Add v0.3 source license review and quality mix workflow

README.md

@@ -73,6 +73,37 @@ OCR garble); heavy quality filtering and mix-weighting are left to downstream tr
 Evaluation-set decontamination is applied/marked separately. Schema:
 `id, text, source, added, created, token_count`.
 
+## v0.3 quality roadmap
+
+The v0.2 raw corpus is intentionally provenance-first, but its token mix is too
+heavy in legal/parliamentary language for natural general pretraining. The v0.3
+workflow therefore separates **source inclusion** from **training mix**:
+
+- cap `eurlex + parliamentary + dziennik_ustaw` to roughly **10-20%** of training
+  tokens combined;
+- use source-level temperature sampling (`sqrt`, alpha `0.5`) instead of raw
+  token-proportional sampling;
+- add traceably licensed contemporary/natural Polish: open web, guides,
+  technical documentation/blogs, Q&A, and dialogue/instruction data;
+- run aggressive exact, normalized, and near-duplicate removal;
+- reserve the final **5-15%** of training for higher-quality sources rather than
+  the largest sources;
+- evaluate per-source perplexity and style contamination, not only global loss;
+- treat GPT-2 124M as a cheap dataset probe, not proof of final model quality.
+
+Current review artifacts:
+
+- `configs/source_candidates_v0_3.json` — candidate decisions and license policy.
+- `artifacts/source_license_review_v0_3.md` — source-by-source license review.
+- `artifacts/source_candidate_audit_v0_3.md` — generated Hugging Face metadata audit.
+- `artifacts/training_mix_v0_3.md` — example 1B-token training mix with legal sources capped at 15%.
+
+TVP Info-derived news data is currently **blocked**: the HF upload
+`WiktorS/polish-news` has an `apache-2.0` tag, but its README says the articles
+were obtained from `tvp.info.pl`, and no upstream TVP Info open license has been
+verified. It should only be included with explicit permission or authoritative
+upstream open-license evidence.
+
 ## Excluded sources (transparency)
 Sources we reviewed and **deliberately left out** — part of the curation:
 

artifacts/hf_readmes/PleIAs__Polish-PD__README.md

@@ -0,0 +1,46 @@
+# 🇵🇱 Polish Public Domain 🇵🇱
+
+**Polish-Public Domain** or **Polish-PD** is a large collection aiming to aggregate all Polish monographies and periodicals in the public domain. As of March 2024, it is the biggest Polish open corpus. 
+
+## Dataset summary
+The collection contains 247,491 individual texts making up 2,697,414,811 words recovered from multiple sources, including Internet Archive and various European national libraries and cultural heritage institutions. Each parquet file has the full text of 2,000 books selected at random.
+
+## Curation method 
+The composition of the dataset adheres to the criteria for public domain works in the EU and, consequently, all Berne-countries for EU authors: any publication whose author is dead for more than 70 years. Additionally, the initial consolidation of public domain status for cultural heritage operates in the EU under the 2019 Copyright Directive (art. 14).
+
+As of March 2024, to limit rights verification, we have retained exclusively titles published prior to 1884. 
+
+The corpus will be expanded at a later stage to encompass late 19th century and early 20th century publications, after checking for public domain validity.
+
+## Uses
+The collection aims to expand the availability of open works for the training of Large Language Models. The text can be used for model training and republished without restriction for reproducibility purposes.
+
+The rationales for creation of this collection are multifold: 
+
+* **Scientific**: We observe that the closure of training corpora represents a major barrier to AI research. Large language models face a real crisis of reproducibility. 
+* **Legal**: With the adoption of the AI Act with its obligations in terms of copyright law compliance for the pretraining corpora, the European AI ecosystem will have to change its provenance practices. 
+* **Cultural**: The linguistic diversity of the European Union is currently underrepresented. Unlike web archives, open, heritage, administrative, or scientific texts are often of high quality: they are long, multilingual, and editorialized publications.
+* **Economical**: Today, value capture is concentrated on players whose financial resources are already considerable, allowing them to collect or purchase data at a high price. Making a royalty-free corpus available to as many people as possible frees innovation in uses and minimizes economic dependencies on dominant actors.
+
+## License
+The entire collection is in the public domain in all regions. This means that the patrimonial rights of each individual or collective right holders have expired.
+
+There has been a debate for years in Europe over the definition of public domain and the possibility to restrict its use. Since 2019, the EU Copyright Directive states that "Member States shall provide that, when the term of protection of a work of visual art has expired, any material resulting from an act of reproduction of that work is not subject to copyright or related rights, unless the material resulting from that act of reproduction is original in the sense that it is the author's own intellectual creation." (art. 14)
+
+## Future work
+This dataset is not a one-time work but will continue to evolve significantly in three directions: 
+
+* Expansion of the dataset to the late 19th and early 20th century works and its further enhancement with currently unexploited collections coming from European patrimonial data repositories. 
+* Correction of computer generated errors in the text. All the texts have been transcribed automatically through the use of Optical Character Recognition (OCR) software. The original files have been digitized over a long time period (since the mid-2000s) and some documents should be. Future versions will strive either to re-OCRize the original text or use experimental LLM models for partial OCR correction.
+* Enhancement of the structure/editorial presentation of the original text. Some parts of the original documents are likely unwanted for large scale analysis or model training (header, page count…). Additionally, some advanced document structures like tables or multi-column layout are unlikely to be well-formatted. 
+
+## Acknowledgements
+The corpus was stored and processed with the generous support of Scaleway. It was built up with the support and concerted efforts of the state start-up LANGU:IA (start-up d’Etat), supported by the French Ministry of Culture and DINUM, as part of the prefiguration of the service offering of the Alliance for Language technologies EDIC (ALT-EDIC).
+
+Corpus collection has been largely facilitated thanks to the open science LLM community insights, cooperation and support (Occiglot, Eleuther AI, OpenLLM France, Allen AI).
+
+<div style="text-align: center;">
+  <img src="https://github.com/mch-dd/datasetlogo/blob/main/scaleway.jpeg?raw=true" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+  <img src="https://github.com/mch-dd/datasetlogo/blob/main/ministere.png?raw=true" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+  <img src="https://github.com/mch-dd/datasetlogo/blob/main/occiglot.jpg?raw=true" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+</div>

artifacts/hf_readmes/PleIAs__common_corpus__README.md

@@ -0,0 +1,134 @@
+---
+language:
+- en
+- fr
+- de
+- zh
+- it
+- es
+- ja
+- pl
+- la
+- nl
+- ru
+- ar
+- ko
+
+configs:
+  - config_name: default
+    data_files:
+      - split: train
+        path: "common_corpus_1/subset_100_1.parquet"
+---
+
+
+# Common Corpus
+
+<p align="center">
+  <a href="https://iclr.cc/virtual/2026/poster/10011885"><b>Full paper - ICLR 2026 oral</b></a>
+</p>
+
+Common Corpus is the largest open licensed text dataset, comprising 2.27 trillion tokens (2,267,302,720,836 tokens). It is a diverse dataset, consisting of books, newspapers, scientific articles, government and legal documents, code, and more. Common Corpus has been created by Pleias in association with several partners.
+
+Common Corpus differs from existing open datasets in that it is:
+*  **Truly Open**: contains only data that is either uncopyrighted or freely licensed
+*  **Traceable**: each individual document is associated with documented contextual information, including licensed use or lack of copyright.
+*  **Multilingual**: mostly representing English and French data, but contains data for 8 languages with more than 10 billion tokens (German, Spanish, Italian, Polish, Greek, Latin) and 33 languages with more than 1 billion tokens.
+*  **Diverse**: consisting of scientific articles, government and legal documents, code, and cultural heritage data, including books and newspapers
+*  **Extensively Curated**: spelling and formatting has been corrected from digitized texts, harmful and toxic content has been removed, and content with low educational content has also been removed.
+
+The dataset in its entirety meets the requirements of the Code of Conduct of the AI Act and goes further than the current requirements for data transparency. It aims to set a new standard of openness in AI, showing that detailed provenance at a granular document level is a realistic objective, even at the scale of 2.3 trillion tokens.
+
+Common Corpus makes it possible to train model compatible with [the Open Source Initiative’s definition](https://opensource.org/ai/open-source-ai-definition#:~:text=An%20Open%20Source%20AI%20is,including%20to%20change%20its%20output.) of open-source AI, which includes openness of use, meaning use is permitted for “any purpose and without having to ask for permission". Based on the available licensing information Common Corpus can be filtered to only include public domain works or a subset of free licenses (like attribution only).
+
+
+# About Common Corpus
+
+Common Corpus is made of six carefully curated collections:
+*  **OpenCulture**: our largest collection at 967,018,390,906 tokens, featuring public domain books, newspapers from cultural heritage repositories and open projets like Wikisource ad Gutenberg. We're developing innovative tools of OCR correction based on Pleias Models to correct historical digitization errors, while implementing advanced toxicity filtering to ensure content meets modern ethical standards.
+*  **OpenGovernment**: 579,150,518,908 tokens of financial and legal documents, including Finance Commons (from sources like SEC and WTO) and Legal Commons (including Europarl, Caselaw Access Project, Chinese Case Law), providing enterprise-grade training data from regulatory bodies and administrative sources.
+*  **OpenSource**: 283,227,402,898 tokens of high-quality code in open source from GitHub, filtered using ArmoRM to ensure only the top 80% of submissions by quality rating are included.
+*  **OpenScience**: 281,193,563,789 tokens of academic content from Open Alex and other open science reposiories, processed using vision-language models to preserve crucial document structure and formatting.
+*  **OpenWeb**: 88,517,032,065 tokens from Wikipedia (official releases from the [Wikimedia Foundation](https://huggingface.co/datasets/wikimedia/wikipedia) on Huggingface), YouTube Commons and Stack-Exchange.
+*  **Open Semantic**: 67,958,671,827 tokens from Wikidata (official releases from the [Wikimedia Foundation](https://huggingface.co/datasets/wikimedia/wikipedia) on Huggingface). The data has been reprocessed thanks to support and help of Wikidata and Wikimedia Germany. It includes the transcriptions of all the semantic triplets into natural language statements in over 300 languages.
+
+| Collection 	| Domain               	| Sources                                                                               	|
+|----------------|--------------------------|-------------------------------------------------------------------------------------------|
+| OpenGovernment | legal and administrative | [Finance Commons](https://huggingface.co/collections/PleIAs/finance-commons-66925e1095c7fa6e6828e26c) (e.g. SEC, WTO) and Legal Commons (e.g. Europarl, Caselaw Access Project, Chinese CaseLaw) |
+| OpenCulture	| cultural heritage    	| public domain books and newspapers, Wikisource                                                    	|
+| OpenScience	| academic             	| OpenAlex                                                             	|
+| OpenWeb    	| web text             	| [YouTube Commons](https://huggingface.co/datasets/PleIAs/YouTube-Commons), MOSEL, Stack Exchange, CCCC                                                       	|
+| OpenSource 	| code                 	| GitHub                                                                                	|
+| OpenSemantic 	| Semantic data                 	| Wikidata                                                                                	|
+
+The first version of [Common Corpus](https://huggingface.co/datasets/PleIAs/common_corpus) was released in November of 2024.  The second version added Wikidata and detailed document-level information, including licensing and other core metadata whenever available. The third ongoing version dramatically expand the language coverage of Common Corpus beyond the US and Europe with the integration of large collection of documents in Chinese, Japanese, Arabic, Korean and Hindi.
+
+The dataset release is accompanied by a comprehensive technical report (ICRL 2026 - oral) detailing our methodologies and data sources will accompany the release, ensuring full transparency and reproducibility.
+
+## Dataset Structure
+
+<details >
+  <summary>Data Fields</summary>
+
+*  `identifier`: unique text identifier. In many cases, this is also the link to the original resources.
+*  `collection`: name of one of the XX sub-collections curated for Common corpus.
+*  `open type`: one of the six leading collection groupings:
+*  `license`: sharing rights for the content either uncopyrighted (public domain, US federal public domain, CC0 on Wikidata) or various free licenses (Creative Commons, MIT, French Licence ouverte, etc.)
+*  `date`: date of creation of the resource where known. Due to the significance of public domain and other cultural heritage content, more than half of Common Corpus predates the 21st century.
+*  `title`: title of the resource when known or alternatively the filename.
+*  `creator`: institution publishing/collecting/curating the resource.
+*  `language`: automatically identified language.
+*  `word_count`: number of space delimited words.
+*  `token_count`: number of tokens as calculated by Pleias official tokenizer and Gemma-3 tokenizer for Chinese, Japanese, Arabic, Korean and few additional non-Western languages.
+*  `text`: full text, without formatting.
+
+</details > 
+<br />
+
+
+## Provenance
+
+The provenance of the datasets that make up Refined Common Corpus is detailed in the technical report [link]. Additionally, the original source URL is available in the metadata for each document for most of the dataset. 
+
+
+## How to Use
+
+### Considerations for Using the Data
+
+All data in Common Corpus are either uncopyrighted or freely licensed and may be used for both commercial and non-commercial purposes. 
+
+The dataset is multilingual. The language text is included in the metadata, so data can be filtered by language. Additionally, some of the text data are historical. The year each text is written is included in the metadata, therefore it is possible to construct a dataset with a custom date cutoff if desired.
+
+### Discussion of Bias
+
+Some of the dataset sources contain biased and toxic content, such as stereotypes about certain minoritized groups. We have removed texts which had high toxicity scores according to our toxicity classifier, [Celadon](https://huggingface.co/PleIAs/celadon), or which contain offensive terms and slurs. See our [preprint](https://arxiv.org/pdf/2410.22587) for more details.
+
+### Personal and Sensitive Information
+
+We have attempted to remove personally identifiable information (PII). We primarily use [Microsoft Presidio](https://microsoft.github.io/presidio/), but make additional modifications to account for language- and country-specific considerations, such as European phone number formats. 
+
+Some small parts of the French administrative common crawl have been entirely dropped using our unreleased small reasoning model for GDPR-filtering, due to the heightened risk of transmitting identifiable indirect personal information.
+
+## Using Common Corpus
+from datasets import load_dataset
+data = load_dataset('PleIAs/common_corpus')
+
+
+# Acknowledgements
+
+The Corpus was built up with the support and concerted efforts of the AI Alliance, the French Ministry of Culture as part of the prefiguration of the service offering of the Alliance for Language technologies EDIC (ALT-EDIC). 
+
+This dataset was also made in partnership with Wikimedia Enterprise and Wikidata/Wikimedia Germany. We're also thankful to our partner Libraries Without Borders for continuous assistance on extending low resource language support.
+
+The corpus was stored and processed with the generous support of the AI Alliance, Jean Zay (Eviden, Idris), Tracto AI, Mozilla. Generation of OCR correction at scale were performed using HPC resources from two GENCI–IDRIS grants: 2023-AD011014736 and GC011015451.
+
+Some parts of the corpus have been built on top of other similar open science LLM community initiatives such as German-Commons, MOSEL, kl3m, AI4Bharat, Creative Commons Common Crawl. We included a new curator field to properly acknowledge this data work.
+
+<div style="text-align: center;">
+  <img src="https://huggingface.co/datasets/PleIAs/common_corpus/resolve/main/logo/ai_alliance.png" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+  <img src="https://huggingface.co/datasets/PleIAs/common_corpus/resolve/main/logo/logo-genci-header.svg" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+  <img src="https://huggingface.co/datasets/PleIAs/common_corpus/resolve/main/logo/Nvidia_(logo).svg.png" style="width: 33%; margin: 0 auto; display: inline-block;"/>   
+  <img src="https://huggingface.co/datasets/PleIAs/common_corpus/resolve/main/logo/tractoAI.png" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+  <img src="https://huggingface.co/datasets/PleIAs/common_corpus/resolve/main/logo/mozilla.png" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+  <img src="https://huggingface.co/datasets/PleIAs/common_corpus/resolve/main/logo/wikimedia_logo.png" style="width: 33%; margin: 0 auto; display: inline-block;"/>
+</div>

artifacts/hf_readmes/WiktorS__polish-news__README.md

@@ -0,0 +1,32 @@
+---
+license: apache-2.0
+task_categories:
+- text-classification
+- summarization
+- text-generation
+language:
+- pl
+size_categories:
+- 100K<n<1M
+---
+
+This dataset contains more than 250k articles obtained from polish news site `tvp.info.pl`.
+Main purpouse of collecting the data was to create a transformer-based model for text summarization.
+
+Columns:
+* `link` - link to article
+* `title` - original title of the article
+* `headline` - lead/headline of the article - first paragraph of the article visible directly from the page
+* `content` - full textual contents of the article
+
+
+
+Link to original repo: https://github.com/WiktorSob/scraper-tvp
+
+Download the data:
+
+```python
+from datasets import load_dataset
+
+dataset = load_dataset("WiktorS/polish-news")
+```

artifacts/hf_readmes/allegro__polish-question-passage-pairs__README.md

@@ -0,0 +1,3 @@
+README_DOWNLOAD_ERROR: RemoteEntryNotFoundError: 404 Client Error. (Request ID: Root=1-6a30c2e8-69709ad67822f4c61c22b757;e5d36c2d-c02b-4c78-ac85-86048ae92297)
+
+Entry Not Found for url: https://huggingface.co/datasets/allegro/polish-question-passage-pairs/resolve/main/README.md.

artifacts/hf_readmes/allegro__summarization-polish-summaries-corpus__README.md

@@ -0,0 +1,3 @@
+README_DOWNLOAD_ERROR: RemoteEntryNotFoundError: 404 Client Error. (Request ID: Root=1-6a30c2e8-2fb9443220aedb731cc645fc;f438cb69-dba3-430b-84e7-7f81180b703c)
+
+Entry Not Found for url: https://huggingface.co/datasets/allegro/summarization-polish-summaries-corpus/resolve/main/README.md.

artifacts/hf_readmes/clarin-knext__wsd_polish_datasets__README.md

@@ -0,0 +1,378 @@
+---
+annotations_creators:
+- expert-generated
+language:
+- pl
+language_creators:
+- expert-generated
+- found
+license:
+- cc-by-4.0
+multilinguality:
+- monolingual
+pretty_name: wsd-polish-datasets
+size_categories:
+- 1M<n<10M
+source_datasets:
+- original
+tags: []
+task_categories:
+- token-classification
+task_ids:
+- word-sense-disambiguation
+---
+# Word Sense Disambiguation Corpora for Polish
+
+## Table of Contents
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-fields)
+  - [Data Splits](#data-splits)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+  - [Contributions](#contributions)
+
+## Dataset Description
+
+- **Homepage:**
+- **Repository:**
+- **Paper:** https://link.springer.com/chapter/10.1007/978-3-031-08754-7_70
+- **Point of Contact:** arkadiusz.janz@pwr.edu.pl
+
+### Dataset Summary
+
+`WSD Polish Datasets` is a comprehensive benchmark for word sense disambiguation (WSD) classification task in Polish language.
+It consists of 7 distinct datasets, manually annotated with senses from plWordNet-4.5 sense inventory. The following datasets
+were annotated and included into our benchmark:
+- KPWr
+- KPWr-100
+- Sherlock (SPEC)
+- Skladnica
+- WikiGlex (a subset of GLEX corpus)
+- EmoGlex (a subset of GLEX corpus)
+- Walenty
+
+For more details, please check the following publication:
+
+```
+@InProceedings{10.1007/978-3-031-08754-7_70,
+  author="Janz, Arkadiusz
+  and Dziob, Agnieszka
+  and Oleksy, Marcin
+  and Baran, Joanna",
+  editor="Groen, Derek
+  and de Mulatier, Cl{\'e}llia
+  and Paszynski, Maciej
+  and Krzhizhanovskaya, Valeria V.
+  and Dongarra, Jack J.
+  and Sloot, Peter M. A.",
+  title="A Unified Sense Inventory for Word Sense Disambiguation in Polish",
+  booktitle="Computational Science -- ICCS 2022",
+  year="2022",
+  publisher="Springer International Publishing",
+  address="Cham",
+  pages="682--689",
+  isbn="978-3-031-08754-7"
+}
+```
+
+**A new publication on Polish WSD corpora will be available soon**
+
+### Supported Tasks and Leaderboards
+
+Word sense disambiguation task. We do not provide a leaderboard. However, we provide an example evaluation script for evaluating WSD models.
+
+### Languages
+
+Polish language, PL
+
+## Dataset Structure
+
+### Data Instances
+
+Data are structured in JSONL format, each single text sample is divided by sentence.
+
+```
+{
+  "text": "Wpierw pani Hudson została zerwana z łóżka, po czym odegrała się na mnie, a ja - na tobie.",
+  "tokens": [
+    {
+      "index": 0,
+      "position": [ 0, 6 ],
+      "orth": "Wpierw",
+      "lemma": "wpierw",
+      "pos": "adv",
+      "ctag": "adv"
+    },
+    {
+      "index": 1,
+      "position": [ 7, 11 ],
+      "orth": "pani",
+      "lemma": "pani",
+      "pos": "noun",
+      "ctag": "subst:nom:f:sg"
+    },
+    {
+      "index": 2,
+      "position": [ 12, 18 ],
+      "orth": "Hudson",
+      "lemma": "Hudson",
+      "pos": "noun",
+      "ctag": "subst:nom:f:sg"
+    },
+    {
+      "index": 3,
+      "position": [ 19, 26 ],
+      "orth": "została",
+      "lemma": "zostać",
+      "pos": "verb",
+      "ctag": "praet:perf:f:sg"
+    },
+    {
+      "index": 4,
+      "position": [ 27, 34 ],
+      "orth": "zerwana",
+      "lemma": "zerwać",
+      "pos": "verb",
+      "ctag": "ppas:perf:nom:f:aff:sg"
+    },
+   <...>
+  ],
+  "phrases": [
+    {
+      "indices": [ 10, 11 ],
+      "head": 10,
+      "lemma": "odegrać się"
+    }
+  ],
+  "wsd": [
+      {
+        "index": 0,
+        "pl_sense": "wpierw.1.r",
+        "plWN_syn_id": "01a4a067-aac5-11ed-aae5-0242ac130002",
+        "plWN_lex_id": "f2757c30-aac4-11ed-aae5-0242ac130002",
+        "plWN_syn_legacy_id": "477654",
+        "plWN_lex_legacy_id": "718454",
+        "PWN_syn_id": "00102736-r",
+        "bn_syn_id": "bn:00115376r",
+        "mapping_relation": "synonymy"
+      },
+      {
+        "index": 1,
+        "pl_sense": "pani.2.n",
+        "plWN_syn_id": "f35fb1ed-aac4-11ed-aae5-0242ac130002",
+        "plWN_lex_id": "d5145565-aac4-11ed-aae5-0242ac130002",
+        "plWN_syn_legacy_id": "129",
+        "plWN_lex_legacy_id": "20695",
+        "PWN_syn_id": "10787470-n",
+        "bn_syn_id": "bn:00001530n",
+        "mapping_relation": "synonymy"
+      },
+   <...>
+  ]
+}
+```
+
+### Data Fields
+
+Description of json keys:
+- `text`: text of the sentence
+- `tokens`: list of tokens made by tokenization process
+  - `index`: token order index in sentence
+  - `position`: token chars span indices <included, excluded>
+  - `orth`: word
+  - `lemma`: lemmatised word
+  - `pos`: part of speech
+  - `ctag`: morphosyntactic tag
+- `phrases`: list of multi-word 
+- `wsd`: annotation labels for the WSD task
+
+### Data Splits
+
+We do not specify an exact data split for training and evaluation. However, we suggest to use GLEX and Składnica for training and other datasets for testing.
+
+## Dataset Creation
+
+### Curation Rationale
+
+[More Information Needed]
+
+### Source Data
+
+#### Initial Data Collection, Normalization and Post-processing
+
+Source corpora were initially pre-processed using morphosyntactic tagging and multi-word expression recognition tools.
+To tokenize and tag the datasets we used [MorphoDiTa](https://clarin-pl.eu/dspace/handle/11321/425) adapted to Polish language. To recognize multi-word expressions
+we applied pattern-based matching tool [Corpus2-MWE](https://clarin-pl.eu/dspace/handle/11321/533) - only MWEs from plWordNet were included. After manual annotation, 
+sense indices of plWordNet 4.5 were mapped automatically to Princeton WordNet 3.0 and BabelNet 4.0 indices using plWordNet's interlingual mapping.
+
+### Annotations
+
+#### Annotation process
+
+* 2+1 annotation process with inter-annotator agreement score over 0.6 PSA
+* annotated with [plWordNet 4.5](http://plwordnet.pwr.wroc.pl/wordnet/)
+* software: [WordNet-Loom](https://clarin-pl.eu/dspace/handle/11321/275) and [Inforex](https://clarin-pl.eu/dspace/handle/11321/13)
+* both single-word and multi-word expressions annotated
+* full-text sense annotation (excluding KPWr)
+
+#### Who are the annotators?
+
+- professional linguists from CLARIN-PL project
+
+### Personal and Sensitive Information
+
+The datasets do not contain any personal or sensitive information.
+
+## Considerations for Using the Data
+
+### Discussion of Biases
+
+Some datasets are biased towards most frequent senses. No information about other biases - needs further analysis.
+
+### Other Known Limitations
+
+* sense inventories are usually incomplete therefore some word senses might be missing in plWordNet
+* single-word and multi-word terms expressing novel senses (missing in plWordNet) were not marked
+
+## Additional Information
+
+### Dataset Curators
+
+Arkadiusz Janz (arkadiusz.janz@pwr.edu.pl)
+
+### Licensing Information
+
+KPWR-100 [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)  
+KPWR [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)  
+Walenty [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)  
+Sherlock [CC-BY 4.0](https://creativecommons.org/licenses/by/4.0/)  
+Skladnica [GNU GPL 3](http://www.gnu.org/licenses/gpl-3.0.en.html)  
+GLEX [plWordNet License](http://plwordnet.pwr.wroc.pl/wordnet/licence)  
+
+
+### Citation Information
+
+Main source (all corpora as a unified benchmark) and published here on HuggingFace:
+
+````
+@InProceedings{10.1007/978-3-031-08754-7_70,
+  author="Janz, Arkadiusz
+  and Dziob, Agnieszka
+  and Oleksy, Marcin
+  and Baran, Joanna",
+  editor="Groen, Derek
+  and de Mulatier, Cl{\'e}llia
+  and Paszynski, Maciej
+  and Krzhizhanovskaya, Valeria V.
+  and Dongarra, Jack J.
+  and Sloot, Peter M. A.",
+  title="A Unified Sense Inventory for Word Sense Disambiguation in Polish",
+  booktitle="Computational Science -- ICCS 2022",
+  year="2022",
+  publisher="Springer International Publishing",
+  address="Cham",
+  pages="682--689",
+  isbn="978-3-031-08754-7"
+}
+
+````
+
+Related work
+------------
+
+KPWr-100, Składnica, SPEC
+````
+@article{janzresults,
+  title={Results of the PolEval 2020 Shared Task 3: Word Sense Disambiguation},
+  author={Janz, Arkadiusz and Chlebus, Joanna and Dziob, Agnieszka and Piasecki, Maciej},
+  journal={Proceedings of the PolEval 2020 Workshop},
+  pages={65--77},
+  year={2020}
+}
+````
+
+GLEX (EmoGLEX)
+
+````
+@article{janz2017plwordnet,
+  title={{plWordNet} as a basis for large emotive lexicons of Polish},
+  author={Janz, Arkadiusz and Kocon, Jan and Piasecki, Maciej and Zasko-Zielinska, Monika},
+  journal={Proceedings of Human Language Technologies as a Challenge for Computer Science and Linguistics Poznan: Fundacja Uniwersytetu im. Adama Mickiewicza w Poznaniu},
+  pages={189--193},
+  year={2017}
+}
+````
+
+KPWr
+````
+@conference{broda2012,
+	address = {Istanbul, Turkey},
+	author = {Bartosz Broda and Micha{\l} Marci{\'n}czuk and Marek Maziarz and Adam Radziszewski and Adam Wardy{\'n}ski},
+	booktitle = {Proceedings of LREC'12},
+	owner = {Marlena},
+	publisher = {ELRA},
+	timestamp = {2014.06.20},
+	title = {KPWr: Towards a Free Corpus of Polish},
+	year = {2012}
+}
+````
+
+Składnica
+````
+@inproceedings{hajnicz-2014-lexico,
+    title = "Lexico-Semantic Annotation of Sk{\l}adnica Treebank by means of {PLWN} Lexical Units",
+    author = "Hajnicz, El{\.z}bieta",
+    booktitle = "Proceedings of the Seventh Global {W}ordnet Conference",
+    month = jan,
+    year = "2014",
+    address = "Tartu, Estonia",
+    publisher = "University of Tartu Press",
+    url = "https://aclanthology.org/W14-0104",
+    pages = "23--31",
+}
+````
+
+Walenty
+````
+@inproceedings{haj:and:bar:lrec16,
+	author = {Hajnicz, El{\.z}bieta and Andrzejczuk, Anna and Bartosiak, Tomasz},
+	crossref = {lrec:16},
+	pages = {2625--2632},
+	pdf = {http://www.lrec-conf.org/proceedings/lrec2016/pdf/382_Paper.pdf},
+	title = {Semantic Layer of the Valence Dictionary of {P}olish \emph{{W}alenty}}
+}
+````
+
+Mapping plWordNet onto Princeton WordNet
+````
+@inproceedings{rudnicka-etal-2021-non,
+    title = "A (Non)-Perfect Match: Mapping pl{W}ord{N}et onto {P}rinceton{W}ord{N}et",
+    author = "Rudnicka, Ewa  and
+      Witkowski, Wojciech  and
+      Piasecki, Maciej",
+    booktitle = "Proceedings of the 11th Global Wordnet Conference",
+    month = jan,
+    year = "2021",
+    address = "University of South Africa (UNISA)",
+    publisher = "Global Wordnet Association",
+    url = "https://aclanthology.org/2021.gwc-1.16",
+    pages = "137--146"
+}
+````
+

artifacts/hf_readmes/clarin-pl__ComplexQA__README.md

@@ -0,0 +1,29 @@
+---
+license: cc-by-sa-4.0
+task_categories:
+- question-answering
+language:
+- pl
+pretty_name: 'ComplexQA: Complex Question Answering on Long Documents in Polish Language'
+---
+
+# Complex Question Answering dataset
+
+Part of publication: Towards Complex Question Answering on Long Documents in Polish Language
+
+
+```
+@InProceedings{10.1007/978-3-032-09318-9_18,
+author="Wojtasik, Konrad
+and Domaga{\l}a, Aleksandra
+and Oleksy, Marcin
+and Piasecki, Maciej",
+title="Towards Complex Question Answering in Polish Language",
+booktitle="Computational Collective Intelligence",
+year="2026",
+publisher="Springer Nature Switzerland",
+address="Cham",
+pages="256--268",
+isbn="978-3-032-09318-9"
+}
+```

artifacts/hf_readmes/clarin-pl__PUGG__README.md

@@ -0,0 +1,173 @@
+---
+annotations_creators:
+- expert-generated
+language_creators: []
+language:
+- pl
+license:
+- cc-by-sa-4.0
+multilinguality:
+- monolingual
+size_categories:
+- 1K<n<10K
+- 10K<n<100K
+source_datasets:
+- original
+task_categories:
+- question-answering
+- text-retrieval
+task_ids:
+- extractive-qa
+- document-retrieval
+pretty_name: 'PUGG: KBQA, MRC, IR dataset for Polish'
+tags:
+- knowledge graph
+- KBQA
+- wikipedia
+- wikidata
+configs:
+- config_name: kbqa_all
+  data_files:
+  - split: train
+    path: kbqa/*/train.jsonl
+  - split: test
+    path: kbqa/*/test.jsonl
+- config_name: kbqa_natural
+  data_files:
+  - split: train
+    path: kbqa/natural/train.jsonl
+  - split: test
+    path: kbqa/natural/test.jsonl
+- config_name: kbqa_template-based
+  data_files:
+  - split: train
+    path: kbqa/template-based/train.jsonl
+  - split: test
+    path: kbqa/template-based/test.jsonl
+- config_name: mrc
+  data_files:
+  - split: train
+    path: mrc/train.jsonl
+  - split: test
+    path: mrc/test.jsonl
+- config_name: ir_corpus
+  data_files:
+  - split: test
+    path: ir/corpus.jsonl
+- config_name: ir_queries
+  data_files:
+  - split: test
+    path: ir/queries.jsonl
+- config_name: ir_qrels
+  data_files:
+  - split: test
+    path: ir/qrels/test.jsonl
+---
+# PUGG: KBQA, MRC, IR Dataset for Polish
+
+## Description
+
+This repository contains the PUGG dataset designed for three NLP tasks in the Polish language:
+
+- KBQA (Knowledge Base Question Answering)
+- MRC (Machine Reading Comprehension)
+- IR (Information Retrieval)
+
+## Paper
+
+For more detailed information, please refer to our research paper titled:
+
+**"Developing PUGG for Polish: A Modern Approach to KBQA, MRC, and IR Dataset Construction"** 
+
+Authored by:
+* Albert Sawczyn
+* Katsiaryna Viarenich
+* Konrad Wojtasik
+* Aleksandra Domogała
+* Marcin Oleksy
+* Maciej Piasecki
+* Tomasz Kajdanowicz
+  
+**The paper was accepted for ACL 2024 (findings).**
+
+## Repositories
+
+The dataset is available in the following repositories:
+
+* [General](https://huggingface.co/datasets/clarin-pl/PUGG) **(this repository)** - contains all tasks (KBQA, MRC, IR*)
+
+For more straightforward usage, the tasks are also available in separate repositories:
+
+* [KBQA](https://huggingface.co/datasets/clarin-pl/PUGG_KBQA) 
+* [MRC](https://huggingface.co/datasets/clarin-pl/PUGG_MRC) 
+* [IR](https://huggingface.co/datasets/clarin-pl/PUGG_IR) 
+
+The knowledge graph for KBQA task is available in the following repository:
+
+* [Knowledge Graph](https://huggingface.co/datasets/clarin-pl/PUGG_KG) 
+
+Note: If you want to utilize the IR task in the BEIR format (`qrels` in `.tsv` format), please 
+download the [IR](https://huggingface.co/datasets/clarin-pl/PUGG_IR) repository.
+
+## Links
+
+* Code:
+  * [Github](https://github.com/CLARIN-PL/PUGG)
+* Paper:
+  * ACL - TBA
+  * [Arxiv](https://arxiv.org/abs/2408.02337)
+
+## Citation
+
+```bibtex
+@misc{sawczyn2024developingpuggpolishmodern,
+      title={Developing PUGG for Polish: A Modern Approach to KBQA, MRC, and IR Dataset Construction}, 
+      author={Albert Sawczyn and Katsiaryna Viarenich and Konrad Wojtasik and Aleksandra Domogała and Marcin Oleksy and Maciej Piasecki and Tomasz Kajdanowicz},
+      year={2024},
+      eprint={2408.02337},
+      archivePrefix={arXiv},
+      primaryClass={cs.AI},
+      url={https://arxiv.org/abs/2408.02337}, 
+}
+```
+
+## Contact
+
+albert.sawczyn@pwr.edu.pl
+
+## Usage 
+
+```python
+from datasets import load_dataset
+
+# loading KBQA (all)
+dataset = load_dataset("clarin-pl/PUGG", "kbqa_all")
+print(dataset)
+
+# loading KBQA (natural)
+dataset = load_dataset("clarin-pl/PUGG", "kbqa_natural")
+print(dataset)
+
+# loading KBQA (template-based)
+dataset = load_dataset("clarin-pl/PUGG", "kbqa_template-based")
+print(dataset)
+
+# loading MRC
+
+dataset = load_dataset("clarin-pl/PUGG", "mrc")
+print(dataset)
+
+# loading IR
+
+## corpus
+dataset = load_dataset("clarin-pl/PUGG", "ir_corpus")
+print(dataset)
+
+## queries
+dataset = load_dataset("clarin-pl/PUGG", "ir_queries")
+print(dataset)
+
+## qrels
+dataset = load_dataset("clarin-pl/PUGG", "ir_qrels")
+print(dataset)
+```

artifacts/hf_readmes/clarin-pl__poquad__README.md

@@ -0,0 +1,24 @@
+---
+annotations_creators:
+- expert-generated
+language_creators:
+- found
+language:
+- pl
+license:
+- cc-by-4.0
+multilinguality:
+- monolingual
+pretty_name: PoQuaD
+size_categories:
+- 10K<n<100K
+source_datasets:
+- original
+task_categories:
+- question-answering
+task_ids:
+- extractive-qa
+- open-domain-qa
+---
+
+PoQuaD dataset

