Update services/whisper_service.py

services/whisper_service.py

@@ -1,46 +1,13 @@
 import whisper
-import tempfile
-import os
 import torch
-from config import settings
 
+# Checking if NVIDIA GPU is available
+DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
 
-class WhisperService:
-    def __init__(self):
-        if settings.TORCH_DEVICE == "cuda" and not torch.cuda.is_available():
-            print("WARNING: CUDA requested but not available. Falling back to CPU.")
-            self.device = "cpu"
-        else:
-            self.device = settings.TORCH_DEVICE
+# Load the Whisper model
+model = whisper.load_model("base", device=DEVICE)
 
-        self.model = whisper.load_model(settings.WHISPER_MODEL)
-        if settings.FORCE_FP32 or self.device == "cpu":
-            self.model = self.model.float()
-
-    async def transcribe(self, audio_file: bytes, output_language: str = None) -> dict:
-        try:
-            # Create a temporary file to store the uploaded audio
-            with tempfile.NamedTemporaryFile(delete=False, suffix='.mp3') as temp_audio:
-                temp_audio.write(audio_file)
-                temp_audio_path = temp_audio.name
-
-            try:
-                # Transcribe the audio
-                transcription_options = {"fp16": not settings.FORCE_FP32 and self.device == "cuda"}
-                if output_language:
-                    transcription_options["language"] = output_language
-
-                result = self.model.transcribe(temp_audio_path, **transcription_options)
-
-                return {
-                    "text": result["text"],
-                    "language": result.get("language"),
-                    "segments": result.get("segments")
-                }
-            finally:
-                # Clean up the temporary file
-                if os.path.exists(temp_audio_path):
-                    os.remove(temp_audio_path)
-
-        except Exception as e:
-            raise Exception(f"Transcription failed: {str(e)}")
+def transcribe_audio(file_path: str) -> str:
+    """Transcribes the given audio file and returns the text."""
+    result = model.transcribe(file_path)
+    return result['text']