Evers

if chinese, please try PR: https://github.com/ming024/FastSpeech2/pull/153

I meet the same issue, yes, I use my own lexicon and did MFA train my own dataset, my total sentence is 88770 (aishell3 + my own 735), however duration/energy/mel/pitch...

https://github.com/ming024/FastSpeech2/pull/153 I have verified, my training is ongoing now...

got same issue when I try "python3 preprocess.py config/AISHELL3/preprocess.yaml" on my updated AISHELL3 dataset (one more speaker was added)

it can run on the first time, but failed for second try

aarch64/arm passed: evers@raspberrypi:~/jemalloc $ getconf PAGESIZE 4096 evers@raspberrypi:~/jemalloc $ lscpu Architecture: aarch64 Byte Order: Little Endian CPU(s): 4 On-line CPU(s) list: 0-3 Thread(s) per core: 1 Core(s) per socket: 4...

myenv: evers@raspberrypi:~/jemalloc $ uname -a Linux raspberrypi 6.1.21-v8+ #1642 SMP PREEMPT Mon Apr 3 17:24:16 BST 2023 aarch64 GNU/Linux

there's 78 cores? maybe you can restrict arena number, it will use much more memory if increase arena number.