154268] tot_loss[loss=2.271, over 5391834.60 frames. , ppl: 9.690848753226666], batch size: 70
+2022-12-13 17:06:32,745 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:06:33,530 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634267951557288
+2022-12-13 17:08:12,525 INFO [train.py:421] (2/8) Epoch 10, batch 15200, loss[loss=2.177, over 6020.00 frames. , ppl: 8.819964549275205] tot_loss[loss=2.272, over 5356739.93 frames. , ppl: 9.698304799241935], batch size: 70
+2022-12-13 17:09:58,749 INFO [train.py:421] (2/8) Epoch 10, batch 15400, loss[loss=2.413, over 1610.00 frames. , ppl: 11.171828346914465] tot_loss[loss=2.271, over 5389178.65 frames. , ppl: 9.684915728847198], batch size: 70
+2022-12-13 17:11:43,376 INFO [train.py:421] (2/8) Epoch 10, batch 15600, loss[loss=2.619, over 770.00 frames. , ppl: 13.723826612415456] tot_loss[loss=2.269, over 5430786.82 frames. , ppl: 9.672147866952827], batch size: 70
+2022-12-13 17:13:24,531 INFO [train.py:421] (2/8) Epoch 10, batch 15800, loss[loss=2.212, over 3570.00 frames. , ppl: 9.138426702854058] tot_loss[loss=2.269, over 5412817.05 frames. , ppl: 9.67327774841489], batch size: 70
+2022-12-13 17:15:07,631 INFO [train.py:421] (2/8) Epoch 10, batch 16000, loss[loss=2.208, over 3990.00 frames. , ppl: 9.100480501766011] tot_loss[loss=2.271, over 5378994.08 frames. , ppl: 9.686443558773831], batch size: 70
+2022-12-13 17:15:07,632 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:15:08,400 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61156377960389
+2022-12-13 17:16:51,483 INFO [train.py:421] (2/8) Epoch 10, batch 16200, loss[loss=2.225, over 2450.00 frames. , ppl: 9.250958431707595] tot_loss[loss=2.27, over 5401925.60 frames. , ppl: 9.681626669258517], batch size: 70
+2022-12-13 17:18:33,685 INFO [train.py:421] (2/8) Epoch 10, batch 16400, loss[loss=2.266, over 3990.00 frames. , ppl: 9.637089326448315] tot_loss[loss=2.269, over 5446645.60 frames. , ppl: 9.673693711529694], batch size: 70
+2022-12-13 17:20:17,176 INFO [train.py:421] (2/8) Epoch 10, batch 16600, loss[loss=2.215, over 7000.00 frames. , ppl: 9.160276786435464] tot_loss[loss=2.269, over 5444875.06 frames. , ppl: 9.669750445107363], batch size: 70
+2022-12-13 17:21:59,443 INFO [train.py:421] (2/8) Epoch 10, batch 16800, loss[loss=2.32, over 1890.00 frames. , ppl: 10.175699543392097] tot_loss[loss=2.269, over 5438360.04 frames. , ppl: 9.6733647543338], batch size: 70
+2022-12-13 17:23:43,885 INFO [train.py:421] (2/8) Epoch 10, batch 17000, loss[loss=2.271, over 5320.00 frames. , ppl: 9.68710343290128] tot_loss[loss=2.271, over 5410565.51 frames. , ppl: 9.686437495152648], batch size: 70
+2022-12-13 17:23:43,886 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:23:44,636 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.629349702040027
+2022-12-13 17:25:25,465 INFO [train.py:421] (2/8) Epoch 10, batch 17200, loss[loss=2.35, over 2030.00 frames. , ppl: 10.490549575154768] tot_loss[loss=2.271, over 5405596.55 frames. , ppl: 9.689853092216202], batch size: 70
+2022-12-13 17:27:07,627 INFO [train.py:421] (2/8) Epoch 10, batch 17400, loss[loss=2.232, over 4550.00 frames. , ppl: 9.31534412457966] tot_loss[loss=2.272, over 5385240.53 frames. , ppl: 9.695599219343157], batch size: 70
+2022-12-13 17:28:48,475 INFO [train.py:421] (2/8) Epoch 10, batch 17600, loss[loss=2.801, over 700.00 frames. , ppl: 16.46922426300079] tot_loss[loss=2.271, over 5430806.84 frames. , ppl: 9.684360736071111], batch size: 70
+2022-12-13 17:30:34,236 INFO [train.py:421] (2/8) Epoch 10, batch 17800, loss[loss=2.25, over 2170.00 frames. , ppl: 9.489025389813278] tot_loss[loss=2.27, over 5455609.85 frames. , ppl: 9.676177577317572], batch size: 70
+2022-12-13 17:32:15,320 INFO [train.py:421] (2/8) Epoch 10, batch 18000, loss[loss=2.234, over 2240.00 frames. , ppl: 9.336656296135601] tot_loss[loss=2.27, over 5446266.03 frames. , ppl: 9.678380345695754], batch size: 70
+2022-12-13 17:32:15,321 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:32:16,072 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622539580598964
+2022-12-13 17:33:56,551 INFO [train.py:421] (2/8) Epoch 10, batch 18200, loss[loss=2.334, over 1470.00 frames. , ppl: 10.317445803963807] tot_loss[loss=2.27, over 5429000.16 frames. , ppl: 9.681418435437195], batch size: 70
+2022-12-13 17:35:35,468 INFO [train.py:421] (2/8) Epoch 10, batch 18400, loss[loss=2.316, over 1960.00 frames. , ppl: 10.137362723247042] tot_loss[loss=2.271, over 5406024.55 frames. , ppl: 9.68538608000572], batch size: 70
+2022-12-13 17:37:19,302 INFO [train.py:421] (2/8) Epoch 10, batch 18600, loss[loss=2.464, over 1540.00 frames. , ppl: 11.752660277388557] tot_loss[loss=2.269, over 5454993.31 frames. , ppl: 9.671587957111376], batch size: 70
+2022-12-13 17:39:02,830 INFO [train.py:421] (2/8) Epoch 10, batch 18800, loss[loss=2.67, over 840.00 frames. , ppl: 14.435626909590818] tot_loss[loss=2.269, over 5468695.46 frames. , ppl: 9.666632732733692], batch size: 70
+2022-12-13 17:40:48,047 INFO [train.py:421] (2/8) Epoch 10, batch 19000, loss[loss=2.583, over 1050.00 frames. , ppl: 13.23017228050329] tot_loss[loss=2.269, over 5486632.85 frames. , ppl: 9.665481923410711], batch size: 70
+2022-12-13 17:40:48,048 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:40:48,813 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.63137799747174
+2022-12-13 17:42:30,791 INFO [train.py:421] (2/8) Epoch 10, batch 19200, loss[loss=2.215, over 3640.00 frames. , ppl: 9.16004164841985] tot_loss[loss=2.268, over 5505301.13 frames. , ppl: 9.661085136467026], batch size: 70
+2022-12-13 17:44:15,229 INFO [train.py:421] (2/8) Epoch 10, batch 19400, loss[loss=2.233, over 5600.00 frames. , ppl: 9.331339348754431] tot_loss[loss=2.269, over 5484245.19 frames. , ppl: 9.671366985129826], batch size: 70
+2022-12-13 17:45:57,147 INFO [train.py:421] (2/8) Epoch 10, batch 19600, loss[loss=2.562, over 1190.00 frames. , ppl: 12.95763920275099] tot_loss[loss=2.271, over 5430346.73 frames. , ppl: 9.686049819658278], batch size: 70
+2022-12-13 17:47:36,296 INFO [train.py:421] (2/8) Epoch 10, batch 19800, loss[loss=2.274, over 3430.00 frames. , ppl: 9.72119384326381] tot_loss[loss=2.272, over 5395732.15 frames. , ppl: 9.69461709000022], batch size: 70
+2022-12-13 17:49:20,635 INFO [train.py:421] (2/8) Epoch 10, batch 20000, loss[loss=2.143, over 8820.00 frames. , ppl: 8.527426597645556] tot_loss[loss=2.271, over 5423431.31 frames. , ppl: 9.689856215293014], batch size: 70
+2022-12-13 17:49:20,636 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:49:21,384 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.641742794080622
+2022-12-13 17:51:06,008 INFO [train.py:421] (2/8) Epoch 10, batch 20200, loss[loss=2.839, over 700.00 frames. , ppl: 17.09391320103847] tot_loss[loss=2.271, over 5448716.17 frames. , ppl: 9.688407579387988], batch size: 70
+2022-12-13 17:52:43,975 INFO [train.py:421] (2/8) Epoch 10, batch 20400, loss[loss=2.362, over 1750.00 frames. , ppl: 10.6119265583111] tot_loss[loss=2.271, over 5453945.05 frames. , ppl: 9.688324970504818], batch size: 70
+2022-12-13 17:54:27,188 INFO [train.py:421] (2/8) Epoch 10, batch 20600, loss[loss=2.144, over 4900.00 frames. , ppl: 8.533055509297178] tot_loss[loss=2.27, over 5469475.24 frames. , ppl: 9.683540321411076], batch size: 70
+2022-12-13 17:56:10,934 INFO [train.py:421] (2/8) Epoch 10, batch 20800, loss[loss=2.296, over 3710.00 frames. , ppl: 9.93123882177216] tot_loss[loss=2.268, over 5514597.44 frames. , ppl: 9.664239394761076], batch size: 70
+2022-12-13 17:57:56,900 INFO [train.py:421] (2/8) Epoch 10, batch 21000, loss[loss=2.194, over 4690.00 frames. , ppl: 8.974841417498526] tot_loss[loss=2.268, over 5520632.76 frames. , ppl: 9.66373747401754], batch size: 70
+2022-12-13 17:57:56,901 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 17:57:57,648 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622576610069983
+2022-12-13 17:59:39,846 INFO [train.py:421] (2/8) Epoch 10, batch 21200, loss[loss=2.366, over 840.00 frames. , ppl: 10.65491659641141] tot_loss[loss=2.268, over 5544466.00 frames. , ppl: 9.65692432362664], batch size: 70
+2022-12-13 18:01:19,796 INFO [train.py:421] (2/8) Epoch 10, batch 21400, loss[loss=2.166, over 6020.00 frames. , ppl: 8.723812381940881] tot_loss[loss=2.268, over 5546084.40 frames. , ppl: 9.656476968867567], batch size: 70
+2022-12-13 18:03:00,324 INFO [train.py:421] (2/8) Epoch 10, batch 21600, loss[loss=2.226, over 5460.00 frames. , ppl: 9.2605310590801] tot_loss[loss=2.267, over 5588220.08 frames. , ppl: 9.647978490440794], batch size: 70
+2022-12-13 18:04:42,565 INFO [train.py:421] (2/8) Epoch 10, batch 21800, loss[loss=2.235, over 2800.00 frames. , ppl: 9.343258471721414] tot_loss[loss=2.267, over 5571650.55 frames. , ppl: 9.651612914325838], batch size: 70
+2022-12-13 18:06:24,589 INFO [train.py:421] (2/8) Epoch 10, batch 22000, loss[loss=2.419, over 1330.00 frames. , ppl: 11.237424286967293] tot_loss[loss=2.267, over 5553146.61 frames. , ppl: 9.651183911037785], batch size: 70
+2022-12-13 18:06:24,589 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:06:25,356 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.620783692748795
+2022-12-13 18:08:04,766 INFO [train.py:421] (2/8) Epoch 10, batch 22200, loss[loss=2.274, over 3430.00 frames. , ppl: 9.718694893932174] tot_loss[loss=2.268, over 5517428.35 frames. , ppl: 9.657374559447986], batch size: 70
+2022-12-13 18:09:43,513 INFO [train.py:421] (2/8) Epoch 10, batch 22400, loss[loss=2.403, over 2170.00 frames. , ppl: 11.05404809139365] tot_loss[loss=2.268, over 5497296.73 frames. , ppl: 9.662642365521972], batch size: 70
+2022-12-13 18:11:22,900 INFO [train.py:421] (2/8) Epoch 10, batch 22600, loss[loss=3.02, over 560.00 frames. , ppl: 20.493769091885852] tot_loss[loss=2.268, over 5492297.44 frames. , ppl: 9.655734790269916], batch size: 70
+2022-12-13 18:13:02,879 INFO [train.py:421] (2/8) Epoch 10, batch 22800, loss[loss=2.569, over 840.00 frames. , ppl: 13.0490548264667] tot_loss[loss=2.268, over 5487726.92 frames. , ppl: 9.659569242379252], batch size: 70
+2022-12-13 18:14:48,430 INFO [train.py:421] (2/8) Epoch 10, batch 23000, loss[loss=2.368, over 1540.00 frames. , ppl: 10.67916380499969] tot_loss[loss=2.268, over 5467951.50 frames. , ppl: 9.660132693694884], batch size: 70
+2022-12-13 18:14:48,430 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:14:49,200 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.624770148321595
+2022-12-13 18:16:34,754 INFO [train.py:421] (2/8) Epoch 10, batch 23200, loss[loss=2.685, over 770.00 frames. , ppl: 14.657099007778816] tot_loss[loss=2.268, over 5462888.47 frames. , ppl: 9.663187000153853], batch size: 70
+2022-12-13 18:18:16,613 INFO [train.py:421] (2/8) Epoch 10, batch 23400, loss[loss=2.383, over 2100.00 frames. , ppl: 10.832246112183807] tot_loss[loss=2.268, over 5456278.57 frames. , ppl: 9.662983492639595], batch size: 70
+2022-12-13 18:19:54,568 INFO [train.py:421] (2/8) Epoch 10, batch 23600, loss[loss=2.356, over 1960.00 frames. , ppl: 10.552290263093552] tot_loss[loss=2.27, over 5417740.37 frames. , ppl: 9.677029010007631], batch size: 70
+2022-12-13 18:21:36,245 INFO [train.py:421] (2/8) Epoch 10, batch 23800, loss[loss=2.633, over 770.00 frames. , ppl: 13.919119067067962] tot_loss[loss=2.269, over 5441442.28 frames. , ppl: 9.666027924835097], batch size: 70
+2022-12-13 18:23:16,493 INFO [train.py:421] (2/8) Epoch 10, batch 24000, loss[loss=2.541, over 980.00 frames. , ppl: 12.687162062940693] tot_loss[loss=2.269, over 5454703.34 frames. , ppl: 9.670448356294541], batch size: 70
+2022-12-13 18:23:16,494 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:23:17,301 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.631834173341721
+2022-12-13 18:24:57,935 INFO [train.py:421] (2/8) Epoch 10, batch 24200, loss[loss=2.106, over 8120.00 frames. , ppl: 8.212562750472445] tot_loss[loss=2.269, over 5461267.70 frames. , ppl: 9.6704086024433], batch size: 70
+2022-12-13 18:26:41,699 INFO [train.py:421] (2/8) Epoch 10, batch 24400, loss[loss=2.314, over 2590.00 frames. , ppl: 10.114968042855018] tot_loss[loss=2.269, over 5470806.07 frames. , ppl: 9.6739140012992], batch size: 70
+2022-12-13 18:28:25,417 INFO [train.py:421] (2/8) Epoch 10, batch 24600, loss[loss=2.344, over 1610.00 frames. , ppl: 10.418889943544487] tot_loss[loss=2.27, over 5435775.71 frames. , ppl: 9.680524135366195], batch size: 70
+2022-12-13 18:30:06,363 INFO [train.py:421] (2/8) Epoch 10, batch 24800, loss[loss=2.145, over 5880.00 frames. , ppl: 8.542415209471915] tot_loss[loss=2.27, over 5436637.12 frames. , ppl: 9.678391218850866], batch size: 70
+2022-12-13 18:31:48,361 INFO [train.py:421] (2/8) Epoch 10, batch 25000, loss[loss=2.606, over 1120.00 frames. , ppl: 13.54495485703612] tot_loss[loss=2.269, over 5465558.86 frames. , ppl: 9.666132627684174], batch size: 70
+2022-12-13 18:31:48,362 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:31:49,120 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617161851458512
+2022-12-13 18:33:31,039 INFO [train.py:421] (2/8) Epoch 10, batch 25200, loss[loss=2.447, over 1260.00 frames. , ppl: 11.5561970556718] tot_loss[loss=2.268, over 5503114.96 frames. , ppl: 9.659078501589999], batch size: 70
+2022-12-13 18:35:13,476 INFO [train.py:421] (2/8) Epoch 10, batch 25400, loss[loss=2.278, over 2100.00 frames. , ppl: 9.760313275093898] tot_loss[loss=2.268, over 5512696.14 frames. , ppl: 9.657731516363366], batch size: 70
+2022-12-13 18:36:55,993 INFO [train.py:421] (2/8) Epoch 10, batch 25600, loss[loss=2.235, over 5530.00 frames. , ppl: 9.342037705221037] tot_loss[loss=2.269, over 5478797.39 frames. , ppl: 9.670275721759342], batch size: 70
+2022-12-13 18:38:38,947 INFO [train.py:421] (2/8) Epoch 10, batch 25800, loss[loss=2.152, over 6020.00 frames. , ppl: 8.601923950180637] tot_loss[loss=2.269, over 5490382.22 frames. , ppl: 9.670544598390874], batch size: 70
+2022-12-13 18:40:24,876 INFO [train.py:421] (2/8) Epoch 10, batch 26000, loss[loss=2.295, over 3150.00 frames. , ppl: 9.92572218634838] tot_loss[loss=2.269, over 5535736.72 frames. , ppl: 9.665857491000951], batch size: 70
+2022-12-13 18:40:24,877 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:40:25,674 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637863426039562
+2022-12-13 18:42:06,995 INFO [train.py:421] (2/8) Epoch 10, batch 26200, loss[loss=2.163, over 5460.00 frames. , ppl: 8.696462906275288] tot_loss[loss=2.269, over 5502025.81 frames. , ppl: 9.673012830107634], batch size: 70
+2022-12-13 18:43:50,151 INFO [train.py:421] (2/8) Epoch 10, batch 26400, loss[loss=2.204, over 3640.00 frames. , ppl: 9.05894125183826] tot_loss[loss=2.269, over 5488540.73 frames. , ppl: 9.673834678472497], batch size: 70
+2022-12-13 18:45:37,800 INFO [train.py:421] (2/8) Epoch 10, batch 26600, loss[loss=2.263, over 7070.00 frames. , ppl: 9.608013771585417] tot_loss[loss=2.269, over 5506506.76 frames. , ppl: 9.670367701215838], batch size: 70
+2022-12-13 18:47:19,988 INFO [train.py:421] (2/8) Epoch 10, batch 26800, loss[loss=2.143, over 7700.00 frames. , ppl: 8.521071589357396] tot_loss[loss=2.27, over 5483343.34 frames. , ppl: 9.682797324689663], batch size: 70
+2022-12-13 18:49:02,367 INFO [train.py:421] (2/8) Epoch 10, batch 27000, loss[loss=2.254, over 2380.00 frames. , ppl: 9.523607503983076] tot_loss[loss=2.271, over 5462040.64 frames. , ppl: 9.68550201551465], batch size: 70
+2022-12-13 18:49:02,367 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:49:03,131 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621924342548795
+2022-12-13 18:50:45,266 INFO [train.py:421] (2/8) Epoch 10, batch 27200, loss[loss=2.235, over 3850.00 frames. , ppl: 9.350361219438192] tot_loss[loss=2.272, over 5426456.27 frames. , ppl: 9.697912101439613], batch size: 70
+2022-12-13 18:52:26,329 INFO [train.py:421] (2/8) Epoch 10, batch 27400, loss[loss=2.184, over 4620.00 frames. , ppl: 8.88070933890293] tot_loss[loss=2.272, over 5426490.25 frames. , ppl: 9.694771932204366], batch size: 70
+2022-12-13 18:54:07,431 INFO [train.py:421] (2/8) Epoch 10, batch 27600, loss[loss=2.406, over 1890.00 frames. , ppl: 11.087794485379545] tot_loss[loss=2.27, over 5445125.58 frames. , ppl: 9.683915523170795], batch size: 70
+2022-12-13 18:55:51,741 INFO [train.py:421] (2/8) Epoch 10, batch 27800, loss[loss=2.195, over 7420.00 frames. , ppl: 8.984463044052195] tot_loss[loss=2.27, over 5491478.27 frames. , ppl: 9.679013142094774], batch size: 70
+2022-12-13 18:57:33,170 INFO [train.py:421] (2/8) Epoch 10, batch 28000, loss[loss=2.262, over 1960.00 frames. , ppl: 9.600674405890235] tot_loss[loss=2.272, over 5452650.42 frames. , ppl: 9.695459471178665], batch size: 70
+2022-12-13 18:57:33,171 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 18:57:33,932 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617768243976228
+2022-12-13 18:59:13,885 INFO [train.py:421] (2/8) Epoch 10, batch 28200, loss[loss=2.253, over 3780.00 frames. , ppl: 9.514680356079458] tot_loss[loss=2.273, over 5393799.48 frames. , ppl: 9.707850617124125], batch size: 70
+2022-12-13 19:00:57,579 INFO [train.py:421] (2/8) Epoch 10, batch 28400, loss[loss=2.198, over 4480.00 frames. , ppl: 9.009304550655296] tot_loss[loss=2.271, over 5434937.13 frames. , ppl: 9.691196806772066], batch size: 70
+2022-12-13 19:02:36,590 INFO [train.py:421] (2/8) Epoch 10, batch 28600, loss[loss=2.287, over 2240.00 frames. , ppl: 9.844996204746469] tot_loss[loss=2.271, over 5429098.92 frames. , ppl: 9.690842294690265], batch size: 70
+2022-12-13 19:04:16,654 INFO [train.py:421] (2/8) Epoch 10, batch 28800, loss[loss=2.2, over 5040.00 frames. , ppl: 9.020974887191713] tot_loss[loss=2.272, over 5411049.35 frames. , ppl: 9.697106079351826], batch size: 70
+2022-12-13 19:06:00,378 INFO [train.py:421] (2/8) Epoch 10, batch 29000, loss[loss=2.198, over 3710.00 frames. , ppl: 9.010324043452636] tot_loss[loss=2.271, over 5447430.87 frames. , ppl: 9.689701287330553], batch size: 70
+2022-12-13 19:06:00,378 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:06:01,145 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.614629935321982
+2022-12-13 19:07:42,003 INFO [train.py:421] (2/8) Epoch 10, batch 29200, loss[loss=2.338, over 1470.00 frames. , ppl: 10.356691842815744] tot_loss[loss=2.271, over 5452397.88 frames. , ppl: 9.6895824483715], batch size: 70
+2022-12-13 19:09:24,501 INFO [train.py:421] (2/8) Epoch 10, batch 29400, loss[loss=2.515, over 1120.00 frames. , ppl: 12.365234584151342] tot_loss[loss=2.271, over 5452079.30 frames. , ppl: 9.688303874307657], batch size: 70
+2022-12-13 19:11:08,716 INFO [train.py:421] (2/8) Epoch 10, batch 29600, loss[loss=2.179, over 4130.00 frames. , ppl: 8.835708222973238] tot_loss[loss=2.271, over 5466787.02 frames. , ppl: 9.689776191994756], batch size: 70
+2022-12-13 19:12:55,138 INFO [train.py:421] (2/8) Epoch 10, batch 29800, loss[loss=2.399, over 1890.00 frames. , ppl: 11.006708266349532] tot_loss[loss=2.27, over 5519303.83 frames. , ppl: 9.675052030701993], batch size: 70
+2022-12-13 19:14:36,432 INFO [train.py:421] (2/8) Epoch 10, batch 30000, loss[loss=2.193, over 3780.00 frames. , ppl: 8.962024365219932] tot_loss[loss=2.269, over 5534515.14 frames. , ppl: 9.67224763553603], batch size: 70
+2022-12-13 19:14:36,433 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:14:37,218 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.638813505699115
+2022-12-13 19:16:16,747 INFO [train.py:421] (2/8) Epoch 10, batch 30200, loss[loss=2.547, over 1890.00 frames. , ppl: 12.764096704338987] tot_loss[loss=2.268, over 5578251.17 frames. , ppl: 9.65569386335523], batch size: 70
+2022-12-13 19:18:02,135 INFO [train.py:421] (2/8) Epoch 10, batch 30400, loss[loss=2.398, over 1330.00 frames. , ppl: 11.003113817910283] tot_loss[loss=2.268, over 5574943.38 frames. , ppl: 9.65790098245716], batch size: 70
+2022-12-13 19:19:43,287 INFO [train.py:421] (2/8) Epoch 10, batch 30600, loss[loss=2.221, over 10570.00 frames. , ppl: 9.220817316083886] tot_loss[loss=2.268, over 5559318.05 frames. , ppl: 9.660797377384709], batch size: 70
+2022-12-13 19:21:25,050 INFO [train.py:421] (2/8) Epoch 10, batch 30800, loss[loss=2.177, over 5600.00 frames. , ppl: 8.821897877003476] tot_loss[loss=2.268, over 5549260.83 frames. , ppl: 9.660713997384114], batch size: 70
+2022-12-13 19:23:07,025 INFO [train.py:421] (2/8) Epoch 10, batch 31000, loss[loss=2.209, over 3220.00 frames. , ppl: 9.109704504962636] tot_loss[loss=2.268, over 5538954.60 frames. , ppl: 9.664755725166696], batch size: 70
+2022-12-13 19:23:07,026 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:23:07,797 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.624022295019808
