Add smooth quant

[mht-sharma]

What does this PR do?

Integrates the SmoothQuant into the Optimum ONNXRuntime Quantizer.

1. The implementation uses INC behind the scenes for the quantisation.
2. To apply smooth quant call the quantizer.apply_smooth_quant before the calibration step in the regular quantization process.

Usage

# Load calibration dataset
calibration_dataset = quantizer.get_calibration_dataset(...)

# Apply smooth quantization to the model
quantizer.apply_smooth_quant(
    dataset=calibration_dataset,
    save_dir=save_dir,
    quantization_config=qconfig,
)

# Perform calibration
quantizer.fit(...)

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?

[fxmarty]

This is not recommended, reference: https://peps.python.org/pep-0008/#imports

[mht-sharma]

This is something done by the ONNXRuntime source code too! Are u suggesting importing it at top level? https://github.com/microsoft/onnxruntime/blob/0f72739b6db129373d221483d61d6637ec11fb28/onnxruntime/python/tools/quantization/quantize.py#L421