Update app.py

app.py

@@ -1,70 +1,74 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
 
-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")
+# ----------------------------------------------------
+# LOAD YOUR FINE–TUNED MODEL  (LOCAL)
+# ----------------------------------------------------
+MODEL_PATH = "smol-medical-meadow-FT"   # change if your folder name is different
 
-    messages = [{"role": "system", "content": system_message}]
+tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
+tokenizer.pad_token = tokenizer.eos_token
 
-    messages.extend(history)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_PATH,
+    device_map="auto",
+    torch_dtype=torch.float32,
+)
+model.config.pad_token_id = tokenizer.eos_token_id
+model.config.use_cache = False  # safer for smaller models
+
+
+# ----------------------------------------------------
+# CHAT FUNCTION (LOCAL GENERATION)
+# ----------------------------------------------------
+def respond(message, history, system_message, max_tokens, temperature, top_p):
+    
+    # Convert gradio history to simple text conversation
+    conversation = system_message + "\n"
 
-    messages.append({"role": "user", "content": message})
+    for turn in history:
+        conversation += f"User: {turn['user']}\nAssistant: {turn['assistant']}\n"
 
-    response = ""
+    # Current user message
+    prompt = conversation + f"User: {message}\nAssistant:"
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+
+    output_stream = model.generate(
+        **inputs,
+        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,
+    )
+
+    # Decode only the assistant's generated part
+    generated = output_stream[0][inputs["input_ids"].shape[1]:]
+    answer = tokenizer.decode(generated, skip_special_tokens=True).strip()
 
-        response += token
-        yield response
+    yield answer
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# ----------------------------------------------------
+# GRADIO UI
+# ----------------------------------------------------
 chatbot = gr.ChatInterface(
     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(value="You are a helpful medical assistant.", label="System message"),
+        gr.Slider(minimum=10, maximum=512, value=150, step=1, label="Max new tokens"),
+        gr.Slider(minimum=0.1, maximum=1.5, value=0.7, step=0.05, label="Temperature"),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.9, step=0.05, label="Top-p"),
     ],
 )
 
-with gr.Blocks() as demo:
-    with gr.Sidebar():
-        gr.LoginButton()
+demo = gr.Blocks()
+with demo:
     chatbot.render()
 
-
 if __name__ == "__main__":
     demo.launch()