artifacts/hf_readmes/ipipan__polqa__README.md

@@ -0,0 +1,236 @@
+---
+task_categories:
+- question-answering
+- text-retrieval
+- text2text-generation
+task_ids:
+- open-domain-qa
+- document-retrieval
+- abstractive-qa
+language:
+- pl
+pretty_name: PolQA
+size_categories:
+- 10K<n<100K
+annotations_creators:
+- expert-generated
+license: cc-by-sa-4.0
+---
+
+# Dataset Card for PolQA Dataset
+
+## Dataset Description
+
+- **Paper:** [Improving Question Answering Performance through Manual Annotation: Costs, Benefits and Strategies](https://arxiv.org/abs/2212.08897)
+- **Point of Contact:** [Piotr Rybak](mailto:piotr.cezary.rybak@gmail.com)
+
+
+### Dataset Summary
+
+PolQA is the first Polish dataset for open-domain question answering. It consists of 7,000 questions, 87,525 manually labeled evidence passages, and a corpus of over 7 million candidate passages. The dataset can be used to train both a passage retriever and an abstractive reader.
+
+### Supported Tasks and Leaderboards
+
+- `open-domain-qa`: The dataset can be used to train a model for open-domain question answering. Success on this task is typically measured using [metric defined during PolEval 2021](https://2021.poleval.pl/tasks/task4).
+- `document-retrieval`: The dataset can be used to train a model for document retrieval. Success on this task is typically measured by [top-k retrieval accuracy](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.top_k_accuracy_score.html) or [NDCG](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.ndcg_score.html).
+- `abstractive-qa`: The dataset can be used to train a model for abstractive question answering. Success on this task is typically measured using [metric defined during PolEval 2021](https://2021.poleval.pl/tasks/task4).
+
+### Languages
+
+The text is in Polish, as spoken by the host of the [Jeden z Dziesięciu](https://pl.wikipedia.org/wiki/Jeden_z_dziesi%C4%99ciu) TV show (questions) and [Polish Wikipedia](https://pl.wikipedia.org/) editors (passages). The BCP-47 code for Polish is pl-PL.
+
+## Dataset Structure
+
+### Data Instances
+
+The main part of the dataset consists of manually annotated question-passage pairs. For each instance, there is a `question`, a passage (`passage_id`, `passage_title`, `passage_text`), and a boolean indicator if the passage is `relevant` for the given question (i.e. does it contain the answers).
+
+For each `question` there is a list of possible `answers` formulated in a natural language, in a way a Polish
+speaker would answer the questions. It means that the answers might
+contain prepositions, be inflected, and contain punctuation. In some
+cases, the answer might have multiple correct variants, e.g. numbers
+are written as numerals and words, synonyms, abbreviations and their
+expansions.
+
+Additionally, we provide a classification of each question-answer pair based on the `question_formulation`, the `question_type`, and the `entity_type/entity_subtype`, according to the taxonomy proposed by
+[Maciej Ogrodniczuk and Piotr Przybyła (2021)](http://nlp.ipipan.waw.pl/Bib/ogr:prz:21:poleval.pdf).
+
+```
+{
+  'question_id': 6,
+  'passage_title': 'Mumbaj',
+  'passage_text': 'Mumbaj lub Bombaj (marathi मुंबई, trb.: Mumbaj; ang. Mumbai; do 1995 Bombay) – stolica indyjskiego stanu Maharasztra, położona na wyspie Salsette, na Morzu Arabskim.',
+  'passage_wiki': 'Mumbaj lub Bombaj (mr. मुंबई, trb.: "Mumbaj"; ang. Mumbai; do 1995 Bombay) – stolica indyjskiego stanu Maharasztra, położona na wyspie Salsette, na Morzu Arabskim. Wraz z miastami satelitarnymi tworzy najludniejszą po Delhi aglomerację liczącą 23 miliony mieszkańców. Dzięki naturalnemu położeniu jest to największy port morski kraju. Znajdują się tutaj także najsilniejsze giełdy Azji Południowej: National Stock Exchange of India i Bombay Stock Exchange.',
+  'passage_id': '42609-0',
+  'duplicate': False,
+  'question': 'W którym państwie leży Bombaj?',
+  'relevant': True,
+  'annotated_by': 'Igor',
+  'answers': "['w Indiach', 'Indie']",
+  'question_formulation': 'QUESTION',
+  'question_type': 'SINGLE ENTITY',
+  'entity_type': 'NAMED',
+  'entity_subtype': 'COUNTRY',
+  'split': 'train',
+  'passage_source': 'human'
+}
+```
+
+The second part of the dataset is a corpus of Polish Wikipedia (March 2022 snapshot) passages. The raw Wikipedia snapshot was parsed using [WikiExtractor](https://github.com/attardi/wikiextractor) and split into passages at the ends of the paragraphs or if the passage was longer than 500 characters.
+
+```
+{
+  'id': '42609-0',
+  'title': 'Mumbaj',
+  'text': 'Mumbaj lub Bombaj (mr. मुंबई, trb.: "Mumbaj"; ang. Mumbai; do 1995 Bombay) – stolica indyjskiego stanu Maharasztra, położona na wyspie Salsette, na Morzu Arabskim. Wraz z miastami satelitarnymi tworzy najludniejszą po Delhi aglomerację liczącą 23 miliony mieszkańców. Dzięki naturalnemu położeniu jest to największy port morski kraju. Znajdują się tutaj także najsilniejsze giełdy Azji Południowej: National Stock Exchange of India i Bombay Stock Exchange.'
+}
+```
+
+### Data Fields
+
+Question-passage pairs:
+
+- `question_id`: an integer id of the question
+- `passage_title`: a string containing the title of the Wikipedia article
+- `passage_text`: a string containing the passage text as extracted by the human annotator
+- `passage_wiki`: a string containing the passage text as it can be found in the provided Wikipedia corpus. Empty if the passage doesn't exist in the corpus.
+- `passage_id`: a string containing the id of the passage from the provided Wikipedia corpus. Empty if the passage doesn't exist in the corpus.
+- `duplicate`: a boolean flag representing whether a question-passage pair is duplicated in the dataset. This occurs when the same passage was found in multiple passage sources.
+- `question`: a string containing the question
+- `relevant`: a boolean flag representing whether a passage is relevant to the question (i.e. does it contain the answers)
+- `annotated_by`: a string containing the name of the annotator who verified the relevance of the pair
+- `answers`: a string containing a list of possible short answers to the question
+- `question_formulation`: a string containing a kind of expression used to request information. One of the following:
+  - `QUESTION`, e.g. *What is the name of the first letter of the Greek alphabet?*
+  - `COMMAND`, e.g. *Expand the abbreviation ’CIA’.*
+  - `COMPOUND`, e.g. *This French writer, born in the 19th century, is
+  considered a pioneer of sci-fi literature. What is his name?*
+- `question_type`: a string indicating what type of information is sought by the question. One of the following:
+  - `SINGLE ENTITY`, e.g. *Who is the hero in the Tomb Rider video game series?*
+  - `MULTIPLE ENTITIES`, e.g. *Which two seas are linked by the Corinth Canal?*
+  - `ENTITY CHOICE`, e.g. *Is "Sombrero" a type of dance, a hat, or a dish?*
+  - `YES/NO`, e.g. *When the term of office of the Polish Sejm is terminated, does it apply to the Senate as well?*
+  - `OTHER NAME`, e.g. *What was the nickname of Louis I, the King of the Franks?*
+  - `GAP FILLING`, e.g. *Finish the proverb: "If you fly with the crows... ".*
+- `entity_type`: a string containing a type of the sought entity. One of the following: `NAMED`, `UNNAMED`, or `YES/NO`.
+- `entity_subtype`: a string containing a subtype of the sought entity. Can take one of the 34 different values.
+- `split`: a string containing the split of the dataset. One of the following: `train`, `valid`, or `test`.
+- `passage_source`: a string containing the source of the passage. One of the following: 
+  - `human`: the passage was proposed by a human annotator using any
+internal (i.e. Wikipedia search) or external (e.g. Google) search engines and any keywords or queries they considered useful
+  - `hard-negatives`: the passage was proposed using a neural retriever trained on the passages found by the human annotators
+  - `zero-shot`: the passage was proposed by the BM25 retriever and re-ranked using [multilingual cross-encoder](https://huggingface.co/unicamp-dl/mMiniLM-L6-v2-mmarco-v2)
+
+Corpus of passages:
+
+- `id`: a string representing the Wikipedia article id and the index of extracted passage. Matches the `passage_id` from the main part of the dataset.
+- `title`: a string containing the title of the Wikipedia article. Matches the `passage_title` from the main part of the dataset.
+- `text`: a string containing the passage text. Matches the `passage_wiki` from the main part of the dataset.
+
+### Data Splits
+
+The questions are assigned into one of three splits: `train`, `validation`, and `test`. The `validation` and `test` questions are randomly sampled from the `test-B` dataset from the [PolEval 2021](https://2021.poleval.pl/tasks/task4) competition.
+
+|            | # questions | # positive passages | # negative passages |
+|------------|------------:|--------------------:|--------------------:|
+| train      |       5,000 |              27,131 |              34,904 |
+| validation |       1,000 |               5,839 |               6,927 |
+| test       |       1,000 |               5,938 |               6,786 |
+
+
+## Dataset Creation
+
+### Curation Rationale
+
+The PolQA dataset was created to support and promote the research in the open-domain question answering for Polish. It also serves as a benchmark to evaluate OpenQA systems.
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+The majority of questions come from two existing resources, the
+6,000 questions from the [PolEval 2021 shared task on QA](https://2021.poleval.pl/tasks/task4) and additional 1,000 questions gathered by one of the shared
+task [participants](http://poleval.pl/files/poleval2021.pdf#page=151). Originally, the questions come from collections associated with TV shows, both officially published and gathered online by their fans, as well as questions used in actual quiz competitions, on TV or online.
+
+The evidence passages come from the Polish Wikipedia (March 2022 snapshot). The raw Wikipedia snapshot was parsed using [WikiExtractor](https://github.com/attardi/wikiextractor) and split into passages at the ends of the paragraphs or if the passage was longer than 500 characters.
+
+#### Who are the source language producers?
+
+The questions come from various sources and their authors are unknown but are mostly analogous (or even identical) to questions asked during the [Jeden z Dziesięciu](https://pl.wikipedia.org/wiki/Jeden_z_dziesi%C4%99ciu) TV show.
+
+The passages were written by the editors of the Polish Wikipedia.
+
+### Annotations
+
+#### Annotation process
+
+Two approaches were used to annotate the question-passage pairs. Each of them consists of two phases: the retrieval of candidate passages and the manual verification of their relevance.
+
+In the first approach, we asked annotators to use internal (i.e. Wikipedia search) or external (e.g. Google) search engines to find up to five relevant passages using any keywords or queries they consider useful  (`passage_source="human"`). Based on those passages, we trained the neural retriever to extend the number of relevant passages, as well as to retrieve the hard negatives (`passage_source="hard-negatives"`).
+
+In the second approach, the passage candidates were proposed by the BM25 retriever and re-ranked using [multilingual cross-encoder](https://huggingface.co/unicamp-dl/mMiniLM-L6-v2-mmarco-v2)  (`passage_source="zero-shot"`).
+
+In both cases, all proposed question-passage pairs were manually verified by the annotators.
+
+We release the annotation guidelines [here](https://docs.google.com/document/d/1LDW7EJFH0bm-FRlxM_uHb0mqJzKHiewOFBHe5qZnTW8/edit?usp=sharing).
+
+#### Who are the annotators?
+
+The annotation team consisted of 16 annotators, all native Polish
+speakers, most of them having linguistic backgrounds and previous
+experience as an annotator.
+
+### Personal and Sensitive Information
+
+The dataset does not contain any personal or sensitive information.
+
+## Considerations for Using the Data
+
+### Social Impact of Dataset
+
+This dataset was created to promote the research in the open-domain question answering for Polish and allow developing question answering systems.
+
+### Discussion of Biases
+
+The passages proposed by the `hard-negative` and `zero-shot` methods are bound to be easier to retrieve by retrievers since they were proposed by such. To mitigate this bias, we include the passages found by the human annotators in an unconstrained way (`passage_source="human"`). We hypothesize that it will result in more unbiased and diverse examples. Moreover, we asked the annotators to find not one but up to five passages, preferably from different articles to even further increase passage diversity.
+
+### Other Known Limitations
+
+The PolQA dataset focuses on trivia questions which might limit its usefulness in real-world applications since neural retrievers generalize poorly to other domains.
+
+## Additional Information
+
+### Dataset Curators
+
+The PolQA dataset was developed by Piotr Rybak, Piotr Przybyła, and Maciej Ogrodniczuk from the [Institute of Computer Science, Polish Academy of Sciences](http://zil.ipipan.waw.pl/).
+
+This work was supported by the European Regional Development Fund as a part of 2014–2020 Smart Growth Operational Programme, CLARIN — Common Language Resources and Technology Infrastructure, project no. POIR.04.02.00-00C002/19.
+
+### Licensing Information
+
+CC BY-SA 4.0
+
+### Citation Information
+
+```
+@inproceedings{rybak-etal-2024-polqa-polish,
+    title = "{P}ol{QA}: {P}olish Question Answering Dataset",
+    author = "Rybak, Piotr  and
+      Przyby{\l}a, Piotr  and
+      Ogrodniczuk, Maciej",
+    editor = "Calzolari, Nicoletta  and
+      Kan, Min-Yen  and
+      Hoste, Veronique  and
+      Lenci, Alessandro  and
+      Sakti, Sakriani  and
+      Xue, Nianwen",
+    booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)",
+    month = may,
+    year = "2024",
+    address = "Torino, Italia",
+    publisher = "ELRA and ICCL",
+    url = "https://aclanthology.org/2024.lrec-main.1125",
+    pages = "12846--12855",
+    abstract = "Recently proposed systems for open-domain question answering (OpenQA) require large amounts of training data to achieve state-of-the-art performance. However, data annotation is known to be time-consuming and therefore expensive to acquire. As a result, the appropriate datasets are available only for a handful of languages (mainly English and Chinese). In this work, we introduce and publicly release PolQA, the first Polish dataset for OpenQA. It consists of 7,000 questions, 87,525 manually labeled evidence passages, and a corpus of over 7,097,322 candidate passages. Each question is classified according to its formulation, type, as well as entity type of the answer. This resource allows us to evaluate the impact of different annotation choices on the performance of the QA system and propose an efficient annotation strategy that increases the passage retrieval accuracy@10 by 10.55 p.p. while reducing the annotation cost by 82{\%}.",
+}
+```

artifacts/hf_readmes/michaljunczyk__pl-asr-bigos__README.md

@@ -0,0 +1,210 @@
+---
+annotations_creators:
+- crowdsourced
+- expert-generated
+- other
+- machine-generated
+language:
+- pl
+language_creators:
+- crowdsourced
+- expert-generated
+- other
+license:
+- cc-by-sa-4.0
+multilinguality:
+- monolingual
+pretty_name: pl-asr-bigos
+size_categories:
+- 1K<n<10K
+source_datasets:
+- original
+- extended|librispeech_asr
+- extended|common_voice
+tags:
+- benchmark
+- polish
+- asr
+- speech
+task_categories:
+- automatic-speech-recognition
+task_ids: []
+extra_gated_prompt: |-
+    Original datasets used for curation of BIGOS have specific terms of usage that must be understood and agreed to before use. Below are the links to the license terms and datasets the specific license type applies to:
+    * [Creative Commons 0](https://creativecommons.org/share-your-work/public-domain/cc0) which applies to [Common Voice](https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0)
+    * [Creative Commons By Attribution Share Alike 4.0](https://creativecommons.org/licenses/by-sa/4.0/), which applies to [Clarin Cyfry](https://clarin-pl.eu/dspace/handle/11321/317), [Azon acoustic speech resources corpus](https://zasobynauki.pl/zasoby/korpus-nagran-probek-mowy-do-celow-budowy-modeli-akustycznych-dla-automatycznego-rozpoznawania-mowy,53293/).
+    * [Creative Commons By Attribution 3.0](https://creativecommons.org/licenses/by/3.0/), which applies to [CLARIN Mobile database](https://clarin-pl.eu/dspace/handle/11321/237), [CLARIN Studio database](https://clarin-pl.eu/dspace/handle/11321/236), [PELCRA Spelling and Numbers Voice Database](http://pelcra.pl/new/snuv) and [FLEURS dataset](https://huggingface.co/datasets/google/fleurs)
+    * [Creative Commons By Attribution 4.0](https://creativecommons.org/licenses/by/4.0/), which applies to [Multilingual Librispeech](https://huggingface.co/datasets/facebook/multilingual_librispeech) and [Poly AI Minds 14](https://huggingface.co/datasets/PolyAI/minds14)
+    * [Proprietiary License of Munich AI Labs dataset](https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset)
+    * Public domain mark, which applies to [PWR datasets](https://www.ii.pwr.edu.pl/~sas/ASR/)
+    To use selected dataset, you also need to fill in the access forms on the specific datasets pages:
+    * Common Voice: https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0
+
+extra_gated_fields:
+ I hereby confirm that I have read and accepted the license terms of datasets comprising BIGOS corpora: checkbox
+ I hereby confirm that I have registered on the original Common Voice page and agree to not attempt to determine the identity of speakers in the Common Voice dataset: checkbox
+---
+
+# Dataset Card for Polish ASR BIGOS corpora
+
+## Table of Contents
+- [Table of Contents](#table-of-contents)
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-fields)
+  - [Data Splits](#data-splits)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+  - [Contributions](#contributions)
+
+## Dataset Description
+
+- **Homepage:** https://huggingface.co/datasets/michaljunczyk/pl-asr-bigos
+- **Repository:** https://github.com/goodmike31/pl-asr-bigos-tools
+- **Paper:** https://annals-csis.org/proceedings/2023/drp/1609.html
+- **Leaderboard:** https://huggingface.co/spaces/michaljunczyk/pl-asr-bigos-benchmark
+- **Point of Contact:** michal.junczyk@amu.edu.pl
+
+### Dataset Summary
+
+The BIGOS (Benchmark Intended Grouping of Open Speech) corpora aims at simplifying the access and use of publicly available ASR speech datasets for Polish.<br>
+The initial release consist of test split with 1900 recordings and original transcriptions extracted from 10 publicly available datasets.
+
+### Supported Tasks and Leaderboards
+The leaderboard with benchmark of publicly available ASR systems supporting Polish is [under construction](https://huggingface.co/spaces/michaljunczyk/pl-asr-bigos-benchmark/).<br>
+Evaluation results of 3 commercial and 5 freely available can be found in the [paper](https://annals-csis.org/proceedings/2023/drp/1609.html).
+
+### Languages
+Polish
+
+## Dataset Structure
+Dataset consists audio recordings in WAV format and corresponding metadata.<br>
+Audio and metadata can be used in raw format (TSV) or via hugging face datasets library.
+
+
+### Data Instances
+1900 audio files with original transcriptions are available in "test" split.<br>
+This consitutes 1.6% of the total available transcribed speech in 10 source datasets considered in the initial release.
+
+### Data Fields
+Available fields:
+* file_id - file identifier
+* dataset_id - source dataset identifier
+* audio - binary representation of audio file
+* ref_original - original transcription of audio file
+* hyp_whisper_cloud - ASR hypothesis (output) from Whisper Cloud system
+* hyp_google_default - ASR hypothesis (output) from Google ASR system, default model
+* hyp_azure_default - ASR hypothesis (output) from Azure ASR system, default model
+* hyp_whisper_tiny - ASR hypothesis (output) from Whisper tiny model
+* hyp_whisper_base - ASR hypothesis (output) from Whisper base model
+* hyp_whisper_small - ASR hypothesis (output) from Whisper small model
+* hyp_whisper_medium - ASR hypothesis (output) from Whisper medium model
+* hyp_whisper_large - ASR hypothesis (output) from Whisper large (V2) model
+<br><br>
+
+Fields to be added in the next release:
+* ref_spoken - manual transcription in a spoken format (without normalization)
+* ref_written - manual transcription in a written format (with normalization)
+
+
+### Data Splits
+Initial release contains only "test" split.<br>
+"Dev" and "train" splits will be added in the next release.
+
+## Dataset Creation
+
+### Curation Rationale
+[Polish ASR Speech Data Catalog](https://github.com/goodmike31/pl-asr-speech-data-survey) was used to identify suitable datasets which can be repurposed and included in the BIGOS corpora.<br>
+The following mandatory criteria were considered:
+* Dataset must be downloadable.
+* The license must allow for free, noncommercial use.
+* Transcriptions must be available and align with the recordings.
+* The sampling rate of audio recordings must be at least 8 kHz.
+* Audio encoding using a minimum of 16 bits per sample.
+
+### Source Data
+10 datasets that meet the criteria were chosen as sources for the BIGOS dataset.
+* The Common Voice dataset (mozilla-common-voice-19)
+* The Multilingual LibriSpeech (MLS) dataset (fair-mls-20)
+* The Clarin Studio Corpus (clarin-pjatk-studio-15)
+* The Clarin Mobile Corpus (clarin-pjatk-mobile-15)
+* The Jerzy Sas PWR datasets from Politechnika Wrocławska (pwr-viu-unk, pwr-shortwords-unk, pwr-maleset-unk). More info [here](https://www.ii.pwr.edu.pl/)
+* The Munich-AI Labs Speech corpus (mailabs-19)
+* The AZON Read and Spontaneous Speech Corpora (pwr-azon-spont-20, pwr-azon-read-20) More info [here](https://zasobynauki.pl/zasoby/korpus-nagran-probek-mowy-do-celow-budowy-modeli-akustycznych-dla-automatycznego-rozpoznawania-mowy)
+
+#### Initial Data Collection and Normalization
+Source text and audio files were extracted and encoded in a unified format.<br>
+Dataset-specific transcription norms are preserved, including punctuation and casing. <br>
+To strike a balance in the evaluation dataset and to facilitate the comparison of Word Error Rate (WER) scores across multiple datasets, 200 samples are randomly selected from each corpus. <br>
+The only exception is ’pwr-azon-spont-20’, which contains significantly longer recordings and utterances, therefore only 100 samples are selected. <br>
+#### Who are the source language producers?
+1. Clarin corpora - Polish Japanese Academy of Technology
+2. Common Voice - Mozilla foundation
+3. Multlingual librispeech - Facebook AI research lab
+4. Jerzy Sas and AZON datasets - Politechnika Wrocławska
+
+Please refer to the [paper](https://www.researchgate.net/publication/374713542_BIGOS_-_Benchmark_Intended_Grouping_of_Open_Speech_Corpora_for_Polish_Automatic_Speech_Recognition) for more details.
+
+### Annotations
+
+#### Annotation process
+
+Current release contains original transcriptions.
+Manual transcriptions are planned for subsequent releases.
+
+#### Who are the annotators?
+Depends on the source dataset.
+
+### Personal and Sensitive Information
+This corpus does not contain PII or Sensitive Information.
+All IDs pf speakers are anonymized.
+
+## Considerations for Using the Data
+
+### Social Impact of Dataset
+To be updated.
+### Discussion of Biases
+To be updated.
+
+### Other Known Limitations
+The dataset in the initial release contains only a subset of recordings from original datasets.
+
+## Additional Information
+
+### Dataset Curators
+Original authors of the source datasets - please refer to [source-data](#source-data) for details.
+
+Michał Junczyk (michal.junczyk@amu.edu.pl) - curator of BIGOS corpora.
+
+### Licensing Information
+The BIGOS corpora is available under [Creative Commons By Attribution Share Alike 4.0 license.](https://creativecommons.org/licenses/by-sa/4.0/)
+
+Original datasets used for curation of BIGOS have specific terms of usage that must be understood and agreed to before use. Below are the links to the license terms and datasets the specific license type applies to:
+* [Creative Commons 0](https://creativecommons.org/share-your-work/public-domain/cc0) which applies to [Common Voice](https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0)
+* [Creative Commons By Attribution Share Alike 4.0](https://creativecommons.org/licenses/by-sa/4.0/), which applies to [Clarin Cyfry](https://clarin-pl.eu/dspace/handle/11321/317), [Azon acoustic speech resources corpus](https://zasobynauki.pl/zasoby/korpus-nagran-probek-mowy-do-celow-budowy-modeli-akustycznych-dla-automatycznego-rozpoznawania-mowy,53293/).
+* [Creative Commons By Attribution 3.0](https://creativecommons.org/licenses/by/3.0/), which applies to [CLARIN Mobile database](https://clarin-pl.eu/dspace/handle/11321/237), [CLARIN Studio database](https://clarin-pl.eu/dspace/handle/11321/236), [PELCRA Spelling and Numbers Voice Database](http://pelcra.pl/new/snuv) and [FLEURS dataset](https://huggingface.co/datasets/google/fleurs)
+* [Creative Commons By Attribution 4.0](https://creativecommons.org/licenses/by/4.0/), which applies to [Multilingual Librispeech](https://huggingface.co/datasets/facebook/multilingual_librispeech) and [Poly AI Minds 14](https://huggingface.co/datasets/PolyAI/minds14)
+* [Proprietiary License of Munich AI Labs dataset](https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset)
+* Public domain mark, which applies to [PWR datasets](https://www.ii.pwr.edu.pl/~sas/ASR/)
+
+### Citation Information
+Please cite [BIGOS V1 paper](https://annals-csis.org/proceedings/2023/drp/1609.html).
+
+### Contributions
+
+Thanks to [@goodmike31](https://github.com/goodmike31) for adding this dataset.

