speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge pull request #101 from alibaba-damo-academy/dev_lhn

Browse Source 
Dev lhn
This commit is contained in:
hnluo 2023-02-13 17:28:06 +08:00 committed by GitHub
commit 7656297365
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 5 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
funasr/datasets/large_datasets/dataset.py
View File

@ -1,9 +1,10 @@
import os
import random
import soundfile
import numpy
from functools import partial
import torch
import torchaudio
import torch.distributed as dist
from kaldiio import ReadHelper
from torch.utils.data import IterableDataset
@ -117,7 +118,9 @@ class AudioDataset(IterableDataset):
sample_dict["key"] = key
elif data_type == "sound":
key, path = item.strip().split()
mat, sampling_rate = soundfile.read(path)
waveform, sampling_rate = torchaudio.load(path)
waveform = waveform.numpy()
mat = waveform[0]
sample_dict[data_name] = mat
sample_dict["sampling_rate"] = sampling_rate
if data_name == "speech":