🎨 Data Designer 101: Structured Outputs and Jinja Expressions¶

Click here to download this notebook to your computer.

📚 What you'll learn¶

In this notebook, we will continue our exploration of Data Designer, demonstrating more advanced data generation using structured outputs and Jinja expressions.

If this is your first time using Data Designer, we recommend starting with the first notebook in this 101 series.

📦 Import the essentials¶

The essentials module provides quick access to the most commonly used objects.

In [1]:

Copied!





from data_designer.essentials import (
    CategorySamplerParams,
    DataDesigner,
    DataDesignerConfigBuilder,
    ExpressionColumnConfig,
    InferenceParameters,
    LLMStructuredColumnConfig,
    ModelConfig,
    PersonFromFakerSamplerParams,
    SamplerColumnConfig,
    SamplerType,
    SubcategorySamplerParams,
)
from data_designer.essentials import (
    CategorySamplerParams,
    DataDesigner,
    DataDesignerConfigBuilder,
    ExpressionColumnConfig,
    InferenceParameters,
    LLMStructuredColumnConfig,
    ModelConfig,
    PersonFromFakerSamplerParams,
    SamplerColumnConfig,
    SamplerType,
    SubcategorySamplerParams,
)

⚙️ Initialize the Data Designer interface¶

DataDesigner is the main object that is used to interface with the library.

In [2]:

Copied!

data_designer_client = DataDesigner()
data_designer_client = DataDesigner()

[16:29:58] [INFO] ♻️ Using default model providers from '/Users/amanoel/.data-designer/model_providers.yaml'

🎛️ Define model configurations¶

Each ModelConfig defines a model that can be used during the generation process.
The "model alias" is used to reference the model in the Data Designer config (as we will see below).
The "model provider" is the external service that hosts the model (see the model config docs for more details).
By default, we use build.nvidia.com as the model provider.

In [3]:

Copied!





# This name is set in the model provider configuration.
MODEL_PROVIDER = "nvidia"

# The model ID is from build.nvidia.com.
MODEL_ID = "nvidia/nvidia-nemotron-nano-9b-v2"

# We choose this alias to be descriptive for our use case.
MODEL_ALIAS = "nemotron-nano-v2"

# This sets reasoning to False for the nemotron-nano-v2 model.
SYSTEM_PROMPT = "/no_think"

model_configs = [
    ModelConfig(
        alias=MODEL_ALIAS,
        model=MODEL_ID,
        provider=MODEL_PROVIDER,
        inference_parameters=InferenceParameters(
            temperature=0.5,
            top_p=1.0,
            max_tokens=1024,
        ),
    )
]
# This name is set in the model provider configuration.
MODEL_PROVIDER = "nvidia"

# The model ID is from build.nvidia.com.
MODEL_ID = "nvidia/nvidia-nemotron-nano-9b-v2"

# We choose this alias to be descriptive for our use case.
MODEL_ALIAS = "nemotron-nano-v2"

# This sets reasoning to False for the nemotron-nano-v2 model.
SYSTEM_PROMPT = "/no_think"

model_configs = [
    ModelConfig(
        alias=MODEL_ALIAS,
        model=MODEL_ID,
        provider=MODEL_PROVIDER,
        inference_parameters=InferenceParameters(
            temperature=0.5,
            top_p=1.0,
            max_tokens=1024,
        ),
    )
]

🏗️ Initialize the Data Designer Config Builder¶

The Data Designer config defines the dataset schema and generation process.
The config builder provides an intuitive interface for building this configuration.
The list of model configs is provided to the builder at initialization.

In [4]:

Copied!

config_builder = DataDesignerConfigBuilder(model_configs=model_configs)
config_builder = DataDesignerConfigBuilder(model_configs=model_configs)

🧑‍🎨 Designing our data¶

We will again create a product review dataset, but this time we will use structured outputs and Jinja expressions.
Structured outputs let you specify the exact schema of the data you want to generate.
Data Designer supports schemas specified using either json schema or Pydantic data models (recommended).

We'll define our structured outputs using Pydantic data models

💡 Why Pydantic?

Pydantic models provide better IDE support and type validation.

They are more Pythonic than raw JSON schemas.

They integrate seamlessly with Data Designer's structured output system.

