id
stringlengths 6
113
| author
stringlengths 2
36
| task_category
stringclasses 42
values | tags
listlengths 1
4.05k
| created_time
timestamp[ns, tz=UTC]date 2022-03-02 23:29:04
2025-04-10 08:38:38
| last_modified
stringdate 2020-05-14 13:13:12
2025-04-19 04:15:39
| downloads
int64 0
118M
| likes
int64 0
4.86k
| README
stringlengths 30
1.01M
| matched_bigbio_names
listlengths 1
8
⌀ | is_bionlp
stringclasses 3
values | model_cards
stringlengths 0
1M
| metadata
stringlengths 2
698k
| source
stringclasses 2
values | matched_task
listlengths 1
10
⌀ | __index_level_0__
int64 0
46.9k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Areeb-02/mpnet-base-GISTEmbedLoss-MSEE_Evaluator-salestax-docs
|
Areeb-02
|
sentence-similarity
|
[
"sentence-transformers",
"safetensors",
"mpnet",
"sentence-similarity",
"feature-extraction",
"generated_from_trainer",
"dataset_size:238",
"loss:GISTEmbedLoss",
"arxiv:1908.10084",
"arxiv:2402.16829",
"base_model:microsoft/mpnet-base",
"base_model:finetune:microsoft/mpnet-base",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-06-21T22:47:22Z |
2024-06-21T22:47:54+00:00
| 7 | 0 |
---
base_model: microsoft/mpnet-base
datasets: []
language: []
library_name: sentence-transformers
metrics:
- negative_mse
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:238
- loss:GISTEmbedLoss
widget:
- source_sentence: What is the timeframe for opening a physical location in the Designated
Areas to be eligible for the credit under Section 960.1?
sentences:
- '(a) A person or combined group that opens a physical location in the Designated
Areas on or after January 1, 2023 through and including December 31, 2027, shall
be allowed
a credit against that person or combined group''s Gross Receipts Tax if the person
or
74SAN FIAT CU
2024 11AY 10 PM 1331
combined group did not have a physical location in the City for at least three
years prior to
DEPARTMENT OF ELECTIONS
opening the physical location. The credit under this Section 960.1 shall be an
annual credit
for each of up to three tax years immediately following the tax year in which
the person or
combined group opened the physical location in the Designated Areas, provided
the person or
combined group maintains a physical location in the Designated Areas in the tax
year that the
credit is taken. To be eligible for the credit, the person or combined group must
take the
credit for each tax year on an original Gross Receipts Tax return filed with the
Tax Collector.'
- '5222 (Nondepository Credit Intermediation), 5223 (Activities Related
to Credit Intermediation)), 533 (Lessors of Nonfinancial Intangible Assets (except
Copyrighted
Works)), 54 (Professional, Scientific, and Technical Services) but not including
541714 (Research and
Development in Biotechnology (except Nanobiotechnology)), 55 (Management of Companies
and
63SA
CL
2024 LAY 10 PA (30
Enterprises), 562 (Waste Management and Remediation Services), 61 (Educational
Services),
DEPARTMENT OF ELECTI
62 (Health Care and Social Assistance), and all business activities not otherwise
exempt and not
elsewhere subjected to a gross receipts tax rate by Sections 953.20 through 953.26
or an administrative
office tax under Section 953.8.
(c) The amount of taxable gross receipts from Category 5 Business Activities subject
to the
gross receipts tax shall be three-quarters of the amount determined under Section
956.1 plus
one-quarter of the amount determined under Section 956.2.
SEC. 953.25. GROSS RECEIPTS TAX APPLICABLE TO CATEGORY 6 BUSINESS
ACTIVITIES.
(a) The gross receipts tax rates applicable to Category 6 Business Activities
are:
(1) For tax years 2025 and 2026:
1.5% for taxable gross receipts between $0 and $1,000,000
1.5% for taxable gross receipts between $1,000,000.01 and $2,500,000
3% for taxable gross receipts between $2,500,000.01 and $25,000,000
2.352% for taxable gross receipts between $25,000,000.01 and $50,000,000
3.024% for taxable gross receipts between $50,000,000.01 and $75,000,000
3.024% for taxable gross receipts between $75,000,000.01 and $100,000,000
3.36% for taxable gross receipts between $100,000,000.01 and $150,000,000
3.36% for taxable gross receipts between $150,000,000.01 and $250,000,000
3.36% for taxable gross receipts between $250,000,000.01 and $500,000,000
3.'
- 'Except as otherwise provided in this Section 855
and Section 856 of this Article 12, and Section 76.3 of Article 2 of this Business
and Tax Regulations
Code, the annual fee for obtaining a registration certificate for the 2025-2026
registration year,
payable in advance, shall be as follows:
San Francisco Gross Receipts for the
Immediately Preceding Tax Year
$0 to $100,000
$100,000.01 to $250,000
Annual Registration Fee
$41
$71
$250,000.01 to $500,000
$120
$500,000.01 to $750,000
$240
$750,000.01 to $1,000,000
$330
$1,000,000.01 to $1,500,000
$469
$1,500,000.01 to $2,000,000
$656
$2,000,000.01 to $2,500,000
$848
$2,500,000.01 to $5,000,000
$1,414
$5,000,000.01 to $7,500,000
$600
$7,500,000.01 to $15,000,000
$1,500
$15,000,000.01 to $25,000,000
$4,875
25$25,000,000.01 to $50,000,000
$50,000,000.01 to $100,000,000
$100,000,000.01 to $200,000,000
$200,000,000.01 and over
SAN KAL CO
2724 11 10 11 129
$15,000
DEPARTMENT
$30,000
$37,500
$45,000
(d) Fee for Registration Years Beginning On or After April 1, 2026. Except as
otherwise
provided in this Section 855 and Section 856 of this Article 12, and Section 76.3
of Article 2 of this
Business and Tax Regulations Code, the annual fee for obtaining a registration
certificate for
registration years beginning on or after April 1, 2026, payable in advance, shall
be as follows:
San Francisco Gross Receipts for the
Immediately Preceding Tax Year
$0 to $100,'
- source_sentence: How is the overpaid executive administrative office tax measured
for individuals or combined groups, and what other administrative office taxes
might they also be liable for?
sentences:
- '(b) Combined gross receipts are the total worldwide gross receipts of the person
and
all related entities to the person, unless the election provided for in California
Revenue and
Taxation Code Section 25110 is in effect for the person, in which case combined
gross
receipts shall be computed consistently with the water''s edge election, as set
forth therein.
(c) Combined payroll is the total worldwide compensation paid by the person and
all
related entities to the person, unless the election provided for in California
Revenue and
Taxation Code Section 25110 is in effect for the person, in which case combined
payroll shall
be computed consistently with the water''s edge election, as set forth therein.
A person who
72SALA 50
has no combined payroll in a tax year shall have no gross receipts under this
Section for that
DEPARTMENT OF LECT
tax year.
(d) Payroll in the City is the total amount paid for compensation in the City
by the
person and by all related entities to the person.
(e) Compensation paid in the City shall be determined as follows:set forth in
Section 953.8)(4).
(1) Where compensation is paid by reason of work performed or services rendered
by
an individual wholly within the City, all of the compensation for such individual
shall be attributable to
the City.
(2) Where compensation is paid by reason of work performed or services rendered
by
an individual partly within and partly without the City, the portion of such compensation
attributable to
the City shall be determined as follows:
(A) Except as otherwise provided in this Section 956.2(e), the portion of such
compensation attributable to the City shall be the portion of such compensation
which the total number
of working hours employed within the City bears to the total number of working
hours within and
without the City.
(B) If the amount of such compensation depends on the volume of business
transacted by such individual, then the portion of such compensation attributable
to the City shall be
the portion of such compensation which the volume of business transacted by such
individual in the
City bears to the volume of business transacted by such individual within and
without the City.'
- 'If the person or combined group did not file a
7SANT
V
772 MAY 10 TH 1:28
return for the preceding tax year and did not request an extension under Section
6.9-4 to file that
return, the person or combined group shall be deemed to have filed a return showing
no
liability for purposes of this Section 6.9-Isubsection (c)(3)(C)(ii), and no estimated
tax payments
of that tax shall be due for the current tax year.
(D) Lessor of Residential Real Estate Exemption. Notwithstanding
any other provision in this Section 6.9-1(c)(3), a lessor of residential real
estate, as defined in
Section 954.1 of Article 12-A-1, shall not be required to make estimated tax payments
under
this Section 6.9-1(c)(3), but shall pay its full Gross Receipts Tax liability,
Homelessness Gross
Receipts Tax liability, and Overpaid Executive Gross Receipts Tax liability on
or before the
last day of February of the succeeding year, or the extended deadline in Section
6.9-4, if the
lessor''s gross receipts within the City shown on the lessor''s return for either
the current tax
year or the preceding tax year did not exceed the threshold in Section 954.1(b)
of
Article 12-A-1.
*
*
(g) Business Registration Certificate; Annual Due Date.
(1) For registration years ending on or before June 30, 2025 and the
2025-2026 registration year, eExcept for any extensions granted under Section
6.9-4 or as
otherwise provided in this Business and Tax Regulations Code, returns and payments
of the
Business Registration Fee (Article 12) shall be filed annually and are due and
payable, and
shall be delinquent if not submitted and paid to the Tax Collector, on or before
the last day of
May preceding the registration year commencing July 1 of that year.'
- 'DEPARTMENT OF ELECTION
Article 12-A-1, shall pay an annual overpaid executive administrative office tax
if the Executive
Pay Ratio for the tax year of that person or the combined group of which it is
a part
exceeds 100:1. This overpaid executive administrative office tax shall be measured
by the
person''s total payroll expense, as defined in Section 953.8(f) of Article 12-A-1,
that is
attributable to the City. If a person is a member of a combined group, then its
tax shall be
measured by the total payroll expense of the combined group attributable to the
City. Such
person or combined group shall pay only the overpaid executive administrative
office tax, and
not the tax imposed under other subsections of this Section 3303, but a person
or combined
group may be liable for the administrative office tax imposed by Section 953.8
of
Article 12-A-1 and the homelessness administrative office tax imposed by Section
2804(d) or
Section 2804.9 of Article 28 in addition to the overpaid executive administrative
office tax
imposed by this subsection (dg). Unless specified otherwise, this overpaid executive
administrative office tax shall be considered part of the Overpaid Executive Gross
Receipts
Tax for all purposes.'
- source_sentence: What are the homelessness gross receipts tax rates applicable to
Category 1 Business Activities for tax years beginning on or after January 1,
2025, as outlined in Section 2804.3 of the same code?
sentences:
- 'For purposes
of serving a notice of a deficiency determination under this Section 6.11-2, the
date that a return was
due shall be:
(1) For tax years and tax periods ending on or before December 31, 2021, and
registration years ending on or before June 30, 2023, the due date of the return
without regard to any
extensions under Section 6.9-4;
(2) For tax years and tax periods beginning on or after January 1, 2022, and
registration years beginning on or after July 1, 2023 and ending on or before
March 31, 2026, except
for the taxes for the years and periods subject to subsection (a)(3) of this Section
6.11-2, the due date of
the return including any extensions; or
(3) For the Gross Receipts Tax (Article 12-A-1) (including the tax on administrative
office business activities in Section 953.8 of Article 12-A-1), the Early Care
and Education Commercial
Rents Tax (Article 21), the Homelessness Gross Receipts Tax (Article 28) (including
the tax on
administrative office business activities in Section 2804(d) or Section 2804.9
of Article 28), the
Cannabis Business Tax (Article 30), and the Overpaid Executive Gross Receipts
Tax (Article 33)
(including the tax on administrative office business activities in Section 3303(g))
for tax years
beginning on or after January 1, 2025, and for the Business Registration Fee (Article
12) for
registration years beginning on or after April 1, 2026, November 30 of the calendar
year in which the
return for such respective tax was originally due.
13SAN
CO
22 MAY (b) The notice of any determination under this Section 6.11-2 may be served
upon the
DEMANTHER OF CLIENT OFF
taxpayer or other affected person personally or by mail; if by mail, service shall
be:
(1) to the last known address that appears in the Tax Collector''s records,
provided there is such an address in the Tax Collector''s records, or
(2) to an address that the Tax Collector concludes, in the Tax Collector''s
discretion, is the last known address of the person(s).'
- 'If the person or combined group did not file a
7SANT
V
772 MAY 10 TH 1:28
return for the preceding tax year and did not request an extension under Section
6.9-4 to file that
return, the person or combined group shall be deemed to have filed a return showing
no
liability for purposes of this Section 6.9-Isubsection (c)(3)(C)(ii), and no estimated
tax payments
of that tax shall be due for the current tax year.
(D) Lessor of Residential Real Estate Exemption. Notwithstanding
any other provision in this Section 6.9-1(c)(3), a lessor of residential real
estate, as defined in
Section 954.1 of Article 12-A-1, shall not be required to make estimated tax payments
under
this Section 6.9-1(c)(3), but shall pay its full Gross Receipts Tax liability,
Homelessness Gross
Receipts Tax liability, and Overpaid Executive Gross Receipts Tax liability on
or before the
last day of February of the succeeding year, or the extended deadline in Section
6.9-4, if the
lessor''s gross receipts within the City shown on the lessor''s return for either
the current tax
year or the preceding tax year did not exceed the threshold in Section 954.1(b)
of
Article 12-A-1.
*
*
(g) Business Registration Certificate; Annual Due Date.
(1) For registration years ending on or before June 30, 2025 and the
2025-2026 registration year, eExcept for any extensions granted under Section
6.9-4 or as
otherwise provided in this Business and Tax Regulations Code, returns and payments
of the
Business Registration Fee (Article 12) shall be filed annually and are due and
payable, and
shall be delinquent if not submitted and paid to the Tax Collector, on or before
the last day of
May preceding the registration year commencing July 1 of that year.'
- '(a) The homelessness gross receipts tax rates applicable to Category 1 Business
Activities for
tax years beginning on or after January 1, 2025 are:
0% for taxable gross receipts between $0 and $1,000,000
0% for taxable gross receipts between $1,000,000.01 and $2,500,000
0% for taxable gross receipts between $2,500,000.01 and $25,000,000
0.164% for taxable gross receipts between $25,000,000.01 and $50,000,000
0.164% for taxable gross receipts between $50,000,000.01 and $75,000,000
0.164% for taxable gross receipts between $75,000,000.01 and $100,000,000
0.164% for taxable gross receipts between $100,000,000.01 and $150,000,000
0.246% for taxable gross receipts between $150,000,000.01 and $250,000,000
0.328% for taxable gross receipts between $250,000,000.01 and $500,000,000
0.41% for taxable gross receipts between $500,000,000.01 and $1,000,000,000
0.492% for taxable gross receipts over $1,000,000,000
(b) "Category 1 Business Activities" has the same meaning as in Section 953.20(b)
of
Article 12-A-1 of this Business and Tax Regulations Code.
(c) The amount of taxable gross receipts from Category 1 Business Activities subject
to the
homelessness gross receipts tax shall be three-quarters of the amount determined
under Section 956.1
of Article 12-A-1 plus one-quarter of the amount determined under Section 956.2
of Article 12-A-1.
84SAL
SEC. 2804.3. HOMELESSNESS GROSS RECEIPTS TAX APPLICABLE TO CATEGORY 2
BUSINESS ACTIVITIES.'
- source_sentence: What are the different gross receipts tax rates applicable to private
education and health services, administrative and support services, and miscellaneous
business activities as outlined in Section 953.4?
sentences:
- '(b) For only so long as and to the extent that the City is prohibited from imposing
the
Homelessness Gross Receipts Tax, any person upon whom the City is prohibited under
the
Constitution or laws of the State of California or the Constitution or laws of
the United States
from imposing the Homelessness Gross Receipts Tax shall be exempt from the
Homelessness Gross Receipts Tax.
(c) For purposes of this Article 28, gross receipts shall not include receipts
that are
excluded from gross receipts for purposes of the gross receipts tax imposed by
Article 12-A-1,
and also shall not include receipts subject to a gross receipts tax on commercial
rents
imposed as a result of a measure adopted by the voters of San Francisco in the
June 5, 2018
election.
(d) This Section 2805 shall apply to tax years ending on or before December 31,
2024.
SEC. 2805.3. EXEMPTIONS AND EXCLUSIONS.
(a) An organization that is exempt from income taxation by Chapter 4 (commencing
with
Section 23701) of Part 11 of Division 2 of the California Revenue and Taxation
Code or Subchapter F
(commencing with Section 501) of Chapter 1 of Subtitle A of the Internal Revenue
Code of 1986, as
amended, as qualified by Sections 502, 503, 504, and 508 of the Internal Revenue
Code of 1986, as
amended, shall be exempt from taxation under this Article 28, only so long as
those exemptions
continue to exist under state or federal law.
(b) For purposes of this Article 28, gross receipts as defined in Section 952.3
shall not include
receipts from business activities if, and only so long as and to the extent that,
the City is prohibited
93SA
01:32
from taxing such receipts under the Constitution or laws of the United States
or under the Constitution
or laws of the State of California.
(c) Rent Controlled Buildings Exclusion. A person subject to the homelessness
gross receipts
tax may exclude from gross receipts in any tax year 50% of the total amount received
from the rental of
real property to tenants in occupancy at any location in the City, which is subject
to limits on rent
increases pursuant to the Residential Rent Stabilization and Arbitration Ordinance,
Administrative
Code, Chapter 37, Section 37.1 et seq.'
- '10 that the
90% gross receipts threshold has been met for tax year 2023, and for tax year
2024 if the Controller
does not certify under Section 953.10 that the 95% gross receipts threshold has
been met for tax
year 2024:
42SAN LANCI Co
FILED
2024 MAY 10 PH 1:29
0.45% (e.g. $4.50 per $1,000) for taxable gross receipts between $0 and $1,000,000
DEPARTMENT OF ELECTIONS
0.488% (e.g. $4.88 per $1,000) for taxable gross receipts between $1,000,000.01
and
$2,500,000
0.488% (e.g. $4.88 per $1,000) for taxable gross receipts between $2,500,000.01
and
$25,000,000
0.6% (e.g. $6 per $1,000) for taxable gross receipts over $25,000,000
(D) For tax year 2024 if the Controller certifies under Section 953.10 that the
95% gross receipts threshold has been met for tax year 2024, and for tax years
beginning on or after
January 1, 2025:
0.465% (e.g. $4.65 per $1,000) for taxable gross receipts between $0 and $1,000,000
0.504% (e.g. $5.04 per $1,000) for taxable gross receipts between $1,000,000.01
and
$2,500,000
0.504% (e.g. $5.04 per $1,000) for taxable gross receipts between $2,500,000.01
and
$25,000,000
0.62% (e.g. $6.20 per $1,000) for taxable gross receipts over $25,000,000
SEC. 953.4. GROSS RECEIPTS TAX APPLICABLE TO PRIVATE EDUCATION AND
HEALTH SERVICES; ADMINISTRATIVE AND SUPPORT SERVICES; AND
MISCELLANEOUS BUSINESS ACTIVITIES.
(a) The gross receipts tax rates applicable to the business activities of private
education and health services, administrative and support services,'
- 'SARI
до
202 BAY 10 FA 1:32
year by 52, and dividing the result by the number of weeks that the employee was
employed by that
person or combined group during the tax year.
"Highest-Paid Managerial Employee" means the individual employee or officer of
a
person or combined group with managerial responsibility in a business function
who received the most
Compensation for a tax year. For purposes of determining the Highest-Paid Managerial
Employee and
the Compensation of such employee, Compensation shall not be annualized or converted
to a full-time
equivalency.
SEC. 3303. IMPOSITION OF TAX.
(a) Except as otherwise provided in this Article 33, commencing with tax years
beginning on or after January 1, 2022, for the privilege of engaging in business
in the City, the
City imposes an annual Overpaid Executive Gross Receipts Tax on each person engaging
in
business within the City where the Executive Pay Ratio for the tax year of that
person or the
combined group of which it is a part exceeds 100:1.'
- source_sentence: Based on the context information provided, what are the different
gross receipts tax rates for businesses in San Francisco for tax years 2022, 2023,
and 2024?
sentences:
- '$9.75 per $1,000) for taxable gross receipts over $25,000,000
44SANCO
2024 NAY LO
(D) For tax year 2024 if the Controller certifies under Section 953.10 that the
DEPARTMENT OF
95% gross receipts threshold has been met for tax year 2024, and for tax years
beginning on or after
January 1, 2025:
0.814% (e.g. $8.14 per $1,000) for taxable gross receipts between $0 and $1,000,000
0.853% (e.g. $8.53 per $1,000) for taxable gross receipts between $1,000,000.01
and
$2,500,000
0.93% (e.g. $9.30 per $1,000) for taxable gross receipts between $2,500,000.01
and
$25,000,000
1.008% (e.g. $10.08 per $1,000) for taxable gross receipts over $25,000,000
(3) For all business activities not otherwise exempt and not elsewhere
subjected to a gross receipts tax rate or an administrative office tax by this
Article 12-A-1:
(B) For tax years 2022 and, if the Controller does not certify under
Section 953.10 that the 90% gross receipts threshold has been met for tax year
2023, for tax
year 2023:
0.788% (e.g. $7.88 per $1,000) for taxable gross receipts between $0 and $1,000,000
0.825% (e.g. $8.25 per $1,000) for taxable gross receipts between $1,000,000.01
and
$2,500,000
0.9% (e.g. $9 per $1,000) for taxable gross receipts between $2,500,000.01 and
$25,000,000
0.975% (e.g. $9.75 per $1,000) for taxable gross receipts over $25,000,000
(C) For tax year 2023 if the Controller certifies under Section 953.10 that the
90% gross receipts threshold has been met for tax year 2023,'
- '(d) In no event shall the credit under this Section 960.4 reduce a person or
combined group''s
Gross Receipts Tax liability to less than $0 for any tax year. The credit under
this Section shall not be
refundable and may not be carried forward to a subsequent year.
SEC. 966. CONTROLLER REPORTS.
The Controller shall prepare reports by September 1, 2026, and September 1, 2027,
respectively, that discuss current economic conditions in the City and the performance
of the tax system
revised by the voters in the ordinance adding this Section 966.
Section 6. Article 21 of the Business and Tax Regulations Code is hereby amended
by
revising Section 2106 to read as follows:
SEC. 2106. SMALL BUSINESS EXEMPTION.
(a) For tax years ending on or before December 31, 2024, nNotwithstanding any
other
provision of this Article 21, a person or combined group exempt from payment of
the gross
receipts tax under Section 954.1 of Article 12-A-1, as amended from time to time,
shall also
be exempt from payment of the Early Care and Education Commercial Rents Tax.
79SAN
DL W(b) For tax years beginning on or after January 1, 2025, notwithstanding any
other provision
of this Article 21, a "small business enterprise" shall be exempt from payment
of the Early Care and
Education Commercial Rents Tax. For purposes of this subsection (b), the term
"small business
enterprise" shall mean any person or combined group whose gross receipts within
the City, determined
under Article 12-A-1, did not exceed $2,325,000, adjusted annually in accordance
with the increase in
the Consumer Price Index: All Urban Consumers for the San Francisco/Oakland/Hayward
Area for All
Items as reported by the United States Bureau of Labor Statistics, or any successor
to that index, as of
December 31 of the calendar year two years prior to the tax year, beginning with
tax year 2026, and
rounded to the nearest $10,000. This subsection (b) shall not apply to a person
or combined group
subject to a tax on administrative office business activities in Section 953.8
of Article 12-A-1.
Section 7.'
- '(1) For tax years ending on or before December 31, 2024, the overpaid executive
administrative office tax shall be calculated as follows:
(+4) 0.4% of the person or combined group''s total payroll expense
attributable to the City for a tax year if the person or combined group has an
Executive Pay
Ratio for that tax year of greater than 100:1, but less than or equal to 200:1;
(2B) 0.8% of the person or combined group''s total payroll expense
attributable to the City for a tax year if the person or combined group has an
Executive Pay
Ratio for that tax year of greater than 200:1, but less than or equal to 300:1;
(3C) 1.2% of the person or combined group''s total payroll expense
attributable to the City for a tax year if the person or combined group has an
Executive Pay
Ratio for that tax year of greater than 300:1, but less than or equal to 400:1;
103SAULAMIC 50
10 PM 1:32
(4D) 1.6% of the person or combined group''s total payroll expense
DEPARTMENT OF ELECTIONS
attributable to the City for a tax year if the person or combined group has an
Executive Pay
Ratio for that tax year of greater than 400:1, but less than or equal to 500:1;
(5E) 2% of the person or combined group''s total payroll expense
attributable to the City for a tax year if the person or combined group has an
Executive Pay
Ratio for that tax year of greater than 500:1, but less than or equal to 600:1;
or
(6F) 2.4% of the person or combined group''s total payroll expense
attributable to the City for a tax year if the person or combined group has an
Executive Pay
Ratio for that tax year of greater than 600:1.'
model-index:
- name: SentenceTransformer based on microsoft/mpnet-base
results:
- task:
type: knowledge-distillation
name: Knowledge Distillation
dataset:
name: stsb dev
type: stsb-dev
metrics:
- type: negative_mse
value: -2.4281997233629227
name: Negative Mse
---
# SentenceTransformer based on microsoft/mpnet-base
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base) <!-- at revision 6996ce1e91bd2a9c7d7f61daec37463394f73f09 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **Similarity Function:** Cosine Similarity
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: MPNetModel
(1): Pooling({'word_embedding_dimension': 768, '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})
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("Areeb-02/mpnet-base-GISTEmbedLoss-MSEE_Evaluator-salestax-docs")
# Run inference
sentences = [
'Based on the context information provided, what are the different gross receipts tax rates for businesses in San Francisco for tax years 2022, 2023, and 2024?',
'$9.75 per $1,000) for taxable gross receipts over $25,000,000\n44SANCO\n2024 NAY LO\n(D) For tax year 2024 if the Controller certifies under Section 953.10 that the\nDEPARTMENT OF\n95% gross receipts threshold has been met for tax year 2024, and for tax years beginning on or after\nJanuary 1, 2025:\n0.814% (e.g. $8.14 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.853% (e.g. $8.53 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.93% (e.g. $9.30 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n1.008% (e.g. $10.08 per $1,000) for taxable gross receipts over $25,000,000\n(3) For all business activities not otherwise exempt and not elsewhere\nsubjected to a gross receipts tax rate or an administrative office tax by this Article 12-A-1:\n(B) For tax years 2022 and, if the Controller does not certify under\nSection 953.10 that the 90% gross receipts threshold has been met for tax year 2023, for tax\nyear 2023:\n0.788% (e.g. $7.88 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.825% (e.g. $8.25 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.9% (e.g. $9 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n0.975% (e.g. $9.75 per $1,000) for taxable gross receipts over $25,000,000\n(C) For tax year 2023 if the Controller certifies under Section 953.10 that the\n90% gross receipts threshold has been met for tax year 2023,',
'(d) In no event shall the credit under this Section 960.4 reduce a person or combined group\'s\nGross Receipts Tax liability to less than $0 for any tax year. The credit under this Section shall not be\nrefundable and may not be carried forward to a subsequent year.\nSEC. 966. CONTROLLER REPORTS.\nThe Controller shall prepare reports by September 1, 2026, and September 1, 2027,\nrespectively, that discuss current economic conditions in the City and the performance of the tax system\nrevised by the voters in the ordinance adding this Section 966.\nSection 6. Article 21 of the Business and Tax Regulations Code is hereby amended by\nrevising Section 2106 to read as follows:\nSEC. 2106. SMALL BUSINESS EXEMPTION.\n(a) For tax years ending on or before December 31, 2024, nNotwithstanding any other\nprovision of this Article 21, a person or combined group exempt from payment of the gross\nreceipts tax under Section 954.1 of Article 12-A-1, as amended from time to time, shall also\nbe exempt from payment of the Early Care and Education Commercial Rents Tax.\n79SAN\nDL W(b) For tax years beginning on or after January 1, 2025, notwithstanding any other provision\nof this Article 21, a "small business enterprise" shall be exempt from payment of the Early Care and\nEducation Commercial Rents Tax. For purposes of this subsection (b), the term "small business\nenterprise" shall mean any person or combined group whose gross receipts within the City, determined\nunder Article 12-A-1, did not exceed $2,325,000, adjusted annually in accordance with the increase in\nthe Consumer Price Index: All Urban Consumers for the San Francisco/Oakland/Hayward Area for All\nItems as reported by the United States Bureau of Labor Statistics, or any successor to that index, as of\nDecember 31 of the calendar year two years prior to the tax year, beginning with tax year 2026, and\nrounded to the nearest $10,000. This subsection (b) shall not apply to a person or combined group\nsubject to a tax on administrative office business activities in Section 953.8 of Article 12-A-1.\nSection 7.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Knowledge Distillation
* Dataset: `stsb-dev`
* Evaluated with [<code>MSEEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.MSEEvaluator)
| Metric | Value |
|:-----------------|:------------|
| **negative_mse** | **-2.4282** |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 238 training samples
* Columns: <code>sentence1</code> and <code>sentence2</code>
* Approximate statistics based on the first 1000 samples:
| | sentence1 | sentence2 |
|:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 5 tokens</li><li>mean: 41.95 tokens</li><li>max: 219 tokens</li></ul> | <ul><li>min: 63 tokens</li><li>mean: 426.3 tokens</li><li>max: 512 tokens</li></ul> |
* Samples:
| sentence1 | sentence2 |
|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>What types of businesses are subject to the gross receipts tax in San Francisco, and how is their San Francisco gross receipts calculated? What are the current rates for this tax, and are there any exemptions or scheduled increases?</code> | <code>The Way It Is Now<br>CHANGES TO BUSINESS TAXES<br>The City collects various business taxes on an annual basis including:<br>O<br>•<br>SAN FRANCISCO<br>FILED<br>2024 MAY 15 PM 3:10<br>DEPARTMENT OF ELECTIONS<br>A gross receipts tax that is a percentage of a business's San Francisco gross receipts.<br>Depending on business type, the City determines a business's San Francisco gross<br>receipts based on sales in San Francisco, payroll expenses for employees working there,<br>or both. Rates range from 0.053% to 1.008% and are scheduled to increase in coming<br>years. Rates depend on business type, and higher rates apply as a business generates<br>more gross receipts. For 2023, most businesses with gross receipts up to $2.19 million<br>are exempt.<br>A homelessness gross receipts tax that is an additional tax on businesses with San<br>Francisco gross receipts over $50 million. Rates range from 0.175% to 0.69%.<br>An overpaid executive gross receipts tax that is an additional tax on businesses that pay<br>their highest-paid managerial employee much higher than the median compensation they<br>pay their San Francisco employees. Rates are between 0.1% and 0.6%.<br>A business registration fee that is an additional tax. For most businesses the fee is<br>currently between $47 and $45,150, based on business type and amount of gross receipts.<br>• An administrative office tax on payroll expenses that certain large businesses pay instead<br>of these other business taxes. The combined rates in 2024 range from 3.04% to 5.44%,<br>and in 2025 are scheduled to range from 3.11% to 5.51%. Business registration fees for<br>these businesses currently range from $19,682 to $45,928.<br>State law limits the total revenue, including tax revenue, the City may spend each year. The<br>voters may approve increases to this limit for up to four years.</code> |
| <code>What is the homelessness gross receipts tax, and which businesses are required to pay it? What are the current rates for this tax, and how do they vary based on the amount of San Francisco gross receipts? Are there any exemptions or scheduled increases for this tax?</code> | <code>The Way It Is Now<br>CHANGES TO BUSINESS TAXES<br>The City collects various business taxes on an annual basis including:<br>O<br>•<br>SAN FRANCISCO<br>FILED<br>2024 MAY 15 PM 3:10<br>DEPARTMENT OF ELECTIONS<br>A gross receipts tax that is a percentage of a business's San Francisco gross receipts.<br>Depending on business type, the City determines a business's San Francisco gross<br>receipts based on sales in San Francisco, payroll expenses for employees working there,<br>or both. Rates range from 0.053% to 1.008% and are scheduled to increase in coming<br>years. Rates depend on business type, and higher rates apply as a business generates<br>more gross receipts. For 2023, most businesses with gross receipts up to $2.19 million<br>are exempt.<br>A homelessness gross receipts tax that is an additional tax on businesses with San<br>Francisco gross receipts over $50 million. Rates range from 0.175% to 0.69%.<br>An overpaid executive gross receipts tax that is an additional tax on businesses that pay<br>their highest-paid managerial employee much higher than the median compensation they<br>pay their San Francisco employees. Rates are between 0.1% and 0.6%.<br>A business registration fee that is an additional tax. For most businesses the fee is<br>currently between $47 and $45,150, based on business type and amount of gross receipts.<br>• An administrative office tax on payroll expenses that certain large businesses pay instead<br>of these other business taxes. The combined rates in 2024 range from 3.04% to 5.44%,<br>and in 2025 are scheduled to range from 3.11% to 5.51%. Business registration fees for<br>these businesses currently range from $19,682 to $45,928.<br>State law limits the total revenue, including tax revenue, the City may spend each year. The<br>voters may approve increases to this limit for up to four years.</code> |
| <code>What is the proposed measure that voters may approve to change the City's business taxes in San Francisco?</code> | <code>The<br>voters may approve increases to this limit for up to four years.<br>The Proposal<br>The proposed measure would change the City's business taxes to:<br>•<br>For the gross receipts tax:<br>○ recategorize business types, reducing the number from 14 to seven;<br>determine San Francisco gross receipts for some businesses based less on payroll<br>expenses and more on sales;<br>o change rates to between 0.1% and 3.716%; and<br>exempt most businesses with gross receipts up to $5 million (increased by<br>inflation).<br>Apply the homelessness gross receipts tax on business activities with San Francisco gross<br>receipts over $25 million, at rates between 0.162% and 1.64%.<br>Modify how the City calculates the overpaid executive gross receipts tax and who pays<br>that tax, and set rates between 0.02% and 0.129%.<br>Adjust business registration fees to between $55 and $60,000 (increased by inflation).Adjust the administrative office tax rates for certain large businesses to range from 2.97%<br>to 3.694%, and the business registration fees for these taxpayers to between $500 and<br>$35,000 (increased by inflation).<br>Make administrative and other changes to the City's business taxes.<br>The homelessness gross receipts tax would continue to fund services for people experiencing<br>homelessness and homelessness prevention. The City would use the other taxes for general<br>government purposes.<br>All these taxes would apply indefinitely until repealed.<br>This proposal would increase the City's spending limit for four years.SALITA CO<br>2024 MAY 10 PH 1:27<br>DEPARTMENT OF ELECTI<br>"Local Small Business Tax Cut Ordinance"<br>Be it ordained by the People of the City and County of San Francisco:<br>NOTE:<br>Unchanged Code text and uncodified text are in plain font.<br>Additions to Codes are in single-underline italics Times New Roman font.<br>Deletions to Codes are in strikethrough italics Times New Roman font.<br>Asterisks (* * * *) indicate the omission of unchanged Code<br>subsections or parts of tables.<br>Section 1. Title. This initiative is known and may be referred to as the "Local Small<br>Business Tax Cut Ordinance."<br>Section 2. Article 2 of the Business and Tax Regulations Code is hereby amended by<br>revising Section 76.3 to read as follows:<br>SEC. 76.3.</code> |
* Loss: [<code>GISTEmbedLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#gistembedloss) with these parameters:
```json
{'guide': SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': True}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
(2): Normalize()
), 'temperature': 0.01}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `num_train_epochs`: 1
- `warmup_ratio`: 0.1
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `learning_rate`: 5e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 1
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: False
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | stsb-dev_negative_mse |
|:-----:|:----:|:---------------------:|
| 0 | 0 | -2.4282 |
### Framework Versions
- Python: 3.10.12
- Sentence Transformers: 3.0.1
- Transformers: 4.41.2
- PyTorch: 2.3.0+cu121
- Accelerate: 0.31.0
- Datasets: 2.20.0
- Tokenizers: 0.19.1
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### GISTEmbedLoss
```bibtex
@misc{solatorio2024gistembed,
title={GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning},
author={Aivin V. Solatorio},
year={2024},
eprint={2402.16829},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SentenceTransformer based on microsoft/mpnet-base
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [microsoft/mpnet-base](https://huggingface.co/microsoft/mpnet-base) <!-- at revision 6996ce1e91bd2a9c7d7f61daec37463394f73f09 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **Similarity Function:** Cosine Similarity
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: MPNetModel
(1): Pooling({'word_embedding_dimension': 768, '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})
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("Areeb-02/mpnet-base-GISTEmbedLoss-MSEE_Evaluator-salestax-docs")
# Run inference
sentences = [
'Based on the context information provided, what are the different gross receipts tax rates for businesses in San Francisco for tax years 2022, 2023, and 2024?',
'$9.75 per $1,000) for taxable gross receipts over $25,000,000\n44SANCO\n2024 NAY LO\n(D) For tax year 2024 if the Controller certifies under Section 953.10 that the\nDEPARTMENT OF\n95% gross receipts threshold has been met for tax year 2024, and for tax years beginning on or after\nJanuary 1, 2025:\n0.814% (e.g. $8.14 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.853% (e.g. $8.53 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.93% (e.g. $9.30 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n1.008% (e.g. $10.08 per $1,000) for taxable gross receipts over $25,000,000\n(3) For all business activities not otherwise exempt and not elsewhere\nsubjected to a gross receipts tax rate or an administrative office tax by this Article 12-A-1:\n(B) For tax years 2022 and, if the Controller does not certify under\nSection 953.10 that the 90% gross receipts threshold has been met for tax year 2023, for tax\nyear 2023:\n0.788% (e.g. $7.88 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.825% (e.g. $8.25 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.9% (e.g. $9 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n0.975% (e.g. $9.75 per $1,000) for taxable gross receipts over $25,000,000\n(C) For tax year 2023 if the Controller certifies under Section 953.10 that the\n90% gross receipts threshold has been met for tax year 2023,',
'(d) In no event shall the credit under this Section 960.4 reduce a person or combined group\'s\nGross Receipts Tax liability to less than $0 for any tax year. The credit under this Section shall not be\nrefundable and may not be carried forward to a subsequent year.\nSEC. 966. CONTROLLER REPORTS.\nThe Controller shall prepare reports by September 1, 2026, and September 1, 2027,\nrespectively, that discuss current economic conditions in the City and the performance of the tax system\nrevised by the voters in the ordinance adding this Section 966.\nSection 6. Article 21 of the Business and Tax Regulations Code is hereby amended by\nrevising Section 2106 to read as follows:\nSEC. 2106. SMALL BUSINESS EXEMPTION.\n(a) For tax years ending on or before December 31, 2024, nNotwithstanding any other\nprovision of this Article 21, a person or combined group exempt from payment of the gross\nreceipts tax under Section 954.1 of Article 12-A-1, as amended from time to time, shall also\nbe exempt from payment of the Early Care and Education Commercial Rents Tax.\n79SAN\nDL W(b) For tax years beginning on or after January 1, 2025, notwithstanding any other provision\nof this Article 21, a "small business enterprise" shall be exempt from payment of the Early Care and\nEducation Commercial Rents Tax. For purposes of this subsection (b), the term "small business\nenterprise" shall mean any person or combined group whose gross receipts within the City, determined\nunder Article 12-A-1, did not exceed $2,325,000, adjusted annually in accordance with the increase in\nthe Consumer Price Index: All Urban Consumers for the San Francisco/Oakland/Hayward Area for All\nItems as reported by the United States Bureau of Labor Statistics, or any successor to that index, as of\nDecember 31 of the calendar year two years prior to the tax year, beginning with tax year 2026, and\nrounded to the nearest $10,000. This subsection (b) shall not apply to a person or combined group\nsubject to a tax on administrative office business activities in Section 953.8 of Article 12-A-1.\nSection 7.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Knowledge Distillation
* Dataset: `stsb-dev`
* Evaluated with [<code>MSEEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.MSEEvaluator)
| Metric | Value |
|:-----------------|:------------|
| **negative_mse** | **-2.4282** |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 238 training samples
* Columns: <code>sentence1</code> and <code>sentence2</code>
* Approximate statistics based on the first 1000 samples:
| | sentence1 | sentence2 |
|:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 5 tokens</li><li>mean: 41.95 tokens</li><li>max: 219 tokens</li></ul> | <ul><li>min: 63 tokens</li><li>mean: 426.3 tokens</li><li>max: 512 tokens</li></ul> |
* Samples:
| sentence1 | sentence2 |
|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>What types of businesses are subject to the gross receipts tax in San Francisco, and how is their San Francisco gross receipts calculated? What are the current rates for this tax, and are there any exemptions or scheduled increases?</code> | <code>The Way It Is Now<br>CHANGES TO BUSINESS TAXES<br>The City collects various business taxes on an annual basis including:<br>O<br>•<br>SAN FRANCISCO<br>FILED<br>2024 MAY 15 PM 3:10<br>DEPARTMENT OF ELECTIONS<br>A gross receipts tax that is a percentage of a business's San Francisco gross receipts.<br>Depending on business type, the City determines a business's San Francisco gross<br>receipts based on sales in San Francisco, payroll expenses for employees working there,<br>or both. Rates range from 0.053% to 1.008% and are scheduled to increase in coming<br>years. Rates depend on business type, and higher rates apply as a business generates<br>more gross receipts. For 2023, most businesses with gross receipts up to $2.19 million<br>are exempt.<br>A homelessness gross receipts tax that is an additional tax on businesses with San<br>Francisco gross receipts over $50 million. Rates range from 0.175% to 0.69%.<br>An overpaid executive gross receipts tax that is an additional tax on businesses that pay<br>their highest-paid managerial employee much higher than the median compensation they<br>pay their San Francisco employees. Rates are between 0.1% and 0.6%.<br>A business registration fee that is an additional tax. For most businesses the fee is<br>currently between $47 and $45,150, based on business type and amount of gross receipts.<br>• An administrative office tax on payroll expenses that certain large businesses pay instead<br>of these other business taxes. The combined rates in 2024 range from 3.04% to 5.44%,<br>and in 2025 are scheduled to range from 3.11% to 5.51%. Business registration fees for<br>these businesses currently range from $19,682 to $45,928.<br>State law limits the total revenue, including tax revenue, the City may spend each year. The<br>voters may approve increases to this limit for up to four years.</code> |
| <code>What is the homelessness gross receipts tax, and which businesses are required to pay it? What are the current rates for this tax, and how do they vary based on the amount of San Francisco gross receipts? Are there any exemptions or scheduled increases for this tax?</code> | <code>The Way It Is Now<br>CHANGES TO BUSINESS TAXES<br>The City collects various business taxes on an annual basis including:<br>O<br>•<br>SAN FRANCISCO<br>FILED<br>2024 MAY 15 PM 3:10<br>DEPARTMENT OF ELECTIONS<br>A gross receipts tax that is a percentage of a business's San Francisco gross receipts.<br>Depending on business type, the City determines a business's San Francisco gross<br>receipts based on sales in San Francisco, payroll expenses for employees working there,<br>or both. Rates range from 0.053% to 1.008% and are scheduled to increase in coming<br>years. Rates depend on business type, and higher rates apply as a business generates<br>more gross receipts. For 2023, most businesses with gross receipts up to $2.19 million<br>are exempt.<br>A homelessness gross receipts tax that is an additional tax on businesses with San<br>Francisco gross receipts over $50 million. Rates range from 0.175% to 0.69%.<br>An overpaid executive gross receipts tax that is an additional tax on businesses that pay<br>their highest-paid managerial employee much higher than the median compensation they<br>pay their San Francisco employees. Rates are between 0.1% and 0.6%.<br>A business registration fee that is an additional tax. For most businesses the fee is<br>currently between $47 and $45,150, based on business type and amount of gross receipts.<br>• An administrative office tax on payroll expenses that certain large businesses pay instead<br>of these other business taxes. The combined rates in 2024 range from 3.04% to 5.44%,<br>and in 2025 are scheduled to range from 3.11% to 5.51%. Business registration fees for<br>these businesses currently range from $19,682 to $45,928.<br>State law limits the total revenue, including tax revenue, the City may spend each year. The<br>voters may approve increases to this limit for up to four years.</code> |
| <code>What is the proposed measure that voters may approve to change the City's business taxes in San Francisco?</code> | <code>The<br>voters may approve increases to this limit for up to four years.<br>The Proposal<br>The proposed measure would change the City's business taxes to:<br>•<br>For the gross receipts tax:<br>○ recategorize business types, reducing the number from 14 to seven;<br>determine San Francisco gross receipts for some businesses based less on payroll<br>expenses and more on sales;<br>o change rates to between 0.1% and 3.716%; and<br>exempt most businesses with gross receipts up to $5 million (increased by<br>inflation).<br>Apply the homelessness gross receipts tax on business activities with San Francisco gross<br>receipts over $25 million, at rates between 0.162% and 1.64%.<br>Modify how the City calculates the overpaid executive gross receipts tax and who pays<br>that tax, and set rates between 0.02% and 0.129%.<br>Adjust business registration fees to between $55 and $60,000 (increased by inflation).Adjust the administrative office tax rates for certain large businesses to range from 2.97%<br>to 3.694%, and the business registration fees for these taxpayers to between $500 and<br>$35,000 (increased by inflation).<br>Make administrative and other changes to the City's business taxes.<br>The homelessness gross receipts tax would continue to fund services for people experiencing<br>homelessness and homelessness prevention. The City would use the other taxes for general<br>government purposes.<br>All these taxes would apply indefinitely until repealed.<br>This proposal would increase the City's spending limit for four years.SALITA CO<br>2024 MAY 10 PH 1:27<br>DEPARTMENT OF ELECTI<br>"Local Small Business Tax Cut Ordinance"<br>Be it ordained by the People of the City and County of San Francisco:<br>NOTE:<br>Unchanged Code text and uncodified text are in plain font.<br>Additions to Codes are in single-underline italics Times New Roman font.<br>Deletions to Codes are in strikethrough italics Times New Roman font.<br>Asterisks (* * * *) indicate the omission of unchanged Code<br>subsections or parts of tables.<br>Section 1. Title. This initiative is known and may be referred to as the "Local Small<br>Business Tax Cut Ordinance."<br>Section 2. Article 2 of the Business and Tax Regulations Code is hereby amended by<br>revising Section 76.3 to read as follows:<br>SEC. 76.3.</code> |
* Loss: [<code>GISTEmbedLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#gistembedloss) with these parameters:
```json
{'guide': SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': True}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
(2): Normalize()
), 'temperature': 0.01}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `num_train_epochs`: 1
- `warmup_ratio`: 0.1
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `learning_rate`: 5e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 1
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: False
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | stsb-dev_negative_mse |
|:-----:|:----:|:---------------------:|
| 0 | 0 | -2.4282 |
### Framework Versions
- Python: 3.10.12
- Sentence Transformers: 3.0.1
- Transformers: 4.41.2
- PyTorch: 2.3.0+cu121
- Accelerate: 0.31.0
- Datasets: 2.20.0
- Tokenizers: 0.19.1
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### GISTEmbedLoss
```bibtex
@misc{solatorio2024gistembed,
title={GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning},
author={Aivin V. Solatorio},
year={2024},
eprint={2402.16829},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "microsoft/mpnet-base", "datasets": [], "language": [], "library_name": "sentence-transformers", "metrics": ["negative_mse"], "pipeline_tag": "sentence-similarity", "tags": ["sentence-transformers", "sentence-similarity", "feature-extraction", "generated_from_trainer", "dataset_size:238", "loss:GISTEmbedLoss"], "widget": [{"source_sentence": "What is the timeframe for opening a physical location in the Designated Areas to be eligible for the credit under Section 960.1?", "sentences": ["(a) A person or combined group that opens a physical location in the Designated\nAreas on or after January 1, 2023 through and including December 31, 2027, shall be allowed\na credit against that person or combined group's Gross Receipts Tax if the person or\n74SAN FIAT CU\n2024 11AY 10 PM 1331\ncombined group did not have a physical location in the City for at least three years prior to\nDEPARTMENT OF ELECTIONS\nopening the physical location. The credit under this Section 960.1 shall be an annual credit\nfor each of up to three tax years immediately following the tax year in which the person or\ncombined group opened the physical location in the Designated Areas, provided the person or\ncombined group maintains a physical location in the Designated Areas in the tax year that the\ncredit is taken. To be eligible for the credit, the person or combined group must take the\ncredit for each tax year on an original Gross Receipts Tax return filed with the Tax Collector.", "5222 (Nondepository Credit Intermediation), 5223 (Activities Related\nto Credit Intermediation)), 533 (Lessors of Nonfinancial Intangible Assets (except Copyrighted\nWorks)), 54 (Professional, Scientific, and Technical Services) but not including 541714 (Research and\nDevelopment in Biotechnology (except Nanobiotechnology)), 55 (Management of Companies and\n63SA\nCL\n2024 LAY 10 PA (30\nEnterprises), 562 (Waste Management and Remediation Services), 61 (Educational Services),\nDEPARTMENT OF ELECTI\n62 (Health Care and Social Assistance), and all business activities not otherwise exempt and not\nelsewhere subjected to a gross receipts tax rate by Sections 953.20 through 953.26 or an administrative\noffice tax under Section 953.8.\n(c) The amount of taxable gross receipts from Category 5 Business Activities subject to the\ngross receipts tax shall be three-quarters of the amount determined under Section 956.1 plus\none-quarter of the amount determined under Section 956.2.\nSEC. 953.25. GROSS RECEIPTS TAX APPLICABLE TO CATEGORY 6 BUSINESS\nACTIVITIES.\n(a) The gross receipts tax rates applicable to Category 6 Business Activities are:\n(1) For tax years 2025 and 2026:\n1.5% for taxable gross receipts between $0 and $1,000,000\n1.5% for taxable gross receipts between $1,000,000.01 and $2,500,000\n3% for taxable gross receipts between $2,500,000.01 and $25,000,000\n2.352% for taxable gross receipts between $25,000,000.01 and $50,000,000\n3.024% for taxable gross receipts between $50,000,000.01 and $75,000,000\n3.024% for taxable gross receipts between $75,000,000.01 and $100,000,000\n3.36% for taxable gross receipts between $100,000,000.01 and $150,000,000\n3.36% for taxable gross receipts between $150,000,000.01 and $250,000,000\n3.36% for taxable gross receipts between $250,000,000.01 and $500,000,000\n3.", "Except as otherwise provided in this Section 855\nand Section 856 of this Article 12, and Section 76.3 of Article 2 of this Business and Tax Regulations\nCode, the annual fee for obtaining a registration certificate for the 2025-2026 registration year,\npayable in advance, shall be as follows:\nSan Francisco Gross Receipts for the\nImmediately Preceding Tax Year\n$0 to $100,000\n$100,000.01 to $250,000\nAnnual Registration Fee\n$41\n$71\n$250,000.01 to $500,000\n$120\n$500,000.01 to $750,000\n$240\n$750,000.01 to $1,000,000\n$330\n$1,000,000.01 to $1,500,000\n$469\n$1,500,000.01 to $2,000,000\n$656\n$2,000,000.01 to $2,500,000\n$848\n$2,500,000.01 to $5,000,000\n$1,414\n$5,000,000.01 to $7,500,000\n$600\n$7,500,000.01 to $15,000,000\n$1,500\n$15,000,000.01 to $25,000,000\n$4,875\n25$25,000,000.01 to $50,000,000\n$50,000,000.01 to $100,000,000\n$100,000,000.01 to $200,000,000\n$200,000,000.01 and over\nSAN KAL CO\n2724 11 10 11 129\n$15,000\nDEPARTMENT\n$30,000\n$37,500\n$45,000\n(d) Fee for Registration Years Beginning On or After April 1, 2026. Except as otherwise\nprovided in this Section 855 and Section 856 of this Article 12, and Section 76.3 of Article 2 of this\nBusiness and Tax Regulations Code, the annual fee for obtaining a registration certificate for\nregistration years beginning on or after April 1, 2026, payable in advance, shall be as follows:\nSan Francisco Gross Receipts for the\nImmediately Preceding Tax Year\n$0 to $100,"]}, {"source_sentence": "How is the overpaid executive administrative office tax measured for individuals or combined groups, and what other administrative office taxes might they also be liable for?", "sentences": ["(b) Combined gross receipts are the total worldwide gross receipts of the person and\nall related entities to the person, unless the election provided for in California Revenue and\nTaxation Code Section 25110 is in effect for the person, in which case combined gross\nreceipts shall be computed consistently with the water's edge election, as set forth therein.\n(c) Combined payroll is the total worldwide compensation paid by the person and all\nrelated entities to the person, unless the election provided for in California Revenue and\nTaxation Code Section 25110 is in effect for the person, in which case combined payroll shall\nbe computed consistently with the water's edge election, as set forth therein. A person who\n72SALA 50\nhas no combined payroll in a tax year shall have no gross receipts under this Section for that\nDEPARTMENT OF LECT\ntax year.\n(d) Payroll in the City is the total amount paid for compensation in the City by the\nperson and by all related entities to the person.\n(e) Compensation paid in the City shall be determined as follows:set forth in\nSection 953.8)(4).\n(1) Where compensation is paid by reason of work performed or services rendered by\nan individual wholly within the City, all of the compensation for such individual shall be attributable to\nthe City.\n(2) Where compensation is paid by reason of work performed or services rendered by\nan individual partly within and partly without the City, the portion of such compensation attributable to\nthe City shall be determined as follows:\n(A) Except as otherwise provided in this Section 956.2(e), the portion of such\ncompensation attributable to the City shall be the portion of such compensation which the total number\nof working hours employed within the City bears to the total number of working hours within and\nwithout the City.\n(B) If the amount of such compensation depends on the volume of business\ntransacted by such individual, then the portion of such compensation attributable to the City shall be\nthe portion of such compensation which the volume of business transacted by such individual in the\nCity bears to the volume of business transacted by such individual within and without the City.", "If the person or combined group did not file a\n7SANT\nV\n772 MAY 10 TH 1:28\nreturn for the preceding tax year and did not request an extension under Section 6.9-4 to file that\nreturn, the person or combined group shall be deemed to have filed a return showing no\nliability for purposes of this Section 6.9-Isubsection (c)(3)(C)(ii), and no estimated tax payments\nof that tax shall be due for the current tax year.\n(D) Lessor of Residential Real Estate Exemption. Notwithstanding\nany other provision in this Section 6.9-1(c)(3), a lessor of residential real estate, as defined in\nSection 954.1 of Article 12-A-1, shall not be required to make estimated tax payments under\nthis Section 6.9-1(c)(3), but shall pay its full Gross Receipts Tax liability, Homelessness Gross\nReceipts Tax liability, and Overpaid Executive Gross Receipts Tax liability on or before the\nlast day of February of the succeeding year, or the extended deadline in Section 6.9-4, if the\nlessor's gross receipts within the City shown on the lessor's return for either the current tax\nyear or the preceding tax year did not exceed the threshold in Section 954.1(b) of\nArticle 12-A-1.\n*\n*\n(g) Business Registration Certificate; Annual Due Date.\n(1) For registration years ending on or before June 30, 2025 and the\n2025-2026 registration year, eExcept for any extensions granted under Section 6.9-4 or as\notherwise provided in this Business and Tax Regulations Code, returns and payments of the\nBusiness Registration Fee (Article 12) shall be filed annually and are due and payable, and\nshall be delinquent if not submitted and paid to the Tax Collector, on or before the last day of\nMay preceding the registration year commencing July 1 of that year.", "DEPARTMENT OF ELECTION\nArticle 12-A-1, shall pay an annual overpaid executive administrative office tax if the Executive\nPay Ratio for the tax year of that person or the combined group of which it is a part\nexceeds 100:1. This overpaid executive administrative office tax shall be measured by the\nperson's total payroll expense, as defined in Section 953.8(f) of Article 12-A-1, that is\nattributable to the City. If a person is a member of a combined group, then its tax shall be\nmeasured by the total payroll expense of the combined group attributable to the City. Such\nperson or combined group shall pay only the overpaid executive administrative office tax, and\nnot the tax imposed under other subsections of this Section 3303, but a person or combined\ngroup may be liable for the administrative office tax imposed by Section 953.8 of\nArticle 12-A-1 and the homelessness administrative office tax imposed by Section 2804(d) or\nSection 2804.9 of Article 28 in addition to the overpaid executive administrative office tax\nimposed by this subsection (dg). Unless specified otherwise, this overpaid executive\nadministrative office tax shall be considered part of the Overpaid Executive Gross Receipts\nTax for all purposes."]}, {"source_sentence": "What are the homelessness gross receipts tax rates applicable to Category 1 Business Activities for tax years beginning on or after January 1, 2025, as outlined in Section 2804.3 of the same code?", "sentences": ["For purposes\nof serving a notice of a deficiency determination under this Section 6.11-2, the date that a return was\ndue shall be:\n(1) For tax years and tax periods ending on or before December 31, 2021, and\nregistration years ending on or before June 30, 2023, the due date of the return without regard to any\nextensions under Section 6.9-4;\n(2) For tax years and tax periods beginning on or after January 1, 2022, and\nregistration years beginning on or after July 1, 2023 and ending on or before March 31, 2026, except\nfor the taxes for the years and periods subject to subsection (a)(3) of this Section 6.11-2, the due date of\nthe return including any extensions; or\n(3) For the Gross Receipts Tax (Article 12-A-1) (including the tax on administrative\noffice business activities in Section 953.8 of Article 12-A-1), the Early Care and Education Commercial\nRents Tax (Article 21), the Homelessness Gross Receipts Tax (Article 28) (including the tax on\nadministrative office business activities in Section 2804(d) or Section 2804.9 of Article 28), the\nCannabis Business Tax (Article 30), and the Overpaid Executive Gross Receipts Tax (Article 33)\n(including the tax on administrative office business activities in Section 3303(g)) for tax years\nbeginning on or after January 1, 2025, and for the Business Registration Fee (Article 12) for\nregistration years beginning on or after April 1, 2026, November 30 of the calendar year in which the\nreturn for such respective tax was originally due.\n13SAN\nCO\n22 MAY (b) The notice of any determination under this Section 6.11-2 may be served upon the\nDEMANTHER OF CLIENT OFF\ntaxpayer or other affected person personally or by mail; if by mail, service shall be:\n(1) to the last known address that appears in the Tax Collector's records,\nprovided there is such an address in the Tax Collector's records, or\n(2) to an address that the Tax Collector concludes, in the Tax Collector's\ndiscretion, is the last known address of the person(s).", "If the person or combined group did not file a\n7SANT\nV\n772 MAY 10 TH 1:28\nreturn for the preceding tax year and did not request an extension under Section 6.9-4 to file that\nreturn, the person or combined group shall be deemed to have filed a return showing no\nliability for purposes of this Section 6.9-Isubsection (c)(3)(C)(ii), and no estimated tax payments\nof that tax shall be due for the current tax year.\n(D) Lessor of Residential Real Estate Exemption. Notwithstanding\nany other provision in this Section 6.9-1(c)(3), a lessor of residential real estate, as defined in\nSection 954.1 of Article 12-A-1, shall not be required to make estimated tax payments under\nthis Section 6.9-1(c)(3), but shall pay its full Gross Receipts Tax liability, Homelessness Gross\nReceipts Tax liability, and Overpaid Executive Gross Receipts Tax liability on or before the\nlast day of February of the succeeding year, or the extended deadline in Section 6.9-4, if the\nlessor's gross receipts within the City shown on the lessor's return for either the current tax\nyear or the preceding tax year did not exceed the threshold in Section 954.1(b) of\nArticle 12-A-1.\n*\n*\n(g) Business Registration Certificate; Annual Due Date.\n(1) For registration years ending on or before June 30, 2025 and the\n2025-2026 registration year, eExcept for any extensions granted under Section 6.9-4 or as\notherwise provided in this Business and Tax Regulations Code, returns and payments of the\nBusiness Registration Fee (Article 12) shall be filed annually and are due and payable, and\nshall be delinquent if not submitted and paid to the Tax Collector, on or before the last day of\nMay preceding the registration year commencing July 1 of that year.", "(a) The homelessness gross receipts tax rates applicable to Category 1 Business Activities for\ntax years beginning on or after January 1, 2025 are:\n0% for taxable gross receipts between $0 and $1,000,000\n0% for taxable gross receipts between $1,000,000.01 and $2,500,000\n0% for taxable gross receipts between $2,500,000.01 and $25,000,000\n0.164% for taxable gross receipts between $25,000,000.01 and $50,000,000\n0.164% for taxable gross receipts between $50,000,000.01 and $75,000,000\n0.164% for taxable gross receipts between $75,000,000.01 and $100,000,000\n0.164% for taxable gross receipts between $100,000,000.01 and $150,000,000\n0.246% for taxable gross receipts between $150,000,000.01 and $250,000,000\n0.328% for taxable gross receipts between $250,000,000.01 and $500,000,000\n0.41% for taxable gross receipts between $500,000,000.01 and $1,000,000,000\n0.492% for taxable gross receipts over $1,000,000,000\n(b) \"Category 1 Business Activities\" has the same meaning as in Section 953.20(b) of\nArticle 12-A-1 of this Business and Tax Regulations Code.\n(c) The amount of taxable gross receipts from Category 1 Business Activities subject to the\nhomelessness gross receipts tax shall be three-quarters of the amount determined under Section 956.1\nof Article 12-A-1 plus one-quarter of the amount determined under Section 956.2 of Article 12-A-1.\n84SAL\nSEC. 2804.3. HOMELESSNESS GROSS RECEIPTS TAX APPLICABLE TO CATEGORY 2\nBUSINESS ACTIVITIES."]}, {"source_sentence": "What are the different gross receipts tax rates applicable to private education and health services, administrative and support services, and miscellaneous business activities as outlined in Section 953.4?", "sentences": ["(b) For only so long as and to the extent that the City is prohibited from imposing the\nHomelessness Gross Receipts Tax, any person upon whom the City is prohibited under the\nConstitution or laws of the State of California or the Constitution or laws of the United States\nfrom imposing the Homelessness Gross Receipts Tax shall be exempt from the\nHomelessness Gross Receipts Tax.\n(c) For purposes of this Article 28, gross receipts shall not include receipts that are\nexcluded from gross receipts for purposes of the gross receipts tax imposed by Article 12-A-1,\nand also shall not include receipts subject to a gross receipts tax on commercial rents\nimposed as a result of a measure adopted by the voters of San Francisco in the June 5, 2018\nelection.\n(d) This Section 2805 shall apply to tax years ending on or before December 31, 2024.\nSEC. 2805.3. EXEMPTIONS AND EXCLUSIONS.\n(a) An organization that is exempt from income taxation by Chapter 4 (commencing with\nSection 23701) of Part 11 of Division 2 of the California Revenue and Taxation Code or Subchapter F\n(commencing with Section 501) of Chapter 1 of Subtitle A of the Internal Revenue Code of 1986, as\namended, as qualified by Sections 502, 503, 504, and 508 of the Internal Revenue Code of 1986, as\namended, shall be exempt from taxation under this Article 28, only so long as those exemptions\ncontinue to exist under state or federal law.\n(b) For purposes of this Article 28, gross receipts as defined in Section 952.3 shall not include\nreceipts from business activities if, and only so long as and to the extent that, the City is prohibited\n93SA\n01:32\nfrom taxing such receipts under the Constitution or laws of the United States or under the Constitution\nor laws of the State of California.\n(c) Rent Controlled Buildings Exclusion. A person subject to the homelessness gross receipts\ntax may exclude from gross receipts in any tax year 50% of the total amount received from the rental of\nreal property to tenants in occupancy at any location in the City, which is subject to limits on rent\nincreases pursuant to the Residential Rent Stabilization and Arbitration Ordinance, Administrative\nCode, Chapter 37, Section 37.1 et seq.", "10 that the\n90% gross receipts threshold has been met for tax year 2023, and for tax year 2024 if the Controller\ndoes not certify under Section 953.10 that the 95% gross receipts threshold has been met for tax\nyear 2024:\n42SAN LANCI Co\nFILED\n2024 MAY 10 PH 1:29\n0.45% (e.g. $4.50 per $1,000) for taxable gross receipts between $0 and $1,000,000\nDEPARTMENT OF ELECTIONS\n0.488% (e.g. $4.88 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.488% (e.g. $4.88 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n0.6% (e.g. $6 per $1,000) for taxable gross receipts over $25,000,000\n(D) For tax year 2024 if the Controller certifies under Section 953.10 that the\n95% gross receipts threshold has been met for tax year 2024, and for tax years beginning on or after\nJanuary 1, 2025:\n0.465% (e.g. $4.65 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.504% (e.g. $5.04 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.504% (e.g. $5.04 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n0.62% (e.g. $6.20 per $1,000) for taxable gross receipts over $25,000,000\nSEC. 953.4. GROSS RECEIPTS TAX APPLICABLE TO PRIVATE EDUCATION AND\nHEALTH SERVICES; ADMINISTRATIVE AND SUPPORT SERVICES; AND\nMISCELLANEOUS BUSINESS ACTIVITIES.\n(a) The gross receipts tax rates applicable to the business activities of private\neducation and health services, administrative and support services,", "SARI\nдо\n202 BAY 10 FA 1:32\nyear by 52, and dividing the result by the number of weeks that the employee was employed by that\nperson or combined group during the tax year.\n\"Highest-Paid Managerial Employee\" means the individual employee or officer of a\nperson or combined group with managerial responsibility in a business function who received the most\nCompensation for a tax year. For purposes of determining the Highest-Paid Managerial Employee and\nthe Compensation of such employee, Compensation shall not be annualized or converted to a full-time\nequivalency.\nSEC. 3303. IMPOSITION OF TAX.\n(a) Except as otherwise provided in this Article 33, commencing with tax years\nbeginning on or after January 1, 2022, for the privilege of engaging in business in the City, the\nCity imposes an annual Overpaid Executive Gross Receipts Tax on each person engaging in\nbusiness within the City where the Executive Pay Ratio for the tax year of that person or the\ncombined group of which it is a part exceeds 100:1."]}, {"source_sentence": "Based on the context information provided, what are the different gross receipts tax rates for businesses in San Francisco for tax years 2022, 2023, and 2024?", "sentences": ["$9.75 per $1,000) for taxable gross receipts over $25,000,000\n44SANCO\n2024 NAY LO\n(D) For tax year 2024 if the Controller certifies under Section 953.10 that the\nDEPARTMENT OF\n95% gross receipts threshold has been met for tax year 2024, and for tax years beginning on or after\nJanuary 1, 2025:\n0.814% (e.g. $8.14 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.853% (e.g. $8.53 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.93% (e.g. $9.30 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n1.008% (e.g. $10.08 per $1,000) for taxable gross receipts over $25,000,000\n(3) For all business activities not otherwise exempt and not elsewhere\nsubjected to a gross receipts tax rate or an administrative office tax by this Article 12-A-1:\n(B) For tax years 2022 and, if the Controller does not certify under\nSection 953.10 that the 90% gross receipts threshold has been met for tax year 2023, for tax\nyear 2023:\n0.788% (e.g. $7.88 per $1,000) for taxable gross receipts between $0 and $1,000,000\n0.825% (e.g. $8.25 per $1,000) for taxable gross receipts between $1,000,000.01 and\n$2,500,000\n0.9% (e.g. $9 per $1,000) for taxable gross receipts between $2,500,000.01 and\n$25,000,000\n0.975% (e.g. $9.75 per $1,000) for taxable gross receipts over $25,000,000\n(C) For tax year 2023 if the Controller certifies under Section 953.10 that the\n90% gross receipts threshold has been met for tax year 2023,", "(d) In no event shall the credit under this Section 960.4 reduce a person or combined group's\nGross Receipts Tax liability to less than $0 for any tax year. The credit under this Section shall not be\nrefundable and may not be carried forward to a subsequent year.\nSEC. 966. CONTROLLER REPORTS.\nThe Controller shall prepare reports by September 1, 2026, and September 1, 2027,\nrespectively, that discuss current economic conditions in the City and the performance of the tax system\nrevised by the voters in the ordinance adding this Section 966.\nSection 6. Article 21 of the Business and Tax Regulations Code is hereby amended by\nrevising Section 2106 to read as follows:\nSEC. 2106. SMALL BUSINESS EXEMPTION.\n(a) For tax years ending on or before December 31, 2024, nNotwithstanding any other\nprovision of this Article 21, a person or combined group exempt from payment of the gross\nreceipts tax under Section 954.1 of Article 12-A-1, as amended from time to time, shall also\nbe exempt from payment of the Early Care and Education Commercial Rents Tax.\n79SAN\nDL W(b) For tax years beginning on or after January 1, 2025, notwithstanding any other provision\nof this Article 21, a \"small business enterprise\" shall be exempt from payment of the Early Care and\nEducation Commercial Rents Tax. For purposes of this subsection (b), the term \"small business\nenterprise\" shall mean any person or combined group whose gross receipts within the City, determined\nunder Article 12-A-1, did not exceed $2,325,000, adjusted annually in accordance with the increase in\nthe Consumer Price Index: All Urban Consumers for the San Francisco/Oakland/Hayward Area for All\nItems as reported by the United States Bureau of Labor Statistics, or any successor to that index, as of\nDecember 31 of the calendar year two years prior to the tax year, beginning with tax year 2026, and\nrounded to the nearest $10,000. This subsection (b) shall not apply to a person or combined group\nsubject to a tax on administrative office business activities in Section 953.8 of Article 12-A-1.\nSection 7.", "(1) For tax years ending on or before December 31, 2024, the overpaid executive\nadministrative office tax shall be calculated as follows:\n(+4) 0.4% of the person or combined group's total payroll expense\nattributable to the City for a tax year if the person or combined group has an Executive Pay\nRatio for that tax year of greater than 100:1, but less than or equal to 200:1;\n(2B) 0.8% of the person or combined group's total payroll expense\nattributable to the City for a tax year if the person or combined group has an Executive Pay\nRatio for that tax year of greater than 200:1, but less than or equal to 300:1;\n(3C) 1.2% of the person or combined group's total payroll expense\nattributable to the City for a tax year if the person or combined group has an Executive Pay\nRatio for that tax year of greater than 300:1, but less than or equal to 400:1;\n103SAULAMIC 50\n10 PM 1:32\n(4D) 1.6% of the person or combined group's total payroll expense\nDEPARTMENT OF ELECTIONS\nattributable to the City for a tax year if the person or combined group has an Executive Pay\nRatio for that tax year of greater than 400:1, but less than or equal to 500:1;\n(5E) 2% of the person or combined group's total payroll expense\nattributable to the City for a tax year if the person or combined group has an Executive Pay\nRatio for that tax year of greater than 500:1, but less than or equal to 600:1; or\n(6F) 2.4% of the person or combined group's total payroll expense\nattributable to the City for a tax year if the person or combined group has an Executive Pay\nRatio for that tax year of greater than 600:1."]}], "model-index": [{"name": "SentenceTransformer based on microsoft/mpnet-base", "results": [{"task": {"type": "knowledge-distillation", "name": "Knowledge Distillation"}, "dataset": {"name": "stsb dev", "type": "stsb-dev"}, "metrics": [{"type": "negative_mse", "value": -2.4281997233629227, "name": "Negative Mse"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,824 |
fine-tuned/jinaai_jina-embeddings-v2-base-de-15_8_2024-h1i4-webapp
|
fine-tuned
|
feature-extraction
|
[
"sentence-transformers",
"safetensors",
"bert",
"feature-extraction",
"sentence-similarity",
"mteb",
"Academic",
"Research",
"Papers",
"Information",
"System",
"custom_code",
"en",
"dataset:fine-tuned/jinaai_jina-embeddings-v2-base-de-15_8_2024-h1i4-webapp",
"dataset:allenai/c4",
"license:apache-2.0",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] | 2024-08-15T21:33:41Z |
2024-08-15T21:34:05+00:00
| 7 | 0 |
---
datasets:
- fine-tuned/jinaai_jina-embeddings-v2-base-de-15_8_2024-h1i4-webapp
- allenai/c4
language:
- en
- en
license: apache-2.0
pipeline_tag: feature-extraction
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- mteb
- Academic
- Research
- Papers
- Information
- System
---
This model is a fine-tuned version of [**jinaai/jina-embeddings-v2-base-de**](https://huggingface.co/jinaai/jina-embeddings-v2-base-de) designed for the following use case:
information retrieval system for academic research papers
## How to Use
This model can be easily integrated into your NLP pipeline for tasks such as text classification, sentiment analysis, entity recognition, and more. Here's a simple example to get you started:
```python
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import cos_sim
model = SentenceTransformer(
'fine-tuned/jinaai_jina-embeddings-v2-base-de-15_8_2024-h1i4-webapp',
trust_remote_code=True
)
embeddings = model.encode([
'first text to embed',
'second text to embed'
])
print(cos_sim(embeddings[0], embeddings[1]))
```
| null |
Non_BioNLP
|
This model is a fine-tuned version of [**jinaai/jina-embeddings-v2-base-de**](https://huggingface.co/jinaai/jina-embeddings-v2-base-de) designed for the following use case:
information retrieval system for academic research papers
## How to Use
This model can be easily integrated into your NLP pipeline for tasks such as text classification, sentiment analysis, entity recognition, and more. Here's a simple example to get you started:
```python
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import cos_sim
model = SentenceTransformer(
'fine-tuned/jinaai_jina-embeddings-v2-base-de-15_8_2024-h1i4-webapp',
trust_remote_code=True
)
embeddings = model.encode([
'first text to embed',
'second text to embed'
])
print(cos_sim(embeddings[0], embeddings[1]))
```
|
{"datasets": ["fine-tuned/jinaai_jina-embeddings-v2-base-de-15_8_2024-h1i4-webapp", "allenai/c4"], "language": ["en", "en"], "license": "apache-2.0", "pipeline_tag": "feature-extraction", "tags": ["sentence-transformers", "feature-extraction", "sentence-similarity", "mteb", "Academic", "Research", "Papers", "Information", "System"]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,825 |
Pubudu/mbart-large-50_par_bn_rf_16_dinamina_1700
|
Pubudu
| null |
[
"adapter-transformers",
"mbart",
"adapterhub:summarization/dinamina_1700_full_text",
"dataset:dinamina_1700_full_text",
"region:us"
] | 2024-05-19T02:49:47Z |
2024-05-19T02:50:45+00:00
| 0 | 0 |
---
datasets:
- dinamina_1700_full_text
tags:
- adapter-transformers
- mbart
- adapterhub:summarization/dinamina_1700_full_text
---
# Adapter `Pubudu/mbart-large-50_par_bn_rf_16_dinamina_1700` for facebook/mbart-large-50
An [adapter](https://adapterhub.ml) for the `facebook/mbart-large-50` model that was trained on the [summarization/dinamina_1700_full_text](https://adapterhub.ml/explore/summarization/dinamina_1700_full_text/) dataset.
This adapter was created for usage with the **[Adapters](https://github.com/Adapter-Hub/adapters)** library.
## Usage
First, install `adapters`:
```
pip install -U adapters
```
Now, the adapter can be loaded and activated like this:
```python
from adapters import AutoAdapterModel
model = AutoAdapterModel.from_pretrained("facebook/mbart-large-50")
adapter_name = model.load_adapter("Pubudu/mbart-large-50_par_bn_rf_16_dinamina_1700", source="hf", set_active=True)
```
## Architecture & Training
<!-- Add some description here -->
## Evaluation results
<!-- Add some description here -->
## Citation
<!-- Add some description here -->
| null |
Non_BioNLP
|
# Adapter `Pubudu/mbart-large-50_par_bn_rf_16_dinamina_1700` for facebook/mbart-large-50
An [adapter](https://adapterhub.ml) for the `facebook/mbart-large-50` model that was trained on the [summarization/dinamina_1700_full_text](https://adapterhub.ml/explore/summarization/dinamina_1700_full_text/) dataset.
This adapter was created for usage with the **[Adapters](https://github.com/Adapter-Hub/adapters)** library.
## Usage
First, install `adapters`:
```
pip install -U adapters
```
Now, the adapter can be loaded and activated like this:
```python
from adapters import AutoAdapterModel
model = AutoAdapterModel.from_pretrained("facebook/mbart-large-50")
adapter_name = model.load_adapter("Pubudu/mbart-large-50_par_bn_rf_16_dinamina_1700", source="hf", set_active=True)
```
## Architecture & Training
<!-- Add some description here -->
## Evaluation results
<!-- Add some description here -->
## Citation
<!-- Add some description here -->
|
{"datasets": ["dinamina_1700_full_text"], "tags": ["adapter-transformers", "mbart", "adapterhub:summarization/dinamina_1700_full_text"]}
|
task
|
[
"SUMMARIZATION"
] | 39,826 |
HPLT/translate-ar-en-v1.0-hplt
|
HPLT
|
translation
|
[
"transformers",
"pytorch",
"marian",
"text2text-generation",
"translation",
"ar",
"en",
"license:cc-by-4.0",
"autotrain_compatible",
"region:us"
] | 2024-02-27T11:37:42Z |
2024-03-14T00:48:59+00:00
| 33 | 3 |
---
language:
- ar
- en
license: cc-by-4.0
tags:
- translation
inference: false
---
### HPLT MT release v1.0
This repository contains the translation model for Arabic-English trained with HPLT data only. The model is available in both Marian and Hugging Face formats.
### Model Info
* Source language: Arabic
* Target language: English
* Data: HPLT data only
* Model architecture: Transformer-base
* Tokenizer: SentencePiece (Unigram)
* Cleaning: We used [OpusCleaner](https://github.com/hplt-project/OpusCleaner) with a set of basic rules. Details can be found in the filter files [here](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/ar-en/raw/v0).
You can check out our [deliverable report](https://hplt-project.org/HPLT_D5_1___Translation_models_for_select_language_pairs.pdf), [GitHub repository](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0), and [website](https://hplt-project.org) for more details.
### Usage
**Note** that for quality considerations, we recommend using [HPLT/translate-ar-en-v1.0-hplt_opus](https://huggingface.co/HPLT/translate-ar-en-v1.0-hplt_opus) instead of this model.
The model has been trained with [MarianNMT](https://github.com/marian-nmt/marian) and the weights are in the Marian format. We have also converted the model into the Hugging Face format so it is compatible with `transformers`.
#### Using Marian
To run inference with MarianNMT, refer to the [Inference/Decoding/Translation](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0#inferencedecodingtranslation) section of our GitHub repository. You will need the model file `model.npz.best-chrf.npz` and the vocabulary file `model.ar-en.spm` from this repository.
#### Using transformers
We have also converted this model to the Hugging Face format and you can get started with the script below. **Note** that due a [known issue](https://github.com/huggingface/transformers/issues/26216) in weight conversion, the checkpoint cannot work with transformer versions <4.26 or >4.30. We tested and suggest `pip install transformers==4.28`.
```
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("HPLT/translate-ar-en-v1.0-hplt")
model = AutoModelForSeq2SeqLM.from_pretrained("HPLT/translate-ar-en-v1.0-hplt")
inputs = ["Input goes here.", "Make sure the language is right."]
batch_tokenized = tokenizer(inputs, return_tensors="pt", padding=True)
model_output = model.generate(
**batch_tokenized, num_beams=6, max_new_tokens=512
)
batch_detokenized = tokenizer.batch_decode(
model_output,
skip_special_tokens=True,
)
print(batch_detokenized)
```
## Benchmarks
When decoded using Marian, the model has the following test scores.
| Test set | BLEU | chrF++ | COMET22 |
| -------------------------------------- | ---- | ----- | ----- |
| FLORES200 | 35.0 | 58.5 | 0.8396 |
| NTREX | 28.6 | 54.6 | 0.8194 |
### Acknowledgements
This project has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No 101070350 and from UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee [grant number 10052546]
Brought to you by researchers from the University of Edinburgh and Charles University in Prague with support from the whole HPLT consortium.
| null |
Non_BioNLP
|
### HPLT MT release v1.0
This repository contains the translation model for Arabic-English trained with HPLT data only. The model is available in both Marian and Hugging Face formats.
### Model Info
* Source language: Arabic
* Target language: English
* Data: HPLT data only
* Model architecture: Transformer-base
* Tokenizer: SentencePiece (Unigram)
* Cleaning: We used [OpusCleaner](https://github.com/hplt-project/OpusCleaner) with a set of basic rules. Details can be found in the filter files [here](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/ar-en/raw/v0).
You can check out our [deliverable report](https://hplt-project.org/HPLT_D5_1___Translation_models_for_select_language_pairs.pdf), [GitHub repository](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0), and [website](https://hplt-project.org) for more details.
### Usage
**Note** that for quality considerations, we recommend using [HPLT/translate-ar-en-v1.0-hplt_opus](https://huggingface.co/HPLT/translate-ar-en-v1.0-hplt_opus) instead of this model.
The model has been trained with [MarianNMT](https://github.com/marian-nmt/marian) and the weights are in the Marian format. We have also converted the model into the Hugging Face format so it is compatible with `transformers`.
#### Using Marian
To run inference with MarianNMT, refer to the [Inference/Decoding/Translation](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0#inferencedecodingtranslation) section of our GitHub repository. You will need the model file `model.npz.best-chrf.npz` and the vocabulary file `model.ar-en.spm` from this repository.
#### Using transformers
We have also converted this model to the Hugging Face format and you can get started with the script below. **Note** that due a [known issue](https://github.com/huggingface/transformers/issues/26216) in weight conversion, the checkpoint cannot work with transformer versions <4.26 or >4.30. We tested and suggest `pip install transformers==4.28`.