artifacts/hf_readmes/openlanguagedata__flores_plus__README.md

@@ -0,0 +1,2016 @@
+---
+annotations_creators:
+- found
+language_creators:
+- expert-generated
+language:
+- ace
+- acm
+- acq
+- aeb
+- af
+- ajp
+- ak
+- als
+- am
+- apc
+- ar
+- ars
+- ary
+- arz
+- as
+- ast
+- awa
+- ayr
+- azb
+- azj
+- ba
+- bm
+- ban
+- be
+- bem
+- bn
+- bho
+- bjn
+- bo
+- bs
+- bug
+- bg
+- ca
+- ceb
+- cs
+- cjk
+- ckb
+- crh
+- cy
+- da
+- de
+- dar
+- dik
+- dyu
+- dz
+- el
+- en
+- eo
+- et
+- eu
+- ee
+- fo
+- fj
+- fi
+- fon
+- fr
+- fur
+- fuv
+- gaz
+- gd
+- ga
+- gl
+- gn
+- gu
+- ht
+- ha
+- he
+- hi
+- hne
+- hr
+- hu
+- hy
+- ig
+- ilo
+- id
+- is
+- it
+- jv
+- ja
+- kab
+- kac
+- kam
+- kn
+- ks
+- ka
+- kk
+- kbp
+- kea
+- khk
+- km
+- ki
+- rw
+- kjh
+- ky
+- kmb
+- kmr
+- knc
+- kg
+- ko
+- lo
+- lij
+- li
+- lld
+- ln
+- lt
+- lmo
+- ltg
+- lb
+- lua
+- lg
+- luo
+- lus
+- lvs
+- mag
+- mai
+- ml
+- mar
+- mfe
+- mhr
+- min
+- mk
+- mt
+- mni
+- mos
+- mi
+- my
+- nl
+- nn
+- nb
+- npi
+- nso
+- nus
+- ny
+- oc
+- ory
+- pag
+- pa
+- pap
+- pbt
+- pes
+- plt
+- pl
+- pt
+- prs
+- quy
+- ro
+- rn
+- ru
+- sg
+- sa
+- sat
+- scn
+- shn
+- si
+- sk
+- sl
+- sm
+- sn
+- sd
+- so
+- st
+- es
+- sc
+- sr
+- ss
+- su
+- sv
+- swh
+- szl
+- ta
+- taq
+- tt
+- te
+- tg
+- tl
+- th
+- ti
+- tpi
+- tn
+- ts
+- tk
+- tum
+- tr
+- tw
+- tzm
+- udm
+- ug
+- uk
+- umb
+- ur
+- uzn
+- uzs
+- vec
+- vi
+- war
+- wo
+- xh
+- ydd
+- yo
+- yue
+- zgh
+- zh
+- zsm
+- zu
+license:
+- cc-by-sa-4.0
+multilinguality:
+- multilingual
+- translation
+size_categories:
+- unknown
+source_datasets:
+- extended|flores
+task_categories:
+- text-generation
+- translation
+task_ids: []
+pretty_name: flores+
+language_details: ace_Arab, ace_Latn, acm_Arab, acq_Arab, aeb_Arab, afr_Latn, ajp_Arab,
+  aka_Latn, amh_Ethi, apc_Arab, arb_Arab, ars_Arab, ary_Arab, arz_Arab, asm_Beng,
+  ast_Latn, awa_Deva, ayr_Latn, azb_Arab, azj_Latn, bak_Cyrl, bam_Latn, ban_Latn, bel_Cyrl,
+  bem_Latn, ben_Beng, bho_Deva, bjn_Arab, bjn_Latn, bod_Tibt, bos_Latn, bug_Latn,
+  bul_Cyrl, cat_Latn, ceb_Latn, ces_Latn, cjk_Latn, ckb_Arab, crh_Latn, cym_Latn,
+  dan_Latn, deu_Latn, dar_Cyrl, dik_Latn, dyu_Latn, dzo_Tibt, ell_Grek, eng_Latn, epo_Latn,
+  est_Latn, eus_Latn, ewe_Latn, fao_Latn, pes_Arab, fij_Latn, fin_Latn, fon_Latn,
+  fra_Latn, fur_Latn, fuv_Latn, gla_Latn, gle_Latn, glg_Latn, grn_Latn, guj_Gujr,
+  hat_Latn, hau_Latn, heb_Hebr, hin_Deva, hne_Deva, hrv_Latn, hun_Latn, hye_Armn,
+  ibo_Latn, ilo_Latn, ind_Latn, isl_Latn, ita_Latn, jav_Latn, jpn_Jpan, kab_Latn,
+  kac_Latn, kam_Latn, kan_Knda, kas_Arab, kas_Deva, kat_Geor, knc_Arab, knc_Latn,
+  kaz_Cyrl, kbp_Latn, kea_Latn, khm_Khmr, kik_Latn, kin_Latn, kir_Cyrl, kjh_Cyrl, kmb_Latn,
+  kon_Latn, kor_Hang, kmr_Latn, lao_Laoo, lvs_Latn, lij_Latn, lim_Latn, lin_Latn,
+  lit_Latn, lld_Latn, lmo_Latn, ltg_Latn, ltz_Latn, lua_Latn, lug_Latn, luo_Latn, lus_Latn,
+  mag_Deva, mai_Deva, mal_Mlym, mar_Deva, mfe_Latn, mhr_Cyrl, min_Latn, mkd_Cyrl, plt_Latn, mlt_Latn,
+  mni_Beng, khk_Cyrl, khk_Mong, mos_Latn, mri_Latn, zsm_Latn, mya_Mymr, nld_Latn, nno_Latn,
+  nob_Latn, npi_Deva, nso_Latn, nus_Latn, nya_Latn, oci_Latn, gaz_Latn, ory_Orya,
+  pag_Latn, pan_Guru, pap_Latn, pol_Latn, por_Latn, prs_Arab, pbt_Arab, quy_Latn,
+  ron_Latn, run_Latn, rus_Cyrl, sag_Latn, san_Deva, sat_Beng, scn_Latn, shn_Mymr,
+  sin_Sinh, slk_Latn, slv_Latn, smo_Latn, sna_Latn, snd_Arab, som_Latn, sot_Latn,
+  spa_Latn, als_Latn, srd_Latn, srp_Cyrl, ssw_Latn, sun_Latn, swe_Latn, swh_Latn,
+  szl_Latn, tam_Taml, tat_Cyrl, tel_Telu, tgk_Cyrl, tgl_Latn, tha_Thai, tir_Ethi,
+  taq_Latn, taq_Tfng, tpi_Latn, tsn_Latn, tso_Latn, tuk_Latn, tum_Latn, tur_Latn,
+  twi_Latn, tzm_Tfng, udm_Cyrl, uig_Arab, ukr_Cyrl, umb_Latn, urd_Arab, uzn_Latn, uzs_Arab, vec_Latn,
+  vie_Latn, war_Latn, wol_Latn, xho_Latn, ydd_Hebr, yor_Latn, yue_Hant, zgh_Tfng, zho_Hans,
+  zho_Hant, zul_Latn
+tags:
+- text
+configs:
+- config_name: default
+  data_files:
+  - split: dev
+    path: "dev/*.jsonl"
+  - split: devtest
+    path: "devtest/*.jsonl"
+# all configs below were generated automatically based on the filenames of the dataset contents
+- config_name: ace_Arab
+  data_files:
+  - path: dev/ace_Arab.jsonl
+    split: dev
+  - path: devtest/ace_Arab.jsonl
+    split: devtest
+- config_name: ace_Latn
+  data_files:
+  - path: dev/ace_Latn.jsonl
+    split: dev
+  - path: devtest/ace_Latn.jsonl
+    split: devtest
+- config_name: acm_Arab
+  data_files:
+  - path: dev/acm_Arab.jsonl
+    split: dev
+  - path: devtest/acm_Arab.jsonl
+    split: devtest
+- config_name: acq_Arab
+  data_files:
+  - path: dev/acq_Arab.jsonl
+    split: dev
+  - path: devtest/acq_Arab.jsonl
+    split: devtest
+- config_name: aeb_Arab
+  data_files:
+  - path: dev/aeb_Arab.jsonl
+    split: dev
+  - path: devtest/aeb_Arab.jsonl
+    split: devtest
+- config_name: afr_Latn
+  data_files:
+  - path: dev/afr_Latn.jsonl
+    split: dev
+  - path: devtest/afr_Latn.jsonl
+    split: devtest
+- config_name: als_Latn
+  data_files:
+  - path: dev/als_Latn.jsonl
+    split: dev
+  - path: devtest/als_Latn.jsonl
+    split: devtest
+- config_name: amh_Ethi
+  data_files:
+  - path: dev/amh_Ethi.jsonl
+    split: dev
+  - path: devtest/amh_Ethi.jsonl
+    split: devtest
+- config_name: apc_Arab_nort3139
+  data_files:
+  - path: dev/apc_Arab_nort3139.jsonl
+    split: dev
+  - path: devtest/apc_Arab_nort3139.jsonl
+    split: devtest
+- config_name: apc_Arab_sout3123
+  data_files:
+  - path: dev/apc_Arab_sout3123.jsonl
+    split: dev
+  - path: devtest/apc_Arab_sout3123.jsonl
+    split: devtest
+- config_name: arb_Arab
+  data_files:
+  - path: dev/arb_Arab.jsonl
+    split: dev
+  - path: devtest/arb_Arab.jsonl
+    split: devtest
+- config_name: arb_Latn
+  data_files:
+  - path: dev/arb_Latn.jsonl
+    split: dev
+  - path: devtest/arb_Latn.jsonl
+    split: devtest
+- config_name: arg_Latn
+  data_files:
+  - path: dev/arg_Latn.jsonl
+    split: dev
+  - path: devtest/arg_Latn.jsonl
+    split: devtest
+- config_name: ars_Arab
+  data_files:
+  - path: dev/ars_Arab.jsonl
+    split: dev
+  - path: devtest/ars_Arab.jsonl
+    split: devtest
+- config_name: ary_Arab
+  data_files:
+  - path: dev/ary_Arab.jsonl
+    split: dev
+  - path: devtest/ary_Arab.jsonl
+    split: devtest
+- config_name: arz_Arab
+  data_files:
+  - path: dev/arz_Arab.jsonl
+    split: dev
+  - path: devtest/arz_Arab.jsonl
+    split: devtest
+- config_name: asm_Beng
+  data_files:
+  - path: dev/asm_Beng.jsonl
+    split: dev
+  - path: devtest/asm_Beng.jsonl
+    split: devtest
+- config_name: ast_Latn
+  data_files:
+  - path: dev/ast_Latn.jsonl
+    split: dev
+  - path: devtest/ast_Latn.jsonl
+    split: devtest
+- config_name: awa_Deva
+  data_files:
+  - path: dev/awa_Deva.jsonl
+    split: dev
+  - path: devtest/awa_Deva.jsonl
+    split: devtest
+- config_name: ayr_Latn
+  data_files:
+  - path: dev/ayr_Latn.jsonl
+    split: dev
+  - path: devtest/ayr_Latn.jsonl
+    split: devtest
+- config_name: azb_Arab
+  data_files:
+  - path: dev/azb_Arab.jsonl
+    split: dev
+  - path: devtest/azb_Arab.jsonl
+    split: devtest
+- config_name: azj_Latn
+  data_files:
+  - path: dev/azj_Latn.jsonl
+    split: dev
+  - path: devtest/azj_Latn.jsonl
+    split: devtest
+- config_name: bak_Cyrl
+  data_files:
+  - path: dev/bak_Cyrl.jsonl
+    split: dev
+  - path: devtest/bak_Cyrl.jsonl
+    split: devtest
+- config_name: bam_Latn
+  data_files:
+  - path: dev/bam_Latn.jsonl
+    split: dev
+  - path: devtest/bam_Latn.jsonl
+    split: devtest
+- config_name: ban_Latn
+  data_files:
+  - path: dev/ban_Latn.jsonl
+    split: dev
+  - path: devtest/ban_Latn.jsonl
+    split: devtest
+- config_name: bel_Cyrl
+  data_files:
+  - path: dev/bel_Cyrl.jsonl
+    split: dev
+  - path: devtest/bel_Cyrl.jsonl
+    split: devtest
+- config_name: bem_Latn
+  data_files:
+  - path: dev/bem_Latn.jsonl
+    split: dev
+  - path: devtest/bem_Latn.jsonl
+    split: devtest
+- config_name: ben_Beng
+  data_files:
+  - path: dev/ben_Beng.jsonl
+    split: dev
+  - path: devtest/ben_Beng.jsonl
+    split: devtest
+- config_name: bho_Deva
+  data_files:
+  - path: dev/bho_Deva.jsonl
+    split: dev
+  - path: devtest/bho_Deva.jsonl
+    split: devtest
+- config_name: bjn_Arab
+  data_files:
+  - path: dev/bjn_Arab.jsonl
+    split: dev
+  - path: devtest/bjn_Arab.jsonl
+    split: devtest
+- config_name: bjn_Latn
+  data_files:
+  - path: dev/bjn_Latn.jsonl
+    split: dev
+  - path: devtest/bjn_Latn.jsonl
+    split: devtest
+- config_name: bod_Tibt
+  data_files:
+  - path: dev/bod_Tibt.jsonl
+    split: dev
+  - path: devtest/bod_Tibt.jsonl
+    split: devtest
+- config_name: bos_Latn
+  data_files:
+  - path: dev/bos_Latn.jsonl
+    split: dev
+  - path: devtest/bos_Latn.jsonl
+    split: devtest
+- config_name: brx_Deva
+  data_files:
+  - path: dev/brx_Deva.jsonl
+    split: dev
+- config_name: bug_Latn
+  data_files:
+  - path: dev/bug_Latn.jsonl
+    split: dev
+  - path: devtest/bug_Latn.jsonl
+    split: devtest
+- config_name: bul_Cyrl
+  data_files:
+  - path: dev/bul_Cyrl.jsonl
+    split: dev
+  - path: devtest/bul_Cyrl.jsonl
+    split: devtest
+- config_name: cat_Latn
+  data_files:
+  - path: dev/cat_Latn.jsonl
+    split: dev
+  - path: devtest/cat_Latn.jsonl
+    split: devtest
+- config_name: cat_Latn_vale1252
+  data_files:
+  - path: devtest/cat_Latn_vale1252.jsonl
+    split: devtest
+- config_name: ceb_Latn
+  data_files:
+  - path: dev/ceb_Latn.jsonl
+    split: dev
+  - path: devtest/ceb_Latn.jsonl
+    split: devtest
+- config_name: ces_Latn
+  data_files:
+  - path: dev/ces_Latn.jsonl
+    split: dev
+  - path: devtest/ces_Latn.jsonl
+    split: devtest
+- config_name: chv_Cyrl
+  data_files:
+  - path: dev/chv_Cyrl.jsonl
+    split: dev
+  - path: devtest/chv_Cyrl.jsonl
+    split: devtest
+- config_name: cjk_Latn
+  data_files:
+  - path: dev/cjk_Latn.jsonl
+    split: dev
+  - path: devtest/cjk_Latn.jsonl
+    split: devtest
+- config_name: ckb_Arab
+  data_files:
+  - path: dev/ckb_Arab.jsonl
+    split: dev
+  - path: devtest/ckb_Arab.jsonl
+    split: devtest
+- config_name: cmn_Hans
+  data_files:
+  - path: dev/cmn_Hans.jsonl
+    split: dev
+  - path: devtest/cmn_Hans.jsonl
+    split: devtest
+- config_name: cmn_Hant
+  data_files:
+  - path: dev/cmn_Hant.jsonl
+    split: dev
+  - path: devtest/cmn_Hant.jsonl
+    split: devtest
+- config_name: crh_Latn
+  data_files:
+  - path: dev/crh_Latn.jsonl
+    split: dev
+  - path: devtest/crh_Latn.jsonl
+    split: devtest
+- config_name: cym_Latn
+  data_files:
+  - path: dev/cym_Latn.jsonl
+    split: dev
+  - path: devtest/cym_Latn.jsonl
+    split: devtest
+- config_name: dan_Latn
+  data_files:
+  - path: dev/dan_Latn.jsonl
+    split: dev
+  - path: devtest/dan_Latn.jsonl
+    split: devtest
+- config_name: dar_Cyrl
+  data_files:
+  - path: dev/dar_Cyrl.jsonl
+    split: dev
+- config_name: deu_Latn
+  data_files:
+  - path: dev/deu_Latn.jsonl
+    split: dev
+  - path: devtest/deu_Latn.jsonl
+    split: devtest
+- config_name: dgo_Deva
+  data_files:
+  - path: dev/dgo_Deva.jsonl
+    split: dev
+- config_name: dik_Latn
+  data_files:
+  - path: dev/dik_Latn.jsonl
+    split: dev
+  - path: devtest/dik_Latn.jsonl
+    split: devtest
+- config_name: dyu_Latn
+  data_files:
+  - path: dev/dyu_Latn.jsonl
+    split: dev
+  - path: devtest/dyu_Latn.jsonl
+    split: devtest
+- config_name: dzo_Tibt
+  data_files:
+  - path: dev/dzo_Tibt.jsonl
+    split: dev
+  - path: devtest/dzo_Tibt.jsonl
+    split: devtest
+- config_name: ekk_Latn
+  data_files:
+  - path: dev/ekk_Latn.jsonl
+    split: dev
+  - path: devtest/ekk_Latn.jsonl
+    split: devtest
+- config_name: ell_Grek
+  data_files:
+  - path: dev/ell_Grek.jsonl
+    split: dev
+  - path: devtest/ell_Grek.jsonl
+    split: devtest
+- config_name: eng_Latn
+  data_files:
+  - path: dev/eng_Latn.jsonl
+    split: dev
+  - path: devtest/eng_Latn.jsonl
+    split: devtest
+- config_name: epo_Latn
+  data_files:
+  - path: dev/epo_Latn.jsonl
+    split: dev
+  - path: devtest/epo_Latn.jsonl
+    split: devtest
+- config_name: eus_Latn
+  data_files:
+  - path: dev/eus_Latn.jsonl
+    split: dev
+  - path: devtest/eus_Latn.jsonl
+    split: devtest
+- config_name: ewe_Latn
+  data_files:
+  - path: dev/ewe_Latn.jsonl
+    split: dev
+  - path: devtest/ewe_Latn.jsonl
+    split: devtest
+- config_name: fao_Latn
+  data_files:
+  - path: dev/fao_Latn.jsonl
+    split: dev
+  - path: devtest/fao_Latn.jsonl
+    split: devtest
+- config_name: fij_Latn
+  data_files:
+  - path: dev/fij_Latn.jsonl
+    split: dev
+  - path: devtest/fij_Latn.jsonl
+    split: devtest
+- config_name: fil_Latn
+  data_files:
+  - path: dev/fil_Latn.jsonl
+    split: dev
+  - path: devtest/fil_Latn.jsonl
+    split: devtest
+- config_name: fin_Latn
+  data_files:
+  - path: dev/fin_Latn.jsonl
+    split: dev
+  - path: devtest/fin_Latn.jsonl
+    split: devtest
+- config_name: fon_Latn
+  data_files:
+  - path: dev/fon_Latn.jsonl
+    split: dev
+  - path: devtest/fon_Latn.jsonl
+    split: devtest
+- config_name: fra_Latn
+  data_files:
+  - path: dev/fra_Latn.jsonl
+    split: dev
+  - path: devtest/fra_Latn.jsonl
+    split: devtest
+- config_name: fur_Latn
+  data_files:
+  - path: dev/fur_Latn.jsonl
+    split: dev
+  - path: devtest/fur_Latn.jsonl
+    split: devtest
+- config_name: fuv_Latn
+  data_files:
+  - path: dev/fuv_Latn.jsonl
+    split: dev
+  - path: devtest/fuv_Latn.jsonl
+    split: devtest
+- config_name: gaz_Latn
+  data_files:
+  - path: dev/gaz_Latn.jsonl
+    split: dev
+  - path: devtest/gaz_Latn.jsonl
+    split: devtest
+- config_name: gla_Latn
+  data_files:
+  - path: dev/gla_Latn.jsonl
+    split: dev
+  - path: devtest/gla_Latn.jsonl
+    split: devtest
+- config_name: gle_Latn
+  data_files:
+  - path: dev/gle_Latn.jsonl
+    split: dev
+  - path: devtest/gle_Latn.jsonl
+    split: devtest
+- config_name: glg_Latn
+  data_files:
+  - path: dev/glg_Latn.jsonl
+    split: dev
+  - path: devtest/glg_Latn.jsonl
+    split: devtest
+- config_name: gom_Deva
+  data_files:
+  - path: dev/gom_Deva.jsonl
+    split: dev
+- config_name: gug_Latn
+  data_files:
+  - path: dev/gug_Latn.jsonl
+    split: dev
+  - path: devtest/gug_Latn.jsonl
+    split: devtest
+- config_name: guj_Gujr
+  data_files:
+  - path: dev/guj_Gujr.jsonl
+    split: dev
+  - path: devtest/guj_Gujr.jsonl
+    split: devtest
+- config_name: hat_Latn
+  data_files:
+  - path: dev/hat_Latn.jsonl
+    split: dev
+  - path: devtest/hat_Latn.jsonl
+    split: devtest
+- config_name: hau_Latn
+  data_files:
+  - path: dev/hau_Latn.jsonl
+    split: dev
+  - path: devtest/hau_Latn.jsonl
+    split: devtest
+- config_name: heb_Hebr
+  data_files:
+  - path: dev/heb_Hebr.jsonl
+    split: dev
+  - path: devtest/heb_Hebr.jsonl
+    split: devtest
+- config_name: hin_Deva
+  data_files:
+  - path: dev/hin_Deva.jsonl
+    split: dev
+  - path: devtest/hin_Deva.jsonl
+    split: devtest
+- config_name: hne_Deva
+  data_files:
+  - path: dev/hne_Deva.jsonl
+    split: dev
+  - path: devtest/hne_Deva.jsonl
+    split: devtest
+- config_name: hrv_Latn
+  data_files:
+  - path: dev/hrv_Latn.jsonl
+    split: dev
+  - path: devtest/hrv_Latn.jsonl
+    split: devtest
+- config_name: hun_Latn
+  data_files:
+  - path: dev/hun_Latn.jsonl
+    split: dev
+  - path: devtest/hun_Latn.jsonl
+    split: devtest
+- config_name: hye_Armn
+  data_files:
+  - path: dev/hye_Armn.jsonl
+    split: dev
+  - path: devtest/hye_Armn.jsonl
+    split: devtest
+- config_name: ibo_Latn
+  data_files:
+  - path: dev/ibo_Latn.jsonl
+    split: dev
+  - path: devtest/ibo_Latn.jsonl
+    split: devtest
+- config_name: ilo_Latn
+  data_files:
+  - path: dev/ilo_Latn.jsonl
+    split: dev
+  - path: devtest/ilo_Latn.jsonl
+    split: devtest
+- config_name: ind_Latn
+  data_files:
+  - path: dev/ind_Latn.jsonl
+    split: dev
+  - path: devtest/ind_Latn.jsonl
+    split: devtest
+- config_name: isl_Latn
+  data_files:
+  - path: dev/isl_Latn.jsonl
+    split: dev
+  - path: devtest/isl_Latn.jsonl
+    split: devtest
+- config_name: ita_Latn
+  data_files:
+  - path: dev/ita_Latn.jsonl
+    split: dev
+  - path: devtest/ita_Latn.jsonl
+    split: devtest
+- config_name: jav_Latn
+  data_files:
+  - path: dev/jav_Latn.jsonl
+    split: dev
+  - path: devtest/jav_Latn.jsonl
+    split: devtest
+- config_name: jpn_Jpan
+  data_files:
+  - path: dev/jpn_Jpan.jsonl
+    split: dev
+  - path: devtest/jpn_Jpan.jsonl
+    split: devtest
+- config_name: kaa_Latn
+  data_files:
+  - path: devtest/kaa_Latn.jsonl
+    split: devtest
+- config_name: kab_Latn
+  data_files:
+  - path: dev/kab_Latn.jsonl
+    split: dev
+  - path: devtest/kab_Latn.jsonl
+    split: devtest
+- config_name: kac_Latn
+  data_files:
+  - path: dev/kac_Latn.jsonl
+    split: dev
+  - path: devtest/kac_Latn.jsonl
+    split: devtest
+- config_name: kam_Latn
+  data_files:
+  - path: dev/kam_Latn.jsonl
+    split: dev
+  - path: devtest/kam_Latn.jsonl
+    split: devtest
+- config_name: kan_Knda
+  data_files:
+  - path: dev/kan_Knda.jsonl
+    split: dev
+  - path: devtest/kan_Knda.jsonl
+    split: devtest
+- config_name: kas_Arab
+  data_files:
+  - path: dev/kas_Arab.jsonl
+    split: dev
+  - path: devtest/kas_Arab.jsonl
+    split: devtest
+- config_name: kas_Deva
+  data_files:
+  - path: dev/kas_Deva.jsonl
+    split: dev
+  - path: devtest/kas_Deva.jsonl
+    split: devtest
+- config_name: kat_Geor
+  data_files:
+  - path: dev/kat_Geor.jsonl
+    split: dev
+  - path: devtest/kat_Geor.jsonl
+    split: devtest
+- config_name: kaz_Cyrl
+  data_files:
+  - path: dev/kaz_Cyrl.jsonl
+    split: dev
+  - path: devtest/kaz_Cyrl.jsonl
+    split: devtest
+- config_name: kbp_Latn
+  data_files:
+  - path: dev/kbp_Latn.jsonl
+    split: dev
+  - path: devtest/kbp_Latn.jsonl
+    split: devtest
+- config_name: kea_Latn
+  data_files:
+  - path: dev/kea_Latn.jsonl
+    split: dev
+  - path: devtest/kea_Latn.jsonl
+    split: devtest
+- config_name: khk_Cyrl
+  data_files:
+  - path: dev/khk_Cyrl.jsonl
+    split: dev
+  - path: devtest/khk_Cyrl.jsonl
+    split: devtest
+- config_name: khk_Mong
+  data_files:
+  - path: devtest/khk_Mong.jsonl
+    split: devtest
+- config_name: khm_Khmr
+  data_files:
+  - path: dev/khm_Khmr.jsonl
+    split: dev
+  - path: devtest/khm_Khmr.jsonl
+    split: devtest
+- config_name: kik_Latn
+  data_files:
+  - path: dev/kik_Latn.jsonl
+    split: dev
+  - path: devtest/kik_Latn.jsonl
+    split: devtest
+- config_name: kin_Latn
+  data_files:
+  - path: dev/kin_Latn.jsonl
+    split: dev
+  - path: devtest/kin_Latn.jsonl
+    split: devtest
+- config_name: kir_Cyrl
+  data_files:
+  - path: dev/kir_Cyrl.jsonl
+    split: dev
+  - path: devtest/kir_Cyrl.jsonl
+    split: devtest
+- config_name: kjh_Cyrl
+  data_files:
+  - path: dev/kjh_Cyrl.jsonl
+    split: dev
+  - path: devtest/kjh_Cyrl.jsonl
+    split: devtest
+- config_name: kmb_Latn
+  data_files:
+  - path: dev/kmb_Latn.jsonl
+    split: dev
+  - path: devtest/kmb_Latn.jsonl
+    split: devtest
+- config_name: kmr_Latn
+  data_files:
+  - path: dev/kmr_Latn.jsonl
+    split: dev
+  - path: devtest/kmr_Latn.jsonl
+    split: devtest
+- config_name: knc_Arab
+  data_files:
+  - path: dev/knc_Arab.jsonl
+    split: dev
+  - path: devtest/knc_Arab.jsonl
+    split: devtest
+- config_name: knc_Latn
+  data_files:
+  - path: dev/knc_Latn.jsonl
+    split: dev
+  - path: devtest/knc_Latn.jsonl
+    split: devtest
+- config_name: kor_Hang
+  data_files:
+  - path: dev/kor_Hang.jsonl
+    split: dev
+  - path: devtest/kor_Hang.jsonl
+    split: devtest
+- config_name: ktu_Latn
+  data_files:
+  - path: dev/ktu_Latn.jsonl
+    split: dev
+  - path: devtest/ktu_Latn.jsonl
+    split: devtest
+- config_name: lao_Laoo
+  data_files:
+  - path: dev/lao_Laoo.jsonl
+    split: dev
+  - path: devtest/lao_Laoo.jsonl
+    split: devtest
+- config_name: lij_Latn
+  data_files:
+  - path: dev/lij_Latn.jsonl
+    split: dev
+  - path: devtest/lij_Latn.jsonl
+    split: devtest
+- config_name: lim_Latn
+  data_files:
+  - path: dev/lim_Latn.jsonl
+    split: dev
+  - path: devtest/lim_Latn.jsonl
+    split: devtest
+- config_name: lin_Latn
+  data_files:
+  - path: dev/lin_Latn.jsonl
+    split: dev
+  - path: devtest/lin_Latn.jsonl
+    split: devtest
+- config_name: lit_Latn
+  data_files:
+  - path: dev/lit_Latn.jsonl
+    split: dev
+  - path: devtest/lit_Latn.jsonl
+    split: devtest
+- config_name: lld_Latn
+  data_files:
+  - path: dev/lld_Latn.jsonl
+    split: dev
+  - path: devtest/lld_Latn.jsonl
+    split: devtest
+- config_name: lld_Latn_gard1241
+  data_files:
+  - path: dev/lld_Latn_gard1241.jsonl
+    split: dev
+- config_name: lmo_Latn
+  data_files:
+  - path: dev/lmo_Latn.jsonl
+    split: dev
+  - path: devtest/lmo_Latn.jsonl
+    split: devtest
+- config_name: ltg_Latn
+  data_files:
+  - path: dev/ltg_Latn.jsonl
+    split: dev
+  - path: devtest/ltg_Latn.jsonl
+    split: devtest
+- config_name: ltz_Latn
+  data_files:
+  - path: dev/ltz_Latn.jsonl
+    split: dev
+  - path: devtest/ltz_Latn.jsonl
+    split: devtest
+- config_name: lua_Latn
+  data_files:
+  - path: dev/lua_Latn.jsonl
+    split: dev
+  - path: devtest/lua_Latn.jsonl
+    split: devtest
+- config_name: lug_Latn
+  data_files:
+  - path: dev/lug_Latn.jsonl
+    split: dev
+  - path: devtest/lug_Latn.jsonl
+    split: devtest
+- config_name: luo_Latn
+  data_files:
+  - path: dev/luo_Latn.jsonl
+    split: dev
+  - path: devtest/luo_Latn.jsonl
+    split: devtest
+- config_name: lus_Latn
+  data_files:
+  - path: dev/lus_Latn.jsonl
+    split: dev
+  - path: devtest/lus_Latn.jsonl
+    split: devtest
+- config_name: lvs_Latn
+  data_files:
+  - path: dev/lvs_Latn.jsonl
+    split: dev
+  - path: devtest/lvs_Latn.jsonl
+    split: devtest
+- config_name: mag_Deva
+  data_files:
+  - path: dev/mag_Deva.jsonl
+    split: dev
+  - path: devtest/mag_Deva.jsonl
+    split: devtest
+- config_name: mai_Deva
+  data_files:
+  - path: dev/mai_Deva.jsonl
+    split: dev
+  - path: devtest/mai_Deva.jsonl
+    split: devtest
+- config_name: mal_Mlym
+  data_files:
+  - path: dev/mal_Mlym.jsonl
+    split: dev
+  - path: devtest/mal_Mlym.jsonl
+    split: devtest
+- config_name: mar_Deva
+  data_files:
+  - path: dev/mar_Deva.jsonl
+    split: dev
+  - path: devtest/mar_Deva.jsonl
+    split: devtest
+- config_name: mfe_Latn
+  data_files:
+  - path: dev/mfe_Latn.jsonl
+    split: dev
+  - path: devtest/mfe_Latn.jsonl
+    split: devtest
+- config_name: mhr_Cyrl
+  data_files:
+  - path: dev/mhr_Cyrl.jsonl
+    split: dev
+  - path: devtest/mhr_Cyrl.jsonl
+    split: devtest
+- config_name: min_Arab
+  data_files:
+  - path: dev/min_Arab.jsonl
+    split: dev
+  - path: devtest/min_Arab.jsonl
+    split: devtest
+- config_name: min_Latn
+  data_files:
+  - path: dev/min_Latn.jsonl
+    split: dev
+  - path: devtest/min_Latn.jsonl
+    split: devtest
+- config_name: mkd_Cyrl
+  data_files:
+  - path: dev/mkd_Cyrl.jsonl
+    split: dev
+  - path: devtest/mkd_Cyrl.jsonl
+    split: devtest
+- config_name: mlt_Latn
+  data_files:
+  - path: dev/mlt_Latn.jsonl
+    split: dev
+  - path: devtest/mlt_Latn.jsonl
+    split: devtest
+- config_name: mni_Beng
+  data_files:
+  - path: dev/mni_Beng.jsonl
+    split: dev
+  - path: devtest/mni_Beng.jsonl
+    split: devtest
+- config_name: mni_Mtei
+  data_files:
+  - path: dev/mni_Mtei.jsonl
+    split: dev
+- config_name: mos_Latn
+  data_files:
+  - path: dev/mos_Latn.jsonl
+    split: dev
+  - path: devtest/mos_Latn.jsonl
+    split: devtest
+- config_name: mri_Latn
+  data_files:
+  - path: dev/mri_Latn.jsonl
+    split: dev
+  - path: devtest/mri_Latn.jsonl
+    split: devtest
+- config_name: mya_Mymr
+  data_files:
+  - path: dev/mya_Mymr.jsonl
+    split: dev
+  - path: devtest/mya_Mymr.jsonl
+    split: devtest
+- config_name: myv_Cyrl
+  data_files:
+  - path: dev/myv_Cyrl.jsonl
+    split: dev
+  - path: devtest/myv_Cyrl.jsonl
+    split: devtest
+- config_name: nld_Latn
+  data_files:
+  - path: dev/nld_Latn.jsonl
+    split: dev
+  - path: devtest/nld_Latn.jsonl
+    split: devtest
+- config_name: nno_Latn
+  data_files:
+  - path: dev/nno_Latn.jsonl
+    split: dev
+  - path: devtest/nno_Latn.jsonl
+    split: devtest
+- config_name: nob_Latn
+  data_files:
+  - path: dev/nob_Latn.jsonl
+    split: dev
+  - path: devtest/nob_Latn.jsonl
+    split: devtest
+- config_name: nob_Latn_radical
+  data_files:
+  - path: dev/nob_Latn_radical.jsonl
+    split: dev
+  - path: devtest/nob_Latn_radical.jsonl
+    split: devtest
+- config_name: npi_Deva
+  data_files:
+  - path: dev/npi_Deva.jsonl
+    split: dev
+  - path: devtest/npi_Deva.jsonl
+    split: devtest
+- config_name: nqo_Nkoo
+  data_files:
+  - path: dev/nqo_Nkoo.jsonl
+    split: dev
+  - path: devtest/nqo_Nkoo.jsonl
+    split: devtest
+- config_name: nso_Latn
+  data_files:
+  - path: dev/nso_Latn.jsonl
+    split: dev
+  - path: devtest/nso_Latn.jsonl
+    split: devtest
+- config_name: nus_Latn
+  data_files:
+  - path: dev/nus_Latn.jsonl
+    split: dev
+  - path: devtest/nus_Latn.jsonl
+    split: devtest
+- config_name: nya_Latn
+  data_files:
+  - path: dev/nya_Latn.jsonl
+    split: dev
+  - path: devtest/nya_Latn.jsonl
+    split: devtest
+- config_name: oci_Latn
+  data_files:
+  - path: dev/oci_Latn.jsonl
+    split: dev
+  - path: devtest/oci_Latn.jsonl
+    split: devtest
+- config_name: oci_Latn_aran1260
+  data_files:
+  - path: dev/oci_Latn_aran1260.jsonl
+    split: dev
+  - path: devtest/oci_Latn_aran1260.jsonl
+    split: devtest
+- config_name: ory_Orya
+  data_files:
+  - path: dev/ory_Orya.jsonl
+    split: dev
+  - path: devtest/ory_Orya.jsonl
+    split: devtest
+- config_name: pag_Latn
+  data_files:
+  - path: dev/pag_Latn.jsonl
+    split: dev
+  - path: devtest/pag_Latn.jsonl
+    split: devtest
+- config_name: pan_Guru
+  data_files:
+  - path: dev/pan_Guru.jsonl
+    split: dev
+  - path: devtest/pan_Guru.jsonl
+    split: devtest
+- config_name: pap_Latn
+  data_files:
+  - path: dev/pap_Latn.jsonl
+    split: dev
+  - path: devtest/pap_Latn.jsonl
+    split: devtest
+- config_name: pbt_Arab
+  data_files:
+  - path: dev/pbt_Arab.jsonl
+    split: dev
+  - path: devtest/pbt_Arab.jsonl
+    split: devtest
+- config_name: pes_Arab
+  data_files:
+  - path: dev/pes_Arab.jsonl
+    split: dev
+  - path: devtest/pes_Arab.jsonl
+    split: devtest
+- config_name: plt_Latn
+  data_files:
+  - path: dev/plt_Latn.jsonl
+    split: dev
+  - path: devtest/plt_Latn.jsonl
+    split: devtest
+- config_name: pol_Latn
+  data_files:
+  - path: dev/pol_Latn.jsonl
+    split: dev
+  - path: devtest/pol_Latn.jsonl
+    split: devtest
+- config_name: por_Latn
+  data_files:
+  - path: dev/por_Latn.jsonl
+    split: dev
+  - path: devtest/por_Latn.jsonl
+    split: devtest
+- config_name: prs_Arab
+  data_files:
+  - path: dev/prs_Arab.jsonl
+    split: dev
+  - path: devtest/prs_Arab.jsonl
+    split: devtest
+- config_name: quy_Latn
+  data_files:
+  - path: dev/quy_Latn.jsonl
+    split: dev
+  - path: devtest/quy_Latn.jsonl
+    split: devtest
+- config_name: ron_Latn
+  data_files:
+  - path: dev/ron_Latn.jsonl
+    split: dev
+  - path: devtest/ron_Latn.jsonl
+    split: devtest
+- config_name: run_Latn
+  data_files:
+  - path: dev/run_Latn.jsonl
+    split: dev
+  - path: devtest/run_Latn.jsonl
+    split: devtest
+- config_name: rus_Cyrl
+  data_files:
+  - path: dev/rus_Cyrl.jsonl
+    split: dev
+  - path: devtest/rus_Cyrl.jsonl
+    split: devtest
+- config_name: sag_Latn
+  data_files:
+  - path: dev/sag_Latn.jsonl
+    split: dev
+  - path: devtest/sag_Latn.jsonl
+    split: devtest
+- config_name: san_Deva
+  data_files:
+  - path: dev/san_Deva.jsonl
+    split: dev
+  - path: devtest/san_Deva.jsonl
+    split: devtest
+- config_name: sat_Olck
+  data_files:
+  - path: dev/sat_Olck.jsonl
+    split: dev
+  - path: devtest/sat_Olck.jsonl
+    split: devtest
+- config_name: scn_Latn
+  data_files:
+  - path: dev/scn_Latn.jsonl
+    split: dev
+  - path: devtest/scn_Latn.jsonl
+    split: devtest
+- config_name: shn_Mymr
+  data_files:
+  - path: dev/shn_Mymr.jsonl
+    split: dev
+  - path: devtest/shn_Mymr.jsonl
+    split: devtest
+- config_name: sin_Sinh
+  data_files:
+  - path: dev/sin_Sinh.jsonl
+    split: dev
+  - path: devtest/sin_Sinh.jsonl
+    split: devtest
+- config_name: slk_Latn
+  data_files:
+  - path: dev/slk_Latn.jsonl
+    split: dev
+  - path: devtest/slk_Latn.jsonl
+    split: devtest
+- config_name: slv_Latn
+  data_files:
+  - path: dev/slv_Latn.jsonl
+    split: dev
+  - path: devtest/slv_Latn.jsonl
+    split: devtest
+- config_name: smo_Latn
+  data_files:
+  - path: dev/smo_Latn.jsonl
+    split: dev
+  - path: devtest/smo_Latn.jsonl
+    split: devtest
+- config_name: sna_Latn
+  data_files:
+  - path: dev/sna_Latn.jsonl
+    split: dev
+  - path: devtest/sna_Latn.jsonl
+    split: devtest
+- config_name: snd_Arab
+  data_files:
+  - path: dev/snd_Arab.jsonl
+    split: dev
+  - path: devtest/snd_Arab.jsonl
+    split: devtest
+- config_name: snd_Deva
+  data_files:
+  - path: dev/snd_Deva.jsonl
+    split: dev
+- config_name: som_Latn
+  data_files:
+  - path: dev/som_Latn.jsonl
+    split: dev
+  - path: devtest/som_Latn.jsonl
+    split: devtest
+- config_name: sot_Latn
+  data_files:
+  - path: dev/sot_Latn.jsonl
+    split: dev
+  - path: devtest/sot_Latn.jsonl
+    split: devtest
+- config_name: spa_Latn
+  data_files:
+  - path: dev/spa_Latn.jsonl
+    split: dev
+  - path: devtest/spa_Latn.jsonl
+    split: devtest
+- config_name: srd_Latn
+  data_files:
+  - path: dev/srd_Latn.jsonl
+    split: dev
+  - path: devtest/srd_Latn.jsonl
+    split: devtest
+- config_name: srp_Cyrl
+  data_files:
+  - path: dev/srp_Cyrl.jsonl
+    split: dev
+  - path: devtest/srp_Cyrl.jsonl
+    split: devtest
+- config_name: ssw_Latn
+  data_files:
+  - path: dev/ssw_Latn.jsonl
+    split: dev
+  - path: devtest/ssw_Latn.jsonl
+    split: devtest
+- config_name: sun_Latn
+  data_files:
+  - path: dev/sun_Latn.jsonl
+    split: dev
+  - path: devtest/sun_Latn.jsonl
+    split: devtest
+- config_name: swe_Latn
+  data_files:
+  - path: dev/swe_Latn.jsonl
+    split: dev
+  - path: devtest/swe_Latn.jsonl
+    split: devtest
+- config_name: swh_Latn
+  data_files:
+  - path: dev/swh_Latn.jsonl
+    split: dev
+  - path: devtest/swh_Latn.jsonl
+    split: devtest
+- config_name: szl_Latn
+  data_files:
+  - path: dev/szl_Latn.jsonl
+    split: dev
+  - path: devtest/szl_Latn.jsonl
+    split: devtest
+- config_name: tam_Taml
+  data_files:
+  - path: dev/tam_Taml.jsonl
+    split: dev
+  - path: devtest/tam_Taml.jsonl
+    split: devtest
+- config_name: taq_Latn
+  data_files:
+  - path: dev/taq_Latn.jsonl
+    split: dev
+  - path: devtest/taq_Latn.jsonl
+    split: devtest
+- config_name: taq_Tfng
+  data_files:
+  - path: dev/taq_Tfng.jsonl
+    split: dev
+  - path: devtest/taq_Tfng.jsonl
+    split: devtest
+- config_name: tat_Cyrl
+  data_files:
+  - path: dev/tat_Cyrl.jsonl
+    split: dev
+  - path: devtest/tat_Cyrl.jsonl
+    split: devtest
+- config_name: tel_Telu
+  data_files:
+  - path: dev/tel_Telu.jsonl
+    split: dev
+  - path: devtest/tel_Telu.jsonl
+    split: devtest
+- config_name: tgk_Cyrl
+  data_files:
+  - path: dev/tgk_Cyrl.jsonl
+    split: dev
+  - path: devtest/tgk_Cyrl.jsonl
+    split: devtest
+- config_name: tha_Thai
+  data_files:
+  - path: dev/tha_Thai.jsonl
+    split: dev
+  - path: devtest/tha_Thai.jsonl
+    split: devtest
+- config_name: tir_Ethi
+  data_files:
+  - path: dev/tir_Ethi.jsonl
+    split: dev
+  - path: devtest/tir_Ethi.jsonl
+    split: devtest
+- config_name: tpi_Latn
+  data_files:
+  - path: dev/tpi_Latn.jsonl
+    split: dev
+  - path: devtest/tpi_Latn.jsonl
+    split: devtest
+- config_name: tsn_Latn
+  data_files:
+  - path: dev/tsn_Latn.jsonl
+    split: dev
+  - path: devtest/tsn_Latn.jsonl
+    split: devtest
+- config_name: tso_Latn
+  data_files:
+  - path: dev/tso_Latn.jsonl
+    split: dev
+  - path: devtest/tso_Latn.jsonl
+    split: devtest
+- config_name: tuk_Latn
+  data_files:
+  - path: dev/tuk_Latn.jsonl
+    split: dev
+  - path: devtest/tuk_Latn.jsonl
+    split: devtest
+- config_name: tum_Latn
+  data_files:
+  - path: dev/tum_Latn.jsonl
+    split: dev
+  - path: devtest/tum_Latn.jsonl
+    split: devtest
+- config_name: tur_Latn
+  data_files:
+  - path: dev/tur_Latn.jsonl
+    split: dev
+  - path: devtest/tur_Latn.jsonl
+    split: devtest
+- config_name: twi_Latn_akua1239
+  data_files:
+  - path: dev/twi_Latn_akua1239.jsonl
+    split: dev
+  - path: devtest/twi_Latn_akua1239.jsonl
+    split: devtest
+- config_name: twi_Latn_asan1239
+  data_files:
+  - path: dev/twi_Latn_asan1239.jsonl
+    split: dev
+  - path: devtest/twi_Latn_asan1239.jsonl
+    split: devtest
+- config_name: tyv_Cyrl
+  data_files:
+  - path: dev/tyv_Cyrl.jsonl
+    split: dev
+  - path: devtest/tyv_Cyrl.jsonl
+    split: devtest
+- config_name: udm_Cyrl
+  data_files:
+  - path: dev/udm_Cyrl.jsonl
+    split: dev
+- config_name: uig_Arab
+  data_files:
+  - path: dev/uig_Arab.jsonl
+    split: dev
+  - path: devtest/uig_Arab.jsonl
+    split: devtest
+- config_name: ukr_Cyrl
+  data_files:
+  - path: dev/ukr_Cyrl.jsonl
+    split: dev
+  - path: devtest/ukr_Cyrl.jsonl
+    split: devtest
+- config_name: umb_Latn
+  data_files:
+  - path: dev/umb_Latn.jsonl
+    split: dev
+  - path: devtest/umb_Latn.jsonl
+    split: devtest
+- config_name: urd_Arab
+  data_files:
+  - path: dev/urd_Arab.jsonl
+    split: dev
+  - path: devtest/urd_Arab.jsonl
+    split: devtest
+- config_name: uzn_Latn
+  data_files:
+  - path: dev/uzn_Latn.jsonl
+    split: dev
+  - path: devtest/uzn_Latn.jsonl
+    split: devtest
+- config_name: uzs_Arab
+  data_files:
+  - path: dev/uzs_Arab.jsonl
+    split: dev
+- config_name: vec_Latn
+  data_files:
+  - path: dev/vec_Latn.jsonl
+    split: dev
+  - path: devtest/vec_Latn.jsonl
+    split: devtest
+- config_name: vie_Latn
+  data_files:
+  - path: dev/vie_Latn.jsonl
+    split: dev
+  - path: devtest/vie_Latn.jsonl
+    split: devtest
+- config_name: vmw_Latn
+  data_files:
+  - path: dev/vmw_Latn.jsonl
+    split: dev
+  - path: devtest/vmw_Latn.jsonl
+    split: devtest
+- config_name: war_Latn
+  data_files:
+  - path: dev/war_Latn.jsonl
+    split: dev
+  - path: devtest/war_Latn.jsonl
+    split: devtest
+- config_name: wol_Latn
+  data_files:
+  - path: dev/wol_Latn.jsonl
+    split: dev
+  - path: devtest/wol_Latn.jsonl
+    split: devtest
+- config_name: wuu_Hans
+  data_files:
+  - path: dev/wuu_Hans.jsonl
+    split: dev
+- config_name: xho_Latn
+  data_files:
+  - path: dev/xho_Latn.jsonl
+    split: dev
+  - path: devtest/xho_Latn.jsonl
+    split: devtest
+- config_name: ydd_Hebr
+  data_files:
+  - path: dev/ydd_Hebr.jsonl
+    split: dev
+  - path: devtest/ydd_Hebr.jsonl
+    split: devtest
+- config_name: yor_Latn
+  data_files:
+  - path: dev/yor_Latn.jsonl
+    split: dev
+  - path: devtest/yor_Latn.jsonl
+    split: devtest
+- config_name: yue_Hant
+  data_files:
+  - path: dev/yue_Hant.jsonl
+    split: dev
+  - path: devtest/yue_Hant.jsonl
+    split: devtest
+- config_name: zgh_Tfng
+  data_files:
+  - path: dev/zgh_Tfng.jsonl
+    split: dev
+  - path: devtest/zgh_Tfng.jsonl
+    split: devtest
+- config_name: zsm_Latn
+  data_files:
+  - path: dev/zsm_Latn.jsonl
+    split: dev
+  - path: devtest/zsm_Latn.jsonl
+    split: devtest
+- config_name: zul_Latn
+  data_files:
+  - path: dev/zul_Latn.jsonl
+    split: dev
+  - path: devtest/zul_Latn.jsonl
+    split: devtest
+extra_gated_heading: "Protecting the integrity of FLORES+ for evaluation"
+extra_gated_fields:
+  I agree not to re-host FLORES+ in places where it could be picked up by web crawlers: checkbox
+  If I evaluate using FLORES+, I will ensure that its contents are not in the training data: checkbox
+---
+# Dataset Card for FLORES+
+
+FLORES+ is an evaluation benchmark dataset for multilingual machine translation.
+
+## Dataset Details
+
+### Dataset Description
+
+FLORES+ is a multilingual machine translation benchmark released under [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/). This dataset was originally released by FAIR researchers at Meta under the name FLORES. Further information about these initial releases can be found in [Dataset Sources](#dataset-sources) below. The data is now being managed by OLDI, [the Open Language Data Initiative](https://oldi.org/). The + has been added to the name to disambiguate between the original datasets and this new actively developed version.
+
+The data consists of translations primarily from English into over 200 language varieties. The original English sentences were sampled in equal amounts from [Wikinews](https://en.wikinews.org/) (an international news source), [Wikijunior](https://en.wikibooks.org/wiki/Wikijunior) (a collection of age-appropriate non-fiction books), and [Wikivoyage](https://en.wikivoyage.org/) (a travel guide).
+
+For each language, the dataset has 997 sentences for the dev split and 1012 sentences for the devtest split. The separate blind test set, originally developed by Meta, is not managed by OLDI and not part of this repository.
+
+- **Curated by:** [The Open Language Data Initiative](https://oldi.org/)
+- **Languages:** Currently 229 language varieties, see the full list in the table [below](#language-coverage).
+- **License:** [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
+
+The current version of the dataset is `4.5`. For the full list of versions, see [CHANGELOG.md](CHANGELOG.md).
+
+### Dataset Sources
+
+FLORES+ is based on FLORES-200, described in the following paper:
+
+```bibtex
+@article{nllb-24,