In [5]:

Copied!





from decimal import Decimal
from typing import Literal

from pydantic import BaseModel, Field


# We define a Product schema so that the name, description, and price are generated
# in one go, with the types and constraints specified.
class Product(BaseModel):
    name: str = Field(description="The name of the product")
    description: str = Field(description="A description of the product")
    price: Decimal = Field(description="The price of the product", ge=10, le=1000, decimal_places=2)


class ProductReview(BaseModel):
    rating: int = Field(description="The rating of the product", ge=1, le=5)
    customer_mood: Literal["irritated", "mad", "happy", "neutral", "excited"] = Field(
        description="The mood of the customer"
    )
    review: str = Field(description="A review of the product")
from decimal import Decimal
from typing import Literal

from pydantic import BaseModel, Field


# We define a Product schema so that the name, description, and price are generated
# in one go, with the types and constraints specified.
class Product(BaseModel):
    name: str = Field(description="The name of the product")
    description: str = Field(description="A description of the product")
    price: Decimal = Field(description="The price of the product", ge=10, le=1000, decimal_places=2)


class ProductReview(BaseModel):
    rating: int = Field(description="The rating of the product", ge=1, le=5)
    customer_mood: Literal["irritated", "mad", "happy", "neutral", "excited"] = Field(
        description="The mood of the customer"
    )
    review: str = Field(description="A review of the product")

Next, let's design our product review dataset using a few more tricks compared to the previous notebook.

In [ ]:

Copied!





# Since we often only want a few attributes from Person objects, we can
# set drop=True in the column config to drop the column from the final dataset.
config_builder.add_column(
    SamplerColumnConfig(
        name="customer",
        sampler_type=SamplerType.PERSON_FROM_FAKER,
        params=PersonFromFakerSamplerParams(),
        drop=True,
    )
)

config_builder.add_column(
    SamplerColumnConfig(
        name="product_category",
        sampler_type=SamplerType.CATEGORY,
        params=CategorySamplerParams(
            values=[
                "Electronics",
                "Clothing",
                "Home & Kitchen",
                "Books",
                "Home Office",
            ],
        ),
    )
)

config_builder.add_column(
    SamplerColumnConfig(
        name="product_subcategory",
        sampler_type=SamplerType.SUBCATEGORY,
        params=SubcategorySamplerParams(
            category="product_category",
            values={
                "Electronics": [
                    "Smartphones",
                    "Laptops",
                    "Headphones",
                    "Cameras",
                    "Accessories",
                ],
                "Clothing": [
                    "Men's Clothing",
                    "Women's Clothing",
                    "Winter Coats",
                    "Activewear",
                    "Accessories",
                ],
                "Home & Kitchen": [
                    "Appliances",
                    "Cookware",
                    "Furniture",
                    "Decor",
                    "Organization",
                ],
                "Books": [
                    "Fiction",
                    "Non-Fiction",
                    "Self-Help",
                    "Textbooks",
                    "Classics",
                ],
                "Home Office": [
                    "Desks",
                    "Chairs",
                    "Storage",
                    "Office Supplies",
                    "Lighting",
                ],
            },
        ),
    )
)

config_builder.add_column(
    SamplerColumnConfig(
        name="target_age_range",
        sampler_type=SamplerType.CATEGORY,
        params=CategorySamplerParams(values=["18-25", "25-35", "35-50", "50-65", "65+"]),
    )
)

# Sampler columns support conditional params, which are used if the condition is met.
# In this example, we set the review style to rambling if the target age range is 18-25.
# Note conditional parameters are only supported for Sampler column types.
config_builder.add_column(
    SamplerColumnConfig(
        name="review_style",
        sampler_type=SamplerType.CATEGORY,
        params=CategorySamplerParams(
            values=["rambling", "brief", "detailed", "structured with bullet points"],
            weights=[1, 2, 2, 1],
        ),
        conditional_params={
            "target_age_range == '18-25'": CategorySamplerParams(values=["rambling"]),
        },
    )
)

# Optionally validate that the columns are configured correctly.
config_builder.validate()
# Since we often only want a few attributes from Person objects, we can
# set drop=True in the column config to drop the column from the final dataset.
config_builder.add_column(
    SamplerColumnConfig(
        name="customer",
        sampler_type=SamplerType.PERSON_FROM_FAKER,
        params=PersonFromFakerSamplerParams(),
        drop=True,
    )
)

config_builder.add_column(
    SamplerColumnConfig(
        name="product_category",
        sampler_type=SamplerType.CATEGORY,
        params=CategorySamplerParams(
            values=[
                "Electronics",
                "Clothing",
                "Home & Kitchen",
                "Books",
                "Home Office",
            ],
        ),
    )
)

