def test_generate_dict_textgrid(
multilingual_ipa_tg_corpus_dir,
english_g2p_model,
generated_dir,
temp_dir,
g2p_config_path,
):
if G2P_DISABLED:
pytest.skip("No Pynini found")
output_file = os.path.join(generated_dir, "tg_g2pped.dict")
command = [
"g2p",
english_g2p_model,
multilingual_ipa_tg_corpus_dir,
output_file,
"-t",
temp_dir,
"-q",
"--clean",
"--debug",
"--config_path",
g2p_config_path,
]
args, unknown = parser.parse_known_args(command)
run_g2p(args, unknown)
assert os.path.exists(output_file)
d = MultispeakerDictionary(dictionary_path=output_file, temporary_directory=temp_dir)
d.dictionary_setup()
assert len(d.word_mapping()) > 0
def test_generate_pretrained_threshold(
english_g2p_model, basic_corpus_dir, temp_dir, generated_dir
):
if G2P_DISABLED:
pytest.skip("No Pynini found")
output_path = os.path.join(generated_dir, "g2p_out.txt")
command = [
"g2p",
english_g2p_model,
basic_corpus_dir,
output_path,
"-t",
temp_dir,
"-q",
"--clean",
"--g2p_threshold",
"0.95",
]
args, unknown = parser.parse_known_args(command)
run_g2p(args, unknown)
assert os.path.exists(output_path)
d = MultispeakerDictionary(output_path, temporary_directory=temp_dir)
d.dictionary_setup()
assert len(d.word_mapping(1)) > 0
def test_generate_dict_text_only(
basic_split_dir,
basic_g2p_model_path,
g2p_basic_output,
temp_dir,
g2p_config_path,
):
if G2P_DISABLED:
pytest.skip("No Pynini found")
text_dir = basic_split_dir[1]
command = [
"g2p",
basic_g2p_model_path,
text_dir,
g2p_basic_output,
"-t",
temp_dir,
"-q",
"--clean",
"--debug",
"--config_path",
g2p_config_path,
]
args, unknown = parser.parse_known_args(command)
run_g2p(args, unknown)
assert os.path.exists(g2p_basic_output)
d = MultispeakerDictionary(dictionary_path=g2p_basic_output, temporary_directory=temp_dir)
d.dictionary_setup()
assert len(d.word_mapping()) > 0
def test_classify(basic_corpus_dir, sick_dict_path, english_ivector_model,
generated_dir, transcription_acoustic_model,
transcription_language_model, temp_dir):
output_path = os.path.join(generated_dir, 'classify_test')
command = [
'classify_speakers', basic_corpus_dir, 'english_ivector', output_path,
'-t', temp_dir, '-q', '--clean', '-d', '--disable_mp'
]
args, unknown = parser.parse_known_args(command)
run_classify_speakers(args)
def test_create_segments(basic_corpus_dir, sick_dict_path,
english_acoustic_model, generated_dir,
transcription_acoustic_model,
transcription_language_model, temp_dir,
basic_segment_config):
output_path = os.path.join(generated_dir, 'segment_output')
command = [
'create_segments', basic_corpus_dir, output_path, '-t', temp_dir, '-q',
'--clean', '-d', '--config', basic_segment_config
]
args, unknown = parser.parse_known_args(command)
run_create_segments(args)
