INFO - 09/12/18 00:13:15 - 0:00:00 - ============ Initialized logger ============ INFO - 09/12/18 00:13:15 - 0:00:00 - attention: True attention_dropout: 0 back_dataset: {} back_directions: [] batch_size: 32 beam_size: 0 clip_grad_norm: 5 command: python main.py --exp_name 'base1' --transformer 'True' --n_enc_layer '4' --n_dec_layer '4' --share_enc '3' --share_dec '3' --share_lang_emb 'True' --share_output_emb 'True' --langs 'en,fr' --n_mono '-1' --mono_dataset 'en:./data/mono/all.en.tok.60000.pth,,;fr:./data/mono/all.fr.tok.60000.pth,,' --para_dataset 'en-fr:,./data/para/dev/newstest2013-ref.XX.60000.pth,./data/para/dev/newstest2014-fren-src.XX.60000.pth' --mono_directions 'en,fr' --word_dropout '0.1' --word_blank '0.2' --pivo_directions 'fr-en-fr,en-fr-en' --pretrained_emb './data/mono/all.en-fr.60000.vec' --pretrained_out 'True' --lambda_xe_mono '0:1,100000:0.1,300000:0' --lambda_xe_otfd '1' --otf_num_processes '15' --otf_sync_params_every '1000' --enc_optimizer 'adam,lr=0.0001' --epoch_size '500000' --max_len '100' --stopping_criterion 'bleu_en_fr_valid,10' --exp_id "2un3e1yus0" dec_optimizer: enc_optimizer decoder_attention_heads: 8 decoder_normalize_before: False dis_clip: 0 dis_dropout: 0 dis_hidden_dim: 128 dis_input_proj: True dis_layers: 3 dis_optimizer: rmsprop,lr=0.0005 dis_smooth: 0 dropout: 0 dump_path: ./dumped/base1/2un3e1yus0 emb_dim: 512 enc_optimizer: adam,lr=0.0001 encoder_attention_heads: 8 encoder_normalize_before: False epoch_size: 500000 eval_only: False exp_id: 2un3e1yus0 exp_name: base1 freeze_dec_emb: False freeze_enc_emb: False group_by_size: True hidden_dim: 512 id2lang: {0: 'en', 1: 'fr'} label_smoothing: 0 lambda_dis: 0 lambda_lm: 0 lambda_xe_back: 0 lambda_xe_mono: 0:1,100000:0.1,300000:0 lambda_xe_otfa: 0 lambda_xe_otfd: 1 lambda_xe_para: 0 lang2id: {'en': 0, 'fr': 1} langs: ['en', 'fr'] length_penalty: 1.0 lm_after: 0 lm_before: 0 lm_share_dec: 0 lm_share_emb: False lm_share_enc: 0 lm_share_proj: False lstm_proj: False max_epoch: 100000 max_len: 100 max_vocab: -1 mono_dataset: {'en': ('./data/mono/all.en.tok.60000.pth', '', ''), 'fr': ('./data/mono/all.fr.tok.60000.pth', '', '')} mono_directions: ['en', 'fr'] n_back: 0 n_dec_layers: 4 n_dis: 0 n_enc_layers: 4 n_langs: 2 n_mono: -1 n_para: 0 otf_backprop_temperature: -1 otf_num_processes: 15 otf_sample: -1 otf_sync_params_every: 1000 otf_update_dec: True otf_update_enc: True para_dataset: {('en', 'fr'): ('', './data/para/dev/newstest2013-ref.XX.60000.pth', './data/para/dev/newstest2014-fren-src.XX.60000.pth')} para_directions: [] pivo_directions: [('fr', 'en', 'fr'), ('en', 'fr', 'en')] pretrained_emb: ./data/mono/all.en-fr.60000.vec pretrained_out: True reload_dec: False reload_dis: False reload_enc: False reload_model: relu_dropout: 0 save_periodic: False seed: -1 share_dec: 3 share_decpro_emb: False share_enc: 3 share_encdec_emb: False share_lang_emb: True share_lstm_proj: False share_output_emb: True stopping_criterion: bleu_en_fr_valid,10 transformer: True transformer_ffn_emb_dim: 2048 vocab: {} vocab_min_count: 0 word_blank: 0.2 word_dropout: 0.1 word_shuffle: 0 INFO - 09/12/18 00:13:15 - 0:00:00 - The experiment will be stored in ./dumped/base1/2un3e1yus0 INFO - 09/12/18 00:13:15 - 0:00:00 - Running command: python main.py --exp_name 'base1' --transformer 'True' --n_enc_layer '4' --n_dec_layer '4' --share_enc '3' --share_dec '3' --share_lang_emb 'True' --share_output_emb 'True' --langs 'en,fr' --n_mono '-1' --mono_dataset 'en:./data/mono/all.en.tok.60000.pth,,;fr:./data/mono/all.fr.tok.60000.pth,,' --para_dataset 'en-fr:,./data/para/dev/newstest2013-ref.XX.60000.pth,./data/para/dev/newstest2014-fren-src.XX.60000.pth' --mono_directions 'en,fr' --word_dropout '0.1' --word_blank '0.2' --pivo_directions 'fr-en-fr,en-fr-en' --pretrained_emb './data/mono/all.en-fr.60000.vec' --pretrained_out 'True' --lambda_xe_mono '0:1,100000:0.1,300000:0' --lambda_xe_otfd '1' --otf_num_processes '15' --otf_sync_params_every '1000' --enc_optimizer 'adam,lr=0.0001' --epoch_size '500000' --max_len '100' --stopping_criterion 'bleu_en_fr_valid,10' --exp_id "2un3e1yus0" INFO - 09/12/18 00:13:15 - 0:00:00 - ============ Parallel data (en - fr) INFO - 09/12/18 00:13:15 - 0:00:00 - Loading data from ./data/para/dev/newstest2013-ref.en.60000.pth ... INFO - 09/12/18 00:13:15 - 0:00:00 - 69880 words (60536 unique) in 3000 sentences. 1 unknown words (1 unique). INFO - 09/12/18 00:13:15 - 0:00:00 - Loading data from ./data/para/dev/newstest2013-ref.fr.60000.pth ... INFO - 09/12/18 00:13:15 - 0:00:00 - 79997 words (60536 unique) in 3000 sentences. 9 unknown words (6 unique). INFO - 09/12/18 00:13:16 - 0:00:00 - Removed 0 empty sentences. INFO - 09/12/18 00:13:16 - 0:00:00 - Removed 7 too long sentences. INFO - 09/12/18 00:13:16 - 0:00:00 - Loading data from ./data/para/dev/newstest2014-fren-src.en.60000.pth ... INFO - 09/12/18 00:13:16 - 0:00:00 - 76331 words (60536 unique) in 3003 sentences. 