Update quantization.md for QDQ and dynamic quantization old doc

[manickavela29]

• doc suggests that QDQ model are created with dynamic quant, which is not the case anymore.
• updating and restructuring the doc

Description

• QDQ model format representation doesn't come for dynamic quantization, but the doc was suggesting.
• May be a couple of years back this support was there but not anymore

Motivation and Context

• Clears misunderstanding that QDQ Onnx models can be created with dynamic quantization.

#20125

[manickavela29]

Hi @yufenglee ,
PR is stuck on checklinks action failure, but other similar documentations have got committed recently,
Could you please retrigger the action?
Checking 1206 external links...
Ran on 276 files!

• ./_site/blogs/accelerating-llama-2.html
  • External link https://github.com/microsoft/onnxruntime-inference-examples/blob/main/python/models/llama2/README.md failed: 404 No error

HTML-Proofer found 1 failure!

Thank you