[common] Generalized MXFP8 gated kernels w.r.t. input tensor dimensions

[Oleg-Goncharov]

Description

This PR lifts the restrictions on the shape of the input tensor for gated MXFP8 fused kernels. Similar to PR#1437, it allows an arbitrary number of rows (or the product of all tensor dimensions except the last of high-dimensional tensors).
The number of columns (or the dimensionality of the last dimension of high-dimensional tensors) must be a multiple of 32.

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

• Removed the isFullTile requirement in the quantize_gated function.
• Added second tensor of scaling factors for gate- part of the gated output tensor.
• Gated MXFP8 Test Suite: Added the alignment/padding requirement for dimensions of the tensors with scaling factors to be multiples of [128,4] for row-wise and [4,128] for column-wise scaling.
• Restricted the last dimension of the gated input tensor to multiple of 32.
• Modified exp2f_rcp function in the testing suites, so it returns 1.0 if the biased_exp==0 to avoid NaN in the amax=0 scenario.
• Commented out NVTE_CHECK in the CheckScaleTensorShape of the transformer_engine.cpp to suppress errors generated for the tensor shapes that are now supported.
• Added additional tensor descriptor into gated kernels to properly handle boundaries of the activation and the gate- part of the input/output tensors.
• Refactored the test suite.

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

[timmoon10]

LGTM, pending testing. Started pipeline 23308831.