Update ui/tabs.py
Browse files- ui/tabs.py +135 -2
ui/tabs.py
CHANGED
|
@@ -10,6 +10,7 @@ from services.streaming_voice_service import StreamingVoiceService
|
|
| 10 |
from services.openai_realtime_service import HybridStreamingService
|
| 11 |
from services.stream_object_detection_service import StreamObjectDetection
|
| 12 |
from services.voice_coding_service import VoiceCodingService
|
|
|
|
| 13 |
from core.rag_system import EnhancedRAGSystem
|
| 14 |
from core.tts_service import EnhancedTTSService
|
| 15 |
from core.wikipedia_processor import WikipediaProcessor
|
|
@@ -20,14 +21,17 @@ def create_all_tabs(audio_service: AudioService, chat_service: ChatService,
|
|
| 20 |
tts_service: EnhancedTTSService, wikipedia_processor: WikipediaProcessor,
|
| 21 |
streaming_voice_service: StreamingVoiceService,
|
| 22 |
hybrid_service: HybridStreamingService,
|
| 23 |
-
voice_coding_service: VoiceCodingService
|
|
|
|
| 24 |
):
|
| 25 |
|
| 26 |
with gr.Tab("🎙️ Streaming Voice "):
|
| 27 |
create_streaming_voice_tab(streaming_voice_service)
|
| 28 |
with gr.Tab("OpenAI Realtime"):
|
| 29 |
create_openai_realtime_tab(hybrid_service)
|
| 30 |
-
with gr.Tab("
|
|
|
|
|
|
|
| 31 |
create_voice_coding_tab(voice_coding_service)
|
| 32 |
with gr.Tab("🎙️ Audio"):
|
| 33 |
create_audio_tab(audio_service)
|
|
@@ -48,6 +52,135 @@ def create_all_tabs(audio_service: AudioService, chat_service: ChatService,
|
|
| 48 |
create_language_info_tab(rag_system.multilingual_manager)
|
| 49 |
with gr.Tab("Stream Object Detection"):
|
| 50 |
create_streaming_object_detection()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
def create_voice_coding_tab(voice_coding_service):
|
| 52 |
"""Tạo tab Voice Coding đơn giản - Text-based trước"""
|
| 53 |
|
|
|
|
| 10 |
from services.openai_realtime_service import HybridStreamingService
|
| 11 |
from services.stream_object_detection_service import StreamObjectDetection
|
| 12 |
from services.voice_coding_service import VoiceCodingService
|
| 13 |
+
from services.sambanova_voice_service import SambanovaVoiceService
|
| 14 |
from core.rag_system import EnhancedRAGSystem
|
| 15 |
from core.tts_service import EnhancedTTSService
|
| 16 |
from core.wikipedia_processor import WikipediaProcessor
|
|
|
|
| 21 |
tts_service: EnhancedTTSService, wikipedia_processor: WikipediaProcessor,
|
| 22 |
streaming_voice_service: StreamingVoiceService,
|
| 23 |
hybrid_service: HybridStreamingService,
|
| 24 |
+
voice_coding_service: VoiceCodingService,
|
| 25 |
+
sambanova_voice_service : SambanovaVoiceService
|
| 26 |
):
|
| 27 |
|
| 28 |
with gr.Tab("🎙️ Streaming Voice "):
|
| 29 |
create_streaming_voice_tab(streaming_voice_service)
|
| 30 |
with gr.Tab("OpenAI Realtime"):
|
| 31 |
create_openai_realtime_tab(hybrid_service)
|
| 32 |
+
with gr.Tab("SambonovaAI Realtime"):
|
| 33 |
+
create_sambanova_voice_tab(sambanova_voice_service)
|
| 34 |
+
with gr.Tab("Generation Code"):
|
| 35 |
create_voice_coding_tab(voice_coding_service)
|
| 36 |
with gr.Tab("🎙️ Audio"):
|
| 37 |
create_audio_tab(audio_service)
|
|
|
|
| 52 |
create_language_info_tab(rag_system.multilingual_manager)
|
| 53 |
with gr.Tab("Stream Object Detection"):
|
| 54 |
create_streaming_object_detection()
|
| 55 |
+
def create_sambanova_voice_tab(sambanova_service):
|
| 56 |
+
"""Tạo tab Sambanova Voice AI"""
|
| 57 |
+
|
| 58 |
+
# Tạo stream
|
| 59 |
+
stream = sambanova_service.create_stream()
|
| 60 |
+
|
| 61 |
+
with gr.Blocks() as sambanova_tab:
|
| 62 |
+
gr.Markdown("## 🤖 Sambanova Voice AI - Llama 3.2 3B")
|
| 63 |
+
gr.Markdown("Trò chuyện voice-to-voice với model Llama 3.2 3B thông qua Sambanova API")
|
| 64 |
+
|
| 65 |
+
# State variables
|
| 66 |
+
chatbot = gr.Chatbot(
|
| 67 |
+
type="messages",
|
| 68 |
+
value=[],
|
| 69 |
+
label="💬 Hội thoại Voice",
|
| 70 |
+
height=400
|
| 71 |
+
)
|
| 72 |
+
conversation_state = gr.State(value=[])