config_builder.add_column(
    SamplerColumnConfig(
        name="product_subcategory",
        sampler_type=SamplerType.SUBCATEGORY,
        params=SubcategorySamplerParams(
            category="product_category",
            values={
                "Electronics": [
                    "Smartphones",
                    "Laptops",
                    "Headphones",
                    "Cameras",
                    "Accessories",
                ],
                "Clothing": [
                    "Men's Clothing",
                    "Women's Clothing",
                    "Winter Coats",
                    "Activewear",
                    "Accessories",
                ],
                "Home & Kitchen": [
                    "Appliances",
                    "Cookware",
                    "Furniture",
                    "Decor",
                    "Organization",
                ],
                "Books": [
                    "Fiction",
                    "Non-Fiction",
                    "Self-Help",
                    "Textbooks",
                    "Classics",
                ],
                "Home Office": [
                    "Desks",
                    "Chairs",
                    "Storage",
                    "Office Supplies",
                    "Lighting",
                ],
            },
        ),
    )
)

config_builder.add_column(
    SamplerColumnConfig(
        name="target_age_range",
        sampler_type=SamplerType.CATEGORY,
        params=CategorySamplerParams(values=["18-25", "25-35", "35-50", "50-65", "65+"]),
    )
)

# Sampler columns support conditional params, which are used if the condition is met.
# In this example, we set the review style to rambling if the target age range is 18-25.
# Note conditional parameters are only supported for Sampler column types.
config_builder.add_column(
    SamplerColumnConfig(
        name="review_style",
        sampler_type=SamplerType.CATEGORY,
        params=CategorySamplerParams(
            values=["rambling", "brief", "detailed", "structured with bullet points"],
            weights=[1, 2, 2, 1],
        ),
        conditional_params={
            "target_age_range == '18-25'": CategorySamplerParams(values=["rambling"]),
        },
    )
)

# Optionally validate that the columns are configured correctly.
config_builder.validate()

[16:29:58] [INFO] ✅ Validation passed

Out[ ]:

DataDesignerConfigBuilder(
    sampler_columns: [
        "customer",
        "product_category",
        "product_subcategory",
        "target_age_range",
        "review_style"
    ]
)

Next, we will use more advanced Jinja expressions to create new columns.

Jinja expressions let you:

Access nested attributes: {{ customer.first_name }}
Combine values: {{ customer.first_name }} {{ customer.last_name }}
Use conditional logic: {% if condition %}...{% endif %}

In [7]:

Copied!





# We can create new columns using Jinja expressions that reference
# existing columns, including attributes of nested objects.
config_builder.add_column(
    ExpressionColumnConfig(name="customer_name", expr="{{ customer.first_name }} {{ customer.last_name }}")
)

config_builder.add_column(ExpressionColumnConfig(name="customer_age", expr="{{ customer.age }}"))

config_builder.add_column(
    LLMStructuredColumnConfig(
        name="product",
        prompt=(
            "Create a product in the '{{ product_category }}' category, focusing on products  "
            "related to '{{ product_subcategory }}'. The target age range of the ideal customer is "
            "{{ target_age_range }} years old. The product should be priced between $10 and $1000."
        ),
        system_prompt=SYSTEM_PROMPT,
        output_format=Product,
        model_alias=MODEL_ALIAS,
    )
)

# We can even use if/else logic in our Jinja expressions to create more complex prompt patterns.
config_builder.add_column(
    LLMStructuredColumnConfig(
        name="customer_review",
        prompt=(
            "Your task is to write a review for the following product:\n\n"
            "Product Name: {{ product.name }}\n"
            "Product Description: {{ product.description }}\n"
            "Price: {{ product.price }}\n\n"
            "Imagine your name is {{ customer_name }} and you are from {{ customer.city }}, {{ customer.state }}. "
            "Write the review in a style that is '{{ review_style }}'."
            "{% if target_age_range == '18-25' %}"
            "Make sure the review is more informal and conversational."
            "{% else %}"
            "Make sure the review is more formal and structured."
            "{% endif %}"
        ),
        system_prompt=SYSTEM_PROMPT,
        output_format=ProductReview,
        model_alias=MODEL_ALIAS,
    )
)

config_builder.validate()
# We can create new columns using Jinja expressions that reference
# existing columns, including attributes of nested objects.
config_builder.add_column(
    ExpressionColumnConfig(name="customer_name", expr="{{ customer.first_name }} {{ customer.last_name }}")
)

config_builder.add_column(ExpressionColumnConfig(name="customer_age", expr="{{ customer.age }}"))

config_builder.add_column(
    LLMStructuredColumnConfig(
        name="product",
        prompt=(
            "Create a product in the '{{ product_category }}' category, focusing on products  "
            "related to '{{ product_subcategory }}'. The target age range of the ideal customer is "
            "{{ target_age_range }} years old. The product should be priced between $10 and $1000."
        ),
        system_prompt=SYSTEM_PROMPT,
        output_format=Product,
        model_alias=MODEL_ALIAS,
    )
)

