Commit
•
824c01e
1
Parent(s):
6bf18bb
Add new SentenceTransformer model.
Browse files- 1_Pooling/config.json +10 -0
- README.md +1028 -0
- config.json +26 -0
- config_sentence_transformers.json +10 -0
- model.safetensors +3 -0
- modules.json +20 -0
- sentence_bert_config.json +4 -0
- special_tokens_map.json +37 -0
- tokenizer.json +0 -0
- tokenizer_config.json +64 -0
- vocab.txt +0 -0
1_Pooling/config.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"word_embedding_dimension": 384,
|
3 |
+
"pooling_mode_cls_token": false,
|
4 |
+
"pooling_mode_mean_tokens": true,
|
5 |
+
"pooling_mode_max_tokens": false,
|
6 |
+
"pooling_mode_mean_sqrt_len_tokens": false,
|
7 |
+
"pooling_mode_weightedmean_tokens": false,
|
8 |
+
"pooling_mode_lasttoken": false,
|
9 |
+
"include_prompt": true
|
10 |
+
}
|
README.md
ADDED
@@ -0,0 +1,1028 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: sentence-transformers/all-MiniLM-L6-v2
|
3 |
+
datasets: []
|
4 |
+
language: []
|
5 |
+
library_name: sentence-transformers
|
6 |
+
metrics:
|
7 |
+
- pearson_cosine
|
8 |
+
- spearman_cosine
|
9 |
+
- pearson_manhattan
|
10 |
+
- spearman_manhattan
|
11 |
+
- pearson_euclidean
|
12 |
+
- spearman_euclidean
|
13 |
+
- pearson_dot
|
14 |
+
- spearman_dot
|
15 |
+
- pearson_max
|
16 |
+
- spearman_max
|
17 |
+
pipeline_tag: sentence-similarity
|
18 |
+
tags:
|
19 |
+
- sentence-transformers
|
20 |
+
- sentence-similarity
|
21 |
+
- feature-extraction
|
22 |
+
- generated_from_trainer
|
23 |
+
- dataset_size:28450
|
24 |
+
- loss:MultipleNegativesRankingLoss
|
25 |
+
widget:
|
26 |
+
- source_sentence: What are the five criteria that community projects must meet to
|
27 |
+
be considered for funding by the Community Ownership Fund?
|
28 |
+
sentences:
|
29 |
+
- 'We want to fund community projects that do at least 1 of these 5 things:
|
30 |
+
|
31 |
+
increase feelings of pride in, and improve perceptions of, the local area as a
|
32 |
+
place to live
|
33 |
+
|
34 |
+
improve social trust, cohesion, and sense of belonging
|
35 |
+
|
36 |
+
increase local participation in community life, arts, culture, or sport
|
37 |
+
|
38 |
+
improve local economic outcomes – including creating jobs, volunteering opportunities,
|
39 |
+
and improving employability and skills levels in the local community
|
40 |
+
|
41 |
+
improve social and wellbeing outcomes – including having a positive impact on
|
42 |
+
physical and mental health of local people, and reducing loneliness and social
|
43 |
+
isolation
|
44 |
+
|
45 |
+
Strengthening community ownership across the UK
|
46 |
+
|
47 |
+
The Fund will be delivered directly by the UK government to communities in England,
|
48 |
+
Scotland, Wales, and Northern Ireland. The UK government is committed to fair
|
49 |
+
opportunities to access funding through the Community Ownership Fund across the
|
50 |
+
UK.
|
51 |
+
|
52 |
+
A minimum target of spending in line with per-capita allocations has therefore
|
53 |
+
been set in Scotland, Wales, and Northern Ireland. The Community Ownership Fund
|
54 |
+
will target a minimum of £12.3 million in Scotland, £7.1 million in Wales, and
|
55 |
+
£4.3 million in Northern Ireland of the total Fund over the 4 years until March
|
56 |
+
2025.
|
57 |
+
|
58 |
+
The design of the Fund recognises the different landscapes for community ownership
|
59 |
+
across the UK, with different legislation in England and Wales, Scotland, and
|
60 |
+
Northern Ireland. We have engaged widely with local stakeholders to ensure the
|
61 |
+
Fund is effective, accessible and achieves its objectives.
|
62 |
+
|
63 |
+
Applications will be assessed against a consistent framework. Eligibility for
|
64 |
+
the Fund and the bidding assessment criteria are consistent in all 4 nations.
|
65 |
+
|
66 |
+
Glossary
|
67 |
+
|
68 |
+
Community asset
|
69 |
+
|
70 |
+
For this fund, an asset is physical building or space. It must be used by the
|
71 |
+
community and accessible to as many people as possible.
|
72 |
+
|
73 |
+
Community Asset Transfer
|
74 |
+
|
75 |
+
Across the United Kingdom, Community Asset Transfer (CAT) policy frameworks support
|
76 |
+
the transfer of community assets from public authorities to community organisations.
|
77 |
+
The legislation and policy contexts work slightly differently in parts of the
|
78 |
+
United Kingdom.
|
79 |
+
|
80 |
+
England
|
81 |
+
|
82 |
+
Asset of community value
|
83 |
+
|
84 |
+
In England the Localism Act 2011 introduced a right for community groups to nominate
|
85 |
+
buildings or land to their local authority as an asset of community value.
|
86 |
+
|
87 |
+
If the local authority agreed that the nomination met the test of being land of
|
88 |
+
community value, the council would place the asset on a list of assets of community
|
89 |
+
value for a period of 5 years.
|
90 |
+
|
91 |
+
What this did was introduce a community right to bid. If the owner of a listed
|
92 |
+
asset decided that they wish to sell the asset during the 5-year period of listing,
|
93 |
+
then they must notify the local authority who would inform the nominating community
|
94 |
+
group.'
|
95 |
+
- "In designated catchments, water companies have a duty to ensure wastewater treatment\
|
96 |
+
\ works serving a population equivalent over 2,000 meet specified nutrient removal\
|
97 |
+
\ standards by 1 April 2030 where the designation takes effect from 25 January\
|
98 |
+
\ 2024. For designations that take effect subsequent to that date, the upgrade\
|
99 |
+
\ date is specified in the notice. Competent authorities (including local planning\
|
100 |
+
\ authorities) considering planning proposals for development draining via a sewer\
|
101 |
+
\ to a wastewater treatment works subject to the upgrade duty are required to\
|
102 |
+
\ consider that the nutrient pollution standard will be met by the upgrade date\
|
103 |
+
\ for the purposes of Habitats Regulations Assessments. \nWhilst the upgrade\
|
104 |
+
\ date under the Water Industry Act 1991 for this catchment is 16 May 2031, the\
|
105 |
+
\ sewerage undertaker has committed to the delivery of the wastewater treatment\
|
106 |
+
\ work upgrades by 1 April 2030. The Environment Agency has also committed to\
|
107 |
+
\ varying Environmental Permits for the relevant wastewater treatment works so\
|
108 |
+
\ that the permits will require compliance with the nutrient pollution standard\
|
109 |
+
\ by 1 April 2030. ↩"
|
110 |
+
- 'https://gcscc.ox.ac.uk/cmm-reviews#/ ↩
|
111 |
+
|
112 |
+
World Bank, ‘Green Digital Transformation: How to Sustainably Close the Digital
|
113 |
+
Divide and Harness Digital Tools for Climate Action’ https://openknowledge.worldbank.org/entities/
|
114 |
+
publication/6be73f14-f899-4a6d-a26e-56d98393acf3 ↩
|
115 |
+
|
116 |
+
Ritchie, 2020 https://ourworldindata.org/ghg-emissions-by-sector ↩
|
117 |
+
|
118 |
+
WHO, e-waste factsheet, 2023: https://www.who.int/news-room/fact-sheets/detail/
|
119 |
+
electronic-waste-(e-waste) ↩
|
120 |
+
|
121 |
+
International development in a contested world: ending extreme poverty and tackling
|
122 |
+
climate change https://www.gov.uk/government/publications/international-development-in-a-contested-world-ending-extreme-poverty-and-tackling-climate-change
|
123 |
+
↩
|
124 |
+
|
125 |
+
https://www.gov.uk/government/publications/greening-government-ict-and-digitalservices-strategy-2020-2025
|
126 |
+
↩
|
127 |
+
|
128 |
+
UK Government’s Department for Environment, Food & Rural Affairs ↩
|
129 |
+
|
130 |
+
https://digitalprinciples.org/ ↩
|
131 |
+
|
132 |
+
https://www.dynamicspectrumalliance.org/ ↩
|
133 |
+
|
134 |
+
https://www.itu.int/itu-d/sites/partner2connect/ ↩
|
135 |
+
|
136 |
+
https://www.govstack.global/ ↩'
|
137 |
+
- source_sentence: What specific actions is the UK government implementing as part
|
138 |
+
of the third National Adaptation Programme (NAP3) to address the impacts of climate
|
139 |
+
change?
