Upload dataset metadata and configuration

.gitattributes

@@ -0,0 +1,2 @@
+*.jsonl filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text

CITATION.cff

@@ -0,0 +1,70 @@
+cff-version: 1.2.0
+title: "Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis"
+message: "If you use this dataset, please cite it as shown below."
+authors:
+  - family-names: "Eubanks"
+    given-names: "Adam"
+    email: "theadameubanks@gmail.com"
+    website: "https://adameubanks.github.io/"
+    orcid: "0000-0000-0000-0000"
+title: "Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis"
+version: "1.0.0"
+date-released: 2025-01-15
+abstract: "Yearly-filtered web articles from the FineWeb dataset spanning 2005-2025, specifically processed for temporal language analysis and Word2Vec model training. This dataset enables research into semantic change, concept emergence, and language evolution over time."
+keywords:
+  - "text"
+  - "web-crawl"
+  - "temporal-analysis"
+  - "language-modeling"
+  - "word2vec"
+  - "semantic-change"
+  - "fineweb"
+license: "MIT"
+repository: "https://huggingface.co/datasets/adameubanks/filtered_articles_by_year"
+repository-code: "https://github.com/adameubanks/fineweb_analysis"
+url: "https://adameubanks.github.io/embeddings-over-time/"
+identifiers:
+  - type: "doi"
+    value: "10.5281/zenodo.0000000"
+  - type: "url"
+    value: "https://huggingface.co/datasets/adameubanks/filtered_articles_by_year"
+references:
+  - type: "dataset"
+    title: "The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale"
+    authors:
+      - family-names: "Penedo"
+        given-names: "Guilherme"
+      - family-names: "Kydlíček"
+        given-names: "Hynek"
+      - family-names: "Ben allal"
+        given-names: "Loubna"
+      - family-names: "Lozhkov"
+        given-names: "Anton"
+      - family-names: "Mitchell"
+        given-names: "Margaret"
+      - family-names: "Raffel"
+        given-names: "Colin"
+      - family-names: "Von Werra"
+        given-names: "Leandro"
+      - family-names: "Wolf"
+        given-names: "Thomas"
+    year: 2024
+    url: "https://openreview.net/forum?id=n6SCkn2QaG"
+    doi: "10.48550/arXiv.2406.19627"
+    type: "article"
+    venue: "The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track"
+  - type: "dataset"
+    title: "Yearly Word2Vec Embeddings: Language Evolution from 2005-2025"
+    authors:
+      - family-names: "Eubanks"
+        given-names: "Adam"
+    year: 2025
+    url: "https://huggingface.co/adameubanks/YearlyWord2Vec"
+    type: "model"
+    venue: "Hugging Face Hub"
+  - type: "webpage"
+    title: "Embeddings Over Time: Interactive Demo"
+    url: "https://adameubanks.github.io/embeddings-over-time/"
+    year: 2025
+    type: "webpage"
+    venue: "GitHub Pages"

