(self.webpackChunk_N_E=self.webpackChunk_N_E||[]).push([[2239],{6684:function(e,t,a){(window.__NEXT_P=window.__NEXT_P||[]).push(["/install/compliance",function(){return a(8592)}])},8592:function(e,t,a){"use strict";a.r(t),a.d(t,{__toc:function(){return d}});var n=a(5893),o=a(2673),r=a(7913),i=a(3215);a(9128);var s=a(2643);let d=[{depth:2,value:"Are MOSTLY AI container images OCI Compliant?",id:"are-mostly-ai-container-images-oci-compliant"},{depth:2,value:"Can I run MOSTLY AI on any Kubernetes distribution?",id:"can-i-run-mostly-ai-on-any-kubernetes-distribution"},{depth:2,value:"How do I make sure to run MOSTLY AI in an OCI Compliant environment?",id:"how-do-i-make-sure-to-run-mostly-ai-in-an-oci-compliant-environment"},{depth:2,value:"Do you scan your images for vulnerabilities?",id:"do-you-scan-your-images-for-vulnerabilities"}];function _createMdxContent(e){let t=Object.assign({h1:"h1",p:"p",h2:"h2",a:"a",ul:"ul",li:"li"},(0,s.a)(),e.components);return(0,n.jsxs)(n.Fragment,{children:[(0,n.jsx)(t.h1,{children:"OCI compliance and security of MOSTLY AI container images"}),"\n",(0,n.jsx)(t.p,{children:"Learn about the OCI compliance and security scans of MOSTLY AI container images."}),"\n",(0,n.jsx)(t.h2,{id:"are-mostly-ai-container-images-oci-compliant",children:"Are MOSTLY AI container images OCI Compliant?"}),"\n",(0,n.jsxs)(t.p,{children:["While MOSTLY AI images do not go through a formal OCI compliance check, we consider the images compliant on the ",(0,n.jsx)(t.a,{href:"https://github.com/opencontainers/image-spec",children:"Image Specification"})," and ",(0,n.jsx)(t.a,{href:"https://github.com/opencontainers/distribution-spec",children:"Distribution Specification"}),"."]}),"\n",(0,n.jsx)(t.p,{children:"MOSTLY AI uses Red Hat Universal Base Images (UBI) as a base image, builds images with Docker, and stores them in DockerHub."}),"\n",(0,n.jsx)(t.h2,{id:"can-i-run-mostly-ai-on-any-kubernetes-distribution",children:"Can I run MOSTLY AI on any Kubernetes distribution?"}),"\n",(0,n.jsx)(t.p,{children:"MOSTLY AI does not impose a Kubernetes distribution to customers. You are free to run MOSTLY AI on any Kubernetes distribution you like."}),"\n",(0,n.jsx)(t.p,{children:"The list of the officially supported distributions are:"}),"\n",(0,n.jsxs)(t.ul,{children:["\n",(0,n.jsx)(t.li,{children:"Amazon Elastic Kubernetes Service (EKS)"}),"\n",(0,n.jsx)(t.li,{children:"Google Kubernetes Engine (GKE)"}),"\n",(0,n.jsx)(t.li,{children:"Azure Kubernetes Service (AKS)"}),"\n",(0,n.jsx)(t.li,{children:"Red Hat OpenShift"}),"\n",(0,n.jsx)(t.li,{children:"Vanilla Kubernetes"}),"\n"]}),"\n",(0,n.jsx)(t.h2,{id:"how-do-i-make-sure-to-run-mostly-ai-in-an-oci-compliant-environment",children:"How do I make sure to run MOSTLY AI in an OCI Compliant environment?"}),"\n",(0,n.jsxs)(t.p,{children:["MOSTLY AI assumes that the officially supported Kubernetes distributions are compliant with the ",(0,n.jsx)(t.a,{href:"https://github.com/opencontainers/runtime-spec",children:"Runtime Specification"})," of OCI."]}),"\n",(0,n.jsx)(t.p,{children:"If you are using a different Kubernetes distribution, check with the vendor if the runtime environment is OCI compliant."}),"\n",(0,n.jsx)(t.h2,{id:"do-you-scan-your-images-for-vulnerabilities",children:"Do you scan your images for vulnerabilities?"}),"\n",(0,n.jsxs)(t.p,{children:["MOSTLY AI uses ",(0,n.jsx)(t.a,{href:"https://trivy.dev/",children:"Trivy Open Source"})," vulnerability scanner in CI/CD pipelines and the scans of DockerHub, when pushing to the MOSTLY AI repository."]})]})}let c={MDXContent:function(){let e=arguments.length>0&&void 0!==arguments[0]?arguments[0]:{},{wrapper:t}=Object.assign({},(0,s.a)(),e.components);return t?(0,n.jsx)(t,{...e,children:(0,n.jsx)(_createMdxContent,{...e})}):_createMdxContent(e)},pageOpts:{filePath:"pages/install/compliance.mdx",route:"/install/compliance",frontMatter:{title:"OCI compliance and security of MOSTLY AI container images",description:""},timestamp:1725279046e3,pageMap:[{kind:"Meta",data:{"-- Welcome":{type:"separator",title:"INTRODUCTION"},index:{title:"Welcome"},"quick-start":"Quick start",concepts:"Concepts","best-practices":"Best practices","-- Assistant":{type:"separator",title:"ASSISTANT"},assistant:"Assistant","-- Create synthetic data":{type:"separator",title:"TRAINING & GENERATION"},generators:"Generators","synthetic-datasets":"Synthetic datasets",connectors:"Connectors",datasets:"Demo datasets","-- Global features":{type:"separator",title:"GLOBAL FEATURES"},search:"Search",usage:"Usage and credits","-- Installation and administration":{type:"separator",title:"INSTALLATION AND ADMINISTRATION"},install:"Installation",administration:"Administration","-- Releases and support":{type:"separator",title:"RELEASES AND SUPPORT"},"whats-new":"Release notes","release-lifecycle":"Release support lifecycle",support:"Support","pythonclient-header":{title:"Python Client",type:"page",href:"/python-client"},"python-client":{title:"Python Client",display:"hidden"},api:{title:"REST API ↗",type:"page",href:"https://api-docs.mostly.ai",newWindow:!0},"tutorials-header":{title:"Tutorials",type:"page",href:"/tutorials"},tutorials:{title:"Tutorials",display:"hidden"}}},{kind:"Folder",name:"administration",route:"/administration",children:[{kind:"Meta",data:{compute:"Compute",models:"Models","users-groups":"Users and groups","backup-restore":"Backup and restore"}},{kind:"MdxPage",name:"backup-restore",route:"/administration/backup-restore",frontMatter:{title:"Backup and restore",description:"You can backup and restore MOSTLY AI just like any other Kubernetes application in your infrastructure by using Velero."