# We can even use if/else logic in our Jinja expressions to create more complex prompt patterns.
config_builder.add_column(
    LLMStructuredColumnConfig(
        name="customer_review",
        prompt=(
            "Your task is to write a review for the following product:\n\n"
            "Product Name: {{ product.name }}\n"
            "Product Description: {{ product.description }}\n"
            "Price: {{ product.price }}\n\n"
            "Imagine your name is {{ customer_name }} and you are from {{ customer.city }}, {{ customer.state }}. "
            "Write the review in a style that is '{{ review_style }}'."
            "{% if target_age_range == '18-25' %}"
            "Make sure the review is more informal and conversational."
            "{% else %}"
            "Make sure the review is more formal and structured."
            "{% endif %}"
        ),
        system_prompt=SYSTEM_PROMPT,
        output_format=ProductReview,
        model_alias=MODEL_ALIAS,
    )
)

config_builder.validate()

[16:29:58] [INFO] ✅ Validation passed

Out[7]:

DataDesignerConfigBuilder(
    sampler_columns: [
        "customer",
        "product_category",
        "product_subcategory",
        "target_age_range",
        "review_style"
    ]
    llm_structured_columns: ['product', 'customer_review']
    expression_columns: ['customer_name', 'customer_age']
)

🔁 Iteration is key – preview the dataset!¶

Use the preview method to generate a sample of records quickly.
Inspect the results for quality and format issues.
Adjust column configurations, prompts, or parameters as needed.
Re-run the preview until satisfied.

In [8]:

Copied!

preview = data_designer_client.preview(config_builder)
preview = data_designer_client.preview(config_builder)

[16:29:58] [INFO] 🧐 Preview generation in progress
[16:29:58] [INFO] ✅ Validation passed
[16:29:58] [INFO] ⛓️ Sorting column configs into a Directed Acyclic Graph
[16:29:58] [INFO] 🩺 Running health checks for models...
[16:29:58] [INFO]   |-- 👀 Checking 'nvidia/nvidia-nemotron-nano-9b-v2' in provider named 'nvidia' for model alias 'nemotron-nano-v2'...
[16:30:00] [INFO]   |-- ✅ Passed!
[16:30:00] [INFO] 🎲 Preparing samplers to generate 10 records across 5 columns
[16:30:00] [INFO] 🧩 Generating column `customer_name` from expression
[16:30:00] [INFO] 🧩 Generating column `customer_age` from expression
[16:30:00] [INFO] 🗂️ Preparing llm-structured column generation
[16:30:00] [INFO]   |-- column name: 'product'
[16:30:00] [INFO]   |-- model config:
{
    "alias": "nemotron-nano-v2",
    "model": "nvidia/nvidia-nemotron-nano-9b-v2",
    "inference_parameters": {
        "temperature": 0.5,
        "top_p": 1.0,
        "max_tokens": 1024,
        "max_parallel_requests": 4,
        "timeout": null,
        "extra_body": null
    },
    "provider": "nvidia"
}
[16:30:00] [INFO] 🐙 Processing llm-structured column 'product' with 4 concurrent workers
[16:30:05] [INFO] 🗂️ Preparing llm-structured column generation
[16:30:05] [INFO]   |-- column name: 'customer_review'
[16:30:05] [INFO]   |-- model config:
{
    "alias": "nemotron-nano-v2",
    "model": "nvidia/nvidia-nemotron-nano-9b-v2",
    "inference_parameters": {
        "temperature": 0.5,
        "top_p": 1.0,
        "max_tokens": 1024,
        "max_parallel_requests": 4,
        "timeout": null,
        "extra_body": null
    },
    "provider": "nvidia"
}
[16:30:05] [INFO] 🐙 Processing llm-structured column 'customer_review' with 4 concurrent workers
[16:30:13] [INFO] 📊 Model usage summary:
{
    "nvidia/nvidia-nemotron-nano-9b-v2": {
        "token_usage": {
            "prompt_tokens": 6304,
            "completion_tokens": 3167,
            "total_tokens": 9471
        },
        "request_usage": {
            "successful_requests": 20,
            "failed_requests": 0,
            "total_requests": 20
        },
        "tokens_per_second": 724,
        "requests_per_minute": 91
    }
}
[16:30:13] [INFO] 📐 Measuring dataset column statistics:
[16:30:13] [INFO]   |-- 🎲 column: 'customer'
[16:30:13] [INFO]   |-- 🎲 column: 'product_category'
[16:30:13] [INFO]   |-- 🎲 column: 'product_subcategory'
[16:30:13] [INFO]   |-- 🎲 column: 'target_age_range'
[16:30:13] [INFO]   |-- 🎲 column: 'review_style'
[16:30:13] [INFO]   |-- 🧩 column: 'customer_name'
[16:30:13] [INFO]   |-- 🧩 column: 'customer_age'
[16:30:13] [INFO]   |-- 🗂️ column: 'product'
[16:30:13] [INFO]   |-- 🗂️ column: 'customer_review'
[16:30:13] [INFO] ✅ Preview complete!

