DiffSinger I can't get checkpoint files.

According to README-SVS-opencpop-cascade.md, I made my own datasets and tried training.

CUDA_VISIBLE_DEVICES=0 python tasks/run.py --config usr/configs/midi/cascade/mydatasets/aux_rel.yaml --exp_name MY_DATASETS_FS_EXP --reset

The training went to Epoch 50 but I didn't any checkpoint files. Which is the checkpoint file, or could anyone tell me which config should be changed to make checkpoint files (in shorter interval)?

Sep 10 '22 05:09 nakasako

It shows Epoch 81: : 160001batch [16:58:42, 1.31batch/s, batch_size=8, f0=0.0126, l1=0.0221, lr=0.000313, pdur=0.025, sdur=0.00533, ssim=0.0182, step=160000, uv=4.2e-5, wdur=0.0204] | Training end.. But it seems there are no files found.

Sep 10 '22 05:09 nakasako

Me too. I try PopCS and OpenCPOP, but both of them don't generate checkpoint files.

Sep 16 '22 12:09 South-Twilight

Please check if "save_ckpt" is set to true in the hparams of your experiments. Otherwise, the saving period is quite large.

Oct 09 '22 08:10 MrZixi

Yes, "save_ckpt" is set to true. I'll try on another environment again.

Oct 17 '22 03:10 nakasako

@nakasako Hey dude, has your issue been solved yet?

Dec 23 '22 01:12 ZJ-CAI

I don't know why but I got a checkpoint. Thank you. I'll close this issue.

Dec 23 '22 05:12 nakasako