Wulf's Various Things

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.

====== Speech to Text ======

OpenAI Whisper
https://github.com/openai/whisper

https://hub.docker.com/r/fedirz/faster-whisper-server

https://github.com/speaches-ai/speaches


OpenAI/Whisper example python script with changed base url

<code python>
mkdir transcribe
cd transcribe
python -m venv venv
source venv/bin/activate
pip install openai

cat >> trans.py << EOD
#!/home/user/transcribe/venv/bin/python3
from openai import OpenAI
import sys,os

def usage():
    print("Usage: "+os.path.basename(__file__)+" inputfile [outputfile]")
    print(" inputfile \t\tInput file to transcribe")
    print(" outputfile\tOptional: Filename to store transcription to")
    exit(1)
    
try:
    inputfile = sys.argv[1]
except Exception as e:
    usage()

client = OpenAI(base_url="http://192.168.xx.xx:8010/v1", api_key="not needed")
audio_file= open(inputfile, "rb")

transcription = client.audio.transcriptions.create(
    model="Systran/faster-whisper-large-v3", 
    file=audio_file
)

try:
    outputfile = sys.argv[2]
    with open(sys.argv[2], "w") as text_file:
        text_file.write(transcription.text)
except:
    print(transcription.text)
EOD
</code>