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
Build the TF Serving GPU image as part of a release workflow. #339
Conversation
The presubmit timed out waiting for the workflow to finish but it looks like it eventually succeeded. The problem is that building the GPU image is relatively slow since we have to recompile everything. |
* This creates a workflow for building the GPU image. * This is the first step in building a single workflow that builds all the artifacts needed in a release. * We don't want to build the GPU image on each presubmit because building the GPU image is too slow.
/uncc @wbuchwalter |
#347 should be submitted first but this is ready for review. |
/hold |
/lgtm |
/lgtm |
/hold cancel |
/approve |
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: jlewi, lluunn, yupbank 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 |
…kubeflow#339) * Suggestion for Neural Architecture Search with Reinforcement Learning * Add NAS RL Suggestion * Fix new line * set json format for GetSuggestion() * finish trial return in GetSuggestion(), finish GetEvaluationHistory, and fix bugs * fix a bug in GetEvaluationResult() * fix bigs in GetEvaluationResult * fix an error in GetEvaluatinResult * Add python Katib api * Remove unnecessary requirements * add about for suggestion * rename to README * Add picture explanations; make the printouts more organized * fix typos * fix some small problems * Fix several problems * Fix a typo * fix some problems * small fixes * Suggestion do not need to handle uncompleted trials * fix a small problem
changed my account from karkumar to kramachandran
Create an initial release Argo workflow to build all our release artifacts.
Build the TF GPU serving image as part of this workflow (as well as CPU version).
This is the first step towards creating a separate Argo workflow to build all our release artifacts.
I think it makes sense to have a separate workflow for our releases rather than repurposing our E2E tests.
* Build all artifacts
* Update ksonnet configs
* run all E2E tests
Immediate use of this workflow is to start automating builds of the GPU image for TF serving.
Fix #338
This change is