search

PlayVoice / lora-svc

singing voice change based on whisper, and lora for singing voice clone
MIT License
630 stars 78 forks source link

ValueError: not enough values to unpack (expected 4, got 1) #63

Closed IsEddy closed 4 months ago

IsEddy commented 1 year ago

按照readme跑的，上周还能跑起来，这周就不行了（

MaxMax2016 commented 1 year ago

有更详细的错误信息没有？

IsEddy commented 1 year ago

F:\Desktop\lora-svc>python svc_trainer.py -c config/maxgan.yaml -n lora Batch size per GPU : 4 ----------5---------- 2023-05-21 11:08:57,209 - INFO - Start from pretrain model: model_pretrain/maxgan_pretrain_48K_5L.pth 2023-05-21 11:08:58,589 - INFO - Starting new training run. Traceback (most recent call last): File "F:\Desktop\lora-svc\svc_trainer.py", line 44, in train(0, args, args.checkpoint_path, hp, hp_str) File "F:\Desktop\lora-svc\utils\train.py", line 107, in train trainloader = create_dataloader(hp, True) File "F:\Desktop\lora-svc\utils\dataloader.py", line 33, in create_dataloader dataset = FeatureFromDisk(hp, train) File "F:\Desktop\lora-svc\utils\dataloader.py", line 53, in init self._filter() File "F:\Desktop\lora-svc\utils\dataloader.py", line 58, in _filter for wavpath, pitch, ppg, spk in self.items: ValueError: not enough values to unpack (expected 4, got 1)

MaxMax2016 commented 1 year ago

训练文件train.txt里面多了空白行？

IsEddy commented 1 year ago

是的是的，怪不得😂