diff --git "a/exp/log/log-train-2022-12-09-10-39-23-4" "b/exp/log/log-train-2022-12-09-10-39-23-4"
new file mode 100644--- /dev/null
+++ "b/exp/log/log-train-2022-12-09-10-39-23-4"
@@ -0,0 +1,6042 @@
+2022-12-09 10:39:23,963 INFO [train.py:493] (4/8) Training started
+2022-12-09 10:39:23,963 INFO [train.py:494] (4/8) {'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 10:39:23,963 INFO [train.py:505] (4/8) Device: cuda:4
+2022-12-09 10:39:23,963 INFO [train.py:507] (4/8) About to create model
+2022-12-09 10:39:24,420 INFO [model.py:64] (4/8) Tying weights
+2022-12-09 10:39:24,421 INFO [train.py:520] (4/8) Number of model parameters: 98611638
+2022-12-09 10:39:27,368 INFO [train.py:539] (4/8) Loading LM training data from ./transformer_lm/libri_lm_training_bpe500/sorted-lm-data-libri-lm_maxlen200.pt
+2022-12-09 10:39:41,315 INFO [train.py:546] (4/8) Loading LM validation data from ./transformer_lm/libri_lm_training_bpe500/sorted_lm_data-test.pt
+2022-12-09 10:39:44,738 INFO [train.py:421] (4/8) Epoch 0, batch 0, loss[loss=78.42, over 2520.00 frames. , ppl: 1.143889694218019e+34] tot_loss[loss=78.42, over 2520.00 frames. , ppl: 1.143889694218019e+34], batch size: 70
+2022-12-09 10:39:44,761 INFO [distributed.py:995] (4/8) Reducer buckets have been rebuilt in this iteration.
+2022-12-09 10:41:21,499 INFO [train.py:421] (4/8) Epoch 0, batch 200, loss[loss=7.968, over 1540.00 frames. , ppl: 2888.0089394154215] tot_loss[loss=17.98, over 527255.55 frames. , ppl: 64397219.834707424], batch size: 70
+2022-12-09 10:43:01,687 INFO [train.py:421] (4/8) Epoch 0, batch 400, loss[loss=6.639, over 630.00 frames. , ppl: 764.2577985687722] tot_loss[loss=12.61, over 978757.45 frames. , ppl: 299815.32478287467], batch size: 70
+2022-12-09 10:44:42,143 INFO [train.py:421] (4/8) Epoch 0, batch 600, loss[loss=6.639, over 910.00 frames. , ppl: 764.2521485992542] tot_loss[loss=10.32, over 1422908.16 frames. , ppl: 30191.29167458764], batch size: 70
+2022-12-09 10:46:22,415 INFO [train.py:421] (4/8) Epoch 0, batch 800, loss[loss=5.373, over 12250.00 frames. , ppl: 215.44699652028976] tot_loss[loss=9.017, over 1856333.67 frames. , ppl: 8239.805780999899], batch size: 70
+2022-12-09 10:48:02,281 INFO [train.py:421] (4/8) Epoch 0, batch 1000, loss[loss=5.928, over 490.00 frames. , ppl: 375.22023754567175] tot_loss[loss=8.15, over 2215461.83 frames. , ppl: 3463.7590448985366], batch size: 70
+2022-12-09 10:48:02,281 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 10:48:03,029 INFO [train.py:452] (4/8) Epoch 0, validation: loss=5.053, over 211138.00 frames. , ppl: 156.50471436114304
+2022-12-09 10:49:43,715 INFO [train.py:421] (4/8) Epoch 0, batch 1200, loss[loss=5.176, over 1400.00 frames. , ppl: 176.99033827894272] tot_loss[loss=7.599, over 2571723.42 frames. , ppl: 1995.2692913328206], batch size: 70
+2022-12-09 10:51:23,207 INFO [train.py:421] (4/8) Epoch 0, batch 1400, loss[loss=5.226, over 1190.00 frames. , ppl: 185.980025683674] tot_loss[loss=7.211, over 2835928.82 frames. , ppl: 1354.2587324767828], batch size: 70
+2022-12-09 10:53:02,876 INFO [train.py:421] (4/8) Epoch 0, batch 1600, loss[loss=4.974, over 1890.00 frames. , ppl: 144.65872922500543] tot_loss[loss=6.861, over 3064175.81 frames. , ppl: 954.062729690253], batch size: 70
+2022-12-09 10:54:41,883 INFO [train.py:421] (4/8) Epoch 0, batch 1800, loss[loss=4.485, over 8050.00 frames. , ppl: 88.68762298256667] tot_loss[loss=6.521, over 3288817.52 frames. , ppl: 679.2716234223385], batch size: 70
+2022-12-09 10:56:19,726 INFO [train.py:421] (4/8) Epoch 0, batch 2000, loss[loss=4.456, over 2520.00 frames. , ppl: 86.13701318890303] tot_loss[loss=6.208, over 3516662.68 frames. , ppl: 496.93163902503176], batch size: 70
+2022-12-09 10:56:19,726 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 10:56:20,475 INFO [train.py:452] (4/8) Epoch 0, validation: loss=4.224, over 211138.00 frames. , ppl: 68.31401714385908
+2022-12-09 10:58:02,726 INFO [train.py:421] (4/8) Epoch 0, batch 2200, loss[loss=4.404, over 1470.00 frames. , ppl: 81.73937950379137] tot_loss[loss=5.943, over 3764324.77 frames. , ppl: 380.9161930921382], batch size: 70
+2022-12-09 10:59:45,477 INFO [train.py:421] (4/8) Epoch 0, batch 2400, loss[loss=4.254, over 1120.00 frames. , ppl: 70.4132497610174] tot_loss[loss=5.743, over 3917126.09 frames. , ppl: 311.8821781631969], batch size: 70
+2022-12-09 11:01:26,791 INFO [train.py:421] (4/8) Epoch 0, batch 2600, loss[loss=3.92, over 1050.00 frames. , ppl: 50.39880416528986] tot_loss[loss=5.562, over 4021552.53 frames. , ppl: 260.43225655259397], batch size: 70
+2022-12-09 11:03:09,418 INFO [train.py:421] (4/8) Epoch 0, batch 2800, loss[loss=4.157, over 1190.00 frames. , ppl: 63.892654722048846] tot_loss[loss=5.365, over 4197626.47 frames. , ppl: 213.86450028866614], batch size: 70
+2022-12-09 11:04:49,031 INFO [train.py:421] (4/8) Epoch 0, batch 3000, loss[loss=3.704, over 5180.00 frames. , ppl: 40.598599079019785] tot_loss[loss=5.178, over 4341516.36 frames. , ppl: 177.37231352059052], batch size: 70
+2022-12-09 11:04:49,031 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:04:49,779 INFO [train.py:452] (4/8) Epoch 0, validation: loss=3.608, over 211138.00 frames. , ppl: 36.89311779083857
+2022-12-09 11:06:26,626 INFO [train.py:421] (4/8) Epoch 0, batch 3200, loss[loss=3.676, over 840.00 frames. , ppl: 39.505911208947765] tot_loss[loss=4.995, over 4462408.71 frames. , ppl: 147.6393141541928], batch size: 70
+2022-12-09 11:08:06,512 INFO [train.py:421] (4/8) Epoch 0, batch 3400, loss[loss=3.368, over 910.00 frames. , ppl: 29.02050224276307] tot_loss[loss=4.827, over 4558674.05 frames. , ppl: 124.82704276198676], batch size: 70
+2022-12-09 11:09:49,633 INFO [train.py:421] (4/8) Epoch 0, batch 3600, loss[loss=3.305, over 1330.00 frames. , ppl: 27.23716772915142] tot_loss[loss=4.664, over 4650237.17 frames. , ppl: 106.11061850521405], batch size: 70
+2022-12-09 11:11:30,325 INFO [train.py:421] (4/8) Epoch 0, batch 3800, loss[loss=3.176, over 2590.00 frames. , ppl: 23.9490053266254] tot_loss[loss=4.512, over 4732080.02 frames. , ppl: 91.0993751913577], batch size: 70
+2022-12-09 11:13:10,810 INFO [train.py:421] (4/8) Epoch 0, batch 4000, loss[loss=3.369, over 910.00 frames. , ppl: 29.063090678279206] tot_loss[loss=4.368, over 4819135.14 frames. , ppl: 78.87709496742148], batch size: 70
+2022-12-09 11:13:10,811 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:13:11,561 INFO [train.py:452] (4/8) Epoch 0, validation: loss=3.125, over 211138.00 frames. , ppl: 22.762158928396982
+2022-12-09 11:14:48,442 INFO [train.py:421] (4/8) Epoch 0, batch 4200, loss[loss=3.162, over 4900.00 frames. , ppl: 23.60792897555145] tot_loss[loss=4.237, over 4889538.41 frames. , ppl: 69.20535696386378], batch size: 70
+2022-12-09 11:16:24,342 INFO [train.py:421] (4/8) Epoch 0, batch 4400, loss[loss=3.027, over 7140.00 frames. , ppl: 20.641430719988982] tot_loss[loss=4.124, over 4924444.88 frames. , ppl: 61.78988362007549], batch size: 70
+2022-12-09 11:18:03,018 INFO [train.py:421] (4/8) Epoch 0, batch 4600, loss[loss=2.985, over 1960.00 frames. , ppl: 19.785359663528507] tot_loss[loss=4.017, over 4960939.53 frames. , ppl: 55.56142546771177], batch size: 70
+2022-12-09 11:19:43,496 INFO [train.py:421] (4/8) Epoch 0, batch 4800, loss[loss=3.023, over 2730.00 frames. , ppl: 20.555453600923368] tot_loss[loss=3.912, over 5037987.57 frames. , ppl: 49.97442495220485], batch size: 70
+2022-12-09 11:21:24,000 INFO [train.py:421] (4/8) Epoch 0, batch 5000, loss[loss=3.072, over 3010.00 frames. , ppl: 21.59554174407963] tot_loss[loss=3.822, over 5064941.08 frames. , ppl: 45.69435552958405], batch size: 70
+2022-12-09 11:21:24,000 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:21:24,746 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.954, over 211138.00 frames. , ppl: 19.189241084043
+2022-12-09 11:23:04,385 INFO [train.py:421] (4/8) Epoch 0, batch 5200, loss[loss=3.044, over 910.00 frames. , ppl: 20.997424548681092] tot_loss[loss=3.728, over 5166084.68 frames. , ppl: 41.61440590295809], batch size: 70
+2022-12-09 11:24:42,543 INFO [train.py:421] (4/8) Epoch 0, batch 5400, loss[loss=3.011, over 2870.00 frames. , ppl: 20.313276829821046] tot_loss[loss=3.656, over 5178324.05 frames. , ppl: 38.700318886336575], batch size: 70
+2022-12-09 11:26:20,377 INFO [train.py:421] (4/8) Epoch 0, batch 5600, loss[loss=3.083, over 1120.00 frames. , ppl: 21.814333843027708] tot_loss[loss=3.59, over 5177229.57 frames. , ppl: 36.23533942392434], batch size: 70
+2022-12-09 11:28:00,853 INFO [train.py:421] (4/8) Epoch 0, batch 5800, loss[loss=2.864, over 5880.00 frames. , ppl: 17.52765922759183] tot_loss[loss=3.527, over 5186134.98 frames. , ppl: 34.02120962449334], batch size: 70
+2022-12-09 11:29:42,214 INFO [train.py:421] (4/8) Epoch 0, batch 6000, loss[loss=3.214, over 840.00 frames. , ppl: 24.88935835158874] tot_loss[loss=3.473, over 5164231.59 frames. , ppl: 32.23062562055949], batch size: 70
+2022-12-09 11:29:42,214 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:29:42,974 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.875, over 211138.00 frames. , ppl: 17.725366404663895
+2022-12-09 11:31:26,510 INFO [train.py:421] (4/8) Epoch 0, batch 6200, loss[loss=2.805, over 5390.00 frames. , ppl: 16.52781313951741] tot_loss[loss=3.411, over 5248162.79 frames. , ppl: 30.29437202861338], batch size: 70
+2022-12-09 11:33:07,631 INFO [train.py:421] (4/8) Epoch 0, batch 6400, loss[loss=2.936, over 1470.00 frames. , ppl: 18.836891419890183] tot_loss[loss=3.36, over 5271991.53 frames. , ppl: 28.79120399961592], batch size: 70
+2022-12-09 11:34:47,526 INFO [train.py:421] (4/8) Epoch 0, batch 6600, loss[loss=2.852, over 6580.00 frames. , ppl: 17.32433610950534] tot_loss[loss=3.313, over 5302873.34 frames. , ppl: 27.461887252130808], batch size: 70
+2022-12-09 11:36:26,660 INFO [train.py:421] (4/8) Epoch 0, batch 6800, loss[loss=2.957, over 1330.00 frames. , ppl: 19.230554259147702] tot_loss[loss=3.272, over 5293537.29 frames. , ppl: 26.373864531932632], batch size: 70
+2022-12-09 11:38:05,181 INFO [train.py:421] (4/8) Epoch 0, batch 7000, loss[loss=2.962, over 1260.00 frames. , ppl: 19.33677297495803] tot_loss[loss=3.233, over 5297388.76 frames. , ppl: 25.356650220510005], batch size: 70
+2022-12-09 11:38:05,182 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:38:05,970 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.816, over 211138.00 frames. , ppl: 16.709705528719027
+2022-12-09 11:39:47,610 INFO [train.py:421] (4/8) Epoch 0, batch 7200, loss[loss=2.889, over 1470.00 frames. , ppl: 17.974621705249472] tot_loss[loss=3.198, over 5287516.02 frames. , ppl: 24.49526016502389], batch size: 70
+2022-12-09 11:41:23,910 INFO [train.py:421] (4/8) Epoch 0, batch 7400, loss[loss=2.908, over 2310.00 frames. , ppl: 18.315870041543995] tot_loss[loss=3.163, over 5303730.90 frames. , ppl: 23.647704049232477], batch size: 70
+2022-12-09 11:43:02,498 INFO [train.py:421] (4/8) Epoch 0, batch 7600, loss[loss=3.145, over 910.00 frames. , ppl: 23.215239230722382] tot_loss[loss=3.135, over 5273259.34 frames. , ppl: 22.986131038129802], batch size: 70
+2022-12-09 11:44:44,304 INFO [train.py:421] (4/8) Epoch 0, batch 7800, loss[loss=2.739, over 980.00 frames. , ppl: 15.475456560993724] tot_loss[loss=3.101, over 5354851.91 frames. , ppl: 22.212104116219066], batch size: 70
+2022-12-09 11:46:23,635 INFO [train.py:421] (4/8) Epoch 0, batch 8000, loss[loss=2.805, over 1610.00 frames. , ppl: 16.528061360738135] tot_loss[loss=3.071, over 5393791.58 frames. , ppl: 21.56732634292846], batch size: 70
+2022-12-09 11:46:23,636 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:46:24,382 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.773, over 211138.00 frames. , ppl: 16.01179093613074
+2022-12-09 11:47:58,403 INFO [train.py:421] (4/8) Epoch 0, batch 8200, loss[loss=2.666, over 4830.00 frames. , ppl: 14.37760672089973] tot_loss[loss=3.047, over 5382164.64 frames. , ppl: 21.043732003876563], batch size: 70
+2022-12-09 11:49:42,155 INFO [train.py:421] (4/8) Epoch 0, batch 8400, loss[loss=2.897, over 1190.00 frames. , ppl: 18.12319089034297] tot_loss[loss=3.025, over 5351974.52 frames. , ppl: 20.591478857023805], batch size: 70
+2022-12-09 11:51:19,108 INFO [train.py:421] (4/8) Epoch 0, batch 8600, loss[loss=2.911, over 1960.00 frames. , ppl: 18.378008054214288] tot_loss[loss=3.001, over 5383201.22 frames. , ppl: 20.10838630718875], batch size: 70
+2022-12-09 11:53:00,420 INFO [train.py:421] (4/8) Epoch 0, batch 8800, loss[loss=2.992, over 840.00 frames. , ppl: 19.92334592355185] tot_loss[loss=2.98, over 5407201.23 frames. , ppl: 19.686320736067348], batch size: 70
+2022-12-09 11:54:40,008 INFO [train.py:421] (4/8) Epoch 0, batch 9000, loss[loss=2.72, over 2030.00 frames. , ppl: 15.185147192263154] tot_loss[loss=2.96, over 5423434.92 frames. , ppl: 19.301567337728113], batch size: 70
+2022-12-09 11:54:40,008 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 11:54:40,785 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.745, over 211138.00 frames. , ppl: 15.565240729798518
+2022-12-09 11:56:21,411 INFO [train.py:421] (4/8) Epoch 0, batch 9200, loss[loss=2.679, over 2170.00 frames. , ppl: 14.574195699746078] tot_loss[loss=2.943, over 5405286.08 frames. , ppl: 18.977432807771454], batch size: 70
+2022-12-09 11:58:04,161 INFO [train.py:421] (4/8) Epoch 0, batch 9400, loss[loss=2.749, over 4900.00 frames. , ppl: 15.62083668422492] tot_loss[loss=2.923, over 5482538.08 frames. , ppl: 18.60354787861859], batch size: 70
+2022-12-09 11:59:43,942 INFO [train.py:421] (4/8) Epoch 0, batch 9600, loss[loss=2.923, over 770.00 frames. , ppl: 18.589972528308433] tot_loss[loss=2.908, over 5463866.71 frames. , ppl: 18.328465412990358], batch size: 70
+2022-12-09 12:01:21,907 INFO [train.py:421] (4/8) Epoch 0, batch 9800, loss[loss=2.858, over 2100.00 frames. , ppl: 17.43210815615129] tot_loss[loss=2.893, over 5498263.63 frames. , ppl: 18.04561543108665], batch size: 70
+2022-12-09 12:03:00,591 INFO [train.py:421] (4/8) Epoch 0, batch 10000, loss[loss=2.846, over 1050.00 frames. , ppl: 17.226149885139023] tot_loss[loss=2.881, over 5456719.53 frames. , ppl: 17.83210700825054], batch size: 70
+2022-12-09 12:03:00,591 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:03:01,337 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.716, over 211138.00 frames. , ppl: 15.11237530515154
+2022-12-09 12:04:40,066 INFO [train.py:421] (4/8) Epoch 0, batch 10200, loss[loss=2.655, over 5320.00 frames. , ppl: 14.223058079368421] tot_loss[loss=2.87, over 5421009.21 frames. , ppl: 17.63767849904128], batch size: 70
+2022-12-09 12:06:22,433 INFO [train.py:421] (4/8) Epoch 0, batch 10400, loss[loss=2.974, over 630.00 frames. , ppl: 19.571533554814145] tot_loss[loss=2.858, over 5425196.47 frames. , ppl: 17.426243217065], batch size: 70
+2022-12-09 12:08:01,855 INFO [train.py:421] (4/8) Epoch 0, batch 10600, loss[loss=2.884, over 1470.00 frames. , ppl: 17.89296974900547] tot_loss[loss=2.847, over 5413210.11 frames. , ppl: 17.233250651013797], batch size: 70
+2022-12-09 12:09:45,721 INFO [train.py:421] (4/8) Epoch 0, batch 10800, loss[loss=2.879, over 1330.00 frames. , ppl: 17.79100094601459] tot_loss[loss=2.835, over 5436137.18 frames. , ppl: 17.0263154823821], batch size: 70
+2022-12-09 12:11:29,859 INFO [train.py:421] (4/8) Epoch 0, batch 11000, loss[loss=2.65, over 3220.00 frames. , ppl: 14.148433566610032] tot_loss[loss=2.824, over 5457388.78 frames. , ppl: 16.846369122993075], batch size: 70
+2022-12-09 12:11:29,859 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:11:30,607 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.692, over 211138.00 frames. , ppl: 14.756337129748474
+2022-12-09 12:13:13,813 INFO [train.py:421] (4/8) Epoch 0, batch 11200, loss[loss=2.777, over 1120.00 frames. , ppl: 16.07323414287345] tot_loss[loss=2.813, over 5516879.72 frames. , ppl: 16.661076571462058], batch size: 70
+2022-12-09 12:14:55,285 INFO [train.py:421] (4/8) Epoch 0, batch 11400, loss[loss=3.043, over 910.00 frames. , ppl: 20.963640548973906] tot_loss[loss=2.805, over 5471281.58 frames. , ppl: 16.531502426508943], batch size: 70
+2022-12-09 12:16:35,945 INFO [train.py:421] (4/8) Epoch 0, batch 11600, loss[loss=2.72, over 1750.00 frames. , ppl: 15.176468355542896] tot_loss[loss=2.797, over 5457766.19 frames. , ppl: 16.397228364583874], batch size: 70
+2022-12-09 12:18:18,166 INFO [train.py:421] (4/8) Epoch 0, batch 11800, loss[loss=2.901, over 630.00 frames. , ppl: 18.191253655947815] tot_loss[loss=2.787, over 5483676.68 frames. , ppl: 16.236884087295856], batch size: 70
+2022-12-09 12:19:58,700 INFO [train.py:421] (4/8) Epoch 0, batch 12000, loss[loss=2.791, over 1750.00 frames. , ppl: 16.292048656996087] tot_loss[loss=2.78, over 5473415.98 frames. , ppl: 16.11733304087621], batch size: 70
+2022-12-09 12:19:58,701 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:19:59,461 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.67, over 211138.00 frames. , ppl: 14.443332034874045
+2022-12-09 12:21:38,247 INFO [train.py:421] (4/8) Epoch 0, batch 12200, loss[loss=2.621, over 4200.00 frames. , ppl: 13.75520655127654] tot_loss[loss=2.772, over 5467591.68 frames. , ppl: 15.998183805265104], batch size: 70
+2022-12-09 12:23:19,260 INFO [train.py:421] (4/8) Epoch 0, batch 12400, loss[loss=4.269, over 350.00 frames. , ppl: 71.46969453948633] tot_loss[loss=2.766, over 5449540.07 frames. , ppl: 15.90209159023916], batch size: 70
+2022-12-09 12:25:01,000 INFO [train.py:421] (4/8) Epoch 0, batch 12600, loss[loss=2.659, over 2030.00 frames. , ppl: 14.285373943147892] tot_loss[loss=2.76, over 5445524.85 frames. , ppl: 15.79260442399321], batch size: 70
+2022-12-09 12:26:41,331 INFO [train.py:421] (4/8) Epoch 0, batch 12800, loss[loss=2.684, over 1750.00 frames. , ppl: 14.638574831947267] tot_loss[loss=2.753, over 5440337.62 frames. , ppl: 15.694060470765683], batch size: 70
+2022-12-09 12:28:17,833 INFO [train.py:421] (4/8) Epoch 0, batch 13000, loss[loss=2.662, over 2030.00 frames. , ppl: 14.326570203179463] tot_loss[loss=2.747, over 5424124.83 frames. , ppl: 15.598956126650338], batch size: 70
+2022-12-09 12:28:17,833 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:28:18,579 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.655, over 211138.00 frames. , ppl: 14.220370751795132
+2022-12-09 12:29:56,788 INFO [train.py:421] (4/8) Epoch 0, batch 13200, loss[loss=2.611, over 6020.00 frames. , ppl: 13.616364227391728] tot_loss[loss=2.743, over 5388121.71 frames. , ppl: 15.532722994064896], batch size: 70
+2022-12-09 12:31:35,020 INFO [train.py:421] (4/8) Epoch 0, batch 13400, loss[loss=2.639, over 9240.00 frames. , ppl: 13.994143213070636] tot_loss[loss=2.736, over 5400812.09 frames. , ppl: 15.427912763937439], batch size: 70
+2022-12-09 12:33:14,489 INFO [train.py:421] (4/8) Epoch 0, batch 13600, loss[loss=3.073, over 630.00 frames. , ppl: 21.60790201963068] tot_loss[loss=2.73, over 5415543.40 frames. , ppl: 15.33449390710946], batch size: 70
+2022-12-09 12:34:55,724 INFO [train.py:421] (4/8) Epoch 0, batch 13800, loss[loss=2.822, over 1260.00 frames. , ppl: 16.818422927263466] tot_loss[loss=2.724, over 5427070.07 frames. , ppl: 15.243384661165985], batch size: 70
+2022-12-09 12:36:40,250 INFO [train.py:421] (4/8) Epoch 0, batch 14000, loss[loss=2.642, over 3570.00 frames. , ppl: 14.039312526823027] tot_loss[loss=2.718, over 5442882.23 frames. , ppl: 15.147632563855849], batch size: 70
+2022-12-09 12:36:40,250 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:36:41,012 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.637, over 211138.00 frames. , ppl: 13.971456542635757
+2022-12-09 12:38:24,308 INFO [train.py:421] (4/8) Epoch 0, batch 14200, loss[loss=2.989, over 630.00 frames. , ppl: 19.85594958118977] tot_loss[loss=2.713, over 5481580.39 frames. , ppl: 15.078303756566504], batch size: 70
+2022-12-09 12:40:02,569 INFO [train.py:421] (4/8) Epoch 0, batch 14400, loss[loss=2.733, over 1680.00 frames. , ppl: 15.376849082452162] tot_loss[loss=2.709, over 5482422.43 frames. , ppl: 15.00982204075809], batch size: 70
+2022-12-09 12:41:41,073 INFO [train.py:421] (4/8) Epoch 0, batch 14600, loss[loss=2.622, over 5670.00 frames. , ppl: 13.765034260311412] tot_loss[loss=2.703, over 5542406.73 frames. , ppl: 14.926618690748489], batch size: 70
+2022-12-09 12:43:22,501 INFO [train.py:421] (4/8) Epoch 0, batch 14800, loss[loss=2.596, over 2100.00 frames. , ppl: 13.404987780424802] tot_loss[loss=2.699, over 5533840.01 frames. , ppl: 14.859560218044592], batch size: 70
+2022-12-09 12:45:00,509 INFO [train.py:421] (4/8) Epoch 0, batch 15000, loss[loss=2.68, over 4900.00 frames. , ppl: 14.579699297312171] tot_loss[loss=2.695, over 5526748.01 frames. , ppl: 14.800045747405507], batch size: 70
+2022-12-09 12:45:00,509 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:45:01,254 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.624, over 211138.00 frames. , ppl: 13.790181361567887
+2022-12-09 12:46:44,176 INFO [train.py:421] (4/8) Epoch 0, batch 15200, loss[loss=2.946, over 700.00 frames. , ppl: 19.024100309532027] tot_loss[loss=2.69, over 5532747.84 frames. , ppl: 14.733973733984657], batch size: 70
+2022-12-09 12:48:23,216 INFO [train.py:421] (4/8) Epoch 0, batch 15400, loss[loss=2.712, over 2240.00 frames. , ppl: 15.06175990424465] tot_loss[loss=2.687, over 5509378.67 frames. , ppl: 14.694159508485823], batch size: 70
+2022-12-09 12:50:02,332 INFO [train.py:421] (4/8) Epoch 0, batch 15600, loss[loss=2.677, over 1120.00 frames. , ppl: 14.539517563072701] tot_loss[loss=2.683, over 5529320.67 frames. , ppl: 14.629075169077149], batch size: 70
+2022-12-09 12:51:41,375 INFO [train.py:421] (4/8) Epoch 0, batch 15800, loss[loss=2.636, over 3710.00 frames. , ppl: 13.96103472980252] tot_loss[loss=2.68, over 5518079.74 frames. , ppl: 14.583948050491228], batch size: 70
+2022-12-09 12:53:18,247 INFO [train.py:421] (4/8) Epoch 0, batch 16000, loss[loss=2.759, over 2730.00 frames. , ppl: 15.785599942805359] tot_loss[loss=2.677, over 5481615.79 frames. , ppl: 14.54800277999886], batch size: 70
+2022-12-09 12:53:18,248 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 12:53:18,996 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.614, over 211138.00 frames. , ppl: 13.652930714419472
+2022-12-09 12:54:59,738 INFO [train.py:421] (4/8) Epoch 0, batch 16200, loss[loss=2.505, over 3220.00 frames. , ppl: 12.237962922175493] tot_loss[loss=2.673, over 5532013.34 frames. , ppl: 14.482697914870538], batch size: 70
+2022-12-09 12:56:38,652 INFO [train.py:421] (4/8) Epoch 0, batch 16400, loss[loss=2.561, over 3850.00 frames. , ppl: 12.944016359571783] tot_loss[loss=2.669, over 5544652.56 frames. , ppl: 14.431901506027872], batch size: 70
+2022-12-09 12:58:19,165 INFO [train.py:421] (4/8) Epoch 0, batch 16600, loss[loss=2.522, over 2660.00 frames. , ppl: 12.45967317708922] tot_loss[loss=2.665, over 5561722.92 frames. , ppl: 14.374620668339295], batch size: 70
+2022-12-09 13:00:02,388 INFO [train.py:421] (4/8) Epoch 0, batch 16800, loss[loss=2.625, over 3710.00 frames. , ppl: 13.810723776934921] tot_loss[loss=2.663, over 5544567.54 frames. , ppl: 14.34221504251771], batch size: 70
+2022-12-09 13:01:41,561 INFO [train.py:421] (4/8) Epoch 0, batch 17000, loss[loss=2.718, over 2170.00 frames. , ppl: 15.14298172853558] tot_loss[loss=2.661, over 5490558.50 frames. , ppl: 14.313510331774008], batch size: 70
+2022-12-09 13:01:41,561 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:01:42,309 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.601, over 211138.00 frames. , ppl: 13.482567378311968
+2022-12-09 13:03:23,547 INFO [train.py:421] (4/8) Epoch 0, batch 17200, loss[loss=2.588, over 4830.00 frames. , ppl: 13.307381173736138] tot_loss[loss=2.659, over 5460172.09 frames. , ppl: 14.281552685432517], batch size: 70
+2022-12-09 13:05:01,411 INFO [train.py:421] (4/8) Epoch 0, batch 17400, loss[loss=2.542, over 1120.00 frames. , ppl: 12.70228972040549] tot_loss[loss=2.657, over 5417193.57 frames. , ppl: 14.252117321558288], batch size: 70
+2022-12-09 13:06:45,178 INFO [train.py:421] (4/8) Epoch 0, batch 17600, loss[loss=2.785, over 1050.00 frames. , ppl: 16.193694382260965] tot_loss[loss=2.654, over 5419344.74 frames. , ppl: 14.205369066197054], batch size: 70
+2022-12-09 13:08:28,801 INFO [train.py:421] (4/8) Epoch 0, batch 17800, loss[loss=2.663, over 1190.00 frames. , ppl: 14.345262281992436] tot_loss[loss=2.651, over 5427081.01 frames. , ppl: 14.163588022185444], batch size: 70
+2022-12-09 13:10:04,650 INFO [train.py:421] (4/8) Epoch 0, batch 18000, loss[loss=2.654, over 1680.00 frames. , ppl: 14.20783368259324] tot_loss[loss=2.648, over 5401426.98 frames. , ppl: 14.12530931336947], batch size: 70
+2022-12-09 13:10:04,651 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:10:05,409 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.595, over 211138.00 frames. , ppl: 13.393325029986839
+2022-12-09 13:11:44,553 INFO [train.py:421] (4/8) Epoch 0, batch 18200, loss[loss=2.766, over 1610.00 frames. , ppl: 15.89225426129287] tot_loss[loss=2.646, over 5362944.56 frames. , ppl: 14.104270071234616], batch size: 70
+2022-12-09 13:13:26,207 INFO [train.py:421] (4/8) Epoch 0, batch 18400, loss[loss=2.629, over 2240.00 frames. , ppl: 13.857963338185954] tot_loss[loss=2.644, over 5360152.29 frames. , ppl: 14.076339488061892], batch size: 70
+2022-12-09 13:15:03,885 INFO [train.py:421] (4/8) Epoch 0, batch 18600, loss[loss=2.78, over 1680.00 frames. , ppl: 16.124886560671435] tot_loss[loss=2.642, over 5413041.87 frames. , ppl: 14.035267962716944], batch size: 70
+2022-12-09 13:16:42,956 INFO [train.py:421] (4/8) Epoch 0, batch 18800, loss[loss=2.638, over 3710.00 frames. , ppl: 13.985667138772582] tot_loss[loss=2.638, over 5458044.24 frames. , ppl: 13.984365252112443], batch size: 70
+2022-12-09 13:18:25,094 INFO [train.py:421] (4/8) Epoch 0, batch 19000, loss[loss=2.76, over 770.00 frames. , ppl: 15.795158675149203] tot_loss[loss=2.635, over 5478789.71 frames. , ppl: 13.942975218433858], batch size: 70
+2022-12-09 13:18:25,094 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:18:25,840 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.582, over 211138.00 frames. , ppl: 13.223891354355008
+2022-12-09 13:20:07,272 INFO [train.py:421] (4/8) Epoch 0, batch 19200, loss[loss=2.56, over 5950.00 frames. , ppl: 12.939174422432208] tot_loss[loss=2.633, over 5495839.49 frames. , ppl: 13.91154239433458], batch size: 70
+2022-12-09 13:21:45,761 INFO [train.py:421] (4/8) Epoch 0, batch 19400, loss[loss=2.775, over 1610.00 frames. , ppl: 16.03257221072118] tot_loss[loss=2.631, over 5469186.42 frames. , ppl: 13.882878184728908], batch size: 70
+2022-12-09 13:23:23,703 INFO [train.py:421] (4/8) Epoch 0, batch 19600, loss[loss=2.714, over 1470.00 frames. , ppl: 15.089323351354517] tot_loss[loss=2.629, over 5422459.39 frames. , ppl: 13.864790315333016], batch size: 70
+2022-12-09 13:25:02,643 INFO [train.py:421] (4/8) Epoch 0, batch 19800, loss[loss=2.721, over 1610.00 frames. , ppl: 15.195234205502109] tot_loss[loss=2.626, over 5442047.35 frames. , ppl: 13.81548014080636], batch size: 70
+2022-12-09 13:26:43,273 INFO [train.py:421] (4/8) Epoch 0, batch 20000, loss[loss=2.546, over 6160.00 frames. , ppl: 12.75850849140761] tot_loss[loss=2.623, over 5430280.15 frames. , ppl: 13.781676198341728], batch size: 70
+2022-12-09 13:26:43,273 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:26:44,026 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.574, over 211138.00 frames. , ppl: 13.115298232215096
+2022-12-09 13:28:25,521 INFO [train.py:421] (4/8) Epoch 0, batch 20200, loss[loss=2.542, over 3850.00 frames. , ppl: 12.704221689313327] tot_loss[loss=2.621, over 5429912.43 frames. , ppl: 13.752260053831804], batch size: 70
+2022-12-09 13:30:02,466 INFO [train.py:421] (4/8) Epoch 0, batch 20400, loss[loss=2.564, over 9030.00 frames. , ppl: 12.992651277333396] tot_loss[loss=2.619, over 5426940.53 frames. , ppl: 13.719066112627354], batch size: 70
+2022-12-09 13:31:42,412 INFO [train.py:421] (4/8) Epoch 0, batch 20600, loss[loss=2.661, over 2940.00 frames. , ppl: 14.313772199236375] tot_loss[loss=2.617, over 5409150.38 frames. , ppl: 13.700531827512028], batch size: 70
+2022-12-09 13:33:22,041 INFO [train.py:421] (4/8) Epoch 0, batch 20800, loss[loss=2.554, over 5320.00 frames. , ppl: 12.853717976743708] tot_loss[loss=2.616, over 5378622.76 frames. , ppl: 13.676551370968676], batch size: 70
+2022-12-09 13:35:01,485 INFO [train.py:421] (4/8) Epoch 0, batch 21000, loss[loss=2.564, over 4550.00 frames. , ppl: 12.98786547075801] tot_loss[loss=2.614, over 5367470.03 frames. , ppl: 13.6498105993985], batch size: 70
+2022-12-09 13:35:01,485 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:35:02,241 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.567, over 211138.00 frames. , ppl: 13.024711781587353
+2022-12-09 13:36:41,726 INFO [train.py:421] (4/8) Epoch 0, batch 21200, loss[loss=2.776, over 980.00 frames. , ppl: 16.050486300255262] tot_loss[loss=2.613, over 5358605.68 frames. , ppl: 13.633239342402153], batch size: 70
+2022-12-09 13:38:26,908 INFO [train.py:421] (4/8) Epoch 0, batch 21400, loss[loss=2.531, over 5950.00 frames. , ppl: 12.560750475030046] tot_loss[loss=2.61, over 5378773.17 frames. , ppl: 13.600536121238115], batch size: 70
+2022-12-09 13:40:04,867 INFO [train.py:421] (4/8) Epoch 0, batch 21600, loss[loss=2.706, over 3290.00 frames. , ppl: 14.9634955022995] tot_loss[loss=2.608, over 5352206.87 frames. , ppl: 13.574074425465497], batch size: 70
+2022-12-09 13:41:41,313 INFO [train.py:421] (4/8) Epoch 0, batch 21800, loss[loss=2.807, over 2030.00 frames. , ppl: 16.55199914610614] tot_loss[loss=2.606, over 5366355.88 frames. , ppl: 13.543826400120215], batch size: 70
+2022-12-09 13:43:22,197 INFO [train.py:421] (4/8) Epoch 0, batch 22000, loss[loss=2.589, over 1680.00 frames. , ppl: 13.321903075726393] tot_loss[loss=2.603, over 5412527.75 frames. , ppl: 13.500769899728594], batch size: 70
+2022-12-09 13:43:22,198 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:43:22,962 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.558, over 211138.00 frames. , ppl: 12.907422572347382
+2022-12-09 13:45:08,023 INFO [train.py:421] (4/8) Epoch 0, batch 22200, loss[loss=2.556, over 7980.00 frames. , ppl: 12.878737103073561] tot_loss[loss=2.599, over 5444717.25 frames. , ppl: 13.453648303435592], batch size: 70
+2022-12-09 13:46:46,060 INFO [train.py:421] (4/8) Epoch 0, batch 22400, loss[loss=2.684, over 1400.00 frames. , ppl: 14.64617689675686] tot_loss[loss=2.598, over 5438521.40 frames. , ppl: 13.439012834072544], batch size: 70
+2022-12-09 13:48:21,209 INFO [train.py:421] (4/8) Epoch 0, batch 22600, loss[loss=2.702, over 1960.00 frames. , ppl: 14.910560415226424] tot_loss[loss=2.597, over 5427172.73 frames. , ppl: 13.427211494607242], batch size: 70
+2022-12-09 13:50:00,273 INFO [train.py:421] (4/8) Epoch 0, batch 22800, loss[loss=2.642, over 2520.00 frames. , ppl: 14.046803005130652] tot_loss[loss=2.597, over 5409884.43 frames. , ppl: 13.422611793853303], batch size: 70
+2022-12-09 13:51:39,299 INFO [train.py:421] (4/8) Epoch 0, batch 23000, loss[loss=2.651, over 2030.00 frames. , ppl: 14.166514854384083] tot_loss[loss=2.596, over 5395276.10 frames. , ppl: 13.40620881718681], batch size: 70
+2022-12-09 13:51:39,299 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 13:51:40,059 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.55, over 211138.00 frames. , ppl: 12.811174547900729
+2022-12-09 13:53:19,944 INFO [train.py:421] (4/8) Epoch 0, batch 23200, loss[loss=3.17, over 560.00 frames. , ppl: 23.79661769757921] tot_loss[loss=2.594, over 5386307.55 frames. , ppl: 13.380708806144622], batch size: 70
+2022-12-09 13:55:01,929 INFO [train.py:421] (4/8) Epoch 0, batch 23400, loss[loss=3.349, over 490.00 frames. , ppl: 28.462075721872516] tot_loss[loss=2.591, over 5426532.55 frames. , ppl: 13.34512363877208], batch size: 70
+2022-12-09 13:56:43,808 INFO [train.py:421] (4/8) Epoch 0, batch 23600, loss[loss=2.829, over 1120.00 frames. , ppl: 16.92582996127332] tot_loss[loss=2.59, over 5428076.13 frames. , ppl: 13.332971529412676], batch size: 70
+2022-12-09 13:58:26,280 INFO [train.py:421] (4/8) Epoch 0, batch 23800, loss[loss=2.679, over 980.00 frames. , ppl: 14.5635902202146] tot_loss[loss=2.589, over 5415006.30 frames. , ppl: 13.310577277575511], batch size: 70
+2022-12-09 14:00:07,995 INFO [train.py:421] (4/8) Epoch 0, batch 24000, loss[loss=2.506, over 4270.00 frames. , ppl: 12.255775349679483] tot_loss[loss=2.587, over 5403340.74 frames. , ppl: 13.294452245432538], batch size: 70
+2022-12-09 14:00:07,995 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:00:08,746 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.543, over 211138.00 frames. , ppl: 12.713002157513388
+2022-12-09 14:01:49,803 INFO [train.py:421] (4/8) Epoch 0, batch 24200, loss[loss=2.624, over 2520.00 frames. , ppl: 13.793494232978535] tot_loss[loss=2.586, over 5416336.69 frames. , ppl: 13.27746775592023], batch size: 70
+2022-12-09 14:03:30,659 INFO [train.py:421] (4/8) Epoch 0, batch 24400, loss[loss=2.67, over 980.00 frames. , ppl: 14.444460797617607] tot_loss[loss=2.584, over 5428049.59 frames. , ppl: 13.256126209979545], batch size: 70
+2022-12-09 14:05:11,531 INFO [train.py:421] (4/8) Epoch 0, batch 24600, loss[loss=2.639, over 2380.00 frames. , ppl: 14.006019221260187] tot_loss[loss=2.582, over 5438116.74 frames. , ppl: 13.229626486558477], batch size: 70
+2022-12-09 14:06:52,051 INFO [train.py:421] (4/8) Epoch 0, batch 24800, loss[loss=2.657, over 1120.00 frames. , ppl: 14.255298872075526] tot_loss[loss=2.581, over 5456106.62 frames. , ppl: 13.212646093211678], batch size: 70
+2022-12-09 14:08:35,024 INFO [train.py:421] (4/8) Epoch 0, batch 25000, loss[loss=2.718, over 1750.00 frames. , ppl: 15.156997895295575] tot_loss[loss=2.578, over 5512720.74 frames. , ppl: 13.172236510755546], batch size: 70
+2022-12-09 14:08:35,025 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:08:35,787 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.538, over 211138.00 frames. , ppl: 12.65462202047239
+2022-12-09 14:10:12,803 INFO [train.py:421] (4/8) Epoch 0, batch 25200, loss[loss=2.716, over 1330.00 frames. , ppl: 15.12076894787025] tot_loss[loss=2.577, over 5504005.24 frames. , ppl: 13.156433935998717], batch size: 70
+2022-12-09 14:11:51,083 INFO [train.py:421] (4/8) Epoch 0, batch 25400, loss[loss=2.557, over 5530.00 frames. , ppl: 12.903371416958132] tot_loss[loss=2.576, over 5504595.46 frames. , ppl: 13.143991469374816], batch size: 70
+2022-12-09 14:13:29,055 INFO [train.py:421] (4/8) Epoch 0, batch 25600, loss[loss=2.63, over 980.00 frames. , ppl: 13.873270824797052] tot_loss[loss=2.575, over 5500771.55 frames. , ppl: 13.124911602633222], batch size: 70
+2022-12-09 14:15:06,161 INFO [train.py:421] (4/8) Epoch 0, batch 25800, loss[loss=2.68, over 980.00 frames. , ppl: 14.590535815916642] tot_loss[loss=2.573, over 5492107.34 frames. , ppl: 13.108609503785749], batch size: 70
+2022-12-09 14:16:45,358 INFO [train.py:421] (4/8) Epoch 0, batch 26000, loss[loss=2.711, over 1050.00 frames. , ppl: 15.049151555912578] tot_loss[loss=2.572, over 5471744.95 frames. , ppl: 13.097550014481135], batch size: 70
+2022-12-09 14:16:45,359 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:16:46,124 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.534, over 211138.00 frames. , ppl: 12.606376393324691
+2022-12-09 14:18:23,663 INFO [train.py:421] (4/8) Epoch 0, batch 26200, loss[loss=2.481, over 7770.00 frames. , ppl: 11.958995679590505] tot_loss[loss=2.57, over 5463573.30 frames. , ppl: 13.071077640532698], batch size: 70
+2022-12-09 14:20:04,884 INFO [train.py:421] (4/8) Epoch 0, batch 26400, loss[loss=2.711, over 2030.00 frames. , ppl: 15.039515324090718] tot_loss[loss=2.569, over 5456020.30 frames. , ppl: 13.057656118342766], batch size: 70
+2022-12-09 14:21:42,869 INFO [train.py:421] (4/8) Epoch 0, batch 26600, loss[loss=2.535, over 1120.00 frames. , ppl: 12.62175573826831] tot_loss[loss=2.569, over 5419920.98 frames. , ppl: 13.05108918609833], batch size: 70
+2022-12-09 14:23:20,634 INFO [train.py:421] (4/8) Epoch 0, batch 26800, loss[loss=2.574, over 2450.00 frames. , ppl: 13.112148176636541] tot_loss[loss=2.568, over 5406020.66 frames. , ppl: 13.03991344173], batch size: 70
+2022-12-09 14:25:00,562 INFO [train.py:421] (4/8) Epoch 0, batch 27000, loss[loss=2.633, over 910.00 frames. , ppl: 13.921092620251438] tot_loss[loss=2.566, over 5427560.85 frames. , ppl: 13.017505455785656], batch size: 70
+2022-12-09 14:25:00,562 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:25:01,326 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.527, over 211138.00 frames. , ppl: 12.51153246345996
+2022-12-09 14:26:38,297 INFO [train.py:421] (4/8) Epoch 0, batch 27200, loss[loss=2.994, over 770.00 frames. , ppl: 19.973371966590758] tot_loss[loss=2.564, over 5423994.33 frames. , ppl: 12.990892070196766], batch size: 70
+2022-12-09 14:28:19,687 INFO [train.py:421] (4/8) Epoch 0, batch 27400, loss[loss=2.494, over 3430.00 frames. , ppl: 12.10578427178319] tot_loss[loss=2.563, over 5421204.75 frames. , ppl: 12.977483215113459], batch size: 70
+2022-12-09 14:30:01,585 INFO [train.py:421] (4/8) Epoch 0, batch 27600, loss[loss=2.576, over 1680.00 frames. , ppl: 13.14913336171025] tot_loss[loss=2.561, over 5468850.16 frames. , ppl: 12.947874841563753], batch size: 70
+2022-12-09 14:31:41,964 INFO [train.py:421] (4/8) Epoch 0, batch 27800, loss[loss=2.487, over 6440.00 frames. , ppl: 12.029467659782243] tot_loss[loss=2.561, over 5464309.63 frames. , ppl: 12.943829588178824], batch size: 70
+2022-12-09 14:33:19,871 INFO [train.py:421] (4/8) Epoch 0, batch 28000, loss[loss=2.462, over 3430.00 frames. , ppl: 11.73209248110337] tot_loss[loss=2.559, over 5473916.45 frames. , ppl: 12.928015354127108], batch size: 70
+2022-12-09 14:33:19,871 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:33:20,620 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.521, over 211138.00 frames. , ppl: 12.4409233226362
+2022-12-09 14:35:00,358 INFO [train.py:421] (4/8) Epoch 0, batch 28200, loss[loss=2.623, over 2100.00 frames. , ppl: 13.777696103339963] tot_loss[loss=2.558, over 5471639.38 frames. , ppl: 12.911876684492395], batch size: 70
+2022-12-09 14:36:41,343 INFO [train.py:421] (4/8) Epoch 0, batch 28400, loss[loss=2.727, over 1050.00 frames. , ppl: 15.284214924098201] tot_loss[loss=2.557, over 5474757.98 frames. , ppl: 12.897098572246689], batch size: 70
+2022-12-09 14:38:25,487 INFO [train.py:421] (4/8) Epoch 0, batch 28600, loss[loss=2.497, over 3080.00 frames. , ppl: 12.14879652738181] tot_loss[loss=2.555, over 5503591.02 frames. , ppl: 12.867232826115178], batch size: 70
+2022-12-09 14:40:02,969 INFO [train.py:421] (4/8) Epoch 0, batch 28800, loss[loss=2.445, over 6440.00 frames. , ppl: 11.533282469309922] tot_loss[loss=2.554, over 5500044.24 frames. , ppl: 12.85876410045246], batch size: 70
+2022-12-09 14:41:43,067 INFO [train.py:421] (4/8) Epoch 0, batch 29000, loss[loss=2.441, over 3570.00 frames. , ppl: 11.487891282259461] tot_loss[loss=2.553, over 5497241.65 frames. , ppl: 12.846122445331424], batch size: 70
+2022-12-09 14:41:43,067 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:41:43,812 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.513, over 211138.00 frames. , ppl: 12.34438956706025
+2022-12-09 14:43:21,607 INFO [train.py:421] (4/8) Epoch 0, batch 29200, loss[loss=2.536, over 4760.00 frames. , ppl: 12.631337380946006] tot_loss[loss=2.552, over 5450838.28 frames. , ppl: 12.837213530219612], batch size: 70
+2022-12-09 14:45:00,642 INFO [train.py:421] (4/8) Epoch 0, batch 29400, loss[loss=2.58, over 1540.00 frames. , ppl: 13.195123125508756] tot_loss[loss=2.551, over 5461686.27 frames. , ppl: 12.818876004189306], batch size: 70
+2022-12-09 14:46:41,805 INFO [train.py:421] (4/8) Epoch 0, batch 29600, loss[loss=2.556, over 1330.00 frames. , ppl: 12.879031828093504] tot_loss[loss=2.55, over 5497144.68 frames. , ppl: 12.806367927121713], batch size: 70
+2022-12-09 14:48:22,719 INFO [train.py:421] (4/8) Epoch 0, batch 29800, loss[loss=2.437, over 7700.00 frames. , ppl: 11.441549475636963] tot_loss[loss=2.549, over 5489166.43 frames. , ppl: 12.79296830216721], batch size: 70
+2022-12-09 14:50:00,643 INFO [train.py:421] (4/8) Epoch 0, batch 30000, loss[loss=2.629, over 1750.00 frames. , ppl: 13.860289788241486] tot_loss[loss=2.548, over 5487396.61 frames. , ppl: 12.777048479293644], batch size: 70
+2022-12-09 14:50:00,643 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:50:01,403 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.512, over 211138.00 frames. , ppl: 12.331413290958714
+2022-12-09 14:51:44,211 INFO [train.py:421] (4/8) Epoch 0, batch 30200, loss[loss=2.471, over 3360.00 frames. , ppl: 11.829122364997477] tot_loss[loss=2.546, over 5546663.18 frames. , ppl: 12.750707309166081], batch size: 70
+2022-12-09 14:53:26,920 INFO [train.py:421] (4/8) Epoch 0, batch 30400, loss[loss=2.762, over 840.00 frames. , ppl: 15.8320094891246] tot_loss[loss=2.545, over 5542719.94 frames. , ppl: 12.739150929393558], batch size: 70
+2022-12-09 14:55:12,236 INFO [train.py:421] (4/8) Epoch 0, batch 30600, loss[loss=2.689, over 1750.00 frames. , ppl: 14.717128179533608] tot_loss[loss=2.546, over 5482714.10 frames. , ppl: 12.755309199357024], batch size: 70
+2022-12-09 14:56:52,346 INFO [train.py:421] (4/8) Epoch 0, batch 30800, loss[loss=2.445, over 4760.00 frames. , ppl: 11.528499270633082] tot_loss[loss=2.544, over 5498837.04 frames. , ppl: 12.72992124040917], batch size: 70
+2022-12-09 14:58:33,196 INFO [train.py:421] (4/8) Epoch 0, batch 31000, loss[loss=2.924, over 770.00 frames. , ppl: 18.61823330852555] tot_loss[loss=2.543, over 5499762.45 frames. , ppl: 12.715820866732], batch size: 70
+2022-12-09 14:58:33,196 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 14:58:33,957 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.506, over 211138.00 frames. , ppl: 12.259548686709403
+2022-12-09 15:00:11,903 INFO [train.py:421] (4/8) Epoch 0, batch 31200, loss[loss=2.456, over 4200.00 frames. , ppl: 11.660119535916797] tot_loss[loss=2.541, over 5540785.40 frames. , ppl: 12.687330030823695], batch size: 70
+2022-12-09 15:01:49,406 INFO [train.py:421] (4/8) Epoch 0, batch 31400, loss[loss=2.552, over 2170.00 frames. , ppl: 12.829535028292577] tot_loss[loss=2.538, over 5607807.44 frames. , ppl: 12.652463858914142], batch size: 70
+2022-12-09 15:03:31,993 INFO [train.py:421] (4/8) Epoch 0, batch 31600, loss[loss=2.48, over 3640.00 frames. , ppl: 11.943113724467757] tot_loss[loss=2.537, over 5601195.85 frames. , ppl: 12.640169405866622], batch size: 70
+2022-12-09 15:05:12,059 INFO [train.py:421] (4/8) Epoch 0, batch 31800, loss[loss=2.497, over 6020.00 frames. , ppl: 12.148479057764568] tot_loss[loss=2.536, over 5585888.86 frames. , ppl: 12.630488100802893], batch size: 70
+2022-12-09 15:06:52,419 INFO [train.py:421] (4/8) Epoch 0, batch 32000, loss[loss=2.515, over 3710.00 frames. , ppl: 12.371063987671208] tot_loss[loss=2.534, over 5612394.08 frames. , ppl: 12.605087144857764], batch size: 70
+2022-12-09 15:06:52,420 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:06:53,178 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.502, over 211138.00 frames. , ppl: 12.211804295938897
+2022-12-09 15:08:31,772 INFO [train.py:421] (4/8) Epoch 0, batch 32200, loss[loss=3.6, over 490.00 frames. , ppl: 36.60111567606822] tot_loss[loss=2.534, over 5595658.60 frames. , ppl: 12.60134422489166], batch size: 70
+2022-12-09 15:10:13,715 INFO [train.py:421] (4/8) Epoch 0, batch 32400, loss[loss=2.539, over 4200.00 frames. , ppl: 12.664891358873849] tot_loss[loss=2.534, over 5560742.41 frames. , ppl: 12.602125128120145], batch size: 70
+2022-12-09 15:11:52,336 INFO [train.py:421] (4/8) Epoch 0, batch 32600, loss[loss=2.488, over 3150.00 frames. , ppl: 12.037275447293979] tot_loss[loss=2.534, over 5507842.74 frames. , ppl: 12.608931282428571], batch size: 70
+2022-12-09 15:13:32,200 INFO [train.py:421] (4/8) Epoch 0, batch 32800, loss[loss=2.51, over 3780.00 frames. , ppl: 12.30681788188467] tot_loss[loss=2.532, over 5533101.19 frames. , ppl: 12.581937785173372], batch size: 70
+2022-12-09 15:15:11,963 INFO [train.py:421] (4/8) Epoch 0, batch 33000, loss[loss=2.533, over 3780.00 frames. , ppl: 12.595902581337954] tot_loss[loss=2.531, over 5544991.12 frames. , ppl: 12.571771777268765], batch size: 70
+2022-12-09 15:15:11,964 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:15:12,723 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.498, over 211138.00 frames. , ppl: 12.152444554503326
+2022-12-09 15:16:47,646 INFO [train.py:421] (4/8) Epoch 0, batch 33200, loss[loss=2.534, over 1120.00 frames. , ppl: 12.60528056431434] tot_loss[loss=2.532, over 5504666.71 frames. , ppl: 12.57434766032979], batch size: 70
+2022-12-09 15:18:24,715 INFO [train.py:421] (4/8) Epoch 0, batch 33400, loss[loss=2.521, over 3710.00 frames. , ppl: 12.445958249424708] tot_loss[loss=2.532, over 5466723.59 frames. , ppl: 12.58113324261561], batch size: 70
+2022-12-09 15:20:01,825 INFO [train.py:421] (4/8) Epoch 0, batch 33600, loss[loss=2.498, over 2590.00 frames. , ppl: 12.156059968400392] tot_loss[loss=2.53, over 5510492.53 frames. , ppl: 12.554383525328635], batch size: 70
+2022-12-09 15:21:40,604 INFO [train.py:421] (4/8) Epoch 0, batch 33800, loss[loss=2.545, over 2100.00 frames. , ppl: 12.73999587378047] tot_loss[loss=2.53, over 5497022.06 frames. , ppl: 12.551449178386907], batch size: 70
+2022-12-09 15:23:22,962 INFO [train.py:421] (4/8) Epoch 0, batch 34000, loss[loss=2.522, over 3430.00 frames. , ppl: 12.458957832435582] tot_loss[loss=2.529, over 5524910.42 frames. , ppl: 12.54296439581802], batch size: 70
+2022-12-09 15:23:22,962 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:23:23,718 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.494, over 211138.00 frames. , ppl: 12.106278320487071
+2022-12-09 15:25:05,072 INFO [train.py:421] (4/8) Epoch 0, batch 34200, loss[loss=2.635, over 2100.00 frames. , ppl: 13.946966707076202] tot_loss[loss=2.528, over 5493404.37 frames. , ppl: 12.530743647996312], batch size: 70
+2022-12-09 15:26:48,280 INFO [train.py:421] (4/8) Epoch 0, batch 34400, loss[loss=2.41, over 8890.00 frames. , ppl: 11.134037464309696] tot_loss[loss=2.527, over 5484015.19 frames. , ppl: 12.514757041155791], batch size: 70
+2022-12-09 15:28:27,299 INFO [train.py:421] (4/8) Epoch 0, batch 34600, loss[loss=2.637, over 1190.00 frames. , ppl: 13.964673556155407] tot_loss[loss=2.526, over 5456388.50 frames. , ppl: 12.50569141186187], batch size: 70
+2022-12-09 15:30:05,467 INFO [train.py:421] (4/8) Epoch 0, batch 34800, loss[loss=2.599, over 2520.00 frames. , ppl: 13.45289254690142] tot_loss[loss=2.525, over 5474567.73 frames. , ppl: 12.493219090823631], batch size: 70
+2022-12-09 15:31:45,504 INFO [train.py:421] (4/8) Epoch 0, batch 35000, loss[loss=2.595, over 1610.00 frames. , ppl: 13.396300522273087] tot_loss[loss=2.524, over 5498380.47 frames. , ppl: 12.479355817646042], batch size: 70
+2022-12-09 15:31:45,504 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:31:46,264 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.491, over 211138.00 frames. , ppl: 12.072202954749265
+2022-12-09 15:33:26,023 INFO [train.py:421] (4/8) Epoch 0, batch 35200, loss[loss=2.55, over 3710.00 frames. , ppl: 12.802954578973765] tot_loss[loss=2.523, over 5470841.72 frames. , ppl: 12.466714422954624], batch size: 70
+2022-12-09 15:35:05,217 INFO [train.py:421] (4/8) Epoch 0, batch 35400, loss[loss=2.479, over 2940.00 frames. , ppl: 11.931087648481947] tot_loss[loss=2.523, over 5424682.43 frames. , ppl: 12.470834078858648], batch size: 70
+2022-12-09 15:36:46,623 INFO [train.py:421] (4/8) Epoch 0, batch 35600, loss[loss=2.739, over 1190.00 frames. , ppl: 15.47729859378148] tot_loss[loss=2.522, over 5443407.07 frames. , ppl: 12.458013681524326], batch size: 70
+2022-12-09 15:38:27,733 INFO [train.py:421] (4/8) Epoch 0, batch 35800, loss[loss=2.421, over 3710.00 frames. , ppl: 11.252448246895614] tot_loss[loss=2.521, over 5485581.76 frames. , ppl: 12.443359108789863], batch size: 70
+2022-12-09 15:40:08,922 INFO [train.py:421] (4/8) Epoch 0, batch 36000, loss[loss=2.418, over 4760.00 frames. , ppl: 11.222081539713672] tot_loss[loss=2.52, over 5467527.78 frames. , ppl: 12.43146920424355], batch size: 70
+2022-12-09 15:40:08,922 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:40:09,682 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.487, over 211138.00 frames. , ppl: 12.026263233378076
+2022-12-09 15:41:50,311 INFO [train.py:421] (4/8) Epoch 0, batch 36200, loss[loss=2.519, over 1820.00 frames. , ppl: 12.412537793693806] tot_loss[loss=2.518, over 5484969.81 frames. , ppl: 12.408816941716424], batch size: 70
+2022-12-09 15:43:29,538 INFO [train.py:421] (4/8) Epoch 0, batch 36400, loss[loss=2.785, over 700.00 frames. , ppl: 16.195804014939924] tot_loss[loss=2.517, over 5468190.51 frames. , ppl: 12.394658303657124], batch size: 70
+2022-12-09 15:45:09,064 INFO [train.py:421] (4/8) Epoch 0, batch 36600, loss[loss=2.442, over 5530.00 frames. , ppl: 11.493222279584487] tot_loss[loss=2.516, over 5479136.46 frames. , ppl: 12.375587865274904], batch size: 70
+2022-12-09 15:46:47,616 INFO [train.py:421] (4/8) Epoch 0, batch 36800, loss[loss=3.139, over 560.00 frames. , ppl: 23.075355154836178] tot_loss[loss=2.515, over 5456043.52 frames. , ppl: 12.369854798241965], batch size: 70
+2022-12-09 15:48:30,571 INFO [train.py:421] (4/8) Epoch 0, batch 37000, loss[loss=2.484, over 4270.00 frames. , ppl: 11.98363612707072] tot_loss[loss=2.514, over 5485248.87 frames. , ppl: 12.351290802789615], batch size: 70
+2022-12-09 15:48:30,571 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:48:31,316 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.483, over 211138.00 frames. , ppl: 11.975344738756288
+2022-12-09 15:50:06,979 INFO [train.py:421] (4/8) Epoch 0, batch 37200, loss[loss=2.552, over 2730.00 frames. , ppl: 12.830774278482778] tot_loss[loss=2.514, over 5488781.77 frames. , ppl: 12.358137816624806], batch size: 70
+2022-12-09 15:51:48,848 INFO [train.py:421] (4/8) Epoch 0, batch 37400, loss[loss=2.527, over 1400.00 frames. , ppl: 12.515317565926846] tot_loss[loss=2.512, over 5532641.11 frames. , ppl: 12.334832886705712], batch size: 70
+2022-12-09 15:53:29,771 INFO [train.py:421] (4/8) Epoch 0, batch 37600, loss[loss=2.635, over 1330.00 frames. , ppl: 13.943272023192558] tot_loss[loss=2.512, over 5531268.44 frames. , ppl: 12.323463276168328], batch size: 70
+2022-12-09 15:55:11,151 INFO [train.py:421] (4/8) Epoch 0, batch 37800, loss[loss=2.469, over 3010.00 frames. , ppl: 11.810775694703844] tot_loss[loss=2.511, over 5516332.95 frames. , ppl: 12.312550642522982], batch size: 70
+2022-12-09 15:56:53,633 INFO [train.py:421] (4/8) Epoch 0, batch 38000, loss[loss=2.428, over 2310.00 frames. , ppl: 11.337110459974815] tot_loss[loss=2.51, over 5530427.09 frames. , ppl: 12.301723739788757], batch size: 70
+2022-12-09 15:56:53,633 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 15:56:54,399 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.477, over 211138.00 frames. , ppl: 11.90581072793103
+2022-12-09 15:58:34,659 INFO [train.py:421] (4/8) Epoch 0, batch 38200, loss[loss=2.497, over 2730.00 frames. , ppl: 12.148789020986424] tot_loss[loss=2.509, over 5512500.46 frames. , ppl: 12.298639950622341], batch size: 70
+2022-12-09 16:00:16,310 INFO [train.py:421] (4/8) Epoch 0, batch 38400, loss[loss=2.504, over 1330.00 frames. , ppl: 12.233250964225727] tot_loss[loss=2.509, over 5503851.78 frames. , ppl: 12.29059035898598], batch size: 70
+2022-12-09 16:01:53,958 INFO [train.py:421] (4/8) Epoch 0, batch 38600, loss[loss=2.622, over 770.00 frames. , ppl: 13.768740664411666] tot_loss[loss=2.509, over 5499765.60 frames. , ppl: 12.289845078311137], batch size: 70
+2022-12-09 16:03:33,396 INFO [train.py:421] (4/8) Epoch 0, batch 38800, loss[loss=2.671, over 1540.00 frames. , ppl: 14.450437421303821] tot_loss[loss=2.508, over 5485781.77 frames. , ppl: 12.284581285969999], batch size: 70
+2022-12-09 16:05:12,805 INFO [train.py:421] (4/8) Epoch 0, batch 39000, loss[loss=3.273, over 490.00 frames. , ppl: 26.3949033950473] tot_loss[loss=2.508, over 5441526.41 frames. , ppl: 12.281954604431139], batch size: 70
+2022-12-09 16:05:12,806 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:05:13,565 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.476, over 211138.00 frames. , ppl: 11.897084242143393
+2022-12-09 16:06:53,444 INFO [train.py:421] (4/8) Epoch 0, batch 39200, loss[loss=2.473, over 1540.00 frames. , ppl: 11.857184841615517] tot_loss[loss=2.508, over 5413590.85 frames. , ppl: 12.277090707105028], batch size: 70
+2022-12-09 16:08:35,763 INFO [train.py:421] (4/8) Epoch 0, batch 39400, loss[loss=2.741, over 910.00 frames. , ppl: 15.502272871851094] tot_loss[loss=2.509, over 5319584.28 frames. , ppl: 12.29201452129229], batch size: 70
+2022-12-09 16:10:13,106 INFO [train.py:421] (4/8) Epoch 0, batch 39600, loss[loss=2.477, over 2800.00 frames. , ppl: 11.908482681616315] tot_loss[loss=2.508, over 5340165.05 frames. , ppl: 12.276879366812942], batch size: 70
+2022-12-09 16:11:57,756 INFO [train.py:421] (4/8) Epoch 0, batch 39800, loss[loss=5.111, over 280.00 frames. , ppl: 165.90036814880716] tot_loss[loss=2.505, over 5388892.90 frames. , ppl: 12.242433486793779], batch size: 70
+2022-12-09 16:13:37,684 INFO [train.py:421] (4/8) Epoch 0, batch 40000, loss[loss=2.455, over 4970.00 frames. , ppl: 11.65212762541124] tot_loss[loss=2.505, over 5355253.81 frames. , ppl: 12.246082841576088], batch size: 70
+2022-12-09 16:13:37,684 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:13:38,430 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.473, over 211138.00 frames. , ppl: 11.854175758684555
+2022-12-09 16:15:17,667 INFO [train.py:421] (4/8) Epoch 0, batch 40200, loss[loss=2.401, over 7980.00 frames. , ppl: 11.033905950580973] tot_loss[loss=2.503, over 5359707.12 frames. , ppl: 12.224527269104389], batch size: 70
+2022-12-09 16:16:59,417 INFO [train.py:421] (4/8) Epoch 0, batch 40400, loss[loss=2.525, over 2100.00 frames. , ppl: 12.489234103592747] tot_loss[loss=2.501, over 5416300.10 frames. , ppl: 12.195761493547622], batch size: 70
+2022-12-09 16:18:39,708 INFO [train.py:421] (4/8) Epoch 0, batch 40600, loss[loss=2.412, over 1890.00 frames. , ppl: 11.157673410871741] tot_loss[loss=2.5, over 5439759.17 frames. , ppl: 12.182817817976483], batch size: 70
+2022-12-09 16:20:18,283 INFO [train.py:421] (4/8) Epoch 0, batch 40800, loss[loss=2.694, over 1540.00 frames. , ppl: 14.786807749315795] tot_loss[loss=2.5, over 5438550.00 frames. , ppl: 12.184965797664951], batch size: 70
+2022-12-09 16:21:58,922 INFO [train.py:421] (4/8) Epoch 0, batch 41000, loss[loss=2.672, over 980.00 frames. , ppl: 14.471741176780016] tot_loss[loss=2.499, over 5456117.59 frames. , ppl: 12.164677386487423], batch size: 70
+2022-12-09 16:21:58,922 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:21:59,668 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.47, over 211138.00 frames. , ppl: 11.826051100421541
+2022-12-09 16:23:39,489 INFO [train.py:421] (4/8) Epoch 0, batch 41200, loss[loss=2.436, over 4200.00 frames. , ppl: 11.423250650784855] tot_loss[loss=2.499, over 5426082.06 frames. , ppl: 12.17007782246551], batch size: 70
+2022-12-09 16:25:18,888 INFO [train.py:421] (4/8) Epoch 0, batch 41400, loss[loss=2.523, over 2450.00 frames. , ppl: 12.462125899077058] tot_loss[loss=2.5, over 5387345.41 frames. , ppl: 12.18022217294721], batch size: 70
+2022-12-09 16:26:57,209 INFO [train.py:421] (4/8) Epoch 0, batch 41600, loss[loss=2.534, over 1960.00 frames. , ppl: 12.598418987754757] tot_loss[loss=2.5, over 5355891.44 frames. , ppl: 12.179001850928318], batch size: 70
+2022-12-09 16:28:34,642 INFO [train.py:421] (4/8) Epoch 0, batch 41800, loss[loss=2.693, over 1330.00 frames. , ppl: 14.778242321298475] tot_loss[loss=2.498, over 5382904.03 frames. , ppl: 12.160458092327023], batch size: 70
+2022-12-09 16:30:16,915 INFO [train.py:421] (4/8) Epoch 0, batch 42000, loss[loss=2.448, over 11970.00 frames. , ppl: 11.566278833541721] tot_loss[loss=2.498, over 5387200.80 frames. , ppl: 12.153582757109987], batch size: 70
+2022-12-09 16:30:16,916 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:30:17,662 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.468, over 211138.00 frames. , ppl: 11.797221378196006
+2022-12-09 16:31:58,193 INFO [train.py:421] (4/8) Epoch 0, batch 42200, loss[loss=2.598, over 1890.00 frames. , ppl: 13.439515716159582] tot_loss[loss=2.496, over 5414301.75 frames. , ppl: 12.13992458421147], batch size: 70
+2022-12-09 16:33:38,751 INFO [train.py:421] (4/8) Epoch 0, batch 42400, loss[loss=2.594, over 2100.00 frames. , ppl: 13.385116961401746] tot_loss[loss=2.496, over 5394109.31 frames. , ppl: 12.138859817435076], batch size: 70
+2022-12-09 16:35:21,173 INFO [train.py:421] (4/8) Epoch 0, batch 42600, loss[loss=2.874, over 630.00 frames. , ppl: 17.69989655589339] tot_loss[loss=2.495, over 5422837.17 frames. , ppl: 12.119161758793098], batch size: 70
+2022-12-09 16:37:02,072 INFO [train.py:421] (4/8) Epoch 0, batch 42800, loss[loss=2.518, over 910.00 frames. , ppl: 12.399394133031361] tot_loss[loss=2.494, over 5403074.37 frames. , ppl: 12.113383601015071], batch size: 70
+2022-12-09 16:38:43,436 INFO [train.py:421] (4/8) Epoch 0, batch 43000, loss[loss=2.487, over 1540.00 frames. , ppl: 12.021745858259024] tot_loss[loss=2.492, over 5419127.95 frames. , ppl: 12.090419486294918], batch size: 70
+2022-12-09 16:38:43,437 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:38:44,197 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.462, over 211138.00 frames. , ppl: 11.728416107172176
+2022-12-09 16:40:25,949 INFO [train.py:421] (4/8) Epoch 0, batch 43200, loss[loss=2.898, over 630.00 frames. , ppl: 18.143678560589667] tot_loss[loss=2.492, over 5466812.24 frames. , ppl: 12.079808223991108], batch size: 70
+2022-12-09 16:42:04,146 INFO [train.py:421] (4/8) Epoch 0, batch 43400, loss[loss=2.456, over 1120.00 frames. , ppl: 11.658875861585454] tot_loss[loss=2.491, over 5454211.12 frames. , ppl: 12.06957190568955], batch size: 70
+2022-12-09 16:43:44,124 INFO [train.py:421] (4/8) Epoch 0, batch 43600, loss[loss=2.503, over 2870.00 frames. , ppl: 12.218324500359145] tot_loss[loss=2.489, over 5488549.29 frames. , ppl: 12.053852903590894], batch size: 70
+2022-12-09 16:45:29,464 INFO [train.py:421] (4/8) Epoch 0, batch 43800, loss[loss=2.472, over 2730.00 frames. , ppl: 11.847497177439891] tot_loss[loss=2.487, over 5552572.09 frames. , ppl: 12.029805715852007], batch size: 70
+2022-12-09 16:47:12,055 INFO [train.py:421] (4/8) Epoch 0, batch 44000, loss[loss=3.143, over 630.00 frames. , ppl: 23.178378109488687] tot_loss[loss=2.488, over 5541664.21 frames. , ppl: 12.031163469448073], batch size: 70
+2022-12-09 16:47:12,055 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:47:12,815 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.46, over 211138.00 frames. , ppl: 11.703982858703393
+2022-12-09 16:48:52,629 INFO [train.py:421] (4/8) Epoch 0, batch 44200, loss[loss=2.432, over 3290.00 frames. , ppl: 11.377379381387565] tot_loss[loss=2.487, over 5535062.62 frames. , ppl: 12.024202638098444], batch size: 70
+2022-12-09 16:50:32,160 INFO [train.py:421] (4/8) Epoch 0, batch 44400, loss[loss=2.364, over 6230.00 frames. , ppl: 10.635188261688242] tot_loss[loss=2.487, over 5524608.36 frames. , ppl: 12.019227865365435], batch size: 70
+2022-12-09 16:52:13,144 INFO [train.py:421] (4/8) Epoch 0, batch 44600, loss[loss=2.471, over 2100.00 frames. , ppl: 11.830047238948287] tot_loss[loss=2.485, over 5527111.07 frames. , ppl: 12.00499038485004], batch size: 70
+2022-12-09 16:53:54,414 INFO [train.py:421] (4/8) Epoch 0, batch 44800, loss[loss=2.56, over 1960.00 frames. , ppl: 12.941207287830121] tot_loss[loss=2.485, over 5556877.20 frames. , ppl: 11.997368990725064], batch size: 70
+2022-12-09 16:55:31,012 INFO [train.py:421] (4/8) Epoch 0, batch 45000, loss[loss=2.431, over 1330.00 frames. , ppl: 11.375064690561763] tot_loss[loss=2.484, over 5578033.28 frames. , ppl: 11.984163563652], batch size: 70
+2022-12-09 16:55:31,012 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 16:55:31,771 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.459, over 211138.00 frames. , ppl: 11.693083254495477
+2022-12-09 16:57:12,294 INFO [train.py:421] (4/8) Epoch 0, batch 45200, loss[loss=2.661, over 1330.00 frames. , ppl: 14.316030024281034] tot_loss[loss=2.482, over 5614993.84 frames. , ppl: 11.966594774136391], batch size: 70
+2022-12-09 16:58:52,876 INFO [train.py:421] (4/8) Epoch 0, batch 45400, loss[loss=2.362, over 3780.00 frames. , ppl: 10.614966100681707] tot_loss[loss=2.481, over 5647258.95 frames. , ppl: 11.953186383760615], batch size: 70
+2022-12-09 17:00:33,543 INFO [train.py:421] (4/8) Epoch 0, batch 45600, loss[loss=2.739, over 1120.00 frames. , ppl: 15.470597726452713] tot_loss[loss=2.481, over 5626435.63 frames. , ppl: 11.955708833803287], batch size: 70
+2022-12-09 17:02:15,306 INFO [train.py:421] (4/8) Epoch 0, batch 45800, loss[loss=2.631, over 1050.00 frames. , ppl: 13.886240218337383] tot_loss[loss=2.48, over 5650540.03 frames. , ppl: 11.942300193688295], batch size: 70
+2022-12-09 17:03:53,497 INFO [train.py:421] (4/8) Epoch 0, batch 46000, loss[loss=2.514, over 2310.00 frames. , ppl: 12.348375860027224] tot_loss[loss=2.48, over 5621834.74 frames. , ppl: 11.93661340739009], batch size: 70
+2022-12-09 17:03:53,497 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:03:54,253 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.455, over 211138.00 frames. , ppl: 11.65209699143015
+2022-12-09 17:05:29,358 INFO [train.py:421] (4/8) Epoch 0, batch 46200, loss[loss=2.526, over 1540.00 frames. , ppl: 12.5035513004598] tot_loss[loss=2.48, over 5603019.68 frames. , ppl: 11.939894189717085], batch size: 70
+2022-12-09 17:07:09,588 INFO [train.py:421] (4/8) Epoch 0, batch 46400, loss[loss=2.41, over 4130.00 frames. , ppl: 11.131730953644519] tot_loss[loss=2.48, over 5602413.96 frames. , ppl: 11.938946854825247], batch size: 70
+2022-12-09 17:08:51,898 INFO [train.py:421] (4/8) Epoch 0, batch 46600, loss[loss=2.588, over 1120.00 frames. , ppl: 13.305691514250729] tot_loss[loss=2.48, over 5592527.35 frames. , ppl: 11.937985077610078], batch size: 70
+2022-12-09 17:10:31,042 INFO [train.py:421] (4/8) Epoch 0, batch 46800, loss[loss=2.474, over 2590.00 frames. , ppl: 11.875266602842363] tot_loss[loss=2.48, over 5588572.78 frames. , ppl: 11.935650502761394], batch size: 70
+2022-12-09 17:12:09,904 INFO [train.py:421] (4/8) Epoch 0, batch 47000, loss[loss=2.596, over 1890.00 frames. , ppl: 13.406876037406617] tot_loss[loss=2.48, over 5568329.00 frames. , ppl: 11.9353314528007], batch size: 70
+2022-12-09 17:12:09,904 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:12:10,666 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.453, over 211138.00 frames. , ppl: 11.621817980016578
+2022-12-09 17:13:49,886 INFO [train.py:421] (4/8) Epoch 0, batch 47200, loss[loss=2.701, over 840.00 frames. , ppl: 14.900444135745088] tot_loss[loss=2.48, over 5520912.48 frames. , ppl: 11.94408290154208], batch size: 70
+2022-12-09 17:15:28,650 INFO [train.py:421] (4/8) Epoch 0, batch 47400, loss[loss=2.485, over 1190.00 frames. , ppl: 12.00301035160684] tot_loss[loss=2.48, over 5529264.56 frames. , ppl: 11.936151563539344], batch size: 70
+2022-12-09 17:17:08,453 INFO [train.py:421] (4/8) Epoch 0, batch 47600, loss[loss=3.05, over 630.00 frames. , ppl: 21.112390667358905] tot_loss[loss=2.48, over 5504832.19 frames. , ppl: 11.936414898444704], batch size: 70
+2022-12-09 17:18:46,707 INFO [train.py:421] (4/8) Epoch 0, batch 47800, loss[loss=2.733, over 980.00 frames. , ppl: 15.371743569653646] tot_loss[loss=2.477, over 5540159.90 frames. , ppl: 11.909664518869034], batch size: 70
+2022-12-09 17:20:33,558 INFO [train.py:421] (4/8) Epoch 0, batch 48000, loss[loss=2.45, over 6580.00 frames. , ppl: 11.585222136684859] tot_loss[loss=2.478, over 5512301.30 frames. , ppl: 11.916192482164842], batch size: 70
+2022-12-09 17:20:33,559 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:20:34,309 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.453, over 211138.00 frames. , ppl: 11.621976231748523
+2022-12-09 17:22:13,233 INFO [train.py:421] (4/8) Epoch 0, batch 48200, loss[loss=2.53, over 4060.00 frames. , ppl: 12.556105615400211] tot_loss[loss=2.479, over 5487293.84 frames. , ppl: 11.924341546874853], batch size: 70
+2022-12-09 17:23:50,812 INFO [train.py:421] (4/8) Epoch 0, batch 48400, loss[loss=2.787, over 770.00 frames. , ppl: 16.237155264928592] tot_loss[loss=2.478, over 5446537.78 frames. , ppl: 11.919039000230304], batch size: 70
+2022-12-09 17:25:30,262 INFO [train.py:421] (4/8) Epoch 0, batch 48600, loss[loss=2.407, over 4480.00 frames. , ppl: 11.097387558370606] tot_loss[loss=2.477, over 5464087.21 frames. , ppl: 11.909358419949406], batch size: 70
+2022-12-09 17:27:09,468 INFO [train.py:421] (4/8) Epoch 0, batch 48800, loss[loss=2.532, over 1610.00 frames. , ppl: 12.58416545375911] tot_loss[loss=2.476, over 5466021.79 frames. , ppl: 11.893275123940498], batch size: 70
+2022-12-09 17:28:45,156 INFO [train.py:421] (4/8) Epoch 0, batch 49000, loss[loss=2.509, over 4690.00 frames. , ppl: 12.289537011912673] tot_loss[loss=2.476, over 5459443.44 frames. , ppl: 11.897327142869502], batch size: 70
+2022-12-09 17:28:45,156 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:28:45,901 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.449, over 211138.00 frames. , ppl: 11.574441333093757
+2022-12-09 17:30:31,205 INFO [train.py:421] (4/8) Epoch 0, batch 49200, loss[loss=2.466, over 2030.00 frames. , ppl: 11.778285223139758] tot_loss[loss=2.475, over 5522511.79 frames. , ppl: 11.8778603353507], batch size: 70
+2022-12-09 17:32:12,942 INFO [train.py:421] (4/8) Epoch 0, batch 49400, loss[loss=2.55, over 1960.00 frames. , ppl: 12.804363397540683] tot_loss[loss=2.474, over 5513700.84 frames. , ppl: 11.871038503609986], batch size: 70
+2022-12-09 17:33:55,712 INFO [train.py:421] (4/8) Epoch 0, batch 49600, loss[loss=2.656, over 1400.00 frames. , ppl: 14.240575499749523] tot_loss[loss=2.474, over 5490043.48 frames. , ppl: 11.869255681372843], batch size: 70
+2022-12-09 17:35:36,204 INFO [train.py:421] (4/8) Epoch 0, batch 49800, loss[loss=2.415, over 4200.00 frames. , ppl: 11.192065369997195] tot_loss[loss=2.473, over 5481513.86 frames. , ppl: 11.860014077282562], batch size: 70
+2022-12-09 17:37:18,308 INFO [train.py:421] (4/8) Epoch 0, batch 50000, loss[loss=2.351, over 7910.00 frames. , ppl: 10.492663615338786] tot_loss[loss=2.471, over 5537204.34 frames. , ppl: 11.83104400121607], batch size: 70
+2022-12-09 17:37:18,309 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:37:19,072 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.445, over 211138.00 frames. , ppl: 11.528131384960666
+2022-12-09 17:38:55,888 INFO [train.py:421] (4/8) Epoch 0, batch 50200, loss[loss=2.524, over 2590.00 frames. , ppl: 12.476675340480094] tot_loss[loss=2.471, over 5493350.47 frames. , ppl: 11.837216189959738], batch size: 70
+2022-12-09 17:40:41,944 INFO [train.py:421] (4/8) Epoch 0, batch 50400, loss[loss=2.488, over 1890.00 frames. , ppl: 12.039555789531347] tot_loss[loss=2.47, over 5516963.24 frames. , ppl: 11.824765830038132], batch size: 70
+2022-12-09 17:42:23,292 INFO [train.py:421] (4/8) Epoch 0, batch 50600, loss[loss=2.546, over 1680.00 frames. , ppl: 12.754585841320516] tot_loss[loss=2.472, over 5465356.27 frames. , ppl: 11.841190413017653], batch size: 70
+2022-12-09 17:44:03,412 INFO [train.py:421] (4/8) Epoch 0, batch 50800, loss[loss=2.378, over 4970.00 frames. , ppl: 10.778455493768162] tot_loss[loss=2.471, over 5440952.78 frames. , ppl: 11.838854327326725], batch size: 70
+2022-12-09 17:45:43,905 INFO [train.py:421] (4/8) Epoch 0, batch 51000, loss[loss=2.399, over 7070.00 frames. , ppl: 11.006978635158683] tot_loss[loss=2.471, over 5468177.27 frames. , ppl: 11.832913885145445], batch size: 70
+2022-12-09 17:45:43,906 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:45:44,651 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.442, over 211138.00 frames. , ppl: 11.496775701837747
+2022-12-09 17:47:23,507 INFO [train.py:421] (4/8) Epoch 0, batch 51200, loss[loss=4.173, over 350.00 frames. , ppl: 64.93840750804405] tot_loss[loss=2.47, over 5468389.39 frames. , ppl: 11.825136686870897], batch size: 70
+2022-12-09 17:49:00,218 INFO [train.py:421] (4/8) Epoch 0, batch 51400, loss[loss=2.362, over 2310.00 frames. , ppl: 10.616126687399309] tot_loss[loss=2.469, over 5481510.87 frames. , ppl: 11.81641117984742], batch size: 70
+2022-12-09 17:50:40,893 INFO [train.py:421] (4/8) Epoch 0, batch 51600, loss[loss=2.605, over 1260.00 frames. , ppl: 13.534647764589941] tot_loss[loss=2.469, over 5448585.92 frames. , ppl: 11.815255461208098], batch size: 70
+2022-12-09 17:52:21,992 INFO [train.py:421] (4/8) Epoch 0, batch 51800, loss[loss=2.461, over 4130.00 frames. , ppl: 11.720611201777292] tot_loss[loss=2.469, over 5456388.97 frames. , ppl: 11.811504861169066], batch size: 70
+2022-12-09 17:54:03,060 INFO [train.py:421] (4/8) Epoch 0, batch 52000, loss[loss=2.514, over 910.00 frames. , ppl: 12.351606863396844] tot_loss[loss=2.47, over 5442079.68 frames. , ppl: 11.81917479406886], batch size: 70
+2022-12-09 17:54:03,061 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 17:54:03,822 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.441, over 211138.00 frames. , ppl: 11.488961168872661
+2022-12-09 17:55:46,396 INFO [train.py:421] (4/8) Epoch 0, batch 52200, loss[loss=2.407, over 6370.00 frames. , ppl: 11.103373102365657] tot_loss[loss=2.469, over 5410744.19 frames. , ppl: 11.816359133792064], batch size: 70
+2022-12-09 17:57:27,052 INFO [train.py:421] (4/8) Epoch 0, batch 52400, loss[loss=2.419, over 2940.00 frames. , ppl: 11.229145429393512] tot_loss[loss=2.468, over 5452272.33 frames. , ppl: 11.803304774832318], batch size: 70
+2022-12-09 17:59:04,826 INFO [train.py:421] (4/8) Epoch 0, batch 52600, loss[loss=2.401, over 2310.00 frames. , ppl: 11.037403505383944] tot_loss[loss=2.468, over 5432866.01 frames. , ppl: 11.804543865400714], batch size: 70
+2022-12-09 18:00:44,978 INFO [train.py:421] (4/8) Epoch 0, batch 52800, loss[loss=2.482, over 2800.00 frames. , ppl: 11.967543916705958] tot_loss[loss=2.467, over 5456977.80 frames. , ppl: 11.788014030898637], batch size: 70
+2022-12-09 18:02:25,843 INFO [train.py:421] (4/8) Epoch 0, batch 53000, loss[loss=2.498, over 2380.00 frames. , ppl: 12.155378602766419] tot_loss[loss=2.466, over 5447904.39 frames. , ppl: 11.77906508051687], batch size: 70
+2022-12-09 18:02:25,843 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:02:26,589 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.439, over 211138.00 frames. , ppl: 11.45966239282214
+2022-12-09 18:04:04,857 INFO [train.py:421] (4/8) Epoch 0, batch 53200, loss[loss=2.869, over 630.00 frames. , ppl: 17.61236461766046] tot_loss[loss=2.465, over 5462117.73 frames. , ppl: 11.763668425859336], batch size: 70
+2022-12-09 18:05:43,609 INFO [train.py:421] (4/8) Epoch 0, batch 53400, loss[loss=2.601, over 1610.00 frames. , ppl: 13.476075208415475] tot_loss[loss=2.465, over 5437201.97 frames. , ppl: 11.76588835384707], batch size: 70
+2022-12-09 18:07:22,503 INFO [train.py:421] (4/8) Epoch 0, batch 53600, loss[loss=2.931, over 770.00 frames. , ppl: 18.748153581528932] tot_loss[loss=2.463, over 5475380.33 frames. , ppl: 11.73861193340433], batch size: 70
+2022-12-09 18:09:01,090 INFO [train.py:421] (4/8) Epoch 0, batch 53800, loss[loss=2.424, over 3990.00 frames. , ppl: 11.293425113545576] tot_loss[loss=2.462, over 5480716.68 frames. , ppl: 11.729589373581877], batch size: 70
+2022-12-09 18:10:37,935 INFO [train.py:421] (4/8) Epoch 0, batch 54000, loss[loss=2.596, over 1190.00 frames. , ppl: 13.406854542328517] tot_loss[loss=2.463, over 5454349.64 frames. , ppl: 11.741760462998872], batch size: 70
+2022-12-09 18:10:37,936 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:10:38,697 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.437, over 211138.00 frames. , ppl: 11.437813533531314
+2022-12-09 18:12:17,929 INFO [train.py:421] (4/8) Epoch 0, batch 54200, loss[loss=2.624, over 1890.00 frames. , ppl: 13.797532323161352] tot_loss[loss=2.463, over 5431848.39 frames. , ppl: 11.736738665179514], batch size: 70
+2022-12-09 18:13:57,230 INFO [train.py:421] (4/8) Epoch 0, batch 54400, loss[loss=2.453, over 1890.00 frames. , ppl: 11.62789391834522] tot_loss[loss=2.462, over 5457590.39 frames. , ppl: 11.722382952440183], batch size: 70
+2022-12-09 18:15:38,992 INFO [train.py:421] (4/8) Epoch 0, batch 54600, loss[loss=2.601, over 980.00 frames. , ppl: 13.472481730474044] tot_loss[loss=2.461, over 5467438.38 frames. , ppl: 11.718760012525026], batch size: 70
+2022-12-09 18:17:17,700 INFO [train.py:421] (4/8) Epoch 0, batch 54800, loss[loss=2.387, over 4900.00 frames. , ppl: 10.884105257367553] tot_loss[loss=2.461, over 5459012.34 frames. , ppl: 11.71813711642593], batch size: 70
+2022-12-09 18:18:54,917 INFO [train.py:421] (4/8) Epoch 0, batch 55000, loss[loss=2.491, over 1820.00 frames. , ppl: 12.076412037323296] tot_loss[loss=2.461, over 5469074.11 frames. , ppl: 11.712246647205971], batch size: 70
+2022-12-09 18:18:54,917 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:18:55,677 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.435, over 211138.00 frames. , ppl: 11.413029548583495
+2022-12-09 18:20:34,862 INFO [train.py:421] (4/8) Epoch 0, batch 55200, loss[loss=2.534, over 1330.00 frames. , ppl: 12.602686826535745] tot_loss[loss=2.462, over 5440290.62 frames. , ppl: 11.724649165482173], batch size: 70
+2022-12-09 18:22:16,307 INFO [train.py:421] (4/8) Epoch 0, batch 55400, loss[loss=2.422, over 2520.00 frames. , ppl: 11.269218280699071] tot_loss[loss=2.459, over 5517345.58 frames. , ppl: 11.698611629319688], batch size: 70
+2022-12-09 18:23:55,792 INFO [train.py:421] (4/8) Epoch 0, batch 55600, loss[loss=2.422, over 3640.00 frames. , ppl: 11.270831872277437] tot_loss[loss=2.46, over 5473146.87 frames. , ppl: 11.70408812393754], batch size: 70
+2022-12-09 18:25:34,335 INFO [train.py:421] (4/8) Epoch 0, batch 55800, loss[loss=2.417, over 4550.00 frames. , ppl: 11.215451812866656] tot_loss[loss=2.461, over 5451844.02 frames. , ppl: 11.713151824441736], batch size: 70
+2022-12-09 18:27:11,405 INFO [train.py:421] (4/8) Epoch 0, batch 56000, loss[loss=2.403, over 5880.00 frames. , ppl: 11.059288398941185] tot_loss[loss=2.461, over 5423814.77 frames. , ppl: 11.711953672375088], batch size: 70
+2022-12-09 18:27:11,406 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:27:12,150 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.435, over 211138.00 frames. , ppl: 11.411955259320733
+2022-12-09 18:28:51,297 INFO [train.py:421] (4/8) Epoch 0, batch 56200, loss[loss=2.483, over 3990.00 frames. , ppl: 11.974667890045588] tot_loss[loss=2.459, over 5457404.65 frames. , ppl: 11.695656757507482], batch size: 70
+2022-12-09 18:30:31,855 INFO [train.py:421] (4/8) Epoch 0, batch 56400, loss[loss=2.637, over 1120.00 frames. , ppl: 13.969015414770526] tot_loss[loss=2.459, over 5453954.95 frames. , ppl: 11.689562692578038], batch size: 70
+2022-12-09 18:32:12,542 INFO [train.py:421] (4/8) Epoch 0, batch 56600, loss[loss=2.496, over 2100.00 frames. , ppl: 12.13495377281688] tot_loss[loss=2.458, over 5457006.89 frames. , ppl: 11.67558957626152], batch size: 70
+2022-12-09 18:33:53,508 INFO [train.py:421] (4/8) Epoch 0, batch 56800, loss[loss=2.447, over 4620.00 frames. , ppl: 11.549203640633428] tot_loss[loss=2.457, over 5482387.65 frames. , ppl: 11.66562222916427], batch size: 70
+2022-12-09 18:35:37,613 INFO [train.py:421] (4/8) Epoch 0, batch 57000, loss[loss=2.381, over 6230.00 frames. , ppl: 10.812673577740583] tot_loss[loss=2.456, over 5486132.01 frames. , ppl: 11.658853058224446], batch size: 70
+2022-12-09 18:35:37,613 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:35:38,358 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.428, over 211138.00 frames. , ppl: 11.336884940977864
+2022-12-09 18:37:20,806 INFO [train.py:421] (4/8) Epoch 0, batch 57200, loss[loss=2.679, over 770.00 frames. , ppl: 14.563708959385728] tot_loss[loss=2.456, over 5461708.63 frames. , ppl: 11.661452014612884], batch size: 70
+2022-12-09 18:39:00,405 INFO [train.py:421] (4/8) Epoch 0, batch 57400, loss[loss=2.442, over 4690.00 frames. , ppl: 11.501618407216831] tot_loss[loss=2.456, over 5409897.31 frames. , ppl: 11.656379174380952], batch size: 70
+2022-12-09 18:40:42,488 INFO [train.py:421] (4/8) Epoch 0, batch 57600, loss[loss=2.407, over 2940.00 frames. , ppl: 11.101310204068433] tot_loss[loss=2.456, over 5389206.91 frames. , ppl: 11.657856213932611], batch size: 70
+2022-12-09 18:42:20,984 INFO [train.py:421] (4/8) Epoch 0, batch 57800, loss[loss=2.545, over 1820.00 frames. , ppl: 12.738107017033085] tot_loss[loss=2.455, over 5456038.51 frames. , ppl: 11.645712625892912], batch size: 70
+2022-12-09 18:44:01,540 INFO [train.py:421] (4/8) Epoch 0, batch 58000, loss[loss=3.147, over 490.00 frames. , ppl: 23.25710508505822] tot_loss[loss=2.455, over 5457707.15 frames. , ppl: 11.647102099462664], batch size: 70
+2022-12-09 18:44:01,541 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:44:02,301 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.427, over 211138.00 frames. , ppl: 11.325093455985686
+2022-12-09 18:45:45,018 INFO [train.py:421] (4/8) Epoch 0, batch 58200, loss[loss=2.906, over 630.00 frames. , ppl: 18.276190590335478] tot_loss[loss=2.455, over 5458348.03 frames. , ppl: 11.640929534462188], batch size: 70
+2022-12-09 18:47:28,520 INFO [train.py:421] (4/8) Epoch 0, batch 58400, loss[loss=2.997, over 560.00 frames. , ppl: 20.024735744665833] tot_loss[loss=2.454, over 5475739.89 frames. , ppl: 11.634024541787664], batch size: 70
+2022-12-09 18:49:11,522 INFO [train.py:421] (4/8) Epoch 0, batch 58600, loss[loss=2.419, over 4970.00 frames. , ppl: 11.234026774071806] tot_loss[loss=2.452, over 5535399.33 frames. , ppl: 11.61180084299054], batch size: 70
+2022-12-09 18:50:48,038 INFO [train.py:421] (4/8) Epoch 0, batch 58800, loss[loss=2.515, over 1400.00 frames. , ppl: 12.371330401968097] tot_loss[loss=2.453, over 5501219.47 frames. , ppl: 11.618485323163608], batch size: 70
+2022-12-09 18:52:28,262 INFO [train.py:421] (4/8) Epoch 0, batch 59000, loss[loss=2.544, over 1610.00 frames. , ppl: 12.730771841279084] tot_loss[loss=2.452, over 5490480.88 frames. , ppl: 11.611726074253983], batch size: 70
+2022-12-09 18:52:28,262 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 18:52:29,026 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.429, over 211138.00 frames. , ppl: 11.349824230730555
+2022-12-09 18:54:09,183 INFO [train.py:421] (4/8) Epoch 0, batch 59200, loss[loss=2.47, over 2450.00 frames. , ppl: 11.818368980572572] tot_loss[loss=2.452, over 5519108.94 frames. , ppl: 11.610344567859899], batch size: 70
+2022-12-09 18:55:49,453 INFO [train.py:421] (4/8) Epoch 0, batch 59400, loss[loss=2.354, over 5600.00 frames. , ppl: 10.524079800558308] tot_loss[loss=2.45, over 5552553.94 frames. , ppl: 11.586463630496691], batch size: 70
+2022-12-09 18:57:31,357 INFO [train.py:421] (4/8) Epoch 0, batch 59600, loss[loss=2.517, over 1820.00 frames. , ppl: 12.389162443007745] tot_loss[loss=2.449, over 5573224.64 frames. , ppl: 11.57742816989945], batch size: 70
+2022-12-09 18:59:09,089 INFO [train.py:421] (4/8) Epoch 0, batch 59800, loss[loss=2.435, over 3640.00 frames. , ppl: 11.419848111631378] tot_loss[loss=2.449, over 5570939.56 frames. , ppl: 11.570991195384261], batch size: 70
+2022-12-09 19:00:50,610 INFO [train.py:421] (4/8) Epoch 0, batch 60000, loss[loss=2.429, over 3570.00 frames. , ppl: 11.35202845065292] tot_loss[loss=2.449, over 5573683.19 frames. , ppl: 11.573621539301678], batch size: 70
+2022-12-09 19:00:50,611 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:00:51,357 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.426, over 211138.00 frames. , ppl: 11.316449249268606
+2022-12-09 19:02:33,547 INFO [train.py:421] (4/8) Epoch 0, batch 60200, loss[loss=2.511, over 2310.00 frames. , ppl: 12.315091605841504] tot_loss[loss=2.448, over 5593775.73 frames. , ppl: 11.569610522696028], batch size: 70
+2022-12-09 19:04:11,496 INFO [train.py:421] (4/8) Epoch 0, batch 60400, loss[loss=2.735, over 910.00 frames. , ppl: 15.415046077702904] tot_loss[loss=2.449, over 5561117.30 frames. , ppl: 11.5734912183011], batch size: 70
+2022-12-09 19:05:47,668 INFO [train.py:421] (4/8) Epoch 0, batch 60600, loss[loss=2.476, over 1960.00 frames. , ppl: 11.891917603795111] tot_loss[loss=2.449, over 5519918.57 frames. , ppl: 11.580463340809663], batch size: 70
+2022-12-09 19:07:33,507 INFO [train.py:421] (4/8) Epoch 0, batch 60800, loss[loss=2.394, over 3640.00 frames. , ppl: 10.955954189270242] tot_loss[loss=2.448, over 5548106.93 frames. , ppl: 11.569656831049906], batch size: 70
+2022-12-09 19:09:14,068 INFO [train.py:421] (4/8) Epoch 0, batch 61000, loss[loss=2.498, over 2380.00 frames. , ppl: 12.161347730768668] tot_loss[loss=2.447, over 5558741.81 frames. , ppl: 11.554777833898232], batch size: 70
+2022-12-09 19:09:14,069 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:09:14,814 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.424, over 211138.00 frames. , ppl: 11.292349414932128
+2022-12-09 19:10:56,138 INFO [train.py:421] (4/8) Epoch 0, batch 61200, loss[loss=2.327, over 5880.00 frames. , ppl: 10.249386599066382] tot_loss[loss=2.447, over 5572404.84 frames. , ppl: 11.549403944983727], batch size: 70
+2022-12-09 19:12:36,985 INFO [train.py:421] (4/8) Epoch 0, batch 61400, loss[loss=2.391, over 4340.00 frames. , ppl: 10.925852089810176] tot_loss[loss=2.446, over 5567274.86 frames. , ppl: 11.547712436768148], batch size: 70
+2022-12-09 19:14:17,498 INFO [train.py:421] (4/8) Epoch 0, batch 61600, loss[loss=2.822, over 700.00 frames. , ppl: 16.813387891449405] tot_loss[loss=2.446, over 5584988.33 frames. , ppl: 11.542221361070679], batch size: 70
+2022-12-09 19:16:00,806 INFO [train.py:421] (4/8) Epoch 0, batch 61800, loss[loss=2.937, over 700.00 frames. , ppl: 18.854623480005987] tot_loss[loss=2.446, over 5598377.38 frames. , ppl: 11.538884519307379], batch size: 70
+2022-12-09 19:17:37,907 INFO [train.py:421] (4/8) Epoch 0, batch 62000, loss[loss=2.406, over 6930.00 frames. , ppl: 11.088654919482442] tot_loss[loss=2.446, over 5600728.16 frames. , ppl: 11.538037280878985], batch size: 70
+2022-12-09 19:17:37,908 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:17:38,663 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.422, over 211138.00 frames. , ppl: 11.26730360842701
+2022-12-09 19:19:16,403 INFO [train.py:421] (4/8) Epoch 0, batch 62200, loss[loss=3.33, over 560.00 frames. , ppl: 27.92768245153858] tot_loss[loss=2.445, over 5606895.44 frames. , ppl: 11.526976239341069], batch size: 70
+2022-12-09 19:20:55,497 INFO [train.py:421] (4/8) Epoch 0, batch 62400, loss[loss=2.356, over 4480.00 frames. , ppl: 10.547076303108389] tot_loss[loss=2.444, over 5625154.92 frames. , ppl: 11.518362901800746], batch size: 70
+2022-12-09 19:22:38,366 INFO [train.py:421] (4/8) Epoch 0, batch 62600, loss[loss=2.427, over 2450.00 frames. , ppl: 11.329504217399233] tot_loss[loss=2.444, over 5601604.71 frames. , ppl: 11.518367913073492], batch size: 70
+2022-12-09 19:24:18,378 INFO [train.py:421] (4/8) Epoch 0, batch 62800, loss[loss=2.373, over 4340.00 frames. , ppl: 10.730446064015101] tot_loss[loss=2.443, over 5637946.11 frames. , ppl: 11.510396604000723], batch size: 70
+2022-12-09 19:25:57,788 INFO [train.py:421] (4/8) Epoch 0, batch 63000, loss[loss=2.404, over 7070.00 frames. , ppl: 11.066292359020665] tot_loss[loss=2.443, over 5651473.78 frames. , ppl: 11.50702530657366], batch size: 70
+2022-12-09 19:25:57,788 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:25:58,533 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.419, over 211138.00 frames. , ppl: 11.238974307884806
+2022-12-09 19:27:42,798 INFO [train.py:421] (4/8) Epoch 0, batch 63200, loss[loss=2.463, over 2030.00 frames. , ppl: 11.744473091906386] tot_loss[loss=2.442, over 5621023.45 frames. , ppl: 11.496994678558629], batch size: 70
+2022-12-09 19:29:19,322 INFO [train.py:421] (4/8) Epoch 0, batch 63400, loss[loss=2.613, over 1330.00 frames. , ppl: 13.635508688978037] tot_loss[loss=2.442, over 5581131.66 frames. , ppl: 11.500889763052594], batch size: 70
+2022-12-09 19:30:58,088 INFO [train.py:421] (4/8) Epoch 0, batch 63600, loss[loss=2.328, over 3290.00 frames. , ppl: 10.255105037367343] tot_loss[loss=2.444, over 5545221.19 frames. , ppl: 11.5135427713643], batch size: 70
+2022-12-09 19:32:37,344 INFO [train.py:421] (4/8) Epoch 0, batch 63800, loss[loss=2.318, over 5670.00 frames. , ppl: 10.150395562129978] tot_loss[loss=2.444, over 5503730.93 frames. , ppl: 11.514560420730607], batch size: 70
+2022-12-09 19:34:18,413 INFO [train.py:421] (4/8) Epoch 0, batch 64000, loss[loss=2.333, over 5530.00 frames. , ppl: 10.310821163421982] tot_loss[loss=2.442, over 5526618.97 frames. , ppl: 11.500339567953851], batch size: 70
+2022-12-09 19:34:18,414 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:34:19,175 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.418, over 211138.00 frames. , ppl: 11.224827241902219
+2022-12-09 19:35:57,968 INFO [train.py:421] (4/8) Epoch 0, batch 64200, loss[loss=2.466, over 2800.00 frames. , ppl: 11.7760741549945] tot_loss[loss=2.443, over 5547163.26 frames. , ppl: 11.505245048116805], batch size: 70
+2022-12-09 19:37:37,733 INFO [train.py:421] (4/8) Epoch 0, batch 64400, loss[loss=2.626, over 1190.00 frames. , ppl: 13.822774234268413] tot_loss[loss=2.442, over 5567642.72 frames. , ppl: 11.493875424991012], batch size: 70
+2022-12-09 19:39:14,489 INFO [train.py:421] (4/8) Epoch 0, batch 64600, loss[loss=2.501, over 1190.00 frames. , ppl: 12.193796320063436] tot_loss[loss=2.444, over 5506796.20 frames. , ppl: 11.514379913955858], batch size: 70
+2022-12-09 19:40:56,610 INFO [train.py:421] (4/8) Epoch 0, batch 64800, loss[loss=2.813, over 910.00 frames. , ppl: 16.656700262685458] tot_loss[loss=2.443, over 5504433.88 frames. , ppl: 11.508672536012535], batch size: 70
+2022-12-09 19:42:37,261 INFO [train.py:421] (4/8) Epoch 0, batch 65000, loss[loss=2.503, over 2590.00 frames. , ppl: 12.222038477762762] tot_loss[loss=2.442, over 5551090.76 frames. , ppl: 11.497669340988473], batch size: 70
+2022-12-09 19:42:37,262 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:42:38,022 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.417, over 211138.00 frames. , ppl: 11.211114467296362
+2022-12-09 19:44:16,899 INFO [train.py:421] (4/8) Epoch 0, batch 65200, loss[loss=2.434, over 2380.00 frames. , ppl: 11.403186952354309] tot_loss[loss=2.442, over 5521254.72 frames. , ppl: 11.492037404597182], batch size: 70
+2022-12-09 19:45:56,839 INFO [train.py:421] (4/8) Epoch 0, batch 65400, loss[loss=2.709, over 910.00 frames. , ppl: 15.019546667288278] tot_loss[loss=2.441, over 5510343.56 frames. , ppl: 11.484463529218354], batch size: 70
+2022-12-09 19:47:35,890 INFO [train.py:421] (4/8) Epoch 0, batch 65600, loss[loss=2.41, over 1610.00 frames. , ppl: 11.135677321769505] tot_loss[loss=2.44, over 5530256.70 frames. , ppl: 11.471951611089029], batch size: 70
+2022-12-09 19:49:13,440 INFO [train.py:421] (4/8) Epoch 0, batch 65800, loss[loss=2.352, over 3360.00 frames. , ppl: 10.50742521693642] tot_loss[loss=2.439, over 5553544.58 frames. , ppl: 11.459094134667923], batch size: 70
+2022-12-09 19:50:53,677 INFO [train.py:421] (4/8) Epoch 0, batch 66000, loss[loss=2.41, over 1750.00 frames. , ppl: 11.130538226801564] tot_loss[loss=2.438, over 5561898.07 frames. , ppl: 11.452459106130389], batch size: 70
+2022-12-09 19:50:53,678 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:50:54,438 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.415, over 211138.00 frames. , ppl: 11.192550351531178
+2022-12-09 19:52:35,267 INFO [train.py:421] (4/8) Epoch 0, batch 66200, loss[loss=2.379, over 3780.00 frames. , ppl: 10.795316829562069] tot_loss[loss=2.438, over 5568056.71 frames. , ppl: 11.455616848181842], batch size: 70
+2022-12-09 19:54:17,413 INFO [train.py:421] (4/8) Epoch 0, batch 66400, loss[loss=2.383, over 6790.00 frames. , ppl: 10.837650423581964] tot_loss[loss=2.438, over 5534218.28 frames. , ppl: 11.455098254139562], batch size: 70
+2022-12-09 19:56:01,926 INFO [train.py:421] (4/8) Epoch 0, batch 66600, loss[loss=2.493, over 1820.00 frames. , ppl: 12.098603291109875] tot_loss[loss=2.438, over 5567360.29 frames. , ppl: 11.448569541295539], batch size: 70
+2022-12-09 19:57:44,179 INFO [train.py:421] (4/8) Epoch 0, batch 66800, loss[loss=2.468, over 2240.00 frames. , ppl: 11.79592129857603] tot_loss[loss=2.437, over 5626716.28 frames. , ppl: 11.435261117262995], batch size: 70
+2022-12-09 19:59:25,913 INFO [train.py:421] (4/8) Epoch 0, batch 67000, loss[loss=2.524, over 2310.00 frames. , ppl: 12.483525748691772] tot_loss[loss=2.436, over 5618894.39 frames. , ppl: 11.426587062511889], batch size: 70
+2022-12-09 19:59:25,913 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 19:59:26,673 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.412, over 211138.00 frames. , ppl: 11.1568600153867
+2022-12-09 20:01:08,429 INFO [train.py:421] (4/8) Epoch 0, batch 67200, loss[loss=2.544, over 1330.00 frames. , ppl: 12.735785513596513] tot_loss[loss=2.436, over 5605464.95 frames. , ppl: 11.422805412848733], batch size: 70
+2022-12-09 20:02:47,732 INFO [train.py:421] (4/8) Epoch 0, batch 67400, loss[loss=2.37, over 3850.00 frames. , ppl: 10.69617945353114] tot_loss[loss=2.435, over 5567178.27 frames. , ppl: 11.421310436193451], batch size: 70
+2022-12-09 20:04:25,980 INFO [train.py:421] (4/8) Epoch 0, batch 67600, loss[loss=2.444, over 1750.00 frames. , ppl: 11.51901838456809] tot_loss[loss=2.437, over 5538721.05 frames. , ppl: 11.436187830447777], batch size: 70
+2022-12-09 20:06:07,214 INFO [train.py:421] (4/8) Epoch 0, batch 67800, loss[loss=2.385, over 5530.00 frames. , ppl: 10.857647154779967] tot_loss[loss=2.436, over 5555395.71 frames. , ppl: 11.42386801908592], batch size: 70
+2022-12-09 20:07:47,435 INFO [train.py:421] (4/8) Epoch 0, batch 68000, loss[loss=2.383, over 2940.00 frames. , ppl: 10.840358148805231] tot_loss[loss=2.435, over 5543533.90 frames. , ppl: 11.420081102959918], batch size: 70
+2022-12-09 20:07:47,436 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 20:07:48,196 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.413, over 211138.00 frames. , ppl: 11.167173836606553
+2022-12-09 20:09:29,909 INFO [train.py:421] (4/8) Epoch 0, batch 68200, loss[loss=2.433, over 3430.00 frames. , ppl: 11.396674077353124] tot_loss[loss=2.435, over 5542109.24 frames. , ppl: 11.419854858543962], batch size: 70
+2022-12-09 20:11:13,456 INFO [train.py:421] (4/8) Epoch 0, batch 68400, loss[loss=2.359, over 3290.00 frames. , ppl: 10.584593371090415] tot_loss[loss=2.433, over 5637003.83 frames. , ppl: 11.394564508400762], batch size: 70
+2022-12-09 20:12:55,461 INFO [train.py:421] (4/8) Epoch 0, batch 68600, loss[loss=2.652, over 840.00 frames. , ppl: 14.1882819749568] tot_loss[loss=2.432, over 5701687.06 frames. , ppl: 11.380309507479998], batch size: 70
+2022-12-09 20:14:35,016 INFO [train.py:421] (4/8) Epoch 0, batch 68800, loss[loss=2.726, over 770.00 frames. , ppl: 15.2696544843139] tot_loss[loss=2.431, over 5689913.43 frames. , ppl: 11.373270587264644], batch size: 70
+2022-12-09 20:16:15,817 INFO [train.py:421] (4/8) Epoch 0, batch 69000, loss[loss=2.424, over 2590.00 frames. , ppl: 11.288132590930596] tot_loss[loss=2.43, over 5757663.67 frames. , ppl: 11.353278245622839], batch size: 70
+2022-12-09 20:16:15,817 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 20:16:16,582 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.409, over 211138.00 frames. , ppl: 11.12763404873676
+2022-12-09 20:17:58,590 INFO [train.py:421] (4/8) Epoch 0, batch 69200, loss[loss=2.543, over 1680.00 frames. , ppl: 12.723425528591186] tot_loss[loss=2.43, over 5710426.91 frames. , ppl: 11.364011346299497], batch size: 70
+2022-12-09 20:19:40,453 INFO [train.py:421] (4/8) Epoch 0, batch 69400, loss[loss=2.783, over 700.00 frames. , ppl: 16.165172156994995] tot_loss[loss=2.43, over 5692157.88 frames. , ppl: 11.358581610694936], batch size: 70
+2022-12-09 20:21:21,272 INFO [train.py:421] (4/8) Epoch 0, batch 69600, loss[loss=2.4, over 12180.00 frames. , ppl: 11.026245400356025] tot_loss[loss=2.43, over 5674807.27 frames. , ppl: 11.354960595538884], batch size: 70
+2022-12-09 20:23:02,999 INFO [train.py:421] (4/8) Epoch 0, batch 69800, loss[loss=2.406, over 4340.00 frames. , ppl: 11.094238963211913] tot_loss[loss=2.428, over 5699021.58 frames. , ppl: 11.339723726973052], batch size: 70
+2022-12-09 20:24:45,759 INFO [train.py:421] (4/8) Epoch 0, batch 70000, loss[loss=2.431, over 1890.00 frames. , ppl: 11.364924196421281] tot_loss[loss=2.428, over 5675877.00 frames. , ppl: 11.337353600366825], batch size: 70
+2022-12-09 20:24:45,759 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 20:24:46,509 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.41, over 211138.00 frames. , ppl: 11.135786255375706
+2022-12-09 20:26:29,629 INFO [train.py:421] (4/8) Epoch 0, batch 70200, loss[loss=2.442, over 4270.00 frames. , ppl: 11.497176777387365] tot_loss[loss=2.429, over 5658532.96 frames. , ppl: 11.34434716407395], batch size: 70
+2022-12-09 20:28:07,730 INFO [train.py:421] (4/8) Epoch 0, batch 70400, loss[loss=2.411, over 3500.00 frames. , ppl: 11.142137541970103] tot_loss[loss=2.43, over 5593521.15 frames. , ppl: 11.360203840940105], batch size: 70
+2022-12-09 20:29:47,732 INFO [train.py:421] (4/8) Epoch 0, batch 70600, loss[loss=2.355, over 7350.00 frames. , ppl: 10.53938347552359] tot_loss[loss=2.429, over 5610962.97 frames. , ppl: 11.346889359896076], batch size: 70
+2022-12-09 20:31:33,135 INFO [train.py:421] (4/8) Epoch 0, batch 70800, loss[loss=2.399, over 4690.00 frames. , ppl: 11.008349729514196] tot_loss[loss=2.429, over 5609934.18 frames. , ppl: 11.34565763158949], batch size: 70
+2022-12-09 20:33:13,790 INFO [train.py:421] (4/8) Epoch 0, batch 71000, loss[loss=2.535, over 2240.00 frames. , ppl: 12.617568917091175] tot_loss[loss=2.427, over 5652727.09 frames. , ppl: 11.328900003622035], batch size: 70
+2022-12-09 20:33:13,791 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 20:33:14,538 INFO [train.py:452] (4/8) Epoch 0, validation: loss=2.406, over 211138.00 frames. , ppl: 11.087587669142305
+2022-12-09 20:34:53,380 INFO [train.py:421] (4/8) Epoch 0, batch 71200, loss[loss=2.323, over 3360.00 frames. , ppl: 10.207683825170482] tot_loss[loss=2.426, over 5688762.86 frames. , ppl: 11.316349366040086], batch size: 70
+2022-12-09 20:36:32,756 INFO [train.py:421] (4/8) Epoch 0, batch 71400, loss[loss=2.996, over 630.00 frames. , ppl: 20.011379280270276] tot_loss[loss=2.426, over 5676243.45 frames. , ppl: 11.312363138635325], batch size: 70
+2022-12-09 20:38:12,563 INFO [train.py:421] (4/8) Epoch 0, batch 71600, loss[loss=2.317, over 6720.00 frames. , ppl: 10.141639424264902] tot_loss[loss=2.426, over 5676865.78 frames. , ppl: 11.308336993204387], batch size: 70
+2022-12-09 20:39:51,524 INFO [train.py:421] (4/8) Epoch 0, batch 71800, loss[loss=2.412, over 3570.00 frames. , ppl: 11.160359284484267] tot_loss[loss=2.427, over 5663292.35 frames. , ppl: 11.319822536570411], batch size: 70
+2022-12-09 20:41:07,495 INFO [train.py:421] (4/8) Epoch 1, batch 0, loss[loss=2.405, over 910.00 frames. , ppl: 11.07436863463692] tot_loss[loss=2.405, over 910.00 frames. , ppl: 11.07436863463692], batch size: 70
+2022-12-09 20:42:47,122 INFO [train.py:421] (4/8) Epoch 1, batch 200, loss[loss=2.51, over 1610.00 frames. , ppl: 12.300164301060953] tot_loss[loss=2.422, over 545045.52 frames. , ppl: 11.266419469782738], batch size: 70
+2022-12-09 20:44:26,253 INFO [train.py:421] (4/8) Epoch 1, batch 400, loss[loss=2.719, over 770.00 frames. , ppl: 15.162925907919801] tot_loss[loss=2.429, over 952199.59 frames. , ppl: 11.35137565286151], batch size: 70
+2022-12-09 20:46:09,602 INFO [train.py:421] (4/8) Epoch 1, batch 600, loss[loss=2.561, over 1470.00 frames. , ppl: 12.947449292052186] tot_loss[loss=2.426, over 1375086.50 frames. , ppl: 11.31902914729741], batch size: 70
+2022-12-09 20:47:54,142 INFO [train.py:421] (4/8) Epoch 1, batch 800, loss[loss=2.361, over 2170.00 frames. , ppl: 10.598650286053989] tot_loss[loss=2.423, over 1810276.09 frames. , ppl: 11.276938363418216], batch size: 70
+2022-12-09 20:49:35,988 INFO [train.py:421] (4/8) Epoch 1, batch 1000, loss[loss=2.327, over 5740.00 frames. , ppl: 10.252044346603501] tot_loss[loss=2.419, over 2212505.29 frames. , ppl: 11.232540659204282], batch size: 70
+2022-12-09 20:49:35,988 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 20:49:36,749 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.403, over 211138.00 frames. , ppl: 11.059317373352997
+2022-12-09 20:51:14,564 INFO [train.py:421] (4/8) Epoch 1, batch 1200, loss[loss=2.406, over 2310.00 frames. , ppl: 11.086769030976592] tot_loss[loss=2.419, over 2505836.10 frames. , ppl: 11.239093905092206], batch size: 70
+2022-12-09 20:52:52,417 INFO [train.py:421] (4/8) Epoch 1, batch 1400, loss[loss=2.546, over 2030.00 frames. , ppl: 12.751019657362825] tot_loss[loss=2.42, over 2766126.04 frames. , ppl: 11.246115776522895], batch size: 70
+2022-12-09 20:54:34,022 INFO [train.py:421] (4/8) Epoch 1, batch 1600, loss[loss=2.368, over 5880.00 frames. , ppl: 10.671696665481418] tot_loss[loss=2.415, over 3093167.82 frames. , ppl: 11.1934741474798], batch size: 70
+2022-12-09 20:56:14,665 INFO [train.py:421] (4/8) Epoch 1, batch 1800, loss[loss=2.358, over 4550.00 frames. , ppl: 10.573242705938437] tot_loss[loss=2.416, over 3321020.53 frames. , ppl: 11.196774966746258], batch size: 70
+2022-12-09 20:57:57,558 INFO [train.py:421] (4/8) Epoch 1, batch 2000, loss[loss=2.393, over 12460.00 frames. , ppl: 10.945888329242452] tot_loss[loss=2.417, over 3498119.89 frames. , ppl: 11.212541221496803], batch size: 70
+2022-12-09 20:57:57,558 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 20:57:58,304 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.404, over 211138.00 frames. , ppl: 11.068564594231407
+2022-12-09 20:59:42,339 INFO [train.py:421] (4/8) Epoch 1, batch 2200, loss[loss=2.338, over 4270.00 frames. , ppl: 10.363612971648038] tot_loss[loss=2.418, over 3681949.17 frames. , ppl: 11.218994575206677], batch size: 70
+2022-12-09 21:01:20,653 INFO [train.py:421] (4/8) Epoch 1, batch 2400, loss[loss=2.437, over 1540.00 frames. , ppl: 11.441393484592917] tot_loss[loss=2.418, over 3842216.40 frames. , ppl: 11.223801695443395], batch size: 70
+2022-12-09 21:03:02,874 INFO [train.py:421] (4/8) Epoch 1, batch 2600, loss[loss=2.375, over 1890.00 frames. , ppl: 10.753802180695853] tot_loss[loss=2.418, over 3996894.70 frames. , ppl: 11.223744965077268], batch size: 70
+2022-12-09 21:04:41,924 INFO [train.py:421] (4/8) Epoch 1, batch 2800, loss[loss=2.726, over 1050.00 frames. , ppl: 15.279140929520686] tot_loss[loss=2.419, over 4108931.66 frames. , ppl: 11.23318210645118], batch size: 70
+2022-12-09 21:06:24,186 INFO [train.py:421] (4/8) Epoch 1, batch 3000, loss[loss=2.339, over 3850.00 frames. , ppl: 10.367421318492465] tot_loss[loss=2.419, over 4238070.09 frames. , ppl: 11.236329201303592], batch size: 70
+2022-12-09 21:06:24,186 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:06:24,931 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.404, over 211138.00 frames. , ppl: 11.068386028582267
+2022-12-09 21:08:06,436 INFO [train.py:421] (4/8) Epoch 1, batch 3200, loss[loss=2.421, over 2450.00 frames. , ppl: 11.258184152506487] tot_loss[loss=2.417, over 4410822.79 frames. , ppl: 11.21215992242814], batch size: 70
+2022-12-09 21:09:45,260 INFO [train.py:421] (4/8) Epoch 1, batch 3400, loss[loss=2.476, over 2870.00 frames. , ppl: 11.891049969227051] tot_loss[loss=2.418, over 4493238.81 frames. , ppl: 11.222881672795895], batch size: 70
+2022-12-09 21:11:24,244 INFO [train.py:421] (4/8) Epoch 1, batch 3600, loss[loss=2.435, over 3430.00 frames. , ppl: 11.41621134690133] tot_loss[loss=2.42, over 4535631.53 frames. , ppl: 11.244268865520452], batch size: 70
+2022-12-09 21:13:06,616 INFO [train.py:421] (4/8) Epoch 1, batch 3800, loss[loss=2.35, over 4480.00 frames. , ppl: 10.487681899108818] tot_loss[loss=2.419, over 4655267.98 frames. , ppl: 11.233043727213452], batch size: 70
+2022-12-09 21:14:40,525 INFO [train.py:421] (4/8) Epoch 1, batch 4000, loss[loss=2.321, over 4690.00 frames. , ppl: 10.180881110804306] tot_loss[loss=2.42, over 4650213.65 frames. , ppl: 11.247451800045518], batch size: 70
+2022-12-09 21:14:40,525 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:14:41,285 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.405, over 211138.00 frames. , ppl: 11.079054234738054
+2022-12-09 21:16:21,687 INFO [train.py:421] (4/8) Epoch 1, batch 4200, loss[loss=2.403, over 3920.00 frames. , ppl: 11.052596391336468] tot_loss[loss=2.421, over 4714149.90 frames. , ppl: 11.253526802157138], batch size: 70
+2022-12-09 21:18:00,239 INFO [train.py:421] (4/8) Epoch 1, batch 4400, loss[loss=2.326, over 4200.00 frames. , ppl: 10.233704303695628] tot_loss[loss=2.421, over 4785950.40 frames. , ppl: 11.25734541418174], batch size: 70
+2022-12-09 21:19:45,667 INFO [train.py:421] (4/8) Epoch 1, batch 4600, loss[loss=2.81, over 910.00 frames. , ppl: 16.604087839123757] tot_loss[loss=2.419, over 4900540.02 frames. , ppl: 11.23822056621541], batch size: 70
+2022-12-09 21:21:25,293 INFO [train.py:421] (4/8) Epoch 1, batch 4800, loss[loss=2.338, over 3990.00 frames. , ppl: 10.362399671186518] tot_loss[loss=2.418, over 4971410.43 frames. , ppl: 11.227446084613897], batch size: 70
+2022-12-09 21:23:05,013 INFO [train.py:421] (4/8) Epoch 1, batch 5000, loss[loss=2.396, over 4970.00 frames. , ppl: 10.98386947787172] tot_loss[loss=2.418, over 5040150.90 frames. , ppl: 11.219808493447028], batch size: 70
+2022-12-09 21:23:05,014 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:23:05,774 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.399, over 211138.00 frames. , ppl: 11.009013517670198
+2022-12-09 21:24:42,079 INFO [train.py:421] (4/8) Epoch 1, batch 5200, loss[loss=2.432, over 3360.00 frames. , ppl: 11.385382801443248] tot_loss[loss=2.417, over 5079610.52 frames. , ppl: 11.211598704648534], batch size: 70
+2022-12-09 21:26:20,285 INFO [train.py:421] (4/8) Epoch 1, batch 5400, loss[loss=2.389, over 4200.00 frames. , ppl: 10.89764577534649] tot_loss[loss=2.417, over 5097614.00 frames. , ppl: 11.215521225981407], batch size: 70
+2022-12-09 21:28:00,127 INFO [train.py:421] (4/8) Epoch 1, batch 5600, loss[loss=2.308, over 3290.00 frames. , ppl: 10.049394735718407] tot_loss[loss=2.418, over 5109695.30 frames. , ppl: 11.222751954514223], batch size: 70
+2022-12-09 21:29:38,430 INFO [train.py:421] (4/8) Epoch 1, batch 5800, loss[loss=2.491, over 3640.00 frames. , ppl: 12.073525598351626] tot_loss[loss=2.418, over 5166144.90 frames. , ppl: 11.219824022815532], batch size: 70
+2022-12-09 21:31:19,572 INFO [train.py:421] (4/8) Epoch 1, batch 6000, loss[loss=2.604, over 1400.00 frames. , ppl: 13.515632232660913] tot_loss[loss=2.418, over 5211297.12 frames. , ppl: 11.2197680485121], batch size: 70
+2022-12-09 21:31:19,572 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:31:20,318 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.4, over 211138.00 frames. , ppl: 11.020588453227733
+2022-12-09 21:33:01,336 INFO [train.py:421] (4/8) Epoch 1, batch 6200, loss[loss=2.544, over 2450.00 frames. , ppl: 12.734905649102288] tot_loss[loss=2.418, over 5245713.32 frames. , ppl: 11.219131397352525], batch size: 70
+2022-12-09 21:34:44,230 INFO [train.py:421] (4/8) Epoch 1, batch 6400, loss[loss=2.316, over 6860.00 frames. , ppl: 10.130355936734865] tot_loss[loss=2.416, over 5325724.00 frames. , ppl: 11.197681704493963], batch size: 70
+2022-12-09 21:36:22,210 INFO [train.py:421] (4/8) Epoch 1, batch 6600, loss[loss=2.3, over 3990.00 frames. , ppl: 9.973096177233257] tot_loss[loss=2.416, over 5329331.57 frames. , ppl: 11.20045959433986], batch size: 70
+2022-12-09 21:38:05,940 INFO [train.py:421] (4/8) Epoch 1, batch 6800, loss[loss=3.48, over 420.00 frames. , ppl: 32.44815590806631] tot_loss[loss=2.416, over 5356158.17 frames. , ppl: 11.195559994281858], batch size: 70
+2022-12-09 21:39:45,054 INFO [train.py:421] (4/8) Epoch 1, batch 7000, loss[loss=2.573, over 980.00 frames. , ppl: 13.10459811339081] tot_loss[loss=2.415, over 5373627.85 frames. , ppl: 11.19506404655639], batch size: 70
+2022-12-09 21:39:45,054 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:39:45,806 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.398, over 211138.00 frames. , ppl: 11.000016301250483
+2022-12-09 21:41:28,160 INFO [train.py:421] (4/8) Epoch 1, batch 7200, loss[loss=2.304, over 5390.00 frames. , ppl: 10.01027447455842] tot_loss[loss=2.414, over 5412182.32 frames. , ppl: 11.18177933705378], batch size: 70
+2022-12-09 21:43:10,170 INFO [train.py:421] (4/8) Epoch 1, batch 7400, loss[loss=2.384, over 3360.00 frames. , ppl: 10.853167376600224] tot_loss[loss=2.414, over 5398821.16 frames. , ppl: 11.181914187217744], batch size: 70
+2022-12-09 21:44:48,881 INFO [train.py:421] (4/8) Epoch 1, batch 7600, loss[loss=2.376, over 2660.00 frames. , ppl: 10.766824667188276] tot_loss[loss=2.412, over 5447104.68 frames. , ppl: 11.160530402207518], batch size: 70
+2022-12-09 21:46:31,987 INFO [train.py:421] (4/8) Epoch 1, batch 7800, loss[loss=2.452, over 2800.00 frames. , ppl: 11.606880569114665] tot_loss[loss=2.413, over 5454650.40 frames. , ppl: 11.169143254702822], batch size: 70
+2022-12-09 21:48:11,093 INFO [train.py:421] (4/8) Epoch 1, batch 8000, loss[loss=2.401, over 3990.00 frames. , ppl: 11.036026106514203] tot_loss[loss=2.414, over 5422170.08 frames. , ppl: 11.18232415487553], batch size: 70
+2022-12-09 21:48:11,093 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:48:11,857 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.397, over 211138.00 frames. , ppl: 10.99293966703717
+2022-12-09 21:49:52,602 INFO [train.py:421] (4/8) Epoch 1, batch 8200, loss[loss=2.538, over 2310.00 frames. , ppl: 12.656574831557164] tot_loss[loss=2.415, over 5380136.90 frames. , ppl: 11.190826554546485], batch size: 70
+2022-12-09 21:51:31,106 INFO [train.py:421] (4/8) Epoch 1, batch 8400, loss[loss=2.317, over 2660.00 frames. , ppl: 10.147520121981287] tot_loss[loss=2.417, over 5335989.18 frames. , ppl: 11.208841242522997], batch size: 70
+2022-12-09 21:53:08,554 INFO [train.py:421] (4/8) Epoch 1, batch 8600, loss[loss=2.29, over 3290.00 frames. , ppl: 9.878706690733456] tot_loss[loss=2.417, over 5346451.19 frames. , ppl: 11.210630466167617], batch size: 70
+2022-12-09 21:54:48,828 INFO [train.py:421] (4/8) Epoch 1, batch 8800, loss[loss=2.405, over 2940.00 frames. , ppl: 11.08383620668995] tot_loss[loss=2.417, over 5366472.41 frames. , ppl: 11.209062154503565], batch size: 70
+2022-12-09 21:56:32,520 INFO [train.py:421] (4/8) Epoch 1, batch 9000, loss[loss=2.402, over 3780.00 frames. , ppl: 11.047703561887209] tot_loss[loss=2.416, over 5389959.95 frames. , ppl: 11.201693065046088], batch size: 70
+2022-12-09 21:56:32,521 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 21:56:33,273 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.396, over 211138.00 frames. , ppl: 10.981821240677286
+2022-12-09 21:58:14,745 INFO [train.py:421] (4/8) Epoch 1, batch 9200, loss[loss=2.462, over 2030.00 frames. , ppl: 11.72910988518284] tot_loss[loss=2.415, over 5436959.58 frames. , ppl: 11.189194708122233], batch size: 70
+2022-12-09 21:59:56,481 INFO [train.py:421] (4/8) Epoch 1, batch 9400, loss[loss=2.77, over 770.00 frames. , ppl: 15.961496166345118] tot_loss[loss=2.415, over 5422080.64 frames. , ppl: 11.184930417944827], batch size: 70
+2022-12-09 22:01:34,866 INFO [train.py:421] (4/8) Epoch 1, batch 9600, loss[loss=2.378, over 1610.00 frames. , ppl: 10.784070658072041] tot_loss[loss=2.414, over 5444917.91 frames. , ppl: 11.182836264136617], batch size: 70
+2022-12-09 22:03:13,418 INFO [train.py:421] (4/8) Epoch 1, batch 9800, loss[loss=2.771, over 910.00 frames. , ppl: 15.967811012352048] tot_loss[loss=2.413, over 5451821.17 frames. , ppl: 11.171467717084985], batch size: 70
+2022-12-09 22:04:51,393 INFO [train.py:421] (4/8) Epoch 1, batch 10000, loss[loss=2.508, over 2100.00 frames. , ppl: 12.278724767803766] tot_loss[loss=2.413, over 5430520.31 frames. , ppl: 11.169477696589977], batch size: 70
+2022-12-09 22:04:51,394 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:04:52,141 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.395, over 211138.00 frames. , ppl: 10.972836503018678
+2022-12-09 22:06:33,531 INFO [train.py:421] (4/8) Epoch 1, batch 10200, loss[loss=2.477, over 2800.00 frames. , ppl: 11.906798617625181] tot_loss[loss=2.414, over 5394092.54 frames. , ppl: 11.180757979063621], batch size: 70
+2022-12-09 22:08:13,457 INFO [train.py:421] (4/8) Epoch 1, batch 10400, loss[loss=2.842, over 840.00 frames. , ppl: 17.14395185135798] tot_loss[loss=2.413, over 5451296.20 frames. , ppl: 11.167436339753753], batch size: 70
+2022-12-09 22:09:54,329 INFO [train.py:421] (4/8) Epoch 1, batch 10600, loss[loss=2.323, over 2240.00 frames. , ppl: 10.210573153551293] tot_loss[loss=2.412, over 5483962.30 frames. , ppl: 11.157267019776675], batch size: 70
+2022-12-09 22:11:32,870 INFO [train.py:421] (4/8) Epoch 1, batch 10800, loss[loss=3.455, over 490.00 frames. , ppl: 31.6529071801682] tot_loss[loss=2.412, over 5490900.53 frames. , ppl: 11.157650013537294], batch size: 70
+2022-12-09 22:13:12,648 INFO [train.py:421] (4/8) Epoch 1, batch 11000, loss[loss=2.426, over 6720.00 frames. , ppl: 11.31834522163239] tot_loss[loss=2.412, over 5474666.75 frames. , ppl: 11.157159297778742], batch size: 70
+2022-12-09 22:13:12,649 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:13:13,393 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.394, over 211138.00 frames. , ppl: 10.959451037049323
+2022-12-09 22:14:51,704 INFO [train.py:421] (4/8) Epoch 1, batch 11200, loss[loss=3.191, over 560.00 frames. , ppl: 24.30112792930937] tot_loss[loss=2.413, over 5445440.20 frames. , ppl: 11.162605799345105], batch size: 70
+2022-12-09 22:16:32,959 INFO [train.py:421] (4/8) Epoch 1, batch 11400, loss[loss=2.432, over 2310.00 frames. , ppl: 11.381468221027742] tot_loss[loss=2.412, over 5446829.52 frames. , ppl: 11.152356941912243], batch size: 70
+2022-12-09 22:18:09,022 INFO [train.py:421] (4/8) Epoch 1, batch 11600, loss[loss=2.37, over 4200.00 frames. , ppl: 10.695002253062796] tot_loss[loss=2.411, over 5412737.61 frames. , ppl: 11.149465184615755], batch size: 70
+2022-12-09 22:19:53,238 INFO [train.py:421] (4/8) Epoch 1, batch 11800, loss[loss=2.473, over 2730.00 frames. , ppl: 11.855579737104302] tot_loss[loss=2.412, over 5404916.95 frames. , ppl: 11.153279706735256], batch size: 70
+2022-12-09 22:21:30,851 INFO [train.py:421] (4/8) Epoch 1, batch 12000, loss[loss=2.356, over 5320.00 frames. , ppl: 10.546317754932302] tot_loss[loss=2.409, over 5425868.59 frames. , ppl: 11.127482546594612], batch size: 70
+2022-12-09 22:21:30,852 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:21:31,611 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.393, over 211138.00 frames. , ppl: 10.943533776107909
+2022-12-09 22:23:19,196 INFO [train.py:421] (4/8) Epoch 1, batch 12200, loss[loss=2.485, over 1540.00 frames. , ppl: 12.001315460313922] tot_loss[loss=2.408, over 5482158.21 frames. , ppl: 11.113884007049906], batch size: 70
+2022-12-09 22:24:55,362 INFO [train.py:421] (4/8) Epoch 1, batch 12400, loss[loss=2.485, over 1540.00 frames. , ppl: 11.997368208309801] tot_loss[loss=2.409, over 5459797.03 frames. , ppl: 11.12272702220907], batch size: 70
+2022-12-09 22:26:36,488 INFO [train.py:421] (4/8) Epoch 1, batch 12600, loss[loss=2.403, over 4200.00 frames. , ppl: 11.059025012217202] tot_loss[loss=2.408, over 5488673.62 frames. , ppl: 11.11170003066351], batch size: 70
+2022-12-09 22:28:14,905 INFO [train.py:421] (4/8) Epoch 1, batch 12800, loss[loss=2.268, over 4340.00 frames. , ppl: 9.660381151457614] tot_loss[loss=2.408, over 5479279.28 frames. , ppl: 11.111519630647654], batch size: 70
+2022-12-09 22:29:53,721 INFO [train.py:421] (4/8) Epoch 1, batch 13000, loss[loss=2.495, over 2660.00 frames. , ppl: 12.12041276021002] tot_loss[loss=2.41, over 5423472.40 frames. , ppl: 11.129908221911181], batch size: 70
+2022-12-09 22:29:53,722 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:29:54,481 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.391, over 211138.00 frames. , ppl: 10.927798136667734
+2022-12-09 22:31:36,555 INFO [train.py:421] (4/8) Epoch 1, batch 13200, loss[loss=2.334, over 5880.00 frames. , ppl: 10.31538290489497] tot_loss[loss=2.41, over 5418468.86 frames. , ppl: 11.137908542745826], batch size: 70
+2022-12-09 22:33:15,726 INFO [train.py:421] (4/8) Epoch 1, batch 13400, loss[loss=2.692, over 1190.00 frames. , ppl: 14.761368389040365] tot_loss[loss=2.41, over 5415118.77 frames. , ppl: 11.137079107623489], batch size: 70
+2022-12-09 22:34:54,052 INFO [train.py:421] (4/8) Epoch 1, batch 13600, loss[loss=2.627, over 1680.00 frames. , ppl: 13.838004220209928] tot_loss[loss=2.41, over 5447648.18 frames. , ppl: 11.130500302590416], batch size: 70
+2022-12-09 22:36:30,646 INFO [train.py:421] (4/8) Epoch 1, batch 13800, loss[loss=2.611, over 1330.00 frames. , ppl: 13.605880449616821] tot_loss[loss=2.409, over 5459672.35 frames. , ppl: 11.126845500661371], batch size: 70
+2022-12-09 22:38:12,937 INFO [train.py:421] (4/8) Epoch 1, batch 14000, loss[loss=2.381, over 4200.00 frames. , ppl: 10.820877820389859] tot_loss[loss=2.41, over 5432554.78 frames. , ppl: 11.13575755938485], batch size: 70
+2022-12-09 22:38:12,937 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:38:13,682 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.39, over 211138.00 frames. , ppl: 10.914509612851669
+2022-12-09 22:39:52,639 INFO [train.py:421] (4/8) Epoch 1, batch 14200, loss[loss=2.49, over 840.00 frames. , ppl: 12.058326224897284] tot_loss[loss=2.41, over 5406223.83 frames. , ppl: 11.13262678563628], batch size: 70
+2022-12-09 22:41:35,426 INFO [train.py:421] (4/8) Epoch 1, batch 14400, loss[loss=2.39, over 3920.00 frames. , ppl: 10.909973117063043] tot_loss[loss=2.409, over 5460301.74 frames. , ppl: 11.117836762830871], batch size: 70
+2022-12-09 22:43:15,960 INFO [train.py:421] (4/8) Epoch 1, batch 14600, loss[loss=2.46, over 2450.00 frames. , ppl: 11.710028733657657] tot_loss[loss=2.407, over 5529396.59 frames. , ppl: 11.098888792559906], batch size: 70
+2022-12-09 22:44:53,639 INFO [train.py:421] (4/8) Epoch 1, batch 14800, loss[loss=2.596, over 1400.00 frames. , ppl: 13.415556538593506] tot_loss[loss=2.407, over 5495861.14 frames. , ppl: 11.099739796917342], batch size: 70
+2022-12-09 22:46:32,345 INFO [train.py:421] (4/8) Epoch 1, batch 15000, loss[loss=2.409, over 2380.00 frames. , ppl: 11.117286316499573] tot_loss[loss=2.407, over 5502696.00 frames. , ppl: 11.09977934519465], batch size: 70
+2022-12-09 22:46:32,345 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:46:33,092 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.389, over 211138.00 frames. , ppl: 10.905875331584172
+2022-12-09 22:48:15,322 INFO [train.py:421] (4/8) Epoch 1, batch 15200, loss[loss=3.008, over 560.00 frames. , ppl: 20.24781829816559] tot_loss[loss=2.406, over 5549795.23 frames. , ppl: 11.084654655226574], batch size: 70
+2022-12-09 22:49:57,178 INFO [train.py:421] (4/8) Epoch 1, batch 15400, loss[loss=2.282, over 3360.00 frames. , ppl: 9.793019487533755] tot_loss[loss=2.405, over 5573228.06 frames. , ppl: 11.08050964967202], batch size: 70
+2022-12-09 22:51:38,846 INFO [train.py:421] (4/8) Epoch 1, batch 15600, loss[loss=2.449, over 1400.00 frames. , ppl: 11.57520654227778] tot_loss[loss=2.405, over 5599155.30 frames. , ppl: 11.074642357050202], batch size: 70
+2022-12-09 22:53:18,733 INFO [train.py:421] (4/8) Epoch 1, batch 15800, loss[loss=2.795, over 770.00 frames. , ppl: 16.359560840390102] tot_loss[loss=2.405, over 5601278.13 frames. , ppl: 11.077125671748627], batch size: 70
+2022-12-09 22:55:01,063 INFO [train.py:421] (4/8) Epoch 1, batch 16000, loss[loss=2.367, over 3010.00 frames. , ppl: 10.666278770823087] tot_loss[loss=2.404, over 5623813.42 frames. , ppl: 11.064421022469725], batch size: 70
+2022-12-09 22:55:01,063 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 22:55:01,823 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.387, over 211138.00 frames. , ppl: 10.886231072245774
+2022-12-09 22:56:40,411 INFO [train.py:421] (4/8) Epoch 1, batch 16200, loss[loss=2.51, over 2450.00 frames. , ppl: 12.301882162368209] tot_loss[loss=2.404, over 5595240.51 frames. , ppl: 11.070641570638204], batch size: 70
+2022-12-09 22:58:20,171 INFO [train.py:421] (4/8) Epoch 1, batch 16400, loss[loss=2.498, over 2870.00 frames. , ppl: 12.152366040245788] tot_loss[loss=2.404, over 5582505.62 frames. , ppl: 11.068742595263423], batch size: 70
+2022-12-09 23:00:00,604 INFO [train.py:421] (4/8) Epoch 1, batch 16600, loss[loss=2.402, over 2660.00 frames. , ppl: 11.041767426391026] tot_loss[loss=2.402, over 5626517.63 frames. , ppl: 11.04985020021296], batch size: 70
+2022-12-09 23:01:39,180 INFO [train.py:421] (4/8) Epoch 1, batch 16800, loss[loss=2.499, over 1750.00 frames. , ppl: 12.176118855151131] tot_loss[loss=2.404, over 5582732.21 frames. , ppl: 11.065008769675412], batch size: 70
+2022-12-09 23:03:20,353 INFO [train.py:421] (4/8) Epoch 1, batch 17000, loss[loss=2.479, over 2030.00 frames. , ppl: 11.931926894177595] tot_loss[loss=2.405, over 5550105.02 frames. , ppl: 11.078926416952061], batch size: 70
+2022-12-09 23:03:20,353 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:03:21,099 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.388, over 211138.00 frames. , ppl: 10.88762972140702
+2022-12-09 23:04:58,074 INFO [train.py:421] (4/8) Epoch 1, batch 17200, loss[loss=2.954, over 630.00 frames. , ppl: 19.183516180883284] tot_loss[loss=2.407, over 5505106.41 frames. , ppl: 11.097577658714489], batch size: 70
+2022-12-09 23:06:36,551 INFO [train.py:421] (4/8) Epoch 1, batch 17400, loss[loss=2.41, over 1960.00 frames. , ppl: 11.128692865745975] tot_loss[loss=2.408, over 5467169.48 frames. , ppl: 11.109142627419303], batch size: 70
+2022-12-09 23:08:16,406 INFO [train.py:421] (4/8) Epoch 1, batch 17600, loss[loss=2.659, over 840.00 frames. , ppl: 14.27546466652839] tot_loss[loss=2.407, over 5496479.76 frames. , ppl: 11.096889771022123], batch size: 70
+2022-12-09 23:09:57,151 INFO [train.py:421] (4/8) Epoch 1, batch 17800, loss[loss=2.339, over 3150.00 frames. , ppl: 10.366172567306489] tot_loss[loss=2.406, over 5505224.40 frames. , ppl: 11.086600889082623], batch size: 70
+2022-12-09 23:11:36,545 INFO [train.py:421] (4/8) Epoch 1, batch 18000, loss[loss=2.941, over 560.00 frames. , ppl: 18.93202161510724] tot_loss[loss=2.406, over 5489173.87 frames. , ppl: 11.086537361695976], batch size: 70
+2022-12-09 23:11:36,545 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:11:37,293 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.386, over 211138.00 frames. , ppl: 10.869570686577362
+2022-12-09 23:13:24,070 INFO [train.py:421] (4/8) Epoch 1, batch 18200, loss[loss=2.452, over 3290.00 frames. , ppl: 11.614918170595258] tot_loss[loss=2.406, over 5479075.04 frames. , ppl: 11.086251874124834], batch size: 70
+2022-12-09 23:15:06,573 INFO [train.py:421] (4/8) Epoch 1, batch 18400, loss[loss=2.459, over 3360.00 frames. , ppl: 11.694092211741513] tot_loss[loss=2.406, over 5435303.42 frames. , ppl: 11.089557475042527], batch size: 70
+2022-12-09 23:16:40,257 INFO [train.py:421] (4/8) Epoch 1, batch 18600, loss[loss=2.991, over 630.00 frames. , ppl: 19.900648156876038] tot_loss[loss=2.406, over 5414947.81 frames. , ppl: 11.093822827031703], batch size: 70
+2022-12-09 23:18:19,348 INFO [train.py:421] (4/8) Epoch 1, batch 18800, loss[loss=2.566, over 1330.00 frames. , ppl: 13.009337948969081] tot_loss[loss=2.406, over 5436615.18 frames. , ppl: 11.088156107455477], batch size: 70
+2022-12-09 23:19:58,913 INFO [train.py:421] (4/8) Epoch 1, batch 19000, loss[loss=2.433, over 1890.00 frames. , ppl: 11.397638536207243] tot_loss[loss=2.405, over 5426638.00 frames. , ppl: 11.08070962900957], batch size: 70
+2022-12-09 23:19:58,913 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:19:59,659 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.385, over 211138.00 frames. , ppl: 10.86154424255148
+2022-12-09 23:21:39,313 INFO [train.py:421] (4/8) Epoch 1, batch 19200, loss[loss=2.303, over 5530.00 frames. , ppl: 10.005273312584283] tot_loss[loss=2.405, over 5456093.88 frames. , ppl: 11.078722023645218], batch size: 70
+2022-12-09 23:23:22,546 INFO [train.py:421] (4/8) Epoch 1, batch 19400, loss[loss=2.457, over 2590.00 frames. , ppl: 11.670259618256264] tot_loss[loss=2.403, over 5481846.35 frames. , ppl: 11.058181966855756], batch size: 70
+2022-12-09 23:25:00,329 INFO [train.py:421] (4/8) Epoch 1, batch 19600, loss[loss=2.459, over 1400.00 frames. , ppl: 11.696529414652227] tot_loss[loss=2.404, over 5482275.52 frames. , ppl: 11.063598107216455], batch size: 70
+2022-12-09 23:26:38,753 INFO [train.py:421] (4/8) Epoch 1, batch 19800, loss[loss=2.589, over 910.00 frames. , ppl: 13.321962048266547] tot_loss[loss=2.403, over 5483852.57 frames. , ppl: 11.06028448971776], batch size: 70
+2022-12-09 23:28:20,850 INFO [train.py:421] (4/8) Epoch 1, batch 20000, loss[loss=2.331, over 4690.00 frames. , ppl: 10.293308629624361] tot_loss[loss=2.402, over 5509642.88 frames. , ppl: 11.048546576163377], batch size: 70
+2022-12-09 23:28:20,851 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:28:21,595 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.383, over 211138.00 frames. , ppl: 10.840206792981473
+2022-12-09 23:30:07,793 INFO [train.py:421] (4/8) Epoch 1, batch 20200, loss[loss=2.357, over 3710.00 frames. , ppl: 10.561930178517324] tot_loss[loss=2.402, over 5513865.65 frames. , ppl: 11.047048559634188], batch size: 70
+2022-12-09 23:31:46,078 INFO [train.py:421] (4/8) Epoch 1, batch 20400, loss[loss=2.313, over 2870.00 frames. , ppl: 10.100009944267809] tot_loss[loss=2.401, over 5544873.33 frames. , ppl: 11.037092139674456], batch size: 70
+2022-12-09 23:33:26,949 INFO [train.py:421] (4/8) Epoch 1, batch 20600, loss[loss=2.556, over 980.00 frames. , ppl: 12.88923540662622] tot_loss[loss=2.401, over 5562236.27 frames. , ppl: 11.038950110602759], batch size: 70
+2022-12-09 23:35:04,530 INFO [train.py:421] (4/8) Epoch 1, batch 20800, loss[loss=2.333, over 8750.00 frames. , ppl: 10.311017115168505] tot_loss[loss=2.403, over 5485567.36 frames. , ppl: 11.05295418760141], batch size: 70
+2022-12-09 23:36:44,985 INFO [train.py:421] (4/8) Epoch 1, batch 21000, loss[loss=2.373, over 2520.00 frames. , ppl: 10.729468032899051] tot_loss[loss=2.402, over 5489617.20 frames. , ppl: 11.04929938951868], batch size: 70
+2022-12-09 23:36:44,985 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:36:45,739 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.381, over 211138.00 frames. , ppl: 10.817335716309938
+2022-12-09 23:38:28,716 INFO [train.py:421] (4/8) Epoch 1, batch 21200, loss[loss=2.202, over 6440.00 frames. , ppl: 9.042517839618883] tot_loss[loss=2.4, over 5568705.12 frames. , ppl: 11.024436023717907], batch size: 70
+2022-12-09 23:40:05,767 INFO [train.py:421] (4/8) Epoch 1, batch 21400, loss[loss=2.64, over 1050.00 frames. , ppl: 14.00900105893467] tot_loss[loss=2.401, over 5519491.09 frames. , ppl: 11.037017825177394], batch size: 70
+2022-12-09 23:41:45,963 INFO [train.py:421] (4/8) Epoch 1, batch 21600, loss[loss=2.474, over 3500.00 frames. , ppl: 11.874783660557387] tot_loss[loss=2.402, over 5483267.49 frames. , ppl: 11.042629865159682], batch size: 70
+2022-12-09 23:43:25,697 INFO [train.py:421] (4/8) Epoch 1, batch 21800, loss[loss=2.483, over 1750.00 frames. , ppl: 11.979454774692007] tot_loss[loss=2.402, over 5468332.78 frames. , ppl: 11.040161945234722], batch size: 70
+2022-12-09 23:45:06,450 INFO [train.py:421] (4/8) Epoch 1, batch 22000, loss[loss=2.475, over 1470.00 frames. , ppl: 11.87879876834839] tot_loss[loss=2.4, over 5516383.17 frames. , ppl: 11.022240100370004], batch size: 70
+2022-12-09 23:45:06,450 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:45:07,207 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.382, over 211138.00 frames. , ppl: 10.821137268370757
+2022-12-09 23:46:46,113 INFO [train.py:421] (4/8) Epoch 1, batch 22200, loss[loss=2.586, over 1120.00 frames. , ppl: 13.271967743779635] tot_loss[loss=2.401, over 5498006.74 frames. , ppl: 11.034090159133472], batch size: 70
+2022-12-09 23:48:27,624 INFO [train.py:421] (4/8) Epoch 1, batch 22400, loss[loss=2.497, over 1680.00 frames. , ppl: 12.147353088414143] tot_loss[loss=2.402, over 5463789.73 frames. , ppl: 11.040805723790484], batch size: 70
+2022-12-09 23:50:05,039 INFO [train.py:421] (4/8) Epoch 1, batch 22600, loss[loss=2.743, over 770.00 frames. , ppl: 15.541060209583073] tot_loss[loss=2.402, over 5422162.93 frames. , ppl: 11.044857740949725], batch size: 70
+2022-12-09 23:51:43,332 INFO [train.py:421] (4/8) Epoch 1, batch 22800, loss[loss=2.472, over 1330.00 frames. , ppl: 11.841469444736166] tot_loss[loss=2.403, over 5387821.70 frames. , ppl: 11.056465342562811], batch size: 70
+2022-12-09 23:53:27,630 INFO [train.py:421] (4/8) Epoch 1, batch 23000, loss[loss=2.325, over 6720.00 frames. , ppl: 10.222074028728729] tot_loss[loss=2.403, over 5381029.80 frames. , ppl: 11.061488957273461], batch size: 70
+2022-12-09 23:53:27,631 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-09 23:53:28,375 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.38, over 211138.00 frames. , ppl: 10.800644888892075
+2022-12-09 23:55:09,816 INFO [train.py:421] (4/8) Epoch 1, batch 23200, loss[loss=2.592, over 1120.00 frames. , ppl: 13.357617107617932] tot_loss[loss=2.403, over 5398239.10 frames. , ppl: 11.056545664779469], batch size: 70
+2022-12-09 23:56:52,367 INFO [train.py:421] (4/8) Epoch 1, batch 23400, loss[loss=2.348, over 7840.00 frames. , ppl: 10.462182718596578] tot_loss[loss=2.404, over 5393180.54 frames. , ppl: 11.063069403697149], batch size: 70
+2022-12-09 23:58:28,672 INFO [train.py:421] (4/8) Epoch 1, batch 23600, loss[loss=2.342, over 4760.00 frames. , ppl: 10.403663013097256] tot_loss[loss=2.402, over 5432740.12 frames. , ppl: 11.05030319038819], batch size: 70
+2022-12-10 00:00:06,981 INFO [train.py:421] (4/8) Epoch 1, batch 23800, loss[loss=2.423, over 2100.00 frames. , ppl: 11.283838331160878] tot_loss[loss=2.402, over 5419221.79 frames. , ppl: 11.047723206973526], batch size: 70
+2022-12-10 00:01:47,289 INFO [train.py:421] (4/8) Epoch 1, batch 24000, loss[loss=2.263, over 4830.00 frames. , ppl: 9.608075319951045] tot_loss[loss=2.403, over 5408325.14 frames. , ppl: 11.05078809002534], batch size: 70
+2022-12-10 00:01:47,290 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:01:48,035 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.379, over 211138.00 frames. , ppl: 10.78979920368168
+2022-12-10 00:03:26,231 INFO [train.py:421] (4/8) Epoch 1, batch 24200, loss[loss=2.315, over 4340.00 frames. , ppl: 10.120671679228723] tot_loss[loss=2.402, over 5404340.19 frames. , ppl: 11.050018170170425], batch size: 70
+2022-12-10 00:05:05,802 INFO [train.py:421] (4/8) Epoch 1, batch 24400, loss[loss=2.469, over 1890.00 frames. , ppl: 11.812150923176482] tot_loss[loss=2.403, over 5375025.12 frames. , ppl: 11.051541394576372], batch size: 70
+2022-12-10 00:06:43,341 INFO [train.py:421] (4/8) Epoch 1, batch 24600, loss[loss=2.36, over 2870.00 frames. , ppl: 10.594156380457603] tot_loss[loss=2.402, over 5383734.75 frames. , ppl: 11.045368570868883], batch size: 70
+2022-12-10 00:08:25,288 INFO [train.py:421] (4/8) Epoch 1, batch 24800, loss[loss=2.492, over 770.00 frames. , ppl: 12.079587740454443] tot_loss[loss=2.401, over 5425076.24 frames. , ppl: 11.035134429104783], batch size: 70
+2022-12-10 00:10:06,626 INFO [train.py:421] (4/8) Epoch 1, batch 25000, loss[loss=2.89, over 630.00 frames. , ppl: 17.99093759307168] tot_loss[loss=2.4, over 5445072.01 frames. , ppl: 11.023996038060638], batch size: 70
+2022-12-10 00:10:06,627 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:10:07,386 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.38, over 211138.00 frames. , ppl: 10.800197296869092
+2022-12-10 00:11:43,292 INFO [train.py:421] (4/8) Epoch 1, batch 25200, loss[loss=2.458, over 3710.00 frames. , ppl: 11.676087631384473] tot_loss[loss=2.4, over 5418933.02 frames. , ppl: 11.023502620729861], batch size: 70
+2022-12-10 00:13:24,100 INFO [train.py:421] (4/8) Epoch 1, batch 25400, loss[loss=2.272, over 5880.00 frames. , ppl: 9.697595706506446] tot_loss[loss=2.399, over 5474070.04 frames. , ppl: 11.009502427989991], batch size: 70
+2022-12-10 00:15:07,567 INFO [train.py:421] (4/8) Epoch 1, batch 25600, loss[loss=2.341, over 3220.00 frames. , ppl: 10.390457037590286] tot_loss[loss=2.398, over 5483928.49 frames. , ppl: 10.999798197139082], batch size: 70
+2022-12-10 00:16:48,181 INFO [train.py:421] (4/8) Epoch 1, batch 25800, loss[loss=2.428, over 2520.00 frames. , ppl: 11.340904206298198] tot_loss[loss=2.397, over 5479036.26 frames. , ppl: 10.990454007211849], batch size: 70
+2022-12-10 00:18:27,901 INFO [train.py:421] (4/8) Epoch 1, batch 26000, loss[loss=2.42, over 1190.00 frames. , ppl: 11.249258799494198] tot_loss[loss=2.395, over 5515288.32 frames. , ppl: 10.96833026715018], batch size: 70
+2022-12-10 00:18:27,901 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:18:28,648 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.379, over 211138.00 frames. , ppl: 10.799350119895008
+2022-12-10 00:20:08,713 INFO [train.py:421] (4/8) Epoch 1, batch 26200, loss[loss=2.325, over 1960.00 frames. , ppl: 10.229029336927603] tot_loss[loss=2.394, over 5502367.43 frames. , ppl: 10.960769114833564], batch size: 70
+2022-12-10 00:21:44,700 INFO [train.py:421] (4/8) Epoch 1, batch 26400, loss[loss=2.289, over 9660.00 frames. , ppl: 9.865006086661504] tot_loss[loss=2.395, over 5480190.02 frames. , ppl: 10.967301646569043], batch size: 70
+2022-12-10 00:23:22,416 INFO [train.py:421] (4/8) Epoch 1, batch 26600, loss[loss=2.429, over 1470.00 frames. , ppl: 11.352796708756063] tot_loss[loss=2.396, over 5451795.83 frames. , ppl: 10.975214160983796], batch size: 70
+2022-12-10 00:25:04,487 INFO [train.py:421] (4/8) Epoch 1, batch 26800, loss[loss=2.427, over 1820.00 frames. , ppl: 11.325328295480789] tot_loss[loss=2.396, over 5432457.45 frames. , ppl: 10.98206681608419], batch size: 70
+2022-12-10 00:26:48,138 INFO [train.py:421] (4/8) Epoch 1, batch 27000, loss[loss=2.329, over 3570.00 frames. , ppl: 10.270649967103537] tot_loss[loss=2.395, over 5451665.20 frames. , ppl: 10.973429817720302], batch size: 70
+2022-12-10 00:26:48,139 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:26:48,899 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.379, over 211138.00 frames. , ppl: 10.791004984020308
+2022-12-10 00:28:29,579 INFO [train.py:421] (4/8) Epoch 1, batch 27200, loss[loss=2.486, over 840.00 frames. , ppl: 12.01727590199064] tot_loss[loss=2.396, over 5440614.13 frames. , ppl: 10.980396442556874], batch size: 70
+2022-12-10 00:30:09,129 INFO [train.py:421] (4/8) Epoch 1, batch 27400, loss[loss=2.572, over 1260.00 frames. , ppl: 13.098041056606476] tot_loss[loss=2.396, over 5457900.25 frames. , ppl: 10.978767000731557], batch size: 70
+2022-12-10 00:31:48,115 INFO [train.py:421] (4/8) Epoch 1, batch 27600, loss[loss=2.41, over 2240.00 frames. , ppl: 11.131216048769916] tot_loss[loss=2.397, over 5460661.55 frames. , ppl: 10.990699148324069], batch size: 70
+2022-12-10 00:33:29,296 INFO [train.py:421] (4/8) Epoch 1, batch 27800, loss[loss=2.51, over 1820.00 frames. , ppl: 12.302458996354273] tot_loss[loss=2.396, over 5486504.60 frames. , ppl: 10.9822281941566], batch size: 70
+2022-12-10 00:35:07,717 INFO [train.py:421] (4/8) Epoch 1, batch 28000, loss[loss=2.487, over 2870.00 frames. , ppl: 12.028299370587554] tot_loss[loss=2.396, over 5476188.29 frames. , ppl: 10.981588069084774], batch size: 70
+2022-12-10 00:35:07,717 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:35:08,463 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.377, over 211138.00 frames. , ppl: 10.77500384045564
+2022-12-10 00:36:52,798 INFO [train.py:421] (4/8) Epoch 1, batch 28200, loss[loss=2.317, over 5250.00 frames. , ppl: 10.146733040176562] tot_loss[loss=2.395, over 5515015.18 frames. , ppl: 10.973588214323028], batch size: 70
+2022-12-10 00:38:30,675 INFO [train.py:421] (4/8) Epoch 1, batch 28400, loss[loss=2.333, over 2520.00 frames. , ppl: 10.305175573115525] tot_loss[loss=2.395, over 5525338.35 frames. , ppl: 10.968302050598618], batch size: 70
+2022-12-10 00:40:09,191 INFO [train.py:421] (4/8) Epoch 1, batch 28600, loss[loss=2.291, over 910.00 frames. , ppl: 9.883465786192382] tot_loss[loss=2.395, over 5512166.51 frames. , ppl: 10.965703151616367], batch size: 70
+2022-12-10 00:41:48,510 INFO [train.py:421] (4/8) Epoch 1, batch 28800, loss[loss=2.57, over 770.00 frames. , ppl: 13.059686778222217] tot_loss[loss=2.395, over 5492672.69 frames. , ppl: 10.969777588193304], batch size: 70
+2022-12-10 00:43:30,618 INFO [train.py:421] (4/8) Epoch 1, batch 29000, loss[loss=4.243, over 350.00 frames. , ppl: 69.59707333982291] tot_loss[loss=2.395, over 5477025.07 frames. , ppl: 10.972860843949473], batch size: 70
+2022-12-10 00:43:30,618 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:43:31,364 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.375, over 211138.00 frames. , ppl: 10.749719593753966
+2022-12-10 00:45:08,474 INFO [train.py:421] (4/8) Epoch 1, batch 29200, loss[loss=2.337, over 2870.00 frames. , ppl: 10.346253736955738] tot_loss[loss=2.396, over 5450611.63 frames. , ppl: 10.977405978123622], batch size: 70
+2022-12-10 00:46:48,119 INFO [train.py:421] (4/8) Epoch 1, batch 29400, loss[loss=2.383, over 2660.00 frames. , ppl: 10.834097351729165] tot_loss[loss=2.395, over 5463393.27 frames. , ppl: 10.971633633485153], batch size: 70
+2022-12-10 00:48:31,855 INFO [train.py:421] (4/8) Epoch 1, batch 29600, loss[loss=2.58, over 1190.00 frames. , ppl: 13.194641514115403] tot_loss[loss=2.395, over 5461753.73 frames. , ppl: 10.965927373806037], batch size: 70
+2022-12-10 00:50:13,023 INFO [train.py:421] (4/8) Epoch 1, batch 29800, loss[loss=2.315, over 4340.00 frames. , ppl: 10.121277458326345] tot_loss[loss=2.394, over 5474409.57 frames. , ppl: 10.960837633346992], batch size: 70
+2022-12-10 00:51:49,940 INFO [train.py:421] (4/8) Epoch 1, batch 30000, loss[loss=2.375, over 2730.00 frames. , ppl: 10.751070873243838] tot_loss[loss=2.393, over 5518306.55 frames. , ppl: 10.944242322694247], batch size: 70
+2022-12-10 00:51:49,940 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 00:51:50,686 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.373, over 211138.00 frames. , ppl: 10.73370182407569
+2022-12-10 00:53:27,749 INFO [train.py:421] (4/8) Epoch 1, batch 30200, loss[loss=2.444, over 1820.00 frames. , ppl: 11.517313353879235] tot_loss[loss=2.393, over 5512112.68 frames. , ppl: 10.943707750089823], batch size: 70
+2022-12-10 00:55:09,215 INFO [train.py:421] (4/8) Epoch 1, batch 30400, loss[loss=2.423, over 2310.00 frames. , ppl: 11.275565863323513] tot_loss[loss=2.394, over 5471522.93 frames. , ppl: 10.95837172064673], batch size: 70
+2022-12-10 00:56:50,132 INFO [train.py:421] (4/8) Epoch 1, batch 30600, loss[loss=2.283, over 2520.00 frames. , ppl: 9.804368003244559] tot_loss[loss=2.392, over 5559604.27 frames. , ppl: 10.930287184778857], batch size: 70
+2022-12-10 00:58:29,427 INFO [train.py:421] (4/8) Epoch 1, batch 30800, loss[loss=3.037, over 560.00 frames. , ppl: 20.839876403487462] tot_loss[loss=2.391, over 5537163.73 frames. , ppl: 10.92875383127299], batch size: 70
+2022-12-10 01:00:09,321 INFO [train.py:421] (4/8) Epoch 1, batch 31000, loss[loss=2.419, over 1820.00 frames. , ppl: 11.233106584747983] tot_loss[loss=2.391, over 5541343.01 frames. , ppl: 10.920472088172698], batch size: 70
+2022-12-10 01:00:09,322 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:00:10,085 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.373, over 211138.00 frames. , ppl: 10.729127435277624
+2022-12-10 01:01:48,639 INFO [train.py:421] (4/8) Epoch 1, batch 31200, loss[loss=2.423, over 4550.00 frames. , ppl: 11.282263437599772] tot_loss[loss=2.392, over 5504755.69 frames. , ppl: 10.930287966582545], batch size: 70
+2022-12-10 01:03:31,208 INFO [train.py:421] (4/8) Epoch 1, batch 31400, loss[loss=2.806, over 630.00 frames. , ppl: 16.536093553201344] tot_loss[loss=2.391, over 5521969.06 frames. , ppl: 10.921409224770086], batch size: 70
+2022-12-10 01:05:10,005 INFO [train.py:421] (4/8) Epoch 1, batch 31600, loss[loss=2.43, over 3990.00 frames. , ppl: 11.354560379765728] tot_loss[loss=2.391, over 5454818.61 frames. , ppl: 10.928788672696], batch size: 70
+2022-12-10 01:06:49,277 INFO [train.py:421] (4/8) Epoch 1, batch 31800, loss[loss=2.337, over 2870.00 frames. , ppl: 10.354537208500176] tot_loss[loss=2.391, over 5499206.59 frames. , ppl: 10.921018879190314], batch size: 70
+2022-12-10 01:08:31,499 INFO [train.py:421] (4/8) Epoch 1, batch 32000, loss[loss=2.999, over 630.00 frames. , ppl: 20.074452182552523] tot_loss[loss=2.389, over 5530757.86 frames. , ppl: 10.90781761334326], batch size: 70
+2022-12-10 01:08:31,499 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:08:32,284 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.372, over 211138.00 frames. , ppl: 10.718152095318677
+2022-12-10 01:10:13,323 INFO [train.py:421] (4/8) Epoch 1, batch 32200, loss[loss=2.411, over 2940.00 frames. , ppl: 11.146257018494092] tot_loss[loss=2.39, over 5523416.02 frames. , ppl: 10.913119996334212], batch size: 70
+2022-12-10 01:11:53,782 INFO [train.py:421] (4/8) Epoch 1, batch 32400, loss[loss=2.273, over 7280.00 frames. , ppl: 9.709654053010969] tot_loss[loss=2.39, over 5529777.80 frames. , ppl: 10.91327251001037], batch size: 70
+2022-12-10 01:13:33,493 INFO [train.py:421] (4/8) Epoch 1, batch 32600, loss[loss=3.216, over 490.00 frames. , ppl: 24.931721136292445] tot_loss[loss=2.392, over 5471241.56 frames. , ppl: 10.936588573256113], batch size: 70
+2022-12-10 01:15:16,739 INFO [train.py:421] (4/8) Epoch 1, batch 32800, loss[loss=2.481, over 1890.00 frames. , ppl: 11.95423596331584] tot_loss[loss=2.391, over 5504991.91 frames. , ppl: 10.926063610597215], batch size: 70
+2022-12-10 01:16:56,205 INFO [train.py:421] (4/8) Epoch 1, batch 33000, loss[loss=2.77, over 840.00 frames. , ppl: 15.962777454699623] tot_loss[loss=2.391, over 5489713.54 frames. , ppl: 10.929406034952574], batch size: 70
+2022-12-10 01:16:56,205 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:16:56,953 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.371, over 211138.00 frames. , ppl: 10.707864689524108
