# Setup Guide for Phase 4 Testing

## Quick Start

1. **Clone the repository:**
```bash
git clone https://huggingface.co/jmurray10/phase4-quantum-compression
cd phase4-quantum-compression
```

2. **Install dependencies:**
```bash
pip install -r requirements.txt
```

3. **Test compressed models:**
```python
import torch

# Load compressed model
model = torch.load('models/mlp_compressed_int8.pth')
print(f"Model loaded successfully!")

# Test inference
test_input = torch.randn(1, 784)
output = model(test_input)
print(f"Output shape: {output.shape}")
```

4. **Run validation tests:**
```bash
python tests/test_saved_models.py
python tests/test_compressed_model_usability.py
```

## Available Models

| Model | Type | Size | Path |
|-------|------|------|------|
| MLP Original | FP32 | 943KB | `models/mlp_original_fp32.pth` |
| MLP Compressed | INT8 | 241KB | `models/mlp_compressed_int8.pth` |
| CNN Original | FP32 | 1.69MB | `models/cnn_original_fp32.pth` |
| CNN Compressed | INT8 | 483KB | `models/cnn_compressed_int8.pth` |

## Running Quantum Experiments

```python
# Example: Run Grover's algorithm
from src.quantum.qiskit.grover_aer import run_grover_experiment

result = run_grover_experiment(n_qubits=3, marked_state=5)
print(f"Success probability: {result['success_rate']:.3f}")
```

## Energy Measurement

```python
# Example: Measure model energy consumption
from src.energy.energy_logger_nvml import EnergyLogger

logger = EnergyLogger()
energy = logger.measure_inference_energy(model, test_data)
print(f"Energy consumed: {energy:.2f} J")
```

## Reproducing Results

All results can be reproduced by running the scripts in the `src/` directory.
No hardcoded values - everything is computed at runtime!