}},{kind:"MdxPage",name:"compute",route:"/administration/compute",frontMatter:{title:"Compute",description:"With flexible compute, you can add any compute resources needed to complete your GenAI workloads in MOSTLY AI."}},{kind:"MdxPage",name:"models",route:"/administration/models",frontMatter:{title:"Models configuration",description:"Unstructured text generation in MOSTLY AI is available via the configuration of any text generation model available on Hugging Face as well as via the built-in LSTM model (that is not pre-trained) provided by MOSTLY AI."}},{kind:"MdxPage",name:"users-groups",route:"/administration/users-groups",frontMatter:{title:"Manage users and groups",description:"When you deploy and host MOSTLY AI for your organization, you can manage access to MOSTLY AI with the Keycloak and Access Management services."}}]},{kind:"Folder",name:"assistant",route:"/assistant",children:[{kind:"Meta",data:{"prompt-examples":"Prompt examples","manage-threads":"Manage threads","python-session-status":"Python session status",configuration:"Configuration"}},{kind:"MdxPage",name:"configuration",route:"/assistant/configuration",frontMatter:{title:"Assistant configuration",description:"When you deploy MOSTLY AI, the Assistant is included as part of the deployment. For the Assistant to work, you need to configure access to the Language Learning Model (LLM) and DataLLM."}},{kind:"MdxPage",name:"manage-threads",route:"/assistant/manage-threads",frontMatter:{title:"Manage threads",description:"Find out how to manage each new thread you have with the MOSTLY AI Assistant. You can share, delete, and rename each thread."}},{kind:"MdxPage",name:"prompt-examples",route:"/assistant/prompt-examples",frontMatter:{title:"Examples",description:"The MOSTLY AI Assistant provides LLM capabilities that literally empower users to talk to their data and get answers to a broad set of questions about data. Check out several examples to get a feeling of how easy it is and what users can accomplish with the Assistant."}},{kind:"MdxPage",name:"python-session-status",route:"/assistant/python-session-status",frontMatter:{title:"Python session status",description:"Each new Assistant thread runs a Python environment where MOSTLY AI runs the Python code to navigate you and show you data insights. The Python session expires after a period of inactivity. Learn how to check and manage your Python session status."}}]},{kind:"MdxPage",name:"assistant",route:"/assistant",frontMatter:{title:"Assistant",description:"The MOSTLY AI Assistant provides by far the easiest interaction you can have about your data. With natural language, you can get rich data insights instantly and use all MOSTLY AI and DataLLM features from a single place without writing a single line of code."}},{kind:"MdxPage",name:"best-practices",route:"/best-practices",frontMatter:{title:"Best practices",description:"MOSTLY AI has years of experience managing synthetic data projects. To generate synthetic data of the highest quality, get to know our best practices."}},{kind:"Folder",name:"concepts",route:"/concepts",children:[{kind:"Meta",data:{"what-is-synthetic-data":"What is synthetic data?","privacy-protection":"Privacy protection"}},{kind:"MdxPage",name:"privacy-protection",route:"/concepts/privacy-protection",frontMatter:{title:"Privacy-protection mechanisms",description:"With synthetic data, you can unlock the utility of your original data and at the same time protect the privacy of your subjects. While specific cases in the original data might increase the risk of re-identification, MOSTLY AI employs a number of privacy-protection mechanisms to avoid such risks."}},{kind:"MdxPage",name:"what-is-synthetic-data",route:"/concepts/what-is-synthetic-data",frontMatter:{title:"What is synthetic data?",description:"Synthetic data is an artificial version of your original data that protects the privacy of your data subjects and at the same time maintains the utility of your original data by preserving its statistical properties. AI-generated synthetic data can help to unlock the analysis and sharing of otherwise protected datasets that can take months or years to become "}}]},{kind:"Folder",name:"connectors",route:"/connectors",children:[{kind:"Meta",data:{create:"Create a connector",use:"Use connectors",manage:"Manage connectors"}},{kind:"Folder",name:"create",route:"/connectors/create",children:[{kind:"Meta",data:{mysql:"MySQL",postgresql:"PostgreSQL","sql-server":"SQL Server",oracle:"Oracle",mariadb:"MariaDB","apache-hive":"Apache Hive",bigquery:"BigQuery",snowflake:"Snowflake",databricks:"Databricks",s3:"S3 storage",azure:"Microsoft Azure blob storage","google-cloud":"Google Cloud object storage"}},{kind:"MdxPage",name:"apache-hive",route:"/connectors/create/apache-hive",frontMatter:{title:"Use Apache Hive for synthetic data",description:"With MOSTLY AI, you can use Apache Hive as a source for original data as well as a destination to deliver synthetic data. To do so, you need to create Apache Hive connectors."