+2022-12-13 19:24:49,277 INFO [train.py:421] (2/8) Epoch 10, batch 31200, loss[loss=2.209, over 3920.00 frames. , ppl: 9.109961843098876] tot_loss[loss=2.269, over 5533599.70 frames. , ppl: 9.666204880983571], batch size: 70
+2022-12-13 19:26:30,755 INFO [train.py:421] (2/8) Epoch 10, batch 31400, loss[loss=2.457, over 980.00 frames. , ppl: 11.669364351772044] tot_loss[loss=2.27, over 5457557.32 frames. , ppl: 9.683918702788505], batch size: 70
+2022-12-13 19:28:10,486 INFO [train.py:421] (2/8) Epoch 10, batch 31600, loss[loss=2.309, over 2100.00 frames. , ppl: 10.059673322383572] tot_loss[loss=2.269, over 5488806.63 frames. , ppl: 9.671869282927414], batch size: 70
+2022-12-13 19:29:51,153 INFO [train.py:421] (2/8) Epoch 10, batch 31800, loss[loss=2.209, over 4480.00 frames. , ppl: 9.104944505688023] tot_loss[loss=2.27, over 5484671.83 frames. , ppl: 9.676168948963811], batch size: 70
+2022-12-13 19:31:35,513 INFO [train.py:421] (2/8) Epoch 10, batch 32000, loss[loss=2.66, over 770.00 frames. , ppl: 14.301601695072982] tot_loss[loss=2.268, over 5556030.09 frames. , ppl: 9.657404511007906], batch size: 70
+2022-12-13 19:31:35,513 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:31:36,277 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616229561960207
+2022-12-13 19:33:19,150 INFO [train.py:421] (2/8) Epoch 10, batch 32200, loss[loss=2.251, over 2170.00 frames. , ppl: 9.49267295654991] tot_loss[loss=2.268, over 5573263.97 frames. , ppl: 9.656048447239819], batch size: 70
+2022-12-13 19:35:01,703 INFO [train.py:421] (2/8) Epoch 10, batch 32400, loss[loss=2.203, over 8680.00 frames. , ppl: 9.04965173346733] tot_loss[loss=2.269, over 5539532.17 frames. , ppl: 9.67124799706491], batch size: 70
+2022-12-13 19:36:43,496 INFO [train.py:421] (2/8) Epoch 10, batch 32600, loss[loss=2.338, over 2030.00 frames. , ppl: 10.358380012366094] tot_loss[loss=2.271, over 5489390.54 frames. , ppl: 9.690438674904243], batch size: 70
+2022-12-13 19:38:26,633 INFO [train.py:421] (2/8) Epoch 10, batch 32800, loss[loss=3.381, over 420.00 frames. , ppl: 29.404635845973644] tot_loss[loss=2.273, over 5451528.60 frames. , ppl: 9.705366020560414], batch size: 70
+2022-12-13 19:40:07,395 INFO [train.py:421] (2/8) Epoch 10, batch 33000, loss[loss=3.707, over 420.00 frames. , ppl: 40.73206235813763] tot_loss[loss=2.272, over 5498703.01 frames. , ppl: 9.697612341165998], batch size: 70
+2022-12-13 19:40:07,395 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:40:08,160 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.611626373467763
+2022-12-13 19:41:48,437 INFO [train.py:421] (2/8) Epoch 10, batch 33200, loss[loss=2.621, over 840.00 frames. , ppl: 13.75301349860503] tot_loss[loss=2.274, over 5441610.83 frames. , ppl: 9.71776934135376], batch size: 70
+2022-12-13 19:43:28,483 INFO [train.py:421] (2/8) Epoch 10, batch 33400, loss[loss=2.298, over 1890.00 frames. , ppl: 9.958738605015032] tot_loss[loss=2.272, over 5471930.98 frames. , ppl: 9.70352008873111], batch size: 70
+2022-12-13 19:45:10,998 INFO [train.py:421] (2/8) Epoch 10, batch 33600, loss[loss=2.342, over 2240.00 frames. , ppl: 10.405038602689478] tot_loss[loss=2.272, over 5483784.58 frames. , ppl: 9.702007662788402], batch size: 70
+2022-12-13 19:46:54,764 INFO [train.py:421] (2/8) Epoch 10, batch 33800, loss[loss=2.398, over 1540.00 frames. , ppl: 10.996180355061416] tot_loss[loss=2.271, over 5528131.21 frames. , ppl: 9.687774905004703], batch size: 70
+2022-12-13 19:48:42,877 INFO [train.py:421] (2/8) Epoch 10, batch 34000, loss[loss=2.293, over 3080.00 frames. , ppl: 9.908619475851749] tot_loss[loss=2.271, over 5513139.58 frames. , ppl: 9.693362324940164], batch size: 70
+2022-12-13 19:48:42,877 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:48:43,659 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612292169791925
+2022-12-13 19:50:27,173 INFO [train.py:421] (2/8) Epoch 10, batch 34200, loss[loss=2.186, over 4690.00 frames. , ppl: 8.903463486380376] tot_loss[loss=2.27, over 5572352.44 frames. , ppl: 9.675561237020712], batch size: 70
+2022-12-13 19:52:09,002 INFO [train.py:421] (2/8) Epoch 10, batch 34400, loss[loss=2.2, over 5390.00 frames. , ppl: 9.028336747717242] tot_loss[loss=2.269, over 5587759.25 frames. , ppl: 9.667210310491313], batch size: 70
+2022-12-13 19:53:50,199 INFO [train.py:421] (2/8) Epoch 10, batch 34600, loss[loss=2.219, over 2520.00 frames. , ppl: 9.196520259201776] tot_loss[loss=2.27, over 5562849.90 frames. , ppl: 9.677719054512085], batch size: 70
+2022-12-13 19:55:28,935 INFO [train.py:421] (2/8) Epoch 10, batch 34800, loss[loss=2.383, over 2310.00 frames. , ppl: 10.840759315629066] tot_loss[loss=2.27, over 5512808.36 frames. , ppl: 9.683375721473599], batch size: 70
+2022-12-13 19:57:12,629 INFO [train.py:421] (2/8) Epoch 10, batch 35000, loss[loss=2.08, over 8610.00 frames. , ppl: 8.004623982047027] tot_loss[loss=2.269, over 5532392.58 frames. , ppl: 9.674000491900905], batch size: 70
+2022-12-13 19:57:12,629 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 19:57:13,388 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616518490872702
+2022-12-13 19:58:57,909 INFO [train.py:421] (2/8) Epoch 10, batch 35200, loss[loss=2.281, over 2380.00 frames. , ppl: 9.782488482217529] tot_loss[loss=2.27, over 5476801.38 frames. , ppl: 9.683994190771294], batch size: 70
+2022-12-13 20:00:39,613 INFO [train.py:421] (2/8) Epoch 10, batch 35400, loss[loss=2.348, over 2310.00 frames. , ppl: 10.469019567756602] tot_loss[loss=2.272, over 5436592.30 frames. , ppl: 9.696508203711765], batch size: 70
+2022-12-13 20:02:22,788 INFO [train.py:421] (2/8) Epoch 10, batch 35600, loss[loss=2.407, over 3290.00 frames. , ppl: 11.101796488288608] tot_loss[loss=2.272, over 5416350.38 frames. , ppl: 9.703377870005363], batch size: 70
+2022-12-13 20:04:05,804 INFO [train.py:421] (2/8) Epoch 10, batch 35800, loss[loss=2.374, over 1610.00 frames. , ppl: 10.741960765350761] tot_loss[loss=2.271, over 5424546.54 frames. , ppl: 9.69359764078827], batch size: 70
+2022-12-13 20:05:47,822 INFO [train.py:421] (2/8) Epoch 10, batch 36000, loss[loss=2.207, over 3570.00 frames. , ppl: 9.08906468986311] tot_loss[loss=2.269, over 5491545.20 frames. , ppl: 9.67348180309415], batch size: 70
+2022-12-13 20:05:47,822 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:05:48,581 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.601890816854365
+2022-12-13 20:07:33,622 INFO [train.py:421] (2/8) Epoch 10, batch 36200, loss[loss=2.415, over 1680.00 frames. , ppl: 11.184991893356507] tot_loss[loss=2.27, over 5476180.34 frames. , ppl: 9.67636459180641], batch size: 70
+2022-12-13 20:09:17,733 INFO [train.py:421] (2/8) Epoch 10, batch 36400, loss[loss=2.315, over 1890.00 frames. , ppl: 10.121228029712013] tot_loss[loss=2.269, over 5526728.02 frames. , ppl: 9.665756216134731], batch size: 70
+2022-12-13 20:11:03,402 INFO [train.py:421] (2/8) Epoch 10, batch 36600, loss[loss=2.252, over 2450.00 frames. , ppl: 9.502880713593598] tot_loss[loss=2.268, over 5547812.37 frames. , ppl: 9.658372729814621], batch size: 70
+2022-12-13 20:12:48,783 INFO [train.py:421] (2/8) Epoch 10, batch 36800, loss[loss=2.234, over 2870.00 frames. , ppl: 9.340306821518304] tot_loss[loss=2.268, over 5533433.52 frames. , ppl: 9.658433479462394], batch size: 70
+2022-12-13 20:14:30,488 INFO [train.py:421] (2/8) Epoch 10, batch 37000, loss[loss=2.21, over 6020.00 frames. , ppl: 9.118543015586136] tot_loss[loss=2.266, over 5576900.82 frames. , ppl: 9.636119259919239], batch size: 70
+2022-12-13 20:14:30,489 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:14:31,281 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.600985586014557
+2022-12-13 20:16:16,936 INFO [train.py:421] (2/8) Epoch 10, batch 37200, loss[loss=2.284, over 1960.00 frames. , ppl: 9.815298826836178] tot_loss[loss=2.265, over 5610169.53 frames. , ppl: 9.630075276254178], batch size: 70
+2022-12-13 20:17:58,370 INFO [train.py:421] (2/8) Epoch 10, batch 37400, loss[loss=2.627, over 770.00 frames. , ppl: 13.827364099307381] tot_loss[loss=2.266, over 5566984.94 frames. , ppl: 9.640172789438065], batch size: 70
+2022-12-13 20:19:44,288 INFO [train.py:421] (2/8) Epoch 10, batch 37600, loss[loss=2.405, over 2030.00 frames. , ppl: 11.083544534770214] tot_loss[loss=2.267, over 5544805.63 frames. , ppl: 9.648463938157745], batch size: 70
+2022-12-13 20:21:26,441 INFO [train.py:421] (2/8) Epoch 10, batch 37800, loss[loss=2.339, over 2380.00 frames. , ppl: 10.369827871967908] tot_loss[loss=2.267, over 5552788.88 frames. , ppl: 9.650455722519979], batch size: 70
+2022-12-13 20:23:05,393 INFO [train.py:421] (2/8) Epoch 10, batch 38000, loss[loss=2.267, over 4060.00 frames. , ppl: 9.64946495587744] tot_loss[loss=2.269, over 5488603.94 frames. , ppl: 9.666024346630357], batch size: 70
+2022-12-13 20:23:05,393 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:23:06,141 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589634108742406
+2022-12-13 20:24:47,970 INFO [train.py:421] (2/8) Epoch 10, batch 38200, loss[loss=2.145, over 8680.00 frames. , ppl: 8.54014012507785] tot_loss[loss=2.269, over 5482652.79 frames. , ppl: 9.668718573755084], batch size: 70
+2022-12-13 20:26:26,942 INFO [train.py:421] (2/8) Epoch 10, batch 38400, loss[loss=2.422, over 1190.00 frames. , ppl: 11.27011845335133] tot_loss[loss=2.27, over 5454427.67 frames. , ppl: 9.675292851375284], batch size: 70
+2022-12-13 20:28:07,271 INFO [train.py:421] (2/8) Epoch 10, batch 38600, loss[loss=2.37, over 1540.00 frames. , ppl: 10.692685872883908] tot_loss[loss=2.268, over 5509473.68 frames. , ppl: 9.661611399331905], batch size: 70
+2022-12-13 20:29:50,034 INFO [train.py:421] (2/8) Epoch 10, batch 38800, loss[loss=2.762, over 630.00 frames. , ppl: 15.83451288793082] tot_loss[loss=2.268, over 5494800.16 frames. , ppl: 9.661760837812364], batch size: 70
+2022-12-13 20:31:30,514 INFO [train.py:421] (2/8) Epoch 10, batch 39000, loss[loss=2.391, over 3080.00 frames. , ppl: 10.921470068282476] tot_loss[loss=2.268, over 5491799.22 frames. , ppl: 9.663987431223921], batch size: 70
+2022-12-13 20:31:30,515 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:31:31,277 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59960729794989
+2022-12-13 20:33:11,828 INFO [train.py:421] (2/8) Epoch 10, batch 39200, loss[loss=2.237, over 3220.00 frames. , ppl: 9.368638390084307] tot_loss[loss=2.267, over 5530506.95 frames. , ppl: 9.654420521194272], batch size: 70
+2022-12-13 20:34:51,223 INFO [train.py:421] (2/8) Epoch 10, batch 39400, loss[loss=2.364, over 1610.00 frames. , ppl: 10.629943825370647] tot_loss[loss=2.269, over 5489883.52 frames. , ppl: 9.67452356801408], batch size: 70
+2022-12-13 20:36:33,968 INFO [train.py:421] (2/8) Epoch 10, batch 39600, loss[loss=2.19, over 4130.00 frames. , ppl: 8.938482610683476] tot_loss[loss=2.269, over 5495718.40 frames. , ppl: 9.667537930843084], batch size: 70
+2022-12-13 20:38:17,862 INFO [train.py:421] (2/8) Epoch 10, batch 39800, loss[loss=2.279, over 1820.00 frames. , ppl: 9.771444304819928] tot_loss[loss=2.268, over 5529064.45 frames. , ppl: 9.660396819263037], batch size: 70
+2022-12-13 20:40:01,099 INFO [train.py:421] (2/8) Epoch 10, batch 40000, loss[loss=2.188, over 7980.00 frames. , ppl: 8.91460904253861] tot_loss[loss=2.267, over 5568385.01 frames. , ppl: 9.647829669611918], batch size: 70
+2022-12-13 20:40:01,100 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:40:01,849 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.261, over 211138.00 frames. , ppl: 9.595198105641067
+2022-12-13 20:41:45,209 INFO [train.py:421] (2/8) Epoch 10, batch 40200, loss[loss=2.407, over 2450.00 frames. , ppl: 11.0988811732534] tot_loss[loss=2.267, over 5575226.73 frames. , ppl: 9.64848181355177], batch size: 70
+2022-12-13 20:43:28,822 INFO [train.py:421] (2/8) Epoch 10, batch 40400, loss[loss=5.155, over 280.00 frames. , ppl: 173.3640578369529] tot_loss[loss=2.266, over 5599011.23 frames. , ppl: 9.644061565108599], batch size: 70
+2022-12-13 20:45:07,757 INFO [train.py:421] (2/8) Epoch 10, batch 40600, loss[loss=2.38, over 2170.00 frames. , ppl: 10.80972561226645] tot_loss[loss=2.267, over 5597415.06 frames. , ppl: 9.647417271091753], batch size: 70
+2022-12-13 20:46:47,451 INFO [train.py:421] (2/8) Epoch 10, batch 40800, loss[loss=2.202, over 4830.00 frames. , ppl: 9.046305911186154] tot_loss[loss=2.268, over 5547213.13 frames. , ppl: 9.656404584500743], batch size: 70
+2022-12-13 20:48:25,962 INFO [train.py:421] (2/8) Epoch 10, batch 41000, loss[loss=2.227, over 3850.00 frames. , ppl: 9.274836136860657] tot_loss[loss=2.267, over 5519284.12 frames. , ppl: 9.654185100557571], batch size: 70
+2022-12-13 20:48:25,963 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:48:26,717 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60531213793815
+2022-12-13 20:50:09,272 INFO [train.py:421] (2/8) Epoch 10, batch 41200, loss[loss=2.291, over 4130.00 frames. , ppl: 9.881279671884542] tot_loss[loss=2.268, over 5514808.01 frames. , ppl: 9.661375704846929], batch size: 70
+2022-12-13 20:51:52,023 INFO [train.py:421] (2/8) Epoch 10, batch 41400, loss[loss=2.193, over 4340.00 frames. , ppl: 8.96039369420179] tot_loss[loss=2.268, over 5503582.34 frames. , ppl: 9.66436498018154], batch size: 70
+2022-12-13 20:53:33,166 INFO [train.py:421] (2/8) Epoch 10, batch 41600, loss[loss=2.296, over 3080.00 frames. , ppl: 9.929928835694351] tot_loss[loss=2.269, over 5501393.17 frames. , ppl: 9.668278996301256], batch size: 70
+2022-12-13 20:55:14,357 INFO [train.py:421] (2/8) Epoch 10, batch 41800, loss[loss=2.297, over 2450.00 frames. , ppl: 9.94053155756237] tot_loss[loss=2.27, over 5479640.77 frames. , ppl: 9.676947092258422], batch size: 70
+2022-12-13 20:56:55,444 INFO [train.py:421] (2/8) Epoch 10, batch 42000, loss[loss=2.539, over 770.00 frames. , ppl: 12.666847510263013] tot_loss[loss=2.27, over 5469117.24 frames. , ppl: 9.681120972027077], batch size: 70
+2022-12-13 20:56:55,444 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 20:56:56,194 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599615822834881
+2022-12-13 20:58:38,510 INFO [train.py:421] (2/8) Epoch 10, batch 42200, loss[loss=2.322, over 1470.00 frames. , ppl: 10.194587039797437] tot_loss[loss=2.27, over 5451463.10 frames. , ppl: 9.680334953311897], batch size: 70
+2022-12-13 21:00:23,423 INFO [train.py:421] (2/8) Epoch 10, batch 42400, loss[loss=2.464, over 1120.00 frames. , ppl: 11.748094502952146] tot_loss[loss=2.27, over 5482011.89 frames. , ppl: 9.681918407966204], batch size: 70
+2022-12-13 21:02:03,395 INFO [train.py:421] (2/8) Epoch 10, batch 42600, loss[loss=2.757, over 630.00 frames. , ppl: 15.747336493696197] tot_loss[loss=2.271, over 5462669.18 frames. , ppl: 9.685385934877065], batch size: 70
+2022-12-13 21:03:47,729 INFO [train.py:421] (2/8) Epoch 10, batch 42800, loss[loss=2.322, over 2450.00 frames. , ppl: 10.19184215544889] tot_loss[loss=2.27, over 5485648.69 frames. , ppl: 9.680972605829394], batch size: 70
+2022-12-13 21:05:29,142 INFO [train.py:421] (2/8) Epoch 10, batch 43000, loss[loss=2.164, over 7910.00 frames. , ppl: 8.701570064932302] tot_loss[loss=2.267, over 5552936.64 frames. , ppl: 9.653492649631369], batch size: 70
+2022-12-13 21:05:29,143 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:05:29,892 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.605019280876657
+2022-12-13 21:07:11,583 INFO [train.py:421] (2/8) Epoch 10, batch 43200, loss[loss=2.476, over 1330.00 frames. , ppl: 11.89416530109778] tot_loss[loss=2.267, over 5554176.19 frames. , ppl: 9.654051077208008], batch size: 70
+2022-12-13 21:08:55,559 INFO [train.py:421] (2/8) Epoch 10, batch 43400, loss[loss=2.228, over 10150.00 frames. , ppl: 9.284163636621829] tot_loss[loss=2.268, over 5547331.96 frames. , ppl: 9.658213750392022], batch size: 70
+2022-12-13 21:10:37,369 INFO [train.py:421] (2/8) Epoch 10, batch 43600, loss[loss=3.668, over 420.00 frames. , ppl: 39.17854146355111] tot_loss[loss=2.268, over 5549878.36 frames. , ppl: 9.66120902186981], batch size: 70
+2022-12-13 21:12:19,476 INFO [train.py:421] (2/8) Epoch 10, batch 43800, loss[loss=2.953, over 560.00 frames. , ppl: 19.162678605569155] tot_loss[loss=2.27, over 5529081.32 frames. , ppl: 9.675196887852767], batch size: 70
+2022-12-13 21:13:59,359 INFO [train.py:421] (2/8) Epoch 10, batch 44000, loss[loss=2.311, over 2590.00 frames. , ppl: 10.087600638839861] tot_loss[loss=2.269, over 5559962.68 frames. , ppl: 9.669987723572591], batch size: 70
+2022-12-13 21:13:59,360 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:14:00,109 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.620145785284024
+2022-12-13 21:15:38,909 INFO [train.py:421] (2/8) Epoch 10, batch 44200, loss[loss=2.337, over 1610.00 frames. , ppl: 10.345688256970481] tot_loss[loss=2.269, over 5557151.13 frames. , ppl: 9.67272902507332], batch size: 70
+2022-12-13 21:17:20,593 INFO [train.py:421] (2/8) Epoch 10, batch 44400, loss[loss=2.535, over 1540.00 frames. , ppl: 12.61568202569382] tot_loss[loss=2.269, over 5597185.59 frames. , ppl: 9.668405866836917], batch size: 70
+2022-12-13 21:19:00,939 INFO [train.py:421] (2/8) Epoch 10, batch 44600, loss[loss=2.252, over 2870.00 frames. , ppl: 9.507116023211157] tot_loss[loss=2.269, over 5590093.03 frames. , ppl: 9.671855314938073], batch size: 70
+2022-12-13 21:20:45,693 INFO [train.py:421] (2/8) Epoch 10, batch 44800, loss[loss=2.494, over 1400.00 frames. , ppl: 12.109842536177208] tot_loss[loss=2.268, over 5616337.93 frames. , ppl: 9.657724411027347], batch size: 70
+2022-12-13 21:22:25,934 INFO [train.py:421] (2/8) Epoch 10, batch 45000, loss[loss=2.276, over 1890.00 frames. , ppl: 9.741844595230965] tot_loss[loss=2.268, over 5619275.19 frames. , ppl: 9.655660253127694], batch size: 70
+2022-12-13 21:22:25,935 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:22:26,698 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.598352801576091
+2022-12-13 21:24:05,946 INFO [train.py:421] (2/8) Epoch 10, batch 45200, loss[loss=2.561, over 910.00 frames. , ppl: 12.951473201523811] tot_loss[loss=2.268, over 5585075.92 frames. , ppl: 9.661819661586451], batch size: 70
+2022-12-13 21:25:48,530 INFO [train.py:421] (2/8) Epoch 10, batch 45400, loss[loss=2.317, over 2940.00 frames. , ppl: 10.145932673757814] tot_loss[loss=2.269, over 5563815.76 frames. , ppl: 9.669053623330784], batch size: 70
+2022-12-13 21:27:30,577 INFO [train.py:421] (2/8) Epoch 10, batch 45600, loss[loss=2.161, over 3990.00 frames. , ppl: 8.6760423669709] tot_loss[loss=2.269, over 5572305.39 frames. , ppl: 9.667763280162584], batch size: 70
+2022-12-13 21:29:13,811 INFO [train.py:421] (2/8) Epoch 10, batch 45800, loss[loss=2.33, over 2240.00 frames. , ppl: 10.277830409136813] tot_loss[loss=2.268, over 5598481.68 frames. , ppl: 9.664257088329006], batch size: 70
+2022-12-13 21:30:58,387 INFO [train.py:421] (2/8) Epoch 10, batch 46000, loss[loss=2.323, over 2030.00 frames. , ppl: 10.20373823507494] tot_loss[loss=2.269, over 5594695.79 frames. , ppl: 9.666870193864053], batch size: 70
+2022-12-13 21:30:58,388 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:30:59,139 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.603299281294177
+2022-12-13 21:32:38,607 INFO [train.py:421] (2/8) Epoch 10, batch 46200, loss[loss=2.754, over 630.00 frames. , ppl: 15.708669888262934] tot_loss[loss=2.269, over 5580779.11 frames. , ppl: 9.665483711746562], batch size: 70
+2022-12-13 21:34:24,216 INFO [train.py:421] (2/8) Epoch 10, batch 46400, loss[loss=3.044, over 560.00 frames. , ppl: 20.978991959582128] tot_loss[loss=2.269, over 5578901.13 frames. , ppl: 9.669158653138584], batch size: 70
+2022-12-13 21:36:10,344 INFO [train.py:421] (2/8) Epoch 10, batch 46600, loss[loss=2.286, over 1680.00 frames. , ppl: 9.832352715994418] tot_loss[loss=2.269, over 5557170.79 frames. , ppl: 9.673393177649517], batch size: 70
+2022-12-13 21:37:49,341 INFO [train.py:421] (2/8) Epoch 10, batch 46800, loss[loss=2.236, over 3150.00 frames. , ppl: 9.358252043359009] tot_loss[loss=2.269, over 5545646.71 frames. , ppl: 9.673088348459737], batch size: 70
+2022-12-13 21:39:32,506 INFO [train.py:421] (2/8) Epoch 10, batch 47000, loss[loss=2.238, over 3710.00 frames. , ppl: 9.373708360486033] tot_loss[loss=2.27, over 5552239.99 frames. , ppl: 9.67465052384487], batch size: 70
+2022-12-13 21:39:32,506 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:39:33,262 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.62268485087924
+2022-12-13 21:41:16,069 INFO [train.py:421] (2/8) Epoch 10, batch 47200, loss[loss=2.263, over 3990.00 frames. , ppl: 9.611950107301409] tot_loss[loss=2.271, over 5516159.74 frames. , ppl: 9.68443108499043], batch size: 70
+2022-12-13 21:42:58,724 INFO [train.py:421] (2/8) Epoch 10, batch 47400, loss[loss=2.248, over 3850.00 frames. , ppl: 9.469509016534907] tot_loss[loss=2.271, over 5470715.47 frames. , ppl: 9.69286635324809], batch size: 70
