Update app.py

app.py

@@ -1,36 +1,48 @@
+# app.py
+
 import gradio as gr
-from transformers import pipeline
-import os
-
-# Set your Hugging Face API key
-os.environ["HUGGING_FACE_HUB_TOKEN"] = "your_huggingface_api_key_here"
-
-# Load the model using Hugging Face's Inference API
-model_name = "kingabzpro/Gemma-2-9b-it-chat-doctor"
-generator = pipeline("text-generation", model=model_name, use_auth_token=True)
-
-# Define the response generation function
-def generate_response(user_input):
-    # Generate response using the Hugging Face API
-    response = generator(user_input, max_length=500, do_sample=True, top_p=0.9, temperature=0.8)
-    return response[0]["generated_text"]
-
-# Define Gradio interface for the chatbot
-with gr.Blocks() as healthcare_chatbot:
-    gr.Markdown("<h1>Healthcare Chatbot</h1><p>Ask any healthcare-related questions!</p>")
-    
-    chatbox = gr.Chatbot()
-    msg = gr.Textbox(placeholder="Type your healthcare question here...")
-    submit_button = gr.Button("Ask")
-
-    # Define the chatbot interaction function
-    def respond(chat_history, message):
-        response = generate_response(message)
-        chat_history.append((message, response))
-        return chat_history
-
-    # Set up the components for interaction
-    submit_button.click(respond, [chatbox, msg], chatbox)
-
-# Launch the Gradio interface
-healthcare_chatbot.launch()
+from transformers import AutoTokenizer, AutoModelForMaskedLM, pipeline
+import torch
+
+# Step 1: Set up Bio_ClinicalBERT model for medical text generation
+device = 0 if torch.cuda.is_available() else -1  # Set to 0 for GPU, -1 for CPU
+model_name = "emilyalsentzer/Bio_ClinicalBERT"  # Bio_ClinicalBERT model for medical context
+
+# Load tokenizer and model
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForMaskedLM.from_pretrained(model_name)
+
+# Create a text generation pipeline with the loaded model
+chatbot = pipeline(
+    "text-generation",
+    model=model,
+    tokenizer=tokenizer,
+    device=device,
+    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32
+)
+
+# Step 2: Function to generate chatbot responses
+def get_response(user_input):
+    response = chatbot(
+        user_input,
+        max_length=75,  # Adjusted to manage memory usage
+        num_return_sequences=1,
+        truncation=True
+    )
+    return response[0]['generated_text']
+
+# Step 3: Create Gradio interface
+def chatbot_interface(user_input):
+    return get_response(user_input)
+
+iface = gr.Interface(
+    fn=chatbot_interface,
+    inputs=gr.Textbox(lines=2, placeholder="Enter your symptoms here..."),
+    outputs="text",
+    title="Health Chatbot",
+    description="Ask your symptoms and get advice!",
+    theme="default"
+)
+
+# Step 4: Launch the Gradio app
+iface.launch()