2022-12-13 11:05:28,709 INFO [test.py:98] Computing perplexity started
2022-12-13 11:05:28,709 INFO [test.py:99] {'max_sent_len': 100, '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': '1.18', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': '2d82a1d9350263ae48a6953034ce570e3d5208c1', 'k2-git-date': 'Mon Aug 15 02:09:05 2022', 'lhotse-version': '1.5.0', 'torch-version': '1.10.1', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.9', 'icefall-git-branch': 'transformer_lm', 'icefall-git-sha1': '7131129-dirty', 'icefall-git-date': 'Fri Dec 9 15:46:21 2022', 'icefall-path': '/ceph-data4/yangxiaoyu/softwares/icefall_development/icefall_transformer_lm', 'k2-path': '/ceph-data4/yangxiaoyu/softwares/anaconda3/envs/k2_test/lib/python3.9/site-packages/k2/__init__.py', 'lhotse-path': '/ceph-data4/yangxiaoyu/softwares/anaconda3/envs/k2_test/lib/python3.9/site-packages/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-2-0602201035-5fb6d86964-mclm7', 'IP address': '10.177.74.202'}, 'epoch': 8, 'avg': 2, 'exp_dir': PosixPath('transformer_lm/exp_full_libri_16layer_maxlen200_8gpu'), 'lm_data': PosixPath('transformer_lm/libri_lm_training_bpe500/sorted_lm_data-test.pt'), 'vocab_size': 500, 'num_layers': 16, 'tie_weights': False, 'batch_size': 50}
2022-12-13 11:05:28,710 INFO [test.py:105] Device: cuda:0
2022-12-13 11:05:28,710 INFO [test.py:107] About to create model
2022-12-13 11:05:29,147 INFO [model.py:68] Not tying weights
2022-12-13 11:05:29,147 INFO [test.py:128] averaging ['transformer_lm/exp_full_libri_16layer_maxlen200_8gpu/epoch-7.pt', 'transformer_lm/exp_full_libri_16layer_maxlen200_8gpu/epoch-8.pt']
2022-12-13 11:05:35,810 INFO [test.py:138] Number of model parameters: 98995638
2022-12-13 11:05:35,810 INFO [test.py:139] Number of model parameters (requires_grad): 98995638 (100.0%)
2022-12-13 11:05:35,811 INFO [test.py:145] Loading LM test data from transformer_lm/libri_lm_training_bpe500/sorted_lm_data-test.pt
2022-12-13 11:05:42,507 INFO [test.py:169] total nll: 476117.77613830566, num tokens: 211138, num sentences: 5559, ppl: 9.535