```
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("HPLT/translate-ar-en-v1.0-hplt")
model = AutoModelForSeq2SeqLM.from_pretrained("HPLT/translate-ar-en-v1.0-hplt")
inputs = ["Input goes here.", "Make sure the language is right."]
batch_tokenized = tokenizer(inputs, return_tensors="pt", padding=True)
model_output = model.generate(
**batch_tokenized, num_beams=6, max_new_tokens=512
)
batch_detokenized = tokenizer.batch_decode(
model_output,
skip_special_tokens=True,
)
print(batch_detokenized)
```
## Benchmarks
When decoded using Marian, the model has the following test scores.
| Test set | BLEU | chrF++ | COMET22 |
| -------------------------------------- | ---- | ----- | ----- |
| FLORES200 | 35.0 | 58.5 | 0.8396 |
| NTREX | 28.6 | 54.6 | 0.8194 |
### Acknowledgements
This project has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No 101070350 and from UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee [grant number 10052546]
Brought to you by researchers from the University of Edinburgh and Charles University in Prague with support from the whole HPLT consortium.
|
{"language": ["ar", "en"], "license": "cc-by-4.0", "tags": ["translation"], "inference": false}
|
task
|
[
"TRANSLATION"
] | 39,827 |
abbasmahmudiai/MT5_en_to_persian
|
abbasmahmudiai
|
translation
|
[
"transformers",
"safetensors",
"mt5",
"text2text-generation",
"translation",
"en",
"fa",
"dataset:persiannlp/parsinlu_translation_en_fa",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-06-26T17:03:21Z |
2024-06-26T19:19:40+00:00
| 17 | 1 |
---
datasets:
- persiannlp/parsinlu_translation_en_fa
language:
- en
- fa
license: mit
pipeline_tag: translation
---
| null |
Non_BioNLP
|
{"datasets": ["persiannlp/parsinlu_translation_en_fa"], "language": ["en", "fa"], "license": "mit", "pipeline_tag": "translation"}
|
task
|
[
"TRANSLATION"
] | 39,828 |
|
RMWeerasinghe/t5-small-finetuned
|
RMWeerasinghe
|
summarization
|
[
"transformers",
"safetensors",
"t5",
"text2text-generation",
"summarization",
"generated_from_trainer",
"base_model:google-t5/t5-small",
"base_model:finetune:google-t5/t5-small",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2024-02-13T04:59:25Z |
2024-02-13T05:08:36+00:00
| 16 | 0 |
---
base_model: google-t5/t5-small
license: apache-2.0
metrics:
- rouge
pipeline_tag: summarization
tags:
- summarization
- generated_from_trainer
model-index:
- name: t5-small-finetuned
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-small-finetuned
This model is a fine-tuned version of [google-t5/t5-small](https://huggingface.co/google-t5/t5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 13.3545
- Rouge1: 0.0324
- Rouge2: 0.0035
- Rougel: 0.0283
- Rougelsum: 0.0297
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
| No log | 0.67 | 1 | 25.3754 | 0.0458 | 0.0078 | 0.038 | 0.0396 |
| No log | 2.0 | 3 | 23.7399 | 0.0458 | 0.0078 | 0.038 | 0.0396 |
| No log | 2.67 | 4 | 22.8640 | 0.0442 | 0.0053 | 0.0367 | 0.0384 |
| No log | 4.0 | 6 | 21.0827 | 0.0442 | 0.0053 | 0.0367 | 0.0384 |
| No log | 4.67 | 7 | 20.1867 | 0.0442 | 0.0053 | 0.0367 | 0.0384 |
| No log | 6.0 | 9 | 18.3401 | 0.0431 | 0.0109 | 0.0368 | 0.0388 |
| No log | 6.67 | 10 | 17.5540 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 8.0 | 12 | 16.5123 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 8.67 | 13 | 16.2865 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 10.0 | 15 | 15.9394 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 10.67 | 16 | 15.7787 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 12.0 | 18 | 15.4614 | 0.0406 | 0.004 | 0.0331 | 0.0361 |
| No log | 12.67 | 19 | 15.3169 | 0.037 | 0.0012 | 0.0288 | 0.032 |
| 17.4357 | 14.0 | 21 | 15.0546 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 14.67 | 22 | 14.9349 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 16.0 | 24 | 14.7097 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 16.67 | 25 | 14.6033 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 18.0 | 27 | 14.4049 | 0.0365 | 0.0023 | 0.0298 | 0.0337 |
| 17.4357 | 18.67 | 28 | 14.3124 | 0.0365 | 0.0023 | 0.0298 | 0.0337 |
| 17.4357 | 20.0 | 30 | 14.1419 | 0.0324 | 0.0023 | 0.0271 | 0.0296 |
| 17.4357 | 20.67 | 31 | 14.0635 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 22.0 | 33 | 13.9163 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 22.67 | 34 | 13.8491 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 24.0 | 36 | 13.7281 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 24.67 | 37 | 13.6752 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 26.0 | 39 | 13.5841 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 26.67 | 40 | 13.5448 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 28.0 | 42 | 13.4779 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 28.67 | 43 | 13.4500 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 30.0 | 45 | 13.4051 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 30.67 | 46 | 13.3881 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 32.0 | 48 | 13.3645 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 32.67 | 49 | 13.3578 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 33.33 | 50 | 13.3545 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
### Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.2.0
- Datasets 2.16.1
- Tokenizers 0.15.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-small-finetuned
This model is a fine-tuned version of [google-t5/t5-small](https://huggingface.co/google-t5/t5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 13.3545
- Rouge1: 0.0324
- Rouge2: 0.0035
- Rougel: 0.0283
- Rougelsum: 0.0297
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
| No log | 0.67 | 1 | 25.3754 | 0.0458 | 0.0078 | 0.038 | 0.0396 |
| No log | 2.0 | 3 | 23.7399 | 0.0458 | 0.0078 | 0.038 | 0.0396 |
| No log | 2.67 | 4 | 22.8640 | 0.0442 | 0.0053 | 0.0367 | 0.0384 |
| No log | 4.0 | 6 | 21.0827 | 0.0442 | 0.0053 | 0.0367 | 0.0384 |
| No log | 4.67 | 7 | 20.1867 | 0.0442 | 0.0053 | 0.0367 | 0.0384 |
| No log | 6.0 | 9 | 18.3401 | 0.0431 | 0.0109 | 0.0368 | 0.0388 |
| No log | 6.67 | 10 | 17.5540 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 8.0 | 12 | 16.5123 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 8.67 | 13 | 16.2865 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 10.0 | 15 | 15.9394 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 10.67 | 16 | 15.7787 | 0.0405 | 0.0054 | 0.0343 | 0.0346 |
| No log | 12.0 | 18 | 15.4614 | 0.0406 | 0.004 | 0.0331 | 0.0361 |
| No log | 12.67 | 19 | 15.3169 | 0.037 | 0.0012 | 0.0288 | 0.032 |
| 17.4357 | 14.0 | 21 | 15.0546 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 14.67 | 22 | 14.9349 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 16.0 | 24 | 14.7097 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 16.67 | 25 | 14.6033 | 0.0372 | 0.0023 | 0.0302 | 0.0345 |
| 17.4357 | 18.0 | 27 | 14.4049 | 0.0365 | 0.0023 | 0.0298 | 0.0337 |
| 17.4357 | 18.67 | 28 | 14.3124 | 0.0365 | 0.0023 | 0.0298 | 0.0337 |
| 17.4357 | 20.0 | 30 | 14.1419 | 0.0324 | 0.0023 | 0.0271 | 0.0296 |
| 17.4357 | 20.67 | 31 | 14.0635 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 22.0 | 33 | 13.9163 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 22.67 | 34 | 13.8491 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 24.0 | 36 | 13.7281 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 24.67 | 37 | 13.6752 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 17.4357 | 26.0 | 39 | 13.5841 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 26.67 | 40 | 13.5448 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 28.0 | 42 | 13.4779 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 28.67 | 43 | 13.4500 | 0.0324 | 0.0023 | 0.0272 | 0.0297 |
| 13.2934 | 30.0 | 45 | 13.4051 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 30.67 | 46 | 13.3881 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 32.0 | 48 | 13.3645 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 32.67 | 49 | 13.3578 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
| 13.2934 | 33.33 | 50 | 13.3545 | 0.0324 | 0.0035 | 0.0283 | 0.0297 |
### Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.2.0
- Datasets 2.16.1
- Tokenizers 0.15.1
|
{"base_model": "google-t5/t5-small", "license": "apache-2.0", "metrics": ["rouge"], "pipeline_tag": "summarization", "tags": ["summarization", "generated_from_trainer"], "model-index": [{"name": "t5-small-finetuned", "results": []}]}
|
task
|
[
"SUMMARIZATION"
] | 39,829 |
SotirisLegkas/Llama3_ALL_BCE_translations_19_shuffled_special_tokens
|
SotirisLegkas
| null |
[
"peft",
"safetensors",
"generated_from_trainer",
"base_model:meta-llama/Meta-Llama-3-8B-Instruct",
"base_model:adapter:meta-llama/Meta-Llama-3-8B-Instruct",
"license:llama3",
"region:us"
] | 2024-05-14T16:27:29Z |
2024-05-14T16:28:10+00:00
| 0 | 0 |
---
base_model: meta-llama/Meta-Llama-3-8B-Instruct
library_name: peft
license: llama3
tags:
- generated_from_trainer
model-index:
- name: Llama3_ALL_BCE_translations_19_shuffled_special_tokens
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# Llama3_ALL_BCE_translations_19_shuffled_special_tokens
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4776
- F1 Macro 0.1: 0.0818
- F1 Macro 0.15: 0.0922
- F1 Macro 0.2: 0.1027
- F1 Macro 0.25: 0.1130
- F1 Macro 0.3: 0.1230
- F1 Macro 0.35: 0.1336
- F1 Macro 0.4: 0.1440
- F1 Macro 0.45: 0.1551
- F1 Macro 0.5: 0.1663
- F1 Macro 0.55: 0.1778
- F1 Macro 0.6: 0.1879
- F1 Macro 0.65: 0.1987
- F1 Macro 0.7: 0.2090
- F1 Macro 0.75: 0.2178
- F1 Macro 0.8: 0.2211
- F1 Macro 0.85: 0.2205
- F1 Macro 0.9: 0.2010
- F1 Macro 0.95: 0.1457
- Threshold 0: 0.65
- Threshold 1: 0.75
- Threshold 2: 0.7
- Threshold 3: 0.85
- Threshold 4: 0.8
- Threshold 5: 0.85
- Threshold 6: 0.8
- Threshold 7: 0.8
- Threshold 8: 0.85
- Threshold 9: 0.75
- Threshold 10: 0.85
- Threshold 11: 0.8
- Threshold 12: 0.85
- Threshold 13: 0.95
- Threshold 14: 0.85
- Threshold 15: 0.75
- Threshold 16: 0.85
- Threshold 17: 0.8
- Threshold 18: 0.9
- 0: 0.0619
- 1: 0.1388
- 2: 0.1978
- 3: 0.1328
- 4: 0.2961
- 5: 0.3489
- 6: 0.3179
- 7: 0.1268
- 8: 0.2043
- 9: 0.3668
- 10: 0.3216
- 11: 0.3669
- 12: 0.1276
- 13: 0.1205
- 14: 0.2264
- 15: 0.1576
- 16: 0.3078
- 17: 0.3722
- 18: 0.125
- Max F1: 0.2211
- Mean F1: 0.2273
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 2024
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Macro 0.1 | F1 Macro 0.15 | F1 Macro 0.2 | F1 Macro 0.25 | F1 Macro 0.3 | F1 Macro 0.35 | F1 Macro 0.4 | F1 Macro 0.45 | F1 Macro 0.5 | F1 Macro 0.55 | F1 Macro 0.6 | F1 Macro 0.65 | F1 Macro 0.7 | F1 Macro 0.75 | F1 Macro 0.8 | F1 Macro 0.85 | F1 Macro 0.9 | F1 Macro 0.95 | Threshold 0 | Threshold 1 | Threshold 2 | Threshold 3 | Threshold 4 | Threshold 5 | Threshold 6 | Threshold 7 | Threshold 8 | Threshold 9 | Threshold 10 | Threshold 11 | Threshold 12 | Threshold 13 | Threshold 14 | Threshold 15 | Threshold 16 | Threshold 17 | Threshold 18 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | Max F1 | Mean F1 |
|:-------------:|:-----:|:-----:|:---------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:-------:|
| 3.3824 | 1.0 | 5595 | 4.3847 | 0.0700 | 0.0761 | 0.0818 | 0.0877 | 0.0936 | 0.1000 | 0.1064 | 0.1134 | 0.1196 | 0.1265 | 0.1327 | 0.1381 | 0.1432 | 0.1483 | 0.1465 | 0.1417 | 0.1291 | 0.0836 | 0.65 | 0.9 | 0.85 | 0.9 | 0.75 | 0.6 | 0.8 | 0.75 | 0.9 | 0.9 | 0.9 | 0.85 | 0.9 | 0.0 | 0.85 | 0.75 | 0.6 | 0.6 | 0.9 | 0.0649 | 0.0879 | 0.1603 | 0.0899 | 0.2589 | 0.2876 | 0.2683 | 0.1036 | 0.1245 | 0.2856 | 0.2387 | 0.3033 | 0.0726 | 0.0 | 0.1779 | 0.1109 | 0.2192 | 0.2743 | 0.0641 | 0.1483 | 0.1680 |
| 2.4859 | 2.0 | 11190 | 1.7537 | 0.0881 | 0.0994 | 0.1111 | 0.1210 | 0.1310 | 0.1401 | 0.1472 | 0.1541 | 0.1607 | 0.1676 | 0.1697 | 0.1731 | 0.1768 | 0.1761 | 0.1713 | 0.1575 | 0.1365 | 0.0927 | 0.55 | 0.7 | 0.85 | 0.8 | 0.4 | 0.35 | 0.95 | 0.75 | 0.7 | 0.85 | 0.8 | 0.65 | 0.8 | 0.95 | 0.8 | 0.7 | 0.85 | 0.6 | 0.75 | 0.0534 | 0.1241 | 0.1924 | 0.1020 | 0.2738 | 0.3163 | 0.3072 | 0.1109 | 0.1793 | 0.3414 | 0.2889 | 0.3332 | 0.0831 | 0.0870 | 0.2137 | 0.1305 | 0.2881 | 0.3396 | 0.1254 | 0.1768 | 0.2048 |
| 1.7561 | 3.0 | 16785 | 1.4633 | 0.0840 | 0.0954 | 0.1062 | 0.1164 | 0.1271 | 0.1382 | 0.1485 | 0.1597 | 0.1713 | 0.1809 | 0.1895 | 0.1976 | 0.2056 | 0.2113 | 0.2115 | 0.1995 | 0.1805 | 0.1184 | 0.6 | 0.75 | 0.75 | 0.95 | 0.8 | 0.7 | 0.9 | 0.8 | 0.8 | 0.7 | 0.8 | 0.8 | 0.9 | 0.95 | 0.75 | 0.8 | 0.7 | 0.7 | 0.8 | 0.0581 | 0.1395 | 0.1946 | 0.1235 | 0.2818 | 0.3391 | 0.3151 | 0.1202 | 0.1997 | 0.3656 | 0.3056 | 0.3630 | 0.1340 | 0.1087 | 0.2272 | 0.1482 | 0.2953 | 0.3589 | 0.1233 | 0.2115 | 0.2211 |
| 1.2709 | 4.0 | 22380 | 1.4776 | 0.0818 | 0.0922 | 0.1027 | 0.1130 | 0.1230 | 0.1336 | 0.1440 | 0.1551 | 0.1663 | 0.1778 | 0.1879 | 0.1987 | 0.2090 | 0.2178 | 0.2211 | 0.2205 | 0.2010 | 0.1457 | 0.65 | 0.75 | 0.7 | 0.85 | 0.8 | 0.85 | 0.8 | 0.8 | 0.85 | 0.75 | 0.85 | 0.8 | 0.85 | 0.95 | 0.85 | 0.75 | 0.85 | 0.8 | 0.9 | 0.0619 | 0.1388 | 0.1978 | 0.1328 | 0.2961 | 0.3489 | 0.3179 | 0.1268 | 0.2043 | 0.3668 | 0.3216 | 0.3669 | 0.1276 | 0.1205 | 0.2264 | 0.1576 | 0.3078 | 0.3722 | 0.125 | 0.2211 | 0.2273 |
### Framework versions
- PEFT 0.10.0
- Transformers 4.40.2
- Pytorch 2.2.2+cu121
- Datasets 2.18.0
- Tokenizers 0.19.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# Llama3_ALL_BCE_translations_19_shuffled_special_tokens
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4776
- F1 Macro 0.1: 0.0818
- F1 Macro 0.15: 0.0922
- F1 Macro 0.2: 0.1027
- F1 Macro 0.25: 0.1130
- F1 Macro 0.3: 0.1230
- F1 Macro 0.35: 0.1336
- F1 Macro 0.4: 0.1440
- F1 Macro 0.45: 0.1551
- F1 Macro 0.5: 0.1663
- F1 Macro 0.55: 0.1778
- F1 Macro 0.6: 0.1879
- F1 Macro 0.65: 0.1987
- F1 Macro 0.7: 0.2090
- F1 Macro 0.75: 0.2178
- F1 Macro 0.8: 0.2211
- F1 Macro 0.85: 0.2205
- F1 Macro 0.9: 0.2010
- F1 Macro 0.95: 0.1457
- Threshold 0: 0.65
- Threshold 1: 0.75
- Threshold 2: 0.7
- Threshold 3: 0.85
- Threshold 4: 0.8
- Threshold 5: 0.85
- Threshold 6: 0.8
- Threshold 7: 0.8
- Threshold 8: 0.85
- Threshold 9: 0.75
- Threshold 10: 0.85
- Threshold 11: 0.8
- Threshold 12: 0.85
- Threshold 13: 0.95
- Threshold 14: 0.85
- Threshold 15: 0.75
- Threshold 16: 0.85
- Threshold 17: 0.8
- Threshold 18: 0.9
- 0: 0.0619
- 1: 0.1388
- 2: 0.1978
- 3: 0.1328
- 4: 0.2961
- 5: 0.3489
- 6: 0.3179
- 7: 0.1268
- 8: 0.2043
- 9: 0.3668
- 10: 0.3216
- 11: 0.3669
- 12: 0.1276
- 13: 0.1205
- 14: 0.2264
- 15: 0.1576
- 16: 0.3078
- 17: 0.3722
- 18: 0.125
- Max F1: 0.2211
- Mean F1: 0.2273
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 2024
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Macro 0.1 | F1 Macro 0.15 | F1 Macro 0.2 | F1 Macro 0.25 | F1 Macro 0.3 | F1 Macro 0.35 | F1 Macro 0.4 | F1 Macro 0.45 | F1 Macro 0.5 | F1 Macro 0.55 | F1 Macro 0.6 | F1 Macro 0.65 | F1 Macro 0.7 | F1 Macro 0.75 | F1 Macro 0.8 | F1 Macro 0.85 | F1 Macro 0.9 | F1 Macro 0.95 | Threshold 0 | Threshold 1 | Threshold 2 | Threshold 3 | Threshold 4 | Threshold 5 | Threshold 6 | Threshold 7 | Threshold 8 | Threshold 9 | Threshold 10 | Threshold 11 | Threshold 12 | Threshold 13 | Threshold 14 | Threshold 15 | Threshold 16 | Threshold 17 | Threshold 18 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | Max F1 | Mean F1 |
|:-------------:|:-----:|:-----:|:---------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:------------:|:-------------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:-----------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:------:|:-------:|
| 3.3824 | 1.0 | 5595 | 4.3847 | 0.0700 | 0.0761 | 0.0818 | 0.0877 | 0.0936 | 0.1000 | 0.1064 | 0.1134 | 0.1196 | 0.1265 | 0.1327 | 0.1381 | 0.1432 | 0.1483 | 0.1465 | 0.1417 | 0.1291 | 0.0836 | 0.65 | 0.9 | 0.85 | 0.9 | 0.75 | 0.6 | 0.8 | 0.75 | 0.9 | 0.9 | 0.9 | 0.85 | 0.9 | 0.0 | 0.85 | 0.75 | 0.6 | 0.6 | 0.9 | 0.0649 | 0.0879 | 0.1603 | 0.0899 | 0.2589 | 0.2876 | 0.2683 | 0.1036 | 0.1245 | 0.2856 | 0.2387 | 0.3033 | 0.0726 | 0.0 | 0.1779 | 0.1109 | 0.2192 | 0.2743 | 0.0641 | 0.1483 | 0.1680 |
| 2.4859 | 2.0 | 11190 | 1.7537 | 0.0881 | 0.0994 | 0.1111 | 0.1210 | 0.1310 | 0.1401 | 0.1472 | 0.1541 | 0.1607 | 0.1676 | 0.1697 | 0.1731 | 0.1768 | 0.1761 | 0.1713 | 0.1575 | 0.1365 | 0.0927 | 0.55 | 0.7 | 0.85 | 0.8 | 0.4 | 0.35 | 0.95 | 0.75 | 0.7 | 0.85 | 0.8 | 0.65 | 0.8 | 0.95 | 0.8 | 0.7 | 0.85 | 0.6 | 0.75 | 0.0534 | 0.1241 | 0.1924 | 0.1020 | 0.2738 | 0.3163 | 0.3072 | 0.1109 | 0.1793 | 0.3414 | 0.2889 | 0.3332 | 0.0831 | 0.0870 | 0.2137 | 0.1305 | 0.2881 | 0.3396 | 0.1254 | 0.1768 | 0.2048 |
| 1.7561 | 3.0 | 16785 | 1.4633 | 0.0840 | 0.0954 | 0.1062 | 0.1164 | 0.1271 | 0.1382 | 0.1485 | 0.1597 | 0.1713 | 0.1809 | 0.1895 | 0.1976 | 0.2056 | 0.2113 | 0.2115 | 0.1995 | 0.1805 | 0.1184 | 0.6 | 0.75 | 0.75 | 0.95 | 0.8 | 0.7 | 0.9 | 0.8 | 0.8 | 0.7 | 0.8 | 0.8 | 0.9 | 0.95 | 0.75 | 0.8 | 0.7 | 0.7 | 0.8 | 0.0581 | 0.1395 | 0.1946 | 0.1235 | 0.2818 | 0.3391 | 0.3151 | 0.1202 | 0.1997 | 0.3656 | 0.3056 | 0.3630 | 0.1340 | 0.1087 | 0.2272 | 0.1482 | 0.2953 | 0.3589 | 0.1233 | 0.2115 | 0.2211 |
| 1.2709 | 4.0 | 22380 | 1.4776 | 0.0818 | 0.0922 | 0.1027 | 0.1130 | 0.1230 | 0.1336 | 0.1440 | 0.1551 | 0.1663 | 0.1778 | 0.1879 | 0.1987 | 0.2090 | 0.2178 | 0.2211 | 0.2205 | 0.2010 | 0.1457 | 0.65 | 0.75 | 0.7 | 0.85 | 0.8 | 0.85 | 0.8 | 0.8 | 0.85 | 0.75 | 0.85 | 0.8 | 0.85 | 0.95 | 0.85 | 0.75 | 0.85 | 0.8 | 0.9 | 0.0619 | 0.1388 | 0.1978 | 0.1328 | 0.2961 | 0.3489 | 0.3179 | 0.1268 | 0.2043 | 0.3668 | 0.3216 | 0.3669 | 0.1276 | 0.1205 | 0.2264 | 0.1576 | 0.3078 | 0.3722 | 0.125 | 0.2211 | 0.2273 |
### Framework versions
- PEFT 0.10.0
- Transformers 4.40.2
- Pytorch 2.2.2+cu121
- Datasets 2.18.0
- Tokenizers 0.19.1
|
{"base_model": "meta-llama/Meta-Llama-3-8B-Instruct", "library_name": "peft", "license": "llama3", "tags": ["generated_from_trainer"], "model-index": [{"name": "Llama3_ALL_BCE_translations_19_shuffled_special_tokens", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,830 |
EleutherAI/pile-t5-xl
|
EleutherAI
|
text2text-generation
|
[
"transformers",
"safetensors",
"umt5",
"text2text-generation",
"t5x",
"encoder-decoder",
"en",
"dataset:EleutherAI/pile",
"arxiv:2101.00027",
"arxiv:2201.07311",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-01-17T03:08:51Z |
2024-04-17T03:50:57+00:00
| 75 | 12 |
---
datasets:
- EleutherAI/pile
language:
- en
pipeline_tag: text2text-generation
tags:
- t5x
- encoder-decoder
---
Pile-T5 XL is an Encoder-Decoder model trained on [the Pile](https://pile.eleuther.ai/) using the [T5x](https://github.com/google-research/t5x) library. The model was trained for 2 million steps or roughly 2 trillion tokens using MLM-objective similar to the original T5 model.
The HF version of Pile-T5 XL borrows UMT5's model implementation as it uses scalable model implementation from T5x and uses `LlamaTokenizer`.
### Model Details
- Developed by: [EleutherAI](http://eleuther.ai)
- Model type: Transformer-based Language Model
- Language: English
- Learn more: [Blogpost](). For details about the training dataset,
see [the Pile paper](https://arxiv.org/abs/2101.00027), and [its data
sheet](https://arxiv.org/abs/2201.07311).
- License: Apache 2.0
- Contact: to ask questions about this model, join the [EleutherAI
Discord](https://discord.gg/zBGx3azzUn), and post them in `#release-discussion`.
Please read the existing GPT-NeoX-20B documentation before asking about the model
on Discord. For general correspondence: [contact@eleuther.
ai](mailto:[email protected]).
<figure style="width:30em">
| Hyperparameter | Value |
| -------------------------- | ----------- |
| n<sub>parameters</sub> | 2849804288 |
| n<sub>encoder layers</sub> | 24 |
| n<sub>decoder layers</sub> | 24 |
| d<sub>model</sub> | 5120 |
| d<sub>emb</sub> | 2048 |
| n<sub>heads</sub> | 32 |
| d<sub>head</sub> | 64 |
| n<sub>vocab</sub> | 32128 |
| Sequence Length | 512 |
</figure>
### Uses and limitations
#### Intended use
Pile-T5 was developed primarily for research purposes. It learns an inner
representation of the English language that can be used to extract features
useful for downstream tasks.
In addition to scientific uses, you may also further fine-tune and adapt
Pile-T5 for deployment, as long as your use is in accordance with the
Apache 2.0 license. This model works with the [Transformers
Library](https://huggingface.co/docs/transformers/index). If you decide to use
pre-trained Pile-T5 as a basis for your fine-tuned model, please note that
you need to conduct your own risk and bias assessment.
#### Out-of-scope use
Pile-T5 is **not** intended for deployment as-is. It is not a product
and cannot be used for human-facing interactions without supervision.
Pile-T5 has not been fine-tuned for downstream tasks for which language
models are commonly deployed, such as writing genre prose, or commercial
chatbots. This means Pile-T5 will likely **not** respond to a given prompt
the way products such as ChatGPT do. This is because, unlike Pile-T5,
ChatGPT was fine-tuned using methods such as Reinforcement Learning from Human
Feedback (RLHF) to better “understand” human instructions and dialogue.
This model is English-language only, and thus cannot be used for translation
or generating text in other languages.
#### Limitations and biases
The core functionality of Pile-T5 is to take a string of text that has been
partially replaced with mask tokens and predict a sequence of tokens that would
replace those mask tokens. Remember that the statistically most likely sequence
of tokens need not result in the most “accurate” text. Never rely on Pile-T5 to produce
factually accurate output.
This model was trained on [the Pile](https://pile.eleuther.ai/), a dataset
known to contain profanity and texts that are lewd or otherwise offensive.
See [Section 6 of the Pile paper](https://arxiv.org/abs/2101.00027) for a
discussion of documented biases with regards to gender, religion, and race.
Pile-T5 may produce socially unacceptable or undesirable text, *even if*
the prompt itself does not include anything explicitly offensive.
We recommend curating the outputs of this model before presenting it to a human
reader. Please inform your audience that you are using artificially generated
text.
#### How to use
Pile-T5 can be loaded using the `AutoModelForSeq2SeqLM` functionality:
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/pile-t5-xl")
model = AutoModelForSeq2SeqLM.from_pretrained("EleutherAI/pile-t5-xl")
```
### Training
#### Training dataset
The Pile is a 825GiB general-purpose dataset in English. It was created by
EleutherAI specifically for training large language models. It contains texts
from 22 diverse sources, roughly broken down into five categories: academic
writing (e.g. arXiv), internet (e.g. CommonCrawl), prose (e.g. Project
Gutenberg), dialogue (e.g. YouTube subtitles), and miscellaneous (e.g. GitHub,
Enron Emails). See [the Pile paper](https://arxiv.org/abs/2101.00027) for
a breakdown of all data sources, methodology, and a discussion of ethical
implications. Consult [the datasheet](https://arxiv.org/abs/2201.07311) for
more detailed documentation about the Pile and its component datasets. The
Pile can be downloaded from the [official website](https://pile.eleuther.ai/),
or from a [community mirror](https://the-eye.eu/public/AI/pile/).
The Pile was deduplicated before being used to train Pile-T5.
#### Training procedure
Pile-T5 was trained with a batch size of approximately 1M tokens
(2048 sequences of 512 tokens each), for a total of 2,000,000 steps. Pile-T5 was trained
with the span-corruption objective.
#### Training checkpoints
Intermediate checkpoints for Pile-T5 are accessible within this repository.
There are in total 200 checkpoints that are spaced 10,000 steps. For T5x-native
checkpoints that can be used for finetuning with the T5x library, refer to [here](https://huggingface.co/lintang/pile-t5-xl-t5x)
The training loss (in tfevent format) and validation perplexity (in jsonl) can be found [here](https://huggingface.co/EleutherAI/pile-t5-xl/blob/main/xl.zip).
### Evaluations
Pile-T5 XL was evaluated on SuperGLUE, CodeXGLUE. A Flan-finetuned version was evaluated on Flan Held In tasks, MMLU and BBH.
Results can be seen in the [blogpost](https://blog.eleuther.ai/pile-t5/)
### BibTeX
```
@misc{2024PileT5,
author = {Lintang Sutawika and Aran Komatsuzaki and Colin Raffel},
title = {Pile-T5},
year = {2024},
url = {https://blog.eleuther.ai/pile-t5/},
note = {Blog post},
}
```
| null |
Non_BioNLP
|
Pile-T5 XL is an Encoder-Decoder model trained on [the Pile](https://pile.eleuther.ai/) using the [T5x](https://github.com/google-research/t5x) library. The model was trained for 2 million steps or roughly 2 trillion tokens using MLM-objective similar to the original T5 model.
The HF version of Pile-T5 XL borrows UMT5's model implementation as it uses scalable model implementation from T5x and uses `LlamaTokenizer`.
### Model Details
- Developed by: [EleutherAI](http://eleuther.ai)
- Model type: Transformer-based Language Model
- Language: English
- Learn more: [Blogpost](). For details about the training dataset,
see [the Pile paper](https://arxiv.org/abs/2101.00027), and [its data
sheet](https://arxiv.org/abs/2201.07311).
- License: Apache 2.0
- Contact: to ask questions about this model, join the [EleutherAI
Discord](https://discord.gg/zBGx3azzUn), and post them in `#release-discussion`.
Please read the existing GPT-NeoX-20B documentation before asking about the model
on Discord. For general correspondence: [contact@eleuther.
ai](mailto:[email protected]).
<figure style="width:30em">
| Hyperparameter | Value |
| -------------------------- | ----------- |
| n<sub>parameters</sub> | 2849804288 |
| n<sub>encoder layers</sub> | 24 |
| n<sub>decoder layers</sub> | 24 |
| d<sub>model</sub> | 5120 |
| d<sub>emb</sub> | 2048 |
| n<sub>heads</sub> | 32 |
| d<sub>head</sub> | 64 |
| n<sub>vocab</sub> | 32128 |
| Sequence Length | 512 |
</figure>
### Uses and limitations
#### Intended use
Pile-T5 was developed primarily for research purposes. It learns an inner
representation of the English language that can be used to extract features
useful for downstream tasks.
In addition to scientific uses, you may also further fine-tune and adapt
Pile-T5 for deployment, as long as your use is in accordance with the
Apache 2.0 license. This model works with the [Transformers
Library](https://huggingface.co/docs/transformers/index). If you decide to use
pre-trained Pile-T5 as a basis for your fine-tuned model, please note that
you need to conduct your own risk and bias assessment.
#### Out-of-scope use
Pile-T5 is **not** intended for deployment as-is. It is not a product
and cannot be used for human-facing interactions without supervision.
Pile-T5 has not been fine-tuned for downstream tasks for which language
models are commonly deployed, such as writing genre prose, or commercial
chatbots. This means Pile-T5 will likely **not** respond to a given prompt
the way products such as ChatGPT do. This is because, unlike Pile-T5,
ChatGPT was fine-tuned using methods such as Reinforcement Learning from Human
Feedback (RLHF) to better “understand” human instructions and dialogue.
This model is English-language only, and thus cannot be used for translation
or generating text in other languages.
#### Limitations and biases
The core functionality of Pile-T5 is to take a string of text that has been
partially replaced with mask tokens and predict a sequence of tokens that would
replace those mask tokens. Remember that the statistically most likely sequence
of tokens need not result in the most “accurate” text. Never rely on Pile-T5 to produce
factually accurate output.
This model was trained on [the Pile](https://pile.eleuther.ai/), a dataset
known to contain profanity and texts that are lewd or otherwise offensive.
See [Section 6 of the Pile paper](https://arxiv.org/abs/2101.00027) for a
discussion of documented biases with regards to gender, religion, and race.
Pile-T5 may produce socially unacceptable or undesirable text, *even if*
the prompt itself does not include anything explicitly offensive.
We recommend curating the outputs of this model before presenting it to a human
reader. Please inform your audience that you are using artificially generated
text.
#### How to use
Pile-T5 can be loaded using the `AutoModelForSeq2SeqLM` functionality:
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/pile-t5-xl")
model = AutoModelForSeq2SeqLM.from_pretrained("EleutherAI/pile-t5-xl")
```
### Training
#### Training dataset
The Pile is a 825GiB general-purpose dataset in English. It was created by
EleutherAI specifically for training large language models. It contains texts
from 22 diverse sources, roughly broken down into five categories: academic
writing (e.g. arXiv), internet (e.g. CommonCrawl), prose (e.g. Project
Gutenberg), dialogue (e.g. YouTube subtitles), and miscellaneous (e.g. GitHub,
Enron Emails). See [the Pile paper](https://arxiv.org/abs/2101.00027) for
a breakdown of all data sources, methodology, and a discussion of ethical
implications. Consult [the datasheet](https://arxiv.org/abs/2201.07311) for
more detailed documentation about the Pile and its component datasets. The
Pile can be downloaded from the [official website](https://pile.eleuther.ai/),
or from a [community mirror](https://the-eye.eu/public/AI/pile/).
The Pile was deduplicated before being used to train Pile-T5.
#### Training procedure
Pile-T5 was trained with a batch size of approximately 1M tokens
(2048 sequences of 512 tokens each), for a total of 2,000,000 steps. Pile-T5 was trained
with the span-corruption objective.
#### Training checkpoints
Intermediate checkpoints for Pile-T5 are accessible within this repository.
There are in total 200 checkpoints that are spaced 10,000 steps. For T5x-native
checkpoints that can be used for finetuning with the T5x library, refer to [here](https://huggingface.co/lintang/pile-t5-xl-t5x)
The training loss (in tfevent format) and validation perplexity (in jsonl) can be found [here](https://huggingface.co/EleutherAI/pile-t5-xl/blob/main/xl.zip).
### Evaluations
Pile-T5 XL was evaluated on SuperGLUE, CodeXGLUE. A Flan-finetuned version was evaluated on Flan Held In tasks, MMLU and BBH.
Results can be seen in the [blogpost](https://blog.eleuther.ai/pile-t5/)
### BibTeX
```
@misc{2024PileT5,
author = {Lintang Sutawika and Aran Komatsuzaki and Colin Raffel},
title = {Pile-T5},
year = {2024},
url = {https://blog.eleuther.ai/pile-t5/},
note = {Blog post},
}
```
|
{"datasets": ["EleutherAI/pile"], "language": ["en"], "pipeline_tag": "text2text-generation", "tags": ["t5x", "encoder-decoder"]}
|
task
|
[
"TRANSLATION"
] | 39,831 |
myrkur/persian-question-generator
|
myrkur
|
summarization
|
[
"transformers",
"safetensors",
"mt5",
"text2text-generation",
"summarization",
"fa",
"base_model:google/mt5-base",
"base_model:finetune:google/mt5-base",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-12-08T06:17:27Z |
2024-12-10T12:19:08+00:00
| 40 | 1 |
---
base_model:
- google/mt5-base
language:
- fa
library_name: transformers
license: apache-2.0
pipeline_tag: summarization
---
# mT5-Based Persian Question Generator
This repository contains model for generating questions from Persian text.
---
## Features
- Preprocess Persian text data to ensure high-quality training samples.
- Fine-tune the mT5 model on question-generation tasks.
- Evaluate and generate questions from Persian texts using the fine-tuned model.
---
## Generating Questions
Use the fine-tuned model to generate questions from Persian text:
```python
from transformers import pipeline
pipe = pipeline("summarization", model="myrkur/persian-question-generator", device_map="auto")
sample_text = """شبکههای اجتماعی،هوموفیلی و اگزیستانسیالیم..."""
generated_question = pipe([sample_text], temperature=0.3, do_sample=True, repetition_penalty=1.1)
print(generated_question)
```
---
## Results and Usage
The fine-tuned model generates natural and contextually relevant questions from Persian text. This can be utilized for:
- Educational tools
- Conversational AI
- Persian text comprehension applications
| null |
Non_BioNLP
|
# mT5-Based Persian Question Generator
This repository contains model for generating questions from Persian text.
---
## Features
- Preprocess Persian text data to ensure high-quality training samples.
- Fine-tune the mT5 model on question-generation tasks.
- Evaluate and generate questions from Persian texts using the fine-tuned model.
---
## Generating Questions
Use the fine-tuned model to generate questions from Persian text:
```python
from transformers import pipeline
pipe = pipeline("summarization", model="myrkur/persian-question-generator", device_map="auto")
sample_text = """شبکههای اجتماعی،هوموفیلی و اگزیستانسیالیم..."""
generated_question = pipe([sample_text], temperature=0.3, do_sample=True, repetition_penalty=1.1)
print(generated_question)
```
---
## Results and Usage
The fine-tuned model generates natural and contextually relevant questions from Persian text. This can be utilized for:
- Educational tools
- Conversational AI
- Persian text comprehension applications
|
{"base_model": ["google/mt5-base"], "language": ["fa"], "library_name": "transformers", "license": "apache-2.0", "pipeline_tag": "summarization"}
|
task
|
[
"SUMMARIZATION"
] | 39,832 |
RichardErkhov/facebook_-_bart-large-cnn-4bits
|
RichardErkhov
|
text-generation
|
[
"transformers",
"safetensors",
"bart",
"text-generation",
"arxiv:1910.13461",
"autotrain_compatible",
"endpoints_compatible",
"4-bit",
"bitsandbytes",
"region:us"
] | 2024-05-09T18:00:56Z |
2024-05-09T18:01:23+00:00
| 14 | 0 |
---
{}
---
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
bart-large-cnn - bnb 4bits
- Model creator: https://huggingface.co/facebook/
- Original model: https://huggingface.co/facebook/bart-large-cnn/
Original model description:
---
language:
- en
pipeline_tag: summarization
license: mit
thumbnail: https://huggingface.co/front/thumbnails/facebook.png
datasets:
- cnn_dailymail
model-index:
- name: facebook/bart-large-cnn
results:
- task:
type: summarization
name: Summarization
dataset:
name: cnn_dailymail
type: cnn_dailymail
config: 3.0.0
split: train
metrics:
- name: ROUGE-1
type: rouge
value: 42.9486
verified: true
- name: ROUGE-2
type: rouge
value: 20.8149
verified: true
- name: ROUGE-L
type: rouge
value: 30.6186
verified: true
- name: ROUGE-LSUM
type: rouge
value: 40.0376
verified: true
- name: loss
type: loss
value: 2.529000997543335
verified: true
- name: gen_len
type: gen_len
value: 78.5866
verified: true
---
# BART (large-sized model), fine-tuned on CNN Daily Mail
BART model pre-trained on English language, and fine-tuned on [CNN Daily Mail](https://huggingface.co/datasets/cnn_dailymail). It was introduced in the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Lewis et al. and first released in [this repository (https://github.com/pytorch/fairseq/tree/master/examples/bart).
Disclaimer: The team releasing BART did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
BART is a transformer encoder-encoder (seq2seq) model with a bidirectional (BERT-like) encoder and an autoregressive (GPT-like) decoder. BART is pre-trained by (1) corrupting text with an arbitrary noising function, and (2) learning a model to reconstruct the original text.
BART is particularly effective when fine-tuned for text generation (e.g. summarization, translation) but also works well for comprehension tasks (e.g. text classification, question answering). This particular checkpoint has been fine-tuned on CNN Daily Mail, a large collection of text-summary pairs.
## Intended uses & limitations
You can use this model for text summarization.
### How to use
Here is how to use this model with the [pipeline API](https://huggingface.co/transformers/main_classes/pipelines.html):
```python
from transformers import pipeline
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
ARTICLE = """ New York (CNN)When Liana Barrientos was 23 years old, she got married in Westchester County, New York.
A year later, she got married again in Westchester County, but to a different man and without divorcing her first husband.
Only 18 days after that marriage, she got hitched yet again. Then, Barrientos declared "I do" five more times, sometimes only within two weeks of each other.
In 2010, she married once more, this time in the Bronx. In an application for a marriage license, she stated it was her "first and only" marriage.
Barrientos, now 39, is facing two criminal counts of "offering a false instrument for filing in the first degree," referring to her false statements on the
2010 marriage license application, according to court documents.
Prosecutors said the marriages were part of an immigration scam.
On Friday, she pleaded not guilty at State Supreme Court in the Bronx, according to her attorney, Christopher Wright, who declined to comment further.
After leaving court, Barrientos was arrested and charged with theft of service and criminal trespass for allegedly sneaking into the New York subway through an emergency exit, said Detective
Annette Markowski, a police spokeswoman. In total, Barrientos has been married 10 times, with nine of her marriages occurring between 1999 and 2002.
All occurred either in Westchester County, Long Island, New Jersey or the Bronx. She is believed to still be married to four men, and at one time, she was married to eight men at once, prosecutors say.
Prosecutors said the immigration scam involved some of her husbands, who filed for permanent residence status shortly after the marriages.
Any divorces happened only after such filings were approved. It was unclear whether any of the men will be prosecuted.
The case was referred to the Bronx District Attorney\'s Office by Immigration and Customs Enforcement and the Department of Homeland Security\'s
Investigation Division. Seven of the men are from so-called "red-flagged" countries, including Egypt, Turkey, Georgia, Pakistan and Mali.
Her eighth husband, Rashid Rajput, was deported in 2006 to his native Pakistan after an investigation by the Joint Terrorism Task Force.
If convicted, Barrientos faces up to four years in prison. Her next court appearance is scheduled for May 18.
"""
print(summarizer(ARTICLE, max_length=130, min_length=30, do_sample=False))
>>> [{'summary_text': 'Liana Barrientos, 39, is charged with two counts of "offering a false instrument for filing in the first degree" In total, she has been married 10 times, with nine of her marriages occurring between 1999 and 2002. She is believed to still be married to four men.'}]
```
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-1910-13461,
author = {Mike Lewis and
Yinhan Liu and
Naman Goyal and
Marjan Ghazvininejad and
Abdelrahman Mohamed and
Omer Levy and
Veselin Stoyanov and
Luke Zettlemoyer},
title = {{BART:} Denoising Sequence-to-Sequence Pre-training for Natural Language
Generation, Translation, and Comprehension},
journal = {CoRR},
volume = {abs/1910.13461},
year = {2019},
url = {http://arxiv.org/abs/1910.13461},
eprinttype = {arXiv},
eprint = {1910.13461},
timestamp = {Thu, 31 Oct 2019 14:02:26 +0100},
biburl = {https://dblp.org/rec/journals/corr/abs-1910-13461.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
| null |
Non_BioNLP
|
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
bart-large-cnn - bnb 4bits
- Model creator: https://huggingface.co/facebook/
- Original model: https://huggingface.co/facebook/bart-large-cnn/
Original model description:
---
language:
- en
pipeline_tag: summarization
license: mit
thumbnail: https://huggingface.co/front/thumbnails/facebook.png
datasets:
- cnn_dailymail
model-index:
- name: facebook/bart-large-cnn
results:
- task:
type: summarization
name: Summarization
dataset:
name: cnn_dailymail
type: cnn_dailymail
config: 3.0.0
split: train
metrics:
- name: ROUGE-1
type: rouge
value: 42.9486
verified: true
- name: ROUGE-2
type: rouge
value: 20.8149
verified: true
- name: ROUGE-L
type: rouge
value: 30.6186
verified: true
- name: ROUGE-LSUM
type: rouge
value: 40.0376
verified: true
- name: loss
type: loss
value: 2.529000997543335
verified: true
- name: gen_len
type: gen_len
value: 78.5866
verified: true
---
# BART (large-sized model), fine-tuned on CNN Daily Mail
BART model pre-trained on English language, and fine-tuned on [CNN Daily Mail](https://huggingface.co/datasets/cnn_dailymail). It was introduced in the paper [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://arxiv.org/abs/1910.13461) by Lewis et al. and first released in [this repository (https://github.com/pytorch/fairseq/tree/master/examples/bart).
Disclaimer: The team releasing BART did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
BART is a transformer encoder-encoder (seq2seq) model with a bidirectional (BERT-like) encoder and an autoregressive (GPT-like) decoder. BART is pre-trained by (1) corrupting text with an arbitrary noising function, and (2) learning a model to reconstruct the original text.
BART is particularly effective when fine-tuned for text generation (e.g. summarization, translation) but also works well for comprehension tasks (e.g. text classification, question answering). This particular checkpoint has been fine-tuned on CNN Daily Mail, a large collection of text-summary pairs.
## Intended uses & limitations
You can use this model for text summarization.
### How to use
Here is how to use this model with the [pipeline API](https://huggingface.co/transformers/main_classes/pipelines.html):
```python
from transformers import pipeline
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
ARTICLE = """ New York (CNN)When Liana Barrientos was 23 years old, she got married in Westchester County, New York.
A year later, she got married again in Westchester County, but to a different man and without divorcing her first husband.
Only 18 days after that marriage, she got hitched yet again. Then, Barrientos declared "I do" five more times, sometimes only within two weeks of each other.
In 2010, she married once more, this time in the Bronx. In an application for a marriage license, she stated it was her "first and only" marriage.
Barrientos, now 39, is facing two criminal counts of "offering a false instrument for filing in the first degree," referring to her false statements on the
2010 marriage license application, according to court documents.
Prosecutors said the marriages were part of an immigration scam.
On Friday, she pleaded not guilty at State Supreme Court in the Bronx, according to her attorney, Christopher Wright, who declined to comment further.
After leaving court, Barrientos was arrested and charged with theft of service and criminal trespass for allegedly sneaking into the New York subway through an emergency exit, said Detective
Annette Markowski, a police spokeswoman. In total, Barrientos has been married 10 times, with nine of her marriages occurring between 1999 and 2002.
All occurred either in Westchester County, Long Island, New Jersey or the Bronx. She is believed to still be married to four men, and at one time, she was married to eight men at once, prosecutors say.
Prosecutors said the immigration scam involved some of her husbands, who filed for permanent residence status shortly after the marriages.
Any divorces happened only after such filings were approved. It was unclear whether any of the men will be prosecuted.
The case was referred to the Bronx District Attorney\'s Office by Immigration and Customs Enforcement and the Department of Homeland Security\'s
Investigation Division. Seven of the men are from so-called "red-flagged" countries, including Egypt, Turkey, Georgia, Pakistan and Mali.
Her eighth husband, Rashid Rajput, was deported in 2006 to his native Pakistan after an investigation by the Joint Terrorism Task Force.
If convicted, Barrientos faces up to four years in prison. Her next court appearance is scheduled for May 18.
"""
print(summarizer(ARTICLE, max_length=130, min_length=30, do_sample=False))
>>> [{'summary_text': 'Liana Barrientos, 39, is charged with two counts of "offering a false instrument for filing in the first degree" In total, she has been married 10 times, with nine of her marriages occurring between 1999 and 2002. She is believed to still be married to four men.'}]
```
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-1910-13461,
author = {Mike Lewis and
Yinhan Liu and
Naman Goyal and
Marjan Ghazvininejad and
Abdelrahman Mohamed and
Omer Levy and
Veselin Stoyanov and
Luke Zettlemoyer},
title = {{BART:} Denoising Sequence-to-Sequence Pre-training for Natural Language
Generation, Translation, and Comprehension},
journal = {CoRR},
volume = {abs/1910.13461},
year = {2019},
url = {http://arxiv.org/abs/1910.13461},
eprinttype = {arXiv},
eprint = {1910.13461},
timestamp = {Thu, 31 Oct 2019 14:02:26 +0100},
biburl = {https://dblp.org/rec/journals/corr/abs-1910-13461.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
|
{}
|
task
|
[
"TEXT_CLASSIFICATION",
"QUESTION_ANSWERING",
"TRANSLATION",
"SUMMARIZATION"
] | 39,834 |
Livyatan/mT5-small-Hebrew-ParaShoot-QA
|
Livyatan
|
text2text-generation
|
[
"transformers",
"pytorch",
"mt5",
"text2text-generation",
"arxiv:2109.11314",
"arxiv:2010.11934",
"arxiv:2212.09682",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-08-19T14:04:30Z |
2023-08-19T15:06:21+00:00
| 24 | 0 |
---
{}
---
---
datasets: https://github.com/omrikeren/ParaShoot/
metrics: f1 49.612 exact_match 26.439
language: he
pipeline_tag: question-answering
license: unknown
---
## mT5-small-Hebrew-ParaShoot-QA
This repository contains a fine-tuned mT5-small (Multilingual Text-to-Text Transfer Transformer) model on the [ParaShoot](https://arxiv.org/pdf/2109.11314.pdf) dataset ([github](https://github.com/omrikeren/ParaShoot/tree/main/data)).
To enhance its performance, a "domain-specific" fine-tuning approach was employed.
Initially, the model was pretrained on a Hebrew dataset to capture Hebrew linguistic nuances.
Subsequently, I further fine-tuned the model on the ParaShoot dataset, aiming to improve its proficiency in the Question-Answering task.
This model builds upon the original work by [imvladikon](https://huggingface.co/imvladikon/) who initially fine-tuned the mT5-small model for the [summarization task](https://huggingface.co/imvladikon/het5_small_summarization).
# Model Details
[Google's mT5](https://github.com/google-research/multilingual-t5)
mT5 is pretrained on the mC4 corpus, covering 101 languages.
Note: mT5 was only pre-trained on mC4 excluding any supervised training. Therefore, this model has to be fine-tuned before it is useable on a downstream task.
# Related papers:
Paper: [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934)
Authors: Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel
Paper: [Multilingual Sequence-to-Sequence Models for Hebrew NLP](https://arxiv.org/pdf/2212.09682.pdf)
Authors: Matan Eyal, Hila Noga, Roee Aharoni, Idan Szpektor, Reut Tsarfaty
Paper: [PARASHOOT: A Hebrew Question Answering Dataset](https://arxiv.org/pdf/2109.11314.pdf)
Authors: Omri Keren, Omer Levy
# This model achieves the following results on the test set:
- Overall F1: 49.612
- Overall EM: 26.439
- Loss: 1.346
Note: In the paper [Multilingual Sequence-to-Sequence Models for Hebrew NLP](https://arxiv.org/pdf/2212.09682.pdf) the results were F1 - 48.71, EM - 24.52.
# How to use the model:
Use the code below to get started with the model.
```python
from transformers import MT5ForConditionalGeneration, AutoTokenizer
MODEL_NAME = "Livyatan/mT5-small-Hebrew-ParaShoot-QA"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = MT5ForConditionalGeneration.from_pretrained(MODEL_NAME)
```
```python
def generate_answer(question, context):
input_encoding = tokenizer(
question,
context,
max_length = len(context),
padding="max_length",
truncation="only_second",
return_attention_mask=True,
add_special_tokens=True,
return_tensors="pt"
).to(DEVICE)
with torch.no_grad():
generated_ids = model.generate(
input_ids = input_encoding['input_ids'].to(DEVICE),
attention_mask = input_encoding['attention_mask'].to(DEVICE),
max_length=20,
)
preds = [
tokenizer.decode(generated_id, skip_special_tokens=True, clean_up_tokenization_spaces=True)
for generated_id in generated_ids
]
return "".join(preds)
context = 'סדרת הלווייתנאים כוללת כ-90 מינים, שכולם חיים באוקיינוסים מלבד חמישה מיני דולפינים החיים במים מתוקים. הלווייתנאים החיים מחולקים לשתי תת-סדרות: לווייתני מזיפות (Mysticeti) ולווייתני שיניים (Odontoceti; ובהם גם דולפינים); בעבר התקיימה תת-סדרה נוספת: לווייתנים קדומים (Archaeoceti), שנכחדה. במרבית המקרים לווייתני המזיפות גדולים באופן משמעותי מלווייתני השיניים, הקטנים והמהירים יותר, וכמה מלווייתני המזיפות הם מבעלי החיים הגדולים ביותר בכדור הארץ. לווייתני השיניים מתאפיינים בשיניים חדות, והם ציידים מהירים שניזונים מדגים ומיצורים ימיים אחרים. לעומתם לווייתני המזיפות הם חסרי שיניים ובמקום זאת יש להם מזיפות ארוכות דמויות מסננת, שבעזרתן הם מסננים פלנקטון מהמים.'
question = 'כמה מינים כוללת סדרת הלווייתנאים?'
answer = generate_answer(question, context)
print(answer)
>>> 'כ-90 מינים'
```
| null |
Non_BioNLP
|
---
datasets: https://github.com/omrikeren/ParaShoot/
metrics: f1 49.612 exact_match 26.439
language: he
pipeline_tag: question-answering
license: unknown
---
## mT5-small-Hebrew-ParaShoot-QA
This repository contains a fine-tuned mT5-small (Multilingual Text-to-Text Transfer Transformer) model on the [ParaShoot](https://arxiv.org/pdf/2109.11314.pdf) dataset ([github](https://github.com/omrikeren/ParaShoot/tree/main/data)).
To enhance its performance, a "domain-specific" fine-tuning approach was employed.
Initially, the model was pretrained on a Hebrew dataset to capture Hebrew linguistic nuances.
Subsequently, I further fine-tuned the model on the ParaShoot dataset, aiming to improve its proficiency in the Question-Answering task.
This model builds upon the original work by [imvladikon](https://huggingface.co/imvladikon/) who initially fine-tuned the mT5-small model for the [summarization task](https://huggingface.co/imvladikon/het5_small_summarization).
# Model Details
[Google's mT5](https://github.com/google-research/multilingual-t5)
mT5 is pretrained on the mC4 corpus, covering 101 languages.
Note: mT5 was only pre-trained on mC4 excluding any supervised training. Therefore, this model has to be fine-tuned before it is useable on a downstream task.
# Related papers:
Paper: [mT5: A massively multilingual pre-trained text-to-text transformer](https://arxiv.org/abs/2010.11934)
Authors: Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, Colin Raffel
Paper: [Multilingual Sequence-to-Sequence Models for Hebrew NLP](https://arxiv.org/pdf/2212.09682.pdf)
Authors: Matan Eyal, Hila Noga, Roee Aharoni, Idan Szpektor, Reut Tsarfaty
Paper: [PARASHOOT: A Hebrew Question Answering Dataset](https://arxiv.org/pdf/2109.11314.pdf)
Authors: Omri Keren, Omer Levy
# This model achieves the following results on the test set:
- Overall F1: 49.612
- Overall EM: 26.439
- Loss: 1.346
Note: In the paper [Multilingual Sequence-to-Sequence Models for Hebrew NLP](https://arxiv.org/pdf/2212.09682.pdf) the results were F1 - 48.71, EM - 24.52.
# How to use the model:
Use the code below to get started with the model.
```python
from transformers import MT5ForConditionalGeneration, AutoTokenizer
MODEL_NAME = "Livyatan/mT5-small-Hebrew-ParaShoot-QA"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = MT5ForConditionalGeneration.from_pretrained(MODEL_NAME)
```
```python
def generate_answer(question, context):
input_encoding = tokenizer(
question,
context,
max_length = len(context),
padding="max_length",
truncation="only_second",
return_attention_mask=True,
add_special_tokens=True,
return_tensors="pt"
).to(DEVICE)
with torch.no_grad():
generated_ids = model.generate(
input_ids = input_encoding['input_ids'].to(DEVICE),
attention_mask = input_encoding['attention_mask'].to(DEVICE),
max_length=20,
)
preds = [
tokenizer.decode(generated_id, skip_special_tokens=True, clean_up_tokenization_spaces=True)
for generated_id in generated_ids
]
return "".join(preds)
context = 'סדרת הלווייתנאים כוללת כ-90 מינים, שכולם חיים באוקיינוסים מלבד חמישה מיני דולפינים החיים במים מתוקים. הלווייתנאים החיים מחולקים לשתי תת-סדרות: לווייתני מזיפות (Mysticeti) ולווייתני שיניים (Odontoceti; ובהם גם דולפינים); בעבר התקיימה תת-סדרה נוספת: לווייתנים קדומים (Archaeoceti), שנכחדה. במרבית המקרים לווייתני המזיפות גדולים באופן משמעותי מלווייתני השיניים, הקטנים והמהירים יותר, וכמה מלווייתני המזיפות הם מבעלי החיים הגדולים ביותר בכדור הארץ. לווייתני השיניים מתאפיינים בשיניים חדות, והם ציידים מהירים שניזונים מדגים ומיצורים ימיים אחרים. לעומתם לווייתני המזיפות הם חסרי שיניים ובמקום זאת יש להם מזיפות ארוכות דמויות מסננת, שבעזרתן הם מסננים פלנקטון מהמים.'
question = 'כמה מינים כוללת סדרת הלווייתנאים?'
answer = generate_answer(question, context)
print(answer)
>>> 'כ-90 מינים'
```
|
{}
|
task
|
[
"QUESTION_ANSWERING",
"SUMMARIZATION"
] | 39,835 |
deepali1021/legal-ft-v0
|
deepali1021
|
sentence-similarity
|
[
"sentence-transformers",
"safetensors",
"bert",
"sentence-similarity",
"feature-extraction",
"generated_from_trainer",
"dataset_size:156",
"loss:MatryoshkaLoss",
"loss:MultipleNegativesRankingLoss",
"arxiv:1908.10084",
"arxiv:2205.13147",
"arxiv:1705.00652",
"base_model:Snowflake/snowflake-arctic-embed-l",
"base_model:finetune:Snowflake/snowflake-arctic-embed-l",
"model-index",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] | 2025-02-18T00:46:47Z |
2025-02-18T00:47:32+00:00
| 8 | 0 |
---
base_model: Snowflake/snowflake-arctic-embed-l
library_name: sentence-transformers
metrics:
- cosine_accuracy@1
- cosine_accuracy@3
- cosine_accuracy@5
- cosine_accuracy@10
- cosine_precision@1
- cosine_precision@3
- cosine_precision@5
- cosine_precision@10
- cosine_recall@1
- cosine_recall@3
- cosine_recall@5
- cosine_recall@10
- cosine_ndcg@10
- cosine_mrr@10
- cosine_map@100
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:156
- loss:MatryoshkaLoss
- loss:MultipleNegativesRankingLoss
widget:
- source_sentence: What new shape of LLM was introduced in the final quarter of 2024,
and what were the names of the initial models released?
sentences:
- '17th: AI for Data Journalism: demonstrating what we can do with this stuff right
now
22nd: Options for accessing Llama 3 from the terminal using LLM
May
8th: Slop is the new name for unwanted AI-generated content
15th: ChatGPT in “4o” mode is not running the new features yet
29th: Training is not the same as chatting: ChatGPT and other LLMs don’t remember
everything you say
June
6th: Accidental prompt injection against RAG applications
10th: Thoughts on the WWDC 2024 keynote on Apple Intelligence
17th: Language models on the command-line
21st: Building search-based RAG using Claude, Datasette and Val Town
27th: Open challenges for AI engineering
July
14th: Imitation Intelligence, my keynote for PyCon US 2024'
- 'Now that those features are rolling out they’re pretty weak. As an LLM power-user
I know what these models are capable of, and Apple’s LLM features offer a pale
imitation of what a frontier LLM can do. Instead we’re getting notification summaries
that misrepresent news headlines and writing assistant tools that I’ve not found
useful at all. Genmoji are kind of fun though.
The rise of inference-scaling “reasoning” models
The most interesting development in the final quarter of 2024 was the introduction
of a new shape of LLM, exemplified by OpenAI’s o1 models—initially released as
o1-preview and o1-mini on September 12th.'
- 'Terminology aside, I remain skeptical as to their utility based, once again,
on the challenge of gullibility. LLMs believe anything you tell them. Any systems
that attempts to make meaningful decisions on your behalf will run into the same
roadblock: how good is a travel agent, or a digital assistant, or even a research
tool if it can’t distinguish truth from fiction?
Just the other day Google Search was caught serving up an entirely fake description
of the non-existant movie “Encanto 2”. It turned out to be summarizing an imagined
movie listing from a fan fiction wiki.'
- source_sentence: In what year does the author expect the prompt-driven custom interface
feature to be widely adopted in various products?
sentences:
- 'The models may have got more capable, but most of the limitations remained the
same. OpenAI’s o1 may finally be able to (mostly) count the Rs in strawberry,
but its abilities are still limited by its nature as an LLM and the constraints
placed on it by the harness it’s running in. o1 can’t run web searches or use
Code Interpreter, but GPT-4o can—both in that same ChatGPT UI. (o1 will pretend
to do those things if you ask it to, a regression to the URL hallucinations bug
from early 2023).
What are we doing about this? Not much. Most users are thrown in at the deep end.
The default LLM chat UI is like taking brand new computer users, dropping them
into a Linux terminal and expecting them to figure it all out.'
- 'This prompt-driven custom interface feature is so powerful and easy to build
(once you’ve figured out the gnarly details of browser sandboxing) that I expect
it to show up as a feature in a wide range of products in 2025.
Universal access to the best models lasted for just a few short months
For a few short months this year all three of the best available models—GPT-4o,
Claude 3.5 Sonnet and Gemini 1.5 Pro—were freely available to most of the world.'
- 'Against this photo of butterflies at the California Academy of Sciences:
A shallow dish, likely a hummingbird or butterfly feeder, is red. Pieces of orange
slices of fruit are visible inside the dish.
Two butterflies are positioned in the feeder, one is a dark brown/black butterfly
with white/cream-colored markings. The other is a large, brown butterfly with
patterns of lighter brown, beige, and black markings, including prominent eye
spots. The larger brown butterfly appears to be feeding on the fruit.'
- source_sentence: What is the license under which Alibaba's QwQ model was released?
sentences:
- The most recent twist, again from December (December was a lot) is live video.
ChatGPT voice mode now provides the option to share your camera feed with the
model and talk about what you can see in real time. Google Gemini have a preview
of the same feature, which they managed to ship the day before ChatGPT did.
- 'OpenAI are not the only game in town here. Google released their first entrant
in the category, gemini-2.0-flash-thinking-exp, on December 19th.
Alibaba’s Qwen team released their QwQ model on November 28th—under an Apache
2.0 license, and that one I could run on my own machine. They followed that up
with a vision reasoning model called QvQ on December 24th, which I also ran locally.
DeepSeek made their DeepSeek-R1-Lite-Preview model available to try out through
their chat interface on November 20th.
To understand more about inference scaling I recommend Is AI progress slowing
down? by Arvind Narayanan and Sayash Kapoor.'
- 'Stuff we figured out about AI in 2023
Simon Willison’s Weblog
Subscribe
Stuff we figured out about AI in 2023
31st December 2023
2023 was the breakthrough year for Large Language Models (LLMs). I think it’s
OK to call these AI—they’re the latest and (currently) most interesting development
in the academic field of Artificial Intelligence that dates back to the 1950s.
Here’s my attempt to round up the highlights in one place!'
- source_sentence: What is the cost of using 260 input tokens and 92 output tokens
as mentioned in the context?
sentences:
- 'I think people who complain that LLM improvement has slowed are often missing
the enormous advances in these multi-modal models. Being able to run prompts against
images (and audio and video) is a fascinating new way to apply these models.
Voice and live camera mode are science fiction come to life
The audio and live video modes that have started to emerge deserve a special mention.
The ability to talk to ChatGPT first arrived in September 2023, but it was mostly
an illusion: OpenAI used their excellent Whisper speech-to-text model and a new
text-to-speech model (creatively named tts-1) to enable conversations with the
ChatGPT mobile apps, but the actual model just saw text.'
- 'I like people who are skeptical of this stuff. The hype has been deafening for
more than two years now, and there are enormous quantities of snake oil and misinformation
out there. A lot of very bad decisions are being made based on that hype. Being
critical is a virtue.
If we want people with decision-making authority to make good decisions about
how to apply these tools we first need to acknowledge that there ARE good applications,
and then help explain how to put those into practice while avoiding the many unintiutive
traps.
(If you still don’t think there are any good applications at all I’m not sure
why you made it to this point in the article!)'
- '260 input tokens, 92 output tokens. Cost approximately 0.0024 cents (that’s less
than a 400th of a cent).
This increase in efficiency and reduction in price is my single favourite trend
from 2024. I want the utility of LLMs at a fraction of the energy cost and it
looks like that’s what we’re getting.
Multimodal vision is common, audio and video are starting to emerge
My butterfly example above illustrates another key trend from 2024: the rise of
multi-modal LLMs.
A year ago the single most notable example of these was GPT-4 Vision, released
at OpenAI’s DevDay in November 2023. Google’s multi-modal Gemini 1.0 was announced
on December 7th 2023 so it also (just) makes it into the 2023 window.'
- source_sentence: How does the author feel about their current experience as a Mac
user compared to the previous year?
sentences:
- 'I’m still trying to figure out the best patterns for doing this for my own work.
Everyone knows that evals are important, but there remains a lack of great guidance
for how to best implement them—I’m tracking this under my evals tag. My SVG pelican
riding a bicycle benchmark is a pale imitation of what a real eval suite should
look like.
Apple Intelligence is bad, Apple’s MLX library is excellent
As a Mac user I’ve been feeling a lot better about my choice of platform this
year.
Last year it felt like my lack of a Linux/Windows machine with an NVIDIA GPU
was a huge disadvantage in terms of trying out new models.'
- Structured and Gradual Learning. In organic datasets, the relationship between
tokens is often complex and indirect. Many reasoning steps may be required to
connect the current token to the next, making it challenging for the model to
learn effectively from next-token prediction. By contrast, each token generated
by a language model is by definition predicted by the preceding tokens, making
it easier for a model to follow the resulting reasoning patterns.
- 'The May 13th announcement of GPT-4o included a demo of a brand new voice mode,
where the true multi-modal GPT-4o (the o is for “omni”) model could accept audio
input and output incredibly realistic sounding speech without needing separate
TTS or STT models.
The demo also sounded conspicuously similar to Scarlett Johansson... and after
she complained the voice from the demo, Skye, never made it to a production product.
The delay in releasing the new voice mode after the initial demo caused quite
a lot of confusion. I wrote about that in ChatGPT in “4o” mode is not running
the new features yet.'
model-index:
- name: SentenceTransformer based on Snowflake/snowflake-arctic-embed-l
results:
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: Unknown
type: unknown
metrics:
- type: cosine_accuracy@1
value: 0.875
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 1.0
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 1.0
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 1.0
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.875
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.3333333333333333
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.20000000000000004
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.10000000000000002
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.875
name: Cosine Recall@1
- type: cosine_recall@3
value: 1.0
name: Cosine Recall@3
- type: cosine_recall@5
value: 1.0
name: Cosine Recall@5
- type: cosine_recall@10
value: 1.0
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.9538662191964322
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.9375
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.9375
name: Cosine Map@100
---
# SentenceTransformer based on Snowflake/snowflake-arctic-embed-l
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Snowflake/snowflake-arctic-embed-l](https://huggingface.co/Snowflake/snowflake-arctic-embed-l). It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [Snowflake/snowflake-arctic-embed-l](https://huggingface.co/Snowflake/snowflake-arctic-embed-l) <!-- at revision d8fb21ca8d905d2832ee8b96c894d3298964346b -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 1024 dimensions
- **Similarity Function:** Cosine Similarity
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
(2): Normalize()
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("deepali1021/legal-ft-v0")
# Run inference
sentences = [
'How does the author feel about their current experience as a Mac user compared to the previous year?',
'I’m still trying to figure out the best patterns for doing this for my own work. Everyone knows that evals are important, but there remains a lack of great guidance for how to best implement them—I’m tracking this under my evals tag. My SVG pelican riding a bicycle benchmark is a pale imitation of what a real eval suite should look like.\nApple Intelligence is bad, Apple’s MLX library is excellent\nAs a Mac user I’ve been feeling a lot better about my choice of platform this year.\nLast year it felt like my lack of a Linux/Windows machine with an NVIDIA GPU was a huge disadvantage in terms of trying out new models.',
'The May 13th announcement of GPT-4o included a demo of a brand new voice mode, where the true multi-modal GPT-4o (the o is for “omni”) model could accept audio input and output incredibly realistic sounding speech without needing separate TTS or STT models.\nThe demo also sounded conspicuously similar to Scarlett Johansson... and after she complained the voice from the demo, Skye, never made it to a production product.\nThe delay in releasing the new voice mode after the initial demo caused quite a lot of confusion. I wrote about that in ChatGPT in “4o” mode is not running the new features yet.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Information Retrieval
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | Value |
|:--------------------|:-----------|
| cosine_accuracy@1 | 0.875 |
| cosine_accuracy@3 | 1.0 |
| cosine_accuracy@5 | 1.0 |
| cosine_accuracy@10 | 1.0 |
| cosine_precision@1 | 0.875 |
| cosine_precision@3 | 0.3333 |
| cosine_precision@5 | 0.2 |
| cosine_precision@10 | 0.1 |
| cosine_recall@1 | 0.875 |
| cosine_recall@3 | 1.0 |
| cosine_recall@5 | 1.0 |
| cosine_recall@10 | 1.0 |
| **cosine_ndcg@10** | **0.9539** |
| cosine_mrr@10 | 0.9375 |
| cosine_map@100 | 0.9375 |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 156 training samples
* Columns: <code>sentence_0</code> and <code>sentence_1</code>
* Approximate statistics based on the first 156 samples:
| | sentence_0 | sentence_1 |
|:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 13 tokens</li><li>mean: 20.15 tokens</li><li>max: 29 tokens</li></ul> | <ul><li>min: 43 tokens</li><li>mean: 130.5 tokens</li><li>max: 204 tokens</li></ul> |
* Samples:
| sentence_0 | sentence_1 |
|:----------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>What key themes and pivotal moments in the field of Large Language Models were identified in 2024?</code> | <code>Things we learned about LLMs in 2024<br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br>Simon Willison’s Weblog<br>Subscribe<br><br><br><br><br><br><br>Things we learned about LLMs in 2024<br>31st December 2024<br>A lot has happened in the world of Large Language Models over the course of 2024. Here’s a review of things we figured out about the field in the past twelve months, plus my attempt at identifying key themes and pivotal moments.<br>This is a sequel to my review of 2023.<br>In this article:</code> |
| <code>How does the review of 2024 compare to the previous year's review of 2023?</code> | <code>Things we learned about LLMs in 2024<br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br>Simon Willison’s Weblog<br>Subscribe<br><br><br><br><br><br><br>Things we learned about LLMs in 2024<br>31st December 2024<br>A lot has happened in the world of Large Language Models over the course of 2024. Here’s a review of things we figured out about the field in the past twelve months, plus my attempt at identifying key themes and pivotal moments.<br>This is a sequel to my review of 2023.<br>In this article:</code> |
| <code>What advancements have been made in multimodal vision and audio/video capabilities in LLMs?</code> | <code>The GPT-4 barrier was comprehensively broken<br>Some of those GPT-4 models run on my laptop<br>LLM prices crashed, thanks to competition and increased efficiency<br>Multimodal vision is common, audio and video are starting to emerge<br>Voice and live camera mode are science fiction come to life<br>Prompt driven app generation is a commodity already<br>Universal access to the best models lasted for just a few short months<br>“Agents” still haven’t really happened yet<br>Evals really matter<br>Apple Intelligence is bad, Apple’s MLX library is excellent<br>The rise of inference-scaling “reasoning” models<br>Was the best currently available LLM trained in China for less than $6m?<br>The environmental impact got better<br>The environmental impact got much, much worse</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "MultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64
],
"matryoshka_weights": [
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 10
- `per_device_eval_batch_size`: 10
- `num_train_epochs`: 10
- `multi_dataset_batch_sampler`: round_robin
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 10
- `per_device_eval_batch_size`: 10
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 5e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1
- `num_train_epochs`: 10
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.0
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: None
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: round_robin
</details>
### Training Logs
| Epoch | Step | cosine_ndcg@10 |
|:-----:|:----:|:--------------:|
| 1.0 | 16 | 0.9039 |
| 2.0 | 32 | 0.9455 |
| 3.0 | 48 | 0.9692 |
| 3.125 | 50 | 0.9692 |
| 4.0 | 64 | 0.9692 |
| 5.0 | 80 | 0.9692 |
| 6.0 | 96 | 0.9539 |
| 6.25 | 100 | 0.9539 |
| 7.0 | 112 | 0.9539 |
| 8.0 | 128 | 0.9539 |
| 9.0 | 144 | 0.9539 |
| 9.375 | 150 | 0.9539 |
| 10.0 | 160 | 0.9539 |
### Framework Versions
- Python: 3.11.11
- Sentence Transformers: 3.4.1
- Transformers: 4.48.3
- PyTorch: 2.5.1+cu124
- Accelerate: 1.3.0
- Datasets: 3.3.1
- Tokenizers: 0.21.0
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MatryoshkaLoss
```bibtex
@misc{kusupati2024matryoshka,
title={Matryoshka Representation Learning},
author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
year={2024},
eprint={2205.13147},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
title={Efficient Natural Language Response Suggestion for Smart Reply},
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},
year={2017},
eprint={1705.00652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SentenceTransformer based on Snowflake/snowflake-arctic-embed-l
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Snowflake/snowflake-arctic-embed-l](https://huggingface.co/Snowflake/snowflake-arctic-embed-l). It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [Snowflake/snowflake-arctic-embed-l](https://huggingface.co/Snowflake/snowflake-arctic-embed-l) <!-- at revision d8fb21ca8d905d2832ee8b96c894d3298964346b -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 1024 dimensions
- **Similarity Function:** Cosine Similarity
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
(2): Normalize()
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("deepali1021/legal-ft-v0")
# Run inference
sentences = [
'How does the author feel about their current experience as a Mac user compared to the previous year?',
'I’m still trying to figure out the best patterns for doing this for my own work. Everyone knows that evals are important, but there remains a lack of great guidance for how to best implement them—I’m tracking this under my evals tag. My SVG pelican riding a bicycle benchmark is a pale imitation of what a real eval suite should look like.\nApple Intelligence is bad, Apple’s MLX library is excellent\nAs a Mac user I’ve been feeling a lot better about my choice of platform this year.\nLast year it felt like my lack of a Linux/Windows machine with an NVIDIA GPU was a huge disadvantage in terms of trying out new models.',
'The May 13th announcement of GPT-4o included a demo of a brand new voice mode, where the true multi-modal GPT-4o (the o is for “omni”) model could accept audio input and output incredibly realistic sounding speech without needing separate TTS or STT models.\nThe demo also sounded conspicuously similar to Scarlett Johansson... and after she complained the voice from the demo, Skye, never made it to a production product.\nThe delay in releasing the new voice mode after the initial demo caused quite a lot of confusion. I wrote about that in ChatGPT in “4o” mode is not running the new features yet.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Information Retrieval
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | Value |
|:--------------------|:-----------|
| cosine_accuracy@1 | 0.875 |
| cosine_accuracy@3 | 1.0 |
| cosine_accuracy@5 | 1.0 |
| cosine_accuracy@10 | 1.0 |
| cosine_precision@1 | 0.875 |
| cosine_precision@3 | 0.3333 |
| cosine_precision@5 | 0.2 |
| cosine_precision@10 | 0.1 |
| cosine_recall@1 | 0.875 |
| cosine_recall@3 | 1.0 |
| cosine_recall@5 | 1.0 |
| cosine_recall@10 | 1.0 |
| **cosine_ndcg@10** | **0.9539** |
| cosine_mrr@10 | 0.9375 |
| cosine_map@100 | 0.9375 |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 156 training samples
* Columns: <code>sentence_0</code> and <code>sentence_1</code>
* Approximate statistics based on the first 156 samples:
| | sentence_0 | sentence_1 |
|:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 13 tokens</li><li>mean: 20.15 tokens</li><li>max: 29 tokens</li></ul> | <ul><li>min: 43 tokens</li><li>mean: 130.5 tokens</li><li>max: 204 tokens</li></ul> |
* Samples:
| sentence_0 | sentence_1 |
|:----------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>What key themes and pivotal moments in the field of Large Language Models were identified in 2024?</code> | <code>Things we learned about LLMs in 2024<br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br>Simon Willison’s Weblog<br>Subscribe<br><br><br><br><br><br><br>Things we learned about LLMs in 2024<br>31st December 2024<br>A lot has happened in the world of Large Language Models over the course of 2024. Here’s a review of things we figured out about the field in the past twelve months, plus my attempt at identifying key themes and pivotal moments.<br>This is a sequel to my review of 2023.<br>In this article:</code> |
| <code>How does the review of 2024 compare to the previous year's review of 2023?</code> | <code>Things we learned about LLMs in 2024<br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br><br>Simon Willison’s Weblog<br>Subscribe<br><br><br><br><br><br><br>Things we learned about LLMs in 2024<br>31st December 2024<br>A lot has happened in the world of Large Language Models over the course of 2024. Here’s a review of things we figured out about the field in the past twelve months, plus my attempt at identifying key themes and pivotal moments.<br>This is a sequel to my review of 2023.<br>In this article:</code> |
| <code>What advancements have been made in multimodal vision and audio/video capabilities in LLMs?</code> | <code>The GPT-4 barrier was comprehensively broken<br>Some of those GPT-4 models run on my laptop<br>LLM prices crashed, thanks to competition and increased efficiency<br>Multimodal vision is common, audio and video are starting to emerge<br>Voice and live camera mode are science fiction come to life<br>Prompt driven app generation is a commodity already<br>Universal access to the best models lasted for just a few short months<br>“Agents” still haven’t really happened yet<br>Evals really matter<br>Apple Intelligence is bad, Apple’s MLX library is excellent<br>The rise of inference-scaling “reasoning” models<br>Was the best currently available LLM trained in China for less than $6m?<br>The environmental impact got better<br>The environmental impact got much, much worse</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "MultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64
],
"matryoshka_weights": [
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 10
- `per_device_eval_batch_size`: 10
- `num_train_epochs`: 10
- `multi_dataset_batch_sampler`: round_robin
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 10
- `per_device_eval_batch_size`: 10
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 5e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1
- `num_train_epochs`: 10
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.0
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: None
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: round_robin
</details>
### Training Logs
| Epoch | Step | cosine_ndcg@10 |
|:-----:|:----:|:--------------:|
| 1.0 | 16 | 0.9039 |
| 2.0 | 32 | 0.9455 |
| 3.0 | 48 | 0.9692 |
| 3.125 | 50 | 0.9692 |
| 4.0 | 64 | 0.9692 |
| 5.0 | 80 | 0.9692 |
| 6.0 | 96 | 0.9539 |
| 6.25 | 100 | 0.9539 |
| 7.0 | 112 | 0.9539 |
| 8.0 | 128 | 0.9539 |
| 9.0 | 144 | 0.9539 |
| 9.375 | 150 | 0.9539 |
| 10.0 | 160 | 0.9539 |
### Framework Versions
- Python: 3.11.11
- Sentence Transformers: 3.4.1
- Transformers: 4.48.3
- PyTorch: 2.5.1+cu124
- Accelerate: 1.3.0
- Datasets: 3.3.1
- Tokenizers: 0.21.0
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MatryoshkaLoss
```bibtex
@misc{kusupati2024matryoshka,
title={Matryoshka Representation Learning},
author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
year={2024},
eprint={2205.13147},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
title={Efficient Natural Language Response Suggestion for Smart Reply},
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},
year={2017},
eprint={1705.00652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "Snowflake/snowflake-arctic-embed-l", "library_name": "sentence-transformers", "metrics": ["cosine_accuracy@1", "cosine_accuracy@3", "cosine_accuracy@5", "cosine_accuracy@10", "cosine_precision@1", "cosine_precision@3", "cosine_precision@5", "cosine_precision@10", "cosine_recall@1", "cosine_recall@3", "cosine_recall@5", "cosine_recall@10", "cosine_ndcg@10", "cosine_mrr@10", "cosine_map@100"], "pipeline_tag": "sentence-similarity", "tags": ["sentence-transformers", "sentence-similarity", "feature-extraction", "generated_from_trainer", "dataset_size:156", "loss:MatryoshkaLoss", "loss:MultipleNegativesRankingLoss"], "widget": [{"source_sentence": "What new shape of LLM was introduced in the final quarter of 2024, and what were the names of the initial models released?", "sentences": ["17th: AI for Data Journalism: demonstrating what we can do with this stuff right now\n\n22nd: Options for accessing Llama 3 from the terminal using LLM\n\n\n\nMay\n\n8th: Slop is the new name for unwanted AI-generated content\n\n15th: ChatGPT in “4o” mode is not running the new features yet\n\n29th: Training is not the same as chatting: ChatGPT and other LLMs don’t remember everything you say\n\n\n\nJune\n\n6th: Accidental prompt injection against RAG applications\n\n10th: Thoughts on the WWDC 2024 keynote on Apple Intelligence\n\n17th: Language models on the command-line\n\n21st: Building search-based RAG using Claude, Datasette and Val Town\n\n27th: Open challenges for AI engineering\n\n\n\nJuly\n\n14th: Imitation Intelligence, my keynote for PyCon US 2024", "Now that those features are rolling out they’re pretty weak. As an LLM power-user I know what these models are capable of, and Apple’s LLM features offer a pale imitation of what a frontier LLM can do. Instead we’re getting notification summaries that misrepresent news headlines and writing assistant tools that I’ve not found useful at all. Genmoji are kind of fun though.\nThe rise of inference-scaling “reasoning” models\nThe most interesting development in the final quarter of 2024 was the introduction of a new shape of LLM, exemplified by OpenAI’s o1 models—initially released as o1-preview and o1-mini on September 12th.", "Terminology aside, I remain skeptical as to their utility based, once again, on the challenge of gullibility. LLMs believe anything you tell them. Any systems that attempts to make meaningful decisions on your behalf will run into the same roadblock: how good is a travel agent, or a digital assistant, or even a research tool if it can’t distinguish truth from fiction?\nJust the other day Google Search was caught serving up an entirely fake description of the non-existant movie “Encanto 2”. It turned out to be summarizing an imagined movie listing from a fan fiction wiki."]}, {"source_sentence": "In what year does the author expect the prompt-driven custom interface feature to be widely adopted in various products?", "sentences": ["The models may have got more capable, but most of the limitations remained the same. OpenAI’s o1 may finally be able to (mostly) count the Rs in strawberry, but its abilities are still limited by its nature as an LLM and the constraints placed on it by the harness it’s running in. o1 can’t run web searches or use Code Interpreter, but GPT-4o can—both in that same ChatGPT UI. (o1 will pretend to do those things if you ask it to, a regression to the URL hallucinations bug from early 2023).\nWhat are we doing about this? Not much. Most users are thrown in at the deep end. The default LLM chat UI is like taking brand new computer users, dropping them into a Linux terminal and expecting them to figure it all out.", "This prompt-driven custom interface feature is so powerful and easy to build (once you’ve figured out the gnarly details of browser sandboxing) that I expect it to show up as a feature in a wide range of products in 2025.\nUniversal access to the best models lasted for just a few short months\nFor a few short months this year all three of the best available models—GPT-4o, Claude 3.5 Sonnet and Gemini 1.5 Pro—were freely available to most of the world.", "Against this photo of butterflies at the California Academy of Sciences:\n\n\nA shallow dish, likely a hummingbird or butterfly feeder, is red. Pieces of orange slices of fruit are visible inside the dish.\nTwo butterflies are positioned in the feeder, one is a dark brown/black butterfly with white/cream-colored markings. The other is a large, brown butterfly with patterns of lighter brown, beige, and black markings, including prominent eye spots. The larger brown butterfly appears to be feeding on the fruit."]}, {"source_sentence": "What is the license under which Alibaba's QwQ model was released?", "sentences": ["The most recent twist, again from December (December was a lot) is live video. ChatGPT voice mode now provides the option to share your camera feed with the model and talk about what you can see in real time. Google Gemini have a preview of the same feature, which they managed to ship the day before ChatGPT did.", "OpenAI are not the only game in town here. Google released their first entrant in the category, gemini-2.0-flash-thinking-exp, on December 19th.\nAlibaba’s Qwen team released their QwQ model on November 28th—under an Apache 2.0 license, and that one I could run on my own machine. They followed that up with a vision reasoning model called QvQ on December 24th, which I also ran locally.\nDeepSeek made their DeepSeek-R1-Lite-Preview model available to try out through their chat interface on November 20th.\nTo understand more about inference scaling I recommend Is AI progress slowing down? by Arvind Narayanan and Sayash Kapoor.", "Stuff we figured out about AI in 2023\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nSimon Willison’s Weblog\nSubscribe\n\n\n\n\n\n\nStuff we figured out about AI in 2023\n31st December 2023\n2023 was the breakthrough year for Large Language Models (LLMs). I think it’s OK to call these AI—they’re the latest and (currently) most interesting development in the academic field of Artificial Intelligence that dates back to the 1950s.\nHere’s my attempt to round up the highlights in one place!"]}, {"source_sentence": "What is the cost of using 260 input tokens and 92 output tokens as mentioned in the context?", "sentences": ["I think people who complain that LLM improvement has slowed are often missing the enormous advances in these multi-modal models. Being able to run prompts against images (and audio and video) is a fascinating new way to apply these models.\nVoice and live camera mode are science fiction come to life\nThe audio and live video modes that have started to emerge deserve a special mention.\nThe ability to talk to ChatGPT first arrived in September 2023, but it was mostly an illusion: OpenAI used their excellent Whisper speech-to-text model and a new text-to-speech model (creatively named tts-1) to enable conversations with the ChatGPT mobile apps, but the actual model just saw text.", "I like people who are skeptical of this stuff. The hype has been deafening for more than two years now, and there are enormous quantities of snake oil and misinformation out there. A lot of very bad decisions are being made based on that hype. Being critical is a virtue.\nIf we want people with decision-making authority to make good decisions about how to apply these tools we first need to acknowledge that there ARE good applications, and then help explain how to put those into practice while avoiding the many unintiutive traps.\n(If you still don’t think there are any good applications at all I’m not sure why you made it to this point in the article!)", "260 input tokens, 92 output tokens. Cost approximately 0.0024 cents (that’s less than a 400th of a cent).\nThis increase in efficiency and reduction in price is my single favourite trend from 2024. I want the utility of LLMs at a fraction of the energy cost and it looks like that’s what we’re getting.\nMultimodal vision is common, audio and video are starting to emerge\nMy butterfly example above illustrates another key trend from 2024: the rise of multi-modal LLMs.\nA year ago the single most notable example of these was GPT-4 Vision, released at OpenAI’s DevDay in November 2023. Google’s multi-modal Gemini 1.0 was announced on December 7th 2023 so it also (just) makes it into the 2023 window."]}, {"source_sentence": "How does the author feel about their current experience as a Mac user compared to the previous year?", "sentences": ["I’m still trying to figure out the best patterns for doing this for my own work. Everyone knows that evals are important, but there remains a lack of great guidance for how to best implement them—I’m tracking this under my evals tag. My SVG pelican riding a bicycle benchmark is a pale imitation of what a real eval suite should look like.\nApple Intelligence is bad, Apple’s MLX library is excellent\nAs a Mac user I’ve been feeling a lot better about my choice of platform this year.\nLast year it felt like my lack of a Linux/Windows machine with an NVIDIA GPU was a huge disadvantage in terms of trying out new models.", "Structured and Gradual Learning. In organic datasets, the relationship between tokens is often complex and indirect. Many reasoning steps may be required to connect the current token to the next, making it challenging for the model to learn effectively from next-token prediction. By contrast, each token generated by a language model is by definition predicted by the preceding tokens, making it easier for a model to follow the resulting reasoning patterns.", "The May 13th announcement of GPT-4o included a demo of a brand new voice mode, where the true multi-modal GPT-4o (the o is for “omni”) model could accept audio input and output incredibly realistic sounding speech without needing separate TTS or STT models.\nThe demo also sounded conspicuously similar to Scarlett Johansson... and after she complained the voice from the demo, Skye, never made it to a production product.\nThe delay in releasing the new voice mode after the initial demo caused quite a lot of confusion. I wrote about that in ChatGPT in “4o” mode is not running the new features yet."]}], "model-index": [{"name": "SentenceTransformer based on Snowflake/snowflake-arctic-embed-l", "results": [{"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "Unknown", "type": "unknown"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.875, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 1.0, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 1.0, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 1.0, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.875, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.3333333333333333, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.20000000000000004, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.10000000000000002, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.875, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 1.0, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 1.0, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 1.0, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.9538662191964322, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.9375, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.9375, "name": "Cosine Map@100"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,836 |
sadhaklal/bert-base-cased-finetuned-conll2003-ner-v2
|
sadhaklal
|
token-classification
|
[
"transformers",
"safetensors",
"bert",
"token-classification",
"en",
"dataset:conll2003",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-02-22T16:57:14Z |
2024-02-26T08:49:37+00:00
| 10 | 0 |
---
datasets:
- conll2003
language:
- en
library_name: transformers
license: apache-2.0
metrics:
- accuracy
- precision
- recall
- f1
pipeline_tag: token-classification
---
# bert-base-cased-finetuned-conll2003-ner-v2
BERT ("bert-base-cased") finetuned on CoNLL-2003 (Conference on Computational Natural Language Learning).
