updated

bot/bot_websocket_server.py

@@ -89,14 +89,14 @@ async def run_bot_websocket_server(websocket_client):
     )
 
     stt = WhisperSTTService(
-        model="tiny",
+        model="smollm",
         device="cpu",
         compute_type="default",
         language="en",
     )
 
     llm = OLLamaLLMService(
-        model="llama3.1",
+        model="smoll",
         # params=OLLamaLLMService.InputParams(temperature=0.7, max_tokens=1000),
     )
 

setup.sh

@@ -1,9 +1,9 @@
-mkdir -p bot/assets
-wget https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/kokoro-v1.0.onnx -O bot/assets/kokoro-v1.0.int8.onnx
-wget https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/kokoro-v1.0.fp16-gpu.onnx -O bot/assets/kokoro-v1.0.fp16-gpu.onnx
-wget https://huggingface.co/NeuML/kokoro-base-onnx/resolve/main/voices.json -O bot/assets/voices.json
+# mkdir -p bot/assets
+# wget https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/kokoro-v1.0.onnx -O bot/assets/kokoro-v1.0.int8.onnx
+# wget https://github.com/thewh1teagle/kokoro-onnx/releases/download/model-files-v1.0/kokoro-v1.0.fp16-gpu.onnx -O bot/assets/kokoro-v1.0.fp16-gpu.onnx
+# wget https://huggingface.co/NeuML/kokoro-base-onnx/resolve/main/voices.json -O bot/assets/voices.json
 curl -fsSL https://ollama.com/install.sh | sh 
 sleep 100
 ollama serve &
 sleep 100
-ollama pull llama3.1
+ollama pull smollm