+    author="{NLLB Team} and Costa-juss{\`a}, Marta R. and Cross, James and {\c{C}}elebi, Onur and Elbayad, Maha and Heafield, Kenneth and Heffernan, Kevin and Kalbassi, Elahe and Lam, Janice and Licht, Daniel and Maillard, Jean and Sun, Anna and Wang, Skyler and Wenzek, Guillaume and Youngblood, Al and Akula, Bapi and Barrault, Loic and Gonzalez, Gabriel Mejia and Hansanti, Prangthip and Hoffman, John and Jarrett, Semarley and Sadagopan, Kaushik Ram and Rowe, Dirk and Spruit, Shannon and Tran, Chau and Andrews, Pierre and Ayan, Necip Fazil and Bhosale, Shruti and Edunov, Sergey and Fan, Angela and Gao, Cynthia and Goswami, Vedanuj and Guzm{\'a}n, Francisco and Koehn, Philipp and Mourachko, Alexandre and Ropers, Christophe and Saleem, Safiyyah and Schwenk, Holger and Wang, Jeff",
+    title="Scaling neural machine translation to 200 languages",
+    journal="Nature",
+    year="2024",
+    volume="630",
+    number="8018",
+    pages="841--846",
+    issn="1476-4687",
+    doi="10.1038/s41586-024-07335-x",
+    url="https://doi.org/10.1038/s41586-024-07335-x"
+}
+```
+
+Other authors have since contributed to the dataset. If you use this dataset in your work, please cite the relevant papers listed in [bibliography.bib](bibliography.bib).
+
+## Uses
+
+FLORES+ is intended to be used to evaluate multilingual NLP applications like machine translation. It should not be used as training data.
+
+To load the dataset using Python code, please follow the steps below:
+1. Install the [datasets](https://huggingface.co/docs/datasets) package: `pip install datasets`.
+2. Log in on the current website (Huggingface.co).
+3. Accept the terms of use of the FLORES+ dataset on this page.
+4. Get your Huggingface access token by clicking on your user icon in the top left corner, choosing "Access Tokens", and creating a token or copying an existing one.
+5. Make sure you are logged into Hugginface hub in your system by running in Python:
+```Python
+import huggingface_hub
+huggingface_hub.login() # now you will be prompted to enter your token; enter it.
+```
+6. After having logged in, you can access the full FLORES+ dataset or a part of it as follows:
+```Python
+from datasets import load_dataset
+# load dev and devtests splits for all languages
+ds_full = load_dataset("openlanguagedata/flores_plus")
+# load only the dev split for all languages
+ds_dev = load_dataset("openlanguagedata/flores_plus", split="dev")
+# load dev and devtests splits for French only
+ds_fra = load_dataset("openlanguagedata/flores_plus", "fra_Latn")
+# load dev split for French only
+ds_fra_dev = load_dataset("openlanguagedata/flores_plus", "fra_Latn", split="dev")
+```
+7. You can convert the dataset to a [Pandas](https://pandas.pydata.org) DataFrame, if this formar is familiar to you (e.g. `df = ds_fra_dev.to_pandas()`).
+8. If you want, you can save the file to disk, e.g. to the csv format (`df.to_csv("flores_plus_fra_dev.csv")`) or [other formats](https://pandas.pydata.org/docs/user_guide/io.html) supported by Pandas.
+But please do not redistribute this file publicly, unless you protect it from automatic scrapping!
+
+
+## Dataset Structure
+
+Each instance in the dataset has the structure as the following example:
+
+```json
+{
+  "id": "26",
+  "iso_639_3": "gla",
+  "iso_15924": "Latn",
+  "glottocode": "scot1245",
+  "variant": "",
+  "text": "Thuirt an aithris cuideachd gum biodh aig an Tuirc cuideachd faire luchd-sabaid ISIS a chaidh an glacadh a ghabhail os làimh is nàiseanan na Roinn-Eòrpa air an diùltadh o ath-dhùthachadh.",
+  "url": "https://en.wikinews.org/wiki/US_President_Trump_announces_troop_withdrawal_from_Syria",
+  "domain": "wikinews",
+  "topic": "politics",
+  "has_image": "yes",
+  "has_hyperlink": "yes",
+  "last_updated": "1.0",
+  "split": "devtest"
+}
+```
+
+Each languoid is uniquely identified by a combination of 4 values: `iso_639_3`, `iso_15924`, `glottocode`, and `variant` — roughly speaking, a combination of language, dialect, and orthography.
+
+Within each languoid, the values of `id` and `split` uniquely identify a sentence and are aligned across all languoids (i.e. joining any pair of langioids by `id` and `split` would result in a parallel dataset).
+
+### Data Fields
+
+- `id`: ID number for each line of data. Lines with the same ID in the same split are translations of each other.
+- `text`: A line of text in the indicated language.
+- `iso_639_3`: The ISO 639-3 code indicating the language variety.
+- `iso_15924`: The ISO 15924 indicating the writing script.
+- `glottocode`: The [Glottocode](https://glottolog.org/glottolog/language) corresponding to the language variety.
+- `variant`: An additional tag for the language variety (usually an empty string; currently used only for Radical Bokmål)
+- `url`: The URL for the English article from which the text was extracted.
+- `domain`: The domain of the text.
+- `topic`: The topic of the text.
+- `has_image`: Whether the original article contains an image.
+- `has_hyperlink`: Whether the text contains a hyperlink.
+- `last_updated`: The FLORES+ version where the given row was last updated.
+
+## Dataset Creation
+
+See the [NLLB Nature paper](https://doi.org/10.1038/s41586-024-07335-x) and the longer [NLLB technical paper](https://arxiv.org/pdf/2207.04672#page=18.71) for more details.
+
+### Additional Dataset Cards
+
+The datasets for some language varieties have individual datacards describing their creation. These can be found in the [dataset_cards](dataset_cards/) directory.
+
+## Contact
+
+For more information about the FLORES+ dataset, please see [oldi.org](https://oldi.org/).
+
+## Contributing
+
+Fixes and new language contributions are most welcome.
+
+By contributing to this project you agree to the [Developer Certificate of
+Origin (DCO)](DCO). This document was created by the Linux Kernel community and is a
+simple statement that you, as a contributor, have the legal right to make the
+contribution.
+
+In order to show your agreement with the DCO you should include at the end of commit message,
+the following line: `Signed-off-by: John Doe <john.doe@example.com>`, using your real name.
+
+This can be done easily using the `-s` flag on the `git commit`.
+
+Please see the [Contribution guidelines](https://oldi.org/guidelines) for further information.
+
+### How to add a pull request
+
+1. Go to https://huggingface.co/datasets/openlanguagedata/flores_plus/discussions, press "New pull request".
+2. In the popup window, enter a branch name and press "Create branch".
+3. On your computer, do `git clone https://huggingface.co/datasets/openlanguagedata/flores_plus`.
+4. Checkout to your newly created branch (e.g. `cd flores_plus && git fetch origin refs/pr/4:pr/4 && git checkout pr/4`).
+5. Check that you are logged in to the HF CLI tool (`hf auth whoami`). If not, please log into it (`hf auth login` and enter your token).
+6. Modify a file (for adding new languages, see the instructions below) and add the changes to git (e.g. `git add dev/rus_Cyrl.jsonl`).
+7. Commit with an -s flag (e.g. `git commit -s -m "fix a few typos in the Russian dev set"`).
+8. Push (e.g. `git push origin pr/4:refs/pr/4`) — important to make sure to push to the same "refs" branch!
+9. Go to the pull request page and see if it reflects your changes.
+10. When your pull request is ready, press the "Publish" button in its web interface.
+
+If you find this difficult, please contact us by email `info@oldi.org` or in our [Discord group](https://discord.gg/jJmrw3E77u)!
+
+### Testing your changes
+
+After contributing new translations or modifying existing ones, you can check that the data format is OK.
+Assuming that you have the Python packages `pytest` and `dataset` installed, you can type
+```
+pytest
+```
+in your console (in the `flores_plus` directory), and the tests will run.
+If any of them fails, please inspect the translations, following the hints in the test output.
+
+## Changelog
+
+See [CHANGELOG.md](CHANGELOG.md) for information about the latest changes.
+
+## Language Coverage
+
+| Code  | Script | Glottocode | Name                                |                           Notes                            |
+|-------|--------|------------|-------------------------------------|------------------------------------------------------------|
+| `ace` | `Arab` | `achi1257` | Acehnese (Jawi script)              |                                                            |
+| `ace` | `Latn` | `achi1257` | Acehnese (Latin script)             |                                                            |
+| `acm` | `Arab` | `meso1252` | Mesopotamian Arabic                 |                                                            |
+| `acq` | `Arab` | `taiz1242` | Taʽizzi-Adeni Arabic                |                                                            |
+| `aeb` | `Arab` | `tuni1259` | Tunisian Arabic                     |                                                            |
+| `afr` | `Latn` | `afri1274` | Afrikaans                           |                                                            |
+| `als` | `Latn` | `tosk1239` | Albanian (Tosk)                     |                                                            |
+| `amh` | `Ethi` | `amha1245` | Amharic                             |                                                            |
+| `apc` | `Arab` | `nort3139` | Levantine Arabic (North)            |                                                            |
+| `apc` | `Arab` | `sout3123` | Levantine Arabic (South)            |                                                            |
+| `arb` | `Arab` | `stan1318` | Modern Standard Arabic              |                                                            |
+| `arb` | `Latn` | `stan1318` | Modern Standard Arabic (Romanized)  |                                                            |
+| `arg` | `Latn` | `arag1245` | [Aragonese](dataset_cards/arg_Latn.md)          |                                                |
+| `ars` | `Arab` | `najd1235` | Najdi Arabic                        |                                                            |
+| `ary` | `Arab` | `moro1292` | Moroccan Arabic                     |                                                            |
+| `arz` | `Arab` | `egyp1253` | Egyptian Arabic                     |                                                            |
+| `asm` | `Beng` | `assa1263` | Assamese                            |                                                            |
+| `ast` | `Latn` | `astu1245` | [Asturian](dataset_cards/ast_Latn.md)           |                                                |
+| `awa` | `Deva` | `awad1243` | Awadhi                              |                                                            |
+| `ayr` | `Latn` | `cent2142` | Central Aymara                      |                                                            |
+| `azb` | `Arab` | `sout2697` | South Azerbaijani                   |                                                            |
+| `azj` | `Latn` | `nort2697` | North Azerbaijani                   |                                                            |
+| `bak` | `Cyrl` | `bash1264` | Bashkir                             |                                                            |
+| `bam` | `Latn` | `bamb1269` | Bambara                             |                                                            |
+| `ban` | `Latn` | `bali1278` | Balinese                            |                                                            |
+| `bel` | `Cyrl` | `bela1254` | Belarusian                          |                                                            |
+| `bem` | `Latn` | `bemb1257` | Bemba                               |                                                            |
+| `ben` | `Beng` | `beng1280` | Bengali                             |                                                            |
+| `bho` | `Deva` | `bhoj1244` | Bhojpuri                            |                                                            |
+| `bjn` | `Arab` | `banj1239` | Banjar (Jawi script)                |                                                            |
+| `bjn` | `Latn` | `banj1239` | Banjar (Latin script)               |                                                            |
+| `bod` | `Tibt` | `utsa1239` | Lhasa Tibetan                       |                                                            |
+| `bos` | `Latn` | `bosn1245` | Bosnian                             |                                                            |
+| `brx` | `Deva` | `bodo1269` | Bodo                                |                         `dev` only                         |
+| `bug` | `Latn` | `bugi1244` | Buginese                            |                                                            |
+| `bul` | `Cyrl` | `bulg1262` | Bulgarian                           |                                                            |
+| `cat` | `Latn` | `stan1289` | Catalan                             |                                                            |
+| `cat` | `Latn` | `vale1252` | [Valencian](dataset_cards/cat_Latn_vale1252.md) |          `devtest` only                        |
+| `ceb` | `Latn` | `cebu1242` | Cebuano                             |                                                            |
+| `ces` | `Latn` | `czec1258` | Czech                               |                                                            |
+| `chv` | `Cyrl` | `chuv1255` | [Chuvash](dataset_cards/chv_Cyrl.md)|                                                            |
+| `cjk` | `Latn` | `chok1245` | Chokwe                              |                                                            |
+| `ckb` | `Arab` | `cent1972` | Central Kurdish                     |                                                            |
+| `cmn` | `Hans` | `beij1234` | [Mandarin Chinese (Standard Beijing)](dataset_cards/cmn_Hans.md) |                               |
+| `cmn` | `Hant` | `taib1240` | [Mandarin Chinese (Taiwanese)](dataset_cards/cmn_Hant.md)       |                                |
+| `crh` | `Latn` | `crim1257` | Crimean Tatar                       |                                                            |
+| `cym` | `Latn` | `wels1247` | Welsh                               |                                                            |
+| `dan` | `Latn` | `dani1285` | Danish                              |                                                            |
+| `dar` | `Cyrl` | `darg1241` | [Dargwa](dataset_cards/dar_Cyrl.md) |                         `dev` only                         |
+| `deu` | `Latn` | `stan1295` | German                              |                                                            |
+| `dgo` | `Deva` | `dogr1250` | Dogri                               |                         `dev` only                         |
+| `dik` | `Latn` | `sout2832` | Southwestern Dinka                  |                                                            |
+| `dyu` | `Latn` | `dyul1238` | Dyula                               |                                                            |
+| `dzo` | `Tibt` | `dzon1239` | Dzongkha                            |                                                            |
+| `ekk` | `Latn` | `esto1258` | Estonian                            |                                                            |
+| `ell` | `Grek` | `mode1248` | Greek                               |                                                            |
+| `eng` | `Latn` | `stan1293` | English                             |                                                            |
+| `epo` | `Latn` | `espe1235` | Esperanto                           |                                                            |
+| `eus` | `Latn` | `basq1248` | Basque                              |                                                            |
+| `ewe` | `Latn` | `ewee1241` | Ewe                                 |                                                            |
+| `fao` | `Latn` | `faro1244` | Faroese                             |                                                            |
+| `fij` | `Latn` | `fiji1243` | Fijian                              |                                                            |
+| `fil` | `Latn` | `fili1244` | [Filipino](dataset_cards/fil_Latn.md)           |                                                |
+| `fin` | `Latn` | `finn1318` | Finnish                             |                                                            |
+| `fon` | `Latn` | `fonn1241` | Fon                                 |                                                            |
+| `fra` | `Latn` | `stan1290` | French                              |                                                            |
+| `fur` | `Latn` | `east2271` | Friulian                            |                                                            |
+| `fuv` | `Latn` | `nige1253` | Nigerian Fulfulde                   |                                                            |
+| `gaz` | `Latn` | `west2721` | West Central Oromo                  |                                                            |
+| `gla` | `Latn` | `scot1245` | Scottish Gaelic                     |                                                            |
+| `gle` | `Latn` | `iris1253` | Irish                               |                                                            |
+| `glg` | `Latn` | `gali1258` | Galician                            |                                                            |
+| `gom` | `Deva` | `goan1235` | Goan Konkani                        |                         `dev` only                         |
+| `gug` | `Latn` | `para1311` | Paraguayan Guaraní                  |                                                            |
+| `guj` | `Gujr` | `guja1252` | Gujarati                            |                                                            |
+| `hat` | `Latn` | `hait1244` | Haitian Creole                      |                                                            |
+| `hau` | `Latn` | `haus1257` | Hausa                               |                                                            |
+| `heb` | `Hebr` | `hebr1245` | Hebrew                              |                                                            |
+| `hin` | `Deva` | `hind1269` | Hindi                               |                                                            |
+| `hne` | `Deva` | `chha1249` | Chhattisgarhi                       |                                                            |
+| `hrv` | `Latn` | `croa1245` | Croatian                            |                                                            |
+| `hun` | `Latn` | `hung1274` | Hungarian                           |                                                            |
+| `hye` | `Armn` | `nucl1235` | Armenian                            |                                                            |
+| `ibo` | `Latn` | `nucl1417` | Igbo                                |                                                            |
+| `ilo` | `Latn` | `ilok1237` | Ilocano                             |                                                            |
+| `ind` | `Latn` | `indo1316` | Indonesian                          |                                                            |
+| `isl` | `Latn` | `icel1247` | Icelandic                           |                                                            |
+| `ita` | `Latn` | `ital1282` | Italian                             |                                                            |
+| `jav` | `Latn` | `java1254` | Javanese                            |                                                            |
+| `jpn` | `Jpan` | `nucl1643` | Japanese                            |                                                            |
+| `kaa` | `Latn` | `kara1467` | [Karakalpak](dataset_cards/kaa_Latn.md)         |           `devtest` only                       |
+| `kab` | `Latn` | `kaby1243` | Kabyle                              |                                                            |
+| `kac` | `Latn` | `kach1280` | Jingpho                             |                                                            |
+| `kam` | `Latn` | `kamb1297` | Kamba                               |                                                            |
+| `kan` | `Knda` | `nucl1305` | Kannada                             |                                                            |
+| `kas` | `Arab` | `kash1277` | Kashmiri (Arabic script)            |                                                            |
+| `kas` | `Deva` | `kash1277` | Kashmiri (Devanagari script)        |                                                            |
+| `kat` | `Geor` | `nucl1302` | Georgian                            |                                                            |
+| `kaz` | `Cyrl` | `kaza1248` | Kazakh                              |                                                            |
+| `kbp` | `Latn` | `kabi1261` | Kabiyè                              |                                                            |
+| `kea` | `Latn` | `kabu1256` | Kabuverdianu                        |                                                            |
+| `khk` | `Cyrl` | `halh1238` | Halh Mongolian (Cyrillic script)    |                                                            |
+| `khk` | `Mong` | `halh1238` | [Halh Mongolian (traditional Mongolian script)](dataset_cards/khk_Mong.md)| `devtest` only     |
+| `khm` | `Khmr` | `cent1989` | Khmer (Central)                     |                                                            |
+| `kik` | `Latn` | `kiku1240` | Kikuyu                              |                                                            |
+| `kin` | `Latn` | `kiny1244` | Kinyarwanda                         |                                                            |
+| `kir` | `Cyrl` | `kirg1245` | Kyrgyz                              |                                                            |
+| `kjh` | `Cyrl` | `khak1248` | [Khakas](dataset_cards/kjh_Cyrl.md) |                                                            |
+| `kmb` | `Latn` | `kimb1241` | Kimbundu                            |                                                            |
+| `kmr` | `Latn` | `nort2641` | Northern Kurdish                    |                                                            |
+| `knc` | `Arab` | `cent2050` | Central Kanuri (Arabic script)      |                                                            |
+| `knc` | `Latn` | `cent2050` | Central Kanuri (Latin script)       |                                                            |
+| `kor` | `Hang` | `kore1280` | Korean                              |                                                            |
+| `ktu` | `Latn` | `kitu1246` | Kituba (DRC)                        |                                                            |
+| `lao` | `Laoo` | `laoo1244` | Lao                                 |                                                            |
+| `lij` | `Latn` | `geno1240` | Ligurian (Genoese)                  |                                                            |
+| `lim` | `Latn` | `limb1263` | Limburgish                          |                                                            |
+| `lin` | `Latn` | `ling1263` | Lingala                             |                                                            |
+| `lit` | `Latn` | `lith1251` | Lithuanian                          |                                                            |
+| `lld` | `Latn` | `ladi1250` | [Ladin (Val Badia)](dataset_cards/lld_Latn.md)  |                                                |
+| `lld` | `Latn` | `gard1241` | [Ladin (Gherdëina)](dataset_cards/lld_Latn_gard1241.md)  |                                       |
+| `lmo` | `Latn` | `lomb1257` | Lombard                             | [[1]](https://github.com/openlanguagedata/flores/issues/5) |
+| `ltg` | `Latn` | `east2282` | Latgalian                           |                                                            |
+| `ltz` | `Latn` | `luxe1241` | Luxembourgish                       |                                                            |
+| `lua` | `Latn` | `luba1249` | Luba-Kasai                          |                                                            |
+| `lug` | `Latn` | `gand1255` | Ganda                               |                                                            |
+| `luo` | `Latn` | `luok1236` | Luo                                 |                                                            |
+| `lus` | `Latn` | `lush1249` | Mizo                                |                                                            |
+| `lvs` | `Latn` | `stan1325` | Standard Latvian                    |                                                            |
+| `mag` | `Deva` | `maga1260` | Magahi                              |                                                            |
+| `mai` | `Deva` | `mait1250` | Maithili                            |                                                            |
+| `mal` | `Mlym` | `mala1464` | Malayalam                           |                                                            |
+| `mar` | `Deva` | `mara1378` | Marathi                             |                                                            |
+| `mfe` | `Latn` | `mori1278` | [Mauritian Creole](dataset_cards/mfe_Latn.md)   |                                                |
+| `mhr` | `Cyrl` | `gras1239` | [Meadow Mari](dataset_cards/mhr_Cyrl.md)        |                                                |
+| `min` | `Arab` | `mina1268` | Minangkabau (Jawi script)           |                                                            |
+| `min` | `Latn` | `mina1268` | Minangkabau (Latin script)          |                                                            |
+| `mkd` | `Cyrl` | `mace1250` | Macedonian                          |                                                            |
+| `mlt` | `Latn` | `malt1254` | Maltese                             |                                                            |
+| `mni` | `Beng` | `mani1292` | Meitei (Manipuri, Bengali script)   |                                                            |
+| `mni` | `Mtei` | `mani1292` | Meitei (Manipuri, Meitei script)    |                         `dev` only                         |
+| `mos` | `Latn` | `moss1236` | Mossi                               |                                                            |
+| `mri` | `Latn` | `maor1246` | Maori                               |                                                            |
+| `mya` | `Mymr` | `nucl1310` | Burmese                             |                                                            |
+| `myv` | `Cyrl` | `erzy1239` | [Erzya](dataset_cards/myv_Cyrl.md)  |                                                            |
+| `nld` | `Latn` | `dutc1256` | Dutch                               |                                                            |
+| `nno` | `Latn` | `norw1262` | Norwegian Nynorsk                   |                                                            |
+| `nob` | `Latn` | `norw1259` | [Norwegian Bokmål, moderate variety](dataset_cards/nob_Latn.md)  |                               |
+| `nob` | `Latn` | `norw1259` (with `variant=radical`) | [Norwegian Bokmål, radical variety](dataset_cards/nob_Latn_radical.md)  | |
+| `npi` | `Deva` | `nepa1254` | Nepali                              |                                                            |
+| `nqo` | `Nkoo` | `nkoa1234` | Nko                                 |                                                            |
+| `nso` | `Latn` | `pedi1238` | Northern Sotho                      |                                                            |
+| `nus` | `Latn` | `nuer1246` | Nuer                                |                                                            |
+| `nya` | `Latn` | `nyan1308` | Nyanja                              |                                                            |
+| `oci` | `Latn` | `occi1239` | Occitan                             |                                                            |
+| `oci` | `Latn` | `aran1260` | [Aranese](dataset_cards/oci_Latn_aran1260.md)   |                                                |
+| `ory` | `Orya` | `oriy1255` | Odia                                |                                                            |
+| `pag` | `Latn` | `pang1290` | Pangasinan                          |                                                            |
+| `pan` | `Guru` | `panj1256` | Eastern Panjabi                     |                                                            |
+| `pap` | `Latn` | `papi1253` | Papiamento                          |                                                            |
+| `pbt` | `Arab` | `sout2649` | Southern Pashto                     |                                                            |
+| `pes` | `Arab` | `west2369` | Western Persian                     |                                                            |
+| `plt` | `Latn` | `plat1254` | Plateau Malagasy                    |                                                            |
+| `pol` | `Latn` | `poli1260` | Polish                              |                                                            |
+| `por` | `Latn` | `braz1246` | Portuguese (Brazilian)              |                                                            |
+| `prs` | `Arab` | `dari1249` | Dari                                |                                                            |
+| `quy` | `Latn` | `ayac1239` | Ayacucho Quechua                    |                                                            |
+| `ron` | `Latn` | `roma1327` | Romanian                            |                                                            |
+| `run` | `Latn` | `rund1242` | Rundi                               |                                                            |
+| `rus` | `Cyrl` | `russ1263` | Russian                             |                                                            |
+| `sag` | `Latn` | `sang1328` | Sango                               |                                                            |
+| `san` | `Deva` | `sans1269` | Sanskrit                            |                                                            |
+| `sat` | `Olck` | `sant1410` | Santali                             |                                                            |
+| `scn` | `Latn` | `sici1248` | Sicilian                            |                                                            |
+| `shn` | `Mymr` | `shan1277` | Shan                                |                                                            |
+| `sin` | `Sinh` | `sinh1246` | Sinhala                             |                                                            |
+| `slk` | `Latn` | `slov1269` | Slovak                              |                                                            |
+| `slv` | `Latn` | `slov1268` | Slovenian                           |                                                            |
+| `smo` | `Latn` | `samo1305` | Samoan                              |                                                            |
+| `sna` | `Latn` | `shon1251` | Shona                               |                                                            |
+| `snd` | `Arab` | `sind1272` | Sindhi (Arabic script)              |                                                            |
+| `snd` | `Deva` | `sind1272` | Sindhi (Devanagari script)          |                         `dev` only                         |
+| `som` | `Latn` | `soma1255` | Somali                              |                                                            |
+| `sot` | `Latn` | `sout2807` | Southern Sotho                      |                                                            |
+| `spa` | `Latn` | `amer1254` | Spanish (Latin American)            |                                                            |
+| `srd` | `Latn` | `sard1257` | Sardinian                           | [[1]](https://github.com/openlanguagedata/flores/issues/6) |
+| `srp` | `Cyrl` | `serb1264` | Serbian                             |                                                            |
+| `ssw` | `Latn` | `swat1243` | Swati                               |                                                            |
+| `sun` | `Latn` | `sund1252` | Sundanese                           |                                                            |
+| `swe` | `Latn` | `swed1254` | Swedish                             |                                                            |
+| `swh` | `Latn` | `swah1253` | Swahili                             |                                                            |
+| `szl` | `Latn` | `sile1253` | Silesian                            |                                                            |
+| `tam` | `Taml` | `tami1289` | Tamil                               |                                                            |
+| `taq` | `Latn` | `tama1365` | Tamasheq (Latin script)             |                                                            |
+| `taq` | `Tfng` | `tama1365` | Tamasheq (Tifinagh script)          |                                                            |
+| `tat` | `Cyrl` | `tata1255` | Tatar                               |                                                            |
+| `tel` | `Telu` | `telu1262` | Telugu                              |                                                            |
+| `tgk` | `Cyrl` | `taji1245` | Tajik                               |                                                            |
+| `tha` | `Thai` | `thai1261` | Thai                                |                                                            |
+| `tir` | `Ethi` | `tigr1271` | Tigrinya                            |                                                            |
+| `tpi` | `Latn` | `tokp1240` | Tok Pisin                           |                                                            |
+| `tsn` | `Latn` | `tswa1253` | Tswana                              |                                                            |
+| `tso` | `Latn` | `tson1249` | Tsonga                              |                                                            |
+| `tuk` | `Latn` | `turk1304` | Turkmen                             |                                                            |
+| `tum` | `Latn` | `tumb1250` | Tumbuka                             |                                                            |
+| `tur` | `Latn` | `nucl1301` | Turkish                             |                                                            |
+| `twi` | `Latn` | `akua1239` | Akuapem Twi                         |                                                            |
+| `twi` | `Latn` | `asan1239` | Asante Twi                          |                                                            |
+| `tyv` | `Cyrl` | `tuvi1240` | [Tuvan](dataset_cards/tyv_Cyrl.md)  |                                                            |
+| `udm` | `Cyrl` | `udmu1245` | [Udmurt](dataset_cards/udm_Cyrl.md) |                           `dev` only                       |
+| `uig` | `Arab` | `uigh1240` | Uyghur                              |                                                            |
+| `ukr` | `Cyrl` | `ukra1253` | Ukrainian                           |                                                            |
+| `umb` | `Latn` | `umbu1257` | Umbundu                             |                                                            |
+| `urd` | `Arab` | `urdu1245` | Urdu                                |                                                            |
+| `uzn` | `Latn` | `nort2690` | Northern Uzbek                      |                                                            |
+| `uzs` | `Arab` | `sout2699` | [Southern Uzbek](dataset_cards/uzs_Arab.md) |                   `dev` only                       |
+| `vec` | `Latn` | `vene1259` | Venetian                            |                                                            |
+| `vie` | `Latn` | `viet1252` | Vietnamese                          |                                                            |
+| `vmw` | `Latn` | `cent2033` | [Emakhuwa (Central)](dataset_cards/vmw_Latn.md) |                                                |
+| `war` | `Latn` | `wara1300` | Waray                               |                                                            |
+| `wol` | `Latn` | `nucl1347` | Wolof                               |                                                            |
+| `wuu` | `Hans` | `suhu1238` | [Wu Chinese](dataset_cards/wuu_Hans.md)         |               `dev` only                       |
+| `xho` | `Latn` | `xhos1239` | Xhosa                               |                                                            |
+| `ydd` | `Hebr` | `east2295` | Eastern Yiddish                     |                                                            |
+| `yor` | `Latn` | `yoru1245` | Yoruba                              |                                                            |
+| `yue` | `Hant` | `xian1255` | [Yue Chinese (Hong Kong Cantonese)](dataset_cards/yue_Hant.md)   |                               |
+| `zgh` | `Tfng` | `stan1324` | [Standard Moroccan Tamazight](dataset_cards/zgh_Tfng.md)         |                               |
+| `zsm` | `Latn` | `stan1306` | Standard Malay                      |                                                            |
+| `zul` | `Latn` | `zulu1248` | Zulu                                |                                                            |

