File size: 970 Bytes
06c46fb 233adb5 06c46fb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 |
import soundfile as sf
import torch
import torchaudio
import numpy as np
from src.model.feature_extractor import processor # type: ignore
from src.config import DEVICE
# Resampler
resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
def preprocess_audio(batch):
speech, sample_rate = sf.read(batch["path"], dtype="float32")
if sample_rate != 16000:
speech = torch.tensor(speech).unsqueeze(0)
speech = resampler(speech).squeeze(0).numpy()
batch["speech"] = speech.tolist()
batch["sampling_rate"] = 16000
return batch
def prepare_features(batch, max_length):
features = processor(
batch["speech"],
sampling_rate=16000,
padding=True,
truncation=True,
max_length=max_length,
return_tensors="pt"
)
batch["input_values"] = features.input_values.squeeze(0)
batch["label"] = torch.tensor(batch["label"], dtype=torch.long)
return batch
|