Update app.py

app.py

@@ -1,12 +1,26 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
+!pip install bitsandbytes accelerate gradio
+from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
+import torch
 
-"""
-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")
+# Define BitsAndBytesConfig
+bnb_config = BitsAndBytesConfig(load_in_4bit=True,
+                                bnb_4bit_quant_type="nf4",
+                                bnb_4bit_compute_dtype=torch.float16)
 
+# Model name
+model_name = "ruslanmv/Medical-Llama3-v2"
 
+# Load tokenizer and model with BitsAndBytesConfig
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True, bnb_config=bnb_config)
+model = AutoModelForCausalLM.from_pretrained(model_name, config=bnb_config)
+
+# Ensure model is on the correct device
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+
+# Define the respond function
 def respond(
     message,
     history: list[tuple[str, str]],
@@ -25,39 +39,49 @@ def respond(
 
     messages.append({"role": "user", "content": message})
 
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+    # Format the conversation as a single string for the model
+    prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+    inputs = tokenizer(prompt, return_tensors="pt", truncation=True, padding=True, max_length=1000)
+    
+    # Move inputs to device
+    input_ids = inputs['input_ids'].to(device)
+    attention_mask = inputs['attention_mask'].to(device)
+    
+    # Generate the response
+    with torch.no_grad():
+        outputs = model.generate(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            max_length=max_tokens,
+            temperature=temperature,
+            top_p=top_p,
+            use_cache=True
+        )
+    
+    # Extract the response
+    response_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)[0]
+    
+    # Remove the prompt and system message from the response
+    response_text = response_text.replace(system_message, '').strip()
+    response_text = response_text.replace(f"Human: {message}\n\nAssistant: ", '').strip()
+    
+    return response_text
 
-        response += token
-        yield response
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Create the Gradio interface
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
+        gr.Textbox(value="You are a Medical AI Assistant. Please be thorough and provide an informative answer. If you don't know the answer to a specific medical inquiry, advise seeking professional help.", label="System message", lines=3),
         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.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
     ],
-)
+    title="Medical AI Assistant",
+    description="Give me your symptoms and ask me a health problem. The AI will provide informative answers. If the AI doesn't know the answer, it will advise seeking professional help.",
+   
+    examples=[["I have a headache and a fever. What should I do?"], ["What are the symptoms of diabetes?"], ["How can I improve my sleep?"]],
 
+)
 
 if __name__ == "__main__":
     demo.launch()