+2022-12-10 01:18:40,595 INFO [train.py:421] (4/8) Epoch 1, batch 33200, loss[loss=2.316, over 4270.00 frames. , ppl: 10.132133490607314] tot_loss[loss=2.391, over 5502378.56 frames. , ppl: 10.92491047261111], batch size: 70
+2022-12-10 01:20:20,452 INFO [train.py:421] (4/8) Epoch 1, batch 33400, loss[loss=2.542, over 1120.00 frames. , ppl: 12.710114213687833] tot_loss[loss=2.392, over 5485700.51 frames. , ppl: 10.930061343298688], batch size: 70
+2022-12-10 01:21:57,275 INFO [train.py:421] (4/8) Epoch 1, batch 33600, loss[loss=2.426, over 1820.00 frames. , ppl: 11.309223915347381] tot_loss[loss=2.391, over 5500655.49 frames. , ppl: 10.921287953126969], batch size: 70
+2022-12-10 01:23:38,628 INFO [train.py:421] (4/8) Epoch 1, batch 33800, loss[loss=2.328, over 3220.00 frames. , ppl: 10.259040026243758] tot_loss[loss=2.389, over 5521751.91 frames. , ppl: 10.905170771283771], batch size: 70
+2022-12-10 01:25:17,021 INFO [train.py:421] (4/8) Epoch 1, batch 34000, loss[loss=2.561, over 1050.00 frames. , ppl: 12.944940089585385] tot_loss[loss=2.389, over 5538755.35 frames. , ppl: 10.904442731271176], batch size: 70
+2022-12-10 01:25:17,021 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:25:17,782 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.373, over 211138.00 frames. , ppl: 10.73305843282348
+2022-12-10 01:27:03,132 INFO [train.py:421] (4/8) Epoch 1, batch 34200, loss[loss=2.306, over 4480.00 frames. , ppl: 10.039009418204166] tot_loss[loss=2.388, over 5564664.03 frames. , ppl: 10.896572489458752], batch size: 70
+2022-12-10 01:28:41,223 INFO [train.py:421] (4/8) Epoch 1, batch 34400, loss[loss=2.307, over 3430.00 frames. , ppl: 10.045721707396678] tot_loss[loss=2.388, over 5583550.42 frames. , ppl: 10.887171764638444], batch size: 70
+2022-12-10 01:30:22,448 INFO [train.py:421] (4/8) Epoch 1, batch 34600, loss[loss=2.491, over 2100.00 frames. , ppl: 12.076626091066291] tot_loss[loss=2.387, over 5593813.93 frames. , ppl: 10.88398490431259], batch size: 70
+2022-12-10 01:32:03,740 INFO [train.py:421] (4/8) Epoch 1, batch 34800, loss[loss=2.652, over 980.00 frames. , ppl: 14.18306699998132] tot_loss[loss=2.388, over 5572982.18 frames. , ppl: 10.891780256507896], batch size: 70
+2022-12-10 01:33:43,070 INFO [train.py:421] (4/8) Epoch 1, batch 35000, loss[loss=2.755, over 840.00 frames. , ppl: 15.719582475450673] tot_loss[loss=2.388, over 5551359.95 frames. , ppl: 10.896551375794749], batch size: 70
+2022-12-10 01:33:43,070 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:33:43,828 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.372, over 211138.00 frames. , ppl: 10.721890226540799
+2022-12-10 01:35:22,932 INFO [train.py:421] (4/8) Epoch 1, batch 35200, loss[loss=2.519, over 980.00 frames. , ppl: 12.41092160377494] tot_loss[loss=2.388, over 5559454.06 frames. , ppl: 10.892428515126943], batch size: 70
+2022-12-10 01:37:03,047 INFO [train.py:421] (4/8) Epoch 1, batch 35400, loss[loss=3.247, over 490.00 frames. , ppl: 25.707334679106722] tot_loss[loss=2.388, over 5556023.30 frames. , ppl: 10.89116954033829], batch size: 70
+2022-12-10 01:38:46,643 INFO [train.py:421] (4/8) Epoch 1, batch 35600, loss[loss=2.33, over 3640.00 frames. , ppl: 10.282167338757114] tot_loss[loss=2.387, over 5564371.83 frames. , ppl: 10.879213390614952], batch size: 70
+2022-12-10 01:40:24,015 INFO [train.py:421] (4/8) Epoch 1, batch 35800, loss[loss=2.535, over 1050.00 frames. , ppl: 12.614673799155387] tot_loss[loss=2.388, over 5500003.42 frames. , ppl: 10.896648562066641], batch size: 70
+2022-12-10 01:42:02,432 INFO [train.py:421] (4/8) Epoch 1, batch 36000, loss[loss=2.365, over 2520.00 frames. , ppl: 10.641588126267662] tot_loss[loss=2.389, over 5493232.20 frames. , ppl: 10.900826128142384], batch size: 70
+2022-12-10 01:42:02,433 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:42:03,181 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.371, over 211138.00 frames. , ppl: 10.709730213282327
+2022-12-10 01:43:40,117 INFO [train.py:421] (4/8) Epoch 1, batch 36200, loss[loss=2.447, over 1750.00 frames. , ppl: 11.559373332525976] tot_loss[loss=2.389, over 5484858.28 frames. , ppl: 10.899610247258883], batch size: 70
+2022-12-10 01:45:21,067 INFO [train.py:421] (4/8) Epoch 1, batch 36400, loss[loss=2.323, over 6160.00 frames. , ppl: 10.20254851119819] tot_loss[loss=2.388, over 5495785.50 frames. , ppl: 10.895336348041658], batch size: 70
+2022-12-10 01:47:02,481 INFO [train.py:421] (4/8) Epoch 1, batch 36600, loss[loss=2.309, over 3780.00 frames. , ppl: 10.066457738413002] tot_loss[loss=2.388, over 5520559.92 frames. , ppl: 10.890420016886733], batch size: 70
+2022-12-10 01:48:44,159 INFO [train.py:421] (4/8) Epoch 1, batch 36800, loss[loss=2.462, over 1540.00 frames. , ppl: 11.72953115423634] tot_loss[loss=2.388, over 5500884.02 frames. , ppl: 10.894137298487587], batch size: 70
+2022-12-10 01:50:20,592 INFO [train.py:421] (4/8) Epoch 1, batch 37000, loss[loss=2.498, over 1120.00 frames. , ppl: 12.162503293840789] tot_loss[loss=2.388, over 5503344.39 frames. , ppl: 10.891082175355464], batch size: 70
+2022-12-10 01:50:20,592 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:50:21,353 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.371, over 211138.00 frames. , ppl: 10.706359193662431
+2022-12-10 01:52:04,207 INFO [train.py:421] (4/8) Epoch 1, batch 37200, loss[loss=2.306, over 13370.00 frames. , ppl: 10.035373369621263] tot_loss[loss=2.387, over 5516997.74 frames. , ppl: 10.885652566314173], batch size: 70
+2022-12-10 01:53:44,571 INFO [train.py:421] (4/8) Epoch 1, batch 37400, loss[loss=2.382, over 2100.00 frames. , ppl: 10.827270386290007] tot_loss[loss=2.388, over 5509393.68 frames. , ppl: 10.889153707221315], batch size: 70
+2022-12-10 01:55:24,444 INFO [train.py:421] (4/8) Epoch 1, batch 37600, loss[loss=2.349, over 2450.00 frames. , ppl: 10.47357302283877] tot_loss[loss=2.388, over 5483222.10 frames. , ppl: 10.889546841920968], batch size: 70
+2022-12-10 01:57:05,223 INFO [train.py:421] (4/8) Epoch 1, batch 37800, loss[loss=2.4, over 3430.00 frames. , ppl: 11.01977798308859] tot_loss[loss=2.389, over 5447367.21 frames. , ppl: 10.905587056509638], batch size: 70
+2022-12-10 01:58:47,483 INFO [train.py:421] (4/8) Epoch 1, batch 38000, loss[loss=2.325, over 3010.00 frames. , ppl: 10.221611533544184] tot_loss[loss=2.388, over 5461414.81 frames. , ppl: 10.891910988492624], batch size: 70
+2022-12-10 01:58:47,483 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 01:58:48,230 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.367, over 211138.00 frames. , ppl: 10.660825546799813
+2022-12-10 02:00:27,318 INFO [train.py:421] (4/8) Epoch 1, batch 38200, loss[loss=2.41, over 3150.00 frames. , ppl: 11.13292394284493] tot_loss[loss=2.387, over 5467625.08 frames. , ppl: 10.884051938707007], batch size: 70
+2022-12-10 02:02:09,572 INFO [train.py:421] (4/8) Epoch 1, batch 38400, loss[loss=2.697, over 1050.00 frames. , ppl: 14.836347440899473] tot_loss[loss=2.388, over 5474657.73 frames. , ppl: 10.887762451306925], batch size: 70
+2022-12-10 02:03:49,946 INFO [train.py:421] (4/8) Epoch 1, batch 38600, loss[loss=2.495, over 1820.00 frames. , ppl: 12.11932980739012] tot_loss[loss=2.387, over 5477375.12 frames. , ppl: 10.880749404233637], batch size: 70
+2022-12-10 02:05:31,055 INFO [train.py:421] (4/8) Epoch 1, batch 38800, loss[loss=2.476, over 1050.00 frames. , ppl: 11.891103917565394] tot_loss[loss=2.388, over 5434629.01 frames. , ppl: 10.89356301037895], batch size: 70
+2022-12-10 02:07:09,882 INFO [train.py:421] (4/8) Epoch 1, batch 39000, loss[loss=2.754, over 770.00 frames. , ppl: 15.712272440584462] tot_loss[loss=2.389, over 5423001.12 frames. , ppl: 10.900513425774752], batch size: 70
+2022-12-10 02:07:09,882 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:07:10,643 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.368, over 211138.00 frames. , ppl: 10.674933246579991
+2022-12-10 02:08:52,861 INFO [train.py:421] (4/8) Epoch 1, batch 39200, loss[loss=2.606, over 1120.00 frames. , ppl: 13.549729997521199] tot_loss[loss=2.39, over 5364607.58 frames. , ppl: 10.909176488997689], batch size: 70
+2022-12-10 02:10:35,768 INFO [train.py:421] (4/8) Epoch 1, batch 39400, loss[loss=2.465, over 1540.00 frames. , ppl: 11.766342135142922] tot_loss[loss=2.389, over 5353256.00 frames. , ppl: 10.903470470721874], batch size: 70
+2022-12-10 02:12:14,145 INFO [train.py:421] (4/8) Epoch 1, batch 39600, loss[loss=2.557, over 1260.00 frames. , ppl: 12.903115134386319] tot_loss[loss=2.388, over 5385495.09 frames. , ppl: 10.890371072831012], batch size: 70
+2022-12-10 02:13:56,725 INFO [train.py:421] (4/8) Epoch 1, batch 39800, loss[loss=2.515, over 1190.00 frames. , ppl: 12.364301215710922] tot_loss[loss=2.387, over 5408313.74 frames. , ppl: 10.87680613664702], batch size: 70
+2022-12-10 02:15:38,369 INFO [train.py:421] (4/8) Epoch 1, batch 40000, loss[loss=3.413, over 490.00 frames. , ppl: 30.363300659190944] tot_loss[loss=2.386, over 5454834.43 frames. , ppl: 10.86648383226236], batch size: 70
+2022-12-10 02:15:38,369 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:15:39,131 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.368, over 211138.00 frames. , ppl: 10.68034759574071
+2022-12-10 02:17:21,026 INFO [train.py:421] (4/8) Epoch 1, batch 40200, loss[loss=2.394, over 4270.00 frames. , ppl: 10.951821013334786] tot_loss[loss=2.385, over 5501922.80 frames. , ppl: 10.860877370383875], batch size: 70
+2022-12-10 02:19:03,094 INFO [train.py:421] (4/8) Epoch 1, batch 40400, loss[loss=2.366, over 2100.00 frames. , ppl: 10.656479967320028] tot_loss[loss=2.385, over 5452374.38 frames. , ppl: 10.863631149686075], batch size: 70
+2022-12-10 02:20:42,139 INFO [train.py:421] (4/8) Epoch 1, batch 40600, loss[loss=2.35, over 4970.00 frames. , ppl: 10.482801013818682] tot_loss[loss=2.385, over 5476418.11 frames. , ppl: 10.856132444569615], batch size: 70
+2022-12-10 02:22:24,726 INFO [train.py:421] (4/8) Epoch 1, batch 40800, loss[loss=2.3, over 4270.00 frames. , ppl: 9.97082064020566] tot_loss[loss=2.384, over 5524379.46 frames. , ppl: 10.848648478852676], batch size: 70
+2022-12-10 02:24:04,605 INFO [train.py:421] (4/8) Epoch 1, batch 41000, loss[loss=2.399, over 2310.00 frames. , ppl: 11.007192202878981] tot_loss[loss=2.385, over 5532573.13 frames. , ppl: 10.854199150868414], batch size: 70
+2022-12-10 02:24:04,606 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:24:05,368 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.367, over 211138.00 frames. , ppl: 10.668684715535875
+2022-12-10 02:25:40,536 INFO [train.py:421] (4/8) Epoch 1, batch 41200, loss[loss=2.459, over 1960.00 frames. , ppl: 11.697440036195607] tot_loss[loss=2.384, over 5533098.49 frames. , ppl: 10.852944494300141], batch size: 70
+2022-12-10 02:27:22,733 INFO [train.py:421] (4/8) Epoch 1, batch 41400, loss[loss=2.485, over 2100.00 frames. , ppl: 12.002727655358491] tot_loss[loss=2.385, over 5526206.36 frames. , ppl: 10.853749511511792], batch size: 70
+2022-12-10 02:29:04,770 INFO [train.py:421] (4/8) Epoch 1, batch 41600, loss[loss=2.576, over 1680.00 frames. , ppl: 13.141522715806648] tot_loss[loss=2.384, over 5526715.24 frames. , ppl: 10.852712131903457], batch size: 70
+2022-12-10 02:30:42,769 INFO [train.py:421] (4/8) Epoch 1, batch 41800, loss[loss=2.64, over 840.00 frames. , ppl: 14.015559541014493] tot_loss[loss=2.385, over 5506686.73 frames. , ppl: 10.859967383988884], batch size: 70
+2022-12-10 02:32:23,880 INFO [train.py:421] (4/8) Epoch 1, batch 42000, loss[loss=2.388, over 1680.00 frames. , ppl: 10.89466382451886] tot_loss[loss=2.385, over 5518758.00 frames. , ppl: 10.854429452990834], batch size: 70
+2022-12-10 02:32:23,880 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:32:24,639 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.364, over 211138.00 frames. , ppl: 10.63275418399596
+2022-12-10 02:34:04,507 INFO [train.py:421] (4/8) Epoch 1, batch 42200, loss[loss=2.322, over 4270.00 frames. , ppl: 10.197793764509555] tot_loss[loss=2.384, over 5512671.26 frames. , ppl: 10.853133806289389], batch size: 70
+2022-12-10 02:35:44,518 INFO [train.py:421] (4/8) Epoch 1, batch 42400, loss[loss=2.467, over 1400.00 frames. , ppl: 11.792450545123273] tot_loss[loss=2.385, over 5473415.71 frames. , ppl: 10.863494917777999], batch size: 70
+2022-12-10 02:37:26,135 INFO [train.py:421] (4/8) Epoch 1, batch 42600, loss[loss=2.432, over 2170.00 frames. , ppl: 11.38494697078548] tot_loss[loss=2.385, over 5462908.64 frames. , ppl: 10.859197610459352], batch size: 70
+2022-12-10 02:39:06,181 INFO [train.py:421] (4/8) Epoch 1, batch 42800, loss[loss=2.29, over 4480.00 frames. , ppl: 9.87328723182059] tot_loss[loss=2.385, over 5499715.40 frames. , ppl: 10.855558407383048], batch size: 70
+2022-12-10 02:40:45,345 INFO [train.py:421] (4/8) Epoch 1, batch 43000, loss[loss=2.863, over 700.00 frames. , ppl: 17.51031574795875] tot_loss[loss=2.385, over 5457956.42 frames. , ppl: 10.863716949612355], batch size: 70
+2022-12-10 02:40:45,345 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:40:46,105 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.364, over 211138.00 frames. , ppl: 10.630343507571958
+2022-12-10 02:42:23,863 INFO [train.py:421] (4/8) Epoch 1, batch 43200, loss[loss=2.693, over 770.00 frames. , ppl: 14.776735830450555] tot_loss[loss=2.386, over 5434770.49 frames. , ppl: 10.868881296251807], batch size: 70
+2022-12-10 02:44:06,618 INFO [train.py:421] (4/8) Epoch 1, batch 43400, loss[loss=2.408, over 3220.00 frames. , ppl: 11.112668407788814] tot_loss[loss=2.385, over 5463418.48 frames. , ppl: 10.854758022217812], batch size: 70
+2022-12-10 02:45:46,629 INFO [train.py:421] (4/8) Epoch 1, batch 43600, loss[loss=2.466, over 1680.00 frames. , ppl: 11.772752608551817] tot_loss[loss=2.386, over 5424551.94 frames. , ppl: 10.864880947650354], batch size: 70
+2022-12-10 02:47:26,684 INFO [train.py:421] (4/8) Epoch 1, batch 43800, loss[loss=2.417, over 3360.00 frames. , ppl: 11.211854960528045] tot_loss[loss=2.384, over 5435775.95 frames. , ppl: 10.852732722578214], batch size: 70
+2022-12-10 02:49:05,961 INFO [train.py:421] (4/8) Epoch 1, batch 44000, loss[loss=2.283, over 5740.00 frames. , ppl: 9.806648300514656] tot_loss[loss=2.385, over 5416317.03 frames. , ppl: 10.85744563273759], batch size: 70
+2022-12-10 02:49:05,961 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:49:06,722 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.365, over 211138.00 frames. , ppl: 10.648347077171886
+2022-12-10 02:50:46,016 INFO [train.py:421] (4/8) Epoch 1, batch 44200, loss[loss=2.49, over 840.00 frames. , ppl: 12.058308701533226] tot_loss[loss=2.385, over 5392688.35 frames. , ppl: 10.863355858521569], batch size: 70
+2022-12-10 02:52:25,664 INFO [train.py:421] (4/8) Epoch 1, batch 44400, loss[loss=2.382, over 3640.00 frames. , ppl: 10.825533643224993] tot_loss[loss=2.385, over 5393769.34 frames. , ppl: 10.862316791204066], batch size: 70
+2022-12-10 02:54:07,177 INFO [train.py:421] (4/8) Epoch 1, batch 44600, loss[loss=2.618, over 1260.00 frames. , ppl: 13.710826330915433] tot_loss[loss=2.385, over 5399177.69 frames. , ppl: 10.860928367651603], batch size: 70
+2022-12-10 02:55:46,318 INFO [train.py:421] (4/8) Epoch 1, batch 44800, loss[loss=2.606, over 910.00 frames. , ppl: 13.546717880693393] tot_loss[loss=2.385, over 5397749.20 frames. , ppl: 10.858230950219088], batch size: 70
+2022-12-10 02:57:25,620 INFO [train.py:421] (4/8) Epoch 1, batch 45000, loss[loss=2.573, over 1190.00 frames. , ppl: 13.101162607341742] tot_loss[loss=2.385, over 5391278.15 frames. , ppl: 10.85845396196222], batch size: 70
+2022-12-10 02:57:25,620 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 02:57:26,378 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.364, over 211138.00 frames. , ppl: 10.628203939461411
+2022-12-10 02:59:10,686 INFO [train.py:421] (4/8) Epoch 1, batch 45200, loss[loss=2.399, over 2100.00 frames. , ppl: 11.016832088944309] tot_loss[loss=2.385, over 5385571.52 frames. , ppl: 10.86132990836274], batch size: 70
+2022-12-10 03:00:53,880 INFO [train.py:421] (4/8) Epoch 1, batch 45400, loss[loss=2.403, over 1960.00 frames. , ppl: 11.052621172476616] tot_loss[loss=2.384, over 5404480.35 frames. , ppl: 10.85261453093706], batch size: 70
+2022-12-10 03:02:33,640 INFO [train.py:421] (4/8) Epoch 1, batch 45600, loss[loss=2.365, over 3010.00 frames. , ppl: 10.639187627959451] tot_loss[loss=2.383, over 5413967.96 frames. , ppl: 10.838357261263104], batch size: 70
+2022-12-10 03:04:17,137 INFO [train.py:421] (4/8) Epoch 1, batch 45800, loss[loss=2.547, over 980.00 frames. , ppl: 12.768957118155036] tot_loss[loss=2.382, over 5439841.81 frames. , ppl: 10.827738495395328], batch size: 70
+2022-12-10 03:05:58,452 INFO [train.py:421] (4/8) Epoch 1, batch 46000, loss[loss=2.517, over 1120.00 frames. , ppl: 12.389530564178003] tot_loss[loss=2.382, over 5462907.55 frames. , ppl: 10.823613735239745], batch size: 70
+2022-12-10 03:05:58,453 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:05:59,200 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.363, over 211138.00 frames. , ppl: 10.624262598332244
+2022-12-10 03:07:41,815 INFO [train.py:421] (4/8) Epoch 1, batch 46200, loss[loss=2.369, over 3990.00 frames. , ppl: 10.689280993354442] tot_loss[loss=2.381, over 5489366.48 frames. , ppl: 10.814116676251839], batch size: 70
+2022-12-10 03:09:19,254 INFO [train.py:421] (4/8) Epoch 1, batch 46400, loss[loss=2.386, over 2940.00 frames. , ppl: 10.867179262217586] tot_loss[loss=2.381, over 5464067.51 frames. , ppl: 10.820331313271609], batch size: 70
+2022-12-10 03:11:00,986 INFO [train.py:421] (4/8) Epoch 1, batch 46600, loss[loss=2.374, over 3780.00 frames. , ppl: 10.739168026268004] tot_loss[loss=2.382, over 5455361.25 frames. , ppl: 10.82186266810536], batch size: 70
+2022-12-10 03:12:37,299 INFO [train.py:421] (4/8) Epoch 1, batch 46800, loss[loss=2.535, over 1120.00 frames. , ppl: 12.622358292695514] tot_loss[loss=2.381, over 5447639.18 frames. , ppl: 10.821096190771744], batch size: 70
+2022-12-10 03:14:18,442 INFO [train.py:421] (4/8) Epoch 1, batch 47000, loss[loss=2.327, over 3640.00 frames. , ppl: 10.246860525524609] tot_loss[loss=2.381, over 5470839.17 frames. , ppl: 10.81280932233049], batch size: 70
+2022-12-10 03:14:18,443 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:14:19,231 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.363, over 211138.00 frames. , ppl: 10.618849962774785
+2022-12-10 03:15:58,272 INFO [train.py:421] (4/8) Epoch 1, batch 47200, loss[loss=2.393, over 1610.00 frames. , ppl: 10.950912220254587] tot_loss[loss=2.381, over 5484264.88 frames. , ppl: 10.812916261030235], batch size: 70
+2022-12-10 03:17:35,094 INFO [train.py:421] (4/8) Epoch 1, batch 47400, loss[loss=2.398, over 2240.00 frames. , ppl: 10.998339589669778] tot_loss[loss=2.38, over 5499231.40 frames. , ppl: 10.805029303053217], batch size: 70
+2022-12-10 03:19:18,529 INFO [train.py:421] (4/8) Epoch 1, batch 47600, loss[loss=2.318, over 3640.00 frames. , ppl: 10.152819376650411] tot_loss[loss=2.38, over 5503930.19 frames. , ppl: 10.801048479799555], batch size: 70
+2022-12-10 03:20:57,306 INFO [train.py:421] (4/8) Epoch 1, batch 47800, loss[loss=2.354, over 10220.00 frames. , ppl: 10.523054974304792] tot_loss[loss=2.38, over 5504280.93 frames. , ppl: 10.802256205967348], batch size: 70
+2022-12-10 03:22:37,620 INFO [train.py:421] (4/8) Epoch 1, batch 48000, loss[loss=2.528, over 1050.00 frames. , ppl: 12.532096937512613] tot_loss[loss=2.379, over 5516019.46 frames. , ppl: 10.794219334492109], batch size: 70
+2022-12-10 03:22:37,621 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:22:38,366 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.361, over 211138.00 frames. , ppl: 10.600952749654093
+2022-12-10 03:24:15,508 INFO [train.py:421] (4/8) Epoch 1, batch 48200, loss[loss=2.304, over 3850.00 frames. , ppl: 10.01372066991439] tot_loss[loss=2.379, over 5508929.41 frames. , ppl: 10.796341316921025], batch size: 70
+2022-12-10 03:25:55,537 INFO [train.py:421] (4/8) Epoch 1, batch 48400, loss[loss=2.407, over 1610.00 frames. , ppl: 11.100404088601307] tot_loss[loss=2.379, over 5535249.02 frames. , ppl: 10.791805872584444], batch size: 70
+2022-12-10 03:27:34,551 INFO [train.py:421] (4/8) Epoch 1, batch 48600, loss[loss=2.318, over 7210.00 frames. , ppl: 10.158776123421456] tot_loss[loss=2.378, over 5543388.53 frames. , ppl: 10.783112845305567], batch size: 70
+2022-12-10 03:29:11,885 INFO [train.py:421] (4/8) Epoch 1, batch 48800, loss[loss=2.331, over 3010.00 frames. , ppl: 10.286714524312602] tot_loss[loss=2.378, over 5538922.65 frames. , ppl: 10.782411887048909], batch size: 70
+2022-12-10 03:30:50,523 INFO [train.py:421] (4/8) Epoch 1, batch 49000, loss[loss=2.45, over 2310.00 frames. , ppl: 11.592523889639917] tot_loss[loss=2.378, over 5563175.80 frames. , ppl: 10.778279138935163], batch size: 70
+2022-12-10 03:30:50,524 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:30:51,283 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.36, over 211138.00 frames. , ppl: 10.585974491786418
+2022-12-10 03:32:29,119 INFO [train.py:421] (4/8) Epoch 1, batch 49200, loss[loss=2.983, over 560.00 frames. , ppl: 19.749141980670306] tot_loss[loss=2.377, over 5585297.71 frames. , ppl: 10.769760787751832], batch size: 70
+2022-12-10 03:34:05,518 INFO [train.py:421] (4/8) Epoch 1, batch 49400, loss[loss=2.421, over 2940.00 frames. , ppl: 11.262102776346776] tot_loss[loss=2.378, over 5535516.43 frames. , ppl: 10.782884728315818], batch size: 70
+2022-12-10 03:35:47,737 INFO [train.py:421] (4/8) Epoch 1, batch 49600, loss[loss=2.372, over 4200.00 frames. , ppl: 10.71573194738782] tot_loss[loss=2.377, over 5578648.95 frames. , ppl: 10.768591171967204], batch size: 70
+2022-12-10 03:37:28,142 INFO [train.py:421] (4/8) Epoch 1, batch 49800, loss[loss=2.331, over 2380.00 frames. , ppl: 10.290229079180003] tot_loss[loss=2.376, over 5579534.68 frames. , ppl: 10.76703392964804], batch size: 70
+2022-12-10 03:39:06,578 INFO [train.py:421] (4/8) Epoch 1, batch 50000, loss[loss=2.292, over 4900.00 frames. , ppl: 9.893001292255406] tot_loss[loss=2.377, over 5561745.15 frames. , ppl: 10.771823582184274], batch size: 70
+2022-12-10 03:39:06,578 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:39:07,339 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.359, over 211138.00 frames. , ppl: 10.58257820728752
+2022-12-10 03:40:43,902 INFO [train.py:421] (4/8) Epoch 1, batch 50200, loss[loss=2.706, over 840.00 frames. , ppl: 14.962537077675139] tot_loss[loss=2.377, over 5568931.42 frames. , ppl: 10.769850276130654], batch size: 70
+2022-12-10 03:42:22,998 INFO [train.py:421] (4/8) Epoch 1, batch 50400, loss[loss=2.539, over 2100.00 frames. , ppl: 12.673280857552626] tot_loss[loss=2.377, over 5567864.25 frames. , ppl: 10.770071300899112], batch size: 70
+2022-12-10 03:44:03,135 INFO [train.py:421] (4/8) Epoch 1, batch 50600, loss[loss=2.273, over 4620.00 frames. , ppl: 9.711238555482291] tot_loss[loss=2.376, over 5540281.26 frames. , ppl: 10.766370688076417], batch size: 70
+2022-12-10 03:45:44,370 INFO [train.py:421] (4/8) Epoch 1, batch 50800, loss[loss=2.367, over 2870.00 frames. , ppl: 10.666936171018106] tot_loss[loss=2.376, over 5568600.10 frames. , ppl: 10.76209684260263], batch size: 70
+2022-12-10 03:47:26,021 INFO [train.py:421] (4/8) Epoch 1, batch 51000, loss[loss=2.474, over 1960.00 frames. , ppl: 11.866623616743015] tot_loss[loss=2.375, over 5579962.45 frames. , ppl: 10.754435605876944], batch size: 70
+2022-12-10 03:47:26,021 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:47:26,769 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.361, over 211138.00 frames. , ppl: 10.598020656585552
+2022-12-10 03:49:07,095 INFO [train.py:421] (4/8) Epoch 1, batch 51200, loss[loss=2.708, over 770.00 frames. , ppl: 14.999991961147572] tot_loss[loss=2.376, over 5545263.12 frames. , ppl: 10.75833577077368], batch size: 70
+2022-12-10 03:50:49,015 INFO [train.py:421] (4/8) Epoch 1, batch 51400, loss[loss=2.375, over 3010.00 frames. , ppl: 10.754199993738307] tot_loss[loss=2.375, over 5557141.21 frames. , ppl: 10.754956384702181], batch size: 70
+2022-12-10 03:52:29,258 INFO [train.py:421] (4/8) Epoch 1, batch 51600, loss[loss=2.288, over 5250.00 frames. , ppl: 9.858152078920968] tot_loss[loss=2.375, over 5574324.99 frames. , ppl: 10.75281253919359], batch size: 70
+2022-12-10 03:54:10,138 INFO [train.py:421] (4/8) Epoch 1, batch 51800, loss[loss=2.529, over 1190.00 frames. , ppl: 12.538203729010906] tot_loss[loss=2.375, over 5601660.25 frames. , ppl: 10.748721092445049], batch size: 70
+2022-12-10 03:55:47,567 INFO [train.py:421] (4/8) Epoch 1, batch 52000, loss[loss=2.794, over 700.00 frames. , ppl: 16.354395866450737] tot_loss[loss=2.375, over 5568053.31 frames. , ppl: 10.747604128889117], batch size: 70
+2022-12-10 03:55:47,568 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 03:55:48,316 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.359, over 211138.00 frames. , ppl: 10.580972870980991
+2022-12-10 03:57:25,692 INFO [train.py:421] (4/8) Epoch 1, batch 52200, loss[loss=2.449, over 1680.00 frames. , ppl: 11.580138251613365] tot_loss[loss=2.376, over 5486613.41 frames. , ppl: 10.763695150676828], batch size: 70
+2022-12-10 03:59:05,788 INFO [train.py:421] (4/8) Epoch 1, batch 52400, loss[loss=2.554, over 1190.00 frames. , ppl: 12.86439560800429] tot_loss[loss=2.378, over 5441944.93 frames. , ppl: 10.780331267823758], batch size: 70
+2022-12-10 04:00:46,728 INFO [train.py:421] (4/8) Epoch 1, batch 52600, loss[loss=2.447, over 1680.00 frames. , ppl: 11.55498349658126] tot_loss[loss=2.376, over 5487267.46 frames. , ppl: 10.764321545075692], batch size: 70
+2022-12-10 04:02:26,556 INFO [train.py:421] (4/8) Epoch 1, batch 52800, loss[loss=2.465, over 2870.00 frames. , ppl: 11.766219523008237] tot_loss[loss=2.376, over 5529283.15 frames. , ppl: 10.757738356144888], batch size: 70
+2022-12-10 04:04:06,247 INFO [train.py:421] (4/8) Epoch 1, batch 53000, loss[loss=2.429, over 3500.00 frames. , ppl: 11.345306446571227] tot_loss[loss=2.374, over 5548690.85 frames. , ppl: 10.745609083176879], batch size: 70
+2022-12-10 04:04:06,247 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:04:07,006 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.358, over 211138.00 frames. , ppl: 10.567887116951528
+2022-12-10 04:05:52,840 INFO [train.py:421] (4/8) Epoch 1, batch 53200, loss[loss=2.377, over 3220.00 frames. , ppl: 10.767189620943912] tot_loss[loss=2.373, over 5574026.87 frames. , ppl: 10.732951255722261], batch size: 70
+2022-12-10 04:07:33,562 INFO [train.py:421] (4/8) Epoch 1, batch 53400, loss[loss=2.277, over 8260.00 frames. , ppl: 9.747020391272915] tot_loss[loss=2.373, over 5576742.58 frames. , ppl: 10.726406415386911], batch size: 70
+2022-12-10 04:09:19,589 INFO [train.py:421] (4/8) Epoch 1, batch 53600, loss[loss=2.249, over 6230.00 frames. , ppl: 9.474679619475836] tot_loss[loss=2.373, over 5598800.07 frames. , ppl: 10.724761730508586], batch size: 70
+2022-12-10 04:10:59,120 INFO [train.py:421] (4/8) Epoch 1, batch 53800, loss[loss=2.314, over 4130.00 frames. , ppl: 10.117171886205801] tot_loss[loss=2.373, over 5612220.11 frames. , ppl: 10.724226606000162], batch size: 70
+2022-12-10 04:12:36,568 INFO [train.py:421] (4/8) Epoch 1, batch 54000, loss[loss=2.458, over 1190.00 frames. , ppl: 11.676499207218585] tot_loss[loss=2.373, over 5584716.22 frames. , ppl: 10.733687978168383], batch size: 70
+2022-12-10 04:12:36,568 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:12:37,329 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.356, over 211138.00 frames. , ppl: 10.55320704955568
+2022-12-10 04:14:15,872 INFO [train.py:421] (4/8) Epoch 1, batch 54200, loss[loss=2.332, over 3640.00 frames. , ppl: 10.301514788183557] tot_loss[loss=2.374, over 5542147.70 frames. , ppl: 10.744550815548598], batch size: 70
+2022-12-10 04:15:51,797 INFO [train.py:421] (4/8) Epoch 1, batch 54400, loss[loss=2.396, over 2450.00 frames. , ppl: 10.983948161287246] tot_loss[loss=2.376, over 5502161.22 frames. , ppl: 10.76006496737798], batch size: 70
+2022-12-10 04:17:31,785 INFO [train.py:421] (4/8) Epoch 1, batch 54600, loss[loss=2.558, over 980.00 frames. , ppl: 12.905721613484184] tot_loss[loss=2.376, over 5477483.95 frames. , ppl: 10.761738866259414], batch size: 70
+2022-12-10 04:19:11,953 INFO [train.py:421] (4/8) Epoch 1, batch 54800, loss[loss=2.452, over 1540.00 frames. , ppl: 11.610037826191054] tot_loss[loss=2.376, over 5487627.69 frames. , ppl: 10.762408498812961], batch size: 70
+2022-12-10 04:20:51,927 INFO [train.py:421] (4/8) Epoch 1, batch 55000, loss[loss=2.492, over 1050.00 frames. , ppl: 12.090634874278397] tot_loss[loss=2.375, over 5486930.10 frames. , ppl: 10.754808473469907], batch size: 70
+2022-12-10 04:20:51,927 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:20:52,675 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.355, over 211138.00 frames. , ppl: 10.542236908630983
+2022-12-10 04:22:35,395 INFO [train.py:421] (4/8) Epoch 1, batch 55200, loss[loss=2.405, over 3640.00 frames. , ppl: 11.074261675463251] tot_loss[loss=2.373, over 5538106.79 frames. , ppl: 10.73470363825005], batch size: 70
+2022-12-10 04:24:13,633 INFO [train.py:421] (4/8) Epoch 1, batch 55400, loss[loss=2.66, over 1680.00 frames. , ppl: 14.293939984052557] tot_loss[loss=2.374, over 5518397.00 frames. , ppl: 10.74544647907056], batch size: 70
+2022-12-10 04:25:54,098 INFO [train.py:421] (4/8) Epoch 1, batch 55600, loss[loss=2.349, over 3150.00 frames. , ppl: 10.472032659222153] tot_loss[loss=2.374, over 5505461.09 frames. , ppl: 10.740896176700653], batch size: 70
+2022-12-10 04:27:33,001 INFO [train.py:421] (4/8) Epoch 1, batch 55800, loss[loss=2.586, over 980.00 frames. , ppl: 13.279430628033092] tot_loss[loss=2.374, over 5504240.32 frames. , ppl: 10.742008045313852], batch size: 70
+2022-12-10 04:29:13,169 INFO [train.py:421] (4/8) Epoch 1, batch 56000, loss[loss=2.518, over 1120.00 frames. , ppl: 12.400465132973345] tot_loss[loss=2.375, over 5497097.78 frames. , ppl: 10.751710238461255], batch size: 70
+2022-12-10 04:29:13,169 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:29:13,929 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.355, over 211138.00 frames. , ppl: 10.539062115723747
+2022-12-10 04:30:54,312 INFO [train.py:421] (4/8) Epoch 1, batch 56200, loss[loss=2.339, over 3780.00 frames. , ppl: 10.365936897246032] tot_loss[loss=2.375, over 5489001.97 frames. , ppl: 10.750641033002797], batch size: 70
+2022-12-10 04:32:32,226 INFO [train.py:421] (4/8) Epoch 1, batch 56400, loss[loss=3.278, over 490.00 frames. , ppl: 26.530113672179784] tot_loss[loss=2.376, over 5485794.61 frames. , ppl: 10.758655128972359], batch size: 70
+2022-12-10 04:34:13,605 INFO [train.py:421] (4/8) Epoch 1, batch 56600, loss[loss=2.254, over 3360.00 frames. , ppl: 9.529535583938394] tot_loss[loss=2.374, over 5511821.15 frames. , ppl: 10.73892205271816], batch size: 70
+2022-12-10 04:35:52,337 INFO [train.py:421] (4/8) Epoch 1, batch 56800, loss[loss=2.409, over 2030.00 frames. , ppl: 11.118938149433065] tot_loss[loss=2.374, over 5506577.56 frames. , ppl: 10.734910401016817], batch size: 70
+2022-12-10 04:37:33,496 INFO [train.py:421] (4/8) Epoch 1, batch 57000, loss[loss=2.259, over 3430.00 frames. , ppl: 9.572953806336736] tot_loss[loss=2.374, over 5482413.80 frames. , ppl: 10.738398290414908], batch size: 70
+2022-12-10 04:37:33,496 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:37:34,257 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.354, over 211138.00 frames. , ppl: 10.531887643629647
+2022-12-10 04:39:15,037 INFO [train.py:421] (4/8) Epoch 1, batch 57200, loss[loss=2.373, over 2170.00 frames. , ppl: 10.730757539666818] tot_loss[loss=2.374, over 5503797.55 frames. , ppl: 10.73652823910075], batch size: 70
+2022-12-10 04:40:55,985 INFO [train.py:421] (4/8) Epoch 1, batch 57400, loss[loss=2.613, over 980.00 frames. , ppl: 13.643310963860978] tot_loss[loss=2.375, over 5477641.86 frames. , ppl: 10.746332989058919], batch size: 70
+2022-12-10 04:42:35,563 INFO [train.py:421] (4/8) Epoch 1, batch 57600, loss[loss=2.73, over 700.00 frames. , ppl: 15.331908424489296] tot_loss[loss=2.375, over 5434707.97 frames. , ppl: 10.756009577941606], batch size: 70
+2022-12-10 04:44:14,443 INFO [train.py:421] (4/8) Epoch 1, batch 57800, loss[loss=2.253, over 4760.00 frames. , ppl: 9.511622227852424] tot_loss[loss=2.375, over 5452733.37 frames. , ppl: 10.751840767545128], batch size: 70
+2022-12-10 04:45:53,663 INFO [train.py:421] (4/8) Epoch 1, batch 58000, loss[loss=2.537, over 1050.00 frames. , ppl: 12.640731933262565] tot_loss[loss=2.375, over 5466679.14 frames. , ppl: 10.746832444465717], batch size: 70
+2022-12-10 04:45:53,664 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:45:54,423 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.353, over 211138.00 frames. , ppl: 10.515346212674748
+2022-12-10 04:47:35,224 INFO [train.py:421] (4/8) Epoch 1, batch 58200, loss[loss=2.297, over 3360.00 frames. , ppl: 9.941036181384687] tot_loss[loss=2.375, over 5458509.78 frames. , ppl: 10.75033616733558], batch size: 70
+2022-12-10 04:49:15,814 INFO [train.py:421] (4/8) Epoch 1, batch 58400, loss[loss=2.302, over 4410.00 frames. , ppl: 9.996794318966245] tot_loss[loss=2.375, over 5468143.11 frames. , ppl: 10.748488694664276], batch size: 70
+2022-12-10 04:50:58,466 INFO [train.py:421] (4/8) Epoch 1, batch 58600, loss[loss=2.309, over 6790.00 frames. , ppl: 10.065191024679848] tot_loss[loss=2.374, over 5475345.10 frames. , ppl: 10.739901686284352], batch size: 70
+2022-12-10 04:52:37,254 INFO [train.py:421] (4/8) Epoch 1, batch 58800, loss[loss=2.393, over 3430.00 frames. , ppl: 10.942152581190092] tot_loss[loss=2.374, over 5447873.90 frames. , ppl: 10.739210891831664], batch size: 70
+2022-12-10 04:54:20,790 INFO [train.py:421] (4/8) Epoch 1, batch 59000, loss[loss=2.269, over 6440.00 frames. , ppl: 9.671209341088678] tot_loss[loss=2.374, over 5472424.39 frames. , ppl: 10.73536071285044], batch size: 70
+2022-12-10 04:54:20,791 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 04:54:21,536 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.352, over 211138.00 frames. , ppl: 10.503966120477648
+2022-12-10 04:55:59,455 INFO [train.py:421] (4/8) Epoch 1, batch 59200, loss[loss=2.933, over 560.00 frames. , ppl: 18.791723434675664] tot_loss[loss=2.374, over 5440694.59 frames. , ppl: 10.74310672205456], batch size: 70
+2022-12-10 04:57:42,858 INFO [train.py:421] (4/8) Epoch 1, batch 59400, loss[loss=2.402, over 1960.00 frames. , ppl: 11.04406121866418] tot_loss[loss=2.374, over 5450599.24 frames. , ppl: 10.74024864668901], batch size: 70
+2022-12-10 04:59:24,822 INFO [train.py:421] (4/8) Epoch 1, batch 59600, loss[loss=2.406, over 3290.00 frames. , ppl: 11.087928570584438] tot_loss[loss=2.375, over 5416384.69 frames. , ppl: 10.746060972291236], batch size: 70
+2022-12-10 05:01:09,075 INFO [train.py:421] (4/8) Epoch 1, batch 59800, loss[loss=2.428, over 2450.00 frames. , ppl: 11.331988237807481] tot_loss[loss=2.374, over 5462800.56 frames. , ppl: 10.744566748412375], batch size: 70
+2022-12-10 05:02:50,323 INFO [train.py:421] (4/8) Epoch 1, batch 60000, loss[loss=2.363, over 3500.00 frames. , ppl: 10.626940135265762] tot_loss[loss=2.374, over 5478364.97 frames. , ppl: 10.741154985077445], batch size: 70
+2022-12-10 05:02:50,324 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:02:51,100 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.353, over 211138.00 frames. , ppl: 10.521841308858221
+2022-12-10 05:04:30,580 INFO [train.py:421] (4/8) Epoch 1, batch 60200, loss[loss=2.291, over 7490.00 frames. , ppl: 9.885544263942279] tot_loss[loss=2.374, over 5483549.54 frames. , ppl: 10.73868625291688], batch size: 70
+2022-12-10 05:06:11,840 INFO [train.py:421] (4/8) Epoch 1, batch 60400, loss[loss=2.458, over 1260.00 frames. , ppl: 11.685562037704466] tot_loss[loss=2.373, over 5514377.57 frames. , ppl: 10.734503410338325], batch size: 70
+2022-12-10 05:07:53,808 INFO [train.py:421] (4/8) Epoch 1, batch 60600, loss[loss=2.42, over 2380.00 frames. , ppl: 11.240681298875499] tot_loss[loss=2.372, over 5551941.96 frames. , ppl: 10.721241585277234], batch size: 70
+2022-12-10 05:09:33,764 INFO [train.py:421] (4/8) Epoch 1, batch 60800, loss[loss=2.432, over 1400.00 frames. , ppl: 11.376779114291537] tot_loss[loss=2.372, over 5537658.97 frames. , ppl: 10.720313082308452], batch size: 70
+2022-12-10 05:11:18,590 INFO [train.py:421] (4/8) Epoch 1, batch 61000, loss[loss=2.475, over 1330.00 frames. , ppl: 11.877310913268047] tot_loss[loss=2.372, over 5513844.92 frames. , ppl: 10.717529953256479], batch size: 70
+2022-12-10 05:11:18,590 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:11:19,337 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.352, over 211138.00 frames. , ppl: 10.508142781873898
+2022-12-10 05:12:58,120 INFO [train.py:421] (4/8) Epoch 1, batch 61200, loss[loss=2.419, over 4060.00 frames. , ppl: 11.232405144382492] tot_loss[loss=2.371, over 5547144.60 frames. , ppl: 10.704011660249394], batch size: 70
+2022-12-10 05:14:39,407 INFO [train.py:421] (4/8) Epoch 1, batch 61400, loss[loss=2.292, over 6230.00 frames. , ppl: 9.890748628123927] tot_loss[loss=2.371, over 5545412.78 frames. , ppl: 10.705901411425016], batch size: 70
+2022-12-10 05:16:19,246 INFO [train.py:421] (4/8) Epoch 1, batch 61600, loss[loss=2.626, over 700.00 frames. , ppl: 13.816871470194698] tot_loss[loss=2.37, over 5549985.31 frames. , ppl: 10.699217228992573], batch size: 70
+2022-12-10 05:17:59,811 INFO [train.py:421] (4/8) Epoch 1, batch 61800, loss[loss=2.508, over 1330.00 frames. , ppl: 12.281874522058141] tot_loss[loss=2.37, over 5583524.79 frames. , ppl: 10.6924031235059], batch size: 70
+2022-12-10 05:19:41,042 INFO [train.py:421] (4/8) Epoch 1, batch 62000, loss[loss=2.626, over 770.00 frames. , ppl: 13.822307868290252] tot_loss[loss=2.369, over 5592388.27 frames. , ppl: 10.685664990779973], batch size: 70
+2022-12-10 05:19:41,042 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:19:41,789 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.351, over 211138.00 frames. , ppl: 10.500484247218756
+2022-12-10 05:21:23,028 INFO [train.py:421] (4/8) Epoch 1, batch 62200, loss[loss=2.395, over 2030.00 frames. , ppl: 10.964476704441573] tot_loss[loss=2.369, over 5572520.78 frames. , ppl: 10.687619710104423], batch size: 70
+2022-12-10 05:23:03,310 INFO [train.py:421] (4/8) Epoch 1, batch 62400, loss[loss=3.261, over 490.00 frames. , ppl: 26.069761207568266] tot_loss[loss=2.369, over 5567995.16 frames. , ppl: 10.683496337686426], batch size: 70
+2022-12-10 05:24:44,528 INFO [train.py:421] (4/8) Epoch 1, batch 62600, loss[loss=2.406, over 2310.00 frames. , ppl: 11.087575220778408] tot_loss[loss=2.369, over 5569473.90 frames. , ppl: 10.683814118812858], batch size: 70
+2022-12-10 05:26:22,732 INFO [train.py:421] (4/8) Epoch 1, batch 62800, loss[loss=2.397, over 1610.00 frames. , ppl: 10.985337270540779] tot_loss[loss=2.369, over 5559823.20 frames. , ppl: 10.690947130522938], batch size: 70
+2022-12-10 05:28:05,213 INFO [train.py:421] (4/8) Epoch 1, batch 63000, loss[loss=2.353, over 4550.00 frames. , ppl: 10.514869447052968] tot_loss[loss=2.368, over 5578392.33 frames. , ppl: 10.681190772196276], batch size: 70
+2022-12-10 05:28:05,214 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:28:05,959 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.352, over 211138.00 frames. , ppl: 10.50832786224836
+2022-12-10 05:29:47,018 INFO [train.py:421] (4/8) Epoch 1, batch 63200, loss[loss=2.366, over 2870.00 frames. , ppl: 10.659561575093587] tot_loss[loss=2.37, over 5499267.13 frames. , ppl: 10.697866972250713], batch size: 70
+2022-12-10 05:31:24,707 INFO [train.py:421] (4/8) Epoch 1, batch 63400, loss[loss=2.373, over 2240.00 frames. , ppl: 10.726345358489676] tot_loss[loss=2.37, over 5505666.64 frames. , ppl: 10.698339472156587], batch size: 70
+2022-12-10 05:33:03,872 INFO [train.py:421] (4/8) Epoch 1, batch 63600, loss[loss=2.381, over 2310.00 frames. , ppl: 10.817756582054958] tot_loss[loss=2.371, over 5470858.14 frames. , ppl: 10.70352145575299], batch size: 70
+2022-12-10 05:34:41,284 INFO [train.py:421] (4/8) Epoch 1, batch 63800, loss[loss=2.345, over 1680.00 frames. , ppl: 10.430867430523643] tot_loss[loss=2.371, over 5458538.82 frames. , ppl: 10.708303378536659], batch size: 70
+2022-12-10 05:36:22,761 INFO [train.py:421] (4/8) Epoch 1, batch 64000, loss[loss=2.276, over 2450.00 frames. , ppl: 9.735925504272423] tot_loss[loss=2.369, over 5501637.65 frames. , ppl: 10.68889894976589], batch size: 70
+2022-12-10 05:36:22,762 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:36:23,521 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.351, over 211138.00 frames. , ppl: 10.500185854649434
+2022-12-10 05:38:06,581 INFO [train.py:421] (4/8) Epoch 1, batch 64200, loss[loss=2.223, over 8470.00 frames. , ppl: 9.237701955558856] tot_loss[loss=2.37, over 5491156.49 frames. , ppl: 10.6963322260943], batch size: 70
+2022-12-10 05:39:47,998 INFO [train.py:421] (4/8) Epoch 1, batch 64400, loss[loss=2.417, over 2520.00 frames. , ppl: 11.210742185445637] tot_loss[loss=2.369, over 5519942.27 frames. , ppl: 10.686130933479047], batch size: 70
+2022-12-10 05:41:29,578 INFO [train.py:421] (4/8) Epoch 1, batch 64600, loss[loss=2.42, over 2590.00 frames. , ppl: 11.242174281055338] tot_loss[loss=2.368, over 5532856.66 frames. , ppl: 10.680073149026532], batch size: 70
+2022-12-10 05:43:06,385 INFO [train.py:421] (4/8) Epoch 1, batch 64800, loss[loss=2.249, over 10290.00 frames. , ppl: 9.473778729146419] tot_loss[loss=2.367, over 5567706.78 frames. , ppl: 10.660981038902428], batch size: 70
+2022-12-10 05:44:45,897 INFO [train.py:421] (4/8) Epoch 1, batch 65000, loss[loss=2.407, over 2030.00 frames. , ppl: 11.096321814545833] tot_loss[loss=2.366, over 5541328.90 frames. , ppl: 10.65915968404463], batch size: 70
+2022-12-10 05:44:45,897 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:44:46,659 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.351, over 211138.00 frames. , ppl: 10.499536258484806
+2022-12-10 05:46:27,198 INFO [train.py:421] (4/8) Epoch 1, batch 65200, loss[loss=2.426, over 1470.00 frames. , ppl: 11.312321444374055] tot_loss[loss=2.366, over 5569236.11 frames. , ppl: 10.650013127669338], batch size: 70
+2022-12-10 05:48:06,794 INFO [train.py:421] (4/8) Epoch 1, batch 65400, loss[loss=2.377, over 4340.00 frames. , ppl: 10.77507657925034] tot_loss[loss=2.366, over 5543778.48 frames. , ppl: 10.650039645677808], batch size: 70
+2022-12-10 05:49:47,266 INFO [train.py:421] (4/8) Epoch 1, batch 65600, loss[loss=2.665, over 1120.00 frames. , ppl: 14.364940669651986] tot_loss[loss=2.366, over 5526941.41 frames. , ppl: 10.657949774695286], batch size: 70
+2022-12-10 05:51:25,721 INFO [train.py:421] (4/8) Epoch 1, batch 65800, loss[loss=2.548, over 770.00 frames. , ppl: 12.777314714338976] tot_loss[loss=2.367, over 5524278.31 frames. , ppl: 10.660585945314137], batch size: 70
+2022-12-10 05:53:05,372 INFO [train.py:421] (4/8) Epoch 1, batch 66000, loss[loss=2.417, over 2310.00 frames. , ppl: 11.214677754454778] tot_loss[loss=2.367, over 5507383.17 frames. , ppl: 10.661280750376696], batch size: 70
+2022-12-10 05:53:05,372 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 05:53:06,133 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.349, over 211138.00 frames. , ppl: 10.479744419827652
+2022-12-10 05:54:45,806 INFO [train.py:421] (4/8) Epoch 1, batch 66200, loss[loss=2.311, over 3990.00 frames. , ppl: 10.083661606390361] tot_loss[loss=2.369, over 5426471.73 frames. , ppl: 10.684381489852418], batch size: 70
+2022-12-10 05:56:26,059 INFO [train.py:421] (4/8) Epoch 1, batch 66400, loss[loss=2.393, over 2030.00 frames. , ppl: 10.950323468247463] tot_loss[loss=2.369, over 5435996.19 frames. , ppl: 10.68995998028201], batch size: 70
+2022-12-10 05:58:06,924 INFO [train.py:421] (4/8) Epoch 1, batch 66600, loss[loss=2.279, over 5180.00 frames. , ppl: 9.768364686663102] tot_loss[loss=2.368, over 5491859.67 frames. , ppl: 10.67427452456624], batch size: 70
+2022-12-10 05:59:44,019 INFO [train.py:421] (4/8) Epoch 1, batch 66800, loss[loss=2.427, over 1750.00 frames. , ppl: 11.323080808376824] tot_loss[loss=2.367, over 5483949.89 frames. , ppl: 10.667940737380158], batch size: 70
+2022-12-10 06:01:21,583 INFO [train.py:421] (4/8) Epoch 1, batch 67000, loss[loss=2.399, over 4060.00 frames. , ppl: 11.011579811094881] tot_loss[loss=2.368, over 5490405.25 frames. , ppl: 10.67559371392806], batch size: 70
+2022-12-10 06:01:21,584 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:01:22,329 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.348, over 211138.00 frames. , ppl: 10.464386029268178
+2022-12-10 06:02:59,567 INFO [train.py:421] (4/8) Epoch 1, batch 67200, loss[loss=2.495, over 2940.00 frames. , ppl: 12.12014842662844] tot_loss[loss=2.37, over 5423973.15 frames. , ppl: 10.693115467658819], batch size: 70
+2022-12-10 06:04:38,155 INFO [train.py:421] (4/8) Epoch 1, batch 67400, loss[loss=2.257, over 4130.00 frames. , ppl: 9.553563290963407] tot_loss[loss=2.369, over 5423265.73 frames. , ppl: 10.686131997785218], batch size: 70
+2022-12-10 06:06:18,488 INFO [train.py:421] (4/8) Epoch 1, batch 67600, loss[loss=2.478, over 2240.00 frames. , ppl: 11.917988817496026] tot_loss[loss=2.368, over 5466002.87 frames. , ppl: 10.680516329457053], batch size: 70
+2022-12-10 06:08:00,614 INFO [train.py:421] (4/8) Epoch 1, batch 67800, loss[loss=2.549, over 980.00 frames. , ppl: 12.793898547435163] tot_loss[loss=2.368, over 5457308.13 frames. , ppl: 10.681284222560162], batch size: 70
+2022-12-10 06:09:41,558 INFO [train.py:421] (4/8) Epoch 1, batch 68000, loss[loss=2.263, over 9660.00 frames. , ppl: 9.61631558793531] tot_loss[loss=2.367, over 5476759.67 frames. , ppl: 10.669574882379292], batch size: 70
+2022-12-10 06:09:41,558 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:09:42,303 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.35, over 211138.00 frames. , ppl: 10.481405758605055
+2022-12-10 06:11:26,714 INFO [train.py:421] (4/8) Epoch 1, batch 68200, loss[loss=2.293, over 2520.00 frames. , ppl: 9.899886478341179] tot_loss[loss=2.368, over 5484946.22 frames. , ppl: 10.673696748625165], batch size: 70
+2022-12-10 06:13:04,726 INFO [train.py:421] (4/8) Epoch 1, batch 68400, loss[loss=2.443, over 1610.00 frames. , ppl: 11.50673839645576] tot_loss[loss=2.369, over 5452968.57 frames. , ppl: 10.682412183482652], batch size: 70
+2022-12-10 06:14:45,539 INFO [train.py:421] (4/8) Epoch 1, batch 68600, loss[loss=2.401, over 2870.00 frames. , ppl: 11.029457022811282] tot_loss[loss=2.37, over 5413934.28 frames. , ppl: 10.694179998349538], batch size: 70
+2022-12-10 06:16:28,483 INFO [train.py:421] (4/8) Epoch 1, batch 68800, loss[loss=2.529, over 1050.00 frames. , ppl: 12.543918370099533] tot_loss[loss=2.369, over 5425336.15 frames. , ppl: 10.688672721870265], batch size: 70
+2022-12-10 06:18:08,251 INFO [train.py:421] (4/8) Epoch 1, batch 69000, loss[loss=2.296, over 8680.00 frames. , ppl: 9.933723075496175] tot_loss[loss=2.369, over 5432565.00 frames. , ppl: 10.686157899431338], batch size: 70
+2022-12-10 06:18:08,251 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:18:08,996 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.348, over 211138.00 frames. , ppl: 10.467406639264686
+2022-12-10 06:19:44,538 INFO [train.py:421] (4/8) Epoch 1, batch 69200, loss[loss=2.564, over 1610.00 frames. , ppl: 12.981666961258815] tot_loss[loss=2.368, over 5439962.74 frames. , ppl: 10.67984212254364], batch size: 70
+2022-12-10 06:21:26,218 INFO [train.py:421] (4/8) Epoch 1, batch 69400, loss[loss=2.608, over 1050.00 frames. , ppl: 13.574542309399094] tot_loss[loss=2.368, over 5453519.64 frames. , ppl: 10.673288155905322], batch size: 70
+2022-12-10 06:23:06,586 INFO [train.py:421] (4/8) Epoch 1, batch 69600, loss[loss=2.257, over 7560.00 frames. , ppl: 9.550012391683364] tot_loss[loss=2.368, over 5447198.54 frames. , ppl: 10.680013172411256], batch size: 70
+2022-12-10 06:24:43,354 INFO [train.py:421] (4/8) Epoch 1, batch 69800, loss[loss=2.902, over 630.00 frames. , ppl: 18.21173357873054] tot_loss[loss=2.368, over 5458991.54 frames. , ppl: 10.675836855186189], batch size: 70
+2022-12-10 06:26:25,267 INFO [train.py:421] (4/8) Epoch 1, batch 70000, loss[loss=2.328, over 5180.00 frames. , ppl: 10.258228160410045] tot_loss[loss=2.367, over 5482578.77 frames. , ppl: 10.669472449254284], batch size: 70
+2022-12-10 06:26:25,267 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:26:26,013 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.348, over 211138.00 frames. , ppl: 10.463410326196138
+2022-12-10 06:28:04,612 INFO [train.py:421] (4/8) Epoch 1, batch 70200, loss[loss=2.374, over 3500.00 frames. , ppl: 10.74559704806556] tot_loss[loss=2.367, over 5492564.67 frames. , ppl: 10.661550023731207], batch size: 70
+2022-12-10 06:29:39,960 INFO [train.py:421] (4/8) Epoch 1, batch 70400, loss[loss=2.336, over 2590.00 frames. , ppl: 10.339877363015809] tot_loss[loss=2.367, over 5466832.79 frames. , ppl: 10.664091210828127], batch size: 70
+2022-12-10 06:31:16,084 INFO [train.py:421] (4/8) Epoch 1, batch 70600, loss[loss=2.27, over 2940.00 frames. , ppl: 9.680146114444831] tot_loss[loss=2.367, over 5436912.46 frames. , ppl: 10.669722169067505], batch size: 70
+2022-12-10 06:32:52,996 INFO [train.py:421] (4/8) Epoch 1, batch 70800, loss[loss=2.458, over 1680.00 frames. , ppl: 11.682202414474672] tot_loss[loss=2.367, over 5425160.11 frames. , ppl: 10.668982589462946], batch size: 70
+2022-12-10 06:34:35,556 INFO [train.py:421] (4/8) Epoch 1, batch 71000, loss[loss=2.421, over 4900.00 frames. , ppl: 11.258933586231972] tot_loss[loss=2.366, over 5488918.58 frames. , ppl: 10.654501267834432], batch size: 70
+2022-12-10 06:34:35,557 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:34:36,305 INFO [train.py:452] (4/8) Epoch 1, validation: loss=2.347, over 211138.00 frames. , ppl: 10.45096802929164
+2022-12-10 06:36:12,870 INFO [train.py:421] (4/8) Epoch 1, batch 71200, loss[loss=2.405, over 1960.00 frames. , ppl: 11.083188254763378] tot_loss[loss=2.366, over 5497756.52 frames. , ppl: 10.653669571636305], batch size: 70
+2022-12-10 06:37:52,675 INFO [train.py:421] (4/8) Epoch 1, batch 71400, loss[loss=2.373, over 3430.00 frames. , ppl: 10.728691503002912] tot_loss[loss=2.367, over 5476182.85 frames. , ppl: 10.667584767346538], batch size: 70
+2022-12-10 06:39:34,126 INFO [train.py:421] (4/8) Epoch 1, batch 71600, loss[loss=2.503, over 1190.00 frames. , ppl: 12.224951969602477] tot_loss[loss=2.368, over 5462426.88 frames. , ppl: 10.672083414783032], batch size: 70
+2022-12-10 06:41:16,765 INFO [train.py:421] (4/8) Epoch 1, batch 71800, loss[loss=2.466, over 1890.00 frames. , ppl: 11.776199120068801] tot_loss[loss=2.368, over 5442745.40 frames. , ppl: 10.671242001985288], batch size: 70
+2022-12-10 06:42:32,558 INFO [train.py:421] (4/8) Epoch 2, batch 0, loss[loss=2.349, over 5040.00 frames. , ppl: 10.470087352088488] tot_loss[loss=2.349, over 5040.00 frames. , ppl: 10.470087352088488], batch size: 70
+2022-12-10 06:44:11,579 INFO [train.py:421] (4/8) Epoch 2, batch 200, loss[loss=2.482, over 1260.00 frames. , ppl: 11.969356938803891] tot_loss[loss=2.35, over 538020.32 frames. , ppl: 10.482444616919764], batch size: 70
+2022-12-10 06:45:50,262 INFO [train.py:421] (4/8) Epoch 2, batch 400, loss[loss=2.596, over 840.00 frames. , ppl: 13.4050321957411] tot_loss[loss=2.359, over 983030.89 frames. , ppl: 10.584582125553924], batch size: 70
+2022-12-10 06:47:30,606 INFO [train.py:421] (4/8) Epoch 2, batch 600, loss[loss=2.332, over 3640.00 frames. , ppl: 10.302266557334308] tot_loss[loss=2.356, over 1449080.49 frames. , ppl: 10.548987690762543], batch size: 70
+2022-12-10 06:49:11,810 INFO [train.py:421] (4/8) Epoch 2, batch 800, loss[loss=2.369, over 1890.00 frames. , ppl: 10.683355425781265] tot_loss[loss=2.358, over 1820076.50 frames. , ppl: 10.567754810511486], batch size: 70
+2022-12-10 06:50:52,992 INFO [train.py:421] (4/8) Epoch 2, batch 1000, loss[loss=2.409, over 3500.00 frames. , ppl: 11.12838313592859] tot_loss[loss=2.356, over 2216336.86 frames. , ppl: 10.545407099751843], batch size: 70
+2022-12-10 06:50:52,992 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:50:53,753 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.346, over 211138.00 frames. , ppl: 10.441904495381296
+2022-12-10 06:52:33,130 INFO [train.py:421] (4/8) Epoch 2, batch 1200, loss[loss=2.939, over 560.00 frames. , ppl: 18.897717313260845] tot_loss[loss=2.359, over 2477508.74 frames. , ppl: 10.579960775636385], batch size: 70
+2022-12-10 06:54:14,266 INFO [train.py:421] (4/8) Epoch 2, batch 1400, loss[loss=2.349, over 1610.00 frames. , ppl: 10.472339461548117] tot_loss[loss=2.359, over 2745639.47 frames. , ppl: 10.584500092560193], batch size: 70
+2022-12-10 06:55:55,748 INFO [train.py:421] (4/8) Epoch 2, batch 1600, loss[loss=2.412, over 2100.00 frames. , ppl: 11.160076591201216] tot_loss[loss=2.355, over 3044123.62 frames. , ppl: 10.542291822871455], batch size: 70
+2022-12-10 06:57:35,263 INFO [train.py:421] (4/8) Epoch 2, batch 1800, loss[loss=2.468, over 1960.00 frames. , ppl: 11.794404325124273] tot_loss[loss=2.355, over 3298517.80 frames. , ppl: 10.54195283754715], batch size: 70
+2022-12-10 06:59:14,992 INFO [train.py:421] (4/8) Epoch 2, batch 2000, loss[loss=2.596, over 1190.00 frames. , ppl: 13.40705258359372] tot_loss[loss=2.355, over 3494674.28 frames. , ppl: 10.541592122660713], batch size: 70
+2022-12-10 06:59:14,993 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 06:59:15,755 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.345, over 211138.00 frames. , ppl: 10.430619322580018
+2022-12-10 07:00:58,121 INFO [train.py:421] (4/8) Epoch 2, batch 2200, loss[loss=2.277, over 3570.00 frames. , ppl: 9.748879973811663] tot_loss[loss=2.355, over 3720298.12 frames. , ppl: 10.542010426346387], batch size: 70
+2022-12-10 07:02:37,018 INFO [train.py:421] (4/8) Epoch 2, batch 2400, loss[loss=2.323, over 2590.00 frames. , ppl: 10.208513359555345] tot_loss[loss=2.357, over 3861590.28 frames. , ppl: 10.560188499985937], batch size: 70
+2022-12-10 07:04:16,524 INFO [train.py:421] (4/8) Epoch 2, batch 2600, loss[loss=2.371, over 1750.00 frames. , ppl: 10.706607230877552] tot_loss[loss=2.355, over 4052146.34 frames. , ppl: 10.539199849453949], batch size: 70
+2022-12-10 07:05:57,378 INFO [train.py:421] (4/8) Epoch 2, batch 2800, loss[loss=2.343, over 3080.00 frames. , ppl: 10.414686906547102] tot_loss[loss=2.356, over 4144888.49 frames. , ppl: 10.552127199427758], batch size: 70
+2022-12-10 07:07:34,871 INFO [train.py:421] (4/8) Epoch 2, batch 3000, loss[loss=2.493, over 770.00 frames. , ppl: 12.097170757787849] tot_loss[loss=2.356, over 4273894.43 frames. , ppl: 10.55315721948215], batch size: 70
+2022-12-10 07:07:34,871 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:07:35,617 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.346, over 211138.00 frames. , ppl: 10.439906381190209
+2022-12-10 07:09:13,504 INFO [train.py:421] (4/8) Epoch 2, batch 3200, loss[loss=2.313, over 4900.00 frames. , ppl: 10.105144824495758] tot_loss[loss=2.356, over 4416772.44 frames. , ppl: 10.553089128181366], batch size: 70
+2022-12-10 07:10:54,075 INFO [train.py:421] (4/8) Epoch 2, batch 3400, loss[loss=2.36, over 2100.00 frames. , ppl: 10.595631338344285] tot_loss[loss=2.355, over 4539545.94 frames. , ppl: 10.540174002844129], batch size: 70
+2022-12-10 07:12:38,050 INFO [train.py:421] (4/8) Epoch 2, batch 3600, loss[loss=2.313, over 6020.00 frames. , ppl: 10.105638767035954] tot_loss[loss=2.355, over 4645368.54 frames. , ppl: 10.537181942198115], batch size: 70
+2022-12-10 07:14:17,882 INFO [train.py:421] (4/8) Epoch 2, batch 3800, loss[loss=2.285, over 4830.00 frames. , ppl: 9.825103762956282] tot_loss[loss=2.355, over 4713864.04 frames. , ppl: 10.540814748173219], batch size: 70
+2022-12-10 07:15:59,649 INFO [train.py:421] (4/8) Epoch 2, batch 4000, loss[loss=2.309, over 4060.00 frames. , ppl: 10.060698189610399] tot_loss[loss=2.355, over 4810100.06 frames. , ppl: 10.537261591802224], batch size: 70
+2022-12-10 07:15:59,649 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:16:00,396 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.345, over 211138.00 frames. , ppl: 10.432709850132296
+2022-12-10 07:17:44,062 INFO [train.py:421] (4/8) Epoch 2, batch 4200, loss[loss=2.353, over 3360.00 frames. , ppl: 10.515246150864224] tot_loss[loss=2.355, over 4874948.41 frames. , ppl: 10.539321462763429], batch size: 70
+2022-12-10 07:19:26,558 INFO [train.py:421] (4/8) Epoch 2, batch 4400, loss[loss=2.33, over 3710.00 frames. , ppl: 10.281637239377803] tot_loss[loss=2.355, over 4938825.74 frames. , ppl: 10.542401232830775], batch size: 70
+2022-12-10 07:21:09,714 INFO [train.py:421] (4/8) Epoch 2, batch 4600, loss[loss=2.802, over 630.00 frames. , ppl: 16.48229196062281] tot_loss[loss=2.355, over 5031098.70 frames. , ppl: 10.533679463987825], batch size: 70
+2022-12-10 07:22:49,446 INFO [train.py:421] (4/8) Epoch 2, batch 4800, loss[loss=2.562, over 1540.00 frames. , ppl: 12.968189093429409] tot_loss[loss=2.355, over 5070572.13 frames. , ppl: 10.541451154983935], batch size: 70
+2022-12-10 07:24:27,422 INFO [train.py:421] (4/8) Epoch 2, batch 5000, loss[loss=2.414, over 2380.00 frames. , ppl: 11.17880552322546] tot_loss[loss=2.356, over 5102793.18 frames. , ppl: 10.545396287978177], batch size: 70
+2022-12-10 07:24:27,422 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:24:28,168 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.417051893133054
+2022-12-10 07:26:10,650 INFO [train.py:421] (4/8) Epoch 2, batch 5200, loss[loss=2.276, over 1540.00 frames. , ppl: 9.739041507436063] tot_loss[loss=2.357, over 5104998.49 frames. , ppl: 10.561211807250146], batch size: 70
+2022-12-10 07:27:50,428 INFO [train.py:421] (4/8) Epoch 2, batch 5400, loss[loss=2.296, over 4270.00 frames. , ppl: 9.938637547794416] tot_loss[loss=2.358, over 5128433.64 frames. , ppl: 10.570347185760454], batch size: 70
+2022-12-10 07:29:28,145 INFO [train.py:421] (4/8) Epoch 2, batch 5600, loss[loss=2.398, over 2170.00 frames. , ppl: 10.997623472134773] tot_loss[loss=2.358, over 5146949.50 frames. , ppl: 10.56703115685762], batch size: 70
+2022-12-10 07:31:06,307 INFO [train.py:421] (4/8) Epoch 2, batch 5800, loss[loss=2.492, over 1680.00 frames. , ppl: 12.081788289373053] tot_loss[loss=2.358, over 5157181.93 frames. , ppl: 10.567355107655933], batch size: 70
+2022-12-10 07:32:43,819 INFO [train.py:421] (4/8) Epoch 2, batch 6000, loss[loss=2.32, over 2730.00 frames. , ppl: 10.17539257448557] tot_loss[loss=2.358, over 5162472.92 frames. , ppl: 10.569839579174722], batch size: 70
+2022-12-10 07:32:43,820 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:32:44,568 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.413985701421621
+2022-12-10 07:34:23,914 INFO [train.py:421] (4/8) Epoch 2, batch 6200, loss[loss=2.285, over 4620.00 frames. , ppl: 9.8266150690876] tot_loss[loss=2.359, over 5150514.25 frames. , ppl: 10.578468959528184], batch size: 70
+2022-12-10 07:36:04,956 INFO [train.py:421] (4/8) Epoch 2, batch 6400, loss[loss=2.374, over 1890.00 frames. , ppl: 10.737245495765796] tot_loss[loss=2.358, over 5178900.08 frames. , ppl: 10.57423040885354], batch size: 70
+2022-12-10 07:37:46,138 INFO [train.py:421] (4/8) Epoch 2, batch 6600, loss[loss=2.331, over 3290.00 frames. , ppl: 10.28710160074556] tot_loss[loss=2.358, over 5224843.52 frames. , ppl: 10.570839332261102], batch size: 70
+2022-12-10 07:39:25,813 INFO [train.py:421] (4/8) Epoch 2, batch 6800, loss[loss=2.505, over 1540.00 frames. , ppl: 12.246251049651649] tot_loss[loss=2.358, over 5239005.31 frames. , ppl: 10.570839092403919], batch size: 70
+2022-12-10 07:41:05,167 INFO [train.py:421] (4/8) Epoch 2, batch 7000, loss[loss=2.309, over 6020.00 frames. , ppl: 10.06521028130956] tot_loss[loss=2.359, over 5243265.32 frames. , ppl: 10.578854206179779], batch size: 70
+2022-12-10 07:41:05,168 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:41:05,947 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.402274966255137
+2022-12-10 07:42:45,113 INFO [train.py:421] (4/8) Epoch 2, batch 7200, loss[loss=2.248, over 7000.00 frames. , ppl: 9.470906342985119] tot_loss[loss=2.358, over 5287322.71 frames. , ppl: 10.567574159534187], batch size: 70
+2022-12-10 07:44:25,815 INFO [train.py:421] (4/8) Epoch 2, batch 7400, loss[loss=2.352, over 2520.00 frames. , ppl: 10.503409067272612] tot_loss[loss=2.359, over 5225346.75 frames. , ppl: 10.584076606205628], batch size: 70
+2022-12-10 07:46:04,877 INFO [train.py:421] (4/8) Epoch 2, batch 7600, loss[loss=2.306, over 4760.00 frames. , ppl: 10.03130636267069] tot_loss[loss=2.36, over 5217751.15 frames. , ppl: 10.592701829520166], batch size: 70
+2022-12-10 07:47:41,317 INFO [train.py:421] (4/8) Epoch 2, batch 7800, loss[loss=2.247, over 2590.00 frames. , ppl: 9.460376346291737] tot_loss[loss=2.36, over 5233994.58 frames. , ppl: 10.588107238862841], batch size: 70
+2022-12-10 07:49:26,414 INFO [train.py:421] (4/8) Epoch 2, batch 8000, loss[loss=2.36, over 4900.00 frames. , ppl: 10.593180650000109] tot_loss[loss=2.359, over 5269661.29 frames. , ppl: 10.585220742849843], batch size: 70
+2022-12-10 07:49:26,414 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:49:27,161 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.408854278509716
+2022-12-10 07:51:05,866 INFO [train.py:421] (4/8) Epoch 2, batch 8200, loss[loss=2.334, over 3290.00 frames. , ppl: 10.320640779515108] tot_loss[loss=2.36, over 5267355.71 frames. , ppl: 10.59309100360825], batch size: 70
+2022-12-10 07:52:44,237 INFO [train.py:421] (4/8) Epoch 2, batch 8400, loss[loss=2.312, over 2380.00 frames. , ppl: 10.09876786054819] tot_loss[loss=2.359, over 5288916.92 frames. , ppl: 10.584879146335094], batch size: 70
+2022-12-10 07:54:24,081 INFO [train.py:421] (4/8) Epoch 2, batch 8600, loss[loss=2.26, over 5040.00 frames. , ppl: 9.587406183234016] tot_loss[loss=2.359, over 5316688.85 frames. , ppl: 10.578694789771903], batch size: 70
+2022-12-10 07:56:01,956 INFO [train.py:421] (4/8) Epoch 2, batch 8800, loss[loss=2.797, over 630.00 frames. , ppl: 16.390002568697188] tot_loss[loss=2.359, over 5290779.99 frames. , ppl: 10.585576649478513], batch size: 70
+2022-12-10 07:57:42,022 INFO [train.py:421] (4/8) Epoch 2, batch 9000, loss[loss=2.27, over 6300.00 frames. , ppl: 9.681471591898545] tot_loss[loss=2.359, over 5304093.41 frames. , ppl: 10.579622393590887], batch size: 70
+2022-12-10 07:57:42,023 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 07:57:42,768 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.399154636076556
+2022-12-10 07:59:22,462 INFO [train.py:421] (4/8) Epoch 2, batch 9200, loss[loss=2.974, over 700.00 frames. , ppl: 19.573853777015202] tot_loss[loss=2.359, over 5299552.16 frames. , ppl: 10.576276213917732], batch size: 70
+2022-12-10 08:00:56,951 INFO [train.py:421] (4/8) Epoch 2, batch 9400, loss[loss=2.329, over 3850.00 frames. , ppl: 10.2665030561731] tot_loss[loss=2.359, over 5308902.29 frames. , ppl: 10.576374984918287], batch size: 70
+2022-12-10 08:02:37,352 INFO [train.py:421] (4/8) Epoch 2, batch 9600, loss[loss=2.448, over 1890.00 frames. , ppl: 11.56476126917269] tot_loss[loss=2.359, over 5310081.85 frames. , ppl: 10.577777297427803], batch size: 70
+2022-12-10 08:04:17,627 INFO [train.py:421] (4/8) Epoch 2, batch 9800, loss[loss=2.373, over 1890.00 frames. , ppl: 10.72808699438715] tot_loss[loss=2.358, over 5321993.49 frames. , ppl: 10.5737309262498], batch size: 70
+2022-12-10 08:06:01,226 INFO [train.py:421] (4/8) Epoch 2, batch 10000, loss[loss=2.369, over 6230.00 frames. , ppl: 10.68256695077743] tot_loss[loss=2.358, over 5365806.85 frames. , ppl: 10.566757574871103], batch size: 70
+2022-12-10 08:06:01,227 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:06:01,975 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.343, over 211138.00 frames. , ppl: 10.415013831030345
+2022-12-10 08:07:39,460 INFO [train.py:421] (4/8) Epoch 2, batch 10200, loss[loss=2.36, over 2380.00 frames. , ppl: 10.59622320862524] tot_loss[loss=2.358, over 5349425.54 frames. , ppl: 10.572406073972278], batch size: 70
+2022-12-10 08:09:21,337 INFO [train.py:421] (4/8) Epoch 2, batch 10400, loss[loss=2.817, over 630.00 frames. , ppl: 16.730017333837125] tot_loss[loss=2.358, over 5351538.92 frames. , ppl: 10.569554654752054], batch size: 70
+2022-12-10 08:11:01,467 INFO [train.py:421] (4/8) Epoch 2, batch 10600, loss[loss=2.366, over 2450.00 frames. , ppl: 10.656744740187541] tot_loss[loss=2.357, over 5372640.95 frames. , ppl: 10.56375906329315], batch size: 70
+2022-12-10 08:12:42,386 INFO [train.py:421] (4/8) Epoch 2, batch 10800, loss[loss=2.477, over 1680.00 frames. , ppl: 11.908286091379283] tot_loss[loss=2.357, over 5419439.83 frames. , ppl: 10.555056862192655], batch size: 70
+2022-12-10 08:14:21,837 INFO [train.py:421] (4/8) Epoch 2, batch 11000, loss[loss=2.484, over 1330.00 frames. , ppl: 11.984523928916744] tot_loss[loss=2.356, over 5428494.72 frames. , ppl: 10.54782654807046], batch size: 70
+2022-12-10 08:14:21,837 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:14:22,596 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.400310603854379
+2022-12-10 08:16:01,434 INFO [train.py:421] (4/8) Epoch 2, batch 11200, loss[loss=2.451, over 1540.00 frames. , ppl: 11.599897021039263] tot_loss[loss=2.356, over 5446420.27 frames. , ppl: 10.545912054487598], batch size: 70
+2022-12-10 08:17:42,428 INFO [train.py:421] (4/8) Epoch 2, batch 11400, loss[loss=2.393, over 1470.00 frames. , ppl: 10.946599627173502] tot_loss[loss=2.356, over 5447651.52 frames. , ppl: 10.545579504010867], batch size: 70
+2022-12-10 08:19:25,307 INFO [train.py:421] (4/8) Epoch 2, batch 11600, loss[loss=2.373, over 3780.00 frames. , ppl: 10.725362857651453] tot_loss[loss=2.355, over 5504988.61 frames. , ppl: 10.536621819449703], batch size: 70
+2022-12-10 08:21:08,341 INFO [train.py:421] (4/8) Epoch 2, batch 11800, loss[loss=2.401, over 1890.00 frames. , ppl: 11.029922119886562] tot_loss[loss=2.354, over 5531015.24 frames. , ppl: 10.527582950520605], batch size: 70
+2022-12-10 08:22:51,113 INFO [train.py:421] (4/8) Epoch 2, batch 12000, loss[loss=2.293, over 3010.00 frames. , ppl: 9.904715205306731] tot_loss[loss=2.354, over 5549088.29 frames. , ppl: 10.529294027086213], batch size: 70
+2022-12-10 08:22:51,113 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:22:51,875 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.342, over 211138.00 frames. , ppl: 10.405858265858988
+2022-12-10 08:24:33,719 INFO [train.py:421] (4/8) Epoch 2, batch 12200, loss[loss=2.415, over 2240.00 frames. , ppl: 11.186297997119658] tot_loss[loss=2.355, over 5517310.69 frames. , ppl: 10.533689752502621], batch size: 70
+2022-12-10 08:26:16,320 INFO [train.py:421] (4/8) Epoch 2, batch 12400, loss[loss=2.598, over 910.00 frames. , ppl: 13.441498429847156] tot_loss[loss=2.353, over 5572530.66 frames. , ppl: 10.517075597178186], batch size: 70
+2022-12-10 08:27:55,050 INFO [train.py:421] (4/8) Epoch 2, batch 12600, loss[loss=2.48, over 2310.00 frames. , ppl: 11.941902025212013] tot_loss[loss=2.353, over 5549942.02 frames. , ppl: 10.522329589029427], batch size: 70
+2022-12-10 08:29:33,559 INFO [train.py:421] (4/8) Epoch 2, batch 12800, loss[loss=2.274, over 7980.00 frames. , ppl: 9.715301117495676] tot_loss[loss=2.353, over 5560246.09 frames. , ppl: 10.52011525071371], batch size: 70
+2022-12-10 08:31:14,850 INFO [train.py:421] (4/8) Epoch 2, batch 13000, loss[loss=2.382, over 1540.00 frames. , ppl: 10.830494870285747] tot_loss[loss=2.354, over 5532895.74 frames. , ppl: 10.528264615833034], batch size: 70
+2022-12-10 08:31:14,850 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:31:15,609 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.345, over 211138.00 frames. , ppl: 10.435761475361208
+2022-12-10 08:32:58,530 INFO [train.py:421] (4/8) Epoch 2, batch 13200, loss[loss=2.4, over 2940.00 frames. , ppl: 11.018466862545162] tot_loss[loss=2.352, over 5592090.72 frames. , ppl: 10.507210962429435], batch size: 70
+2022-12-10 08:34:37,206 INFO [train.py:421] (4/8) Epoch 2, batch 13400, loss[loss=2.944, over 770.00 frames. , ppl: 18.990387819860878] tot_loss[loss=2.353, over 5545885.81 frames. , ppl: 10.513975359294548], batch size: 70
+2022-12-10 08:36:17,992 INFO [train.py:421] (4/8) Epoch 2, batch 13600, loss[loss=2.361, over 3710.00 frames. , ppl: 10.597676637452322] tot_loss[loss=2.352, over 5579224.85 frames. , ppl: 10.504929222204863], batch size: 70
+2022-12-10 08:38:00,730 INFO [train.py:421] (4/8) Epoch 2, batch 13800, loss[loss=2.242, over 8190.00 frames. , ppl: 9.410700868638816] tot_loss[loss=2.351, over 5622135.56 frames. , ppl: 10.498439636183372], batch size: 70
+2022-12-10 08:39:39,900 INFO [train.py:421] (4/8) Epoch 2, batch 14000, loss[loss=2.956, over 560.00 frames. , ppl: 19.229059343297973] tot_loss[loss=2.353, over 5608331.72 frames. , ppl: 10.514058583738203], batch size: 70
+2022-12-10 08:39:39,900 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:39:40,646 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.34, over 211138.00 frames. , ppl: 10.38519312924439
+2022-12-10 08:41:22,077 INFO [train.py:421] (4/8) Epoch 2, batch 14200, loss[loss=2.388, over 1470.00 frames. , ppl: 10.896895053260558] tot_loss[loss=2.352, over 5628081.62 frames. , ppl: 10.504661606320681], batch size: 70
+2022-12-10 08:43:05,914 INFO [train.py:421] (4/8) Epoch 2, batch 14400, loss[loss=2.491, over 1330.00 frames. , ppl: 12.074822014970717] tot_loss[loss=2.353, over 5548745.63 frames. , ppl: 10.518740203704935], batch size: 70
+2022-12-10 08:44:46,250 INFO [train.py:421] (4/8) Epoch 2, batch 14600, loss[loss=2.295, over 11970.00 frames. , ppl: 9.92288673558068] tot_loss[loss=2.352, over 5556442.91 frames. , ppl: 10.506303291490328], batch size: 70
+2022-12-10 08:46:23,523 INFO [train.py:421] (4/8) Epoch 2, batch 14800, loss[loss=2.287, over 2940.00 frames. , ppl: 9.845941937960513] tot_loss[loss=2.352, over 5547339.20 frames. , ppl: 10.501992731587082], batch size: 70
+2022-12-10 08:48:03,521 INFO [train.py:421] (4/8) Epoch 2, batch 15000, loss[loss=2.557, over 770.00 frames. , ppl: 12.89781456594272] tot_loss[loss=2.351, over 5543360.52 frames. , ppl: 10.491651441411818], batch size: 70
+2022-12-10 08:48:03,521 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:48:04,267 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.339, over 211138.00 frames. , ppl: 10.36868412657422
+2022-12-10 08:49:43,366 INFO [train.py:421] (4/8) Epoch 2, batch 15200, loss[loss=2.569, over 1050.00 frames. , ppl: 13.051662895524474] tot_loss[loss=2.351, over 5522412.19 frames. , ppl: 10.495705780514884], batch size: 70
+2022-12-10 08:51:24,495 INFO [train.py:421] (4/8) Epoch 2, batch 15400, loss[loss=2.332, over 1820.00 frames. , ppl: 10.297027425542716] tot_loss[loss=2.351, over 5545592.98 frames. , ppl: 10.495596607541387], batch size: 70
+2022-12-10 08:53:07,426 INFO [train.py:421] (4/8) Epoch 2, batch 15600, loss[loss=2.445, over 2100.00 frames. , ppl: 11.534652295840186] tot_loss[loss=2.352, over 5517961.16 frames. , ppl: 10.507428665109131], batch size: 70
+2022-12-10 08:54:45,435 INFO [train.py:421] (4/8) Epoch 2, batch 15800, loss[loss=2.54, over 1190.00 frames. , ppl: 12.684446412799788] tot_loss[loss=2.353, over 5507768.59 frames. , ppl: 10.52231243563374], batch size: 70
+2022-12-10 08:56:24,652 INFO [train.py:421] (4/8) Epoch 2, batch 16000, loss[loss=2.587, over 910.00 frames. , ppl: 13.285163980136169] tot_loss[loss=2.354, over 5476183.97 frames. , ppl: 10.526879546353559], batch size: 70
+2022-12-10 08:56:24,653 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 08:56:25,397 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.34, over 211138.00 frames. , ppl: 10.380303272822726
+2022-12-10 08:58:07,170 INFO [train.py:421] (4/8) Epoch 2, batch 16200, loss[loss=2.547, over 980.00 frames. , ppl: 12.768505418178977] tot_loss[loss=2.355, over 5461633.29 frames. , ppl: 10.535817323957293], batch size: 70
+2022-12-10 08:59:47,494 INFO [train.py:421] (4/8) Epoch 2, batch 16400, loss[loss=2.329, over 3990.00 frames. , ppl: 10.271843632333853] tot_loss[loss=2.356, over 5432908.01 frames. , ppl: 10.550529470698244], batch size: 70
+2022-12-10 09:01:27,415 INFO [train.py:421] (4/8) Epoch 2, batch 16600, loss[loss=2.399, over 3080.00 frames. , ppl: 11.011931484439366] tot_loss[loss=2.356, over 5432822.90 frames. , ppl: 10.548045506765511], batch size: 70
+2022-12-10 09:03:04,477 INFO [train.py:421] (4/8) Epoch 2, batch 16800, loss[loss=2.381, over 6230.00 frames. , ppl: 10.815933369295767] tot_loss[loss=2.355, over 5479428.39 frames. , ppl: 10.536972278775979], batch size: 70
+2022-12-10 09:04:43,329 INFO [train.py:421] (4/8) Epoch 2, batch 17000, loss[loss=2.244, over 3150.00 frames. , ppl: 9.432987548026562] tot_loss[loss=2.355, over 5450557.65 frames. , ppl: 10.535306670457308], batch size: 70
+2022-12-10 09:04:43,330 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:04:44,079 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.34, over 211138.00 frames. , ppl: 10.380625913933697
+2022-12-10 09:06:22,856 INFO [train.py:421] (4/8) Epoch 2, batch 17200, loss[loss=2.399, over 1890.00 frames. , ppl: 11.009830961857151] tot_loss[loss=2.356, over 5429467.06 frames. , ppl: 10.549284453173936], batch size: 70
+2022-12-10 09:08:07,700 INFO [train.py:421] (4/8) Epoch 2, batch 17400, loss[loss=2.253, over 4900.00 frames. , ppl: 9.512070615814505] tot_loss[loss=2.355, over 5481147.84 frames. , ppl: 10.533653923458134], batch size: 70
+2022-12-10 09:09:46,649 INFO [train.py:421] (4/8) Epoch 2, batch 17600, loss[loss=2.468, over 1330.00 frames. , ppl: 11.797118505632113] tot_loss[loss=2.356, over 5421168.30 frames. , ppl: 10.548760490266087], batch size: 70
+2022-12-10 09:11:23,680 INFO [train.py:421] (4/8) Epoch 2, batch 17800, loss[loss=2.341, over 2170.00 frames. , ppl: 10.394006985753887] tot_loss[loss=2.356, over 5430965.28 frames. , ppl: 10.546173510297342], batch size: 70
+2022-12-10 09:13:01,348 INFO [train.py:421] (4/8) Epoch 2, batch 18000, loss[loss=2.802, over 910.00 frames. , ppl: 16.474453909241234] tot_loss[loss=2.354, over 5465819.09 frames. , ppl: 10.529613835007574], batch size: 70
+2022-12-10 09:13:01,349 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:13:02,108 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.339, over 211138.00 frames. , ppl: 10.371116820501767
+2022-12-10 09:14:39,901 INFO [train.py:421] (4/8) Epoch 2, batch 18200, loss[loss=2.351, over 1680.00 frames. , ppl: 10.500022080080138] tot_loss[loss=2.355, over 5431301.81 frames. , ppl: 10.536214895774696], batch size: 70
+2022-12-10 09:16:18,845 INFO [train.py:421] (4/8) Epoch 2, batch 18400, loss[loss=2.781, over 700.00 frames. , ppl: 16.134806445961747] tot_loss[loss=2.356, over 5394160.78 frames. , ppl: 10.550942298394933], batch size: 70
+2022-12-10 09:17:58,062 INFO [train.py:421] (4/8) Epoch 2, batch 18600, loss[loss=2.593, over 1050.00 frames. , ppl: 13.368340067765779] tot_loss[loss=2.356, over 5399560.88 frames. , ppl: 10.547293569550472], batch size: 70
+2022-12-10 09:19:35,458 INFO [train.py:421] (4/8) Epoch 2, batch 18800, loss[loss=2.349, over 3010.00 frames. , ppl: 10.473278950927137] tot_loss[loss=2.356, over 5385938.72 frames. , ppl: 10.549258813896628], batch size: 70
+2022-12-10 09:21:19,016 INFO [train.py:421] (4/8) Epoch 2, batch 19000, loss[loss=2.342, over 4200.00 frames. , ppl: 10.40368250746175] tot_loss[loss=2.354, over 5423630.78 frames. , ppl: 10.532657948802264], batch size: 70
+2022-12-10 09:21:19,017 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:21:19,781 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.337, over 211138.00 frames. , ppl: 10.346198405029323
+2022-12-10 09:23:00,682 INFO [train.py:421] (4/8) Epoch 2, batch 19200, loss[loss=2.29, over 4270.00 frames. , ppl: 9.876674114399789] tot_loss[loss=2.355, over 5420534.41 frames. , ppl: 10.538343869135725], batch size: 70
+2022-12-10 09:24:40,865 INFO [train.py:421] (4/8) Epoch 2, batch 19400, loss[loss=2.288, over 4340.00 frames. , ppl: 9.858543145165537] tot_loss[loss=2.355, over 5439063.01 frames. , ppl: 10.538514813602093], batch size: 70
+2022-12-10 09:26:20,689 INFO [train.py:421] (4/8) Epoch 2, batch 19600, loss[loss=3.632, over 420.00 frames. , ppl: 37.7868956829839] tot_loss[loss=2.355, over 5424973.31 frames. , ppl: 10.541567936425038], batch size: 70
+2022-12-10 09:28:00,785 INFO [train.py:421] (4/8) Epoch 2, batch 19800, loss[loss=2.391, over 4480.00 frames. , ppl: 10.929566553451235] tot_loss[loss=2.354, over 5455109.68 frames. , ppl: 10.52973380638619], batch size: 70
+2022-12-10 09:29:37,936 INFO [train.py:421] (4/8) Epoch 2, batch 20000, loss[loss=2.552, over 1190.00 frames. , ppl: 12.828914884241918] tot_loss[loss=2.355, over 5408796.93 frames. , ppl: 10.542653837825647], batch size: 70
+2022-12-10 09:29:37,936 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:29:38,694 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.339, over 211138.00 frames. , ppl: 10.367246266120773
+2022-12-10 09:31:19,562 INFO [train.py:421] (4/8) Epoch 2, batch 20200, loss[loss=2.493, over 1330.00 frames. , ppl: 12.101649345024416] tot_loss[loss=2.355, over 5414541.32 frames. , ppl: 10.542135867618791], batch size: 70
+2022-12-10 09:32:58,803 INFO [train.py:421] (4/8) Epoch 2, batch 20400, loss[loss=2.361, over 2940.00 frames. , ppl: 10.596627048426996] tot_loss[loss=2.355, over 5414560.59 frames. , ppl: 10.54074292757647], batch size: 70
+2022-12-10 09:34:39,613 INFO [train.py:421] (4/8) Epoch 2, batch 20600, loss[loss=2.289, over 3920.00 frames. , ppl: 9.860750297906664] tot_loss[loss=2.355, over 5421899.81 frames. , ppl: 10.542892497469927], batch size: 70
+2022-12-10 09:36:18,783 INFO [train.py:421] (4/8) Epoch 2, batch 20800, loss[loss=2.353, over 2870.00 frames. , ppl: 10.512302185838035] tot_loss[loss=2.355, over 5428612.62 frames. , ppl: 10.542029775835884], batch size: 70
+2022-12-10 09:37:55,972 INFO [train.py:421] (4/8) Epoch 2, batch 21000, loss[loss=2.525, over 1750.00 frames. , ppl: 12.494241487258792] tot_loss[loss=2.356, over 5414697.25 frames. , ppl: 10.543826663861644], batch size: 70
+2022-12-10 09:37:55,972 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:37:56,718 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.338, over 211138.00 frames. , ppl: 10.362088796314158
+2022-12-10 09:39:34,404 INFO [train.py:421] (4/8) Epoch 2, batch 21200, loss[loss=3.031, over 560.00 frames. , ppl: 20.727873031885718] tot_loss[loss=2.356, over 5407755.33 frames. , ppl: 10.548495534255524], batch size: 70
+2022-12-10 09:41:12,855 INFO [train.py:421] (4/8) Epoch 2, batch 21400, loss[loss=2.338, over 2870.00 frames. , ppl: 10.357918360663938] tot_loss[loss=2.356, over 5426091.93 frames. , ppl: 10.545739497690112], batch size: 70
+2022-12-10 09:42:51,082 INFO [train.py:421] (4/8) Epoch 2, batch 21600, loss[loss=3.282, over 490.00 frames. , ppl: 26.6246355052534] tot_loss[loss=2.356, over 5389537.07 frames. , ppl: 10.547792513913436], batch size: 70
+2022-12-10 09:44:31,993 INFO [train.py:421] (4/8) Epoch 2, batch 21800, loss[loss=2.702, over 840.00 frames. , ppl: 14.908016155408323] tot_loss[loss=2.356, over 5393002.87 frames. , ppl: 10.547581051053612], batch size: 70
+2022-12-10 09:46:13,455 INFO [train.py:421] (4/8) Epoch 2, batch 22000, loss[loss=2.471, over 1260.00 frames. , ppl: 11.83740296131368] tot_loss[loss=2.356, over 5398429.87 frames. , ppl: 10.546992645141625], batch size: 70
+2022-12-10 09:46:13,456 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:46:14,216 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.337, over 211138.00 frames. , ppl: 10.353811838414043
+2022-12-10 09:47:52,620 INFO [train.py:421] (4/8) Epoch 2, batch 22200, loss[loss=2.345, over 3710.00 frames. , ppl: 10.434396570713057] tot_loss[loss=2.355, over 5402506.65 frames. , ppl: 10.538779494337115], batch size: 70
+2022-12-10 09:49:28,392 INFO [train.py:421] (4/8) Epoch 2, batch 22400, loss[loss=2.447, over 1400.00 frames. , ppl: 11.550161194869366] tot_loss[loss=2.356, over 5367622.70 frames. , ppl: 10.546031011209923], batch size: 70
+2022-12-10 09:51:04,824 INFO [train.py:421] (4/8) Epoch 2, batch 22600, loss[loss=2.299, over 6160.00 frames. , ppl: 9.959382010919606] tot_loss[loss=2.355, over 5384932.21 frames. , ppl: 10.536751553186553], batch size: 70
+2022-12-10 09:52:46,947 INFO [train.py:421] (4/8) Epoch 2, batch 22800, loss[loss=2.473, over 1190.00 frames. , ppl: 11.862470117102788] tot_loss[loss=2.355, over 5399323.20 frames. , ppl: 10.535166291063307], batch size: 70
+2022-12-10 09:54:25,159 INFO [train.py:421] (4/8) Epoch 2, batch 23000, loss[loss=2.298, over 4620.00 frames. , ppl: 9.951116798400347] tot_loss[loss=2.354, over 5433723.28 frames. , ppl: 10.528804332115149], batch size: 70
+2022-12-10 09:54:25,160 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 09:54:25,907 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.337, over 211138.00 frames. , ppl: 10.35141844156331
+2022-12-10 09:56:05,788 INFO [train.py:421] (4/8) Epoch 2, batch 23200, loss[loss=2.255, over 7070.00 frames. , ppl: 9.53853863949403] tot_loss[loss=2.355, over 5414879.96 frames. , ppl: 10.535450803217566], batch size: 70
+2022-12-10 09:57:48,632 INFO [train.py:421] (4/8) Epoch 2, batch 23400, loss[loss=2.236, over 6510.00 frames. , ppl: 9.354038654741935] tot_loss[loss=2.354, over 5435264.64 frames. , ppl: 10.526208801491652], batch size: 70
+2022-12-10 09:59:32,032 INFO [train.py:421] (4/8) Epoch 2, batch 23600, loss[loss=2.305, over 5460.00 frames. , ppl: 10.02128994737437] tot_loss[loss=2.352, over 5492096.54 frames. , ppl: 10.50975134397624], batch size: 70
+2022-12-10 10:01:14,603 INFO [train.py:421] (4/8) Epoch 2, batch 23800, loss[loss=2.532, over 980.00 frames. , ppl: 12.584184263810021] tot_loss[loss=2.351, over 5514290.49 frames. , ppl: 10.49923263436533], batch size: 70
+2022-12-10 10:02:54,344 INFO [train.py:421] (4/8) Epoch 2, batch 24000, loss[loss=2.754, over 910.00 frames. , ppl: 15.704934012880988] tot_loss[loss=2.351, over 5507610.61 frames. , ppl: 10.496625352109403], batch size: 70
+2022-12-10 10:02:54,345 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:02:55,105 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.32757755196688
+2022-12-10 10:04:39,373 INFO [train.py:421] (4/8) Epoch 2, batch 24200, loss[loss=2.61, over 1190.00 frames. , ppl: 13.595649171052495] tot_loss[loss=2.35, over 5527159.10 frames. , ppl: 10.490073161889692], batch size: 70
+2022-12-10 10:06:16,094 INFO [train.py:421] (4/8) Epoch 2, batch 24400, loss[loss=2.391, over 1750.00 frames. , ppl: 10.923583842104122] tot_loss[loss=2.351, over 5493863.92 frames. , ppl: 10.498348889353014], batch size: 70
+2022-12-10 10:07:52,407 INFO [train.py:421] (4/8) Epoch 2, batch 24600, loss[loss=2.234, over 3360.00 frames. , ppl: 9.33284169858757] tot_loss[loss=2.352, over 5451144.33 frames. , ppl: 10.50637390042353], batch size: 70
+2022-12-10 10:09:36,242 INFO [train.py:421] (4/8) Epoch 2, batch 24800, loss[loss=2.378, over 2380.00 frames. , ppl: 10.777982719837079] tot_loss[loss=2.351, over 5469242.55 frames. , ppl: 10.498432995634174], batch size: 70
+2022-12-10 10:11:14,046 INFO [train.py:421] (4/8) Epoch 2, batch 25000, loss[loss=2.284, over 7630.00 frames. , ppl: 9.81230304016146] tot_loss[loss=2.35, over 5507628.47 frames. , ppl: 10.490771966989525], batch size: 70
+2022-12-10 10:11:14,046 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:11:14,793 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.336, over 211138.00 frames. , ppl: 10.337237971792588
+2022-12-10 10:12:54,463 INFO [train.py:421] (4/8) Epoch 2, batch 25200, loss[loss=2.28, over 5110.00 frames. , ppl: 9.781073638132627] tot_loss[loss=2.351, over 5499747.21 frames. , ppl: 10.496215653438037], batch size: 70
+2022-12-10 10:14:32,509 INFO [train.py:421] (4/8) Epoch 2, batch 25400, loss[loss=2.384, over 2730.00 frames. , ppl: 10.850791157584682] tot_loss[loss=2.348, over 5559547.07 frames. , ppl: 10.468013571939645], batch size: 70
+2022-12-10 10:16:11,533 INFO [train.py:421] (4/8) Epoch 2, batch 25600, loss[loss=2.202, over 4690.00 frames. , ppl: 9.043151032739416] tot_loss[loss=2.349, over 5533105.26 frames. , ppl: 10.471040623669639], batch size: 70
+2022-12-10 10:17:49,159 INFO [train.py:421] (4/8) Epoch 2, batch 25800, loss[loss=2.63, over 980.00 frames. , ppl: 13.870443980460093] tot_loss[loss=2.349, over 5511785.42 frames. , ppl: 10.476224611634663], batch size: 70
+2022-12-10 10:19:31,238 INFO [train.py:421] (4/8) Epoch 2, batch 26000, loss[loss=2.497, over 1120.00 frames. , ppl: 12.151924245835165] tot_loss[loss=2.35, over 5511353.59 frames. , ppl: 10.484559847049658], batch size: 70
+2022-12-10 10:19:31,238 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:19:31,994 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.33015502328756
+2022-12-10 10:21:14,003 INFO [train.py:421] (4/8) Epoch 2, batch 26200, loss[loss=2.22, over 4900.00 frames. , ppl: 9.211306348940663] tot_loss[loss=2.348, over 5575336.27 frames. , ppl: 10.464804159895706], batch size: 70
+2022-12-10 10:22:55,797 INFO [train.py:421] (4/8) Epoch 2, batch 26400, loss[loss=2.397, over 2520.00 frames. , ppl: 10.989136177561923] tot_loss[loss=2.347, over 5630781.26 frames. , ppl: 10.452180800527618], batch size: 70
+2022-12-10 10:24:38,308 INFO [train.py:421] (4/8) Epoch 2, batch 26600, loss[loss=2.489, over 1330.00 frames. , ppl: 12.048901226735735] tot_loss[loss=2.347, over 5644381.76 frames. , ppl: 10.450505940083746], batch size: 70
+2022-12-10 10:26:19,546 INFO [train.py:421] (4/8) Epoch 2, batch 26800, loss[loss=2.38, over 2520.00 frames. , ppl: 10.804415488587173] tot_loss[loss=2.347, over 5639240.92 frames. , ppl: 10.453385111586586], batch size: 70
+2022-12-10 10:27:59,904 INFO [train.py:421] (4/8) Epoch 2, batch 27000, loss[loss=2.315, over 3500.00 frames. , ppl: 10.127438924341586] tot_loss[loss=2.348, over 5591324.47 frames. , ppl: 10.466932759011598], batch size: 70
+2022-12-10 10:27:59,905 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:28:00,666 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.332974773515254
+2022-12-10 10:29:43,680 INFO [train.py:421] (4/8) Epoch 2, batch 27200, loss[loss=2.314, over 4620.00 frames. , ppl: 10.11386595502432] tot_loss[loss=2.348, over 5576810.04 frames. , ppl: 10.466786300185852], batch size: 70
+2022-12-10 10:31:24,575 INFO [train.py:421] (4/8) Epoch 2, batch 27400, loss[loss=2.27, over 1890.00 frames. , ppl: 9.681515613240158] tot_loss[loss=2.349, over 5559054.77 frames. , ppl: 10.47085068551953], batch size: 70
+2022-12-10 10:33:03,927 INFO [train.py:421] (4/8) Epoch 2, batch 27600, loss[loss=3.092, over 630.00 frames. , ppl: 22.01675711453804] tot_loss[loss=2.349, over 5563573.58 frames. , ppl: 10.475854300063919], batch size: 70
+2022-12-10 10:34:44,431 INFO [train.py:421] (4/8) Epoch 2, batch 27800, loss[loss=2.252, over 10080.00 frames. , ppl: 9.502421030769886] tot_loss[loss=2.349, over 5591107.63 frames. , ppl: 10.471734493487533], batch size: 70
+2022-12-10 10:36:25,384 INFO [train.py:421] (4/8) Epoch 2, batch 28000, loss[loss=2.496, over 1400.00 frames. , ppl: 12.13875147772927] tot_loss[loss=2.348, over 5621545.33 frames. , ppl: 10.464536301861072], batch size: 70
+2022-12-10 10:36:25,385 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:36:26,133 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.333, over 211138.00 frames. , ppl: 10.313298466854016
+2022-12-10 10:38:05,396 INFO [train.py:421] (4/8) Epoch 2, batch 28200, loss[loss=2.506, over 1260.00 frames. , ppl: 12.253316021379993] tot_loss[loss=2.349, over 5618374.67 frames. , ppl: 10.472423991822177], batch size: 70
+2022-12-10 10:39:46,467 INFO [train.py:421] (4/8) Epoch 2, batch 28400, loss[loss=2.293, over 1960.00 frames. , ppl: 9.906325863341054] tot_loss[loss=2.348, over 5635083.23 frames. , ppl: 10.460729766499515], batch size: 70
+2022-12-10 10:41:25,661 INFO [train.py:421] (4/8) Epoch 2, batch 28600, loss[loss=2.627, over 1610.00 frames. , ppl: 13.827339033237346] tot_loss[loss=2.347, over 5646658.66 frames. , ppl: 10.454384310547045], batch size: 70
+2022-12-10 10:43:03,861 INFO [train.py:421] (4/8) Epoch 2, batch 28800, loss[loss=3.993, over 350.00 frames. , ppl: 54.205733709479105] tot_loss[loss=2.345, over 5719248.38 frames. , ppl: 10.434732916857989], batch size: 70
+2022-12-10 10:44:44,078 INFO [train.py:421] (4/8) Epoch 2, batch 29000, loss[loss=2.289, over 7560.00 frames. , ppl: 9.861749705638916] tot_loss[loss=2.346, over 5687784.67 frames. , ppl: 10.448193571633178], batch size: 70
+2022-12-10 10:44:44,079 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:44:44,826 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.335, over 211138.00 frames. , ppl: 10.328367846981726
+2022-12-10 10:46:24,218 INFO [train.py:421] (4/8) Epoch 2, batch 29200, loss[loss=2.507, over 980.00 frames. , ppl: 12.2690102396226] tot_loss[loss=2.346, over 5658759.18 frames. , ppl: 10.443584861119376], batch size: 70
+2022-12-10 10:48:02,851 INFO [train.py:421] (4/8) Epoch 2, batch 29400, loss[loss=2.33, over 4690.00 frames. , ppl: 10.273074394459252] tot_loss[loss=2.347, over 5623814.71 frames. , ppl: 10.455073555102775], batch size: 70
+2022-12-10 10:49:40,055 INFO [train.py:421] (4/8) Epoch 2, batch 29600, loss[loss=2.12, over 1680.00 frames. , ppl: 8.333372005512459] tot_loss[loss=2.347, over 5616739.16 frames. , ppl: 10.457100970656743], batch size: 70
+2022-12-10 10:51:22,089 INFO [train.py:421] (4/8) Epoch 2, batch 29800, loss[loss=2.322, over 5950.00 frames. , ppl: 10.191668189039708] tot_loss[loss=2.347, over 5657291.20 frames. , ppl: 10.454439626235345], batch size: 70
+2022-12-10 10:53:03,452 INFO [train.py:421] (4/8) Epoch 2, batch 30000, loss[loss=2.472, over 1610.00 frames. , ppl: 11.85080242643778] tot_loss[loss=2.348, over 5627154.74 frames. , ppl: 10.46236491613724], batch size: 70
+2022-12-10 10:53:03,453 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 10:53:04,215 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.333, over 211138.00 frames. , ppl: 10.311935441408778
+2022-12-10 10:54:45,306 INFO [train.py:421] (4/8) Epoch 2, batch 30200, loss[loss=2.734, over 910.00 frames. , ppl: 15.396124706796996] tot_loss[loss=2.348, over 5587751.59 frames. , ppl: 10.469763203413386], batch size: 70
+2022-12-10 10:56:29,753 INFO [train.py:421] (4/8) Epoch 2, batch 30400, loss[loss=2.25, over 4130.00 frames. , ppl: 9.491548190125659] tot_loss[loss=2.347, over 5627175.36 frames. , ppl: 10.45709491081922], batch size: 70
+2022-12-10 10:58:09,981 INFO [train.py:421] (4/8) Epoch 2, batch 30600, loss[loss=2.237, over 3220.00 frames. , ppl: 9.367920983427988] tot_loss[loss=2.346, over 5666269.06 frames. , ppl: 10.444143513498911], batch size: 70
+2022-12-10 10:59:47,280 INFO [train.py:421] (4/8) Epoch 2, batch 30800, loss[loss=2.985, over 560.00 frames. , ppl: 19.789330979501567] tot_loss[loss=2.348, over 5595033.29 frames. , ppl: 10.464174522008332], batch size: 70
+2022-12-10 11:01:28,601 INFO [train.py:421] (4/8) Epoch 2, batch 31000, loss[loss=2.286, over 6860.00 frames. , ppl: 9.833653351847506] tot_loss[loss=2.347, over 5631675.03 frames. , ppl: 10.451352231144755], batch size: 70
+2022-12-10 11:01:28,601 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:01:29,350 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.334, over 211138.00 frames. , ppl: 10.318448451765484
+2022-12-10 11:03:12,445 INFO [train.py:421] (4/8) Epoch 2, batch 31200, loss[loss=2.226, over 6160.00 frames. , ppl: 9.262056878002777] tot_loss[loss=2.346, over 5671901.10 frames. , ppl: 10.448006929167018], batch size: 70
+2022-12-10 11:04:53,783 INFO [train.py:421] (4/8) Epoch 2, batch 31400, loss[loss=2.691, over 630.00 frames. , ppl: 14.74611963700694] tot_loss[loss=2.346, over 5695154.46 frames. , ppl: 10.440692250172768], batch size: 70
+2022-12-10 11:06:34,443 INFO [train.py:421] (4/8) Epoch 2, batch 31600, loss[loss=2.735, over 630.00 frames. , ppl: 15.417306604405324] tot_loss[loss=2.347, over 5652011.83 frames. , ppl: 10.45317472722326], batch size: 70
+2022-12-10 11:08:17,186 INFO [train.py:421] (4/8) Epoch 2, batch 31800, loss[loss=2.319, over 4200.00 frames. , ppl: 10.162345924885011] tot_loss[loss=2.346, over 5640320.46 frames. , ppl: 10.446719664333113], batch size: 70
+2022-12-10 11:09:56,529 INFO [train.py:421] (4/8) Epoch 2, batch 32000, loss[loss=2.245, over 8470.00 frames. , ppl: 9.436813915679535] tot_loss[loss=2.347, over 5597968.76 frames. , ppl: 10.45831567393895], batch size: 70
+2022-12-10 11:09:56,530 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:09:57,291 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.334, over 211138.00 frames. , ppl: 10.31484792506888
+2022-12-10 11:11:36,906 INFO [train.py:421] (4/8) Epoch 2, batch 32200, loss[loss=2.269, over 3500.00 frames. , ppl: 9.670141756323009] tot_loss[loss=2.347, over 5603735.79 frames. , ppl: 10.454264027421054], batch size: 70
+2022-12-10 11:13:14,855 INFO [train.py:421] (4/8) Epoch 2, batch 32400, loss[loss=2.257, over 3220.00 frames. , ppl: 9.552152161379158] tot_loss[loss=2.348, over 5558177.78 frames. , ppl: 10.463832977917258], batch size: 70
+2022-12-10 11:14:55,253 INFO [train.py:421] (4/8) Epoch 2, batch 32600, loss[loss=2.335, over 2800.00 frames. , ppl: 10.333881302016014] tot_loss[loss=2.349, over 5507571.30 frames. , ppl: 10.47034322571804], batch size: 70
+2022-12-10 11:16:38,284 INFO [train.py:421] (4/8) Epoch 2, batch 32800, loss[loss=2.378, over 3290.00 frames. , ppl: 10.77822059252135] tot_loss[loss=2.348, over 5533818.92 frames. , ppl: 10.459743509190671], batch size: 70
+2022-12-10 11:18:17,211 INFO [train.py:421] (4/8) Epoch 2, batch 33000, loss[loss=2.582, over 1260.00 frames. , ppl: 13.220072831857033] tot_loss[loss=2.347, over 5548865.67 frames. , ppl: 10.454360258647728], batch size: 70
+2022-12-10 11:18:17,212 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:18:17,973 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.332, over 211138.00 frames. , ppl: 10.303007731813349
+2022-12-10 11:20:00,440 INFO [train.py:421] (4/8) Epoch 2, batch 33200, loss[loss=2.376, over 1400.00 frames. , ppl: 10.763180199488598] tot_loss[loss=2.345, over 5596394.64 frames. , ppl: 10.433884569027192], batch size: 70
+2022-12-10 11:21:39,783 INFO [train.py:421] (4/8) Epoch 2, batch 33400, loss[loss=2.272, over 11830.00 frames. , ppl: 9.696993552657752] tot_loss[loss=2.345, over 5607881.62 frames. , ppl: 10.430086303111738], batch size: 70
+2022-12-10 11:23:22,971 INFO [train.py:421] (4/8) Epoch 2, batch 33600, loss[loss=2.322, over 3570.00 frames. , ppl: 10.195284736120765] tot_loss[loss=2.344, over 5631831.91 frames. , ppl: 10.423929045701888], batch size: 70
+2022-12-10 11:25:04,054 INFO [train.py:421] (4/8) Epoch 2, batch 33800, loss[loss=2.944, over 630.00 frames. , ppl: 18.987281802945322] tot_loss[loss=2.345, over 5609282.02 frames. , ppl: 10.431236902024331], batch size: 70
+2022-12-10 11:26:42,438 INFO [train.py:421] (4/8) Epoch 2, batch 34000, loss[loss=2.628, over 840.00 frames. , ppl: 13.851210483545675] tot_loss[loss=2.344, over 5584987.84 frames. , ppl: 10.427540222514512], batch size: 70
+2022-12-10 11:26:42,438 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:26:43,186 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.291255765342557
+2022-12-10 11:28:20,403 INFO [train.py:421] (4/8) Epoch 2, batch 34200, loss[loss=2.499, over 1540.00 frames. , ppl: 12.170153249755673] tot_loss[loss=2.344, over 5596561.31 frames. , ppl: 10.426834025582949], batch size: 70
+2022-12-10 11:30:00,226 INFO [train.py:421] (4/8) Epoch 2, batch 34400, loss[loss=2.352, over 2240.00 frames. , ppl: 10.505661728503672] tot_loss[loss=2.345, over 5568769.03 frames. , ppl: 10.430548113791518], batch size: 70
+2022-12-10 11:31:39,796 INFO [train.py:421] (4/8) Epoch 2, batch 34600, loss[loss=2.356, over 2170.00 frames. , ppl: 10.545316615287746] tot_loss[loss=2.345, over 5573810.39 frames. , ppl: 10.433171582410322], batch size: 70
+2022-12-10 11:33:17,810 INFO [train.py:421] (4/8) Epoch 2, batch 34800, loss[loss=2.32, over 3150.00 frames. , ppl: 10.173124082330325] tot_loss[loss=2.345, over 5551620.59 frames. , ppl: 10.436686463563579], batch size: 70
+2022-12-10 11:35:00,057 INFO [train.py:421] (4/8) Epoch 2, batch 35000, loss[loss=2.255, over 4830.00 frames. , ppl: 9.535413612548734] tot_loss[loss=2.346, over 5539674.02 frames. , ppl: 10.440822447020254], batch size: 70
+2022-12-10 11:35:00,058 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:35:00,819 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.332, over 211138.00 frames. , ppl: 10.297171983552662
+2022-12-10 11:36:40,680 INFO [train.py:421] (4/8) Epoch 2, batch 35200, loss[loss=2.424, over 2310.00 frames. , ppl: 11.29399980845919] tot_loss[loss=2.345, over 5567390.10 frames. , ppl: 10.432295065960776], batch size: 70
+2022-12-10 11:38:22,509 INFO [train.py:421] (4/8) Epoch 2, batch 35400, loss[loss=2.652, over 910.00 frames. , ppl: 14.183647021456817] tot_loss[loss=2.346, over 5555502.82 frames. , ppl: 10.443202480425516], batch size: 70
+2022-12-10 11:40:03,279 INFO [train.py:421] (4/8) Epoch 2, batch 35600, loss[loss=2.293, over 7560.00 frames. , ppl: 9.9018925063044] tot_loss[loss=2.345, over 5589777.73 frames. , ppl: 10.436709780941062], batch size: 70
+2022-12-10 11:41:42,466 INFO [train.py:421] (4/8) Epoch 2, batch 35800, loss[loss=2.563, over 1330.00 frames. , ppl: 12.969968529242943] tot_loss[loss=2.346, over 5543344.10 frames. , ppl: 10.444724038891481], batch size: 70
+2022-12-10 11:43:21,034 INFO [train.py:421] (4/8) Epoch 2, batch 36000, loss[loss=2.492, over 1820.00 frames. , ppl: 12.080381612566198] tot_loss[loss=2.346, over 5530394.37 frames. , ppl: 10.442763161188674], batch size: 70
+2022-12-10 11:43:21,035 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:43:21,794 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.288832667020564
+2022-12-10 11:45:02,829 INFO [train.py:421] (4/8) Epoch 2, batch 36200, loss[loss=2.758, over 700.00 frames. , ppl: 15.770658520837067] tot_loss[loss=2.348, over 5460306.71 frames. , ppl: 10.46628381752426], batch size: 70
+2022-12-10 11:46:46,966 INFO [train.py:421] (4/8) Epoch 2, batch 36400, loss[loss=2.347, over 2240.00 frames. , ppl: 10.455095790163806] tot_loss[loss=2.348, over 5445058.58 frames. , ppl: 10.468484742539141], batch size: 70
+2022-12-10 11:48:25,554 INFO [train.py:421] (4/8) Epoch 2, batch 36600, loss[loss=2.345, over 2800.00 frames. , ppl: 10.438273687409643] tot_loss[loss=2.347, over 5493498.83 frames. , ppl: 10.453040045472207], batch size: 70
+2022-12-10 11:50:04,140 INFO [train.py:421] (4/8) Epoch 2, batch 36800, loss[loss=2.277, over 3850.00 frames. , ppl: 9.743758515415582] tot_loss[loss=2.346, over 5512251.84 frames. , ppl: 10.44445434969955], batch size: 70
+2022-12-10 11:51:43,180 INFO [train.py:421] (4/8) Epoch 2, batch 37000, loss[loss=2.542, over 840.00 frames. , ppl: 12.708394720237385] tot_loss[loss=2.346, over 5530940.03 frames. , ppl: 10.44040894059515], batch size: 70
+2022-12-10 11:51:43,181 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 11:51:43,927 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.333, over 211138.00 frames. , ppl: 10.313814418221853
+2022-12-10 11:53:27,750 INFO [train.py:421] (4/8) Epoch 2, batch 37200, loss[loss=2.435, over 1540.00 frames. , ppl: 11.420765196490812] tot_loss[loss=2.347, over 5482881.66 frames. , ppl: 10.452318048622535], batch size: 70
+2022-12-10 11:55:05,711 INFO [train.py:421] (4/8) Epoch 2, batch 37400, loss[loss=2.648, over 770.00 frames. , ppl: 14.12790588570118] tot_loss[loss=2.348, over 5447475.59 frames. , ppl: 10.461807244993476], batch size: 70
+2022-12-10 11:56:46,031 INFO [train.py:421] (4/8) Epoch 2, batch 37600, loss[loss=2.274, over 4620.00 frames. , ppl: 9.71684619063349] tot_loss[loss=2.348, over 5447465.88 frames. , ppl: 10.468437643803206], batch size: 70
+2022-12-10 11:58:28,997 INFO [train.py:421] (4/8) Epoch 2, batch 37800, loss[loss=2.279, over 4970.00 frames. , ppl: 9.765524802367445] tot_loss[loss=2.347, over 5456874.44 frames. , ppl: 10.458027334310543], batch size: 70
+2022-12-10 12:00:10,577 INFO [train.py:421] (4/8) Epoch 2, batch 38000, loss[loss=3.027, over 700.00 frames. , ppl: 20.631409805516142] tot_loss[loss=2.347, over 5477324.09 frames. , ppl: 10.459351296137369], batch size: 70
+2022-12-10 12:00:10,578 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:00:11,327 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.269646244641864
+2022-12-10 12:01:54,474 INFO [train.py:421] (4/8) Epoch 2, batch 38200, loss[loss=2.508, over 2730.00 frames. , ppl: 12.279188821641768] tot_loss[loss=2.347, over 5508857.48 frames. , ppl: 10.452744065335475], batch size: 70
+2022-12-10 12:03:35,690 INFO [train.py:421] (4/8) Epoch 2, batch 38400, loss[loss=2.371, over 3220.00 frames. , ppl: 10.712564983661125] tot_loss[loss=2.346, over 5549387.06 frames. , ppl: 10.448060857323433], batch size: 70
+2022-12-10 12:05:13,549 INFO [train.py:421] (4/8) Epoch 2, batch 38600, loss[loss=2.365, over 2660.00 frames. , ppl: 10.638949821740594] tot_loss[loss=2.346, over 5546309.85 frames. , ppl: 10.44328815667666], batch size: 70
+2022-12-10 12:06:50,161 INFO [train.py:421] (4/8) Epoch 2, batch 38800, loss[loss=2.243, over 9100.00 frames. , ppl: 9.420126879814497] tot_loss[loss=2.346, over 5523178.12 frames. , ppl: 10.447076520258687], batch size: 70
+2022-12-10 12:08:30,349 INFO [train.py:421] (4/8) Epoch 2, batch 39000, loss[loss=2.365, over 3780.00 frames. , ppl: 10.643508652689071] tot_loss[loss=2.345, over 5548966.13 frames. , ppl: 10.43328060730669], batch size: 70
+2022-12-10 12:08:30,350 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:08:31,096 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.258752219498415
+2022-12-10 12:10:14,098 INFO [train.py:421] (4/8) Epoch 2, batch 39200, loss[loss=2.207, over 5740.00 frames. , ppl: 9.092551494821787] tot_loss[loss=2.344, over 5570893.27 frames. , ppl: 10.419760159209348], batch size: 70
+2022-12-10 12:11:55,246 INFO [train.py:421] (4/8) Epoch 2, batch 39400, loss[loss=2.313, over 2520.00 frames. , ppl: 10.109353268574292] tot_loss[loss=2.343, over 5593813.16 frames. , ppl: 10.413237423209052], batch size: 70
+2022-12-10 12:13:32,407 INFO [train.py:421] (4/8) Epoch 2, batch 39600, loss[loss=2.374, over 1960.00 frames. , ppl: 10.7453366349661] tot_loss[loss=2.343, over 5569083.19 frames. , ppl: 10.415287333652264], batch size: 70
+2022-12-10 12:15:17,688 INFO [train.py:421] (4/8) Epoch 2, batch 39800, loss[loss=2.376, over 3080.00 frames. , ppl: 10.761858566805506] tot_loss[loss=2.344, over 5550006.48 frames. , ppl: 10.418516264664191], batch size: 70
+2022-12-10 12:16:56,231 INFO [train.py:421] (4/8) Epoch 2, batch 40000, loss[loss=2.223, over 7560.00 frames. , ppl: 9.233055882356776] tot_loss[loss=2.345, over 5496721.65 frames. , ppl: 10.435985744009766], batch size: 70
+2022-12-10 12:16:56,231 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:16:56,977 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.292756209643162
+2022-12-10 12:18:37,689 INFO [train.py:421] (4/8) Epoch 2, batch 40200, loss[loss=2.464, over 1680.00 frames. , ppl: 11.752035506628326] tot_loss[loss=2.347, over 5445554.60 frames. , ppl: 10.453147103112236], batch size: 70
+2022-12-10 12:20:17,609 INFO [train.py:421] (4/8) Epoch 2, batch 40400, loss[loss=2.401, over 3360.00 frames. , ppl: 11.02879730596489] tot_loss[loss=2.346, over 5455818.06 frames. , ppl: 10.448090200900465], batch size: 70
+2022-12-10 12:22:00,252 INFO [train.py:421] (4/8) Epoch 2, batch 40600, loss[loss=2.434, over 2030.00 frames. , ppl: 11.40644429633392] tot_loss[loss=2.345, over 5478317.77 frames. , ppl: 10.437427807950705], batch size: 70
+2022-12-10 12:23:41,676 INFO [train.py:421] (4/8) Epoch 2, batch 40800, loss[loss=2.428, over 2870.00 frames. , ppl: 11.330739329855305] tot_loss[loss=2.345, over 5468515.26 frames. , ppl: 10.433144694847728], batch size: 70
+2022-12-10 12:25:27,337 INFO [train.py:421] (4/8) Epoch 2, batch 41000, loss[loss=2.506, over 1190.00 frames. , ppl: 12.25998467863248] tot_loss[loss=2.346, over 5433142.23 frames. , ppl: 10.447702330489324], batch size: 70
+2022-12-10 12:25:27,338 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:25:28,086 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.27130012072034
+2022-12-10 12:27:08,687 INFO [train.py:421] (4/8) Epoch 2, batch 41200, loss[loss=2.455, over 1400.00 frames. , ppl: 11.643277838493509] tot_loss[loss=2.346, over 5452958.16 frames. , ppl: 10.444873132456394], batch size: 70
+2022-12-10 12:28:47,117 INFO [train.py:421] (4/8) Epoch 2, batch 41400, loss[loss=2.439, over 1470.00 frames. , ppl: 11.458774248406213] tot_loss[loss=2.345, over 5486580.44 frames. , ppl: 10.436024906094213], batch size: 70
+2022-12-10 12:30:25,486 INFO [train.py:421] (4/8) Epoch 2, batch 41600, loss[loss=2.337, over 2660.00 frames. , ppl: 10.34616535488048] tot_loss[loss=2.346, over 5435620.20 frames. , ppl: 10.446898619502871], batch size: 70
+2022-12-10 12:32:02,259 INFO [train.py:421] (4/8) Epoch 2, batch 41800, loss[loss=2.401, over 1610.00 frames. , ppl: 11.038873427940365] tot_loss[loss=2.345, over 5453910.61 frames. , ppl: 10.435232486231865], batch size: 70
+2022-12-10 12:33:43,236 INFO [train.py:421] (4/8) Epoch 2, batch 42000, loss[loss=2.314, over 6510.00 frames. , ppl: 10.111184355440704] tot_loss[loss=2.346, over 5440269.14 frames. , ppl: 10.444843616448525], batch size: 70
+2022-12-10 12:33:43,237 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:33:43,997 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.331, over 211138.00 frames. , ppl: 10.283807052673856
+2022-12-10 12:35:25,662 INFO [train.py:421] (4/8) Epoch 2, batch 42200, loss[loss=2.154, over 6440.00 frames. , ppl: 8.619436202762381] tot_loss[loss=2.344, over 5522583.70 frames. , ppl: 10.421367011057693], batch size: 70
+2022-12-10 12:37:03,124 INFO [train.py:421] (4/8) Epoch 2, batch 42400, loss[loss=2.436, over 2660.00 frames. , ppl: 11.427310554809715] tot_loss[loss=2.346, over 5464804.95 frames. , ppl: 10.443703206218036], batch size: 70
+2022-12-10 12:38:43,240 INFO [train.py:421] (4/8) Epoch 2, batch 42600, loss[loss=2.438, over 2170.00 frames. , ppl: 11.454962120292011] tot_loss[loss=2.346, over 5433614.50 frames. , ppl: 10.448584276579481], batch size: 70
+2022-12-10 12:40:22,491 INFO [train.py:421] (4/8) Epoch 2, batch 42800, loss[loss=2.543, over 1540.00 frames. , ppl: 12.724095217175996] tot_loss[loss=2.346, over 5429232.43 frames. , ppl: 10.445136473767823], batch size: 70
+2022-12-10 12:42:04,985 INFO [train.py:421] (4/8) Epoch 2, batch 43000, loss[loss=2.427, over 2310.00 frames. , ppl: 11.320625595902271] tot_loss[loss=2.346, over 5450953.92 frames. , ppl: 10.439856492050012], batch size: 70
+2022-12-10 12:42:04,985 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:42:05,730 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.261654238592559
+2022-12-10 12:43:44,226 INFO [train.py:421] (4/8) Epoch 2, batch 43200, loss[loss=2.472, over 1330.00 frames. , ppl: 11.840413087268132] tot_loss[loss=2.346, over 5411813.21 frames. , ppl: 10.448329018630186], batch size: 70
+2022-12-10 12:45:21,821 INFO [train.py:421] (4/8) Epoch 2, batch 43400, loss[loss=2.307, over 3640.00 frames. , ppl: 10.045142656042175] tot_loss[loss=2.346, over 5410454.28 frames. , ppl: 10.447919732939516], batch size: 70
+2022-12-10 12:47:02,706 INFO [train.py:421] (4/8) Epoch 2, batch 43600, loss[loss=2.484, over 1260.00 frames. , ppl: 11.99229036993723] tot_loss[loss=2.345, over 5455504.66 frames. , ppl: 10.42897265752518], batch size: 70
+2022-12-10 12:48:41,287 INFO [train.py:421] (4/8) Epoch 2, batch 43800, loss[loss=2.403, over 1330.00 frames. , ppl: 11.057596495513387] tot_loss[loss=2.343, over 5462543.13 frames. , ppl: 10.417483652293445], batch size: 70
+2022-12-10 12:50:21,826 INFO [train.py:421] (4/8) Epoch 2, batch 44000, loss[loss=2.294, over 3710.00 frames. , ppl: 9.913958693952864] tot_loss[loss=2.344, over 5468663.94 frames. , ppl: 10.421348615475901], batch size: 70
+2022-12-10 12:50:21,827 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:50:22,572 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.270837981475065
+2022-12-10 12:52:01,323 INFO [train.py:421] (4/8) Epoch 2, batch 44200, loss[loss=2.659, over 770.00 frames. , ppl: 14.282931554384463] tot_loss[loss=2.345, over 5457526.86 frames. , ppl: 10.430707905377831], batch size: 70
+2022-12-10 12:53:43,812 INFO [train.py:421] (4/8) Epoch 2, batch 44400, loss[loss=2.405, over 1330.00 frames. , ppl: 11.078405065510857] tot_loss[loss=2.344, over 5452330.66 frames. , ppl: 10.422876884586378], batch size: 70
+2022-12-10 12:55:21,483 INFO [train.py:421] (4/8) Epoch 2, batch 44600, loss[loss=2.487, over 1260.00 frames. , ppl: 12.023330352580988] tot_loss[loss=2.345, over 5422554.67 frames. , ppl: 10.428920727628864], batch size: 70
+2022-12-10 12:57:01,753 INFO [train.py:421] (4/8) Epoch 2, batch 44800, loss[loss=2.509, over 840.00 frames. , ppl: 12.292739322044028] tot_loss[loss=2.344, over 5460732.26 frames. , ppl: 10.420946201740854], batch size: 70
+2022-12-10 12:58:44,440 INFO [train.py:421] (4/8) Epoch 2, batch 45000, loss[loss=3.558, over 420.00 frames. , ppl: 35.10504874712361] tot_loss[loss=2.344, over 5456343.75 frames. , ppl: 10.420838294209945], batch size: 70
+2022-12-10 12:58:44,441 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 12:58:45,186 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.267509369063996
+2022-12-10 13:00:27,504 INFO [train.py:421] (4/8) Epoch 2, batch 45200, loss[loss=2.3, over 6160.00 frames. , ppl: 9.970912995140644] tot_loss[loss=2.343, over 5476801.36 frames. , ppl: 10.416774428239435], batch size: 70
+2022-12-10 13:02:10,365 INFO [train.py:421] (4/8) Epoch 2, batch 45400, loss[loss=2.344, over 4410.00 frames. , ppl: 10.427478074697332] tot_loss[loss=2.343, over 5460037.30 frames. , ppl: 10.415560937192513], batch size: 70
+2022-12-10 13:03:49,609 INFO [train.py:421] (4/8) Epoch 2, batch 45600, loss[loss=2.525, over 980.00 frames. , ppl: 12.492933639510532] tot_loss[loss=2.342, over 5511036.39 frames. , ppl: 10.401159637235022], batch size: 70
+2022-12-10 13:05:28,544 INFO [train.py:421] (4/8) Epoch 2, batch 45800, loss[loss=2.511, over 1400.00 frames. , ppl: 12.314061720861543] tot_loss[loss=2.342, over 5523558.96 frames. , ppl: 10.39765048624421], batch size: 70
+2022-12-10 13:07:05,423 INFO [train.py:421] (4/8) Epoch 2, batch 46000, loss[loss=2.273, over 6160.00 frames. , ppl: 9.710455469009338] tot_loss[loss=2.342, over 5544516.96 frames. , ppl: 10.398966278339335], batch size: 70
+2022-12-10 13:07:05,423 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:07:06,169 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.269302733935248
+2022-12-10 13:08:48,310 INFO [train.py:421] (4/8) Epoch 2, batch 46200, loss[loss=2.482, over 2030.00 frames. , ppl: 11.969448108148464] tot_loss[loss=2.341, over 5564769.46 frames. , ppl: 10.390654651256384], batch size: 70
+2022-12-10 13:10:28,307 INFO [train.py:421] (4/8) Epoch 2, batch 46400, loss[loss=2.385, over 2870.00 frames. , ppl: 10.86308057546154] tot_loss[loss=2.341, over 5544270.56 frames. , ppl: 10.39363575318864], batch size: 70
+2022-12-10 13:12:09,472 INFO [train.py:421] (4/8) Epoch 2, batch 46600, loss[loss=2.397, over 3150.00 frames. , ppl: 10.990595258038962] tot_loss[loss=2.341, over 5517883.01 frames. , ppl: 10.396315859812814], batch size: 70
+2022-12-10 13:13:46,920 INFO [train.py:421] (4/8) Epoch 2, batch 46800, loss[loss=2.422, over 3360.00 frames. , ppl: 11.270774168033423] tot_loss[loss=2.343, over 5488362.57 frames. , ppl: 10.40817720790898], batch size: 70
+2022-12-10 13:15:22,163 INFO [train.py:421] (4/8) Epoch 2, batch 47000, loss[loss=2.355, over 2800.00 frames. , ppl: 10.533080061383979] tot_loss[loss=2.342, over 5497905.41 frames. , ppl: 10.404178513031605], batch size: 70
+2022-12-10 13:15:22,164 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:15:22,923 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.265565897614955
+2022-12-10 13:17:01,354 INFO [train.py:421] (4/8) Epoch 2, batch 47200, loss[loss=2.277, over 4200.00 frames. , ppl: 9.74968910725001] tot_loss[loss=2.343, over 5471895.31 frames. , ppl: 10.40920068008903], batch size: 70
+2022-12-10 13:18:45,163 INFO [train.py:421] (4/8) Epoch 2, batch 47400, loss[loss=2.363, over 2240.00 frames. , ppl: 10.621253657199166] tot_loss[loss=2.341, over 5546483.24 frames. , ppl: 10.396524515979186], batch size: 70
+2022-12-10 13:20:26,219 INFO [train.py:421] (4/8) Epoch 2, batch 47600, loss[loss=2.278, over 4270.00 frames. , ppl: 9.758187768691293] tot_loss[loss=2.341, over 5541184.49 frames. , ppl: 10.389634312849669], batch size: 70
+2022-12-10 13:22:08,948 INFO [train.py:421] (4/8) Epoch 2, batch 47800, loss[loss=2.496, over 1610.00 frames. , ppl: 12.12790536220277] tot_loss[loss=2.341, over 5525144.27 frames. , ppl: 10.39531975504538], batch size: 70
+2022-12-10 13:23:44,809 INFO [train.py:421] (4/8) Epoch 2, batch 48000, loss[loss=2.415, over 1050.00 frames. , ppl: 11.190467658495132] tot_loss[loss=2.341, over 5544698.99 frames. , ppl: 10.389162380058494], batch size: 70
+2022-12-10 13:23:44,809 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:23:45,556 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.26050912621634
+2022-12-10 13:25:25,335 INFO [train.py:421] (4/8) Epoch 2, batch 48200, loss[loss=2.339, over 2520.00 frames. , ppl: 10.37508643403529] tot_loss[loss=2.342, over 5502005.09 frames. , ppl: 10.398947555102382], batch size: 70
+2022-12-10 13:27:08,696 INFO [train.py:421] (4/8) Epoch 2, batch 48400, loss[loss=2.377, over 1750.00 frames. , ppl: 10.771322482288824] tot_loss[loss=2.34, over 5545126.31 frames. , ppl: 10.384763930974305], batch size: 70
+2022-12-10 13:28:49,067 INFO [train.py:421] (4/8) Epoch 2, batch 48600, loss[loss=2.614, over 980.00 frames. , ppl: 13.653606606010873] tot_loss[loss=2.34, over 5547838.29 frames. , ppl: 10.384076577276943], batch size: 70
+2022-12-10 13:30:27,744 INFO [train.py:421] (4/8) Epoch 2, batch 48800, loss[loss=2.265, over 5110.00 frames. , ppl: 9.628250770657843] tot_loss[loss=2.341, over 5534610.38 frames. , ppl: 10.38970482225548], batch size: 70
+2022-12-10 13:32:08,063 INFO [train.py:421] (4/8) Epoch 2, batch 49000, loss[loss=2.274, over 5180.00 frames. , ppl: 9.720508963573812] tot_loss[loss=2.341, over 5510534.78 frames. , ppl: 10.395774841281922], batch size: 70
+2022-12-10 13:32:08,064 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:32:08,810 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.328, over 211138.00 frames. , ppl: 10.253641134938565
+2022-12-10 13:33:44,602 INFO [train.py:421] (4/8) Epoch 2, batch 49200, loss[loss=2.438, over 2310.00 frames. , ppl: 11.455415774808166] tot_loss[loss=2.343, over 5466500.90 frames. , ppl: 10.409487301091342], batch size: 70
+2022-12-10 13:35:23,413 INFO [train.py:421] (4/8) Epoch 2, batch 49400, loss[loss=3.093, over 560.00 frames. , ppl: 22.05286762355286] tot_loss[loss=2.342, over 5478759.53 frames. , ppl: 10.405957959983795], batch size: 70
+2022-12-10 13:37:03,027 INFO [train.py:421] (4/8) Epoch 2, batch 49600, loss[loss=2.23, over 6720.00 frames. , ppl: 9.300365597424005] tot_loss[loss=2.343, over 5442747.86 frames. , ppl: 10.415205900356177], batch size: 70
+2022-12-10 13:38:43,472 INFO [train.py:421] (4/8) Epoch 2, batch 49800, loss[loss=2.336, over 4340.00 frames. , ppl: 10.337080325120402] tot_loss[loss=2.343, over 5475317.35 frames. , ppl: 10.41650142851898], batch size: 70
+2022-12-10 13:40:22,316 INFO [train.py:421] (4/8) Epoch 2, batch 50000, loss[loss=2.39, over 1260.00 frames. , ppl: 10.915561396038672] tot_loss[loss=2.343, over 5472358.31 frames. , ppl: 10.412281911387645], batch size: 70
+2022-12-10 13:40:22,316 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:40:23,076 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.327, over 211138.00 frames. , ppl: 10.24865242939489
+2022-12-10 13:42:06,023 INFO [train.py:421] (4/8) Epoch 2, batch 50200, loss[loss=2.425, over 1050.00 frames. , ppl: 11.302605220450662] tot_loss[loss=2.342, over 5503179.83 frames. , ppl: 10.401964294520365], batch size: 70
+2022-12-10 13:43:50,060 INFO [train.py:421] (4/8) Epoch 2, batch 50400, loss[loss=2.257, over 8470.00 frames. , ppl: 9.556563315944016] tot_loss[loss=2.341, over 5521794.31 frames. , ppl: 10.395679533951624], batch size: 70
+2022-12-10 13:45:26,741 INFO [train.py:421] (4/8) Epoch 2, batch 50600, loss[loss=2.337, over 5040.00 frames. , ppl: 10.348724720933447] tot_loss[loss=2.343, over 5481013.03 frames. , ppl: 10.411872432978928], batch size: 70
+2022-12-10 13:47:06,675 INFO [train.py:421] (4/8) Epoch 2, batch 50800, loss[loss=2.303, over 5320.00 frames. , ppl: 10.005556643024715] tot_loss[loss=2.343, over 5456634.81 frames. , ppl: 10.416282301802976], batch size: 70
+2022-12-10 13:48:47,679 INFO [train.py:421] (4/8) Epoch 2, batch 51000, loss[loss=2.295, over 2380.00 frames. , ppl: 9.925595027604153] tot_loss[loss=2.343, over 5442215.84 frames. , ppl: 10.41456228190114], batch size: 70
+2022-12-10 13:48:47,680 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:48:48,425 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.265193656171405
+2022-12-10 13:50:27,915 INFO [train.py:421] (4/8) Epoch 2, batch 51200, loss[loss=2.393, over 2170.00 frames. , ppl: 10.94331628147992] tot_loss[loss=2.341, over 5471932.53 frames. , ppl: 10.396617424067038], batch size: 70
+2022-12-10 13:52:05,819 INFO [train.py:421] (4/8) Epoch 2, batch 51400, loss[loss=2.281, over 4970.00 frames. , ppl: 9.782208018243045] tot_loss[loss=2.342, over 5431888.93 frames. , ppl: 10.402386339578596], batch size: 70
+2022-12-10 13:53:44,960 INFO [train.py:421] (4/8) Epoch 2, batch 51600, loss[loss=2.34, over 2450.00 frames. , ppl: 10.383976233249859] tot_loss[loss=2.341, over 5475784.96 frames. , ppl: 10.393336512974214], batch size: 70
+2022-12-10 13:55:22,666 INFO [train.py:421] (4/8) Epoch 2, batch 51800, loss[loss=2.287, over 3500.00 frames. , ppl: 9.848122535399105] tot_loss[loss=2.341, over 5456264.14 frames. , ppl: 10.392830226658605], batch size: 70
+2022-12-10 13:57:00,774 INFO [train.py:421] (4/8) Epoch 2, batch 52000, loss[loss=2.296, over 2870.00 frames. , ppl: 9.93794926907821] tot_loss[loss=2.341, over 5486570.36 frames. , ppl: 10.38744010278776], batch size: 70
+2022-12-10 13:57:00,774 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 13:57:01,533 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.329, over 211138.00 frames. , ppl: 10.267428826974491
+2022-12-10 13:58:39,504 INFO [train.py:421] (4/8) Epoch 2, batch 52200, loss[loss=2.232, over 3430.00 frames. , ppl: 9.31565867976432] tot_loss[loss=2.342, over 5498922.54 frames. , ppl: 10.397147923938133], batch size: 70
+2022-12-10 14:00:16,956 INFO [train.py:421] (4/8) Epoch 2, batch 52400, loss[loss=2.463, over 980.00 frames. , ppl: 11.739925937499308] tot_loss[loss=2.341, over 5514906.40 frames. , ppl: 10.391700660524636], batch size: 70
+2022-12-10 14:01:58,880 INFO [train.py:421] (4/8) Epoch 2, batch 52600, loss[loss=2.253, over 6650.00 frames. , ppl: 9.512100847944529] tot_loss[loss=2.341, over 5507088.30 frames. , ppl: 10.390364487550587], batch size: 70
+2022-12-10 14:03:38,027 INFO [train.py:421] (4/8) Epoch 2, batch 52800, loss[loss=2.36, over 3150.00 frames. , ppl: 10.589602058912979] tot_loss[loss=2.342, over 5464241.17 frames. , ppl: 10.403368168506992], batch size: 70
+2022-12-10 14:05:17,808 INFO [train.py:421] (4/8) Epoch 2, batch 53000, loss[loss=2.541, over 840.00 frames. , ppl: 12.692717620858314] tot_loss[loss=2.342, over 5455645.71 frames. , ppl: 10.40549362855172], batch size: 70
+2022-12-10 14:05:17,809 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:05:18,555 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.326, over 211138.00 frames. , ppl: 10.233852338507358
+2022-12-10 14:06:58,238 INFO [train.py:421] (4/8) Epoch 2, batch 53200, loss[loss=2.268, over 2870.00 frames. , ppl: 9.662100031508631] tot_loss[loss=2.341, over 5477611.61 frames. , ppl: 10.39606457260365], batch size: 70
+2022-12-10 14:08:37,562 INFO [train.py:421] (4/8) Epoch 2, batch 53400, loss[loss=2.451, over 1050.00 frames. , ppl: 11.601826868832829] tot_loss[loss=2.341, over 5497947.79 frames. , ppl: 10.396113862430724], batch size: 70
+2022-12-10 14:10:19,462 INFO [train.py:421] (4/8) Epoch 2, batch 53600, loss[loss=3.059, over 560.00 frames. , ppl: 21.315867105304488] tot_loss[loss=2.343, over 5443034.36 frames. , ppl: 10.414684080014517], batch size: 70
+2022-12-10 14:11:57,911 INFO [train.py:421] (4/8) Epoch 2, batch 53800, loss[loss=2.277, over 3290.00 frames. , ppl: 9.752244919386602] tot_loss[loss=2.343, over 5437632.02 frames. , ppl: 10.41215588576134], batch size: 70
+2022-12-10 14:13:35,070 INFO [train.py:421] (4/8) Epoch 2, batch 54000, loss[loss=2.293, over 2730.00 frames. , ppl: 9.90631358715562] tot_loss[loss=2.342, over 5469728.45 frames. , ppl: 10.404252668417376], batch size: 70
+2022-12-10 14:13:35,070 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:13:35,833 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.325, over 211138.00 frames. , ppl: 10.222168662360373
+2022-12-10 14:15:13,891 INFO [train.py:421] (4/8) Epoch 2, batch 54200, loss[loss=2.296, over 2450.00 frames. , ppl: 9.933239702642046] tot_loss[loss=2.342, over 5461446.55 frames. , ppl: 10.40329289988713], batch size: 70
+2022-12-10 14:16:54,955 INFO [train.py:421] (4/8) Epoch 2, batch 54400, loss[loss=2.37, over 2800.00 frames. , ppl: 10.696052955352867] tot_loss[loss=2.343, over 5426892.34 frames. , ppl: 10.416895798712188], batch size: 70
+2022-12-10 14:18:34,362 INFO [train.py:421] (4/8) Epoch 2, batch 54600, loss[loss=2.405, over 2380.00 frames. , ppl: 11.078500286275423] tot_loss[loss=2.343, over 5441045.32 frames. , ppl: 10.412190554212804], batch size: 70
+2022-12-10 14:20:12,829 INFO [train.py:421] (4/8) Epoch 2, batch 54800, loss[loss=2.448, over 1610.00 frames. , ppl: 11.565151307620233] tot_loss[loss=2.342, over 5432725.01 frames. , ppl: 10.404168298116351], batch size: 70
+2022-12-10 14:21:50,925 INFO [train.py:421] (4/8) Epoch 2, batch 55000, loss[loss=2.479, over 910.00 frames. , ppl: 11.927567581707677] tot_loss[loss=2.342, over 5425671.27 frames. , ppl: 10.400062934882428], batch size: 70
+2022-12-10 14:21:50,926 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:21:51,690 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.326, over 211138.00 frames. , ppl: 10.232852694220295
+2022-12-10 14:23:32,569 INFO [train.py:421] (4/8) Epoch 2, batch 55200, loss[loss=3.548, over 490.00 frames. , ppl: 34.75883426496409] tot_loss[loss=2.341, over 5448216.42 frames. , ppl: 10.390459083033539], batch size: 70
+2022-12-10 14:25:08,970 INFO [train.py:421] (4/8) Epoch 2, batch 55400, loss[loss=2.935, over 630.00 frames. , ppl: 18.826275343562322] tot_loss[loss=2.341, over 5442141.65 frames. , ppl: 10.389779166054481], batch size: 70
+2022-12-10 14:26:48,398 INFO [train.py:421] (4/8) Epoch 2, batch 55600, loss[loss=2.47, over 1120.00 frames. , ppl: 11.821616027380848] tot_loss[loss=2.341, over 5477191.73 frames. , ppl: 10.388100144551938], batch size: 70
+2022-12-10 14:28:34,501 INFO [train.py:421] (4/8) Epoch 2, batch 55800, loss[loss=2.638, over 770.00 frames. , ppl: 13.986251902956297] tot_loss[loss=2.341, over 5467539.39 frames. , ppl: 10.390722915253404], batch size: 70
+2022-12-10 14:30:15,194 INFO [train.py:421] (4/8) Epoch 2, batch 56000, loss[loss=2.412, over 3570.00 frames. , ppl: 11.158472762938008] tot_loss[loss=2.34, over 5506964.04 frames. , ppl: 10.382390204529042], batch size: 70
+2022-12-10 14:30:15,194 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:30:15,958 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.327, over 211138.00 frames. , ppl: 10.2458754038067
+2022-12-10 14:31:57,753 INFO [train.py:421] (4/8) Epoch 2, batch 56200, loss[loss=2.416, over 2030.00 frames. , ppl: 11.201722712054643] tot_loss[loss=2.339, over 5511862.56 frames. , ppl: 10.376003735979488], batch size: 70
+2022-12-10 14:33:41,514 INFO [train.py:421] (4/8) Epoch 2, batch 56400, loss[loss=2.385, over 1960.00 frames. , ppl: 10.863124535516672] tot_loss[loss=2.339, over 5516182.09 frames. , ppl: 10.37000559376256], batch size: 70
+2022-12-10 14:35:23,939 INFO [train.py:421] (4/8) Epoch 2, batch 56600, loss[loss=2.274, over 3990.00 frames. , ppl: 9.715660178619599] tot_loss[loss=2.339, over 5537745.90 frames. , ppl: 10.365766086176475], batch size: 70
+2022-12-10 14:37:04,930 INFO [train.py:421] (4/8) Epoch 2, batch 56800, loss[loss=2.413, over 2380.00 frames. , ppl: 11.165779465613753] tot_loss[loss=2.338, over 5557146.14 frames. , ppl: 10.361230610383164], batch size: 70
+2022-12-10 14:38:48,269 INFO [train.py:421] (4/8) Epoch 2, batch 57000, loss[loss=2.427, over 1890.00 frames. , ppl: 11.32393947981157] tot_loss[loss=2.338, over 5530372.12 frames. , ppl: 10.365060888815822], batch size: 70
+2022-12-10 14:38:48,270 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:38:49,030 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.216746149309023
+2022-12-10 14:40:28,409 INFO [train.py:421] (4/8) Epoch 2, batch 57200, loss[loss=2.178, over 8750.00 frames. , ppl: 8.830665299575669] tot_loss[loss=2.339, over 5530556.77 frames. , ppl: 10.367377665634619], batch size: 70
+2022-12-10 14:42:06,873 INFO [train.py:421] (4/8) Epoch 2, batch 57400, loss[loss=2.247, over 3080.00 frames. , ppl: 9.463202355037065] tot_loss[loss=2.34, over 5496174.61 frames. , ppl: 10.377672275803798], batch size: 70
+2022-12-10 14:43:46,025 INFO [train.py:421] (4/8) Epoch 2, batch 57600, loss[loss=2.281, over 4270.00 frames. , ppl: 9.781995831075605] tot_loss[loss=2.341, over 5451002.01 frames. , ppl: 10.39026337573984], batch size: 70
+2022-12-10 14:45:24,407 INFO [train.py:421] (4/8) Epoch 2, batch 57800, loss[loss=2.204, over 10500.00 frames. , ppl: 9.064898055650962] tot_loss[loss=2.341, over 5433364.45 frames. , ppl: 10.396273147807163], batch size: 70
+2022-12-10 14:47:02,177 INFO [train.py:421] (4/8) Epoch 2, batch 58000, loss[loss=2.232, over 7210.00 frames. , ppl: 9.318655723994523] tot_loss[loss=2.341, over 5447848.70 frames. , ppl: 10.388629351396704], batch size: 70
+2022-12-10 14:47:02,178 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:47:02,926 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.325, over 211138.00 frames. , ppl: 10.22471528659933
+2022-12-10 14:48:40,205 INFO [train.py:421] (4/8) Epoch 2, batch 58200, loss[loss=2.378, over 1750.00 frames. , ppl: 10.783200322196691] tot_loss[loss=2.342, over 5402224.88 frames. , ppl: 10.399759153750193], batch size: 70
+2022-12-10 14:50:20,524 INFO [train.py:421] (4/8) Epoch 2, batch 58400, loss[loss=2.422, over 3710.00 frames. , ppl: 11.269415049182356] tot_loss[loss=2.341, over 5441130.10 frames. , ppl: 10.39573080402768], batch size: 70
+2022-12-10 14:51:59,910 INFO [train.py:421] (4/8) Epoch 2, batch 58600, loss[loss=2.438, over 2240.00 frames. , ppl: 11.455415774808166] tot_loss[loss=2.341, over 5437115.79 frames. , ppl: 10.392527406055242], batch size: 70
+2022-12-10 14:53:41,810 INFO [train.py:421] (4/8) Epoch 2, batch 58800, loss[loss=2.274, over 6090.00 frames. , ppl: 9.717694425540403] tot_loss[loss=2.341, over 5420701.76 frames. , ppl: 10.394957028844317], batch size: 70
+2022-12-10 14:55:21,055 INFO [train.py:421] (4/8) Epoch 2, batch 59000, loss[loss=2.437, over 1190.00 frames. , ppl: 11.439452630263352] tot_loss[loss=2.342, over 5407107.22 frames. , ppl: 10.4032884060291], batch size: 70
+2022-12-10 14:55:21,055 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 14:55:21,801 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.325, over 211138.00 frames. , ppl: 10.225733811230693
+2022-12-10 14:57:02,577 INFO [train.py:421] (4/8) Epoch 2, batch 59200, loss[loss=2.381, over 2310.00 frames. , ppl: 10.81165072591562] tot_loss[loss=2.342, over 5442869.87 frames. , ppl: 10.403059450400802], batch size: 70
+2022-12-10 14:58:41,089 INFO [train.py:421] (4/8) Epoch 2, batch 59400, loss[loss=2.253, over 5180.00 frames. , ppl: 9.513251633367739] tot_loss[loss=2.343, over 5408890.95 frames. , ppl: 10.409498332562798], batch size: 70
+2022-12-10 15:00:19,436 INFO [train.py:421] (4/8) Epoch 2, batch 59600, loss[loss=2.495, over 1400.00 frames. , ppl: 12.116980373810105] tot_loss[loss=2.343, over 5420857.87 frames. , ppl: 10.408919367669114], batch size: 70
+2022-12-10 15:02:02,512 INFO [train.py:421] (4/8) Epoch 2, batch 59800, loss[loss=2.351, over 3500.00 frames. , ppl: 10.491914477010301] tot_loss[loss=2.342, over 5408336.54 frames. , ppl: 10.406888382443515], batch size: 70
+2022-12-10 15:03:44,826 INFO [train.py:421] (4/8) Epoch 2, batch 60000, loss[loss=2.443, over 1330.00 frames. , ppl: 11.511347740189892] tot_loss[loss=2.342, over 5391576.41 frames. , ppl: 10.404596299314615], batch size: 70
+2022-12-10 15:03:44,827 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:03:45,573 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.21343657885031
+2022-12-10 15:05:24,002 INFO [train.py:421] (4/8) Epoch 2, batch 60200, loss[loss=2.297, over 2940.00 frames. , ppl: 9.947861808505623] tot_loss[loss=2.342, over 5412073.75 frames. , ppl: 10.400776454718104], batch size: 70
+2022-12-10 15:07:02,397 INFO [train.py:421] (4/8) Epoch 2, batch 60400, loss[loss=2.404, over 910.00 frames. , ppl: 11.063096070575781] tot_loss[loss=2.342, over 5383171.58 frames. , ppl: 10.401142151383143], batch size: 70
+2022-12-10 15:08:40,177 INFO [train.py:421] (4/8) Epoch 2, batch 60600, loss[loss=2.279, over 4830.00 frames. , ppl: 9.771756463702522] tot_loss[loss=2.342, over 5406212.70 frames. , ppl: 10.400381728751658], batch size: 70
+2022-12-10 15:10:20,018 INFO [train.py:421] (4/8) Epoch 2, batch 60800, loss[loss=2.568, over 980.00 frames. , ppl: 13.042777918017558] tot_loss[loss=2.341, over 5408598.03 frames. , ppl: 10.390667819286733], batch size: 70
+2022-12-10 15:11:55,175 INFO [train.py:421] (4/8) Epoch 2, batch 61000, loss[loss=2.267, over 4270.00 frames. , ppl: 9.646316922629008] tot_loss[loss=2.342, over 5375395.97 frames. , ppl: 10.398234135654135], batch size: 70
+2022-12-10 15:11:55,176 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:11:55,926 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.211778417126597
+2022-12-10 15:13:35,397 INFO [train.py:421] (4/8) Epoch 2, batch 61200, loss[loss=2.522, over 1400.00 frames. , ppl: 12.457002184035614] tot_loss[loss=2.341, over 5377352.35 frames. , ppl: 10.393524224571898], batch size: 70
+2022-12-10 15:15:14,585 INFO [train.py:421] (4/8) Epoch 2, batch 61400, loss[loss=2.292, over 5320.00 frames. , ppl: 9.895494614829188] tot_loss[loss=2.341, over 5370639.75 frames. , ppl: 10.396374871987673], batch size: 70
+2022-12-10 15:16:52,145 INFO [train.py:421] (4/8) Epoch 2, batch 61600, loss[loss=2.325, over 3290.00 frames. , ppl: 10.223550902306453] tot_loss[loss=2.342, over 5335062.40 frames. , ppl: 10.406958577249721], batch size: 70
+2022-12-10 15:18:32,314 INFO [train.py:421] (4/8) Epoch 2, batch 61800, loss[loss=2.258, over 3150.00 frames. , ppl: 9.563036966565699] tot_loss[loss=2.341, over 5376070.69 frames. , ppl: 10.394552548995438], batch size: 70
+2022-12-10 15:20:15,006 INFO [train.py:421] (4/8) Epoch 2, batch 62000, loss[loss=2.279, over 7840.00 frames. , ppl: 9.770967195164264] tot_loss[loss=2.34, over 5419505.35 frames. , ppl: 10.383512607250575], batch size: 70
+2022-12-10 15:20:15,007 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:20:15,753 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.219063032959546
+2022-12-10 15:21:54,084 INFO [train.py:421] (4/8) Epoch 2, batch 62200, loss[loss=2.663, over 980.00 frames. , ppl: 14.342550493139866] tot_loss[loss=2.34, over 5413774.42 frames. , ppl: 10.383004716501278], batch size: 70
+2022-12-10 15:23:35,865 INFO [train.py:421] (4/8) Epoch 2, batch 62400, loss[loss=2.391, over 2100.00 frames. , ppl: 10.920677572036361] tot_loss[loss=2.339, over 5464806.71 frames. , ppl: 10.369852305575552], batch size: 70
+2022-12-10 15:25:13,327 INFO [train.py:421] (4/8) Epoch 2, batch 62600, loss[loss=2.276, over 3710.00 frames. , ppl: 9.735147782455698] tot_loss[loss=2.339, over 5450644.64 frames. , ppl: 10.374647074119798], batch size: 70
+2022-12-10 15:26:54,710 INFO [train.py:421] (4/8) Epoch 2, batch 62800, loss[loss=2.384, over 2800.00 frames. , ppl: 10.847524984792585] tot_loss[loss=2.339, over 5442341.96 frames. , ppl: 10.373332349515618], batch size: 70
+2022-12-10 15:28:32,254 INFO [train.py:421] (4/8) Epoch 2, batch 63000, loss[loss=2.278, over 4760.00 frames. , ppl: 9.761706263207715] tot_loss[loss=2.34, over 5412885.73 frames. , ppl: 10.382360036720401], batch size: 70
+2022-12-10 15:28:32,254 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:28:33,000 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.20283624301792
+2022-12-10 15:30:14,780 INFO [train.py:421] (4/8) Epoch 2, batch 63200, loss[loss=2.225, over 5040.00 frames. , ppl: 9.254587343682186] tot_loss[loss=2.338, over 5450178.49 frames. , ppl: 10.361909822252779], batch size: 70
+2022-12-10 15:31:59,079 INFO [train.py:421] (4/8) Epoch 2, batch 63400, loss[loss=2.204, over 12460.00 frames. , ppl: 9.060830654379924] tot_loss[loss=2.336, over 5516514.61 frames. , ppl: 10.343341815346873], batch size: 70
+2022-12-10 15:33:37,719 INFO [train.py:421] (4/8) Epoch 2, batch 63600, loss[loss=2.316, over 4550.00 frames. , ppl: 10.137151830847907] tot_loss[loss=2.338, over 5484464.11 frames. , ppl: 10.355394735548265], batch size: 70
+2022-12-10 15:35:17,819 INFO [train.py:421] (4/8) Epoch 2, batch 63800, loss[loss=2.699, over 700.00 frames. , ppl: 14.86198337509066] tot_loss[loss=2.338, over 5466133.15 frames. , ppl: 10.362216266745905], batch size: 70
+2022-12-10 15:36:57,203 INFO [train.py:421] (4/8) Epoch 2, batch 64000, loss[loss=2.453, over 1890.00 frames. , ppl: 11.617729603875574] tot_loss[loss=2.338, over 5474611.06 frames. , ppl: 10.355974012475333], batch size: 70
+2022-12-10 15:36:57,203 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:36:57,966 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.324, over 211138.00 frames. , ppl: 10.220371427142487
+2022-12-10 15:38:37,720 INFO [train.py:421] (4/8) Epoch 2, batch 64200, loss[loss=2.256, over 3150.00 frames. , ppl: 9.549565830881027] tot_loss[loss=2.338, over 5481517.98 frames. , ppl: 10.359803593013982], batch size: 70
+2022-12-10 15:40:17,493 INFO [train.py:421] (4/8) Epoch 2, batch 64400, loss[loss=2.329, over 2170.00 frames. , ppl: 10.27129517651795] tot_loss[loss=2.338, over 5503166.22 frames. , ppl: 10.358482377756694], batch size: 70
+2022-12-10 15:41:55,772 INFO [train.py:421] (4/8) Epoch 2, batch 64600, loss[loss=2.5, over 1120.00 frames. , ppl: 12.18741041800426] tot_loss[loss=2.336, over 5555216.64 frames. , ppl: 10.339349282489163], batch size: 70
+2022-12-10 15:43:33,789 INFO [train.py:421] (4/8) Epoch 2, batch 64800, loss[loss=2.549, over 1470.00 frames. , ppl: 12.788259410762906] tot_loss[loss=2.336, over 5552781.48 frames. , ppl: 10.34424957370479], batch size: 70
+2022-12-10 15:45:14,653 INFO [train.py:421] (4/8) Epoch 2, batch 65000, loss[loss=2.568, over 840.00 frames. , ppl: 13.038614118278984] tot_loss[loss=2.337, over 5527694.40 frames. , ppl: 10.348223079175344], batch size: 70
+2022-12-10 15:45:14,654 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:45:15,399 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.211031803223284
+2022-12-10 15:46:58,768 INFO [train.py:421] (4/8) Epoch 2, batch 65200, loss[loss=2.346, over 3430.00 frames. , ppl: 10.438937780217369] tot_loss[loss=2.337, over 5514399.93 frames. , ppl: 10.348537975369553], batch size: 70
+2022-12-10 15:48:38,204 INFO [train.py:421] (4/8) Epoch 2, batch 65400, loss[loss=2.443, over 2170.00 frames. , ppl: 11.504169483139815] tot_loss[loss=2.336, over 5533669.73 frames. , ppl: 10.335631702241637], batch size: 70
+2022-12-10 15:50:19,932 INFO [train.py:421] (4/8) Epoch 2, batch 65600, loss[loss=2.464, over 1260.00 frames. , ppl: 11.75300559237812] tot_loss[loss=2.335, over 5538673.26 frames. , ppl: 10.32518264527471], batch size: 70
+2022-12-10 15:51:59,509 INFO [train.py:421] (4/8) Epoch 2, batch 65800, loss[loss=2.278, over 3290.00 frames. , ppl: 9.756894971341978] tot_loss[loss=2.335, over 5502496.53 frames. , ppl: 10.334349948644403], batch size: 70
+2022-12-10 15:53:41,830 INFO [train.py:421] (4/8) Epoch 2, batch 66000, loss[loss=2.419, over 1820.00 frames. , ppl: 11.230683839503216] tot_loss[loss=2.335, over 5504074.02 frames. , ppl: 10.3276099643207], batch size: 70
+2022-12-10 15:53:41,830 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 15:53:42,595 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.321, over 211138.00 frames. , ppl: 10.187981682837393
+2022-12-10 15:55:24,764 INFO [train.py:421] (4/8) Epoch 2, batch 66200, loss[loss=2.518, over 1190.00 frames. , ppl: 12.405499143252747] tot_loss[loss=2.334, over 5524255.01 frames. , ppl: 10.322194499770628], batch size: 70
+2022-12-10 15:57:04,889 INFO [train.py:421] (4/8) Epoch 2, batch 66400, loss[loss=2.226, over 7840.00 frames. , ppl: 9.262856109236534] tot_loss[loss=2.335, over 5527088.83 frames. , ppl: 10.324731246559766], batch size: 70
+2022-12-10 15:58:43,335 INFO [train.py:421] (4/8) Epoch 2, batch 66600, loss[loss=2.371, over 2310.00 frames. , ppl: 10.710396112986508] tot_loss[loss=2.333, over 5583803.79 frames. , ppl: 10.31278725179407], batch size: 70
+2022-12-10 16:00:20,766 INFO [train.py:421] (4/8) Epoch 2, batch 66800, loss[loss=2.305, over 5110.00 frames. , ppl: 10.021620733594983] tot_loss[loss=2.335, over 5564192.26 frames. , ppl: 10.324326115398193], batch size: 70
+2022-12-10 16:02:02,736 INFO [train.py:421] (4/8) Epoch 2, batch 67000, loss[loss=2.3, over 11200.00 frames. , ppl: 9.973161503111573] tot_loss[loss=2.335, over 5574552.59 frames. , ppl: 10.32525219507615], batch size: 70
+2022-12-10 16:02:02,737 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 16:02:03,483 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.203262098954827
+2022-12-10 16:03:41,303 INFO [train.py:421] (4/8) Epoch 2, batch 67200, loss[loss=2.273, over 6510.00 frames. , ppl: 9.708515356562604] tot_loss[loss=2.335, over 5566516.95 frames. , ppl: 10.326964250412242], batch size: 70
+2022-12-10 16:05:20,727 INFO [train.py:421] (4/8) Epoch 2, batch 67400, loss[loss=2.324, over 3850.00 frames. , ppl: 10.218776556374754] tot_loss[loss=2.335, over 5547371.82 frames. , ppl: 10.332487254848816], batch size: 70
+2022-12-10 16:07:01,226 INFO [train.py:421] (4/8) Epoch 2, batch 67600, loss[loss=4.071, over 350.00 frames. , ppl: 58.60405664441665] tot_loss[loss=2.336, over 5515242.45 frames. , ppl: 10.341013763780147], batch size: 70
+2022-12-10 16:08:44,345 INFO [train.py:421] (4/8) Epoch 2, batch 67800, loss[loss=2.252, over 7840.00 frames. , ppl: 9.503769945744232] tot_loss[loss=2.334, over 5593345.25 frames. , ppl: 10.320905938843842], batch size: 70
