modify the shell scripts

egs/librispeech/asr/conf/train_ctc_debug.yaml

@@ -12,7 +12,7 @@ log-interval: 100
 seed: 1
 report-accuracy: True
 
-arch: s2t_conformer_s
+arch: s2t_transformer_s
 share-decoder-input-output-embed: True
 optimizer: adam
 clip-norm: 10.0

egs/lower/mt/train.sh

@@ -12,7 +12,7 @@ extra_parameter=
 #extra_tag="${extra_tag}"
 #extra_parameter="${extra_parameter} "
 
-exp_tag=baseline_997
+exp_tag=baseline
 train_config=train.yaml
 
 cmd="./run.sh

egs/mustc/asr/run.sh

@@ -93,8 +93,8 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     ### Task dependent. You have to make data the following preparation part by yourself.
     ### But you can utilize Kaldi recipes in most cases
     echo "stage 0: ASR Data Preparation"
-    if [[ ! -e ${data_dir} ]]; then
-        mkdir -p ${data_dir}
+    if [[ ! -e ${data_dir}/${lang} ]]; then
+        mkdir -p ${data_dir}/${lang}
     fi
 
     cmd="python ${root_dir}/examples/speech_to_text/prep_mustc_data.py

egs/mustc/mt/conf/train.yaml

 train-subset: train
-valid-subset: dev
+valid-subset: valid
 
 max-epoch: 50
 max-update: 100000

egs/mustc/mt/decode.sh

@@ -2,7 +2,7 @@
 
 gpu_num=1
 
-test_subset=(tst-COMMON)
+test_subset=(test)
 
 exp_name=
 if [ "$#" -eq 1 ]; then

egs/mustc/mt/run.sh

@@ -43,7 +43,7 @@ data_dir=~/st/data/${dataset}/mt/${lang}
 train_prefix=train
 valid_prefix=dev
 test_prefix=tst-COMMON
-test_subset=(tst-COMMON)
+test_subset=(test)
 
 # exp
 extra_tag=
@@ -148,9 +148,9 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
 
     cmd="python ${root_dir}/fairseq_cli/preprocess.py
         --source-lang ${src_lang} --target-lang ${tgt_lang}
-        --trainpref ${data_dir}/data/train/${train_prefix}
-        --validpref ${data_dir}/data/dev/${valid_prefix}
-        --testpref ${data_dir}/data/test/${test_prefix}
+        --trainpref ${data_dir}/data/${train_prefix}
+        --validpref ${data_dir}/data/${valid_prefix}
+        --testpref ${data_dir}/data/${test_prefix}
         --destdir ${data_dir}/data-bin
         --srcdict ${data_dir}/${src_vocab_prefix}.txt
         --tgtdict ${data_dir}/${tgt_vocab_prefix}.txt

egs/mustc/st/conf/train.yaml

@@ -33,7 +33,7 @@ dropout: 0.1
 activation-fn: relu
 encoder-embed-dim: 256
 encoder-ffn-embed-dim: 2048
-encoder-layers: 6
+encoder-layers: 12
 decoder-layers: 6
 encoder-attention-heads: 4
 

egs/mustc/st/conf/train_ctc.yaml

@@ -34,7 +34,7 @@ dropout: 0.1
 activation-fn: relu
 encoder-embed-dim: 256
 encoder-ffn-embed-dim: 2048
-encoder-layers: 6
+encoder-layers: 12
 decoder-layers: 6
 encoder-attention-heads: 4
 

egs/mustc/st/conf/train_ctc_conformer.yaml

@@ -34,7 +34,7 @@ dropout: 0.1
 activation-fn: relu
 encoder-embed-dim: 256
 encoder-ffn-embed-dim: 2048
-encoder-layers: 6
+encoder-layers: 12
 decoder-layers: 6
 encoder-attention-heads: 4
 

egs/mustc/st/conf/train_ctc_enc_rpr.yaml

@@ -34,7 +34,7 @@ dropout: 0.1
 activation-fn: relu
 encoder-embed-dim: 256
 encoder-ffn-embed-dim: 2048
-encoder-layers: 6
+encoder-layers: 12
 decoder-layers: 6
 encoder-attention-heads: 4
 encoder-attention-type: rel_selfattn