Captioning & lyrics provider fixes and improvements

[scragnog]

Captioning & lyrics provider fixes and improvements

Gemini/OpenAI masked-key fix. The frontend receives masked API keys (••••aldc, Unicode bullets) from /api/settings for display. When a caption task echoed one back, httpx rejected the non-ASCII x-goog-api-key header with an opaque encoding error. Task manager now detects masked/empty keys and falls back to the real key from settings.json or env — the same pattern already used for the HF token.

Genius lyrics provider. Search reliability was poor for anything that wasn't an exact title match:

• title/artist cleaning strips feat., remaster/deluxe suffixes and bracketed noise before querying
• artist/title swap retry when the first search misses
• fuzzy title validation rejects wrong-song results
• artist-profile pages returned as "songs" are detected and rejected
  Compatible with the Music Genius give error (Lyrics error: Genius.__init__() got an unexpected keyword argument 'verbose') #47 lyricsgenius >= 3.6 fix (no verbose kwarg; provider logger silenced instead).

local_12gb caption tier. New local captioner tier loads the model 8-bit (BitsAndBytes) for 12 GB GPUs, surfaced in GUI, wizard and CLI. The local captioner also resolves model weights from the configured checkpoint dir before falling back to the HF cache.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

@scragnog, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 59 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 6c8c2016-733e-4195-9658-6e1b1086982c

📥 Commits

Reviewing files that changed from the base of the PR and between ecd13bd and e3c37fa.

📒 Files selected for processing (9)

• frontend/index.html
• frontend/js/workspace-lab.js
• sidestep_engine/cli/args.py
• sidestep_engine/data/caption_provider_local.py
• sidestep_engine/data/lyrics_provider_genius.py
• sidestep_engine/gui/task_manager.py
• sidestep_engine/ui/flows/build_dataset_ai.py
• sidestep_engine/ui/flows/build_dataset_ai_batch.py
• train.py

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}