# Oumi

## Docs

- [WELCOME TO OUMI](https://docs.oumi.ai/guides/overview.md): Custom AI in hours not months
- [EXAMPLES](https://docs.oumi.ai/guides/examples.md): Explore a curated set of prompts that demonstrate Oumi’s capabilities
- [OUMI QUICKSTART](https://docs.oumi.ai/guides/quickstart.md): Learn how to quickly create your first custom AI model
- [OVERVIEW](https://docs.oumi.ai/guides/intro.md): Introducing the Oumi workflow
- [OUMI AGENT](https://docs.oumi.ai/guides/intro/agent.md): Build, fine-tune, and evaluate custom AI models with guided assistance
- [RECIPES](https://docs.oumi.ai/guides/intro/recipes.md): Reusable configurations for end-to-end model workflows
- [JOBS](https://docs.oumi.ai/guides/jobs.md): Track long-running jobs and their operations
- [NEXT STEPS](https://docs.oumi.ai/guides/intro/next-steps.md): Where to go from here
- [OUMI DATASETS](https://docs.oumi.ai/guides/datasets.md): Upload, generate, and validate datasets in Oumi
- [CREATE DATASETS](https://docs.oumi.ai/guides/datasets/create.md): Bringing your datasets into the Oumi Platform
- [IMPORT OBSERVABILITY LOGS](https://docs.oumi.ai/guides/datasets/import-observability-logs.md): Turn Langfuse or LangSmith traces into a Conversation dataset you can train, evaluate, and synthesize with
- [ANALYZE DATASETS](https://docs.oumi.ai/guides/datasets/analyze.md): Measure, track, and improve the quality of your datasets in Oumi
- [ADDING CONTEXT FILES](https://docs.oumi.ai/guides/datasets/add-files.md): Uploading context files to Oumi
- [DATA EXPLORER](https://docs.oumi.ai/guides/datasets/exploring.md): Exploring & validating your datasets
- [NEXT STEPS](https://docs.oumi.ai/guides/datasets/next-steps.md): What to do after creating training data
- [SYNTHETIC DATA GENERATION](https://docs.oumi.ai/guides/synthesis.md): Generate high quality synthetic datasets at scale
- [HOW IT WORKS](https://docs.oumi.ai/guides/synthesis/how-it-works.md): Creating synthetic datasets with Oumi
- [DATA SYNTHESIS RECIPES](https://docs.oumi.ai/guides/synthesis/recipes.md): Reusable recipes for generating structured synthetic datasets.
- [NEXT STEPS](https://docs.oumi.ai/guides/synthesis/next-steps.md): Getting the most out of Oumi's data synthesis capabilities
- [EVALUATORS](https://docs.oumi.ai/guides/evaluators.md): Define and manage model evaluators
- [DEFINING EVALUATORS](https://docs.oumi.ai/guides/evaluators/defining.md): Establish criteria for measuring model performance
- [SAVED EVALUATORS](https://docs.oumi.ai/guides/evaluators/recipes.md): Save and reuse evaluator configurations
- [NEXT STEPS](https://docs.oumi.ai/guides/evaluators/next-steps.md): Apply and iterate with evaluators
- [EVALUATE MODEL](https://docs.oumi.ai/guides/evaluations.md): Determine how well your model performs and identify failure modes
- [RUNNING EVALUATIONS](https://docs.oumi.ai/guides/evaluations/running.md): Using the Oumi Builder to perform model evaluation
- [UNDERSTANDING RESULTS](https://docs.oumi.ai/guides/evaluations/results.md): How to read the results from your evaluation run
- [DIAGNOSING FAILURE MODES](https://docs.oumi.ai/guides/evaluations/failure-modes.md): Identify patterns in your model results
- [EVALUATION RECIPES](https://docs.oumi.ai/guides/evaluations/recipes.md): Making evaluations consistent and repeatable
- [NEXT STEPS](https://docs.oumi.ai/guides/evaluations/next-steps.md): What to do after model evaluation
- [TRAIN MODEL](https://docs.oumi.ai/guides/training.md): Fine-tune your model to improve its task-specific performance.