+2022-12-13 21:44:41,590 INFO [train.py:421] (2/8) Epoch 10, batch 47600, loss[loss=2.251, over 3710.00 frames. , ppl: 9.493763985250048] tot_loss[loss=2.271, over 5470432.09 frames. , ppl: 9.687542688268762], batch size: 70
+2022-12-13 21:46:26,614 INFO [train.py:421] (2/8) Epoch 10, batch 47800, loss[loss=2.171, over 3640.00 frames. , ppl: 8.765509615126836] tot_loss[loss=2.271, over 5485172.27 frames. , ppl: 9.68940286694529], batch size: 70
+2022-12-13 21:48:08,393 INFO [train.py:421] (2/8) Epoch 10, batch 48000, loss[loss=2.133, over 5740.00 frames. , ppl: 8.44140862087467] tot_loss[loss=2.273, over 5412440.22 frames. , ppl: 9.706338781884735], batch size: 70
+2022-12-13 21:48:08,394 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:48:09,144 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621020070987242
+2022-12-13 21:49:55,460 INFO [train.py:421] (2/8) Epoch 10, batch 48200, loss[loss=2.206, over 12810.00 frames. , ppl: 9.079828432858918] tot_loss[loss=2.272, over 5452076.60 frames. , ppl: 9.698454421118667], batch size: 70
+2022-12-13 21:51:37,487 INFO [train.py:421] (2/8) Epoch 10, batch 48400, loss[loss=2.179, over 4200.00 frames. , ppl: 8.835573710824711] tot_loss[loss=2.272, over 5418758.54 frames. , ppl: 9.703036058914316], batch size: 70
+2022-12-13 21:53:19,244 INFO [train.py:421] (2/8) Epoch 10, batch 48600, loss[loss=2.457, over 1890.00 frames. , ppl: 11.667292736660405] tot_loss[loss=2.272, over 5441753.86 frames. , ppl: 9.698347419668591], batch size: 70
+2022-12-13 21:55:03,437 INFO [train.py:421] (2/8) Epoch 10, batch 48800, loss[loss=2.149, over 6510.00 frames. , ppl: 8.573391664410304] tot_loss[loss=2.271, over 5468407.45 frames. , ppl: 9.686617773225437], batch size: 70
+2022-12-13 21:56:44,857 INFO [train.py:421] (2/8) Epoch 10, batch 49000, loss[loss=2.429, over 1540.00 frames. , ppl: 11.342544414822259] tot_loss[loss=2.273, over 5399768.41 frames. , ppl: 9.706408018746783], batch size: 70
+2022-12-13 21:56:44,857 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 21:56:45,606 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608761699260915
+2022-12-13 21:58:25,257 INFO [train.py:421] (2/8) Epoch 10, batch 49200, loss[loss=2.963, over 560.00 frames. , ppl: 19.352241015785655] tot_loss[loss=2.272, over 5412471.20 frames. , ppl: 9.694386367601224], batch size: 70
+2022-12-13 22:00:03,861 INFO [train.py:421] (2/8) Epoch 10, batch 49400, loss[loss=2.44, over 1120.00 frames. , ppl: 11.473391377754794] tot_loss[loss=2.273, over 5391868.92 frames. , ppl: 9.705404885385231], batch size: 70
+2022-12-13 22:01:44,563 INFO [train.py:421] (2/8) Epoch 10, batch 49600, loss[loss=2.273, over 3920.00 frames. , ppl: 9.708076686870067] tot_loss[loss=2.273, over 5375294.20 frames. , ppl: 9.710564441323664], batch size: 70
+2022-12-13 22:03:22,301 INFO [train.py:421] (2/8) Epoch 10, batch 49800, loss[loss=2.526, over 1050.00 frames. , ppl: 12.500076284177194] tot_loss[loss=2.273, over 5400460.72 frames. , ppl: 9.703924707074643], batch size: 70
+2022-12-13 22:05:00,318 INFO [train.py:421] (2/8) Epoch 10, batch 50000, loss[loss=2.156, over 6370.00 frames. , ppl: 8.637771993374324] tot_loss[loss=2.272, over 5411382.46 frames. , ppl: 9.695787870304663], batch size: 70
+2022-12-13 22:05:00,318 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:05:01,070 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608694857576053
+2022-12-13 22:06:42,909 INFO [train.py:421] (2/8) Epoch 10, batch 50200, loss[loss=2.46, over 980.00 frames. , ppl: 11.706617228540216] tot_loss[loss=2.27, over 5440818.69 frames. , ppl: 9.68244482919756], batch size: 70
+2022-12-13 22:08:19,153 INFO [train.py:421] (2/8) Epoch 10, batch 50400, loss[loss=2.272, over 1890.00 frames. , ppl: 9.701937056987859] tot_loss[loss=2.271, over 5417129.19 frames. , ppl: 9.690286070295386], batch size: 70
+2022-12-13 22:10:01,748 INFO [train.py:421] (2/8) Epoch 10, batch 50600, loss[loss=2.2, over 5670.00 frames. , ppl: 9.024444604270453] tot_loss[loss=2.271, over 5433611.24 frames. , ppl: 9.686523973263498], batch size: 70
+2022-12-13 22:11:42,222 INFO [train.py:421] (2/8) Epoch 10, batch 50800, loss[loss=2.221, over 2800.00 frames. , ppl: 9.21627214712523] tot_loss[loss=2.27, over 5462450.77 frames. , ppl: 9.679028722186075], batch size: 70
+2022-12-13 22:13:22,635 INFO [train.py:421] (2/8) Epoch 10, batch 51000, loss[loss=2.147, over 6720.00 frames. , ppl: 8.557134910095979] tot_loss[loss=2.271, over 5449980.70 frames. , ppl: 9.687711362002382], batch size: 70
+2022-12-13 22:13:22,635 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:13:23,385 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617215941256719
+2022-12-13 22:15:05,584 INFO [train.py:421] (2/8) Epoch 10, batch 51200, loss[loss=2.231, over 5180.00 frames. , ppl: 9.308998747389765] tot_loss[loss=2.271, over 5462120.58 frames. , ppl: 9.685876360203778], batch size: 70
+2022-12-13 22:16:46,787 INFO [train.py:421] (2/8) Epoch 10, batch 51400, loss[loss=2.341, over 1470.00 frames. , ppl: 10.38790628750642] tot_loss[loss=2.269, over 5511392.61 frames. , ppl: 9.674105971320168], batch size: 70
+2022-12-13 22:18:32,939 INFO [train.py:421] (2/8) Epoch 10, batch 51600, loss[loss=4.065, over 350.00 frames. , ppl: 58.27435864097975] tot_loss[loss=2.269, over 5508602.91 frames. , ppl: 9.66648759380431], batch size: 70
+2022-12-13 22:20:11,661 INFO [train.py:421] (2/8) Epoch 10, batch 51800, loss[loss=2.296, over 2520.00 frames. , ppl: 9.938878109642648] tot_loss[loss=2.27, over 5442378.42 frames. , ppl: 9.681310864348443], batch size: 70
+2022-12-13 22:21:49,901 INFO [train.py:421] (2/8) Epoch 10, batch 52000, loss[loss=2.399, over 1680.00 frames. , ppl: 11.013641145904693] tot_loss[loss=2.27, over 5452402.40 frames. , ppl: 9.679652967228234], batch size: 70
+2022-12-13 22:21:49,902 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:21:50,672 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599999450495101
+2022-12-13 22:23:31,974 INFO [train.py:421] (2/8) Epoch 10, batch 52200, loss[loss=2.377, over 2030.00 frames. , ppl: 10.770321837096597] tot_loss[loss=2.27, over 5451827.12 frames. , ppl: 9.681850286391569], batch size: 70
+2022-12-13 22:25:13,177 INFO [train.py:421] (2/8) Epoch 10, batch 52400, loss[loss=2.222, over 3080.00 frames. , ppl: 9.222187920861595] tot_loss[loss=2.27, over 5477693.06 frames. , ppl: 9.680233262109311], batch size: 70
+2022-12-13 22:26:53,088 INFO [train.py:421] (2/8) Epoch 10, batch 52600, loss[loss=2.9, over 630.00 frames. , ppl: 18.16939905309852] tot_loss[loss=2.27, over 5480263.01 frames. , ppl: 9.679707220926547], batch size: 70
+2022-12-13 22:28:32,258 INFO [train.py:421] (2/8) Epoch 10, batch 52800, loss[loss=2.159, over 5250.00 frames. , ppl: 8.66640982290213] tot_loss[loss=2.269, over 5509659.72 frames. , ppl: 9.667826649534597], batch size: 70
+2022-12-13 22:30:12,636 INFO [train.py:421] (2/8) Epoch 10, batch 53000, loss[loss=2.587, over 910.00 frames. , ppl: 13.295318782895514] tot_loss[loss=2.268, over 5526629.37 frames. , ppl: 9.661926105876876], batch size: 70
+2022-12-13 22:30:12,637 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:30:13,403 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621981307382367
+2022-12-13 22:31:58,567 INFO [train.py:421] (2/8) Epoch 10, batch 53200, loss[loss=2.193, over 2870.00 frames. , ppl: 8.963846664848386] tot_loss[loss=2.268, over 5536325.98 frames. , ppl: 9.657302924682545], batch size: 70
+2022-12-13 22:33:42,494 INFO [train.py:421] (2/8) Epoch 10, batch 53400, loss[loss=2.171, over 6930.00 frames. , ppl: 8.766045170205587] tot_loss[loss=2.268, over 5514516.32 frames. , ppl: 9.661335241658156], batch size: 70
+2022-12-13 22:35:23,717 INFO [train.py:421] (2/8) Epoch 10, batch 53600, loss[loss=2.439, over 770.00 frames. , ppl: 11.464712323947785] tot_loss[loss=2.267, over 5542174.09 frames. , ppl: 9.648811830898588], batch size: 70
+2022-12-13 22:37:03,239 INFO [train.py:421] (2/8) Epoch 10, batch 53800, loss[loss=2.343, over 2030.00 frames. , ppl: 10.411938863480753] tot_loss[loss=2.268, over 5506815.61 frames. , ppl: 9.663736042372268], batch size: 70
+2022-12-13 22:38:45,588 INFO [train.py:421] (2/8) Epoch 10, batch 54000, loss[loss=2.77, over 630.00 frames. , ppl: 15.95322607937666] tot_loss[loss=2.269, over 5467172.40 frames. , ppl: 9.673198573604468], batch size: 70
+2022-12-13 22:38:45,588 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:38:46,336 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606318724465547
+2022-12-13 22:40:29,332 INFO [train.py:421] (2/8) Epoch 10, batch 54200, loss[loss=2.4, over 1610.00 frames. , ppl: 11.01890972410026] tot_loss[loss=2.269, over 5486436.58 frames. , ppl: 9.670237851905883], batch size: 70
+2022-12-13 22:42:07,612 INFO [train.py:421] (2/8) Epoch 10, batch 54400, loss[loss=2.466, over 1120.00 frames. , ppl: 11.771383162702813] tot_loss[loss=2.269, over 5500858.40 frames. , ppl: 9.670324150067954], batch size: 70
+2022-12-13 22:43:49,809 INFO [train.py:421] (2/8) Epoch 10, batch 54600, loss[loss=2.171, over 4270.00 frames. , ppl: 8.770644344971247] tot_loss[loss=2.27, over 5459270.21 frames. , ppl: 9.679201368429686], batch size: 70
+2022-12-13 22:45:29,927 INFO [train.py:421] (2/8) Epoch 10, batch 54800, loss[loss=2.382, over 1050.00 frames. , ppl: 10.828468783313147] tot_loss[loss=2.27, over 5481953.79 frames. , ppl: 9.675082843833833], batch size: 70
+2022-12-13 22:47:09,488 INFO [train.py:421] (2/8) Epoch 10, batch 55000, loss[loss=2.219, over 8050.00 frames. , ppl: 9.201624965183234] tot_loss[loss=2.27, over 5468101.02 frames. , ppl: 9.677170272238142], batch size: 70
+2022-12-13 22:47:09,488 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:47:10,233 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.6300680375108
+2022-12-13 22:48:52,916 INFO [train.py:421] (2/8) Epoch 10, batch 55200, loss[loss=2.16, over 7140.00 frames. , ppl: 8.67020718730946] tot_loss[loss=2.268, over 5485640.05 frames. , ppl: 9.658954900873837], batch size: 70
+2022-12-13 22:50:33,381 INFO [train.py:421] (2/8) Epoch 10, batch 55400, loss[loss=2.249, over 3710.00 frames. , ppl: 9.475773329070678] tot_loss[loss=2.269, over 5452059.07 frames. , ppl: 9.665887857220484], batch size: 70
+2022-12-13 22:52:12,074 INFO [train.py:421] (2/8) Epoch 10, batch 55600, loss[loss=2.153, over 3150.00 frames. , ppl: 8.608364742292858] tot_loss[loss=2.27, over 5399501.12 frames. , ppl: 9.675584892566869], batch size: 70
+2022-12-13 22:53:51,906 INFO [train.py:421] (2/8) Epoch 10, batch 55800, loss[loss=2.996, over 560.00 frames. , ppl: 19.99879363478001] tot_loss[loss=2.271, over 5348692.91 frames. , ppl: 9.691376644858412], batch size: 70
+2022-12-13 22:55:32,376 INFO [train.py:421] (2/8) Epoch 10, batch 56000, loss[loss=2.243, over 2520.00 frames. , ppl: 9.41896116833786] tot_loss[loss=2.271, over 5372527.86 frames. , ppl: 9.6884628311667], batch size: 70
+2022-12-13 22:55:32,376 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 22:55:33,125 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.602260323357212
+2022-12-13 22:57:14,379 INFO [train.py:421] (2/8) Epoch 10, batch 56200, loss[loss=2.13, over 7980.00 frames. , ppl: 8.414208190010735] tot_loss[loss=2.271, over 5371793.11 frames. , ppl: 9.686005789121483], batch size: 70
+2022-12-13 22:58:56,428 INFO [train.py:421] (2/8) Epoch 10, batch 56400, loss[loss=2.3, over 3010.00 frames. , ppl: 9.972651935995883] tot_loss[loss=2.269, over 5413297.42 frames. , ppl: 9.673895811792736], batch size: 70
+2022-12-13 23:00:37,873 INFO [train.py:421] (2/8) Epoch 10, batch 56600, loss[loss=2.409, over 2030.00 frames. , ppl: 11.127657656150873] tot_loss[loss=2.27, over 5418577.01 frames. , ppl: 9.675424221827752], batch size: 70
+2022-12-13 23:02:17,338 INFO [train.py:421] (2/8) Epoch 10, batch 56800, loss[loss=2.313, over 2730.00 frames. , ppl: 10.104737552088777] tot_loss[loss=2.27, over 5436337.77 frames. , ppl: 9.677674735590593], batch size: 70
+2022-12-13 23:03:57,927 INFO [train.py:421] (2/8) Epoch 10, batch 57000, loss[loss=2.424, over 2030.00 frames. , ppl: 11.29406039648827] tot_loss[loss=2.27, over 5411187.95 frames. , ppl: 9.682926320475694], batch size: 70
+2022-12-13 23:03:57,928 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:03:58,673 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.62573033470701
+2022-12-13 23:05:40,807 INFO [train.py:421] (2/8) Epoch 10, batch 57200, loss[loss=2.34, over 2170.00 frames. , ppl: 10.37915219011016] tot_loss[loss=2.269, over 5453650.74 frames. , ppl: 9.673391500852324], batch size: 70
+2022-12-13 23:07:18,394 INFO [train.py:421] (2/8) Epoch 10, batch 57400, loss[loss=2.544, over 770.00 frames. , ppl: 12.731534761065632] tot_loss[loss=2.27, over 5455045.53 frames. , ppl: 9.67849621808994], batch size: 70
+2022-12-13 23:08:53,716 INFO [train.py:421] (2/8) Epoch 10, batch 57600, loss[loss=2.151, over 11130.00 frames. , ppl: 8.594677323669858] tot_loss[loss=2.269, over 5493312.10 frames. , ppl: 9.669109155191498], batch size: 70
+2022-12-13 23:10:36,350 INFO [train.py:421] (2/8) Epoch 10, batch 57800, loss[loss=2.138, over 6370.00 frames. , ppl: 8.484549096578602] tot_loss[loss=2.268, over 5505230.80 frames. , ppl: 9.660743997855302], batch size: 70
+2022-12-13 23:12:20,210 INFO [train.py:421] (2/8) Epoch 10, batch 58000, loss[loss=2.34, over 2380.00 frames. , ppl: 10.378652139232127] tot_loss[loss=2.268, over 5511729.60 frames. , ppl: 9.660761570407235], batch size: 70
+2022-12-13 23:12:20,211 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:12:20,972 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.609582325499408
+2022-12-13 23:14:05,369 INFO [train.py:421] (2/8) Epoch 10, batch 58200, loss[loss=2.428, over 980.00 frames. , ppl: 11.331312982001648] tot_loss[loss=2.268, over 5523671.85 frames. , ppl: 9.661816433098544], batch size: 70
+2022-12-13 23:15:40,237 INFO [train.py:421] (2/8) Epoch 10, batch 58400, loss[loss=2.219, over 5810.00 frames. , ppl: 9.199072265203702] tot_loss[loss=2.268, over 5518441.44 frames. , ppl: 9.655328976205949], batch size: 70
+2022-12-13 23:17:18,353 INFO [train.py:421] (2/8) Epoch 10, batch 58600, loss[loss=2.14, over 5810.00 frames. , ppl: 8.501046969075615] tot_loss[loss=2.268, over 5512989.23 frames. , ppl: 9.65865760343959], batch size: 70
+2022-12-13 23:18:57,519 INFO [train.py:421] (2/8) Epoch 10, batch 58800, loss[loss=2.556, over 910.00 frames. , ppl: 12.878177495682145] tot_loss[loss=2.268, over 5536461.46 frames. , ppl: 9.65819375142199], batch size: 70
+2022-12-13 23:20:40,941 INFO [train.py:421] (2/8) Epoch 10, batch 59000, loss[loss=2.364, over 1820.00 frames. , ppl: 10.631029575087267] tot_loss[loss=2.268, over 5556515.90 frames. , ppl: 9.65788024319233], batch size: 70
+2022-12-13 23:20:40,942 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:20:41,703 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606206402379026
+2022-12-13 23:22:20,233 INFO [train.py:421] (2/8) Epoch 10, batch 59200, loss[loss=2.271, over 2660.00 frames. , ppl: 9.693103182778875] tot_loss[loss=2.269, over 5507087.67 frames. , ppl: 9.674377914663953], batch size: 70
+2022-12-13 23:24:00,444 INFO [train.py:421] (2/8) Epoch 10, batch 59400, loss[loss=2.29, over 4340.00 frames. , ppl: 9.872250756135754] tot_loss[loss=2.27, over 5475604.43 frames. , ppl: 9.67971522923746], batch size: 70
+2022-12-13 23:25:41,349 INFO [train.py:421] (2/8) Epoch 10, batch 59600, loss[loss=2.276, over 2310.00 frames. , ppl: 9.74088207955319] tot_loss[loss=2.269, over 5491830.19 frames. , ppl: 9.674185390904963], batch size: 70
+2022-12-13 23:27:19,481 INFO [train.py:421] (2/8) Epoch 10, batch 59800, loss[loss=2.363, over 1680.00 frames. , ppl: 10.625383321535043] tot_loss[loss=2.27, over 5480218.90 frames. , ppl: 9.675814511014536], batch size: 70
+2022-12-13 23:28:53,372 INFO [train.py:421] (2/8) Epoch 10, batch 60000, loss[loss=2.185, over 3500.00 frames. , ppl: 8.887215996977408] tot_loss[loss=2.269, over 5500916.12 frames. , ppl: 9.673616556013911], batch size: 70
+2022-12-13 23:28:53,373 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:28:54,130 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612700490778435
+2022-12-13 23:30:35,175 INFO [train.py:421] (2/8) Epoch 10, batch 60200, loss[loss=2.629, over 770.00 frames. , ppl: 13.86310843451136] tot_loss[loss=2.269, over 5517463.65 frames. , ppl: 9.67271802717875], batch size: 70
+2022-12-13 23:32:13,575 INFO [train.py:421] (2/8) Epoch 10, batch 60400, loss[loss=2.157, over 2100.00 frames. , ppl: 8.641462165321622] tot_loss[loss=2.27, over 5493377.61 frames. , ppl: 9.67470433097202], batch size: 70
+2022-12-13 23:33:56,148 INFO [train.py:421] (2/8) Epoch 10, batch 60600, loss[loss=2.222, over 2940.00 frames. , ppl: 9.225787489700478] tot_loss[loss=2.269, over 5512451.85 frames. , ppl: 9.667741535807961], batch size: 70
+2022-12-13 23:35:35,556 INFO [train.py:421] (2/8) Epoch 10, batch 60800, loss[loss=2.321, over 3290.00 frames. , ppl: 10.187429921162817] tot_loss[loss=2.269, over 5494077.69 frames. , ppl: 9.671464956798767], batch size: 70
+2022-12-13 23:37:16,235 INFO [train.py:421] (2/8) Epoch 10, batch 61000, loss[loss=2.558, over 1050.00 frames. , ppl: 12.912801304836606] tot_loss[loss=2.267, over 5544262.10 frames. , ppl: 9.655202485464464], batch size: 70
+2022-12-13 23:37:16,236 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:37:16,983 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60258010007721
+2022-12-13 23:38:59,952 INFO [train.py:421] (2/8) Epoch 10, batch 61200, loss[loss=2.228, over 2590.00 frames. , ppl: 9.282284171180933] tot_loss[loss=2.268, over 5553382.39 frames. , ppl: 9.656677222084816], batch size: 70
+2022-12-13 23:40:40,175 INFO [train.py:421] (2/8) Epoch 10, batch 61400, loss[loss=2.425, over 2450.00 frames. , ppl: 11.301121235236105] tot_loss[loss=2.268, over 5553078.67 frames. , ppl: 9.655247362741521], batch size: 70
+2022-12-13 23:42:19,008 INFO [train.py:421] (2/8) Epoch 10, batch 61600, loss[loss=2.201, over 3640.00 frames. , ppl: 9.031272277842948] tot_loss[loss=2.267, over 5553838.43 frames. , ppl: 9.64950958493658], batch size: 70
+2022-12-13 23:44:03,592 INFO [train.py:421] (2/8) Epoch 10, batch 61800, loss[loss=2.232, over 4270.00 frames. , ppl: 9.315632591793596] tot_loss[loss=2.268, over 5544890.00 frames. , ppl: 9.658098641408229], batch size: 70
+2022-12-13 23:45:42,346 INFO [train.py:421] (2/8) Epoch 10, batch 62000, loss[loss=2.469, over 1120.00 frames. , ppl: 11.813759070746986] tot_loss[loss=2.266, over 5579655.47 frames. , ppl: 9.644525916127504], batch size: 70
+2022-12-13 23:45:42,346 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:45:43,104 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61007587277309
+2022-12-13 23:47:24,544 INFO [train.py:421] (2/8) Epoch 10, batch 62200, loss[loss=2.242, over 6860.00 frames. , ppl: 9.416413421181787] tot_loss[loss=2.265, over 5610498.22 frames. , ppl: 9.635174820378118], batch size: 70
+2022-12-13 23:49:09,235 INFO [train.py:421] (2/8) Epoch 10, batch 62400, loss[loss=2.237, over 4830.00 frames. , ppl: 9.363792346356833] tot_loss[loss=2.265, over 5598536.59 frames. , ppl: 9.634754215873889], batch size: 70
+2022-12-13 23:50:45,969 INFO [train.py:421] (2/8) Epoch 10, batch 62600, loss[loss=2.322, over 2730.00 frames. , ppl: 10.194488968506878] tot_loss[loss=2.266, over 5571243.37 frames. , ppl: 9.642998121450447], batch size: 70
+2022-12-13 23:52:24,484 INFO [train.py:421] (2/8) Epoch 10, batch 62800, loss[loss=2.313, over 2240.00 frames. , ppl: 10.10795428351182] tot_loss[loss=2.267, over 5553665.96 frames. , ppl: 9.647696973238117], batch size: 70
+2022-12-13 23:54:03,434 INFO [train.py:421] (2/8) Epoch 10, batch 63000, loss[loss=2.459, over 1610.00 frames. , ppl: 11.690929969923502] tot_loss[loss=2.266, over 5574630.54 frames. , ppl: 9.635999400004383], batch size: 70
+2022-12-13 23:54:03,434 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-13 23:54:04,195 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606716838640049
+2022-12-13 23:55:42,811 INFO [train.py:421] (2/8) Epoch 10, batch 63200, loss[loss=2.243, over 2450.00 frames. , ppl: 9.423148272406848] tot_loss[loss=2.265, over 5598549.48 frames. , ppl: 9.633262614907354], batch size: 70
+2022-12-13 23:57:21,563 INFO [train.py:421] (2/8) Epoch 10, batch 63400, loss[loss=2.21, over 5460.00 frames. , ppl: 9.119283787110566] tot_loss[loss=2.265, over 5592940.66 frames. , ppl: 9.633602288170486], batch size: 70
+2022-12-13 23:59:00,619 INFO [train.py:421] (2/8) Epoch 10, batch 63600, loss[loss=2.242, over 2800.00 frames. , ppl: 9.40907974570767] tot_loss[loss=2.266, over 5566276.92 frames. , ppl: 9.641546635546737], batch size: 70
+2022-12-14 00:00:37,636 INFO [train.py:421] (2/8) Epoch 10, batch 63800, loss[loss=2.306, over 2100.00 frames. , ppl: 10.031774780030858] tot_loss[loss=2.267, over 5544883.22 frames. , ppl: 9.648213124653017], batch size: 70
