Skip to main content
Ctrl+K
We're pleased to announce NeMo Platform, an integrated set of open source tools that are designed to make your agent measurably better, on multiple dimensions such as quality, performance, cost and safety. Visit the GitHub repo or view the latest docs.
NVIDIA NeMo Microservices Documentation - Home NVIDIA NeMo Microservices Documentation - Home

NVIDIA NeMo Microservices Documentation

NVIDIA NeMo Microservices Documentation - Home NVIDIA NeMo Microservices Documentation - Home

NVIDIA NeMo Microservices Documentation

Table of Contents

NeMo Microservices

  • Getting Started
  • Quickstart Installation
  • Core Concepts
    • Workspaces
    • Secrets
    • Files
    • Projects
    • Entities
    • Entity References
  • Studio UI

Example Applications

  • About
  • Tool-calling fine-tuning with Synthetic Data
  • Fine-tune and evaluate an embedding model

Models and Inference

  • About
  • Tutorials
    • Deploy Models
    • Run Inference

Design Synthetic Data

  • About
  • Quickstart
  • Tutorials
    • The Basics
    • Seeding
  • SDK Resources
  • Migrating from Standalone Library

Guardrail Models

  • About
  • Core Concepts
    • Guardrails Architecture
    • Guardrail Configurations
      • Configuration Structure
      • Manage Configurations
      • Default Guardrail Configurations
    • Running Checks with Guardrails
    • Running Inference with Guardrails
  • Tutorials
    • Deploy NemoGuard NIMs
    • Improving Content Safety with NemoGuard NIM
    • Running Rails in Parallel
    • Adding Safety Checks to Multimodal Data
    • Detecting Injection Attacks
  • Terminology
  • Observability

Evaluate Models

  • About
  • Tutorials
    • Run an Evaluation
    • Run an LLM Judge Eval
  • Metrics
    • LLM-as-a-Judge
    • Agentic Metrics
    • RAG Metrics
    • Retriever Metrics
    • Similarity Metrics
    • Bring Your Own Metric
    • Model Configuration
    • Manage Metrics
    • Manage Jobs
    • Results
  • Benchmarks
    • Custom
    • Agentic
    • Industry
    • Manage Benchmarks
    • Manage Jobs
    • Results

Fine-tune Models

  • About
  • Customization Concepts
  • Tutorials
    • Understanding Model Entities
    • Format Training Dataset
    • LoRA Customization
    • SFT Customization
    • DPO Customization
    • Embedding Model Customization
    • Checking Your Customization Job Metrics
    • Optimize Throughput
  • Model Catalog
    • Llama Models
    • Llama Nemotron Models
    • Phi Models
    • Qwen Models
    • Mistral Models
    • Embedding Models
    • GPT-OSS Models
    • Dataset Format Requirements
  • Manage Model Entities
    • Create FileSet
    • Create Model Entity
  • Manage Jobs
    • Create Job
    • Get Job Status
    • List Active Jobs
    • Cancel Job
    • Training Configuration

Synthesize Safe Data

  • About
    • Data Synthesis
    • PII Replacement
    • Evaluation
    • Safe Synthesizer Jobs
    • Parameters Reference
  • Getting Started
  • Tutorials
    • Safe Synthesizer 101
    • Differential Privacy

Audit Model Safety

  • About
  • Tutorials
    • Run a Simple Job
    • Notebook
  • Manage Targets
    • Basic Target
    • NVIDIA Hosted NIM
    • OpenAI
    • List Targets
    • Update and Delete a Target
    • Target Schema
  • Manage Configs
    • Create Config
    • Get a Config
    • List Configs
    • Update and Delete a Config
    • Probes and Plugins
    • Config Schema
  • Manage Audit Jobs
  • Viewing Audit Job Results

Platform

  • About
  • Authentication and Authorization
    • Security Model
    • Concepts
    • Authentication
      • OIDC Setup
      • Provider Guides
        • Azure AD (Entra ID)
        • Generic OIDC Provider
      • Using Authentication
    • Authorization
      • Roles & Permissions
      • Managing Access
      • API Scopes
      • Permissions Reference
      • Policy Engine
    • Configuration Reference
      • Gateway Integration
      • Hardening
      • Credential Propagation
    • Troubleshooting
  • Deploying on Kubernetes
    • Prerequisites
    • Install
    • Database Setup
    • Ingress
    • Persistent Volumes
    • File Storage
    • Multinode Networking
    • OpenShift
    • Backup and Restore
  • Jobs
  • Observability
  • Security
  • Milvus

Reference

  • Release Notes
  • System Requirements
  • API Reference
  • SDK Reference
    • Client APIs
    • API Reference
      • quickstart
        • quickstart.cluster
        • quickstart.config
        • quickstart.container
        • quickstart.gpu_config
        • quickstart.platform_config
        • quickstart.preflight
        • quickstart.prompts
        • quickstart.storage
        • quickstart.validators
      • filesets
        • filesets.resources
      • jobs
        • results
        • steps
        • tasks
      • secrets
        • admin
      • models
        • models.resources
      • inference
        • deployment_configs
        • deployments
        • gateway
        • models
        • providers
      • audit
        • configs
        • jobs
        • plugins
        • targets
      • customization
        • jobs
      • data_designer
        • data_designer.plugins
      • guardrail
        • chat
        • completions
        • configs
      • evaluation
        • benchmark_job_results
        • benchmark_jobs
        • benchmarks
        • metric_job_results
        • metric_jobs
        • metrics
      • safe_synthesizer
        • safe_synthesizer.config
        • safe_synthesizer.job
        • safe_synthesizer.job_builder
  • CLI Reference
    • Configuration
    • Working with Resources
    • Troubleshooting
    • Full CLI Reference
  • Config Reference
  • Helm Reference
  • Troubleshooting
    • Troubleshoot Platform Setup
    • Troubleshoot Auditor
    • Troubleshoot Customizer
    • Troubleshoot Data Designer
    • Troubleshoot Evaluator
    • Troubleshoot Guardrails
    • Troubleshoot Studio
  • EULA
  • OSS License Acknowledgements
  • Governing Terms

Governing Terms#

The software and materials are governed by the NVIDIA Software License Agreement and the Product-Specific Terms for NVIDIA AI Products.

previous

Troubleshooting Studio

next

OSS License Acknowledgements
NVIDIA NVIDIA
Privacy Policy | Your Privacy Choices | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2025, NVIDIA Corporation.