Add Quantization Introduction and Bug Fix

[Uxito-Ada]

Description

• Add a quantization guide that connects the underlying concepts to TeleFuser's implemented quantization paths.
• Fix the offline converter's Linear dtype option.
• Configure MathJax so the formulas render correctly.

Motivation

Explain quantization clearly and fix bug

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Performance improvement
• Code refactoring
• Documentation update
• Other (please describe):

Changes Made

• Refine quantization docs
• Add MathJax configuration so formulas render in docs
• fix the converter's linear_type argument to linear_dtype, which was not aligned with CLI option and docs.

Testing

• Unit tests pass (pytest tests/)
• Manual testing performed
• Benchmarks added/updated (if applicable)

Test commands:

The targeted test run passed all 17 tests. The strict English and Chinese documentation build also passed.

Checklist

• Code follows the project's coding standards (ruff)
• Pre-commit hooks pass (pre-commit run --all-files)
• All tests pass (pytest tests/)
• New tests added for new functionality
• Documentation updated (README, CLAUDE.md, docstrings)
• Commit messages are clear and descriptive
• PR title follows the convention: [TYPE] Brief description

Related Issues

N/A

Additional Notes

• This PR does not change TorchAO API selection or add a quantization backend.

GPU Architecture Support

• SM80 (Ampere, Ada Lovelace)
• SM90 (Hopper H100)
• SM100+ (Blackwell)

N/A: this PR does not add or modify GPU kernels.

[Uxito-Ada]

This change is because we have used linear_dytpe in docs, e.g. here, so changed the codes to make an alignment.

[lzx1413]

LGTM