Update app.py

app.py

@@ -1,64 +1,77 @@
+import requests
+import torch
+from PIL import Image
+from transformers import MllamaForConditionalGeneration, AutoProcessor
 import gradio as gr
-from huggingface_hub import InferenceClient
-# import spaces
+import spaces
 
 
-"""
-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("meta-llama/Llama-3.2-90B-Vision-Instruct")
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+class VisionInstructChat:
+    def __init__(self):
+        # Initialize the model and processor
+        self.model_id = "meta-llama/Llama-3.2-90B-Vision-Instruct"
+        self.model = MllamaForConditionalGeneration.from_pretrained(
+            self.model_id,
+            torch_dtype=torch.bfloat16,
+            device_map="auto",
+        )
+        self.processor = AutoProcessor.from_pretrained(self.model_id)
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+    # Method to handle the model's response to an image and text input
+    @spaces
+    def chat_with_model(self, history, image, user_text):
+        if image is None or not user_text.strip():
+            return history + [["Please upload an image and enter a prompt."]]
 
-    messages.append({"role": "user", "content": message})
+        # Prepare messages for the model
+        messages = [
+            {"role": "user", "content": [
+                {"type": "image"},
+                {"type": "text", "text": user_text}
+            ]}
+        ]
+        input_text = self.processor.apply_chat_template(messages, add_generation_prompt=True)
+        inputs = self.processor(image, input_text, return_tensors="pt").to(self.model.device)
 
-    response = ""
+        # Generate response
+        output = self.model.generate(**inputs, max_new_tokens=100)
+        response = self.processor.decode(output[0])
 
-    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
+        # Add user prompt and model response to chat history
+        history.append([user_text, response])
+        return history
 
-        response += token
-        yield response
+    # Method to reset the chat history
+    @spaces
+    def reset_chat(self):
+        return []
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    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)",
-        ),
-    ],
-)
+    # Method to create the Gradio interface
+    def launch_interface(self):
+        with gr.Blocks() as demo:
+            gr.Markdown("### Chat with Vision-Instruct Model")
 
+            # Chat history
+            chat_history = gr.Chatbot(label="Chat History")
 
-if __name__ == "__main__":
-    demo.launch()
+            # Inputs: Image and Text
+            with gr.Row():
+                with gr.Column(scale=3):
+                    image_input = gr.Image(type="pil", label="Upload Image")
+                with gr.Column(scale=7):
+                    user_input = gr.Textbox(placeholder="Type your message here...", label="Your Prompt")
+
+            # Submit and Clear buttons
+            submit_button = gr.Button("Send")
+            clear_button = gr.Button("Clear Chat")
+
+            # Button actions
+            submit_button.click(fn=self.chat_with_model, inputs=[chat_history, image_input, user_input], outputs=chat_history)
+            clear_button.click(fn=self.reset_chat, outputs=chat_history)
+
+        demo.launch()
+
+# Create an instance of the class and launch the interface
+chat_app = VisionInstructChat()
+chat_app.launch_interface()