artifacts/hf_readmes/oscar-corpus__OSCAR-2301__README.md

@@ -0,0 +1,531 @@
+---
+license: cc0-1.0
+size_categories:
+- n>1T
+multilinguality:
+- multilingual
+source_datasets:
+- original
+task_categories:
+- fill-mask
+- text-generation
+task_ids:
+- language-modeling
+paperswithcode_id: oscar
+extra_gated_prompt: "**[IMPORTANT: We are forced to temporarily suspend access to OSCAR. The temporary nature of this suspension has led us to choose to implement it as gated access with manual control, and we will not grant any access until the situation has been clarified. We sincerely apologize for this situation and hope to be able to restore access as soon as possible. In the meantime, we remind you that we have always prohibited access to or use of OSCAR that violates the legislation in force where you are located. In France, for example, any use of OSCAR that does not fall within the framework of the so-called 'TDM' or the so-called 'research' exceptions to copyright has always been prohibited.]** By filling the form below, you understand that only the metadata and the annotations of OSCAR 23.01 have a cc0-1.0 license, and that the rest of the content is crawled data derived from the November/December 2022 snapshot of Common Crawl, for which the authors of OSCAR **do not** hold any copyright whatsoever."
+extra_gated_fields:
+ Name: text
+ Email: text
+ Affiliation: text
+ Country: text
+ Usecase: text
+ I have explicitly check with my jurisdiction and I confirm that downloading OSCAR 2301 is legal in the country/region where I am located right now, and for the use case that I have described above: checkbox
+---
+
+# Dataset Card for "OSCAR 23.01"
+
+## IMPORTANT NOTE: THIS DATASET CARD IS STILL BEING WRITTEN, PLEASE BE PATIENT WHILE WE COMPLETE ALL THE INFORMATION ABOUT THE CORPUS
+
+## Table of Contents
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-fields)
+  - [Data Splits](#data-splits)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+  - [Contributions](#contributions)
+
+## Dataset Description
+
+- **Homepage:** [https://oscar-project.org](https://oscar-project.org)
+- **Repository:** [https://github.com/oscar-project](https://github.com/oscar-project)
+- **Papers:** [Towards a Cleaner Document-Oriented Multilingual Crawled Corpus](https://aclanthology.org/2022.lrec-1.463/), [Perplexed by Quality: A Perplexity-based Method for Adult and Harmful Content Detection in Multilingual Heterogeneous Web Data](https://arxiv.org/abs/2212.10440)
+- **Point of Contact:** [Contact](https://oscar-project.org/#contact)
+
+### Dataset Summary
+
+The OSCAR project (**O**pen **S**uper-large **C**rawled **A**ggregated co**R**pus) is an Open Source project aiming to provide web-based multilingual resources and datasets for Machine Learning (ML) and Artificial Intelligence (AI) applications. The project focuses specifically in providing large quantities of unannotated raw data that is commonly used in the pre-training of large deep learning models. The OSCAR project has developed [high-performance data pipelines](https://github.com/oscar-corpus/ungoliant) specifically conceived to classify and filter large amounts of [web data](https://commoncrawl.org/). The project has also put special attention in improving the data quality of web-based corpora as well as providing data for low-resource languages, so that these new ML/AI technologies are accessible to as many communities as possible.
+
+OSCAR 23.01 is the January 2023 version of the OSCAR Corpus based on the [November/December 2022 dump of Common Crawl](https://commoncrawl.org/2022/12/nov-dec-2022-crawl-archive-now-available/). While being quite similar to OSCAR 22.01, it contains several new features, including [KenLM](https://kheafield.com/code/kenlm/)-based adult content detection, precomputed [Locality-Sensitive Hashes](https://fr.wikipedia.org/wiki/Locality_sensitive_hashing) for near deduplication, and [blocklist](https://dsi.ut-capitole.fr/blacklists/index_en.php)-based categories. OSCAR 23.01 has also moved from gzip to [Zstandard compression](https://facebook.github.io/zstd/). You might already have `zstd` installed on your system, but if not, please check the [Zstandard website](https://facebook.github.io/zstd/) for installation instructions.
+
+### Supported Tasks and Leaderboards
+
+OSCAR is mainly intended to pretrain language models and word representations.
+
+### Languages
+
+All the data is distributed by language, both the original and the deduplicated versions of the data are available. 151 different languages are available. The table in subsection [Data Splits Sample Size](#data-splits-sample-size) provides the language code for each subcorpus as well as the number of words (space separated tokens), lines and sizes for both the original and the deduplicated versions of OSCAR.
+
+### Issues
+
+OSCAR 23.01 may have quality issues on low size subcorpora, as it has been the case before.
+
+Note that since the documents are identified as a whole, it is expected to have lines in other languages in a given language subcorpus.
+As an example, it is known and expected that the German subcorpus contains documents holding lines identified as Swiss German / Alemannic.
+
+**If you encounter something that is unexpected, please file an issue here: https://github.com/oscar-corpus/corpus/issues.**
+
+|Language code|Language|Issues|
+|-------------|--------|------|
+|             |        |      |
+## Dataset Structure
+
+We show detailed information for all the configurations of the dataset.
+
+### Data Instances
+
+TODO
+
+### Layout
+
+```js
+{
+   "content":"English sentence\nphrase en français\n????????????", // (1)
+   "warc_headers":{ // (2)
+      "warc-identified-content-language":"fra,eng",
+      "warc-target-uri":"https://fr.wikipedia.org/wiki/...",
+      "warc-record-id":"<urn:uuid:29eaa920-d299-4b1d-b687-c72bd8d68116>",
+      "warc-type":"conversion",
+      "content-length":"35298", // (3)
+      "warc-refers-to":"<urn:uuid:39e42055-0d94-4e45-9c6c-9e7056635d64>",
+      "warc-block-digest":"sha1:WFH2A5WHCS2H365GIAFYQPI7UOAMFGHB", // (3)
+      "warc-date":"2022-11-26T09:45:47Z",
+      "content-type":"text/plain"
+   },
+   "metadata":{
+      "identification":{ // (4)
+         "label":"fr",
+         "prob":0.8938327
+      },
+      "harmful_pp":4063.1814, // (5)
+      "tlsh":"tlsh:T125315FF2B6088901EEA097015DB39B4600B...", // (6)
+      "quality_warnings":[ // (7)
+         "short_sentences",
+         "header",
+         "footer"
+      ],
+      "categories":[ // (8)
+         "examen_pix",
+         "liste_bu"
+      ],
+      "sentence_identifications":[ // (9)
+         {
+            "label":"fr",
+            "prob":0.99837273
+         },
+         {
+            "label":"en",
+            "prob":0.9992377
+         },
+         null
+      ]
+   }
+}
+```
+
+### Data Splits
+
+
+<details>
+  <summary>Click to expand the number of samples per configuration</summary>
+</details>
+
+## Table
+
+|     | Code   | Language                 | # docs        | # words         | Content Length :  |
+|----:|:-------|:-------------------------|:--------------|:----------------|:-----------------|
+|   0 | af     | Afrikaans                | 23,994        | 6,217,024       | 37.2 MB          |
+|   1 | sq     | Albanian                 | 1,342,790     | 462,694,599     | 3.2 GB           |
+|   2 | am     | Amharic                  | 119,434       | 40,262,809      | 512.9 MB         |
+|   3 | ar     | Arabic                   | 25,012,116    | 10,081,452,882  | 110.7 GB         |
+|   4 | an     | Aragonese                | 34            | 264             | 11.0 kB          |
+|   5 | hy     | Armenian                 | 1,056,974     | 336,045,041     | 4.9 GB           |
+|   6 | as     | Assamese                 | 89,542        | 24,395,215      | 412.1 MB         |
+|   7 | ast    | Asturian                 | 440           | 10,917          | 74.1 kB          |
+|   8 | av     | Avaric                   | 44            | 1,073           | 18.6 kB          |
+|   9 | az     | Azerbaijani              | 1,159,994     | 316,850,330     | 3.0 GB           |
+|  10 | bn     | Bangla                   | 3,474,086     | 1,092,983,765   | 19.1 GB          |
+|  11 | ba     | Bashkir                  | 128,248       | 26,036,637      | 363.7 MB         |
+|  12 | eu     | Basque                   | 678,474       | 136,672,615     | 1.2 GB           |
+|  13 | be     | Belarusian               | 445,612       | 164,729,607     | 2.3 GB           |
+|  14 | bh     | Bihari languages         | 48            | 507             | 6.8 kB           |
+|  15 | bpy    | Bishnupriya              | 2,346         | 346,947         | 5.4 MB           |
+|  16 | bs     | Bosnian                  | 20            | 395             | 3.0 kB           |
+|  17 | br     | Breton                   | 36,338        | 4,759,407       | 31.4 MB          |
+|  18 | bg     | Bulgarian                | 8,933,998     | 3,635,273,738   | 44.1 GB          |
+|  19 | my     | Burmese                  | 430,276       | 82,433,836      | 3.0 GB           |
+|  20 | ca     | Catalan                  | 6,953,898     | 2,240,460,836   | 15.3 GB          |
+|  21 | ceb    | Cebuano                  | 16,174        | 6,263,404       | 41.1 MB          |
+|  22 | ckb    | Central Kurdish          | 182,508       | 61,334,746      | 772.9 MB         |
+|  23 | ce     | Chechen                  | 11,686        | 1,051,752       | 13.9 MB          |
+|  24 | zh     | Chinese                  | 138,478,270   | 44,378,380,161  | 1.4 TB           |
+|  25 | cv     | Chuvash                  | 16,652        | 3,039,925       | 42.3 MB          |
+|  26 | kw     | Cornish                  | 8             | 80              | 432 Bytes        |
+|  27 | hr     | Croatian                 | 31,808        | 3,542,961       | 26.5 MB          |
+|  28 | cs     | Czech                    | 34,859,632    | 9,717,378,559   | 77.0 GB          |
+|  29 | da     | Danish                   | 7,214,338     | 2,217,634,340   | 14.8 GB          |
+|  30 | dv     | Divehi                   | 77,060        | 10,655,359      | 200.1 MB         |
+|  31 | nl     | Dutch                    | 72,552,688    | 19,564,553,306  | 135.0 GB         |
+|  32 | mhr    | Eastern Mari             | 9,502         | 1,615,215       | 22.9 MB          |
+|  33 | arz    | Egyptian Arabic          | 3,958         | 385,511         | 3.7 MB           |
+|  34 | en     | English                  | 1,235,510,986 | 523,869,288,690 | 3.4 TB           |
+|  35 | eo     | Esperanto                | 226,924       | 67,774,923      | 474.8 MB         |
+|  36 | et     | Estonian                 | 3,601,904     | 938,296,892     | 8.0 GB           |
+|  37 | tl     | Filipino                 | 250,558       | 110,560,444     | 719.2 MB         |
+|  38 | fi     | Finnish                  | 14,471,710    | 4,198,143,883   | 41.1 GB          |
+|  39 | fr     | French                   | 158,334,998   | 62,127,088,294  | 430.5 GB         |
+|  40 | gl     | Galician                 | 248,762       | 38,345,625      | 255.7 MB         |
+|  41 | ka     | Georgian                 | 1,343,036     | 373,935,158     | 8.4 GB           |
+|  42 | de     | German                   | 206,598,430   | 73,848,586,648  | 594.7 GB         |
+|  43 | gom    | Goan Konkani             | 398           | 121,035         | 2.3 MB           |
+|  44 | el     | Greek                    | 20,282,864    | 7,691,622,692   | 95.7 GB          |
+|  45 | gn     | Guarani                  | 14            | 260             | 2.2 kB           |
+|  46 | gu     | Gujarati                 | 425,552       | 417,001,705     | 5.6 GB           |
+|  47 | ht     | Haitian Creole           | 2             | 20,671          | 93.1 kB          |
+|  48 | he     | Hebrew                   | 3,997,888     | 1,697,158,891   | 18.0 GB          |
+|  49 | hi     | Hindi                    | 5,514,454     | 2,475,605,444   | 32.6 GB          |
+|  50 | hu     | Hungarian                | 21,349,372    | 16,013,364,289  | 150.1 GB         |
+|  51 | is     | Icelandic                | 1,210,232     | 294,471,539     | 2.2 GB           |
+|  52 | io     | Ido                      | 224           | 2,598           | 16.1 kB          |
+|  53 | ilo    | Iloko                    | 144           | 4,411           | 28.0 kB          |
+|  54 | id     | Indonesian               | 7,109,778     | 3,228,020,221   | 23.4 GB          |
+|  55 | ia     | Interlingua              | 34            | 9,384           | 33.5 kB          |
+|  56 | ie     | Interlingue              | 2             | 0               | 881 Bytes        |
+|  57 | ga     | Irish                    | 29,894        | 9,054,923       | 63.2 MB          |
+|  58 | it     | Italian                  | 89,021,606    | 36,327,274,203  | 259.4 GB         |
+|  59 | ja     | Japanese                 | 94,236,404    | 4,401,059,165   | 181.2 GB         |
+|  60 | jv     | Javanese                 | 172           | 3,286           | 25.7 kB          |
+|  61 | xal    | Kalmyk                   | 2             | 27              | 315 Bytes        |
+|  62 | kn     | Kannada                  | 448,500       | 124,924,350     | 2.6 GB           |
+|  63 | krc    | Karachay-Balkar          | 496           | 8,385           | 122.4 kB         |
+|  64 | kk     | Kazakh                   | 677,622       | 214,679,857     | 3.3 GB           |
+|  65 | km     | Khmer                    | 450,660       | 59,880,231      | 3.2 GB           |
+|  66 | kv     | Komi                     | 460           | 5,909           | 70.3 kB          |
+|  67 | ko     | Korean                   | 15,147,698    | 3,435,866,935   | 38.1 GB          |
+|  68 | ku     | Kurdish                  | 80,338        | 25,921,607      | 174.1 MB         |
+|  69 | ky     | Kyrgyz                   | 144,288       | 32,062,783      | 489.3 MB         |
+|  70 | lo     | Lao                      | 118,374       | 10,659,203      | 472.1 MB         |
+|  71 | la     | Latin                    | 14,384        | 307,865         | 2.0 MB           |
+|  72 | lv     | Latvian                  | 2,435,882     | 845,459,899     | 7.4 GB           |
+|  73 | lez    | Lezghian                 | 676           | 60,634          | 856.6 kB         |
+|  74 | li     | Limburgish               | 6             | 169             | 1.4 kB           |
+|  75 | lt     | Lithuanian               | 5,182,028     | 1,674,362,574   | 14.5 GB          |
+|  76 | jbo    | Lojban                   | 572           | 312,315         | 1.5 MB           |
+|  77 | lmo    | Lombard                  | 112           | 3,269           | 21.0 kB          |
+|  78 | nds    | Low German               | 5,248         | 1,612,175       | 10.7 MB          |
+|  79 | dsb    | Lower Sorbian            | 8             | 84              | 664 Bytes        |
+|  80 | lb     | Luxembourgish            | 18,090        | 2,514,838       | 18.4 MB          |
+|  81 | mk     | Macedonian               | 1,063,298     | 389,344,425     | 4.7 GB           |
+|  82 | mai    | Maithili                 | 46            | 467             | 6.8 kB           |
+|  83 | mg     | Malagasy                 | 10,830        | 1,416,430       | 11.2 MB          |
+|  84 | ms     | Malay                    | 11,500        | 238,477         | 2.6 MB           |
+|  85 | ml     | Malayalam                | 800,936       | 236,597,838     | 5.8 GB           |
+|  86 | mt     | Maltese                  | 5,180         | 149,886         | 1.3 MB           |
+|  87 | mr     | Marathi                  | 729,578       | 252,706,331     | 4.5 GB           |
+|  88 | mzn    | Mazanderani              | 384           | 16,115          | 169.2 kB         |
+|  89 | min    | Minangkabau              | 2,436         | 305,589         | 3.8 MB           |
+|  90 | xmf    | Mingrelian               | 7,318         | 283,316         | 6.1 MB           |
+|  91 | mwl    | Mirandese                | 4             | 54              | 423 Bytes        |
+|  92 | mn     | Mongolian                | 1,061,710     | 454,350,415     | 5.8 GB           |
+|  93 | multi  | **Multilingual**         | 2,948,202     | 1,251,676,406   | 11.9 GB          |
+|  94 | nah    | Nahuatl languages        | 38            | 279             | 2.4 kB           |
+|  95 | ne     | Nepali                   | 1,152,156     | 278,901,036     | 4.9 GB           |
+|  96 | new    | Newari                   | 1,996         | 229,703         | 4.0 MB           |
+|  97 | no     | Norwegian                | 2,797,378     | 373,160,033     | 2.6 GB           |
+|  98 | nn     | Norwegian Nynorsk        | 19,470        | 575,518         | 3.7 MB           |
+|  99 | oc     | Occitan                  | 920           | 34,701          | 405.0 kB         |
+| 100 | or     | Odia                     | 158,426       | 31,963,340      | 543.1 MB         |
+| 101 | os     | Ossetic                  | 8,628         | 3,935,964       | 50.7 MB          |
+| 102 | ps     | Pashto                   | 87,408        | 30,196,179      | 261.6 MB         |
+| 103 | fa     | Persian                  | 23,813,882    | 9,609,206,698   | 93.2 GB          |
+| 104 | pms    | Piedmontese              | 2,524         | 510,087         | 3.1 MB           |
+| 105 | pl     | Polish                   | 57,184,826    | 18,073,705,588  | 147.1 GB         |
+| 106 | pt     | Portuguese               | 36,062,800    | 15,172,557,311  | 105.0 GB         |
+| 107 | pa     | Punjabi                  | 222,058       | 104,235,418     | 1.4 GB           |
+| 108 | qu     | Quechua                  | 2             | 13              | 143 Bytes        |
+| 109 | ro     | Romanian                 | 11,985,668    | 6,302,600,833   | 45.6 GB          |
+| 110 | bxr    | Russia Buriat            | 72            | 698             | 8.2 kB           |
+| 111 | ru     | Russian                  | 194,143,422   | 78,032,029,344  | 1.1 TB           |
+| 112 | sah    | Sakha                    | 17,566        | 4,288,051       | 68.8 MB          |
+| 113 | sa     | Sanskrit                 | 16,802        | 2,479,345       | 56.3 MB          |
+| 114 | gd     | Scottish Gaelic          | 776           | 18,458          | 146.1 kB         |
+| 115 | sr     | Serbian                  | 1,677,896     | 632,781,822     | 7.7 GB           |
+| 116 | sh     | Serbian (Latin)          | 3,214         | 166,517         | 816.4 kB         |
+| 117 | sd     | Sindhi                   | 48,566        | 14,667,207      | 131.6 MB         |
+| 118 | si     | Sinhala                  | 301,066       | 172,755,385     | 2.6 GB           |
+| 119 | sk     | Slovak                   | 8,931,784     | 2,704,716,280   | 21.5 GB          |
+| 120 | sl     | Slovenian                | 1,112,560     | 192,816,743     | 1.4 GB           |
+| 121 | so     | Somali                   | 6             | 51              | 503 Bytes        |
+| 122 | azb    | South Azerbaijani        | 26,364        | 2,029,729       | 28.4 MB          |
+| 123 | es     | Spanish                  | 153,574,556   | 63,388,237,965  | 429.9 GB         |
+| 124 | su     | Sundanese                | 18            | 258             | 2.0 kB           |
+| 125 | sw     | Swahili                  | 1,664         | 164,459         | 1.0 MB           |
+| 126 | sv     | Swedish                  | 21,891,348    | 6,993,719,601   | 50.0 GB          |
+| 127 | gsw    | Swiss German             | 342           | 34,328          | 232.7 kB         |
+| 128 | tg     | Tajik                    | 144,932       | 76,987,285      | 1.0 GB           |
+| 129 | ta     | Tamil                    | 1,638,238     | 738,824,392     | 15.8 GB          |
+| 130 | tt     | Tatar                    | 262,654       | 59,253,765      | 833.8 MB         |
+| 131 | te     | Telugu                   | 644,712       | 201,575,815     | 3.9 GB           |
+| 132 | th     | Thai                     | 14,845,900    | 2,224,483,018   | 92.0 GB          |
+| 133 | bo     | Tibetan                  | 62,352        | 6,062,558       | 531.6 MB         |
+| 134 | tr     | Turkish                  | 26,654,330    | 8,290,890,087   | 73.7 GB          |
+| 135 | tk     | Turkmen                  | 4,576         | 325,786         | 3.3 MB           |
+| 136 | uk     | Ukrainian                | 10,059,992    | 3,183,842,018   | 44.7 GB          |
+| 137 | x-eml  | Emiliano-Romagnol        | 4             | 329             | 1.8 kB           |
+| 138 | hsb    | Upper Sorbian            | 402           | 15,827          | 123.2 kB         |
+| 139 | ur     | Urdu                     | 887,004       | 434,023,273     | 3.8 GB           |
+| 140 | ug     | Uyghur                   | 51,304        | 14,659,554      | 219.8 MB         |
+| 141 | uz     | Uzbek                    | 15,806        | 1,665,960       | 15.3 MB          |
+| 142 | vi     | Vietnamese               | 33,933,994    | 22,424,984,210  | 140.8 GB         |
+| 143 | vo     | Volapük                  | 896           | 49,968          | 371.9 kB         |
+| 144 | wa     | Walloon                  | 390           | 6,347           | 34.3 kB          |
+| 145 | war    | Waray                    | 1,494         | 19,665          | 126.8 kB         |
+| 146 | cy     | Welsh                    | 151,512       | 52,250,043      | 333.0 MB         |
+| 147 | fy     | Western Frisian          | 45,458        | 9,885,788       | 70.4 MB          |
+| 148 | mrj    | Western Mari             | 496           | 60,180          | 765.8 kB         |
+| 149 | pnb    | Western Panjabi          | 12,904        | 11,844,695      | 105.8 MB         |
+| 150 | wuu    | Wu Chinese               | 136           | 1,199           | 26.8 kB          |
+| 151 | yi     | Yiddish                  | 47,438        | 14,287,370      | 171.7 MB         |
+| 152 | yo     | Yoruba                   | 128           | 2,396           | 16.6 kB          |
+
+
+## Dataset Creation
+
+### Curation Rationale
+
+OSCAR was constructed using [`Ungoliant`](https://github.com/oscar-corpus/ungoliant), a new pipeline derived from [goclassy](https://github.com/oscar-corpus/goclassy), itself being derived from [fastText's one](https://github.com/facebookresearch/fastText).
+
+The pipeline works on documents rather than lines. 
+`Ungoliant` is implemented in the [Rust programming language](https://rust-lang.org), and uses [rayon](https://github.com/rayon-rs/rayon) as its data parallelism strategy. 
+Threading is done at shard, record and sentence level, making the whole generation process much more efficient.
+
+Filtering will be explained in a future blog post at our [website](https://oscar-corpus.com)
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+[Common Crawl](https://commoncrawl.org/) is a non-profit foundation which produces and maintains an open repository of web crawled data that is both accessible and analysable. Common Crawl's complete web archive consists of petabytes of data collected over 8 years of web crawling. The repository contains raw web page HTML data (WARC files), metdata extracts (WAT files) and plain text extracts (WET files). The organisation's crawlers has always respected [nofollow](http://microformats.org/wiki/rel-nofollow) and [robots.txt](https://www.robotstxt.org/) policies.
+
+Each monthly Common Crawl snapshot is in itself a massive multilingual corpus, where every single file contains data coming from multiple web pages written in a large variety of languages and covering all possible types of topics.
+
+To construct OSCAR the WET files of Common Crawl were used. These contain the extracted plain texts from the websites mostly converted to UTF-8, as well as headers containing the metatada of each crawled document. Each WET file comes compressed in gzip format and is stored on Amazon Web Services. In the case of OSCAR 22.01, the **November/December 2021** snapshot was used. It is composed by 64 000 compressed text files containing documents and their headers.
+
+#### Who are the source language producers?
+
+The data comes from multiple web pages in a large variety of languages.
+
+### Annotations
+
+The dataset does not contain any additional annotations.
+
+#### Annotation process
+
+N/A
+
+#### Who are the annotators?
+
+N/A
+
+### Personal and Sensitive Information
+
+Being constructed from Common Crawl, Personal and sensitive information might be present. This **must** be considered before training deep learning models with OSCAR, specially in the case of text-generation models.
+
+## Considerations for Using the Data
+
+### Social Impact of Dataset
+
+OSCAR is intended to bring more data to a wide variety of lanuages, the aim of the corpus is to make large amounts of data available to lower resource languages in order to facilitate the pre-training of state-of-the-art language modeling architectures.
+
+### Discussion of Biases
+
+OSCAR is not properly filtered yet and this can be reflected on the models trained with it. Care is advised specially concerning biases of the resulting models.
+
+### Other Known Limitations
+
+The [fastText linear classifier](https://fasttext.cc) is limed both in performance and the variety of languages it can recognize, so the quality of some OSCAR sub-corpora might be lower than expected, specially for the lowest-resource langiuages. Some audits have already been done by [third parties](https://arxiv.org/abs/2010.14571).
+
+## Additional Information
+
+### Dataset Curators
+
+This release of OSCAR was made possible by [Julien Abadji](https://ujj.space), [Pedro Ortiz Suarez](https://portizs.eu/), [Rua Ismail](https://oscar-project.org/authors/rua/), [Sotaro Takeshita](https://sotaro.io/about), [Sebastian Nagel](https://www.polver.uni-konstanz.de/cnc/people/nagel/) and [Benoit Sagot](http://pauillac.inria.fr/~sagot/).
+
+### Licensing Information
+
+    These data are released under this licensing scheme
+    We do not own any of the text from which these data has been extracted.
+    We license the actual packaging, the metadata and the annotations of these data under the Creative Commons CC0 license ("no rights reserved") http://creativecommons.org/publicdomain/zero/1.0/
+    To the extent possible under law, the OSCAR project, Inria, the Univertity of Mannheim and DFKI GmbH have waived all copyright and related or neighboring rights to OSCAR
+    This work is published from: France and Germany.
+
+    Should you consider that our data contains material that is owned by you and should therefore not be reproduced here, please:
+    * Clearly identify yourself, with detailed contact data such as an address, telephone number or email address at which you can be contacted.
+    * Clearly identify the copyrighted work claimed to be infringed.
+    * Clearly identify the material that is claimed to be infringing and information reasonably sufficient to allow us to locate the material.
+
+    We will comply to legitimate requests by removing the affected sources from the next release of the corpus.
+
+### Citation Information
+
+```
+@ARTICLE{2022arXiv221210440J,
+       author = {{Jansen}, Tim and {Tong}, Yangling and {Zevallos}, Victoria and {Ortiz Suarez}, Pedro},
+        title = "{Perplexed by Quality: A Perplexity-based Method for Adult and Harmful Content Detection in Multilingual Heterogeneous Web Data}",
+      journal = {arXiv e-prints},
+     keywords = {Computer Science - Computation and Language},
+         year = 2022,
+        month = dec,
+          eid = {arXiv:2212.10440},
+        pages = {arXiv:2212.10440},
+          doi = {10.48550/arXiv.2212.10440},
+archivePrefix = {arXiv},
+       eprint = {2212.10440},
+ primaryClass = {cs.CL},
+       adsurl = {https://ui.adsabs.harvard.edu/abs/2022arXiv221210440J},
+      adsnote = {Provided by the SAO/NASA Astrophysics Data System}
+}
+
+@inproceedings{abadji-etal-2022-towards,
+    title = "Towards a Cleaner Document-Oriented Multilingual Crawled Corpus",
+    author = "Abadji, Julien  and
+      Ortiz Suarez, Pedro  and
+      Romary, Laurent  and
+      Sagot, Beno{\^\i}t",
+    booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference",
+    month = jun,
+    year = "2022",
+    address = "Marseille, France",
+    publisher = "European Language Resources Association",
+    url = "https://aclanthology.org/2022.lrec-1.463",
+    pages = "4344--4355",
+    abstract = "The need for large corpora raw corpora has dramatically increased in recent years with the introduction of transfer learning and semi-supervised learning methods to Natural Language Processing. And while there have been some recent attempts to manually curate the amount of data necessary to train large language models, the main way to obtain this data is still through automatic web crawling. In this paper we take the existing multilingual web corpus OSCAR and its pipeline Ungoliant that extracts and classifies data from Common Crawl at the line level, and propose a set of improvements and automatic annotations in order to produce a new document-oriented version of OSCAR that could prove more suitable to pre-train large generative language models as well as hopefully other applications in Natural Language Processing and Digital Humanities.",
+}
+
+      
+@inproceedings{AbadjiOrtizSuarezRomaryetal.2021,
+  author    = {Julien Abadji and Pedro Javier Ortiz Su{\'a}rez and Laurent Romary and Beno{\^i}t Sagot},
+  title     = {Ungoliant: An optimized pipeline for the generation of a very large-scale multilingual web corpus},
+  series = {Proceedings of the Workshop on Challenges in the Management of Large Corpora (CMLC-9) 2021. Limerick, 12 July 2021 (Online-Event)},
+  editor    = {Harald L{\"u}ngen and Marc Kupietz and Piotr Bański and Adrien Barbaresi and Simon Clematide and Ines Pisetta},
+  publisher = {Leibniz-Institut f{\"u}r Deutsche Sprache},
+  address   = {Mannheim},
+  doi       = {10.14618/ids-pub-10468},
+  url       = {https://nbn-resolving.org/urn:nbn:de:bsz:mh39-104688},
+  pages     = {1 -- 9},
+  year      = {2021},
+  abstract  = {Since the introduction of large language models in Natural Language Processing, large raw corpora have played a crucial role in Computational Linguistics. However, most of these large raw corpora are either available only for English or not available to the general public due to copyright issues. Nevertheless, there are some examples of freely available multilingual corpora for training Deep Learning NLP models, such as the OSCAR and Paracrawl corpora. However, they have quality issues, especially for low-resource languages. Moreover, recreating or updating these corpora is very complex. In this work, we try to reproduce and improve the goclassy pipeline used to create the OSCAR corpus. We propose a new pipeline that is faster, modular, parameterizable, and well documented. We use it to create a corpus similar to OSCAR but larger and based on recent data. Also, unlike OSCAR, the metadata information is at the document level. We release our pipeline under an open source license and publish the corpus under a research-only license.},
+  language  = {en}
+}
+
+@article{kreutzer-etal-2022-quality,
+    title = "Quality at a Glance: An Audit of Web-Crawled Multilingual Datasets",
+    author = {Kreutzer, Julia  and
+      Caswell, Isaac  and
+      Wang, Lisa  and
+      Wahab, Ahsan  and
+      van Esch, Daan  and
+      Ulzii-Orshikh, Nasanbayar  and
+      Tapo, Allahsera  and
+      Subramani, Nishant  and
+      Sokolov, Artem  and
+      Sikasote, Claytone  and
+      Setyawan, Monang  and
+      Sarin, Supheakmungkol  and
+      Samb, Sokhar  and
+      Sagot, Beno{\^\i}t  and
+      Rivera, Clara  and
+      Rios, Annette  and
+      Papadimitriou, Isabel  and
+      Osei, Salomey  and
+      Suarez, Pedro Ortiz  and
+      Orife, Iroro  and
+      Ogueji, Kelechi  and
+      Rubungo, Andre Niyongabo  and
+      Nguyen, Toan Q.  and
+      M{\"u}ller, Mathias  and
+      M{\"u}ller, Andr{\'e}  and
+      Muhammad, Shamsuddeen Hassan  and
+      Muhammad, Nanda  and
+      Mnyakeni, Ayanda  and
+      Mirzakhalov, Jamshidbek  and
+      Matangira, Tapiwanashe  and
+      Leong, Colin  and
+      Lawson, Nze  and
+      Kudugunta, Sneha  and
+      Jernite, Yacine  and
+      Jenny, Mathias  and
+      Firat, Orhan  and
+      Dossou, Bonaventure F. P.  and
+      Dlamini, Sakhile  and
+      de Silva, Nisansa  and
+      {\c{C}}abuk Ball{\i}, Sakine  and
+      Biderman, Stella  and
+      Battisti, Alessia  and
+      Baruwa, Ahmed  and
+      Bapna, Ankur  and
+      Baljekar, Pallavi  and
+      Azime, Israel Abebe  and
+      Awokoya, Ayodele  and
+      Ataman, Duygu  and
+      Ahia, Orevaoghene  and
+      Ahia, Oghenefego  and
+      Agrawal, Sweta  and
+      Adeyemi, Mofetoluwa},
+    journal = "Transactions of the Association for Computational Linguistics",
+    volume = "10",
+    year = "2022",
+    address = "Cambridge, MA",
+    publisher = "MIT Press",
+    url = "https://aclanthology.org/2022.tacl-1.4",
+    doi = "10.1162/tacl_a_00447",
+    pages = "50--72",
+    abstract = "With the success of large-scale pre-training and multilingual modeling in Natural Language Processing (NLP), recent years have seen a proliferation of large, Web-mined text datasets covering hundreds of languages. We manually audit the quality of 205 language-specific corpora released with five major public datasets (CCAligned, ParaCrawl, WikiMatrix, OSCAR, mC4). Lower-resource corpora have systematic issues: At least 15 corpora have no usable text, and a significant fraction contains less than 50{\%} sentences of acceptable quality. In addition, many are mislabeled or use nonstandard/ambiguous language codes. We demonstrate that these issues are easy to detect even for non-proficient speakers, and supplement the human audit with automatic analyses. Finally, we recommend techniques to evaluate and improve multilingual corpora and discuss potential risks that come with low-quality data releases.",
+}
+
+@inproceedings{ortiz-suarez-etal-2020-monolingual,
+    title = "A Monolingual Approach to Contextualized Word Embeddings for Mid-Resource Languages",
+    author = "Ortiz Su{'a}rez, Pedro Javier  and
+      Romary, Laurent  and
+      Sagot, Benoit",
+    booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
+    month = jul,
+    year = "2020",
+    address = "Online",
+    publisher = "Association for Computational Linguistics",
+    url = "https://www.aclweb.org/anthology/2020.acl-main.156",
+    pages = "1703--1714",
+    abstract = "We use the multilingual OSCAR corpus, extracted from Common Crawl via language classification, filtering and cleaning, to train monolingual contextualized word embeddings (ELMo) for five mid-resource languages. We then compare the performance of OSCAR-based and Wikipedia-based ELMo embeddings for these languages on the part-of-speech tagging and parsing tasks. We show that, despite the noise in the Common-Crawl-based OSCAR data, embeddings trained on OSCAR perform much better than monolingual embeddings trained on Wikipedia. They actually equal or improve the current state of the art in tagging and parsing for all five languages. In particular, they also improve over multilingual Wikipedia-based contextual embeddings (multilingual BERT), which almost always constitutes the previous state of the art, thereby showing that the benefit of a larger, more diverse corpus surpasses the cross-lingual benefit of multilingual embedding architectures.",
+}
+
+@inproceedings{OrtizSuarezSagotRomary2019,
+  author    = {Pedro Javier {Ortiz Su{'a}rez} and Benoit Sagot and Laurent Romary},
+  title     = {Asynchronous pipelines for processing huge corpora on medium to low resource infrastructures},
+  series = {Proceedings of the Workshop on Challenges in the Management of Large Corpora (CMLC-7) 2019. Cardiff, 22nd July 2019},
+  editor    = {Piotr Bański and Adrien Barbaresi and Hanno Biber and Evelyn Breiteneder and Simon Clematide and Marc Kupietz and Harald L{"u}ngen and Caroline Iliadi},
+  publisher = {Leibniz-Institut f{"u}r Deutsche Sprache},
+  address   = {Mannheim},
+  doi       = {10.14618/ids-pub-9021},
+  url       = {http://nbn-resolving.de/urn:nbn:de:bsz:mh39-90215},
+  pages     = {9 -- 16},
+  year      = {2019},
+  abstract  = {Common Crawl is a considerably large, heterogeneous multilingual corpus comprised of crawled documents from the internet, surpassing 20TB of data and distributed as a set of more than 50 thousand plain text files where each contains many documents written in a wide variety of languages. Even though each document has a metadata block associated to it, this data lacks any information about the language in which each document is written, making it extremely difficult to use Common Crawl for monolingual applications. We propose a general, highly parallel, multithreaded pipeline to clean and classify Common Crawl by language; we specifically design it so that it runs efficiently on medium to low resource infrastructures where I/O speeds are the main constraint. We develop the pipeline so that it can be easily reapplied to any kind of heterogeneous corpus and so that it can be parameterised to a wide range of infrastructures. We also distribute a 6.3TB version of Common Crawl, filtered, classified by language, shuffled at line level in order to avoid copyright issues, and ready to be used for NLP applications.},
+  language  = {en}
+}
+
+```

artifacts/hf_readmes/oscar-corpus__mOSCAR__README.md

@@ -0,0 +1,697 @@
+---
+license: cc-by-4.0
+configs:
+- config_name: "ace_Latn"
+  data_files:
+  - split: train
+    path: data/ace_Latn/*
+- config_name: "acm_Arab"
+  data_files:
+  - split: train
+    path: data/acm_Arab/*
+- config_name: "aeb_Arab"
+  data_files:
+  - split: train
+    path: data/aeb_Arab/*
+- config_name: "afr_Latn"
+  data_files:
+  - split: train
+    path: data/afr_Latn/*
+- config_name: "ajp_Arab"
+  data_files:
+  - split: train
+    path: data/ajp_Arab/*
+- config_name: "als_Latn"
+  data_files:
+  - split: train
+    path: data/als_Latn/*
+- config_name: "amh_Ethi"
+  data_files:
+  - split: train
+    path: data/amh_Ethi/*
+- config_name: "apc_Arab"
+  data_files:
+  - split: train
+    path: data/apc_Arab/*
+- config_name: "arb_Arab"
+  data_files:
+  - split: train
+    path: data/arb_Arab/*
+- config_name: "ars_Arab"
+  data_files:
+  - split: train
+    path: data/ars_Arab/*
+- config_name: "ary_Arab"
+  data_files:
+  - split: train
+    path: data/ary_Arab/*
+- config_name: "arz_Arab"
+  data_files:
+  - split: train
+    path: data/arz_Arab/*
+- config_name: "asm_Beng"
+  data_files:
+  - split: train
+    path: data/asm_Beng/*
+- config_name: "ast_Latn"
+  data_files:
+  - split: train
+    path: data/ast_Latn/*
+- config_name: "awa_Deva"
+  data_files:
+  - split: train
+    path: data/awa_Deva/*
+- config_name: "ayr_Latn"
+  data_files:
+  - split: train
+    path: data/ayr_Latn/*
+- config_name: "azb_Arab"
+  data_files:
+  - split: train
+    path: data/azb_Arab/*
+- config_name: "azj_Latn"
+  data_files:
+  - split: train
+    path: data/azj_Latn/*
+- config_name: "bak_Cyrl"
+  data_files:
+  - split: train
+    path: data/bak_Cyrl/*
+- config_name: "bam_Latn"
+  data_files:
+  - split: train
+    path: data/bam_Latn/*
+- config_name: "ban_Latn"
+  data_files:
+  - split: train
+    path: data/ban_Latn/*
+- config_name: "bel_Cyrl"
+  data_files:
+  - split: train
+    path: data/bel_Cyrl/*
+- config_name: "bem_Latn"
+  data_files:
+  - split: train
+    path: data/bem_Latn/*
+- config_name: "ben_Beng"
+  data_files:
+  - split: train
+    path: data/ben_Beng/*
+- config_name: "bho_Deva"
+  data_files:
+  - split: train
+    path: data/bho_Deva/*
+- config_name: "bjn_Latn"
+  data_files:
+  - split: train
+    path: data/bjn_Latn/*
+- config_name: "bos_Latn"
+  data_files:
+  - split: train
+    path: data/bos_Latn/*
+- config_name: "bug_Latn"
+  data_files:
+  - split: train
+    path: data/bug_Latn/*
+- config_name: "bul_Cyrl"
+  data_files:
+  - split: train
+    path: data/bul_Cyrl/*
+- config_name: "cat_Latn"
+  data_files:
+  - split: train
+    path: data/cat_Latn/*
+- config_name: "ceb_Latn"
+  data_files:
+  - split: train
+    path: data/ceb_Latn/*
+- config_name: "ces_Latn"
+  data_files:
+  - split: train
+    path: data/ces_Latn/*
+- config_name: "ckb_Arab"
+  data_files:
+  - split: train
+    path: data/ckb_Arab/*
+- config_name: "crh_Latn"
+  data_files:
+  - split: train
+    path: data/crh_Latn/*
+- config_name: "cym_Latn"
+  data_files:
+  - split: train
+    path: data/cym_Latn/*
+- config_name: "dan_Latn"
+  data_files:
+  - split: train
+    path: data/dan_Latn/*
+- config_name: "deu_Latn"
+  data_files:
+  - split: train
+    path: data/deu_Latn/*
+- config_name: "dik_Latn"
+  data_files:
+  - split: train
+    path: data/dik_Latn/*
+- config_name: "ell_Grek"
+  data_files:
+  - split: train
+    path: data/ell_Grek/*
+- config_name: "eng_Latn"
+  data_files:
+  - split: train
+    path: data/eng_Latn/*
+- config_name: "epo_Latn"
+  data_files:
+  - split: train
+    path: data/epo_Latn/*
+- config_name: "est_Latn"
+  data_files:
+  - split: train
+    path: data/est_Latn/*
+- config_name: "eus_Latn"
+  data_files:
+  - split: train
+    path: data/eus_Latn/*
+- config_name: "fao_Latn"
+  data_files:
+  - split: train
+    path: data/fao_Latn/*
+- config_name: "fij_Latn"
+  data_files:
+  - split: train
+    path: data/fij_Latn/*
+- config_name: "fin_Latn"
+  data_files:
+  - split: train
+    path: data/fin_Latn/*
+- config_name: "fra_Latn"
+  data_files:
+  - split: train
+    path: data/fra_Latn/*
+- config_name: "fur_Latn"
+  data_files:
+  - split: train
+    path: data/fur_Latn/*
+- config_name: "fuv_Latn"
+  data_files:
+  - split: train
+    path: data/fuv_Latn/*
+- config_name: "gaz_Latn"
+  data_files:
+  - split: train
+    path: data/gaz_Latn/*
+- config_name: "gla_Latn"
+  data_files:
+  - split: train
+    path: data/gla_Latn/*
+- config_name: "gle_Latn"
+  data_files:
+  - split: train
+    path: data/gle_Latn/*
+- config_name: "glg_Latn"
+  data_files:
+  - split: train
+    path: data/glg_Latn/*
+- config_name: "grn_Latn"
+  data_files:
+  - split: train
+    path: data/grn_Latn/*
+- config_name: "guj_Gujr"
+  data_files:
+  - split: train
+    path: data/guj_Gujr/*
+- config_name: "hat_Latn"
+  data_files:
+  - split: train
+    path: data/hat_Latn/*
+- config_name: "hau_Latn"
+  data_files:
+  - split: train
+    path: data/hau_Latn/*
+- config_name: "heb_Hebr"
+  data_files:
+  - split: train
+    path: data/heb_Hebr/*
+- config_name: "hin_Deva"
+  data_files:
+  - split: train
+    path: data/hin_Deva/*
+- config_name: "hne_Deva"
+  data_files:
+  - split: train
+    path: data/hne_Deva/*
+- config_name: "hrv_Latn"
+  data_files:
+  - split: train
+    path: data/hrv_Latn/*
+- config_name: "hun_Latn"
+  data_files:
+  - split: train
+    path: data/hun_Latn/*
+- config_name: "hye_Armn"
+  data_files:
+  - split: train
+    path: data/hye_Armn/*
+- config_name: "ibo_Latn"
+  data_files:
+  - split: train
+    path: data/ibo_Latn/*
+- config_name: "ilo_Latn"
+  data_files:
+  - split: train
+    path: data/ilo_Latn/*
+- config_name: "ind_Latn"
+  data_files:
+  - split: train
+    path: data/ind_Latn/*
+- config_name: "isl_Latn"
+  data_files:
+  - split: train
+    path: data/isl_Latn/*
+- config_name: "ita_Latn"
+  data_files:
+  - split: train
+    path: data/ita_Latn/*
+- config_name: "jav_Latn"
+  data_files:
+  - split: train
+    path: data/jav_Latn/*
+- config_name: "jpn_Jpan"
+  data_files:
+  - split: train
+    path: data/jpn_Jpan/*
+- config_name: "kab_Latn"
+  data_files:
+  - split: train
+    path: data/kab_Latn/*
+- config_name: "kan_Knda"
+  data_files:
+  - split: train
+    path: data/kan_Knda/*
+- config_name: "kas_Arab"
+  data_files:
+  - split: train
+    path: data/kas_Arab/*
+- config_name: "kat_Geor"
+  data_files:
+  - split: train
+    path: data/kat_Geor/*
+- config_name: "kaz_Cyrl"
+  data_files:
+  - split: train
+    path: data/kaz_Cyrl/*
+- config_name: "khk_Cyrl"
+  data_files:
+  - split: train
+    path: data/khk_Cyrl/*
+- config_name: "khm_Khmr"
+  data_files:
+  - split: train
+    path: data/khm_Khmr/*
+- config_name: "kin_Latn"
+  data_files:
+  - split: train
+    path: data/kin_Latn/*
+- config_name: "kir_Cyrl"
+  data_files:
+  - split: train
+    path: data/kir_Cyrl/*
+- config_name: "kmr_Latn"
+  data_files:
+  - split: train
+    path: data/kmr_Latn/*
+- config_name: "kor_Hang"
+  data_files:
+  - split: train
+    path: data/kor_Hang/*
+- config_name: "lao_Laoo"
+  data_files:
+  - split: train
+    path: data/lao_Laoo/*
+- config_name: "lij_Latn"
+  data_files:
+  - split: train
+    path: data/lij_Latn/*
+- config_name: "lim_Latn"
+  data_files:
+  - split: train
+    path: data/lim_Latn/*
+- config_name: "lin_Latn"
+  data_files:
+  - split: train
+    path: data/lin_Latn/*
+- config_name: "lit_Latn"
+  data_files:
+  - split: train
+    path: data/lit_Latn/*
+- config_name: "lmo_Latn"
+  data_files:
+  - split: train
+    path: data/lmo_Latn/*
+- config_name: "ltg_Latn"
+  data_files:
+  - split: train
+    path: data/ltg_Latn/*
+- config_name: "ltz_Latn"
+  data_files:
+  - split: train
+    path: data/ltz_Latn/*
+- config_name: "lug_Latn"
+  data_files:
+  - split: train
+    path: data/lug_Latn/*
+- config_name: "lus_Latn"
+  data_files:
+  - split: train
+    path: data/lus_Latn/*
+- config_name: "lvs_Latn"
+  data_files:
+  - split: train
+    path: data/lvs_Latn/*
+- config_name: "mag_Deva"
+  data_files:
+  - split: train
+    path: data/mag_Deva/*
+- config_name: "mal_Mlym"
+  data_files:
+  - split: train
+    path: data/mal_Mlym/*
+- config_name: "mar_Deva"
+  data_files:
+  - split: train
+    path: data/mar_Deva/*
+- config_name: "min_Latn"
+  data_files:
+  - split: train
+    path: data/min_Latn/*
+- config_name: "mkd_Cyrl"
+  data_files:
+  - split: train
+    path: data/mkd_Cyrl/*
+- config_name: "mlt_Latn"
+  data_files:
+  - split: train
+    path: data/mlt_Latn/*
+- config_name: "mri_Latn"
+  data_files:
+  - split: train
+    path: data/mri_Latn/*
+- config_name: "mya_Mymr"
+  data_files:
+  - split: train
+    path: data/mya_Mymr/*
+- config_name: "nld_Latn"
+  data_files:
+  - split: train
+    path: data/nld_Latn/*
+- config_name: "nno_Latn"
+  data_files:
+  - split: train
+    path: data/nno_Latn/*
+- config_name: "nob_Latn"
+  data_files:
+  - split: train
+    path: data/nob_Latn/*
+- config_name: "npi_Deva"
+  data_files:
+  - split: train
+    path: data/npi_Deva/*
+- config_name: "nya_Latn"
+  data_files:
+  - split: train
+    path: data/nya_Latn/*
+- config_name: "oci_Latn"
+  data_files:
+  - split: train
+    path: data/oci_Latn/*
+- config_name: "ory_Orya"
+  data_files:
+  - split: train
+    path: data/ory_Orya/*
+- config_name: "pag_Latn"
+  data_files:
+  - split: train
+    path: data/pag_Latn/*
+- config_name: "pan_Guru"
+  data_files:
+  - split: train
+    path: data/pan_Guru/*
+- config_name: "pap_Latn"
+  data_files:
+  - split: train
+    path: data/pap_Latn/*
+- config_name: "pbt_Arab"
+  data_files:
+  - split: train
+    path: data/pbt_Arab/*
+- config_name: "pes_Arab"
+  data_files:
+  - split: train
+    path: data/pes_Arab/*
+- config_name: "plt_Latn"
+  data_files:
+  - split: train
+    path: data/plt_Latn/*
+- config_name: "pol_Latn"
+  data_files:
+  - split: train
+    path: data/pol_Latn/*
+- config_name: "por_Latn"
+  data_files:
+  - split: train
+    path: data/por_Latn/*
+- config_name: "prs_Arab"
+  data_files:
+  - split: train
+    path: data/prs_Arab/*
+- config_name: "quy_Latn"
+  data_files:
+  - split: train
+    path: data/quy_Latn/*
+- config_name: "ron_Latn"
+  data_files:
+  - split: train
+    path: data/ron_Latn/*
+- config_name: "run_Latn"
+  data_files:
+  - split: train
+    path: data/run_Latn/*
+- config_name: "rus_Cyrl"
+  data_files:
+  - split: train
+    path: data/rus_Cyrl/*
+- config_name: "sag_Latn"
+  data_files:
+  - split: train
+    path: data/sag_Latn/*
+- config_name: "scn_Latn"
+  data_files:
+  - split: train
+    path: data/scn_Latn/*
+- config_name: "sin_Sinh"
+  data_files:
+  - split: train
+    path: data/sin_Sinh/*
+- config_name: "slk_Latn"
+  data_files:
+  - split: train
+    path: data/slk_Latn/*
+- config_name: "slv_Latn"
+  data_files:
+  - split: train
+    path: data/slv_Latn/*
+- config_name: "smo_Latn"
+  data_files:
+  - split: train
+    path: data/smo_Latn/*
+- config_name: "sna_Latn"
+  data_files:
+  - split: train
+    path: data/sna_Latn/*
+- config_name: "snd_Arab"
+  data_files:
+  - split: train
+    path: data/snd_Arab/*
+- config_name: "som_Latn"
+  data_files:
+  - split: train
+    path: data/som_Latn/*
+- config_name: "sot_Latn"
+  data_files:
+  - split: train
+    path: data/sot_Latn/*
+- config_name: "spa_Latn"
+  data_files:
+  - split: train
+    path: data/spa_Latn/*
+- config_name: "srd_Latn"
+  data_files:
+  - split: train
+    path: data/srd_Latn/*
+- config_name: "srp_Cyrl"
+  data_files:
+  - split: train
+    path: data/srp_Cyrl/*
+- config_name: "sun_Latn"
+  data_files:
+  - split: train
+    path: data/sun_Latn/*
+- config_name: "swe_Latn"
+  data_files:
+  - split: train
+    path: data/swe_Latn/*
+- config_name: "swh_Latn"
+  data_files:
+  - split: train
+    path: data/swh_Latn/*
+- config_name: "szl_Latn"
+  data_files:
+  - split: train
+    path: data/szl_Latn/*
+- config_name: "tam_Taml"
+  data_files:
+  - split: train
+    path: data/tam_Taml/*
+- config_name: "tat_Cyrl"
+  data_files:
+  - split: train
+    path: data/tat_Cyrl/*
+- config_name: "tel_Telu"
+  data_files:
+  - split: train
+    path: data/tel_Telu/*
+- config_name: "tgk_Cyrl"
+  data_files:
+  - split: train
+    path: data/tgk_Cyrl/*
+- config_name: "tgl_Latn"
+  data_files:
+  - split: train
+    path: data/tgl_Latn/*
+- config_name: "tha_Thai"
+  data_files:
+  - split: train
+    path: data/tha_Thai/*
+- config_name: "tir_Ethi"
+  data_files:
+  - split: train
+    path: data/tir_Ethi/*
+- config_name: "tpi_Latn"
+  data_files:
+  - split: train
+    path: data/tpi_Latn/*
+- config_name: "tuk_Latn"
+  data_files:
+  - split: train
+    path: data/tuk_Latn/*
+- config_name: "tur_Latn"
+  data_files:
+  - split: train
+    path: data/tur_Latn/*
+- config_name: "twi_Latn"
+  data_files:
+  - split: train
+    path: data/twi_Latn/*
+- config_name: "uig_Arab"
+  data_files:
+  - split: train
+    path: data/uig_Arab/*
+- config_name: "ukr_Cyrl"
+  data_files:
+  - split: train
+    path: data/ukr_Cyrl/*
+- config_name: "urd_Arab"
+  data_files:
+  - split: train
+    path: data/urd_Arab/*
+- config_name: "uzn_Latn"
+  data_files:
+  - split: train
+    path: data/uzn_Latn/*
+- config_name: "vec_Latn"
+  data_files:
+  - split: train
+    path: data/vec_Latn/*
+- config_name: "vie_Latn"
+  data_files:
+  - split: train
+    path: data/vie_Latn/*
+- config_name: "wol_Latn"
+  data_files:
+  - split: train
+    path: data/wol_Latn/*
+- config_name: "xho_Latn"
+  data_files:
+  - split: train
+    path: data/xho_Latn/*
+- config_name: "ydd_Hebr"
+  data_files:
+  - split: train
+    path: data/ydd_Hebr/*
+- config_name: "yor_Latn"
+  data_files:
+  - split: train
+    path: data/yor_Latn/*
+- config_name: "yue_Hant"
+  data_files:
+  - split: train
+    path: data/yue_Hant/*
+- config_name: "zho_Hans"
+  data_files:
+  - split: train
+    path: data/zho_Hans/*
+- config_name: "zho_Hant"
+  data_files:
+  - split: train
+    path: data/zho_Hant/*
+- config_name: "zsm_Latn"
+  data_files:
+  - split: train
+    path: data/zsm_Latn/*
+- config_name: "zul_Latn"
+  data_files:
+  - split: train
+    path: data/zul_Latn/*
+---
+
+More info can be found here: https://oscar-project.github.io/documentation/versions/mOSCAR/
+
+Paper link: https://arxiv.org/abs/2406.08707
+
+**New features:**
+- Additional filtering steps were applied to remove toxic content (more details in the next version of the paper, coming soon).
+- Spanish split is now complete.
+- Face detection in images to blur them once downloaded (coordinates are reported on images of size 256 respecting aspect ratio).
+- Additional language identification of the documents to improve document-language matching.
+- Replace most of Personal Identifiable Information by generic strings.
+
+Previous version remains available, to continue using it:
+```
+dataset = load_dataset("oscar-corpus/mOSCAR", revision="v1")
+```
+
+# Layout
+```
+{
+    'images': [{'img_idx': '#000002',
+                'sha512': '65c1e5605d48f8753256f758bd442cbdd43e6987691227b1ea6b81430ff36609f46d448c8171546232fe0c258d9e44ce4378f32e8ada5c43c314df5a5e230de2',
+                'url': 'https://actuconsommation.fr/wp-content/uploads/2020/05/Disneylands-Japon-1068x712.jpg',
+                'faces_loc': [x0, y0, x1, y1]}],
+    'metadata': [{'node_order': 'img_#000002|txt_#000000|txt_#000001|txt_#000002|txt_#000003|txt_#000004|txt_#000005|txt_#000006|txt_#000009',
+                  'url': 'https://actuconsommation.fr/2020/05/11/disneyland-une-reouverture-sous-haute-securite-a-shanghai-ce-lundi/'}],
+    'text': [{'text': 'Disneyland : une réouverture sous haute sécurité à Shanghai ce lundi', 'text_idx': '#000000'},
+             {'text': 'Des milliers de visiteurs ont pu pénétrer lundi dans le Disneyland de Shanghai, le premier des six parcs de [...]', text_idx': '#000001'},
+             [...] ]
+}
+```
+
+# Citation
+```
+@article{futeral2024moscar,
+  title={mOSCAR: A Large-scale Multilingual and Multimodal Document-level Corpus},
+  author={Futeral, Matthieu and Zebaze, Armel and Suarez, Pedro Ortiz and Abadji, Julien and Lacroix, R{\'e}mi and Schmid, Cordelia and Bawden, Rachel and Sagot, Beno{\^\i}t},
+  journal={arXiv preprint arXiv:2406.08707},
+  year={2024}
+}
+```