- [MODEL SELECTION](https://docs.oumi.ai/guides/training/model-selection.md): Choosing the right base model for performance and efficiency
- [RUNNING TRAINING](https://docs.oumi.ai/guides/training/launch-jobs.md): Easily launch fine-tuning jobs
- [ON-POLICY DISTILLATION](https://docs.oumi.ai/guides/training/on-policy-distillation.md): Train student models via knowledge distillation
- [TRAINING RECIPES](https://docs.oumi.ai/guides/training/recipes.md): Using recipes to streamline model training in Oumi
- [NEXT STEPS](https://docs.oumi.ai/guides/training/next-steps.md): What to do after fine-tuning
- [DEPLOYMENT OPTIONS](https://docs.oumi.ai/guides/deployment.md): Serve a model on a managed inference endpoint, or export it to run on your own infrastructure
- [DEPLOY A MODEL](https://docs.oumi.ai/guides/deployment/deploy-a-model.md): Stand up a managed inference endpoint for your own model or an external one
- [CALLING YOUR ENDPOINT](https://docs.oumi.ai/guides/deployment/calling-your-endpoint.md): Call your Oumi deployment from your own code using the OpenAI or Anthropic SDKs
- [IMPORTING MODELS](https://docs.oumi.ai/guides/deployment/importing-models.md): Bring your own model into Oumi to evaluate and deploy it
- [SELF-HOSTING](https://docs.oumi.ai/guides/deployment/self-hosting.md): Export your trained model from Oumi and serve it on your own hardware or cloud GPU instance
- [INFERENCE LOGS](https://docs.oumi.ai/guides/deployment/inference-logs.md): Review the requests your deployment has served, and turn real production traffic into a training dataset
- [QUALITY MONITORING](https://docs.oumi.ai/guides/deployment/monitoring.md): Score a sample of production traffic with LLM judges and track your deployment's quality over time
- [COMPOUNDING INTELLIGENCE](https://docs.oumi.ai/guides/deployment/compounding-intelligence.md): Turn monitoring insights into targeted training data, retrain your model, and improve it with every production cycle
- [HEALTH METRICS](https://docs.oumi.ai/guides/deployment/health-metrics.md): Track requests, errors, latency, and token throughput for every managed deployment
- [NEXT STEPS](https://docs.oumi.ai/guides/deployment/next-steps.md): Where to go from here
- [MODEL PLAYGROUND](https://docs.oumi.ai/guides/playground.md)
- [MODEL PARAMETERS AND PROMPTS](https://docs.oumi.ai/guides/playground/system-prompts-and-parameters.md): System prompts and model parameters to customize your model's behavior
- [METRICS AND MANAGEMENT](https://docs.oumi.ai/guides/playground/metrics-and-management.md): Comparing model performance and managing your conversations
- [INSTALLATION](https://docs.oumi.ai/cli/installation.md): Install the Oumi CLI on macOS, Linux or Windows
- [QUICKSTART](https://docs.oumi.ai/cli/quickstart.md): Authenticate, select your workspace, and view commands
- [CONFIGURATION AND PROFILES](https://docs.oumi.ai/cli/basics/configuration.md): How the CLI remembers your workspace and settings
- [RECIPES](https://docs.oumi.ai/cli/basics/recipes.md): Create a reusable configuration and launch a run from it.
