# Quick Start Series
Welcome to the NVIDIA FLARE Quick Start Series! This guide provides a set of hello-world examples to help you quickly learn how to build federated learning programs using NVIDIA FLARE.
Make sure you have completed the [Installation](installation.md#installation) steps before proceeding.
## Run Modes
FLARE supports three modes for different stages of your workflow:
- **Simulator** ([NVIDIA FLARE FL Simulator](user_guide/nvflare_cli/fl_simulator.md#fl-simulator)) – Runs jobs on a single system for fast testing and algorithm development.
- **POC** ([Proof Of Concept (POC) Command](user_guide/nvflare_cli/poc_command.md#poc-command)) – Simulates deployment on one host with separate processes for clients and server.
- **Production** ([Provisioning and startup package distribution](user_guide/admin_guide/deployment/overview.md#provisioned-setup)) – Distributed deployment using startup kits from provisioning.
Start with the **Simulator** for development, then validate with **POC** before going to **Production**.
## Convert Your ML Code to Federated
Converting existing training code to federated learning requires just 3 changes:
**Step 1: Add FLARE imports to your training script**
```python
import nvflare.client as flare
```
**Step 2: Initialize FLARE and wrap your training loop**
```python
flare.init()
while flare.is_running():
input_model = flare.receive() # receive global model
model.load_state_dict(input_model.params)
# ... your existing training code here ...
output_model = flare.FLModel(
params=model.cpu().state_dict(),
metrics={"accuracy": accuracy},
)
flare.send(output_model) # send updated model back
```
**Step 3: Create a job recipe to define the FL workflow**
```python
from model import MyModel
from nvflare.app_opt.pt.recipes import FedAvgRecipe
from nvflare.recipe import SimEnv
recipe = FedAvgRecipe(
name="my-fedavg-job",
min_clients=2,
num_rounds=5,
model=MyModel(),
train_script="train.py",
)
env = SimEnv(num_clients=2)
run = recipe.execute(env)
```
That’s it. Your training logic stays the same – FLARE handles the communication, aggregation, and orchestration.
For the full Client API reference, see [Client API](programming_guide/execution_api_type/client_api.md#client-api). For pre-built recipes, see [Available Recipes](user_guide/data_scientist_guide/available_recipes.md#available-recipes).
## Hello-world Examples
The following hello-world examples demonstrate different federated learning algorithms and workflows. Each example includes instructions and code to help you get started.
1. **Hello PyTorch** - Federated averaging with PyTorch models and training loops. [Hello PyTorch](hello-world/hello-pt/index.md)
2. **Hello Lightning** - Example using PyTorch Lightning for streamlined model training. [Hello Pytorch Lightning](hello-world/hello-lightning/index.md)
3. **Hello Differential Privacy** - [Federated learning with differential privacy using PyTorch and Opacus for privacy-preserving training.](hello-world/hello-dp/index.html)
4. **Hello TensorFlow** - [Federated averaging using TensorFlow models.](hello-world/hello-tf/index.html)
5. **Hello JAX** - [Federated averaging using JAX, Flax, and Optax on MNIST.](hello-world/hello-jax/index.html)
6. **Hello HuggingFace** - [Federated Qwen SFT/PEFT using HuggingFace Trainer and TRL.](hello-world/hello-huggingface/index.html)
7. **Hello Logistic Regression** - [Federated logistic regression example using scikit-learn.](hello-world/hello-lr/index.html)
8. **Hello Cyclic** - [Cyclic federated learning workflow example.](hello-world/hello-cyclic/index.html)
9. **Hello Tabular Statistics** - [Federated statistics computation example.](hello-world/hello-tabular-stats/index.html)
10. **Hello Flower** - [Running Flower apps in FLARE.](hello-world/hello-flower/index.html)
11. **Hello XGBoost** - [Federated XGBoost example demonstrating gradient boosting for tabular data in a federated setting.](hello-world/hello-xgboost/index.html)
Let’s start with Hello PyTorch: [Hello PyTorch](hello-world/hello-pt/index.md)