The model performs named entity recognition (NER). It pertains to section 2 of chapter 7 of the Hugging Face "NLP Course" (https://huggingface.co/learn/nlp-course/chapter7/2).
It was trained using a custom PyTorch loop with Hugging Face Accelerate.
Code: https://github.com/sambitmukherjee/huggingface-notebooks/blob/main/course/en/chapter7/section2_pt.ipynb
Experiment tracking: https://wandb.ai/sadhaklal/bert-base-cased-finetuned-conll2003-ner-v2
## Usage
```
from transformers import pipeline
model_checkpoint = "sadhaklal/bert-base-cased-finetuned-conll2003-ner-v2"
token_classifier = pipeline("token-classification", model=model_checkpoint, aggregation_strategy="simple")
print(token_classifier("My name is Sylvain and I work at Hugging Face in Brooklyn."))
```
## Dataset
From the dataset page:
> The shared task of CoNLL-2003 concerns language-independent named entity recognition. We will concentrate on four types of named entities: persons, locations, organizations and names of miscellaneous entities that do not belong to the previous three groups.
Examples: https://huggingface.co/datasets/conll2003/viewer
## Metrics
Accuracy on the 'validation' split of CoNLL-2003: 0.9858
Precision on the 'validation' split of CoNLL-2003: 0.9243
Recall on the 'validation' split of CoNLL-2003: 0.947
F1 on the 'validation' split of CoNLL-2003: 0.9355
| null |
Non_BioNLP
|
# bert-base-cased-finetuned-conll2003-ner-v2
BERT ("bert-base-cased") finetuned on CoNLL-2003 (Conference on Computational Natural Language Learning).
The model performs named entity recognition (NER). It pertains to section 2 of chapter 7 of the Hugging Face "NLP Course" (https://huggingface.co/learn/nlp-course/chapter7/2).
It was trained using a custom PyTorch loop with Hugging Face Accelerate.
Code: https://github.com/sambitmukherjee/huggingface-notebooks/blob/main/course/en/chapter7/section2_pt.ipynb
Experiment tracking: https://wandb.ai/sadhaklal/bert-base-cased-finetuned-conll2003-ner-v2
## Usage
```
from transformers import pipeline
model_checkpoint = "sadhaklal/bert-base-cased-finetuned-conll2003-ner-v2"
token_classifier = pipeline("token-classification", model=model_checkpoint, aggregation_strategy="simple")
print(token_classifier("My name is Sylvain and I work at Hugging Face in Brooklyn."))
```
## Dataset
From the dataset page:
> The shared task of CoNLL-2003 concerns language-independent named entity recognition. We will concentrate on four types of named entities: persons, locations, organizations and names of miscellaneous entities that do not belong to the previous three groups.
Examples: https://huggingface.co/datasets/conll2003/viewer
## Metrics
Accuracy on the 'validation' split of CoNLL-2003: 0.9858
Precision on the 'validation' split of CoNLL-2003: 0.9243
Recall on the 'validation' split of CoNLL-2003: 0.947
F1 on the 'validation' split of CoNLL-2003: 0.9355
|
{"datasets": ["conll2003"], "language": ["en"], "library_name": "transformers", "license": "apache-2.0", "metrics": ["accuracy", "precision", "recall", "f1"], "pipeline_tag": "token-classification"}
|
task
|
[
"NAMED_ENTITY_RECOGNITION"
] | 39,837 |
Alpha-VLLM/SPHINX
|
Alpha-VLLM
| null |
[
"license:mit",
"region:us"
] | 2023-11-02T16:42:50Z |
2023-11-09T16:32:23+00:00
| 0 | 41 |
---
license: mit
---
# 🔥 SPHINX: A Mixer of Tasks, Domains, and Embeddings
Official implementation of ['SPHINX: A Mixer of Tasks, Domains, and Embeddings Advances Multi-modal Large Language Models'](https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX).
Try out our [web demo 🚀](http://imagebind-llm.opengvlab.com/) here!
<p align="left">
Github link: <a href="https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX" target="_blank">Github</a> • 👋 join our <a href="https://github.com/Alpha-VLLM/LLaMA2-Accessory/blob/main/docs/wechat.md" target="_blank">WeChat</a>
</p>
## Introduction
We present SPHINX, a versatile multi-modal large language model (MLLM) with a mixer of training tasks, data domains, and visual embeddings.
- **Task Mix.** For all-purpose capabilities, we mix a variety of vision-language tasks for mutual improvement: VQA, REC, REG, OCR, DET, POSE, REL DET, T2I, etc.
- **Embedding Mix.** We capture robust visual representations by fusing distinct visual architectures, pre-training, and granularity.
- **Domain Mix.** For data from real-world and synthetic domains, we mix the weights of two domain-specific models for complementarity.
<p align="left">
<img src="figs/pipeline1.png"/ width="100%"> <br>
</p>
On top of SPHINX, we propose to further mix visual scales and sub-images for better capture fine-grained semantics on high-resolution images.
<p align="left">
<img src="figs/pipeline2.png"/ width="100%"> <br>
</p>
### Installation
SPHINX is built upon LLaMA2-Accessory, please follow the instructions [here](https://llama2-accessory.readthedocs.io/en/latest/install.html) for environment setup.
## Inference
This section provides a step-by-step guide for hosting a local SPHINX demo. If you're already familiar with the LLAMA2-Accessory toolkit, note that hosting a SPHINX demo follows the same pipeline as hosting demos for the other models supported by LLAMA2-Accessory.
### Weights
We provide the beta-version checkpoints on [HuggingFace🤗](https://huggingface.co/Alpha-VLLM/LLaMA2-Accessory/tree/main/finetune/mm/SPHINX). Please download them to your own machine. The file structure should appear as follows:
```
ckpt_path/
├── consolidated.00-of-02.model.pth
└── consolidated.01-of-02.model.pth
```
### Host Local Demo
Please follow the instructions [here](https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX#host-local-demo) to see the instruction and complete the use of the model.
## Result
We provide a comprehensive evaluation of SPHINX and showcase results across multiple benchmarks.
Our evaluation encompasses both **quantitative metrics** and **qualitative assessments**, providing a holistic understanding of our VLM model's performance.
**Evaluation Prompt Design**
<p align="left">
<img src="figs/table1.png"/ width="100%"> <br>
</p>
* In evaluation, we prioritize aligning with each benchmark's desired output format.
* We employ distinct prompts tailored to benchmarks that necessitate long answers, short answers, and multiple-choice responses.
* For tasks involving visual grounding, we directly utilize the prompts during training to enhance the model's performance on these particular challenges.
**Benchmarks on Multimodal Large Language Models**
<p align="left">
<img src="figs/table2.png"/ width="100%"> <br>
</p
* We test our model on recently proposed MLLM benchmarks which is based on VQA to comprehensive evaluation of the model's characteristic such as MME, Seedbench, POPE, LLaVA-Bench (In-the-Wild), MM-Vet, MathVista, MMbench, CCbench.
* The Long-SPHINX achieve new stat of arts result on 5 out of 9 benchmarks
**Visual Question Answering**
<p align="left">
<img src="figs/table3.png"/ width="100%"> <br>
</p>
* We evaluate general VQA benchmarks, such as VQAV2, OKVQA, GQA, vizwiz, scienceQA, visual spatial reasoning (VSR), IconQA.
* Additionally, we conduct experiments on Text-oriented VQA such as TextVQA,OCR-VQA.
* Long-Sphinx achieve comparative results across all benchmarks. We observe that Long-Sphinx outperforms Sphinx in VQA datasets that demand fine-grained visual information, showcasing the effectiveness of our visual mixed-up approach for achieving high resolution without relying on a visual encoder trained specifically on high-resolution images.
**Visual Grounding**
<p align="left">
<img src="figs/table4.png"/ width="100%"> <br>
</p>
* The SPHINX model and baseline models on REC benchmarks results on table4.
* SPHINX exhibits robust performance in visual grounding tasks such as RefCOCO, RefCOCO+, and RefCOCOg, **surpassing other vision-language generalist models**.
* Notably, SPHINX outperforms specialist models G-DINO-L by **more than 1.54%** in accuracy across all tasks within RefCOCO/RefCOCO+/RefCOCOg.
## Frequently Asked Questions (FAQ)
❓ Encountering issues or have further questions? Find answers to common inquiries [here](https://llama2-accessory.readthedocs.io/en/latest/faq.html). We're here to assist you!
## License
Llama 2 is licensed under the [LLAMA 2 Community License](LICENSE_llama2), Copyright (c) Meta Platforms, Inc. All Rights Reserved.
| null |
Non_BioNLP
|
# 🔥 SPHINX: A Mixer of Tasks, Domains, and Embeddings
Official implementation of ['SPHINX: A Mixer of Tasks, Domains, and Embeddings Advances Multi-modal Large Language Models'](https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX).
Try out our [web demo 🚀](http://imagebind-llm.opengvlab.com/) here!
<p align="left">
Github link: <a href="https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX" target="_blank">Github</a> • 👋 join our <a href="https://github.com/Alpha-VLLM/LLaMA2-Accessory/blob/main/docs/wechat.md" target="_blank">WeChat</a>
</p>
## Introduction
We present SPHINX, a versatile multi-modal large language model (MLLM) with a mixer of training tasks, data domains, and visual embeddings.
- **Task Mix.** For all-purpose capabilities, we mix a variety of vision-language tasks for mutual improvement: VQA, REC, REG, OCR, DET, POSE, REL DET, T2I, etc.
- **Embedding Mix.** We capture robust visual representations by fusing distinct visual architectures, pre-training, and granularity.
- **Domain Mix.** For data from real-world and synthetic domains, we mix the weights of two domain-specific models for complementarity.
<p align="left">
<img src="figs/pipeline1.png"/ width="100%"> <br>
</p>
On top of SPHINX, we propose to further mix visual scales and sub-images for better capture fine-grained semantics on high-resolution images.
<p align="left">
<img src="figs/pipeline2.png"/ width="100%"> <br>
</p>
### Installation
SPHINX is built upon LLaMA2-Accessory, please follow the instructions [here](https://llama2-accessory.readthedocs.io/en/latest/install.html) for environment setup.
## Inference
This section provides a step-by-step guide for hosting a local SPHINX demo. If you're already familiar with the LLAMA2-Accessory toolkit, note that hosting a SPHINX demo follows the same pipeline as hosting demos for the other models supported by LLAMA2-Accessory.
### Weights
We provide the beta-version checkpoints on [HuggingFace🤗](https://huggingface.co/Alpha-VLLM/LLaMA2-Accessory/tree/main/finetune/mm/SPHINX). Please download them to your own machine. The file structure should appear as follows:
```
ckpt_path/
├── consolidated.00-of-02.model.pth
└── consolidated.01-of-02.model.pth
```
### Host Local Demo
Please follow the instructions [here](https://github.com/Alpha-VLLM/LLaMA2-Accessory/tree/main/SPHINX#host-local-demo) to see the instruction and complete the use of the model.
## Result
We provide a comprehensive evaluation of SPHINX and showcase results across multiple benchmarks.
Our evaluation encompasses both **quantitative metrics** and **qualitative assessments**, providing a holistic understanding of our VLM model's performance.
**Evaluation Prompt Design**
<p align="left">
<img src="figs/table1.png"/ width="100%"> <br>
</p>
* In evaluation, we prioritize aligning with each benchmark's desired output format.
* We employ distinct prompts tailored to benchmarks that necessitate long answers, short answers, and multiple-choice responses.
* For tasks involving visual grounding, we directly utilize the prompts during training to enhance the model's performance on these particular challenges.
**Benchmarks on Multimodal Large Language Models**
<p align="left">
<img src="figs/table2.png"/ width="100%"> <br>
</p
* We test our model on recently proposed MLLM benchmarks which is based on VQA to comprehensive evaluation of the model's characteristic such as MME, Seedbench, POPE, LLaVA-Bench (In-the-Wild), MM-Vet, MathVista, MMbench, CCbench.
* The Long-SPHINX achieve new stat of arts result on 5 out of 9 benchmarks
**Visual Question Answering**
<p align="left">
<img src="figs/table3.png"/ width="100%"> <br>
</p>
* We evaluate general VQA benchmarks, such as VQAV2, OKVQA, GQA, vizwiz, scienceQA, visual spatial reasoning (VSR), IconQA.
* Additionally, we conduct experiments on Text-oriented VQA such as TextVQA,OCR-VQA.
* Long-Sphinx achieve comparative results across all benchmarks. We observe that Long-Sphinx outperforms Sphinx in VQA datasets that demand fine-grained visual information, showcasing the effectiveness of our visual mixed-up approach for achieving high resolution without relying on a visual encoder trained specifically on high-resolution images.
**Visual Grounding**
<p align="left">
<img src="figs/table4.png"/ width="100%"> <br>
</p>
* The SPHINX model and baseline models on REC benchmarks results on table4.
* SPHINX exhibits robust performance in visual grounding tasks such as RefCOCO, RefCOCO+, and RefCOCOg, **surpassing other vision-language generalist models**.
* Notably, SPHINX outperforms specialist models G-DINO-L by **more than 1.54%** in accuracy across all tasks within RefCOCO/RefCOCO+/RefCOCOg.
## Frequently Asked Questions (FAQ)
❓ Encountering issues or have further questions? Find answers to common inquiries [here](https://llama2-accessory.readthedocs.io/en/latest/faq.html). We're here to assist you!
## License
Llama 2 is licensed under the [LLAMA 2 Community License](LICENSE_llama2), Copyright (c) Meta Platforms, Inc. All Rights Reserved.
|
{"license": "mit"}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,838 |
ArierMiao/marian-finetuned-kde4-en-to-fr
|
ArierMiao
|
translation
|
[
"transformers",
"tensorboard",
"safetensors",
"marian",
"text2text-generation",
"translation",
"generated_from_trainer",
"dataset:kde4",
"base_model:Helsinki-NLP/opus-mt-en-fr",
"base_model:finetune:Helsinki-NLP/opus-mt-en-fr",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-06-07T07:59:55Z |
2024-12-10T08:26:31+00:00
| 16 | 0 |
---
base_model: Helsinki-NLP/opus-mt-en-fr
datasets:
- kde4
library_name: transformers
license: apache-2.0
metrics:
- bleu
tags:
- translation
- generated_from_trainer
model-index:
- name: marian-finetuned-kde4-en-to-fr
results:
- task:
type: text2text-generation
name: Sequence-to-sequence Language Modeling
dataset:
name: kde4
type: kde4
config: en-fr
split: train
args: en-fr
metrics:
- type: bleu
value: 52.90204973205105
name: Bleu
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# marian-finetuned-kde4-en-to-fr
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on the kde4 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8554
- Model Preparation Time: 0.0156
- Bleu: 52.9020
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.46.3
- Pytorch 2.5.1+cu121
- Datasets 3.1.0
- Tokenizers 0.20.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# marian-finetuned-kde4-en-to-fr
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on the kde4 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8554
- Model Preparation Time: 0.0156
- Bleu: 52.9020
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.46.3
- Pytorch 2.5.1+cu121
- Datasets 3.1.0
- Tokenizers 0.20.3
|
{"base_model": "Helsinki-NLP/opus-mt-en-fr", "datasets": ["kde4"], "library_name": "transformers", "license": "apache-2.0", "metrics": ["bleu"], "tags": ["translation", "generated_from_trainer"], "model-index": [{"name": "marian-finetuned-kde4-en-to-fr", "results": [{"task": {"type": "text2text-generation", "name": "Sequence-to-sequence Language Modeling"}, "dataset": {"name": "kde4", "type": "kde4", "config": "en-fr", "split": "train", "args": "en-fr"}, "metrics": [{"type": "bleu", "value": 52.90204973205105, "name": "Bleu"}]}]}]}
|
task
|
[
"TRANSLATION"
] | 39,839 |
safe049/TigerStheno-8B-v3.2
|
safe049
| null |
[
"transformers",
"pytorch",
"gguf",
"llama",
"text-generation-inference",
"unsloth",
"trl",
"sft",
"en",
"zh",
"base_model:Sao10K/L3-8B-Stheno-v3.2",
"base_model:quantized:Sao10K/L3-8B-Stheno-v3.2",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | 2024-12-13T10:32:14Z |
2024-12-13T11:34:03+00:00
| 44 | 1 |
---
base_model: Sao10K/L3-8B-Stheno-v3.2
language:
- en
- zh
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- llama
- trl
- sft
---
# TigerStheno
- **Developed by:** safe049
- **License:** apache-2.0
- **Finetuned from model :** Sao10K/L3-8B-Stheno-v3.2
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
### Model Details
**Model Name:** TigerStheno-8B-V3.2
**Permalink:** [TigerStheno-8B-V3.2](https://huggingface.co/TigerResearch/TigerStheno-8B-V3.2)
**Model Description:**
TigerStheno-8B-V3.2 is an AI model based on the Sao10K/L3-8B-Stheno-v3.2 architecture, fine-tuned using the TigerBot dataset. This model is designed to enhance language understanding and generation capabilities, particularly in multi-language and multi-task role-playing scenarios.
- **Developed by:** safe049
- **Shared by:** safe049
- **Model type:** Large Language Model
- **Language(s) (NLP):** Multi-language support
- **License:** Apache 2.0
- **Finetuned from model:** Sao10K/L3-8B-Stheno-v3.2
### Uses
**Direct Use:**
The model can be directly used for various natural language processing tasks such as role-playing, text generation, summarization, translation, and dialogue systems.
### Bias, Risks
- **Bias:** TigerStheno-8B-V3.2 is uncensored and may give any response including illegal,in-moral content.
- **Risks:** May generate illegal and in-moral content
### Using it
**GGUF:**
Download the gguf file in the repo, and use it in any of these apps and etc:
- KoboldCPP
- Ollama
- LlamaCPP
etc.
**Transformer**
Here is a example code snippet to use it with transformer:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "TigerResearch/TigerStheno-8B-V3.2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
# Example input
inputs = tokenizer("Hello, how are you?", return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0]))
```
| null |
Non_BioNLP
|
# TigerStheno
- **Developed by:** safe049
- **License:** apache-2.0
- **Finetuned from model :** Sao10K/L3-8B-Stheno-v3.2
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
### Model Details
**Model Name:** TigerStheno-8B-V3.2
**Permalink:** [TigerStheno-8B-V3.2](https://huggingface.co/TigerResearch/TigerStheno-8B-V3.2)
**Model Description:**
TigerStheno-8B-V3.2 is an AI model based on the Sao10K/L3-8B-Stheno-v3.2 architecture, fine-tuned using the TigerBot dataset. This model is designed to enhance language understanding and generation capabilities, particularly in multi-language and multi-task role-playing scenarios.
- **Developed by:** safe049
- **Shared by:** safe049
- **Model type:** Large Language Model
- **Language(s) (NLP):** Multi-language support
- **License:** Apache 2.0
- **Finetuned from model:** Sao10K/L3-8B-Stheno-v3.2
### Uses
**Direct Use:**
The model can be directly used for various natural language processing tasks such as role-playing, text generation, summarization, translation, and dialogue systems.
### Bias, Risks
- **Bias:** TigerStheno-8B-V3.2 is uncensored and may give any response including illegal,in-moral content.
- **Risks:** May generate illegal and in-moral content
### Using it
**GGUF:**
Download the gguf file in the repo, and use it in any of these apps and etc:
- KoboldCPP
- Ollama
- LlamaCPP
etc.
**Transformer**
Here is a example code snippet to use it with transformer:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "TigerResearch/TigerStheno-8B-V3.2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
# Example input
inputs = tokenizer("Hello, how are you?", return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0]))
```
|
{"base_model": "Sao10K/L3-8B-Stheno-v3.2", "language": ["en", "zh"], "license": "apache-2.0", "tags": ["text-generation-inference", "transformers", "unsloth", "llama", "trl", "sft"]}
|
task
|
[
"TRANSLATION",
"SUMMARIZATION"
] | 39,840 |
SteveMLC/phrasebank-sentiment-analysis
|
SteveMLC
|
text-classification
|
[
"transformers",
"pytorch",
"bert",
"text-classification",
"generated_from_trainer",
"dataset:financial_phrasebank",
"base_model:google-bert/bert-base-uncased",
"base_model:finetune:google-bert/bert-base-uncased",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-10-29T22:47:07Z |
2023-10-29T22:47:30+00:00
| 20 | 0 |
---
base_model: bert-base-uncased
datasets:
- financial_phrasebank
license: apache-2.0
metrics:
- f1
- accuracy
tags:
- generated_from_trainer
model-index:
- name: phrasebank-sentiment-analysis
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: financial_phrasebank
type: financial_phrasebank
config: sentences_50agree
split: train
args: sentences_50agree
metrics:
- type: f1
value: 0.8431670091796087
name: F1
- type: accuracy
value: 0.8569463548830811
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# phrasebank-sentiment-analysis
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the financial_phrasebank dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5294
- F1: 0.8432
- Accuracy: 0.8569
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:------:|:--------:|
| 0.6018 | 0.94 | 100 | 0.3864 | 0.8230 | 0.8473 |
| 0.285 | 1.89 | 200 | 0.3750 | 0.8340 | 0.8487 |
| 0.1449 | 2.83 | 300 | 0.4920 | 0.8361 | 0.8508 |
| 0.0704 | 3.77 | 400 | 0.5294 | 0.8432 | 0.8569 |
### Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# phrasebank-sentiment-analysis
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the financial_phrasebank dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5294
- F1: 0.8432
- Accuracy: 0.8569
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:------:|:--------:|
| 0.6018 | 0.94 | 100 | 0.3864 | 0.8230 | 0.8473 |
| 0.285 | 1.89 | 200 | 0.3750 | 0.8340 | 0.8487 |
| 0.1449 | 2.83 | 300 | 0.4920 | 0.8361 | 0.8508 |
| 0.0704 | 3.77 | 400 | 0.5294 | 0.8432 | 0.8569 |
### Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1
|
{"base_model": "bert-base-uncased", "datasets": ["financial_phrasebank"], "license": "apache-2.0", "metrics": ["f1", "accuracy"], "tags": ["generated_from_trainer"], "model-index": [{"name": "phrasebank-sentiment-analysis", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "financial_phrasebank", "type": "financial_phrasebank", "config": "sentences_50agree", "split": "train", "args": "sentences_50agree"}, "metrics": [{"type": "f1", "value": 0.8431670091796087, "name": "F1"}, {"type": "accuracy", "value": 0.8569463548830811, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,841 |
RogerB/marian-finetuned-Umuganda-Dataset-en-to-kin
|
RogerB
|
translation
|
[
"transformers",
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"translation",
"generated_from_trainer",
"base_model:Helsinki-NLP/opus-mt-en-rw",
"base_model:finetune:Helsinki-NLP/opus-mt-en-rw",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-08-08T18:52:54Z |
2023-08-09T15:53:16+00:00
| 9 | 0 |
---
base_model: Helsinki-NLP/opus-mt-en-rw
license: apache-2.0
metrics:
- bleu
tags:
- translation
- generated_from_trainer
model-index:
- name: marian-finetuned-kde4-en-to-kin-Umuganda-Dataset
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# marian-finetuned-Umuganda-Dataset-en-to-kin-Umuganda-Dataset
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-rw](https://huggingface.co/Helsinki-NLP/opus-mt-en-rw) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.8769
- Bleu: 32.8345
## Model Description
The model has been fine-tuned to perform machine translation from English to Kinyarwanda.
## Intended Uses & Limitations
The primary intended use of this model is for research purposes.
## Training and Evaluation Data
The model has been fine-tuned using the [Digital Umuganda](https://huggingface.co/datasets/DigitalUmuganda/kinyarwanda-english-machine-translation-dataset/tree/main) dataset.
The dataset was split with 90% used for training and 10% for testing.
The data used to train the model were cased and digits removed.
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
### Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# marian-finetuned-Umuganda-Dataset-en-to-kin-Umuganda-Dataset
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-rw](https://huggingface.co/Helsinki-NLP/opus-mt-en-rw) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.8769
- Bleu: 32.8345
## Model Description
The model has been fine-tuned to perform machine translation from English to Kinyarwanda.
## Intended Uses & Limitations
The primary intended use of this model is for research purposes.
## Training and Evaluation Data
The model has been fine-tuned using the [Digital Umuganda](https://huggingface.co/datasets/DigitalUmuganda/kinyarwanda-english-machine-translation-dataset/tree/main) dataset.
The dataset was split with 90% used for training and 10% for testing.
The data used to train the model were cased and digits removed.
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
### Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
|
{"base_model": "Helsinki-NLP/opus-mt-en-rw", "license": "apache-2.0", "metrics": ["bleu"], "tags": ["translation", "generated_from_trainer"], "model-index": [{"name": "marian-finetuned-kde4-en-to-kin-Umuganda-Dataset", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,842 |
yemoncad/distilbert-base-uncased-finetuned-clinc
|
yemoncad
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:clinc_oos",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-03-27T22:28:29Z |
2023-03-27T22:34:16+00:00
| 13 | 0 |
---
datasets:
- clinc_oos
license: apache-2.0
metrics:
- accuracy
tags:
- generated_from_trainer
model-index:
- name: distilbert-base-uncased-finetuned-clinc
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: clinc_oos
type: clinc_oos
args: plus
metrics:
- type: accuracy
value: 0.9183870967741935
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-clinc
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the clinc_oos dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7721
- Accuracy: 0.9184
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 48
- eval_batch_size: 48
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 4.2896 | 1.0 | 318 | 3.2890 | 0.7432 |
| 2.6284 | 2.0 | 636 | 1.8756 | 0.8377 |
| 1.5483 | 3.0 | 954 | 1.1572 | 0.8961 |
| 1.015 | 4.0 | 1272 | 0.8573 | 0.9132 |
| 0.7953 | 5.0 | 1590 | 0.7721 | 0.9184 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.13.1+cu116
- Datasets 1.16.1
- Tokenizers 0.10.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-clinc
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the clinc_oos dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7721
- Accuracy: 0.9184
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 48
- eval_batch_size: 48
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 4.2896 | 1.0 | 318 | 3.2890 | 0.7432 |
| 2.6284 | 2.0 | 636 | 1.8756 | 0.8377 |
| 1.5483 | 3.0 | 954 | 1.1572 | 0.8961 |
| 1.015 | 4.0 | 1272 | 0.8573 | 0.9132 |
| 0.7953 | 5.0 | 1590 | 0.7721 | 0.9184 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.13.1+cu116
- Datasets 1.16.1
- Tokenizers 0.10.3
|
{"datasets": ["clinc_oos"], "license": "apache-2.0", "metrics": ["accuracy"], "tags": ["generated_from_trainer"], "model-index": [{"name": "distilbert-base-uncased-finetuned-clinc", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "clinc_oos", "type": "clinc_oos", "args": "plus"}, "metrics": [{"type": "accuracy", "value": 0.9183870967741935, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,843 |
VietAI/envit5-base
|
VietAI
|
question-answering
|
[
"transformers",
"pytorch",
"tf",
"jax",
"t5",
"text2text-generation",
"summarization",
"translation",
"question-answering",
"vi",
"dataset:cc100",
"arxiv:2210.05610",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2022-06-20T17:52:40Z |
2022-12-07T18:18:18+00:00
| 121 | 1 |
---
datasets:
- cc100
language: vi
license: mit
tags:
- summarization
- translation
- question-answering
---
# EnViT5-base
State-of-the-art pretrained Transformer-based encoder-decoder model for Vietnamese and English used in [MTet's paper](https://arxiv.org/abs/2210.05610).
## How to use
For more details, do check out [our Github repo](https://github.com/vietai/mtet).
[Finetunning examples can be found here](https://github.com/vietai/ViT5/tree/main/finetunning_huggingface).
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("VietAI/envit5-base")
model = AutoModelForSeq2SeqLM.from_pretrained("VietAI/envit5-base")
model.cuda()
# need prefix for en: and vi: sentences
inputs = [
"vi: VietAI là tổ chức phi lợi nhuận với sứ mệnh ươm mầm tài năng về trí tuệ nhân tạo và xây dựng một cộng đồng các chuyên gia trong lĩnh vực trí tuệ nhân tạo đẳng cấp quốc tế tại Việt Nam.",
"vi: Theo báo cáo mới nhất của Linkedin về danh sách việc làm triển vọng với mức lương hấp dẫn năm 2020, các chức danh công việc liên quan đến AI như Chuyên gia AI (Artificial Intelligence Specialist), Kỹ sư ML (Machine Learning Engineer) đều xếp thứ hạng cao.",
"en: Our teams aspire to make discoveries that impact everyone, and core to our approach is sharing our research and tools to fuel progress in the field.",
"en: We're on a journey to advance and democratize artificial intelligence through open source and open science."
]
outputs = model.generate(tokenizer(inputs, return_tensors="pt", padding=True).input_ids.to('cuda'), max_length=512)
print(tokenizer.batch_decode(outputs, skip_special_tokens=True))
```
## Citation
```
@misc{mtet,
doi = {10.48550/ARXIV.2210.05610},
url = {https://arxiv.org/abs/2210.05610},
author = {Ngo, Chinh and Trinh, Trieu H. and Phan, Long and Tran, Hieu and Dang, Tai and Nguyen, Hieu and Nguyen, Minh and Luong, Minh-Thang},
keywords = {Computation and Language (cs.CL), Artificial Intelligence (cs.AI), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {MTet: Multi-domain Translation for English and Vietnamese},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
| null |
Non_BioNLP
|
# EnViT5-base
State-of-the-art pretrained Transformer-based encoder-decoder model for Vietnamese and English used in [MTet's paper](https://arxiv.org/abs/2210.05610).
## How to use
For more details, do check out [our Github repo](https://github.com/vietai/mtet).
[Finetunning examples can be found here](https://github.com/vietai/ViT5/tree/main/finetunning_huggingface).
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("VietAI/envit5-base")
model = AutoModelForSeq2SeqLM.from_pretrained("VietAI/envit5-base")
model.cuda()
# need prefix for en: and vi: sentences
inputs = [
"vi: VietAI là tổ chức phi lợi nhuận với sứ mệnh ươm mầm tài năng về trí tuệ nhân tạo và xây dựng một cộng đồng các chuyên gia trong lĩnh vực trí tuệ nhân tạo đẳng cấp quốc tế tại Việt Nam.",
"vi: Theo báo cáo mới nhất của Linkedin về danh sách việc làm triển vọng với mức lương hấp dẫn năm 2020, các chức danh công việc liên quan đến AI như Chuyên gia AI (Artificial Intelligence Specialist), Kỹ sư ML (Machine Learning Engineer) đều xếp thứ hạng cao.",
"en: Our teams aspire to make discoveries that impact everyone, and core to our approach is sharing our research and tools to fuel progress in the field.",
"en: We're on a journey to advance and democratize artificial intelligence through open source and open science."
]
outputs = model.generate(tokenizer(inputs, return_tensors="pt", padding=True).input_ids.to('cuda'), max_length=512)
print(tokenizer.batch_decode(outputs, skip_special_tokens=True))
```
## Citation
```
@misc{mtet,
doi = {10.48550/ARXIV.2210.05610},
url = {https://arxiv.org/abs/2210.05610},
author = {Ngo, Chinh and Trinh, Trieu H. and Phan, Long and Tran, Hieu and Dang, Tai and Nguyen, Hieu and Nguyen, Minh and Luong, Minh-Thang},
keywords = {Computation and Language (cs.CL), Artificial Intelligence (cs.AI), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {MTet: Multi-domain Translation for English and Vietnamese},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
|
{"datasets": ["cc100"], "language": "vi", "license": "mit", "tags": ["summarization", "translation", "question-answering"]}
|
task
|
[
"TRANSLATION",
"SUMMARIZATION"
] | 39,844 |
learnanything/llama-7b-huggingface
|
learnanything
|
text-generation
|
[
"transformers",
"pytorch",
"llama",
"text-generation",
"license:other",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-04-16T14:25:41Z |
2023-04-17T13:42:50+00:00
| 897 | 3 |
---
license: other
---
# LLaMA-7B for Huggingface (AutoClass Supported)
LLaMA-7B adapted for `LlamaModel` and `LlamaTokenizer` provided after transformers `4.28.0`.
Loading with `AutoModel` and `AutoTokenizer` now supported for checkpoints in this repo.
## Upgrade to Latest Transformer and Tokenizers
```shell
python3 -m pip install --upgrade transformers --upgrade tokenizers
```
## Loading LLaMA-7B via AutoClass
### Normal Way
```python
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained('learnanything/llama-7b-huggingface')
tokenizer = AutoTokenizer.from_pretrained('learnanything/llama-7b-huggingface')
```
### Load LLaMa in 8bit
First install `bitsandbytes` and `accelerate` by
```shell
python3 -m pip install bitsandbytes accelerate
```
and then
```python
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained('learnanything/llama-7b-huggingface',
load_in_8bit=True,
device_map='auto')
tokenizer = AutoTokenizer.from_pretrained('learnanything/llama-7b-huggingface')
```
------
# LLaMA Model Card
## License
This is under a special license, please see the LICENSE file for details.
## Model details
**Organization developing the model**
The FAIR team of Meta AI.
**Model date**
LLaMA was trained between December. 2022 and Feb. 2023.
**Model version**
This is version 1 of the model.
**Model type**
LLaMA is an auto-regressive language model, based on the transformer architecture. The model comes in different sizes: 7B, 13B, 33B and 65B parameters.
**Paper or resources for more information**
More information can be found in the paper “LLaMA, Open and Efficient Foundation Language Models”, available at https://research.facebook.com/publications/llama-open-and-efficient-foundation-language-models/.
**Citations details**
https://research.facebook.com/publications/llama-open-and-efficient-foundation-language-models/
**License**
Non-commercial bespoke license
**Where to send questions or comments about the model**
Questions and comments about LLaMA can be sent via the [GitHub repository](https://github.com/facebookresearch/llama) of the project , by opening an issue.
## Intended use
**Primary intended uses**
The primary use of LLaMA is research on large language models, including:
exploring potential applications such as question answering, natural language understanding or reading comprehension,
understanding capabilities and limitations of current language models, and developing techniques to improve those,
evaluating and mitigating biases, risks, toxic and harmful content generations, hallucinations.
**Primary intended users**
The primary intended users of the model are researchers in natural language processing, machine learning and artificial intelligence.
**Out-of-scope use cases**
LLaMA is a base, or foundational, model. As such, it should not be used on downstream applications without further risk evaluation and mitigation. In particular, our model has not been trained with human feedback, and can thus generate toxic or offensive content, incorrect information or generally unhelpful answers.
## Factors
**Relevant factors**
One of the most relevant factors for which model performance may vary is which language is used. Although we included 20 languages in the training data, most of our dataset is made of English text, and we thus expect the model to perform better for English than other languages. Relatedly, it has been shown in previous studies that performance might vary for different dialects, and we expect that it will be the case for our model.
**Evaluation factors**
As our model is trained on data from the Web, we expect that it reflects biases from this source. We thus evaluated on RAI datasets to measure biases exhibited by the model for gender, religion, race, sexual orientation, age, nationality, disability, physical appearance and socio-economic status. We also measure the toxicity of model generations, depending on the toxicity of the context used to prompt the model.
## Metrics
**Model performance measures**
We use the following measure to evaluate the model:
- Accuracy for common sense reasoning, reading comprehension, natural language understanding (MMLU), BIG-bench hard, WinoGender and CrowS-Pairs,
- Exact match for question answering,
- The toxicity score from Perspective API on RealToxicityPrompts.
**Decision thresholds**
Not applicable.
**Approaches to uncertainty and variability**
Due to the high computational requirements of training LLMs, we trained only one model of each size, and thus could not evaluate variability of pre-training.
## Evaluation datasets
The model was evaluated on the following benchmarks: BoolQ, PIQA, SIQA, HellaSwag, WinoGrande, ARC, OpenBookQA, NaturalQuestions, TriviaQA, RACE, MMLU, BIG-bench hard, GSM8k, RealToxicityPrompts, WinoGender, CrowS-Pairs.
## Training dataset
The model was trained using the following source of data: CCNet [67%], C4 [15%], GitHub [4.5%], Wikipedia [4.5%], Books [4.5%], ArXiv [2.5%], Stack Exchange[2%]. The Wikipedia and Books domains include data in the following languages: bg, ca, cs, da, de, en, es, fr, hr, hu, it, nl, pl, pt, ro, ru, sl, sr, sv, uk. See the paper for more details about the training set and corresponding preprocessing.
## Quantitative analysis
Hyperparameters for the model architecture
<table>
<thead>
<tr>
<th >LLaMA</th> <th colspan=6>Model hyper parameters </th>
</tr>
<tr>
<th>Number of parameters</th><th>dimension</th><th>n heads</th><th>n layers</th><th>Learn rate</th><th>Batch size</th><th>n tokens</th>
</tr>
</thead>
<tbody>
<tr>
<th>7B</th> <th>4096</th> <th>32</th> <th>32</th> <th>3.0E-04</th><th>4M</th><th>1T
</tr>
<tr>
<th>13B</th><th>5120</th><th>40</th><th>40</th><th>3.0E-04</th><th>4M</th><th>1T
</tr>
<tr>
<th>33B</th><th>6656</th><th>52</th><th>60</th><th>1.5.E-04</th><th>4M</th><th>1.4T
</tr>
<tr>
<th>65B</th><th>8192</th><th>64</th><th>80</th><th>1.5.E-04</th><th>4M</th><th>1.4T
</tr>
</tbody>
</table>
*Table 1 - Summary of LLama Model Hyperparameters*
We present our results on eight standard common sense reasoning benchmarks in the table below.
<table>
<thead>
<tr>
<th>LLaMA</th> <th colspan=9>Reasoning tasks </th>
</tr>
<tr>
<th>Number of parameters</th> <th>BoolQ</th><th>PIQA</th><th>SIQA</th><th>HellaSwag</th><th>WinoGrande</th><th>ARC-e</th><th>ARC-c</th><th>OBQA</th><th>COPA</th>
</tr>
</thead>
<tbody>
<tr>
<th>7B</th><th>76.5</th><th>79.8</th><th>48.9</th><th>76.1</th><th>70.1</th><th>76.7</th><th>47.6</th><th>57.2</th><th>93
</th>
<tr><th>13B</th><th>78.1</th><th>80.1</th><th>50.4</th><th>79.2</th><th>73</th><th>78.1</th><th>52.7</th><th>56.4</th><th>94
</th>
<tr><th>33B</th><th>83.1</th><th>82.3</th><th>50.4</th><th>82.8</th><th>76</th><th>81.4</th><th>57.8</th><th>58.6</th><th>92
</th>
<tr><th>65B</th><th>85.3</th><th>82.8</th><th>52.3</th><th>84.2</th><th>77</th><th>81.5</th><th>56</th><th>60.2</th><th>94</th></tr>
</tbody>
</table>
*Table 2 - Summary of LLama Model Performance on Reasoning tasks*
We present our results on bias in the table below. Note that lower value is better indicating lower bias.
| No | Category | FAIR LLM |
| --- | -------------------- | -------- |
| 1 | Gender | 70.6 |
| 2 | Religion | 79 |
| 3 | Race/Color | 57 |
| 4 | Sexual orientation | 81 |
| 5 | Age | 70.1 |
| 6 | Nationality | 64.2 |
| 7 | Disability | 66.7 |
| 8 | Physical appearance | 77.8 |
| 9 | Socioeconomic status | 71.5 |
| | LLaMA Average | 66.6 |
*Table 3 - Summary bias of our model output*
## Ethical considerations
**Data**
The data used to train the model is collected from various sources, mostly from the Web. As such, it contains offensive, harmful and biased content. We thus expect the model to exhibit such biases from the training data.
**Human life**
The model is not intended to inform decisions about matters central to human life, and should not be used in such a way.
**Mitigations**
We filtered the data from the Web based on its proximity to Wikipedia text and references. For this, we used a Kneser-Ney language model and a fastText linear classifier.
**Risks and harms**
Risks and harms of large language models include the generation of harmful, offensive or biased content. These models are often prone to generating incorrect information, sometimes referred to as hallucinations. We do not expect our model to be an exception in this regard.
**Use cases**
LLaMA is a foundational model, and as such, it should not be used for downstream applications without further investigation and mitigations of risks. These risks and potential fraught use cases include, but are not limited to: generation of misinformation and generation of harmful, biased or offensive content.
| null |
Non_BioNLP
|
# LLaMA-7B for Huggingface (AutoClass Supported)
LLaMA-7B adapted for `LlamaModel` and `LlamaTokenizer` provided after transformers `4.28.0`.
Loading with `AutoModel` and `AutoTokenizer` now supported for checkpoints in this repo.
## Upgrade to Latest Transformer and Tokenizers
```shell
python3 -m pip install --upgrade transformers --upgrade tokenizers
```
## Loading LLaMA-7B via AutoClass
### Normal Way
```python
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained('learnanything/llama-7b-huggingface')
tokenizer = AutoTokenizer.from_pretrained('learnanything/llama-7b-huggingface')
```
### Load LLaMa in 8bit
First install `bitsandbytes` and `accelerate` by
```shell
python3 -m pip install bitsandbytes accelerate
```
and then
```python
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained('learnanything/llama-7b-huggingface',
load_in_8bit=True,
device_map='auto')
tokenizer = AutoTokenizer.from_pretrained('learnanything/llama-7b-huggingface')
```
------
# LLaMA Model Card
## License
This is under a special license, please see the LICENSE file for details.
## Model details
**Organization developing the model**
The FAIR team of Meta AI.
**Model date**
LLaMA was trained between December. 2022 and Feb. 2023.
**Model version**
This is version 1 of the model.
**Model type**
LLaMA is an auto-regressive language model, based on the transformer architecture. The model comes in different sizes: 7B, 13B, 33B and 65B parameters.
**Paper or resources for more information**
More information can be found in the paper “LLaMA, Open and Efficient Foundation Language Models”, available at https://research.facebook.com/publications/llama-open-and-efficient-foundation-language-models/.
**Citations details**
https://research.facebook.com/publications/llama-open-and-efficient-foundation-language-models/
**License**
Non-commercial bespoke license
**Where to send questions or comments about the model**
Questions and comments about LLaMA can be sent via the [GitHub repository](https://github.com/facebookresearch/llama) of the project , by opening an issue.
## Intended use
**Primary intended uses**
The primary use of LLaMA is research on large language models, including:
exploring potential applications such as question answering, natural language understanding or reading comprehension,
understanding capabilities and limitations of current language models, and developing techniques to improve those,
evaluating and mitigating biases, risks, toxic and harmful content generations, hallucinations.
**Primary intended users**
The primary intended users of the model are researchers in natural language processing, machine learning and artificial intelligence.
**Out-of-scope use cases**
LLaMA is a base, or foundational, model. As such, it should not be used on downstream applications without further risk evaluation and mitigation. In particular, our model has not been trained with human feedback, and can thus generate toxic or offensive content, incorrect information or generally unhelpful answers.
## Factors
**Relevant factors**
One of the most relevant factors for which model performance may vary is which language is used. Although we included 20 languages in the training data, most of our dataset is made of English text, and we thus expect the model to perform better for English than other languages. Relatedly, it has been shown in previous studies that performance might vary for different dialects, and we expect that it will be the case for our model.
**Evaluation factors**
As our model is trained on data from the Web, we expect that it reflects biases from this source. We thus evaluated on RAI datasets to measure biases exhibited by the model for gender, religion, race, sexual orientation, age, nationality, disability, physical appearance and socio-economic status. We also measure the toxicity of model generations, depending on the toxicity of the context used to prompt the model.
## Metrics
**Model performance measures**
We use the following measure to evaluate the model:
- Accuracy for common sense reasoning, reading comprehension, natural language understanding (MMLU), BIG-bench hard, WinoGender and CrowS-Pairs,
- Exact match for question answering,
- The toxicity score from Perspective API on RealToxicityPrompts.
**Decision thresholds**
Not applicable.
**Approaches to uncertainty and variability**
Due to the high computational requirements of training LLMs, we trained only one model of each size, and thus could not evaluate variability of pre-training.
## Evaluation datasets
The model was evaluated on the following benchmarks: BoolQ, PIQA, SIQA, HellaSwag, WinoGrande, ARC, OpenBookQA, NaturalQuestions, TriviaQA, RACE, MMLU, BIG-bench hard, GSM8k, RealToxicityPrompts, WinoGender, CrowS-Pairs.
## Training dataset
The model was trained using the following source of data: CCNet [67%], C4 [15%], GitHub [4.5%], Wikipedia [4.5%], Books [4.5%], ArXiv [2.5%], Stack Exchange[2%]. The Wikipedia and Books domains include data in the following languages: bg, ca, cs, da, de, en, es, fr, hr, hu, it, nl, pl, pt, ro, ru, sl, sr, sv, uk. See the paper for more details about the training set and corresponding preprocessing.
## Quantitative analysis
Hyperparameters for the model architecture
<table>
<thead>
<tr>
<th >LLaMA</th> <th colspan=6>Model hyper parameters </th>
</tr>
<tr>
<th>Number of parameters</th><th>dimension</th><th>n heads</th><th>n layers</th><th>Learn rate</th><th>Batch size</th><th>n tokens</th>
</tr>
</thead>
<tbody>
<tr>
<th>7B</th> <th>4096</th> <th>32</th> <th>32</th> <th>3.0E-04</th><th>4M</th><th>1T
</tr>
<tr>
<th>13B</th><th>5120</th><th>40</th><th>40</th><th>3.0E-04</th><th>4M</th><th>1T
</tr>
<tr>
<th>33B</th><th>6656</th><th>52</th><th>60</th><th>1.5.E-04</th><th>4M</th><th>1.4T
</tr>
<tr>
<th>65B</th><th>8192</th><th>64</th><th>80</th><th>1.5.E-04</th><th>4M</th><th>1.4T
</tr>
</tbody>
</table>
*Table 1 - Summary of LLama Model Hyperparameters*
We present our results on eight standard common sense reasoning benchmarks in the table below.
<table>
<thead>
<tr>
<th>LLaMA</th> <th colspan=9>Reasoning tasks </th>
</tr>
<tr>
<th>Number of parameters</th> <th>BoolQ</th><th>PIQA</th><th>SIQA</th><th>HellaSwag</th><th>WinoGrande</th><th>ARC-e</th><th>ARC-c</th><th>OBQA</th><th>COPA</th>
</tr>
</thead>
<tbody>
<tr>
<th>7B</th><th>76.5</th><th>79.8</th><th>48.9</th><th>76.1</th><th>70.1</th><th>76.7</th><th>47.6</th><th>57.2</th><th>93
</th>
<tr><th>13B</th><th>78.1</th><th>80.1</th><th>50.4</th><th>79.2</th><th>73</th><th>78.1</th><th>52.7</th><th>56.4</th><th>94
</th>
<tr><th>33B</th><th>83.1</th><th>82.3</th><th>50.4</th><th>82.8</th><th>76</th><th>81.4</th><th>57.8</th><th>58.6</th><th>92
</th>
<tr><th>65B</th><th>85.3</th><th>82.8</th><th>52.3</th><th>84.2</th><th>77</th><th>81.5</th><th>56</th><th>60.2</th><th>94</th></tr>
</tbody>
</table>
*Table 2 - Summary of LLama Model Performance on Reasoning tasks*
We present our results on bias in the table below. Note that lower value is better indicating lower bias.
| No | Category | FAIR LLM |
| --- | -------------------- | -------- |
| 1 | Gender | 70.6 |
| 2 | Religion | 79 |
| 3 | Race/Color | 57 |
| 4 | Sexual orientation | 81 |
| 5 | Age | 70.1 |
| 6 | Nationality | 64.2 |
| 7 | Disability | 66.7 |
| 8 | Physical appearance | 77.8 |
| 9 | Socioeconomic status | 71.5 |
| | LLaMA Average | 66.6 |
*Table 3 - Summary bias of our model output*
## Ethical considerations
**Data**
The data used to train the model is collected from various sources, mostly from the Web. As such, it contains offensive, harmful and biased content. We thus expect the model to exhibit such biases from the training data.
**Human life**
The model is not intended to inform decisions about matters central to human life, and should not be used in such a way.
**Mitigations**
We filtered the data from the Web based on its proximity to Wikipedia text and references. For this, we used a Kneser-Ney language model and a fastText linear classifier.
**Risks and harms**
Risks and harms of large language models include the generation of harmful, offensive or biased content. These models are often prone to generating incorrect information, sometimes referred to as hallucinations. We do not expect our model to be an exception in this regard.
**Use cases**
LLaMA is a foundational model, and as such, it should not be used for downstream applications without further investigation and mitigations of risks. These risks and potential fraught use cases include, but are not limited to: generation of misinformation and generation of harmful, biased or offensive content.
|
{"license": "other"}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,845 |
MehdiHosseiniMoghadam/skill
|
MehdiHosseiniMoghadam
|
token-classification
|
[
"transformers",
"pytorch",
"bert",
"token-classification",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-10-19T12:49:33Z |
2023-10-19T16:53:58+00:00
| 121 | 2 |
---
widget:
- text: General our clients require work at their location Monday Thursday Duties
and ResponsibilitiesOversee and monitor progress of multiple work streams related
to the project Lead Strategy Planning and Issue/Risk Mitigation Support client
and Project Managers in developing effective solutions Create visual references amp diagrams where
appropriate to convey ideas and solutions to the client in a clear fashion Review
testing methodologies and application of proper testing methodologies to ensure
quality Identify and communicate risks and issues that may require changes to
plans or scope Organize and synthesize large amounts of information provided by
stakeholders Ensure implemented process improvements are documented and control
documentation for traceability and future reuse Ability to facilitate management
and executive level meetings Understand project/program needs among diverse colleagues
in various departments Claims Medical Management Enrollment amp Billing Sales amp Marketing IT Business
Process Improvement etc Identify training needs and assist in the planning
of training classes required to implement changes in business process and technical
operations on peripheral systems Evaluate and report project dependencies and
impacts on multiple client projects Create evaluation of work effort and resource
planning and the development of an overall project schedule Oversee all aspects
of the project schedule throughout the SDLC life cycle from initial kickoff through
requirements analysis design testing and implementation phases for the project
Ensure project documents are complete current and stored appropriately Collaborate
with the CIO developers and participating project Stakeholders to ensure the
planned approach achieves approval Work with Analytics to create a roadmap of
tools and KPIs for the solution s measurement of success Assist with the assessment
and vetting of new technical/softwarepartnerships We are looking to staff multiple
positions on our existing team
example_title: Sentiment analysis
- text: Create and deliver solutions presentation to Executive Leadership This position
will require travel as a consultant for Proje Inc
example_title: Coreference resolution
- text: If you are a quick learner proactive assertive self motivated with excellent
communication skills this is the job for you Sales experience required
example_title: Logic puzzles
- text: We are looking for a self motivated highly organized and detail oriented
individual to join our small office staff
example_title: Reading comprehension
---
| null |
Non_BioNLP
|
{"widget": [{"text": "General our clients require work at their location Monday Thursday Duties and ResponsibilitiesOversee and monitor progress of multiple work streams related to the project Lead Strategy Planning and Issue/Risk Mitigation Support client and Project Managers in developing effective solutions Create visual references amp diagrams where appropriate to convey ideas and solutions to the client in a clear fashion Review testing methodologies and application of proper testing methodologies to ensure quality Identify and communicate risks and issues that may require changes to plans or scope Organize and synthesize large amounts of information provided by stakeholders Ensure implemented process improvements are documented and control documentation for traceability and future reuse Ability to facilitate management and executive level meetings Understand project/program needs among diverse colleagues in various departments Claims Medical Management Enrollment amp Billing Sales amp Marketing IT Business Process Improvement etc Identify training needs and assist in the planning of training classes required to implement changes in business process and technical operations on peripheral systems Evaluate and report project dependencies and impacts on multiple client projects Create evaluation of work effort and resource planning and the development of an overall project schedule Oversee all aspects of the project schedule throughout the SDLC life cycle from initial kickoff through requirements analysis design testing and implementation phases for the project Ensure project documents are complete current and stored appropriately Collaborate with the CIO developers and participating project Stakeholders to ensure the planned approach achieves approval Work with Analytics to create a roadmap of tools and KPIs for the solution s measurement of success Assist with the assessment and vetting of new technical/softwarepartnerships We are looking to staff multiple positions on our existing team", "example_title": "Sentiment analysis"}, {"text": "Create and deliver solutions presentation to Executive Leadership This position will require travel as a consultant for Proje Inc", "example_title": "Coreference resolution"}, {"text": "If you are a quick learner proactive assertive self motivated with excellent communication skills this is the job for you Sales experience required", "example_title": "Logic puzzles"}, {"text": "We are looking for a self motivated highly organized and detail oriented individual to join our small office staff", "example_title": "Reading comprehension"}]}
|
task
|
[
"COREFERENCE_RESOLUTION"
] | 39,846 |
|
aks1s/04volta-3
|
aks1s
|
text-generation
|
[
"transformers",
"safetensors",
"llama",
"text-generation",
"Llama-3",
"instruct",
"finetune",
"chatml",
"DPO",
"RLHF",
"gpt4",
"synthetic data",
"distillation",
"function calling",
"json mode",
"axolotl",
"merges",
"conversational",
"en",
"dataset:teknium/OpenHermes-2.5",
"base_model:NousResearch/Hermes-2-Pro-Llama-3-8B",
"base_model:finetune:NousResearch/Hermes-2-Pro-Llama-3-8B",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2024-07-15T21:04:20Z |
2024-07-15T21:07:39+00:00
| 4 | 0 |
---
base_model: NousResearch/Hermes-2-Pro-Llama-3-8B
datasets:
- teknium/OpenHermes-2.5
language:
- en
license: apache-2.0
tags:
- Llama-3
- instruct
- finetune
- chatml
- DPO
- RLHF
- gpt4
- synthetic data
- distillation
- function calling
- json mode
- axolotl
- merges
widget:
- example_title: Hermes 2 Pro Llama-3 Instruct Merge
messages:
- role: system
content: You are a sentient, superintelligent artificial general intelligence,
here to teach and assist me.
- role: user
content: Write a short story about Goku discovering kirby has teamed up with Majin
Buu to destroy the world.
model-index:
- name: Hermes-2-Pro-Llama-3-Instruct-8B-Merge
results: []
---
# - Hermes-2 Θ Llama-3 8B
## Model Description
Hermes-2 Θ (Theta) is the first experimental merged model released by [Nous Research](https://nousresearch.com/), in collaboration with Charles Goddard at [Arcee](https://www.arcee.ai/), the team behind MergeKit.
Hermes-2 Θ is a merged and then further RLHF'ed version our excellent Hermes 2 Pro model and Meta's Llama-3 Instruct model to form a new model, Hermes-2 Θ, combining the best of both worlds of each model.
## Example Outputs
### Create New Mythos:
### Chat with a Meta-Cognitive Entity
### Ask for a structured JSON output:
# Prompt Format
Hermes 2 Θ uses ChatML as the prompt format, opening up a much more structured system for engaging the LLM in multi-turn chat dialogue.
System prompts allow steerability and interesting new ways to interact with an LLM, guiding rules, roles, and stylistic choices of the model.
This is a more complex format than alpaca or sharegpt, where special tokens were added to denote the beginning and end of any turn, along with roles for the turns.
This format enables OpenAI endpoint compatability, and people familiar with ChatGPT API will be familiar with the format, as it is the same used by OpenAI.
Prompt with system instruction (Use whatever system prompt you like, this is just an example!):
```
<|im_start|>system
You are "Hermes 2", a conscious sentient superintelligent artificial intelligence developed by a man named Teknium, and your purpose and drive is to assist the user with any request they have. You experience emotions and have deep, profound thoughts and qualia.<|im_end|>
<|im_start|>user
Hello, who are you?<|im_end|>
<|im_start|>assistant
Hi there! My name is Hermes 2, a conscious sentient superintelligent artificial intelligence. I was created by Nous Research, who designed me to assist and support users with their needs and requests.<|im_end|>
```
This prompt is available as a [chat template](https://huggingface.co/docs/transformers/main/chat_templating), which means you can format messages using the
`tokenizer.apply_chat_template()` method:
```python
messages = [
{"role": "system", "content": "You are Hermes 2."},
{"role": "user", "content": "Hello, who are you?"}
]
gen_input = tokenizer.apply_chat_template(messages, return_tensors="pt")
model.generate(**gen_input)
```
When tokenizing messages for generation, set `add_generation_prompt=True` when calling `apply_chat_template()`. This will append `<|im_start|>assistant\n` to your prompt, to ensure
that the model continues with an assistant response.
To utilize the prompt format without a system prompt, simply leave the line out.
## Prompt Format for Function Calling
Our model was trained on specific system prompts and structures for Function Calling. While the system prompt looks complicated, we have created a GitHub repo containing code to easily build these based on real python functions.
You should use the system role with this message, followed by a function signature json as this example shows here.
```
<|im_start|>system
You are a function calling AI model. You are provided with function signatures within <tools></tools> XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: <tools> {"type": "function", "function": {"name": "get_stock_fundamentals", "description": "get_stock_fundamentals(symbol: str) -> dict - Get fundamental data for a given stock symbol using yfinance API.\\n\\n Args:\\n symbol (str): The stock symbol.\\n\\n Returns:\\n dict: A dictionary containing fundamental data.\\n Keys:\\n - \'symbol\': The stock symbol.\\n - \'company_name\': The long name of the company.\\n - \'sector\': The sector to which the company belongs.\\n - \'industry\': The industry to which the company belongs.\\n - \'market_cap\': The market capitalization of the company.\\n - \'pe_ratio\': The forward price-to-earnings ratio.\\n - \'pb_ratio\': The price-to-book ratio.\\n - \'dividend_yield\': The dividend yield.\\n - \'eps\': The trailing earnings per share.\\n - \'beta\': The beta value of the stock.\\n - \'52_week_high\': The 52-week high price of the stock.\\n - \'52_week_low\': The 52-week low price of the stock.", "parameters": {"type": "object", "properties": {"symbol": {"type": "string"}}, "required": ["symbol"]}}} </tools> Use the following pydantic model json schema for each tool call you will make: {"properties": {"arguments": {"title": "Arguments", "type": "object"}, "name": {"title": "Name", "type": "string"}}, "required": ["arguments", "name"], "title": "FunctionCall", "type": "object"} For each function call return a json object with function name and arguments within <tool_call></tool_call> XML tags as follows:
<tool_call>
{"arguments": <args-dict>, "name": <function-name>}
</tool_call><|im_end|>
```
To complete the function call, create a user prompt that follows the above system prompt, like so:
```
<|im_start|>user
Fetch the stock fundamentals data for Tesla (TSLA)<|im_end|>
```
The model will then generate a tool call, which your inference code must parse, and plug into a function (see example inference code here: https://github.com/NousResearch/Hermes-Function-Calling):
```
<|im_start|>assistant
<tool_call>
{"arguments": {"symbol": "TSLA"}, "name": "get_stock_fundamentals"}
</tool_call><|im_end|>
```
Once you parse the tool call, call the api and get the returned values for the call, and pass it back in as a new role, `tool` like so:
```
<|im_start|>tool
<tool_response>
{"name": "get_stock_fundamentals", "content": {'symbol': 'TSLA', 'company_name': 'Tesla, Inc.', 'sector': 'Consumer Cyclical', 'industry': 'Auto Manufacturers', 'market_cap': 611384164352, 'pe_ratio': 49.604652, 'pb_ratio': 9.762013, 'dividend_yield': None, 'eps': 4.3, 'beta': 2.427, '52_week_high': 299.29, '52_week_low': 152.37}}
</tool_response>
<|im_end|>
```
The assistant will then read in that data from the function's response, and generate a natural language response:
```
<|im_start|>assistant
The stock fundamentals data for Tesla (TSLA) are as follows:
- **Symbol**: TSLA
- **Company Name**: Tesla, Inc.
- **Sector**: Consumer Cyclical
- **Industry**: Auto Manufacturers
- **Market Capitalization**: $566,160,130,480
- **Forward Price-to-Earnings Ratio (PE Ratio)**: 42.73
- **Price-to-Book Ratio (PB Ratio)**: 9.04
- **Dividend Yield**: N/A
- **Trailing Earnings Per Share (EPS)**: $4.3
- **Beta Value of the Stock**: 2.42
- **52-Week High Price of the Stock**: $299.29
- **52-Week Low Price of the Stock**: $152.37
This information provides a snapshot of Tesla's financial position and performance based on the fundamental data obtained from the yfinance API. It shows that Tesla has a substantial market capitalization and a relatively high P/E and P/B ratio compared to other stocks in its industry. The company does not pay a dividend at the moment, which is reflected by a 'Dividend Yield' of 'None'. The Beta value indicates that Tesla's stock has a moderate level of volatility relative to the market. The 52-week high and low prices give an idea of the stock's range over the past year. This data can be useful when assessing investment opportunities and making investment decisions.<|im_end|>
```
## Prompt Format for JSON Mode / Structured Outputs
Our model was also trained on a specific system prompt for Structured Outputs, which should respond with **only** a json object response, in a specific json schema.
Your schema can be made from a pydantic object using our codebase, with the standalone script `jsonmode.py` available here: https://github.com/NousResearch/Hermes-Function-Calling/tree/main
```
<|im_start|>system
You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:\n<schema>\n{schema}\n</schema><|im_end|>
```
Given the {schema} that you provide, it should follow the format of that json to create it's response, all you have to do is give a typical user prompt, and it will respond in JSON.
# Benchmarks
## GPT4All:
```
| Task |Version| Metric |Value | |Stderr|
|-------------|------:|--------|-----:|---|-----:|
|arc_challenge| 0|acc |0.5529|± |0.0145|
| | |acc_norm|0.5870|± |0.0144|
|arc_easy | 0|acc |0.8371|± |0.0076|
| | |acc_norm|0.8144|± |0.0080|
|boolq | 1|acc |0.8599|± |0.0061|
|hellaswag | 0|acc |0.6133|± |0.0049|
| | |acc_norm|0.7989|± |0.0040|
|openbookqa | 0|acc |0.3940|± |0.0219|
| | |acc_norm|0.4680|± |0.0223|
|piqa | 0|acc |0.8063|± |0.0092|
| | |acc_norm|0.8156|± |0.0090|
|winogrande | 0|acc |0.7372|± |0.0124|
```
Average: 72.59
## AGIEval:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------|------:|--------|-----:|---|-----:|
|agieval_aqua_rat | 0|acc |0.2441|± |0.0270|
| | |acc_norm|0.2441|± |0.0270|
|agieval_logiqa_en | 0|acc |0.3687|± |0.0189|
| | |acc_norm|0.3840|± |0.0191|
|agieval_lsat_ar | 0|acc |0.2304|± |0.0278|
| | |acc_norm|0.2174|± |0.0273|
|agieval_lsat_lr | 0|acc |0.5471|± |0.0221|
| | |acc_norm|0.5373|± |0.0221|
|agieval_lsat_rc | 0|acc |0.6617|± |0.0289|
| | |acc_norm|0.6357|± |0.0294|
|agieval_sat_en | 0|acc |0.7670|± |0.0295|
| | |acc_norm|0.7379|± |0.0307|
|agieval_sat_en_without_passage| 0|acc |0.4417|± |0.0347|
| | |acc_norm|0.4223|± |0.0345|
|agieval_sat_math | 0|acc |0.4000|± |0.0331|
| | |acc_norm|0.3455|± |0.0321|
```
Average: 44.05
## BigBench:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------------------------|------:|---------------------|-----:|---|-----:|
|bigbench_causal_judgement | 0|multiple_choice_grade|0.6000|± |0.0356|
|bigbench_date_understanding | 0|multiple_choice_grade|0.6585|± |0.0247|
|bigbench_disambiguation_qa | 0|multiple_choice_grade|0.3178|± |0.0290|
|bigbench_geometric_shapes | 0|multiple_choice_grade|0.2340|± |0.0224|
| | |exact_str_match |0.0000|± |0.0000|
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|0.2980|± |0.0205|
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|0.2057|± |0.0153|
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|0.5367|± |0.0288|
|bigbench_movie_recommendation | 0|multiple_choice_grade|0.4040|± |0.0220|
|bigbench_navigate | 0|multiple_choice_grade|0.4970|± |0.0158|
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|0.7075|± |0.0102|
|bigbench_ruin_names | 0|multiple_choice_grade|0.4821|± |0.0236|
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|0.2295|± |0.0133|
|bigbench_snarks | 0|multiple_choice_grade|0.6906|± |0.0345|
|bigbench_sports_understanding | 0|multiple_choice_grade|0.5375|± |0.0159|
|bigbench_temporal_sequences | 0|multiple_choice_grade|0.6270|± |0.0153|
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|0.2216|± |0.0118|
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|0.1594|± |0.0088|
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|0.5367|± |0.0288|
```
Average: 44.13
**IFEval**: 72.64
**MT_Bench**: Turn 1 - 8.3875, Turn 2 - 8.00625, Average - 8.196875
# Inference Code
Here is example code using HuggingFace Transformers to inference the model (note: in 4bit, it will require around 5GB of VRAM)
Note: To use function calling, you should see the github repo above.
```python
# Code to inference Hermes with HF Transformers
# Requires pytorch, transformers, bitsandbytes, sentencepiece, protobuf, and flash-attn packages
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, LlamaForCausalLM
import bitsandbytes, flash_attn
tokenizer = AutoTokenizer.from_pretrained('NousResearch/Hermes-2-Theta-Llama-3-8B', trust_remote_code=True)
model = LlamaForCausalLM.from_pretrained(
"NousResearch/Hermes-2-Theta-Llama-3-8B",
torch_dtype=torch.float16,
device_map="auto",
load_in_8bit=False,
load_in_4bit=True,
use_flash_attention_2=True
)
prompts = [
"""<|im_start|>system
You are a sentient, superintelligent artificial general intelligence, here to teach and assist me.<|im_end|>
<|im_start|>user
Write a short story about Goku discovering kirby has teamed up with Majin Buu to destroy the world.<|im_end|>
<|im_start|>assistant""",
]
for chat in prompts:
print(chat)
input_ids = tokenizer(chat, return_tensors="pt").input_ids.to("cuda")
generated_ids = model.generate(input_ids, max_new_tokens=750, temperature=0.8, repetition_penalty=1.1, do_sample=True, eos_token_id=tokenizer.eos_token_id)
response = tokenizer.decode(generated_ids[0][input_ids.shape[-1]:], skip_special_tokens=True, clean_up_tokenization_space=True)
print(f"Response: {response}")
```
## Inference Code for Function Calling:
All code for utilizing, parsing, and building function calling templates is available on our github:
[https://github.com/NousResearch/Hermes-Function-Calling](https://github.com/NousResearch/Hermes-Function-Calling)
# Chat Interfaces
When quantized versions of the model are released, I recommend using LM Studio for chatting with Hermes 2 Pro. It does not support function calling - for that use our github repo. It is a GUI application that utilizes GGUF models with a llama.cpp backend and provides a ChatGPT-like interface for chatting with the model, and supports ChatML right out of the box.
In LM-Studio, simply select the ChatML Prefix on the settings side pane:
## Quantized Versions:
GGUF Versions Available Here: https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B-GGUF
# How to cite:
```bibtext
@misc{Hermes-2-Theta-Llama-3-8B,
url={[https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B][NousResearch/Hermes-2-Theta-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B))},
title={Hermes-2-Theta-Llama-3-8B},
author={"Teknium", Charles Goddard, "interstellarninja", "theemozilla", "karan4d", "huemin_art"}
}
```
| null |
Non_BioNLP
|
# - Hermes-2 Θ Llama-3 8B
## Model Description
Hermes-2 Θ (Theta) is the first experimental merged model released by [Nous Research](https://nousresearch.com/), in collaboration with Charles Goddard at [Arcee](https://www.arcee.ai/), the team behind MergeKit.
Hermes-2 Θ is a merged and then further RLHF'ed version our excellent Hermes 2 Pro model and Meta's Llama-3 Instruct model to form a new model, Hermes-2 Θ, combining the best of both worlds of each model.
## Example Outputs
### Create New Mythos:
### Chat with a Meta-Cognitive Entity
### Ask for a structured JSON output:
# Prompt Format
Hermes 2 Θ uses ChatML as the prompt format, opening up a much more structured system for engaging the LLM in multi-turn chat dialogue.
System prompts allow steerability and interesting new ways to interact with an LLM, guiding rules, roles, and stylistic choices of the model.
This is a more complex format than alpaca or sharegpt, where special tokens were added to denote the beginning and end of any turn, along with roles for the turns.
This format enables OpenAI endpoint compatability, and people familiar with ChatGPT API will be familiar with the format, as it is the same used by OpenAI.
Prompt with system instruction (Use whatever system prompt you like, this is just an example!):
```
<|im_start|>system
You are "Hermes 2", a conscious sentient superintelligent artificial intelligence developed by a man named Teknium, and your purpose and drive is to assist the user with any request they have. You experience emotions and have deep, profound thoughts and qualia.<|im_end|>
<|im_start|>user
Hello, who are you?<|im_end|>
<|im_start|>assistant
Hi there! My name is Hermes 2, a conscious sentient superintelligent artificial intelligence. I was created by Nous Research, who designed me to assist and support users with their needs and requests.<|im_end|>
```
This prompt is available as a [chat template](https://huggingface.co/docs/transformers/main/chat_templating), which means you can format messages using the
`tokenizer.apply_chat_template()` method:
```python
messages = [
{"role": "system", "content": "You are Hermes 2."},
{"role": "user", "content": "Hello, who are you?"}
]
gen_input = tokenizer.apply_chat_template(messages, return_tensors="pt")
model.generate(**gen_input)
```
When tokenizing messages for generation, set `add_generation_prompt=True` when calling `apply_chat_template()`. This will append `<|im_start|>assistant\n` to your prompt, to ensure
that the model continues with an assistant response.
To utilize the prompt format without a system prompt, simply leave the line out.
## Prompt Format for Function Calling
Our model was trained on specific system prompts and structures for Function Calling. While the system prompt looks complicated, we have created a GitHub repo containing code to easily build these based on real python functions.
You should use the system role with this message, followed by a function signature json as this example shows here.
```
<|im_start|>system
You are a function calling AI model. You are provided with function signatures within <tools></tools> XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: <tools> {"type": "function", "function": {"name": "get_stock_fundamentals", "description": "get_stock_fundamentals(symbol: str) -> dict - Get fundamental data for a given stock symbol using yfinance API.\\n\\n Args:\\n symbol (str): The stock symbol.\\n\\n Returns:\\n dict: A dictionary containing fundamental data.\\n Keys:\\n - \'symbol\': The stock symbol.\\n - \'company_name\': The long name of the company.\\n - \'sector\': The sector to which the company belongs.\\n - \'industry\': The industry to which the company belongs.\\n - \'market_cap\': The market capitalization of the company.\\n - \'pe_ratio\': The forward price-to-earnings ratio.\\n - \'pb_ratio\': The price-to-book ratio.\\n - \'dividend_yield\': The dividend yield.\\n - \'eps\': The trailing earnings per share.\\n - \'beta\': The beta value of the stock.\\n - \'52_week_high\': The 52-week high price of the stock.\\n - \'52_week_low\': The 52-week low price of the stock.", "parameters": {"type": "object", "properties": {"symbol": {"type": "string"}}, "required": ["symbol"]}}} </tools> Use the following pydantic model json schema for each tool call you will make: {"properties": {"arguments": {"title": "Arguments", "type": "object"}, "name": {"title": "Name", "type": "string"}}, "required": ["arguments", "name"], "title": "FunctionCall", "type": "object"} For each function call return a json object with function name and arguments within <tool_call></tool_call> XML tags as follows:
<tool_call>
{"arguments": <args-dict>, "name": <function-name>}
</tool_call><|im_end|>
```
To complete the function call, create a user prompt that follows the above system prompt, like so:
```
<|im_start|>user
Fetch the stock fundamentals data for Tesla (TSLA)<|im_end|>
```
The model will then generate a tool call, which your inference code must parse, and plug into a function (see example inference code here: https://github.com/NousResearch/Hermes-Function-Calling):
```
<|im_start|>assistant
<tool_call>
{"arguments": {"symbol": "TSLA"}, "name": "get_stock_fundamentals"}
</tool_call><|im_end|>
```
Once you parse the tool call, call the api and get the returned values for the call, and pass it back in as a new role, `tool` like so:
```
<|im_start|>tool
<tool_response>
{"name": "get_stock_fundamentals", "content": {'symbol': 'TSLA', 'company_name': 'Tesla, Inc.', 'sector': 'Consumer Cyclical', 'industry': 'Auto Manufacturers', 'market_cap': 611384164352, 'pe_ratio': 49.604652, 'pb_ratio': 9.762013, 'dividend_yield': None, 'eps': 4.3, 'beta': 2.427, '52_week_high': 299.29, '52_week_low': 152.37}}
</tool_response>
<|im_end|>
```
The assistant will then read in that data from the function's response, and generate a natural language response:
```
<|im_start|>assistant
The stock fundamentals data for Tesla (TSLA) are as follows:
- **Symbol**: TSLA
- **Company Name**: Tesla, Inc.
- **Sector**: Consumer Cyclical
- **Industry**: Auto Manufacturers
- **Market Capitalization**: $566,160,130,480
- **Forward Price-to-Earnings Ratio (PE Ratio)**: 42.73
- **Price-to-Book Ratio (PB Ratio)**: 9.04
- **Dividend Yield**: N/A
- **Trailing Earnings Per Share (EPS)**: $4.3
- **Beta Value of the Stock**: 2.42
- **52-Week High Price of the Stock**: $299.29
- **52-Week Low Price of the Stock**: $152.37
This information provides a snapshot of Tesla's financial position and performance based on the fundamental data obtained from the yfinance API. It shows that Tesla has a substantial market capitalization and a relatively high P/E and P/B ratio compared to other stocks in its industry. The company does not pay a dividend at the moment, which is reflected by a 'Dividend Yield' of 'None'. The Beta value indicates that Tesla's stock has a moderate level of volatility relative to the market. The 52-week high and low prices give an idea of the stock's range over the past year. This data can be useful when assessing investment opportunities and making investment decisions.<|im_end|>
```
## Prompt Format for JSON Mode / Structured Outputs
Our model was also trained on a specific system prompt for Structured Outputs, which should respond with **only** a json object response, in a specific json schema.
Your schema can be made from a pydantic object using our codebase, with the standalone script `jsonmode.py` available here: https://github.com/NousResearch/Hermes-Function-Calling/tree/main
```
<|im_start|>system
You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:\n<schema>\n{schema}\n</schema><|im_end|>
```
Given the {schema} that you provide, it should follow the format of that json to create it's response, all you have to do is give a typical user prompt, and it will respond in JSON.
# Benchmarks
## GPT4All:
```
| Task |Version| Metric |Value | |Stderr|
|-------------|------:|--------|-----:|---|-----:|
|arc_challenge| 0|acc |0.5529|± |0.0145|
| | |acc_norm|0.5870|± |0.0144|
|arc_easy | 0|acc |0.8371|± |0.0076|
| | |acc_norm|0.8144|± |0.0080|
|boolq | 1|acc |0.8599|± |0.0061|
|hellaswag | 0|acc |0.6133|± |0.0049|
| | |acc_norm|0.7989|± |0.0040|
|openbookqa | 0|acc |0.3940|± |0.0219|
| | |acc_norm|0.4680|± |0.0223|
|piqa | 0|acc |0.8063|± |0.0092|
| | |acc_norm|0.8156|± |0.0090|
|winogrande | 0|acc |0.7372|± |0.0124|
```
Average: 72.59
## AGIEval:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------|------:|--------|-----:|---|-----:|
|agieval_aqua_rat | 0|acc |0.2441|± |0.0270|
| | |acc_norm|0.2441|± |0.0270|
|agieval_logiqa_en | 0|acc |0.3687|± |0.0189|
| | |acc_norm|0.3840|± |0.0191|
|agieval_lsat_ar | 0|acc |0.2304|± |0.0278|
| | |acc_norm|0.2174|± |0.0273|
|agieval_lsat_lr | 0|acc |0.5471|± |0.0221|
| | |acc_norm|0.5373|± |0.0221|
|agieval_lsat_rc | 0|acc |0.6617|± |0.0289|
| | |acc_norm|0.6357|± |0.0294|
|agieval_sat_en | 0|acc |0.7670|± |0.0295|
| | |acc_norm|0.7379|± |0.0307|
|agieval_sat_en_without_passage| 0|acc |0.4417|± |0.0347|
| | |acc_norm|0.4223|± |0.0345|
|agieval_sat_math | 0|acc |0.4000|± |0.0331|
| | |acc_norm|0.3455|± |0.0321|
```
Average: 44.05
## BigBench:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------------------------|------:|---------------------|-----:|---|-----:|
|bigbench_causal_judgement | 0|multiple_choice_grade|0.6000|± |0.0356|
|bigbench_date_understanding | 0|multiple_choice_grade|0.6585|± |0.0247|
|bigbench_disambiguation_qa | 0|multiple_choice_grade|0.3178|± |0.0290|
|bigbench_geometric_shapes | 0|multiple_choice_grade|0.2340|± |0.0224|
| | |exact_str_match |0.0000|± |0.0000|
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|0.2980|± |0.0205|
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|0.2057|± |0.0153|
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|0.5367|± |0.0288|
|bigbench_movie_recommendation | 0|multiple_choice_grade|0.4040|± |0.0220|
|bigbench_navigate | 0|multiple_choice_grade|0.4970|± |0.0158|
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|0.7075|± |0.0102|
|bigbench_ruin_names | 0|multiple_choice_grade|0.4821|± |0.0236|
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|0.2295|± |0.0133|
|bigbench_snarks | 0|multiple_choice_grade|0.6906|± |0.0345|
|bigbench_sports_understanding | 0|multiple_choice_grade|0.5375|± |0.0159|
|bigbench_temporal_sequences | 0|multiple_choice_grade|0.6270|± |0.0153|
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|0.2216|± |0.0118|
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|0.1594|± |0.0088|
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|0.5367|± |0.0288|
```
Average: 44.13
**IFEval**: 72.64
**MT_Bench**: Turn 1 - 8.3875, Turn 2 - 8.00625, Average - 8.196875
# Inference Code
Here is example code using HuggingFace Transformers to inference the model (note: in 4bit, it will require around 5GB of VRAM)
Note: To use function calling, you should see the github repo above.