0 unknown words (0 unique). INFO - 09/12/18 00:13:16 - 0:00:00 - Loading data from ./data/para/dev/newstest2014-fren-src.fr.60000.pth ... INFO - 09/12/18 00:13:16 - 0:00:00 - 86843 words (60536 unique) in 3003 sentences. 0 unknown words (0 unique). INFO - 09/12/18 00:13:16 - 0:00:00 - Removed 0 empty sentences. INFO - 09/12/18 00:13:16 - 0:00:00 - ============ Monolingual data (en) INFO - 09/12/18 00:13:16 - 0:00:00 - Loading data from ./data/mono/all.en.tok.60000.pth ... INFO - 09/12/18 00:13:46 - 0:00:30 - 257504282 words (60536 unique) in 10000000 sentences. 0 unknown words (0 unique). INFO - 09/12/18 00:13:50 - 0:00:35 - Removed 8 empty sentences. INFO - 09/12/18 00:13:55 - 0:00:40 - Removed 23610 too long sentences. INFO - 09/12/18 00:13:55 - 0:00:40 - ============ Monolingual data (fr) INFO - 09/12/18 00:13:55 - 0:00:40 - Loading data from ./data/mono/all.fr.tok.60000.pth ... INFO - 09/12/18 00:14:16 - 0:01:00 - 264586174 words (60536 unique) in 10000000 sentences. 0 unknown words (0 unique). INFO - 09/12/18 00:14:25 - 0:01:10 - Removed 0 empty sentences. INFO - 09/12/18 00:14:30 - 0:01:15 - Removed 30272 too long sentences. INFO - 09/12/18 00:14:30 - 0:01:15 - ============ Data summary INFO - 09/12/18 00:14:30 - 0:01:15 - Parallel data - valid - en -> fr: 2993 INFO - 09/12/18 00:14:30 - 0:01:15 - Parallel data - test - en -> fr: 3003 INFO - 09/12/18 00:14:30 - 0:01:15 - Monolingual data - train - en: 9976382 INFO - 09/12/18 00:14:30 - 0:01:15 - Monolingual data - valid - en: 0 INFO - 09/12/18 00:14:30 - 0:01:15 - Monolingual data - test - en: 0 INFO - 09/12/18 00:14:30 - 0:01:15 - Monolingual data - train - fr: 9969728 INFO - 09/12/18 00:14:30 - 0:01:15 - Monolingual data - valid - fr: 0 INFO - 09/12/18 00:14:30 - 0:01:15 - Monolingual data - test - fr: 0 INFO - 09/12/18 00:14:30 - 0:01:15 - ============ Building transformer attention model - Encoder ... INFO - 09/12/18 00:14:30 - 0:01:15 - Sharing encoder input embeddings INFO - 09/12/18 00:14:32 - 0:01:17 - Sharing encoder transformer parameters for layer 1 INFO - 09/12/18 00:14:32 - 0:01:17 - Sharing encoder transformer parameters for layer 2 INFO - 09/12/18 00:14:33 - 0:01:17 - Sharing encoder transformer parameters for layer 3 INFO - 09/12/18 00:14:33 - 0:01:17 - ============ Building transformer attention model - Decoder ... INFO - 09/12/18 00:14:33 - 0:01:17 - Sharing decoder input embeddings INFO - 09/12/18 00:14:36 - 0:01:20 - Sharing decoder transformer parameters for layer 0 INFO - 09/12/18 00:14:36 - 0:01:21 - Sharing decoder transformer parameters for layer 1 INFO - 09/12/18 00:14:36 - 0:01:21 - Sharing decoder transformer parameters for layer 2 INFO - 09/12/18 00:14:39 - 0:01:24 - Sharing decoder projection matrices INFO - 09/12/18 00:14:48 - 0:01:33 - Reloading embeddings from ./data/mono/all.en-fr.60000.vec ... INFO - 09/12/18 00:14:59 - 0:01:44 - Reloaded 60523 embeddings. INFO - 09/12/18 00:15:07 - 0:01:51 - Initialized 60523 / 60536 word embeddings for "en" (including 0 after lowercasing). INFO - 09/12/18 00:15:07 - 0:01:51 - Initialized 60523 / 60536 word embeddings for "fr" (including 0 after lowercasing). INFO - 09/12/18 00:15:07 - 0:01:51 - ============ Model summary INFO - 09/12/18 00:15:07 - 0:01:51 - Number of enc+dec parameters: 129825912 INFO - 09/12/18 00:15:07 - 0:01:51 - Encoder: TransformerEncoder( (embeddings): ModuleList( (0): Embedding(60536, 512, padding_idx=2) (1): Embedding(60536, 512, padding_idx=2) ) (embed_positions): SinusoidalPositionalEmbedding() (layers): ModuleList( (0): ModuleList( (0): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) (1): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) ) (1): ModuleList( (0): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) (1): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) ) (2): ModuleList( (0): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) (1): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) ) (3): ModuleList( (0): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) (1): TransformerEncoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() ) ) ) ) ) INFO - 09/12/18 00:15:07 - 0:01:51 - Decoder: TransformerDecoder( (embeddings): ModuleList( (0): Embedding(60536, 512, padding_idx=2) (1): Embedding(60536, 512, padding_idx=2) ) (embed_positions): SinusoidalPositionalEmbedding() (layers): ModuleList( (0): ModuleList( (0): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) (1): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) ) (1): ModuleList( (0): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) (1): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) ) (2): ModuleList( (0): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) (1): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) ) (3): ModuleList( (0): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) (1): TransformerDecoderLayer( (self_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (encoder_attn): MultiheadAttention( (out_proj): Linear(in_features=512, out_features=512, bias=True) ) (fc1): Linear(in_features=512, out_features=2048, bias=True) (fc2): Linear(in_features=2048, out_features=512, bias=True) (layer_norms): ModuleList( (0): LayerNorm() (1): LayerNorm() (2): LayerNorm() ) ) ) ) (proj): ModuleList( (0): Linear(in_features=512, out_features=60536, bias=True) (1): Linear(in_features=512, out_features=60536, bias=True) ) (loss_fn): ModuleList( (0): CrossEntropyLoss() (1): CrossEntropyLoss() ) ) INFO - 09/12/18 00:15:07 - 0:01:51 - Discriminator: None INFO - 09/12/18 00:15:07 - 0:01:51 - LM: None INFO - 09/12/18 00:15:10 - 0:01:55 - Starting subprocesses for OTF generation ... INFO - 09/12/18 00:15:11 - 0:01:55 - Stopping criterion: bleu_en_fr_valid,10 INFO - 09/12/18 00:15:11 - 0:01:56 - Test: Parameters are shared correctly. INFO - 09/12/18 00:15:21 - 0:02:05 - ====================== Starting epoch 0 ... ====================== INFO - 09/12/18 00:15:21 - 0:02:05 - Creating new training encdec,en iterator ... INFO - 09/12/18 00:15:43 - 0:02:27 - Creating new training encdec,fr iterator ... INFO - 09/12/18 00:15:57 - 0:02:42 - Populating initial OTF generation cache ... INFO - 09/12/18 00:15:57 - 0:02:42 - Creating new training otf,fr iterator ... INFO - 09/12/18 00:16:07 - 0:02:52 - Creating new training otf,en iterator ... INFO - 09/12/18 00:28:59 - 0:15:43 - 50 - 7.73 sent/s - 208.00 words/s - XE-en-en: 9.1661 || XE-fr-fr: 10.3221 || XE-fr-en-fr: 8.8589 || XE-en-fr-en: 9.7598 || ENC-L2-en: 4.3974 || ENC-L2-fr: 4.4282 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 677.15s (81.78%) INFO - 09/12/18 00:38:42 - 0:25:26 - 100 - 10.98 sent/s - 300.00 words/s - XE-en-en: 6.7314 || XE-fr-fr: 6.6711 || XE-fr-en-fr: 6.5773 || XE-en-fr-en: 6.9626 || ENC-L2-en: 4.0510 || ENC-L2-fr: 4.0565 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 474.84s (81.43%) INFO - 09/12/18 00:47:49 - 0:34:33 - 150 - 11.70 sent/s - 304.00 words/s - XE-en-en: 6.1906 || XE-fr-fr: 6.2093 || XE-fr-en-fr: 6.1644 || XE-en-fr-en: 6.4241 || ENC-L2-en: 4.2035 || ENC-L2-fr: 4.0626 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 436.20s (79.71%) INFO - 09/12/18 00:56:44 - 0:43:28 - 200 - 11.97 sent/s - 315.00 words/s - XE-en-en: 6.0179 || XE-fr-fr: 5.8178 || XE-fr-en-fr: 5.8872 || XE-en-fr-en: 6.3037 || ENC-L2-en: 4.2361 || ENC-L2-fr: 4.1174 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 426.92s (79.85%) INFO - 09/12/18 01:06:53 - 0:53:37 - 250 - 10.51 sent/s - 295.00 words/s - XE-en-en: 5.8070 || XE-fr-fr: 5.6257 || XE-fr-en-fr: 5.7490 || XE-en-fr-en: 6.2022 || ENC-L2-en: 4.3129 || ENC-L2-fr: 4.1616 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 504.13s (82.75%) INFO - 09/12/18 01:16:49 - 1:03:33 - 300 - 10.74 sent/s - 297.00 words/s - XE-en-en: 5.6088 || XE-fr-fr: 5.3637 || XE-fr-en-fr: 5.5591 || XE-en-fr-en: 6.1296 || ENC-L2-en: 4.4052 || ENC-L2-fr: 4.2818 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 488.31s (81.94%) INFO - 09/12/18 01:26:54 - 1:13:39 - 350 - 10.56 sent/s - 305.00 words/s - XE-en-en: 5.5705 || XE-fr-fr: 5.3075 || XE-fr-en-fr: 5.5083 || XE-en-fr-en: 6.0392 || ENC-L2-en: 4.4234 || ENC-L2-fr: 4.3310 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 500.56s (82.63%) INFO - 09/12/18 01:36:42 - 1:23:27 - 400 - 10.89 sent/s - 286.00 words/s - XE-en-en: 5.3607 || XE-fr-fr: 5.0535 || XE-fr-en-fr: 5.5065 || XE-en-fr-en: 5.8965 || ENC-L2-en: 4.5841 || ENC-L2-fr: 4.4252 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 483.47s (82.25%) INFO - 09/12/18 01:46:10 - 1:32:54 - 450 - 11.28 sent/s - 317.00 words/s - XE-en-en: 5.1190 || XE-fr-fr: 4.9407 || XE-fr-en-fr: 5.4213 || XE-en-fr-en: 5.8902 || ENC-L2-en: 4.6964 || ENC-L2-fr: 4.6057 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 461.19s (81.28%) INFO - 09/12/18 01:55:21 - 1:42:06 - 500 - 11.60 sent/s - 295.00 words/s - XE-en-en: 5.0081 || XE-fr-fr: 4.5900 || XE-fr-en-fr: 5.3086 || XE-en-fr-en: 5.7548 || ENC-L2-en: 4.8601 || ENC-L2-fr: 4.8254 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 447.69s (81.16%) INFO - 09/12/18 02:05:28 - 1:52:13 - 550 - 10.54 sent/s - 287.00 words/s - XE-en-en: 4.8493 || XE-fr-fr: 4.5845 || XE-fr-en-fr: 5.3413 || XE-en-fr-en: 5.8398 || ENC-L2-en: 4.9412 || ENC-L2-fr: 4.8388 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 502.77s (82.80%) INFO - 09/12/18 02:14:04 - 2:00:49 - 600 - 12.41 sent/s - 345.00 words/s - XE-en-en: 4.8164 || XE-fr-fr: 4.6268 || XE-fr-en-fr: 5.2810 || XE-en-fr-en: 5.7997 || ENC-L2-en: 4.9693 || ENC-L2-fr: 4.8868 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 415.46s (80.55%) INFO - 09/12/18 02:23:47 - 2:10:31 - 650 - 10.98 sent/s - 304.00 words/s - XE-en-en: 4.6208 || XE-fr-fr: 4.2996 || XE-fr-en-fr: 5.3366 || XE-en-fr-en: 5.6974 || ENC-L2-en: 5.0687 || ENC-L2-fr: 5.0444 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 483.02s (82.90%) INFO - 09/12/18 02:33:32 - 2:20:16 - 700 - 10.95 sent/s - 288.00 words/s - XE-en-en: 4.4145 || XE-fr-fr: 4.2215 || XE-fr-en-fr: 5.2344 || XE-en-fr-en: 5.6750 || ENC-L2-en: 5.0999 || ENC-L2-fr: 5.0910 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 482.54s (82.52%) INFO - 09/12/18 02:42:59 - 2:29:44 - 750 - 11.28 sent/s - 323.00 words/s - XE-en-en: 