recognize_whisper()

recognize_whisper(
    audio_data: AudioData,
    model: str = "base",
    show_dict: bool = False,
    load_options: dict | None = None,
    language: str | None = None,
    task: Literal["transcribe", "translate"] = "transcribe",
    **transcribe_options
) -> str | dict

import speech_recognition as sr

# Initialize recognizer
r = sr.Recognizer()

# Record audio
with sr.Microphone() as source:
    print("Say something!")
    audio = r.listen(source)

# Recognize with Whisper (offline)
try:
    text = r.recognize_whisper(audio)
    print(f"You said: {text}")
except sr.RequestError as e:
    print(f"Could not process audio; {e}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    audio = r.listen(source)

# Use tiny model for speed
text_tiny = r.recognize_whisper(audio, model="tiny")
print(f"Tiny model: {text_tiny}")

# Use large model for accuracy
text_large = r.recognize_whisper(audio, model="large")
print(f"Large model: {text_large}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    print("Parlez maintenant...")
    audio = r.listen(source)

# Specify French language
text = r.recognize_whisper(audio, language="french")
print(f"Vous avez dit: {text}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    audio = r.listen(source)

# Get full response to see detected language
result = r.recognize_whisper(audio, show_dict=True)
print(f"Detected language: {result['language']}")
print(f"Transcript: {result['text']}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    print("Speak in any language...")
    audio = r.listen(source)

# Transcribe and translate to English
text = r.recognize_whisper(audio, task="translate")
print(f"English translation: {text}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    audio = r.listen(source)

# Get detailed segments
result = r.recognize_whisper(audio, show_dict=True)

print(f"Full text: {result['text']}")
print("\nSegments:")
for segment in result['segments']:
    start = segment['start']
    end = segment['end']
    text = segment['text']
    print(f"  [{start:.2f}s - {end:.2f}s]: {text}")

import speech_recognition as sr
import torch

r = sr.Recognizer()

with sr.Microphone() as source:
    audio = r.listen(source)

# Explicitly use GPU
load_options = {
    "device": "cuda" if torch.cuda.is_available() else "cpu"
}

text = r.recognize_whisper(
    audio,
    model="large",
    load_options=load_options
)
print(f"Transcript: {text}")

import speech_recognition as sr

r = sr.Recognizer()

# Load audio file
with sr.AudioFile("speech.wav") as source:
    audio = r.record(source)

# Transcribe with Whisper
text = r.recognize_whisper(audio, model="medium")
print(text)

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    audio = r.listen(source)

# Use single temperature for deterministic results
text = r.recognize_whisper(audio, temperature=0.0)
print(text)

# Or use multiple temperatures for fallback
text = r.recognize_whisper(
    audio,
    temperature=(0.0, 0.2, 0.4, 0.6, 0.8, 1.0)
)
print(text)

pip install openai-whisper

# Install PyTorch with CUDA support first
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

# Then install Whisper
pip install openai-whisper