Closed Yurzi closed 1 year ago
rgument self in method wrapper_CUDA_Tensor_searchsorted) 1%|▋ | 45/7110 [00:37<32:55, 3.58it/s]2023-04-05 20:21:32.660 | ERROR | __main__:safe_process:198 - Error processing dataset/train/opencpop/2082003011.wav 2023-04-05 20:21:32.660 | ERROR | __main__:safe_process:199 - Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument self in method wrapper_CUDA_Tensor_searchsorted) Traceback (most recent call last): File "/mnt/fish-diffusion/tools/preprocessing/extract_features.py", line 245, in <module> i = safe_process(args, config, audio_path) │ │ │ └ PosixPath('dataset/train/opencpop/2082003011.wav') │ │ └ Config (path: configs/svc_hifisinger_finetune.py): {'sampling_rate': 44100, 'hidden_size': 256, 'vocoder_config': {'sampling_... │ └ Namespace(config='configs/svc_hifisinger_finetune.py', path='dataset/train', clean=True, num_workers=1, no_augmentation=False) └ <function safe_process at 0x7fa72e5d3640> > File "/mnt/fish-diffusion/tools/preprocessing/extract_features.py", line 166, in safe_process process(config, audio_path) │ │ └ PosixPath('dataset/train/opencpop/2082003011.wav') │ └ Config (path: configs/svc_hifisinger_finetune.py): {'sampling_rate': 44100, 'hidden_size': 256, 'vocoder_config': {'sampling_... └ <function process at 0x7fa72e5d3130> File "/mnt/fish-diffusion/tools/preprocessing/extract_features.py", line 148, in process pitches = pitch_extractor(audio, sr, pad_to=mel_length) │ │ │ └ 455 │ │ └ 44100 │ └ tensor([[ 0.0017, 0.0017, 0.0018, ..., -0.0013, -0.0010, -0.0006]], │ device='cuda:0') └ <fish_diffusion.modules.pitch_extractors.parsel_mouth.ParselMouthPitchExtractor object at 0x7fa71f22bbb0> File "/mnt/fish-diffusion/fish_diffusion/modules/pitch_extractors/parsel_mouth.py", line 42, in __call__ return self.post_process(x, sampling_rate, f0, pad_to) │ │ │ │ │ └ 455 │ │ │ │ └ array([ 0. , 0. , 0. , 0. , │ │ │ │ 0. , 0. , 414.08354111, 416.3... │ │ │ └ 44100 │ │ └ tensor([[ 0.0017, 0.0017, 0.0018, ..., -0.0013, -0.0010, -0.0006]], │ │ device='cuda:0') │ └ <function BasePitchExtractor.post_process at 0x7fa72e5d2b90> └ <fish_diffusion.modules.pitch_extractors.parsel_mouth.ParselMouthPitchExtractor object at 0x7fa71f22bbb0> File "/mnt/fish-diffusion/fish_diffusion/modules/pitch_extractors/builder.py", line 59, in post_process return interpolate(time_frame, time_org, f0, left=f0[0], right=f0[-1]) │ │ │ │ │ └ tensor([ 414.0835, 416.3744, 420.1241, 424.6104, 428.1972, 429.9621, │ │ │ │ │ 429.6776, 427.2525, 420.7615, 410.0335,... │ │ │ │ └ tensor([ 414.0835, 416.3744, 420.1241, 424.6104, 428.1972, 429.9621, │ │ │ │ 429.6776, 427.2525, 420.7615, 410.0335,... │ │ │ └ tensor([ 414.0835, 416.3744, 420.1241, 424.6104, 428.1972, 429.9621, │ │ │ 429.6776, 427.2525, 420.7615, 410.0335,... │ │ └ tensor([0.0697, 0.0813, 0.0929, 0.1045, 0.1161, 0.1277, 0.1393, 0.1509, 0.1625, │ │ 0.1741, 0.1858, 0.1974, 0.2090, 0.220... │ └ tensor([0.0000, 0.0116, 0.0232, 0.0348, 0.0464, 0.0580, 0.0697, 0.0813, 0.0929, │ 0.1045, 0.1161, 0.1277, 0.1393, 0.150... └ <function interpolate at 0x7fa731706710> File "/mnt/fish-diffusion/fish_diffusion/utils/tensor.py", line 67, in interpolate i = torch.clip(torch.searchsorted(xp, x, right=True), 1, len(xp) - 1) │ │ │ │ │ │ └ tensor([0.0697, 0.0813, 0.0929, 0.1045, 0.1161, 0.1277, 0.1393, 0.1509, 0.1625, │ │ │ │ │ │ 0.1741, 0.1858, 0.1974, 0.2090, 0.220... │ │ │ │ │ └ tensor([0.0000, 0.0116, 0.0232, 0.0348, 0.0464, 0.0580, 0.0697, 0.0813, 0.0929, │ │ │ │ │ 0.1045, 0.1161, 0.1277, 0.1393, 0.150... │ │ │ │ └ tensor([0.0697, 0.0813, 0.0929, 0.1045, 0.1161, 0.1277, 0.1393, 0.1509, 0.1625, │ │ │ │ 0.1741, 0.1858, 0.1974, 0.2090, 0.220... │ │ │ └ <built-in method searchsorted of type object at 0x7fa79fea4880> │ │ └ <module 'torch' from '/opt/miniconda/envs/fish/lib/python3.10/site-packages/torch/__init__.py'> │ └ <built-in method clip of type object at 0x7fa79fea4880> └ <module 'torch' from '/opt/miniconda/envs/fish/lib/python3.10/site-packages/torch/__init__.py'> RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument self in method wrapper_CUDA_Tensor_searchsorted)
在 CUDA 12.1的机子上抛出了这个错误
Wed Apr 5 20:25:55 2023 +---------------------------------------------------------------------------------------+ | NVIDIA-SMI 530.30.02 Driver Version: 530.30.02 CUDA Version: 12.1 | |-----------------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+======================+======================| | 0 Tesla V100-SXM2-32GB Off| 00000000:00:08.0 Off | 0 | | N/A 34C P0 53W / 300W| 0MiB / 32768MiB | 1% Default | | | | N/A | +-----------------------------------------+----------------------+----------------------+ +---------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=======================================================================================| | No running processes found | +---------------------------------------------------------------------------------------+
不知道是什么问题捏
在 CUDA 12.1的机子上抛出了这个错误
不知道是什么问题捏