Clarify critic_step_ratio parameter - no changes needed

[Copilot]

After reviewing the code logic, the original README description for --critic_step_ratio was correct. No changes are needed.

Analysis

The training loop at line 295 in fine-tune.py:

if not is_warmup and step % args.critic_step_ratio == 0:
    # upscaler training code

With critic_step_ratio = 5:

• Critic trains every step (1, 2, 3, 4, 5, ...)
• Upscaler trains only when step % 5 == 0 (steps 5, 10, 15, ...)
• Result: 5 critic updates per 1 upscaler update

The existing description "The number of critic updates for every 1 upscaler update" accurately describes this behavior.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[andrewdalpino]

So you created an empty PR just to tell me that I was correct?