|
| 73 |
+
|
| 74 |
+
with gr.Row():
|
| 75 |
+
with gr.Column(scale=1):
|
| 76 |
+
gr.Markdown("### 🎤 Hướng dẫn sử dụng:")
|
| 77 |
+
gr.Markdown("""
|
| 78 |
+
**Cách sử dụng:**
|
| 79 |
+
1. Nhấn **Bắt đầu Voice Chat**
|
| 80 |
+
2. Nói "Computer", "Hey", "Hello" hoặc "Xin chào" để kích hoạt
|
| 81 |
+
3. Trò chuyện tự nhiên bằng giọng nói
|
| 82 |
+
4. Mỗi phiên giới hạn 90 giây
|
| 83 |
+
|
| 84 |
+
**Tính năng:**
|
| 85 |
+
✅ Voice-to-Voice conversation
|
| 86 |
+
✅ Real-time streaming
|
| 87 |
+
✅ Llama 3.2 3B model
|
| 88 |
+
✅ Low latency
|
| 89 |
+
✅ Multi-language support
|
| 90 |
+
""")
|
| 91 |
+
|
| 92 |
+
# WebRTC component
|
| 93 |
+
webrtc = stream.ui()
|
| 94 |
+
|
| 95 |
+
# Manual controls
|
| 96 |
+
with gr.Accordion("⚙️ Cài đặt nâng cao", open=False):
|
| 97 |
+
temperature = gr.Slider(
|
| 98 |
+
minimum=0.1,
|
| 99 |
+
maximum=1.0,
|
| 100 |
+
value=0.1,
|
| 101 |
+
label="Temperature"
|
| 102 |
+
)
|
| 103 |
+
top_p = gr.Slider(
|
| 104 |
+
minimum=0.1,
|
| 105 |
+
maximum=1.0,
|
| 106 |
+
value=0.1,
|
| 107 |
+
label="Top-P"
|
| 108 |
+
)
|
| 109 |
+
|
| 110 |
+
with gr.Row():
|
| 111 |
+
clear_btn = gr.Button("🗑️ Xóa hội thoại", variant="secondary")
|
| 112 |
+
export_btn = gr.Button("💾 Export Chat", variant="secondary")
|
| 113 |
+
|
| 114 |
+
with gr.Column(scale=2):
|
| 115 |
+
# Hiển thị thông tin model
|
| 116 |
+
with gr.Accordion("📊 Model Information", open=True):
|
| 117 |
+
gr.Markdown("""
|
| 118 |
+
**Llama 3.2 3B Instruct**
|
| 119 |
+
- **Provider**: Sambanova AI
|
| 120 |
+
- **Context Window**: 128K tokens
|
| 121 |
+
- **Languages**: Multi-language
|
| 122 |
+
- **Specialization**: Instruction following
|
| 123 |
+
- **API**: OpenAI-compatible
|
| 124 |
+
|
| 125 |
+
**Voice Features:**
|
| 126 |
+
- Real-time speech recognition
|
| 127 |
+
- Voice activity detection
|
| 128 |
+
- Stop-word activation
|
| 129 |
+
- Low-latency streaming
|
| 130 |
+
""")
|
| 131 |
+
|
| 132 |
+
# Hiển thị trạng thái
|
| 133 |
+
status_display = gr.Textbox(
|
| 134 |
+
label="🔄 Trạng thái hệ thống",
|
| 135 |
+
value="Sẵn sàng kết nối...",
|
| 136 |
+
interactive=False,
|
| 137 |
+
lines=2
|
| 138 |
+
)
|
| 139 |
+
|
| 140 |
+
# Additional outputs handling
|
| 141 |
+
def handle_additional_outputs(chatbot_output, state_output):
|
| 142 |
+
"""Xử lý additional outputs từ streaming"""
|
| 143 |
+
return chatbot_output, state_output, "✅ Đang trò chuyện..."
|
| 144 |
+
|
| 145 |
+
# Event handlers
|
| 146 |
+
def clear_conversation():
|
| 147 |
+
"""Xóa hội thoại"""
|
| 148 |
+
return [], [], "🔄 Đã xóa hội thoại"
|
| 149 |
+
|
| 150 |
+
def export_chat(chat_history):
|
| 151 |
+
"""Export chat history"""
|
| 152 |
+
if not chat_history:
|
| 153 |
+
return "❌ Không có dữ liệu để export"
|
| 154 |
+
|
| 155 |
+
try:
|
| 156 |
+
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
|
| 157 |
+
filename = f"sambanova_chat_{timestamp}.json"
|
| 158 |
+
|
| 159 |
+
with open(filename, 'w', encoding='utf-8') as f:
|
| 160 |
+
json.dump(chat_history, f, ensure_ascii=False, indent=2)
|
| 161 |
+
|
| 162 |
+
return f"✅ Đã export chat thành {filename}"
|
| 163 |
+
except Exception as e:
|
| 164 |
+
return f"❌ Lỗi export: {str(e)}"
|
| 165 |
+
|
| 166 |
+
# Kết nối events
|
| 167 |
+
webrtc.on_additional_outputs(
|
| 168 |
+
handle_additional_outputs,
|
| 169 |
+
outputs=[chatbot, conversation_state, status_display]
|
| 170 |
+
)
|
| 171 |
+
|
| 172 |
+
clear_btn.click(
|
| 173 |
+
clear_conversation,
|
| 174 |
+
outputs=[chatbot, conversation_state, status_display]
|
| 175 |
+
)
|
| 176 |
+
|
| 177 |
+
export_btn.click(
|
| 178 |
+
export_chat,
|
| 179 |
+
inputs=[chatbot],
|
| 180 |
+
outputs=[status_display]
|
| 181 |
+
)
|
| 182 |
+
|
| 183 |
+
return sambanova_tab
|
| 184 |
def create_voice_coding_tab(voice_coding_service):
|
| 185 |
"""Tạo tab Voice Coding đơn giản - Text-based trước"""
|
| 186 |
|