temporary test fix for gpt-oss nan loss

[shuningjin]

Description

Temporary fix for gpt-oss test: b/497864549

• Issue: We encounter NaN loss when finetuning gpt-oss-20b, while pretraining is mostly fine.
• Use abort_on_nan_loss=false for finetune/SFT as a temporary fix; the check is kept for pretrain. Will keep investigating cause in the bug.
• Add gcs_metrics=true to facilitate metric monitoring (e.g., loss, grad norm) and debugging.

Tests

N/A

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!