In [9]:

Copied!

# Run this cell multiple times to cycle through the 10 preview records.
preview.display_sample_record()
# Run this cell multiple times to cycle through the 10 preview records.
preview.display_sample_record()

                                                                                                                   
                                                 Generated Columns                                                 
┏━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Name                ┃ Value                                                                                     ┃
┡━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ customer            │ {                                                                                         │
│                     │     'uuid': '02a9fa30-045c-4443-9223-d314546ce65b',                                       │
│                     │     'locale': 'en_US',                                                                    │
│                     │     'first_name': 'Steven',                                                               │
│                     │     'last_name': 'Caldwell',                                                              │
│                     │     'middle_name': None,                                                                  │
│                     │     'sex': 'Male',                                                                        │
│                     │     'street_number': '9869',                                                              │
│                     │     'street_name': 'Reynolds Alley',                                                      │
│                     │     'city': 'West Nicholas',                                                              │
│                     │     'state': 'Arizona',                                                                   │
│                     │     'postcode': '05872',                                                                  │
│                     │     'age': 71,                                                                            │
│                     │     'birth_date': '1954-01-09',                                                           │
│                     │     'country': 'Saint Vincent and the Grenadines',                                        │
│                     │     'marital_status': 'separated',                                                        │
│                     │     'education_level': 'some_college',                                                    │
│                     │     'unit': '',                                                                           │
│                     │     'occupation': 'Programmer, systems',                                                  │
│                     │     'phone_number': '001-301-970-1903x043',                                               │
│                     │     'bachelors_field': 'no_degree'                                                        │
│                     │ }                                                                                         │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ product_category    │ Home Office                                                                               │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ product_subcategory │ Storage                                                                                   │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ target_age_range    │ 35-50                                                                                     │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ review_style        │ brief                                                                                     │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ customer_name       │ Steven Caldwell                                                                           │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ customer_age        │ 71                                                                                        │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ product             │ {                                                                                         │
│                     │     'name': 'Ergonomic Under-Desk Storage Cabinet',                                       │
│                     │     'description': 'A sleek and durable storage cabinet designed to fit under desks in a  │
│                     │ home office setup. It features adjustable shelves, a modern design to match various       │
│                     │ decors, and provides ample space for organizing office supplies, documents, and           │
│                     │ electronics. Ideal for professionals aged 35-50 looking to maximize space efficiency.',   │
│                     │     'price': 149.99                                                                       │
│                     │ }                                                                                         │
├─────────────────────┼───────────────────────────────────────────────────────────────────────────────────────────┤
│ customer_review     │ {                                                                                         │
│                     │     'rating': 5,                                                                          │
│                     │     'customer_mood': 'happy',                                                             │
│                     │     'review': 'The Ergonomic Under-Desk Storage Cabinet is a well-designed, durable       │
│                     │ solution for optimizing space in a home office. Its adjustable shelves and modern         │
│                     │ aesthetics make it both functional and visually appealing, particularly suited for        │
│                     │ professionals seeking efficiency.'                                                        │
│                     │ }                                                                                         │
└─────────────────────┴───────────────────────────────────────────────────────────────────────────────────────────┘
                                                                                                                   
                                                    [index: 0]

In [10]:

Copied!

# The preview dataset is available as a pandas DataFrame.
preview.dataset
# The preview dataset is available as a pandas DataFrame.
preview.dataset

Out[10]:

