Team Ai
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
20likes82kdownloads
DWS_150Pages_TextNeedles_page_116.txt64 linesDownload Raw Back to Text_TextNeedles
1determine fair value. For some instruments, modelling techniques follow industry standard 2models, for example, discounted cash flow analysis and standard option pricing models. 3These models are dependent upon estimated future cash flows, discount factors and volatility 4levels. 5Frequently, valuation models require multiple parameter inputs. Where possible, parameter 6inputs are based on observable data or are derived from the prices of relevant instruments 7traded in active markets. Where observable data is not available for parameter inputs, then 8other market information is considered. For example, indicative broker quotes and consensus 9pricing information are used to support parameter inputs where they are available. 10Level 3 – Valuation techniques using unobservable market data – Where no observable 11information is available to support parameter inputs, then valuation models used are based 12on other relevant sources of information such as prices for similar transactions, historic data, 13economic fundamentals, and research information, with appropriate adjustment to reflect the 14terms of the actual instrument being valued and current market conditions. 15Significant unobservable inputs and valuation adjustments are subject to regular reviews. If 16third party information, such as broker quotes or pricing services, is used to measure fair 17values, then the valuation control group assesses the evidence obtained from the third parties 18to support the conclusion that these valuations meet the requirements of IFRS.19Validation and control – The Group has an established valuation control framework which 20governs internal control standards, methodologies, and procedures over the valuation 21process. The PVCC develops and governs the valuation control framework and ensures 22review and appropriateness of various detailed aspects of the controls such as independent 23price verification classification, testing thresholds and market data approvals. In addition, the 24PVCC reviews the results of completeness controls and ensures that all fair value assets and 25liabilities have been subject to the appropriate valuation control process.26Independent specialised valuation control groups, including a group within Deutsche Bank 27Group´s Risk function, execute the valuation control processes which covers the valuation of 28financial instruments across all levels of the fair value hierarchy. A key focus of these 29specialists is directed to areas where management judgment forms part of the valuation 30process, including regular review of significant unobservable inputs and valuation 31adjustments mentioned above. 32The PVCC oversees the valuation control processes performed by these specialist valuation 33control groups on behalf of the Group. Results of the valuation control processes are collected 34and analysed as part of a standard monthly reporting cycle. Variances outside of pre-set and 35approved tolerance levels are escalated both within the Finance function and Senior Business 36Management for review, resolution and, if required, adjustment. This process is summarised 37in the Valuation Control Report and reviewed by the PVCC.38For instruments where fair value is determined from valuation models, the assumptions and 39techniques used within the models are in scope for validation by Deutsche Bank Group’s 40independent model validation. 41Transfers42Transfers between levels take place when there is a change in the inputs that is relevant to 43categorization in the fair value hierarchy. Where applicable, transfers between levels 1, 2 and 443 are assumed to take place at the beginning of the year.45In 2022 and 2023, level transfers were made to reflect changes in current market liquidity and 46price transparency.47In 2023, there were transfers from level 2 into level 1 of € 6 million driven by corporate bonds 48denominated in EUR and debt instruments held by consolidated guaranteed funds. In 2022, 49there were transfers from level 1 into level 2 of € 73 million and from level 2 into level 1 of 50€ 1 million largely driven by corporate bonds denominated in EUR. 51There were no transfers into and out of level 3 in 2023 and in 2022. 52         53To our 54Shareholders55Summarised 56Management Report57Consolidated 58Financial Statements Compensation Report Corporate Govern-59ance Statement60Supplementary 61Information DWS 2023 Annual Report62 63Notes to the Consolidated Balance Sheet6494 09 – Financial Instruments
AmazonScience/document-haystack · Team Ai