CoolFace
Datasetpublic

AmazonScience/document-haystack

Document Haystack Dataset This repository contains the dataset for the paper “Document Haystack: A Long Context Multimodal Image/Document Understanding Vision LLM Benchmark”. 📑 Abstract Paper The proliferation of multimodal Large Language Models has significantly advanced the ability to analyze and understand complex data inputs from different modalities. However, the processing of long documents remains under-explored, largely due to a lack of suitable… See the full description on the dataset page: https://huggingface.co/datasets/AmazonScience/document-haystack.

sourceHugging Faceupdated 1y agoView on Hugging Face
20likes90kdownloads
Entain_200Pages_TextNeedles_page_57.txt119 linesDownload Raw Back to Text_TextNeedles
1Pillar Data point 2023 2022 20212Positive impact 3on communities 4(including 5Streamlined 6Energy & 7Carbon 8Reporting Data) 9Total energy consumption (kWh) 17,1810UK11Rest of the world (RoW)12 124,771,8151377,957,3131446,814,50215125,026,0961682,641,3451742,384,75018110,509,7361985,336,2392025,173,49721Absolute direct emissions (scope 1) – (tCO 2e) 5,899 4,414 3,66322Absolute indirect emissions (scope 2, location-based) – (tCO 2e) 27,202 26,846 24,76723% of purchased electricity from renewable sources 19 70.3% 66.4% 67.4%24Total GHG emissions – direct & indirect: location based (tCO 2e)2025UK 26RoW2733,1012814,8852918,2163031,2593115,5693215,6903328,4303418,2863510,14436Absolute GHG emissions intensity per employee (tCO 2e/headcount) 1.12 1.08 1.1337Absolute indirect emissions (scope 2, market-based) – (tCO 2e) 9,171 12,151 12,67738Total GHG emissions – direct and indirect: market based (tCO 2e)39UK40RoW4115,07142625 4314,4454416,565451,9804614,5854716,340484,9324911,40850Waste generated 21 (tonnes) 3,738 4,384 3,85851Total Scope 3 GHG emissions (tCO 2e)2252Category 1: Purchased Goods & Services (EEIO methodology)53Category 1: Purchased Goods & Services (Supplier specific)54Category 4: Upstream Transportation & Distribution55Category 5: Waste56Category 6: Business Travel57Category 7: Employee Commuting58346,05159312,6036015,726617,87362101635,292644,45665315,55066288,5246712,100686,3996983704,398714,04672Supplier spend £2.8bn £2.7bn £2.1bn73Number of suppliers 12,613 12,006 10,380741. Data covers all Great Britain licenses.752. Data covers all UK licenses. 763. This figure includes all ARC TM real-time packages and risk-based interceptors, as well as ARC TM emails. It is a count of the number of customer interactions, not at a distinct 77customer level. This figure includes the 1,807,892 interactions reported under ‘Customer interactions regarding problem gambling’.784. Data only includes self-exclusions made via Entain’s own processes (e.g., via customer services) and does not include third-party self-exclusion schemes such as, for example, 79GAMSTOP (National Online Self-Exclusion Scheme) and the Multi-operator Self Exclusion Scheme. This information has been obtained from Entain’s Regulatory Returns.805. We use employee headcount to evaluate the scope of our ISO27001 certification. 816. The 2023 figures under the ‘Foster an inclusive culture’ pillar do not include our latest acquisitions 365 Scores and STS as data isn’t yet available for these new subsidiaries at 82the time of publication. Unless stated otherwise, the 2022 figures do not include employees from our November 2022 acquisitions, SuperSport, Puni Broj, and Minus. All figures 83are global unless stated otherwise. The snapshot date for all figures is 31 December 2023 unless otherwise stated.847. The 2022 figures have been revised from the 2022 annual report to include employees from SuperSport, Puni Broj, and Minus 5. The 2022 figures do not include employees from 85SuperSport, Puni Broj, and Minus 5 who have left the business between 1/01/2023 and 31/04/2023.868. Data covers UK colleagues only. Data is based on a snapshot date of 5 April for the year stated, as per the requirements of the UK’s Gender Pay Gap Reporting.879. For the 2021 and 2022 figures, technical colleagues were those employees that rolled up to our Chief Technology Officer based on our Business Process Flow Manager. 88Following changes to the Group’s functions in 2023, technical roles are defined for 2023 as all roles in our Product & Technology function excluding customer operations. 8910. For the 2021 and 2022 figures, revenue-generating functions included our digital and retail/stadia functions. Following changes in the business, revenue-generating functions 90are defined for 2023 as the following functions: Ladbrokes.au/Neds, Core, BetCity, Crystalbet, Enlabs, Eurobet, Labrokes.be, Latam, Retail & Stadia, and BetMGM.9111. This 2023 data is based on a sample of 47% of UK-based Entain employees who have provided us with their ethnicity information. To prevent us from over or understating the 92ethnic diversity of our employees, we report this data in two ways. We report on both the percentage of the sample that identifies as being from ethnic minority backgrounds, as 93well as the number of those confirmed to be identifying as from an ethnic minority background as a proportion of all UK employees.9412. As a percentage of the total number of employees excluding contractors.9513. As a percentage of the total number of employees. 9614. Our methodology to measure customer satisfaction changed in 2023, as we stopped using email surveys and replaced them with digital pop-up surveys shared with customers 97whilst online.9815. Data covers UK retail colleagues only. 9916. We measure employee engagement based on the results of the annual Your Voice survey. The 2023 survey was postponed to January 2024, which is the basis for the 2023 data. 10017. Coverage of energy consumption and emissions data is 100% for the UK, and 87% globally, by employee headcount. Global and ROW energy and emissions data are scaled up 101based on this coverage to estimate totals across global operations. This data includes energy consumption related to both scope 1 (company vehicles, gas, and fuel) and scope 1022 emissions (purchased electricity). Global coverage is below 100% due to limited availability at the time of reporting. Any updates to figures will be provided in our forthcoming 103ESG Report and CDP submission. 10418. Recent acquisitions of 365Scores and STS are not included in the figures due to no data availability at the time of reporting – we will include these entities in our 2024 reporting 105and restate previous years according to our rebaselining policy.10619. Energy from renewable sources only includes electricity purchased that was actively sourced from renewables. All remaining electricity used by Entain is sourced from the local 107grids where we operate.10820. Emissions are calculated using the GHG Protocol Corporate Accounting and Reporting Standard. Consumption data has been converted to GHG emissions using 2023 BEIS 109emissions factors and 2023 IEA emissions factors for non-UK grid electricity. Emissions reported above are calculated using both the location-based and market-based 110methods, using an operational control boundary. 2021 and 2022 GHG emissions (Scope 1 & 2) data has been assured to limited assurance by the Carbon Trust based on ISO 11114064-3: 2019. Verification statements are available on our website. 2021 Scope 1 emissions data has been restated due to a methodology change that arose in the 2022 112assurance process.11321. Waste data is sourced from our operations in the UK. This makes up 49% of our overall headcount. These figures are not prorated to 100% coverage.11422. Scope 3 emissions data disclosed has been verified by the Carbon Trust to ISO 14064-3 for 2022 and 2021. 2023 data was not available at the time of reporting but will be 115disclosed later in 2024.116Entain plc Annual Report 2023 551171 Overview 8 S trategic report 88 Governance 140 F inancial statements118Our ESG Key  119Performance Indicators
AmazonScience/document-haystack · CoolFace