4.3274 || XE-fr-fr: 4.0474 || XE-fr-en-fr: 5.2802 || XE-en-fr-en: 5.6882 || ENC-L2-en: 5.1056 || ENC-L2-fr: 5.1181 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 460.53s (81.16%) INFO - 09/12/18 02:52:43 - 2:39:27 - 800 - 10.97 sent/s - 303.00 words/s - XE-en-en: 4.2553 || XE-fr-fr: 3.8836 || XE-fr-en-fr: 5.1004 || XE-en-fr-en: 5.6223 || ENC-L2-en: 5.1231 || ENC-L2-fr: 5.1132 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 476.91s (81.73%) INFO - 09/12/18 03:02:43 - 2:49:28 - 850 - 10.66 sent/s - 288.00 words/s - XE-en-en: 4.0454 || XE-fr-fr: 3.7294 || XE-fr-en-fr: 5.1255 || XE-en-fr-en: 5.6609 || ENC-L2-en: 5.1577 || ENC-L2-fr: 5.1205 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 490.82s (81.73%) INFO - 09/12/18 03:12:35 - 2:59:19 - 900 - 10.82 sent/s - 300.00 words/s - XE-en-en: 4.0419 || XE-fr-fr: 3.5819 || XE-fr-en-fr: 5.1324 || XE-en-fr-en: 5.5564 || ENC-L2-en: 5.1338 || ENC-L2-fr: 5.1236 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 485.61s (82.06%) INFO - 09/12/18 03:23:25 - 3:10:10 - 950 - 9.84 sent/s - 276.00 words/s - XE-en-en: 3.8012 || XE-fr-fr: 3.5361 || XE-fr-en-fr: 5.0602 || XE-en-fr-en: 5.5578 || ENC-L2-en: 5.1254 || ENC-L2-fr: 5.0901 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 542.39s (83.41%) INFO - 09/12/18 03:33:25 - 3:20:09 - 1000 - 10.67 sent/s - 297.00 words/s - XE-en-en: 3.7472 || XE-fr-fr: 3.5208 || XE-fr-en-fr: 5.0302 || XE-en-fr-en: 5.5621 || ENC-L2-en: 5.1286 || ENC-L2-fr: 5.1080 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 496.39s (82.76%) INFO - 09/12/18 03:42:25 - 3:29:09 - 1050 - 11.86 sent/s - 328.00 words/s - XE-en-en: 3.6961 || XE-fr-fr: 3.5080 || XE-fr-en-fr: 4.9843 || XE-en-fr-en: 5.4201 || ENC-L2-en: 5.0306 || ENC-L2-fr: 4.9839 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 420.52s (77.92%) INFO - 09/12/18 03:52:56 - 3:39:40 - 1100 - 10.14 sent/s - 276.00 words/s - XE-en-en: 3.6204 || XE-fr-fr: 3.3347 || XE-fr-en-fr: 4.8431 || XE-en-fr-en: 5.3124 || ENC-L2-en: 5.0046 || ENC-L2-fr: 5.0139 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 518.68s (82.15%) INFO - 09/12/18 04:02:33 - 3:49:17 - 1150 - 11.10 sent/s - 300.00 words/s - XE-en-en: 3.5403 || XE-fr-fr: 3.0586 || XE-fr-en-fr: 4.8230 || XE-en-fr-en: 5.2239 || ENC-L2-en: 5.0438 || ENC-L2-fr: 5.0055 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 467.54s (81.07%) INFO - 09/12/18 04:13:21 - 4:00:05 - 1200 - 9.87 sent/s - 288.00 words/s - XE-en-en: 3.6330 || XE-fr-fr: 3.1748 || XE-fr-en-fr: 4.8212 || XE-en-fr-en: 5.2040 || ENC-L2-en: 5.0675 || ENC-L2-fr: 5.0078 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 539.82s (83.29%) INFO - 09/12/18 04:23:00 - 4:09:45 - 1250 - 11.05 sent/s - 307.00 words/s - XE-en-en: 3.4313 || XE-fr-fr: 3.1089 || XE-fr-en-fr: 4.7378 || XE-en-fr-en: 5.1445 || ENC-L2-en: 5.0429 || ENC-L2-fr: 5.0171 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 473.55s (81.76%) INFO - 09/12/18 04:32:51 - 4:19:35 - 1300 - 10.83 sent/s - 293.00 words/s - XE-en-en: 3.3948 || XE-fr-fr: 2.9642 || XE-fr-en-fr: 4.7311 || XE-en-fr-en: 5.1657 || ENC-L2-en: 5.0408 || ENC-L2-fr: 4.9928 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 486.71s (82.37%) INFO - 09/12/18 04:42:15 - 4:29:00 - 1350 - 11.34 sent/s - 291.00 words/s - XE-en-en: 3.3029 || XE-fr-fr: 2.7672 || XE-fr-en-fr: 4.7236 || XE-en-fr-en: 5.1264 || ENC-L2-en: 5.0386 || ENC-L2-fr: 5.0000 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 458.94s (81.32%) INFO - 09/12/18 04:52:44 - 4:39:28 - 1400 - 10.19 sent/s - 289.00 words/s - XE-en-en: 3.2432 || XE-fr-fr: 3.0275 || XE-fr-en-fr: 4.7498 || XE-en-fr-en: 5.0745 || ENC-L2-en: 5.0144 || ENC-L2-fr: 4.9981 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 519.00s (82.60%) INFO - 09/12/18 05:02:47 - 4:49:31 - 1450 - 10.61 sent/s - 319.00 words/s - XE-en-en: 3.3478 || XE-fr-fr: 2.9084 || XE-fr-en-fr: 4.7582 || XE-en-fr-en: 5.0673 || ENC-L2-en: 5.0186 || ENC-L2-fr: 4.9808 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 491.95s (81.56%) INFO - 09/12/18 05:13:39 - 5:00:24 - 1500 - 9.81 sent/s - 275.00 words/s - XE-en-en: 3.2033 || XE-fr-fr: 2.8267 || XE-fr-en-fr: 4.7247 || XE-en-fr-en: 5.0272 || ENC-L2-en: 4.9895 || ENC-L2-fr: 4.9998 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 551.96s (84.61%) INFO - 09/12/18 05:22:40 - 5:09:25 - 1550 - 11.83 sent/s - 331.00 words/s - XE-en-en: 3.0620 || XE-fr-fr: 2.8688 || XE-fr-en-fr: 4.6226 || XE-en-fr-en: 5.0337 || ENC-L2-en: 4.9562 || ENC-L2-fr: 4.9575 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 438.78s (81.08%) INFO - 09/12/18 05:33:47 - 5:20:31 - 1600 - 9.60 sent/s - 263.00 words/s - XE-en-en: 3.0383 || XE-fr-fr: 2.6930 || XE-fr-en-fr: 4.7487 || XE-en-fr-en: 5.0797 || ENC-L2-en: 4.9621 || ENC-L2-fr: 4.9623 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 566.79s (85.03%) INFO - 09/12/18 05:43:21 - 5:30:05 - 1650 - 11.15 sent/s - 302.00 words/s - XE-en-en: 3.1398 || XE-fr-fr: 2.6422 || XE-fr-en-fr: 4.6337 || XE-en-fr-en: 5.0125 || ENC-L2-en: 4.9555 || ENC-L2-fr: 4.9081 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 