+2022-12-14 00:02:18,824 INFO [train.py:421] (2/8) Epoch 10, batch 64000, loss[loss=2.303, over 1610.00 frames. , ppl: 10.004960543597173] tot_loss[loss=2.267, over 5540562.42 frames. , ppl: 9.652483632010213], batch size: 70
+2022-12-14 00:02:18,824 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:02:19,585 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616598196927988
+2022-12-14 00:03:57,270 INFO [train.py:421] (2/8) Epoch 10, batch 64200, loss[loss=2.319, over 2240.00 frames. , ppl: 10.163828898409111] tot_loss[loss=2.267, over 5576761.67 frames. , ppl: 9.649762678041794], batch size: 70
+2022-12-14 00:05:36,743 INFO [train.py:421] (2/8) Epoch 10, batch 64400, loss[loss=2.397, over 1470.00 frames. , ppl: 10.994368763177649] tot_loss[loss=2.267, over 5578582.62 frames. , ppl: 9.651794727959992], batch size: 70
+2022-12-14 00:07:12,118 INFO [train.py:421] (2/8) Epoch 10, batch 64600, loss[loss=3.119, over 490.00 frames. , ppl: 22.633680272431313] tot_loss[loss=2.267, over 5581332.92 frames. , ppl: 9.654245711768715], batch size: 70
+2022-12-14 00:08:51,348 INFO [train.py:421] (2/8) Epoch 10, batch 64800, loss[loss=2.817, over 630.00 frames. , ppl: 16.72449771746581] tot_loss[loss=2.268, over 5535026.62 frames. , ppl: 9.660974297958221], batch size: 70
+2022-12-14 00:10:28,813 INFO [train.py:421] (2/8) Epoch 10, batch 65000, loss[loss=2.697, over 630.00 frames. , ppl: 14.831304878204996] tot_loss[loss=2.269, over 5540142.73 frames. , ppl: 9.665176510090621], batch size: 70
+2022-12-14 00:10:28,813 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:10:29,560 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599982400051333
+2022-12-14 00:12:08,438 INFO [train.py:421] (2/8) Epoch 10, batch 65200, loss[loss=2.472, over 1050.00 frames. , ppl: 11.847813338684386] tot_loss[loss=2.268, over 5558638.88 frames. , ppl: 9.661927831766755], batch size: 70
+2022-12-14 00:13:46,654 INFO [train.py:421] (2/8) Epoch 10, batch 65400, loss[loss=2.788, over 700.00 frames. , ppl: 16.255393636730407] tot_loss[loss=2.269, over 5533919.95 frames. , ppl: 9.672206342093146], batch size: 70
+2022-12-14 00:15:25,095 INFO [train.py:421] (2/8) Epoch 10, batch 65600, loss[loss=2.423, over 2660.00 frames. , ppl: 11.274127512176726] tot_loss[loss=2.269, over 5533457.36 frames. , ppl: 9.671651899389454], batch size: 70
+2022-12-14 00:17:06,084 INFO [train.py:421] (2/8) Epoch 10, batch 65800, loss[loss=2.373, over 1750.00 frames. , ppl: 10.727063103399672] tot_loss[loss=2.269, over 5533162.62 frames. , ppl: 9.670354202463878], batch size: 70
+2022-12-14 00:18:46,934 INFO [train.py:421] (2/8) Epoch 10, batch 66000, loss[loss=2.226, over 2940.00 frames. , ppl: 9.25852500403814] tot_loss[loss=2.269, over 5536192.36 frames. , ppl: 9.67443881481067], batch size: 70
+2022-12-14 00:18:46,935 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:18:47,681 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599644239172868
+2022-12-14 00:20:26,481 INFO [train.py:421] (2/8) Epoch 10, batch 66200, loss[loss=2.269, over 3920.00 frames. , ppl: 9.674235608046367] tot_loss[loss=2.27, over 5488078.74 frames. , ppl: 9.683655089164647], batch size: 70
+2022-12-14 00:22:05,233 INFO [train.py:421] (2/8) Epoch 10, batch 66400, loss[loss=2.309, over 1960.00 frames. , ppl: 10.060257426984379] tot_loss[loss=2.27, over 5492369.09 frames. , ppl: 9.68078590094172], batch size: 70
+2022-12-14 00:23:48,624 INFO [train.py:421] (2/8) Epoch 10, batch 66600, loss[loss=2.353, over 1540.00 frames. , ppl: 10.520342886241568] tot_loss[loss=2.269, over 5547987.92 frames. , ppl: 9.665286036232681], batch size: 70
+2022-12-14 00:25:27,835 INFO [train.py:421] (2/8) Epoch 10, batch 66800, loss[loss=2.806, over 700.00 frames. , ppl: 16.53608201854189] tot_loss[loss=2.268, over 5546834.56 frames. , ppl: 9.6640412241278], batch size: 70
+2022-12-14 00:27:09,560 INFO [train.py:421] (2/8) Epoch 10, batch 67000, loss[loss=2.588, over 840.00 frames. , ppl: 13.30502443654127] tot_loss[loss=2.27, over 5503283.59 frames. , ppl: 9.677107277675217], batch size: 70
+2022-12-14 00:27:09,560 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:27:10,307 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608690591101315
+2022-12-14 00:28:53,522 INFO [train.py:421] (2/8) Epoch 10, batch 67200, loss[loss=2.248, over 3920.00 frames. , ppl: 9.467536650649102] tot_loss[loss=2.27, over 5484785.13 frames. , ppl: 9.675792541146626], batch size: 70
+2022-12-14 00:30:34,764 INFO [train.py:421] (2/8) Epoch 10, batch 67400, loss[loss=2.221, over 4060.00 frames. , ppl: 9.217395807910643] tot_loss[loss=2.268, over 5504940.05 frames. , ppl: 9.662976620308799], batch size: 70
+2022-12-14 00:32:15,545 INFO [train.py:421] (2/8) Epoch 10, batch 67600, loss[loss=2.382, over 2100.00 frames. , ppl: 10.8314049129373] tot_loss[loss=2.268, over 5496525.63 frames. , ppl: 9.663868821669478], batch size: 70
+2022-12-14 00:33:55,293 INFO [train.py:421] (2/8) Epoch 10, batch 67800, loss[loss=3.224, over 490.00 frames. , ppl: 25.119683596868065] tot_loss[loss=2.268, over 5515326.08 frames. , ppl: 9.661367714489428], batch size: 70
+2022-12-14 00:35:36,606 INFO [train.py:421] (2/8) Epoch 10, batch 68000, loss[loss=3.631, over 420.00 frames. , ppl: 37.752952986804175] tot_loss[loss=2.268, over 5527128.43 frames. , ppl: 9.659638219892733], batch size: 70
+2022-12-14 00:35:36,607 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:35:37,368 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606830588577454
+2022-12-14 00:37:17,019 INFO [train.py:421] (2/8) Epoch 10, batch 68200, loss[loss=2.493, over 1330.00 frames. , ppl: 12.09807336834398] tot_loss[loss=2.268, over 5535449.89 frames. , ppl: 9.661869129352793], batch size: 70
+2022-12-14 00:38:58,132 INFO [train.py:421] (2/8) Epoch 10, batch 68400, loss[loss=2.902, over 560.00 frames. , ppl: 18.210196432225665] tot_loss[loss=2.268, over 5550003.64 frames. , ppl: 9.658496420845026], batch size: 70
+2022-12-14 00:40:37,840 INFO [train.py:421] (2/8) Epoch 10, batch 68600, loss[loss=2.381, over 1400.00 frames. , ppl: 10.816917708407102] tot_loss[loss=2.268, over 5535516.02 frames. , ppl: 9.66217361760269], batch size: 70
+2022-12-14 00:42:17,523 INFO [train.py:421] (2/8) Epoch 10, batch 68800, loss[loss=2.339, over 2240.00 frames. , ppl: 10.368500013161292] tot_loss[loss=2.267, over 5558237.81 frames. , ppl: 9.654677390175097], batch size: 70
+2022-12-14 00:44:01,098 INFO [train.py:421] (2/8) Epoch 10, batch 69000, loss[loss=2.388, over 1050.00 frames. , ppl: 10.893060444980435] tot_loss[loss=2.268, over 5569415.31 frames. , ppl: 9.655757360316489], batch size: 70
+2022-12-14 00:44:01,099 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:44:01,845 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60023105535575
+2022-12-14 00:45:40,104 INFO [train.py:421] (2/8) Epoch 10, batch 69200, loss[loss=2.273, over 2660.00 frames. , ppl: 9.708148223996224] tot_loss[loss=2.268, over 5588394.96 frames. , ppl: 9.657585584428672], batch size: 70
+2022-12-14 00:47:14,231 INFO [train.py:421] (2/8) Epoch 10, batch 69400, loss[loss=2.244, over 3290.00 frames. , ppl: 9.429625787457068] tot_loss[loss=2.269, over 5507180.25 frames. , ppl: 9.67039228672611], batch size: 70
+2022-12-14 00:48:53,974 INFO [train.py:421] (2/8) Epoch 10, batch 69600, loss[loss=2.227, over 4970.00 frames. , ppl: 9.271431575485684] tot_loss[loss=2.269, over 5509241.59 frames. , ppl: 9.667917945886678], batch size: 70
+2022-12-14 00:50:33,906 INFO [train.py:421] (2/8) Epoch 10, batch 69800, loss[loss=2.209, over 2800.00 frames. , ppl: 9.10683614069927] tot_loss[loss=2.268, over 5507826.02 frames. , ppl: 9.663257871687003], batch size: 70
+2022-12-14 00:52:14,676 INFO [train.py:421] (2/8) Epoch 10, batch 70000, loss[loss=2.387, over 1400.00 frames. , ppl: 10.88310856226002] tot_loss[loss=2.268, over 5529929.09 frames. , ppl: 9.662595302017776], batch size: 70
+2022-12-14 00:52:14,676 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 00:52:15,422 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.601824022965605
+2022-12-14 00:53:54,452 INFO [train.py:421] (2/8) Epoch 10, batch 70200, loss[loss=2.361, over 2380.00 frames. , ppl: 10.596623218036145] tot_loss[loss=2.268, over 5546632.57 frames. , ppl: 9.65889388029094], batch size: 70
+2022-12-14 00:55:36,204 INFO [train.py:421] (2/8) Epoch 10, batch 70400, loss[loss=2.444, over 1050.00 frames. , ppl: 11.521201976874451] tot_loss[loss=2.267, over 5528146.76 frames. , ppl: 9.654638078255436], batch size: 70
+2022-12-14 00:57:16,764 INFO [train.py:421] (2/8) Epoch 10, batch 70600, loss[loss=2.568, over 700.00 frames. , ppl: 13.046068855790057] tot_loss[loss=2.267, over 5528684.18 frames. , ppl: 9.653203135884617], batch size: 70
+2022-12-14 00:58:55,099 INFO [train.py:421] (2/8) Epoch 10, batch 70800, loss[loss=2.371, over 1890.00 frames. , ppl: 10.706592626140544] tot_loss[loss=2.268, over 5481568.82 frames. , ppl: 9.660293728608922], batch size: 70
+2022-12-14 01:00:36,975 INFO [train.py:421] (2/8) Epoch 10, batch 71000, loss[loss=2.503, over 1330.00 frames. , ppl: 12.22025815343647] tot_loss[loss=2.268, over 5491259.29 frames. , ppl: 9.66063151187901], batch size: 70
+2022-12-14 01:00:36,976 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:00:37,735 INFO [train.py:452] (2/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.609851142052824
+2022-12-14 01:02:17,478 INFO [train.py:421] (2/8) Epoch 10, batch 71200, loss[loss=2.321, over 1610.00 frames. , ppl: 10.182704975463347] tot_loss[loss=2.268, over 5494952.57 frames. , ppl: 9.661230477880961], batch size: 70
+2022-12-14 01:03:55,203 INFO [train.py:421] (2/8) Epoch 10, batch 71400, loss[loss=3.005, over 630.00 frames. , ppl: 20.182708720258468] tot_loss[loss=2.268, over 5526259.45 frames. , ppl: 9.656634283443799], batch size: 70
+2022-12-14 01:05:31,328 INFO [train.py:421] (2/8) Epoch 10, batch 71600, loss[loss=2.353, over 1680.00 frames. , ppl: 10.518192729836215] tot_loss[loss=2.267, over 5526247.49 frames. , ppl: 9.654861158467991], batch size: 70
+2022-12-14 01:07:12,759 INFO [train.py:421] (2/8) Epoch 10, batch 71800, loss[loss=2.358, over 2240.00 frames. , ppl: 10.56463518385544] tot_loss[loss=2.268, over 5477689.26 frames. , ppl: 9.66415309500225], batch size: 70
+2022-12-14 01:08:27,258 INFO [train.py:421] (2/8) Epoch 11, batch 0, loss[loss=2.386, over 3010.00 frames. , ppl: 10.867754476864917] tot_loss[loss=2.386, over 3010.00 frames. , ppl: 10.867754476864917], batch size: 70
+2022-12-14 01:10:06,184 INFO [train.py:421] (2/8) Epoch 11, batch 200, loss[loss=2.172, over 4130.00 frames. , ppl: 8.779156268682295] tot_loss[loss=2.256, over 530598.05 frames. , ppl: 9.541530473056334], batch size: 70
+2022-12-14 01:11:46,144 INFO [train.py:421] (2/8) Epoch 11, batch 400, loss[loss=3.662, over 420.00 frames. , ppl: 38.94779441744867] tot_loss[loss=2.264, over 976416.14 frames. , ppl: 9.625261054030139], batch size: 70
+2022-12-14 01:13:23,773 INFO [train.py:421] (2/8) Epoch 11, batch 600, loss[loss=2.401, over 1190.00 frames. , ppl: 11.033720088965296] tot_loss[loss=2.265, over 1373052.52 frames. , ppl: 9.634685314371959], batch size: 70
+2022-12-14 01:15:02,879 INFO [train.py:421] (2/8) Epoch 11, batch 800, loss[loss=2.203, over 4900.00 frames. , ppl: 9.05067738961395] tot_loss[loss=2.263, over 1760035.87 frames. , ppl: 9.61437962702303], batch size: 70
+2022-12-14 01:16:43,479 INFO [train.py:421] (2/8) Epoch 11, batch 1000, loss[loss=2.396, over 1540.00 frames. , ppl: 10.974908826343533] tot_loss[loss=2.266, over 2085148.60 frames. , ppl: 9.636338446696081], batch size: 70
+2022-12-14 01:16:43,479 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:16:44,237 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.619912276163067
+2022-12-14 01:18:27,827 INFO [train.py:421] (2/8) Epoch 11, batch 1200, loss[loss=2.626, over 770.00 frames. , ppl: 13.821468629628882] tot_loss[loss=2.26, over 2479438.56 frames. , ppl: 9.584454130448089], batch size: 70
+2022-12-14 01:20:08,976 INFO [train.py:421] (2/8) Epoch 11, batch 1400, loss[loss=2.279, over 2170.00 frames. , ppl: 9.764966837590245] tot_loss[loss=2.261, over 2775777.27 frames. , ppl: 9.594481949685845], batch size: 70
+2022-12-14 01:21:47,194 INFO [train.py:421] (2/8) Epoch 11, batch 1600, loss[loss=2.676, over 770.00 frames. , ppl: 14.527132359326105] tot_loss[loss=2.262, over 3010977.73 frames. , ppl: 9.605519818945577], batch size: 70
+2022-12-14 01:23:27,106 INFO [train.py:421] (2/8) Epoch 11, batch 1800, loss[loss=2.176, over 2870.00 frames. , ppl: 8.810677277051274] tot_loss[loss=2.262, over 3247604.85 frames. , ppl: 9.600688550859163], batch size: 70
+2022-12-14 01:25:04,658 INFO [train.py:421] (2/8) Epoch 11, batch 2000, loss[loss=2.19, over 5460.00 frames. , ppl: 8.936103957201004] tot_loss[loss=2.262, over 3447616.71 frames. , ppl: 9.605171636384243], batch size: 70
+2022-12-14 01:25:04,658 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:25:05,404 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.623191889826966
+2022-12-14 01:26:47,629 INFO [train.py:421] (2/8) Epoch 11, batch 2200, loss[loss=2.354, over 1400.00 frames. , ppl: 10.53108178710747] tot_loss[loss=2.263, over 3625112.95 frames. , ppl: 9.61240318418439], batch size: 70
+2022-12-14 01:28:26,720 INFO [train.py:421] (2/8) Epoch 11, batch 2400, loss[loss=2.344, over 1540.00 frames. , ppl: 10.424809581221252] tot_loss[loss=2.262, over 3810075.37 frames. , ppl: 9.603852895997825], batch size: 70
+2022-12-14 01:30:12,132 INFO [train.py:421] (2/8) Epoch 11, batch 2600, loss[loss=2.322, over 2940.00 frames. , ppl: 10.198737765776011] tot_loss[loss=2.263, over 3960705.16 frames. , ppl: 9.610073561112753], batch size: 70
+2022-12-14 01:31:49,627 INFO [train.py:421] (2/8) Epoch 11, batch 2800, loss[loss=2.128, over 5950.00 frames. , ppl: 8.3957986557124] tot_loss[loss=2.265, over 4053611.49 frames. , ppl: 9.628631302098114], batch size: 70
+2022-12-14 01:33:29,896 INFO [train.py:421] (2/8) Epoch 11, batch 3000, loss[loss=2.211, over 2660.00 frames. , ppl: 9.120823459621294] tot_loss[loss=2.264, over 4224010.92 frames. , ppl: 9.617148177845557], batch size: 70
+2022-12-14 01:33:29,897 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:33:30,660 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612494194035516
+2022-12-14 01:35:15,619 INFO [train.py:421] (2/8) Epoch 11, batch 3200, loss[loss=3.023, over 700.00 frames. , ppl: 20.56056806353132] tot_loss[loss=2.265, over 4298563.32 frames. , ppl: 9.631167351547694], batch size: 70
+2022-12-14 01:36:56,901 INFO [train.py:421] (2/8) Epoch 11, batch 3400, loss[loss=2.185, over 3850.00 frames. , ppl: 8.893770209635981] tot_loss[loss=2.266, over 4376864.32 frames. , ppl: 9.645005876014288], batch size: 70
+2022-12-14 01:38:36,970 INFO [train.py:421] (2/8) Epoch 11, batch 3600, loss[loss=2.388, over 1820.00 frames. , ppl: 10.89612926532941] tot_loss[loss=2.266, over 4487041.55 frames. , ppl: 9.638554631423656], batch size: 70
+2022-12-14 01:40:18,033 INFO [train.py:421] (2/8) Epoch 11, batch 3800, loss[loss=4.191, over 350.00 frames. , ppl: 66.1031346835429] tot_loss[loss=2.265, over 4578354.68 frames. , ppl: 9.635505479964136], batch size: 70
+2022-12-14 01:42:00,390 INFO [train.py:421] (2/8) Epoch 11, batch 4000, loss[loss=2.164, over 12110.00 frames. , ppl: 8.706383851091253] tot_loss[loss=2.266, over 4643218.03 frames. , ppl: 9.641225250485778], batch size: 70
+2022-12-14 01:42:00,390 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:42:01,137 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606853338726559
+2022-12-14 01:43:42,450 INFO [train.py:421] (2/8) Epoch 11, batch 4200, loss[loss=2.142, over 6440.00 frames. , ppl: 8.51764104690255] tot_loss[loss=2.264, over 4767708.18 frames. , ppl: 9.61897068670743], batch size: 70
+2022-12-14 01:45:19,904 INFO [train.py:421] (2/8) Epoch 11, batch 4400, loss[loss=2.63, over 840.00 frames. , ppl: 13.867109298327378] tot_loss[loss=2.265, over 4795231.29 frames. , ppl: 9.62660455971994], batch size: 70
+2022-12-14 01:46:59,747 INFO [train.py:421] (2/8) Epoch 11, batch 4600, loss[loss=2.381, over 1610.00 frames. , ppl: 10.818889913500495] tot_loss[loss=2.266, over 4811794.88 frames. , ppl: 9.638654427093783], batch size: 70
+2022-12-14 01:48:42,457 INFO [train.py:421] (2/8) Epoch 11, batch 4800, loss[loss=2.169, over 6440.00 frames. , ppl: 8.746685131624627] tot_loss[loss=2.265, over 4888711.23 frames. , ppl: 9.632603962356812], batch size: 70
+2022-12-14 01:50:23,264 INFO [train.py:421] (2/8) Epoch 11, batch 5000, loss[loss=2.421, over 1330.00 frames. , ppl: 11.259741060650937] tot_loss[loss=2.267, over 4897617.91 frames. , ppl: 9.650523992429934], batch size: 70
+2022-12-14 01:50:23,265 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:50:24,027 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.615150780970064
+2022-12-14 01:52:05,450 INFO [train.py:421] (2/8) Epoch 11, batch 5200, loss[loss=2.481, over 700.00 frames. , ppl: 11.955478944564291] tot_loss[loss=2.267, over 4958282.84 frames. , ppl: 9.646584983620967], batch size: 70
+2022-12-14 01:53:46,205 INFO [train.py:421] (2/8) Epoch 11, batch 5400, loss[loss=2.106, over 5040.00 frames. , ppl: 8.218390569731593] tot_loss[loss=2.267, over 4962170.49 frames. , ppl: 9.655169620520459], batch size: 70
+2022-12-14 01:55:25,271 INFO [train.py:421] (2/8) Epoch 11, batch 5600, loss[loss=2.16, over 4760.00 frames. , ppl: 8.672163474531358] tot_loss[loss=2.267, over 5019439.20 frames. , ppl: 9.64660790756475], batch size: 70
+2022-12-14 01:57:09,546 INFO [train.py:421] (2/8) Epoch 11, batch 5800, loss[loss=2.824, over 840.00 frames. , ppl: 16.8517541056236] tot_loss[loss=2.266, over 5060775.58 frames. , ppl: 9.642310370811119], batch size: 70
+2022-12-14 01:58:49,576 INFO [train.py:421] (2/8) Epoch 11, batch 6000, loss[loss=2.492, over 910.00 frames. , ppl: 12.081820703211523] tot_loss[loss=2.264, over 5150560.96 frames. , ppl: 9.623489393710956], batch size: 70
+2022-12-14 01:58:49,577 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 01:58:50,322 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606113986457235
+2022-12-14 02:00:27,433 INFO [train.py:421] (2/8) Epoch 11, batch 6200, loss[loss=2.473, over 1470.00 frames. , ppl: 11.854177438315068] tot_loss[loss=2.264, over 5185733.01 frames. , ppl: 9.619218451689566], batch size: 70
+2022-12-14 02:02:09,157 INFO [train.py:421] (2/8) Epoch 11, batch 6400, loss[loss=2.287, over 1750.00 frames. , ppl: 9.8414682340845] tot_loss[loss=2.265, over 5171071.52 frames. , ppl: 9.629904052015469], batch size: 70
+2022-12-14 02:03:48,306 INFO [train.py:421] (2/8) Epoch 11, batch 6600, loss[loss=2.232, over 3570.00 frames. , ppl: 9.317784075127358] tot_loss[loss=2.265, over 5172009.55 frames. , ppl: 9.632461823192791], batch size: 70
+2022-12-14 02:05:25,882 INFO [train.py:421] (2/8) Epoch 11, batch 6800, loss[loss=2.505, over 980.00 frames. , ppl: 12.24698964864637] tot_loss[loss=2.265, over 5175238.17 frames. , ppl: 9.635141333984631], batch size: 70
+2022-12-14 02:07:09,115 INFO [train.py:421] (2/8) Epoch 11, batch 7000, loss[loss=2.455, over 1540.00 frames. , ppl: 11.641827465963086] tot_loss[loss=2.266, over 5192235.08 frames. , ppl: 9.63639854648319], batch size: 70
+2022-12-14 02:07:09,116 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:07:09,879 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616433092260849
+2022-12-14 02:08:51,379 INFO [train.py:421] (2/8) Epoch 11, batch 7200, loss[loss=2.227, over 3150.00 frames. , ppl: 9.271376781244737] tot_loss[loss=2.266, over 5197920.75 frames. , ppl: 9.64010856311593], batch size: 70
+2022-12-14 02:10:28,062 INFO [train.py:421] (2/8) Epoch 11, batch 7400, loss[loss=2.26, over 3010.00 frames. , ppl: 9.579009611302054] tot_loss[loss=2.266, over 5229462.57 frames. , ppl: 9.645551887580565], batch size: 70
+2022-12-14 02:12:08,226 INFO [train.py:421] (2/8) Epoch 11, batch 7600, loss[loss=2.642, over 770.00 frames. , ppl: 14.04385764653429] tot_loss[loss=2.266, over 5261048.83 frames. , ppl: 9.644581613033463], batch size: 70
+2022-12-14 02:13:46,076 INFO [train.py:421] (2/8) Epoch 11, batch 7800, loss[loss=2.358, over 2100.00 frames. , ppl: 10.574305371136191] tot_loss[loss=2.266, over 5281406.88 frames. , ppl: 9.641362703019997], batch size: 70
+2022-12-14 02:15:21,877 INFO [train.py:421] (2/8) Epoch 11, batch 8000, loss[loss=2.354, over 3150.00 frames. , ppl: 10.527410293629076] tot_loss[loss=2.267, over 5279135.41 frames. , ppl: 9.646575449832762], batch size: 70
+2022-12-14 02:15:21,878 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:15:22,633 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.595916733986652