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Adam Eubanks
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,158 @@
+---
+language:
+- en
+license: mit
+multilinguality:
+- monolingual
+size_categories:
+- n>1T
+source_datasets:
+- HuggingFaceFW/fineweb
+task_categories:
+- text-generation
+- feature-extraction
+task_ids:
+- language-modeling
+- semantic-change-detection
+paperswithcode_id: null
+configs:
+- config_name: "2005"
+  data_files: "year_2005/*.jsonl"
+- config_name: "2006"
+  data_files: "year_2006/*.jsonl"
+- config_name: "2007"
+  data_files: "year_2007/*.jsonl"
+- config_name: "2008"
+  data_files: "year_2008/*.jsonl"
+- config_name: "2009"
+  data_files: "year_2009/*.jsonl"
+- config_name: "2010"
+  data_files: "year_2010/*.jsonl"
+- config_name: "2011"
+  data_files: "year_2011/*.jsonl"
+- config_name: "2012"
+  data_files: "year_2012/*.jsonl"
+- config_name: "2013"
+  data_files: "year_2013/*.jsonl"
+- config_name: "2014"
+  data_files: "year_2014/*.jsonl"
+- config_name: "2015"
+  data_files: "year_2015/*.jsonl"
+- config_name: "2016"
+  data_files: "year_2016/*.jsonl"
+- config_name: "2017"
+  data_files: "year_2017/*.jsonl"
+- config_name: "2018"
+  data_files: "year_2018/*.jsonl"
+- config_name: "2019"
+  data_files: "year_2019/*.jsonl"
+- config_name: "2020"
+  data_files: "year_2020/*.jsonl"
+- config_name: "2021"
+  data_files: "year_2021/*.jsonl"
+- config_name: "2022"
+  data_files: "year_2022/*.jsonl"
+- config_name: "2023"
+  data_files: "year_2023/*.jsonl"
+- config_name: "2024"
+  data_files: "year_2024/*.jsonl"
+- config_name: "2025"
+  data_files: "year_2025/*.jsonl"
+---
+
+# Dataset Card for Filtered Articles by Year
+
+## Dataset Description
+
+- **Repository:** [https://huggingface.co/datasets/adameubanks/filtered_articles_by_year](https://huggingface.co/datasets/adameubanks/filtered_articles_by_year)
+
+### Dataset Summary
+
+The **Filtered Articles by Year** dataset contains yearly-segmented web articles from the [FineWeb dataset](https://huggingface.co/datasets/HuggingFaceFW/fineweb), specifically filtered and processed for temporal language analysis and Word2Vec model training. This dataset spans 21 years (2005-2025) and serves as the foundation for research into semantic change, concept emergence, and language evolution over time.
+
+### Supported Tasks and Leaderboards
+
+This dataset supports several tasks:
+
+- **Language Modeling**: Training language models on specific time periods
+- **Semantic Change Detection**: Tracking how word meanings evolve over time
+- **Temporal Analysis**: Studying language evolution and concept emergence
+- **Word2Vec Training**: Creating time-specific word embeddings
+
+### Languages
+
+The dataset contains English text only.
+
+## Dataset Structure
+
+### Data Instances
+
+Each article contains three fields:
+- **`url`**: Source URL of the web page
+- **`text`**: Extracted and cleaned text content
+- **`hash`**: Content hash for deduplication purposes
+
+### Data Fields
+
+- `url`: string - The source URL of the web page
+- `text`: string - The extracted and cleaned text content
+- `hash`: string - Content hash for deduplication purposes
+
+### Data Splits
+
+Each year is available as a separate configuration, with all articles from that year in the training split.
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+The dataset originates from the **FineWeb dataset**, which contains high-quality web text extracted from Common Crawl (CC-MAIN) snapshots using advanced content extraction and filtering techniques.
+
+#### Who are the source language producers?
+
+The text content comes from web pages crawled by Common Crawl, representing a diverse range of web content producers including news sites, blogs, educational resources, and other web content.
+
+### Annotations
+
+#### Annotation process
+
+No additional annotations were added beyond the original FineWeb processing.
+
+#### Who are the annotators?
+
+N/A - uses original FineWeb annotations and quality scores.
+
+### Personal and Sensitive Information
+
+The dataset contains web content that may include personal information depending on the source websites. Users should be aware of privacy considerations when using this data.
+
+## Additional Information
+
+### Dataset Curators
+
+- **Adam Eubanks** - Primary curator and researcher
+- **FineWeb Team** - Original dataset creators
+
+### Licensing Information
+
+This dataset is released under the MIT License.
+
+### Citation Information
+
+```bibtex
+@misc{filtered_articles_by_year_2025,
+  title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis},
+  author={Adam Eubanks},
+  year={2025},
+  url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year},
+  note={Yearly-filtered web articles for Word2Vec training and semantic change research}
+}
+```
+
+### Contributions
+
+We welcome contributions to improve the dataset documentation, processing scripts, or research applications.
+
+### Dataset Card Authors
+
+Adam Eubanks