+2022-12-10 16:10:23,887 INFO [train.py:421] (4/8) Epoch 2, batch 68000, loss[loss=2.235, over 6370.00 frames. , ppl: 9.343424880227964] tot_loss[loss=2.336, over 5534361.03 frames. , ppl: 10.33659611964191], batch size: 70
+2022-12-10 16:10:23,888 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 16:10:24,633 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.323, over 211138.00 frames. , ppl: 10.201776210751579
+2022-12-10 16:12:06,495 INFO [train.py:421] (4/8) Epoch 2, batch 68200, loss[loss=2.3, over 4620.00 frames. , ppl: 9.969439875273824] tot_loss[loss=2.336, over 5512463.16 frames. , ppl: 10.340250502912928], batch size: 70
+2022-12-10 16:13:47,728 INFO [train.py:421] (4/8) Epoch 2, batch 68400, loss[loss=2.266, over 5040.00 frames. , ppl: 9.644334048864389] tot_loss[loss=2.336, over 5504393.50 frames. , ppl: 10.344516687800967], batch size: 70
+2022-12-10 16:15:28,277 INFO [train.py:421] (4/8) Epoch 2, batch 68600, loss[loss=2.611, over 910.00 frames. , ppl: 13.6103890911406] tot_loss[loss=2.336, over 5515391.28 frames. , ppl: 10.344075634495415], batch size: 70
+2022-12-10 16:17:07,753 INFO [train.py:421] (4/8) Epoch 2, batch 68800, loss[loss=2.309, over 3220.00 frames. , ppl: 10.067517905673629] tot_loss[loss=2.338, over 5463732.91 frames. , ppl: 10.357519755631607], batch size: 70
+2022-12-10 16:18:50,959 INFO [train.py:421] (4/8) Epoch 2, batch 69000, loss[loss=2.806, over 700.00 frames. , ppl: 16.54807086945777] tot_loss[loss=2.337, over 5466115.25 frames. , ppl: 10.350338271677614], batch size: 70
+2022-12-10 16:18:50,959 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 16:18:51,729 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.32, over 211138.00 frames. , ppl: 10.176567267443412
+2022-12-10 16:20:33,667 INFO [train.py:421] (4/8) Epoch 2, batch 69200, loss[loss=2.237, over 5040.00 frames. , ppl: 9.368624657031901] tot_loss[loss=2.337, over 5477302.22 frames. , ppl: 10.348713910147033], batch size: 70
+2022-12-10 16:22:12,410 INFO [train.py:421] (4/8) Epoch 2, batch 69400, loss[loss=2.402, over 2170.00 frames. , ppl: 11.042202363142502] tot_loss[loss=2.337, over 5492143.18 frames. , ppl: 10.3470623210561], batch size: 70
+2022-12-10 16:23:54,609 INFO [train.py:421] (4/8) Epoch 2, batch 69600, loss[loss=2.359, over 2520.00 frames. , ppl: 10.579802776504362] tot_loss[loss=2.337, over 5441230.31 frames. , ppl: 10.353572789948448], batch size: 70
+2022-12-10 16:25:37,452 INFO [train.py:421] (4/8) Epoch 2, batch 69800, loss[loss=2.409, over 1820.00 frames. , ppl: 11.126953844805886] tot_loss[loss=2.338, over 5457344.62 frames. , ppl: 10.355937584349697], batch size: 70
+2022-12-10 16:27:18,878 INFO [train.py:421] (4/8) Epoch 2, batch 70000, loss[loss=2.37, over 2170.00 frames. , ppl: 10.70052867110216] tot_loss[loss=2.337, over 5480833.82 frames. , ppl: 10.352121638651417], batch size: 70
+2022-12-10 16:27:18,879 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 16:27:19,638 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.322, over 211138.00 frames. , ppl: 10.191266412638488
+2022-12-10 16:29:02,809 INFO [train.py:421] (4/8) Epoch 2, batch 70200, loss[loss=2.299, over 2240.00 frames. , ppl: 9.966814642022394] tot_loss[loss=2.337, over 5472049.67 frames. , ppl: 10.351910914564929], batch size: 70
+2022-12-10 16:30:41,310 INFO [train.py:421] (4/8) Epoch 2, batch 70400, loss[loss=2.292, over 2800.00 frames. , ppl: 9.892093877351325] tot_loss[loss=2.338, over 5465016.45 frames. , ppl: 10.356289137727492], batch size: 70
+2022-12-10 16:32:24,246 INFO [train.py:421] (4/8) Epoch 2, batch 70600, loss[loss=2.416, over 1610.00 frames. , ppl: 11.20330149562109] tot_loss[loss=2.337, over 5465606.62 frames. , ppl: 10.35427642410709], batch size: 70
+2022-12-10 16:34:01,869 INFO [train.py:421] (4/8) Epoch 2, batch 70800, loss[loss=2.377, over 2380.00 frames. , ppl: 10.773671716932917] tot_loss[loss=2.338, over 5473170.11 frames. , ppl: 10.356881748711167], batch size: 70
+2022-12-10 16:35:42,867 INFO [train.py:421] (4/8) Epoch 2, batch 71000, loss[loss=2.269, over 6370.00 frames. , ppl: 9.673930274846093] tot_loss[loss=2.338, over 5487885.89 frames. , ppl: 10.355395071430648], batch size: 70
+2022-12-10 16:35:42,868 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 16:35:43,613 INFO [train.py:452] (4/8) Epoch 2, validation: loss=2.319, over 211138.00 frames. , ppl: 10.169496572570102
+2022-12-10 16:37:26,629 INFO [train.py:421] (4/8) Epoch 2, batch 71200, loss[loss=2.278, over 4830.00 frames. , ppl: 9.76056251958357] tot_loss[loss=2.339, over 5451418.69 frames. , ppl: 10.372024771281168], batch size: 70
+2022-12-10 16:39:05,521 INFO [train.py:421] (4/8) Epoch 2, batch 71400, loss[loss=2.2, over 3220.00 frames. , ppl: 9.022720093858956] tot_loss[loss=2.338, over 5481314.80 frames. , ppl: 10.364160559440679], batch size: 70
+2022-12-10 16:40:47,631 INFO [train.py:421] (4/8) Epoch 2, batch 71600, loss[loss=2.449, over 2590.00 frames. , ppl: 11.577240661482573] tot_loss[loss=2.337, over 5530736.43 frames. , ppl: 10.354518332474008], batch size: 70
+2022-12-10 16:42:29,323 INFO [train.py:421] (4/8) Epoch 2, batch 71800, loss[loss=2.65, over 840.00 frames. , ppl: 14.151463652875885] tot_loss[loss=2.337, over 5537099.64 frames. , ppl: 10.350524222998601], batch size: 70
+2022-12-10 16:43:45,605 INFO [train.py:421] (4/8) Epoch 3, batch 0, loss[loss=2.266, over 8820.00 frames. , ppl: 9.642973724425172] tot_loss[loss=2.266, over 8820.00 frames. , ppl: 9.642973724425172], batch size: 70
+2022-12-10 16:45:28,193 INFO [train.py:421] (4/8) Epoch 3, batch 200, loss[loss=3.551, over 420.00 frames. , ppl: 34.83322581089898] tot_loss[loss=2.337, over 521959.28 frames. , ppl: 10.350930641893825], batch size: 70
+2022-12-10 16:47:09,531 INFO [train.py:421] (4/8) Epoch 3, batch 400, loss[loss=2.405, over 1400.00 frames. , ppl: 11.073242346953966] tot_loss[loss=2.325, over 1022732.75 frames. , ppl: 10.23171457072998], batch size: 70
+2022-12-10 16:48:48,730 INFO [train.py:421] (4/8) Epoch 3, batch 600, loss[loss=2.401, over 2380.00 frames. , ppl: 11.030614756933153] tot_loss[loss=2.326, over 1429132.82 frames. , ppl: 10.24151193049517], batch size: 70
+2022-12-10 16:50:28,169 INFO [train.py:421] (4/8) Epoch 3, batch 800, loss[loss=2.306, over 2660.00 frames. , ppl: 10.037908402874146] tot_loss[loss=2.327, over 1851913.91 frames. , ppl: 10.245731810937764], batch size: 70
+2022-12-10 16:52:09,324 INFO [train.py:421] (4/8) Epoch 3, batch 1000, loss[loss=2.255, over 6230.00 frames. , ppl: 9.536642198039102] tot_loss[loss=2.325, over 2240675.45 frames. , ppl: 10.2239597494214], batch size: 70
+2022-12-10 16:52:09,324 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 16:52:10,090 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.164404381725806
+2022-12-10 16:53:53,057 INFO [train.py:421] (4/8) Epoch 3, batch 1200, loss[loss=2.335, over 2100.00 frames. , ppl: 10.328955573522393] tot_loss[loss=2.326, over 2522911.16 frames. , ppl: 10.236852909876509], batch size: 70
+2022-12-10 16:55:35,512 INFO [train.py:421] (4/8) Epoch 3, batch 1400, loss[loss=2.349, over 2800.00 frames. , ppl: 10.478479223300349] tot_loss[loss=2.325, over 2819987.14 frames. , ppl: 10.222185223533783], batch size: 70
+2022-12-10 16:57:16,467 INFO [train.py:421] (4/8) Epoch 3, batch 1600, loss[loss=2.296, over 2800.00 frames. , ppl: 9.937136971095748] tot_loss[loss=2.326, over 3063745.89 frames. , ppl: 10.235138695310587], batch size: 70
+2022-12-10 16:58:58,235 INFO [train.py:421] (4/8) Epoch 3, batch 1800, loss[loss=2.469, over 1540.00 frames. , ppl: 11.813486103218963] tot_loss[loss=2.326, over 3339981.94 frames. , ppl: 10.232793398408234], batch size: 70
+2022-12-10 17:00:33,686 INFO [train.py:421] (4/8) Epoch 3, batch 2000, loss[loss=2.236, over 3010.00 frames. , ppl: 9.357210757178375] tot_loss[loss=2.324, over 3572753.06 frames. , ppl: 10.220335152770428], batch size: 70
+2022-12-10 17:00:33,686 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:00:34,445 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.321, over 211138.00 frames. , ppl: 10.184881917574328
+2022-12-10 17:02:13,490 INFO [train.py:421] (4/8) Epoch 3, batch 2200, loss[loss=2.427, over 1050.00 frames. , ppl: 11.32228257616708] tot_loss[loss=2.325, over 3756989.56 frames. , ppl: 10.222076721170744], batch size: 70
+2022-12-10 17:03:54,066 INFO [train.py:421] (4/8) Epoch 3, batch 2400, loss[loss=2.285, over 8470.00 frames. , ppl: 9.828257265671297] tot_loss[loss=2.324, over 3939310.08 frames. , ppl: 10.219585192279965], batch size: 70
+2022-12-10 17:05:32,316 INFO [train.py:421] (4/8) Epoch 3, batch 2600, loss[loss=2.255, over 3780.00 frames. , ppl: 9.534257242831803] tot_loss[loss=2.325, over 4091608.26 frames. , ppl: 10.22399895180621], batch size: 70
+2022-12-10 17:07:12,597 INFO [train.py:421] (4/8) Epoch 3, batch 2800, loss[loss=2.471, over 1050.00 frames. , ppl: 11.831554697774992] tot_loss[loss=2.324, over 4246080.01 frames. , ppl: 10.217464369803348], batch size: 70
+2022-12-10 17:08:55,682 INFO [train.py:421] (4/8) Epoch 3, batch 3000, loss[loss=2.345, over 1960.00 frames. , ppl: 10.43773489533557] tot_loss[loss=2.324, over 4355616.23 frames. , ppl: 10.21986427903782], batch size: 70
+2022-12-10 17:08:55,682 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:08:56,443 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.32, over 211138.00 frames. , ppl: 10.171098190408838
+2022-12-10 17:10:38,822 INFO [train.py:421] (4/8) Epoch 3, batch 3200, loss[loss=2.334, over 1610.00 frames. , ppl: 10.323052995535226] tot_loss[loss=2.327, over 4403172.19 frames. , ppl: 10.248417690992232], batch size: 70
+2022-12-10 17:12:18,589 INFO [train.py:421] (4/8) Epoch 3, batch 3400, loss[loss=2.32, over 2940.00 frames. , ppl: 10.174031085752198] tot_loss[loss=2.326, over 4548013.51 frames. , ppl: 10.239100190286848], batch size: 70
+2022-12-10 17:13:58,759 INFO [train.py:421] (4/8) Epoch 3, batch 3600, loss[loss=2.412, over 1400.00 frames. , ppl: 11.154333645171395] tot_loss[loss=2.326, over 4646102.79 frames. , ppl: 10.241855780491644], batch size: 70
+2022-12-10 17:15:38,952 INFO [train.py:421] (4/8) Epoch 3, batch 3800, loss[loss=2.504, over 2240.00 frames. , ppl: 12.235342628594962] tot_loss[loss=2.327, over 4710435.72 frames. , ppl: 10.24354974909689], batch size: 70
+2022-12-10 17:17:17,268 INFO [train.py:421] (4/8) Epoch 3, batch 4000, loss[loss=2.432, over 1400.00 frames. , ppl: 11.376900136057914] tot_loss[loss=2.328, over 4749753.11 frames. , ppl: 10.259077655926102], batch size: 70
+2022-12-10 17:17:17,269 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:17:18,024 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.163163321129922
+2022-12-10 17:18:57,020 INFO [train.py:421] (4/8) Epoch 3, batch 4200, loss[loss=2.358, over 2310.00 frames. , ppl: 10.574720449736198] tot_loss[loss=2.327, over 4842042.21 frames. , ppl: 10.250014899540599], batch size: 70
+2022-12-10 17:20:36,279 INFO [train.py:421] (4/8) Epoch 3, batch 4400, loss[loss=2.311, over 4620.00 frames. , ppl: 10.084095535254447] tot_loss[loss=2.326, over 4933674.29 frames. , ppl: 10.237373808927025], batch size: 70
+2022-12-10 17:22:15,217 INFO [train.py:421] (4/8) Epoch 3, batch 4600, loss[loss=2.668, over 910.00 frames. , ppl: 14.404912173159518] tot_loss[loss=2.327, over 4970778.98 frames. , ppl: 10.24907689994663], batch size: 70
+2022-12-10 17:23:55,352 INFO [train.py:421] (4/8) Epoch 3, batch 4800, loss[loss=2.331, over 3290.00 frames. , ppl: 10.28439961282209] tot_loss[loss=2.326, over 5065144.73 frames. , ppl: 10.237824785110899], batch size: 70
+2022-12-10 17:25:31,817 INFO [train.py:421] (4/8) Epoch 3, batch 5000, loss[loss=2.783, over 630.00 frames. , ppl: 16.161397354046073] tot_loss[loss=2.327, over 5085430.04 frames. , ppl: 10.247826841061778], batch size: 70
+2022-12-10 17:25:31,818 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:25:32,565 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.167579177724292
+2022-12-10 17:27:12,749 INFO [train.py:421] (4/8) Epoch 3, batch 5200, loss[loss=2.521, over 910.00 frames. , ppl: 12.441249901514466] tot_loss[loss=2.325, over 5165506.96 frames. , ppl: 10.229312174716139], batch size: 70
+2022-12-10 17:28:53,461 INFO [train.py:421] (4/8) Epoch 3, batch 5400, loss[loss=2.492, over 980.00 frames. , ppl: 12.08776131251085] tot_loss[loss=2.325, over 5202361.93 frames. , ppl: 10.231630153469034], batch size: 70
+2022-12-10 17:30:38,110 INFO [train.py:421] (4/8) Epoch 3, batch 5600, loss[loss=2.366, over 1470.00 frames. , ppl: 10.658341304246731] tot_loss[loss=2.326, over 5225458.81 frames. , ppl: 10.234182081639474], batch size: 70
+2022-12-10 17:32:21,541 INFO [train.py:421] (4/8) Epoch 3, batch 5800, loss[loss=2.419, over 1680.00 frames. , ppl: 11.231265622329976] tot_loss[loss=2.326, over 5257572.71 frames. , ppl: 10.234401454527237], batch size: 70
+2022-12-10 17:34:02,533 INFO [train.py:421] (4/8) Epoch 3, batch 6000, loss[loss=2.283, over 4690.00 frames. , ppl: 9.805798283232] tot_loss[loss=2.326, over 5270555.13 frames. , ppl: 10.232979140632207], batch size: 70
+2022-12-10 17:34:02,534 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:34:03,279 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.146697899157441
+2022-12-10 17:35:42,386 INFO [train.py:421] (4/8) Epoch 3, batch 6200, loss[loss=2.305, over 1470.00 frames. , ppl: 10.019644930062254] tot_loss[loss=2.325, over 5266046.71 frames. , ppl: 10.230427663327179], batch size: 70
+2022-12-10 17:37:19,187 INFO [train.py:421] (4/8) Epoch 3, batch 6400, loss[loss=2.291, over 3360.00 frames. , ppl: 9.881670043126125] tot_loss[loss=2.326, over 5298282.93 frames. , ppl: 10.232707742698683], batch size: 70
+2022-12-10 17:38:59,188 INFO [train.py:421] (4/8) Epoch 3, batch 6600, loss[loss=2.562, over 1680.00 frames. , ppl: 12.955704924103651] tot_loss[loss=2.326, over 5281762.60 frames. , ppl: 10.237191839664977], batch size: 70
+2022-12-10 17:40:37,939 INFO [train.py:421] (4/8) Epoch 3, batch 6800, loss[loss=2.472, over 910.00 frames. , ppl: 11.852009419253106] tot_loss[loss=2.325, over 5338292.02 frames. , ppl: 10.231673008210597], batch size: 70
+2022-12-10 17:42:18,035 INFO [train.py:421] (4/8) Epoch 3, batch 7000, loss[loss=2.284, over 3010.00 frames. , ppl: 9.811741680039797] tot_loss[loss=2.324, over 5404600.54 frames. , ppl: 10.220935058778393], batch size: 70
+2022-12-10 17:42:18,035 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:42:18,803 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.159209485439936
+2022-12-10 17:44:00,387 INFO [train.py:421] (4/8) Epoch 3, batch 7200, loss[loss=2.454, over 2170.00 frames. , ppl: 11.631485202466965] tot_loss[loss=2.324, over 5421745.90 frames. , ppl: 10.212975319224736], batch size: 70
+2022-12-10 17:45:38,827 INFO [train.py:421] (4/8) Epoch 3, batch 7400, loss[loss=2.327, over 2660.00 frames. , ppl: 10.243871394936788] tot_loss[loss=2.326, over 5383958.26 frames. , ppl: 10.235280939460925], batch size: 70
+2022-12-10 17:47:16,706 INFO [train.py:421] (4/8) Epoch 3, batch 7600, loss[loss=2.311, over 4830.00 frames. , ppl: 10.087905422878636] tot_loss[loss=2.327, over 5360867.47 frames. , ppl: 10.24358777603914], batch size: 70
+2022-12-10 17:48:56,787 INFO [train.py:421] (4/8) Epoch 3, batch 7800, loss[loss=2.269, over 3430.00 frames. , ppl: 9.674378492145259] tot_loss[loss=2.327, over 5364052.05 frames. , ppl: 10.250526865026384], batch size: 70
+2022-12-10 17:50:40,363 INFO [train.py:421] (4/8) Epoch 3, batch 8000, loss[loss=2.351, over 1820.00 frames. , ppl: 10.494981071787928] tot_loss[loss=2.326, over 5406515.44 frames. , ppl: 10.232679491872348], batch size: 70
+2022-12-10 17:50:40,363 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:50:41,126 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.153032911836572
+2022-12-10 17:52:20,457 INFO [train.py:421] (4/8) Epoch 3, batch 8200, loss[loss=2.283, over 2730.00 frames. , ppl: 9.80573823397519] tot_loss[loss=2.327, over 5367915.03 frames. , ppl: 10.248916973109116], batch size: 70
+2022-12-10 17:53:58,390 INFO [train.py:421] (4/8) Epoch 3, batch 8400, loss[loss=2.372, over 1400.00 frames. , ppl: 10.722028117845367] tot_loss[loss=2.326, over 5396168.38 frames. , ppl: 10.236683971147201], batch size: 70
+2022-12-10 17:55:40,041 INFO [train.py:421] (4/8) Epoch 3, batch 8600, loss[loss=2.276, over 3360.00 frames. , ppl: 9.739511010719049] tot_loss[loss=2.327, over 5355928.12 frames. , ppl: 10.248965621769484], batch size: 70
+2022-12-10 17:57:23,065 INFO [train.py:421] (4/8) Epoch 3, batch 8800, loss[loss=2.404, over 2170.00 frames. , ppl: 11.068162686761443] tot_loss[loss=2.326, over 5406169.32 frames. , ppl: 10.237895943279725], batch size: 70
+2022-12-10 17:59:05,018 INFO [train.py:421] (4/8) Epoch 3, batch 9000, loss[loss=2.721, over 630.00 frames. , ppl: 15.196372398591024] tot_loss[loss=2.327, over 5380607.54 frames. , ppl: 10.245098414120278], batch size: 70
+2022-12-10 17:59:05,018 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 17:59:05,778 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.158504303301722
+2022-12-10 18:00:47,946 INFO [train.py:421] (4/8) Epoch 3, batch 9200, loss[loss=2.433, over 3290.00 frames. , ppl: 11.388108478994594] tot_loss[loss=2.328, over 5358675.88 frames. , ppl: 10.257597251187397], batch size: 70
+2022-12-10 18:02:28,019 INFO [train.py:421] (4/8) Epoch 3, batch 9400, loss[loss=2.202, over 3360.00 frames. , ppl: 9.04186402525658] tot_loss[loss=2.327, over 5382998.70 frames. , ppl: 10.251681191817966], batch size: 70
+2022-12-10 18:04:09,126 INFO [train.py:421] (4/8) Epoch 3, batch 9600, loss[loss=2.557, over 980.00 frames. , ppl: 12.893185549314918] tot_loss[loss=2.328, over 5391228.86 frames. , ppl: 10.252590791557402], batch size: 70
+2022-12-10 18:05:47,189 INFO [train.py:421] (4/8) Epoch 3, batch 9800, loss[loss=2.349, over 3780.00 frames. , ppl: 10.479273317520686] tot_loss[loss=2.329, over 5352870.54 frames. , ppl: 10.264248057002803], batch size: 70
+2022-12-10 18:07:29,482 INFO [train.py:421] (4/8) Epoch 3, batch 10000, loss[loss=2.6, over 770.00 frames. , ppl: 13.463392259339383] tot_loss[loss=2.327, over 5424283.86 frames. , ppl: 10.250222604386403], batch size: 70
+2022-12-10 18:07:29,483 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:07:30,247 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.15942601173964
+2022-12-10 18:09:07,694 INFO [train.py:421] (4/8) Epoch 3, batch 10200, loss[loss=2.213, over 3990.00 frames. , ppl: 9.146885294404465] tot_loss[loss=2.329, over 5429837.94 frames. , ppl: 10.262690422974597], batch size: 70
+2022-12-10 18:10:48,660 INFO [train.py:421] (4/8) Epoch 3, batch 10400, loss[loss=2.587, over 980.00 frames. , ppl: 13.295584523566696] tot_loss[loss=2.327, over 5473865.36 frames. , ppl: 10.25153973919972], batch size: 70
+2022-12-10 18:12:27,603 INFO [train.py:421] (4/8) Epoch 3, batch 10600, loss[loss=2.764, over 770.00 frames. , ppl: 15.860116776545837] tot_loss[loss=2.326, over 5518754.95 frames. , ppl: 10.237533633569939], batch size: 70
+2022-12-10 18:14:05,263 INFO [train.py:421] (4/8) Epoch 3, batch 10800, loss[loss=2.199, over 6860.00 frames. , ppl: 9.020315051517377] tot_loss[loss=2.327, over 5497955.92 frames. , ppl: 10.247190674936181], batch size: 70
+2022-12-10 18:15:47,984 INFO [train.py:421] (4/8) Epoch 3, batch 11000, loss[loss=2.274, over 1120.00 frames. , ppl: 9.717451409667838] tot_loss[loss=2.326, over 5500712.81 frames. , ppl: 10.23865812466485], batch size: 70
+2022-12-10 18:15:47,985 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:15:48,745 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.319, over 211138.00 frames. , ppl: 10.163211456412954
+2022-12-10 18:17:26,886 INFO [train.py:421] (4/8) Epoch 3, batch 11200, loss[loss=2.385, over 3080.00 frames. , ppl: 10.854428637813307] tot_loss[loss=2.327, over 5500157.17 frames. , ppl: 10.250400234395512], batch size: 70
+2022-12-10 18:19:04,367 INFO [train.py:421] (4/8) Epoch 3, batch 11400, loss[loss=2.438, over 1260.00 frames. , ppl: 11.451623057076779] tot_loss[loss=2.328, over 5470230.96 frames. , ppl: 10.252981187496056], batch size: 70
+2022-12-10 18:20:41,827 INFO [train.py:421] (4/8) Epoch 3, batch 11600, loss[loss=2.468, over 1960.00 frames. , ppl: 11.801537634549037] tot_loss[loss=2.329, over 5388298.32 frames. , ppl: 10.272706195338172], batch size: 70
+2022-12-10 18:22:26,235 INFO [train.py:421] (4/8) Epoch 3, batch 11800, loss[loss=2.258, over 3920.00 frames. , ppl: 9.564536285126117] tot_loss[loss=2.329, over 5403782.43 frames. , ppl: 10.27161598795567], batch size: 70
+2022-12-10 18:24:10,841 INFO [train.py:421] (4/8) Epoch 3, batch 12000, loss[loss=2.572, over 1190.00 frames. , ppl: 13.089209876342629] tot_loss[loss=2.329, over 5424581.30 frames. , ppl: 10.271378177859614], batch size: 70
+2022-12-10 18:24:10,841 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:24:11,627 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.149768016214665
+2022-12-10 18:25:51,545 INFO [train.py:421] (4/8) Epoch 3, batch 12200, loss[loss=2.491, over 1400.00 frames. , ppl: 12.07312362494842] tot_loss[loss=2.329, over 5430723.44 frames. , ppl: 10.269367866478344], batch size: 70
+2022-12-10 18:27:29,145 INFO [train.py:421] (4/8) Epoch 3, batch 12400, loss[loss=2.429, over 1330.00 frames. , ppl: 11.348856408694733] tot_loss[loss=2.33, over 5415718.34 frames. , ppl: 10.273953182048361], batch size: 70
+2022-12-10 18:29:06,470 INFO [train.py:421] (4/8) Epoch 3, batch 12600, loss[loss=2.385, over 3150.00 frames. , ppl: 10.855338243980214] tot_loss[loss=2.331, over 5379717.48 frames. , ppl: 10.284284916477693], batch size: 70
+2022-12-10 18:30:46,101 INFO [train.py:421] (4/8) Epoch 3, batch 12800, loss[loss=2.513, over 1400.00 frames. , ppl: 12.345191086837225] tot_loss[loss=2.329, over 5441662.98 frames. , ppl: 10.266632272899672], batch size: 70
+2022-12-10 18:32:27,344 INFO [train.py:421] (4/8) Epoch 3, batch 13000, loss[loss=2.388, over 2940.00 frames. , ppl: 10.894709059817448] tot_loss[loss=2.329, over 5436590.83 frames. , ppl: 10.269220117337142], batch size: 70
+2022-12-10 18:32:27,345 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:32:28,107 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.318, over 211138.00 frames. , ppl: 10.154893455412658
+2022-12-10 18:34:13,397 INFO [train.py:421] (4/8) Epoch 3, batch 13200, loss[loss=2.438, over 1400.00 frames. , ppl: 11.45547970029777] tot_loss[loss=2.331, over 5406140.47 frames. , ppl: 10.2837686288533], batch size: 70
+2022-12-10 18:35:54,378 INFO [train.py:421] (4/8) Epoch 3, batch 13400, loss[loss=2.281, over 10500.00 frames. , ppl: 9.784849239794895] tot_loss[loss=2.33, over 5411373.48 frames. , ppl: 10.276896081828896], batch size: 70
+2022-12-10 18:37:33,219 INFO [train.py:421] (4/8) Epoch 3, batch 13600, loss[loss=2.708, over 840.00 frames. , ppl: 15.004217058563745] tot_loss[loss=2.331, over 5368422.71 frames. , ppl: 10.287093341641631], batch size: 70
+2022-12-10 18:39:15,490 INFO [train.py:421] (4/8) Epoch 3, batch 13800, loss[loss=2.676, over 770.00 frames. , ppl: 14.52488017250786] tot_loss[loss=2.332, over 5343492.22 frames. , ppl: 10.299418091373743], batch size: 70
+2022-12-10 18:40:59,171 INFO [train.py:421] (4/8) Epoch 3, batch 14000, loss[loss=2.412, over 2450.00 frames. , ppl: 11.157854109137464] tot_loss[loss=2.331, over 5378253.35 frames. , ppl: 10.285085689346369], batch size: 70
+2022-12-10 18:40:59,171 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:40:59,950 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.146131741243801
+2022-12-10 18:42:39,824 INFO [train.py:421] (4/8) Epoch 3, batch 14200, loss[loss=2.422, over 1120.00 frames. , ppl: 11.263816925012346] tot_loss[loss=2.331, over 5358338.43 frames. , ppl: 10.291658922698668], batch size: 70
+2022-12-10 18:44:22,939 INFO [train.py:421] (4/8) Epoch 3, batch 14400, loss[loss=2.511, over 1260.00 frames. , ppl: 12.314688778194844] tot_loss[loss=2.33, over 5427453.73 frames. , ppl: 10.27435396815317], batch size: 70
+2022-12-10 18:46:05,781 INFO [train.py:421] (4/8) Epoch 3, batch 14600, loss[loss=2.779, over 630.00 frames. , ppl: 16.09570457574241] tot_loss[loss=2.329, over 5463470.74 frames. , ppl: 10.262738642006866], batch size: 70
+2022-12-10 18:47:46,146 INFO [train.py:421] (4/8) Epoch 3, batch 14800, loss[loss=2.356, over 3780.00 frames. , ppl: 10.550440223444191] tot_loss[loss=2.328, over 5460152.26 frames. , ppl: 10.261518822136454], batch size: 70
+2022-12-10 18:49:30,929 INFO [train.py:421] (4/8) Epoch 3, batch 15000, loss[loss=2.367, over 2030.00 frames. , ppl: 10.666301324156652] tot_loss[loss=2.326, over 5548573.18 frames. , ppl: 10.23771670078776], batch size: 70
+2022-12-10 18:49:30,929 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:49:31,678 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.126931133917406
+2022-12-10 18:51:11,718 INFO [train.py:421] (4/8) Epoch 3, batch 15200, loss[loss=2.284, over 7280.00 frames. , ppl: 9.8205166146157] tot_loss[loss=2.326, over 5544466.91 frames. , ppl: 10.2341550759247], batch size: 70
+2022-12-10 18:52:50,273 INFO [train.py:421] (4/8) Epoch 3, batch 15400, loss[loss=2.3, over 4200.00 frames. , ppl: 9.974340157978975] tot_loss[loss=2.326, over 5538941.42 frames. , ppl: 10.239524670003577], batch size: 70
+2022-12-10 18:54:28,574 INFO [train.py:421] (4/8) Epoch 3, batch 15600, loss[loss=2.511, over 1470.00 frames. , ppl: 12.319650528598476] tot_loss[loss=2.327, over 5504221.49 frames. , ppl: 10.2521434424315], batch size: 70
+2022-12-10 18:56:05,397 INFO [train.py:421] (4/8) Epoch 3, batch 15800, loss[loss=2.49, over 1610.00 frames. , ppl: 12.061359521929067] tot_loss[loss=2.328, over 5461772.67 frames. , ppl: 10.258774595919409], batch size: 70
+2022-12-10 18:57:45,622 INFO [train.py:421] (4/8) Epoch 3, batch 16000, loss[loss=2.387, over 2450.00 frames. , ppl: 10.877666850609003] tot_loss[loss=2.329, over 5418957.12 frames. , ppl: 10.26983083839731], batch size: 70
+2022-12-10 18:57:45,623 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 18:57:46,369 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.316, over 211138.00 frames. , ppl: 10.134522713241985
+2022-12-10 18:59:25,143 INFO [train.py:421] (4/8) Epoch 3, batch 16200, loss[loss=2.269, over 2660.00 frames. , ppl: 9.673459975388917] tot_loss[loss=2.328, over 5455116.68 frames. , ppl: 10.26028071273736], batch size: 70
+2022-12-10 19:01:05,960 INFO [train.py:421] (4/8) Epoch 3, batch 16400, loss[loss=2.522, over 1120.00 frames. , ppl: 12.45257902192687] tot_loss[loss=2.329, over 5454758.40 frames. , ppl: 10.267102117553538], batch size: 70
+2022-12-10 19:02:45,402 INFO [train.py:421] (4/8) Epoch 3, batch 16600, loss[loss=2.469, over 1190.00 frames. , ppl: 11.813852081101212] tot_loss[loss=2.329, over 5471518.99 frames. , ppl: 10.264038322761003], batch size: 70
+2022-12-10 19:04:28,700 INFO [train.py:421] (4/8) Epoch 3, batch 16800, loss[loss=2.47, over 1610.00 frames. , ppl: 11.821734789814702] tot_loss[loss=2.327, over 5508473.97 frames. , ppl: 10.250247623286622], batch size: 70
+2022-12-10 19:06:10,463 INFO [train.py:421] (4/8) Epoch 3, batch 17000, loss[loss=2.25, over 5460.00 frames. , ppl: 9.487961533880823] tot_loss[loss=2.328, over 5518145.25 frames. , ppl: 10.262292890943655], batch size: 70
+2022-12-10 19:06:10,464 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:06:11,210 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.149083017239782
+2022-12-10 19:07:50,801 INFO [train.py:421] (4/8) Epoch 3, batch 17200, loss[loss=2.509, over 770.00 frames. , ppl: 12.298639765228119] tot_loss[loss=2.329, over 5470838.79 frames. , ppl: 10.272354441309965], batch size: 70
+2022-12-10 19:09:33,685 INFO [train.py:421] (4/8) Epoch 3, batch 17400, loss[loss=2.64, over 1400.00 frames. , ppl: 14.018487902195679] tot_loss[loss=2.329, over 5493949.15 frames. , ppl: 10.265121243407819], batch size: 70
+2022-12-10 19:11:12,084 INFO [train.py:421] (4/8) Epoch 3, batch 17600, loss[loss=2.276, over 3010.00 frames. , ppl: 9.74141920606809] tot_loss[loss=2.33, over 5469243.11 frames. , ppl: 10.27563207111392], batch size: 70
+2022-12-10 19:12:53,522 INFO [train.py:421] (4/8) Epoch 3, batch 17800, loss[loss=2.346, over 1960.00 frames. , ppl: 10.447539955913735] tot_loss[loss=2.328, over 5515075.36 frames. , ppl: 10.257078285269822], batch size: 70
+2022-12-10 19:14:33,720 INFO [train.py:421] (4/8) Epoch 3, batch 18000, loss[loss=2.266, over 8890.00 frames. , ppl: 9.638096637605203] tot_loss[loss=2.329, over 5480172.39 frames. , ppl: 10.270243693427322], batch size: 70
+2022-12-10 19:14:33,721 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:14:34,466 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.129355082846555
+2022-12-10 19:16:15,835 INFO [train.py:421] (4/8) Epoch 3, batch 18200, loss[loss=2.687, over 770.00 frames. , ppl: 14.681852626061053] tot_loss[loss=2.33, over 5463276.82 frames. , ppl: 10.280270302327505], batch size: 70
+2022-12-10 19:17:56,507 INFO [train.py:421] (4/8) Epoch 3, batch 18400, loss[loss=2.476, over 1120.00 frames. , ppl: 11.893386008895169] tot_loss[loss=2.33, over 5456845.21 frames. , ppl: 10.2749566639767], batch size: 70
+2022-12-10 19:19:35,454 INFO [train.py:421] (4/8) Epoch 3, batch 18600, loss[loss=2.261, over 4830.00 frames. , ppl: 9.592426622399815] tot_loss[loss=2.329, over 5452940.90 frames. , ppl: 10.271924282177217], batch size: 70
+2022-12-10 19:21:19,554 INFO [train.py:421] (4/8) Epoch 3, batch 18800, loss[loss=2.211, over 6790.00 frames. , ppl: 9.122531708245788] tot_loss[loss=2.328, over 5483670.71 frames. , ppl: 10.260319301542106], batch size: 70
+2022-12-10 19:22:57,474 INFO [train.py:421] (4/8) Epoch 3, batch 19000, loss[loss=2.381, over 1750.00 frames. , ppl: 10.814729420450798] tot_loss[loss=2.329, over 5439969.53 frames. , ppl: 10.27171830083427], batch size: 70
+2022-12-10 19:22:57,474 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:22:58,219 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.317, over 211138.00 frames. , ppl: 10.14855577978015
+2022-12-10 19:24:36,563 INFO [train.py:421] (4/8) Epoch 3, batch 19200, loss[loss=2.407, over 2100.00 frames. , ppl: 11.095366877588335] tot_loss[loss=2.331, over 5386341.20 frames. , ppl: 10.28677893214436], batch size: 70
+2022-12-10 19:26:16,122 INFO [train.py:421] (4/8) Epoch 3, batch 19400, loss[loss=2.409, over 3360.00 frames. , ppl: 11.123237629802382] tot_loss[loss=2.332, over 5345282.08 frames. , ppl: 10.30162249557107], batch size: 70
+2022-12-10 19:27:55,650 INFO [train.py:421] (4/8) Epoch 3, batch 19600, loss[loss=2.511, over 1960.00 frames. , ppl: 12.322106060775992] tot_loss[loss=2.332, over 5357997.92 frames. , ppl: 10.298689166296542], batch size: 70
+2022-12-10 19:29:36,902 INFO [train.py:421] (4/8) Epoch 3, batch 19800, loss[loss=2.458, over 1400.00 frames. , ppl: 11.683586444121426] tot_loss[loss=2.333, over 5328187.50 frames. , ppl: 10.304511569701603], batch size: 70
+2022-12-10 19:31:19,065 INFO [train.py:421] (4/8) Epoch 3, batch 20000, loss[loss=2.495, over 1330.00 frames. , ppl: 12.118170292623125] tot_loss[loss=2.332, over 5318354.70 frames. , ppl: 10.302594148213155], batch size: 70
+2022-12-10 19:31:19,066 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:31:19,811 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.12899827473271
+2022-12-10 19:33:00,567 INFO [train.py:421] (4/8) Epoch 3, batch 20200, loss[loss=2.435, over 1750.00 frames. , ppl: 11.41056431446478] tot_loss[loss=2.333, over 5295850.37 frames. , ppl: 10.310855067263034], batch size: 70
+2022-12-10 19:34:42,566 INFO [train.py:421] (4/8) Epoch 3, batch 20400, loss[loss=2.621, over 840.00 frames. , ppl: 13.746575475878371] tot_loss[loss=2.333, over 5309258.54 frames. , ppl: 10.310783060952758], batch size: 70
+2022-12-10 19:36:23,578 INFO [train.py:421] (4/8) Epoch 3, batch 20600, loss[loss=2.365, over 2520.00 frames. , ppl: 10.642309829621466] tot_loss[loss=2.333, over 5316161.03 frames. , ppl: 10.309836972597553], batch size: 70
+2022-12-10 19:38:05,419 INFO [train.py:421] (4/8) Epoch 3, batch 20800, loss[loss=2.986, over 630.00 frames. , ppl: 19.811689210408527] tot_loss[loss=2.332, over 5323707.03 frames. , ppl: 10.298238072709088], batch size: 70
+2022-12-10 19:39:45,185 INFO [train.py:421] (4/8) Epoch 3, batch 21000, loss[loss=2.258, over 8330.00 frames. , ppl: 9.559248329865351] tot_loss[loss=2.332, over 5314061.74 frames. , ppl: 10.295000625175312], batch size: 70
+2022-12-10 19:39:45,186 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:39:45,934 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.115766279020953
+2022-12-10 19:41:26,043 INFO [train.py:421] (4/8) Epoch 3, batch 21200, loss[loss=2.198, over 5600.00 frames. , ppl: 9.003125970604072] tot_loss[loss=2.332, over 5303642.81 frames. , ppl: 10.29853320718995], batch size: 70
+2022-12-10 19:43:03,101 INFO [train.py:421] (4/8) Epoch 3, batch 21400, loss[loss=2.901, over 560.00 frames. , ppl: 18.19654247728769] tot_loss[loss=2.33, over 5346993.58 frames. , ppl: 10.277806858557291], batch size: 70
+2022-12-10 19:44:44,717 INFO [train.py:421] (4/8) Epoch 3, batch 21600, loss[loss=2.702, over 840.00 frames. , ppl: 14.915583352735524] tot_loss[loss=2.329, over 5379651.16 frames. , ppl: 10.271563096786256], batch size: 70
+2022-12-10 19:46:24,994 INFO [train.py:421] (4/8) Epoch 3, batch 21800, loss[loss=2.438, over 1470.00 frames. , ppl: 11.455442410385483] tot_loss[loss=2.33, over 5365533.22 frames. , ppl: 10.274949276873977], batch size: 70
+2022-12-10 19:48:03,093 INFO [train.py:421] (4/8) Epoch 3, batch 22000, loss[loss=2.323, over 2660.00 frames. , ppl: 10.206026768787014] tot_loss[loss=2.33, over 5322484.31 frames. , ppl: 10.281639128637668], batch size: 70
+2022-12-10 19:48:03,094 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:48:03,838 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.119331761662655
+2022-12-10 19:49:42,923 INFO [train.py:421] (4/8) Epoch 3, batch 22200, loss[loss=2.168, over 7910.00 frames. , ppl: 8.739303327375074] tot_loss[loss=2.33, over 5332808.65 frames. , ppl: 10.280541013093806], batch size: 70
+2022-12-10 19:51:27,707 INFO [train.py:421] (4/8) Epoch 3, batch 22400, loss[loss=2.431, over 1190.00 frames. , ppl: 11.370499886531269] tot_loss[loss=2.329, over 5363469.30 frames. , ppl: 10.272385805005088], batch size: 70
+2022-12-10 19:53:09,801 INFO [train.py:421] (4/8) Epoch 3, batch 22600, loss[loss=2.541, over 700.00 frames. , ppl: 12.687551778917511] tot_loss[loss=2.328, over 5385384.69 frames. , ppl: 10.262265997760188], batch size: 70
+2022-12-10 19:54:48,481 INFO [train.py:421] (4/8) Epoch 3, batch 22800, loss[loss=2.204, over 7980.00 frames. , ppl: 9.057219078649252] tot_loss[loss=2.327, over 5422217.71 frames. , ppl: 10.250685042783074], batch size: 70
+2022-12-10 19:56:25,441 INFO [train.py:421] (4/8) Epoch 3, batch 23000, loss[loss=2.303, over 5530.00 frames. , ppl: 10.001273925157136] tot_loss[loss=2.327, over 5430405.72 frames. , ppl: 10.245485255961368], batch size: 70
+2022-12-10 19:56:25,442 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 19:56:26,223 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.111284628584807
+2022-12-10 19:58:04,904 INFO [train.py:421] (4/8) Epoch 3, batch 23200, loss[loss=2.233, over 6510.00 frames. , ppl: 9.325763585055626] tot_loss[loss=2.327, over 5443458.65 frames. , ppl: 10.245443143717132], batch size: 70
+2022-12-10 19:59:44,528 INFO [train.py:421] (4/8) Epoch 3, batch 23400, loss[loss=2.913, over 630.00 frames. , ppl: 18.412427995610777] tot_loss[loss=2.327, over 5437491.97 frames. , ppl: 10.249472911381265], batch size: 70
+2022-12-10 20:01:22,837 INFO [train.py:421] (4/8) Epoch 3, batch 23600, loss[loss=2.266, over 4620.00 frames. , ppl: 9.645570873969213] tot_loss[loss=2.328, over 5434396.10 frames. , ppl: 10.254920347622267], batch size: 70
+2022-12-10 20:03:03,117 INFO [train.py:421] (4/8) Epoch 3, batch 23800, loss[loss=2.492, over 2030.00 frames. , ppl: 12.08087581961683] tot_loss[loss=2.328, over 5438652.03 frames. , ppl: 10.254485467269015], batch size: 70
+2022-12-10 20:04:45,087 INFO [train.py:421] (4/8) Epoch 3, batch 24000, loss[loss=2.339, over 1960.00 frames. , ppl: 10.373167634967814] tot_loss[loss=2.327, over 5434179.09 frames. , ppl: 10.25038105880789], batch size: 70
+2022-12-10 20:04:45,087 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:04:45,874 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.115662972133224
+2022-12-10 20:06:27,302 INFO [train.py:421] (4/8) Epoch 3, batch 24200, loss[loss=3.279, over 490.00 frames. , ppl: 26.553203263200388] tot_loss[loss=2.328, over 5399531.64 frames. , ppl: 10.258087184890801], batch size: 70
+2022-12-10 20:08:06,570 INFO [train.py:421] (4/8) Epoch 3, batch 24400, loss[loss=2.565, over 1120.00 frames. , ppl: 12.996615563301557] tot_loss[loss=2.328, over 5422492.85 frames. , ppl: 10.261086139098163], batch size: 70
+2022-12-10 20:09:46,778 INFO [train.py:421] (4/8) Epoch 3, batch 24600, loss[loss=2.464, over 1190.00 frames. , ppl: 11.748736849815861] tot_loss[loss=2.328, over 5446201.52 frames. , ppl: 10.254155642038276], batch size: 70
+2022-12-10 20:11:29,178 INFO [train.py:421] (4/8) Epoch 3, batch 24800, loss[loss=2.295, over 4060.00 frames. , ppl: 9.920362459623599] tot_loss[loss=2.327, over 5480578.70 frames. , ppl: 10.249400984019132], batch size: 70
+2022-12-10 20:13:09,469 INFO [train.py:421] (4/8) Epoch 3, batch 25000, loss[loss=2.302, over 3920.00 frames. , ppl: 9.99013269623165] tot_loss[loss=2.326, over 5490741.64 frames. , ppl: 10.240087699410191], batch size: 70
+2022-12-10 20:13:09,470 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:13:10,215 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.315, over 211138.00 frames. , ppl: 10.122748675677348
+2022-12-10 20:14:53,081 INFO [train.py:421] (4/8) Epoch 3, batch 25200, loss[loss=2.468, over 980.00 frames. , ppl: 11.80031170343094] tot_loss[loss=2.327, over 5464603.15 frames. , ppl: 10.244365690372376], batch size: 70
+2022-12-10 20:16:34,342 INFO [train.py:421] (4/8) Epoch 3, batch 25400, loss[loss=2.272, over 9940.00 frames. , ppl: 9.702802506821827] tot_loss[loss=2.325, over 5498427.22 frames. , ppl: 10.230262490267265], batch size: 70
+2022-12-10 20:18:12,371 INFO [train.py:421] (4/8) Epoch 3, batch 25600, loss[loss=2.414, over 1330.00 frames. , ppl: 11.181981672610469] tot_loss[loss=2.327, over 5447593.40 frames. , ppl: 10.248258025721809], batch size: 70
+2022-12-10 20:19:49,966 INFO [train.py:421] (4/8) Epoch 3, batch 25800, loss[loss=2.573, over 910.00 frames. , ppl: 13.10527667755332] tot_loss[loss=2.327, over 5469261.83 frames. , ppl: 10.249078922179997], batch size: 70
+2022-12-10 20:21:34,321 INFO [train.py:421] (4/8) Epoch 3, batch 26000, loss[loss=2.224, over 4900.00 frames. , ppl: 9.247260704996219] tot_loss[loss=2.325, over 5537201.21 frames. , ppl: 10.224716372049587], batch size: 70
+2022-12-10 20:21:34,321 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:21:35,067 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.316, over 211138.00 frames. , ppl: 10.132153014018082
+2022-12-10 20:23:08,526 INFO [train.py:421] (4/8) Epoch 3, batch 26200, loss[loss=2.435, over 1680.00 frames. , ppl: 11.41999542981506] tot_loss[loss=2.326, over 5512198.53 frames. , ppl: 10.237111703297618], batch size: 70
+2022-12-10 20:24:45,539 INFO [train.py:421] (4/8) Epoch 3, batch 26400, loss[loss=2.306, over 2520.00 frames. , ppl: 10.03120449096482] tot_loss[loss=2.326, over 5487549.11 frames. , ppl: 10.23764985502242], batch size: 70
+2022-12-10 20:26:27,607 INFO [train.py:421] (4/8) Epoch 3, batch 26600, loss[loss=2.474, over 1260.00 frames. , ppl: 11.86882162785055] tot_loss[loss=2.326, over 5466307.59 frames. , ppl: 10.23703081555514], batch size: 70
+2022-12-10 20:28:07,421 INFO [train.py:421] (4/8) Epoch 3, batch 26800, loss[loss=2.79, over 910.00 frames. , ppl: 16.277406139368765] tot_loss[loss=2.327, over 5426088.17 frames. , ppl: 10.250411968357348], batch size: 70
+2022-12-10 20:29:49,943 INFO [train.py:421] (4/8) Epoch 3, batch 27000, loss[loss=2.441, over 2100.00 frames. , ppl: 11.484194281394354] tot_loss[loss=2.327, over 5449457.61 frames. , ppl: 10.242944234884227], batch size: 70
+2022-12-10 20:29:49,944 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:29:50,703 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.101093802370023
+2022-12-10 20:31:29,520 INFO [train.py:421] (4/8) Epoch 3, batch 27200, loss[loss=2.493, over 1610.00 frames. , ppl: 12.096461192879799] tot_loss[loss=2.325, over 5480005.24 frames. , ppl: 10.228744930837529], batch size: 70
+2022-12-10 20:33:11,514 INFO [train.py:421] (4/8) Epoch 3, batch 27400, loss[loss=2.279, over 4480.00 frames. , ppl: 9.76544166289031] tot_loss[loss=2.325, over 5498087.08 frames. , ppl: 10.221844499113786], batch size: 70
+2022-12-10 20:34:50,136 INFO [train.py:421] (4/8) Epoch 3, batch 27600, loss[loss=2.266, over 3990.00 frames. , ppl: 9.6453813697707] tot_loss[loss=2.324, over 5521659.44 frames. , ppl: 10.216867254940896], batch size: 70
+2022-12-10 20:36:32,938 INFO [train.py:421] (4/8) Epoch 3, batch 27800, loss[loss=2.215, over 9730.00 frames. , ppl: 9.16505397996692] tot_loss[loss=2.323, over 5595097.60 frames. , ppl: 10.203582648235532], batch size: 70
+2022-12-10 20:38:09,408 INFO [train.py:421] (4/8) Epoch 3, batch 28000, loss[loss=2.421, over 1400.00 frames. , ppl: 11.254256051183486] tot_loss[loss=2.324, over 5550092.63 frames. , ppl: 10.211539946185102], batch size: 70
+2022-12-10 20:38:09,408 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:38:10,154 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.106116881121231
+2022-12-10 20:39:49,677 INFO [train.py:421] (4/8) Epoch 3, batch 28200, loss[loss=2.45, over 1190.00 frames. , ppl: 11.593373775375467] tot_loss[loss=2.324, over 5557822.08 frames. , ppl: 10.21292214141284], batch size: 70
+2022-12-10 20:41:31,327 INFO [train.py:421] (4/8) Epoch 3, batch 28400, loss[loss=2.518, over 1330.00 frames. , ppl: 12.404364877741427] tot_loss[loss=2.323, over 5592079.88 frames. , ppl: 10.204677235756437], batch size: 70
+2022-12-10 20:43:09,535 INFO [train.py:421] (4/8) Epoch 3, batch 28600, loss[loss=2.358, over 1960.00 frames. , ppl: 10.57359202578521] tot_loss[loss=2.323, over 5603981.05 frames. , ppl: 10.201344686108602], batch size: 70
+2022-12-10 20:44:46,042 INFO [train.py:421] (4/8) Epoch 3, batch 28800, loss[loss=2.529, over 2030.00 frames. , ppl: 12.540962645665731] tot_loss[loss=2.324, over 5566605.50 frames. , ppl: 10.220760100868736], batch size: 70
+2022-12-10 20:46:28,959 INFO [train.py:421] (4/8) Epoch 3, batch 29000, loss[loss=3.274, over 490.00 frames. , ppl: 26.407261801779747] tot_loss[loss=2.323, over 5604140.63 frames. , ppl: 10.207191722138825], batch size: 70
+2022-12-10 20:46:28,959 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:46:29,722 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.10355792431578
+2022-12-10 20:48:07,969 INFO [train.py:421] (4/8) Epoch 3, batch 29200, loss[loss=2.379, over 1330.00 frames. , ppl: 10.792041462655554] tot_loss[loss=2.323, over 5607819.86 frames. , ppl: 10.205899725815517], batch size: 70
+2022-12-10 20:49:48,613 INFO [train.py:421] (4/8) Epoch 3, batch 29400, loss[loss=2.491, over 1050.00 frames. , ppl: 12.07582303111856] tot_loss[loss=2.323, over 5622512.59 frames. , ppl: 10.20217059195424], batch size: 70
+2022-12-10 20:51:26,256 INFO [train.py:421] (4/8) Epoch 3, batch 29600, loss[loss=2.383, over 1820.00 frames. , ppl: 10.836192018497998] tot_loss[loss=2.324, over 5571186.71 frames. , ppl: 10.211957068272598], batch size: 70
+2022-12-10 20:53:01,799 INFO [train.py:421] (4/8) Epoch 3, batch 29800, loss[loss=2.222, over 6160.00 frames. , ppl: 9.225160686451243] tot_loss[loss=2.324, over 5561193.75 frames. , ppl: 10.218229260963483], batch size: 70
+2022-12-10 20:54:41,555 INFO [train.py:421] (4/8) Epoch 3, batch 30000, loss[loss=2.983, over 560.00 frames. , ppl: 19.750549756461208] tot_loss[loss=2.325, over 5522288.39 frames. , ppl: 10.231076357215855], batch size: 70
+2022-12-10 20:54:41,555 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 20:54:42,324 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.100455441623417
+2022-12-10 20:56:19,484 INFO [train.py:421] (4/8) Epoch 3, batch 30200, loss[loss=2.355, over 2100.00 frames. , ppl: 10.541699512463634] tot_loss[loss=2.326, over 5497148.03 frames. , ppl: 10.238356447450828], batch size: 70
+2022-12-10 20:58:04,768 INFO [train.py:421] (4/8) Epoch 3, batch 30400, loss[loss=5.021, over 280.00 frames. , ppl: 151.55413774538474] tot_loss[loss=2.327, over 5487455.79 frames. , ppl: 10.245878044911764], batch size: 70
+2022-12-10 20:59:43,369 INFO [train.py:421] (4/8) Epoch 3, batch 30600, loss[loss=2.557, over 1400.00 frames. , ppl: 12.893133193981251] tot_loss[loss=2.327, over 5481294.13 frames. , ppl: 10.2518259131518], batch size: 70
+2022-12-10 21:01:20,859 INFO [train.py:421] (4/8) Epoch 3, batch 30800, loss[loss=2.32, over 3430.00 frames. , ppl: 10.177630351640387] tot_loss[loss=2.328, over 5445556.79 frames. , ppl: 10.260331155336477], batch size: 70
+2022-12-10 21:02:59,411 INFO [train.py:421] (4/8) Epoch 3, batch 31000, loss[loss=2.387, over 2870.00 frames. , ppl: 10.880780086561012] tot_loss[loss=2.328, over 5451937.03 frames. , ppl: 10.252696462117624], batch size: 70
+2022-12-10 21:02:59,412 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:03:00,157 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.091295598195819
+2022-12-10 21:04:37,260 INFO [train.py:421] (4/8) Epoch 3, batch 31200, loss[loss=2.296, over 4620.00 frames. , ppl: 9.938266780549412] tot_loss[loss=2.327, over 5467506.11 frames. , ppl: 10.247691518240584], batch size: 70
+2022-12-10 21:06:17,344 INFO [train.py:421] (4/8) Epoch 3, batch 31400, loss[loss=2.267, over 3010.00 frames. , ppl: 9.653631314566734] tot_loss[loss=2.327, over 5472383.69 frames. , ppl: 10.24252340732936], batch size: 70
+2022-12-10 21:07:59,103 INFO [train.py:421] (4/8) Epoch 3, batch 31600, loss[loss=2.443, over 1330.00 frames. , ppl: 11.505567822898781] tot_loss[loss=2.327, over 5454236.47 frames. , ppl: 10.247344163330828], batch size: 70
+2022-12-10 21:09:38,631 INFO [train.py:421] (4/8) Epoch 3, batch 31800, loss[loss=3.19, over 490.00 frames. , ppl: 24.291802818323276] tot_loss[loss=2.327, over 5463864.94 frames. , ppl: 10.242260416142079], batch size: 70
+2022-12-10 21:11:16,593 INFO [train.py:421] (4/8) Epoch 3, batch 32000, loss[loss=2.282, over 3640.00 frames. , ppl: 9.796559474451028] tot_loss[loss=2.327, over 5469775.10 frames. , ppl: 10.243788084570381], batch size: 70
+2022-12-10 21:11:16,594 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:11:17,340 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.101464578420366
+2022-12-10 21:12:58,747 INFO [train.py:421] (4/8) Epoch 3, batch 32200, loss[loss=2.246, over 2660.00 frames. , ppl: 9.448732747835278] tot_loss[loss=2.325, over 5482425.91 frames. , ppl: 10.225182875544343], batch size: 70
+2022-12-10 21:14:34,252 INFO [train.py:421] (4/8) Epoch 3, batch 32400, loss[loss=2.246, over 3220.00 frames. , ppl: 9.450849844433913] tot_loss[loss=2.326, over 5476191.94 frames. , ppl: 10.23628616126285], batch size: 70
+2022-12-10 21:16:14,624 INFO [train.py:421] (4/8) Epoch 3, batch 32600, loss[loss=2.635, over 1190.00 frames. , ppl: 13.943122519778578] tot_loss[loss=2.326, over 5450419.17 frames. , ppl: 10.238051865632249], batch size: 70
+2022-12-10 21:17:51,468 INFO [train.py:421] (4/8) Epoch 3, batch 32800, loss[loss=2.368, over 2450.00 frames. , ppl: 10.676219312632494] tot_loss[loss=2.327, over 5430222.14 frames. , ppl: 10.245400595618639], batch size: 70
+2022-12-10 21:19:29,969 INFO [train.py:421] (4/8) Epoch 3, batch 33000, loss[loss=2.319, over 1680.00 frames. , ppl: 10.162717497176535] tot_loss[loss=2.327, over 5405161.93 frames. , ppl: 10.245058222156274], batch size: 70
+2022-12-10 21:19:29,970 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:19:30,729 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.314, over 211138.00 frames. , ppl: 10.111637819514526
+2022-12-10 21:21:05,638 INFO [train.py:421] (4/8) Epoch 3, batch 33200, loss[loss=2.47, over 910.00 frames. , ppl: 11.825805632333182] tot_loss[loss=2.326, over 5411377.31 frames. , ppl: 10.237733998148586], batch size: 70
+2022-12-10 21:22:50,813 INFO [train.py:421] (4/8) Epoch 3, batch 33400, loss[loss=2.444, over 1540.00 frames. , ppl: 11.51401686129846] tot_loss[loss=2.325, over 5447966.74 frames. , ppl: 10.229979526593956], batch size: 70
+2022-12-10 21:24:31,309 INFO [train.py:421] (4/8) Epoch 3, batch 33600, loss[loss=2.373, over 2100.00 frames. , ppl: 10.725515812205456] tot_loss[loss=2.327, over 5410542.14 frames. , ppl: 10.2426513656091], batch size: 70
+2022-12-10 21:26:13,403 INFO [train.py:421] (4/8) Epoch 3, batch 33800, loss[loss=2.437, over 1190.00 frames. , ppl: 11.438664092233248] tot_loss[loss=2.326, over 5420705.73 frames. , ppl: 10.241646949152976], batch size: 70
+2022-12-10 21:27:54,537 INFO [train.py:421] (4/8) Epoch 3, batch 34000, loss[loss=2.411, over 1680.00 frames. , ppl: 11.145806724819094] tot_loss[loss=2.326, over 5435499.62 frames. , ppl: 10.237168683405978], batch size: 70
+2022-12-10 21:27:54,537 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:27:55,295 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.071958753555176
+2022-12-10 21:29:32,475 INFO [train.py:421] (4/8) Epoch 3, batch 34200, loss[loss=2.457, over 1050.00 frames. , ppl: 11.666500620349764] tot_loss[loss=2.326, over 5422822.73 frames. , ppl: 10.241650006145802], batch size: 70
+2022-12-10 21:31:11,666 INFO [train.py:421] (4/8) Epoch 3, batch 34400, loss[loss=2.473, over 1540.00 frames. , ppl: 11.85407425751827] tot_loss[loss=2.327, over 5411398.51 frames. , ppl: 10.24322851517933], batch size: 70
+2022-12-10 21:32:50,594 INFO [train.py:421] (4/8) Epoch 3, batch 34600, loss[loss=2.686, over 770.00 frames. , ppl: 14.674704114009147] tot_loss[loss=2.326, over 5418460.87 frames. , ppl: 10.240391508417451], batch size: 70
+2022-12-10 21:34:29,939 INFO [train.py:421] (4/8) Epoch 3, batch 34800, loss[loss=2.322, over 2590.00 frames. , ppl: 10.191022509381032] tot_loss[loss=2.328, over 5377822.09 frames. , ppl: 10.253871631586314], batch size: 70
+2022-12-10 21:36:14,331 INFO [train.py:421] (4/8) Epoch 3, batch 35000, loss[loss=2.424, over 2030.00 frames. , ppl: 11.286016722304545] tot_loss[loss=2.327, over 5418068.53 frames. , ppl: 10.24875740063115], batch size: 70
+2022-12-10 21:36:14,332 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:36:15,077 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.083123013689438
+2022-12-10 21:37:54,319 INFO [train.py:421] (4/8) Epoch 3, batch 35200, loss[loss=2.207, over 3710.00 frames. , ppl: 9.08513748869419] tot_loss[loss=2.327, over 5420587.78 frames. , ppl: 10.2510329337838], batch size: 70
+2022-12-10 21:39:36,771 INFO [train.py:421] (4/8) Epoch 3, batch 35400, loss[loss=2.305, over 2870.00 frames. , ppl: 10.025684611892542] tot_loss[loss=2.328, over 5388496.51 frames. , ppl: 10.257003974832132], batch size: 70
+2022-12-10 21:41:19,937 INFO [train.py:421] (4/8) Epoch 3, batch 35600, loss[loss=2.258, over 8190.00 frames. , ppl: 9.560147132452137] tot_loss[loss=2.327, over 5412333.43 frames. , ppl: 10.248783706225511], batch size: 70
+2022-12-10 21:42:54,213 INFO [train.py:421] (4/8) Epoch 3, batch 35800, loss[loss=2.34, over 1610.00 frames. , ppl: 10.379875902916057] tot_loss[loss=2.327, over 5430878.20 frames. , ppl: 10.243800563126008], batch size: 70
+2022-12-10 21:44:36,435 INFO [train.py:421] (4/8) Epoch 3, batch 36000, loss[loss=2.4, over 1680.00 frames. , ppl: 11.022671791221251] tot_loss[loss=2.327, over 5425835.10 frames. , ppl: 10.24427220778548], batch size: 70
+2022-12-10 21:44:36,435 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:44:37,181 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.086282871808761
+2022-12-10 21:46:17,616 INFO [train.py:421] (4/8) Epoch 3, batch 36200, loss[loss=2.325, over 3990.00 frames. , ppl: 10.231016210459908] tot_loss[loss=2.327, over 5413384.01 frames. , ppl: 10.245368312161144], batch size: 70
+2022-12-10 21:47:56,619 INFO [train.py:421] (4/8) Epoch 3, batch 36400, loss[loss=2.378, over 1120.00 frames. , ppl: 10.782086406543204] tot_loss[loss=2.328, over 5390735.19 frames. , ppl: 10.25480463961725], batch size: 70
+2022-12-10 21:49:39,008 INFO [train.py:421] (4/8) Epoch 3, batch 36600, loss[loss=2.331, over 2590.00 frames. , ppl: 10.290548665901277] tot_loss[loss=2.328, over 5422956.07 frames. , ppl: 10.253804208648729], batch size: 70
+2022-12-10 21:51:17,824 INFO [train.py:421] (4/8) Epoch 3, batch 36800, loss[loss=2.339, over 2870.00 frames. , ppl: 10.37526255756793] tot_loss[loss=2.327, over 5434270.31 frames. , ppl: 10.248364535723415], batch size: 70
+2022-12-10 21:53:00,121 INFO [train.py:421] (4/8) Epoch 3, batch 37000, loss[loss=2.487, over 1260.00 frames. , ppl: 12.029305108833176] tot_loss[loss=2.327, over 5440811.91 frames. , ppl: 10.242684510328953], batch size: 70
+2022-12-10 21:53:00,122 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 21:53:00,870 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.091421060289873
+2022-12-10 21:54:40,044 INFO [train.py:421] (4/8) Epoch 3, batch 37200, loss[loss=2.32, over 4830.00 frames. , ppl: 10.173264567654066] tot_loss[loss=2.326, over 5473852.19 frames. , ppl: 10.235136253000832], batch size: 70
+2022-12-10 21:56:18,550 INFO [train.py:421] (4/8) Epoch 3, batch 37400, loss[loss=2.422, over 1680.00 frames. , ppl: 11.271691423306004] tot_loss[loss=2.325, over 5484992.78 frames. , ppl: 10.226833489578818], batch size: 70
+2022-12-10 21:57:58,919 INFO [train.py:421] (4/8) Epoch 3, batch 37600, loss[loss=2.856, over 630.00 frames. , ppl: 17.389132186589112] tot_loss[loss=2.325, over 5505971.36 frames. , ppl: 10.22319734463344], batch size: 70
+2022-12-10 21:59:41,113 INFO [train.py:421] (4/8) Epoch 3, batch 37800, loss[loss=2.292, over 3640.00 frames. , ppl: 9.897539881041668] tot_loss[loss=2.325, over 5486923.92 frames. , ppl: 10.223913530234015], batch size: 70
+2022-12-10 22:01:19,423 INFO [train.py:421] (4/8) Epoch 3, batch 38000, loss[loss=2.562, over 1050.00 frames. , ppl: 12.955602503106105] tot_loss[loss=2.324, over 5524843.16 frames. , ppl: 10.212921737696396], batch size: 70
+2022-12-10 22:01:19,423 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:01:20,188 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.088428318438387
+2022-12-10 22:02:59,176 INFO [train.py:421] (4/8) Epoch 3, batch 38200, loss[loss=2.257, over 4690.00 frames. , ppl: 9.557622489165405] tot_loss[loss=2.322, over 5577111.33 frames. , ppl: 10.193169619103552], batch size: 70
+2022-12-10 22:04:43,431 INFO [train.py:421] (4/8) Epoch 3, batch 38400, loss[loss=2.229, over 7350.00 frames. , ppl: 9.286972548856356] tot_loss[loss=2.321, over 5624657.00 frames. , ppl: 10.183159268638699], batch size: 70
+2022-12-10 22:06:21,541 INFO [train.py:421] (4/8) Epoch 3, batch 38600, loss[loss=2.581, over 910.00 frames. , ppl: 13.203878374391564] tot_loss[loss=2.32, over 5643299.63 frames. , ppl: 10.178433492195163], batch size: 70
+2022-12-10 22:08:03,093 INFO [train.py:421] (4/8) Epoch 3, batch 38800, loss[loss=2.179, over 6720.00 frames. , ppl: 8.837496326991781] tot_loss[loss=2.32, over 5623240.52 frames. , ppl: 10.177316351485706], batch size: 70
+2022-12-10 22:09:44,350 INFO [train.py:421] (4/8) Epoch 3, batch 39000, loss[loss=2.363, over 1610.00 frames. , ppl: 10.624727300761432] tot_loss[loss=2.32, over 5624822.27 frames. , ppl: 10.177308865431488], batch size: 70
+2022-12-10 22:09:44,351 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:09:45,098 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.087220422135697
+2022-12-10 22:11:24,494 INFO [train.py:421] (4/8) Epoch 3, batch 39200, loss[loss=2.453, over 1750.00 frames. , ppl: 11.627854384918926] tot_loss[loss=2.32, over 5613065.41 frames. , ppl: 10.179253707043834], batch size: 70
+2022-12-10 22:13:02,104 INFO [train.py:421] (4/8) Epoch 3, batch 39400, loss[loss=2.4, over 840.00 frames. , ppl: 11.024916189840022] tot_loss[loss=2.32, over 5605714.55 frames. , ppl: 10.177372481195947], batch size: 70
+2022-12-10 22:14:41,112 INFO [train.py:421] (4/8) Epoch 3, batch 39600, loss[loss=2.438, over 2030.00 frames. , ppl: 11.450765601428403] tot_loss[loss=2.32, over 5626357.05 frames. , ppl: 10.173377939790436], batch size: 70
+2022-12-10 22:16:21,162 INFO [train.py:421] (4/8) Epoch 3, batch 39800, loss[loss=2.266, over 4410.00 frames. , ppl: 9.644870116378323] tot_loss[loss=2.32, over 5610382.32 frames. , ppl: 10.172904796047106], batch size: 70
+2022-12-10 22:17:59,551 INFO [train.py:421] (4/8) Epoch 3, batch 40000, loss[loss=2.463, over 1610.00 frames. , ppl: 11.73766054127557] tot_loss[loss=2.321, over 5571482.23 frames. , ppl: 10.187401099783306], batch size: 70
+2022-12-10 22:17:59,552 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:18:00,311 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.0863605000554
+2022-12-10 22:19:42,202 INFO [train.py:421] (4/8) Epoch 3, batch 40200, loss[loss=2.353, over 2730.00 frames. , ppl: 10.519259457616739] tot_loss[loss=2.321, over 5579717.11 frames. , ppl: 10.18475050004325], batch size: 70
+2022-12-10 22:21:21,717 INFO [train.py:421] (4/8) Epoch 3, batch 40400, loss[loss=2.403, over 1680.00 frames. , ppl: 11.0615848036394] tot_loss[loss=2.321, over 5596129.40 frames. , ppl: 10.18215755365516], batch size: 70
+2022-12-10 22:23:02,836 INFO [train.py:421] (4/8) Epoch 3, batch 40600, loss[loss=2.461, over 1260.00 frames. , ppl: 11.71709077393825] tot_loss[loss=2.321, over 5582784.99 frames. , ppl: 10.188318825643702], batch size: 70
+2022-12-10 22:24:42,985 INFO [train.py:421] (4/8) Epoch 3, batch 40800, loss[loss=2.414, over 2170.00 frames. , ppl: 11.17582910080337] tot_loss[loss=2.323, over 5525999.64 frames. , ppl: 10.203753220301845], batch size: 70
+2022-12-10 22:26:20,554 INFO [train.py:421] (4/8) Epoch 3, batch 41000, loss[loss=2.226, over 3780.00 frames. , ppl: 9.259150346275653] tot_loss[loss=2.322, over 5551595.61 frames. , ppl: 10.200386828250435], batch size: 70
+2022-12-10 22:26:20,554 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:26:21,301 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.09838814732227
+2022-12-10 22:28:02,987 INFO [train.py:421] (4/8) Epoch 3, batch 41200, loss[loss=2.397, over 3220.00 frames. , ppl: 10.992205919099455] tot_loss[loss=2.322, over 5597666.99 frames. , ppl: 10.192822148759172], batch size: 70
+2022-12-10 22:29:45,178 INFO [train.py:421] (4/8) Epoch 3, batch 41400, loss[loss=2.521, over 1120.00 frames. , ppl: 12.439285333425758] tot_loss[loss=2.321, over 5578322.99 frames. , ppl: 10.190624491847595], batch size: 70
+2022-12-10 22:31:24,818 INFO [train.py:421] (4/8) Epoch 3, batch 41600, loss[loss=2.192, over 7210.00 frames. , ppl: 8.956267407960517] tot_loss[loss=2.321, over 5579167.91 frames. , ppl: 10.188826037199316], batch size: 70
+2022-12-10 22:32:59,977 INFO [train.py:421] (4/8) Epoch 3, batch 41800, loss[loss=2.416, over 1610.00 frames. , ppl: 11.198742655783986] tot_loss[loss=2.323, over 5522019.74 frames. , ppl: 10.206750088945217], batch size: 70
+2022-12-10 22:34:40,203 INFO [train.py:421] (4/8) Epoch 3, batch 42000, loss[loss=2.264, over 3920.00 frames. , ppl: 9.61901749255198] tot_loss[loss=2.324, over 5483831.34 frames. , ppl: 10.215088586964468], batch size: 70
+2022-12-10 22:34:40,204 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:34:40,954 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.07229715383208
+2022-12-10 22:36:24,002 INFO [train.py:421] (4/8) Epoch 3, batch 42200, loss[loss=2.298, over 2520.00 frames. , ppl: 9.954572121185397] tot_loss[loss=2.323, over 5515314.44 frames. , ppl: 10.207486095642324], batch size: 70
+2022-12-10 22:38:04,419 INFO [train.py:421] (4/8) Epoch 3, batch 42400, loss[loss=2.195, over 4340.00 frames. , ppl: 8.978824723344173] tot_loss[loss=2.325, over 5476944.09 frames. , ppl: 10.22188017702066], batch size: 70
+2022-12-10 22:39:45,497 INFO [train.py:421] (4/8) Epoch 3, batch 42600, loss[loss=2.346, over 2590.00 frames. , ppl: 10.443995296375189] tot_loss[loss=2.324, over 5472632.69 frames. , ppl: 10.217574287997778], batch size: 70
+2022-12-10 22:41:24,909 INFO [train.py:421] (4/8) Epoch 3, batch 42800, loss[loss=2.21, over 9450.00 frames. , ppl: 9.113180834228544] tot_loss[loss=2.325, over 5453970.70 frames. , ppl: 10.222616373709931], batch size: 70
+2022-12-10 22:43:08,205 INFO [train.py:421] (4/8) Epoch 3, batch 43000, loss[loss=2.386, over 1470.00 frames. , ppl: 10.870049342309025] tot_loss[loss=2.324, over 5505101.80 frames. , ppl: 10.213040968924236], batch size: 70
+2022-12-10 22:43:08,206 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:43:08,978 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.073643413765264
+2022-12-10 22:44:46,762 INFO [train.py:421] (4/8) Epoch 3, batch 43200, loss[loss=2.285, over 4270.00 frames. , ppl: 9.827266557874403] tot_loss[loss=2.324, over 5498718.75 frames. , ppl: 10.21843442706569], batch size: 70
+2022-12-10 22:46:25,133 INFO [train.py:421] (4/8) Epoch 3, batch 43400, loss[loss=2.325, over 2240.00 frames. , ppl: 10.228686598019435] tot_loss[loss=2.324, over 5499125.28 frames. , ppl: 10.214327236518512], batch size: 70
+2022-12-10 22:48:04,841 INFO [train.py:421] (4/8) Epoch 3, batch 43600, loss[loss=2.364, over 2030.00 frames. , ppl: 10.635676268268792] tot_loss[loss=2.323, over 5492211.44 frames. , ppl: 10.210272474547999], batch size: 70
+2022-12-10 22:49:41,095 INFO [train.py:421] (4/8) Epoch 3, batch 43800, loss[loss=2.297, over 2380.00 frames. , ppl: 9.945665065224912] tot_loss[loss=2.324, over 5505711.46 frames. , ppl: 10.214342690765928], batch size: 70
+2022-12-10 22:51:19,561 INFO [train.py:421] (4/8) Epoch 3, batch 44000, loss[loss=2.334, over 4200.00 frames. , ppl: 10.31420086461132] tot_loss[loss=2.324, over 5482157.82 frames. , ppl: 10.220363850700416], batch size: 70
+2022-12-10 22:51:19,561 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:51:20,323 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.312, over 211138.00 frames. , ppl: 10.093368910010755
+2022-12-10 22:53:02,999 INFO [train.py:421] (4/8) Epoch 3, batch 44200, loss[loss=2.258, over 3710.00 frames. , ppl: 9.56768805439058] tot_loss[loss=2.323, over 5511208.80 frames. , ppl: 10.2045116195712], batch size: 70
+2022-12-10 22:54:43,986 INFO [train.py:421] (4/8) Epoch 3, batch 44400, loss[loss=2.361, over 4760.00 frames. , ppl: 10.605322808058887] tot_loss[loss=2.323, over 5509443.69 frames. , ppl: 10.208556359998042], batch size: 70
+2022-12-10 22:56:24,870 INFO [train.py:421] (4/8) Epoch 3, batch 44600, loss[loss=2.295, over 2100.00 frames. , ppl: 9.926886963840118] tot_loss[loss=2.321, over 5541538.94 frames. , ppl: 10.190123699541727], batch size: 70
+2022-12-10 22:58:04,170 INFO [train.py:421] (4/8) Epoch 3, batch 44800, loss[loss=2.245, over 6160.00 frames. , ppl: 9.438109851828939] tot_loss[loss=2.321, over 5542400.78 frames. , ppl: 10.189397689556866], batch size: 70
+2022-12-10 22:59:45,431 INFO [train.py:421] (4/8) Epoch 3, batch 45000, loss[loss=2.274, over 3920.00 frames. , ppl: 9.716724276611318] tot_loss[loss=2.319, over 5586157.03 frames. , ppl: 10.170108636019021], batch size: 70
+2022-12-10 22:59:45,431 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 22:59:46,181 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.313, over 211138.00 frames. , ppl: 10.10122686156436
+2022-12-10 23:01:30,009 INFO [train.py:421] (4/8) Epoch 3, batch 45200, loss[loss=2.288, over 3010.00 frames. , ppl: 9.851188824863398] tot_loss[loss=2.32, over 5569538.66 frames. , ppl: 10.175087167652677], batch size: 70
+2022-12-10 23:03:11,265 INFO [train.py:421] (4/8) Epoch 3, batch 45400, loss[loss=2.285, over 2310.00 frames. , ppl: 9.826739697199702] tot_loss[loss=2.319, over 5609728.05 frames. , ppl: 10.163620095955867], batch size: 70
+2022-12-10 23:04:51,283 INFO [train.py:421] (4/8) Epoch 3, batch 45600, loss[loss=2.368, over 2590.00 frames. , ppl: 10.677061994044797] tot_loss[loss=2.32, over 5590021.61 frames. , ppl: 10.175437554208116], batch size: 70
+2022-12-10 23:06:30,767 INFO [train.py:421] (4/8) Epoch 3, batch 45800, loss[loss=2.418, over 1890.00 frames. , ppl: 11.223419184617343] tot_loss[loss=2.319, over 5591316.30 frames. , ppl: 10.167311076113258], batch size: 70
+2022-12-10 23:08:09,340 INFO [train.py:421] (4/8) Epoch 3, batch 46000, loss[loss=2.385, over 2450.00 frames. , ppl: 10.858692593955421] tot_loss[loss=2.32, over 5564717.28 frames. , ppl: 10.1762211572139], batch size: 70
+2022-12-10 23:08:09,340 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:08:10,128 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.07448734063421
+2022-12-10 23:09:50,389 INFO [train.py:421] (4/8) Epoch 3, batch 46200, loss[loss=2.272, over 2660.00 frames. , ppl: 9.69613559865262] tot_loss[loss=2.32, over 5559954.18 frames. , ppl: 10.17379810707905], batch size: 70
+2022-12-10 23:11:28,888 INFO [train.py:421] (4/8) Epoch 3, batch 46400, loss[loss=2.671, over 770.00 frames. , ppl: 14.458457688962007] tot_loss[loss=2.32, over 5542415.17 frames. , ppl: 10.172486596194497], batch size: 70
+2022-12-10 23:13:11,866 INFO [train.py:421] (4/8) Epoch 3, batch 46600, loss[loss=2.252, over 6300.00 frames. , ppl: 9.51069756644202] tot_loss[loss=2.321, over 5488887.62 frames. , ppl: 10.18752323757203], batch size: 70
+2022-12-10 23:14:52,146 INFO [train.py:421] (4/8) Epoch 3, batch 46800, loss[loss=2.357, over 1820.00 frames. , ppl: 10.558751653526974] tot_loss[loss=2.32, over 5527429.58 frames. , ppl: 10.177511973688777], batch size: 70
+2022-12-10 23:16:28,906 INFO [train.py:421] (4/8) Epoch 3, batch 47000, loss[loss=2.393, over 1960.00 frames. , ppl: 10.945085309777124] tot_loss[loss=2.32, over 5571091.06 frames. , ppl: 10.171083299706181], batch size: 70
+2022-12-10 23:16:28,907 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:16:29,653 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.048629133994561
+2022-12-10 23:18:11,341 INFO [train.py:421] (4/8) Epoch 3, batch 47200, loss[loss=2.345, over 910.00 frames. , ppl: 10.436768835740832] tot_loss[loss=2.32, over 5570515.62 frames. , ppl: 10.171991256041375], batch size: 70
+2022-12-10 23:19:51,609 INFO [train.py:421] (4/8) Epoch 3, batch 47400, loss[loss=2.268, over 5460.00 frames. , ppl: 9.660329651036259] tot_loss[loss=2.32, over 5545851.21 frames. , ppl: 10.173516934327697], batch size: 70
+2022-12-10 23:21:30,718 INFO [train.py:421] (4/8) Epoch 3, batch 47600, loss[loss=2.335, over 2800.00 frames. , ppl: 10.329904863722133] tot_loss[loss=2.32, over 5561652.53 frames. , ppl: 10.172005005501465], batch size: 70
+2022-12-10 23:23:11,823 INFO [train.py:421] (4/8) Epoch 3, batch 47800, loss[loss=2.189, over 5740.00 frames. , ppl: 8.925938186647604] tot_loss[loss=2.32, over 5563079.88 frames. , ppl: 10.1807518656664], batch size: 70
+2022-12-10 23:24:53,395 INFO [train.py:421] (4/8) Epoch 3, batch 48000, loss[loss=2.243, over 3850.00 frames. , ppl: 9.418587838809575] tot_loss[loss=2.32, over 5577287.08 frames. , ppl: 10.178743192790554], batch size: 70
+2022-12-10 23:24:53,396 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:24:54,156 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.0626251790151
+2022-12-10 23:26:31,207 INFO [train.py:421] (4/8) Epoch 3, batch 48200, loss[loss=2.473, over 840.00 frames. , ppl: 11.862425499232147] tot_loss[loss=2.321, over 5517893.18 frames. , ppl: 10.189840106309056], batch size: 70
+2022-12-10 23:28:10,222 INFO [train.py:421] (4/8) Epoch 3, batch 48400, loss[loss=2.31, over 2590.00 frames. , ppl: 10.072249826616474] tot_loss[loss=2.322, over 5490843.88 frames. , ppl: 10.194060922369305], batch size: 70
+2022-12-10 23:29:50,970 INFO [train.py:421] (4/8) Epoch 3, batch 48600, loss[loss=2.352, over 1820.00 frames. , ppl: 10.50494760870591] tot_loss[loss=2.322, over 5487507.65 frames. , ppl: 10.192228509044513], batch size: 70
+2022-12-10 23:31:26,390 INFO [train.py:421] (4/8) Epoch 3, batch 48800, loss[loss=2.178, over 4970.00 frames. , ppl: 8.832896651470664] tot_loss[loss=2.322, over 5494370.46 frames. , ppl: 10.193610075707273], batch size: 70
+2022-12-10 23:33:09,328 INFO [train.py:421] (4/8) Epoch 3, batch 49000, loss[loss=2.35, over 2870.00 frames. , ppl: 10.481016315493518] tot_loss[loss=2.322, over 5498169.85 frames. , ppl: 10.195878570036943], batch size: 70
+2022-12-10 23:33:09,329 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:33:10,087 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.064253163288134
+2022-12-10 23:34:51,167 INFO [train.py:421] (4/8) Epoch 3, batch 49200, loss[loss=2.436, over 1120.00 frames. , ppl: 11.432495541340344] tot_loss[loss=2.321, over 5543583.26 frames. , ppl: 10.185603245849146], batch size: 70
+2022-12-10 23:36:30,854 INFO [train.py:421] (4/8) Epoch 3, batch 49400, loss[loss=2.351, over 2030.00 frames. , ppl: 10.499975882785318] tot_loss[loss=2.321, over 5540396.16 frames. , ppl: 10.187086851332662], batch size: 70
+2022-12-10 23:38:10,928 INFO [train.py:421] (4/8) Epoch 3, batch 49600, loss[loss=2.531, over 840.00 frames. , ppl: 12.570311133541658] tot_loss[loss=2.32, over 5551287.45 frames. , ppl: 10.17739798760546], batch size: 70
+2022-12-10 23:39:52,726 INFO [train.py:421] (4/8) Epoch 3, batch 49800, loss[loss=2.23, over 4410.00 frames. , ppl: 9.3005040997278] tot_loss[loss=2.319, over 5585493.33 frames. , ppl: 10.169328683129585], batch size: 70
+2022-12-10 23:41:29,023 INFO [train.py:421] (4/8) Epoch 3, batch 50000, loss[loss=2.379, over 2450.00 frames. , ppl: 10.790332021972633] tot_loss[loss=2.32, over 5558433.34 frames. , ppl: 10.177307037955469], batch size: 70
+2022-12-10 23:41:29,023 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:41:29,779 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.31, over 211138.00 frames. , ppl: 10.074415768120824
+2022-12-10 23:43:13,634 INFO [train.py:421] (4/8) Epoch 3, batch 50200, loss[loss=2.331, over 3080.00 frames. , ppl: 10.287488985484197] tot_loss[loss=2.32, over 5591680.33 frames. , ppl: 10.175079284179697], batch size: 70
+2022-12-10 23:44:54,564 INFO [train.py:421] (4/8) Epoch 3, batch 50400, loss[loss=2.353, over 2660.00 frames. , ppl: 10.518620081286882] tot_loss[loss=2.32, over 5573388.64 frames. , ppl: 10.174474615601794], batch size: 70
+2022-12-10 23:46:36,439 INFO [train.py:421] (4/8) Epoch 3, batch 50600, loss[loss=2.495, over 1540.00 frames. , ppl: 12.120310011824316] tot_loss[loss=2.319, over 5595071.05 frames. , ppl: 10.16856756552914], batch size: 70
+2022-12-10 23:48:17,653 INFO [train.py:421] (4/8) Epoch 3, batch 50800, loss[loss=2.319, over 2380.00 frames. , ppl: 10.169941582998032] tot_loss[loss=2.319, over 5605518.01 frames. , ppl: 10.16476695147555], batch size: 70
+2022-12-10 23:49:58,956 INFO [train.py:421] (4/8) Epoch 3, batch 51000, loss[loss=2.472, over 1260.00 frames. , ppl: 11.847816093477611] tot_loss[loss=2.32, over 5587856.38 frames. , ppl: 10.170796584358676], batch size: 70
+2022-12-10 23:49:58,956 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:49:59,716 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.069029904707316
+2022-12-10 23:51:40,243 INFO [train.py:421] (4/8) Epoch 3, batch 51200, loss[loss=2.265, over 5950.00 frames. , ppl: 9.627243089292469] tot_loss[loss=2.318, over 5627953.50 frames. , ppl: 10.155000011440618], batch size: 70
+2022-12-10 23:53:19,787 INFO [train.py:421] (4/8) Epoch 3, batch 51400, loss[loss=2.312, over 5320.00 frames. , ppl: 10.093577972944756] tot_loss[loss=2.319, over 5609031.74 frames. , ppl: 10.164497742024015], batch size: 70
+2022-12-10 23:55:01,898 INFO [train.py:421] (4/8) Epoch 3, batch 51600, loss[loss=2.416, over 1400.00 frames. , ppl: 11.206470676489623] tot_loss[loss=2.318, over 5615565.36 frames. , ppl: 10.153254907505818], batch size: 70
+2022-12-10 23:56:41,059 INFO [train.py:421] (4/8) Epoch 3, batch 51800, loss[loss=2.427, over 1890.00 frames. , ppl: 11.31991465459032] tot_loss[loss=2.318, over 5615357.81 frames. , ppl: 10.152245144838961], batch size: 70
+2022-12-10 23:58:20,734 INFO [train.py:421] (4/8) Epoch 3, batch 52000, loss[loss=2.4, over 1120.00 frames. , ppl: 11.020732941082933] tot_loss[loss=2.318, over 5593467.28 frames. , ppl: 10.156473520238452], batch size: 70
+2022-12-10 23:58:20,734 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-10 23:58:21,496 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.064753676155952
+2022-12-11 00:00:01,560 INFO [train.py:421] (4/8) Epoch 3, batch 52200, loss[loss=2.425, over 1820.00 frames. , ppl: 11.304782849086338] tot_loss[loss=2.319, over 5586461.96 frames. , ppl: 10.161526477110472], batch size: 70
+2022-12-11 00:01:45,812 INFO [train.py:421] (4/8) Epoch 3, batch 52400, loss[loss=2.454, over 2450.00 frames. , ppl: 11.630359315291017] tot_loss[loss=2.318, over 5614243.34 frames. , ppl: 10.158031204902754], batch size: 70
+2022-12-11 00:03:24,670 INFO [train.py:421] (4/8) Epoch 3, batch 52600, loss[loss=2.525, over 1470.00 frames. , ppl: 12.491573647971517] tot_loss[loss=2.318, over 5624456.14 frames. , ppl: 10.151494819711347], batch size: 70
+2022-12-11 00:05:04,004 INFO [train.py:421] (4/8) Epoch 3, batch 52800, loss[loss=2.298, over 3290.00 frames. , ppl: 9.949902472215797] tot_loss[loss=2.317, over 5653199.33 frames. , ppl: 10.145378239890123], batch size: 70
+2022-12-11 00:06:42,494 INFO [train.py:421] (4/8) Epoch 3, batch 53000, loss[loss=2.387, over 1750.00 frames. , ppl: 10.883122226885932] tot_loss[loss=2.318, over 5622293.38 frames. , ppl: 10.150464409588551], batch size: 70
+2022-12-11 00:06:42,494 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:06:43,239 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.311, over 211138.00 frames. , ppl: 10.088516415422001
+2022-12-11 00:08:25,723 INFO [train.py:421] (4/8) Epoch 3, batch 53200, loss[loss=2.173, over 5180.00 frames. , ppl: 8.781251364617848] tot_loss[loss=2.317, over 5649178.88 frames. , ppl: 10.141026748751118], batch size: 70
+2022-12-11 00:10:06,550 INFO [train.py:421] (4/8) Epoch 3, batch 53400, loss[loss=2.658, over 770.00 frames. , ppl: 14.266958962637878] tot_loss[loss=2.315, over 5667511.42 frames. , ppl: 10.12974841944654], batch size: 70
+2022-12-11 00:11:48,855 INFO [train.py:421] (4/8) Epoch 3, batch 53600, loss[loss=2.61, over 1050.00 frames. , ppl: 13.599553616213125] tot_loss[loss=2.316, over 5647622.79 frames. , ppl: 10.136508302037617], batch size: 70
+2022-12-11 00:13:30,371 INFO [train.py:421] (4/8) Epoch 3, batch 53800, loss[loss=2.301, over 2800.00 frames. , ppl: 9.98205093998622] tot_loss[loss=2.317, over 5602822.95 frames. , ppl: 10.147882260195965], batch size: 70
+2022-12-11 00:15:11,539 INFO [train.py:421] (4/8) Epoch 3, batch 54000, loss[loss=2.611, over 1120.00 frames. , ppl: 13.617032073435611] tot_loss[loss=2.319, over 5550651.45 frames. , ppl: 10.163940076438996], batch size: 70
+2022-12-11 00:15:11,539 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:15:12,298 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.032199299773053
+2022-12-11 00:16:54,791 INFO [train.py:421] (4/8) Epoch 3, batch 54200, loss[loss=2.173, over 12110.00 frames. , ppl: 8.78503138526725] tot_loss[loss=2.318, over 5575284.67 frames. , ppl: 10.158354964384966], batch size: 70
+2022-12-11 00:18:34,608 INFO [train.py:421] (4/8) Epoch 3, batch 54400, loss[loss=2.809, over 770.00 frames. , ppl: 16.6008034663615] tot_loss[loss=2.32, over 5517128.58 frames. , ppl: 10.174146282234542], batch size: 70
+2022-12-11 00:20:14,992 INFO [train.py:421] (4/8) Epoch 3, batch 54600, loss[loss=2.303, over 5110.00 frames. , ppl: 10.00413478888473] tot_loss[loss=2.318, over 5569122.75 frames. , ppl: 10.158605512723936], batch size: 70
+2022-12-11 00:21:53,922 INFO [train.py:421] (4/8) Epoch 3, batch 54800, loss[loss=2.701, over 700.00 frames. , ppl: 14.89226738468065] tot_loss[loss=2.319, over 5554518.30 frames. , ppl: 10.163466851646985], batch size: 70
+2022-12-11 00:23:33,753 INFO [train.py:421] (4/8) Epoch 3, batch 55000, loss[loss=2.506, over 910.00 frames. , ppl: 12.251754606795323] tot_loss[loss=2.319, over 5528973.41 frames. , ppl: 10.168074260040996], batch size: 70
+2022-12-11 00:23:33,754 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:23:34,513 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.042009979542168
+2022-12-11 00:25:14,234 INFO [train.py:421] (4/8) Epoch 3, batch 55200, loss[loss=2.267, over 5950.00 frames. , ppl: 9.653213844011301] tot_loss[loss=2.319, over 5524804.77 frames. , ppl: 10.169256255065008], batch size: 70
+2022-12-11 00:26:51,652 INFO [train.py:421] (4/8) Epoch 3, batch 55400, loss[loss=2.208, over 5390.00 frames. , ppl: 9.097986338989026] tot_loss[loss=2.32, over 5528740.05 frames. , ppl: 10.172796805085872], batch size: 70
+2022-12-11 00:28:28,803 INFO [train.py:421] (4/8) Epoch 3, batch 55600, loss[loss=2.407, over 2100.00 frames. , ppl: 11.10021803306727] tot_loss[loss=2.32, over 5521838.21 frames. , ppl: 10.175790893329337], batch size: 70
+2022-12-11 00:30:05,107 INFO [train.py:421] (4/8) Epoch 3, batch 55800, loss[loss=2.347, over 2240.00 frames. , ppl: 10.451254051551619] tot_loss[loss=2.319, over 5534450.43 frames. , ppl: 10.167628849305142], batch size: 70
+2022-12-11 00:31:49,154 INFO [train.py:421] (4/8) Epoch 3, batch 56000, loss[loss=2.767, over 700.00 frames. , ppl: 15.90654140710941] tot_loss[loss=2.319, over 5551502.04 frames. , ppl: 10.162317859167675], batch size: 70
+2022-12-11 00:31:49,155 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:31:49,915 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.029381468540643
+2022-12-11 00:33:27,283 INFO [train.py:421] (4/8) Epoch 3, batch 56200, loss[loss=2.355, over 1820.00 frames. , ppl: 10.533770729622553] tot_loss[loss=2.32, over 5495544.71 frames. , ppl: 10.175938715070638], batch size: 70
+2022-12-11 00:35:04,950 INFO [train.py:421] (4/8) Epoch 3, batch 56400, loss[loss=2.762, over 910.00 frames. , ppl: 15.838159666144993] tot_loss[loss=2.321, over 5473371.36 frames. , ppl: 10.180972367582186], batch size: 70
+2022-12-11 00:36:45,126 INFO [train.py:421] (4/8) Epoch 3, batch 56600, loss[loss=2.741, over 910.00 frames. , ppl: 15.499478242851929] tot_loss[loss=2.32, over 5499452.61 frames. , ppl: 10.175882127914965], batch size: 70
+2022-12-11 00:38:27,987 INFO [train.py:421] (4/8) Epoch 3, batch 56800, loss[loss=2.375, over 1960.00 frames. , ppl: 10.749505394510042] tot_loss[loss=2.32, over 5533604.55 frames. , ppl: 10.170979813561367], batch size: 70
+2022-12-11 00:40:12,339 INFO [train.py:421] (4/8) Epoch 3, batch 57000, loss[loss=2.253, over 5460.00 frames. , ppl: 9.518096994843626] tot_loss[loss=2.319, over 5572599.12 frames. , ppl: 10.162683145664289], batch size: 70
+2022-12-11 00:40:12,340 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:40:13,085 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.04296571105421
+2022-12-11 00:41:53,208 INFO [train.py:421] (4/8) Epoch 3, batch 57200, loss[loss=2.4, over 1190.00 frames. , ppl: 11.019612806068373] tot_loss[loss=2.319, over 5559804.49 frames. , ppl: 10.162789477400782], batch size: 70
+2022-12-11 00:43:33,444 INFO [train.py:421] (4/8) Epoch 3, batch 57400, loss[loss=2.444, over 1330.00 frames. , ppl: 11.513720966229762] tot_loss[loss=2.319, over 5530656.18 frames. , ppl: 10.161484305622949], batch size: 70
+2022-12-11 00:45:12,037 INFO [train.py:421] (4/8) Epoch 3, batch 57600, loss[loss=2.335, over 3360.00 frames. , ppl: 10.325580625795252] tot_loss[loss=2.319, over 5518459.32 frames. , ppl: 10.16822684452196], batch size: 70
+2022-12-11 00:46:53,806 INFO [train.py:421] (4/8) Epoch 3, batch 57800, loss[loss=2.633, over 980.00 frames. , ppl: 13.911541644259794] tot_loss[loss=2.319, over 5538747.07 frames. , ppl: 10.1688551376028], batch size: 70
+2022-12-11 00:48:34,004 INFO [train.py:421] (4/8) Epoch 3, batch 58000, loss[loss=2.258, over 6650.00 frames. , ppl: 9.559805721521407] tot_loss[loss=2.321, over 5511819.40 frames. , ppl: 10.182249004329766], batch size: 70
+2022-12-11 00:48:34,005 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:48:34,750 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.037739283472558
+2022-12-11 00:50:13,898 INFO [train.py:421] (4/8) Epoch 3, batch 58200, loss[loss=2.528, over 1260.00 frames. , ppl: 12.528470629281538] tot_loss[loss=2.321, over 5507092.74 frames. , ppl: 10.18575836676764], batch size: 70
+2022-12-11 00:51:51,052 INFO [train.py:421] (4/8) Epoch 3, batch 58400, loss[loss=2.318, over 2100.00 frames. , ppl: 10.159470686481576] tot_loss[loss=2.321, over 5522010.81 frames. , ppl: 10.183818228771468], batch size: 70
+2022-12-11 00:53:28,866 INFO [train.py:421] (4/8) Epoch 3, batch 58600, loss[loss=2.348, over 1610.00 frames. , ppl: 10.459828523700944] tot_loss[loss=2.321, over 5478512.90 frames. , ppl: 10.19049273235436], batch size: 70
+2022-12-11 00:55:08,253 INFO [train.py:421] (4/8) Epoch 3, batch 58800, loss[loss=2.228, over 6650.00 frames. , ppl: 9.28483233464041] tot_loss[loss=2.321, over 5508437.45 frames. , ppl: 10.18960032470868], batch size: 70
+2022-12-11 00:56:50,214 INFO [train.py:421] (4/8) Epoch 3, batch 59000, loss[loss=3.007, over 630.00 frames. , ppl: 20.226688812396212] tot_loss[loss=2.321, over 5509277.51 frames. , ppl: 10.188936379975472], batch size: 70
+2022-12-11 00:56:50,215 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 00:56:50,962 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.309, over 211138.00 frames. , ppl: 10.063399667509817
+2022-12-11 00:58:29,069 INFO [train.py:421] (4/8) Epoch 3, batch 59200, loss[loss=2.285, over 2660.00 frames. , ppl: 9.820790905734352] tot_loss[loss=2.321, over 5500818.21 frames. , ppl: 10.190875222882621], batch size: 70
+2022-12-11 01:00:05,132 INFO [train.py:421] (4/8) Epoch 3, batch 59400, loss[loss=2.362, over 2660.00 frames. , ppl: 10.61731188438462] tot_loss[loss=2.321, over 5512833.46 frames. , ppl: 10.19028357974126], batch size: 70
+2022-12-11 01:01:44,071 INFO [train.py:421] (4/8) Epoch 3, batch 59600, loss[loss=2.263, over 5390.00 frames. , ppl: 9.614738618208778] tot_loss[loss=2.321, over 5543355.38 frames. , ppl: 10.184207924775327], batch size: 70
+2022-12-11 01:03:24,733 INFO [train.py:421] (4/8) Epoch 3, batch 59800, loss[loss=2.45, over 1190.00 frames. , ppl: 11.58921452731768] tot_loss[loss=2.322, over 5497476.54 frames. , ppl: 10.193089888600795], batch size: 70
+2022-12-11 01:05:04,163 INFO [train.py:421] (4/8) Epoch 3, batch 60000, loss[loss=2.47, over 910.00 frames. , ppl: 11.825621617000559] tot_loss[loss=2.321, over 5509436.53 frames. , ppl: 10.183606226417643], batch size: 70
+2022-12-11 01:05:04,163 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:05:04,908 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.045027606490608
+2022-12-11 01:06:48,968 INFO [train.py:421] (4/8) Epoch 3, batch 60200, loss[loss=2.387, over 3220.00 frames. , ppl: 10.884956743144752] tot_loss[loss=2.321, over 5503551.96 frames. , ppl: 10.182614781474815], batch size: 70
+2022-12-11 01:08:28,495 INFO [train.py:421] (4/8) Epoch 3, batch 60400, loss[loss=2.388, over 1750.00 frames. , ppl: 10.893187085649211] tot_loss[loss=2.322, over 5487073.27 frames. , ppl: 10.191605292607521], batch size: 70
+2022-12-11 01:10:08,958 INFO [train.py:421] (4/8) Epoch 3, batch 60600, loss[loss=2.334, over 3360.00 frames. , ppl: 10.321646980446669] tot_loss[loss=2.322, over 5463392.62 frames. , ppl: 10.195232907450094], batch size: 70
+2022-12-11 01:11:50,388 INFO [train.py:421] (4/8) Epoch 3, batch 60800, loss[loss=2.286, over 3780.00 frames. , ppl: 9.838820888660749] tot_loss[loss=2.322, over 5456712.81 frames. , ppl: 10.199750876563867], batch size: 70
+2022-12-11 01:13:29,073 INFO [train.py:421] (4/8) Epoch 3, batch 61000, loss[loss=2.557, over 910.00 frames. , ppl: 12.894647828992083] tot_loss[loss=2.323, over 5439112.91 frames. , ppl: 10.202089797278623], batch size: 70
+2022-12-11 01:13:29,073 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:13:29,818 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.0369311167839
+2022-12-11 01:15:11,780 INFO [train.py:421] (4/8) Epoch 3, batch 61200, loss[loss=2.255, over 6790.00 frames. , ppl: 9.539097236759556] tot_loss[loss=2.322, over 5441287.21 frames. , ppl: 10.199234847990708], batch size: 70
+2022-12-11 01:16:52,649 INFO [train.py:421] (4/8) Epoch 3, batch 61400, loss[loss=2.531, over 910.00 frames. , ppl: 12.560497221299476] tot_loss[loss=2.322, over 5423779.23 frames. , ppl: 10.193700118793645], batch size: 70
+2022-12-11 01:18:30,899 INFO [train.py:421] (4/8) Epoch 3, batch 61600, loss[loss=2.241, over 3570.00 frames. , ppl: 9.399918753965233] tot_loss[loss=2.321, over 5447803.36 frames. , ppl: 10.188378015218149], batch size: 70
+2022-12-11 01:20:08,761 INFO [train.py:421] (4/8) Epoch 3, batch 61800, loss[loss=2.518, over 1120.00 frames. , ppl: 12.401249052946609] tot_loss[loss=2.322, over 5428177.32 frames. , ppl: 10.19273329156953], batch size: 70
+2022-12-11 01:21:46,207 INFO [train.py:421] (4/8) Epoch 3, batch 62000, loss[loss=2.408, over 1330.00 frames. , ppl: 11.10854327420354] tot_loss[loss=2.321, over 5434815.76 frames. , ppl: 10.186312813018132], batch size: 70
+2022-12-11 01:21:46,208 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:21:46,967 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.307, over 211138.00 frames. , ppl: 10.04664679581944
+2022-12-11 01:23:28,145 INFO [train.py:421] (4/8) Epoch 3, batch 62200, loss[loss=2.259, over 3640.00 frames. , ppl: 9.572303159197416] tot_loss[loss=2.321, over 5435286.23 frames. , ppl: 10.182428381756724], batch size: 70
+2022-12-11 01:25:03,203 INFO [train.py:421] (4/8) Epoch 3, batch 62400, loss[loss=2.322, over 1750.00 frames. , ppl: 10.195105830478749] tot_loss[loss=2.322, over 5413260.30 frames. , ppl: 10.194410257970095], batch size: 70
+2022-12-11 01:26:46,722 INFO [train.py:421] (4/8) Epoch 3, batch 62600, loss[loss=2.498, over 1050.00 frames. , ppl: 12.159277369437014] tot_loss[loss=2.322, over 5416516.38 frames. , ppl: 10.194185866353674], batch size: 70
+2022-12-11 01:28:27,836 INFO [train.py:421] (4/8) Epoch 3, batch 62800, loss[loss=2.336, over 2100.00 frames. , ppl: 10.344578009198221] tot_loss[loss=2.324, over 5352708.52 frames. , ppl: 10.212003607309471], batch size: 70
+2022-12-11 01:30:02,598 INFO [train.py:421] (4/8) Epoch 3, batch 63000, loss[loss=2.238, over 5670.00 frames. , ppl: 9.375887409327273] tot_loss[loss=2.324, over 5387478.88 frames. , ppl: 10.211636876320405], batch size: 70
+2022-12-11 01:30:02,599 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:30:03,344 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.036726114257576
+2022-12-11 01:31:44,152 INFO [train.py:421] (4/8) Epoch 3, batch 63200, loss[loss=2.474, over 1750.00 frames. , ppl: 11.866274902628057] tot_loss[loss=2.323, over 5417025.92 frames. , ppl: 10.201459110366422], batch size: 70
+2022-12-11 01:33:23,261 INFO [train.py:421] (4/8) Epoch 3, batch 63400, loss[loss=2.444, over 1890.00 frames. , ppl: 11.51409271498901] tot_loss[loss=2.324, over 5407464.92 frames. , ppl: 10.212431966269282], batch size: 70
+2022-12-11 01:35:03,957 INFO [train.py:421] (4/8) Epoch 3, batch 63600, loss[loss=2.244, over 6020.00 frames. , ppl: 9.429231173163338] tot_loss[loss=2.323, over 5431996.67 frames. , ppl: 10.201714529440265], batch size: 70
+2022-12-11 01:36:42,933 INFO [train.py:421] (4/8) Epoch 3, batch 63800, loss[loss=2.203, over 5180.00 frames. , ppl: 9.052130967956598] tot_loss[loss=2.321, over 5464105.11 frames. , ppl: 10.18275522500688], batch size: 70
+2022-12-11 01:38:28,651 INFO [train.py:421] (4/8) Epoch 3, batch 64000, loss[loss=4.925, over 280.00 frames. , ppl: 137.67073428563552] tot_loss[loss=2.32, over 5485228.10 frames. , ppl: 10.17801870850001], batch size: 70
+2022-12-11 01:38:28,652 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:38:29,412 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.306, over 211138.00 frames. , ppl: 10.036485464460732
+2022-12-11 01:40:09,937 INFO [train.py:421] (4/8) Epoch 3, batch 64200, loss[loss=2.449, over 1120.00 frames. , ppl: 11.573056475391214] tot_loss[loss=2.319, over 5498654.18 frames. , ppl: 10.170560614653724], batch size: 70
+2022-12-11 01:41:46,095 INFO [train.py:421] (4/8) Epoch 3, batch 64400, loss[loss=2.272, over 4690.00 frames. , ppl: 9.69649657264346] tot_loss[loss=2.319, over 5503630.77 frames. , ppl: 10.16854951087362], batch size: 70
+2022-12-11 01:43:25,726 INFO [train.py:421] (4/8) Epoch 3, batch 64600, loss[loss=2.374, over 1610.00 frames. , ppl: 10.742534157829652] tot_loss[loss=2.32, over 5488079.89 frames. , ppl: 10.176373763301742], batch size: 70
+2022-12-11 01:45:05,816 INFO [train.py:421] (4/8) Epoch 3, batch 64800, loss[loss=2.321, over 2800.00 frames. , ppl: 10.183911303149973] tot_loss[loss=2.32, over 5519319.15 frames. , ppl: 10.173633648257999], batch size: 70
+2022-12-11 01:46:48,564 INFO [train.py:421] (4/8) Epoch 3, batch 65000, loss[loss=2.662, over 910.00 frames. , ppl: 14.32507815955047] tot_loss[loss=2.319, over 5509657.88 frames. , ppl: 10.166507962642774], batch size: 70
+2022-12-11 01:46:48,565 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:46:49,313 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.022263217124767
+2022-12-11 01:48:29,575 INFO [train.py:421] (4/8) Epoch 3, batch 65200, loss[loss=2.386, over 1750.00 frames. , ppl: 10.866967447849344] tot_loss[loss=2.32, over 5519005.70 frames. , ppl: 10.170702843146067], batch size: 70
+2022-12-11 01:50:10,536 INFO [train.py:421] (4/8) Epoch 3, batch 65400, loss[loss=2.382, over 2170.00 frames. , ppl: 10.830707889848636] tot_loss[loss=2.319, over 5545047.41 frames. , ppl: 10.16152484050955], batch size: 70
+2022-12-11 01:51:49,289 INFO [train.py:421] (4/8) Epoch 3, batch 65600, loss[loss=2.2, over 6790.00 frames. , ppl: 9.023305479561618] tot_loss[loss=2.319, over 5538261.06 frames. , ppl: 10.16058763035119], batch size: 70
+2022-12-11 01:53:27,735 INFO [train.py:421] (4/8) Epoch 3, batch 65800, loss[loss=2.628, over 910.00 frames. , ppl: 13.84801872882006] tot_loss[loss=2.319, over 5504955.91 frames. , ppl: 10.169614984949968], batch size: 70
+2022-12-11 01:55:11,465 INFO [train.py:421] (4/8) Epoch 3, batch 66000, loss[loss=2.275, over 3850.00 frames. , ppl: 9.729503284207473] tot_loss[loss=2.319, over 5515929.05 frames. , ppl: 10.164504999934499], batch size: 70
+2022-12-11 01:55:11,465 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 01:55:12,223 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.026417513222102
+2022-12-11 01:56:53,388 INFO [train.py:421] (4/8) Epoch 3, batch 66200, loss[loss=2.271, over 3640.00 frames. , ppl: 9.690466606268306] tot_loss[loss=2.319, over 5506971.81 frames. , ppl: 10.169418707295689], batch size: 70
+2022-12-11 01:58:33,167 INFO [train.py:421] (4/8) Epoch 3, batch 66400, loss[loss=3.361, over 490.00 frames. , ppl: 28.82789718711574] tot_loss[loss=2.318, over 5542316.66 frames. , ppl: 10.156769221814297], batch size: 70
+2022-12-11 02:00:07,972 INFO [train.py:421] (4/8) Epoch 3, batch 66600, loss[loss=2.265, over 2940.00 frames. , ppl: 9.62681188169607] tot_loss[loss=2.318, over 5528396.98 frames. , ppl: 10.160057658926258], batch size: 70
+2022-12-11 02:01:49,359 INFO [train.py:421] (4/8) Epoch 3, batch 66800, loss[loss=3.262, over 490.00 frames. , ppl: 26.095505434097763] tot_loss[loss=2.318, over 5549680.91 frames. , ppl: 10.153191891717706], batch size: 70
+2022-12-11 02:03:29,974 INFO [train.py:421] (4/8) Epoch 3, batch 67000, loss[loss=2.238, over 6090.00 frames. , ppl: 9.377690578857907] tot_loss[loss=2.317, over 5587202.36 frames. , ppl: 10.14176429190508], batch size: 70
+2022-12-11 02:03:29,974 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 02:03:30,724 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.023177014556284
+2022-12-11 02:05:08,154 INFO [train.py:421] (4/8) Epoch 3, batch 67200, loss[loss=2.454, over 1120.00 frames. , ppl: 11.640071728911693] tot_loss[loss=2.316, over 5594266.36 frames. , ppl: 10.137969036089054], batch size: 70
+2022-12-11 02:06:47,974 INFO [train.py:421] (4/8) Epoch 3, batch 67400, loss[loss=2.427, over 1890.00 frames. , ppl: 11.322125791345588] tot_loss[loss=2.316, over 5580036.94 frames. , ppl: 10.137448556983427], batch size: 70
+2022-12-11 02:08:29,692 INFO [train.py:421] (4/8) Epoch 3, batch 67600, loss[loss=2.24, over 3780.00 frames. , ppl: 9.389134405646221] tot_loss[loss=2.315, over 5624352.14 frames. , ppl: 10.125781240659737], batch size: 70
+2022-12-11 02:10:11,841 INFO [train.py:421] (4/8) Epoch 3, batch 67800, loss[loss=2.478, over 1260.00 frames. , ppl: 11.915588874003985] tot_loss[loss=2.315, over 5653524.27 frames. , ppl: 10.12278122714892], batch size: 70
+2022-12-11 02:11:53,378 INFO [train.py:421] (4/8) Epoch 3, batch 68000, loss[loss=2.311, over 1960.00 frames. , ppl: 10.088054900595997] tot_loss[loss=2.315, over 5625004.31 frames. , ppl: 10.126057785707989], batch size: 70
+2022-12-11 02:11:53,378 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 02:11:54,123 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.025138400067002
+2022-12-11 02:13:31,881 INFO [train.py:421] (4/8) Epoch 3, batch 68200, loss[loss=2.416, over 840.00 frames. , ppl: 11.197236401647261] tot_loss[loss=2.315, over 5599865.00 frames. , ppl: 10.127893204511752], batch size: 70
+2022-12-11 02:15:11,929 INFO [train.py:421] (4/8) Epoch 3, batch 68400, loss[loss=2.322, over 4690.00 frames. , ppl: 10.194992364874098] tot_loss[loss=2.315, over 5616914.60 frames. , ppl: 10.124612753799784], batch size: 70
+2022-12-11 02:16:58,787 INFO [train.py:421] (4/8) Epoch 3, batch 68600, loss[loss=2.333, over 2870.00 frames. , ppl: 10.305572107177802] tot_loss[loss=2.314, over 5657133.49 frames. , ppl: 10.115485735324805], batch size: 70
+2022-12-11 02:18:39,312 INFO [train.py:421] (4/8) Epoch 3, batch 68800, loss[loss=2.744, over 630.00 frames. , ppl: 15.547652462989474] tot_loss[loss=2.315, over 5645459.97 frames. , ppl: 10.121425613804284], batch size: 70
+2022-12-11 02:20:20,721 INFO [train.py:421] (4/8) Epoch 3, batch 69000, loss[loss=2.223, over 3360.00 frames. , ppl: 9.235513128808247] tot_loss[loss=2.315, over 5639388.68 frames. , ppl: 10.128927166117759], batch size: 70
+2022-12-11 02:20:20,722 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 02:20:21,480 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.019322126504301
+2022-12-11 02:22:00,180 INFO [train.py:421] (4/8) Epoch 3, batch 69200, loss[loss=2.363, over 2380.00 frames. , ppl: 10.619509736509675] tot_loss[loss=2.316, over 5633518.75 frames. , ppl: 10.131083067155432], batch size: 70
+2022-12-11 02:23:42,015 INFO [train.py:421] (4/8) Epoch 3, batch 69400, loss[loss=2.195, over 10640.00 frames. , ppl: 8.975560354099002] tot_loss[loss=2.316, over 5593857.48 frames. , ppl: 10.13269742055951], batch size: 70
+2022-12-11 02:25:20,934 INFO [train.py:421] (4/8) Epoch 3, batch 69600, loss[loss=2.313, over 2800.00 frames. , ppl: 10.109550522923461] tot_loss[loss=2.316, over 5569032.98 frames. , ppl: 10.132039592780412], batch size: 70
+2022-12-11 02:27:00,942 INFO [train.py:421] (4/8) Epoch 3, batch 69800, loss[loss=2.229, over 3710.00 frames. , ppl: 9.288829434259005] tot_loss[loss=2.317, over 5518120.88 frames. , ppl: 10.14846844184939], batch size: 70
+2022-12-11 02:28:40,242 INFO [train.py:421] (4/8) Epoch 3, batch 70000, loss[loss=2.297, over 4970.00 frames. , ppl: 9.942354562132621] tot_loss[loss=2.317, over 5495642.80 frames. , ppl: 10.150074057807949], batch size: 70
+2022-12-11 02:28:40,242 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 02:28:40,991 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.305, over 211138.00 frames. , ppl: 10.023684386112267
+2022-12-11 02:30:18,975 INFO [train.py:421] (4/8) Epoch 3, batch 70200, loss[loss=2.448, over 1960.00 frames. , ppl: 11.5671213025095] tot_loss[loss=2.317, over 5487968.69 frames. , ppl: 10.148770078976233], batch size: 70
+2022-12-11 02:31:59,181 INFO [train.py:421] (4/8) Epoch 3, batch 70400, loss[loss=2.831, over 630.00 frames. , ppl: 16.9707032898819] tot_loss[loss=2.317, over 5499577.03 frames. , ppl: 10.145010044085236], batch size: 70
+2022-12-11 02:33:37,911 INFO [train.py:421] (4/8) Epoch 3, batch 70600, loss[loss=2.265, over 4130.00 frames. , ppl: 9.633117931283484] tot_loss[loss=2.318, over 5499203.00 frames. , ppl: 10.151268276455164], batch size: 70
+2022-12-11 02:35:16,211 INFO [train.py:421] (4/8) Epoch 3, batch 70800, loss[loss=2.225, over 8680.00 frames. , ppl: 9.250618185761015] tot_loss[loss=2.317, over 5515493.59 frames. , ppl: 10.142622607310834], batch size: 70
+2022-12-11 02:36:54,122 INFO [train.py:421] (4/8) Epoch 3, batch 71000, loss[loss=2.327, over 2590.00 frames. , ppl: 10.244927426578364] tot_loss[loss=2.317, over 5485454.80 frames. , ppl: 10.14397992291303], batch size: 70
+2022-12-11 02:36:54,122 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 02:36:54,881 INFO [train.py:452] (4/8) Epoch 3, validation: loss=2.303, over 211138.00 frames. , ppl: 10.006615510581163
+2022-12-11 02:38:35,001 INFO [train.py:421] (4/8) Epoch 3, batch 71200, loss[loss=2.211, over 6020.00 frames. , ppl: 9.120441035893897] tot_loss[loss=2.317, over 5479248.71 frames. , ppl: 10.144785663347918], batch size: 70
+2022-12-11 02:40:17,314 INFO [train.py:421] (4/8) Epoch 3, batch 71400, loss[loss=2.277, over 5460.00 frames. , ppl: 9.750567676296278] tot_loss[loss=2.318, over 5491007.93 frames. , ppl: 10.150747765893076], batch size: 70
+2022-12-11 02:41:59,128 INFO [train.py:421] (4/8) Epoch 3, batch 71600, loss[loss=2.545, over 840.00 frames. , ppl: 12.74124748107548] tot_loss[loss=2.317, over 5517868.82 frames. , ppl: 10.14831729050412], batch size: 70
+2022-12-11 02:43:42,979 INFO [train.py:421] (4/8) Epoch 3, batch 71800, loss[loss=2.486, over 1400.00 frames. , ppl: 12.008602358088803] tot_loss[loss=2.316, over 5554036.72 frames. , ppl: 10.133781666322145], batch size: 70
+2022-12-11 02:44:57,829 INFO [train.py:421] (4/8) Epoch 4, batch 0, loss[loss=2.237, over 4900.00 frames. , ppl: 9.361968445500755] tot_loss[loss=2.237, over 4900.00 frames. , ppl: 9.361968445500755], batch size: 70
+2022-12-11 02:46:37,238 INFO [train.py:421] (4/8) Epoch 4, batch 200, loss[loss=2.469, over 1960.00 frames. , ppl: 11.813768635761267] tot_loss[loss=2.304, over 543217.66 frames. , ppl: 10.01789883267699], batch size: 70
+2022-12-11 02:48:15,755 INFO [train.py:421] (4/8) Epoch 4, batch 400, loss[loss=2.316, over 2170.00 frames. , ppl: 10.138790110927356] tot_loss[loss=2.311, over 997941.51 frames. , ppl: 10.082918971610928], batch size: 70
+2022-12-11 02:49:54,209 INFO [train.py:421] (4/8) Epoch 4, batch 600, loss[loss=2.443, over 1470.00 frames. , ppl: 11.509399651366364] tot_loss[loss=2.314, over 1378183.94 frames. , ppl: 10.11723127262876], batch size: 70
+2022-12-11 02:51:35,082 INFO [train.py:421] (4/8) Epoch 4, batch 800, loss[loss=2.552, over 980.00 frames. , ppl: 12.837897600239375] tot_loss[loss=2.311, over 1775410.39 frames. , ppl: 10.088592852968217], batch size: 70
+2022-12-11 02:53:15,884 INFO [train.py:421] (4/8) Epoch 4, batch 1000, loss[loss=2.248, over 6790.00 frames. , ppl: 9.472808381346015] tot_loss[loss=2.31, over 2129653.11 frames. , ppl: 10.075180813662492], batch size: 70
+2022-12-11 02:53:15,884 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 02:53:16,642 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.006100117136109
+2022-12-11 02:54:58,858 INFO [train.py:421] (4/8) Epoch 4, batch 1200, loss[loss=2.523, over 840.00 frames. , ppl: 12.470603313231383] tot_loss[loss=2.308, over 2491841.02 frames. , ppl: 10.049688637207892], batch size: 70
+2022-12-11 02:56:36,318 INFO [train.py:421] (4/8) Epoch 4, batch 1400, loss[loss=2.361, over 3360.00 frames. , ppl: 10.601147020088371] tot_loss[loss=2.308, over 2773865.42 frames. , ppl: 10.052780773607376], batch size: 70
+2022-12-11 02:58:17,502 INFO [train.py:421] (4/8) Epoch 4, batch 1600, loss[loss=2.43, over 2240.00 frames. , ppl: 11.362709196485593] tot_loss[loss=2.308, over 3033190.57 frames. , ppl: 10.050362251215686], batch size: 70
+2022-12-11 02:59:58,155 INFO [train.py:421] (4/8) Epoch 4, batch 1800, loss[loss=2.301, over 4620.00 frames. , ppl: 9.986470949914299] tot_loss[loss=2.309, over 3243450.02 frames. , ppl: 10.064051024876264], batch size: 70
+2022-12-11 03:01:36,803 INFO [train.py:421] (4/8) Epoch 4, batch 2000, loss[loss=2.154, over 3850.00 frames. , ppl: 8.617061781807553] tot_loss[loss=2.308, over 3463368.33 frames. , ppl: 10.056246762116436], batch size: 70
+2022-12-11 03:01:36,804 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:01:37,550 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.020806653868545
+2022-12-11 03:03:19,536 INFO [train.py:421] (4/8) Epoch 4, batch 2200, loss[loss=2.467, over 1470.00 frames. , ppl: 11.788175883023134] tot_loss[loss=2.309, over 3652736.53 frames. , ppl: 10.06229192522486], batch size: 70
+2022-12-11 03:04:59,303 INFO [train.py:421] (4/8) Epoch 4, batch 2400, loss[loss=2.503, over 980.00 frames. , ppl: 12.220416781042198] tot_loss[loss=2.309, over 3837395.16 frames. , ppl: 10.063909258392703], batch size: 70
+2022-12-11 03:06:39,921 INFO [train.py:421] (4/8) Epoch 4, batch 2600, loss[loss=2.465, over 1470.00 frames. , ppl: 11.763337969558746] tot_loss[loss=2.308, over 4020365.84 frames. , ppl: 10.051373172155477], batch size: 70
+2022-12-11 03:08:21,018 INFO [train.py:421] (4/8) Epoch 4, batch 2800, loss[loss=2.173, over 3220.00 frames. , ppl: 8.781621770434487] tot_loss[loss=2.309, over 4088872.66 frames. , ppl: 10.068926993025276], batch size: 70
+2022-12-11 03:09:57,239 INFO [train.py:421] (4/8) Epoch 4, batch 3000, loss[loss=4.139, over 350.00 frames. , ppl: 62.728261108640716] tot_loss[loss=2.309, over 4215022.38 frames. , ppl: 10.06721207056468], batch size: 70
+2022-12-11 03:09:57,239 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:09:58,000 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.00204158109846
+2022-12-11 03:11:38,666 INFO [train.py:421] (4/8) Epoch 4, batch 3200, loss[loss=2.408, over 1890.00 frames. , ppl: 11.111848218293208] tot_loss[loss=2.308, over 4382314.93 frames. , ppl: 10.05332419564185], batch size: 70
+2022-12-11 03:13:21,675 INFO [train.py:421] (4/8) Epoch 4, batch 3400, loss[loss=2.807, over 700.00 frames. , ppl: 16.55691518402029] tot_loss[loss=2.308, over 4489123.97 frames. , ppl: 10.05728166155997], batch size: 70
+2022-12-11 03:15:01,796 INFO [train.py:421] (4/8) Epoch 4, batch 3600, loss[loss=3.487, over 420.00 frames. , ppl: 32.69517051431339] tot_loss[loss=2.308, over 4610596.55 frames. , ppl: 10.051580235260358], batch size: 70
+2022-12-11 03:16:43,481 INFO [train.py:421] (4/8) Epoch 4, batch 3800, loss[loss=2.311, over 2590.00 frames. , ppl: 10.0828967550576] tot_loss[loss=2.308, over 4657373.79 frames. , ppl: 10.0575718886875], batch size: 70
+2022-12-11 03:18:23,716 INFO [train.py:421] (4/8) Epoch 4, batch 4000, loss[loss=2.413, over 1610.00 frames. , ppl: 11.16391294009391] tot_loss[loss=2.307, over 4783362.37 frames. , ppl: 10.043932593241866], batch size: 70
+2022-12-11 03:18:23,717 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:18:24,476 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.024831259142056
+2022-12-11 03:20:05,985 INFO [train.py:421] (4/8) Epoch 4, batch 4200, loss[loss=2.244, over 5040.00 frames. , ppl: 9.4306431910283] tot_loss[loss=2.307, over 4854304.82 frames. , ppl: 10.045016823997072], batch size: 70
+2022-12-11 03:21:49,026 INFO [train.py:421] (4/8) Epoch 4, batch 4400, loss[loss=2.391, over 2660.00 frames. , ppl: 10.919896054800187] tot_loss[loss=2.308, over 4910595.69 frames. , ppl: 10.050534591566361], batch size: 70
+2022-12-11 03:23:27,758 INFO [train.py:421] (4/8) Epoch 4, batch 4600, loss[loss=2.488, over 1610.00 frames. , ppl: 12.039652550956074] tot_loss[loss=2.31, over 4884460.93 frames. , ppl: 10.072470227481546], batch size: 70
+2022-12-11 03:25:07,823 INFO [train.py:421] (4/8) Epoch 4, batch 4800, loss[loss=2.299, over 2870.00 frames. , ppl: 9.96377040315998] tot_loss[loss=2.309, over 4990045.24 frames. , ppl: 10.064245532129942], batch size: 70
+2022-12-11 03:26:49,291 INFO [train.py:421] (4/8) Epoch 4, batch 5000, loss[loss=2.626, over 840.00 frames. , ppl: 13.82086534419735] tot_loss[loss=2.309, over 5039639.33 frames. , ppl: 10.063869517681232], batch size: 70
+2022-12-11 03:26:49,291 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:26:50,036 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.007444935075092
+2022-12-11 03:28:31,282 INFO [train.py:421] (4/8) Epoch 4, batch 5200, loss[loss=2.43, over 1330.00 frames. , ppl: 11.355339237232288] tot_loss[loss=2.309, over 5086646.64 frames. , ppl: 10.06093046962282], batch size: 70
+2022-12-11 03:30:12,462 INFO [train.py:421] (4/8) Epoch 4, batch 5400, loss[loss=2.207, over 8820.00 frames. , ppl: 9.089406738166472] tot_loss[loss=2.309, over 5134439.08 frames. , ppl: 10.061285154201057], batch size: 70
+2022-12-11 03:31:52,290 INFO [train.py:421] (4/8) Epoch 4, batch 5600, loss[loss=2.628, over 840.00 frames. , ppl: 13.846629919741906] tot_loss[loss=2.309, over 5186738.93 frames. , ppl: 10.059805101555009], batch size: 70
+2022-12-11 03:33:32,843 INFO [train.py:421] (4/8) Epoch 4, batch 5800, loss[loss=2.255, over 10850.00 frames. , ppl: 9.532458135431543] tot_loss[loss=2.308, over 5251265.77 frames. , ppl: 10.051964445872498], batch size: 70
+2022-12-11 03:35:14,355 INFO [train.py:421] (4/8) Epoch 4, batch 6000, loss[loss=2.353, over 1400.00 frames. , ppl: 10.512472719535157] tot_loss[loss=2.307, over 5275792.31 frames. , ppl: 10.04886762088429], batch size: 70
+2022-12-11 03:35:14,355 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:35:15,103 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.002629307959912
+2022-12-11 03:36:56,273 INFO [train.py:421] (4/8) Epoch 4, batch 6200, loss[loss=2.278, over 4200.00 frames. , ppl: 9.754006331430249] tot_loss[loss=2.307, over 5323257.90 frames. , ppl: 10.046350074245389], batch size: 70
+2022-12-11 03:38:34,622 INFO [train.py:421] (4/8) Epoch 4, batch 6400, loss[loss=2.244, over 3150.00 frames. , ppl: 9.43276237088995] tot_loss[loss=2.308, over 5307281.22 frames. , ppl: 10.058791882162472], batch size: 70
+2022-12-11 03:40:16,694 INFO [train.py:421] (4/8) Epoch 4, batch 6600, loss[loss=2.365, over 2100.00 frames. , ppl: 10.64380617969499] tot_loss[loss=2.308, over 5321236.45 frames. , ppl: 10.058586537537108], batch size: 70
+2022-12-11 03:41:57,721 INFO [train.py:421] (4/8) Epoch 4, batch 6800, loss[loss=2.284, over 3500.00 frames. , ppl: 9.816991161060903] tot_loss[loss=2.308, over 5338811.09 frames. , ppl: 10.059100716384137], batch size: 70
+2022-12-11 03:43:38,826 INFO [train.py:421] (4/8) Epoch 4, batch 7000, loss[loss=2.256, over 4340.00 frames. , ppl: 9.546004040492111] tot_loss[loss=2.31, over 5320544.55 frames. , ppl: 10.071380709807313], batch size: 70
+2022-12-11 03:43:38,826 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:43:39,587 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.304, over 211138.00 frames. , ppl: 10.018172918207874
+2022-12-11 03:45:21,082 INFO [train.py:421] (4/8) Epoch 4, batch 7200, loss[loss=2.262, over 3710.00 frames. , ppl: 9.602874887189543] tot_loss[loss=2.308, over 5380111.15 frames. , ppl: 10.056482323836542], batch size: 70
+2022-12-11 03:47:01,073 INFO [train.py:421] (4/8) Epoch 4, batch 7400, loss[loss=2.305, over 3290.00 frames. , ppl: 10.019944186267642] tot_loss[loss=2.309, over 5369286.91 frames. , ppl: 10.062301493619461], batch size: 70
+2022-12-11 03:48:40,876 INFO [train.py:421] (4/8) Epoch 4, batch 7600, loss[loss=2.205, over 4620.00 frames. , ppl: 9.068738221660075] tot_loss[loss=2.308, over 5381066.41 frames. , ppl: 10.058344909854144], batch size: 70
+2022-12-11 03:50:23,892 INFO [train.py:421] (4/8) Epoch 4, batch 7800, loss[loss=2.34, over 4130.00 frames. , ppl: 10.385570510138496] tot_loss[loss=2.308, over 5400188.54 frames. , ppl: 10.057688186978579], batch size: 70
+2022-12-11 03:52:01,577 INFO [train.py:421] (4/8) Epoch 4, batch 8000, loss[loss=2.314, over 4060.00 frames. , ppl: 10.119297306100973] tot_loss[loss=2.309, over 5375068.12 frames. , ppl: 10.06354484504915], batch size: 70
+2022-12-11 03:52:01,577 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 03:52:02,323 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.304, over 211138.00 frames. , ppl: 10.015572487058913
+2022-12-11 03:53:42,801 INFO [train.py:421] (4/8) Epoch 4, batch 8200, loss[loss=2.281, over 3290.00 frames. , ppl: 9.784412038680808] tot_loss[loss=2.31, over 5367486.03 frames. , ppl: 10.074825096487293], batch size: 70
+2022-12-11 03:55:20,981 INFO [train.py:421] (4/8) Epoch 4, batch 8400, loss[loss=2.656, over 770.00 frames. , ppl: 14.233792352421164] tot_loss[loss=2.311, over 5371781.70 frames. , ppl: 10.081066415686962], batch size: 70
+2022-12-11 03:56:59,259 INFO [train.py:421] (4/8) Epoch 4, batch 8600, loss[loss=2.45, over 1540.00 frames. , ppl: 11.586190124717486] tot_loss[loss=2.31, over 5413778.88 frames. , ppl: 10.073144291196432], batch size: 70
+2022-12-11 03:58:39,457 INFO [train.py:421] (4/8) Epoch 4, batch 8800, loss[loss=2.554, over 840.00 frames. , ppl: 12.859281075720771] tot_loss[loss=2.31, over 5413694.34 frames. , ppl: 10.072556205615284], batch size: 70
+2022-12-11 04:00:15,936 INFO [train.py:421] (4/8) Epoch 4, batch 9000, loss[loss=2.712, over 700.00 frames. , ppl: 15.058600480991082] tot_loss[loss=2.31, over 5399847.65 frames. , ppl: 10.07938138951895], batch size: 70
+2022-12-11 04:00:15,936 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:00:16,697 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.02282839694073
+2022-12-11 04:01:59,809 INFO [train.py:421] (4/8) Epoch 4, batch 9200, loss[loss=2.149, over 4690.00 frames. , ppl: 8.580483568313529] tot_loss[loss=2.31, over 5401366.13 frames. , ppl: 10.072470657981375], batch size: 70
+2022-12-11 04:03:46,666 INFO [train.py:421] (4/8) Epoch 4, batch 9400, loss[loss=2.554, over 980.00 frames. , ppl: 12.857530458209212] tot_loss[loss=2.31, over 5466791.12 frames. , ppl: 10.069941285759679], batch size: 70
+2022-12-11 04:05:23,921 INFO [train.py:421] (4/8) Epoch 4, batch 9600, loss[loss=2.376, over 1540.00 frames. , ppl: 10.760283943994684] tot_loss[loss=2.311, over 5429873.79 frames. , ppl: 10.081601790572204], batch size: 70
+2022-12-11 04:07:04,976 INFO [train.py:421] (4/8) Epoch 4, batch 9800, loss[loss=2.33, over 1890.00 frames. , ppl: 10.278478980793613] tot_loss[loss=2.311, over 5448370.94 frames. , ppl: 10.080096152579241], batch size: 70
+2022-12-11 04:08:44,055 INFO [train.py:421] (4/8) Epoch 4, batch 10000, loss[loss=2.384, over 1890.00 frames. , ppl: 10.850239783957258] tot_loss[loss=2.312, over 5407785.49 frames. , ppl: 10.09867678615571], batch size: 70
+2022-12-11 04:08:44,055 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:08:44,815 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.305, over 211138.00 frames. , ppl: 10.02091344153873
+2022-12-11 04:10:25,960 INFO [train.py:421] (4/8) Epoch 4, batch 10200, loss[loss=2.234, over 4550.00 frames. , ppl: 9.338180588308614] tot_loss[loss=2.312, over 5415505.50 frames. , ppl: 10.095535377776269], batch size: 70
+2022-12-11 04:12:05,325 INFO [train.py:421] (4/8) Epoch 4, batch 10400, loss[loss=2.373, over 2240.00 frames. , ppl: 10.72471111074329] tot_loss[loss=2.312, over 5441573.25 frames. , ppl: 10.090121146227428], batch size: 70
+2022-12-11 04:13:45,272 INFO [train.py:421] (4/8) Epoch 4, batch 10600, loss[loss=2.383, over 2520.00 frames. , ppl: 10.841231372633377] tot_loss[loss=2.312, over 5411202.48 frames. , ppl: 10.096194120624665], batch size: 70
+2022-12-11 04:15:25,275 INFO [train.py:421] (4/8) Epoch 4, batch 10800, loss[loss=2.248, over 2310.00 frames. , ppl: 9.471731495769241] tot_loss[loss=2.311, over 5467537.78 frames. , ppl: 10.084707313475782], batch size: 70
+2022-12-11 04:17:08,163 INFO [train.py:421] (4/8) Epoch 4, batch 11000, loss[loss=2.197, over 6860.00 frames. , ppl: 9.002301779572356] tot_loss[loss=2.311, over 5463901.39 frames. , ppl: 10.084472038990624], batch size: 70
+2022-12-11 04:17:08,163 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:17:08,910 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.303, over 211138.00 frames. , ppl: 10.005364098386623
+2022-12-11 04:18:47,976 INFO [train.py:421] (4/8) Epoch 4, batch 11200, loss[loss=2.317, over 2170.00 frames. , ppl: 10.148444916074514] tot_loss[loss=2.312, over 5419413.54 frames. , ppl: 10.097332735926582], batch size: 70
+2022-12-11 04:20:23,515 INFO [train.py:421] (4/8) Epoch 4, batch 11400, loss[loss=2.286, over 2870.00 frames. , ppl: 9.834175964855032] tot_loss[loss=2.312, over 5416665.53 frames. , ppl: 10.097580782349207], batch size: 70
+2022-12-11 04:22:05,513 INFO [train.py:421] (4/8) Epoch 4, batch 11600, loss[loss=2.369, over 2520.00 frames. , ppl: 10.68930083310423] tot_loss[loss=2.31, over 5478036.96 frames. , ppl: 10.077605008790368], batch size: 70
+2022-12-11 04:23:44,973 INFO [train.py:421] (4/8) Epoch 4, batch 11800, loss[loss=2.619, over 770.00 frames. , ppl: 13.722543022623002] tot_loss[loss=2.31, over 5479448.28 frames. , ppl: 10.071002829862776], batch size: 70
+2022-12-11 04:25:25,250 INFO [train.py:421] (4/8) Epoch 4, batch 12000, loss[loss=2.253, over 2940.00 frames. , ppl: 9.517140546296535] tot_loss[loss=2.307, over 5582119.40 frames. , ppl: 10.047263986895706], batch size: 70
+2022-12-11 04:25:25,251 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:25:25,998 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.98219907371343
+2022-12-11 04:27:07,827 INFO [train.py:421] (4/8) Epoch 4, batch 12200, loss[loss=2.358, over 2030.00 frames. , ppl: 10.571612438436318] tot_loss[loss=2.308, over 5558460.36 frames. , ppl: 10.051244454451064], batch size: 70
+2022-12-11 04:28:46,443 INFO [train.py:421] (4/8) Epoch 4, batch 12400, loss[loss=2.505, over 910.00 frames. , ppl: 12.248228185768792] tot_loss[loss=2.307, over 5569871.34 frames. , ppl: 10.045198061772535], batch size: 70
+2022-12-11 04:30:27,175 INFO [train.py:421] (4/8) Epoch 4, batch 12600, loss[loss=2.424, over 1960.00 frames. , ppl: 11.293763127825363] tot_loss[loss=2.309, over 5510338.32 frames. , ppl: 10.065030798779784], batch size: 70
+2022-12-11 04:32:07,001 INFO [train.py:421] (4/8) Epoch 4, batch 12800, loss[loss=2.232, over 2380.00 frames. , ppl: 9.313886422760246] tot_loss[loss=2.311, over 5439511.06 frames. , ppl: 10.08751204403478], batch size: 70
+2022-12-11 04:33:50,382 INFO [train.py:421] (4/8) Epoch 4, batch 13000, loss[loss=2.565, over 1400.00 frames. , ppl: 13.00446317868699] tot_loss[loss=2.311, over 5432773.99 frames. , ppl: 10.085107883116603], batch size: 70
+2022-12-11 04:33:50,382 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:33:51,146 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.97099750372906
+2022-12-11 04:35:32,578 INFO [train.py:421] (4/8) Epoch 4, batch 13200, loss[loss=2.559, over 980.00 frames. , ppl: 12.924183124982651] tot_loss[loss=2.312, over 5427281.35 frames. , ppl: 10.091080913723042], batch size: 70
+2022-12-11 04:37:11,847 INFO [train.py:421] (4/8) Epoch 4, batch 13400, loss[loss=2.452, over 1400.00 frames. , ppl: 11.607884555887022] tot_loss[loss=2.312, over 5439439.01 frames. , ppl: 10.092549313078642], batch size: 70
+2022-12-11 04:38:49,025 INFO [train.py:421] (4/8) Epoch 4, batch 13600, loss[loss=2.207, over 5180.00 frames. , ppl: 9.087119715466857] tot_loss[loss=2.313, over 5404979.82 frames. , ppl: 10.106277147169651], batch size: 70
+2022-12-11 04:40:27,700 INFO [train.py:421] (4/8) Epoch 4, batch 13800, loss[loss=2.404, over 1610.00 frames. , ppl: 11.068362307229778] tot_loss[loss=2.313, over 5392945.51 frames. , ppl: 10.102220952604496], batch size: 70
+2022-12-11 04:42:07,520 INFO [train.py:421] (4/8) Epoch 4, batch 14000, loss[loss=2.557, over 840.00 frames. , ppl: 12.89867815147704] tot_loss[loss=2.312, over 5414618.59 frames. , ppl: 10.09460386843109], batch size: 70
+2022-12-11 04:42:07,520 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:42:08,266 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.986692979928648
+2022-12-11 04:43:46,542 INFO [train.py:421] (4/8) Epoch 4, batch 14200, loss[loss=2.27, over 6440.00 frames. , ppl: 9.676124977302914] tot_loss[loss=2.312, over 5412115.58 frames. , ppl: 10.091056909961479], batch size: 70
+2022-12-11 04:45:27,608 INFO [train.py:421] (4/8) Epoch 4, batch 14400, loss[loss=2.448, over 1820.00 frames. , ppl: 11.568478633150292] tot_loss[loss=2.312, over 5385219.71 frames. , ppl: 10.096860081817255], batch size: 70
+2022-12-11 04:47:07,723 INFO [train.py:421] (4/8) Epoch 4, batch 14600, loss[loss=2.264, over 3920.00 frames. , ppl: 9.624559378674864] tot_loss[loss=2.309, over 5479696.81 frames. , ppl: 10.069341773146808], batch size: 70
+2022-12-11 04:48:49,492 INFO [train.py:421] (4/8) Epoch 4, batch 14800, loss[loss=2.242, over 5180.00 frames. , ppl: 9.413443813037851] tot_loss[loss=2.309, over 5497266.25 frames. , ppl: 10.062087746789656], batch size: 70
+2022-12-11 04:50:23,944 INFO [train.py:421] (4/8) Epoch 4, batch 15000, loss[loss=2.258, over 3850.00 frames. , ppl: 9.567326411965515] tot_loss[loss=2.31, over 5476060.60 frames. , ppl: 10.071378586296952], batch size: 70
+2022-12-11 04:50:23,945 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:50:24,691 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.975186652961758
+2022-12-11 04:52:04,459 INFO [train.py:421] (4/8) Epoch 4, batch 15200, loss[loss=2.505, over 980.00 frames. , ppl: 12.249061458209441] tot_loss[loss=2.31, over 5457672.80 frames. , ppl: 10.079108283360446], batch size: 70
+2022-12-11 04:53:41,644 INFO [train.py:421] (4/8) Epoch 4, batch 15400, loss[loss=2.399, over 1750.00 frames. , ppl: 11.011316856827218] tot_loss[loss=2.311, over 5431910.97 frames. , ppl: 10.083599894180068], batch size: 70
+2022-12-11 04:55:25,232 INFO [train.py:421] (4/8) Epoch 4, batch 15600, loss[loss=2.272, over 2310.00 frames. , ppl: 9.69724284584185] tot_loss[loss=2.309, over 5501059.21 frames. , ppl: 10.06362807044785], batch size: 70
+2022-12-11 04:57:01,735 INFO [train.py:421] (4/8) Epoch 4, batch 15800, loss[loss=2.315, over 3500.00 frames. , ppl: 10.122007903053614] tot_loss[loss=2.31, over 5463256.58 frames. , ppl: 10.074123292821389], batch size: 70
+2022-12-11 04:58:46,138 INFO [train.py:421] (4/8) Epoch 4, batch 16000, loss[loss=2.32, over 2380.00 frames. , ppl: 10.175914805419442] tot_loss[loss=2.309, over 5522565.50 frames. , ppl: 10.063998124388185], batch size: 70
+2022-12-11 04:58:46,138 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 04:58:46,898 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.980866511809946
+2022-12-11 05:00:26,707 INFO [train.py:421] (4/8) Epoch 4, batch 16200, loss[loss=2.285, over 3500.00 frames. , ppl: 9.822174936366906] tot_loss[loss=2.309, over 5510053.19 frames. , ppl: 10.061814977142543], batch size: 70
+2022-12-11 05:02:10,985 INFO [train.py:421] (4/8) Epoch 4, batch 16400, loss[loss=2.278, over 2310.00 frames. , ppl: 9.761221497200626] tot_loss[loss=2.307, over 5550707.13 frames. , ppl: 10.044343541078716], batch size: 70
+2022-12-11 05:03:48,549 INFO [train.py:421] (4/8) Epoch 4, batch 16600, loss[loss=2.36, over 2380.00 frames. , ppl: 10.587789570842366] tot_loss[loss=2.306, over 5593292.38 frames. , ppl: 10.035530859573019], batch size: 70
+2022-12-11 05:05:30,768 INFO [train.py:421] (4/8) Epoch 4, batch 16800, loss[loss=2.403, over 1470.00 frames. , ppl: 11.057174502463285] tot_loss[loss=2.307, over 5565150.97 frames. , ppl: 10.040510313155261], batch size: 70
+2022-12-11 05:07:12,257 INFO [train.py:421] (4/8) Epoch 4, batch 17000, loss[loss=2.511, over 1680.00 frames. , ppl: 12.313912119434699] tot_loss[loss=2.307, over 5551038.59 frames. , ppl: 10.046263679725003], batch size: 70
+2022-12-11 05:07:12,257 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:07:13,003 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.993913175932885
+2022-12-11 05:08:54,242 INFO [train.py:421] (4/8) Epoch 4, batch 17200, loss[loss=2.252, over 5110.00 frames. , ppl: 9.504964008961716] tot_loss[loss=2.307, over 5579978.49 frames. , ppl: 10.039475549092998], batch size: 70
+2022-12-11 05:10:34,567 INFO [train.py:421] (4/8) Epoch 4, batch 17400, loss[loss=2.206, over 5530.00 frames. , ppl: 9.083447389385109] tot_loss[loss=2.308, over 5555747.50 frames. , ppl: 10.051096528202427], batch size: 70
+2022-12-11 05:12:12,659 INFO [train.py:421] (4/8) Epoch 4, batch 17600, loss[loss=2.374, over 2380.00 frames. , ppl: 10.73667040275554] tot_loss[loss=2.309, over 5506972.50 frames. , ppl: 10.065254573973387], batch size: 70
+2022-12-11 05:13:51,762 INFO [train.py:421] (4/8) Epoch 4, batch 17800, loss[loss=2.336, over 4410.00 frames. , ppl: 10.338743100027811] tot_loss[loss=2.31, over 5465623.01 frames. , ppl: 10.0734462900166], batch size: 70
+2022-12-11 05:15:30,827 INFO [train.py:421] (4/8) Epoch 4, batch 18000, loss[loss=2.319, over 3780.00 frames. , ppl: 10.164737720262478] tot_loss[loss=2.31, over 5476558.82 frames. , ppl: 10.07367675111878], batch size: 70
+2022-12-11 05:15:30,828 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:15:31,578 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.995456073238609
+2022-12-11 05:17:10,754 INFO [train.py:421] (4/8) Epoch 4, batch 18200, loss[loss=2.902, over 630.00 frames. , ppl: 18.207806504314608] tot_loss[loss=2.31, over 5492490.57 frames. , ppl: 10.071339412993858], batch size: 70
+2022-12-11 05:18:48,983 INFO [train.py:421] (4/8) Epoch 4, batch 18400, loss[loss=2.193, over 3430.00 frames. , ppl: 8.958457470420614] tot_loss[loss=2.31, over 5474034.68 frames. , ppl: 10.07353473672556], batch size: 70
+2022-12-11 05:20:26,278 INFO [train.py:421] (4/8) Epoch 4, batch 18600, loss[loss=2.095, over 5180.00 frames. , ppl: 8.122398666177753] tot_loss[loss=2.309, over 5494154.51 frames. , ppl: 10.06529868279985], batch size: 70
+2022-12-11 05:22:06,457 INFO [train.py:421] (4/8) Epoch 4, batch 18800, loss[loss=2.264, over 4480.00 frames. , ppl: 9.622347155335264] tot_loss[loss=2.309, over 5490796.93 frames. , ppl: 10.068658244904608], batch size: 70
+2022-12-11 05:23:48,712 INFO [train.py:421] (4/8) Epoch 4, batch 19000, loss[loss=2.242, over 4410.00 frames. , ppl: 9.412794811430812] tot_loss[loss=2.311, over 5472525.42 frames. , ppl: 10.08055515796572], batch size: 70
+2022-12-11 05:23:48,712 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:23:49,478 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.977298131379836
+2022-12-11 05:25:29,342 INFO [train.py:421] (4/8) Epoch 4, batch 19200, loss[loss=2.349, over 3710.00 frames. , ppl: 10.474899254423551] tot_loss[loss=2.311, over 5476211.84 frames. , ppl: 10.084538892394974], batch size: 70
+2022-12-11 05:27:11,514 INFO [train.py:421] (4/8) Epoch 4, batch 19400, loss[loss=2.573, over 770.00 frames. , ppl: 13.110507882835682] tot_loss[loss=2.311, over 5456359.45 frames. , ppl: 10.08552984673666], batch size: 70
+2022-12-11 05:28:50,710 INFO [train.py:421] (4/8) Epoch 4, batch 19600, loss[loss=2.327, over 2030.00 frames. , ppl: 10.251086279095993] tot_loss[loss=2.309, over 5518925.61 frames. , ppl: 10.067298585779897], batch size: 70
+2022-12-11 05:30:32,726 INFO [train.py:421] (4/8) Epoch 4, batch 19800, loss[loss=2.298, over 3360.00 frames. , ppl: 9.959051370643094] tot_loss[loss=2.311, over 5450093.21 frames. , ppl: 10.084681834450498], batch size: 70
+2022-12-11 05:32:10,663 INFO [train.py:421] (4/8) Epoch 4, batch 20000, loss[loss=2.308, over 2590.00 frames. , ppl: 10.054002978157142] tot_loss[loss=2.311, over 5446288.07 frames. , ppl: 10.08249890113763], batch size: 70
+2022-12-11 05:32:10,663 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:32:11,424 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.990693533416067
+2022-12-11 05:33:50,978 INFO [train.py:421] (4/8) Epoch 4, batch 20200, loss[loss=2.765, over 700.00 frames. , ppl: 15.880031353924501] tot_loss[loss=2.31, over 5474307.41 frames. , ppl: 10.075477220151587], batch size: 70
+2022-12-11 05:35:32,027 INFO [train.py:421] (4/8) Epoch 4, batch 20400, loss[loss=2.263, over 7350.00 frames. , ppl: 9.608935553477778] tot_loss[loss=2.31, over 5485863.62 frames. , ppl: 10.075975558293244], batch size: 70
+2022-12-11 05:37:16,469 INFO [train.py:421] (4/8) Epoch 4, batch 20600, loss[loss=2.277, over 5040.00 frames. , ppl: 9.746805220596883] tot_loss[loss=2.309, over 5543695.93 frames. , ppl: 10.063301930943199], batch size: 70
+2022-12-11 05:38:53,403 INFO [train.py:421] (4/8) Epoch 4, batch 20800, loss[loss=2.278, over 3780.00 frames. , ppl: 9.757887316549677] tot_loss[loss=2.31, over 5514962.40 frames. , ppl: 10.077335214448425], batch size: 70
+2022-12-11 05:40:32,944 INFO [train.py:421] (4/8) Epoch 4, batch 21000, loss[loss=2.24, over 5460.00 frames. , ppl: 9.397153213002113] tot_loss[loss=2.31, over 5524338.72 frames. , ppl: 10.069773862859803], batch size: 70
+2022-12-11 05:40:32,945 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:40:33,720 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.982080879215324
+2022-12-11 05:42:16,404 INFO [train.py:421] (4/8) Epoch 4, batch 21200, loss[loss=2.286, over 2940.00 frames. , ppl: 9.835331717918734] tot_loss[loss=2.31, over 5507565.64 frames. , ppl: 10.073093309795158], batch size: 70
+2022-12-11 05:43:57,340 INFO [train.py:421] (4/8) Epoch 4, batch 21400, loss[loss=2.504, over 1050.00 frames. , ppl: 12.226882335325334] tot_loss[loss=2.309, over 5530341.02 frames. , ppl: 10.064161554550003], batch size: 70
+2022-12-11 05:45:35,240 INFO [train.py:421] (4/8) Epoch 4, batch 21600, loss[loss=2.212, over 3500.00 frames. , ppl: 9.132465102910572] tot_loss[loss=2.308, over 5527383.16 frames. , ppl: 10.058108996940387], batch size: 70
+2022-12-11 05:47:14,431 INFO [train.py:421] (4/8) Epoch 4, batch 21800, loss[loss=2.357, over 1890.00 frames. , ppl: 10.556539387433752] tot_loss[loss=2.309, over 5519039.07 frames. , ppl: 10.062589280644183], batch size: 70
+2022-12-11 05:48:55,726 INFO [train.py:421] (4/8) Epoch 4, batch 22000, loss[loss=2.863, over 700.00 frames. , ppl: 17.509805811409148] tot_loss[loss=2.309, over 5550342.24 frames. , ppl: 10.062701683628337], batch size: 70
+2022-12-11 05:48:55,727 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:48:56,472 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.988403293971805
+2022-12-11 05:50:35,529 INFO [train.py:421] (4/8) Epoch 4, batch 22200, loss[loss=2.367, over 1820.00 frames. , ppl: 10.666996129259148] tot_loss[loss=2.31, over 5519188.22 frames. , ppl: 10.06992205343409], batch size: 70
+2022-12-11 05:52:14,307 INFO [train.py:421] (4/8) Epoch 4, batch 22400, loss[loss=2.229, over 4060.00 frames. , ppl: 9.289731194312392] tot_loss[loss=2.309, over 5518542.60 frames. , ppl: 10.061958806314557], batch size: 70
+2022-12-11 05:53:54,670 INFO [train.py:421] (4/8) Epoch 4, batch 22600, loss[loss=2.336, over 2310.00 frames. , ppl: 10.339878308146215] tot_loss[loss=2.308, over 5552471.73 frames. , ppl: 10.059142017603193], batch size: 70
+2022-12-11 05:55:36,838 INFO [train.py:421] (4/8) Epoch 4, batch 22800, loss[loss=2.313, over 2170.00 frames. , ppl: 10.101869584699088] tot_loss[loss=2.308, over 5578931.76 frames. , ppl: 10.05733892718976], batch size: 70
+2022-12-11 05:57:12,502 INFO [train.py:421] (4/8) Epoch 4, batch 23000, loss[loss=2.294, over 2240.00 frames. , ppl: 9.916911687391192] tot_loss[loss=2.309, over 5536193.04 frames. , ppl: 10.066384060986689], batch size: 70
+2022-12-11 05:57:12,502 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 05:57:13,264 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.298, over 211138.00 frames. , ppl: 9.957022795141143
+2022-12-11 05:58:56,279 INFO [train.py:421] (4/8) Epoch 4, batch 23200, loss[loss=2.353, over 3640.00 frames. , ppl: 10.512943447068531] tot_loss[loss=2.309, over 5545258.11 frames. , ppl: 10.064005501808685], batch size: 70
+2022-12-11 06:00:34,256 INFO [train.py:421] (4/8) Epoch 4, batch 23400, loss[loss=2.271, over 2590.00 frames. , ppl: 9.688305623573736] tot_loss[loss=2.309, over 5529912.43 frames. , ppl: 10.067072610126527], batch size: 70
+2022-12-11 06:02:14,832 INFO [train.py:421] (4/8) Epoch 4, batch 23600, loss[loss=2.355, over 2030.00 frames. , ppl: 10.53486155067933] tot_loss[loss=2.308, over 5564351.56 frames. , ppl: 10.053394771799075], batch size: 70
+2022-12-11 06:03:55,549 INFO [train.py:421] (4/8) Epoch 4, batch 23800, loss[loss=2.519, over 1260.00 frames. , ppl: 12.415290809085892] tot_loss[loss=2.309, over 5546120.67 frames. , ppl: 10.060045889289823], batch size: 70
+2022-12-11 06:05:34,129 INFO [train.py:421] (4/8) Epoch 4, batch 24000, loss[loss=2.552, over 840.00 frames. , ppl: 12.82705968999159] tot_loss[loss=2.307, over 5567872.19 frames. , ppl: 10.046589917100324], batch size: 70
+2022-12-11 06:05:34,129 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:05:34,888 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.975346105661082
+2022-12-11 06:07:17,272 INFO [train.py:421] (4/8) Epoch 4, batch 24200, loss[loss=2.289, over 2660.00 frames. , ppl: 9.869027916036217] tot_loss[loss=2.306, over 5604560.20 frames. , ppl: 10.03603960871211], batch size: 70
+2022-12-11 06:08:57,886 INFO [train.py:421] (4/8) Epoch 4, batch 24400, loss[loss=2.373, over 1750.00 frames. , ppl: 10.725713326313032] tot_loss[loss=2.307, over 5557127.92 frames. , ppl: 10.043923604861693], batch size: 70
+2022-12-11 06:10:39,376 INFO [train.py:421] (4/8) Epoch 4, batch 24600, loss[loss=2.161, over 6930.00 frames. , ppl: 8.683728660289603] tot_loss[loss=2.308, over 5529959.22 frames. , ppl: 10.052371819646762], batch size: 70
+2022-12-11 06:12:21,116 INFO [train.py:421] (4/8) Epoch 4, batch 24800, loss[loss=2.23, over 9170.00 frames. , ppl: 9.30445629524353] tot_loss[loss=2.308, over 5514580.99 frames. , ppl: 10.053542952724305], batch size: 70
+2022-12-11 06:14:01,869 INFO [train.py:421] (4/8) Epoch 4, batch 25000, loss[loss=2.406, over 2170.00 frames. , ppl: 11.09201243039394] tot_loss[loss=2.307, over 5542913.75 frames. , ppl: 10.046157848688084], batch size: 70
+2022-12-11 06:14:01,869 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:14:02,628 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.974898758710204
+2022-12-11 06:15:42,489 INFO [train.py:421] (4/8) Epoch 4, batch 25200, loss[loss=3.265, over 490.00 frames. , ppl: 26.181564148360675] tot_loss[loss=2.308, over 5522729.06 frames. , ppl: 10.052620693109818], batch size: 70
+2022-12-11 06:17:23,860 INFO [train.py:421] (4/8) Epoch 4, batch 25400, loss[loss=2.236, over 2380.00 frames. , ppl: 9.35300294737738] tot_loss[loss=2.309, over 5488973.03 frames. , ppl: 10.061958167391118], batch size: 70
+2022-12-11 06:19:05,299 INFO [train.py:421] (4/8) Epoch 4, batch 25600, loss[loss=2.282, over 2870.00 frames. , ppl: 9.795296184816102] tot_loss[loss=2.308, over 5511670.30 frames. , ppl: 10.058304706523138], batch size: 70
+2022-12-11 06:20:48,746 INFO [train.py:421] (4/8) Epoch 4, batch 25800, loss[loss=2.687, over 700.00 frames. , ppl: 14.683509008540401] tot_loss[loss=2.308, over 5547271.23 frames. , ppl: 10.050550268324931], batch size: 70
+2022-12-11 06:22:27,270 INFO [train.py:421] (4/8) Epoch 4, batch 26000, loss[loss=2.31, over 2940.00 frames. , ppl: 10.07233138822347] tot_loss[loss=2.309, over 5546898.11 frames. , ppl: 10.06018541093724], batch size: 70
+2022-12-11 06:22:27,271 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:22:28,032 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.983234812864332
+2022-12-11 06:24:08,778 INFO [train.py:421] (4/8) Epoch 4, batch 26200, loss[loss=2.233, over 4130.00 frames. , ppl: 9.328601712178154] tot_loss[loss=2.31, over 5505042.08 frames. , ppl: 10.077732745440906], batch size: 70
+2022-12-11 06:25:49,600 INFO [train.py:421] (4/8) Epoch 4, batch 26400, loss[loss=2.398, over 1260.00 frames. , ppl: 11.002705942168175] tot_loss[loss=2.311, over 5489151.09 frames. , ppl: 10.082413347281017], batch size: 70
+2022-12-11 06:27:29,120 INFO [train.py:421] (4/8) Epoch 4, batch 26600, loss[loss=3.169, over 490.00 frames. , ppl: 23.789036626528606] tot_loss[loss=2.312, over 5446994.04 frames. , ppl: 10.091236235727369], batch size: 70
+2022-12-11 06:29:04,088 INFO [train.py:421] (4/8) Epoch 4, batch 26800, loss[loss=2.22, over 5320.00 frames. , ppl: 9.206912397228514] tot_loss[loss=2.312, over 5442169.87 frames. , ppl: 10.089770185156425], batch size: 70
+2022-12-11 06:30:43,400 INFO [train.py:421] (4/8) Epoch 4, batch 27000, loss[loss=2.47, over 1680.00 frames. , ppl: 11.827716282554178] tot_loss[loss=2.313, over 5386224.92 frames. , ppl: 10.107615329892392], batch size: 70
+2022-12-11 06:30:43,401 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:30:44,146 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.968636532001739
+2022-12-11 06:32:23,960 INFO [train.py:421] (4/8) Epoch 4, batch 27200, loss[loss=2.327, over 2520.00 frames. , ppl: 10.249151692746283] tot_loss[loss=2.313, over 5411539.48 frames. , ppl: 10.10624376471383], batch size: 70
+2022-12-11 06:34:06,620 INFO [train.py:421] (4/8) Epoch 4, batch 27400, loss[loss=2.354, over 2170.00 frames. , ppl: 10.525955360315558] tot_loss[loss=2.311, over 5454071.30 frames. , ppl: 10.084778353478963], batch size: 70
+2022-12-11 06:35:42,978 INFO [train.py:421] (4/8) Epoch 4, batch 27600, loss[loss=2.332, over 3220.00 frames. , ppl: 10.295239128427555] tot_loss[loss=2.311, over 5471496.38 frames. , ppl: 10.088156373497418], batch size: 70
+2022-12-11 06:37:23,603 INFO [train.py:421] (4/8) Epoch 4, batch 27800, loss[loss=2.238, over 4060.00 frames. , ppl: 9.377807121014596] tot_loss[loss=2.312, over 5464991.64 frames. , ppl: 10.08955842494655], batch size: 70
+2022-12-11 06:39:01,492 INFO [train.py:421] (4/8) Epoch 4, batch 28000, loss[loss=2.542, over 980.00 frames. , ppl: 12.706906281125498] tot_loss[loss=2.311, over 5501446.34 frames. , ppl: 10.081841546706652], batch size: 70
+2022-12-11 06:39:01,492 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:39:02,238 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.969008347954837
+2022-12-11 06:40:41,649 INFO [train.py:421] (4/8) Epoch 4, batch 28200, loss[loss=2.829, over 630.00 frames. , ppl: 16.921401455165245] tot_loss[loss=2.311, over 5502891.34 frames. , ppl: 10.081540497682225], batch size: 70
+2022-12-11 06:42:23,115 INFO [train.py:421] (4/8) Epoch 4, batch 28400, loss[loss=2.617, over 630.00 frames. , ppl: 13.696825008202289] tot_loss[loss=2.312, over 5486697.40 frames. , ppl: 10.090827013196451], batch size: 70
+2022-12-11 06:44:01,476 INFO [train.py:421] (4/8) Epoch 4, batch 28600, loss[loss=2.232, over 13230.00 frames. , ppl: 9.321667836468976] tot_loss[loss=2.312, over 5471754.23 frames. , ppl: 10.096384073598895], batch size: 70
+2022-12-11 06:45:43,568 INFO [train.py:421] (4/8) Epoch 4, batch 28800, loss[loss=2.46, over 980.00 frames. , ppl: 11.6995966707313] tot_loss[loss=2.313, over 5447216.59 frames. , ppl: 10.1064156093189], batch size: 70
+2022-12-11 06:47:27,939 INFO [train.py:421] (4/8) Epoch 4, batch 29000, loss[loss=2.512, over 1120.00 frames. , ppl: 12.32438323678948] tot_loss[loss=2.313, over 5431023.23 frames. , ppl: 10.106110250066155], batch size: 70
+2022-12-11 06:47:27,939 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:47:28,683 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.302, over 211138.00 frames. , ppl: 9.996510942600489
+2022-12-11 06:49:06,727 INFO [train.py:421] (4/8) Epoch 4, batch 29200, loss[loss=2.344, over 3360.00 frames. , ppl: 10.423313742212772] tot_loss[loss=2.314, over 5415520.71 frames. , ppl: 10.116110781173038], batch size: 70
+2022-12-11 06:50:43,147 INFO [train.py:421] (4/8) Epoch 4, batch 29400, loss[loss=2.303, over 2800.00 frames. , ppl: 10.003177895215991] tot_loss[loss=2.313, over 5422909.52 frames. , ppl: 10.107488229391908], batch size: 70
+2022-12-11 06:52:19,513 INFO [train.py:421] (4/8) Epoch 4, batch 29600, loss[loss=2.375, over 2240.00 frames. , ppl: 10.756240539293687] tot_loss[loss=2.314, over 5404680.86 frames. , ppl: 10.115233146762614], batch size: 70
+2022-12-11 06:53:56,661 INFO [train.py:421] (4/8) Epoch 4, batch 29800, loss[loss=2.164, over 7630.00 frames. , ppl: 8.70240952932318] tot_loss[loss=2.314, over 5388911.00 frames. , ppl: 10.117862336474623], batch size: 70
+2022-12-11 06:55:37,409 INFO [train.py:421] (4/8) Epoch 4, batch 30000, loss[loss=2.734, over 700.00 frames. , ppl: 15.389097731459065] tot_loss[loss=2.314, over 5402813.62 frames. , ppl: 10.114642797847338], batch size: 70
+2022-12-11 06:55:37,409 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 06:55:38,158 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.970644798055698
+2022-12-11 06:57:17,485 INFO [train.py:421] (4/8) Epoch 4, batch 30200, loss[loss=2.394, over 910.00 frames. , ppl: 10.96189031121238] tot_loss[loss=2.314, over 5380181.19 frames. , ppl: 10.119600648186255], batch size: 70
+2022-12-11 06:58:56,929 INFO [train.py:421] (4/8) Epoch 4, batch 30400, loss[loss=2.401, over 1890.00 frames. , ppl: 11.035691175810637] tot_loss[loss=2.315, over 5389457.24 frames. , ppl: 10.120572804818526], batch size: 70
+2022-12-11 07:00:39,366 INFO [train.py:421] (4/8) Epoch 4, batch 30600, loss[loss=2.436, over 1120.00 frames. , ppl: 11.429535326178465] tot_loss[loss=2.314, over 5397800.51 frames. , ppl: 10.112243199002783], batch size: 70
+2022-12-11 07:02:17,072 INFO [train.py:421] (4/8) Epoch 4, batch 30800, loss[loss=2.898, over 700.00 frames. , ppl: 18.146707119050873] tot_loss[loss=2.313, over 5430144.68 frames. , ppl: 10.099906626803259], batch size: 70
+2022-12-11 07:04:02,218 INFO [train.py:421] (4/8) Epoch 4, batch 31000, loss[loss=2.563, over 910.00 frames. , ppl: 12.974621494208057] tot_loss[loss=2.311, over 5485985.20 frames. , ppl: 10.086121710010424], batch size: 70
+2022-12-11 07:04:02,219 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:04:02,984 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.976530269338536
+2022-12-11 07:05:48,983 INFO [train.py:421] (4/8) Epoch 4, batch 31200, loss[loss=2.912, over 630.00 frames. , ppl: 18.395207937622168] tot_loss[loss=2.31, over 5526347.21 frames. , ppl: 10.077980750848724], batch size: 70
+2022-12-11 07:07:34,367 INFO [train.py:421] (4/8) Epoch 4, batch 31400, loss[loss=2.28, over 3010.00 frames. , ppl: 9.777647264652163] tot_loss[loss=2.31, over 5529521.53 frames. , ppl: 10.073005595374866], batch size: 70
+2022-12-11 07:09:13,736 INFO [train.py:421] (4/8) Epoch 4, batch 31600, loss[loss=2.553, over 770.00 frames. , ppl: 12.847781275733524] tot_loss[loss=2.311, over 5480103.84 frames. , ppl: 10.084016579976218], batch size: 70
+2022-12-11 07:10:54,254 INFO [train.py:421] (4/8) Epoch 4, batch 31800, loss[loss=2.331, over 2870.00 frames. , ppl: 10.287653031987224] tot_loss[loss=2.312, over 5446351.20 frames. , ppl: 10.092596058184954], batch size: 70
+2022-12-11 07:12:34,096 INFO [train.py:421] (4/8) Epoch 4, batch 32000, loss[loss=2.669, over 980.00 frames. , ppl: 14.426770869813495] tot_loss[loss=2.311, over 5466578.92 frames. , ppl: 10.086113554160768], batch size: 70
+2022-12-11 07:12:34,097 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:12:34,843 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.965888179968418
