Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #9 from eginhard/disable-wavegrad-test 

test(vocoder): disable wavegrad training test in CI
This commit is contained in:
Enno Hermann 2024-03-08 19:23:40 +01:00 committed by GitHub
parent f782a65c4c dca564a705
commit ec2346099d
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 43 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

75
tests/vocoder_tests/test_wavegrad_train.py
View File

@ -1,43 +1,54 @@
import glob import glob
import os import os
import shutil import shutil
import unittest
from tests import get_device_id, get_tests_output_path, run_cli from tests import get_device_id, get_tests_output_path, run_cli
from TTS.vocoder.configs import WavegradConfig from TTS.vocoder.configs import WavegradConfig
config_path = os.path.join(get_tests_output_path(), "test_vocoder_config.json")
output_path = os.path.join(get_tests_output_path(), "train_outputs")
config = WavegradConfig( class WavegradTrainingTest(unittest.TestCase):
batch_size=8, # TODO: Reactivate after improving CI run times
eval_batch_size=8, # This test currently takes ~2h on CI (15min/step vs 8sec/step locally)
num_loader_workers=0, if os.getenv("GITHUB_ACTIONS") == "true":
num_eval_loader_workers=0, __test__ = False
run_eval=True,
test_delay_epochs=-1,
epochs=1,
seq_len=8192,
eval_split_size=1,
print_step=1,
print_eval=True,
data_path="tests/data/ljspeech",
output_path=output_path,
test_noise_schedule={"min_val": 1e-6, "max_val": 1e-2, "num_steps": 2},
)
config.audio.do_trim_silence = True
config.audio.trim_db = 60
config.save_json(config_path)
# train the model for one epoch def test_train(self): # pylint: disable=no-self-use
command_train = f"CUDA_VISIBLE_DEVICES='{get_device_id()}' python TTS/bin/train_vocoder.py --config_path {config_path} " config_path = os.path.join(get_tests_output_path(), "test_vocoder_config.json")
run_cli(command_train) output_path = os.path.join(get_tests_output_path(), "train_outputs")
# Find latest folder config = WavegradConfig(
continue_path = max(glob.glob(os.path.join(output_path, "*/")), key=os.path.getmtime) batch_size=8,
eval_batch_size=8,
num_loader_workers=0,
num_eval_loader_workers=0,
run_eval=True,
test_delay_epochs=-1,
epochs=1,
seq_len=8192,
eval_split_size=1,
print_step=1,
print_eval=True,
data_path="tests/data/ljspeech",
output_path=output_path,
test_noise_schedule={"min_val": 1e-6, "max_val": 1e-2, "num_steps": 2},
)
config.audio.do_trim_silence = True
config.audio.trim_db = 60
config.save_json(config_path)
# restore the model and continue training for one more epoch # train the model for one epoch
command_train = ( command_train = (
f"CUDA_VISIBLE_DEVICES='{get_device_id()}' python TTS/bin/train_vocoder.py --continue_path {continue_path} " f"CUDA_VISIBLE_DEVICES='{get_device_id()}' python TTS/bin/train_vocoder.py --config_path {config_path} "
) )
run_cli(command_train) run_cli(command_train)
shutil.rmtree(continue_path)
# Find latest folder
continue_path = max(glob.glob(os.path.join(output_path, "*/")), key=os.path.getmtime)
# restore the model and continue training for one more epoch
command_train = (
f"CUDA_VISIBLE_DEVICES='{get_device_id()}' python TTS/bin/train_vocoder.py --continue_path {continue_path} "
)
run_cli(command_train)
shutil.rmtree(continue_path)