Delete whisper_server.py

whisper_server.py

@@ -1,62 +0,0 @@
-
-import os
-import tempfile
-from flask import request, jsonify
-from transformers import pipeline
-import torch
-
-# Define a writable directory for the model cache.
-# This now respects the HF_HOME environment variable set in the Dockerfile.
-cache_dir = os.environ.get("HF_HOME", "/tmp/huggingface")
-os.makedirs(cache_dir, exist_ok=True)
-
-print("Loading collabora/whisper-tiny-hindi model via transformers pipeline...")
-
-# Determine device
-device = "cuda:0" if torch.cuda.is_available() else "cpu"
-
-# Initialize the ASR pipeline with the specified model
-# Using the transformers pipeline is the correct way to load custom models from the Hub.
-model = pipeline(
-    "automatic-speech-recognition",
-    model="collabora/whisper-tiny-hindi",
-    device=device,
-    model_kwargs={"cache_dir": cache_dir}
-)
-
-print("Whisper model loaded.")
-
-def handle_transcribe():
-    if 'file' not in request.files:
-        return jsonify({'error': 'No file part in the request'}), 400
-    
-    file = request.files['file']
-    
-    if file.filename == '':
-        return jsonify({'error': 'No selected file'}), 400
-
-    if file:
-        # Use a temporary file to save the upload
-        with tempfile.NamedTemporaryFile(delete=True, suffix=".webm") as temp_audio:
-            file.save(temp_audio.name)
-            
-            try:
-                print(f"Transcribing file: {temp_audio.name} with collabora/whisper-tiny-hindi pipeline")
-                
-                # The pipeline expects a file path and handles the processing.
-                result = model(temp_audio.name)
-                
-                transcribed_text = result.get('text', '')
-
-                print("Transcription successful.")
-                return jsonify({'text': transcribed_text})
-            except Exception as e:
-                print(f"Error during transcription: {e}")
-                # Provide a more specific error if possible
-                error_message = f"An unexpected error occurred during transcription: {str(e)}"
-                if "out of memory" in str(e).lower():
-                    error_message = "The model ran out of memory. Please try a smaller audio file or check server resources."
-                
-                return jsonify({'error': error_message}), 500
-
-    return jsonify({'error': 'File processing failed'}), 500