+2022-12-11 07:14:15,365 INFO [train.py:421] (4/8) Epoch 4, batch 32200, loss[loss=2.225, over 4690.00 frames. , ppl: 9.25451368932151] tot_loss[loss=2.311, over 5455481.41 frames. , ppl: 10.0808144314662], batch size: 70
+2022-12-11 07:15:52,080 INFO [train.py:421] (4/8) Epoch 4, batch 32400, loss[loss=2.234, over 4340.00 frames. , ppl: 9.34108351016015] tot_loss[loss=2.311, over 5464271.03 frames. , ppl: 10.08041126840281], batch size: 70
+2022-12-11 07:17:32,125 INFO [train.py:421] (4/8) Epoch 4, batch 32600, loss[loss=2.297, over 1540.00 frames. , ppl: 9.94466974421544] tot_loss[loss=2.31, over 5468556.34 frames. , ppl: 10.07307027349425], batch size: 70
+2022-12-11 07:19:13,615 INFO [train.py:421] (4/8) Epoch 4, batch 32800, loss[loss=2.337, over 2310.00 frames. , ppl: 10.354083621070968] tot_loss[loss=2.309, over 5488632.93 frames. , ppl: 10.066605574428475], batch size: 70
+2022-12-11 07:20:53,530 INFO [train.py:421] (4/8) Epoch 4, batch 33000, loss[loss=2.215, over 3500.00 frames. , ppl: 9.157217907955365] tot_loss[loss=2.309, over 5529691.25 frames. , ppl: 10.061596299419385], batch size: 70
+2022-12-11 07:20:53,530 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:20:54,297 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.301, over 211138.00 frames. , ppl: 9.98494896820417
+2022-12-11 07:22:34,966 INFO [train.py:421] (4/8) Epoch 4, batch 33200, loss[loss=2.354, over 3080.00 frames. , ppl: 10.524171229966512] tot_loss[loss=2.309, over 5502543.27 frames. , ppl: 10.06583840926554], batch size: 70
+2022-12-11 07:24:11,943 INFO [train.py:421] (4/8) Epoch 4, batch 33400, loss[loss=2.357, over 2100.00 frames. , ppl: 10.556367297433468] tot_loss[loss=2.308, over 5512459.42 frames. , ppl: 10.057616304881742], batch size: 70
+2022-12-11 07:25:47,474 INFO [train.py:421] (4/8) Epoch 4, batch 33600, loss[loss=2.609, over 840.00 frames. , ppl: 13.588050026467894] tot_loss[loss=2.31, over 5440777.40 frames. , ppl: 10.07482797369435], batch size: 70
+2022-12-11 07:27:27,208 INFO [train.py:421] (4/8) Epoch 4, batch 33800, loss[loss=2.222, over 3150.00 frames. , ppl: 9.222248903539796] tot_loss[loss=2.309, over 5500933.01 frames. , ppl: 10.060267275279362], batch size: 70
+2022-12-11 07:29:07,748 INFO [train.py:421] (4/8) Epoch 4, batch 34000, loss[loss=2.276, over 5530.00 frames. , ppl: 9.741141986164045] tot_loss[loss=2.309, over 5482002.57 frames. , ppl: 10.065108806839886], batch size: 70
+2022-12-11 07:29:07,749 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:29:08,495 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.963913316067512
+2022-12-11 07:30:52,335 INFO [train.py:421] (4/8) Epoch 4, batch 34200, loss[loss=2.476, over 1120.00 frames. , ppl: 11.89787715741491] tot_loss[loss=2.309, over 5487264.28 frames. , ppl: 10.062542147780784], batch size: 70
+2022-12-11 07:32:31,251 INFO [train.py:421] (4/8) Epoch 4, batch 34400, loss[loss=2.445, over 1050.00 frames. , ppl: 11.531911639656318] tot_loss[loss=2.308, over 5509578.80 frames. , ppl: 10.053358887368715], batch size: 70
+2022-12-11 07:34:12,826 INFO [train.py:421] (4/8) Epoch 4, batch 34600, loss[loss=2.283, over 6090.00 frames. , ppl: 9.808222698820197] tot_loss[loss=2.308, over 5502167.63 frames. , ppl: 10.053831687538658], batch size: 70
+2022-12-11 07:35:50,515 INFO [train.py:421] (4/8) Epoch 4, batch 34800, loss[loss=2.253, over 3570.00 frames. , ppl: 9.514237471648045] tot_loss[loss=2.307, over 5515148.71 frames. , ppl: 10.049007026383011], batch size: 70
+2022-12-11 07:37:31,269 INFO [train.py:421] (4/8) Epoch 4, batch 35000, loss[loss=2.478, over 1260.00 frames. , ppl: 11.920378273185335] tot_loss[loss=2.308, over 5502669.68 frames. , ppl: 10.055943110070052], batch size: 70
+2022-12-11 07:37:31,270 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:37:32,028 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.3, over 211138.00 frames. , ppl: 9.97713864747801
+2022-12-11 07:39:12,241 INFO [train.py:421] (4/8) Epoch 4, batch 35200, loss[loss=2.344, over 1750.00 frames. , ppl: 10.425910320891402] tot_loss[loss=2.308, over 5501535.97 frames. , ppl: 10.052690621811468], batch size: 70
+2022-12-11 07:40:54,303 INFO [train.py:421] (4/8) Epoch 4, batch 35400, loss[loss=2.4, over 1610.00 frames. , ppl: 11.02433984501821] tot_loss[loss=2.308, over 5488330.92 frames. , ppl: 10.053598810562614], batch size: 70
+2022-12-11 07:42:36,895 INFO [train.py:421] (4/8) Epoch 4, batch 35600, loss[loss=2.346, over 2100.00 frames. , ppl: 10.443731130880945] tot_loss[loss=2.307, over 5490212.65 frames. , ppl: 10.048606216015523], batch size: 70
+2022-12-11 07:44:17,399 INFO [train.py:421] (4/8) Epoch 4, batch 35800, loss[loss=2.785, over 630.00 frames. , ppl: 16.19360527099967] tot_loss[loss=2.309, over 5459864.46 frames. , ppl: 10.059365524737023], batch size: 70
+2022-12-11 07:46:03,625 INFO [train.py:421] (4/8) Epoch 4, batch 36000, loss[loss=2.319, over 2520.00 frames. , ppl: 10.164508599227979] tot_loss[loss=2.31, over 5440337.85 frames. , ppl: 10.072292174448435], batch size: 70
+2022-12-11 07:46:03,626 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:46:04,385 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.299, over 211138.00 frames. , ppl: 9.962801429018057
+2022-12-11 07:47:46,413 INFO [train.py:421] (4/8) Epoch 4, batch 36200, loss[loss=2.23, over 5250.00 frames. , ppl: 9.301177425368383] tot_loss[loss=2.31, over 5449577.30 frames. , ppl: 10.073328638379474], batch size: 70
+2022-12-11 07:49:21,365 INFO [train.py:421] (4/8) Epoch 4, batch 36400, loss[loss=2.307, over 2730.00 frames. , ppl: 10.04563854371334] tot_loss[loss=2.308, over 5496043.15 frames. , ppl: 10.055411968936863], batch size: 70
+2022-12-11 07:51:00,542 INFO [train.py:421] (4/8) Epoch 4, batch 36600, loss[loss=2.42, over 2030.00 frames. , ppl: 11.247918546483534] tot_loss[loss=2.308, over 5495420.95 frames. , ppl: 10.054281539530619], batch size: 70
+2022-12-11 07:52:42,134 INFO [train.py:421] (4/8) Epoch 4, batch 36800, loss[loss=2.231, over 1960.00 frames. , ppl: 9.313312369910854] tot_loss[loss=2.307, over 5528906.08 frames. , ppl: 10.041169743995273], batch size: 70
+2022-12-11 07:54:20,219 INFO [train.py:421] (4/8) Epoch 4, batch 37000, loss[loss=2.352, over 2170.00 frames. , ppl: 10.509471736892502] tot_loss[loss=2.307, over 5539872.13 frames. , ppl: 10.039818060149312], batch size: 70
+2022-12-11 07:54:20,220 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 07:54:20,987 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.297, over 211138.00 frames. , ppl: 9.946271967014194
+2022-12-11 07:56:00,690 INFO [train.py:421] (4/8) Epoch 4, batch 37200, loss[loss=2.245, over 3290.00 frames. , ppl: 9.440138985837194] tot_loss[loss=2.306, over 5533842.61 frames. , ppl: 10.034700294992186], batch size: 70
+2022-12-11 07:57:39,427 INFO [train.py:421] (4/8) Epoch 4, batch 37400, loss[loss=3.561, over 420.00 frames. , ppl: 35.19014318635338] tot_loss[loss=2.306, over 5519258.85 frames. , ppl: 10.03492474090983], batch size: 70
+2022-12-11 07:59:17,919 INFO [train.py:421] (4/8) Epoch 4, batch 37600, loss[loss=2.23, over 4340.00 frames. , ppl: 9.304513239138519] tot_loss[loss=2.306, over 5518574.96 frames. , ppl: 10.030579147259292], batch size: 70
+2022-12-11 08:00:55,523 INFO [train.py:421] (4/8) Epoch 4, batch 37800, loss[loss=2.375, over 1820.00 frames. , ppl: 10.75623224275683] tot_loss[loss=2.306, over 5492020.70 frames. , ppl: 10.035921031724296], batch size: 70
+2022-12-11 08:02:37,238 INFO [train.py:421] (4/8) Epoch 4, batch 38000, loss[loss=2.196, over 3290.00 frames. , ppl: 8.990021769408198] tot_loss[loss=2.306, over 5513059.61 frames. , ppl: 10.033363525512812], batch size: 70
+2022-12-11 08:02:37,239 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:02:37,986 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.297, over 211138.00 frames. , ppl: 9.945532988975891
+2022-12-11 08:04:17,113 INFO [train.py:421] (4/8) Epoch 4, batch 38200, loss[loss=2.359, over 2870.00 frames. , ppl: 10.58164960883981] tot_loss[loss=2.306, over 5502288.57 frames. , ppl: 10.034553558470076], batch size: 70
+2022-12-11 08:05:57,529 INFO [train.py:421] (4/8) Epoch 4, batch 38400, loss[loss=2.202, over 5670.00 frames. , ppl: 9.046549041836988] tot_loss[loss=2.306, over 5499821.04 frames. , ppl: 10.038705951549138], batch size: 70
+2022-12-11 08:07:37,178 INFO [train.py:421] (4/8) Epoch 4, batch 38600, loss[loss=2.733, over 700.00 frames. , ppl: 15.379662243401073] tot_loss[loss=2.308, over 5458277.25 frames. , ppl: 10.050961338578377], batch size: 70
+2022-12-11 08:09:17,878 INFO [train.py:421] (4/8) Epoch 4, batch 38800, loss[loss=2.208, over 5670.00 frames. , ppl: 9.101473548441062] tot_loss[loss=2.307, over 5511357.01 frames. , ppl: 10.045011541697303], batch size: 70
+2022-12-11 08:10:56,256 INFO [train.py:421] (4/8) Epoch 4, batch 39000, loss[loss=2.836, over 700.00 frames. , ppl: 17.040949560954534] tot_loss[loss=2.306, over 5552409.78 frames. , ppl: 10.03399925725461], batch size: 70
+2022-12-11 08:10:56,256 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:10:57,005 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.298, over 211138.00 frames. , ppl: 9.949938227560008
+2022-12-11 08:12:35,868 INFO [train.py:421] (4/8) Epoch 4, batch 39200, loss[loss=2.569, over 1050.00 frames. , ppl: 13.047706238961668] tot_loss[loss=2.305, over 5580089.69 frames. , ppl: 10.029161136309583], batch size: 70
+2022-12-11 08:14:16,246 INFO [train.py:421] (4/8) Epoch 4, batch 39400, loss[loss=2.644, over 980.00 frames. , ppl: 14.067018698042183] tot_loss[loss=2.306, over 5585489.58 frames. , ppl: 10.034206725842095], batch size: 70
+2022-12-11 08:15:56,341 INFO [train.py:421] (4/8) Epoch 4, batch 39600, loss[loss=2.715, over 630.00 frames. , ppl: 15.107784710263081] tot_loss[loss=2.306, over 5581263.71 frames. , ppl: 10.034892834550424], batch size: 70
+2022-12-11 08:17:35,780 INFO [train.py:421] (4/8) Epoch 4, batch 39800, loss[loss=2.22, over 4200.00 frames. , ppl: 9.208375655778656] tot_loss[loss=2.306, over 5563317.54 frames. , ppl: 10.035635156682162], batch size: 70
+2022-12-11 08:19:17,000 INFO [train.py:421] (4/8) Epoch 4, batch 40000, loss[loss=2.701, over 770.00 frames. , ppl: 14.888674743173851] tot_loss[loss=2.306, over 5571751.07 frames. , ppl: 10.036674358449421], batch size: 70
+2022-12-11 08:19:17,000 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:19:17,760 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.93333160275137
+2022-12-11 08:20:58,979 INFO [train.py:421] (4/8) Epoch 4, batch 40200, loss[loss=2.349, over 3290.00 frames. , ppl: 10.470897226881032] tot_loss[loss=2.308, over 5529350.92 frames. , ppl: 10.049575984907067], batch size: 70
+2022-12-11 08:22:39,534 INFO [train.py:421] (4/8) Epoch 4, batch 40400, loss[loss=2.31, over 2870.00 frames. , ppl: 10.072941134097835] tot_loss[loss=2.309, over 5472901.98 frames. , ppl: 10.061639950551436], batch size: 70
+2022-12-11 08:24:20,707 INFO [train.py:421] (4/8) Epoch 4, batch 40600, loss[loss=2.502, over 1190.00 frames. , ppl: 12.21083966782584] tot_loss[loss=2.309, over 5479285.74 frames. , ppl: 10.06618279848468], batch size: 70
+2022-12-11 08:26:03,834 INFO [train.py:421] (4/8) Epoch 4, batch 40800, loss[loss=2.665, over 840.00 frames. , ppl: 14.36746263817432] tot_loss[loss=2.31, over 5479742.49 frames. , ppl: 10.070012057655914], batch size: 70
+2022-12-11 08:27:44,150 INFO [train.py:421] (4/8) Epoch 4, batch 41000, loss[loss=3.515, over 420.00 frames. , ppl: 33.625635095879126] tot_loss[loss=2.31, over 5455186.28 frames. , ppl: 10.075443509433946], batch size: 70
+2022-12-11 08:27:44,151 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:27:44,908 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.931032464771791
+2022-12-11 08:29:23,713 INFO [train.py:421] (4/8) Epoch 4, batch 41200, loss[loss=2.374, over 3010.00 frames. , ppl: 10.735791648977033] tot_loss[loss=2.312, over 5396135.86 frames. , ppl: 10.093305091215253], batch size: 70
+2022-12-11 08:31:02,554 INFO [train.py:421] (4/8) Epoch 4, batch 41400, loss[loss=2.258, over 4550.00 frames. , ppl: 9.562789414714452] tot_loss[loss=2.312, over 5404316.61 frames. , ppl: 10.092332725727207], batch size: 70
+2022-12-11 08:32:46,802 INFO [train.py:421] (4/8) Epoch 4, batch 41600, loss[loss=2.187, over 11760.00 frames. , ppl: 8.905093522200819] tot_loss[loss=2.311, over 5415860.27 frames. , ppl: 10.086385184838912], batch size: 70
+2022-12-11 08:34:26,753 INFO [train.py:421] (4/8) Epoch 4, batch 41800, loss[loss=2.158, over 6510.00 frames. , ppl: 8.652183786534948] tot_loss[loss=2.311, over 5403614.91 frames. , ppl: 10.08251362997387], batch size: 70
+2022-12-11 08:36:06,525 INFO [train.py:421] (4/8) Epoch 4, batch 42000, loss[loss=2.225, over 3500.00 frames. , ppl: 9.257635409597182] tot_loss[loss=2.309, over 5437818.51 frames. , ppl: 10.067776899837959], batch size: 70
+2022-12-11 08:36:06,526 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:36:07,272 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.298, over 211138.00 frames. , ppl: 9.952867789688982
+2022-12-11 08:37:48,692 INFO [train.py:421] (4/8) Epoch 4, batch 42200, loss[loss=3.619, over 420.00 frames. , ppl: 37.29216061460649] tot_loss[loss=2.308, over 5487619.58 frames. , ppl: 10.057614545432466], batch size: 70
+2022-12-11 08:39:25,695 INFO [train.py:421] (4/8) Epoch 4, batch 42400, loss[loss=2.319, over 3220.00 frames. , ppl: 10.166685129271244] tot_loss[loss=2.308, over 5473592.05 frames. , ppl: 10.051263473242859], batch size: 70
+2022-12-11 08:41:06,767 INFO [train.py:421] (4/8) Epoch 4, batch 42600, loss[loss=2.693, over 700.00 frames. , ppl: 14.775890410876851] tot_loss[loss=2.308, over 5477445.29 frames. , ppl: 10.050498521456657], batch size: 70
+2022-12-11 08:42:48,004 INFO [train.py:421] (4/8) Epoch 4, batch 42800, loss[loss=2.211, over 7070.00 frames. , ppl: 9.124273946225072] tot_loss[loss=2.308, over 5484261.49 frames. , ppl: 10.055596989312384], batch size: 70
+2022-12-11 08:44:22,384 INFO [train.py:421] (4/8) Epoch 4, batch 43000, loss[loss=2.268, over 2730.00 frames. , ppl: 9.660421226109632] tot_loss[loss=2.308, over 5491514.72 frames. , ppl: 10.053474670171042], batch size: 70
+2022-12-11 08:44:22,385 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:44:23,129 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.932262819602796
+2022-12-11 08:46:01,989 INFO [train.py:421] (4/8) Epoch 4, batch 43200, loss[loss=2.338, over 2730.00 frames. , ppl: 10.364665936930495] tot_loss[loss=2.308, over 5517889.22 frames. , ppl: 10.050134349536942], batch size: 70
+2022-12-11 08:47:41,483 INFO [train.py:421] (4/8) Epoch 4, batch 43400, loss[loss=2.307, over 2520.00 frames. , ppl: 10.044412643739761] tot_loss[loss=2.309, over 5477659.21 frames. , ppl: 10.061569009548462], batch size: 70
+2022-12-11 08:49:23,891 INFO [train.py:421] (4/8) Epoch 4, batch 43600, loss[loss=2.218, over 2730.00 frames. , ppl: 9.187125739401036] tot_loss[loss=2.31, over 5442677.75 frames. , ppl: 10.071777642067435], batch size: 70
+2022-12-11 08:51:04,765 INFO [train.py:421] (4/8) Epoch 4, batch 43800, loss[loss=2.286, over 3920.00 frames. , ppl: 9.833201900896432] tot_loss[loss=2.309, over 5449441.28 frames. , ppl: 10.066873424997981], batch size: 70
+2022-12-11 08:52:41,190 INFO [train.py:421] (4/8) Epoch 4, batch 44000, loss[loss=2.466, over 1680.00 frames. , ppl: 11.773642277722015] tot_loss[loss=2.31, over 5416227.91 frames. , ppl: 10.074537389485286], batch size: 70
+2022-12-11 08:52:41,191 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 08:52:41,937 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.938447770769514
+2022-12-11 08:54:20,560 INFO [train.py:421] (4/8) Epoch 4, batch 44200, loss[loss=2.5, over 1960.00 frames. , ppl: 12.176562096658559] tot_loss[loss=2.309, over 5458644.67 frames. , ppl: 10.06570311225846], batch size: 70
+2022-12-11 08:55:59,749 INFO [train.py:421] (4/8) Epoch 4, batch 44400, loss[loss=2.98, over 630.00 frames. , ppl: 19.69501335749835] tot_loss[loss=2.31, over 5424867.77 frames. , ppl: 10.076344111482078], batch size: 70
+2022-12-11 08:57:40,790 INFO [train.py:421] (4/8) Epoch 4, batch 44600, loss[loss=2.291, over 3850.00 frames. , ppl: 9.881281880926096] tot_loss[loss=2.31, over 5417877.98 frames. , ppl: 10.074777002825055], batch size: 70
+2022-12-11 08:59:24,320 INFO [train.py:421] (4/8) Epoch 4, batch 44800, loss[loss=2.33, over 2870.00 frames. , ppl: 10.28206419583119] tot_loss[loss=2.309, over 5432803.95 frames. , ppl: 10.068057831270329], batch size: 70
+2022-12-11 09:01:09,940 INFO [train.py:421] (4/8) Epoch 4, batch 45000, loss[loss=2.281, over 6580.00 frames. , ppl: 9.78568129251192] tot_loss[loss=2.308, over 5475916.67 frames. , ppl: 10.058938492645112], batch size: 70
+2022-12-11 09:01:09,941 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:01:10,691 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.909560376874897
+2022-12-11 09:02:55,724 INFO [train.py:421] (4/8) Epoch 4, batch 45200, loss[loss=2.233, over 6090.00 frames. , ppl: 9.330351787283861] tot_loss[loss=2.308, over 5490938.54 frames. , ppl: 10.052665070942036], batch size: 70
+2022-12-11 09:04:33,464 INFO [train.py:421] (4/8) Epoch 4, batch 45400, loss[loss=2.209, over 7700.00 frames. , ppl: 9.103010330488349] tot_loss[loss=2.309, over 5462710.25 frames. , ppl: 10.064177529355016], batch size: 70
+2022-12-11 09:06:16,158 INFO [train.py:421] (4/8) Epoch 4, batch 45600, loss[loss=2.526, over 980.00 frames. , ppl: 12.497743034812217] tot_loss[loss=2.308, over 5505823.79 frames. , ppl: 10.055245993328462], batch size: 70
+2022-12-11 09:07:57,231 INFO [train.py:421] (4/8) Epoch 4, batch 45800, loss[loss=3.391, over 490.00 frames. , ppl: 29.707270114716135] tot_loss[loss=2.307, over 5527184.08 frames. , ppl: 10.048416846741524], batch size: 70
+2022-12-11 09:09:38,445 INFO [train.py:421] (4/8) Epoch 4, batch 46000, loss[loss=2.261, over 4690.00 frames. , ppl: 9.590639588611852] tot_loss[loss=2.307, over 5561368.76 frames. , ppl: 10.041479404648424], batch size: 70
+2022-12-11 09:09:38,446 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:09:39,192 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.295, over 211138.00 frames. , ppl: 9.921327269150712
+2022-12-11 09:11:18,492 INFO [train.py:421] (4/8) Epoch 4, batch 46200, loss[loss=2.298, over 3150.00 frames. , ppl: 9.949902767556274] tot_loss[loss=2.308, over 5573280.72 frames. , ppl: 10.049742175038162], batch size: 70
+2022-12-11 09:12:54,785 INFO [train.py:421] (4/8) Epoch 4, batch 46400, loss[loss=2.494, over 1330.00 frames. , ppl: 12.113822244931033] tot_loss[loss=2.308, over 5559203.39 frames. , ppl: 10.052125683333692], batch size: 70
+2022-12-11 09:14:38,861 INFO [train.py:421] (4/8) Epoch 4, batch 46600, loss[loss=2.364, over 2170.00 frames. , ppl: 10.62888312688127] tot_loss[loss=2.308, over 5583550.33 frames. , ppl: 10.05069197878164], batch size: 70
+2022-12-11 09:16:20,806 INFO [train.py:421] (4/8) Epoch 4, batch 46800, loss[loss=2.872, over 560.00 frames. , ppl: 17.674494795016873] tot_loss[loss=2.306, over 5604692.36 frames. , ppl: 10.035976310749387], batch size: 70
+2022-12-11 09:18:01,896 INFO [train.py:421] (4/8) Epoch 4, batch 47000, loss[loss=2.167, over 5530.00 frames. , ppl: 8.735177518022773] tot_loss[loss=2.306, over 5630026.84 frames. , ppl: 10.030243745490385], batch size: 70
+2022-12-11 09:18:01,897 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:18:02,643 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.912724529894838
+2022-12-11 09:19:43,173 INFO [train.py:421] (4/8) Epoch 4, batch 47200, loss[loss=2.617, over 700.00 frames. , ppl: 13.696021156486632] tot_loss[loss=2.306, over 5622440.41 frames. , ppl: 10.032806877059741], batch size: 70
+2022-12-11 09:21:18,453 INFO [train.py:421] (4/8) Epoch 4, batch 47400, loss[loss=2.249, over 3010.00 frames. , ppl: 9.478794817165292] tot_loss[loss=2.306, over 5621260.48 frames. , ppl: 10.03038666220357], batch size: 70
+2022-12-11 09:22:56,912 INFO [train.py:421] (4/8) Epoch 4, batch 47600, loss[loss=2.418, over 1050.00 frames. , ppl: 11.219650088158906] tot_loss[loss=2.306, over 5613012.78 frames. , ppl: 10.035448877256796], batch size: 70
+2022-12-11 09:24:37,051 INFO [train.py:421] (4/8) Epoch 4, batch 47800, loss[loss=2.291, over 2240.00 frames. , ppl: 9.887245535169944] tot_loss[loss=2.307, over 5607083.87 frames. , ppl: 10.046080697795267], batch size: 70
+2022-12-11 09:26:19,532 INFO [train.py:421] (4/8) Epoch 4, batch 48000, loss[loss=2.694, over 840.00 frames. , ppl: 14.786877293934577] tot_loss[loss=2.307, over 5588410.66 frames. , ppl: 10.048379222673464], batch size: 70
+2022-12-11 09:26:19,533 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:26:20,321 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.295, over 211138.00 frames. , ppl: 9.92101743515498
+2022-12-11 09:28:03,658 INFO [train.py:421] (4/8) Epoch 4, batch 48200, loss[loss=2.198, over 12180.00 frames. , ppl: 9.006635343588085] tot_loss[loss=2.306, over 5602487.54 frames. , ppl: 10.03747176693339], batch size: 70
+2022-12-11 09:29:44,757 INFO [train.py:421] (4/8) Epoch 4, batch 48400, loss[loss=2.446, over 770.00 frames. , ppl: 11.54748025525901] tot_loss[loss=2.307, over 5602013.85 frames. , ppl: 10.040850497239346], batch size: 70
+2022-12-11 09:31:30,699 INFO [train.py:421] (4/8) Epoch 4, batch 48600, loss[loss=2.451, over 1330.00 frames. , ppl: 11.602369240564903] tot_loss[loss=2.307, over 5578993.75 frames. , ppl: 10.043750382924298], batch size: 70
+2022-12-11 09:33:10,103 INFO [train.py:421] (4/8) Epoch 4, batch 48800, loss[loss=2.696, over 630.00 frames. , ppl: 14.815909578183033] tot_loss[loss=2.306, over 5587301.02 frames. , ppl: 10.036110618497823], batch size: 70
+2022-12-11 09:34:51,422 INFO [train.py:421] (4/8) Epoch 4, batch 49000, loss[loss=2.17, over 9100.00 frames. , ppl: 8.757893054645018] tot_loss[loss=2.306, over 5599123.89 frames. , ppl: 10.030737473324569], batch size: 70
+2022-12-11 09:34:51,423 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:34:52,168 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.912736267159561
+2022-12-11 09:36:32,382 INFO [train.py:421] (4/8) Epoch 4, batch 49200, loss[loss=2.468, over 1820.00 frames. , ppl: 11.799786407971881] tot_loss[loss=2.306, over 5588844.29 frames. , ppl: 10.030528228569098], batch size: 70
+2022-12-11 09:38:12,932 INFO [train.py:421] (4/8) Epoch 4, batch 49400, loss[loss=2.226, over 6300.00 frames. , ppl: 9.263764421706034] tot_loss[loss=2.306, over 5567962.74 frames. , ppl: 10.032731875871045], batch size: 70
+2022-12-11 09:39:55,533 INFO [train.py:421] (4/8) Epoch 4, batch 49600, loss[loss=2.16, over 4970.00 frames. , ppl: 8.66877003440251] tot_loss[loss=2.306, over 5549407.99 frames. , ppl: 10.031303231071707], batch size: 70
+2022-12-11 09:41:34,985 INFO [train.py:421] (4/8) Epoch 4, batch 49800, loss[loss=2.393, over 1050.00 frames. , ppl: 10.94918954083952] tot_loss[loss=2.305, over 5567597.93 frames. , ppl: 10.020628403912625], batch size: 70
+2022-12-11 09:43:18,289 INFO [train.py:421] (4/8) Epoch 4, batch 50000, loss[loss=2.215, over 13650.00 frames. , ppl: 9.156844488499786] tot_loss[loss=2.304, over 5577792.39 frames. , ppl: 10.01857685527227], batch size: 70
+2022-12-11 09:43:18,289 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:43:19,051 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.911439383482614
+2022-12-11 09:44:57,840 INFO [train.py:421] (4/8) Epoch 4, batch 50200, loss[loss=2.428, over 2030.00 frames. , ppl: 11.339000473059201] tot_loss[loss=2.306, over 5523211.62 frames. , ppl: 10.037542342695087], batch size: 70
+2022-12-11 09:46:40,421 INFO [train.py:421] (4/8) Epoch 4, batch 50400, loss[loss=2.443, over 1470.00 frames. , ppl: 11.507356434141757] tot_loss[loss=2.307, over 5475458.88 frames. , ppl: 10.046419049697624], batch size: 70
+2022-12-11 09:48:16,333 INFO [train.py:421] (4/8) Epoch 4, batch 50600, loss[loss=2.56, over 770.00 frames. , ppl: 12.941408053764814] tot_loss[loss=2.308, over 5482471.11 frames. , ppl: 10.055549548816048], batch size: 70
+2022-12-11 09:49:58,423 INFO [train.py:421] (4/8) Epoch 4, batch 50800, loss[loss=2.326, over 2800.00 frames. , ppl: 10.233985682083448] tot_loss[loss=2.307, over 5525439.14 frames. , ppl: 10.045713553570188], batch size: 70
+2022-12-11 09:51:36,250 INFO [train.py:421] (4/8) Epoch 4, batch 51000, loss[loss=2.363, over 2660.00 frames. , ppl: 10.619083637083401] tot_loss[loss=2.307, over 5515520.18 frames. , ppl: 10.045099654787936], batch size: 70
+2022-12-11 09:51:36,250 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:51:36,999 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.909437175772835
+2022-12-11 09:53:14,568 INFO [train.py:421] (4/8) Epoch 4, batch 51200, loss[loss=2.472, over 980.00 frames. , ppl: 11.851048497486273] tot_loss[loss=2.309, over 5454321.59 frames. , ppl: 10.059737095991322], batch size: 70
+2022-12-11 09:54:54,289 INFO [train.py:421] (4/8) Epoch 4, batch 51400, loss[loss=2.575, over 1120.00 frames. , ppl: 13.125193051038769] tot_loss[loss=2.308, over 5466374.41 frames. , ppl: 10.057821620830625], batch size: 70
+2022-12-11 09:56:32,680 INFO [train.py:421] (4/8) Epoch 4, batch 51600, loss[loss=2.298, over 4130.00 frames. , ppl: 9.95460765733847] tot_loss[loss=2.309, over 5453732.92 frames. , ppl: 10.060366468125112], batch size: 70
+2022-12-11 09:58:13,022 INFO [train.py:421] (4/8) Epoch 4, batch 51800, loss[loss=2.324, over 2310.00 frames. , ppl: 10.217669824105586] tot_loss[loss=2.309, over 5417707.62 frames. , ppl: 10.067688475989872], batch size: 70
+2022-12-11 09:59:53,302 INFO [train.py:421] (4/8) Epoch 4, batch 52000, loss[loss=2.351, over 2380.00 frames. , ppl: 10.497582567770825] tot_loss[loss=2.308, over 5457005.29 frames. , ppl: 10.05520210666412], batch size: 70
+2022-12-11 09:59:53,303 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 09:59:54,056 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.914090547381605
+2022-12-11 10:01:33,675 INFO [train.py:421] (4/8) Epoch 4, batch 52200, loss[loss=2.428, over 1540.00 frames. , ppl: 11.334252506693899] tot_loss[loss=2.307, over 5493202.32 frames. , ppl: 10.04921764193998], batch size: 70
+2022-12-11 10:03:14,831 INFO [train.py:421] (4/8) Epoch 4, batch 52400, loss[loss=2.202, over 5040.00 frames. , ppl: 9.043044932357205] tot_loss[loss=2.308, over 5460270.75 frames. , ppl: 10.059002186692478], batch size: 70
+2022-12-11 10:04:58,011 INFO [train.py:421] (4/8) Epoch 4, batch 52600, loss[loss=2.252, over 5670.00 frames. , ppl: 9.504516560951455] tot_loss[loss=2.31, over 5417308.36 frames. , ppl: 10.071643106335365], batch size: 70
+2022-12-11 10:06:38,642 INFO [train.py:421] (4/8) Epoch 4, batch 52800, loss[loss=2.228, over 2660.00 frames. , ppl: 9.2812081997328] tot_loss[loss=2.31, over 5427232.49 frames. , ppl: 10.07862200617619], batch size: 70
+2022-12-11 10:08:18,150 INFO [train.py:421] (4/8) Epoch 4, batch 53000, loss[loss=2.151, over 5810.00 frames. , ppl: 8.591617037809296] tot_loss[loss=2.31, over 5445324.85 frames. , ppl: 10.071732314714703], batch size: 70
+2022-12-11 10:08:18,150 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:08:18,896 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.296, over 211138.00 frames. , ppl: 9.929699384943623
+2022-12-11 10:09:58,043 INFO [train.py:421] (4/8) Epoch 4, batch 53200, loss[loss=2.6, over 770.00 frames. , ppl: 13.460412980223158] tot_loss[loss=2.31, over 5423873.09 frames. , ppl: 10.078414903143502], batch size: 70
+2022-12-11 10:11:38,308 INFO [train.py:421] (4/8) Epoch 4, batch 53400, loss[loss=2.221, over 2590.00 frames. , ppl: 9.211957286816466] tot_loss[loss=2.31, over 5415259.52 frames. , ppl: 10.073830385515734], batch size: 70
+2022-12-11 10:13:16,442 INFO [train.py:421] (4/8) Epoch 4, batch 53600, loss[loss=2.22, over 4340.00 frames. , ppl: 9.204154957523363] tot_loss[loss=2.31, over 5419241.92 frames. , ppl: 10.072392277034542], batch size: 70
+2022-12-11 10:14:54,533 INFO [train.py:421] (4/8) Epoch 4, batch 53800, loss[loss=2.275, over 2870.00 frames. , ppl: 9.727303357725448] tot_loss[loss=2.309, over 5442042.81 frames. , ppl: 10.068055701502079], batch size: 70
+2022-12-11 10:16:36,543 INFO [train.py:421] (4/8) Epoch 4, batch 54000, loss[loss=2.415, over 1190.00 frames. , ppl: 11.185795837420777] tot_loss[loss=2.309, over 5489506.41 frames. , ppl: 10.060937166560597], batch size: 70
+2022-12-11 10:16:36,543 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:16:37,305 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.903191128032574
+2022-12-11 10:18:18,565 INFO [train.py:421] (4/8) Epoch 4, batch 54200, loss[loss=2.421, over 1540.00 frames. , ppl: 11.257195319022102] tot_loss[loss=2.308, over 5524400.34 frames. , ppl: 10.057018270600828], batch size: 70
+2022-12-11 10:19:56,530 INFO [train.py:421] (4/8) Epoch 4, batch 54400, loss[loss=2.194, over 7420.00 frames. , ppl: 8.972790244936839] tot_loss[loss=2.308, over 5532748.30 frames. , ppl: 10.055402180190944], batch size: 70
+2022-12-11 10:21:35,656 INFO [train.py:421] (4/8) Epoch 4, batch 54600, loss[loss=2.509, over 1400.00 frames. , ppl: 12.295292714382382] tot_loss[loss=2.308, over 5516438.22 frames. , ppl: 10.057971033577632], batch size: 70
+2022-12-11 10:23:17,361 INFO [train.py:421] (4/8) Epoch 4, batch 54800, loss[loss=2.369, over 2240.00 frames. , ppl: 10.686458251381024] tot_loss[loss=2.309, over 5467770.45 frames. , ppl: 10.069218533555384], batch size: 70
+2022-12-11 10:24:58,184 INFO [train.py:421] (4/8) Epoch 4, batch 55000, loss[loss=3.641, over 420.00 frames. , ppl: 38.1148982633827] tot_loss[loss=2.31, over 5446369.35 frames. , ppl: 10.075788992917996], batch size: 70
+2022-12-11 10:24:58,185 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:24:58,943 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.91692148777633
+2022-12-11 10:26:38,703 INFO [train.py:421] (4/8) Epoch 4, batch 55200, loss[loss=2.247, over 4830.00 frames. , ppl: 9.458293575900989] tot_loss[loss=2.311, over 5438595.11 frames. , ppl: 10.080282906687913], batch size: 70
+2022-12-11 10:28:18,394 INFO [train.py:421] (4/8) Epoch 4, batch 55400, loss[loss=2.555, over 1400.00 frames. , ppl: 12.865729824045582] tot_loss[loss=2.31, over 5461438.70 frames. , ppl: 10.077589772087077], batch size: 70
+2022-12-11 10:29:59,489 INFO [train.py:421] (4/8) Epoch 4, batch 55600, loss[loss=3.827, over 420.00 frames. , ppl: 45.90971373708212] tot_loss[loss=2.309, over 5502485.72 frames. , ppl: 10.064311458439263], batch size: 70
+2022-12-11 10:31:38,226 INFO [train.py:421] (4/8) Epoch 4, batch 55800, loss[loss=2.516, over 1050.00 frames. , ppl: 12.381385760702937] tot_loss[loss=2.31, over 5470644.50 frames. , ppl: 10.076573058339159], batch size: 70
+2022-12-11 10:33:18,378 INFO [train.py:421] (4/8) Epoch 4, batch 56000, loss[loss=2.432, over 1820.00 frames. , ppl: 11.378894509876632] tot_loss[loss=2.311, over 5445592.33 frames. , ppl: 10.081852011678807], batch size: 70
+2022-12-11 10:33:18,378 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:33:19,123 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.91445299177917
+2022-12-11 10:34:53,562 INFO [train.py:421] (4/8) Epoch 4, batch 56200, loss[loss=2.241, over 5110.00 frames. , ppl: 9.401941859856823] tot_loss[loss=2.31, over 5438877.41 frames. , ppl: 10.076359495780146], batch size: 70
+2022-12-11 10:36:32,181 INFO [train.py:421] (4/8) Epoch 4, batch 56400, loss[loss=2.273, over 5250.00 frames. , ppl: 9.710859463953676] tot_loss[loss=2.31, over 5437126.93 frames. , ppl: 10.070902745548327], batch size: 70
+2022-12-11 10:38:13,265 INFO [train.py:421] (4/8) Epoch 4, batch 56600, loss[loss=2.57, over 840.00 frames. , ppl: 13.064173393012796] tot_loss[loss=2.309, over 5452599.61 frames. , ppl: 10.06659483021864], batch size: 70
+2022-12-11 10:39:51,813 INFO [train.py:421] (4/8) Epoch 4, batch 56800, loss[loss=2.34, over 2940.00 frames. , ppl: 10.379687023934656] tot_loss[loss=2.311, over 5394394.61 frames. , ppl: 10.08328283098794], batch size: 70
+2022-12-11 10:41:31,980 INFO [train.py:421] (4/8) Epoch 4, batch 57000, loss[loss=2.312, over 2590.00 frames. , ppl: 10.0925122062356] tot_loss[loss=2.31, over 5411663.81 frames. , ppl: 10.072069934321677], batch size: 70
+2022-12-11 10:41:31,981 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:41:32,726 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.90575504999001
+2022-12-11 10:43:12,819 INFO [train.py:421] (4/8) Epoch 4, batch 57200, loss[loss=2.487, over 910.00 frames. , ppl: 12.021294035613021] tot_loss[loss=2.31, over 5434364.65 frames. , ppl: 10.077628667842864], batch size: 70
+2022-12-11 10:44:55,967 INFO [train.py:421] (4/8) Epoch 4, batch 57400, loss[loss=2.321, over 3150.00 frames. , ppl: 10.185594237554191] tot_loss[loss=2.308, over 5492371.20 frames. , ppl: 10.055625271635904], batch size: 70
+2022-12-11 10:46:33,832 INFO [train.py:421] (4/8) Epoch 4, batch 57600, loss[loss=2.291, over 3080.00 frames. , ppl: 9.889527030915376] tot_loss[loss=2.309, over 5494674.15 frames. , ppl: 10.063000885058498], batch size: 70
+2022-12-11 10:48:10,497 INFO [train.py:421] (4/8) Epoch 4, batch 57800, loss[loss=2.365, over 3290.00 frames. , ppl: 10.645840168666009] tot_loss[loss=2.31, over 5478344.77 frames. , ppl: 10.07660741548257], batch size: 70
+2022-12-11 10:49:55,353 INFO [train.py:421] (4/8) Epoch 4, batch 58000, loss[loss=2.375, over 1610.00 frames. , ppl: 10.745989821207244] tot_loss[loss=2.308, over 5547530.55 frames. , ppl: 10.04957197511793], batch size: 70
+2022-12-11 10:49:55,353 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:49:56,110 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.915101609561932
+2022-12-11 10:51:34,570 INFO [train.py:421] (4/8) Epoch 4, batch 58200, loss[loss=2.332, over 4130.00 frames. , ppl: 10.296996522460214] tot_loss[loss=2.307, over 5565568.00 frames. , ppl: 10.046234638825956], batch size: 70
+2022-12-11 10:53:13,739 INFO [train.py:421] (4/8) Epoch 4, batch 58400, loss[loss=2.49, over 1820.00 frames. , ppl: 12.065210293240547] tot_loss[loss=2.306, over 5610600.37 frames. , ppl: 10.029688991571723], batch size: 70
+2022-12-11 10:54:51,409 INFO [train.py:421] (4/8) Epoch 4, batch 58600, loss[loss=2.323, over 3710.00 frames. , ppl: 10.20685992824152] tot_loss[loss=2.304, over 5617628.32 frames. , ppl: 10.018557674878009], batch size: 70
+2022-12-11 10:56:31,627 INFO [train.py:421] (4/8) Epoch 4, batch 58800, loss[loss=2.272, over 4340.00 frames. , ppl: 9.69895462573592] tot_loss[loss=2.305, over 5603151.18 frames. , ppl: 10.022636572738358], batch size: 70
+2022-12-11 10:58:12,932 INFO [train.py:421] (4/8) Epoch 4, batch 59000, loss[loss=2.514, over 1050.00 frames. , ppl: 12.349127088270839] tot_loss[loss=2.304, over 5594231.85 frames. , ppl: 10.017191455421939], batch size: 70
+2022-12-11 10:58:12,932 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 10:58:13,677 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.911222274742915
+2022-12-11 10:59:52,976 INFO [train.py:421] (4/8) Epoch 4, batch 59200, loss[loss=2.321, over 1120.00 frames. , ppl: 10.185635436228416] tot_loss[loss=2.304, over 5596154.42 frames. , ppl: 10.009971702582527], batch size: 70
+2022-12-11 11:01:31,036 INFO [train.py:421] (4/8) Epoch 4, batch 59400, loss[loss=2.178, over 7000.00 frames. , ppl: 8.826207727484297] tot_loss[loss=2.306, over 5536338.52 frames. , ppl: 10.030157348108576], batch size: 70
+2022-12-11 11:03:11,584 INFO [train.py:421] (4/8) Epoch 4, batch 59600, loss[loss=2.369, over 1680.00 frames. , ppl: 10.684581645455317] tot_loss[loss=2.306, over 5525580.52 frames. , ppl: 10.033896247412661], batch size: 70
+2022-12-11 11:04:50,759 INFO [train.py:421] (4/8) Epoch 4, batch 59800, loss[loss=2.527, over 1330.00 frames. , ppl: 12.514750244127052] tot_loss[loss=2.305, over 5554220.40 frames. , ppl: 10.021002760872864], batch size: 70
+2022-12-11 11:06:31,810 INFO [train.py:421] (4/8) Epoch 4, batch 60000, loss[loss=2.283, over 4060.00 frames. , ppl: 9.803468495338883] tot_loss[loss=2.305, over 5527777.21 frames. , ppl: 10.024632210803002], batch size: 70
+2022-12-11 11:06:31,811 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:06:32,558 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.908699467886942
+2022-12-11 11:08:12,353 INFO [train.py:421] (4/8) Epoch 4, batch 60200, loss[loss=2.276, over 3080.00 frames. , ppl: 9.737221350786323] tot_loss[loss=2.306, over 5475699.14 frames. , ppl: 10.029859979294526], batch size: 70
+2022-12-11 11:09:54,641 INFO [train.py:421] (4/8) Epoch 4, batch 60400, loss[loss=2.264, over 3150.00 frames. , ppl: 9.62610074360782] tot_loss[loss=2.306, over 5463834.25 frames. , ppl: 10.037916041124301], batch size: 70
+2022-12-11 11:11:33,931 INFO [train.py:421] (4/8) Epoch 4, batch 60600, loss[loss=3.263, over 490.00 frames. , ppl: 26.120728466438486] tot_loss[loss=2.306, over 5459342.92 frames. , ppl: 10.036881825422164], batch size: 70
+2022-12-11 11:13:09,650 INFO [train.py:421] (4/8) Epoch 4, batch 60800, loss[loss=2.795, over 630.00 frames. , ppl: 16.365224933571564] tot_loss[loss=2.307, over 5424535.92 frames. , ppl: 10.043017739276982], batch size: 70
+2022-12-11 11:14:48,151 INFO [train.py:421] (4/8) Epoch 4, batch 61000, loss[loss=2.423, over 1400.00 frames. , ppl: 11.28469761263386] tot_loss[loss=2.307, over 5439719.99 frames. , ppl: 10.043593104643842], batch size: 70
+2022-12-11 11:14:48,152 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:14:48,898 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.292, over 211138.00 frames. , ppl: 9.890057963547763
+2022-12-11 11:16:30,418 INFO [train.py:421] (4/8) Epoch 4, batch 61200, loss[loss=2.185, over 7000.00 frames. , ppl: 8.88660849321089] tot_loss[loss=2.307, over 5453050.19 frames. , ppl: 10.04098625113119], batch size: 70
+2022-12-11 11:18:08,905 INFO [train.py:421] (4/8) Epoch 4, batch 61400, loss[loss=2.45, over 1330.00 frames. , ppl: 11.585815618546375] tot_loss[loss=2.306, over 5468591.59 frames. , ppl: 10.03723464058962], batch size: 70
+2022-12-11 11:19:47,063 INFO [train.py:421] (4/8) Epoch 4, batch 61600, loss[loss=3.04, over 560.00 frames. , ppl: 20.902200304408378] tot_loss[loss=2.306, over 5474511.40 frames. , ppl: 10.030212589833681], batch size: 70
+2022-12-11 11:21:28,011 INFO [train.py:421] (4/8) Epoch 4, batch 61800, loss[loss=2.258, over 7770.00 frames. , ppl: 9.565353623666466] tot_loss[loss=2.305, over 5481939.23 frames. , ppl: 10.026075395092894], batch size: 70
+2022-12-11 11:23:08,281 INFO [train.py:421] (4/8) Epoch 4, batch 62000, loss[loss=2.33, over 2240.00 frames. , ppl: 10.27455772162752] tot_loss[loss=2.307, over 5447415.08 frames. , ppl: 10.04026994085937], batch size: 70
+2022-12-11 11:23:08,282 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:23:09,027 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.916335861485702
+2022-12-11 11:24:49,036 INFO [train.py:421] (4/8) Epoch 4, batch 62200, loss[loss=2.536, over 1050.00 frames. , ppl: 12.625663040167586] tot_loss[loss=2.306, over 5475789.10 frames. , ppl: 10.03067975367699], batch size: 70
+2022-12-11 11:26:31,323 INFO [train.py:421] (4/8) Epoch 4, batch 62400, loss[loss=2.577, over 910.00 frames. , ppl: 13.16251115484472] tot_loss[loss=2.306, over 5510355.39 frames. , ppl: 10.038348415426642], batch size: 70
+2022-12-11 11:28:12,735 INFO [train.py:421] (4/8) Epoch 4, batch 62600, loss[loss=2.811, over 700.00 frames. , ppl: 16.628396828136026] tot_loss[loss=2.304, over 5573021.44 frames. , ppl: 10.01164736718772], batch size: 70
+2022-12-11 11:29:56,769 INFO [train.py:421] (4/8) Epoch 4, batch 62800, loss[loss=2.341, over 1540.00 frames. , ppl: 10.394836868090975] tot_loss[loss=2.304, over 5570572.23 frames. , ppl: 10.01497338129389], batch size: 70
+2022-12-11 11:31:39,087 INFO [train.py:421] (4/8) Epoch 4, batch 63000, loss[loss=2.289, over 1960.00 frames. , ppl: 9.867257363335419] tot_loss[loss=2.305, over 5581228.55 frames. , ppl: 10.022206682674362], batch size: 70
+2022-12-11 11:31:39,088 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:31:39,834 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.912125947821703
+2022-12-11 11:33:18,808 INFO [train.py:421] (4/8) Epoch 4, batch 63200, loss[loss=2.355, over 2450.00 frames. , ppl: 10.53896898836316] tot_loss[loss=2.304, over 5582710.18 frames. , ppl: 10.019166106188338], batch size: 70
+2022-12-11 11:34:57,558 INFO [train.py:421] (4/8) Epoch 4, batch 63400, loss[loss=2.323, over 2730.00 frames. , ppl: 10.20386718271968] tot_loss[loss=2.307, over 5489582.01 frames. , ppl: 10.045174683442331], batch size: 70
+2022-12-11 11:36:40,498 INFO [train.py:421] (4/8) Epoch 4, batch 63600, loss[loss=2.513, over 700.00 frames. , ppl: 12.343307503879553] tot_loss[loss=2.306, over 5529280.62 frames. , ppl: 10.029829786331208], batch size: 70
+2022-12-11 11:38:20,357 INFO [train.py:421] (4/8) Epoch 4, batch 63800, loss[loss=2.448, over 3010.00 frames. , ppl: 11.570182743188104] tot_loss[loss=2.305, over 5522195.01 frames. , ppl: 10.025829766637171], batch size: 70
+2022-12-11 11:39:59,621 INFO [train.py:421] (4/8) Epoch 4, batch 64000, loss[loss=2.317, over 3570.00 frames. , ppl: 10.140660607355844] tot_loss[loss=2.304, over 5530967.25 frames. , ppl: 10.0191247923541], batch size: 70
+2022-12-11 11:39:59,621 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:40:00,404 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.292, over 211138.00 frames. , ppl: 9.899215356849181
+2022-12-11 11:41:39,829 INFO [train.py:421] (4/8) Epoch 4, batch 64200, loss[loss=2.362, over 1960.00 frames. , ppl: 10.614864829170477] tot_loss[loss=2.305, over 5509067.86 frames. , ppl: 10.024967386042565], batch size: 70
+2022-12-11 11:43:17,850 INFO [train.py:421] (4/8) Epoch 4, batch 64400, loss[loss=2.371, over 2730.00 frames. , ppl: 10.707794647326633] tot_loss[loss=2.305, over 5507276.54 frames. , ppl: 10.023472435586505], batch size: 70
+2022-12-11 11:44:57,183 INFO [train.py:421] (4/8) Epoch 4, batch 64600, loss[loss=2.301, over 1540.00 frames. , ppl: 9.983565174768806] tot_loss[loss=2.304, over 5511986.75 frames. , ppl: 10.016848990322702], batch size: 70
+2022-12-11 11:46:37,646 INFO [train.py:421] (4/8) Epoch 4, batch 64800, loss[loss=2.259, over 3430.00 frames. , ppl: 9.5776456027738] tot_loss[loss=2.306, over 5466911.65 frames. , ppl: 10.03250145625726], batch size: 70
+2022-12-11 11:48:19,715 INFO [train.py:421] (4/8) Epoch 4, batch 65000, loss[loss=2.498, over 840.00 frames. , ppl: 12.160532716775728] tot_loss[loss=2.304, over 5509145.21 frames. , ppl: 10.014336832427377], batch size: 70
+2022-12-11 11:48:19,716 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:48:20,473 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.916086357031197
+2022-12-11 11:50:01,971 INFO [train.py:421] (4/8) Epoch 4, batch 65200, loss[loss=2.29, over 2030.00 frames. , ppl: 9.8788823252694] tot_loss[loss=2.302, over 5571685.04 frames. , ppl: 9.995154355354774], batch size: 70
+2022-12-11 11:51:41,797 INFO [train.py:421] (4/8) Epoch 4, batch 65400, loss[loss=2.521, over 770.00 frames. , ppl: 12.440474034704463] tot_loss[loss=2.301, over 5605040.00 frames. , ppl: 9.985574439068772], batch size: 70
+2022-12-11 11:53:19,725 INFO [train.py:421] (4/8) Epoch 4, batch 65600, loss[loss=2.384, over 1120.00 frames. , ppl: 10.850534557546116] tot_loss[loss=2.302, over 5580612.05 frames. , ppl: 9.997780842367007], batch size: 70
+2022-12-11 11:55:00,756 INFO [train.py:421] (4/8) Epoch 4, batch 65800, loss[loss=3.193, over 490.00 frames. , ppl: 24.363736134435975] tot_loss[loss=2.304, over 5548154.03 frames. , ppl: 10.014024551754698], batch size: 70
+2022-12-11 11:56:40,731 INFO [train.py:421] (4/8) Epoch 4, batch 66000, loss[loss=2.29, over 2520.00 frames. , ppl: 9.879254470598822] tot_loss[loss=2.305, over 5527704.81 frames. , ppl: 10.020243925462543], batch size: 70
+2022-12-11 11:56:40,732 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 11:56:41,491 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.919058803417657
+2022-12-11 11:58:22,192 INFO [train.py:421] (4/8) Epoch 4, batch 66200, loss[loss=2.568, over 1260.00 frames. , ppl: 13.034703848105478] tot_loss[loss=2.305, over 5533640.61 frames. , ppl: 10.021311610673608], batch size: 70
+2022-12-11 12:00:04,131 INFO [train.py:421] (4/8) Epoch 4, batch 66400, loss[loss=2.334, over 3500.00 frames. , ppl: 10.322627281066996] tot_loss[loss=2.305, over 5502385.18 frames. , ppl: 10.022971968468623], batch size: 70
+2022-12-11 12:01:41,417 INFO [train.py:421] (4/8) Epoch 4, batch 66600, loss[loss=2.301, over 1890.00 frames. , ppl: 9.981915937497533] tot_loss[loss=2.306, over 5484491.90 frames. , ppl: 10.032330016738708], batch size: 70
+2022-12-11 12:03:22,240 INFO [train.py:421] (4/8) Epoch 4, batch 66800, loss[loss=2.31, over 2450.00 frames. , ppl: 10.071636182603992] tot_loss[loss=2.306, over 5456957.29 frames. , ppl: 10.035283021247201], batch size: 70
+2022-12-11 12:05:01,127 INFO [train.py:421] (4/8) Epoch 4, batch 67000, loss[loss=2.713, over 770.00 frames. , ppl: 15.079914413919209] tot_loss[loss=2.306, over 5467096.71 frames. , ppl: 10.031173330599163], batch size: 70
+2022-12-11 12:05:01,128 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 12:05:01,878 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.293, over 211138.00 frames. , ppl: 9.901741611125681
+2022-12-11 12:06:40,981 INFO [train.py:421] (4/8) Epoch 4, batch 67200, loss[loss=2.211, over 3920.00 frames. , ppl: 9.129157058279413] tot_loss[loss=2.306, over 5466641.54 frames. , ppl: 10.030510088777664], batch size: 70
+2022-12-11 12:08:24,020 INFO [train.py:421] (4/8) Epoch 4, batch 67400, loss[loss=4.016, over 350.00 frames. , ppl: 55.487605304754396] tot_loss[loss=2.304, over 5522656.29 frames. , ppl: 10.014287897707293], batch size: 70
+2022-12-11 12:10:06,522 INFO [train.py:421] (4/8) Epoch 4, batch 67600, loss[loss=2.397, over 1820.00 frames. , ppl: 10.98890064880404] tot_loss[loss=2.304, over 5531609.72 frames. , ppl: 10.0098931362077], batch size: 70
+2022-12-11 12:11:44,333 INFO [train.py:421] (4/8) Epoch 4, batch 67800, loss[loss=2.244, over 3430.00 frames. , ppl: 9.435086077804288] tot_loss[loss=2.305, over 5480755.56 frames. , ppl: 10.020764595565959], batch size: 70
+2022-12-11 12:13:27,419 INFO [train.py:421] (4/8) Epoch 4, batch 68000, loss[loss=2.319, over 2590.00 frames. , ppl: 10.161334208447323] tot_loss[loss=2.305, over 5497590.56 frames. , ppl: 10.021653323038661], batch size: 70
+2022-12-11 12:13:27,420 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 12:13:28,169 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.910991968643046
+2022-12-11 12:15:09,924 INFO [train.py:421] (4/8) Epoch 4, batch 68200, loss[loss=2.597, over 770.00 frames. , ppl: 13.42749493456386] tot_loss[loss=2.305, over 5502661.21 frames. , ppl: 10.023360567270833], batch size: 70
+2022-12-11 12:16:49,666 INFO [train.py:421] (4/8) Epoch 4, batch 68400, loss[loss=2.25, over 1820.00 frames. , ppl: 9.488754063193678] tot_loss[loss=2.306, over 5494991.53 frames. , ppl: 10.031495319307648], batch size: 70
+2022-12-11 12:18:30,823 INFO [train.py:421] (4/8) Epoch 4, batch 68600, loss[loss=2.278, over 3430.00 frames. , ppl: 9.7612311820798] tot_loss[loss=2.306, over 5489656.73 frames. , ppl: 10.031213865890198], batch size: 70
+2022-12-11 12:20:12,563 INFO [train.py:421] (4/8) Epoch 4, batch 68800, loss[loss=2.259, over 6930.00 frames. , ppl: 9.571959711899831] tot_loss[loss=2.306, over 5485711.63 frames. , ppl: 10.03422235942407], batch size: 70
+2022-12-11 12:21:53,589 INFO [train.py:421] (4/8) Epoch 4, batch 69000, loss[loss=2.751, over 770.00 frames. , ppl: 15.662989985028636] tot_loss[loss=2.306, over 5495060.87 frames. , ppl: 10.035518191305666], batch size: 70
+2022-12-11 12:21:53,590 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 12:21:54,350 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.294, over 211138.00 frames. , ppl: 9.910044396033411
+2022-12-11 12:23:33,505 INFO [train.py:421] (4/8) Epoch 4, batch 69200, loss[loss=2.46, over 1050.00 frames. , ppl: 11.702386896457295] tot_loss[loss=2.306, over 5535229.20 frames. , ppl: 10.03026753792436], batch size: 70
+2022-12-11 12:25:12,763 INFO [train.py:421] (4/8) Epoch 4, batch 69400, loss[loss=2.725, over 700.00 frames. , ppl: 15.260163042436501] tot_loss[loss=2.306, over 5526367.46 frames. , ppl: 10.029711942862159], batch size: 70
+2022-12-11 12:26:51,729 INFO [train.py:421] (4/8) Epoch 4, batch 69600, loss[loss=2.4, over 1400.00 frames. , ppl: 11.02538723514017] tot_loss[loss=2.306, over 5495877.84 frames. , ppl: 10.030429585698393], batch size: 70
+2022-12-11 12:28:30,365 INFO [train.py:421] (4/8) Epoch 4, batch 69800, loss[loss=2.408, over 2240.00 frames. , ppl: 11.115048482657988] tot_loss[loss=2.306, over 5478139.52 frames. , ppl: 10.03438692801827], batch size: 70
+2022-12-11 12:30:08,269 INFO [train.py:421] (4/8) Epoch 4, batch 70000, loss[loss=2.26, over 3780.00 frames. , ppl: 9.57999195915236] tot_loss[loss=2.306, over 5446060.47 frames. , ppl: 10.037867578975847], batch size: 70
+2022-12-11 12:30:08,270 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 12:30:09,018 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.292, over 211138.00 frames. , ppl: 9.896056980662939
+2022-12-11 12:31:48,775 INFO [train.py:421] (4/8) Epoch 4, batch 70200, loss[loss=2.369, over 1400.00 frames. , ppl: 10.689418477036446] tot_loss[loss=2.305, over 5501177.16 frames. , ppl: 10.024386380915443], batch size: 70
+2022-12-11 12:33:26,587 INFO [train.py:421] (4/8) Epoch 4, batch 70400, loss[loss=3.606, over 420.00 frames. , ppl: 36.81750116613861] tot_loss[loss=2.304, over 5515119.60 frames. , ppl: 10.017685360209965], batch size: 70
+2022-12-11 12:35:03,418 INFO [train.py:421] (4/8) Epoch 4, batch 70600, loss[loss=2.443, over 1540.00 frames. , ppl: 11.51253112212051] tot_loss[loss=2.303, over 5563560.94 frames. , ppl: 10.002366377891692], batch size: 70
+2022-12-11 12:36:42,005 INFO [train.py:421] (4/8) Epoch 4, batch 70800, loss[loss=2.411, over 1330.00 frames. , ppl: 11.143403817974777] tot_loss[loss=2.304, over 5535616.12 frames. , ppl: 10.010608120851359], batch size: 70
+2022-12-11 12:38:21,443 INFO [train.py:421] (4/8) Epoch 4, batch 71000, loss[loss=2.578, over 910.00 frames. , ppl: 13.166657580711828] tot_loss[loss=2.304, over 5495778.84 frames. , ppl: 10.018842009844176], batch size: 70
+2022-12-11 12:38:21,443 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 12:38:22,205 INFO [train.py:452] (4/8) Epoch 4, validation: loss=2.295, over 211138.00 frames. , ppl: 9.920506450462009
+2022-12-11 12:40:01,637 INFO [train.py:421] (4/8) Epoch 4, batch 71200, loss[loss=2.339, over 2380.00 frames. , ppl: 10.368138793722258] tot_loss[loss=2.306, over 5444530.54 frames. , ppl: 10.029578423780476], batch size: 70
+2022-12-11 12:41:40,328 INFO [train.py:421] (4/8) Epoch 4, batch 71400, loss[loss=2.517, over 1050.00 frames. , ppl: 12.389051648709675] tot_loss[loss=2.305, over 5484814.91 frames. , ppl: 10.025474355952074], batch size: 70
+2022-12-11 12:43:20,028 INFO [train.py:421] (4/8) Epoch 4, batch 71600, loss[loss=2.412, over 1890.00 frames. , ppl: 11.161692455866202] tot_loss[loss=2.307, over 5452637.88 frames. , ppl: 10.04343850071127], batch size: 70
+2022-12-11 12:45:06,136 INFO [train.py:421] (4/8) Epoch 4, batch 71800, loss[loss=3.217, over 490.00 frames. , ppl: 24.946625940266834] tot_loss[loss=2.308, over 5414207.16 frames. , ppl: 10.049466846365831], batch size: 70
+2022-12-11 12:46:20,323 INFO [train.py:421] (4/8) Epoch 5, batch 0, loss[loss=2.452, over 1260.00 frames. , ppl: 11.60724131063845] tot_loss[loss=2.452, over 1260.00 frames. , ppl: 11.60724131063845], batch size: 70
+2022-12-11 12:48:02,166 INFO [train.py:421] (4/8) Epoch 5, batch 200, loss[loss=2.579, over 770.00 frames. , ppl: 13.186800145603145] tot_loss[loss=2.284, over 535745.26 frames. , ppl: 9.820059320554126], batch size: 70
+2022-12-11 12:49:43,355 INFO [train.py:421] (4/8) Epoch 5, batch 400, loss[loss=2.26, over 4550.00 frames. , ppl: 9.583222860347155] tot_loss[loss=2.29, over 1010066.68 frames. , ppl: 9.872214735326994], batch size: 70
+2022-12-11 12:51:24,423 INFO [train.py:421] (4/8) Epoch 5, batch 600, loss[loss=2.193, over 3220.00 frames. , ppl: 8.957686673576635] tot_loss[loss=2.29, over 1469929.21 frames. , ppl: 9.87649220167715], batch size: 70
+2022-12-11 12:53:04,445 INFO [train.py:421] (4/8) Epoch 5, batch 800, loss[loss=2.284, over 4830.00 frames. , ppl: 9.818306302628313] tot_loss[loss=2.29, over 1859900.76 frames. , ppl: 9.874311103899913], batch size: 70
+2022-12-11 12:54:44,630 INFO [train.py:421] (4/8) Epoch 5, batch 1000, loss[loss=2.318, over 3920.00 frames. , ppl: 10.151588459591704] tot_loss[loss=2.29, over 2236268.14 frames. , ppl: 9.873925673676565], batch size: 70
+2022-12-11 12:54:44,630 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 12:54:45,418 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.90861440768603
+2022-12-11 12:56:25,126 INFO [train.py:421] (4/8) Epoch 5, batch 1200, loss[loss=2.261, over 3080.00 frames. , ppl: 9.596273943035156] tot_loss[loss=2.292, over 2529172.60 frames. , ppl: 9.895523835475137], batch size: 70
+2022-12-11 12:58:03,058 INFO [train.py:421] (4/8) Epoch 5, batch 1400, loss[loss=2.409, over 910.00 frames. , ppl: 11.12675383750653] tot_loss[loss=2.294, over 2771675.02 frames. , ppl: 9.9174971258905], batch size: 70
+2022-12-11 12:59:45,897 INFO [train.py:421] (4/8) Epoch 5, batch 1600, loss[loss=2.226, over 2380.00 frames. , ppl: 9.260276038586307] tot_loss[loss=2.293, over 3049202.95 frames. , ppl: 9.905290976232452], batch size: 70
+2022-12-11 13:01:25,826 INFO [train.py:421] (4/8) Epoch 5, batch 1800, loss[loss=2.197, over 4130.00 frames. , ppl: 8.996924556933564] tot_loss[loss=2.295, over 3289868.63 frames. , ppl: 9.91969443454471], batch size: 70
+2022-12-11 13:03:08,689 INFO [train.py:421] (4/8) Epoch 5, batch 2000, loss[loss=2.278, over 1820.00 frames. , ppl: 9.759159606043761] tot_loss[loss=2.294, over 3514471.88 frames. , ppl: 9.91687982670235], batch size: 70
+2022-12-11 13:03:08,690 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:03:09,438 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.893647858407924
+2022-12-11 13:04:49,705 INFO [train.py:421] (4/8) Epoch 5, batch 2200, loss[loss=2.953, over 560.00 frames. , ppl: 19.167792097811432] tot_loss[loss=2.293, over 3706972.40 frames. , ppl: 9.907058357665013], batch size: 70
+2022-12-11 13:06:31,486 INFO [train.py:421] (4/8) Epoch 5, batch 2400, loss[loss=2.33, over 2310.00 frames. , ppl: 10.280582790275709] tot_loss[loss=2.295, over 3860932.02 frames. , ppl: 9.92777288477752], batch size: 70
+2022-12-11 13:08:14,048 INFO [train.py:421] (4/8) Epoch 5, batch 2600, loss[loss=2.211, over 4270.00 frames. , ppl: 9.126071600408906] tot_loss[loss=2.294, over 4060420.46 frames. , ppl: 9.911290954957153], batch size: 70
+2022-12-11 13:09:54,954 INFO [train.py:421] (4/8) Epoch 5, batch 2800, loss[loss=2.388, over 1610.00 frames. , ppl: 10.88962232807487] tot_loss[loss=2.293, over 4192525.46 frames. , ppl: 9.909154875048504], batch size: 70
+2022-12-11 13:11:38,206 INFO [train.py:421] (4/8) Epoch 5, batch 3000, loss[loss=2.2, over 5600.00 frames. , ppl: 9.026886560346686] tot_loss[loss=2.294, over 4327593.33 frames. , ppl: 9.910869868594672], batch size: 70
+2022-12-11 13:11:38,207 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:11:38,953 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.878481675447484
+2022-12-11 13:13:17,482 INFO [train.py:421] (4/8) Epoch 5, batch 3200, loss[loss=2.19, over 3150.00 frames. , ppl: 8.9335055153371] tot_loss[loss=2.293, over 4438807.00 frames. , ppl: 9.903338210925643], batch size: 70
+2022-12-11 13:14:57,420 INFO [train.py:421] (4/8) Epoch 5, batch 3400, loss[loss=2.203, over 7420.00 frames. , ppl: 9.049711939048285] tot_loss[loss=2.293, over 4562629.60 frames. , ppl: 9.902390285620719], batch size: 70
+2022-12-11 13:16:37,748 INFO [train.py:421] (4/8) Epoch 5, batch 3600, loss[loss=3.662, over 420.00 frames. , ppl: 38.94136522948087] tot_loss[loss=2.294, over 4635040.57 frames. , ppl: 9.91704435162463], batch size: 70
+2022-12-11 13:18:14,022 INFO [train.py:421] (4/8) Epoch 5, batch 3800, loss[loss=3.082, over 560.00 frames. , ppl: 21.8004437263907] tot_loss[loss=2.294, over 4718949.51 frames. , ppl: 9.916107464478769], batch size: 70
+2022-12-11 13:19:56,337 INFO [train.py:421] (4/8) Epoch 5, batch 4000, loss[loss=2.242, over 840.00 frames. , ppl: 9.415874618117783] tot_loss[loss=2.294, over 4821726.99 frames. , ppl: 9.917946281991812], batch size: 70
+2022-12-11 13:19:56,338 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:19:57,099 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.905227258803862
+2022-12-11 13:21:39,464 INFO [train.py:421] (4/8) Epoch 5, batch 4200, loss[loss=2.168, over 4970.00 frames. , ppl: 8.736576355619443] tot_loss[loss=2.293, over 4888571.82 frames. , ppl: 9.908933945828615], batch size: 70
+2022-12-11 13:23:20,068 INFO [train.py:421] (4/8) Epoch 5, batch 4400, loss[loss=2.254, over 2380.00 frames. , ppl: 9.524721272220795] tot_loss[loss=2.292, over 4992277.26 frames. , ppl: 9.890231665857609], batch size: 70
+2022-12-11 13:25:01,536 INFO [train.py:421] (4/8) Epoch 5, batch 4600, loss[loss=2.367, over 2450.00 frames. , ppl: 10.660156226023343] tot_loss[loss=2.293, over 5043755.84 frames. , ppl: 9.905526318357554], batch size: 70
+2022-12-11 13:26:39,713 INFO [train.py:421] (4/8) Epoch 5, batch 4800, loss[loss=2.336, over 2170.00 frames. , ppl: 10.341028287152445] tot_loss[loss=2.293, over 5097303.85 frames. , ppl: 9.90724231400567], batch size: 70
+2022-12-11 13:28:19,898 INFO [train.py:421] (4/8) Epoch 5, batch 5000, loss[loss=2.201, over 4550.00 frames. , ppl: 9.033676181512826] tot_loss[loss=2.291, over 5174703.71 frames. , ppl: 9.888967813745252], batch size: 70
+2022-12-11 13:28:19,899 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:28:20,645 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.89573182475821
+2022-12-11 13:30:01,900 INFO [train.py:421] (4/8) Epoch 5, batch 5200, loss[loss=2.844, over 700.00 frames. , ppl: 17.177080439335906] tot_loss[loss=2.292, over 5197847.36 frames. , ppl: 9.894674975258349], batch size: 70
+2022-12-11 13:31:43,064 INFO [train.py:421] (4/8) Epoch 5, batch 5400, loss[loss=2.988, over 630.00 frames. , ppl: 19.83795220786746] tot_loss[loss=2.293, over 5221253.68 frames. , ppl: 9.906799544296003], batch size: 70
+2022-12-11 13:33:23,011 INFO [train.py:421] (4/8) Epoch 5, batch 5600, loss[loss=2.349, over 2310.00 frames. , ppl: 10.471341169480105] tot_loss[loss=2.294, over 5241279.31 frames. , ppl: 9.912777429617465], batch size: 70
+2022-12-11 13:35:01,443 INFO [train.py:421] (4/8) Epoch 5, batch 5800, loss[loss=2.791, over 630.00 frames. , ppl: 16.294586026979115] tot_loss[loss=2.295, over 5232852.00 frames. , ppl: 9.923216493237454], batch size: 70
+2022-12-11 13:36:38,803 INFO [train.py:421] (4/8) Epoch 5, batch 6000, loss[loss=2.372, over 1960.00 frames. , ppl: 10.718592826261194] tot_loss[loss=2.295, over 5259208.22 frames. , ppl: 9.925557444271098], batch size: 70
+2022-12-11 13:36:38,803 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:36:39,553 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.896351387792528
+2022-12-11 13:38:21,637 INFO [train.py:421] (4/8) Epoch 5, batch 6200, loss[loss=2.18, over 11340.00 frames. , ppl: 8.848955635072796] tot_loss[loss=2.295, over 5297040.93 frames. , ppl: 9.925612878678981], batch size: 70
+2022-12-11 13:39:57,535 INFO [train.py:421] (4/8) Epoch 5, batch 6400, loss[loss=2.252, over 3920.00 frames. , ppl: 9.505488984969416] tot_loss[loss=2.296, over 5307498.75 frames. , ppl: 9.93441079603025], batch size: 70
+2022-12-11 13:41:36,935 INFO [train.py:421] (4/8) Epoch 5, batch 6600, loss[loss=2.534, over 770.00 frames. , ppl: 12.607812479925947] tot_loss[loss=2.296, over 5328304.19 frames. , ppl: 9.936425979502145], batch size: 70
+2022-12-11 13:43:15,780 INFO [train.py:421] (4/8) Epoch 5, batch 6800, loss[loss=2.762, over 700.00 frames. , ppl: 15.830897812459662] tot_loss[loss=2.297, over 5323249.75 frames. , ppl: 9.942945301664317], batch size: 70
+2022-12-11 13:44:54,863 INFO [train.py:421] (4/8) Epoch 5, batch 7000, loss[loss=2.285, over 2310.00 frames. , ppl: 9.82249699922946] tot_loss[loss=2.296, over 5370335.57 frames. , ppl: 9.939318549704781], batch size: 70
+2022-12-11 13:44:54,864 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:44:55,610 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.88711469496257
+2022-12-11 13:46:38,710 INFO [train.py:421] (4/8) Epoch 5, batch 7200, loss[loss=2.352, over 1470.00 frames. , ppl: 10.506631228801469] tot_loss[loss=2.298, over 5345649.16 frames. , ppl: 9.950879345575604], batch size: 70
+2022-12-11 13:48:18,508 INFO [train.py:421] (4/8) Epoch 5, batch 7400, loss[loss=2.509, over 1190.00 frames. , ppl: 12.288581279222493] tot_loss[loss=2.3, over 5311416.09 frames. , ppl: 9.970555878188033], batch size: 70
+2022-12-11 13:49:58,390 INFO [train.py:421] (4/8) Epoch 5, batch 7600, loss[loss=2.316, over 3990.00 frames. , ppl: 10.1375383543848] tot_loss[loss=2.3, over 5343379.22 frames. , ppl: 9.976382840180149], batch size: 70
+2022-12-11 13:51:37,173 INFO [train.py:421] (4/8) Epoch 5, batch 7800, loss[loss=2.245, over 6370.00 frames. , ppl: 9.442526501517033] tot_loss[loss=2.3, over 5344226.27 frames. , ppl: 9.975582334491149], batch size: 70
+2022-12-11 13:53:18,459 INFO [train.py:421] (4/8) Epoch 5, batch 8000, loss[loss=2.421, over 2520.00 frames. , ppl: 11.259974736177059] tot_loss[loss=2.299, over 5359538.94 frames. , ppl: 9.967502038102488], batch size: 70
+2022-12-11 13:53:18,459 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 13:53:19,219 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.903350895657939
+2022-12-11 13:54:59,219 INFO [train.py:421] (4/8) Epoch 5, batch 8200, loss[loss=2.485, over 1330.00 frames. , ppl: 11.998754929969053] tot_loss[loss=2.299, over 5367779.20 frames. , ppl: 9.964913314030326], batch size: 70
+2022-12-11 13:56:40,075 INFO [train.py:421] (4/8) Epoch 5, batch 8400, loss[loss=2.343, over 2030.00 frames. , ppl: 10.415783844285432] tot_loss[loss=2.3, over 5368048.52 frames. , ppl: 9.972051247854482], batch size: 70
+2022-12-11 13:58:20,775 INFO [train.py:421] (4/8) Epoch 5, batch 8600, loss[loss=2.249, over 2310.00 frames. , ppl: 9.480428638111134] tot_loss[loss=2.3, over 5412323.89 frames. , ppl: 9.969840060807075], batch size: 70
+2022-12-11 14:00:00,780 INFO [train.py:421] (4/8) Epoch 5, batch 8800, loss[loss=2.262, over 3990.00 frames. , ppl: 9.605861287591424] tot_loss[loss=2.301, over 5375915.14 frames. , ppl: 9.982294653873984], batch size: 70
+2022-12-11 14:01:38,184 INFO [train.py:421] (4/8) Epoch 5, batch 9000, loss[loss=2.284, over 4200.00 frames. , ppl: 9.819273109730535] tot_loss[loss=2.301, over 5350928.41 frames. , ppl: 9.988489326169237], batch size: 70
+2022-12-11 14:01:38,185 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:01:38,944 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.898942841256256
+2022-12-11 14:03:18,628 INFO [train.py:421] (4/8) Epoch 5, batch 9200, loss[loss=2.48, over 1190.00 frames. , ppl: 11.945554424288998] tot_loss[loss=2.302, over 5322321.52 frames. , ppl: 9.99868919077532], batch size: 70
+2022-12-11 14:04:53,968 INFO [train.py:421] (4/8) Epoch 5, batch 9400, loss[loss=2.224, over 3290.00 frames. , ppl: 9.245734895161421] tot_loss[loss=2.302, over 5343242.45 frames. , ppl: 9.990528580875514], batch size: 70
+2022-12-11 14:06:31,800 INFO [train.py:421] (4/8) Epoch 5, batch 9600, loss[loss=2.468, over 1750.00 frames. , ppl: 11.795935492198018] tot_loss[loss=2.302, over 5318570.50 frames. , ppl: 9.99739074937757], batch size: 70
+2022-12-11 14:08:10,034 INFO [train.py:421] (4/8) Epoch 5, batch 9800, loss[loss=2.226, over 5320.00 frames. , ppl: 9.25958282292371] tot_loss[loss=2.303, over 5293596.76 frames. , ppl: 10.005136674658337], batch size: 70
+2022-12-11 14:09:51,391 INFO [train.py:421] (4/8) Epoch 5, batch 10000, loss[loss=2.19, over 5460.00 frames. , ppl: 8.936447596182616] tot_loss[loss=2.303, over 5301984.34 frames. , ppl: 10.001128282468795], batch size: 70
+2022-12-11 14:09:51,392 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:09:52,185 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.293, over 211138.00 frames. , ppl: 9.904837297826665
+2022-12-11 14:11:34,860 INFO [train.py:421] (4/8) Epoch 5, batch 10200, loss[loss=2.3, over 2730.00 frames. , ppl: 9.978762927311443] tot_loss[loss=2.302, over 5346639.38 frames. , ppl: 9.99780881093043], batch size: 70
+2022-12-11 14:13:12,042 INFO [train.py:421] (4/8) Epoch 5, batch 10400, loss[loss=2.417, over 1260.00 frames. , ppl: 11.207838303033698] tot_loss[loss=2.303, over 5330480.55 frames. , ppl: 10.006202964288532], batch size: 70
+2022-12-11 14:14:53,715 INFO [train.py:421] (4/8) Epoch 5, batch 10600, loss[loss=2.172, over 2870.00 frames. , ppl: 8.780032967766614] tot_loss[loss=2.304, over 5295991.62 frames. , ppl: 10.017548847489872], batch size: 70
+2022-12-11 14:16:33,812 INFO [train.py:421] (4/8) Epoch 5, batch 10800, loss[loss=2.543, over 1960.00 frames. , ppl: 12.722484162801898] tot_loss[loss=2.304, over 5318831.38 frames. , ppl: 10.00992188736663], batch size: 70
+2022-12-11 14:18:14,047 INFO [train.py:421] (4/8) Epoch 5, batch 11000, loss[loss=2.319, over 2380.00 frames. , ppl: 10.162491452562039] tot_loss[loss=2.304, over 5336214.36 frames. , ppl: 10.010004178863847], batch size: 70
+2022-12-11 14:18:14,048 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:18:14,795 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.8904824753545
+2022-12-11 14:19:51,235 INFO [train.py:421] (4/8) Epoch 5, batch 11200, loss[loss=2.214, over 6790.00 frames. , ppl: 9.14811021182707] tot_loss[loss=2.303, over 5352692.98 frames. , ppl: 10.00412147308396], batch size: 70
+2022-12-11 14:21:32,948 INFO [train.py:421] (4/8) Epoch 5, batch 11400, loss[loss=2.276, over 7770.00 frames. , ppl: 9.741350429846875] tot_loss[loss=2.301, over 5414076.32 frames. , ppl: 9.988715586610073], batch size: 70
+2022-12-11 14:23:12,624 INFO [train.py:421] (4/8) Epoch 5, batch 11600, loss[loss=2.827, over 700.00 frames. , ppl: 16.889242779188606] tot_loss[loss=2.301, over 5417253.18 frames. , ppl: 9.987389221912652], batch size: 70
+2022-12-11 14:24:53,833 INFO [train.py:421] (4/8) Epoch 5, batch 11800, loss[loss=2.383, over 3150.00 frames. , ppl: 10.838794923755845] tot_loss[loss=2.301, over 5415769.92 frames. , ppl: 9.980497968924599], batch size: 70
+2022-12-11 14:26:33,014 INFO [train.py:421] (4/8) Epoch 5, batch 12000, loss[loss=2.385, over 1750.00 frames. , ppl: 10.861192874127408] tot_loss[loss=2.301, over 5397688.10 frames. , ppl: 9.980929840808281], batch size: 70
+2022-12-11 14:26:33,014 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:26:33,774 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.883088329655019
+2022-12-11 14:28:11,315 INFO [train.py:421] (4/8) Epoch 5, batch 12200, loss[loss=2.235, over 4900.00 frames. , ppl: 9.342755253528653] tot_loss[loss=2.3, over 5423277.69 frames. , ppl: 9.977803084865332], batch size: 70
+2022-12-11 14:29:52,535 INFO [train.py:421] (4/8) Epoch 5, batch 12400, loss[loss=2.989, over 560.00 frames. , ppl: 19.85879585231802] tot_loss[loss=2.3, over 5410451.66 frames. , ppl: 9.972176622154699], batch size: 70
+2022-12-11 14:31:29,844 INFO [train.py:421] (4/8) Epoch 5, batch 12600, loss[loss=2.18, over 5320.00 frames. , ppl: 8.848195991849524] tot_loss[loss=2.301, over 5370877.21 frames. , ppl: 9.987209413335211], batch size: 70
+2022-12-11 14:33:09,806 INFO [train.py:421] (4/8) Epoch 5, batch 12800, loss[loss=2.205, over 4480.00 frames. , ppl: 9.072213911636016] tot_loss[loss=2.3, over 5403746.06 frames. , ppl: 9.977257955096734], batch size: 70
+2022-12-11 14:34:47,875 INFO [train.py:421] (4/8) Epoch 5, batch 13000, loss[loss=2.251, over 3570.00 frames. , ppl: 9.492803294602043] tot_loss[loss=2.3, over 5440939.65 frames. , ppl: 9.971061538756286], batch size: 70
+2022-12-11 14:34:47,876 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:34:48,636 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.882500313269402
+2022-12-11 14:36:27,618 INFO [train.py:421] (4/8) Epoch 5, batch 13200, loss[loss=2.439, over 1330.00 frames. , ppl: 11.457541909249992] tot_loss[loss=2.3, over 5404168.96 frames. , ppl: 9.974112169346633], batch size: 70
+2022-12-11 14:38:12,162 INFO [train.py:421] (4/8) Epoch 5, batch 13400, loss[loss=2.35, over 2660.00 frames. , ppl: 10.485669813614585] tot_loss[loss=2.299, over 5437658.88 frames. , ppl: 9.968302762925592], batch size: 70
+2022-12-11 14:39:51,871 INFO [train.py:421] (4/8) Epoch 5, batch 13600, loss[loss=2.392, over 1610.00 frames. , ppl: 10.935438088347318] tot_loss[loss=2.3, over 5399372.03 frames. , ppl: 9.973487710857233], batch size: 70
+2022-12-11 14:41:35,232 INFO [train.py:421] (4/8) Epoch 5, batch 13800, loss[loss=2.268, over 2310.00 frames. , ppl: 9.655584799993754] tot_loss[loss=2.3, over 5416796.91 frames. , ppl: 9.978377079992839], batch size: 70
+2022-12-11 14:43:18,804 INFO [train.py:421] (4/8) Epoch 5, batch 14000, loss[loss=2.18, over 5250.00 frames. , ppl: 8.847770891219717] tot_loss[loss=2.298, over 5494124.53 frames. , ppl: 9.957808410476014], batch size: 70
+2022-12-11 14:43:18,804 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:43:19,550 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.88920899459524
+2022-12-11 14:45:00,093 INFO [train.py:421] (4/8) Epoch 5, batch 14200, loss[loss=2.45, over 1680.00 frames. , ppl: 11.592790066834057] tot_loss[loss=2.299, over 5474365.43 frames. , ppl: 9.966441298635079], batch size: 70
+2022-12-11 14:46:37,410 INFO [train.py:421] (4/8) Epoch 5, batch 14400, loss[loss=2.191, over 6790.00 frames. , ppl: 8.946989252683649] tot_loss[loss=2.298, over 5520200.08 frames. , ppl: 9.954581546247729], batch size: 70
+2022-12-11 14:48:13,461 INFO [train.py:421] (4/8) Epoch 5, batch 14600, loss[loss=2.222, over 6090.00 frames. , ppl: 9.224921446160364] tot_loss[loss=2.298, over 5517284.65 frames. , ppl: 9.956601985939797], batch size: 70
+2022-12-11 14:49:54,267 INFO [train.py:421] (4/8) Epoch 5, batch 14800, loss[loss=2.6, over 840.00 frames. , ppl: 13.468818261014526] tot_loss[loss=2.3, over 5444804.01 frames. , ppl: 9.9766916275785], batch size: 70
+2022-12-11 14:51:36,210 INFO [train.py:421] (4/8) Epoch 5, batch 15000, loss[loss=2.207, over 4690.00 frames. , ppl: 9.092025596062228] tot_loss[loss=2.299, over 5475406.15 frames. , ppl: 9.96549119072181], batch size: 70
+2022-12-11 14:51:36,211 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 14:51:36,973 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.882491537170422
+2022-12-11 14:53:18,612 INFO [train.py:421] (4/8) Epoch 5, batch 15200, loss[loss=2.615, over 910.00 frames. , ppl: 13.665641975452925] tot_loss[loss=2.299, over 5463891.98 frames. , ppl: 9.967438876579276], batch size: 70
+2022-12-11 14:55:01,722 INFO [train.py:421] (4/8) Epoch 5, batch 15400, loss[loss=2.296, over 3640.00 frames. , ppl: 9.934822354020396] tot_loss[loss=2.298, over 5527526.21 frames. , ppl: 9.951275027498829], batch size: 70
+2022-12-11 14:56:39,993 INFO [train.py:421] (4/8) Epoch 5, batch 15600, loss[loss=2.385, over 1470.00 frames. , ppl: 10.857992884103888] tot_loss[loss=2.298, over 5509733.90 frames. , ppl: 9.958598310879765], batch size: 70
+2022-12-11 14:58:22,472 INFO [train.py:421] (4/8) Epoch 5, batch 15800, loss[loss=2.556, over 770.00 frames. , ppl: 12.88800973630084] tot_loss[loss=2.297, over 5553062.51 frames. , ppl: 9.940444409096946], batch size: 70
+2022-12-11 15:00:04,356 INFO [train.py:421] (4/8) Epoch 5, batch 16000, loss[loss=2.175, over 4130.00 frames. , ppl: 8.80482255959373] tot_loss[loss=2.297, over 5552598.51 frames. , ppl: 9.942076271230667], batch size: 70
+2022-12-11 15:00:04,357 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:00:05,117 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.876905669210421
+2022-12-11 15:01:47,812 INFO [train.py:421] (4/8) Epoch 5, batch 16200, loss[loss=2.436, over 1890.00 frames. , ppl: 11.427588846056667] tot_loss[loss=2.298, over 5520422.00 frames. , ppl: 9.949427887123852], batch size: 70
+2022-12-11 15:03:30,620 INFO [train.py:421] (4/8) Epoch 5, batch 16400, loss[loss=2.49, over 1470.00 frames. , ppl: 12.064027175592473] tot_loss[loss=2.298, over 5493457.03 frames. , ppl: 9.956318673739688], batch size: 70
+2022-12-11 15:05:08,590 INFO [train.py:421] (4/8) Epoch 5, batch 16600, loss[loss=2.284, over 2100.00 frames. , ppl: 9.812503689195074] tot_loss[loss=2.298, over 5501601.78 frames. , ppl: 9.95728196231693], batch size: 70
+2022-12-11 15:06:50,339 INFO [train.py:421] (4/8) Epoch 5, batch 16800, loss[loss=2.289, over 1680.00 frames. , ppl: 9.869472825319276] tot_loss[loss=2.299, over 5472056.09 frames. , ppl: 9.968958402643889], batch size: 70
+2022-12-11 15:08:31,261 INFO [train.py:421] (4/8) Epoch 5, batch 17000, loss[loss=2.382, over 980.00 frames. , ppl: 10.824418432782506] tot_loss[loss=2.298, over 5499531.66 frames. , ppl: 9.958271582904949], batch size: 70
+2022-12-11 15:08:31,262 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:08:32,009 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.891302273858114
+2022-12-11 15:10:09,655 INFO [train.py:421] (4/8) Epoch 5, batch 17200, loss[loss=2.48, over 910.00 frames. , ppl: 11.942277464071969] tot_loss[loss=2.299, over 5498506.92 frames. , ppl: 9.960788342283092], batch size: 70
+2022-12-11 15:11:47,999 INFO [train.py:421] (4/8) Epoch 5, batch 17400, loss[loss=2.38, over 1890.00 frames. , ppl: 10.808957919713366] tot_loss[loss=2.298, over 5488628.01 frames. , ppl: 9.958674551184957], batch size: 70
+2022-12-11 15:13:29,109 INFO [train.py:421] (4/8) Epoch 5, batch 17600, loss[loss=2.396, over 1470.00 frames. , ppl: 10.980286798878703] tot_loss[loss=2.299, over 5457786.06 frames. , ppl: 9.964710422977179], batch size: 70
+2022-12-11 15:15:10,238 INFO [train.py:421] (4/8) Epoch 5, batch 17800, loss[loss=2.3, over 2310.00 frames. , ppl: 9.973145218063186] tot_loss[loss=2.3, over 5458822.01 frames. , ppl: 9.970691161181271], batch size: 70
+2022-12-11 15:16:49,288 INFO [train.py:421] (4/8) Epoch 5, batch 18000, loss[loss=2.309, over 1960.00 frames. , ppl: 10.06823043550128] tot_loss[loss=2.3, over 5473069.81 frames. , ppl: 9.971341855208466], batch size: 70
+2022-12-11 15:16:49,288 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:16:50,060 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.884169376428208
+2022-12-11 15:18:27,178 INFO [train.py:421] (4/8) Epoch 5, batch 18200, loss[loss=2.28, over 2520.00 frames. , ppl: 9.775068826779517] tot_loss[loss=2.301, over 5422035.64 frames. , ppl: 9.985546779173625], batch size: 70
+2022-12-11 15:20:07,209 INFO [train.py:421] (4/8) Epoch 5, batch 18400, loss[loss=2.804, over 700.00 frames. , ppl: 16.51730273632504] tot_loss[loss=2.301, over 5452216.85 frames. , ppl: 9.97990485281201], batch size: 70
+2022-12-11 15:21:46,262 INFO [train.py:421] (4/8) Epoch 5, batch 18600, loss[loss=2.247, over 2940.00 frames. , ppl: 9.459983554753405] tot_loss[loss=2.298, over 5533383.78 frames. , ppl: 9.954540138641917], batch size: 70
+2022-12-11 15:23:25,436 INFO [train.py:421] (4/8) Epoch 5, batch 18800, loss[loss=2.172, over 4760.00 frames. , ppl: 8.774868847930984] tot_loss[loss=2.299, over 5517531.62 frames. , ppl: 9.9607011761592], batch size: 70
+2022-12-11 15:25:02,768 INFO [train.py:421] (4/8) Epoch 5, batch 19000, loss[loss=2.39, over 2170.00 frames. , ppl: 10.911030178527357] tot_loss[loss=2.298, over 5544932.85 frames. , ppl: 9.955520081705865], batch size: 70
+2022-12-11 15:25:02,769 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:25:03,528 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.892699015651312
+2022-12-11 15:26:43,246 INFO [train.py:421] (4/8) Epoch 5, batch 19200, loss[loss=2.434, over 1820.00 frames. , ppl: 11.403802072222048] tot_loss[loss=2.299, over 5506385.25 frames. , ppl: 9.967682687040135], batch size: 70
+2022-12-11 15:28:24,621 INFO [train.py:421] (4/8) Epoch 5, batch 19400, loss[loss=2.277, over 4410.00 frames. , ppl: 9.747144628504765] tot_loss[loss=2.3, over 5515727.73 frames. , ppl: 9.970689161926822], batch size: 70
+2022-12-11 15:30:03,434 INFO [train.py:421] (4/8) Epoch 5, batch 19600, loss[loss=2.782, over 630.00 frames. , ppl: 16.157035803373816] tot_loss[loss=2.298, over 5580564.07 frames. , ppl: 9.954695588279975], batch size: 70
+2022-12-11 15:31:37,911 INFO [train.py:421] (4/8) Epoch 5, batch 19800, loss[loss=2.245, over 3290.00 frames. , ppl: 9.44126409318891] tot_loss[loss=2.298, over 5587167.04 frames. , ppl: 9.95491863006878], batch size: 70
+2022-12-11 15:33:17,500 INFO [train.py:421] (4/8) Epoch 5, batch 20000, loss[loss=2.313, over 1120.00 frames. , ppl: 10.105808439480816] tot_loss[loss=2.298, over 5593290.92 frames. , ppl: 9.953283152101106], batch size: 70
+2022-12-11 15:33:17,501 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:33:18,262 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.292, over 211138.00 frames. , ppl: 9.897619928623687
+2022-12-11 15:34:59,004 INFO [train.py:421] (4/8) Epoch 5, batch 20200, loss[loss=2.64, over 770.00 frames. , ppl: 14.014832989742152] tot_loss[loss=2.298, over 5574621.62 frames. , ppl: 9.95880064543805], batch size: 70
+2022-12-11 15:36:41,035 INFO [train.py:421] (4/8) Epoch 5, batch 20400, loss[loss=2.835, over 630.00 frames. , ppl: 17.02420562659681] tot_loss[loss=2.299, over 5531434.11 frames. , ppl: 9.967015490916523], batch size: 70
+2022-12-11 15:38:20,076 INFO [train.py:421] (4/8) Epoch 5, batch 20600, loss[loss=2.484, over 980.00 frames. , ppl: 11.989451531571536] tot_loss[loss=2.3, over 5539924.37 frames. , ppl: 9.973134768158621], batch size: 70
+2022-12-11 15:40:02,095 INFO [train.py:421] (4/8) Epoch 5, batch 20800, loss[loss=2.364, over 1820.00 frames. , ppl: 10.63554835767582] tot_loss[loss=2.298, over 5593105.01 frames. , ppl: 9.956141193384541], batch size: 70
+2022-12-11 15:41:40,040 INFO [train.py:421] (4/8) Epoch 5, batch 21000, loss[loss=2.192, over 6650.00 frames. , ppl: 8.953961065356165] tot_loss[loss=2.299, over 5559461.78 frames. , ppl: 9.960338855289153], batch size: 70
+2022-12-11 15:41:40,041 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:41:40,786 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.8614396619666
+2022-12-11 15:43:22,020 INFO [train.py:421] (4/8) Epoch 5, batch 21200, loss[loss=2.243, over 4410.00 frames. , ppl: 9.418184515254493] tot_loss[loss=2.298, over 5570284.27 frames. , ppl: 9.95512851263113], batch size: 70
+2022-12-11 15:45:08,223 INFO [train.py:421] (4/8) Epoch 5, batch 21400, loss[loss=2.338, over 980.00 frames. , ppl: 10.355496469624283] tot_loss[loss=2.299, over 5527882.51 frames. , ppl: 9.9608008459118], batch size: 70
+2022-12-11 15:46:45,918 INFO [train.py:421] (4/8) Epoch 5, batch 21600, loss[loss=2.33, over 2310.00 frames. , ppl: 10.276752363224485] tot_loss[loss=2.3, over 5482284.20 frames. , ppl: 9.970314705217866], batch size: 70
+2022-12-11 15:48:20,256 INFO [train.py:421] (4/8) Epoch 5, batch 21800, loss[loss=2.15, over 3290.00 frames. , ppl: 8.584736083387426] tot_loss[loss=2.299, over 5476425.51 frames. , ppl: 9.96618706835181], batch size: 70
+2022-12-11 15:49:59,709 INFO [train.py:421] (4/8) Epoch 5, batch 22000, loss[loss=2.409, over 1470.00 frames. , ppl: 11.121461986943485] tot_loss[loss=2.298, over 5514140.13 frames. , ppl: 9.956655175099591], batch size: 70
+2022-12-11 15:49:59,710 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:50:00,468 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.291, over 211138.00 frames. , ppl: 9.884660933415132
+2022-12-11 15:51:45,352 INFO [train.py:421] (4/8) Epoch 5, batch 22200, loss[loss=2.363, over 1750.00 frames. , ppl: 10.622208879508626] tot_loss[loss=2.299, over 5484426.85 frames. , ppl: 9.960254428815956], batch size: 70
+2022-12-11 15:53:28,205 INFO [train.py:421] (4/8) Epoch 5, batch 22400, loss[loss=2.23, over 5320.00 frames. , ppl: 9.302682577885726] tot_loss[loss=2.298, over 5496534.72 frames. , ppl: 9.953039782832459], batch size: 70
+2022-12-11 15:55:07,005 INFO [train.py:421] (4/8) Epoch 5, batch 22600, loss[loss=2.447, over 1260.00 frames. , ppl: 11.553371585654592] tot_loss[loss=2.297, over 5519517.02 frames. , ppl: 9.94482720899599], batch size: 70
+2022-12-11 15:56:53,258 INFO [train.py:421] (4/8) Epoch 5, batch 22800, loss[loss=2.328, over 2590.00 frames. , ppl: 10.257270905565848] tot_loss[loss=2.296, over 5536452.44 frames. , ppl: 9.93550236885725], batch size: 70
+2022-12-11 15:58:36,998 INFO [train.py:421] (4/8) Epoch 5, batch 23000, loss[loss=2.271, over 3500.00 frames. , ppl: 9.688975567995703] tot_loss[loss=2.296, over 5538755.14 frames. , ppl: 9.935867101345696], batch size: 70
+2022-12-11 15:58:36,998 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 15:58:37,746 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.87439159731824
+2022-12-11 16:00:16,678 INFO [train.py:421] (4/8) Epoch 5, batch 23200, loss[loss=2.432, over 1610.00 frames. , ppl: 11.380385304335817] tot_loss[loss=2.298, over 5476158.53 frames. , ppl: 9.95110619461683], batch size: 70
+2022-12-11 16:01:54,226 INFO [train.py:421] (4/8) Epoch 5, batch 23400, loss[loss=2.308, over 3920.00 frames. , ppl: 10.054577379589402] tot_loss[loss=2.297, over 5473312.28 frames. , ppl: 9.945523874021852], batch size: 70
+2022-12-11 16:03:33,364 INFO [train.py:421] (4/8) Epoch 5, batch 23600, loss[loss=2.382, over 2240.00 frames. , ppl: 10.827047275019702] tot_loss[loss=2.296, over 5500482.19 frames. , ppl: 9.936599444349591], batch size: 70
+2022-12-11 16:05:15,464 INFO [train.py:421] (4/8) Epoch 5, batch 23800, loss[loss=2.219, over 5320.00 frames. , ppl: 9.201801381369524] tot_loss[loss=2.296, over 5522295.70 frames. , ppl: 9.931266275496045], batch size: 70
+2022-12-11 16:06:54,574 INFO [train.py:421] (4/8) Epoch 5, batch 24000, loss[loss=2.253, over 2590.00 frames. , ppl: 9.517756156064982] tot_loss[loss=2.296, over 5492666.00 frames. , ppl: 9.937883633116613], batch size: 70
+2022-12-11 16:06:54,575 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:06:55,332 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.869746633606848
+2022-12-11 16:08:33,924 INFO [train.py:421] (4/8) Epoch 5, batch 24200, loss[loss=2.21, over 4340.00 frames. , ppl: 9.119360815239494] tot_loss[loss=2.297, over 5469752.82 frames. , ppl: 9.9492168147376], batch size: 70
+2022-12-11 16:10:16,216 INFO [train.py:421] (4/8) Epoch 5, batch 24400, loss[loss=2.211, over 4410.00 frames. , ppl: 9.125923346892634] tot_loss[loss=2.298, over 5441296.91 frames. , ppl: 9.958125707505554], batch size: 70
+2022-12-11 16:11:57,882 INFO [train.py:421] (4/8) Epoch 5, batch 24600, loss[loss=2.276, over 2800.00 frames. , ppl: 9.734810591593368] tot_loss[loss=2.297, over 5472591.69 frames. , ppl: 9.947485916766112], batch size: 70
+2022-12-11 16:13:31,672 INFO [train.py:421] (4/8) Epoch 5, batch 24800, loss[loss=2.3, over 2030.00 frames. , ppl: 9.970344607136191] tot_loss[loss=2.298, over 5485318.42 frames. , ppl: 9.949988938297437], batch size: 70
+2022-12-11 16:15:09,416 INFO [train.py:421] (4/8) Epoch 5, batch 25000, loss[loss=2.458, over 1680.00 frames. , ppl: 11.678407021594316] tot_loss[loss=2.297, over 5515390.49 frames. , ppl: 9.945922866689646], batch size: 70
+2022-12-11 16:15:09,416 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:15:10,164 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.867343916412231
+2022-12-11 16:16:49,415 INFO [train.py:421] (4/8) Epoch 5, batch 25200, loss[loss=2.239, over 3500.00 frames. , ppl: 9.380710931861381] tot_loss[loss=2.297, over 5536194.00 frames. , ppl: 9.941902236090122], batch size: 70
+2022-12-11 16:18:26,295 INFO [train.py:421] (4/8) Epoch 5, batch 25400, loss[loss=2.552, over 770.00 frames. , ppl: 12.835128824707397] tot_loss[loss=2.298, over 5530097.26 frames. , ppl: 9.949546880829805], batch size: 70
+2022-12-11 16:20:05,519 INFO [train.py:421] (4/8) Epoch 5, batch 25600, loss[loss=2.192, over 4410.00 frames. , ppl: 8.948655002305921] tot_loss[loss=2.299, over 5498187.92 frames. , ppl: 9.963331811495541], batch size: 70
+2022-12-11 16:21:46,552 INFO [train.py:421] (4/8) Epoch 5, batch 25800, loss[loss=2.796, over 630.00 frames. , ppl: 16.3817254368173] tot_loss[loss=2.299, over 5495636.69 frames. , ppl: 9.96444482249886], batch size: 70
+2022-12-11 16:23:25,042 INFO [train.py:421] (4/8) Epoch 5, batch 26000, loss[loss=2.532, over 1400.00 frames. , ppl: 12.57434668504609] tot_loss[loss=2.3, over 5459288.95 frames. , ppl: 9.974476164591042], batch size: 70
+2022-12-11 16:23:25,042 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:23:25,801 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.864416167537827
+2022-12-11 16:25:05,345 INFO [train.py:421] (4/8) Epoch 5, batch 26200, loss[loss=2.251, over 4900.00 frames. , ppl: 9.497132543304211] tot_loss[loss=2.3, over 5439121.19 frames. , ppl: 9.978210643790884], batch size: 70
+2022-12-11 16:26:45,398 INFO [train.py:421] (4/8) Epoch 5, batch 26400, loss[loss=2.386, over 1540.00 frames. , ppl: 10.871265785858824] tot_loss[loss=2.299, over 5471443.68 frames. , ppl: 9.967424401192465], batch size: 70
+2022-12-11 16:28:28,141 INFO [train.py:421] (4/8) Epoch 5, batch 26600, loss[loss=2.483, over 910.00 frames. , ppl: 11.972153163840005] tot_loss[loss=2.3, over 5465419.71 frames. , ppl: 9.969559523743849], batch size: 70
+2022-12-11 16:30:12,513 INFO [train.py:421] (4/8) Epoch 5, batch 26800, loss[loss=2.238, over 2660.00 frames. , ppl: 9.372240833617857] tot_loss[loss=2.3, over 5483035.56 frames. , ppl: 9.973369598194365], batch size: 70
+2022-12-11 16:31:49,994 INFO [train.py:421] (4/8) Epoch 5, batch 27000, loss[loss=2.695, over 700.00 frames. , ppl: 14.801868428470577] tot_loss[loss=2.3, over 5456302.28 frames. , ppl: 9.978480075193994], batch size: 70
+2022-12-11 16:31:49,995 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:31:50,754 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.875375224811005
+2022-12-11 16:33:32,081 INFO [train.py:421] (4/8) Epoch 5, batch 27200, loss[loss=2.358, over 2030.00 frames. , ppl: 10.573362041847222] tot_loss[loss=2.3, over 5466931.60 frames. , ppl: 9.972896819412771], batch size: 70
+2022-12-11 16:35:10,671 INFO [train.py:421] (4/8) Epoch 5, batch 27400, loss[loss=2.193, over 5600.00 frames. , ppl: 8.960761611065028] tot_loss[loss=2.297, over 5536606.61 frames. , ppl: 9.946335493545421], batch size: 70
+2022-12-11 16:36:49,937 INFO [train.py:421] (4/8) Epoch 5, batch 27600, loss[loss=2.728, over 700.00 frames. , ppl: 15.309065556969784] tot_loss[loss=2.297, over 5531640.43 frames. , ppl: 9.943063531765812], batch size: 70
+2022-12-11 16:38:28,348 INFO [train.py:421] (4/8) Epoch 5, batch 27800, loss[loss=2.244, over 3780.00 frames. , ppl: 9.428912283842134] tot_loss[loss=2.299, over 5480085.29 frames. , ppl: 9.963127203053233], batch size: 70
+2022-12-11 16:40:08,913 INFO [train.py:421] (4/8) Epoch 5, batch 28000, loss[loss=2.185, over 5180.00 frames. , ppl: 8.89446890405144] tot_loss[loss=2.299, over 5489519.58 frames. , ppl: 9.964684468458877], batch size: 70
+2022-12-11 16:40:08,914 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:40:09,677 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.858846353025392
+2022-12-11 16:41:47,858 INFO [train.py:421] (4/8) Epoch 5, batch 28200, loss[loss=2.868, over 630.00 frames. , ppl: 17.608993277687443] tot_loss[loss=2.3, over 5433493.79 frames. , ppl: 9.977118435003597], batch size: 70
+2022-12-11 16:43:28,937 INFO [train.py:421] (4/8) Epoch 5, batch 28400, loss[loss=2.313, over 2520.00 frames. , ppl: 10.103357163438107] tot_loss[loss=2.302, over 5410735.86 frames. , ppl: 9.989627661421432], batch size: 70
+2022-12-11 16:45:09,674 INFO [train.py:421] (4/8) Epoch 5, batch 28600, loss[loss=2.251, over 5600.00 frames. , ppl: 9.493079817605667] tot_loss[loss=2.302, over 5371587.91 frames. , ppl: 9.99262002265243], batch size: 70
+2022-12-11 16:46:50,655 INFO [train.py:421] (4/8) Epoch 5, batch 28800, loss[loss=4.049, over 350.00 frames. , ppl: 57.335832725811414] tot_loss[loss=2.302, over 5382891.35 frames. , ppl: 9.989278725687802], batch size: 70
+2022-12-11 16:48:28,318 INFO [train.py:421] (4/8) Epoch 5, batch 29000, loss[loss=2.27, over 3080.00 frames. , ppl: 9.683469944091861] tot_loss[loss=2.302, over 5360025.10 frames. , ppl: 9.993103257288737], batch size: 70
+2022-12-11 16:48:28,318 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:48:29,078 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.86421176864417
+2022-12-11 16:50:05,458 INFO [train.py:421] (4/8) Epoch 5, batch 29200, loss[loss=2.18, over 2940.00 frames. , ppl: 8.843744910250608] tot_loss[loss=2.302, over 5340647.14 frames. , ppl: 9.992799681229037], batch size: 70
+2022-12-11 16:51:46,725 INFO [train.py:421] (4/8) Epoch 5, batch 29400, loss[loss=2.241, over 4830.00 frames. , ppl: 9.407303126682189] tot_loss[loss=2.303, over 5339937.90 frames. , ppl: 9.999634574833419], batch size: 70
+2022-12-11 16:53:26,108 INFO [train.py:421] (4/8) Epoch 5, batch 29600, loss[loss=2.212, over 5740.00 frames. , ppl: 9.137909200324403] tot_loss[loss=2.302, over 5344014.83 frames. , ppl: 9.997935653144348], batch size: 70
+2022-12-11 16:55:06,784 INFO [train.py:421] (4/8) Epoch 5, batch 29800, loss[loss=4.147, over 350.00 frames. , ppl: 63.217797514135285] tot_loss[loss=2.302, over 5353543.19 frames. , ppl: 9.991635295614502], batch size: 70
+2022-12-11 16:56:46,812 INFO [train.py:421] (4/8) Epoch 5, batch 30000, loss[loss=2.405, over 1540.00 frames. , ppl: 11.081614551999776] tot_loss[loss=2.301, over 5363152.39 frames. , ppl: 9.986517835861575], batch size: 70
+2022-12-11 16:56:46,812 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 16:56:47,572 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.86519146183926
+2022-12-11 16:58:29,824 INFO [train.py:421] (4/8) Epoch 5, batch 30200, loss[loss=2.232, over 3430.00 frames. , ppl: 9.320768373300897] tot_loss[loss=2.302, over 5374401.03 frames. , ppl: 9.98920767692988], batch size: 70
+2022-12-11 17:00:10,447 INFO [train.py:421] (4/8) Epoch 5, batch 30400, loss[loss=2.283, over 1540.00 frames. , ppl: 9.801305272983733] tot_loss[loss=2.299, over 5478023.34 frames. , ppl: 9.961357523171168], batch size: 70
+2022-12-11 17:01:51,087 INFO [train.py:421] (4/8) Epoch 5, batch 30600, loss[loss=2.277, over 4900.00 frames. , ppl: 9.747300684564777] tot_loss[loss=2.297, over 5541670.99 frames. , ppl: 9.943193891537415], batch size: 70
+2022-12-11 17:03:31,212 INFO [train.py:421] (4/8) Epoch 5, batch 30800, loss[loss=2.205, over 9100.00 frames. , ppl: 9.068227184075539] tot_loss[loss=2.298, over 5498274.94 frames. , ppl: 9.955948057820876], batch size: 70
+2022-12-11 17:05:13,224 INFO [train.py:421] (4/8) Epoch 5, batch 31000, loss[loss=2.361, over 2450.00 frames. , ppl: 10.603678535596723] tot_loss[loss=2.298, over 5492103.25 frames. , ppl: 9.95888678433788], batch size: 70
+2022-12-11 17:05:13,224 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:05:13,971 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.862906635580453
+2022-12-11 17:06:53,705 INFO [train.py:421] (4/8) Epoch 5, batch 31200, loss[loss=2.371, over 980.00 frames. , ppl: 10.708858850992721] tot_loss[loss=2.299, over 5469691.98 frames. , ppl: 9.967649492541893], batch size: 70
+2022-12-11 17:08:34,075 INFO [train.py:421] (4/8) Epoch 5, batch 31400, loss[loss=2.365, over 2520.00 frames. , ppl: 10.64935622492263] tot_loss[loss=2.298, over 5508920.10 frames. , ppl: 9.954565090223133], batch size: 70
+2022-12-11 17:10:13,743 INFO [train.py:421] (4/8) Epoch 5, batch 31600, loss[loss=2.309, over 2450.00 frames. , ppl: 10.062284680053715] tot_loss[loss=2.299, over 5458754.61 frames. , ppl: 9.966204146102802], batch size: 70
+2022-12-11 17:11:57,205 INFO [train.py:421] (4/8) Epoch 5, batch 31800, loss[loss=2.596, over 840.00 frames. , ppl: 13.41279154946406] tot_loss[loss=2.299, over 5455697.81 frames. , ppl: 9.964463287856228], batch size: 70
+2022-12-11 17:13:39,197 INFO [train.py:421] (4/8) Epoch 5, batch 32000, loss[loss=2.849, over 560.00 frames. , ppl: 17.275678090369436] tot_loss[loss=2.299, over 5484217.54 frames. , ppl: 9.959403882883821], batch size: 70
+2022-12-11 17:13:39,198 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:13:39,950 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.866756836775595
+2022-12-11 17:15:20,468 INFO [train.py:421] (4/8) Epoch 5, batch 32200, loss[loss=2.166, over 8190.00 frames. , ppl: 8.723332360838713] tot_loss[loss=2.298, over 5507792.06 frames. , ppl: 9.951058355496885], batch size: 70
+2022-12-11 17:17:01,718 INFO [train.py:421] (4/8) Epoch 5, batch 32400, loss[loss=2.187, over 5040.00 frames. , ppl: 8.90492960384097] tot_loss[loss=2.296, over 5549057.39 frames. , ppl: 9.936582848546443], batch size: 70
+2022-12-11 17:18:38,659 INFO [train.py:421] (4/8) Epoch 5, batch 32600, loss[loss=2.197, over 6160.00 frames. , ppl: 9.00183921661339] tot_loss[loss=2.297, over 5521885.48 frames. , ppl: 9.942584291340795], batch size: 70
+2022-12-11 17:20:17,508 INFO [train.py:421] (4/8) Epoch 5, batch 32800, loss[loss=2.169, over 3010.00 frames. , ppl: 8.751921888942881] tot_loss[loss=2.297, over 5510790.19 frames. , ppl: 9.948623714577526], batch size: 70
+2022-12-11 17:21:54,550 INFO [train.py:421] (4/8) Epoch 5, batch 33000, loss[loss=2.222, over 3570.00 frames. , ppl: 9.221236424299228] tot_loss[loss=2.298, over 5465584.67 frames. , ppl: 9.956713165350223], batch size: 70
+2022-12-11 17:21:54,551 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:21:55,298 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.29, over 211138.00 frames. , ppl: 9.874875360260138
+2022-12-11 17:23:34,440 INFO [train.py:421] (4/8) Epoch 5, batch 33200, loss[loss=2.228, over 10920.00 frames. , ppl: 9.279842282951227] tot_loss[loss=2.297, over 5487493.05 frames. , ppl: 9.939808700168006], batch size: 70
+2022-12-11 17:25:20,359 INFO [train.py:421] (4/8) Epoch 5, batch 33400, loss[loss=2.15, over 8470.00 frames. , ppl: 8.585373111356482] tot_loss[loss=2.295, over 5544515.03 frames. , ppl: 9.925319651059903], batch size: 70
+2022-12-11 17:27:01,179 INFO [train.py:421] (4/8) Epoch 5, batch 33600, loss[loss=2.416, over 1330.00 frames. , ppl: 11.20127233423954] tot_loss[loss=2.295, over 5527335.35 frames. , ppl: 9.928143463227585], batch size: 70
+2022-12-11 17:28:43,675 INFO [train.py:421] (4/8) Epoch 5, batch 33800, loss[loss=2.486, over 1120.00 frames. , ppl: 12.015770623461215] tot_loss[loss=2.293, over 5582329.43 frames. , ppl: 9.907175844709423], batch size: 70
+2022-12-11 17:30:21,119 INFO [train.py:421] (4/8) Epoch 5, batch 34000, loss[loss=2.398, over 2170.00 frames. , ppl: 11.0057902864097] tot_loss[loss=2.295, over 5531169.76 frames. , ppl: 9.924445383124887], batch size: 70
+2022-12-11 17:30:21,119 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:30:21,882 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.858973302750663
+2022-12-11 17:32:01,801 INFO [train.py:421] (4/8) Epoch 5, batch 34200, loss[loss=2.418, over 1330.00 frames. , ppl: 11.228879332261005] tot_loss[loss=2.295, over 5536426.35 frames. , ppl: 9.921331329735837], batch size: 70
+2022-12-11 17:33:39,848 INFO [train.py:421] (4/8) Epoch 5, batch 34400, loss[loss=2.623, over 1260.00 frames. , ppl: 13.780190802916676] tot_loss[loss=2.295, over 5525223.38 frames. , ppl: 9.928360677474473], batch size: 70
+2022-12-11 17:35:26,817 INFO [train.py:421] (4/8) Epoch 5, batch 34600, loss[loss=2.464, over 1610.00 frames. , ppl: 11.74814349382833] tot_loss[loss=2.296, over 5527467.74 frames. , ppl: 9.932298999484866], batch size: 70
+2022-12-11 17:37:05,670 INFO [train.py:421] (4/8) Epoch 5, batch 34800, loss[loss=2.318, over 3500.00 frames. , ppl: 10.151347398645491] tot_loss[loss=2.296, over 5508579.27 frames. , ppl: 9.938207581325605], batch size: 70
+2022-12-11 17:38:42,833 INFO [train.py:421] (4/8) Epoch 5, batch 35000, loss[loss=2.511, over 1050.00 frames. , ppl: 12.31961042563468] tot_loss[loss=2.297, over 5490780.57 frames. , ppl: 9.946924909711], batch size: 70
+2022-12-11 17:38:42,834 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:38:43,595 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.860366938807065
+2022-12-11 17:40:23,339 INFO [train.py:421] (4/8) Epoch 5, batch 35200, loss[loss=2.247, over 2730.00 frames. , ppl: 9.457921485478739] tot_loss[loss=2.298, over 5468774.25 frames. , ppl: 9.958340526117327], batch size: 70
+2022-12-11 17:42:03,890 INFO [train.py:421] (4/8) Epoch 5, batch 35400, loss[loss=2.744, over 630.00 frames. , ppl: 15.551927859834255] tot_loss[loss=2.299, over 5472775.60 frames. , ppl: 9.960679974453639], batch size: 70
+2022-12-11 17:44:04,291 INFO [train.py:421] (4/8) Epoch 5, batch 35600, loss[loss=2.616, over 770.00 frames. , ppl: 13.685807189420531] tot_loss[loss=2.298, over 5481346.58 frames. , ppl: 9.95457813516517], batch size: 70
+2022-12-11 17:45:45,024 INFO [train.py:421] (4/8) Epoch 5, batch 35800, loss[loss=2.473, over 1890.00 frames. , ppl: 11.853218250489189] tot_loss[loss=2.299, over 5484892.02 frames. , ppl: 9.963058303185184], batch size: 70
+2022-12-11 17:47:22,426 INFO [train.py:421] (4/8) Epoch 5, batch 36000, loss[loss=2.188, over 6930.00 frames. , ppl: 8.919434005977417] tot_loss[loss=2.299, over 5495325.06 frames. , ppl: 9.960862596670285], batch size: 70
+2022-12-11 17:47:22,427 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:47:23,172 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.84657788756117
+2022-12-11 17:49:05,398 INFO [train.py:421] (4/8) Epoch 5, batch 36200, loss[loss=2.526, over 1400.00 frames. , ppl: 12.49774552559352] tot_loss[loss=2.3, over 5456896.43 frames. , ppl: 9.969921287852495], batch size: 70
+2022-12-11 17:50:49,468 INFO [train.py:421] (4/8) Epoch 5, batch 36400, loss[loss=2.411, over 1890.00 frames. , ppl: 11.141398695182007] tot_loss[loss=2.299, over 5471923.90 frames. , ppl: 9.964012092682076], batch size: 70
+2022-12-11 17:52:28,479 INFO [train.py:421] (4/8) Epoch 5, batch 36600, loss[loss=2.162, over 3850.00 frames. , ppl: 8.691914383024871] tot_loss[loss=2.298, over 5486352.50 frames. , ppl: 9.953431162649414], batch size: 70
+2022-12-11 17:54:09,081 INFO [train.py:421] (4/8) Epoch 5, batch 36800, loss[loss=2.242, over 3360.00 frames. , ppl: 9.416140078239364] tot_loss[loss=2.297, over 5520403.54 frames. , ppl: 9.944678472337625], batch size: 70
+2022-12-11 17:55:52,436 INFO [train.py:421] (4/8) Epoch 5, batch 37000, loss[loss=2.22, over 6300.00 frames. , ppl: 9.20739651858715] tot_loss[loss=2.297, over 5525208.72 frames. , ppl: 9.948634290633253], batch size: 70
+2022-12-11 17:55:52,437 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 17:55:53,182 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.841108451353056
+2022-12-11 17:57:35,869 INFO [train.py:421] (4/8) Epoch 5, batch 37200, loss[loss=2.431, over 2310.00 frames. , ppl: 11.369710022584114] tot_loss[loss=2.298, over 5511600.81 frames. , ppl: 9.958506038044362], batch size: 70
+2022-12-11 17:59:14,079 INFO [train.py:421] (4/8) Epoch 5, batch 37400, loss[loss=2.404, over 1960.00 frames. , ppl: 11.071076505826264] tot_loss[loss=2.299, over 5487157.88 frames. , ppl: 9.960582239651186], batch size: 70
+2022-12-11 18:00:54,462 INFO [train.py:421] (4/8) Epoch 5, batch 37600, loss[loss=2.383, over 1820.00 frames. , ppl: 10.841658938641936] tot_loss[loss=2.3, over 5445250.45 frames. , ppl: 9.970121915809404], batch size: 70
+2022-12-11 18:02:36,711 INFO [train.py:421] (4/8) Epoch 5, batch 37800, loss[loss=2.48, over 910.00 frames. , ppl: 11.943706511553353] tot_loss[loss=2.298, over 5491108.66 frames. , ppl: 9.957179112960835], batch size: 70
+2022-12-11 18:04:14,228 INFO [train.py:421] (4/8) Epoch 5, batch 38000, loss[loss=2.294, over 3850.00 frames. , ppl: 9.916519560007792] tot_loss[loss=2.299, over 5472704.76 frames. , ppl: 9.963468019032476], batch size: 70
+2022-12-11 18:04:14,229 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:04:14,993 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.84971901441481
+2022-12-11 18:05:55,141 INFO [train.py:421] (4/8) Epoch 5, batch 38200, loss[loss=2.291, over 3710.00 frames. , ppl: 9.886829987595474] tot_loss[loss=2.299, over 5485391.66 frames. , ppl: 9.961732691653388], batch size: 70
+2022-12-11 18:07:31,754 INFO [train.py:421] (4/8) Epoch 5, batch 38400, loss[loss=2.392, over 2520.00 frames. , ppl: 10.937105126516165] tot_loss[loss=2.3, over 5446918.67 frames. , ppl: 9.97065214688034], batch size: 70
+2022-12-11 18:09:11,040 INFO [train.py:421] (4/8) Epoch 5, batch 38600, loss[loss=2.764, over 910.00 frames. , ppl: 15.858368819470403] tot_loss[loss=2.299, over 5480949.81 frames. , ppl: 9.959420198761379], batch size: 70
+2022-12-11 18:10:51,067 INFO [train.py:421] (4/8) Epoch 5, batch 38800, loss[loss=2.491, over 1470.00 frames. , ppl: 12.069823031966246] tot_loss[loss=2.299, over 5518212.69 frames. , ppl: 9.964076069101337], batch size: 70
+2022-12-11 18:12:31,786 INFO [train.py:421] (4/8) Epoch 5, batch 39000, loss[loss=2.501, over 1190.00 frames. , ppl: 12.199366334991552] tot_loss[loss=2.3, over 5489311.49 frames. , ppl: 9.97317391384239], batch size: 70
+2022-12-11 18:12:31,787 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:12:32,547 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.839695691985794
+2022-12-11 18:14:11,127 INFO [train.py:421] (4/8) Epoch 5, batch 39200, loss[loss=2.326, over 3150.00 frames. , ppl: 10.240322213899766] tot_loss[loss=2.3, over 5507410.49 frames. , ppl: 9.969676196566196], batch size: 70
+2022-12-11 18:15:49,010 INFO [train.py:421] (4/8) Epoch 5, batch 39400, loss[loss=2.994, over 630.00 frames. , ppl: 19.97403939209564] tot_loss[loss=2.301, over 5467810.16 frames. , ppl: 9.981942213894902], batch size: 70
+2022-12-11 18:17:29,910 INFO [train.py:421] (4/8) Epoch 5, batch 39600, loss[loss=2.578, over 910.00 frames. , ppl: 13.168593497495399] tot_loss[loss=2.302, over 5437840.76 frames. , ppl: 9.99203155900321], batch size: 70
+2022-12-11 18:19:09,420 INFO [train.py:421] (4/8) Epoch 5, batch 39800, loss[loss=2.383, over 3150.00 frames. , ppl: 10.840579358855997] tot_loss[loss=2.303, over 5431034.99 frames. , ppl: 10.004845938127316], batch size: 70
+2022-12-11 18:20:46,138 INFO [train.py:421] (4/8) Epoch 5, batch 40000, loss[loss=2.367, over 3150.00 frames. , ppl: 10.66356219212582] tot_loss[loss=2.303, over 5409053.04 frames. , ppl: 10.003827200829209], batch size: 70
+2022-12-11 18:20:46,139 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:20:46,898 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.289, over 211138.00 frames. , ppl: 9.863040936604593
+2022-12-11 18:22:26,580 INFO [train.py:421] (4/8) Epoch 5, batch 40200, loss[loss=2.324, over 5110.00 frames. , ppl: 10.214452847064665] tot_loss[loss=2.302, over 5430881.10 frames. , ppl: 9.997173224653322], batch size: 70
+2022-12-11 18:24:04,874 INFO [train.py:421] (4/8) Epoch 5, batch 40400, loss[loss=3.592, over 420.00 frames. , ppl: 36.28920021592332] tot_loss[loss=2.303, over 5384521.25 frames. , ppl: 10.006597574083147], batch size: 70
+2022-12-11 18:25:44,573 INFO [train.py:421] (4/8) Epoch 5, batch 40600, loss[loss=2.25, over 3850.00 frames. , ppl: 9.491779767178766] tot_loss[loss=2.302, over 5400041.43 frames. , ppl: 9.998124621386724], batch size: 70
+2022-12-11 18:27:25,885 INFO [train.py:421] (4/8) Epoch 5, batch 40800, loss[loss=2.416, over 1330.00 frames. , ppl: 11.198829888892565] tot_loss[loss=2.302, over 5425308.55 frames. , ppl: 9.991764287774739], batch size: 70
+2022-12-11 18:29:07,910 INFO [train.py:421] (4/8) Epoch 5, batch 41000, loss[loss=2.163, over 5180.00 frames. , ppl: 8.700892969742055] tot_loss[loss=2.303, over 5402651.53 frames. , ppl: 9.999210221313213], batch size: 70
+2022-12-11 18:29:07,911 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:29:08,657 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.844880201345818
+2022-12-11 18:30:49,865 INFO [train.py:421] (4/8) Epoch 5, batch 41200, loss[loss=2.377, over 2100.00 frames. , ppl: 10.7776937918814] tot_loss[loss=2.301, over 5447186.32 frames. , ppl: 9.981088125857225], batch size: 70
+2022-12-11 18:32:28,300 INFO [train.py:421] (4/8) Epoch 5, batch 41400, loss[loss=2.654, over 770.00 frames. , ppl: 14.208762456135661] tot_loss[loss=2.302, over 5413504.76 frames. , ppl: 9.99115161470857], batch size: 70
+2022-12-11 18:34:08,974 INFO [train.py:421] (4/8) Epoch 5, batch 41600, loss[loss=2.422, over 2170.00 frames. , ppl: 11.27337596468741] tot_loss[loss=2.301, over 5429602.34 frames. , ppl: 9.987668830587841], batch size: 70
+2022-12-11 18:35:50,123 INFO [train.py:421] (4/8) Epoch 5, batch 41800, loss[loss=2.277, over 2380.00 frames. , ppl: 9.750785571537367] tot_loss[loss=2.301, over 5459690.06 frames. , ppl: 9.979199160241743], batch size: 70
+2022-12-11 18:37:33,725 INFO [train.py:421] (4/8) Epoch 5, batch 42000, loss[loss=2.538, over 1190.00 frames. , ppl: 12.648190431566611] tot_loss[loss=2.301, over 5454877.29 frames. , ppl: 9.979856346424894], batch size: 70
+2022-12-11 18:37:33,725 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:37:34,472 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.285, over 211138.00 frames. , ppl: 9.829064270772456
+2022-12-11 18:39:16,852 INFO [train.py:421] (4/8) Epoch 5, batch 42200, loss[loss=2.245, over 6930.00 frames. , ppl: 9.441692223140949] tot_loss[loss=2.3, over 5474231.76 frames. , ppl: 9.976226197157207], batch size: 70
+2022-12-11 18:40:58,476 INFO [train.py:421] (4/8) Epoch 5, batch 42400, loss[loss=2.4, over 1610.00 frames. , ppl: 11.018767697749686] tot_loss[loss=2.3, over 5446812.86 frames. , ppl: 9.977041639706849], batch size: 70
+2022-12-11 18:42:36,705 INFO [train.py:421] (4/8) Epoch 5, batch 42600, loss[loss=3.131, over 490.00 frames. , ppl: 22.90706349356313] tot_loss[loss=2.301, over 5431559.13 frames. , ppl: 9.979224125952562], batch size: 70
+2022-12-11 18:44:18,990 INFO [train.py:421] (4/8) Epoch 5, batch 42800, loss[loss=2.297, over 5110.00 frames. , ppl: 9.94110477277744] tot_loss[loss=2.3, over 5448701.84 frames. , ppl: 9.978164210274077], batch size: 70
+2022-12-11 18:45:56,834 INFO [train.py:421] (4/8) Epoch 5, batch 43000, loss[loss=2.184, over 4200.00 frames. , ppl: 8.878435615189812] tot_loss[loss=2.3, over 5470385.68 frames. , ppl: 9.969880404622723], batch size: 70
+2022-12-11 18:45:56,835 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:45:57,597 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.842544721854633
+2022-12-11 18:47:37,836 INFO [train.py:421] (4/8) Epoch 5, batch 43200, loss[loss=2.94, over 630.00 frames. , ppl: 18.919796322365073] tot_loss[loss=2.297, over 5522878.57 frames. , ppl: 9.949232772675366], batch size: 70
+2022-12-11 18:49:19,128 INFO [train.py:421] (4/8) Epoch 5, batch 43400, loss[loss=2.549, over 1610.00 frames. , ppl: 12.79531695475226] tot_loss[loss=2.298, over 5527512.31 frames. , ppl: 9.949954680518996], batch size: 70
+2022-12-11 18:51:01,641 INFO [train.py:421] (4/8) Epoch 5, batch 43600, loss[loss=2.285, over 2310.00 frames. , ppl: 9.827026347856032] tot_loss[loss=2.297, over 5539794.21 frames. , ppl: 9.947809611800066], batch size: 70
+2022-12-11 18:52:42,485 INFO [train.py:421] (4/8) Epoch 5, batch 43800, loss[loss=2.307, over 2520.00 frames. , ppl: 10.046329862756075] tot_loss[loss=2.296, over 5556035.30 frames. , ppl: 9.934863361382783], batch size: 70
+2022-12-11 18:54:22,154 INFO [train.py:421] (4/8) Epoch 5, batch 44000, loss[loss=2.415, over 1960.00 frames. , ppl: 11.186940713280793] tot_loss[loss=2.295, over 5584474.06 frames. , ppl: 9.927685515005319], batch size: 70
+2022-12-11 18:54:22,154 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 18:54:22,915 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.849284590078504
+2022-12-11 18:56:01,422 INFO [train.py:421] (4/8) Epoch 5, batch 44200, loss[loss=2.386, over 1610.00 frames. , ppl: 10.870398324147724] tot_loss[loss=2.297, over 5526470.14 frames. , ppl: 9.94323912517892], batch size: 70
+2022-12-11 18:57:41,930 INFO [train.py:421] (4/8) Epoch 5, batch 44400, loss[loss=2.304, over 2030.00 frames. , ppl: 10.015624278967158] tot_loss[loss=2.297, over 5503191.92 frames. , ppl: 9.94740899851826], batch size: 70
+2022-12-11 18:59:21,017 INFO [train.py:421] (4/8) Epoch 5, batch 44600, loss[loss=2.358, over 1330.00 frames. , ppl: 10.565345715379783] tot_loss[loss=2.298, over 5466091.02 frames. , ppl: 9.956861114532522], batch size: 70
+2022-12-11 19:01:02,198 INFO [train.py:421] (4/8) Epoch 5, batch 44800, loss[loss=2.956, over 560.00 frames. , ppl: 19.214762913473862] tot_loss[loss=2.298, over 5491058.61 frames. , ppl: 9.952051067728734], batch size: 70
+2022-12-11 19:02:40,862 INFO [train.py:421] (4/8) Epoch 5, batch 45000, loss[loss=2.748, over 840.00 frames. , ppl: 15.61148805201316] tot_loss[loss=2.3, over 5442096.32 frames. , ppl: 9.970726380583903], batch size: 70
+2022-12-11 19:02:40,862 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:02:41,620 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.835643508977666
+2022-12-11 19:04:22,576 INFO [train.py:421] (4/8) Epoch 5, batch 45200, loss[loss=2.248, over 4830.00 frames. , ppl: 9.465117362364483] tot_loss[loss=2.299, over 5472169.17 frames. , ppl: 9.964203399628687], batch size: 70
+2022-12-11 19:06:01,352 INFO [train.py:421] (4/8) Epoch 5, batch 45400, loss[loss=2.655, over 630.00 frames. , ppl: 14.221836545602445] tot_loss[loss=2.299, over 5467089.27 frames. , ppl: 9.965330692145848], batch size: 70
+2022-12-11 19:07:43,218 INFO [train.py:421] (4/8) Epoch 5, batch 45600, loss[loss=2.332, over 1890.00 frames. , ppl: 10.302009205346598] tot_loss[loss=2.299, over 5484174.90 frames. , ppl: 9.960964061890564], batch size: 70
+2022-12-11 19:09:21,551 INFO [train.py:421] (4/8) Epoch 5, batch 45800, loss[loss=2.386, over 1260.00 frames. , ppl: 10.8669124494631] tot_loss[loss=2.299, over 5509481.47 frames. , ppl: 9.962580674567358], batch size: 70
+2022-12-11 19:11:03,159 INFO [train.py:421] (4/8) Epoch 5, batch 46000, loss[loss=2.581, over 1050.00 frames. , ppl: 13.209055574687772] tot_loss[loss=2.298, over 5493478.95 frames. , ppl: 9.959135877980916], batch size: 70
+2022-12-11 19:11:03,159 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:11:03,906 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.285, over 211138.00 frames. , ppl: 9.824885591047718
+2022-12-11 19:12:44,281 INFO [train.py:421] (4/8) Epoch 5, batch 46200, loss[loss=2.318, over 1610.00 frames. , ppl: 10.153732504084207] tot_loss[loss=2.3, over 5446688.74 frames. , ppl: 9.97170760335541], batch size: 70
+2022-12-11 19:14:23,630 INFO [train.py:421] (4/8) Epoch 5, batch 46400, loss[loss=2.536, over 1330.00 frames. , ppl: 12.625851386987309] tot_loss[loss=2.299, over 5479235.71 frames. , ppl: 9.96262742528925], batch size: 70
+2022-12-11 19:15:58,261 INFO [train.py:421] (4/8) Epoch 5, batch 46600, loss[loss=2.328, over 2800.00 frames. , ppl: 10.257282411432033] tot_loss[loss=2.3, over 5448869.29 frames. , ppl: 9.970599948847996], batch size: 70
+2022-12-11 19:17:38,452 INFO [train.py:421] (4/8) Epoch 5, batch 46800, loss[loss=2.406, over 1190.00 frames. , ppl: 11.084370384047856] tot_loss[loss=2.299, over 5459028.92 frames. , ppl: 9.966071300957484], batch size: 70
+2022-12-11 19:19:20,585 INFO [train.py:421] (4/8) Epoch 5, batch 47000, loss[loss=2.204, over 6790.00 frames. , ppl: 9.065200842821303] tot_loss[loss=2.3, over 5438094.05 frames. , ppl: 9.969504802492303], batch size: 70
+2022-12-11 19:19:20,586 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:19:21,331 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.837504131725721
+2022-12-11 19:21:03,345 INFO [train.py:421] (4/8) Epoch 5, batch 47200, loss[loss=2.305, over 5040.00 frames. , ppl: 10.019541480087678] tot_loss[loss=2.298, over 5514644.39 frames. , ppl: 9.953604499857997], batch size: 70
+2022-12-11 19:22:43,309 INFO [train.py:421] (4/8) Epoch 5, batch 47400, loss[loss=2.382, over 1190.00 frames. , ppl: 10.826137421814257] tot_loss[loss=2.298, over 5485562.48 frames. , ppl: 9.957331591127613], batch size: 70
+2022-12-11 19:24:25,849 INFO [train.py:421] (4/8) Epoch 5, batch 47600, loss[loss=2.278, over 3430.00 frames. , ppl: 9.759002565752454] tot_loss[loss=2.297, over 5515397.19 frames. , ppl: 9.94484087215752], batch size: 70
+2022-12-11 19:26:04,799 INFO [train.py:421] (4/8) Epoch 5, batch 47800, loss[loss=2.269, over 2590.00 frames. , ppl: 9.666434344693661] tot_loss[loss=2.298, over 5487183.20 frames. , ppl: 9.953424291126078], batch size: 70
+2022-12-11 19:27:44,266 INFO [train.py:421] (4/8) Epoch 5, batch 48000, loss[loss=2.196, over 5950.00 frames. , ppl: 8.990223744654724] tot_loss[loss=2.296, over 5510306.45 frames. , ppl: 9.939163860496794], batch size: 70
+2022-12-11 19:27:44,267 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:27:45,017 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.847490241584369
+2022-12-11 19:29:31,063 INFO [train.py:421] (4/8) Epoch 5, batch 48200, loss[loss=2.427, over 1050.00 frames. , ppl: 11.327393604431625] tot_loss[loss=2.297, over 5483974.56 frames. , ppl: 9.945340605064413], batch size: 70
+2022-12-11 19:31:11,713 INFO [train.py:421] (4/8) Epoch 5, batch 48400, loss[loss=2.289, over 3990.00 frames. , ppl: 9.861237480720098] tot_loss[loss=2.297, over 5500064.38 frames. , ppl: 9.940288185280435], batch size: 70
+2022-12-11 19:32:50,746 INFO [train.py:421] (4/8) Epoch 5, batch 48600, loss[loss=2.38, over 1890.00 frames. , ppl: 10.809421483466414] tot_loss[loss=2.298, over 5434491.56 frames. , ppl: 9.949902925332076], batch size: 70
+2022-12-11 19:34:30,309 INFO [train.py:421] (4/8) Epoch 5, batch 48800, loss[loss=2.318, over 2310.00 frames. , ppl: 10.157898423220573] tot_loss[loss=2.298, over 5427769.12 frames. , ppl: 9.953800201457264], batch size: 70
+2022-12-11 19:36:12,832 INFO [train.py:421] (4/8) Epoch 5, batch 49000, loss[loss=2.333, over 2660.00 frames. , ppl: 10.312592930270986] tot_loss[loss=2.297, over 5447763.78 frames. , ppl: 9.947675934744394], batch size: 70
+2022-12-11 19:36:12,833 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:36:13,593 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.8508430669708
+2022-12-11 19:37:58,725 INFO [train.py:421] (4/8) Epoch 5, batch 49200, loss[loss=2.395, over 3290.00 frames. , ppl: 10.969633249172743] tot_loss[loss=2.297, over 5477195.55 frames. , ppl: 9.940647388874867], batch size: 70
+2022-12-11 19:39:40,033 INFO [train.py:421] (4/8) Epoch 5, batch 49400, loss[loss=2.393, over 1330.00 frames. , ppl: 10.942590544438263] tot_loss[loss=2.296, over 5480621.56 frames. , ppl: 9.937465680675416], batch size: 70
+2022-12-11 19:41:17,702 INFO [train.py:421] (4/8) Epoch 5, batch 49600, loss[loss=2.775, over 910.00 frames. , ppl: 16.04024498477591] tot_loss[loss=2.298, over 5437715.09 frames. , ppl: 9.951215979134576], batch size: 70
+2022-12-11 19:42:57,230 INFO [train.py:421] (4/8) Epoch 5, batch 49800, loss[loss=2.335, over 2310.00 frames. , ppl: 10.332611510182225] tot_loss[loss=2.298, over 5418073.92 frames. , ppl: 9.95878488561751], batch size: 70
+2022-12-11 19:44:39,905 INFO [train.py:421] (4/8) Epoch 5, batch 50000, loss[loss=2.255, over 2590.00 frames. , ppl: 9.530561999662261] tot_loss[loss=2.298, over 5444339.04 frames. , ppl: 9.95274161976204], batch size: 70
+2022-12-11 19:44:39,906 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:44:40,663 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.837191091537502
+2022-12-11 19:46:20,685 INFO [train.py:421] (4/8) Epoch 5, batch 50200, loss[loss=2.807, over 630.00 frames. , ppl: 16.56362662184319] tot_loss[loss=2.298, over 5478965.82 frames. , ppl: 9.950743087314908], batch size: 70
+2022-12-11 19:48:00,103 INFO [train.py:421] (4/8) Epoch 5, batch 50400, loss[loss=2.223, over 5600.00 frames. , ppl: 9.232488482575286] tot_loss[loss=2.298, over 5499033.52 frames. , ppl: 9.951764816243395], batch size: 70
+2022-12-11 19:49:38,962 INFO [train.py:421] (4/8) Epoch 5, batch 50600, loss[loss=2.727, over 700.00 frames. , ppl: 15.281773653391815] tot_loss[loss=2.298, over 5497921.89 frames. , ppl: 9.952384264128169], batch size: 70
+2022-12-11 19:51:17,597 INFO [train.py:421] (4/8) Epoch 5, batch 50800, loss[loss=2.464, over 2030.00 frames. , ppl: 11.74999712694134] tot_loss[loss=2.298, over 5494697.32 frames. , ppl: 9.949426735621108], batch size: 70
+2022-12-11 19:52:59,276 INFO [train.py:421] (4/8) Epoch 5, batch 51000, loss[loss=2.173, over 4900.00 frames. , ppl: 8.78167891727247] tot_loss[loss=2.298, over 5492644.04 frames. , ppl: 9.952955907984794], batch size: 70
+2022-12-11 19:52:59,276 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 19:53:00,024 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.288, over 211138.00 frames. , ppl: 9.852051821719042
+2022-12-11 19:54:41,942 INFO [train.py:421] (4/8) Epoch 5, batch 51200, loss[loss=2.302, over 2800.00 frames. , ppl: 9.996995852310281] tot_loss[loss=2.298, over 5495612.84 frames. , ppl: 9.955993135089809], batch size: 70
+2022-12-11 19:56:21,646 INFO [train.py:421] (4/8) Epoch 5, batch 51400, loss[loss=2.467, over 1470.00 frames. , ppl: 11.79067234808132] tot_loss[loss=2.298, over 5490193.14 frames. , ppl: 9.954867539871461], batch size: 70
+2022-12-11 19:58:02,206 INFO [train.py:421] (4/8) Epoch 5, batch 51600, loss[loss=2.296, over 2450.00 frames. , ppl: 9.932190527935774] tot_loss[loss=2.3, over 5474356.31 frames. , ppl: 9.970946420057558], batch size: 70
+2022-12-11 19:59:40,709 INFO [train.py:421] (4/8) Epoch 5, batch 51800, loss[loss=2.502, over 1120.00 frames. , ppl: 12.207745382031243] tot_loss[loss=2.3, over 5465921.08 frames. , ppl: 9.972496134188255], batch size: 70
+2022-12-11 20:01:19,072 INFO [train.py:421] (4/8) Epoch 5, batch 52000, loss[loss=2.297, over 2520.00 frames. , ppl: 9.945397089140835] tot_loss[loss=2.3, over 5474097.74 frames. , ppl: 9.970265592178047], batch size: 70
+2022-12-11 20:01:19,072 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:01:19,832 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.840256402098536
+2022-12-11 20:03:00,929 INFO [train.py:421] (4/8) Epoch 5, batch 52200, loss[loss=3.228, over 490.00 frames. , ppl: 25.234772522158323] tot_loss[loss=2.298, over 5538636.01 frames. , ppl: 9.955791525076386], batch size: 70
+2022-12-11 20:04:40,909 INFO [train.py:421] (4/8) Epoch 5, batch 52400, loss[loss=2.535, over 1190.00 frames. , ppl: 12.617861273068412] tot_loss[loss=2.298, over 5524860.70 frames. , ppl: 9.958018352570967], batch size: 70
+2022-12-11 20:06:20,201 INFO [train.py:421] (4/8) Epoch 5, batch 52600, loss[loss=2.396, over 1190.00 frames. , ppl: 10.984655580141183] tot_loss[loss=2.298, over 5510293.46 frames. , ppl: 9.952639473352233], batch size: 70
+2022-12-11 20:07:56,000 INFO [train.py:421] (4/8) Epoch 5, batch 52800, loss[loss=2.261, over 3360.00 frames. , ppl: 9.597328532675668] tot_loss[loss=2.298, over 5516917.58 frames. , ppl: 9.949924432654203], batch size: 70
+2022-12-11 20:09:33,171 INFO [train.py:421] (4/8) Epoch 5, batch 53000, loss[loss=2.288, over 1960.00 frames. , ppl: 9.85166037703486] tot_loss[loss=2.298, over 5486988.60 frames. , ppl: 9.958138516020085], batch size: 70
+2022-12-11 20:09:33,172 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:09:33,930 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.842675832035669
+2022-12-11 20:11:15,343 INFO [train.py:421] (4/8) Epoch 5, batch 53200, loss[loss=2.307, over 2870.00 frames. , ppl: 10.042084653440027] tot_loss[loss=2.297, over 5506241.82 frames. , ppl: 9.94612704861288], batch size: 70
+2022-12-11 20:12:50,608 INFO [train.py:421] (4/8) Epoch 5, batch 53400, loss[loss=2.471, over 1400.00 frames. , ppl: 11.833424158468276] tot_loss[loss=2.297, over 5524975.60 frames. , ppl: 9.946164646921913], batch size: 70
+2022-12-11 20:14:31,569 INFO [train.py:421] (4/8) Epoch 5, batch 53600, loss[loss=2.95, over 560.00 frames. , ppl: 19.100856732934947] tot_loss[loss=2.297, over 5530641.68 frames. , ppl: 9.94101307723609], batch size: 70
+2022-12-11 20:16:11,456 INFO [train.py:421] (4/8) Epoch 5, batch 53800, loss[loss=2.284, over 2870.00 frames. , ppl: 9.813132305039986] tot_loss[loss=2.297, over 5514333.91 frames. , ppl: 9.948582072322148], batch size: 70
+2022-12-11 20:17:55,393 INFO [train.py:421] (4/8) Epoch 5, batch 54000, loss[loss=2.265, over 6090.00 frames. , ppl: 9.62961367882041] tot_loss[loss=2.296, over 5549420.25 frames. , ppl: 9.93655108073102], batch size: 70
+2022-12-11 20:17:55,394 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:17:56,155 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.832751361696745
+2022-12-11 20:19:35,175 INFO [train.py:421] (4/8) Epoch 5, batch 54200, loss[loss=2.245, over 2730.00 frames. , ppl: 9.438297076847185] tot_loss[loss=2.296, over 5587605.97 frames. , ppl: 9.934761181178736], batch size: 70
+2022-12-11 20:21:11,752 INFO [train.py:421] (4/8) Epoch 5, batch 54400, loss[loss=2.38, over 2450.00 frames. , ppl: 10.80878614287537] tot_loss[loss=2.296, over 5578429.62 frames. , ppl: 9.932923797999011], batch size: 70
+2022-12-11 20:22:53,112 INFO [train.py:421] (4/8) Epoch 5, batch 54600, loss[loss=2.374, over 1820.00 frames. , ppl: 10.742072534385237] tot_loss[loss=2.296, over 5572501.52 frames. , ppl: 9.934059526454432], batch size: 70
+2022-12-11 20:24:36,518 INFO [train.py:421] (4/8) Epoch 5, batch 54800, loss[loss=2.597, over 910.00 frames. , ppl: 13.417020417724755] tot_loss[loss=2.297, over 5526569.25 frames. , ppl: 9.9423954731209], batch size: 70
+2022-12-11 20:26:21,050 INFO [train.py:421] (4/8) Epoch 5, batch 55000, loss[loss=2.71, over 700.00 frames. , ppl: 15.025305379395665] tot_loss[loss=2.296, over 5547938.49 frames. , ppl: 9.935857323317945], batch size: 70
+2022-12-11 20:26:21,051 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:26:21,809 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.833813803126963
+2022-12-11 20:27:59,954 INFO [train.py:421] (4/8) Epoch 5, batch 55200, loss[loss=2.355, over 2240.00 frames. , ppl: 10.540201995513334] tot_loss[loss=2.297, over 5511734.51 frames. , ppl: 9.943126335354952], batch size: 70
+2022-12-11 20:29:36,219 INFO [train.py:421] (4/8) Epoch 5, batch 55400, loss[loss=2.263, over 3990.00 frames. , ppl: 9.614053516112836] tot_loss[loss=2.297, over 5515879.85 frames. , ppl: 9.946739618392114], batch size: 70
+2022-12-11 20:31:18,002 INFO [train.py:421] (4/8) Epoch 5, batch 55600, loss[loss=2.287, over 3570.00 frames. , ppl: 9.850242100074151] tot_loss[loss=2.297, over 5515775.58 frames. , ppl: 9.944617265724736], batch size: 70
+2022-12-11 20:32:57,832 INFO [train.py:421] (4/8) Epoch 5, batch 55800, loss[loss=2.205, over 5250.00 frames. , ppl: 9.06649163481887] tot_loss[loss=2.297, over 5553147.04 frames. , ppl: 9.941263805192627], batch size: 70
+2022-12-11 20:34:40,845 INFO [train.py:421] (4/8) Epoch 5, batch 56000, loss[loss=2.299, over 6650.00 frames. , ppl: 9.96710449303592] tot_loss[loss=2.297, over 5533849.42 frames. , ppl: 9.946398443572724], batch size: 70
+2022-12-11 20:34:40,845 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:34:41,592 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.835016101318848
+2022-12-11 20:36:19,909 INFO [train.py:421] (4/8) Epoch 5, batch 56200, loss[loss=2.889, over 560.00 frames. , ppl: 17.975257577336926] tot_loss[loss=2.298, over 5487995.88 frames. , ppl: 9.954115335198876], batch size: 70
+2022-12-11 20:38:00,524 INFO [train.py:421] (4/8) Epoch 5, batch 56400, loss[loss=2.146, over 3220.00 frames. , ppl: 8.550858858436001] tot_loss[loss=2.297, over 5487559.07 frames. , ppl: 9.947845136413962], batch size: 70
+2022-12-11 20:39:43,937 INFO [train.py:421] (4/8) Epoch 5, batch 56600, loss[loss=2.2, over 1400.00 frames. , ppl: 9.0249977610733] tot_loss[loss=2.297, over 5484818.62 frames. , ppl: 9.943245907070208], batch size: 70
+2022-12-11 20:41:26,617 INFO [train.py:421] (4/8) Epoch 5, batch 56800, loss[loss=2.417, over 1330.00 frames. , ppl: 11.208907381074587] tot_loss[loss=2.296, over 5508700.82 frames. , ppl: 9.933031517650337], batch size: 70
+2022-12-11 20:43:05,303 INFO [train.py:421] (4/8) Epoch 5, batch 57000, loss[loss=2.437, over 1470.00 frames. , ppl: 11.438469089440133] tot_loss[loss=2.294, over 5546093.36 frames. , ppl: 9.918987270876864], batch size: 70
+2022-12-11 20:43:05,304 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:43:06,065 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.842181992787078
+2022-12-11 20:44:43,547 INFO [train.py:421] (4/8) Epoch 5, batch 57200, loss[loss=2.447, over 910.00 frames. , ppl: 11.549468799672665] tot_loss[loss=2.294, over 5526580.03 frames. , ppl: 9.916555230586281], batch size: 70
+2022-12-11 20:46:22,291 INFO [train.py:421] (4/8) Epoch 5, batch 57400, loss[loss=2.772, over 770.00 frames. , ppl: 15.997899778445717] tot_loss[loss=2.295, over 5546477.72 frames. , ppl: 9.920687522320991], batch size: 70
+2022-12-11 20:48:03,932 INFO [train.py:421] (4/8) Epoch 5, batch 57600, loss[loss=2.329, over 2940.00 frames. , ppl: 10.264990976852713] tot_loss[loss=2.294, over 5554854.63 frames. , ppl: 9.917684308480943], batch size: 70
+2022-12-11 20:49:40,763 INFO [train.py:421] (4/8) Epoch 5, batch 57800, loss[loss=2.378, over 2310.00 frames. , ppl: 10.78135313528395] tot_loss[loss=2.295, over 5553712.54 frames. , ppl: 9.922940867082467], batch size: 70
+2022-12-11 20:51:21,984 INFO [train.py:421] (4/8) Epoch 5, batch 58000, loss[loss=2.568, over 840.00 frames. , ppl: 13.045732910495525] tot_loss[loss=2.296, over 5524449.23 frames. , ppl: 9.935210861041718], batch size: 70
+2022-12-11 20:51:21,985 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:51:22,729 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.846384059654563
+2022-12-11 20:53:06,173 INFO [train.py:421] (4/8) Epoch 5, batch 58200, loss[loss=2.294, over 4620.00 frames. , ppl: 9.915091362752799] tot_loss[loss=2.294, over 5580274.35 frames. , ppl: 9.909914981614557], batch size: 70
+2022-12-11 20:54:48,411 INFO [train.py:421] (4/8) Epoch 5, batch 58400, loss[loss=2.454, over 910.00 frames. , ppl: 11.63070629418433] tot_loss[loss=2.294, over 5583453.93 frames. , ppl: 9.911712648621242], batch size: 70
+2022-12-11 20:56:30,697 INFO [train.py:421] (4/8) Epoch 5, batch 58600, loss[loss=2.333, over 2100.00 frames. , ppl: 10.31190124223426] tot_loss[loss=2.293, over 5597884.89 frames. , ppl: 9.906455158208859], batch size: 70
+2022-12-11 20:58:11,550 INFO [train.py:421] (4/8) Epoch 5, batch 58800, loss[loss=2.289, over 3010.00 frames. , ppl: 9.867092725047502] tot_loss[loss=2.293, over 5591894.23 frames. , ppl: 9.905750614284285], batch size: 70
+2022-12-11 20:59:49,673 INFO [train.py:421] (4/8) Epoch 5, batch 59000, loss[loss=2.157, over 5740.00 frames. , ppl: 8.646883793762308] tot_loss[loss=2.294, over 5553068.85 frames. , ppl: 9.919402280045777], batch size: 70
+2022-12-11 20:59:49,674 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 20:59:50,433 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.833901132185455
+2022-12-11 21:01:27,627 INFO [train.py:421] (4/8) Epoch 5, batch 59200, loss[loss=2.511, over 980.00 frames. , ppl: 12.313765382958694] tot_loss[loss=2.294, over 5575878.03 frames. , ppl: 9.91640510145949], batch size: 70
+2022-12-11 21:03:06,559 INFO [train.py:421] (4/8) Epoch 5, batch 59400, loss[loss=2.297, over 4970.00 frames. , ppl: 9.942460056425467] tot_loss[loss=2.295, over 5530623.49 frames. , ppl: 9.924348473408283], batch size: 70
+2022-12-11 21:04:48,229 INFO [train.py:421] (4/8) Epoch 5, batch 59600, loss[loss=2.221, over 8610.00 frames. , ppl: 9.216696010851237] tot_loss[loss=2.294, over 5562635.51 frames. , ppl: 9.915081540996065], batch size: 70
+2022-12-11 21:06:24,265 INFO [train.py:421] (4/8) Epoch 5, batch 59800, loss[loss=2.252, over 5320.00 frames. , ppl: 9.508701831994285] tot_loss[loss=2.296, over 5500177.35 frames. , ppl: 9.936239821631837], batch size: 70
+2022-12-11 21:08:10,094 INFO [train.py:421] (4/8) Epoch 5, batch 60000, loss[loss=2.64, over 700.00 frames. , ppl: 14.016107034502227] tot_loss[loss=2.295, over 5541817.71 frames. , ppl: 9.92422396792917], batch size: 70
+2022-12-11 21:08:10,094 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:08:10,853 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.831021137755421
+2022-12-11 21:09:53,968 INFO [train.py:421] (4/8) Epoch 5, batch 60200, loss[loss=2.256, over 3920.00 frames. , ppl: 9.542692416963469] tot_loss[loss=2.294, over 5536043.36 frames. , ppl: 9.919304181987146], batch size: 70
+2022-12-11 21:11:35,809 INFO [train.py:421] (4/8) Epoch 5, batch 60400, loss[loss=2.508, over 1330.00 frames. , ppl: 12.282077430055276] tot_loss[loss=2.295, over 5529870.38 frames. , ppl: 9.923953841781321], batch size: 70
+2022-12-11 21:13:18,738 INFO [train.py:421] (4/8) Epoch 5, batch 60600, loss[loss=2.357, over 1260.00 frames. , ppl: 10.557399879517272] tot_loss[loss=2.296, over 5505402.39 frames. , ppl: 9.930556272059682], batch size: 70
+2022-12-11 21:14:58,853 INFO [train.py:421] (4/8) Epoch 5, batch 60800, loss[loss=2.328, over 3220.00 frames. , ppl: 10.257108050809736] tot_loss[loss=2.296, over 5490276.50 frames. , ppl: 9.938470357752104], batch size: 70
+2022-12-11 21:16:39,557 INFO [train.py:421] (4/8) Epoch 5, batch 61000, loss[loss=2.226, over 2940.00 frames. , ppl: 9.259410747137808] tot_loss[loss=2.297, over 5467017.08 frames. , ppl: 9.94845790385863], batch size: 70
+2022-12-11 21:16:39,558 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:16:40,311 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816234263345795
+2022-12-11 21:18:20,764 INFO [train.py:421] (4/8) Epoch 5, batch 61200, loss[loss=2.511, over 1330.00 frames. , ppl: 12.313179961380797] tot_loss[loss=2.297, over 5496142.90 frames. , ppl: 9.945622981291447], batch size: 70
+2022-12-11 21:20:01,590 INFO [train.py:421] (4/8) Epoch 5, batch 61400, loss[loss=2.39, over 3010.00 frames. , ppl: 10.909641914342096] tot_loss[loss=2.297, over 5490704.06 frames. , ppl: 9.944928494793354], batch size: 70
+2022-12-11 21:21:39,916 INFO [train.py:421] (4/8) Epoch 5, batch 61600, loss[loss=2.379, over 2590.00 frames. , ppl: 10.79612370081435] tot_loss[loss=2.297, over 5485164.99 frames. , ppl: 9.942009579251518], batch size: 70
+2022-12-11 21:23:20,543 INFO [train.py:421] (4/8) Epoch 5, batch 61800, loss[loss=2.324, over 3990.00 frames. , ppl: 10.21997673174528] tot_loss[loss=2.297, over 5467673.81 frames. , ppl: 9.946712460435904], batch size: 70
+2022-12-11 21:24:59,404 INFO [train.py:421] (4/8) Epoch 5, batch 62000, loss[loss=2.217, over 4410.00 frames. , ppl: 9.181156493435694] tot_loss[loss=2.296, over 5502798.48 frames. , ppl: 9.934896530682165], batch size: 70
+2022-12-11 21:24:59,404 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:25:00,150 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.287, over 211138.00 frames. , ppl: 9.841092429233514
+2022-12-11 21:26:42,941 INFO [train.py:421] (4/8) Epoch 5, batch 62200, loss[loss=2.286, over 2450.00 frames. , ppl: 9.839991810785115] tot_loss[loss=2.296, over 5517649.35 frames. , ppl: 9.935334920322134], batch size: 70
+2022-12-11 21:28:21,527 INFO [train.py:421] (4/8) Epoch 5, batch 62400, loss[loss=2.268, over 1820.00 frames. , ppl: 9.663629493250529] tot_loss[loss=2.296, over 5543455.65 frames. , ppl: 9.929528177571296], batch size: 70
+2022-12-11 21:30:03,693 INFO [train.py:421] (4/8) Epoch 5, batch 62600, loss[loss=2.31, over 3430.00 frames. , ppl: 10.075821045375406] tot_loss[loss=2.295, over 5556424.99 frames. , ppl: 9.924932134237967], batch size: 70
+2022-12-11 21:31:48,069 INFO [train.py:421] (4/8) Epoch 5, batch 62800, loss[loss=2.361, over 1470.00 frames. , ppl: 10.603487636161654] tot_loss[loss=2.294, over 5580591.47 frames. , ppl: 9.916247843766158], batch size: 70
+2022-12-11 21:33:23,566 INFO [train.py:421] (4/8) Epoch 5, batch 63000, loss[loss=2.448, over 980.00 frames. , ppl: 11.567245213562618] tot_loss[loss=2.295, over 5539976.68 frames. , ppl: 9.927366777642597], batch size: 70
+2022-12-11 21:33:23,567 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:33:24,313 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.831210297961553
+2022-12-11 21:35:04,615 INFO [train.py:421] (4/8) Epoch 5, batch 63200, loss[loss=3.221, over 560.00 frames. , ppl: 25.065402696418225] tot_loss[loss=2.296, over 5530364.49 frames. , ppl: 9.931013494921178], batch size: 70
+2022-12-11 21:36:45,661 INFO [train.py:421] (4/8) Epoch 5, batch 63400, loss[loss=2.33, over 3920.00 frames. , ppl: 10.276063854034938] tot_loss[loss=2.296, over 5522392.39 frames. , ppl: 9.934892783252884], batch size: 70
+2022-12-11 21:38:24,036 INFO [train.py:421] (4/8) Epoch 5, batch 63600, loss[loss=2.342, over 2450.00 frames. , ppl: 10.405558792450444] tot_loss[loss=2.297, over 5492794.94 frames. , ppl: 9.94344653176695], batch size: 70
+2022-12-11 21:40:04,571 INFO [train.py:421] (4/8) Epoch 5, batch 63800, loss[loss=2.214, over 4130.00 frames. , ppl: 9.152153425973252] tot_loss[loss=2.297, over 5484368.19 frames. , ppl: 9.943012991689342], batch size: 70
+2022-12-11 21:41:46,258 INFO [train.py:421] (4/8) Epoch 5, batch 64000, loss[loss=2.867, over 840.00 frames. , ppl: 17.579685067801684] tot_loss[loss=2.298, over 5465390.61 frames. , ppl: 9.953546996098437], batch size: 70
+2022-12-11 21:41:46,259 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:41:47,016 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.285, over 211138.00 frames. , ppl: 9.828019797907846
+2022-12-11 21:43:27,826 INFO [train.py:421] (4/8) Epoch 5, batch 64200, loss[loss=2.137, over 6790.00 frames. , ppl: 8.474765464729025] tot_loss[loss=2.299, over 5452560.79 frames. , ppl: 9.95947143888403], batch size: 70
+2022-12-11 21:45:08,808 INFO [train.py:421] (4/8) Epoch 5, batch 64400, loss[loss=2.799, over 840.00 frames. , ppl: 16.431565619803106] tot_loss[loss=2.297, over 5505171.90 frames. , ppl: 9.945914491540513], batch size: 70
+2022-12-11 21:46:48,251 INFO [train.py:421] (4/8) Epoch 5, batch 64600, loss[loss=2.353, over 1960.00 frames. , ppl: 10.515173675774642] tot_loss[loss=2.298, over 5489813.03 frames. , ppl: 9.955117473509226], batch size: 70
+2022-12-11 21:48:30,047 INFO [train.py:421] (4/8) Epoch 5, batch 64800, loss[loss=2.219, over 2940.00 frames. , ppl: 9.194207083120395] tot_loss[loss=2.298, over 5505144.87 frames. , ppl: 9.94940460210642], batch size: 70
+2022-12-11 21:50:08,871 INFO [train.py:421] (4/8) Epoch 5, batch 65000, loss[loss=2.58, over 910.00 frames. , ppl: 13.197220302128676] tot_loss[loss=2.296, over 5551017.53 frames. , ppl: 9.934910084975366], batch size: 70
+2022-12-11 21:50:08,872 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:50:09,631 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.286, over 211138.00 frames. , ppl: 9.835764336858247
+2022-12-11 21:51:51,527 INFO [train.py:421] (4/8) Epoch 5, batch 65200, loss[loss=2.513, over 1260.00 frames. , ppl: 12.337645570076674] tot_loss[loss=2.297, over 5515290.91 frames. , ppl: 9.94035518870037], batch size: 70
+2022-12-11 21:53:29,546 INFO [train.py:421] (4/8) Epoch 5, batch 65400, loss[loss=2.225, over 5880.00 frames. , ppl: 9.256014321429822] tot_loss[loss=2.298, over 5491697.35 frames. , ppl: 9.952562406753174], batch size: 70
+2022-12-11 21:55:05,996 INFO [train.py:421] (4/8) Epoch 5, batch 65600, loss[loss=2.232, over 5810.00 frames. , ppl: 9.315865847562083] tot_loss[loss=2.298, over 5469190.47 frames. , ppl: 9.959089157834564], batch size: 70
+2022-12-11 21:56:43,518 INFO [train.py:421] (4/8) Epoch 5, batch 65800, loss[loss=2.323, over 3570.00 frames. , ppl: 10.206544114821298] tot_loss[loss=2.298, over 5464936.53 frames. , ppl: 9.953898151244019], batch size: 70
+2022-12-11 21:58:21,357 INFO [train.py:421] (4/8) Epoch 5, batch 66000, loss[loss=2.329, over 1890.00 frames. , ppl: 10.267772500994564] tot_loss[loss=2.299, over 5460871.46 frames. , ppl: 9.960564275155908], batch size: 70
+2022-12-11 21:58:21,358 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 21:58:22,104 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816399892600549
+2022-12-11 22:00:01,629 INFO [train.py:421] (4/8) Epoch 5, batch 66200, loss[loss=2.341, over 3430.00 frames. , ppl: 10.390755796902274] tot_loss[loss=2.299, over 5445606.18 frames. , ppl: 9.965773044080551], batch size: 70
+2022-12-11 22:01:40,817 INFO [train.py:421] (4/8) Epoch 5, batch 66400, loss[loss=2.318, over 2170.00 frames. , ppl: 10.153824092611433] tot_loss[loss=2.298, over 5449343.40 frames. , ppl: 9.958120527175167], batch size: 70
+2022-12-11 22:03:20,545 INFO [train.py:421] (4/8) Epoch 5, batch 66600, loss[loss=2.281, over 3570.00 frames. , ppl: 9.781840919012684] tot_loss[loss=2.299, over 5430420.54 frames. , ppl: 9.964117678449032], batch size: 70
+2022-12-11 22:05:00,222 INFO [train.py:421] (4/8) Epoch 5, batch 66800, loss[loss=2.331, over 1400.00 frames. , ppl: 10.29139998700028] tot_loss[loss=2.3, over 5394239.62 frames. , ppl: 9.971436501338175], batch size: 70
+2022-12-11 22:06:37,492 INFO [train.py:421] (4/8) Epoch 5, batch 67000, loss[loss=2.364, over 2380.00 frames. , ppl: 10.635369435909652] tot_loss[loss=2.3, over 5364782.12 frames. , ppl: 9.974657473570117], batch size: 70
+2022-12-11 22:06:37,493 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 22:06:38,242 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.812880145960932
+2022-12-11 22:08:16,524 INFO [train.py:421] (4/8) Epoch 5, batch 67200, loss[loss=2.361, over 1610.00 frames. , ppl: 10.597153121427704] tot_loss[loss=2.3, over 5393720.46 frames. , ppl: 9.970526135017893], batch size: 70
+2022-12-11 22:09:55,865 INFO [train.py:421] (4/8) Epoch 5, batch 67400, loss[loss=2.469, over 1890.00 frames. , ppl: 11.809966131230446] tot_loss[loss=2.299, over 5398209.01 frames. , ppl: 9.966823440622663], batch size: 70
+2022-12-11 22:11:35,229 INFO [train.py:421] (4/8) Epoch 5, batch 67600, loss[loss=2.201, over 3990.00 frames. , ppl: 9.032548974408357] tot_loss[loss=2.299, over 5392655.53 frames. , ppl: 9.967513672051004], batch size: 70
+2022-12-11 22:13:14,516 INFO [train.py:421] (4/8) Epoch 5, batch 67800, loss[loss=2.255, over 3990.00 frames. , ppl: 9.536295024112524] tot_loss[loss=2.299, over 5399824.52 frames. , ppl: 9.965710294266401], batch size: 70
+2022-12-11 22:14:52,339 INFO [train.py:421] (4/8) Epoch 5, batch 68000, loss[loss=2.183, over 4480.00 frames. , ppl: 8.870517650346049] tot_loss[loss=2.299, over 5415825.29 frames. , ppl: 9.964960778145995], batch size: 70
+2022-12-11 22:14:52,340 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 22:14:53,100 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.814543261416496
+2022-12-11 22:16:33,228 INFO [train.py:421] (4/8) Epoch 5, batch 68200, loss[loss=2.425, over 1120.00 frames. , ppl: 11.297184907236907] tot_loss[loss=2.3, over 5396259.45 frames. , ppl: 9.96922410279795], batch size: 70
+2022-12-11 22:18:14,671 INFO [train.py:421] (4/8) Epoch 5, batch 68400, loss[loss=2.199, over 5250.00 frames. , ppl: 9.01828250455737] tot_loss[loss=2.299, over 5386003.93 frames. , ppl: 9.963709203015938], batch size: 70
+2022-12-11 22:19:55,895 INFO [train.py:421] (4/8) Epoch 5, batch 68600, loss[loss=2.315, over 1680.00 frames. , ppl: 10.127767186029127] tot_loss[loss=2.298, over 5424812.17 frames. , ppl: 9.950741071711825], batch size: 70
+2022-12-11 22:21:39,031 INFO [train.py:421] (4/8) Epoch 5, batch 68800, loss[loss=2.368, over 2100.00 frames. , ppl: 10.678403675716545] tot_loss[loss=2.298, over 5434761.20 frames. , ppl: 9.955336066291329], batch size: 70