}},{kind:"MdxPage",name:"azure",route:"/connectors/create/azure",frontMatter:{title:"Use Azure blob storage for synthetic data",description:"To use Azure blob storage as a data source or destination for your synthetic data, you need to create an Azure connector in MOSTLY AI."}},{kind:"MdxPage",name:"bigquery",route:"/connectors/create/bigquery",frontMatter:{title:"Use BigQuery for synthetic data",description:"With MOSTLY AI, you can connect to BigQuery and use it as a data source or destination for your synthetic data."}},{kind:"MdxPage",name:"databricks",route:"/connectors/create/databricks",frontMatter:{title:"Use Databricks for synthetic data",description:"With MOSTLY AI, you can connect to a Databricks SQL Warehouse and use it as a data source or destination for your synthetic data."}},{kind:"MdxPage",name:"google-cloud",route:"/connectors/create/google-cloud",frontMatter:{title:"Connect to Google Cloud object storage",description:"To use datasets that you keep in Google Cloud storage buckets as a data source for synthetic data, you need to create a Google Cloud connector in MOSTLY AI."}},{kind:"MdxPage",name:"mariadb",route:"/connectors/create/mariadb",frontMatter:{title:"Use a MariaDB database for synthetic data",description:"MOSTLY AI can use a MariaDB database as a source for original data as well as a destination to deliver synthetic data. To do so, you need to create MariaDB connectors."}},{kind:"MdxPage",name:"mysql",route:"/connectors/create/mysql",frontMatter:{title:"Use a MySQL database for synthetic data",description:"MOSTLY AI can use a MySQL database as a source for original data as well as a destination to deliver synthetic data. To do so, you need to create MySQL connectors."}},{kind:"MdxPage",name:"oracle",route:"/connectors/create/oracle",frontMatter:{title:"Use an Oracle database for synthetic data",description:"MOSTLY AI can use a Oracle database as a source for original data as well as a destination to deliver synthetic data. To do so, you need to create Oracle connectors."}},{kind:"MdxPage",name:"postgresql",route:"/connectors/create/postgresql",frontMatter:{title:"Use a PostgreSQL database for synthetic data",description:"MOSTLY AI can use a PostgreSQL database as a source for original data as well as a destination to deliver synthetic data. To do so, you need to create PostgreSQL connectors."}},{kind:"MdxPage",name:"s3",route:"/connectors/create/s3",frontMatter:{title:"Use S3 storage for synthetic data",description:"If you keep datasets in S3 storage (Amazon AWS S3 or any S3-compatible storage service), you can synthesize them via an S3 connector in MOSTLY AI. If you want to store the generated synthetic data in a separate S3 bucket, you need to create a second destination S3 connector that points to that bucket."}},{kind:"MdxPage",name:"snowflake",route:"/connectors/create/snowflake",frontMatter:{title:"Use Snowflake for synthetic data",description:"With MOSTLY AI, you can connect to Snowflake and use it as a data source or destination for your synthetic data."}},{kind:"MdxPage",name:"sql-server",route:"/connectors/create/sql-server",frontMatter:{title:"Use a SQL Server database for synthetic data",description:"MOSTLY AI can use a Microsoft SQL Server database as a source for original data as well as a destination to deliver synthetic data. To do so, you need to create SQL Server connectors."}}]},{kind:"MdxPage",name:"create",route:"/connectors/create",frontMatter:{title:"Create a connector",description:"MOSTLY AI integrates with popular databases. You can create connectors to all supported databases and use them as data sources or destinations for synthetic data."}},{kind:"MdxPage",name:"manage",route:"/connectors/manage",frontMatter:{title:"Manage connectors",description:"In MOSTLY AI, you can manage the connections to all of your synthetic data sources and destinations from the Connectors page."}},{kind:"Folder",name:"use",route:"/connectors/use",children:[{kind:"Meta",data:{"data-source":"Use as a data source","data-destination":"Use as a data destination","local-db":"Use a local database"}},{kind:"MdxPage",name:"data-destination",route:"/connectors/use/data-destination",frontMatter:{title:"Use a connector as a data destination",description:"When you configure a data synthesis job, you can specify a destination where you want the job to deliver the generated synthetic data."}},{kind:"MdxPage",name:"data-source",route:"/connectors/use/data-source",frontMatter:{title:"Use a connector as a data source",description:"When you configure a new generator in MOSTLY AI, you can specify any of your source connectors as a data source for original data."}},{kind:"MdxPage",name:"local-db",route:"/connectors/use/local-db",frontMatter:{title:"Local database as a source or destination",description:"If you want to use a database running in your local environment as a data source or destination for synthetic data, you can use `ngrok` to expose the local database and then add the ngrok forwarding address in a new MOSTLY AI connector."}}]}]},{kind:"MdxPage",name:"connectors",route:"/connectors",frontMatter:{title:"Connectors",description:"MOSTLY AI integrates with popular databases and cloud storage providers so that you can utilize them as data sources for training generators or data destinations for synthetic data."}},{kind:"MdxPage",name:"datasets",route:"/datasets",frontMatter:{title:"Datasets",description:"The datasets below are popular for machine learning use cases. You can use them in MOSTLY AI to generate synthetic data versions and compare them to the original versions."