artifacts/hf_readmes/pelcra__PLLuMIC__README.md

@@ -0,0 +1,174 @@
+---
+license: cc-by-sa-4.0
+task_categories:
+- text-generation
+language:
+- pl
+tags:
+- llm
+- sft
+- fine-tuning
+pretty_name: PLLuM Instruction Corpus
+size_categories:
+- 1K<n<10K
+---
+
+# Dataset Card for PLLuMIC
+
+PLLuMIC - Polish Large Language Model (PLLuM) Instruction Corpus
+
+## Dataset Details
+
+### Dataset Description
+
+We release the first representative subset of the PLLuM Instruction Corpus (PLLuMIC), which we believe to be useful in guiding and planning the development of similar LLM datasets. PLLuMIC is a hand-crafted set of LLM fine-tuning Polish language instructions, developed in line with the annotation guidelines and covering a functional typology. Each instruction is designed to be unique in some way - no two samples, especially within given subtype, are the same. Every single row provides a new insight into the dataset, making the corpus ideal for extensive analysis. The methodology is described in more detail in a paper titled The PLLuM Instruction Corpus (link below). We plan regular updates and significant extensions of the corpus.
+
+
+- **Curated by:** PELCRA (Polish and English Language Corpora for Research and Applications) Team
+- **Language(s) (NLP):** Polish
+- **License:** CC-BY-SA-4.0
+
+### Dataset Sources
+
+- **Paper:** https://arxiv.org/abs/2511.17161
+
+## Uses
+
+### Direct Use
+
+We believe the dataset to be useful in guiding and planning the development of similar, bigger, LLM datasets. This first sample is designed to be a representative guidance, on how to properly structure and build your own dataset.
+
+It is also a great foundation for synthetic extensions that will combine high quality, diversity and scale. We are currently working on such corpus extension ourselves and are planning to make it available alongside this organic component.
+
+### Out-of-Scope Use
+
+Current scale of the dataset will not be sufficient to perform a full LLM fine-tuning. However, with only 10k synthetic samples that are based around the corpus, one can already expect very interesting results. We will provide more details (and data) on that topic in future updates.
+
+## Dataset Structure
+
+### Statistics
+
+**Total instructions:** 1,278
+
+All instructions were annotated by professional annotators. Each sample was developed in accordance with comprehensive annotation guidelines and subsequently reviewed by a senior annotator to ensure full compliance with quality standards. The annotation process followed a functional typology designed to encompass key areas of model competence.
+
+There are both single-turn and multi-turn instructions available.
+
+#### Type & Thematic distributions
+
+<table>
+  <tr>
+    <td>
+
+| Type | Number of samples |
+|---|---|
+| Generation | 392 |
+| Adversarial | 125 |
+| Dialogue | 124 |
+| NLP | 102 |
+| Data manipulation | 88 |
+| Formatting | 87 |
+| Knowledge (QA) | 80 |
+| Extraction | 71 |
+| Identity | 68 |
+| Translation | 61 |
+| CoT | 50 |
+| Programming | 30 |
+
+</td>
+<td>
+
+| Topic | Number of samples |
+|---|---|
+| Languages | 185 |
+| Society | 169 |
+| Computer science | 163 |
+| Technology | 87 |
+| Entertainment | 85 |
+| Biology | 78 |
+| Other | 73 |
+| Home | 60 |
+| Geography | 59 |
+| Culture | 55 |
+| Culinary | 52 |
+| Literature | 50 |
+| History | 48 |
+| Politics | 42 |
+| Medicine | 36 |
+| Law and administration | 31 |
+| Sports | 26 |
+| Travel | 25 |
+| Industry | 20 |
+| Economy | 19 |
+| Psychology | 19 |
+| Mathematics | 15 |
+| Art | 14 |
+| Physics | 8 |
+| Chemistry | 7 |
+| Religion | 7 |
+| Automotive | 6 |
+| Philosophy | 5 |
+| Astronomy | 5 |
+| Ecology | 4 |
+| Hobby | 4 |
+
+</td>
+  </tr>
+</table>
+
+### Data format explanation
+
+The PLLuMIC dataset is distributed as a JSON file storing rows with conversations between a user and an AI assistant. Each conversation is a JSON structure described by following fields:
+
+#### Top-Level Fields
+- dataset_name: Name of the dataset (PLLuMIC).
+- dataset_source: Source organization (CLARIN-BIZ-bis).
+- conv_id: Unique identifier for the conversation (3242183cbce2).
+- messages: Array of dialogue messages (user/assistant/system exchanges).
+
+#### Message Object Fields
+Each entry in messages contains:
+
+- instruction_id: Unique ID for the instruction/task (2a07c2eca0cb).
+- seq: Sequence number (-1 for system, 0,1,2,… for user/assistant turns).
+- role: Speaker role (system, user, or assistant).
+- content: Text of the message (empty for some system prompts).
+- type: Interaction type (e.g., Dialog, Generation).
+- subtype: List of task subtype (e.g., [System prompt, Text simplification]).
+- topic: List of relevant topics (e.g., [Geography]).
+- language: Language code (e.g., pol for Polish).
+- source: References (e.g., Wikipedia URLs).
+
+## Dataset Creation
+
+### Curation Rationale
+
+Most instruction-tuning datasets for LLMs are either private or poorly documented, making it hard to understand how models are trained or to build comparable resources. Even when public, such datasets often mix data from many sources without clear structure or balance.
+
+There’s also little research on how different instruction types shape model behavior, and while distilling data from strong LLMs is common, it doesn’t always transfer well across languages and cultures.
+
+That’s why we created this dataset — to offer a transparent, well-documented, and balanced resource for instruction tuning, designed with linguistic and cultural diversity in mind. The results and findings are well-described in the paper linked above.
+
+### Annotation
+
+#### Annotation process
+
+All instructions were annotated by professional annotators. Each sample was developed in accordance with comprehensive annotation guidelines and subsequently reviewed by a senior annotator to ensure full compliance with quality standards. The annotation process followed a functional typology designed to encompass key areas of model competence.
+
+#### Who are the annotators?
+
+All annotators (over 50 in total) were university graduates, with at least a bachelor’s or master’s degree in linguistics or other humanities with the exception of technical instructions annotators who had a university degree in computer science. All of the super-annotators had a PhD degree.
+
+## Citation
+
+```bibtex
+@misc{pęzik2025plluminstructioncorpus,
+      title={The PLLuM Instruction Corpus}, 
+      author={Piotr Pęzik and Filip Żarnecki and Konrad Kaczyński and Anna Cichosz and Zuzanna Deckert and Monika Garnys and Izabela Grabarczyk and Wojciech Janowski and Sylwia Karasińska and Aleksandra Kujawiak and Piotr Misztela and Maria Szymańska and Karolina Walkusz and Igor Siek and Maciej Chrabąszcz and Anna Kołos and Agnieszka Karlińska and Karolina Seweryn and Aleksandra Krasnodębska and Paula Betscher and Zofia Cieślińska and Katarzyna Kowol and Artur Wilczek and Maciej Trzciński and Katarzyna Dziewulska and Roman Roszko and Tomasz Bernaś and Jurgita Vaičenonienė and Danuta Roszko and Paweł Levchuk and Paweł Kowalski and Irena Prawdzic-Jankowska and Marek Kozłowski and Sławomir Dadas and Rafał Poświata and Alina Wróblewska and Katarzyna Krasnowska-Kieraś and Maciej Ogrodniczuk and Michał Rudolf and Piotr Rybak and Karolina Saputa and Joanna Wołoszyn and Marcin Oleksy and Bartłomiej Koptyra and Teddy Ferdinan and Stanisław Woźniak and Maciej Piasecki and Paweł Walkowiak and Konrad Wojtasik and Arkadiusz Janz and Przemysław Kazienko and Julia Moska and Jan Kocoń},
+      year={2025},
+      eprint={2511.17161},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL},
+      url={https://arxiv.org/abs/2511.17161}, 
+}
+```

artifacts/hf_readmes/ptaszynski__PolishCyberbullyingDataset__README.md

@@ -0,0 +1,51 @@
+---
+license: cc-by-4.0
+language:
+- pl
+tags:
+- cyberbullying
+- hate-speech
+pretty_name: PolishCyberbullyingDataset
+---
+
+# Expert-annotated dataset to study cyberbullying in Polish language
+
+This the first publically available expert-annotated dataset containing annotations of cyberbullying and hate-speech in Polish language.
+
+Please, read [the paper](https://www.mdpi.com/2306-5729/9/1/1) about the dataset for all necessary details.
+
+## Model
+The classification model which achieved the highest classification results for the dataset is also released under the following URL.
+[Polbert-CB - Polish BERT trained for Automatic Cyberbullying Detection](https://huggingface.co/ptaszynski/bert-base-polish-cyberbullying)
+
+## Citations
+Whenever you use the dataset, please, cite it using the following citation to [the paper](https://www.mdpi.com/2306-5729/9/1/1). 
+```
+@article{ptaszynski2023expert,
+  title={Expert-Annotated Dataset to Study Cyberbullying in Polish Language},
+  author={Ptaszynski, Michal and Pieciukiewicz, Agata and Dybala, Pawel and Skrzek, Pawel and Soliwoda, Kamil and Fortuna, Marcin and Leliwa, Gniewosz and Wroczynski, Michal},
+  journal={Data},
+  volume={9},
+  number={1},
+  pages={1},
+  year={2023},
+  publisher={MDPI}
+}
+```
+
+## Licences
+The dataset is licensed under [CC BY 4.0](http://creativecommons.org/licenses/by/4.0/), or Creative Commons Attribution 4.0 International License.
+
+<a rel="license" href="http://creativecommons.org/licenses/by/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by/4.0/88x31.png" /></a>
+
+
+## Bundle
+
+The whole bundle containing (1) the old version of the dataset, (2) current version of the dataset, as well as (3) the model trained on this dataset can be found on [Zenodo](https://zenodo.org/records/7188178). 
+
+## Author
+Michal Ptaszynski - contact me on:
+- Twitter: [@mich_ptaszynski](https://twitter.com/mich_ptaszynski)
+- GitHub: [ptaszynski](https://github.com/ptaszynski)
+- LinkedIn: [michalptaszynski](https://jp.linkedin.com/in/michalptaszynski)
+- HuggingFace: [ptaszynski](https://huggingface.co/ptaszynski)

artifacts/source_candidate_audit_v0_3.json

@@ -0,0 +1,952 @@
+[
+  {
+    "repo_id": "PleIAs/common_corpus",
+    "configured_decision": "accept_after_subset_review",
+    "target_bucket": "bulk_pretraining",
+    "priority": "high",
+    "notes": "Large open/traceable corpus with Polish coverage. Filter to Polish, non-legal, non-code, high-quality domains where metadata license is acceptable.",
+    "gated": false,
+    "private": false,
+    "downloads": 83583,
+    "license_tags": [],
+    "language_tags": [
+      "en",
+      "fr",
+      "de",
+      "zh",
+      "it",
+      "es",
+      "ja",
+      "pl",
+      "la",
+      "nl",
+      "ru",
+      "ar",
+      "ko"
+    ],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [],
+    "readme_license_lines": [
+      "Common Corpus is the largest open licensed text dataset, comprising 2.27 trillion tokens (2,267,302,720,836 tokens). It is a diverse dataset, consisting of books, newspapers, scientific articles, government and legal documents, code, and more. Common Corpus has been created by Pleias in association with several partners.",
+      "* **Truly Open**: contains only data that is either uncopyrighted or freely licensed",
+      "* **Traceable**: each individual document is associated with documented contextual information, including licensed use or lack of copyright.",
+      "Common Corpus makes it possible to train model compatible with [the Open Source Initiative’s definition](https://opensource.org/ai/open-source-ai-definition#:~:text=An%20Open%20Source%20AI%20is,including%20to%20change%20its%20output.) of open-source AI, which includes openness of use, meaning use is permitted for “any purpose and without having to ask for permission\". Based on the available licensing information Common Corpus can be filtered to only include public domain works or a subset of free licenses (like attribution only).",
+      "* **OpenCulture**: our largest collection at 967,018,390,906 tokens, featuring public domain books, newspapers from cultural heritage repositories and open projets like Wikisource ad Gutenberg. We're developing innovative tools of OCR correction based on Pleias Models to correct historical digitization errors, while implementing advanced toxicity filtering to ensure content meets modern ethical standards.",
+      "| OpenCulture | cultural heritage | public domain books and newspapers, Wikisource |",
+      "The first version of [Common Corpus](https://huggingface.co/datasets/PleIAs/common_corpus) was released in November of 2024. The second version added Wikidata and detailed document-level information, including licensing and other core metadata whenever available. The third ongoing version dramatically expand the language coverage of Common Corpus beyond the US and Europe with the integration of large collection of documents in Chinese, Japanese, Arabic, Korean and Hindi.",
+      "* `license`: sharing rights for the content either uncopyrighted (public domain, US federal public domain, CC0 on Wikidata) or various free licenses (Creative Commons, MIT, French Licence ouverte, etc.)",
+      "* `date`: date of creation of the resource where known. Due to the significance of public domain and other cultural heritage content, more than half of Common Corpus predates the 21st century.",
+      "* `word_count`: number of space delimited words.",
+      "All data in Common Corpus are either uncopyrighted or freely licensed and may be used for both commercial and non-commercial purposes.",
+      "Some small parts of the French administrative common crawl have been entirely dropped using our unreleased small reasoning model for GDPR-filtering, due to the heightened risk of transmitting identifiable indirect personal information."
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      ".gitkeep",
+      "README.md",
+      "common_corpus_1/.gitkeep",
+      "common_corpus_1/subset_100_1.parquet",
+      "common_corpus_1/subset_100_10.parquet",
+      "common_corpus_1/subset_100_2.parquet",
+      "common_corpus_1/subset_100_3.parquet",
+      "common_corpus_1/subset_100_4.parquet",
+      "common_corpus_1/subset_100_5.parquet",
+      "common_corpus_1/subset_100_6.parquet",
+      "common_corpus_1/subset_100_7.parquet",
+      "common_corpus_1/subset_100_8.parquet",
+      "common_corpus_1/subset_100_9.parquet",
+      "common_corpus_1/subset_10_1.parquet",
+      "common_corpus_1/subset_10_10.parquet",
+      "common_corpus_1/subset_10_2.parquet",
+      "common_corpus_1/subset_10_3.parquet",
+      "common_corpus_1/subset_10_4.parquet",
+      "common_corpus_1/subset_10_5.parquet",
+      "common_corpus_1/subset_10_6.parquet",
+      "common_corpus_1/subset_10_7.parquet",
+      "common_corpus_1/subset_10_8.parquet",
+      "common_corpus_1/subset_10_9.parquet",
+      "common_corpus_1/subset_11_1.parquet",
+      "common_corpus_1/subset_11_10.parquet",
+      "common_corpus_1/subset_11_2.parquet",
+      "common_corpus_1/subset_11_3.parquet",
+      "common_corpus_1/subset_11_4.parquet",
+      "common_corpus_1/subset_11_5.parquet"
+    ]
+  },
+  {
+    "repo_id": "PleIAs/Polish-PD",
+    "configured_decision": "accept_after_ocr_review",
+    "target_bucket": "bulk_pretraining",
+    "priority": "high",
+    "notes": "Large public-domain Polish books/newspapers. Useful for scale, but OCR garbage must be measured aggressively.",
+    "gated": false,
+    "private": false,
+    "downloads": 518,
+    "license_tags": [],
+    "language_tags": [],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [],
+    "readme_license_lines": [
+      "# 🇵🇱 Polish Public Domain 🇵🇱",
+      "**Polish-Public Domain** or **Polish-PD** is a large collection aiming to aggregate all Polish monographies and periodicals in the public domain. As of March 2024, it is the biggest Polish open corpus.",
+      "The composition of the dataset adheres to the criteria for public domain works in the EU and, consequently, all Berne-countries for EU authors: any publication whose author is dead for more than 70 years. Additionally, the initial consolidation of public domain status for cultural heritage operates in the EU under the 2019 Copyright Directive (art. 14).",
+      "As of March 2024, to limit rights verification, we have retained exclusively titles published prior to 1884.",
+      "The corpus will be expanded at a later stage to encompass late 19th century and early 20th century publications, after checking for public domain validity.",
+      "* **Legal**: With the adoption of the AI Act with its obligations in terms of copyright law compliance for the pretraining corpora, the European AI ecosystem will have to change its provenance practices.",
+      "## License",
+      "The entire collection is in the public domain in all regions. This means that the patrimonial rights of each individual or collective right holders have expired.",
+      "There has been a debate for years in Europe over the definition of public domain and the possibility to restrict its use. Since 2019, the EU Copyright Directive states that \"Member States shall provide that, when the term of protection of a work of visual art has expired, any material resulting from an act of reproduction of that work is not subject to copyright or related rights, unless the material resulting from that act of reproduction is original in the sense that it is the author's own intellectual creation.\" (art. 14)"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "pol_pd_1.parquet",
+      "pol_pd_10.parquet",
+      "pol_pd_11.parquet",
+      "pol_pd_12.parquet",
+      "pol_pd_13.parquet",
+      "pol_pd_14.parquet",
+      "pol_pd_15.parquet",
+      "pol_pd_16.parquet",
+      "pol_pd_17.parquet",
+      "pol_pd_18.parquet",
+      "pol_pd_19.parquet",
+      "pol_pd_2.parquet",
+      "pol_pd_20.parquet",
+      "pol_pd_21.parquet",
+      "pol_pd_22.parquet",
+      "pol_pd_23.parquet",
+      "pol_pd_24.parquet",
+      "pol_pd_25.parquet",
+      "pol_pd_26.parquet",
+      "pol_pd_27.parquet",
+      "pol_pd_28.parquet",
+      "pol_pd_29.parquet",
+      "pol_pd_3.parquet",
+      "pol_pd_30.parquet",
+      "pol_pd_31.parquet",
+      "pol_pd_32.parquet",
+      "pol_pd_33.parquet",
+      "pol_pd_34.parquet"
+    ]
+  },
+  {
+    "repo_id": "pelcra/PLLuMIC",
+    "configured_decision": "accept_gated",
+    "target_bucket": "final_phase_sft",
+    "priority": "high",
+    "notes": "High-quality hand-crafted Polish instruction/dialogue data. Use for SFT/final phase, not bulk pretraining.",
+    "gated": "auto",
+    "private": false,
+    "downloads": 33,
+    "license_tags": [
+      "cc-by-sa-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "n<1K"
+    ],
+    "task_tags": [
+      "text-generation"
+    ],
+    "readme_license_lines": [
+      "license: cc-by-sa-4.0",
+      "- **License:** CC-BY-SA-4.0"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "pllumic.json"
+    ]
+  },
+  {
+    "repo_id": "ipipan/polqa",
+    "configured_decision": "accept_with_dedup",
+    "target_bucket": "qa_eval_final_phase",
+    "priority": "medium",
+    "notes": "CC-BY-SA Polish QA; passages are Wikipedia-heavy, so dedup against existing Wikipedia.",
+    "gated": false,
+    "private": false,
+    "downloads": 214,
+    "license_tags": [
+      "cc-by-sa-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [
+      "question-answering",
+      "text-retrieval"
+    ],
+    "readme_license_lines": [
+      "license: cc-by-sa-4.0",
+      "The passages proposed by the `hard-negative` and `zero-shot` methods are bound to be easier to retrieve by retrievers since they were proposed by such. To mitigate this bias, we include the passages found by the human annotators in an unconstrained way (`passage_source=\"human\"`). We hypothesize that it will result in more unbiased and diverse examples. Moreover, we asked the annotators to find not one but up to five passages, preferably from different articles to even further increase passage diversity.",
+      "### Other Known Limitations",
+      "The PolQA dataset focuses on trivia questions which might limit its usefulness in real-world applications since neural retrievers generalize poorly to other domains.",
+      "### Licensing Information"
+    ],
+    "warnings": [
+      "unknown"
+    ],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "data/passages.jsonl",
+      "data/test.csv",
+      "data/train.csv",
+      "data/valid.csv",
+      "polqa.py"
+    ]
+  },
+  {
+    "repo_id": "clarin-pl/poquad",
+    "configured_decision": "accept_with_dedup",
+    "target_bucket": "qa_eval_final_phase",
+    "priority": "medium",
+    "notes": "CC-BY Polish QA. Keep as QA/eval/final-phase source.",
+    "gated": false,
+    "private": false,
+    "downloads": 217,
+    "license_tags": [
+      "cc-by-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [
+      "question-answering"
+    ],
+    "readme_license_lines": [
+      "license:",
+      "- cc-by-4.0"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "poquad-dev.json",
+      "poquad-train.json",
+      "poquad.py"
+    ]
+  },
+  {
+    "repo_id": "clarin-pl/PUGG",
+    "configured_decision": "accept_with_dedup",
+    "target_bucket": "qa_eval_final_phase",
+    "priority": "medium",
+    "notes": "CC-BY-SA Polish QA/retrieval data.",
+    "gated": false,
+    "private": false,
+    "downloads": 44,
+    "license_tags": [
+      "cc-by-sa-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [
+      "question-answering",
+      "text-retrieval"
+    ],
+    "readme_license_lines": [
+      "license:",
+      "- cc-by-sa-4.0"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "ir/corpus.jsonl",
+      "ir/qrels/test.jsonl",
+      "ir/queries.jsonl",
+      "kbqa/natural/test.jsonl",
+      "kbqa/natural/train.jsonl",
+      "kbqa/template-based/test.jsonl",
+      "kbqa/template-based/train.jsonl",
+      "mrc/test.jsonl",
+      "mrc/train.jsonl"
+    ]
+  },
+  {
+    "repo_id": "clarin-pl/ComplexQA",
+    "configured_decision": "accept_with_dedup",
+    "target_bucket": "qa_eval",
+    "priority": "medium",
+    "notes": "Small CC-BY-SA complex QA set; useful for evaluation and final phase.",
+    "gated": false,
+    "private": false,
+    "downloads": 21,
+    "license_tags": [
+      "cc-by-sa-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "n<1K"
+    ],
+    "task_tags": [
+      "question-answering"
+    ],
+    "readme_license_lines": [
+      "license: cc-by-sa-4.0"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "test.jsonl"
+    ]
+  },
+  {
+    "repo_id": "openlanguagedata/flores_plus",
+    "configured_decision": "eval_only",
+    "target_bucket": "eval",
+    "priority": "low",
+    "notes": "Clean CC-BY-SA multilingual dev/devtest data. Too small for pretraining; useful for eval.",
+    "gated": "auto",
+    "private": false,
+    "downloads": 17481,
+    "license_tags": [
+      "cc-by-sa-4.0"
+    ],
+    "language_tags": [
+      "ace",
+      "acm",
+      "acq",
+      "aeb",
+      "af",
+      "ajp",
+      "ak",
+      "als",
+      "am",
+      "apc",
+      "ar",
+      "ars",
+      "ary",
+      "arz",
+      "as",
+      "ast",
+      "awa",
+      "ayr",
+      "azb",
+      "azj",
+      "ba",
+      "bm",
+      "ban",
+      "be",
+      "bem",
+      "bn",
+      "bho",
+      "bjn",
+      "bo",
+      "bs",
+      "bug",
+      "bg",
+      "ca",
+      "ceb",
+      "cs",
+      "cjk",
+      "ckb",
+      "crh",
+      "cy",
+      "da",
+      "de",
+      "dar",
+      "dik",
+      "dyu",
+      "dz",
+      "el",
+      "en",
+      "eo",
+      "et",
+      "eu",
+      "ee",
+      "fo",
+      "fj",
+      "fi",
+      "fon",
+      "fr",
+      "fur",
+      "fuv",
+      "gaz",
+      "gd",
+      "ga",
+      "gl",
+      "gn",
+      "gu",
+      "ht",
+      "ha",
+      "he",
+      "hi",
+      "hne",
+      "hr",
+      "hu",
+      "hy",
+      "ig",
+      "ilo",
+      "id",
+      "is",
+      "it",
+      "jv",
+      "ja",
+      "kab",
+      "kac",
+      "kam",
+      "kn",
+      "ks",
+      "ka",
+      "kk",
+      "kbp",
+      "kea",
+      "khk",
+      "km",
+      "ki",
+      "rw",
+      "kjh",
+      "ky",
+      "kmb",
+      "kmr",
+      "knc",
+      "kg",
+      "ko",
+      "lo",
+      "lij",
+      "li",
+      "lld",
+      "ln",
+      "lt",
+      "lmo",
+      "ltg",
+      "lb",
+      "lua",
+      "lg",
+      "luo",
+      "lus",
+      "lvs",
+      "mag",
+      "mai",
+      "ml",
+      "mar",
+      "mfe",
+      "mhr",
+      "min",
+      "mk",
+      "mt",
+      "mni",
+      "mos",
+      "mi",
+      "my",
+      "nl",
+      "nn",
+      "nb",
+      "npi",
+      "nso",
+      "nus",
+      "ny",
+      "oc",
+      "ory",
+      "pag",
+      "pa",
+      "pap",
+      "pbt",
+      "pes",
+      "plt",
+      "pl",
+      "pt",
+      "prs",
+      "quy",
+      "ro",
+      "rn",
+      "ru",
+      "sg",
+      "sa",
+      "sat",
+      "scn",
+      "shn",
+      "si",
+      "sk",
+      "sl",
+      "sm",
+      "sn",
+      "sd",
+      "so",
+      "st",
+      "es",
+      "sc",
+      "sr",
+      "ss",
+      "su",
+      "sv",
+      "swh",
+      "szl",
+      "ta",
+      "taq",
+      "tt",
+      "te",
+      "tg",
+      "tl",
+      "th",
+      "ti",
+      "tpi",
+      "tn",
+      "ts",
+      "tk",
+      "tum",
+      "tr",
+      "tw",
+      "tzm",
+      "udm",
+      "ug",
+      "uk",
+      "umb",
+      "ur",
+      "uzn",
+      "uzs",
+      "vec",
+      "vi",
+      "war",
+      "wo",
+      "xh",
+      "ydd",
+      "yo",
+      "yue",
+      "zgh",
+      "zh",
+      "zsm",
+      "zu"
+    ],
+    "size_tags": [
+      "100K<n<1M"
+    ],
+    "task_tags": [
+      "text-generation",
+      "translation"
+    ],
+    "readme_license_lines": [
+      "license:",
+      "- cc-by-sa-4.0",
+      "FLORES+ is a multilingual machine translation benchmark released under [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/). This dataset was originally released by FAIR researchers at Meta under the name FLORES. Further information about these initial releases can be found in [Dataset Sources](#dataset-sources) below. The data is now being managed by OLDI, [the Open Language Data Initiative](https://oldi.org/). The + has been added to the name to disambiguate between the original datasets and this new actively developed version.",
+      "- **License:** [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)",
+      "In order to show your agreement with the DCO you should include at the end of commit message,",
+      "This can be done easily using the `-s` flag on the `git commit`.",
+      "7. Commit with an -s flag (e.g. `git commit -s -m \"fix a few typos in the Russian dev set\"`)."
+    ],
+    "warnings": [
+      "unknown"
+    ],
+    "sample_files": [
+      ".gitattributes",
+      ".gitignore",
+      "CHANGELOG.md",
+      "DCO",
+      "README.md",
+      "bibliography.bib",
+      "convert_parquet_to_jsonl.py",
+      "dataset_cards/arg_Latn.md",
+      "dataset_cards/ast_Latn.md",
+      "dataset_cards/cat_Latn_vale1252.md",
+      "dataset_cards/chv_Cyrl.md",
+      "dataset_cards/cmn_Hans.md",
+      "dataset_cards/cmn_Hant.md",
+      "dataset_cards/dar_Cyrl.md",
+      "dataset_cards/fil_Latn.md",
+      "dataset_cards/kaa_Latn.md",
+      "dataset_cards/khk_Mong.md",
+      "dataset_cards/kjh_Cyrl.md",
+      "dataset_cards/lij_Latn.md",
+      "dataset_cards/lld_Latn.md",
+      "dataset_cards/lld_Latn_gard1241.md",
+      "dataset_cards/mfe_Latn.md",
+      "dataset_cards/mhr_Cyrl.md",
+      "dataset_cards/myv_Cyrl.md",
+      "dataset_cards/nob_Latn.md",
+      "dataset_cards/nob_Latn_radical.md",
+      "dataset_cards/oci_Latn_aran1260.md",
+      "dataset_cards/tyv_Cyrl.md",
+      "dataset_cards/udm_Cyrl.md",
+      "dataset_cards/uzs_Arab.md"
+    ]
+  },
+  {
+    "repo_id": "oscar-corpus/mOSCAR",
+    "configured_decision": "maybe_after_legal_review",
+    "target_bucket": "web_fallback",
+    "priority": "medium",
+    "notes": "CC-BY tag but web-crawl copyright/provenance risk remains. Only use after filtering and legal review.",
+    "gated": false,
+    "private": false,
+    "downloads": 1741,
+    "license_tags": [
+      "cc-by-4.0"
+    ],
+    "language_tags": [],
+    "size_tags": [
+      "100M<n<1B"
+    ],
+    "task_tags": [],
+    "readme_license_lines": [
+      "license: cc-by-4.0"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "data/ace_Latn/00000.parquet",
+      "data/acm_Arab/00000.parquet",
+      "data/aeb_Arab/00000.parquet",
+      "data/afr_Latn/00000.parquet",
+      "data/ajp_Arab/00000.parquet",
+      "data/als_Latn/00000.parquet",
+      "data/als_Latn/00001.parquet",
+      "data/als_Latn/00002.parquet",
+      "data/als_Latn/00003.parquet",
+      "data/als_Latn/00004.parquet",
+      "data/als_Latn/00005.parquet",
+      "data/als_Latn/00006.parquet",
+      "data/als_Latn/00007.parquet",
+      "data/als_Latn/00008.parquet",
+      "data/amh_Ethi/00000.parquet",
+      "data/apc_Arab/00000.parquet",
+      "data/arb_Arab/00000.parquet",
+      "data/arb_Arab/00001.parquet",
+      "data/arb_Arab/00002.parquet",
+      "data/arb_Arab/00003.parquet",
+      "data/arb_Arab/00004.parquet",
+      "data/arb_Arab/00005.parquet",
+      "data/arb_Arab/00006.parquet",
+      "data/arb_Arab/00007.parquet",
+      "data/arb_Arab/00008.parquet",
+      "data/arb_Arab/00009.parquet",
+      "data/arb_Arab/00010.parquet",
+      "data/arb_Arab/00011.parquet"
+    ]
+  },
+  {
+    "repo_id": "oscar-corpus/OSCAR-2301",
+    "configured_decision": "hold",
+    "target_bucket": "web_fallback",
+    "priority": "low",
+    "notes": "HF tag says CC0 but card text mentions research-only license; do not include until resolved.",
+    "gated": "manual",
+    "private": false,
+    "downloads": 1931,
+    "license_tags": [
+      "cc0-1.0"
+    ],
+    "language_tags": [],
+    "size_tags": [
+      "n>1T"
+    ],
+    "task_tags": [
+      "fill-mask",
+      "text-generation"
+    ],
+    "readme_license_lines": [
+      "license: cc0-1.0",
+      "extra_gated_prompt: \"**[IMPORTANT: We are forced to temporarily suspend access to OSCAR. The temporary nature of this suspension has led us to choose to implement it as gated access with manual control, and we will not grant any access until the situation has been clarified. We sincerely apologize for this situation and hope to be able to restore access as soon as possible. In the meantime, we remind you that we have always prohibited access to or use of OSCAR that violates the legislation in force where you are located. In France, for example, any use of OSCAR that does not fall within the framework of the so-called 'TDM' or the so-called 'research' exceptions to copyright has always been prohibited.]** By filling the form below, you understand that only the metadata and the annotations of OSCAR 23.01 have a cc0-1.0 license, and that the rest of the content is crawled data derived from the November/December 2022 snapshot of Common Crawl, for which the authors of OSCAR **do not** hold any copyright whatsoever.\"",
+      "- [Other Known Limitations](#other-known-limitations)",
+      "- [Licensing Information](#licensing-information)",
+      "### Other Known Limitations",
+      "### Licensing Information",
+      "These data are released under this licensing scheme",
+      "We license the actual packaging, the metadata and the annotations of these data under the Creative Commons CC0 license (\"no rights reserved\") http://creativecommons.org/publicdomain/zero/1.0/",
+      "To the extent possible under law, the OSCAR project, Inria, the Univertity of Mannheim and DFKI GmbH have waived all copyright and related or neighboring rights to OSCAR",
+      "* Clearly identify the copyrighted work claimed to be infringed.",
+      "abstract = {Since the introduction of large language models in Natural Language Processing, large raw corpora have played a crucial role in Computational Linguistics. However, most of these large raw corpora are either available only for English or not available to the general public due to copyright issues. Nevertheless, there are some examples of freely available multilingual corpora for training Deep Learning NLP models, such as the OSCAR and Paracrawl corpora. However, they have quality issues, especially for low-resource languages. Moreover, recreating or updating these corpora is very complex. In this work, we try to reproduce and improve the goclassy pipeline used to create the OSCAR corpus. We propose a new pipeline that is faster, modular, parameterizable, and well documented. We use it to create a corpus similar to OSCAR but larger and based on recent data. Also, unlike OSCAR, the metadata information is at the document level. We release our pipeline under an open source license and publish the corpus under a research-only license.},",
+      "Papadimitriou, Isabel and"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      ".gitignore",
+      "OSCAR-2301.py",
+      "README.md",
+      "af_meta/af_meta.jsonl.zst",
+      "af_meta/checksum.sha256",
+      "am_meta/am_meta.jsonl.zst",
+      "am_meta/checksum.sha256",
+      "an_meta/an_meta.jsonl.zst",
+      "an_meta/checksum.sha256",
+      "ar_meta/ar_meta_part_1.jsonl.zst",
+      "ar_meta/ar_meta_part_10.jsonl.zst",
+      "ar_meta/ar_meta_part_11.jsonl.zst",
+      "ar_meta/ar_meta_part_12.jsonl.zst",
+      "ar_meta/ar_meta_part_13.jsonl.zst",
+      "ar_meta/ar_meta_part_14.jsonl.zst",
+      "ar_meta/ar_meta_part_15.jsonl.zst",
+      "ar_meta/ar_meta_part_16.jsonl.zst",
+      "ar_meta/ar_meta_part_17.jsonl.zst",
+      "ar_meta/ar_meta_part_18.jsonl.zst",
+      "ar_meta/ar_meta_part_19.jsonl.zst",
+      "ar_meta/ar_meta_part_2.jsonl.zst",
+      "ar_meta/ar_meta_part_20.jsonl.zst",
+      "ar_meta/ar_meta_part_21.jsonl.zst",
+      "ar_meta/ar_meta_part_22.jsonl.zst",
+      "ar_meta/ar_meta_part_23.jsonl.zst",
+      "ar_meta/ar_meta_part_24.jsonl.zst",
+      "ar_meta/ar_meta_part_25.jsonl.zst",
+      "ar_meta/ar_meta_part_26.jsonl.zst",
+      "ar_meta/ar_meta_part_27.jsonl.zst"
+    ]
+  },
+  {
+    "repo_id": "WiktorS/polish-news",
+    "configured_decision": "reject_unless_permission",
+    "target_bucket": "contemporary_news",
+    "priority": "medium",
+    "notes": "Apache-2.0 tag is not enough: README says articles were obtained from tvp.info.pl, while TVP/TVP Info pages do not expose an open license for article reuse and TVP has separate licensing/copyright channels. Include only with explicit permission or a documented upstream open license.",
+    "gated": false,
+    "private": false,
+    "downloads": 117,
+    "license_tags": [
+      "apache-2.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "100K<n<1M"
+    ],
+    "task_tags": [
+      "text-classification",
+      "summarization",
+      "text-generation"
+    ],
+    "readme_license_lines": [
+      "license: apache-2.0"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "full_data.csv"
+    ]
+  },
+  {
+    "repo_id": "ptaszynski/PolishCyberbullyingDataset",
+    "configured_decision": "eval_only",
+    "target_bucket": "safety_eval",
+    "priority": "low",
+    "notes": "CC-BY Polish social/toxicity data. Useful for safety probes, not bulk pretraining.",
+    "gated": false,
+    "private": false,
+    "downloads": 168,
+    "license_tags": [
+      "cc-by-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [],
+    "readme_license_lines": [
+      "license: cc-by-4.0",
+      "## Licences",
+      "The dataset is licensed under [CC BY 4.0](http://creativecommons.org/licenses/by/4.0/), or Creative Commons Attribution 4.0 International License.",
+      "<a rel=\"license\" href=\"http://creativecommons.org/licenses/by/4.0/\"><img alt=\"Creative Commons License\" style=\"border-width:0\" src=\"https://i.creativecommons.org/l/by/4.0/88x31.png\" /></a>"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "v2_test_anonymized.tsv",
+      "v2_training_anonymized.tsv"
+    ]
+  },
+  {
+    "repo_id": "michaljunczyk/pl-asr-bigos",
+    "configured_decision": "partial_only",
+    "target_bucket": "spoken_transcripts",
+    "priority": "low",
+    "notes": "Mixed upstream licenses include proprietary and noncommercial-compatible criteria; only compatible transcript subsets may be used.",
+    "gated": "auto",
+    "private": false,
+    "downloads": 6,
+    "license_tags": [
+      "cc-by-sa-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "1K<n<10K"
+    ],
+    "task_tags": [
+      "automatic-speech-recognition"
+    ],
+    "readme_license_lines": [
+      "license:",
+      "- cc-by-sa-4.0",
+      "Original datasets used for curation of BIGOS have specific terms of usage that must be understood and agreed to before use. Below are the links to the license terms and datasets the specific license type applies to:",
+      "* [Creative Commons 0](https://creativecommons.org/share-your-work/public-domain/cc0) which applies to [Common Voice](https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0)",
+      "* [Creative Commons By Attribution Share Alike 4.0](https://creativecommons.org/licenses/by-sa/4.0/), which applies to [Clarin Cyfry](https://clarin-pl.eu/dspace/handle/11321/317), [Azon acoustic speech resources corpus](https://zasobynauki.pl/zasoby/korpus-nagran-probek-mowy-do-celow-budowy-modeli-akustycznych-dla-automatycznego-rozpoznawania-mowy,53293/).",
+      "* [Creative Commons By Attribution 3.0](https://creativecommons.org/licenses/by/3.0/), which applies to [CLARIN Mobile database](https://clarin-pl.eu/dspace/handle/11321/237), [CLARIN Studio database](https://clarin-pl.eu/dspace/handle/11321/236), [PELCRA Spelling and Numbers Voice Database](http://pelcra.pl/new/snuv) and [FLEURS dataset](https://huggingface.co/datasets/google/fleurs)",
+      "* [Creative Commons By Attribution 4.0](https://creativecommons.org/licenses/by/4.0/), which applies to [Multilingual Librispeech](https://huggingface.co/datasets/facebook/multilingual_librispeech) and [Poly AI Minds 14](https://huggingface.co/datasets/PolyAI/minds14)",
+      "* [Proprietiary License of Munich AI Labs dataset](https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset)",
+      "* Public domain mark, which applies to [PWR datasets](https://www.ii.pwr.edu.pl/~sas/ASR/)",
+      "I hereby confirm that I have read and accepted the license terms of datasets comprising BIGOS corpora: checkbox",
+      "- [Other Known Limitations](#other-known-limitations)",
+      "- [Licensing Information](#licensing-information)"
+    ],
+    "warnings": [
+      "noncommercial"
+    ],
+    "sample_files": [
+      ".gitattributes",
+      ".gitignore",
+      ".python-version",
+      "README.md",
+      "data/clarin-pjatk-mobile-15/test.tar.gz",
+      "data/clarin-pjatk-mobile-15/test.tsv",
+      "data/clarin-pjatk-studio-15/test.tar.gz",
+      "data/clarin-pjatk-studio-15/test.tsv",
+      "data/fair-mls-20/test.tar.gz",
+      "data/fair-mls-20/test.tsv",
+      "data/mailabs-19/test.tar.gz",
+      "data/mailabs-19/test.tsv",
+      "data/mozilla-common-voice-19/test.tar.gz",
+      "data/mozilla-common-voice-19/test.tsv",
+      "data/pwr-azon-read-20/test.tar.gz",
+      "data/pwr-azon-read-20/test.tsv",
+      "data/pwr-azon-spont-20/test.tar.gz",
+      "data/pwr-azon-spont-20/test.tsv",
+      "data/pwr-maleset-unk/test.tar.gz",
+      "data/pwr-maleset-unk/test.tsv",
+      "data/pwr-shortwords-unk/test.tar.gz",
+      "data/pwr-shortwords-unk/test.tsv",
+      "data/pwr-viu-unk/test.tar.gz",
+      "data/pwr-viu-unk/test.tsv",
+      "pl-asr-bigos.py",
+      "test.py"
+    ]
+  },
+  {
+    "repo_id": "clarin-knext/wsd_polish_datasets",
+    "configured_decision": "partial_only",
+    "target_bucket": "nlp_eval_final_phase",
+    "priority": "low",
+    "notes": "Mixed licenses including GPL and plWordNet license; inspect each subcorpus before inclusion.",
+    "gated": false,
+    "private": false,
+    "downloads": 48,
+    "license_tags": [
+      "cc-by-4.0"
+    ],
+    "language_tags": [
+      "pl"
+    ],
+    "size_tags": [
+      "1M<n<10M"
+    ],
+    "task_tags": [
+      "token-classification"
+    ],
+    "readme_license_lines": [
+      "license:",
+      "- cc-by-4.0",
+      "- [Other Known Limitations](#other-known-limitations)",
+      "- [Licensing Information](#licensing-information)",
+      "### Other Known Limitations",
+      "### Licensing Information",
+      "KPWR-100 [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)",
+      "KPWR [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)",
+      "Walenty [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)",
+      "Sherlock [CC-BY 4.0](https://creativecommons.org/licenses/by/4.0/)",
+      "Skladnica [GNU GPL 3](http://www.gnu.org/licenses/gpl-3.0.en.html)",
+      "GLEX [plWordNet License](http://plwordnet.pwr.wroc.pl/wordnet/licence)"
+    ],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "README.md",
+      "data/emoglex_sentences.jsonl",
+      "data/emoglex_text.jsonl",
+      "data/kpwr-100_sentences.jsonl",
+      "data/kpwr-100_text.jsonl",
+      "data/kpwr_sentences.jsonl",
+      "data/kpwr_text.jsonl",
+      "data/sherlock_sentences.jsonl",
+      "data/sherlock_text.jsonl",
+      "data/skladnica_sentences.jsonl",
+      "data/skladnica_text.jsonl",
+      "data/walenty_sentences.jsonl",
+      "data/walenty_text.jsonl",
+      "data/wikiglex_sentences.jsonl",
+      "data/wikiglex_text.jsonl",
+      "wsd_polish_datasets.py"
+    ]
+  },
+  {
+    "repo_id": "allegro/summarization-polish-summaries-corpus",
+    "configured_decision": "reject_until_license",
+    "target_bucket": "summarization",
+    "priority": "low",
+    "notes": "No HF license tag; hold until provenance and redistribution rights are documented.",
+    "gated": false,
+    "private": false,
+    "downloads": 279,
+    "license_tags": [],
+    "language_tags": [],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [],
+    "readme_license_lines": [],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "abstract/dev.csv",
+      "abstract/test.csv",
+      "abstract/train.csv",
+      "extract/dev.csv",
+      "extract/test.csv",
+      "extract/train.csv",
+      "whole/dev.csv",
+      "whole/test.csv",
+      "whole/train.csv"
+    ]
+  },
+  {
+    "repo_id": "allegro/polish-question-passage-pairs",
+    "configured_decision": "reject_until_license",
+    "target_bucket": "qa_eval_final_phase",
+    "priority": "low",
+    "notes": "No HF license tag; hold until provenance and redistribution rights are documented.",
+    "gated": false,
+    "private": false,
+    "downloads": 175,
+    "license_tags": [],
+    "language_tags": [],
+    "size_tags": [
+      "10K<n<100K"
+    ],
+    "task_tags": [],
+    "readme_license_lines": [],
+    "warnings": [],
+    "sample_files": [
+      ".gitattributes",
+      "all.csv"
+    ]
+  }
+]

