fix: resolve custom embedding postprocessing by canonical model name

[serhiizghama]

Fixes #650. TextEmbedding resolves model names case-insensitively, but CustomTextEmbedding.__init__ then looked up the postprocessing config with the raw user-provided model_name. So registering a model as Org/Model and instantiating it as org/model blew up with KeyError: 'org/model' even though the parent had already resolved the name fine.

Switched the lookup to self.model_description.model — the canonical name the parent resolves to — so registration and instantiation stay case-insensitive.

Added a test that registers under one casing and instantiates under another; it reproduced the KeyError before the change and passes now. It uses the existing lazy_load=True / specific_model_path="./" trick so nothing gets downloaded. ruff format, mypy and the offline custom-model tests are clean.

[coderabbitai]

📝 Walkthrough

Walkthrough

CustomTextEmbedding.__init__ now uses the canonical model name from self.model_description.model to resolve pooling and normalization settings. A regression test covers mixed-case custom model registration and lowercase instantiation.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related issues

• Issue 650 — Addresses the same case-sensitive custom model postprocessing lookup.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly describes the main change: resolving custom embedding postprocessing by canonical model name.
Description check	✅ Passed	The description accurately explains the case-insensitive model lookup fix, regression test, and validation performed.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/test_custom_models.py`:
- Around line 206-207: Update the restore_custom_models_fixture to clear both
CustomTextEmbedding.SUPPORTED_MODELS and
CustomTextEmbedding.POSTPROCESSING_MAPPING during fixture setup and teardown,
ensuring cleanup also runs when construction or assertions fail and preventing
stale configuration across tests.

🪄 Autofix

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: a95557db-2f8c-47b4-9e80-c8eb052d9a57

📥 Commits

Reviewing files that changed from the base of the PR and between f613647 and e0e78a7.

📒 Files selected for processing (2)

• fastembed/text/custom_text_embedding.py
• tests/test_custom_models.py

[coderabbitai]

🩺 Stability & Availability | 🟡 Minor | ⚡ Quick win

Make custom-model cleanup exception-safe.

If construction or an assertion fails before these lines, CustomTextEmbedding.POSTPROCESSING_MAPPING retains the registered entry. The shown restore_custom_models_fixture resets SUPPORTED_MODELS but not POSTPROCESSING_MAPPING. A later test can observe stale configuration. Clear both registries in the fixture setup and teardown.

Suggested fixture fix

 def restore_custom_models_fixture():
     CustomTextEmbedding.SUPPORTED_MODELS = []
+    CustomTextEmbedding.POSTPROCESSING_MAPPING.clear()
     CustomTextCrossEncoder.SUPPORTED_MODELS = []
     yield
     CustomTextEmbedding.SUPPORTED_MODELS = []
+    CustomTextEmbedding.POSTPROCESSING_MAPPING.clear()
     CustomTextCrossEncoder.SUPPORTED_MODELS = []

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@tests/test_custom_models.py` around lines 206 - 207, Update the
restore_custom_models_fixture to clear both CustomTextEmbedding.SUPPORTED_MODELS
and CustomTextEmbedding.POSTPROCESSING_MAPPING during fixture setup and
teardown, ensuring cleanup also runs when construction or assertions fail and
preventing stale configuration across tests.