fix tiny and local script, test=asr

examples/aishell/asr0/local/export.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: $0 config_path ckpt_prefix jit_model_path model_type"
+if [ $# != 3 ];then
+    echo "usage: $0 config_path ckpt_prefix jit_model_path"
     exit -1
 fi
 
@@ -11,14 +11,12 @@ echo "using $ngpu gpus..."
 config_path=$1
 ckpt_path_prefix=$2
 jit_model_export_path=$3
-model_type=$4
 
 python3 -u ${BIN_DIR}/export.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --checkpoint_path ${ckpt_path_prefix} \
---export_path ${jit_model_export_path} \
---model_type ${model_type}
+--export_path ${jit_model_export_path}
 
 if [ $? -ne 0 ]; then
     echo "Failed in export!"

examples/aishell/asr0/local/test.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix model_type"
+if [ $# != 3 ];then
+    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix"
     exit -1
 fi
 
@@ -13,7 +13,6 @@ echo "using $ngpu gpus..."
 config_path=$1
 decode_config_path=$2
 ckpt_prefix=$3
-model_type=$4
 
 # download language model
 bash local/download_lm_ch.sh
@@ -23,7 +22,7 @@ fi
 
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     # format the reference test file
-    python utils/format_rsl.py \
+    python3 utils/format_rsl.py \
         --origin_ref data/manifest.test.raw \
         --trans_ref data/manifest.test.text
 
@@ -32,8 +31,7 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     --config ${config_path} \
     --decode_cfg ${decode_config_path} \
     --result_file ${ckpt_prefix}.rsl \
-    --checkpoint_path ${ckpt_prefix} \
-    --model_type ${model_type}
+    --checkpoint_path ${ckpt_prefix}
 
     if [ $? -ne 0 ]; then
         echo "Failed in evaluation!"
@@ -41,25 +39,25 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     fi
 
     # format the hyp file
-    python utils/format_rsl.py \
+    python3 utils/format_rsl.py \
         --origin_hyp ${ckpt_prefix}.rsl \
         --trans_hyp ${ckpt_prefix}.rsl.text
 
-    python utils/compute-wer.py --char=1 --v=1 \
-        data/manifest.test.text ${ckpt_prefix}.rsl.text > ${ckpt_prefix}.error 
+    python3 utils/compute-wer.py --char=1 --v=1 \
+        data/manifest.test.text ${ckpt_prefix}.rsl.text > ${ckpt_prefix}.error
 fi
 
 if [ ${stage} -le 101 ] && [ ${stop_stage} -ge 101 ]; then
-    python utils/format_rsl.py \
+    python3 utils/format_rsl.py \
         --origin_ref data/manifest.test.raw \
         --trans_ref_sclite data/manifest.test.text.sclite
 
-        python utils/format_rsl.py \
-            --origin_hyp ${ckpt_prefix}.rsl \
-            --trans_hyp_sclite ${ckpt_prefix}.rsl.text.sclite
+    python3 utils/format_rsl.py \
+        --origin_hyp ${ckpt_prefix}.rsl \
+        --trans_hyp_sclite ${ckpt_prefix}.rsl.text.sclite
 
-        mkdir -p ${ckpt_prefix}_sclite
-        sclite -i wsj -r data/manifest.test.text.sclite -h  ${ckpt_prefix}.rsl.text.sclite  -e utf-8 -o all -O ${ckpt_prefix}_sclite -c NOASCII
+    mkdir -p ${ckpt_prefix}_sclite
+    sclite -i wsj -r data/manifest.test.text.sclite -h  ${ckpt_prefix}.rsl.text.sclite  -e utf-8 -o all -O ${ckpt_prefix}_sclite -c NOASCII
 fi
 
 exit 0

examples/aishell/asr0/local/test_export.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix model_type"
+if [ $# != 3 ];then
+    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix"
     exit -1
 fi
 
@@ -11,7 +11,6 @@ echo "using $ngpu gpus..."
 config_path=$1
 decode_config_path=$2
 jit_model_export_path=$3
-model_type=$4
 