|
140 |
+
sentences:
|
141 |
+
- '(The Thames Barrier in London, shown at low tide. Photo by mikeinlondon via Getty
|
142 |
+
Images.)
|
143 |
+
|
144 |
+
The government is taking action to adapt the UK to climate change. This can help
|
145 |
+
reduce the costs from climate change impacts and make our economy and society
|
146 |
+
more resilient.
|
147 |
+
|
148 |
+
This page explains more about:
|
149 |
+
|
150 |
+
climate change and adaptation
|
151 |
+
|
152 |
+
the risks and opportunities of climate change
|
153 |
+
|
154 |
+
what the government is doing to make sure that the UK is prepared for climate
|
155 |
+
change – including the third National Adaptation Programme (NAP3)
|
156 |
+
|
157 |
+
Climate change
|
158 |
+
|
159 |
+
Our climate is changing. The main cause is human activity: in particular, burning
|
160 |
+
fossil fuels for energy, which emits greenhouse gases into the atmosphere and
|
161 |
+
causes the world’s temperature to rise.
|
162 |
+
|
163 |
+
In the UK we can see the effects of climate change already. In 2022 the UK recorded
|
164 |
+
the warmest year on record with temperatures reaching over 40°C, which had impacts
|
165 |
+
on public health and the environment. These temperatures would not have been possible
|
166 |
+
without climate change caused by human activity. The frequency of hotter summers
|
167 |
+
will increase in the future, and we can expect the winters to become wetter, which
|
168 |
+
will make flooding more likely across the UK.
|
169 |
+
|
170 |
+
The government is taking action to limit climate change through its commitment
|
171 |
+
to reach net zero greenhouse gas emissions by 2050. One of these actions is reducing
|
172 |
+
our reliance on fossil fuels. Achieving ‘net zero’ in the UK and across the world
|
173 |
+
will help to limit temperature rises in the future and reduce the level of climate
|
174 |
+
change we need to adapt to.
|
175 |
+
|
176 |
+
Climate adaptation
|
177 |
+
|
178 |
+
Climate adaptation relates to actions that protect us against the impacts of climate
|
179 |
+
change. This includes reacting to the changes we have seen already, as well as
|
180 |
+
preparing for what will happen in the future.
|
181 |
+
|
182 |
+
The UK government is taking steps to address the impacts of climate change to
|
183 |
+
protect communities, our economy and the environment.
|
184 |
+
|
185 |
+
Examples of the government’s approach to climate adaptation include:
|
186 |
+
|
187 |
+
building new flood defences to protect against rising sea levels
|
188 |
+
|
189 |
+
planning for more green spaces in urban areas to help keep them cool and planting
|
190 |
+
more drought-resistant crops
|
191 |
+
|
192 |
+
building infrastructure that can withstand expected climate impacts such as extreme
|
193 |
+
heat and flooding
|
194 |
+
|
195 |
+
Many of the actions in NAP3 can help to improve our standard of living too, by
|
196 |
+
upgrading our buildings and infrastructure, improving the sustainability and productivity
|
197 |
+
of important sectors such as agriculture and forestry, and restoring our natural
|
198 |
+
environment.
|
199 |
+
|
200 |
+
Climate risks and opportunities
|
201 |
+
|
202 |
+
Climate change can lead to both risks and opportunities, although there are more
|
203 |
+
risks than opportunities. Without measures to adapt to climate change, we would
|
204 |
+
experience additional issues including:
|
205 |
+
|
206 |
+
health risks
|
207 |
+
|
208 |
+
damage to houses and infrastructure'
|
209 |
+
- 'We will help shape an international order in which all citizens are well informed,
|
210 |
+
able to participate in democratic processes and enjoy their rights in offline
|
211 |
+
and online public spaces, as well as freedom of expression; and we will promote
|
212 |
+
an information ecosystem that supports accountability and inclusive deliberative
|
213 |
+
democracy.
|
214 |
+
|
215 |
+
The UK commits to an open, free, global, interoperable, reliable and secure Internet;
|
216 |
+
and to ensuring emerging tech supports, rather than erodes, the enjoyment of democracy,
|
217 |
+
human rights and fundamental freedoms. Working collectively with international
|
218 |
+
partners, civil society and the tech sector is critical in ensuring that the online
|
219 |
+
world and technologies promote freedom, democracy and inclusion, and protect human
|
220 |
+
rights and fundamental freedoms.
|
221 |
+
|
222 |
+
We will strengthen our collaboration in the multi-stakeholder spaces that support
|
223 |
+
digital democracy. We will enhance our advisory support to the Freedom Online
|
224 |
+
Coalition (FOC) and will bid to continue as a member of the FOC Steering Committee
|
225 |
+
and to maintain our role as co-chairs of the Taskforce on Internet Shutdowns (TFIS).
|
226 |
+
|
227 |
+
We will support our overseas network to better understand the threat posed by
|
228 |
+
information disorder through digital platforms. In doing so, we will identify
|
229 |
+
international best practice and increase our understanding of information disorder
|
230 |
+
in elections, independent media as well as gendered disinformation impacts on
|
231 |
+
women’s political empowerment and participation in electoral processes.
|
232 |
+
|
233 |
+
We will champion the importance of a vibrant, independent, and pluralistic civic
|
234 |
+
space online and offline, where people can exercise their freedoms. We will work
|
235 |
+
in collaboration with other donors, civil society, academia and the private sector
|
236 |
+
to leverage the opportunities and mitigate the risks that digital transformation
|
237 |
+
provides for civil society and civic space.
|
238 |
+
|
239 |
+
We will support open and accountable use of emerging digital technologies, especially
|
240 |
+
the need for democratic and human rights safeguards. This includes grant support
|
241 |
+
for the Open Government Partnership to help enable open and accountable use of
|
242 |
+
emerging digital technologies by driving digital governance reforms in 10 countries
|
243 |
+
(Ghana, Indonesia, Kenya, Nigeria, Dominic Republic, Armenia, Colombia, Zambia,
|
244 |
+
the Philippines and Ukraine), accelerating collective action and norm-raising
|
245 |
+
on digital governance and increasing impact through better connection between
|
246 |
+
global pledges and country action.
|
247 |
+
|
248 |
+
Chapter 3 – Digital inclusion: leaving no one behind in a digital world
|
249 |
+
|
250 |
+
The benefits of digital transformation are not evenly distributed. A third of
|
251 |
+
the world’s population is offline, and that is concentrated within the poorest
|
252 |
+
and most marginalised groups.'
|
253 |
+
- 'Estimated one-off impact on administrative burden (£ million)
|
254 |
+
|
255 |
+
One-off impact (£ million) £30,000 to £50,000 threshold Above £50,000 threshold
|
256 |
+
Total mandated population above £30,000
|
257 |
+
|
258 |
+
Costs 338 223 561
|
259 |
+
|
260 |
+
Savings — — —
|
261 |
+
|
262 |
+
Estimated continuing impact on administrative burden (£ million)
|
263 |
+
|
264 |
+
Continuing average annual impact (£ million) £30,000 to £50,000 threshold Above
|
265 |
+
£50,000 threshold Total mandated population above £30,000
|
266 |
+
|
267 |
+
Costs 110 90 201
|
268 |
+
|
269 |
+
Savings 2 3 5
|
270 |
+
|
271 |
+
Net impact on annual administrative burden +108 +88 +196
|
272 |
+
|
273 |
+
Numbers do not sum due to rounding.
|
274 |
+
|
275 |
+
Operational impact (£ million) (HMRC or other)
|
276 |
+
|
277 |
+
There will be both IT and resource costs for HMRC in developing, applying, and
|
278 |
+
policing this measure, and in updating guidance.
|
279 |
+
|
280 |
+
HMRC IT and non-IT costs for this next phase of MTD expansion are expected to
|
281 |
+
be in the region of £0.5bn to the end of March 2028.
