--- language: - en - multilingual license: other license_name: private-upon.request.only license_details: This dataset is currently private (we are working on a more permissive license). Please contact dataset creator for more information regarding access and usage rights. pretty_name: Synthetic Dataset ENG-ML extra_gated_prompt: 'Unfortunately, we receive a high volume of requests from “bot” accounts (we experienced many request made from accounts without any activity and with fake/anonymous emails) that request access purely to scrape large numbers of datasets The quick form you’ll see below helps us confirm that a real person behind each request. It allows us to keep this dataset freely available rather than locking it down, while protecting it from misuse. We gladly approve legitimate requests. Thanks for your understanding!' extra_gated_fields: Complete Name: text Huggingface account tag: text Country: country current date: date_picker Personal, Professional or Institutional email: type: text label: please provide a valid Personal or, if possible, Professional / Institutional email address (requests that provide only an anonymous email will be rejected) Applicant_Information: type: text label: 'Your Information: Please provide (again) your full name, your role or position, and the name of your affiliated institution or company if relevant. If you are an independent researcher, please state so. ' Request_motive: type: text label: 'Request motive: Please describe your project. Explain the primary research questions or goals, how you plan to use this dataset, and the type of models you intend to train. ' Planned data-retention period: type: select options: - Up to 1 month - 1 month to 6 months - 6 month to 1 year - 1 year or more - label: Other (you can provide a more specific time frame on the 'Request motive' section) value: other ? I agree that access to this dataset is granted solely for the purpose I have described in my request. I will not use the dataset for any other research, commercial, or personal project without obtaining separate, explicit permission. : type: checkbox ? I agree not to copy, redistribute, sublicense, or otherwise disclose the original dataset, or any part thereof, to any third party without prior written consent from the dataset authors. I will implement reasonable security measures to protect the data from unauthorized access." : type: checkbox ? I acknowledge that the dataset is provided 'AS IS,' without any warranty, express or implied. The authors assume no liability for any damages, losses, or harms (including those resulting from data inaccuracies or biases) arising from my use of the dataset. I also I assume all risks." : type: checkbox ? I acknowledge and agree that no warranty, express or implied, is provided regarding the suitability of this dataset for training models intended for public release or commercial applications. The responsibility for ensuring the legal and ethical viability of any downstream use rests entirely with me. I understand that I must contact the dataset authors directly to inquire about licensing viability for commercial use or **to request access to subsets with more permissive conditions**. : type: checkbox ? I declare that all information I have provided is true and accurate. I understand that any breach of this agreement, misrepresentation in my request, or violation of the dataset license will result in the immediate and permanent revocation of access and may lead to further legal action. : type: checkbox configs: - config_name: CrossRep data_files: - split: train path: CrossRep/train-* - config_name: DescA data_files: - split: train path: DescA/train-* - config_name: LongA data_files: - split: train path: LongA/train-* - config_name: QA data_files: - split: train path: QA/train-* dataset_info: - config_name: CrossRep features: - name: anchor dtype: string - name: positive dtype: string - name: semi_hard_negative dtype: string - name: rephrased_anchor