}},{kind:"Folder",name:"generators",route:"/generators",children:[{kind:"Meta",data:{"tabular-synthetic-data":"Tabular synthetic data","prepare-data":"Prepare your data",train:"Train a new generator",configure:"Configure generators","evaluate-quality":"Evaluate generator quality",share:"Share a generator","export-import":"Export and import generators",manage:"Manage generators","live-probing":"Live-probe generators","rare-values":"_RARE_ values"}},{kind:"Folder",name:"configure",route:"/generators/configure",children:[{kind:"Meta",data:{"add-data":"Add data","set-table-relationships":"Set table relationships","set-encoding-types":"Set encoding types","improve-model-accuracy":"Improve model accuracy","speed-up-training":"Speed up training","configure-time-series":"Configure time-series models","fine-tune-privacy-mechanisms":"Fine-tune privacy mechanisms","enable-flexible-generation":"Enable flexible generation"}},{kind:"MdxPage",name:"add-data",route:"/generators/configure/add-data",frontMatter:{title:"Add data to a generator",description:"Before you start training a generator, you can add tabular data to it. You can add data from multiple sources: file upload (CSV, Parquet), databases, cloud storage buckets."}},{kind:"MdxPage",name:"configure-time-series",route:"/generators/configure/configure-time-series",frontMatter:{title:"Configure AI models for time-series and events data",description:"MOSTLY AI supports specific AI models for training on time-series. You can configure how these models train on your time-series data."}},{kind:"MdxPage",name:"enable-flexible-generation",route:"/generators/configure/enable-flexible-generation",frontMatter:{title:"Flexible generation",description:"Generators in MOSTLY AI have flexible generation on by default. You can use such generators to rebalance your synthetic datasets and set the generation temperature to be more conservative or creative."}},{kind:"MdxPage",name:"fine-tune-privacy-mechanisms",route:"/generators/configure/fine-tune-privacy-mechanisms",frontMatter:{title:"Fine-tune privacy mechanisms",description:"When enabled, the **Value protection** setting replaces rare categories or removes numeric and date-time outliers from your original dataset before generator training."}},{kind:"MdxPage",name:"improve-model-accuracy",route:"/generators/configure/improve-model-accuracy",frontMatter:{title:"Improve model accuracy",description:"You can adjust the model settings in MOSTLY AI to improve the accuracy of your tabular AI models."}},{kind:"MdxPage",name:"set-encoding-types",route:"/generators/configure/set-encoding-types",frontMatter:{title:"Set encoding types",description:"MOSTLY AI supports well-known data types and implements logic on how to encode the data in each column for AI model training. Setting the correct encoding type ensures that the generated synthetic data accurately represents your original data."}},{kind:"Folder",name:"set-table-relationships",route:"/generators/configure/set-table-relationships",children:[{kind:"Meta",data:{"two-table":"Two-table scenario","multi-table":"Multi-table scenario","manage-pks":"Manage primary keys","relationship-diagram":"Relationship diagram"}},{kind:"MdxPage",name:"manage-pks",route:"/generators/configure/set-table-relationships/manage-pks",frontMatter:{title:"Manage primary keys",description:"MOSTLY AI provides easy ways to set and remove primary keys for each table in a generator."}},{kind:"MdxPage",name:"multi-table",route:"/generators/configure/set-table-relationships/multi-table",frontMatter:{title:"Multi-table relationships",description:"In a multi-table scenario with MOSTLY AI, where you add 3 or more tables to a generator, you can have multiple tables with multiple relationships to other tables."}},{kind:"MdxPage",name:"relationship-diagram",route:"/generators/configure/set-table-relationships/relationship-diagram"},{kind:"MdxPage",name:"two-table",route:"/generators/configure/set-table-relationships/two-table",frontMatter:{title:"Two-table relationship",description:"In MOSTLY AI, a two-table scenario includes a subject table and a linked table that have a context relationship."}}]},{kind:"MdxPage",name:"set-table-relationships",route:"/generators/configure/set-table-relationships",frontMatter:{title:"Set table relationships",description:"With MOSTLY AI, you can train a generator on multi-table datasets with existing relationships between tables by setting foreign keys."}},{kind:"MdxPage",name:"speed-up-training",route:"/generators/configure/speed-up-training",frontMatter:{title:"Speed up training",description:"You can adjust the model settings in MOSTLY AI to get a quicker completion of AI model training for your tabular data."}}]},{kind:"MdxPage",name:"evaluate-quality",route:"/generators/evaluate-quality",frontMatter:{title:"Evaluate generator quality",description:"MOSTLY AI calculates generator quality metrics at the end of the training of each generator. The metrics are available on the page of a generator after training completes and detailed charts and metrics are also available in the model report for each table."}},{kind:"MdxPage",name:"export-import",route:"/generators/export-import",frontMatter:{title:"Export and import generators",description:"You can export already trained generators of MOSTLY AI from one environment and import them in another to avoid training a new generator with the same data from scratch."