lite

app.py

@@ -3,6 +3,8 @@ import os
 import uuid
 import subprocess
 from werkzeug.utils import secure_filename
+import time
+import random
 
 app = Flask(__name__)
 UPLOAD_FOLDER = '/data'
@@ -11,6 +13,11 @@ ALLOWED_EXTENSIONS = {'wav', 'mp3', 'ogg', 'flac'}
 def allowed_file(filename):
     return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS
 
+def generate_filename():
+    timestamp = int(time.time())
+    random_num = random.randint(1000, 9999)
+    return f"{timestamp}-{random_num}"
+
 @app.route('/stt', methods=['POST'])
 def transcribe():
     if 'audio' not in request.files:
@@ -24,12 +31,12 @@ if not file or not allowed_file(file.filename):
         return jsonify({"error": "Invalid file type"}), 400
 
     try:
-# Generate unique filename
-file_id = str(uuid.uuid4())
+        # Generate filename with timestamp + random number
+        file_base = generate_filename()
         orig_ext = secure_filename(file.filename).rsplit('.', 1)[1].lower()
-orig_path = os.path.join(UPLOAD_FOLDER, f"{file_id}.{orig_ext}")
-wav_path = os.path.join(UPLOAD_FOLDER, f"{file_id}.wav")
-output_base = os.path.join(UPLOAD_FOLDER, file_id)
+        orig_path = os.path.join(UPLOAD_FOLDER, f"{file_base}.{orig_ext}")
+        wav_path = os.path.join(UPLOAD_FOLDER, f"{file_base}.wav")
+        output_base = os.path.join(UPLOAD_FOLDER, file_base)
 
         # Save original file
         file.save(orig_path)
@@ -48,7 +55,7 @@ audio_path = orig_path
         # Run whisper.cpp
         subprocess.run([
             '/app/whisper.cpp/build/bin/main',
-'-m', '/app/whisper.cpp/models/ggml-base.bin',
+            '-m', '/app/whisper.cpp/models/ggml-tiny.en-q5_1.bin',
             '-f', audio_path,
             '-otxt', '-of', output_base
         ], check=True)
@@ -58,12 +65,14 @@ output_file = f"{output_base}.txt"
         with open(output_file, 'r') as f:
             transcription = f.read()
 
-# Cleanup
-for f in [orig_path, wav_path, audio_path, output_file]:
-if os.path.exists(f):
-os.remove(f)
+        # Return the filenames in the response
+        response = {
+            "text": transcription.strip(),
+            "audio_file": f"{file_base}.{orig_ext}",
+            "text_file": f"{file_base}.txt"
+        }
 
-return jsonify({"text": transcription.strip()})
+        return jsonify(response)
 
     except subprocess.CalledProcessError as e:
         return jsonify({"error": f"Whisper processing failed: {e.stderr}"}), 500

info

@@ -0,0 +1,6 @@
+docker build -t whisper-api .
+
+docker run -itd --name stt-41-lite -p 4004:4004 \
+-v $(pwd)/whisper.cpp-1.5.2/models:/models \
+-v $(pwd)/data:/data \
+whisper-api