File size: 1,922 Bytes
53a770a |
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 |
import os
import base64
import tempfile
from pathlib import Path
from typing import Dict, Any
from groq import Groq
class EndpointHandler:
def __init__(self, path=""):
# Get Groq API key from HF endpoint environment variables
self.groq_api_key = os.getenv("GROQ_API_KEY")
if not self.groq_api_key:
raise ValueError("GROQ_API_KEY environment variable is required")
# Initialize Groq client
self.client = Groq(api_key=self.groq_api_key)
def __call__(self, data: Dict[str, Any]) -> Dict[str, Any]:
try:
# Get audio data from request
if "inputs" in data:
audio_data = data["inputs"]
else:
audio_data = data
# Handle base64 encoded audio
if isinstance(audio_data, str):
audio_bytes = base64.b64decode(audio_data)
else:
audio_bytes = audio_data
# Save to temp file
with tempfile.NamedTemporaryFile(delete=False, suffix='.wav') as temp_file:
temp_file.write(audio_bytes)
temp_file.flush()
# Forward to Groq API
with open(temp_file.name, "rb") as file:
transcription = self.client.audio.transcriptions.create(
file=(Path(temp_file.name).name, file.read()),
model="whisper-large-v3-turbo",
language="ta",
response_format="text",
temperature=0.0
)
# Clean up
os.unlink(temp_file.name)
return {"text": transcription}
except Exception as e:
return {"error": str(e)} |