}},{kind:"MdxPage",name:"live-probing",route:"/generators/live-probing",frontMatter:{title:"Live probing",description:"You can use the MOSTLY AI Python client to probe a trained generator for a number of synthetic samples in real-time without the need to generate a synthetic dataset."}},{kind:"MdxPage",name:"manage",route:"/generators/manage",frontMatter:{title:"Manage generators",description:"Train generators with MOSTLY AI and you can share them with peers and colleagues, generate synthetic datasets, transfer ownership, manage access, and more."}},{kind:"Folder",name:"prepare-data",route:"/generators/prepare-data",children:[{kind:"Meta",data:{"csv-requirements":"CSV requirements","subject-linked-requirements":"Privacy best practices for original data"}},{kind:"MdxPage",name:"csv-requirements",route:"/generators/prepare-data/csv-requirements",frontMatter:{title:"CSV requirements",description:"To successfully synthesize your dataset, the content must be encoded in UTF-8, have commas (,) or semicolons (;) as comma separators, and contain the column names as a header row."}},{kind:"MdxPage",name:"subject-linked-requirements",route:"/generators/prepare-data/subject-linked-requirements",frontMatter:{title:"Subject and linked tables requirements",description:"To generate synthetic data from CSV or Parquet files, get acquainted with the guidelines to learn how to best prepare your subject and table files."}}]},{kind:"MdxPage",name:"prepare-data",route:"/generators/prepare-data",frontMatter:{title:"Prepare your data",description:"Before you train a generator with MOSTLY AI, you can review some of the considerations and requirements that can help you avoid unexpected errors, maintain the privacy of the subjects (people, companies, or any other entities), and ensure higher accuracy of a trained generator."}},{kind:"MdxPage",name:"rare-values",route:"/generators/rare-values",frontMatter:{title:"_RARE_ values in MOSTLY AI",description:"In MOSTLY AI, the Rare category protection mechanism with the Constant method is on by default for all categorical columns. Depending on your data, MOSTLY AI masks rare categories with the _RARE_ token."}},{kind:"MdxPage",name:"share",route:"/generators/share",frontMatter:{title:"Share a generator",description:"Once trained, a generator is a shareable asset that you and others can use to create synthetic data. You can share your generators with anyone inside and outside your organization."}},{kind:"MdxPage",name:"tabular-synthetic-data",route:"/generators/tabular-synthetic-data",frontMatter:{title:"Tabular synthetic data",description:"MOSTLY AI generates tabular synthetic data that provides privacy protection for data subjects while maintaining referential integrity and retaining the correlations between columns and tables (depending on the table relationship schema)."}},{kind:"MdxPage",name:"train",route:"/generators/train",frontMatter:{title:"Train a new generator",description:"With generators, MOSTLY AI makes it easy to use Generative AI to train your own AI models on single- and multi-table datasets. You can then use a trained generator to generate endless amounts of high-quality and privacy-protected synthetic data."}}]},{kind:"MdxPage",name:"generators",route:"/generators",frontMatter:{title:"Generators",description:"With MOSTLY AI, your journey to synthetic data starts with the training of a generator. You can train generators on tabular data that also includes columns with text data."}},{kind:"MdxPage",name:"index",route:"/",frontMatter:{title:"MOSTLY AI Documentation",description:"MOSTLY AI official documentation helps you to get started, learn how to train Generative AI models with tabular data, and how to generate multi-table synthetic data that is better than real data.",toc:!1,breadcrumb:!1}},{kind:"Folder",name:"install",route:"/install",children:[{kind:"Meta",data:{architecture:"Architecture",requirements:"Requirements","hardware-profiling":"Hardware profiling",config:"Configure",deploy:"Deploy",troubleshoot:"Troubleshoot",compliance:"Compliance"}},{kind:"MdxPage",name:"architecture",route:"/install/architecture",frontMatter:{title:"Architecture",description:"MOSTLY AI runs as a set of containerized applications and services that you can deploy in a Kubernetes cluster and maintain a fault-tolerant and highly available application."}},{kind:"MdxPage",name:"compliance",route:"/install/compliance",frontMatter:{title:"OCI compliance and security of MOSTLY AI container images",description:""}},{kind:"Folder",name:"config",route:"/install/config",children:[{kind:"Meta",data:{"compute-resources":"Compute resources","external-postgresql-db":"External PostgreSQL database","internal-image-repo":"Internal image repository","domain-ssl-cert":"Domain SSL certificate"}},{kind:"MdxPage",name:"compute-resources",route:"/install/config/compute-resources",frontMatter:{title:"Configure compute resources",description:"The MOSTLY AI Platform runs AI tasks using Kubernetes jobs. These jobs may require a lot of resources, especially memory and CPU. It is very important that these jobs are assigned on a node that has enough resources."}},{kind:"MdxPage",name:"domain-ssl-cert",route:"/install/config/domain-ssl-cert",frontMatter:{title:"Configure your domain SSL certificate",description:"To deploy MOSTLY AI, you need to configure SSL certificates for your fully-qualified domain name (FQDN) to enable secure access to the MOSTLY AI web app."