milvus-io/milvus

[Feature]: Implement serving ONNX model in Milvus

Open

#24,955 opened on Jun 16, 2023

View on GitHub
 (1 comment) (3 reactions) (1 assignee)Go (44,298 stars) (4,000 forks)batch import
good first issuekind/feature

Description

Is there an existing issue for this?

  • I have searched the existing issues

Is your feature request related to a problem? Please describe.

ONNX is a computation backend serving models.

Milvus will need to run ONNX cases in following scenarios:

  1. User defined metrics
  2. Ranking/Reranking
  3. Embedding finetuning -> UDF function to finetune the embeddings ..... more suggestion on the usages

Describe the solution you'd like.

No response

Describe an alternate solution.

No response

Anything else? (Additional Context)

No response

Contributor guide

Tech stack
go
Domain
backendmachine learningdatabase
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
5
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
over 1 week
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
blocked
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
needs investigation
Prerequisites
Familiarity with ONNX runtimeUnderstanding of Milvus architecture
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
15
Research direction
Start by examining the Milvus codebase's existing model serving or embedding components, particularly any Python or C++ bindings. Review the ONNX runtime documentation for integration patterns. Check the assignee's recent activity and any related issues or PRs linked to this feature. Focus on the specific use cases mentioned: user defined metrics, ranking/reranking, and embedding finetuning.