	customer	product_category	product_subcategory	target_age_range	review_style	customer_name	customer_age	product	customer_review
0	{'uuid': '02a9fa30-045c-4443-9223-d314546ce65b...	Home Office	Storage	35-50	brief	Steven Caldwell	71	{'name': 'Ergonomic Under-Desk Storage Cabinet...	{'rating': 5, 'customer_mood': 'happy', 'revie...
1	{'uuid': 'c2631f8f-5d17-4a41-b9f1-36297b9b9e8a...	Home Office	Storage	50-65	detailed	Robert May	61	{'name': 'Ergonomic Wall-Mounted File Cabinet'...	{'rating': 5, 'customer_mood': 'happy', 'revie...
2	{'uuid': '5fb4e814-4186-4d55-a6ca-edef9c3df344...	Electronics	Cameras	50-65	structured with bullet points	Jennifer Morse	21	{'name': 'RetroView Digital Camera', 'descript...	{'rating': 5, 'customer_mood': 'happy', 'revie...
3	{'uuid': '194eb1db-006e-45e8-ad48-48f9b4bf22ed...	Electronics	Headphones	65+	structured with bullet points	David Mays	24	{'name': 'EarComfort Wireless Headphones', 'de...	{'rating': 5, 'customer_mood': 'happy', 'revie...
4	{'uuid': '727adea0-9197-46e2-a3ea-8b17585e9b6c...	Home Office	Lighting	50-65	rambling	Kerri Diaz	102	{'name': 'Adjustable LED Desk Lamp', 'descript...	{'rating': 4, 'customer_mood': 'happy', 'revie...
5	{'uuid': '5fc887bf-fdda-4698-8251-be51f9402a4a...	Clothing	Women's Clothing	35-50	structured with bullet points	Danielle Romero	21	{'name': 'Elegant Floral Maxi Dress', 'descrip...	{'rating': 5, 'customer_mood': 'happy', 'revie...
6	{'uuid': '74c81e95-b43f-487d-b31e-7cf429fffbeb...	Electronics	Headphones	35-50	brief	Ryan Martinez	48	{'name': 'Premium Noise-Cancelling Over-Ear He...	{'rating': 5, 'customer_mood': 'happy', 'revie...
7	{'uuid': '8bc6fdf7-1ba7-49cb-b06b-2d1e0f994a7e...	Home & Kitchen	Decor	50-65	rambling	Mary Weiss	59	{'name': 'Timeless Heritage Wall Art Set', 'de...	{'rating': 5, 'customer_mood': 'happy', 'revie...
8	{'uuid': '94b5f377-8593-40af-81d2-cd5542b6ff33...	Home & Kitchen	Decor	35-50	detailed	Norma Meadows	90	{'name': 'Modern Geometric Wall Art Set', 'des...	{'rating': 5, 'customer_mood': 'happy', 'revie...
9	{'uuid': '78b40c52-9f46-4486-a19a-bb9eca51f87d...	Home & Kitchen	Furniture	50-65	detailed	Brian Miller	29	{'name': 'Ergonomic Memory Foam Reading Chair'...	{'rating': 5, 'customer_mood': 'happy', 'revie...

📊 Analyze the generated data¶

Data Designer automatically generates a basic statistical analysis of the generated data.
This analysis is available via the analysis property of generation result objects.

In [11]:

Copied!

# Print the analysis as a table.
preview.analysis.to_report()
# Print the analysis as a table.
preview.analysis.to_report()

──────────────────────────────────────── 🎨 Data Designer Dataset Profile ─────────────────────────────────────────

                                                                                                                   
                                                 Dataset Overview                                                  
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ number of records               ┃ number of columns               ┃ percent complete records                    ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ 10                              │ 9                               │ 100.0%                                      │
└─────────────────────────────────┴─────────────────────────────────┴─────────────────────────────────────────────┘
                                                                                                                   
                                                                                                                   
                                                🎲 Sampler Columns                                                 
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ column name                    ┃       data type ┃            number unique values ┃               sampler type ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ customer                       │            dict │                     10 (100.0%) │          person_from_faker │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ product_category               │          string │                       4 (40.0%) │                   category │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ product_subcategory            │          string │                       7 (70.0%) │                subcategory │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ target_age_range               │          string │                       3 (30.0%) │                   category │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ review_style                   │          string │                       4 (40.0%) │                   category │
└────────────────────────────────┴─────────────────┴─────────────────────────────────┴────────────────────────────┘
                                                                                                                   
                                                                                                                   
                                             🗂️ LLM-Structured Columns                                              
┏━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┓
┃                       ┃               ┃                            ┃     prompt tokens ┃      completion tokens ┃
┃ column name           ┃     data type ┃       number unique values ┃        per record ┃             per record ┃
┡━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━┩
│ product               │          dict │                10 (100.0%) │     268.0 +/- 0.7 │           77.0 +/- 8.7 │
├───────────────────────┼───────────────┼────────────────────────────┼───────────────────┼────────────────────────┤
│ customer_review       │          dict │                10 (100.0%) │     321.0 +/- 7.1 │        176.0 +/- 137.6 │
└───────────────────────┴───────────────┴────────────────────────────┴───────────────────┴────────────────────────┘
                                                                                                                   