+2022-12-11 22:23:22,204 INFO [train.py:421] (4/8) Epoch 5, batch 69000, loss[loss=2.233, over 7560.00 frames. , ppl: 9.328892637500568] tot_loss[loss=2.297, over 5475546.00 frames. , ppl: 9.942951211735613], batch size: 70
+2022-12-11 22:23:22,205 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 22:23:22,952 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816581506837656
+2022-12-11 22:25:00,975 INFO [train.py:421] (4/8) Epoch 5, batch 69200, loss[loss=2.262, over 4830.00 frames. , ppl: 9.603604434251741] tot_loss[loss=2.298, over 5445882.05 frames. , ppl: 9.949367967215013], batch size: 70
+2022-12-11 22:26:43,990 INFO [train.py:421] (4/8) Epoch 5, batch 69400, loss[loss=2.472, over 1330.00 frames. , ppl: 11.85075248726545] tot_loss[loss=2.297, over 5470153.95 frames. , ppl: 9.943003555180255], batch size: 70
+2022-12-11 22:28:25,073 INFO [train.py:421] (4/8) Epoch 5, batch 69600, loss[loss=2.364, over 1260.00 frames. , ppl: 10.637897894257055] tot_loss[loss=2.297, over 5459870.70 frames. , ppl: 9.944584684753007], batch size: 70
+2022-12-11 22:30:05,464 INFO [train.py:421] (4/8) Epoch 5, batch 69800, loss[loss=2.238, over 2310.00 frames. , ppl: 9.37700215526677] tot_loss[loss=2.299, over 5412752.92 frames. , ppl: 9.962013687638416], batch size: 70
+2022-12-11 22:31:48,577 INFO [train.py:421] (4/8) Epoch 5, batch 70000, loss[loss=3.369, over 490.00 frames. , ppl: 29.056103527605462] tot_loss[loss=2.299, over 5439688.56 frames. , ppl: 9.960064016379953], batch size: 70
+2022-12-11 22:31:48,578 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 22:31:49,328 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.283, over 211138.00 frames. , ppl: 9.80977109406468
+2022-12-11 22:33:34,167 INFO [train.py:421] (4/8) Epoch 5, batch 70200, loss[loss=2.244, over 5950.00 frames. , ppl: 9.431061588447246] tot_loss[loss=2.296, over 5520363.13 frames. , ppl: 9.931181920245946], batch size: 70
+2022-12-11 22:35:14,439 INFO [train.py:421] (4/8) Epoch 5, batch 70400, loss[loss=2.23, over 3220.00 frames. , ppl: 9.301333119541352] tot_loss[loss=2.296, over 5503962.56 frames. , ppl: 9.931184705034916], batch size: 70
+2022-12-11 22:36:54,989 INFO [train.py:421] (4/8) Epoch 5, batch 70600, loss[loss=2.262, over 4270.00 frames. , ppl: 9.605029318662517] tot_loss[loss=2.297, over 5490086.93 frames. , ppl: 9.939456627251616], batch size: 70
+2022-12-11 22:38:32,965 INFO [train.py:421] (4/8) Epoch 5, batch 70800, loss[loss=2.174, over 6860.00 frames. , ppl: 8.793832535226324] tot_loss[loss=2.297, over 5473819.47 frames. , ppl: 9.943034384277107], batch size: 70
+2022-12-11 22:40:16,508 INFO [train.py:421] (4/8) Epoch 5, batch 71000, loss[loss=2.806, over 560.00 frames. , ppl: 16.550633617916727] tot_loss[loss=2.295, over 5541645.92 frames. , ppl: 9.920705637097605], batch size: 70
+2022-12-11 22:40:16,509 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 22:40:17,255 INFO [train.py:452] (4/8) Epoch 5, validation: loss=2.283, over 211138.00 frames. , ppl: 9.803261396653303
+2022-12-11 22:41:56,396 INFO [train.py:421] (4/8) Epoch 5, batch 71200, loss[loss=2.474, over 1190.00 frames. , ppl: 11.865152364987985] tot_loss[loss=2.295, over 5528990.88 frames. , ppl: 9.928358280278598], batch size: 70
+2022-12-11 22:43:36,034 INFO [train.py:421] (4/8) Epoch 5, batch 71400, loss[loss=2.585, over 770.00 frames. , ppl: 13.265801174226272] tot_loss[loss=2.296, over 5484642.75 frames. , ppl: 9.930268319834228], batch size: 70
+2022-12-11 22:45:15,377 INFO [train.py:421] (4/8) Epoch 5, batch 71600, loss[loss=2.228, over 2450.00 frames. , ppl: 9.283201839271609] tot_loss[loss=2.296, over 5458030.23 frames. , ppl: 9.935071375013809], batch size: 70
+2022-12-11 22:46:51,071 INFO [train.py:421] (4/8) Epoch 5, batch 71800, loss[loss=2.319, over 1470.00 frames. , ppl: 10.168022512385095] tot_loss[loss=2.296, over 5446466.13 frames. , ppl: 9.937771685583499], batch size: 70
+2022-12-11 22:48:06,680 INFO [train.py:421] (4/8) Epoch 6, batch 0, loss[loss=2.43, over 980.00 frames. , ppl: 11.36143367795183] tot_loss[loss=2.43, over 980.00 frames. , ppl: 11.36143367795183], batch size: 70
+2022-12-11 22:49:46,706 INFO [train.py:421] (4/8) Epoch 6, batch 200, loss[loss=2.574, over 980.00 frames. , ppl: 13.123631819839957] tot_loss[loss=2.281, over 529275.11 frames. , ppl: 9.782438636880368], batch size: 70
+2022-12-11 22:51:26,931 INFO [train.py:421] (4/8) Epoch 6, batch 400, loss[loss=2.19, over 5740.00 frames. , ppl: 8.931184955957653] tot_loss[loss=2.283, over 1009598.51 frames. , ppl: 9.805359830254947], batch size: 70
+2022-12-11 22:53:08,652 INFO [train.py:421] (4/8) Epoch 6, batch 600, loss[loss=2.222, over 5110.00 frames. , ppl: 9.222886852964692] tot_loss[loss=2.283, over 1444641.40 frames. , ppl: 9.809882839798673], batch size: 70
+2022-12-11 22:54:51,849 INFO [train.py:421] (4/8) Epoch 6, batch 800, loss[loss=2.486, over 700.00 frames. , ppl: 12.014915737158294] tot_loss[loss=2.283, over 1860115.86 frames. , ppl: 9.80519685494817], batch size: 70
+2022-12-11 22:56:35,370 INFO [train.py:421] (4/8) Epoch 6, batch 1000, loss[loss=3.224, over 490.00 frames. , ppl: 25.12521618356619] tot_loss[loss=2.284, over 2226106.89 frames. , ppl: 9.812751613788283], batch size: 70
+2022-12-11 22:56:35,371 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 22:56:36,121 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.816725347697853
+2022-12-11 22:58:21,089 INFO [train.py:421] (4/8) Epoch 6, batch 1200, loss[loss=2.22, over 7840.00 frames. , ppl: 9.20415776825814] tot_loss[loss=2.284, over 2541668.44 frames. , ppl: 9.815218192053976], batch size: 70
+2022-12-11 22:59:58,474 INFO [train.py:421] (4/8) Epoch 6, batch 1400, loss[loss=2.391, over 1540.00 frames. , ppl: 10.920770600221534] tot_loss[loss=2.286, over 2830174.83 frames. , ppl: 9.830807510818783], batch size: 70
+2022-12-11 23:01:38,947 INFO [train.py:421] (4/8) Epoch 6, batch 1600, loss[loss=2.28, over 3290.00 frames. , ppl: 9.777159829647703] tot_loss[loss=2.283, over 3136442.58 frames. , ppl: 9.80162818095843], batch size: 70
+2022-12-11 23:03:19,723 INFO [train.py:421] (4/8) Epoch 6, batch 1800, loss[loss=2.348, over 2380.00 frames. , ppl: 10.461649652598167] tot_loss[loss=2.284, over 3353076.76 frames. , ppl: 9.815460448375722], batch size: 70
+2022-12-11 23:04:57,205 INFO [train.py:421] (4/8) Epoch 6, batch 2000, loss[loss=2.644, over 700.00 frames. , ppl: 14.075175845491765] tot_loss[loss=2.285, over 3542583.72 frames. , ppl: 9.826511100692992], batch size: 70
+2022-12-11 23:04:57,206 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:04:57,952 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.286, over 211138.00 frames. , ppl: 9.840409328583648
+2022-12-11 23:06:35,144 INFO [train.py:421] (4/8) Epoch 6, batch 2200, loss[loss=2.171, over 7210.00 frames. , ppl: 8.762830060458878] tot_loss[loss=2.286, over 3716414.38 frames. , ppl: 9.835141981182778], batch size: 70
+2022-12-11 23:08:19,874 INFO [train.py:421] (4/8) Epoch 6, batch 2400, loss[loss=2.236, over 5110.00 frames. , ppl: 9.352112899759506] tot_loss[loss=2.287, over 3873363.78 frames. , ppl: 9.849140179978216], batch size: 70
+2022-12-11 23:10:02,604 INFO [train.py:421] (4/8) Epoch 6, batch 2600, loss[loss=2.418, over 1400.00 frames. , ppl: 11.226022429446036] tot_loss[loss=2.288, over 4021402.04 frames. , ppl: 9.852060573429949], batch size: 70
+2022-12-11 23:11:44,591 INFO [train.py:421] (4/8) Epoch 6, batch 2800, loss[loss=2.372, over 2730.00 frames. , ppl: 10.723692776795776] tot_loss[loss=2.29, over 4120984.74 frames. , ppl: 9.871304961767928], batch size: 70
+2022-12-11 23:13:24,433 INFO [train.py:421] (4/8) Epoch 6, batch 3000, loss[loss=2.373, over 1260.00 frames. , ppl: 10.724853717582555] tot_loss[loss=2.288, over 4280519.74 frames. , ppl: 9.85767535339425], batch size: 70
+2022-12-11 23:13:24,433 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:13:25,195 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.824575860616017
+2022-12-11 23:15:01,075 INFO [train.py:421] (4/8) Epoch 6, batch 3200, loss[loss=2.199, over 7070.00 frames. , ppl: 9.015046176104992] tot_loss[loss=2.288, over 4401589.34 frames. , ppl: 9.855499888426875], batch size: 70
+2022-12-11 23:16:39,607 INFO [train.py:421] (4/8) Epoch 6, batch 3400, loss[loss=2.277, over 2100.00 frames. , ppl: 9.747544810466644] tot_loss[loss=2.288, over 4496411.75 frames. , ppl: 9.85507760783024], batch size: 70
+2022-12-11 23:18:19,908 INFO [train.py:421] (4/8) Epoch 6, batch 3600, loss[loss=2.35, over 1610.00 frames. , ppl: 10.486422017190415] tot_loss[loss=2.287, over 4595798.71 frames. , ppl: 9.85011539366285], batch size: 70
+2022-12-11 23:19:59,208 INFO [train.py:421] (4/8) Epoch 6, batch 3800, loss[loss=2.428, over 1330.00 frames. , ppl: 11.33856775782665] tot_loss[loss=2.289, over 4649165.62 frames. , ppl: 9.862632674660016], batch size: 70
+2022-12-11 23:21:41,576 INFO [train.py:421] (4/8) Epoch 6, batch 4000, loss[loss=2.218, over 6860.00 frames. , ppl: 9.188804425309248] tot_loss[loss=2.287, over 4776986.58 frames. , ppl: 9.84690338312415], batch size: 70
+2022-12-11 23:21:41,576 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:21:42,322 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.808357026550453
+2022-12-11 23:23:26,051 INFO [train.py:421] (4/8) Epoch 6, batch 4200, loss[loss=2.379, over 2240.00 frames. , ppl: 10.797741412817675] tot_loss[loss=2.286, over 4912541.98 frames. , ppl: 9.837444538658644], batch size: 70
+2022-12-11 23:25:03,905 INFO [train.py:421] (4/8) Epoch 6, batch 4400, loss[loss=2.307, over 2450.00 frames. , ppl: 10.043717808232893] tot_loss[loss=2.287, over 4978598.53 frames. , ppl: 9.8473022363407], batch size: 70
+2022-12-11 23:26:43,142 INFO [train.py:421] (4/8) Epoch 6, batch 4600, loss[loss=2.362, over 3360.00 frames. , ppl: 10.609105549542873] tot_loss[loss=2.288, over 5040984.18 frames. , ppl: 9.85450729562851], batch size: 70
+2022-12-11 23:28:17,781 INFO [train.py:421] (4/8) Epoch 6, batch 4800, loss[loss=2.381, over 1610.00 frames. , ppl: 10.820596249144007] tot_loss[loss=2.29, over 5025530.11 frames. , ppl: 9.87510407689779], batch size: 70
+2022-12-11 23:29:55,651 INFO [train.py:421] (4/8) Epoch 6, batch 5000, loss[loss=2.312, over 2660.00 frames. , ppl: 10.099264029910966] tot_loss[loss=2.291, over 5035570.31 frames. , ppl: 9.883175606942919], batch size: 70
+2022-12-11 23:29:55,651 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:29:56,411 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.82324543954202
+2022-12-11 23:31:38,998 INFO [train.py:421] (4/8) Epoch 6, batch 5200, loss[loss=2.235, over 8470.00 frames. , ppl: 9.345343094968701] tot_loss[loss=2.291, over 5068505.81 frames. , ppl: 9.884521180439393], batch size: 70
+2022-12-11 23:33:16,576 INFO [train.py:421] (4/8) Epoch 6, batch 5400, loss[loss=2.459, over 1400.00 frames. , ppl: 11.6983795805296] tot_loss[loss=2.292, over 5078581.84 frames. , ppl: 9.890837397315964], batch size: 70
+2022-12-11 23:34:57,715 INFO [train.py:421] (4/8) Epoch 6, batch 5600, loss[loss=2.472, over 1540.00 frames. , ppl: 11.84127401347602] tot_loss[loss=2.291, over 5131261.89 frames. , ppl: 9.880604343252601], batch size: 70
+2022-12-11 23:36:40,248 INFO [train.py:421] (4/8) Epoch 6, batch 5800, loss[loss=3.208, over 490.00 frames. , ppl: 24.720592953670142] tot_loss[loss=2.29, over 5181911.51 frames. , ppl: 9.878226670871957], batch size: 70
+2022-12-11 23:38:22,677 INFO [train.py:421] (4/8) Epoch 6, batch 6000, loss[loss=4.922, over 280.00 frames. , ppl: 137.21097272142455] tot_loss[loss=2.291, over 5192876.19 frames. , ppl: 9.88068342860185], batch size: 70
+2022-12-11 23:38:22,677 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:38:23,436 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.813196769801683
+2022-12-11 23:40:02,703 INFO [train.py:421] (4/8) Epoch 6, batch 6200, loss[loss=2.248, over 2940.00 frames. , ppl: 9.470762954166004] tot_loss[loss=2.289, over 5256926.74 frames. , ppl: 9.86600157130723], batch size: 70
+2022-12-11 23:41:42,417 INFO [train.py:421] (4/8) Epoch 6, batch 6400, loss[loss=2.219, over 4760.00 frames. , ppl: 9.198140290013377] tot_loss[loss=2.29, over 5256945.49 frames. , ppl: 9.870658195793238], batch size: 70
+2022-12-11 23:43:22,494 INFO [train.py:421] (4/8) Epoch 6, batch 6600, loss[loss=2.215, over 4550.00 frames. , ppl: 9.163289092230091] tot_loss[loss=2.289, over 5258313.91 frames. , ppl: 9.868511992358425], batch size: 70
+2022-12-11 23:45:02,938 INFO [train.py:421] (4/8) Epoch 6, batch 6800, loss[loss=2.367, over 1610.00 frames. , ppl: 10.661452004874342] tot_loss[loss=2.29, over 5280863.00 frames. , ppl: 9.87425806195512], batch size: 70
+2022-12-11 23:46:38,670 INFO [train.py:421] (4/8) Epoch 6, batch 7000, loss[loss=2.455, over 1190.00 frames. , ppl: 11.6519992770066] tot_loss[loss=2.291, over 5272803.15 frames. , ppl: 9.880436620547131], batch size: 70
+2022-12-11 23:46:38,671 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:46:39,429 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.822367314949352
+2022-12-11 23:48:19,181 INFO [train.py:421] (4/8) Epoch 6, batch 7200, loss[loss=2.565, over 910.00 frames. , ppl: 13.000897641801899] tot_loss[loss=2.291, over 5280602.69 frames. , ppl: 9.886327041216402], batch size: 70
+2022-12-11 23:50:03,958 INFO [train.py:421] (4/8) Epoch 6, batch 7400, loss[loss=2.318, over 2240.00 frames. , ppl: 10.159573739205841] tot_loss[loss=2.292, over 5282880.19 frames. , ppl: 9.893188088224395], batch size: 70
+2022-12-11 23:51:43,015 INFO [train.py:421] (4/8) Epoch 6, batch 7600, loss[loss=2.281, over 2030.00 frames. , ppl: 9.788739374975796] tot_loss[loss=2.291, over 5336661.55 frames. , ppl: 9.885838765610169], batch size: 70
+2022-12-11 23:53:23,567 INFO [train.py:421] (4/8) Epoch 6, batch 7800, loss[loss=2.22, over 7700.00 frames. , ppl: 9.205294567014553] tot_loss[loss=2.291, over 5366978.62 frames. , ppl: 9.881169429751006], batch size: 70
+2022-12-11 23:55:07,598 INFO [train.py:421] (4/8) Epoch 6, batch 8000, loss[loss=2.445, over 910.00 frames. , ppl: 11.532210303666698] tot_loss[loss=2.289, over 5404858.20 frames. , ppl: 9.865589522467326], batch size: 70
+2022-12-11 23:55:07,598 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-11 23:55:08,354 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.793563654653626
+2022-12-11 23:56:47,538 INFO [train.py:421] (4/8) Epoch 6, batch 8200, loss[loss=2.481, over 840.00 frames. , ppl: 11.958640025422563] tot_loss[loss=2.289, over 5431312.88 frames. , ppl: 9.863019256616509], batch size: 70
+2022-12-11 23:58:28,578 INFO [train.py:421] (4/8) Epoch 6, batch 8400, loss[loss=2.279, over 2940.00 frames. , ppl: 9.762811251392847] tot_loss[loss=2.29, over 5403080.71 frames. , ppl: 9.873786245937696], batch size: 70
+2022-12-12 00:00:05,798 INFO [train.py:421] (4/8) Epoch 6, batch 8600, loss[loss=2.465, over 1960.00 frames. , ppl: 11.763126974078352] tot_loss[loss=2.291, over 5365385.79 frames. , ppl: 9.88966449438127], batch size: 70
+2022-12-12 00:01:46,887 INFO [train.py:421] (4/8) Epoch 6, batch 8800, loss[loss=2.292, over 2660.00 frames. , ppl: 9.894444756454073] tot_loss[loss=2.29, over 5438483.75 frames. , ppl: 9.874121749726212], batch size: 70
+2022-12-12 00:03:27,014 INFO [train.py:421] (4/8) Epoch 6, batch 9000, loss[loss=2.521, over 1050.00 frames. , ppl: 12.43626864907392] tot_loss[loss=2.288, over 5453269.38 frames. , ppl: 9.858587022433206], batch size: 70
+2022-12-12 00:03:27,015 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:03:27,760 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.82595299982713
+2022-12-12 00:05:06,496 INFO [train.py:421] (4/8) Epoch 6, batch 9200, loss[loss=2.239, over 5110.00 frames. , ppl: 9.380600518267842] tot_loss[loss=2.289, over 5462969.17 frames. , ppl: 9.862175685274854], batch size: 70
+2022-12-12 00:06:43,069 INFO [train.py:421] (4/8) Epoch 6, batch 9400, loss[loss=3.145, over 560.00 frames. , ppl: 23.22355087153952] tot_loss[loss=2.289, over 5450802.22 frames. , ppl: 9.864520503640646], batch size: 70
+2022-12-12 00:08:21,540 INFO [train.py:421] (4/8) Epoch 6, batch 9600, loss[loss=2.327, over 2520.00 frames. , ppl: 10.244017441146632] tot_loss[loss=2.29, over 5415556.04 frames. , ppl: 9.870706975418253], batch size: 70
+2022-12-12 00:10:01,616 INFO [train.py:421] (4/8) Epoch 6, batch 9800, loss[loss=2.308, over 2240.00 frames. , ppl: 10.058092266590446] tot_loss[loss=2.292, over 5353643.03 frames. , ppl: 9.894053849375993], batch size: 70
+2022-12-12 00:11:46,128 INFO [train.py:421] (4/8) Epoch 6, batch 10000, loss[loss=2.537, over 1120.00 frames. , ppl: 12.646707368052397] tot_loss[loss=2.29, over 5425922.06 frames. , ppl: 9.876569708653498], batch size: 70
+2022-12-12 00:11:46,129 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:11:46,874 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.804028982352971
+2022-12-12 00:13:23,523 INFO [train.py:421] (4/8) Epoch 6, batch 10200, loss[loss=2.771, over 700.00 frames. , ppl: 15.973983872798641] tot_loss[loss=2.291, over 5428854.66 frames. , ppl: 9.880282163533948], batch size: 70
+2022-12-12 00:15:02,150 INFO [train.py:421] (4/8) Epoch 6, batch 10400, loss[loss=2.361, over 2240.00 frames. , ppl: 10.605789797275891] tot_loss[loss=2.291, over 5407760.26 frames. , ppl: 9.88965892727014], batch size: 70
+2022-12-12 00:16:43,003 INFO [train.py:421] (4/8) Epoch 6, batch 10600, loss[loss=2.334, over 1680.00 frames. , ppl: 10.31546899355521] tot_loss[loss=2.291, over 5408608.56 frames. , ppl: 9.882496201318022], batch size: 70
+2022-12-12 00:18:24,444 INFO [train.py:421] (4/8) Epoch 6, batch 10800, loss[loss=2.18, over 11200.00 frames. , ppl: 8.84365634281668] tot_loss[loss=2.291, over 5401029.59 frames. , ppl: 9.884601464852757], batch size: 70
+2022-12-12 00:20:06,164 INFO [train.py:421] (4/8) Epoch 6, batch 11000, loss[loss=2.234, over 3430.00 frames. , ppl: 9.334639144687197] tot_loss[loss=2.29, over 5435718.64 frames. , ppl: 9.874839564828035], batch size: 70
+2022-12-12 00:20:06,165 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:20:06,914 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.804163932742426
+2022-12-12 00:21:45,894 INFO [train.py:421] (4/8) Epoch 6, batch 11200, loss[loss=2.387, over 1260.00 frames. , ppl: 10.87958501110074] tot_loss[loss=2.291, over 5400301.57 frames. , ppl: 9.883405425408132], batch size: 70
+2022-12-12 00:23:23,935 INFO [train.py:421] (4/8) Epoch 6, batch 11400, loss[loss=2.207, over 5880.00 frames. , ppl: 9.092912676028243] tot_loss[loss=2.29, over 5426375.44 frames. , ppl: 9.877921251937293], batch size: 70
+2022-12-12 00:25:02,700 INFO [train.py:421] (4/8) Epoch 6, batch 11600, loss[loss=2.302, over 1680.00 frames. , ppl: 9.991649303500125] tot_loss[loss=2.29, over 5417043.79 frames. , ppl: 9.872725324979148], batch size: 70
+2022-12-12 00:26:41,476 INFO [train.py:421] (4/8) Epoch 6, batch 11800, loss[loss=2.218, over 2800.00 frames. , ppl: 9.190512808251736] tot_loss[loss=2.289, over 5450031.98 frames. , ppl: 9.865018839248153], batch size: 70
+2022-12-12 00:28:20,379 INFO [train.py:421] (4/8) Epoch 6, batch 12000, loss[loss=2.889, over 630.00 frames. , ppl: 17.97926339204424] tot_loss[loss=2.29, over 5445570.33 frames. , ppl: 9.87358141921839], batch size: 70
+2022-12-12 00:28:20,379 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:28:21,131 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.80987272894533
+2022-12-12 00:30:01,072 INFO [train.py:421] (4/8) Epoch 6, batch 12200, loss[loss=2.388, over 1120.00 frames. , ppl: 10.892532210969623] tot_loss[loss=2.291, over 5417901.29 frames. , ppl: 9.883760432155281], batch size: 70
+2022-12-12 00:31:40,362 INFO [train.py:421] (4/8) Epoch 6, batch 12400, loss[loss=2.27, over 2870.00 frames. , ppl: 9.677330698691723] tot_loss[loss=2.29, over 5431114.36 frames. , ppl: 9.879502321150563], batch size: 70
+2022-12-12 00:33:20,726 INFO [train.py:421] (4/8) Epoch 6, batch 12600, loss[loss=2.242, over 4900.00 frames. , ppl: 9.414911127220169] tot_loss[loss=2.291, over 5426518.68 frames. , ppl: 9.885605790509254], batch size: 70
+2022-12-12 00:35:01,335 INFO [train.py:421] (4/8) Epoch 6, batch 12800, loss[loss=2.387, over 1540.00 frames. , ppl: 10.877302993992728] tot_loss[loss=2.293, over 5387558.77 frames. , ppl: 9.899706163113757], batch size: 70
+2022-12-12 00:36:42,459 INFO [train.py:421] (4/8) Epoch 6, batch 13000, loss[loss=2.57, over 1050.00 frames. , ppl: 13.06603710256821] tot_loss[loss=2.292, over 5425676.31 frames. , ppl: 9.892185966854415], batch size: 70
+2022-12-12 00:36:42,460 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:36:43,214 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.805265371495329
+2022-12-12 00:38:26,660 INFO [train.py:421] (4/8) Epoch 6, batch 13200, loss[loss=2.374, over 2660.00 frames. , ppl: 10.742483144549109] tot_loss[loss=2.291, over 5432200.95 frames. , ppl: 9.885392991561377], batch size: 70
+2022-12-12 00:40:05,803 INFO [train.py:421] (4/8) Epoch 6, batch 13400, loss[loss=2.479, over 1400.00 frames. , ppl: 11.92722902607032] tot_loss[loss=2.291, over 5450652.16 frames. , ppl: 9.886872046643113], batch size: 70
+2022-12-12 00:41:44,769 INFO [train.py:421] (4/8) Epoch 6, batch 13600, loss[loss=2.154, over 4340.00 frames. , ppl: 8.622904552857932] tot_loss[loss=2.292, over 5419243.03 frames. , ppl: 9.898012956383957], batch size: 70
+2022-12-12 00:43:25,861 INFO [train.py:421] (4/8) Epoch 6, batch 13800, loss[loss=2.193, over 6160.00 frames. , ppl: 8.958597057862761] tot_loss[loss=2.292, over 5434949.17 frames. , ppl: 9.891576671362206], batch size: 70
+2022-12-12 00:45:05,058 INFO [train.py:421] (4/8) Epoch 6, batch 14000, loss[loss=2.529, over 980.00 frames. , ppl: 12.53705321075541] tot_loss[loss=2.292, over 5424648.61 frames. , ppl: 9.898432353856615], batch size: 70
+2022-12-12 00:45:05,058 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:45:05,804 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.285, over 211138.00 frames. , ppl: 9.822797644344924
+2022-12-12 00:46:45,803 INFO [train.py:421] (4/8) Epoch 6, batch 14200, loss[loss=2.317, over 4410.00 frames. , ppl: 10.14143007645974] tot_loss[loss=2.293, over 5408136.63 frames. , ppl: 9.903248117997483], batch size: 70
+2022-12-12 00:48:26,187 INFO [train.py:421] (4/8) Epoch 6, batch 14400, loss[loss=2.347, over 3780.00 frames. , ppl: 10.453209936313115] tot_loss[loss=2.294, over 5384222.17 frames. , ppl: 9.919205970074666], batch size: 70
+2022-12-12 00:50:08,143 INFO [train.py:421] (4/8) Epoch 6, batch 14600, loss[loss=2.282, over 2730.00 frames. , ppl: 9.79277547693279] tot_loss[loss=2.293, over 5439148.73 frames. , ppl: 9.901995167159335], batch size: 70
+2022-12-12 00:51:47,359 INFO [train.py:421] (4/8) Epoch 6, batch 14800, loss[loss=2.209, over 4060.00 frames. , ppl: 9.105516739090739] tot_loss[loss=2.293, over 5447515.81 frames. , ppl: 9.899963921361325], batch size: 70
+2022-12-12 00:53:30,892 INFO [train.py:421] (4/8) Epoch 6, batch 15000, loss[loss=2.396, over 1820.00 frames. , ppl: 10.975150137500089] tot_loss[loss=2.293, over 5451307.26 frames. , ppl: 9.900364504124088], batch size: 70
+2022-12-12 00:53:30,892 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 00:53:31,639 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.800679035400885
+2022-12-12 00:55:13,350 INFO [train.py:421] (4/8) Epoch 6, batch 15200, loss[loss=2.227, over 5880.00 frames. , ppl: 9.27012017668495] tot_loss[loss=2.291, over 5503503.39 frames. , ppl: 9.885462852682082], batch size: 70
+2022-12-12 00:56:51,707 INFO [train.py:421] (4/8) Epoch 6, batch 15400, loss[loss=2.462, over 1540.00 frames. , ppl: 11.729956991193582] tot_loss[loss=2.29, over 5524421.46 frames. , ppl: 9.879658545293365], batch size: 70
+2022-12-12 00:58:34,883 INFO [train.py:421] (4/8) Epoch 6, batch 15600, loss[loss=2.219, over 3500.00 frames. , ppl: 9.196149623182992] tot_loss[loss=2.291, over 5479214.56 frames. , ppl: 9.888571369180168], batch size: 70
+2022-12-12 01:00:14,970 INFO [train.py:421] (4/8) Epoch 6, batch 15800, loss[loss=2.398, over 2800.00 frames. , ppl: 10.997012442663301] tot_loss[loss=2.291, over 5468795.81 frames. , ppl: 9.889731944781671], batch size: 70
+2022-12-12 01:01:54,259 INFO [train.py:421] (4/8) Epoch 6, batch 16000, loss[loss=2.441, over 1470.00 frames. , ppl: 11.486854536009572] tot_loss[loss=2.292, over 5453386.67 frames. , ppl: 9.89640076411927], batch size: 70
+2022-12-12 01:01:54,260 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:01:55,007 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.805294396587579
+2022-12-12 01:03:38,892 INFO [train.py:421] (4/8) Epoch 6, batch 16200, loss[loss=2.266, over 2940.00 frames. , ppl: 9.643089034908824] tot_loss[loss=2.291, over 5493028.14 frames. , ppl: 9.883977164732825], batch size: 70
+2022-12-12 01:05:19,348 INFO [train.py:421] (4/8) Epoch 6, batch 16400, loss[loss=2.296, over 2520.00 frames. , ppl: 9.934923342734518] tot_loss[loss=2.292, over 5448453.63 frames. , ppl: 9.892441278479229], batch size: 70
+2022-12-12 01:07:02,718 INFO [train.py:421] (4/8) Epoch 6, batch 16600, loss[loss=2.584, over 770.00 frames. , ppl: 13.246246575538267] tot_loss[loss=2.292, over 5450162.62 frames. , ppl: 9.890494708667415], batch size: 70
+2022-12-12 01:08:42,210 INFO [train.py:421] (4/8) Epoch 6, batch 16800, loss[loss=2.458, over 2030.00 frames. , ppl: 11.681752596822749] tot_loss[loss=2.29, over 5478838.41 frames. , ppl: 9.87860505953708], batch size: 70
+2022-12-12 01:10:19,698 INFO [train.py:421] (4/8) Epoch 6, batch 17000, loss[loss=2.438, over 1050.00 frames. , ppl: 11.448487301734861] tot_loss[loss=2.291, over 5425511.35 frames. , ppl: 9.886558727894643], batch size: 70
+2022-12-12 01:10:19,698 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:10:20,459 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.80433371167732
+2022-12-12 01:12:00,281 INFO [train.py:421] (4/8) Epoch 6, batch 17200, loss[loss=2.35, over 1470.00 frames. , ppl: 10.483495845033083] tot_loss[loss=2.291, over 5435245.21 frames. , ppl: 9.887500295207731], batch size: 70
+2022-12-12 01:13:40,302 INFO [train.py:421] (4/8) Epoch 6, batch 17400, loss[loss=2.346, over 1260.00 frames. , ppl: 10.442765917133425] tot_loss[loss=2.291, over 5430439.72 frames. , ppl: 9.8845047634607], batch size: 70
+2022-12-12 01:15:20,967 INFO [train.py:421] (4/8) Epoch 6, batch 17600, loss[loss=2.197, over 8330.00 frames. , ppl: 8.993694905380464] tot_loss[loss=2.29, over 5463174.63 frames. , ppl: 9.877081778145717], batch size: 70
+2022-12-12 01:17:03,064 INFO [train.py:421] (4/8) Epoch 6, batch 17800, loss[loss=2.653, over 910.00 frames. , ppl: 14.195848187502037] tot_loss[loss=2.29, over 5487820.16 frames. , ppl: 9.870993172961265], batch size: 70
+2022-12-12 01:18:44,998 INFO [train.py:421] (4/8) Epoch 6, batch 18000, loss[loss=2.156, over 7350.00 frames. , ppl: 8.639346377562177] tot_loss[loss=2.289, over 5494478.11 frames. , ppl: 9.865256776276944], batch size: 70
+2022-12-12 01:18:44,999 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:18:45,748 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.802139871930427
+2022-12-12 01:20:22,555 INFO [train.py:421] (4/8) Epoch 6, batch 18200, loss[loss=2.374, over 1960.00 frames. , ppl: 10.743489724820842] tot_loss[loss=2.291, over 5433078.24 frames. , ppl: 9.883720433603635], batch size: 70
+2022-12-12 01:22:05,359 INFO [train.py:421] (4/8) Epoch 6, batch 18400, loss[loss=2.561, over 1050.00 frames. , ppl: 12.954524119395456] tot_loss[loss=2.291, over 5445775.09 frames. , ppl: 9.88361085068663], batch size: 70
+2022-12-12 01:23:45,394 INFO [train.py:421] (4/8) Epoch 6, batch 18600, loss[loss=2.367, over 2030.00 frames. , ppl: 10.664864688516808] tot_loss[loss=2.29, over 5459947.79 frames. , ppl: 9.876112101093042], batch size: 70
+2022-12-12 01:25:24,440 INFO [train.py:421] (4/8) Epoch 6, batch 18800, loss[loss=2.495, over 1050.00 frames. , ppl: 12.125028773535613] tot_loss[loss=2.29, over 5467956.55 frames. , ppl: 9.87462077042273], batch size: 70
+2022-12-12 01:27:04,218 INFO [train.py:421] (4/8) Epoch 6, batch 19000, loss[loss=2.25, over 11970.00 frames. , ppl: 9.49153562975115] tot_loss[loss=2.289, over 5474092.03 frames. , ppl: 9.866876929055675], batch size: 70
+2022-12-12 01:27:04,218 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:27:04,968 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.798580280075496
+2022-12-12 01:28:43,251 INFO [train.py:421] (4/8) Epoch 6, batch 19200, loss[loss=2.309, over 2030.00 frames. , ppl: 10.068130106855405] tot_loss[loss=2.29, over 5470821.70 frames. , ppl: 9.873292085943921], batch size: 70
+2022-12-12 01:30:21,610 INFO [train.py:421] (4/8) Epoch 6, batch 19400, loss[loss=3.688, over 420.00 frames. , ppl: 39.97115292689049] tot_loss[loss=2.289, over 5482211.13 frames. , ppl: 9.869154236632061], batch size: 70
+2022-12-12 01:32:02,249 INFO [train.py:421] (4/8) Epoch 6, batch 19600, loss[loss=2.515, over 770.00 frames. , ppl: 12.37018213603391] tot_loss[loss=2.288, over 5495407.61 frames. , ppl: 9.857041500849636], batch size: 70
+2022-12-12 01:33:40,022 INFO [train.py:421] (4/8) Epoch 6, batch 19800, loss[loss=2.238, over 3850.00 frames. , ppl: 9.372354131093212] tot_loss[loss=2.288, over 5475792.60 frames. , ppl: 9.859134643238939], batch size: 70
+2022-12-12 01:35:20,120 INFO [train.py:421] (4/8) Epoch 6, batch 20000, loss[loss=2.45, over 1190.00 frames. , ppl: 11.58773810355066] tot_loss[loss=2.288, over 5481395.39 frames. , ppl: 9.85823360010494], batch size: 70
+2022-12-12 01:35:20,120 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:35:20,866 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.811063387394697
+2022-12-12 01:36:57,458 INFO [train.py:421] (4/8) Epoch 6, batch 20200, loss[loss=2.601, over 840.00 frames. , ppl: 13.474425171943752] tot_loss[loss=2.289, over 5471085.23 frames. , ppl: 9.8617691539977], batch size: 70
+2022-12-12 01:38:35,707 INFO [train.py:421] (4/8) Epoch 6, batch 20400, loss[loss=2.293, over 2240.00 frames. , ppl: 9.903580711161343] tot_loss[loss=2.288, over 5521326.37 frames. , ppl: 9.850750964491466], batch size: 70
+2022-12-12 01:40:16,777 INFO [train.py:421] (4/8) Epoch 6, batch 20600, loss[loss=2.281, over 2730.00 frames. , ppl: 9.788030028690482] tot_loss[loss=2.287, over 5525093.73 frames. , ppl: 9.848386572444392], batch size: 70
+2022-12-12 01:41:54,631 INFO [train.py:421] (4/8) Epoch 6, batch 20800, loss[loss=2.275, over 2940.00 frames. , ppl: 9.724623679904747] tot_loss[loss=2.289, over 5492618.35 frames. , ppl: 9.863264123499299], batch size: 70
+2022-12-12 01:43:37,329 INFO [train.py:421] (4/8) Epoch 6, batch 21000, loss[loss=2.164, over 3080.00 frames. , ppl: 8.706358512007881] tot_loss[loss=2.288, over 5520421.10 frames. , ppl: 9.852563376592057], batch size: 70
+2022-12-12 01:43:37,330 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:43:38,089 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.790676635383708
+2022-12-12 01:45:21,893 INFO [train.py:421] (4/8) Epoch 6, batch 21200, loss[loss=2.208, over 5600.00 frames. , ppl: 9.097647232940846] tot_loss[loss=2.288, over 5507363.60 frames. , ppl: 9.859521474832686], batch size: 70
+2022-12-12 01:47:02,204 INFO [train.py:421] (4/8) Epoch 6, batch 21400, loss[loss=2.737, over 700.00 frames. , ppl: 15.436784705677763] tot_loss[loss=2.288, over 5516541.51 frames. , ppl: 9.85777507125163], batch size: 70
+2022-12-12 01:48:39,309 INFO [train.py:421] (4/8) Epoch 6, batch 21600, loss[loss=2.311, over 1820.00 frames. , ppl: 10.079708453311103] tot_loss[loss=2.29, over 5466436.63 frames. , ppl: 9.8724698546117], batch size: 70
+2022-12-12 01:50:16,059 INFO [train.py:421] (4/8) Epoch 6, batch 21800, loss[loss=2.322, over 1610.00 frames. , ppl: 10.197656719353764] tot_loss[loss=2.291, over 5406738.30 frames. , ppl: 9.889142701118066], batch size: 70
+2022-12-12 01:51:58,032 INFO [train.py:421] (4/8) Epoch 6, batch 22000, loss[loss=2.183, over 6090.00 frames. , ppl: 8.870114844343338] tot_loss[loss=2.29, over 5429573.98 frames. , ppl: 9.87458202285517], batch size: 70
+2022-12-12 01:51:58,032 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 01:51:58,792 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.776022789235363
+2022-12-12 01:53:38,893 INFO [train.py:421] (4/8) Epoch 6, batch 22200, loss[loss=2.196, over 6230.00 frames. , ppl: 8.986514549287982] tot_loss[loss=2.29, over 5446654.17 frames. , ppl: 9.878818859183859], batch size: 70
+2022-12-12 01:55:20,543 INFO [train.py:421] (4/8) Epoch 6, batch 22400, loss[loss=2.344, over 1750.00 frames. , ppl: 10.42782171903759] tot_loss[loss=2.29, over 5466302.84 frames. , ppl: 9.872769600324633], batch size: 70
+2022-12-12 01:57:04,222 INFO [train.py:421] (4/8) Epoch 6, batch 22600, loss[loss=2.366, over 3430.00 frames. , ppl: 10.6534517290036] tot_loss[loss=2.29, over 5472363.75 frames. , ppl: 9.87808164846209], batch size: 70
+2022-12-12 01:58:41,918 INFO [train.py:421] (4/8) Epoch 6, batch 22800, loss[loss=2.916, over 560.00 frames. , ppl: 18.475841529207766] tot_loss[loss=2.292, over 5426588.30 frames. , ppl: 9.892710420778632], batch size: 70
+2022-12-12 02:00:22,243 INFO [train.py:421] (4/8) Epoch 6, batch 23000, loss[loss=2.372, over 1610.00 frames. , ppl: 10.720930942225687] tot_loss[loss=2.291, over 5467718.55 frames. , ppl: 9.88320705902241], batch size: 70
+2022-12-12 02:00:22,244 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:00:23,004 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.793408557185776
+2022-12-12 02:01:58,413 INFO [train.py:421] (4/8) Epoch 6, batch 23200, loss[loss=2.313, over 2170.00 frames. , ppl: 10.10058992952339] tot_loss[loss=2.292, over 5422037.02 frames. , ppl: 9.892670506316009], batch size: 70
+2022-12-12 02:03:43,836 INFO [train.py:421] (4/8) Epoch 6, batch 23400, loss[loss=2.246, over 4270.00 frames. , ppl: 9.450370580928949] tot_loss[loss=2.291, over 5446897.40 frames. , ppl: 9.883634891440108], batch size: 70
+2022-12-12 02:05:24,908 INFO [train.py:421] (4/8) Epoch 6, batch 23600, loss[loss=2.313, over 2450.00 frames. , ppl: 10.106411672094177] tot_loss[loss=2.29, over 5454764.87 frames. , ppl: 9.878360325880648], batch size: 70
+2022-12-12 02:07:03,693 INFO [train.py:421] (4/8) Epoch 6, batch 23800, loss[loss=2.352, over 2800.00 frames. , ppl: 10.505387841792718] tot_loss[loss=2.291, over 5450936.35 frames. , ppl: 9.886631775737923], batch size: 70
+2022-12-12 02:08:44,621 INFO [train.py:421] (4/8) Epoch 6, batch 24000, loss[loss=2.334, over 2100.00 frames. , ppl: 10.31792594978089] tot_loss[loss=2.29, over 5498938.98 frames. , ppl: 9.876010881794825], batch size: 70
+2022-12-12 02:08:44,621 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:08:45,409 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.790136138531322
+2022-12-12 02:10:29,007 INFO [train.py:421] (4/8) Epoch 6, batch 24200, loss[loss=3.289, over 490.00 frames. , ppl: 26.818952941908112] tot_loss[loss=2.288, over 5553524.73 frames. , ppl: 9.852956023179006], batch size: 70
+2022-12-12 02:12:08,544 INFO [train.py:421] (4/8) Epoch 6, batch 24400, loss[loss=2.175, over 3780.00 frames. , ppl: 8.804304785779257] tot_loss[loss=2.287, over 5549789.97 frames. , ppl: 9.849104051124117], batch size: 70
+2022-12-12 02:13:52,338 INFO [train.py:421] (4/8) Epoch 6, batch 24600, loss[loss=2.429, over 1330.00 frames. , ppl: 11.34955640086586] tot_loss[loss=2.288, over 5529642.58 frames. , ppl: 9.857289467498449], batch size: 70
+2022-12-12 02:15:29,117 INFO [train.py:421] (4/8) Epoch 6, batch 24800, loss[loss=2.655, over 770.00 frames. , ppl: 14.225673438925861] tot_loss[loss=2.288, over 5525680.61 frames. , ppl: 9.858484512285415], batch size: 70
+2022-12-12 02:17:09,490 INFO [train.py:421] (4/8) Epoch 6, batch 25000, loss[loss=2.651, over 1050.00 frames. , ppl: 14.171623631555876] tot_loss[loss=2.288, over 5547302.62 frames. , ppl: 9.853966486929101], batch size: 70
+2022-12-12 02:17:09,491 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:17:10,254 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.788068615004098
+2022-12-12 02:18:49,731 INFO [train.py:421] (4/8) Epoch 6, batch 25200, loss[loss=2.133, over 5530.00 frames. , ppl: 8.43771160744294] tot_loss[loss=2.287, over 5575447.01 frames. , ppl: 9.846601693223148], batch size: 70
+2022-12-12 02:20:33,642 INFO [train.py:421] (4/8) Epoch 6, batch 25400, loss[loss=2.584, over 770.00 frames. , ppl: 13.24640407388942] tot_loss[loss=2.288, over 5565091.56 frames. , ppl: 9.850415450028311], batch size: 70
+2022-12-12 02:22:12,266 INFO [train.py:421] (4/8) Epoch 6, batch 25600, loss[loss=2.534, over 1120.00 frames. , ppl: 12.607866446365412] tot_loss[loss=2.288, over 5533232.70 frames. , ppl: 9.858991936097924], batch size: 70
+2022-12-12 02:23:50,883 INFO [train.py:421] (4/8) Epoch 6, batch 25800, loss[loss=2.358, over 2240.00 frames. , ppl: 10.568131576203495] tot_loss[loss=2.287, over 5541919.83 frames. , ppl: 9.84860656480111], batch size: 70
+2022-12-12 02:25:30,634 INFO [train.py:421] (4/8) Epoch 6, batch 26000, loss[loss=2.283, over 1470.00 frames. , ppl: 9.803608688272764] tot_loss[loss=2.288, over 5532739.32 frames. , ppl: 9.853725498227265], batch size: 70
+2022-12-12 02:25:30,634 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:25:31,395 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.806832849432856
+2022-12-12 02:27:11,365 INFO [train.py:421] (4/8) Epoch 6, batch 26200, loss[loss=2.571, over 1050.00 frames. , ppl: 13.074625447725987] tot_loss[loss=2.288, over 5537960.00 frames. , ppl: 9.857202831133245], batch size: 70
+2022-12-12 02:28:48,580 INFO [train.py:421] (4/8) Epoch 6, batch 26400, loss[loss=2.275, over 2450.00 frames. , ppl: 9.723821338523102] tot_loss[loss=2.289, over 5509873.17 frames. , ppl: 9.868547666014173], batch size: 70
+2022-12-12 02:30:30,700 INFO [train.py:421] (4/8) Epoch 6, batch 26600, loss[loss=2.271, over 2590.00 frames. , ppl: 9.693011813314946] tot_loss[loss=2.29, over 5518470.14 frames. , ppl: 9.871886424947917], batch size: 70
+2022-12-12 02:32:10,389 INFO [train.py:421] (4/8) Epoch 6, batch 26800, loss[loss=2.219, over 2730.00 frames. , ppl: 9.194388209746599] tot_loss[loss=2.291, over 5458997.07 frames. , ppl: 9.888043934181798], batch size: 70
+2022-12-12 02:33:50,899 INFO [train.py:421] (4/8) Epoch 6, batch 27000, loss[loss=2.991, over 560.00 frames. , ppl: 19.90058742501022] tot_loss[loss=2.292, over 5419921.91 frames. , ppl: 9.89674699620861], batch size: 70
+2022-12-12 02:33:50,899 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:33:51,644 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.810623407744394
+2022-12-12 02:35:28,235 INFO [train.py:421] (4/8) Epoch 6, batch 27200, loss[loss=2.193, over 5040.00 frames. , ppl: 8.958230329871675] tot_loss[loss=2.293, over 5390404.14 frames. , ppl: 9.908523856402875], batch size: 70
+2022-12-12 02:37:07,150 INFO [train.py:421] (4/8) Epoch 6, batch 27400, loss[loss=2.151, over 3290.00 frames. , ppl: 8.589608322661782] tot_loss[loss=2.293, over 5378706.91 frames. , ppl: 9.904508772857548], batch size: 70
+2022-12-12 02:38:49,954 INFO [train.py:421] (4/8) Epoch 6, batch 27600, loss[loss=2.551, over 1400.00 frames. , ppl: 12.817819561575837] tot_loss[loss=2.293, over 5364852.30 frames. , ppl: 9.909253655316991], batch size: 70
+2022-12-12 02:40:23,826 INFO [train.py:421] (4/8) Epoch 6, batch 27800, loss[loss=2.39, over 2240.00 frames. , ppl: 10.913677599750073] tot_loss[loss=2.293, over 5374280.05 frames. , ppl: 9.908387228033193], batch size: 70
+2022-12-12 02:42:02,135 INFO [train.py:421] (4/8) Epoch 6, batch 28000, loss[loss=2.508, over 1190.00 frames. , ppl: 12.278458081398204] tot_loss[loss=2.293, over 5391940.36 frames. , ppl: 9.903559748185456], batch size: 70
+2022-12-12 02:42:02,135 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:42:02,898 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.284, over 211138.00 frames. , ppl: 9.819234910695574
+2022-12-12 02:43:40,808 INFO [train.py:421] (4/8) Epoch 6, batch 28200, loss[loss=2.52, over 1120.00 frames. , ppl: 12.422680054185474] tot_loss[loss=2.292, over 5417728.15 frames. , ppl: 9.891810062670832], batch size: 70
+2022-12-12 02:45:19,070 INFO [train.py:421] (4/8) Epoch 6, batch 28400, loss[loss=2.52, over 1050.00 frames. , ppl: 12.427513023193587] tot_loss[loss=2.29, over 5492824.77 frames. , ppl: 9.87575066221518], batch size: 70
+2022-12-12 02:46:59,636 INFO [train.py:421] (4/8) Epoch 6, batch 28600, loss[loss=2.27, over 3430.00 frames. , ppl: 9.68210219221796] tot_loss[loss=2.288, over 5560687.01 frames. , ppl: 9.859692894704525], batch size: 70
+2022-12-12 02:48:38,764 INFO [train.py:421] (4/8) Epoch 6, batch 28800, loss[loss=2.241, over 4410.00 frames. , ppl: 9.401098206183113] tot_loss[loss=2.288, over 5581248.15 frames. , ppl: 9.855741721941591], batch size: 70
+2022-12-12 02:50:20,987 INFO [train.py:421] (4/8) Epoch 6, batch 29000, loss[loss=2.858, over 630.00 frames. , ppl: 17.42733546217443] tot_loss[loss=2.287, over 5581744.07 frames. , ppl: 9.844837210650372], batch size: 70
+2022-12-12 02:50:20,988 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:50:21,747 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.796046998105865
+2022-12-12 02:51:59,584 INFO [train.py:421] (4/8) Epoch 6, batch 29200, loss[loss=2.301, over 2240.00 frames. , ppl: 9.986657107693668] tot_loss[loss=2.287, over 5579109.06 frames. , ppl: 9.844083287659096], batch size: 70
+2022-12-12 02:53:38,991 INFO [train.py:421] (4/8) Epoch 6, batch 29400, loss[loss=2.308, over 2520.00 frames. , ppl: 10.056591740849402] tot_loss[loss=2.287, over 5584634.16 frames. , ppl: 9.845227809506145], batch size: 70
+2022-12-12 02:55:17,412 INFO [train.py:421] (4/8) Epoch 6, batch 29600, loss[loss=2.499, over 980.00 frames. , ppl: 12.164773703528358] tot_loss[loss=2.287, over 5594103.75 frames. , ppl: 9.843067695389031], batch size: 70
+2022-12-12 02:57:01,361 INFO [train.py:421] (4/8) Epoch 6, batch 29800, loss[loss=2.159, over 7070.00 frames. , ppl: 8.661635575940684] tot_loss[loss=2.287, over 5615846.61 frames. , ppl: 9.843912982169995], batch size: 70
+2022-12-12 02:58:45,206 INFO [train.py:421] (4/8) Epoch 6, batch 30000, loss[loss=2.191, over 5040.00 frames. , ppl: 8.944223227953575] tot_loss[loss=2.288, over 5591771.20 frames. , ppl: 9.852401665862867], batch size: 70
+2022-12-12 02:58:45,207 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 02:58:45,956 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.793146201640242
+2022-12-12 03:00:24,701 INFO [train.py:421] (4/8) Epoch 6, batch 30200, loss[loss=2.431, over 1190.00 frames. , ppl: 11.37144936515153] tot_loss[loss=2.29, over 5525038.72 frames. , ppl: 9.872152379824712], batch size: 70
+2022-12-12 03:02:01,508 INFO [train.py:421] (4/8) Epoch 6, batch 30400, loss[loss=2.297, over 3430.00 frames. , ppl: 9.942320115259015] tot_loss[loss=2.289, over 5524477.74 frames. , ppl: 9.864290361439686], batch size: 70
+2022-12-12 03:03:40,077 INFO [train.py:421] (4/8) Epoch 6, batch 30600, loss[loss=4.197, over 350.00 frames. , ppl: 66.51365244812784] tot_loss[loss=2.29, over 5511204.85 frames. , ppl: 9.87125720794326], batch size: 70
+2022-12-12 03:05:19,243 INFO [train.py:421] (4/8) Epoch 6, batch 30800, loss[loss=2.713, over 840.00 frames. , ppl: 15.067570105315829] tot_loss[loss=2.29, over 5490848.37 frames. , ppl: 9.874742825607512], batch size: 70
+2022-12-12 03:07:01,550 INFO [train.py:421] (4/8) Epoch 6, batch 31000, loss[loss=2.225, over 4550.00 frames. , ppl: 9.253433152368121] tot_loss[loss=2.289, over 5517615.62 frames. , ppl: 9.866595985154898], batch size: 70
+2022-12-12 03:07:01,550 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:07:02,295 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.792647600527868
+2022-12-12 03:08:42,301 INFO [train.py:421] (4/8) Epoch 6, batch 31200, loss[loss=2.313, over 1750.00 frames. , ppl: 10.107263509947876] tot_loss[loss=2.289, over 5533121.98 frames. , ppl: 9.863204340760536], batch size: 70
+2022-12-12 03:10:23,241 INFO [train.py:421] (4/8) Epoch 6, batch 31400, loss[loss=2.265, over 6510.00 frames. , ppl: 9.627685816758573] tot_loss[loss=2.287, over 5585192.89 frames. , ppl: 9.850018895735637], batch size: 70
+2022-12-12 03:12:05,139 INFO [train.py:421] (4/8) Epoch 6, batch 31600, loss[loss=2.274, over 4340.00 frames. , ppl: 9.718463090987825] tot_loss[loss=2.288, over 5529476.46 frames. , ppl: 9.859001478758781], batch size: 70
+2022-12-12 03:13:47,714 INFO [train.py:421] (4/8) Epoch 6, batch 31800, loss[loss=2.26, over 3780.00 frames. , ppl: 9.579836036072978] tot_loss[loss=2.287, over 5589953.59 frames. , ppl: 9.849896696507038], batch size: 70
+2022-12-12 03:15:28,300 INFO [train.py:421] (4/8) Epoch 6, batch 32000, loss[loss=2.139, over 4900.00 frames. , ppl: 8.486835709462257] tot_loss[loss=2.287, over 5592456.96 frames. , ppl: 9.847957715762982], batch size: 70
+2022-12-12 03:15:28,301 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:15:29,060 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.786621463485243
+2022-12-12 03:17:08,909 INFO [train.py:421] (4/8) Epoch 6, batch 32200, loss[loss=2.475, over 1820.00 frames. , ppl: 11.883109784081357] tot_loss[loss=2.287, over 5601355.76 frames. , ppl: 9.844724478712639], batch size: 70
+2022-12-12 03:18:47,922 INFO [train.py:421] (4/8) Epoch 6, batch 32400, loss[loss=2.431, over 1330.00 frames. , ppl: 11.36904224183438] tot_loss[loss=2.287, over 5585798.23 frames. , ppl: 9.849272761115353], batch size: 70
+2022-12-12 03:20:28,895 INFO [train.py:421] (4/8) Epoch 6, batch 32600, loss[loss=2.547, over 1260.00 frames. , ppl: 12.763414118833042] tot_loss[loss=2.287, over 5600326.88 frames. , ppl: 9.841569338631887], batch size: 70
+2022-12-12 03:22:12,184 INFO [train.py:421] (4/8) Epoch 6, batch 32800, loss[loss=2.295, over 4900.00 frames. , ppl: 9.922632308696095] tot_loss[loss=2.287, over 5591985.05 frames. , ppl: 9.84544416287834], batch size: 70
+2022-12-12 03:23:51,259 INFO [train.py:421] (4/8) Epoch 6, batch 33000, loss[loss=2.297, over 4760.00 frames. , ppl: 9.947364909090652] tot_loss[loss=2.287, over 5571754.64 frames. , ppl: 9.8490597719134], batch size: 70
+2022-12-12 03:23:51,259 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:23:52,003 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.283, over 211138.00 frames. , ppl: 9.805240700234473
+2022-12-12 03:25:32,458 INFO [train.py:421] (4/8) Epoch 6, batch 33200, loss[loss=2.555, over 840.00 frames. , ppl: 12.87285898446551] tot_loss[loss=2.288, over 5569904.13 frames. , ppl: 9.851198446757959], batch size: 70
+2022-12-12 03:27:11,786 INFO [train.py:421] (4/8) Epoch 6, batch 33400, loss[loss=2.291, over 2380.00 frames. , ppl: 9.887434311383588] tot_loss[loss=2.286, over 5593443.92 frames. , ppl: 9.839103520608703], batch size: 70
+2022-12-12 03:28:50,124 INFO [train.py:421] (4/8) Epoch 6, batch 33600, loss[loss=2.155, over 5530.00 frames. , ppl: 8.631432431535767] tot_loss[loss=2.287, over 5560294.40 frames. , ppl: 9.844829880765323], batch size: 70
+2022-12-12 03:30:30,181 INFO [train.py:421] (4/8) Epoch 6, batch 33800, loss[loss=2.641, over 700.00 frames. , ppl: 14.031235107331472] tot_loss[loss=2.286, over 5565179.15 frames. , ppl: 9.837674734986168], batch size: 70
+2022-12-12 03:32:08,721 INFO [train.py:421] (4/8) Epoch 6, batch 34000, loss[loss=2.267, over 2450.00 frames. , ppl: 9.653431583367595] tot_loss[loss=2.287, over 5554167.40 frames. , ppl: 9.845732544548564], batch size: 70
+2022-12-12 03:32:08,721 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:32:09,480 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.78966811844687
+2022-12-12 03:33:48,818 INFO [train.py:421] (4/8) Epoch 6, batch 34200, loss[loss=2.342, over 2380.00 frames. , ppl: 10.399488930096128] tot_loss[loss=2.287, over 5554689.15 frames. , ppl: 9.85015482571971], batch size: 70
+2022-12-12 03:35:27,527 INFO [train.py:421] (4/8) Epoch 6, batch 34400, loss[loss=2.345, over 2100.00 frames. , ppl: 10.43398838998698] tot_loss[loss=2.289, over 5512125.50 frames. , ppl: 9.862533439215856], batch size: 70
+2022-12-12 03:37:07,539 INFO [train.py:421] (4/8) Epoch 6, batch 34600, loss[loss=2.304, over 3570.00 frames. , ppl: 10.011823464985413] tot_loss[loss=2.289, over 5508024.63 frames. , ppl: 9.863382040286067], batch size: 70
+2022-12-12 03:38:50,946 INFO [train.py:421] (4/8) Epoch 6, batch 34800, loss[loss=2.675, over 700.00 frames. , ppl: 14.513901274084128] tot_loss[loss=2.288, over 5562570.58 frames. , ppl: 9.855419405943286], batch size: 70
+2022-12-12 03:40:30,319 INFO [train.py:421] (4/8) Epoch 6, batch 35000, loss[loss=2.321, over 2310.00 frames. , ppl: 10.1825797632086] tot_loss[loss=2.289, over 5541327.43 frames. , ppl: 9.863517830345327], batch size: 70
+2022-12-12 03:40:30,319 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:40:31,078 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.778649309689856
+2022-12-12 03:42:08,421 INFO [train.py:421] (4/8) Epoch 6, batch 35200, loss[loss=2.891, over 630.00 frames. , ppl: 18.01621116135012] tot_loss[loss=2.288, over 5545086.30 frames. , ppl: 9.858077984064106], batch size: 70
+2022-12-12 03:43:48,388 INFO [train.py:421] (4/8) Epoch 6, batch 35400, loss[loss=2.267, over 3220.00 frames. , ppl: 9.647269831706685] tot_loss[loss=2.288, over 5564378.01 frames. , ppl: 9.852857873909343], batch size: 70
+2022-12-12 03:45:30,989 INFO [train.py:421] (4/8) Epoch 6, batch 35600, loss[loss=2.36, over 1890.00 frames. , ppl: 10.58753944980575] tot_loss[loss=2.289, over 5533348.91 frames. , ppl: 9.860781801170212], batch size: 70
+2022-12-12 03:47:10,358 INFO [train.py:421] (4/8) Epoch 6, batch 35800, loss[loss=3.231, over 490.00 frames. , ppl: 25.30078591749663] tot_loss[loss=2.288, over 5574476.60 frames. , ppl: 9.854637177917391], batch size: 70
+2022-12-12 03:48:50,030 INFO [train.py:421] (4/8) Epoch 6, batch 36000, loss[loss=2.419, over 2940.00 frames. , ppl: 11.239360726774152] tot_loss[loss=2.289, over 5565990.65 frames. , ppl: 9.862128601664665], batch size: 70
+2022-12-12 03:48:50,030 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:48:50,790 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.788203345702467
+2022-12-12 03:50:31,023 INFO [train.py:421] (4/8) Epoch 6, batch 36200, loss[loss=2.392, over 1890.00 frames. , ppl: 10.932353500354681] tot_loss[loss=2.288, over 5574147.66 frames. , ppl: 9.857071031064839], batch size: 70
+2022-12-12 03:52:13,148 INFO [train.py:421] (4/8) Epoch 6, batch 36400, loss[loss=2.619, over 700.00 frames. , ppl: 13.726255753918233] tot_loss[loss=2.289, over 5539586.79 frames. , ppl: 9.862426412729798], batch size: 70
+2022-12-12 03:53:51,589 INFO [train.py:421] (4/8) Epoch 6, batch 36600, loss[loss=2.261, over 3290.00 frames. , ppl: 9.593508346411484] tot_loss[loss=2.289, over 5539991.26 frames. , ppl: 9.869594467754041], batch size: 70
+2022-12-12 03:55:34,328 INFO [train.py:421] (4/8) Epoch 6, batch 36800, loss[loss=2.664, over 770.00 frames. , ppl: 14.359024284496543] tot_loss[loss=2.289, over 5517839.76 frames. , ppl: 9.866918704243055], batch size: 70
+2022-12-12 03:57:15,031 INFO [train.py:421] (4/8) Epoch 6, batch 37000, loss[loss=2.608, over 910.00 frames. , ppl: 13.573588174276141] tot_loss[loss=2.29, over 5477732.93 frames. , ppl: 9.877142162378693], batch size: 70
+2022-12-12 03:57:15,031 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 03:57:15,781 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.776730360844793
+2022-12-12 03:58:58,749 INFO [train.py:421] (4/8) Epoch 6, batch 37200, loss[loss=2.324, over 2870.00 frames. , ppl: 10.212037190585082] tot_loss[loss=2.29, over 5489082.88 frames. , ppl: 9.87709613371691], batch size: 70
+2022-12-12 04:00:39,203 INFO [train.py:421] (4/8) Epoch 6, batch 37400, loss[loss=2.507, over 980.00 frames. , ppl: 12.262000646211344] tot_loss[loss=2.289, over 5529805.69 frames. , ppl: 9.864370475756289], batch size: 70
+2022-12-12 04:02:19,599 INFO [train.py:421] (4/8) Epoch 6, batch 37600, loss[loss=2.327, over 2590.00 frames. , ppl: 10.24262155764664] tot_loss[loss=2.288, over 5564019.49 frames. , ppl: 9.850587771675508], batch size: 70
+2022-12-12 04:04:01,661 INFO [train.py:421] (4/8) Epoch 6, batch 37800, loss[loss=2.531, over 1050.00 frames. , ppl: 12.563784789375202] tot_loss[loss=2.286, over 5596660.44 frames. , ppl: 9.837180245254315], batch size: 70
+2022-12-12 04:05:40,692 INFO [train.py:421] (4/8) Epoch 6, batch 38000, loss[loss=2.31, over 1960.00 frames. , ppl: 10.077334402569344] tot_loss[loss=2.287, over 5576486.44 frames. , ppl: 9.845811296565117], batch size: 70
+2022-12-12 04:05:40,693 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:05:41,464 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.785526464019718
+2022-12-12 04:07:23,631 INFO [train.py:421] (4/8) Epoch 6, batch 38200, loss[loss=2.617, over 1120.00 frames. , ppl: 13.691507233925314] tot_loss[loss=2.287, over 5591989.55 frames. , ppl: 9.844402154719898], batch size: 70
+2022-12-12 04:08:58,703 INFO [train.py:421] (4/8) Epoch 6, batch 38400, loss[loss=2.328, over 1960.00 frames. , ppl: 10.255448362520413] tot_loss[loss=2.287, over 5609407.22 frames. , ppl: 9.842391664037628], batch size: 70
+2022-12-12 04:10:38,745 INFO [train.py:421] (4/8) Epoch 6, batch 38600, loss[loss=2.175, over 4410.00 frames. , ppl: 8.799154669037705] tot_loss[loss=2.286, over 5655185.49 frames. , ppl: 9.83693770449724], batch size: 70
+2022-12-12 04:12:18,447 INFO [train.py:421] (4/8) Epoch 6, batch 38800, loss[loss=2.306, over 3010.00 frames. , ppl: 10.033235922461111] tot_loss[loss=2.286, over 5648444.40 frames. , ppl: 9.832869927218502], batch size: 70
+2022-12-12 04:13:53,332 INFO [train.py:421] (4/8) Epoch 6, batch 39000, loss[loss=2.455, over 1680.00 frames. , ppl: 11.648111933661115] tot_loss[loss=2.287, over 5620623.54 frames. , ppl: 9.840747777239404], batch size: 70
+2022-12-12 04:13:53,332 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:13:54,080 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.790575199383582
+2022-12-12 04:15:37,675 INFO [train.py:421] (4/8) Epoch 6, batch 39200, loss[loss=2.352, over 2240.00 frames. , ppl: 10.506848042683218] tot_loss[loss=2.287, over 5586515.62 frames. , ppl: 9.849640552173451], batch size: 70
+2022-12-12 04:17:15,042 INFO [train.py:421] (4/8) Epoch 6, batch 39400, loss[loss=2.55, over 770.00 frames. , ppl: 12.805840969114916] tot_loss[loss=2.289, over 5551184.51 frames. , ppl: 9.860913935762063], batch size: 70
+2022-12-12 04:18:56,787 INFO [train.py:421] (4/8) Epoch 6, batch 39600, loss[loss=2.555, over 1050.00 frames. , ppl: 12.868714165070983] tot_loss[loss=2.289, over 5527979.29 frames. , ppl: 9.865101706163353], batch size: 70
+2022-12-12 04:20:36,867 INFO [train.py:421] (4/8) Epoch 6, batch 39800, loss[loss=2.35, over 1820.00 frames. , ppl: 10.48446140781446] tot_loss[loss=2.289, over 5502974.43 frames. , ppl: 9.863745632847627], batch size: 70
+2022-12-12 04:22:14,667 INFO [train.py:421] (4/8) Epoch 6, batch 40000, loss[loss=2.25, over 2870.00 frames. , ppl: 9.488116789410292] tot_loss[loss=2.29, over 5482262.21 frames. , ppl: 9.870398409657263], batch size: 70
+2022-12-12 04:22:14,667 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:22:15,413 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.796226785877156
+2022-12-12 04:23:58,060 INFO [train.py:421] (4/8) Epoch 6, batch 40200, loss[loss=2.265, over 3780.00 frames. , ppl: 9.628614832251985] tot_loss[loss=2.29, over 5479872.00 frames. , ppl: 9.877340391456931], batch size: 70
+2022-12-12 04:25:35,491 INFO [train.py:421] (4/8) Epoch 6, batch 40400, loss[loss=2.434, over 1680.00 frames. , ppl: 11.406938730617078] tot_loss[loss=2.292, over 5416055.63 frames. , ppl: 9.899410739269722], batch size: 70
+2022-12-12 04:27:15,613 INFO [train.py:421] (4/8) Epoch 6, batch 40600, loss[loss=2.213, over 2870.00 frames. , ppl: 9.139603935160514] tot_loss[loss=2.293, over 5439052.25 frames. , ppl: 9.899707938858082], batch size: 70
+2022-12-12 04:28:54,809 INFO [train.py:421] (4/8) Epoch 6, batch 40800, loss[loss=2.25, over 3430.00 frames. , ppl: 9.485072754812988] tot_loss[loss=2.29, over 5521247.38 frames. , ppl: 9.87510170298489], batch size: 70
+2022-12-12 04:30:38,979 INFO [train.py:421] (4/8) Epoch 6, batch 41000, loss[loss=2.615, over 980.00 frames. , ppl: 13.67109433598532] tot_loss[loss=2.29, over 5514425.94 frames. , ppl: 9.875748601103222], batch size: 70
+2022-12-12 04:30:38,979 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:30:39,735 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.77442840863681
+2022-12-12 04:32:21,592 INFO [train.py:421] (4/8) Epoch 6, batch 41200, loss[loss=2.348, over 2030.00 frames. , ppl: 10.46493639758801] tot_loss[loss=2.289, over 5533448.48 frames. , ppl: 9.86488196129088], batch size: 70
+2022-12-12 04:34:01,967 INFO [train.py:421] (4/8) Epoch 6, batch 41400, loss[loss=2.394, over 1120.00 frames. , ppl: 10.962384030299672] tot_loss[loss=2.29, over 5495320.28 frames. , ppl: 9.873258476575858], batch size: 70
+2022-12-12 04:35:43,840 INFO [train.py:421] (4/8) Epoch 6, batch 41600, loss[loss=2.203, over 6090.00 frames. , ppl: 9.048935638871386] tot_loss[loss=2.29, over 5490165.84 frames. , ppl: 9.872275776670886], batch size: 70
+2022-12-12 04:37:22,153 INFO [train.py:421] (4/8) Epoch 6, batch 41800, loss[loss=2.248, over 7070.00 frames. , ppl: 9.466995990364738] tot_loss[loss=2.29, over 5484421.83 frames. , ppl: 9.871554190502557], batch size: 70
+2022-12-12 04:39:01,064 INFO [train.py:421] (4/8) Epoch 6, batch 42000, loss[loss=2.292, over 2870.00 frames. , ppl: 9.896638519302392] tot_loss[loss=2.29, over 5477752.29 frames. , ppl: 9.87539222323836], batch size: 70
+2022-12-12 04:39:01,064 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:39:01,825 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.774656988070976
+2022-12-12 04:40:40,495 INFO [train.py:421] (4/8) Epoch 6, batch 42200, loss[loss=2.244, over 4200.00 frames. , ppl: 9.433194452536805] tot_loss[loss=2.29, over 5456736.13 frames. , ppl: 9.87482024087542], batch size: 70
+2022-12-12 04:42:19,271 INFO [train.py:421] (4/8) Epoch 6, batch 42400, loss[loss=2.234, over 3220.00 frames. , ppl: 9.338302499514867] tot_loss[loss=2.291, over 5436518.13 frames. , ppl: 9.88089771875179], batch size: 70
+2022-12-12 04:43:58,780 INFO [train.py:421] (4/8) Epoch 6, batch 42600, loss[loss=3.319, over 490.00 frames. , ppl: 27.624557948465622] tot_loss[loss=2.29, over 5436963.43 frames. , ppl: 9.87919982286589], batch size: 70
+2022-12-12 04:45:37,699 INFO [train.py:421] (4/8) Epoch 6, batch 42800, loss[loss=2.519, over 980.00 frames. , ppl: 12.414032391749867] tot_loss[loss=2.289, over 5481835.82 frames. , ppl: 9.863930009154068], batch size: 70
+2022-12-12 04:47:16,068 INFO [train.py:421] (4/8) Epoch 6, batch 43000, loss[loss=2.702, over 630.00 frames. , ppl: 14.915545781752728] tot_loss[loss=2.288, over 5511538.99 frames. , ppl: 9.854394495835722], batch size: 70
+2022-12-12 04:47:16,068 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:47:16,826 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.774992633382912
+2022-12-12 04:48:53,986 INFO [train.py:421] (4/8) Epoch 6, batch 43200, loss[loss=2.278, over 3010.00 frames. , ppl: 9.758394802621572] tot_loss[loss=2.289, over 5494519.64 frames. , ppl: 9.867085981248024], batch size: 70
+2022-12-12 04:50:34,705 INFO [train.py:421] (4/8) Epoch 6, batch 43400, loss[loss=3.166, over 490.00 frames. , ppl: 23.719987814986265] tot_loss[loss=2.29, over 5473264.05 frames. , ppl: 9.876617260765766], batch size: 70
+2022-12-12 04:52:11,333 INFO [train.py:421] (4/8) Epoch 6, batch 43600, loss[loss=2.736, over 770.00 frames. , ppl: 15.425189644678824] tot_loss[loss=2.29, over 5469008.15 frames. , ppl: 9.878974232003245], batch size: 70
+2022-12-12 04:53:48,620 INFO [train.py:421] (4/8) Epoch 6, batch 43800, loss[loss=2.295, over 1750.00 frames. , ppl: 9.926323329870577] tot_loss[loss=2.291, over 5438244.10 frames. , ppl: 9.887447240148862], batch size: 70
+2022-12-12 04:55:27,839 INFO [train.py:421] (4/8) Epoch 6, batch 44000, loss[loss=2.196, over 4760.00 frames. , ppl: 8.989828306272626] tot_loss[loss=2.29, over 5470588.84 frames. , ppl: 9.876710241096792], batch size: 70
+2022-12-12 04:55:27,839 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 04:55:28,598 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.775564124978258
+2022-12-12 04:57:06,981 INFO [train.py:421] (4/8) Epoch 6, batch 44200, loss[loss=2.71, over 770.00 frames. , ppl: 15.035461152858744] tot_loss[loss=2.29, over 5464940.06 frames. , ppl: 9.876377708952072], batch size: 70
+2022-12-12 04:58:47,001 INFO [train.py:421] (4/8) Epoch 6, batch 44400, loss[loss=3.028, over 560.00 frames. , ppl: 20.652896447691024] tot_loss[loss=2.29, over 5480195.72 frames. , ppl: 9.87340444162371], batch size: 70
+2022-12-12 05:00:25,090 INFO [train.py:421] (4/8) Epoch 6, batch 44600, loss[loss=2.418, over 2030.00 frames. , ppl: 11.223946418607866] tot_loss[loss=2.291, over 5425557.70 frames. , ppl: 9.879908134019812], batch size: 70
+2022-12-12 05:02:08,397 INFO [train.py:421] (4/8) Epoch 6, batch 44800, loss[loss=2.307, over 1960.00 frames. , ppl: 10.04633236552909] tot_loss[loss=2.291, over 5402268.38 frames. , ppl: 9.886773724951665], batch size: 70
+2022-12-12 05:03:45,269 INFO [train.py:421] (4/8) Epoch 6, batch 45000, loss[loss=2.197, over 3290.00 frames. , ppl: 9.000203801532255] tot_loss[loss=2.29, over 5434167.28 frames. , ppl: 9.875352757972703], batch size: 70
+2022-12-12 05:03:45,269 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:03:46,014 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.282, over 211138.00 frames. , ppl: 9.792734564010422
+2022-12-12 05:05:26,480 INFO [train.py:421] (4/8) Epoch 6, batch 45200, loss[loss=2.392, over 1960.00 frames. , ppl: 10.931389625005304] tot_loss[loss=2.292, over 5384987.39 frames. , ppl: 9.894380336341426], batch size: 70
+2022-12-12 05:07:06,361 INFO [train.py:421] (4/8) Epoch 6, batch 45400, loss[loss=2.312, over 2240.00 frames. , ppl: 10.091556978776882] tot_loss[loss=2.293, over 5363626.99 frames. , ppl: 9.899882676759871], batch size: 70
+2022-12-12 05:08:44,607 INFO [train.py:421] (4/8) Epoch 6, batch 45600, loss[loss=2.171, over 6510.00 frames. , ppl: 8.77089691619501] tot_loss[loss=2.292, over 5367466.09 frames. , ppl: 9.89821835593179], batch size: 70
+2022-12-12 05:10:22,563 INFO [train.py:421] (4/8) Epoch 6, batch 45800, loss[loss=2.485, over 1470.00 frames. , ppl: 11.999049130525142] tot_loss[loss=2.293, over 5363361.16 frames. , ppl: 9.904915054284706], batch size: 70
+2022-12-12 05:12:00,167 INFO [train.py:421] (4/8) Epoch 6, batch 46000, loss[loss=2.474, over 980.00 frames. , ppl: 11.865928918350836] tot_loss[loss=2.293, over 5369446.85 frames. , ppl: 9.90214843317354], batch size: 70
+2022-12-12 05:12:00,167 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:12:00,928 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.7772122333999
+2022-12-12 05:13:38,802 INFO [train.py:421] (4/8) Epoch 6, batch 46200, loss[loss=2.387, over 1610.00 frames. , ppl: 10.88343499968088] tot_loss[loss=2.292, over 5383359.38 frames. , ppl: 9.8931591524519], batch size: 70
+2022-12-12 05:15:18,368 INFO [train.py:421] (4/8) Epoch 6, batch 46400, loss[loss=2.335, over 1050.00 frames. , ppl: 10.327281768951595] tot_loss[loss=2.292, over 5400397.48 frames. , ppl: 9.890364274315818], batch size: 70
+2022-12-12 05:16:59,461 INFO [train.py:421] (4/8) Epoch 6, batch 46600, loss[loss=2.432, over 1610.00 frames. , ppl: 11.379134223775381] tot_loss[loss=2.292, over 5393469.27 frames. , ppl: 9.893402635511748], batch size: 70
+2022-12-12 05:18:44,138 INFO [train.py:421] (4/8) Epoch 6, batch 46800, loss[loss=3.143, over 490.00 frames. , ppl: 23.16914111996864] tot_loss[loss=2.291, over 5445804.83 frames. , ppl: 9.88240781445897], batch size: 70
+2022-12-12 05:20:24,482 INFO [train.py:421] (4/8) Epoch 6, batch 47000, loss[loss=2.154, over 6160.00 frames. , ppl: 8.620896422913521] tot_loss[loss=2.29, over 5462450.17 frames. , ppl: 9.87574625621719], batch size: 70
+2022-12-12 05:20:24,483 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:20:25,228 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.782431870183903
+2022-12-12 05:22:04,498 INFO [train.py:421] (4/8) Epoch 6, batch 47200, loss[loss=2.286, over 3360.00 frames. , ppl: 9.839479569681897] tot_loss[loss=2.289, over 5493258.50 frames. , ppl: 9.867926757341731], batch size: 70
+2022-12-12 05:23:42,974 INFO [train.py:421] (4/8) Epoch 6, batch 47400, loss[loss=2.219, over 4760.00 frames. , ppl: 9.196542061785719] tot_loss[loss=2.291, over 5466009.96 frames. , ppl: 9.887402781476382], batch size: 70
+2022-12-12 05:25:19,133 INFO [train.py:421] (4/8) Epoch 6, batch 47600, loss[loss=2.297, over 2590.00 frames. , ppl: 9.94734223965318] tot_loss[loss=2.293, over 5420884.50 frames. , ppl: 9.904902411359528], batch size: 70
+2022-12-12 05:26:59,803 INFO [train.py:421] (4/8) Epoch 6, batch 47800, loss[loss=2.2, over 6440.00 frames. , ppl: 9.022653051987765] tot_loss[loss=2.293, over 5412982.13 frames. , ppl: 9.908705732340676], batch size: 70
+2022-12-12 05:28:40,353 INFO [train.py:421] (4/8) Epoch 6, batch 48000, loss[loss=2.345, over 1610.00 frames. , ppl: 10.42839118382286] tot_loss[loss=2.293, over 5420058.18 frames. , ppl: 9.906943618179996], batch size: 70
+2022-12-12 05:28:40,353 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:28:41,113 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.767148491772168
+2022-12-12 05:30:22,489 INFO [train.py:421] (4/8) Epoch 6, batch 48200, loss[loss=2.723, over 630.00 frames. , ppl: 15.22950743354391] tot_loss[loss=2.294, over 5385440.53 frames. , ppl: 9.919167606326472], batch size: 70
+2022-12-12 05:31:59,718 INFO [train.py:421] (4/8) Epoch 6, batch 48400, loss[loss=2.278, over 2730.00 frames. , ppl: 9.759884888190165] tot_loss[loss=2.294, over 5408007.32 frames. , ppl: 9.91225858307533], batch size: 70
+2022-12-12 05:33:38,676 INFO [train.py:421] (4/8) Epoch 6, batch 48600, loss[loss=2.444, over 1400.00 frames. , ppl: 11.515099162282658] tot_loss[loss=2.294, over 5406480.49 frames. , ppl: 9.913486528062926], batch size: 70
+2022-12-12 05:35:20,109 INFO [train.py:421] (4/8) Epoch 6, batch 48800, loss[loss=2.154, over 8260.00 frames. , ppl: 8.622449684835816] tot_loss[loss=2.293, over 5408422.76 frames. , ppl: 9.908228234510277], batch size: 70
+2022-12-12 05:36:57,680 INFO [train.py:421] (4/8) Epoch 6, batch 49000, loss[loss=2.865, over 630.00 frames. , ppl: 17.54362233131364] tot_loss[loss=2.293, over 5419596.53 frames. , ppl: 9.904746482190752], batch size: 70
+2022-12-12 05:36:57,681 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:36:58,441 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.767533031836354
+2022-12-12 05:38:37,513 INFO [train.py:421] (4/8) Epoch 6, batch 49200, loss[loss=2.36, over 1050.00 frames. , ppl: 10.590848025678518] tot_loss[loss=2.294, over 5388934.94 frames. , ppl: 9.912579284145483], batch size: 70
+2022-12-12 05:40:16,673 INFO [train.py:421] (4/8) Epoch 6, batch 49400, loss[loss=2.344, over 1680.00 frames. , ppl: 10.424262009280195] tot_loss[loss=2.293, over 5399753.00 frames. , ppl: 9.908005812129023], batch size: 70
+2022-12-12 05:41:57,410 INFO [train.py:421] (4/8) Epoch 6, batch 49600, loss[loss=2.408, over 2240.00 frames. , ppl: 11.113226619991478] tot_loss[loss=2.293, over 5409050.46 frames. , ppl: 9.901236446742242], batch size: 70
+2022-12-12 05:43:39,741 INFO [train.py:421] (4/8) Epoch 6, batch 49800, loss[loss=2.259, over 4200.00 frames. , ppl: 9.573632852231501] tot_loss[loss=2.292, over 5415919.82 frames. , ppl: 9.895985640811322], batch size: 70
+2022-12-12 05:45:15,853 INFO [train.py:421] (4/8) Epoch 6, batch 50000, loss[loss=2.285, over 2590.00 frames. , ppl: 9.822868767524517] tot_loss[loss=2.292, over 5404533.72 frames. , ppl: 9.897934641002808], batch size: 70
+2022-12-12 05:45:15,853 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:45:16,606 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.785662608076919
+2022-12-12 05:46:53,731 INFO [train.py:421] (4/8) Epoch 6, batch 50200, loss[loss=2.337, over 2730.00 frames. , ppl: 10.351927047975101] tot_loss[loss=2.292, over 5424976.37 frames. , ppl: 9.896616017001497], batch size: 70
+2022-12-12 05:48:35,516 INFO [train.py:421] (4/8) Epoch 6, batch 50400, loss[loss=2.171, over 9870.00 frames. , ppl: 8.7650427466021] tot_loss[loss=2.291, over 5456466.52 frames. , ppl: 9.887125956392964], batch size: 70
+2022-12-12 05:50:15,397 INFO [train.py:421] (4/8) Epoch 6, batch 50600, loss[loss=2.335, over 3360.00 frames. , ppl: 10.330676188718343] tot_loss[loss=2.291, over 5467678.45 frames. , ppl: 9.882875665853053], batch size: 70
+2022-12-12 05:51:57,533 INFO [train.py:421] (4/8) Epoch 6, batch 50800, loss[loss=2.799, over 630.00 frames. , ppl: 16.433495127761912] tot_loss[loss=2.291, over 5455775.94 frames. , ppl: 9.885696363064241], batch size: 70
+2022-12-12 05:53:42,723 INFO [train.py:421] (4/8) Epoch 6, batch 51000, loss[loss=2.359, over 2030.00 frames. , ppl: 10.58415087427641] tot_loss[loss=2.29, over 5499230.53 frames. , ppl: 9.873174855035003], batch size: 70
+2022-12-12 05:53:42,723 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 05:53:43,469 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.76539078711788
+2022-12-12 05:55:23,427 INFO [train.py:421] (4/8) Epoch 6, batch 51200, loss[loss=2.793, over 700.00 frames. , ppl: 16.33184655797598] tot_loss[loss=2.289, over 5486607.44 frames. , ppl: 9.869725649591272], batch size: 70
+2022-12-12 05:57:03,660 INFO [train.py:421] (4/8) Epoch 6, batch 51400, loss[loss=2.238, over 4410.00 frames. , ppl: 9.377387253804656] tot_loss[loss=2.289, over 5506139.08 frames. , ppl: 9.863784712457496], batch size: 70
+2022-12-12 05:58:44,078 INFO [train.py:421] (4/8) Epoch 6, batch 51600, loss[loss=2.342, over 1750.00 frames. , ppl: 10.399268748919948] tot_loss[loss=2.289, over 5483192.05 frames. , ppl: 9.869321861484519], batch size: 70
+2022-12-12 06:00:26,223 INFO [train.py:421] (4/8) Epoch 6, batch 51800, loss[loss=2.284, over 2450.00 frames. , ppl: 9.819292679411053] tot_loss[loss=2.289, over 5487007.82 frames. , ppl: 9.86589160288073], batch size: 70
+2022-12-12 06:02:05,611 INFO [train.py:421] (4/8) Epoch 6, batch 52000, loss[loss=2.431, over 1190.00 frames. , ppl: 11.3693242292653] tot_loss[loss=2.289, over 5510264.76 frames. , ppl: 9.862785132217468], batch size: 70
+2022-12-12 06:02:05,611 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:02:06,356 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758165262568566
+2022-12-12 06:03:45,495 INFO [train.py:421] (4/8) Epoch 6, batch 52200, loss[loss=2.336, over 2590.00 frames. , ppl: 10.338776049448098] tot_loss[loss=2.287, over 5548165.31 frames. , ppl: 9.844390349119955], batch size: 70
+2022-12-12 06:05:23,973 INFO [train.py:421] (4/8) Epoch 6, batch 52400, loss[loss=2.839, over 700.00 frames. , ppl: 17.105113286391294] tot_loss[loss=2.288, over 5518741.58 frames. , ppl: 9.851443465809579], batch size: 70
+2022-12-12 06:07:01,650 INFO [train.py:421] (4/8) Epoch 6, batch 52600, loss[loss=2.383, over 910.00 frames. , ppl: 10.836380988337964] tot_loss[loss=2.287, over 5519933.83 frames. , ppl: 9.848681522049093], batch size: 70
+2022-12-12 06:08:40,078 INFO [train.py:421] (4/8) Epoch 6, batch 52800, loss[loss=2.284, over 1960.00 frames. , ppl: 9.819750131804257] tot_loss[loss=2.287, over 5515978.32 frames. , ppl: 9.846977115195665], batch size: 70
+2022-12-12 06:10:20,320 INFO [train.py:421] (4/8) Epoch 6, batch 53000, loss[loss=2.233, over 3780.00 frames. , ppl: 9.32371954860021] tot_loss[loss=2.287, over 5538904.70 frames. , ppl: 9.845645528254263], batch size: 70
+2022-12-12 06:10:20,320 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:10:21,067 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764269259359716
+2022-12-12 06:11:59,976 INFO [train.py:421] (4/8) Epoch 6, batch 53200, loss[loss=2.484, over 1120.00 frames. , ppl: 11.990717649113039] tot_loss[loss=2.287, over 5543839.40 frames. , ppl: 9.840681953333702], batch size: 70
+2022-12-12 06:13:40,687 INFO [train.py:421] (4/8) Epoch 6, batch 53400, loss[loss=2.27, over 3010.00 frames. , ppl: 9.6839670592964] tot_loss[loss=2.286, over 5543559.36 frames. , ppl: 9.837518180913648], batch size: 70
+2022-12-12 06:15:21,197 INFO [train.py:421] (4/8) Epoch 6, batch 53600, loss[loss=2.182, over 3920.00 frames. , ppl: 8.86153584263889] tot_loss[loss=2.287, over 5532466.40 frames. , ppl: 9.844538141894406], batch size: 70
+2022-12-12 06:17:02,504 INFO [train.py:421] (4/8) Epoch 6, batch 53800, loss[loss=2.237, over 6370.00 frames. , ppl: 9.366295434434791] tot_loss[loss=2.287, over 5541850.73 frames. , ppl: 9.842826977400188], batch size: 70
+2022-12-12 06:18:44,476 INFO [train.py:421] (4/8) Epoch 6, batch 54000, loss[loss=2.262, over 2800.00 frames. , ppl: 9.59846109675975] tot_loss[loss=2.287, over 5554840.00 frames. , ppl: 9.849054869585856], batch size: 70
+2022-12-12 06:18:44,477 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:18:45,222 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.773898935036751
+2022-12-12 06:20:25,788 INFO [train.py:421] (4/8) Epoch 6, batch 54200, loss[loss=2.611, over 840.00 frames. , ppl: 13.610888136204753] tot_loss[loss=2.287, over 5558589.46 frames. , ppl: 9.844037503543085], batch size: 70
+2022-12-12 06:22:07,243 INFO [train.py:421] (4/8) Epoch 6, batch 54400, loss[loss=2.362, over 3150.00 frames. , ppl: 10.611163656119475] tot_loss[loss=2.285, over 5602256.51 frames. , ppl: 9.829095320595101], batch size: 70
+2022-12-12 06:23:46,090 INFO [train.py:421] (4/8) Epoch 6, batch 54600, loss[loss=2.714, over 840.00 frames. , ppl: 15.092144818763694] tot_loss[loss=2.286, over 5575763.58 frames. , ppl: 9.83808234588335], batch size: 70
+2022-12-12 06:25:25,862 INFO [train.py:421] (4/8) Epoch 6, batch 54800, loss[loss=2.338, over 1680.00 frames. , ppl: 10.361599048127102] tot_loss[loss=2.287, over 5534851.48 frames. , ppl: 9.849978779410744], batch size: 70
+2022-12-12 06:27:05,727 INFO [train.py:421] (4/8) Epoch 6, batch 55000, loss[loss=2.313, over 2310.00 frames. , ppl: 10.103282772983853] tot_loss[loss=2.287, over 5549937.25 frames. , ppl: 9.842486032007884], batch size: 70
+2022-12-12 06:27:05,728 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:27:06,473 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758666441048378
+2022-12-12 06:28:46,934 INFO [train.py:421] (4/8) Epoch 6, batch 55200, loss[loss=2.269, over 4130.00 frames. , ppl: 9.665574286029653] tot_loss[loss=2.287, over 5560950.42 frames. , ppl: 9.840997127043787], batch size: 70
+2022-12-12 06:30:30,149 INFO [train.py:421] (4/8) Epoch 6, batch 55400, loss[loss=3.591, over 420.00 frames. , ppl: 36.27954024483227] tot_loss[loss=2.286, over 5592411.53 frames. , ppl: 9.833200840267097], batch size: 70
+2022-12-12 06:32:09,447 INFO [train.py:421] (4/8) Epoch 6, batch 55600, loss[loss=2.31, over 2660.00 frames. , ppl: 10.070904562297644] tot_loss[loss=2.286, over 5590474.60 frames. , ppl: 9.831842301813825], batch size: 70
+2022-12-12 06:33:49,115 INFO [train.py:421] (4/8) Epoch 6, batch 55800, loss[loss=2.705, over 630.00 frames. , ppl: 14.947894704964993] tot_loss[loss=2.285, over 5592054.35 frames. , ppl: 9.827486646826584], batch size: 70
+2022-12-12 06:35:30,414 INFO [train.py:421] (4/8) Epoch 6, batch 56000, loss[loss=2.534, over 1050.00 frames. , ppl: 12.608559769095885] tot_loss[loss=2.285, over 5616493.32 frames. , ppl: 9.826772680497761], batch size: 70
+2022-12-12 06:35:30,415 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:35:31,174 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.763342939887206
+2022-12-12 06:37:13,842 INFO [train.py:421] (4/8) Epoch 6, batch 56200, loss[loss=2.272, over 2170.00 frames. , ppl: 9.701925333070664] tot_loss[loss=2.287, over 5562974.72 frames. , ppl: 9.845279149673944], batch size: 70
+2022-12-12 06:38:54,618 INFO [train.py:421] (4/8) Epoch 6, batch 56400, loss[loss=2.372, over 1120.00 frames. , ppl: 10.718870535747495] tot_loss[loss=2.287, over 5560834.92 frames. , ppl: 9.849149797915606], batch size: 70
+2022-12-12 06:40:33,674 INFO [train.py:421] (4/8) Epoch 6, batch 56600, loss[loss=2.376, over 1680.00 frames. , ppl: 10.762666865995008] tot_loss[loss=2.288, over 5554156.89 frames. , ppl: 9.855396269880142], batch size: 70
+2022-12-12 06:42:12,585 INFO [train.py:421] (4/8) Epoch 6, batch 56800, loss[loss=2.428, over 1050.00 frames. , ppl: 11.34034343326407] tot_loss[loss=2.289, over 5528245.41 frames. , ppl: 9.861902388076455], batch size: 70
+2022-12-12 06:43:53,583 INFO [train.py:421] (4/8) Epoch 6, batch 57000, loss[loss=2.385, over 1260.00 frames. , ppl: 10.855659641595835] tot_loss[loss=2.289, over 5545183.44 frames. , ppl: 9.864057400948859], batch size: 70
+2022-12-12 06:43:53,583 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:43:54,329 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.763007694592458
+2022-12-12 06:45:35,860 INFO [train.py:421] (4/8) Epoch 6, batch 57200, loss[loss=2.494, over 840.00 frames. , ppl: 12.106981400201244] tot_loss[loss=2.289, over 5512188.16 frames. , ppl: 9.864073765180818], batch size: 70
+2022-12-12 06:47:18,363 INFO [train.py:421] (4/8) Epoch 6, batch 57400, loss[loss=2.305, over 3570.00 frames. , ppl: 10.023494400233616] tot_loss[loss=2.29, over 5490749.73 frames. , ppl: 9.874740749219898], batch size: 70
+2022-12-12 06:48:59,494 INFO [train.py:421] (4/8) Epoch 6, batch 57600, loss[loss=2.336, over 1680.00 frames. , ppl: 10.339247504502302] tot_loss[loss=2.287, over 5591982.04 frames. , ppl: 9.844438921121355], batch size: 70
+2022-12-12 06:50:45,868 INFO [train.py:421] (4/8) Epoch 6, batch 57800, loss[loss=2.771, over 700.00 frames. , ppl: 15.968215850432939] tot_loss[loss=2.287, over 5593447.86 frames. , ppl: 9.843188460967808], batch size: 70
+2022-12-12 06:52:25,579 INFO [train.py:421] (4/8) Epoch 6, batch 58000, loss[loss=2.308, over 2590.00 frames. , ppl: 10.059099203262967] tot_loss[loss=2.284, over 5644076.45 frames. , ppl: 9.819423445619195], batch size: 70
+2022-12-12 06:52:25,580 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 06:52:26,339 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758448345808779
+2022-12-12 06:54:02,192 INFO [train.py:421] (4/8) Epoch 6, batch 58200, loss[loss=2.733, over 770.00 frames. , ppl: 15.381678995990123] tot_loss[loss=2.285, over 5626500.12 frames. , ppl: 9.82277480732246], batch size: 70
+2022-12-12 06:55:40,804 INFO [train.py:421] (4/8) Epoch 6, batch 58400, loss[loss=2.478, over 1330.00 frames. , ppl: 11.911734542121936] tot_loss[loss=2.286, over 5572309.54 frames. , ppl: 9.837504685269645], batch size: 70
+2022-12-12 06:57:22,165 INFO [train.py:421] (4/8) Epoch 6, batch 58600, loss[loss=2.254, over 2730.00 frames. , ppl: 9.523030079198177] tot_loss[loss=2.287, over 5541679.57 frames. , ppl: 9.847605959458011], batch size: 70
+2022-12-12 06:59:04,023 INFO [train.py:421] (4/8) Epoch 6, batch 58800, loss[loss=2.34, over 2590.00 frames. , ppl: 10.38358381671473] tot_loss[loss=2.287, over 5555957.68 frames. , ppl: 9.846433260309997], batch size: 70
+2022-12-12 07:00:45,054 INFO [train.py:421] (4/8) Epoch 6, batch 59000, loss[loss=2.261, over 3080.00 frames. , ppl: 9.597425655077501] tot_loss[loss=2.287, over 5569143.47 frames. , ppl: 9.848921569127945], batch size: 70
+2022-12-12 07:00:45,054 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:00:45,814 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.762743263537212
+2022-12-12 07:02:25,671 INFO [train.py:421] (4/8) Epoch 6, batch 59200, loss[loss=2.678, over 910.00 frames. , ppl: 14.553624655974124] tot_loss[loss=2.287, over 5572677.20 frames. , ppl: 9.842901597151444], batch size: 70
+2022-12-12 07:04:06,627 INFO [train.py:421] (4/8) Epoch 6, batch 59400, loss[loss=2.202, over 6230.00 frames. , ppl: 9.046512574981643] tot_loss[loss=2.287, over 5541139.97 frames. , ppl: 9.847226011910093], batch size: 70
+2022-12-12 07:05:46,487 INFO [train.py:421] (4/8) Epoch 6, batch 59600, loss[loss=2.139, over 5390.00 frames. , ppl: 8.492345507962815] tot_loss[loss=2.285, over 5602729.81 frames. , ppl: 9.826743482694239], batch size: 70
+2022-12-12 07:07:27,937 INFO [train.py:421] (4/8) Epoch 6, batch 59800, loss[loss=2.383, over 910.00 frames. , ppl: 10.838840804073188] tot_loss[loss=2.285, over 5597899.40 frames. , ppl: 9.823620768127098], batch size: 70
+2022-12-12 07:09:08,960 INFO [train.py:421] (4/8) Epoch 6, batch 60000, loss[loss=2.255, over 1540.00 frames. , ppl: 9.53663091846671] tot_loss[loss=2.286, over 5567280.61 frames. , ppl: 9.836383076264013], batch size: 70
+2022-12-12 07:09:08,960 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:09:09,719 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.763438313496728
+2022-12-12 07:10:50,271 INFO [train.py:421] (4/8) Epoch 6, batch 60200, loss[loss=2.484, over 1050.00 frames. , ppl: 11.985353478538862] tot_loss[loss=2.285, over 5574576.05 frames. , ppl: 9.830384561763886], batch size: 70
+2022-12-12 07:12:28,651 INFO [train.py:421] (4/8) Epoch 6, batch 60400, loss[loss=2.182, over 4690.00 frames. , ppl: 8.863219422907516] tot_loss[loss=2.286, over 5564535.03 frames. , ppl: 9.837764632831075], batch size: 70
+2022-12-12 07:14:10,685 INFO [train.py:421] (4/8) Epoch 6, batch 60600, loss[loss=2.239, over 3710.00 frames. , ppl: 9.381562880259331] tot_loss[loss=2.287, over 5528863.13 frames. , ppl: 9.846615468973582], batch size: 70
+2022-12-12 07:15:51,737 INFO [train.py:421] (4/8) Epoch 6, batch 60800, loss[loss=2.258, over 5250.00 frames. , ppl: 9.566838803960673] tot_loss[loss=2.287, over 5498532.44 frames. , ppl: 9.846070599114626], batch size: 70
+2022-12-12 07:17:31,929 INFO [train.py:421] (4/8) Epoch 6, batch 61000, loss[loss=2.206, over 6160.00 frames. , ppl: 9.078477108907116] tot_loss[loss=2.287, over 5512756.65 frames. , ppl: 9.846821665614193], batch size: 70
+2022-12-12 07:17:31,930 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:17:32,674 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.767613989568469
+2022-12-12 07:19:10,822 INFO [train.py:421] (4/8) Epoch 6, batch 61200, loss[loss=2.291, over 2450.00 frames. , ppl: 9.889196720179783] tot_loss[loss=2.286, over 5548591.39 frames. , ppl: 9.837466052394735], batch size: 70
+2022-12-12 07:20:50,819 INFO [train.py:421] (4/8) Epoch 6, batch 61400, loss[loss=2.23, over 3570.00 frames. , ppl: 9.296243183374122] tot_loss[loss=2.286, over 5557541.22 frames. , ppl: 9.835729552244644], batch size: 70
+2022-12-12 07:22:30,434 INFO [train.py:421] (4/8) Epoch 6, batch 61600, loss[loss=2.316, over 2800.00 frames. , ppl: 10.130217710720146] tot_loss[loss=2.286, over 5562249.11 frames. , ppl: 9.839815310250634], batch size: 70
+2022-12-12 07:24:09,143 INFO [train.py:421] (4/8) Epoch 6, batch 61800, loss[loss=2.307, over 3780.00 frames. , ppl: 10.040894477070864] tot_loss[loss=2.288, over 5532900.09 frames. , ppl: 9.850717767528035], batch size: 70
+2022-12-12 07:25:49,776 INFO [train.py:421] (4/8) Epoch 6, batch 62000, loss[loss=2.33, over 2940.00 frames. , ppl: 10.277703241163337] tot_loss[loss=2.286, over 5546760.71 frames. , ppl: 9.83969976535346], batch size: 70
+2022-12-12 07:25:49,776 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:25:50,536 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.776349799606294
+2022-12-12 07:27:29,140 INFO [train.py:421] (4/8) Epoch 6, batch 62200, loss[loss=4.088, over 350.00 frames. , ppl: 59.642585347042775] tot_loss[loss=2.288, over 5514427.25 frames. , ppl: 9.852272630774207], batch size: 70
+2022-12-12 07:29:06,005 INFO [train.py:421] (4/8) Epoch 6, batch 62400, loss[loss=2.246, over 4760.00 frames. , ppl: 9.44962409351575] tot_loss[loss=2.287, over 5523608.48 frames. , ppl: 9.843693735197602], batch size: 70
+2022-12-12 07:30:49,310 INFO [train.py:421] (4/8) Epoch 6, batch 62600, loss[loss=2.418, over 910.00 frames. , ppl: 11.222491360934605] tot_loss[loss=2.286, over 5574515.76 frames. , ppl: 9.831467940442321], batch size: 70
+2022-12-12 07:32:26,314 INFO [train.py:421] (4/8) Epoch 6, batch 62800, loss[loss=2.218, over 3080.00 frames. , ppl: 9.184651133949549] tot_loss[loss=2.287, over 5514310.51 frames. , ppl: 9.848234544009413], batch size: 70
+2022-12-12 07:34:06,412 INFO [train.py:421] (4/8) Epoch 6, batch 63000, loss[loss=2.361, over 2310.00 frames. , ppl: 10.601996612746131] tot_loss[loss=2.29, over 5459171.01 frames. , ppl: 9.870426045874163], batch size: 70
+2022-12-12 07:34:06,412 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:34:07,172 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.784635781017636
+2022-12-12 07:35:46,484 INFO [train.py:421] (4/8) Epoch 6, batch 63200, loss[loss=2.215, over 8610.00 frames. , ppl: 9.164452530231921] tot_loss[loss=2.29, over 5418589.96 frames. , ppl: 9.874975832300608], batch size: 70
+2022-12-12 07:37:31,525 INFO [train.py:421] (4/8) Epoch 6, batch 63400, loss[loss=2.269, over 4970.00 frames. , ppl: 9.667711833562686] tot_loss[loss=2.29, over 5414061.42 frames. , ppl: 9.87317921401541], batch size: 70
+2022-12-12 07:39:13,150 INFO [train.py:421] (4/8) Epoch 6, batch 63600, loss[loss=2.416, over 1540.00 frames. , ppl: 11.2034636616779] tot_loss[loss=2.289, over 5456542.79 frames. , ppl: 9.86079854011949], batch size: 70
+2022-12-12 07:40:49,344 INFO [train.py:421] (4/8) Epoch 6, batch 63800, loss[loss=2.818, over 630.00 frames. , ppl: 16.735548500014804] tot_loss[loss=2.29, over 5406466.01 frames. , ppl: 9.875754838741237], batch size: 70
+2022-12-12 07:42:29,285 INFO [train.py:421] (4/8) Epoch 6, batch 64000, loss[loss=2.231, over 3640.00 frames. , ppl: 9.306360907176401] tot_loss[loss=2.291, over 5371208.75 frames. , ppl: 9.883943217969804], batch size: 70
+2022-12-12 07:42:29,286 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:42:30,050 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.768704092947166
+2022-12-12 07:44:07,948 INFO [train.py:421] (4/8) Epoch 6, batch 64200, loss[loss=3.112, over 560.00 frames. , ppl: 22.475177972717464] tot_loss[loss=2.292, over 5379165.12 frames. , ppl: 9.894699536326751], batch size: 70
+2022-12-12 07:45:49,265 INFO [train.py:421] (4/8) Epoch 6, batch 64400, loss[loss=2.393, over 1540.00 frames. , ppl: 10.95043858269897] tot_loss[loss=2.291, over 5424108.48 frames. , ppl: 9.88238508744282], batch size: 70
+2022-12-12 07:47:28,533 INFO [train.py:421] (4/8) Epoch 6, batch 64600, loss[loss=2.281, over 2660.00 frames. , ppl: 9.783478394574425] tot_loss[loss=2.289, over 5478541.58 frames. , ppl: 9.86664562945809], batch size: 70
+2022-12-12 07:49:11,547 INFO [train.py:421] (4/8) Epoch 6, batch 64800, loss[loss=2.468, over 1050.00 frames. , ppl: 11.793226534080096] tot_loss[loss=2.289, over 5476025.83 frames. , ppl: 9.867738153697008], batch size: 70
+2022-12-12 07:50:54,147 INFO [train.py:421] (4/8) Epoch 6, batch 65000, loss[loss=2.262, over 3430.00 frames. , ppl: 9.602940685277746] tot_loss[loss=2.29, over 5447361.13 frames. , ppl: 9.877519298798848], batch size: 70
+2022-12-12 07:50:54,148 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:50:54,908 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.281, over 211138.00 frames. , ppl: 9.786995181821556
+2022-12-12 07:52:32,871 INFO [train.py:421] (4/8) Epoch 6, batch 65200, loss[loss=2.359, over 1960.00 frames. , ppl: 10.576442534152667] tot_loss[loss=2.289, over 5481784.64 frames. , ppl: 9.863251694113954], batch size: 70
+2022-12-12 07:54:12,249 INFO [train.py:421] (4/8) Epoch 6, batch 65400, loss[loss=2.477, over 910.00 frames. , ppl: 11.903441583048822] tot_loss[loss=2.288, over 5501789.93 frames. , ppl: 9.859708993769443], batch size: 70
+2022-12-12 07:55:56,355 INFO [train.py:421] (4/8) Epoch 6, batch 65600, loss[loss=2.272, over 2660.00 frames. , ppl: 9.70223177240842] tot_loss[loss=2.287, over 5537846.57 frames. , ppl: 9.846965024283357], batch size: 70
+2022-12-12 07:57:38,642 INFO [train.py:421] (4/8) Epoch 6, batch 65800, loss[loss=2.265, over 1540.00 frames. , ppl: 9.635387592979207] tot_loss[loss=2.287, over 5550285.57 frames. , ppl: 9.843635281887494], batch size: 70
+2022-12-12 07:59:17,514 INFO [train.py:421] (4/8) Epoch 6, batch 66000, loss[loss=2.452, over 1470.00 frames. , ppl: 11.611071748180946] tot_loss[loss=2.288, over 5518105.70 frames. , ppl: 9.853470616418534], batch size: 70
+2022-12-12 07:59:17,514 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 07:59:18,273 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.756986800195627
+2022-12-12 08:00:55,094 INFO [train.py:421] (4/8) Epoch 6, batch 66200, loss[loss=2.249, over 9380.00 frames. , ppl: 9.480705455904317] tot_loss[loss=2.288, over 5534783.19 frames. , ppl: 9.860007930100409], batch size: 70
+2022-12-12 08:02:34,946 INFO [train.py:421] (4/8) Epoch 6, batch 66400, loss[loss=2.493, over 770.00 frames. , ppl: 12.097374046215354] tot_loss[loss=2.289, over 5493917.65 frames. , ppl: 9.86767438095395], batch size: 70
+2022-12-12 08:04:12,889 INFO [train.py:421] (4/8) Epoch 6, batch 66600, loss[loss=2.442, over 910.00 frames. , ppl: 11.496404202881891] tot_loss[loss=2.289, over 5487031.94 frames. , ppl: 9.869591721192402], batch size: 70
+2022-12-12 08:05:56,269 INFO [train.py:421] (4/8) Epoch 6, batch 66800, loss[loss=2.302, over 3150.00 frames. , ppl: 9.997301715907927] tot_loss[loss=2.289, over 5508934.27 frames. , ppl: 9.862417058483942], batch size: 70
+2022-12-12 08:07:41,384 INFO [train.py:421] (4/8) Epoch 6, batch 67000, loss[loss=2.297, over 4200.00 frames. , ppl: 9.943440946455304] tot_loss[loss=2.288, over 5544662.59 frames. , ppl: 9.856311247855485], batch size: 70
+2022-12-12 08:07:41,384 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 08:07:42,129 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.766049889291654
+2022-12-12 08:09:21,214 INFO [train.py:421] (4/8) Epoch 6, batch 67200, loss[loss=2.377, over 2170.00 frames. , ppl: 10.773781948719478] tot_loss[loss=2.289, over 5498309.88 frames. , ppl: 9.869060330829647], batch size: 70
+2022-12-12 08:11:01,749 INFO [train.py:421] (4/8) Epoch 6, batch 67400, loss[loss=2.293, over 2170.00 frames. , ppl: 9.900197668954783] tot_loss[loss=2.29, over 5448512.38 frames. , ppl: 9.874348803675186], batch size: 70
+2022-12-12 08:12:41,351 INFO [train.py:421] (4/8) Epoch 6, batch 67600, loss[loss=2.366, over 2240.00 frames. , ppl: 10.650920946906183] tot_loss[loss=2.291, over 5410892.15 frames. , ppl: 9.883841555793778], batch size: 70
+2022-12-12 08:14:15,982 INFO [train.py:421] (4/8) Epoch 6, batch 67800, loss[loss=2.303, over 1960.00 frames. , ppl: 9.999783838020875] tot_loss[loss=2.291, over 5401608.62 frames. , ppl: 9.884887419438893], batch size: 70
+2022-12-12 08:15:54,492 INFO [train.py:421] (4/8) Epoch 6, batch 68000, loss[loss=2.319, over 2310.00 frames. , ppl: 10.164488009025858] tot_loss[loss=2.29, over 5405404.57 frames. , ppl: 9.872568158062855], batch size: 70
+2022-12-12 08:15:54,493 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 08:15:55,254 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.28, over 211138.00 frames. , ppl: 9.779122592522853
+2022-12-12 08:17:39,591 INFO [train.py:421] (4/8) Epoch 6, batch 68200, loss[loss=2.313, over 3850.00 frames. , ppl: 10.108222776408455] tot_loss[loss=2.289, over 5454028.13 frames. , ppl: 9.867144788101028], batch size: 70
+2022-12-12 08:19:22,896 INFO [train.py:421] (4/8) Epoch 6, batch 68400, loss[loss=2.229, over 4830.00 frames. , ppl: 9.289035033052604] tot_loss[loss=2.289, over 5483784.72 frames. , ppl: 9.860971612792818], batch size: 70
+2022-12-12 08:21:04,974 INFO [train.py:421] (4/8) Epoch 6, batch 68600, loss[loss=2.26, over 2800.00 frames. , ppl: 9.578604827884744] tot_loss[loss=2.288, over 5499691.52 frames. , ppl: 9.855951582626524], batch size: 70
+2022-12-12 08:22:43,487 INFO [train.py:421] (4/8) Epoch 6, batch 68800, loss[loss=2.243, over 3710.00 frames. , ppl: 9.423372290848716] tot_loss[loss=2.288, over 5519677.25 frames. , ppl: 9.855715713866317], batch size: 70
+2022-12-12 08:24:23,647 INFO [train.py:421] (4/8) Epoch 6, batch 69000, loss[loss=2.493, over 1610.00 frames. , ppl: 12.095503721442833] tot_loss[loss=2.288, over 5496948.21 frames. , ppl: 9.858048697943744], batch size: 70
+2022-12-12 08:24:23,648 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 08:24:24,395 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.279, over 211138.00 frames. , ppl: 9.762691245156288
+2022-12-12 08:26:05,591 INFO [train.py:421] (4/8) Epoch 6, batch 69200, loss[loss=2.463, over 1050.00 frames. , ppl: 11.742653036988122] tot_loss[loss=2.287, over 5534922.34 frames. , ppl: 9.840519339526296], batch size: 70
+2022-12-12 08:27:44,995 INFO [train.py:421] (4/8) Epoch 6, batch 69400, loss[loss=2.325, over 3990.00 frames. , ppl: 10.222816175300492] tot_loss[loss=2.287, over 5521990.09 frames. , ppl: 9.844899805690622], batch size: 70
+2022-12-12 08:29:30,567 INFO [train.py:421] (4/8) Epoch 6, batch 69600, loss[loss=2.262, over 3640.00 frames. , ppl: 9.603219196754948] tot_loss[loss=2.286, over 5567775.77 frames. , ppl: 9.832510875559446], batch size: 70
+2022-12-12 08:31:11,134 INFO [train.py:421] (4/8) Epoch 6, batch 69800, loss[loss=2.209, over 4760.00 frames. , ppl: 9.104942287471385] tot_loss[loss=2.286, over 5576530.96 frames. , ppl: 9.832061802761189], batch size: 70
+2022-12-12 08:32:55,420 INFO [train.py:421] (4/8) Epoch 6, batch 70000, loss[loss=2.214, over 3360.00 frames. , ppl: 9.15529332298375] tot_loss[loss=2.286, over 5566804.47 frames. , ppl: 9.833807574797534], batch size: 70
+2022-12-12 08:32:55,421 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 08:32:56,166 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.761897999189102
+2022-12-12 08:34:38,123 INFO [train.py:421] (4/8) Epoch 6, batch 70200, loss[loss=2.439, over 1960.00 frames. , ppl: 11.457088228685791] tot_loss[loss=2.286, over 5572969.42 frames. , ppl: 9.83900302114277], batch size: 70
+2022-12-12 08:36:17,125 INFO [train.py:421] (4/8) Epoch 6, batch 70400, loss[loss=2.243, over 2870.00 frames. , ppl: 9.42581241280919] tot_loss[loss=2.287, over 5563206.52 frames. , ppl: 9.844362734967445], batch size: 70
+2022-12-12 08:37:57,468 INFO [train.py:421] (4/8) Epoch 6, batch 70600, loss[loss=2.189, over 5740.00 frames. , ppl: 8.92846700582759] tot_loss[loss=2.288, over 5560332.08 frames. , ppl: 9.854659584042887], batch size: 70
+2022-12-12 08:39:34,984 INFO [train.py:421] (4/8) Epoch 6, batch 70800, loss[loss=2.334, over 1890.00 frames. , ppl: 10.321325993928452] tot_loss[loss=2.289, over 5525487.67 frames. , ppl: 9.865468398484868], batch size: 70
+2022-12-12 08:41:17,309 INFO [train.py:421] (4/8) Epoch 6, batch 71000, loss[loss=2.368, over 2940.00 frames. , ppl: 10.671565510364013] tot_loss[loss=2.29, over 5516785.50 frames. , ppl: 9.871735439999137], batch size: 70
+2022-12-12 08:41:17,310 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 08:41:18,069 INFO [train.py:452] (4/8) Epoch 6, validation: loss=2.278, over 211138.00 frames. , ppl: 9.760633851500042
+2022-12-12 08:42:55,817 INFO [train.py:421] (4/8) Epoch 6, batch 71200, loss[loss=2.311, over 1750.00 frames. , ppl: 10.080461560623364] tot_loss[loss=2.291, over 5482674.70 frames. , ppl: 9.882990577402401], batch size: 70
+2022-12-12 08:44:37,451 INFO [train.py:421] (4/8) Epoch 6, batch 71400, loss[loss=2.823, over 700.00 frames. , ppl: 16.833880544644718] tot_loss[loss=2.291, over 5457568.49 frames. , ppl: 9.88207555931043], batch size: 70
+2022-12-12 08:46:15,622 INFO [train.py:421] (4/8) Epoch 6, batch 71600, loss[loss=2.555, over 840.00 frames. , ppl: 12.873936556822917] tot_loss[loss=2.292, over 5413785.92 frames. , ppl: 9.892984803008597], batch size: 70
+2022-12-12 08:47:54,855 INFO [train.py:421] (4/8) Epoch 6, batch 71800, loss[loss=2.31, over 3780.00 frames. , ppl: 10.071884193989003] tot_loss[loss=2.291, over 5403038.24 frames. , ppl: 9.887129742352936], batch size: 70
+2022-12-12 08:49:08,129 INFO [train.py:421] (4/8) Epoch 7, batch 0, loss[loss=2.564, over 910.00 frames. , ppl: 12.991424296648793] tot_loss[loss=2.564, over 910.00 frames. , ppl: 12.991424296648793], batch size: 70
+2022-12-12 08:50:48,795 INFO [train.py:421] (4/8) Epoch 7, batch 200, loss[loss=2.247, over 2100.00 frames. , ppl: 9.461707245258676] tot_loss[loss=2.296, over 488641.65 frames. , ppl: 9.931626304960373], batch size: 70
+2022-12-12 08:52:29,295 INFO [train.py:421] (4/8) Epoch 7, batch 400, loss[loss=2.399, over 2590.00 frames. , ppl: 11.006725558413535] tot_loss[loss=2.284, over 991625.19 frames. , ppl: 9.814596530435923], batch size: 70
+2022-12-12 08:54:08,470 INFO [train.py:421] (4/8) Epoch 7, batch 600, loss[loss=2.147, over 4620.00 frames. , ppl: 8.555453585476773] tot_loss[loss=2.286, over 1389417.75 frames. , ppl: 9.83996936032109], batch size: 70
+2022-12-12 08:55:44,620 INFO [train.py:421] (4/8) Epoch 7, batch 800, loss[loss=2.356, over 1890.00 frames. , ppl: 10.547466898517863] tot_loss[loss=2.288, over 1753779.89 frames. , ppl: 9.852355753036425], batch size: 70
+2022-12-12 08:57:20,593 INFO [train.py:421] (4/8) Epoch 7, batch 1000, loss[loss=2.198, over 3780.00 frames. , ppl: 9.007110517604385] tot_loss[loss=2.283, over 2143231.91 frames. , ppl: 9.807785816017681], batch size: 70
+2022-12-12 08:57:20,593 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 08:57:21,340 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.766622303585429
+2022-12-12 08:58:56,585 INFO [train.py:421] (4/8) Epoch 7, batch 1200, loss[loss=2.196, over 4060.00 frames. , ppl: 8.988630445794312] tot_loss[loss=2.283, over 2456213.48 frames. , ppl: 9.80920132215615], batch size: 70
+2022-12-12 09:00:32,854 INFO [train.py:421] (4/8) Epoch 7, batch 1400, loss[loss=2.384, over 1820.00 frames. , ppl: 10.845082983412249] tot_loss[loss=2.283, over 2738785.33 frames. , ppl: 9.808995622329439], batch size: 70
+2022-12-12 09:02:12,540 INFO [train.py:421] (4/8) Epoch 7, batch 1600, loss[loss=2.439, over 1330.00 frames. , ppl: 11.46567575299292] tot_loss[loss=2.282, over 3009007.87 frames. , ppl: 9.797550933594291], batch size: 70
+2022-12-12 09:03:50,400 INFO [train.py:421] (4/8) Epoch 7, batch 1800, loss[loss=2.192, over 6020.00 frames. , ppl: 8.955791382832963] tot_loss[loss=2.283, over 3231139.34 frames. , ppl: 9.805165656881714], batch size: 70
+2022-12-12 09:05:32,156 INFO [train.py:421] (4/8) Epoch 7, batch 2000, loss[loss=2.784, over 630.00 frames. , ppl: 16.180250403916368] tot_loss[loss=2.282, over 3446340.14 frames. , ppl: 9.798131924360431], batch size: 70
+2022-12-12 09:05:32,156 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:05:32,901 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764543848320063
+2022-12-12 09:07:08,131 INFO [train.py:421] (4/8) Epoch 7, batch 2200, loss[loss=2.356, over 2170.00 frames. , ppl: 10.5486272579583] tot_loss[loss=2.283, over 3613498.35 frames. , ppl: 9.81090467186696], batch size: 70
+2022-12-12 09:08:47,628 INFO [train.py:421] (4/8) Epoch 7, batch 2400, loss[loss=2.159, over 4060.00 frames. , ppl: 8.659341925871534] tot_loss[loss=2.283, over 3781956.98 frames. , ppl: 9.807260360285701], batch size: 70
+2022-12-12 09:10:26,432 INFO [train.py:421] (4/8) Epoch 7, batch 2600, loss[loss=2.358, over 1960.00 frames. , ppl: 10.567851190390991] tot_loss[loss=2.283, over 3913937.75 frames. , ppl: 9.810557682835856], batch size: 70
+2022-12-12 09:12:07,859 INFO [train.py:421] (4/8) Epoch 7, batch 2800, loss[loss=2.202, over 7630.00 frames. , ppl: 9.045169542077083] tot_loss[loss=2.282, over 4100008.20 frames. , ppl: 9.796378005290746], batch size: 70
+2022-12-12 09:13:50,922 INFO [train.py:421] (4/8) Epoch 7, batch 3000, loss[loss=2.473, over 1470.00 frames. , ppl: 11.85530559501746] tot_loss[loss=2.281, over 4243716.46 frames. , ppl: 9.784832930402356], batch size: 70
+2022-12-12 09:13:50,922 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:13:51,668 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.743588919093323
+2022-12-12 09:15:28,747 INFO [train.py:421] (4/8) Epoch 7, batch 3200, loss[loss=2.185, over 5040.00 frames. , ppl: 8.888654610262344] tot_loss[loss=2.281, over 4358334.35 frames. , ppl: 9.787212800718942], batch size: 70
+2022-12-12 09:17:10,260 INFO [train.py:421] (4/8) Epoch 7, batch 3400, loss[loss=2.351, over 2030.00 frames. , ppl: 10.50052647515863] tot_loss[loss=2.282, over 4456320.66 frames. , ppl: 9.801121700111416], batch size: 70
+2022-12-12 09:18:51,745 INFO [train.py:421] (4/8) Epoch 7, batch 3600, loss[loss=2.339, over 2310.00 frames. , ppl: 10.371585444883282] tot_loss[loss=2.283, over 4547247.67 frames. , ppl: 9.805683077286563], batch size: 70
+2022-12-12 09:20:35,630 INFO [train.py:421] (4/8) Epoch 7, batch 3800, loss[loss=2.396, over 2100.00 frames. , ppl: 10.980384910537161] tot_loss[loss=2.282, over 4661153.08 frames. , ppl: 9.796366892865567], batch size: 70
+2022-12-12 09:22:13,512 INFO [train.py:421] (4/8) Epoch 7, batch 4000, loss[loss=2.416, over 980.00 frames. , ppl: 11.201012380566315] tot_loss[loss=2.283, over 4708424.58 frames. , ppl: 9.808026183398049], batch size: 70
+2022-12-12 09:22:13,513 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:22:14,273 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764824226068129
+2022-12-12 09:23:52,017 INFO [train.py:421] (4/8) Epoch 7, batch 4200, loss[loss=2.524, over 910.00 frames. , ppl: 12.47619243072937] tot_loss[loss=2.283, over 4792742.77 frames. , ppl: 9.803729276246266], batch size: 70
+2022-12-12 09:25:30,245 INFO [train.py:421] (4/8) Epoch 7, batch 4400, loss[loss=2.319, over 1050.00 frames. , ppl: 10.166978257789172] tot_loss[loss=2.281, over 4902173.32 frames. , ppl: 9.788361988865686], batch size: 70
+2022-12-12 09:27:09,447 INFO [train.py:421] (4/8) Epoch 7, batch 4600, loss[loss=2.278, over 3150.00 frames. , ppl: 9.761786995148078] tot_loss[loss=2.281, over 4946410.90 frames. , ppl: 9.78856785038435], batch size: 70
+2022-12-12 09:28:50,536 INFO [train.py:421] (4/8) Epoch 7, batch 4800, loss[loss=2.305, over 3710.00 frames. , ppl: 10.028703959160415] tot_loss[loss=2.281, over 5012941.00 frames. , ppl: 9.78822022688762], batch size: 70
+2022-12-12 09:30:30,393 INFO [train.py:421] (4/8) Epoch 7, batch 5000, loss[loss=2.262, over 3220.00 frames. , ppl: 9.604737496200116] tot_loss[loss=2.281, over 5069344.96 frames. , ppl: 9.786506321021777], batch size: 70
+2022-12-12 09:30:30,394 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:30:31,152 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.770388643082995
+2022-12-12 09:32:12,050 INFO [train.py:421] (4/8) Epoch 7, batch 5200, loss[loss=2.288, over 1680.00 frames. , ppl: 9.858048473514422] tot_loss[loss=2.28, over 5157262.95 frames. , ppl: 9.772045560118343], batch size: 70
+2022-12-12 09:33:52,020 INFO [train.py:421] (4/8) Epoch 7, batch 5400, loss[loss=2.37, over 3220.00 frames. , ppl: 10.699625898979027] tot_loss[loss=2.28, over 5169671.79 frames. , ppl: 9.779601632767113], batch size: 70
+2022-12-12 09:35:30,665 INFO [train.py:421] (4/8) Epoch 7, batch 5600, loss[loss=2.338, over 2450.00 frames. , ppl: 10.358480172293769] tot_loss[loss=2.28, over 5237464.80 frames. , ppl: 9.778893557860133], batch size: 70
+2022-12-12 09:37:13,202 INFO [train.py:421] (4/8) Epoch 7, batch 5800, loss[loss=2.243, over 2730.00 frames. , ppl: 9.422435422610217] tot_loss[loss=2.282, over 5212864.65 frames. , ppl: 9.79190397315476], batch size: 70
+2022-12-12 09:38:55,653 INFO [train.py:421] (4/8) Epoch 7, batch 6000, loss[loss=2.271, over 3710.00 frames. , ppl: 9.685520655912587] tot_loss[loss=2.281, over 5227867.15 frames. , ppl: 9.78882384518625], batch size: 70
+2022-12-12 09:38:55,653 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:38:56,400 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.766977911177703
+2022-12-12 09:40:36,648 INFO [train.py:421] (4/8) Epoch 7, batch 6200, loss[loss=2.515, over 980.00 frames. , ppl: 12.37247232974552] tot_loss[loss=2.282, over 5240346.10 frames. , ppl: 9.7916765307546], batch size: 70
+2022-12-12 09:42:17,027 INFO [train.py:421] (4/8) Epoch 7, batch 6400, loss[loss=2.193, over 2170.00 frames. , ppl: 8.96165391882688] tot_loss[loss=2.281, over 5287863.46 frames. , ppl: 9.786931890205839], batch size: 70
+2022-12-12 09:43:57,320 INFO [train.py:421] (4/8) Epoch 7, batch 6600, loss[loss=2.174, over 4480.00 frames. , ppl: 8.792945562267912] tot_loss[loss=2.281, over 5320323.40 frames. , ppl: 9.782199268229846], batch size: 70
+2022-12-12 09:45:38,137 INFO [train.py:421] (4/8) Epoch 7, batch 6800, loss[loss=2.221, over 2940.00 frames. , ppl: 9.214041177836908] tot_loss[loss=2.28, over 5376162.82 frames. , ppl: 9.778150599614957], batch size: 70
+2022-12-12 09:47:22,573 INFO [train.py:421] (4/8) Epoch 7, batch 7000, loss[loss=2.265, over 5250.00 frames. , ppl: 9.629884960243649] tot_loss[loss=2.28, over 5419386.67 frames. , ppl: 9.775080354815763], batch size: 70
+2022-12-12 09:47:22,573 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:47:23,332 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.761162606501355
+2022-12-12 09:49:03,422 INFO [train.py:421] (4/8) Epoch 7, batch 7200, loss[loss=2.104, over 3220.00 frames. , ppl: 8.201594345831976] tot_loss[loss=2.28, over 5408758.97 frames. , ppl: 9.779458031469213], batch size: 70
+2022-12-12 09:50:43,930 INFO [train.py:421] (4/8) Epoch 7, batch 7400, loss[loss=2.307, over 2380.00 frames. , ppl: 10.046903897436419] tot_loss[loss=2.281, over 5420762.32 frames. , ppl: 9.786166135467289], batch size: 70
+2022-12-12 09:52:24,698 INFO [train.py:421] (4/8) Epoch 7, batch 7600, loss[loss=2.688, over 700.00 frames. , ppl: 14.70508510350802] tot_loss[loss=2.28, over 5438449.75 frames. , ppl: 9.779146162994994], batch size: 70
+2022-12-12 09:54:01,291 INFO [train.py:421] (4/8) Epoch 7, batch 7800, loss[loss=2.248, over 2730.00 frames. , ppl: 9.471772462062493] tot_loss[loss=2.28, over 5452853.37 frames. , ppl: 9.774591925506966], batch size: 70
+2022-12-12 09:55:38,140 INFO [train.py:421] (4/8) Epoch 7, batch 8000, loss[loss=2.155, over 7560.00 frames. , ppl: 8.62939913067592] tot_loss[loss=2.28, over 5449386.90 frames. , ppl: 9.77753947441804], batch size: 70
+2022-12-12 09:55:38,141 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 09:55:38,887 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.762243320567698
+2022-12-12 09:57:20,375 INFO [train.py:421] (4/8) Epoch 7, batch 8200, loss[loss=2.166, over 6720.00 frames. , ppl: 8.719986885034658] tot_loss[loss=2.28, over 5471158.50 frames. , ppl: 9.773405742566315], batch size: 70
+2022-12-12 09:59:00,379 INFO [train.py:421] (4/8) Epoch 7, batch 8400, loss[loss=2.314, over 2520.00 frames. , ppl: 10.113651816953803] tot_loss[loss=2.28, over 5447625.36 frames. , ppl: 9.78122132195104], batch size: 70
+2022-12-12 10:00:40,150 INFO [train.py:421] (4/8) Epoch 7, batch 8600, loss[loss=2.327, over 1190.00 frames. , ppl: 10.25165703682846] tot_loss[loss=2.28, over 5444901.22 frames. , ppl: 9.77952693817445], batch size: 70
+2022-12-12 10:02:20,359 INFO [train.py:421] (4/8) Epoch 7, batch 8800, loss[loss=2.281, over 2940.00 frames. , ppl: 9.791277923303635] tot_loss[loss=2.28, over 5437225.20 frames. , ppl: 9.77900806476215], batch size: 70
+2022-12-12 10:04:02,209 INFO [train.py:421] (4/8) Epoch 7, batch 9000, loss[loss=2.308, over 1820.00 frames. , ppl: 10.052891001502505] tot_loss[loss=2.279, over 5457368.97 frames. , ppl: 9.770071296824144], batch size: 70
+2022-12-12 10:04:02,210 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:04:02,970 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.758306803162158
+2022-12-12 10:05:41,185 INFO [train.py:421] (4/8) Epoch 7, batch 9200, loss[loss=2.46, over 1120.00 frames. , ppl: 11.707413794281386] tot_loss[loss=2.279, over 5442543.14 frames. , ppl: 9.771189460158249], batch size: 70
+2022-12-12 10:07:18,510 INFO [train.py:421] (4/8) Epoch 7, batch 9400, loss[loss=2.205, over 8330.00 frames. , ppl: 9.069582110271897] tot_loss[loss=2.28, over 5433373.05 frames. , ppl: 9.780928366274319], batch size: 70
+2022-12-12 10:08:57,690 INFO [train.py:421] (4/8) Epoch 7, batch 9600, loss[loss=2.244, over 2100.00 frames. , ppl: 9.435333219091989] tot_loss[loss=2.28, over 5483353.35 frames. , ppl: 9.777303617535827], batch size: 70
+2022-12-12 10:10:41,235 INFO [train.py:421] (4/8) Epoch 7, batch 9800, loss[loss=2.395, over 1750.00 frames. , ppl: 10.9672922408431] tot_loss[loss=2.28, over 5475094.11 frames. , ppl: 9.778837152753628], batch size: 70
+2022-12-12 10:12:17,286 INFO [train.py:421] (4/8) Epoch 7, batch 10000, loss[loss=2.268, over 1750.00 frames. , ppl: 9.661627471793574] tot_loss[loss=2.281, over 5466391.72 frames. , ppl: 9.787356225637753], batch size: 70
+2022-12-12 10:12:17,286 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:12:18,035 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.743937919324686
+2022-12-12 10:13:56,296 INFO [train.py:421] (4/8) Epoch 7, batch 10200, loss[loss=2.194, over 4410.00 frames. , ppl: 8.970463442451619] tot_loss[loss=2.281, over 5445720.52 frames. , ppl: 9.790420893464304], batch size: 70
+2022-12-12 10:15:40,296 INFO [train.py:421] (4/8) Epoch 7, batch 10400, loss[loss=2.416, over 1470.00 frames. , ppl: 11.201233756929042] tot_loss[loss=2.281, over 5445371.56 frames. , ppl: 9.790096889406737], batch size: 70
+2022-12-12 10:17:21,296 INFO [train.py:421] (4/8) Epoch 7, batch 10600, loss[loss=2.147, over 8050.00 frames. , ppl: 8.555286519856045] tot_loss[loss=2.282, over 5433890.72 frames. , ppl: 9.7971783701798], batch size: 70
+2022-12-12 10:18:59,591 INFO [train.py:421] (4/8) Epoch 7, batch 10800, loss[loss=2.589, over 980.00 frames. , ppl: 13.319603350271572] tot_loss[loss=2.283, over 5421513.86 frames. , ppl: 9.805697510141771], batch size: 70
+2022-12-12 10:20:44,636 INFO [train.py:421] (4/8) Epoch 7, batch 11000, loss[loss=2.718, over 630.00 frames. , ppl: 15.15226923715606] tot_loss[loss=2.284, over 5367748.86 frames. , ppl: 9.820530682083968], batch size: 70
+2022-12-12 10:20:44,637 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:20:45,398 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.764081387151732
+2022-12-12 10:22:26,536 INFO [train.py:421] (4/8) Epoch 7, batch 11200, loss[loss=2.227, over 6090.00 frames. , ppl: 9.272983214100387] tot_loss[loss=2.285, over 5365766.14 frames. , ppl: 9.828245755608352], batch size: 70
+2022-12-12 10:24:09,637 INFO [train.py:421] (4/8) Epoch 7, batch 11400, loss[loss=2.477, over 1190.00 frames. , ppl: 11.904823898881405] tot_loss[loss=2.284, over 5395818.89 frames. , ppl: 9.82022671903433], batch size: 70
+2022-12-12 10:25:51,819 INFO [train.py:421] (4/8) Epoch 7, batch 11600, loss[loss=2.259, over 5180.00 frames. , ppl: 9.572162954821367] tot_loss[loss=2.285, over 5393156.17 frames. , ppl: 9.82309617695717], batch size: 70
+2022-12-12 10:27:33,032 INFO [train.py:421] (4/8) Epoch 7, batch 11800, loss[loss=2.303, over 3710.00 frames. , ppl: 10.005574351251727] tot_loss[loss=2.284, over 5412202.18 frames. , ppl: 9.818412926431037], batch size: 70
+2022-12-12 10:29:14,367 INFO [train.py:421] (4/8) Epoch 7, batch 12000, loss[loss=2.561, over 1050.00 frames. , ppl: 12.953147653855417] tot_loss[loss=2.284, over 5422560.42 frames. , ppl: 9.813958424832773], batch size: 70
+2022-12-12 10:29:14,367 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:29:15,131 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.770513007653056
+2022-12-12 10:30:57,657 INFO [train.py:421] (4/8) Epoch 7, batch 12200, loss[loss=2.373, over 1540.00 frames. , ppl: 10.734867713728322] tot_loss[loss=2.284, over 5420202.46 frames. , ppl: 9.814279059766276], batch size: 70
+2022-12-12 10:32:39,418 INFO [train.py:421] (4/8) Epoch 7, batch 12400, loss[loss=2.417, over 1750.00 frames. , ppl: 11.209228858254392] tot_loss[loss=2.284, over 5400707.23 frames. , ppl: 9.817918253762715], batch size: 70
+2022-12-12 10:34:19,730 INFO [train.py:421] (4/8) Epoch 7, batch 12600, loss[loss=2.367, over 2030.00 frames. , ppl: 10.664238786369529] tot_loss[loss=2.283, over 5421526.42 frames. , ppl: 9.808720762676861], batch size: 70
+2022-12-12 10:35:58,810 INFO [train.py:421] (4/8) Epoch 7, batch 12800, loss[loss=2.355, over 2660.00 frames. , ppl: 10.540074299114554] tot_loss[loss=2.282, over 5444203.55 frames. , ppl: 9.799914723997368], batch size: 70
+2022-12-12 10:37:35,199 INFO [train.py:421] (4/8) Epoch 7, batch 13000, loss[loss=2.3, over 3430.00 frames. , ppl: 9.974491994593958] tot_loss[loss=2.284, over 5396792.58 frames. , ppl: 9.817625652078812], batch size: 70
+2022-12-12 10:37:35,200 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:37:35,946 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.279, over 211138.00 frames. , ppl: 9.769163881244294
+2022-12-12 10:39:15,582 INFO [train.py:421] (4/8) Epoch 7, batch 13200, loss[loss=2.542, over 1050.00 frames. , ppl: 12.710789146119403] tot_loss[loss=2.284, over 5406282.55 frames. , ppl: 9.81344237099415], batch size: 70
+2022-12-12 10:40:55,457 INFO [train.py:421] (4/8) Epoch 7, batch 13400, loss[loss=2.375, over 2170.00 frames. , ppl: 10.75575089167011] tot_loss[loss=2.284, over 5413636.53 frames. , ppl: 9.811865843364648], batch size: 70
+2022-12-12 10:42:34,882 INFO [train.py:421] (4/8) Epoch 7, batch 13600, loss[loss=2.263, over 2310.00 frames. , ppl: 9.609879659574723] tot_loss[loss=2.284, over 5404259.13 frames. , ppl: 9.815812474677228], batch size: 70
+2022-12-12 10:44:12,330 INFO [train.py:421] (4/8) Epoch 7, batch 13800, loss[loss=2.239, over 4970.00 frames. , ppl: 9.383076222458145] tot_loss[loss=2.284, over 5422528.91 frames. , ppl: 9.812628382382393], batch size: 70
+2022-12-12 10:45:49,784 INFO [train.py:421] (4/8) Epoch 7, batch 14000, loss[loss=2.332, over 1960.00 frames. , ppl: 10.30186263253118] tot_loss[loss=2.284, over 5414868.73 frames. , ppl: 9.817946965192144], batch size: 70
+2022-12-12 10:45:49,785 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:45:50,534 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.282, over 211138.00 frames. , ppl: 9.796524022897163
+2022-12-12 10:47:32,824 INFO [train.py:421] (4/8) Epoch 7, batch 14200, loss[loss=2.336, over 2030.00 frames. , ppl: 10.340003868673652] tot_loss[loss=2.284, over 5425496.80 frames. , ppl: 9.817097092696633], batch size: 70
+2022-12-12 10:49:07,720 INFO [train.py:421] (4/8) Epoch 7, batch 14400, loss[loss=2.165, over 4410.00 frames. , ppl: 8.712601730925481] tot_loss[loss=2.285, over 5390363.55 frames. , ppl: 9.82540296571202], batch size: 70
+2022-12-12 10:50:46,721 INFO [train.py:421] (4/8) Epoch 7, batch 14600, loss[loss=2.355, over 2100.00 frames. , ppl: 10.53416261826365] tot_loss[loss=2.284, over 5409761.91 frames. , ppl: 9.815928587431673], batch size: 70
+2022-12-12 10:52:28,728 INFO [train.py:421] (4/8) Epoch 7, batch 14800, loss[loss=2.181, over 6370.00 frames. , ppl: 8.855839485080601] tot_loss[loss=2.283, over 5445762.38 frames. , ppl: 9.807704995015701], batch size: 70
+2022-12-12 10:54:11,258 INFO [train.py:421] (4/8) Epoch 7, batch 15000, loss[loss=2.282, over 1960.00 frames. , ppl: 9.793948434892911] tot_loss[loss=2.282, over 5486654.64 frames. , ppl: 9.798365164083169], batch size: 70
+2022-12-12 10:54:11,259 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 10:54:12,022 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.747404083511375
+2022-12-12 10:55:53,710 INFO [train.py:421] (4/8) Epoch 7, batch 15200, loss[loss=2.459, over 1260.00 frames. , ppl: 11.694710812081762] tot_loss[loss=2.282, over 5500825.01 frames. , ppl: 9.800195843119194], batch size: 70
+2022-12-12 10:57:32,864 INFO [train.py:421] (4/8) Epoch 7, batch 15400, loss[loss=2.247, over 3010.00 frames. , ppl: 9.457080472443861] tot_loss[loss=2.281, over 5533065.54 frames. , ppl: 9.787077117719429], batch size: 70
+2022-12-12 10:59:08,564 INFO [train.py:421] (4/8) Epoch 7, batch 15600, loss[loss=2.085, over 3220.00 frames. , ppl: 8.04127944947087] tot_loss[loss=2.281, over 5518176.80 frames. , ppl: 9.790154391251614], batch size: 70
+2022-12-12 11:00:48,237 INFO [train.py:421] (4/8) Epoch 7, batch 15800, loss[loss=2.207, over 3430.00 frames. , ppl: 9.084430775467872] tot_loss[loss=2.283, over 5477193.31 frames. , ppl: 9.80783058357206], batch size: 70
+2022-12-12 11:02:25,848 INFO [train.py:421] (4/8) Epoch 7, batch 16000, loss[loss=2.451, over 1190.00 frames. , ppl: 11.60469143611769] tot_loss[loss=2.284, over 5473996.06 frames. , ppl: 9.811596513482714], batch size: 70
+2022-12-12 11:02:25,849 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:02:26,595 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.759092535215544
+2022-12-12 11:04:07,742 INFO [train.py:421] (4/8) Epoch 7, batch 16200, loss[loss=2.29, over 2520.00 frames. , ppl: 9.876913647041302] tot_loss[loss=2.285, over 5419657.71 frames. , ppl: 9.828322978038361], batch size: 70
+2022-12-12 11:05:46,866 INFO [train.py:421] (4/8) Epoch 7, batch 16400, loss[loss=2.515, over 1890.00 frames. , ppl: 12.363533100623556] tot_loss[loss=2.287, over 5384481.57 frames. , ppl: 9.841873128162488], batch size: 70
+2022-12-12 11:07:27,128 INFO [train.py:421] (4/8) Epoch 7, batch 16600, loss[loss=2.253, over 3570.00 frames. , ppl: 9.511580597944276] tot_loss[loss=2.287, over 5370837.86 frames. , ppl: 9.847218333776675], batch size: 70
+2022-12-12 11:09:03,975 INFO [train.py:421] (4/8) Epoch 7, batch 16800, loss[loss=2.671, over 700.00 frames. , ppl: 14.460632635762208] tot_loss[loss=2.288, over 5345561.87 frames. , ppl: 9.856148973190033], batch size: 70
+2022-12-12 11:10:44,190 INFO [train.py:421] (4/8) Epoch 7, batch 17000, loss[loss=2.254, over 9310.00 frames. , ppl: 9.521161843366832] tot_loss[loss=2.287, over 5391152.39 frames. , ppl: 9.845417055823939], batch size: 70
+2022-12-12 11:10:44,191 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:10:44,952 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.755632321970248
+2022-12-12 11:12:25,660 INFO [train.py:421] (4/8) Epoch 7, batch 17200, loss[loss=2.447, over 1190.00 frames. , ppl: 11.556112758323332] tot_loss[loss=2.286, over 5394200.19 frames. , ppl: 9.838690136443647], batch size: 70
+2022-12-12 11:14:04,105 INFO [train.py:421] (4/8) Epoch 7, batch 17400, loss[loss=2.259, over 4200.00 frames. , ppl: 9.572764746872204] tot_loss[loss=2.285, over 5426587.53 frames. , ppl: 9.823162676013352], batch size: 70
+2022-12-12 11:15:49,431 INFO [train.py:421] (4/8) Epoch 7, batch 17600, loss[loss=2.404, over 2100.00 frames. , ppl: 11.068095443573808] tot_loss[loss=2.285, over 5413630.93 frames. , ppl: 9.825341764363493], batch size: 70
+2022-12-12 11:17:26,212 INFO [train.py:421] (4/8) Epoch 7, batch 17800, loss[loss=2.353, over 1820.00 frames. , ppl: 10.516023862179692] tot_loss[loss=2.286, over 5375736.15 frames. , ppl: 9.837949893984122], batch size: 70
+2022-12-12 11:19:03,125 INFO [train.py:421] (4/8) Epoch 7, batch 18000, loss[loss=2.154, over 10500.00 frames. , ppl: 8.616621584994782] tot_loss[loss=2.287, over 5342107.06 frames. , ppl: 9.847277727977694], batch size: 70
+2022-12-12 11:19:03,126 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:19:03,871 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.756920371505483
+2022-12-12 11:20:45,119 INFO [train.py:421] (4/8) Epoch 7, batch 18200, loss[loss=2.288, over 2380.00 frames. , ppl: 9.859761662149253] tot_loss[loss=2.286, over 5391485.51 frames. , ppl: 9.835094150235054], batch size: 70
+2022-12-12 11:22:26,377 INFO [train.py:421] (4/8) Epoch 7, batch 18400, loss[loss=2.36, over 1120.00 frames. , ppl: 10.593567007503786] tot_loss[loss=2.288, over 5354783.92 frames. , ppl: 9.853104949246337], batch size: 70
+2022-12-12 11:24:05,728 INFO [train.py:421] (4/8) Epoch 7, batch 18600, loss[loss=2.352, over 2310.00 frames. , ppl: 10.503546904403322] tot_loss[loss=2.288, over 5363510.11 frames. , ppl: 9.852083318278444], batch size: 70
+2022-12-12 11:25:43,136 INFO [train.py:421] (4/8) Epoch 7, batch 18800, loss[loss=2.482, over 910.00 frames. , ppl: 11.965493399597802] tot_loss[loss=2.287, over 5339567.35 frames. , ppl: 9.846840554114774], batch size: 70
+2022-12-12 11:27:23,641 INFO [train.py:421] (4/8) Epoch 7, batch 19000, loss[loss=2.633, over 700.00 frames. , ppl: 13.917494193737156] tot_loss[loss=2.286, over 5384662.35 frames. , ppl: 9.833621649937594], batch size: 70
+2022-12-12 11:27:23,642 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:27:24,402 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.753232844662795
+2022-12-12 11:29:02,259 INFO [train.py:421] (4/8) Epoch 7, batch 19200, loss[loss=2.277, over 2940.00 frames. , ppl: 9.751646003414566] tot_loss[loss=2.287, over 5374981.22 frames. , ppl: 9.845939059258244], batch size: 70
+2022-12-12 11:30:43,927 INFO [train.py:421] (4/8) Epoch 7, batch 19400, loss[loss=2.28, over 4970.00 frames. , ppl: 9.773299205923296] tot_loss[loss=2.286, over 5400421.64 frames. , ppl: 9.835021437171216], batch size: 70
+2022-12-12 11:32:24,645 INFO [train.py:421] (4/8) Epoch 7, batch 19600, loss[loss=2.258, over 3850.00 frames. , ppl: 9.561707646542752] tot_loss[loss=2.285, over 5414802.10 frames. , ppl: 9.825254250156409], batch size: 70
+2022-12-12 11:34:05,671 INFO [train.py:421] (4/8) Epoch 7, batch 19800, loss[loss=2.226, over 3710.00 frames. , ppl: 9.266767126368567] tot_loss[loss=2.285, over 5400107.35 frames. , ppl: 9.828064048631271], batch size: 70
+2022-12-12 11:35:44,818 INFO [train.py:421] (4/8) Epoch 7, batch 20000, loss[loss=2.392, over 1540.00 frames. , ppl: 10.940521810072626] tot_loss[loss=2.284, over 5450090.89 frames. , ppl: 9.81943583102805], batch size: 70
+2022-12-12 11:35:44,818 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:35:45,578 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.73576853706518
+2022-12-12 11:37:27,967 INFO [train.py:421] (4/8) Epoch 7, batch 20200, loss[loss=2.404, over 1190.00 frames. , ppl: 11.069685070503674] tot_loss[loss=2.283, over 5467847.45 frames. , ppl: 9.810383320966782], batch size: 70
+2022-12-12 11:39:05,123 INFO [train.py:421] (4/8) Epoch 7, batch 20400, loss[loss=2.138, over 6930.00 frames. , ppl: 8.4788356631158] tot_loss[loss=2.282, over 5494395.45 frames. , ppl: 9.799343835279684], batch size: 70
+2022-12-12 11:40:47,485 INFO [train.py:421] (4/8) Epoch 7, batch 20600, loss[loss=2.378, over 2240.00 frames. , ppl: 10.780222773127228] tot_loss[loss=2.282, over 5527401.09 frames. , ppl: 9.79362702989244], batch size: 70
+2022-12-12 11:42:27,078 INFO [train.py:421] (4/8) Epoch 7, batch 20800, loss[loss=2.369, over 2240.00 frames. , ppl: 10.691749771573704] tot_loss[loss=2.282, over 5496818.89 frames. , ppl: 9.799949644568008], batch size: 70
+2022-12-12 11:44:06,627 INFO [train.py:421] (4/8) Epoch 7, batch 21000, loss[loss=2.215, over 2870.00 frames. , ppl: 9.162706940968956] tot_loss[loss=2.282, over 5492857.87 frames. , ppl: 9.79621678777076], batch size: 70
+2022-12-12 11:44:06,628 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:44:07,376 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.750929208941372
+2022-12-12 11:45:45,792 INFO [train.py:421] (4/8) Epoch 7, batch 21200, loss[loss=2.272, over 2870.00 frames. , ppl: 9.699345443696991] tot_loss[loss=2.282, over 5487890.81 frames. , ppl: 9.795235282481674], batch size: 70
+2022-12-12 11:47:26,537 INFO [train.py:421] (4/8) Epoch 7, batch 21400, loss[loss=2.713, over 700.00 frames. , ppl: 15.068950521647523] tot_loss[loss=2.282, over 5501760.69 frames. , ppl: 9.791828416440255], batch size: 70
+2022-12-12 11:49:07,973 INFO [train.py:421] (4/8) Epoch 7, batch 21600, loss[loss=2.18, over 4900.00 frames. , ppl: 8.849099379166676] tot_loss[loss=2.281, over 5530473.95 frames. , ppl: 9.782913178111123], batch size: 70
+2022-12-12 11:50:45,600 INFO [train.py:421] (4/8) Epoch 7, batch 21800, loss[loss=2.41, over 1330.00 frames. , ppl: 11.13934008702432] tot_loss[loss=2.28, over 5556279.20 frames. , ppl: 9.77766108251617], batch size: 70
+2022-12-12 11:52:26,437 INFO [train.py:421] (4/8) Epoch 7, batch 22000, loss[loss=2.514, over 840.00 frames. , ppl: 12.355095925519649] tot_loss[loss=2.28, over 5563222.50 frames. , ppl: 9.772469830070031], batch size: 70
+2022-12-12 11:52:26,437 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 11:52:27,187 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.739122243408765
+2022-12-12 11:54:04,986 INFO [train.py:421] (4/8) Epoch 7, batch 22200, loss[loss=2.201, over 3500.00 frames. , ppl: 9.032117448463971] tot_loss[loss=2.279, over 5579926.81 frames. , ppl: 9.766273694840098], batch size: 70
+2022-12-12 11:55:47,110 INFO [train.py:421] (4/8) Epoch 7, batch 22400, loss[loss=2.514, over 1260.00 frames. , ppl: 12.352746484015368] tot_loss[loss=2.279, over 5624945.37 frames. , ppl: 9.76384901707164], batch size: 70
+2022-12-12 11:57:29,541 INFO [train.py:421] (4/8) Epoch 7, batch 22600, loss[loss=2.374, over 2170.00 frames. , ppl: 10.743436713948014] tot_loss[loss=2.28, over 5594845.64 frames. , ppl: 9.776915656770596], batch size: 70
+2022-12-12 11:59:10,057 INFO [train.py:421] (4/8) Epoch 7, batch 22800, loss[loss=2.241, over 4480.00 frames. , ppl: 9.402970520788367] tot_loss[loss=2.28, over 5582976.57 frames. , ppl: 9.776615759003413], batch size: 70
+2022-12-12 12:00:50,386 INFO [train.py:421] (4/8) Epoch 7, batch 23000, loss[loss=2.292, over 2450.00 frames. , ppl: 9.891187531327203] tot_loss[loss=2.28, over 5599478.91 frames. , ppl: 9.774755273462457], batch size: 70
+2022-12-12 12:00:50,387 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:00:51,133 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.741689827028287
+2022-12-12 12:02:30,432 INFO [train.py:421] (4/8) Epoch 7, batch 23200, loss[loss=2.18, over 6720.00 frames. , ppl: 8.846130909917338] tot_loss[loss=2.28, over 5609990.85 frames. , ppl: 9.772929921807224], batch size: 70
+2022-12-12 12:04:11,331 INFO [train.py:421] (4/8) Epoch 7, batch 23400, loss[loss=2.381, over 1260.00 frames. , ppl: 10.819000197360031] tot_loss[loss=2.281, over 5569544.96 frames. , ppl: 9.78469942489222], batch size: 70
+2022-12-12 12:05:50,807 INFO [train.py:421] (4/8) Epoch 7, batch 23600, loss[loss=2.553, over 910.00 frames. , ppl: 12.845565276485702] tot_loss[loss=2.28, over 5593375.60 frames. , ppl: 9.781218216220545], batch size: 70
+2022-12-12 12:07:34,132 INFO [train.py:421] (4/8) Epoch 7, batch 23800, loss[loss=2.262, over 1750.00 frames. , ppl: 9.604300217482344] tot_loss[loss=2.28, over 5600646.69 frames. , ppl: 9.779552363010257], batch size: 70
+2022-12-12 12:09:14,676 INFO [train.py:421] (4/8) Epoch 7, batch 24000, loss[loss=2.327, over 1330.00 frames. , ppl: 10.245509362897648] tot_loss[loss=2.281, over 5562726.24 frames. , ppl: 9.787115344568551], batch size: 70
+2022-12-12 12:09:14,677 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:09:15,425 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.74665824206101
+2022-12-12 12:10:54,504 INFO [train.py:421] (4/8) Epoch 7, batch 24200, loss[loss=2.268, over 2800.00 frames. , ppl: 9.661937826907252] tot_loss[loss=2.281, over 5592572.98 frames. , ppl: 9.78253133793865], batch size: 70
+2022-12-12 12:12:37,632 INFO [train.py:421] (4/8) Epoch 7, batch 24400, loss[loss=2.166, over 5880.00 frames. , ppl: 8.723728112100224] tot_loss[loss=2.282, over 5587796.01 frames. , ppl: 9.793116551341383], batch size: 70
+2022-12-12 12:14:18,797 INFO [train.py:421] (4/8) Epoch 7, batch 24600, loss[loss=2.367, over 1610.00 frames. , ppl: 10.66542983878039] tot_loss[loss=2.283, over 5544582.79 frames. , ppl: 9.803602065179152], batch size: 70
+2022-12-12 12:15:58,739 INFO [train.py:421] (4/8) Epoch 7, batch 24800, loss[loss=2.328, over 1680.00 frames. , ppl: 10.25545538841244] tot_loss[loss=2.281, over 5570258.28 frames. , ppl: 9.789654786995301], batch size: 70
+2022-12-12 12:17:36,489 INFO [train.py:421] (4/8) Epoch 7, batch 25000, loss[loss=2.21, over 4200.00 frames. , ppl: 9.11696276759036] tot_loss[loss=2.281, over 5575711.88 frames. , ppl: 9.789807312218533], batch size: 70
+2022-12-12 12:17:36,489 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:17:37,246 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.73495874762003
+2022-12-12 12:19:18,780 INFO [train.py:421] (4/8) Epoch 7, batch 25200, loss[loss=2.271, over 4410.00 frames. , ppl: 9.687026292889474] tot_loss[loss=2.282, over 5567018.13 frames. , ppl: 9.800171070797475], batch size: 70
+2022-12-12 12:20:57,119 INFO [train.py:421] (4/8) Epoch 7, batch 25400, loss[loss=2.383, over 1190.00 frames. , ppl: 10.836797246650248] tot_loss[loss=2.283, over 5523926.70 frames. , ppl: 9.810843749529155], batch size: 70
+2022-12-12 12:22:35,788 INFO [train.py:421] (4/8) Epoch 7, batch 25600, loss[loss=2.295, over 2800.00 frames. , ppl: 9.929148053197347] tot_loss[loss=2.284, over 5500640.69 frames. , ppl: 9.818469790939616], batch size: 70
+2022-12-12 12:24:16,860 INFO [train.py:421] (4/8) Epoch 7, batch 25800, loss[loss=2.355, over 1400.00 frames. , ppl: 10.541830647236646] tot_loss[loss=2.283, over 5536079.67 frames. , ppl: 9.803888540764659], batch size: 70
+2022-12-12 12:25:56,646 INFO [train.py:421] (4/8) Epoch 7, batch 26000, loss[loss=2.375, over 1890.00 frames. , ppl: 10.746450684447593] tot_loss[loss=2.284, over 5532773.01 frames. , ppl: 9.81166131513333], batch size: 70
+2022-12-12 12:25:56,647 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:25:57,404 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.75006332120515
+2022-12-12 12:27:34,575 INFO [train.py:421] (4/8) Epoch 7, batch 26200, loss[loss=2.274, over 3360.00 frames. , ppl: 9.721426739968992] tot_loss[loss=2.284, over 5500139.04 frames. , ppl: 9.811087370425183], batch size: 70
+2022-12-12 12:29:16,217 INFO [train.py:421] (4/8) Epoch 7, batch 26400, loss[loss=2.17, over 5950.00 frames. , ppl: 8.758640543276307] tot_loss[loss=2.283, over 5535692.00 frames. , ppl: 9.806103746992497], batch size: 70
+2022-12-12 12:30:55,036 INFO [train.py:421] (4/8) Epoch 7, batch 26600, loss[loss=2.453, over 1680.00 frames. , ppl: 11.618401945644754] tot_loss[loss=2.283, over 5528854.64 frames. , ppl: 9.802021925978327], batch size: 70
+2022-12-12 12:32:37,070 INFO [train.py:421] (4/8) Epoch 7, batch 26800, loss[loss=2.262, over 4270.00 frames. , ppl: 9.603100809190138] tot_loss[loss=2.282, over 5543263.32 frames. , ppl: 9.799721610583706], batch size: 70
+2022-12-12 12:34:13,501 INFO [train.py:421] (4/8) Epoch 7, batch 27000, loss[loss=2.97, over 560.00 frames. , ppl: 19.48447524336497] tot_loss[loss=2.283, over 5497059.33 frames. , ppl: 9.810377849232257], batch size: 70
+2022-12-12 12:34:13,502 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:34:14,248 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.741972432318025
+2022-12-12 12:35:56,493 INFO [train.py:421] (4/8) Epoch 7, batch 27200, loss[loss=2.382, over 1190.00 frames. , ppl: 10.827405739862936] tot_loss[loss=2.284, over 5505542.95 frames. , ppl: 9.812691365801323], batch size: 70
+2022-12-12 12:37:39,587 INFO [train.py:421] (4/8) Epoch 7, batch 27400, loss[loss=2.609, over 1190.00 frames. , ppl: 13.588585745216534] tot_loss[loss=2.284, over 5480294.90 frames. , ppl: 9.813935548532314], batch size: 70
+2022-12-12 12:39:19,441 INFO [train.py:421] (4/8) Epoch 7, batch 27600, loss[loss=2.388, over 1400.00 frames. , ppl: 10.887635908173582] tot_loss[loss=2.285, over 5472902.55 frames. , ppl: 9.821925666736048], batch size: 70
+2022-12-12 12:41:01,764 INFO [train.py:421] (4/8) Epoch 7, batch 27800, loss[loss=2.681, over 910.00 frames. , ppl: 14.592941718678714] tot_loss[loss=2.286, over 5442753.14 frames. , ppl: 9.83192924272879], batch size: 70
+2022-12-12 12:42:41,588 INFO [train.py:421] (4/8) Epoch 7, batch 28000, loss[loss=2.6, over 980.00 frames. , ppl: 13.460964456886403] tot_loss[loss=2.285, over 5435893.14 frames. , ppl: 9.828660098098425], batch size: 70
+2022-12-12 12:42:41,588 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:42:42,348 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.745110478587499
+2022-12-12 12:44:19,031 INFO [train.py:421] (4/8) Epoch 7, batch 28200, loss[loss=2.148, over 5600.00 frames. , ppl: 8.563697835947782] tot_loss[loss=2.285, over 5401790.00 frames. , ppl: 9.830419726662875], batch size: 70
+2022-12-12 12:45:59,452 INFO [train.py:421] (4/8) Epoch 7, batch 28400, loss[loss=2.377, over 1610.00 frames. , ppl: 10.768308104658423] tot_loss[loss=2.284, over 5445963.64 frames. , ppl: 9.816958225191456], batch size: 70
+2022-12-12 12:47:41,287 INFO [train.py:421] (4/8) Epoch 7, batch 28600, loss[loss=2.215, over 1890.00 frames. , ppl: 9.162074693676715] tot_loss[loss=2.283, over 5484716.72 frames. , ppl: 9.805684318211709], batch size: 70
+2022-12-12 12:49:21,364 INFO [train.py:421] (4/8) Epoch 7, batch 28800, loss[loss=2.204, over 4480.00 frames. , ppl: 9.059587910755464] tot_loss[loss=2.283, over 5473587.47 frames. , ppl: 9.810619106532043], batch size: 70
+2022-12-12 12:50:59,786 INFO [train.py:421] (4/8) Epoch 7, batch 29000, loss[loss=2.112, over 7210.00 frames. , ppl: 8.265767997925014] tot_loss[loss=2.282, over 5533503.44 frames. , ppl: 9.795499581939684], batch size: 70
+2022-12-12 12:50:59,787 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:51:00,533 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.757074891112701
+2022-12-12 12:52:40,871 INFO [train.py:421] (4/8) Epoch 7, batch 29200, loss[loss=6.215, over 210.00 frames. , ppl: 500.2329749154324] tot_loss[loss=2.283, over 5532987.67 frames. , ppl: 9.803377563798332], batch size: 70
+2022-12-12 12:54:23,367 INFO [train.py:421] (4/8) Epoch 7, batch 29400, loss[loss=2.623, over 770.00 frames. , ppl: 13.77899708485271] tot_loss[loss=2.283, over 5545142.44 frames. , ppl: 9.801197417048916], batch size: 70
+2022-12-12 12:56:02,431 INFO [train.py:421] (4/8) Epoch 7, batch 29600, loss[loss=2.334, over 3080.00 frames. , ppl: 10.321493985627503] tot_loss[loss=2.283, over 5530392.93 frames. , ppl: 9.802525601191867], batch size: 70
+2022-12-12 12:57:39,009 INFO [train.py:421] (4/8) Epoch 7, batch 29800, loss[loss=2.428, over 1190.00 frames. , ppl: 11.336597006319186] tot_loss[loss=2.283, over 5509027.52 frames. , ppl: 9.80428250312761], batch size: 70
+2022-12-12 12:59:14,357 INFO [train.py:421] (4/8) Epoch 7, batch 30000, loss[loss=2.322, over 980.00 frames. , ppl: 10.196130451900016] tot_loss[loss=2.282, over 5516690.04 frames. , ppl: 9.799261363442524], batch size: 70
+2022-12-12 12:59:14,357 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 12:59:15,118 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.278, over 211138.00 frames. , ppl: 9.757277069995297
+2022-12-12 13:00:54,969 INFO [train.py:421] (4/8) Epoch 7, batch 30200, loss[loss=2.239, over 3220.00 frames. , ppl: 9.381913925394121] tot_loss[loss=2.282, over 5505760.00 frames. , ppl: 9.800347412074755], batch size: 70
+2022-12-12 13:02:35,774 INFO [train.py:421] (4/8) Epoch 7, batch 30400, loss[loss=2.332, over 2030.00 frames. , ppl: 10.29459228541268] tot_loss[loss=2.282, over 5513935.50 frames. , ppl: 9.79172026999713], batch size: 70
+2022-12-12 13:04:16,900 INFO [train.py:421] (4/8) Epoch 7, batch 30600, loss[loss=2.416, over 1050.00 frames. , ppl: 11.203651694872494] tot_loss[loss=2.282, over 5495850.02 frames. , ppl: 9.79570110621617], batch size: 70
+2022-12-12 13:05:54,568 INFO [train.py:421] (4/8) Epoch 7, batch 30800, loss[loss=2.418, over 2030.00 frames. , ppl: 11.220512893166505] tot_loss[loss=2.282, over 5492537.13 frames. , ppl: 9.795879946045174], batch size: 70
+2022-12-12 13:07:37,486 INFO [train.py:421] (4/8) Epoch 7, batch 31000, loss[loss=2.401, over 840.00 frames. , ppl: 11.033237747007218] tot_loss[loss=2.281, over 5509914.95 frames. , ppl: 9.790656250260422], batch size: 70
+2022-12-12 13:07:37,487 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:07:38,249 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.74524606034524
+2022-12-12 13:09:21,954 INFO [train.py:421] (4/8) Epoch 7, batch 31200, loss[loss=2.276, over 2520.00 frames. , ppl: 9.7403069491898] tot_loss[loss=2.282, over 5497135.05 frames. , ppl: 9.794770866485385], batch size: 70
+2022-12-12 13:10:57,985 INFO [train.py:421] (4/8) Epoch 7, batch 31400, loss[loss=2.212, over 4620.00 frames. , ppl: 9.13114851428917] tot_loss[loss=2.284, over 5430989.40 frames. , ppl: 9.815898481084886], batch size: 70
+2022-12-12 13:12:37,921 INFO [train.py:421] (4/8) Epoch 7, batch 31600, loss[loss=2.388, over 1330.00 frames. , ppl: 10.894286521829311] tot_loss[loss=2.283, over 5466263.62 frames. , ppl: 9.808268476905084], batch size: 70
+2022-12-12 13:14:17,601 INFO [train.py:421] (4/8) Epoch 7, batch 31800, loss[loss=2.281, over 1960.00 frames. , ppl: 9.790430732896905] tot_loss[loss=2.284, over 5448191.57 frames. , ppl: 9.813791598189779], batch size: 70
+2022-12-12 13:15:58,458 INFO [train.py:421] (4/8) Epoch 7, batch 32000, loss[loss=2.698, over 770.00 frames. , ppl: 14.855454606939764] tot_loss[loss=2.285, over 5394892.62 frames. , ppl: 9.828849651098215], batch size: 70
+2022-12-12 13:15:58,458 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:15:59,205 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.747510843049605
+2022-12-12 13:17:36,138 INFO [train.py:421] (4/8) Epoch 7, batch 32200, loss[loss=2.129, over 5180.00 frames. , ppl: 8.408863411340953] tot_loss[loss=2.285, over 5395290.69 frames. , ppl: 9.826872543335291], batch size: 70
+2022-12-12 13:19:19,513 INFO [train.py:421] (4/8) Epoch 7, batch 32400, loss[loss=2.207, over 8960.00 frames. , ppl: 9.089720549059168] tot_loss[loss=2.285, over 5421446.09 frames. , ppl: 9.825791642436776], batch size: 70
+2022-12-12 13:20:59,435 INFO [train.py:421] (4/8) Epoch 7, batch 32600, loss[loss=2.565, over 980.00 frames. , ppl: 13.004026472840986] tot_loss[loss=2.284, over 5453502.43 frames. , ppl: 9.818277244142239], batch size: 70
+2022-12-12 13:22:38,758 INFO [train.py:421] (4/8) Epoch 7, batch 32800, loss[loss=2.234, over 6650.00 frames. , ppl: 9.336958500480266] tot_loss[loss=2.284, over 5446344.78 frames. , ppl: 9.817521569904587], batch size: 70
+2022-12-12 13:24:19,352 INFO [train.py:421] (4/8) Epoch 7, batch 33000, loss[loss=2.32, over 2170.00 frames. , ppl: 10.177998131801955] tot_loss[loss=2.283, over 5469279.43 frames. , ppl: 9.808867925758346], batch size: 70
+2022-12-12 13:24:19,353 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:24:20,117 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.737587206525212
+2022-12-12 13:26:00,218 INFO [train.py:421] (4/8) Epoch 7, batch 33200, loss[loss=2.385, over 1960.00 frames. , ppl: 10.85606350549362] tot_loss[loss=2.283, over 5457404.12 frames. , ppl: 9.81031424255115], batch size: 70
+2022-12-12 13:27:39,652 INFO [train.py:421] (4/8) Epoch 7, batch 33400, loss[loss=2.21, over 7630.00 frames. , ppl: 9.112073224128636] tot_loss[loss=2.283, over 5470504.20 frames. , ppl: 9.80650575977256], batch size: 70
+2022-12-12 13:29:21,157 INFO [train.py:421] (4/8) Epoch 7, batch 33600, loss[loss=2.236, over 2730.00 frames. , ppl: 9.353193753272912] tot_loss[loss=2.282, over 5492747.69 frames. , ppl: 9.79972382708963], batch size: 70
+2022-12-12 13:31:03,138 INFO [train.py:421] (4/8) Epoch 7, batch 33800, loss[loss=2.244, over 3640.00 frames. , ppl: 9.433447535710782] tot_loss[loss=2.282, over 5521518.52 frames. , ppl: 9.792855722015155], batch size: 70
+2022-12-12 13:32:47,144 INFO [train.py:421] (4/8) Epoch 7, batch 34000, loss[loss=2.409, over 2310.00 frames. , ppl: 11.120446813531812] tot_loss[loss=2.281, over 5570465.85 frames. , ppl: 9.78381892226058], batch size: 70
+2022-12-12 13:32:47,145 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:32:47,895 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.749546711482623
