experiments
/
STT
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Make language model scoring optional in Python inference code

This commit is contained in:
Reuben Morais 2019-09-30 11:43:00 +02:00
parent 14c0db7294
commit 4302a5f767
3 changed files with 12 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
DeepSpeech.py
View File

@ -870,9 +870,12 @@ def do_single_file_inference(input_file_path):
logits = np.squeeze(logits)
scorer = Scorer(FLAGS.lm_alpha, FLAGS.lm_beta,
FLAGS.lm_binary_path, FLAGS.lm_trie_path,
Config.alphabet)
if FLAGS.lm_binary_path:
scorer = Scorer(FLAGS.lm_alpha, FLAGS.lm_beta,
FLAGS.lm_binary_path, FLAGS.lm_trie_path,
Config.alphabet)
else:
scorer = None
decoded = ctc_beam_search_decoder(logits, Config.alphabet, FLAGS.beam_width, scorer=scorer)
# Print highest probability result
print(decoded[0][1])

9
evaluate.py
View File

@ -42,9 +42,12 @@ def sparse_tuple_to_texts(sp_tuple, alphabet):
def evaluate(test_csvs, create_model, try_loading):
scorer = Scorer(FLAGS.lm_alpha, FLAGS.lm_beta,
FLAGS.lm_binary_path, FLAGS.lm_trie_path,
Config.alphabet)
if FLAGS.lm_binary_path:
scorer = Scorer(FLAGS.lm_alpha, FLAGS.lm_beta,
FLAGS.lm_binary_path, FLAGS.lm_trie_path,
Config.alphabet)
else:
scorer = None
test_csvs = FLAGS.test_files.split(',')
test_sets = [create_dataset([csv], batch_size=FLAGS.test_batch_size, train_phase=False) for csv in test_csvs]

8
util/flags.py
View File

@ -143,14 +143,6 @@ def create_flags():
# Register validators for paths which require a file to be specified
f.register_validator('lm_binary_path',
os.path.isfile,
message='The file pointed to by --lm_binary_path must exist and be readable.')
f.register_validator('lm_trie_path',
os.path.isfile,
message='The file pointed to by --lm_trie_path must exist and be readable.')
f.register_validator('alphabet_config_path',
os.path.isfile,
message='The file pointed to by --alphabet_config_path must exist and be readable.')