refactor(trainer): extract the sync-loop template into BaseTrainer

[haonan3]

Summary

Extracts the synchronous training loop into a BaseTrainer template method. Nine trainers hand-copy the same outer skeleton today (~700–900 duplicated lines repo-wide, measured); the diffusion and AR train() bodies differ in exactly four lines. BaseTrainer.train() now owns the loop — load checkpoint → data fast-forward → wandb → baseline eval → per-rollout build / sync-gate / train_step / log / periodic eval / save → finally finish-wandb + on_train_end — with four hooks for the observed divergences: _wandb_extra, evaluate_baseline, evaluate_periodic, on_train_end (AR routes _shutdown_runtime through it, preserving its nested-finally ordering).

DiffusionTrainer and ARTrainer drop their train() bodies (line-equivalent to the template, including the resume force-sync semantics and per-module progress-logger names). PETrainer / UnifiedModelTrainer / SFTTrainer keep their own loops and migrate in follow-ups; async/agentic loops are intentionally out of scope (producer/consumer shapes, not this template). Independent of #210 (based on main).

Test Plan

Static (CPU): DiffusionTrainer.train is BaseTrainer.train and ARTrainer.train is BaseTrainer.train assert-verified; ARTrainer.on_train_end override confirmed; compose-check diffusion/sd3/sd3_trainside and ar/qwen3_grpo_4b_base_dapo_sglang rc=0; pre-commit run --all-files green.

Pending GPU parity before merge:

• diffusion smoke (sd3_trainside, ~10 rollouts) — identical logs/cadence vs main, incl. a resume-from-checkpoint step (force-sync path);
• AR smoke (qwen3 grpo, ~5 rollouts) with eval_interval>0 — baseline + periodic eval fire at the same steps; teardown runs on both clean exit and Ctrl-C.

Compatibility / Risk

No config or entrypoint changes; train(**kwargs) signature is unchanged for both migrated trainers. Behavioral risk is confined to the four hook seams — hence the explicit parity checklist above. Not-yet-migrated trainers are untouched.

AI-assisted; reviewed and directed by the maintainer.

[haonan3]

Closing by triage, not by verdict on the idea. The duplication is real and has grown since this was cut — ar / diffusion / pe / unified_model (and experimental refl) all hand-roll the same load→wandb→loop→save→finish skeleton. But the trainer surface is in an active evolution burst (AR PPO/GAE #259, value head #256, hi3 it2i #186), a template landed now adds hook-friction to every in-flight change, and this draft only migrated 2 of the 5 candidates with no GPU parity smoke.

Re-derive when the AR-PPO wave settles, scoped to all five sync trainers in one pass — the branch (refactor/base-trainer-sync-template) stays as reference.