}},{kind:"MdxPage",name:"external-postgresql-db",route:"/install/config/external-postgresql-db",frontMatter:{title:"Configure an external PostgreSQL database",description:"The MOSTLY AI Helm charts include a PostgreSQL container that the application uses to store data about generators, synthetic datasets, and connectors. If necessary, you can configure an external PostgreSQL database that your company hosts internally."}},{kind:"MdxPage",name:"internal-image-repo",route:"/install/config/internal-image-repo",frontMatter:{title:"Configure an internal image repository",description:"Due to internal IT and security policies, you might need to host the MOSTLY AI container images in an internal repository."}}]},{kind:"MdxPage",name:"config",route:"/install/config",frontMatter:{title:"Configure MOSTLY AI deployments",description:"If you Kubernetes environment has specific requirements, you may need to configure the deployment of MOSTLY AI in a custom way."}},{kind:"Folder",name:"deploy",route:"/install/deploy",children:[{kind:"Meta",data:{checklist:"Deployment checklist","helm-chart":"MOSTLY AI Helm chart","aws-eks":"Deploy to AWS EKS",openshift:"Deploy to OpenShift",gke:"Deploy to Google GKE","azure-aks":"Deploy to Azure AKS",minikube:"Deploy to minikube"}},{kind:"MdxPage",name:"aws-eks",route:"/install/deploy/aws-eks",frontMatter:{title:"Deploy MOSTLY AI to an AWS EKS cluster",description:"You can deploy MOSTLY AI to an Amazon AWS Elastic Kubernetes Services (EKS) cluster. This page covers the list of prerequisites you need and the steps for a successful installation."}},{kind:"MdxPage",name:"azure-aks",route:"/install/deploy/azure-aks",frontMatter:{title:"Deploy MOSTLY AI to an Azure AKS cluster",description:"You can deploy MOSTLY AI to an Azure AKS cluster. This page covers the list of prerequisites you need to have in place and the pre-deployment and deployment steps for a successful installation."}},{kind:"MdxPage",name:"checklist",route:"/install/deploy/checklist",frontMatter:{title:"Deployment checklist",description:"The checklist below provides a list of prerequisites to ensure a successful installation process. Before you contact MOSTLY AI to complete the installation or troubleshoot installation issues, make sure to complete the checklist."}},{kind:"MdxPage",name:"gke",route:"/install/deploy/gke",frontMatter:{title:"Deploy MOSTLY AI to a GKE cluster",description:"You can deploy MOSTLY AI to a Google Kubernetes Engine (GKE) cluster. This page covers the list of prerequisites you need to have in place and the pre-deployment and deployments steps for a successful installation."}},{kind:"MdxPage",name:"helm-chart",route:"/install/deploy/helm-chart",frontMatter:{title:"MOSTLY AI Helm chart",description:"To deploy MOSTLY AI in a Kubernetes cluster that your organization owns and configures, MOSTLY AI provides a Helm chart."}},{kind:"MdxPage",name:"minikube",route:"/install/deploy/minikube",frontMatter:{title:"Deploy MOSTLY AI to a minikube cluster",description:"One of the options to install MOSTLY AI without the need of a multi-node Kubernetes cluster is to do so on a 1local Kubernetes cluster with `minikube`. More specifically, this page covers the steps to install MOSTLY AI on a Ubuntu Server LTS."}},{kind:"MdxPage",name:"openshift",route:"/install/deploy/openshift",frontMatter:{title:"Deploy MOSTLY AI to an OpenShift cluster",description:"You can install MOSTLY AI in an OpenShift cluster. This page covers the list of prerequisites you need and the steps for a successful installation."}}]},{kind:"MdxPage",name:"deploy",route:"/install/deploy",frontMatter:{title:"Deploy MOSTLY AI",description:"A system administrator of a Kubernetes cluster can best navigate the deployment of MOSTLY AI. The deployment requires the management and mapping of permissions for the storage, nodes, ingress, and secrets to internal repositories, as well as the definition of limits and requests for the pods."}},{kind:"MdxPage",name:"hardware-profiling",route:"/install/hardware-profiling",frontMatter:{title:"Hardware profiling",description:"Depending on the datasets you need to synthesize, your hardware requirements for MOSTLY AI can differ. Review the MOSTLY AI hardware profiling benchmarks to get insights on what hardware can cover your synthetic data needs best."}},{kind:"MdxPage",name:"requirements",route:"/install/requirements",frontMatter:{title:"Requirements",description:"Before you deploy MOSTLY AI in your organization's cloud, make sure that your environment meets the requirements."}},{kind:"Folder",name:"troubleshoot",route:"/install/troubleshoot",children:[{kind:"Meta",data:{common:"Common issues","aws-eks":"AWS EKS",openshift:"OpenShift","azure-aks":"Azure AKS",minikube:"minikube"}},{kind:"MdxPage",name:"aws-eks",route:"/install/troubleshoot/aws-eks",frontMatter:{title:"Troubleshoot MOSTLY AI deployments to Amazon AWS EKS",description:"Learn how you can troubleshoot various issues when you deploy MOSTLY AI to an Amazon AWS EKS cluster."}},{kind:"MdxPage",name:"azure-aks",route:"/install/troubleshoot/azure-aks",frontMatter:{title:"Troubleshoot MOSTLY AI deployments to Azure AKS",description:"Learn how you can troubleshoot various issues when you deploy MOSTLY AI to an Azure AKS cluster."