 # download language model
 bash local/download_lm_ch.sh > /dev/null 2>&1
@@ -24,8 +23,7 @@ python3 -u ${BIN_DIR}/test_export.py \
 --config ${config_path} \
 --decode_cfg ${decode_config_path} \
 --result_file ${jit_model_export_path}.rsl \
---export_path ${jit_model_export_path} \
---model_type ${model_type}
+--export_path ${jit_model_export_path}
 
 if [ $? -ne 0 ]; then
     echo "Failed in evaluation!"

examples/aishell/asr0/local/test_wav.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 5 ];then
-    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix model_type audio_file"
+if [ $# != 4 ];then
+    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix audio_file"
     exit -1
 fi
 
@@ -11,8 +11,7 @@ echo "using $ngpu gpus..."
 config_path=$1
 decode_config_path=$2
 ckpt_prefix=$3
-model_type=$4
-audio_file=$5
+audio_file=$4
 
 mkdir -p data
 wget -nc https://paddlespeech.bj.bcebos.com/datasets/single_wav/zh/demo_01_03.wav -P data/
@@ -37,7 +36,6 @@ python3 -u ${BIN_DIR}/test_wav.py \
 --decode_cfg ${decode_config_path} \
 --result_file ${ckpt_prefix}.rsl \
 --checkpoint_path ${ckpt_prefix} \
---model_type ${model_type} \
 --audio_file ${audio_file}
 
 if [ $? -ne 0 ]; then

examples/aishell/asr0/local/train.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 3 ];then
-    echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name model_type"
+if [ $# != 2 ];then
+    echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name"
     exit -1
 fi
 
@@ -10,7 +10,6 @@ echo "using $ngpu gpus..."
 
 config_path=$1
 ckpt_name=$2
-model_type=$3
 
 mkdir -p exp
 
@@ -25,14 +24,12 @@ python3 -u ${BIN_DIR}/train.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --output exp/${ckpt_name} \
---model_type ${model_type} \
 --seed ${seed}
 else
 python3 -m paddle.distributed.launch --gpus=${CUDA_VISIBLE_DEVICES} ${BIN_DIR}/train.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --output exp/${ckpt_name} \
---model_type ${model_type} \
 --seed ${seed}
 fi
 

examples/aishell/asr0/run.sh

@@ -24,7 +24,7 @@ fi
 
 if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
     # train model, all `ckpt` under `exp` dir
-    CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path}  ${ckpt} ${model_type}
+    CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path}  ${ckpt}
 fi
 
 if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
@@ -34,21 +34,21 @@ fi
 
 if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
     # test ckpt avg_n
-    CUDA_VISIBLE_DEVICES=0 ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${model_type}|| exit -1
+    CUDA_VISIBLE_DEVICES=0 ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}|| exit -1
 fi
 
 if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
     # export ckpt avg_n
-    CUDA_VISIBLE_DEVICES=0 ./local/export.sh ${conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} exp/${ckpt}/checkpoints/${avg_ckpt}.jit ${model_type}
+    CUDA_VISIBLE_DEVICES=0 ./local/export.sh ${conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} exp/${ckpt}/checkpoints/${avg_ckpt}.jit
 fi
 
 if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
     # test export ckpt avg_n
-    CUDA_VISIBLE_DEVICES=0 ./local/test_export.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}.jit ${model_type}|| exit -1
+    CUDA_VISIBLE_DEVICES=0 ./local/test_export.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}.jit|| exit -1
 fi
 
 # Optionally, you can add LM and test it with runtime.
 if [ ${stage} -le 6 ] && [ ${stop_stage} -ge 6 ]; then
     # test a single .wav file
-    CUDA_VISIBLE_DEVICES=0 ./local/test_wav.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${model_type} ${audio_file} || exit -1
+    CUDA_VISIBLE_DEVICES=0 ./local/test_wav.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${audio_file} || exit -1
 fi

examples/librispeech/asr0/local/export.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: $0 config_path ckpt_prefix jit_model_path model_type"
+if [ $# != 3 ];then
+    echo "usage: $0 config_path ckpt_prefix jit_model_path"
     exit -1
 fi
 
