Error in phonemes production

Hi, maybe someone had the same issue with phonemes creation.
Phonemes are generated during the 1st epoch. I am trying training with bigger dataset (~20h), however, training is suspended during the 1st epoch:

! Run is removed from /home/marta/Desktop/inz/test/TTS/./results/ljspeech-January-15-2020_05+52PM-7eb291c
Traceback (most recent call last):
File “/home/marta/Desktop/inz/test/TTS/datasets/”, line 94, in _load_or_generate_phoneme_sequence
phonemes = np.load(cache_path)
File “/home/marta/Desktop/Desktop/inż/TTS/.eggs/numpy-1.15.4-py3.6-linux-x86_64.egg/numpy/lib/”, line 384, in load
fid = open(file, “rb”)
FileNotFoundError: [Errno 2] No such file or directory: ‘mozilla_us_phonemes/lalka_2_08_f000163_phoneme.npy’

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File “”, line 704, in
File “”, line 615, in main
global_step, epoch)
File “”, line 100, in train
for num_iter, data in enumerate(data_loader):
File “/home/marta/.local/lib/python3.6/site-packages/torch/utils/data/”, line 346, in next
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File “/home/marta/.local/lib/python3.6/site-packages/torch/utils/data/_utils/”, line 44, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File “/home/marta/.local/lib/python3.6/site-packages/torch/utils/data/_utils/”, line 44, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File “/home/marta/Desktop/inz/test/TTS/datasets/”, line 164, in getitem
return self.load_data(idx)
File “/home/marta/Desktop/inz/test/TTS/datasets/”, line 113, in load_data
text = self._load_or_generate_phoneme_sequence(wav_file, text)
File “/home/marta/Desktop/inz/test/TTS/datasets/”, line 97, in _load_or_generate_phoneme_sequence
File “/home/marta/Desktop/inz/test/TTS/datasets/”, line 84, in _generate_and_cache_phoneme_sequence
File “/home/marta/Desktop/inz/test/TTS/utils/text/”, line 57, in phoneme_to_sequence
to_phonemes = text2phone(clean_text, language)
File “/home/marta/Desktop/inz/test/TTS/utils/text/”, line 31, in text2phone
ph = phonemize(text, separator=seperator, strip=False, njobs=1, backend=‘espeak’, language=language)
File “/home/marta/.local/lib/python3.6/site-packages/phonemizer/”, line 149, in phonemize
File “/home/marta/.local/lib/python3.6/site-packages/phonemizer/backend/”, line 42, in init
super(self.class, self).init(language, logger=logger)
File “/home/marta/.local/lib/python3.6/site-packages/phonemizer/backend/”, line 43, in init
‘initializing backend %s-%s’,, self.version())
File “/home/marta/.local/lib/python3.6/site-packages/phonemizer/backend/”, line 104, in version
long_version = cls.long_version()
File “/home/marta/.local/lib/python3.6/site-packages/phonemizer/backend/”, line 92, in long_version
‘{} --help’.format(cls.espeak_exe()), posix=False)).decode(
File “/usr/lib/python3.6/”, line 356, in check_output
File “/usr/lib/python3.6/”, line 423, in run
with Popen(*popenargs, **kwargs) as process:
File “/usr/lib/python3.6/”, line 729, in init
restore_signals, start_new_session)
File “/usr/lib/python3.6/”, line 1295, in _execute_child
restore_signals, start_new_session, preexec_fn)
OSError: [Errno 12] Cannot allocate memory

I checked the dataset with the notebook and everything was fine.
I checked the file with phonemes generated and some that should be generated, are not. I am also attaching config.json file

Have you read the error message?

You are out of memory
“OSError: [Errno 12] Cannot allocate memory”

How large is your system memory ?

I have read it, however, I still control the memory with System Monitor and it hasn’t reached the peak. Maybe NVIDIA memory is full. And swap memory is empty, i.e. 0, so I have no idea, what I should do.

Still, the same, after

sudo apt purge ‘nvidia*’

Anyway, it crashed everything, so I am not able to train anything anymore.