|
282 |
+
|
283 |
+
Other impacts
|
284 |
+
|
285 |
+
HMRC is required to consider the justice impact test and rural proofing measures
|
286 |
+
in relation to their impacts on rural communities and the justice system.
|
287 |
+
|
288 |
+
HMRC’s assessments suggest any impact is likely to be negligible. Mitigations
|
289 |
+
are in place for those whose rural location impacts their internet access to the
|
290 |
+
point where it is not feasible to operate MTD, as discussed in the ‘Equalities
|
291 |
+
impacts’ section.
|
292 |
+
|
293 |
+
This measure does not fall within the scope of the environmental principles duty.
|
294 |
+
|
295 |
+
Other impacts have been considered and none have been identified.
|
296 |
+
|
297 |
+
Monitoring and evaluation
|
298 |
+
|
299 |
+
HMRC’s communications programme includes work to build software developer, agent
|
300 |
+
and taxpayer readiness, to promote inclusion in the large-scale public beta testing
|
301 |
+
programme beginning in 2025 and encourage voluntary early adoption of MTD for
|
302 |
+
ITSA.
|
303 |
+
|
304 |
+
HMRC is committed to monitoring and formally evaluating the impact of MTD for
|
305 |
+
ITSA, including both customer and revenue impacts. This will build on HMRC’s track
|
306 |
+
record in successfully evaluating MTD for VAT and publishing the findings. Independent
|
307 |
+
social research will be undertaken both before and after MTD for ITSA is introduced
|
308 |
+
to gather evidence of customer impacts and behaviour change. Self Assessment data
|
309 |
+
will be used to monitor take-up and estimate additional tax revenue due to MTD.
|
310 |
+
The evaluation will take until at least 2029, when all data for the 2027 to 28
|
311 |
+
tax year becomes available for analysis.
|
312 |
+
|
313 |
+
Further advice'
|
314 |
+
- source_sentence: Who are the joint leaders of the new Anti-social Behaviour Taskforce
|
315 |
+
responsible for overseeing the implementation and delivery of the action plan?
|
316 |
+
sentences:
|
317 |
+
- '80. It is also vital that we measure the overall success of this plan in tackling
|
318 |
+
anti-social behaviour to ensure that it is meeting the commitments we have set
|
319 |
+
out. We will assess the impact of our proposals on both communities’ experience
|
320 |
+
and perceptions of anti-social behaviour and their effectiveness in tackling it.
|
321 |
+
To achieve this, we will draw from the wide range of data enhancements outlined
|
322 |
+
throughout this plan, alongside wider measures, to monitor and evaluate its success
|
323 |
+
and to further inform our understanding of what works in driving down anti-social
|
324 |
+
behaviour.
|
325 |
+
|
326 |
+
81. We will oversee the implementation and delivery to this action plan with a
|
327 |
+
new Anti-social Behaviour Taskforce jointly led by the Home Secretary and the
|
328 |
+
Secretary of State for Levelling Up that will bring together national and local
|
329 |
+
partners, with a sole focus of addressing anti-social behaviour and restoring
|
330 |
+
pride in place in communities.
|
331 |
+
|
332 |
+
Home Office. Anti-social behaviour: impacts on individuals and local communities.
|
333 |
+
2023 ↩
|
334 |
+
|
335 |
+
Home Office. Guidance: Anti-social behaviour principles. 2022. ↩
|
336 |
+
|
337 |
+
Home Office. Anti-social behaviour: impacts on individuals and local communities.
|
338 |
+
2023. ↩
|
339 |
+
|
340 |
+
YouGov. Anti-Social Behaviour. 2023. ↩
|
341 |
+
|
342 |
+
A legal definition of ASB can be found in the Anti-Social Behaviour Act 2014:
|
343 |
+
a) conduct that has caused, or is likely to cause, harassment, alarm or distress
|
344 |
+
to any person, b) conduct capable of causing nuisance or annoyance to a person
|
345 |
+
in relation to that person’s occupation of residential premises, or c) conduct
|
346 |
+
capable of causing housing-related nuisance or annoyance to any person. ↩
|
347 |
+
|
348 |
+
Ipsos. Ipsos Levelling Up Index: Levelling up Panel. 2022. ↩
|
349 |
+
|
350 |
+
Public First. Levelling Up Poll. 2021. ↩
|
351 |
+
|
352 |
+
Office for National Statistics. Crime in England and Wales: Other related tables
|
353 |
+
. 2022. ↩
|
354 |
+
|
355 |
+
Office for National Statistics. Crime Survey for England and Wales (CSEW) estimates
|
356 |
+
of personal and household crime, anti-social behaviour, and public perceptions,
|
357 |
+
by police force area, year ending September 2022. ↩
|
358 |
+
|
359 |
+
Office for National Statistics. Crime in England and Wales: Police Force Area
|
360 |
+
data tables. 2023. Office for National Statistics. Crime in England and Wales:
|
361 |
+
Other related tables. 2023. Office for National Statistics. Crime in England and
|
362 |
+
Wales: Annual Trend and Demographic Tables. 2022. ↩'
|
363 |
+
- '323. Similarly, DCMS Ministers in both Houses of Parliament expressed at the
|
364 |
+
dispatch box their disappointment about the proposed changes to BBC local radio
|
365 |
+
services. There have also been several instances over the Charter period where
|
366 |
+
a lack of effective transparency in engaging the public has been highlighted in
|
367 |
+
the media and by Parliamentarians. For example, the BBC’s failure to explain how
|
368 |
+
it was dealing with complaints about the anti-semitic incident on a bus on Oxford
|
369 |
+
Street at the end of 2021 in the face of significant public pressure received
|
370 |
+
widespread media coverage. The announcement of the closure of BBC Singers led
|
371 |
+
to Parliamentary discussions and media reports raising concerns about how the
|
372 |
+
decision had been made and communicated, including internally within the BBC.
|
373 |
+
|
374 |
+
The government’s response
|
375 |
+
|
376 |
+
324. When considering how the BBC communicates with audiences, it is our view
|
377 |
+
that the BBC should be held to a higher standard than other organisations given
|
378 |
+
the extent of its public funding. This higher standard needs to go beyond publication
|
379 |
+
of more data and information, to straightforward and open communication with audiences.
|
380 |
+
The BBC Board has overall responsibility for ensuring that the BBC communicates
|
381 |
+
changes that have an impact on audiences effectively with those audiences. This
|
382 |
+
has to be accompanied by equally effective communication with its workforce. Evidence
|
383 |
+
received indicates that the BBC has not always achieved this.
|
384 |
+
|
385 |
+
7.1 We recommend that the BBC continues to learn from recent experiences where
|
386 |
+
announcements about service changes have led to criticism about the BBC’s approach
|
387 |
+
to transparency.
|
388 |
+
|
389 |
+
7.2 We also recommend that the BBC publishes details of its strategy for communicating
|
390 |
+
with audiences which explains improvements to its communications approach already
|
391 |
+
made, but also how it identifies any changes needed so that audiences and staff
|
392 |
+
can be confident that future service changes and their impact will be explained
|
393 |
+
clearly.
|
394 |
+
|
395 |
+
Understanding audience needs
|
396 |
+
|
397 |
+
What we learnt
|
398 |
+
|
399 |
+
325. During evidence gathering, many stakeholders made proposals regarding how
|
400 |
+
the BBC could improve its transparency in specific ways to help audiences hold
|
401 |
+
it to account. All of these proposals related to individual specific themes in
|
402 |
+
previous chapters. Ofcom’s research suggests that there are perception issues
|
403 |
+
with the BBC’s impartiality that more effective transparency could help address.
|
404 |
+
|
405 |
+
The government’s response
|
406 |
+
|
407 |
+
326. It is important that licence fee payers do not just have the opportunity
|
408 |
+
to shape the services that the BBC provides, but that they also have the opportunity
|
409 |
+
to tell the BBC how they would like the BBC to be more transparent.'
|
410 |
+
- '67. Building on our Fraud Plan, DWP is investing £70 million between 2022/23
|
411 |
+
and 2024/25 in advanced analytics to tackle fraud and error, which it expects
|
412 |
+
will help it to generate savings of around £1.6 billion by 2030/31[footnote 24].