+2022-12-12 13:34:29,525 INFO [train.py:421] (4/8) Epoch 7, batch 34200, loss[loss=2.232, over 3920.00 frames. , ppl: 9.317607978487889] tot_loss[loss=2.282, over 5549501.46 frames. , ppl: 9.798052325655924], batch size: 70
+2022-12-12 13:36:09,713 INFO [train.py:421] (4/8) Epoch 7, batch 34400, loss[loss=2.183, over 4970.00 frames. , ppl: 8.876086151508762] tot_loss[loss=2.282, over 5546971.29 frames. , ppl: 9.791718859426465], batch size: 70
+2022-12-12 13:37:49,610 INFO [train.py:421] (4/8) Epoch 7, batch 34600, loss[loss=2.413, over 1680.00 frames. , ppl: 11.166706217371344] tot_loss[loss=2.282, over 5536307.48 frames. , ppl: 9.796919798725076], batch size: 70
+2022-12-12 13:39:31,007 INFO [train.py:421] (4/8) Epoch 7, batch 34800, loss[loss=2.207, over 4620.00 frames. , ppl: 9.092614216488695] tot_loss[loss=2.282, over 5544424.54 frames. , ppl: 9.797853374897787], batch size: 70
+2022-12-12 13:41:13,651 INFO [train.py:421] (4/8) Epoch 7, batch 35000, loss[loss=2.376, over 2730.00 frames. , ppl: 10.764659048655522] tot_loss[loss=2.281, over 5567951.26 frames. , ppl: 9.79043668916048], batch size: 70
+2022-12-12 13:41:13,652 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:41:14,435 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.743404328984035
+2022-12-12 13:42:53,077 INFO [train.py:421] (4/8) Epoch 7, batch 35200, loss[loss=2.127, over 4480.00 frames. , ppl: 8.391885912657365] tot_loss[loss=2.281, over 5545969.41 frames. , ppl: 9.78999475100402], batch size: 70
+2022-12-12 13:44:36,901 INFO [train.py:421] (4/8) Epoch 7, batch 35400, loss[loss=3.179, over 490.00 frames. , ppl: 24.011265083238506] tot_loss[loss=2.283, over 5505955.36 frames. , ppl: 9.807471001916683], batch size: 70
+2022-12-12 13:46:17,869 INFO [train.py:421] (4/8) Epoch 7, batch 35600, loss[loss=2.476, over 1400.00 frames. , ppl: 11.893614674215446] tot_loss[loss=2.283, over 5513843.04 frames. , ppl: 9.80207304923921], batch size: 70
+2022-12-12 13:48:00,259 INFO [train.py:421] (4/8) Epoch 7, batch 35800, loss[loss=2.171, over 9800.00 frames. , ppl: 8.76341735154874] tot_loss[loss=2.281, over 5563708.15 frames. , ppl: 9.789087886506252], batch size: 70
+2022-12-12 13:49:39,568 INFO [train.py:421] (4/8) Epoch 7, batch 36000, loss[loss=2.366, over 2240.00 frames. , ppl: 10.65064698788081] tot_loss[loss=2.281, over 5575335.22 frames. , ppl: 9.789588108189772], batch size: 70
+2022-12-12 13:49:39,569 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:49:40,314 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730916571608322
+2022-12-12 13:51:17,496 INFO [train.py:421] (4/8) Epoch 7, batch 36200, loss[loss=2.132, over 5740.00 frames. , ppl: 8.43273149149304] tot_loss[loss=2.282, over 5562827.65 frames. , ppl: 9.793351425598933], batch size: 70
+2022-12-12 13:52:55,997 INFO [train.py:421] (4/8) Epoch 7, batch 36400, loss[loss=2.315, over 910.00 frames. , ppl: 10.120864375414099] tot_loss[loss=2.282, over 5539285.62 frames. , ppl: 9.799328103382388], batch size: 70
+2022-12-12 13:54:35,749 INFO [train.py:421] (4/8) Epoch 7, batch 36600, loss[loss=2.287, over 5460.00 frames. , ppl: 9.843328048494229] tot_loss[loss=2.283, over 5513803.85 frames. , ppl: 9.807474638942182], batch size: 70
+2022-12-12 13:56:13,097 INFO [train.py:421] (4/8) Epoch 7, batch 36800, loss[loss=2.157, over 6370.00 frames. , ppl: 8.646522940677109] tot_loss[loss=2.283, over 5496801.38 frames. , ppl: 9.810822864345123], batch size: 70
+2022-12-12 13:57:51,835 INFO [train.py:421] (4/8) Epoch 7, batch 37000, loss[loss=2.33, over 2660.00 frames. , ppl: 10.281324123364751] tot_loss[loss=2.285, over 5443130.37 frames. , ppl: 9.825484522704608], batch size: 70
+2022-12-12 13:57:51,835 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 13:57:52,594 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730660210790171
+2022-12-12 13:59:32,639 INFO [train.py:421] (4/8) Epoch 7, batch 37200, loss[loss=2.241, over 7280.00 frames. , ppl: 9.402703434871002] tot_loss[loss=2.286, over 5429578.15 frames. , ppl: 9.83135194894702], batch size: 70
+2022-12-12 14:01:13,842 INFO [train.py:421] (4/8) Epoch 7, batch 37400, loss[loss=2.222, over 3360.00 frames. , ppl: 9.225723518995308] tot_loss[loss=2.285, over 5426331.06 frames. , ppl: 9.830404972561025], batch size: 70
+2022-12-12 14:03:00,114 INFO [train.py:421] (4/8) Epoch 7, batch 37600, loss[loss=2.345, over 2380.00 frames. , ppl: 10.430719860821466] tot_loss[loss=2.286, over 5415227.78 frames. , ppl: 9.832443657136302], batch size: 70
+2022-12-12 14:04:41,204 INFO [train.py:421] (4/8) Epoch 7, batch 37800, loss[loss=2.31, over 1750.00 frames. , ppl: 10.074265837908003] tot_loss[loss=2.285, over 5413530.87 frames. , ppl: 9.827071609172066], batch size: 70
+2022-12-12 14:06:18,527 INFO [train.py:421] (4/8) Epoch 7, batch 38000, loss[loss=2.308, over 2520.00 frames. , ppl: 10.059172004119437] tot_loss[loss=2.285, over 5409118.87 frames. , ppl: 9.829518399603465], batch size: 70
+2022-12-12 14:06:18,528 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:06:19,290 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.735293029761026
+2022-12-12 14:08:00,975 INFO [train.py:421] (4/8) Epoch 7, batch 38200, loss[loss=3.034, over 560.00 frames. , ppl: 20.776157706856214] tot_loss[loss=2.285, over 5450003.31 frames. , ppl: 9.822699251773168], batch size: 70
+2022-12-12 14:09:41,637 INFO [train.py:421] (4/8) Epoch 7, batch 38400, loss[loss=2.228, over 4830.00 frames. , ppl: 9.27679775341224] tot_loss[loss=2.285, over 5435731.80 frames. , ppl: 9.826833382688354], batch size: 70
+2022-12-12 14:11:19,544 INFO [train.py:421] (4/8) Epoch 7, batch 38600, loss[loss=2.269, over 2730.00 frames. , ppl: 9.66631668571755] tot_loss[loss=2.285, over 5435793.90 frames. , ppl: 9.825879828139794], batch size: 70
+2022-12-12 14:12:58,079 INFO [train.py:421] (4/8) Epoch 7, batch 38800, loss[loss=2.428, over 1050.00 frames. , ppl: 11.33197468712812] tot_loss[loss=2.284, over 5456829.93 frames. , ppl: 9.81460793200896], batch size: 70
+2022-12-12 14:14:39,697 INFO [train.py:421] (4/8) Epoch 7, batch 39000, loss[loss=2.522, over 770.00 frames. , ppl: 12.45758299929893] tot_loss[loss=2.284, over 5466955.05 frames. , ppl: 9.81176307061865], batch size: 70
+2022-12-12 14:14:39,698 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:14:40,458 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.733431570068024
+2022-12-12 14:16:18,498 INFO [train.py:421] (4/8) Epoch 7, batch 39200, loss[loss=2.726, over 700.00 frames. , ppl: 15.275320556199834] tot_loss[loss=2.284, over 5436290.50 frames. , ppl: 9.818038015815418], batch size: 70
+2022-12-12 14:17:57,353 INFO [train.py:421] (4/8) Epoch 7, batch 39400, loss[loss=2.212, over 8820.00 frames. , ppl: 9.132297373385196] tot_loss[loss=2.284, over 5433606.03 frames. , ppl: 9.818393140841149], batch size: 70
+2022-12-12 14:19:39,322 INFO [train.py:421] (4/8) Epoch 7, batch 39600, loss[loss=2.374, over 1960.00 frames. , ppl: 10.735099651272549] tot_loss[loss=2.286, over 5401049.74 frames. , ppl: 9.835249444414478], batch size: 70
+2022-12-12 14:21:22,520 INFO [train.py:421] (4/8) Epoch 7, batch 39800, loss[loss=2.813, over 700.00 frames. , ppl: 16.651521097743146] tot_loss[loss=2.285, over 5445143.31 frames. , ppl: 9.825563614216229], batch size: 70
+2022-12-12 14:23:03,816 INFO [train.py:421] (4/8) Epoch 7, batch 40000, loss[loss=2.293, over 3220.00 frames. , ppl: 9.906433378283886] tot_loss[loss=2.286, over 5420102.47 frames. , ppl: 9.835903277410905], batch size: 70
+2022-12-12 14:23:03,816 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:23:04,577 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.732923044306935
+2022-12-12 14:24:40,622 INFO [train.py:421] (4/8) Epoch 7, batch 40200, loss[loss=2.349, over 1680.00 frames. , ppl: 10.471380734122837] tot_loss[loss=2.287, over 5391949.62 frames. , ppl: 9.840976314694833], batch size: 70
+2022-12-12 14:26:24,155 INFO [train.py:421] (4/8) Epoch 7, batch 40400, loss[loss=2.811, over 630.00 frames. , ppl: 16.62892523765851] tot_loss[loss=2.287, over 5376494.05 frames. , ppl: 9.84408539611387], batch size: 70
+2022-12-12 14:28:05,238 INFO [train.py:421] (4/8) Epoch 7, batch 40600, loss[loss=2.318, over 2100.00 frames. , ppl: 10.156926885320106] tot_loss[loss=2.285, over 5416725.37 frames. , ppl: 9.829845087857588], batch size: 70
+2022-12-12 14:29:48,173 INFO [train.py:421] (4/8) Epoch 7, batch 40800, loss[loss=2.368, over 2310.00 frames. , ppl: 10.674674352516327] tot_loss[loss=2.286, over 5390890.34 frames. , ppl: 9.83244368313401], batch size: 70
+2022-12-12 14:31:28,831 INFO [train.py:421] (4/8) Epoch 7, batch 41000, loss[loss=2.459, over 1470.00 frames. , ppl: 11.695260490702806] tot_loss[loss=2.284, over 5446709.54 frames. , ppl: 9.816688481114332], batch size: 70
+2022-12-12 14:31:28,832 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:31:29,592 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.72567260083667
+2022-12-12 14:33:15,370 INFO [train.py:421] (4/8) Epoch 7, batch 41200, loss[loss=2.12, over 2870.00 frames. , ppl: 8.334902669380726] tot_loss[loss=2.283, over 5473270.94 frames. , ppl: 9.80823146188833], batch size: 70
+2022-12-12 14:34:58,546 INFO [train.py:421] (4/8) Epoch 7, batch 41400, loss[loss=2.339, over 1610.00 frames. , ppl: 10.367983236997537] tot_loss[loss=2.283, over 5483588.21 frames. , ppl: 9.801861312066212], batch size: 70
+2022-12-12 14:36:43,130 INFO [train.py:421] (4/8) Epoch 7, batch 41600, loss[loss=2.152, over 6370.00 frames. , ppl: 8.598012502277944] tot_loss[loss=2.283, over 5467322.19 frames. , ppl: 9.80501552597066], batch size: 70
+2022-12-12 14:38:19,999 INFO [train.py:421] (4/8) Epoch 7, batch 41800, loss[loss=2.144, over 6720.00 frames. , ppl: 8.531868768224657] tot_loss[loss=2.282, over 5510992.46 frames. , ppl: 9.794104028804583], batch size: 70
+2022-12-12 14:39:58,260 INFO [train.py:421] (4/8) Epoch 7, batch 42000, loss[loss=2.304, over 2660.00 frames. , ppl: 10.010698785892233] tot_loss[loss=2.283, over 5473170.72 frames. , ppl: 9.802348653008387], batch size: 70
+2022-12-12 14:39:58,261 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:39:59,022 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.732604689011115
+2022-12-12 14:41:41,114 INFO [train.py:421] (4/8) Epoch 7, batch 42200, loss[loss=2.354, over 1050.00 frames. , ppl: 10.522693663426836] tot_loss[loss=2.282, over 5471206.58 frames. , ppl: 9.80069832458638], batch size: 70
+2022-12-12 14:43:22,130 INFO [train.py:421] (4/8) Epoch 7, batch 42400, loss[loss=2.334, over 1680.00 frames. , ppl: 10.318845434833433] tot_loss[loss=2.283, over 5438316.80 frames. , ppl: 9.806993509764954], batch size: 70
+2022-12-12 14:45:04,978 INFO [train.py:421] (4/8) Epoch 7, batch 42600, loss[loss=2.731, over 840.00 frames. , ppl: 15.353970495966221] tot_loss[loss=2.282, over 5479889.40 frames. , ppl: 9.791833169326553], batch size: 70
+2022-12-12 14:46:47,201 INFO [train.py:421] (4/8) Epoch 7, batch 42800, loss[loss=2.191, over 7420.00 frames. , ppl: 8.940591091714337] tot_loss[loss=2.281, over 5497573.91 frames. , ppl: 9.786893599735823], batch size: 70
+2022-12-12 14:48:26,743 INFO [train.py:421] (4/8) Epoch 7, batch 43000, loss[loss=2.477, over 1680.00 frames. , ppl: 11.8997403233363] tot_loss[loss=2.281, over 5488628.22 frames. , ppl: 9.790509703536982], batch size: 70
+2022-12-12 14:48:26,743 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:48:27,491 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.722768179823841
+2022-12-12 14:50:02,821 INFO [train.py:421] (4/8) Epoch 7, batch 43200, loss[loss=2.199, over 3850.00 frames. , ppl: 9.01629579155962] tot_loss[loss=2.283, over 5445171.98 frames. , ppl: 9.807785908334857], batch size: 70
+2022-12-12 14:51:39,671 INFO [train.py:421] (4/8) Epoch 7, batch 43400, loss[loss=2.157, over 5390.00 frames. , ppl: 8.644819391735084] tot_loss[loss=2.282, over 5480655.60 frames. , ppl: 9.795615836686666], batch size: 70
+2022-12-12 14:53:19,379 INFO [train.py:421] (4/8) Epoch 7, batch 43600, loss[loss=2.371, over 1540.00 frames. , ppl: 10.710226319345324] tot_loss[loss=2.282, over 5494084.92 frames. , ppl: 9.800861786788706], batch size: 70
+2022-12-12 14:54:58,929 INFO [train.py:421] (4/8) Epoch 7, batch 43800, loss[loss=2.372, over 1540.00 frames. , ppl: 10.717523080825083] tot_loss[loss=2.282, over 5494413.67 frames. , ppl: 9.800104860413766], batch size: 70
+2022-12-12 14:56:40,138 INFO [train.py:421] (4/8) Epoch 7, batch 44000, loss[loss=2.537, over 770.00 frames. , ppl: 12.643677045695101] tot_loss[loss=2.283, over 5460275.83 frames. , ppl: 9.808473351016165], batch size: 70
+2022-12-12 14:56:40,139 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 14:56:40,926 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.738247314799137
+2022-12-12 14:58:24,199 INFO [train.py:421] (4/8) Epoch 7, batch 44200, loss[loss=2.265, over 2030.00 frames. , ppl: 9.634995026876341] tot_loss[loss=2.283, over 5485709.35 frames. , ppl: 9.808101245250407], batch size: 70
+2022-12-12 15:00:05,404 INFO [train.py:421] (4/8) Epoch 7, batch 44400, loss[loss=2.315, over 2170.00 frames. , ppl: 10.128786662347173] tot_loss[loss=2.283, over 5465647.99 frames. , ppl: 9.810092502039728], batch size: 70
+2022-12-12 15:01:46,541 INFO [train.py:421] (4/8) Epoch 7, batch 44600, loss[loss=3.331, over 490.00 frames. , ppl: 27.955855982859386] tot_loss[loss=2.284, over 5437546.50 frames. , ppl: 9.817569522042081], batch size: 70
+2022-12-12 15:03:25,945 INFO [train.py:421] (4/8) Epoch 7, batch 44800, loss[loss=2.223, over 5670.00 frames. , ppl: 9.232594724512339] tot_loss[loss=2.285, over 5411806.09 frames. , ppl: 9.826773901640673], batch size: 70
+2022-12-12 15:05:07,366 INFO [train.py:421] (4/8) Epoch 7, batch 45000, loss[loss=2.238, over 5320.00 frames. , ppl: 9.370764838438866] tot_loss[loss=2.283, over 5466197.29 frames. , ppl: 9.809155281573775], batch size: 70
+2022-12-12 15:05:07,367 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:05:08,112 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.739091972737263
+2022-12-12 15:06:49,436 INFO [train.py:421] (4/8) Epoch 7, batch 45200, loss[loss=2.278, over 3570.00 frames. , ppl: 9.752631516010661] tot_loss[loss=2.283, over 5494019.74 frames. , ppl: 9.80745893871796], batch size: 70
+2022-12-12 15:08:26,838 INFO [train.py:421] (4/8) Epoch 7, batch 45400, loss[loss=2.338, over 1890.00 frames. , ppl: 10.3614628608319] tot_loss[loss=2.284, over 5474274.99 frames. , ppl: 9.814301143840781], batch size: 70
+2022-12-12 15:10:04,369 INFO [train.py:421] (4/8) Epoch 7, batch 45600, loss[loss=2.355, over 1890.00 frames. , ppl: 10.534857984340505] tot_loss[loss=2.284, over 5441218.79 frames. , ppl: 9.818841558649813], batch size: 70
+2022-12-12 15:11:43,506 INFO [train.py:421] (4/8) Epoch 7, batch 45800, loss[loss=2.219, over 6930.00 frames. , ppl: 9.199771740110176] tot_loss[loss=2.284, over 5472420.81 frames. , ppl: 9.811679690798433], batch size: 70
+2022-12-12 15:13:23,835 INFO [train.py:421] (4/8) Epoch 7, batch 46000, loss[loss=2.246, over 6160.00 frames. , ppl: 9.448574441109736] tot_loss[loss=2.284, over 5462152.32 frames. , ppl: 9.820672813221504], batch size: 70
+2022-12-12 15:13:23,836 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:13:24,592 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.725171677432357
+2022-12-12 15:15:04,841 INFO [train.py:421] (4/8) Epoch 7, batch 46200, loss[loss=2.329, over 1890.00 frames. , ppl: 10.2684648732086] tot_loss[loss=2.284, over 5470643.41 frames. , ppl: 9.81675106997631], batch size: 70
+2022-12-12 15:16:41,451 INFO [train.py:421] (4/8) Epoch 7, batch 46400, loss[loss=2.373, over 2310.00 frames. , ppl: 10.730044678375746] tot_loss[loss=2.286, over 5409555.61 frames. , ppl: 9.833886034145493], batch size: 70
+2022-12-12 15:18:27,480 INFO [train.py:421] (4/8) Epoch 7, batch 46600, loss[loss=2.297, over 2240.00 frames. , ppl: 9.939848028363519] tot_loss[loss=2.284, over 5461403.27 frames. , ppl: 9.820600783010347], batch size: 70
+2022-12-12 15:20:06,538 INFO [train.py:421] (4/8) Epoch 7, batch 46800, loss[loss=2.158, over 12950.00 frames. , ppl: 8.655558949186167] tot_loss[loss=2.284, over 5456914.45 frames. , ppl: 9.817457911803828], batch size: 70
+2022-12-12 15:21:46,508 INFO [train.py:421] (4/8) Epoch 7, batch 47000, loss[loss=2.292, over 10430.00 frames. , ppl: 9.896768462190552] tot_loss[loss=2.284, over 5475648.37 frames. , ppl: 9.816425784782295], batch size: 70
+2022-12-12 15:21:46,509 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:21:47,256 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.277, over 211138.00 frames. , ppl: 9.746437530099742
+2022-12-12 15:23:25,432 INFO [train.py:421] (4/8) Epoch 7, batch 47200, loss[loss=2.224, over 3640.00 frames. , ppl: 9.240419988809789] tot_loss[loss=2.284, over 5486164.78 frames. , ppl: 9.812692320906956], batch size: 70
+2022-12-12 15:25:07,547 INFO [train.py:421] (4/8) Epoch 7, batch 47400, loss[loss=2.254, over 3990.00 frames. , ppl: 9.523713561743257] tot_loss[loss=2.283, over 5472013.67 frames. , ppl: 9.809568383870149], batch size: 70
+2022-12-12 15:26:45,395 INFO [train.py:421] (4/8) Epoch 7, batch 47600, loss[loss=2.392, over 2380.00 frames. , ppl: 10.934911850063324] tot_loss[loss=2.286, over 5425354.11 frames. , ppl: 9.831529079519433], batch size: 70
+2022-12-12 15:28:25,599 INFO [train.py:421] (4/8) Epoch 7, batch 47800, loss[loss=2.269, over 3920.00 frames. , ppl: 9.667010491163618] tot_loss[loss=2.285, over 5465743.00 frames. , ppl: 9.822275456452866], batch size: 70
+2022-12-12 15:30:08,467 INFO [train.py:421] (4/8) Epoch 7, batch 48000, loss[loss=2.279, over 2450.00 frames. , ppl: 9.767585253074058] tot_loss[loss=2.285, over 5436208.96 frames. , ppl: 9.827403984333996], batch size: 70
+2022-12-12 15:30:08,467 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:30:09,214 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730277122379839
+2022-12-12 15:31:50,593 INFO [train.py:421] (4/8) Epoch 7, batch 48200, loss[loss=2.367, over 1470.00 frames. , ppl: 10.669952788676424] tot_loss[loss=2.285, over 5460690.30 frames. , ppl: 9.826927303715037], batch size: 70
+2022-12-12 15:33:31,840 INFO [train.py:421] (4/8) Epoch 7, batch 48400, loss[loss=2.224, over 4550.00 frames. , ppl: 9.247992958119994] tot_loss[loss=2.284, over 5483740.20 frames. , ppl: 9.818256407560218], batch size: 70
+2022-12-12 15:35:15,561 INFO [train.py:421] (4/8) Epoch 7, batch 48600, loss[loss=2.335, over 1330.00 frames. , ppl: 10.33175297324686] tot_loss[loss=2.283, over 5520345.16 frames. , ppl: 9.80527054073112], batch size: 70
+2022-12-12 15:37:00,084 INFO [train.py:421] (4/8) Epoch 7, batch 48800, loss[loss=2.609, over 1120.00 frames. , ppl: 13.586333185547936] tot_loss[loss=2.282, over 5536582.31 frames. , ppl: 9.800825572682498], batch size: 70
+2022-12-12 15:38:39,938 INFO [train.py:421] (4/8) Epoch 7, batch 49000, loss[loss=2.396, over 1470.00 frames. , ppl: 10.982648651930889] tot_loss[loss=2.283, over 5518027.87 frames. , ppl: 9.807171776194984], batch size: 70
+2022-12-12 15:38:39,938 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:38:40,684 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71475889319782
+2022-12-12 15:40:19,125 INFO [train.py:421] (4/8) Epoch 7, batch 49200, loss[loss=2.337, over 3640.00 frames. , ppl: 10.348686468077279] tot_loss[loss=2.284, over 5488795.04 frames. , ppl: 9.820231343993715], batch size: 70
+2022-12-12 15:42:00,529 INFO [train.py:421] (4/8) Epoch 7, batch 49400, loss[loss=2.167, over 4340.00 frames. , ppl: 8.731251424364928] tot_loss[loss=2.284, over 5520195.85 frames. , ppl: 9.811435214799545], batch size: 70
+2022-12-12 15:43:37,682 INFO [train.py:421] (4/8) Epoch 7, batch 49600, loss[loss=2.407, over 1610.00 frames. , ppl: 11.096442385618435] tot_loss[loss=2.284, over 5514493.58 frames. , ppl: 9.81156953968905], batch size: 70
+2022-12-12 15:45:24,853 INFO [train.py:421] (4/8) Epoch 7, batch 49800, loss[loss=2.516, over 1190.00 frames. , ppl: 12.382116334412945] tot_loss[loss=2.283, over 5559944.22 frames. , ppl: 9.802517060991498], batch size: 70
+2022-12-12 15:47:06,649 INFO [train.py:421] (4/8) Epoch 7, batch 50000, loss[loss=2.37, over 2310.00 frames. , ppl: 10.69770975911852] tot_loss[loss=2.282, over 5580450.27 frames. , ppl: 9.793364993403346], batch size: 70
+2022-12-12 15:47:06,650 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:47:07,406 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714458385768214
+2022-12-12 15:48:44,227 INFO [train.py:421] (4/8) Epoch 7, batch 50200, loss[loss=2.35, over 1890.00 frames. , ppl: 10.485778315609766] tot_loss[loss=2.28, over 5588763.35 frames. , ppl: 9.781064865435585], batch size: 70
+2022-12-12 15:50:22,977 INFO [train.py:421] (4/8) Epoch 7, batch 50400, loss[loss=2.34, over 1820.00 frames. , ppl: 10.38014929922187] tot_loss[loss=2.28, over 5627271.66 frames. , ppl: 9.77909175337856], batch size: 70
+2022-12-12 15:52:02,485 INFO [train.py:421] (4/8) Epoch 7, batch 50600, loss[loss=2.24, over 3920.00 frames. , ppl: 9.393305078407673] tot_loss[loss=2.28, over 5664450.64 frames. , ppl: 9.776899806943886], batch size: 70
+2022-12-12 15:53:41,810 INFO [train.py:421] (4/8) Epoch 7, batch 50800, loss[loss=2.323, over 2660.00 frames. , ppl: 10.21055218486924] tot_loss[loss=2.28, over 5680015.39 frames. , ppl: 9.775704753256708], batch size: 70
+2022-12-12 15:55:22,387 INFO [train.py:421] (4/8) Epoch 7, batch 51000, loss[loss=2.267, over 3640.00 frames. , ppl: 9.653541691081644] tot_loss[loss=2.28, over 5681526.84 frames. , ppl: 9.77361004208444], batch size: 70
+2022-12-12 15:55:22,387 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 15:55:23,149 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.712544846014454
+2022-12-12 15:57:03,533 INFO [train.py:421] (4/8) Epoch 7, batch 51200, loss[loss=2.165, over 7840.00 frames. , ppl: 8.711161113579154] tot_loss[loss=2.279, over 5688481.99 frames. , ppl: 9.763704917477394], batch size: 70
+2022-12-12 15:58:38,215 INFO [train.py:421] (4/8) Epoch 7, batch 51400, loss[loss=2.296, over 2310.00 frames. , ppl: 9.937197570239197] tot_loss[loss=2.28, over 5648747.92 frames. , ppl: 9.773852696809355], batch size: 70
+2022-12-12 16:00:15,721 INFO [train.py:421] (4/8) Epoch 7, batch 51600, loss[loss=2.838, over 560.00 frames. , ppl: 17.089002804682085] tot_loss[loss=2.28, over 5642375.46 frames. , ppl: 9.777341942206226], batch size: 70
+2022-12-12 16:01:57,240 INFO [train.py:421] (4/8) Epoch 7, batch 51800, loss[loss=2.341, over 3710.00 frames. , ppl: 10.387108929470065] tot_loss[loss=2.281, over 5615193.91 frames. , ppl: 9.782256367439683], batch size: 70
+2022-12-12 16:03:36,215 INFO [train.py:421] (4/8) Epoch 7, batch 52000, loss[loss=2.349, over 2240.00 frames. , ppl: 10.47924489071105] tot_loss[loss=2.282, over 5580523.58 frames. , ppl: 9.791655437018452], batch size: 70
+2022-12-12 16:03:36,215 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:03:36,974 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714971698339436
+2022-12-12 16:05:14,234 INFO [train.py:421] (4/8) Epoch 7, batch 52200, loss[loss=3.211, over 490.00 frames. , ppl: 24.807042407777367] tot_loss[loss=2.283, over 5512698.08 frames. , ppl: 9.805892653472863], batch size: 70
+2022-12-12 16:06:54,094 INFO [train.py:421] (4/8) Epoch 7, batch 52400, loss[loss=2.259, over 2660.00 frames. , ppl: 9.57238777232647] tot_loss[loss=2.285, over 5446130.60 frames. , ppl: 9.821625791450291], batch size: 70
+2022-12-12 16:08:37,710 INFO [train.py:421] (4/8) Epoch 7, batch 52600, loss[loss=2.146, over 10150.00 frames. , ppl: 8.549393434771904] tot_loss[loss=2.284, over 5420123.70 frames. , ppl: 9.81965775131349], batch size: 70
+2022-12-12 16:10:20,600 INFO [train.py:421] (4/8) Epoch 7, batch 52800, loss[loss=2.241, over 2870.00 frames. , ppl: 9.40720292254974] tot_loss[loss=2.284, over 5436916.11 frames. , ppl: 9.81744500706336], batch size: 70
+2022-12-12 16:12:02,790 INFO [train.py:421] (4/8) Epoch 7, batch 53000, loss[loss=2.363, over 3150.00 frames. , ppl: 10.618307936421983] tot_loss[loss=2.285, over 5421487.61 frames. , ppl: 9.822582880746209], batch size: 70
+2022-12-12 16:12:02,791 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:12:03,557 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.710289624670212
+2022-12-12 16:13:44,886 INFO [train.py:421] (4/8) Epoch 7, batch 53200, loss[loss=2.355, over 3220.00 frames. , ppl: 10.543277607416805] tot_loss[loss=2.286, over 5396957.66 frames. , ppl: 9.83514019558566], batch size: 70
+2022-12-12 16:15:28,443 INFO [train.py:421] (4/8) Epoch 7, batch 53400, loss[loss=2.723, over 910.00 frames. , ppl: 15.220767996760738] tot_loss[loss=2.286, over 5401531.27 frames. , ppl: 9.840341782325522], batch size: 70
+2022-12-12 16:17:11,372 INFO [train.py:421] (4/8) Epoch 7, batch 53600, loss[loss=2.268, over 3220.00 frames. , ppl: 9.655281143907459] tot_loss[loss=2.287, over 5381968.04 frames. , ppl: 9.841160018443144], batch size: 70
+2022-12-12 16:18:49,882 INFO [train.py:421] (4/8) Epoch 7, batch 53800, loss[loss=2.157, over 6020.00 frames. , ppl: 8.649483240951978] tot_loss[loss=2.284, over 5424580.89 frames. , ppl: 9.817332193019213], batch size: 70
+2022-12-12 16:20:30,724 INFO [train.py:421] (4/8) Epoch 7, batch 54000, loss[loss=2.245, over 2240.00 frames. , ppl: 9.440443543224852] tot_loss[loss=2.284, over 5425275.97 frames. , ppl: 9.82040063236461], batch size: 70
+2022-12-12 16:20:30,725 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:20:31,473 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.730233917866439
+2022-12-12 16:22:14,273 INFO [train.py:421] (4/8) Epoch 7, batch 54200, loss[loss=2.392, over 2170.00 frames. , ppl: 10.936738716181575] tot_loss[loss=2.285, over 5417606.86 frames. , ppl: 9.823630694224349], batch size: 70
+2022-12-12 16:23:57,484 INFO [train.py:421] (4/8) Epoch 7, batch 54400, loss[loss=2.159, over 4760.00 frames. , ppl: 8.659764326780536] tot_loss[loss=2.283, over 5485257.80 frames. , ppl: 9.805645840569355], batch size: 70
+2022-12-12 16:25:40,588 INFO [train.py:421] (4/8) Epoch 7, batch 54600, loss[loss=2.445, over 1190.00 frames. , ppl: 11.534356966480553] tot_loss[loss=2.282, over 5515010.75 frames. , ppl: 9.79706447279174], batch size: 70
+2022-12-12 16:27:23,732 INFO [train.py:421] (4/8) Epoch 7, batch 54800, loss[loss=2.29, over 3780.00 frames. , ppl: 9.873119361193954] tot_loss[loss=2.282, over 5540216.66 frames. , ppl: 9.792093445489721], batch size: 70
+2022-12-12 16:29:03,510 INFO [train.py:421] (4/8) Epoch 7, batch 55000, loss[loss=2.332, over 2520.00 frames. , ppl: 10.298620313686152] tot_loss[loss=2.282, over 5512485.59 frames. , ppl: 9.793709208429366], batch size: 70
+2022-12-12 16:29:03,511 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:29:04,267 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.276, over 211138.00 frames. , ppl: 9.738142098037239
+2022-12-12 16:30:45,226 INFO [train.py:421] (4/8) Epoch 7, batch 55200, loss[loss=2.61, over 840.00 frames. , ppl: 13.594228104131384] tot_loss[loss=2.283, over 5471759.25 frames. , ppl: 9.807276382294035], batch size: 70
+2022-12-12 16:32:27,263 INFO [train.py:421] (4/8) Epoch 7, batch 55400, loss[loss=2.146, over 7630.00 frames. , ppl: 8.552424657267816] tot_loss[loss=2.284, over 5445619.04 frames. , ppl: 9.81431147160649], batch size: 70
+2022-12-12 16:34:09,958 INFO [train.py:421] (4/8) Epoch 7, batch 55600, loss[loss=2.196, over 7840.00 frames. , ppl: 8.98843763853189] tot_loss[loss=2.284, over 5430292.63 frames. , ppl: 9.817627292125971], batch size: 70
+2022-12-12 16:35:52,341 INFO [train.py:421] (4/8) Epoch 7, batch 55800, loss[loss=2.278, over 3780.00 frames. , ppl: 9.757587328242263] tot_loss[loss=2.284, over 5444809.37 frames. , ppl: 9.81531675947324], batch size: 70
+2022-12-12 16:37:34,622 INFO [train.py:421] (4/8) Epoch 7, batch 56000, loss[loss=2.508, over 1260.00 frames. , ppl: 12.28330597178825] tot_loss[loss=2.284, over 5468813.21 frames. , ppl: 9.81234698008054], batch size: 70
+2022-12-12 16:37:34,623 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:37:35,393 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.72342008782373
+2022-12-12 16:39:17,250 INFO [train.py:421] (4/8) Epoch 7, batch 56200, loss[loss=2.353, over 3500.00 frames. , ppl: 10.517710461408411] tot_loss[loss=2.283, over 5480952.54 frames. , ppl: 9.802809941212468], batch size: 70
+2022-12-12 16:40:58,439 INFO [train.py:421] (4/8) Epoch 7, batch 56400, loss[loss=2.403, over 2520.00 frames. , ppl: 11.052772308040256] tot_loss[loss=2.282, over 5503333.24 frames. , ppl: 9.795901689738189], batch size: 70
+2022-12-12 16:42:37,536 INFO [train.py:421] (4/8) Epoch 7, batch 56600, loss[loss=2.206, over 3500.00 frames. , ppl: 9.082237577711298] tot_loss[loss=2.282, over 5461273.16 frames. , ppl: 9.798540586286315], batch size: 70
+2022-12-12 16:44:14,650 INFO [train.py:421] (4/8) Epoch 7, batch 56800, loss[loss=2.307, over 4200.00 frames. , ppl: 10.04459792649784] tot_loss[loss=2.283, over 5450338.14 frames. , ppl: 9.80143426518573], batch size: 70
+2022-12-12 16:45:52,794 INFO [train.py:421] (4/8) Epoch 7, batch 57000, loss[loss=2.387, over 1470.00 frames. , ppl: 10.885228230331933] tot_loss[loss=2.283, over 5419091.19 frames. , ppl: 9.807225342704541], batch size: 70
+2022-12-12 16:45:52,795 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:45:53,554 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.720317799677936
+2022-12-12 16:47:33,760 INFO [train.py:421] (4/8) Epoch 7, batch 57200, loss[loss=2.314, over 3430.00 frames. , ppl: 10.116402973399014] tot_loss[loss=2.284, over 5398623.45 frames. , ppl: 9.819781680362881], batch size: 70
+2022-12-12 16:49:17,346 INFO [train.py:421] (4/8) Epoch 7, batch 57400, loss[loss=2.669, over 700.00 frames. , ppl: 14.41861110957854] tot_loss[loss=2.284, over 5407406.13 frames. , ppl: 9.8176422112151], batch size: 70
+2022-12-12 16:50:58,688 INFO [train.py:421] (4/8) Epoch 7, batch 57600, loss[loss=2.829, over 630.00 frames. , ppl: 16.92657608431769] tot_loss[loss=2.284, over 5436355.39 frames. , ppl: 9.814211188562084], batch size: 70
+2022-12-12 16:52:38,249 INFO [train.py:421] (4/8) Epoch 7, batch 57800, loss[loss=2.431, over 1400.00 frames. , ppl: 11.374614827642908] tot_loss[loss=2.284, over 5407061.52 frames. , ppl: 9.816371055851334], batch size: 70
+2022-12-12 16:54:19,003 INFO [train.py:421] (4/8) Epoch 7, batch 58000, loss[loss=2.4, over 1050.00 frames. , ppl: 11.018507487946296] tot_loss[loss=2.286, over 5375434.30 frames. , ppl: 9.837917402678658], batch size: 70
+2022-12-12 16:54:19,004 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 16:54:19,765 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714061557664111
+2022-12-12 16:56:00,042 INFO [train.py:421] (4/8) Epoch 7, batch 58200, loss[loss=2.269, over 3780.00 frames. , ppl: 9.67121107921451] tot_loss[loss=2.285, over 5419999.33 frames. , ppl: 9.822317954330044], batch size: 70
+2022-12-12 16:57:41,654 INFO [train.py:421] (4/8) Epoch 7, batch 58400, loss[loss=2.156, over 11480.00 frames. , ppl: 8.633592162870084] tot_loss[loss=2.284, over 5434997.39 frames. , ppl: 9.813815901170763], batch size: 70
+2022-12-12 16:59:24,636 INFO [train.py:421] (4/8) Epoch 7, batch 58600, loss[loss=2.381, over 1610.00 frames. , ppl: 10.812965768389253] tot_loss[loss=2.284, over 5452393.49 frames. , ppl: 9.812345905560925], batch size: 70
+2022-12-12 17:01:06,528 INFO [train.py:421] (4/8) Epoch 7, batch 58800, loss[loss=2.368, over 3080.00 frames. , ppl: 10.677369377341408] tot_loss[loss=2.284, over 5468977.65 frames. , ppl: 9.815757659718654], batch size: 70
+2022-12-12 17:02:51,795 INFO [train.py:421] (4/8) Epoch 7, batch 59000, loss[loss=2.34, over 2660.00 frames. , ppl: 10.38525211043209] tot_loss[loss=2.283, over 5492344.55 frames. , ppl: 9.807263812878992], batch size: 70
+2022-12-12 17:02:51,795 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:02:52,548 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.716621096251133
+2022-12-12 17:04:37,258 INFO [train.py:421] (4/8) Epoch 7, batch 59200, loss[loss=4.068, over 350.00 frames. , ppl: 58.43948468365342] tot_loss[loss=2.283, over 5515000.45 frames. , ppl: 9.801996126637023], batch size: 70
+2022-12-12 17:06:15,593 INFO [train.py:421] (4/8) Epoch 7, batch 59400, loss[loss=2.148, over 2870.00 frames. , ppl: 8.565349903594884] tot_loss[loss=2.282, over 5518950.12 frames. , ppl: 9.801041203554101], batch size: 70
+2022-12-12 17:07:59,263 INFO [train.py:421] (4/8) Epoch 7, batch 59600, loss[loss=2.314, over 1540.00 frames. , ppl: 10.110130234720843] tot_loss[loss=2.283, over 5497936.15 frames. , ppl: 9.806095554045832], batch size: 70
+2022-12-12 17:09:44,695 INFO [train.py:421] (4/8) Epoch 7, batch 59800, loss[loss=2.426, over 1470.00 frames. , ppl: 11.313672363266914] tot_loss[loss=2.281, over 5591314.34 frames. , ppl: 9.787376727796463], batch size: 70
+2022-12-12 17:11:28,836 INFO [train.py:421] (4/8) Epoch 7, batch 60000, loss[loss=2.686, over 1050.00 frames. , ppl: 14.672079523209169] tot_loss[loss=2.281, over 5573849.45 frames. , ppl: 9.790805630328208], batch size: 70
+2022-12-12 17:11:28,836 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:11:29,585 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71337577339601
+2022-12-12 17:13:08,942 INFO [train.py:421] (4/8) Epoch 7, batch 60200, loss[loss=2.281, over 3780.00 frames. , ppl: 9.785811411359596] tot_loss[loss=2.283, over 5533008.01 frames. , ppl: 9.807131703452498], batch size: 70
+2022-12-12 17:14:49,948 INFO [train.py:421] (4/8) Epoch 7, batch 60400, loss[loss=2.14, over 5110.00 frames. , ppl: 8.50003767075534] tot_loss[loss=2.283, over 5541012.13 frames. , ppl: 9.808231105473038], batch size: 70
+2022-12-12 17:16:33,754 INFO [train.py:421] (4/8) Epoch 7, batch 60600, loss[loss=2.371, over 1190.00 frames. , ppl: 10.70572696984717] tot_loss[loss=2.283, over 5561274.68 frames. , ppl: 9.803327559758722], batch size: 70
+2022-12-12 17:18:13,205 INFO [train.py:421] (4/8) Epoch 7, batch 60800, loss[loss=2.349, over 1400.00 frames. , ppl: 10.480189717343741] tot_loss[loss=2.282, over 5575144.73 frames. , ppl: 9.79507863047382], batch size: 70
+2022-12-12 17:19:56,253 INFO [train.py:421] (4/8) Epoch 7, batch 61000, loss[loss=2.322, over 1750.00 frames. , ppl: 10.192864519015203] tot_loss[loss=2.283, over 5551436.58 frames. , ppl: 9.801449830395763], batch size: 70
+2022-12-12 17:19:56,254 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:19:57,004 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.70608961125915
+2022-12-12 17:21:39,553 INFO [train.py:421] (4/8) Epoch 7, batch 61200, loss[loss=2.336, over 2100.00 frames. , ppl: 10.341316979462793] tot_loss[loss=2.284, over 5516991.16 frames. , ppl: 9.812172013062103], batch size: 70
+2022-12-12 17:23:24,470 INFO [train.py:421] (4/8) Epoch 7, batch 61400, loss[loss=2.312, over 2170.00 frames. , ppl: 10.094044177477475] tot_loss[loss=2.284, over 5526830.48 frames. , ppl: 9.813541170948804], batch size: 70
+2022-12-12 17:25:04,986 INFO [train.py:421] (4/8) Epoch 7, batch 61600, loss[loss=2.344, over 1750.00 frames. , ppl: 10.420588182019818] tot_loss[loss=2.283, over 5569166.15 frames. , ppl: 9.801384164981492], batch size: 70
+2022-12-12 17:26:46,420 INFO [train.py:421] (4/8) Epoch 7, batch 61800, loss[loss=2.519, over 1400.00 frames. , ppl: 12.410529563754839] tot_loss[loss=2.282, over 5605281.24 frames. , ppl: 9.795132459796609], batch size: 70
+2022-12-12 17:28:30,621 INFO [train.py:421] (4/8) Epoch 7, batch 62000, loss[loss=2.284, over 2590.00 frames. , ppl: 9.820689367730546] tot_loss[loss=2.282, over 5590552.84 frames. , ppl: 9.794649410548907], batch size: 70
+2022-12-12 17:28:30,622 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:28:31,386 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71544765103413
+2022-12-12 17:30:11,249 INFO [train.py:421] (4/8) Epoch 7, batch 62200, loss[loss=2.17, over 5670.00 frames. , ppl: 8.755726656384523] tot_loss[loss=2.281, over 5603456.39 frames. , ppl: 9.78950986803643], batch size: 70
+2022-12-12 17:31:56,803 INFO [train.py:421] (4/8) Epoch 7, batch 62400, loss[loss=2.357, over 1680.00 frames. , ppl: 10.561411604963707] tot_loss[loss=2.281, over 5595626.29 frames. , ppl: 9.784855223133242], batch size: 70
+2022-12-12 17:33:40,849 INFO [train.py:421] (4/8) Epoch 7, batch 62600, loss[loss=2.292, over 4270.00 frames. , ppl: 9.891195481157276] tot_loss[loss=2.281, over 5607049.27 frames. , ppl: 9.786346665612193], batch size: 70
+2022-12-12 17:35:24,964 INFO [train.py:421] (4/8) Epoch 7, batch 62800, loss[loss=2.243, over 3640.00 frames. , ppl: 9.424192906158256] tot_loss[loss=2.281, over 5617869.50 frames. , ppl: 9.78190359434854], batch size: 70
+2022-12-12 17:37:08,813 INFO [train.py:421] (4/8) Epoch 7, batch 63000, loss[loss=2.376, over 1400.00 frames. , ppl: 10.759066712056987] tot_loss[loss=2.281, over 5601679.45 frames. , ppl: 9.782727576329048], batch size: 70
+2022-12-12 17:37:08,813 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:37:09,562 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.722217042257858
+2022-12-12 17:38:51,299 INFO [train.py:421] (4/8) Epoch 7, batch 63200, loss[loss=2.738, over 630.00 frames. , ppl: 15.46197747457492] tot_loss[loss=2.28, over 5594001.16 frames. , ppl: 9.77788037920395], batch size: 70
+2022-12-12 17:40:30,753 INFO [train.py:421] (4/8) Epoch 7, batch 63400, loss[loss=2.336, over 2310.00 frames. , ppl: 10.338191151677696] tot_loss[loss=2.279, over 5607757.43 frames. , ppl: 9.768454602891504], batch size: 70
+2022-12-12 17:42:12,622 INFO [train.py:421] (4/8) Epoch 7, batch 63600, loss[loss=2.617, over 980.00 frames. , ppl: 13.699903533433659] tot_loss[loss=2.28, over 5598220.40 frames. , ppl: 9.772782615104374], batch size: 70
+2022-12-12 17:43:53,054 INFO [train.py:421] (4/8) Epoch 7, batch 63800, loss[loss=2.265, over 3780.00 frames. , ppl: 9.627704431493822] tot_loss[loss=2.281, over 5536908.26 frames. , ppl: 9.789389004384077], batch size: 70
+2022-12-12 17:45:29,446 INFO [train.py:421] (4/8) Epoch 7, batch 64000, loss[loss=2.438, over 1330.00 frames. , ppl: 11.445122959276658] tot_loss[loss=2.281, over 5547037.18 frames. , ppl: 9.789511989404009], batch size: 70
+2022-12-12 17:45:29,446 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:45:30,209 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.717427922051451
+2022-12-12 17:47:12,552 INFO [train.py:421] (4/8) Epoch 7, batch 64200, loss[loss=2.179, over 4340.00 frames. , ppl: 8.834274032433433] tot_loss[loss=2.282, over 5539388.70 frames. , ppl: 9.794226438966426], batch size: 70
+2022-12-12 17:48:56,533 INFO [train.py:421] (4/8) Epoch 7, batch 64400, loss[loss=2.338, over 1540.00 frames. , ppl: 10.361516641975623] tot_loss[loss=2.28, over 5592032.07 frames. , ppl: 9.77878112592439], batch size: 70
+2022-12-12 17:50:37,245 INFO [train.py:421] (4/8) Epoch 7, batch 64600, loss[loss=2.416, over 1260.00 frames. , ppl: 11.204648591911502] tot_loss[loss=2.28, over 5592691.72 frames. , ppl: 9.779242940540179], batch size: 70
+2022-12-12 17:52:15,807 INFO [train.py:421] (4/8) Epoch 7, batch 64800, loss[loss=2.408, over 910.00 frames. , ppl: 11.11554425393335] tot_loss[loss=2.281, over 5575045.09 frames. , ppl: 9.78400187059679], batch size: 70
+2022-12-12 17:53:56,995 INFO [train.py:421] (4/8) Epoch 7, batch 65000, loss[loss=2.209, over 9870.00 frames. , ppl: 9.110556278818128] tot_loss[loss=2.28, over 5603280.36 frames. , ppl: 9.774807709956026], batch size: 70
+2022-12-12 17:53:56,995 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 17:53:57,744 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.728521733447238
+2022-12-12 17:55:38,086 INFO [train.py:421] (4/8) Epoch 7, batch 65200, loss[loss=2.49, over 910.00 frames. , ppl: 12.0600387847954] tot_loss[loss=2.28, over 5617614.58 frames. , ppl: 9.774113246293645], batch size: 70
+2022-12-12 17:57:16,335 INFO [train.py:421] (4/8) Epoch 7, batch 65400, loss[loss=2.23, over 4200.00 frames. , ppl: 9.295538047676674] tot_loss[loss=2.28, over 5612977.34 frames. , ppl: 9.776085427916465], batch size: 70
+2022-12-12 17:58:55,635 INFO [train.py:421] (4/8) Epoch 7, batch 65600, loss[loss=2.337, over 2800.00 frames. , ppl: 10.348746577568376] tot_loss[loss=2.28, over 5618129.26 frames. , ppl: 9.780922858596158], batch size: 70
+2022-12-12 18:00:40,205 INFO [train.py:421] (4/8) Epoch 7, batch 65800, loss[loss=2.194, over 3150.00 frames. , ppl: 8.969957906689483] tot_loss[loss=2.28, over 5611516.20 frames. , ppl: 9.779060927714918], batch size: 70
+2022-12-12 18:02:18,980 INFO [train.py:421] (4/8) Epoch 7, batch 66000, loss[loss=2.412, over 3220.00 frames. , ppl: 11.160131648003693] tot_loss[loss=2.281, over 5573067.97 frames. , ppl: 9.788665571522959], batch size: 70
+2022-12-12 18:02:18,981 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 18:02:19,737 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.275, over 211138.00 frames. , ppl: 9.725315618297568
+2022-12-12 18:04:01,116 INFO [train.py:421] (4/8) Epoch 7, batch 66200, loss[loss=2.211, over 2310.00 frames. , ppl: 9.1239616609036] tot_loss[loss=2.281, over 5550658.46 frames. , ppl: 9.784580448951086], batch size: 70
+2022-12-12 18:05:46,669 INFO [train.py:421] (4/8) Epoch 7, batch 66400, loss[loss=2.161, over 4690.00 frames. , ppl: 8.675573882196165] tot_loss[loss=2.281, over 5537834.99 frames. , ppl: 9.784042282107514], batch size: 70
+2022-12-12 18:07:29,663 INFO [train.py:421] (4/8) Epoch 7, batch 66600, loss[loss=2.211, over 4410.00 frames. , ppl: 9.126970217337105] tot_loss[loss=2.28, over 5524288.17 frames. , ppl: 9.777175194700447], batch size: 70
+2022-12-12 18:09:07,081 INFO [train.py:421] (4/8) Epoch 7, batch 66800, loss[loss=2.352, over 2450.00 frames. , ppl: 10.508054516966132] tot_loss[loss=2.281, over 5499634.37 frames. , ppl: 9.789629642672288], batch size: 70
+2022-12-12 18:10:47,191 INFO [train.py:421] (4/8) Epoch 7, batch 67000, loss[loss=2.213, over 5180.00 frames. , ppl: 9.140077996091932] tot_loss[loss=2.281, over 5499559.20 frames. , ppl: 9.789032016536167], batch size: 70
+2022-12-12 18:10:47,191 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 18:10:47,941 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.274, over 211138.00 frames. , ppl: 9.71459066540528
+2022-12-12 18:12:28,623 INFO [train.py:421] (4/8) Epoch 7, batch 67200, loss[loss=2.845, over 630.00 frames. , ppl: 17.202581268415617] tot_loss[loss=2.283, over 5443046.10 frames. , ppl: 9.805530942913995], batch size: 70
+2022-12-12 18:14:14,270 INFO [train.py:421] (4/8) Epoch 7, batch 67400, loss[loss=2.223, over 4690.00 frames. , ppl: 9.239551112464834] tot_loss[loss=2.282, over 5477303.37 frames. , ppl: 9.792920261819752], batch size: 70
+2022-12-12 18:15:54,840 INFO [train.py:421] (4/8) Epoch 7, batch 67600, loss[loss=2.358, over 2730.00 frames. , ppl: 10.568039666082422] tot_loss[loss=2.282, over 5484164.47 frames. , ppl: 9.799661830979822], batch size: 70
+2022-12-12 18:17:35,752 INFO [train.py:421] (4/8) Epoch 7, batch 67800, loss[loss=2.378, over 1750.00 frames. , ppl: 10.78564968657053] tot_loss[loss=2.283, over 5435393.36 frames. , ppl: 9.81010345331193], batch size: 70
+2022-12-12 18:19:18,270 INFO [train.py:421] (4/8) Epoch 7, batch 68000, loss[loss=2.19, over 7210.00 frames. , ppl: 8.937653765741794] tot_loss[loss=2.282, over 5477153.99 frames. , ppl: 9.79706116199177], batch size: 70
+2022-12-12 18:19:18,270 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 18:19:19,022 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.710605812793908
+2022-12-12 18:20:59,031 INFO [train.py:421] (4/8) Epoch 7, batch 68200, loss[loss=2.59, over 770.00 frames. , ppl: 13.32470594089376] tot_loss[loss=2.283, over 5477849.63 frames. , ppl: 9.803091701649175], batch size: 70
+2022-12-12 18:22:38,124 INFO [train.py:421] (4/8) Epoch 7, batch 68400, loss[loss=2.148, over 5320.00 frames. , ppl: 8.57130398364342] tot_loss[loss=2.282, over 5489289.35 frames. , ppl: 9.796949925650392], batch size: 70
+2022-12-12 18:24:19,216 INFO [train.py:421] (4/8) Epoch 7, batch 68600, loss[loss=2.246, over 5110.00 frames. , ppl: 9.44624689415712] tot_loss[loss=2.282, over 5506904.63 frames. , ppl: 9.798739435503043], batch size: 70
+2022-12-12 18:25:59,214 INFO [train.py:421] (4/8) Epoch 7, batch 68800, loss[loss=2.329, over 1470.00 frames. , ppl: 10.270823154457617] tot_loss[loss=2.284, over 5457062.46 frames. , ppl: 9.817461494935348], batch size: 70
+2022-12-12 18:27:41,645 INFO [train.py:421] (4/8) Epoch 7, batch 69000, loss[loss=2.351, over 1680.00 frames. , ppl: 10.491361694925633] tot_loss[loss=2.284, over 5469077.27 frames. , ppl: 9.811454295699676], batch size: 70
+2022-12-12 18:27:41,646 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 18:27:42,411 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.71098525213279
+2022-12-12 18:29:24,658 INFO [train.py:421] (4/8) Epoch 7, batch 69200, loss[loss=2.24, over 9030.00 frames. , ppl: 9.390531194452702] tot_loss[loss=2.283, over 5465564.46 frames. , ppl: 9.81010085409491], batch size: 70
+2022-12-12 18:31:04,101 INFO [train.py:421] (4/8) Epoch 7, batch 69400, loss[loss=2.265, over 3850.00 frames. , ppl: 9.629556348631795] tot_loss[loss=2.283, over 5477040.75 frames. , ppl: 9.807791737207772], batch size: 70
+2022-12-12 18:32:46,724 INFO [train.py:421] (4/8) Epoch 7, batch 69600, loss[loss=2.281, over 2170.00 frames. , ppl: 9.781766126584039] tot_loss[loss=2.282, over 5513692.06 frames. , ppl: 9.799092886023663], batch size: 70
+2022-12-12 18:34:31,065 INFO [train.py:421] (4/8) Epoch 7, batch 69800, loss[loss=2.145, over 9660.00 frames. , ppl: 8.543175318837495] tot_loss[loss=2.282, over 5515271.66 frames. , ppl: 9.79792359847446], batch size: 70
+2022-12-12 18:36:12,716 INFO [train.py:421] (4/8) Epoch 7, batch 70000, loss[loss=2.337, over 2240.00 frames. , ppl: 10.348532725413193] tot_loss[loss=2.282, over 5525801.97 frames. , ppl: 9.799137661648784], batch size: 70
+2022-12-12 18:36:12,717 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 18:36:13,480 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.71248878254001
+2022-12-12 18:37:52,814 INFO [train.py:421] (4/8) Epoch 7, batch 70200, loss[loss=2.269, over 3500.00 frames. , ppl: 9.667123006449437] tot_loss[loss=2.282, over 5514803.19 frames. , ppl: 9.800444290309033], batch size: 70
+2022-12-12 18:39:34,533 INFO [train.py:421] (4/8) Epoch 7, batch 70400, loss[loss=3.497, over 420.00 frames. , ppl: 32.99990623345811] tot_loss[loss=2.282, over 5517295.93 frames. , ppl: 9.796977506556711], batch size: 70
+2022-12-12 18:41:16,497 INFO [train.py:421] (4/8) Epoch 7, batch 70600, loss[loss=2.201, over 4690.00 frames. , ppl: 9.034677803934066] tot_loss[loss=2.283, over 5510965.44 frames. , ppl: 9.801947484728657], batch size: 70
+2022-12-12 18:43:01,189 INFO [train.py:421] (4/8) Epoch 7, batch 70800, loss[loss=2.442, over 1120.00 frames. , ppl: 11.496686807991642] tot_loss[loss=2.282, over 5526351.57 frames. , ppl: 9.795842109514359], batch size: 70
+2022-12-12 18:44:42,633 INFO [train.py:421] (4/8) Epoch 7, batch 71000, loss[loss=2.298, over 1890.00 frames. , ppl: 9.950966520514841] tot_loss[loss=2.282, over 5517078.11 frames. , ppl: 9.800100086237405], batch size: 70
+2022-12-12 18:44:42,634 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 18:44:43,417 INFO [train.py:452] (4/8) Epoch 7, validation: loss=2.273, over 211138.00 frames. , ppl: 9.707774858509012
+2022-12-12 18:46:25,670 INFO [train.py:421] (4/8) Epoch 7, batch 71200, loss[loss=2.193, over 3850.00 frames. , ppl: 8.959060633624766] tot_loss[loss=2.282, over 5527379.66 frames. , ppl: 9.791397850297724], batch size: 70
+2022-12-12 18:48:08,695 INFO [train.py:421] (4/8) Epoch 7, batch 71400, loss[loss=2.55, over 980.00 frames. , ppl: 12.81350561588308] tot_loss[loss=2.282, over 5504592.56 frames. , ppl: 9.795197399490938], batch size: 70
+2022-12-12 18:49:50,960 INFO [train.py:421] (4/8) Epoch 7, batch 71600, loss[loss=2.777, over 630.00 frames. , ppl: 16.076298993260792] tot_loss[loss=2.283, over 5469186.71 frames. , ppl: 9.808340554347216], batch size: 70
+2022-12-12 18:51:32,658 INFO [train.py:421] (4/8) Epoch 7, batch 71800, loss[loss=2.449, over 840.00 frames. , ppl: 11.579714182049809] tot_loss[loss=2.283, over 5460027.65 frames. , ppl: 9.81017024320273], batch size: 70
+2022-12-12 18:52:48,002 INFO [train.py:421] (4/8) Epoch 8, batch 0, loss[loss=2.492, over 1120.00 frames. , ppl: 12.09032687034645] tot_loss[loss=2.492, over 1120.00 frames. , ppl: 12.09032687034645], batch size: 70
+2022-12-12 18:54:30,706 INFO [train.py:421] (4/8) Epoch 8, batch 200, loss[loss=2.334, over 2590.00 frames. , ppl: 10.315142689267542] tot_loss[loss=2.293, over 473178.15 frames. , ppl: 9.903474821903934], batch size: 70
+2022-12-12 18:56:11,896 INFO [train.py:421] (4/8) Epoch 8, batch 400, loss[loss=2.235, over 1400.00 frames. , ppl: 9.34881289566682] tot_loss[loss=2.282, over 948128.40 frames. , ppl: 9.797181761990577], batch size: 70
+2022-12-12 18:57:54,089 INFO [train.py:421] (4/8) Epoch 8, batch 600, loss[loss=2.23, over 2800.00 frames. , ppl: 9.298614158905812] tot_loss[loss=2.276, over 1404078.54 frames. , ppl: 9.734930586197018], batch size: 70
+2022-12-12 18:59:34,601 INFO [train.py:421] (4/8) Epoch 8, batch 800, loss[loss=2.609, over 770.00 frames. , ppl: 13.582382197245456] tot_loss[loss=2.274, over 1804428.56 frames. , ppl: 9.714694232579644], batch size: 70
+2022-12-12 19:01:16,011 INFO [train.py:421] (4/8) Epoch 8, batch 1000, loss[loss=2.312, over 1190.00 frames. , ppl: 10.092040684236926] tot_loss[loss=2.273, over 2166026.24 frames. , ppl: 9.706257016323326], batch size: 70
+2022-12-12 19:01:16,012 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:01:16,794 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.274, over 211138.00 frames. , ppl: 9.714002609944576
+2022-12-12 19:03:00,009 INFO [train.py:421] (4/8) Epoch 8, batch 1200, loss[loss=2.109, over 10080.00 frames. , ppl: 8.241279911810677] tot_loss[loss=2.273, over 2476462.82 frames. , ppl: 9.707211373713273], batch size: 70
+2022-12-12 19:04:36,368 INFO [train.py:421] (4/8) Epoch 8, batch 1400, loss[loss=2.134, over 8820.00 frames. , ppl: 8.451547275088076] tot_loss[loss=2.275, over 2764886.20 frames. , ppl: 9.723455204664273], batch size: 70
+2022-12-12 19:06:17,625 INFO [train.py:421] (4/8) Epoch 8, batch 1600, loss[loss=2.369, over 1190.00 frames. , ppl: 10.689086894060244] tot_loss[loss=2.276, over 3002005.28 frames. , ppl: 9.733659914647331], batch size: 70
+2022-12-12 19:08:00,658 INFO [train.py:421] (4/8) Epoch 8, batch 1800, loss[loss=2.195, over 4060.00 frames. , ppl: 8.979990688905414] tot_loss[loss=2.275, over 3238514.56 frames. , ppl: 9.730668036497905], batch size: 70
+2022-12-12 19:09:38,945 INFO [train.py:421] (4/8) Epoch 8, batch 2000, loss[loss=2.242, over 3220.00 frames. , ppl: 9.407699708924788] tot_loss[loss=2.274, over 3448628.34 frames. , ppl: 9.722401616927048], batch size: 70
+2022-12-12 19:09:38,945 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:09:39,713 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.70567014079976
+2022-12-12 19:11:23,707 INFO [train.py:421] (4/8) Epoch 8, batch 2200, loss[loss=2.295, over 2800.00 frames. , ppl: 9.925962593229787] tot_loss[loss=2.276, over 3626461.41 frames. , ppl: 9.739838421606953], batch size: 70
+2022-12-12 19:13:02,999 INFO [train.py:421] (4/8) Epoch 8, batch 2400, loss[loss=3.197, over 490.00 frames. , ppl: 24.4638707347068] tot_loss[loss=2.278, over 3758143.26 frames. , ppl: 9.759793972215691], batch size: 70
+2022-12-12 19:14:41,152 INFO [train.py:421] (4/8) Epoch 8, batch 2600, loss[loss=2.439, over 980.00 frames. , ppl: 11.45652596012185] tot_loss[loss=2.279, over 3894751.03 frames. , ppl: 9.768481825523489], batch size: 70
+2022-12-12 19:16:22,764 INFO [train.py:421] (4/8) Epoch 8, batch 2800, loss[loss=2.985, over 560.00 frames. , ppl: 19.779058337503756] tot_loss[loss=2.281, over 4002926.14 frames. , ppl: 9.786295407195444], batch size: 70
+2022-12-12 19:18:01,848 INFO [train.py:421] (4/8) Epoch 8, batch 3000, loss[loss=2.381, over 1470.00 frames. , ppl: 10.812978261193729] tot_loss[loss=2.281, over 4112447.58 frames. , ppl: 9.789154145919214], batch size: 70
+2022-12-12 19:18:01,848 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:18:02,632 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.276, over 211138.00 frames. , ppl: 9.739168370327212
+2022-12-12 19:19:44,665 INFO [train.py:421] (4/8) Epoch 8, batch 3200, loss[loss=2.638, over 770.00 frames. , ppl: 13.985828408753912] tot_loss[loss=2.279, over 4295847.33 frames. , ppl: 9.76720056349398], batch size: 70
+2022-12-12 19:21:28,603 INFO [train.py:421] (4/8) Epoch 8, batch 3400, loss[loss=2.137, over 6510.00 frames. , ppl: 8.477989193614436] tot_loss[loss=2.28, over 4389300.95 frames. , ppl: 9.779156248053294], batch size: 70
+2022-12-12 19:23:11,767 INFO [train.py:421] (4/8) Epoch 8, batch 3600, loss[loss=2.309, over 1750.00 frames. , ppl: 10.062415031763818] tot_loss[loss=2.278, over 4539600.37 frames. , ppl: 9.754616643810701], batch size: 70
+2022-12-12 19:24:50,908 INFO [train.py:421] (4/8) Epoch 8, batch 3800, loss[loss=2.164, over 6160.00 frames. , ppl: 8.709616497351728] tot_loss[loss=2.277, over 4646938.99 frames. , ppl: 9.750382289986879], batch size: 70
+2022-12-12 19:26:35,947 INFO [train.py:421] (4/8) Epoch 8, batch 4000, loss[loss=2.183, over 5390.00 frames. , ppl: 8.868613264805933] tot_loss[loss=2.277, over 4724829.99 frames. , ppl: 9.748142324894864], batch size: 70
+2022-12-12 19:26:35,947 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:26:36,699 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.709655842207155
+2022-12-12 19:28:16,983 INFO [train.py:421] (4/8) Epoch 8, batch 4200, loss[loss=3.117, over 560.00 frames. , ppl: 22.574773366768735] tot_loss[loss=2.278, over 4786484.83 frames. , ppl: 9.752575344366319], batch size: 70
+2022-12-12 19:29:59,662 INFO [train.py:421] (4/8) Epoch 8, batch 4400, loss[loss=2.153, over 4760.00 frames. , ppl: 8.611867618669566] tot_loss[loss=2.277, over 4854632.22 frames. , ppl: 9.749682398620525], batch size: 70
+2022-12-12 19:31:43,099 INFO [train.py:421] (4/8) Epoch 8, batch 4600, loss[loss=2.172, over 4620.00 frames. , ppl: 8.776279378272754] tot_loss[loss=2.278, over 4919730.97 frames. , ppl: 9.753123213039498], batch size: 70
+2022-12-12 19:33:23,009 INFO [train.py:421] (4/8) Epoch 8, batch 4800, loss[loss=2.216, over 2590.00 frames. , ppl: 9.16676694616359] tot_loss[loss=2.278, over 4948827.46 frames. , ppl: 9.76155604912291], batch size: 70
+2022-12-12 19:35:03,760 INFO [train.py:421] (4/8) Epoch 8, batch 5000, loss[loss=2.213, over 6090.00 frames. , ppl: 9.139653836577038] tot_loss[loss=2.278, over 5027828.46 frames. , ppl: 9.75723297581847], batch size: 70
+2022-12-12 19:35:03,760 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:35:04,513 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.275, over 211138.00 frames. , ppl: 9.73047586561236
+2022-12-12 19:36:45,801 INFO [train.py:421] (4/8) Epoch 8, batch 5200, loss[loss=2.334, over 1610.00 frames. , ppl: 10.321481464412772] tot_loss[loss=2.278, over 5063729.62 frames. , ppl: 9.757078380495356], batch size: 70
+2022-12-12 19:38:27,894 INFO [train.py:421] (4/8) Epoch 8, batch 5400, loss[loss=3.607, over 420.00 frames. , ppl: 36.86671477263306] tot_loss[loss=2.277, over 5159323.88 frames. , ppl: 9.747706728221385], batch size: 70
+2022-12-12 19:40:10,516 INFO [train.py:421] (4/8) Epoch 8, batch 5600, loss[loss=2.52, over 910.00 frames. , ppl: 12.432774731404423] tot_loss[loss=2.276, over 5237711.66 frames. , ppl: 9.733046942668649], batch size: 70
+2022-12-12 19:41:53,315 INFO [train.py:421] (4/8) Epoch 8, batch 5800, loss[loss=2.647, over 1260.00 frames. , ppl: 14.114593189485447] tot_loss[loss=2.275, over 5298321.04 frames. , ppl: 9.723620115725106], batch size: 70
+2022-12-12 19:43:36,498 INFO [train.py:421] (4/8) Epoch 8, batch 6000, loss[loss=2.158, over 4410.00 frames. , ppl: 8.651830082214452] tot_loss[loss=2.275, over 5320278.75 frames. , ppl: 9.723738137250365], batch size: 70
+2022-12-12 19:43:36,498 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:43:37,249 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.709706140891356
+2022-12-12 19:45:21,824 INFO [train.py:421] (4/8) Epoch 8, batch 6200, loss[loss=2.264, over 4060.00 frames. , ppl: 9.622585695265105] tot_loss[loss=2.275, over 5341966.05 frames. , ppl: 9.726860267749382], batch size: 70
+2022-12-12 19:47:05,535 INFO [train.py:421] (4/8) Epoch 8, batch 6400, loss[loss=2.297, over 2030.00 frames. , ppl: 9.947487287666839] tot_loss[loss=2.275, over 5370701.84 frames. , ppl: 9.7256655028756], batch size: 70
+2022-12-12 19:48:44,539 INFO [train.py:421] (4/8) Epoch 8, batch 6600, loss[loss=2.172, over 8890.00 frames. , ppl: 8.772223157348378] tot_loss[loss=2.276, over 5345418.13 frames. , ppl: 9.735983434749183], batch size: 70
+2022-12-12 19:50:24,296 INFO [train.py:421] (4/8) Epoch 8, batch 6800, loss[loss=2.174, over 4830.00 frames. , ppl: 8.791129542077307] tot_loss[loss=2.276, over 5366524.43 frames. , ppl: 9.74023960936621], batch size: 70
+2022-12-12 19:52:06,684 INFO [train.py:421] (4/8) Epoch 8, batch 7000, loss[loss=2.282, over 2940.00 frames. , ppl: 9.792035741131803] tot_loss[loss=2.276, over 5369848.77 frames. , ppl: 9.740657614646357], batch size: 70
+2022-12-12 19:52:06,684 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 19:52:07,437 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.703498378235237
+2022-12-12 19:53:49,599 INFO [train.py:421] (4/8) Epoch 8, batch 7200, loss[loss=2.291, over 2380.00 frames. , ppl: 9.88820517443104] tot_loss[loss=2.277, over 5340741.81 frames. , ppl: 9.743162854910796], batch size: 70
+2022-12-12 19:55:33,510 INFO [train.py:421] (4/8) Epoch 8, batch 7400, loss[loss=2.439, over 1120.00 frames. , ppl: 11.460438508466623] tot_loss[loss=2.277, over 5387289.69 frames. , ppl: 9.749711487135567], batch size: 70
+2022-12-12 19:57:12,619 INFO [train.py:421] (4/8) Epoch 8, batch 7600, loss[loss=2.593, over 1330.00 frames. , ppl: 13.372053902269206] tot_loss[loss=2.277, over 5401802.62 frames. , ppl: 9.74500494075115], batch size: 70
+2022-12-12 19:58:56,551 INFO [train.py:421] (4/8) Epoch 8, batch 7800, loss[loss=2.419, over 2590.00 frames. , ppl: 11.229060498447318] tot_loss[loss=2.277, over 5407869.58 frames. , ppl: 9.74590130325694], batch size: 70
+2022-12-12 20:00:37,106 INFO [train.py:421] (4/8) Epoch 8, batch 8000, loss[loss=2.217, over 3430.00 frames. , ppl: 9.182082036469073] tot_loss[loss=2.277, over 5414609.14 frames. , ppl: 9.74935952898934], batch size: 70
+2022-12-12 20:00:37,106 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:00:37,873 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.274, over 211138.00 frames. , ppl: 9.715900618606277
+2022-12-12 20:02:19,423 INFO [train.py:421] (4/8) Epoch 8, batch 8200, loss[loss=2.33, over 1610.00 frames. , ppl: 10.273465110047026] tot_loss[loss=2.279, over 5378985.88 frames. , ppl: 9.767880421815835], batch size: 70
+2022-12-12 20:04:00,775 INFO [train.py:421] (4/8) Epoch 8, batch 8400, loss[loss=2.321, over 2730.00 frames. , ppl: 10.190303389205393] tot_loss[loss=2.279, over 5390064.44 frames. , ppl: 9.765994338880285], batch size: 70
+2022-12-12 20:05:43,632 INFO [train.py:421] (4/8) Epoch 8, batch 8600, loss[loss=2.234, over 3220.00 frames. , ppl: 9.33394771006695] tot_loss[loss=2.277, over 5437367.37 frames. , ppl: 9.751335264268187], batch size: 70
+2022-12-12 20:07:22,568 INFO [train.py:421] (4/8) Epoch 8, batch 8800, loss[loss=2.23, over 6370.00 frames. , ppl: 9.295569351691094] tot_loss[loss=2.278, over 5421247.93 frames. , ppl: 9.757555979010961], batch size: 70
+2022-12-12 20:09:06,511 INFO [train.py:421] (4/8) Epoch 8, batch 9000, loss[loss=2.323, over 1960.00 frames. , ppl: 10.209405984745878] tot_loss[loss=2.28, over 5397557.18 frames. , ppl: 9.774952326851404], batch size: 70
+2022-12-12 20:09:06,512 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:09:07,262 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.70899192399486
+2022-12-12 20:10:45,538 INFO [train.py:421] (4/8) Epoch 8, batch 9200, loss[loss=2.455, over 1260.00 frames. , ppl: 11.643604516143085] tot_loss[loss=2.28, over 5403758.60 frames. , ppl: 9.774158934324204], batch size: 70
+2022-12-12 20:12:29,544 INFO [train.py:421] (4/8) Epoch 8, batch 9400, loss[loss=2.234, over 5250.00 frames. , ppl: 9.337344988943327] tot_loss[loss=2.279, over 5452633.93 frames. , ppl: 9.766072233410272], batch size: 70
+2022-12-12 20:14:10,749 INFO [train.py:421] (4/8) Epoch 8, batch 9600, loss[loss=2.113, over 7770.00 frames. , ppl: 8.2711140894494] tot_loss[loss=2.278, over 5471638.00 frames. , ppl: 9.761426267543978], batch size: 70
+2022-12-12 20:15:50,146 INFO [train.py:421] (4/8) Epoch 8, batch 9800, loss[loss=2.19, over 10850.00 frames. , ppl: 8.939125700321576] tot_loss[loss=2.276, over 5517105.56 frames. , ppl: 9.740980041200729], batch size: 70
+2022-12-12 20:17:30,581 INFO [train.py:421] (4/8) Epoch 8, batch 10000, loss[loss=2.214, over 11620.00 frames. , ppl: 9.149074380165548] tot_loss[loss=2.276, over 5508797.20 frames. , ppl: 9.736458508469523], batch size: 70
+2022-12-12 20:17:30,582 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:17:31,332 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.276, over 211138.00 frames. , ppl: 9.734638884938837
+2022-12-12 20:19:15,342 INFO [train.py:421] (4/8) Epoch 8, batch 10200, loss[loss=2.298, over 4410.00 frames. , ppl: 9.950140950787505] tot_loss[loss=2.275, over 5555912.85 frames. , ppl: 9.724371821828015], batch size: 70
+2022-12-12 20:20:52,758 INFO [train.py:421] (4/8) Epoch 8, batch 10400, loss[loss=2.258, over 3220.00 frames. , ppl: 9.565739958787812] tot_loss[loss=2.276, over 5510077.21 frames. , ppl: 9.736157662118462], batch size: 70
+2022-12-12 20:22:35,029 INFO [train.py:421] (4/8) Epoch 8, batch 10600, loss[loss=2.279, over 1540.00 frames. , ppl: 9.76263547546335] tot_loss[loss=2.275, over 5507095.37 frames. , ppl: 9.730605302660061], batch size: 70
+2022-12-12 20:24:13,571 INFO [train.py:421] (4/8) Epoch 8, batch 10800, loss[loss=2.196, over 13160.00 frames. , ppl: 8.992067396439413] tot_loss[loss=2.276, over 5486072.00 frames. , ppl: 9.735768581957526], batch size: 70
+2022-12-12 20:25:55,032 INFO [train.py:421] (4/8) Epoch 8, batch 11000, loss[loss=2.426, over 1470.00 frames. , ppl: 11.312886969370329] tot_loss[loss=2.276, over 5466191.72 frames. , ppl: 9.742373986594211], batch size: 70
+2022-12-12 20:25:55,033 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:25:55,810 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.706705920947643
+2022-12-12 20:27:38,289 INFO [train.py:421] (4/8) Epoch 8, batch 11200, loss[loss=2.816, over 700.00 frames. , ppl: 16.70797925235234] tot_loss[loss=2.276, over 5474537.52 frames. , ppl: 9.742471323817236], batch size: 70
+2022-12-12 20:29:17,860 INFO [train.py:421] (4/8) Epoch 8, batch 11400, loss[loss=2.239, over 2730.00 frames. , ppl: 9.383609396160688] tot_loss[loss=2.277, over 5488025.17 frames. , ppl: 9.744281395145586], batch size: 70
+2022-12-12 20:30:55,493 INFO [train.py:421] (4/8) Epoch 8, batch 11600, loss[loss=2.554, over 840.00 frames. , ppl: 12.85693042337297] tot_loss[loss=2.277, over 5464314.94 frames. , ppl: 9.748147541753392], batch size: 70
+2022-12-12 20:32:40,056 INFO [train.py:421] (4/8) Epoch 8, batch 11800, loss[loss=2.421, over 2100.00 frames. , ppl: 11.259590316673522] tot_loss[loss=2.276, over 5492578.36 frames. , ppl: 9.734468102555887], batch size: 70
+2022-12-12 20:34:17,560 INFO [train.py:421] (4/8) Epoch 8, batch 12000, loss[loss=2.428, over 1540.00 frames. , ppl: 11.338768894397717] tot_loss[loss=2.275, over 5506866.45 frames. , ppl: 9.727774276744642], batch size: 70
+2022-12-12 20:34:17,560 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:34:18,324 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.712665599212405
+2022-12-12 20:36:00,924 INFO [train.py:421] (4/8) Epoch 8, batch 12200, loss[loss=2.408, over 1470.00 frames. , ppl: 11.109717099929759] tot_loss[loss=2.275, over 5514177.58 frames. , ppl: 9.723266597430113], batch size: 70
+2022-12-12 20:37:42,035 INFO [train.py:421] (4/8) Epoch 8, batch 12400, loss[loss=2.179, over 6860.00 frames. , ppl: 8.839663043352447] tot_loss[loss=2.275, over 5507934.67 frames. , ppl: 9.73038408362234], batch size: 70
+2022-12-12 20:39:23,002 INFO [train.py:421] (4/8) Epoch 8, batch 12600, loss[loss=2.141, over 6160.00 frames. , ppl: 8.506016220989666] tot_loss[loss=2.275, over 5498736.43 frames. , ppl: 9.73176043841532], batch size: 70
+2022-12-12 20:41:03,256 INFO [train.py:421] (4/8) Epoch 8, batch 12800, loss[loss=2.308, over 1750.00 frames. , ppl: 10.05765307655896] tot_loss[loss=2.276, over 5458408.71 frames. , ppl: 9.740135274700569], batch size: 70
+2022-12-12 20:42:44,548 INFO [train.py:421] (4/8) Epoch 8, batch 13000, loss[loss=2.377, over 1960.00 frames. , ppl: 10.76985294618835] tot_loss[loss=2.276, over 5462058.95 frames. , ppl: 9.741174737067539], batch size: 70
+2022-12-12 20:42:44,549 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:42:45,301 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.274, over 211138.00 frames. , ppl: 9.722002639478895
+2022-12-12 20:44:28,071 INFO [train.py:421] (4/8) Epoch 8, batch 13200, loss[loss=2.341, over 2520.00 frames. , ppl: 10.394022642740744] tot_loss[loss=2.277, over 5440335.07 frames. , ppl: 9.74434919453241], batch size: 70
+2022-12-12 20:46:10,752 INFO [train.py:421] (4/8) Epoch 8, batch 13400, loss[loss=2.154, over 9310.00 frames. , ppl: 8.620869047943742] tot_loss[loss=2.275, over 5477202.72 frames. , ppl: 9.7310891375291], batch size: 70
+2022-12-12 20:47:53,585 INFO [train.py:421] (4/8) Epoch 8, batch 13600, loss[loss=2.388, over 2940.00 frames. , ppl: 10.89069289739276] tot_loss[loss=2.276, over 5484387.13 frames. , ppl: 9.738070092855997], batch size: 70
+2022-12-12 20:49:37,900 INFO [train.py:421] (4/8) Epoch 8, batch 13800, loss[loss=2.747, over 840.00 frames. , ppl: 15.599654517109625] tot_loss[loss=2.277, over 5455469.72 frames. , ppl: 9.74698192593992], batch size: 70
+2022-12-12 20:51:19,926 INFO [train.py:421] (4/8) Epoch 8, batch 14000, loss[loss=2.291, over 2800.00 frames. , ppl: 9.887582190714822] tot_loss[loss=2.278, over 5417921.44 frames. , ppl: 9.758305419514102], batch size: 70
+2022-12-12 20:51:19,927 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:51:20,693 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.699218039613152
+2022-12-12 20:53:03,141 INFO [train.py:421] (4/8) Epoch 8, batch 14200, loss[loss=2.344, over 1050.00 frames. , ppl: 10.421475016460628] tot_loss[loss=2.278, over 5424812.84 frames. , ppl: 9.759748030893894], batch size: 70
+2022-12-12 20:54:47,606 INFO [train.py:421] (4/8) Epoch 8, batch 14400, loss[loss=2.295, over 3080.00 frames. , ppl: 9.928203640291576] tot_loss[loss=2.278, over 5451685.36 frames. , ppl: 9.752962697954748], batch size: 70
+2022-12-12 20:56:30,676 INFO [train.py:421] (4/8) Epoch 8, batch 14600, loss[loss=2.453, over 1330.00 frames. , ppl: 11.626603744070701] tot_loss[loss=2.277, over 5475287.52 frames. , ppl: 9.744669597686457], batch size: 70
+2022-12-12 20:58:11,810 INFO [train.py:421] (4/8) Epoch 8, batch 14800, loss[loss=2.241, over 2310.00 frames. , ppl: 9.404406776361524] tot_loss[loss=2.277, over 5494358.27 frames. , ppl: 9.743395937446945], batch size: 70
+2022-12-12 20:59:52,450 INFO [train.py:421] (4/8) Epoch 8, batch 15000, loss[loss=2.15, over 3500.00 frames. , ppl: 8.584814083709578] tot_loss[loss=2.276, over 5505308.22 frames. , ppl: 9.740488630198385], batch size: 70
+2022-12-12 20:59:52,450 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 20:59:53,203 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.698958207336654
+2022-12-12 21:01:33,668 INFO [train.py:421] (4/8) Epoch 8, batch 15200, loss[loss=2.268, over 3920.00 frames. , ppl: 9.658233905218017] tot_loss[loss=2.277, over 5503826.87 frames. , ppl: 9.744215161939852], batch size: 70
+2022-12-12 21:03:13,958 INFO [train.py:421] (4/8) Epoch 8, batch 15400, loss[loss=2.52, over 1330.00 frames. , ppl: 12.424336562246522] tot_loss[loss=2.278, over 5452685.45 frames. , ppl: 9.76093370763074], batch size: 70
+2022-12-12 21:04:56,825 INFO [train.py:421] (4/8) Epoch 8, batch 15600, loss[loss=2.429, over 2030.00 frames. , ppl: 11.343179619301631] tot_loss[loss=2.277, over 5488551.41 frames. , ppl: 9.750177421705041], batch size: 70
+2022-12-12 21:06:37,398 INFO [train.py:421] (4/8) Epoch 8, batch 15800, loss[loss=2.268, over 3080.00 frames. , ppl: 9.656795171141184] tot_loss[loss=2.278, over 5479528.04 frames. , ppl: 9.753274208595329], batch size: 70
+2022-12-12 21:08:18,882 INFO [train.py:421] (4/8) Epoch 8, batch 16000, loss[loss=2.178, over 4270.00 frames. , ppl: 8.827400870598721] tot_loss[loss=2.278, over 5467345.39 frames. , ppl: 9.758321521325898], batch size: 70
+2022-12-12 21:08:18,883 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:08:19,648 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.704328531673617
+2022-12-12 21:09:58,160 INFO [train.py:421] (4/8) Epoch 8, batch 16200, loss[loss=2.309, over 2940.00 frames. , ppl: 10.064057608000997] tot_loss[loss=2.277, over 5484393.78 frames. , ppl: 9.747650931290725], batch size: 70
+2022-12-12 21:11:38,668 INFO [train.py:421] (4/8) Epoch 8, batch 16400, loss[loss=2.364, over 1610.00 frames. , ppl: 10.637954891751841] tot_loss[loss=2.278, over 5496881.70 frames. , ppl: 9.752601553461934], batch size: 70
+2022-12-12 21:13:18,887 INFO [train.py:421] (4/8) Epoch 8, batch 16600, loss[loss=2.289, over 1400.00 frames. , ppl: 9.868583630649503] tot_loss[loss=2.277, over 5520908.11 frames. , ppl: 9.748336528621024], batch size: 70
+2022-12-12 21:14:59,021 INFO [train.py:421] (4/8) Epoch 8, batch 16800, loss[loss=2.244, over 3010.00 frames. , ppl: 9.431403465154553] tot_loss[loss=2.278, over 5477026.53 frames. , ppl: 9.757517771247988], batch size: 70
+2022-12-12 21:16:43,666 INFO [train.py:421] (4/8) Epoch 8, batch 17000, loss[loss=2.291, over 4480.00 frames. , ppl: 9.886316666706946] tot_loss[loss=2.277, over 5499400.89 frames. , ppl: 9.747414481935214], batch size: 70
+2022-12-12 21:16:43,667 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:16:44,419 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.701667408509223
+2022-12-12 21:18:25,170 INFO [train.py:421] (4/8) Epoch 8, batch 17200, loss[loss=2.433, over 1120.00 frames. , ppl: 11.388016313270223] tot_loss[loss=2.277, over 5473510.91 frames. , ppl: 9.752112912324257], batch size: 70
+2022-12-12 21:20:08,369 INFO [train.py:421] (4/8) Epoch 8, batch 17400, loss[loss=2.253, over 2800.00 frames. , ppl: 9.519499607231683] tot_loss[loss=2.278, over 5474998.48 frames. , ppl: 9.760043652506495], batch size: 70
+2022-12-12 21:21:48,828 INFO [train.py:421] (4/8) Epoch 8, batch 17600, loss[loss=2.254, over 2030.00 frames. , ppl: 9.521729283718285] tot_loss[loss=2.28, over 5420435.11 frames. , ppl: 9.774553791512705], batch size: 70
+2022-12-12 21:23:31,881 INFO [train.py:421] (4/8) Epoch 8, batch 17800, loss[loss=2.216, over 4480.00 frames. , ppl: 9.169071318471223] tot_loss[loss=2.28, over 5405325.57 frames. , ppl: 9.773675999705093], batch size: 70
+2022-12-12 21:25:12,513 INFO [train.py:421] (4/8) Epoch 8, batch 18000, loss[loss=2.9, over 630.00 frames. , ppl: 18.174194670034474] tot_loss[loss=2.28, over 5416768.95 frames. , ppl: 9.7744227189545], batch size: 70
+2022-12-12 21:25:12,514 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:25:13,280 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.693148970049764
+2022-12-12 21:26:52,972 INFO [train.py:421] (4/8) Epoch 8, batch 18200, loss[loss=2.267, over 2170.00 frames. , ppl: 9.649288485765787] tot_loss[loss=2.28, over 5420001.51 frames. , ppl: 9.772217285703537], batch size: 70
+2022-12-12 21:28:32,811 INFO [train.py:421] (4/8) Epoch 8, batch 18400, loss[loss=2.209, over 5390.00 frames. , ppl: 9.103615639050243] tot_loss[loss=2.28, over 5419689.10 frames. , ppl: 9.774901009284582], batch size: 70
+2022-12-12 21:30:13,694 INFO [train.py:421] (4/8) Epoch 8, batch 18600, loss[loss=2.22, over 6230.00 frames. , ppl: 9.210812099035001] tot_loss[loss=2.28, over 5417562.33 frames. , ppl: 9.781273509467072], batch size: 70
+2022-12-12 21:31:51,693 INFO [train.py:421] (4/8) Epoch 8, batch 18800, loss[loss=2.216, over 3850.00 frames. , ppl: 9.17338923382275] tot_loss[loss=2.279, over 5434297.29 frames. , ppl: 9.77037009120243], batch size: 70
+2022-12-12 21:33:32,792 INFO [train.py:421] (4/8) Epoch 8, batch 19000, loss[loss=2.475, over 1750.00 frames. , ppl: 11.884638118830216] tot_loss[loss=2.28, over 5412987.55 frames. , ppl: 9.777912878942141], batch size: 70
+2022-12-12 21:33:32,792 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:33:33,556 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.703235558989775
+2022-12-12 21:35:14,717 INFO [train.py:421] (4/8) Epoch 8, batch 19200, loss[loss=3.708, over 420.00 frames. , ppl: 40.761515384176484] tot_loss[loss=2.28, over 5432516.74 frames. , ppl: 9.772183492654928], batch size: 70
+2022-12-12 21:36:52,519 INFO [train.py:421] (4/8) Epoch 8, batch 19400, loss[loss=2.333, over 1890.00 frames. , ppl: 10.304335638841826] tot_loss[loss=2.279, over 5410359.04 frames. , ppl: 9.76978213117615], batch size: 70
+2022-12-12 21:38:36,742 INFO [train.py:421] (4/8) Epoch 8, batch 19600, loss[loss=2.255, over 3010.00 frames. , ppl: 9.53623319960199] tot_loss[loss=2.278, over 5447655.66 frames. , ppl: 9.754348901856241], batch size: 70
+2022-12-12 21:40:17,794 INFO [train.py:421] (4/8) Epoch 8, batch 19800, loss[loss=2.438, over 1470.00 frames. , ppl: 11.453793023968398] tot_loss[loss=2.278, over 5446870.86 frames. , ppl: 9.75621098965811], batch size: 70
+2022-12-12 21:42:00,376 INFO [train.py:421] (4/8) Epoch 8, batch 20000, loss[loss=2.324, over 1680.00 frames. , ppl: 10.220548544128235] tot_loss[loss=2.276, over 5504488.86 frames. , ppl: 9.740329665655814], batch size: 70
+2022-12-12 21:42:00,377 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:42:01,125 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.70147643313627
+2022-12-12 21:43:41,756 INFO [train.py:421] (4/8) Epoch 8, batch 20200, loss[loss=2.373, over 1470.00 frames. , ppl: 10.732873030827863] tot_loss[loss=2.276, over 5512706.76 frames. , ppl: 9.738141335103354], batch size: 70
+2022-12-12 21:45:25,099 INFO [train.py:421] (4/8) Epoch 8, batch 20400, loss[loss=2.595, over 700.00 frames. , ppl: 13.396984517799025] tot_loss[loss=2.276, over 5521875.32 frames. , ppl: 9.736590150222053], batch size: 70
+2022-12-12 21:47:09,833 INFO [train.py:421] (4/8) Epoch 8, batch 20600, loss[loss=2.366, over 1470.00 frames. , ppl: 10.657431481223352] tot_loss[loss=2.276, over 5514590.11 frames. , ppl: 9.737482163906444], batch size: 70
+2022-12-12 21:48:48,198 INFO [train.py:421] (4/8) Epoch 8, batch 20800, loss[loss=2.986, over 560.00 frames. , ppl: 19.806632288993523] tot_loss[loss=2.276, over 5515166.17 frames. , ppl: 9.741797118576585], batch size: 70
+2022-12-12 21:50:31,029 INFO [train.py:421] (4/8) Epoch 8, batch 21000, loss[loss=2.29, over 3010.00 frames. , ppl: 9.873700764305283] tot_loss[loss=2.276, over 5557956.95 frames. , ppl: 9.732876805693158], batch size: 70
+2022-12-12 21:50:31,030 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:50:31,777 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.699197941840623
+2022-12-12 21:52:13,986 INFO [train.py:421] (4/8) Epoch 8, batch 21200, loss[loss=2.209, over 6230.00 frames. , ppl: 9.110594928601719] tot_loss[loss=2.276, over 5576784.22 frames. , ppl: 9.735090649575737], batch size: 70
+2022-12-12 21:53:55,495 INFO [train.py:421] (4/8) Epoch 8, batch 21400, loss[loss=2.694, over 770.00 frames. , ppl: 14.786132636533157] tot_loss[loss=2.275, over 5624066.43 frames. , ppl: 9.723394579163172], batch size: 70
+2022-12-12 21:55:37,177 INFO [train.py:421] (4/8) Epoch 8, batch 21600, loss[loss=2.35, over 2660.00 frames. , ppl: 10.486990306450375] tot_loss[loss=2.274, over 5640542.11 frames. , ppl: 9.722403830308503], batch size: 70
+2022-12-12 21:57:17,898 INFO [train.py:421] (4/8) Epoch 8, batch 21800, loss[loss=2.172, over 6720.00 frames. , ppl: 8.774945659093918] tot_loss[loss=2.275, over 5612977.36 frames. , ppl: 9.72405120450265], batch size: 70
+2022-12-12 21:59:00,465 INFO [train.py:421] (4/8) Epoch 8, batch 22000, loss[loss=2.181, over 4130.00 frames. , ppl: 8.852080146162363] tot_loss[loss=2.274, over 5622776.72 frames. , ppl: 9.719812615014963], batch size: 70
+2022-12-12 21:59:00,466 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 21:59:01,248 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.689339266199932
+2022-12-12 22:00:41,755 INFO [train.py:421] (4/8) Epoch 8, batch 22200, loss[loss=2.216, over 2660.00 frames. , ppl: 9.168346766140568] tot_loss[loss=2.274, over 5607411.69 frames. , ppl: 9.720062969510684], batch size: 70
+2022-12-12 22:02:23,779 INFO [train.py:421] (4/8) Epoch 8, batch 22400, loss[loss=2.6, over 980.00 frames. , ppl: 13.465495712783873] tot_loss[loss=2.274, over 5581982.19 frames. , ppl: 9.72164986758379], batch size: 70
+2022-12-12 22:04:05,302 INFO [train.py:421] (4/8) Epoch 8, batch 22600, loss[loss=2.316, over 2240.00 frames. , ppl: 10.135993389883525] tot_loss[loss=2.275, over 5553424.55 frames. , ppl: 9.728886549350191], batch size: 70
+2022-12-12 22:05:46,665 INFO [train.py:421] (4/8) Epoch 8, batch 22800, loss[loss=2.52, over 1330.00 frames. , ppl: 12.428926108982573] tot_loss[loss=2.277, over 5507898.29 frames. , ppl: 9.746404433840976], batch size: 70
+2022-12-12 22:07:29,981 INFO [train.py:421] (4/8) Epoch 8, batch 23000, loss[loss=2.709, over 770.00 frames. , ppl: 15.019828005088023] tot_loss[loss=2.277, over 5512223.44 frames. , ppl: 9.751255601036705], batch size: 70
+2022-12-12 22:07:29,981 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:07:30,755 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.695581021092998
+2022-12-12 22:09:13,794 INFO [train.py:421] (4/8) Epoch 8, batch 23200, loss[loss=2.212, over 4480.00 frames. , ppl: 9.135634240927619] tot_loss[loss=2.275, over 5585316.04 frames. , ppl: 9.724968655810525], batch size: 70
+2022-12-12 22:10:53,393 INFO [train.py:421] (4/8) Epoch 8, batch 23400, loss[loss=2.281, over 4270.00 frames. , ppl: 9.789755187004573] tot_loss[loss=2.275, over 5547204.00 frames. , ppl: 9.732783579542966], batch size: 70
+2022-12-12 22:12:32,785 INFO [train.py:421] (4/8) Epoch 8, batch 23600, loss[loss=2.229, over 2940.00 frames. , ppl: 9.289505514217494] tot_loss[loss=2.275, over 5533894.33 frames. , ppl: 9.732780119850515], batch size: 70
+2022-12-12 22:14:11,811 INFO [train.py:421] (4/8) Epoch 8, batch 23800, loss[loss=2.408, over 1190.00 frames. , ppl: 11.106794915354827] tot_loss[loss=2.275, over 5530222.87 frames. , ppl: 9.730140898065489], batch size: 70
+2022-12-12 22:15:55,040 INFO [train.py:421] (4/8) Epoch 8, batch 24000, loss[loss=2.371, over 1400.00 frames. , ppl: 10.704106377098707] tot_loss[loss=2.274, over 5571550.16 frames. , ppl: 9.717462062113462], batch size: 70
+2022-12-12 22:15:55,041 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:15:55,805 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.708671477552679
+2022-12-12 22:17:36,361 INFO [train.py:421] (4/8) Epoch 8, batch 24200, loss[loss=2.214, over 4060.00 frames. , ppl: 9.150176230258822] tot_loss[loss=2.274, over 5569385.19 frames. , ppl: 9.717713290119558], batch size: 70
+2022-12-12 22:19:17,608 INFO [train.py:421] (4/8) Epoch 8, batch 24400, loss[loss=2.341, over 1680.00 frames. , ppl: 10.392782110774508] tot_loss[loss=2.274, over 5544747.98 frames. , ppl: 9.717752149484653], batch size: 70
+2022-12-12 22:20:57,978 INFO [train.py:421] (4/8) Epoch 8, batch 24600, loss[loss=2.937, over 560.00 frames. , ppl: 18.866711521624914] tot_loss[loss=2.275, over 5497108.66 frames. , ppl: 9.729082794589283], batch size: 70
+2022-12-12 22:22:37,693 INFO [train.py:421] (4/8) Epoch 8, batch 24800, loss[loss=2.265, over 3640.00 frames. , ppl: 9.627842574198118] tot_loss[loss=2.277, over 5440505.80 frames. , ppl: 9.744698303571708], batch size: 70
+2022-12-12 22:24:20,891 INFO [train.py:421] (4/8) Epoch 8, batch 25000, loss[loss=2.2, over 5600.00 frames. , ppl: 9.02322579716589] tot_loss[loss=2.278, over 5434683.63 frames. , ppl: 9.756950387022535], batch size: 70
+2022-12-12 22:24:20,891 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:24:21,639 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.690235616742381
+2022-12-12 22:26:02,834 INFO [train.py:421] (4/8) Epoch 8, batch 25200, loss[loss=2.389, over 700.00 frames. , ppl: 10.899066099122107] tot_loss[loss=2.277, over 5465416.36 frames. , ppl: 9.748646305672148], batch size: 70
+2022-12-12 22:27:41,827 INFO [train.py:421] (4/8) Epoch 8, batch 25400, loss[loss=2.426, over 2100.00 frames. , ppl: 11.311473022022712] tot_loss[loss=2.278, over 5439700.20 frames. , ppl: 9.75235542493014], batch size: 70
+2022-12-12 22:29:21,120 INFO [train.py:421] (4/8) Epoch 8, batch 25600, loss[loss=2.781, over 770.00 frames. , ppl: 16.128981886593632] tot_loss[loss=2.279, over 5399810.11 frames. , ppl: 9.766422345714028], batch size: 70
+2022-12-12 22:31:04,405 INFO [train.py:421] (4/8) Epoch 8, batch 25800, loss[loss=2.287, over 2030.00 frames. , ppl: 9.845666827730247] tot_loss[loss=2.278, over 5420352.10 frames. , ppl: 9.758917358910123], batch size: 70
+2022-12-12 22:32:42,674 INFO [train.py:421] (4/8) Epoch 8, batch 26000, loss[loss=2.188, over 9660.00 frames. , ppl: 8.92119636757486] tot_loss[loss=2.278, over 5447926.74 frames. , ppl: 9.752685415817172], batch size: 70
+2022-12-12 22:32:42,674 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:32:43,419 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.708037800704728
+2022-12-12 22:34:22,025 INFO [train.py:421] (4/8) Epoch 8, batch 26200, loss[loss=2.191, over 5880.00 frames. , ppl: 8.94366272628352] tot_loss[loss=2.278, over 5416121.46 frames. , ppl: 9.76057371840547], batch size: 70
+2022-12-12 22:36:01,414 INFO [train.py:421] (4/8) Epoch 8, batch 26400, loss[loss=2.149, over 3850.00 frames. , ppl: 8.576605015450577] tot_loss[loss=2.277, over 5450223.59 frames. , ppl: 9.751332271155713], batch size: 70
+2022-12-12 22:37:43,186 INFO [train.py:421] (4/8) Epoch 8, batch 26600, loss[loss=2.185, over 6860.00 frames. , ppl: 8.888309456578083] tot_loss[loss=2.277, over 5469786.55 frames. , ppl: 9.75210866237557], batch size: 70
+2022-12-12 22:39:21,936 INFO [train.py:421] (4/8) Epoch 8, batch 26800, loss[loss=2.277, over 2520.00 frames. , ppl: 9.747267929852617] tot_loss[loss=2.277, over 5458554.79 frames. , ppl: 9.749225986317452], batch size: 70
+2022-12-12 22:40:59,519 INFO [train.py:421] (4/8) Epoch 8, batch 27000, loss[loss=2.501, over 910.00 frames. , ppl: 12.199017099737615] tot_loss[loss=2.278, over 5444209.53 frames. , ppl: 9.752861076614387], batch size: 70
+2022-12-12 22:40:59,520 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:41:00,266 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.710776846115149
+2022-12-12 22:42:40,999 INFO [train.py:421] (4/8) Epoch 8, batch 27200, loss[loss=2.645, over 700.00 frames. , ppl: 14.082654286877661] tot_loss[loss=2.277, over 5480491.55 frames. , ppl: 9.74776431703382], batch size: 70
+2022-12-12 22:44:23,273 INFO [train.py:421] (4/8) Epoch 8, batch 27400, loss[loss=2.263, over 2520.00 frames. , ppl: 9.608774690947833] tot_loss[loss=2.277, over 5515834.81 frames. , ppl: 9.744161493266548], batch size: 70
+2022-12-12 22:46:01,804 INFO [train.py:421] (4/8) Epoch 8, batch 27600, loss[loss=2.911, over 630.00 frames. , ppl: 18.37396629250742] tot_loss[loss=2.279, over 5451384.50 frames. , ppl: 9.762959950208147], batch size: 70
+2022-12-12 22:47:40,352 INFO [train.py:421] (4/8) Epoch 8, batch 27800, loss[loss=2.735, over 840.00 frames. , ppl: 15.416111732973574] tot_loss[loss=2.28, over 5417169.82 frames. , ppl: 9.777977425867045], batch size: 70
+2022-12-12 22:49:20,551 INFO [train.py:421] (4/8) Epoch 8, batch 28000, loss[loss=2.485, over 1190.00 frames. , ppl: 11.997288773052125] tot_loss[loss=2.278, over 5455320.60 frames. , ppl: 9.760722528095283], batch size: 70
+2022-12-12 22:49:20,552 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:49:21,297 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.709539437680133
+2022-12-12 22:51:00,491 INFO [train.py:421] (4/8) Epoch 8, batch 28200, loss[loss=2.365, over 2100.00 frames. , ppl: 10.644402633648056] tot_loss[loss=2.277, over 5501335.16 frames. , ppl: 9.74731565508925], batch size: 70
+2022-12-12 22:52:39,726 INFO [train.py:421] (4/8) Epoch 8, batch 28400, loss[loss=2.339, over 2380.00 frames. , ppl: 10.36871313454454] tot_loss[loss=2.277, over 5481356.37 frames. , ppl: 9.744785603374943], batch size: 70
+2022-12-12 22:54:17,935 INFO [train.py:421] (4/8) Epoch 8, batch 28600, loss[loss=2.209, over 3710.00 frames. , ppl: 9.10597395971038] tot_loss[loss=2.276, over 5507770.55 frames. , ppl: 9.734138629526162], batch size: 70
+2022-12-12 22:55:59,819 INFO [train.py:421] (4/8) Epoch 8, batch 28800, loss[loss=2.152, over 4690.00 frames. , ppl: 8.598941598978952] tot_loss[loss=2.277, over 5471443.20 frames. , ppl: 9.748098608437424], batch size: 70
+2022-12-12 22:57:42,746 INFO [train.py:421] (4/8) Epoch 8, batch 29000, loss[loss=2.297, over 2590.00 frames. , ppl: 9.940748895552522] tot_loss[loss=2.276, over 5541626.44 frames. , ppl: 9.733112627313973], batch size: 70
+2022-12-12 22:57:42,747 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 22:57:43,506 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.696305732449282
+2022-12-12 22:59:22,325 INFO [train.py:421] (4/8) Epoch 8, batch 29200, loss[loss=2.187, over 3920.00 frames. , ppl: 8.906554870635386] tot_loss[loss=2.275, over 5551493.39 frames. , ppl: 9.729339625681574], batch size: 70
+2022-12-12 23:01:03,745 INFO [train.py:421] (4/8) Epoch 8, batch 29400, loss[loss=2.304, over 2100.00 frames. , ppl: 10.01518085662897] tot_loss[loss=2.276, over 5539469.85 frames. , ppl: 9.738255586474141], batch size: 70
+2022-12-12 23:02:44,759 INFO [train.py:421] (4/8) Epoch 8, batch 29600, loss[loss=2.291, over 1890.00 frames. , ppl: 9.889255628419244] tot_loss[loss=2.276, over 5520431.01 frames. , ppl: 9.736627538039306], batch size: 70
+2022-12-12 23:04:24,090 INFO [train.py:421] (4/8) Epoch 8, batch 29800, loss[loss=2.397, over 1890.00 frames. , ppl: 10.99107115493802] tot_loss[loss=2.278, over 5450483.40 frames. , ppl: 9.755694926768276], batch size: 70
+2022-12-12 23:06:04,137 INFO [train.py:421] (4/8) Epoch 8, batch 30000, loss[loss=2.372, over 1750.00 frames. , ppl: 10.718338941999743] tot_loss[loss=2.277, over 5477227.56 frames. , ppl: 9.747691010758682], batch size: 70
+2022-12-12 23:06:04,137 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:06:04,900 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.273, over 211138.00 frames. , ppl: 9.705356986317165
+2022-12-12 23:07:46,652 INFO [train.py:421] (4/8) Epoch 8, batch 30200, loss[loss=2.639, over 770.00 frames. , ppl: 13.993386714197575] tot_loss[loss=2.277, over 5475525.53 frames. , ppl: 9.744943201699769], batch size: 70
+2022-12-12 23:09:28,046 INFO [train.py:421] (4/8) Epoch 8, batch 30400, loss[loss=2.535, over 770.00 frames. , ppl: 12.615166036163407] tot_loss[loss=2.279, over 5419450.13 frames. , ppl: 9.765583768230671], batch size: 70
+2022-12-12 23:11:07,278 INFO [train.py:421] (4/8) Epoch 8, batch 30600, loss[loss=2.516, over 1190.00 frames. , ppl: 12.375919511297745] tot_loss[loss=2.279, over 5422493.14 frames. , ppl: 9.769207074909021], batch size: 70
+2022-12-12 23:12:43,210 INFO [train.py:421] (4/8) Epoch 8, batch 30800, loss[loss=2.356, over 2240.00 frames. , ppl: 10.547048714182322] tot_loss[loss=2.278, over 5435805.70 frames. , ppl: 9.75699284671254], batch size: 70
+2022-12-12 23:14:23,185 INFO [train.py:421] (4/8) Epoch 8, batch 31000, loss[loss=2.069, over 6580.00 frames. , ppl: 7.919672772687001] tot_loss[loss=2.277, over 5454200.23 frames. , ppl: 9.749439990004557], batch size: 70
+2022-12-12 23:14:23,185 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:14:23,945 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.697824214248042
+2022-12-12 23:16:06,465 INFO [train.py:421] (4/8) Epoch 8, batch 31200, loss[loss=2.374, over 1260.00 frames. , ppl: 10.741928277646391] tot_loss[loss=2.276, over 5516181.48 frames. , ppl: 9.734557094479584], batch size: 70
+2022-12-12 23:17:46,239 INFO [train.py:421] (4/8) Epoch 8, batch 31400, loss[loss=2.219, over 4270.00 frames. , ppl: 9.201738440986517] tot_loss[loss=2.276, over 5517760.18 frames. , ppl: 9.73946207701717], batch size: 70
+2022-12-12 23:19:24,625 INFO [train.py:421] (4/8) Epoch 8, batch 31600, loss[loss=2.254, over 3290.00 frames. , ppl: 9.52974802807737] tot_loss[loss=2.277, over 5509119.33 frames. , ppl: 9.7479845374253], batch size: 70
+2022-12-12 23:21:02,599 INFO [train.py:421] (4/8) Epoch 8, batch 31800, loss[loss=2.196, over 5670.00 frames. , ppl: 8.991830542273084] tot_loss[loss=2.278, over 5479255.35 frames. , ppl: 9.75363885984196], batch size: 70
+2022-12-12 23:22:42,623 INFO [train.py:421] (4/8) Epoch 8, batch 32000, loss[loss=2.399, over 1610.00 frames. , ppl: 11.015433111023766] tot_loss[loss=2.277, over 5509025.32 frames. , ppl: 9.750201917191276], batch size: 70
+2022-12-12 23:22:42,624 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:22:43,385 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.681927833421199
+2022-12-12 23:24:20,872 INFO [train.py:421] (4/8) Epoch 8, batch 32200, loss[loss=2.382, over 1820.00 frames. , ppl: 10.830701703454004] tot_loss[loss=2.278, over 5497276.79 frames. , ppl: 9.753431658776163], batch size: 70
+2022-12-12 23:26:03,836 INFO [train.py:421] (4/8) Epoch 8, batch 32400, loss[loss=2.196, over 5040.00 frames. , ppl: 8.989856176567274] tot_loss[loss=2.278, over 5497977.33 frames. , ppl: 9.756746275933006], batch size: 70
+2022-12-12 23:27:44,838 INFO [train.py:421] (4/8) Epoch 8, batch 32600, loss[loss=2.355, over 1470.00 frames. , ppl: 10.536376013053205] tot_loss[loss=2.278, over 5474483.53 frames. , ppl: 9.755953026363958], batch size: 70
+2022-12-12 23:29:26,644 INFO [train.py:421] (4/8) Epoch 8, batch 32800, loss[loss=2.268, over 2590.00 frames. , ppl: 9.66038596890314] tot_loss[loss=2.278, over 5483305.71 frames. , ppl: 9.75529984377207], batch size: 70
+2022-12-12 23:31:04,207 INFO [train.py:421] (4/8) Epoch 8, batch 33000, loss[loss=2.191, over 5810.00 frames. , ppl: 8.94077382090885] tot_loss[loss=2.278, over 5462841.01 frames. , ppl: 9.759909434661349], batch size: 70
+2022-12-12 23:31:04,207 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:31:04,954 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.272, over 211138.00 frames. , ppl: 9.70102126639035
+2022-12-12 23:32:46,895 INFO [train.py:421] (4/8) Epoch 8, batch 33200, loss[loss=2.161, over 5600.00 frames. , ppl: 8.677142878798803] tot_loss[loss=2.278, over 5460518.91 frames. , ppl: 9.760547150528577], batch size: 70
+2022-12-12 23:34:29,611 INFO [train.py:421] (4/8) Epoch 8, batch 33400, loss[loss=2.106, over 6650.00 frames. , ppl: 8.21314702904822] tot_loss[loss=2.277, over 5515877.07 frames. , ppl: 9.742525077472898], batch size: 70
+2022-12-12 23:36:09,713 INFO [train.py:421] (4/8) Epoch 8, batch 33600, loss[loss=2.278, over 3780.00 frames. , ppl: 9.752982782972728] tot_loss[loss=2.276, over 5523488.20 frames. , ppl: 9.740176498268433], batch size: 70
+2022-12-12 23:37:51,013 INFO [train.py:421] (4/8) Epoch 8, batch 33800, loss[loss=3.109, over 560.00 frames. , ppl: 22.39335874840977] tot_loss[loss=2.276, over 5555340.81 frames. , ppl: 9.738374879291177], batch size: 70
+2022-12-12 23:39:28,192 INFO [train.py:421] (4/8) Epoch 8, batch 34000, loss[loss=2.235, over 3640.00 frames. , ppl: 9.349196401099766] tot_loss[loss=2.275, over 5583103.93 frames. , ppl: 9.72882048360033], batch size: 70
+2022-12-12 23:39:28,192 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:39:28,937 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.680815891149308
+2022-12-12 23:41:01,892 INFO [train.py:421] (4/8) Epoch 8, batch 34200, loss[loss=2.285, over 2800.00 frames. , ppl: 9.82207661918993] tot_loss[loss=2.276, over 5536655.13 frames. , ppl: 9.73810727992978], batch size: 70
+2022-12-12 23:42:44,182 INFO [train.py:421] (4/8) Epoch 8, batch 34400, loss[loss=2.435, over 1610.00 frames. , ppl: 11.41718012870535] tot_loss[loss=2.277, over 5510881.11 frames. , ppl: 9.743150393947538], batch size: 70
+2022-12-12 23:44:28,311 INFO [train.py:421] (4/8) Epoch 8, batch 34600, loss[loss=2.265, over 2380.00 frames. , ppl: 9.633453310663024] tot_loss[loss=2.277, over 5506970.41 frames. , ppl: 9.746376203776585], batch size: 70
+2022-12-12 23:46:06,947 INFO [train.py:421] (4/8) Epoch 8, batch 34800, loss[loss=2.27, over 2660.00 frames. , ppl: 9.68392985312914] tot_loss[loss=2.278, over 5468853.69 frames. , ppl: 9.758276655696587], batch size: 70
+2022-12-12 23:47:51,625 INFO [train.py:421] (4/8) Epoch 8, batch 35000, loss[loss=2.26, over 3080.00 frames. , ppl: 9.581165401108999] tot_loss[loss=2.277, over 5511885.03 frames. , ppl: 9.744483192248609], batch size: 70
+2022-12-12 23:47:51,625 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:47:52,385 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.680259967903076
+2022-12-12 23:49:30,413 INFO [train.py:421] (4/8) Epoch 8, batch 35200, loss[loss=2.5, over 1610.00 frames. , ppl: 12.178962329775777] tot_loss[loss=2.277, over 5481088.85 frames. , ppl: 9.750602146211829], batch size: 70
+2022-12-12 23:51:13,489 INFO [train.py:421] (4/8) Epoch 8, batch 35400, loss[loss=2.234, over 4410.00 frames. , ppl: 9.334354186176665] tot_loss[loss=2.278, over 5475406.86 frames. , ppl: 9.75258702012674], batch size: 70
+2022-12-12 23:52:48,344 INFO [train.py:421] (4/8) Epoch 8, batch 35600, loss[loss=2.34, over 2450.00 frames. , ppl: 10.383351259866068] tot_loss[loss=2.279, over 5448365.77 frames. , ppl: 9.763501301175948], batch size: 70
+2022-12-12 23:54:26,509 INFO [train.py:421] (4/8) Epoch 8, batch 35800, loss[loss=2.392, over 1400.00 frames. , ppl: 10.937730944105327] tot_loss[loss=2.279, over 5432883.85 frames. , ppl: 9.77093630542166], batch size: 70
+2022-12-12 23:56:06,756 INFO [train.py:421] (4/8) Epoch 8, batch 36000, loss[loss=2.351, over 1960.00 frames. , ppl: 10.492379321608382] tot_loss[loss=2.278, over 5476881.72 frames. , ppl: 9.752843444117117], batch size: 70
+2022-12-12 23:56:06,757 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-12 23:56:07,522 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675932595464811
+2022-12-12 23:57:48,184 INFO [train.py:421] (4/8) Epoch 8, batch 36200, loss[loss=2.337, over 2310.00 frames. , ppl: 10.349674497189406] tot_loss[loss=2.277, over 5481078.89 frames. , ppl: 9.746174624206564], batch size: 70
+2022-12-12 23:59:26,453 INFO [train.py:421] (4/8) Epoch 8, batch 36400, loss[loss=2.855, over 560.00 frames. , ppl: 17.36904261962567] tot_loss[loss=2.277, over 5509923.78 frames. , ppl: 9.743088236072595], batch size: 70
+2022-12-13 00:01:08,130 INFO [train.py:421] (4/8) Epoch 8, batch 36600, loss[loss=2.609, over 770.00 frames. , ppl: 13.587438978915968] tot_loss[loss=2.277, over 5488454.36 frames. , ppl: 9.747809701129924], batch size: 70
+2022-12-13 00:02:44,810 INFO [train.py:421] (4/8) Epoch 8, batch 36800, loss[loss=2.98, over 560.00 frames. , ppl: 19.696214528305056] tot_loss[loss=2.278, over 5456491.33 frames. , ppl: 9.758066486799912], batch size: 70
+2022-12-13 00:04:22,094 INFO [train.py:421] (4/8) Epoch 8, batch 37000, loss[loss=2.862, over 630.00 frames. , ppl: 17.500141539601824] tot_loss[loss=2.279, over 5442214.40 frames. , ppl: 9.765117276430301], batch size: 70
+2022-12-13 00:04:22,095 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:04:22,851 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.678712720778323
+2022-12-13 00:06:04,170 INFO [train.py:421] (4/8) Epoch 8, batch 37200, loss[loss=3.217, over 490.00 frames. , ppl: 24.960774791438485] tot_loss[loss=2.278, over 5455312.60 frames. , ppl: 9.760787501398722], batch size: 70
+2022-12-13 00:07:45,618 INFO [train.py:421] (4/8) Epoch 8, batch 37400, loss[loss=2.228, over 3010.00 frames. , ppl: 9.283467659130249] tot_loss[loss=2.278, over 5445715.82 frames. , ppl: 9.761601904491307], batch size: 70
+2022-12-13 00:09:28,550 INFO [train.py:421] (4/8) Epoch 8, batch 37600, loss[loss=2.205, over 3850.00 frames. , ppl: 9.066164157877845] tot_loss[loss=2.28, over 5409312.35 frames. , ppl: 9.773506050434602], batch size: 70
+2022-12-13 00:11:11,010 INFO [train.py:421] (4/8) Epoch 8, batch 37800, loss[loss=2.227, over 2450.00 frames. , ppl: 9.270583300124164] tot_loss[loss=2.278, over 5465845.77 frames. , ppl: 9.75932642336838], batch size: 70
+2022-12-13 00:12:52,620 INFO [train.py:421] (4/8) Epoch 8, batch 38000, loss[loss=2.177, over 5250.00 frames. , ppl: 8.818662051675249] tot_loss[loss=2.279, over 5446159.19 frames. , ppl: 9.766201038933394], batch size: 70
+2022-12-13 00:12:52,620 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:12:53,370 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.666419470350528
+2022-12-13 00:14:30,994 INFO [train.py:421] (4/8) Epoch 8, batch 38200, loss[loss=2.255, over 3780.00 frames. , ppl: 9.533343449034879] tot_loss[loss=2.279, over 5428238.10 frames. , ppl: 9.768807926182772], batch size: 70
+2022-12-13 00:16:13,250 INFO [train.py:421] (4/8) Epoch 8, batch 38400, loss[loss=2.194, over 5110.00 frames. , ppl: 8.969223578891288] tot_loss[loss=2.278, over 5459514.80 frames. , ppl: 9.76099613962873], batch size: 70
+2022-12-13 00:17:51,819 INFO [train.py:421] (4/8) Epoch 8, batch 38600, loss[loss=2.317, over 3360.00 frames. , ppl: 10.146060317657048] tot_loss[loss=2.279, over 5442089.19 frames. , ppl: 9.766262614454591], batch size: 70
+2022-12-13 00:19:32,517 INFO [train.py:421] (4/8) Epoch 8, batch 38800, loss[loss=2.604, over 980.00 frames. , ppl: 13.511801066006548] tot_loss[loss=2.278, over 5493744.54 frames. , ppl: 9.753187773622516], batch size: 70
+2022-12-13 00:21:14,870 INFO [train.py:421] (4/8) Epoch 8, batch 39000, loss[loss=2.326, over 1540.00 frames. , ppl: 10.232158185586577] tot_loss[loss=2.277, over 5468826.36 frames. , ppl: 9.750852593412082], batch size: 70
+2022-12-13 00:21:14,870 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:21:15,629 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.679752787430928
+2022-12-13 00:22:54,873 INFO [train.py:421] (4/8) Epoch 8, batch 39200, loss[loss=2.205, over 5460.00 frames. , ppl: 9.07128469678062] tot_loss[loss=2.275, over 5512144.06 frames. , ppl: 9.729461951454317], batch size: 70
+2022-12-13 00:24:36,587 INFO [train.py:421] (4/8) Epoch 8, batch 39400, loss[loss=2.863, over 630.00 frames. , ppl: 17.510564784245695] tot_loss[loss=2.276, over 5479491.67 frames. , ppl: 9.740222889121721], batch size: 70
+2022-12-13 00:26:21,910 INFO [train.py:421] (4/8) Epoch 8, batch 39600, loss[loss=2.373, over 1680.00 frames. , ppl: 10.732675839413004] tot_loss[loss=2.278, over 5436391.04 frames. , ppl: 9.754676746754216], batch size: 70
+2022-12-13 00:28:01,276 INFO [train.py:421] (4/8) Epoch 8, batch 39800, loss[loss=2.186, over 3710.00 frames. , ppl: 8.89779639840054] tot_loss[loss=2.279, over 5406161.05 frames. , ppl: 9.766706059813307], batch size: 70
+2022-12-13 00:29:43,020 INFO [train.py:421] (4/8) Epoch 8, batch 40000, loss[loss=2.22, over 5320.00 frames. , ppl: 9.21121629859626] tot_loss[loss=2.28, over 5392049.13 frames. , ppl: 9.775217486087715], batch size: 70
+2022-12-13 00:29:43,020 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:29:43,770 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.689682020856184
+2022-12-13 00:31:23,419 INFO [train.py:421] (4/8) Epoch 8, batch 40200, loss[loss=2.181, over 3920.00 frames. , ppl: 8.856645121973084] tot_loss[loss=2.281, over 5354169.82 frames. , ppl: 9.782705915351578], batch size: 70
+2022-12-13 00:33:03,369 INFO [train.py:421] (4/8) Epoch 8, batch 40400, loss[loss=2.269, over 3710.00 frames. , ppl: 9.66880295956063] tot_loss[loss=2.279, over 5397057.72 frames. , ppl: 9.769439409092701], batch size: 70
+2022-12-13 00:34:47,610 INFO [train.py:421] (4/8) Epoch 8, batch 40600, loss[loss=2.272, over 3010.00 frames. , ppl: 9.701462583417646] tot_loss[loss=2.28, over 5406152.32 frames. , ppl: 9.772810132024809], batch size: 70
+2022-12-13 00:36:29,478 INFO [train.py:421] (4/8) Epoch 8, batch 40800, loss[loss=2.339, over 2380.00 frames. , ppl: 10.370198059372678] tot_loss[loss=2.279, over 5434242.27 frames. , ppl: 9.76600983917977], batch size: 70
+2022-12-13 00:38:06,221 INFO [train.py:421] (4/8) Epoch 8, batch 41000, loss[loss=2.219, over 3010.00 frames. , ppl: 9.202233689156945] tot_loss[loss=2.28, over 5386766.27 frames. , ppl: 9.777164303012107], batch size: 70
+2022-12-13 00:38:06,222 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:38:06,968 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.67065241409767
+2022-12-13 00:39:49,692 INFO [train.py:421] (4/8) Epoch 8, batch 41200, loss[loss=2.257, over 2800.00 frames. , ppl: 9.556247919564811] tot_loss[loss=2.28, over 5407834.24 frames. , ppl: 9.773709728610761], batch size: 70
+2022-12-13 00:41:29,186 INFO [train.py:421] (4/8) Epoch 8, batch 41400, loss[loss=2.214, over 2800.00 frames. , ppl: 9.153378187761035] tot_loss[loss=2.279, over 5410168.78 frames. , ppl: 9.769225461409865], batch size: 70
+2022-12-13 00:43:12,258 INFO [train.py:421] (4/8) Epoch 8, batch 41600, loss[loss=2.246, over 3780.00 frames. , ppl: 9.448713026508264] tot_loss[loss=2.278, over 5428785.52 frames. , ppl: 9.76131939083255], batch size: 70
+2022-12-13 00:44:46,373 INFO [train.py:421] (4/8) Epoch 8, batch 41800, loss[loss=2.192, over 4900.00 frames. , ppl: 8.953909408615468] tot_loss[loss=2.28, over 5407217.74 frames. , ppl: 9.77243819920658], batch size: 70
+2022-12-13 00:46:26,562 INFO [train.py:421] (4/8) Epoch 8, batch 42000, loss[loss=2.662, over 700.00 frames. , ppl: 14.320684683646617] tot_loss[loss=2.28, over 5386328.82 frames. , ppl: 9.775570663454669], batch size: 70
+2022-12-13 00:46:26,563 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:46:27,308 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675724941707049
+2022-12-13 00:48:05,873 INFO [train.py:421] (4/8) Epoch 8, batch 42200, loss[loss=2.411, over 2100.00 frames. , ppl: 11.142201791832393] tot_loss[loss=2.281, over 5335910.16 frames. , ppl: 9.78670096741537], batch size: 70
+2022-12-13 00:49:47,952 INFO [train.py:421] (4/8) Epoch 8, batch 42400, loss[loss=2.419, over 980.00 frames. , ppl: 11.231869533230679] tot_loss[loss=2.28, over 5385370.76 frames. , ppl: 9.771911719786155], batch size: 70
+2022-12-13 00:51:28,040 INFO [train.py:421] (4/8) Epoch 8, batch 42600, loss[loss=2.198, over 5530.00 frames. , ppl: 9.004897207514478] tot_loss[loss=2.279, over 5378503.35 frames. , ppl: 9.771499310970443], batch size: 70
+2022-12-13 00:53:10,398 INFO [train.py:421] (4/8) Epoch 8, batch 42800, loss[loss=2.635, over 910.00 frames. , ppl: 13.937943203131669] tot_loss[loss=2.279, over 5383267.89 frames. , ppl: 9.770430881482206], batch size: 70
+2022-12-13 00:54:50,419 INFO [train.py:421] (4/8) Epoch 8, batch 43000, loss[loss=2.667, over 700.00 frames. , ppl: 14.396365734796953] tot_loss[loss=2.278, over 5399328.60 frames. , ppl: 9.761428636568702], batch size: 70
+2022-12-13 00:54:50,420 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 00:54:51,176 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675021816214633
+2022-12-13 00:56:31,399 INFO [train.py:421] (4/8) Epoch 8, batch 43200, loss[loss=2.494, over 1610.00 frames. , ppl: 12.11112987768643] tot_loss[loss=2.278, over 5390908.94 frames. , ppl: 9.761029046448899], batch size: 70
+2022-12-13 00:58:13,083 INFO [train.py:421] (4/8) Epoch 8, batch 43400, loss[loss=2.394, over 1610.00 frames. , ppl: 10.960411636588947] tot_loss[loss=2.278, over 5423852.02 frames. , ppl: 9.752828933125356], batch size: 70
+2022-12-13 00:59:53,237 INFO [train.py:421] (4/8) Epoch 8, batch 43600, loss[loss=2.335, over 2240.00 frames. , ppl: 10.331326260148176] tot_loss[loss=2.279, over 5378312.38 frames. , ppl: 9.767444274194537], batch size: 70
+2022-12-13 01:01:34,004 INFO [train.py:421] (4/8) Epoch 8, batch 43800, loss[loss=2.282, over 3780.00 frames. , ppl: 9.796468361040333] tot_loss[loss=2.278, over 5393191.51 frames. , ppl: 9.757584770748645], batch size: 70
+2022-12-13 01:03:12,893 INFO [train.py:421] (4/8) Epoch 8, batch 44000, loss[loss=2.473, over 910.00 frames. , ppl: 11.856652738694754] tot_loss[loss=2.278, over 5403237.33 frames. , ppl: 9.759187056918623], batch size: 70
+2022-12-13 01:03:12,893 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:03:13,638 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674165532792536
+2022-12-13 01:04:54,833 INFO [train.py:421] (4/8) Epoch 8, batch 44200, loss[loss=2.346, over 1890.00 frames. , ppl: 10.439915313982858] tot_loss[loss=2.279, over 5413484.42 frames. , ppl: 9.763624958431665], batch size: 70
+2022-12-13 01:06:33,909 INFO [train.py:421] (4/8) Epoch 8, batch 44400, loss[loss=2.321, over 1820.00 frames. , ppl: 10.185924590709766] tot_loss[loss=2.28, over 5358124.86 frames. , ppl: 9.775539240322246], batch size: 70
+2022-12-13 01:08:12,922 INFO [train.py:421] (4/8) Epoch 8, batch 44600, loss[loss=2.183, over 13790.00 frames. , ppl: 8.876204784816053] tot_loss[loss=2.279, over 5410211.92 frames. , ppl: 9.767951981237651], batch size: 70
+2022-12-13 01:09:58,344 INFO [train.py:421] (4/8) Epoch 8, batch 44800, loss[loss=2.908, over 560.00 frames. , ppl: 18.32185027394669] tot_loss[loss=2.277, over 5470251.70 frames. , ppl: 9.749967787016924], batch size: 70
+2022-12-13 01:11:44,374 INFO [train.py:421] (4/8) Epoch 8, batch 45000, loss[loss=2.403, over 1190.00 frames. , ppl: 11.0554796367464] tot_loss[loss=2.276, over 5506488.14 frames. , ppl: 9.740034388048533], batch size: 70
+2022-12-13 01:11:44,374 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:11:45,133 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674410382020383
+2022-12-13 01:13:24,886 INFO [train.py:421] (4/8) Epoch 8, batch 45200, loss[loss=2.546, over 1050.00 frames. , ppl: 12.761166790906568] tot_loss[loss=2.277, over 5503555.29 frames. , ppl: 9.747677195200767], batch size: 70
+2022-12-13 01:15:00,537 INFO [train.py:421] (4/8) Epoch 8, batch 45400, loss[loss=2.311, over 2800.00 frames. , ppl: 10.080547346166288] tot_loss[loss=2.276, over 5534336.13 frames. , ppl: 9.73694550990539], batch size: 70
+2022-12-13 01:16:39,283 INFO [train.py:421] (4/8) Epoch 8, batch 45600, loss[loss=2.241, over 3290.00 frames. , ppl: 9.403056259062959] tot_loss[loss=2.277, over 5504507.04 frames. , ppl: 9.748483824380775], batch size: 70
+2022-12-13 01:18:18,207 INFO [train.py:421] (4/8) Epoch 8, batch 45800, loss[loss=2.31, over 2030.00 frames. , ppl: 10.073187915404231] tot_loss[loss=2.278, over 5475573.38 frames. , ppl: 9.753412902024913], batch size: 70
+2022-12-13 01:19:57,220 INFO [train.py:421] (4/8) Epoch 8, batch 46000, loss[loss=3.391, over 490.00 frames. , ppl: 29.686451654917754] tot_loss[loss=2.278, over 5459504.25 frames. , ppl: 9.759835452103715], batch size: 70
+2022-12-13 01:19:57,221 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:19:57,983 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.68747512264899
+2022-12-13 01:21:36,866 INFO [train.py:421] (4/8) Epoch 8, batch 46200, loss[loss=2.2, over 3080.00 frames. , ppl: 9.023632920364175] tot_loss[loss=2.279, over 5434278.25 frames. , ppl: 9.76492108263267], batch size: 70
+2022-12-13 01:23:16,620 INFO [train.py:421] (4/8) Epoch 8, batch 46400, loss[loss=2.353, over 1540.00 frames. , ppl: 10.519100433575119] tot_loss[loss=2.279, over 5420483.70 frames. , ppl: 9.770909815897154], batch size: 70
+2022-12-13 01:24:55,148 INFO [train.py:421] (4/8) Epoch 8, batch 46600, loss[loss=2.435, over 840.00 frames. , ppl: 11.41647103795828] tot_loss[loss=2.28, over 5389296.02 frames. , ppl: 9.781557983183507], batch size: 70
+2022-12-13 01:26:38,493 INFO [train.py:421] (4/8) Epoch 8, batch 46800, loss[loss=2.156, over 6930.00 frames. , ppl: 8.638314154158273] tot_loss[loss=2.279, over 5419821.26 frames. , ppl: 9.769205486786921], batch size: 70
+2022-12-13 01:28:17,302 INFO [train.py:421] (4/8) Epoch 8, batch 47000, loss[loss=2.321, over 1960.00 frames. , ppl: 10.182659093535923] tot_loss[loss=2.28, over 5421387.75 frames. , ppl: 9.775039015000655], batch size: 70
+2022-12-13 01:28:17,303 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:28:18,064 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.67573496626884
+2022-12-13 01:29:59,145 INFO [train.py:421] (4/8) Epoch 8, batch 47200, loss[loss=2.591, over 1120.00 frames. , ppl: 13.348491954087248] tot_loss[loss=2.28, over 5419947.47 frames. , ppl: 9.77549357195589], batch size: 70
+2022-12-13 01:31:41,108 INFO [train.py:421] (4/8) Epoch 8, batch 47400, loss[loss=2.213, over 6930.00 frames. , ppl: 9.146259016098435] tot_loss[loss=2.28, over 5441246.46 frames. , ppl: 9.773073248263412], batch size: 70
+2022-12-13 01:33:21,465 INFO [train.py:421] (4/8) Epoch 8, batch 47600, loss[loss=2.196, over 9590.00 frames. , ppl: 8.98665775981069] tot_loss[loss=2.279, over 5455823.89 frames. , ppl: 9.765901795526773], batch size: 70
+2022-12-13 01:35:04,432 INFO [train.py:421] (4/8) Epoch 8, batch 47800, loss[loss=2.453, over 840.00 frames. , ppl: 11.623782456115283] tot_loss[loss=2.279, over 5454901.01 frames. , ppl: 9.763277708095359], batch size: 70
+2022-12-13 01:36:41,911 INFO [train.py:421] (4/8) Epoch 8, batch 48000, loss[loss=2.266, over 1820.00 frames. , ppl: 9.64152679363016] tot_loss[loss=2.279, over 5438075.91 frames. , ppl: 9.763815519310175], batch size: 70
+2022-12-13 01:36:41,911 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:36:42,686 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664405251487315
+2022-12-13 01:38:24,558 INFO [train.py:421] (4/8) Epoch 8, batch 48200, loss[loss=2.29, over 2800.00 frames. , ppl: 9.879450489497229] tot_loss[loss=2.279, over 5424933.20 frames. , ppl: 9.76547995120184], batch size: 70
+2022-12-13 01:40:06,222 INFO [train.py:421] (4/8) Epoch 8, batch 48400, loss[loss=2.377, over 2100.00 frames. , ppl: 10.776536093988252] tot_loss[loss=2.279, over 5419145.13 frames. , ppl: 9.766004970083902], batch size: 70
+2022-12-13 01:41:49,536 INFO [train.py:421] (4/8) Epoch 8, batch 48600, loss[loss=2.286, over 3640.00 frames. , ppl: 9.831440928831606] tot_loss[loss=2.278, over 5456605.11 frames. , ppl: 9.754777325565888], batch size: 70
+2022-12-13 01:43:30,068 INFO [train.py:421] (4/8) Epoch 8, batch 48800, loss[loss=2.438, over 910.00 frames. , ppl: 11.44761831249332] tot_loss[loss=2.279, over 5417777.48 frames. , ppl: 9.766208278345275], batch size: 70
+2022-12-13 01:45:08,311 INFO [train.py:421] (4/8) Epoch 8, batch 49000, loss[loss=2.455, over 980.00 frames. , ppl: 11.644088296941723] tot_loss[loss=2.279, over 5391509.82 frames. , ppl: 9.769172608231807], batch size: 70
+2022-12-13 01:45:08,312 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:45:09,059 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.66599885304498
+2022-12-13 01:46:48,495 INFO [train.py:421] (4/8) Epoch 8, batch 49200, loss[loss=2.308, over 1540.00 frames. , ppl: 10.05029958931407] tot_loss[loss=2.279, over 5403940.53 frames. , ppl: 9.77078049117443], batch size: 70
+2022-12-13 01:48:27,316 INFO [train.py:421] (4/8) Epoch 8, batch 49400, loss[loss=2.219, over 3640.00 frames. , ppl: 9.202148112873154] tot_loss[loss=2.279, over 5447301.04 frames. , ppl: 9.762486565741096], batch size: 70
+2022-12-13 01:50:08,163 INFO [train.py:421] (4/8) Epoch 8, batch 49600, loss[loss=2.517, over 980.00 frames. , ppl: 12.393313669568597] tot_loss[loss=2.278, over 5475914.71 frames. , ppl: 9.752495876848641], batch size: 70
+2022-12-13 01:51:45,763 INFO [train.py:421] (4/8) Epoch 8, batch 49800, loss[loss=2.2, over 4830.00 frames. , ppl: 9.025747074355133] tot_loss[loss=2.277, over 5487537.99 frames. , ppl: 9.751464195635256], batch size: 70
+2022-12-13 01:53:28,038 INFO [train.py:421] (4/8) Epoch 8, batch 50000, loss[loss=2.259, over 2450.00 frames. , ppl: 9.57568545539975] tot_loss[loss=2.277, over 5514424.36 frames. , ppl: 9.743653831432797], batch size: 70
+2022-12-13 01:53:28,038 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 01:53:28,798 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.6543676004614
+2022-12-13 01:55:10,390 INFO [train.py:421] (4/8) Epoch 8, batch 50200, loss[loss=2.659, over 770.00 frames. , ppl: 14.288534566372451] tot_loss[loss=2.277, over 5513560.35 frames. , ppl: 9.744571158220062], batch size: 70
+2022-12-13 01:56:54,450 INFO [train.py:421] (4/8) Epoch 8, batch 50400, loss[loss=2.343, over 2100.00 frames. , ppl: 10.41394659296454] tot_loss[loss=2.278, over 5488077.75 frames. , ppl: 9.755579949356497], batch size: 70
+2022-12-13 01:58:36,243 INFO [train.py:421] (4/8) Epoch 8, batch 50600, loss[loss=2.428, over 1540.00 frames. , ppl: 11.339595805650807] tot_loss[loss=2.277, over 5492324.44 frames. , ppl: 9.751893062822317], batch size: 70
+2022-12-13 02:00:17,742 INFO [train.py:421] (4/8) Epoch 8, batch 50800, loss[loss=2.309, over 2450.00 frames. , ppl: 10.060293516939435] tot_loss[loss=2.277, over 5503313.24 frames. , ppl: 9.744647915621146], batch size: 70
+2022-12-13 02:01:59,729 INFO [train.py:421] (4/8) Epoch 8, batch 51000, loss[loss=2.455, over 910.00 frames. , ppl: 11.651862833915468] tot_loss[loss=2.277, over 5512765.86 frames. , ppl: 9.74514715749156], batch size: 70
+2022-12-13 02:01:59,729 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:02:00,491 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66433230115276
+2022-12-13 02:03:40,185 INFO [train.py:421] (4/8) Epoch 8, batch 51200, loss[loss=2.54, over 770.00 frames. , ppl: 12.679965259429018] tot_loss[loss=2.278, over 5476068.08 frames. , ppl: 9.754377412212301], batch size: 70
+2022-12-13 02:05:20,066 INFO [train.py:421] (4/8) Epoch 8, batch 51400, loss[loss=2.542, over 1540.00 frames. , ppl: 12.70966639428852] tot_loss[loss=2.279, over 5432440.85 frames. , ppl: 9.762590475499506], batch size: 70
+2022-12-13 02:07:01,059 INFO [train.py:421] (4/8) Epoch 8, batch 51600, loss[loss=2.4, over 2870.00 frames. , ppl: 11.021903054670487] tot_loss[loss=2.279, over 5415104.81 frames. , ppl: 9.764710833953908], batch size: 70
+2022-12-13 02:08:42,082 INFO [train.py:421] (4/8) Epoch 8, batch 51800, loss[loss=2.272, over 910.00 frames. , ppl: 9.696291954218127] tot_loss[loss=2.278, over 5426334.62 frames. , ppl: 9.761174847385323], batch size: 70
+2022-12-13 02:10:19,803 INFO [train.py:421] (4/8) Epoch 8, batch 52000, loss[loss=2.266, over 3010.00 frames. , ppl: 9.641252127102211] tot_loss[loss=2.278, over 5447471.92 frames. , ppl: 9.757280223903033], batch size: 70
+2022-12-13 02:10:19,803 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:10:20,563 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.669551784931777
+2022-12-13 02:11:58,346 INFO [train.py:421] (4/8) Epoch 8, batch 52200, loss[loss=2.636, over 980.00 frames. , ppl: 13.956578199597134] tot_loss[loss=2.279, over 5413079.78 frames. , ppl: 9.768867554286125], batch size: 70
+2022-12-13 02:13:38,891 INFO [train.py:421] (4/8) Epoch 8, batch 52400, loss[loss=2.398, over 1750.00 frames. , ppl: 11.000838973174945] tot_loss[loss=2.279, over 5436785.96 frames. , ppl: 9.76393961653259], batch size: 70
+2022-12-13 02:15:21,731 INFO [train.py:421] (4/8) Epoch 8, batch 52600, loss[loss=2.18, over 5390.00 frames. , ppl: 8.850104374392647] tot_loss[loss=2.278, over 5455545.23 frames. , ppl: 9.757727244225283], batch size: 70
+2022-12-13 02:17:02,386 INFO [train.py:421] (4/8) Epoch 8, batch 52800, loss[loss=2.101, over 8050.00 frames. , ppl: 8.172821898870255] tot_loss[loss=2.278, over 5471502.65 frames. , ppl: 9.759338188500735], batch size: 70
+2022-12-13 02:18:40,326 INFO [train.py:421] (4/8) Epoch 8, batch 53000, loss[loss=2.598, over 770.00 frames. , ppl: 13.430980067615094] tot_loss[loss=2.278, over 5478316.70 frames. , ppl: 9.76083814181423], batch size: 70
+2022-12-13 02:18:40,327 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:18:41,091 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.271, over 211138.00 frames. , ppl: 9.68564430928278
+2022-12-13 02:20:19,613 INFO [train.py:421] (4/8) Epoch 8, batch 53200, loss[loss=2.351, over 1750.00 frames. , ppl: 10.49413370475339] tot_loss[loss=2.278, over 5458360.56 frames. , ppl: 9.761262683451424], batch size: 70
+2022-12-13 02:21:58,982 INFO [train.py:421] (4/8) Epoch 8, batch 53400, loss[loss=2.811, over 840.00 frames. , ppl: 16.6221586767088] tot_loss[loss=2.279, over 5460101.20 frames. , ppl: 9.763229155418887], batch size: 70
+2022-12-13 02:23:40,432 INFO [train.py:421] (4/8) Epoch 8, batch 53600, loss[loss=2.429, over 1680.00 frames. , ppl: 11.344237664480866] tot_loss[loss=2.278, over 5484712.86 frames. , ppl: 9.75817695894504], batch size: 70
+2022-12-13 02:25:20,304 INFO [train.py:421] (4/8) Epoch 8, batch 53800, loss[loss=2.18, over 4760.00 frames. , ppl: 8.849257431332463] tot_loss[loss=2.278, over 5490047.92 frames. , ppl: 9.758225970593223], batch size: 70
+2022-12-13 02:26:59,831 INFO [train.py:421] (4/8) Epoch 8, batch 54000, loss[loss=2.307, over 1820.00 frames. , ppl: 10.046819070470198] tot_loss[loss=2.278, over 5499673.04 frames. , ppl: 9.752283943432762], batch size: 70
+2022-12-13 02:26:59,832 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:27:00,592 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66121883652427
+2022-12-13 02:28:44,769 INFO [train.py:421] (4/8) Epoch 8, batch 54200, loss[loss=2.225, over 2380.00 frames. , ppl: 9.251030336359413] tot_loss[loss=2.277, over 5523907.30 frames. , ppl: 9.747708537952562], batch size: 70
+2022-12-13 02:30:23,554 INFO [train.py:421] (4/8) Epoch 8, batch 54400, loss[loss=2.34, over 2170.00 frames. , ppl: 10.384997152969754] tot_loss[loss=2.277, over 5507365.30 frames. , ppl: 9.752167591934228], batch size: 70
+2022-12-13 02:32:02,924 INFO [train.py:421] (4/8) Epoch 8, batch 54600, loss[loss=2.312, over 1890.00 frames. , ppl: 10.097675049195416] tot_loss[loss=2.277, over 5500870.29 frames. , ppl: 9.750144781777523], batch size: 70
+2022-12-13 02:33:42,229 INFO [train.py:421] (4/8) Epoch 8, batch 54800, loss[loss=4.171, over 350.00 frames. , ppl: 64.77260442496053] tot_loss[loss=2.277, over 5494730.72 frames. , ppl: 9.752066762173296], batch size: 70
+2022-12-13 02:35:25,399 INFO [train.py:421] (4/8) Epoch 8, batch 55000, loss[loss=2.306, over 1820.00 frames. , ppl: 10.038973058318962] tot_loss[loss=2.277, over 5550512.92 frames. , ppl: 9.744916124623456], batch size: 70
+2022-12-13 02:35:25,399 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:35:26,159 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.668597244433377
+2022-12-13 02:37:08,637 INFO [train.py:421] (4/8) Epoch 8, batch 55200, loss[loss=2.465, over 1470.00 frames. , ppl: 11.757847458570957] tot_loss[loss=2.276, over 5548375.93 frames. , ppl: 9.741316911330834], batch size: 70
+2022-12-13 02:38:47,759 INFO [train.py:421] (4/8) Epoch 8, batch 55400, loss[loss=2.218, over 3640.00 frames. , ppl: 9.189473332302455] tot_loss[loss=2.278, over 5497645.55 frames. , ppl: 9.758661467916872], batch size: 70
+2022-12-13 02:40:28,442 INFO [train.py:421] (4/8) Epoch 8, batch 55600, loss[loss=2.153, over 6090.00 frames. , ppl: 8.609155745027897] tot_loss[loss=2.278, over 5480466.89 frames. , ppl: 9.760933388382021], batch size: 70
+2022-12-13 02:42:10,518 INFO [train.py:421] (4/8) Epoch 8, batch 55800, loss[loss=2.304, over 1680.00 frames. , ppl: 10.01613682480954] tot_loss[loss=2.278, over 5516501.56 frames. , ppl: 9.753804893677092], batch size: 70
+2022-12-13 02:43:52,137 INFO [train.py:421] (4/8) Epoch 8, batch 56000, loss[loss=2.315, over 2100.00 frames. , ppl: 10.128589064322409] tot_loss[loss=2.278, over 5499540.78 frames. , ppl: 9.756690238637887], batch size: 70
+2022-12-13 02:43:52,137 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:43:52,898 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.676379424179105
+2022-12-13 02:45:35,148 INFO [train.py:421] (4/8) Epoch 8, batch 56200, loss[loss=2.304, over 4340.00 frames. , ppl: 10.016936383411934] tot_loss[loss=2.279, over 5486476.34 frames. , ppl: 9.76733959729053], batch size: 70
+2022-12-13 02:47:16,351 INFO [train.py:421] (4/8) Epoch 8, batch 56400, loss[loss=2.522, over 1120.00 frames. , ppl: 12.447626141740585] tot_loss[loss=2.279, over 5484990.48 frames. , ppl: 9.766252135122075], batch size: 70
+2022-12-13 02:48:53,785 INFO [train.py:421] (4/8) Epoch 8, batch 56600, loss[loss=2.256, over 2730.00 frames. , ppl: 9.54069227684828] tot_loss[loss=2.279, over 5490237.15 frames. , ppl: 9.769400814808176], batch size: 70
+2022-12-13 02:50:34,339 INFO [train.py:421] (4/8) Epoch 8, batch 56800, loss[loss=2.184, over 5740.00 frames. , ppl: 8.88486235525844] tot_loss[loss=2.28, over 5465454.64 frames. , ppl: 9.780204376757508], batch size: 70
+2022-12-13 02:52:17,207 INFO [train.py:421] (4/8) Epoch 8, batch 57000, loss[loss=2.395, over 1400.00 frames. , ppl: 10.9736429598772] tot_loss[loss=2.28, over 5472014.69 frames. , ppl: 9.77195453118968], batch size: 70
+2022-12-13 02:52:17,207 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 02:52:17,969 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66042525657999
+2022-12-13 02:54:00,655 INFO [train.py:421] (4/8) Epoch 8, batch 57200, loss[loss=2.158, over 4550.00 frames. , ppl: 8.655923675440548] tot_loss[loss=2.28, over 5488101.42 frames. , ppl: 9.773033137579993], batch size: 70
+2022-12-13 02:55:44,472 INFO [train.py:421] (4/8) Epoch 8, batch 57400, loss[loss=2.167, over 6160.00 frames. , ppl: 8.732114621359194] tot_loss[loss=2.279, over 5492663.07 frames. , ppl: 9.766277905136294], batch size: 70
+2022-12-13 02:57:27,576 INFO [train.py:421] (4/8) Epoch 8, batch 57600, loss[loss=2.196, over 4270.00 frames. , ppl: 8.992244433429525] tot_loss[loss=2.278, over 5497463.36 frames. , ppl: 9.759125685937367], batch size: 70
+2022-12-13 02:59:04,483 INFO [train.py:421] (4/8) Epoch 8, batch 57800, loss[loss=2.145, over 3220.00 frames. , ppl: 8.545177514671982] tot_loss[loss=2.279, over 5437787.16 frames. , ppl: 9.771044763291545], batch size: 70
+2022-12-13 03:00:43,920 INFO [train.py:421] (4/8) Epoch 8, batch 58000, loss[loss=2.362, over 1470.00 frames. , ppl: 10.614351826436303] tot_loss[loss=2.279, over 5434232.70 frames. , ppl: 9.77163078186459], batch size: 70
+2022-12-13 03:00:43,921 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:00:44,670 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.27, over 211138.00 frames. , ppl: 9.679271420376821
+2022-12-13 03:02:24,460 INFO [train.py:421] (4/8) Epoch 8, batch 58200, loss[loss=2.601, over 1050.00 frames. , ppl: 13.479440478569936] tot_loss[loss=2.279, over 5446588.07 frames. , ppl: 9.767196389260915], batch size: 70
+2022-12-13 03:04:02,655 INFO [train.py:421] (4/8) Epoch 8, batch 58400, loss[loss=2.344, over 2170.00 frames. , ppl: 10.422647573398715] tot_loss[loss=2.279, over 5449641.80 frames. , ppl: 9.766107972110532], batch size: 70
+2022-12-13 03:05:40,979 INFO [train.py:421] (4/8) Epoch 8, batch 58600, loss[loss=2.574, over 980.00 frames. , ppl: 13.118996629864332] tot_loss[loss=2.278, over 5476645.21 frames. , ppl: 9.757152158220665], batch size: 70
+2022-12-13 03:07:18,316 INFO [train.py:421] (4/8) Epoch 8, batch 58800, loss[loss=2.217, over 3080.00 frames. , ppl: 9.1763552886612] tot_loss[loss=2.276, over 5520458.99 frames. , ppl: 9.737538591963052], batch size: 70
+2022-12-13 03:09:03,624 INFO [train.py:421] (4/8) Epoch 8, batch 59000, loss[loss=2.358, over 1260.00 frames. , ppl: 10.571468840839666] tot_loss[loss=2.274, over 5590644.49 frames. , ppl: 9.719126209881331], batch size: 70
+2022-12-13 03:09:03,625 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:09:04,370 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.666656969839083
+2022-12-13 03:10:48,155 INFO [train.py:421] (4/8) Epoch 8, batch 59200, loss[loss=2.229, over 3430.00 frames. , ppl: 9.287618616380508] tot_loss[loss=2.274, over 5592348.44 frames. , ppl: 9.716963598283957], batch size: 70
+2022-12-13 03:12:28,716 INFO [train.py:421] (4/8) Epoch 8, batch 59400, loss[loss=2.409, over 1470.00 frames. , ppl: 11.126967633643432] tot_loss[loss=2.275, over 5560100.75 frames. , ppl: 9.723883273175774], batch size: 70
+2022-12-13 03:14:12,138 INFO [train.py:421] (4/8) Epoch 8, batch 59600, loss[loss=2.308, over 1610.00 frames. , ppl: 10.056353676880258] tot_loss[loss=2.275, over 5562425.29 frames. , ppl: 9.72631726442498], batch size: 70
+2022-12-13 03:15:48,387 INFO [train.py:421] (4/8) Epoch 8, batch 59800, loss[loss=2.449, over 1540.00 frames. , ppl: 11.577006132922406] tot_loss[loss=2.276, over 5539693.38 frames. , ppl: 9.735494656513461], batch size: 70
+2022-12-13 03:17:30,967 INFO [train.py:421] (4/8) Epoch 8, batch 60000, loss[loss=2.174, over 11550.00 frames. , ppl: 8.791281434641073] tot_loss[loss=2.276, over 5520845.54 frames. , ppl: 9.739732210509338], batch size: 70
+2022-12-13 03:17:30,967 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:17:31,713 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674444747320207
+2022-12-13 03:19:10,631 INFO [train.py:421] (4/8) Epoch 8, batch 60200, loss[loss=2.224, over 5040.00 frames. , ppl: 9.2480301748561] tot_loss[loss=2.276, over 5527071.10 frames. , ppl: 9.740299683506723], batch size: 70
+2022-12-13 03:20:51,023 INFO [train.py:421] (4/8) Epoch 8, batch 60400, loss[loss=2.461, over 980.00 frames. , ppl: 11.71864183743341] tot_loss[loss=2.276, over 5517458.19 frames. , ppl: 9.742329369675748], batch size: 70
+2022-12-13 03:22:27,997 INFO [train.py:421] (4/8) Epoch 8, batch 60600, loss[loss=2.242, over 2800.00 frames. , ppl: 9.411196632951851] tot_loss[loss=2.277, over 5510415.47 frames. , ppl: 9.743735087709451], batch size: 70
+2022-12-13 03:24:04,404 INFO [train.py:421] (4/8) Epoch 8, batch 60800, loss[loss=2.956, over 630.00 frames. , ppl: 19.220949837629867] tot_loss[loss=2.277, over 5483918.84 frames. , ppl: 9.749970093979947], batch size: 70
+2022-12-13 03:25:46,091 INFO [train.py:421] (4/8) Epoch 8, batch 61000, loss[loss=3.096, over 490.00 frames. , ppl: 22.11078722335913] tot_loss[loss=2.278, over 5467145.90 frames. , ppl: 9.758586975349123], batch size: 70
+2022-12-13 03:25:46,091 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:25:46,839 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66320950762563
+2022-12-13 03:27:25,696 INFO [train.py:421] (4/8) Epoch 8, batch 61200, loss[loss=2.417, over 1330.00 frames. , ppl: 11.215431732081152] tot_loss[loss=2.28, over 5406408.04 frames. , ppl: 9.77918798657011], batch size: 70
+2022-12-13 03:29:02,363 INFO [train.py:421] (4/8) Epoch 8, batch 61400, loss[loss=2.363, over 1750.00 frames. , ppl: 10.625419144711632] tot_loss[loss=2.279, over 5464908.41 frames. , ppl: 9.764892907657478], batch size: 70
+2022-12-13 03:30:39,175 INFO [train.py:421] (4/8) Epoch 8, batch 61600, loss[loss=2.369, over 1750.00 frames. , ppl: 10.68738607019283] tot_loss[loss=2.279, over 5426062.41 frames. , ppl: 9.76497168223485], batch size: 70
+2022-12-13 03:32:18,644 INFO [train.py:421] (4/8) Epoch 8, batch 61800, loss[loss=2.562, over 910.00 frames. , ppl: 12.95972129970518] tot_loss[loss=2.278, over 5459071.60 frames. , ppl: 9.759164219367335], batch size: 70
+2022-12-13 03:33:59,898 INFO [train.py:421] (4/8) Epoch 8, batch 62000, loss[loss=2.587, over 910.00 frames. , ppl: 13.291944873307528] tot_loss[loss=2.278, over 5450692.52 frames. , ppl: 9.759427521860726], batch size: 70
+2022-12-13 03:33:59,899 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:34:00,657 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.655050647650823
+2022-12-13 03:35:43,438 INFO [train.py:421] (4/8) Epoch 8, batch 62200, loss[loss=2.529, over 1190.00 frames. , ppl: 12.540621962015077] tot_loss[loss=2.278, over 5470677.20 frames. , ppl: 9.759000798147056], batch size: 70
+2022-12-13 03:37:22,806 INFO [train.py:421] (4/8) Epoch 8, batch 62400, loss[loss=2.244, over 4060.00 frames. , ppl: 9.432832555394887] tot_loss[loss=2.279, over 5488003.49 frames. , ppl: 9.762578867728763], batch size: 70
+2022-12-13 03:39:04,712 INFO [train.py:421] (4/8) Epoch 8, batch 62600, loss[loss=2.202, over 7350.00 frames. , ppl: 9.04430024699677] tot_loss[loss=2.281, over 5429455.17 frames. , ppl: 9.781590702225527], batch size: 70
+2022-12-13 03:40:42,483 INFO [train.py:421] (4/8) Epoch 8, batch 62800, loss[loss=2.376, over 3360.00 frames. , ppl: 10.76499911739705] tot_loss[loss=2.28, over 5445644.19 frames. , ppl: 9.778694053126767], batch size: 70
+2022-12-13 03:42:19,089 INFO [train.py:421] (4/8) Epoch 8, batch 63000, loss[loss=2.255, over 2660.00 frames. , ppl: 9.537500044315738] tot_loss[loss=2.28, over 5426787.95 frames. , ppl: 9.781203746514729], batch size: 70
+2022-12-13 03:42:19,090 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:42:19,855 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.660179331522313
+2022-12-13 03:44:04,024 INFO [train.py:421] (4/8) Epoch 8, batch 63200, loss[loss=2.37, over 2240.00 frames. , ppl: 10.699196352641563] tot_loss[loss=2.28, over 5451245.91 frames. , ppl: 9.774390922040682], batch size: 70
+2022-12-13 03:45:42,073 INFO [train.py:421] (4/8) Epoch 8, batch 63400, loss[loss=2.192, over 5460.00 frames. , ppl: 8.956468564873179] tot_loss[loss=2.279, over 5478359.44 frames. , ppl: 9.76947923159364], batch size: 70
+2022-12-13 03:47:24,377 INFO [train.py:421] (4/8) Epoch 8, batch 63600, loss[loss=2.302, over 1820.00 frames. , ppl: 9.995816755033305] tot_loss[loss=2.28, over 5465577.04 frames. , ppl: 9.77830952265908], batch size: 70
+2022-12-13 03:49:06,228 INFO [train.py:421] (4/8) Epoch 8, batch 63800, loss[loss=2.315, over 2660.00 frames. , ppl: 10.125266389265047] tot_loss[loss=2.28, over 5453167.35 frames. , ppl: 9.779768501919978], batch size: 70
+2022-12-13 03:50:46,485 INFO [train.py:421] (4/8) Epoch 8, batch 64000, loss[loss=2.225, over 2520.00 frames. , ppl: 9.253701549691694] tot_loss[loss=2.279, over 5488736.34 frames. , ppl: 9.762919978175248], batch size: 70
+2022-12-13 03:50:46,485 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:50:47,232 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.672815393681503
+2022-12-13 03:52:29,193 INFO [train.py:421] (4/8) Epoch 8, batch 64200, loss[loss=2.966, over 630.00 frames. , ppl: 19.413278729469997] tot_loss[loss=2.278, over 5480400.95 frames. , ppl: 9.75694582670687], batch size: 70
+2022-12-13 03:54:10,488 INFO [train.py:421] (4/8) Epoch 8, batch 64400, loss[loss=2.225, over 3290.00 frames. , ppl: 9.253529497724736] tot_loss[loss=2.279, over 5454871.99 frames. , ppl: 9.765731913027109], batch size: 70
+2022-12-13 03:55:52,715 INFO [train.py:421] (4/8) Epoch 8, batch 64600, loss[loss=2.442, over 1820.00 frames. , ppl: 11.498134641619687] tot_loss[loss=2.279, over 5452397.21 frames. , ppl: 9.764858407633753], batch size: 70
+2022-12-13 03:57:33,013 INFO [train.py:421] (4/8) Epoch 8, batch 64800, loss[loss=2.549, over 980.00 frames. , ppl: 12.789389375715071] tot_loss[loss=2.279, over 5466584.75 frames. , ppl: 9.762546355456694], batch size: 70
+2022-12-13 03:59:15,150 INFO [train.py:421] (4/8) Epoch 8, batch 65000, loss[loss=3.364, over 490.00 frames. , ppl: 28.890344927991514] tot_loss[loss=2.278, over 5461411.88 frames. , ppl: 9.761195576193238], batch size: 70
+2022-12-13 03:59:15,150 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 03:59:15,897 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65988194214127
+2022-12-13 04:00:56,727 INFO [train.py:421] (4/8) Epoch 8, batch 65200, loss[loss=2.239, over 3430.00 frames. , ppl: 9.386091015072685] tot_loss[loss=2.279, over 5444246.02 frames. , ppl: 9.769224178335827], batch size: 70
+2022-12-13 04:02:38,100 INFO [train.py:421] (4/8) Epoch 8, batch 65400, loss[loss=2.238, over 6650.00 frames. , ppl: 9.37030522866728] tot_loss[loss=2.278, over 5497781.80 frames. , ppl: 9.754834201622433], batch size: 70
+2022-12-13 04:04:12,154 INFO [train.py:421] (4/8) Epoch 8, batch 65600, loss[loss=2.471, over 1190.00 frames. , ppl: 11.831665386091098] tot_loss[loss=2.279, over 5448729.14 frames. , ppl: 9.76931648920873], batch size: 70
+2022-12-13 04:05:51,084 INFO [train.py:421] (4/8) Epoch 8, batch 65800, loss[loss=2.537, over 910.00 frames. , ppl: 12.647054512950165] tot_loss[loss=2.279, over 5447114.84 frames. , ppl: 9.763258395624463], batch size: 70
+2022-12-13 04:07:30,778 INFO [train.py:421] (4/8) Epoch 8, batch 66000, loss[loss=2.133, over 5250.00 frames. , ppl: 8.441929848450762] tot_loss[loss=2.279, over 5448404.68 frames. , ppl: 9.767098587630256], batch size: 70
+2022-12-13 04:07:30,778 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 04:07:31,525 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.665475252994222
+2022-12-13 04:09:13,156 INFO [train.py:421] (4/8) Epoch 8, batch 66200, loss[loss=2.232, over 2870.00 frames. , ppl: 9.315543500333611] tot_loss[loss=2.279, over 5449743.76 frames. , ppl: 9.769141792826815], batch size: 70
+2022-12-13 04:10:50,316 INFO [train.py:421] (4/8) Epoch 8, batch 66400, loss[loss=3.332, over 490.00 frames. , ppl: 27.99120199493557] tot_loss[loss=2.28, over 5413070.98 frames. , ppl: 9.776989545910224], batch size: 70
+2022-12-13 04:12:32,867 INFO [train.py:421] (4/8) Epoch 8, batch 66600, loss[loss=2.489, over 1190.00 frames. , ppl: 12.053561632601367] tot_loss[loss=2.279, over 5433110.78 frames. , ppl: 9.767890566864889], batch size: 70
+2022-12-13 04:14:11,380 INFO [train.py:421] (4/8) Epoch 8, batch 66800, loss[loss=2.92, over 630.00 frames. , ppl: 18.533402651638937] tot_loss[loss=2.278, over 5469275.06 frames. , ppl: 9.755917129105624], batch size: 70
+2022-12-13 04:15:48,896 INFO [train.py:421] (4/8) Epoch 8, batch 67000, loss[loss=2.302, over 2590.00 frames. , ppl: 9.998065612147535] tot_loss[loss=2.279, over 5442205.91 frames. , ppl: 9.762306917287775], batch size: 70
+2022-12-13 04:15:48,896 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 04:15:49,642 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.651940173203029
+2022-12-13 04:17:32,964 INFO [train.py:421] (4/8) Epoch 8, batch 67200, loss[loss=2.318, over 2660.00 frames. , ppl: 10.154269148909298] tot_loss[loss=2.278, over 5491982.99 frames. , ppl: 9.755193010593914], batch size: 70
+2022-12-13 04:19:12,944 INFO [train.py:421] (4/8) Epoch 8, batch 67400, loss[loss=2.374, over 1260.00 frames. , ppl: 10.740898043056257] tot_loss[loss=2.279, over 5446048.16 frames. , ppl: 9.767212916192994], batch size: 70
+2022-12-13 04:20:51,764 INFO [train.py:421] (4/8) Epoch 8, batch 67600, loss[loss=2.295, over 2100.00 frames. , ppl: 9.928181919851554] tot_loss[loss=2.278, over 5491245.88 frames. , ppl: 9.757560617935434], batch size: 70
+2022-12-13 04:22:29,390 INFO [train.py:421] (4/8) Epoch 8, batch 67800, loss[loss=2.508, over 1680.00 frames. , ppl: 12.277717712711674] tot_loss[loss=2.278, over 5504888.70 frames. , ppl: 9.754473593034549], batch size: 70
+2022-12-13 04:24:11,409 INFO [train.py:421] (4/8) Epoch 8, batch 68000, loss[loss=2.361, over 840.00 frames. , ppl: 10.60544918594959] tot_loss[loss=2.278, over 5498200.77 frames. , ppl: 9.753293498386272], batch size: 70
+2022-12-13 04:24:11,409 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 04:24:12,154 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.269, over 211138.00 frames. , ppl: 9.674181283141165
+2022-12-13 04:25:51,983 INFO [train.py:421] (4/8) Epoch 8, batch 68200, loss[loss=2.199, over 4270.00 frames. , ppl: 9.017218889770303] tot_loss[loss=2.277, over 5527655.34 frames. , ppl: 9.75066827953822], batch size: 70
+2022-12-13 04:27:29,340 INFO [train.py:421] (4/8) Epoch 8, batch 68400, loss[loss=2.52, over 1330.00 frames. , ppl: 12.428168671370182] tot_loss[loss=2.277, over 5554844.51 frames. , ppl: 9.744041587869635], batch size: 70
+2022-12-13 04:29:12,426 INFO [train.py:421] (4/8) Epoch 8, batch 68600, loss[loss=2.579, over 700.00 frames. , ppl: 13.190534799253497] tot_loss[loss=2.277, over 5545323.48 frames. , ppl: 9.74670388267633], batch size: 70
+2022-12-13 04:30:48,901 INFO [train.py:421] (4/8) Epoch 8, batch 68800, loss[loss=2.309, over 3220.00 frames. , ppl: 10.062824582538374] tot_loss[loss=2.275, over 5610520.99 frames. , ppl: 9.731264540989324], batch size: 70
+2022-12-13 04:32:30,673 INFO [train.py:421] (4/8) Epoch 8, batch 69000, loss[loss=2.234, over 3430.00 frames. , ppl: 9.335047108204098] tot_loss[loss=2.275, over 5604568.45 frames. , ppl: 9.729630913679104], batch size: 70
+2022-12-13 04:32:30,673 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 04:32:31,419 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.265, over 211138.00 frames. , ppl: 9.629115969513046
+2022-12-13 04:34:10,754 INFO [train.py:421] (4/8) Epoch 8, batch 69200, loss[loss=2.311, over 3570.00 frames. , ppl: 10.085520763557145] tot_loss[loss=2.277, over 5567690.56 frames. , ppl: 9.744249270690245], batch size: 70
+2022-12-13 04:35:51,600 INFO [train.py:421] (4/8) Epoch 8, batch 69400, loss[loss=2.309, over 1680.00 frames. , ppl: 10.060043772089962] tot_loss[loss=2.277, over 5532165.83 frames. , ppl: 9.751765430175606], batch size: 70
+2022-12-13 04:37:25,625 INFO [train.py:421] (4/8) Epoch 8, batch 69600, loss[loss=2.382, over 1610.00 frames. , ppl: 10.823740549068004] tot_loss[loss=2.277, over 5513425.09 frames. , ppl: 9.749606996570426], batch size: 70
+2022-12-13 04:39:03,668 INFO [train.py:421] (4/8) Epoch 8, batch 69800, loss[loss=2.362, over 770.00 frames. , ppl: 10.6076314736333] tot_loss[loss=2.277, over 5516447.51 frames. , ppl: 9.747636423293619], batch size: 70
+2022-12-13 04:40:44,763 INFO [train.py:421] (4/8) Epoch 8, batch 70000, loss[loss=2.524, over 1330.00 frames. , ppl: 12.472753399976733] tot_loss[loss=2.277, over 5537002.99 frames. , ppl: 9.74357139996357], batch size: 70
+2022-12-13 04:40:44,764 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 04:40:45,525 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.267, over 211138.00 frames. , ppl: 9.655207841082497
+2022-12-13 04:42:28,625 INFO [train.py:421] (4/8) Epoch 8, batch 70200, loss[loss=2.645, over 980.00 frames. , ppl: 14.083790324868973] tot_loss[loss=2.278, over 5474439.97 frames. , ppl: 9.76084927213047], batch size: 70
+2022-12-13 04:44:05,957 INFO [train.py:421] (4/8) Epoch 8, batch 70400, loss[loss=2.333, over 1400.00 frames. , ppl: 10.308837474576366] tot_loss[loss=2.279, over 5448843.94 frames. , ppl: 9.767354604466272], batch size: 70
+2022-12-13 04:45:49,700 INFO [train.py:421] (4/8) Epoch 8, batch 70600, loss[loss=2.216, over 4130.00 frames. , ppl: 9.169289484136382] tot_loss[loss=2.278, over 5464102.09 frames. , ppl: 9.759578618266863], batch size: 70
+2022-12-13 04:47:27,337 INFO [train.py:421] (4/8) Epoch 8, batch 70800, loss[loss=2.379, over 1050.00 frames. , ppl: 10.799306907563324] tot_loss[loss=2.278, over 5469510.25 frames. , ppl: 9.761498893470854], batch size: 70
+2022-12-13 04:49:05,280 INFO [train.py:421] (4/8) Epoch 8, batch 71000, loss[loss=2.243, over 3640.00 frames. , ppl: 9.42458607788457] tot_loss[loss=2.279, over 5471846.26 frames. , ppl: 9.765285764346668], batch size: 70
+2022-12-13 04:49:05,281 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 04:49:06,043 INFO [train.py:452] (4/8) Epoch 8, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664858700229336
