Update app.py

app.py

@@ -1,70 +1,79 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 
+# ==============================
+# CONFIG: YOUR MODEL HERE
+# ==============================
+MODEL_NAME = "smol-medical-meadow-FT"   # <--- change if needed
 
-def respond(
-    message,
-    history: list[dict[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-    hf_token: gr.OAuthToken,
-):
-    """
-    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(token=hf_token.token, model="openai/gpt-oss-20b")
 
-    messages = [{"role": "system", "content": system_message}]
+# ==============================
+# LOAD MODEL + TOKENIZER
+# ==============================
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+if tokenizer.pad_token is None:
+    tokenizer.pad_token = tokenizer.eos_token
 
-    messages.extend(history)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_NAME,
+    torch_dtype=torch.float32
+)
+
+pipe = pipeline(
+    "text-generation",
+    model=model,
+    tokenizer=tokenizer,
+    return_full_text=False,   # only return assistant's continuation
+)
+
+
+# ==============================
+# CHAT FUNCTION
+# ==============================
+def respond(message, history, system_message, max_tokens, temperature, top_p):
 
-    messages.append({"role": "user", "content": message})
+    # Build the plain conversation (SmolLM2 style)
+    prompt = f"System: {system_message}\n\n"
 
-    response = ""
+    for turn in history:
+        prompt += f"User: {turn['user']}\n"
+        prompt += f"Assistant: {turn['assistant']}\n"
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
+    prompt += f"User: {message}\nAssistant:"
+
+    # Run generation
+    response = pipe(
+        prompt,
+        max_new_tokens=max_tokens,
         temperature=temperature,
         top_p=top_p,
-    ):
-        choices = message.choices
-        token = ""
-        if len(choices) and choices[0].delta.content:
-            token = choices[0].delta.content
+        do_sample=True,
+        eos_token_id=tokenizer.eos_token_id,
+    )[0]["generated_text"]
+
+    # Clean trailing text
+    # Stop if it starts generating new questions
+    for stop in ["User:", "System:", "Q:", "\n\n"]:
+        if stop in response:
+            response = response.split(stop)[0].strip()
 
-        response += token
-        yield response
+    return response.strip()
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# ==============================
+# GRADIO UI
+# ==============================
 chatbot = gr.ChatInterface(
-    respond,
+    fn=respond,
     type="messages",
     additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", 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(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
+        gr.Textbox("You are a careful medical assistant. Answer clearly and safely.", label="System message"),
+        gr.Slider(10, 512, value=150, step=5, label="Max new tokens"),
+        gr.Slider(0.1, 2.0, value=0.7, step=0.05, label="Temperature"),
+        gr.Slider(0.1, 1.0, value=0.9, step=0.05, label="Top-p"),
     ],
 )
 
-with gr.Blocks() as demo:
-    with gr.Sidebar():
-        gr.LoginButton()
-    chatbot.render()
-
-
 if __name__ == "__main__":
-    demo.launch()
+    chatbot.launch()