.1998285672285072:0f1da57489c9b5e26090b4e19672e818_69e5fb9e6da72146aef923a9.69e5ff096da72146aef9241c.69e5ff0867df5f5a4bc90285:Trae CN.T(2026/4/20 18:25:13)

requirements_webui.txt

@@ -0,0 +1,7 @@
+flask
+espeakng_loader
+phonemizer
+onnxruntime
+soundfile
+numpy
+huggingface_hub

run_webui.py

@@ -0,0 +1,18 @@
+import os
+import sys
+
+script_dir = os.path.dirname(os.path.abspath(__file__))
+sys.path.insert(0, script_dir)
+
+from webui.app import app
+
+if __name__ == '__main__':
+    print("=" * 60)
+    print("[KittenTTS] WebUI 启动中...")
+    print("=" * 60)
+    print()
+    print("请在浏览器中访问: http://localhost:5000")
+    print()
+    print("=" * 60)
+
+    app.run(host='0.0.0.0', port=5000, debug=False)

webui/__init__.py

@@ -0,0 +1,3 @@
+from .app import app
+
+__all__ = ['app']

webui/app.py

@@ -0,0 +1,308 @@
+import os
+import uuid
+import json
+import threading
+from datetime import datetime
+from flask import Flask, render_template, request, jsonify, send_file, Response
+from werkzeug.utils import secure_filename
+from kittentts import KittenTTS
+
+app = Flask(__name__)
+
+BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+AUDIO_DIR = os.path.join(BASE_DIR, 'static', 'audio')
+HISTORY_FILE = os.path.join(BASE_DIR, 'history.json')
+
+os.makedirs(AUDIO_DIR, exist_ok=True)
+
+AVAILABLE_MODELS = [
+    {"id": "KittenML/kitten-tts-mini-0.8", "name": "Mini (80M)", "description": "最高质量版本"},
+    {"id": "KittenML/kitten-tts-micro-0.8", "name": "Micro (40M)", "description": "速度与质量平衡"},
+    {"id": "KittenML/kitten-tts-nano-0.8", "name": "Nano (15M)", "description": "最小最快版本"},
+]
+
+MODELS = {}
+MODELS_LOCK = threading.Lock()
+TASK_QUEUE = []
+TASK_QUEUE_LOCK = threading.Lock()
+HISTORY = []
+HISTORY_LOCK = threading.Lock()
+
+def load_model(model_id):
+    with MODELS_LOCK:
+        if model_id in MODELS:
+            return MODELS[model_id]
+        print(f"Loading model: {model_id}")
+        model = KittenTTS(model_id)
+        MODELS[model_id] = model
+        return model
+
+def load_history():
+    global HISTORY
+    if os.path.exists(HISTORY_FILE):
+        try:
+            with open(HISTORY_FILE, 'r', encoding='utf-8') as f:
+                HISTORY = json.load(f)
+        except:
+            HISTORY = []
+    else:
+        HISTORY = []
+
+def save_history():
+    with open(HISTORY_FILE, 'w', encoding='utf-8') as f:
+        json.dump(HISTORY, f, ensure_ascii=False, indent=2)
+
+def add_to_history(task):
+    global HISTORY
+    with HISTORY_LOCK:
+        HISTORY.insert(0, task)
+        if len(HISTORY) > 100:
+            HISTORY = HISTORY[:100]
+        save_history()
+
+def process_task(task):
+    try:
+        task['status'] = 'processing'
+        model = load_model(task['model_id'])
+        audio = model.generate(
+            text=task['text'],
+            voice=task['voice'],
+            speed=task['speed'],
+            clean_text=task.get('clean_text', True)
+        )
+
+        audio_filename = f"{task['id']}.wav"
+        audio_path = os.path.join(AUDIO_DIR, audio_filename)
+        import soundfile as sf
+        sf.write(audio_path, audio, 24000)
+
+        task['status'] = 'completed'
+        task['audio_file'] = audio_filename
+        task['completed_at'] = datetime.now().isoformat()
+
+        with TASK_QUEUE_LOCK:
+            for i, t in enumerate(TASK_QUEUE):
+                if t['id'] == task['id']:
+                    TASK_QUEUE[i] = task
+                    break
+
+        add_to_history(task)
+    except Exception as e:
+        task['status'] = 'failed'
+        task['error'] = str(e)
+        with TASK_QUEUE_LOCK:
+            for i, t in enumerate(TASK_QUEUE):
+                if t['id'] == task['id']:
+                    TASK_QUEUE[i] = task
+                    break
+
+def task_worker():
+    while True:
+        task_to_process = None
+        with TASK_QUEUE_LOCK:
+            for task in TASK_QUEUE:
+                if task['status'] == 'pending':
+                    task_to_process = task
+                    break
+
+        if task_to_process:
+            process_task(task_to_process)
+        else:
+            import time
+            time.sleep(1)
+
+worker_thread = threading.Thread(target=task_worker, daemon=True)
+worker_thread.start()
+
+load_history()
+
+@app.route('/')
+def index():
+    return render_template('index.html')
+
+@app.route('/api/models', methods=['GET'])
+def get_models():
+    return jsonify(AVAILABLE_MODELS)
+
+@app.route('/api/voices', methods=['GET'])
+def get_voices():
+    model_id = request.args.get('model', 'KittenML/kitten-tts-mini-0.8')
+    try:
+        model = load_model(model_id)
+        voices = [
+            {'id': 'Bella', 'name': 'Bella', 'description': '女声'},
+            {'id': 'Jasper', 'name': 'Jasper', 'description': '男声'},
+            {'id': 'Luna', 'name': 'Luna', 'description': '女声'},
+            {'id': 'Bruno', 'name': 'Bruno', 'description': '男声'},
+            {'id': 'Rosie', 'name': 'Rosie', 'description': '女声'},
+            {'id': 'Hugo', 'name': 'Hugo', 'description': '男声'},
+            {'id': 'Kiki', 'name': 'Kiki', 'description': '女声'},
+            {'id': 'Leo', 'name': 'Leo', 'description': '男声'},
