Merge pull request #347 from kuke/coding_convert

Convert decoding results to unicode in DS2

Author: Yibing Liu

deep_speech_2/decoders/swig_wrapper.py

@@ -35,7 +35,8 @@ def ctc_greedy_decoder(probs_seq, vocabulary):
  :return: Decoding result string.
  :rtype: basestring
  """
- return swig_decoders.ctc_greedy_decoder(probs_seq.tolist(), vocabulary)
+ result = swig_decoders.ctc_greedy_decoder(probs_seq.tolist(), vocabulary)
+ return result.decode('utf-8')
 
 
 def ctc_beam_search_decoder(probs_seq,
@@ -69,9 +70,11 @@ def ctc_beam_search_decoder(probs_seq,
  results, in descending order of the probability.
  :rtype: list
  """
- return swig_decoders.ctc_beam_search_decoder(probs_seq.tolist(), vocabulary,
- beam_size, cutoff_prob,
- cutoff_top_n, ext_scoring_func)
+ beam_results = swig_decoders.ctc_beam_search_decoder(
+ probs_seq.tolist(), vocabulary, beam_size, cutoff_prob, cutoff_top_n,
+ ext_scoring_func)
+ beam_results = [(res[0], res[1].decode('utf-8')) for res in beam_results]
+ return beam_results
 
 
 def ctc_beam_search_decoder_batch(probs_split,
@@ -111,6 +114,11 @@ def ctc_beam_search_decoder_batch(probs_split,
  """
  probs_split = [probs_seq.tolist() for probs_seq in probs_split]
 
- return swig_decoders.ctc_beam_search_decoder_batch(
+ batch_beam_results = swig_decoders.ctc_beam_search_decoder_batch(
  probs_split, vocabulary, beam_size, num_processes, cutoff_prob,
  cutoff_top_n, ext_scoring_func)
+ batch_beam_results = [
+ [(res[0], res[1].decode("utf-8")) for res in beam_results]
+ for beam_results in batch_beam_results
+ ]
+ return batch_beam_results

deep_speech_2/examples/tiny/run_train.sh

@@ -33,7 +33,7 @@ python -u train.py \
 --shuffle_method='batch_shuffle_clipped'
 
 if [ $? -ne 0 ]; then
- echo "Fail to do inference!"
+ echo "Fail in training!"
  exit 1
 fi
 

deep_speech_2/train.py

@@ -118,7 +118,9 @@ def train():
 
 def main():
  print_arguments(args)
- paddle.init(use_gpu=args.use_gpu, trainer_count=args.trainer_count)
+ paddle.init(use_gpu=args.use_gpu,
+ trainer_count=args.trainer_count,
+ log_clipping=True)
  train()