469.38s (81.80%) INFO - 09/12/18 05:53:29 - 5:40:14 - 1700 - 10.51 sent/s - 284.00 words/s - XE-en-en: 2.9949 || XE-fr-fr: 2.7213 || XE-fr-en-fr: 4.6445 || XE-en-fr-en: 5.0430 || ENC-L2-en: 4.9435 || ENC-L2-fr: 4.9039 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 501.13s (82.32%) INFO - 09/12/18 06:02:27 - 5:49:12 - 1750 - 11.90 sent/s - 341.00 words/s - XE-en-en: 3.1402 || XE-fr-fr: 2.6921 || XE-fr-en-fr: 4.6446 || XE-en-fr-en: 5.0308 || ENC-L2-en: 4.9122 || ENC-L2-fr: 4.9261 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 433.97s (80.69%) INFO - 09/12/18 06:12:11 - 5:58:55 - 1800 - 10.97 sent/s - 314.00 words/s - XE-en-en: 3.0966 || XE-fr-fr: 2.6610 || XE-fr-en-fr: 4.5980 || XE-en-fr-en: 5.0058 || ENC-L2-en: 4.9202 || ENC-L2-fr: 4.8994 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 478.16s (81.94%) INFO - 09/12/18 06:21:42 - 6:08:27 - 1850 - 11.20 sent/s - 299.00 words/s - XE-en-en: 2.9677 || XE-fr-fr: 2.5909 || XE-fr-en-fr: 4.6525 || XE-en-fr-en: 4.9898 || ENC-L2-en: 4.8995 || ENC-L2-fr: 4.8829 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 465.48s (81.43%) INFO - 09/12/18 06:31:45 - 6:18:30 - 1900 - 10.62 sent/s - 304.00 words/s - XE-en-en: 2.9639 || XE-fr-fr: 2.4422 || XE-fr-en-fr: 4.6129 || XE-en-fr-en: 4.9739 || ENC-L2-en: 4.8933 || ENC-L2-fr: 4.8388 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 497.63s (82.59%) INFO - 09/12/18 06:42:13 - 6:28:58 - 1950 - 10.18 sent/s - 296.00 words/s - XE-en-en: 2.9480 || XE-fr-fr: 2.4710 || XE-fr-en-fr: 4.6153 || XE-en-fr-en: 4.9021 || ENC-L2-en: 4.8988 || ENC-L2-fr: 4.8996 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 521.18s (82.94%) INFO - 09/12/18 06:52:45 - 6:39:29 - 2000 - 10.14 sent/s - 276.00 words/s - XE-en-en: 2.7779 || XE-fr-fr: 2.4025 || XE-fr-en-fr: 4.5986 || XE-en-fr-en: 4.9312 || ENC-L2-en: 4.8520 || ENC-L2-fr: 4.8394 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 528.49s (83.69%) INFO - 09/12/18 07:01:49 - 6:48:34 - 2050 - 11.76 sent/s - 328.00 words/s - XE-en-en: 2.8612 || XE-fr-fr: 2.4049 || XE-fr-en-fr: 4.4924 || XE-en-fr-en: 4.8329 || ENC-L2-en: 4.8330 || ENC-L2-fr: 4.8005 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 434.22s (79.80%) INFO - 09/12/18 07:12:28 - 6:59:13 - 2100 - 10.01 sent/s - 269.00 words/s - XE-en-en: 2.7221 || XE-fr-fr: 2.3442 || XE-fr-en-fr: 4.3868 || XE-en-fr-en: 4.7984 || ENC-L2-en: 4.8134 || ENC-L2-fr: 4.7380 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 533.86s (83.53%) INFO - 09/12/18 07:23:32 - 7:10:17 - 2150 - 9.64 sent/s - 288.00 words/s - XE-en-en: 2.7103 || XE-fr-fr: 2.4878 || XE-fr-en-fr: 4.4820 || XE-en-fr-en: 4.7232 || ENC-L2-en: 4.8404 || ENC-L2-fr: 4.7818 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 555.75s (83.67%) INFO - 09/12/18 07:32:23 - 7:19:07 - 2200 - 12.06 sent/s - 321.00 words/s - XE-en-en: 2.9058 || XE-fr-fr: 2.2955 || XE-fr-en-fr: 4.3114 || XE-en-fr-en: 4.7381 || ENC-L2-en: 4.8253 || ENC-L2-fr: 4.7277 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 431.47s (81.32%) INFO - 09/12/18 07:41:57 - 7:28:42 - 2250 - 11.15 sent/s - 304.00 words/s - XE-en-en: 2.6717 || XE-fr-fr: 2.3233 || XE-fr-en-fr: 4.3324 || XE-en-fr-en: 4.7045 || ENC-L2-en: 4.8161 || ENC-L2-fr: 4.7164 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 468.93s (81.68%) INFO - 09/12/18 07:51:32 - 7:38:17 - 2300 - 11.13 sent/s - 306.00 words/s - XE-en-en: 2.6747 || XE-fr-fr: 2.3387 || XE-fr-en-fr: 4.2133 || XE-en-fr-en: 4.7302 || ENC-L2-en: 4.8328 || ENC-L2-fr: 4.7369 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 471.81s (82.04%) INFO - 09/12/18 08:01:44 - 7:48:29 - 2350 - 10.45 sent/s - 279.00 words/s - XE-en-en: 2.5885 || XE-fr-fr: 2.2188 || XE-fr-en-fr: 4.3310 || XE-en-fr-en: 4.6589 || ENC-L2-en: 4.8133 || ENC-L2-fr: 4.7106 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 507.29s (82.84%) INFO - 09/12/18 08:11:14 - 7:57:58 - 2400 - 11.24 sent/s - 308.00 words/s - XE-en-en: 2.7912 || XE-fr-fr: 2.2262 || XE-fr-en-fr: 4.2290 || XE-en-fr-en: 4.7165 || ENC-L2-en: 4.7941 || ENC-L2-fr: 4.6939 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 465.17s (81.72%) INFO - 09/12/18 08:19:53 - 8:06:37 - 2450 - 12.33 sent/s - 327.00 words/s - XE-en-en: 2.5534 || XE-fr-fr: 2.2518 || XE-fr-en-fr: 4.2078 || XE-en-fr-en: 4.6361 || ENC-L2-en: 4.7722 || ENC-L2-fr: 4.6753 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 416.16s (80.20%) INFO - 09/12/18 08:29:39 - 8:16:24 - 2500 - 10.91 sent/s - 287.00 words/s - XE-en-en: 2.5518 || XE-fr-fr: 2.1439 || XE-fr-en-fr: 4.2018 || XE-en-fr-en: 4.6456 || ENC-L2-en: 4.7501 || ENC-L2-fr: 4.6758 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 481.69s (82.10%) INFO - 09/12/18 08:39:38 - 8:26:23 - 2550 - 10.69 sent/s - 290.00 words/s - XE-en-en: 2.3991 || XE-fr-fr: 2.2788 || XE-fr-en-fr: 4.2382 || XE-en-fr-en: 4.7055 || ENC-L2-en: 4.7275 || ENC-L2-fr: 4.6791 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 497.92s (83.15%) INFO - 09/12/18 08:50:39 - 8:37:23 - 2600 - 9.69 sent/s - 268.00 words/s - XE-en-en: 2.4991 || XE-fr-fr: 