}},{kind:"MdxPage",name:"common",route:"/install/troubleshoot/common",frontMatter:{title:"Troubleshoot deployment issues",description:"Learn how you can troubleshoot various issues when you deploy MOSTLY AI to an OpenShift cluster."}},{kind:"MdxPage",name:"minikube",route:"/install/troubleshoot/minikube",frontMatter:{title:"Troubleshoot MOSTLY AI deployments to minikube",description:"Learn how you can troubleshoot various issues when you deploy MOSTLY AI to a minikube cluster."}},{kind:"MdxPage",name:"openshift",route:"/install/troubleshoot/openshift",frontMatter:{title:"Troubleshoot MOSTLY AI deployments to OpenShift",description:"Learn how you can troubleshoot various issues when you deploy MOSTLY AI to an OpenShift cluster."}}]},{kind:"MdxPage",name:"troubleshoot",route:"/install/troubleshoot",frontMatter:{title:"Troubleshoot MOSTLY AI deployments",description:"Due to different Kubernetes cluster configurations, you might need to take into account our troubleshooting information in case you run into any issue deploying MOSTLY AI."}}]},{kind:"MdxPage",name:"python-client",route:"/python-client",frontMatter:{title:"MOSTLY AI Python client",description:"A Python wrapper for the MOSTLY AI platform."}},{kind:"Folder",name:"quick-start",route:"/quick-start",children:[{kind:"Meta",data:{"model-creators":"For Model creators","data-consumers":"For Data consumers","fine-tuning-llms":"For Fine-tuning LLMs"}},{kind:"MdxPage",name:"data-consumers",route:"/quick-start/data-consumers",frontMatter:{title:"Quick start for Data consumers",description:"As a Data consumer, you can use a trained generator in MOSTLY AI to generate new synthetic data. Use the new data for analysis, downstream machine learning (ML) tasks, or share it with others."}},{kind:"MdxPage",name:"fine-tuning-llms",route:"/quick-start/fine-tuning-llms",frontMatter:{title:"Quick start for fine-tuning LLMs",description:"Learn how easy it is to fine-tune an LLM with MOSTLY AI to train and generate unstructured synthetic text data."}},{kind:"MdxPage",name:"model-creators",route:"/quick-start/model-creators",frontMatter:{title:"Quick start for model creators",description:"As a model creator, you can train brand new generators and AI models on tabular data. You can then share with peers and data consumers in your teams and organization and they can share brand new privacy-safe synthetic data."}}]},{kind:"MdxPage",name:"quick-start",route:"/quick-start",frontMatter:{title:"Quick start",description:"Take your first steps in training generators on your tabular data and generating synthetic data."}},{kind:"MdxPage",name:"release-lifecycle",route:"/release-lifecycle",frontMatter:{title:"Release lifecycle and support",description:"Our guide to software release management and support policies, including LTS and regular releases, end-of-support details, security updates, and user best practices."}},{kind:"MdxPage",name:"search",route:"/search",frontMatter:{title:"Search",description:"The Search bar offers direct access to all of your generators, synthetic datasets, and connectors in MOSTLY AI."}},{kind:"MdxPage",name:"support",route:"/support",frontMatter:{title:"Support",description:"The team at MOSTLY AI is open to your feedback. If you have questions or run into any issues, send us an email."}},{kind:"Folder",name:"synthetic-datasets",route:"/synthetic-datasets",children:[{kind:"Meta",data:{generate:"Generate a new synthetic dataset","set-size-temp":"Set sample size and temperature","rebalance-columns":"Rebalance columns","data-imputation":"Data imputation","evaluate-quality":"Evaluate synthetic data quality",deliver:"Deliver synthetic data","seeded-generation":"Seeded generation",fairness:"Fair synthetic data",manage:"Manage synthetic datasets"}},{kind:"MdxPage",name:"data-imputation",route:"/synthetic-datasets/data-imputation",frontMatter:{title:"Data imputation",description:"MOSTLY AI can fill any null or empty values in your datasets with highly-correlated values based on the learnings of your generators."}},{kind:"MdxPage",name:"deliver",route:"/synthetic-datasets/deliver",frontMatter:{title:"Deliver synthetic data to databases or cloud buckets",description:"When you configure the generation of a synthetic dataset, you can select to deliver the generated synthetic data to a database or cloud storage bucket."}},{kind:"MdxPage",name:"evaluate-quality",route:"/synthetic-datasets/evaluate-quality",frontMatter:{title:"Evaluate synthetic dataset quality",description:"MOSTLY AI calculates quality metrics for each synthetic dataset. The metrics are available on the page of a synthetic dataset after generation completes and detailed charts and metrics are also available in the data for each model."}},{kind:"MdxPage",name:"fairness",route:"/synthetic-datasets/fairness",frontMatter:{title:"Fair synthetic data",description:"MOSTLY AI supports the generation of fair synthetic data**. You can generate statistical parity synthetic data where you can target a specific table column for fairness (for example, income) and easily remove biases based on other sensitive columns in your datasets, such as race, sex, age, or any other attribute that you define as sensitive."}},{kind:"MdxPage",name:"generate",route:"/synthetic-datasets/generate",frontMatter:{title:"Generate a synthetic dataset",description:"When you configure the generation of a synthetic dataset, you can select to deliver the generated synthetic data to a database or cloud storage bucket."