+2022-12-14 02:17:02,599 INFO [train.py:421] (2/8) Epoch 11, batch 8200, loss[loss=2.452, over 1680.00 frames. , ppl: 11.610502886863117] tot_loss[loss=2.266, over 5312078.29 frames. , ppl: 9.638593515016707], batch size: 70
+2022-12-14 02:18:44,665 INFO [train.py:421] (2/8) Epoch 11, batch 8400, loss[loss=2.195, over 9240.00 frames. , ppl: 8.98317611274824] tot_loss[loss=2.266, over 5341829.09 frames. , ppl: 9.640527661158723], batch size: 70
+2022-12-14 02:20:24,716 INFO [train.py:421] (2/8) Epoch 11, batch 8600, loss[loss=2.184, over 3290.00 frames. , ppl: 8.882836357690394] tot_loss[loss=2.266, over 5356801.59 frames. , ppl: 9.642057952640664], batch size: 70
+2022-12-14 02:22:03,173 INFO [train.py:421] (2/8) Epoch 11, batch 8800, loss[loss=2.323, over 2100.00 frames. , ppl: 10.210240931909873] tot_loss[loss=2.267, over 5340532.74 frames. , ppl: 9.650502972333726], batch size: 70
+2022-12-14 02:23:44,978 INFO [train.py:421] (2/8) Epoch 11, batch 9000, loss[loss=2.37, over 1540.00 frames. , ppl: 10.699073380708569] tot_loss[loss=2.267, over 5345356.47 frames. , ppl: 9.649429036815297], batch size: 70
+2022-12-14 02:23:44,978 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:23:45,723 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.615415483833607
+2022-12-14 02:25:22,447 INFO [train.py:421] (2/8) Epoch 11, batch 9200, loss[loss=2.416, over 840.00 frames. , ppl: 11.195685784988024] tot_loss[loss=2.267, over 5365243.39 frames. , ppl: 9.650961708914418], batch size: 70
+2022-12-14 02:27:01,801 INFO [train.py:421] (2/8) Epoch 11, batch 9400, loss[loss=2.302, over 1890.00 frames. , ppl: 9.995939122662987] tot_loss[loss=2.269, over 5349438.27 frames. , ppl: 9.66571373833882], batch size: 70
+2022-12-14 02:28:45,755 INFO [train.py:421] (2/8) Epoch 11, batch 9600, loss[loss=2.272, over 1610.00 frames. , ppl: 9.699575682020146] tot_loss[loss=2.268, over 5386417.86 frames. , ppl: 9.660197255769225], batch size: 70
+2022-12-14 02:30:24,416 INFO [train.py:421] (2/8) Epoch 11, batch 9800, loss[loss=2.116, over 8610.00 frames. , ppl: 8.30161652340162] tot_loss[loss=2.269, over 5375712.09 frames. , ppl: 9.667627944520602], batch size: 70
+2022-12-14 02:32:06,052 INFO [train.py:421] (2/8) Epoch 11, batch 10000, loss[loss=2.21, over 3150.00 frames. , ppl: 9.112803198047374] tot_loss[loss=2.268, over 5401416.23 frames. , ppl: 9.662886263949948], batch size: 70
+2022-12-14 02:32:06,052 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:32:06,813 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.607311196929732
+2022-12-14 02:33:49,164 INFO [train.py:421] (2/8) Epoch 11, batch 10200, loss[loss=2.195, over 6300.00 frames. , ppl: 8.982693567603679] tot_loss[loss=2.268, over 5408042.27 frames. , ppl: 9.661876438569465], batch size: 70
+2022-12-14 02:35:28,712 INFO [train.py:421] (2/8) Epoch 11, batch 10400, loss[loss=2.45, over 1050.00 frames. , ppl: 11.586256004511611] tot_loss[loss=2.269, over 5369028.69 frames. , ppl: 9.66787164385342], batch size: 70
+2022-12-14 02:37:09,001 INFO [train.py:421] (2/8) Epoch 11, batch 10600, loss[loss=2.795, over 630.00 frames. , ppl: 16.37057205438153] tot_loss[loss=2.27, over 5319810.79 frames. , ppl: 9.682985981698872], batch size: 70
+2022-12-14 02:38:52,752 INFO [train.py:421] (2/8) Epoch 11, batch 10800, loss[loss=2.204, over 13510.00 frames. , ppl: 9.064863039111096] tot_loss[loss=2.27, over 5343258.69 frames. , ppl: 9.67747496025384], batch size: 70
+2022-12-14 02:40:34,269 INFO [train.py:421] (2/8) Epoch 11, batch 11000, loss[loss=2.217, over 7350.00 frames. , ppl: 9.18003713654969] tot_loss[loss=2.268, over 5419130.35 frames. , ppl: 9.658558793743829], batch size: 70
+2022-12-14 02:40:34,270 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:40:35,016 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604206151775301
+2022-12-14 02:42:14,871 INFO [train.py:421] (2/8) Epoch 11, batch 11200, loss[loss=2.551, over 980.00 frames. , ppl: 12.813534345175928] tot_loss[loss=2.267, over 5456349.10 frames. , ppl: 9.648663272457146], batch size: 70
+2022-12-14 02:43:54,215 INFO [train.py:421] (2/8) Epoch 11, batch 11400, loss[loss=2.43, over 910.00 frames. , ppl: 11.364311465890358] tot_loss[loss=2.265, over 5507271.60 frames. , ppl: 9.63269603749169], batch size: 70
+2022-12-14 02:45:34,727 INFO [train.py:421] (2/8) Epoch 11, batch 11600, loss[loss=2.254, over 2310.00 frames. , ppl: 9.528923500219063] tot_loss[loss=2.266, over 5453145.81 frames. , ppl: 9.64051568697319], batch size: 70
+2022-12-14 02:47:16,923 INFO [train.py:421] (2/8) Epoch 11, batch 11800, loss[loss=2.272, over 2030.00 frames. , ppl: 9.697452420000907] tot_loss[loss=2.264, over 5496711.46 frames. , ppl: 9.624462196468668], batch size: 70
+2022-12-14 02:48:54,556 INFO [train.py:421] (2/8) Epoch 11, batch 12000, loss[loss=2.302, over 2380.00 frames. , ppl: 9.995285784278959] tot_loss[loss=2.265, over 5475831.81 frames. , ppl: 9.635917530586301], batch size: 70
+2022-12-14 02:48:54,557 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:48:55,303 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621384617486425
+2022-12-14 02:50:34,274 INFO [train.py:421] (2/8) Epoch 11, batch 12200, loss[loss=2.525, over 980.00 frames. , ppl: 12.489302138774635] tot_loss[loss=2.264, over 5529655.85 frames. , ppl: 9.617171815701884], batch size: 70
+2022-12-14 02:52:10,839 INFO [train.py:421] (2/8) Epoch 11, batch 12400, loss[loss=2.327, over 1750.00 frames. , ppl: 10.247971659325438] tot_loss[loss=2.265, over 5471580.38 frames. , ppl: 9.631597833292714], batch size: 70
+2022-12-14 02:53:50,841 INFO [train.py:421] (2/8) Epoch 11, batch 12600, loss[loss=2.171, over 4270.00 frames. , ppl: 8.763770903309727] tot_loss[loss=2.264, over 5508450.23 frames. , ppl: 9.621240777912833], batch size: 70
+2022-12-14 02:55:29,957 INFO [train.py:421] (2/8) Epoch 11, batch 12800, loss[loss=2.188, over 4830.00 frames. , ppl: 8.913587794791711] tot_loss[loss=2.265, over 5454610.68 frames. , ppl: 9.62966733484706], batch size: 70
+2022-12-14 02:57:14,314 INFO [train.py:421] (2/8) Epoch 11, batch 13000, loss[loss=2.569, over 1260.00 frames. , ppl: 13.051611305627082] tot_loss[loss=2.264, over 5482124.01 frames. , ppl: 9.621346757113546], batch size: 70
+2022-12-14 02:57:14,314 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 02:57:15,076 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.593847627631558
+2022-12-14 02:58:53,991 INFO [train.py:421] (2/8) Epoch 11, batch 13200, loss[loss=2.382, over 1680.00 frames. , ppl: 10.826570543496063] tot_loss[loss=2.265, over 5486246.34 frames. , ppl: 9.627783608898808], batch size: 70
+2022-12-14 03:00:37,502 INFO [train.py:421] (2/8) Epoch 11, batch 13400, loss[loss=2.206, over 3150.00 frames. , ppl: 9.08114319122563] tot_loss[loss=2.265, over 5489866.36 frames. , ppl: 9.632869337472027], batch size: 70
+2022-12-14 03:02:16,685 INFO [train.py:421] (2/8) Epoch 11, batch 13600, loss[loss=2.174, over 7490.00 frames. , ppl: 8.794023392576502] tot_loss[loss=2.265, over 5508625.23 frames. , ppl: 9.628324431204051], batch size: 70
+2022-12-14 03:03:48,771 INFO [train.py:421] (2/8) Epoch 11, batch 13800, loss[loss=2.469, over 1260.00 frames. , ppl: 11.81154289405115] tot_loss[loss=2.266, over 5501932.55 frames. , ppl: 9.637959214390468], batch size: 70
+2022-12-14 03:05:29,080 INFO [train.py:421] (2/8) Epoch 11, batch 14000, loss[loss=2.399, over 1680.00 frames. , ppl: 11.010586172870472] tot_loss[loss=2.265, over 5510525.55 frames. , ppl: 9.629799435013984], batch size: 70
+2022-12-14 03:05:29,080 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:05:29,843 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617450808382541
+2022-12-14 03:07:07,756 INFO [train.py:421] (2/8) Epoch 11, batch 14200, loss[loss=2.368, over 1960.00 frames. , ppl: 10.676031008049293] tot_loss[loss=2.266, over 5478218.47 frames. , ppl: 9.637528466544664], batch size: 70
+2022-12-14 03:08:46,878 INFO [train.py:421] (2/8) Epoch 11, batch 14400, loss[loss=2.323, over 1400.00 frames. , ppl: 10.203778486199937] tot_loss[loss=2.265, over 5475232.47 frames. , ppl: 9.63472276876402], batch size: 70
+2022-12-14 03:10:30,210 INFO [train.py:421] (2/8) Epoch 11, batch 14600, loss[loss=2.105, over 7350.00 frames. , ppl: 8.204059060033405] tot_loss[loss=2.265, over 5487157.01 frames. , ppl: 9.630225231065134], batch size: 70
+2022-12-14 03:12:09,168 INFO [train.py:421] (2/8) Epoch 11, batch 14800, loss[loss=2.331, over 2240.00 frames. , ppl: 10.284372186322647] tot_loss[loss=2.264, over 5493246.73 frames. , ppl: 9.618848422954674], batch size: 70
+2022-12-14 03:13:48,745 INFO [train.py:421] (2/8) Epoch 11, batch 15000, loss[loss=2.446, over 980.00 frames. , ppl: 11.540493973917947] tot_loss[loss=2.264, over 5504314.27 frames. , ppl: 9.619598649084999], batch size: 70
+2022-12-14 03:13:48,746 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:13:49,508 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.603182730483441
+2022-12-14 03:15:28,759 INFO [train.py:421] (2/8) Epoch 11, batch 15200, loss[loss=2.133, over 5810.00 frames. , ppl: 8.444187011009538] tot_loss[loss=2.264, over 5498743.96 frames. , ppl: 9.625648581182444], batch size: 70
+2022-12-14 03:17:08,551 INFO [train.py:421] (2/8) Epoch 11, batch 15400, loss[loss=2.265, over 3430.00 frames. , ppl: 9.632888479854115] tot_loss[loss=2.266, over 5472209.77 frames. , ppl: 9.64261550412026], batch size: 70
+2022-12-14 03:18:44,459 INFO [train.py:421] (2/8) Epoch 11, batch 15600, loss[loss=2.359, over 1050.00 frames. , ppl: 10.585434306854724] tot_loss[loss=2.268, over 5428229.53 frames. , ppl: 9.658061550550363], batch size: 70
+2022-12-14 03:20:25,762 INFO [train.py:421] (2/8) Epoch 11, batch 15800, loss[loss=2.256, over 4620.00 frames. , ppl: 9.542470537743858] tot_loss[loss=2.266, over 5465873.65 frames. , ppl: 9.64313454873313], batch size: 70
+2022-12-14 03:22:04,990 INFO [train.py:421] (2/8) Epoch 11, batch 16000, loss[loss=3.594, over 420.00 frames. , ppl: 36.39393734232458] tot_loss[loss=2.266, over 5456901.83 frames. , ppl: 9.64276373690179], batch size: 70
+2022-12-14 03:22:04,991 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:22:05,752 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608784453983127
+2022-12-14 03:23:48,825 INFO [train.py:421] (2/8) Epoch 11, batch 16200, loss[loss=2.343, over 3220.00 frames. , ppl: 10.40836666176323] tot_loss[loss=2.266, over 5461720.43 frames. , ppl: 9.640172736870085], batch size: 70
+2022-12-14 03:25:29,632 INFO [train.py:421] (2/8) Epoch 11, batch 16400, loss[loss=2.206, over 2240.00 frames. , ppl: 9.082693253161121] tot_loss[loss=2.265, over 5483776.85 frames. , ppl: 9.634177579028094], batch size: 70
+2022-12-14 03:27:08,106 INFO [train.py:421] (2/8) Epoch 11, batch 16600, loss[loss=2.503, over 1120.00 frames. , ppl: 12.21839166977328] tot_loss[loss=2.266, over 5484238.87 frames. , ppl: 9.637340155714135], batch size: 70
+2022-12-14 03:28:48,329 INFO [train.py:421] (2/8) Epoch 11, batch 16800, loss[loss=2.447, over 980.00 frames. , ppl: 11.55891155197307] tot_loss[loss=2.265, over 5516108.13 frames. , ppl: 9.631285940552704], batch size: 70
+2022-12-14 03:30:30,504 INFO [train.py:421] (2/8) Epoch 11, batch 17000, loss[loss=2.19, over 4970.00 frames. , ppl: 8.93131422036726] tot_loss[loss=2.264, over 5545238.03 frames. , ppl: 9.62378562906941], batch size: 70
+2022-12-14 03:30:30,504 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:30:31,251 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599325980992132
+2022-12-14 03:32:09,722 INFO [train.py:421] (2/8) Epoch 11, batch 17200, loss[loss=2.295, over 2310.00 frames. , ppl: 9.924678101259133] tot_loss[loss=2.265, over 5533130.95 frames. , ppl: 9.629390532428197], batch size: 70
+2022-12-14 03:33:49,128 INFO [train.py:421] (2/8) Epoch 11, batch 17400, loss[loss=2.215, over 3920.00 frames. , ppl: 9.16319588476052] tot_loss[loss=2.265, over 5523565.19 frames. , ppl: 9.632763413809958], batch size: 70
+2022-12-14 03:35:30,987 INFO [train.py:421] (2/8) Epoch 11, batch 17600, loss[loss=2.383, over 1330.00 frames. , ppl: 10.840196588786734] tot_loss[loss=2.265, over 5531899.96 frames. , ppl: 9.632100167751556], batch size: 70
+2022-12-14 03:37:07,956 INFO [train.py:421] (2/8) Epoch 11, batch 17800, loss[loss=2.598, over 840.00 frames. , ppl: 13.433429640359627] tot_loss[loss=2.266, over 5492612.28 frames. , ppl: 9.645436486956209], batch size: 70
+2022-12-14 03:38:46,541 INFO [train.py:421] (2/8) Epoch 11, batch 18000, loss[loss=2.177, over 6440.00 frames. , ppl: 8.816540195184555] tot_loss[loss=2.267, over 5485942.95 frames. , ppl: 9.653814763341066], batch size: 70
+2022-12-14 03:38:46,542 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:38:47,290 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.611424367463018
+2022-12-14 03:40:27,927 INFO [train.py:421] (2/8) Epoch 11, batch 18200, loss[loss=2.172, over 4690.00 frames. , ppl: 8.773639656283631] tot_loss[loss=2.267, over 5477843.77 frames. , ppl: 9.652476189211189], batch size: 70
+2022-12-14 03:42:05,032 INFO [train.py:421] (2/8) Epoch 11, batch 18400, loss[loss=2.222, over 4550.00 frames. , ppl: 9.223169573298076] tot_loss[loss=2.267, over 5445672.88 frames. , ppl: 9.653027816565295], batch size: 70
+2022-12-14 03:43:44,909 INFO [train.py:421] (2/8) Epoch 11, batch 18600, loss[loss=2.492, over 1050.00 frames. , ppl: 12.082159108878297] tot_loss[loss=2.268, over 5407890.57 frames. , ppl: 9.660346036627235], batch size: 70
+2022-12-14 03:45:24,640 INFO [train.py:421] (2/8) Epoch 11, batch 18800, loss[loss=2.219, over 3920.00 frames. , ppl: 9.196889637293065] tot_loss[loss=2.268, over 5412913.40 frames. , ppl: 9.660337725966274], batch size: 70
+2022-12-14 03:47:03,784 INFO [train.py:421] (2/8) Epoch 11, batch 19000, loss[loss=2.222, over 4900.00 frames. , ppl: 9.229642179195805] tot_loss[loss=2.269, over 5378687.26 frames. , ppl: 9.673130052085178], batch size: 70
+2022-12-14 03:47:03,784 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:47:04,530 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604402319978165
+2022-12-14 03:48:43,945 INFO [train.py:421] (2/8) Epoch 11, batch 19200, loss[loss=2.256, over 3990.00 frames. , ppl: 9.546444124465966] tot_loss[loss=2.27, over 5383260.51 frames. , ppl: 9.675355693279426], batch size: 70
+2022-12-14 03:50:21,980 INFO [train.py:421] (2/8) Epoch 11, batch 19400, loss[loss=2.327, over 1330.00 frames. , ppl: 10.25025549462815] tot_loss[loss=2.27, over 5375375.61 frames. , ppl: 9.681726346950935], batch size: 70
+2022-12-14 03:52:05,991 INFO [train.py:421] (2/8) Epoch 11, batch 19600, loss[loss=2.261, over 1960.00 frames. , ppl: 9.58989606152005] tot_loss[loss=2.269, over 5390987.36 frames. , ppl: 9.671657452726423], batch size: 70
+2022-12-14 03:53:44,117 INFO [train.py:421] (2/8) Epoch 11, batch 19800, loss[loss=2.288, over 2940.00 frames. , ppl: 9.85368127155703] tot_loss[loss=2.271, over 5357225.82 frames. , ppl: 9.684311944366575], batch size: 70
+2022-12-14 03:55:25,027 INFO [train.py:421] (2/8) Epoch 11, batch 20000, loss[loss=2.239, over 5110.00 frames. , ppl: 9.38521070544731] tot_loss[loss=2.27, over 5369848.52 frames. , ppl: 9.679877285272665], batch size: 70
+2022-12-14 03:55:25,027 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 03:55:25,775 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61202043965363
+2022-12-14 03:57:05,991 INFO [train.py:421] (2/8) Epoch 11, batch 20200, loss[loss=2.747, over 630.00 frames. , ppl: 15.603448374641575] tot_loss[loss=2.269, over 5390816.29 frames. , ppl: 9.672148797553664], batch size: 70
+2022-12-14 03:58:46,598 INFO [train.py:421] (2/8) Epoch 11, batch 20400, loss[loss=2.3, over 1820.00 frames. , ppl: 9.975683768738017] tot_loss[loss=2.269, over 5377060.31 frames. , ppl: 9.66995992413354], batch size: 70
+2022-12-14 04:00:28,898 INFO [train.py:421] (2/8) Epoch 11, batch 20600, loss[loss=2.428, over 2170.00 frames. , ppl: 11.338142430398053] tot_loss[loss=2.268, over 5435707.17 frames. , ppl: 9.657112326170273], batch size: 70
+2022-12-14 04:02:14,437 INFO [train.py:421] (2/8) Epoch 11, batch 20800, loss[loss=2.323, over 2170.00 frames. , ppl: 10.208209344753913] tot_loss[loss=2.266, over 5476160.26 frames. , ppl: 9.638280676853077], batch size: 70
+2022-12-14 04:03:51,808 INFO [train.py:421] (2/8) Epoch 11, batch 21000, loss[loss=2.348, over 3080.00 frames. , ppl: 10.466862201193461] tot_loss[loss=2.266, over 5458061.22 frames. , ppl: 9.64453094805651], batch size: 70
+2022-12-14 04:03:51,808 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:03:52,554 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59960729794989
+2022-12-14 04:05:34,808 INFO [train.py:421] (2/8) Epoch 11, batch 21200, loss[loss=2.322, over 2100.00 frames. , ppl: 10.195435337224243] tot_loss[loss=2.266, over 5469780.16 frames. , ppl: 9.64527413644055], batch size: 70
+2022-12-14 04:07:12,357 INFO [train.py:421] (2/8) Epoch 11, batch 21400, loss[loss=2.387, over 1540.00 frames. , ppl: 10.88557986110606] tot_loss[loss=2.267, over 5457489.83 frames. , ppl: 9.652144360542623], batch size: 70
+2022-12-14 04:08:53,885 INFO [train.py:421] (2/8) Epoch 11, batch 21600, loss[loss=2.317, over 2030.00 frames. , ppl: 10.149814742552044] tot_loss[loss=2.268, over 5423699.92 frames. , ppl: 9.659307302644608], batch size: 70
+2022-12-14 04:10:28,568 INFO [train.py:421] (2/8) Epoch 11, batch 21800, loss[loss=2.349, over 1330.00 frames. , ppl: 10.479058301611689] tot_loss[loss=2.268, over 5424266.17 frames. , ppl: 9.662198793022458], batch size: 70
+2022-12-14 04:12:08,385 INFO [train.py:421] (2/8) Epoch 11, batch 22000, loss[loss=2.157, over 7210.00 frames. , ppl: 8.645959180702938] tot_loss[loss=2.267, over 5450450.90 frames. , ppl: 9.652352872518092], batch size: 70
+2022-12-14 04:12:08,385 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:12:09,131 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59981047643547
+2022-12-14 04:13:46,450 INFO [train.py:421] (2/8) Epoch 11, batch 22200, loss[loss=2.171, over 3850.00 frames. , ppl: 8.770057336445694] tot_loss[loss=2.269, over 5397191.02 frames. , ppl: 9.670272330087524], batch size: 70
+2022-12-14 04:15:26,607 INFO [train.py:421] (2/8) Epoch 11, batch 22400, loss[loss=2.309, over 3290.00 frames. , ppl: 10.066480591117548] tot_loss[loss=2.269, over 5395428.80 frames. , ppl: 9.67294975426599], batch size: 70
+2022-12-14 04:17:04,240 INFO [train.py:421] (2/8) Epoch 11, batch 22600, loss[loss=2.312, over 3010.00 frames. , ppl: 10.095357182973657] tot_loss[loss=2.269, over 5400551.34 frames. , ppl: 9.668039286068476], batch size: 70
+2022-12-14 04:18:46,595 INFO [train.py:421] (2/8) Epoch 11, batch 22800, loss[loss=2.747, over 630.00 frames. , ppl: 15.593571156408832] tot_loss[loss=2.268, over 5418270.57 frames. , ppl: 9.660198459061046], batch size: 70
+2022-12-14 04:20:26,418 INFO [train.py:421] (2/8) Epoch 11, batch 23000, loss[loss=2.859, over 700.00 frames. , ppl: 17.448133861456586] tot_loss[loss=2.268, over 5405453.40 frames. , ppl: 9.662245767204103], batch size: 70
+2022-12-14 04:20:26,418 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:20:27,166 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589605722035854
+2022-12-14 04:22:05,159 INFO [train.py:421] (2/8) Epoch 11, batch 23200, loss[loss=2.174, over 6650.00 frames. , ppl: 8.789289118945044] tot_loss[loss=2.268, over 5406988.06 frames. , ppl: 9.659479830255208], batch size: 70
+2022-12-14 04:23:43,566 INFO [train.py:421] (2/8) Epoch 11, batch 23400, loss[loss=2.256, over 2030.00 frames. , ppl: 9.543176743663558] tot_loss[loss=2.269, over 5375592.25 frames. , ppl: 9.670988482162477], batch size: 70
+2022-12-14 04:25:27,479 INFO [train.py:421] (2/8) Epoch 11, batch 23600, loss[loss=2.108, over 6440.00 frames. , ppl: 8.228566997063766] tot_loss[loss=2.268, over 5416233.30 frames. , ppl: 9.656812235169888], batch size: 70
+2022-12-14 04:27:09,128 INFO [train.py:421] (2/8) Epoch 11, batch 23800, loss[loss=2.352, over 1400.00 frames. , ppl: 10.501768444784181] tot_loss[loss=2.266, over 5489693.53 frames. , ppl: 9.64205575415617], batch size: 70
+2022-12-14 04:28:49,639 INFO [train.py:421] (2/8) Epoch 11, batch 24000, loss[loss=2.482, over 980.00 frames. , ppl: 11.962710268689568] tot_loss[loss=2.267, over 5454659.38 frames. , ppl: 9.646833743517005], batch size: 70
+2022-12-14 04:28:49,640 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:28:50,387 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589086260144834
+2022-12-14 04:30:27,487 INFO [train.py:421] (2/8) Epoch 11, batch 24200, loss[loss=2.24, over 3710.00 frames. , ppl: 9.394776359103911] tot_loss[loss=2.267, over 5462977.97 frames. , ppl: 9.64733144274022], batch size: 70
+2022-12-14 04:32:06,789 INFO [train.py:421] (2/8) Epoch 11, batch 24400, loss[loss=2.469, over 1050.00 frames. , ppl: 11.805778319448397] tot_loss[loss=2.267, over 5421675.35 frames. , ppl: 9.654741703022776], batch size: 70