+        ]
+        return jsonify(voices)
+    except Exception as e:
+        return jsonify({'error': str(e)}), 500
+
+@app.route('/api/generate', methods=['POST'])
+def generate():
+    try:
+        data = request.get_json()
+        text = data.get('text', '')
+        model_id = data.get('model', 'KittenML/kitten-tts-mini-0.8')
+        voice = data.get('voice', 'Bruno')
+        speed = float(data.get('speed', 1.0))
+        clean_text = data.get('clean_text', True)
+
+        if not text:
+            return jsonify({'error': '文本不能为空'}), 400
+
+        model = load_model(model_id)
+        audio = model.generate(
+            text=text,
+            voice=voice,
+            speed=speed,
+            clean_text=clean_text
+        )
+
+        task_id = str(uuid.uuid4())
+        audio_filename = f"{task_id}.wav"
+        audio_path = os.path.join(AUDIO_DIR, audio_filename)
+        import soundfile as sf
+        sf.write(audio_path, audio, 24000)
+
+        task = {
+            'id': task_id,
+            'text': text[:100] + ('...' if len(text) > 100 else ''),
+            'model_id': model_id,
+            'voice': voice,
+            'speed': speed,
+            'status': 'completed',
+            'audio_file': audio_filename,
+            'created_at': datetime.now().isoformat(),
+            'completed_at': datetime.now().isoformat(),
+        }
+
+        add_to_history(task)
+
+        return jsonify({
+            'success': True,
+            'task': task,
+            'audio_url': f'/api/audio/{audio_filename}'
+        })
+    except Exception as e:
+        return jsonify({'error': str(e)}), 500
+
+@app.route('/api/tasks', methods=['POST'])
+def add_task():
+    try:
+        data = request.get_json()
+        text = data.get('text', '')
+        model_id = data.get('model', 'KittenML/kitten-tts-mini-0.8')
+        voice = data.get('voice', 'Bruno')
+        speed = float(data.get('speed', 1.0))
+        clean_text = data.get('clean_text', True)
+
+        if not text:
+            return jsonify({'error': '文本不能为空'}), 400
+
+        task_id = str(uuid.uuid4())
+        task = {
+            'id': task_id,
+            'text': text,
+            'model_id': model_id,
+            'voice': voice,
+            'speed': speed,
+            'clean_text': clean_text,
+            'status': 'pending',
+            'created_at': datetime.now().isoformat(),
+        }
+
+        with TASK_QUEUE_LOCK:
+            TASK_QUEUE.append(task)
+
+        return jsonify({
+            'success': True,
+            'task': task
+        })
+    except Exception as e:
+        return jsonify({'error': str(e)}), 500
+
+@app.route('/api/tasks', methods=['GET'])
+def get_tasks():
+    with TASK_QUEUE_LOCK:
+        return jsonify(TASK_QUEUE)
+
+@app.route('/api/tasks/<task_id>', methods=['GET'])
+def get_task(task_id):
+    with TASK_QUEUE_LOCK:
+        for task in TASK_QUEUE:
+            if task['id'] == task_id:
+                return jsonify(task)
+    return jsonify({'error': '任务不存在'}), 404
+
+@app.route('/api/history', methods=['GET'])
+def get_history():
+    with HISTORY_LOCK:
+        return jsonify(HISTORY)
+
+@app.route('/api/history/<task_id>', methods=['DELETE'])
+def delete_history_item(task_id):
+    global HISTORY
+    with HISTORY_LOCK:
+        HISTORY = [h for h in HISTORY if h['id'] != task_id]
+        save_history()
+        return jsonify({'success': True})
+
+@app.route('/api/history', methods=['DELETE'])
+def clear_history():
+    global HISTORY
+    with HISTORY_LOCK:
+        for item in HISTORY:
+            audio_file = item.get('audio_file')
+            if audio_file:
+                audio_path = os.path.join(AUDIO_DIR, audio_file)
+                if os.path.exists(audio_path):
+                    os.remove(audio_path)
+        HISTORY = []
+        save_history()
+        return jsonify({'success': True})
+
+@app.route('/api/audio/<filename>', methods=['GET'])
+def get_audio(filename):
+    audio_path = os.path.join(AUDIO_DIR, secure_filename(filename))
+    if os.path.exists(audio_path):
+        return send_file(audio_path, mimetype='audio/wav')
+    return jsonify({'error': '音频文件不存在'}), 404
+
+@app.route('/api/status', methods=['GET'])
+def get_status():
+    model_info = {}
+    with MODELS_LOCK:
+        for model_id, model in MODELS.items():
+            model_info[model_id] = 'loaded'
+
+    pending_count = 0
+    processing_count = 0
+    completed_count = 0
+    with TASK_QUEUE_LOCK:
+        for task in TASK_QUEUE:
+            if task['status'] == 'pending':
+                pending_count += 1
+            elif task['status'] == 'processing':
+                processing_count += 1
+            elif task['status'] == 'completed':
+                completed_count += 1
+
+    return jsonify({
+        'loaded_models': model_info,
+        'queue_stats': {
+            'pending': pending_count,
+            'processing': processing_count,
+            'completed': completed_count,
+            'total': len(TASK_QUEUE)
+        },
+        'history_count': len(HISTORY)
+    })
+
+if __name__ == '__main__':
+    app.run(host='0.0.0.0', port=5000, debug=True)

