NucliaDB Dataset

6.13.0.post6240 verified Sat May 09 auth: no python

Python client for exporting and managing datasets from NucliaDB, a semantic search engine for unstructured data. Version 6.13.0.post6240 supports Python 3.10+. Actively maintained on GitHub.

pip install nucliadb-dataset

Common errors

error ImportError: cannot import name 'NucliaDataset' from 'nucliadb_dataset' ↓

cause Package not installed or installed with wrong name. Also possible version mismatch.

fix

Run pip install nucliadb-dataset and verify version >=6.0.0.

error TypeError: __init__() missing 1 required positional argument: 'api_key' ↓

cause API key not provided or passed incorrectly.

fix

Pass api_key as keyword argument: NucliaDataset(api_key='your-key').

Warnings

breaking In v6.0+, the dataset API was restructured. Methods like `list_datasets()` were renamed to `list()` and `get_dataset()` to `get()`. Old code using deprecated method names will fail. ↓

fix Update method calls: `list()` and `get()` instead of `list_datasets()` and `get_dataset()`.

deprecated The `NucliaDataset` constructor no longer accepts a `region` parameter; use `endpoint` instead. ↓

fix Replace `region='europe-1'` with `endpoint='https://europe-1.nucliadb.com'`.

gotcha API key is required and must be passed as keyword argument `api_key`. Passing it as positional argument causes TypeError. ↓

fix Always use keyword argument: `NucliaDataset(api_key='...', endpoint='...')`.

Imports

NucliaDataset
wrong
```
from nucliadb.dataset import ...
```
correct
```
from nucliadb_dataset import NucliaDataset
```
Wrong package path; correct import is from nucliadb_dataset directly

Quickstart

Basic usage: initialize client, list and retrieve datasets.

from nucliadb_dataset import NucliaDataset
import os

# Initialize with your NucliaDB API key
api_key = os.environ.get('NUCLIADB_API_KEY', '')
dataset = NucliaDataset(api_key=api_key, endpoint='https://your-nucliadb-instance')

# List available datasets
datasets = dataset.list()
print(datasets)

# Fetch a specific dataset by ID
if datasets:
    ds = dataset.get(datasets[0]['id'])
    print(ds)