+2022-12-14 04:33:49,455 INFO [train.py:421] (2/8) Epoch 11, batch 24600, loss[loss=2.326, over 2520.00 frames. , ppl: 10.241429451450353] tot_loss[loss=2.266, over 5473420.63 frames. , ppl: 9.644677382340591], batch size: 70
+2022-12-14 04:35:31,656 INFO [train.py:421] (2/8) Epoch 11, batch 24800, loss[loss=2.253, over 3780.00 frames. , ppl: 9.513935575957868] tot_loss[loss=2.268, over 5439381.63 frames. , ppl: 9.65895429460805], batch size: 70
+2022-12-14 04:37:15,799 INFO [train.py:421] (2/8) Epoch 11, batch 25000, loss[loss=2.21, over 5180.00 frames. , ppl: 9.113179828840837] tot_loss[loss=2.266, over 5489139.34 frames. , ppl: 9.643050062355133], batch size: 70
+2022-12-14 04:37:15,800 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:37:16,563 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606207824169385
+2022-12-14 04:38:53,653 INFO [train.py:421] (2/8) Epoch 11, batch 25200, loss[loss=2.329, over 1540.00 frames. , ppl: 10.272156866064941] tot_loss[loss=2.265, over 5519572.84 frames. , ppl: 9.634224683707902], batch size: 70
+2022-12-14 04:40:34,679 INFO [train.py:421] (2/8) Epoch 11, batch 25400, loss[loss=2.309, over 2730.00 frames. , ppl: 10.063296222760835] tot_loss[loss=2.265, over 5517749.51 frames. , ppl: 9.629146580097283], batch size: 70
+2022-12-14 04:42:14,011 INFO [train.py:421] (2/8) Epoch 11, batch 25600, loss[loss=2.207, over 4830.00 frames. , ppl: 9.092193732379927] tot_loss[loss=2.264, over 5525811.97 frames. , ppl: 9.623505043273074], batch size: 70
+2022-12-14 04:43:56,157 INFO [train.py:421] (2/8) Epoch 11, batch 25800, loss[loss=2.441, over 1960.00 frames. , ppl: 11.482770747012442] tot_loss[loss=2.265, over 5510646.20 frames. , ppl: 9.634540511579981], batch size: 70
+2022-12-14 04:45:33,980 INFO [train.py:421] (2/8) Epoch 11, batch 26000, loss[loss=2.231, over 3570.00 frames. , ppl: 9.313452035559155] tot_loss[loss=2.265, over 5531319.44 frames. , ppl: 9.626562795983336], batch size: 70
+2022-12-14 04:45:33,980 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:45:34,741 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612918174211162
+2022-12-14 04:47:14,157 INFO [train.py:421] (2/8) Epoch 11, batch 26200, loss[loss=2.254, over 5320.00 frames. , ppl: 9.527253188389897] tot_loss[loss=2.264, over 5562559.38 frames. , ppl: 9.617203450931543], batch size: 70
+2022-12-14 04:48:55,148 INFO [train.py:421] (2/8) Epoch 11, batch 26400, loss[loss=2.454, over 1260.00 frames. , ppl: 11.63150686751489] tot_loss[loss=2.264, over 5538258.65 frames. , ppl: 9.619939140006071], batch size: 70
+2022-12-14 04:50:33,170 INFO [train.py:421] (2/8) Epoch 11, batch 26600, loss[loss=2.376, over 1960.00 frames. , ppl: 10.766896665057617] tot_loss[loss=2.264, over 5528455.43 frames. , ppl: 9.62579240893154], batch size: 70
+2022-12-14 04:52:12,404 INFO [train.py:421] (2/8) Epoch 11, batch 26800, loss[loss=2.271, over 2310.00 frames. , ppl: 9.686102569533547] tot_loss[loss=2.266, over 5488168.44 frames. , ppl: 9.638697767039512], batch size: 70
+2022-12-14 04:53:51,698 INFO [train.py:421] (2/8) Epoch 11, batch 27000, loss[loss=2.441, over 1400.00 frames. , ppl: 11.478930430441338] tot_loss[loss=2.267, over 5452062.86 frames. , ppl: 9.649817056314633], batch size: 70
+2022-12-14 04:53:51,699 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 04:53:52,445 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.594008084573185
+2022-12-14 04:55:35,141 INFO [train.py:421] (2/8) Epoch 11, batch 27200, loss[loss=2.44, over 1890.00 frames. , ppl: 11.478176234045016] tot_loss[loss=2.267, over 5471980.97 frames. , ppl: 9.64576331135484], batch size: 70
+2022-12-14 04:57:14,469 INFO [train.py:421] (2/8) Epoch 11, batch 27400, loss[loss=2.135, over 7910.00 frames. , ppl: 8.460590508479825] tot_loss[loss=2.267, over 5460701.43 frames. , ppl: 9.646177624451928], batch size: 70
+2022-12-14 04:58:50,881 INFO [train.py:421] (2/8) Epoch 11, batch 27600, loss[loss=2.199, over 4410.00 frames. , ppl: 9.019716967770604] tot_loss[loss=2.267, over 5442881.39 frames. , ppl: 9.65029118723069], batch size: 70
+2022-12-14 05:00:29,842 INFO [train.py:421] (2/8) Epoch 11, batch 27800, loss[loss=2.192, over 5950.00 frames. , ppl: 8.955232975933646] tot_loss[loss=2.268, over 5390283.58 frames. , ppl: 9.663872263204349], batch size: 70
+2022-12-14 05:02:14,491 INFO [train.py:421] (2/8) Epoch 11, batch 28000, loss[loss=2.241, over 4690.00 frames. , ppl: 9.404556254019754] tot_loss[loss=2.268, over 5432863.43 frames. , ppl: 9.659673076924989], batch size: 70
+2022-12-14 05:02:14,491 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:02:15,236 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.594497991840212
+2022-12-14 05:04:00,693 INFO [train.py:421] (2/8) Epoch 11, batch 28200, loss[loss=2.398, over 1820.00 frames. , ppl: 11.004395365111858] tot_loss[loss=2.269, over 5408665.21 frames. , ppl: 9.666116098618495], batch size: 70
+2022-12-14 05:05:42,767 INFO [train.py:421] (2/8) Epoch 11, batch 28400, loss[loss=2.207, over 3360.00 frames. , ppl: 9.092249169447879] tot_loss[loss=2.268, over 5448756.74 frames. , ppl: 9.657452589927003], batch size: 70
+2022-12-14 05:07:20,249 INFO [train.py:421] (2/8) Epoch 11, batch 28600, loss[loss=2.275, over 1890.00 frames. , ppl: 9.726499154465671] tot_loss[loss=2.268, over 5439260.61 frames. , ppl: 9.661747009085547], batch size: 70
+2022-12-14 05:09:04,634 INFO [train.py:421] (2/8) Epoch 11, batch 28800, loss[loss=2.607, over 700.00 frames. , ppl: 13.556724170644543] tot_loss[loss=2.268, over 5443454.44 frames. , ppl: 9.663304290815942], batch size: 70
+2022-12-14 05:10:45,350 INFO [train.py:421] (2/8) Epoch 11, batch 29000, loss[loss=2.27, over 2310.00 frames. , ppl: 9.679604189036617] tot_loss[loss=2.268, over 5442954.54 frames. , ppl: 9.659433709419282], batch size: 70
+2022-12-14 05:10:45,351 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:10:46,097 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.587935312328966
+2022-12-14 05:12:24,279 INFO [train.py:421] (2/8) Epoch 11, batch 29200, loss[loss=2.296, over 2450.00 frames. , ppl: 9.934356304616726] tot_loss[loss=2.267, over 5454737.68 frames. , ppl: 9.653378199783882], batch size: 70
+2022-12-14 05:14:04,310 INFO [train.py:421] (2/8) Epoch 11, batch 29400, loss[loss=2.166, over 3430.00 frames. , ppl: 8.719727499952189] tot_loss[loss=2.269, over 5386850.69 frames. , ppl: 9.66946923868123], batch size: 70
+2022-12-14 05:15:46,598 INFO [train.py:421] (2/8) Epoch 11, batch 29600, loss[loss=2.168, over 7910.00 frames. , ppl: 8.739639965525132] tot_loss[loss=2.269, over 5380859.78 frames. , ppl: 9.673449101207815], batch size: 70
+2022-12-14 05:17:28,220 INFO [train.py:421] (2/8) Epoch 11, batch 29800, loss[loss=2.151, over 5180.00 frames. , ppl: 8.591256955780587] tot_loss[loss=2.27, over 5375406.73 frames. , ppl: 9.681924165474086], batch size: 70
+2022-12-14 05:19:07,019 INFO [train.py:421] (2/8) Epoch 11, batch 30000, loss[loss=2.314, over 2240.00 frames. , ppl: 10.112935102024556] tot_loss[loss=2.27, over 5412268.94 frames. , ppl: 9.676142357057515], batch size: 70
+2022-12-14 05:19:07,019 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:19:07,779 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.613103137864265
+2022-12-14 05:20:48,039 INFO [train.py:421] (2/8) Epoch 11, batch 30200, loss[loss=2.484, over 1190.00 frames. , ppl: 11.994896563797951] tot_loss[loss=2.27, over 5401607.44 frames. , ppl: 9.675849973287528], batch size: 70
+2022-12-14 05:22:30,468 INFO [train.py:421] (2/8) Epoch 11, batch 30400, loss[loss=2.244, over 2870.00 frames. , ppl: 9.434748922531956] tot_loss[loss=2.266, over 5487448.72 frames. , ppl: 9.643690796801517], batch size: 70
+2022-12-14 05:24:12,352 INFO [train.py:421] (2/8) Epoch 11, batch 30600, loss[loss=2.465, over 910.00 frames. , ppl: 11.765242059340505] tot_loss[loss=2.264, over 5529200.56 frames. , ppl: 9.626082573027578], batch size: 70
+2022-12-14 05:25:49,474 INFO [train.py:421] (2/8) Epoch 11, batch 30800, loss[loss=2.219, over 6790.00 frames. , ppl: 9.195742925391134] tot_loss[loss=2.264, over 5534633.15 frames. , ppl: 9.62414682880796], batch size: 70
+2022-12-14 05:27:28,454 INFO [train.py:421] (2/8) Epoch 11, batch 31000, loss[loss=2.244, over 2380.00 frames. , ppl: 9.43559648541039] tot_loss[loss=2.265, over 5521366.00 frames. , ppl: 9.627748891788068], batch size: 70
+2022-12-14 05:27:28,454 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:27:29,214 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.613066144848931
+2022-12-14 05:29:07,427 INFO [train.py:421] (2/8) Epoch 11, batch 31200, loss[loss=2.196, over 4060.00 frames. , ppl: 8.986559431921997] tot_loss[loss=2.265, over 5515994.49 frames. , ppl: 9.627037260312086], batch size: 70
+2022-12-14 05:30:46,537 INFO [train.py:421] (2/8) Epoch 11, batch 31400, loss[loss=2.313, over 3430.00 frames. , ppl: 10.102700772482343] tot_loss[loss=2.265, over 5498521.18 frames. , ppl: 9.631311304439178], batch size: 70
+2022-12-14 05:32:28,444 INFO [train.py:421] (2/8) Epoch 11, batch 31600, loss[loss=2.399, over 1890.00 frames. , ppl: 11.017311398964509] tot_loss[loss=2.266, over 5480589.76 frames. , ppl: 9.637441302897034], batch size: 70
+2022-12-14 05:34:09,689 INFO [train.py:421] (2/8) Epoch 11, batch 31800, loss[loss=2.13, over 7980.00 frames. , ppl: 8.418340357796975] tot_loss[loss=2.265, over 5505406.32 frames. , ppl: 9.631502484186573], batch size: 70
+2022-12-14 05:35:44,348 INFO [train.py:421] (2/8) Epoch 11, batch 32000, loss[loss=2.324, over 1470.00 frames. , ppl: 10.21598872633957] tot_loss[loss=2.265, over 5471081.05 frames. , ppl: 9.633943847676472], batch size: 70
+2022-12-14 05:35:44,348 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:35:45,109 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604530257921299
+2022-12-14 05:37:24,350 INFO [train.py:421] (2/8) Epoch 11, batch 32200, loss[loss=2.125, over 13580.00 frames. , ppl: 8.375034047979753] tot_loss[loss=2.265, over 5503457.25 frames. , ppl: 9.63082439052897], batch size: 70
+2022-12-14 05:39:05,741 INFO [train.py:421] (2/8) Epoch 11, batch 32400, loss[loss=2.603, over 840.00 frames. , ppl: 13.498186438639225] tot_loss[loss=2.265, over 5485804.73 frames. , ppl: 9.632230607640034], batch size: 70
+2022-12-14 05:40:47,672 INFO [train.py:421] (2/8) Epoch 11, batch 32600, loss[loss=2.317, over 1540.00 frames. , ppl: 10.149368704475226] tot_loss[loss=2.266, over 5470738.03 frames. , ppl: 9.637706665905894], batch size: 70
+2022-12-14 05:42:26,787 INFO [train.py:421] (2/8) Epoch 11, batch 32800, loss[loss=4.049, over 350.00 frames. , ppl: 57.35695360698438] tot_loss[loss=2.267, over 5437155.76 frames. , ppl: 9.647299148846571], batch size: 70
+2022-12-14 05:44:07,638 INFO [train.py:421] (2/8) Epoch 11, batch 33000, loss[loss=2.53, over 1260.00 frames. , ppl: 12.555686240578353] tot_loss[loss=2.266, over 5466749.38 frames. , ppl: 9.637241261913262], batch size: 70
+2022-12-14 05:44:07,639 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:44:08,396 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.5747582284789
+2022-12-14 05:45:47,673 INFO [train.py:421] (2/8) Epoch 11, batch 33200, loss[loss=2.237, over 2730.00 frames. , ppl: 9.363442468371643] tot_loss[loss=2.266, over 5446716.13 frames. , ppl: 9.644698695701326], batch size: 70
+2022-12-14 05:47:26,446 INFO [train.py:421] (2/8) Epoch 11, batch 33400, loss[loss=2.213, over 4760.00 frames. , ppl: 9.139464173954613] tot_loss[loss=2.267, over 5406175.95 frames. , ppl: 9.65408693908256], batch size: 70
+2022-12-14 05:49:07,561 INFO [train.py:421] (2/8) Epoch 11, batch 33600, loss[loss=2.324, over 2310.00 frames. , ppl: 10.21628117860258] tot_loss[loss=2.267, over 5386655.98 frames. , ppl: 9.654889275614515], batch size: 70
+2022-12-14 05:50:49,688 INFO [train.py:421] (2/8) Epoch 11, batch 33800, loss[loss=2.284, over 1540.00 frames. , ppl: 9.81774612776578] tot_loss[loss=2.267, over 5408705.23 frames. , ppl: 9.649017438062996], batch size: 70
+2022-12-14 05:52:29,250 INFO [train.py:421] (2/8) Epoch 11, batch 34000, loss[loss=2.245, over 2310.00 frames. , ppl: 9.44017889451778] tot_loss[loss=2.268, over 5406255.97 frames. , ppl: 9.656252083752255], batch size: 70
+2022-12-14 05:52:29,251 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 05:52:30,011 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.580360386976093
+2022-12-14 05:54:12,839 INFO [train.py:421] (2/8) Epoch 11, batch 34200, loss[loss=2.189, over 3920.00 frames. , ppl: 8.92604416352823] tot_loss[loss=2.266, over 5431640.87 frames. , ppl: 9.645190928744633], batch size: 70
+2022-12-14 05:55:56,791 INFO [train.py:421] (2/8) Epoch 11, batch 34400, loss[loss=2.616, over 700.00 frames. , ppl: 13.679307809876134] tot_loss[loss=2.267, over 5419285.19 frames. , ppl: 9.651023130805868], batch size: 70
+2022-12-14 05:57:35,020 INFO [train.py:421] (2/8) Epoch 11, batch 34600, loss[loss=2.283, over 3010.00 frames. , ppl: 9.8061692779615] tot_loss[loss=2.267, over 5415896.15 frames. , ppl: 9.654670957602765], batch size: 70
+2022-12-14 05:59:12,284 INFO [train.py:421] (2/8) Epoch 11, batch 34800, loss[loss=2.132, over 6020.00 frames. , ppl: 8.432783941054312] tot_loss[loss=2.268, over 5403306.81 frames. , ppl: 9.657889281675294], batch size: 70
+2022-12-14 06:00:51,153 INFO [train.py:421] (2/8) Epoch 11, batch 35000, loss[loss=2.362, over 2310.00 frames. , ppl: 10.609977646380367] tot_loss[loss=2.266, over 5465038.41 frames. , ppl: 9.641714044784864], batch size: 70
+2022-12-14 06:00:51,154 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:00:51,940 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.571281208986392
+2022-12-14 06:02:35,354 INFO [train.py:421] (2/8) Epoch 11, batch 35200, loss[loss=2.704, over 770.00 frames. , ppl: 14.940195583752972] tot_loss[loss=2.265, over 5524581.42 frames. , ppl: 9.631461481706701], batch size: 70
+2022-12-14 06:04:14,594 INFO [train.py:421] (2/8) Epoch 11, batch 35400, loss[loss=2.245, over 3500.00 frames. , ppl: 9.439912467362738] tot_loss[loss=2.265, over 5538586.66 frames. , ppl: 9.629845353901729], batch size: 70
+2022-12-14 06:05:56,229 INFO [train.py:421] (2/8) Epoch 11, batch 35600, loss[loss=2.274, over 4340.00 frames. , ppl: 9.718036675483674] tot_loss[loss=2.265, over 5538849.13 frames. , ppl: 9.630455903397406], batch size: 70
+2022-12-14 06:07:32,111 INFO [train.py:421] (2/8) Epoch 11, batch 35800, loss[loss=3.234, over 490.00 frames. , ppl: 25.374019221257917] tot_loss[loss=2.266, over 5506922.31 frames. , ppl: 9.636447659289637], batch size: 70
+2022-12-14 06:09:12,202 INFO [train.py:421] (2/8) Epoch 11, batch 36000, loss[loss=2.238, over 5110.00 frames. , ppl: 9.37097681846879] tot_loss[loss=2.265, over 5512860.90 frames. , ppl: 9.631587233064373], batch size: 70
+2022-12-14 06:09:12,202 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:09:12,963 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.593320836541546
+2022-12-14 06:10:52,349 INFO [train.py:421] (2/8) Epoch 11, batch 36200, loss[loss=2.28, over 2940.00 frames. , ppl: 9.772327181401138] tot_loss[loss=2.264, over 5506775.58 frames. , ppl: 9.622293646937814], batch size: 70
+2022-12-14 06:12:34,282 INFO [train.py:421] (2/8) Epoch 11, batch 36400, loss[loss=2.246, over 3430.00 frames. , ppl: 9.453764808659297] tot_loss[loss=2.265, over 5461530.74 frames. , ppl: 9.631414594442514], batch size: 70
+2022-12-14 06:14:15,795 INFO [train.py:421] (2/8) Epoch 11, batch 36600, loss[loss=2.227, over 4270.00 frames. , ppl: 9.27423732656709] tot_loss[loss=2.265, over 5488599.71 frames. , ppl: 9.631105760936395], batch size: 70
+2022-12-14 06:15:56,234 INFO [train.py:421] (2/8) Epoch 11, batch 36800, loss[loss=2.3, over 1750.00 frames. , ppl: 9.974718191892508] tot_loss[loss=2.264, over 5520442.93 frames. , ppl: 9.624256135177237], batch size: 70
+2022-12-14 06:17:38,557 INFO [train.py:421] (2/8) Epoch 11, batch 37000, loss[loss=2.387, over 1260.00 frames. , ppl: 10.87705985488635] tot_loss[loss=2.263, over 5545204.53 frames. , ppl: 9.614189796315369], batch size: 70
+2022-12-14 06:17:38,558 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:17:39,307 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.585279151337225
+2022-12-14 06:19:19,619 INFO [train.py:421] (2/8) Epoch 11, batch 37200, loss[loss=2.201, over 3780.00 frames. , ppl: 9.03163770463613] tot_loss[loss=2.264, over 5524806.18 frames. , ppl: 9.616845440605587], batch size: 70
+2022-12-14 06:21:00,637 INFO [train.py:421] (2/8) Epoch 11, batch 37400, loss[loss=2.214, over 5180.00 frames. , ppl: 9.155953256767846] tot_loss[loss=2.264, over 5522548.12 frames. , ppl: 9.621101905313829], batch size: 70
+2022-12-14 06:22:44,800 INFO [train.py:421] (2/8) Epoch 11, batch 37600, loss[loss=2.242, over 4200.00 frames. , ppl: 9.41468971729316] tot_loss[loss=2.264, over 5559698.70 frames. , ppl: 9.619096185203631], batch size: 70
+2022-12-14 06:24:22,307 INFO [train.py:421] (2/8) Epoch 11, batch 37800, loss[loss=2.123, over 5390.00 frames. , ppl: 8.359749319812455] tot_loss[loss=2.264, over 5535582.48 frames. , ppl: 9.625217081860125], batch size: 70
+2022-12-14 06:26:03,854 INFO [train.py:421] (2/8) Epoch 11, batch 38000, loss[loss=2.107, over 7280.00 frames. , ppl: 8.22436098660149] tot_loss[loss=2.265, over 5528402.74 frames. , ppl: 9.630633501991538], batch size: 70
+2022-12-14 06:26:03,855 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:26:04,620 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.584332929905047
+2022-12-14 06:27:46,868 INFO [train.py:421] (2/8) Epoch 11, batch 38200, loss[loss=2.444, over 1190.00 frames. , ppl: 11.517175579796183] tot_loss[loss=2.266, over 5510756.15 frames. , ppl: 9.637218874527496], batch size: 70
+2022-12-14 06:29:24,550 INFO [train.py:421] (2/8) Epoch 11, batch 38400, loss[loss=3, over 560.00 frames. , ppl: 20.0917463116125] tot_loss[loss=2.267, over 5479604.72 frames. , ppl: 9.6467057778732], batch size: 70
+2022-12-14 06:31:03,585 INFO [train.py:421] (2/8) Epoch 11, batch 38600, loss[loss=2.25, over 1960.00 frames. , ppl: 9.486582462818564] tot_loss[loss=2.265, over 5506555.21 frames. , ppl: 9.633969734560576], batch size: 70
+2022-12-14 06:32:43,275 INFO [train.py:421] (2/8) Epoch 11, batch 38800, loss[loss=2.468, over 980.00 frames. , ppl: 11.802807798225038] tot_loss[loss=2.263, over 5534718.91 frames. , ppl: 9.613662533202506], batch size: 70
+2022-12-14 06:34:25,390 INFO [train.py:421] (2/8) Epoch 11, batch 39000, loss[loss=2.451, over 1400.00 frames. , ppl: 11.603739623041982] tot_loss[loss=2.264, over 5540935.23 frames. , ppl: 9.62349330070632], batch size: 70
+2022-12-14 06:34:25,390 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:34:26,137 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.58167067609052
+2022-12-14 06:36:07,551 INFO [train.py:421] (2/8) Epoch 11, batch 39200, loss[loss=2.85, over 630.00 frames. , ppl: 17.28990569483796] tot_loss[loss=2.265, over 5522749.95 frames. , ppl: 9.628523759549976], batch size: 70
+2022-12-14 06:37:47,055 INFO [train.py:421] (2/8) Epoch 11, batch 39400, loss[loss=2.379, over 2310.00 frames. , ppl: 10.791585834362628] tot_loss[loss=2.266, over 5485291.72 frames. , ppl: 9.64209025624351], batch size: 70
+2022-12-14 06:39:24,955 INFO [train.py:421] (2/8) Epoch 11, batch 39600, loss[loss=2.201, over 4200.00 frames. , ppl: 9.03275170144958] tot_loss[loss=2.266, over 5475292.03 frames. , ppl: 9.64256845482149], batch size: 70
+2022-12-14 06:41:07,681 INFO [train.py:421] (2/8) Epoch 11, batch 39800, loss[loss=2.201, over 2030.00 frames. , ppl: 9.031406628214631] tot_loss[loss=2.266, over 5502283.80 frames. , ppl: 9.637409600657799], batch size: 70
+2022-12-14 06:42:42,716 INFO [train.py:421] (2/8) Epoch 11, batch 40000, loss[loss=2.539, over 910.00 frames. , ppl: 12.671790128353027] tot_loss[loss=2.268, over 5433142.36 frames. , ppl: 9.65808569478198], batch size: 70
+2022-12-14 06:42:42,717 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:42:43,479 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.58547067678701
+2022-12-14 06:44:22,143 INFO [train.py:421] (2/8) Epoch 11, batch 40200, loss[loss=2.411, over 1190.00 frames. , ppl: 11.14864161050726] tot_loss[loss=2.268, over 5437100.37 frames. , ppl: 9.656128581461672], batch size: 70
+2022-12-14 06:46:06,984 INFO [train.py:421] (2/8) Epoch 11, batch 40400, loss[loss=2.494, over 1260.00 frames. , ppl: 12.107224200986337] tot_loss[loss=2.266, over 5460813.21 frames. , ppl: 9.643395707916167], batch size: 70
+2022-12-14 06:47:50,938 INFO [train.py:421] (2/8) Epoch 11, batch 40600, loss[loss=2.719, over 630.00 frames. , ppl: 15.166659236249075] tot_loss[loss=2.266, over 5483170.33 frames. , ppl: 9.637757727601103], batch size: 70