|
413 |
+
|
414 |
+
68. Investing in advanced analytics, such as machine learning, is essential to
|
415 |
+
enable the public sector to keep up with offenders. Sophisticated crimminals already
|
416 |
+
utilise such tools to analyse large amounts of data to exploit existing weaknesses
|
417 |
+
and vulnerabilities in public sector systems. In DWP these tools can play a crucial
|
418 |
+
role in detecting and preventing fraudulent activities in DWPs benefit systems.
|
419 |
+
Going forward we want to maximise the benefits that advanced analytics and machine
|
420 |
+
learning can offer.
|
421 |
+
|
422 |
+
69. Where these tools are used to assist in the prevention and detection of fraud,
|
423 |
+
DWP always ensures appropriate safeguards are in place to ensure the proportionate,
|
424 |
+
ethical, and lawful use of data with human input. In decision making, any final
|
425 |
+
decision will always be made by a member of DWP staff and DWP seeks to ensure
|
426 |
+
compliance using internal monitoring protocols. DWPs Personal Information Charter
|
427 |
+
sets out in more detail how the Department uses these tools, as well as Artificial
|
428 |
+
Intelligence and automated decision making.
|
429 |
+
|
430 |
+
Continuous improvement to Universal Credit (UC)
|
431 |
+
|
432 |
+
70. As we complete the Move to UC, the Department’s spending on UC alone is forecast
|
433 |
+
to double (relative to 2022/23 in nominal terms) to reach over £85 billion by
|
434 |
+
2028/29[footnote 25].
|
435 |
+
|
436 |
+
71. We are constantly improving UC to reduce fraud and error and to ensure the
|
437 |
+
right support reaches the right people.
|
438 |
+
|
439 |
+
72. Building on our previous Fraud Plan our UC Continuous Improvement plan brings
|
440 |
+
together multi-disciplinary teams to look at the largest areas of loss within
|
441 |
+
UC and considers how we can improve our processes to reduce these.
|
442 |
+
|
443 |
+
73. These teams focus on understanding the root-causes and scale of the losses,
|
444 |
+
design and test solutions with a view to implementing them more widely if the
|
445 |
+
tests are successful. The implementation of these solutions may involve changes
|
446 |
+
to policy, improvements to the operation of UC service or greater use of data
|
447 |
+
and automation to prevent the fraud.'
|
448 |
+
- source_sentence: What is the date and time of the next meeting?
|
449 |
+
sentences:
|
450 |
+
- 'Defra is working with the British Standards Institution (BSI) to develop a suite
|
451 |
+
of nature investment standards that will support best practice standardisation
|
452 |
+
of methodologies with regards to best practices for assessing the baseline, monitoring,
|
453 |
+
and verifying the delivery of nature-based carbon removals. This will be critical
|
454 |
+
for the purposes of supplying and selling credits into nature markets, and for
|
455 |
+
quantifying within value chain mitigation of environmental impacts. These standards
|
456 |
+
will build on and aim to align with the work of international integrity initiatives,
|
457 |
+
including the Integrity Council for Voluntary Carbon Markets (ICVCM) and the Voluntary
|
458 |
+
Carbon Markets Initiative (VCMI).
|
459 |
+
|
460 |
+
As part of this programme, BSI is developing the ‘Nature markets - Overarching
|
461 |
+
principles and framework’, which will apply to nature-based environmental improvement
|
462 |
+
projects and the quantification of ecosystem services. These principles will set
|
463 |
+
the basis by which nature markets can be more effectively designed and governed.
|
464 |
+
A first draft of the BSI Flex 701 standard was published for consultation in March
|
465 |
+
2024.
|
466 |
+
|
467 |
+
Further to this, BSI will be developing more specific thematic and market specific
|
468 |
+
standards to follow over the course of 2024 to 2025, for example, for nature-based
|
469 |
+
carbon and biodiversity. This will include a certification mechanism to allow
|
470 |
+
methodologies which meet these standards to become certified as offering high
|
471 |
+
integrity.
|
472 |
+
|
473 |
+
1.2 A standardised approach to product level impact quantification
|
474 |
+
|
475 |
+
Increasingly, businesses are seeing the benefits of communicating product level
|
476 |
+
impact data to consumers and other businesses in the supply chain. Product level
|
477 |
+
accounting can help improve understanding of the impacts of specific products
|
478 |
+
and supply chains to inform changes at the supplier and product level to reduce
|
479 |
+
impacts. Product level data can also enable more accurate reporting of company
|
480 |
+
impacts from the ‘bottom-up’, by summing up the impact of all products sold by
|
481 |
+
the company, in addition to any energy use or emissions on site.
|
482 |
+
|
483 |
+
Product level impact data is generated through lifecycle assessments (LCAs). Although
|
484 |
+
there are many commonalities between Scope 3 and product carbon footprinting,
|
485 |
+
there are a number of practical and methodological differences summarised in section
|
486 |
+
4.1 of the WRAP Protocol.
|
487 |
+
|
488 |
+
Relevant priorities
|
489 |
+
|
490 |
+
1.3 – A standardised product level accounting method (including multi-metric approach)
|
491 |
+
|
492 |
+
Developing a product level accounting method'
|
493 |
+
- 'To enable efficient and extensive use of genomic AMR data, the design and implementation
|
494 |
+
of data handling solutions will be explored. The design should accommodate complexities
|
495 |
+
such as AMR outbreaks caused by the same AMR-causing mobile genetic element transferred
|
496 |
+
among different pathogen species, or longer-term trends in AMR epidemiology. These
|
497 |
+
should provide new or use existing open standards, for the handling of AMR-related
|
498 |
+
information, to facilitate working with international partners and allow convenient
|
499 |
+
and effective querying for surveillance and response planning. Few countries offer
|
500 |
+
large scale sequencing and analysis of AMR associated isolates so UK data would
|
501 |
+
provide vital insight into the molecular epidemiology of these infections and
|
502 |
+
position the UK to exploit the knowledge these new methods can provide.
|
503 |
+
|
504 |
+
Theme 2 - Optimising the use of antimicrobials
|
505 |
+
|
506 |
+
Outcome 4 - Antimicrobial stewardship and disposal
|
507 |
+
|
508 |
+
By 2029, the UK has strengthened antimicrobial stewardship and diagnostic stewardship
|
509 |
+
by improved targeting of antimicrobials and diagnostic tools for humans, animals
|
510 |
+
and plants, and improved the disposal of antimicrobials, informed by the right
|
511 |
+
data, risk stratification and guidance.
|
512 |
+
|
513 |
+
This outcome has:
|
514 |
+
|
515 |
+
3 commitments:
|
516 |
+
|
517 |
+
clinical decision support
|
518 |
+
|
519 |
+
appropriate prescribing and disposal
|
520 |
+
|
521 |
+
behavioural interventions
|
522 |
+
|
523 |
+
2 human health targets (see appendix B):
|
524 |
+
|
525 |
+
target 4a: by 2029, we aim to reduce total antibiotic use in human populations
|
526 |
+
by 5% from the 2019 baseline
|
527 |
+
|
528 |
+
target 4b: by 2029, we aim to achieve 70% of total use of antibiotics from the
|
529 |
+
Access category (new UK category) across the human healthcare system
|
530 |
+
|
531 |
+
While all use of antimicrobials drives AMR, there is an opportunity to reduce
|
532 |
+
inappropriate use of antimicrobials occurring, for example, when antimicrobials
|
533 |
+
are taken when they are not needed, or when taken for longer than necessary.
|
534 |
+
|
535 |
+
According to the National Institute for Health and Care Excellence’s NICE guideline
|
536 |
+
(NG15):
|
537 |
+
|
538 |
+
The term ‘antimicrobial stewardship’ is defined as ‘an organisational or healthcare‑system‑wide
|
539 |
+
approach to promoting and monitoring judicious use of antimicrobials to preserve
|
540 |
+
their future effectiveness’.'
|
541 |
+
- 'None.
|
542 |
+
|
543 |
+
Date of next meeting: 1 December 2021 at 11am to 12.30pm'
|
544 |
+
- source_sentence: How much funding has the government committed to expand the Public
|
545 |
+
Sector Fraud Authority to deploy AI in combating fraud?
|
546 |
+
sentences:
|
547 |
+
- '2) Embracing the opportunities presented by making greater use of cutting-edge
|
548 |
+
technology, such as AI, across the public sector. The government is:
|
549 |
+
|
550 |
+
More than doubling the size of i.AI, the AI incubator team, ensuring that the
|
551 |
+
UK government has the in-house expertise consisting of the most talented technology
|
552 |
+
professionals in the UK, who can apply their skills and expertise to appropriately
|
553 |
+
seize the benefits of AI across the public sector and Civil Service.