2.1442 || XE-fr-en-fr: 4.2355 || XE-en-fr-en: 4.6633 || ENC-L2-en: 4.7650 || ENC-L2-fr: 4.6801 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 552.72s (83.68%) INFO - 09/12/18 09:00:43 - 8:47:28 - 2650 - 10.59 sent/s - 299.00 words/s - XE-en-en: 2.4932 || XE-fr-fr: 2.2504 || XE-fr-en-fr: 4.2403 || XE-en-fr-en: 4.6802 || ENC-L2-en: 4.7526 || ENC-L2-fr: 4.6582 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 502.07s (83.07%) INFO - 09/12/18 09:11:21 - 8:58:06 - 2700 - 10.03 sent/s - 286.00 words/s - XE-en-en: 2.5313 || XE-fr-fr: 2.1144 || XE-fr-en-fr: 4.2631 || XE-en-fr-en: 4.6904 || ENC-L2-en: 4.7721 || ENC-L2-fr: 4.5998 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 532.25s (83.39%) INFO - 09/12/18 09:20:48 - 9:07:32 - 2750 - 11.30 sent/s - 317.00 words/s - XE-en-en: 2.4987 || XE-fr-fr: 2.1388 || XE-fr-en-fr: 4.2393 || XE-en-fr-en: 4.6174 || ENC-L2-en: 4.7235 || ENC-L2-fr: 4.6385 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 461.46s (81.48%) INFO - 09/12/18 09:30:50 - 9:17:35 - 2800 - 10.62 sent/s - 298.00 words/s - XE-en-en: 2.4183 || XE-fr-fr: 2.3156 || XE-fr-en-fr: 4.1154 || XE-en-fr-en: 4.5997 || ENC-L2-en: 4.6819 || ENC-L2-fr: 4.6356 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 500.90s (83.12%) INFO - 09/12/18 09:41:21 - 9:28:06 - 2850 - 10.15 sent/s - 282.00 words/s - XE-en-en: 2.4332 || XE-fr-fr: 2.0810 || XE-fr-en-fr: 4.1634 || XE-en-fr-en: 4.6895 || ENC-L2-en: 4.7198 || ENC-L2-fr: 4.6262 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 524.02s (83.07%) INFO - 09/12/18 09:52:06 - 9:38:51 - 2900 - 9.92 sent/s - 273.00 words/s - XE-en-en: 2.4564 || XE-fr-fr: 2.1034 || XE-fr-en-fr: 4.1954 || XE-en-fr-en: 4.5364 || ENC-L2-en: 4.7143 || ENC-L2-fr: 4.6042 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 449.40s (69.64%) INFO - 09/12/18 10:01:50 - 9:48:35 - 2950 - 10.96 sent/s - 318.00 words/s - XE-en-en: 2.4861 || XE-fr-fr: 2.1129 || XE-fr-en-fr: 4.1567 || XE-en-fr-en: 4.6506 || ENC-L2-en: 4.7036 || ENC-L2-fr: 4.5820 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 396.58s (67.93%) INFO - 09/12/18 10:11:31 - 9:58:16 - 3000 - 11.01 sent/s - 308.00 words/s - XE-en-en: 2.5044 || XE-fr-fr: 2.0395 || XE-fr-en-fr: 4.1785 || XE-en-fr-en: 4.5932 || ENC-L2-en: 4.6905 || ENC-L2-fr: 4.6105 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 476.50s (81.97%) INFO - 09/12/18 10:21:03 - 10:07:48 - 3050 - 11.19 sent/s - 313.00 words/s - XE-en-en: 2.4341 || XE-fr-fr: 2.0570 || XE-fr-en-fr: 4.1284 || XE-en-fr-en: 4.3460 || ENC-L2-en: 4.6961 || ENC-L2-fr: 4.6266 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 458.35s (80.15%) INFO - 09/12/18 10:30:43 - 10:17:27 - 3100 - 11.04 sent/s - 320.00 words/s - XE-en-en: 2.4132 || XE-fr-fr: 2.0611 || XE-fr-en-fr: 3.9949 || XE-en-fr-en: 4.3123 || ENC-L2-en: 4.6433 || ENC-L2-fr: 4.6369 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 466.46s (80.49%) INFO - 09/12/18 10:40:09 - 10:26:53 - 3150 - 11.31 sent/s - 314.00 words/s - XE-en-en: 2.3880 || XE-fr-fr: 2.1118 || XE-fr-en-fr: 3.9230 || XE-en-fr-en: 4.1259 || ENC-L2-en: 4.6623 || ENC-L2-fr: 4.6434 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 457.01s (80.78%) INFO - 09/12/18 10:46:16 - 10:33:01 - 3200 - 17.40 sent/s - 490.00 words/s - XE-en-en: 2.2972 || XE-fr-fr: 1.9885 || XE-fr-en-fr: 3.9749 || XE-en-fr-en: 4.1820 || ENC-L2-en: 4.6630 || ENC-L2-fr: 4.6333 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 282.36s (76.75%) INFO - 09/12/18 10:51:29 - 10:38:13 - 3250 - 20.49 sent/s - 568.00 words/s - XE-en-en: 2.3694 || XE-fr-fr: 2.0380 || XE-fr-en-fr: 3.9955 || XE-en-fr-en: 4.1473 || ENC-L2-en: 4.6422 || ENC-L2-fr: 4.5769 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 236.45s (75.68%) INFO - 09/12/18 10:57:08 - 10:43:53 - 3300 - 18.85 sent/s - 500.00 words/s - XE-en-en: 2.2998 || XE-fr-fr: 2.0128 || XE-fr-en-fr: 3.9162 || XE-en-fr-en: 4.0999 || ENC-L2-en: 4.6426 || ENC-L2-fr: 4.5942 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 266.65s (78.54%) INFO - 09/12/18 11:02:49 - 10:49:34 - 3350 - 18.77 sent/s - 549.00 words/s - XE-en-en: 2.2248 || XE-fr-fr: 2.0117 || XE-fr-en-fr: 4.0487 || XE-en-fr-en: 4.2262 || ENC-L2-en: 4.6246 || ENC-L2-fr: 4.5877 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 262.68s (77.04%) INFO - 09/12/18 11:08:02 - 10:54:47 - 3400 - 20.46 sent/s - 574.00 words/s - XE-en-en: 2.2398 || XE-fr-fr: 2.0246 || XE-fr-en-fr: 3.9525 || XE-en-fr-en: 4.1087 || ENC-L2-en: 4.6369 || ENC-L2-fr: 4.5788 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 234.22s (74.86%) INFO - 09/12/18 11:14:02 - 11:00:47 - 3450 - 17.78 sent/s - 529.00 words/s - XE-en-en: 2.2870 || XE-fr-fr: 2.0339 || XE-fr-en-fr: 3.9009 || XE-en-fr-en: 4.1717 || ENC-L2-en: 4.6446 || ENC-L2-fr: 4.6211 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 281.18s (78.13%) INFO - 09/12/18 11:19:28 - 11:06:13 - 3500 - 19.62 sent/s - 535.00 words/s - XE-en-en: 2.0867 || XE-fr-fr: 1.8984 || XE-fr-en-fr: 4.0485 || XE-en-fr-en: 3.9540 || ENC-L2-en: 4.5922 || ENC-L2-fr: 4.5811 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 