Spaces:
Running
Running
Create transcribe.py
Browse files- transcribe.py +31 -0
transcribe.py
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
from fastapi import APIRouter, File, UploadFile, HTTPException
|
3 |
+
from groq import Groq
|
4 |
+
from dotenv import load_dotenv
|
5 |
+
|
6 |
+
# Load environment variables from .env file
|
7 |
+
load_dotenv()
|
8 |
+
GROQ_API_KEY = os.getenv("GROQ_API_KEY")
|
9 |
+
if not GROQ_API_KEY:
|
10 |
+
raise Exception("GROQ_API_KEY is not set in the environment.")
|
11 |
+
|
12 |
+
router = APIRouter()
|
13 |
+
|
14 |
+
@router.post("/transcribe")
|
15 |
+
async def transcribe_audio(file: UploadFile = File(...)):
|
16 |
+
"""
|
17 |
+
Transcribe an uploaded audio file using the Groq client with the Whisper model.
|
18 |
+
Returns the transcribed text.
|
19 |
+
"""
|
20 |
+
try:
|
21 |
+
# Initialize the Groq client with the API key from the environment.
|
22 |
+
client = Groq(api_key=GROQ_API_KEY)
|
23 |
+
file_bytes = await file.read()
|
24 |
+
transcription = client.audio.transcriptions.create(
|
25 |
+
file=(file.filename, file_bytes),
|
26 |
+
model="whisper-large-v3",
|
27 |
+
response_format="verbose_json",
|
28 |
+
)
|
29 |
+
return {"transcription": transcription.text}
|
30 |
+
except Exception as e:
|
31 |
+
raise HTTPException(status_code=500, detail=str(e))
|