Layernorm forward optimization

[eliotwang]

Description

Optimized the LayerNorm kernel for specific shapes, with performance results shown in the figure. The optimizations also demonstrate good generalizability.

Fixes # (issue)

Type of change

• Documentation change (change only to the documentation, either a fix or a new content)
• 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)
• Infra/Build change
• Code refactoring

Changes

• Optimized performance when shapes match the tuned branch by offline tuning.
• The performance improvement of the LayerNorm forward general kernel relies on algorithm optimization.
  Changed from first calculating the mean then calculating the variance to obtaining both mean and variance in a single pass based on Welford algorithm.
  1.Added a Stats_ge struct in utils.cu.
  2.Extended the kernel_traits class with an additional template parameter (defaulting to the Stats struct used by the tuned kernel).
  3.Modified kernel launches for the general case to explicitly specify the kernel_traits parameter.

Checklist:

• I have read and followed the contributing guidelines
• The functionality is complete
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes