fix tools

env.sh

@@ -0,0 +1,10 @@
+export MAIN_ROOT=${PWD}
+
+export PATH=${MAIN_ROOT}:${PWD}/tools:${PATH}
+export LC_ALL=C
+
+# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
+export PYTHONIOENCODING=UTF-8 
+export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
+
+export export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/lib/

examples/aishell/conf/deepspeech2.yaml

@@ -29,9 +29,9 @@ model:
   use_gru: True 
   share_rnn_weights: False 
 training:
-  n_epoch: 20
+  n_epoch: 30
   lr: 5e-4
-  lr_decay: 1.0
+  lr_decay: 0.83
   weight_decay: 1e-06
   global_grad_clip: 5.0
   max_iteration: 500000

examples/aishell/local/tune.sh

@@ -0,0 +1,28 @@
+#! /usr/bin/env bash
+
+# grid-search for hyper-parameters in language model
+python3 -u ${MAIN_ROOT}/tune.py \
+--device 'gpu' \
+--nproc 1 \
+--config conf/deepspeech2.yaml \
+--num_batches=-1 \
+--batch_size=128 \
+--beam_size=500 \
+--num_proc_bsearch=12 \
+--num_alphas=45 \
+--num_betas=8 \
+--alpha_from=1.0 \
+--alpha_to=3.2 \
+--beta_from=0.1 \
+--beta_to=0.45 \
+--cutoff_prob=1.0 \
+--cutoff_top_n=40 \
+--checkpoint_path ${1} 
+
+if [ $? -ne 0 ]; then
+    echo "Failed in tuning!"
+    exit 1
+fi
+
+
+exit 0

examples/aishell/path.sh

@@ -6,3 +6,5 @@ export LC_ALL=C
 # Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
 export PYTHONIOENCODING=UTF-8 
 export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
+
+export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/lib/

examples/librispeech/path.sh

@@ -6,3 +6,5 @@ export LC_ALL=C
 # Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
 export PYTHONIOENCODING=UTF-8 
 export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
+
+export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/lib/

examples/tiny/path.sh

@@ -6,3 +6,5 @@ export LC_ALL=C
 # Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
 export PYTHONIOENCODING=UTF-8 
 export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
+
+export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/lib/

model_utils/network.py

@@ -281,8 +281,8 @@ class GRUCellShare(nn.RNNCellBase):
         self.hidden_size = hidden_size
         self.input_size = input_size
         self._gate_activation = F.sigmoid
-        #self._activation = paddle.tanh
-        self._activation = F.relu
+        self._activation = paddle.tanh
+        #self._activation = F.relu
 
     def forward(self, inputs, states=None):
         if states is None:

setup.sh

@@ -29,7 +29,7 @@ if [ $? != 0 ]; then
     fi
     tar -zxvf libsndfile-1.0.28.tar.gz
     cd libsndfile-1.0.28
-    ./configure > /dev/null && make > /dev/null && make install > /dev/null
+    ./configure > /dev/null; make > /dev/null; make install > /dev/null
     cd ..
     rm -rf libsndfile-1.0.28
     rm libsndfile-1.0.28.tar.gz

tests/network_test.py

@@ -27,7 +27,7 @@ if __name__ == '__main__':
     text = np.array([[1, 2], [1, 2]], dtype='int32')
     text_len = np.array([2] * batch_size, dtype='int32')
 
-    place = paddle.CUDAPinnedPlace()
+    place = paddle.CUDAPlace(0)
     audio = paddle.to_tensor(
         audio, dtype='float32', place=place, stop_gradient=True)
     audio_len = paddle.to_tensor(

tools/build_vocab.py

@@ -22,7 +22,6 @@ import codecs
 import json
 from collections import Counter
 import os.path
-import _init_paths
 from data_utils.utility import read_manifest
 from utils.utility import add_arguments, print_arguments
 

tools/compute_mean_std.py

@@ -15,7 +15,6 @@
 
 import argparse
 import functools
-import _init_paths
 from data_utils.normalizer import FeatureNormalizer
 from data_utils.augmentor.augmentation import AugmentationPipeline
 from data_utils.featurizer.audio_featurizer import AudioFeaturizer