MINT-1T-PDF-CC-2023-06
DatasetFreeDataset by mlfoundations. 5,39,406 downloads.
Capabilities6 decomposed
large-scale multimodal document-image-text dataset curation and indexing
Medium confidenceProvides a curated dataset of 1 trillion tokens spanning 539,406 PDF documents with aligned image-to-text pairs extracted from Common Crawl 2023-06 snapshot. The dataset uses a hierarchical indexing structure that maps document boundaries, page-level image coordinates, and corresponding OCR/text extractions, enabling efficient retrieval of multimodal training samples at scale without requiring full dataset materialization in memory.
Combines 1 trillion tokens of document text with aligned page-level images from a single Common Crawl snapshot, providing temporally-consistent multimodal pairs at unprecedented scale — most competing datasets either use synthetic image-text pairs or lack document-level coherence across modalities
Larger and more document-focused than LAION-5B (which emphasizes web images) and more naturally-paired than synthetic datasets like Synthetic Docvqa, with real-world OCR challenges that improve model robustness
streaming dataset access with lazy loading and batching
Medium confidenceImplements HuggingFace Datasets streaming protocol that enables on-demand loading of document samples without downloading the full 1T token dataset upfront. The architecture uses memory-mapped file access and configurable batch sampling strategies, allowing training loops to fetch and cache only the samples needed for each epoch while maintaining deterministic shuffling across distributed workers.
Uses HuggingFace's streaming protocol with deterministic shuffling and worker-aware sharding, enabling true distributed training without pre-downloading — avoids the storage bottleneck that limits competitors like LAION-5B when used in multi-node setups
More practical for large-scale training than downloading full datasets upfront, and more deterministic than ad-hoc web scraping approaches that lack reproducibility
document-level metadata and provenance tracking
Medium confidenceMaintains structured metadata for each document including source URL, Common Crawl snapshot date (2023-06), document hash, page count, and extraction quality scores. This metadata is queryable and filterable within the dataset, allowing users to select subsets based on source domain, quality thresholds, or temporal characteristics without scanning the full corpus.
Embeds Common Crawl provenance (URLs, crawl dates, document hashes) directly in the dataset schema, enabling reproducible filtering and bias analysis — most competing datasets either lack this metadata or store it separately, making it harder to correlate quality with source
Provides better auditability and reproducibility than datasets without source tracking, and more granular filtering than datasets with only aggregate statistics
image-text pair extraction with layout-aware alignment
Medium confidenceExtracts page-level images from PDF documents and aligns them with corresponding OCR/text content using spatial layout information (bounding boxes, reading order). The extraction pipeline preserves document structure (headers, footers, tables, body text) by analyzing PDF internal structure and image coordinates, creating naturally-aligned multimodal pairs suitable for vision-language model training without requiring post-hoc alignment.
Preserves document layout structure through PDF internal coordinate systems rather than post-hoc image analysis, enabling structurally-aware alignment that captures reading order and spatial relationships — most competing datasets either discard layout information or infer it from image analysis alone
More accurate layout alignment than image-only document datasets, and more scalable than manually-annotated document datasets like DocVQA
common crawl snapshot integration and temporal consistency
Medium confidenceDataset is derived from a single Common Crawl snapshot (2023-06), ensuring temporal consistency across all documents — all PDFs were crawled within a specific time window, avoiding temporal distribution shifts that occur when combining data from multiple crawl dates. The integration includes Common Crawl metadata (WARC records, crawl IDs) enabling users to trace documents back to original crawl artifacts for verification or re-extraction.
Anchors entire dataset to a single Common Crawl snapshot (2023-06) with traceable WARC references, ensuring temporal consistency and reproducibility — most competing web-derived datasets either combine multiple crawl dates or lack explicit Common Crawl integration
More reproducible than datasets combining multiple crawl dates, and more verifiable than proprietary datasets without public provenance
cc-by-4.0 licensed dataset with commercial use rights
Medium confidenceDataset is released under Creative Commons Attribution 4.0 (CC-BY-4.0) license, permitting commercial use, modification, and redistribution with attribution. The license is applied at the dataset level, though individual documents may have different licenses — users are responsible for verifying compliance for derived works, but the dataset itself imposes minimal legal restrictions on model training and deployment.
Explicitly licensed under CC-BY-4.0 with clear commercial use rights, reducing legal friction for commercial model training — many competing datasets either lack explicit licensing or use more restrictive licenses (e.g., non-commercial only)
More commercially-friendly than datasets with non-commercial restrictions, and more legally transparent than datasets with unclear licensing
Capabilities are decomposed by AI analysis. Each maps to specific user intents and improves with match feedback.
Related Artifactssharing capabilities
Artifacts that share capabilities with MINT-1T-PDF-CC-2023-06, ranked by overlap. Discovered automatically through the match graph.
MINT-1T-PDF-CC-2024-18
Dataset by mlfoundations. 10,34,415 downloads.
FineFineWeb
Dataset by m-a-p. 5,55,725 downloads.
documentation-images
Dataset by huggingface. 24,44,926 downloads.
MINT-1T-PDF-CC-2023-14
Dataset by mlfoundations. 5,72,108 downloads.
MINT-1T-PDF-CC-2023-50
Dataset by mlfoundations. 7,96,577 downloads.
MINT-1T-PDF-CC-2023-23
Dataset by mlfoundations. 6,33,111 downloads.
Best For
- ✓ML researchers training multimodal foundation models at scale
- ✓Teams building document understanding and OCR systems
- ✓Organizations developing enterprise document processing pipelines
- ✓Teams with limited local storage but access to high-bandwidth cloud infrastructure
- ✓Researchers iterating rapidly on model architectures and hyperparameters
- ✓Distributed training setups requiring deterministic data sharding across nodes
- ✓Researchers studying dataset bias and composition effects on model performance
- ✓Teams building production document systems that need quality guarantees
Known Limitations
- ⚠1T token size requires distributed storage infrastructure — not suitable for single-machine training without streaming/sharding
- ⚠PDF extraction quality varies by source document; OCR errors propagate into training data
- ⚠No built-in deduplication across documents — may contain near-duplicate content from web crawl
- ⚠Image resolution and quality varies significantly across source PDFs; no normalization applied
- ⚠English-language dominant; multilingual coverage limited to incidental non-English content in PDFs
- ⚠Streaming introduces network latency — slower than local SSD access by 2-5x depending on connection quality
Requirements
Input / Output
UnfragileRank
UnfragileRank is computed from adoption signals, documentation quality, ecosystem connectivity, match graph feedback, and freshness. No artifact can pay for a higher rank.
About
MINT-1T-PDF-CC-2023-06 — a dataset on HuggingFace with 5,39,406 downloads
Categories
Alternatives to MINT-1T-PDF-CC-2023-06
Are you the builder of MINT-1T-PDF-CC-2023-06?
Claim this artifact to get a verified badge, access match analytics, see which intents users search for, and manage your listing.
Get the weekly brief
New tools, rising stars, and what's actually worth your time. No spam.
Data Sources
Looking for something else?
Search →