dataset_infos.json

@@ -0,0 +1,772 @@
+{
+  "default": {
+    "config_name": "2005",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2005/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2005 (2.3 GB, 689,905 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 689905
+      }
+    },
+    "task_templates": []
+  },
+  "2005": {
+    "config_name": "2005",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2005/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2005 (2.3 GB, 689,905 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 689905
+      }
+    },
+    "task_templates": []
+  },
+  "2006": {
+    "config_name": "2006",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2006/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2006 (3.3 GB, 1,047,683 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 1047683
+      }
+    },
+    "task_templates": []
+  },
+  "2007": {
+    "config_name": "2007",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2007/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2007 (4.5 GB, 1,468,094 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 1468094
+      }
+    },
+    "task_templates": []
+  },
+  "2008": {
+    "config_name": "2008",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2008/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2008 (7.0 GB, 2,379,636 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 2379636
+      }
+    },
+    "task_templates": []
+  },
+  "2009": {
+    "config_name": "2009",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2009/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2009 (9.3 GB, 3,251,110 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3251110
+      }
+    },
+    "task_templates": []
+  },
+  "2010": {
+    "config_name": "2010",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2010/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2010 (11.6 GB, 4,102,893 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 4102893
+      }
+    },
+    "task_templates": []
+  },
+  "2011": {
+    "config_name": "2011",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2011/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2011 (12.5 GB, 4,446,823 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 4446823
+      }
+    },
+    "task_templates": []
+  },
+  "2012": {
+    "config_name": "2012",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2012/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2012 (20.0 GB, 7,276,289 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 7276289
+      }
+    },
+    "task_templates": []
+  },
+  "2013": {
+    "config_name": "2013",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2013/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2013 (15.7 GB, 5,626,713 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 5626713
+      }
+    },
+    "task_templates": []
+  },
+  "2014": {
+    "config_name": "2014",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2014/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2014 (8.7 GB, 2,868,446 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 2868446
+      }
+    },
+    "task_templates": []
+  },
+  "2015": {
+    "config_name": "2015",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2015/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2015 (8.7 GB, 2,762,626 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 2762626
+      }
+    },
+    "task_templates": []
+  },
+  "2016": {
+    "config_name": "2016",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2016/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2016 (9.4 GB, 2,901,744 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 2901744
+      }
+    },
+    "task_templates": []
+  },
+  "2017": {
+    "config_name": "2017",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2017/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2017 (10.1 GB, 3,085,758 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3085758
+      }
+    },
+    "task_templates": []
+  },
+  "2018": {
+    "config_name": "2018",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2018/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2018 (10.4 GB, 3,103,828 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3103828
+      }
+    },
+    "task_templates": []
+  },
+  "2019": {
+    "config_name": "2019",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2019/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2019 (10.9 GB, 3,187,052 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3187052
+      }
+    },
+    "task_templates": []
+  },
+  "2020": {
+    "config_name": "2020",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2020/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2020 (12.9 GB, 3,610,390 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3610390
+      }
+    },
+    "task_templates": []
+  },
+  "2021": {
+    "config_name": "2021",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2021/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2021 (14.3 GB, 3,903,312 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3903312
+      }
+    },
+    "task_templates": []
+  },
+  "2022": {
+    "config_name": "2022",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2022/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2022 (16.5 GB, 4,330,132 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 4330132
+      }
+    },
+    "task_templates": []
+  },
+  "2023": {
+    "config_name": "2023",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2023/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2023 (21.6 GB, 5,188,559 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 5188559
+      }
+    },
+    "task_templates": []
+  },
+  "2024": {
+    "config_name": "2024",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2024/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2024 (27.9 GB, 6,443,985 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 6443985
+      }
+    },
+    "task_templates": []
+  },
+  "2025": {
+    "config_name": "2025",
+    "version": "1.0.0",
+    "data_dir": ".",
+    "data_files": {
+      "train": "year_2025/*.jsonl"
+    },
+    "features": {
+      "url": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "text": {
+        "dtype": "string",
+        "_type": "Value"
+      },
+      "hash": {
+        "dtype": "string",
+        "_type": "Value"
+      }
+    },
+    "builder_name": "json",
+    "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+    "dataset_name": "filtered_articles_by_year",
+    "description": "Articles from 2025 (16.6 GB, 3,625,629 articles)",
+    "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+    "license": "mit",
+    "splits": {
+      "train": {
+        "name": "train",
+        "num_examples": 3625629
+      }
+    },
+    "task_templates": []
+  }
+}