+2022-12-14 06:49:30,439 INFO [train.py:421] (2/8) Epoch 11, batch 40800, loss[loss=2.254, over 4130.00 frames. , ppl: 9.527613372323199] tot_loss[loss=2.267, over 5447509.36 frames. , ppl: 9.648530077494668], batch size: 70
+2022-12-14 06:51:11,967 INFO [train.py:421] (2/8) Epoch 11, batch 41000, loss[loss=2.469, over 1540.00 frames. , ppl: 11.810812010537287] tot_loss[loss=2.267, over 5458992.26 frames. , ppl: 9.652192831673977], batch size: 70
+2022-12-14 06:51:11,967 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:51:12,725 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.258, over 211138.00 frames. , ppl: 9.567495329829377
+2022-12-14 06:52:55,088 INFO [train.py:421] (2/8) Epoch 11, batch 41200, loss[loss=2.465, over 910.00 frames. , ppl: 11.761855671160871] tot_loss[loss=2.269, over 5407847.29 frames. , ppl: 9.666433547634051], batch size: 70
+2022-12-14 06:54:38,130 INFO [train.py:421] (2/8) Epoch 11, batch 41400, loss[loss=2.203, over 5180.00 frames. , ppl: 9.056166170746716] tot_loss[loss=2.268, over 5435405.33 frames. , ppl: 9.66020780222024], batch size: 70
+2022-12-14 06:56:20,580 INFO [train.py:421] (2/8) Epoch 11, batch 41600, loss[loss=2.398, over 910.00 frames. , ppl: 11.000292392011872] tot_loss[loss=2.268, over 5445104.98 frames. , ppl: 9.658303642518854], batch size: 70
+2022-12-14 06:58:01,897 INFO [train.py:421] (2/8) Epoch 11, batch 41800, loss[loss=2.419, over 1680.00 frames. , ppl: 11.233854506982787] tot_loss[loss=2.267, over 5468658.19 frames. , ppl: 9.648399282004084], batch size: 70
+2022-12-14 06:59:44,359 INFO [train.py:421] (2/8) Epoch 11, batch 42000, loss[loss=2.272, over 3220.00 frames. , ppl: 9.695934553364205] tot_loss[loss=2.267, over 5454230.33 frames. , ppl: 9.648509375266862], batch size: 70
+2022-12-14 06:59:44,360 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 06:59:45,115 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.592491660652186
+2022-12-14 07:01:25,038 INFO [train.py:421] (2/8) Epoch 11, batch 42200, loss[loss=2.269, over 2800.00 frames. , ppl: 9.667175941105803] tot_loss[loss=2.269, over 5390443.49 frames. , ppl: 9.666586132066255], batch size: 70
+2022-12-14 07:03:05,993 INFO [train.py:421] (2/8) Epoch 11, batch 42400, loss[loss=2.192, over 10360.00 frames. , ppl: 8.957090049645167] tot_loss[loss=2.268, over 5427738.48 frames. , ppl: 9.656111766594282], batch size: 70
+2022-12-14 07:04:49,207 INFO [train.py:421] (2/8) Epoch 11, batch 42600, loss[loss=2.53, over 980.00 frames. , ppl: 12.547708084459028] tot_loss[loss=2.266, over 5473841.08 frames. , ppl: 9.645260051288188], batch size: 70
+2022-12-14 07:06:29,113 INFO [train.py:421] (2/8) Epoch 11, batch 42800, loss[loss=2.173, over 4690.00 frames. , ppl: 8.78277143411646] tot_loss[loss=2.267, over 5438251.27 frames. , ppl: 9.651206520266696], batch size: 70
+2022-12-14 07:08:09,283 INFO [train.py:421] (2/8) Epoch 11, batch 43000, loss[loss=2.249, over 2170.00 frames. , ppl: 9.481243762076158] tot_loss[loss=2.266, over 5449672.64 frames. , ppl: 9.643858619556518], batch size: 70
+2022-12-14 07:08:09,284 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:08:10,045 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.569812285675939
+2022-12-14 07:09:48,736 INFO [train.py:421] (2/8) Epoch 11, batch 43200, loss[loss=2.228, over 2940.00 frames. , ppl: 9.28488283266271] tot_loss[loss=2.266, over 5468013.33 frames. , ppl: 9.639108823459646], batch size: 70
+2022-12-14 07:11:27,392 INFO [train.py:421] (2/8) Epoch 11, batch 43400, loss[loss=2.648, over 840.00 frames. , ppl: 14.126117195245818] tot_loss[loss=2.268, over 5395889.90 frames. , ppl: 9.655940780756046], batch size: 70
+2022-12-14 07:13:07,427 INFO [train.py:421] (2/8) Epoch 11, batch 43600, loss[loss=2.287, over 1960.00 frames. , ppl: 9.843329180279031] tot_loss[loss=2.267, over 5397347.40 frames. , ppl: 9.652135806484866], batch size: 70
+2022-12-14 07:14:45,248 INFO [train.py:421] (2/8) Epoch 11, batch 43800, loss[loss=2.264, over 3990.00 frames. , ppl: 9.62584488407434] tot_loss[loss=2.267, over 5429948.76 frames. , ppl: 9.646918607688221], batch size: 70
+2022-12-14 07:16:27,745 INFO [train.py:421] (2/8) Epoch 11, batch 44000, loss[loss=2.527, over 1330.00 frames. , ppl: 12.520786607683211] tot_loss[loss=2.265, over 5512642.20 frames. , ppl: 9.629946065996506], batch size: 70
+2022-12-14 07:16:27,746 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:16:28,507 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.577322170162937
+2022-12-14 07:18:06,698 INFO [train.py:421] (2/8) Epoch 11, batch 44200, loss[loss=2.174, over 3990.00 frames. , ppl: 8.790511946388339] tot_loss[loss=2.266, over 5476957.68 frames. , ppl: 9.642325860124133], batch size: 70
+2022-12-14 07:19:47,740 INFO [train.py:421] (2/8) Epoch 11, batch 44400, loss[loss=2.196, over 5110.00 frames. , ppl: 8.993033001797686] tot_loss[loss=2.266, over 5505792.57 frames. , ppl: 9.637312989430164], batch size: 70
+2022-12-14 07:21:28,931 INFO [train.py:421] (2/8) Epoch 11, batch 44600, loss[loss=2.246, over 3360.00 frames. , ppl: 9.45396370886702] tot_loss[loss=2.266, over 5488457.91 frames. , ppl: 9.643068597846415], batch size: 70
+2022-12-14 07:23:10,233 INFO [train.py:421] (2/8) Epoch 11, batch 44800, loss[loss=4.88, over 280.00 frames. , ppl: 131.6891651164087] tot_loss[loss=2.266, over 5481152.29 frames. , ppl: 9.642856237358261], batch size: 70
+2022-12-14 07:24:47,389 INFO [train.py:421] (2/8) Epoch 11, batch 45000, loss[loss=2.214, over 4130.00 frames. , ppl: 9.155547332324948] tot_loss[loss=2.267, over 5475104.60 frames. , ppl: 9.650140923007754], batch size: 70
+2022-12-14 07:24:47,389 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:24:48,135 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.587963694090844
+2022-12-14 07:26:33,405 INFO [train.py:421] (2/8) Epoch 11, batch 45200, loss[loss=2.266, over 3220.00 frames. , ppl: 9.639727074875871] tot_loss[loss=2.265, over 5551423.78 frames. , ppl: 9.633056152444816], batch size: 70
+2022-12-14 07:28:14,100 INFO [train.py:421] (2/8) Epoch 11, batch 45400, loss[loss=2.489, over 2030.00 frames. , ppl: 12.048577019189437] tot_loss[loss=2.266, over 5535652.65 frames. , ppl: 9.640734364482997], batch size: 70
+2022-12-14 07:29:58,380 INFO [train.py:421] (2/8) Epoch 11, batch 45600, loss[loss=2.259, over 3080.00 frames. , ppl: 9.570496307998859] tot_loss[loss=2.265, over 5569065.47 frames. , ppl: 9.629487060738295], batch size: 70
+2022-12-14 07:31:41,339 INFO [train.py:421] (2/8) Epoch 11, batch 45800, loss[loss=2.439, over 1190.00 frames. , ppl: 11.455911676742778] tot_loss[loss=2.265, over 5543689.86 frames. , ppl: 9.635097464895765], batch size: 70
+2022-12-14 07:33:22,374 INFO [train.py:421] (2/8) Epoch 11, batch 46000, loss[loss=2.265, over 980.00 frames. , ppl: 9.628238153270111] tot_loss[loss=2.267, over 5517517.99 frames. , ppl: 9.645701455680463], batch size: 70
+2022-12-14 07:33:22,375 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:33:23,121 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604798933149176
+2022-12-14 07:35:07,447 INFO [train.py:421] (2/8) Epoch 11, batch 46200, loss[loss=2.609, over 770.00 frames. , ppl: 13.592099000418829] tot_loss[loss=2.265, over 5560709.63 frames. , ppl: 9.627613151134835], batch size: 70
+2022-12-14 07:36:50,560 INFO [train.py:421] (2/8) Epoch 11, batch 46400, loss[loss=2.341, over 1610.00 frames. , ppl: 10.392128894668433] tot_loss[loss=2.264, over 5591136.32 frames. , ppl: 9.618032357891392], batch size: 70
+2022-12-14 07:38:31,816 INFO [train.py:421] (2/8) Epoch 11, batch 46600, loss[loss=2.667, over 840.00 frames. , ppl: 14.39543017548774] tot_loss[loss=2.263, over 5586346.61 frames. , ppl: 9.612089976447384], batch size: 70
+2022-12-14 07:40:11,794 INFO [train.py:421] (2/8) Epoch 11, batch 46800, loss[loss=2.226, over 1680.00 frames. , ppl: 9.265752204374504] tot_loss[loss=2.262, over 5593093.99 frames. , ppl: 9.600877863480218], batch size: 70
+2022-12-14 07:41:53,103 INFO [train.py:421] (2/8) Epoch 11, batch 47000, loss[loss=3.558, over 420.00 frames. , ppl: 35.07701174324866] tot_loss[loss=2.263, over 5549809.01 frames. , ppl: 9.610039384515662], batch size: 70
+2022-12-14 07:41:53,104 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:41:53,854 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.593797929123168
+2022-12-14 07:43:37,145 INFO [train.py:421] (2/8) Epoch 11, batch 47200, loss[loss=2.144, over 10780.00 frames. , ppl: 8.531818626237193] tot_loss[loss=2.262, over 5590942.96 frames. , ppl: 9.600320370867957], batch size: 70
+2022-12-14 07:45:10,242 INFO [train.py:421] (2/8) Epoch 11, batch 47400, loss[loss=2.349, over 1610.00 frames. , ppl: 10.470294282478054] tot_loss[loss=2.262, over 5598219.61 frames. , ppl: 9.598714401738428], batch size: 70
+2022-12-14 07:46:47,936 INFO [train.py:421] (2/8) Epoch 11, batch 47600, loss[loss=3.256, over 490.00 frames. , ppl: 25.94073551204826] tot_loss[loss=2.262, over 5578256.87 frames. , ppl: 9.598661924695078], batch size: 70
+2022-12-14 07:48:29,963 INFO [train.py:421] (2/8) Epoch 11, batch 47800, loss[loss=2.469, over 1260.00 frames. , ppl: 11.808643552060811] tot_loss[loss=2.263, over 5542227.27 frames. , ppl: 9.60786618571362], batch size: 70
+2022-12-14 07:50:08,691 INFO [train.py:421] (2/8) Epoch 11, batch 48000, loss[loss=2.522, over 980.00 frames. , ppl: 12.457147110241873] tot_loss[loss=2.262, over 5561033.50 frames. , ppl: 9.605638084741287], batch size: 70
+2022-12-14 07:50:08,691 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:50:09,455 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.597875482483978
+2022-12-14 07:51:49,510 INFO [train.py:421] (2/8) Epoch 11, batch 48200, loss[loss=2.228, over 4340.00 frames. , ppl: 9.283723530321092] tot_loss[loss=2.264, over 5530192.59 frames. , ppl: 9.621199546704549], batch size: 70
+2022-12-14 07:53:28,410 INFO [train.py:421] (2/8) Epoch 11, batch 48400, loss[loss=2.948, over 560.00 frames. , ppl: 19.06920588783114] tot_loss[loss=2.264, over 5522817.48 frames. , ppl: 9.625611257019248], batch size: 70
+2022-12-14 07:55:09,671 INFO [train.py:421] (2/8) Epoch 11, batch 48600, loss[loss=2.268, over 3360.00 frames. , ppl: 9.66144135336769] tot_loss[loss=2.266, over 5494975.00 frames. , ppl: 9.637488143790932], batch size: 70
+2022-12-14 07:56:48,790 INFO [train.py:421] (2/8) Epoch 11, batch 48800, loss[loss=2.651, over 770.00 frames. , ppl: 14.172052751681688] tot_loss[loss=2.265, over 5498372.48 frames. , ppl: 9.630585286406422], batch size: 70
+2022-12-14 07:58:30,730 INFO [train.py:421] (2/8) Epoch 11, batch 49000, loss[loss=2.129, over 5670.00 frames. , ppl: 8.406387637672886] tot_loss[loss=2.266, over 5472845.15 frames. , ppl: 9.643368960331692], batch size: 70
+2022-12-14 07:58:30,730 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 07:58:31,475 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.595209466935325
+2022-12-14 08:00:11,580 INFO [train.py:421] (2/8) Epoch 11, batch 49200, loss[loss=2.398, over 1960.00 frames. , ppl: 11.006563853972976] tot_loss[loss=2.267, over 5450074.68 frames. , ppl: 9.650296862040928], batch size: 70
+2022-12-14 08:01:52,649 INFO [train.py:421] (2/8) Epoch 11, batch 49400, loss[loss=2.265, over 4060.00 frames. , ppl: 9.62982137025344] tot_loss[loss=2.267, over 5449444.47 frames. , ppl: 9.649586075863247], batch size: 70
+2022-12-14 08:03:32,903 INFO [train.py:421] (2/8) Epoch 11, batch 49600, loss[loss=2.129, over 9240.00 frames. , ppl: 8.4039288081115] tot_loss[loss=2.265, over 5500591.61 frames. , ppl: 9.630398382131926], batch size: 70
+2022-12-14 08:05:11,615 INFO [train.py:421] (2/8) Epoch 11, batch 49800, loss[loss=2.486, over 1400.00 frames. , ppl: 12.017103361658663] tot_loss[loss=2.264, over 5535374.49 frames. , ppl: 9.61897836236197], batch size: 70
+2022-12-14 08:06:50,133 INFO [train.py:421] (2/8) Epoch 11, batch 50000, loss[loss=2.321, over 2240.00 frames. , ppl: 10.183324374143496] tot_loss[loss=2.263, over 5568593.67 frames. , ppl: 9.608079139135713], batch size: 70
+2022-12-14 08:06:50,134 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:06:50,881 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.581751511481997
+2022-12-14 08:08:32,385 INFO [train.py:421] (2/8) Epoch 11, batch 50200, loss[loss=2.608, over 910.00 frames. , ppl: 13.573384245697271] tot_loss[loss=2.263, over 5567383.73 frames. , ppl: 9.608387051960436], batch size: 70
+2022-12-14 08:10:12,766 INFO [train.py:421] (2/8) Epoch 11, batch 50400, loss[loss=2.332, over 1750.00 frames. , ppl: 10.294421231945432] tot_loss[loss=2.264, over 5574566.86 frames. , ppl: 9.617149378723001], batch size: 70
+2022-12-14 08:11:48,474 INFO [train.py:421] (2/8) Epoch 11, batch 50600, loss[loss=2.164, over 4410.00 frames. , ppl: 8.708693759864472] tot_loss[loss=2.264, over 5560087.97 frames. , ppl: 9.62017783494286], batch size: 70
+2022-12-14 08:13:30,527 INFO [train.py:421] (2/8) Epoch 11, batch 50800, loss[loss=2.401, over 840.00 frames. , ppl: 11.035514767304251] tot_loss[loss=2.264, over 5562416.56 frames. , ppl: 9.622990382743719], batch size: 70
+2022-12-14 08:15:09,635 INFO [train.py:421] (2/8) Epoch 11, batch 51000, loss[loss=2.254, over 2730.00 frames. , ppl: 9.522967058553169] tot_loss[loss=2.264, over 5553689.45 frames. , ppl: 9.617049631752847], batch size: 70
+2022-12-14 08:15:09,636 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:15:10,400 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599885781441973
+2022-12-14 08:16:50,480 INFO [train.py:421] (2/8) Epoch 11, batch 51200, loss[loss=2.301, over 1540.00 frames. , ppl: 9.984631987249335] tot_loss[loss=2.264, over 5527632.04 frames. , ppl: 9.624883732575531], batch size: 70
+2022-12-14 08:18:32,233 INFO [train.py:421] (2/8) Epoch 11, batch 51400, loss[loss=2.419, over 840.00 frames. , ppl: 11.23485039074123] tot_loss[loss=2.263, over 5553322.82 frames. , ppl: 9.615711633897336], batch size: 70
+2022-12-14 08:20:10,263 INFO [train.py:421] (2/8) Epoch 11, batch 51600, loss[loss=2.292, over 2100.00 frames. , ppl: 9.89086687196323] tot_loss[loss=2.265, over 5505684.75 frames. , ppl: 9.634920168036304], batch size: 70
+2022-12-14 08:21:49,063 INFO [train.py:421] (2/8) Epoch 11, batch 51800, loss[loss=2.204, over 4760.00 frames. , ppl: 9.057429674919167] tot_loss[loss=2.265, over 5496550.78 frames. , ppl: 9.632582048387404], batch size: 70
+2022-12-14 08:23:27,648 INFO [train.py:421] (2/8) Epoch 11, batch 52000, loss[loss=2.491, over 840.00 frames. , ppl: 12.068487052642219] tot_loss[loss=2.266, over 5479073.22 frames. , ppl: 9.638990428909006], batch size: 70
+2022-12-14 08:23:27,649 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:23:28,401 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.587526624273279
+2022-12-14 08:25:07,202 INFO [train.py:421] (2/8) Epoch 11, batch 52200, loss[loss=2.462, over 1050.00 frames. , ppl: 11.724799268124688] tot_loss[loss=2.266, over 5495352.90 frames. , ppl: 9.639178304358015], batch size: 70
+2022-12-14 08:26:45,634 INFO [train.py:421] (2/8) Epoch 11, batch 52400, loss[loss=2.243, over 2590.00 frames. , ppl: 9.425412724790991] tot_loss[loss=2.266, over 5503983.29 frames. , ppl: 9.640287344584928], batch size: 70
+2022-12-14 08:28:22,604 INFO [train.py:421] (2/8) Epoch 11, batch 52600, loss[loss=2.385, over 2380.00 frames. , ppl: 10.853887717855525] tot_loss[loss=2.267, over 5485132.26 frames. , ppl: 9.646646047463141], batch size: 70
+2022-12-14 08:30:00,798 INFO [train.py:421] (2/8) Epoch 11, batch 52800, loss[loss=2.233, over 3220.00 frames. , ppl: 9.32569391132076] tot_loss[loss=2.266, over 5524437.18 frames. , ppl: 9.638674461190885], batch size: 70
+2022-12-14 08:31:42,284 INFO [train.py:421] (2/8) Epoch 11, batch 53000, loss[loss=2.348, over 2660.00 frames. , ppl: 10.463027902906811] tot_loss[loss=2.266, over 5531036.58 frames. , ppl: 9.640491890479785], batch size: 70
+2022-12-14 08:31:42,284 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:31:43,029 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.581084994414798
+2022-12-14 08:33:24,598 INFO [train.py:421] (2/8) Epoch 11, batch 53200, loss[loss=2.437, over 2170.00 frames. , ppl: 11.438639943356879] tot_loss[loss=2.265, over 5530590.32 frames. , ppl: 9.635175186257111], batch size: 70
+2022-12-14 08:35:06,681 INFO [train.py:421] (2/8) Epoch 11, batch 53400, loss[loss=2.506, over 980.00 frames. , ppl: 12.260369519570967] tot_loss[loss=2.265, over 5512439.89 frames. , ppl: 9.635361265238574], batch size: 70
+2022-12-14 08:36:47,426 INFO [train.py:421] (2/8) Epoch 11, batch 53600, loss[loss=2.467, over 1470.00 frames. , ppl: 11.783558326796909] tot_loss[loss=2.265, over 5522801.76 frames. , ppl: 9.630769156049345], batch size: 70
+2022-12-14 08:38:27,754 INFO [train.py:421] (2/8) Epoch 11, batch 53800, loss[loss=2.596, over 910.00 frames. , ppl: 13.405564171998494] tot_loss[loss=2.266, over 5500288.59 frames. , ppl: 9.639305916130569], batch size: 70
+2022-12-14 08:40:09,517 INFO [train.py:421] (2/8) Epoch 11, batch 54000, loss[loss=2.294, over 1820.00 frames. , ppl: 9.915580991547893] tot_loss[loss=2.265, over 5528739.26 frames. , ppl: 9.628542266628564], batch size: 70
+2022-12-14 08:40:09,517 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:40:10,265 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.59367723381749
+2022-12-14 08:41:46,161 INFO [train.py:421] (2/8) Epoch 11, batch 54200, loss[loss=2.187, over 4480.00 frames. , ppl: 8.908159782886546] tot_loss[loss=2.265, over 5535305.71 frames. , ppl: 9.63077855444283], batch size: 70
+2022-12-14 08:43:25,433 INFO [train.py:421] (2/8) Epoch 11, batch 54400, loss[loss=2.338, over 1960.00 frames. , ppl: 10.35776693654627] tot_loss[loss=2.265, over 5533256.51 frames. , ppl: 9.632530881481147], batch size: 70
+2022-12-14 08:45:04,355 INFO [train.py:421] (2/8) Epoch 11, batch 54600, loss[loss=2.314, over 1260.00 frames. , ppl: 10.112491773377384] tot_loss[loss=2.265, over 5506178.83 frames. , ppl: 9.634162866103749], batch size: 70
+2022-12-14 08:46:40,892 INFO [train.py:421] (2/8) Epoch 11, batch 54800, loss[loss=2.711, over 630.00 frames. , ppl: 15.04505985436549] tot_loss[loss=2.266, over 5463105.61 frames. , ppl: 9.644417580485404], batch size: 70
+2022-12-14 08:48:23,526 INFO [train.py:421] (2/8) Epoch 11, batch 55000, loss[loss=2.179, over 9310.00 frames. , ppl: 8.836205825159519] tot_loss[loss=2.266, over 5505960.36 frames. , ppl: 9.637427103777211], batch size: 70
+2022-12-14 08:48:23,526 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:48:24,276 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.600448456413142
+2022-12-14 08:50:04,126 INFO [train.py:421] (2/8) Epoch 11, batch 55200, loss[loss=2.584, over 1260.00 frames. , ppl: 13.256368541448536] tot_loss[loss=2.267, over 5449066.01 frames. , ppl: 9.64789410921481], batch size: 70
+2022-12-14 08:51:43,843 INFO [train.py:421] (2/8) Epoch 11, batch 55400, loss[loss=2.179, over 5530.00 frames. , ppl: 8.840703481784834] tot_loss[loss=2.266, over 5460386.15 frames. , ppl: 9.639576431157767], batch size: 70
+2022-12-14 08:53:24,613 INFO [train.py:421] (2/8) Epoch 11, batch 55600, loss[loss=2.289, over 1680.00 frames. , ppl: 9.86710946505352] tot_loss[loss=2.265, over 5504981.66 frames. , ppl: 9.632432896349117], batch size: 70
+2022-12-14 08:55:02,629 INFO [train.py:421] (2/8) Epoch 11, batch 55800, loss[loss=2.311, over 2030.00 frames. , ppl: 10.08110674011984] tot_loss[loss=2.265, over 5493240.91 frames. , ppl: 9.632671685204919], batch size: 70
+2022-12-14 08:56:44,614 INFO [train.py:421] (2/8) Epoch 11, batch 56000, loss[loss=2.219, over 3990.00 frames. , ppl: 9.19411401739668] tot_loss[loss=2.265, over 5503852.64 frames. , ppl: 9.628154504386883], batch size: 70
+2022-12-14 08:56:44,615 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 08:56:45,361 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589785979049779
+2022-12-14 08:58:29,535 INFO [train.py:421] (2/8) Epoch 11, batch 56200, loss[loss=2.214, over 5600.00 frames. , ppl: 9.154222628402158] tot_loss[loss=2.265, over 5517743.06 frames. , ppl: 9.628563999189012], batch size: 70
+2022-12-14 09:00:08,805 INFO [train.py:421] (2/8) Epoch 11, batch 56400, loss[loss=2.705, over 560.00 frames. , ppl: 14.957747678990886] tot_loss[loss=2.266, over 5490057.46 frames. , ppl: 9.63734202259811], batch size: 70
+2022-12-14 09:01:50,866 INFO [train.py:421] (2/8) Epoch 11, batch 56600, loss[loss=2.227, over 4200.00 frames. , ppl: 9.269694025024615] tot_loss[loss=2.267, over 5442369.98 frames. , ppl: 9.650627040593193], batch size: 70
+2022-12-14 09:03:29,242 INFO [train.py:421] (2/8) Epoch 11, batch 56800, loss[loss=2.154, over 7770.00 frames. , ppl: 8.616112691207032] tot_loss[loss=2.268, over 5420910.29 frames. , ppl: 9.661937526491354], batch size: 70
+2022-12-14 09:05:09,669 INFO [train.py:421] (2/8) Epoch 11, batch 57000, loss[loss=2.274, over 2870.00 frames. , ppl: 9.716449742182323] tot_loss[loss=2.268, over 5440442.85 frames. , ppl: 9.658720628195853], batch size: 70