|
554 |
+
|
555 |
+
Committing £34 million to expand the Public Sector Fraud Authority by deploying
|
556 |
+
AI to help combat fraud across the public sector, making it easier to spot, stop
|
557 |
+
and catch fraudsters thereby saving £100 million for the public purse.
|
558 |
+
|
559 |
+
Committing £17 million to accelerate DWP’s digital transformation, replacing paper-based
|
560 |
+
processes with simplified online services, such as a new system for the Child
|
561 |
+
Maintenance Service.
|
562 |
+
|
563 |
+
Committing £14 million for public sector research and innovation infrastructure.
|
564 |
+
This includes funding to develop the next generation of health and security technologies,
|
565 |
+
unlocking productivity improvements in the public and private sector alike.
|
566 |
+
|
567 |
+
3) Strengthening preventative action to reduce demand on public services. The
|
568 |
+
government is:
|
569 |
+
|
570 |
+
Committing an initial £105 million towards a wave of 15 new special free schools
|
571 |
+
to create over 2,000 additional places for children with special educational needs
|
572 |
+
and disabilities (SEND) across England. This will help more children receive a
|
573 |
+
world-class education and builds on the significant levels of capital funding
|
574 |
+
for SEND invested at the 2021 Spending Review. The locations of these special
|
575 |
+
free schools will be announced by May 2024.
|
576 |
+
|
577 |
+
Confirming the location of 20 Alternative Provision (AP) free schools, which will
|
578 |
+
create over 1,600 additional AP places across England as part of the Spending
|
579 |
+
Review 2021 commitment to invest £2.6 billion capital in high needs provision.
|
580 |
+
This will support early intervention, helping improve outcomes for children requiring
|
581 |
+
alternative provision, and helping them to fulfil their potential.'
|
582 |
+
- "We will help build the UKDev (UK International Development) approach and brand\
|
583 |
+
\ by leveraging the UK’s comparative advantage within both the public and private\
|
584 |
+
\ sectors. We will build first and foremost on existing successful partnerships,\
|
585 |
+
\ through which we share UK models and expertise to support digital transformation\
|
586 |
+
\ in partner countries. For example, through our collaboration with the British\
|
587 |
+
\ Standards Institution (BSI) we will expand our collaboration to build the capacity\
|
588 |
+
\ of partner countries in Africa and South-East Asia (including through ASEAN)\
|
589 |
+
\ on digital standards, working with local private sector and national standards-setting\
|
590 |
+
\ bodies.\nWe will strengthen our delivery of peer learning activities in collaboration\
|
591 |
+
\ with Ofcom, exchanging experiences and sharing the UK models on spectrum management,\
|
592 |
+
\ local networks and other technical areas with telecoms regulators in partner\
|
593 |
+
\ countries, building on the positive peer-learning experience with Kenya and\
|
594 |
+
\ South Africa.\nWe will collaborate with Government Digital Service (GDS) to\
|
595 |
+
\ share know-how with partner countries on digitalisation in the public sector,\
|
596 |
+
\ building on our advisory role in GovStack[footnote 56]. We will leverage the\
|
597 |
+
\ UK experience of DPI for public or regulated services (health, transport, banking,\
|
598 |
+
\ land registries) based on the significant demand for this expertise from developing\
|
599 |
+
\ countries and riding the momentum on DPI generated by the G20 India presidency\
|
600 |
+
\ of 2023.\n 6.4 Enhancing FCDO’s digital development capability\nThe UK government\
|
601 |
+
\ will also enhance its own digital development capability to keep up with the\
|
602 |
+
\ pace of technological change, to be forward-looking and anticipate emergent\
|
603 |
+
\ benefits and risks of digital transformation. We will invest in new research\
|
604 |
+
\ on digital technologies and on their inclusive business models to build the\
|
605 |
+
\ global evidence base, share lessons learned and improve knowledge management\
|
606 |
+
\ through our portfolio of digital development and technology programmes, including\
|
607 |
+
\ the FCDO’s new Technology Centre for Expertise (Tech CoE), which will complement\
|
608 |
+
\ and support our programming portfolio.\nSince all sectors within international\
|
609 |
+
\ development are underpinned by digital technologies, we will ensure that digital\
|
610 |
+
\ development skills are mainstreamed across the FCDO. We will raise awareness\
|
611 |
+
\ and upgrade staff knowledge through new training opportunities on best practice\
|
612 |
+
\ in the complex and evolving area of digital development, through partnering\
|
613 |
+
\ with existing FCDO capability initiatives, ie the International Academy’s Development\
|
614 |
+
\ Faculty, the Cyber Network and the International Technology curriculum."
|
615 |
+
- "The Burma (Sanctions) (EU Exit) Regulations 2019 (S.I. 2019/136) (revoked) 29\
|
616 |
+
\ January 2019 To ensure that the UK continues to operate an effective sanctions\
|
617 |
+
\ regime in relation to Burma after end of the Transition Period, replacing with\
|
618 |
+
\ substantially the same effect the EU sanctions regime relating to Burma that\
|
619 |
+
\ was previously in force in the UK under EU legislation and related UK legislation.\
|
620 |
+
\ Section 2(4) report (PDF, 74 KB) and section 18 report (PDF, 65 KB).\nThe Burma\
|
621 |
+
\ (Sanctions) (Overseas Territories) Order 2020 (S.I. 2020/1264) (revoked)[footnote\
|
622 |
+
\ 81] 11 November 2020 To extend with modifications The Burma (Sanctions) (EU\
|
623 |
+
\ Exit) Regulations 2019 (S.I. 2019/136) as amended from time to time to all British\
|
624 |
+
\ Overseas Territories except Bermuda and Gibraltar (which implement sanctions\
|
625 |
+
\ under their own legislative arrangements). \nThe Myanmar (Sanctions) Regulations\
|
626 |
+
\ 2021 (S.I. 2021/496) 26 April 2021 To establish a UK autonomous sanctions regime\
|
627 |
+
\ in respect of Myanmar comprising financial, immigration and trade sanctions,\
|
628 |
+
\ replacing the existing sanctions regime established by The Burma (Sanctions)\
|
629 |
+
\ (EU Exit) Regulations 2019 (S.I. 2019/136). \nThe Myanmar (Sanctions) (Overseas\
|
630 |
+
\ Territories) Order 2021 (S.I. 2021/528) 28 April 2021 To extend with modifications\
|
631 |
+
\ The Myanmar (Sanctions) Regulations 2021 (S.I. 2021/496) as amended from time\
|
632 |
+
\ to time to all British Overseas Territories except Bermuda and Gibraltar (which\
|
633 |
+
\ implement sanctions under their own legislative arrangements). \nThe Myanmar\
|
634 |
+
\ (Sanctions) (Isle of Man) Order 2021 (S.I. 2021/529) 28 April 2021 To extend\
|
635 |
+
\ to the Isle of Man with modifications The Myanmar (Sanctions) Regulations 2021\
|
636 |
+
\ (S.I. 2021/496) as amended from time to time. \nSee also in section (C) of\
|
637 |
+
\ this Annex:\nthe Sanctions Regulations (Commencement No. 1) (EU Exit) Regulations\
|
638 |
+
\ 2019 (S.I. 2019/627)\nthe Sanctions (EU Exit) (Miscellaneous Amendments) (No.\
|
639 |
+
\ 2) Regulations 2020 (S.I. 2020/590)\nthe Sanctions (EU Exit) (Miscellaneous\
|
640 |
+
\ Amendments) (No. 4) Regulations 2020 (S.I. 2020/951)\nthe Sanctions (EU Exit)\
|
641 |
+
\ (Miscellaneous Amendments) (No. 2) Regulations 2022 (S.I. 2022/818)\nStatutory\
|
642 |
+
\ guidance for this regime was published on 29 April 2021.\n19. Nicaragua"
|
643 |
+
model-index:
|
644 |
+
- name: SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2
|
645 |
+
results:
|
646 |
+
- task:
|
647 |
+
type: semantic-similarity
|
648 |
+
name: Semantic Similarity
|
649 |
+
dataset:
|
650 |
+
name: sts dev
|
651 |
+
type: sts-dev
|
652 |
+
metrics:
|
653 |
+
- type: pearson_cosine
|
654 |
+
value: 0.8601045098831278
|
655 |
+
name: Pearson Cosine
|
656 |
+
- type: spearman_cosine
|
657 |
+
value: 0.8581596602965272
|
658 |
+
name: Spearman Cosine
|
659 |
+
- type: pearson_manhattan
|
660 |
+
value: 0.8604789808039027
|
661 |
+
name: Pearson Manhattan
|
662 |