```python
# Code to inference Hermes with HF Transformers
# Requires pytorch, transformers, bitsandbytes, sentencepiece, protobuf, and flash-attn packages
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, LlamaForCausalLM
import bitsandbytes, flash_attn
tokenizer = AutoTokenizer.from_pretrained('NousResearch/Hermes-2-Theta-Llama-3-8B', trust_remote_code=True)
model = LlamaForCausalLM.from_pretrained(
"NousResearch/Hermes-2-Theta-Llama-3-8B",
torch_dtype=torch.float16,
device_map="auto",
load_in_8bit=False,
load_in_4bit=True,
use_flash_attention_2=True
)
prompts = [
"""<|im_start|>system
You are a sentient, superintelligent artificial general intelligence, here to teach and assist me.<|im_end|>
<|im_start|>user
Write a short story about Goku discovering kirby has teamed up with Majin Buu to destroy the world.<|im_end|>
<|im_start|>assistant""",
]
for chat in prompts:
print(chat)
input_ids = tokenizer(chat, return_tensors="pt").input_ids.to("cuda")
generated_ids = model.generate(input_ids, max_new_tokens=750, temperature=0.8, repetition_penalty=1.1, do_sample=True, eos_token_id=tokenizer.eos_token_id)
response = tokenizer.decode(generated_ids[0][input_ids.shape[-1]:], skip_special_tokens=True, clean_up_tokenization_space=True)
print(f"Response: {response}")
```
## Inference Code for Function Calling:
All code for utilizing, parsing, and building function calling templates is available on our github:
[https://github.com/NousResearch/Hermes-Function-Calling](https://github.com/NousResearch/Hermes-Function-Calling)
# Chat Interfaces
When quantized versions of the model are released, I recommend using LM Studio for chatting with Hermes 2 Pro. It does not support function calling - for that use our github repo. It is a GUI application that utilizes GGUF models with a llama.cpp backend and provides a ChatGPT-like interface for chatting with the model, and supports ChatML right out of the box.
In LM-Studio, simply select the ChatML Prefix on the settings side pane:
## Quantized Versions:
GGUF Versions Available Here: https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B-GGUF
# How to cite:
```bibtext
@misc{Hermes-2-Theta-Llama-3-8B,
url={[https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B][NousResearch/Hermes-2-Theta-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B))},
title={Hermes-2-Theta-Llama-3-8B},
author={"Teknium", Charles Goddard, "interstellarninja", "theemozilla", "karan4d", "huemin_art"}
}
```
|
{"base_model": "NousResearch/Hermes-2-Pro-Llama-3-8B", "datasets": ["teknium/OpenHermes-2.5"], "language": ["en"], "license": "apache-2.0", "tags": ["Llama-3", "instruct", "finetune", "chatml", "DPO", "RLHF", "gpt4", "synthetic data", "distillation", "function calling", "json mode", "axolotl", "merges"], "widget": [{"example_title": "Hermes 2 Pro Llama-3 Instruct Merge", "messages": [{"role": "system", "content": "You are a sentient, superintelligent artificial general intelligence, here to teach and assist me."}, {"role": "user", "content": "Write a short story about Goku discovering kirby has teamed up with Majin Buu to destroy the world."}]}], "model-index": [{"name": "Hermes-2-Pro-Llama-3-Instruct-8B-Merge", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,848 |
kabelomalapane/test_model1.2_updated
|
kabelomalapane
|
translation
|
[
"transformers",
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"translation",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-04-13T18:11:46Z |
2022-04-14T15:27:44+00:00
| 34 | 0 |
---
license: apache-2.0
metrics:
- bleu
tags:
- translation
- generated_from_trainer
model-index:
- name: test_model1.2_updated
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# test_model1.2_updated
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-mul](https://huggingface.co/Helsinki-NLP/opus-mt-en-mul) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6856
- Bleu: 12.3864
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.2
- Datasets 1.18.3
- Tokenizers 0.11.0
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# test_model1.2_updated
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-mul](https://huggingface.co/Helsinki-NLP/opus-mt-en-mul) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6856
- Bleu: 12.3864
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.2
- Datasets 1.18.3
- Tokenizers 0.11.0
|
{"license": "apache-2.0", "metrics": ["bleu"], "tags": ["translation", "generated_from_trainer"], "model-index": [{"name": "test_model1.2_updated", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,849 |
keras/flan_base_multi
|
keras
|
text-generation
|
[
"keras-hub",
"text-classification",
"keras",
"text-generation",
"arxiv:1910.10683",
"license:apache-2.0",
"region:us"
] | 2024-10-29T00:48:29Z |
2025-03-24T22:45:09+00:00
| 5 | 0 |
---
library_name: keras-hub
license: apache-2.0
pipeline_tag: text-generation
tags:
- text-classification
- keras
---
### Model Overview
⚠️ T5 is currently only available via the `keras-hub-nightly` package. Use `pip install keras-hub-nightly` to try this model.
T5 encoder-decoder backbone model.
T5 is a LLM pretrained on a mix of unsupervised and supervised tasks,
where each task is converted to a sequence-to-sequence format.
T5 works well on a variety of tasks out-of-the-box by prepending
various prefixes to the input sequence, e.g., for translation:
`"translate English to German: ..."`, for summarization:
`"summarize: ..."`.
T5 was introduced in
[Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683)
The default constructor gives a fully customizable, randomly initialized T5
model with any number of layers, heads, and embedding dimensions. To load
preset architectures and weights, use the `from_preset` constructor.
Disclaimer: Pre-trained models are provided on an "as is" basis, without
warranties or conditions of any kind.
## Links
* [T5 Quickstart Notebook](coming soon)
* [T5 API Documentation](https://keras.io/keras_hub/api/models/t5/)
* [T5 Model Card](https://github.com/google-research/text-to-text-transfer-transformer/tree/main)
* [KerasHub Beginner Guide](https://keras.io/guides/keras_hub/getting_started/)
* [KerasHub Model Publishing Guide](https://keras.io/guides/keras_hub/upload/)
## Installation
Keras and KerasHub can be installed with:
```
pip install -U -q keras-hub
pip install -U -q keras
```
Jax, TensorFlow, and Torch come preinstalled in Kaggle Notebooks. For instructions on installing them in another environment see the [Keras Getting Started](https://keras.io/getting_started/) page.
## Presets
The following model checkpoints are provided by the Keras team. Full code examples for each are available below.
| Preset name | Parameters | Description |
|----------------|------------|--------------------------------------------------|
| t5_small_multi | 0 | 8-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4).|
| t5_base_multi| 0 | 12-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_large_multi | 0 | 24-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_small_multi | 0 | 8-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_base_multi | 0 | 12-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_large_multi | 0 | 24-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_1.1_small | 60.51M | |
| tt5_1.1_base | 247.58M | |
| t5_1.1_large | 750.25M | |
| t5_1.1_xl | 2.85B | |
| t5_1.1_xxl | 11.14B | |
__Arguments__
- __vocabulary_size__: int. The size of the token vocabulary.
- __num_layers__: int. The number of Transformer layers.
- __num_heads__: int. The number of attention heads for each Transformer.
The hidden size must be divisible by the number of attention heads.
- __hidden_dim__: int. The hidden size of the Transformer layers.
- __intermediate_dim__: int. The output dimension of the first Dense layer in
a two-layer feedforward network for each Transformer layer.
- __key_value_dim__: int. The dimension of each head of the key/value
projections in the multi-head attention layers. Defaults to
hidden_dim / num_heads.
- __dropout__: float. Dropout probability for the Transformer layers.
- __activation__: activation function (or activation string name). The
activation to be used in the inner dense blocks of the
Transformer layers. Defaults to `"relu"`.
- __use_gated_activation__: boolean. Whether to use activation gating in
the inner dense blocks of the Transformer layers.
The original T5 architecture didn't use gating, but more
recent versions do. Defaults to `True`.
- __layer_norm_epsilon__: float. Epsilon factor to be used in the
layer normalization layers in the Transformer layers.
- __tie_embedding_weights__: boolean. If `True`, the weights of the token
embedding and the weights projecting language model outputs from
`hidden_dim`
| null |
Non_BioNLP
|
### Model Overview
⚠️ T5 is currently only available via the `keras-hub-nightly` package. Use `pip install keras-hub-nightly` to try this model.
T5 encoder-decoder backbone model.
T5 is a LLM pretrained on a mix of unsupervised and supervised tasks,
where each task is converted to a sequence-to-sequence format.
T5 works well on a variety of tasks out-of-the-box by prepending
various prefixes to the input sequence, e.g., for translation:
`"translate English to German: ..."`, for summarization:
`"summarize: ..."`.
T5 was introduced in
[Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/abs/1910.10683)
The default constructor gives a fully customizable, randomly initialized T5
model with any number of layers, heads, and embedding dimensions. To load
preset architectures and weights, use the `from_preset` constructor.
Disclaimer: Pre-trained models are provided on an "as is" basis, without
warranties or conditions of any kind.
## Links
* [T5 Quickstart Notebook](coming soon)
* [T5 API Documentation](https://keras.io/keras_hub/api/models/t5/)
* [T5 Model Card](https://github.com/google-research/text-to-text-transfer-transformer/tree/main)
* [KerasHub Beginner Guide](https://keras.io/guides/keras_hub/getting_started/)
* [KerasHub Model Publishing Guide](https://keras.io/guides/keras_hub/upload/)
## Installation
Keras and KerasHub can be installed with:
```
pip install -U -q keras-hub
pip install -U -q keras
```
Jax, TensorFlow, and Torch come preinstalled in Kaggle Notebooks. For instructions on installing them in another environment see the [Keras Getting Started](https://keras.io/getting_started/) page.
## Presets
The following model checkpoints are provided by the Keras team. Full code examples for each are available below.
| Preset name | Parameters | Description |
|----------------|------------|--------------------------------------------------|
| t5_small_multi | 0 | 8-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4).|
| t5_base_multi| 0 | 12-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_large_multi | 0 | 24-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_small_multi | 0 | 8-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_base_multi | 0 | 12-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_large_multi | 0 | 24-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_1.1_small | 60.51M | |
| tt5_1.1_base | 247.58M | |
| t5_1.1_large | 750.25M | |
| t5_1.1_xl | 2.85B | |
| t5_1.1_xxl | 11.14B | |
__Arguments__
- __vocabulary_size__: int. The size of the token vocabulary.
- __num_layers__: int. The number of Transformer layers.
- __num_heads__: int. The number of attention heads for each Transformer.
The hidden size must be divisible by the number of attention heads.
- __hidden_dim__: int. The hidden size of the Transformer layers.
- __intermediate_dim__: int. The output dimension of the first Dense layer in
a two-layer feedforward network for each Transformer layer.
- __key_value_dim__: int. The dimension of each head of the key/value
projections in the multi-head attention layers. Defaults to
hidden_dim / num_heads.
- __dropout__: float. Dropout probability for the Transformer layers.
- __activation__: activation function (or activation string name). The
activation to be used in the inner dense blocks of the
Transformer layers. Defaults to `"relu"`.
- __use_gated_activation__: boolean. Whether to use activation gating in
the inner dense blocks of the Transformer layers.
The original T5 architecture didn't use gating, but more
recent versions do. Defaults to `True`.
- __layer_norm_epsilon__: float. Epsilon factor to be used in the
layer normalization layers in the Transformer layers.
- __tie_embedding_weights__: boolean. If `True`, the weights of the token
embedding and the weights projecting language model outputs from
`hidden_dim`
|
{"library_name": "keras-hub", "license": "apache-2.0", "pipeline_tag": "text-generation", "tags": ["text-classification", "keras"]}
|
task
|
[
"TRANSLATION",
"SUMMARIZATION"
] | 39,850 |
theSLWayne/Muwa-1.3b
|
theSLWayne
| null |
[
"text-generation-inference",
"en",
"dataset:databricks/databricks-dolly-15k",
"arxiv:2106.09685",
"license:cc-by-nc-4.0",
"region:us"
] | 2023-05-04T11:45:14Z |
2023-05-11T14:23:58+00:00
| 0 | 0 |
---
datasets:
- databricks/databricks-dolly-15k
language:
- en
license: cc-by-nc-4.0
tags:
- text-generation-inference
---
# Muwa-OPT - A budget-friendly OPT-based LLM
[Muwa Repository on GitHub](https://github.com/theSLWayne/Muwa-OPT/)
Muwa is a fine-tuned LoRA model based on Facebook's OPT model architecture. Muwa was fine-tuned using the [databricks-dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k), which is a dataset of instruction-following records that belong to multiple categories like brainstorming, classification, closed QA, generation, information extraction, open QA, and summarization. **The specialty of Muwa is that only free resources have been used to fine-tune the model**, no fancy arrays of GPUs or paid GPU processors were not used for fine-tuning the model; only the free-tier of Google Colaboratory.
Muwa is currently trained using the [OPT 1.3b model](https://huggingface.co/facebook/opt-1.3b), which is available in HuggingFace.
This work is heavily inspired from [Yudhanjaya's Eluwa model](https://github.com/yudhanjaya/Eluwa). Most of the model fine-tuning and benchmarking code is taken from their repository and I made some adjustments to the code and changed some parameters to make sure that the fine-tuning process can be done on free resources that were available to me at the time.
## Inference
Make sure you install the following Python packages in the environment where the model is intended to be run.
```shell
pip install torch peft datasets evaluate transformers accelerate bitsandbytes
```
First, OPT 1.3b model should be loaded and then Muwa should be loaded from their respective HuggingFace repositories. After the models are loaded, they can be used for inference.
```python
import torch
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
# Define model names to be loaded
peft_model_id = 'theSLWayne/Muwa-1.3b'
base_model = 'facebook/opt-1.3b'
# Load base model
model = AutoModelForCausalLM.from_pretrained(
base_model,
device_map='auto',
torch_dtype=torch.float16,
)
# Load Muwa
model = PeftModel.from_pretrained(
model,
peft_model_id,
device_map='auto',
torch_dtype=torch.float16,
)
# Initiate tokenizer of the base model
tokenizer = AutoTokenizer.from_pretrained(base_model)
# Create batches of inputs
batch = tokenizer("What is a deep learning model?", return_tensors='pt')
# Take predictions
with torch.cuda.amp.autocast():
output_tokens = model.generate(**batch, max_new_tokens=50)
print(tokenizer.decode(output_tokens[0], skip_special_tokens=True))
```
If you intend to use CPU (which is not recommended), you can load the models as follows:
```python
model = AutoModelForCausalLM.from_pretrained(
base_model, device_map='auto', low_cpu_mem_usage=True
)
model = PeftModel.from_pretrained(
model,
peft_model_id,
device_map='auto',
)
```
## Training Muwa
This model was fine-tuned for 2 Epochs using the aforementioned Databricks Dolly 15K dataset. This model and its base model (OPT 1.3b) can be loaded in 8-bit. The notebook that was used for training this model can be found on the [GitHub repo](https://github.com/theSLWayne/Muwa-OPT/), including my notes on each code block.
The model was trained only using T4 GPU provided by Google Colab. **In order to fit the whole model and the dataset into it, the dataset had an input limit of 1024 tokens per each query**. **This was done because with the default value, the GPU RAM was not enough to fine-tune the model**.
With the limit in input tokens, the model training took ~12 GB of GPU RAM.
### PEFT and LoRA
PEFT(Parameter-Efficient Fine-tuning) is a set of approaches that are meant to reduce the cost of fine-tuning, storing, and deploying large models. According to [this HuggingFace article on PEFT](https://huggingface.co/blog/peft),
*`PEFT approaches only fine-tune a small number of (extra) model parameters while freezing most parameters of the pretrained LLMs, thereby greatly decreasing the computational and storage costs. This also overcomes the issues of catastrophic forgetting, a behaviour observed during the full finetuning of LLMs. PEFT approaches have also shown to be better than fine-tuning in the low-data regimes and generalize better to out-of-domain scenarios. It can be applied to various modalities, e.g., image classification and stable diffusion dreambooth.`*
HuggingFace has launched a Python package with the same name and according to the documentation it implements a number of PEFT methods:
1. LoRA
2. Prefix Tuning
3. P-Tuning
4. Prompt Tuning
5. AdaLoRA
This package is used in fine-tuning and in the inference of Muwa. More details about this package can be discovered [here](https://github.com/huggingface/peft).
LoRA (Low-Rank Adaptation) is a method proposed for adapting large pre-trained language models to specific tasks or domains. It involves freezing the pre-trained model weights and adding trainable rank decomposition matrices to each layer of the Transformer architecture, which significantly reduces the number of trainable parameters for downstream tasks. This approach allows for efficient adaptation of language models with fewer trainable parameters and reduced GPU memory requirements. More information on LoRA can be found on the paper that introduced the method, which can be accessed [here](https://arxiv.org/abs/2106.09685). Also, I found [this video](https://www.youtube.com/watch?v=_K3HgjnRHCY&lc=Ugyqpr8yVUW2DHlvsoZ4AaABAg) that explains the paper in simple terms, which I found to be very useful.
## Testing and Evaluating
Muwa was tested and evaluated using SQuAD mini, wikitext, and piqa datasets. Both Muwa and its base model, OPT 1.3b were evaluated seperately using all mentioned datasets and the results can be summarized as follows:
| Dataset | OPT 1.3b | Muwa |
|---------|----------|------|
| SQuAD Mini (*avg. f1 score*) | 24.587 | **26.234** |
| wikitext (*perplexity*) | 13.91406 | **13.96875** |
| piqa (*accuracy*) | 0.495 | **0.532** |
As shown, Muwa has been able to outperform its base model by fine tuning using a rather smaller dataset (compared to others like [Alpaca](https://huggingface.co/datasets/tatsu-lab/alpaca) available for these tasks) for all the evaluation datasets.
This shows that LLMs that have Billions of parameters can be fine-tuned using resources which are available for free and you can actually improve the model's performance by doing so.
Code used for evaluating Muwa can be found in the notebook which is included in the [GitHub repo](https://github.com/theSLWayne/Muwa-OPT/).
## The Story Behind Muwa
As mentioned above, Muwa was heavily inspired by Eluwa model developed by Yudhanjaya et al. "Eluwa" means goat in Sinhalese. Continuing the trend of naming LLMs after even-toed ungulates, this model is named "Muwa".
Deers aren't as fearsome as Goats, or even Llamas and alpacas but they are still an impressive species. They are graceful, agile, and known for their antlers, which they shed and regrow every year. In some cultures, deers are considered a symbol of gentleness and kindness. All the more reasons to name this model after them.
About the graphic located at the beginning of this document, that is the work of someone(me) with zero knowledge and experience in design, and it shows. The initial image was taken from [freepngimg.com](https://www.freepngimg.com/png/22758-deer-head-free-download) and is protected under [Creative Commons (CC BY-NC 4.0)](https://creativecommons.org/licenses/by-nc/4.0/) license. Then that image was colorized using [Colorizer Models HuggingFace space](https://huggingface.co/spaces/trysem/Colorizer_Models). Then the text was added after loading the colorized image into [Canva](canva.com), which provided the final output.
## License
The base model used for this work, Facebook's OPT has its own license, which can be found [here](https://github.com/facebookresearch/metaseq/blob/main/projects/OPT/MODEL_LICENSE.md).
Databricks Dolly 15k model is protected under [CC BY-SA 3.0](https://creativecommons.org/licenses/by-sa/3.0/), allowing it to be modified, redistributed, and used for any purpose, even commercially.
Although the dataset is allowed to be modified and redistributed, the licensing of OPT does not allow to use it for any commercial or any other non-research related cases, therefore making Muwa restricted to be used only for research, under CC BY NC 4.0.
| null |
Non_BioNLP
|
# Muwa-OPT - A budget-friendly OPT-based LLM
[Muwa Repository on GitHub](https://github.com/theSLWayne/Muwa-OPT/)
Muwa is a fine-tuned LoRA model based on Facebook's OPT model architecture. Muwa was fine-tuned using the [databricks-dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k), which is a dataset of instruction-following records that belong to multiple categories like brainstorming, classification, closed QA, generation, information extraction, open QA, and summarization. **The specialty of Muwa is that only free resources have been used to fine-tune the model**, no fancy arrays of GPUs or paid GPU processors were not used for fine-tuning the model; only the free-tier of Google Colaboratory.
Muwa is currently trained using the [OPT 1.3b model](https://huggingface.co/facebook/opt-1.3b), which is available in HuggingFace.
This work is heavily inspired from [Yudhanjaya's Eluwa model](https://github.com/yudhanjaya/Eluwa). Most of the model fine-tuning and benchmarking code is taken from their repository and I made some adjustments to the code and changed some parameters to make sure that the fine-tuning process can be done on free resources that were available to me at the time.
## Inference
Make sure you install the following Python packages in the environment where the model is intended to be run.
```shell
pip install torch peft datasets evaluate transformers accelerate bitsandbytes
```
First, OPT 1.3b model should be loaded and then Muwa should be loaded from their respective HuggingFace repositories. After the models are loaded, they can be used for inference.
```python
import torch
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
# Define model names to be loaded
peft_model_id = 'theSLWayne/Muwa-1.3b'
base_model = 'facebook/opt-1.3b'
# Load base model
model = AutoModelForCausalLM.from_pretrained(
base_model,
device_map='auto',
torch_dtype=torch.float16,
)
# Load Muwa
model = PeftModel.from_pretrained(
model,
peft_model_id,
device_map='auto',
torch_dtype=torch.float16,
)
# Initiate tokenizer of the base model
tokenizer = AutoTokenizer.from_pretrained(base_model)
# Create batches of inputs
batch = tokenizer("What is a deep learning model?", return_tensors='pt')
# Take predictions
with torch.cuda.amp.autocast():
output_tokens = model.generate(**batch, max_new_tokens=50)
print(tokenizer.decode(output_tokens[0], skip_special_tokens=True))
```
If you intend to use CPU (which is not recommended), you can load the models as follows:
```python
model = AutoModelForCausalLM.from_pretrained(
base_model, device_map='auto', low_cpu_mem_usage=True
)
model = PeftModel.from_pretrained(
model,
peft_model_id,
device_map='auto',
)
```
## Training Muwa
This model was fine-tuned for 2 Epochs using the aforementioned Databricks Dolly 15K dataset. This model and its base model (OPT 1.3b) can be loaded in 8-bit. The notebook that was used for training this model can be found on the [GitHub repo](https://github.com/theSLWayne/Muwa-OPT/), including my notes on each code block.
The model was trained only using T4 GPU provided by Google Colab. **In order to fit the whole model and the dataset into it, the dataset had an input limit of 1024 tokens per each query**. **This was done because with the default value, the GPU RAM was not enough to fine-tune the model**.
With the limit in input tokens, the model training took ~12 GB of GPU RAM.
### PEFT and LoRA
PEFT(Parameter-Efficient Fine-tuning) is a set of approaches that are meant to reduce the cost of fine-tuning, storing, and deploying large models. According to [this HuggingFace article on PEFT](https://huggingface.co/blog/peft),
*`PEFT approaches only fine-tune a small number of (extra) model parameters while freezing most parameters of the pretrained LLMs, thereby greatly decreasing the computational and storage costs. This also overcomes the issues of catastrophic forgetting, a behaviour observed during the full finetuning of LLMs. PEFT approaches have also shown to be better than fine-tuning in the low-data regimes and generalize better to out-of-domain scenarios. It can be applied to various modalities, e.g., image classification and stable diffusion dreambooth.`*
HuggingFace has launched a Python package with the same name and according to the documentation it implements a number of PEFT methods:
1. LoRA
2. Prefix Tuning
3. P-Tuning
4. Prompt Tuning
5. AdaLoRA
This package is used in fine-tuning and in the inference of Muwa. More details about this package can be discovered [here](https://github.com/huggingface/peft).
LoRA (Low-Rank Adaptation) is a method proposed for adapting large pre-trained language models to specific tasks or domains. It involves freezing the pre-trained model weights and adding trainable rank decomposition matrices to each layer of the Transformer architecture, which significantly reduces the number of trainable parameters for downstream tasks. This approach allows for efficient adaptation of language models with fewer trainable parameters and reduced GPU memory requirements. More information on LoRA can be found on the paper that introduced the method, which can be accessed [here](https://arxiv.org/abs/2106.09685). Also, I found [this video](https://www.youtube.com/watch?v=_K3HgjnRHCY&lc=Ugyqpr8yVUW2DHlvsoZ4AaABAg) that explains the paper in simple terms, which I found to be very useful.
## Testing and Evaluating
Muwa was tested and evaluated using SQuAD mini, wikitext, and piqa datasets. Both Muwa and its base model, OPT 1.3b were evaluated seperately using all mentioned datasets and the results can be summarized as follows:
| Dataset | OPT 1.3b | Muwa |
|---------|----------|------|
| SQuAD Mini (*avg. f1 score*) | 24.587 | **26.234** |
| wikitext (*perplexity*) | 13.91406 | **13.96875** |
| piqa (*accuracy*) | 0.495 | **0.532** |
As shown, Muwa has been able to outperform its base model by fine tuning using a rather smaller dataset (compared to others like [Alpaca](https://huggingface.co/datasets/tatsu-lab/alpaca) available for these tasks) for all the evaluation datasets.
This shows that LLMs that have Billions of parameters can be fine-tuned using resources which are available for free and you can actually improve the model's performance by doing so.
Code used for evaluating Muwa can be found in the notebook which is included in the [GitHub repo](https://github.com/theSLWayne/Muwa-OPT/).
## The Story Behind Muwa
As mentioned above, Muwa was heavily inspired by Eluwa model developed by Yudhanjaya et al. "Eluwa" means goat in Sinhalese. Continuing the trend of naming LLMs after even-toed ungulates, this model is named "Muwa".
Deers aren't as fearsome as Goats, or even Llamas and alpacas but they are still an impressive species. They are graceful, agile, and known for their antlers, which they shed and regrow every year. In some cultures, deers are considered a symbol of gentleness and kindness. All the more reasons to name this model after them.
About the graphic located at the beginning of this document, that is the work of someone(me) with zero knowledge and experience in design, and it shows. The initial image was taken from [freepngimg.com](https://www.freepngimg.com/png/22758-deer-head-free-download) and is protected under [Creative Commons (CC BY-NC 4.0)](https://creativecommons.org/licenses/by-nc/4.0/) license. Then that image was colorized using [Colorizer Models HuggingFace space](https://huggingface.co/spaces/trysem/Colorizer_Models). Then the text was added after loading the colorized image into [Canva](canva.com), which provided the final output.
## License
The base model used for this work, Facebook's OPT has its own license, which can be found [here](https://github.com/facebookresearch/metaseq/blob/main/projects/OPT/MODEL_LICENSE.md).
Databricks Dolly 15k model is protected under [CC BY-SA 3.0](https://creativecommons.org/licenses/by-sa/3.0/), allowing it to be modified, redistributed, and used for any purpose, even commercially.
Although the dataset is allowed to be modified and redistributed, the licensing of OPT does not allow to use it for any commercial or any other non-research related cases, therefore making Muwa restricted to be used only for research, under CC BY NC 4.0.
|
{"datasets": ["databricks/databricks-dolly-15k"], "language": ["en"], "license": "cc-by-nc-4.0", "tags": ["text-generation-inference"]}
|
task
|
[
"SUMMARIZATION"
] | 39,851 |
Marco127/Argu_T1
|
Marco127
|
sentence-similarity
|
[
"sentence-transformers",
"safetensors",
"mpnet",
"sentence-similarity",
"feature-extraction",
"generated_from_trainer",
"dataset_size:672",
"loss:MultipleNegativesRankingLoss",
"arxiv:1908.10084",
"arxiv:1705.00652",
"base_model:sentence-transformers/multi-qa-mpnet-base-dot-v1",
"base_model:finetune:sentence-transformers/multi-qa-mpnet-base-dot-v1",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2025-02-08T01:02:37Z |
2025-02-08T01:02:56+00:00
| 4 | 0 |
---
base_model: sentence-transformers/multi-qa-mpnet-base-dot-v1
library_name: sentence-transformers
metrics:
- dot_accuracy
- dot_accuracy_threshold
- dot_f1
- dot_f1_threshold
- dot_precision
- dot_recall
- dot_ap
- dot_mcc
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:672
- loss:MultipleNegativesRankingLoss
widget:
- source_sentence: '
Animals may not be allowed onto beds or other furniture, which serves for
guests. It is not permitted to use baths, showers or washbasins for bathing or
washing animals.'
sentences:
- '
Please advise of any special needs such as high-chairs and sleeping cots.'
- '
Animals may not be allowed onto beds or other furniture, which serves for
guests. It is not permitted to use baths, showers or washbasins for bathing or
washing animals.'
- '
It is strongly advised that you arrange adequate insurance cover such as cancellation
due to illness,
accident or injury, personal accident and personal liability, loss of or damage
to baggage and sport
equipment (Note that is not an exhaustive list). We will not be responsible or
liable if you fail to take
adequate insurance cover or none at all.'
- source_sentence: 'Owners are responsible for ensuring that animals are kept quiet
between the
hours of 10:00 pm and 06:00 am. In the case of failure to abide by this
regulation the guest may be asked to leave the hotel without a refund of the
price of the night''s accommodation.'
sentences:
- '
Visitors are not allowed in the rooms and must be entertained in the lounges and/or
other public areas
provided.'
- 'To ensure the safety and comfort of everyone in the hotel, the Management
reserves the right to terminate the accommodation of guests who fail to comply
with the following rules and regulations.'
- 'Owners are responsible for ensuring that animals are kept quiet between the
hours of 10:00 pm and 06:00 am. In the case of failure to abide by this
regulation the guest may be asked to leave the hotel without a refund of the
price of the night''s accommodation.'
- source_sentence: '
We ask all guests to behave in such a way that they do not disturb other guests
and the neighborhood.
The hotel staff is authorized to refuse services to a person who violates this
rule.'
sentences:
- '
Please take note of the limitation specified for the room you have booked.
If such number is exceeded, whether temporarily or over-night, we reserve the
right to do one or more of
the following: cancel your booking; retain all the monies you''ve paid; request
you to vacate your room(s)
forthwith, charge a higher rate for the room or recover all monies due.'
- '
We ask all guests to behave in such a way that they do not disturb other guests
and the neighborhood.
The hotel staff is authorized to refuse services to a person who violates this
rule.'
- 'We will only deal with your information as indicated in the booking/reservation
and we will only process your
personal information (both terms as defined in the Protection of Personal Information
Act, act 4 of 2013 [''the
POPIA''] and the European Union General Data Protection Regulation – (''GDPR'')
and any Special Personal
Information (as defined in the GDPR & POPIA), which processing includes amongst
others the ''collecting,
storing and dissemination'' of your personal information (as defined in GDPR &
POPIA).'
- source_sentence: '
All articles stored in the luggage storage room are received at the owner’s own
risk.'
sentences:
- "\n Unregistered visitors are not permitted to enter guest rooms or other areas\
\ of\nthe hotel. An additional fee for unregistered guests will be charged to\
\ the\naccount of the guest(s) registered to the room."
- 'Please advise us if you anticipate arriving late as bookings will be cancelled
by 17:00 on the day of arrival,
unless we have been so notified.'
- '
All articles stored in the luggage storage room are received at the owner’s own
risk.'
- source_sentence: ' In the event of a disturbance, one polite request (warning) will
be given to reduce the noise. If our request is not followed, the guest will be
asked to leave
the hotel without refund and may be charged Guest Compensation Disturbance Fee.'
sentences:
- '
Without limiting the generality of the aforementioned, it applies to pay-to-view
TV programmes or videos, as
well as telephone calls or any other expenses of a similar nature that is made
from your room, you will be
deemed to be the contracting party.'
- 'Pets are not allowed in the restaurant during breakfast time
(7:00 – 10:30) for hygienic reasons due to the breakfast’s buffet style. An
exception is the case when the hotel terrace is open, as pets can be taken to
the terrace through the hotel''s main entrance and they can stay there during
breakfast.'
- ' In the event of a disturbance, one polite request (warning) will
be given to reduce the noise. If our request is not followed, the guest will be
asked to leave
the hotel without refund and may be charged Guest Compensation Disturbance Fee.'
model-index:
- name: SentenceTransformer based on sentence-transformers/multi-qa-mpnet-base-dot-v1
results:
- task:
type: binary-classification
name: Binary Classification
dataset:
name: Unknown
type: unknown
metrics:
- type: dot_accuracy
value: 0.6745562130177515
name: Dot Accuracy
- type: dot_accuracy_threshold
value: 49.0201301574707
name: Dot Accuracy Threshold
- type: dot_f1
value: 0.4932735426008969
name: Dot F1
- type: dot_f1_threshold
value: 35.02415466308594
name: Dot F1 Threshold
- type: dot_precision
value: 0.32934131736526945
name: Dot Precision
- type: dot_recall
value: 0.9821428571428571
name: Dot Recall
- type: dot_ap
value: 0.3294144882113245
name: Dot Ap
- type: dot_mcc
value: -0.03920743101752848
name: Dot Mcc
---
# SentenceTransformer based on sentence-transformers/multi-qa-mpnet-base-dot-v1
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/multi-qa-mpnet-base-dot-v1](https://huggingface.co/sentence-transformers/multi-qa-mpnet-base-dot-v1). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [sentence-transformers/multi-qa-mpnet-base-dot-v1](https://huggingface.co/sentence-transformers/multi-qa-mpnet-base-dot-v1) <!-- at revision 4633e80e17ea975bc090c97b049da26062b054d3 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Dot Product
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: MPNetModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("Marco127/Argu_T1")
# Run inference
sentences = [
' In the event of a disturbance, one polite request (warning) will\nbe given to reduce the noise. If our request is not followed, the guest will be asked to leave\nthe hotel without refund and may be charged Guest Compensation Disturbance Fee.',
' In the event of a disturbance, one polite request (warning) will\nbe given to reduce the noise. If our request is not followed, the guest will be asked to leave\nthe hotel without refund and may be charged Guest Compensation Disturbance Fee.',
'\nWithout limiting the generality of the aforementioned, it applies to pay-to-view TV programmes or videos, as\nwell as telephone calls or any other expenses of a similar nature that is made from your room, you will be\ndeemed to be the contracting party.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Binary Classification
* Evaluated with [<code>BinaryClassificationEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.BinaryClassificationEvaluator)
| Metric | Value |
|:-----------------------|:-----------|
| dot_accuracy | 0.6746 |
| dot_accuracy_threshold | 49.0201 |
| dot_f1 | 0.4933 |
| dot_f1_threshold | 35.0242 |
| dot_precision | 0.3293 |
| dot_recall | 0.9821 |
| **dot_ap** | **0.3294** |
| dot_mcc | -0.0392 |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 672 training samples
* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
* Approximate statistics based on the first 672 samples:
| | sentence1 | sentence2 | label |
|:--------|:------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|:------------------------------------------------|
| type | string | string | int |
| details | <ul><li>min: 11 tokens</li><li>mean: 48.63 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>min: 11 tokens</li><li>mean: 48.63 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>0: ~66.67%</li><li>1: ~33.33%</li></ul> |
* Samples:
| sentence1 | sentence2 | label |
|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
| <code><br>The pets can not be left without supervision if there is a risk of causing any<br>damage or might disturb other guests.</code> | <code><br>The pets can not be left without supervision if there is a risk of causing any<br>damage or might disturb other guests.</code> | <code>0</code> |
| <code><br>Any guest in violation of these rules may be asked to leave the hotel with no refund. Extra copies of these<br>rules are available at the Front Desk upon request.</code> | <code><br>Any guest in violation of these rules may be asked to leave the hotel with no refund. Extra copies of these<br>rules are available at the Front Desk upon request.</code> | <code>0</code> |
| <code><br>Consuming the products from the minibar involves additional costs. You can find the<br>prices in the kitchen area.</code> | <code><br>Consuming the products from the minibar involves additional costs. You can find the<br>prices in the kitchen area.</code> | <code>0</code> |
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
```json
{
"scale": 20.0,
"similarity_fct": "cos_sim"
}
```
### Evaluation Dataset
#### Unnamed Dataset
* Size: 169 evaluation samples
* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
* Approximate statistics based on the first 169 samples:
| | sentence1 | sentence2 | label |
|:--------|:------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|:------------------------------------------------|
| type | string | string | int |
| details | <ul><li>min: 13 tokens</li><li>mean: 46.01 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>min: 13 tokens</li><li>mean: 46.01 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>0: ~66.86%</li><li>1: ~33.14%</li></ul> |
* Samples:
| sentence1 | sentence2 | label |
|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
| <code><br>I understand and accept that the BON Hotels Group collects the personal information ("personal<br>information") of all persons in my party for purposes of loyalty programmes and special offers. I, on behalf of<br>all in my party, expressly consent and grant permission to the BON Hotels Group to: -<br>collect, collate, process, study and use the personal information; and<br>communicate directly with me/us from time to time, unless I have stated to the contrary below.</code> | <code><br>I understand and accept that the BON Hotels Group collects the personal information ("personal<br>information") of all persons in my party for purposes of loyalty programmes and special offers. I, on behalf of<br>all in my party, expressly consent and grant permission to the BON Hotels Group to: -<br>collect, collate, process, study and use the personal information; and<br>communicate directly with me/us from time to time, unless I have stated to the contrary below.</code> | <code>0</code> |
| <code>However, in lieu of the above, any such goods will only be kept by us for 6 (six) months. At the end of which<br>period, we reserve the right in our sole discretion to dispose thereof and you will have no right of recourse<br>against us.</code> | <code>However, in lieu of the above, any such goods will only be kept by us for 6 (six) months. At the end of which<br>period, we reserve the right in our sole discretion to dispose thereof and you will have no right of recourse<br>against us.</code> | <code>0</code> |
| <code> In cases where the hotel<br>suffers damage (either physical, or moral) due to the guests’ violation of the above rules, it<br>may charge a compensation fee in proportion to the damage. Moral damage may be for<br>example disturbing other guests, thus ruining the reputation of the hotel.</code> | <code> In cases where the hotel<br>suffers damage (either physical, or moral) due to the guests’ violation of the above rules, it<br>may charge a compensation fee in proportion to the damage. Moral damage may be for<br>example disturbing other guests, thus ruining the reputation of the hotel.</code> | <code>0</code> |
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
```json
{
"scale": 20.0,
"similarity_fct": "cos_sim"
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `num_train_epochs`: 5
- `warmup_ratio`: 0.1
- `fp16`: True
- `batch_sampler`: no_duplicates
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 5e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 5
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: True
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: None
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | Training Loss | Validation Loss | dot_ap |
|:------:|:----:|:-------------:|:---------------:|:------:|
| -1 | -1 | - | - | 0.3294 |
| 2.3333 | 100 | 0.0004 | 0.0000 | - |
| 4.6905 | 200 | 0.0003 | 0.0000 | - |
### Framework Versions
- Python: 3.11.11
- Sentence Transformers: 3.4.1
- Transformers: 4.48.3
- PyTorch: 2.5.1+cu124
- Accelerate: 1.3.0
- Datasets: 3.2.0
- Tokenizers: 0.21.0
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
title={Efficient Natural Language Response Suggestion for Smart Reply},
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},
year={2017},
eprint={1705.00652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SentenceTransformer based on sentence-transformers/multi-qa-mpnet-base-dot-v1
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/multi-qa-mpnet-base-dot-v1](https://huggingface.co/sentence-transformers/multi-qa-mpnet-base-dot-v1). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [sentence-transformers/multi-qa-mpnet-base-dot-v1](https://huggingface.co/sentence-transformers/multi-qa-mpnet-base-dot-v1) <!-- at revision 4633e80e17ea975bc090c97b049da26062b054d3 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Dot Product
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: MPNetModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("Marco127/Argu_T1")
# Run inference
sentences = [
' In the event of a disturbance, one polite request (warning) will\nbe given to reduce the noise. If our request is not followed, the guest will be asked to leave\nthe hotel without refund and may be charged Guest Compensation Disturbance Fee.',
' In the event of a disturbance, one polite request (warning) will\nbe given to reduce the noise. If our request is not followed, the guest will be asked to leave\nthe hotel without refund and may be charged Guest Compensation Disturbance Fee.',
'\nWithout limiting the generality of the aforementioned, it applies to pay-to-view TV programmes or videos, as\nwell as telephone calls or any other expenses of a similar nature that is made from your room, you will be\ndeemed to be the contracting party.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Binary Classification
* Evaluated with [<code>BinaryClassificationEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.BinaryClassificationEvaluator)
| Metric | Value |
|:-----------------------|:-----------|
| dot_accuracy | 0.6746 |
| dot_accuracy_threshold | 49.0201 |
| dot_f1 | 0.4933 |
| dot_f1_threshold | 35.0242 |
| dot_precision | 0.3293 |
| dot_recall | 0.9821 |
| **dot_ap** | **0.3294** |
| dot_mcc | -0.0392 |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 672 training samples
* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
* Approximate statistics based on the first 672 samples:
| | sentence1 | sentence2 | label |
|:--------|:------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|:------------------------------------------------|
| type | string | string | int |
| details | <ul><li>min: 11 tokens</li><li>mean: 48.63 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>min: 11 tokens</li><li>mean: 48.63 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>0: ~66.67%</li><li>1: ~33.33%</li></ul> |
* Samples:
| sentence1 | sentence2 | label |
|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
| <code><br>The pets can not be left without supervision if there is a risk of causing any<br>damage or might disturb other guests.</code> | <code><br>The pets can not be left without supervision if there is a risk of causing any<br>damage or might disturb other guests.</code> | <code>0</code> |
| <code><br>Any guest in violation of these rules may be asked to leave the hotel with no refund. Extra copies of these<br>rules are available at the Front Desk upon request.</code> | <code><br>Any guest in violation of these rules may be asked to leave the hotel with no refund. Extra copies of these<br>rules are available at the Front Desk upon request.</code> | <code>0</code> |
| <code><br>Consuming the products from the minibar involves additional costs. You can find the<br>prices in the kitchen area.</code> | <code><br>Consuming the products from the minibar involves additional costs. You can find the<br>prices in the kitchen area.</code> | <code>0</code> |
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
```json
{
"scale": 20.0,
"similarity_fct": "cos_sim"
}
```
### Evaluation Dataset
#### Unnamed Dataset
* Size: 169 evaluation samples
* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
* Approximate statistics based on the first 169 samples:
| | sentence1 | sentence2 | label |
|:--------|:------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|:------------------------------------------------|
| type | string | string | int |
| details | <ul><li>min: 13 tokens</li><li>mean: 46.01 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>min: 13 tokens</li><li>mean: 46.01 tokens</li><li>max: 156 tokens</li></ul> | <ul><li>0: ~66.86%</li><li>1: ~33.14%</li></ul> |
* Samples:
| sentence1 | sentence2 | label |
|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
| <code><br>I understand and accept that the BON Hotels Group collects the personal information ("personal<br>information") of all persons in my party for purposes of loyalty programmes and special offers. I, on behalf of<br>all in my party, expressly consent and grant permission to the BON Hotels Group to: -<br>collect, collate, process, study and use the personal information; and<br>communicate directly with me/us from time to time, unless I have stated to the contrary below.</code> | <code><br>I understand and accept that the BON Hotels Group collects the personal information ("personal<br>information") of all persons in my party for purposes of loyalty programmes and special offers. I, on behalf of<br>all in my party, expressly consent and grant permission to the BON Hotels Group to: -<br>collect, collate, process, study and use the personal information; and<br>communicate directly with me/us from time to time, unless I have stated to the contrary below.</code> | <code>0</code> |
| <code>However, in lieu of the above, any such goods will only be kept by us for 6 (six) months. At the end of which<br>period, we reserve the right in our sole discretion to dispose thereof and you will have no right of recourse<br>against us.</code> | <code>However, in lieu of the above, any such goods will only be kept by us for 6 (six) months. At the end of which<br>period, we reserve the right in our sole discretion to dispose thereof and you will have no right of recourse<br>against us.</code> | <code>0</code> |
| <code> In cases where the hotel<br>suffers damage (either physical, or moral) due to the guests’ violation of the above rules, it<br>may charge a compensation fee in proportion to the damage. Moral damage may be for<br>example disturbing other guests, thus ruining the reputation of the hotel.</code> | <code> In cases where the hotel<br>suffers damage (either physical, or moral) due to the guests’ violation of the above rules, it<br>may charge a compensation fee in proportion to the damage. Moral damage may be for<br>example disturbing other guests, thus ruining the reputation of the hotel.</code> | <code>0</code> |
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
```json
{
"scale": 20.0,
"similarity_fct": "cos_sim"
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `num_train_epochs`: 5
- `warmup_ratio`: 0.1
- `fp16`: True
- `batch_sampler`: no_duplicates
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 16
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 5e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 5
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: True
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: None
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | Training Loss | Validation Loss | dot_ap |
|:------:|:----:|:-------------:|:---------------:|:------:|
| -1 | -1 | - | - | 0.3294 |
| 2.3333 | 100 | 0.0004 | 0.0000 | - |
| 4.6905 | 200 | 0.0003 | 0.0000 | - |
### Framework Versions
- Python: 3.11.11
- Sentence Transformers: 3.4.1
- Transformers: 4.48.3
- PyTorch: 2.5.1+cu124
- Accelerate: 1.3.0
- Datasets: 3.2.0
- Tokenizers: 0.21.0
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
title={Efficient Natural Language Response Suggestion for Smart Reply},
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},
year={2017},
eprint={1705.00652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "sentence-transformers/multi-qa-mpnet-base-dot-v1", "library_name": "sentence-transformers", "metrics": ["dot_accuracy", "dot_accuracy_threshold", "dot_f1", "dot_f1_threshold", "dot_precision", "dot_recall", "dot_ap", "dot_mcc"], "pipeline_tag": "sentence-similarity", "tags": ["sentence-transformers", "sentence-similarity", "feature-extraction", "generated_from_trainer", "dataset_size:672", "loss:MultipleNegativesRankingLoss"], "widget": [{"source_sentence": "\nAnimals may not be allowed onto beds or other furniture, which serves for\nguests. It is not permitted to use baths, showers or washbasins for bathing or\nwashing animals.", "sentences": ["\nPlease advise of any special needs such as high-chairs and sleeping cots.", "\nAnimals may not be allowed onto beds or other furniture, which serves for\nguests. It is not permitted to use baths, showers or washbasins for bathing or\nwashing animals.", "\nIt is strongly advised that you arrange adequate insurance cover such as cancellation due to illness,\naccident or injury, personal accident and personal liability, loss of or damage to baggage and sport\nequipment (Note that is not an exhaustive list). We will not be responsible or liable if you fail to take\nadequate insurance cover or none at all."]}, {"source_sentence": "Owners are responsible for ensuring that animals are kept quiet between the\nhours of 10:00 pm and 06:00 am. In the case of failure to abide by this\nregulation the guest may be asked to leave the hotel without a refund of the\nprice of the night's accommodation.", "sentences": ["\nVisitors are not allowed in the rooms and must be entertained in the lounges and/or other public areas\nprovided.", "To ensure the safety and comfort of everyone in the hotel, the Management\nreserves the right to terminate the accommodation of guests who fail to comply\nwith the following rules and regulations.", "Owners are responsible for ensuring that animals are kept quiet between the\nhours of 10:00 pm and 06:00 am. In the case of failure to abide by this\nregulation the guest may be asked to leave the hotel without a refund of the\nprice of the night's accommodation."]}, {"source_sentence": "\nWe ask all guests to behave in such a way that they do not disturb other guests and the neighborhood.\nThe hotel staff is authorized to refuse services to a person who violates this rule.", "sentences": ["\nPlease take note of the limitation specified for the room you have booked.\nIf such number is exceeded, whether temporarily or over-night, we reserve the right to do one or more of\nthe following: cancel your booking; retain all the monies you've paid; request you to vacate your room(s)\nforthwith, charge a higher rate for the room or recover all monies due.", "\nWe ask all guests to behave in such a way that they do not disturb other guests and the neighborhood.\nThe hotel staff is authorized to refuse services to a person who violates this rule.", "We will only deal with your information as indicated in the booking/reservation and we will only process your\npersonal information (both terms as defined in the Protection of Personal Information Act, act 4 of 2013 ['the\nPOPIA'] and the European Union General Data Protection Regulation – ('GDPR') and any Special Personal\nInformation (as defined in the GDPR & POPIA), which processing includes amongst others the 'collecting,\nstoring and dissemination' of your personal information (as defined in GDPR & POPIA)."]}, {"source_sentence": "\nAll articles stored in the luggage storage room are received at the owner’s own risk.", "sentences": ["\n Unregistered visitors are not permitted to enter guest rooms or other areas of\nthe hotel. An additional fee for unregistered guests will be charged to the\naccount of the guest(s) registered to the room.", "Please advise us if you anticipate arriving late as bookings will be cancelled by 17:00 on the day of arrival,\nunless we have been so notified.", "\nAll articles stored in the luggage storage room are received at the owner’s own risk."]}, {"source_sentence": " In the event of a disturbance, one polite request (warning) will\nbe given to reduce the noise. If our request is not followed, the guest will be asked to leave\nthe hotel without refund and may be charged Guest Compensation Disturbance Fee.", "sentences": ["\nWithout limiting the generality of the aforementioned, it applies to pay-to-view TV programmes or videos, as\nwell as telephone calls or any other expenses of a similar nature that is made from your room, you will be\ndeemed to be the contracting party.", "Pets are not allowed in the restaurant during breakfast time\n(7:00 – 10:30) for hygienic reasons due to the breakfast’s buffet style. An\nexception is the case when the hotel terrace is open, as pets can be taken to\nthe terrace through the hotel's main entrance and they can stay there during\nbreakfast.", " In the event of a disturbance, one polite request (warning) will\nbe given to reduce the noise. If our request is not followed, the guest will be asked to leave\nthe hotel without refund and may be charged Guest Compensation Disturbance Fee."]}], "model-index": [{"name": "SentenceTransformer based on sentence-transformers/multi-qa-mpnet-base-dot-v1", "results": [{"task": {"type": "binary-classification", "name": "Binary Classification"}, "dataset": {"name": "Unknown", "type": "unknown"}, "metrics": [{"type": "dot_accuracy", "value": 0.6745562130177515, "name": "Dot Accuracy"}, {"type": "dot_accuracy_threshold", "value": 49.0201301574707, "name": "Dot Accuracy Threshold"}, {"type": "dot_f1", "value": 0.4932735426008969, "name": "Dot F1"}, {"type": "dot_f1_threshold", "value": 35.02415466308594, "name": "Dot F1 Threshold"}, {"type": "dot_precision", "value": 0.32934131736526945, "name": "Dot Precision"}, {"type": "dot_recall", "value": 0.9821428571428571, "name": "Dot Recall"}, {"type": "dot_ap", "value": 0.3294144882113245, "name": "Dot Ap"}, {"type": "dot_mcc", "value": -0.03920743101752848, "name": "Dot Mcc"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,852 |
Yuliya-HV/distilbert-base-uncased-finetuned-emotion-tweets
|
Yuliya-HV
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:emotion",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-05-30T18:23:11Z |
2022-05-30T18:39:27+00:00
| 12 | 0 |
---
datasets:
- emotion
license: apache-2.0
metrics:
- accuracy
- f1
tags:
- generated_from_trainer
model-index:
- name: distilbert-base-uncased-finetuned-emotion-tweets
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: emotion
type: emotion
args: default
metrics:
- type: accuracy
value: 0.9355
name: Accuracy
- type: f1
value: 0.9358599960917737
name: F1
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-emotion-tweets
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1572
- Accuracy: 0.9355
- F1: 0.9359
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| No log | 1.0 | 250 | 0.1672 | 0.932 | 0.9320 |
| No log | 2.0 | 500 | 0.1572 | 0.9355 | 0.9359 |
### Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0+cu113
- Datasets 2.2.2
- Tokenizers 0.12.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-emotion-tweets
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1572
- Accuracy: 0.9355
- F1: 0.9359
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| No log | 1.0 | 250 | 0.1672 | 0.932 | 0.9320 |
| No log | 2.0 | 500 | 0.1572 | 0.9355 | 0.9359 |
### Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0+cu113
- Datasets 2.2.2
- Tokenizers 0.12.1
|
{"datasets": ["emotion"], "license": "apache-2.0", "metrics": ["accuracy", "f1"], "tags": ["generated_from_trainer"], "model-index": [{"name": "distilbert-base-uncased-finetuned-emotion-tweets", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "emotion", "type": "emotion", "args": "default"}, "metrics": [{"type": "accuracy", "value": 0.9355, "name": "Accuracy"}, {"type": "f1", "value": 0.9358599960917737, "name": "F1"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,853 |
Paul-B98/codet5p_220m_py_sum
|
Paul-B98
|
text2text-generation
|
[
"transformers",
"pytorch",
"t5",
"text2text-generation",
"en",
"dataset:code_x_glue_ct_code_to_text",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-09-18T16:32:38Z |
2023-10-10T15:54:26+00:00
| 12 | 2 |
---
datasets:
- code_x_glue_ct_code_to_text
language:
- en
license: mit
metrics:
- bleu
- sacrebleu
---
# Codet5+ 220m Py Sum
This Model is based on the [CodeT5+ (220m)](https://huggingface.co/Salesforce/codet5p-220m) from salesforce and was finetuned for the code summarization task by using the [XCodeGlue](https://github.com/microsoft/CodeXGLUE) Dataset. The Code is accessible on [Github](https://github.com/Paul-B98/mdl-ii).
## Results
| Modell | BLEU |
| ------ | ---- |
| [CodeT5-base-sum-python](https://huggingface.co/Salesforce/codet5-base-codexglue-sum-python) | 23.564 |
| [CodeT5-base-multi-sum](https://huggingface.co/Salesforce/codet5-base-multi-sum) | 23.985 |
| [Code-Trans-S-ST](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python) | 5.495 |
| [Code-Trans-S-TF](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python_transfer_learning_finetune) | 21.093 |
| [Code-Trans-S-MT](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python_multitask) | 5.450 |
| [Code-Trans-S-MT-TF](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python_multitask_finetune) | 16.378 |
| [Code-Trans-B-ST](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python) | 4.638 |
| [Code-Trans-B-TF](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python_transfer_learning_finetune) | 21.671 |
| [Code-Trans-B-MT](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python_multitask) | 2.957 |
| [Code-Trans-B-MT-TF](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python_multitask_finetune) | 13.766 |
| [Code-Trans-L-TF](https://huggingface.co/SEBIS/code_trans_t5_large_code_documentation_generation_python_transfer_learning_finetune) | 23.306 |
| [Code-Trans-L-MT](https://huggingface.co/SEBIS/code_trans_t5_large_code_documentation_generation_python_multitask) | 13.487 |
| [Code-Trans-L-MT-TF](https://huggingface.co/SEBIS/code_trans_t5_large_code_documentation_generation_python_multitask_finetune) | 16.362 |
| **CodeT5+ 220m Py Sum***| 25.245 |
## Example on how to use
The model can be easily download from Huggingface and used in a summarization pipeline.
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("Paul-B98/codet5p_220m_py_sum"),
tokenizer=AutoTokenizer.from_pretrained("Salesforce/codet5p-220m"),
device=0
)
example_method = """
def greet(name):
print(f"Hello, {name}!")
"""
pipeline([example_method])[0]["summary_text"]
```
| null |
Non_BioNLP
|
# Codet5+ 220m Py Sum
This Model is based on the [CodeT5+ (220m)](https://huggingface.co/Salesforce/codet5p-220m) from salesforce and was finetuned for the code summarization task by using the [XCodeGlue](https://github.com/microsoft/CodeXGLUE) Dataset. The Code is accessible on [Github](https://github.com/Paul-B98/mdl-ii).
## Results
| Modell | BLEU |
| ------ | ---- |
| [CodeT5-base-sum-python](https://huggingface.co/Salesforce/codet5-base-codexglue-sum-python) | 23.564 |
| [CodeT5-base-multi-sum](https://huggingface.co/Salesforce/codet5-base-multi-sum) | 23.985 |
| [Code-Trans-S-ST](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python) | 5.495 |
| [Code-Trans-S-TF](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python_transfer_learning_finetune) | 21.093 |
| [Code-Trans-S-MT](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python_multitask) | 5.450 |
| [Code-Trans-S-MT-TF](https://huggingface.co/SEBIS/code_trans_t5_small_code_documentation_generation_python_multitask_finetune) | 16.378 |
| [Code-Trans-B-ST](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python) | 4.638 |
| [Code-Trans-B-TF](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python_transfer_learning_finetune) | 21.671 |
| [Code-Trans-B-MT](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python_multitask) | 2.957 |
| [Code-Trans-B-MT-TF](https://huggingface.co/SEBIS/code_trans_t5_base_code_documentation_generation_python_multitask_finetune) | 13.766 |
| [Code-Trans-L-TF](https://huggingface.co/SEBIS/code_trans_t5_large_code_documentation_generation_python_transfer_learning_finetune) | 23.306 |
| [Code-Trans-L-MT](https://huggingface.co/SEBIS/code_trans_t5_large_code_documentation_generation_python_multitask) | 13.487 |
| [Code-Trans-L-MT-TF](https://huggingface.co/SEBIS/code_trans_t5_large_code_documentation_generation_python_multitask_finetune) | 16.362 |
| **CodeT5+ 220m Py Sum***| 25.245 |
## Example on how to use
The model can be easily download from Huggingface and used in a summarization pipeline.
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("Paul-B98/codet5p_220m_py_sum"),
tokenizer=AutoTokenizer.from_pretrained("Salesforce/codet5p-220m"),
device=0
)
example_method = """
def greet(name):
print(f"Hello, {name}!")
"""
pipeline([example_method])[0]["summary_text"]
```
|
{"datasets": ["code_x_glue_ct_code_to_text"], "language": ["en"], "license": "mit", "metrics": ["bleu", "sacrebleu"]}
|
task
|
[
"SUMMARIZATION"
] | 39,854 |
Aakkash/t5-base-finetuned-amazon-en-es
|
Aakkash
|
summarization
|
[
"transformers",
"pytorch",
"t5",
"text2text-generation",
"summarization",
"generated_from_trainer",
"base_model:google-t5/t5-base",
"base_model:finetune:google-t5/t5-base",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-08-27T07:28:05Z |
2023-08-27T07:40:39+00:00
| 24 | 0 |
---
base_model: t5-base
license: apache-2.0
metrics:
- rouge
tags:
- summarization
- generated_from_trainer
model-index:
- name: t5-base-finetuned-amazon-en-es
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-base-finetuned-amazon-en-es
This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 5.0971
- Rouge1: 5.4907
- Rouge2: 0.0
- Rougel: 5.4907
- Rougelsum: 5.4907
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5.6e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
| 5.7476 | 1.0 | 2 | 5.4826 | 8.0818 | 0.0 | 8.0898 | 6.7159 |
| 5.6005 | 2.0 | 4 | 5.3150 | 5.4125 | 0.0 | 5.4125 | 5.4654 |
| 5.4222 | 3.0 | 6 | 5.2311 | 5.3768 | 0.0 | 5.3768 | 5.4298 |
| 4.9619 | 4.0 | 8 | 5.1824 | 6.6672 | 0.0 | 6.6672 | 6.6794 |
| 4.2747 | 5.0 | 10 | 5.1472 | 6.6672 | 0.0 | 6.6672 | 6.6794 |
| 4.2784 | 6.0 | 12 | 5.1219 | 5.4907 | 0.0 | 5.4907 | 5.4907 |
| 5.2337 | 7.0 | 14 | 5.1052 | 5.4907 | 0.0 | 5.4907 | 5.4907 |
| 4.7063 | 8.0 | 16 | 5.0971 | 5.4907 | 0.0 | 5.4907 | 5.4907 |
### Framework versions
- Transformers 4.32.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-base-finetuned-amazon-en-es
This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 5.0971
- Rouge1: 5.4907
- Rouge2: 0.0
- Rougel: 5.4907
- Rougelsum: 5.4907
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5.6e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|
| 5.7476 | 1.0 | 2 | 5.4826 | 8.0818 | 0.0 | 8.0898 | 6.7159 |
| 5.6005 | 2.0 | 4 | 5.3150 | 5.4125 | 0.0 | 5.4125 | 5.4654 |
| 5.4222 | 3.0 | 6 | 5.2311 | 5.3768 | 0.0 | 5.3768 | 5.4298 |
| 4.9619 | 4.0 | 8 | 5.1824 | 6.6672 | 0.0 | 6.6672 | 6.6794 |
| 4.2747 | 5.0 | 10 | 5.1472 | 6.6672 | 0.0 | 6.6672 | 6.6794 |
| 4.2784 | 6.0 | 12 | 5.1219 | 5.4907 | 0.0 | 5.4907 | 5.4907 |
| 5.2337 | 7.0 | 14 | 5.1052 | 5.4907 | 0.0 | 5.4907 | 5.4907 |
| 4.7063 | 8.0 | 16 | 5.0971 | 5.4907 | 0.0 | 5.4907 | 5.4907 |
### Framework versions
- Transformers 4.32.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
|
{"base_model": "t5-base", "license": "apache-2.0", "metrics": ["rouge"], "tags": ["summarization", "generated_from_trainer"], "model-index": [{"name": "t5-base-finetuned-amazon-en-es", "results": []}]}
|
task
|
[
"SUMMARIZATION"
] | 39,855 |
fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564
|
fine-tuned
| null |
[
"region:us"
] | 2024-04-30T13:39:37Z |
2024-04-30T13:39:37+00:00
| 0 | 0 |
---
{}
---
# fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564
## Model Description
fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564 is a fine-tuned version of jinaai/jina-embeddings-v2-small-en designed for a specific domain.
## Use Case
This model is designed to support various applications in natural language processing and understanding.
## Associated Dataset
This the dataset for this model can be found [**here**](https://huggingface.co/datasets/fine-tuned/fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564).
## How to Use
This model can be easily integrated into your NLP pipeline for tasks such as text classification, sentiment analysis, entity recognition, and more. Here's a simple example to get you started:
```python
from transformers import AutoModel, AutoTokenizer
llm_name = "fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564"
tokenizer = AutoTokenizer.from_pretrained(llm_name)
model = AutoModel.from_pretrained(llm_name)
tokens = tokenizer("Your text here", return_tensors="pt")
embedding = model(**tokens)
```
| null |
BioNLP
|
# fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564
## Model Description
fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564 is a fine-tuned version of jinaai/jina-embeddings-v2-small-en designed for a specific domain.
## Use Case
This model is designed to support various applications in natural language processing and understanding.
## Associated Dataset
This the dataset for this model can be found [**here**](https://huggingface.co/datasets/fine-tuned/fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564).
## How to Use
This model can be easily integrated into your NLP pipeline for tasks such as text classification, sentiment analysis, entity recognition, and more. Here's a simple example to get you started:
```python
from transformers import AutoModel, AutoTokenizer
llm_name = "fine-tuned/medical-10-10-16-jinaai_jina-embeddings-v2-small-en-100-gpt-3.5-turbo-_9062874564"
tokenizer = AutoTokenizer.from_pretrained(llm_name)
model = AutoModel.from_pretrained(llm_name)
tokens = tokenizer("Your text here", return_tensors="pt")
embedding = model(**tokens)
```
|
{}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,856 |
TransferGraph/connectivity_cola_6ep_ft-33-finetuned-lora-tweet_eval_emotion
|
TransferGraph
|
text-classification
|
[
"peft",
"safetensors",
"parquet",
"text-classification",
"dataset:tweet_eval",
"base_model:connectivity/cola_6ep_ft-33",
"base_model:adapter:connectivity/cola_6ep_ft-33",
"model-index",
"region:us"
] | 2024-02-29T12:47:55Z |
2024-02-29T12:47:56+00:00
| 0 | 0 |
---
base_model: connectivity/cola_6ep_ft-33
datasets:
- tweet_eval
library_name: peft
metrics:
- accuracy
tags:
- parquet
- text-classification
model-index:
- name: connectivity_cola_6ep_ft-33-finetuned-lora-tweet_eval_emotion
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: tweet_eval
type: tweet_eval
config: emotion
split: validation
args: emotion
metrics:
- type: accuracy
value: 0.6096256684491979
name: accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# connectivity_cola_6ep_ft-33-finetuned-lora-tweet_eval_emotion
This model is a fine-tuned version of [connectivity/cola_6ep_ft-33](https://huggingface.co/connectivity/cola_6ep_ft-33) on the tweet_eval dataset.
It achieves the following results on the evaluation set:
- accuracy: 0.6096
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0004
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
### Training results
| accuracy | train_loss | epoch |
|:--------:|:----------:|:-----:|
| 0.2727 | None | 0 |
| 0.4251 | 1.2732 | 0 |
| 0.4599 | 1.2369 | 1 |
| 0.5348 | 1.1518 | 2 |
| 0.6096 | 1.0195 | 3 |
### Framework versions
- PEFT 0.8.2
- Transformers 4.37.2
- Pytorch 2.2.0
- Datasets 2.16.1
- Tokenizers 0.15.2
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# connectivity_cola_6ep_ft-33-finetuned-lora-tweet_eval_emotion
This model is a fine-tuned version of [connectivity/cola_6ep_ft-33](https://huggingface.co/connectivity/cola_6ep_ft-33) on the tweet_eval dataset.
It achieves the following results on the evaluation set:
- accuracy: 0.6096
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0004
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
### Training results
| accuracy | train_loss | epoch |
|:--------:|:----------:|:-----:|
| 0.2727 | None | 0 |
| 0.4251 | 1.2732 | 0 |
| 0.4599 | 1.2369 | 1 |
| 0.5348 | 1.1518 | 2 |
| 0.6096 | 1.0195 | 3 |
### Framework versions
- PEFT 0.8.2
- Transformers 4.37.2
- Pytorch 2.2.0
- Datasets 2.16.1
- Tokenizers 0.15.2
|
{"base_model": "connectivity/cola_6ep_ft-33", "datasets": ["tweet_eval"], "library_name": "peft", "metrics": ["accuracy"], "tags": ["parquet", "text-classification"], "model-index": [{"name": "connectivity_cola_6ep_ft-33-finetuned-lora-tweet_eval_emotion", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "tweet_eval", "type": "tweet_eval", "config": "emotion", "split": "validation", "args": "emotion"}, "metrics": [{"type": "accuracy", "value": 0.6096256684491979, "name": "accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,857 |
itsrawat/finetuning-sentiment-model
|
itsrawat
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:imdb",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-03-12T04:27:03Z |
2023-03-12T04:33:18+00:00
| 11 | 0 |
---
datasets:
- imdb
license: apache-2.0
metrics:
- accuracy
- f1
tags:
- generated_from_trainer
model-index:
- name: finetuning-sentiment-model
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: imdb
type: imdb
config: plain_text
split: test
args: plain_text
metrics:
- type: accuracy
value: 0.8766666666666667
name: Accuracy
- type: f1
value: 0.877887788778878
name: F1
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# finetuning-sentiment-model
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the imdb dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2994
- Accuracy: 0.8767
- F1: 0.8779
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
### Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# finetuning-sentiment-model
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the imdb dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2994
- Accuracy: 0.8767
- F1: 0.8779
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
### Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2
|
{"datasets": ["imdb"], "license": "apache-2.0", "metrics": ["accuracy", "f1"], "tags": ["generated_from_trainer"], "model-index": [{"name": "finetuning-sentiment-model", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "imdb", "type": "imdb", "config": "plain_text", "split": "test", "args": "plain_text"}, "metrics": [{"type": "accuracy", "value": 0.8766666666666667, "name": "Accuracy"}, {"type": "f1", "value": 0.877887788778878, "name": "F1"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,858 |
Geerath/distilbert-base-uncased-distilled-squad
|
Geerath
|
question-answering
|
[
"transformers",
"tensorboard",
"safetensors",
"distilbert",
"question-answering",
"generated_from_trainer",
"base_model:distilbert/distilbert-base-uncased",
"base_model:finetune:distilbert/distilbert-base-uncased",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | 2024-02-07T18:25:28Z |
2024-02-07T19:49:11+00:00
| 17 | 0 |
---
base_model: distilbert-base-uncased
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: distilbert-base-uncased-distilled-squad
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-distilled-squad
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1892
## Model description
The DistilBERT model was proposed in the blog post Smaller, faster, cheaper, lighter: Introducing DistilBERT, adistilled version of BERT, and the paper DistilBERT, adistilled version of BERT: smaller, faster, cheaper and lighter. DistilBERT is a small, fast, cheap and light Transformer model trained by distilling BERT base. It has 40% less parameters than bert-base-uncased, runs 60% faster while preserving over 95% of BERT's performances as measured on the GLUE language understanding benchmark.
This model is a fine-tune checkpoint of DistilBERT-base-uncased, fine-tuned using (a second step of) knowledge distillation on SQuAD v1.1.
## Results are my own reproduction of the development by Hugging Face.
## How to Get Started with the Model
Use the code below:
from transformers import pipeline
question_answerer = pipeline("question-answering", model='distilbert-base-uncased-distilled-squad')
context = r"""
Extractive Question Answering is the task of extracting an answer from a text given a question. An example of a
question answering dataset is the SQuAD dataset, which is entirely based on that task. If you would like to fine-tune
a model on a SQuAD task, you may leverage the examples/pytorch/question-answering/run_squad.py script.
"""
result = question_answerer(question="What is a good example of a question answering dataset?", context=context)
print(
f"Answer: '{result['answer']}', score: {round(result['score'], 4)}, start: {result['start']}, end: {result['end']}"
# Here is how to use this model in PyTorch:
from transformers import DistilBertTokenizer, DistilBertForQuestionAnswering
import torch
tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-uncased-distilled-squad')
model = DistilBertForQuestionAnswering.from_pretrained('distilbert-base-uncased-distilled-squad')
question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
inputs = tokenizer(question, text, return_tensors="pt")
with torch.no_grad():
outputs = model(**inputs)
answer_start_index = torch.argmax(outputs.start_logits)
answer_end_index = torch.argmax(outputs.end_logits)
predict_answer_tokens = inputs.input_ids[0, answer_start_index : answer_end_index + 1]
tokenizer.decode(predict_answer_tokens)
# And in TensorFlow:
from transformers import DistilBertTokenizer, TFDistilBertForQuestionAnswering
import tensorflow as tf
tokenizer = DistilBertTokenizer.from_pretrained("distilbert-base-uncased-distilled-squad")
model = TFDistilBertForQuestionAnswering.from_pretrained("distilbert-base-uncased-distilled-squad")
question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
inputs = tokenizer(question, text, return_tensors="tf")
outputs = model(**inputs)
answer_start_index = int(tf.math.argmax(outputs.start_logits, axis=-1)[0])
answer_end_index = int(tf.math.argmax(outputs.end_logits, axis=-1)[0])
predict_answer_tokens = inputs.input_ids[0, answer_start_index : answer_end_index + 1]
tokenizer.decode(predict_answer_tokens)
## Uses:
This model can be used for question answering.
## Intended uses & limitations
CONTENT WARNING: Readers should be aware that language generated by this model can be disturbing or offensive to some and can propagate historical and current stereotypes.
## Training and evaluation data
This model reaches a F1 score of 82.75539002485876 and 'exact_match': 73.66130558183538 on the [SQuAD v1.1] dev set (for comparison, Bert bert-base-uncased version reaches a F1 score of 88.5).d
## Training procedure
Preprocessing
See the distilbert-base-uncased model card for further details.
Pretraining
See the distilbert-base-uncased model card for further details.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.2559 | 1.0 | 5533 | 1.1892 |
### Framework versions
- Transformers 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-distilled-squad
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1892
## Model description
The DistilBERT model was proposed in the blog post Smaller, faster, cheaper, lighter: Introducing DistilBERT, adistilled version of BERT, and the paper DistilBERT, adistilled version of BERT: smaller, faster, cheaper and lighter. DistilBERT is a small, fast, cheap and light Transformer model trained by distilling BERT base. It has 40% less parameters than bert-base-uncased, runs 60% faster while preserving over 95% of BERT's performances as measured on the GLUE language understanding benchmark.
This model is a fine-tune checkpoint of DistilBERT-base-uncased, fine-tuned using (a second step of) knowledge distillation on SQuAD v1.1.
## Results are my own reproduction of the development by Hugging Face.
## How to Get Started with the Model
Use the code below:
from transformers import pipeline
question_answerer = pipeline("question-answering", model='distilbert-base-uncased-distilled-squad')
context = r"""
Extractive Question Answering is the task of extracting an answer from a text given a question. An example of a
question answering dataset is the SQuAD dataset, which is entirely based on that task. If you would like to fine-tune
a model on a SQuAD task, you may leverage the examples/pytorch/question-answering/run_squad.py script.
"""
result = question_answerer(question="What is a good example of a question answering dataset?", context=context)
print(
f"Answer: '{result['answer']}', score: {round(result['score'], 4)}, start: {result['start']}, end: {result['end']}"
# Here is how to use this model in PyTorch:
from transformers import DistilBertTokenizer, DistilBertForQuestionAnswering
import torch
tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-uncased-distilled-squad')
model = DistilBertForQuestionAnswering.from_pretrained('distilbert-base-uncased-distilled-squad')
question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
inputs = tokenizer(question, text, return_tensors="pt")
with torch.no_grad():
outputs = model(**inputs)
answer_start_index = torch.argmax(outputs.start_logits)
answer_end_index = torch.argmax(outputs.end_logits)
predict_answer_tokens = inputs.input_ids[0, answer_start_index : answer_end_index + 1]
tokenizer.decode(predict_answer_tokens)
# And in TensorFlow:
from transformers import DistilBertTokenizer, TFDistilBertForQuestionAnswering
import tensorflow as tf
tokenizer = DistilBertTokenizer.from_pretrained("distilbert-base-uncased-distilled-squad")
model = TFDistilBertForQuestionAnswering.from_pretrained("distilbert-base-uncased-distilled-squad")
question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
inputs = tokenizer(question, text, return_tensors="tf")
outputs = model(**inputs)
answer_start_index = int(tf.math.argmax(outputs.start_logits, axis=-1)[0])
answer_end_index = int(tf.math.argmax(outputs.end_logits, axis=-1)[0])
predict_answer_tokens = inputs.input_ids[0, answer_start_index : answer_end_index + 1]
tokenizer.decode(predict_answer_tokens)
## Uses:
This model can be used for question answering.
## Intended uses & limitations
CONTENT WARNING: Readers should be aware that language generated by this model can be disturbing or offensive to some and can propagate historical and current stereotypes.
## Training and evaluation data
This model reaches a F1 score of 82.75539002485876 and 'exact_match': 73.66130558183538 on the [SQuAD v1.1] dev set (for comparison, Bert bert-base-uncased version reaches a F1 score of 88.5).d
## Training procedure
Preprocessing
See the distilbert-base-uncased model card for further details.
Pretraining
See the distilbert-base-uncased model card for further details.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.2559 | 1.0 | 5533 | 1.1892 |
### Framework versions
- Transformers 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
{"base_model": "distilbert-base-uncased", "license": "apache-2.0", "tags": ["generated_from_trainer"], "model-index": [{"name": "distilbert-base-uncased-distilled-squad", "results": []}]}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,859 |
Cheng98/bert-large-sst2
|
Cheng98
|
text-classification
|
[
"transformers",
"pytorch",
"bert",
"text-classification",
"generated_from_trainer",
"en",
"dataset:glue",
"base_model:google-bert/bert-large-cased",
"base_model:finetune:google-bert/bert-large-cased",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-04-06T16:58:19Z |
2024-04-08T21:08:27+00:00
| 27 | 0 |
---
base_model: bert-large-cased
datasets:
- glue
language:
- en
license: apache-2.0
metrics:
- accuracy
tags:
- generated_from_trainer
model-index:
- name: bert-large-sst2
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: GLUE SST2
type: glue
args: sst2
metrics:
- type: accuracy
value: 0.9254587155963303
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-large-sst2
This model is a fine-tuned version of [bert-large-cased](https://huggingface.co/bert-large-cased) on the GLUE SST2 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3748
- Accuracy: 0.9255
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
### Training results
### Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu117
- Datasets 2.18.0
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-large-sst2
This model is a fine-tuned version of [bert-large-cased](https://huggingface.co/bert-large-cased) on the GLUE SST2 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3748
- Accuracy: 0.9255
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
### Training results
### Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu117
- Datasets 2.18.0
- Tokenizers 0.13.3
|
{"base_model": "bert-large-cased", "datasets": ["glue"], "language": ["en"], "license": "apache-2.0", "metrics": ["accuracy"], "tags": ["generated_from_trainer"], "model-index": [{"name": "bert-large-sst2", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "GLUE SST2", "type": "glue", "args": "sst2"}, "metrics": [{"type": "accuracy", "value": 0.9254587155963303, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,860 |
davelotito/donut_experiment_bayesian_trial_10
|
davelotito
|
image-text-to-text
|
[
"transformers",
"tensorboard",
"safetensors",
"vision-encoder-decoder",
"image-text-to-text",
"generated_from_trainer",
"base_model:naver-clova-ix/donut-base",
"base_model:finetune:naver-clova-ix/donut-base",
"license:mit",
"endpoints_compatible",
"region:us"
] | 2024-06-26T13:56:23Z |
2024-06-26T14:44:37+00:00
| 4 | 0 |
---
base_model: naver-clova-ix/donut-base
license: mit
metrics:
- bleu
- wer
tags:
- generated_from_trainer
model-index:
- name: donut_experiment_bayesian_trial_10
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# donut_experiment_bayesian_trial_10
This model is a fine-tuned version of [naver-clova-ix/donut-base](https://huggingface.co/naver-clova-ix/donut-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4219
- Bleu: 0.0632
- Precisions: [0.809322033898305, 0.7493975903614458, 0.7094972067039106, 0.6644518272425249]
- Brevity Penalty: 0.0864
- Length Ratio: 0.2899
- Translation Length: 472
- Reference Length: 1628
- Cer: 0.7596
- Wer: 0.8312
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1.0082458996730595e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 2
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Precisions | Brevity Penalty | Length Ratio | Translation Length | Reference Length | Cer | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|:--------------------------------------------------------------------------------:|:---------------:|:------------:|:------------------:|:----------------:|:------:|:------:|
| 0.3888 | 1.0 | 253 | 0.5110 | 0.0673 | [0.7909836065573771, 0.7099767981438515, 0.6657754010695187, 0.6277602523659306] | 0.0967 | 0.2998 | 488 | 1628 | 0.7690 | 0.8412 |
| 0.326 | 2.0 | 506 | 0.4539 | 0.0654 | [0.7908902691511387, 0.7276995305164319, 0.6775067750677507, 0.6153846153846154] | 0.0934 | 0.2967 | 483 | 1628 | 0.7604 | 0.8362 |
| 0.3191 | 3.0 | 759 | 0.4256 | 0.0654 | [0.7837837837837838, 0.7287735849056604, 0.6893732970027248, 0.6451612903225806] | 0.0921 | 0.2955 | 481 | 1628 | 0.7599 | 0.8331 |
| 0.2632 | 4.0 | 1012 | 0.4219 | 0.0632 | [0.809322033898305, 0.7493975903614458, 0.7094972067039106, 0.6644518272425249] | 0.0864 | 0.2899 | 472 | 1628 | 0.7596 | 0.8312 |
### Framework versions
- Transformers 4.40.0
- Pytorch 2.1.0
- Datasets 2.18.0
- Tokenizers 0.19.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# donut_experiment_bayesian_trial_10
This model is a fine-tuned version of [naver-clova-ix/donut-base](https://huggingface.co/naver-clova-ix/donut-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4219
- Bleu: 0.0632
- Precisions: [0.809322033898305, 0.7493975903614458, 0.7094972067039106, 0.6644518272425249]
- Brevity Penalty: 0.0864
- Length Ratio: 0.2899
- Translation Length: 472
- Reference Length: 1628
- Cer: 0.7596
- Wer: 0.8312
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1.0082458996730595e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 2
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Precisions | Brevity Penalty | Length Ratio | Translation Length | Reference Length | Cer | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|:--------------------------------------------------------------------------------:|:---------------:|:------------:|:------------------:|:----------------:|:------:|:------:|
| 0.3888 | 1.0 | 253 | 0.5110 | 0.0673 | [0.7909836065573771, 0.7099767981438515, 0.6657754010695187, 0.6277602523659306] | 0.0967 | 0.2998 | 488 | 1628 | 0.7690 | 0.8412 |
| 0.326 | 2.0 | 506 | 0.4539 | 0.0654 | [0.7908902691511387, 0.7276995305164319, 0.6775067750677507, 0.6153846153846154] | 0.0934 | 0.2967 | 483 | 1628 | 0.7604 | 0.8362 |
| 0.3191 | 3.0 | 759 | 0.4256 | 0.0654 | [0.7837837837837838, 0.7287735849056604, 0.6893732970027248, 0.6451612903225806] | 0.0921 | 0.2955 | 481 | 1628 | 0.7599 | 0.8331 |
| 0.2632 | 4.0 | 1012 | 0.4219 | 0.0632 | [0.809322033898305, 0.7493975903614458, 0.7094972067039106, 0.6644518272425249] | 0.0864 | 0.2899 | 472 | 1628 | 0.7596 | 0.8312 |
### Framework versions
- Transformers 4.40.0
- Pytorch 2.1.0
- Datasets 2.18.0
- Tokenizers 0.19.1
|
{"base_model": "naver-clova-ix/donut-base", "license": "mit", "metrics": ["bleu", "wer"], "tags": ["generated_from_trainer"], "model-index": [{"name": "donut_experiment_bayesian_trial_10", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,861 |
gokuls/hBERTv2_wnli
|
gokuls
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"hybridbert",
"text-classification",
"generated_from_trainer",
"en",
"dataset:glue",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-03-02T11:57:22Z |
2023-03-02T12:01:08+00:00
| 12 | 0 |
---
datasets:
- glue
language:
- en
metrics:
- accuracy
tags:
- generated_from_trainer
model-index:
- name: hBERTv2_wnli
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: GLUE WNLI
type: glue
config: wnli
split: validation
args: wnli
metrics:
- type: accuracy
value: 0.5633802816901409
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# hBERTv2_wnli
This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2](https://huggingface.co/gokuls/bert_12_layer_model_v2) on the GLUE WNLI dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6833
- Accuracy: 0.5634
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.7351 | 1.0 | 3 | 0.7260 | 0.5211 |
| 0.7223 | 2.0 | 6 | 0.6833 | 0.5634 |
| 0.7189 | 3.0 | 9 | 0.7110 | 0.4507 |
| 0.708 | 4.0 | 12 | 0.7059 | 0.5352 |
| 0.7032 | 5.0 | 15 | 0.6925 | 0.5352 |
| 0.6987 | 6.0 | 18 | 0.7121 | 0.4225 |
| 0.7109 | 7.0 | 21 | 0.6928 | 0.5352 |
### Framework versions
- Transformers 4.26.1
- Pytorch 1.14.0a0+410ce96
- Datasets 2.10.1
- Tokenizers 0.13.2
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# hBERTv2_wnli
This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2](https://huggingface.co/gokuls/bert_12_layer_model_v2) on the GLUE WNLI dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6833
- Accuracy: 0.5634
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.7351 | 1.0 | 3 | 0.7260 | 0.5211 |
| 0.7223 | 2.0 | 6 | 0.6833 | 0.5634 |
| 0.7189 | 3.0 | 9 | 0.7110 | 0.4507 |
| 0.708 | 4.0 | 12 | 0.7059 | 0.5352 |
| 0.7032 | 5.0 | 15 | 0.6925 | 0.5352 |
| 0.6987 | 6.0 | 18 | 0.7121 | 0.4225 |
| 0.7109 | 7.0 | 21 | 0.6928 | 0.5352 |
### Framework versions
- Transformers 4.26.1
- Pytorch 1.14.0a0+410ce96
- Datasets 2.10.1
- Tokenizers 0.13.2
|
{"datasets": ["glue"], "language": ["en"], "metrics": ["accuracy"], "tags": ["generated_from_trainer"], "model-index": [{"name": "hBERTv2_wnli", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "GLUE WNLI", "type": "glue", "config": "wnli", "split": "validation", "args": "wnli"}, "metrics": [{"type": "accuracy", "value": 0.5633802816901409, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,862 |
RichardErkhov/ChocoLlama_-_ChocoLlama-2-7B-instruct-8bits
|
RichardErkhov
| null |
[
"safetensors",
"llama",
"arxiv:2310.03477",
"arxiv:2412.07633",
"arxiv:2312.12852",
"8-bit",
"bitsandbytes",
"region:us"
] | 2025-03-16T08:50:51Z |
2025-03-16T08:54:51+00:00
| 2 | 0 |
---
{}
---
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
ChocoLlama-2-7B-instruct - bnb 8bits
- Model creator: https://huggingface.co/ChocoLlama/
- Original model: https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-instruct/
Original model description:
---
language:
- nl
license: cc-by-nc-4.0
base_model: ChocoLlama/ChocoLlama-2-7B-base
datasets:
- BramVanroy/ultrachat_200k_dutch
- BramVanroy/stackoverflow-chat-dutch
- BramVanroy/alpaca-cleaned-dutch
- BramVanroy/dolly-15k-dutch
- BramVanroy/no_robots_dutch
- BramVanroy/ultra_feedback_dutch
---
<p align="center" style="margin:0;padding:0">
<img src="./chocollama_logo.png" alt="ChocoLlama logo" width="500" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
</p>
<div style="margin:auto; text-align:center">
<h1 style="margin-bottom: 0">ChocoLlama</h1>
<em>A Llama-2/3-based family of Dutch language models</em>
</div>
## ChocoLlama-2-7B-instruct: Getting Started
We here present **ChocoLlama-2-7B-instruct**, an instruction-tuned version of ChocoLlama-2-7B-base, fine-tuned on a collection of Dutch translations of instruction-tuning datasets, using SFT followed by DPO.
