admin-valentin/WhisperLiveKit
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

specify dtype for librosa.load, instead of cast

Browse source
This commit is contained in:
koiking213 2024-02-20 23:29:25 +09:00
parent 24926c98e0
commit 4405c451ce
2 changed files with 5 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
whisper_online.py
View file

@ -10,8 +10,8 @@ import math
@lru_cache @lru_cache
def load_audio(fname): def load_audio(fname):
a, _ = librosa.load(fname, sr=16000) a, _ = librosa.load(fname, sr=16000, dtype=np.float32)
return a.astype('float32') return a
def load_audio_chunk(fname, beg, end): def load_audio_chunk(fname, beg, end):
audio = load_audio(fname) audio = load_audio(fname)

5
whisper_online_server.py
View file

@ -4,6 +4,7 @@ from whisper_online import *
import sys import sys
import argparse import argparse
import os import os
import numpy as np
parser = argparse.ArgumentParser() parser = argparse.ArgumentParser()
# server options # server options
@ -137,8 +138,8 @@ class ServerProcessor:
if not raw_bytes: if not raw_bytes:
break break
sf = soundfile.SoundFile(io.BytesIO(raw_bytes), channels=1,endian="LITTLE",samplerate=SAMPLING_RATE, subtype="PCM_16",format="RAW") sf = soundfile.SoundFile(io.BytesIO(raw_bytes), channels=1,endian="LITTLE",samplerate=SAMPLING_RATE, subtype="PCM_16",format="RAW")
audio, _ = librosa.load(sf,sr=SAMPLING_RATE) audio, _ = librosa.load(sf,sr=SAMPLING_RATE,dtype=np.float32)
out.append(audio.astype('float32')) out.append(audio)
if not out: if not out:
return None return None
return np.concatenate(out) return np.concatenate(out)