Publish corrected independent-builder census, complete appendix and Zenodo package (v1.1.0)
7b53a50 verified Download data/context/census_context.json from edeneldith/uk-small-model-census: direct link, hf CLI and curl.
- Browser
- Download file 4.76 kB
-
https://huggingface.co/datasets/edeneldith/uk-small-model-census/resolve/main/data/context/census_context.json
- Command line
-
hf download hf://datasets/edeneldith/uk-small-model-census/data/context/census_context.json
-
curl -L -o census_context.json https://huggingface.co/datasets/edeneldith/uk-small-model-census/resolve/main/data/context/census_context.json
4.76 kB
| { | |
| "date_fixed": "2026-10-30", | |
| "compute": { | |
| "currency": "GBP", | |
| "price_date": "2026-10-30", | |
| "bands": [ | |
| { | |
| "label": "Under GBP 500", | |
| "minimum_inclusive": 0, | |
| "maximum_exclusive": 500 | |
| }, | |
| { | |
| "label": "GBP 500-1,999", | |
| "minimum_inclusive": 500, | |
| "maximum_exclusive": 2000 | |
| }, | |
| { | |
| "label": "GBP 2,000-9,999", | |
| "minimum_inclusive": 2000, | |
| "maximum_exclusive": 10000 | |
| }, | |
| { | |
| "label": "GBP 10,000-99,999", | |
| "minimum_inclusive": 10000, | |
| "maximum_exclusive": 100000 | |
| }, | |
| { | |
| "label": "GBP 100,000 or more", | |
| "minimum_inclusive": 100000, | |
| "maximum_exclusive": null | |
| } | |
| ], | |
| "basis": "Current publicly listed UK purchase price including VAT for the documented training accelerator model and memory variant, multiplied by the actual simultaneous device count. Where only a complete CPU laptop is usable and its make/model is known, price that device and label the basis. The accelerator price excludes its host system. Record new/used condition; use a used UK listing when new hardware is no longer sold. Rental charges are recorded separately, not used as the hardware purchase price.", | |
| "missing": "Unknown if actual hardware, count, memory variant or comparable UK price cannot be established; do not borrow hardware from another model or treat a recommended device as an actual run.", | |
| "multiple": "Per-model run records retained. A person appears once in each evidenced band used for their UK-built cohort models; report multiple-response totals and unknowns explicitly. Sequentially replaced devices are separate runs, not a simultaneous array." | |
| }, | |
| "ai_centre_cities": [ | |
| "Cambridge", | |
| "Oxford", | |
| "London", | |
| "Southampton", | |
| "Bristol", | |
| "Edinburgh", | |
| "Leeds", | |
| "Manchester" | |
| ], | |
| "ai_centre_rule": "A training-period locality within one of the eight city areas explicitly named for AI specialisation in the UK government's UK Technology capabilities 2024 brochure: Cambridge, Oxford, Greater London, Southampton, Bristol, Edinburgh, Leeds and Manchester. Use city boundaries (Greater London for London), not the whole South East or a surrounding county. Elsewhere is 'outside the fixed centre list'; a location too broad to decide is unknown.", | |
| "ai_centre_source": "https://paas-s3-broker-prod-lon-8db94025-3db3-41df-9a2a-fbb251b24433.s3.amazonaws.com/documents/UK_Tech_capabilities_2024_HPO_Brochure_A5_Digital.pdf", | |
| "deprivation_rule": "Most deprived 20% of local authorities within the relevant UK nation's latest published official multiple-deprivation index, using rank of average score. Compare rank/N <= 0.20. Use the builder's documented training-period locality, never a current place substituted for it. If only a larger area is known or no comparable authority summary is available, classify unknown. Record neighbourhood deprivation separately only when a public training locality supports it; do not infer a neighbourhood from a town.", | |
| "local_authority_deprivation": { | |
| "Cambridge": { | |
| "rank_of_average_score": 257, | |
| "authority_count": 296, | |
| "most_deprived_fifth": false, | |
| "source_url": "https://assets.publishing.service.gov.uk/media/6917412ebc34c86ce4e6e7fc/File_10_-_IoD2025_Local_Authority_District_Summaries__lower-tier__v2.xlsx" | |
| }, | |
| "Gosport": { | |
| "rank_of_average_score": 101, | |
| "authority_count": 296, | |
| "most_deprived_fifth": false, | |
| "source_url": "https://assets.publishing.service.gov.uk/media/6917412ebc34c86ce4e6e7fc/File_10_-_IoD2025_Local_Authority_District_Summaries__lower-tier__v2.xlsx" | |
| }, | |
| "Oxford": { | |
| "rank_of_average_score": 208, | |
| "authority_count": 296, | |
| "most_deprived_fifth": false, | |
| "source_url": "https://assets.publishing.service.gov.uk/media/6917412ebc34c86ce4e6e7fc/File_10_-_IoD2025_Local_Authority_District_Summaries__lower-tier__v2.xlsx" | |
| } | |
| }, | |
| "schooling": { | |
| "definition": "Primary or secondary schooling in the UK; university study is recorded under route, not substituted for schooling.", | |
| "categories": [ | |
| "Yes", | |
| "No", | |
| "Unknown" | |
| ], | |
| "no_requires": "A documented schooling history locating schooling outside the UK, not a guess from name or nationality." | |
| }, | |
| "education_context_source": "https://explore-education-statistics.service.gov.uk/find-statistics/widening-participation-in-higher-education/2024-25", | |
| "education_context_scope": "England state-school pupils aged15 in2020/21 followed toage19 in2024/25; source release9July2026. The three cited rows are context, not developer demographics." | |
| } | |