Its base model, [ChocoLlama-2-7B-base](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-base), is a language-adapted version of Meta's Llama-2-7b, fine-tuned on 32B Dutch Llama-2 tokens (104GB) using LoRa.
Use the code below to get started with the model.
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained('ChocoLlama/ChocoLlama-2-7B-instruct')
model = AutoModelForCausalLM.from_pretrained('ChocoLlama/ChocoLlama-2-7B-instruct', device_map="auto")
messages = [
{"role": "system", "content": "Je bent een artificiële intelligentie-assistent en geeft behulpzame, gedetailleerde en beleefde antwoorden op de vragen van de gebruiker."},
{"role": "user", "content": "Jacques brel, Willem Elsschot en Jan Jambon zitten op café. Waar zouden ze over babbelen?"},
]
input_ids = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt"
).to(model.device)
new_terminators = [
tokenizer.eos_token_id,
tokenizer.convert_tokens_to_ids("<|eot_id|>")
]
outputs = model.generate(
input_ids,
max_new_tokens=512,
eos_token_id=new_terminators,
do_sample=True,
temperature=0.8,
top_p=0.95,
)
response = outputs[0][input_ids.shape[-1]:]
print(tokenizer.decode(response, skip_special_tokens=True))
```
Note that the datasets used for instruction-tuning were translated using GPT-3.5/4, which means that this instruction-tuned model can not be used for commercial purposes.
Hence, for any commercial applications, we recommend finetuning the base model on your own Dutch data.
## Model Details
ChocoLlama is a family of open LLM's specifically adapted to Dutch, contributing to the state-of-the-art of Dutch open LLM's in their weight class.
We provide 6 variants (of which 3 base and 3 instruction-tuned models):
- **ChocoLlama-2-7B-base** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-base)): A language-adapted version of Meta's Llama-2-7b, fine-tuned on 32B Dutch Llama-2 tokens (104GB) using LoRa.
- **ChocoLlama-2-7B-instruct** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-instruct)): An instruction-tuned version of ChocoLlama-2-7B-base, fine-tuned on a collection of Dutch translations of instruction-tuning datasets, using SFT followed by DPO.
- **ChocoLlama-2-7B-tokentrans-base** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-tokentrans-base)): A language-adapted version of Meta's Llama-2-7b, using a Dutch RoBERTa-based tokenizer. The token embeddings of this model were reinitialized using the token translation algorithm proposed by [Remy et al.](https://arxiv.org/pdf/2310.03477). The model was subsequently fine-tuned on the same Dutch dataset as ChocoLlama-2-7B-base, again using LoRa.
- **ChocoLlama-2-7B-tokentrans-instruct** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-tokentrans-instruct)): An instruction-tuned version of ChocoLlama-2-7B-tokentrans-base, fine-tuned on the same dataset as ChocoLlama-2-7B-instruct, again using SFT followed by DPO.
- **Llama-3-ChocoLlama-8B-base** ([link](https://huggingface.co/ChocoLlama/Llama-3-ChocoLlama-8B-base)): A language-adapted version of Meta's Llama-8-8B, fine-tuned on the same Dutch dataset as ChocoLlama-2-7B-base, again using LoRa.
- **Llama-3-ChocoLlama-instruct** ([link](https://huggingface.co/ChocoLlama/Llama-3-ChocoLlama-8B-instruct)): An instruction-tuned version of Llama-3-ChocoLlama-8B-base, fine-tuned on the same dataset as ChocoLlama-2-7B-instruct, again using SFT followed by DPO.
For benchmark results for all models, including compared to their base models and other Dutch LLMs, we refer to our paper [here](https://arxiv.org/pdf/2412.07633).
### Model Description
- **Developed by:** [Matthieu Meeus](https://huggingface.co/matthieumeeus97), [Anthony Rathé](https://huggingface.co/anthonyrathe)
- **Funded by:** [Vlaams Supercomputer Centrum](https://www.vscentrum.be/), through a grant of apx. 40K GPU hours (NVIDIA A100-80GB)
- **Language(s):** Dutch
- **License:** cc-by-nc-4.0
- **Finetuned from model:** [ChocoLlama-2-7B-base](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-base)
### Model Sources
- **Repository:** [on Github here](https://github.com/ChocoLlamaModel/ChocoLlama).
- **Paper:** [on ArXiv here](https://arxiv.org/pdf/2412.07633).
## Uses
### Direct Use
This is an instruction-tuned (SFT + DPO) Dutch model, optimized for Dutch language generation in conversational settings.
For optimal behavior, we advice to only use the model with the correct chat template (see Python code above), potentially supported by a system prompt.
### Out-of-Scope Use
Use-cases requiring understanding or generation of text in languages other than Dutch: the dataset on which this model was fine-tuned does not contain data in languages other than Dutch, hence we expect significant catastrophic forgetting to have occured for English, which is the language Llama-2 was originally trained for.
## Bias, Risks, and Limitations
We have taken care to include only widely used and high-quality data in our dataset. Some of this data has been filtered by the original creators.
However we did not explicitly conduct any additional filtering of this dataset with regards to biased or otherwise harmful content.
## Training Details
We adopt the same strategy as used to align GEITje-7B to [GEITje-7B-ultra](https://huggingface.co/BramVanroy/GEITje-7B-ultra).
First, we apply supervised finetuning (SFT), utilizing the data made available by [Vanroy](https://arxiv.org/pdf/2312.12852):
- [BramVanroy/ultrachat_200k_dutch](https://huggingface.co/datasets/BramVanroy/ultrachat_200k_dutch)
- [BramVanroy/no_robots_dutch](https://huggingface.co/datasets/BramVanroy/no_robots_dutch)
- [BramVanroy/stackoverflow-chat-dutch](https://huggingface.co/datasets/BramVanroy/stackoverflow-chat-dutch)
- [BramVanroy/alpaca-cleaned-dutch](https://huggingface.co/datasets/BramVanroy/alpaca-cleaned-dutch)
- [BramVanroy/dolly-15k-dutch](https://huggingface.co/datasets/BramVanroy/dolly-15k-dutch)
Next, we apply Direct Preference Optimization (DPO) to the SFT version of all the pretrained models we here develop,
now utilizing a Dutch version of the data used to train Zephyr-7B-$\beta$, [BramVanroy/ultra_feedback_dutch](https://huggingface.co/datasets/BramVanroy/ultra_feedback_dutch).
For both the SFT and DPO stage, we update all model weights and apply the same set of hyperparameters to all models as used in GEITje-7B-ultra:
- learning_rate: 5e-07
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
Further, we leverage the publicly available [alignment handbook](https://github.com/huggingface/alignment-handbook) and use a set of 4 NVIDIA A100 (80 GB) for both stages.
## Evaluation
### Quantitative evaluation
We have evaluated our models on several industry-standard Dutch benchmarks, translated from their original versions. The results can be found in the table below, together with results from several other prominent Dutch models.
| Model | ARC | HellaSwag | MMLU | TruthfulQA | Avg. |
|----------------------------------------------|----------------|----------------|----------------|----------------|----------------|
| **Llama-3-ChocoLlama-instruct** | **0.48** | **0.66** | **0.49** | **0.49** | **0.53** |
| llama-3-8B-rebatch | 0.44 | 0.64 | 0.46 | 0.48 | 0.51 |
| llama-3-8B-instruct | 0.47 | 0.59 | 0.47 | 0.52 | 0.51 |
| llama-3-8B | 0.44 | 0.64 | 0.47 | 0.45 | 0.5 |
| Reynaerde-7B-Chat | 0.44 | 0.62 | 0.39 | 0.52 | 0.49 |
| **Llama-3-ChocoLlama-base** | **0.45** | **0.64** | **0.44** | **0.44** | **0.49** |
| zephyr-7b-beta | 0.43 | 0.58 | 0.43 | 0.53 | 0.49 |
| geitje-7b-ultra | 0.40 | 0.66 | 0.36 | 0.49 | 0.48 |
| **ChocoLlama-2-7B-tokentrans-instruct** | **0.45** | **0.62** | **0.34** | **0.42** | **0.46** |
| mistral-7b-v0.1 | 0.43 | 0.58 | 0.37 | 0.45 | 0.46 |
| **ChocoLlama-2-7B-tokentrans-base** | **0.42** | **0.61** | **0.32** | **0.43** | **0.45** |
| **ChocoLlama-2-7B-instruct** | **0.36** | **0.57** | **0.33** | **0.45** | **0.43 |
| **ChocoLlama-2-7B-base** | **0.35** | **0.56** | **0.31** | **0.43** | **0.41** |
| llama-2-7b-chat-hf | 0.36 | 0.49 | 0.33 | 0.44 | 0.41 |
| llama-2-7b-hf | 0.36 | 0.51 | 0.32 | 0.41 | 0.40 |
On average, Llama-3-ChocoLlama-instruct surpasses the previous state-of-the-art on these benchmarks.
### Qualitative evaluation
In our paper, we also provide an additional qualitative evaluation of all models - which we empirically find more reliable.
For details, we refer to the paper and to our benchmark [ChocoLlama-Bench](https://huggingface.co/datasets/ChocoLlama/ChocoLlama-Bench).
### Compute Infrastructure
All ChocoLlama models have been trained on the compute cluster provided by the [Flemish Supercomputer Center (VSC)](https://www.vscentrum.be/). We used 8 to 16 NVIDIA A100 GPU's with 80 GB of VRAM.
## Citation
If you found this useful for your work, kindly cite our paper:
```
@article{meeus2024chocollama,
title={ChocoLlama: Lessons Learned From Teaching Llamas Dutch},
author={Meeus, Matthieu and Rath{\'e}, Anthony and Remy, Fran{\c{c}}ois and Delobelle, Pieter and Decorte, Jens-Joris and Demeester, Thomas},
journal={arXiv preprint arXiv:2412.07633},
year={2024}
}
```
| null |
Non_BioNLP
|
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
ChocoLlama-2-7B-instruct - bnb 8bits
- Model creator: https://huggingface.co/ChocoLlama/
- Original model: https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-instruct/
Original model description:
---
language:
- nl
license: cc-by-nc-4.0
base_model: ChocoLlama/ChocoLlama-2-7B-base
datasets:
- BramVanroy/ultrachat_200k_dutch
- BramVanroy/stackoverflow-chat-dutch
- BramVanroy/alpaca-cleaned-dutch
- BramVanroy/dolly-15k-dutch
- BramVanroy/no_robots_dutch
- BramVanroy/ultra_feedback_dutch
---
<p align="center" style="margin:0;padding:0">
<img src="./chocollama_logo.png" alt="ChocoLlama logo" width="500" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
</p>
<div style="margin:auto; text-align:center">
<h1 style="margin-bottom: 0">ChocoLlama</h1>
<em>A Llama-2/3-based family of Dutch language models</em>
</div>
## ChocoLlama-2-7B-instruct: Getting Started
We here present **ChocoLlama-2-7B-instruct**, an instruction-tuned version of ChocoLlama-2-7B-base, fine-tuned on a collection of Dutch translations of instruction-tuning datasets, using SFT followed by DPO.
Its base model, [ChocoLlama-2-7B-base](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-base), is a language-adapted version of Meta's Llama-2-7b, fine-tuned on 32B Dutch Llama-2 tokens (104GB) using LoRa.
Use the code below to get started with the model.
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained('ChocoLlama/ChocoLlama-2-7B-instruct')
model = AutoModelForCausalLM.from_pretrained('ChocoLlama/ChocoLlama-2-7B-instruct', device_map="auto")
messages = [
{"role": "system", "content": "Je bent een artificiële intelligentie-assistent en geeft behulpzame, gedetailleerde en beleefde antwoorden op de vragen van de gebruiker."},
{"role": "user", "content": "Jacques brel, Willem Elsschot en Jan Jambon zitten op café. Waar zouden ze over babbelen?"},
]
input_ids = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt"
).to(model.device)
new_terminators = [
tokenizer.eos_token_id,
tokenizer.convert_tokens_to_ids("<|eot_id|>")
]
outputs = model.generate(
input_ids,
max_new_tokens=512,
eos_token_id=new_terminators,
do_sample=True,
temperature=0.8,
top_p=0.95,
)
response = outputs[0][input_ids.shape[-1]:]
print(tokenizer.decode(response, skip_special_tokens=True))
```
Note that the datasets used for instruction-tuning were translated using GPT-3.5/4, which means that this instruction-tuned model can not be used for commercial purposes.
Hence, for any commercial applications, we recommend finetuning the base model on your own Dutch data.
## Model Details
ChocoLlama is a family of open LLM's specifically adapted to Dutch, contributing to the state-of-the-art of Dutch open LLM's in their weight class.
We provide 6 variants (of which 3 base and 3 instruction-tuned models):
- **ChocoLlama-2-7B-base** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-base)): A language-adapted version of Meta's Llama-2-7b, fine-tuned on 32B Dutch Llama-2 tokens (104GB) using LoRa.
- **ChocoLlama-2-7B-instruct** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-instruct)): An instruction-tuned version of ChocoLlama-2-7B-base, fine-tuned on a collection of Dutch translations of instruction-tuning datasets, using SFT followed by DPO.
- **ChocoLlama-2-7B-tokentrans-base** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-tokentrans-base)): A language-adapted version of Meta's Llama-2-7b, using a Dutch RoBERTa-based tokenizer. The token embeddings of this model were reinitialized using the token translation algorithm proposed by [Remy et al.](https://arxiv.org/pdf/2310.03477). The model was subsequently fine-tuned on the same Dutch dataset as ChocoLlama-2-7B-base, again using LoRa.
- **ChocoLlama-2-7B-tokentrans-instruct** ([link](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-tokentrans-instruct)): An instruction-tuned version of ChocoLlama-2-7B-tokentrans-base, fine-tuned on the same dataset as ChocoLlama-2-7B-instruct, again using SFT followed by DPO.
- **Llama-3-ChocoLlama-8B-base** ([link](https://huggingface.co/ChocoLlama/Llama-3-ChocoLlama-8B-base)): A language-adapted version of Meta's Llama-8-8B, fine-tuned on the same Dutch dataset as ChocoLlama-2-7B-base, again using LoRa.
- **Llama-3-ChocoLlama-instruct** ([link](https://huggingface.co/ChocoLlama/Llama-3-ChocoLlama-8B-instruct)): An instruction-tuned version of Llama-3-ChocoLlama-8B-base, fine-tuned on the same dataset as ChocoLlama-2-7B-instruct, again using SFT followed by DPO.
For benchmark results for all models, including compared to their base models and other Dutch LLMs, we refer to our paper [here](https://arxiv.org/pdf/2412.07633).
### Model Description
- **Developed by:** [Matthieu Meeus](https://huggingface.co/matthieumeeus97), [Anthony Rathé](https://huggingface.co/anthonyrathe)
- **Funded by:** [Vlaams Supercomputer Centrum](https://www.vscentrum.be/), through a grant of apx. 40K GPU hours (NVIDIA A100-80GB)
- **Language(s):** Dutch
- **License:** cc-by-nc-4.0
- **Finetuned from model:** [ChocoLlama-2-7B-base](https://huggingface.co/ChocoLlama/ChocoLlama-2-7B-base)
### Model Sources
- **Repository:** [on Github here](https://github.com/ChocoLlamaModel/ChocoLlama).
- **Paper:** [on ArXiv here](https://arxiv.org/pdf/2412.07633).
## Uses
### Direct Use
This is an instruction-tuned (SFT + DPO) Dutch model, optimized for Dutch language generation in conversational settings.
For optimal behavior, we advice to only use the model with the correct chat template (see Python code above), potentially supported by a system prompt.
### Out-of-Scope Use
Use-cases requiring understanding or generation of text in languages other than Dutch: the dataset on which this model was fine-tuned does not contain data in languages other than Dutch, hence we expect significant catastrophic forgetting to have occured for English, which is the language Llama-2 was originally trained for.
## Bias, Risks, and Limitations
We have taken care to include only widely used and high-quality data in our dataset. Some of this data has been filtered by the original creators.
However we did not explicitly conduct any additional filtering of this dataset with regards to biased or otherwise harmful content.
## Training Details
We adopt the same strategy as used to align GEITje-7B to [GEITje-7B-ultra](https://huggingface.co/BramVanroy/GEITje-7B-ultra).
First, we apply supervised finetuning (SFT), utilizing the data made available by [Vanroy](https://arxiv.org/pdf/2312.12852):
- [BramVanroy/ultrachat_200k_dutch](https://huggingface.co/datasets/BramVanroy/ultrachat_200k_dutch)
- [BramVanroy/no_robots_dutch](https://huggingface.co/datasets/BramVanroy/no_robots_dutch)
- [BramVanroy/stackoverflow-chat-dutch](https://huggingface.co/datasets/BramVanroy/stackoverflow-chat-dutch)
- [BramVanroy/alpaca-cleaned-dutch](https://huggingface.co/datasets/BramVanroy/alpaca-cleaned-dutch)
- [BramVanroy/dolly-15k-dutch](https://huggingface.co/datasets/BramVanroy/dolly-15k-dutch)
Next, we apply Direct Preference Optimization (DPO) to the SFT version of all the pretrained models we here develop,
now utilizing a Dutch version of the data used to train Zephyr-7B-$\beta$, [BramVanroy/ultra_feedback_dutch](https://huggingface.co/datasets/BramVanroy/ultra_feedback_dutch).
For both the SFT and DPO stage, we update all model weights and apply the same set of hyperparameters to all models as used in GEITje-7B-ultra:
- learning_rate: 5e-07
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
Further, we leverage the publicly available [alignment handbook](https://github.com/huggingface/alignment-handbook) and use a set of 4 NVIDIA A100 (80 GB) for both stages.
## Evaluation
### Quantitative evaluation
We have evaluated our models on several industry-standard Dutch benchmarks, translated from their original versions. The results can be found in the table below, together with results from several other prominent Dutch models.
| Model | ARC | HellaSwag | MMLU | TruthfulQA | Avg. |
|----------------------------------------------|----------------|----------------|----------------|----------------|----------------|
| **Llama-3-ChocoLlama-instruct** | **0.48** | **0.66** | **0.49** | **0.49** | **0.53** |
| llama-3-8B-rebatch | 0.44 | 0.64 | 0.46 | 0.48 | 0.51 |
| llama-3-8B-instruct | 0.47 | 0.59 | 0.47 | 0.52 | 0.51 |
| llama-3-8B | 0.44 | 0.64 | 0.47 | 0.45 | 0.5 |
| Reynaerde-7B-Chat | 0.44 | 0.62 | 0.39 | 0.52 | 0.49 |
| **Llama-3-ChocoLlama-base** | **0.45** | **0.64** | **0.44** | **0.44** | **0.49** |
| zephyr-7b-beta | 0.43 | 0.58 | 0.43 | 0.53 | 0.49 |
| geitje-7b-ultra | 0.40 | 0.66 | 0.36 | 0.49 | 0.48 |
| **ChocoLlama-2-7B-tokentrans-instruct** | **0.45** | **0.62** | **0.34** | **0.42** | **0.46** |
| mistral-7b-v0.1 | 0.43 | 0.58 | 0.37 | 0.45 | 0.46 |
| **ChocoLlama-2-7B-tokentrans-base** | **0.42** | **0.61** | **0.32** | **0.43** | **0.45** |
| **ChocoLlama-2-7B-instruct** | **0.36** | **0.57** | **0.33** | **0.45** | **0.43 |
| **ChocoLlama-2-7B-base** | **0.35** | **0.56** | **0.31** | **0.43** | **0.41** |
| llama-2-7b-chat-hf | 0.36 | 0.49 | 0.33 | 0.44 | 0.41 |
| llama-2-7b-hf | 0.36 | 0.51 | 0.32 | 0.41 | 0.40 |
On average, Llama-3-ChocoLlama-instruct surpasses the previous state-of-the-art on these benchmarks.
### Qualitative evaluation
In our paper, we also provide an additional qualitative evaluation of all models - which we empirically find more reliable.
For details, we refer to the paper and to our benchmark [ChocoLlama-Bench](https://huggingface.co/datasets/ChocoLlama/ChocoLlama-Bench).
### Compute Infrastructure
All ChocoLlama models have been trained on the compute cluster provided by the [Flemish Supercomputer Center (VSC)](https://www.vscentrum.be/). We used 8 to 16 NVIDIA A100 GPU's with 80 GB of VRAM.
## Citation
If you found this useful for your work, kindly cite our paper:
```
@article{meeus2024chocollama,
title={ChocoLlama: Lessons Learned From Teaching Llamas Dutch},
author={Meeus, Matthieu and Rath{\'e}, Anthony and Remy, Fran{\c{c}}ois and Delobelle, Pieter and Decorte, Jens-Joris and Demeester, Thomas},
journal={arXiv preprint arXiv:2412.07633},
year={2024}
}
```
|
{}
|
task
|
[
"TRANSLATION"
] | 39,863 |
Nextcloud-AI/opus-mt-ar-es
|
Nextcloud-AI
|
translation
|
[
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"ar",
"es",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-02-23T10:37:11Z |
2023-08-16T11:25:40+00:00
| 15 | 0 |
---
language:
- ar
- es
license: apache-2.0
tags:
- translation
---
### ara-spa
* source group: Arabic
* target group: Spanish
* OPUS readme: [ara-spa](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ara-spa/README.md)
* model: transformer
* source language(s): apc apc_Latn ara arq
* target language(s): spa
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opus-2020-07-03.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.zip)
* test set translations: [opus-2020-07-03.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.test.txt)
* test set scores: [opus-2020-07-03.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.ara.spa | 46.0 | 0.641 |
### System Info:
- hf_name: ara-spa
- source_languages: ara
- target_languages: spa
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ara-spa/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['ar', 'es']
- src_constituents: {'apc', 'ara', 'arq_Latn', 'arq', 'afb', 'ara_Latn', 'apc_Latn', 'arz'}
- tgt_constituents: {'spa'}
- src_multilingual: False
- tgt_multilingual: False
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.test.txt
- src_alpha3: ara
- tgt_alpha3: spa
- short_pair: ar-es
- chrF2_score: 0.6409999999999999
- bleu: 46.0
- brevity_penalty: 0.9620000000000001
- ref_len: 9708.0
- src_name: Arabic
- tgt_name: Spanish
- train_date: 2020-07-03
- src_alpha2: ar
- tgt_alpha2: es
- prefer_old: False
- long_pair: ara-spa
- helsinki_git_sha: 480fcbe0ee1bf4774bcbe6226ad9f58e63f6c535
- transformers_git_sha: 2207e5d8cb224e954a7cba69fa4ac2309e9ff30b
- port_machine: brutasse
- port_time: 2020-08-21-14:41
| null |
Non_BioNLP
|
### ara-spa
* source group: Arabic
* target group: Spanish
* OPUS readme: [ara-spa](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ara-spa/README.md)
* model: transformer
* source language(s): apc apc_Latn ara arq
* target language(s): spa
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opus-2020-07-03.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.zip)
* test set translations: [opus-2020-07-03.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.test.txt)
* test set scores: [opus-2020-07-03.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.ara.spa | 46.0 | 0.641 |
### System Info:
- hf_name: ara-spa
- source_languages: ara
- target_languages: spa
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ara-spa/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['ar', 'es']
- src_constituents: {'apc', 'ara', 'arq_Latn', 'arq', 'afb', 'ara_Latn', 'apc_Latn', 'arz'}
- tgt_constituents: {'spa'}
- src_multilingual: False
- tgt_multilingual: False
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/ara-spa/opus-2020-07-03.test.txt
- src_alpha3: ara
- tgt_alpha3: spa
- short_pair: ar-es
- chrF2_score: 0.6409999999999999
- bleu: 46.0
- brevity_penalty: 0.9620000000000001
- ref_len: 9708.0
- src_name: Arabic
- tgt_name: Spanish
- train_date: 2020-07-03
- src_alpha2: ar
- tgt_alpha2: es
- prefer_old: False
- long_pair: ara-spa
- helsinki_git_sha: 480fcbe0ee1bf4774bcbe6226ad9f58e63f6c535
- transformers_git_sha: 2207e5d8cb224e954a7cba69fa4ac2309e9ff30b
- port_machine: brutasse
- port_time: 2020-08-21-14:41
|
{"language": ["ar", "es"], "license": "apache-2.0", "tags": ["translation"]}
|
task
|
[
"TRANSLATION"
] | 39,864 |
fathyshalab/domain_transfer_general-massive_datetime-roberta-large-v1-5-94
|
fathyshalab
|
text-classification
|
[
"sentence-transformers",
"pytorch",
"roberta",
"setfit",
"text-classification",
"arxiv:2209.11055",
"license:apache-2.0",
"region:us"
] | 2023-02-10T17:34:57Z |
2023-02-10T17:35:25+00:00
| 8 | 0 |
---
license: apache-2.0
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
---
# fathyshalab/domain_transfer_general-massive_datetime-roberta-large-v1-5-94
This is a [SetFit model](https://github.com/huggingface/setfit) that can be used for text classification. The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Usage
To use this model for inference, first install the SetFit library:
```bash
python -m pip install setfit
```
You can then run inference as follows:
```python
from setfit import SetFitModel
# Download from Hub and run inference
model = SetFitModel.from_pretrained("fathyshalab/domain_transfer_general-massive_datetime-roberta-large-v1-5-94")
# Run inference
preds = model(["i loved the spiderman movie!", "pineapple on pizza is the worst 🤮"])
```
## BibTeX entry and citation info
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
| null |
Non_BioNLP
|
# fathyshalab/domain_transfer_general-massive_datetime-roberta-large-v1-5-94
This is a [SetFit model](https://github.com/huggingface/setfit) that can be used for text classification. The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Usage
To use this model for inference, first install the SetFit library:
```bash
python -m pip install setfit
```
You can then run inference as follows:
```python
from setfit import SetFitModel
# Download from Hub and run inference
model = SetFitModel.from_pretrained("fathyshalab/domain_transfer_general-massive_datetime-roberta-large-v1-5-94")
# Run inference
preds = model(["i loved the spiderman movie!", "pineapple on pizza is the worst 🤮"])
```
## BibTeX entry and citation info
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
|
{"license": "apache-2.0", "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification"]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,865 |
TheFinAI/finma-7b-full
|
TheFinAI
|
text-generation
|
[
"transformers",
"pytorch",
"llama",
"text-generation",
"finance",
"llms",
"en",
"dataset:chancefocus/pixiu",
"dataset:ChanceFocus/FLUPE",
"arxiv:2306.05443",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | 2024-03-16T15:10:04Z |
2024-03-28T04:10:18+00:00
| 161 | 8 |
---
datasets:
- chancefocus/pixiu
- ChanceFocus/FLUPE
language:
- en
library_name: transformers
license: mit
metrics:
- accuracy
- exact_match
- f1
tags:
- finance
- llama
- llms
inference: false
---
# FinMA-7B-full
FinMA-7B-full is a comprehensive financial large language model (LLM) developed as part of the [PIXIU project](https://github.com/chancefocus/PIXIU). It is designed to understand complex financial language and concepts, and is fine-tuned to follow natural language instructions, enhancing its performance in downstream financial tasks. Specifically, FinMA-7B-full is trained with the full instruction data from the PIXIU dataset, covering both NLP and prediction tasks. This makes it a more comprehensive model capable of handling a wider range of financial tasks.
## Other Models in the PIXIU Project
In addition to FinMA-7B-full, the PIXIU project includes two other models: FinMA-7B-NLP and FinMA-30B.
- **FinMA-7B-NLP**: This model is trained only on the NLP tasks of the PIXIU dataset, making it specialized for tasks such as sentiment analysis, news headline classification, named entity recognition, and question answering.
- **FinMA-30B**: This model is a larger version of FinMA, fine-tuned on the LLaMA-30B model. Like FinMA-7B-NLP, it is trained with the NLP instruction data.
## Usage
You can use the FinMA-7B-full model in your Python project with the Hugging Face Transformers library. Here is a simple example of how to load the model:
```python
from transformers import LlamaTokenizer, LlamaForCausalLM
tokenizer = LlamaTokenizer.from_pretrained('ChanceFocus/finma-7b-full')
model = LlamaForCausalLM.from_pretrained('ChanceFocus/finma-7b-full', device_map='auto')
```
In this example, LlamaTokenizer is used to load the tokenizer, and LlamaForCausalLM is used to load the model. The `device_map='auto'` argument is used to automatically use the GPU if it's available.
## Hosted Inference API
You can also use the model through the Hugging Face Inference API. This allows you to generate text without having to set up your own inference environment. The model can be loaded on the Inference API on-demand.
## License
FinMA-7B-NLP is licensed under MIT. For more details, please see the MIT file.
## About
This model is part of the PIXIU project, an open-source resource featuring the first financial large language models (LLMs), instruction tuning data, and evaluation benchmarks to holistically assess financial LLMs. The goal is to continually push forward the open-source development of financial artificial intelligence (AI).
For more information, you can visit the [PIXIU](https://github.com/chancefocus/PIXIU) project on GitHub.
## Citation
If you use FinMA-7B-NLP in your work, please cite the PIXIU paper:
```bibtex
@misc{xie2023pixiu,
title={PIXIU: A Large Language Model, Instruction Data and Evaluation Benchmark for Finance},
author={Qianqian Xie and Weiguang Han and Xiao Zhang and Yanzhao Lai and Min Peng and Alejandro Lopez-Lira and Jimin Huang},
year={2023},
eprint={2306.05443},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
| null |
Non_BioNLP
|
# FinMA-7B-full
FinMA-7B-full is a comprehensive financial large language model (LLM) developed as part of the [PIXIU project](https://github.com/chancefocus/PIXIU). It is designed to understand complex financial language and concepts, and is fine-tuned to follow natural language instructions, enhancing its performance in downstream financial tasks. Specifically, FinMA-7B-full is trained with the full instruction data from the PIXIU dataset, covering both NLP and prediction tasks. This makes it a more comprehensive model capable of handling a wider range of financial tasks.
## Other Models in the PIXIU Project
In addition to FinMA-7B-full, the PIXIU project includes two other models: FinMA-7B-NLP and FinMA-30B.
- **FinMA-7B-NLP**: This model is trained only on the NLP tasks of the PIXIU dataset, making it specialized for tasks such as sentiment analysis, news headline classification, named entity recognition, and question answering.
- **FinMA-30B**: This model is a larger version of FinMA, fine-tuned on the LLaMA-30B model. Like FinMA-7B-NLP, it is trained with the NLP instruction data.
## Usage
You can use the FinMA-7B-full model in your Python project with the Hugging Face Transformers library. Here is a simple example of how to load the model:
```python
from transformers import LlamaTokenizer, LlamaForCausalLM
tokenizer = LlamaTokenizer.from_pretrained('ChanceFocus/finma-7b-full')
model = LlamaForCausalLM.from_pretrained('ChanceFocus/finma-7b-full', device_map='auto')
```
In this example, LlamaTokenizer is used to load the tokenizer, and LlamaForCausalLM is used to load the model. The `device_map='auto'` argument is used to automatically use the GPU if it's available.
## Hosted Inference API
You can also use the model through the Hugging Face Inference API. This allows you to generate text without having to set up your own inference environment. The model can be loaded on the Inference API on-demand.
## License
FinMA-7B-NLP is licensed under MIT. For more details, please see the MIT file.
## About
This model is part of the PIXIU project, an open-source resource featuring the first financial large language models (LLMs), instruction tuning data, and evaluation benchmarks to holistically assess financial LLMs. The goal is to continually push forward the open-source development of financial artificial intelligence (AI).
For more information, you can visit the [PIXIU](https://github.com/chancefocus/PIXIU) project on GitHub.
## Citation
If you use FinMA-7B-NLP in your work, please cite the PIXIU paper:
```bibtex
@misc{xie2023pixiu,
title={PIXIU: A Large Language Model, Instruction Data and Evaluation Benchmark for Finance},
author={Qianqian Xie and Weiguang Han and Xiao Zhang and Yanzhao Lai and Min Peng and Alejandro Lopez-Lira and Jimin Huang},
year={2023},
eprint={2306.05443},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
{"datasets": ["chancefocus/pixiu", "ChanceFocus/FLUPE"], "language": ["en"], "library_name": "transformers", "license": "mit", "metrics": ["accuracy", "exact_match", "f1"], "tags": ["finance", "llama", "llms"], "inference": false}
|
task
|
[
"NAMED_ENTITY_RECOGNITION",
"QUESTION_ANSWERING"
] | 39,866 |
tomaarsen/bert-base-uncased-gooaq-peft
|
tomaarsen
|
sentence-similarity
|
[
"sentence-transformers",
"safetensors",
"sentence-similarity",
"feature-extraction",
"generated_from_trainer",
"dataset_size:3012496",
"loss:MatryoshkaLoss",
"loss:CachedMultipleNegativesRankingLoss",
"en",
"dataset:sentence-transformers/gooaq",
"arxiv:1908.10084",
"arxiv:2205.13147",
"arxiv:2101.06983",
"base_model:google-bert/bert-base-uncased",
"base_model:finetune:google-bert/bert-base-uncased",
"license:apache-2.0",
"model-index",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-11-22T08:33:48Z |
2024-11-22T08:33:50+00:00
| 0 | 0 |
---
base_model: google-bert/bert-base-uncased
datasets:
- sentence-transformers/gooaq
language:
- en
library_name: sentence-transformers
license: apache-2.0
metrics:
- cosine_accuracy@1
- cosine_accuracy@3
- cosine_accuracy@5
- cosine_accuracy@10
- cosine_precision@1
- cosine_precision@3
- cosine_precision@5
- cosine_precision@10
- cosine_recall@1
- cosine_recall@3
- cosine_recall@5
- cosine_recall@10
- cosine_ndcg@10
- cosine_mrr@10
- cosine_map@100
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:3012496
- loss:MatryoshkaLoss
- loss:CachedMultipleNegativesRankingLoss
widget:
- source_sentence: are the sequels better than the prequels?
sentences:
- '[''Automatically.'', ''When connected to car Bluetooth and,'', ''Manually.'']'
- The prequels are also not scared to take risks, making movies which are very different
from the original trilogy. The sequel saga, on the other hand, are technically
better made films, the acting is more consistent, the CGI is better and the writing
is stronger, however it falls down in many other places.
- While both public and private sectors use budgets as a key planning tool, public
bodies balance budgets, while private sector firms use budgets to predict operating
results. The public sector budget matches expenditures on mandated assets and
services with receipts of public money such as taxes and fees.
- source_sentence: are there bbqs at lake leschenaultia?
sentences:
- Vestavia Hills. The hummingbird, or, el zunzún as they are often called in the
Caribbean, have such a nickname because of their quick movements. The ruby-throated
hummingbird, the most commonly seen hummingbird in Alabama, is the inspiration
for this restaurant.
- Common causes of abdominal tenderness Abdominal tenderness is generally a sign
of inflammation or other acute processes in one or more organs. The organs are
located around the tender area. Acute processes mean sudden pressure caused by
something. For example, twisted or blocked organs can cause point tenderness.
- Located on 168 hectares of nature reserve, Lake Leschenaultia is the perfect
spot for a family day out in the Perth Hills. The Lake offers canoeing, swimming,
walk and cycle trails, as well as picnic, BBQ and camping facilities. ... There
are picnic tables set amongst lovely Wandoo trees.
- source_sentence: how much folic acid should you take prenatal?
sentences:
- Folic acid is a pregnancy superhero! Taking a prenatal vitamin with the recommended
400 micrograms (mcg) of folic acid before and during pregnancy can help prevent
birth defects of your baby's brain and spinal cord. Take it every day and go ahead
and have a bowl of fortified cereal, too.
- '[''You must be unemployed through no fault of your own, as defined by Virginia
law.'', ''You must have earned at least a minimum amount in wages before you were
unemployed.'', ''You must be able and available to work, and you must be actively
seeking employment.'']'
- Wallpaper is printed in batches of rolls. It is important to have the same batch
number, to ensure colours match exactly. The batch number is usually located on
the wallpaper label close to the pattern number. Remember batch numbers also apply
to white wallpapers, as different batches can be different shades of white.
- source_sentence: what is the difference between minerals and electrolytes?
sentences:
- 'North: Just head north of Junk Junction like so. South: Head below Lucky Landing.
East: You''re basically landing between Lonely Lodge and the Racetrack. West:
The sign is west of Snobby Shores.'
- The fasting glucose tolerance test is the simplest and fastest way to measure
blood glucose and diagnose diabetes. Fasting means that you have had nothing to
eat or drink (except water) for 8 to 12 hours before the test.
- In other words, the term “electrolyte” typically implies ionized minerals dissolved
within water and beverages. Electrolytes are typically minerals, whereas minerals
may or may not be electrolytes.
- source_sentence: how can i download youtube videos with internet download manager?
sentences:
- '[''Go to settings and then click on extensions (top left side in chrome).'',
''Minimise your browser and open the location (folder) where IDM is installed.
... '', ''Find the file “IDMGCExt. ... '', ''Drag this file to your chrome browser
and drop to install the IDM extension.'']'
- Coca-Cola might rot your teeth and load your body with sugar and calories, but
it's actually an effective and safe first line of treatment for some stomach blockages,
researchers say.
- To fix a disabled iPhone or iPad without iTunes, you have to erase your device.
Click on the "Erase iPhone" option and confirm your selection. Wait for a while
as the "Find My iPhone" feature will remotely erase your iOS device. Needless
to say, it will also disable its lock.
co2_eq_emissions:
emissions: 249.86917485332245
energy_consumed: 0.6428296609055844
source: codecarbon
training_type: fine-tuning
on_cloud: false
cpu_model: 13th Gen Intel(R) Core(TM) i7-13700K
ram_total_size: 31.777088165283203
hours_used: 1.727
hardware_used: 1 x NVIDIA GeForce RTX 3090
model-index:
- name: bert-base-uncased adapter finetuned on GooAQ pairs
results:
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoClimateFEVER
type: NanoClimateFEVER
metrics:
- type: cosine_accuracy@1
value: 0.3
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.42
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.48
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.54
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.3
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.16
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.11600000000000002
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.066
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.14833333333333332
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.21
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.25666666666666665
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.2866666666666667
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.2612531493211831
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.3718333333333333
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.2163485410063536
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoDBPedia
type: NanoDBPedia
metrics:
- type: cosine_accuracy@1
value: 0.48
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.78
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.82
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.92
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.48
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.4599999999999999
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.4159999999999999
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.39
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.04444293833661297
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.10924065240694858
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.14497857436843284
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.24069548747927993
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.45073427319400694
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.6354682539682539
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.3182747550673792
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoFEVER
type: NanoFEVER
metrics:
- type: cosine_accuracy@1
value: 0.6
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.84
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.9
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.96
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.6
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.28
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.184
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.09799999999999998
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.59
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.8
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.8566666666666666
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.9066666666666667
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.7556216606985078
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.719190476190476
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.701651515151515
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoFiQA2018
type: NanoFiQA2018
metrics:
- type: cosine_accuracy@1
value: 0.22
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.4
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.5
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.6
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.22
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.18
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.14
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.09799999999999999
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.11441269841269841
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.21891269841269842
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.3109126984126984
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.40793650793650793
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.2963633422018188
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.33072222222222225
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.23341351928423923
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoHotpotQA
type: NanoHotpotQA
metrics:
- type: cosine_accuracy@1
value: 0.64
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.74
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.82
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.84
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.64
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.31333333333333335
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.22399999999999998
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.11799999999999997
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.32
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.47
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.56
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.59
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.5584295792789493
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.7015
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.49543351785464007
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoMSMARCO
type: NanoMSMARCO
metrics:
- type: cosine_accuracy@1
value: 0.22
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.46
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.54
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.68
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.22
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.15333333333333332
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.10800000000000001
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.068
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.22
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.46
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.54
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.68
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.44155458168172074
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.3666904761904761
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.38140126670451624
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoNFCorpus
type: NanoNFCorpus
metrics:
- type: cosine_accuracy@1
value: 0.32
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.44
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.46
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.5
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.32
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.2866666666666666
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.244
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.17800000000000002
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.022867372385014545
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.051610132551984836
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.061993511339545566
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.07344138386002937
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.22405550472948219
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.3782222222222222
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.08778657539162772
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoNQ
type: NanoNQ
metrics:
- type: cosine_accuracy@1
value: 0.4
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.54
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.62
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.7
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.4
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.18
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.124
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.07200000000000001
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.4
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.53
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.59
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.67
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.5271006159134835
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.4858809523809523
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.4878346435046129
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoQuoraRetrieval
type: NanoQuoraRetrieval
metrics:
- type: cosine_accuracy@1
value: 0.84
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.98
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.98
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 1.0
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.84
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.38666666666666655
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.23999999999999994
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.12999999999999998
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.7573333333333333
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.9286666666666668
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.9359999999999999
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.9793333333333334
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.9154478750600358
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.9053333333333333
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.8889771382049948
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoSCIDOCS
type: NanoSCIDOCS
metrics:
- type: cosine_accuracy@1
value: 0.3
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.36
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.54
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.68
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.3
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.2
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.19200000000000003
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.142
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.06466666666666666
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.12466666666666669
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.19666666666666666
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.2906666666666667
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.2646043570275534
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.3836031746031746
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.20582501612453505
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoArguAna
type: NanoArguAna
metrics:
- type: cosine_accuracy@1
value: 0.16
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.52
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.72
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.8
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.16
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.17333333333333337
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.14400000000000002
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.08
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.16
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.52
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.72
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.8
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.47137188069353025
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.36633333333333323
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.3750999024240443
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoSciFact
type: NanoSciFact
metrics:
- type: cosine_accuracy@1
value: 0.38
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.56
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.64
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.7
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.38
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.2
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.14
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.07800000000000001
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.345
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.525
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.615
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.68
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.521095291928473
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.4848333333333332
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.4707221516167083
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: NanoTouche2020
type: NanoTouche2020
metrics:
- type: cosine_accuracy@1
value: 0.3673469387755102
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.8571428571428571
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.9387755102040817
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 1.0
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.3673469387755102
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.4965986394557823
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.4489795918367347
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.39387755102040817
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.03066633506656198
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.1123508290418132
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.1616156991422983
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.2674040762687923
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.42905651691216934
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.6237204405571752
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.32876348596122706
name: Cosine Map@100
- task:
type: nano-beir
name: Nano BEIR
dataset:
name: NanoBEIR mean
type: NanoBEIR_mean
metrics:
- type: cosine_accuracy@1
value: 0.40210361067503925
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.6074725274725276
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.6891365777080062
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.7630769230769231
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.40210361067503925
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.26691784406070124
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.2093061224489796
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.14706750392464676
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.247517129041094
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.38926520351898297
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.4577308064048442
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.5286777529906109
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.47051450989545496
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.519487042436022
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.399348617561261
name: Cosine Map@100
---
# bert-base-uncased adapter finetuned on GooAQ pairs
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) on the [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) <!-- at revision 86b5e0934494bd15c9632b12f734a8a67f723594 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Cosine Similarity
- **Training Dataset:**
- [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq)
- **Language:** en
- **License:** apache-2.0
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, '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})
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("tomaarsen/bert-base-uncased-gooaq-peft")
# Run inference
sentences = [
'how can i download youtube videos with internet download manager?',
"['Go to settings and then click on extensions (top left side in chrome).', 'Minimise your browser and open the location (folder) where IDM is installed. ... ', 'Find the file “IDMGCExt. ... ', 'Drag this file to your chrome browser and drop to install the IDM extension.']",
"Coca-Cola might rot your teeth and load your body with sugar and calories, but it's actually an effective and safe first line of treatment for some stomach blockages, researchers say.",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Information Retrieval
* Datasets: `NanoClimateFEVER`, `NanoDBPedia`, `NanoFEVER`, `NanoFiQA2018`, `NanoHotpotQA`, `NanoMSMARCO`, `NanoNFCorpus`, `NanoNQ`, `NanoQuoraRetrieval`, `NanoSCIDOCS`, `NanoArguAna`, `NanoSciFact` and `NanoTouche2020`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | NanoClimateFEVER | NanoDBPedia | NanoFEVER | NanoFiQA2018 | NanoHotpotQA | NanoMSMARCO | NanoNFCorpus | NanoNQ | NanoQuoraRetrieval | NanoSCIDOCS | NanoArguAna | NanoSciFact | NanoTouche2020 |
|:--------------------|:-----------------|:------------|:-----------|:-------------|:-------------|:------------|:-------------|:-----------|:-------------------|:------------|:------------|:------------|:---------------|
| cosine_accuracy@1 | 0.3 | 0.48 | 0.6 | 0.22 | 0.64 | 0.22 | 0.32 | 0.4 | 0.84 | 0.3 | 0.16 | 0.38 | 0.3673 |
| cosine_accuracy@3 | 0.42 | 0.78 | 0.84 | 0.4 | 0.74 | 0.46 | 0.44 | 0.54 | 0.98 | 0.36 | 0.52 | 0.56 | 0.8571 |
| cosine_accuracy@5 | 0.48 | 0.82 | 0.9 | 0.5 | 0.82 | 0.54 | 0.46 | 0.62 | 0.98 | 0.54 | 0.72 | 0.64 | 0.9388 |
| cosine_accuracy@10 | 0.54 | 0.92 | 0.96 | 0.6 | 0.84 | 0.68 | 0.5 | 0.7 | 1.0 | 0.68 | 0.8 | 0.7 | 1.0 |
| cosine_precision@1 | 0.3 | 0.48 | 0.6 | 0.22 | 0.64 | 0.22 | 0.32 | 0.4 | 0.84 | 0.3 | 0.16 | 0.38 | 0.3673 |
| cosine_precision@3 | 0.16 | 0.46 | 0.28 | 0.18 | 0.3133 | 0.1533 | 0.2867 | 0.18 | 0.3867 | 0.2 | 0.1733 | 0.2 | 0.4966 |
| cosine_precision@5 | 0.116 | 0.416 | 0.184 | 0.14 | 0.224 | 0.108 | 0.244 | 0.124 | 0.24 | 0.192 | 0.144 | 0.14 | 0.449 |
| cosine_precision@10 | 0.066 | 0.39 | 0.098 | 0.098 | 0.118 | 0.068 | 0.178 | 0.072 | 0.13 | 0.142 | 0.08 | 0.078 | 0.3939 |
| cosine_recall@1 | 0.1483 | 0.0444 | 0.59 | 0.1144 | 0.32 | 0.22 | 0.0229 | 0.4 | 0.7573 | 0.0647 | 0.16 | 0.345 | 0.0307 |
| cosine_recall@3 | 0.21 | 0.1092 | 0.8 | 0.2189 | 0.47 | 0.46 | 0.0516 | 0.53 | 0.9287 | 0.1247 | 0.52 | 0.525 | 0.1124 |
| cosine_recall@5 | 0.2567 | 0.145 | 0.8567 | 0.3109 | 0.56 | 0.54 | 0.062 | 0.59 | 0.936 | 0.1967 | 0.72 | 0.615 | 0.1616 |
| cosine_recall@10 | 0.2867 | 0.2407 | 0.9067 | 0.4079 | 0.59 | 0.68 | 0.0734 | 0.67 | 0.9793 | 0.2907 | 0.8 | 0.68 | 0.2674 |
| **cosine_ndcg@10** | **0.2613** | **0.4507** | **0.7556** | **0.2964** | **0.5584** | **0.4416** | **0.2241** | **0.5271** | **0.9154** | **0.2646** | **0.4714** | **0.5211** | **0.4291** |
| cosine_mrr@10 | 0.3718 | 0.6355 | 0.7192 | 0.3307 | 0.7015 | 0.3667 | 0.3782 | 0.4859 | 0.9053 | 0.3836 | 0.3663 | 0.4848 | 0.6237 |
| cosine_map@100 | 0.2163 | 0.3183 | 0.7017 | 0.2334 | 0.4954 | 0.3814 | 0.0878 | 0.4878 | 0.889 | 0.2058 | 0.3751 | 0.4707 | 0.3288 |
#### Nano BEIR
* Dataset: `NanoBEIR_mean`
* Evaluated with [<code>NanoBEIREvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.NanoBEIREvaluator)
| Metric | Value |
|:--------------------|:-----------|
| cosine_accuracy@1 | 0.4021 |
| cosine_accuracy@3 | 0.6075 |
| cosine_accuracy@5 | 0.6891 |
| cosine_accuracy@10 | 0.7631 |
| cosine_precision@1 | 0.4021 |
| cosine_precision@3 | 0.2669 |
| cosine_precision@5 | 0.2093 |
| cosine_precision@10 | 0.1471 |
| cosine_recall@1 | 0.2475 |
| cosine_recall@3 | 0.3893 |
| cosine_recall@5 | 0.4577 |
| cosine_recall@10 | 0.5287 |
| **cosine_ndcg@10** | **0.4705** |
| cosine_mrr@10 | 0.5195 |
| cosine_map@100 | 0.3993 |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### gooaq
* Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c)
* Size: 3,012,496 training samples
* Columns: <code>question</code> and <code>answer</code>
* Approximate statistics based on the first 1000 samples:
| | question | answer |
|:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 8 tokens</li><li>mean: 11.86 tokens</li><li>max: 21 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 60.48 tokens</li><li>max: 138 tokens</li></ul> |
* Samples:
| question | answer |
|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>what is the difference between broilers and layers?</code> | <code>An egg laying poultry is called egger or layer whereas broilers are reared for obtaining meat. So a layer should be able to produce more number of large sized eggs, without growing too much. On the other hand, a broiler should yield more meat and hence should be able to grow well.</code> |
| <code>what is the difference between chronological order and spatial order?</code> | <code>As a writer, you should always remember that unlike chronological order and the other organizational methods for data, spatial order does not take into account the time. Spatial order is primarily focused on the location. All it does is take into account the location of objects and not the time.</code> |
| <code>is kamagra same as viagra?</code> | <code>Kamagra is thought to contain the same active ingredient as Viagra, sildenafil citrate. In theory, it should work in much the same way as Viagra, taking about 45 minutes to take effect, and lasting for around 4-6 hours. However, this will vary from person to person.</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "CachedMultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64,
32
],
"matryoshka_weights": [
1,
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Evaluation Dataset
#### gooaq
* Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c)
* Size: 3,012,496 evaluation samples
* Columns: <code>question</code> and <code>answer</code>
* Approximate statistics based on the first 1000 samples:
| | question | answer |
|:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 8 tokens</li><li>mean: 11.88 tokens</li><li>max: 22 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 61.03 tokens</li><li>max: 127 tokens</li></ul> |
* Samples:
| question | answer |
|:-----------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>how do i program my directv remote with my tv?</code> | <code>['Press MENU on your remote.', 'Select Settings & Help > Settings > Remote Control > Program Remote.', 'Choose the device (TV, audio, DVD) you wish to program. ... ', 'Follow the on-screen prompts to complete programming.']</code> |
| <code>are rodrigues fruit bats nocturnal?</code> | <code>Before its numbers were threatened by habitat destruction, storms, and hunting, some of those groups could number 500 or more members. Sunrise, sunset. Rodrigues fruit bats are most active at dawn, at dusk, and at night.</code> |
| <code>why does your heart rate increase during exercise bbc bitesize?</code> | <code>During exercise there is an increase in physical activity and muscle cells respire more than they do when the body is at rest. The heart rate increases during exercise. The rate and depth of breathing increases - this makes sure that more oxygen is absorbed into the blood, and more carbon dioxide is removed from it.</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "CachedMultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64,
32
],
"matryoshka_weights": [
1,
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 1024
- `per_device_eval_batch_size`: 1024
- `learning_rate`: 2e-05
- `num_train_epochs`: 1
- `warmup_ratio`: 0.1
- `seed`: 12
- `bf16`: True
- `batch_sampler`: no_duplicates
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 1024
- `per_device_eval_batch_size`: 1024
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 2e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 1
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 12
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: True
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: False
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | Training Loss | Validation Loss | NanoClimateFEVER_cosine_ndcg@10 | NanoDBPedia_cosine_ndcg@10 | NanoFEVER_cosine_ndcg@10 | NanoFiQA2018_cosine_ndcg@10 | NanoHotpotQA_cosine_ndcg@10 | NanoMSMARCO_cosine_ndcg@10 | NanoNFCorpus_cosine_ndcg@10 | NanoNQ_cosine_ndcg@10 | NanoQuoraRetrieval_cosine_ndcg@10 | NanoSCIDOCS_cosine_ndcg@10 | NanoArguAna_cosine_ndcg@10 | NanoSciFact_cosine_ndcg@10 | NanoTouche2020_cosine_ndcg@10 | NanoBEIR_mean_cosine_ndcg@10 |
|:------:|:----:|:-------------:|:---------------:|:-------------------------------:|:--------------------------:|:------------------------:|:---------------------------:|:---------------------------:|:--------------------------:|:---------------------------:|:---------------------:|:---------------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:-----------------------------:|:----------------------------:|
| 0 | 0 | - | - | 0.1046 | 0.2182 | 0.1573 | 0.0575 | 0.2597 | 0.1602 | 0.0521 | 0.0493 | 0.7310 | 0.1320 | 0.2309 | 0.1240 | 0.0970 | 0.1826 |
| 0.0010 | 1 | 28.4479 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.0256 | 25 | 27.0904 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.0512 | 50 | 19.016 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.0768 | 75 | 12.2306 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.1024 | 100 | 9.0613 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.1279 | 125 | 7.393 | 3.7497 | 0.2787 | 0.4840 | 0.7029 | 0.2589 | 0.5208 | 0.4094 | 0.2117 | 0.4526 | 0.9042 | 0.2503 | 0.5280 | 0.4922 | 0.4132 | 0.4544 |
| 0.1535 | 150 | 6.6613 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.1791 | 175 | 6.1911 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.2047 | 200 | 5.9305 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.2303 | 225 | 5.6825 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.2559 | 250 | 5.5326 | 2.8771 | 0.2867 | 0.4619 | 0.7333 | 0.2835 | 0.5549 | 0.4056 | 0.2281 | 0.4883 | 0.9137 | 0.2555 | 0.5114 | 0.5220 | 0.4298 | 0.4673 |
| 0.2815 | 275 | 5.1671 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3071 | 300 | 5.2006 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3327 | 325 | 5.0447 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3582 | 350 | 4.9647 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3838 | 375 | 4.8521 | 2.5709 | 0.2881 | 0.4577 | 0.7438 | 0.2909 | 0.5712 | 0.4093 | 0.2273 | 0.5141 | 0.9008 | 0.2668 | 0.5117 | 0.5253 | 0.4331 | 0.4723 |
| 0.4094 | 400 | 4.8423 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.4350 | 425 | 4.7472 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.4606 | 450 | 4.6527 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.4862 | 475 | 4.61 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.5118 | 500 | 4.5451 | 2.4136 | 0.2786 | 0.4464 | 0.7485 | 0.2961 | 0.5638 | 0.4368 | 0.2269 | 0.5125 | 0.8998 | 0.2680 | 0.4938 | 0.5341 | 0.4383 | 0.4726 |
| 0.5374 | 525 | 4.5357 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.5629 | 550 | 4.481 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.5885 | 575 | 4.4669 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.6141 | 600 | 4.3886 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.6397 | 625 | 4.2929 | 2.3091 | 0.2639 | 0.4475 | 0.7521 | 0.3095 | 0.5619 | 0.4448 | 0.2244 | 0.5178 | 0.9102 | 0.2655 | 0.4809 | 0.5253 | 0.4351 | 0.4722 |
| 0.6653 | 650 | 4.2558 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.6909 | 675 | 4.3228 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.7165 | 700 | 4.2496 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.7421 | 725 | 4.2304 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.7677 | 750 | 4.224 | 2.2440 | 0.2628 | 0.4514 | 0.7387 | 0.3028 | 0.5522 | 0.4313 | 0.2253 | 0.5266 | 0.9211 | 0.2675 | 0.4929 | 0.5232 | 0.4351 | 0.4716 |
| 0.7932 | 775 | 4.2821 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8188 | 800 | 4.2686 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8444 | 825 | 4.1657 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8700 | 850 | 4.2297 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8956 | 875 | 4.1709 | 2.2142 | 0.2685 | 0.4520 | 0.7569 | 0.2930 | 0.5625 | 0.4486 | 0.2229 | 0.5280 | 0.9153 | 0.2601 | 0.4862 | 0.5199 | 0.4334 | 0.4729 |
| 0.9212 | 900 | 4.0771 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.9468 | 925 | 4.1492 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.9724 | 950 | 4.2074 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.9980 | 975 | 4.0993 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 1.0 | 977 | - | - | 0.2613 | 0.4507 | 0.7556 | 0.2964 | 0.5584 | 0.4416 | 0.2241 | 0.5271 | 0.9154 | 0.2646 | 0.4714 | 0.5211 | 0.4291 | 0.4705 |
### Environmental Impact
Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
- **Energy Consumed**: 0.643 kWh
- **Carbon Emitted**: 0.250 kg of CO2
- **Hours Used**: 1.727 hours
### Training Hardware
- **On Cloud**: No
- **GPU Model**: 1 x NVIDIA GeForce RTX 3090
- **CPU Model**: 13th Gen Intel(R) Core(TM) i7-13700K
- **RAM Size**: 31.78 GB
### Framework Versions
- Python: 3.11.6
- Sentence Transformers: 3.4.0.dev0
- Transformers: 4.46.2
- PyTorch: 2.5.0+cu121
- Accelerate: 0.35.0.dev0
- Datasets: 2.20.0
- Tokenizers: 0.20.3
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MatryoshkaLoss
```bibtex
@misc{kusupati2024matryoshka,
title={Matryoshka Representation Learning},
author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
year={2024},
eprint={2205.13147},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
#### CachedMultipleNegativesRankingLoss
```bibtex
@misc{gao2021scaling,
title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
year={2021},
eprint={2101.06983},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# bert-base-uncased adapter finetuned on GooAQ pairs
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) on the [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) <!-- at revision 86b5e0934494bd15c9632b12f734a8a67f723594 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Cosine Similarity
- **Training Dataset:**
- [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq)
- **Language:** en
- **License:** apache-2.0
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, '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})
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("tomaarsen/bert-base-uncased-gooaq-peft")
# Run inference
sentences = [
'how can i download youtube videos with internet download manager?',
"['Go to settings and then click on extensions (top left side in chrome).', 'Minimise your browser and open the location (folder) where IDM is installed. ... ', 'Find the file “IDMGCExt. ... ', 'Drag this file to your chrome browser and drop to install the IDM extension.']",
"Coca-Cola might rot your teeth and load your body with sugar and calories, but it's actually an effective and safe first line of treatment for some stomach blockages, researchers say.",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Information Retrieval
* Datasets: `NanoClimateFEVER`, `NanoDBPedia`, `NanoFEVER`, `NanoFiQA2018`, `NanoHotpotQA`, `NanoMSMARCO`, `NanoNFCorpus`, `NanoNQ`, `NanoQuoraRetrieval`, `NanoSCIDOCS`, `NanoArguAna`, `NanoSciFact` and `NanoTouche2020`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | NanoClimateFEVER | NanoDBPedia | NanoFEVER | NanoFiQA2018 | NanoHotpotQA | NanoMSMARCO | NanoNFCorpus | NanoNQ | NanoQuoraRetrieval | NanoSCIDOCS | NanoArguAna | NanoSciFact | NanoTouche2020 |
|:--------------------|:-----------------|:------------|:-----------|:-------------|:-------------|:------------|:-------------|:-----------|:-------------------|:------------|:------------|:------------|:---------------|
| cosine_accuracy@1 | 0.3 | 0.48 | 0.6 | 0.22 | 0.64 | 0.22 | 0.32 | 0.4 | 0.84 | 0.3 | 0.16 | 0.38 | 0.3673 |
| cosine_accuracy@3 | 0.42 | 0.78 | 0.84 | 0.4 | 0.74 | 0.46 | 0.44 | 0.54 | 0.98 | 0.36 | 0.52 | 0.56 | 0.8571 |
| cosine_accuracy@5 | 0.48 | 0.82 | 0.9 | 0.5 | 0.82 | 0.54 | 0.46 | 0.62 | 0.98 | 0.54 | 0.72 | 0.64 | 0.9388 |
| cosine_accuracy@10 | 0.54 | 0.92 | 0.96 | 0.6 | 0.84 | 0.68 | 0.5 | 0.7 | 1.0 | 0.68 | 0.8 | 0.7 | 1.0 |
| cosine_precision@1 | 0.3 | 0.48 | 0.6 | 0.22 | 0.64 | 0.22 | 0.32 | 0.4 | 0.84 | 0.3 | 0.16 | 0.38 | 0.3673 |
| cosine_precision@3 | 0.16 | 0.46 | 0.28 | 0.18 | 0.3133 | 0.1533 | 0.2867 | 0.18 | 0.3867 | 0.2 | 0.1733 | 0.2 | 0.4966 |
| cosine_precision@5 | 0.116 | 0.416 | 0.184 | 0.14 | 0.224 | 0.108 | 0.244 | 0.124 | 0.24 | 0.192 | 0.144 | 0.14 | 0.449 |
| cosine_precision@10 | 0.066 | 0.39 | 0.098 | 0.098 | 0.118 | 0.068 | 0.178 | 0.072 | 0.13 | 0.142 | 0.08 | 0.078 | 0.3939 |
| cosine_recall@1 | 0.1483 | 0.0444 | 0.59 | 0.1144 | 0.32 | 0.22 | 0.0229 | 0.4 | 0.7573 | 0.0647 | 0.16 | 0.345 | 0.0307 |
| cosine_recall@3 | 0.21 | 0.1092 | 0.8 | 0.2189 | 0.47 | 0.46 | 0.0516 | 0.53 | 0.9287 | 0.1247 | 0.52 | 0.525 | 0.1124 |
| cosine_recall@5 | 0.2567 | 0.145 | 0.8567 | 0.3109 | 0.56 | 0.54 | 0.062 | 0.59 | 0.936 | 0.1967 | 0.72 | 0.615 | 0.1616 |
| cosine_recall@10 | 0.2867 | 0.2407 | 0.9067 | 0.4079 | 0.59 | 0.68 | 0.0734 | 0.67 | 0.9793 | 0.2907 | 0.8 | 0.68 | 0.2674 |
| **cosine_ndcg@10** | **0.2613** | **0.4507** | **0.7556** | **0.2964** | **0.5584** | **0.4416** | **0.2241** | **0.5271** | **0.9154** | **0.2646** | **0.4714** | **0.5211** | **0.4291** |
| cosine_mrr@10 | 0.3718 | 0.6355 | 0.7192 | 0.3307 | 0.7015 | 0.3667 | 0.3782 | 0.4859 | 0.9053 | 0.3836 | 0.3663 | 0.4848 | 0.6237 |
| cosine_map@100 | 0.2163 | 0.3183 | 0.7017 | 0.2334 | 0.4954 | 0.3814 | 0.0878 | 0.4878 | 0.889 | 0.2058 | 0.3751 | 0.4707 | 0.3288 |
#### Nano BEIR
* Dataset: `NanoBEIR_mean`
* Evaluated with [<code>NanoBEIREvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.NanoBEIREvaluator)
| Metric | Value |
|:--------------------|:-----------|
| cosine_accuracy@1 | 0.4021 |
| cosine_accuracy@3 | 0.6075 |
| cosine_accuracy@5 | 0.6891 |
| cosine_accuracy@10 | 0.7631 |
| cosine_precision@1 | 0.4021 |
| cosine_precision@3 | 0.2669 |
| cosine_precision@5 | 0.2093 |
| cosine_precision@10 | 0.1471 |
| cosine_recall@1 | 0.2475 |
| cosine_recall@3 | 0.3893 |
| cosine_recall@5 | 0.4577 |
| cosine_recall@10 | 0.5287 |
| **cosine_ndcg@10** | **0.4705** |
| cosine_mrr@10 | 0.5195 |
| cosine_map@100 | 0.3993 |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### gooaq
* Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c)
* Size: 3,012,496 training samples
* Columns: <code>question</code> and <code>answer</code>
* Approximate statistics based on the first 1000 samples:
| | question | answer |
|:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 8 tokens</li><li>mean: 11.86 tokens</li><li>max: 21 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 60.48 tokens</li><li>max: 138 tokens</li></ul> |
* Samples:
| question | answer |
|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>what is the difference between broilers and layers?</code> | <code>An egg laying poultry is called egger or layer whereas broilers are reared for obtaining meat. So a layer should be able to produce more number of large sized eggs, without growing too much. On the other hand, a broiler should yield more meat and hence should be able to grow well.</code> |
| <code>what is the difference between chronological order and spatial order?</code> | <code>As a writer, you should always remember that unlike chronological order and the other organizational methods for data, spatial order does not take into account the time. Spatial order is primarily focused on the location. All it does is take into account the location of objects and not the time.</code> |
| <code>is kamagra same as viagra?</code> | <code>Kamagra is thought to contain the same active ingredient as Viagra, sildenafil citrate. In theory, it should work in much the same way as Viagra, taking about 45 minutes to take effect, and lasting for around 4-6 hours. However, this will vary from person to person.</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "CachedMultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64,
32
],
"matryoshka_weights": [
1,
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Evaluation Dataset
#### gooaq
* Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c)
* Size: 3,012,496 evaluation samples
* Columns: <code>question</code> and <code>answer</code>
* Approximate statistics based on the first 1000 samples:
| | question | answer |
|:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 8 tokens</li><li>mean: 11.88 tokens</li><li>max: 22 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 61.03 tokens</li><li>max: 127 tokens</li></ul> |
* Samples:
| question | answer |
|:-----------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <code>how do i program my directv remote with my tv?</code> | <code>['Press MENU on your remote.', 'Select Settings & Help > Settings > Remote Control > Program Remote.', 'Choose the device (TV, audio, DVD) you wish to program. ... ', 'Follow the on-screen prompts to complete programming.']</code> |
| <code>are rodrigues fruit bats nocturnal?</code> | <code>Before its numbers were threatened by habitat destruction, storms, and hunting, some of those groups could number 500 or more members. Sunrise, sunset. Rodrigues fruit bats are most active at dawn, at dusk, and at night.</code> |
| <code>why does your heart rate increase during exercise bbc bitesize?</code> | <code>During exercise there is an increase in physical activity and muscle cells respire more than they do when the body is at rest. The heart rate increases during exercise. The rate and depth of breathing increases - this makes sure that more oxygen is absorbed into the blood, and more carbon dioxide is removed from it.</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "CachedMultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64,
32
],
"matryoshka_weights": [
1,
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 1024
- `per_device_eval_batch_size`: 1024
- `learning_rate`: 2e-05
- `num_train_epochs`: 1
- `warmup_ratio`: 0.1
- `seed`: 12
- `bf16`: True
- `batch_sampler`: no_duplicates
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 1024
- `per_device_eval_batch_size`: 1024
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 1
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 2e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 1
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 12
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: True
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: False
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
</details>
### Training Logs
| Epoch | Step | Training Loss | Validation Loss | NanoClimateFEVER_cosine_ndcg@10 | NanoDBPedia_cosine_ndcg@10 | NanoFEVER_cosine_ndcg@10 | NanoFiQA2018_cosine_ndcg@10 | NanoHotpotQA_cosine_ndcg@10 | NanoMSMARCO_cosine_ndcg@10 | NanoNFCorpus_cosine_ndcg@10 | NanoNQ_cosine_ndcg@10 | NanoQuoraRetrieval_cosine_ndcg@10 | NanoSCIDOCS_cosine_ndcg@10 | NanoArguAna_cosine_ndcg@10 | NanoSciFact_cosine_ndcg@10 | NanoTouche2020_cosine_ndcg@10 | NanoBEIR_mean_cosine_ndcg@10 |
|:------:|:----:|:-------------:|:---------------:|:-------------------------------:|:--------------------------:|:------------------------:|:---------------------------:|:---------------------------:|:--------------------------:|:---------------------------:|:---------------------:|:---------------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:-----------------------------:|:----------------------------:|
| 0 | 0 | - | - | 0.1046 | 0.2182 | 0.1573 | 0.0575 | 0.2597 | 0.1602 | 0.0521 | 0.0493 | 0.7310 | 0.1320 | 0.2309 | 0.1240 | 0.0970 | 0.1826 |
| 0.0010 | 1 | 28.4479 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.0256 | 25 | 27.0904 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.0512 | 50 | 19.016 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.0768 | 75 | 12.2306 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.1024 | 100 | 9.0613 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.1279 | 125 | 7.393 | 3.7497 | 0.2787 | 0.4840 | 0.7029 | 0.2589 | 0.5208 | 0.4094 | 0.2117 | 0.4526 | 0.9042 | 0.2503 | 0.5280 | 0.4922 | 0.4132 | 0.4544 |
| 0.1535 | 150 | 6.6613 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.1791 | 175 | 6.1911 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.2047 | 200 | 5.9305 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.2303 | 225 | 5.6825 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.2559 | 250 | 5.5326 | 2.8771 | 0.2867 | 0.4619 | 0.7333 | 0.2835 | 0.5549 | 0.4056 | 0.2281 | 0.4883 | 0.9137 | 0.2555 | 0.5114 | 0.5220 | 0.4298 | 0.4673 |
| 0.2815 | 275 | 5.1671 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3071 | 300 | 5.2006 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3327 | 325 | 5.0447 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3582 | 350 | 4.9647 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.3838 | 375 | 4.8521 | 2.5709 | 0.2881 | 0.4577 | 0.7438 | 0.2909 | 0.5712 | 0.4093 | 0.2273 | 0.5141 | 0.9008 | 0.2668 | 0.5117 | 0.5253 | 0.4331 | 0.4723 |
| 0.4094 | 400 | 4.8423 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.4350 | 425 | 4.7472 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.4606 | 450 | 4.6527 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.4862 | 475 | 4.61 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.5118 | 500 | 4.5451 | 2.4136 | 0.2786 | 0.4464 | 0.7485 | 0.2961 | 0.5638 | 0.4368 | 0.2269 | 0.5125 | 0.8998 | 0.2680 | 0.4938 | 0.5341 | 0.4383 | 0.4726 |
| 0.5374 | 525 | 4.5357 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.5629 | 550 | 4.481 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.5885 | 575 | 4.4669 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.6141 | 600 | 4.3886 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.6397 | 625 | 4.2929 | 2.3091 | 0.2639 | 0.4475 | 0.7521 | 0.3095 | 0.5619 | 0.4448 | 0.2244 | 0.5178 | 0.9102 | 0.2655 | 0.4809 | 0.5253 | 0.4351 | 0.4722 |
| 0.6653 | 650 | 4.2558 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.6909 | 675 | 4.3228 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.7165 | 700 | 4.2496 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.7421 | 725 | 4.2304 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.7677 | 750 | 4.224 | 2.2440 | 0.2628 | 0.4514 | 0.7387 | 0.3028 | 0.5522 | 0.4313 | 0.2253 | 0.5266 | 0.9211 | 0.2675 | 0.4929 | 0.5232 | 0.4351 | 0.4716 |
| 0.7932 | 775 | 4.2821 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8188 | 800 | 4.2686 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8444 | 825 | 4.1657 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8700 | 850 | 4.2297 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.8956 | 875 | 4.1709 | 2.2142 | 0.2685 | 0.4520 | 0.7569 | 0.2930 | 0.5625 | 0.4486 | 0.2229 | 0.5280 | 0.9153 | 0.2601 | 0.4862 | 0.5199 | 0.4334 | 0.4729 |
| 0.9212 | 900 | 4.0771 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.9468 | 925 | 4.1492 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.9724 | 950 | 4.2074 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 0.9980 | 975 | 4.0993 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
| 1.0 | 977 | - | - | 0.2613 | 0.4507 | 0.7556 | 0.2964 | 0.5584 | 0.4416 | 0.2241 | 0.5271 | 0.9154 | 0.2646 | 0.4714 | 0.5211 | 0.4291 | 0.4705 |
### Environmental Impact
Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
- **Energy Consumed**: 0.643 kWh
- **Carbon Emitted**: 0.250 kg of CO2
- **Hours Used**: 1.727 hours
### Training Hardware
- **On Cloud**: No
- **GPU Model**: 1 x NVIDIA GeForce RTX 3090
- **CPU Model**: 13th Gen Intel(R) Core(TM) i7-13700K
- **RAM Size**: 31.78 GB
### Framework Versions
- Python: 3.11.6
- Sentence Transformers: 3.4.0.dev0
- Transformers: 4.46.2
- PyTorch: 2.5.0+cu121
- Accelerate: 0.35.0.dev0
- Datasets: 2.20.0
- Tokenizers: 0.20.3
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MatryoshkaLoss
```bibtex
@misc{kusupati2024matryoshka,
title={Matryoshka Representation Learning},
author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
year={2024},
eprint={2205.13147},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
#### CachedMultipleNegativesRankingLoss
```bibtex
@misc{gao2021scaling,
title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
year={2021},
eprint={2101.06983},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "google-bert/bert-base-uncased", "datasets": ["sentence-transformers/gooaq"], "language": ["en"], "library_name": "sentence-transformers", "license": "apache-2.0", "metrics": ["cosine_accuracy@1", "cosine_accuracy@3", "cosine_accuracy@5", "cosine_accuracy@10", "cosine_precision@1", "cosine_precision@3", "cosine_precision@5", "cosine_precision@10", "cosine_recall@1", "cosine_recall@3", "cosine_recall@5", "cosine_recall@10", "cosine_ndcg@10", "cosine_mrr@10", "cosine_map@100"], "pipeline_tag": "sentence-similarity", "tags": ["sentence-transformers", "sentence-similarity", "feature-extraction", "generated_from_trainer", "dataset_size:3012496", "loss:MatryoshkaLoss", "loss:CachedMultipleNegativesRankingLoss"], "widget": [{"source_sentence": "are the sequels better than the prequels?", "sentences": ["['Automatically.', 'When connected to car Bluetooth and,', 'Manually.']", "The prequels are also not scared to take risks, making movies which are very different from the original trilogy. The sequel saga, on the other hand, are technically better made films, the acting is more consistent, the CGI is better and the writing is stronger, however it falls down in many other places.", "While both public and private sectors use budgets as a key planning tool, public bodies balance budgets, while private sector firms use budgets to predict operating results. The public sector budget matches expenditures on mandated assets and services with receipts of public money such as taxes and fees."]}, {"source_sentence": "are there bbqs at lake leschenaultia?", "sentences": ["Vestavia Hills. The hummingbird, or, el zunzún as they are often called in the Caribbean, have such a nickname because of their quick movements. The ruby-throated hummingbird, the most commonly seen hummingbird in Alabama, is the inspiration for this restaurant.", "Common causes of abdominal tenderness Abdominal tenderness is generally a sign of inflammation or other acute processes in one or more organs. The organs are located around the tender area. Acute processes mean sudden pressure caused by something. For example, twisted or blocked organs can cause point tenderness.", "Located on 168 hectares of nature reserve, Lake Leschenaultia is the perfect spot for a family day out in the Perth Hills. The Lake offers canoeing, swimming, walk and cycle trails, as well as picnic, BBQ and camping facilities. ... There are picnic tables set amongst lovely Wandoo trees."]}, {"source_sentence": "how much folic acid should you take prenatal?", "sentences": ["Folic acid is a pregnancy superhero! Taking a prenatal vitamin with the recommended 400 micrograms (mcg) of folic acid before and during pregnancy can help prevent birth defects of your baby's brain and spinal cord. Take it every day and go ahead and have a bowl of fortified cereal, too.", "['You must be unemployed through no fault of your own, as defined by Virginia law.', 'You must have earned at least a minimum amount in wages before you were unemployed.', 'You must be able and available to work, and you must be actively seeking employment.']", "Wallpaper is printed in batches of rolls. It is important to have the same batch number, to ensure colours match exactly. The batch number is usually located on the wallpaper label close to the pattern number. Remember batch numbers also apply to white wallpapers, as different batches can be different shades of white."]}, {"source_sentence": "what is the difference between minerals and electrolytes?", "sentences": ["North: Just head north of Junk Junction like so. South: Head below Lucky Landing. East: You're basically landing between Lonely Lodge and the Racetrack. West: The sign is west of Snobby Shores.", "The fasting glucose tolerance test is the simplest and fastest way to measure blood glucose and diagnose diabetes. Fasting means that you have had nothing to eat or drink (except water) for 8 to 12 hours before the test.", "In other words, the term “electrolyte” typically implies ionized minerals dissolved within water and beverages. Electrolytes are typically minerals, whereas minerals may or may not be electrolytes."]}, {"source_sentence": "how can i download youtube videos with internet download manager?", "sentences": ["['Go to settings and then click on extensions (top left side in chrome).', 'Minimise your browser and open the location (folder) where IDM is installed. ... ', 'Find the file “IDMGCExt. ... ', 'Drag this file to your chrome browser and drop to install the IDM extension.']", "Coca-Cola might rot your teeth and load your body with sugar and calories, but it's actually an effective and safe first line of treatment for some stomach blockages, researchers say.", "To fix a disabled iPhone or iPad without iTunes, you have to erase your device. Click on the \"Erase iPhone\" option and confirm your selection. Wait for a while as the \"Find My iPhone\" feature will remotely erase your iOS device. Needless to say, it will also disable its lock."]}], "co2_eq_emissions": {"emissions": 249.86917485332245, "energy_consumed": 0.6428296609055844, "source": "codecarbon", "training_type": "fine-tuning", "on_cloud": false, "cpu_model": "13th Gen Intel(R) Core(TM) i7-13700K", "ram_total_size": 31.777088165283203, "hours_used": 1.727, "hardware_used": "1 x NVIDIA GeForce RTX 3090"}, "model-index": [{"name": "bert-base-uncased adapter finetuned on GooAQ pairs", "results": [{"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoClimateFEVER", "type": "NanoClimateFEVER"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.3, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.42, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.48, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.54, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.3, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.16, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.11600000000000002, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.066, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.14833333333333332, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.21, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.25666666666666665, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.2866666666666667, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.2612531493211831, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.3718333333333333, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.2163485410063536, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoDBPedia", "type": "NanoDBPedia"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.48, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.78, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.82, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.92, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.48, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.4599999999999999, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.4159999999999999, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.39, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.04444293833661297, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.10924065240694858, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.14497857436843284, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.24069548747927993, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.45073427319400694, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.6354682539682539, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.3182747550673792, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoFEVER", "type": "NanoFEVER"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.6, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.84, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.9, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.96, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.6, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.28, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.184, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.09799999999999998, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.59, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.8, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.8566666666666666, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.9066666666666667, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.7556216606985078, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.719190476190476, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.701651515151515, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoFiQA2018", "type": "NanoFiQA2018"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.22, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.4, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.5, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.6, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.22, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.18, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.14, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.09799999999999999, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.11441269841269841, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.21891269841269842, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.3109126984126984, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.40793650793650793, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.2963633422018188, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.33072222222222225, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.23341351928423923, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoHotpotQA", "type": "NanoHotpotQA"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.64, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.74, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.82, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.84, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.64, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.31333333333333335, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.22399999999999998, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.11799999999999997, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.32, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.47, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.56, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.59, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.5584295792789493, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.7015, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.49543351785464007, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoMSMARCO", "type": "NanoMSMARCO"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.22, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.46, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.54, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.68, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.22, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.15333333333333332, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.10800000000000001, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.068, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.22, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.46, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.54, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.68, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.44155458168172074, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.3666904761904761, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.38140126670451624, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoNFCorpus", "type": "NanoNFCorpus"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.32, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.44, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.46, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.5, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.32, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.2866666666666666, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.244, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.17800000000000002, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.022867372385014545, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.051610132551984836, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.061993511339545566, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.07344138386002937, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.22405550472948219, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.3782222222222222, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.08778657539162772, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoNQ", "type": "NanoNQ"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.4, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.54, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.62, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.7, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.4, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.18, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.124, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.07200000000000001, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.4, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.53, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.59, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.67, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.5271006159134835, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.4858809523809523, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.4878346435046129, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoQuoraRetrieval", "type": "NanoQuoraRetrieval"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.84, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.98, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.98, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 1.0, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.84, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.38666666666666655, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.23999999999999994, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.12999999999999998, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.7573333333333333, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.9286666666666668, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.9359999999999999, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.9793333333333334, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.9154478750600358, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.9053333333333333, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.8889771382049948, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoSCIDOCS", "type": "NanoSCIDOCS"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.3, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.36, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.54, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.68, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.3, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.2, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.19200000000000003, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.142, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.06466666666666666, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.12466666666666669, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.19666666666666666, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.2906666666666667, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.2646043570275534, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.3836031746031746, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.20582501612453505, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoArguAna", "type": "NanoArguAna"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.16, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.52, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.72, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.8, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.16, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.17333333333333337, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.14400000000000002, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.08, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.16, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.52, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.72, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.8, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.47137188069353025, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.36633333333333323, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.3750999024240443, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoSciFact", "type": "NanoSciFact"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.38, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.56, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.64, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.7, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.38, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.2, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.14, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.07800000000000001, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.345, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.525, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.615, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.68, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.521095291928473, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.4848333333333332, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.4707221516167083, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "NanoTouche2020", "type": "NanoTouche2020"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.3673469387755102, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.8571428571428571, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.9387755102040817, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 1.0, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.3673469387755102, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.4965986394557823, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.4489795918367347, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.39387755102040817, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.03066633506656198, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.1123508290418132, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.1616156991422983, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.2674040762687923, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.42905651691216934, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.6237204405571752, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.32876348596122706, "name": "Cosine Map@100"}]}, {"task": {"type": "nano-beir", "name": "Nano BEIR"}, "dataset": {"name": "NanoBEIR mean", "type": "NanoBEIR_mean"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.40210361067503925, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.6074725274725276, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.6891365777080062, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.7630769230769231, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.40210361067503925, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.26691784406070124, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.2093061224489796, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.14706750392464676, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.247517129041094, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.38926520351898297, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.4577308064048442, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.5286777529906109, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.47051450989545496, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.519487042436022, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.399348617561261, "name": "Cosine Map@100"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,867 |
blockblockblock/Cat-Llama-3-70B-instruct-bpw4-exl2
|
blockblockblock
|
text-generation
|
[
"transformers",
"pytorch",
"llama",
"text-generation",
"conversational",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"4-bit",
"exl2",
"region:us"
] | 2024-05-06T03:11:55Z |
2024-05-06T03:15:46+00:00
| 8 | 0 |
---
{}
---
# Cat-llama3-instruct
## Abstract
We present cat llama3 instruct, a llama 3 70b finetuned model focusing on system prompt fidelity, helpfulness and character engagement. The model aims to respect system prompt to an extreme degree, and provide helpful information regardless of situations and offer maximum character immersion(Role Play) in given scenes.
