For detail please refer to Kubeflow DocsĪdl submit trial jobs to AdaptDL, NNI support AdaptDL on Kubernetes cluster.
Kubeflow submit trial jobs to kubeflow, NNI support kubeflow based on normal kubernetes and azure kubernetes. For more details of pai configuration, please refer to Guide to PAI Mode Pai submit trial jobs to OpenPAI of Microsoft. Remote submit trial jobs to remote ubuntu machines, and machineList field should be filed in order to set up SSH connection to remote machine. Local run an experiment on local ubuntu machine. Specifies the platform to run the experiment, including local, remote, pai, kubeflow, frameworkcontroller. If current trial job reach the max duration time, this trial job will stop. Ma圎xecDuration specifies the max duration time of an experiment. If trialGpuNum is bigger than the free gpu numbers, and the trial jobs running simultaneously can not reach trialConcurrency number, some trial jobs will be put into a queue to wait for gpu allocation. Specifies the max num of trial jobs run simultaneously. The name of the author who create the experiment. Light weight (without Annotation and Assessor)ĪuthorName : experimentName : trialConcurrency : ma圎xecDuration : maxTrialNum : #choice: local, remote, pai, kubeflow trainingServicePlatform : #choice: true, false, default: false useAnnotation : #choice: true, false, default: false multiThread : tuner : #choice: TPE, Random, Anneal, Evolution builtinTunerName : classArgs : #choice: maximize, minimize optimize_mode : gpuIndices : assessor : #choice: Medianstop builtinAssessorName : classArgs : #choice: maximize, minimize optimize_mode : trial : command : codeDir : gpuNum : #machineList can be empty if the platform is local machineList : - ip : port : username : passwd : Configuration Spec ¶ authorName ¶ This document describes the rules to write the config file, and provides some examples and templates. The path of the config file is provided to nnictl. It is still supported for now, but we recommend users to use the new version of experiment configuration (V2).Ī config file is needed when creating an experiment. PM> Install-Package -Version 3.1.7Īfter that, we need to add the values we’ve written down while creating the vault in our appsettings.This is the previous version (V1) of experiment configuration specification.
To complete step 5, go to the “Certificates & secrets” in your app, and then “Upload certificate”.įirst, we need to install a NuGet package to support our KeyVault integration:
Openssl pkcs12 -export -out -inkey -in pfx file from the key and certificate files: Openssl req -x509 -newkey rsa:4096 -keyout key.pem -out cert.pem -days 365
If you don’t have OpenSSL on Windows download it and Install it.Next, since we want to connect to the app securely, we need to create and upload a certificate.Ĭertificate creation has a few steps, and we recommend using OpenSSL to do it since it has all the tools we need. Write it down, since we’re going to need it later. Click on it, and check out the “Application (client) ID”. Once the app creation is over, you can find it in the apps list. Just give your app a decent name, and click “Register”. To do that, head on to “Azure Active Directory”, then “App Registration” and click on “New Registration”. In order to connect to our Azure Key Vault, we’ll need to register an Azure app.