Dataset Viewer
Auto-converted to Parquet Duplicate
text
stringlengths
1
1.03M
meta
stringlengths
96
6.55k
red_pajama_subset
stringclasses
6 values
CNHS Biology Majors and Minors Faculty Directory and Research Marine Science Field Station Student Presentations at Professional Scientific Meetings The Department of Biology offers five majors and three minors, in which multiple tracks or "concentrations" are available, and some of which are closely linked to other de...
{"pred_label": "__label__cc", "pred_label_prob": 0.5600966215133667, "wiki_prob": 0.4399033784866333, "source": "cc/2019-30/en_head_0000.json.gz/line1775"}
common_crawl
Delta Nigeria - The Rape of ParadiseGeorge OsodiPrice: £34.99 For more than five centuries the fortunes of the Niger Delta have been closely tied to that of the global economy. For its slave ports, then palm oil industry, and most recently, through the discovery of crude oil in the 1950s. Oil multinationals soon came t...
{"pred_label": "__label__cc", "pred_label_prob": 0.5701277852058411, "wiki_prob": 0.42987221479415894, "source": "cc/2019-30/en_head_0000.json.gz/line14477"}
common_crawl
Which got me thinking about other things that have been forced down our throats by the zeitgeist, people, places, and trends that the pop culture puppeteers insist on making relevant. Here are ten items I could truly do without. If you know me, I'm not usually about spreading the hateration, but sometimes we all need a...
{'timestamp': '2019-04-18T11:08:15Z', 'url': 'http://www.thefirstecho.com/2012/08/10-things-pop-culture-should-stop.html', 'language': 'en', 'source': 'c4'}
c4
Microsoft CFO Peter Klein to depart after 4 years on the job Sean Ludwig April 18, 2013 1:17 PM Updated at 2:32 p.m. PT with internal memos. Microsoft chief financial officer Peter Klein will leave the company at the end of Microsoft’s current fiscal year, the company said today along with announcing its Q3 2013 earni...
{"pred_label": "__label__cc", "pred_label_prob": 0.705028772354126, "wiki_prob": 0.294971227645874, "source": "cc/2019-30/en_head_0000.json.gz/line8208"}
common_crawl
Help - Scheme information The LGPS Information page provides all key communication literature that the LGPS has issued to members. To access any of the literature just click on the relevant document. Many of these documents are in pdf format and require Adobe Reader for you to view them. If you need to download a free ...
{"pred_label": "__label__cc", "pred_label_prob": 0.5906679034233093, "wiki_prob": 0.4093320965766907, "source": "cc/2019-30/en_head_0000.json.gz/line8060"}
common_crawl
"Location: Dubai Financial Services Authority (DFSA) > Media Releases\n2 July 2019 — DFSA accepts (...TRUNCATED)
"{\"pred_label\": \"__label__wiki\", \"pred_label_prob\": 0.8464577198028564, \"wiki_prob\": 0.84645(...TRUNCATED)
common_crawl
"India's growing environmental crisis\nA longstanding ALERT fan, Dr Shaju Thomas from the Tropical I(...TRUNCATED)
"{\"pred_label\": \"__label__cc\", \"pred_label_prob\": 0.6384994387626648, \"wiki_prob\": 0.3615005(...TRUNCATED)
common_crawl
"\\section{Introduction}\nThe problem of sampling complex free energy landscapes that exhibit long l(...TRUNCATED)
"{'timestamp': '2016-07-19T02:06:58', 'yymm': '1607', 'arxiv_id': '1607.04846', 'language': 'en', 'u(...TRUNCATED)
arxiv
"Tucker, H.H. \"Jinx\"\nSportswriter (Waco)\nTucker is probably the only person ever to play footbal(...TRUNCATED)
"{\"pred_label\": \"__label__cc\", \"pred_label_prob\": 0.5364937782287598, \"wiki_prob\": 0.4635062(...TRUNCATED)
common_crawl
"LOS ANGELES -- USC pitcher Ian Kennedy (Huntington Beach, Calif/La Quinta HS) was named today to th(...TRUNCATED)
"{'timestamp': '2019-04-22T04:21:42Z', 'url': 'https://pac-12.com/article/2004/06/25/ian-kennedy-ear(...TRUNCATED)
c4
End of preview. Expand in Data Studio

RedPajama-10B-Weighted

A canonical 10 Billion token weighted subset of the RedPajama-Data-1T dataset.

Dataset Description

This dataset is a faithful reproduction of the original RedPajama-Data-1T distribution, scaled down to exactly 10 Billion tokens. It is designed to preserve the exact domain ratios of the original dataset (excluding the defunct 'Books' subset). This allows researchers and developers to prototype, debug, and test on a representative slice of the data without needing to download or process the full 1 Terabyte dataset.

It serves as both a standalone dataset for medium-scale experiments and the parent source for smaller slices (like the 1B subset).

Dataset Details

Motivation

The original RedPajama dataset is a standard for open-source LLM training, but its size (1TB+) makes it unwieldy for quick iteration, debugging, or educational purposes. Randomly sampling without care can destroy the delicate balance of data sources (CommonCrawl vs. C4 vs. GitHub).

This RedPajama 10B subset solves this by using a weighted interleaving strategy that strictly adheres to the original mixing ratios. It ensures that even at a smaller scale, the data seen by the model is distributionally equivalent to the full run.

Dataset Creation Process

The creation process involved a precise streaming and interleaving pipeline:

1. Source Streaming

We streamed the data directly from togethercomputer/RedPajama-Data-1T to avoid local storage bottlenecks.

2. Weighted Interleaving

We defined the target probabilities based on the original token counts:

  • CommonCrawl: 74.16%
  • C4: 14.78%
  • GitHub: 4.98%
  • ArXiv: 2.36%
  • Wikipedia: 2.03%
  • StackExchange: 1.69%

An interleaving algorithm sampled from these streams according to these probabilities to construct a single, unified stream.

3. Buffer Shuffling

To avoid burstiness (e.g., seeing 1000 Wikipedia articles in a row), we implemented a buffer shuffle with a size of 10,000 documents. This ensures a healthy mixture of domains throughout the dataset.

4. Verification

The process ran until exactly 10 Billion tokens were collected. We verified that the final composition matches the target weights.

Composition

Subset Weight Approx. Tokens
CommonCrawl 74.16% ~7.42 B
C4 14.78% ~1.48 B
GitHub 4.98% ~0.50 B
ArXiv 2.36% ~0.24 B
Wikipedia 2.03% ~0.20 B
StackExchange 1.69% ~0.17 B

Usage

from datasets import load_dataset

# Load the 10B weighted subset
ds = load_dataset("krisbailey/RedPajama-10B-Weighted", split="train")

print(ds)

Citation

If you use this dataset, please cite the original RedPajama work:

@software{together2023redpajama,
  author = {Together Computer},
  title = {RedPajama: An Open Source Recipe to Reproduce LLaMA training dataset},
  month = April,
  year = 2023,
  url = {https://github.com/togethercomputer/RedPajama-Data}
}
Downloads last month
183

Collection including krisbailey/RedPajama-10B-Weighted