app.py

import gradio as gr
import random
import time

def benchmark_model(model_name, device, precision, batch_size):
    # Simulate benchmark
    time.sleep(1)
    latency = round(random.uniform(5, 50), 2)
    throughput = round(random.uniform(10, 200), 2)
    power = round(random.uniform(1, 15), 2)
    accuracy = round(random.uniform(85, 99), 2)
    
    results = f"""### Benchmark Results for {model_name}

**Device**: {device}  
**Precision**: {precision}  
**Batch Size**: {batch_size}

📊 **Performance Metrics**:
- Latency: {latency}ms
- Throughput: {throughput} FPS
- Power Consumption: {power}W
- Accuracy: {accuracy}%

**Anktechsol** - Edge AI Solutions
🔗 [Visit anktechsol.com](https://anktechsol.com)"""
    return results

with gr.Blocks(title="Edge AI Benchmarker") as demo:
    gr.Markdown("# 🚀 Edge AI Model Benchmarker")
    gr.Markdown("Benchmark AI models on edge devices - **Anktechsol**")
    
    with gr.Row():
        with gr.Column():
            model = gr.Dropdown(["MobileNetV2", "EfficientNet-B0", "YOLOv5n", "TinyBERT"], label="Model", value="MobileNetV2")
            device = gr.Dropdown(["Raspberry Pi 4", "Jetson Nano", "Coral TPU", "iPhone 13"], label="Device", value="Raspberry Pi 4")
            precision = gr.Radio(["FP32", "FP16", "INT8"], label="Precision", value="FP32")
            batch = gr.Slider(1, 32, value=1, step=1, label="Batch Size")
            btn = gr.Button("Run Benchmark")
        
        with gr.Column():
            output = gr.Markdown()
    
    btn.click(benchmark_model, inputs=[model, device, precision, batch], outputs=output)
    
    gr.Markdown("""---
### About Anktechsol
Expert AIoT & Edge AI consulting services. [Learn more →](https://anktechsol.com)""")

demo.launch()