Update app.py

app.py

@@ -50,6 +50,11 @@ hf_hub_download(repo_id="TheBloke/Mistral-7B-Instruct-v0.1-GGUF", local_dir=".",
 mistral_model_path="./mistral-7b-instruct-v0.1.Q5_K_M.gguf"
 mistral_llm = Llama(model_path=mistral_model_path,n_gpu_layers=35,max_new_tokens=256, context_window=4096, n_ctx=4096,n_batch=128,verbose=False)
 
+# Load Saul-Instruct-v1-GGUF.Q4_K_M
+print("Loading Saul-Instruct-v1-GGUF.Q4_K_M")
+hf_hub_download(repo_id="MaziyarPanahi/Saul-Instruct-v1-GGUF", local_dir=".", filename="Saul-Instruct-v1-GGUF.Q4_K_M.gguf")
+saul_model_path="./Saul-Instruct-v1-GGUF.Q4_K_M.gguf"
+saul_instruct_llm = Llama(model_path=saul_model_path,n_gpu_layers=35,max_new_tokens=256, context_window=4096, n_ctx=32768,n_batch=128,verbose=False)
 
 # Load XTTS Model
 print("Loading XTTS model")
@@ -71,13 +76,13 @@ xtts_model.cuda()
 
 ###### Set up Gradio Interface ######
 
-with gr.Blocks(title="Voice chat with LLM") as demo:
-    DESCRIPTION = """# Voice chat with LLM"""
+with gr.Blocks(title="Voice chat with Saul-Instruct-v1-GGUF") as demo:
+    DESCRIPTION = """# Voice chat with Saul-Instruct-v1-GGUF"""
     gr.Markdown(DESCRIPTION)
 
     # Define chatbot component
     chatbot = gr.Chatbot(
-        value=[(None, "Hi friend, I'm Amy, an AI coach. How can I help you today?")],  # Initial greeting from the chatbot
+        value=[(None, "Hi friend, I'm you data protection assistant. How can I help you today?")],  # Initial greeting from the chatbot
         elem_id="chatbot",
         avatar_images=("examples/hf-logo.png", "examples/ai-chat-logo.png"),
         bubble_full_width=False,
@@ -152,7 +157,7 @@ with gr.Blocks(title="Voice chat with LLM") as demo:
                 yield from handle_speech_generation(sentence, chatbot_history, chatbot_voice)
         else:
             # Continuously get and process sentences from a generator function
-            for sentence, chatbot_history in get_sentence(chatbot_history, mistral_llm):
+            for sentence, chatbot_history in get_sentence(chatbot_history, saul_instruct_llm):
                 print("Inserting sentence to queue")
                 yield from handle_speech_generation(sentence, chatbot_history, chatbot_voice)
 
@@ -170,6 +175,7 @@ with gr.Blocks(title="Voice chat with LLM") as demo:
             This Space demonstrates how to speak to an llm chatbot, based solely on open accessible models.
             It relies on the following models :
             - Speech to Text Model: [Faster-Whisper-large-v3](https://huggingface.co/Systran/faster-whisper-large-v3) an ASR model, to transcribe recorded audio to text.
+            - Legal Large Language Model: [MaziyarPanahi/Saul-Instruct-v1-GGUF](https://huggingface.co/MaziyarPanahi/Saul-Instruct-v1-GGUF/blob/main/Saul-Instruct-v1.Q4_K_M.gguf) a LLM to generate legal chatbot responses. 
             - Large Language Model: [Mistral-7b-instruct-v0.1-quantized](https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF) a LLM to generate the chatbot responses. 
             - Text to Speech Model: [XTTS-v2](https://huggingface.co/spaces/coqui/xtts) a TTS model, to generate the voice of the chatbot.