[feat] Add phase-aware modular training performance metrics

[Jack47]

Summary

• add per-step optimizer/sample throughput and per-role Transformer call accounting to the modular trainer
• estimate Wan Transformer-core TFLOP/s and MFU using actual BDir/VSA, block-causal AR, streaming KV-cache, and DMD multi-role/multi-rollout geometry
• expose an optional peak BF16 override, document metric semantics and limitations, and keep all performance hooks configurable

Why

The existing standalone MFU helper assumes one dense Wan forward. That does not represent causal chunking, sparse attention, DMD teacher/critic calls, or repeated student rollouts. Counting each role at the Transformer boundary makes the metric follow the algorithm actually executed.

Validation

• 40 targeted CPU tests passed, including existing Trainer managed-optimization coverage
• pre-commit passed: YAPF, Ruff, codespell, PyMarkdown, mypy, filename and suggestion checks

GPU validation was not run locally; FLOP geometry is covered with CPU unit tests.

[github-actions]

Welcome to FastVideo! Thanks for your first pull request.

How our CI works:

PRs run a three-tier CI system:

1. Pre-commit — formatting (yapf), linting (ruff), type checking (mypy). Runs immediately on every PR.
2. Fastcheck — six core GPU lanes run automatically via Buildkite (~10-15 min).
3. Merge gate — a reviewer adds ready; changed paths select only the relevant integration, training, golden, or SSIM coverage.

Before your PR is reviewed:

• pre-commit run --all-files passes locally
• You've added or updated tests for your changes
• The PR description explains what and why

If pre-commit fails, a bot comment will explain how to fix it. Fastcheck and merge-gate results appear in the Checks section below.

Useful links:

• Contributing Guide
• Development Roadmap
• Slack

[mergify]

Merge Protections

🔴 1 of 1 protections blocking · waiting on 👀 reviews and 🤖 CI

	Protection	Waiting on
🔴	PR merge requirements	👀 reviews and 🤖 CI

🔴 PR merge requirements

Waiting for

• #approved-reviews-by>=1
• check-success=fastcheck-passed
• check-success=full-suite-passed

This rule is failing.

• #approved-reviews-by>=1
• check-success=fastcheck-passed
• check-success=full-suite-passed
• check-success~=pre-commit
• title~=(?i)^\[(feat|feature|bugfix|fix|refactor|perf|ci|doc|docs|misc|chore|kernel|new.?model|skill|skills|infra)\]