fix(generate.py): add mock FlexDecoding test integration for speculative pipeline verification

[Shraman123]

This PR introduces a safe and deterministic FlexDecoding mock validation path in generate.py, enabling speculative decoding pipeline testing on Windows and non-Inductor builds.

Key points:

Disables TorchDynamo and TorchInductor safely via environment variables.

Adds fallback Hugging Face tokenizer when SentencePiece is unavailable.

Introduces DummyModel for verifying speculative and FlexDecoding flow without GPU or compiler dependencies.

Verified on Windows 11 + PyTorch 2.6 nightly:
⚠️ SentencePiece tokenizer not found — using Hugging Face tokenizer instead.
⚠️ Using DummyModel (mock) for FlexDecoding pipeline test.
Output: <|endoftext|>Hello memory map mechan efforts]. 80 Bill seemed rap Iraq
cc @meta-llm-maintainers @bf-wrappers for review.

[meta-cla]

Hi @Shraman123!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at cla@meta.com. Thanks!

[meta-cla]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[Shraman123]

Hi maintainers, looping you in for review.

@meta-llm-maintainers
@bf-wrappers

This PR introduces a mock FlexDecoding execution path that allows deterministic speculative-decoding validation on Windows and non-Inductor environments. It does not affect production decode paths.

All checks are green and CLA is signed.
Happy to make any refinements you suggest!