Training gets killed on epoch 0

[DeltaC3f]

Hello everybody !

I'm new to piper tts and to AI in general, but I wanted to make a personnal voice assistant. I bought a Raspberrypi 5 with 8 Go of RAM to train and use the model on it.

The problem is when I start the training, the process is killed with no error message...

I follow all the step of the training guide and plus, I configure the lightning.py file to use 4 workers (I modify both train_dataloader, val_dataloader and test_dataloader).

Python Version : 3.11.2

Argument used :
--devices 1
--batch-size 32
--validation-split 0.01
--num-test-exemples 1
--quality high (I also try in medium and nothing changes)
--max_epochs 10000
--precision 32
--accelerator 'cpu'

Result of the training :

I also tried to finetune a medium model, but I encounter the same problem...
Do you have any Idea from where the error comes from ?

Thanks in advance !