@@ -11,14 +11,12 @@ echo "using $ngpu gpus..."
 config_path=$1
 ckpt_path_prefix=$2
 jit_model_export_path=$3
-model_type=$4
 
 python3 -u ${BIN_DIR}/export.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --checkpoint_path ${ckpt_path_prefix} \
---export_path ${jit_model_export_path} \
---model_type ${model_type}
+--export_path ${jit_model_export_path}
 
 if [ $? -ne 0 ]; then
     echo "Failed in export!"

examples/librispeech/asr0/local/test.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix model_type"
+if [ $# != 3 ];then
+    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix"
     exit -1
 fi
 stage=0
@@ -13,7 +13,6 @@ echo "using $ngpu gpus..."
 config_path=$1
 decode_config_path=$2
 ckpt_prefix=$3
-model_type=$4
 
 # download language model
 bash local/download_lm_en.sh
@@ -23,7 +22,7 @@ fi
 
 if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     # format the reference test file
-    python utils/format_rsl.py \
+    python3 utils/format_rsl.py \
         --origin_ref data/manifest.test-clean.raw \
         --trans_ref data/manifest.test-clean.text
 
@@ -32,33 +31,32 @@ if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
     --config ${config_path} \
     --decode_cfg ${decode_config_path} \
     --result_file ${ckpt_prefix}.rsl \
-    --checkpoint_path ${ckpt_prefix} \
-    --model_type ${model_type}
+    --checkpoint_path ${ckpt_prefix}
 
     if [ $? -ne 0 ]; then
         echo "Failed in evaluation!"
         exit 1
     fi
 
-    python utils/format_rsl.py \
+    python3 utils/format_rsl.py \
         --origin_hyp ${ckpt_prefix}.rsl \
         --trans_hyp ${ckpt_prefix}.rsl.text
 
-    python utils/compute-wer.py --char=1 --v=1 \
+    python3 utils/compute-wer.py --char=1 --v=1 \
         data/manifest.test-clean.text ${ckpt_prefix}.rsl.text > ${ckpt_prefix}.error
 fi
 
 if [ ${stage} -le 101 ] && [ ${stop_stage} -ge 101 ]; then
-    python utils/format_rsl.py \
+    python3 utils/format_rsl.py \
         --origin_ref data/manifest.test-clean.raw \
         --trans_ref_sclite data/manifest.test.text-clean.sclite
 
-        python utils/format_rsl.py \
-            --origin_hyp ${ckpt_prefix}.rsl \
-            --trans_hyp_sclite ${ckpt_prefix}.rsl.text.sclite
+    python3 utils/format_rsl.py \
+        --origin_hyp ${ckpt_prefix}.rsl \
+        --trans_hyp_sclite ${ckpt_prefix}.rsl.text.sclite
 
-        mkdir -p ${ckpt_prefix}_sclite
-        sclite -i wsj -r data/manifest.test-clean.text.sclite -h  ${ckpt_prefix}.rsl.text.sclite  -e utf-8 -o all -O ${ckpt_prefix}_sclite -c NOASCII
+    mkdir -p ${ckpt_prefix}_sclite
+    sclite -i wsj -r data/manifest.test-clean.text.sclite -h  ${ckpt_prefix}.rsl.text.sclite  -e utf-8 -o all -O ${ckpt_prefix}_sclite -c NOASCII
 fi
 
 

examples/librispeech/asr0/local/test_wav.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 5 ];then
-    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix model_type audio_file"
+if [ $# != 4 ];then
+    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix audio_file"
     exit -1
 fi
 
@@ -11,8 +11,7 @@ echo "using $ngpu gpus..."
 config_path=$1
 decode_config_path=$2
 ckpt_prefix=$3
-model_type=$4
-audio_file=$5
+audio_file=$4
 
 mkdir -p data
 wget -nc https://paddlespeech.bj.bcebos.com/datasets/single_wav/en/demo_002_en.wav -P data/
@@ -37,7 +36,6 @@ python3 -u ${BIN_DIR}/test_wav.py \
 --decode_cfg ${decode_config_path} \
 --result_file ${ckpt_prefix}.rsl \
 --checkpoint_path ${ckpt_prefix} \
---model_type ${model_type} \
 --audio_file ${audio_file}
 
 if [ $? -ne 0 ]; then

examples/librispeech/asr0/local/train.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 3 ];then
-    echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name model_type"
+if [ $# != 2 ];then
+    echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name"
     exit -1
 fi
 