                                                                                                                   
                                               🧩 Expression Columns                                               
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ column name                       ┃                data type ┃                             number unique values ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ customer_name                     │                   string │                                      10 (100.0%) │
├───────────────────────────────────┼──────────────────────────┼──────────────────────────────────────────────────┤
│ customer_age                      │                   string │                                        9 (90.0%) │
└───────────────────────────────────┴──────────────────────────┴──────────────────────────────────────────────────┘
                                                                                                                   
                                                                                                                   
╭────────────────────────────────────────────────── Table Notes ──────────────────────────────────────────────────╮
│                                                                                                                 │
│  1. All token statistics are based on a sample of max(1000, len(dataset)) records.                              │
│  2. Tokens are calculated using tiktoken's cl100k_base tokenizer.                                               │
│                                                                                                                 │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
                                                                                                                   
───────────────────────────────────────────────────────────────────────────────────────────────────────────────────

🆙 Scale up!¶

Happy with your preview data?
Use the create method to submit larger Data Designer generation jobs.

In [12]:

Copied!

job_results = data_designer_client.create(config_builder, num_records=20)
job_results = data_designer_client.create(config_builder, num_records=20)

[16:30:13] [INFO] 🎨 Creating Data Designer dataset
[16:30:13] [INFO] ✅ Validation passed
[16:30:13] [INFO] ⛓️ Sorting column configs into a Directed Acyclic Graph
[16:30:13] [INFO] 🩺 Running health checks for models...
[16:30:13] [INFO]   |-- 👀 Checking 'nvidia/nvidia-nemotron-nano-9b-v2' in provider named 'nvidia' for model alias 'nemotron-nano-v2'...
[16:30:14] [INFO]   |-- ✅ Passed!
[16:30:14] [INFO] ⏳ Processing batch 1 of 1
[16:30:14] [INFO] 🎲 Preparing samplers to generate 20 records across 5 columns
[16:30:14] [INFO] 🧩 Generating column `customer_name` from expression
[16:30:14] [INFO] 🧩 Generating column `customer_age` from expression
[16:30:14] [INFO] 🗂️ Preparing llm-structured column generation
[16:30:14] [INFO]   |-- column name: 'product'
[16:30:14] [INFO]   |-- model config:
{
    "alias": "nemotron-nano-v2",
    "model": "nvidia/nvidia-nemotron-nano-9b-v2",
    "inference_parameters": {
        "temperature": 0.5,
        "top_p": 1.0,
        "max_tokens": 1024,
        "max_parallel_requests": 4,
        "timeout": null,
        "extra_body": null
    },
    "provider": "nvidia"
}
[16:30:14] [INFO] 🐙 Processing llm-structured column 'product' with 4 concurrent workers
[16:30:22] [INFO] 🗂️ Preparing llm-structured column generation
[16:30:22] [INFO]   |-- column name: 'customer_review'
[16:30:22] [INFO]   |-- model config:
{
    "alias": "nemotron-nano-v2",
    "model": "nvidia/nvidia-nemotron-nano-9b-v2",
    "inference_parameters": {
        "temperature": 0.5,
        "top_p": 1.0,
        "max_tokens": 1024,
        "max_parallel_requests": 4,
        "timeout": null,
        "extra_body": null
    },
    "provider": "nvidia"
}
[16:30:22] [INFO] 🐙 Processing llm-structured column 'customer_review' with 4 concurrent workers
[16:30:35] [INFO] 📊 Model usage summary:
{
    "nvidia/nvidia-nemotron-nano-9b-v2": {
        "token_usage": {
            "prompt_tokens": 12414,
            "completion_tokens": 5547,
            "total_tokens": 17961
        },
        "request_usage": {
            "successful_requests": 40,
            "failed_requests": 0,
            "total_requests": 40
        },
        "tokens_per_second": 876,
        "requests_per_minute": 117
    }
}
[16:30:35] [INFO] 📐 Measuring dataset column statistics:
[16:30:35] [INFO]   |-- 🎲 column: 'customer'
[16:30:35] [INFO]   |-- 🎲 column: 'product_category'
[16:30:35] [INFO]   |-- 🎲 column: 'product_subcategory'
[16:30:35] [INFO]   |-- 🎲 column: 'target_age_range'
[16:30:35] [INFO]   |-- 🎲 column: 'review_style'
[16:30:35] [INFO]   |-- 🧩 column: 'customer_name'
[16:30:35] [INFO]   |-- 🧩 column: 'customer_age'
[16:30:35] [INFO]   |-- 🗂️ column: 'product'
[16:30:35] [INFO]   |-- 🗂️ column: 'customer_review'

In [13]:

Copied!

