huggingface / huggingface/datasets

NotImplementedError when using streaming=True in Google Colab environment

Open
#7,698 2 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
22k
Forks
3.4k
Avg merge
5d 7h
Merged PRs (30d)
17

Description

### Describe the bug

When attempting to load a large dataset (like tiiuae/falcon-refinedweb or allenai/c4) using streaming=True in a standard Google Colab notebook, the process fails with a NotImplementedError: Loading a streaming dataset cached in a LocalFileSystem is not supported yet. This issue persists even after upgrading datasets and huggingface_hub and restarting the session.

### Steps to reproduce the bug

Open a new Google Colab notebook.

(Optional but recommended) Run !pip install --upgrade datasets huggingface_hub and restart the runtime.

Run the following code:

Python

from datasets import load_dataset

try:
print("Attempting to load a stream...")
streaming_dataset = load_dataset('tiiuae/falcon-refinedweb', streaming=True)
print("Success!")
except Exception as e:
print(e)

### Expected behavior

The load_dataset command should return a StreamingDataset object without raising an error, allowing iteration over the dataset.

Actual Behavior
The code fails and prints the following error traceback:

[PASTE THE FULL ERROR TRACEBACK HERE]
(Note: Copy the entire error message you received, from Traceback... to the final error line, and paste it in this section.)

### Environment info

Platform: Google Colab

datasets version: [Run !pip show datasets in Colab and paste the version here]

huggingface_hub version: [Run !pip show huggingface_hub and paste the version here]

Python version: [Run !python --version and paste the version here]

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.