@@ -10,7 +10,6 @@ echo "using $ngpu gpus..."
 
 config_path=$1
 ckpt_name=$2
-model_type=$3
 
 mkdir -p exp
 
@@ -25,14 +24,12 @@ python3 -u ${BIN_DIR}/train.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --output exp/${ckpt_name} \
---model_type ${model_type} \
 --seed ${seed}
 else
 python3 -m paddle.distributed.launch --gpus=${CUDA_VISIBLE_DEVICES} ${BIN_DIR}/train.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --output exp/${ckpt_name} \
---model_type ${model_type} \
 --seed ${seed}
 fi
 

examples/librispeech/asr0/run.sh

@@ -23,7 +23,7 @@ fi
 
 if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
     # train model, all `ckpt` under `exp` dir
-    CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path}  ${ckpt} ${model_type}
+    CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path}  ${ckpt}
 fi
 
 if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
@@ -33,20 +33,20 @@ fi
 
 if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
     # test ckpt avg_n
-    CUDA_VISIBLE_DEVICES=0 ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${model_type} || exit -1
+    CUDA_VISIBLE_DEVICES=0 ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}|| exit -1
 fi
 
 if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
     # export ckpt avg_n
-    CUDA_VISIBLE_DEVICES= ./local/export.sh ${conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} exp/${ckpt}/checkpoints/${avg_ckpt}.jit ${model_type}
+    CUDA_VISIBLE_DEVICES= ./local/export.sh ${conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} exp/${ckpt}/checkpoints/${avg_ckpt}.jit
 fi
 
 if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
     # test export ckpt avg_n
-    CUDA_VISIBLE_DEVICES=0 ./local/test_export.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}.jit ${model_type}|| exit -1
+    CUDA_VISIBLE_DEVICES=0 ./local/test_export.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}.jit|| exit -1
 fi
 
 if [ ${stage} -le 6 ] && [ ${stop_stage} -ge 6 ]; then
     # test a single .wav file
-    CUDA_VISIBLE_DEVICES=0 ./local/test_wav.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${model_type} ${audio_file} || exit -1
+    CUDA_VISIBLE_DEVICES=0 ./local/test_wav.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${audio_file} || exit -1
 fi

examples/tiny/asr0/conf/augmentation.json

@@ -1,36 +0,0 @@
-[
-  {
-    "type": "speed",
-    "params": {
-      "min_speed_rate": 0.9,
-      "max_speed_rate": 1.1,
-      "num_rates": 3
-    },
-    "prob": 0.0
-  },
-  {
-    "type": "shift",
-    "params": {
-      "min_shift_ms": -5,
-      "max_shift_ms": 5
-    },
-    "prob": 1.0
-  },
-  {
-    "type": "specaug",
-    "params": {
-      "W": 5,
-      "warp_mode": "PIL",
-      "F": 30,
-      "n_freq_masks": 2,
-      "T": 40,
-      "n_time_masks": 2,
-      "p": 1.0,
-      "adaptive_number_ratio": 0,
-      "adaptive_size_ratio": 0,
-      "max_n_time_masks": 20,
-      "replace_with_zero": true
-    },
-    "prob": 1.0
-  }
-]

