The Basics

1
import data_designer.config as dd
2
3
MODEL_ALIAS = "text"
4
5
model_configs = [
6
    dd.ModelConfig(
7
        provider="default/nvidia-build",
8
        model="nvidia/nemotron-3-nano-30b-a3b",  # Use the `served_model_name` from the provider
9
        alias=MODEL_ALIAS,
10
        inference_parameters=dd.ChatCompletionInferenceParams(
11
            temperature=1.0,
12
            top_p=1.0,
13
        ),
14
    )
15
]
16
17
config_builder = dd.DataDesignerConfigBuilder(model_configs)

1
# Product category sampler
2
config_builder.add_column(
3
    dd.SamplerColumnConfig(
4
        name="product_category",
5
        sampler_type=dd.SamplerType.CATEGORY,
6
        params=dd.CategorySamplerParams(
7
            values=[
8
                "Electronics",
9
                "Clothing",
10
                "Home & Kitchen",
11
                "Books",
12
                "Home Office",
13
            ],
14
        ),
15
    )
16
)
17
18
# Product subcategory sampler (conditional on category)
19
config_builder.add_column(
20
    dd.SamplerColumnConfig(
21
        name="product_subcategory",
22
        sampler_type=dd.SamplerType.SUBCATEGORY,
23
        params=dd.SubcategorySamplerParams(
24
            category="product_category",
25
            values={
26
                "Electronics": [
27
                    "Smartphones",
28
                    "Laptops",
29
                    "Headphones",
30
                    "Cameras",
31
                    "Accessories",
32
                ],
33
                "Clothing": [
34
                    "Men's Clothing",
35
                    "Women's Clothing",
36
                    "Winter Coats",
37
                    "Activewear",
38
                    "Accessories",
39
                ],
40
                "Home & Kitchen": [
41
                    "Appliances",
42
                    "Cookware",
43
                    "Furniture",
44
                    "Decor",
45
                    "Organization",
46
                ],
47
                "Books": [
48
                    "Fiction",
49
                    "Non-Fiction",
50
                    "Self-Help",
51
                    "Textbooks",
52
                    "Classics",
53
                ],
54
                "Home Office": [
55
                    "Desks",
56
                    "Chairs",
57
                    "Storage",
58
                    "Office Supplies",
59
                    "Lighting",
60
                ],
61
            },
62
        ),
63
    )
64
)
65
66
# Target age range
67
config_builder.add_column(
68
    dd.SamplerColumnConfig(
69
        name="target_age_range",
70
        sampler_type=dd.SamplerType.CATEGORY,
71
        params=dd.CategorySamplerParams(
72
            values=["18-25", "25-35", "35-50", "50-65", "65+"]
73
        ),
74
    )
75
)
76
77
# Customer details using Faker
78
config_builder.add_column(
79
    dd.SamplerColumnConfig(
80
        name="customer",
81
        sampler_type=dd.SamplerType.PERSON_FROM_FAKER,
82
        params=dd.PersonFromFakerSamplerParams(age_range=[18, 70], locale="en_US"),
83
    )
84
)
85
86
# Star rating
87
config_builder.add_column(
88
    dd.SamplerColumnConfig(
89
        name="number_of_stars",
90
        sampler_type=dd.SamplerType.UNIFORM,
91
        params=dd.UniformSamplerParams(low=1, high=5),
92
        convert_to="int",  # Convert the sampled float to an integer
93
    )
94
)
95
96
# Review style
97
config_builder.add_column(
98
    dd.SamplerColumnConfig(
99
        name="review_style",
100
        sampler_type=dd.SamplerType.CATEGORY,
101
        params=dd.CategorySamplerParams(
102
            values=["rambling", "brief", "detailed", "structured with bullet points"],
103
            weights=[1, 2, 2, 1],
104
        ),
105
    )
106
)
107
108
# LLM-generated product name
109
config_builder.add_column(
110
    dd.LLMTextColumnConfig(
111
        name="product_name",
112
        prompt=(
113
            "You are a helpful assistant that generates product names. DO NOT add quotes around the product name.\n\n"
114
            "Come up with a creative product name for a product in the '{{ product_category }}' category, focusing "
115
            "on products related to '{{ product_subcategory }}'. The target age range of the ideal customer is "
116
            "{{ target_age_range }} years old. Respond with only the product name, no other text."
117
        ),
118
        model_alias=MODEL_ALIAS,
119
    )
120
)
121
122
# LLM-generated customer review
123
config_builder.add_column(
124
    dd.LLMTextColumnConfig(
125
        name="customer_review",
126
        prompt=(
127
            "You are a customer named {{ customer.first_name }} from {{ customer.city }}, {{ customer.state }}. "
128
            "You are {{ customer.age }} years old and recently purchased a product called {{ product_name }}. "
129
            "Write a review of this product, which you gave a rating of {{ number_of_stars }} stars. "
130
            "The style of the review should be '{{ review_style }}'. "
131
            "Respond with only the review, no other text."
132
        ),
133
        model_alias=MODEL_ALIAS,
134
    )
135
)

1
def load_config_builder() -> dd.DataDesignerConfigBuilder:
2
    return config_builder

1
import os
2
from nemo_platform import NeMoPlatform
3
4
base_url = os.environ.get("NMP_BASE_URL", "http://localhost:8080")
5
client = NeMoPlatform(base_url=base_url, workspace="default")
6
7
data_designer = client.data_designer

1
preview = data_designer.preview(config_builder)
2
3
# Display a random sample record
4
preview.display_sample_record()
5
6
# Access the full preview dataset as a pandas DataFrame
7
df = preview.dataset
8
print(df.head())
9
10
# View statistical analysis
11
preview.analysis.to_report()

1
# Defaulting to 30 for demo speed purposes. Happy with the output? Scale it up!
2
job = data_designer.create(config_builder, num_records=30)
3
4
# Block until the job completes
5
job.wait_until_done()
6
7
# Download the generated artifacts
8
results = job.download_artifacts()
9
10
# Load the dataset as a pandas DataFrame
11
dataset = results.load_dataset()
12
print(dataset.head())
13
14
# Load the full analysis report
15
analysis = results.load_analysis()
16
analysis.to_report()