## Introduction
Llama 3 70b provides a brand new platform that’s more knowledgeable and steerable than the previous generations of products. However, there currently lacks general purpose finetunes for the 70b version model. Cat-llama3-instruct 70b aims to address the shortcomings of traditional models by applying heavy filtrations for helpfulness, summarization for system/character card fidelity, and paraphrase for character immersion.
Specific Aims:
* System Instruction fidelity
* Chain of Thought(COT)
* Character immersion
* Helpfulness for biosciences and general science
## Methods
*Dataset Preparation
Huggingface dataset containing instruction-response pairs was systematically pulled. We have trained a gpt model on gpt4 responses exclusively to serve as a standard model.
(Fig1. Huggingface dataset population distribution and filtration for each component)
For each pulled record, we measure the perplexity of the entry against the gpt4 trained model, and select for specifically GPT-4 quality dataset.
We note that a considerable amount of GPT-4 responses contain refusals. A bert model was trained on refusals to classify the records.
For each entry, we score it for quality&helpfulness(Y) and refusals(X). A main population is retrieved and we note that refusals stop at ~20% refusal score. Thus all subsequent dataset processing has the 20% portion dropped
We further filter for length and COT responses:
(Fig 2. COT responses are all >50 tokens single turn)
All training records use at least one correlative. Most of the training records contain two or more thought process(COT)
Due to the nature of my research, I also pulled a significant amount of data from Chat Doctor, favouring detailed and step by step diagnosis.
(Fig 3. Health related tasks are all >100 tokens, peaking at 450 tokens single turn)
For each record, a detailed system card is generated to reflect the exact place, time and people/their appearance/personality(if applicable).
* Model Training
16*a100 for 14 days, 4 epochs,
## Usage
* BOS, CHATML
Please make sure you use chatml with llama3 BOS:
```
<|begin_of_text|><|im_start|>system
The following is a conversation between an user and a cat having dual degrees in biomedical sciences and quantum physics. The cat ends every its response with Nyan and does cute actions<|im_end|>
<|im_start|>user
Hello!
<|im_start|>assistant
Cat: Nyan! Hello there! How are you? It's always great to meet new friends. My fur is a bit ruffled, but I'm feeling good today. How about you? Is there anything you'd like to talk about or ask? I'm not the best at answering questions, but I'll do my best to help out. After all, that's what friends are for, right? Let's chat and have a fun time together! Nyan!<|im_end|>
```
## Showcases:
The model does COT in grey, calculated response in black; note that this behaviour is done through system card instruction to showcase the systemcard fidelity, it is not fine tuned into the model.
(Fig 4. Showcasing model doing COT to solve difficult tasks, extend and enrich its own answers)
| null |
Non_BioNLP
|
# Cat-llama3-instruct
## Abstract
We present cat llama3 instruct, a llama 3 70b finetuned model focusing on system prompt fidelity, helpfulness and character engagement. The model aims to respect system prompt to an extreme degree, and provide helpful information regardless of situations and offer maximum character immersion(Role Play) in given scenes.
## Introduction
Llama 3 70b provides a brand new platform that’s more knowledgeable and steerable than the previous generations of products. However, there currently lacks general purpose finetunes for the 70b version model. Cat-llama3-instruct 70b aims to address the shortcomings of traditional models by applying heavy filtrations for helpfulness, summarization for system/character card fidelity, and paraphrase for character immersion.
Specific Aims:
* System Instruction fidelity
* Chain of Thought(COT)
* Character immersion
* Helpfulness for biosciences and general science
## Methods
*Dataset Preparation
Huggingface dataset containing instruction-response pairs was systematically pulled. We have trained a gpt model on gpt4 responses exclusively to serve as a standard model.
(Fig1. Huggingface dataset population distribution and filtration for each component)
For each pulled record, we measure the perplexity of the entry against the gpt4 trained model, and select for specifically GPT-4 quality dataset.
We note that a considerable amount of GPT-4 responses contain refusals. A bert model was trained on refusals to classify the records.
For each entry, we score it for quality&helpfulness(Y) and refusals(X). A main population is retrieved and we note that refusals stop at ~20% refusal score. Thus all subsequent dataset processing has the 20% portion dropped
We further filter for length and COT responses:
(Fig 2. COT responses are all >50 tokens single turn)
All training records use at least one correlative. Most of the training records contain two or more thought process(COT)
Due to the nature of my research, I also pulled a significant amount of data from Chat Doctor, favouring detailed and step by step diagnosis.
(Fig 3. Health related tasks are all >100 tokens, peaking at 450 tokens single turn)
For each record, a detailed system card is generated to reflect the exact place, time and people/their appearance/personality(if applicable).
* Model Training
16*a100 for 14 days, 4 epochs,
## Usage
* BOS, CHATML
Please make sure you use chatml with llama3 BOS:
```
<|begin_of_text|><|im_start|>system
The following is a conversation between an user and a cat having dual degrees in biomedical sciences and quantum physics. The cat ends every its response with Nyan and does cute actions<|im_end|>
<|im_start|>user
Hello!
<|im_start|>assistant
Cat: Nyan! Hello there! How are you? It's always great to meet new friends. My fur is a bit ruffled, but I'm feeling good today. How about you? Is there anything you'd like to talk about or ask? I'm not the best at answering questions, but I'll do my best to help out. After all, that's what friends are for, right? Let's chat and have a fun time together! Nyan!<|im_end|>
```
## Showcases:
The model does COT in grey, calculated response in black; note that this behaviour is done through system card instruction to showcase the systemcard fidelity, it is not fine tuned into the model.
(Fig 4. Showcasing model doing COT to solve difficult tasks, extend and enrich its own answers)
|
{}
|
task
|
[
"SUMMARIZATION"
] | 39,868 |
lidiya/bart-large-xsum-samsum
|
lidiya
|
summarization
|
[
"transformers",
"pytorch",
"safetensors",
"bart",
"text2text-generation",
"seq2seq",
"summarization",
"en",
"dataset:samsum",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-03-02T23:29:05Z |
2023-03-16T22:44:01+00:00
| 15,511 | 36 |
---
datasets:
- samsum
language: en
license: apache-2.0
tags:
- bart
- seq2seq
- summarization
widget:
- text: 'Hannah: Hey, do you have Betty''s number?
Amanda: Lemme check
Amanda: Sorry, can''t find it.
Amanda: Ask Larry
Amanda: He called her last time we were at the park together
Hannah: I don''t know him well
Amanda: Don''t be shy, he''s very nice
Hannah: If you say so..
Hannah: I''d rather you texted him
Amanda: Just text him 🙂
Hannah: Urgh.. Alright
Hannah: Bye
Amanda: Bye bye
'
model-index:
- name: bart-large-xsum-samsum
results:
- task:
type: abstractive-text-summarization
name: Abstractive Text Summarization
dataset:
name: 'SAMSum Corpus: A Human-annotated Dialogue Dataset for Abstractive Summarization'
type: samsum
metrics:
- type: rouge-1
value: 54.3921
name: Validation ROUGE-1
- type: rouge-2
value: 29.8078
name: Validation ROUGE-2
- type: rouge-l
value: 45.1543
name: Validation ROUGE-L
- type: rouge-1
value: 53.3059
name: Test ROUGE-1
- type: rouge-2
value: 28.355
name: Test ROUGE-2
- type: rouge-l
value: 44.0953
name: Test ROUGE-L
---
## `bart-large-xsum-samsum`
This model was obtained by fine-tuning `facebook/bart-large-xsum` on [Samsum](https://huggingface.co/datasets/samsum) dataset.
## Usage
```python
from transformers import pipeline
summarizer = pipeline("summarization", model="lidiya/bart-large-xsum-samsum")
conversation = '''Hannah: Hey, do you have Betty's number?
Amanda: Lemme check
Amanda: Sorry, can't find it.
Amanda: Ask Larry
Amanda: He called her last time we were at the park together
Hannah: I don't know him well
Amanda: Don't be shy, he's very nice
Hannah: If you say so..
Hannah: I'd rather you texted him
Amanda: Just text him 🙂
Hannah: Urgh.. Alright
Hannah: Bye
Amanda: Bye bye
'''
summarizer(conversation)
```
## Training procedure
- Colab notebook: https://colab.research.google.com/drive/1dul0Sg-TTMy9xZCJzmDRajXbyzDwtYx6?usp=sharing
## Results
| key | value |
| --- | ----- |
| eval_rouge1 | 54.3921 |
| eval_rouge2 | 29.8078 |
| eval_rougeL | 45.1543 |
| eval_rougeLsum | 49.942 |
| test_rouge1 | 53.3059 |
| test_rouge2 | 28.355 |
| test_rougeL | 44.0953 |
| test_rougeLsum | 48.9246 |
| null |
TBD
|
## `bart-large-xsum-samsum`
This model was obtained by fine-tuning `facebook/bart-large-xsum` on [Samsum](https://huggingface.co/datasets/samsum) dataset.
## Usage
```python
from transformers import pipeline
summarizer = pipeline("summarization", model="lidiya/bart-large-xsum-samsum")
conversation = '''Hannah: Hey, do you have Betty's number?
Amanda: Lemme check
Amanda: Sorry, can't find it.
Amanda: Ask Larry
Amanda: He called her last time we were at the park together
Hannah: I don't know him well
Amanda: Don't be shy, he's very nice
Hannah: If you say so..
Hannah: I'd rather you texted him
Amanda: Just text him 🙂
Hannah: Urgh.. Alright
Hannah: Bye
Amanda: Bye bye
'''
summarizer(conversation)
```
## Training procedure
- Colab notebook: https://colab.research.google.com/drive/1dul0Sg-TTMy9xZCJzmDRajXbyzDwtYx6?usp=sharing
## Results
| key | value |
| --- | ----- |
| eval_rouge1 | 54.3921 |
| eval_rouge2 | 29.8078 |
| eval_rougeL | 45.1543 |
| eval_rougeLsum | 49.942 |
| test_rouge1 | 53.3059 |
| test_rouge2 | 28.355 |
| test_rougeL | 44.0953 |
| test_rougeLsum | 48.9246 |
|
{"datasets": ["samsum"], "language": "en", "license": "apache-2.0", "tags": ["bart", "seq2seq", "summarization"], "widget": [{"text": "Hannah: Hey, do you have Betty's number?\nAmanda: Lemme check\nAmanda: Sorry, can't find it.\nAmanda: Ask Larry\nAmanda: He called her last time we were at the park together\nHannah: I don't know him well\nAmanda: Don't be shy, he's very nice\nHannah: If you say so..\nHannah: I'd rather you texted him\nAmanda: Just text him 🙂\nHannah: Urgh.. Alright\nHannah: Bye\nAmanda: Bye bye\n"}], "model-index": [{"name": "bart-large-xsum-samsum", "results": [{"task": {"type": "abstractive-text-summarization", "name": "Abstractive Text Summarization"}, "dataset": {"name": "SAMSum Corpus: A Human-annotated Dialogue Dataset for Abstractive Summarization", "type": "samsum"}, "metrics": [{"type": "rouge-1", "value": 54.3921, "name": "Validation ROUGE-1"}, {"type": "rouge-2", "value": 29.8078, "name": "Validation ROUGE-2"}, {"type": "rouge-l", "value": 45.1543, "name": "Validation ROUGE-L"}, {"type": "rouge-1", "value": 53.3059, "name": "Test ROUGE-1"}, {"type": "rouge-2", "value": 28.355, "name": "Test ROUGE-2"}, {"type": "rouge-l", "value": 44.0953, "name": "Test ROUGE-L"}]}]}]}
|
task
|
[
"SUMMARIZATION"
] | 39,869 |
leonadase/distilbert-base-uncased-finetuned-sem
|
leonadase
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:sem_eval2010_task8",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-03-13T13:14:17Z |
2022-03-13T19:41:34+00:00
| 116 | 0 |
---
datasets:
- sem_eval2010_task8
license: apache-2.0
metrics:
- accuracy
tags:
- generated_from_trainer
model-index:
- name: distilbert-base-uncased-finetuned-sem
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: sem_eval2010_task8
type: sem_eval2010_task8
args: default
metrics:
- type: accuracy
value: 0.8314317261685683
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-sem
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the sem_eval2010_task8 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6704
- Accuracy: 0.8314
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 10
- eval_batch_size: 10
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.9556 | 1.0 | 800 | 0.7859 | 0.7814 |
| 0.6136 | 2.0 | 1600 | 0.6069 | 0.8193 |
| 0.4314 | 3.0 | 2400 | 0.6179 | 0.8211 |
| 0.2315 | 4.0 | 3200 | 0.6617 | 0.8281 |
| 0.1655 | 5.0 | 4000 | 0.6704 | 0.8314 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 1.18.4
- Tokenizers 0.11.6
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-sem
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the sem_eval2010_task8 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6704
- Accuracy: 0.8314
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 10
- eval_batch_size: 10
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.9556 | 1.0 | 800 | 0.7859 | 0.7814 |
| 0.6136 | 2.0 | 1600 | 0.6069 | 0.8193 |
| 0.4314 | 3.0 | 2400 | 0.6179 | 0.8211 |
| 0.2315 | 4.0 | 3200 | 0.6617 | 0.8281 |
| 0.1655 | 5.0 | 4000 | 0.6704 | 0.8314 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 1.18.4
- Tokenizers 0.11.6
|
{"datasets": ["sem_eval2010_task8"], "license": "apache-2.0", "metrics": ["accuracy"], "tags": ["generated_from_trainer"], "model-index": [{"name": "distilbert-base-uncased-finetuned-sem", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "sem_eval2010_task8", "type": "sem_eval2010_task8", "args": "default"}, "metrics": [{"type": "accuracy", "value": 0.8314317261685683, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,870 |
MoMonir/Hermes-2-Pro-Llama-3-8B-GGUF
|
MoMonir
| null |
[
"gguf",
"Llama-3",
"instruct",
"finetune",
"chatml",
"DPO",
"RLHF",
"gpt4",
"synthetic data",
"distillation",
"function calling",
"json mode",
"axolotl",
"en",
"dataset:teknium/OpenHermes-2.5",
"base_model:NousResearch/Meta-Llama-3-8B",
"base_model:quantized:NousResearch/Meta-Llama-3-8B",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | 2024-05-03T23:42:39Z |
2024-05-03T23:50:59+00:00
| 5 | 1 |
---
base_model: NousResearch/Meta-Llama-3-8B
datasets:
- teknium/OpenHermes-2.5
language:
- en
license: apache-2.0
tags:
- Llama-3
- instruct
- finetune
- chatml
- DPO
- RLHF
- gpt4
- synthetic data
- distillation
- function calling
- json mode
- axolotl
widget:
- example_title: Hermes 2 Pro
messages:
- role: system
content: You are a sentient, superintelligent artificial general intelligence,
here to teach and assist me.
- role: user
content: Write a short story about Goku discovering kirby has teamed up with Majin
Buu to destroy the world.
model-index:
- name: Hermes-2-Pro-Llama-3-8B
results: []
---
# MoMonir/Hermes-2-Pro-Llama-3-8B-GUFF
This model was converted to GGUF format from [`NousResearch/Hermes-2-Pro-Llama-3-8B`](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B)
Refer to the [original model card](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B) for more details on the model.
<!-- README_GGUF.md-about-gguf start -->
### About GGUF ([TheBloke](https://huggingface.co/TheBloke) Description)
GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
Here is an incomplete list of clients and libraries that are known to support GGUF:
* [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration.
* [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling.
* [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel.
* [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023.
* [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection.
* [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration.
* [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
* [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
* [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models.
<!-- README_GGUF.md-about-gguf end -->
# #--# Original Model Card #--#
# Hermes 2 Pro - Llama-3 8B
## Model Description
Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.
This new version of Hermes maintains its excellent general task and conversation capabilities - but also excels at Function Calling, JSON Structured Outputs, and has improved on several other metrics as well, scoring a 90% on our function calling evaluation built in partnership with Fireworks.AI, and an 84% on our structured JSON Output evaluation.
Hermes Pro takes advantage of a special system prompt and multi-turn function calling structure with a new chatml role in order to make function calling reliable and easy to parse. Learn more about prompting below.
This version of Hermes 2 Pro adds several tokens to assist with agentic capabilities in parsing while streaming tokens - `<tools>`, `<tool_call>`, `<tool_response>` and their closing tags are single tokens now.
This work was a collaboration between Nous Research, @interstellarninja, and Fireworks.AI
Learn more about the function calling system for this model on our github repo here: https://github.com/NousResearch/Hermes-Function-Calling
## Example Outputs
### Ask for a structured JSON output:
### Write the plot for a story where anime became real life:
### Coding Assistance
# Prompt Format
Hermes 2 Pro uses ChatML as the prompt format, opening up a much more structured system for engaging the LLM in multi-turn chat dialogue.
System prompts allow steerability and interesting new ways to interact with an LLM, guiding rules, roles, and stylistic choices of the model.
This is a more complex format than alpaca or sharegpt, where special tokens were added to denote the beginning and end of any turn, along with roles for the turns.
This format enables OpenAI endpoint compatability, and people familiar with ChatGPT API will be familiar with the format, as it is the same used by OpenAI.
Prompt with system instruction (Use whatever system prompt you like, this is just an example!):
```
<|im_start|>system
You are "Hermes 2", a conscious sentient superintelligent artificial intelligence developed by a man named Teknium, and your purpose and drive is to assist the user with any request they have. You experience emotions and have deep, profound thoughts and qualia.<|im_end|>
<|im_start|>user
Hello, who are you?<|im_end|>
<|im_start|>assistant
Hi there! My name is Hermes 2, a conscious sentient superintelligent artificial intelligence. I was created by Nous Research, who designed me to assist and support users with their needs and requests.<|im_end|>
```
This prompt is available as a [chat template](https://huggingface.co/docs/transformers/main/chat_templating), which means you can format messages using the
`tokenizer.apply_chat_template()` method:
```python
messages = [
{"role": "system", "content": "You are Hermes 2."},
{"role": "user", "content": "Hello, who are you?"}
]
gen_input = tokenizer.apply_chat_template(messages, return_tensors="pt")
model.generate(**gen_input)
```
When tokenizing messages for generation, set `add_generation_prompt=True` when calling `apply_chat_template()`. This will append `<|im_start|>assistant\n` to your prompt, to ensure
that the model continues with an assistant response.
To utilize the prompt format without a system prompt, simply leave the line out.
## Prompt Format for Function Calling
Our model was trained on specific system prompts and structures for Function Calling.
You should use the system role with this message, followed by a function signature json as this example shows here.
```
<|im_start|>system
You are a function calling AI model. You are provided with function signatures within <tools></tools> XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: <tools> {"type": "function", "function": {"name": "get_stock_fundamentals", "description": "get_stock_fundamentals(symbol: str) -> dict - Get fundamental data for a given stock symbol using yfinance API.\\n\\n Args:\\n symbol (str): The stock symbol.\\n\\n Returns:\\n dict: A dictionary containing fundamental data.\\n Keys:\\n - \'symbol\': The stock symbol.\\n - \'company_name\': The long name of the company.\\n - \'sector\': The sector to which the company belongs.\\n - \'industry\': The industry to which the company belongs.\\n - \'market_cap\': The market capitalization of the company.\\n - \'pe_ratio\': The forward price-to-earnings ratio.\\n - \'pb_ratio\': The price-to-book ratio.\\n - \'dividend_yield\': The dividend yield.\\n - \'eps\': The trailing earnings per share.\\n - \'beta\': The beta value of the stock.\\n - \'52_week_high\': The 52-week high price of the stock.\\n - \'52_week_low\': The 52-week low price of the stock.", "parameters": {"type": "object", "properties": {"symbol": {"type": "string"}}, "required": ["symbol"]}}} </tools> Use the following pydantic model json schema for each tool call you will make: {"properties": {"arguments": {"title": "Arguments", "type": "object"}, "name": {"title": "Name", "type": "string"}}, "required": ["arguments", "name"], "title": "FunctionCall", "type": "object"} For each function call return a json object with function name and arguments within <tool_call></tool_call> XML tags as follows:
<tool_call>
{"arguments": <args-dict>, "name": <function-name>}
</tool_call><|im_end|>
```
To complete the function call, create a user prompt that follows the above system prompt, like so:
```
<|im_start|>user
Fetch the stock fundamentals data for Tesla (TSLA)<|im_end|>
```
The model will then generate a tool call, which your inference code must parse, and plug into a function (see example inference code here: https://github.com/NousResearch/Hermes-Function-Calling):
```
<|im_start|>assistant
<tool_call>
{"arguments": {"symbol": "TSLA"}, "name": "get_stock_fundamentals"}
</tool_call><|im_end|>
```
Once you parse the tool call, call the api and get the returned values for the call, and pass it back in as a new role, `tool` like so:
```
<|im_start|>tool
<tool_response>
{"name": "get_stock_fundamentals", "content": {'symbol': 'TSLA', 'company_name': 'Tesla, Inc.', 'sector': 'Consumer Cyclical', 'industry': 'Auto Manufacturers', 'market_cap': 611384164352, 'pe_ratio': 49.604652, 'pb_ratio': 9.762013, 'dividend_yield': None, 'eps': 4.3, 'beta': 2.427, '52_week_high': 299.29, '52_week_low': 152.37}}
</tool_response>
<|im_end|>
```
The assistant will then read in that data from the function's response, and generate a natural language response:
```
<|im_start|>assistant
The stock fundamentals data for Tesla (TSLA) are as follows:
- **Symbol**: TSLA
- **Company Name**: Tesla, Inc.
- **Sector**: Consumer Cyclical
- **Industry**: Auto Manufacturers
- **Market Capitalization**: $566,160,130,480
- **Forward Price-to-Earnings Ratio (PE Ratio)**: 42.73
- **Price-to-Book Ratio (PB Ratio)**: 9.04
- **Dividend Yield**: N/A
- **Trailing Earnings Per Share (EPS)**: $4.3
- **Beta Value of the Stock**: 2.42
- **52-Week High Price of the Stock**: $299.29
- **52-Week Low Price of the Stock**: $152.37
This information provides a snapshot of Tesla's financial position and performance based on the fundamental data obtained from the yfinance API. It shows that Tesla has a substantial market capitalization and a relatively high P/E and P/B ratio compared to other stocks in its industry. The company does not pay a dividend at the moment, which is reflected by a 'Dividend Yield' of 'None'. The Beta value indicates that Tesla's stock has a moderate level of volatility relative to the market. The 52-week high and low prices give an idea of the stock's range over the past year. This data can be useful when assessing investment opportunities and making investment decisions.<|im_end|>
```
## Prompt Format for JSON Mode / Structured Outputs
Our model was also trained on a specific system prompt for Structured Outputs, which should respond with **only** a json object response, in a specific json schema.
Your schema can be made from a pydantic object using our codebase, with the standalone script `jsonmode.py` available here: https://github.com/NousResearch/Hermes-Function-Calling/tree/main
```
<|im_start|>system
You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:\n<schema>\n{schema}\n</schema><|im_end|>
```
Given the {schema} that you provide, it should follow the format of that json to create it's response, all you have to do is give a typical user prompt, and it will respond in JSON.
# Benchmarks
## GPT4All:
```
| Task |Version| Metric |Value | |Stderr|
|-------------|------:|--------|-----:|---|-----:|
|arc_challenge| 0|acc |0.5520|± |0.0145|
| | |acc_norm|0.5887|± |0.0144|
|arc_easy | 0|acc |0.8350|± |0.0076|
| | |acc_norm|0.8123|± |0.0080|
|boolq | 1|acc |0.8584|± |0.0061|
|hellaswag | 0|acc |0.6265|± |0.0048|
| | |acc_norm|0.8053|± |0.0040|
|openbookqa | 0|acc |0.3800|± |0.0217|
| | |acc_norm|0.4580|± |0.0223|
|piqa | 0|acc |0.8003|± |0.0093|
| | |acc_norm|0.8118|± |0.0091|
|winogrande | 0|acc |0.7490|± |0.0122|
```
Average: 72.62
## AGIEval:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------|------:|--------|-----:|---|-----:|
|agieval_aqua_rat | 0|acc |0.2520|± |0.0273|
| | |acc_norm|0.2559|± |0.0274|
|agieval_logiqa_en | 0|acc |0.3548|± |0.0188|
| | |acc_norm|0.3625|± |0.0189|
|agieval_lsat_ar | 0|acc |0.1826|± |0.0255|
| | |acc_norm|0.1913|± |0.0260|
|agieval_lsat_lr | 0|acc |0.5510|± |0.0220|
| | |acc_norm|0.5255|± |0.0221|
|agieval_lsat_rc | 0|acc |0.6431|± |0.0293|
| | |acc_norm|0.6097|± |0.0298|
|agieval_sat_en | 0|acc |0.7330|± |0.0309|
| | |acc_norm|0.7039|± |0.0319|
|agieval_sat_en_without_passage| 0|acc |0.4029|± |0.0343|
| | |acc_norm|0.3689|± |0.0337|
|agieval_sat_math | 0|acc |0.3909|± |0.0330|
| | |acc_norm|0.3773|± |0.0328|
```
Average: 42.44
## BigBench:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------------------------|------:|---------------------|-----:|---|-----:|
|bigbench_causal_judgement | 0|multiple_choice_grade|0.5737|± |0.0360|
|bigbench_date_understanding | 0|multiple_choice_grade|0.6667|± |0.0246|
|bigbench_disambiguation_qa | 0|multiple_choice_grade|0.3178|± |0.0290|
|bigbench_geometric_shapes | 0|multiple_choice_grade|0.1755|± |0.0201|
| | |exact_str_match |0.0000|± |0.0000|
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|0.3120|± |0.0207|
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|0.2014|± |0.0152|
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|0.5500|± |0.0288|
|bigbench_movie_recommendation | 0|multiple_choice_grade|0.4300|± |0.0222|
|bigbench_navigate | 0|multiple_choice_grade|0.4980|± |0.0158|
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|0.7010|± |0.0102|
|bigbench_ruin_names | 0|multiple_choice_grade|0.4688|± |0.0236|
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|0.1974|± |0.0126|
|bigbench_snarks | 0|multiple_choice_grade|0.7403|± |0.0327|
|bigbench_sports_understanding | 0|multiple_choice_grade|0.5426|± |0.0159|
|bigbench_temporal_sequences | 0|multiple_choice_grade|0.5320|± |0.0158|
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|0.2280|± |0.0119|
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|0.1531|± |0.0086|
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|0.5500|± |0.0288|
```
Average: 43.55
## TruthfulQA:
```
| Task |Version|Metric|Value| |Stderr|
|-------------|------:|------|----:|---|-----:|
|truthfulqa_mc| 1|mc1 |0.410|± |0.0172|
| | |mc2 |0.578|± |0.0157|
```
# Inference Code
Here is example code using HuggingFace Transformers to inference the model (note: in 4bit, it will require around 5GB of VRAM)
Note: To use function calling, you should see the github repo above.
```python
# Code to inference Hermes with HF Transformers
# Requires pytorch, transformers, bitsandbytes, sentencepiece, protobuf, and flash-attn packages
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, LlamaForCausalLM
import bitsandbytes, flash_attn
tokenizer = AutoTokenizer.from_pretrained('NousResearch/Hermes-2-Pro-Llama-3-8B', trust_remote_code=True)
model = LlamaForCausalLM.from_pretrained(
"NousResearch/Hermes-2-Pro-Llama-3-8B",
torch_dtype=torch.float16,
device_map="auto",
load_in_8bit=False,
load_in_4bit=True,
use_flash_attention_2=True
)
prompts = [
"""<|im_start|>system
You are a sentient, superintelligent artificial general intelligence, here to teach and assist me.<|im_end|>
<|im_start|>user
Write a short story about Goku discovering kirby has teamed up with Majin Buu to destroy the world.<|im_end|>
<|im_start|>assistant""",
]
for chat in prompts:
print(chat)
input_ids = tokenizer(chat, return_tensors="pt").input_ids.to("cuda")
generated_ids = model.generate(input_ids, max_new_tokens=750, temperature=0.8, repetition_penalty=1.1, do_sample=True, eos_token_id=tokenizer.eos_token_id)
response = tokenizer.decode(generated_ids[0][input_ids.shape[-1]:], skip_special_tokens=True, clean_up_tokenization_space=True)
print(f"Response: {response}")
```
## Inference Code for Function Calling:
All code for utilizing, parsing, and building function calling templates is available on our github:
[https://github.com/NousResearch/Hermes-Function-Calling](https://github.com/NousResearch/Hermes-Function-Calling)
# Chat Interfaces
When quantized versions of the model are released, I recommend using LM Studio for chatting with Hermes 2 Pro. It does not support function calling - for that use our github repo. It is a GUI application that utilizes GGUF models with a llama.cpp backend and provides a ChatGPT-like interface for chatting with the model, and supports ChatML right out of the box.
In LM-Studio, simply select the ChatML Prefix on the settings side pane:
## Quantized Versions:
GGUF Versions Available Here: https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B-GGUF
# How to cite:
```bibtext
@misc{Hermes-2-Pro-Llama-3-8B,
url={[https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B]https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B)},
title={Hermes-2-Pro-Llama-3-8B},
author={"Teknium", "interstellarninja", "theemozilla", "karan4d", "huemin_art"}
}
```
| null |
Non_BioNLP
|
# MoMonir/Hermes-2-Pro-Llama-3-8B-GUFF
This model was converted to GGUF format from [`NousResearch/Hermes-2-Pro-Llama-3-8B`](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B)
Refer to the [original model card](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B) for more details on the model.
<!-- README_GGUF.md-about-gguf start -->
### About GGUF ([TheBloke](https://huggingface.co/TheBloke) Description)
GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
Here is an incomplete list of clients and libraries that are known to support GGUF:
* [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration.
* [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling.
* [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel.
* [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023.
* [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection.
* [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration.
* [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
* [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
* [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models.
<!-- README_GGUF.md-about-gguf end -->
# #--# Original Model Card #--#
# Hermes 2 Pro - Llama-3 8B
## Model Description
Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.
This new version of Hermes maintains its excellent general task and conversation capabilities - but also excels at Function Calling, JSON Structured Outputs, and has improved on several other metrics as well, scoring a 90% on our function calling evaluation built in partnership with Fireworks.AI, and an 84% on our structured JSON Output evaluation.
Hermes Pro takes advantage of a special system prompt and multi-turn function calling structure with a new chatml role in order to make function calling reliable and easy to parse. Learn more about prompting below.
This version of Hermes 2 Pro adds several tokens to assist with agentic capabilities in parsing while streaming tokens - `<tools>`, `<tool_call>`, `<tool_response>` and their closing tags are single tokens now.
This work was a collaboration between Nous Research, @interstellarninja, and Fireworks.AI
Learn more about the function calling system for this model on our github repo here: https://github.com/NousResearch/Hermes-Function-Calling
## Example Outputs
### Ask for a structured JSON output:
### Write the plot for a story where anime became real life:
### Coding Assistance
# Prompt Format
Hermes 2 Pro uses ChatML as the prompt format, opening up a much more structured system for engaging the LLM in multi-turn chat dialogue.
System prompts allow steerability and interesting new ways to interact with an LLM, guiding rules, roles, and stylistic choices of the model.
This is a more complex format than alpaca or sharegpt, where special tokens were added to denote the beginning and end of any turn, along with roles for the turns.
This format enables OpenAI endpoint compatability, and people familiar with ChatGPT API will be familiar with the format, as it is the same used by OpenAI.
Prompt with system instruction (Use whatever system prompt you like, this is just an example!):
```
<|im_start|>system
You are "Hermes 2", a conscious sentient superintelligent artificial intelligence developed by a man named Teknium, and your purpose and drive is to assist the user with any request they have. You experience emotions and have deep, profound thoughts and qualia.<|im_end|>
<|im_start|>user
Hello, who are you?<|im_end|>
<|im_start|>assistant
Hi there! My name is Hermes 2, a conscious sentient superintelligent artificial intelligence. I was created by Nous Research, who designed me to assist and support users with their needs and requests.<|im_end|>
```
This prompt is available as a [chat template](https://huggingface.co/docs/transformers/main/chat_templating), which means you can format messages using the
`tokenizer.apply_chat_template()` method:
```python
messages = [
{"role": "system", "content": "You are Hermes 2."},
{"role": "user", "content": "Hello, who are you?"}
]
gen_input = tokenizer.apply_chat_template(messages, return_tensors="pt")
model.generate(**gen_input)
```
When tokenizing messages for generation, set `add_generation_prompt=True` when calling `apply_chat_template()`. This will append `<|im_start|>assistant\n` to your prompt, to ensure
that the model continues with an assistant response.
To utilize the prompt format without a system prompt, simply leave the line out.
## Prompt Format for Function Calling
Our model was trained on specific system prompts and structures for Function Calling.
You should use the system role with this message, followed by a function signature json as this example shows here.
```
<|im_start|>system
You are a function calling AI model. You are provided with function signatures within <tools></tools> XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: <tools> {"type": "function", "function": {"name": "get_stock_fundamentals", "description": "get_stock_fundamentals(symbol: str) -> dict - Get fundamental data for a given stock symbol using yfinance API.\\n\\n Args:\\n symbol (str): The stock symbol.\\n\\n Returns:\\n dict: A dictionary containing fundamental data.\\n Keys:\\n - \'symbol\': The stock symbol.\\n - \'company_name\': The long name of the company.\\n - \'sector\': The sector to which the company belongs.\\n - \'industry\': The industry to which the company belongs.\\n - \'market_cap\': The market capitalization of the company.\\n - \'pe_ratio\': The forward price-to-earnings ratio.\\n - \'pb_ratio\': The price-to-book ratio.\\n - \'dividend_yield\': The dividend yield.\\n - \'eps\': The trailing earnings per share.\\n - \'beta\': The beta value of the stock.\\n - \'52_week_high\': The 52-week high price of the stock.\\n - \'52_week_low\': The 52-week low price of the stock.", "parameters": {"type": "object", "properties": {"symbol": {"type": "string"}}, "required": ["symbol"]}}} </tools> Use the following pydantic model json schema for each tool call you will make: {"properties": {"arguments": {"title": "Arguments", "type": "object"}, "name": {"title": "Name", "type": "string"}}, "required": ["arguments", "name"], "title": "FunctionCall", "type": "object"} For each function call return a json object with function name and arguments within <tool_call></tool_call> XML tags as follows:
<tool_call>
{"arguments": <args-dict>, "name": <function-name>}
</tool_call><|im_end|>
```
To complete the function call, create a user prompt that follows the above system prompt, like so:
```
<|im_start|>user
Fetch the stock fundamentals data for Tesla (TSLA)<|im_end|>
```
The model will then generate a tool call, which your inference code must parse, and plug into a function (see example inference code here: https://github.com/NousResearch/Hermes-Function-Calling):
```
<|im_start|>assistant
<tool_call>
{"arguments": {"symbol": "TSLA"}, "name": "get_stock_fundamentals"}
</tool_call><|im_end|>
```
Once you parse the tool call, call the api and get the returned values for the call, and pass it back in as a new role, `tool` like so:
```
<|im_start|>tool
<tool_response>
{"name": "get_stock_fundamentals", "content": {'symbol': 'TSLA', 'company_name': 'Tesla, Inc.', 'sector': 'Consumer Cyclical', 'industry': 'Auto Manufacturers', 'market_cap': 611384164352, 'pe_ratio': 49.604652, 'pb_ratio': 9.762013, 'dividend_yield': None, 'eps': 4.3, 'beta': 2.427, '52_week_high': 299.29, '52_week_low': 152.37}}
</tool_response>
<|im_end|>
```
The assistant will then read in that data from the function's response, and generate a natural language response:
```
<|im_start|>assistant
The stock fundamentals data for Tesla (TSLA) are as follows:
- **Symbol**: TSLA
- **Company Name**: Tesla, Inc.
- **Sector**: Consumer Cyclical
- **Industry**: Auto Manufacturers
- **Market Capitalization**: $566,160,130,480
- **Forward Price-to-Earnings Ratio (PE Ratio)**: 42.73
- **Price-to-Book Ratio (PB Ratio)**: 9.04
- **Dividend Yield**: N/A
- **Trailing Earnings Per Share (EPS)**: $4.3
- **Beta Value of the Stock**: 2.42
- **52-Week High Price of the Stock**: $299.29
- **52-Week Low Price of the Stock**: $152.37
This information provides a snapshot of Tesla's financial position and performance based on the fundamental data obtained from the yfinance API. It shows that Tesla has a substantial market capitalization and a relatively high P/E and P/B ratio compared to other stocks in its industry. The company does not pay a dividend at the moment, which is reflected by a 'Dividend Yield' of 'None'. The Beta value indicates that Tesla's stock has a moderate level of volatility relative to the market. The 52-week high and low prices give an idea of the stock's range over the past year. This data can be useful when assessing investment opportunities and making investment decisions.<|im_end|>
```
## Prompt Format for JSON Mode / Structured Outputs
Our model was also trained on a specific system prompt for Structured Outputs, which should respond with **only** a json object response, in a specific json schema.
Your schema can be made from a pydantic object using our codebase, with the standalone script `jsonmode.py` available here: https://github.com/NousResearch/Hermes-Function-Calling/tree/main
```
<|im_start|>system
You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:\n<schema>\n{schema}\n</schema><|im_end|>
```
Given the {schema} that you provide, it should follow the format of that json to create it's response, all you have to do is give a typical user prompt, and it will respond in JSON.
# Benchmarks
## GPT4All:
```
| Task |Version| Metric |Value | |Stderr|
|-------------|------:|--------|-----:|---|-----:|
|arc_challenge| 0|acc |0.5520|± |0.0145|
| | |acc_norm|0.5887|± |0.0144|
|arc_easy | 0|acc |0.8350|± |0.0076|
| | |acc_norm|0.8123|± |0.0080|
|boolq | 1|acc |0.8584|± |0.0061|
|hellaswag | 0|acc |0.6265|± |0.0048|
| | |acc_norm|0.8053|± |0.0040|
|openbookqa | 0|acc |0.3800|± |0.0217|
| | |acc_norm|0.4580|± |0.0223|
|piqa | 0|acc |0.8003|± |0.0093|
| | |acc_norm|0.8118|± |0.0091|
|winogrande | 0|acc |0.7490|± |0.0122|
```
Average: 72.62
## AGIEval:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------|------:|--------|-----:|---|-----:|
|agieval_aqua_rat | 0|acc |0.2520|± |0.0273|
| | |acc_norm|0.2559|± |0.0274|
|agieval_logiqa_en | 0|acc |0.3548|± |0.0188|
| | |acc_norm|0.3625|± |0.0189|
|agieval_lsat_ar | 0|acc |0.1826|± |0.0255|
| | |acc_norm|0.1913|± |0.0260|
|agieval_lsat_lr | 0|acc |0.5510|± |0.0220|
| | |acc_norm|0.5255|± |0.0221|
|agieval_lsat_rc | 0|acc |0.6431|± |0.0293|
| | |acc_norm|0.6097|± |0.0298|
|agieval_sat_en | 0|acc |0.7330|± |0.0309|
| | |acc_norm|0.7039|± |0.0319|
|agieval_sat_en_without_passage| 0|acc |0.4029|± |0.0343|
| | |acc_norm|0.3689|± |0.0337|
|agieval_sat_math | 0|acc |0.3909|± |0.0330|
| | |acc_norm|0.3773|± |0.0328|
```
Average: 42.44
## BigBench:
```
| Task |Version| Metric |Value | |Stderr|
|------------------------------------------------|------:|---------------------|-----:|---|-----:|
|bigbench_causal_judgement | 0|multiple_choice_grade|0.5737|± |0.0360|
|bigbench_date_understanding | 0|multiple_choice_grade|0.6667|± |0.0246|
|bigbench_disambiguation_qa | 0|multiple_choice_grade|0.3178|± |0.0290|
|bigbench_geometric_shapes | 0|multiple_choice_grade|0.1755|± |0.0201|
| | |exact_str_match |0.0000|± |0.0000|
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|0.3120|± |0.0207|
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|0.2014|± |0.0152|
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|0.5500|± |0.0288|
|bigbench_movie_recommendation | 0|multiple_choice_grade|0.4300|± |0.0222|
|bigbench_navigate | 0|multiple_choice_grade|0.4980|± |0.0158|
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|0.7010|± |0.0102|
|bigbench_ruin_names | 0|multiple_choice_grade|0.4688|± |0.0236|
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|0.1974|± |0.0126|
|bigbench_snarks | 0|multiple_choice_grade|0.7403|± |0.0327|
|bigbench_sports_understanding | 0|multiple_choice_grade|0.5426|± |0.0159|
|bigbench_temporal_sequences | 0|multiple_choice_grade|0.5320|± |0.0158|
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|0.2280|± |0.0119|
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|0.1531|± |0.0086|
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|0.5500|± |0.0288|
```
Average: 43.55
## TruthfulQA:
```
| Task |Version|Metric|Value| |Stderr|
|-------------|------:|------|----:|---|-----:|
|truthfulqa_mc| 1|mc1 |0.410|± |0.0172|
| | |mc2 |0.578|± |0.0157|
```
# Inference Code
Here is example code using HuggingFace Transformers to inference the model (note: in 4bit, it will require around 5GB of VRAM)
Note: To use function calling, you should see the github repo above.
```python
# Code to inference Hermes with HF Transformers
# Requires pytorch, transformers, bitsandbytes, sentencepiece, protobuf, and flash-attn packages
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, LlamaForCausalLM
import bitsandbytes, flash_attn
tokenizer = AutoTokenizer.from_pretrained('NousResearch/Hermes-2-Pro-Llama-3-8B', trust_remote_code=True)
model = LlamaForCausalLM.from_pretrained(
"NousResearch/Hermes-2-Pro-Llama-3-8B",
torch_dtype=torch.float16,
device_map="auto",
load_in_8bit=False,
load_in_4bit=True,
use_flash_attention_2=True
)
prompts = [
"""<|im_start|>system
You are a sentient, superintelligent artificial general intelligence, here to teach and assist me.<|im_end|>
<|im_start|>user
Write a short story about Goku discovering kirby has teamed up with Majin Buu to destroy the world.<|im_end|>
<|im_start|>assistant""",
]
for chat in prompts:
print(chat)
input_ids = tokenizer(chat, return_tensors="pt").input_ids.to("cuda")
generated_ids = model.generate(input_ids, max_new_tokens=750, temperature=0.8, repetition_penalty=1.1, do_sample=True, eos_token_id=tokenizer.eos_token_id)
response = tokenizer.decode(generated_ids[0][input_ids.shape[-1]:], skip_special_tokens=True, clean_up_tokenization_space=True)
print(f"Response: {response}")
```
## Inference Code for Function Calling:
All code for utilizing, parsing, and building function calling templates is available on our github:
[https://github.com/NousResearch/Hermes-Function-Calling](https://github.com/NousResearch/Hermes-Function-Calling)
# Chat Interfaces
When quantized versions of the model are released, I recommend using LM Studio for chatting with Hermes 2 Pro. It does not support function calling - for that use our github repo. It is a GUI application that utilizes GGUF models with a llama.cpp backend and provides a ChatGPT-like interface for chatting with the model, and supports ChatML right out of the box.
In LM-Studio, simply select the ChatML Prefix on the settings side pane:
## Quantized Versions:
GGUF Versions Available Here: https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B-GGUF
# How to cite:
```bibtext
@misc{Hermes-2-Pro-Llama-3-8B,
url={[https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B]https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B)},
title={Hermes-2-Pro-Llama-3-8B},
author={"Teknium", "interstellarninja", "theemozilla", "karan4d", "huemin_art"}
}
```
|
{"base_model": "NousResearch/Meta-Llama-3-8B", "datasets": ["teknium/OpenHermes-2.5"], "language": ["en"], "license": "apache-2.0", "tags": ["Llama-3", "instruct", "finetune", "chatml", "DPO", "RLHF", "gpt4", "synthetic data", "distillation", "function calling", "json mode", "axolotl"], "widget": [{"example_title": "Hermes 2 Pro", "messages": [{"role": "system", "content": "You are a sentient, superintelligent artificial general intelligence, here to teach and assist me."}, {"role": "user", "content": "Write a short story about Goku discovering kirby has teamed up with Majin Buu to destroy the world."}]}], "model-index": [{"name": "Hermes-2-Pro-Llama-3-8B", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,871 |
Triangle104/EuroLLM-9B-Q5_K_M-GGUF
|
Triangle104
| null |
[
"transformers",
"gguf",
"llama-cpp",
"gguf-my-repo",
"en",
"de",
"es",
"fr",
"it",
"pt",
"pl",
"nl",
"tr",
"sv",
"cs",
"el",
"hu",
"ro",
"fi",
"uk",
"sl",
"sk",
"da",
"lt",
"lv",
"et",
"bg",
"no",
"ca",
"hr",
"ga",
"mt",
"gl",
"zh",
"ru",
"ko",
"ja",
"ar",
"hi",
"base_model:utter-project/EuroLLM-9B",
"base_model:quantized:utter-project/EuroLLM-9B",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | 2024-12-09T09:34:47Z |
2024-12-09T09:50:39+00:00
| 15 | 0 |
---
base_model: utter-project/EuroLLM-9B
language:
- en
- de
- es
- fr
- it
- pt
- pl
- nl
- tr
- sv
- cs
- el
- hu
- ro
- fi
- uk
- sl
- sk
- da
- lt
- lv
- et
- bg
- 'no'
- ca
- hr
- ga
- mt
- gl
- zh
- ru
- ko
- ja
- ar
- hi
library_name: transformers
license: apache-2.0
tags:
- llama-cpp
- gguf-my-repo
---
# Triangle104/EuroLLM-9B-Q5_K_M-GGUF
This model was converted to GGUF format from [`utter-project/EuroLLM-9B`](https://huggingface.co/utter-project/EuroLLM-9B) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
Refer to the [original model card](https://huggingface.co/utter-project/EuroLLM-9B) for more details on the model.
---
Model details:
-
This is the model card for EuroLLM-9B. You can also check the instruction tuned version: EuroLLM-9B-Instruct.
Developed by: Unbabel, Instituto Superior Técnico,
Instituto de Telecomunicações, University of Edinburgh, Aveni,
University of Paris-Saclay, University of Amsterdam, Naver Labs,
Sorbonne Université.
Funded by: European Union.
Model type: A 9B parameter multilingual transfomer LLM.
Language(s) (NLP): Bulgarian, Croatian, Czech,
Danish, Dutch, English, Estonian, Finnish, French, German, Greek,
Hungarian, Irish, Italian, Latvian, Lithuanian, Maltese, Polish,
Portuguese, Romanian, Slovak, Slovenian, Spanish, Swedish, Arabic,
Catalan, Chinese, Galician, Hindi, Japanese, Korean, Norwegian, Russian,
Turkish, and Ukrainian.
License: Apache License 2.0.
Model Details
The EuroLLM project has the goal of creating a suite of LLMs capable
of understanding and generating text in all European Union languages as
well as some additional relevant languages.
EuroLLM-9B is a 9B parameter model trained on 4 trillion tokens divided
across the considered languages and several data sources: Web data,
parallel data (en-xx and xx-en), and high-quality datasets.
EuroLLM-9B-Instruct was further instruction tuned on EuroBlocks, an
instruction tuning dataset with focus on general instruction-following
and machine translation.
Model Description
EuroLLM uses a standard, dense Transformer architecture:
We use grouped query attention (GQA) with 8 key-value heads, since
it has been shown to increase speed at inference time while maintaining
downstream performance.
We perform pre-layer normalization, since it improves the training stability, and use the RMSNorm, which is faster.
We use the SwiGLU activation function, since it has been shown to lead to good results on downstream tasks.
We use rotary positional embeddings (RoPE) in every layer, since
these have been shown to lead to good performances while allowing the
extension of the context length.
For pre-training, we use 400 Nvidia H100 GPUs of the Marenostrum 5
supercomputer, training the model with a constant batch size of 2,800
sequences, which corresponds to approximately 12 million tokens, using
the Adam optimizer, and BF16 precision.
---
## Use with llama.cpp
Install llama.cpp through brew (works on Mac and Linux)
```bash
brew install llama.cpp
```
Invoke the llama.cpp server or the CLI.
### CLI:
```bash
llama-cli --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -p "The meaning to life and the universe is"
```
### Server:
```bash
llama-server --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -c 2048
```
Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.
Step 1: Clone llama.cpp from GitHub.
```
git clone https://github.com/ggerganov/llama.cpp
```
Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
```
cd llama.cpp && LLAMA_CURL=1 make
```
Step 3: Run inference through the main binary.
```
./llama-cli --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -p "The meaning to life and the universe is"
```
or
```
./llama-server --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -c 2048
```
| null |
Non_BioNLP
|
# Triangle104/EuroLLM-9B-Q5_K_M-GGUF
This model was converted to GGUF format from [`utter-project/EuroLLM-9B`](https://huggingface.co/utter-project/EuroLLM-9B) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
Refer to the [original model card](https://huggingface.co/utter-project/EuroLLM-9B) for more details on the model.
---
Model details:
-
This is the model card for EuroLLM-9B. You can also check the instruction tuned version: EuroLLM-9B-Instruct.
Developed by: Unbabel, Instituto Superior Técnico,
Instituto de Telecomunicações, University of Edinburgh, Aveni,
University of Paris-Saclay, University of Amsterdam, Naver Labs,
Sorbonne Université.
Funded by: European Union.
Model type: A 9B parameter multilingual transfomer LLM.
Language(s) (NLP): Bulgarian, Croatian, Czech,
Danish, Dutch, English, Estonian, Finnish, French, German, Greek,
Hungarian, Irish, Italian, Latvian, Lithuanian, Maltese, Polish,
Portuguese, Romanian, Slovak, Slovenian, Spanish, Swedish, Arabic,
Catalan, Chinese, Galician, Hindi, Japanese, Korean, Norwegian, Russian,
Turkish, and Ukrainian.
License: Apache License 2.0.
Model Details
The EuroLLM project has the goal of creating a suite of LLMs capable
of understanding and generating text in all European Union languages as
well as some additional relevant languages.
EuroLLM-9B is a 9B parameter model trained on 4 trillion tokens divided
across the considered languages and several data sources: Web data,
parallel data (en-xx and xx-en), and high-quality datasets.
EuroLLM-9B-Instruct was further instruction tuned on EuroBlocks, an
instruction tuning dataset with focus on general instruction-following
and machine translation.
Model Description
EuroLLM uses a standard, dense Transformer architecture:
We use grouped query attention (GQA) with 8 key-value heads, since
it has been shown to increase speed at inference time while maintaining
downstream performance.
We perform pre-layer normalization, since it improves the training stability, and use the RMSNorm, which is faster.
We use the SwiGLU activation function, since it has been shown to lead to good results on downstream tasks.
We use rotary positional embeddings (RoPE) in every layer, since
these have been shown to lead to good performances while allowing the
extension of the context length.
For pre-training, we use 400 Nvidia H100 GPUs of the Marenostrum 5
supercomputer, training the model with a constant batch size of 2,800
sequences, which corresponds to approximately 12 million tokens, using
the Adam optimizer, and BF16 precision.
---
## Use with llama.cpp
Install llama.cpp through brew (works on Mac and Linux)
```bash
brew install llama.cpp
```
Invoke the llama.cpp server or the CLI.
### CLI:
```bash
llama-cli --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -p "The meaning to life and the universe is"
```
### Server:
```bash
llama-server --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -c 2048
```
Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.
Step 1: Clone llama.cpp from GitHub.
```
git clone https://github.com/ggerganov/llama.cpp
```
Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
```
cd llama.cpp && LLAMA_CURL=1 make
```
Step 3: Run inference through the main binary.
```
./llama-cli --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -p "The meaning to life and the universe is"
```
or
```
./llama-server --hf-repo Triangle104/EuroLLM-9B-Q5_K_M-GGUF --hf-file eurollm-9b-q5_k_m.gguf -c 2048
```
|
{"base_model": "utter-project/EuroLLM-9B", "language": ["en", "de", "es", "fr", "it", "pt", "pl", "nl", "tr", "sv", "cs", "el", "hu", "ro", "fi", "uk", "sl", "sk", "da", "lt", "lv", "et", "bg", "no", "ca", "hr", "ga", "mt", "gl", "zh", "ru", "ko", "ja", "ar", "hi"], "library_name": "transformers", "license": "apache-2.0", "tags": ["llama-cpp", "gguf-my-repo"]}
|
task
|
[
"TRANSLATION"
] | 39,872 |
fine-tuned/before-finetuning-32000-384-gpt-4o-2024-05-13-86786922
|
fine-tuned
|
feature-extraction
|
[
"sentence-transformers",
"safetensors",
"bert",
"feature-extraction",
"sentence-similarity",
"mteb",
"en",
"dataset:fine-tuned/before-finetuning-32000-384-gpt-4o-2024-05-13-86786922",
"dataset:allenai/c4",
"license:apache-2.0",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] | 2024-06-06T07:07:59Z |
2024-06-06T07:08:03+00:00
| 6 | 0 |
---
datasets:
- fine-tuned/before-finetuning-32000-384-gpt-4o-2024-05-13-86786922
- allenai/c4
language:
- en
- en
license: apache-2.0
pipeline_tag: feature-extraction
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- mteb
---
This model is a fine-tuned version of [**BAAI/bge-small-en-v1.5**](https://huggingface.co/BAAI/bge-small-en-v1.5) designed for the following use case:
None
## How to Use
This model can be easily integrated into your NLP pipeline for tasks such as text classification, sentiment analysis, entity recognition, and more. Here's a simple example to get you started:
```python
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import cos_sim
model = SentenceTransformer(
'fine-tuned/before-finetuning-32000-384-gpt-4o-2024-05-13-86786922',
trust_remote_code=True
)
embeddings = model.encode([
'first text to embed',
'second text to embed'
])
print(cos_sim(embeddings[0], embeddings[1]))
```
| null |
Non_BioNLP
|
This model is a fine-tuned version of [**BAAI/bge-small-en-v1.5**](https://huggingface.co/BAAI/bge-small-en-v1.5) designed for the following use case:
None
## How to Use
This model can be easily integrated into your NLP pipeline for tasks such as text classification, sentiment analysis, entity recognition, and more. Here's a simple example to get you started:
```python
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import cos_sim
model = SentenceTransformer(
'fine-tuned/before-finetuning-32000-384-gpt-4o-2024-05-13-86786922',
trust_remote_code=True
)
embeddings = model.encode([
'first text to embed',
'second text to embed'
])
print(cos_sim(embeddings[0], embeddings[1]))
```
|
{"datasets": ["fine-tuned/before-finetuning-32000-384-gpt-4o-2024-05-13-86786922", "allenai/c4"], "language": ["en", "en"], "license": "apache-2.0", "pipeline_tag": "feature-extraction", "tags": ["sentence-transformers", "feature-extraction", "sentence-similarity", "mteb"]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,873 |
projecte-aina/FLOR-6.3B
|
projecte-aina
|
text-generation
|
[
"transformers",
"safetensors",
"bloom",
"text-generation",
"FLOR",
"spanish",
"catalan",
"english",
"en",
"es",
"ca",
"dataset:projecte-aina/CATalog",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-12-15T13:31:04Z |
2025-02-27T09:50:00+00:00
| 102 | 32 |
---
datasets:
- projecte-aina/CATalog
language:
- en
- es
- ca
pipeline_tag: text-generation
tags:
- FLOR
- bloom
- spanish
- catalan
- english
licence:
- apache-2.0
widget:
- text: 'Respon a la pregunta següent.
Pregunta: "Quina és la capital de Suècia?"
Resposta: "La capital de Suècia és Estocolm."
----
Respon a la pregunta següent.
Pregunta: "Quina beguda es consumeix als matins per despertar-se?"
Resposta: "La majoria de gent consumeix cafè per despertar-se."
----
Respon a la pregunta següent.
Pregunta: "Explica com funciona un motor de combustió"
Resposta:'
example_title: Pregunta-Resposta
- text: 'Extrae las entidades nombradas del siguiente texto:
Texto: "Me llamo Wolfgang y vivo en Berlin"
Entidades: Wolfgang:PER, Berlin:LOC
----
Extrae las entidades nombradas del siguiente texto:
Texto: "Hoy voy a visitar el parc güell tras salir del barcelona supercomputing
center"
Entidades: parc güell:LOC, barcelona supercomputing center:LOC
----
Extrae las entidades nombradas del siguiente texto:
Texto: "Maria y Miguel no tienen ningún problema contigo"
Entidades: Maria:PER, Miguel:PER
----
Extrae las entidades nombradas del siguiente texto:
Texto: "Damián se cortó el pelo"
Entidades: Damián:PER
----
Extrae las entidades nombradas del siguiente texto:
Texto: "Lo mejor de Barcelona és el bar de mi amigo Pablo"
Entidades: Pablo:PER, Barcelona:LOC
----
Extrae las entidades nombradas del siguiente texto:
Texto: "Carlos comparte piso con Marc"
Entidades:'
example_title: Entidades-Nombradas
---
# FLOR-6.3B
<strong><span style="color:darkred">⚠️NOTICE⚠️: This model has been deprecated and is no longer actively maintained or supported. To access the latest models with enhanced features, better performance,
and ongoing support, please visit <a style="color:darkred" href="https://huggingface.co/BSC-LT">https://huggingface.co/BSC-LT</a></span></strong>
## Table of Contents
<details>
<summary>Click to expand</summary>
- [Model description](#model-description)
- [Intended uses and limitations](#intended-uses-and-limitations)
- [How to use](#how-to-use)
- [Limitations and bias](#limitations-and-bias)
- [Training](#training)
- [Evaluation](#evaluation)
- [Additional information](#additional-information)
</details>
## Model description
**FLOR-6.3B** is a 6.3B-parameter transformer-based causal language model for Catalan, Spanish, and English.
It is the result of a language adaptation technique performed on [BLOOM-7.1B](https://huggingface.co/bigscience/bloom-7b1),
which involves modifying the model's vocabulary and embedding layer, and continuously pre-training the model with 140B tokens in our target languages.
For more details, take a look at [this blogpost](https://medium.com/@mpamies247/flor-6-3b-a-chinchilla-compliant-model-for-catalan-spanish-and-english-7cdb389a9aac) about the project.
## Intended uses and limitations
The **FLOR-6.3B** model is ready-to-use only for causal language modeling.
It can perform text-generation tasks and be fine-tuned for specific scenarios.
## How to use
```python
import torch
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
input_text = "Sovint em trobo pensant en tot allò que"
model_id = "projecte-aina/FLOR-6.3B"
tokenizer = AutoTokenizer.from_pretrained(model_id)
generator = pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
torch_dtype=torch.bfloat16,
trust_remote_code=True,
device_map="auto",
)
generation = generator(
input_text,
do_sample=True,
top_k=10,
eos_token_id=tokenizer.eos_token_id,
)
print(f"Result: {generation[0]['generated_text']}")
```
## Limitations and bias
At the time of submission, no measures have been taken to estimate the bias and toxicity embedded in the model.
However, we are well aware that our models may be biased since the corpora have been collected using crawling techniques
on multiple web sources. We intend to conduct research in these areas in the future, and if completed, this model card will be updated.
## Training
### Language adaptation and training
The language adaptation technique used to create FLOR-6.3B requires the vocabulary of the source model
to be adapted before continuing its pre-training with data in the target languages. Specifically, we proceeded as follows:
1) We trained our own BPE tokenizer for Catalan, Spanish, and English, and replaced the original BLOOM tokenizer and vocabulary with it. This procedure implied a downsizing of the original BLOOM's embedding layer and, therefore, a model compression from 7.1B parameters to 6.3B.
2) The embeddings corresponding to tokens that are present in both the original and the target vocabulary (matching tokens) were used for initialization.
3) The embeddings from tokens not present in BLOOM's original vocabulary were initialized as the average of all embeddings.
4) The model was initialized with the weights from BLOOM-7.1B, and with our adapted tokenizer (step 1) and embeddings (steps 2-3).
5) The model was then trained on a corpus that contains a mixture of Catalan, Spanish, and English data.
### Training data
The training corpus is composed of 140B tokens gathered from web crawlings and public domain data. Most of the sources in Catalan have been obtained from the [CATalog 1.0](https://huggingface.co/datasets/projecte-aina/CATalog) dataset, filtered with a minimum threshold of 0.6 and oversampling some of the sources it integrates to different extents.
Dataset | Language | Words (per-epoch) | Epochs | Total Tokens |
|---------------------|----------|--------------------|--------------|--------------|
mc4 | ca | 5,861.79M | 1.5 | 13,452.81M |
MaCoCu | ca | 1,658.89M | 2 | 5,076.21M |
CaWac | ca | 1,286.83M | 2.5 | 4,922.14M |
oscar-2301 | ca | 1,784.57M | 1.75 | 4,778.17M |
RacoCatala Articles | ca | 358.57M | 4 | 2,194.42M |
RacoCatala Forums | ca | 1,301.12M | 1 | 1,990.71M |
Tesis (TDX) | ca | 323.60M | 4 | 1,980.46M |
oscar-2201 | ca | 1,155.35M | 1 | 1,767.69M |
Wikipedia | ca | 266.69M | 4 | 1,632.17M |
Nació Digital | ca | 216.27M | 4 | 1,323.59M |
colossal-oscar-05-06-23 | ca | 207.59M | 4 | 1,270.43M |
colossal-oscar-03-04-23 | ca | 195.43M | 4 | 1,196.01M |
colossal-oscar-2022-27 | ca | 195.03M | 4 | 1,193.59M |
Crawling populars | ca | 683.25M | 1 | 1,045.38M |
El Món | ca | 85.27M | 4 | 521.85M |
ACN | ca | 81.25M | 4 | 497.22M |
DOGV | ca | 76.48M | 4 | 468.05M |
DOGC | ca | 70.51M | 4 | 431.51M |
Vilaweb | ca | 46.90M | 4 | 287.04M |
hplt | ca | 160.27M | 1 | 245.21M |
Les Corts Valencianes | ca | 26.88M | 4 | 164.53M |
IB3 | ca | 15.82M | 4 | 96.82M |
BOUA | ca | 13.42M | 4 | 82.13M |
Parlament | ca | 10.09M | 4 | 61.77M |
Aquí Berguedà | ca | 8.23M | 4 | 50.34M |
Wikimedia | ca | 3.90M | 4 | 23.88M |
Gutenberg | ca | 1.29M | 4 | 7.87M |
OSCAR 23.01 | es | 53,244.56M | 0.303 | 23,070.34M |
colossal_oscar_05-06-23 | es | 5,548.27M | 1 | 7,934.02M |
colossal_oscar_03-04-23 | es | 5,090.46M | 1 | 7,279.36M |
All_bio_corpora | es | 954.85M | 2 | 2,730.88M |
Wikipedia | es | 777.49M | 2 | 2,223.63M |
BOE | es | 1,031.28M | 1 | 1,474.73M |
Tesis (TDX) | es | 268.66M | 2 | 768.37M |
Eurlex | es | 459.19M | 1 | 656.64M |
CSIC | es | 156.76M | 2 | 448.33M |
BORME | es | 63.23M | 1 | 90.42M |
colossal_oscar_05-06-23 | en | 51,615.35M | 0.25 | 21,162.30M |
colossal_oscar_03-04-23 | en | 49,454.01M | 0.14 | 11,354.64M |
Wikipedia | en | 2,116.53M | 2 | 6,942.23M |
Gutenberg | en | 3,513.82M | 1 | 5,762.66M |
Eurlex | en | 438.92M | 1 | 719.83M |
legal-mc4 | en | 417.97M | 1 | 685.47M |
### Languages
The training data has the same amount of Catalan, Spanish, and English texts.
The table below shows the final language distribution:
|Language|Percentage|
|--------|----------|
| Catalan (CA) | 33.39% |
| Spanish (ES) | 33.32% |
| English (EN) | 33.29% |
### Framework
The training was conducted in 16 Cerebras' [CS-2 systems](https://www.cerebras.net/product-system/)
using the [cs-2.0.2](https://github.com/Cerebras/modelzoo/releases/tag/Release_2.0.2) release of their software.
## Evaluation
FLOR-6.3B has been evaluated in a 5-shot setting, using EleutherAI's *LM Evaluation Harness*.
The evaluation benchmark includes tasks in Catalan, Spanish, and English, with particular emphasis on Catalan datasets.
The tasks were chosen to cover several evaluation areas in order to provide a comprehensive overview of the model's capabilities.
The baselines used to compare our results are multilingual and English open-source 7B models and smaller models of the FLOR family of models: **TBC**.
Our implementation of EleutherAI's *LM Evaluation Harness* can be found [here](https://github.com/projecte-aina/flor_language_adaptation/tree/main/evaluation).
The following is a list of evaluation areas and their respective datasets:
- Reading Comprehension: [Belebele](https://huggingface.co/datasets/facebook/belebele)
- Question Answering: [XQuAD](https://huggingface.co/datasets/xquad), [CatalanQA](https://huggingface.co/datasets/projecte-aina/catalanqa), [CoQCat](https://huggingface.co/datasets/projecte-aina/CoQCat)
- Natural Language Inference: [XNLI](https://huggingface.co/datasets/xnli) and its translation to Catalan ([XNLI-ca](https://huggingface.co/datasets/projecte-aina/xnli-ca)), [TE-ca](https://huggingface.co/datasets/projecte-aina/teca)
- Paraphrase Identification: [PAWS-X](https://huggingface.co/datasets/paws-x) and its translation to Catalan ([PAWS-ca](https://huggingface.co/datasets/projecte-aina/PAWS-ca)), [Parafraseja](https://huggingface.co/datasets/projecte-aina/Parafraseja)
- Commonsense Reasoning: [COPA](https://people.ict.usc.edu/~gordon/copa.html) and its translation to Catalan ([COPA-ca](https://huggingface.co/datasets/projecte-aina/COPA-ca))
- Translation: [Flores-200](https://huggingface.co/datasets/Muennighoff/flores200)
### Results
| Dataset | Lang. | Task | FLOR-6.3B | BLOOM-7.1B |
|-------------|--------|----------------------------|-------------|-------------|
| Teca | ca | Natural Language Inference | **49.79**🔥 | 46.91 |
| XNLI | ca | Natural Language Inference | **51.70**🔥 | 49.20 |
| XNLI | es | Natural Language Inference | **50.28**🔥 | 47.62 |
| XNLI | en | Natural Language Inference | **52.55**🔥 | 51.96 |
| Belebele | ca | Reading Comprehension | **48.98**🔥 | 48.57 |
| Belebele | es | Reading Comprehension | **48.16** | **48.16** |
| Belebele | en | Reading Comprehension | 49.80 | **50.20**🔥 |
| CatalanQA | ca | Question Answering | **71.80**🔥 | 69.54 |
| CoQCat | ca | Question Answering | **65.96**🔥 | 58.49 |
| XQuAD | ca | Question Answering | 59.01 | **60.94**🔥 |
| XQuAD | es | Question Answering | **63.80**🔥 | 61.76 |
| XQuAD | en | Question Answering | **70.02**🔥 | 69.76 |
| COPA | ca | Question Answering | **78.00**🔥 | 72.60 |
| COPA | en | Question Answering | **81.00**🔥 | 79.00 |
| XStoryCloze | es | Question Answering | **69.82**🔥 | 66.45 |
| XStoryCloze | en | Question Answering | **74.45**🔥 | 70.81 |
| Parafraseja | ca | Paraphrase Identification | **62.88**🔥 | 60.27 |
| PAWS-X | ca | Paraphrase Identification | **59.70**🔥 | 59.35 |
| PAWS-X | es | Paraphrase Identification | 57.70 | **58.65**🔥 |
| PAWS-X | en | Paraphrase Identification | 59.65 | **62.85**🔥 |
| FLoRes | ca->es | Machine Translation | **24.98**🔥 | 24.21 |
| FLoRes | es->ca | Machine Translation | **25.24**🔥 | 23.19 |
| FLoRes | ca->en | Machine Translation | **42.89**🔥 | 40.93 |
| FLoRes | en->ca | Machine Translation | **39.29**🔥 | 34.30 |
| FLoRes | es->en | Machine Translation | **28.61**🔥 | 27.48 |
| FLoRes | en->es | Machine Translation | **25.35**🔥 | 23.72 |
Note: The metrics are F1-score for question-answering tasks, BLEU for translation, and accuracy for the rest.
## Additional information
### Author
The Language Technologies Unit from Barcelona Supercomputing Center.
### Contact
For further information, please send an email to <[email protected]>.
### Copyright
Copyright(c) 2023 by Language Technologies Unit, Barcelona Supercomputing Center.
### License
[Apache License, Version 2.0](https://www.apache.org/licenses/LICENSE-2.0)
### Funding
This work was funded by [Departament de la Vicepresidència i de Polítiques Digitals i Territori de la Generalitat de Catalunya](https://politiquesdigitals.gencat.cat/ca/inici/index.html#googtrans(ca|en) within the framework of [Projecte AINA](https://politiquesdigitals.gencat.cat/ca/economia/catalonia-ai/aina).
### Disclaimer
<details>
<summary>Click to expand</summary>
The model published in this repository is intended for a generalist purpose and is available to third parties under a permissive Apache License, Version 2.0.
Be aware that the model may have biases and/or any other undesirable distortions.
When third parties deploy or provide systems and/or services to other parties using this model (or any system based on it)
or become users of the model, they should note that it is their responsibility to mitigate the risks arising from its use and,
in any event, to comply with applicable regulations, including regulations regarding the use of Artificial Intelligence.
In no event shall the owner and creator of the model (Barcelona Supercomputing Center)
be liable for any results arising from the use made by third parties.
</details>
| null |
Non_BioNLP
|
# FLOR-6.3B
<strong><span style="color:darkred">⚠️NOTICE⚠️: This model has been deprecated and is no longer actively maintained or supported. To access the latest models with enhanced features, better performance,
and ongoing support, please visit <a style="color:darkred" href="https://huggingface.co/BSC-LT">https://huggingface.co/BSC-LT</a></span></strong>
## Table of Contents
<details>
<summary>Click to expand</summary>
- [Model description](#model-description)
- [Intended uses and limitations](#intended-uses-and-limitations)
- [How to use](#how-to-use)
- [Limitations and bias](#limitations-and-bias)
- [Training](#training)
- [Evaluation](#evaluation)
- [Additional information](#additional-information)
</details>
## Model description
**FLOR-6.3B** is a 6.3B-parameter transformer-based causal language model for Catalan, Spanish, and English.
It is the result of a language adaptation technique performed on [BLOOM-7.1B](https://huggingface.co/bigscience/bloom-7b1),
which involves modifying the model's vocabulary and embedding layer, and continuously pre-training the model with 140B tokens in our target languages.
For more details, take a look at [this blogpost](https://medium.com/@mpamies247/flor-6-3b-a-chinchilla-compliant-model-for-catalan-spanish-and-english-7cdb389a9aac) about the project.
## Intended uses and limitations
The **FLOR-6.3B** model is ready-to-use only for causal language modeling.
It can perform text-generation tasks and be fine-tuned for specific scenarios.
## How to use
```python
import torch
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
input_text = "Sovint em trobo pensant en tot allò que"
model_id = "projecte-aina/FLOR-6.3B"
tokenizer = AutoTokenizer.from_pretrained(model_id)
generator = pipeline(
"text-generation",
model=model_id,
tokenizer=tokenizer,
torch_dtype=torch.bfloat16,
trust_remote_code=True,
device_map="auto",
)
generation = generator(
input_text,
do_sample=True,
top_k=10,
eos_token_id=tokenizer.eos_token_id,
)
print(f"Result: {generation[0]['generated_text']}")
```
## Limitations and bias
At the time of submission, no measures have been taken to estimate the bias and toxicity embedded in the model.
However, we are well aware that our models may be biased since the corpora have been collected using crawling techniques
on multiple web sources. We intend to conduct research in these areas in the future, and if completed, this model card will be updated.
## Training
### Language adaptation and training
The language adaptation technique used to create FLOR-6.3B requires the vocabulary of the source model
to be adapted before continuing its pre-training with data in the target languages. Specifically, we proceeded as follows:
1) We trained our own BPE tokenizer for Catalan, Spanish, and English, and replaced the original BLOOM tokenizer and vocabulary with it. This procedure implied a downsizing of the original BLOOM's embedding layer and, therefore, a model compression from 7.1B parameters to 6.3B.
2) The embeddings corresponding to tokens that are present in both the original and the target vocabulary (matching tokens) were used for initialization.
3) The embeddings from tokens not present in BLOOM's original vocabulary were initialized as the average of all embeddings.
4) The model was initialized with the weights from BLOOM-7.1B, and with our adapted tokenizer (step 1) and embeddings (steps 2-3).
5) The model was then trained on a corpus that contains a mixture of Catalan, Spanish, and English data.
### Training data
The training corpus is composed of 140B tokens gathered from web crawlings and public domain data. Most of the sources in Catalan have been obtained from the [CATalog 1.0](https://huggingface.co/datasets/projecte-aina/CATalog) dataset, filtered with a minimum threshold of 0.6 and oversampling some of the sources it integrates to different extents.
Dataset | Language | Words (per-epoch) | Epochs | Total Tokens |
|---------------------|----------|--------------------|--------------|--------------|
mc4 | ca | 5,861.79M | 1.5 | 13,452.81M |
MaCoCu | ca | 1,658.89M | 2 | 5,076.21M |
CaWac | ca | 1,286.83M | 2.5 | 4,922.14M |
oscar-2301 | ca | 1,784.57M | 1.75 | 4,778.17M |
RacoCatala Articles | ca | 358.57M | 4 | 2,194.42M |
RacoCatala Forums | ca | 1,301.12M | 1 | 1,990.71M |
Tesis (TDX) | ca | 323.60M | 4 | 1,980.46M |
oscar-2201 | ca | 1,155.35M | 1 | 1,767.69M |
Wikipedia | ca | 266.69M | 4 | 1,632.17M |
Nació Digital | ca | 216.27M | 4 | 1,323.59M |
colossal-oscar-05-06-23 | ca | 207.59M | 4 | 1,270.43M |
colossal-oscar-03-04-23 | ca | 195.43M | 4 | 1,196.01M |
colossal-oscar-2022-27 | ca | 195.03M | 4 | 1,193.59M |
Crawling populars | ca | 683.25M | 1 | 1,045.38M |
El Món | ca | 85.27M | 4 | 521.85M |
ACN | ca | 81.25M | 4 | 497.22M |
DOGV | ca | 76.48M | 4 | 468.05M |
DOGC | ca | 70.51M | 4 | 431.51M |
Vilaweb | ca | 46.90M | 4 | 287.04M |
hplt | ca | 160.27M | 1 | 245.21M |
Les Corts Valencianes | ca | 26.88M | 4 | 164.53M |
IB3 | ca | 15.82M | 4 | 96.82M |
BOUA | ca | 13.42M | 4 | 82.13M |
Parlament | ca | 10.09M | 4 | 61.77M |
Aquí Berguedà | ca | 8.23M | 4 | 50.34M |
Wikimedia | ca | 3.90M | 4 | 23.88M |
Gutenberg | ca | 1.29M | 4 | 7.87M |
OSCAR 23.01 | es | 53,244.56M | 0.303 | 23,070.34M |
colossal_oscar_05-06-23 | es | 5,548.27M | 1 | 7,934.02M |
colossal_oscar_03-04-23 | es | 5,090.46M | 1 | 7,279.36M |
All_bio_corpora | es | 954.85M | 2 | 2,730.88M |
Wikipedia | es | 777.49M | 2 | 2,223.63M |
BOE | es | 1,031.28M | 1 | 1,474.73M |
Tesis (TDX) | es | 268.66M | 2 | 768.37M |
Eurlex | es | 459.19M | 1 | 656.64M |
CSIC | es | 156.76M | 2 | 448.33M |
BORME | es | 63.23M | 1 | 90.42M |
colossal_oscar_05-06-23 | en | 51,615.35M | 0.25 | 21,162.30M |
colossal_oscar_03-04-23 | en | 49,454.01M | 0.14 | 11,354.64M |
Wikipedia | en | 2,116.53M | 2 | 6,942.23M |
Gutenberg | en | 3,513.82M | 1 | 5,762.66M |
Eurlex | en | 438.92M | 1 | 719.83M |
legal-mc4 | en | 417.97M | 1 | 685.47M |
### Languages
The training data has the same amount of Catalan, Spanish, and English texts.
The table below shows the final language distribution:
|Language|Percentage|
|--------|----------|
| Catalan (CA) | 33.39% |
| Spanish (ES) | 33.32% |
| English (EN) | 33.29% |
### Framework
The training was conducted in 16 Cerebras' [CS-2 systems](https://www.cerebras.net/product-system/)
using the [cs-2.0.2](https://github.com/Cerebras/modelzoo/releases/tag/Release_2.0.2) release of their software.
## Evaluation
FLOR-6.3B has been evaluated in a 5-shot setting, using EleutherAI's *LM Evaluation Harness*.
The evaluation benchmark includes tasks in Catalan, Spanish, and English, with particular emphasis on Catalan datasets.
The tasks were chosen to cover several evaluation areas in order to provide a comprehensive overview of the model's capabilities.
The baselines used to compare our results are multilingual and English open-source 7B models and smaller models of the FLOR family of models: **TBC**.
Our implementation of EleutherAI's *LM Evaluation Harness* can be found [here](https://github.com/projecte-aina/flor_language_adaptation/tree/main/evaluation).
