search

mozilla / TTS

:robot: :speech_balloon: Deep learning for Text to Speech (Discussion forum: https://discourse.mozilla.org/c/tts)
Mozilla Public License 2.0
9.29k stars 1.24k forks source link

Error during training: json.decoder.JSONDecodeError: Invalid \escape: line 3 column 78 (char 107) #671

Closed luis-vera closed 3 years ago

luis-vera commented 3 years ago

Hi

While I was training I got erro message about RunTimeError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 11.00 GiB total capacity; 8.11 GiB already allocated; 7.74 MiB free; 8.30 GiB reserved in total by PyTorch)

I found this error is solved if I reduce batch size. Now I was using batch_size = 4 and eval_batch_size = 4. But when I reduced this value and I executed:

python TTS/bin/train_tacotron.py --continue_path Models/LJSpeech/ljspeech-ddc-February-19-2021_11+37AM-e9e0784

(README indicates I must type run directory)

The error was:

**2021-02-25 10:32:05.110659: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudart64_101.dll

Using CUDA: True Number of GPUs: 1 Training continues for Models/LJSpeech/ljspeech-ddc-February-19-2021_11+37AM-e9e0784 Traceback (most recent call last): File "TTS/bin/train_tacotron.py", line 688, in c = load_config(args.config_path) File "c:\users\voice-trainner\proyecto\tts\TTS\utils\io.py", line 46, in load_config data = read_json_with_comments(config_path) File "c:\users\voice-trainner\proyecto\tts\TTS\utils\io.py", line 30, in read_json_with_comments data = json.loads(input_str) File "C:\Users\Voice-trainner\anaconda3\envs\TF2\lib\json__init__.py", line 357, in loads return _default_decoder.decode(s) File "C:\Users\Voice-trainner\anaconda3\envs\TF2\lib\json\decoder.py", line 337, in decode obj, end = self.raw_decode(s, idx=_w(s, 0).end()) File "C:\Users\Voice-trainner\anaconda3\envs\TF2\lib\json\decoder.py", line 353, in raw_decode obj, end = self.scan_once(s, idx) json.decoder.JSONDecodeError: Invalid \escape: line 3 column 78 (char 107)**

What it could have happened? Thanks a lot for your help

luis-vera commented 3 years ago

The weird thing that is not a json file problem because I execute train_tacotron.py from zero again and the training stage started without problems. So maybe It's not a json problem.

stale[bot] commented 3 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions. You might also look our discourse page for further help. https://discourse.mozilla.org/c/tts