examples/tiny/asr0/conf/deepspeech2.yaml

@@ -16,28 +16,26 @@ max_output_input_ratio: 10.0
 ###########################################
 #              Dataloader                 #
 ###########################################
-mean_std_filepath: data/mean_std.json
-unit_type: char
-vocab_filepath: data/lang_char/vocab.txt
-augmentation_config: conf/augmentation.json
-random_seed: 0
-spm_model_prefix: 
-spectrum_type: linear
+vocab_filepath: data/lang_char/vocab.txt 
+spm_model_prefix: ''
+unit_type: 'char'
+preprocess_config: conf/preprocess.yaml
 feat_dim: 161
-delta_delta: False
 stride_ms: 10.0
-window_ms: 20.0
-n_fft: None
-max_freq: None
-target_sample_rate: 16000
-use_dB_normalization: True
-target_dB: -20
-dither: 1.0
-keep_transcription_text: False
-sortagrad: True 
-shuffle_method: batch_shuffle
-num_workers: 2
+window_ms: 25.0
+sortagrad: 0 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs 
 batch_size: 4
+maxlen_in: 512  # if input length  > maxlen-in, batchsize is automatically reduced
+maxlen_out: 150  # if output length > maxlen-out, batchsize is automatically reduced
+minibatches: 0 # for debug
+batch_count: auto
+batch_bins: 0 
+batch_frames_in: 0
+batch_frames_out: 0
+batch_frames_inout: 0
+num_workers: 8
+subsampling_factor: 1
+num_encs: 1
   
 ############################################
 #           Network Architecture           #
@@ -45,8 +43,10 @@ batch_size: 4
 num_conv_layers: 2
 num_rnn_layers: 3
 rnn_layer_size: 2048
+rnn_direction: bidirect # [forward, bidirect]
+num_fc_layers: 0
+fc_layers_size_list: -1,
 use_gru: False 
-share_rnn_weights: True 
 blank_id: 0
   
 
@@ -59,6 +59,7 @@ lr: 1.0e-5
 lr_decay: 0.8 
 weight_decay: 1.0e-6
 global_grad_clip: 5.0
+dist_sampler: False
 log_interval: 1
 checkpoint:
   kbest_n: 3

examples/tiny/asr0/conf/deepspeech2_online.yaml

@@ -16,29 +16,27 @@ max_output_input_ratio: 10.0
 ###########################################
 #              Dataloader                 #
 ###########################################
-mean_std_filepath: data/mean_std.json
-unit_type: char
-vocab_filepath: data/lang_char/vocab.txt
-augmentation_config: conf/augmentation.json
-random_seed: 0
-spm_model_prefix: 
-spectrum_type: linear
+vocab_filepath: data/lang_char/vocab.txt 
+spm_model_prefix: ''
+unit_type: 'char'
+preprocess_config: conf/preprocess.yaml
 feat_dim: 161
-delta_delta: False
 stride_ms: 10.0
-window_ms: 20.0
-n_fft: None
-max_freq: None
-target_sample_rate: 16000
-use_dB_normalization: True
-target_dB: -20
-dither: 1.0
-keep_transcription_text: False
-sortagrad: True 
-shuffle_method: batch_shuffle
-num_workers: 0
+window_ms: 25.0
+sortagrad: 0 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs 
 batch_size: 4
-  
+maxlen_in: 512  # if input length  > maxlen-in, batchsize is automatically reduced
+maxlen_out: 150  # if output length > maxlen-out, batchsize is automatically reduced
+minibatches: 0 # for debug
+batch_count: auto
+batch_bins: 0 
+batch_frames_in: 0
+batch_frames_out: 0
+batch_frames_inout: 0
+num_workers: 8
+subsampling_factor: 1
+num_encs: 1
+ 
 ############################################
 #           Network Architecture           #
 ############################################
@@ -61,6 +59,7 @@ lr: 1.0e-5
 lr_decay: 1.0 
 weight_decay: 1.0e-6
 global_grad_clip: 5.0
+dist_sampler: False
 log_interval: 1
 checkpoint:
   kbest_n: 3

examples/tiny/asr0/local/export.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: $0 config_path ckpt_prefix jit_model_path model_type"
+if [ $# != 3 ];then
+    echo "usage: $0 config_path ckpt_prefix jit_model_path"
     exit -1
 fi
 
@@ -11,14 +11,12 @@ echo "using $ngpu gpus..."
 config_path=$1
 ckpt_path_prefix=$2
 jit_model_export_path=$3
-model_type=$4
 
 python3 -u ${BIN_DIR}/export.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --checkpoint_path ${ckpt_path_prefix} \
---export_path ${jit_model_export_path} \
---model_type ${model_type}
+--export_path ${jit_model_export_path}
 
 if [ $? -ne 0 ]; then
     echo "Failed in export!"