The following is a list of evaluation areas and their respective datasets:
- Reading Comprehension: [Belebele](https://huggingface.co/datasets/facebook/belebele)
- Question Answering: [XQuAD](https://huggingface.co/datasets/xquad), [CatalanQA](https://huggingface.co/datasets/projecte-aina/catalanqa), [CoQCat](https://huggingface.co/datasets/projecte-aina/CoQCat)
- Natural Language Inference: [XNLI](https://huggingface.co/datasets/xnli) and its translation to Catalan ([XNLI-ca](https://huggingface.co/datasets/projecte-aina/xnli-ca)), [TE-ca](https://huggingface.co/datasets/projecte-aina/teca)
- Paraphrase Identification: [PAWS-X](https://huggingface.co/datasets/paws-x) and its translation to Catalan ([PAWS-ca](https://huggingface.co/datasets/projecte-aina/PAWS-ca)), [Parafraseja](https://huggingface.co/datasets/projecte-aina/Parafraseja)
- Commonsense Reasoning: [COPA](https://people.ict.usc.edu/~gordon/copa.html) and its translation to Catalan ([COPA-ca](https://huggingface.co/datasets/projecte-aina/COPA-ca))
- Translation: [Flores-200](https://huggingface.co/datasets/Muennighoff/flores200)
### Results
| Dataset | Lang. | Task | FLOR-6.3B | BLOOM-7.1B |
|-------------|--------|----------------------------|-------------|-------------|
| Teca | ca | Natural Language Inference | **49.79**🔥 | 46.91 |
| XNLI | ca | Natural Language Inference | **51.70**🔥 | 49.20 |
| XNLI | es | Natural Language Inference | **50.28**🔥 | 47.62 |
| XNLI | en | Natural Language Inference | **52.55**🔥 | 51.96 |
| Belebele | ca | Reading Comprehension | **48.98**🔥 | 48.57 |
| Belebele | es | Reading Comprehension | **48.16** | **48.16** |
| Belebele | en | Reading Comprehension | 49.80 | **50.20**🔥 |
| CatalanQA | ca | Question Answering | **71.80**🔥 | 69.54 |
| CoQCat | ca | Question Answering | **65.96**🔥 | 58.49 |
| XQuAD | ca | Question Answering | 59.01 | **60.94**🔥 |
| XQuAD | es | Question Answering | **63.80**🔥 | 61.76 |
| XQuAD | en | Question Answering | **70.02**🔥 | 69.76 |
| COPA | ca | Question Answering | **78.00**🔥 | 72.60 |
| COPA | en | Question Answering | **81.00**🔥 | 79.00 |
| XStoryCloze | es | Question Answering | **69.82**🔥 | 66.45 |
| XStoryCloze | en | Question Answering | **74.45**🔥 | 70.81 |
| Parafraseja | ca | Paraphrase Identification | **62.88**🔥 | 60.27 |
| PAWS-X | ca | Paraphrase Identification | **59.70**🔥 | 59.35 |
| PAWS-X | es | Paraphrase Identification | 57.70 | **58.65**🔥 |
| PAWS-X | en | Paraphrase Identification | 59.65 | **62.85**🔥 |
| FLoRes | ca->es | Machine Translation | **24.98**🔥 | 24.21 |
| FLoRes | es->ca | Machine Translation | **25.24**🔥 | 23.19 |
| FLoRes | ca->en | Machine Translation | **42.89**🔥 | 40.93 |
| FLoRes | en->ca | Machine Translation | **39.29**🔥 | 34.30 |
| FLoRes | es->en | Machine Translation | **28.61**🔥 | 27.48 |
| FLoRes | en->es | Machine Translation | **25.35**🔥 | 23.72 |
Note: The metrics are F1-score for question-answering tasks, BLEU for translation, and accuracy for the rest.
## Additional information
### Author
The Language Technologies Unit from Barcelona Supercomputing Center.
### Contact
For further information, please send an email to <[email protected]>.
### Copyright
Copyright(c) 2023 by Language Technologies Unit, Barcelona Supercomputing Center.
### License
[Apache License, Version 2.0](https://www.apache.org/licenses/LICENSE-2.0)
### Funding
This work was funded by [Departament de la Vicepresidència i de Polítiques Digitals i Territori de la Generalitat de Catalunya](https://politiquesdigitals.gencat.cat/ca/inici/index.html#googtrans(ca|en) within the framework of [Projecte AINA](https://politiquesdigitals.gencat.cat/ca/economia/catalonia-ai/aina).
### Disclaimer
<details>
<summary>Click to expand</summary>
The model published in this repository is intended for a generalist purpose and is available to third parties under a permissive Apache License, Version 2.0.
Be aware that the model may have biases and/or any other undesirable distortions.
When third parties deploy or provide systems and/or services to other parties using this model (or any system based on it)
or become users of the model, they should note that it is their responsibility to mitigate the risks arising from its use and,
in any event, to comply with applicable regulations, including regulations regarding the use of Artificial Intelligence.
In no event shall the owner and creator of the model (Barcelona Supercomputing Center)
be liable for any results arising from the use made by third parties.
</details>
|
{"datasets": ["projecte-aina/CATalog"], "language": ["en", "es", "ca"], "pipeline_tag": "text-generation", "tags": ["FLOR", "bloom", "spanish", "catalan", "english"], "licence": ["apache-2.0"], "widget": [{"text": "Respon a la pregunta següent.\nPregunta: \"Quina és la capital de Suècia?\"\nResposta: \"La capital de Suècia és Estocolm.\"\n----\nRespon a la pregunta següent.\nPregunta: \"Quina beguda es consumeix als matins per despertar-se?\"\nResposta: \"La majoria de gent consumeix cafè per despertar-se.\"\n----\nRespon a la pregunta següent.\nPregunta: \"Explica com funciona un motor de combustió\"\nResposta:", "example_title": "Pregunta-Resposta"}, {"text": "Extrae las entidades nombradas del siguiente texto:\nTexto: \"Me llamo Wolfgang y vivo en Berlin\"\nEntidades: Wolfgang:PER, Berlin:LOC\n----\nExtrae las entidades nombradas del siguiente texto:\nTexto: \"Hoy voy a visitar el parc güell tras salir del barcelona supercomputing center\"\nEntidades: parc güell:LOC, barcelona supercomputing center:LOC\n----\nExtrae las entidades nombradas del siguiente texto:\nTexto: \"Maria y Miguel no tienen ningún problema contigo\"\nEntidades: Maria:PER, Miguel:PER\n----\nExtrae las entidades nombradas del siguiente texto:\nTexto: \"Damián se cortó el pelo\"\nEntidades: Damián:PER\n----\nExtrae las entidades nombradas del siguiente texto:\nTexto: \"Lo mejor de Barcelona és el bar de mi amigo Pablo\"\nEntidades: Pablo:PER, Barcelona:LOC\n----\nExtrae las entidades nombradas del siguiente texto:\nTexto: \"Carlos comparte piso con Marc\"\nEntidades:", "example_title": "Entidades-Nombradas"}]}
|
task
|
[
"QUESTION_ANSWERING",
"TRANSLATION"
] | 39,874 |
yongsun-yoon/mdeberta-v3-base-open-ner
|
yongsun-yoon
|
token-classification
|
[
"transformers",
"pytorch",
"deberta-v2",
"token-classification",
"en",
"ko",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-08-20T03:46:14Z |
2023-08-20T22:32:44+00:00
| 14 | 2 |
---
language:
- en
- ko
---
# Open Named Entity Recognition (Multilingual)
You can read the details in [this post](https://medium.com/@yongsun.yoon/cross-encoder-for-open-named-entity-recognition-4a7d485c37cc).
```python
from transformers import pipeline
nlp = pipeline('token-classification', 'yongsun-yoon/deberta-v3-base-open-ner', aggregation_strategy='simple')
text = '30일 전주지법 군산지원 제1형사부는 박 시장에 대한 선고공판에서 벌금 500만원을 선고했다.'
entity_type = 'event'
input_text = f'{text}{nlp.tokenizer.sep_token}{entity_type}' # [SEP]
nlp(input_text) # 선고공판
```
| null |
Non_BioNLP
|
# Open Named Entity Recognition (Multilingual)
You can read the details in [this post](https://medium.com/@yongsun.yoon/cross-encoder-for-open-named-entity-recognition-4a7d485c37cc).
```python
from transformers import pipeline
nlp = pipeline('token-classification', 'yongsun-yoon/deberta-v3-base-open-ner', aggregation_strategy='simple')
text = '30일 전주지법 군산지원 제1형사부는 박 시장에 대한 선고공판에서 벌금 500만원을 선고했다.'
entity_type = 'event'
input_text = f'{text}{nlp.tokenizer.sep_token}{entity_type}' # [SEP]
nlp(input_text) # 선고공판
```
|
{"language": ["en", "ko"]}
|
task
|
[
"NAMED_ENTITY_RECOGNITION"
] | 39,875 |
mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF
|
mradermacher
| null |
[
"transformers",
"gguf",
"German",
"RAG",
"Retrieval",
"Question-Answering",
"Summarization",
"Reasoning",
"en",
"de",
"dataset:avemio/German-RAG-CPT-HESSIAN-AI",
"dataset:avemio/German-RAG-SFT-ShareGPT-HESSIAN-AI",
"dataset:avemio/German-RAG-ORPO-ShareGPT-HESSIAN-AI",
"dataset:VAGOsolutions/SauerkrautLM-Fermented-GER-DPO",
"dataset:VAGOsolutions/SauerkrautLM-Fermented-Irrelevance-GER-DPO",
"base_model:avemio/German-RAG-NEMO-12B-ORPO-HESSIAN-AI",
"base_model:quantized:avemio/German-RAG-NEMO-12B-ORPO-HESSIAN-AI",
"license:apache-2.0",
"endpoints_compatible",
"region:us",
"conversational"
] | 2025-02-09T21:47:05Z |
2025-02-09T22:01:06+00:00
| 994 | 0 |
---
base_model: avemio/German-RAG-NEMO-12B-ORPO-HESSIAN-AI
datasets:
- avemio/German-RAG-CPT-HESSIAN-AI
- avemio/German-RAG-SFT-ShareGPT-HESSIAN-AI
- avemio/German-RAG-ORPO-ShareGPT-HESSIAN-AI
- VAGOsolutions/SauerkrautLM-Fermented-GER-DPO
- VAGOsolutions/SauerkrautLM-Fermented-Irrelevance-GER-DPO
language:
- en
- de
library_name: transformers
license: apache-2.0
tags:
- German
- RAG
- Retrieval
- Question-Answering
- Summarization
- Reasoning
quantized_by: mradermacher
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/avemio/German-RAG-NEMO-12B-ORPO-HESSIAN-AI
<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q2_K.gguf) | Q2_K | 4.9 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q3_K_S.gguf) | Q3_K_S | 5.6 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q3_K_M.gguf) | Q3_K_M | 6.2 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q3_K_L.gguf) | Q3_K_L | 6.7 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.IQ4_XS.gguf) | IQ4_XS | 6.9 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q4_K_S.gguf) | Q4_K_S | 7.2 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q4_K_M.gguf) | Q4_K_M | 7.6 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q5_K_S.gguf) | Q5_K_S | 8.6 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q5_K_M.gguf) | Q5_K_M | 8.8 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q6_K.gguf) | Q6_K | 10.2 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q8_0.gguf) | Q8_0 | 13.1 | fast, best quality |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.
<!-- end -->
| null |
Non_BioNLP
|
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/avemio/German-RAG-NEMO-12B-ORPO-HESSIAN-AI
<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q2_K.gguf) | Q2_K | 4.9 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q3_K_S.gguf) | Q3_K_S | 5.6 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q3_K_M.gguf) | Q3_K_M | 6.2 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q3_K_L.gguf) | Q3_K_L | 6.7 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.IQ4_XS.gguf) | IQ4_XS | 6.9 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q4_K_S.gguf) | Q4_K_S | 7.2 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q4_K_M.gguf) | Q4_K_M | 7.6 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q5_K_S.gguf) | Q5_K_S | 8.6 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q5_K_M.gguf) | Q5_K_M | 8.8 | |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q6_K.gguf) | Q6_K | 10.2 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/German-RAG-NEMO-12B-ORPO-HESSIAN-AI-GGUF/resolve/main/German-RAG-NEMO-12B-ORPO-HESSIAN-AI.Q8_0.gguf) | Q8_0 | 13.1 | fast, best quality |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.
<!-- end -->
|
{"base_model": "avemio/German-RAG-NEMO-12B-ORPO-HESSIAN-AI", "datasets": ["avemio/German-RAG-CPT-HESSIAN-AI", "avemio/German-RAG-SFT-ShareGPT-HESSIAN-AI", "avemio/German-RAG-ORPO-ShareGPT-HESSIAN-AI", "VAGOsolutions/SauerkrautLM-Fermented-GER-DPO", "VAGOsolutions/SauerkrautLM-Fermented-Irrelevance-GER-DPO"], "language": ["en", "de"], "library_name": "transformers", "license": "apache-2.0", "tags": ["German", "RAG", "Retrieval", "Question-Answering", "Summarization", "Reasoning"], "quantized_by": "mradermacher"}
|
task
|
[
"SUMMARIZATION"
] | 39,876 |
sandeep16064/inshorts-news-summary
|
sandeep16064
|
summarization
|
[
"transformers",
"pytorch",
"mt5",
"text2text-generation",
"summarization",
"T2SAM",
"abstractive summarization",
"en",
"dataset:sandeep16064/news_summary",
"base_model:google/mt5-small",
"base_model:finetune:google/mt5-small",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-10-05T07:27:35Z |
2023-10-13T10:19:09+00:00
| 62 | 1 |
---
base_model: google/mt5-small
datasets:
- sandeep16064/news_summary
language:
- en
library_name: transformers
license: apache-2.0
metrics:
- rouge
tags:
- summarization
- T2SAM
- abstractive summarization
model-index:
- name: mt5-small-finetuned-inshorts-news-summary
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# mt5-small-finetuned-inshorts-news-summary
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the [inshorts-news-summary dataset] (https://huggingface.co/datasets/sandeep16064/news_summary).
It achieves the following results on the evaluation set:
- Loss: 1.5399
- Rouge1: 54.613
- Rouge2: 31.1543
- Rougel: 50.7709
- Rougelsum: 50.7907
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5.6e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|:-------:|:---------:|
| 3.3244 | 1.0 | 5511 | 1.8904 | 51.0778 | 28.3112 | 47.4136 | 47.404 |
| 2.2747 | 2.0 | 11022 | 1.7450 | 51.8372 | 28.9814 | 48.0917 | 48.0965 |
| 2.0745 | 3.0 | 16533 | 1.6567 | 52.518 | 29.7276 | 48.727 | 48.7504 |
| 1.9516 | 4.0 | 22044 | 1.6210 | 54.2404 | 30.8927 | 50.4042 | 50.3996 |
| 1.8714 | 5.0 | 27555 | 1.5971 | 53.8556 | 30.6665 | 50.112 | 50.1177 |
| 1.8112 | 6.0 | 33066 | 1.5649 | 54.179 | 31.0178 | 50.407 | 50.4281 |
| 1.7644 | 7.0 | 38577 | 1.5605 | 54.3104 | 30.7997 | 50.4555 | 50.4861 |
| 1.7265 | 8.0 | 44088 | 1.5447 | 54.5593 | 31.0283 | 50.6343 | 50.6605 |
| 1.7013 | 9.0 | 49599 | 1.5440 | 54.7385 | 31.3073 | 50.9111 | 50.9334 |
| 1.6864 | 10.0 | 55110 | 1.5399 | 54.613 | 31.1543 | 50.7709 | 50.7907 |
### Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0
- Datasets 2.12.0
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# mt5-small-finetuned-inshorts-news-summary
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the [inshorts-news-summary dataset] (https://huggingface.co/datasets/sandeep16064/news_summary).
It achieves the following results on the evaluation set:
- Loss: 1.5399
- Rouge1: 54.613
- Rouge2: 31.1543
- Rougel: 50.7709
- Rougelsum: 50.7907
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5.6e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|:-------:|:---------:|
| 3.3244 | 1.0 | 5511 | 1.8904 | 51.0778 | 28.3112 | 47.4136 | 47.404 |
| 2.2747 | 2.0 | 11022 | 1.7450 | 51.8372 | 28.9814 | 48.0917 | 48.0965 |
| 2.0745 | 3.0 | 16533 | 1.6567 | 52.518 | 29.7276 | 48.727 | 48.7504 |
| 1.9516 | 4.0 | 22044 | 1.6210 | 54.2404 | 30.8927 | 50.4042 | 50.3996 |
| 1.8714 | 5.0 | 27555 | 1.5971 | 53.8556 | 30.6665 | 50.112 | 50.1177 |
| 1.8112 | 6.0 | 33066 | 1.5649 | 54.179 | 31.0178 | 50.407 | 50.4281 |
| 1.7644 | 7.0 | 38577 | 1.5605 | 54.3104 | 30.7997 | 50.4555 | 50.4861 |
| 1.7265 | 8.0 | 44088 | 1.5447 | 54.5593 | 31.0283 | 50.6343 | 50.6605 |
| 1.7013 | 9.0 | 49599 | 1.5440 | 54.7385 | 31.3073 | 50.9111 | 50.9334 |
| 1.6864 | 10.0 | 55110 | 1.5399 | 54.613 | 31.1543 | 50.7709 | 50.7907 |
### Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0
- Datasets 2.12.0
- Tokenizers 0.13.3
|
{"base_model": "google/mt5-small", "datasets": ["sandeep16064/news_summary"], "language": ["en"], "library_name": "transformers", "license": "apache-2.0", "metrics": ["rouge"], "tags": ["summarization", "T2SAM", "abstractive summarization"], "model-index": [{"name": "mt5-small-finetuned-inshorts-news-summary", "results": []}]}
|
task
|
[
"SUMMARIZATION"
] | 39,877 |
TroyDoesAI/Mermaid-Llama-3-7B-Pruned
|
TroyDoesAI
|
text-generation
|
[
"transformers",
"safetensors",
"llama",
"text-generation",
"conversational",
"license:cc-by-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2024-04-19T22:17:34Z |
2024-05-03T02:18:42+00:00
| 20 | 1 |
---
license: cc-by-4.0
---
What If I told you I found 5 layers out of the 32 that do absolutely nothing.
Cutting these layers out do not change the output for my model at all individually, but cutting all 5 at once breaks the model completely.
Lets take 4 layers out of 5 out and see if we can recover after removing this redundancy.
Please Give my Mermaid-Llama-3-8B and This Mermaid-Llama-3-Pruned-7B and the rest of the pruned Mermaids.
28/32 Layers Model.
---
license: cc-by-4.0
---
# Mermaid-Llama-3-7B
Introducing Mermaid-LLama-3-7B, a robust language model designed for Python code understanding and crafting captivating story flow maps.
Pruned down to 7 billion parameter.
See MergeKit Notes And Try Triming my model yourself and explore my world of trimming models to fit SMARTER Models with lower requirements f
or specific tasks. Mermaid is just a start, Hire me to solve your problem and I will build the smallest footprint model that solves just that problem.
I wish to specialize in packing models on Edge Devices.
Open For Hire See my links to my Linkedin for more.
## Key Features
1. **Code Understanding:**
- Masters Python intricacies with finesse.
- Generates clear and accurate Mermaid Diagram Flow Charts.
- Ideal for developers seeking visual representations of their code logic.
2. **Storytelling Capabilities:**
- Converts narrative inputs into captivating Mermaid Diagrams.
- Maps character interactions, plot developments, and narrative arcs.
3. **Unmatched Performance:**
- Surpasses GPT-4 in generating well-organized Mermaid Diagrams.
4. **Training Insights:**
- Trained on a diverse dataset, including 800 unique, hand-curated Mermaid Graph examples utilizing 478 complete Python programs.
- Exhibits emergent properties in story-to-flow map translations and step-by-step instruction flow maps.
## Collaboration
Interested in enhancing Mermaid's capabilities? Contact [email protected] for collaboration opportunities.
## Example Use Cases
- **Retrieval-Augmented Generation (RAG):** Utilize Mermaid-LLama-3-8B to create condensed knowledge graphs. This model excels in generating flow diagrams that enhance the retrieval process. These knowledge graphs are stored in a vector database, which allows for quick and efficient retrieval of contextually relevant information. When a query is received, the system retrieves a pertinent knowledge graph, appending it as context to the model. This enriched context enables Mermaid-LLama-3-8B to deliver more accurate and nuanced responses. This approach is particularly beneficial in applications requiring deep, context-aware interactions, such as sophisticated Q&A systems, dynamic data analysis, and complex decision-making tasks.
- **Code Documentation:** Automatic visual flow charts from Python code.
- **Storyboarding:** Visually appealing diagrams for storytelling.
- **Project Planning:** Visual project flow maps for effective team communication.
- **Learning Python:** Helps students visually understand Python code structures.
- **Game Design:** Visualizing game storylines for coherent narrative structure.
## Proof of Concept
Stay tuned for the release of the VSCode Extension that displays the Live Flow Map every time a user stops typing for more than 10 seconds.
## Training Specifications
- **LoRA Rank:** 2048
- **LoRA Alpha:** 4096
- **Batch Size:** 1
- **Micro Batch Size:** 1
- **Cutoff Length:** 4096
- **Save every n steps:** 1000
- **Epochs:** 3
- **Learning Rate:** 1e-6
- **LR Scheduler:** Cosine
**Target Modules:**
- Enable q_proj
- Enable v_proj
- Enable k_proj
- Enable o_proj
- Enable gate_proj
- Enable down_proj
- Enable up_proj
## Getting Started
Start by downloading one of my models.
Load the model.
Use my prompt template to generate a Mermaid code block, which can be viewed in the Mermaid Live Editor or using the Mermaid CLI tool.
Here we open the VLLM GUI Program while still running in Vram the Mermaid-Llama-8B to compare the flow diagram to the actual program and show the lightweight capabilites of small models on consumer hardware.
## More on my VLLM Class and inference GUI : https://github.com/Troys-Code/VLLM
---
Note: This model should be treated as an Auto-Complete Model, Do not try talking to it in chat you are gonna get garbage, those layers have been pruned and replaced, that is all you will hear of my secret sauce on training on small < 1000 entry datasets.
| null |
Non_BioNLP
|
What If I told you I found 5 layers out of the 32 that do absolutely nothing.
Cutting these layers out do not change the output for my model at all individually, but cutting all 5 at once breaks the model completely.
Lets take 4 layers out of 5 out and see if we can recover after removing this redundancy.
Please Give my Mermaid-Llama-3-8B and This Mermaid-Llama-3-Pruned-7B and the rest of the pruned Mermaids.
28/32 Layers Model.
---
license: cc-by-4.0
---
# Mermaid-Llama-3-7B
Introducing Mermaid-LLama-3-7B, a robust language model designed for Python code understanding and crafting captivating story flow maps.
Pruned down to 7 billion parameter.
See MergeKit Notes And Try Triming my model yourself and explore my world of trimming models to fit SMARTER Models with lower requirements f
or specific tasks. Mermaid is just a start, Hire me to solve your problem and I will build the smallest footprint model that solves just that problem.
I wish to specialize in packing models on Edge Devices.
Open For Hire See my links to my Linkedin for more.
## Key Features
1. **Code Understanding:**
- Masters Python intricacies with finesse.
- Generates clear and accurate Mermaid Diagram Flow Charts.
- Ideal for developers seeking visual representations of their code logic.
2. **Storytelling Capabilities:**
- Converts narrative inputs into captivating Mermaid Diagrams.
- Maps character interactions, plot developments, and narrative arcs.
3. **Unmatched Performance:**
- Surpasses GPT-4 in generating well-organized Mermaid Diagrams.
4. **Training Insights:**
- Trained on a diverse dataset, including 800 unique, hand-curated Mermaid Graph examples utilizing 478 complete Python programs.
- Exhibits emergent properties in story-to-flow map translations and step-by-step instruction flow maps.
## Collaboration
Interested in enhancing Mermaid's capabilities? Contact [email protected] for collaboration opportunities.
## Example Use Cases
- **Retrieval-Augmented Generation (RAG):** Utilize Mermaid-LLama-3-8B to create condensed knowledge graphs. This model excels in generating flow diagrams that enhance the retrieval process. These knowledge graphs are stored in a vector database, which allows for quick and efficient retrieval of contextually relevant information. When a query is received, the system retrieves a pertinent knowledge graph, appending it as context to the model. This enriched context enables Mermaid-LLama-3-8B to deliver more accurate and nuanced responses. This approach is particularly beneficial in applications requiring deep, context-aware interactions, such as sophisticated Q&A systems, dynamic data analysis, and complex decision-making tasks.
- **Code Documentation:** Automatic visual flow charts from Python code.
- **Storyboarding:** Visually appealing diagrams for storytelling.
- **Project Planning:** Visual project flow maps for effective team communication.
- **Learning Python:** Helps students visually understand Python code structures.
- **Game Design:** Visualizing game storylines for coherent narrative structure.
## Proof of Concept
Stay tuned for the release of the VSCode Extension that displays the Live Flow Map every time a user stops typing for more than 10 seconds.
## Training Specifications
- **LoRA Rank:** 2048
- **LoRA Alpha:** 4096
- **Batch Size:** 1
- **Micro Batch Size:** 1
- **Cutoff Length:** 4096
- **Save every n steps:** 1000
- **Epochs:** 3
- **Learning Rate:** 1e-6
- **LR Scheduler:** Cosine
**Target Modules:**
- Enable q_proj
- Enable v_proj
- Enable k_proj
- Enable o_proj
- Enable gate_proj
- Enable down_proj
- Enable up_proj
## Getting Started
Start by downloading one of my models.
Load the model.
Use my prompt template to generate a Mermaid code block, which can be viewed in the Mermaid Live Editor or using the Mermaid CLI tool.
Here we open the VLLM GUI Program while still running in Vram the Mermaid-Llama-8B to compare the flow diagram to the actual program and show the lightweight capabilites of small models on consumer hardware.
## More on my VLLM Class and inference GUI : https://github.com/Troys-Code/VLLM
---
Note: This model should be treated as an Auto-Complete Model, Do not try talking to it in chat you are gonna get garbage, those layers have been pruned and replaced, that is all you will hear of my secret sauce on training on small < 1000 entry datasets.
|
{"license": "cc-by-4.0"}
|
task
|
[
"TRANSLATION"
] | 39,878 |
pszemraj/pegasus-x-large-book_synthsumm-bf16
|
pszemraj
|
summarization
|
[
"transformers",
"safetensors",
"pegasus_x",
"text2text-generation",
"generated_from_trainer",
"synthsumm",
"summarization",
"en",
"dataset:pszemraj/synthsumm",
"base_model:pszemraj/pegasus-x-large-book-summary",
"base_model:finetune:pszemraj/pegasus-x-large-book-summary",
"license:bsd-3-clause",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-01-07T16:36:02Z |
2024-01-10T16:54:53+00:00
| 17 | 0 |
---
base_model: pszemraj/pegasus-x-large-book-summary
datasets:
- pszemraj/synthsumm
language:
- en
license: bsd-3-clause
metrics:
- rouge
pipeline_tag: summarization
tags:
- generated_from_trainer
- synthsumm
---
# pegasus-x-large-book_synthsumm - bf16
> this is just a new repo in bf16 (training precision). refer to original repo for details https://huggingface.co/pszemraj/pegasus-x-large-book_synthsumm
Fine-tuned on a synthetic dataset of curated long-context text and `GPT-3.5-turbo-1106` summaries spanning multiple domains + "random" long-context examples from pretraining datasets
Try it: [gradio demo](https://huggingface.co/spaces/pszemraj/document-summarization) | [example outputs .md](evals-outputs/GAUNTLET.md) (gauntlet) | code for free [HF inference api](https://gist.github.com/pszemraj/08f527380ed00ef2f2169e220341c489)
| null |
Non_BioNLP
|
# pegasus-x-large-book_synthsumm - bf16
> this is just a new repo in bf16 (training precision). refer to original repo for details https://huggingface.co/pszemraj/pegasus-x-large-book_synthsumm
Fine-tuned on a synthetic dataset of curated long-context text and `GPT-3.5-turbo-1106` summaries spanning multiple domains + "random" long-context examples from pretraining datasets
Try it: [gradio demo](https://huggingface.co/spaces/pszemraj/document-summarization) | [example outputs .md](evals-outputs/GAUNTLET.md) (gauntlet) | code for free [HF inference api](https://gist.github.com/pszemraj/08f527380ed00ef2f2169e220341c489)
|
{"base_model": "pszemraj/pegasus-x-large-book-summary", "datasets": ["pszemraj/synthsumm"], "language": ["en"], "license": "bsd-3-clause", "metrics": ["rouge"], "pipeline_tag": "summarization", "tags": ["generated_from_trainer", "synthsumm"]}
|
task
|
[
"SUMMARIZATION"
] | 39,879 |
SeoTae/t5-small-custom
|
SeoTae
| null |
[
"region:us"
] | 2024-09-20T04:24:24Z |
2024-09-20T04:24:28+00:00
| 0 | 0 |
---
{}
---
#
# Model Card for t5_small Summarization Model
## Model Details
This model is a t5-small for studing Text Summarization.
## Training Data
The model was trained on the cnn_dailymail dataset.
## Training Procedure
- **Learning Rate** : 2e-5
- **Epochs** : 5
- **Batch Size ** : 4
## How to Use
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("t5-small")
model = AutoModelForSequenceClassification.from_pretrained("t5-small")
input_text = "The movie was fantastic with a gripping storyline!"
inputs = tokenizer.encode(input_text, return_tensors="pt")
outputs = model(inputs)
print(outputs.logits)
```
## Evaluation
- **Accuracy** : i don't know well.
## Limitations
The model may generate biased or inappropriate content
due to the nature of the training data.
It is recommended to use the model with caution and apply necessary filters.
## Ethical Considerations
- **Bias**: The model may inherit biases present in the training data.
- **Misuse**: The model can be misused to generate misleading or harmful content.
| null |
Non_BioNLP
|
#
# Model Card for t5_small Summarization Model
## Model Details
This model is a t5-small for studing Text Summarization.
## Training Data
The model was trained on the cnn_dailymail dataset.
## Training Procedure
- **Learning Rate** : 2e-5
- **Epochs** : 5
- **Batch Size ** : 4
## How to Use
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("t5-small")
model = AutoModelForSequenceClassification.from_pretrained("t5-small")
input_text = "The movie was fantastic with a gripping storyline!"
inputs = tokenizer.encode(input_text, return_tensors="pt")
outputs = model(inputs)
print(outputs.logits)
```
## Evaluation
- **Accuracy** : i don't know well.
## Limitations
The model may generate biased or inappropriate content
due to the nature of the training data.
It is recommended to use the model with caution and apply necessary filters.
## Ethical Considerations
- **Bias**: The model may inherit biases present in the training data.
- **Misuse**: The model can be misused to generate misleading or harmful content.
|
{}
|
task
|
[
"SUMMARIZATION"
] | 39,880 |
YakovElm/IntelDAOS20SetFitModel_balance_ratio_4
|
YakovElm
|
text-classification
|
[
"sentence-transformers",
"pytorch",
"mpnet",
"setfit",
"text-classification",
"arxiv:2209.11055",
"license:apache-2.0",
"region:us"
] | 2023-06-02T07:14:35Z |
2023-06-02T07:15:17+00:00
| 8 | 0 |
---
license: apache-2.0
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
---
# YakovElm/IntelDAOS20SetFitModel_balance_ratio_4
This is a [SetFit model](https://github.com/huggingface/setfit) that can be used for text classification. The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Usage
To use this model for inference, first install the SetFit library:
```bash
python -m pip install setfit
```
You can then run inference as follows:
```python
from setfit import SetFitModel
# Download from Hub and run inference
model = SetFitModel.from_pretrained("YakovElm/IntelDAOS20SetFitModel_balance_ratio_4")
# Run inference
preds = model(["i loved the spiderman movie!", "pineapple on pizza is the worst 🤮"])
```
## BibTeX entry and citation info
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
| null |
Non_BioNLP
|
# YakovElm/IntelDAOS20SetFitModel_balance_ratio_4
This is a [SetFit model](https://github.com/huggingface/setfit) that can be used for text classification. The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Usage
To use this model for inference, first install the SetFit library:
```bash
python -m pip install setfit
```
You can then run inference as follows:
```python
from setfit import SetFitModel
# Download from Hub and run inference
model = SetFitModel.from_pretrained("YakovElm/IntelDAOS20SetFitModel_balance_ratio_4")
# Run inference
preds = model(["i loved the spiderman movie!", "pineapple on pizza is the worst 🤮"])
```
## BibTeX entry and citation info
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
|
{"license": "apache-2.0", "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification"]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,881 |
RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf
|
RichardErkhov
| null |
[
"gguf",
"endpoints_compatible",
"region:us",
"conversational"
] | 2024-09-15T04:15:27Z |
2024-09-15T11:21:20+00:00
| 43 | 0 |
---
{}
---
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
Hercules-Qwen1.5-14B - GGUF
- Model creator: https://huggingface.co/M4-ai/
- Original model: https://huggingface.co/M4-ai/Hercules-Qwen1.5-14B/
| Name | Quant method | Size |
| ---- | ---- | ---- |
| [Hercules-Qwen1.5-14B.Q2_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q2_K.gguf) | Q2_K | 5.51GB |
| [Hercules-Qwen1.5-14B.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ3_XS.gguf) | IQ3_XS | 6.03GB |
| [Hercules-Qwen1.5-14B.IQ3_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ3_S.gguf) | IQ3_S | 6.31GB |
| [Hercules-Qwen1.5-14B.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K_S.gguf) | Q3_K_S | 6.31GB |
| [Hercules-Qwen1.5-14B.IQ3_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ3_M.gguf) | IQ3_M | 6.61GB |
| [Hercules-Qwen1.5-14B.Q3_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K.gguf) | Q3_K | 6.91GB |
| [Hercules-Qwen1.5-14B.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K_M.gguf) | Q3_K_M | 6.91GB |
| [Hercules-Qwen1.5-14B.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K_L.gguf) | Q3_K_L | 7.3GB |
| [Hercules-Qwen1.5-14B.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ4_XS.gguf) | IQ4_XS | 7.37GB |
| [Hercules-Qwen1.5-14B.Q4_0.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_0.gguf) | Q4_0 | 7.62GB |
| [Hercules-Qwen1.5-14B.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ4_NL.gguf) | IQ4_NL | 7.68GB |
| [Hercules-Qwen1.5-14B.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_K_S.gguf) | Q4_K_S | 7.98GB |
| [Hercules-Qwen1.5-14B.Q4_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_K.gguf) | Q4_K | 8.56GB |
| [Hercules-Qwen1.5-14B.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_K_M.gguf) | Q4_K_M | 8.56GB |
| [Hercules-Qwen1.5-14B.Q4_1.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_1.gguf) | Q4_1 | 8.4GB |
| [Hercules-Qwen1.5-14B.Q5_0.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_0.gguf) | Q5_0 | 9.18GB |
| [Hercules-Qwen1.5-14B.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_K_S.gguf) | Q5_K_S | 9.34GB |
| [Hercules-Qwen1.5-14B.Q5_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_K.gguf) | Q5_K | 9.81GB |
| [Hercules-Qwen1.5-14B.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_K_M.gguf) | Q5_K_M | 9.81GB |
| [Hercules-Qwen1.5-14B.Q5_1.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_1.gguf) | Q5_1 | 9.96GB |
| [Hercules-Qwen1.5-14B.Q6_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q6_K.gguf) | Q6_K | 11.46GB |
| [Hercules-Qwen1.5-14B.Q8_0.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q8_0.gguf) | Q8_0 | 14.03GB |
Original model description:
---
library_name: transformers
license: other
datasets:
- Locutusque/hercules-v4.0
language:
- en
---
<style>
body {
font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;
line-height: 1.6;
color: #f5f5f5;
background-color: #1e2a36;
margin: 0;
padding: 0;
}
.container {
max-width: 1200px;
margin: 20px auto;
padding: 20px;
background-color: #2a3f54;
border-radius: 8px;
box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
display: flex;
flex-wrap: wrap;
justify-content: space-between;
}
h1 {
font-size: 2.5rem;
color: #51a3d3;
text-align: center;
margin-bottom: 30px;
width: 100%;
}
h2 {
font-size: 1.75rem;
margin: 20px 0;
color: #63b8ea;
padding-bottom: 10px;
}
h3 {
font-size: 1.25rem;
color: #80c8f4;
}
p, a {
font-size: 1rem;
}
p {
color: #b0c2ce;
margin-bottom: 20px;
}
ul {
list-style-type: none;
padding: 0;
display: flex;
flex-wrap: wrap;
justify-content: space-between;
width: 100%;
}
li {
background-color: #34495e;
padding: 20px;
margin-bottom: 10px;
border-radius: 4px;
cursor: pointer;
transition: background-color 0.3s ease, color 0.3s ease;
overflow: hidden;
color: #b0c2ce;
width: calc(50% - 10px);
box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
}
li:hover {
background-color: #4e6a81;
color: #dfe8f1;
}
.section-content {
margin-top: 15px;
border-top: 1px solid #4e6a81;
padding-top: 10px;
}
a {
color: #a4c8e1;
text-decoration: none;
}
a:hover {
text-decoration: underline;
}
pre {
background-color: #2c3e50;
padding: 10px;
border-radius: 5px;
overflow-x: auto;
color: #b0c2ce;
}
</style>
<div class="container">
<h1>Hercules-Qwen1.5-14B</h1>
</div>
<ul>
<li>
<h2>Model Details</h2>
<div class="section-content">
<h3>Model Description</h3>
<p>This model has capabilities in math, coding, function calling, roleplay, and more. We fine-tuned it using 700,000 examples of Hercules-v4.</p>
<p><strong>Developed by:</strong> M4-ai</p>
<p><strong>Language(s) (NLP):</strong> English and maybe Chinese</p>
<p><strong>License:</strong> tongyi-qianwen license</p>
<p><strong>Finetuned from model:</strong> <a href="https://huggingface.co/Qwen/Qwen1.5-14B">Qwen1.5-14B</a></p>
</div>
</li>
<li>
<h2>Uses</h2>
<div class="section-content">
<p>General purpose assistant, question answering, chain-of-thought, etc..</p>
<h3>Recommendations</h3>
<p>Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.</p>
</div>
</li>
<li>
<h2>Evaluation</h2>
<div class="section-content">
<p>Coming soon</p>
</div>
</li>
<li>
<h2>Training Details</h2>
<div class="section-content">
<h3>Training Data</h3>
<p><a href="https://huggingface.co/datasets/Locutusque/hercules-v4.0">https://huggingface.co/datasets/Locutusque/hercules-v4.0</a></p>
<h4>Training Hyperparameters</h4>
<p><strong>Training regime:</strong> bf16 non-mixed precision</p>
</div>
</li>
<li>
<h2>Technical Specifications</h2>
<div class="section-content">
<h4>Hardware</h4>
<p>We used 8 Kaggle TPUs, and we trained at a global batch size of 128 and sequence length of 1024</p>
</div>
</li>
<li>
<h2>Contributions</h2>
<div class="section-content">
<p>Thanks to @Tonic, @aloobun, @fhai50032, and @Locutusque for their contributions to this model.</p>
</div>
</li>
</ul>
| null |
Non_BioNLP
|
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
Hercules-Qwen1.5-14B - GGUF
- Model creator: https://huggingface.co/M4-ai/
- Original model: https://huggingface.co/M4-ai/Hercules-Qwen1.5-14B/
| Name | Quant method | Size |
| ---- | ---- | ---- |
| [Hercules-Qwen1.5-14B.Q2_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q2_K.gguf) | Q2_K | 5.51GB |
| [Hercules-Qwen1.5-14B.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ3_XS.gguf) | IQ3_XS | 6.03GB |
| [Hercules-Qwen1.5-14B.IQ3_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ3_S.gguf) | IQ3_S | 6.31GB |
| [Hercules-Qwen1.5-14B.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K_S.gguf) | Q3_K_S | 6.31GB |
| [Hercules-Qwen1.5-14B.IQ3_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ3_M.gguf) | IQ3_M | 6.61GB |
| [Hercules-Qwen1.5-14B.Q3_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K.gguf) | Q3_K | 6.91GB |
| [Hercules-Qwen1.5-14B.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K_M.gguf) | Q3_K_M | 6.91GB |
| [Hercules-Qwen1.5-14B.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q3_K_L.gguf) | Q3_K_L | 7.3GB |
| [Hercules-Qwen1.5-14B.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ4_XS.gguf) | IQ4_XS | 7.37GB |
| [Hercules-Qwen1.5-14B.Q4_0.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_0.gguf) | Q4_0 | 7.62GB |
| [Hercules-Qwen1.5-14B.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.IQ4_NL.gguf) | IQ4_NL | 7.68GB |
| [Hercules-Qwen1.5-14B.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_K_S.gguf) | Q4_K_S | 7.98GB |
| [Hercules-Qwen1.5-14B.Q4_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_K.gguf) | Q4_K | 8.56GB |
| [Hercules-Qwen1.5-14B.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_K_M.gguf) | Q4_K_M | 8.56GB |
| [Hercules-Qwen1.5-14B.Q4_1.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q4_1.gguf) | Q4_1 | 8.4GB |
| [Hercules-Qwen1.5-14B.Q5_0.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_0.gguf) | Q5_0 | 9.18GB |
| [Hercules-Qwen1.5-14B.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_K_S.gguf) | Q5_K_S | 9.34GB |
| [Hercules-Qwen1.5-14B.Q5_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_K.gguf) | Q5_K | 9.81GB |
| [Hercules-Qwen1.5-14B.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_K_M.gguf) | Q5_K_M | 9.81GB |
| [Hercules-Qwen1.5-14B.Q5_1.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q5_1.gguf) | Q5_1 | 9.96GB |
| [Hercules-Qwen1.5-14B.Q6_K.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q6_K.gguf) | Q6_K | 11.46GB |
| [Hercules-Qwen1.5-14B.Q8_0.gguf](https://huggingface.co/RichardErkhov/M4-ai_-_Hercules-Qwen1.5-14B-gguf/blob/main/Hercules-Qwen1.5-14B.Q8_0.gguf) | Q8_0 | 14.03GB |
Original model description:
---
library_name: transformers
license: other
datasets:
- Locutusque/hercules-v4.0
language:
- en
---
<style>
body {
font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;
line-height: 1.6;
color: #f5f5f5;
background-color: #1e2a36;
margin: 0;
padding: 0;
}
.container {
max-width: 1200px;
margin: 20px auto;
padding: 20px;
background-color: #2a3f54;
border-radius: 8px;
box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
display: flex;
flex-wrap: wrap;
justify-content: space-between;
}
h1 {
font-size: 2.5rem;
color: #51a3d3;
text-align: center;
margin-bottom: 30px;
width: 100%;
}
h2 {
font-size: 1.75rem;
margin: 20px 0;
color: #63b8ea;
padding-bottom: 10px;
}
h3 {
font-size: 1.25rem;
color: #80c8f4;
}
p, a {
font-size: 1rem;
}
p {
color: #b0c2ce;
margin-bottom: 20px;
}
ul {
list-style-type: none;
padding: 0;
display: flex;
flex-wrap: wrap;
justify-content: space-between;
width: 100%;
}
li {
background-color: #34495e;
padding: 20px;
margin-bottom: 10px;
border-radius: 4px;
cursor: pointer;
transition: background-color 0.3s ease, color 0.3s ease;
overflow: hidden;
color: #b0c2ce;
width: calc(50% - 10px);
box-shadow: 0 2px 4px rgba(0, 0, 0, 0.1);
}
li:hover {
background-color: #4e6a81;
color: #dfe8f1;
}
.section-content {
margin-top: 15px;
border-top: 1px solid #4e6a81;
padding-top: 10px;
}
a {
color: #a4c8e1;
text-decoration: none;
}
a:hover {
text-decoration: underline;
}
pre {
background-color: #2c3e50;
padding: 10px;
border-radius: 5px;
overflow-x: auto;
color: #b0c2ce;
}
</style>
<div class="container">
<h1>Hercules-Qwen1.5-14B</h1>
</div>
<ul>
<li>
<h2>Model Details</h2>
<div class="section-content">
<h3>Model Description</h3>
<p>This model has capabilities in math, coding, function calling, roleplay, and more. We fine-tuned it using 700,000 examples of Hercules-v4.</p>
<p><strong>Developed by:</strong> M4-ai</p>
<p><strong>Language(s) (NLP):</strong> English and maybe Chinese</p>
<p><strong>License:</strong> tongyi-qianwen license</p>
<p><strong>Finetuned from model:</strong> <a href="https://huggingface.co/Qwen/Qwen1.5-14B">Qwen1.5-14B</a></p>
</div>
</li>
<li>
<h2>Uses</h2>
<div class="section-content">
<p>General purpose assistant, question answering, chain-of-thought, etc..</p>
<h3>Recommendations</h3>
<p>Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.</p>
</div>
</li>
<li>
<h2>Evaluation</h2>
<div class="section-content">
<p>Coming soon</p>
</div>
</li>
<li>
<h2>Training Details</h2>
<div class="section-content">
<h3>Training Data</h3>
<p><a href="https://huggingface.co/datasets/Locutusque/hercules-v4.0">https://huggingface.co/datasets/Locutusque/hercules-v4.0</a></p>
<h4>Training Hyperparameters</h4>
<p><strong>Training regime:</strong> bf16 non-mixed precision</p>
</div>
</li>
<li>
<h2>Technical Specifications</h2>
<div class="section-content">
<h4>Hardware</h4>
<p>We used 8 Kaggle TPUs, and we trained at a global batch size of 128 and sequence length of 1024</p>
</div>
</li>
<li>
<h2>Contributions</h2>
<div class="section-content">
<p>Thanks to @Tonic, @aloobun, @fhai50032, and @Locutusque for their contributions to this model.</p>
</div>
</li>
</ul>
|
{}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,882 |
Helsinki-NLP/opus-mt-sv-wls
|
Helsinki-NLP
|
translation
|
[
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"sv",
"wls",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-03-02T23:29:04Z |
2023-08-16T12:06:28+00:00
| 349 | 0 |
---
license: apache-2.0
tags:
- translation
---
### opus-mt-sv-wls
* source languages: sv
* target languages: wls
* OPUS readme: [sv-wls](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/sv-wls/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-16.zip](https://object.pouta.csc.fi/OPUS-MT-models/sv-wls/opus-2020-01-16.zip)
* test set translations: [opus-2020-01-16.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/sv-wls/opus-2020-01-16.test.txt)
* test set scores: [opus-2020-01-16.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/sv-wls/opus-2020-01-16.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| JW300.sv.wls | 29.0 | 0.501 |
| null |
Non_BioNLP
|
### opus-mt-sv-wls
* source languages: sv
* target languages: wls
* OPUS readme: [sv-wls](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/sv-wls/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-16.zip](https://object.pouta.csc.fi/OPUS-MT-models/sv-wls/opus-2020-01-16.zip)
* test set translations: [opus-2020-01-16.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/sv-wls/opus-2020-01-16.test.txt)
* test set scores: [opus-2020-01-16.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/sv-wls/opus-2020-01-16.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| JW300.sv.wls | 29.0 | 0.501 |
|
{"license": "apache-2.0", "tags": ["translation"]}
|
task
|
[
"TRANSLATION"
] | 39,883 |
YakovElm/Hyperledger5SetFitModel_Train_balance_ratio_Half
|
YakovElm
|
text-classification
|
[
"sentence-transformers",
"pytorch",
"mpnet",
"setfit",
"text-classification",
"arxiv:2209.11055",
"license:apache-2.0",
"region:us"
] | 2023-06-09T11:58:00Z |
2023-06-09T11:58:35+00:00
| 9 | 0 |
---
license: apache-2.0
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
---
# YakovElm/Hyperledger5SetFitModel_Train_balance_ratio_Half
This is a [SetFit model](https://github.com/huggingface/setfit) that can be used for text classification. The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Usage
To use this model for inference, first install the SetFit library:
```bash
python -m pip install setfit
```
You can then run inference as follows:
```python
from setfit import SetFitModel
# Download from Hub and run inference
model = SetFitModel.from_pretrained("YakovElm/Hyperledger5SetFitModel_Train_balance_ratio_Half")
# Run inference
preds = model(["i loved the spiderman movie!", "pineapple on pizza is the worst 🤮"])
```
## BibTeX entry and citation info
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
| null |
Non_BioNLP
|
# YakovElm/Hyperledger5SetFitModel_Train_balance_ratio_Half
This is a [SetFit model](https://github.com/huggingface/setfit) that can be used for text classification. The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Usage
To use this model for inference, first install the SetFit library:
```bash
python -m pip install setfit
```
You can then run inference as follows:
```python
from setfit import SetFitModel
# Download from Hub and run inference
model = SetFitModel.from_pretrained("YakovElm/Hyperledger5SetFitModel_Train_balance_ratio_Half")
# Run inference
preds = model(["i loved the spiderman movie!", "pineapple on pizza is the worst 🤮"])
```
## BibTeX entry and citation info
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
|
{"license": "apache-2.0", "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification"]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,884 |
october-sd/pegasus-xsum-finetuned-en-sum
|
october-sd
|
summarization
|
[
"transformers",
"tensorboard",
"safetensors",
"pegasus",
"text2text-generation",
"summarization",
"generated_from_trainer",
"base_model:october-sd/pegasus-xsum-finetuned-en-sum",
"base_model:finetune:october-sd/pegasus-xsum-finetuned-en-sum",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-03-11T13:38:27Z |
2024-04-06T05:15:46+00:00
| 18 | 0 |
---
base_model: october-sd/pegasus-xsum-finetuned-en-sum
tags:
- summarization
- generated_from_trainer
model-index:
- name: pegasus-xsum-finetuned-en-sum
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# pegasus-xsum-finetuned-en-sum
This model is a fine-tuned version of [october-sd/pegasus-xsum-finetuned-en-sum](https://huggingface.co/october-sd/pegasus-xsum-finetuned-en-sum) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5095
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 1.0 | 507 | 1.4958 |
| No log | 2.0 | 1014 | 1.5115 |
| No log | 3.0 | 1522 | 1.5094 |
| No log | 4.0 | 2028 | 1.5095 |
### Framework versions
- Transformers 4.38.1
- Pytorch 2.1.2
- Datasets 2.15.0
- Tokenizers 0.15.2
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# pegasus-xsum-finetuned-en-sum
This model is a fine-tuned version of [october-sd/pegasus-xsum-finetuned-en-sum](https://huggingface.co/october-sd/pegasus-xsum-finetuned-en-sum) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5095
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 1.0 | 507 | 1.4958 |
| No log | 2.0 | 1014 | 1.5115 |
| No log | 3.0 | 1522 | 1.5094 |
| No log | 4.0 | 2028 | 1.5095 |
### Framework versions
- Transformers 4.38.1
- Pytorch 2.1.2
- Datasets 2.15.0
- Tokenizers 0.15.2
|
{"base_model": "october-sd/pegasus-xsum-finetuned-en-sum", "tags": ["summarization", "generated_from_trainer"], "model-index": [{"name": "pegasus-xsum-finetuned-en-sum", "results": []}]}
|
task
|
[
"SUMMARIZATION"
] | 39,885 |
Helsinki-NLP/opus-mt-rw-es
|
Helsinki-NLP
|
translation
|
[
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"rw",
"es",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-03-02T23:29:04Z |
2023-08-16T12:03:45+00:00
| 48 | 0 |
---
license: apache-2.0
tags:
- translation
---
### opus-mt-rw-es
* source languages: rw
* target languages: es
* OPUS readme: [rw-es](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/rw-es/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-16.zip](https://object.pouta.csc.fi/OPUS-MT-models/rw-es/opus-2020-01-16.zip)
* test set translations: [opus-2020-01-16.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/rw-es/opus-2020-01-16.test.txt)
* test set scores: [opus-2020-01-16.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/rw-es/opus-2020-01-16.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| JW300.rw.es | 26.2 | 0.445 |
| null |
Non_BioNLP
|
### opus-mt-rw-es
* source languages: rw
* target languages: es
* OPUS readme: [rw-es](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/rw-es/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-16.zip](https://object.pouta.csc.fi/OPUS-MT-models/rw-es/opus-2020-01-16.zip)
* test set translations: [opus-2020-01-16.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/rw-es/opus-2020-01-16.test.txt)
* test set scores: [opus-2020-01-16.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/rw-es/opus-2020-01-16.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| JW300.rw.es | 26.2 | 0.445 |
|
{"license": "apache-2.0", "tags": ["translation"]}
|
task
|
[
"TRANSLATION"
] | 39,886 |
NTUYG/ComFormer
|
NTUYG
|
summarization
|
[
"transformers",
"pytorch",
"bart",
"text2text-generation",
"summarization",
"en",
"dataset:DeepCom",
"arxiv:2107.03644",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-05-09T04:20:31Z |
2022-05-09T10:55:14+00:00
| 135 | 1 |
---
datasets:
- DeepCom
language:
- en
license: apache-2.0
metrics:
- bleu
tags:
- summarization
---
# How To Use
```PYTHON
from transformers import BartForConditionalGeneration, BartTokenizer
model = BartForConditionalGeneration.from_pretrained("NTUYG/ComFormer")
tokenizer = BartTokenizer.from_pretrained("NTUYG/ComFormer")
code = '''
public static void copyFile( File in, File out )
throws IOException
{
FileChannel inChannel = new FileInputStream( in ).getChannel();
FileChannel outChannel = new FileOutputStream( out ).getChannel();
try
{
// inChannel.transferTo(0, inChannel.size(), outChannel); // original -- apparently has trouble copying large files on Windows
// magic number for Windows, 64Mb - 32Kb)
int maxCount = (64 * 1024 * 1024) - (32 * 1024);
long size = inChannel.size();
long position = 0;
while ( position < size )
{
position += inChannel.transferTo( position, maxCount, outChannel );
}
}
finally
{
if ( inChannel != null )
{
inChannel.close();
}
if ( outChannel != null )
{
outChannel.close();
}
}
}
'''
code_seq, sbt = utils.transformer(code) #can find in https://github.com/NTDXYG/ComFormer
input_text = code_seq + sbt
input_ids = tokenizer.encode(input_text, return_tensors="pt", max_length=256, truncation=True)
summary_text_ids = model.generate(
input_ids=input_ids,
bos_token_id=model.config.bos_token_id,
eos_token_id=model.config.eos_token_id,
length_penalty=2.0,
max_length=30,
min_length=2,
num_beams=5,
)
comment = tokenizer.decode(summary_text_ids[0], skip_special_tokens=True)
print(comment)
```
# BibTeX entry and citation info
```
@misc{yang2021comformer,
title={ComFormer: Code Comment Generation via Transformer and Fusion Method-based Hybrid Code Representation},
author={Guang Yang and Xiang Chen and Jinxin Cao and Shuyuan Xu and Zhanqi Cui and Chi Yu and Ke Liu},
year={2021},
eprint={2107.03644},
archivePrefix={arXiv},
primaryClass={cs.SE}
}
```
| null |
Non_BioNLP
|
# How To Use
```PYTHON
from transformers import BartForConditionalGeneration, BartTokenizer
model = BartForConditionalGeneration.from_pretrained("NTUYG/ComFormer")
tokenizer = BartTokenizer.from_pretrained("NTUYG/ComFormer")
code = '''
public static void copyFile( File in, File out )
throws IOException
{
FileChannel inChannel = new FileInputStream( in ).getChannel();
FileChannel outChannel = new FileOutputStream( out ).getChannel();
try
{
// inChannel.transferTo(0, inChannel.size(), outChannel); // original -- apparently has trouble copying large files on Windows
// magic number for Windows, 64Mb - 32Kb)
int maxCount = (64 * 1024 * 1024) - (32 * 1024);
long size = inChannel.size();
long position = 0;
while ( position < size )
{
position += inChannel.transferTo( position, maxCount, outChannel );
}
}
finally
{
if ( inChannel != null )
{
inChannel.close();
}
if ( outChannel != null )
{
outChannel.close();
}
}
}
'''
code_seq, sbt = utils.transformer(code) #can find in https://github.com/NTDXYG/ComFormer
input_text = code_seq + sbt
input_ids = tokenizer.encode(input_text, return_tensors="pt", max_length=256, truncation=True)
summary_text_ids = model.generate(
input_ids=input_ids,
bos_token_id=model.config.bos_token_id,
eos_token_id=model.config.eos_token_id,
length_penalty=2.0,
max_length=30,
min_length=2,
num_beams=5,
)
comment = tokenizer.decode(summary_text_ids[0], skip_special_tokens=True)
print(comment)
```
# BibTeX entry and citation info
```
@misc{yang2021comformer,
title={ComFormer: Code Comment Generation via Transformer and Fusion Method-based Hybrid Code Representation},
author={Guang Yang and Xiang Chen and Jinxin Cao and Shuyuan Xu and Zhanqi Cui and Chi Yu and Ke Liu},
year={2021},
eprint={2107.03644},
archivePrefix={arXiv},
primaryClass={cs.SE}
}
```
|
{"datasets": ["DeepCom"], "language": ["en"], "license": "apache-2.0", "metrics": ["bleu"], "tags": ["summarization"]}
|
task
|
[
"SUMMARIZATION"
] | 39,887 |
QuantFactory/granite-3.1-2b-base-GGUF
|
QuantFactory
| null |
[
"transformers",
"gguf",
"language",
"granite-3.1",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | 2024-12-19T11:10:04Z |
2024-12-19T11:23:03+00:00
| 107 | 3 |
---
library_name: transformers
license: apache-2.0
tags:
- language
- granite-3.1
---
[](https://hf.co/QuantFactory)
# QuantFactory/granite-3.1-2b-base-GGUF
This is quantized version of [ibm-granite/granite-3.1-2b-base](https://huggingface.co/ibm-granite/granite-3.1-2b-base) created using llama.cpp
# Original Model Card
# Granite-3.1-2B-Base
**Model Summary:**
Granite-3.1-2B-Base extends the context length of Granite-3.0-2B-Base from 4K to 128K using a progressive training strategy by increasing the supported context length in increments while adjusting RoPE theta until the model has successfully adapted to desired length of 128K. This long-context pre-training stage was performed using approximately 500B tokens.
- **Developers:** Granite Team, IBM
- **GitHub Repository:** [ibm-granite/granite-3.1-language-models](https://github.com/ibm-granite/granite-3.1-language-models)
- **Website**: [Granite Docs](https://www.ibm.com/granite/docs/)
- **Paper:** [Granite 3.1 Language Models (coming soon)](https://huggingface.co/collections/ibm-granite/granite-31-language-models-6751dbbf2f3389bec5c6f02d)
- **Release Date**: December 18th, 2024
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)
**Supported Languages:**
English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. Users may finetune Granite 3.1 models for languages beyond these 12 languages.
**Intended Use:**
Prominent use cases of LLMs in text-to-text generation include summarization, text classification, extraction, question-answering, and other long-context tasks. All Granite Base models are able to handle these tasks as they were trained on a large amount of data from various domains. Moreover, they can serve as baseline to create specialized models for specific application scenarios.
**Generation:**
This is a simple example of how to use Granite-3.1-2B-Base model.
Install the following libraries:
```shell
pip install torch torchvision torchaudio
pip install accelerate
pip install transformers
```
Then, copy the code snippet below to run the example.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "auto"
model_path = "ibm-granite/granite-3.1-2b-base"
tokenizer = AutoTokenizer.from_pretrained(model_path)
# drop device_map if running on CPU
model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device)
model.eval()
# change input text as desired
input_text = "Where is the Thomas J. Watson Research Center located?"
# tokenize the text
input_tokens = tokenizer(input_text, return_tensors="pt").to(device)
# generate output tokens
output = model.generate(**input_tokens,
max_length=4000)
# decode output tokens into text
output = tokenizer.batch_decode(output)
# print output
print(output)
```
**Model Architecture:**
Granite-3.1-2B-Base is based on a decoder-only dense transformer architecture. Core components of this architecture are: GQA and RoPE, MLP with SwiGLU, RMSNorm, and shared input/output embeddings.
| Model | 2B Dense | 8B Dense | 1B MoE | 3B MoE |
| :-------- | :--------| :-------- | :------| :------|
| Embedding size | **2048** | 4096 | 1024 | 1536 |
| Number of layers | **40** | 40 | 24 | 32 |
| Attention head size | **64** | 128 | 64 | 64 |
| Number of attention heads | **32** | 32 | 16 | 24 |
| Number of KV heads | **8** | 8 | 8 | 8 |
| MLP hidden size | **8192** | 12800 | 512 | 512 |
| MLP activation | **SwiGLU** | SwiGLU | SwiGLU | SwiGLU |
| Number of experts | **—** | — | 32 | 40 |
| MoE TopK | **—** | — | 8 | 8 |
| Initialization std | **0.1** | 0.1 | 0.1 | 0.1 |
| Sequence length | **128K** | 128K | 128K | 128K |
| Position embedding | **RoPE** | RoPE | RoPE | RoPE |
| # Parameters | **2.5B** | 8.1B | 1.3B | 3.3B |
| # Active parameters | **2.5B** | 8.1B | 400M | 800M |
| # Training tokens | **12T** | 12T | 10T | 10T |
**Training Data:**
This model is trained on a mix of open source and proprietary data following a three-stage training strategy.
* Stage 1 data: The data for stage 1 is sourced from diverse domains, such as: web, code, academic sources, books, and math data.
* Stage 2 data: The data for stage 2 comprises a curated mix of high-quality data from the same domains, plus multilingual and instruction data. The goal of this second training phase is to enhance the model’s performance on specific tasks.
* Stage 3 data: The data for stage 3 consists of original stage-2 pretraining data with additional synthetic long-context data in form of QA/summary pairs where the answer
contains a recitation of the related paragraph before the answer.
A detailed attribution of datasets can be found in the [Granite 3.0 Technical Report](https://github.com/ibm-granite/granite-3.0-language-models/blob/main/paper.pdf), [Granite 3.1 Technical Report (coming soon)](https://huggingface.co/collections/ibm-granite/granite-31-language-models-6751dbbf2f3389bec5c6f02d), and [Accompanying Author List](https://github.com/ibm-granite/granite-3.0-language-models/blob/main/author-ack.pdf).
**Infrastructure:**
We train Granite 3.1 Language Models using IBM's super computing cluster, Blue Vela, which is outfitted with NVIDIA H100 GPUs. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs.
**Ethical Considerations and Limitations:**
The use of Large Language Models involves risks and ethical considerations people must be aware of, including but not limited to: bias and fairness, misinformation, and autonomous decision-making. Granite-3.1-2B-Base model is not the exception in this regard. Even though this model is suited for multiple generative AI tasks, it has not undergone any safety alignment, there it may produce problematic outputs. Additionally, it remains uncertain whether smaller models might exhibit increased susceptibility to hallucination in generation scenarios by copying text verbatim from the training dataset due to their reduced sizes and memorization capacities. This aspect is currently an active area of research, and we anticipate more rigorous exploration, comprehension, and mitigations in this domain. Regarding ethics, a latent risk associated with all Large Language Models is their malicious utilization. We urge the community to use Granite-3.1-2B-Base model with ethical intentions and in a responsible way.
| null |
Non_BioNLP
|
[](https://hf.co/QuantFactory)
# QuantFactory/granite-3.1-2b-base-GGUF
This is quantized version of [ibm-granite/granite-3.1-2b-base](https://huggingface.co/ibm-granite/granite-3.1-2b-base) created using llama.cpp
# Original Model Card
# Granite-3.1-2B-Base
**Model Summary:**
Granite-3.1-2B-Base extends the context length of Granite-3.0-2B-Base from 4K to 128K using a progressive training strategy by increasing the supported context length in increments while adjusting RoPE theta until the model has successfully adapted to desired length of 128K. This long-context pre-training stage was performed using approximately 500B tokens.
- **Developers:** Granite Team, IBM
- **GitHub Repository:** [ibm-granite/granite-3.1-language-models](https://github.com/ibm-granite/granite-3.1-language-models)
- **Website**: [Granite Docs](https://www.ibm.com/granite/docs/)
- **Paper:** [Granite 3.1 Language Models (coming soon)](https://huggingface.co/collections/ibm-granite/granite-31-language-models-6751dbbf2f3389bec5c6f02d)
- **Release Date**: December 18th, 2024
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)
**Supported Languages:**
English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. Users may finetune Granite 3.1 models for languages beyond these 12 languages.
**Intended Use:**
Prominent use cases of LLMs in text-to-text generation include summarization, text classification, extraction, question-answering, and other long-context tasks. All Granite Base models are able to handle these tasks as they were trained on a large amount of data from various domains. Moreover, they can serve as baseline to create specialized models for specific application scenarios.
**Generation:**
This is a simple example of how to use Granite-3.1-2B-Base model.
Install the following libraries:
```shell
pip install torch torchvision torchaudio
pip install accelerate
pip install transformers
```
Then, copy the code snippet below to run the example.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "auto"
model_path = "ibm-granite/granite-3.1-2b-base"
tokenizer = AutoTokenizer.from_pretrained(model_path)
# drop device_map if running on CPU
model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device)
model.eval()
# change input text as desired
input_text = "Where is the Thomas J. Watson Research Center located?"
# tokenize the text
input_tokens = tokenizer(input_text, return_tensors="pt").to(device)
# generate output tokens
output = model.generate(**input_tokens,
max_length=4000)
# decode output tokens into text
output = tokenizer.batch_decode(output)
# print output
print(output)
```
**Model Architecture:**
Granite-3.1-2B-Base is based on a decoder-only dense transformer architecture. Core components of this architecture are: GQA and RoPE, MLP with SwiGLU, RMSNorm, and shared input/output embeddings.
| Model | 2B Dense | 8B Dense | 1B MoE | 3B MoE |
| :-------- | :--------| :-------- | :------| :------|
| Embedding size | **2048** | 4096 | 1024 | 1536 |
| Number of layers | **40** | 40 | 24 | 32 |
| Attention head size | **64** | 128 | 64 | 64 |
| Number of attention heads | **32** | 32 | 16 | 24 |
| Number of KV heads | **8** | 8 | 8 | 8 |
| MLP hidden size | **8192** | 12800 | 512 | 512 |
| MLP activation | **SwiGLU** | SwiGLU | SwiGLU | SwiGLU |
| Number of experts | **—** | — | 32 | 40 |
| MoE TopK | **—** | — | 8 | 8 |
| Initialization std | **0.1** | 0.1 | 0.1 | 0.1 |
| Sequence length | **128K** | 128K | 128K | 128K |
| Position embedding | **RoPE** | RoPE | RoPE | RoPE |
| # Parameters | **2.5B** | 8.1B | 1.3B | 3.3B |
| # Active parameters | **2.5B** | 8.1B | 400M | 800M |
| # Training tokens | **12T** | 12T | 10T | 10T |
**Training Data:**
This model is trained on a mix of open source and proprietary data following a three-stage training strategy.
* Stage 1 data: The data for stage 1 is sourced from diverse domains, such as: web, code, academic sources, books, and math data.
* Stage 2 data: The data for stage 2 comprises a curated mix of high-quality data from the same domains, plus multilingual and instruction data. The goal of this second training phase is to enhance the model’s performance on specific tasks.
* Stage 3 data: The data for stage 3 consists of original stage-2 pretraining data with additional synthetic long-context data in form of QA/summary pairs where the answer
contains a recitation of the related paragraph before the answer.
A detailed attribution of datasets can be found in the [Granite 3.0 Technical Report](https://github.com/ibm-granite/granite-3.0-language-models/blob/main/paper.pdf), [Granite 3.1 Technical Report (coming soon)](https://huggingface.co/collections/ibm-granite/granite-31-language-models-6751dbbf2f3389bec5c6f02d), and [Accompanying Author List](https://github.com/ibm-granite/granite-3.0-language-models/blob/main/author-ack.pdf).
**Infrastructure:**
We train Granite 3.1 Language Models using IBM's super computing cluster, Blue Vela, which is outfitted with NVIDIA H100 GPUs. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs.
**Ethical Considerations and Limitations:**
The use of Large Language Models involves risks and ethical considerations people must be aware of, including but not limited to: bias and fairness, misinformation, and autonomous decision-making. Granite-3.1-2B-Base model is not the exception in this regard. Even though this model is suited for multiple generative AI tasks, it has not undergone any safety alignment, there it may produce problematic outputs. Additionally, it remains uncertain whether smaller models might exhibit increased susceptibility to hallucination in generation scenarios by copying text verbatim from the training dataset due to their reduced sizes and memorization capacities. This aspect is currently an active area of research, and we anticipate more rigorous exploration, comprehension, and mitigations in this domain. Regarding ethics, a latent risk associated with all Large Language Models is their malicious utilization. We urge the community to use Granite-3.1-2B-Base model with ethical intentions and in a responsible way.
|
{"library_name": "transformers", "license": "apache-2.0", "tags": ["language", "granite-3.1"]}
|
task
|
[
"TEXT_CLASSIFICATION",
"SUMMARIZATION"
] | 39,889 |
hyunussarioglu/tr-paraphrase-mt5-base-ost
|
hyunussarioglu
|
text2text-generation
|
[
"transformers",
"pytorch",
"safetensors",
"mt5",
"text2text-generation",
"arxiv:1910.09700",
"license:cc-by-4.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-11-15T15:36:20Z |
2025-03-05T17:06:14+00:00
| 124 | 0 |
---
license: cc-by-4.0
---
# Model Card for tr-paraphrase-mt5-base-ost
<!-- Provide a quick summary of what the model is/does. [Optional] -->
The model is a fine-tuned version of mT5-base provided by [Google](https://huggingface.co/google/mt5-base).
# Table of Contents
- [Model Card for tr-paraphrase-mt5-base-ost](#model-card-for--model_id-)
- [Table of Contents](#table-of-contents)
- [Table of Contents](#table-of-contents-1)
- [Model Details](#model-details)
- [Model Description](#model-description)
- [Uses](#uses)
- [Direct Use](#direct-use)
- [Downstream Use [Optional]](#downstream-use-optional)
- [Out-of-Scope Use](#out-of-scope-use)
- [Bias, Risks, and Limitations](#bias-risks-and-limitations)
- [Recommendations](#recommendations)
- [Training Details](#training-details)
- [Training Data](#training-data)
- [Training Procedure](#training-procedure)
- [Preprocessing](#preprocessing)
- [Speeds, Sizes, Times](#speeds-sizes-times)
- [Evaluation](#evaluation)
- [Testing Data, Factors & Metrics](#testing-data-factors--metrics)
- [Testing Data](#testing-data)
- [Factors](#factors)
- [Metrics](#metrics)
- [Results](#results)
- [Model Examination](#model-examination)
- [Environmental Impact](#environmental-impact)
- [Technical Specifications [optional]](#technical-specifications-optional)
- [Model Architecture and Objective](#model-architecture-and-objective)
- [Compute Infrastructure](#compute-infrastructure)
- [Hardware](#hardware)
- [Software](#software)
- [Citation](#citation)
- [Glossary [optional]](#glossary-optional)
- [More Information [optional]](#more-information-optional)
- [Model Card Authors [optional]](#model-card-authors-optional)
- [Model Card Contact](#model-card-contact)
- [How to Get Started with the Model](#how-to-get-started-with-the-model)
# Model Details
## Model Description
<!-- Provide a longer summary of what this model is/does. -->
The model is a fine-tuned version of mT5-base provided by [Google](https://huggingface.co/google/mt5-base).
- **Developed by:** More information needed
- **Shared by [Optional]:** More information needed
- **Model type:** Language model
- **Language(s) (NLP):** tr
- **License:** cc-by-4.0
- **Parent Model:** More information needed
- **Resources for more information:** More information needed
- [GitHub Repo](https://github.com/mrbesher/semantic-filtering-for-paraphrasing)
# Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
## Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
## Downstream Use [Optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
## Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
# Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
## Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
# Training Details
## Training Data
<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
More information on training data needed
## Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
### Preprocessing
More information needed
### Speeds, Sizes, Times
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
More information needed
# Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
## Testing Data, Factors & Metrics
### Testing Data
<!-- This should link to a Data Card if possible. -->
More information needed
### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
More information needed
### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
More information needed
## Results
More information needed
# Model Examination
More information needed
# Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** More information needed
- **Hours used:** More information needed
- **Cloud Provider:** More information needed
- **Compute Region:** More information needed
- **Carbon Emitted:** More information needed
# Technical Specifications [optional]
## Model Architecture and Objective
More information needed
## Compute Infrastructure
More information needed
### Hardware
More information needed
### Software
More information needed
# Citation
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
More information needed
**APA:**
More information needed
# Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
More information needed
# More Information [optional]
More information needed
# Model Card Authors [optional]
<!-- This section provides another layer of transparency and accountability. Whose views is this model card representing? How many voices were included in its construction? Etc. -->
Besher Alkurdi, Hasan Yunus Sarioglu, Mehmet Fatih Amasyali
# Model Card Contact
More information needed
# How to Get Started with the Model
Use the code below to get started with the model.
<details>
<summary> Click to expand </summary>
More information needed
</details>
| null |
Non_BioNLP
|
# Model Card for tr-paraphrase-mt5-base-ost
<!-- Provide a quick summary of what the model is/does. [Optional] -->
The model is a fine-tuned version of mT5-base provided by [Google](https://huggingface.co/google/mt5-base).
# Table of Contents
- [Model Card for tr-paraphrase-mt5-base-ost](#model-card-for--model_id-)
- [Table of Contents](#table-of-contents)
- [Table of Contents](#table-of-contents-1)
- [Model Details](#model-details)
- [Model Description](#model-description)
- [Uses](#uses)
- [Direct Use](#direct-use)
- [Downstream Use [Optional]](#downstream-use-optional)
- [Out-of-Scope Use](#out-of-scope-use)
- [Bias, Risks, and Limitations](#bias-risks-and-limitations)
- [Recommendations](#recommendations)
- [Training Details](#training-details)
- [Training Data](#training-data)
- [Training Procedure](#training-procedure)
- [Preprocessing](#preprocessing)
- [Speeds, Sizes, Times](#speeds-sizes-times)
- [Evaluation](#evaluation)
- [Testing Data, Factors & Metrics](#testing-data-factors--metrics)
- [Testing Data](#testing-data)
- [Factors](#factors)
- [Metrics](#metrics)
- [Results](#results)
- [Model Examination](#model-examination)
- [Environmental Impact](#environmental-impact)
- [Technical Specifications [optional]](#technical-specifications-optional)
- [Model Architecture and Objective](#model-architecture-and-objective)
- [Compute Infrastructure](#compute-infrastructure)
- [Hardware](#hardware)
- [Software](#software)
- [Citation](#citation)
- [Glossary [optional]](#glossary-optional)
- [More Information [optional]](#more-information-optional)
- [Model Card Authors [optional]](#model-card-authors-optional)
- [Model Card Contact](#model-card-contact)
- [How to Get Started with the Model](#how-to-get-started-with-the-model)
# Model Details
## Model Description
<!-- Provide a longer summary of what this model is/does. -->
The model is a fine-tuned version of mT5-base provided by [Google](https://huggingface.co/google/mt5-base).
- **Developed by:** More information needed
- **Shared by [Optional]:** More information needed
- **Model type:** Language model
- **Language(s) (NLP):** tr
- **License:** cc-by-4.0
- **Parent Model:** More information needed
- **Resources for more information:** More information needed
- [GitHub Repo](https://github.com/mrbesher/semantic-filtering-for-paraphrasing)
# Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
## Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
## Downstream Use [Optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
## Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
# Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
## Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
# Training Details
## Training Data
<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
More information on training data needed
## Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
### Preprocessing
More information needed
### Speeds, Sizes, Times
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
More information needed
# Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
## Testing Data, Factors & Metrics
### Testing Data
<!-- This should link to a Data Card if possible. -->
More information needed
### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
More information needed
### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
More information needed
## Results
More information needed
# Model Examination
More information needed
# Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** More information needed
- **Hours used:** More information needed
- **Cloud Provider:** More information needed
- **Compute Region:** More information needed
- **Carbon Emitted:** More information needed
# Technical Specifications [optional]
## Model Architecture and Objective
More information needed
## Compute Infrastructure
More information needed
### Hardware
More information needed
### Software
More information needed
# Citation
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
More information needed
**APA:**
More information needed
# Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
More information needed
# More Information [optional]
More information needed
# Model Card Authors [optional]
<!-- This section provides another layer of transparency and accountability. Whose views is this model card representing? How many voices were included in its construction? Etc. -->
Besher Alkurdi, Hasan Yunus Sarioglu, Mehmet Fatih Amasyali
# Model Card Contact
More information needed
# How to Get Started with the Model
Use the code below to get started with the model.
