Espnet: failed at loading the model when decoding on fastspeech.v3.single

[JoeyHeisenberg]

I try to train & decode on csmsc/conf/tuning/fastspeech.v3.single, but failed at decoding,
even with the model downloading from
!./espnet/utils/download_from_google_drive.sh \
https://drive.google.com/open?id=1T8thxkAxjGFPXPWPTcKLvHnd6lG0-82R downloads tar.gz > /dev/null 2>&1

[kan-bayashi]

What is your espnet version?
v3 config needs espnet>=0.5.3.

And could you show me your decoding command on the top of decode.*.log?
e.g.

# tts_decode.py --backend pytorch --ngpu 0 --verbose 0 --out exp/train_no_dev_pytorch_train_fastspeech.v3.single/outputs_model.last1.avg.best_decode/eval/feats.10 --json exp/train_no_dev_pytorch_train_fastspeech.v3.single/outputs_model.last1.avg.best_decode/eval/split32utt/data.10.json --model exp/train_no_dev_pytorch_train_fastspeech.v3.single/results/model.last1.avg.best --config conf/decode.yaml 

[JoeyHeisenberg]

tts_decode.py --backend pytorch --ngpu 0 --verbose 0 --out exp/train_no_dev_pytorch_train_fastspeech.v3.single/outputs_model.last1.avg.best_decode/dev/feats.1 --json exp/train_no_dev_pytorch_train_fastspeech.v3.single/outputs_model.last1.avg.best_decode/dev/split16utt/data.1.json --model exp/train_no_dev_pytorch_train_fastspeech.v3.single/results/model.last1.avg.best --config conf/decode.yaml

[JoeyHeisenberg]

Thx， @kan-bayashi After I update the version it did work！