+
- type: spearman_manhattan
|
663 |
+
value: 0.8571595448874573
|
664 |
+
name: Spearman Manhattan
|
665 |
+
- type: pearson_euclidean
|
666 |
+
value: 0.8615938042335468
|
667 |
+
name: Pearson Euclidean
|
668 |
+
- type: spearman_euclidean
|
669 |
+
value: 0.8581596602965272
|
670 |
+
name: Spearman Euclidean
|
671 |
+
- type: pearson_dot
|
672 |
+
value: 0.8601045118561034
|
673 |
+
name: Pearson Dot
|
674 |
+
- type: spearman_dot
|
675 |
+
value: 0.8581596602965272
|
676 |
+
name: Spearman Dot
|
677 |
+
- type: pearson_max
|
678 |
+
value: 0.8615938042335468
|
679 |
+
name: Pearson Max
|
680 |
+
- type: spearman_max
|
681 |
+
value: 0.8581596602965272
|
682 |
+
name: Spearman Max
|
683 |
+
---
|
684 |
+
|
685 |
+
# SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2
|
686 |
+
|
687 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2). It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
|
688 |
+
|
689 |
+
## Model Details
|
690 |
+
|
691 |
+
### Model Description
|
692 |
+
- **Model Type:** Sentence Transformer
|
693 |
+
- **Base model:** [sentence-transformers/all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2) <!-- at revision 8b3219a92973c328a8e22fadcfa821b5dc75636a -->
|
694 |
+
- **Maximum Sequence Length:** 256 tokens
|
695 |
+
- **Output Dimensionality:** 384 tokens
|
696 |
+
- **Similarity Function:** Cosine Similarity
|
697 |
+
<!-- - **Training Dataset:** Unknown -->
|
698 |
+
<!-- - **Language:** Unknown -->
|
699 |
+
<!-- - **License:** Unknown -->
|
700 |
+
|
701 |
+
### Model Sources
|
702 |
+
|
703 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
704 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
705 |
+
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
|
706 |
+
|
707 |
+
### Full Model Architecture
|
708 |
+
|
709 |
+
```
|
710 |
+
SentenceTransformer(
|
711 |
+
(0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
|
712 |
+
(1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
|
713 |
+
(2): Normalize()
|
714 |
+
)
|
715 |
+
```
|
716 |
+
|
717 |
+
## Usage
|
718 |
+
|
719 |
+
### Direct Usage (Sentence Transformers)
|
720 |
+
|
721 |
+
First install the Sentence Transformers library:
|
722 |
+
|
723 |
+
```bash
|
724 |
+
pip install -U sentence-transformers
|
725 |
+
```
|
726 |
+
|
727 |
+
Then you can load this model and run inference.
|
728 |
+
```python
|
729 |
+
from sentence_transformers import SentenceTransformer
|
730 |
+
|
731 |
+
# Download from the 🤗 Hub
|
732 |
+
model = SentenceTransformer("AndreasThinks/all-MiniLM-L6-v2_policy_doc_finetune")
|
733 |
+
# Run inference
|
734 |
+
sentences = [
|
735 |
+
'How much funding has the government committed to expand the Public Sector Fraud Authority to deploy AI in combating fraud?',
|
736 |
+
'2) Embracing the opportunities presented by making greater use of cutting-edge technology, such as AI, across the public sector. The government is:\nMore than doubling the size of i.AI, the AI incubator team, ensuring that the UK government has the in-house expertise consisting of the most talented technology professionals in the UK, who can apply their skills and expertise to appropriately seize the benefits of AI across the public sector and Civil Service.\nCommitting £34 million to expand the Public Sector Fraud Authority by deploying AI to help combat fraud across the public sector, making it easier to spot, stop and catch fraudsters thereby saving £100 million for the public purse.\nCommitting £17 million to accelerate DWP’s digital transformation, replacing paper-based processes with simplified online services, such as a new system for the Child Maintenance Service.\nCommitting £14 million for public sector research and innovation infrastructure. This includes funding to develop the next generation of health and security technologies, unlocking productivity improvements in the public and private sector alike.\n3) Strengthening preventative action to reduce demand on public services. The government is:\nCommitting an initial £105 million towards a wave of 15 new special free schools to create over 2,000 additional places for children with special educational needs and disabilities (SEND) across England. This will help more children receive a world-class education and builds on the significant levels of capital funding for SEND invested at the 2021 Spending Review. The locations of these special free schools will be announced by May 2024.\nConfirming the location of 20 Alternative Provision (AP) free schools, which will create over 1,600 additional AP places across England as part of the Spending Review 2021 commitment to invest £2.6 billion capital in high needs provision. This will support early intervention, helping improve outcomes for children requiring alternative provision, and helping them to fulfil their potential.',
|
737 |
+
'We will help build the UKDev (UK International Development) approach and brand by leveraging the UK’s comparative advantage within both the public and private sectors. We will build first and foremost on existing successful partnerships, through which we share UK models and expertise to support digital transformation in partner countries. For example, through our collaboration with the British Standards Institution (BSI) we will expand our collaboration to build the capacity of partner countries in Africa and South-East Asia (including through ASEAN) on digital standards, working with local private sector and national standards-setting bodies.\nWe will strengthen our delivery of peer learning activities in collaboration with Ofcom, exchanging experiences and sharing the UK models on spectrum management, local networks and other technical areas with telecoms regulators in partner countries, building on the positive peer-learning experience with Kenya and South Africa.\nWe will collaborate with Government Digital Service (GDS) to share know-how with partner countries on digitalisation in the public sector, building on our advisory role in GovStack[footnote 56]. We will leverage the UK experience of DPI for public or regulated services (health, transport, banking, land registries) based on the significant demand for this expertise from developing countries and riding the momentum on DPI generated by the G20 India presidency of 2023.\n 6.4 Enhancing FCDO’s digital development capability\nThe UK government will also enhance its own digital development capability to keep up with the pace of technological change, to be forward-looking and anticipate emergent benefits and risks of digital transformation. We will invest in new research on digital technologies and on their inclusive business models to build the global evidence base, share lessons learned and improve knowledge management through our portfolio of digital development and technology programmes, including the FCDO’s new Technology Centre for Expertise (Tech CoE), which will complement and support our programming portfolio.\nSince all sectors within international development are underpinned by digital technologies, we will ensure that digital development skills are mainstreamed across the FCDO. We will raise awareness and upgrade staff knowledge through new training opportunities on best practice in the complex and evolving area of digital development, through partnering with existing FCDO capability initiatives, ie the International Academy’s Development Faculty, the Cyber Network and the International Technology curriculum.',
|
738 |
+
]
|
739 |
+
embeddings = model.encode(sentences)
|
740 |
+
print(embeddings.shape)
|
741 |
+
# [3, 384]
|
742 |
+
|
743 |
+
# Get the similarity scores for the embeddings
|
744 |
+
similarities = model.similarity(embeddings, embeddings)
|
745 |
+
print(similarities.shape)
|
746 |
+
# [3, 3]
|
747 |
+
```
|
748 |
+
|
749 |
+
<!--
|
750 |
+
### Direct Usage (Transformers)
|
751 |
+
|
752 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
753 |
+
|
754 |
+
</details>
|
755 |
+
-->
|
756 |
+
|
757 |
+
<!--
|
758 |
+
### Downstream Usage (Sentence Transformers)
|
759 |
+
|
760 |
+
You can finetune this model on your own dataset.