+2022-12-13 04:50:47,497 INFO [train.py:421] (4/8) Epoch 8, batch 71200, loss[loss=2.406, over 1260.00 frames. , ppl: 11.085475890608555] tot_loss[loss=2.277, over 5539723.96 frames. , ppl: 9.745043237945563], batch size: 70
+2022-12-13 04:52:24,900 INFO [train.py:421] (4/8) Epoch 8, batch 71400, loss[loss=2.398, over 1260.00 frames. , ppl: 11.002537522510433] tot_loss[loss=2.277, over 5548060.49 frames. , ppl: 9.746790104922749], batch size: 70
+2022-12-13 04:54:03,448 INFO [train.py:421] (4/8) Epoch 8, batch 71600, loss[loss=2.105, over 12740.00 frames. , ppl: 8.204346248045317] tot_loss[loss=2.276, over 5554356.91 frames. , ppl: 9.739127118638878], batch size: 70
+2022-12-13 04:55:49,181 INFO [train.py:421] (4/8) Epoch 8, batch 71800, loss[loss=2.187, over 4480.00 frames. , ppl: 8.906573453342967] tot_loss[loss=2.274, over 5607159.84 frames. , ppl: 9.72149471066259], batch size: 70
+2022-12-13 04:57:06,856 INFO [train.py:421] (4/8) Epoch 9, batch 0, loss[loss=2.722, over 630.00 frames. , ppl: 15.217953120270375] tot_loss[loss=2.722, over 630.00 frames. , ppl: 15.217953120270375], batch size: 70
+2022-12-13 04:58:47,685 INFO [train.py:421] (4/8) Epoch 9, batch 200, loss[loss=2.222, over 9030.00 frames. , ppl: 9.225963663092804] tot_loss[loss=2.251, over 582575.89 frames. , ppl: 9.498812915774272], batch size: 70
+2022-12-13 05:00:29,937 INFO [train.py:421] (4/8) Epoch 9, batch 400, loss[loss=2.257, over 5810.00 frames. , ppl: 9.55372283725895] tot_loss[loss=2.264, over 1029584.58 frames. , ppl: 9.616902085397156], batch size: 70
+2022-12-13 05:02:09,653 INFO [train.py:421] (4/8) Epoch 9, batch 600, loss[loss=2.116, over 5320.00 frames. , ppl: 8.296711472414167] tot_loss[loss=2.264, over 1445720.25 frames. , ppl: 9.617203536846391], batch size: 70
+2022-12-13 05:03:48,157 INFO [train.py:421] (4/8) Epoch 9, batch 800, loss[loss=2.508, over 1190.00 frames. , ppl: 12.278009698546281] tot_loss[loss=2.264, over 1841180.49 frames. , ppl: 9.625126525180256], batch size: 70
+2022-12-13 05:05:26,193 INFO [train.py:421] (4/8) Epoch 9, batch 1000, loss[loss=2.162, over 5110.00 frames. , ppl: 8.69272633419025] tot_loss[loss=2.266, over 2204964.81 frames. , ppl: 9.638043900144428], batch size: 70
+2022-12-13 05:05:26,193 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:05:26,939 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664010468014997
+2022-12-13 05:07:07,627 INFO [train.py:421] (4/8) Epoch 9, batch 1200, loss[loss=2.499, over 1330.00 frames. , ppl: 12.166675906038696] tot_loss[loss=2.267, over 2514885.35 frames. , ppl: 9.652162760702671], batch size: 70
+2022-12-13 05:08:50,641 INFO [train.py:421] (4/8) Epoch 9, batch 1400, loss[loss=2.491, over 1050.00 frames. , ppl: 12.07478137907352] tot_loss[loss=2.265, over 2829595.92 frames. , ppl: 9.631962694578538], batch size: 70
+2022-12-13 05:10:29,028 INFO [train.py:421] (4/8) Epoch 9, batch 1600, loss[loss=2.238, over 4270.00 frames. , ppl: 9.374172686000039] tot_loss[loss=2.266, over 3076979.44 frames. , ppl: 9.64508062429556], batch size: 70
+2022-12-13 05:12:07,732 INFO [train.py:421] (4/8) Epoch 9, batch 1800, loss[loss=2.615, over 770.00 frames. , ppl: 13.670441679269807] tot_loss[loss=2.267, over 3317337.17 frames. , ppl: 9.645810669333152], batch size: 70
+2022-12-13 05:13:46,771 INFO [train.py:421] (4/8) Epoch 9, batch 2000, loss[loss=2.205, over 4410.00 frames. , ppl: 9.074524339775593] tot_loss[loss=2.266, over 3530791.66 frames. , ppl: 9.642433614122154], batch size: 70
+2022-12-13 05:13:46,772 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:13:47,536 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.667164894955036
+2022-12-13 05:15:30,256 INFO [train.py:421] (4/8) Epoch 9, batch 2200, loss[loss=2.41, over 1400.00 frames. , ppl: 11.13682344419894] tot_loss[loss=2.267, over 3723681.06 frames. , ppl: 9.64979282515103], batch size: 70
+2022-12-13 05:17:11,207 INFO [train.py:421] (4/8) Epoch 9, batch 2400, loss[loss=2.606, over 1260.00 frames. , ppl: 13.549152743789158] tot_loss[loss=2.266, over 3891756.06 frames. , ppl: 9.642408273521491], batch size: 70
+2022-12-13 05:18:52,166 INFO [train.py:421] (4/8) Epoch 9, batch 2600, loss[loss=2.344, over 3220.00 frames. , ppl: 10.420144477077432] tot_loss[loss=2.267, over 4029447.82 frames. , ppl: 9.646881664339483], batch size: 70
+2022-12-13 05:20:30,385 INFO [train.py:421] (4/8) Epoch 9, batch 2800, loss[loss=2.198, over 3150.00 frames. , ppl: 9.009998304428919] tot_loss[loss=2.268, over 4128114.47 frames. , ppl: 9.660353444514461], batch size: 70
+2022-12-13 05:22:11,118 INFO [train.py:421] (4/8) Epoch 9, batch 3000, loss[loss=2.574, over 1050.00 frames. , ppl: 13.124327130588059] tot_loss[loss=2.269, over 4236925.58 frames. , ppl: 9.672236049338157], batch size: 70
+2022-12-13 05:22:11,118 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:22:11,878 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65743883465414
+2022-12-13 05:23:52,011 INFO [train.py:421] (4/8) Epoch 9, batch 3200, loss[loss=2.335, over 3430.00 frames. , ppl: 10.32932111877348] tot_loss[loss=2.27, over 4345391.73 frames. , ppl: 9.682555218460816], batch size: 70
+2022-12-13 05:25:30,714 INFO [train.py:421] (4/8) Epoch 9, batch 3400, loss[loss=3.15, over 560.00 frames. , ppl: 23.33380612511284] tot_loss[loss=2.27, over 4481669.22 frames. , ppl: 9.67686435621911], batch size: 70
+2022-12-13 05:27:10,001 INFO [train.py:421] (4/8) Epoch 9, batch 3600, loss[loss=2.565, over 770.00 frames. , ppl: 12.994786580066828] tot_loss[loss=2.27, over 4587360.74 frames. , ppl: 9.681554041178948], batch size: 70
+2022-12-13 05:28:50,468 INFO [train.py:421] (4/8) Epoch 9, batch 3800, loss[loss=3.05, over 560.00 frames. , ppl: 21.11765514442972] tot_loss[loss=2.27, over 4656546.69 frames. , ppl: 9.682730446931156], batch size: 70
+2022-12-13 05:30:30,350 INFO [train.py:421] (4/8) Epoch 9, batch 4000, loss[loss=2.48, over 910.00 frames. , ppl: 11.942655536232131] tot_loss[loss=2.27, over 4743306.14 frames. , ppl: 9.6786711066569], batch size: 70
+2022-12-13 05:30:30,351 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:30:31,100 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.671714518344173
+2022-12-13 05:32:07,771 INFO [train.py:421] (4/8) Epoch 9, batch 4200, loss[loss=2.212, over 2660.00 frames. , ppl: 9.136459350886451] tot_loss[loss=2.271, over 4805067.34 frames. , ppl: 9.684369370292751], batch size: 70
+2022-12-13 05:33:45,394 INFO [train.py:421] (4/8) Epoch 9, batch 4400, loss[loss=2.354, over 4130.00 frames. , ppl: 10.52380675565154] tot_loss[loss=2.273, over 4811288.06 frames. , ppl: 9.70687569806686], batch size: 70
+2022-12-13 05:35:25,479 INFO [train.py:421] (4/8) Epoch 9, batch 4600, loss[loss=2.171, over 3430.00 frames. , ppl: 8.766076481195942] tot_loss[loss=2.273, over 4869855.67 frames. , ppl: 9.709253876174964], batch size: 70
+2022-12-13 05:37:03,949 INFO [train.py:421] (4/8) Epoch 9, batch 4800, loss[loss=2.254, over 2590.00 frames. , ppl: 9.530434430930871] tot_loss[loss=2.273, over 4923150.31 frames. , ppl: 9.708366636165973], batch size: 70
+2022-12-13 05:38:46,307 INFO [train.py:421] (4/8) Epoch 9, batch 5000, loss[loss=2.383, over 1540.00 frames. , ppl: 10.83208835372823] tot_loss[loss=2.271, over 5002284.91 frames. , ppl: 9.693238735540854], batch size: 70
+2022-12-13 05:38:46,307 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:38:47,053 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.655562250208732
+2022-12-13 05:40:30,491 INFO [train.py:421] (4/8) Epoch 9, batch 5200, loss[loss=2.279, over 1890.00 frames. , ppl: 9.762204135645524] tot_loss[loss=2.274, over 4979739.54 frames. , ppl: 9.717364318294743], batch size: 70
+2022-12-13 05:42:13,825 INFO [train.py:421] (4/8) Epoch 9, batch 5400, loss[loss=3.208, over 490.00 frames. , ppl: 24.73206268560737] tot_loss[loss=2.275, over 4997502.53 frames. , ppl: 9.723511779753194], batch size: 70
+2022-12-13 05:43:55,319 INFO [train.py:421] (4/8) Epoch 9, batch 5600, loss[loss=2.195, over 1610.00 frames. , ppl: 8.98065579980547] tot_loss[loss=2.275, over 5040430.21 frames. , ppl: 9.723349253130102], batch size: 70
+2022-12-13 05:45:33,974 INFO [train.py:421] (4/8) Epoch 9, batch 5800, loss[loss=2.404, over 1050.00 frames. , ppl: 11.071652587769778] tot_loss[loss=2.275, over 5059339.94 frames. , ppl: 9.725820536332286], batch size: 70
+2022-12-13 05:47:17,486 INFO [train.py:421] (4/8) Epoch 9, batch 6000, loss[loss=2.507, over 1260.00 frames. , ppl: 12.263199184032262] tot_loss[loss=2.274, over 5110756.01 frames. , ppl: 9.720480482191642], batch size: 70
+2022-12-13 05:47:17,486 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:47:18,250 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.66698032189641
+2022-12-13 05:48:59,167 INFO [train.py:421] (4/8) Epoch 9, batch 6200, loss[loss=2.19, over 4550.00 frames. , ppl: 8.933978164582255] tot_loss[loss=2.273, over 5187530.04 frames. , ppl: 9.708649723500537], batch size: 70
+2022-12-13 05:50:38,216 INFO [train.py:421] (4/8) Epoch 9, batch 6400, loss[loss=2.403, over 1190.00 frames. , ppl: 11.057115089736891] tot_loss[loss=2.272, over 5241143.32 frames. , ppl: 9.696486039737966], batch size: 70
+2022-12-13 05:52:22,372 INFO [train.py:421] (4/8) Epoch 9, batch 6600, loss[loss=2.487, over 1470.00 frames. , ppl: 12.026859321016115] tot_loss[loss=2.271, over 5285672.01 frames. , ppl: 9.69080478166532], batch size: 70
+2022-12-13 05:54:01,401 INFO [train.py:421] (4/8) Epoch 9, batch 6800, loss[loss=2.166, over 3290.00 frames. , ppl: 8.722238970697658] tot_loss[loss=2.27, over 5306940.33 frames. , ppl: 9.68362103958589], batch size: 70
+2022-12-13 05:55:37,121 INFO [train.py:421] (4/8) Epoch 9, batch 7000, loss[loss=2.234, over 4270.00 frames. , ppl: 9.339945739694453] tot_loss[loss=2.27, over 5339859.00 frames. , ppl: 9.68241640985228], batch size: 70
+2022-12-13 05:55:37,122 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 05:55:37,868 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.271, over 211138.00 frames. , ppl: 9.689801055644402
+2022-12-13 05:57:20,761 INFO [train.py:421] (4/8) Epoch 9, batch 7200, loss[loss=2.244, over 3920.00 frames. , ppl: 9.42995667177583] tot_loss[loss=2.27, over 5370492.56 frames. , ppl: 9.68023397480968], batch size: 70
+2022-12-13 05:59:03,231 INFO [train.py:421] (4/8) Epoch 9, batch 7400, loss[loss=2.222, over 2450.00 frames. , ppl: 9.229976965815304] tot_loss[loss=2.271, over 5357009.46 frames. , ppl: 9.691535220165425], batch size: 70
+2022-12-13 06:00:38,563 INFO [train.py:421] (4/8) Epoch 9, batch 7600, loss[loss=2.161, over 3710.00 frames. , ppl: 8.677776664741861] tot_loss[loss=2.271, over 5376330.73 frames. , ppl: 9.690321550855062], batch size: 70
+2022-12-13 06:02:16,060 INFO [train.py:421] (4/8) Epoch 9, batch 7800, loss[loss=2.361, over 2030.00 frames. , ppl: 10.600349668269645] tot_loss[loss=2.271, over 5369236.01 frames. , ppl: 9.691620920129873], batch size: 70
+2022-12-13 06:03:55,032 INFO [train.py:421] (4/8) Epoch 9, batch 8000, loss[loss=2.223, over 3500.00 frames. , ppl: 9.231705724900703] tot_loss[loss=2.273, over 5358090.06 frames. , ppl: 9.705001826444445], batch size: 70
+2022-12-13 06:03:55,033 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:03:55,778 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.27, over 211138.00 frames. , ppl: 9.677170017837494
+2022-12-13 06:05:34,733 INFO [train.py:421] (4/8) Epoch 9, batch 8200, loss[loss=2.39, over 1400.00 frames. , ppl: 10.913975453728053] tot_loss[loss=2.271, over 5387784.15 frames. , ppl: 9.693906932666257], batch size: 70
+2022-12-13 06:07:15,119 INFO [train.py:421] (4/8) Epoch 9, batch 8400, loss[loss=2.249, over 2520.00 frames. , ppl: 9.473873659576924] tot_loss[loss=2.273, over 5344530.59 frames. , ppl: 9.708124125844373], batch size: 70
+2022-12-13 06:08:56,773 INFO [train.py:421] (4/8) Epoch 9, batch 8600, loss[loss=2.374, over 2030.00 frames. , ppl: 10.735017945239035] tot_loss[loss=2.273, over 5379135.07 frames. , ppl: 9.705842506962428], batch size: 70
+2022-12-13 06:10:33,931 INFO [train.py:421] (4/8) Epoch 9, batch 8800, loss[loss=2.331, over 1610.00 frames. , ppl: 10.29206731663379] tot_loss[loss=2.271, over 5410827.64 frames. , ppl: 9.692800757830847], batch size: 70
+2022-12-13 06:12:16,353 INFO [train.py:421] (4/8) Epoch 9, batch 9000, loss[loss=2.587, over 910.00 frames. , ppl: 13.2908537073937] tot_loss[loss=2.271, over 5432381.71 frames. , ppl: 9.6916711546432], batch size: 70
+2022-12-13 06:12:16,353 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:12:17,100 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.667984785319996
+2022-12-13 06:13:54,674 INFO [train.py:421] (4/8) Epoch 9, batch 9200, loss[loss=2.303, over 2590.00 frames. , ppl: 10.000648246475214] tot_loss[loss=2.271, over 5463021.26 frames. , ppl: 9.688327065099877], batch size: 70
+2022-12-13 06:15:37,227 INFO [train.py:421] (4/8) Epoch 9, batch 9400, loss[loss=2.339, over 1260.00 frames. , ppl: 10.371846333806856] tot_loss[loss=2.27, over 5459653.20 frames. , ppl: 9.68073558856711], batch size: 70
+2022-12-13 06:17:15,205 INFO [train.py:421] (4/8) Epoch 9, batch 9600, loss[loss=2.199, over 3850.00 frames. , ppl: 9.016135702501616] tot_loss[loss=2.271, over 5442864.68 frames. , ppl: 9.68832504600834], batch size: 70
+2022-12-13 06:18:55,628 INFO [train.py:421] (4/8) Epoch 9, batch 9800, loss[loss=2.436, over 2450.00 frames. , ppl: 11.431780975739226] tot_loss[loss=2.271, over 5435990.22 frames. , ppl: 9.691038817217848], batch size: 70
+2022-12-13 06:20:37,101 INFO [train.py:421] (4/8) Epoch 9, batch 10000, loss[loss=2.354, over 2170.00 frames. , ppl: 10.524846964636849] tot_loss[loss=2.271, over 5461941.45 frames. , ppl: 9.686524478336619], batch size: 70
+2022-12-13 06:20:37,102 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:20:37,860 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.658802552803898
+2022-12-13 06:22:19,989 INFO [train.py:421] (4/8) Epoch 9, batch 10200, loss[loss=2.499, over 1190.00 frames. , ppl: 12.171469264302537] tot_loss[loss=2.272, over 5440869.80 frames. , ppl: 9.698608992741695], batch size: 70
+2022-12-13 06:23:58,842 INFO [train.py:421] (4/8) Epoch 9, batch 10400, loss[loss=2.328, over 2170.00 frames. , ppl: 10.257534336683145] tot_loss[loss=2.272, over 5471792.43 frames. , ppl: 9.695914330961195], batch size: 70
+2022-12-13 06:25:43,275 INFO [train.py:421] (4/8) Epoch 9, batch 10600, loss[loss=2.454, over 840.00 frames. , ppl: 11.6356195589253] tot_loss[loss=2.271, over 5455219.66 frames. , ppl: 9.69230497763332], batch size: 70
+2022-12-13 06:27:26,722 INFO [train.py:421] (4/8) Epoch 9, batch 10800, loss[loss=2.181, over 4200.00 frames. , ppl: 8.857581124091448] tot_loss[loss=2.27, over 5489351.89 frames. , ppl: 9.683392649874893], batch size: 70
+2022-12-13 06:29:05,717 INFO [train.py:421] (4/8) Epoch 9, batch 11000, loss[loss=2.417, over 1260.00 frames. , ppl: 11.210138324460447] tot_loss[loss=2.271, over 5482163.17 frames. , ppl: 9.686547433577568], batch size: 70
+2022-12-13 06:29:05,717 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:29:06,467 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.671750305558916
+2022-12-13 06:30:49,649 INFO [train.py:421] (4/8) Epoch 9, batch 11200, loss[loss=2.181, over 8610.00 frames. , ppl: 8.857835202106916] tot_loss[loss=2.27, over 5500551.60 frames. , ppl: 9.677616728515677], batch size: 70
+2022-12-13 06:32:30,897 INFO [train.py:421] (4/8) Epoch 9, batch 11400, loss[loss=2.422, over 1260.00 frames. , ppl: 11.268174591755026] tot_loss[loss=2.269, over 5526838.68 frames. , ppl: 9.669843053163941], batch size: 70
+2022-12-13 06:34:12,383 INFO [train.py:421] (4/8) Epoch 9, batch 11600, loss[loss=2.45, over 1190.00 frames. , ppl: 11.593012249201593] tot_loss[loss=2.27, over 5512248.78 frames. , ppl: 9.675599909249053], batch size: 70
+2022-12-13 06:35:50,659 INFO [train.py:421] (4/8) Epoch 9, batch 11800, loss[loss=2.468, over 770.00 frames. , ppl: 11.799313573661848] tot_loss[loss=2.27, over 5500535.82 frames. , ppl: 9.682798290615796], batch size: 70
+2022-12-13 06:37:30,263 INFO [train.py:421] (4/8) Epoch 9, batch 12000, loss[loss=2.163, over 3290.00 frames. , ppl: 8.69558330630908] tot_loss[loss=2.271, over 5479864.96 frames. , ppl: 9.688455054686111], batch size: 70
+2022-12-13 06:37:30,263 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:37:31,009 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.655009206172396
+2022-12-13 06:39:10,529 INFO [train.py:421] (4/8) Epoch 9, batch 12200, loss[loss=2.541, over 770.00 frames. , ppl: 12.697433782606069] tot_loss[loss=2.272, over 5477260.74 frames. , ppl: 9.695447659783643], batch size: 70
+2022-12-13 06:40:53,668 INFO [train.py:421] (4/8) Epoch 9, batch 12400, loss[loss=2.194, over 4200.00 frames. , ppl: 8.967836357681565] tot_loss[loss=2.271, over 5516882.76 frames. , ppl: 9.68643734065653], batch size: 70
+2022-12-13 06:42:37,744 INFO [train.py:421] (4/8) Epoch 9, batch 12600, loss[loss=2.329, over 3150.00 frames. , ppl: 10.268342311967807] tot_loss[loss=2.27, over 5550541.05 frames. , ppl: 9.676544003298037], batch size: 70
+2022-12-13 06:44:16,096 INFO [train.py:421] (4/8) Epoch 9, batch 12800, loss[loss=2.881, over 630.00 frames. , ppl: 17.82510525877907] tot_loss[loss=2.272, over 5490422.91 frames. , ppl: 9.69394337964205], batch size: 70
+2022-12-13 06:45:55,153 INFO [train.py:421] (4/8) Epoch 9, batch 13000, loss[loss=3.261, over 490.00 frames. , ppl: 26.06282591944207] tot_loss[loss=2.273, over 5442758.31 frames. , ppl: 9.70841352701564], batch size: 70
+2022-12-13 06:45:55,154 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:45:55,900 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66105725550274
+2022-12-13 06:47:36,600 INFO [train.py:421] (4/8) Epoch 9, batch 13200, loss[loss=2.948, over 700.00 frames. , ppl: 19.06657234846214] tot_loss[loss=2.273, over 5459399.42 frames. , ppl: 9.70736995759512], batch size: 70
+2022-12-13 06:49:17,417 INFO [train.py:421] (4/8) Epoch 9, batch 13400, loss[loss=2.212, over 9870.00 frames. , ppl: 9.136333662270218] tot_loss[loss=2.274, over 5451580.64 frames. , ppl: 9.713629104906344], batch size: 70
+2022-12-13 06:50:56,810 INFO [train.py:421] (4/8) Epoch 9, batch 13600, loss[loss=2.097, over 5530.00 frames. , ppl: 8.139043148305639] tot_loss[loss=2.272, over 5491602.79 frames. , ppl: 9.701017411444093], batch size: 70
+2022-12-13 06:52:32,853 INFO [train.py:421] (4/8) Epoch 9, batch 13800, loss[loss=2.258, over 6580.00 frames. , ppl: 9.56653242406903] tot_loss[loss=2.273, over 5449264.37 frames. , ppl: 9.713194703052459], batch size: 70
+2022-12-13 06:54:17,171 INFO [train.py:421] (4/8) Epoch 9, batch 14000, loss[loss=2.441, over 1400.00 frames. , ppl: 11.487495179829372] tot_loss[loss=2.273, over 5477071.25 frames. , ppl: 9.704385376866492], batch size: 70
+2022-12-13 06:54:17,172 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 06:54:17,916 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.27, over 211138.00 frames. , ppl: 9.675671954910115
+2022-12-13 06:55:57,100 INFO [train.py:421] (4/8) Epoch 9, batch 14200, loss[loss=2.282, over 2730.00 frames. , ppl: 9.794874013519882] tot_loss[loss=2.274, over 5433306.18 frames. , ppl: 9.714545796889569], batch size: 70
+2022-12-13 06:57:38,458 INFO [train.py:421] (4/8) Epoch 9, batch 14400, loss[loss=2.385, over 1470.00 frames. , ppl: 10.855910251799175] tot_loss[loss=2.272, over 5486168.43 frames. , ppl: 9.700162735094784], batch size: 70
+2022-12-13 06:59:18,391 INFO [train.py:421] (4/8) Epoch 9, batch 14600, loss[loss=2.534, over 980.00 frames. , ppl: 12.601460616623399] tot_loss[loss=2.271, over 5517374.15 frames. , ppl: 9.691848418198399], batch size: 70
+2022-12-13 07:00:59,400 INFO [train.py:421] (4/8) Epoch 9, batch 14800, loss[loss=3.227, over 490.00 frames. , ppl: 25.213061786849465] tot_loss[loss=2.27, over 5551287.58 frames. , ppl: 9.68060953284134], batch size: 70
+2022-12-13 07:02:36,592 INFO [train.py:421] (4/8) Epoch 9, batch 15000, loss[loss=2.537, over 1750.00 frames. , ppl: 12.640987054775783] tot_loss[loss=2.271, over 5498619.60 frames. , ppl: 9.693401783750414], batch size: 70
+2022-12-13 07:02:36,592 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:02:37,339 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.661317502370732
+2022-12-13 07:04:17,421 INFO [train.py:421] (4/8) Epoch 9, batch 15200, loss[loss=2.266, over 2870.00 frames. , ppl: 9.638359312139944] tot_loss[loss=2.273, over 5471518.98 frames. , ppl: 9.703820496667207], batch size: 70
+2022-12-13 07:05:53,691 INFO [train.py:421] (4/8) Epoch 9, batch 15400, loss[loss=2.311, over 2940.00 frames. , ppl: 10.079845211959071] tot_loss[loss=2.272, over 5483834.64 frames. , ppl: 9.700025379911551], batch size: 70
+2022-12-13 07:07:28,823 INFO [train.py:421] (4/8) Epoch 9, batch 15600, loss[loss=2.214, over 2310.00 frames. , ppl: 9.149816354222152] tot_loss[loss=2.271, over 5491027.19 frames. , ppl: 9.689366818648828], batch size: 70
+2022-12-13 07:09:10,482 INFO [train.py:421] (4/8) Epoch 9, batch 15800, loss[loss=2.568, over 840.00 frames. , ppl: 13.038576222436438] tot_loss[loss=2.271, over 5534445.81 frames. , ppl: 9.684586663210718], batch size: 70
+2022-12-13 07:10:50,253 INFO [train.py:421] (4/8) Epoch 9, batch 16000, loss[loss=2.286, over 1470.00 frames. , ppl: 9.830879880727135] tot_loss[loss=2.271, over 5532121.41 frames. , ppl: 9.68724794820643], batch size: 70
+2022-12-13 07:10:50,254 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:10:51,000 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65729732776054
+2022-12-13 07:12:29,286 INFO [train.py:421] (4/8) Epoch 9, batch 16200, loss[loss=2.31, over 1960.00 frames. , ppl: 10.071667295208018] tot_loss[loss=2.271, over 5537086.15 frames. , ppl: 9.693238454761017], batch size: 70
+2022-12-13 07:14:07,469 INFO [train.py:421] (4/8) Epoch 9, batch 16400, loss[loss=2.383, over 1190.00 frames. , ppl: 10.836610492792204] tot_loss[loss=2.27, over 5570000.34 frames. , ppl: 9.681870385644709], batch size: 70
+2022-12-13 07:15:49,881 INFO [train.py:421] (4/8) Epoch 9, batch 16600, loss[loss=2.44, over 1470.00 frames. , ppl: 11.469307017778208] tot_loss[loss=2.272, over 5527659.30 frames. , ppl: 9.698399982570177], batch size: 70
+2022-12-13 07:17:30,242 INFO [train.py:421] (4/8) Epoch 9, batch 16800, loss[loss=2.686, over 770.00 frames. , ppl: 14.673494423906744] tot_loss[loss=2.273, over 5486625.04 frames. , ppl: 9.712874652734618], batch size: 70
+2022-12-13 07:19:11,498 INFO [train.py:421] (4/8) Epoch 9, batch 17000, loss[loss=2.296, over 4480.00 frames. , ppl: 9.93909837327191] tot_loss[loss=2.273, over 5497946.05 frames. , ppl: 9.706981526309558], batch size: 70
+2022-12-13 07:19:11,498 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:19:12,256 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.27, over 211138.00 frames. , ppl: 9.679668259935742
+2022-12-13 07:20:56,017 INFO [train.py:421] (4/8) Epoch 9, batch 17200, loss[loss=2.294, over 1820.00 frames. , ppl: 9.916743573888366] tot_loss[loss=2.272, over 5521534.90 frames. , ppl: 9.697962229877312], batch size: 70
+2022-12-13 07:22:38,926 INFO [train.py:421] (4/8) Epoch 9, batch 17400, loss[loss=2.195, over 4830.00 frames. , ppl: 8.981957801650395] tot_loss[loss=2.272, over 5513628.52 frames. , ppl: 9.702447600900397], batch size: 70
+2022-12-13 07:24:16,738 INFO [train.py:421] (4/8) Epoch 9, batch 17600, loss[loss=2.326, over 2310.00 frames. , ppl: 10.238725607041003] tot_loss[loss=2.272, over 5528978.11 frames. , ppl: 9.703147550812758], batch size: 70
+2022-12-13 07:25:57,040 INFO [train.py:421] (4/8) Epoch 9, batch 17800, loss[loss=2.737, over 700.00 frames. , ppl: 15.443882347708255] tot_loss[loss=2.273, over 5491936.52 frames. , ppl: 9.70496287574786], batch size: 70
+2022-12-13 07:27:36,348 INFO [train.py:421] (4/8) Epoch 9, batch 18000, loss[loss=2.402, over 2450.00 frames. , ppl: 11.045160705106088] tot_loss[loss=2.271, over 5541909.83 frames. , ppl: 9.692994249049299], batch size: 70
+2022-12-13 07:27:36,349 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:27:37,108 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635843748058024
+2022-12-13 07:29:17,364 INFO [train.py:421] (4/8) Epoch 9, batch 18200, loss[loss=2.336, over 1820.00 frames. , ppl: 10.335165853217758] tot_loss[loss=2.272, over 5528892.00 frames. , ppl: 9.698456985747004], batch size: 70
+2022-12-13 07:30:55,921 INFO [train.py:421] (4/8) Epoch 9, batch 18400, loss[loss=2.322, over 2660.00 frames. , ppl: 10.19369746826417] tot_loss[loss=2.273, over 5500947.22 frames. , ppl: 9.706919069294939], batch size: 70
+2022-12-13 07:32:36,191 INFO [train.py:421] (4/8) Epoch 9, batch 18600, loss[loss=2.452, over 1540.00 frames. , ppl: 11.606217426064491] tot_loss[loss=2.274, over 5460018.00 frames. , ppl: 9.721987974463097], batch size: 70
+2022-12-13 07:34:20,274 INFO [train.py:421] (4/8) Epoch 9, batch 18800, loss[loss=2.396, over 1960.00 frames. , ppl: 10.973949695155985] tot_loss[loss=2.274, over 5503069.61 frames. , ppl: 9.718274279705211], batch size: 70
+2022-12-13 07:36:01,308 INFO [train.py:421] (4/8) Epoch 9, batch 19000, loss[loss=2.406, over 1680.00 frames. , ppl: 11.089854811120947] tot_loss[loss=2.273, over 5521900.45 frames. , ppl: 9.706912513068305], batch size: 70
+2022-12-13 07:36:01,309 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:36:02,070 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.654813433317269
+2022-12-13 07:37:39,895 INFO [train.py:421] (4/8) Epoch 9, batch 19200, loss[loss=2.381, over 1260.00 frames. , ppl: 10.81739453888394] tot_loss[loss=2.274, over 5506235.12 frames. , ppl: 9.713855191397084], batch size: 70
+2022-12-13 07:39:21,549 INFO [train.py:421] (4/8) Epoch 9, batch 19400, loss[loss=2.298, over 1470.00 frames. , ppl: 9.95256827705524] tot_loss[loss=2.272, over 5543774.24 frames. , ppl: 9.700075263478842], batch size: 70
+2022-12-13 07:41:01,750 INFO [train.py:421] (4/8) Epoch 9, batch 19600, loss[loss=2.275, over 2380.00 frames. , ppl: 9.731246549640128] tot_loss[loss=2.273, over 5520853.99 frames. , ppl: 9.70455039500768], batch size: 70
+2022-12-13 07:42:40,800 INFO [train.py:421] (4/8) Epoch 9, batch 19800, loss[loss=2.428, over 1260.00 frames. , ppl: 11.335673306417158] tot_loss[loss=2.273, over 5508270.06 frames. , ppl: 9.708433749745513], batch size: 70
+2022-12-13 07:44:22,225 INFO [train.py:421] (4/8) Epoch 9, batch 20000, loss[loss=2.229, over 2940.00 frames. , ppl: 9.291455843647075] tot_loss[loss=2.274, over 5498343.17 frames. , ppl: 9.71741535090453], batch size: 70
+2022-12-13 07:44:22,225 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:44:22,986 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.671224962542615
+2022-12-13 07:46:03,948 INFO [train.py:421] (4/8) Epoch 9, batch 20200, loss[loss=2.428, over 1050.00 frames. , ppl: 11.330984025801081] tot_loss[loss=2.274, over 5482811.32 frames. , ppl: 9.720867872375203], batch size: 70
+2022-12-13 07:47:42,245 INFO [train.py:421] (4/8) Epoch 9, batch 20400, loss[loss=2.303, over 3360.00 frames. , ppl: 10.007727202815161] tot_loss[loss=2.274, over 5487937.22 frames. , ppl: 9.715238071849946], batch size: 70
+2022-12-13 07:49:23,897 INFO [train.py:421] (4/8) Epoch 9, batch 20600, loss[loss=3.431, over 420.00 frames. , ppl: 30.904446248952578] tot_loss[loss=2.273, over 5509149.38 frames. , ppl: 9.706604700378104], batch size: 70
+2022-12-13 07:51:06,183 INFO [train.py:421] (4/8) Epoch 9, batch 20800, loss[loss=2.216, over 6440.00 frames. , ppl: 9.170665549797233] tot_loss[loss=2.272, over 5505773.86 frames. , ppl: 9.69786394139464], batch size: 70
+2022-12-13 07:52:48,694 INFO [train.py:421] (4/8) Epoch 9, batch 21000, loss[loss=2.243, over 4130.00 frames. , ppl: 9.418552893967293] tot_loss[loss=2.272, over 5508001.29 frames. , ppl: 9.69968716842511], batch size: 70
+2022-12-13 07:52:48,695 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 07:52:49,455 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.652867352648064
+2022-12-13 07:54:33,301 INFO [train.py:421] (4/8) Epoch 9, batch 21200, loss[loss=2.321, over 3710.00 frames. , ppl: 10.189955826231538] tot_loss[loss=2.271, over 5550815.32 frames. , ppl: 9.692854343684687], batch size: 70
+2022-12-13 07:56:15,662 INFO [train.py:421] (4/8) Epoch 9, batch 21400, loss[loss=2.311, over 2800.00 frames. , ppl: 10.08434690112868] tot_loss[loss=2.272, over 5513467.60 frames. , ppl: 9.699950766693737], batch size: 70
+2022-12-13 07:57:54,060 INFO [train.py:421] (4/8) Epoch 9, batch 21600, loss[loss=2.159, over 3500.00 frames. , ppl: 8.664814437935268] tot_loss[loss=2.273, over 5499693.39 frames. , ppl: 9.704803014148276], batch size: 70
+2022-12-13 07:59:33,285 INFO [train.py:421] (4/8) Epoch 9, batch 21800, loss[loss=2.395, over 1120.00 frames. , ppl: 10.969923451934262] tot_loss[loss=2.273, over 5487029.42 frames. , ppl: 9.706786786007852], batch size: 70
+2022-12-13 08:01:12,366 INFO [train.py:421] (4/8) Epoch 9, batch 22000, loss[loss=2.165, over 8400.00 frames. , ppl: 8.714642442438013] tot_loss[loss=2.272, over 5518592.28 frames. , ppl: 9.69545013625203], batch size: 70
+2022-12-13 08:01:12,366 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:01:13,114 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.65536646613963
+2022-12-13 08:02:53,702 INFO [train.py:421] (4/8) Epoch 9, batch 22200, loss[loss=2.395, over 1400.00 frames. , ppl: 10.965063588552052] tot_loss[loss=2.272, over 5490603.21 frames. , ppl: 9.698153419107083], batch size: 70
+2022-12-13 08:04:32,686 INFO [train.py:421] (4/8) Epoch 9, batch 22400, loss[loss=2.611, over 700.00 frames. , ppl: 13.613593461442964] tot_loss[loss=2.273, over 5445233.08 frames. , ppl: 9.709136530359345], batch size: 70
+2022-12-13 08:06:15,182 INFO [train.py:421] (4/8) Epoch 9, batch 22600, loss[loss=2.363, over 1610.00 frames. , ppl: 10.627297372092043] tot_loss[loss=2.274, over 5433363.88 frames. , ppl: 9.71761222916998], batch size: 70
+2022-12-13 08:07:55,327 INFO [train.py:421] (4/8) Epoch 9, batch 22800, loss[loss=2.167, over 5810.00 frames. , ppl: 8.735286223863342] tot_loss[loss=2.273, over 5475887.96 frames. , ppl: 9.709091032509845], batch size: 70
+2022-12-13 08:09:36,094 INFO [train.py:421] (4/8) Epoch 9, batch 23000, loss[loss=2.201, over 4550.00 frames. , ppl: 9.036627654103963] tot_loss[loss=2.274, over 5445701.80 frames. , ppl: 9.71656077195799], batch size: 70
+2022-12-13 08:09:36,094 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:09:36,839 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.639090273243687
+2022-12-13 08:11:16,443 INFO [train.py:421] (4/8) Epoch 9, batch 23200, loss[loss=2.322, over 2310.00 frames. , ppl: 10.191026992341666] tot_loss[loss=2.273, over 5481453.66 frames. , ppl: 9.70848606931702], batch size: 70
+2022-12-13 08:12:57,332 INFO [train.py:421] (4/8) Epoch 9, batch 23400, loss[loss=2.239, over 4200.00 frames. , ppl: 9.38642337018822] tot_loss[loss=2.274, over 5458121.67 frames. , ppl: 9.720287222891717], batch size: 70
+2022-12-13 08:14:44,681 INFO [train.py:421] (4/8) Epoch 9, batch 23600, loss[loss=2.569, over 840.00 frames. , ppl: 13.047401347044786] tot_loss[loss=2.273, over 5496289.06 frames. , ppl: 9.708220950859685], batch size: 70
+2022-12-13 08:16:23,460 INFO [train.py:421] (4/8) Epoch 9, batch 23800, loss[loss=2.431, over 980.00 frames. , ppl: 11.375095713160661] tot_loss[loss=2.273, over 5471253.35 frames. , ppl: 9.706258447419334], batch size: 70
+2022-12-13 08:18:03,862 INFO [train.py:421] (4/8) Epoch 9, batch 24000, loss[loss=2.708, over 770.00 frames. , ppl: 14.996406378089024] tot_loss[loss=2.273, over 5459545.71 frames. , ppl: 9.705152123519484], batch size: 70
+2022-12-13 08:18:03,863 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:18:04,611 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.650770254134942
+2022-12-13 08:19:39,211 INFO [train.py:421] (4/8) Epoch 9, batch 24200, loss[loss=2.311, over 2730.00 frames. , ppl: 10.087279586035226] tot_loss[loss=2.274, over 5433016.49 frames. , ppl: 9.714332686083642], batch size: 70
+2022-12-13 08:21:19,435 INFO [train.py:421] (4/8) Epoch 9, batch 24400, loss[loss=2.589, over 840.00 frames. , ppl: 13.310075723568069] tot_loss[loss=2.272, over 5448944.20 frames. , ppl: 9.702910893396677], batch size: 70
+2022-12-13 08:23:01,199 INFO [train.py:421] (4/8) Epoch 9, batch 24600, loss[loss=2.473, over 1470.00 frames. , ppl: 11.862061029664398] tot_loss[loss=2.272, over 5472638.07 frames. , ppl: 9.697214080025688], batch size: 70
+2022-12-13 08:24:43,349 INFO [train.py:421] (4/8) Epoch 9, batch 24800, loss[loss=2.517, over 1050.00 frames. , ppl: 12.38683087520535] tot_loss[loss=2.271, over 5500042.39 frames. , ppl: 9.68994768995308], batch size: 70
+2022-12-13 08:26:25,431 INFO [train.py:421] (4/8) Epoch 9, batch 25000, loss[loss=2.329, over 3080.00 frames. , ppl: 10.266419724505973] tot_loss[loss=2.272, over 5472556.76 frames. , ppl: 9.700094437654638], batch size: 70
+2022-12-13 08:26:25,431 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:26:26,196 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.640315849779926
+2022-12-13 08:28:06,125 INFO [train.py:421] (4/8) Epoch 9, batch 25200, loss[loss=2.438, over 840.00 frames. , ppl: 11.452685956615786] tot_loss[loss=2.272, over 5486895.21 frames. , ppl: 9.697423065276075], batch size: 70
+2022-12-13 08:29:44,792 INFO [train.py:421] (4/8) Epoch 9, batch 25400, loss[loss=2.705, over 770.00 frames. , ppl: 14.955907074721178] tot_loss[loss=2.272, over 5477063.44 frames. , ppl: 9.702914263750174], batch size: 70
+2022-12-13 08:31:25,697 INFO [train.py:421] (4/8) Epoch 9, batch 25600, loss[loss=2.312, over 3710.00 frames. , ppl: 10.096744906443828] tot_loss[loss=2.274, over 5449890.79 frames. , ppl: 9.71353059111366], batch size: 70
+2022-12-13 08:33:08,240 INFO [train.py:421] (4/8) Epoch 9, batch 25800, loss[loss=2.628, over 840.00 frames. , ppl: 13.839681465564984] tot_loss[loss=2.273, over 5476288.03 frames. , ppl: 9.70775388834503], batch size: 70
+2022-12-13 08:34:47,564 INFO [train.py:421] (4/8) Epoch 9, batch 26000, loss[loss=2.338, over 1960.00 frames. , ppl: 10.360466343619427] tot_loss[loss=2.273, over 5470711.91 frames. , ppl: 9.704514350739133], batch size: 70
+2022-12-13 08:34:47,565 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:34:48,310 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.645219714400634
+2022-12-13 08:36:28,484 INFO [train.py:421] (4/8) Epoch 9, batch 26200, loss[loss=2.552, over 1050.00 frames. , ppl: 12.829803121385847] tot_loss[loss=2.273, over 5453084.59 frames. , ppl: 9.710773950632374], batch size: 70
+2022-12-13 08:38:06,093 INFO [train.py:421] (4/8) Epoch 9, batch 26400, loss[loss=2.336, over 2100.00 frames. , ppl: 10.341672853078737] tot_loss[loss=2.272, over 5497704.47 frames. , ppl: 9.701055289135127], batch size: 70
+2022-12-13 08:39:46,906 INFO [train.py:421] (4/8) Epoch 9, batch 26600, loss[loss=2.362, over 1680.00 frames. , ppl: 10.612167875791052] tot_loss[loss=2.272, over 5504354.40 frames. , ppl: 9.697317958300706], batch size: 70
+2022-12-13 08:41:30,300 INFO [train.py:421] (4/8) Epoch 9, batch 26800, loss[loss=2.644, over 770.00 frames. , ppl: 14.070970326238035] tot_loss[loss=2.271, over 5513470.17 frames. , ppl: 9.693267210151049], batch size: 70
+2022-12-13 08:43:11,278 INFO [train.py:421] (4/8) Epoch 9, batch 27000, loss[loss=2.359, over 1330.00 frames. , ppl: 10.58516771570057] tot_loss[loss=2.271, over 5542848.59 frames. , ppl: 9.686534657440827], batch size: 70
+2022-12-13 08:43:11,278 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:43:12,039 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.652885925719394
+2022-12-13 08:44:53,159 INFO [train.py:421] (4/8) Epoch 9, batch 27200, loss[loss=2.304, over 2240.00 frames. , ppl: 10.014578038103076] tot_loss[loss=2.269, over 5595734.12 frames. , ppl: 9.673239160811809], batch size: 70
+2022-12-13 08:46:33,215 INFO [train.py:421] (4/8) Epoch 9, batch 27400, loss[loss=2.525, over 910.00 frames. , ppl: 12.485983462259458] tot_loss[loss=2.269, over 5613338.51 frames. , ppl: 9.674170487000575], batch size: 70
+2022-12-13 08:48:13,489 INFO [train.py:421] (4/8) Epoch 9, batch 27600, loss[loss=2.537, over 980.00 frames. , ppl: 12.646921215937853] tot_loss[loss=2.269, over 5623832.84 frames. , ppl: 9.672264711243825], batch size: 70
+2022-12-13 08:49:50,178 INFO [train.py:421] (4/8) Epoch 9, batch 27800, loss[loss=2.21, over 8610.00 frames. , ppl: 9.118032258452768] tot_loss[loss=2.27, over 5582981.71 frames. , ppl: 9.683576533416145], batch size: 70
+2022-12-13 08:51:34,722 INFO [train.py:421] (4/8) Epoch 9, batch 28000, loss[loss=2.267, over 6720.00 frames. , ppl: 9.647672141265003] tot_loss[loss=2.271, over 5539499.53 frames. , ppl: 9.692417598816958], batch size: 70
+2022-12-13 08:51:34,722 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:51:35,469 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.631405082310689
+2022-12-13 08:53:13,314 INFO [train.py:421] (4/8) Epoch 9, batch 28200, loss[loss=2.201, over 5180.00 frames. , ppl: 9.031850796242109] tot_loss[loss=2.273, over 5504001.26 frames. , ppl: 9.704453819513711], batch size: 70
+2022-12-13 08:54:53,461 INFO [train.py:421] (4/8) Epoch 9, batch 28400, loss[loss=2.292, over 2030.00 frames. , ppl: 9.893090289309496] tot_loss[loss=2.274, over 5435457.23 frames. , ppl: 9.720656803734384], batch size: 70
+2022-12-13 08:56:32,697 INFO [train.py:421] (4/8) Epoch 9, batch 28600, loss[loss=2.289, over 1820.00 frames. , ppl: 9.868993356304964] tot_loss[loss=2.275, over 5405358.73 frames. , ppl: 9.727646397694377], batch size: 70
+2022-12-13 08:58:14,669 INFO [train.py:421] (4/8) Epoch 9, batch 28800, loss[loss=2.292, over 1890.00 frames. , ppl: 9.892219742493436] tot_loss[loss=2.275, over 5398962.92 frames. , ppl: 9.727502836317035], batch size: 70
+2022-12-13 08:59:57,440 INFO [train.py:421] (4/8) Epoch 9, batch 29000, loss[loss=2.213, over 2170.00 frames. , ppl: 9.147462903201042] tot_loss[loss=2.274, over 5435154.31 frames. , ppl: 9.721725096576652], batch size: 70
+2022-12-13 08:59:57,441 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 08:59:58,199 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.650904523356159
+2022-12-13 09:01:42,282 INFO [train.py:421] (4/8) Epoch 9, batch 29200, loss[loss=2.263, over 5390.00 frames. , ppl: 9.614742102214299] tot_loss[loss=2.274, over 5441899.14 frames. , ppl: 9.722995283388595], batch size: 70
+2022-12-13 09:03:21,199 INFO [train.py:421] (4/8) Epoch 9, batch 29400, loss[loss=2.279, over 2590.00 frames. , ppl: 9.762539644106257] tot_loss[loss=2.274, over 5429834.81 frames. , ppl: 9.722569238499855], batch size: 70
+2022-12-13 09:05:00,146 INFO [train.py:421] (4/8) Epoch 9, batch 29600, loss[loss=2.218, over 2730.00 frames. , ppl: 9.193043118421942] tot_loss[loss=2.275, over 5401004.41 frames. , ppl: 9.730187303101724], batch size: 70
+2022-12-13 09:06:41,435 INFO [train.py:421] (4/8) Epoch 9, batch 29800, loss[loss=2.179, over 5670.00 frames. , ppl: 8.837731781622775] tot_loss[loss=2.274, over 5446979.90 frames. , ppl: 9.715317822186371], batch size: 70
+2022-12-13 09:08:24,042 INFO [train.py:421] (4/8) Epoch 9, batch 30000, loss[loss=2.561, over 1260.00 frames. , ppl: 12.952680331383402] tot_loss[loss=2.274, over 5448824.15 frames. , ppl: 9.71658730966429], batch size: 70
+2022-12-13 09:08:24,043 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:08:24,805 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637924764691014
+2022-12-13 09:10:04,498 INFO [train.py:421] (4/8) Epoch 9, batch 30200, loss[loss=2.286, over 2310.00 frames. , ppl: 9.83793774815671] tot_loss[loss=2.273, over 5468839.70 frames. , ppl: 9.712847108247], batch size: 70
+2022-12-13 09:11:45,358 INFO [train.py:421] (4/8) Epoch 9, batch 30400, loss[loss=2.2, over 6580.00 frames. , ppl: 9.027285469039642] tot_loss[loss=2.274, over 5451774.15 frames. , ppl: 9.720925781032294], batch size: 70
+2022-12-13 09:13:24,106 INFO [train.py:421] (4/8) Epoch 9, batch 30600, loss[loss=2.159, over 2450.00 frames. , ppl: 8.663816874329358] tot_loss[loss=2.273, over 5473812.37 frames. , ppl: 9.711042663889367], batch size: 70
+2022-12-13 09:15:01,162 INFO [train.py:421] (4/8) Epoch 9, batch 30800, loss[loss=2.158, over 4410.00 frames. , ppl: 8.658126054104542] tot_loss[loss=2.274, over 5441380.09 frames. , ppl: 9.713953682469164], batch size: 70
+2022-12-13 09:16:34,883 INFO [train.py:421] (4/8) Epoch 9, batch 31000, loss[loss=2.376, over 2030.00 frames. , ppl: 10.758375375002023] tot_loss[loss=2.274, over 5434619.89 frames. , ppl: 9.71780859888689], batch size: 70
+2022-12-13 09:16:34,884 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:16:35,648 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.66225130295525
+2022-12-13 09:18:16,658 INFO [train.py:421] (4/8) Epoch 9, batch 31200, loss[loss=2.587, over 910.00 frames. , ppl: 13.28736685616897] tot_loss[loss=2.274, over 5432487.54 frames. , ppl: 9.722887430806184], batch size: 70
+2022-12-13 09:19:54,608 INFO [train.py:421] (4/8) Epoch 9, batch 31400, loss[loss=2.167, over 4900.00 frames. , ppl: 8.73446406156027] tot_loss[loss=2.274, over 5438460.93 frames. , ppl: 9.717182200015689], batch size: 70
+2022-12-13 09:21:34,364 INFO [train.py:421] (4/8) Epoch 9, batch 31600, loss[loss=2.391, over 1890.00 frames. , ppl: 10.925551704830896] tot_loss[loss=2.274, over 5453251.90 frames. , ppl: 9.715121072738082], batch size: 70
+2022-12-13 09:23:15,091 INFO [train.py:421] (4/8) Epoch 9, batch 31800, loss[loss=2.169, over 910.00 frames. , ppl: 8.745389538089597] tot_loss[loss=2.272, over 5475405.06 frames. , ppl: 9.70097930987425], batch size: 70
+2022-12-13 09:24:55,717 INFO [train.py:421] (4/8) Epoch 9, batch 32000, loss[loss=2.366, over 2030.00 frames. , ppl: 10.657253749473046] tot_loss[loss=2.273, over 5463384.26 frames. , ppl: 9.70571338682433], batch size: 70
+2022-12-13 09:24:55,718 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:24:56,478 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.664727097772253
+2022-12-13 09:26:38,798 INFO [train.py:421] (4/8) Epoch 9, batch 32200, loss[loss=2.312, over 3080.00 frames. , ppl: 10.096571733990428] tot_loss[loss=2.272, over 5491512.15 frames. , ppl: 9.699309089638874], batch size: 70
+2022-12-13 09:28:15,342 INFO [train.py:421] (4/8) Epoch 9, batch 32400, loss[loss=2.24, over 2800.00 frames. , ppl: 9.389343441087739] tot_loss[loss=2.27, over 5554158.48 frames. , ppl: 9.683943457493362], batch size: 70
+2022-12-13 09:29:54,356 INFO [train.py:421] (4/8) Epoch 9, batch 32600, loss[loss=2.472, over 1470.00 frames. , ppl: 11.841827899023977] tot_loss[loss=2.271, over 5544002.88 frames. , ppl: 9.68964519889867], batch size: 70
+2022-12-13 09:31:30,579 INFO [train.py:421] (4/8) Epoch 9, batch 32800, loss[loss=2.358, over 1960.00 frames. , ppl: 10.571769367785357] tot_loss[loss=2.271, over 5544539.82 frames. , ppl: 9.693664919489423], batch size: 70
+2022-12-13 09:33:11,515 INFO [train.py:421] (4/8) Epoch 9, batch 33000, loss[loss=2.235, over 4340.00 frames. , ppl: 9.342147121133788] tot_loss[loss=2.272, over 5556290.06 frames. , ppl: 9.698177285691694], batch size: 70
+2022-12-13 09:33:11,516 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:33:12,263 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.644468844695679
+2022-12-13 09:34:49,445 INFO [train.py:421] (4/8) Epoch 9, batch 33200, loss[loss=2.268, over 2940.00 frames. , ppl: 9.660266462273457] tot_loss[loss=2.272, over 5543032.62 frames. , ppl: 9.701581959068141], batch size: 70
+2022-12-13 09:36:29,258 INFO [train.py:421] (4/8) Epoch 9, batch 33400, loss[loss=2.584, over 980.00 frames. , ppl: 13.250927925734072] tot_loss[loss=2.273, over 5524792.58 frames. , ppl: 9.709518786006036], batch size: 70
+2022-12-13 09:38:10,183 INFO [train.py:421] (4/8) Epoch 9, batch 33600, loss[loss=2.19, over 5390.00 frames. , ppl: 8.93113057029807] tot_loss[loss=2.273, over 5550355.65 frames. , ppl: 9.707269565492783], batch size: 70
+2022-12-13 09:39:46,992 INFO [train.py:421] (4/8) Epoch 9, batch 33800, loss[loss=2.484, over 1330.00 frames. , ppl: 11.988317211358163] tot_loss[loss=2.273, over 5566204.38 frames. , ppl: 9.706413128970189], batch size: 70
+2022-12-13 09:41:27,519 INFO [train.py:421] (4/8) Epoch 9, batch 34000, loss[loss=2.398, over 1120.00 frames. , ppl: 10.996191686231697] tot_loss[loss=2.271, over 5585602.22 frames. , ppl: 9.691218921198649], batch size: 70
+2022-12-13 09:41:27,520 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:41:28,282 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.6631322756689
+2022-12-13 09:43:09,901 INFO [train.py:421] (4/8) Epoch 9, batch 34200, loss[loss=2.279, over 3780.00 frames. , ppl: 9.763780553005073] tot_loss[loss=2.271, over 5592846.02 frames. , ppl: 9.68958991194787], batch size: 70
+2022-12-13 09:44:49,447 INFO [train.py:421] (4/8) Epoch 9, batch 34400, loss[loss=2.218, over 3430.00 frames. , ppl: 9.187782870136104] tot_loss[loss=2.271, over 5630132.31 frames. , ppl: 9.684266144149117], batch size: 70
+2022-12-13 09:46:28,568 INFO [train.py:421] (4/8) Epoch 9, batch 34600, loss[loss=2.333, over 2730.00 frames. , ppl: 10.31339473021858] tot_loss[loss=2.271, over 5616140.55 frames. , ppl: 9.68587191903504], batch size: 70
+2022-12-13 09:48:11,554 INFO [train.py:421] (4/8) Epoch 9, batch 34800, loss[loss=2.216, over 6580.00 frames. , ppl: 9.169941261553285] tot_loss[loss=2.271, over 5599352.74 frames. , ppl: 9.688498936773003], batch size: 70
+2022-12-13 09:49:50,488 INFO [train.py:421] (4/8) Epoch 9, batch 35000, loss[loss=2.209, over 2730.00 frames. , ppl: 9.11016097951145] tot_loss[loss=2.271, over 5608524.34 frames. , ppl: 9.691237635107797], batch size: 70
+2022-12-13 09:49:50,488 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:49:51,240 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.644501676179262
+2022-12-13 09:51:35,363 INFO [train.py:421] (4/8) Epoch 9, batch 35200, loss[loss=2.235, over 1680.00 frames. , ppl: 9.34280878609065] tot_loss[loss=2.271, over 5608246.59 frames. , ppl: 9.685042787425237], batch size: 70
+2022-12-13 09:53:15,453 INFO [train.py:421] (4/8) Epoch 9, batch 35400, loss[loss=2.317, over 2170.00 frames. , ppl: 10.140382634885482] tot_loss[loss=2.27, over 5627343.62 frames. , ppl: 9.677659183224453], batch size: 70
+2022-12-13 09:54:55,419 INFO [train.py:421] (4/8) Epoch 9, batch 35600, loss[loss=4.035, over 350.00 frames. , ppl: 56.52170484784358] tot_loss[loss=2.27, over 5594199.96 frames. , ppl: 9.682899860083651], batch size: 70
+2022-12-13 09:56:35,170 INFO [train.py:421] (4/8) Epoch 9, batch 35800, loss[loss=2.477, over 1540.00 frames. , ppl: 11.91113583450228] tot_loss[loss=2.271, over 5594129.75 frames. , ppl: 9.685200501852513], batch size: 70
+2022-12-13 09:58:16,595 INFO [train.py:421] (4/8) Epoch 9, batch 36000, loss[loss=2.759, over 700.00 frames. , ppl: 15.784976977507618] tot_loss[loss=2.271, over 5576219.18 frames. , ppl: 9.6911186979249], batch size: 70
+2022-12-13 09:58:16,596 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 09:58:17,362 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.647108567176648
+2022-12-13 09:59:55,817 INFO [train.py:421] (4/8) Epoch 9, batch 36200, loss[loss=3.506, over 420.00 frames. , ppl: 33.310986809380836] tot_loss[loss=2.271, over 5548707.85 frames. , ppl: 9.690274758316827], batch size: 70
+2022-12-13 10:01:36,713 INFO [train.py:421] (4/8) Epoch 9, batch 36400, loss[loss=2.29, over 2520.00 frames. , ppl: 9.871902743157534] tot_loss[loss=2.272, over 5549236.86 frames. , ppl: 9.694338049827815], batch size: 70
+2022-12-13 10:03:20,896 INFO [train.py:421] (4/8) Epoch 9, batch 36600, loss[loss=2.614, over 840.00 frames. , ppl: 13.648279307633548] tot_loss[loss=2.272, over 5530095.98 frames. , ppl: 9.69599219053233], batch size: 70
+2022-12-13 10:05:04,601 INFO [train.py:421] (4/8) Epoch 9, batch 36800, loss[loss=2.433, over 1400.00 frames. , ppl: 11.390055531427667] tot_loss[loss=2.273, over 5497384.53 frames. , ppl: 9.706858048648744], batch size: 70
+2022-12-13 10:06:43,622 INFO [train.py:421] (4/8) Epoch 9, batch 37000, loss[loss=2.366, over 3010.00 frames. , ppl: 10.657770916563967] tot_loss[loss=2.274, over 5462261.40 frames. , ppl: 9.715632695617582], batch size: 70
+2022-12-13 10:06:43,623 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:06:44,368 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.65039602425002
+2022-12-13 10:08:24,181 INFO [train.py:421] (4/8) Epoch 9, batch 37200, loss[loss=2.481, over 1120.00 frames. , ppl: 11.954555903973825] tot_loss[loss=2.273, over 5475379.28 frames. , ppl: 9.712661001025669], batch size: 70
+2022-12-13 10:10:01,442 INFO [train.py:421] (4/8) Epoch 9, batch 37400, loss[loss=2.132, over 6510.00 frames. , ppl: 8.428803900725539] tot_loss[loss=2.273, over 5486985.43 frames. , ppl: 9.711962335158521], batch size: 70
+2022-12-13 10:11:39,958 INFO [train.py:421] (4/8) Epoch 9, batch 37600, loss[loss=2.863, over 630.00 frames. , ppl: 17.510286569552022] tot_loss[loss=2.272, over 5528380.12 frames. , ppl: 9.699136286992527], batch size: 70
+2022-12-13 10:13:18,222 INFO [train.py:421] (4/8) Epoch 9, batch 37800, loss[loss=2.342, over 2380.00 frames. , ppl: 10.404457046696184] tot_loss[loss=2.273, over 5489183.05 frames. , ppl: 9.712874248418098], batch size: 70
+2022-12-13 10:14:59,565 INFO [train.py:421] (4/8) Epoch 9, batch 38000, loss[loss=2.225, over 3360.00 frames. , ppl: 9.252364053865701] tot_loss[loss=2.272, over 5519215.21 frames. , ppl: 9.703176562130766], batch size: 70
+2022-12-13 10:14:59,566 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:15:00,328 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637123113571448
+2022-12-13 10:16:42,063 INFO [train.py:421] (4/8) Epoch 9, batch 38200, loss[loss=2.315, over 2170.00 frames. , ppl: 10.126763003428122] tot_loss[loss=2.273, over 5510753.11 frames. , ppl: 9.705496297164867], batch size: 70
+2022-12-13 10:18:21,502 INFO [train.py:421] (4/8) Epoch 9, batch 38400, loss[loss=2.502, over 1050.00 frames. , ppl: 12.211875185376607] tot_loss[loss=2.274, over 5503180.91 frames. , ppl: 9.713807322800506], batch size: 70
+2022-12-13 10:20:03,169 INFO [train.py:421] (4/8) Epoch 9, batch 38600, loss[loss=2.355, over 1400.00 frames. , ppl: 10.542404227844855] tot_loss[loss=2.272, over 5539603.07 frames. , ppl: 9.702719219111053], batch size: 70
+2022-12-13 10:21:45,084 INFO [train.py:421] (4/8) Epoch 9, batch 38800, loss[loss=2.176, over 4130.00 frames. , ppl: 8.81390663278147] tot_loss[loss=2.272, over 5531859.22 frames. , ppl: 9.695845850559508], batch size: 70
+2022-12-13 10:23:26,884 INFO [train.py:421] (4/8) Epoch 9, batch 39000, loss[loss=2.466, over 1890.00 frames. , ppl: 11.775970943818033] tot_loss[loss=2.273, over 5516183.16 frames. , ppl: 9.707843252067136], batch size: 70
+2022-12-13 10:23:26,885 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:23:27,632 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.646513174570288
+2022-12-13 10:25:07,541 INFO [train.py:421] (4/8) Epoch 9, batch 39200, loss[loss=2.532, over 840.00 frames. , ppl: 12.574549885417138] tot_loss[loss=2.273, over 5524428.99 frames. , ppl: 9.704493101248094], batch size: 70
+2022-12-13 10:26:46,237 INFO [train.py:421] (4/8) Epoch 9, batch 39400, loss[loss=2.388, over 1750.00 frames. , ppl: 10.896175219439892] tot_loss[loss=2.272, over 5517715.07 frames. , ppl: 9.70110644869211], batch size: 70
+2022-12-13 10:28:27,571 INFO [train.py:421] (4/8) Epoch 9, batch 39600, loss[loss=2.348, over 1540.00 frames. , ppl: 10.462658273686571] tot_loss[loss=2.273, over 5508124.09 frames. , ppl: 9.708301931527345], batch size: 70
+2022-12-13 10:30:04,664 INFO [train.py:421] (4/8) Epoch 9, batch 39800, loss[loss=2.942, over 560.00 frames. , ppl: 18.95146907533392] tot_loss[loss=2.274, over 5471257.33 frames. , ppl: 9.722641120209717], batch size: 70
+2022-12-13 10:31:44,874 INFO [train.py:421] (4/8) Epoch 9, batch 40000, loss[loss=2.195, over 6790.00 frames. , ppl: 8.983644961860854] tot_loss[loss=2.274, over 5476579.04 frames. , ppl: 9.718312025219303], batch size: 70
+2022-12-13 10:31:44,875 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:31:45,622 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.648730734417011
+2022-12-13 10:33:26,284 INFO [train.py:421] (4/8) Epoch 9, batch 40200, loss[loss=2.169, over 7000.00 frames. , ppl: 8.752665293240005] tot_loss[loss=2.275, over 5473108.84 frames. , ppl: 9.725367517282763], batch size: 70
+2022-12-13 10:35:02,369 INFO [train.py:421] (4/8) Epoch 9, batch 40400, loss[loss=2.188, over 7350.00 frames. , ppl: 8.914390483891939] tot_loss[loss=2.275, over 5460834.27 frames. , ppl: 9.73024441674235], batch size: 70
+2022-12-13 10:36:42,721 INFO [train.py:421] (4/8) Epoch 9, batch 40600, loss[loss=2.363, over 2310.00 frames. , ppl: 10.623653509049566] tot_loss[loss=2.275, over 5433952.59 frames. , ppl: 9.73037251603799], batch size: 70
+2022-12-13 10:38:21,179 INFO [train.py:421] (4/8) Epoch 9, batch 40800, loss[loss=2.419, over 1050.00 frames. , ppl: 11.234438312721727] tot_loss[loss=2.275, over 5414853.81 frames. , ppl: 9.732048823638502], batch size: 70
+2022-12-13 10:39:58,028 INFO [train.py:421] (4/8) Epoch 9, batch 41000, loss[loss=2.76, over 630.00 frames. , ppl: 15.80520255879073] tot_loss[loss=2.276, over 5403241.48 frames. , ppl: 9.736252249952914], batch size: 70
+2022-12-13 10:39:58,028 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:39:58,774 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.268, over 211138.00 frames. , ppl: 9.663169461348792
+2022-12-13 10:41:39,550 INFO [train.py:421] (4/8) Epoch 9, batch 41200, loss[loss=2.323, over 2310.00 frames. , ppl: 10.203990290652156] tot_loss[loss=2.275, over 5403848.51 frames. , ppl: 9.730817418400969], batch size: 70
+2022-12-13 10:43:16,339 INFO [train.py:421] (4/8) Epoch 9, batch 41400, loss[loss=2.247, over 2940.00 frames. , ppl: 9.461617674753485] tot_loss[loss=2.276, over 5395418.46 frames. , ppl: 9.732899573977814], batch size: 70
+2022-12-13 10:44:58,173 INFO [train.py:421] (4/8) Epoch 9, batch 41600, loss[loss=2.209, over 2590.00 frames. , ppl: 9.104293165209462] tot_loss[loss=2.275, over 5417122.32 frames. , ppl: 9.727855223651302], batch size: 70
+2022-12-13 10:46:39,492 INFO [train.py:421] (4/8) Epoch 9, batch 41800, loss[loss=2.647, over 700.00 frames. , ppl: 14.111360114200519] tot_loss[loss=2.275, over 5405351.35 frames. , ppl: 9.729532869004279], batch size: 70
+2022-12-13 10:48:20,323 INFO [train.py:421] (4/8) Epoch 9, batch 42000, loss[loss=2.196, over 4970.00 frames. , ppl: 8.990308786545679] tot_loss[loss=2.276, over 5372164.34 frames. , ppl: 9.737602028951217], batch size: 70
+2022-12-13 10:48:20,324 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:48:21,082 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635082199742264
+2022-12-13 10:50:03,519 INFO [train.py:421] (4/8) Epoch 9, batch 42200, loss[loss=2.451, over 1330.00 frames. , ppl: 11.602000794178949] tot_loss[loss=2.276, over 5395186.59 frames. , ppl: 9.735490515077915], batch size: 70
+2022-12-13 10:51:44,308 INFO [train.py:421] (4/8) Epoch 9, batch 42400, loss[loss=2.182, over 7770.00 frames. , ppl: 8.86087173395469] tot_loss[loss=2.275, over 5421897.96 frames. , ppl: 9.725561906627938], batch size: 70
+2022-12-13 10:53:20,410 INFO [train.py:421] (4/8) Epoch 9, batch 42600, loss[loss=2.125, over 4620.00 frames. , ppl: 8.372872710413311] tot_loss[loss=2.274, over 5458999.46 frames. , ppl: 9.718682675210527], batch size: 70
+2022-12-13 10:54:59,731 INFO [train.py:421] (4/8) Epoch 9, batch 42800, loss[loss=2.359, over 980.00 frames. , ppl: 10.579098782350881] tot_loss[loss=2.273, over 5498938.19 frames. , ppl: 9.705717317573772], batch size: 70
+2022-12-13 10:56:37,378 INFO [train.py:421] (4/8) Epoch 9, batch 43000, loss[loss=2.262, over 2380.00 frames. , ppl: 9.598654574780396] tot_loss[loss=2.274, over 5452915.38 frames. , ppl: 9.718634057830341], batch size: 70
+2022-12-13 10:56:37,378 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 10:56:38,138 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.269, over 211138.00 frames. , ppl: 9.668268114386317
+2022-12-13 10:58:19,733 INFO [train.py:421] (4/8) Epoch 9, batch 43200, loss[loss=2.319, over 3500.00 frames. , ppl: 10.169393107307249] tot_loss[loss=2.275, over 5409631.22 frames. , ppl: 9.730876750021418], batch size: 70
+2022-12-13 11:00:01,371 INFO [train.py:421] (4/8) Epoch 9, batch 43400, loss[loss=2.321, over 2310.00 frames. , ppl: 10.181505789795459] tot_loss[loss=2.275, over 5417839.34 frames. , ppl: 9.729715305715512], batch size: 70
+2022-12-13 11:01:45,159 INFO [train.py:421] (4/8) Epoch 9, batch 43600, loss[loss=2.405, over 1330.00 frames. , ppl: 11.08296330184365] tot_loss[loss=2.276, over 5416337.30 frames. , ppl: 9.733250432093287], batch size: 70
+2022-12-13 11:03:26,870 INFO [train.py:421] (4/8) Epoch 9, batch 43800, loss[loss=2.255, over 1820.00 frames. , ppl: 9.533459776201802] tot_loss[loss=2.276, over 5381043.34 frames. , ppl: 9.738170444770345], batch size: 70
+2022-12-13 11:05:05,109 INFO [train.py:421] (4/8) Epoch 9, batch 44000, loss[loss=2.354, over 1610.00 frames. , ppl: 10.528266128906328] tot_loss[loss=2.276, over 5400067.45 frames. , ppl: 9.73316899328424], batch size: 70
+2022-12-13 11:05:05,109 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:05:05,859 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.623468208814646
+2022-12-13 11:06:45,137 INFO [train.py:421] (4/8) Epoch 9, batch 44200, loss[loss=2.306, over 4060.00 frames. , ppl: 10.034201072656124] tot_loss[loss=2.275, over 5437359.23 frames. , ppl: 9.727748325862033], batch size: 70
+2022-12-13 11:08:25,131 INFO [train.py:421] (4/8) Epoch 9, batch 44400, loss[loss=2.134, over 9240.00 frames. , ppl: 8.452484284911534] tot_loss[loss=2.274, over 5471078.89 frames. , ppl: 9.718679680778168], batch size: 70
+2022-12-13 11:10:02,004 INFO [train.py:421] (4/8) Epoch 9, batch 44600, loss[loss=2.167, over 8400.00 frames. , ppl: 8.73400277093233] tot_loss[loss=2.275, over 5453198.25 frames. , ppl: 9.7269859702063], batch size: 70
+2022-12-13 11:11:38,150 INFO [train.py:421] (4/8) Epoch 9, batch 44800, loss[loss=2.286, over 1890.00 frames. , ppl: 9.833835026846948] tot_loss[loss=2.275, over 5413752.08 frames. , ppl: 9.73058692938066], batch size: 70
+2022-12-13 11:13:19,789 INFO [train.py:421] (4/8) Epoch 9, batch 45000, loss[loss=2.231, over 4410.00 frames. , ppl: 9.311667263190548] tot_loss[loss=2.274, over 5436398.19 frames. , ppl: 9.716753966771648], batch size: 70
+2022-12-13 11:13:19,790 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:13:20,550 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.648610776092447
+2022-12-13 11:15:01,467 INFO [train.py:421] (4/8) Epoch 9, batch 45200, loss[loss=2.265, over 2660.00 frames. , ppl: 9.631444248414063] tot_loss[loss=2.273, over 5448586.71 frames. , ppl: 9.7131398004304], batch size: 70
+2022-12-13 11:16:44,206 INFO [train.py:421] (4/8) Epoch 9, batch 45400, loss[loss=2.413, over 1680.00 frames. , ppl: 11.162419692921652] tot_loss[loss=2.273, over 5459582.33 frames. , ppl: 9.71002363215036], batch size: 70
+2022-12-13 11:18:27,671 INFO [train.py:421] (4/8) Epoch 9, batch 45600, loss[loss=2.126, over 6020.00 frames. , ppl: 8.377833460323973] tot_loss[loss=2.272, over 5487607.96 frames. , ppl: 9.699508484316427], batch size: 70
+2022-12-13 11:20:08,207 INFO [train.py:421] (4/8) Epoch 9, batch 45800, loss[loss=2.329, over 1470.00 frames. , ppl: 10.267974864399317] tot_loss[loss=2.272, over 5453105.43 frames. , ppl: 9.702983740128657], batch size: 70
+2022-12-13 11:21:46,359 INFO [train.py:421] (4/8) Epoch 9, batch 46000, loss[loss=2.365, over 2100.00 frames. , ppl: 10.647504237898582] tot_loss[loss=2.272, over 5479183.84 frames. , ppl: 9.700305069310348], batch size: 70
+2022-12-13 11:21:46,360 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:21:47,106 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.636034857644312
+2022-12-13 11:23:26,570 INFO [train.py:421] (4/8) Epoch 9, batch 46200, loss[loss=2.304, over 3500.00 frames. , ppl: 10.013502485753854] tot_loss[loss=2.272, over 5498802.63 frames. , ppl: 9.694507677649534], batch size: 70
+2022-12-13 11:25:07,660 INFO [train.py:421] (4/8) Epoch 9, batch 46400, loss[loss=2.168, over 11410.00 frames. , ppl: 8.739976219453263] tot_loss[loss=2.27, over 5537054.33 frames. , ppl: 9.682571192076356], batch size: 70
+2022-12-13 11:26:44,518 INFO [train.py:421] (4/8) Epoch 9, batch 46600, loss[loss=2.243, over 3080.00 frames. , ppl: 9.424345414845561] tot_loss[loss=2.271, over 5520141.23 frames. , ppl: 9.691851142057706], batch size: 70
+2022-12-13 11:28:29,186 INFO [train.py:421] (4/8) Epoch 9, batch 46800, loss[loss=2.655, over 840.00 frames. , ppl: 14.220954762638062] tot_loss[loss=2.271, over 5579685.19 frames. , ppl: 9.685772978886137], batch size: 70
+2022-12-13 11:30:14,119 INFO [train.py:421] (4/8) Epoch 9, batch 47000, loss[loss=2.425, over 1190.00 frames. , ppl: 11.30389673609987] tot_loss[loss=2.269, over 5627239.77 frames. , ppl: 9.672054638653245], batch size: 70
+2022-12-13 11:30:14,120 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:30:14,868 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.63486544043961
+2022-12-13 11:31:58,222 INFO [train.py:421] (4/8) Epoch 9, batch 47200, loss[loss=2.413, over 2030.00 frames. , ppl: 11.169748422355692] tot_loss[loss=2.27, over 5606556.65 frames. , ppl: 9.681464600308647], batch size: 70
+2022-12-13 11:33:39,691 INFO [train.py:421] (4/8) Epoch 9, batch 47400, loss[loss=2.239, over 8470.00 frames. , ppl: 9.382873844914222] tot_loss[loss=2.269, over 5656562.67 frames. , ppl: 9.667672184537626], batch size: 70
+2022-12-13 11:35:21,971 INFO [train.py:421] (4/8) Epoch 9, batch 47600, loss[loss=2.319, over 2450.00 frames. , ppl: 10.161702899783403] tot_loss[loss=2.27, over 5599328.11 frames. , ppl: 9.682570705723968], batch size: 70
+2022-12-13 11:37:01,535 INFO [train.py:421] (4/8) Epoch 9, batch 47800, loss[loss=2.502, over 1050.00 frames. , ppl: 12.20182769382701] tot_loss[loss=2.27, over 5584859.29 frames. , ppl: 9.683584158623917], batch size: 70
+2022-12-13 11:38:43,295 INFO [train.py:421] (4/8) Epoch 9, batch 48000, loss[loss=2.188, over 5250.00 frames. , ppl: 8.914950730383723] tot_loss[loss=2.268, over 5680704.39 frames. , ppl: 9.662332881462236], batch size: 70
+2022-12-13 11:38:43,295 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:38:44,056 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.627387380087804
+2022-12-13 11:40:21,734 INFO [train.py:421] (4/8) Epoch 9, batch 48200, loss[loss=2.343, over 2520.00 frames. , ppl: 10.416815537042462] tot_loss[loss=2.269, over 5661380.11 frames. , ppl: 9.666204488359044], batch size: 70
+2022-12-13 11:42:04,578 INFO [train.py:421] (4/8) Epoch 9, batch 48400, loss[loss=2.22, over 3080.00 frames. , ppl: 9.203441973112957] tot_loss[loss=2.269, over 5661353.25 frames. , ppl: 9.66572641603734], batch size: 70
+2022-12-13 11:43:45,249 INFO [train.py:421] (4/8) Epoch 9, batch 48600, loss[loss=4, over 350.00 frames. , ppl: 54.600473246041496] tot_loss[loss=2.27, over 5622394.30 frames. , ppl: 9.676871872624014], batch size: 70
+2022-12-13 11:45:22,550 INFO [train.py:421] (4/8) Epoch 9, batch 48800, loss[loss=2.217, over 4130.00 frames. , ppl: 9.180210357899524] tot_loss[loss=2.271, over 5588101.83 frames. , ppl: 9.686024855260403], batch size: 70
+2022-12-13 11:47:02,933 INFO [train.py:421] (4/8) Epoch 9, batch 49000, loss[loss=2.237, over 4270.00 frames. , ppl: 9.369185151161952] tot_loss[loss=2.272, over 5555544.44 frames. , ppl: 9.695015802505983], batch size: 70
+2022-12-13 11:47:02,933 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:47:03,679 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.644904227803092
+2022-12-13 11:48:42,688 INFO [train.py:421] (4/8) Epoch 9, batch 49200, loss[loss=2.211, over 5950.00 frames. , ppl: 9.12720504855464] tot_loss[loss=2.272, over 5518743.77 frames. , ppl: 9.698801615337226], batch size: 70
+2022-12-13 11:50:20,811 INFO [train.py:421] (4/8) Epoch 9, batch 49400, loss[loss=2.28, over 5460.00 frames. , ppl: 9.77183230127551] tot_loss[loss=2.271, over 5555316.25 frames. , ppl: 9.687406529898757], batch size: 70
+2022-12-13 11:52:05,013 INFO [train.py:421] (4/8) Epoch 9, batch 49600, loss[loss=2.181, over 9100.00 frames. , ppl: 8.85504067520833] tot_loss[loss=2.27, over 5577095.40 frames. , ppl: 9.683387974801848], batch size: 70
+2022-12-13 11:53:47,506 INFO [train.py:421] (4/8) Epoch 9, batch 49800, loss[loss=2.384, over 2730.00 frames. , ppl: 10.843765989755015] tot_loss[loss=2.27, over 5568202.91 frames. , ppl: 9.680685541139242], batch size: 70
+2022-12-13 11:55:28,993 INFO [train.py:421] (4/8) Epoch 9, batch 50000, loss[loss=2.317, over 3010.00 frames. , ppl: 10.143412219414309] tot_loss[loss=2.271, over 5518501.59 frames. , ppl: 9.689044916744116], batch size: 70
+2022-12-13 11:55:28,994 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 11:55:29,755 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.643979240614769
+2022-12-13 11:57:09,484 INFO [train.py:421] (4/8) Epoch 9, batch 50200, loss[loss=2.319, over 2100.00 frames. , ppl: 10.168529142837182] tot_loss[loss=2.272, over 5502313.02 frames. , ppl: 9.69589764807001], batch size: 70
+2022-12-13 11:58:51,076 INFO [train.py:421] (4/8) Epoch 9, batch 50400, loss[loss=2.322, over 2240.00 frames. , ppl: 10.195349403683176] tot_loss[loss=2.272, over 5507429.09 frames. , ppl: 9.696314083184538], batch size: 70
+2022-12-13 12:00:30,170 INFO [train.py:421] (4/8) Epoch 9, batch 50600, loss[loss=2.249, over 2800.00 frames. , ppl: 9.477812196277863] tot_loss[loss=2.273, over 5477239.77 frames. , ppl: 9.704547772319822], batch size: 70
+2022-12-13 12:02:07,593 INFO [train.py:421] (4/8) Epoch 9, batch 50800, loss[loss=2.448, over 1470.00 frames. , ppl: 11.56416031930162] tot_loss[loss=2.275, over 5416017.12 frames. , ppl: 9.724494670898956], batch size: 70
+2022-12-13 12:03:46,555 INFO [train.py:421] (4/8) Epoch 9, batch 51000, loss[loss=2.447, over 910.00 frames. , ppl: 11.549512179670002] tot_loss[loss=2.275, over 5417709.03 frames. , ppl: 9.727842023556763], batch size: 70
+2022-12-13 12:03:46,555 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:03:47,301 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.639113099787668
+2022-12-13 12:05:25,875 INFO [train.py:421] (4/8) Epoch 9, batch 51200, loss[loss=4.207, over 350.00 frames. , ppl: 67.14911922557341] tot_loss[loss=2.275, over 5414009.81 frames. , ppl: 9.731697129622164], batch size: 70
+2022-12-13 12:07:04,769 INFO [train.py:421] (4/8) Epoch 9, batch 51400, loss[loss=2.487, over 1120.00 frames. , ppl: 12.025691121791622] tot_loss[loss=2.275, over 5430110.98 frames. , ppl: 9.728595611895813], batch size: 70
+2022-12-13 12:08:46,931 INFO [train.py:421] (4/8) Epoch 9, batch 51600, loss[loss=2.521, over 840.00 frames. , ppl: 12.438488163559946] tot_loss[loss=2.276, over 5423939.82 frames. , ppl: 9.732979530014198], batch size: 70
+2022-12-13 12:10:24,704 INFO [train.py:421] (4/8) Epoch 9, batch 51800, loss[loss=2.322, over 840.00 frames. , ppl: 10.191934872533647] tot_loss[loss=2.275, over 5435218.74 frames. , ppl: 9.7255692875002], batch size: 70
+2022-12-13 12:12:05,979 INFO [train.py:421] (4/8) Epoch 9, batch 52000, loss[loss=2.31, over 1890.00 frames. , ppl: 10.077303160881648] tot_loss[loss=2.275, over 5439422.72 frames. , ppl: 9.723170599909745], batch size: 70
+2022-12-13 12:12:05,980 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:12:06,730 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.646610262463758
+2022-12-13 12:13:45,404 INFO [train.py:421] (4/8) Epoch 9, batch 52200, loss[loss=3.01, over 560.00 frames. , ppl: 20.283936673773123] tot_loss[loss=2.276, over 5391865.06 frames. , ppl: 9.737305863041287], batch size: 70
+2022-12-13 12:15:24,528 INFO [train.py:421] (4/8) Epoch 9, batch 52400, loss[loss=2.493, over 980.00 frames. , ppl: 12.096289424157556] tot_loss[loss=2.276, over 5389483.11 frames. , ppl: 9.741499587794742], batch size: 70
+2022-12-13 12:17:07,157 INFO [train.py:421] (4/8) Epoch 9, batch 52600, loss[loss=2.445, over 980.00 frames. , ppl: 11.527569450935365] tot_loss[loss=2.275, over 5439143.57 frames. , ppl: 9.729080852938846], batch size: 70
+2022-12-13 12:18:50,859 INFO [train.py:421] (4/8) Epoch 9, batch 52800, loss[loss=2.167, over 8610.00 frames. , ppl: 8.729118562234616] tot_loss[loss=2.276, over 5407590.61 frames. , ppl: 9.738084500028819], batch size: 70
+2022-12-13 12:20:28,986 INFO [train.py:421] (4/8) Epoch 9, batch 53000, loss[loss=2.396, over 1750.00 frames. , ppl: 10.974176498875519] tot_loss[loss=2.276, over 5421670.85 frames. , ppl: 9.736455442356844], batch size: 70
+2022-12-13 12:20:28,987 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:20:29,752 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.642737499169275
+2022-12-13 12:22:10,639 INFO [train.py:421] (4/8) Epoch 9, batch 53200, loss[loss=2.241, over 2800.00 frames. , ppl: 9.405592023215117] tot_loss[loss=2.277, over 5384300.45 frames. , ppl: 9.744541787317578], batch size: 70
+2022-12-13 12:23:51,913 INFO [train.py:421] (4/8) Epoch 9, batch 53400, loss[loss=2.566, over 980.00 frames. , ppl: 13.01213772833235] tot_loss[loss=2.277, over 5366007.28 frames. , ppl: 9.74490919370099], batch size: 70
+2022-12-13 12:25:35,097 INFO [train.py:421] (4/8) Epoch 9, batch 53600, loss[loss=2.41, over 1050.00 frames. , ppl: 11.128487154504867] tot_loss[loss=2.276, over 5370888.13 frames. , ppl: 9.73656190146111], batch size: 70
+2022-12-13 12:27:15,424 INFO [train.py:421] (4/8) Epoch 9, batch 53800, loss[loss=2.19, over 3360.00 frames. , ppl: 8.934725937812265] tot_loss[loss=2.274, over 5442689.96 frames. , ppl: 9.714051662346954], batch size: 70
+2022-12-13 12:28:55,762 INFO [train.py:421] (4/8) Epoch 9, batch 54000, loss[loss=2.317, over 2240.00 frames. , ppl: 10.14288044088573] tot_loss[loss=2.273, over 5456159.60 frames. , ppl: 9.713087849137096], batch size: 70
+2022-12-13 12:28:55,763 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:28:56,510 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.633422404126998
+2022-12-13 12:30:35,815 INFO [train.py:421] (4/8) Epoch 9, batch 54200, loss[loss=2.544, over 1190.00 frames. , ppl: 12.727841685577783] tot_loss[loss=2.275, over 5429621.82 frames. , ppl: 9.724596571406362], batch size: 70
+2022-12-13 12:32:16,552 INFO [train.py:421] (4/8) Epoch 9, batch 54400, loss[loss=2.8, over 630.00 frames. , ppl: 16.451562613625683] tot_loss[loss=2.274, over 5470759.78 frames. , ppl: 9.72174371613905], batch size: 70
+2022-12-13 12:33:55,484 INFO [train.py:421] (4/8) Epoch 9, batch 54600, loss[loss=2.328, over 2940.00 frames. , ppl: 10.259848790657868] tot_loss[loss=2.274, over 5450539.55 frames. , ppl: 9.721527119922227], batch size: 70
+2022-12-13 12:35:38,800 INFO [train.py:421] (4/8) Epoch 9, batch 54800, loss[loss=2.364, over 1330.00 frames. , ppl: 10.629095986020113] tot_loss[loss=2.273, over 5495518.11 frames. , ppl: 9.709674817157497], batch size: 70
+2022-12-13 12:37:17,706 INFO [train.py:421] (4/8) Epoch 9, batch 55000, loss[loss=2.301, over 2100.00 frames. , ppl: 9.98633289722943] tot_loss[loss=2.272, over 5546490.57 frames. , ppl: 9.703542296729808], batch size: 70
+2022-12-13 12:37:17,707 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:37:18,465 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.64445171527035
+2022-12-13 12:39:02,245 INFO [train.py:421] (4/8) Epoch 9, batch 55200, loss[loss=3.141, over 490.00 frames. , ppl: 23.135636007540626] tot_loss[loss=2.271, over 5596541.44 frames. , ppl: 9.684465412288006], batch size: 70
+2022-12-13 12:40:39,118 INFO [train.py:421] (4/8) Epoch 9, batch 55400, loss[loss=2.214, over 2380.00 frames. , ppl: 9.14919337985409] tot_loss[loss=2.271, over 5571072.96 frames. , ppl: 9.68477556653161], batch size: 70
+2022-12-13 12:42:18,509 INFO [train.py:421] (4/8) Epoch 9, batch 55600, loss[loss=2.426, over 1610.00 frames. , ppl: 11.317160239362117] tot_loss[loss=2.27, over 5551193.61 frames. , ppl: 9.683798279338989], batch size: 70
+2022-12-13 12:43:58,844 INFO [train.py:421] (4/8) Epoch 9, batch 55800, loss[loss=2.312, over 2660.00 frames. , ppl: 10.090406441115384] tot_loss[loss=2.272, over 5499654.94 frames. , ppl: 9.697068693835934], batch size: 70
+2022-12-13 12:45:38,588 INFO [train.py:421] (4/8) Epoch 9, batch 56000, loss[loss=2.579, over 770.00 frames. , ppl: 13.189934239417912] tot_loss[loss=2.273, over 5459563.51 frames. , ppl: 9.706101378504796], batch size: 70
+2022-12-13 12:45:38,589 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:45:39,335 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.625552251361995
+2022-12-13 12:47:19,624 INFO [train.py:421] (4/8) Epoch 9, batch 56200, loss[loss=2.516, over 770.00 frames. , ppl: 12.379646006799657] tot_loss[loss=2.273, over 5443471.20 frames. , ppl: 9.711161794324283], batch size: 70
+2022-12-13 12:48:56,802 INFO [train.py:421] (4/8) Epoch 9, batch 56400, loss[loss=2.329, over 2310.00 frames. , ppl: 10.27083462984211] tot_loss[loss=2.275, over 5394004.52 frames. , ppl: 9.72984508297421], batch size: 70
+2022-12-13 12:50:35,595 INFO [train.py:421] (4/8) Epoch 9, batch 56600, loss[loss=2.267, over 2030.00 frames. , ppl: 9.653957157943607] tot_loss[loss=2.275, over 5393912.56 frames. , ppl: 9.730382505193534], batch size: 70
+2022-12-13 12:52:13,282 INFO [train.py:421] (4/8) Epoch 9, batch 56800, loss[loss=2.267, over 3360.00 frames. , ppl: 9.645701106610682] tot_loss[loss=2.275, over 5376341.38 frames. , ppl: 9.732071479833467], batch size: 70
+2022-12-13 12:53:52,633 INFO [train.py:421] (4/8) Epoch 9, batch 57000, loss[loss=2.304, over 2450.00 frames. , ppl: 10.0150484554932] tot_loss[loss=2.275, over 5387932.96 frames. , ppl: 9.731654406918164], batch size: 70
+2022-12-13 12:53:52,633 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 12:53:53,394 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.636780791774152
+2022-12-13 12:55:34,879 INFO [train.py:421] (4/8) Epoch 9, batch 57200, loss[loss=2.163, over 4340.00 frames. , ppl: 8.698544318638271] tot_loss[loss=2.275, over 5415746.86 frames. , ppl: 9.725790021828487], batch size: 70
+2022-12-13 12:57:12,841 INFO [train.py:421] (4/8) Epoch 9, batch 57400, loss[loss=2.199, over 7490.00 frames. , ppl: 9.016705110954259] tot_loss[loss=2.274, over 5436765.36 frames. , ppl: 9.721188456191713], batch size: 70
+2022-12-13 12:58:55,749 INFO [train.py:421] (4/8) Epoch 9, batch 57600, loss[loss=2.278, over 3570.00 frames. , ppl: 9.754824696067281] tot_loss[loss=2.274, over 5488682.14 frames. , ppl: 9.713582812141503], batch size: 70
+2022-12-13 13:00:36,168 INFO [train.py:421] (4/8) Epoch 9, batch 57800, loss[loss=2.582, over 980.00 frames. , ppl: 13.228413683085028] tot_loss[loss=2.274, over 5462247.01 frames. , ppl: 9.720343900227729], batch size: 70
+2022-12-13 13:02:18,867 INFO [train.py:421] (4/8) Epoch 9, batch 58000, loss[loss=2.49, over 1960.00 frames. , ppl: 12.058388307877738] tot_loss[loss=2.274, over 5459090.19 frames. , ppl: 9.720921475626962], batch size: 70
+2022-12-13 13:02:18,867 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:02:19,612 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635025157347723
+2022-12-13 13:04:04,620 INFO [train.py:421] (4/8) Epoch 9, batch 58200, loss[loss=2.165, over 4480.00 frames. , ppl: 8.715113312914745] tot_loss[loss=2.273, over 5518411.06 frames. , ppl: 9.707951287673428], batch size: 70
+2022-12-13 13:05:43,766 INFO [train.py:421] (4/8) Epoch 9, batch 58400, loss[loss=2.766, over 560.00 frames. , ppl: 15.901028581302754] tot_loss[loss=2.273, over 5520400.05 frames. , ppl: 9.705484624704845], batch size: 70
+2022-12-13 13:07:23,478 INFO [train.py:421] (4/8) Epoch 9, batch 58600, loss[loss=2.375, over 1890.00 frames. , ppl: 10.752063140974302] tot_loss[loss=2.273, over 5518774.93 frames. , ppl: 9.707223885057028], batch size: 70
+2022-12-13 13:09:02,105 INFO [train.py:421] (4/8) Epoch 9, batch 58800, loss[loss=2.432, over 1330.00 frames. , ppl: 11.382249873933453] tot_loss[loss=2.274, over 5498442.67 frames. , ppl: 9.714821154653766], batch size: 70
+2022-12-13 13:10:46,532 INFO [train.py:421] (4/8) Epoch 9, batch 59000, loss[loss=3.025, over 560.00 frames. , ppl: 20.589601338853114] tot_loss[loss=2.274, over 5514304.68 frames. , ppl: 9.714533127868757], batch size: 70
+2022-12-13 13:10:46,533 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:10:47,279 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634386305600566
+2022-12-13 13:12:26,044 INFO [train.py:421] (4/8) Epoch 9, batch 59200, loss[loss=2.704, over 770.00 frames. , ppl: 14.940816146819909] tot_loss[loss=2.273, over 5523522.35 frames. , ppl: 9.708644153596184], batch size: 70
+2022-12-13 13:14:06,375 INFO [train.py:421] (4/8) Epoch 9, batch 59400, loss[loss=2.368, over 2170.00 frames. , ppl: 10.670997228261177] tot_loss[loss=2.273, over 5523719.89 frames. , ppl: 9.706896325578667], batch size: 70
+2022-12-13 13:15:48,493 INFO [train.py:421] (4/8) Epoch 9, batch 59600, loss[loss=2.415, over 1750.00 frames. , ppl: 11.187635013631466] tot_loss[loss=2.273, over 5500074.07 frames. , ppl: 9.709766237605775], batch size: 70
+2022-12-13 13:17:26,950 INFO [train.py:421] (4/8) Epoch 9, batch 59800, loss[loss=2.202, over 4970.00 frames. , ppl: 9.044612400625317] tot_loss[loss=2.274, over 5466366.77 frames. , ppl: 9.71775743121163], batch size: 70
+2022-12-13 13:19:05,941 INFO [train.py:421] (4/8) Epoch 9, batch 60000, loss[loss=2.188, over 5600.00 frames. , ppl: 8.918355532612907] tot_loss[loss=2.274, over 5479817.07 frames. , ppl: 9.722750531196338], batch size: 70
+2022-12-13 13:19:05,941 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:19:06,687 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.63870508346662
+2022-12-13 13:20:45,779 INFO [train.py:421] (4/8) Epoch 9, batch 60200, loss[loss=2.525, over 980.00 frames. , ppl: 12.49038805533238] tot_loss[loss=2.275, over 5458062.28 frames. , ppl: 9.726797361791219], batch size: 70
+2022-12-13 13:22:24,271 INFO [train.py:421] (4/8) Epoch 9, batch 60400, loss[loss=2.269, over 2800.00 frames. , ppl: 9.669295251250848] tot_loss[loss=2.274, over 5459028.52 frames. , ppl: 9.722658704979857], batch size: 70
+2022-12-13 13:24:05,150 INFO [train.py:421] (4/8) Epoch 9, batch 60600, loss[loss=2.238, over 3360.00 frames. , ppl: 9.376610330197469] tot_loss[loss=2.275, over 5436120.76 frames. , ppl: 9.729749016370038], batch size: 70
+2022-12-13 13:25:47,347 INFO [train.py:421] (4/8) Epoch 9, batch 60800, loss[loss=2.228, over 4900.00 frames. , ppl: 9.277022619363166] tot_loss[loss=2.275, over 5439695.61 frames. , ppl: 9.725846213999066], batch size: 70
+2022-12-13 13:27:25,582 INFO [train.py:421] (4/8) Epoch 9, batch 61000, loss[loss=2.213, over 5740.00 frames. , ppl: 9.139205877455598] tot_loss[loss=2.274, over 5468829.09 frames. , ppl: 9.71498435633664], batch size: 70
+2022-12-13 13:27:25,583 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:27:26,343 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.267, over 211138.00 frames. , ppl: 9.648386572270214
+2022-12-13 13:29:06,954 INFO [train.py:421] (4/8) Epoch 9, batch 61200, loss[loss=2.561, over 770.00 frames. , ppl: 12.947628064653255] tot_loss[loss=2.275, over 5440781.75 frames. , ppl: 9.727907055238214], batch size: 70
+2022-12-13 13:30:44,585 INFO [train.py:421] (4/8) Epoch 9, batch 61400, loss[loss=2.431, over 1470.00 frames. , ppl: 11.371510577971272] tot_loss[loss=2.275, over 5438884.62 frames. , ppl: 9.727378913550451], batch size: 70
+2022-12-13 13:32:23,424 INFO [train.py:421] (4/8) Epoch 9, batch 61600, loss[loss=2.399, over 1540.00 frames. , ppl: 11.01135016398837] tot_loss[loss=2.274, over 5452371.56 frames. , ppl: 9.720631604243144], batch size: 70
+2022-12-13 13:34:08,411 INFO [train.py:421] (4/8) Epoch 9, batch 61800, loss[loss=2.314, over 2520.00 frames. , ppl: 10.11589782000253] tot_loss[loss=2.274, over 5473596.09 frames. , ppl: 9.719471899725741], batch size: 70
+2022-12-13 13:35:49,228 INFO [train.py:421] (4/8) Epoch 9, batch 62000, loss[loss=2.419, over 1680.00 frames. , ppl: 11.235480618530776] tot_loss[loss=2.275, over 5434625.24 frames. , ppl: 9.726167248608414], batch size: 70
+2022-12-13 13:35:49,228 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:35:49,994 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.263, over 211138.00 frames. , ppl: 9.615221936937656
+2022-12-13 13:37:32,296 INFO [train.py:421] (4/8) Epoch 9, batch 62200, loss[loss=2.345, over 1540.00 frames. , ppl: 10.436040595380629] tot_loss[loss=2.274, over 5483073.49 frames. , ppl: 9.714317939962191], batch size: 70
+2022-12-13 13:39:16,055 INFO [train.py:421] (4/8) Epoch 9, batch 62400, loss[loss=2.209, over 4760.00 frames. , ppl: 9.10982273704973] tot_loss[loss=2.275, over 5448770.43 frames. , ppl: 9.723837858324686], batch size: 70
+2022-12-13 13:40:53,821 INFO [train.py:421] (4/8) Epoch 9, batch 62600, loss[loss=2.439, over 1330.00 frames. , ppl: 11.464518232114209] tot_loss[loss=2.275, over 5437910.76 frames. , ppl: 9.725220382364308], batch size: 70
+2022-12-13 13:42:33,899 INFO [train.py:421] (4/8) Epoch 9, batch 62800, loss[loss=2.362, over 1820.00 frames. , ppl: 10.614582491528981] tot_loss[loss=2.274, over 5466835.16 frames. , ppl: 9.720851522017728], batch size: 70
+2022-12-13 13:44:11,462 INFO [train.py:421] (4/8) Epoch 9, batch 63000, loss[loss=2.812, over 770.00 frames. , ppl: 16.638923659742023] tot_loss[loss=2.274, over 5468073.34 frames. , ppl: 9.720127720242091], batch size: 70
+2022-12-13 13:44:11,462 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:44:12,223 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.266, over 211138.00 frames. , ppl: 9.639786507154147
+2022-12-13 13:45:47,947 INFO [train.py:421] (4/8) Epoch 9, batch 63200, loss[loss=2.396, over 1120.00 frames. , ppl: 10.976592293052528] tot_loss[loss=2.275, over 5439725.91 frames. , ppl: 9.727147703124535], batch size: 70
+2022-12-13 13:47:29,990 INFO [train.py:421] (4/8) Epoch 9, batch 63400, loss[loss=3.071, over 560.00 frames. , ppl: 21.566351406221674] tot_loss[loss=2.276, over 5424960.85 frames. , ppl: 9.739427235187632], batch size: 70
+2022-12-13 13:49:09,473 INFO [train.py:421] (4/8) Epoch 9, batch 63600, loss[loss=2.176, over 3780.00 frames. , ppl: 8.807949434689338] tot_loss[loss=2.276, over 5416781.17 frames. , ppl: 9.733921094672432], batch size: 70
+2022-12-13 13:50:48,800 INFO [train.py:421] (4/8) Epoch 9, batch 63800, loss[loss=2.312, over 2310.00 frames. , ppl: 10.098378482979312] tot_loss[loss=2.276, over 5390909.76 frames. , ppl: 9.740963895376533], batch size: 70
+2022-12-13 13:52:29,000 INFO [train.py:421] (4/8) Epoch 9, batch 64000, loss[loss=2.875, over 630.00 frames. , ppl: 17.721248236220568] tot_loss[loss=2.276, over 5395560.97 frames. , ppl: 9.737539060900103], batch size: 70
+2022-12-13 13:52:29,000 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 13:52:29,747 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.627401629350286
+2022-12-13 13:54:05,564 INFO [train.py:421] (4/8) Epoch 9, batch 64200, loss[loss=2.411, over 1750.00 frames. , ppl: 11.140713777197462] tot_loss[loss=2.274, over 5472477.84 frames. , ppl: 9.720113826863827], batch size: 70
+2022-12-13 13:55:48,758 INFO [train.py:421] (4/8) Epoch 9, batch 64400, loss[loss=2.199, over 3710.00 frames. , ppl: 9.012225191263406] tot_loss[loss=2.273, over 5525335.11 frames. , ppl: 9.708567581198292], batch size: 70
+2022-12-13 13:57:26,866 INFO [train.py:421] (4/8) Epoch 9, batch 64600, loss[loss=2.23, over 5670.00 frames. , ppl: 9.29647516061015] tot_loss[loss=2.272, over 5554511.46 frames. , ppl: 9.698138331623849], batch size: 70
+2022-12-13 13:59:05,983 INFO [train.py:421] (4/8) Epoch 9, batch 64800, loss[loss=2.274, over 2940.00 frames. , ppl: 9.716845310382363] tot_loss[loss=2.272, over 5551613.14 frames. , ppl: 9.694038529542363], batch size: 70
+2022-12-13 14:00:44,182 INFO [train.py:421] (4/8) Epoch 9, batch 65000, loss[loss=2.19, over 4550.00 frames. , ppl: 8.932553581707582] tot_loss[loss=2.272, over 5527579.31 frames. , ppl: 9.699245351263219], batch size: 70
+2022-12-13 14:00:44,182 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:00:44,934 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.633545025291745
+2022-12-13 14:02:29,168 INFO [train.py:421] (4/8) Epoch 9, batch 65200, loss[loss=2.297, over 1680.00 frames. , ppl: 9.941810543555622] tot_loss[loss=2.273, over 5523062.10 frames. , ppl: 9.710453376338902], batch size: 70
+2022-12-13 14:04:10,421 INFO [train.py:421] (4/8) Epoch 9, batch 65400, loss[loss=2.334, over 1540.00 frames. , ppl: 10.323500279611617] tot_loss[loss=2.274, over 5492775.24 frames. , ppl: 9.720004557348997], batch size: 70
+2022-12-13 14:05:48,253 INFO [train.py:421] (4/8) Epoch 9, batch 65600, loss[loss=2.182, over 4970.00 frames. , ppl: 8.867000685537922] tot_loss[loss=2.274, over 5474381.07 frames. , ppl: 9.719003421526109], batch size: 70
+2022-12-13 14:07:28,832 INFO [train.py:421] (4/8) Epoch 9, batch 65800, loss[loss=2.189, over 2730.00 frames. , ppl: 8.928200712953275] tot_loss[loss=2.274, over 5489194.69 frames. , ppl: 9.714004793751426], batch size: 70
+2022-12-13 14:09:11,363 INFO [train.py:421] (4/8) Epoch 9, batch 66000, loss[loss=2.361, over 1330.00 frames. , ppl: 10.603039504992527] tot_loss[loss=2.274, over 5460560.35 frames. , ppl: 9.718127501448198], batch size: 70
+2022-12-13 14:09:11,363 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:09:12,123 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.633080213783197
+2022-12-13 14:10:51,440 INFO [train.py:421] (4/8) Epoch 9, batch 66200, loss[loss=2.446, over 1050.00 frames. , ppl: 11.53850962546883] tot_loss[loss=2.274, over 5453615.17 frames. , ppl: 9.716080743303115], batch size: 70
+2022-12-13 14:12:30,207 INFO [train.py:421] (4/8) Epoch 9, batch 66400, loss[loss=2.198, over 5110.00 frames. , ppl: 9.009500617214277] tot_loss[loss=2.275, over 5429550.21 frames. , ppl: 9.72865620399458], batch size: 70
+2022-12-13 14:14:07,598 INFO [train.py:421] (4/8) Epoch 9, batch 66600, loss[loss=2.196, over 4830.00 frames. , ppl: 8.989141270241248] tot_loss[loss=2.276, over 5420291.04 frames. , ppl: 9.734599051887944], batch size: 70
+2022-12-13 14:15:45,367 INFO [train.py:421] (4/8) Epoch 9, batch 66800, loss[loss=2.224, over 4760.00 frames. , ppl: 9.247870800709372] tot_loss[loss=2.276, over 5401634.84 frames. , ppl: 9.736586604106558], batch size: 70
+2022-12-13 14:17:23,681 INFO [train.py:421] (4/8) Epoch 9, batch 67000, loss[loss=2.171, over 3570.00 frames. , ppl: 8.762727942275514] tot_loss[loss=2.276, over 5395460.84 frames. , ppl: 9.7411079950491], batch size: 70
+2022-12-13 14:17:23,681 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:17:24,429 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.635444426828531
+2022-12-13 14:19:04,981 INFO [train.py:421] (4/8) Epoch 9, batch 67200, loss[loss=2.157, over 11270.00 frames. , ppl: 8.648040791954877] tot_loss[loss=2.275, over 5445294.18 frames. , ppl: 9.727410260216827], batch size: 70
+2022-12-13 14:20:42,270 INFO [train.py:421] (4/8) Epoch 9, batch 67400, loss[loss=2.428, over 840.00 frames. , ppl: 11.331638089689038] tot_loss[loss=2.276, over 5430837.96 frames. , ppl: 9.73568049708399], batch size: 70
+2022-12-13 14:22:19,562 INFO [train.py:421] (4/8) Epoch 9, batch 67600, loss[loss=2.453, over 1050.00 frames. , ppl: 11.625538669653777] tot_loss[loss=2.275, over 5449778.78 frames. , ppl: 9.724065760786557], batch size: 70
+2022-12-13 14:23:58,429 INFO [train.py:421] (4/8) Epoch 9, batch 67800, loss[loss=2.622, over 840.00 frames. , ppl: 13.769083731774456] tot_loss[loss=2.275, over 5400194.17 frames. , ppl: 9.729787077319148], batch size: 70
+2022-12-13 14:25:36,749 INFO [train.py:421] (4/8) Epoch 9, batch 68000, loss[loss=2.19, over 10080.00 frames. , ppl: 8.93924208477048] tot_loss[loss=2.275, over 5414920.42 frames. , ppl: 9.728266694981718], batch size: 70
+2022-12-13 14:25:36,749 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:25:37,495 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621968490265408
+2022-12-13 14:27:17,453 INFO [train.py:421] (4/8) Epoch 9, batch 68200, loss[loss=2.396, over 1330.00 frames. , ppl: 10.984003118956045] tot_loss[loss=2.276, over 5367406.79 frames. , ppl: 9.740788244651801], batch size: 70
+2022-12-13 14:28:57,911 INFO [train.py:421] (4/8) Epoch 9, batch 68400, loss[loss=2.143, over 5180.00 frames. , ppl: 8.526099710520537] tot_loss[loss=2.276, over 5398329.16 frames. , ppl: 9.733560306746815], batch size: 70
+2022-12-13 14:30:42,258 INFO [train.py:421] (4/8) Epoch 9, batch 68600, loss[loss=2.166, over 3850.00 frames. , ppl: 8.720497053326886] tot_loss[loss=2.273, over 5470580.51 frames. , ppl: 9.710150032541062], batch size: 70
+2022-12-13 14:32:21,734 INFO [train.py:421] (4/8) Epoch 9, batch 68800, loss[loss=2.508, over 1050.00 frames. , ppl: 12.281899922949783] tot_loss[loss=2.273, over 5496469.71 frames. , ppl: 9.704940428561097], batch size: 70
+2022-12-13 14:34:01,334 INFO [train.py:421] (4/8) Epoch 9, batch 69000, loss[loss=2.124, over 3360.00 frames. , ppl: 8.367894430911667] tot_loss[loss=2.273, over 5492036.94 frames. , ppl: 9.70509255513955], batch size: 70
+2022-12-13 14:34:01,335 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:34:02,096 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634126784178106
+2022-12-13 14:35:35,880 INFO [train.py:421] (4/8) Epoch 9, batch 69200, loss[loss=2.751, over 840.00 frames. , ppl: 15.657351055232745] tot_loss[loss=2.273, over 5485776.25 frames. , ppl: 9.707516354027417], batch size: 70
+2022-12-13 14:37:12,958 INFO [train.py:421] (4/8) Epoch 9, batch 69400, loss[loss=3.264, over 490.00 frames. , ppl: 26.156816699519066] tot_loss[loss=2.275, over 5426847.80 frames. , ppl: 9.72595791719166], batch size: 70
+2022-12-13 14:38:55,433 INFO [train.py:421] (4/8) Epoch 9, batch 69600, loss[loss=2.907, over 630.00 frames. , ppl: 18.292682492293544] tot_loss[loss=2.274, over 5447530.53 frames. , ppl: 9.720768399032274], batch size: 70
+2022-12-13 14:40:38,305 INFO [train.py:421] (4/8) Epoch 9, batch 69800, loss[loss=2.553, over 980.00 frames. , ppl: 12.841355338034651] tot_loss[loss=2.274, over 5421886.05 frames. , ppl: 9.721973137786861], batch size: 70
+2022-12-13 14:42:20,462 INFO [train.py:421] (4/8) Epoch 9, batch 70000, loss[loss=2.264, over 5180.00 frames. , ppl: 9.624207765641549] tot_loss[loss=2.274, over 5432942.07 frames. , ppl: 9.717127070695755], batch size: 70
+2022-12-13 14:42:20,462 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:42:21,216 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.265, over 211138.00 frames. , ppl: 9.632411551904289
+2022-12-13 14:44:01,660 INFO [train.py:421] (4/8) Epoch 9, batch 70200, loss[loss=2.226, over 3710.00 frames. , ppl: 9.261651015420336] tot_loss[loss=2.273, over 5475357.08 frames. , ppl: 9.704883856865404], batch size: 70
+2022-12-13 14:45:43,982 INFO [train.py:421] (4/8) Epoch 9, batch 70400, loss[loss=2.487, over 840.00 frames. , ppl: 12.025028801493029] tot_loss[loss=2.273, over 5450747.16 frames. , ppl: 9.711895046402873], batch size: 70
+2022-12-13 14:47:24,126 INFO [train.py:421] (4/8) Epoch 9, batch 70600, loss[loss=2.218, over 5810.00 frames. , ppl: 9.189441341426162] tot_loss[loss=2.274, over 5414264.56 frames. , ppl: 9.721822639477086], batch size: 70
+2022-12-13 14:49:06,627 INFO [train.py:421] (4/8) Epoch 9, batch 70800, loss[loss=2.141, over 13510.00 frames. , ppl: 8.503694114578428] tot_loss[loss=2.274, over 5436373.72 frames. , ppl: 9.716968418841324], batch size: 70
+2022-12-13 14:50:45,781 INFO [train.py:421] (4/8) Epoch 9, batch 71000, loss[loss=2.196, over 9870.00 frames. , ppl: 8.988042415656981] tot_loss[loss=2.273, over 5479221.83 frames. , ppl: 9.704176778291487], batch size: 70
+2022-12-13 14:50:45,782 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 14:50:46,533 INFO [train.py:452] (4/8) Epoch 9, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622304589200576
+2022-12-13 14:52:28,161 INFO [train.py:421] (4/8) Epoch 9, batch 71200, loss[loss=2.435, over 1330.00 frames. , ppl: 11.413935399465423] tot_loss[loss=2.271, over 5506030.32 frames. , ppl: 9.69106772870916], batch size: 70
+2022-12-13 14:54:10,854 INFO [train.py:421] (4/8) Epoch 9, batch 71400, loss[loss=2.31, over 2170.00 frames. , ppl: 10.077175675643957] tot_loss[loss=2.271, over 5528776.21 frames. , ppl: 9.690961050692177], batch size: 70
+2022-12-13 14:55:50,275 INFO [train.py:421] (4/8) Epoch 9, batch 71600, loss[loss=2.331, over 1890.00 frames. , ppl: 10.288739422117532] tot_loss[loss=2.272, over 5515755.24 frames. , ppl: 9.69944255555904], batch size: 70
+2022-12-13 14:57:34,045 INFO [train.py:421] (4/8) Epoch 9, batch 71800, loss[loss=2.518, over 1260.00 frames. , ppl: 12.402225568168001] tot_loss[loss=2.273, over 5487552.47 frames. , ppl: 9.706996631787], batch size: 70
+2022-12-13 14:58:47,404 INFO [train.py:421] (4/8) Epoch 10, batch 0, loss[loss=2.535, over 770.00 frames. , ppl: 12.622271747202191] tot_loss[loss=2.535, over 770.00 frames. , ppl: 12.622271747202191], batch size: 70
+2022-12-13 15:00:28,758 INFO [train.py:421] (4/8) Epoch 10, batch 200, loss[loss=2.143, over 5740.00 frames. , ppl: 8.522757351857841] tot_loss[loss=2.267, over 495695.54 frames. , ppl: 9.653849586319657], batch size: 70
+2022-12-13 15:02:15,441 INFO [train.py:421] (4/8) Epoch 10, batch 400, loss[loss=2.25, over 3290.00 frames. , ppl: 9.492395009484426] tot_loss[loss=2.256, over 1028105.27 frames. , ppl: 9.54458360764556], batch size: 70
+2022-12-13 15:03:56,689 INFO [train.py:421] (4/8) Epoch 10, batch 600, loss[loss=2.611, over 980.00 frames. , ppl: 13.615265214923802] tot_loss[loss=2.257, over 1470203.59 frames. , ppl: 9.551825683062745], batch size: 70
+2022-12-13 15:05:40,250 INFO [train.py:421] (4/8) Epoch 10, batch 800, loss[loss=2.303, over 2030.00 frames. , ppl: 10.00424541433146] tot_loss[loss=2.257, over 1880807.81 frames. , ppl: 9.55471541777183], batch size: 70
+2022-12-13 15:07:21,188 INFO [train.py:421] (4/8) Epoch 10, batch 1000, loss[loss=2.367, over 1540.00 frames. , ppl: 10.666040977895785] tot_loss[loss=2.256, over 2270547.81 frames. , ppl: 9.547391564466894], batch size: 70
+2022-12-13 15:07:21,189 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:07:21,948 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.636474138823111
+2022-12-13 15:09:07,555 INFO [train.py:421] (4/8) Epoch 10, batch 1200, loss[loss=2.502, over 910.00 frames. , ppl: 12.202205253073544] tot_loss[loss=2.255, over 2612990.71 frames. , ppl: 9.532676192110953], batch size: 70
+2022-12-13 15:10:54,475 INFO [train.py:421] (4/8) Epoch 10, batch 1400, loss[loss=2.462, over 1190.00 frames. , ppl: 11.725511787628506] tot_loss[loss=2.255, over 2889704.58 frames. , ppl: 9.534612547245876], batch size: 70
+2022-12-13 15:12:36,531 INFO [train.py:421] (4/8) Epoch 10, batch 1600, loss[loss=2.937, over 560.00 frames. , ppl: 18.851825612040862] tot_loss[loss=2.259, over 3114330.21 frames. , ppl: 9.570531842305835], batch size: 70
+2022-12-13 15:14:19,426 INFO [train.py:421] (4/8) Epoch 10, batch 1800, loss[loss=2.197, over 8400.00 frames. , ppl: 8.995390141227205] tot_loss[loss=2.258, over 3351916.87 frames. , ppl: 9.564338814117226], batch size: 70
+2022-12-13 15:16:01,289 INFO [train.py:421] (4/8) Epoch 10, batch 2000, loss[loss=2.375, over 1050.00 frames. , ppl: 10.749885848667526] tot_loss[loss=2.259, over 3566286.12 frames. , ppl: 9.574884710819116], batch size: 70
+2022-12-13 15:16:01,290 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:16:02,057 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634249414308712
+2022-12-13 15:17:44,593 INFO [train.py:421] (4/8) Epoch 10, batch 2200, loss[loss=2.255, over 2660.00 frames. , ppl: 9.53891300040888] tot_loss[loss=2.26, over 3727462.45 frames. , ppl: 9.584989119392823], batch size: 70
+2022-12-13 15:19:27,875 INFO [train.py:421] (4/8) Epoch 10, batch 2400, loss[loss=2.214, over 5320.00 frames. , ppl: 9.153771959112902] tot_loss[loss=2.259, over 3923216.92 frames. , ppl: 9.574383237867949], batch size: 70
+2022-12-13 15:21:09,658 INFO [train.py:421] (4/8) Epoch 10, batch 2600, loss[loss=2.394, over 1960.00 frames. , ppl: 10.958252149232804] tot_loss[loss=2.261, over 4056469.05 frames. , ppl: 9.595578791518921], batch size: 70
+2022-12-13 15:22:53,451 INFO [train.py:421] (4/8) Epoch 10, batch 2800, loss[loss=2.438, over 1750.00 frames. , ppl: 11.451401968615185] tot_loss[loss=2.261, over 4203038.86 frames. , ppl: 9.595849738105885], batch size: 70
+2022-12-13 15:24:36,433 INFO [train.py:421] (4/8) Epoch 10, batch 3000, loss[loss=2.439, over 1120.00 frames. , ppl: 11.459986347288213] tot_loss[loss=2.26, over 4372008.22 frames. , ppl: 9.581832019224816], batch size: 70
+2022-12-13 15:24:36,433 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:24:37,197 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.624782969170553
+2022-12-13 15:26:21,716 INFO [train.py:421] (4/8) Epoch 10, batch 3200, loss[loss=2.27, over 2940.00 frames. , ppl: 9.681667116789834] tot_loss[loss=2.26, over 4514936.03 frames. , ppl: 9.581742897887676], batch size: 70
+2022-12-13 15:28:06,727 INFO [train.py:421] (4/8) Epoch 10, batch 3400, loss[loss=2.272, over 3150.00 frames. , ppl: 9.69599748210328] tot_loss[loss=2.259, over 4618825.14 frames. , ppl: 9.575840368628487], batch size: 70
+2022-12-13 15:29:50,698 INFO [train.py:421] (4/8) Epoch 10, batch 3600, loss[loss=2.508, over 1260.00 frames. , ppl: 12.283001330080653] tot_loss[loss=2.26, over 4706742.89 frames. , ppl: 9.579651711437718], batch size: 70
+2022-12-13 15:31:33,968 INFO [train.py:421] (4/8) Epoch 10, batch 3800, loss[loss=2.659, over 630.00 frames. , ppl: 14.28130637245207] tot_loss[loss=2.258, over 4850352.04 frames. , ppl: 9.562579447427694], batch size: 70
+2022-12-13 15:33:14,208 INFO [train.py:421] (4/8) Epoch 10, batch 4000, loss[loss=2.197, over 6160.00 frames. , ppl: 8.994755089344203] tot_loss[loss=2.26, over 4879044.24 frames. , ppl: 9.579062135319777], batch size: 70
+2022-12-13 15:33:14,209 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:33:14,973 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.6403101424268
+2022-12-13 15:34:56,293 INFO [train.py:421] (4/8) Epoch 10, batch 4200, loss[loss=2.653, over 840.00 frames. , ppl: 14.200571896366693] tot_loss[loss=2.259, over 4960405.28 frames. , ppl: 9.577961743123174], batch size: 70
+2022-12-13 15:36:36,698 INFO [train.py:421] (4/8) Epoch 10, batch 4400, loss[loss=2.29, over 3780.00 frames. , ppl: 9.879235328336724] tot_loss[loss=2.26, over 5017712.28 frames. , ppl: 9.587133580942735], batch size: 70
+2022-12-13 15:38:18,566 INFO [train.py:421] (4/8) Epoch 10, batch 4600, loss[loss=2.161, over 4270.00 frames. , ppl: 8.678216703170078] tot_loss[loss=2.261, over 5047180.35 frames. , ppl: 9.596553601888692], batch size: 70
+2022-12-13 15:39:59,357 INFO [train.py:421] (4/8) Epoch 10, batch 4800, loss[loss=2.243, over 2520.00 frames. , ppl: 9.424538317671395] tot_loss[loss=2.261, over 5126347.65 frames. , ppl: 9.595931054619912], batch size: 70
+2022-12-13 15:41:41,307 INFO [train.py:421] (4/8) Epoch 10, batch 5000, loss[loss=2.319, over 1610.00 frames. , ppl: 10.165952858805877] tot_loss[loss=2.263, over 5124055.06 frames. , ppl: 9.61460014653982], batch size: 70
+2022-12-13 15:41:41,308 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:41:42,117 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634593072176935
+2022-12-13 15:43:26,471 INFO [train.py:421] (4/8) Epoch 10, batch 5200, loss[loss=2.249, over 2800.00 frames. , ppl: 9.481708644903133] tot_loss[loss=2.263, over 5184340.88 frames. , ppl: 9.615962841866859], batch size: 70
+2022-12-13 15:45:06,994 INFO [train.py:421] (4/8) Epoch 10, batch 5400, loss[loss=2.342, over 2730.00 frames. , ppl: 10.405098097350365] tot_loss[loss=2.265, over 5208738.08 frames. , ppl: 9.629499731477969], batch size: 70
+2022-12-13 15:46:49,887 INFO [train.py:421] (4/8) Epoch 10, batch 5600, loss[loss=2.454, over 910.00 frames. , ppl: 11.640139846777288] tot_loss[loss=2.263, over 5313898.64 frames. , ppl: 9.615375389884472], batch size: 70
+2022-12-13 15:48:31,065 INFO [train.py:421] (4/8) Epoch 10, batch 5800, loss[loss=2.236, over 3500.00 frames. , ppl: 9.354826736620641] tot_loss[loss=2.263, over 5349078.42 frames. , ppl: 9.614587089122482], batch size: 70
+2022-12-13 15:50:14,249 INFO [train.py:421] (4/8) Epoch 10, batch 6000, loss[loss=2.48, over 1680.00 frames. , ppl: 11.944300930308124] tot_loss[loss=2.263, over 5356285.38 frames. , ppl: 9.611088517235599], batch size: 70
+2022-12-13 15:50:14,249 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:50:14,999 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622741820215266
+2022-12-13 15:51:54,467 INFO [train.py:421] (4/8) Epoch 10, batch 6200, loss[loss=2.258, over 1890.00 frames. , ppl: 9.563124920662501] tot_loss[loss=2.263, over 5357293.66 frames. , ppl: 9.611434707142545], batch size: 70
+2022-12-13 15:53:34,733 INFO [train.py:421] (4/8) Epoch 10, batch 6400, loss[loss=2.317, over 1680.00 frames. , ppl: 10.148400529741286] tot_loss[loss=2.264, over 5328792.82 frames. , ppl: 9.626173944838555], batch size: 70
+2022-12-13 15:55:17,562 INFO [train.py:421] (4/8) Epoch 10, batch 6600, loss[loss=2.296, over 2100.00 frames. , ppl: 9.93807162533116] tot_loss[loss=2.265, over 5359466.32 frames. , ppl: 9.626602827609338], batch size: 70
+2022-12-13 15:56:56,921 INFO [train.py:421] (4/8) Epoch 10, batch 6800, loss[loss=2.624, over 700.00 frames. , ppl: 13.78661999268225] tot_loss[loss=2.266, over 5333171.84 frames. , ppl: 9.645092445311802], batch size: 70
+2022-12-13 15:58:42,426 INFO [train.py:421] (4/8) Epoch 10, batch 7000, loss[loss=2.19, over 6510.00 frames. , ppl: 8.933680401611976] tot_loss[loss=2.265, over 5399131.38 frames. , ppl: 9.632898841768315], batch size: 70
+2022-12-13 15:58:42,427 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 15:58:43,176 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637175889263702
+2022-12-13 16:00:25,416 INFO [train.py:421] (4/8) Epoch 10, batch 7200, loss[loss=2.496, over 1120.00 frames. , ppl: 12.132179441062297] tot_loss[loss=2.266, over 5399254.56 frames. , ppl: 9.638004438256933], batch size: 70
+2022-12-13 16:02:05,000 INFO [train.py:421] (4/8) Epoch 10, batch 7400, loss[loss=2.256, over 3920.00 frames. , ppl: 9.546454054664665] tot_loss[loss=2.266, over 5424130.06 frames. , ppl: 9.637936353677091], batch size: 70
+2022-12-13 16:03:46,164 INFO [train.py:421] (4/8) Epoch 10, batch 7600, loss[loss=2.173, over 11340.00 frames. , ppl: 8.78203593865661] tot_loss[loss=2.266, over 5433238.13 frames. , ppl: 9.63871589863238], batch size: 70
+2022-12-13 16:05:27,166 INFO [train.py:421] (4/8) Epoch 10, batch 7800, loss[loss=2.487, over 840.00 frames. , ppl: 12.029261988610262] tot_loss[loss=2.266, over 5422224.49 frames. , ppl: 9.640576642898449], batch size: 70
+2022-12-13 16:07:06,113 INFO [train.py:421] (4/8) Epoch 10, batch 8000, loss[loss=2.127, over 5040.00 frames. , ppl: 8.390086088309392] tot_loss[loss=2.267, over 5413251.03 frames. , ppl: 9.647861853752309], batch size: 70
+2022-12-13 16:07:06,114 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:07:06,898 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.623687560464123
+2022-12-13 16:08:45,826 INFO [train.py:421] (4/8) Epoch 10, batch 8200, loss[loss=2.312, over 2730.00 frames. , ppl: 10.090006080431301] tot_loss[loss=2.268, over 5410369.63 frames. , ppl: 9.657501777684857], batch size: 70
+2022-12-13 16:10:28,724 INFO [train.py:421] (4/8) Epoch 10, batch 8400, loss[loss=2.456, over 1400.00 frames. , ppl: 11.654621262243026] tot_loss[loss=2.266, over 5461777.91 frames. , ppl: 9.644651990842819], batch size: 70
+2022-12-13 16:12:10,332 INFO [train.py:421] (4/8) Epoch 10, batch 8600, loss[loss=2.132, over 4270.00 frames. , ppl: 8.434547856613042] tot_loss[loss=2.267, over 5442393.87 frames. , ppl: 9.649224707586056], batch size: 70
+2022-12-13 16:13:50,371 INFO [train.py:421] (4/8) Epoch 10, batch 8800, loss[loss=2.314, over 1890.00 frames. , ppl: 10.111124190222204] tot_loss[loss=2.267, over 5446157.67 frames. , ppl: 9.648654657319764], batch size: 70
+2022-12-13 16:15:29,983 INFO [train.py:421] (4/8) Epoch 10, batch 9000, loss[loss=2.363, over 1890.00 frames. , ppl: 10.620471474650742] tot_loss[loss=2.267, over 5456434.57 frames. , ppl: 9.6503791295152], batch size: 70
+2022-12-13 16:15:29,983 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:15:30,747 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634809825352011
+2022-12-13 16:17:11,734 INFO [train.py:421] (4/8) Epoch 10, batch 9200, loss[loss=2.364, over 1610.00 frames. , ppl: 10.630567659101944] tot_loss[loss=2.268, over 5453874.98 frames. , ppl: 9.660498531809779], batch size: 70
+2022-12-13 16:18:52,513 INFO [train.py:421] (4/8) Epoch 10, batch 9400, loss[loss=2.409, over 1540.00 frames. , ppl: 11.122667772680584] tot_loss[loss=2.268, over 5473078.09 frames. , ppl: 9.657543587003312], batch size: 70
+2022-12-13 16:20:35,315 INFO [train.py:421] (4/8) Epoch 10, batch 9600, loss[loss=2.394, over 1120.00 frames. , ppl: 10.959870448711927] tot_loss[loss=2.267, over 5488886.01 frames. , ppl: 9.653263842294683], batch size: 70
+2022-12-13 16:22:23,362 INFO [train.py:421] (4/8) Epoch 10, batch 9800, loss[loss=2.413, over 1750.00 frames. , ppl: 11.17007136704503] tot_loss[loss=2.268, over 5462259.98 frames. , ppl: 9.658262076736376], batch size: 70
+2022-12-13 16:24:04,262 INFO [train.py:421] (4/8) Epoch 10, batch 10000, loss[loss=2.445, over 1260.00 frames. , ppl: 11.525972211698338] tot_loss[loss=2.268, over 5441109.40 frames. , ppl: 9.664879092677696], batch size: 70
+2022-12-13 16:24:04,263 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:24:05,013 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621310567860883
+2022-12-13 16:25:47,082 INFO [train.py:421] (4/8) Epoch 10, batch 10200, loss[loss=2.339, over 1610.00 frames. , ppl: 10.37247441958905] tot_loss[loss=2.269, over 5430659.88 frames. , ppl: 9.667790519751735], batch size: 70
+2022-12-13 16:27:29,193 INFO [train.py:421] (4/8) Epoch 10, batch 10400, loss[loss=2.747, over 700.00 frames. , ppl: 15.602730645101142] tot_loss[loss=2.269, over 5432462.26 frames. , ppl: 9.670101536969756], batch size: 70
+2022-12-13 16:29:09,732 INFO [train.py:421] (4/8) Epoch 10, batch 10600, loss[loss=2.498, over 1540.00 frames. , ppl: 12.16124475474305] tot_loss[loss=2.27, over 5411535.83 frames. , ppl: 9.67570810071469], batch size: 70
+2022-12-13 16:30:51,417 INFO [train.py:421] (4/8) Epoch 10, batch 10800, loss[loss=2.268, over 2100.00 frames. , ppl: 9.6594136025135] tot_loss[loss=2.268, over 5455054.06 frames. , ppl: 9.660264924917357], batch size: 70
+2022-12-13 16:32:33,750 INFO [train.py:421] (4/8) Epoch 10, batch 11000, loss[loss=2.202, over 3500.00 frames. , ppl: 9.047489398432877] tot_loss[loss=2.267, over 5481482.29 frames. , ppl: 9.648441073855457], batch size: 70
+2022-12-13 16:32:33,751 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:32:34,520 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60629313197587
+2022-12-13 16:34:15,547 INFO [train.py:421] (4/8) Epoch 10, batch 11200, loss[loss=2.303, over 1680.00 frames. , ppl: 10.008835471970219] tot_loss[loss=2.267, over 5477931.51 frames. , ppl: 9.64918408682106], batch size: 70
+2022-12-13 16:35:59,434 INFO [train.py:421] (4/8) Epoch 10, batch 11400, loss[loss=2.459, over 1610.00 frames. , ppl: 11.698459763069321] tot_loss[loss=2.267, over 5472648.01 frames. , ppl: 9.6526129044026], batch size: 70
+2022-12-13 16:37:41,285 INFO [train.py:421] (4/8) Epoch 10, batch 11600, loss[loss=2.222, over 3010.00 frames. , ppl: 9.228951426066821] tot_loss[loss=2.268, over 5447965.30 frames. , ppl: 9.662394189748579], batch size: 70
+2022-12-13 16:39:23,732 INFO [train.py:421] (4/8) Epoch 10, batch 11800, loss[loss=2.308, over 1960.00 frames. , ppl: 10.049496369150186] tot_loss[loss=2.269, over 5411137.48 frames. , ppl: 9.670891333915096], batch size: 70
+2022-12-13 16:41:07,319 INFO [train.py:421] (4/8) Epoch 10, batch 12000, loss[loss=2.323, over 2170.00 frames. , ppl: 10.205747264025664] tot_loss[loss=2.269, over 5441825.23 frames. , ppl: 9.668339727936546], batch size: 70
+2022-12-13 16:41:07,320 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:41:08,102 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.625352801927015
+2022-12-13 16:42:53,922 INFO [train.py:421] (4/8) Epoch 10, batch 12200, loss[loss=2.439, over 1050.00 frames. , ppl: 11.462956114989975] tot_loss[loss=2.268, over 5470392.71 frames. , ppl: 9.657246066367136], batch size: 70
+2022-12-13 16:44:34,115 INFO [train.py:421] (4/8) Epoch 10, batch 12400, loss[loss=2.457, over 2240.00 frames. , ppl: 11.67149073164482] tot_loss[loss=2.267, over 5464161.08 frames. , ppl: 9.652166749942939], batch size: 70
+2022-12-13 16:46:16,870 INFO [train.py:421] (4/8) Epoch 10, batch 12600, loss[loss=2.569, over 770.00 frames. , ppl: 13.047442026468035] tot_loss[loss=2.266, over 5515475.94 frames. , ppl: 9.638587733945826], batch size: 70
+2022-12-13 16:47:56,491 INFO [train.py:421] (4/8) Epoch 10, batch 12800, loss[loss=2.173, over 5530.00 frames. , ppl: 8.78133413953294] tot_loss[loss=2.266, over 5506014.33 frames. , ppl: 9.644884295529966], batch size: 70
+2022-12-13 16:49:37,745 INFO [train.py:421] (4/8) Epoch 10, batch 13000, loss[loss=2.181, over 4200.00 frames. , ppl: 8.850787316777952] tot_loss[loss=2.267, over 5497422.18 frames. , ppl: 9.648082505566636], batch size: 70
+2022-12-13 16:49:37,745 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:49:38,513 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634478993517094
+2022-12-13 16:51:20,368 INFO [train.py:421] (4/8) Epoch 10, batch 13200, loss[loss=2.429, over 1260.00 frames. , ppl: 11.34459211150246] tot_loss[loss=2.267, over 5496274.81 frames. , ppl: 9.648637241715495], batch size: 70
+2022-12-13 16:53:01,609 INFO [train.py:421] (4/8) Epoch 10, batch 13400, loss[loss=3.016, over 560.00 frames. , ppl: 20.40530280563611] tot_loss[loss=2.266, over 5502759.73 frames. , ppl: 9.64365474200285], batch size: 70
+2022-12-13 16:54:46,215 INFO [train.py:421] (4/8) Epoch 10, batch 13600, loss[loss=2.265, over 2380.00 frames. , ppl: 9.628063196956969] tot_loss[loss=2.267, over 5489260.26 frames. , ppl: 9.648535365871508], batch size: 70
+2022-12-13 16:56:30,070 INFO [train.py:421] (4/8) Epoch 10, batch 13800, loss[loss=2.287, over 3010.00 frames. , ppl: 9.842617383521926] tot_loss[loss=2.267, over 5480655.03 frames. , ppl: 9.651442515779204], batch size: 70
+2022-12-13 16:58:09,139 INFO [train.py:421] (4/8) Epoch 10, batch 14000, loss[loss=2.671, over 630.00 frames. , ppl: 14.449745086258991] tot_loss[loss=2.267, over 5478163.71 frames. , ppl: 9.65124336378167], batch size: 70
+2022-12-13 16:58:09,140 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 16:58:09,934 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.641104923874973
+2022-12-13 16:59:47,674 INFO [train.py:421] (4/8) Epoch 10, batch 14200, loss[loss=2.256, over 3010.00 frames. , ppl: 9.548545698815017] tot_loss[loss=2.267, over 5481938.66 frames. , ppl: 9.654426378027921], batch size: 70
+2022-12-13 17:01:29,027 INFO [train.py:421] (4/8) Epoch 10, batch 14400, loss[loss=2.37, over 1470.00 frames. , ppl: 10.69218371897436] tot_loss[loss=2.268, over 5441737.75 frames. , ppl: 9.663916994410021], batch size: 70
+2022-12-13 17:03:09,990 INFO [train.py:421] (4/8) Epoch 10, batch 14600, loss[loss=2.373, over 2100.00 frames. , ppl: 10.732620626466069] tot_loss[loss=2.268, over 5478792.97 frames. , ppl: 9.655939408741911], batch size: 70
+2022-12-13 17:04:50,645 INFO [train.py:421] (4/8) Epoch 10, batch 14800, loss[loss=2.234, over 5460.00 frames. , ppl: 9.340301320813653] tot_loss[loss=2.268, over 5478854.80 frames. , ppl: 9.66103638328173], batch size: 70
+2022-12-13 17:06:32,745 INFO [train.py:421] (4/8) Epoch 10, batch 15000, loss[loss=2.244, over 4410.00 frames. , ppl: 9.43337984533957] tot_loss[loss=2.268, over 5474148.27 frames. , ppl: 9.661026483535734], batch size: 70
+2022-12-13 17:06:32,746 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:06:33,530 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.634267951557288
+2022-12-13 17:08:12,507 INFO [train.py:421] (4/8) Epoch 10, batch 15200, loss[loss=2.546, over 1120.00 frames. , ppl: 12.754876847182935] tot_loss[loss=2.269, over 5457757.79 frames. , ppl: 9.668977977834775], batch size: 70
+2022-12-13 17:09:58,765 INFO [train.py:421] (4/8) Epoch 10, batch 15400, loss[loss=2.279, over 3360.00 frames. , ppl: 9.766259826509726] tot_loss[loss=2.269, over 5457079.21 frames. , ppl: 9.668529934794746], batch size: 70
+2022-12-13 17:11:43,410 INFO [train.py:421] (4/8) Epoch 10, batch 15600, loss[loss=2.164, over 7490.00 frames. , ppl: 8.70948311948673] tot_loss[loss=2.268, over 5456649.79 frames. , ppl: 9.659225720928076], batch size: 70
+2022-12-13 17:13:24,563 INFO [train.py:421] (4/8) Epoch 10, batch 15800, loss[loss=2.167, over 7210.00 frames. , ppl: 8.73117850345691] tot_loss[loss=2.269, over 5436820.12 frames. , ppl: 9.669890209094087], batch size: 70
+2022-12-13 17:15:07,632 INFO [train.py:421] (4/8) Epoch 10, batch 16000, loss[loss=2.275, over 1890.00 frames. , ppl: 9.723629914733028] tot_loss[loss=2.269, over 5483390.36 frames. , ppl: 9.665157488270106], batch size: 70
+2022-12-13 17:15:07,632 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:15:08,400 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61156377960389
+2022-12-13 17:16:51,467 INFO [train.py:421] (4/8) Epoch 10, batch 16200, loss[loss=2.529, over 910.00 frames. , ppl: 12.53583759714428] tot_loss[loss=2.27, over 5445728.93 frames. , ppl: 9.679125975860115], batch size: 70
+2022-12-13 17:18:33,699 INFO [train.py:421] (4/8) Epoch 10, batch 16400, loss[loss=2.244, over 4900.00 frames. , ppl: 9.426679609699955] tot_loss[loss=2.27, over 5434577.74 frames. , ppl: 9.683340513778388], batch size: 70
+2022-12-13 17:20:17,176 INFO [train.py:421] (4/8) Epoch 10, batch 16600, loss[loss=2.266, over 3850.00 frames. , ppl: 9.644171506499871] tot_loss[loss=2.271, over 5408113.95 frames. , ppl: 9.690414876288685], batch size: 70
+2022-12-13 17:21:59,457 INFO [train.py:421] (4/8) Epoch 10, batch 16800, loss[loss=2.164, over 5180.00 frames. , ppl: 8.701657402173366] tot_loss[loss=2.271, over 5427087.01 frames. , ppl: 9.6869399596894], batch size: 70
+2022-12-13 17:23:43,870 INFO [train.py:421] (4/8) Epoch 10, batch 17000, loss[loss=2.239, over 3710.00 frames. , ppl: 9.384311787661725] tot_loss[loss=2.271, over 5429208.25 frames. , ppl: 9.68510428000068], batch size: 70
+2022-12-13 17:23:43,871 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:23:44,636 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.629349702040027
+2022-12-13 17:25:25,474 INFO [train.py:421] (4/8) Epoch 10, batch 17200, loss[loss=2.685, over 840.00 frames. , ppl: 14.652777302209172] tot_loss[loss=2.271, over 5441694.42 frames. , ppl: 9.68589696789215], batch size: 70
+2022-12-13 17:27:07,628 INFO [train.py:421] (4/8) Epoch 10, batch 17400, loss[loss=2.169, over 3850.00 frames. , ppl: 8.752092516820392] tot_loss[loss=2.269, over 5482799.82 frames. , ppl: 9.671525487716908], batch size: 70
+2022-12-13 17:28:48,475 INFO [train.py:421] (4/8) Epoch 10, batch 17600, loss[loss=2.742, over 840.00 frames. , ppl: 15.520484946018028] tot_loss[loss=2.268, over 5503466.21 frames. , ppl: 9.660098426345641], batch size: 70
+2022-12-13 17:30:34,265 INFO [train.py:421] (4/8) Epoch 10, batch 17800, loss[loss=2.175, over 10990.00 frames. , ppl: 8.80361971142682] tot_loss[loss=2.267, over 5541354.22 frames. , ppl: 9.651321085717798], batch size: 70
+2022-12-13 17:32:15,320 INFO [train.py:421] (4/8) Epoch 10, batch 18000, loss[loss=2.606, over 1050.00 frames. , ppl: 13.545189095951722] tot_loss[loss=2.268, over 5515048.57 frames. , ppl: 9.66203179170143], batch size: 70
+2022-12-13 17:32:15,321 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:32:16,072 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622539580598964
+2022-12-13 17:33:56,551 INFO [train.py:421] (4/8) Epoch 10, batch 18200, loss[loss=2.512, over 700.00 frames. , ppl: 12.33002940846225] tot_loss[loss=2.269, over 5501334.35 frames. , ppl: 9.6693488470408], batch size: 70
+2022-12-13 17:35:35,469 INFO [train.py:421] (4/8) Epoch 10, batch 18400, loss[loss=2.35, over 1120.00 frames. , ppl: 10.487080663304683] tot_loss[loss=2.27, over 5462127.10 frames. , ppl: 9.675394537449172], batch size: 70
+2022-12-13 17:37:19,321 INFO [train.py:421] (4/8) Epoch 10, batch 18600, loss[loss=2.369, over 1890.00 frames. , ppl: 10.688225252023305] tot_loss[loss=2.269, over 5490250.17 frames. , ppl: 9.67255314211143], batch size: 70
+2022-12-13 17:39:02,831 INFO [train.py:421] (4/8) Epoch 10, batch 18800, loss[loss=2.215, over 5110.00 frames. , ppl: 9.162528654114388] tot_loss[loss=2.269, over 5490375.34 frames. , ppl: 9.671739935079323], batch size: 70
+2022-12-13 17:40:48,051 INFO [train.py:421] (4/8) Epoch 10, batch 19000, loss[loss=2.109, over 5670.00 frames. , ppl: 8.240836354252744] tot_loss[loss=2.268, over 5532377.01 frames. , ppl: 9.660234547163704], batch size: 70
+2022-12-13 17:40:48,052 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:40:48,813 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.63137799747174
+2022-12-13 17:42:30,791 INFO [train.py:421] (4/8) Epoch 10, batch 19200, loss[loss=2.321, over 1540.00 frames. , ppl: 10.183874492964858] tot_loss[loss=2.267, over 5546162.00 frames. , ppl: 9.650877021350446], batch size: 70
+2022-12-13 17:44:15,187 INFO [train.py:421] (4/8) Epoch 10, batch 19400, loss[loss=3.04, over 560.00 frames. , ppl: 20.89877878959978] tot_loss[loss=2.266, over 5564456.79 frames. , ppl: 9.645573157851503], batch size: 70
+2022-12-13 17:45:57,130 INFO [train.py:421] (4/8) Epoch 10, batch 19600, loss[loss=2.317, over 3220.00 frames. , ppl: 10.149625379568342] tot_loss[loss=2.265, over 5632551.05 frames. , ppl: 9.632600464627295], batch size: 70
+2022-12-13 17:47:36,296 INFO [train.py:421] (4/8) Epoch 10, batch 19800, loss[loss=2.269, over 2940.00 frames. , ppl: 9.673656403922529] tot_loss[loss=2.267, over 5604285.40 frames. , ppl: 9.645769152998332], batch size: 70
+2022-12-13 17:49:20,607 INFO [train.py:421] (4/8) Epoch 10, batch 20000, loss[loss=2.334, over 2030.00 frames. , ppl: 10.323651639313756] tot_loss[loss=2.265, over 5673494.37 frames. , ppl: 9.630791043469715], batch size: 70
+2022-12-13 17:49:20,608 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:49:21,384 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.641742794080622
+2022-12-13 17:51:06,026 INFO [train.py:421] (4/8) Epoch 10, batch 20200, loss[loss=2.147, over 5530.00 frames. , ppl: 8.562826950082352] tot_loss[loss=2.266, over 5658341.46 frames. , ppl: 9.636850449896304], batch size: 70
+2022-12-13 17:52:43,974 INFO [train.py:421] (4/8) Epoch 10, batch 20400, loss[loss=2.326, over 1960.00 frames. , ppl: 10.231802249354555] tot_loss[loss=2.266, over 5624089.55 frames. , ppl: 9.644993249435476], batch size: 70
+2022-12-13 17:54:27,189 INFO [train.py:421] (4/8) Epoch 10, batch 20600, loss[loss=2.455, over 1960.00 frames. , ppl: 11.648974287855891] tot_loss[loss=2.266, over 5603427.01 frames. , ppl: 9.644847908609265], batch size: 70
+2022-12-13 17:56:10,951 INFO [train.py:421] (4/8) Epoch 10, batch 20800, loss[loss=2.306, over 2240.00 frames. , ppl: 10.030697205852968] tot_loss[loss=2.267, over 5616472.54 frames. , ppl: 9.64717158949042], batch size: 70
+2022-12-13 17:57:56,884 INFO [train.py:421] (4/8) Epoch 10, batch 21000, loss[loss=2.257, over 4200.00 frames. , ppl: 9.554901500697447] tot_loss[loss=2.267, over 5614539.89 frames. , ppl: 9.648509450634936], batch size: 70
+2022-12-13 17:57:56,885 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 17:57:57,648 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.622576610069983
+2022-12-13 17:59:39,845 INFO [train.py:421] (4/8) Epoch 10, batch 21200, loss[loss=2.851, over 560.00 frames. , ppl: 17.306223096640714] tot_loss[loss=2.268, over 5564717.84 frames. , ppl: 9.663305964956955], batch size: 70
+2022-12-13 18:01:19,792 INFO [train.py:421] (4/8) Epoch 10, batch 21400, loss[loss=2.532, over 980.00 frames. , ppl: 12.581491577632201] tot_loss[loss=2.268, over 5568334.47 frames. , ppl: 9.66103847413159], batch size: 70
+2022-12-13 18:03:00,309 INFO [train.py:421] (4/8) Epoch 10, batch 21600, loss[loss=2.998, over 560.00 frames. , ppl: 20.046520435101332] tot_loss[loss=2.268, over 5547146.76 frames. , ppl: 9.663461782400876], batch size: 70
+2022-12-13 18:04:42,566 INFO [train.py:421] (4/8) Epoch 10, batch 21800, loss[loss=2.128, over 6440.00 frames. , ppl: 8.399847558118976] tot_loss[loss=2.268, over 5546238.39 frames. , ppl: 9.661789601971334], batch size: 70
+2022-12-13 18:06:24,589 INFO [train.py:421] (4/8) Epoch 10, batch 22000, loss[loss=2.485, over 1540.00 frames. , ppl: 12.000883577198554] tot_loss[loss=2.268, over 5564353.52 frames. , ppl: 9.663451624425896], batch size: 70
+2022-12-13 18:06:24,589 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:06:25,356 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.620783692748795
+2022-12-13 18:08:04,733 INFO [train.py:421] (4/8) Epoch 10, batch 22200, loss[loss=2.322, over 1540.00 frames. , ppl: 10.195207978585923] tot_loss[loss=2.267, over 5591712.13 frames. , ppl: 9.654909842042903], batch size: 70
+2022-12-13 18:09:43,529 INFO [train.py:421] (4/8) Epoch 10, batch 22400, loss[loss=2.304, over 2310.00 frames. , ppl: 10.01388704170633] tot_loss[loss=2.268, over 5530971.40 frames. , ppl: 9.659123855318594], batch size: 70
+2022-12-13 18:11:22,901 INFO [train.py:421] (4/8) Epoch 10, batch 22600, loss[loss=2.285, over 2870.00 frames. , ppl: 9.823712846364613] tot_loss[loss=2.268, over 5522132.61 frames. , ppl: 9.661672652137282], batch size: 70
+2022-12-13 18:13:02,880 INFO [train.py:421] (4/8) Epoch 10, batch 22800, loss[loss=2.4, over 1120.00 frames. , ppl: 11.025964052713483] tot_loss[loss=2.268, over 5518748.48 frames. , ppl: 9.664247375119361], batch size: 70
+2022-12-13 18:14:48,430 INFO [train.py:421] (4/8) Epoch 10, batch 23000, loss[loss=2.272, over 3290.00 frames. , ppl: 9.699540131448412] tot_loss[loss=2.267, over 5582980.75 frames. , ppl: 9.649143990039043], batch size: 70
+2022-12-13 18:14:48,430 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:14:49,200 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.624770148321595
+2022-12-13 18:16:34,755 INFO [train.py:421] (4/8) Epoch 10, batch 23200, loss[loss=2.166, over 6720.00 frames. , ppl: 8.720825814565082] tot_loss[loss=2.266, over 5597485.69 frames. , ppl: 9.641505111340763], batch size: 70
+2022-12-13 18:18:16,613 INFO [train.py:421] (4/8) Epoch 10, batch 23400, loss[loss=2.325, over 1890.00 frames. , ppl: 10.229877479874386] tot_loss[loss=2.266, over 5585979.33 frames. , ppl: 9.640975642190364], batch size: 70
+2022-12-13 18:19:54,568 INFO [train.py:421] (4/8) Epoch 10, batch 23600, loss[loss=2.36, over 2030.00 frames. , ppl: 10.59055303570359] tot_loss[loss=2.267, over 5554759.15 frames. , ppl: 9.648345279130622], batch size: 70
+2022-12-13 18:21:36,228 INFO [train.py:421] (4/8) Epoch 10, batch 23800, loss[loss=2.194, over 4620.00 frames. , ppl: 8.973097041126831] tot_loss[loss=2.267, over 5534678.33 frames. , ppl: 9.651040309834325], batch size: 70
+2022-12-13 18:23:16,517 INFO [train.py:421] (4/8) Epoch 10, batch 24000, loss[loss=2.246, over 5180.00 frames. , ppl: 9.446801001562276] tot_loss[loss=2.266, over 5550623.21 frames. , ppl: 9.635947196257776], batch size: 70
+2022-12-13 18:23:16,518 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:23:17,301 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.631834173341721
+2022-12-13 18:24:57,919 INFO [train.py:421] (4/8) Epoch 10, batch 24200, loss[loss=2.605, over 910.00 frames. , ppl: 13.527106329198173] tot_loss[loss=2.266, over 5556740.14 frames. , ppl: 9.636107924124584], batch size: 70
+2022-12-13 18:26:41,699 INFO [train.py:421] (4/8) Epoch 10, batch 24400, loss[loss=2.53, over 840.00 frames. , ppl: 12.552877864119512] tot_loss[loss=2.265, over 5597739.92 frames. , ppl: 9.633564331063843], batch size: 70
+2022-12-13 18:28:25,417 INFO [train.py:421] (4/8) Epoch 10, batch 24600, loss[loss=2.19, over 6300.00 frames. , ppl: 8.93209176481123] tot_loss[loss=2.267, over 5537421.14 frames. , ppl: 9.648828517875149], batch size: 70
+2022-12-13 18:30:06,363 INFO [train.py:421] (4/8) Epoch 10, batch 24800, loss[loss=2.202, over 4830.00 frames. , ppl: 9.044742213622488] tot_loss[loss=2.267, over 5536458.70 frames. , ppl: 9.647832255895711], batch size: 70
+2022-12-13 18:31:48,362 INFO [train.py:421] (4/8) Epoch 10, batch 25000, loss[loss=2.577, over 1120.00 frames. , ppl: 13.1580447895151] tot_loss[loss=2.268, over 5500769.75 frames. , ppl: 9.660012731196629], batch size: 70
+2022-12-13 18:31:48,362 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:31:49,120 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617161851458512
+2022-12-13 18:33:31,042 INFO [train.py:421] (4/8) Epoch 10, batch 25200, loss[loss=4.854, over 280.00 frames. , ppl: 128.18903799369377] tot_loss[loss=2.269, over 5466514.41 frames. , ppl: 9.672104758378971], batch size: 70
+2022-12-13 18:35:13,475 INFO [train.py:421] (4/8) Epoch 10, batch 25400, loss[loss=2.473, over 1610.00 frames. , ppl: 11.860017723437775] tot_loss[loss=2.269, over 5454439.92 frames. , ppl: 9.670301820272591], batch size: 70
+2022-12-13 18:36:55,994 INFO [train.py:421] (4/8) Epoch 10, batch 25600, loss[loss=2.233, over 3780.00 frames. , ppl: 9.32362078907306] tot_loss[loss=2.27, over 5453327.22 frames. , ppl: 9.677683481053535], batch size: 70
+2022-12-13 18:38:38,936 INFO [train.py:421] (4/8) Epoch 10, batch 25800, loss[loss=2.261, over 3570.00 frames. , ppl: 9.59333444791543] tot_loss[loss=2.269, over 5481344.48 frames. , ppl: 9.670160742430106], batch size: 70
+2022-12-13 18:40:24,891 INFO [train.py:421] (4/8) Epoch 10, batch 26000, loss[loss=2.226, over 4760.00 frames. , ppl: 9.2611595063189] tot_loss[loss=2.268, over 5483388.49 frames. , ppl: 9.663166629263289], batch size: 70
+2022-12-13 18:40:24,892 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:40:25,675 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.637863426039562
+2022-12-13 18:42:06,979 INFO [train.py:421] (4/8) Epoch 10, batch 26200, loss[loss=2.55, over 910.00 frames. , ppl: 12.81049897274048] tot_loss[loss=2.269, over 5466889.74 frames. , ppl: 9.666112103556411], batch size: 70
+2022-12-13 18:43:50,151 INFO [train.py:421] (4/8) Epoch 10, batch 26400, loss[loss=2.476, over 1190.00 frames. , ppl: 11.892205846095335] tot_loss[loss=2.267, over 5492721.94 frames. , ppl: 9.652598923817436], batch size: 70
+2022-12-13 18:45:37,768 INFO [train.py:421] (4/8) Epoch 10, batch 26600, loss[loss=2.565, over 770.00 frames. , ppl: 12.994335425811276] tot_loss[loss=2.267, over 5493290.38 frames. , ppl: 9.65241868538002], batch size: 70
+2022-12-13 18:47:19,961 INFO [train.py:421] (4/8) Epoch 10, batch 26800, loss[loss=2.369, over 1750.00 frames. , ppl: 10.682920010198078] tot_loss[loss=2.268, over 5471379.35 frames. , ppl: 9.656146715209754], batch size: 70
+2022-12-13 18:49:02,367 INFO [train.py:421] (4/8) Epoch 10, batch 27000, loss[loss=2.352, over 1960.00 frames. , ppl: 10.51036262199583] tot_loss[loss=2.269, over 5457809.83 frames. , ppl: 9.66596424724515], batch size: 70
+2022-12-13 18:49:02,367 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:49:03,131 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621924342548795
+2022-12-13 18:50:45,266 INFO [train.py:421] (4/8) Epoch 10, batch 27200, loss[loss=2.272, over 2170.00 frames. , ppl: 9.699417306517987] tot_loss[loss=2.268, over 5458452.90 frames. , ppl: 9.664648616040607], batch size: 70
+2022-12-13 18:52:26,330 INFO [train.py:421] (4/8) Epoch 10, batch 27400, loss[loss=2.36, over 1820.00 frames. , ppl: 10.590509338538343] tot_loss[loss=2.269, over 5456033.30 frames. , ppl: 9.665159047275587], batch size: 70
+2022-12-13 18:54:07,431 INFO [train.py:421] (4/8) Epoch 10, batch 27600, loss[loss=2.322, over 1330.00 frames. , ppl: 10.194573227365265] tot_loss[loss=2.268, over 5465268.26 frames. , ppl: 9.660302986369809], batch size: 70
+2022-12-13 18:55:51,707 INFO [train.py:421] (4/8) Epoch 10, batch 27800, loss[loss=2.316, over 2520.00 frames. , ppl: 10.130927111464942] tot_loss[loss=2.267, over 5507597.29 frames. , ppl: 9.64630097203306], batch size: 70
+2022-12-13 18:57:33,179 INFO [train.py:421] (4/8) Epoch 10, batch 28000, loss[loss=2.263, over 5320.00 frames. , ppl: 9.612739419856814] tot_loss[loss=2.268, over 5484157.57 frames. , ppl: 9.656166963917416], batch size: 70
+2022-12-13 18:57:33,179 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 18:57:33,931 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617768243976228
+2022-12-13 18:59:13,885 INFO [train.py:421] (4/8) Epoch 10, batch 28200, loss[loss=2.353, over 2240.00 frames. , ppl: 10.512247692990918] tot_loss[loss=2.267, over 5505240.53 frames. , ppl: 9.646183993311686], batch size: 70
+2022-12-13 19:00:57,580 INFO [train.py:421] (4/8) Epoch 10, batch 28400, loss[loss=2.192, over 4620.00 frames. , ppl: 8.95190363850801] tot_loss[loss=2.267, over 5515716.57 frames. , ppl: 9.647103741311144], batch size: 70
+2022-12-13 19:02:36,602 INFO [train.py:421] (4/8) Epoch 10, batch 28600, loss[loss=2.602, over 1050.00 frames. , ppl: 13.490392583027916] tot_loss[loss=2.268, over 5501644.96 frames. , ppl: 9.655654631295821], batch size: 70
+2022-12-13 19:04:16,638 INFO [train.py:421] (4/8) Epoch 10, batch 28800, loss[loss=2.265, over 2170.00 frames. , ppl: 9.628159540872616] tot_loss[loss=2.267, over 5508733.17 frames. , ppl: 9.654236567742423], batch size: 70
+2022-12-13 19:06:00,382 INFO [train.py:421] (4/8) Epoch 10, batch 29000, loss[loss=2.322, over 1960.00 frames. , ppl: 10.19621935548979] tot_loss[loss=2.267, over 5500389.91 frames. , ppl: 9.653296015532119], batch size: 70
+2022-12-13 19:06:00,383 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:06:01,145 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.614629935321982
+2022-12-13 19:07:42,003 INFO [train.py:421] (4/8) Epoch 10, batch 29200, loss[loss=2.244, over 3220.00 frames. , ppl: 9.42886547119562] tot_loss[loss=2.268, over 5479685.98 frames. , ppl: 9.658359946727263], batch size: 70
+2022-12-13 19:09:24,501 INFO [train.py:421] (4/8) Epoch 10, batch 29400, loss[loss=2.363, over 1610.00 frames. , ppl: 10.621967784699956] tot_loss[loss=2.269, over 5437859.67 frames. , ppl: 9.674401518958954], batch size: 70
+2022-12-13 19:11:08,697 INFO [train.py:421] (4/8) Epoch 10, batch 29600, loss[loss=2.12, over 7420.00 frames. , ppl: 8.327454570097622] tot_loss[loss=2.268, over 5456128.98 frames. , ppl: 9.664409097237005], batch size: 70
+2022-12-13 19:12:55,138 INFO [train.py:421] (4/8) Epoch 10, batch 29800, loss[loss=2.134, over 6300.00 frames. , ppl: 8.447330264586336] tot_loss[loss=2.269, over 5430884.33 frames. , ppl: 9.66889887739272], batch size: 70
+2022-12-13 19:14:36,456 INFO [train.py:421] (4/8) Epoch 10, batch 30000, loss[loss=2.193, over 5810.00 frames. , ppl: 8.959823948650417] tot_loss[loss=2.268, over 5464992.23 frames. , ppl: 9.664649201011903], batch size: 70
+2022-12-13 19:14:36,456 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:14:37,217 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.266, over 211138.00 frames. , ppl: 9.638813505699115
+2022-12-13 19:16:16,748 INFO [train.py:421] (4/8) Epoch 10, batch 30200, loss[loss=2.28, over 1750.00 frames. , ppl: 9.779747014473717] tot_loss[loss=2.268, over 5487255.10 frames. , ppl: 9.659442780619873], batch size: 70
+2022-12-13 19:18:02,135 INFO [train.py:421] (4/8) Epoch 10, batch 30400, loss[loss=2.468, over 840.00 frames. , ppl: 11.799753167935293] tot_loss[loss=2.269, over 5498730.64 frames. , ppl: 9.66548290874432], batch size: 70
+2022-12-13 19:19:43,254 INFO [train.py:421] (4/8) Epoch 10, batch 30600, loss[loss=2.286, over 2590.00 frames. , ppl: 9.840193381645642] tot_loss[loss=2.269, over 5492829.33 frames. , ppl: 9.670554321272991], batch size: 70
+2022-12-13 19:21:25,066 INFO [train.py:421] (4/8) Epoch 10, batch 30800, loss[loss=2.14, over 10920.00 frames. , ppl: 8.501448164606535] tot_loss[loss=2.268, over 5533483.94 frames. , ppl: 9.66052173847161], batch size: 70
+2022-12-13 19:23:07,026 INFO [train.py:421] (4/8) Epoch 10, batch 31000, loss[loss=2.32, over 3290.00 frames. , ppl: 10.170694248377503] tot_loss[loss=2.27, over 5505074.90 frames. , ppl: 9.676878406584981], batch size: 70
+2022-12-13 19:23:07,026 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:23:07,797 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.624022295019808
+2022-12-13 19:24:49,278 INFO [train.py:421] (4/8) Epoch 10, batch 31200, loss[loss=2.33, over 1260.00 frames. , ppl: 10.27520202191224] tot_loss[loss=2.27, over 5506206.94 frames. , ppl: 9.67988950627957], batch size: 70
+2022-12-13 19:26:30,755 INFO [train.py:421] (4/8) Epoch 10, batch 31400, loss[loss=2.21, over 3990.00 frames. , ppl: 9.11257198477324] tot_loss[loss=2.27, over 5522202.31 frames. , ppl: 9.674826714762084], batch size: 70
+2022-12-13 19:28:10,486 INFO [train.py:421] (4/8) Epoch 10, batch 31600, loss[loss=2.221, over 1610.00 frames. , ppl: 9.219657511542147] tot_loss[loss=2.268, over 5557190.89 frames. , ppl: 9.658212394254791], batch size: 70
+2022-12-13 19:29:51,154 INFO [train.py:421] (4/8) Epoch 10, batch 31800, loss[loss=2.767, over 770.00 frames. , ppl: 15.918244173518547] tot_loss[loss=2.268, over 5563018.02 frames. , ppl: 9.657304102568188], batch size: 70
+2022-12-13 19:31:35,513 INFO [train.py:421] (4/8) Epoch 10, batch 32000, loss[loss=2.439, over 1120.00 frames. , ppl: 11.458112941349034] tot_loss[loss=2.269, over 5533647.76 frames. , ppl: 9.666207494025004], batch size: 70
+2022-12-13 19:31:35,513 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:31:36,277 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616229561960207
+2022-12-13 19:33:19,163 INFO [train.py:421] (4/8) Epoch 10, batch 32200, loss[loss=2.153, over 4480.00 frames. , ppl: 8.610598036708154] tot_loss[loss=2.267, over 5579846.71 frames. , ppl: 9.654442178148829], batch size: 70
+2022-12-13 19:35:01,660 INFO [train.py:421] (4/8) Epoch 10, batch 32400, loss[loss=2.164, over 4550.00 frames. , ppl: 8.709037361913627] tot_loss[loss=2.267, over 5602674.37 frames. , ppl: 9.647005636369762], batch size: 70
+2022-12-13 19:36:43,533 INFO [train.py:421] (4/8) Epoch 10, batch 32600, loss[loss=2.22, over 4830.00 frames. , ppl: 9.206484993130523] tot_loss[loss=2.267, over 5589027.26 frames. , ppl: 9.654931712138005], batch size: 70
+2022-12-13 19:38:26,633 INFO [train.py:421] (4/8) Epoch 10, batch 32800, loss[loss=2.408, over 1260.00 frames. , ppl: 11.108067373446419] tot_loss[loss=2.269, over 5555709.68 frames. , ppl: 9.665314031426366], batch size: 70
+2022-12-13 19:40:07,395 INFO [train.py:421] (4/8) Epoch 10, batch 33000, loss[loss=2.516, over 1260.00 frames. , ppl: 12.383886779893995] tot_loss[loss=2.268, over 5580129.29 frames. , ppl: 9.660990214505151], batch size: 70
+2022-12-13 19:40:07,395 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:40:08,159 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.611626373467763
+2022-12-13 19:41:48,451 INFO [train.py:421] (4/8) Epoch 10, batch 33200, loss[loss=2.122, over 6230.00 frames. , ppl: 8.348272550992709] tot_loss[loss=2.269, over 5545602.46 frames. , ppl: 9.665486935983994], batch size: 70
+2022-12-13 19:43:28,468 INFO [train.py:421] (4/8) Epoch 10, batch 33400, loss[loss=2.192, over 4620.00 frames. , ppl: 8.955033930804365] tot_loss[loss=2.268, over 5569232.63 frames. , ppl: 9.662435620441135], batch size: 70
+2022-12-13 19:45:11,012 INFO [train.py:421] (4/8) Epoch 10, batch 33600, loss[loss=2.241, over 5320.00 frames. , ppl: 9.404830308289853] tot_loss[loss=2.269, over 5560300.06 frames. , ppl: 9.66866934601441], batch size: 70
+2022-12-13 19:46:54,750 INFO [train.py:421] (4/8) Epoch 10, batch 33800, loss[loss=2.262, over 3500.00 frames. , ppl: 9.600936353884581] tot_loss[loss=2.268, over 5568000.78 frames. , ppl: 9.660414213758386], batch size: 70
+2022-12-13 19:48:42,876 INFO [train.py:421] (4/8) Epoch 10, batch 34000, loss[loss=2.252, over 2100.00 frames. , ppl: 9.509127617147739] tot_loss[loss=2.267, over 5582817.86 frames. , ppl: 9.652732806684085], batch size: 70
+2022-12-13 19:48:42,877 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:48:43,659 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612292169791925
+2022-12-13 19:50:27,173 INFO [train.py:421] (4/8) Epoch 10, batch 34200, loss[loss=2.475, over 910.00 frames. , ppl: 11.876910592350008] tot_loss[loss=2.267, over 5568479.00 frames. , ppl: 9.65296067174941], batch size: 70
+2022-12-13 19:52:09,000 INFO [train.py:421] (4/8) Epoch 10, batch 34400, loss[loss=2.183, over 5740.00 frames. , ppl: 8.870720446893474] tot_loss[loss=2.268, over 5534386.56 frames. , ppl: 9.663431953054273], batch size: 70
+2022-12-13 19:53:50,207 INFO [train.py:421] (4/8) Epoch 10, batch 34600, loss[loss=2.363, over 2030.00 frames. , ppl: 10.627528878495012] tot_loss[loss=2.27, over 5495712.30 frames. , ppl: 9.674725177371501], batch size: 70
+2022-12-13 19:55:28,936 INFO [train.py:421] (4/8) Epoch 10, batch 34800, loss[loss=2.194, over 4550.00 frames. , ppl: 8.972138904679632] tot_loss[loss=2.27, over 5466036.46 frames. , ppl: 9.679685768327872], batch size: 70
+2022-12-13 19:57:12,624 INFO [train.py:421] (4/8) Epoch 10, batch 35000, loss[loss=2.291, over 2660.00 frames. , ppl: 9.887719587766835] tot_loss[loss=2.271, over 5419263.91 frames. , ppl: 9.686732421236012], batch size: 70
+2022-12-13 19:57:12,625 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 19:57:13,388 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616518490872702
+2022-12-13 19:58:57,925 INFO [train.py:421] (4/8) Epoch 10, batch 35200, loss[loss=2.268, over 2730.00 frames. , ppl: 9.663073824259172] tot_loss[loss=2.27, over 5433850.76 frames. , ppl: 9.68270182184647], batch size: 70
+2022-12-13 20:00:39,614 INFO [train.py:421] (4/8) Epoch 10, batch 35400, loss[loss=2.61, over 910.00 frames. , ppl: 13.595517431569524] tot_loss[loss=2.269, over 5472404.34 frames. , ppl: 9.668979941271633], batch size: 70
+2022-12-13 20:02:22,797 INFO [train.py:421] (4/8) Epoch 10, batch 35600, loss[loss=2.383, over 2240.00 frames. , ppl: 10.838698317243775] tot_loss[loss=2.269, over 5453143.36 frames. , ppl: 9.672921723039632], batch size: 70
+2022-12-13 20:04:05,805 INFO [train.py:421] (4/8) Epoch 10, batch 35800, loss[loss=2.32, over 2170.00 frames. , ppl: 10.173743836724064] tot_loss[loss=2.27, over 5437717.64 frames. , ppl: 9.680635956296873], batch size: 70
+2022-12-13 20:05:47,822 INFO [train.py:421] (4/8) Epoch 10, batch 36000, loss[loss=2.166, over 2940.00 frames. , ppl: 8.720591905511482] tot_loss[loss=2.269, over 5472467.31 frames. , ppl: 9.670852585381187], batch size: 70
+2022-12-13 20:05:47,823 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:05:48,581 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.601890816854365
+2022-12-13 20:07:33,622 INFO [train.py:421] (4/8) Epoch 10, batch 36200, loss[loss=2.303, over 2450.00 frames. , ppl: 9.999157077526228] tot_loss[loss=2.268, over 5521389.34 frames. , ppl: 9.664335918324163], batch size: 70
+2022-12-13 20:09:17,734 INFO [train.py:421] (4/8) Epoch 10, batch 36400, loss[loss=2.446, over 2170.00 frames. , ppl: 11.540648854177027] tot_loss[loss=2.269, over 5505587.06 frames. , ppl: 9.67176505731002], batch size: 70