# Load the generated dataset as a pandas DataFrame.
dataset = job_results.load_dataset()

dataset.head()
# Load the generated dataset as a pandas DataFrame.
dataset = job_results.load_dataset()

dataset.head()

Out[13]:

	customer	product_category	product_subcategory	target_age_range	review_style	customer_name	customer_age	product	customer_review
0	{'age': 93, 'bachelors_field': 'stem', 'birth_...	Books	Classics	35-50	structured with bullet points	Angela Mitchell	93	{'description': 'A curated selection of classi...	{'customer_mood': 'happy', 'rating': 5, 'revie...
1	{'age': 45, 'bachelors_field': 'no_degree', 'b...	Clothing	Accessories	25-35	detailed	Alex Fletcher	45	{'description': 'A sleek and versatile crossbo...	{'customer_mood': 'happy', 'rating': 5, 'revie...
2	{'age': 101, 'bachelors_field': 'arts_humaniti...	Home Office	Lighting	25-35	detailed	Kathryn Stafford	101	{'description': 'A sleek, adjustable LED desk ...	{'customer_mood': 'happy', 'rating': 5, 'revie...
3	{'age': 42, 'bachelors_field': 'arts_humanitie...	Books	Non-Fiction	35-50	structured with bullet points	Daniel Freeman	42	{'description': 'A non-fiction book exploring ...	{'customer_mood': 'happy', 'rating': 5, 'revie...
4	{'age': 39, 'bachelors_field': 'education', 'b...	Home Office	Desks	50-65	brief	Brittany Acosta	39	{'description': 'A sturdy and comfortable desk...	{'customer_mood': 'happy', 'rating': 5, 'revie...

In [14]:

Copied!

# Load the analysis results into memory.
analysis = job_results.load_analysis()

analysis.to_report()
# Load the analysis results into memory.
analysis = job_results.load_analysis()

analysis.to_report()

──────────────────────────────────────── 🎨 Data Designer Dataset Profile ─────────────────────────────────────────

                                                                                                                   
                                                 Dataset Overview                                                  
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ number of records               ┃ number of columns               ┃ percent complete records                    ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ 20                              │ 9                               │ 100.0%                                      │
└─────────────────────────────────┴─────────────────────────────────┴─────────────────────────────────────────────┘
                                                                                                                   
                                                                                                                   
                                                🎲 Sampler Columns                                                 
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ column name                    ┃       data type ┃            number unique values ┃               sampler type ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ customer                       │            dict │                     20 (100.0%) │          person_from_faker │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ product_category               │          string │                       5 (25.0%) │                   category │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ product_subcategory            │          string │                      11 (55.0%) │                subcategory │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ target_age_range               │          string │                       5 (25.0%) │                   category │
├────────────────────────────────┼─────────────────┼─────────────────────────────────┼────────────────────────────┤
│ review_style                   │          string │                       4 (20.0%) │                   category │
└────────────────────────────────┴─────────────────┴─────────────────────────────────┴────────────────────────────┘
                                                                                                                   
                                                                                                                   
                                             🗂️ LLM-Structured Columns                                              
┏━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┓
┃                       ┃               ┃                            ┃     prompt tokens ┃      completion tokens ┃
┃ column name           ┃     data type ┃       number unique values ┃        per record ┃             per record ┃
┡━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━┩
│ product               │          dict │                20 (100.0%) │     268.0 +/- 0.8 │           63.5 +/- 7.5 │
├───────────────────────┼───────────────┼────────────────────────────┼───────────────────┼────────────────────────┤
│ customer_review       │          dict │                20 (100.0%) │     308.0 +/- 7.1 │         173.5 +/- 93.8 │
└───────────────────────┴───────────────┴────────────────────────────┴───────────────────┴────────────────────────┘
                                                                                                                   
                                                                                                                   
                                               🧩 Expression Columns                                               
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ column name                       ┃                data type ┃                             number unique values ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ customer_name                     │                   string │                                      20 (100.0%) │
├───────────────────────────────────┼──────────────────────────┼──────────────────────────────────────────────────┤
│ customer_age                      │                   string │                                       19 (95.0%) │
└───────────────────────────────────┴──────────────────────────┴──────────────────────────────────────────────────┘
                                                                                                                   
                                                                                                                   
╭────────────────────────────────────────────────── Table Notes ──────────────────────────────────────────────────╮
│                                                                                                                 │
│  1. All token statistics are based on a sample of max(1000, len(dataset)) records.                              │
│  2. Tokens are calculated using tiktoken's cl100k_base tokenizer.                                               │
│                                                                                                                 │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
                                                                                                                   
───────────────────────────────────────────────────────────────────────────────────────────────────────────────────

⏭️ Next Steps¶

Check out the following notebook to learn more about:

Seeding synthetic data generation with an external dataset