|
761 |
+
|
762 |
+
<details><summary>Click to expand</summary>
|
763 |
+
|
764 |
+
</details>
|
765 |
+
-->
|
766 |
+
|
767 |
+
<!--
|
768 |
+
### Out-of-Scope Use
|
769 |
+
|
770 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
771 |
+
-->
|
772 |
+
|
773 |
+
## Evaluation
|
774 |
+
|
775 |
+
### Metrics
|
776 |
+
|
777 |
+
#### Semantic Similarity
|
778 |
+
* Dataset: `sts-dev`
|
779 |
+
* Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
|
780 |
+
|
781 |
+
| Metric | Value |
|
782 |
+
|:--------------------|:-----------|
|
783 |
+
| pearson_cosine | 0.8601 |
|
784 |
+
| **spearman_cosine** | **0.8582** |
|
785 |
+
| pearson_manhattan | 0.8605 |
|
786 |
+
| spearman_manhattan | 0.8572 |
|
787 |
+
| pearson_euclidean | 0.8616 |
|
788 |
+
| spearman_euclidean | 0.8582 |
|
789 |
+
| pearson_dot | 0.8601 |
|
790 |
+
| spearman_dot | 0.8582 |
|
791 |
+
| pearson_max | 0.8616 |
|
792 |
+
| spearman_max | 0.8582 |
|
793 |
+
|
794 |
+
<!--
|
795 |
+
## Bias, Risks and Limitations
|
796 |
+
|
797 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
798 |
+
-->
|
799 |
+
|
800 |
+
<!--
|
801 |
+
### Recommendations
|
802 |
+
|
803 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
804 |
+
-->
|
805 |
+
|
806 |
+
## Training Details
|
807 |
+
|
808 |
+
### Training Hyperparameters
|
809 |
+
#### Non-Default Hyperparameters
|
810 |
+
|
811 |
+
- `eval_strategy`: steps
|
812 |
+
- `per_device_train_batch_size`: 16
|
813 |
+
- `per_device_eval_batch_size`: 16
|
814 |
+
- `learning_rate`: 2e-05
|
815 |
+
- `num_train_epochs`: 2
|
816 |
+
- `warmup_ratio`: 0.1
|
817 |
+
- `use_mps_device`: True
|
818 |
+
- `batch_sampler`: no_duplicates
|
819 |
+
|
820 |
+
#### All Hyperparameters
|
821 |
+
<details><summary>Click to expand</summary>
|
822 |
+
|
823 |
+
- `overwrite_output_dir`: False
|
824 |
+
- `do_predict`: False
|
825 |
+
- `eval_strategy`: steps
|
826 |
+
- `prediction_loss_only`: True
|
827 |
+
- `per_device_train_batch_size`: 16
|
828 |
+
- `per_device_eval_batch_size`: 16
|
829 |
+
- `per_gpu_train_batch_size`: None
|
830 |
+
- `per_gpu_eval_batch_size`: None
|
831 |
+
- `gradient_accumulation_steps`: 1
|
832 |
+
- `eval_accumulation_steps`: None
|
833 |
+
- `learning_rate`: 2e-05
|
834 |
+
- `weight_decay`: 0.0
|
835 |
+
- `adam_beta1`: 0.9
|
836 |
+
- `adam_beta2`: 0.999
|
837 |
+
- `adam_epsilon`: 1e-08
|
838 |
+
- `max_grad_norm`: 1.0
|
839 |
+
- `num_train_epochs`: 2
|
840 |
+
- `max_steps`: -1
|
841 |
+
- `lr_scheduler_type`: linear
|
842 |
+
- `lr_scheduler_kwargs`: {}
|
843 |
+
- `warmup_ratio`: 0.1
|
844 |
+
- `warmup_steps`: 0
|
845 |
+
- `log_level`: passive
|
846 |
+
- `log_level_replica`: warning
|
847 |
+
- `log_on_each_node`: True
|
848 |
+
- `logging_nan_inf_filter`: True
|
849 |
+
- `save_safetensors`: True
|
850 |
+
- `save_on_each_node`: False
|
851 |
+
- `save_only_model`: False
|
852 |
+
- `restore_callback_states_from_checkpoint`: False
|
853 |
+
- `no_cuda`: False
|
854 |
+
- `use_cpu`: False
|
855 |
+
- `use_mps_device`: True
|
856 |
+
- `seed`: 42
|
857 |
+
- `data_seed`: None
|
858 |
+
- `jit_mode_eval`: False
|
859 |
+
- `use_ipex`: False
|
860 |
+
- `bf16`: False
|
861 |
+
- `fp16`: False
|
862 |
+
- `fp16_opt_level`: O1
|
863 |
+
- `half_precision_backend`: auto
|
864 |
+
- `bf16_full_eval`: False
|
865 |
+
- `fp16_full_eval`: False
|
866 |
+
- `tf32`: None
|
867 |
+
- `local_rank`: 0
|
868 |
+
- `ddp_backend`: None
|
869 |
+
- `tpu_num_cores`: None
|
870 |
+
- `tpu_metrics_debug`: False
|
871 |
+
- `debug`: []
|
872 |
+
- `dataloader_drop_last`: False
|
873 |
+
- `dataloader_num_workers`: 0
|
874 |
+
- `dataloader_prefetch_factor`: None
|
875 |
+
- `past_index`: -1
|
876 |
+
- `disable_tqdm`: False
|
877 |
+
- `remove_unused_columns`: True
|
878 |
+
- `label_names`: None
|
879 |
+
- `load_best_model_at_end`: False
|
880 |
+
- `ignore_data_skip`: False
|
881 |
+
- `fsdp`: []
|
882 |
+
- `fsdp_min_num_params`: 0
|
883 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
884 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
885 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
|
886 |
+
- `deepspeed`: None
|
887 |
+
- `label_smoothing_factor`: 0.0
|
888 |
+
- `optim`: adamw_torch
|
889 |
+
- `optim_args`: None
|
890 |
+
- `adafactor`: False
|
891 |
+
- `group_by_length`: False
|
892 |
+
- `length_column_name`: length
|
893 |
+
- `ddp_find_unused_parameters`: None
|
894 |
+
- `ddp_bucket_cap_mb`: None
|
895 |
+
- `ddp_broadcast_buffers`: False
|
896 |
+
- `dataloader_pin_memory`: True
|
897 |
+
- `dataloader_persistent_workers`: False
|
898 |
+
- `skip_memory_metrics`: True
|
899 |
+
- `use_legacy_prediction_loop`: False
|
900 |
+
- `push_to_hub`: False
|
901 |
+
- `resume_from_checkpoint`: None
|
902 |
+
- `hub_model_id`: None
|
903 |
+
- `hub_strategy`: every_save
|
904 |
+
- `hub_private_repo`: False
|
905 |
+
- `hub_always_push`: False
|
906 |
+
- `gradient_checkpointing`: False
|
907 |
+
- `gradient_checkpointing_kwargs`: None
|
908 |
+
- `include_inputs_for_metrics`: False
|
909 |
+
- `eval_do_concat_batches`: True
|
910 |
+
- `fp16_backend`: auto
|
911 |
+
- `push_to_hub_model_id`: None
|
912 |
+
- `push_to_hub_organization`: None
|
913 |
+
- `mp_parameters`:
|
914 |
+
- `auto_find_batch_size`: False
|
915 |
+
- `full_determinism`: False
|
916 |
+
- `torchdynamo`: None
|
917 |
+
- `ray_scope`: last
|
918 |
+
- `ddp_timeout`: 1800
|
919 |
+
- `torch_compile`: False
|
920 |
+
- `torch_compile_backend`: None
|
921 |
+
- `torch_compile_mode`: None
|
922 |
+
- `dispatch_batches`: None
|
923 |
+
- `split_batches`: None
|
924 |
+
- `include_tokens_per_second`: False
|
925 |
+
- `include_num_input_tokens_seen`: False
|
926 |
+
- `neftune_noise_alpha`: None
|
927 |
+
- `optim_target_modules`: None
|
928 |
+
- `batch_eval_metrics`: False
|
929 |
+
- `batch_sampler`: no_duplicates
|
930 |
+
- `multi_dataset_batch_sampler`: proportional
|
931 |
+
|
932 |
+
</details>
|
933 |
+
|
934 |
+
### Training Logs
|
935 |
+
| Epoch | Step | Training Loss | loss | sts-dev_spearman_cosine |
|
936 |
+
|:------:|:----:|:-------------:|:------:|:-----------------------:|
|
937 |
+
| 0.0562 | 100 | 0.3598 | 0.8263 | 0.8672 |
|
938 |
+
| 0.1124 | 200 | 0.1983 | 0.7948 | 0.8666 |
|
939 |
+
| 0.1686 | 300 | 0.2021 | 0.7623 | 0.8666 |
|
940 |
+
| 0.2248 | 400 | 0.1844 | 0.7510 | 0.8657 |
|
