41 lines
584 B
Python
41 lines
584 B
Python
from dataclasses import dataclass
|
|
|
|
import torch
|
|
|
|
# Category
|
|
|
|
BASE_NODE_CATEGORY = "speech-dataset-toolkit"
|
|
|
|
|
|
# Data
|
|
|
|
|
|
@dataclass(frozen=True)
|
|
class AudioData:
|
|
"""
|
|
audio data mainly handled in this extension
|
|
RETURN_TYPES = ("AUDIO", )
|
|
"""
|
|
|
|
waveform: torch.Tensor
|
|
|
|
sample_rate: int
|
|
|
|
def is_stereo(self):
|
|
return self.waveform.size(0) > 1
|
|
|
|
|
|
@dataclass(frozen=True)
|
|
class SpectrogramData:
|
|
"""
|
|
spectrogram data mainly handled for visualization
|
|
RETURN_TYPES = ("AUDIO", )
|
|
"""
|
|
|
|
waveform: torch.Tensor
|
|
|
|
sample_rate: int
|
|
|
|
|
|
# Node Pin
|