artifacts/source_candidate_audit_v0_3.md

@@ -0,0 +1,296 @@
+# Source candidate audit
+
+| repo | decision | bucket | license tags | gated | warnings |
+|---|---|---|---|---:|---|
+| `PleIAs/common_corpus` | `accept_after_subset_review` | `bulk_pretraining` | `none` | `False` | `-` |
+| `PleIAs/Polish-PD` | `accept_after_ocr_review` | `bulk_pretraining` | `none` | `False` | `-` |
+| `pelcra/PLLuMIC` | `accept_gated` | `final_phase_sft` | `cc-by-sa-4.0` | `auto` | `-` |
+| `ipipan/polqa` | `accept_with_dedup` | `qa_eval_final_phase` | `cc-by-sa-4.0` | `False` | `unknown` |
+| `clarin-pl/poquad` | `accept_with_dedup` | `qa_eval_final_phase` | `cc-by-4.0` | `False` | `-` |
+| `clarin-pl/PUGG` | `accept_with_dedup` | `qa_eval_final_phase` | `cc-by-sa-4.0` | `False` | `-` |
+| `clarin-pl/ComplexQA` | `accept_with_dedup` | `qa_eval` | `cc-by-sa-4.0` | `False` | `-` |
+| `openlanguagedata/flores_plus` | `eval_only` | `eval` | `cc-by-sa-4.0` | `auto` | `unknown` |
+| `oscar-corpus/mOSCAR` | `maybe_after_legal_review` | `web_fallback` | `cc-by-4.0` | `False` | `-` |
+| `oscar-corpus/OSCAR-2301` | `hold` | `web_fallback` | `cc0-1.0` | `manual` | `-` |
+| `WiktorS/polish-news` | `reject_unless_permission` | `contemporary_news` | `apache-2.0` | `False` | `-` |
+| `ptaszynski/PolishCyberbullyingDataset` | `eval_only` | `safety_eval` | `cc-by-4.0` | `False` | `-` |
+| `michaljunczyk/pl-asr-bigos` | `partial_only` | `spoken_transcripts` | `cc-by-sa-4.0` | `auto` | `noncommercial` |
+| `clarin-knext/wsd_polish_datasets` | `partial_only` | `nlp_eval_final_phase` | `cc-by-4.0` | `False` | `-` |
+| `allegro/summarization-polish-summaries-corpus` | `reject_until_license` | `summarization` | `none` | `False` | `-` |
+| `allegro/polish-question-passage-pairs` | `reject_until_license` | `qa_eval_final_phase` | `none` | `False` | `-` |
+
+## PleIAs/common_corpus
+
+- decision: `accept_after_subset_review`
+- bucket: `bulk_pretraining`
+- priority: `high`
+- license tags: `none`
+- language tags: `en, fr, de, zh, it, es, ja, pl, la, nl, ru, ar, ko`
+- gated: `False`
+- notes: Large open/traceable corpus with Polish coverage. Filter to Polish, non-legal, non-code, high-quality domains where metadata license is acceptable.
+
+README/license signals:
+- Common Corpus is the largest open licensed text dataset, comprising 2.27 trillion tokens (2,267,302,720,836 tokens). It is a diverse dataset, consisting of books, newspapers, scientific articles, government and legal documents, code, and more. Common Corpus has been created by Pleias in association with several partners.
+- * **Truly Open**: contains only data that is either uncopyrighted or freely licensed
+- * **Traceable**: each individual document is associated with documented contextual information, including licensed use or lack of copyright.
+- Common Corpus makes it possible to train model compatible with [the Open Source Initiative’s definition](https://opensource.org/ai/open-source-ai-definition#:~:text=An%20Open%20Source%20AI%20is,including%20to%20change%20its%20output.) of open-source AI, which includes openness of use, meaning use is permitted for “any purpose and without having to ask for permission". Based on the available licensing information Common Corpus can be filtered to only include public domain works or a subset of free licenses (like attribution only).
+- * **OpenCulture**: our largest collection at 967,018,390,906 tokens, featuring public domain books, newspapers from cultural heritage repositories and open projets like Wikisource ad Gutenberg. We're developing innovative tools of OCR correction based on Pleias Models to correct historical digitization errors, while implementing advanced toxicity filtering to ensure content meets modern ethical standards.
+- | OpenCulture | cultural heritage | public domain books and newspapers, Wikisource |
+- The first version of [Common Corpus](https://huggingface.co/datasets/PleIAs/common_corpus) was released in November of 2024. The second version added Wikidata and detailed document-level information, including licensing and other core metadata whenever available. The third ongoing version dramatically expand the language coverage of Common Corpus beyond the US and Europe with the integration of large collection of documents in Chinese, Japanese, Arabic, Korean and Hindi.
+- * `license`: sharing rights for the content either uncopyrighted (public domain, US federal public domain, CC0 on Wikidata) or various free licenses (Creative Commons, MIT, French Licence ouverte, etc.)
+- * `date`: date of creation of the resource where known. Due to the significance of public domain and other cultural heritage content, more than half of Common Corpus predates the 21st century.
+- * `word_count`: number of space delimited words.
+- All data in Common Corpus are either uncopyrighted or freely licensed and may be used for both commercial and non-commercial purposes.
+- Some small parts of the French administrative common crawl have been entirely dropped using our unreleased small reasoning model for GDPR-filtering, due to the heightened risk of transmitting identifiable indirect personal information.
+
+## PleIAs/Polish-PD
+
+- decision: `accept_after_ocr_review`
+- bucket: `bulk_pretraining`
+- priority: `high`
+- license tags: `none`
+- language tags: `none`
+- gated: `False`
+- notes: Large public-domain Polish books/newspapers. Useful for scale, but OCR garbage must be measured aggressively.
+
+README/license signals:
+- # 🇵🇱 Polish Public Domain 🇵🇱
+- **Polish-Public Domain** or **Polish-PD** is a large collection aiming to aggregate all Polish monographies and periodicals in the public domain. As of March 2024, it is the biggest Polish open corpus.
+- The composition of the dataset adheres to the criteria for public domain works in the EU and, consequently, all Berne-countries for EU authors: any publication whose author is dead for more than 70 years. Additionally, the initial consolidation of public domain status for cultural heritage operates in the EU under the 2019 Copyright Directive (art. 14).
+- As of March 2024, to limit rights verification, we have retained exclusively titles published prior to 1884.
+- The corpus will be expanded at a later stage to encompass late 19th century and early 20th century publications, after checking for public domain validity.
+- * **Legal**: With the adoption of the AI Act with its obligations in terms of copyright law compliance for the pretraining corpora, the European AI ecosystem will have to change its provenance practices.
+- ## License
+- The entire collection is in the public domain in all regions. This means that the patrimonial rights of each individual or collective right holders have expired.
+- There has been a debate for years in Europe over the definition of public domain and the possibility to restrict its use. Since 2019, the EU Copyright Directive states that "Member States shall provide that, when the term of protection of a work of visual art has expired, any material resulting from an act of reproduction of that work is not subject to copyright or related rights, unless the material resulting from that act of reproduction is original in the sense that it is the author's own intellectual creation." (art. 14)
+
+## pelcra/PLLuMIC
+
+- decision: `accept_gated`
+- bucket: `final_phase_sft`
+- priority: `high`
+- license tags: `cc-by-sa-4.0`
+- language tags: `pl`
+- gated: `auto`
+- notes: High-quality hand-crafted Polish instruction/dialogue data. Use for SFT/final phase, not bulk pretraining.
+
+README/license signals:
+- license: cc-by-sa-4.0
+- - **License:** CC-BY-SA-4.0
+
+## ipipan/polqa
+
+- decision: `accept_with_dedup`
+- bucket: `qa_eval_final_phase`
+- priority: `medium`
+- license tags: `cc-by-sa-4.0`
+- language tags: `pl`
+- gated: `False`
+- notes: CC-BY-SA Polish QA; passages are Wikipedia-heavy, so dedup against existing Wikipedia.
+
+README/license signals:
+- license: cc-by-sa-4.0
+- The passages proposed by the `hard-negative` and `zero-shot` methods are bound to be easier to retrieve by retrievers since they were proposed by such. To mitigate this bias, we include the passages found by the human annotators in an unconstrained way (`passage_source="human"`). We hypothesize that it will result in more unbiased and diverse examples. Moreover, we asked the annotators to find not one but up to five passages, preferably from different articles to even further increase passage diversity.
+- ### Other Known Limitations
+- The PolQA dataset focuses on trivia questions which might limit its usefulness in real-world applications since neural retrievers generalize poorly to other domains.
+- ### Licensing Information
+
+## clarin-pl/poquad
+
+- decision: `accept_with_dedup`
+- bucket: `qa_eval_final_phase`
+- priority: `medium`
+- license tags: `cc-by-4.0`
+- language tags: `pl`
+- gated: `False`
+- notes: CC-BY Polish QA. Keep as QA/eval/final-phase source.
+
+README/license signals:
+- license:
+- - cc-by-4.0
+
+## clarin-pl/PUGG
+
+- decision: `accept_with_dedup`
+- bucket: `qa_eval_final_phase`
+- priority: `medium`
+- license tags: `cc-by-sa-4.0`
+- language tags: `pl`
+- gated: `False`
+- notes: CC-BY-SA Polish QA/retrieval data.
+
+README/license signals:
+- license:
+- - cc-by-sa-4.0
+
+## clarin-pl/ComplexQA
+
+- decision: `accept_with_dedup`
+- bucket: `qa_eval`
+- priority: `medium`
+- license tags: `cc-by-sa-4.0`
+- language tags: `pl`
+- gated: `False`
+- notes: Small CC-BY-SA complex QA set; useful for evaluation and final phase.
+
+README/license signals:
+- license: cc-by-sa-4.0
+
+## openlanguagedata/flores_plus
+
+- decision: `eval_only`
+- bucket: `eval`
+- priority: `low`
+- license tags: `cc-by-sa-4.0`
+- language tags: `ace, acm, acq, aeb, af, ajp, ak, als, am, apc, ar, ars, ary, arz, as, ast, awa, ayr, azb, azj, ba, bm, ban, be, bem, bn, bho, bjn, bo, bs, bug, bg, ca, ceb, cs, cjk, ckb, crh, cy, da, de, dar, dik, dyu, dz, el, en, eo, et, eu, ee, fo, fj, fi, fon, fr, fur, fuv, gaz, gd, ga, gl, gn, gu, ht, ha, he, hi, hne, hr, hu, hy, ig, ilo, id, is, it, jv, ja, kab, kac, kam, kn, ks, ka, kk, kbp, kea, khk, km, ki, rw, kjh, ky, kmb, kmr, knc, kg, ko, lo, lij, li, lld, ln, lt, lmo, ltg, lb, lua, lg, luo, lus, lvs, mag, mai, ml, mar, mfe, mhr, min, mk, mt, mni, mos, mi, my, nl, nn, nb, npi, nso, nus, ny, oc, ory, pag, pa, pap, pbt, pes, plt, pl, pt, prs, quy, ro, rn, ru, sg, sa, sat, scn, shn, si, sk, sl, sm, sn, sd, so, st, es, sc, sr, ss, su, sv, swh, szl, ta, taq, tt, te, tg, tl, th, ti, tpi, tn, ts, tk, tum, tr, tw, tzm, udm, ug, uk, umb, ur, uzn, uzs, vec, vi, war, wo, xh, ydd, yo, yue, zgh, zh, zsm, zu`
+- gated: `auto`
+- notes: Clean CC-BY-SA multilingual dev/devtest data. Too small for pretraining; useful for eval.
+
+README/license signals:
+- license:
+- - cc-by-sa-4.0
+- FLORES+ is a multilingual machine translation benchmark released under [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/). This dataset was originally released by FAIR researchers at Meta under the name FLORES. Further information about these initial releases can be found in [Dataset Sources](#dataset-sources) below. The data is now being managed by OLDI, [the Open Language Data Initiative](https://oldi.org/). The + has been added to the name to disambiguate between the original datasets and this new actively developed version.
+- - **License:** [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
+- In order to show your agreement with the DCO you should include at the end of commit message,
+- This can be done easily using the `-s` flag on the `git commit`.
+- 7. Commit with an -s flag (e.g. `git commit -s -m "fix a few typos in the Russian dev set"`).
+
+## oscar-corpus/mOSCAR
+
+- decision: `maybe_after_legal_review`
+- bucket: `web_fallback`
+- priority: `medium`
+- license tags: `cc-by-4.0`
+- language tags: `none`
+- gated: `False`
+- notes: CC-BY tag but web-crawl copyright/provenance risk remains. Only use after filtering and legal review.
+
+README/license signals:
+- license: cc-by-4.0
+
+## oscar-corpus/OSCAR-2301
+
+- decision: `hold`
+- bucket: `web_fallback`
+- priority: `low`
+- license tags: `cc0-1.0`
+- language tags: `none`
+- gated: `manual`
+- notes: HF tag says CC0 but card text mentions research-only license; do not include until resolved.
+
+README/license signals:
+- license: cc0-1.0
+- extra_gated_prompt: "**[IMPORTANT: We are forced to temporarily suspend access to OSCAR. The temporary nature of this suspension has led us to choose to implement it as gated access with manual control, and we will not grant any access until the situation has been clarified. We sincerely apologize for this situation and hope to be able to restore access as soon as possible. In the meantime, we remind you that we have always prohibited access to or use of OSCAR that violates the legislation in force where you are located. In France, for example, any use of OSCAR that does not fall within the framework of the so-called 'TDM' or the so-called 'research' exceptions to copyright has always been prohibited.]** By filling the form below, you understand that only the metadata and the annotations of OSCAR 23.01 have a cc0-1.0 license, and that the rest of the content is crawled data derived from the November/December 2022 snapshot of Common Crawl, for which the authors of OSCAR **do not** hold any copyright whatsoever."
+- - [Other Known Limitations](#other-known-limitations)
+- - [Licensing Information](#licensing-information)
+- ### Other Known Limitations
+- ### Licensing Information
+- These data are released under this licensing scheme
+- We license the actual packaging, the metadata and the annotations of these data under the Creative Commons CC0 license ("no rights reserved") http://creativecommons.org/publicdomain/zero/1.0/
+- To the extent possible under law, the OSCAR project, Inria, the Univertity of Mannheim and DFKI GmbH have waived all copyright and related or neighboring rights to OSCAR
+- * Clearly identify the copyrighted work claimed to be infringed.
+- abstract = {Since the introduction of large language models in Natural Language Processing, large raw corpora have played a crucial role in Computational Linguistics. However, most of these large raw corpora are either available only for English or not available to the general public due to copyright issues. Nevertheless, there are some examples of freely available multilingual corpora for training Deep Learning NLP models, such as the OSCAR and Paracrawl corpora. However, they have quality issues, especially for low-resource languages. Moreover, recreating or updating these corpora is very complex. In this work, we try to reproduce and improve the goclassy pipeline used to create the OSCAR corpus. We propose a new pipeline that is faster, modular, parameterizable, and well documented. We use it to create a corpus similar to OSCAR but larger and based on recent data. Also, unlike OSCAR, the metadata information is at the document level. We release our pipeline under an open source license and publish the corpus under a research-only license.},
+- Papadimitriou, Isabel and
+
+## WiktorS/polish-news
+
+- decision: `reject_unless_permission`
+- bucket: `contemporary_news`
+- priority: `medium`
+- license tags: `apache-2.0`
+- language tags: `pl`
+- gated: `False`
+- notes: Apache-2.0 tag is not enough: README says articles were obtained from tvp.info.pl, while TVP/TVP Info pages do not expose an open license for article reuse and TVP has separate licensing/copyright channels. Include only with explicit permission or a documented upstream open license.
+
+README/license signals:
+- license: apache-2.0
+
+## ptaszynski/PolishCyberbullyingDataset
+
+- decision: `eval_only`
+- bucket: `safety_eval`
+- priority: `low`
+- license tags: `cc-by-4.0`
+- language tags: `pl`
+- gated: `False`
+- notes: CC-BY Polish social/toxicity data. Useful for safety probes, not bulk pretraining.
+
+README/license signals:
+- license: cc-by-4.0
+- ## Licences
+- The dataset is licensed under [CC BY 4.0](http://creativecommons.org/licenses/by/4.0/), or Creative Commons Attribution 4.0 International License.
+- <a rel="license" href="http://creativecommons.org/licenses/by/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by/4.0/88x31.png" /></a>
+
+## michaljunczyk/pl-asr-bigos
+
+- decision: `partial_only`
+- bucket: `spoken_transcripts`
+- priority: `low`
+- license tags: `cc-by-sa-4.0`
+- language tags: `pl`
+- gated: `auto`
+- notes: Mixed upstream licenses include proprietary and noncommercial-compatible criteria; only compatible transcript subsets may be used.
+
+README/license signals:
+- license:
+- - cc-by-sa-4.0
+- Original datasets used for curation of BIGOS have specific terms of usage that must be understood and agreed to before use. Below are the links to the license terms and datasets the specific license type applies to:
+- * [Creative Commons 0](https://creativecommons.org/share-your-work/public-domain/cc0) which applies to [Common Voice](https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0)
+- * [Creative Commons By Attribution Share Alike 4.0](https://creativecommons.org/licenses/by-sa/4.0/), which applies to [Clarin Cyfry](https://clarin-pl.eu/dspace/handle/11321/317), [Azon acoustic speech resources corpus](https://zasobynauki.pl/zasoby/korpus-nagran-probek-mowy-do-celow-budowy-modeli-akustycznych-dla-automatycznego-rozpoznawania-mowy,53293/).
+- * [Creative Commons By Attribution 3.0](https://creativecommons.org/licenses/by/3.0/), which applies to [CLARIN Mobile database](https://clarin-pl.eu/dspace/handle/11321/237), [CLARIN Studio database](https://clarin-pl.eu/dspace/handle/11321/236), [PELCRA Spelling and Numbers Voice Database](http://pelcra.pl/new/snuv) and [FLEURS dataset](https://huggingface.co/datasets/google/fleurs)
+- * [Creative Commons By Attribution 4.0](https://creativecommons.org/licenses/by/4.0/), which applies to [Multilingual Librispeech](https://huggingface.co/datasets/facebook/multilingual_librispeech) and [Poly AI Minds 14](https://huggingface.co/datasets/PolyAI/minds14)
+- * [Proprietiary License of Munich AI Labs dataset](https://www.caito.de/2019/01/03/the-m-ailabs-speech-dataset)
+- * Public domain mark, which applies to [PWR datasets](https://www.ii.pwr.edu.pl/~sas/ASR/)
+- I hereby confirm that I have read and accepted the license terms of datasets comprising BIGOS corpora: checkbox
+- - [Other Known Limitations](#other-known-limitations)
+- - [Licensing Information](#licensing-information)
+
+## clarin-knext/wsd_polish_datasets
+
+- decision: `partial_only`
+- bucket: `nlp_eval_final_phase`
+- priority: `low`
+- license tags: `cc-by-4.0`
+- language tags: `pl`
+- gated: `False`
+- notes: Mixed licenses including GPL and plWordNet license; inspect each subcorpus before inclusion.
+
+README/license signals:
+- license:
+- - cc-by-4.0
+- - [Other Known Limitations](#other-known-limitations)
+- - [Licensing Information](#licensing-information)
+- ### Other Known Limitations
+- ### Licensing Information
+- KPWR-100 [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
+- KPWR [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
+- Walenty [CC-BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
+- Sherlock [CC-BY 4.0](https://creativecommons.org/licenses/by/4.0/)
+- Skladnica [GNU GPL 3](http://www.gnu.org/licenses/gpl-3.0.en.html)
+- GLEX [plWordNet License](http://plwordnet.pwr.wroc.pl/wordnet/licence)
+
+## allegro/summarization-polish-summaries-corpus
+
+- decision: `reject_until_license`
+- bucket: `summarization`
+- priority: `low`
+- license tags: `none`
+- language tags: `none`
+- gated: `False`
+- notes: No HF license tag; hold until provenance and redistribution rights are documented.
+
+README/license signals:
+- no obvious README license lines found
+
+## allegro/polish-question-passage-pairs
+
+- decision: `reject_until_license`
+- bucket: `qa_eval_final_phase`
+- priority: `low`
+- license tags: `none`
+- language tags: `none`
+- gated: `False`
+- notes: No HF license tag; hold until provenance and redistribution rights are documented.
+
+README/license signals:
+- no obvious README license lines found

artifacts/source_license_review_v0_3.md

@@ -0,0 +1,56 @@
+# Source license review for v0.3
+
+Review date: 2026-06-16
+
+Scope: candidate sources for a larger, more diverse, more natural Polish DynaWord v0.3. This is a practical engineering review, not legal advice.
+
+## Decision table
+
+| source | bucket | decision | license evidence | main risk / condition |
+|---|---|---|---|---|
+| `PleIAs/common_corpus` | bulk pretraining | `accept_after_subset_review` | Card says Common Corpus contains only uncopyrighted or freely licensed data, document-level contextual/license metadata, and commercial/non-commercial use. | Filter Polish records by `license`, source/domain, quality, and duplicate overlap. |
+| `PleIAs/Polish-PD` | bulk pretraining | `accept_after_ocr_review` | Card says entire collection is public domain in all regions and describes EU public-domain criteria. | Strong OCR-noise risk; sample first, then aggressive OCR filtering. |
+| `pelcra/PLLuMIC` | final phase / SFT | `accept_gated` | HF card: `CC-BY-SA-4.0`, Polish, hand-crafted instructions, professional annotators and review. | Gated; small and instruction-formatted, so not bulk pretraining. |
+| `ipipan/polqa` | QA/eval/final phase | `accept_with_dedup` | HF card: `CC-BY-SA-4.0`; Polish QA with 7,000 questions, evidence passages, and Wikipedia candidate passages. | Wikipedia-heavy; dedup against existing Wikipedia. |
+| `clarin-pl/poquad` | QA/eval/final phase | `accept_with_dedup` | HF card tag: `CC-BY-4.0`, Polish QA. | Small; schema review needed. |
+| `clarin-pl/PUGG` | QA/eval/final phase | `accept_with_dedup` | HF card tag: `CC-BY-SA-4.0`, Polish QA/retrieval. | Dedup and separate eval/final phase bucket. |
+| `clarin-pl/ComplexQA` | QA/eval | `accept_with_dedup` | HF card tag: `CC-BY-SA-4.0`, Polish QA. | Small; eval/final phase only. |
+| `openlanguagedata/flores_plus` | eval | `eval_only` | HF card tag: `CC-BY-SA-4.0`, multilingual including Polish. | Eval/dev sentences only, not pretraining scale. |
+| `oscar-corpus/mOSCAR` | web fallback | `maybe_after_legal_review` | HF card tag: `CC-BY-4.0`. | Web-crawl copyright/provenance risk; require allowlist or source-level filtering plus boilerplate/near-dedup. |
+| `oscar-corpus/OSCAR-2301` | web fallback | `hold` | HF tag says `CC0-1.0`, but card text references research-only license. | Do not use until license conflict is resolved. |
+| `WiktorS/polish-news` | contemporary news | `reject_unless_permission` | HF tag says `apache-2.0`, README says articles obtained from `tvp.info.pl`. | Upstream TVP Info articles are not shown as Apache/CC/open; TVP has separate licensing/copyright channels. |
+| `ptaszynski/PolishCyberbullyingDataset` | safety eval | `eval_only` | HF tag: `CC-BY-4.0`, Polish. | Toxicity/classification skew; not bulk pretraining. |
+| `michaljunczyk/pl-asr-bigos` | spoken transcripts | `partial_only` | HF tag: `CC-BY-SA-4.0`; card lists mixed upstream licenses. | Includes proprietary/noncommercial-compatible source conditions; use only compatible transcript subsets. |
+| `clarin-knext/wsd_polish_datasets` | NLP eval/final phase | `partial_only` | HF tag: `CC-BY-4.0`; README lists mixed subcorpus licenses including CC-BY-SA, CC-BY, GPL, plWordNet. | Subcorpus-level license review required. |
+| `allegro/summarization-polish-summaries-corpus` | summarization | `reject_until_license` | No HF license tag found. | Need upstream rights and redistribution license. |
+| `allegro/polish-question-passage-pairs` | QA/eval | `reject_until_license` | No HF license tag found. | Need upstream rights and redistribution license. |
+
+## TVP Info / `WiktorS/polish-news`
+
+Decision: **blocked for inclusion** unless explicit permission or authoritative upstream open-license evidence is obtained.
+
+Evidence checked:
+
+- `WiktorS/polish-news` README says the dataset contains more than 250k articles obtained from `tvp.info.pl`.
+- HF metadata tags the uploaded dataset as `apache-2.0`, but that does not establish that TVP Info article text was licensed by TVP under Apache 2.0.
+- TVP pages expose `Telewizja Polska SA` copyright metadata and a separate `licencje.tvp.pl` licensing channel.
+- TVP Stream terms reserve rights to materials and prohibit redistribution outside the granted viewing license. This is not the same service as TVP Info articles, but it is evidence of TVP's general licensing posture for materials.
+
+Operational result: keep `WiktorS/polish-news` out of the clean open corpus. Revisit only if we can cite a TVP page granting open reuse of `tvp.info.pl` article text or get permission from TVP.
+
+## Import order
+
+1. `PleIAs/common_corpus`: inspect Polish subset metadata and select natural, non-legal, non-code records with acceptable license values.
+2. `PleIAs/Polish-PD`: sample OCR quality and filter aggressively.
+3. `pelcra/PLLuMIC`: accept gated terms and import as `dialogue_sft` / final-phase source.
+4. `ipipan/polqa`, `clarin-pl/poquad`, `clarin-pl/PUGG`, `clarin-pl/ComplexQA`: import as QA/eval/final-phase buckets after dedup.
+5. `oscar-corpus/mOSCAR`: only if the above is not enough for modern web diversity and legal review is acceptable.
+
+## Required gates before publish
+
+- License evidence per source and, where needed, per subcorpus.
+- Exact, normalized, and near-dedup against existing v0.2 data.
+- Boilerplate removal for web/news-like data.
+- OCR-noise score for public-domain scans.
+- Per-source phrase/style contamination report.
+- Per-source perplexity probe with small GPT-2 style model.