941 |
+
| 0.2811 | 500 | 0.1704 | 0.7575 | 0.8629 |
|
942 |
+
| 0.3373 | 600 | 0.1643 | 0.7348 | 0.8641 |
|
943 |
+
| 0.3935 | 700 | 0.1808 | 0.7293 | 0.8604 |
|
944 |
+
| 0.4497 | 800 | 0.1494 | 0.7232 | 0.8636 |
|
945 |
+
| 0.5059 | 900 | 0.1563 | 0.7161 | 0.8634 |
|
946 |
+
| 0.5621 | 1000 | 0.1345 | 0.7115 | 0.8643 |
|
947 |
+
| 0.6183 | 1100 | 0.1344 | 0.7142 | 0.8617 |
|
948 |
+
| 0.6745 | 1200 | 0.1584 | 0.7106 | 0.8622 |
|
949 |
+
| 0.7307 | 1300 | 0.1488 | 0.7130 | 0.8592 |
|
950 |
+
| 0.7870 | 1400 | 0.1391 | 0.7034 | 0.8635 |
|
951 |
+
| 0.8432 | 1500 | 0.1433 | 0.7140 | 0.8614 |
|
952 |
+
| 0.8994 | 1600 | 0.1393 | 0.7067 | 0.8612 |
|
953 |
+
| 0.9556 | 1700 | 0.1644 | 0.6950 | 0.8628 |
|
954 |
+
| 1.0118 | 1800 | 0.1399 | 0.7072 | 0.8594 |
|
955 |
+
| 1.0680 | 1900 | 0.12 | 0.7093 | 0.8594 |
|
956 |
+
| 1.1242 | 2000 | 0.0904 | 0.7040 | 0.8587 |
|
957 |
+
| 1.1804 | 2100 | 0.082 | 0.6962 | 0.8585 |
|
958 |
+
| 1.2366 | 2200 | 0.0715 | 0.6985 | 0.8593 |
|
959 |
+
| 1.2929 | 2300 | 0.0624 | 0.7233 | 0.8562 |
|
960 |
+
| 1.3491 | 2400 | 0.0725 | 0.7064 | 0.8581 |
|
961 |
+
| 1.4053 | 2500 | 0.0665 | 0.7034 | 0.8570 |
|
962 |
+
| 1.4615 | 2600 | 0.0616 | 0.6940 | 0.8584 |
|
963 |
+
| 1.5177 | 2700 | 0.0703 | 0.6886 | 0.8599 |
|
964 |
+
| 1.5739 | 2800 | 0.0564 | 0.6860 | 0.8603 |
|
965 |
+
| 1.6301 | 2900 | 0.0603 | 0.6962 | 0.8590 |
|
966 |
+
| 1.6863 | 3000 | 0.0729 | 0.6906 | 0.8589 |
|
967 |
+
| 1.7426 | 3100 | 0.0753 | 0.6946 | 0.8579 |
|
968 |
+
| 1.7988 | 3200 | 0.0711 | 0.6909 | 0.8582 |
|
969 |
+
| 1.8550 | 3300 | 0.0743 | 0.6896 | 0.8583 |
|
970 |
+
| 1.9112 | 3400 | 0.0693 | 0.6902 | 0.8581 |
|
971 |
+
| 1.9674 | 3500 | 0.0845 | 0.6904 | 0.8582 |
|
972 |
+
|
973 |
+
|
974 |
+
### Framework Versions
|
975 |
+
- Python: 3.10.13
|
976 |
+
- Sentence Transformers: 3.0.1
|
977 |
+
- Transformers: 4.41.2
|
978 |
+
- PyTorch: 2.3.1
|
979 |
+
- Accelerate: 0.31.0
|
980 |
+
- Datasets: 2.20.0
|
981 |
+
- Tokenizers: 0.19.1
|
982 |
+
|
983 |
+
## Citation
|
984 |
+
|
985 |
+
### BibTeX
|
986 |
+
|
987 |
+
#### Sentence Transformers
|
988 |
+
```bibtex
|
989 |
+
@inproceedings{reimers-2019-sentence-bert,
|
990 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
991 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
992 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
993 |
+
month = "11",
|
994 |
+
year = "2019",
|
995 |
+
publisher = "Association for Computational Linguistics",
|
996 |
+
url = "https://arxiv.org/abs/1908.10084",
|
997 |
+
}
|
998 |
+
```
|
999 |
+
|
1000 |
+
#### MultipleNegativesRankingLoss
|
1001 |
+
```bibtex
|
1002 |
+
@misc{henderson2017efficient,
|
1003 |
+
title={Efficient Natural Language Response Suggestion for Smart Reply},
|
1004 |
+
author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
|
1005 |
+
year={2017},
|
1006 |
+
eprint={1705.00652},
|
1007 |
+
archivePrefix={arXiv},
|
1008 |
+
primaryClass={cs.CL}
|
1009 |
+
}
|
1010 |
+
```
|
1011 |
+
|
1012 |
+
<!--
|
1013 |
+
## Glossary
|
1014 |
+
|
1015 |
+
*Clearly define terms in order to be accessible across audiences.*
|
1016 |
+
-->
|
1017 |
+
|
1018 |
+
<!--
|
1019 |
+
## Model Card Authors
|
1020 |
+
|
1021 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
1022 |
+
-->
|
1023 |
+
|
1024 |
+
<!--
|
1025 |
+
## Model Card Contact
|
1026 |
+
|
1027 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
1028 |
+
-->
|
config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "sentence-transformers/all-MiniLM-L6-v2",
|
3 |
+
"architectures": [
|
4 |
+
"BertModel"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"gradient_checkpointing": false,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 384,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 1536,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 512,
|
16 |
+
"model_type": "bert",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 6,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.41.2",
|
23 |
+
"type_vocab_size": 2,
|
24 |
+
"use_cache": true,
|
25 |
+
"vocab_size": 30522
|
26 |
+
}
|
config_sentence_transformers.json
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"__version__": {
|
3 |
+
"sentence_transformers": "3.0.1",
|
4 |
+
"transformers": "4.41.2",
|
5 |
+
"pytorch": "2.3.1"
|
6 |
+
},
|
7 |
+
"prompts": {},
|
8 |
+
"default_prompt_name": null,
|
9 |
+
"similarity_fn_name": null
|
10 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:199e37a7161d7ea9bcd8b2fdb1f810660fe559f38072edafa6e209fd49872a22
|
3 |
+
size 90864192
|
modules.json
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"idx": 0,
|
4 |
+
"name": "0",
|
5 |
+
"path": "",
|
6 |
+
"type": "sentence_transformers.models.Transformer"
|
7 |
+
},
|
8 |
+
{
|
9 |
+
"idx": 1,
|
10 |
+
"name": "1",
|
11 |
+
"path": "1_Pooling",
|
12 |
+
"type": "sentence_transformers.models.Pooling"
|
13 |
+
},
|
14 |
+
{
|
15 |
+
"idx": 2,
|
16 |
+
"name": "2",
|
17 |
+
"path": "2_Normalize",
|
18 |
+
"type": "sentence_transformers.models.Normalize"
|
19 |
+
}
|
20 |
+
]
|
sentence_bert_config.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"max_seq_length": 256,
|
3 |
+
"do_lower_case": false
|
4 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": {
|
3 |
+
"content": "[CLS]",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"mask_token": {
|
10 |
+
"content": "[MASK]",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "[PAD]",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"sep_token": {
|
24 |
+
"content": "[SEP]",
|
25 |
+
"lstrip": false,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"unk_token": {
|
31 |
+
"content": "[UNK]",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
}
|
37 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": true,
|
48 |
+
"mask_token": "[MASK]",
|
49 |
+
"max_length": 128,
|
50 |
+
"model_max_length": 256,
|
51 |
+
"never_split": null,
|
52 |
+
"pad_to_multiple_of": null,
|
53 |
+
"pad_token": "[PAD]",
|
54 |
+
"pad_token_type_id": 0,
|
55 |
+
"padding_side": "right",
|
56 |
+
"sep_token": "[SEP]",
|
57 |
+
"stride": 0,
|
58 |
+
"strip_accents": null,
|
59 |
+
"tokenize_chinese_chars": true,
|
60 |
+
"tokenizer_class": "BertTokenizer",
|
61 |
+
"truncation_side": "right",
|
62 |
+
"truncation_strategy": "longest_first",
|
63 |
+
"unk_token": "[UNK]"
|
64 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|