Upload New File

configs/fallout_tuned.yaml

+# Data output:
+overwrite: false
+save_data: ./data/fallout/vocab/fallout_multi
+src_vocab: ./data/fallout/vocab/fallout_multi.vocab.src
+tgt_vocab: ./data/fallout/vocab/fallout_multi.vocab.tgt
+
+# Training corpora:
+data:
+    globalvoices:
+        path_src: ./data/globalvoices/tok/trn.en
+        path_tgt: ./data/globalvoices/tok/trn.fr
+        transforms: [filtertoolong, sentencepiece]
+        weight: 1
+    europarl:
+        path_src: ./data/europarl/tok/trn.en
+        path_tgt: ./data/europarl/tok/trn.fr
+        transforms: [filtertoolong, sentencepiece]
+        weight: 1
+    books:
+        path_src: ./data/books/tok/trn.en
+        path_tgt: ./data/books/tok/trn.fr
+        transforms: [filtertoolong, sentencepiece]
+        weight: 1
+    news:
+        path_src: ./data/news/tok/trn.en
+        path_tgt: ./data/news/tok/trn.fr
+        transforms: [filtertoolong, sentencepiece]
+        weight: 1
+    ted:
+        path_src: ./data/ted/tok/trn.en
+        path_tgt: ./data/ted/tok/trn.fr
+        transforms: [filtertoolong, sentencepiece]
+        weight: 1
+    vg:
+        path_src: ./data/fallout/tok/trn.en
+        path_tgt: ./data/fallout/tok/trn.fr
+        transforms: [filtertoolong, sentencepiece]
+        weight: 10
+    valid:
+        path_src: ./data/fallout/tok/val.en
+        path_tgt: ./data/fallout/tok/val.fr
+        transforms: [filtertoolong, sentencepiece]
+src_seq_length: 200
+tgt_seq_length: 200
+skip_empty_level: silent
+src_subword_model: ./data/fallout/subword/unigram_multi_en.model
+tgt_subword_model: ./data/fallout/subword/unigram_multi_fr.model
+src_subword_vocab: ./data/fallout/subword/unigram_multi_en.vocab
+tgt_subword_vocab: ./data/fallout/subword/unigram_multi_fr.vocab
+src_subword_nbest: 20
+tgt_subword_nbest: 20
+src_subword_alpha: 0.1
+tgt_subword_alpha: 0.1
+
+# Training parameters:
+batch_type: "tokens"
+batch_size: 4096
+valid_batch_size: 16
+batch_size_multiple: 1
+max_generator_batches: 0
+accum_count: [3]
+accum_steps: [0]
+train_steps: 100000
+valid_steps: 5000
+report_every: 100
+save_checkpoint_steps: 10000
+queue_size: 10000
+bucket_size: 32768
+train_from: ./out/fallout/models/fallout_train_step_100000.pt
+reset_optim: all
+
+# Optimization
+model_dtype: "fp32"
+optim: "adam"
+learning_rate: 2
+warmup_steps: 8000
+decay_method: "noam"
+average_decay: 0.0005
+adam_beta2: 0.998
+max_grad_norm: 0
+label_smoothing: 0.1
+param_init: 0
+param_init_glorot: true
+normalization: "tokens"
+
+# Model
+encoder_type: transformer
+decoder_type: transformer
+enc_layers: 6
+dec_layers: 6
+heads: 8
+rnn_size: 512
+word_vec_size: 512
+transformer_ff: 2048
+dropout_steps: [0]
+dropout: [0.1]
+attention_dropout: [0.1]
+position_encoding: true
+
+# Model output:
+save_model: ./out/fallout/models/fallout_tuned
+
+# Logs:
+log_file: ./out/fallout/logs/fallout_tuned
+tensorboard: true
+tensorboard_log_dir: ./out/fallout/tensor/fallout_tuned
+
+# GPU settings:
+world_size: 1
+gpu_ranks: [0]
+
+# Reproducibility:
+seed: 5