Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: Update Triton model support #485

Merged
merged 9 commits into from
Mar 13, 2024
Merged

Conversation

rafvasq
Copy link
Member

@rafvasq rafvasq commented Jan 29, 2024

Motivation

Triton introduced support for more model frameworks last year and can support xgboost, lightgbm, and more. This PR adds examples and docs to advertise this.

Modifications

  • Add newly supported models to Triton runtime config, setting autoSelect: false.
  • Add an example ISVC config for Triton-served XGBoost model.
  • Update example-models doc to reflect example models added in feat: Adds and refactors for Triton FIL examples modelmesh-minio-examples#7
  • Update model-formats README to reflect framework support and framework-specific docs to show example ISVC using Triton.
  • Add FVTs for lightgbm and xgboost deployment on Triton runtime

Result

Closes #185

Signed-off-by: Rafael Vasquez <[email protected]>
Signed-off-by: Rafael Vasquez <[email protected]>
ckadner pushed a commit to kserve/modelmesh-minio-examples that referenced this pull request Mar 5, 2024
Related to kserve/modelmesh-serving#485 and
kserve/modelmesh-serving#185, this PR expands
on `lightgbm` and `xgboost` examples to show that they can be deployed
with Triton (in addition to MLServer).

---------

Signed-off-by: Rafael Vasquez <[email protected]>
@rafvasq rafvasq marked this pull request as ready for review March 5, 2024 20:04
@oss-prow-bot oss-prow-bot bot requested review from njhill March 5, 2024 20:04
Copy link
Member

@ckadner ckadner left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank @rafvasq -- I was looking at the failed FVTs which of course are unrelated but it got me thinking about whether we should add a (few) test(s) for the new features, using the new example models?

Signed-off-by: Rafael Vasquez <[email protected]>
rafvasq added a commit to kserve/modelmesh-minio-examples that referenced this pull request Mar 13, 2024
Towards adding a couple of tests in
kserve/modelmesh-serving#485

Signed-off-by: Rafael Vasquez <[email protected]>
@rafvasq rafvasq requested a review from ckadner March 13, 2024 19:43
Copy link
Member

@ckadner ckadner left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

/lgtm

Thanks @rafvasq

Copy link

oss-prow-bot bot commented Mar 13, 2024

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ckadner, rafvasq

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

Copy link

oss-prow-bot bot commented Mar 13, 2024

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ckadner, rafvasq

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

@rafvasq rafvasq merged commit 3e755a9 into kserve:main Mar 13, 2024
7 checks passed
@rafvasq rafvasq deleted the update-triton-support branch March 13, 2024 19:52
zhlsunshine pushed a commit to zhlsunshine/modelmesh-serving that referenced this pull request Mar 22, 2024
#### Motivation
Triton introduced [support for more model frameworks last
year](https://developer.nvidia.com/blog/real-time-serving-for-xgboost-scikit-learn-randomforest-lightgbm-and-more/)
and can support xgboost, lightgbm, and more. This PR adds examples and
docs to advertise this.

#### Modifications
- Add newly supported models to Triton runtime config, setting
`autoSelect: false`.
- Add an example ISVC config for Triton-served XGBoost model.
- Update example-models doc to reflect example models added in
kserve/modelmesh-minio-examples#7
- Update model-formats README to reflect framework support and
framework-specific docs to show example ISVC using Triton.
- Add FVTs for lightgbm and xgboost deployment on Triton runtime

#### Result
Closes kserve#185

---------

Signed-off-by: Rafael Vasquez <[email protected]>
Signed-off-by: Rafael Vasquez <[email protected]>
Signed-off-by: zhlsunshine <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Can the MLServer runtime be replaced by Triton?
2 participants