}},{kind:"MdxPage",name:"manage",route:"/synthetic-datasets/manage",frontMatter:{title:"Manage synthetic datasets",description:"Generate synthetic datasets with MOSTLY AI and you can then share them with peers and colleagues, clone them, manage access, and more."}},{kind:"MdxPage",name:"rebalance-columns",route:"/synthetic-datasets/rebalance-columns",frontMatter:{title:"Rebalance columns",description:"With MOSTLY AI, you can generate synthetic datasets that rebalance column distributions in your original data and observe impacts of fairer distributions and bias mitigation on the rest of the features from your original data."}},{kind:"MdxPage",name:"seeded-generation",route:"/synthetic-datasets/seeded-generation",frontMatter:{title:"Seeded generation",description:"With seeded generation, you can create a new synthetic dataset that is conditioned on a seed dataset you provide. This is also known as conditional generation."}},{kind:"MdxPage",name:"set-size-temp",route:"/synthetic-datasets/set-size-temp",frontMatter:{title:"Set sample size and temperature",description:"With MOSTLY AI, you can define the sample size as the amount of data you want to generate and specify the temperature and Top P generation parameters to configure how creative or conservative the generated synthetic samples should be."}}]},{kind:"MdxPage",name:"synthetic-datasets",route:"/synthetic-datasets",frontMatter:{title:"Synthetic datasets",description:"To generate synthetic data in MOSTLY AI, you start a new synthetic dataset. You can view all finished, canceled, failed, and in-progress synthetic datasets on the Synthetic datasets page."}},{kind:"MdxPage",name:"tutorials",route:"/tutorials",frontMatter:{title:"Tutorials",description:"MOSTLY AI provides a number of synthetic data tutorials that you can explore and run in Jupyter Lab or run directly in Google Colab."}},{kind:"MdxPage",name:"usage",route:"/usage",frontMatter:{title:"Usage and credits",description:"Daily, MOSTLY AI provides 5 credits for the free generation of synthetic data points. 1 credits equals 1 million synthetic data points."}},{kind:"MdxPage",name:"whats-new",route:"/whats-new",frontMatter:{title:"What is new in MOSTLY AI",description:"Learn about the latest releases and features of the MOSTLY AI Platform.",searchable:!1}}],flexsearch:{codeblocks:!0},title:"OCI compliance and security of MOSTLY AI container images",headings:d},pageNextRoute:"/install/compliance",nextraLayout:r.ZP,themeConfig:i.Z};t.default=(0,o.j)(c)},3215:function(e,t,a){"use strict";a.d(t,{Z:function(){return c}});var n=a(5893),o=a(5379),r=a.n(o),i=a(1163),s=a(7913),d=a(1661);a(6491);var c={logo:(0,n.jsxs)("div",{className:"jsx-8aeb97ea4d6b8a5f",children:[(0,n.jsx)("span",{className:"jsx-8aeb97ea4d6b8a5f mostlyai-logo",children:(0,n.jsx)("a",{href:"https://mostly.ai",className:"jsx-8aeb97ea4d6b8a5f",children:(0,n.jsx)("img",{src:"/docs/logo/MOSTLY.AI_logo_light_gradient.svg",alt:"MOSTLY AI Logo - gradient",width:175,height:30,className:"jsx-8aeb97ea4d6b8a5f"})})}),(0,n.jsx)("span",{className:"jsx-8aeb97ea4d6b8a5f logo-separator",children:"/"}),(0,n.jsx)("span",{className:"jsx-8aeb97ea4d6b8a5f mostlyai-docs-logo",children:(0,n.jsx)("a",{href:"/docs",className:"jsx-8aeb97ea4d6b8a5f",children:(0,n.jsx)("img",{src:"/docs/logo/docs-logo.svg",alt:"MOSTLY AI Docs logo",width:60,height:30,className:"jsx-8aeb97ea4d6b8a5f"})})}),(0,n.jsx)(r(),{id:"8aeb97ea4d6b8a5f",children:'.mostlyai-logo.jsx-8aeb97ea4d6b8a5f a.jsx-8aeb97ea4d6b8a5f img.jsx-8aeb97ea4d6b8a5f{display:inline-block;margin:0;padding:0;margin-right:10px}.mostlyai-docs-logo.jsx-8aeb97ea4d6b8a5f a.jsx-8aeb97ea4d6b8a5f img.jsx-8aeb97ea4d6b8a5f{display:inline-block;margin:0;padding:0;margin-left:10px}.mostlyai-logo.jsx-8aeb97ea4d6b8a5f a.jsx-8aeb97ea4d6b8a5f:hover{background:none}.mostlyai-docs-logo.jsx-8aeb97ea4d6b8a5f a.jsx-8aeb97ea4d6b8a5f:hover{background:none}.logo-separator.jsx-8aeb97ea4d6b8a5f{font-family:"Objectivity-Black";color:lightgrey}'})]}),logoLink:!1,feedback:{content:null},editLink:{text:null},sidebar:{defaultMenuCollapseLevel:1},toc:{title:"ON THIS PAGE"},navigation:{prev:!1,next:!1},useNextSeoProps:()=>({titleTemplate:"%s – MOSTLY AI Docs"}),head:()=>{let{asPath:e}=(0,i.useRouter)(),{frontMatter:t}=(0,s.ZR)(),a="https://mostly.ai/docs"+"".concat(e);return(0,n.jsxs)(n.Fragment,{children:[(0,n.jsx)("link",{rel:"icon",type:"image/x-icon",href:"/docs/favicon/MOSTLY-AI-Docs-favicon.svg"}),(0,n.jsx)("link",{rel:"canonical",href:a}),(0,n.jsx)("meta",{property:"og:url",content:a}),(0,n.jsx)("meta",{property:"og:title",content:t.title||"MOSTLY AI Documentation"}),(0,n.jsx)("meta",{property:"og:description",content:t.description||"MOSTLY AI Documentation "}),(0,n.jsx)("meta",{property:"og:image",content:"/docs/logo/mostly-ai-docs-social.png"}),(0,n.jsx)("meta",{name:"twitter:title",content:t.title||"MOSTLY AI Documentation"}),(0,n.jsx)("meta",{name:"twitter:description",content:t.description||"MOSTLY AI Documentation"}),(0,n.jsx)("meta",{name:"twitter:image",content:"/docs/logo/mostly-ai-docs-social.png"})]})},footer:{component:null},nextThemes:{defaultTheme:"light"},gitTimestamp:null,darkMode:!1,search:{component:function(){return(0,n.jsx)(d.F,{appId:"UPODPFM590",indexName:"mostly",apiKey:"777482884c302e9439e4a4048ccd631b"})}}}},5789:function(){}},function(e){e.O(0,[7560,9774,2888,179],function(){return e(e.s=6684)}),_N_E=e.O()}]);