litellm/.github/pull_request_template.md at ab784590973e401bf00633e9a7e36caf98d2fafb

mirror of https://github.com/BerriAI/litellm.git synced 2026-11-07 07:58:09 +00:00

jaideeppyne ab78459097 fix(cohere): handle multimodal inputs in token counting and guardrails

Response transformation crashed with TypeError when counting tokens for
multimodal (dict-shaped) embedding inputs, since encoding.encode() was
called directly on the dict instead of its text parts. Also removed a
dead, equally-buggy duplicate of the same loop in _transform_response.

Separately, the embeddings guardrail handler silently skipped multimodal
dict inputs instead of scanning them, letting prohibited text/image
content bypass the configured guardrail. It now extracts and scans the
text parts while leaving image parts untouched.

2026-08-13 09:01:47 +05:30

6.8 KiB

Raw Blame History

TLDR

Problem this solves:

How it solves it:

User Flow

Relevant issues

Linear ticket

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

I have added meaningful tests
My PR passes all CI/CD checks (e.g., lint, format, unit tests)
My PR's scope is as isolated as possible; it only solves 1 specific problem
I have received a Greptile Confidence Score of at least 4/5 before requesting a maintainer review (Greptile reviews automatically once the PR is opened; only comment @greptileai to re-request a review after pushing changes)

Delays in PR merge?

If you're seeing a delay in your PR being merged, ping the LiteLLM Team on Slack (#pr-review).

Screenshots / Proof of Fix

Type

🆕 New Feature 🐛 Bug Fix 🧹 Refactoring 📖 Documentation 🚄 Infrastructure ✅ Test

Caveats (if any)

QA runbook

Final Attestation

The tests check the right things, including the edge cases, and regressions in the respective real-world customer use-cases are not possible after this PR

6.8 KiB Raw Blame History