+2022-12-13 20:11:03,422 INFO [train.py:421] (4/8) Epoch 10, batch 36600, loss[loss=2.121, over 8680.00 frames. , ppl: 8.337815330521856] tot_loss[loss=2.268, over 5534189.91 frames. , ppl: 9.66204873527777], batch size: 70
+2022-12-13 20:12:48,784 INFO [train.py:421] (4/8) Epoch 10, batch 36800, loss[loss=2.38, over 2520.00 frames. , ppl: 10.807089205429916] tot_loss[loss=2.268, over 5555267.78 frames. , ppl: 9.656383793888326], batch size: 70
+2022-12-13 20:14:30,489 INFO [train.py:421] (4/8) Epoch 10, batch 37000, loss[loss=2.259, over 3640.00 frames. , ppl: 9.57613042107374] tot_loss[loss=2.268, over 5543158.59 frames. , ppl: 9.66126488951638], batch size: 70
+2022-12-13 20:14:30,489 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:14:31,281 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.600985586014557
+2022-12-13 20:16:16,954 INFO [train.py:421] (4/8) Epoch 10, batch 37200, loss[loss=2.609, over 770.00 frames. , ppl: 13.586017376499113] tot_loss[loss=2.267, over 5589355.24 frames. , ppl: 9.64739201434093], batch size: 70
+2022-12-13 20:17:58,370 INFO [train.py:421] (4/8) Epoch 10, batch 37400, loss[loss=2.348, over 1680.00 frames. , ppl: 10.461211455996333] tot_loss[loss=2.268, over 5561122.48 frames. , ppl: 9.656073540523941], batch size: 70
+2022-12-13 20:19:44,288 INFO [train.py:421] (4/8) Epoch 10, batch 37600, loss[loss=2.153, over 4200.00 frames. , ppl: 8.607720259944442] tot_loss[loss=2.268, over 5528711.06 frames. , ppl: 9.66382331875825], batch size: 70
+2022-12-13 20:21:26,441 INFO [train.py:421] (4/8) Epoch 10, batch 37800, loss[loss=2.203, over 2730.00 frames. , ppl: 9.047926941239668] tot_loss[loss=2.27, over 5464294.08 frames. , ppl: 9.678589518153924], batch size: 70
+2022-12-13 20:23:05,363 INFO [train.py:421] (4/8) Epoch 10, batch 38000, loss[loss=2.274, over 2310.00 frames. , ppl: 9.717417197002307] tot_loss[loss=2.27, over 5494356.68 frames. , ppl: 9.678376420633928], batch size: 70
+2022-12-13 20:23:05,363 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:23:06,141 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589634108742406
+2022-12-13 20:24:47,953 INFO [train.py:421] (4/8) Epoch 10, batch 38200, loss[loss=2.415, over 1190.00 frames. , ppl: 11.194873474304247] tot_loss[loss=2.27, over 5493277.12 frames. , ppl: 9.679289277524784], batch size: 70
+2022-12-13 20:26:26,983 INFO [train.py:421] (4/8) Epoch 10, batch 38400, loss[loss=2.16, over 7700.00 frames. , ppl: 8.671872308083657] tot_loss[loss=2.269, over 5547377.32 frames. , ppl: 9.668750546006239], batch size: 70
+2022-12-13 20:28:07,272 INFO [train.py:421] (4/8) Epoch 10, batch 38600, loss[loss=2.424, over 1050.00 frames. , ppl: 11.293996466248565] tot_loss[loss=2.27, over 5530265.90 frames. , ppl: 9.674996215172701], batch size: 70
+2022-12-13 20:29:50,034 INFO [train.py:421] (4/8) Epoch 10, batch 38800, loss[loss=2.557, over 770.00 frames. , ppl: 12.895455166117443] tot_loss[loss=2.269, over 5566497.95 frames. , ppl: 9.666569655336708], batch size: 70
+2022-12-13 20:31:30,515 INFO [train.py:421] (4/8) Epoch 10, batch 39000, loss[loss=2.579, over 910.00 frames. , ppl: 13.177598543252408] tot_loss[loss=2.269, over 5546393.83 frames. , ppl: 9.672041263586978], batch size: 70
+2022-12-13 20:31:30,515 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:31:31,277 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59960729794989
+2022-12-13 20:33:11,828 INFO [train.py:421] (4/8) Epoch 10, batch 39200, loss[loss=2.469, over 1610.00 frames. , ppl: 11.813020348742796] tot_loss[loss=2.27, over 5510797.50 frames. , ppl: 9.682661790207282], batch size: 70
+2022-12-13 20:34:51,224 INFO [train.py:421] (4/8) Epoch 10, batch 39400, loss[loss=2.481, over 980.00 frames. , ppl: 11.95665903926364] tot_loss[loss=2.27, over 5489211.94 frames. , ppl: 9.680523886432114], batch size: 70
+2022-12-13 20:36:33,968 INFO [train.py:421] (4/8) Epoch 10, batch 39600, loss[loss=2.154, over 6930.00 frames. , ppl: 8.620699621010631] tot_loss[loss=2.27, over 5517546.44 frames. , ppl: 9.679383231868742], batch size: 70
+2022-12-13 20:38:17,862 INFO [train.py:421] (4/8) Epoch 10, batch 39800, loss[loss=2.494, over 1750.00 frames. , ppl: 12.105611250524424] tot_loss[loss=2.27, over 5520049.48 frames. , ppl: 9.675402903056662], batch size: 70
+2022-12-13 20:40:01,100 INFO [train.py:421] (4/8) Epoch 10, batch 40000, loss[loss=2.222, over 5530.00 frames. , ppl: 9.229867480080303] tot_loss[loss=2.27, over 5488982.54 frames. , ppl: 9.679714775819226], batch size: 70
+2022-12-13 20:40:01,100 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:40:01,849 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.261, over 211138.00 frames. , ppl: 9.595198105641067
+2022-12-13 20:41:45,224 INFO [train.py:421] (4/8) Epoch 10, batch 40200, loss[loss=2.267, over 4130.00 frames. , ppl: 9.650077446907543] tot_loss[loss=2.27, over 5519040.06 frames. , ppl: 9.676184659581482], batch size: 70
+2022-12-13 20:43:28,841 INFO [train.py:421] (4/8) Epoch 10, batch 40400, loss[loss=2.257, over 3430.00 frames. , ppl: 9.552312442542075] tot_loss[loss=2.271, over 5448955.83 frames. , ppl: 9.693698289161562], batch size: 70
+2022-12-13 20:45:07,759 INFO [train.py:421] (4/8) Epoch 10, batch 40600, loss[loss=2.267, over 1890.00 frames. , ppl: 9.647498915820067] tot_loss[loss=2.271, over 5460951.96 frames. , ppl: 9.688282863379904], batch size: 70
+2022-12-13 20:46:47,451 INFO [train.py:421] (4/8) Epoch 10, batch 40800, loss[loss=2.4, over 1960.00 frames. , ppl: 11.02255851921991] tot_loss[loss=2.271, over 5447707.62 frames. , ppl: 9.691101934802203], batch size: 70
+2022-12-13 20:48:25,946 INFO [train.py:421] (4/8) Epoch 10, batch 41000, loss[loss=2.419, over 1400.00 frames. , ppl: 11.234255462023691] tot_loss[loss=2.272, over 5439440.08 frames. , ppl: 9.695734357562463], batch size: 70
+2022-12-13 20:48:25,946 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:48:26,717 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60531213793815
+2022-12-13 20:50:09,272 INFO [train.py:421] (4/8) Epoch 10, batch 41200, loss[loss=2.499, over 1050.00 frames. , ppl: 12.174024566174818] tot_loss[loss=2.271, over 5439566.45 frames. , ppl: 9.692779884750838], batch size: 70
+2022-12-13 20:51:52,024 INFO [train.py:421] (4/8) Epoch 10, batch 41400, loss[loss=2.201, over 4760.00 frames. , ppl: 9.02963435780591] tot_loss[loss=2.269, over 5496512.56 frames. , ppl: 9.669800698027565], batch size: 70
+2022-12-13 20:53:33,166 INFO [train.py:421] (4/8) Epoch 10, batch 41600, loss[loss=2.559, over 1120.00 frames. , ppl: 12.92158185918368] tot_loss[loss=2.269, over 5500669.52 frames. , ppl: 9.665078183153955], batch size: 70
+2022-12-13 20:55:14,367 INFO [train.py:421] (4/8) Epoch 10, batch 41800, loss[loss=2.191, over 5530.00 frames. , ppl: 8.947633826652517] tot_loss[loss=2.268, over 5518218.78 frames. , ppl: 9.6588193305486], batch size: 70
+2022-12-13 20:56:55,445 INFO [train.py:421] (4/8) Epoch 10, batch 42000, loss[loss=2.341, over 1680.00 frames. , ppl: 10.39163434782252] tot_loss[loss=2.268, over 5517102.68 frames. , ppl: 9.659187884983448], batch size: 70
+2022-12-13 20:56:55,445 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 20:56:56,194 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599615822834881
+2022-12-13 20:58:38,511 INFO [train.py:421] (4/8) Epoch 10, batch 42200, loss[loss=2.274, over 2870.00 frames. , ppl: 9.718658517678717] tot_loss[loss=2.268, over 5518558.39 frames. , ppl: 9.657893419213826], batch size: 70
+2022-12-13 21:00:23,424 INFO [train.py:421] (4/8) Epoch 10, batch 42400, loss[loss=2.22, over 3920.00 frames. , ppl: 9.20300447751686] tot_loss[loss=2.267, over 5544112.27 frames. , ppl: 9.652210134462184], batch size: 70
+2022-12-13 21:02:03,395 INFO [train.py:421] (4/8) Epoch 10, batch 42600, loss[loss=2.291, over 2800.00 frames. , ppl: 9.880391204616231] tot_loss[loss=2.268, over 5518131.86 frames. , ppl: 9.662154554608957], batch size: 70
+2022-12-13 21:03:47,730 INFO [train.py:421] (4/8) Epoch 10, batch 42800, loss[loss=2.516, over 1120.00 frames. , ppl: 12.381242178570727] tot_loss[loss=2.268, over 5549149.26 frames. , ppl: 9.661104791712425], batch size: 70
+2022-12-13 21:05:29,126 INFO [train.py:421] (4/8) Epoch 10, batch 43000, loss[loss=2.459, over 1330.00 frames. , ppl: 11.687966293508786] tot_loss[loss=2.268, over 5553425.36 frames. , ppl: 9.65776150665654], batch size: 70
+2022-12-13 21:05:29,127 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:05:29,892 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.605019280876657
+2022-12-13 21:07:11,598 INFO [train.py:421] (4/8) Epoch 10, batch 43200, loss[loss=2.243, over 3710.00 frames. , ppl: 9.417123600665706] tot_loss[loss=2.267, over 5550188.68 frames. , ppl: 9.654661025608439], batch size: 70
+2022-12-13 21:08:55,574 INFO [train.py:421] (4/8) Epoch 10, batch 43400, loss[loss=2.194, over 10640.00 frames. , ppl: 8.971641585674993] tot_loss[loss=2.267, over 5551790.98 frames. , ppl: 9.650722745325321], batch size: 70
+2022-12-13 21:10:37,373 INFO [train.py:421] (4/8) Epoch 10, batch 43600, loss[loss=2.35, over 3290.00 frames. , ppl: 10.489809678059983] tot_loss[loss=2.268, over 5513231.66 frames. , ppl: 9.65853794706755], batch size: 70
+2022-12-13 21:12:19,476 INFO [train.py:421] (4/8) Epoch 10, batch 43800, loss[loss=2.358, over 2030.00 frames. , ppl: 10.574091976530687] tot_loss[loss=2.268, over 5550004.20 frames. , ppl: 9.656287824198229], batch size: 70
+2022-12-13 21:13:59,360 INFO [train.py:421] (4/8) Epoch 10, batch 44000, loss[loss=2.226, over 2380.00 frames. , ppl: 9.2663784345867] tot_loss[loss=2.268, over 5526839.11 frames. , ppl: 9.660629219209982], batch size: 70
+2022-12-13 21:13:59,361 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:14:00,110 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.620145785284024
+2022-12-13 21:15:38,911 INFO [train.py:421] (4/8) Epoch 10, batch 44200, loss[loss=2.189, over 4270.00 frames. , ppl: 8.923166160220017] tot_loss[loss=2.27, over 5474586.38 frames. , ppl: 9.676031548453004], batch size: 70
+2022-12-13 21:17:20,593 INFO [train.py:421] (4/8) Epoch 10, batch 44400, loss[loss=2.642, over 770.00 frames. , ppl: 14.039443363242093] tot_loss[loss=2.268, over 5513130.87 frames. , ppl: 9.663111386469607], batch size: 70
+2022-12-13 21:19:00,922 INFO [train.py:421] (4/8) Epoch 10, batch 44600, loss[loss=2.214, over 2940.00 frames. , ppl: 9.149267925867457] tot_loss[loss=2.269, over 5495923.09 frames. , ppl: 9.66599686470941], batch size: 70
+2022-12-13 21:20:45,693 INFO [train.py:421] (4/8) Epoch 10, batch 44800, loss[loss=2.914, over 630.00 frames. , ppl: 18.42752885365845] tot_loss[loss=2.269, over 5491296.99 frames. , ppl: 9.66866408338516], batch size: 70
+2022-12-13 21:22:25,934 INFO [train.py:421] (4/8) Epoch 10, batch 45000, loss[loss=2.306, over 2940.00 frames. , ppl: 10.031605172708707] tot_loss[loss=2.27, over 5460937.52 frames. , ppl: 9.675366737376761], batch size: 70
+2022-12-13 21:22:25,934 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:22:26,698 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.598352801576091
+2022-12-13 21:24:05,946 INFO [train.py:421] (4/8) Epoch 10, batch 45200, loss[loss=2.262, over 3010.00 frames. , ppl: 9.602835678358366] tot_loss[loss=2.268, over 5494130.87 frames. , ppl: 9.659612428533782], batch size: 70
+2022-12-13 21:25:48,527 INFO [train.py:421] (4/8) Epoch 10, batch 45400, loss[loss=3.139, over 490.00 frames. , ppl: 23.08637498430284] tot_loss[loss=2.269, over 5488852.52 frames. , ppl: 9.665320659173013], batch size: 70
+2022-12-13 21:27:30,593 INFO [train.py:421] (4/8) Epoch 10, batch 45600, loss[loss=2.216, over 2240.00 frames. , ppl: 9.169918805981602] tot_loss[loss=2.268, over 5518582.50 frames. , ppl: 9.660087904390402], batch size: 70
+2022-12-13 21:29:13,826 INFO [train.py:421] (4/8) Epoch 10, batch 45800, loss[loss=2.386, over 1190.00 frames. , ppl: 10.868623445300939] tot_loss[loss=2.267, over 5544477.55 frames. , ppl: 9.65356907555735], batch size: 70
+2022-12-13 21:30:58,371 INFO [train.py:421] (4/8) Epoch 10, batch 46000, loss[loss=2.236, over 4410.00 frames. , ppl: 9.353156153158839] tot_loss[loss=2.266, over 5571470.26 frames. , ppl: 9.644145370222962], batch size: 70
+2022-12-13 21:30:58,372 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:30:59,139 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.603299281294177
+2022-12-13 21:32:38,622 INFO [train.py:421] (4/8) Epoch 10, batch 46200, loss[loss=2.18, over 5530.00 frames. , ppl: 8.847350495426985] tot_loss[loss=2.266, over 5603427.28 frames. , ppl: 9.638114504538263], batch size: 70
+2022-12-13 21:34:24,216 INFO [train.py:421] (4/8) Epoch 10, batch 46400, loss[loss=2.287, over 2520.00 frames. , ppl: 9.844339538755335] tot_loss[loss=2.266, over 5612288.20 frames. , ppl: 9.63942309304397], batch size: 70
+2022-12-13 21:36:10,344 INFO [train.py:421] (4/8) Epoch 10, batch 46600, loss[loss=2.35, over 1190.00 frames. , ppl: 10.487015445210345] tot_loss[loss=2.266, over 5618216.40 frames. , ppl: 9.64011192269349], batch size: 70
+2022-12-13 21:37:49,342 INFO [train.py:421] (4/8) Epoch 10, batch 46800, loss[loss=2.227, over 3290.00 frames. , ppl: 9.270910451341209] tot_loss[loss=2.267, over 5568059.80 frames. , ppl: 9.646974079272814], batch size: 70
+2022-12-13 21:39:32,511 INFO [train.py:421] (4/8) Epoch 10, batch 47000, loss[loss=2.107, over 7350.00 frames. , ppl: 8.225380603111065] tot_loss[loss=2.266, over 5591658.84 frames. , ppl: 9.637559648286405], batch size: 70
+2022-12-13 21:39:32,512 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:39:33,262 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.62268485087924
+2022-12-13 21:41:16,053 INFO [train.py:421] (4/8) Epoch 10, batch 47200, loss[loss=4.891, over 280.00 frames. , ppl: 133.07341469235868] tot_loss[loss=2.268, over 5554893.88 frames. , ppl: 9.65553187783624], batch size: 70
+2022-12-13 21:42:58,725 INFO [train.py:421] (4/8) Epoch 10, batch 47400, loss[loss=2.378, over 1050.00 frames. , ppl: 10.783303621578488] tot_loss[loss=2.268, over 5549196.33 frames. , ppl: 9.658828103585357], batch size: 70
+2022-12-13 21:44:41,590 INFO [train.py:421] (4/8) Epoch 10, batch 47600, loss[loss=2.285, over 1820.00 frames. , ppl: 9.823399411552183] tot_loss[loss=2.268, over 5546358.76 frames. , ppl: 9.657588174906472], batch size: 70
+2022-12-13 21:46:26,614 INFO [train.py:421] (4/8) Epoch 10, batch 47800, loss[loss=2.446, over 1050.00 frames. , ppl: 11.545229500973365] tot_loss[loss=2.267, over 5559395.80 frames. , ppl: 9.647064219733045], batch size: 70
+2022-12-13 21:48:08,378 INFO [train.py:421] (4/8) Epoch 10, batch 48000, loss[loss=2.281, over 2800.00 frames. , ppl: 9.789392008394941] tot_loss[loss=2.266, over 5588916.19 frames. , ppl: 9.639342943802001], batch size: 70
+2022-12-13 21:48:08,379 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:48:09,144 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621020070987242
+2022-12-13 21:49:55,425 INFO [train.py:421] (4/8) Epoch 10, batch 48200, loss[loss=2.286, over 3430.00 frames. , ppl: 9.834031326723268] tot_loss[loss=2.265, over 5578724.38 frames. , ppl: 9.63548627508952], batch size: 70
+2022-12-13 21:51:37,488 INFO [train.py:421] (4/8) Epoch 10, batch 48400, loss[loss=2.248, over 4410.00 frames. , ppl: 9.469994874190759] tot_loss[loss=2.266, over 5554520.04 frames. , ppl: 9.641956681826288], batch size: 70
+2022-12-13 21:53:19,230 INFO [train.py:421] (4/8) Epoch 10, batch 48600, loss[loss=2.183, over 4410.00 frames. , ppl: 8.875811699246954] tot_loss[loss=2.265, over 5597968.99 frames. , ppl: 9.632791934227882], batch size: 70
+2022-12-13 21:55:03,402 INFO [train.py:421] (4/8) Epoch 10, batch 48800, loss[loss=3.026, over 560.00 frames. , ppl: 20.604583833737017] tot_loss[loss=2.265, over 5641681.76 frames. , ppl: 9.627372140436833], batch size: 70
+2022-12-13 21:56:44,857 INFO [train.py:421] (4/8) Epoch 10, batch 49000, loss[loss=2.788, over 630.00 frames. , ppl: 16.25561821075779] tot_loss[loss=2.264, over 5650572.91 frames. , ppl: 9.62476944122957], batch size: 70
+2022-12-13 21:56:44,858 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 21:56:45,606 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608761699260915
+2022-12-13 21:58:25,261 INFO [train.py:421] (4/8) Epoch 10, batch 49200, loss[loss=2.635, over 770.00 frames. , ppl: 13.940402899606218] tot_loss[loss=2.264, over 5651738.07 frames. , ppl: 9.625652287084533], batch size: 70
+2022-12-13 22:00:03,893 INFO [train.py:421] (4/8) Epoch 10, batch 49400, loss[loss=2.284, over 4060.00 frames. , ppl: 9.816673050968829] tot_loss[loss=2.265, over 5609349.71 frames. , ppl: 9.635483337909703], batch size: 70
+2022-12-13 22:01:44,548 INFO [train.py:421] (4/8) Epoch 10, batch 49600, loss[loss=2.559, over 770.00 frames. , ppl: 12.924866014751904] tot_loss[loss=2.266, over 5628741.96 frames. , ppl: 9.636842951012966], batch size: 70
+2022-12-13 22:03:22,301 INFO [train.py:421] (4/8) Epoch 10, batch 49800, loss[loss=2.202, over 1960.00 frames. , ppl: 9.046102791584334] tot_loss[loss=2.268, over 5558587.39 frames. , ppl: 9.658917584872015], batch size: 70
+2022-12-13 22:05:00,309 INFO [train.py:421] (4/8) Epoch 10, batch 50000, loss[loss=2.131, over 6720.00 frames. , ppl: 8.423070842216063] tot_loss[loss=2.27, over 5500896.94 frames. , ppl: 9.675577729705617], batch size: 70
+2022-12-13 22:05:00,310 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:05:01,070 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608694857576053
+2022-12-13 22:06:42,910 INFO [train.py:421] (4/8) Epoch 10, batch 50200, loss[loss=2.241, over 3360.00 frames. , ppl: 9.401904745039408] tot_loss[loss=2.27, over 5479540.04 frames. , ppl: 9.681462422435965], batch size: 70
+2022-12-13 22:08:19,153 INFO [train.py:421] (4/8) Epoch 10, batch 50400, loss[loss=2.311, over 2030.00 frames. , ppl: 10.084868361095634] tot_loss[loss=2.271, over 5454239.45 frames. , ppl: 9.689498149142334], batch size: 70
+2022-12-13 22:10:01,734 INFO [train.py:421] (4/8) Epoch 10, batch 50600, loss[loss=2.318, over 2170.00 frames. , ppl: 10.158024343828094] tot_loss[loss=2.269, over 5486925.31 frames. , ppl: 9.669688447011236], batch size: 70
+2022-12-13 22:11:42,227 INFO [train.py:421] (4/8) Epoch 10, batch 50800, loss[loss=2.468, over 1260.00 frames. , ppl: 11.794477455338123] tot_loss[loss=2.27, over 5429507.23 frames. , ppl: 9.68259912120841], batch size: 70
+2022-12-13 22:13:22,635 INFO [train.py:421] (4/8) Epoch 10, batch 51000, loss[loss=2.308, over 4340.00 frames. , ppl: 10.05374104444119] tot_loss[loss=2.27, over 5451504.39 frames. , ppl: 9.681374360333603], batch size: 70
+2022-12-13 22:13:22,635 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:13:23,385 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617215941256719
+2022-12-13 22:15:05,584 INFO [train.py:421] (4/8) Epoch 10, batch 51200, loss[loss=2.308, over 1680.00 frames. , ppl: 10.055373946239575] tot_loss[loss=2.27, over 5451723.82 frames. , ppl: 9.677052523310847], batch size: 70
+2022-12-13 22:16:46,784 INFO [train.py:421] (4/8) Epoch 10, batch 51400, loss[loss=2.511, over 1120.00 frames. , ppl: 12.311559178218975] tot_loss[loss=2.27, over 5440897.00 frames. , ppl: 9.680335791922722], batch size: 70
+2022-12-13 22:18:32,961 INFO [train.py:421] (4/8) Epoch 10, batch 51600, loss[loss=2.199, over 4690.00 frames. , ppl: 9.016989088497722] tot_loss[loss=2.269, over 5469973.49 frames. , ppl: 9.671510166108458], batch size: 70
+2022-12-13 22:20:11,668 INFO [train.py:421] (4/8) Epoch 10, batch 51800, loss[loss=2.209, over 4200.00 frames. , ppl: 9.110251738743042] tot_loss[loss=2.269, over 5488039.31 frames. , ppl: 9.672130451856948], batch size: 70
+2022-12-13 22:21:49,902 INFO [train.py:421] (4/8) Epoch 10, batch 52000, loss[loss=3.247, over 490.00 frames. , ppl: 25.72125497000506] tot_loss[loss=2.27, over 5480936.89 frames. , ppl: 9.676349900020295], batch size: 70
+2022-12-13 22:21:49,902 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:21:50,672 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599999450495101
+2022-12-13 22:23:31,973 INFO [train.py:421] (4/8) Epoch 10, batch 52200, loss[loss=2.131, over 4830.00 frames. , ppl: 8.422052063630263] tot_loss[loss=2.271, over 5469629.63 frames. , ppl: 9.68697164188161], batch size: 70
+2022-12-13 22:25:13,144 INFO [train.py:421] (4/8) Epoch 10, batch 52400, loss[loss=2.192, over 910.00 frames. , ppl: 8.948783388995755] tot_loss[loss=2.27, over 5457207.88 frames. , ppl: 9.6841394608809], batch size: 70
+2022-12-13 22:26:53,088 INFO [train.py:421] (4/8) Epoch 10, batch 52600, loss[loss=2.27, over 3430.00 frames. , ppl: 9.679997747969411] tot_loss[loss=2.271, over 5436663.87 frames. , ppl: 9.693394817161641], batch size: 70
+2022-12-13 22:28:32,244 INFO [train.py:421] (4/8) Epoch 10, batch 52800, loss[loss=2.235, over 2730.00 frames. , ppl: 9.344393476908762] tot_loss[loss=2.272, over 5406387.51 frames. , ppl: 9.703253998850345], batch size: 70
+2022-12-13 22:30:12,654 INFO [train.py:421] (4/8) Epoch 10, batch 53000, loss[loss=2.412, over 2030.00 frames. , ppl: 11.155437545360522] tot_loss[loss=2.272, over 5422664.75 frames. , ppl: 9.69845988300117], batch size: 70
+2022-12-13 22:30:12,655 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:30:13,403 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621981307382367
+2022-12-13 22:31:58,567 INFO [train.py:421] (4/8) Epoch 10, batch 53200, loss[loss=2.23, over 3220.00 frames. , ppl: 9.298852459251728] tot_loss[loss=2.272, over 5409823.58 frames. , ppl: 9.699309637132874], batch size: 70
+2022-12-13 22:33:42,494 INFO [train.py:421] (4/8) Epoch 10, batch 53400, loss[loss=2.553, over 840.00 frames. , ppl: 12.842361487505698] tot_loss[loss=2.273, over 5401362.73 frames. , ppl: 9.70623446899365], batch size: 70
+2022-12-13 22:35:23,717 INFO [train.py:421] (4/8) Epoch 10, batch 53600, loss[loss=2.301, over 2590.00 frames. , ppl: 9.984812310268445] tot_loss[loss=2.272, over 5425113.18 frames. , ppl: 9.699928316538978], batch size: 70
+2022-12-13 22:37:03,239 INFO [train.py:421] (4/8) Epoch 10, batch 53800, loss[loss=2.596, over 840.00 frames. , ppl: 13.40925621923693] tot_loss[loss=2.272, over 5424455.10 frames. , ppl: 9.702583188308981], batch size: 70
+2022-12-13 22:38:45,588 INFO [train.py:421] (4/8) Epoch 10, batch 54000, loss[loss=2.161, over 6300.00 frames. , ppl: 8.677498986300753] tot_loss[loss=2.273, over 5406066.67 frames. , ppl: 9.704083872025462], batch size: 70
+2022-12-13 22:38:45,588 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:38:46,336 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606318724465547
+2022-12-13 22:40:29,332 INFO [train.py:421] (4/8) Epoch 10, batch 54200, loss[loss=2.967, over 560.00 frames. , ppl: 19.436184166774783] tot_loss[loss=2.271, over 5437689.10 frames. , ppl: 9.69303508700382], batch size: 70
+2022-12-13 22:42:07,612 INFO [train.py:421] (4/8) Epoch 10, batch 54400, loss[loss=2.28, over 2870.00 frames. , ppl: 9.780485528261934] tot_loss[loss=2.272, over 5424312.46 frames. , ppl: 9.696528192359152], batch size: 70
+2022-12-13 22:43:49,797 INFO [train.py:421] (4/8) Epoch 10, batch 54600, loss[loss=2.295, over 2590.00 frames. , ppl: 9.919781282799622] tot_loss[loss=2.272, over 5413746.22 frames. , ppl: 9.69742390298817], batch size: 70
+2022-12-13 22:45:29,927 INFO [train.py:421] (4/8) Epoch 10, batch 54800, loss[loss=2.402, over 1820.00 frames. , ppl: 11.04487818243296] tot_loss[loss=2.273, over 5388956.85 frames. , ppl: 9.703847853932078], batch size: 70
+2022-12-13 22:47:09,472 INFO [train.py:421] (4/8) Epoch 10, batch 55000, loss[loss=2.396, over 3360.00 frames. , ppl: 10.97680125814007] tot_loss[loss=2.272, over 5388409.24 frames. , ppl: 9.698733614676572], batch size: 70
+2022-12-13 22:47:09,473 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:47:10,233 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.265, over 211138.00 frames. , ppl: 9.6300680375108
+2022-12-13 22:48:52,900 INFO [train.py:421] (4/8) Epoch 10, batch 55200, loss[loss=2.361, over 1680.00 frames. , ppl: 10.599162941606538] tot_loss[loss=2.272, over 5424685.35 frames. , ppl: 9.697107916430529], batch size: 70
+2022-12-13 22:50:33,380 INFO [train.py:421] (4/8) Epoch 10, batch 55400, loss[loss=2.187, over 3430.00 frames. , ppl: 8.911841054129171] tot_loss[loss=2.271, over 5462890.65 frames. , ppl: 9.686500615780414], batch size: 70
+2022-12-13 22:52:12,077 INFO [train.py:421] (4/8) Epoch 10, batch 55600, loss[loss=2.238, over 2380.00 frames. , ppl: 9.37647511059085] tot_loss[loss=2.27, over 5494398.63 frames. , ppl: 9.682291477182224], batch size: 70
+2022-12-13 22:53:51,906 INFO [train.py:421] (4/8) Epoch 10, batch 55800, loss[loss=2.166, over 2520.00 frames. , ppl: 8.722632363804564] tot_loss[loss=2.268, over 5547436.70 frames. , ppl: 9.663343863317731], batch size: 70
+2022-12-13 22:55:32,376 INFO [train.py:421] (4/8) Epoch 10, batch 56000, loss[loss=2.273, over 2450.00 frames. , ppl: 9.710044582802748] tot_loss[loss=2.269, over 5527407.40 frames. , ppl: 9.666007791630923], batch size: 70
+2022-12-13 22:55:32,376 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 22:55:33,125 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.602260323357212
+2022-12-13 22:57:14,363 INFO [train.py:421] (4/8) Epoch 10, batch 56200, loss[loss=2.221, over 2520.00 frames. , ppl: 9.213075107343593] tot_loss[loss=2.268, over 5539263.48 frames. , ppl: 9.662185877200438], batch size: 70
+2022-12-13 22:58:56,428 INFO [train.py:421] (4/8) Epoch 10, batch 56400, loss[loss=2.356, over 2380.00 frames. , ppl: 10.543528337361801] tot_loss[loss=2.27, over 5511907.64 frames. , ppl: 9.67544766836665], batch size: 70
+2022-12-13 23:00:37,873 INFO [train.py:421] (4/8) Epoch 10, batch 56600, loss[loss=2.465, over 1190.00 frames. , ppl: 11.758387123997917] tot_loss[loss=2.271, over 5469693.24 frames. , ppl: 9.688382410130117], batch size: 70
+2022-12-13 23:02:17,338 INFO [train.py:421] (4/8) Epoch 10, batch 56800, loss[loss=2.318, over 2100.00 frames. , ppl: 10.151822310027638] tot_loss[loss=2.271, over 5451470.33 frames. , ppl: 9.691479668148592], batch size: 70
+2022-12-13 23:03:57,926 INFO [train.py:421] (4/8) Epoch 10, batch 57000, loss[loss=2.498, over 980.00 frames. , ppl: 12.15363253395977] tot_loss[loss=2.27, over 5486568.15 frames. , ppl: 9.676300329485983], batch size: 70
+2022-12-13 23:03:57,927 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:03:58,673 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.264, over 211138.00 frames. , ppl: 9.62573033470701
+2022-12-13 23:05:40,807 INFO [train.py:421] (4/8) Epoch 10, batch 57200, loss[loss=2.384, over 1260.00 frames. , ppl: 10.850049170886704] tot_loss[loss=2.268, over 5527438.27 frames. , ppl: 9.66021973964029], batch size: 70
+2022-12-13 23:07:18,394 INFO [train.py:421] (4/8) Epoch 10, batch 57400, loss[loss=2.153, over 2730.00 frames. , ppl: 8.607679615756892] tot_loss[loss=2.268, over 5536963.09 frames. , ppl: 9.65769913884229], batch size: 70
+2022-12-13 23:08:53,700 INFO [train.py:421] (4/8) Epoch 10, batch 57600, loss[loss=2.334, over 1960.00 frames. , ppl: 10.314808142301422] tot_loss[loss=2.269, over 5507787.31 frames. , ppl: 9.67281862301055], batch size: 70
+2022-12-13 23:10:36,335 INFO [train.py:421] (4/8) Epoch 10, batch 57800, loss[loss=2.16, over 4060.00 frames. , ppl: 8.66693100496572] tot_loss[loss=2.27, over 5510912.12 frames. , ppl: 9.676252181421658], batch size: 70
+2022-12-13 23:12:20,210 INFO [train.py:421] (4/8) Epoch 10, batch 58000, loss[loss=2.513, over 1120.00 frames. , ppl: 12.338858546079766] tot_loss[loss=2.269, over 5527744.99 frames. , ppl: 9.668149609267854], batch size: 70
+2022-12-13 23:12:20,211 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:12:20,972 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.609582325499408
+2022-12-13 23:14:05,369 INFO [train.py:421] (4/8) Epoch 10, batch 58200, loss[loss=2.36, over 1400.00 frames. , ppl: 10.593990064019613] tot_loss[loss=2.269, over 5560103.89 frames. , ppl: 9.667499809008858], batch size: 70
+2022-12-13 23:15:40,222 INFO [train.py:421] (4/8) Epoch 10, batch 58400, loss[loss=2.629, over 1050.00 frames. , ppl: 13.853113996350483] tot_loss[loss=2.268, over 5566314.67 frames. , ppl: 9.663687320109645], batch size: 70
+2022-12-13 23:17:18,338 INFO [train.py:421] (4/8) Epoch 10, batch 58600, loss[loss=2.636, over 700.00 frames. , ppl: 13.951586950905545] tot_loss[loss=2.268, over 5589469.03 frames. , ppl: 9.66480379917301], batch size: 70
+2022-12-13 23:18:57,519 INFO [train.py:421] (4/8) Epoch 10, batch 58800, loss[loss=2.754, over 700.00 frames. , ppl: 15.713141792816335] tot_loss[loss=2.268, over 5589888.57 frames. , ppl: 9.658307817874524], batch size: 70
+2022-12-13 23:20:40,954 INFO [train.py:421] (4/8) Epoch 10, batch 59000, loss[loss=2.309, over 3220.00 frames. , ppl: 10.060117950114252] tot_loss[loss=2.268, over 5547038.54 frames. , ppl: 9.664115072534148], batch size: 70
+2022-12-13 23:20:40,955 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:20:41,703 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606206402379026
+2022-12-13 23:22:20,249 INFO [train.py:421] (4/8) Epoch 10, batch 59200, loss[loss=2.127, over 12180.00 frames. , ppl: 8.391952355950556] tot_loss[loss=2.267, over 5599221.95 frames. , ppl: 9.648953556300468], batch size: 70
+2022-12-13 23:24:00,410 INFO [train.py:421] (4/8) Epoch 10, batch 59400, loss[loss=2.283, over 2310.00 frames. , ppl: 9.801704615194426] tot_loss[loss=2.266, over 5617011.70 frames. , ppl: 9.64248655984706], batch size: 70
+2022-12-13 23:25:41,349 INFO [train.py:421] (4/8) Epoch 10, batch 59600, loss[loss=2.18, over 3220.00 frames. , ppl: 8.842930728654474] tot_loss[loss=2.267, over 5585115.85 frames. , ppl: 9.648738605278874], batch size: 70
+2022-12-13 23:27:19,478 INFO [train.py:421] (4/8) Epoch 10, batch 59800, loss[loss=4.156, over 350.00 frames. , ppl: 63.80770961671402] tot_loss[loss=2.267, over 5570641.51 frames. , ppl: 9.650537033068021], batch size: 70
+2022-12-13 23:28:53,373 INFO [train.py:421] (4/8) Epoch 10, batch 60000, loss[loss=2.35, over 2450.00 frames. , ppl: 10.4841592329561] tot_loss[loss=2.266, over 5599129.99 frames. , ppl: 9.639257544183328], batch size: 70
+2022-12-13 23:28:53,373 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:28:54,130 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612700490778435
+2022-12-13 23:30:35,175 INFO [train.py:421] (4/8) Epoch 10, batch 60200, loss[loss=2.235, over 3360.00 frames. , ppl: 9.345338549323158] tot_loss[loss=2.266, over 5619957.51 frames. , ppl: 9.644831604661679], batch size: 70
+2022-12-13 23:32:13,575 INFO [train.py:421] (4/8) Epoch 10, batch 60400, loss[loss=2.903, over 560.00 frames. , ppl: 18.221898329936398] tot_loss[loss=2.267, over 5618870.48 frames. , ppl: 9.648602662047036], batch size: 70
+2022-12-13 23:33:56,148 INFO [train.py:421] (4/8) Epoch 10, batch 60600, loss[loss=2.343, over 1890.00 frames. , ppl: 10.414076273063095] tot_loss[loss=2.268, over 5587506.24 frames. , ppl: 9.659085430052503], batch size: 70
+2022-12-13 23:35:35,542 INFO [train.py:421] (4/8) Epoch 10, batch 60800, loss[loss=2.339, over 1890.00 frames. , ppl: 10.369226050153738] tot_loss[loss=2.269, over 5554554.03 frames. , ppl: 9.670606729206929], batch size: 70
+2022-12-13 23:37:16,235 INFO [train.py:421] (4/8) Epoch 10, batch 61000, loss[loss=2.459, over 1260.00 frames. , ppl: 11.695077909072682] tot_loss[loss=2.269, over 5561678.17 frames. , ppl: 9.672656726913745], batch size: 70
+2022-12-13 23:37:16,236 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:37:16,983 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60258010007721
+2022-12-13 23:38:59,952 INFO [train.py:421] (4/8) Epoch 10, batch 61200, loss[loss=2.381, over 1680.00 frames. , ppl: 10.8118478379488] tot_loss[loss=2.269, over 5564189.63 frames. , ppl: 9.672663314468917], batch size: 70
+2022-12-13 23:40:40,175 INFO [train.py:421] (4/8) Epoch 10, batch 61400, loss[loss=2.23, over 6370.00 frames. , ppl: 9.302241078679822] tot_loss[loss=2.269, over 5560995.37 frames. , ppl: 9.668799916367592], batch size: 70
+2022-12-13 23:42:19,009 INFO [train.py:421] (4/8) Epoch 10, batch 61600, loss[loss=2.359, over 2030.00 frames. , ppl: 10.578230921279802] tot_loss[loss=2.269, over 5543165.99 frames. , ppl: 9.667912028528562], batch size: 70
+2022-12-13 23:44:03,589 INFO [train.py:421] (4/8) Epoch 10, batch 61800, loss[loss=2.187, over 4900.00 frames. , ppl: 8.908051374245588] tot_loss[loss=2.267, over 5604589.86 frames. , ppl: 9.654984294505567], batch size: 70
+2022-12-13 23:45:42,346 INFO [train.py:421] (4/8) Epoch 10, batch 62000, loss[loss=2.216, over 5600.00 frames. , ppl: 9.17202987009864] tot_loss[loss=2.268, over 5579181.78 frames. , ppl: 9.659392873828441], batch size: 70
+2022-12-13 23:45:42,347 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:45:43,104 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61007587277309
+2022-12-13 23:47:24,544 INFO [train.py:421] (4/8) Epoch 10, batch 62200, loss[loss=2.389, over 2170.00 frames. , ppl: 10.90038743494002] tot_loss[loss=2.268, over 5588154.49 frames. , ppl: 9.657684900576932], batch size: 70
+2022-12-13 23:49:09,220 INFO [train.py:421] (4/8) Epoch 10, batch 62400, loss[loss=2.461, over 980.00 frames. , ppl: 11.719911838384936] tot_loss[loss=2.267, over 5578154.55 frames. , ppl: 9.651622699381576], batch size: 70
+2022-12-13 23:50:45,969 INFO [train.py:421] (4/8) Epoch 10, batch 62600, loss[loss=2.243, over 2800.00 frames. , ppl: 9.417193782423581] tot_loss[loss=2.267, over 5584085.30 frames. , ppl: 9.647711435937085], batch size: 70
+2022-12-13 23:52:24,484 INFO [train.py:421] (4/8) Epoch 10, batch 62800, loss[loss=2.457, over 1190.00 frames. , ppl: 11.671479207997768] tot_loss[loss=2.266, over 5585916.85 frames. , ppl: 9.64318435141344], batch size: 70
+2022-12-13 23:54:03,448 INFO [train.py:421] (4/8) Epoch 10, batch 63000, loss[loss=2.182, over 4200.00 frames. , ppl: 8.864507935427016] tot_loss[loss=2.268, over 5539537.29 frames. , ppl: 9.65744698577962], batch size: 70
+2022-12-13 23:54:03,448 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-13 23:54:04,195 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606716838640049
+2022-12-13 23:55:42,811 INFO [train.py:421] (4/8) Epoch 10, batch 63200, loss[loss=2.257, over 2520.00 frames. , ppl: 9.55480189693741] tot_loss[loss=2.268, over 5549890.85 frames. , ppl: 9.65612423577935], batch size: 70
+2022-12-13 23:57:21,549 INFO [train.py:421] (4/8) Epoch 10, batch 63400, loss[loss=2.573, over 840.00 frames. , ppl: 13.103470766678612] tot_loss[loss=2.268, over 5536456.89 frames. , ppl: 9.657521118222913], batch size: 70
+2022-12-13 23:59:00,619 INFO [train.py:421] (4/8) Epoch 10, batch 63600, loss[loss=2.314, over 2100.00 frames. , ppl: 10.116846543521484] tot_loss[loss=2.267, over 5545638.68 frames. , ppl: 9.654597091323675], batch size: 70
+2022-12-14 00:00:37,649 INFO [train.py:421] (4/8) Epoch 10, batch 63800, loss[loss=2.16, over 5180.00 frames. , ppl: 8.673983865637812] tot_loss[loss=2.268, over 5544951.29 frames. , ppl: 9.658007071594627], batch size: 70
+2022-12-14 00:02:18,824 INFO [train.py:421] (4/8) Epoch 10, batch 64000, loss[loss=2.453, over 1050.00 frames. , ppl: 11.623433137079846] tot_loss[loss=2.268, over 5540841.04 frames. , ppl: 9.657627060263028], batch size: 70
+2022-12-14 00:02:18,825 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:02:19,585 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616598196927988
+2022-12-14 00:03:57,270 INFO [train.py:421] (4/8) Epoch 10, batch 64200, loss[loss=2.353, over 2590.00 frames. , ppl: 10.511882237950529] tot_loss[loss=2.268, over 5509630.72 frames. , ppl: 9.664538988003395], batch size: 70
+2022-12-14 00:05:36,743 INFO [train.py:421] (4/8) Epoch 10, batch 64400, loss[loss=2.392, over 2100.00 frames. , ppl: 10.939331360473279] tot_loss[loss=2.268, over 5514439.05 frames. , ppl: 9.659350081773606], batch size: 70
+2022-12-14 00:07:12,118 INFO [train.py:421] (4/8) Epoch 10, batch 64600, loss[loss=2.462, over 1120.00 frames. , ppl: 11.725672363076374] tot_loss[loss=2.268, over 5513555.01 frames. , ppl: 9.661581371388017], batch size: 70
+2022-12-14 00:08:51,349 INFO [train.py:421] (4/8) Epoch 10, batch 64800, loss[loss=2.483, over 1470.00 frames. , ppl: 11.97419247699176] tot_loss[loss=2.267, over 5576359.04 frames. , ppl: 9.653284004129215], batch size: 70
+2022-12-14 00:10:28,813 INFO [train.py:421] (4/8) Epoch 10, batch 65000, loss[loss=2.527, over 910.00 frames. , ppl: 12.516584654200413] tot_loss[loss=2.268, over 5572254.93 frames. , ppl: 9.656507841769402], batch size: 70
+2022-12-14 00:10:28,814 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:10:29,560 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599982400051333
+2022-12-14 00:12:08,438 INFO [train.py:421] (4/8) Epoch 10, batch 65200, loss[loss=2.552, over 700.00 frames. , ppl: 12.838807844973743] tot_loss[loss=2.268, over 5552851.75 frames. , ppl: 9.658675312964464], batch size: 70
+2022-12-14 00:13:46,654 INFO [train.py:421] (4/8) Epoch 10, batch 65400, loss[loss=2.327, over 2380.00 frames. , ppl: 10.247499790520106] tot_loss[loss=2.268, over 5537341.34 frames. , ppl: 9.660800905490296], batch size: 70
+2022-12-14 00:15:25,095 INFO [train.py:421] (4/8) Epoch 10, batch 65600, loss[loss=2.165, over 3920.00 frames. , ppl: 8.71520070157405] tot_loss[loss=2.269, over 5527854.15 frames. , ppl: 9.669228787087631], batch size: 70
+2022-12-14 00:17:06,085 INFO [train.py:421] (4/8) Epoch 10, batch 65800, loss[loss=2.294, over 2940.00 frames. , ppl: 9.917936702178926] tot_loss[loss=2.268, over 5556988.86 frames. , ppl: 9.656262748858184], batch size: 70
+2022-12-14 00:18:46,934 INFO [train.py:421] (4/8) Epoch 10, batch 66000, loss[loss=2.346, over 2030.00 frames. , ppl: 10.438617414125066] tot_loss[loss=2.268, over 5549480.90 frames. , ppl: 9.658637932696164], batch size: 70
+2022-12-14 00:18:46,935 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:18:47,681 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599644239172868
+2022-12-14 00:20:26,482 INFO [train.py:421] (4/8) Epoch 10, batch 66200, loss[loss=2.262, over 1400.00 frames. , ppl: 9.60610186267773] tot_loss[loss=2.269, over 5527376.50 frames. , ppl: 9.665359521465412], batch size: 70
+2022-12-14 00:22:05,234 INFO [train.py:421] (4/8) Epoch 10, batch 66400, loss[loss=2.147, over 4620.00 frames. , ppl: 8.555634430140083] tot_loss[loss=2.268, over 5537386.78 frames. , ppl: 9.662887632268554], batch size: 70
+2022-12-14 00:23:48,637 INFO [train.py:421] (4/8) Epoch 10, batch 66600, loss[loss=2.173, over 4410.00 frames. , ppl: 8.783864380099033] tot_loss[loss=2.268, over 5528356.21 frames. , ppl: 9.662045202051663], batch size: 70
+2022-12-14 00:25:27,836 INFO [train.py:421] (4/8) Epoch 10, batch 66800, loss[loss=2.375, over 1610.00 frames. , ppl: 10.754399821609864] tot_loss[loss=2.269, over 5509622.95 frames. , ppl: 9.668236784447139], batch size: 70
+2022-12-14 00:27:09,560 INFO [train.py:421] (4/8) Epoch 10, batch 67000, loss[loss=2.219, over 3290.00 frames. , ppl: 9.201428955262484] tot_loss[loss=2.268, over 5544471.50 frames. , ppl: 9.659447078837772], batch size: 70
+2022-12-14 00:27:09,561 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:27:10,307 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608690591101315
+2022-12-14 00:28:53,508 INFO [train.py:421] (4/8) Epoch 10, batch 67200, loss[loss=2.725, over 700.00 frames. , ppl: 15.263742724100469] tot_loss[loss=2.267, over 5574089.74 frames. , ppl: 9.651899252276737], batch size: 70
+2022-12-14 00:30:34,765 INFO [train.py:421] (4/8) Epoch 10, batch 67400, loss[loss=2.184, over 6300.00 frames. , ppl: 8.880629621719107] tot_loss[loss=2.268, over 5557912.55 frames. , ppl: 9.65998539298501], batch size: 70
+2022-12-14 00:32:15,546 INFO [train.py:421] (4/8) Epoch 10, batch 67600, loss[loss=2.44, over 1470.00 frames. , ppl: 11.467705154071355] tot_loss[loss=2.269, over 5522698.25 frames. , ppl: 9.673670517304696], batch size: 70
+2022-12-14 00:33:55,294 INFO [train.py:421] (4/8) Epoch 10, batch 67800, loss[loss=2.361, over 3290.00 frames. , ppl: 10.600022982360818] tot_loss[loss=2.271, over 5478541.99 frames. , ppl: 9.687418102574854], batch size: 70
+2022-12-14 00:35:36,607 INFO [train.py:421] (4/8) Epoch 10, batch 68000, loss[loss=2.397, over 1120.00 frames. , ppl: 10.993847690082513] tot_loss[loss=2.27, over 5473322.87 frames. , ppl: 9.682774205482188], batch size: 70
+2022-12-14 00:35:36,608 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:35:37,368 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606830588577454
+2022-12-14 00:37:17,019 INFO [train.py:421] (4/8) Epoch 10, batch 68200, loss[loss=2.182, over 4060.00 frames. , ppl: 8.864323430674794] tot_loss[loss=2.27, over 5473670.04 frames. , ppl: 9.67793279904354], batch size: 70
+2022-12-14 00:38:58,133 INFO [train.py:421] (4/8) Epoch 10, batch 68400, loss[loss=2.745, over 630.00 frames. , ppl: 15.571753452063962] tot_loss[loss=2.27, over 5462637.49 frames. , ppl: 9.677998274382999], batch size: 70
+2022-12-14 00:40:37,840 INFO [train.py:421] (4/8) Epoch 10, batch 68600, loss[loss=2.385, over 2100.00 frames. , ppl: 10.861512087983785] tot_loss[loss=2.269, over 5499438.88 frames. , ppl: 9.66610059240965], batch size: 70
+2022-12-14 00:42:17,523 INFO [train.py:421] (4/8) Epoch 10, batch 68800, loss[loss=2.333, over 2660.00 frames. , ppl: 10.307320311739778] tot_loss[loss=2.27, over 5445884.40 frames. , ppl: 9.683405944791264], batch size: 70
+2022-12-14 00:44:01,099 INFO [train.py:421] (4/8) Epoch 10, batch 69000, loss[loss=2.23, over 6790.00 frames. , ppl: 9.298380458092842] tot_loss[loss=2.271, over 5469668.82 frames. , ppl: 9.692125490170783], batch size: 70
+2022-12-14 00:44:01,099 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:44:01,845 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60023105535575
+2022-12-14 00:45:40,104 INFO [train.py:421] (4/8) Epoch 10, batch 69200, loss[loss=2.37, over 2100.00 frames. , ppl: 10.695434955824357] tot_loss[loss=2.271, over 5494972.42 frames. , ppl: 9.685493040706561], batch size: 70
+2022-12-14 00:47:14,230 INFO [train.py:421] (4/8) Epoch 10, batch 69400, loss[loss=2.242, over 2800.00 frames. , ppl: 9.413937806770575] tot_loss[loss=2.271, over 5459181.16 frames. , ppl: 9.691825705813159], batch size: 70
+2022-12-14 00:48:53,975 INFO [train.py:421] (4/8) Epoch 10, batch 69600, loss[loss=2.669, over 910.00 frames. , ppl: 14.428300630349304] tot_loss[loss=2.271, over 5464266.12 frames. , ppl: 9.6908478938502], batch size: 70
+2022-12-14 00:50:33,923 INFO [train.py:421] (4/8) Epoch 10, batch 69800, loss[loss=2.22, over 2450.00 frames. , ppl: 9.210414193709406] tot_loss[loss=2.27, over 5502820.63 frames. , ppl: 9.681508820526652], batch size: 70
+2022-12-14 00:52:14,676 INFO [train.py:421] (4/8) Epoch 10, batch 70000, loss[loss=2.318, over 2870.00 frames. , ppl: 10.151842287512299] tot_loss[loss=2.269, over 5532818.79 frames. , ppl: 9.66784774294999], batch size: 70
+2022-12-14 00:52:14,676 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 00:52:15,422 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.262, over 211138.00 frames. , ppl: 9.601824022965605
+2022-12-14 00:53:54,452 INFO [train.py:421] (4/8) Epoch 10, batch 70200, loss[loss=2.274, over 1890.00 frames. , ppl: 9.717715722081223] tot_loss[loss=2.27, over 5509512.55 frames. , ppl: 9.680676102074717], batch size: 70
+2022-12-14 00:55:36,203 INFO [train.py:421] (4/8) Epoch 10, batch 70400, loss[loss=2.246, over 4480.00 frames. , ppl: 9.45006685145532] tot_loss[loss=2.271, over 5477121.44 frames. , ppl: 9.68561440540147], batch size: 70
+2022-12-14 00:57:16,764 INFO [train.py:421] (4/8) Epoch 10, batch 70600, loss[loss=2.143, over 4550.00 frames. , ppl: 8.521727846472741] tot_loss[loss=2.27, over 5479708.21 frames. , ppl: 9.683127529437982], batch size: 70
+2022-12-14 00:58:55,099 INFO [train.py:421] (4/8) Epoch 10, batch 70800, loss[loss=2.149, over 8050.00 frames. , ppl: 8.577841911248022] tot_loss[loss=2.269, over 5494045.11 frames. , ppl: 9.674561044887389], batch size: 70
+2022-12-14 01:00:36,989 INFO [train.py:421] (4/8) Epoch 10, batch 71000, loss[loss=2.138, over 8540.00 frames. , ppl: 8.486153653634647] tot_loss[loss=2.268, over 5526655.20 frames. , ppl: 9.662354982854556], batch size: 70
+2022-12-14 01:00:36,989 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:00:37,735 INFO [train.py:452] (4/8) Epoch 10, validation: loss=2.263, over 211138.00 frames. , ppl: 9.609851142052824
+2022-12-14 01:02:17,478 INFO [train.py:421] (4/8) Epoch 10, batch 71200, loss[loss=2.748, over 700.00 frames. , ppl: 15.6139039395607] tot_loss[loss=2.268, over 5546364.17 frames. , ppl: 9.655390710976096], batch size: 70
+2022-12-14 01:03:55,216 INFO [train.py:421] (4/8) Epoch 10, batch 71400, loss[loss=2.165, over 4480.00 frames. , ppl: 8.718845208798978] tot_loss[loss=2.268, over 5544856.80 frames. , ppl: 9.658560051786118], batch size: 70
+2022-12-14 01:05:31,328 INFO [train.py:421] (4/8) Epoch 10, batch 71600, loss[loss=2.473, over 1120.00 frames. , ppl: 11.85695175446224] tot_loss[loss=2.269, over 5505191.33 frames. , ppl: 9.667300479939295], batch size: 70
+2022-12-14 01:07:12,759 INFO [train.py:421] (4/8) Epoch 10, batch 71800, loss[loss=2.312, over 1400.00 frames. , ppl: 10.096877460577968] tot_loss[loss=2.268, over 5529582.50 frames. , ppl: 9.65738831224233], batch size: 70
+2022-12-14 01:08:27,258 INFO [train.py:421] (4/8) Epoch 11, batch 0, loss[loss=4.11, over 350.00 frames. , ppl: 60.94950223842665] tot_loss[loss=4.11, over 350.00 frames. , ppl: 60.94950223842665], batch size: 70
+2022-12-14 01:10:06,170 INFO [train.py:421] (4/8) Epoch 11, batch 200, loss[loss=2.283, over 2660.00 frames. , ppl: 9.80733361132935] tot_loss[loss=2.262, over 526789.47 frames. , ppl: 9.599149859324104], batch size: 70
+2022-12-14 01:11:46,146 INFO [train.py:421] (4/8) Epoch 11, batch 400, loss[loss=2.286, over 2170.00 frames. , ppl: 9.837761728816153] tot_loss[loss=2.263, over 984669.98 frames. , ppl: 9.609488488240181], batch size: 70
+2022-12-14 01:13:23,774 INFO [train.py:421] (4/8) Epoch 11, batch 600, loss[loss=2.504, over 1400.00 frames. , ppl: 12.230514009578027] tot_loss[loss=2.266, over 1393802.42 frames. , ppl: 9.64233838242519], batch size: 70
+2022-12-14 01:15:02,879 INFO [train.py:421] (4/8) Epoch 11, batch 800, loss[loss=2.253, over 2800.00 frames. , ppl: 9.51255309416403] tot_loss[loss=2.267, over 1754053.47 frames. , ppl: 9.651992325480863], batch size: 70
+2022-12-14 01:16:43,492 INFO [train.py:421] (4/8) Epoch 11, batch 1000, loss[loss=2.233, over 4130.00 frames. , ppl: 9.328714208762783] tot_loss[loss=2.267, over 2091291.03 frames. , ppl: 9.647131508454176], batch size: 70
+2022-12-14 01:16:43,493 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:16:44,237 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.619912276163067
+2022-12-14 01:18:27,828 INFO [train.py:421] (4/8) Epoch 11, batch 1200, loss[loss=2.28, over 2800.00 frames. , ppl: 9.775616598727328] tot_loss[loss=2.264, over 2425153.31 frames. , ppl: 9.62485907985876], batch size: 70
+2022-12-14 01:20:08,975 INFO [train.py:421] (4/8) Epoch 11, batch 1400, loss[loss=2.241, over 3920.00 frames. , ppl: 9.402967592667524] tot_loss[loss=2.261, over 2750014.78 frames. , ppl: 9.595284931220094], batch size: 70
+2022-12-14 01:21:47,194 INFO [train.py:421] (4/8) Epoch 11, batch 1600, loss[loss=2.488, over 840.00 frames. , ppl: 12.040304645867888] tot_loss[loss=2.262, over 2973353.21 frames. , ppl: 9.606542811102607], batch size: 70
+2022-12-14 01:23:27,106 INFO [train.py:421] (4/8) Epoch 11, batch 1800, loss[loss=2.243, over 4550.00 frames. , ppl: 9.420248190608646] tot_loss[loss=2.263, over 3191260.60 frames. , ppl: 9.61617273146173], batch size: 70
+2022-12-14 01:25:04,643 INFO [train.py:421] (4/8) Epoch 11, batch 2000, loss[loss=2.229, over 3990.00 frames. , ppl: 9.291050574112463] tot_loss[loss=2.264, over 3404118.81 frames. , ppl: 9.61897067347717], batch size: 70
+2022-12-14 01:25:04,643 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:25:05,404 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.623191889826966
+2022-12-14 01:26:47,629 INFO [train.py:421] (4/8) Epoch 11, batch 2200, loss[loss=2.26, over 2100.00 frames. , ppl: 9.582304868645808] tot_loss[loss=2.261, over 3655076.64 frames. , ppl: 9.593088926034707], batch size: 70
+2022-12-14 01:28:26,725 INFO [train.py:421] (4/8) Epoch 11, batch 2400, loss[loss=2.276, over 2240.00 frames. , ppl: 9.732837737049516] tot_loss[loss=2.262, over 3836800.41 frames. , ppl: 9.600912855153666], batch size: 70
+2022-12-14 01:30:12,132 INFO [train.py:421] (4/8) Epoch 11, batch 2600, loss[loss=2.358, over 1120.00 frames. , ppl: 10.574837382129756] tot_loss[loss=2.258, over 4071271.39 frames. , ppl: 9.562806792388471], batch size: 70
+2022-12-14 01:31:49,623 INFO [train.py:421] (4/8) Epoch 11, batch 2800, loss[loss=2.142, over 3570.00 frames. , ppl: 8.517067259366637] tot_loss[loss=2.258, over 4210455.98 frames. , ppl: 9.56028699204621], batch size: 70
+2022-12-14 01:33:29,897 INFO [train.py:421] (4/8) Epoch 11, batch 3000, loss[loss=2.333, over 1610.00 frames. , ppl: 10.309430582448512] tot_loss[loss=2.258, over 4335263.21 frames. , ppl: 9.568701286812285], batch size: 70
+2022-12-14 01:33:29,898 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:33:30,660 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612494194035516
+2022-12-14 01:35:15,645 INFO [train.py:421] (4/8) Epoch 11, batch 3200, loss[loss=2.109, over 5390.00 frames. , ppl: 8.240620874542678] tot_loss[loss=2.26, over 4401819.25 frames. , ppl: 9.584133082226883], batch size: 70
+2022-12-14 01:36:56,915 INFO [train.py:421] (4/8) Epoch 11, batch 3400, loss[loss=2.162, over 4970.00 frames. , ppl: 8.687892339828233] tot_loss[loss=2.262, over 4478795.01 frames. , ppl: 9.600588181697926], batch size: 70
+2022-12-14 01:38:36,970 INFO [train.py:421] (4/8) Epoch 11, batch 3600, loss[loss=2.364, over 2800.00 frames. , ppl: 10.629774413280467] tot_loss[loss=2.261, over 4586634.11 frames. , ppl: 9.593230870563835], batch size: 70
+2022-12-14 01:40:18,033 INFO [train.py:421] (4/8) Epoch 11, batch 3800, loss[loss=3.151, over 490.00 frames. , ppl: 23.35911506597411] tot_loss[loss=2.261, over 4696531.72 frames. , ppl: 9.594177095667431], batch size: 70
+2022-12-14 01:42:00,374 INFO [train.py:421] (4/8) Epoch 11, batch 4000, loss[loss=2.267, over 2380.00 frames. , ppl: 9.654073399824956] tot_loss[loss=2.261, over 4786372.77 frames. , ppl: 9.592414500237215], batch size: 70
+2022-12-14 01:42:00,374 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:42:01,137 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606853338726559
+2022-12-14 01:43:42,447 INFO [train.py:421] (4/8) Epoch 11, batch 4200, loss[loss=2.25, over 2170.00 frames. , ppl: 9.491931797608023] tot_loss[loss=2.262, over 4845963.62 frames. , ppl: 9.598620068842182], batch size: 70
+2022-12-14 01:45:19,904 INFO [train.py:421] (4/8) Epoch 11, batch 4400, loss[loss=2.284, over 2940.00 frames. , ppl: 9.810995372578247] tot_loss[loss=2.26, over 4938443.65 frames. , ppl: 9.58750033279163], batch size: 70
+2022-12-14 01:46:59,761 INFO [train.py:421] (4/8) Epoch 11, batch 4600, loss[loss=2.233, over 6230.00 frames. , ppl: 9.32802665995067] tot_loss[loss=2.261, over 4961733.32 frames. , ppl: 9.596850855337957], batch size: 70
+2022-12-14 01:48:42,457 INFO [train.py:421] (4/8) Epoch 11, batch 4800, loss[loss=2.296, over 3640.00 frames. , ppl: 9.931379282750418] tot_loss[loss=2.261, over 5037019.21 frames. , ppl: 9.595862252890125], batch size: 70
+2022-12-14 01:50:23,278 INFO [train.py:421] (4/8) Epoch 11, batch 5000, loss[loss=2.161, over 3430.00 frames. , ppl: 8.680466753367687] tot_loss[loss=2.262, over 5061655.49 frames. , ppl: 9.599616093709221], batch size: 70
+2022-12-14 01:50:23,278 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:50:24,027 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.615150780970064
+2022-12-14 01:52:05,450 INFO [train.py:421] (4/8) Epoch 11, batch 5200, loss[loss=2.387, over 1540.00 frames. , ppl: 10.885578135383941] tot_loss[loss=2.26, over 5137504.25 frames. , ppl: 9.58750971714778], batch size: 70
+2022-12-14 01:53:46,219 INFO [train.py:421] (4/8) Epoch 11, batch 5400, loss[loss=2.193, over 7070.00 frames. , ppl: 8.962967376269752] tot_loss[loss=2.262, over 5175735.59 frames. , ppl: 9.599250987267546], batch size: 70
+2022-12-14 01:55:25,271 INFO [train.py:421] (4/8) Epoch 11, batch 5600, loss[loss=2.198, over 2520.00 frames. , ppl: 9.00400801647699] tot_loss[loss=2.262, over 5215988.05 frames. , ppl: 9.603408524840356], batch size: 70
+2022-12-14 01:57:09,546 INFO [train.py:421] (4/8) Epoch 11, batch 5800, loss[loss=2.326, over 1190.00 frames. , ppl: 10.24188587453096] tot_loss[loss=2.261, over 5267881.46 frames. , ppl: 9.595157020435051], batch size: 70
+2022-12-14 01:58:49,576 INFO [train.py:421] (4/8) Epoch 11, batch 6000, loss[loss=2.533, over 840.00 frames. , ppl: 12.593634643150999] tot_loss[loss=2.261, over 5319700.71 frames. , ppl: 9.589090010502986], batch size: 70
+2022-12-14 01:58:49,577 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 01:58:50,322 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606113986457235
+2022-12-14 02:00:27,433 INFO [train.py:421] (4/8) Epoch 11, batch 6200, loss[loss=2.275, over 4830.00 frames. , ppl: 9.725730146643857] tot_loss[loss=2.261, over 5338227.84 frames. , ppl: 9.597349220665832], batch size: 70
+2022-12-14 02:02:09,172 INFO [train.py:421] (4/8) Epoch 11, batch 6400, loss[loss=2.116, over 6930.00 frames. , ppl: 8.297001243781109] tot_loss[loss=2.26, over 5389616.11 frames. , ppl: 9.586788370196441], batch size: 70
+2022-12-14 02:03:48,306 INFO [train.py:421] (4/8) Epoch 11, batch 6600, loss[loss=2.339, over 3360.00 frames. , ppl: 10.36652708676494] tot_loss[loss=2.259, over 5445289.65 frames. , ppl: 9.572822038818023], batch size: 70
+2022-12-14 02:05:25,882 INFO [train.py:421] (4/8) Epoch 11, batch 6800, loss[loss=2.326, over 1120.00 frames. , ppl: 10.238193895904889] tot_loss[loss=2.258, over 5502934.56 frames. , ppl: 9.562868206716193], batch size: 70
+2022-12-14 02:07:09,116 INFO [train.py:421] (4/8) Epoch 11, batch 7000, loss[loss=2.319, over 1610.00 frames. , ppl: 10.166298175569333] tot_loss[loss=2.259, over 5474905.24 frames. , ppl: 9.572300430970389], batch size: 70
+2022-12-14 02:07:09,116 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:07:09,879 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.616433092260849
+2022-12-14 02:08:51,379 INFO [train.py:421] (4/8) Epoch 11, batch 7200, loss[loss=2.515, over 1260.00 frames. , ppl: 12.371779866135107] tot_loss[loss=2.259, over 5505055.17 frames. , ppl: 9.572337669659696], batch size: 70
+2022-12-14 02:10:28,062 INFO [train.py:421] (4/8) Epoch 11, batch 7400, loss[loss=2.18, over 4200.00 frames. , ppl: 8.846365908959541] tot_loss[loss=2.26, over 5490776.11 frames. , ppl: 9.584175587016391], batch size: 70
+2022-12-14 02:12:08,240 INFO [train.py:421] (4/8) Epoch 11, batch 7600, loss[loss=2.121, over 6930.00 frames. , ppl: 8.336494564859581] tot_loss[loss=2.259, over 5507961.51 frames. , ppl: 9.576933626364143], batch size: 70
+2022-12-14 02:13:46,076 INFO [train.py:421] (4/8) Epoch 11, batch 7800, loss[loss=2.252, over 3430.00 frames. , ppl: 9.506399707873403] tot_loss[loss=2.262, over 5453076.63 frames. , ppl: 9.599508613961966], batch size: 70
+2022-12-14 02:15:21,878 INFO [train.py:421] (4/8) Epoch 11, batch 8000, loss[loss=2.492, over 840.00 frames. , ppl: 12.082069212195558] tot_loss[loss=2.262, over 5468842.85 frames. , ppl: 9.603927581847332], batch size: 70
+2022-12-14 02:15:21,878 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:15:22,633 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.595916733986652
+2022-12-14 02:17:02,599 INFO [train.py:421] (4/8) Epoch 11, batch 8200, loss[loss=2.706, over 770.00 frames. , ppl: 14.973073496163543] tot_loss[loss=2.263, over 5448754.81 frames. , ppl: 9.609906766850202], batch size: 70
+2022-12-14 02:18:44,665 INFO [train.py:421] (4/8) Epoch 11, batch 8400, loss[loss=2.201, over 4690.00 frames. , ppl: 9.037518894601336] tot_loss[loss=2.262, over 5469888.87 frames. , ppl: 9.606556322905927], batch size: 70
+2022-12-14 02:20:24,716 INFO [train.py:421] (4/8) Epoch 11, batch 8600, loss[loss=2.23, over 3430.00 frames. , ppl: 9.303988004311364] tot_loss[loss=2.262, over 5473468.17 frames. , ppl: 9.606391621756579], batch size: 70
+2022-12-14 02:22:03,174 INFO [train.py:421] (4/8) Epoch 11, batch 8800, loss[loss=2.347, over 1820.00 frames. , ppl: 10.45724320756466] tot_loss[loss=2.263, over 5472560.64 frames. , ppl: 9.61186510275738], batch size: 70
+2022-12-14 02:23:44,977 INFO [train.py:421] (4/8) Epoch 11, batch 9000, loss[loss=2.208, over 3990.00 frames. , ppl: 9.095414655948554] tot_loss[loss=2.262, over 5497301.71 frames. , ppl: 9.605066229044326], batch size: 70
+2022-12-14 02:23:44,978 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:23:45,723 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.615415483833607
+2022-12-14 02:25:22,447 INFO [train.py:421] (4/8) Epoch 11, batch 9200, loss[loss=2.324, over 1890.00 frames. , ppl: 10.216019643892862] tot_loss[loss=2.261, over 5516604.03 frames. , ppl: 9.591634778884256], batch size: 70
+2022-12-14 02:27:01,801 INFO [train.py:421] (4/8) Epoch 11, batch 9400, loss[loss=2.326, over 1750.00 frames. , ppl: 10.239934272533455] tot_loss[loss=2.261, over 5517837.64 frames. , ppl: 9.59320482062775], batch size: 70
+2022-12-14 02:28:45,755 INFO [train.py:421] (4/8) Epoch 11, batch 9600, loss[loss=2.235, over 2730.00 frames. , ppl: 9.346093359673004] tot_loss[loss=2.262, over 5507640.82 frames. , ppl: 9.602589609674158], batch size: 70
+2022-12-14 02:30:24,400 INFO [train.py:421] (4/8) Epoch 11, batch 9800, loss[loss=2.131, over 4830.00 frames. , ppl: 8.425182443485552] tot_loss[loss=2.263, over 5509188.45 frames. , ppl: 9.607851852849135], batch size: 70
+2022-12-14 02:32:06,052 INFO [train.py:421] (4/8) Epoch 11, batch 10000, loss[loss=2.246, over 2870.00 frames. , ppl: 9.447041217515721] tot_loss[loss=2.262, over 5534999.51 frames. , ppl: 9.597565430717331], batch size: 70
+2022-12-14 02:32:06,052 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:32:06,813 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.607311196929732
+2022-12-14 02:33:49,149 INFO [train.py:421] (4/8) Epoch 11, batch 10200, loss[loss=2.706, over 910.00 frames. , ppl: 14.970193532544062] tot_loss[loss=2.263, over 5475515.49 frames. , ppl: 9.612349900702364], batch size: 70
+2022-12-14 02:35:28,725 INFO [train.py:421] (4/8) Epoch 11, batch 10400, loss[loss=2.211, over 4060.00 frames. , ppl: 9.123994247603207] tot_loss[loss=2.263, over 5463286.09 frames. , ppl: 9.616603141551504], batch size: 70
+2022-12-14 02:37:09,002 INFO [train.py:421] (4/8) Epoch 11, batch 10600, loss[loss=2.23, over 2520.00 frames. , ppl: 9.29536622186643] tot_loss[loss=2.262, over 5527445.49 frames. , ppl: 9.600936694617719], batch size: 70
+2022-12-14 02:38:52,736 INFO [train.py:421] (4/8) Epoch 11, batch 10800, loss[loss=2.399, over 1680.00 frames. , ppl: 11.009906160631557] tot_loss[loss=2.263, over 5474320.93 frames. , ppl: 9.61498989556506], batch size: 70
+2022-12-14 02:40:34,254 INFO [train.py:421] (4/8) Epoch 11, batch 11000, loss[loss=2.433, over 1260.00 frames. , ppl: 11.388053549189406] tot_loss[loss=2.263, over 5460213.16 frames. , ppl: 9.615009919239176], batch size: 70
+2022-12-14 02:40:34,254 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:40:35,016 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604206151775301
+2022-12-14 02:42:14,871 INFO [train.py:421] (4/8) Epoch 11, batch 11200, loss[loss=2.267, over 2310.00 frames. , ppl: 9.648680626624493] tot_loss[loss=2.264, over 5422857.90 frames. , ppl: 9.625838710637709], batch size: 70
+2022-12-14 02:43:54,215 INFO [train.py:421] (4/8) Epoch 11, batch 11400, loss[loss=4.017, over 350.00 frames. , ppl: 55.526924239617124] tot_loss[loss=2.264, over 5436352.28 frames. , ppl: 9.623066628570694], batch size: 70
+2022-12-14 02:45:34,727 INFO [train.py:421] (4/8) Epoch 11, batch 11600, loss[loss=2.389, over 910.00 frames. , ppl: 10.900728560386904] tot_loss[loss=2.264, over 5432738.67 frames. , ppl: 9.618902305407097], batch size: 70
+2022-12-14 02:47:16,923 INFO [train.py:421] (4/8) Epoch 11, batch 11800, loss[loss=2.151, over 3850.00 frames. , ppl: 8.589306150224484] tot_loss[loss=2.263, over 5466391.30 frames. , ppl: 9.61148141989918], batch size: 70
+2022-12-14 02:48:54,556 INFO [train.py:421] (4/8) Epoch 11, batch 12000, loss[loss=2.105, over 5250.00 frames. , ppl: 8.20640079496174] tot_loss[loss=2.263, over 5470910.53 frames. , ppl: 9.611849447607053], batch size: 70
+2022-12-14 02:48:54,557 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:48:55,303 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.621384617486425
+2022-12-14 02:50:34,275 INFO [train.py:421] (4/8) Epoch 11, batch 12200, loss[loss=2.405, over 910.00 frames. , ppl: 11.076234644202167] tot_loss[loss=2.262, over 5500138.23 frames. , ppl: 9.6064153565446], batch size: 70
+2022-12-14 02:52:10,840 INFO [train.py:421] (4/8) Epoch 11, batch 12400, loss[loss=2.508, over 770.00 frames. , ppl: 12.282668136864162] tot_loss[loss=2.262, over 5510707.80 frames. , ppl: 9.601051472847058], batch size: 70
+2022-12-14 02:53:50,841 INFO [train.py:421] (4/8) Epoch 11, batch 12600, loss[loss=2.134, over 4270.00 frames. , ppl: 8.448620120280935] tot_loss[loss=2.262, over 5520650.13 frames. , ppl: 9.605760649978974], batch size: 70
+2022-12-14 02:55:29,958 INFO [train.py:421] (4/8) Epoch 11, batch 12800, loss[loss=2.343, over 1680.00 frames. , ppl: 10.416370996750212] tot_loss[loss=2.263, over 5489441.88 frames. , ppl: 9.61383507234161], batch size: 70
+2022-12-14 02:57:14,314 INFO [train.py:421] (4/8) Epoch 11, batch 13000, loss[loss=3.561, over 420.00 frames. , ppl: 35.21328605384393] tot_loss[loss=2.263, over 5513055.37 frames. , ppl: 9.611731935959423], batch size: 70
+2022-12-14 02:57:14,315 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 02:57:15,076 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.593847627631558
+2022-12-14 02:58:53,991 INFO [train.py:421] (4/8) Epoch 11, batch 13200, loss[loss=2.2, over 2800.00 frames. , ppl: 9.023002359282408] tot_loss[loss=2.262, over 5549717.36 frames. , ppl: 9.603805061567986], batch size: 70
+2022-12-14 03:00:37,502 INFO [train.py:421] (4/8) Epoch 11, batch 13400, loss[loss=2.413, over 1750.00 frames. , ppl: 11.17099705026436] tot_loss[loss=2.264, over 5477215.96 frames. , ppl: 9.62323727232269], batch size: 70
+2022-12-14 03:02:16,669 INFO [train.py:421] (4/8) Epoch 11, batch 13600, loss[loss=2.385, over 1960.00 frames. , ppl: 10.863119123007788] tot_loss[loss=2.265, over 5450053.35 frames. , ppl: 9.631116445359504], batch size: 70
+2022-12-14 03:03:48,771 INFO [train.py:421] (4/8) Epoch 11, batch 13800, loss[loss=2.34, over 2100.00 frames. , ppl: 10.37619296306955] tot_loss[loss=2.266, over 5440642.02 frames. , ppl: 9.637243139168188], batch size: 70
+2022-12-14 03:05:29,080 INFO [train.py:421] (4/8) Epoch 11, batch 14000, loss[loss=2.418, over 1540.00 frames. , ppl: 11.224893448604648] tot_loss[loss=2.265, over 5464631.96 frames. , ppl: 9.633531831030165], batch size: 70
+2022-12-14 03:05:29,081 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:05:29,843 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.264, over 211138.00 frames. , ppl: 9.617450808382541
+2022-12-14 03:07:07,756 INFO [train.py:421] (4/8) Epoch 11, batch 14200, loss[loss=2.209, over 6440.00 frames. , ppl: 9.107349527757728] tot_loss[loss=2.264, over 5526464.06 frames. , ppl: 9.622183399380248], batch size: 70
+2022-12-14 03:08:46,878 INFO [train.py:421] (4/8) Epoch 11, batch 14400, loss[loss=2.346, over 2660.00 frames. , ppl: 10.441425756674425] tot_loss[loss=2.264, over 5547957.28 frames. , ppl: 9.618351057223016], batch size: 70
+2022-12-14 03:10:30,194 INFO [train.py:421] (4/8) Epoch 11, batch 14600, loss[loss=2.278, over 2870.00 frames. , ppl: 9.758932459551426] tot_loss[loss=2.264, over 5567014.35 frames. , ppl: 9.618295660194693], batch size: 70
+2022-12-14 03:12:09,169 INFO [train.py:421] (4/8) Epoch 11, batch 14800, loss[loss=2.454, over 1190.00 frames. , ppl: 11.631673646114027] tot_loss[loss=2.265, over 5525842.12 frames. , ppl: 9.627179471113445], batch size: 70
+2022-12-14 03:13:48,746 INFO [train.py:421] (4/8) Epoch 11, batch 15000, loss[loss=2.396, over 1960.00 frames. , ppl: 10.983719404385486] tot_loss[loss=2.263, over 5553587.15 frames. , ppl: 9.610365996571502], batch size: 70
+2022-12-14 03:13:48,746 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:13:49,508 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.603182730483441
+2022-12-14 03:15:28,744 INFO [train.py:421] (4/8) Epoch 11, batch 15200, loss[loss=2.464, over 1120.00 frames. , ppl: 11.748412056533732] tot_loss[loss=2.264, over 5550179.26 frames. , ppl: 9.620612091418026], batch size: 70
+2022-12-14 03:17:08,552 INFO [train.py:421] (4/8) Epoch 11, batch 15400, loss[loss=2.434, over 1820.00 frames. , ppl: 11.405913312291476] tot_loss[loss=2.265, over 5524718.57 frames. , ppl: 9.631473302285366], batch size: 70
+2022-12-14 03:18:44,459 INFO [train.py:421] (4/8) Epoch 11, batch 15600, loss[loss=2.46, over 1610.00 frames. , ppl: 11.705188538902416] tot_loss[loss=2.266, over 5507799.27 frames. , ppl: 9.639578023279057], batch size: 70
+2022-12-14 03:20:25,762 INFO [train.py:421] (4/8) Epoch 11, batch 15800, loss[loss=3.201, over 490.00 frames. , ppl: 24.561394008437333] tot_loss[loss=2.266, over 5507141.93 frames. , ppl: 9.642526263999814], batch size: 70
+2022-12-14 03:22:04,991 INFO [train.py:421] (4/8) Epoch 11, batch 16000, loss[loss=2.301, over 2800.00 frames. , ppl: 9.98456139049794] tot_loss[loss=2.266, over 5512884.13 frames. , ppl: 9.638982773896508], batch size: 70
+2022-12-14 03:22:04,991 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:22:05,752 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.608784453983127
+2022-12-14 03:23:48,824 INFO [train.py:421] (4/8) Epoch 11, batch 16200, loss[loss=2.34, over 1820.00 frames. , ppl: 10.38056146563359] tot_loss[loss=2.266, over 5500203.14 frames. , ppl: 9.643688588515584], batch size: 70
+2022-12-14 03:25:29,633 INFO [train.py:421] (4/8) Epoch 11, batch 16400, loss[loss=2.758, over 770.00 frames. , ppl: 15.767868080424783] tot_loss[loss=2.267, over 5480365.46 frames. , ppl: 9.653151734472837], batch size: 70
+2022-12-14 03:27:08,106 INFO [train.py:421] (4/8) Epoch 11, batch 16600, loss[loss=2.259, over 3150.00 frames. , ppl: 9.576454367262878] tot_loss[loss=2.268, over 5480232.31 frames. , ppl: 9.655350645756146], batch size: 70
+2022-12-14 03:28:48,330 INFO [train.py:421] (4/8) Epoch 11, batch 16800, loss[loss=2.182, over 4270.00 frames. , ppl: 8.860983630841305] tot_loss[loss=2.266, over 5535139.41 frames. , ppl: 9.639517335032052], batch size: 70
+2022-12-14 03:30:30,498 INFO [train.py:421] (4/8) Epoch 11, batch 17000, loss[loss=2.385, over 1960.00 frames. , ppl: 10.858140757170569] tot_loss[loss=2.265, over 5582393.79 frames. , ppl: 9.630931547697354], batch size: 70
+2022-12-14 03:30:30,498 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:30:31,251 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599325980992132
+2022-12-14 03:32:09,722 INFO [train.py:421] (4/8) Epoch 11, batch 17200, loss[loss=2.127, over 6300.00 frames. , ppl: 8.393767442532683] tot_loss[loss=2.265, over 5570301.20 frames. , ppl: 9.62922252639846], batch size: 70
+2022-12-14 03:33:49,128 INFO [train.py:421] (4/8) Epoch 11, batch 17400, loss[loss=2.202, over 2380.00 frames. , ppl: 9.04673262648487] tot_loss[loss=2.265, over 5555770.41 frames. , ppl: 9.628944516059791], batch size: 70
+2022-12-14 03:35:30,988 INFO [train.py:421] (4/8) Epoch 11, batch 17600, loss[loss=2.306, over 2800.00 frames. , ppl: 10.03378679486535] tot_loss[loss=2.264, over 5569410.94 frames. , ppl: 9.624796064665233], batch size: 70
+2022-12-14 03:37:07,957 INFO [train.py:421] (4/8) Epoch 11, batch 17800, loss[loss=2.582, over 770.00 frames. , ppl: 13.229285522218774] tot_loss[loss=2.264, over 5559319.47 frames. , ppl: 9.620479052257528], batch size: 70
+2022-12-14 03:38:46,523 INFO [train.py:421] (4/8) Epoch 11, batch 18000, loss[loss=2.496, over 1540.00 frames. , ppl: 12.133209919192842] tot_loss[loss=2.264, over 5574957.03 frames. , ppl: 9.620869970300637], batch size: 70
+2022-12-14 03:38:46,523 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:38:47,290 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.611424367463018
+2022-12-14 03:40:27,911 INFO [train.py:421] (4/8) Epoch 11, batch 18200, loss[loss=2.457, over 1820.00 frames. , ppl: 11.666955725914063] tot_loss[loss=2.264, over 5575102.88 frames. , ppl: 9.620228474706765], batch size: 70
+2022-12-14 03:42:05,024 INFO [train.py:421] (4/8) Epoch 11, batch 18400, loss[loss=2.251, over 3080.00 frames. , ppl: 9.493012244907757] tot_loss[loss=2.266, over 5525846.18 frames. , ppl: 9.63696052600954], batch size: 70
+2022-12-14 03:43:44,909 INFO [train.py:421] (4/8) Epoch 11, batch 18600, loss[loss=2.404, over 1470.00 frames. , ppl: 11.066023972443105] tot_loss[loss=2.265, over 5535577.35 frames. , ppl: 9.634779499936753], batch size: 70
+2022-12-14 03:45:24,640 INFO [train.py:421] (4/8) Epoch 11, batch 18800, loss[loss=2.531, over 1120.00 frames. , ppl: 12.564200346467517] tot_loss[loss=2.266, over 5480644.84 frames. , ppl: 9.643495886616236], batch size: 70
+2022-12-14 03:47:03,770 INFO [train.py:421] (4/8) Epoch 11, batch 19000, loss[loss=2.218, over 4130.00 frames. , ppl: 9.187065860738125] tot_loss[loss=2.266, over 5491912.84 frames. , ppl: 9.640899636698562], batch size: 70
+2022-12-14 03:47:03,770 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:47:04,530 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604402319978165
+2022-12-14 03:48:43,958 INFO [train.py:421] (4/8) Epoch 11, batch 19200, loss[loss=2.132, over 4620.00 frames. , ppl: 8.433971541393348] tot_loss[loss=2.266, over 5496697.94 frames. , ppl: 9.637240224027611], batch size: 70
+2022-12-14 03:50:21,993 INFO [train.py:421] (4/8) Epoch 11, batch 19400, loss[loss=2.211, over 4480.00 frames. , ppl: 9.1293316112287] tot_loss[loss=2.266, over 5470313.50 frames. , ppl: 9.643227955467463], batch size: 70
+2022-12-14 03:52:05,991 INFO [train.py:421] (4/8) Epoch 11, batch 19600, loss[loss=2.184, over 1960.00 frames. , ppl: 8.878230065458549] tot_loss[loss=2.265, over 5516514.25 frames. , ppl: 9.630986339160236], batch size: 70
+2022-12-14 03:53:44,117 INFO [train.py:421] (4/8) Epoch 11, batch 19800, loss[loss=2.58, over 910.00 frames. , ppl: 13.201168700394241] tot_loss[loss=2.265, over 5510376.41 frames. , ppl: 9.628610663046294], batch size: 70
+2022-12-14 03:55:25,012 INFO [train.py:421] (4/8) Epoch 11, batch 20000, loss[loss=2.333, over 1820.00 frames. , ppl: 10.312362711910785] tot_loss[loss=2.265, over 5520356.59 frames. , ppl: 9.628204581448728], batch size: 70
+2022-12-14 03:55:25,013 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 03:55:25,775 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61202043965363
+2022-12-14 03:57:06,005 INFO [train.py:421] (4/8) Epoch 11, batch 20200, loss[loss=2.153, over 5390.00 frames. , ppl: 8.609444099655217] tot_loss[loss=2.265, over 5511635.54 frames. , ppl: 9.63487063288784], batch size: 70
+2022-12-14 03:58:46,599 INFO [train.py:421] (4/8) Epoch 11, batch 20400, loss[loss=2.301, over 2100.00 frames. , ppl: 9.98095201863538] tot_loss[loss=2.266, over 5499716.35 frames. , ppl: 9.636587473413448], batch size: 70
+2022-12-14 04:00:28,898 INFO [train.py:421] (4/8) Epoch 11, batch 20600, loss[loss=2.081, over 3010.00 frames. , ppl: 8.015506606244678] tot_loss[loss=2.266, over 5507120.72 frames. , ppl: 9.639220672045509], batch size: 70
+2022-12-14 04:02:14,438 INFO [train.py:421] (4/8) Epoch 11, batch 20800, loss[loss=2.383, over 2800.00 frames. , ppl: 10.836536236795293] tot_loss[loss=2.265, over 5536075.41 frames. , ppl: 9.62671931864897], batch size: 70
+2022-12-14 04:03:51,808 INFO [train.py:421] (4/8) Epoch 11, batch 21000, loss[loss=2.385, over 1960.00 frames. , ppl: 10.853791966494654] tot_loss[loss=2.266, over 5512699.56 frames. , ppl: 9.635951082645708], batch size: 70
+2022-12-14 04:03:51,808 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:03:52,554 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59960729794989
+2022-12-14 04:05:34,808 INFO [train.py:421] (4/8) Epoch 11, batch 21200, loss[loss=2.293, over 2030.00 frames. , ppl: 9.906005723030542] tot_loss[loss=2.264, over 5535954.71 frames. , ppl: 9.62301678280873], batch size: 70
+2022-12-14 04:07:12,357 INFO [train.py:421] (4/8) Epoch 11, batch 21400, loss[loss=2.31, over 2380.00 frames. , ppl: 10.075250610468341] tot_loss[loss=2.264, over 5527316.90 frames. , ppl: 9.624642515896001], batch size: 70
+2022-12-14 04:08:53,876 INFO [train.py:421] (4/8) Epoch 11, batch 21600, loss[loss=2.163, over 4620.00 frames. , ppl: 8.696077168564788] tot_loss[loss=2.264, over 5528773.36 frames. , ppl: 9.624237570966619], batch size: 70
+2022-12-14 04:10:28,568 INFO [train.py:421] (4/8) Epoch 11, batch 21800, loss[loss=2.349, over 1120.00 frames. , ppl: 10.476330536760216] tot_loss[loss=2.264, over 5543855.03 frames. , ppl: 9.624433195504839], batch size: 70
+2022-12-14 04:12:08,370 INFO [train.py:421] (4/8) Epoch 11, batch 22000, loss[loss=2.451, over 1540.00 frames. , ppl: 11.604182600960801] tot_loss[loss=2.265, over 5530481.67 frames. , ppl: 9.630935237440172], batch size: 70
+2022-12-14 04:12:08,370 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:12:09,131 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59981047643547
+2022-12-14 04:13:46,451 INFO [train.py:421] (4/8) Epoch 11, batch 22200, loss[loss=2.41, over 1400.00 frames. , ppl: 11.136221407028337] tot_loss[loss=2.265, over 5543718.82 frames. , ppl: 9.628381952290303], batch size: 70
+2022-12-14 04:15:26,607 INFO [train.py:421] (4/8) Epoch 11, batch 22400, loss[loss=2.417, over 1330.00 frames. , ppl: 11.217680116536243] tot_loss[loss=2.265, over 5545951.74 frames. , ppl: 9.629717092372347], batch size: 70
+2022-12-14 04:17:04,240 INFO [train.py:421] (4/8) Epoch 11, batch 22600, loss[loss=2.257, over 2870.00 frames. , ppl: 9.557773884339216] tot_loss[loss=2.266, over 5511176.01 frames. , ppl: 9.640012014779089], batch size: 70
+2022-12-14 04:18:46,608 INFO [train.py:421] (4/8) Epoch 11, batch 22800, loss[loss=2.169, over 5880.00 frames. , ppl: 8.74970119969226] tot_loss[loss=2.265, over 5581900.41 frames. , ppl: 9.626691173977024], batch size: 70
+2022-12-14 04:20:26,417 INFO [train.py:421] (4/8) Epoch 11, batch 23000, loss[loss=2.552, over 1120.00 frames. , ppl: 12.83603637306574] tot_loss[loss=2.264, over 5611401.57 frames. , ppl: 9.617343442438179], batch size: 70
+2022-12-14 04:20:26,418 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:20:27,166 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589605722035854
+2022-12-14 04:22:05,160 INFO [train.py:421] (4/8) Epoch 11, batch 23200, loss[loss=2.585, over 770.00 frames. , ppl: 13.25745673245552] tot_loss[loss=2.263, over 5621482.36 frames. , ppl: 9.609441270910587], batch size: 70
+2022-12-14 04:23:43,567 INFO [train.py:421] (4/8) Epoch 11, batch 23400, loss[loss=2.761, over 630.00 frames. , ppl: 15.81943712620202] tot_loss[loss=2.263, over 5632332.43 frames. , ppl: 9.607626102624383], batch size: 70
+2022-12-14 04:25:27,480 INFO [train.py:421] (4/8) Epoch 11, batch 23600, loss[loss=2.261, over 2660.00 frames. , ppl: 9.591735943405572] tot_loss[loss=2.262, over 5655980.30 frames. , ppl: 9.606373874333832], batch size: 70
+2022-12-14 04:27:09,128 INFO [train.py:421] (4/8) Epoch 11, batch 23800, loss[loss=2.379, over 1400.00 frames. , ppl: 10.794522002751494] tot_loss[loss=2.263, over 5627495.91 frames. , ppl: 9.615218540983518], batch size: 70
+2022-12-14 04:28:49,640 INFO [train.py:421] (4/8) Epoch 11, batch 24000, loss[loss=2.236, over 2730.00 frames. , ppl: 9.358871591006556] tot_loss[loss=2.262, over 5653215.99 frames. , ppl: 9.604648720189912], batch size: 70
+2022-12-14 04:28:49,640 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:28:50,387 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589086260144834
+2022-12-14 04:30:27,488 INFO [train.py:421] (4/8) Epoch 11, batch 24200, loss[loss=2.696, over 980.00 frames. , ppl: 14.81829496679929] tot_loss[loss=2.263, over 5637197.43 frames. , ppl: 9.609917813806572], batch size: 70
+2022-12-14 04:32:06,790 INFO [train.py:421] (4/8) Epoch 11, batch 24400, loss[loss=2.434, over 1120.00 frames. , ppl: 11.403469740504626] tot_loss[loss=2.265, over 5561263.10 frames. , ppl: 9.62846168580296], batch size: 70
+2022-12-14 04:33:49,455 INFO [train.py:421] (4/8) Epoch 11, batch 24600, loss[loss=2.37, over 2660.00 frames. , ppl: 10.69603037611195] tot_loss[loss=2.264, over 5582464.28 frames. , ppl: 9.61715284553428], batch size: 70
+2022-12-14 04:35:31,657 INFO [train.py:421] (4/8) Epoch 11, batch 24800, loss[loss=2.274, over 3360.00 frames. , ppl: 9.722141610282424] tot_loss[loss=2.263, over 5601591.33 frames. , ppl: 9.61358670918465], batch size: 70
+2022-12-14 04:37:15,799 INFO [train.py:421] (4/8) Epoch 11, batch 25000, loss[loss=2.509, over 1050.00 frames. , ppl: 12.289841403431193] tot_loss[loss=2.263, over 5613093.24 frames. , ppl: 9.607828412315916], batch size: 70
+2022-12-14 04:37:15,800 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:37:16,563 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.606207824169385
+2022-12-14 04:38:53,653 INFO [train.py:421] (4/8) Epoch 11, batch 25200, loss[loss=2.491, over 1120.00 frames. , ppl: 12.07766211026014] tot_loss[loss=2.264, over 5570672.43 frames. , ppl: 9.619688988483183], batch size: 70
+2022-12-14 04:40:34,680 INFO [train.py:421] (4/8) Epoch 11, batch 25400, loss[loss=2.657, over 910.00 frames. , ppl: 14.256763014750426] tot_loss[loss=2.263, over 5570438.47 frames. , ppl: 9.615785884536255], batch size: 70
+2022-12-14 04:42:13,996 INFO [train.py:421] (4/8) Epoch 11, batch 25600, loss[loss=2.355, over 2590.00 frames. , ppl: 10.536379512859808] tot_loss[loss=2.263, over 5561805.60 frames. , ppl: 9.610235691780586], batch size: 70
+2022-12-14 04:43:56,157 INFO [train.py:421] (4/8) Epoch 11, batch 25800, loss[loss=2.214, over 1400.00 frames. , ppl: 9.148164814141307] tot_loss[loss=2.261, over 5595951.42 frames. , ppl: 9.594818324954739], batch size: 70
+2022-12-14 04:45:33,986 INFO [train.py:421] (4/8) Epoch 11, batch 26000, loss[loss=2.187, over 3710.00 frames. , ppl: 8.91126438554423] tot_loss[loss=2.263, over 5527259.16 frames. , ppl: 9.615759414245531], batch size: 70
+2022-12-14 04:45:33,986 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:45:34,741 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.612918174211162
+2022-12-14 04:47:14,142 INFO [train.py:421] (4/8) Epoch 11, batch 26200, loss[loss=2.243, over 4270.00 frames. , ppl: 9.419292164700655] tot_loss[loss=2.264, over 5514438.03 frames. , ppl: 9.621905474428537], batch size: 70
+2022-12-14 04:48:55,148 INFO [train.py:421] (4/8) Epoch 11, batch 26400, loss[loss=2.507, over 980.00 frames. , ppl: 12.262932379581038] tot_loss[loss=2.264, over 5527321.98 frames. , ppl: 9.61870390997227], batch size: 70
+2022-12-14 04:50:33,171 INFO [train.py:421] (4/8) Epoch 11, batch 26600, loss[loss=2.454, over 1470.00 frames. , ppl: 11.630885493401134] tot_loss[loss=2.264, over 5489405.67 frames. , ppl: 9.621687064570924], batch size: 70
+2022-12-14 04:52:12,404 INFO [train.py:421] (4/8) Epoch 11, batch 26800, loss[loss=2.955, over 560.00 frames. , ppl: 19.20698228237056] tot_loss[loss=2.265, over 5470984.68 frames. , ppl: 9.631611579750698], batch size: 70
+2022-12-14 04:53:51,698 INFO [train.py:421] (4/8) Epoch 11, batch 27000, loss[loss=2.331, over 1330.00 frames. , ppl: 10.286477702301934] tot_loss[loss=2.266, over 5488695.90 frames. , ppl: 9.638911716482534], batch size: 70
+2022-12-14 04:53:51,699 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 04:53:52,445 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.594008084573185
+2022-12-14 04:55:35,141 INFO [train.py:421] (4/8) Epoch 11, batch 27200, loss[loss=2.528, over 1120.00 frames. , ppl: 12.523150273800525] tot_loss[loss=2.265, over 5501376.30 frames. , ppl: 9.630290275624159], batch size: 70
+2022-12-14 04:57:14,463 INFO [train.py:421] (4/8) Epoch 11, batch 27400, loss[loss=2.269, over 3430.00 frames. , ppl: 9.66645591562951] tot_loss[loss=2.265, over 5485968.27 frames. , ppl: 9.630467676796238], batch size: 70
+2022-12-14 04:58:50,882 INFO [train.py:421] (4/8) Epoch 11, batch 27600, loss[loss=2.223, over 3780.00 frames. , ppl: 9.234971524143607] tot_loss[loss=2.265, over 5498055.66 frames. , ppl: 9.633875645408477], batch size: 70
+2022-12-14 05:00:29,827 INFO [train.py:421] (4/8) Epoch 11, batch 27800, loss[loss=2.333, over 1610.00 frames. , ppl: 10.31257803300022] tot_loss[loss=2.265, over 5473152.13 frames. , ppl: 9.63506087589492], batch size: 70
+2022-12-14 05:02:14,476 INFO [train.py:421] (4/8) Epoch 11, batch 28000, loss[loss=2.36, over 2100.00 frames. , ppl: 10.58570751146104] tot_loss[loss=2.265, over 5477510.65 frames. , ppl: 9.63457574307885], batch size: 70
+2022-12-14 05:02:14,477 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:02:15,236 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.594497991840212
+2022-12-14 05:04:00,693 INFO [train.py:421] (4/8) Epoch 11, batch 28200, loss[loss=2.197, over 2730.00 frames. , ppl: 8.994576077752379] tot_loss[loss=2.265, over 5497641.43 frames. , ppl: 9.631666475006941], batch size: 70
+2022-12-14 05:05:42,753 INFO [train.py:421] (4/8) Epoch 11, batch 28400, loss[loss=2.26, over 2800.00 frames. , ppl: 9.579981316702755] tot_loss[loss=2.265, over 5491447.68 frames. , ppl: 9.635663049481309], batch size: 70
+2022-12-14 05:07:20,249 INFO [train.py:421] (4/8) Epoch 11, batch 28600, loss[loss=2.172, over 3150.00 frames. , ppl: 8.778633946048997] tot_loss[loss=2.264, over 5518638.75 frames. , ppl: 9.624601020439885], batch size: 70
+2022-12-14 05:09:04,634 INFO [train.py:421] (4/8) Epoch 11, batch 28800, loss[loss=2.359, over 2240.00 frames. , ppl: 10.575547386500995] tot_loss[loss=2.266, over 5452229.98 frames. , ppl: 9.643750952531919], batch size: 70
+2022-12-14 05:10:45,350 INFO [train.py:421] (4/8) Epoch 11, batch 29000, loss[loss=2.479, over 1680.00 frames. , ppl: 11.92861989616669] tot_loss[loss=2.265, over 5490309.83 frames. , ppl: 9.630733091315431], batch size: 70
+2022-12-14 05:10:45,351 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:10:46,097 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.587935312328966
+2022-12-14 05:12:24,279 INFO [train.py:421] (4/8) Epoch 11, batch 29200, loss[loss=2.633, over 1120.00 frames. , ppl: 13.912665871642911] tot_loss[loss=2.264, over 5499773.96 frames. , ppl: 9.626165360858693], batch size: 70
+2022-12-14 05:14:04,310 INFO [train.py:421] (4/8) Epoch 11, batch 29400, loss[loss=2.364, over 2940.00 frames. , ppl: 10.632281068190778] tot_loss[loss=2.265, over 5494102.81 frames. , ppl: 9.630483977875206], batch size: 70
+2022-12-14 05:15:46,582 INFO [train.py:421] (4/8) Epoch 11, batch 29600, loss[loss=2.322, over 2240.00 frames. , ppl: 10.198567960997245] tot_loss[loss=2.265, over 5512819.62 frames. , ppl: 9.630400533858875], batch size: 70
+2022-12-14 05:17:28,206 INFO [train.py:421] (4/8) Epoch 11, batch 29800, loss[loss=2.247, over 3920.00 frames. , ppl: 9.463959357831936] tot_loss[loss=2.264, over 5539839.47 frames. , ppl: 9.621725414673465], batch size: 70
+2022-12-14 05:19:07,032 INFO [train.py:421] (4/8) Epoch 11, batch 30000, loss[loss=2.289, over 3780.00 frames. , ppl: 9.862427438866632] tot_loss[loss=2.263, over 5573367.55 frames. , ppl: 9.615818027230414], batch size: 70
+2022-12-14 05:19:07,033 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:19:07,779 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.613103137864265
+2022-12-14 05:20:48,039 INFO [train.py:421] (4/8) Epoch 11, batch 30200, loss[loss=2.367, over 1890.00 frames. , ppl: 10.664463092250843] tot_loss[loss=2.264, over 5545339.85 frames. , ppl: 9.620976576138734], batch size: 70
+2022-12-14 05:22:30,481 INFO [train.py:421] (4/8) Epoch 11, batch 30400, loss[loss=2.248, over 6090.00 frames. , ppl: 9.467761043915564] tot_loss[loss=2.264, over 5540344.99 frames. , ppl: 9.624977524768171], batch size: 70
+2022-12-14 05:24:12,352 INFO [train.py:421] (4/8) Epoch 11, batch 30600, loss[loss=2.241, over 4200.00 frames. , ppl: 9.400526247664088] tot_loss[loss=2.264, over 5549996.67 frames. , ppl: 9.623368303699413], batch size: 70
+2022-12-14 05:25:49,459 INFO [train.py:421] (4/8) Epoch 11, batch 30800, loss[loss=2.479, over 2170.00 frames. , ppl: 11.930240091008736] tot_loss[loss=2.264, over 5549408.89 frames. , ppl: 9.625909046568156], batch size: 70
+2022-12-14 05:27:28,454 INFO [train.py:421] (4/8) Epoch 11, batch 31000, loss[loss=3.071, over 560.00 frames. , ppl: 21.572661193917053] tot_loss[loss=2.265, over 5549770.37 frames. , ppl: 9.630158255603838], batch size: 70
+2022-12-14 05:27:28,454 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:27:29,214 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.613066144848931
+2022-12-14 05:29:07,440 INFO [train.py:421] (4/8) Epoch 11, batch 31200, loss[loss=2.197, over 4760.00 frames. , ppl: 8.995263418065127] tot_loss[loss=2.265, over 5553957.63 frames. , ppl: 9.62988967669272], batch size: 70
+2022-12-14 05:30:46,537 INFO [train.py:421] (4/8) Epoch 11, batch 31400, loss[loss=2.282, over 2310.00 frames. , ppl: 9.801033356665313] tot_loss[loss=2.264, over 5574568.09 frames. , ppl: 9.624158392885892], batch size: 70
+2022-12-14 05:32:28,444 INFO [train.py:421] (4/8) Epoch 11, batch 31600, loss[loss=2.497, over 980.00 frames. , ppl: 12.141827016848955] tot_loss[loss=2.265, over 5549050.47 frames. , ppl: 9.631982799571526], batch size: 70
+2022-12-14 05:34:09,689 INFO [train.py:421] (4/8) Epoch 11, batch 31800, loss[loss=2.381, over 2030.00 frames. , ppl: 10.819552048405011] tot_loss[loss=2.266, over 5515419.63 frames. , ppl: 9.644562854412122], batch size: 70
+2022-12-14 05:35:44,348 INFO [train.py:421] (4/8) Epoch 11, batch 32000, loss[loss=2.398, over 1190.00 frames. , ppl: 11.005200685900668] tot_loss[loss=2.267, over 5478657.06 frames. , ppl: 9.650288450171857], batch size: 70
+2022-12-14 05:35:44,349 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:35:45,109 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604530257921299
+2022-12-14 05:37:24,334 INFO [train.py:421] (4/8) Epoch 11, batch 32200, loss[loss=2.298, over 3710.00 frames. , ppl: 9.95372234769024] tot_loss[loss=2.267, over 5471548.87 frames. , ppl: 9.649727821349478], batch size: 70
+2022-12-14 05:39:05,741 INFO [train.py:421] (4/8) Epoch 11, batch 32400, loss[loss=3.11, over 490.00 frames. , ppl: 22.41076478587762] tot_loss[loss=2.267, over 5494484.91 frames. , ppl: 9.648208196994299], batch size: 70
+2022-12-14 05:40:47,672 INFO [train.py:421] (4/8) Epoch 11, batch 32600, loss[loss=2.307, over 1960.00 frames. , ppl: 10.044633126159628] tot_loss[loss=2.267, over 5497257.61 frames. , ppl: 9.650358631623215], batch size: 70
+2022-12-14 05:42:26,787 INFO [train.py:421] (4/8) Epoch 11, batch 32800, loss[loss=2.375, over 1260.00 frames. , ppl: 10.748788620467684] tot_loss[loss=2.266, over 5505992.58 frames. , ppl: 9.64370502776978], batch size: 70
+2022-12-14 05:44:07,652 INFO [train.py:421] (4/8) Epoch 11, batch 33000, loss[loss=2.252, over 3990.00 frames. , ppl: 9.503986189781092] tot_loss[loss=2.265, over 5549054.06 frames. , ppl: 9.633548973357694], batch size: 70
+2022-12-14 05:44:07,652 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:44:08,397 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.5747582284789
+2022-12-14 05:45:47,674 INFO [train.py:421] (4/8) Epoch 11, batch 33200, loss[loss=2.202, over 4830.00 frames. , ppl: 9.042646734152399] tot_loss[loss=2.265, over 5535497.92 frames. , ppl: 9.633213190566005], batch size: 70
+2022-12-14 05:47:26,447 INFO [train.py:421] (4/8) Epoch 11, batch 33400, loss[loss=2.464, over 1120.00 frames. , ppl: 11.74876291200362] tot_loss[loss=2.265, over 5530002.03 frames. , ppl: 9.631502011261805], batch size: 70
+2022-12-14 05:49:07,562 INFO [train.py:421] (4/8) Epoch 11, batch 33600, loss[loss=2.547, over 1400.00 frames. , ppl: 12.764610646065595] tot_loss[loss=2.265, over 5555027.85 frames. , ppl: 9.627480071965826], batch size: 70
+2022-12-14 05:50:49,702 INFO [train.py:421] (4/8) Epoch 11, batch 33800, loss[loss=2.168, over 7980.00 frames. , ppl: 8.740069016239747] tot_loss[loss=2.265, over 5572645.98 frames. , ppl: 9.626399937389028], batch size: 70
+2022-12-14 05:52:29,250 INFO [train.py:421] (4/8) Epoch 11, batch 34000, loss[loss=2.228, over 4130.00 frames. , ppl: 9.285738083882624] tot_loss[loss=2.264, over 5585859.57 frames. , ppl: 9.621045170457853], batch size: 70
+2022-12-14 05:52:29,251 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 05:52:30,011 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.580360386976093
+2022-12-14 05:54:12,838 INFO [train.py:421] (4/8) Epoch 11, batch 34200, loss[loss=2.243, over 2730.00 frames. , ppl: 9.424404028486657] tot_loss[loss=2.263, over 5588703.08 frames. , ppl: 9.610298541833417], batch size: 70
+2022-12-14 05:55:56,791 INFO [train.py:421] (4/8) Epoch 11, batch 34400, loss[loss=2.236, over 2940.00 frames. , ppl: 9.358598748416885] tot_loss[loss=2.263, over 5579920.23 frames. , ppl: 9.614226533755389], batch size: 70
+2022-12-14 05:57:35,021 INFO [train.py:421] (4/8) Epoch 11, batch 34600, loss[loss=2.364, over 1540.00 frames. , ppl: 10.629995846751514] tot_loss[loss=2.263, over 5576470.97 frames. , ppl: 9.61206193766831], batch size: 70
+2022-12-14 05:59:12,269 INFO [train.py:421] (4/8) Epoch 11, batch 34800, loss[loss=2.445, over 1400.00 frames. , ppl: 11.532005486975354] tot_loss[loss=2.264, over 5537077.63 frames. , ppl: 9.619160980990777], batch size: 70
+2022-12-14 06:00:51,154 INFO [train.py:421] (4/8) Epoch 11, batch 35000, loss[loss=2.689, over 700.00 frames. , ppl: 14.714591705633769] tot_loss[loss=2.265, over 5499672.29 frames. , ppl: 9.627898111776505], batch size: 70
+2022-12-14 06:00:51,154 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:00:51,940 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.571281208986392
+2022-12-14 06:02:35,354 INFO [train.py:421] (4/8) Epoch 11, batch 35200, loss[loss=2.359, over 2520.00 frames. , ppl: 10.585515529109255] tot_loss[loss=2.264, over 5508546.39 frames. , ppl: 9.626004502324058], batch size: 70
+2022-12-14 06:04:14,594 INFO [train.py:421] (4/8) Epoch 11, batch 35400, loss[loss=2.302, over 2170.00 frames. , ppl: 9.992288205891914] tot_loss[loss=2.264, over 5536614.14 frames. , ppl: 9.618778179942735], batch size: 70
+2022-12-14 06:05:56,229 INFO [train.py:421] (4/8) Epoch 11, batch 35600, loss[loss=2.312, over 1330.00 frames. , ppl: 10.09075837270041] tot_loss[loss=2.263, over 5564762.81 frames. , ppl: 9.609448372710801], batch size: 70
+2022-12-14 06:07:32,112 INFO [train.py:421] (4/8) Epoch 11, batch 35800, loss[loss=2.476, over 1120.00 frames. , ppl: 11.892750850516657] tot_loss[loss=2.263, over 5563609.80 frames. , ppl: 9.610710232187948], batch size: 70
+2022-12-14 06:09:12,202 INFO [train.py:421] (4/8) Epoch 11, batch 36000, loss[loss=2.234, over 5110.00 frames. , ppl: 9.33699058160449] tot_loss[loss=2.265, over 5521521.42 frames. , ppl: 9.629420691002107], batch size: 70
+2022-12-14 06:09:12,203 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:09:12,963 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.593320836541546
+2022-12-14 06:10:52,363 INFO [train.py:421] (4/8) Epoch 11, batch 36200, loss[loss=2.18, over 3780.00 frames. , ppl: 8.84709339998122] tot_loss[loss=2.264, over 5523847.17 frames. , ppl: 9.62097246082057], batch size: 70
+2022-12-14 06:12:34,282 INFO [train.py:421] (4/8) Epoch 11, batch 36400, loss[loss=2.623, over 840.00 frames. , ppl: 13.779716871360174] tot_loss[loss=2.265, over 5489642.52 frames. , ppl: 9.63584052953621], batch size: 70
+2022-12-14 06:14:15,795 INFO [train.py:421] (4/8) Epoch 11, batch 36600, loss[loss=3.114, over 490.00 frames. , ppl: 22.512674098288596] tot_loss[loss=2.265, over 5492830.65 frames. , ppl: 9.63587069965284], batch size: 70
+2022-12-14 06:15:56,234 INFO [train.py:421] (4/8) Epoch 11, batch 36800, loss[loss=2.895, over 630.00 frames. , ppl: 18.08820008425562] tot_loss[loss=2.265, over 5489632.40 frames. , ppl: 9.635707735676803], batch size: 70
+2022-12-14 06:17:38,558 INFO [train.py:421] (4/8) Epoch 11, batch 37000, loss[loss=2.494, over 980.00 frames. , ppl: 12.1095016379027] tot_loss[loss=2.266, over 5492637.60 frames. , ppl: 9.638437475610251], batch size: 70
+2022-12-14 06:17:38,558 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:17:39,307 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.585279151337225
+2022-12-14 06:19:19,620 INFO [train.py:421] (4/8) Epoch 11, batch 37200, loss[loss=2.366, over 1680.00 frames. , ppl: 10.659206504698298] tot_loss[loss=2.266, over 5478759.37 frames. , ppl: 9.645333234678047], batch size: 70
+2022-12-14 06:21:00,621 INFO [train.py:421] (4/8) Epoch 11, batch 37400, loss[loss=2.212, over 1820.00 frames. , ppl: 9.13103248534374] tot_loss[loss=2.266, over 5479246.94 frames. , ppl: 9.639690469916305], batch size: 70
+2022-12-14 06:22:44,801 INFO [train.py:421] (4/8) Epoch 11, batch 37600, loss[loss=2.387, over 1680.00 frames. , ppl: 10.877369671387559] tot_loss[loss=2.266, over 5489402.84 frames. , ppl: 9.6383939081796], batch size: 70
+2022-12-14 06:24:22,292 INFO [train.py:421] (4/8) Epoch 11, batch 37800, loss[loss=2.22, over 2800.00 frames. , ppl: 9.208747141409805] tot_loss[loss=2.267, over 5427815.61 frames. , ppl: 9.646735412809493], batch size: 70
+2022-12-14 06:26:03,854 INFO [train.py:421] (4/8) Epoch 11, batch 38000, loss[loss=2.113, over 9170.00 frames. , ppl: 8.270937971873876] tot_loss[loss=2.265, over 5472367.10 frames. , ppl: 9.63593371651474], batch size: 70
+2022-12-14 06:26:03,855 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:26:04,620 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.584332929905047
+2022-12-14 06:27:46,869 INFO [train.py:421] (4/8) Epoch 11, batch 38200, loss[loss=2.141, over 9170.00 frames. , ppl: 8.506912752212282] tot_loss[loss=2.265, over 5518305.33 frames. , ppl: 9.628346350258104], batch size: 70
+2022-12-14 06:29:24,564 INFO [train.py:421] (4/8) Epoch 11, batch 38400, loss[loss=2.128, over 8820.00 frames. , ppl: 8.396087469238292] tot_loss[loss=2.265, over 5496274.70 frames. , ppl: 9.633751856272953], batch size: 70
+2022-12-14 06:31:03,586 INFO [train.py:421] (4/8) Epoch 11, batch 38600, loss[loss=2.354, over 1610.00 frames. , ppl: 10.52906441304249] tot_loss[loss=2.265, over 5484560.07 frames. , ppl: 9.632229850225352], batch size: 70
+2022-12-14 06:32:43,274 INFO [train.py:421] (4/8) Epoch 11, batch 38800, loss[loss=2.345, over 1820.00 frames. , ppl: 10.428959638383354] tot_loss[loss=2.264, over 5548603.20 frames. , ppl: 9.620685967317664], batch size: 70
+2022-12-14 06:34:25,390 INFO [train.py:421] (4/8) Epoch 11, batch 39000, loss[loss=2.242, over 4690.00 frames. , ppl: 9.41129696607338] tot_loss[loss=2.264, over 5559860.17 frames. , ppl: 9.620069514394073], batch size: 70
+2022-12-14 06:34:25,390 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:34:26,137 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.58167067609052
+2022-12-14 06:36:07,551 INFO [train.py:421] (4/8) Epoch 11, batch 39200, loss[loss=2.199, over 6090.00 frames. , ppl: 9.017699386954918] tot_loss[loss=2.265, over 5516302.11 frames. , ppl: 9.633611954475365], batch size: 70
+2022-12-14 06:37:47,055 INFO [train.py:421] (4/8) Epoch 11, batch 39400, loss[loss=2.137, over 7420.00 frames. , ppl: 8.476845436844279] tot_loss[loss=2.265, over 5503702.32 frames. , ppl: 9.635643966958881], batch size: 70
+2022-12-14 06:39:24,955 INFO [train.py:421] (4/8) Epoch 11, batch 39600, loss[loss=2.425, over 1820.00 frames. , ppl: 11.298770169256889] tot_loss[loss=2.266, over 5496965.60 frames. , ppl: 9.63964343436818], batch size: 70
+2022-12-14 06:41:07,695 INFO [train.py:421] (4/8) Epoch 11, batch 39800, loss[loss=2.19, over 6930.00 frames. , ppl: 8.934909920534277] tot_loss[loss=2.265, over 5522444.16 frames. , ppl: 9.631438935836982], batch size: 70
+2022-12-14 06:42:42,717 INFO [train.py:421] (4/8) Epoch 11, batch 40000, loss[loss=2.343, over 2100.00 frames. , ppl: 10.408301424485767] tot_loss[loss=2.264, over 5546701.01 frames. , ppl: 9.621932551436936], batch size: 70
+2022-12-14 06:42:42,717 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:42:43,479 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.58547067678701
+2022-12-14 06:44:22,144 INFO [train.py:421] (4/8) Epoch 11, batch 40200, loss[loss=2.469, over 840.00 frames. , ppl: 11.80716326283272] tot_loss[loss=2.265, over 5511988.99 frames. , ppl: 9.631670622086256], batch size: 70
+2022-12-14 06:46:06,984 INFO [train.py:421] (4/8) Epoch 11, batch 40400, loss[loss=2.223, over 3290.00 frames. , ppl: 9.232943923430529] tot_loss[loss=2.264, over 5506888.44 frames. , ppl: 9.624646194197185], batch size: 70
+2022-12-14 06:47:50,938 INFO [train.py:421] (4/8) Epoch 11, batch 40600, loss[loss=2.248, over 3640.00 frames. , ppl: 9.467580919521582] tot_loss[loss=2.264, over 5545574.16 frames. , ppl: 9.61845629523022], batch size: 70
+2022-12-14 06:49:30,439 INFO [train.py:421] (4/8) Epoch 11, batch 40800, loss[loss=2.155, over 4550.00 frames. , ppl: 8.62796054752211] tot_loss[loss=2.262, over 5635358.58 frames. , ppl: 9.600275137859564], batch size: 70
+2022-12-14 06:51:11,966 INFO [train.py:421] (4/8) Epoch 11, batch 41000, loss[loss=2.403, over 1330.00 frames. , ppl: 11.059581799377996] tot_loss[loss=2.262, over 5637126.07 frames. , ppl: 9.598561775701594], batch size: 70
+2022-12-14 06:51:11,967 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:51:12,725 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.258, over 211138.00 frames. , ppl: 9.567495329829377
+2022-12-14 06:52:55,089 INFO [train.py:421] (4/8) Epoch 11, batch 41200, loss[loss=2.295, over 1750.00 frames. , ppl: 9.92615023003834] tot_loss[loss=2.261, over 5648477.78 frames. , ppl: 9.592874806961046], batch size: 70
+2022-12-14 06:54:38,130 INFO [train.py:421] (4/8) Epoch 11, batch 41400, loss[loss=2.36, over 1470.00 frames. , ppl: 10.588669966198829] tot_loss[loss=2.261, over 5646893.33 frames. , ppl: 9.589783669927106], batch size: 70
+2022-12-14 06:56:20,581 INFO [train.py:421] (4/8) Epoch 11, batch 41600, loss[loss=2.33, over 1260.00 frames. , ppl: 10.28223446407421] tot_loss[loss=2.261, over 5677464.19 frames. , ppl: 9.588917771701233], batch size: 70
+2022-12-14 06:58:01,897 INFO [train.py:421] (4/8) Epoch 11, batch 41800, loss[loss=2.23, over 3850.00 frames. , ppl: 9.297672528496214] tot_loss[loss=2.262, over 5643450.61 frames. , ppl: 9.602427570603107], batch size: 70
+2022-12-14 06:59:44,359 INFO [train.py:421] (4/8) Epoch 11, batch 42000, loss[loss=2.426, over 1470.00 frames. , ppl: 11.308551374077641] tot_loss[loss=2.262, over 5654813.60 frames. , ppl: 9.602103358266762], batch size: 70
+2022-12-14 06:59:44,360 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 06:59:45,115 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.592491660652186
+2022-12-14 07:01:25,039 INFO [train.py:421] (4/8) Epoch 11, batch 42200, loss[loss=2.355, over 2870.00 frames. , ppl: 10.537707901459662] tot_loss[loss=2.262, over 5672524.21 frames. , ppl: 9.603785523232151], batch size: 70
+2022-12-14 07:03:05,977 INFO [train.py:421] (4/8) Epoch 11, batch 42400, loss[loss=2.294, over 2870.00 frames. , ppl: 9.911889818654492] tot_loss[loss=2.263, over 5620778.45 frames. , ppl: 9.613902148929293], batch size: 70
+2022-12-14 07:04:49,208 INFO [train.py:421] (4/8) Epoch 11, batch 42600, loss[loss=2.398, over 1680.00 frames. , ppl: 11.00302040349279] tot_loss[loss=2.263, over 5623947.10 frames. , ppl: 9.612966953188462], batch size: 70
+2022-12-14 07:06:29,113 INFO [train.py:421] (4/8) Epoch 11, batch 42800, loss[loss=2.323, over 2380.00 frames. , ppl: 10.207707734099904] tot_loss[loss=2.263, over 5646815.03 frames. , ppl: 9.609142017396147], batch size: 70
+2022-12-14 07:08:09,283 INFO [train.py:421] (4/8) Epoch 11, batch 43000, loss[loss=2.212, over 2940.00 frames. , ppl: 9.13470420486642] tot_loss[loss=2.262, over 5669453.71 frames. , ppl: 9.60502022830462], batch size: 70
+2022-12-14 07:08:09,284 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:08:10,045 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.569812285675939
+2022-12-14 07:09:48,736 INFO [train.py:421] (4/8) Epoch 11, batch 43200, loss[loss=2.485, over 1120.00 frames. , ppl: 12.002843823503119] tot_loss[loss=2.263, over 5638309.76 frames. , ppl: 9.609720346949034], batch size: 70
+2022-12-14 07:11:27,396 INFO [train.py:421] (4/8) Epoch 11, batch 43400, loss[loss=2.455, over 840.00 frames. , ppl: 11.643610156376363] tot_loss[loss=2.263, over 5643140.96 frames. , ppl: 9.61387591233624], batch size: 70
+2022-12-14 07:13:07,427 INFO [train.py:421] (4/8) Epoch 11, batch 43600, loss[loss=2.384, over 1470.00 frames. , ppl: 10.853590074920612] tot_loss[loss=2.265, over 5580489.01 frames. , ppl: 9.627508488754382], batch size: 70
+2022-12-14 07:14:45,248 INFO [train.py:421] (4/8) Epoch 11, batch 43800, loss[loss=2.189, over 3290.00 frames. , ppl: 8.92912198837559] tot_loss[loss=2.265, over 5544246.80 frames. , ppl: 9.632565813998253], batch size: 70
+2022-12-14 07:16:27,745 INFO [train.py:421] (4/8) Epoch 11, batch 44000, loss[loss=2.209, over 3710.00 frames. , ppl: 9.104037458756103] tot_loss[loss=2.266, over 5507467.79 frames. , ppl: 9.642336962101385], batch size: 70
+2022-12-14 07:16:27,746 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:16:28,507 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.577322170162937
+2022-12-14 07:18:06,699 INFO [train.py:421] (4/8) Epoch 11, batch 44200, loss[loss=2.434, over 840.00 frames. , ppl: 11.399620777900926] tot_loss[loss=2.266, over 5529865.25 frames. , ppl: 9.637676608321682], batch size: 70
+2022-12-14 07:19:47,740 INFO [train.py:421] (4/8) Epoch 11, batch 44400, loss[loss=2.277, over 4410.00 frames. , ppl: 9.746078420205038] tot_loss[loss=2.266, over 5506133.46 frames. , ppl: 9.643069115039733], batch size: 70
+2022-12-14 07:21:28,935 INFO [train.py:421] (4/8) Epoch 11, batch 44600, loss[loss=2.211, over 7420.00 frames. , ppl: 9.122763025547087] tot_loss[loss=2.265, over 5524297.20 frames. , ppl: 9.635126466519248], batch size: 70
+2022-12-14 07:23:10,233 INFO [train.py:421] (4/8) Epoch 11, batch 44800, loss[loss=2.38, over 1470.00 frames. , ppl: 10.802459748297668] tot_loss[loss=2.265, over 5530312.43 frames. , ppl: 9.633951797015213], batch size: 70
+2022-12-14 07:24:47,373 INFO [train.py:421] (4/8) Epoch 11, batch 45000, loss[loss=2.339, over 1540.00 frames. , ppl: 10.372811022113266] tot_loss[loss=2.266, over 5499977.61 frames. , ppl: 9.643293999769345], batch size: 70
+2022-12-14 07:24:47,374 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:24:48,135 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.587963694090844
+2022-12-14 07:26:33,406 INFO [train.py:421] (4/8) Epoch 11, batch 45200, loss[loss=2.221, over 3150.00 frames. , ppl: 9.217581207308418] tot_loss[loss=2.266, over 5511320.61 frames. , ppl: 9.641619420137632], batch size: 70
+2022-12-14 07:28:14,119 INFO [train.py:421] (4/8) Epoch 11, batch 45400, loss[loss=2.229, over 3290.00 frames. , ppl: 9.293565265558064] tot_loss[loss=2.265, over 5524070.02 frames. , ppl: 9.62972020598843], batch size: 70
+2022-12-14 07:29:58,381 INFO [train.py:421] (4/8) Epoch 11, batch 45600, loss[loss=2.884, over 700.00 frames. , ppl: 17.882206216960657] tot_loss[loss=2.265, over 5531791.37 frames. , ppl: 9.627857141233958], batch size: 70
+2022-12-14 07:31:41,352 INFO [train.py:421] (4/8) Epoch 11, batch 45800, loss[loss=2.144, over 4830.00 frames. , ppl: 8.535609564593221] tot_loss[loss=2.265, over 5519240.57 frames. , ppl: 9.63282912297055], batch size: 70
+2022-12-14 07:33:22,374 INFO [train.py:421] (4/8) Epoch 11, batch 46000, loss[loss=2.276, over 4060.00 frames. , ppl: 9.739948758253783] tot_loss[loss=2.264, over 5553693.71 frames. , ppl: 9.62370631285343], batch size: 70
+2022-12-14 07:33:22,375 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:33:23,121 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.604798933149176
+2022-12-14 07:35:07,449 INFO [train.py:421] (4/8) Epoch 11, batch 46200, loss[loss=2.257, over 3290.00 frames. , ppl: 9.552300956329892] tot_loss[loss=2.263, over 5589312.69 frames. , ppl: 9.613826564857723], batch size: 70
+2022-12-14 07:36:50,560 INFO [train.py:421] (4/8) Epoch 11, batch 46400, loss[loss=2.42, over 1470.00 frames. , ppl: 11.245003177538498] tot_loss[loss=2.262, over 5635722.04 frames. , ppl: 9.59976597106405], batch size: 70
+2022-12-14 07:38:31,817 INFO [train.py:421] (4/8) Epoch 11, batch 46600, loss[loss=2.318, over 2380.00 frames. , ppl: 10.156201055286814] tot_loss[loss=2.262, over 5606797.50 frames. , ppl: 9.606650176778434], batch size: 70
+2022-12-14 07:40:11,795 INFO [train.py:421] (4/8) Epoch 11, batch 46800, loss[loss=2.231, over 3570.00 frames. , ppl: 9.309110550889606] tot_loss[loss=2.263, over 5587613.03 frames. , ppl: 9.612600852477737], batch size: 70
+2022-12-14 07:41:53,103 INFO [train.py:421] (4/8) Epoch 11, batch 47000, loss[loss=2.945, over 560.00 frames. , ppl: 19.00439604061661] tot_loss[loss=2.263, over 5595636.51 frames. , ppl: 9.61038662533338], batch size: 70
+2022-12-14 07:41:53,104 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:41:53,854 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.593797929123168
+2022-12-14 07:43:37,129 INFO [train.py:421] (4/8) Epoch 11, batch 47200, loss[loss=2.186, over 3430.00 frames. , ppl: 8.895230835709848] tot_loss[loss=2.262, over 5632586.14 frames. , ppl: 9.600159372700757], batch size: 70
+2022-12-14 07:45:10,242 INFO [train.py:421] (4/8) Epoch 11, batch 47400, loss[loss=2.349, over 2030.00 frames. , ppl: 10.472226109888116] tot_loss[loss=2.263, over 5570603.81 frames. , ppl: 9.613767221778913], batch size: 70
+2022-12-14 07:46:47,936 INFO [train.py:421] (4/8) Epoch 11, batch 47600, loss[loss=2.336, over 1050.00 frames. , ppl: 10.33723492720723] tot_loss[loss=2.264, over 5541885.07 frames. , ppl: 9.62456210256191], batch size: 70
+2022-12-14 07:48:29,963 INFO [train.py:421] (4/8) Epoch 11, batch 47800, loss[loss=2.188, over 3780.00 frames. , ppl: 8.91568040650871] tot_loss[loss=2.264, over 5535233.45 frames. , ppl: 9.625156923425275], batch size: 70
+2022-12-14 07:50:08,691 INFO [train.py:421] (4/8) Epoch 11, batch 48000, loss[loss=2.212, over 4970.00 frames. , ppl: 9.134482324812712] tot_loss[loss=2.264, over 5539830.62 frames. , ppl: 9.625497802838886], batch size: 70
+2022-12-14 07:50:08,692 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:50:09,455 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.597875482483978
+2022-12-14 07:51:49,511 INFO [train.py:421] (4/8) Epoch 11, batch 48200, loss[loss=2.16, over 7280.00 frames. , ppl: 8.667289336751145] tot_loss[loss=2.264, over 5542818.01 frames. , ppl: 9.617739528105153], batch size: 70
+2022-12-14 07:53:28,410 INFO [train.py:421] (4/8) Epoch 11, batch 48400, loss[loss=2.276, over 3640.00 frames. , ppl: 9.737057841916972] tot_loss[loss=2.263, over 5576048.24 frames. , ppl: 9.610336566702031], batch size: 70
+2022-12-14 07:55:09,671 INFO [train.py:421] (4/8) Epoch 11, batch 48600, loss[loss=2.279, over 1260.00 frames. , ppl: 9.768093670259994] tot_loss[loss=2.263, over 5587694.24 frames. , ppl: 9.60816867188869], batch size: 70
+2022-12-14 07:56:48,795 INFO [train.py:421] (4/8) Epoch 11, batch 48800, loss[loss=3.651, over 420.00 frames. , ppl: 38.505474749803035] tot_loss[loss=2.261, over 5618878.24 frames. , ppl: 9.593046738955138], batch size: 70
+2022-12-14 07:58:30,715 INFO [train.py:421] (4/8) Epoch 11, batch 49000, loss[loss=2.271, over 1820.00 frames. , ppl: 9.685057889987966] tot_loss[loss=2.26, over 5650369.04 frames. , ppl: 9.587793102268096], batch size: 70
+2022-12-14 07:58:30,715 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 07:58:31,475 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.595209466935325
+2022-12-14 08:00:11,580 INFO [train.py:421] (4/8) Epoch 11, batch 49200, loss[loss=2.315, over 3010.00 frames. , ppl: 10.125370992236775] tot_loss[loss=2.26, over 5666395.24 frames. , ppl: 9.582658742152553], batch size: 70
+2022-12-14 08:01:52,649 INFO [train.py:421] (4/8) Epoch 11, batch 49400, loss[loss=2.49, over 1050.00 frames. , ppl: 12.056988914942169] tot_loss[loss=2.262, over 5621730.24 frames. , ppl: 9.600948452329515], batch size: 70
+2022-12-14 08:03:32,887 INFO [train.py:421] (4/8) Epoch 11, batch 49600, loss[loss=3.699, over 420.00 frames. , ppl: 40.41675095520453] tot_loss[loss=2.262, over 5606234.81 frames. , ppl: 9.601888442374692], batch size: 70
+2022-12-14 08:05:11,621 INFO [train.py:421] (4/8) Epoch 11, batch 49800, loss[loss=2.235, over 4200.00 frames. , ppl: 9.346368846681807] tot_loss[loss=2.263, over 5612705.12 frames. , ppl: 9.609050459454409], batch size: 70
+2022-12-14 08:06:50,133 INFO [train.py:421] (4/8) Epoch 11, batch 50000, loss[loss=2.851, over 700.00 frames. , ppl: 17.30994842183007] tot_loss[loss=2.264, over 5611750.86 frames. , ppl: 9.616870718187112], batch size: 70
+2022-12-14 08:06:50,134 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:06:50,881 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.581751511481997
+2022-12-14 08:08:32,386 INFO [train.py:421] (4/8) Epoch 11, batch 50200, loss[loss=2.53, over 910.00 frames. , ppl: 12.558819167724948] tot_loss[loss=2.266, over 5557337.53 frames. , ppl: 9.636875396613855], batch size: 70
+2022-12-14 08:10:12,767 INFO [train.py:421] (4/8) Epoch 11, batch 50400, loss[loss=2.308, over 2660.00 frames. , ppl: 10.058047955308826] tot_loss[loss=2.265, over 5554417.62 frames. , ppl: 9.632297132139762], batch size: 70
+2022-12-14 08:11:48,457 INFO [train.py:421] (4/8) Epoch 11, batch 50600, loss[loss=2.269, over 2310.00 frames. , ppl: 9.669125918252139] tot_loss[loss=2.265, over 5560635.80 frames. , ppl: 9.633100753738972], batch size: 70
+2022-12-14 08:13:30,528 INFO [train.py:421] (4/8) Epoch 11, batch 50800, loss[loss=2.244, over 3010.00 frames. , ppl: 9.42727347958101] tot_loss[loss=2.265, over 5570652.29 frames. , ppl: 9.6329096053755], batch size: 70
+2022-12-14 08:15:09,636 INFO [train.py:421] (4/8) Epoch 11, batch 51000, loss[loss=5.017, over 280.00 frames. , ppl: 150.99290370836746] tot_loss[loss=2.265, over 5577040.17 frames. , ppl: 9.628490075890225], batch size: 70
+2022-12-14 08:15:09,636 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:15:10,400 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.599885781441973
+2022-12-14 08:16:50,488 INFO [train.py:421] (4/8) Epoch 11, batch 51200, loss[loss=2.172, over 2800.00 frames. , ppl: 8.779688099942735] tot_loss[loss=2.266, over 5537953.19 frames. , ppl: 9.640485059579202], batch size: 70
+2022-12-14 08:18:32,233 INFO [train.py:421] (4/8) Epoch 11, batch 51400, loss[loss=2.363, over 1330.00 frames. , ppl: 10.617477547299991] tot_loss[loss=2.267, over 5503595.27 frames. , ppl: 9.649404499016079], batch size: 70
+2022-12-14 08:20:10,263 INFO [train.py:421] (4/8) Epoch 11, batch 51600, loss[loss=3.074, over 560.00 frames. , ppl: 21.63671906829707] tot_loss[loss=2.267, over 5492752.10 frames. , ppl: 9.652207766163594], batch size: 70
+2022-12-14 08:21:49,055 INFO [train.py:421] (4/8) Epoch 11, batch 51800, loss[loss=2.87, over 560.00 frames. , ppl: 17.642369107844363] tot_loss[loss=2.267, over 5513098.76 frames. , ppl: 9.650319730321636], batch size: 70
+2022-12-14 08:23:27,648 INFO [train.py:421] (4/8) Epoch 11, batch 52000, loss[loss=2.228, over 4060.00 frames. , ppl: 9.281645884509027] tot_loss[loss=2.266, over 5521632.20 frames. , ppl: 9.640644809900843], batch size: 70
+2022-12-14 08:23:27,649 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:23:28,401 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.587526624273279
+2022-12-14 08:25:07,202 INFO [train.py:421] (4/8) Epoch 11, batch 52200, loss[loss=2.404, over 1400.00 frames. , ppl: 11.063587595249336] tot_loss[loss=2.267, over 5500355.12 frames. , ppl: 9.645982110209877], batch size: 70
+2022-12-14 08:26:45,661 INFO [train.py:421] (4/8) Epoch 11, batch 52400, loss[loss=2.197, over 7210.00 frames. , ppl: 8.996103737404574] tot_loss[loss=2.265, over 5530771.68 frames. , ppl: 9.633427153937498], batch size: 70
+2022-12-14 08:28:22,605 INFO [train.py:421] (4/8) Epoch 11, batch 52600, loss[loss=2.567, over 1190.00 frames. , ppl: 13.026803084315144] tot_loss[loss=2.266, over 5515389.93 frames. , ppl: 9.63642118597488], batch size: 70
+2022-12-14 08:30:00,798 INFO [train.py:421] (4/8) Epoch 11, batch 52800, loss[loss=3.224, over 490.00 frames. , ppl: 25.132960095442034] tot_loss[loss=2.265, over 5532183.49 frames. , ppl: 9.632680056340064], batch size: 70
+2022-12-14 08:31:42,278 INFO [train.py:421] (4/8) Epoch 11, batch 53000, loss[loss=2.157, over 4200.00 frames. , ppl: 8.643594262231106] tot_loss[loss=2.265, over 5526554.57 frames. , ppl: 9.635709390982182], batch size: 70
+2022-12-14 08:31:42,278 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:31:43,029 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.581084994414798
+2022-12-14 08:33:24,598 INFO [train.py:421] (4/8) Epoch 11, batch 53200, loss[loss=2.461, over 980.00 frames. , ppl: 11.713478584539148] tot_loss[loss=2.266, over 5527211.19 frames. , ppl: 9.636336379025435], batch size: 70
+2022-12-14 08:35:06,681 INFO [train.py:421] (4/8) Epoch 11, batch 53400, loss[loss=2.229, over 2450.00 frames. , ppl: 9.295003900581753] tot_loss[loss=2.265, over 5522230.88 frames. , ppl: 9.630907785006787], batch size: 70
+2022-12-14 08:36:47,426 INFO [train.py:421] (4/8) Epoch 11, batch 53600, loss[loss=2.629, over 770.00 frames. , ppl: 13.85785239302133] tot_loss[loss=2.265, over 5511714.09 frames. , ppl: 9.62970859235077], batch size: 70
+2022-12-14 08:38:27,754 INFO [train.py:421] (4/8) Epoch 11, batch 53800, loss[loss=2.244, over 2310.00 frames. , ppl: 9.426876706863291] tot_loss[loss=2.265, over 5517375.36 frames. , ppl: 9.628463543136162], batch size: 70
+2022-12-14 08:40:09,517 INFO [train.py:421] (4/8) Epoch 11, batch 54000, loss[loss=2.361, over 2100.00 frames. , ppl: 10.60683944214576] tot_loss[loss=2.265, over 5540386.25 frames. , ppl: 9.62637265558601], batch size: 70
+2022-12-14 08:40:09,518 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:40:10,265 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.59367723381749
+2022-12-14 08:41:46,162 INFO [train.py:421] (4/8) Epoch 11, batch 54200, loss[loss=2.237, over 2800.00 frames. , ppl: 9.365554783324123] tot_loss[loss=2.266, over 5501919.68 frames. , ppl: 9.638662256342853], batch size: 70
+2022-12-14 08:43:25,433 INFO [train.py:421] (4/8) Epoch 11, batch 54400, loss[loss=2.337, over 1190.00 frames. , ppl: 10.346307683199464] tot_loss[loss=2.266, over 5474114.35 frames. , ppl: 9.644665447318287], batch size: 70
+2022-12-14 08:45:04,356 INFO [train.py:421] (4/8) Epoch 11, batch 54600, loss[loss=2.244, over 1540.00 frames. , ppl: 9.430649503535367] tot_loss[loss=2.269, over 5412059.84 frames. , ppl: 9.669590830261555], batch size: 70
+2022-12-14 08:46:40,893 INFO [train.py:421] (4/8) Epoch 11, batch 54800, loss[loss=2.257, over 1960.00 frames. , ppl: 9.556855012752553] tot_loss[loss=2.268, over 5434386.53 frames. , ppl: 9.656988109632413], batch size: 70
+2022-12-14 08:48:23,510 INFO [train.py:421] (4/8) Epoch 11, batch 55000, loss[loss=2.359, over 2380.00 frames. , ppl: 10.580732296936958] tot_loss[loss=2.267, over 5462973.04 frames. , ppl: 9.646456316401578], batch size: 70
+2022-12-14 08:48:23,511 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:48:24,275 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.600448456413142
+2022-12-14 08:50:04,126 INFO [train.py:421] (4/8) Epoch 11, batch 55200, loss[loss=2.41, over 2100.00 frames. , ppl: 11.136920550016434] tot_loss[loss=2.266, over 5462904.70 frames. , ppl: 9.642546413009491], batch size: 70
+2022-12-14 08:51:43,844 INFO [train.py:421] (4/8) Epoch 11, batch 55400, loss[loss=2.327, over 2100.00 frames. , ppl: 10.249406685508834] tot_loss[loss=2.265, over 5480005.48 frames. , ppl: 9.634813755987329], batch size: 70
+2022-12-14 08:53:24,613 INFO [train.py:421] (4/8) Epoch 11, batch 55600, loss[loss=2.394, over 2800.00 frames. , ppl: 10.958890976995342] tot_loss[loss=2.265, over 5494008.27 frames. , ppl: 9.631324236959763], batch size: 70
+2022-12-14 08:55:02,642 INFO [train.py:421] (4/8) Epoch 11, batch 55800, loss[loss=2.216, over 5180.00 frames. , ppl: 9.16989719645389] tot_loss[loss=2.265, over 5519958.07 frames. , ppl: 9.626912342271662], batch size: 70
+2022-12-14 08:56:44,614 INFO [train.py:421] (4/8) Epoch 11, batch 56000, loss[loss=2.376, over 1330.00 frames. , ppl: 10.764552040756273] tot_loss[loss=2.263, over 5547332.13 frames. , ppl: 9.61450717745263], batch size: 70
+2022-12-14 08:56:44,615 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 08:56:45,361 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.589785979049779
+2022-12-14 08:58:29,520 INFO [train.py:421] (4/8) Epoch 11, batch 56200, loss[loss=2.156, over 5250.00 frames. , ppl: 8.636114344191839] tot_loss[loss=2.265, over 5505691.89 frames. , ppl: 9.6270290498217], batch size: 70
+2022-12-14 09:00:08,806 INFO [train.py:421] (4/8) Epoch 11, batch 56400, loss[loss=2.41, over 910.00 frames. , ppl: 11.133807012269315] tot_loss[loss=2.265, over 5512176.10 frames. , ppl: 9.629031076879157], batch size: 70
+2022-12-14 09:01:50,882 INFO [train.py:421] (4/8) Epoch 11, batch 56600, loss[loss=2.168, over 12040.00 frames. , ppl: 8.736736825118689] tot_loss[loss=2.263, over 5566208.92 frames. , ppl: 9.615098612023536], batch size: 70
+2022-12-14 09:03:29,224 INFO [train.py:421] (4/8) Epoch 11, batch 56800, loss[loss=2.139, over 4410.00 frames. , ppl: 8.488968849752089] tot_loss[loss=2.264, over 5524563.19 frames. , ppl: 9.625481822815004], batch size: 70
+2022-12-14 09:05:09,670 INFO [train.py:421] (4/8) Epoch 11, batch 57000, loss[loss=2.832, over 700.00 frames. , ppl: 16.973622221945902] tot_loss[loss=2.265, over 5524772.74 frames. , ppl: 9.630003275167175], batch size: 70
+2022-12-14 09:05:09,670 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:05:10,418 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.587563519007391
+2022-12-14 09:06:48,000 INFO [train.py:421] (4/8) Epoch 11, batch 57200, loss[loss=2.226, over 3570.00 frames. , ppl: 9.267263118270318] tot_loss[loss=2.266, over 5485779.80 frames. , ppl: 9.645305893953452], batch size: 70
+2022-12-14 09:08:32,480 INFO [train.py:421] (4/8) Epoch 11, batch 57400, loss[loss=2.369, over 1260.00 frames. , ppl: 10.689159993479803] tot_loss[loss=2.266, over 5470275.21 frames. , ppl: 9.644825118597181], batch size: 70
+2022-12-14 09:10:14,445 INFO [train.py:421] (4/8) Epoch 11, batch 57600, loss[loss=2.286, over 2380.00 frames. , ppl: 9.837491212942949] tot_loss[loss=2.266, over 5498318.22 frames. , ppl: 9.636686337185958], batch size: 70
+2022-12-14 09:11:54,086 INFO [train.py:421] (4/8) Epoch 11, batch 57800, loss[loss=2.312, over 2520.00 frames. , ppl: 10.093116013385853] tot_loss[loss=2.265, over 5506092.24 frames. , ppl: 9.634882806019112], batch size: 70
+2022-12-14 09:13:32,728 INFO [train.py:421] (4/8) Epoch 11, batch 58000, loss[loss=2.275, over 4760.00 frames. , ppl: 9.729236317667791] tot_loss[loss=2.264, over 5548929.47 frames. , ppl: 9.624958534937663], batch size: 70
+2022-12-14 09:13:32,729 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:13:33,485 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.579467110769087
+2022-12-14 09:15:15,192 INFO [train.py:421] (4/8) Epoch 11, batch 58200, loss[loss=2.349, over 2100.00 frames. , ppl: 10.472013180012853] tot_loss[loss=2.263, over 5583863.61 frames. , ppl: 9.612900625710536], batch size: 70
+2022-12-14 09:16:56,029 INFO [train.py:421] (4/8) Epoch 11, batch 58400, loss[loss=2.561, over 700.00 frames. , ppl: 12.943218546300663] tot_loss[loss=2.263, over 5558343.91 frames. , ppl: 9.614293452127523], batch size: 70
+2022-12-14 09:18:34,148 INFO [train.py:421] (4/8) Epoch 11, batch 58600, loss[loss=2.197, over 2240.00 frames. , ppl: 8.99568950102928] tot_loss[loss=2.264, over 5570521.14 frames. , ppl: 9.618614517820316], batch size: 70
+2022-12-14 09:20:16,267 INFO [train.py:421] (4/8) Epoch 11, batch 58800, loss[loss=2.155, over 9030.00 frames. , ppl: 8.62687915732911] tot_loss[loss=2.264, over 5585208.43 frames. , ppl: 9.618760345277549], batch size: 70
+2022-12-14 09:21:58,265 INFO [train.py:421] (4/8) Epoch 11, batch 59000, loss[loss=3.112, over 560.00 frames. , ppl: 22.469529904925924] tot_loss[loss=2.262, over 5630721.17 frames. , ppl: 9.605025408289805], batch size: 70
+2022-12-14 09:21:58,266 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:21:59,025 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.585608293732157
+2022-12-14 09:23:40,364 INFO [train.py:421] (4/8) Epoch 11, batch 59200, loss[loss=2.495, over 1190.00 frames. , ppl: 12.125522152257178] tot_loss[loss=2.262, over 5635231.83 frames. , ppl: 9.605067016924144], batch size: 70
+2022-12-14 09:25:21,847 INFO [train.py:421] (4/8) Epoch 11, batch 59400, loss[loss=2.184, over 3150.00 frames. , ppl: 8.879963379236434] tot_loss[loss=2.262, over 5627727.38 frames. , ppl: 9.600971389306403], batch size: 70
+2022-12-14 09:27:00,559 INFO [train.py:421] (4/8) Epoch 11, batch 59600, loss[loss=2.399, over 1750.00 frames. , ppl: 11.010782280202067] tot_loss[loss=2.263, over 5573459.83 frames. , ppl: 9.610023709446923], batch size: 70
+2022-12-14 09:28:41,363 INFO [train.py:421] (4/8) Epoch 11, batch 59800, loss[loss=2.27, over 1960.00 frames. , ppl: 9.67586254575352] tot_loss[loss=2.263, over 5545338.23 frames. , ppl: 9.614737119211359], batch size: 70
+2022-12-14 09:30:19,036 INFO [train.py:421] (4/8) Epoch 11, batch 60000, loss[loss=2.234, over 6790.00 frames. , ppl: 9.337037122628363] tot_loss[loss=2.263, over 5551880.43 frames. , ppl: 9.613022941116103], batch size: 70
+2022-12-14 09:30:19,037 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:30:19,782 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.57757590873942
+2022-12-14 09:32:01,996 INFO [train.py:421] (4/8) Epoch 11, batch 60200, loss[loss=2.345, over 2660.00 frames. , ppl: 10.429160503535499] tot_loss[loss=2.265, over 5521820.36 frames. , ppl: 9.626486057060413], batch size: 70
+2022-12-14 09:33:42,601 INFO [train.py:421] (4/8) Epoch 11, batch 60400, loss[loss=2.146, over 8120.00 frames. , ppl: 8.548662207311901] tot_loss[loss=2.266, over 5501351.88 frames. , ppl: 9.637596391286712], batch size: 70
+2022-12-14 09:35:24,852 INFO [train.py:421] (4/8) Epoch 11, batch 60600, loss[loss=2.298, over 3360.00 frames. , ppl: 9.958722846389842] tot_loss[loss=2.265, over 5519512.13 frames. , ppl: 9.629623679995483], batch size: 70
+2022-12-14 09:37:07,335 INFO [train.py:421] (4/8) Epoch 11, batch 60800, loss[loss=2.238, over 2800.00 frames. , ppl: 9.378047465590358] tot_loss[loss=2.265, over 5521028.48 frames. , ppl: 9.626539752349153], batch size: 70
+2022-12-14 09:38:49,724 INFO [train.py:421] (4/8) Epoch 11, batch 61000, loss[loss=2.354, over 2450.00 frames. , ppl: 10.529368233077303] tot_loss[loss=2.266, over 5497337.34 frames. , ppl: 9.63611240722688], batch size: 70
+2022-12-14 09:38:49,724 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:38:50,484 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.587292487923966
+2022-12-14 09:40:29,790 INFO [train.py:421] (4/8) Epoch 11, batch 61200, loss[loss=2.53, over 1330.00 frames. , ppl: 12.547441561157733] tot_loss[loss=2.265, over 5530126.17 frames. , ppl: 9.627824869460127], batch size: 70
+2022-12-14 09:42:09,350 INFO [train.py:421] (4/8) Epoch 11, batch 61400, loss[loss=6.221, over 210.00 frames. , ppl: 503.06851647942807] tot_loss[loss=2.265, over 5498565.54 frames. , ppl: 9.635050201794485], batch size: 70
+2022-12-14 09:43:48,639 INFO [train.py:421] (4/8) Epoch 11, batch 61600, loss[loss=2.565, over 910.00 frames. , ppl: 12.99554471929638] tot_loss[loss=2.267, over 5471382.31 frames. , ppl: 9.646831781826242], batch size: 70
+2022-12-14 09:45:23,760 INFO [train.py:421] (4/8) Epoch 11, batch 61800, loss[loss=2.206, over 2800.00 frames. , ppl: 9.07882699417901] tot_loss[loss=2.268, over 5417836.72 frames. , ppl: 9.658467780323669], batch size: 70
+2022-12-14 09:47:03,482 INFO [train.py:421] (4/8) Epoch 11, batch 62000, loss[loss=2.16, over 7840.00 frames. , ppl: 8.6675961177803] tot_loss[loss=2.269, over 5380330.43 frames. , ppl: 9.670869898555093], batch size: 70
+2022-12-14 09:47:03,483 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:47:04,244 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.582718752746382
+2022-12-14 09:48:44,734 INFO [train.py:421] (4/8) Epoch 11, batch 62200, loss[loss=2.179, over 3500.00 frames. , ppl: 8.839082093791191] tot_loss[loss=2.269, over 5389310.04 frames. , ppl: 9.670368957258205], batch size: 70
+2022-12-14 09:50:24,715 INFO [train.py:421] (4/8) Epoch 11, batch 62400, loss[loss=2.287, over 2520.00 frames. , ppl: 9.842743121795046] tot_loss[loss=2.271, over 5344148.74 frames. , ppl: 9.68540512227472], batch size: 70
+2022-12-14 09:52:03,658 INFO [train.py:421] (4/8) Epoch 11, batch 62600, loss[loss=2.524, over 770.00 frames. , ppl: 12.474367431292949] tot_loss[loss=2.27, over 5377538.53 frames. , ppl: 9.684141315166627], batch size: 70
+2022-12-14 09:53:42,469 INFO [train.py:421] (4/8) Epoch 11, batch 62800, loss[loss=2.383, over 1540.00 frames. , ppl: 10.839547185737032] tot_loss[loss=2.27, over 5363094.99 frames. , ppl: 9.683669109067809], batch size: 70
+2022-12-14 09:55:20,749 INFO [train.py:421] (4/8) Epoch 11, batch 63000, loss[loss=2.233, over 2940.00 frames. , ppl: 9.32941117162211] tot_loss[loss=2.271, over 5335136.03 frames. , ppl: 9.687931876853725], batch size: 70
+2022-12-14 09:55:20,750 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 09:55:21,511 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.60109500498798
+2022-12-14 09:57:02,141 INFO [train.py:421] (4/8) Epoch 11, batch 63200, loss[loss=2.308, over 3080.00 frames. , ppl: 10.055160338126973] tot_loss[loss=2.27, over 5364046.76 frames. , ppl: 9.679267388399278], batch size: 70
+2022-12-14 09:58:42,040 INFO [train.py:421] (4/8) Epoch 11, batch 63400, loss[loss=2.368, over 1260.00 frames. , ppl: 10.679531380115977] tot_loss[loss=2.27, over 5386238.29 frames. , ppl: 9.678285203394598], batch size: 70
+2022-12-14 10:00:19,148 INFO [train.py:421] (4/8) Epoch 11, batch 63600, loss[loss=2.312, over 2660.00 frames. , ppl: 10.09667265896626] tot_loss[loss=2.269, over 5410472.55 frames. , ppl: 9.673529524131482], batch size: 70
+2022-12-14 10:02:00,367 INFO [train.py:421] (4/8) Epoch 11, batch 63800, loss[loss=2.226, over 3640.00 frames. , ppl: 9.26060435136936] tot_loss[loss=2.269, over 5448160.32 frames. , ppl: 9.671844096533158], batch size: 70
+2022-12-14 10:03:38,191 INFO [train.py:421] (4/8) Epoch 11, batch 64000, loss[loss=2.735, over 770.00 frames. , ppl: 15.403963521149612] tot_loss[loss=2.27, over 5442448.74 frames. , ppl: 9.679346365984046], batch size: 70
+2022-12-14 10:03:38,191 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:03:38,952 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.263, over 211138.00 frames. , ppl: 9.61195926586509
+2022-12-14 10:05:19,744 INFO [train.py:421] (4/8) Epoch 11, batch 64200, loss[loss=2.36, over 2100.00 frames. , ppl: 10.587238572245209] tot_loss[loss=2.269, over 5494315.63 frames. , ppl: 9.671145611134484], batch size: 70
+2022-12-14 10:07:02,734 INFO [train.py:421] (4/8) Epoch 11, batch 64400, loss[loss=2.404, over 1680.00 frames. , ppl: 11.069953342561204] tot_loss[loss=2.27, over 5466545.57 frames. , ppl: 9.675538483983592], batch size: 70
+2022-12-14 10:08:42,830 INFO [train.py:421] (4/8) Epoch 11, batch 64600, loss[loss=2.164, over 4480.00 frames. , ppl: 8.703179700336497] tot_loss[loss=2.268, over 5489087.30 frames. , ppl: 9.661871611443555], batch size: 70
+2022-12-14 10:10:23,651 INFO [train.py:421] (4/8) Epoch 11, batch 64800, loss[loss=2.341, over 1890.00 frames. , ppl: 10.39029276692011] tot_loss[loss=2.268, over 5509607.48 frames. , ppl: 9.656141334340191], batch size: 70
+2022-12-14 10:12:04,659 INFO [train.py:421] (4/8) Epoch 11, batch 65000, loss[loss=2.214, over 2660.00 frames. , ppl: 9.153331729403604] tot_loss[loss=2.268, over 5494101.36 frames. , ppl: 9.662846105895316], batch size: 70
+2022-12-14 10:12:04,659 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:12:05,419 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.588153854062595
+2022-12-14 10:13:48,149 INFO [train.py:421] (4/8) Epoch 11, batch 65200, loss[loss=2.252, over 3080.00 frames. , ppl: 9.50632235516146] tot_loss[loss=2.269, over 5454796.76 frames. , ppl: 9.667818245151793], batch size: 70
+2022-12-14 10:15:30,918 INFO [train.py:421] (4/8) Epoch 11, batch 65400, loss[loss=2.178, over 5600.00 frames. , ppl: 8.826351179152306] tot_loss[loss=2.267, over 5493614.98 frames. , ppl: 9.64853104146047], batch size: 70
+2022-12-14 10:17:13,918 INFO [train.py:421] (4/8) Epoch 11, batch 65600, loss[loss=2.33, over 1680.00 frames. , ppl: 10.272992307477551] tot_loss[loss=2.266, over 5526136.05 frames. , ppl: 9.636840441288406], batch size: 70
+2022-12-14 10:18:52,817 INFO [train.py:421] (4/8) Epoch 11, batch 65800, loss[loss=2.417, over 2030.00 frames. , ppl: 11.21505164998098] tot_loss[loss=2.266, over 5492223.07 frames. , ppl: 9.645086593397064], batch size: 70
+2022-12-14 10:20:34,406 INFO [train.py:421] (4/8) Epoch 11, batch 66000, loss[loss=2.31, over 1750.00 frames. , ppl: 10.075178016273822] tot_loss[loss=2.265, over 5520031.07 frames. , ppl: 9.635213033600154], batch size: 70
+2022-12-14 10:20:34,407 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:20:35,171 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.583815172113377
+2022-12-14 10:22:18,351 INFO [train.py:421] (4/8) Epoch 11, batch 66200, loss[loss=2.237, over 2590.00 frames. , ppl: 9.366553398016668] tot_loss[loss=2.266, over 5505022.06 frames. , ppl: 9.640981139887836], batch size: 70
+2022-12-14 10:24:06,459 INFO [train.py:421] (4/8) Epoch 11, batch 66400, loss[loss=2.222, over 6790.00 frames. , ppl: 9.222167316953973] tot_loss[loss=2.264, over 5567619.59 frames. , ppl: 9.621030869932587], batch size: 70
+2022-12-14 10:25:47,258 INFO [train.py:421] (4/8) Epoch 11, batch 66600, loss[loss=2.256, over 1960.00 frames. , ppl: 9.543191664070665] tot_loss[loss=2.264, over 5526817.61 frames. , ppl: 9.625784990454488], batch size: 70
+2022-12-14 10:27:31,529 INFO [train.py:421] (4/8) Epoch 11, batch 66800, loss[loss=2.382, over 2030.00 frames. , ppl: 10.822113167564048] tot_loss[loss=2.264, over 5538400.38 frames. , ppl: 9.624987119296305], batch size: 70
+2022-12-14 10:29:12,953 INFO [train.py:421] (4/8) Epoch 11, batch 67000, loss[loss=2.375, over 1120.00 frames. , ppl: 10.74799279395535] tot_loss[loss=2.265, over 5510783.59 frames. , ppl: 9.633478151772684], batch size: 70
+2022-12-14 10:29:12,953 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:29:13,717 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.262, over 211138.00 frames. , ppl: 9.59918248410303
+2022-12-14 10:30:53,795 INFO [train.py:421] (4/8) Epoch 11, batch 67200, loss[loss=2.627, over 770.00 frames. , ppl: 13.838456145974924] tot_loss[loss=2.265, over 5508967.88 frames. , ppl: 9.632960736353901], batch size: 70
+2022-12-14 10:32:32,048 INFO [train.py:421] (4/8) Epoch 11, batch 67400, loss[loss=2.34, over 3080.00 frames. , ppl: 10.376176074748395] tot_loss[loss=2.265, over 5519210.92 frames. , ppl: 9.630627967965527], batch size: 70
+2022-12-14 10:34:15,945 INFO [train.py:421] (4/8) Epoch 11, batch 67600, loss[loss=2.376, over 1120.00 frames. , ppl: 10.766754505021941] tot_loss[loss=2.266, over 5508891.79 frames. , ppl: 9.63844348982989], batch size: 70
+2022-12-14 10:35:56,743 INFO [train.py:421] (4/8) Epoch 11, batch 67800, loss[loss=2.667, over 770.00 frames. , ppl: 14.395154798819968] tot_loss[loss=2.266, over 5505255.14 frames. , ppl: 9.63648911740083], batch size: 70
+2022-12-14 10:37:37,855 INFO [train.py:421] (4/8) Epoch 11, batch 68000, loss[loss=2.265, over 2940.00 frames. , ppl: 9.62764331477629] tot_loss[loss=2.264, over 5555110.61 frames. , ppl: 9.623240364249662], batch size: 70
+2022-12-14 10:37:37,856 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:37:38,622 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.26, over 211138.00 frames. , ppl: 9.586750448742256
+2022-12-14 10:39:17,198 INFO [train.py:421] (4/8) Epoch 11, batch 68200, loss[loss=2.454, over 1190.00 frames. , ppl: 11.629044178655128] tot_loss[loss=2.266, over 5498451.48 frames. , ppl: 9.640109442790955], batch size: 70
+2022-12-14 10:40:56,989 INFO [train.py:421] (4/8) Epoch 11, batch 68400, loss[loss=2.267, over 2450.00 frames. , ppl: 9.64981914869249] tot_loss[loss=2.266, over 5486599.12 frames. , ppl: 9.644960940468176], batch size: 70
+2022-12-14 10:42:35,925 INFO [train.py:421] (4/8) Epoch 11, batch 68600, loss[loss=2.359, over 1260.00 frames. , ppl: 10.576980347131878] tot_loss[loss=2.266, over 5480530.52 frames. , ppl: 9.641187046296562], batch size: 70
+2022-12-14 10:44:17,967 INFO [train.py:421] (4/8) Epoch 11, batch 68800, loss[loss=2.267, over 3010.00 frames. , ppl: 9.65341520774021] tot_loss[loss=2.265, over 5511445.27 frames. , ppl: 9.632797443501119], batch size: 70
+2022-12-14 10:46:02,186 INFO [train.py:421] (4/8) Epoch 11, batch 69000, loss[loss=2.178, over 4060.00 frames. , ppl: 8.825723953552446] tot_loss[loss=2.265, over 5522528.56 frames. , ppl: 9.628891490726414], batch size: 70
+2022-12-14 10:46:02,187 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:46:02,935 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.261, over 211138.00 frames. , ppl: 9.590407678839963
+2022-12-14 10:47:43,618 INFO [train.py:421] (4/8) Epoch 11, batch 69200, loss[loss=2.207, over 5180.00 frames. , ppl: 9.089941719035842] tot_loss[loss=2.264, over 5560076.03 frames. , ppl: 9.617393639712772], batch size: 70
+2022-12-14 10:49:26,940 INFO [train.py:421] (4/8) Epoch 11, batch 69400, loss[loss=2.212, over 4200.00 frames. , ppl: 9.130938906066833] tot_loss[loss=2.262, over 5581737.92 frames. , ppl: 9.604821032746623], batch size: 70
+2022-12-14 10:51:09,341 INFO [train.py:421] (4/8) Epoch 11, batch 69600, loss[loss=2.3, over 4060.00 frames. , ppl: 9.970047235606074] tot_loss[loss=2.262, over 5568813.69 frames. , ppl: 9.606430244782745], batch size: 70
+2022-12-14 10:52:54,746 INFO [train.py:421] (4/8) Epoch 11, batch 69800, loss[loss=2.322, over 1820.00 frames. , ppl: 10.194257404183498] tot_loss[loss=2.262, over 5570963.47 frames. , ppl: 9.605237472876343], batch size: 70
+2022-12-14 10:54:39,544 INFO [train.py:421] (4/8) Epoch 11, batch 70000, loss[loss=2.35, over 3220.00 frames. , ppl: 10.48197686524722] tot_loss[loss=2.264, over 5520020.60 frames. , ppl: 9.617138089999196], batch size: 70
+2022-12-14 10:54:39,545 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 10:54:40,301 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.575361947313867
+2022-12-14 10:56:19,540 INFO [train.py:421] (4/8) Epoch 11, batch 70200, loss[loss=2.181, over 5390.00 frames. , ppl: 8.8552722490974] tot_loss[loss=2.265, over 5485194.13 frames. , ppl: 9.629339645161192], batch size: 70
+2022-12-14 10:58:03,412 INFO [train.py:421] (4/8) Epoch 11, batch 70400, loss[loss=2.738, over 630.00 frames. , ppl: 15.45899379641327] tot_loss[loss=2.265, over 5489749.51 frames. , ppl: 9.628744991187094], batch size: 70
+2022-12-14 10:59:41,663 INFO [train.py:421] (4/8) Epoch 11, batch 70600, loss[loss=2.458, over 840.00 frames. , ppl: 11.687113133088708] tot_loss[loss=2.266, over 5460457.52 frames. , ppl: 9.64136726572544], batch size: 70
+2022-12-14 11:01:20,631 INFO [train.py:421] (4/8) Epoch 11, batch 70800, loss[loss=2.151, over 6090.00 frames. , ppl: 8.597474797263475] tot_loss[loss=2.266, over 5462338.64 frames. , ppl: 9.640689723883545], batch size: 70
+2022-12-14 11:03:04,092 INFO [train.py:421] (4/8) Epoch 11, batch 71000, loss[loss=2.279, over 1680.00 frames. , ppl: 9.765499137887872] tot_loss[loss=2.265, over 5512730.63 frames. , ppl: 9.632268183638114], batch size: 70
+2022-12-14 11:03:04,093 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 11:03:04,885 INFO [train.py:452] (4/8) Epoch 11, validation: loss=2.259, over 211138.00 frames. , ppl: 9.57697205031597
+2022-12-14 11:04:44,372 INFO [train.py:421] (4/8) Epoch 11, batch 71200, loss[loss=2.249, over 5740.00 frames. , ppl: 9.478465316369233] tot_loss[loss=2.266, over 5510179.04 frames. , ppl: 9.640039592601582], batch size: 70
+2022-12-14 11:06:25,536 INFO [train.py:421] (4/8) Epoch 11, batch 71400, loss[loss=2.531, over 1050.00 frames. , ppl: 12.566472642419878] tot_loss[loss=2.267, over 5505062.09 frames. , ppl: 9.646041604988305], batch size: 70
+2022-12-14 11:08:07,128 INFO [train.py:421] (4/8) Epoch 11, batch 71600, loss[loss=2.225, over 5250.00 frames. , ppl: 9.249937691764528] tot_loss[loss=2.266, over 5510381.22 frames. , ppl: 9.645422911881434], batch size: 70
+2022-12-14 11:09:48,360 INFO [train.py:421] (4/8) Epoch 11, batch 71800, loss[loss=2.261, over 2800.00 frames. , ppl: 9.593276944711903] tot_loss[loss=2.267, over 5491543.68 frames. , ppl: 9.647189068800111], batch size: 70
+2022-12-14 11:11:04,452 INFO [train.py:421] (4/8) Epoch 12, batch 0, loss[loss=2.198, over 5180.00 frames. , ppl: 9.010301191474325] tot_loss[loss=2.198, over 5180.00 frames. , ppl: 9.010301191474325], batch size: 70
+2022-12-14 11:12:45,932 INFO [train.py:421] (4/8) Epoch 12, batch 200, loss[loss=2.339, over 1890.00 frames. , ppl: 10.370932642345542] tot_loss[loss=2.243, over 557919.08 frames. , ppl: 9.424858312287748], batch size: 70
+2022-12-14 11:14:25,249 INFO [train.py:421] (4/8) Epoch 12, batch 400, loss[loss=2.192, over 5390.00 frames. , ppl: 8.953662502049795] tot_loss[loss=2.252, over 1010067.42 frames. , ppl: 9.511228694821863], batch size: 70
+2022-12-14 11:16:08,524 INFO [train.py:421] (4/8) Epoch 12, batch 600, loss[loss=2.29, over 3080.00 frames. , ppl: 9.872333655040267] tot_loss[loss=2.25, over 1471197.31 frames. , ppl: 9.490102452270692], batch size: 70
+2022-12-14 11:17:49,309 INFO [train.py:421] (4/8) Epoch 12, batch 800, loss[loss=2.236, over 4900.00 frames. , ppl: 9.352487390282207] tot_loss[loss=2.251, over 1879163.14 frames. , ppl: 9.497678470064258], batch size: 70
+2022-12-14 11:19:30,219 INFO [train.py:421] (4/8) Epoch 12, batch 1000, loss[loss=2.108, over 5250.00 frames. , ppl: 8.231479550539454] tot_loss[loss=2.252, over 2242006.07 frames. , ppl: 9.506044717530443], batch size: 70
+2022-12-14 11:19:30,219 INFO [train.py:441] (4/8) Computing validation loss
+2022-12-14 11:19:30,968 INFO [train.py:452] (4/8) Epoch 12, validation: loss=2.26, over 211138.00 frames. , ppl: 9.584870576501038
+2022-12-14 11:21:11,007 INFO [train.py:421] (4/8) Epoch 12, batch 1200, loss[loss=2.63, over 1120.00 frames. , ppl: 13.874170317526984] tot_loss[loss=2.256, over 2539748.75 frames. , ppl: 9.541549585672328], batch size: 70