Exact Duplicate Removal

Remove character-for-character duplicate documents using NeMo Curator’s exact duplicate removal workflow. This method computes MD5 hashes for each document’s text and identifies documents with identical hashes as duplicates.

For an overview of all duplicate removal options, refer to Deduplication .

How It Works

Exact deduplication uses MD5 hashing to identify identical documents:

1. Computes MD5 hash for each document’s text content
2. Groups documents by identical hash values
3. Identifies duplicates and saves IDs for removal

This method targets character-for-character duplicates and is recommended for removing exact copies of documents.

Before You Start

Prerequisites:

• Ray cluster with GPU support (required for distributed processing)
• Stable document identifiers for removal (either existing IDs or IDs assigned by the workflow)

Quick Start

Get started with exact deduplication using the following example of identifying duplicates, then remove them:

1
from nemo_curator.core.client import RayClient
2
from nemo_curator.stages.deduplication.exact.workflow import ExactDeduplicationWorkflow
3
from nemo_curator.stages.text.deduplication.removal_workflow import TextDuplicatesRemovalWorkflow
4
5
ray_client = RayClient()
6
ray_client.start()
7
8
# Step 1: Identify duplicates
9
exact_workflow = ExactDeduplicationWorkflow(
10
    input_path="input_data/",
11
    output_path="./results",
12
    text_field="text",
13
    assign_id=True,
14
    perform_removal=False,
15
    input_filetype="parquet"
16
)
17
exact_workflow.run()
18
# Duplicate IDs saved to ./results/ExactDuplicateIds/
19
20
# Step 2: Remove duplicates
21
removal_workflow = TextDuplicatesRemovalWorkflow(
22
    input_path="input_data/",
23
    ids_to_remove_path="./results/ExactDuplicateIds",
24
    output_path="./deduplicated",
25
    input_filetype="parquet",
26
    input_id_field="_curator_dedup_id",
27
    ids_to_remove_duplicate_id_field="_curator_dedup_id",
28
    id_generator_path="./results/exact_id_generator.json"
29
)
30
removal_workflow.run()
31
# Clean dataset saved to ./deduplicated/

Configuration

Configure exact deduplication using these key parameters:

Removing Duplicates

After identifying duplicates, use TextDuplicatesRemovalWorkflow to remove them:

1
from nemo_curator.core.client import RayClient
2
from nemo_curator.stages.text.deduplication.removal_workflow import TextDuplicatesRemovalWorkflow
3
4
ray_client = RayClient()
5
ray_client.start()
6
7
removal_workflow = TextDuplicatesRemovalWorkflow(
8
    input_path="/path/to/input/data",
9
    ids_to_remove_path="/path/to/output/ExactDuplicateIds",
10
    output_path="/path/to/deduplicated",
11
    input_filetype="parquet",
12
    input_id_field="_curator_dedup_id",
13
    ids_to_remove_duplicate_id_field="_curator_dedup_id",
14
    id_generator_path="/path/to/output/exact_id_generator.json"  # Required if assign_id=True
15
)
16
removal_workflow.run()

Output Format

The exact deduplication process produces the following directory structure:

1
output_path/
2
├── ExactDuplicateIds/              # Duplicate identification results
3
│   └── *.parquet                   # Parquet files with document IDs to remove
4
└── exact_id_generator.json         # ID generator mapping (if assign_id=True)

File Formats

The workflow produces these output files:

1. Duplicate IDs (ExactDuplicateIds/*.parquet):

   • Contains document IDs to remove
   • Format: Parquet files with a single ID column
   • Column name depends on assign_id:
     • When assign_id=True: Column is "_curator_dedup_id"
     • When assign_id=False: Column matches the id_field parameter (e.g., "id")
   • Important: Contains only the IDs of documents to remove, not the full document content

2. ID Generator (exact_id_generator.json):

   • JSON file containing ID generator state
   • Required for removal workflow when assign_id=True
   • Ensures consistent ID mapping across workflow stages

For comparison with other deduplication methods and guidance on when to use exact deduplication, refer to the Deduplication overview .