+2022-12-14 09:05:09,670 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:05:10,417 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.587563519007391
+2022-12-14 09:06:48,000 INFO [train.py:421] (2/8) Epoch 11, batch 57200, loss[loss=2.423, over 1610.00 frames. , ppl: 11.279033721202293] tot_loss[loss=2.266, over 5508070.39 frames. , ppl: 9.644222657218487], batch size: 70
+2022-12-14 09:08:32,480 INFO [train.py:421] (2/8) Epoch 11, batch 57400, loss[loss=2.498, over 1330.00 frames. , ppl: 12.159546850455602] tot_loss[loss=2.267, over 5506750.77 frames. , ppl: 9.648524734204655], batch size: 70
+2022-12-14 09:10:14,444 INFO [train.py:421] (2/8) Epoch 11, batch 57600, loss[loss=2.367, over 2030.00 frames. , ppl: 10.670147842320965] tot_loss[loss=2.266, over 5534776.86 frames. , ppl: 9.636781013922906], batch size: 70
+2022-12-14 09:11:54,085 INFO [train.py:421] (2/8) Epoch 11, batch 57800, loss[loss=2.341, over 2100.00 frames. , ppl: 10.386755957935584] tot_loss[loss=2.265, over 5528061.13 frames. , ppl: 9.632375094452234], batch size: 70
+2022-12-14 09:13:32,714 INFO [train.py:421] (2/8) Epoch 11, batch 58000, loss[loss=2.239, over 2940.00 frames. , ppl: 9.388371906262263] tot_loss[loss=2.265, over 5521687.82 frames. , ppl: 9.632875808106023], batch size: 70
+2022-12-14 09:13:32,715 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:13:33,485 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.579467110769087
+2022-12-14 09:15:15,192 INFO [train.py:421] (2/8) Epoch 11, batch 58200, loss[loss=2.279, over 2030.00 frames. , ppl: 9.766406450865114] tot_loss[loss=2.264, over 5538921.80 frames. , ppl: 9.622136072206068], batch size: 70
+2022-12-14 09:16:56,032 INFO [train.py:421] (2/8) Epoch 11, batch 58400, loss[loss=2.407, over 910.00 frames. , ppl: 11.098463710524774] tot_loss[loss=2.265, over 5504787.26 frames. , ppl: 9.63004979610827], batch size: 70
+2022-12-14 09:18:34,152 INFO [train.py:421] (2/8) Epoch 11, batch 58600, loss[loss=2.726, over 910.00 frames. , ppl: 15.26917332282983] tot_loss[loss=2.265, over 5473844.29 frames. , ppl: 9.635840882400997], batch size: 70
+2022-12-14 09:20:16,253 INFO [train.py:421] (2/8) Epoch 11, batch 58800, loss[loss=2.129, over 7560.00 frames. , ppl: 8.407010240279483] tot_loss[loss=2.265, over 5457747.95 frames. , ppl: 9.632793917863527], batch size: 70
+2022-12-14 09:21:58,265 INFO [train.py:421] (2/8) Epoch 11, batch 59000, loss[loss=2.256, over 3360.00 frames. , ppl: 9.549151362192005] tot_loss[loss=2.267, over 5418348.47 frames. , ppl: 9.646514354696953], batch size: 70
+2022-12-14 09:21:58,265 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:21:59,025 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.585608293732157
+2022-12-14 09:23:40,364 INFO [train.py:421] (2/8) Epoch 11, batch 59200, loss[loss=2.772, over 770.00 frames. , ppl: 15.98413924416596] tot_loss[loss=2.267, over 5425531.29 frames. , ppl: 9.649534642251544], batch size: 70
+2022-12-14 09:25:21,847 INFO [train.py:421] (2/8) Epoch 11, batch 59400, loss[loss=2.814, over 630.00 frames. , ppl: 16.67182421580749] tot_loss[loss=2.267, over 5446460.11 frames. , ppl: 9.647795564326538], batch size: 70
+2022-12-14 09:27:00,559 INFO [train.py:421] (2/8) Epoch 11, batch 59600, loss[loss=2.49, over 1050.00 frames. , ppl: 12.066476647432593] tot_loss[loss=2.267, over 5432445.92 frames. , ppl: 9.653459682134631], batch size: 70
+2022-12-14 09:28:41,363 INFO [train.py:421] (2/8) Epoch 11, batch 59800, loss[loss=2.247, over 2170.00 frames. , ppl: 9.457952043845557] tot_loss[loss=2.266, over 5479233.17 frames. , ppl: 9.63894295006274], batch size: 70
+2022-12-14 09:30:19,023 INFO [train.py:421] (2/8) Epoch 11, batch 60000, loss[loss=2.281, over 3080.00 frames. , ppl: 9.785721891812592] tot_loss[loss=2.265, over 5490958.71 frames. , ppl: 9.634889949220298], batch size: 70
+2022-12-14 09:30:19,024 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:30:19,782 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.57757590873942
+2022-12-14 09:32:01,997 INFO [train.py:421] (2/8) Epoch 11, batch 60200, loss[loss=2.704, over 630.00 frames. , ppl: 14.931984893345367] tot_loss[loss=2.267, over 5452102.42 frames. , ppl: 9.646197862999784], batch size: 70
+2022-12-14 09:33:42,587 INFO [train.py:421] (2/8) Epoch 11, batch 60400, loss[loss=2.249, over 3010.00 frames. , ppl: 9.476368720568399] tot_loss[loss=2.266, over 5485378.25 frames. , ppl: 9.637045313732326], batch size: 70
+2022-12-14 09:35:24,852 INFO [train.py:421] (2/8) Epoch 11, batch 60600, loss[loss=2.177, over 4480.00 frames. , ppl: 8.821512511986933] tot_loss[loss=2.266, over 5470488.99 frames. , ppl: 9.645242431308128], batch size: 70
+2022-12-14 09:37:07,334 INFO [train.py:421] (2/8) Epoch 11, batch 60800, loss[loss=2.703, over 700.00 frames. , ppl: 14.929523802250896] tot_loss[loss=2.264, over 5548199.65 frames. , ppl: 9.625010822397936], batch size: 70
+2022-12-14 09:38:49,723 INFO [train.py:421] (2/8) Epoch 11, batch 61000, loss[loss=2.715, over 700.00 frames. , ppl: 15.105400303053884] tot_loss[loss=2.262, over 5587467.53 frames. , ppl: 9.60545767422639], batch size: 70
+2022-12-14 09:38:49,724 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:38:50,484 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.587292487923966
+2022-12-14 09:40:29,790 INFO [train.py:421] (2/8) Epoch 11, batch 61200, loss[loss=2.275, over 2100.00 frames. , ppl: 9.728656415320584] tot_loss[loss=2.262, over 5585449.77 frames. , ppl: 9.604112968124417], batch size: 70
+2022-12-14 09:42:09,350 INFO [train.py:421] (2/8) Epoch 11, batch 61400, loss[loss=2.171, over 3150.00 frames. , ppl: 8.763354864606628] tot_loss[loss=2.263, over 5562442.96 frames. , ppl: 9.609953071284785], batch size: 70
+2022-12-14 09:43:48,639 INFO [train.py:421] (2/8) Epoch 11, batch 61600, loss[loss=2.292, over 2170.00 frames. , ppl: 9.89668077591587] tot_loss[loss=2.263, over 5594279.43 frames. , ppl: 9.608207461219223], batch size: 70
+2022-12-14 09:45:23,760 INFO [train.py:421] (2/8) Epoch 11, batch 61800, loss[loss=2.209, over 3010.00 frames. , ppl: 9.10205885651399] tot_loss[loss=2.262, over 5611393.48 frames. , ppl: 9.605584999138598], batch size: 70
+2022-12-14 09:47:03,482 INFO [train.py:421] (2/8) Epoch 11, batch 62000, loss[loss=2.602, over 840.00 frames. , ppl: 13.49360496367306] tot_loss[loss=2.262, over 5602508.44 frames. , ppl: 9.604459721075047], batch size: 70
+2022-12-14 09:47:03,483 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:47:04,244 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.582718752746382
+2022-12-14 09:48:44,734 INFO [train.py:421] (2/8) Epoch 11, batch 62200, loss[loss=2.333, over 1400.00 frames. , ppl: 10.309581756938787] tot_loss[loss=2.263, over 5578496.55 frames. , ppl: 9.61167428188017], batch size: 70
+2022-12-14 09:50:24,748 INFO [train.py:421] (2/8) Epoch 11, batch 62400, loss[loss=2.147, over 6300.00 frames. , ppl: 8.557372180484096] tot_loss[loss=2.263, over 5579802.77 frames. , ppl: 9.610993465518849], batch size: 70
+2022-12-14 09:52:03,658 INFO [train.py:421] (2/8) Epoch 11, batch 62600, loss[loss=2.327, over 2870.00 frames. , ppl: 10.250776384773422] tot_loss[loss=2.263, over 5566996.98 frames. , ppl: 9.61185036771913], batch size: 70
+2022-12-14 09:53:42,469 INFO [train.py:421] (2/8) Epoch 11, batch 62800, loss[loss=2.672, over 630.00 frames. , ppl: 14.473781889225732] tot_loss[loss=2.263, over 5564816.61 frames. , ppl: 9.610168885097009], batch size: 70
+2022-12-14 09:55:20,749 INFO [train.py:421] (2/8) Epoch 11, batch 63000, loss[loss=2.503, over 1050.00 frames. , ppl: 12.223542346390445] tot_loss[loss=2.264, over 5548706.43 frames. , ppl: 9.622857845383], batch size: 70
+2022-12-14 09:55:20,749 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 09:55:21,511 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60109500498798
+2022-12-14 09:57:02,141 INFO [train.py:421] (2/8) Epoch 11, batch 63200, loss[loss=2.441, over 1190.00 frames. , ppl: 11.483546843510762] tot_loss[loss=2.264, over 5522757.34 frames. , ppl: 9.626152122319326], batch size: 70
+2022-12-14 09:58:42,040 INFO [train.py:421] (2/8) Epoch 11, batch 63400, loss[loss=3.613, over 420.00 frames. , ppl: 37.07793304080155] tot_loss[loss=2.265, over 5491812.34 frames. , ppl: 9.633811693630923], batch size: 70
+2022-12-14 10:00:19,148 INFO [train.py:421] (2/8) Epoch 11, batch 63600, loss[loss=2.202, over 2240.00 frames. , ppl: 9.039029557861992] tot_loss[loss=2.265, over 5485502.05 frames. , ppl: 9.632300628069975], batch size: 70
+2022-12-14 10:02:00,367 INFO [train.py:421] (2/8) Epoch 11, batch 63800, loss[loss=2.17, over 3430.00 frames. , ppl: 8.761131185113465] tot_loss[loss=2.265, over 5487994.88 frames. , ppl: 9.626826360201132], batch size: 70
+2022-12-14 10:03:38,191 INFO [train.py:421] (2/8) Epoch 11, batch 64000, loss[loss=2.733, over 630.00 frames. , ppl: 15.37446272403168] tot_loss[loss=2.266, over 5445573.07 frames. , ppl: 9.63728258087681], batch size: 70
+2022-12-14 10:03:38,192 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:03:38,952 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61195926586509
+2022-12-14 10:05:19,744 INFO [train.py:421] (2/8) Epoch 11, batch 64200, loss[loss=2.605, over 1330.00 frames. , ppl: 13.527727499195127] tot_loss[loss=2.265, over 5501025.50 frames. , ppl: 9.629219703129055], batch size: 70
+2022-12-14 10:07:02,734 INFO [train.py:421] (2/8) Epoch 11, batch 64400, loss[loss=2.275, over 1750.00 frames. , ppl: 9.72779822860101] tot_loss[loss=2.265, over 5492089.81 frames. , ppl: 9.628472131197867], batch size: 70
+2022-12-14 10:08:42,830 INFO [train.py:421] (2/8) Epoch 11, batch 64600, loss[loss=2.493, over 840.00 frames. , ppl: 12.092615563905394] tot_loss[loss=2.268, over 5422105.83 frames. , ppl: 9.656724140241483], batch size: 70
+2022-12-14 10:10:23,650 INFO [train.py:421] (2/8) Epoch 11, batch 64800, loss[loss=2.236, over 4830.00 frames. , ppl: 9.3524001098405] tot_loss[loss=2.267, over 5441715.57 frames. , ppl: 9.652630273971806], batch size: 70
+2022-12-14 10:12:04,673 INFO [train.py:421] (2/8) Epoch 11, batch 65000, loss[loss=2.204, over 5180.00 frames. , ppl: 9.060247582514645] tot_loss[loss=2.266, over 5460596.29 frames. , ppl: 9.64008939338051], batch size: 70
+2022-12-14 10:12:04,674 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:12:05,419 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.588153854062595
+2022-12-14 10:13:48,133 INFO [train.py:421] (2/8) Epoch 11, batch 65200, loss[loss=2.27, over 2310.00 frames. , ppl: 9.67817819912563] tot_loss[loss=2.266, over 5478512.66 frames. , ppl: 9.643525720047794], batch size: 70
+2022-12-14 10:15:30,902 INFO [train.py:421] (2/8) Epoch 11, batch 65400, loss[loss=2.618, over 840.00 frames. , ppl: 13.70143035028974] tot_loss[loss=2.266, over 5480685.15 frames. , ppl: 9.642335403281072], batch size: 70
+2022-12-14 10:17:13,917 INFO [train.py:421] (2/8) Epoch 11, batch 65600, loss[loss=2.14, over 3290.00 frames. , ppl: 8.501288605554933] tot_loss[loss=2.267, over 5450053.71 frames. , ppl: 9.649752875715329], batch size: 70
+2022-12-14 10:18:52,818 INFO [train.py:421] (2/8) Epoch 11, batch 65800, loss[loss=2.304, over 2030.00 frames. , ppl: 10.01334072722094] tot_loss[loss=2.267, over 5423989.26 frames. , ppl: 9.653135916975904], batch size: 70
+2022-12-14 10:20:34,422 INFO [train.py:421] (2/8) Epoch 11, batch 66000, loss[loss=2.209, over 4060.00 frames. , ppl: 9.108907759622392] tot_loss[loss=2.267, over 5423810.06 frames. , ppl: 9.651347836429226], batch size: 70
+2022-12-14 10:20:34,423 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:20:35,171 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.583815172113377
+2022-12-14 10:22:18,362 INFO [train.py:421] (2/8) Epoch 11, batch 66200, loss[loss=2.197, over 4060.00 frames. , ppl: 9.00187415571844] tot_loss[loss=2.266, over 5456414.97 frames. , ppl: 9.639455751674321], batch size: 70
+2022-12-14 10:24:06,432 INFO [train.py:421] (2/8) Epoch 11, batch 66400, loss[loss=2.185, over 5950.00 frames. , ppl: 8.88886850062605] tot_loss[loss=2.264, over 5519082.04 frames. , ppl: 9.619647304416363], batch size: 70
+2022-12-14 10:25:47,274 INFO [train.py:421] (2/8) Epoch 11, batch 66600, loss[loss=2.183, over 7000.00 frames. , ppl: 8.868648888057715] tot_loss[loss=2.264, over 5532101.05 frames. , ppl: 9.617636183800546], batch size: 70
+2022-12-14 10:27:31,529 INFO [train.py:421] (2/8) Epoch 11, batch 66800, loss[loss=2.34, over 1750.00 frames. , ppl: 10.38130028703579] tot_loss[loss=2.264, over 5529702.22 frames. , ppl: 9.616758712628414], batch size: 70
+2022-12-14 10:29:12,961 INFO [train.py:421] (2/8) Epoch 11, batch 67000, loss[loss=2.633, over 770.00 frames. , ppl: 13.91935517933019] tot_loss[loss=2.264, over 5505603.03 frames. , ppl: 9.620589158651994], batch size: 70
+2022-12-14 10:29:12,961 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:29:13,717 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59918248410303
+2022-12-14 10:30:53,847 INFO [train.py:421] (2/8) Epoch 11, batch 67200, loss[loss=2.102, over 8260.00 frames. , ppl: 8.18484603151676] tot_loss[loss=2.264, over 5511855.38 frames. , ppl: 9.622927382029399], batch size: 70
+2022-12-14 10:32:32,048 INFO [train.py:421] (2/8) Epoch 11, batch 67400, loss[loss=2.951, over 630.00 frames. , ppl: 19.12995414817252] tot_loss[loss=2.265, over 5470538.66 frames. , ppl: 9.634111361986124], batch size: 70
+2022-12-14 10:34:15,961 INFO [train.py:421] (2/8) Epoch 11, batch 67600, loss[loss=2.18, over 6790.00 frames. , ppl: 8.850699610790192] tot_loss[loss=2.266, over 5443415.33 frames. , ppl: 9.64273777700554], batch size: 70
+2022-12-14 10:35:56,758 INFO [train.py:421] (2/8) Epoch 11, batch 67800, loss[loss=2.082, over 4760.00 frames. , ppl: 8.01953349195225] tot_loss[loss=2.265, over 5490377.59 frames. , ppl: 9.627382897854455], batch size: 70
+2022-12-14 10:37:37,872 INFO [train.py:421] (2/8) Epoch 11, batch 68000, loss[loss=2.422, over 1260.00 frames. , ppl: 11.264908775658636] tot_loss[loss=2.264, over 5489732.96 frames. , ppl: 9.622898763931225], batch size: 70
+2022-12-14 10:37:37,873 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:37:38,622 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.586750448742256
+2022-12-14 10:39:17,198 INFO [train.py:421] (2/8) Epoch 11, batch 68200, loss[loss=2.877, over 700.00 frames. , ppl: 17.76938408222048] tot_loss[loss=2.266, over 5455531.04 frames. , ppl: 9.638329711729655], batch size: 70
+2022-12-14 10:40:57,005 INFO [train.py:421] (2/8) Epoch 11, batch 68400, loss[loss=2.371, over 1680.00 frames. , ppl: 10.711418044567738] tot_loss[loss=2.266, over 5464576.17 frames. , ppl: 9.637455786650515], batch size: 70
+2022-12-14 10:42:35,941 INFO [train.py:421] (2/8) Epoch 11, batch 68600, loss[loss=2.299, over 2800.00 frames. , ppl: 9.96628019843091] tot_loss[loss=2.265, over 5474512.46 frames. , ppl: 9.63261100217871], batch size: 70
+2022-12-14 10:44:17,983 INFO [train.py:421] (2/8) Epoch 11, batch 68800, loss[loss=2.177, over 6440.00 frames. , ppl: 8.817797009660765] tot_loss[loss=2.265, over 5496507.22 frames. , ppl: 9.629132784704385], batch size: 70
+2022-12-14 10:46:02,171 INFO [train.py:421] (2/8) Epoch 11, batch 69000, loss[loss=2.257, over 2800.00 frames. , ppl: 9.550237160296714] tot_loss[loss=2.264, over 5480598.00 frames. , ppl: 9.624972368316566], batch size: 70
+2022-12-14 10:46:02,172 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:46:02,935 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.590407678839963
+2022-12-14 10:47:43,618 INFO [train.py:421] (2/8) Epoch 11, batch 69200, loss[loss=2.223, over 2240.00 frames. , ppl: 9.233492786059957] tot_loss[loss=2.267, over 5390665.82 frames. , ppl: 9.6486746762254], batch size: 70
+2022-12-14 10:49:26,939 INFO [train.py:421] (2/8) Epoch 11, batch 69400, loss[loss=2.267, over 2450.00 frames. , ppl: 9.652700523677984] tot_loss[loss=2.268, over 5369038.41 frames. , ppl: 9.660266865814327], batch size: 70
+2022-12-14 10:51:09,337 INFO [train.py:421] (2/8) Epoch 11, batch 69600, loss[loss=2.278, over 4060.00 frames. , ppl: 9.759880373631912] tot_loss[loss=2.268, over 5394790.04 frames. , ppl: 9.657512338045663], batch size: 70
+2022-12-14 10:52:54,763 INFO [train.py:421] (2/8) Epoch 11, batch 69800, loss[loss=2.221, over 3850.00 frames. , ppl: 9.220100685012316] tot_loss[loss=2.266, over 5434479.96 frames. , ppl: 9.64352387918208], batch size: 70
+2022-12-14 10:54:39,528 INFO [train.py:421] (2/8) Epoch 11, batch 70000, loss[loss=2.371, over 1470.00 frames. , ppl: 10.706150462934415] tot_loss[loss=2.266, over 5465327.16 frames. , ppl: 9.639116585351385], batch size: 70
+2022-12-14 10:54:39,529 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 10:54:40,301 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.575361947313867
+2022-12-14 10:56:19,535 INFO [train.py:421] (2/8) Epoch 11, batch 70200, loss[loss=2.304, over 3080.00 frames. , ppl: 10.018134585798323] tot_loss[loss=2.267, over 5445589.77 frames. , ppl: 9.65009926914341], batch size: 70
+2022-12-14 10:58:03,424 INFO [train.py:421] (2/8) Epoch 11, batch 70400, loss[loss=2.492, over 1330.00 frames. , ppl: 12.080280289030524] tot_loss[loss=2.266, over 5459659.67 frames. , ppl: 9.642000738086605], batch size: 70
+2022-12-14 10:59:41,666 INFO [train.py:421] (2/8) Epoch 11, batch 70600, loss[loss=2.297, over 1820.00 frames. , ppl: 9.944553079676634] tot_loss[loss=2.266, over 5443024.13 frames. , ppl: 9.64386276938609], batch size: 70
+2022-12-14 11:01:20,612 INFO [train.py:421] (2/8) Epoch 11, batch 70800, loss[loss=2.192, over 4760.00 frames. , ppl: 8.951490530143111] tot_loss[loss=2.267, over 5451802.81 frames. , ppl: 9.646392867191565], batch size: 70
+2022-12-14 11:03:04,092 INFO [train.py:421] (2/8) Epoch 11, batch 71000, loss[loss=2.743, over 630.00 frames. , ppl: 15.537816585698737] tot_loss[loss=2.267, over 5440017.27 frames. , ppl: 9.64943392723691], batch size: 70
+2022-12-14 11:03:04,093 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 11:03:04,885 INFO [train.py:452] (2/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.57697205031597
+2022-12-14 11:04:44,371 INFO [train.py:421] (2/8) Epoch 11, batch 71200, loss[loss=2.394, over 1260.00 frames. , ppl: 10.954090973565958] tot_loss[loss=2.266, over 5448517.22 frames. , ppl: 9.645341677403165], batch size: 70
+2022-12-14 11:06:25,521 INFO [train.py:421] (2/8) Epoch 11, batch 71400, loss[loss=2.18, over 6650.00 frames. , ppl: 8.842094301929443] tot_loss[loss=2.266, over 5456202.15 frames. , ppl: 9.641994973148057], batch size: 70
+2022-12-14 11:08:07,130 INFO [train.py:421] (2/8) Epoch 11, batch 71600, loss[loss=2.224, over 3990.00 frames. , ppl: 9.243541089132234] tot_loss[loss=2.266, over 5464424.52 frames. , ppl: 9.644428863203904], batch size: 70
+2022-12-14 11:09:48,360 INFO [train.py:421] (2/8) Epoch 11, batch 71800, loss[loss=2.334, over 1470.00 frames. , ppl: 10.322152265163338] tot_loss[loss=2.265, over 5523936.31 frames. , ppl: 9.635176978344287], batch size: 70
+2022-12-14 11:11:04,452 INFO [train.py:421] (2/8) Epoch 12, batch 0, loss[loss=2.991, over 560.00 frames. , ppl: 19.89703927060972] tot_loss[loss=2.991, over 560.00 frames. , ppl: 19.89703927060972], batch size: 70
+2022-12-14 11:12:45,932 INFO [train.py:421] (2/8) Epoch 12, batch 200, loss[loss=2.332, over 1470.00 frames. , ppl: 10.295539175093822] tot_loss[loss=2.252, over 525781.25 frames. , ppl: 9.504299922466128], batch size: 70
+2022-12-14 11:14:25,215 INFO [train.py:421] (2/8) Epoch 12, batch 400, loss[loss=2.269, over 2800.00 frames. , ppl: 9.67253834081985] tot_loss[loss=2.26, over 996471.14 frames. , ppl: 9.586634536720933], batch size: 70
+2022-12-14 11:16:08,524 INFO [train.py:421] (2/8) Epoch 12, batch 600, loss[loss=2.261, over 3360.00 frames. , ppl: 9.591539480524498] tot_loss[loss=2.257, over 1457198.30 frames. , ppl: 9.551755200582898], batch size: 70
+2022-12-14 11:17:49,325 INFO [train.py:421] (2/8) Epoch 12, batch 800, loss[loss=2.124, over 7560.00 frames. , ppl: 8.363470440364292] tot_loss[loss=2.255, over 1882573.74 frames. , ppl: 9.531826131723493], batch size: 70
+2022-12-14 11:19:30,204 INFO [train.py:421] (2/8) Epoch 12, batch 1000, loss[loss=2.136, over 4690.00 frames. , ppl: 8.46312818086212] tot_loss[loss=2.256, over 2208236.61 frames. , ppl: 9.543513463972973], batch size: 70
+2022-12-14 11:19:30,205 INFO [train.py:441] (2/8) Computing validation loss
+2022-12-14 11:19:30,968 INFO [train.py:452] (2/8) Epoch 12, validation: loss=2.26, over 211138.00 frames. , ppl: 9.584870576501038
+2022-12-14 11:21:11,006 INFO [train.py:421] (2/8) Epoch 12, batch 1200, loss[loss=2.338, over 3080.00 frames. , ppl: 10.363202128720046] tot_loss[loss=2.261, over 2457075.61 frames. , ppl: 9.590926918832611], batch size: 70