withjustasinglelineofcodeyoucan
deploymachinelearningmodelsstraightfromJupyterNotebook(oranyothercode)implementdatapipelinesquickly,withoutmemorylimitation,allfromaPandas-likeAPIservemodelsanddatafromaneasytouseRESTAPIFurther,omega|mlisthefastestwayto
scalemodeltrainingontheincludedscalablepure-Pythoncomputecluster,onSparkoranyothercloudcollaborateondatascienceprojectseasily,sharingJupyterNotebooksdeploybeautifuldashboardsrightfromyourJupyterNotebook,usingdashserveLinksDocumentation:https://omegaml.github.io/omegaml/Contributions:https://bit.ly/omegaml-contributeGetstartedin<5minutesStarttheomega|mlserverrightfromyourlaptoporvirtualmachine
$wgethttps://raw.githubusercontent.com/omegaml/omegaml/master/docker-compose.yml$docker-composeup-dJupyterNotebookisimmediatelyavailableathttps://localhost:8899(omegamlisfuntologin).Anynotebookyoucreatewillautomaticallybestoredintheintegratedomega|mldatabase,makingcollaborationabreeze.TheRESTAPIisavailableathttps://localhost:5000.
AlreadyhaveaPythonenvironment(e.g.JupyterNotebook)?Leveragethepowerofomega|mlbyinstallingasfollows:
#assumingyouhavestartedtheserverasperabove$pipinstallomega|mlExamplesGetmoreinformationathttps://omegaml.github.io/omegaml/
#transparentlystorePandasSeriesandDataFramesoranyPythonobjectom.datasets.put(df,'stats')om.datasets.get('stats',sales__gte=100)#transparentlystoreandgetmodelsclf=LogisticRegression()om.models.put(clf,'forecast')clf=om.models.get('forecast')#runandscalemodelsdirectlyontheintegratedPythonorSparkcomputeclusterom.runtime.model('forecast').fit('stats[^sales]','stats[sales]')om.runtime.model('forecast').predict('stats')om.runtime.model('forecast').gridsearch(X,Y)#usetheRESTAPItostoreandretrievedata,runpredictionsrequests.put('/v1/dataset/stats',json={...})requests.get('/v1/dataset/stats?sales__gte=100')requests.put('/v1/model/forecast',json={...})UseCasesomega|mlcurrentlysupportsscikit-learn,KerasandTensorflowoutofthebox.Needtodeployamodelfromanotherframework?Openanissueathttps://github.com/omegaml/omegaml/issuesordropusalineatsupport@omegaml.io
MachineLearningDeploymentdeploymodelstoproductionwithasinglelineofcodeserveandusemodelsordatasetsfromaRESTAPIDataScienceCollaborationgetafullyintegrateddatascienceworkplacewithinminuteseasilysharemodels,data,jupyternotebooksandreportswithyourcollaboratorsCentralizedData&Computeclusterperformout-of-corecomputationsonapure-pythonorApacheSparkcomputeclusterhaveasharedNoSQLdatabase(MongoDB),outofthebox,workinglikeaPandasdataframeuseacomputeclustertotrainyourmodelswithnoadditionalsetupScalabilityandExtensibilityscaleyourdatascienceworkfromyourlaptoptoteamtoproductionwithnocodechangesintegrateanymachinelearningframeworkorthirdpartydatascienceplatformwithacommonAPITowardsDataSciencerecentlypublishedanarticleonomega|ml:https://towardsdatascience.com/omega-ml-deploying-data-machine-learning-pipelines-the-easy-way-a3d281569666
Inadditionomega|mlprovidesaneasy-to-useextensionsAPItosupportanykindofmodels,computecluster,databaseanddatasource.
EnterpriseEdition
https://omegaml.io
omega|mlEnterpriseEditionprovidessecurityoneverylevelandisreadymadeforKubernetesdeployment.Itislicensedseparatelyforon-premise,privateorhybridcloud.Signupathttps://omegaml.io
评论