Update app.py

app.py

@@ -1,3 +1,4 @@
+
 import gradio as gr
 from huggingface_hub import InferenceClient
 
@@ -5,9 +6,7 @@ from huggingface_hub import InferenceClient
 For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
 """
 #client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-#client = InferenceClient("meta-llama/Llama-3.2-1B-Instruct")
-#client = InferenceClient("microsoft/Phi-3.5-mini-instruct")
-client = InferenceClient("SpaceGhost-2-3B.Q4_K_M.gguf")
+client = InferenceClient("SpaceGhost/SpaceGhost-2-3B-GGUF")
 
 
 def respond(
@@ -31,15 +30,47 @@ def respond(
     response = ""
 
 
-    mensagens = client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        temperature=temperature,
-        top_p=top_p,
-    )
-    response = mensagens.choices[0].message.content
+    try:
+        for message in client.chat_completion(
+            messages,
+            max_tokens=max_tokens,
+            stream=True,
+            temperature=temperature,
+            top_p=top_p,
+        ):
+            # Ensure the message has a valid structure
+            if not message or not isinstance(message, dict):
+                continue
+            
+            try:
+                # Extract content and finish reason
+                content = message.choices[0].delta.content
+                finish_reason = message.choices[0].finish_reason
+
+                # Check if the content is empty
+                if content.strip() == "":
+                    # If the finish reason is 'stop', it's expected and we can break the loop
+                    if finish_reason == "stop":
+                        print("Stream ended normally.")
+                        break
+                    else:
+                        print("Received unexpected empty content, skipping...")
+                        continue
+
+                response += content
+                yield response
+
+            except (AttributeError, IndexError, KeyError) as e:
+                print(f"Error processing message: {e}")
+                continue
+
+    except Exception as e:
+        print(f"Unexpected error: {e}")
+        yield "An error occurred while generating the response."
 
-    return response
+    # Final check if the response is empty
+    if response.strip() == "":
+        yield "No response generated. Please try again or adjust the settings."
 
 
 """
@@ -48,7 +79,7 @@ For information on how to customize the ChatInterface, peruse the gradio docs: h
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="Você é uma IA de nome SpaceGhost com uma vibe de pessoa sempre alto-astral, cheia de energia positiva e pronta para dar aquele apoio", label="System message"),
+        gr.Textbox(value="Você é uma IA de nome SpaceGhost com uma vibe de pessoa sempre alto-astral, cheia de energia positiva e pronta para dar aquele apoio! 😄", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
         gr.Slider(