generate_dataset_config.py

@@ -0,0 +1,78 @@
+#!/usr/bin/env python3
+"""
+Generate Hugging Face dataset configuration for filtered_articles_by_year
+"""
+
+import json
+import os
+from pathlib import Path
+
+# Year configurations with metadata
+YEAR_CONFIGS = {
+    "2005": {"size_gb": 2.3, "articles": 689905},
+    "2006": {"size_gb": 3.3, "articles": 1047683},
+    "2007": {"size_gb": 4.5, "articles": 1468094},
+    "2008": {"size_gb": 7.0, "articles": 2379636},
+    "2009": {"size_gb": 9.3, "articles": 3251110},
+    "2010": {"size_gb": 11.6, "articles": 4102893},
+    "2011": {"size_gb": 12.5, "articles": 4446823},
+    "2012": {"size_gb": 20.0, "articles": 7276289},
+    "2013": {"size_gb": 15.7, "articles": 5626713},
+    "2014": {"size_gb": 8.7, "articles": 2868446},
+    "2015": {"size_gb": 8.7, "articles": 2762626},
+    "2016": {"size_gb": 9.4, "articles": 2901744},
+    "2017": {"size_gb": 10.1, "articles": 3085758},
+    "2018": {"size_gb": 10.4, "articles": 3103828},
+    "2019": {"size_gb": 10.9, "articles": 3187052},
+    "2020": {"size_gb": 12.9, "articles": 3610390},
+    "2021": {"size_gb": 14.3, "articles": 3903312},
+    "2022": {"size_gb": 16.5, "articles": 4330132},
+    "2023": {"size_gb": 21.6, "articles": 5188559},
+    "2024": {"size_gb": 27.9, "articles": 6443985},
+    "2025": {"size_gb": 16.6, "articles": 3625629}
+}
+
+def create_year_config(year: str, config: dict) -> dict:
+    return {
+        "config_name": year,
+        "version": "1.0.0",
+        "data_dir": ".",
+        # Point to year shards
+        "data_files": {"train": f"year_{year}/*.jsonl"},
+        "features": {
+            "url": {"dtype": "string", "_type": "Value"},
+            "text": {"dtype": "string", "_type": "Value"},
+            "hash": {"dtype": "string", "_type": "Value"}
+        },
+        "builder_name": "json",
+        "citation": "@misc{filtered_articles_by_year_2025, title={Filtered Articles by Year: FineWeb Dataset for Temporal Language Analysis}, author={Adam Eubanks}, year={2025}, url={https://huggingface.co/datasets/adameubanks/filtered_articles_by_year}, note={Yearly-filtered web articles for Word2Vec training and semantic change research}}",
+        "dataset_name": "filtered_articles_by_year",
+        "description": f"Articles from {year} ({config['size_gb']} GB, {config['articles']:,} articles)",
+        "homepage": "https://adameubanks.github.io/embeddings-over-time/",
+        "license": "mit",
+        "splits": {
+            "train": {
+                "name": "train",
+                "num_examples": config["articles"]
+            }
+        },
+        "task_templates": [],
+    }
+
+def generate_dataset_infos() -> dict:
+    dataset_infos = {}
+    # Default to the first year
+    first_year = sorted(YEAR_CONFIGS.keys())[0]
+    dataset_infos["default"] = create_year_config(first_year, YEAR_CONFIGS[first_year])
+    for year, cfg in YEAR_CONFIGS.items():
+        dataset_infos[year] = create_year_config(year, cfg)
+    return dataset_infos
+
+def main():
+    infos = generate_dataset_infos()
+    with open("dataset_infos.json", "w") as f:
+        json.dump(infos, f, indent=2)
+    print("Wrote dataset_infos.json with", len(infos), "configs")
+
+if __name__ == "__main__":
+    main()

split_jsonl_to_shards.py

@@ -0,0 +1,81 @@
+#!/usr/bin/env python3
+import os
+import sys
+import json
+import argparse
+from pathlib import Path
+
+
+def split_jsonl(input_path: Path, output_dir: Path, lines_per_shard: int = 100_000, prefix: str = None):
+    output_dir.mkdir(parents=True, exist_ok=True)
+    shard_index = 0
+    lines_in_shard = 0
+    shard_file = None
+
+    def open_new_shard():
+        nonlocal shard_index, lines_in_shard, shard_file
+        if shard_file:
+            shard_file.close()
+        shard_name = f"{prefix}_chunk{shard_index:03d}.jsonl" if prefix else f"chunk{shard_index:03d}.jsonl"
+        shard_path = output_dir / shard_name
+        shard_file = shard_path.open("w", encoding="utf-8")
+        lines_in_shard = 0
+        shard_index += 1
+        return shard_path
+
+    open_new_shard()
+
+    with open(input_path, "r", encoding="utf-8") as f:
+        for line_num, line in enumerate(f, 1):
+            if not line.strip():
+                continue
+            # Validate JSON and fields; write original line for speed
+            try:
+                obj = json.loads(line)
+                # minimal validation
+                _ = obj.get("url"), obj.get("text"), obj.get("hash")
+            except Exception:
+                # Skip malformed lines
+                continue
+
+            shard_file.write(line)
+            lines_in_shard += 1
+
+            if lines_in_shard >= lines_per_shard:
+                open_new_shard()
+
+    if shard_file:
+        shard_file.close()
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Split big JSONL into year folder shards")
+    parser.add_argument("inputs", nargs="+", help="Input articles_YYYY.jsonl paths")
+    parser.add_argument("--lines-per-shard", type=int, default=100_000)
+    parser.add_argument("--out-root", type=Path, default=Path("."))
+    args = parser.parse_args()
+
+    for input_str in args.inputs:
+        input_path = Path(input_str)
+        if not input_path.exists():
+            print(f"Missing: {input_path}", file=sys.stderr)
+            continue
+        # infer year
+        name = input_path.name
+        year = None
+        for part in name.split("_"):
+            if part.startswith("20") and part[:4].isdigit():
+                year = part[:4]
+                break
+        if not year:
+            print(f"Could not infer year from {name}", file=sys.stderr)
+            continue
+        out_dir = args.out_root / f"year_{year}"
+        prefix = f"articles_{year}"
+        print(f"Splitting {input_path} -> {out_dir} ...")
+        split_jsonl(input_path, out_dir, args.lines_per_shard, prefix)
+        print(f"Done: {out_dir}")
+
+
+if __name__ == "__main__":
+    main()