artifacts/source_scouting_v0_3.md

@@ -0,0 +1,62 @@
+# Source scouting for Polish DynaWord v0.3
+
+Goal: improve training quality by adding contemporary, diverse Polish while keeping traceable licensing.
+
+## Strong candidates
+
+| candidate | type | license/status | why useful | action |
+|---|---|---|---|---|
+| `PleIAs/common_corpus` | multilingual open corpus | open/traceable per card | large Polish component, diverse domains, provenance-oriented | inspect Polish subset metadata, dedup against existing data |
+| `PleIAs/Polish-PD` | Polish public-domain books/newspapers | public-domain collection, license must be verified per card/source | large, traceable cultural/news-ish historical text; already separate Polish subset | sample OCR quality before inclusion |
+| `pelcra/PLLuMIC` | Polish instruction/dialogue | `CC-BY-SA-4.0`, gated | high-quality hand-crafted Polish instructions/dialogues | use for SFT/final quality phase, not bulk pretraining |
+| `ipipan/polqa` | Polish QA + passages | `CC-BY-SA-4.0` | real QA style, manually labeled evidence passages, large candidate passage corpus | use for QA/eval/high-quality mix; avoid duplicate Wikipedia-heavy passages |
+| `clarin-pl/poquad` | Polish QA | `CC-BY-4.0` | native Polish QA, useful for eval and final phase | add as QA-style source after dedup |
+| `clarin-pl/PUGG` | Polish QA/retrieval | `CC-BY-SA-4.0` | natural questions + retrieval/MRC variants | use for eval/final phase |
+| `clarin-pl/ComplexQA` | Polish complex QA | `CC-BY-SA-4.0` | small but useful complex-question evaluation | eval/final phase |
+| `openlanguagedata/flores_plus` | translation/dev sentences | `CC-BY-SA-4.0`, gated | clean multilingual Polish sentence set | eval only or very small high-quality mix |
+
+## Possible candidates, needs review
+
+| candidate | type | license/status | concern | action |
+|---|---|---|---|---|
+| `oscar-corpus/OSCAR-2301` | Common Crawl web | HF tag says `CC0-1.0`, gated/manual; card text mentions research-only license | legal/copyright ambiguity despite useful web scale | use only after license review and heavy filtering |
+| `oscar-corpus/mOSCAR` | multilingual web | `CC-BY-4.0` | web crawl copyright and boilerplate risk | possible fallback with allowlist/blocklist + near-dedup |
+| `WiktorS/polish-news` | Polish news from `tvp.info.pl` | HF tag says `apache-2.0` | upstream TVP Info article rights are not shown as Apache/CC/open; TVP has copyright/licensing pages | reject unless explicit TVP permission or upstream open-license evidence is obtained |
+| `ptaszynski/PolishCyberbullyingDataset` | Polish social/media examples | `CC-BY-4.0` | classification dataset, toxicity/safety skew | eval/safety only, not bulk pretraining |
+| `michaljunczyk/pl-asr-bigos` | Polish ASR transcripts | `CC-BY-SA-4.0`, gated | mixed upstream licenses; speech transcript style | only transcript subset with compatible upstream licenses |
+| `clarin-knext/wsd_polish_datasets` | Polish WSD text/sentences | `CC-BY-4.0` | annotated NLP data, may be sentence fragments | small final/eval source after schema review |
+| `allegro/summarization-polish-summaries-corpus` | Polish summaries | missing HF license | source/license unclear | hold until license provenance is documented |
+| `allegro/polish-question-passage-pairs` | Polish Q/passages | missing HF license | source/license unclear | hold until license provenance is documented |
+
+## Reject or avoid for clean pretraining
+
+| candidate | reason |
+|---|---|
+| `pelcra/PLLuMIC-syn-ext` | synthetic extension; useful for SFT experiments, not clean human-text pretraining |
+| random `*-forum-polish` HF datasets | mostly no license, tiny, likely scraped student datasets |
+| `PLLuMIC` as bulk text | high quality but too small and instruction-formatted; do not treat as web text |
+| extra Wikipedia mirrors | already covered; only use if needed for validation, not diversity |
+| subtitles/movie dialogue | copyright derivative risk |
+| `CC-BY-NC-*` / `NC` datasets | not compatible with the current open commercial-friendly curation stance |
+
+## Next import order
+
+1. Inspect and sample `PleIAs/common_corpus` Polish records.
+2. Inspect and sample `PleIAs/Polish-PD`; measure OCR garbage and duplicate overlap.
+3. Add `pelcra/PLLuMIC` as `dialogue_sft` / high-quality final-phase source after accepting HF access terms.
+4. Add QA/eval buckets from `ipipan/polqa`, `clarin-pl/poquad`, `clarin-pl/PUGG`, `clarin-pl/ComplexQA`.
+5. Review `WiktorS/polish-news` provenance before considering contemporary news.
+6. Treat OSCAR/mOSCAR as fallback web source only after license review, boilerplate removal, language filtering, URL/domain filtering, and near-dedup.
+
+## TVP Info review note
+
+`WiktorS/polish-news` should not be imported into Polish DynaWord v0.3 based only on the HF `apache-2.0` tag. The dataset README says it contains articles obtained from `tvp.info.pl`, but the upstream TVP/TVP Info pages checked during review did not expose an Apache, Creative Commons, CC0, public-domain, or equivalent open license for article reuse. TVP also maintains a separate `licencje.tvp.pl` licensing channel, and TVP Stream terms reserve rights to materials and prohibit redistribution of materials outside the granted viewing license. Treat the dataset as blocked unless TVP grants explicit permission or an authoritative upstream open-license page is found.
+
+## Filters required before inclusion
+
+- Exact and normalized-text dedup.
+- Near-dedup with MinHash/LSH or equivalent.
+- Boilerplate removal for web/news/forum sources.
+- Per-source perplexity with a small GPT-2 probe.
+- Style contamination rates for legal/parliamentary markers.
+- Per-source token caps and sqrt/temperature sampling.

artifacts/training_mix_v0_3.json

@@ -0,0 +1,103 @@
+{
+  "sources": [
+    {
+      "source": "dziennik_ustaw",
+      "path": "data/dziennik_ustaw/dziennik_ustaw.parquet",
+      "tokens": 486126575,
+      "raw_corpus_share": 0.07813788759879016,
+      "target_share": 0.02394963053714845,
+      "target_tokens": 23949631,
+      "sampling_probability": 0.04926624511321974
+    },
+    {
+      "source": "eltec_pol",
+      "path": "data/eltec_pol/eltec_pol.parquet",
+      "tokens": 21486720,
+      "raw_corpus_share": 0.003453682638573455,
+      "target_share": 0.03369541042261511,
+      "target_tokens": 33695410,
+      "sampling_probability": 1.0
+    },
+    {
+      "source": "eurlex",
+      "path": "data/eurlex/eurlex.parquet",
+      "tokens": 2378055718,
+      "raw_corpus_share": 0.38223841269476827,
+      "target_share": 0.0756723305320818,
+      "target_tokens": 75672331,
+      "sampling_probability": 0.031821092511508595
+    },
+    {
+      "source": "parliamentary",
+      "path": "data/parliamentary/parliamentary.parquet",
+      "tokens": 1646835986,
+      "raw_corpus_share": 0.2647053088338377,
+      "target_share": 0.05037803893076972,
+      "target_tokens": 50378039,
+      "sampling_probability": 0.030590805294680997
+    },
+    {
+      "source": "wikibooks",
+      "path": "data/wikibooks/wikibooks.parquet",
+      "tokens": 15571295,
+      "raw_corpus_share": 0.0025028627543713347,
+      "target_share": 0.04971988821765658,
+      "target_tokens": 49719888,
+      "sampling_probability": 1.0
+    },
+    {
+      "source": "wikinews",
+      "path": "data/wikinews/wikinews.parquet",
+      "tokens": 12141355,
+      "raw_corpus_share": 0.001951549002000166,
+      "target_share": 0.04221511889874753,
+      "target_tokens": 42215119,
+      "sampling_probability": 1.0
+    },
+    {
+      "source": "wikipedia",
+      "path": "data/wikipedia/wikipedia.parquet",
+      "tokens": 707194207,
+      "raw_corpus_share": 0.1136713446638492,
+      "target_share": 0.2964091545224541,
+      "target_tokens": 296409155,
+      "sampling_probability": 0.4191340257966791
+    },
+    {
+      "source": "wikiquote",
+      "path": "data/wikiquote/wikiquote.parquet",
+      "tokens": 31896591,
+      "raw_corpus_share": 0.005126920375300572,
+      "target_share": 0.03831727358127462,
+      "target_tokens": 38317274,
+      "sampling_probability": 1.0
+    },
+    {
+      "source": "wikisource",
+      "path": "data/wikisource/wikisource.parquet",
+      "tokens": 801947427,
+      "raw_corpus_share": 0.1289015683025866,
+      "target_share": 0.23330083859999876,
+      "target_tokens": 233300839,
+      "sampling_probability": 0.2909178721013591
+    },
+    {
+      "source": "wikivoyage",
+      "path": "data/wikivoyage/wikivoyage.parquet",
+      "tokens": 17128200,
+      "raw_corpus_share": 0.0027531129446473845,
+      "target_share": 0.04813506180175537,
+      "target_tokens": 48135062,
+      "sampling_probability": 1.0
+    },
+    {
+      "source": "wolne_lektury",
+      "path": "data/wolne_lektury/wolne_lektury.parquet",
+      "tokens": 103009797,
+      "raw_corpus_share": 0.016557350191275168,
+      "target_share": 0.10820725395549792,
+      "target_tokens": 108207254,
+      "sampling_probability": 1.0
+    }
+  ]
+}

artifacts/training_mix_v0_3.md

@@ -0,0 +1,39 @@
+# Polish DynaWord training mix
+
+- config: `polish-dynaword-v0.3-quality-mix`
+- temperature alpha: `0.5`
+- legal/parliamentary raw share: `72.51%`
+- legal/parliamentary target share: `15.00%`
+- token budget: `1,000,000,000`
+
+| source | raw tokens | raw share | target share | sampling multiplier | target tokens |
+|---|---:|---:|---:|---:|---:|
+| `dziennik_ustaw` | 486,126,575 | 7.81% | 2.39% | 0.307 | 23,949,631 |
+| `eltec_pol` | 21,486,720 | 0.35% | 3.37% | 9.756 | 33,695,410 |
+| `eurlex` | 2,378,055,718 | 38.22% | 7.57% | 0.198 | 75,672,331 |
+| `parliamentary` | 1,646,835,986 | 26.47% | 5.04% | 0.190 | 50,378,039 |
+| `wikibooks` | 15,571,295 | 0.25% | 4.97% | 19.865 | 49,719,888 |
+| `wikinews` | 12,141,355 | 0.20% | 4.22% | 21.632 | 42,215,119 |
+| `wikipedia` | 707,194,207 | 11.37% | 29.64% | 2.608 | 296,409,155 |
+| `wikiquote` | 31,896,591 | 0.51% | 3.83% | 7.474 | 38,317,274 |
+| `wikisource` | 801,947,427 | 12.89% | 23.33% | 1.810 | 233,300,839 |
+| `wikivoyage` | 17,128,200 | 0.28% | 4.81% | 17.484 | 48,135,062 |
+| `wolne_lektury` | 103,009,797 | 1.66% | 10.82% | 6.535 | 108,207,254 |
+
+## Final training phase
+
+Reserve the last `10%` of training tokens for higher-quality sources:
+
+- `wikipedia`
+- `wikinews`
+- `wikibooks`
+- `wikivoyage`
+- `wolne_lektury`
+
+## Missing source classes for v0.3+
+
+- licensed contemporary Polish web
+- how-to guides and poradniki
+- technical blogs and documentation
+- Q&A
+- forums with compatible licensing

configs/source_candidates_v0_3.json

@@ -0,0 +1,143 @@
+{
+  "policy": {
+    "accepted_licenses": [
+      "cc0-1.0",
+      "cc-by-3.0",
+      "cc-by-4.0",
+      "cc-by-sa-3.0",
+      "cc-by-sa-4.0",
+      "public-domain",
+      "mit",
+      "apache-2.0",
+      "bsd-3-clause"
+    ],
+    "rejected_license_terms": [
+      "cc-by-nc",
+      "noncommercial",
+      "research-only",
+      "unknown",
+      "proprietary"
+    ],
+    "bulk_pretraining_requires": [
+      "clear license or public-domain basis",
+      "traceable upstream source",
+      "human-authored or public-domain OCR text",
+      "dedup and boilerplate removal",
+      "per-source quality report"
+    ]
+  },
+  "candidates": [
+    {
+      "repo_id": "PleIAs/common_corpus",
+      "target_bucket": "bulk_pretraining",
+      "priority": "high",
+      "decision": "accept_after_subset_review",
+      "notes": "Large open/traceable corpus with Polish coverage. Filter to Polish, non-legal, non-code, high-quality domains where metadata license is acceptable."
+    },
+    {
+      "repo_id": "PleIAs/Polish-PD",
+      "target_bucket": "bulk_pretraining",
+      "priority": "high",
+      "decision": "accept_after_ocr_review",
+      "notes": "Large public-domain Polish books/newspapers. Useful for scale, but OCR garbage must be measured aggressively."
+    },
+    {
+      "repo_id": "pelcra/PLLuMIC",
+      "target_bucket": "final_phase_sft",
+      "priority": "high",
+      "decision": "accept_gated",
+      "notes": "High-quality hand-crafted Polish instruction/dialogue data. Use for SFT/final phase, not bulk pretraining."
+    },
+    {
+      "repo_id": "ipipan/polqa",
+      "target_bucket": "qa_eval_final_phase",
+      "priority": "medium",
+      "decision": "accept_with_dedup",
+      "notes": "CC-BY-SA Polish QA; passages are Wikipedia-heavy, so dedup against existing Wikipedia."
+    },
+    {
+      "repo_id": "clarin-pl/poquad",
+      "target_bucket": "qa_eval_final_phase",
+      "priority": "medium",
+      "decision": "accept_with_dedup",
+      "notes": "CC-BY Polish QA. Keep as QA/eval/final-phase source."
+    },
+    {
+      "repo_id": "clarin-pl/PUGG",
+      "target_bucket": "qa_eval_final_phase",
+      "priority": "medium",
+      "decision": "accept_with_dedup",
+      "notes": "CC-BY-SA Polish QA/retrieval data."
+    },
+    {
+      "repo_id": "clarin-pl/ComplexQA",
+      "target_bucket": "qa_eval",
+      "priority": "medium",
+      "decision": "accept_with_dedup",
+      "notes": "Small CC-BY-SA complex QA set; useful for evaluation and final phase."
+    },
+    {
+      "repo_id": "openlanguagedata/flores_plus",
+      "target_bucket": "eval",
+      "priority": "low",
+      "decision": "eval_only",
+      "notes": "Clean CC-BY-SA multilingual dev/devtest data. Too small for pretraining; useful for eval."
+    },
+    {
+      "repo_id": "oscar-corpus/mOSCAR",
+      "target_bucket": "web_fallback",
+      "priority": "medium",
+      "decision": "maybe_after_legal_review",
+      "notes": "CC-BY tag but web-crawl copyright/provenance risk remains. Only use after filtering and legal review."
+    },
+    {
+      "repo_id": "oscar-corpus/OSCAR-2301",
+      "target_bucket": "web_fallback",
+      "priority": "low",
+      "decision": "hold",
+      "notes": "HF tag says CC0 but card text mentions research-only license; do not include until resolved."
+    },
+    {
+      "repo_id": "WiktorS/polish-news",
+      "target_bucket": "contemporary_news",
+      "priority": "medium",
+      "decision": "reject_unless_permission",
+      "notes": "Apache-2.0 tag is not enough: README says articles were obtained from tvp.info.pl, while TVP/TVP Info pages do not expose an open license for article reuse and TVP has separate licensing/copyright channels. Include only with explicit permission or a documented upstream open license."
+    },
+    {
+      "repo_id": "ptaszynski/PolishCyberbullyingDataset",
+      "target_bucket": "safety_eval",
+      "priority": "low",
+      "decision": "eval_only",
+      "notes": "CC-BY Polish social/toxicity data. Useful for safety probes, not bulk pretraining."
+    },
+    {
+      "repo_id": "michaljunczyk/pl-asr-bigos",
+      "target_bucket": "spoken_transcripts",
+      "priority": "low",
+      "decision": "partial_only",
+      "notes": "Mixed upstream licenses include proprietary and noncommercial-compatible criteria; only compatible transcript subsets may be used."
+    },
+    {
+      "repo_id": "clarin-knext/wsd_polish_datasets",
+      "target_bucket": "nlp_eval_final_phase",
+      "priority": "low",
+      "decision": "partial_only",
+      "notes": "Mixed licenses including GPL and plWordNet license; inspect each subcorpus before inclusion."
+    },
+    {
+      "repo_id": "allegro/summarization-polish-summaries-corpus",
+      "target_bucket": "summarization",
+      "priority": "low",
+      "decision": "reject_until_license",
+      "notes": "No HF license tag; hold until provenance and redistribution rights are documented."
+    },
+    {
+      "repo_id": "allegro/polish-question-passage-pairs",
+      "target_bucket": "qa_eval_final_phase",
+      "priority": "low",
+      "decision": "reject_until_license",
+      "notes": "No HF license tag; hold until provenance and redistribution rights are documented."
+    }
+  ]
+}

configs/training_mix_v0_3.json

@@ -0,0 +1,45 @@
+{
+  "name": "polish-dynaword-v0.3-quality-mix",
+  "description": "Training-oriented source mix for reducing legal/parliamentary style dominance while preserving traceable open licensing.",
+  "temperature_alpha": 0.5,
+  "legal_cap_share": 0.15,
+  "legal_sources": [
+    "eurlex",
+    "parliamentary",
+    "dziennik_ustaw"
+  ],
+  "source_multipliers": {
+    "wikipedia": 1.15,
+    "wikinews": 1.25,
+    "wikibooks": 1.30,
+    "wikivoyage": 1.20,
+    "wolne_lektury": 1.10,
+    "wikisource": 0.85,
+    "wikiquote": 0.70,
+    "eltec_pol": 0.75,
+    "eurlex": 1.00,
+    "parliamentary": 0.80,
+    "dziennik_ustaw": 0.70
+  },
+  "final_phase_share": 0.10,
+  "final_phase_sources": [
+    "wikipedia",
+    "wikinews",
+    "wikibooks",
+    "wikivoyage",
+    "wolne_lektury"
+  ],
+  "target_missing_sources": [
+    "licensed contemporary Polish web",
+    "how-to guides and poradniki",
+    "technical blogs and documentation",
+    "Q&A",
+    "forums with compatible licensing"
+  ],
+  "quality_policy": {
+    "dedup": "Use exact, normalized-text, and near-duplicate removal before release/training.",
+    "boilerplate": "Remove navigation, cookie banners, repeated headers/footers, legal boilerplate, and template remnants aggressively.",
+    "evaluation": "Track per-source perplexity, legal/parliamentary marker rates, and style-contamination probes; global loss alone is insufficient.",
+    "probe_model": "GPT-2 124M is only a cheap dataset probe, not evidence that the final model is high quality."
+  }
+}

src/make_training_mix.py

@@ -0,0 +1,248 @@
+#!/usr/bin/env python3
+"""Build a training-oriented source mix from existing Polish DynaWord parquets.
+
+The raw corpus is provenance-first. This script creates a training view with:
+- temperature/sqrt sampling by source,
+- a hard cap for legal/parliamentary sources,
+- an optional high-quality final-phase manifest,
+- optional sampled parquet materialization.
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import math
+import random
+from pathlib import Path
+
+import pyarrow as pa
+import pyarrow.compute as pc
+import pyarrow.parquet as pq
+
+
+def load_config(path: Path) -> dict:
+    return json.loads(path.read_text(encoding="utf-8"))
+
+
+def source_inventory(data_root: Path) -> dict[str, dict]:
+    inventory = {}
+    for parquet_path in sorted((data_root / "data").glob("*/*.parquet")):
+        source = parquet_path.parent.name
+        stats_path = parquet_path.with_name(f"{source}.stats.json")
+        docs = None
+        tokens = None
+        if stats_path.exists():
+            stats = json.loads(stats_path.read_text(encoding="utf-8"))
+            docs = int(stats.get("kept", 0))
+            tokens = int(stats.get("tokens", 0))
+
+        if not docs or not tokens:
+            pf = pq.ParquetFile(parquet_path)
+            docs = pf.metadata.num_rows
+            tokens = 0
+            for rg in range(pf.num_row_groups):
+                tbl = pf.read_row_group(rg, columns=["token_count"])
+                tokens += int(pc.sum(tbl["token_count"]).as_py())
+
+        inventory[source] = {
+            "path": str(parquet_path),
+            "docs": docs,
+            "tokens": tokens,
+        }
+    return inventory
+
+
+def normalize(weights: dict[str, float], total: float = 1.0) -> dict[str, float]:
+    denom = sum(weights.values())
+    if denom <= 0:
+        return {k: 0.0 for k in weights}
+    return {k: v / denom * total for k, v in weights.items()}
+
+
+def compute_mix(inventory: dict[str, dict], config: dict) -> dict[str, dict]:
+    alpha = float(config.get("temperature_alpha", 0.5))
+    multipliers = config.get("source_multipliers", {})
+    legal_sources = set(config.get("legal_sources", []))
+    legal_cap = float(config.get("legal_cap_share", 0.15))
+
+    base_weights = {}
+    for source, meta in inventory.items():
+        multiplier = float(multipliers.get(source, 1.0))
+        base_weights[source] = math.pow(meta["tokens"], alpha) * multiplier
+
+    legal_weights = {s: w for s, w in base_weights.items() if s in legal_sources}
+    other_weights = {s: w for s, w in base_weights.items() if s not in legal_sources}
+    raw_share = normalize(base_weights)
+    raw_legal_share = sum(raw_share.get(s, 0.0) for s in legal_sources)
+
+    if raw_legal_share > legal_cap and other_weights:
+        legal_share = legal_cap
+    else:
+        legal_share = raw_legal_share
+    other_share = max(0.0, 1.0 - legal_share)
+
+    final_shares = {}
+    final_shares.update(normalize(legal_weights, legal_share))
+    final_shares.update(normalize(other_weights, other_share))
+
+    mix = {}
+    total_tokens = sum(meta["tokens"] for meta in inventory.values())
+    for source, meta in inventory.items():
+        raw_corpus_share = meta["tokens"] / total_tokens if total_tokens else 0.0
+        target_share = final_shares.get(source, 0.0)
+        mix[source] = {
+            **meta,
+            "raw_corpus_share": raw_corpus_share,
+            "target_share": target_share,
+            "sampling_multiplier": target_share / raw_corpus_share if raw_corpus_share else 0.0,
+            "is_legal_capped": source in legal_sources,
+        }
+    return dict(sorted(mix.items()))
+
+
+def add_token_targets(mix: dict[str, dict], token_budget: int | None) -> None:
+    for meta in mix.values():
+        target_tokens = int(round(meta["target_share"] * token_budget)) if token_budget else 0
+        meta["target_tokens"] = target_tokens
+        meta["sampling_probability"] = min(1.0, target_tokens / meta["tokens"]) if token_budget else 0.0
+
+
+def write_report(mix: dict[str, dict], config: dict, out_path: Path, token_budget: int | None) -> None:
+    legal_sources = set(config.get("legal_sources", []))
+    raw_legal = sum(v["raw_corpus_share"] for k, v in mix.items() if k in legal_sources)
+    target_legal = sum(v["target_share"] for k, v in mix.items() if k in legal_sources)
+    lines = [
+        "# Polish DynaWord training mix",
+        "",
+        f"- config: `{config.get('name', 'unnamed')}`",
+        f"- temperature alpha: `{config.get('temperature_alpha', 0.5)}`",
+        f"- legal/parliamentary raw share: `{raw_legal * 100:.2f}%`",
+        f"- legal/parliamentary target share: `{target_legal * 100:.2f}%`",
+    ]
+    if token_budget:
+        lines.append(f"- token budget: `{token_budget:,}`")
+    lines.extend([
+        "",
+        "| source | raw tokens | raw share | target share | sampling multiplier | target tokens |",
+        "|---|---:|---:|---:|---:|---:|",
+    ])
+    for source, meta in mix.items():
+        lines.append(
+            f"| `{source}` | {meta['tokens']:,} | {meta['raw_corpus_share'] * 100:.2f}% | "
+            f"{meta['target_share'] * 100:.2f}% | {meta['sampling_multiplier']:.3f} | "
+            f"{meta['target_tokens']:,} |"
+        )
+
+    final_phase_sources = config.get("final_phase_sources", [])
+    final_phase_share = float(config.get("final_phase_share", 0.10))
+    lines.extend([
+        "",
+        "## Final training phase",
+        "",
+        f"Reserve the last `{final_phase_share * 100:.0f}%` of training tokens for higher-quality sources:",
+        "",
+    ])
+    for source in final_phase_sources:
+        lines.append(f"- `{source}`")
+
+    missing = config.get("target_missing_sources", [])
+    if missing:
+        lines.extend(["", "## Missing source classes for v0.3+", ""])
+        for item in missing:
+            lines.append(f"- {item}")
+
+    out_path.parent.mkdir(parents=True, exist_ok=True)
+    out_path.write_text("\n".join(lines) + "\n", encoding="utf-8")
+
+
+def write_manifest(mix: dict[str, dict], out_path: Path) -> None:
+    payload = {
+        "sources": [
+            {
+                "source": source,
+                "path": meta["path"],
+                "tokens": meta["tokens"],
+                "raw_corpus_share": meta["raw_corpus_share"],
+                "target_share": meta["target_share"],
+                "target_tokens": meta["target_tokens"],
+                "sampling_probability": meta["sampling_probability"],
+            }
+            for source, meta in mix.items()
+        ]
+    }
+    out_path.parent.mkdir(parents=True, exist_ok=True)
+    out_path.write_text(json.dumps(payload, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+
+
+def sample_source(source: str, meta: dict, seed: int) -> pa.Table:
+    rng = random.Random(f"{seed}:{source}")
+    target = meta["target_tokens"]
+    if target <= 0:
+        return pa.table({})
+
+    pf = pq.ParquetFile(meta["path"])
+    batches = []
+    sampled_tokens = 0
+    probability = meta["sampling_probability"]
+    for rg in range(pf.num_row_groups):
+        tbl = pf.read_row_group(rg)
+        keep = [rng.random() < probability for _ in range(tbl.num_rows)]
+        if not any(keep):
+            continue
+        sampled = tbl.filter(pa.array(keep))
+        batches.append(sampled)
+        sampled_tokens += int(pc.sum(sampled["token_count"]).as_py())
+        if sampled_tokens >= target:
+            break
+
+    return pa.concat_tables(batches, promote_options="default") if batches else pa.table({})
+
+
+def write_sampled_parquet(mix: dict[str, dict], out_path: Path, seed: int) -> None:
+    out_path.parent.mkdir(parents=True, exist_ok=True)
+    writer = None
+    try:
+        for source, meta in mix.items():
+            tbl = sample_source(source, meta, seed)
+            if tbl.num_rows == 0:
+                continue
+            if writer is None:
+                writer = pq.ParquetWriter(out_path, tbl.schema, compression="zstd")
+            writer.write_table(tbl)
+            print(f"{source}: wrote {tbl.num_rows:,} docs")
+    finally:
+        if writer is not None:
+            writer.close()
+
+
+def parse_args() -> argparse.Namespace:
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--data-root", type=Path, default=Path("."))
+    ap.add_argument("--config", type=Path, default=Path("configs/training_mix_v0_3.json"))
+    ap.add_argument("--token-budget", type=int, default=None)
+    ap.add_argument("--out-report", type=Path, default=Path("artifacts/training_mix_v0_3.md"))
+    ap.add_argument("--out-manifest", type=Path, default=Path("artifacts/training_mix_v0_3.json"))
+    ap.add_argument("--write-parquet", type=Path, default=None)
+    ap.add_argument("--seed", type=int, default=13)
+    return ap.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    config = load_config(args.config)
+    inventory = source_inventory(args.data_root)
+    mix = compute_mix(inventory, config)
+    add_token_targets(mix, args.token_budget)
+    write_report(mix, config, args.out_report, args.token_budget)
+    write_manifest(mix, args.out_manifest)
+    if args.write_parquet:
+        if not args.token_budget:
+            raise SystemExit("--write-parquet requires --token-budget")
+        write_sampled_parquet(mix, args.write_parquet, args.seed)
+    print(f"wrote: {args.out_report}")
+    print(f"wrote: {args.out_manifest}")
+
+
+if __name__ == "__main__":
+    main()

src/review_source_candidates.py

@@ -0,0 +1,161 @@
+#!/usr/bin/env python3
+"""Review Hugging Face source candidates for Polish DynaWord.
+
+This is an audit helper, not a legal opinion. It records HF metadata, README
+signals, configured curation decisions, and obvious license warnings.
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import re
+from pathlib import Path
+
+from huggingface_hub import HfApi, hf_hub_download
+
+
+WARNING_TERMS = [
+    "cc-by-nc",
+    "noncommercial",
+    "research-only",
+    "proprietary",
+    "unknown",
+    "no license",
+]
+
+
+def read_config(path: Path) -> dict:
+    return json.loads(path.read_text(encoding="utf-8"))
+
+
+def fetch_readme(repo_id: str, cache_dir: Path) -> str:
+    cache_dir.mkdir(parents=True, exist_ok=True)
+    out = cache_dir / f"{repo_id.replace('/', '__')}__README.md"
+    if out.exists():
+        return out.read_text(encoding="utf-8", errors="replace")
+    try:
+        path = hf_hub_download(repo_id=repo_id, repo_type="dataset", filename="README.md")
+        text = Path(path).read_text(encoding="utf-8", errors="replace")
+    except Exception as exc:
+        text = f"README_DOWNLOAD_ERROR: {type(exc).__name__}: {exc}"
+    out.write_text(text, encoding="utf-8")
+    return text
+
+
+def extract_license_tags(tags: list[str]) -> list[str]:
+    return [tag.removeprefix("license:") for tag in tags if tag.startswith("license:")]
+
+
+def extract_language_tags(tags: list[str]) -> list[str]:
+    return [tag.removeprefix("language:") for tag in tags if tag.startswith("language:")]
+
+
+def readme_license_lines(readme: str) -> list[str]:
+    lines = []
+    for line in readme.splitlines():
+        if re.search(r"license|licen|public domain|copyright|cc-by|cc0|apache|mit", line, re.I):
+            clean = re.sub(r"\s+", " ", line).strip()
+            if clean:
+                lines.append(clean)
+        if len(lines) >= 12:
+            break
+    return lines
+
+
+def warnings_for(license_tags: list[str], readme: str) -> list[str]:
+    haystack = " ".join(license_tags).lower() + "\n" + readme[:20000].lower()
+    return sorted({term for term in WARNING_TERMS if term in haystack})
+
+
+def review_candidate(api: HfApi, candidate: dict, cache_dir: Path) -> dict:
+    repo_id = candidate["repo_id"]
+    info = api.dataset_info(repo_id, files_metadata=False)
+    tags = info.tags or []
+    readme = fetch_readme(repo_id, cache_dir)
+    files = [s.rfilename for s in (info.siblings or [])[:30]]
+    return {
+        "repo_id": repo_id,
+        "configured_decision": candidate.get("decision", ""),
+        "target_bucket": candidate.get("target_bucket", ""),
+        "priority": candidate.get("priority", ""),
+        "notes": candidate.get("notes", ""),
+        "gated": getattr(info, "gated", None),
+        "private": info.private,
+        "downloads": getattr(info, "downloads", None),
+        "license_tags": extract_license_tags(tags),
+        "language_tags": extract_language_tags(tags),
+        "size_tags": [tag.removeprefix("size_categories:") for tag in tags if tag.startswith("size_categories:")],
+        "task_tags": [tag.removeprefix("task_categories:") for tag in tags if tag.startswith("task_categories:")],
+        "readme_license_lines": readme_license_lines(readme),
+        "warnings": warnings_for(extract_license_tags(tags), readme),
+        "sample_files": files,
+    }
+
+
+def write_json(rows: list[dict], out_path: Path) -> None:
+    out_path.parent.mkdir(parents=True, exist_ok=True)
+    out_path.write_text(json.dumps(rows, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+
+
+def write_markdown(rows: list[dict], out_path: Path) -> None:
+    lines = [
+        "# Source candidate audit",
+        "",
+        "| repo | decision | bucket | license tags | gated | warnings |",
+        "|---|---|---|---|---:|---|",
+    ]
+    for row in rows:
+        lines.append(
+            f"| `{row['repo_id']}` | `{row['configured_decision']}` | `{row['target_bucket']}` | "
+            f"`{', '.join(row['license_tags']) or 'none'}` | `{row['gated']}` | "
+            f"`{', '.join(row['warnings']) or '-'}` |"
+        )
+
+    for row in rows:
+        lines.extend([
+            "",
+            f"## {row['repo_id']}",
+            "",
+            f"- decision: `{row['configured_decision']}`",
+            f"- bucket: `{row['target_bucket']}`",
+            f"- priority: `{row['priority']}`",
+            f"- license tags: `{', '.join(row['license_tags']) or 'none'}`",
+            f"- language tags: `{', '.join(row['language_tags']) or 'none'}`",
+            f"- gated: `{row['gated']}`",
+            f"- notes: {row['notes']}",
+            "",
+            "README/license signals:",
+        ])
+        if row["readme_license_lines"]:
+            for line in row["readme_license_lines"]:
+                lines.append(f"- {line}")
+        else:
+            lines.append("- no obvious README license lines found")
+
+    out_path.parent.mkdir(parents=True, exist_ok=True)
+    out_path.write_text("\n".join(lines) + "\n", encoding="utf-8")
+
+
+def parse_args() -> argparse.Namespace:
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--config", type=Path, default=Path("configs/source_candidates_v0_3.json"))
+    ap.add_argument("--readme-cache", type=Path, default=Path("artifacts/hf_readmes"))
+    ap.add_argument("--out-json", type=Path, default=Path("artifacts/source_candidate_audit_v0_3.json"))
+    ap.add_argument("--out-md", type=Path, default=Path("artifacts/source_candidate_audit_v0_3.md"))
+    return ap.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    config = read_config(args.config)
+    api = HfApi()
+    rows = [review_candidate(api, c, args.readme_cache) for c in config["candidates"]]
+    write_json(rows, args.out_json)
+    write_markdown(rows, args.out_md)
+    print(f"wrote: {args.out_json}")
+    print(f"wrote: {args.out_md}")
+
+
+if __name__ == "__main__":
+    main()