dtype: string - name: rephrased_positive dtype: string - name: positive_summary dtype: string - name: semantic_keywords_for_positive dtype: string - name: positive_wrong_paraphrase dtype: string - name: entailment_from_positive dtype: string - name: not_entailment_from_positive dtype: string - name: positive_core_claims_extractive dtype: string - name: anchor_translated dtype: string - name: positive_translated dtype: string - name: negative_translated dtype: string - name: positive_wrong_translation dtype: string - name: anchor_type_and_intent dtype: string - name: positive_type_and_description dtype: string - name: domain_topic dtype: string - name: hard_negative_0 dtype: string - name: hard_negative_1 dtype: string - name: hard_negative_2 dtype: string - name: pair_type dtype: string - name: translation_language dtype: string - name: base_language dtype: string - name: generation_params_json dtype: string - name: id dtype: string splits: - name: train num_bytes: 100646897 num_examples: 9129 download_size: 59979675 dataset_size: 100646897 - config_name: DescA features: - name: anchor dtype: string - name: positive dtype: string - name: semi_hard_negative dtype: string - name: rephrased_anchor dtype: string - name: rephrased_positive dtype: string - name: positive_summary dtype: string - name: semantic_keywords_for_positive dtype: string - name: positive_wrong_paraphrase dtype: string - name: entailment_from_positive dtype: string - name: not_entailment_from_positive dtype: string - name: positive_core_claims_extractive dtype: string - name: anchor_translated dtype: string - name: positive_translated dtype: string - name: negative_translated dtype: string - name: positive_wrong_translation dtype: string - name: anchor_type_and_intent dtype: string - name: positive_type_and_description dtype: string - name: domain_topic dtype: string - name: hard_negative_0 dtype: string - name: hard_negative_1 dtype: string - name: hard_negative_2 dtype: string - name: pair_type dtype: string - name: translation_language dtype: string - name: base_language dtype: string - name: generation_params_json dtype: string - name: id dtype: string splits: - name: train num_bytes: 143651906 num_examples: 8002 download_size: 87243580 dataset_size: 143651906 - config_name: LongA features: - name: anchor dtype: string - name: positive dtype: string - name: semi_hard_negative dtype: string - name: rephrased_anchor dtype: string - name: rephrased_positive dtype: string - name: positive_summary dtype: string - name: semantic_keywords_for_positive dtype: string - name: positive_wrong_paraphrase dtype: string - name: entailment_from_positive dtype: string - name: not_entailment_from_positive dtype: string - name: positive_core_claims_extractive dtype: string - name: anchor_translated dtype: string - name: positive_translated dtype: string - name: negative_translated dtype: string - name: positive_wrong_translation dtype: string - name: anchor_type_and_intent dtype: string - name: positive_type_and_description dtype: string - name: domain_topic dtype: string - name: hard_negative_0 dtype: string - name: hard_negative_1 dtype: string - name: hard_negative_2 dtype: string - name: pair_type dtype: string - name: translation_language dtype: string - name: base_language dtype: string - name: generation_params_json dtype: string - name: id dtype: string splits: - name: train num_bytes: 115280377 num_examples: 7033 download_size: 69993312 dataset_size: 115280377 - config_name: QA features: - name: anchor dtype: string - name: positive dtype: string - name: semi_hard_negative dtype: string - name: rephrased_anchor dtype: string - name: rephrased_positive dtype: string - name: positive_summary dtype: string - name: semantic_keywords_for_positive dtype: string - name: positive_wrong_paraphrase dtype: string - name: entailment_from_positive dtype: string - name: not_entailment_from_positive dtype: string - name: positive_core_claims_extractive dtype: string - name: anchor_translated dtype: string - name: positive_translated dtype: string - name: negative_translated dtype: string - name: positive_wrong_translation dtype: string - name: anchor_type_and_intent dtype: string - name: positive_type_and_description dtype: string - name: domain_topic dtype: string - name: hard_negative_0 dtype: string - name: hard_negative_1 dtype: string - name: hard_negative_2 dtype: string - name: pair_type dtype: string - name: translation_language dtype: string - name: base_language dtype: string - name: generation_params_json dtype: string - name: id dtype: string splits: - name: train num_bytes: 161676676 num_examples: 14446 download_size: 99244945 dataset_size: 161676676 --- ## Dataset Statistics **Total items:** 38610 **Total characters (trainable cols):** 464110005 **Average characters per item:** 12020.46 **Estimated Token Count (anchor, positive, negative_0):** ~18,404,798 tokens **Estimated Tokens per item (anchor, positive, negative_0):** ~477 tokens **All columns:** positive_wrong_paraphrase, entailment_from_positive, anchor_translated, hard_negative_1, negative_translated, positive_translated, anchor_type_and_intent, id, hard_negative_0, pair_type, not_entailment_from_positive, base_language, semi_hard_negative, semantic_keywords_for_positive, translation_language, hard_negative_2, positive, rephrased_positive, domain_topic, positive_core_claims_extractive, positive_summary, rephrased_anchor, positive_type_and_description, anchor, positive_wrong_translation, generation_params_json ### Subset Statistics | Split (Config) | Column | Total Characters | Average per Sample | |----------------|--------|------------------|--------------------| | QA | anchor | 4270102 | 295.59 | | QA | positive | 14209849 | 983.65 | | QA | semi_hard_negative | 4446148 | 307.78 | | QA | rephrased_anchor | 4087396 | 282.94 | | QA | rephrased_positive | 13384204 | 926.50 | | QA | positive_summary | 4790387 | 331.61 | | QA | semantic_keywords_for_positive | 3506570 | 242.74 | | QA | positive_wrong_paraphrase | 9205319 | 637.22 | | QA | entailment_from_positive | 1898733 | 131.44 | | QA | not_entailment_from_positive | 1699061 | 117.61 | | QA | positive_core_claims_extractive | 9445783 | 653.87 | | QA | anchor_translated | 4508395 | 312.09 | | QA | positive_translated | 15319202 | 1060.45 | | QA | negative_translated | 11788063 | 816.01 | | QA | positive_wrong_translation | 9372457 | 648.79 | | QA | anchor_type_and_intent | 4027109 | 278.77 | | QA | positive_type_and_description | 5004147 | 346.40 | | QA | hard_negative_0 | 10937142 | 757.11 | | QA | hard_negative_1 | 5703180 | 394.79 | | QA | hard_negative_2 | 5399954 | 373.80 | | CrossRep | anchor | 7633770 | 836.21 | | CrossRep | positive | 6693106 | 733.17 | | CrossRep | semi_hard_negative | 2420802 | 265.18 | | CrossRep | rephrased_anchor | 5976107 | 654.63 | | CrossRep | rephrased_positive | 6391019 | 700.08 | | CrossRep | positive_summary | 2474111 | 271.02 | | CrossRep | semantic_keywords_for_positive | 1978736 | 216.75 | | CrossRep | positive_wrong_paraphrase | 4943957 | 541.57 | | CrossRep | entailment_from_positive | 1105484 | 121.10 | | CrossRep | not_entailment_from_positive | 1003187 | 109.89 | | CrossRep | positive_core_claims_extractive | 4678347 | 512.47 | | CrossRep | anchor_translated | 8139203 | 891.58 | | CrossRep | positive_translated | 7171765 | 785.60 | | CrossRep | negative_translated | 5882189 | 644.34 | | CrossRep | positive_wrong_translation | 5040788 | 552.17 | | CrossRep | anchor_type_and_intent | 2559960 | 280.42 | | CrossRep | positive_type_and_description | 3063178 | 335.54 | | CrossRep | hard_negative_0 | 5511236 | 603.71 | | CrossRep | hard_negative_1 | 3033157 | 332.26 | | CrossRep | hard_negative_2 | 2910877 | 318.86 | | DescA | anchor | 5619329 | 702.24 | | DescA | positive | 13903750 | 1737.53 | | DescA | semi_hard_negative | 3307841 | 413.38 | | DescA | rephrased_anchor | 5224072 | 652.85 | | DescA | rephrased_positive | 13245134 | 1655.23 | | DescA | positive_summary | 3442686 | 430.23 | | DescA | semantic_keywords_for_positive | 2208507 | 275.99 | | DescA | positive_wrong_paraphrase | 9086916 | 1135.58 | | DescA | entailment_from_positive | 1093210 | 136.62 | | DescA | not_entailment_from_positive | 960735 | 120.06 | | DescA | positive_core_claims_extractive | 6867885 | 858.27 | | DescA | anchor_translated | 6141336 | 767.48 | | DescA | positive_translated | 15147294 | 1892.94 | | DescA | negative_translated | 11072172 | 1383.68 | | DescA | positive_wrong_translation | 9330084 | 1165.97 | | DescA | anchor_type_and_intent | 2328597 | 291.00 | | DescA | positive_type_and_description | 2986591 | 373.23 | | DescA | hard_negative_0 | 10119791 | 1264.66 | | DescA | hard_negative_1 | 4522804 | 565.21 | | DescA | hard_negative_2 | 4190393 | 523.67 | | LongA | anchor | 16264687 | 2312.62 | | LongA | positive | 5024599 | 714.43 | | LongA | semi_hard_negative | 2327760 | 330.98 | | LongA | rephrased_anchor | 14040311 | 1996.35 | | LongA | rephrased_positive | 5284395 | 751.37 | | LongA | positive_summary | 2098957 | 298.44 | | LongA | semantic_keywords_for_positive | 1638964 | 233.04 | | LongA | positive_wrong_paraphrase | 4080998 | 580.26 | | LongA | entailment_from_positive | 936921 | 133.22 | | LongA | not_entailment_from_positive | 848355 | 120.62 | | LongA | positive_core_claims_extractive | 3975034 | 565.20 | | LongA | anchor_translated | 16833648 | 2393.52 | | LongA | positive_translated | 5296686 | 753.12 | | LongA | negative_translated | 4516333 | 642.16 | | LongA | positive_wrong_translation | 4123265 | 586.27 | | LongA | anchor_type_and_intent | 2232149 | 317.38 | | LongA | positive_type_and_description | 2217155 | 315.25 | | LongA | hard_negative_0 | 4270883 | 607.26 | | LongA | hard_negative_1 | 2885676 | 410.31 | | LongA | hard_negative_2 | 2799922 | 398.11 | ### ID Format & Pair Type Mapping ID string format: `-------[]`. Below is the mapping for the hashes of distinct `pair_types` encoded in the IDs: | Hash | Original Pair Type | |------|--------------------| | `31380c` | Query to Title | | `8fac93` | Question to Passage | | `eece87` | Instruction to Procedure/Guide/Solution | | `82a3bd` | Keywords to Passage | | `600099` | Query to Passage - WebQuery | | `4743f5` | Request to Example | | `f81e0d` | Question to Answer | | `91b204` | Question to Passage - verify | | `1aff0e` | Problem/Request to Passage | | `d1ab6e` | Multiple Questions to Answer | | `f33a24` | Problem/Request to Procedure/Guide/Solution | | `4b5b56` | Hypotesis to Passage | | `7c0973` | Multiple Questions to Passage | | `cb0b25` | STS | | `314f85` | Contextualized Questions to Answer | | `893226` | Query to Passage | | `63e92f` | Instruction/command to Execution | | `25981f` | Keywords to Notes | | `6835d0` | Keywords to Title | | `6a8e30` | Structured Text to Text | | `47ab92` | Structured Text to Structured Text | | `3e5c30` | Noisy Text to Text | | `ae7cb1` | Code to Text | | `c279c0` | Text to Notes | | `3e414b` | Text to Condensed/Deconstructed Text | | `4ecacb` | Text to Supporting Statements | | `01779a` | Not common Language to Literal/Explanation/Clarification | | `9f745d` | Text to Completion | | `358513` | Structured Query to Passage | | `9d4b28` | Text to Code | | `50e93d` | Noisy Text to Text | | `49c659` | Structured Text to Summary | | `385fc7` | Structured Query to Question | | `0c5aa9` | Formal Language to Natural Language | | `39dae5` | Text to Summary | | `688f64` | Title to Passage | | `86db52` | Notes to Source | | `4f383d` | Summary to Text | | `e12648` | Description/Definition to Description's subject | | `2d54bb` | Description/Definition to Expansion | | `cb1cbe` | Deconstruction/Statements to Text | | `569a52` | Text to Related Text | | `a0c48c` | Text to Supported Statements | | `6f468e` | Text to Deconstruction/Statements | | `5ee92e` | Supported Statements | | `2155a0` | Passage to Description | | `e065ac` | Passage to Title |