Create app.py

app.py

@@ -0,0 +1,59 @@
+import os
+import gradio as gr
+from together import Together
+import base64
+
+# Initialize the Together client
+client = Together(api_key=os.environ.get('TOGETHER_API_KEY'))
+
+def process_image(image):
+    # Convert the image to base64
+    buffered = BytesIO()
+    image.save(buffered, format="PNG")
+    img_str = base64.b64encode(buffered.getvalue()).decode()
+    
+    # Prepare the messages for the API call
+    messages = [
+        {"role": "system", "content": "You are an AI assistant that can analyze images and generate code based on their content."},
+        {"role": "user", "content": [
+            {"type": "image_url", "image_url": f"data:image/png;base64,{img_str}"},
+            {"type": "text", "text": "Analyze this image and generate Python code that could recreate or represent the main elements seen in the image."}
+        ]}
+    ]
+    
+    # Make the API call
+    response = client.chat.completions.create(
+        model="meta-llama/Llama-Vision-Free",
+        messages=messages,
+        max_tokens=512,
+        temperature=0.7,
+        top_p=0.7,
+        top_k=50,
+        repetition_penalty=1,
+        stop=["<|eot_id|>", "<|eom_id|>"]
+    )
+    
+    # Extract the generated code from the response
+    generated_code = response.choices[0].message.content
+    
+    # Generate HTML to display the code with syntax highlighting
+    html_output = f"""
+    <pre><code class="language-python">{generated_code}</code></pre>
+    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.7.0/styles/default.min.css">
+    <script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.7.0/highlight.min.js"></script>
+    <script>hljs.highlightAll();</script>
+    """
+    
+    return html_output
+
+# Create the Gradio interface
+iface = gr.Interface(
+    fn=process_image,
+    inputs=gr.Image(type="pil"),
+    outputs=gr.HTML(),
+    title="Llama Vision Free Code Generation",
+    description="Upload an image, and this demo will use the Llama Vision Free model to analyze it and generate relevant Python code."
+)
+
+# Launch the interface
+iface.launch()