- [OUTPUT FORMATS](https://docs.oumi.ai/cli/basics/output-formats.md): Output modes, and how to page through long result sets
- [TRAIN A MODEL](https://docs.oumi.ai/cli/workflows/train-a-model.md): Upload a dataset and train a model
- [EVALUATE A MODEL](https://docs.oumi.ai/cli/workflows/evaluate-a-model.md): Score a model against your own criteria and find where it fails
- [DEPLOY A MODEL](https://docs.oumi.ai/cli/workflows/deploy-a-model.md): Serve a model behind an Oumi inference endpoint
- [GLOBAL OPTIONS](https://docs.oumi.ai/cli/commands/global-options.md): Flags, environment variables, and exit codes shared by every command
- [AUTH](https://docs.oumi.ai/cli/commands/auth.md): Sign in to Oumi and check your authentication status
- [CONFIG](https://docs.oumi.ai/cli/commands/config.md): Manage local CLI configuration and named profiles
- [DATASETS](https://docs.oumi.ai/cli/commands/datasets.md): Upload, version, analyze, and synthesize datasets
- [DEPLOYMENTS](https://docs.oumi.ai/cli/commands/deployments.md): Serve models behind an Oumi inference endpoint
- [EVALUATIONS](https://docs.oumi.ai/cli/commands/evaluations.md): Run evaluations and inspect how a model scored
- [EVALUATORS](https://docs.oumi.ai/cli/commands/evaluators.md): Create and manage evaluator resources
- [FILES](https://docs.oumi.ai/cli/commands/files.md): Upload and manage project files and data sources
- [MODELS](https://docs.oumi.ai/cli/commands/models.md): Launch training runs and manage trained models
- [OPERATIONS](https://docs.oumi.ai/cli/commands/operations.md): Track, wait on, and cancel long-running operations
- [ORGANIZATIONS](https://docs.oumi.ai/cli/commands/organizations.md): List organizations and set the active one
- [PROJECTS](https://docs.oumi.ai/cli/commands/projects.md): Create, inspect, and manage Oumi projects
- [RECIPES](https://docs.oumi.ai/cli/commands/recipes.md): Create and manage the reusable configs you launch runs from
- [OUMI ADMINISTRATION](https://docs.oumi.ai/admin.md): Manage platform settings, access, billing, and usage limits
- [ORGANIZATION & PROJECTS](https://docs.oumi.ai/admin/org-projects.md): Setting up your organization & project
- [USER MANAGEMENT](https://docs.oumi.ai/admin/users.md): Manage users and roles within your organization
- [PERMISSIONS](https://docs.oumi.ai/admin/permissions.md): Manage users and roles within your organization
- [PROVIDER API KEYS](https://docs.oumi.ai/admin/api-keys.md): Store API keys for external model providers
- [OUMI BILLING](https://docs.oumi.ai/admin/billing.md): View & manage your organization’s billing & usage details.
- [QUOTAS](https://docs.oumi.ai/admin/quotas.md): View usage limits across your organization
- [TROUBLESHOOTING & ERRORS](https://docs.oumi.ai/admin/troubleshooting-errors.md): Common issues, error responses, and how to resolve them
- [REFERENCE](https://docs.oumi.ai/reference.md): Detail technical definitions & configurations
- [KEY TERMS](https://docs.oumi.ai/reference/key-terms.md): Important Oumi definitions and concepts
- [SUPPORTED MODELS](https://docs.oumi.ai/reference/supported-models.md): Models available for training and inference in Oumi
- [PROMPT LIBRARY](https://docs.oumi.ai/reference/prompts.md): Examples for building, evaluating, and managing models with the Oumi Agent.
- [SYNTHESIS RECIPE SCHEMA](https://docs.oumi.ai/reference/schema/datasets.md): Request schema for dataset synthesis recipes
- [EVALUATOR RECIPE SCHEMA DEFINITIONS](https://docs.oumi.ai/reference/schema/evaluators.md): Schema definitions for configuring evaluators
- [EVALUATION RECIPE SCHEMA DEFINITIONS](https://docs.oumi.ai/reference/schema/evaluations.md): Schema definitions for creating evaluations
- [MODEL TRAINING RECIPE SCHEMA DEFINITIONS](https://docs.oumi.ai/reference/schema/models.md): Schema definitions for training models
- [TRAINING](https://docs.oumi.ai/reference/recipes/training.md): Training recipes
- [EVALUATIONS](https://docs.oumi.ai/reference/recipes/evaluations.md): Evaluation recipes
- [SAVED EVALUATORS](https://docs.oumi.ai/reference/recipes/evaluators.md): Saved evaluator examples
- [SYNTHESIS](https://docs.oumi.ai/reference/recipes/synthesis.md): Synthesis recipes