examples/tiny/asr0/local/test.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
-if [ $# != 4 ];then
-    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix model_type"
+if [ $# != 3 ];then
+    echo "usage: ${0} config_path decode_config_path ckpt_path_prefix"
     exit -1
 fi
 
@@ -11,7 +11,6 @@ echo "using $ngpu gpus..."
 config_path=$1
 decode_config_path=$2
 ckpt_prefix=$3
-model_type=$4
 
 # download language model
 bash local/download_lm_en.sh
@@ -24,8 +23,7 @@ python3 -u ${BIN_DIR}/test.py \
 --config ${config_path} \
 --decode_cfg ${decode_config_path} \
 --result_file ${ckpt_prefix}.rsl \
---checkpoint_path ${ckpt_prefix} \
---model_type ${model_type}
+--checkpoint_path ${ckpt_prefix}
 
 if [ $? -ne 0 ]; then
     echo "Failed in evaluation!"

examples/tiny/asr0/local/train.sh

@@ -15,14 +15,13 @@ if [ ${seed} != 0  ]; then
     echo "using seed $seed & FLAGS_cudnn_deterministic=True ..."
 fi
 
-if [ $# != 3 ];then
-    echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name model_type"
+if [ $# != 2 ];then
+    echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name"
     exit -1
 fi
 
 config_path=$1
 ckpt_name=$2
-model_type=$3
 
 mkdir -p exp
 
@@ -31,7 +30,6 @@ python3 -u ${BIN_DIR}/train.py \
 --ngpu ${ngpu} \
 --config ${config_path} \
 --output exp/${ckpt_name} \
---model_type ${model_type} \
 --profiler-options "${profiler_options}" \
 --seed ${seed}
 else
@@ -39,7 +37,6 @@ python3 -m paddle.distributed.launch --gpus=${CUDA_VISIBLE_DEVICES} ${BIN_DIR}/t
 --ngpu ${ngpu} \
 --config ${config_path} \
 --output exp/${ckpt_name} \
---model_type ${model_type} \
 --profiler-options "${profiler_options}" \
 --seed ${seed}
 fi

examples/tiny/asr0/run.sh

@@ -8,8 +8,6 @@ stop_stage=100
 conf_path=conf/deepspeech2.yaml
 decode_conf_path=conf/tuning/decode.yaml
 avg_num=1
-model_type=offline
-
 source ${MAIN_ROOT}/utils/parse_options.sh || exit 1;
 
 avg_ckpt=avg_${avg_num}
@@ -23,7 +21,7 @@ fi
 
 if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
     # train model, all `ckpt` under `exp` dir
-    CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path}  ${ckpt} ${model_type}
+    CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path} ${ckpt}
 fi
 
 if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
@@ -33,10 +31,10 @@ fi
 
 if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
     # test ckpt avg_n
-    CUDA_VISIBLE_DEVICES=${gpus} ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${model_type} || exit -1
+    CUDA_VISIBLE_DEVICES=${gpus} ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt}|| exit -1
 fi
 
 if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
     # export ckpt avg_n
-    CUDA_VISIBLE_DEVICES=${gpus} ./local/export.sh ${conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} exp/${ckpt}/checkpoints/${avg_ckpt}.jit ${model_type}
+    CUDA_VISIBLE_DEVICES=${gpus} ./local/export.sh ${conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} exp/${ckpt}/checkpoints/${avg_ckpt}.jit
 fi

paddlespeech/s2t/exps/deepspeech2/model.py

@@ -248,7 +248,6 @@ class DeepSpeech2Tester(DeepSpeech2Trainer):
         for text, n in zip(texts, texts_len):
             n = n.numpy().item()
             ids = text[:n]
-            #trans.append(''.join([chr(i) for i in ids]))
             trans.append(self._text_featurizer.defeaturize(ids.numpy().tolist()))
         return trans