249.76s (76.56%) INFO - 09/12/18 11:25:08 - 11:11:53 - 3550 - 18.84 sent/s - 486.00 words/s - XE-en-en: 2.1970 || XE-fr-fr: 1.8988 || XE-fr-en-fr: 3.8241 || XE-en-fr-en: 4.0218 || ENC-L2-en: 4.6200 || ENC-L2-fr: 4.5685 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 261.78s (77.04%) INFO - 09/12/18 11:30:21 - 11:17:06 - 3600 - 20.46 sent/s - 544.00 words/s - XE-en-en: 2.1996 || XE-fr-fr: 1.8140 || XE-fr-en-fr: 3.8791 || XE-en-fr-en: 4.0762 || ENC-L2-en: 4.5943 || ENC-L2-fr: 4.5225 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 238.77s (76.31%) INFO - 09/12/18 11:36:00 - 11:22:45 - 3650 - 18.87 sent/s - 521.00 words/s - XE-en-en: 2.1673 || XE-fr-fr: 1.8874 || XE-fr-en-fr: 3.9414 || XE-en-fr-en: 4.0719 || ENC-L2-en: 4.5958 || ENC-L2-fr: 4.5342 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 260.44s (76.80%) INFO - 09/12/18 11:41:15 - 11:28:00 - 3700 - 20.30 sent/s - 574.00 words/s - XE-en-en: 2.2113 || XE-fr-fr: 1.9911 || XE-fr-en-fr: 3.8612 || XE-en-fr-en: 4.0643 || ENC-L2-en: 4.5936 || ENC-L2-fr: 4.5887 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 241.54s (76.61%) INFO - 09/12/18 11:46:35 - 11:33:19 - 3750 - 20.04 sent/s - 576.00 words/s - XE-en-en: 2.2332 || XE-fr-fr: 1.9591 || XE-fr-en-fr: 3.8968 || XE-en-fr-en: 4.0105 || ENC-L2-en: 4.6122 || ENC-L2-fr: 4.5944 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 241.98s (75.76%) INFO - 09/12/18 11:51:43 - 11:38:28 - 3800 - 20.76 sent/s - 556.00 words/s - XE-en-en: 2.0574 || XE-fr-fr: 1.8903 || XE-fr-en-fr: 3.8682 || XE-en-fr-en: 4.0710 || ENC-L2-en: 4.5576 || ENC-L2-fr: 4.5704 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 234.56s (76.09%) INFO - 09/12/18 11:57:13 - 11:43:58 - 3850 - 19.40 sent/s - 523.00 words/s - XE-en-en: 2.1349 || XE-fr-fr: 1.7823 || XE-fr-en-fr: 3.9010 || XE-en-fr-en: 3.9755 || ENC-L2-en: 4.5504 || ENC-L2-fr: 4.5159 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 254.27s (77.06%) INFO - 09/12/18 12:02:54 - 11:49:38 - 3900 - 18.80 sent/s - 524.00 words/s - XE-en-en: 2.0925 || XE-fr-fr: 1.8943 || XE-fr-en-fr: 3.8455 || XE-en-fr-en: 4.1243 || ENC-L2-en: 4.5128 || ENC-L2-fr: 4.5169 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 261.41s (76.77%) INFO - 09/12/18 12:03:28 - 11:50:13 - ====================== End of epoch 0 ====================== INFO - 09/12/18 12:03:28 - 11:50:13 - Evaluating en -> fr (valid) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 12:06:26 - 11:53:11 - BLEU ./dumped/base1/2un3e1yus0/hyp0.en-fr.valid.txt ./dumped/base1/2un3e1yus0/ref.en-fr.valid.txt : 2.260000 INFO - 09/12/18 12:06:26 - 11:53:11 - Evaluating fr -> en (valid) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 12:08:17 - 11:55:01 - BLEU ./dumped/base1/2un3e1yus0/hyp0.fr-en.valid.txt ./dumped/base1/2un3e1yus0/ref.fr-en.valid.txt : 2.620000 INFO - 09/12/18 12:08:17 - 11:55:01 - Evaluating en -> fr (test) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 12:11:16 - 11:58:00 - BLEU ./dumped/base1/2un3e1yus0/hyp0.en-fr.test.txt ./dumped/base1/2un3e1yus0/ref.en-fr.test.txt : 2.580000 INFO - 09/12/18 12:11:16 - 11:58:00 - Evaluating fr -> en (test) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 12:19:42 - 12:06:27 - BLEU ./dumped/base1/2un3e1yus0/hyp0.fr-en.test.txt ./dumped/base1/2un3e1yus0/ref.fr-en.test.txt : 3.010000 INFO - 09/12/18 12:19:42 - 12:06:27 - Evaluating fr -> en -> fr (valid) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 12:37:28 - 12:24:13 - BLEU ./dumped/base1/2un3e1yus0/hyp0.fr-en-fr.valid.txt ./dumped/base1/2un3e1yus0/ref.en-fr.valid.txt : 2.250000 INFO - 09/12/18 12:37:28 - 12:24:13 - Evaluating fr -> en -> fr (test) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 12:57:08 - 12:43:53 - BLEU ./dumped/base1/2un3e1yus0/hyp0.fr-en-fr.test.txt ./dumped/base1/2un3e1yus0/ref.en-fr.test.txt : 2.010000 INFO - 09/12/18 12:57:08 - 12:43:53 - Evaluating en -> fr -> en (valid) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 13:16:13 - 13:02:57 - BLEU ./dumped/base1/2un3e1yus0/hyp0.en-fr-en.valid.txt ./dumped/base1/2un3e1yus0/ref.fr-en.valid.txt : 2.170000 INFO - 09/12/18 13:16:13 - 13:02:57 - Evaluating en -> fr -> en (test) ... It is in-advisable to publish scores from multi-bleu.perl. The scores depend on your tokenizer, which is unlikely to be reproducible from your paper or consistent across research groups. Instead you should detokenize then use mteval-v14.pl, which has a standard tokenization. Scores from multi-bleu.perl can still be used for internal purposes when you have a consistent tokenizer. INFO - 09/12/18 13:35:28 - 13:22:13 - BLEU ./dumped/base1/2un3e1yus0/hyp0.en-fr-en.test.txt ./dumped/base1/2un3e1yus0/ref.fr-en.test.txt : 2.280000 INFO - 09/12/18 13:35:28 - 13:22:13 - epoch -> 0.000000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_en_fr_valid -> 64.527524 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_en_fr_valid -> 2.260000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_fr_en_valid -> 89.226258 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_fr_en_valid -> 2.620000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_en_fr_test -> 51.925330 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_en_fr_test -> 2.580000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_fr_en_test -> 74.360186 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_fr_en_test -> 3.010000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_fr_en_fr_valid -> 64.002322 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_fr_en_fr_valid -> 2.250000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_fr_en_fr_test -> 58.542789 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_fr_en_fr_test -> 2.010000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_en_fr_en_valid -> 87.600623 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_en_fr_en_valid -> 2.170000 INFO - 09/12/18 13:35:28 - 13:22:13 - ppl_en_fr_en_test -> 86.599739 INFO - 09/12/18 13:35:28 - 13:22:13 - bleu_en_fr_en_test -> 2.280000 INFO - 09/12/18 13:35:28 - 13:22:13 - __log__:{"epoch": 0, "ppl_en_fr_valid": 64.52752396544003, "bleu_en_fr_valid": 2.26, "ppl_fr_en_valid": 89.22625815559613, "bleu_fr_en_valid": 2.62, "ppl_en_fr_test": 51.9253300364525, "bleu_en_fr_test": 2.58, "ppl_fr_en_test": 74.36018595597423, "bleu_fr_en_test": 3.01, "ppl_fr_en_fr_valid": 64.00232203749903, "bleu_fr_en_fr_valid": 2.25, "ppl_fr_en_fr_test": 58.54278880639311, "bleu_fr_en_fr_test": 2.01, "ppl_en_fr_en_valid": 87.60062318695631, "bleu_en_fr_en_valid": 2.17, "ppl_en_fr_en_test": 86.59973864520856, "bleu_en_fr_en_test": 2.28} INFO - 09/12/18 13:35:28 - 13:22:13 - New best score for bleu_en_fr_valid: 2.260000 INFO - 09/12/18 13:35:28 - 13:22:13 - Saving model to ./dumped/base1/2un3e1yus0/best-bleu_en_fr_valid.pth ... INFO - 09/12/18 13:35:34 - 13:22:18 - New best validation score: 2.260000 INFO - 09/12/18 13:35:34 - 13:22:18 - Saving checkpoint to ./dumped/base1/2un3e1yus0/checkpoint.pth ... INFO - 09/12/18 13:35:43 - 13:22:28 - Test: Parameters are shared correctly. INFO - 09/12/18 13:35:43 - 13:22:28 - ====================== Starting epoch 1 ... ====================== INFO - 09/12/18 13:41:40 - 13:28:25 - 3950 - 1.08 sent/s - 28.00 words/s - XE-en-en: 2.1020 || XE-fr-fr: 1.8571 || XE-fr-en-fr: 3.9077 || XE-en-fr-en: 3.8841 || ENC-L2-en: 4.5637 || ENC-L2-fr: 4.4674 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 292.87s (4.94%) INFO - 09/12/18 13:51:25 - 13:38:10 - 4000 - 10.95 sent/s - 315.00 words/s - XE-en-en: 2.2067 || XE-fr-fr: 1.7914 || XE-fr-en-fr: 3.9852 || XE-en-fr-en: 3.9281 || ENC-L2-en: 4.5827 || ENC-L2-fr: 4.5026 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 475.74s (81.39%) INFO - 09/12/18 14:00:32 - 13:47:16 - 4050 - 11.70 sent/s - 318.00 words/s - XE-en-en: 2.0842 || XE-fr-fr: 1.7998 || XE-fr-en-fr: 3.6291 || XE-en-fr-en: 3.9301 || ENC-L2-en: 4.5470 || ENC-L2-fr: 4.5034 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 436.40s (79.81%) INFO - 09/12/18 14:10:06 - 13:56:50 - 4100 - 11.15 sent/s - 304.00 words/s - XE-en-en: 2.1037 || XE-fr-fr: 1.8181 || XE-fr-en-fr: 3.6182 || XE-en-fr-en: 3.7607 || ENC-L2-en: 4.5243 || ENC-L2-fr: 4.4946 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 472.35s (82.32%) INFO - 09/12/18 14:19:33 - 14:06:17 - 4150 - 11.29 sent/s - 327.00 words/s - XE-en-en: 2.1165 || XE-fr-fr: 1.8207 || XE-fr-en-fr: 3.6759 || XE-en-fr-en: 3.9170 || ENC-L2-en: 4.5195 || ENC-L2-fr: 4.4797 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 460.23s (81.17%) INFO - 09/12/18 14:29:03 - 14:15:48 - 4200 - 11.21 sent/s - 320.00 words/s - XE-en-en: 1.9359 || XE-fr-fr: 1.8547 || XE-fr-en-fr: 3.6076 || XE-en-fr-en: 3.9188 || ENC-L2-en: 4.5065 || ENC-L2-fr: 4.5197 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 459.59s (80.51%) INFO - 09/12/18 14:37:36 - 14:24:21 - 4250 - 12.47 sent/s - 339.00 words/s - XE-en-en: 2.0330 || XE-fr-fr: 1.7889 || XE-fr-en-fr: 3.5425 || XE-en-fr-en: 3.7453 || ENC-L2-en: 4.5126 || ENC-L2-fr: 4.4746 - LR enc=1.0000e-04,dec=1.0000e-04 - Sentences generation time: 406.67s (79.25%) data_type: valid lang1: en lang2:fr data_type: test lang1: en lang2:fr Traceback (most recent call last): File "main.py", line 317, in batches = next(otf_iterator) File "/mnt/disk-c/liujiqiang/low-resourceMT/UnsupervisedMT/en2fr/baseline/NMT/src/trainer.py", line 561, in otf_bt_gen_async results = cache[0].gen() File "/mnt/disk-c/liujiqiang/low-resourceMT/UnsupervisedMT/en2fr/baseline/NMT/src/multiprocessing_event_loop.py", line 203, in gen return next(self.generator) File "/mnt/disk-c/liujiqiang/low-resourceMT/UnsupervisedMT/en2fr/baseline/NMT/src/multiprocessing_event_loop.py", line 73, in fetch_all_result_generator result_type, result = self.return_pipes[rank].recv() File "/home/liujq/Python-3.6.4/lib/python3.6/multiprocessing/connection.py", line 250, in recv buf = self._recv_bytes() File "/home/liujq/Python-3.6.4/lib/python3.6/multiprocessing/connection.py", line 407, in _recv_bytes buf = self._recv(4) File "/home/liujq/Python-3.6.4/lib/python3.6/multiprocessing/connection.py", line 383, in _recv raise EOFError EOFError