webui/history.json

@@ -0,0 +1,47 @@
+[
+  {
+    "id": "ac570fca-72f7-4aa9-9f24-f5801b45a4df",
+    "text": "This is a batch test message. It will be added to the task queue and processed in the background.",
+    "model_id": "KittenML/kitten-tts-mini-0.8",
+    "voice": "Bella",
+    "speed": 2.0,
+    "status": "completed",
+    "audio_file": "ac570fca-72f7-4aa9-9f24-f5801b45a4df.wav",
+    "created_at": "2026-04-20T18:30:33.216799",
+    "completed_at": "2026-04-20T18:30:33.216799"
+  },
+  {
+    "id": "45e127e1-25be-4411-89ca-8ab228cb2810",
+    "text": "This is a batch test message. It will be added to the task queue and processed in the background.",
+    "model_id": "KittenML/kitten-tts-mini-0.8",
+    "voice": "Bella",
+    "speed": 1.4,
+    "status": "completed",
+    "audio_file": "45e127e1-25be-4411-89ca-8ab228cb2810.wav",
+    "created_at": "2026-04-20T18:30:20.507877",
+    "completed_at": "2026-04-20T18:30:20.507877"
+  },
+  {
+    "id": "cbc3ff0f-c2b2-4a27-98c4-28ac87187485",
+    "text": "This is a batch test message. It will be added to the task queue and processed in the background.",
+    "model_id": "KittenML/kitten-tts-mini-0.8",
+    "voice": "Bruno",
+    "speed": 1.0,
+    "clean_text": true,
+    "status": "completed",
+    "created_at": "2026-04-20T18:27:37.313439",
+    "audio_file": "cbc3ff0f-c2b2-4a27-98c4-28ac87187485.wav",
+    "completed_at": "2026-04-20T18:27:44.202011"
+  },
+  {
+    "id": "641a77d3-231f-410d-bca6-8f6705e43ee1",
+    "text": "Hello, this is a test of KittenTTS WebUI. The quick brown fox jumps over the lazy dog.",
+    "model_id": "KittenML/kitten-tts-mini-0.8",
+    "voice": "Bruno",
+    "speed": 1.0,
+    "status": "completed",
+    "audio_file": "641a77d3-231f-410d-bca6-8f6705e43ee1.wav",
+    "created_at": "2026-04-20T18:27:19.889696",
+    "completed_at": "2026-04-20T18:27:19.889696"
+  }
+]