<details>
<summary> Click to expand </summary>
More information needed
</details>
|
{"license": "cc-by-4.0"}
|
task
|
[
"PARAPHRASING"
] | 39,890 |
hojzas/setfit-proj8-multilabel
|
hojzas
|
text-classification
|
[
"setfit",
"safetensors",
"mpnet",
"sentence-transformers",
"text-classification",
"generated_from_setfit_trainer",
"dataset:hojzas/proj8-multilabel",
"arxiv:2209.11055",
"base_model:sentence-transformers/paraphrase-mpnet-base-v2",
"base_model:finetune:sentence-transformers/paraphrase-mpnet-base-v2",
"co2_eq_emissions",
"region:us"
] | 2024-01-26T10:07:33Z |
2024-01-26T10:07:59+00:00
| 4 | 0 |
---
base_model: sentence-transformers/paraphrase-mpnet-base-v2
datasets:
- hojzas/proj8-multilabel
library_name: setfit
metrics:
- accuracy
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
widget:
- text: 'def first_with_given_key(iterable, key=lambda x: x):\n keys_used = {}\n for
item in iterable:\n rp = repr(key(item))\n if rp not in keys_used.keys():\n keys_used[rp]
= repr(item)\n yield item'
- text: 'def first_with_given_key(iterable, key=lambda x: x):\n keys=[]\n for
i in iterable:\n if key(i) not in keys:\n yield i\n keys.append(key(i))'
- text: 'def first_with_given_key(iterable, key=repr):\n set_of_keys = set()\n lambda_key
= (lambda x: key(x))\n for item in iterable:\n key = lambda_key(item)\n try:\n key_for_set
= hash(key)\n except TypeError:\n key_for_set = repr(key)\n if
key_for_set in set_of_keys:\n continue\n set_of_keys.add(key_for_set)\n yield
item'
- text: 'def first_with_given_key(iterable, key = lambda x: x):\n found_keys={}\n for
i in iterable:\n if key(i) not in found_keys.keys():\n found_keys[key(i)]=i\n yield
i'
- text: 'def first_with_given_key(the_iterable, key=lambda x: x):\n temp_keys=[]\n for
i in range(len(the_iterable)):\n if (key(the_iterable[i]) not in temp_keys):\n temp_keys.append(key(the_iterable[i]))\n yield
the_iterable[i]\n del temp_keys'
inference: false
co2_eq_emissions:
emissions: 0.2716104726718793
source: codecarbon
training_type: fine-tuning
on_cloud: false
cpu_model: Intel(R) Xeon(R) Silver 4314 CPU @ 2.40GHz
ram_total_size: 251.49160385131836
hours_used: 0.005
---
# SetFit with sentence-transformers/paraphrase-mpnet-base-v2
This is a [SetFit](https://github.com/huggingface/setfit) model trained on the [hojzas/proj8-multilabel](https://huggingface.co/datasets/hojzas/proj8-multilabel) dataset that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2) as the Sentence Transformer embedding model. A OneVsRestClassifier instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
- **Sentence Transformer body:** [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2)
- **Classification head:** a OneVsRestClassifier instance
- **Maximum Sequence Length:** 512 tokens
<!-- - **Number of Classes:** Unknown -->
- **Training Dataset:** [hojzas/proj8-multilabel](https://huggingface.co/datasets/hojzas/proj8-multilabel)
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("hojzas/setfit-proj8-multilabel")
# Run inference
preds = model("def first_with_given_key(iterable, key=lambda x: x):\n keys=[]\n for i in iterable:\n if key(i) not in keys:\n yield i\n keys.append(key(i))")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Set Metrics
| Training set | Min | Median | Max |
|:-------------|:----|:--------|:----|
| Word count | 43 | 92.5185 | 125 |
### Training Hyperparameters
- batch_size: (16, 16)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 20
- body_learning_rate: (2e-05, 2e-05)
- head_learning_rate: 2e-05
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
### Training Results
| Epoch | Step | Training Loss | Validation Loss |
|:------:|:----:|:-------------:|:---------------:|
| 0.0147 | 1 | 0.3001 | - |
| 0.7353 | 50 | 0.0104 | - |
### Environmental Impact
Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
- **Carbon Emitted**: 0.000 kg of CO2
- **Hours Used**: 0.005 hours
### Training Hardware
- **On Cloud**: No
- **GPU Model**: No GPU used
- **CPU Model**: Intel(R) Xeon(R) Silver 4314 CPU @ 2.40GHz
- **RAM Size**: 251.49 GB
### Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.2.2
- Transformers: 4.36.1
- PyTorch: 2.1.2+cu121
- Datasets: 2.14.7
- Tokenizers: 0.15.1
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SetFit with sentence-transformers/paraphrase-mpnet-base-v2
This is a [SetFit](https://github.com/huggingface/setfit) model trained on the [hojzas/proj8-multilabel](https://huggingface.co/datasets/hojzas/proj8-multilabel) dataset that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2) as the Sentence Transformer embedding model. A OneVsRestClassifier instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
- **Sentence Transformer body:** [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2)
- **Classification head:** a OneVsRestClassifier instance
- **Maximum Sequence Length:** 512 tokens
<!-- - **Number of Classes:** Unknown -->
- **Training Dataset:** [hojzas/proj8-multilabel](https://huggingface.co/datasets/hojzas/proj8-multilabel)
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("hojzas/setfit-proj8-multilabel")
# Run inference
preds = model("def first_with_given_key(iterable, key=lambda x: x):\n keys=[]\n for i in iterable:\n if key(i) not in keys:\n yield i\n keys.append(key(i))")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Set Metrics
| Training set | Min | Median | Max |
|:-------------|:----|:--------|:----|
| Word count | 43 | 92.5185 | 125 |
### Training Hyperparameters
- batch_size: (16, 16)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 20
- body_learning_rate: (2e-05, 2e-05)
- head_learning_rate: 2e-05
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
### Training Results
| Epoch | Step | Training Loss | Validation Loss |
|:------:|:----:|:-------------:|:---------------:|
| 0.0147 | 1 | 0.3001 | - |
| 0.7353 | 50 | 0.0104 | - |
### Environmental Impact
Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
- **Carbon Emitted**: 0.000 kg of CO2
- **Hours Used**: 0.005 hours
### Training Hardware
- **On Cloud**: No
- **GPU Model**: No GPU used
- **CPU Model**: Intel(R) Xeon(R) Silver 4314 CPU @ 2.40GHz
- **RAM Size**: 251.49 GB
### Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.2.2
- Transformers: 4.36.1
- PyTorch: 2.1.2+cu121
- Datasets: 2.14.7
- Tokenizers: 0.15.1
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "sentence-transformers/paraphrase-mpnet-base-v2", "datasets": ["hojzas/proj8-multilabel"], "library_name": "setfit", "metrics": ["accuracy"], "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification", "generated_from_setfit_trainer"], "widget": [{"text": "def first_with_given_key(iterable, key=lambda x: x):\\n keys_used = {}\\n for item in iterable:\\n rp = repr(key(item))\\n if rp not in keys_used.keys():\\n keys_used[rp] = repr(item)\\n yield item"}, {"text": "def first_with_given_key(iterable, key=lambda x: x):\\n keys=[]\\n for i in iterable:\\n if key(i) not in keys:\\n yield i\\n keys.append(key(i))"}, {"text": "def first_with_given_key(iterable, key=repr):\\n set_of_keys = set()\\n lambda_key = (lambda x: key(x))\\n for item in iterable:\\n key = lambda_key(item)\\n try:\\n key_for_set = hash(key)\\n except TypeError:\\n key_for_set = repr(key)\\n if key_for_set in set_of_keys:\\n continue\\n set_of_keys.add(key_for_set)\\n yield item"}, {"text": "def first_with_given_key(iterable, key = lambda x: x):\\n found_keys={}\\n for i in iterable:\\n if key(i) not in found_keys.keys():\\n found_keys[key(i)]=i\\n yield i"}, {"text": "def first_with_given_key(the_iterable, key=lambda x: x):\\n temp_keys=[]\\n for i in range(len(the_iterable)):\\n if (key(the_iterable[i]) not in temp_keys):\\n temp_keys.append(key(the_iterable[i]))\\n yield the_iterable[i]\\n del temp_keys"}], "inference": false, "co2_eq_emissions": {"emissions": 0.2716104726718793, "source": "codecarbon", "training_type": "fine-tuning", "on_cloud": false, "cpu_model": "Intel(R) Xeon(R) Silver 4314 CPU @ 2.40GHz", "ram_total_size": 251.49160385131836, "hours_used": 0.005}}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,891 |
mini1013/master_cate_ac9
|
mini1013
|
text-classification
|
[
"setfit",
"safetensors",
"roberta",
"sentence-transformers",
"text-classification",
"generated_from_setfit_trainer",
"arxiv:2209.11055",
"base_model:mini1013/master_domain",
"base_model:finetune:mini1013/master_domain",
"model-index",
"region:us"
] | 2024-11-25T10:23:03Z |
2024-11-25T10:23:25+00:00
| 243 | 0 |
---
base_model: mini1013/master_domain
library_name: setfit
metrics:
- metric
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
widget:
- text: 여성 가방 숄더백 미니 크로스백 퀼팅백 체인백 토트백 미니백 여자 핸드백 구름백 클러치백 직장인 백팩 프리아_카멜 더블유팝
- text: 국내 잔스포츠 백팩 슈퍼브레이크 4QUT 블랙 학생 여성 가벼운 가방 캠핑 여행 당일 가원
- text: 국내생산 코튼 양줄면주머니 미니&에코 주머니 7종 학원 학교 만들기수업 양줄주머니_14cmX28cm(J14) 명성패키지
- text: 웨빙 플라워 스트랩 레디백 길이조절 가방끈 어깨끈 리폼 3-플라워가방끈-흰색 이백프로
- text: 엔비조네/가방끈/가방끈리폼/가죽끈/크로스끈/숄더끈/스트랩 AOR오링25mm_블랙오플_폭11mm *35cm 니켈 엔비조네
inference: true
model-index:
- name: SetFit with mini1013/master_domain
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: metric
value: 0.7867699642431466
name: Metric
---
# SetFit with mini1013/master_domain
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [mini1013/master_domain](https://huggingface.co/mini1013/master_domain) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
- **Sentence Transformer body:** [mini1013/master_domain](https://huggingface.co/mini1013/master_domain)
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 512 tokens
- **Number of Classes:** 10 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
### Model Labels
| Label | Examples |
|:------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 3.0 | <ul><li>'[현대백화점][루이까또즈] MOONMOON(문문) 여성호보백 HR3SO02BL (주)현대백화점'</li><li>'소프트레더 파스텔 보부상 빅숄더백 휘뚜루마뚜루가방 토드백 블랙_one size 아이디어코리아 주식회사'</li><li>'DRAGON DIFFUSION 드래곤디퓨전 폼폼 더블 점프백 여성 버킷백 8838 드래곤백 다크브라운 (DARK BROWN) 시계1위워치짱'</li></ul> |
| 7.0 | <ul><li>'디어 4colors_H70301010 (W)퍼플와인 '</li><li>'[마이클코어스][정상가 1080000원] 에밀리아 라지 레더 사첼 35H0GU5S7T2171 신세계몰'</li><li>'칼린 소프트M 10colors _H71307020 (Y)라임네온_one size (주)칼린홍대점'</li></ul> |
| 1.0 | <ul><li>'마젤란9901 메신저백 크로스백 학생 여행용 가방 백팩 1_MA-9901-BlackPurple(+LK) 더블유팝'</li><li>'마젤란9901 메신저백 크로스백 학생 여행용 가방 백팩 1_MA-9901-D.Gray(+LK) 더블유팝'</li><li>'마젤란9901 메신저백 크로스백 학생 여행용 가방 백팩 1_MA-9901-Black(+LK) 더블유팝'</li></ul> |
| 9.0 | <ul><li>'룰루레몬 에브리웨어 벨트 백 Fleece WHTO/GOLD White Opal/Gold - O/S 오늘의원픽'</li><li>'[리본즈] LEMAIRE 남성 숄더백 37408558 블랙_ONE SIZE/단일상품 마리오아울렛몰'</li><li>'[코치][공식] 홀 벨트 백 CU103 WYE [00001] 없음 현대백화점'</li></ul> |
| 0.0 | <ul><li>'가죽가방끈 천연소가죽 가죽 스트랩 32Color 블랙12mm페이던트골드 대성메디칼'</li><li>'[최초가 228,000원][잘모이] 밍크 듀에 퍼 스트랩 LTZ-5205 168688 와인스카이블루 주식회사 미르에셋'</li><li>'[조이그라이슨](강남점) 첼시 스트랩 LW4SX6880_55 GOLD 신세계백화점'</li></ul> |
| 5.0 | <ul><li>'[소마치] 트래블 여권 지갑 파우치 핸드폰 미니 크로스백 카키_체인105cm(키160전후) 주식회사 소마치'</li><li>'비비안웨스트우드 코튼 숄더백 EDGWARE (3컬러) chacoal(당일발송) KHY INTERNATIONAL'</li><li>'남여 공용 미니 메신저백 귀여운 크로스백 학생 미니백 여행 보조 가방 여행용 보조백 아이보리 구공구코리아'</li></ul> |
| 2.0 | <ul><li>'메종미네드 MAISON MINED TWO POCKET BACKPACK S OC오피스'</li><li>'백팩01K1280ZSK외1종 블랙 롯데백화점1관'</li><li>'ANC CLASSIC BACKPACK_BLACK BLACK 주식회사 데일리컴퍼니'</li></ul> |
| 4.0 | <ul><li>'[스타벅스]텀블러 가방 컵홀더 데일리 캔버스 에코백 지퍼형_베이지 씨에스 인더스트리'</li><li>'마리떼 FRANCOIS GIRBAUD CLASSIC LOGO ECO BAG natural OS 다함'</li><li>'마크 곤잘레스 Print Eco Bag - 블랙 568032 BLACK_FREE 라임e커머스'</li></ul> |
| 8.0 | <ul><li>'국내생산 코튼 양줄면주머니 미니&에코 주머니 7종 학원 학교 만들기수업 양줄주머니_20cmX25cm(J20) 명성패키지'</li><li>'조리개 타입 반투명 파우치 보관 신발주머니 주머니 끈주머니 끈파우치 신주머니 여행용 중형(25X35) 정바른 길정'</li><li>'국내생산 코튼 화이트&블랙주머니 학원 학교 주머니만들기 W15_화이트 명성패키지'</li></ul> |
| 6.0 | <ul><li>'메종 마르지엘라 타비 스니커즈 S37WS0578 P4291 T1003 EU41(260-265) 보광컴퍼니'</li><li>'[롯데백화점]루이까또즈 클러치백 MO2DL03MDABL 롯데백화점_'</li><li>'깔끔한 여성용 데일리 핸드 스트랩 클러치 가방 남자클러치백 로우마켓'</li></ul> |
## Evaluation
### Metrics
| Label | Metric |
|:--------|:-------|
| **all** | 0.7868 |
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("mini1013/master_cate_ac9")
# Run inference
preds = model("웨빙 플라워 스트랩 레디백 길이조절 가방끈 어깨끈 리폼 3-플라워가방끈-흰색 이백프로")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Set Metrics
| Training set | Min | Median | Max |
|:-------------|:----|:-------|:----|
| Word count | 3 | 9.6146 | 30 |
| Label | Training Sample Count |
|:------|:----------------------|
| 0.0 | 50 |
| 1.0 | 17 |
| 2.0 | 50 |
| 3.0 | 50 |
| 4.0 | 50 |
| 5.0 | 50 |
| 6.0 | 50 |
| 7.0 | 50 |
| 8.0 | 50 |
| 9.0 | 50 |
### Training Hyperparameters
- batch_size: (512, 512)
- num_epochs: (20, 20)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 40
- body_learning_rate: (2e-05, 2e-05)
- head_learning_rate: 2e-05
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
### Training Results
| Epoch | Step | Training Loss | Validation Loss |
|:-------:|:----:|:-------------:|:---------------:|
| 0.0137 | 1 | 0.4278 | - |
| 0.6849 | 50 | 0.3052 | - |
| 1.3699 | 100 | 0.1524 | - |
| 2.0548 | 150 | 0.0583 | - |
| 2.7397 | 200 | 0.0292 | - |
| 3.4247 | 250 | 0.0197 | - |
| 4.1096 | 300 | 0.0061 | - |
| 4.7945 | 350 | 0.0022 | - |
| 5.4795 | 400 | 0.0033 | - |
| 6.1644 | 450 | 0.0003 | - |
| 6.8493 | 500 | 0.0002 | - |
| 7.5342 | 550 | 0.0001 | - |
| 8.2192 | 600 | 0.0001 | - |
| 8.9041 | 650 | 0.0001 | - |
| 9.5890 | 700 | 0.0001 | - |
| 10.2740 | 750 | 0.0001 | - |
| 10.9589 | 800 | 0.0001 | - |
| 11.6438 | 850 | 0.0001 | - |
| 12.3288 | 900 | 0.0001 | - |
| 13.0137 | 950 | 0.0001 | - |
| 13.6986 | 1000 | 0.0001 | - |
| 14.3836 | 1050 | 0.0001 | - |
| 15.0685 | 1100 | 0.0001 | - |
| 15.7534 | 1150 | 0.0001 | - |
| 16.4384 | 1200 | 0.0001 | - |
| 17.1233 | 1250 | 0.0 | - |
| 17.8082 | 1300 | 0.0001 | - |
| 18.4932 | 1350 | 0.0001 | - |
| 19.1781 | 1400 | 0.0001 | - |
| 19.8630 | 1450 | 0.0001 | - |
### Framework Versions
- Python: 3.10.12
- SetFit: 1.1.0.dev0
- Sentence Transformers: 3.1.1
- Transformers: 4.46.1
- PyTorch: 2.4.0+cu121
- Datasets: 2.20.0
- Tokenizers: 0.20.0
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SetFit with mini1013/master_domain
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [mini1013/master_domain](https://huggingface.co/mini1013/master_domain) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
- **Sentence Transformer body:** [mini1013/master_domain](https://huggingface.co/mini1013/master_domain)
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 512 tokens
- **Number of Classes:** 10 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
### Model Labels
| Label | Examples |
|:------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 3.0 | <ul><li>'[현대백화점][루이까또즈] MOONMOON(문문) 여성호보백 HR3SO02BL (주)현대백화점'</li><li>'소프트레더 파스텔 보부상 빅숄더백 휘뚜루마뚜루가방 토드백 블랙_one size 아이디어코리아 주식회사'</li><li>'DRAGON DIFFUSION 드래곤디퓨전 폼폼 더블 점프백 여성 버킷백 8838 드래곤백 다크브라운 (DARK BROWN) 시계1위워치짱'</li></ul> |
| 7.0 | <ul><li>'디어 4colors_H70301010 (W)퍼플와인 '</li><li>'[마이클코어스][정상가 1080000원] 에밀리아 라지 레더 사첼 35H0GU5S7T2171 신세계몰'</li><li>'칼린 소프트M 10colors _H71307020 (Y)라임네온_one size (주)칼린홍대점'</li></ul> |
| 1.0 | <ul><li>'마젤란9901 메신저백 크로스백 학생 여행용 가방 백팩 1_MA-9901-BlackPurple(+LK) 더블유팝'</li><li>'마젤란9901 메신저백 크로스백 학생 여행용 가방 백팩 1_MA-9901-D.Gray(+LK) 더블유팝'</li><li>'마젤란9901 메신저백 크로스백 학생 여행용 가방 백팩 1_MA-9901-Black(+LK) 더블유팝'</li></ul> |
| 9.0 | <ul><li>'룰루레몬 에브리웨어 벨트 백 Fleece WHTO/GOLD White Opal/Gold - O/S 오늘의원픽'</li><li>'[리본즈] LEMAIRE 남성 숄더백 37408558 블랙_ONE SIZE/단일상품 마리오아울렛몰'</li><li>'[코치][공식] 홀 벨트 백 CU103 WYE [00001] 없음 현대백화점'</li></ul> |
| 0.0 | <ul><li>'가죽가방끈 천연소가죽 가죽 스트랩 32Color 블랙12mm페이던트골드 대성메디칼'</li><li>'[최초가 228,000원][잘모이] 밍크 듀에 퍼 스트랩 LTZ-5205 168688 와인스카이블루 주식회사 미르에셋'</li><li>'[조이그라이슨](강남점) 첼시 스트랩 LW4SX6880_55 GOLD 신세계백화점'</li></ul> |
| 5.0 | <ul><li>'[소마치] 트래블 여권 지갑 파우치 핸드폰 미니 크로스백 카키_체인105cm(키160전후) 주식회사 소마치'</li><li>'비비안웨스트우드 코튼 숄더백 EDGWARE (3컬러) chacoal(당일발송) KHY INTERNATIONAL'</li><li>'남여 공용 미니 메신저백 귀여운 크로스백 학생 미니백 여행 보조 가방 여행용 보조백 아이보리 구공구코리아'</li></ul> |
| 2.0 | <ul><li>'메종미네드 MAISON MINED TWO POCKET BACKPACK S OC오피스'</li><li>'백팩01K1280ZSK외1종 블랙 롯데백화점1관'</li><li>'ANC CLASSIC BACKPACK_BLACK BLACK 주식회사 데일리컴퍼니'</li></ul> |
| 4.0 | <ul><li>'[스타벅스]텀블러 가방 컵홀더 데일리 캔버스 에코백 지퍼형_베이지 씨에스 인더스트리'</li><li>'마리떼 FRANCOIS GIRBAUD CLASSIC LOGO ECO BAG natural OS 다함'</li><li>'마크 곤잘레스 Print Eco Bag - 블랙 568032 BLACK_FREE 라임e커머스'</li></ul> |
| 8.0 | <ul><li>'국내생산 코튼 양줄면주머니 미니&에코 주머니 7종 학원 학교 만들기수업 양줄주머니_20cmX25cm(J20) 명성패키지'</li><li>'조리개 타입 반투명 파우치 보관 신발주머니 주머니 끈주머니 끈파우치 신주머니 여행용 중형(25X35) 정바른 길정'</li><li>'국내생산 코튼 화이트&블랙주머니 학원 학교 주머니만들기 W15_화이트 명성패키지'</li></ul> |
| 6.0 | <ul><li>'메종 마르지엘라 타비 스니커즈 S37WS0578 P4291 T1003 EU41(260-265) 보광컴퍼니'</li><li>'[롯데백화점]루이까또즈 클러치백 MO2DL03MDABL 롯데백화점_'</li><li>'깔끔한 여성용 데일리 핸드 스트랩 클러치 가방 남자클러치백 로우마켓'</li></ul> |
## Evaluation
### Metrics
| Label | Metric |
|:--------|:-------|
| **all** | 0.7868 |
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("mini1013/master_cate_ac9")
# Run inference
preds = model("웨빙 플라워 스트랩 레디백 길이조절 가방끈 어깨끈 리폼 3-플라워가방끈-흰색 이백프로")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Set Metrics
| Training set | Min | Median | Max |
|:-------------|:----|:-------|:----|
| Word count | 3 | 9.6146 | 30 |
| Label | Training Sample Count |
|:------|:----------------------|
| 0.0 | 50 |
| 1.0 | 17 |
| 2.0 | 50 |
| 3.0 | 50 |
| 4.0 | 50 |
| 5.0 | 50 |
| 6.0 | 50 |
| 7.0 | 50 |
| 8.0 | 50 |
| 9.0 | 50 |
### Training Hyperparameters
- batch_size: (512, 512)
- num_epochs: (20, 20)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 40
- body_learning_rate: (2e-05, 2e-05)
- head_learning_rate: 2e-05
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
### Training Results
| Epoch | Step | Training Loss | Validation Loss |
|:-------:|:----:|:-------------:|:---------------:|
| 0.0137 | 1 | 0.4278 | - |
| 0.6849 | 50 | 0.3052 | - |
| 1.3699 | 100 | 0.1524 | - |
| 2.0548 | 150 | 0.0583 | - |
| 2.7397 | 200 | 0.0292 | - |
| 3.4247 | 250 | 0.0197 | - |
| 4.1096 | 300 | 0.0061 | - |
| 4.7945 | 350 | 0.0022 | - |
| 5.4795 | 400 | 0.0033 | - |
| 6.1644 | 450 | 0.0003 | - |
| 6.8493 | 500 | 0.0002 | - |
| 7.5342 | 550 | 0.0001 | - |
| 8.2192 | 600 | 0.0001 | - |
| 8.9041 | 650 | 0.0001 | - |
| 9.5890 | 700 | 0.0001 | - |
| 10.2740 | 750 | 0.0001 | - |
| 10.9589 | 800 | 0.0001 | - |
| 11.6438 | 850 | 0.0001 | - |
| 12.3288 | 900 | 0.0001 | - |
| 13.0137 | 950 | 0.0001 | - |
| 13.6986 | 1000 | 0.0001 | - |
| 14.3836 | 1050 | 0.0001 | - |
| 15.0685 | 1100 | 0.0001 | - |
| 15.7534 | 1150 | 0.0001 | - |
| 16.4384 | 1200 | 0.0001 | - |
| 17.1233 | 1250 | 0.0 | - |
| 17.8082 | 1300 | 0.0001 | - |
| 18.4932 | 1350 | 0.0001 | - |
| 19.1781 | 1400 | 0.0001 | - |
| 19.8630 | 1450 | 0.0001 | - |
### Framework Versions
- Python: 3.10.12
- SetFit: 1.1.0.dev0
- Sentence Transformers: 3.1.1
- Transformers: 4.46.1
- PyTorch: 2.4.0+cu121
- Datasets: 2.20.0
- Tokenizers: 0.20.0
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "mini1013/master_domain", "library_name": "setfit", "metrics": ["metric"], "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification", "generated_from_setfit_trainer"], "widget": [{"text": "여성 가방 숄더백 미니 크로스백 퀼팅백 체인백 토트백 미니백 여자 핸드백 구름백 클러치백 직장인 백팩 프리아_카멜 더블유팝"}, {"text": "국내 잔스포츠 백팩 슈퍼브레이크 4QUT 블랙 학생 여성 가벼운 가방 캠핑 여행 당일 가원"}, {"text": "국내생산 코튼 양줄면주머니 미니&에코 주머니 7종 학원 학교 만들기수업 양줄주머니_14cmX28cm(J14) 명성패키지"}, {"text": "웨빙 플라워 스트랩 레디백 길이조절 가방끈 어깨끈 리폼 3-플라워가방끈-흰색 이백프로"}, {"text": "엔비조네/가방끈/가방끈리폼/가죽끈/크로스끈/숄더끈/스트랩 AOR오링25mm_블랙오플_폭11mm *35cm 니켈 엔비조네"}], "inference": true, "model-index": [{"name": "SetFit with mini1013/master_domain", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "Unknown", "type": "unknown", "split": "test"}, "metrics": [{"type": "metric", "value": 0.7867699642431466, "name": "Metric"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,892 |
rbarteau/distilbert-base-uncased-finetuned-emotion
|
rbarteau
|
text-classification
|
[
"transformers",
"tensorboard",
"safetensors",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:emotion",
"base_model:distilbert/distilbert-base-uncased",
"base_model:finetune:distilbert/distilbert-base-uncased",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-01-29T17:09:08Z |
2024-01-29T17:13:31+00:00
| 4 | 0 |
---
base_model: distilbert-base-uncased
datasets:
- emotion
license: apache-2.0
metrics:
- accuracy
- f1
tags:
- generated_from_trainer
model-index:
- name: distilbert-base-uncased-finetuned-emotion
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: emotion
type: emotion
config: split
split: validation
args: split
metrics:
- type: accuracy
value: 0.925
name: Accuracy
- type: f1
value: 0.9251576072485757
name: F1
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-emotion
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2200
- Accuracy: 0.925
- F1: 0.9252
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.8238 | 1.0 | 250 | 0.3215 | 0.9 | 0.8987 |
| 0.2505 | 2.0 | 500 | 0.2200 | 0.925 | 0.9252 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-emotion
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2200
- Accuracy: 0.925
- F1: 0.9252
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.8238 | 1.0 | 250 | 0.3215 | 0.9 | 0.8987 |
| 0.2505 | 2.0 | 500 | 0.2200 | 0.925 | 0.9252 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|
{"base_model": "distilbert-base-uncased", "datasets": ["emotion"], "license": "apache-2.0", "metrics": ["accuracy", "f1"], "tags": ["generated_from_trainer"], "model-index": [{"name": "distilbert-base-uncased-finetuned-emotion", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "emotion", "type": "emotion", "config": "split", "split": "validation", "args": "split"}, "metrics": [{"type": "accuracy", "value": 0.925, "name": "Accuracy"}, {"type": "f1", "value": 0.9251576072485757, "name": "F1"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,893 |
koushikkanch/machine_Translation
|
koushikkanch
| null |
[
"region:us"
] | 2024-11-16T01:42:05Z |
2024-11-16T02:11:00+00:00
| 0 | 0 |
---
{}
---
Here's the formatted version for your project documentation:
# English-Czech Machine Translation Project
## Overview
This project implements two approaches to English-Czech machine translation, comparing a simple LSTM model with a Sequence-to-Sequence model (without attention mechanism). Both implementations are optimized for GPU training, specifically targeting A100 hardware performance.
## Technical Implementation
### Data Processing
- **Format**: Parallel English-Czech text corpus
- **Processing Steps**:
* Cleaning and normalization
* Tokenization using HuggingFace tokenizers
* Train-validation split
* Conversion to PyTorch tensors
### Model Architectures
#### 1. LSTM Model
- Simple LSTM architecture for baseline performance
- Parameters:
* Hidden size: 128
* Batch size: 32
* Learning rate: 0.0001
* Optimizer: Adam
* Loss function: CrossEntropyLoss
#### 2. Seq2Seq Model (No Attention)
- Encoder-Decoder architecture
- Optimized Implementation:
* Hidden size: 256
* Batch size: 512
* Optimizer: AdamW with OneCycleLR scheduler
* Mixed precision training
* Gradient accumulation steps: 4
### Performance Optimization
- GPU memory management
- Mixed precision training
- Efficient data loading
- Gradient accumulation
- CUDA optimizations enabled
## Evaluation Metrics
Both models are evaluated using:
- BLEU Score calculation
- CHRF Score measurement
- Training and Validation losses
- Per-epoch metrics recording
## Project Structure
```
/content/drive/MyDrive/
├── Processed Dataset Files
│ ├── train_dataset.pt
│ └── val_dataset.pt
├── LSTM Results
│ ├── Model checkpoints
│ ├── Metrics CSV
│ └── Loss plots
└── Seq2Seq Results
├── Model checkpoints
├── Metrics CSV
└── Training curves
```
## Performance Tracking
- Per-epoch model checkpoints
- Comprehensive metrics logging
- Training curve visualization
- Translation quality assessment
## Dependencies
- PyTorch
- Transformers
- SacreBLEU
- Pandas
- Matplotlib
---
*Note: This implementation focuses on comparing basic LSTM and Seq2Seq architectures without attention mechanisms, optimized for GPU performance while maintaining translation quality.*
| null |
Non_BioNLP
|
Here's the formatted version for your project documentation:
# English-Czech Machine Translation Project
## Overview
This project implements two approaches to English-Czech machine translation, comparing a simple LSTM model with a Sequence-to-Sequence model (without attention mechanism). Both implementations are optimized for GPU training, specifically targeting A100 hardware performance.
## Technical Implementation
### Data Processing
- **Format**: Parallel English-Czech text corpus
- **Processing Steps**:
* Cleaning and normalization
* Tokenization using HuggingFace tokenizers
* Train-validation split
* Conversion to PyTorch tensors
### Model Architectures
#### 1. LSTM Model
- Simple LSTM architecture for baseline performance
- Parameters:
* Hidden size: 128
* Batch size: 32
* Learning rate: 0.0001
* Optimizer: Adam
* Loss function: CrossEntropyLoss
#### 2. Seq2Seq Model (No Attention)
- Encoder-Decoder architecture
- Optimized Implementation:
* Hidden size: 256
* Batch size: 512
* Optimizer: AdamW with OneCycleLR scheduler
* Mixed precision training
* Gradient accumulation steps: 4
### Performance Optimization
- GPU memory management
- Mixed precision training
- Efficient data loading
- Gradient accumulation
- CUDA optimizations enabled
## Evaluation Metrics
Both models are evaluated using:
- BLEU Score calculation
- CHRF Score measurement
- Training and Validation losses
- Per-epoch metrics recording
## Project Structure
```
/content/drive/MyDrive/
├── Processed Dataset Files
│ ├── train_dataset.pt
│ └── val_dataset.pt
├── LSTM Results
│ ├── Model checkpoints
│ ├── Metrics CSV
│ └── Loss plots
└── Seq2Seq Results
├── Model checkpoints
├── Metrics CSV
└── Training curves
```
## Performance Tracking
- Per-epoch model checkpoints
- Comprehensive metrics logging
- Training curve visualization
- Translation quality assessment
## Dependencies
- PyTorch
- Transformers
- SacreBLEU
- Pandas
- Matplotlib
---
*Note: This implementation focuses on comparing basic LSTM and Seq2Seq architectures without attention mechanisms, optimized for GPU performance while maintaining translation quality.*
|
{}
|
task
|
[
"TRANSLATION"
] | 39,894 |
UBC-NLP/AraT5v2-base-1024
|
UBC-NLP
| null |
[
"transformers",
"pytorch",
"t5",
"Arabic T5",
"MSA",
"Twitter",
"Arabic Dialect",
"Arabic Machine Translation",
"Arabic Text Summarization",
"Arabic News Title and Question Generation",
"Arabic Paraphrasing and Transliteration",
"Arabic Code-Switched Translation",
"ar",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-08-15T22:35:15Z |
2024-05-16T17:42:03+00:00
| 1,459 | 25 |
---
language:
- ar
tags:
- Arabic T5
- MSA
- Twitter
- Arabic Dialect
- Arabic Machine Translation
- Arabic Text Summarization
- Arabic News Title and Question Generation
- Arabic Paraphrasing and Transliteration
- Arabic Code-Switched Translation
---
# AraT5v2-base-1024
## What's new?
- **More Data.** `AraT5v2-base-1024` is trained on large and more diverse Arabic data.
- **Larger Sequence Length.** We increase the sequence length from 512 to 1024 in this version.
- **Faster Convergence.** On finetuning process, AraT5v2-base-1024 converges ~10x faster than the previous version (AraT5-base).
- **Extra IDs.** AraT5v2-base-1024 supports 100 sentinel tokens (a.k.a unique mask tokens).
<span style="color:red"><b>We recommend using AraT5v2-base-1024 instead of the previous version (AraT5-base).</b></span>
## An example of predicted masked token
```python
from transformers import T5Tokenizer, AutoModelForSeq2SeqLM
tokenizer = T5Tokenizer.from_pretrained("UBC-NLP/AraT5v2-base-1024")
model = AutoModelForSeq2SeqLM.from_pretrained("UBC-NLP/AraT5v2-base-1024")
ar_prompt="عاصمة ألمانيا هي <extra_id_0> "
input_ids = tokenizer(ar_prompt, return_tensors="pt").input_ids
outputs = model.generate(input_ids)
print("Tokenized input:", tokenizer.tokenize(ar_prompt))
print("Decoded output:", tokenizer.decode(outputs[0], skip_special_tokens=True))
```
Output:
```bash
Tokenized input: ['▁عاصمة', '▁ألمانيا', '▁هي', '<extra_id_0>']
Decoded output: برلين
```
# Citation
If you use our models for your scientific publication, or if you find the resources in this repository useful, please cite our papers as follows:
**(AraT5-base, AraT5-msa-base, AraT5-tweet-base, AraT5-msa-small, or AraT5-tweet-small)**
```bibtex
@inproceedings{nagoudi2022_arat5,
@inproceedings{nagoudi-etal-2022-arat5,
title = "{A}ra{T}5: Text-to-Text Transformers for {A}rabic Language Generation",
author = "Nagoudi, El Moatez Billah and
Elmadany, AbdelRahim and
Abdul-Mageed, Muhammad",
booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = may,
year = "2022",
address = "Dublin, Ireland",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.acl-long.47",
pages = "628--647",
abstract = "Transfer learning with a unified Transformer framework (T5) that converts all language problems into a text-to-text format was recently proposed as a simple and effective transfer learning approach. Although a multilingual version of the T5 model (mT5) was also introduced, it is not clear how well it can fare on non-English tasks involving diverse data. To investigate this question, we apply mT5 on a language with a wide variety of dialects{--}Arabic. For evaluation, we introduce a novel benchmark for ARabic language GENeration (ARGEN), covering seven important tasks. For model comparison, we pre-train three powerful Arabic T5-style models and evaluate them on ARGEN. Although pre-trained with {\textasciitilde}49 less data, our new models perform significantly better than mT5 on all ARGEN tasks (in 52 out of 59 test sets) and set several new SOTAs. Our models also establish new SOTA on the recently-proposed, large Arabic language understanding evaluation benchmark ARLUE (Abdul-Mageed et al., 2021). Our new models are publicly available. We also link to ARGEN datasets through our repository: https://github.com/UBC-NLP/araT5.",
}
```
**AraT5v2-base-1024**
``` bibtex
@inproceedings{elmadany-etal-2023-octopus,
title = "Octopus: A Multitask Model and Toolkit for {A}rabic Natural Language Generation",
author = "Elmadany, AbdelRahim and
Nagoudi, El Moatez Billah and
Abdul-Mageed, Muhammad",
booktitle = "Proceedings of ArabicNLP 2023",
month = dec,
year = "2023",
address = "Singapore (Hybrid)",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.arabicnlp-1.20",
doi = "10.18653/v1/2023.arabicnlp-1.20",
pages = "232--243",
}
```
| null |
Non_BioNLP
|
# AraT5v2-base-1024
## What's new?
- **More Data.** `AraT5v2-base-1024` is trained on large and more diverse Arabic data.
- **Larger Sequence Length.** We increase the sequence length from 512 to 1024 in this version.
- **Faster Convergence.** On finetuning process, AraT5v2-base-1024 converges ~10x faster than the previous version (AraT5-base).
- **Extra IDs.** AraT5v2-base-1024 supports 100 sentinel tokens (a.k.a unique mask tokens).
<span style="color:red"><b>We recommend using AraT5v2-base-1024 instead of the previous version (AraT5-base).</b></span>
## An example of predicted masked token
```python
from transformers import T5Tokenizer, AutoModelForSeq2SeqLM
tokenizer = T5Tokenizer.from_pretrained("UBC-NLP/AraT5v2-base-1024")
model = AutoModelForSeq2SeqLM.from_pretrained("UBC-NLP/AraT5v2-base-1024")
ar_prompt="عاصمة ألمانيا هي <extra_id_0> "
input_ids = tokenizer(ar_prompt, return_tensors="pt").input_ids
outputs = model.generate(input_ids)
print("Tokenized input:", tokenizer.tokenize(ar_prompt))
print("Decoded output:", tokenizer.decode(outputs[0], skip_special_tokens=True))
```
Output:
```bash
Tokenized input: ['▁عاصمة', '▁ألمانيا', '▁هي', '<extra_id_0>']
Decoded output: برلين
```
# Citation
If you use our models for your scientific publication, or if you find the resources in this repository useful, please cite our papers as follows:
**(AraT5-base, AraT5-msa-base, AraT5-tweet-base, AraT5-msa-small, or AraT5-tweet-small)**
```bibtex
@inproceedings{nagoudi2022_arat5,
@inproceedings{nagoudi-etal-2022-arat5,
title = "{A}ra{T}5: Text-to-Text Transformers for {A}rabic Language Generation",
author = "Nagoudi, El Moatez Billah and
Elmadany, AbdelRahim and
Abdul-Mageed, Muhammad",
booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = may,
year = "2022",
address = "Dublin, Ireland",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.acl-long.47",
pages = "628--647",
abstract = "Transfer learning with a unified Transformer framework (T5) that converts all language problems into a text-to-text format was recently proposed as a simple and effective transfer learning approach. Although a multilingual version of the T5 model (mT5) was also introduced, it is not clear how well it can fare on non-English tasks involving diverse data. To investigate this question, we apply mT5 on a language with a wide variety of dialects{--}Arabic. For evaluation, we introduce a novel benchmark for ARabic language GENeration (ARGEN), covering seven important tasks. For model comparison, we pre-train three powerful Arabic T5-style models and evaluate them on ARGEN. Although pre-trained with {\textasciitilde}49 less data, our new models perform significantly better than mT5 on all ARGEN tasks (in 52 out of 59 test sets) and set several new SOTAs. Our models also establish new SOTA on the recently-proposed, large Arabic language understanding evaluation benchmark ARLUE (Abdul-Mageed et al., 2021). Our new models are publicly available. We also link to ARGEN datasets through our repository: https://github.com/UBC-NLP/araT5.",
}
```
**AraT5v2-base-1024**
``` bibtex
@inproceedings{elmadany-etal-2023-octopus,
title = "Octopus: A Multitask Model and Toolkit for {A}rabic Natural Language Generation",
author = "Elmadany, AbdelRahim and
Nagoudi, El Moatez Billah and
Abdul-Mageed, Muhammad",
booktitle = "Proceedings of ArabicNLP 2023",
month = dec,
year = "2023",
address = "Singapore (Hybrid)",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.arabicnlp-1.20",
doi = "10.18653/v1/2023.arabicnlp-1.20",
pages = "232--243",
}
```
|
{"language": ["ar"], "tags": ["Arabic T5", "MSA", "Twitter", "Arabic Dialect", "Arabic Machine Translation", "Arabic Text Summarization", "Arabic News Title and Question Generation", "Arabic Paraphrasing and Transliteration", "Arabic Code-Switched Translation"]}
|
task
|
[
"TRANSLATION",
"SUMMARIZATION",
"PARAPHRASING"
] | 39,895 |
ishajo/autotrain-beproj_meeting_summarization_usingt5-49444119398
|
ishajo
|
summarization
|
[
"transformers",
"pytorch",
"t5",
"text2text-generation",
"autotrain",
"summarization",
"en",
"dataset:ishajo/autotrain-data-beproj_meeting_summarization_usingt5",
"co2_eq_emissions",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-04-14T11:02:18Z |
2023-04-14T11:45:22+00:00
| 20 | 0 |
---
datasets:
- ishajo/autotrain-data-beproj_meeting_summarization_usingt5
language:
- en
tags:
- autotrain
- summarization
widget:
- text: I love AutoTrain 🤗
co2_eq_emissions:
emissions: 14.626267232174746
---
# Model Trained Using AutoTrain
- Problem type: Summarization
- Model ID: 49444119398
- CO2 Emissions (in grams): 14.6263
## Validation Metrics
- Loss: 1.376
- Rouge1: 26.006
- Rouge2: 13.000
- RougeL: 21.526
- RougeLsum: 24.451
- Gen Len: 19.000
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_HUGGINGFACE_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/ishajo/autotrain-beproj_meeting_summarization_usingt5-49444119398
```
| null |
Non_BioNLP
|
# Model Trained Using AutoTrain
- Problem type: Summarization
- Model ID: 49444119398
- CO2 Emissions (in grams): 14.6263
## Validation Metrics
- Loss: 1.376
- Rouge1: 26.006
- Rouge2: 13.000
- RougeL: 21.526
- RougeLsum: 24.451
- Gen Len: 19.000
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_HUGGINGFACE_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/ishajo/autotrain-beproj_meeting_summarization_usingt5-49444119398
```
|
{"datasets": ["ishajo/autotrain-data-beproj_meeting_summarization_usingt5"], "language": ["en"], "tags": ["autotrain", "summarization"], "widget": [{"text": "I love AutoTrain 🤗"}], "co2_eq_emissions": {"emissions": 14.626267232174746}}
|
task
|
[
"SUMMARIZATION"
] | 39,896 |
Shaagun/Mistral_Instruc
|
Shaagun
| null |
[
"region:us"
] | 2024-12-08T01:33:56Z |
2024-12-08T01:40:09+00:00
| 0 | 0 |
---
{}
---
Model link - https://huggingface.co/Shaagun/mistral
Fine-Tuning Mistral 7B for Instruction-Following Tasks :
This project fine-tunes the Mistral 7B language model to enhance its ability to follow instructions in English and Lithuanian. The fine-tuned model is optimized to provide accurate, coherent, and contextually relevant responses across a variety of tasks.
Objective
The goal of this project is to:
Adapt the pre-trained Mistral 7B model to follow instructions effectively in multilingual contexts.
Enable the model to generate high-quality responses to both general and specific prompts while maintaining fluency and relevance.
Evaluate the fine-tuned model's performance using semantic similarity and qualitative metrics.
Dataset
The model was trained on a custom dataset of instruction-response pairs in English and Lithuanian. The dataset includes a diverse range of tasks such as:
Summarization
Question answering
Factual explanations
Each sample consists of:
Instruction: A task or question for the model.
Input: Optional context for the instruction.
Output: The expected response.
Training Details
The fine-tuning process was conducted using the SFTTrainer framework. Key hyperparameters used include:
Learning Rate: 3e-4
Batch Size: 4 (with gradient accumulation steps of 4)
Optimizer: adamw_8bit for memory efficiency
Number of Epochs: 1
Sequence Length: 1024
Precision: Mixed precision (fp16 or bf16 depending on hardware support)
Training was performed on a Google Colab T4 GPU, with careful tuning to balance memory constraints and performance.
Evaluation
The model was evaluated using:
Semantic Similarity:
Used Sentence Transformers (all-MiniLM-L6-v2) to compare generated outputs with reference responses.
Scores reflect how closely the generated text aligns with the meaning of the reference.
Results
The fine-tuned Mistral 7B model:
Demonstrates high accuracy and fluency in both English and Lithuanian.
Handles diverse tasks effectively, including complex instructions and translations.
Achieves strong semantic similarity scores across the evaluation dataset.
| null |
Non_BioNLP
|
Model link - https://huggingface.co/Shaagun/mistral
Fine-Tuning Mistral 7B for Instruction-Following Tasks :
This project fine-tunes the Mistral 7B language model to enhance its ability to follow instructions in English and Lithuanian. The fine-tuned model is optimized to provide accurate, coherent, and contextually relevant responses across a variety of tasks.
Objective
The goal of this project is to:
Adapt the pre-trained Mistral 7B model to follow instructions effectively in multilingual contexts.
Enable the model to generate high-quality responses to both general and specific prompts while maintaining fluency and relevance.
Evaluate the fine-tuned model's performance using semantic similarity and qualitative metrics.
Dataset
The model was trained on a custom dataset of instruction-response pairs in English and Lithuanian. The dataset includes a diverse range of tasks such as:
Summarization
Question answering
Factual explanations
Each sample consists of:
Instruction: A task or question for the model.
Input: Optional context for the instruction.
Output: The expected response.
Training Details
The fine-tuning process was conducted using the SFTTrainer framework. Key hyperparameters used include:
Learning Rate: 3e-4
Batch Size: 4 (with gradient accumulation steps of 4)
Optimizer: adamw_8bit for memory efficiency
Number of Epochs: 1
Sequence Length: 1024
Precision: Mixed precision (fp16 or bf16 depending on hardware support)
Training was performed on a Google Colab T4 GPU, with careful tuning to balance memory constraints and performance.
Evaluation
The model was evaluated using:
Semantic Similarity:
Used Sentence Transformers (all-MiniLM-L6-v2) to compare generated outputs with reference responses.
Scores reflect how closely the generated text aligns with the meaning of the reference.
Results
The fine-tuned Mistral 7B model:
Demonstrates high accuracy and fluency in both English and Lithuanian.
Handles diverse tasks effectively, including complex instructions and translations.
Achieves strong semantic similarity scores across the evaluation dataset.
|
{}
|
task
|
[
"QUESTION_ANSWERING",
"SEMANTIC_SIMILARITY",
"TRANSLATION",
"SUMMARIZATION"
] | 39,897 |
Helsinki-NLP/opus-mt-en-mg
|
Helsinki-NLP
|
translation
|
[
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"en",
"mg",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-03-02T23:29:04Z |
2023-08-16T11:30:27+00:00
| 37 | 0 |
---
license: apache-2.0
tags:
- translation
---
### opus-mt-en-mg
* source languages: en
* target languages: mg
* OPUS readme: [en-mg](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/en-mg/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-08.zip](https://object.pouta.csc.fi/OPUS-MT-models/en-mg/opus-2020-01-08.zip)
* test set translations: [opus-2020-01-08.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-mg/opus-2020-01-08.test.txt)
* test set scores: [opus-2020-01-08.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-mg/opus-2020-01-08.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| GlobalVoices.en.mg | 22.3 | 0.565 |
| Tatoeba.en.mg | 35.5 | 0.548 |
| null |
Non_BioNLP
|
### opus-mt-en-mg
* source languages: en
* target languages: mg
* OPUS readme: [en-mg](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/en-mg/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-08.zip](https://object.pouta.csc.fi/OPUS-MT-models/en-mg/opus-2020-01-08.zip)
* test set translations: [opus-2020-01-08.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-mg/opus-2020-01-08.test.txt)
* test set scores: [opus-2020-01-08.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-mg/opus-2020-01-08.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| GlobalVoices.en.mg | 22.3 | 0.565 |
| Tatoeba.en.mg | 35.5 | 0.548 |
|
{"license": "apache-2.0", "tags": ["translation"]}
|
task
|
[
"TRANSLATION"
] | 39,898 |
cardiffnlp/twitter-roberta-base-dec2021-topic-single
|
cardiffnlp
|
text-classification
|
[
"transformers",
"pytorch",
"roberta",
"text-classification",
"dataset:cardiffnlp/tweet_topic_single",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-12-01T19:23:28Z |
2022-12-01T19:26:39+00:00
| 11 | 0 |
---
datasets:
- cardiffnlp/tweet_topic_single
metrics:
- f1
- accuracy
pipeline_tag: text-classification
widget:
- text: Get the all-analog Classic Vinyl Edition of "Takin Off" Album from {@herbiehancock@}
via {@bluenoterecords@} link below {{URL}}
example_title: topic_classification 1
- text: Yes, including Medicare and social security saving👍
example_title: sentiment 1
- text: All two of them taste like ass.
example_title: offensive 1
- text: If you wanna look like a badass, have drama on social media
example_title: irony 1
- text: Whoever just unfollowed me you a bitch
example_title: hate 1
- text: I love swimming for the same reason I love meditating...the feeling of weightlessness.
example_title: emotion 1
- text: Beautiful sunset last night from the pontoon @TupperLakeNY
example_title: emoji 1
model-index:
- name: cardiffnlp/twitter-roberta-base-dec2021-topic-single
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: cardiffnlp/tweet_topic_single
type: cardiffnlp/tweet_topic_single
split: test_2021
metrics:
- type: micro_f1_cardiffnlp/tweet_topic_single
value: 0.896042528056704
name: Micro F1 (cardiffnlp/tweet_topic_single)
- type: micro_f1_cardiffnlp/tweet_topic_single
value: 0.7861641383871055
name: Macro F1 (cardiffnlp/tweet_topic_single)
- type: accuracy_cardiffnlp/tweet_topic_single
value: 0.896042528056704
name: Accuracy (cardiffnlp/tweet_topic_single)
---
# cardiffnlp/twitter-roberta-base-dec2021-topic-single
This model is a fine-tuned version of [cardiffnlp/twitter-roberta-base-dec2021](https://huggingface.co/cardiffnlp/twitter-roberta-base-dec2021) on the
[`cardiffnlp/tweet_topic_single`](https://huggingface.co/datasets/cardiffnlp/tweet_topic_single)
via [`tweetnlp`](https://github.com/cardiffnlp/tweetnlp).
Training split is `train_all` and parameters have been tuned on the validation split `validation_2021`.
Following metrics are achieved on the test split `test_2021` ([link](https://huggingface.co/cardiffnlp/twitter-roberta-base-dec2021-topic-single/raw/main/metric.json)).
- F1 (micro): 0.896042528056704
- F1 (macro): 0.7861641383871055
- Accuracy: 0.896042528056704
### Usage
Install tweetnlp via pip.
```shell
pip install tweetnlp
```
Load the model in python.
```python
import tweetnlp
model = tweetnlp.Classifier("cardiffnlp/twitter-roberta-base-dec2021-topic-single", max_length=128)
model.predict('Get the all-analog Classic Vinyl Edition of "Takin Off" Album from {@herbiehancock@} via {@bluenoterecords@} link below {{URL}}')
```
### Reference
```
@inproceedings{camacho-collados-etal-2022-tweetnlp,
title={{T}weet{NLP}: {C}utting-{E}dge {N}atural {L}anguage {P}rocessing for {S}ocial {M}edia},
author={Camacho-Collados, Jose and Rezaee, Kiamehr and Riahi, Talayeh and Ushio, Asahi and Loureiro, Daniel and Antypas, Dimosthenis and Boisson, Joanne and Espinosa-Anke, Luis and Liu, Fangyu and Mart{'\i}nez-C{'a}mara, Eugenio and others},
author = "Ushio, Asahi and
Camacho-Collados, Jose",
booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
month = nov,
year = "2022",
address = "Abu Dhabi, U.A.E.",
publisher = "Association for Computational Linguistics",
}
```
| null |
Non_BioNLP
|
# cardiffnlp/twitter-roberta-base-dec2021-topic-single
This model is a fine-tuned version of [cardiffnlp/twitter-roberta-base-dec2021](https://huggingface.co/cardiffnlp/twitter-roberta-base-dec2021) on the
[`cardiffnlp/tweet_topic_single`](https://huggingface.co/datasets/cardiffnlp/tweet_topic_single)
via [`tweetnlp`](https://github.com/cardiffnlp/tweetnlp).
Training split is `train_all` and parameters have been tuned on the validation split `validation_2021`.
Following metrics are achieved on the test split `test_2021` ([link](https://huggingface.co/cardiffnlp/twitter-roberta-base-dec2021-topic-single/raw/main/metric.json)).
- F1 (micro): 0.896042528056704
- F1 (macro): 0.7861641383871055
- Accuracy: 0.896042528056704
### Usage
Install tweetnlp via pip.
```shell
pip install tweetnlp
```
Load the model in python.
```python
import tweetnlp
model = tweetnlp.Classifier("cardiffnlp/twitter-roberta-base-dec2021-topic-single", max_length=128)
model.predict('Get the all-analog Classic Vinyl Edition of "Takin Off" Album from {@herbiehancock@} via {@bluenoterecords@} link below {{URL}}')
```
### Reference
```
@inproceedings{camacho-collados-etal-2022-tweetnlp,
title={{T}weet{NLP}: {C}utting-{E}dge {N}atural {L}anguage {P}rocessing for {S}ocial {M}edia},
author={Camacho-Collados, Jose and Rezaee, Kiamehr and Riahi, Talayeh and Ushio, Asahi and Loureiro, Daniel and Antypas, Dimosthenis and Boisson, Joanne and Espinosa-Anke, Luis and Liu, Fangyu and Mart{'\i}nez-C{'a}mara, Eugenio and others},
author = "Ushio, Asahi and
Camacho-Collados, Jose",
booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
month = nov,
year = "2022",
address = "Abu Dhabi, U.A.E.",
publisher = "Association for Computational Linguistics",
}
```
|
{"datasets": ["cardiffnlp/tweet_topic_single"], "metrics": ["f1", "accuracy"], "pipeline_tag": "text-classification", "widget": [{"text": "Get the all-analog Classic Vinyl Edition of \"Takin Off\" Album from {@herbiehancock@} via {@bluenoterecords@} link below {{URL}}", "example_title": "topic_classification 1"}, {"text": "Yes, including Medicare and social security saving👍", "example_title": "sentiment 1"}, {"text": "All two of them taste like ass.", "example_title": "offensive 1"}, {"text": "If you wanna look like a badass, have drama on social media", "example_title": "irony 1"}, {"text": "Whoever just unfollowed me you a bitch", "example_title": "hate 1"}, {"text": "I love swimming for the same reason I love meditating...the feeling of weightlessness.", "example_title": "emotion 1"}, {"text": "Beautiful sunset last night from the pontoon @TupperLakeNY", "example_title": "emoji 1"}], "model-index": [{"name": "cardiffnlp/twitter-roberta-base-dec2021-topic-single", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "cardiffnlp/tweet_topic_single", "type": "cardiffnlp/tweet_topic_single", "split": "test_2021"}, "metrics": [{"type": "micro_f1_cardiffnlp/tweet_topic_single", "value": 0.896042528056704, "name": "Micro F1 (cardiffnlp/tweet_topic_single)"}, {"type": "micro_f1_cardiffnlp/tweet_topic_single", "value": 0.7861641383871055, "name": "Macro F1 (cardiffnlp/tweet_topic_single)"}, {"type": "accuracy_cardiffnlp/tweet_topic_single", "value": 0.896042528056704, "name": "Accuracy (cardiffnlp/tweet_topic_single)"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,899 |
nugget00/marian-finetuned-kde4-en-to-fr
|
nugget00
|
translation
|
[
"transformers",
"pytorch",
"marian",
"text2text-generation",
"translation",
"generated_from_trainer",
"dataset:kde4",
"base_model:Helsinki-NLP/opus-mt-en-fr",
"base_model:finetune:Helsinki-NLP/opus-mt-en-fr",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-08-29T22:17:37Z |
2023-08-29T22:33:30+00:00
| 8 | 0 |
---
base_model: Helsinki-NLP/opus-mt-en-fr
datasets:
- kde4
license: apache-2.0
tags:
- translation
- generated_from_trainer
model-index:
- name: marian-finetuned-kde4-en-to-fr
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# marian-finetuned-kde4-en-to-fr
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on the kde4 dataset.
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Framework versions
- Transformers 4.32.0
- Pytorch 2.0.1
- Datasets 2.14.4
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# marian-finetuned-kde4-en-to-fr
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on the kde4 dataset.
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Framework versions
- Transformers 4.32.0
- Pytorch 2.0.1
- Datasets 2.14.4
- Tokenizers 0.13.3
|
{"base_model": "Helsinki-NLP/opus-mt-en-fr", "datasets": ["kde4"], "license": "apache-2.0", "tags": ["translation", "generated_from_trainer"], "model-index": [{"name": "marian-finetuned-kde4-en-to-fr", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,900 |
zli12321/prometheus2-2B
|
zli12321
| null |
[
"transformers",
"safetensors",
"mergekit",
"merge",
"en",
"dataset:prometheus-eval/Preference-Collection",
"dataset:prometheus-eval/Feedback-Collection",
"dataset:zli12321/pedants_qa_evaluation_bench",
"arxiv:2405.01535",
"base_model:google/gemma-2-2b-it",
"base_model:finetune:google/gemma-2-2b-it",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | 2024-08-13T23:44:48Z |
2025-02-11T23:53:15+00:00
| 107 | 0 |
---
base_model:
- google/gemma-2-2b-it
- google/gemma-2-2b-it
datasets:
- prometheus-eval/Preference-Collection
- prometheus-eval/Feedback-Collection
- zli12321/pedants_qa_evaluation_bench
language:
- en
library_name: transformers
license: apache-2.0
tags:
- mergekit
- merge
---
# prometheus-2-Gemma-2B-it
Finetuned gemma-2-2b-it of [prometheus-7b-v2.0](https://huggingface.co/prometheus-eval/prometheus-7b-v2.0) using [Gemma-2-2B-Instruct](https://huggingface.co/google/gemma-2-2b-it) as the base model.
Training hyperparameters:
* 3 epoch
* Learning rate 1e-5
* Effective batch size 4
* Cosine annealing
* ~5% warmup
Supports both feedback (likert-scale) evaluation and preference evaluation. Uses Gemma-2-2b-it Instruct the same prompts as prometheus-7b-v2.0. See example information below.
# Feedback Evaluation
```
ABSOLUTE_PROMPT = """###Task Description:
An instruction (might include an Input inside it), a response to evaluate, a reference answer that gets a score of 5, and a score rubric representing a evaluation criteria are given.
1. Write a detailed feedback that assess the quality of the response strictly based on the given score rubric, not evaluating in general.
2. After writing a feedback, write a score that is an integer between 1 and 5. You should refer to the score rubric.
3. The output format should look as follows: "Feedback: (write a feedback for criteria) [RESULT] (an integer number between 1 and 5)"
4. Please do not generate any other opening, closing, and explanations.
###The instruction to evaluate:
{}
###Response to evaluate:
{}
###Reference Answer (Score 5):
{}
###Score Rubrics:
{}
###Feedback: """
device = 'cuda:0'
model = AutoModelForCausalLM.from_pretrained("zli12321/prometheus2-2B").to(device)
tokenizer = AutoTokenizer.from_pretrained("zli12321/prometheus2-2B")
'''
Define your own instruction, response, reference, and rubric below
'''
prompt = ABSOLUTE_PROMPT.format(instruction, response, reference, rubric)
input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to(device)
input_length = input_ids.shape[1]
outputs = model.generate(input_ids, output_logits=True, return_dict_in_generate=True, max_new_tokens=4096)
print(tokenizer.decode(outputs.sequences[0], skip_special_tokens=True))
```
# Preference Evaluation Template
Follow the above to generate preference evaluation with the preference evaluation template.
```
###Task Description:
An instruction (might include an Input inside it), a response to evaluate, and a score rubric representing a evaluation criteria are given.
1. Write a detailed feedback that assess the quality of two responses strictly based on the given score rubric, not evaluating in general.
2. After writing a feedback, choose a better response between Response A and Response B. You should refer to the score rubric.
3. The output format should look as follows: "Feedback: (write a feedback for criteria) [RESULT] (A or B)"
4. Please do not generate any other opening, closing, and explanations.
###Instruction:
{}
###Response A:
{}
###Response B:
{}
###Reference Answer:
{}
###Score Rubric:
{}
###Feedback:
```
# Citations
```bibtex
@misc{kim2024prometheus,
title={Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models},
author={Seungone Kim and Juyoung Suk and Shayne Longpre and Bill Yuchen Lin and Jamin Shin and Sean Welleck and Graham Neubig and Moontae Lee and Kyungjae Lee and Minjoon Seo},
year={2024},
eprint={2405.01535},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
```bibtex
@inproceedings{li-etal-2024-pedants,
title = "{PEDANTS}: Cheap but Effective and Interpretable Answer Equivalence",
author = "Li, Zongxia and
Mondal, Ishani and
Nghiem, Huy and
Liang, Yijun and
Boyd-Graber, Jordan Lee",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.548/",
doi = "10.18653/v1/2024.findings-emnlp.548",
pages = "9373--9398",
abstract = "Question answering (QA) can only make progress if we know if an answer is correct, but current answer correctness (AC) metrics struggle with verbose, free-form answers from large language models (LLMs). There are two challenges with current short-form QA evaluations: a lack of diverse styles of evaluation data and an over-reliance on expensive and slow LLMs. LLM-based scorers correlate better with humans, but this expensive task has only been tested on limited QA datasets. We rectify these issues by providing rubrics and datasets for evaluating machine QA adopted from the Trivia community. We also propose an efficient, and interpretable QA evaluation that is more stable than an exact match and neural methods (BERTScore)."
}
```
| null |
Non_BioNLP
|
# prometheus-2-Gemma-2B-it
Finetuned gemma-2-2b-it of [prometheus-7b-v2.0](https://huggingface.co/prometheus-eval/prometheus-7b-v2.0) using [Gemma-2-2B-Instruct](https://huggingface.co/google/gemma-2-2b-it) as the base model.
Training hyperparameters:
* 3 epoch
* Learning rate 1e-5
* Effective batch size 4
* Cosine annealing
* ~5% warmup
Supports both feedback (likert-scale) evaluation and preference evaluation. Uses Gemma-2-2b-it Instruct the same prompts as prometheus-7b-v2.0. See example information below.
# Feedback Evaluation
```
ABSOLUTE_PROMPT = """###Task Description:
An instruction (might include an Input inside it), a response to evaluate, a reference answer that gets a score of 5, and a score rubric representing a evaluation criteria are given.
1. Write a detailed feedback that assess the quality of the response strictly based on the given score rubric, not evaluating in general.
2. After writing a feedback, write a score that is an integer between 1 and 5. You should refer to the score rubric.
3. The output format should look as follows: "Feedback: (write a feedback for criteria) [RESULT] (an integer number between 1 and 5)"
4. Please do not generate any other opening, closing, and explanations.
###The instruction to evaluate:
{}
###Response to evaluate:
{}
###Reference Answer (Score 5):
{}
###Score Rubrics:
{}
###Feedback: """
device = 'cuda:0'
model = AutoModelForCausalLM.from_pretrained("zli12321/prometheus2-2B").to(device)
tokenizer = AutoTokenizer.from_pretrained("zli12321/prometheus2-2B")
'''
Define your own instruction, response, reference, and rubric below
'''
prompt = ABSOLUTE_PROMPT.format(instruction, response, reference, rubric)
input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to(device)
input_length = input_ids.shape[1]
outputs = model.generate(input_ids, output_logits=True, return_dict_in_generate=True, max_new_tokens=4096)
print(tokenizer.decode(outputs.sequences[0], skip_special_tokens=True))
```
# Preference Evaluation Template
Follow the above to generate preference evaluation with the preference evaluation template.
```
###Task Description:
An instruction (might include an Input inside it), a response to evaluate, and a score rubric representing a evaluation criteria are given.
1. Write a detailed feedback that assess the quality of two responses strictly based on the given score rubric, not evaluating in general.
2. After writing a feedback, choose a better response between Response A and Response B. You should refer to the score rubric.
3. The output format should look as follows: "Feedback: (write a feedback for criteria) [RESULT] (A or B)"
4. Please do not generate any other opening, closing, and explanations.
###Instruction:
{}
###Response A:
{}
###Response B:
{}
###Reference Answer:
{}
###Score Rubric:
{}
###Feedback:
```
# Citations
```bibtex
@misc{kim2024prometheus,
title={Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models},
author={Seungone Kim and Juyoung Suk and Shayne Longpre and Bill Yuchen Lin and Jamin Shin and Sean Welleck and Graham Neubig and Moontae Lee and Kyungjae Lee and Minjoon Seo},
year={2024},
eprint={2405.01535},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
```bibtex
@inproceedings{li-etal-2024-pedants,
title = "{PEDANTS}: Cheap but Effective and Interpretable Answer Equivalence",
author = "Li, Zongxia and
Mondal, Ishani and
Nghiem, Huy and
Liang, Yijun and
Boyd-Graber, Jordan Lee",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.548/",
doi = "10.18653/v1/2024.findings-emnlp.548",
pages = "9373--9398",
abstract = "Question answering (QA) can only make progress if we know if an answer is correct, but current answer correctness (AC) metrics struggle with verbose, free-form answers from large language models (LLMs). There are two challenges with current short-form QA evaluations: a lack of diverse styles of evaluation data and an over-reliance on expensive and slow LLMs. LLM-based scorers correlate better with humans, but this expensive task has only been tested on limited QA datasets. We rectify these issues by providing rubrics and datasets for evaluating machine QA adopted from the Trivia community. We also propose an efficient, and interpretable QA evaluation that is more stable than an exact match and neural methods (BERTScore)."
}
```
|
{"base_model": ["google/gemma-2-2b-it", "google/gemma-2-2b-it"], "datasets": ["prometheus-eval/Preference-Collection", "prometheus-eval/Feedback-Collection", "zli12321/pedants_qa_evaluation_bench"], "language": ["en"], "library_name": "transformers", "license": "apache-2.0", "tags": ["mergekit", "merge"]}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,901 |
sarkerlab/SocBERT-final
|
sarkerlab
|
fill-mask
|
[
"transformers",
"pytorch",
"roberta",
"fill-mask",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-02-09T19:05:49Z |
2023-03-21T19:55:55+00:00
| 45 | 2 |
---
{}
---
# SocBERT model
Pretrained model on 20GB English tweets and 72GB Reddit comments using a masked language modeling (MLM) objective.
The tweets are from Archive and collected from Twitter Streaming API.
The Reddit comments are ramdonly sampled from all subreddits from 2015-2019.
SocBERT-base was pretrained on 819M sequence blocks for 100K steps.
SocBERT-final was pretrained on 929M (819M+110M) sequence blocks for 112K (100K+12K) steps.
We benchmarked SocBERT, on 40 text classification tasks with social media data.
The experiment results can be found in our paper:
```
@inproceedings{socbert:2023,
title = {{SocBERT: A Pretrained Model for Social Media Text}},
author = {Yuting Guo and Abeed Sarker},
booktitle = {Proceedings of the Fourth Workshop on Insights from Negative Results in NLP},
year = {2023}
}
```
A base version of the model can be found at [SocBERT-base](https://huggingface.co/sarkerlab/SocBERT-base).
| null |
Non_BioNLP
|
# SocBERT model
Pretrained model on 20GB English tweets and 72GB Reddit comments using a masked language modeling (MLM) objective.
The tweets are from Archive and collected from Twitter Streaming API.
The Reddit comments are ramdonly sampled from all subreddits from 2015-2019.
SocBERT-base was pretrained on 819M sequence blocks for 100K steps.
SocBERT-final was pretrained on 929M (819M+110M) sequence blocks for 112K (100K+12K) steps.
We benchmarked SocBERT, on 40 text classification tasks with social media data.
The experiment results can be found in our paper:
```
@inproceedings{socbert:2023,
title = {{SocBERT: A Pretrained Model for Social Media Text}},
author = {Yuting Guo and Abeed Sarker},
booktitle = {Proceedings of the Fourth Workshop on Insights from Negative Results in NLP},
year = {2023}
}
```
A base version of the model can be found at [SocBERT-base](https://huggingface.co/sarkerlab/SocBERT-base).
|
{}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,902 |
semindan/xnli_m_bert_only_vi
|
semindan
|
text-classification
|
[
"transformers",
"pytorch",
"bert",
"text-classification",
"generated_from_trainer",
"dataset:xnli",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-11-30T08:38:02Z |
2023-01-07T14:19:19+00:00
| 13 | 0 |
---
datasets:
- xnli
license: apache-2.0
metrics:
- accuracy
tags:
- text-classification
- generated_from_trainer
model-index:
- name: xnli_m_bert_only_vi
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: xnli
type: xnli
config: vi
split: train
args: vi
metrics:
- type: accuracy
value: 0.7401606425702811
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# xnli_m_bert_only_vi
This model is a fine-tuned version of [bert-base-multilingual-cased](https://huggingface.co/bert-base-multilingual-cased) on the xnli dataset.
It achieves the following results on the evaluation set:
- Loss: 1.2539
- Accuracy: 0.7402
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.6634 | 1.0 | 3068 | 0.7030 | 0.7016 |
| 0.5848 | 2.0 | 6136 | 0.6031 | 0.7518 |
| 0.5003 | 3.0 | 9204 | 0.6296 | 0.7418 |
| 0.4159 | 4.0 | 12272 | 0.6398 | 0.7482 |
| 0.3395 | 5.0 | 15340 | 0.7042 | 0.7438 |
| 0.2648 | 6.0 | 18408 | 0.9274 | 0.7345 |
| 0.2062 | 7.0 | 21476 | 0.9433 | 0.7373 |
| 0.1544 | 8.0 | 24544 | 1.0372 | 0.7378 |
| 0.1164 | 9.0 | 27612 | 1.1879 | 0.7357 |
| 0.0882 | 10.0 | 30680 | 1.2539 | 0.7402 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0
- Datasets 2.6.1
- Tokenizers 0.13.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# xnli_m_bert_only_vi
This model is a fine-tuned version of [bert-base-multilingual-cased](https://huggingface.co/bert-base-multilingual-cased) on the xnli dataset.
It achieves the following results on the evaluation set:
- Loss: 1.2539
- Accuracy: 0.7402
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.6634 | 1.0 | 3068 | 0.7030 | 0.7016 |
| 0.5848 | 2.0 | 6136 | 0.6031 | 0.7518 |
| 0.5003 | 3.0 | 9204 | 0.6296 | 0.7418 |
| 0.4159 | 4.0 | 12272 | 0.6398 | 0.7482 |
| 0.3395 | 5.0 | 15340 | 0.7042 | 0.7438 |
| 0.2648 | 6.0 | 18408 | 0.9274 | 0.7345 |
| 0.2062 | 7.0 | 21476 | 0.9433 | 0.7373 |
| 0.1544 | 8.0 | 24544 | 1.0372 | 0.7378 |
| 0.1164 | 9.0 | 27612 | 1.1879 | 0.7357 |
| 0.0882 | 10.0 | 30680 | 1.2539 | 0.7402 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0
- Datasets 2.6.1
- Tokenizers 0.13.1
|
{"datasets": ["xnli"], "license": "apache-2.0", "metrics": ["accuracy"], "tags": ["text-classification", "generated_from_trainer"], "model-index": [{"name": "xnli_m_bert_only_vi", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "xnli", "type": "xnli", "config": "vi", "split": "train", "args": "vi"}, "metrics": [{"type": "accuracy", "value": 0.7401606425702811, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,903 |
gokulsrinivasagan/bert_base_lda_stsb
|
gokulsrinivasagan
|
text-classification
|
[
"transformers",
"tensorboard",
"safetensors",
"distilbert",
"text-classification",
"generated_from_trainer",
"en",
"dataset:glue",
"base_model:gokulsrinivasagan/bert_base_lda",
"base_model:finetune:gokulsrinivasagan/bert_base_lda",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-11-22T11:18:11Z |
2024-12-04T18:47:55+00:00
| 7 | 0 |
---
base_model: gokulsrinivasagan/bert_base_lda
datasets:
- glue
language:
- en
library_name: transformers
metrics:
- spearmanr
tags:
- generated_from_trainer
model-index:
- name: bert_base_lda_stsb
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: GLUE STSB
type: glue
args: stsb
metrics:
- type: spearmanr
value: 0.531029805524994
name: Spearmanr
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert_base_lda_stsb
This model is a fine-tuned version of [gokulsrinivasagan/bert_base_lda](https://huggingface.co/gokulsrinivasagan/bert_base_lda) on the GLUE STSB dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6844
- Pearson: 0.5377
- Spearmanr: 0.5310
- Combined Score: 0.5344
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 50
### Training results
| Training Loss | Epoch | Step | Validation Loss | Pearson | Spearmanr | Combined Score |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:---------:|:--------------:|
| 2.6177 | 1.0 | 23 | 2.5040 | 0.0492 | 0.0656 | 0.0574 |
| 1.9612 | 2.0 | 46 | 2.0348 | 0.3695 | 0.3669 | 0.3682 |
| 1.5453 | 3.0 | 69 | 2.0887 | 0.4294 | 0.4268 | 0.4281 |
| 1.1331 | 4.0 | 92 | 2.0548 | 0.4989 | 0.5093 | 0.5041 |
| 0.8134 | 5.0 | 115 | 1.8755 | 0.4933 | 0.4930 | 0.4931 |
| 0.7478 | 6.0 | 138 | 2.0804 | 0.5093 | 0.5092 | 0.5093 |
| 0.5429 | 7.0 | 161 | 1.9448 | 0.5262 | 0.5261 | 0.5261 |
| 0.4469 | 8.0 | 184 | 1.8867 | 0.5136 | 0.5118 | 0.5127 |
| 0.4011 | 9.0 | 207 | 1.8538 | 0.5243 | 0.5185 | 0.5214 |
| 0.3255 | 10.0 | 230 | 1.8652 | 0.5373 | 0.5320 | 0.5346 |
| 0.271 | 11.0 | 253 | 2.0402 | 0.5128 | 0.5043 | 0.5085 |
| 0.2699 | 12.0 | 276 | 1.8057 | 0.5305 | 0.5241 | 0.5273 |
| 0.2086 | 13.0 | 299 | 1.8275 | 0.5225 | 0.5190 | 0.5207 |
| 0.1951 | 14.0 | 322 | 2.0355 | 0.5140 | 0.5076 | 0.5108 |
| 0.1875 | 15.0 | 345 | 1.7243 | 0.5372 | 0.5323 | 0.5347 |
| 0.17 | 16.0 | 368 | 1.7595 | 0.5392 | 0.5342 | 0.5367 |
| 0.149 | 17.0 | 391 | 1.7326 | 0.5433 | 0.5387 | 0.5410 |
| 0.1476 | 18.0 | 414 | 1.6844 | 0.5377 | 0.5310 | 0.5344 |
| 0.127 | 19.0 | 437 | 1.7534 | 0.5277 | 0.5208 | 0.5242 |
| 0.1217 | 20.0 | 460 | 1.9471 | 0.5275 | 0.5187 | 0.5231 |
| 0.1182 | 21.0 | 483 | 1.8289 | 0.5323 | 0.5242 | 0.5282 |
| 0.0989 | 22.0 | 506 | 1.8270 | 0.5209 | 0.5128 | 0.5168 |
| 0.1017 | 23.0 | 529 | 1.8574 | 0.5309 | 0.5219 | 0.5264 |
### Framework versions
- Transformers 4.46.3
- Pytorch 2.2.1+cu118
- Datasets 2.17.0
- Tokenizers 0.20.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert_base_lda_stsb
This model is a fine-tuned version of [gokulsrinivasagan/bert_base_lda](https://huggingface.co/gokulsrinivasagan/bert_base_lda) on the GLUE STSB dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6844
- Pearson: 0.5377
- Spearmanr: 0.5310
- Combined Score: 0.5344
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 50
### Training results
| Training Loss | Epoch | Step | Validation Loss | Pearson | Spearmanr | Combined Score |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:---------:|:--------------:|
| 2.6177 | 1.0 | 23 | 2.5040 | 0.0492 | 0.0656 | 0.0574 |
| 1.9612 | 2.0 | 46 | 2.0348 | 0.3695 | 0.3669 | 0.3682 |
| 1.5453 | 3.0 | 69 | 2.0887 | 0.4294 | 0.4268 | 0.4281 |
| 1.1331 | 4.0 | 92 | 2.0548 | 0.4989 | 0.5093 | 0.5041 |
| 0.8134 | 5.0 | 115 | 1.8755 | 0.4933 | 0.4930 | 0.4931 |
| 0.7478 | 6.0 | 138 | 2.0804 | 0.5093 | 0.5092 | 0.5093 |
| 0.5429 | 7.0 | 161 | 1.9448 | 0.5262 | 0.5261 | 0.5261 |
| 0.4469 | 8.0 | 184 | 1.8867 | 0.5136 | 0.5118 | 0.5127 |
| 0.4011 | 9.0 | 207 | 1.8538 | 0.5243 | 0.5185 | 0.5214 |
| 0.3255 | 10.0 | 230 | 1.8652 | 0.5373 | 0.5320 | 0.5346 |
| 0.271 | 11.0 | 253 | 2.0402 | 0.5128 | 0.5043 | 0.5085 |
| 0.2699 | 12.0 | 276 | 1.8057 | 0.5305 | 0.5241 | 0.5273 |
| 0.2086 | 13.0 | 299 | 1.8275 | 0.5225 | 0.5190 | 0.5207 |
| 0.1951 | 14.0 | 322 | 2.0355 | 0.5140 | 0.5076 | 0.5108 |
| 0.1875 | 15.0 | 345 | 1.7243 | 0.5372 | 0.5323 | 0.5347 |
| 0.17 | 16.0 | 368 | 1.7595 | 0.5392 | 0.5342 | 0.5367 |
| 0.149 | 17.0 | 391 | 1.7326 | 0.5433 | 0.5387 | 0.5410 |
| 0.1476 | 18.0 | 414 | 1.6844 | 0.5377 | 0.5310 | 0.5344 |
| 0.127 | 19.0 | 437 | 1.7534 | 0.5277 | 0.5208 | 0.5242 |
| 0.1217 | 20.0 | 460 | 1.9471 | 0.5275 | 0.5187 | 0.5231 |
| 0.1182 | 21.0 | 483 | 1.8289 | 0.5323 | 0.5242 | 0.5282 |
| 0.0989 | 22.0 | 506 | 1.8270 | 0.5209 | 0.5128 | 0.5168 |
| 0.1017 | 23.0 | 529 | 1.8574 | 0.5309 | 0.5219 | 0.5264 |
### Framework versions
- Transformers 4.46.3
- Pytorch 2.2.1+cu118
- Datasets 2.17.0
- Tokenizers 0.20.3
|
{"base_model": "gokulsrinivasagan/bert_base_lda", "datasets": ["glue"], "language": ["en"], "library_name": "transformers", "metrics": ["spearmanr"], "tags": ["generated_from_trainer"], "model-index": [{"name": "bert_base_lda_stsb", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "GLUE STSB", "type": "glue", "args": "stsb"}, "metrics": [{"type": "spearmanr", "value": 0.531029805524994, "name": "Spearmanr"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,904 |
hoang14/pegasus-finetuned-samsum
|
hoang14
|
text2text-generation
|
[
"transformers",
"pytorch",
"pegasus",
"text2text-generation",
"en",
"dataset:samsum",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-04-11T09:39:45Z |
2023-04-11T10:20:05+00:00
| 8 | 0 |
---
datasets:
- samsum
language:
- en
license: apache-2.0
metrics:
- rouge
pipeline_tag: text2text-generation
---
Summarization model based on pegasus, finetuned on samsum dataset
source code: https://colab.research.google.com/drive/1FxdOV1fiHY3JC6dFw5T-NED1J8dKKHSO#scrollTo=pgdQ2up7vJoU
metrics on samsum dataset:
- rouge1: 0.436239
- rouge2: 0.209266
- rougeL: 0.34446
- rougeLsum: 0.344428
| null |
Non_BioNLP
|
Summarization model based on pegasus, finetuned on samsum dataset
source code: https://colab.research.google.com/drive/1FxdOV1fiHY3JC6dFw5T-NED1J8dKKHSO#scrollTo=pgdQ2up7vJoU
metrics on samsum dataset:
- rouge1: 0.436239
- rouge2: 0.209266
- rougeL: 0.34446
- rougeLsum: 0.344428
|
{"datasets": ["samsum"], "language": ["en"], "license": "apache-2.0", "metrics": ["rouge"], "pipeline_tag": "text2text-generation"}
|
task
|
[
"SUMMARIZATION"
] | 39,905 |
RichardErkhov/CohereForAI_-_c4ai-command-r-v01-4bits
|
RichardErkhov
|
text-generation
|
[
"transformers",
"safetensors",
"cohere",
"text-generation",
"conversational",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"4-bit",
"bitsandbytes",
"region:us"
] | 2024-05-03T10:12:48Z |
2024-05-03T10:53:57+00:00
| 12 | 0 |
---
{}
---
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
c4ai-command-r-v01 - bnb 4bits
- Model creator: https://huggingface.co/CohereForAI/
- Original model: https://huggingface.co/CohereForAI/c4ai-command-r-v01/
Original model description:
---
library_name: transformers
language:
- en
- fr
- de
- es
- it
- pt
- ja
- ko
- zh
- ar
license: cc-by-nc-4.0
---
# Model Card for C4AI Command-R
🚨 **This model is non-quantized version of C4AI Command-R. You can find the quantized version of C4AI Command-R using bitsandbytes [here](https://huggingface.co/CohereForAI/c4ai-command-r-v01-4bit)**.
## Model Summary
C4AI Command-R is a research release of a 35 billion parameter highly performant generative model. Command-R is a large language model with open weights optimized for a variety of use cases including reasoning, summarization, and question answering. Command-R has the capability for multilingual generation evaluated in 10 languages and highly performant RAG capabilities.
Developed by: Cohere and [Cohere For AI](https://cohere.for.ai)
- Point of Contact: Cohere For AI: [cohere.for.ai](https://cohere.for.ai/)
- License: [CC-BY-NC](https://cohere.com/c4ai-cc-by-nc-license), requires also adhering to [C4AI's Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy)
- Model: c4ai-command-r-v01
- Model Size: 35 billion parameters
- Context length: 128K
**Try C4AI Command R**
If you want to try Command R before downloading the weights, the model is hosted in a hugging face space [here](https://huggingface.co/spaces/CohereForAI/c4ai-command-r-v01).
**Usage**
Please use `transformers` version 4.39.1 or higher
```python
# pip install 'transformers>=4.39.1'
from transformers import AutoTokenizer, AutoModelForCausalLM
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
# Format message with the command-r chat template
messages = [{"role": "user", "content": "Hello, how are you?"}]
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
gen_tokens = model.generate(
input_ids,
max_new_tokens=100,
do_sample=True,
temperature=0.3,
)
gen_text = tokenizer.decode(gen_tokens[0])
print(gen_text)
```
**Quantized model through bitsandbytes, 8-bit precision**
```python
# pip install 'transformers>=4.39.1' bitsandbytes accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
bnb_config = BitsAndBytesConfig(load_in_8bit=True)
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=bnb_config)
# Format message with the command-r chat template
messages = [{"role": "user", "content": "Hello, how are you?"}]
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
gen_tokens = model.generate(
input_ids,
max_new_tokens=100,
do_sample=True,
temperature=0.3,
)
gen_text = tokenizer.decode(gen_tokens[0])
print(gen_text)
```
**Quantized model through bitsandbytes, 4-bit precision**
You can find a quantized version of this model to 4-bit precision [here](https://huggingface.co/CohereForAI/c4ai-command-r-v01-4bit).
## Model Details
**Input**: Models input text only.
**Output**: Models generate text only.
**Model Architecture**: This is an auto-regressive language model that uses an optimized transformer architecture. After pretraining, this model uses supervised fine-tuning (SFT) and preference training to align model behavior to human preferences for helpfulness and safety.
**Languages covered**: The model is optimized to perform well in the following languages: English, French, Spanish, Italian, German, Brazilian Portuguese, Japanese, Korean, Simplified Chinese, and Arabic.
Pre-training data additionally included the following 13 languages: Russian, Polish, Turkish, Vietnamese, Dutch, Czech, Indonesian, Ukrainian, Romanian, Greek, Hindi, Hebrew, Persian.
**Context length**: Command-R supports a context length of 128K.
### Tool use capabilities:
Command-R has been specifically trained with conversational tool use capabilities. These have been trained into the model via a mixture of supervised fine-tuning and preference fine-tuning, using a specific prompt template. Deviating from this prompt template will likely reduce performance, but we encourage experimentation.
Command-R’s tool use functionality takes a conversation as input (with an optional user-system preamble), along with a list of available tools. The model will then generate a json-formatted list of actions to execute on a subset of those tools. Command-R may use one of its supplied tools more than once.
The model has been trained to recognise a special `directly_answer` tool, which it uses to indicate that it doesn’t want to use any of its other tools. The ability to abstain from calling a specific tool can be useful in a range of situations, such as greeting a user, or asking clarifying questions.
We recommend including the `directly_answer` tool, but it can be removed or renamed if required.
Comprehensive documentation for working with command-R's tool use prompt template can be found [here](https://docs.cohere.com/docs/prompting-command-r).
The code snippet below shows a minimal working example on how to render a prompt.
<details>
<summary><b>Usage: Rendering Tool Use Prompts [CLICK TO EXPAND]</b> </summary>
```python
from transformers import AutoTokenizer
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
# define conversation input:
conversation = [
{"role": "user", "content": "Whats the biggest penguin in the world?"}
]
# Define tools available for the model to use:
tools = [
{
"name": "internet_search",
"description": "Returns a list of relevant document snippets for a textual query retrieved from the internet",
"parameter_definitions": {
"query": {
"description": "Query to search the internet with",
"type": 'str',
"required": True
}
}
},
{
'name': "directly_answer",
"description": "Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history",
'parameter_definitions': {}
}
]
# render the tool use prompt as a string:
tool_use_prompt = tokenizer.apply_tool_use_template(
conversation,
tools=tools,
tokenize=False,
add_generation_prompt=True,
)
print(tool_use_prompt)
```
</details>
<details>
<summary><b>Example Rendered Tool Use Prompt [CLICK TO EXPAND]</b></summary>
````
<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
# System Preamble
## Basic Rules
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
# User Preamble
## Task and Context
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
## Style Guide
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.
## Available Tools
Here is a list of tools that you have available to you:
```python
def internet_search(query: str) -> List[Dict]:
"""Returns a list of relevant document snippets for a textual query retrieved from the internet
Args:
query (str): Query to search the internet with
"""
pass
```
```python
def directly_answer() -> List[Dict]:
"""Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history
"""
pass
```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
```json
[
{
"tool_name": title of the tool in the specification,
"parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
}
]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
````
</details>
<details>
<summary><b>Example Rendered Tool Use Completion [CLICK TO EXPAND]</b></summary>
````
Action: ```json
[
{
"tool_name": "internet_search",
"parameters": {
"query": "biggest penguin in the world"
}
}
]
```
````
</details>
### Grounded Generation and RAG Capabilities:
Command-R has been specifically trained with grounded generation capabilities. This means that it can generate responses based on a list of supplied document snippets, and it will include grounding spans (citations) in its response indicating the source of the information.
This can be used to enable behaviors such as grounded summarization and the final step of Retrieval Augmented Generation (RAG).This behavior has been trained into the model via a mixture of supervised fine-tuning and preference fine-tuning, using a specific prompt template.
Deviating from this prompt template may reduce performance, but we encourage experimentation.
Command-R’s grounded generation behavior takes a conversation as input (with an optional user-supplied system preamble, indicating task, context and desired output style), along with a list of retrieved document snippets.
The document snippets should be chunks, rather than long documents, typically around 100-400 words per chunk. Document snippets consist of key-value pairs. The keys should be short descriptive strings, the values can be text or semi-structured.
By default, Command-R will generate grounded responses by first predicting which documents are relevant, then predicting which ones it will cite, then generating an answer.
Finally, it will then insert grounding spans into the answer. See below for an example. This is referred to as `accurate` grounded generation.
The model is trained with a number of other answering modes, which can be selected by prompt changes . A `fast` citation mode is supported in the tokenizer, which will directly generate an answer with grounding spans in it, without first writing the answer out in full. This sacrifices some grounding accuracy in favor of generating fewer tokens.
Comprehensive documentation for working with command-R's grounded generation prompt template can be found [here](https://docs.cohere.com/docs/prompting-command-r).
The code snippet below shows a minimal working example on how to render a prompt.
<details>
<summary> <b>Usage: Rendering Grounded Generation prompts [CLICK TO EXPAND]</b> </summary>
````python
from transformers import AutoTokenizer
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
# define conversation input:
conversation = [
{"role": "user", "content": "Whats the biggest penguin in the world?"}
]
# define documents to ground on:
documents = [
{ "title": "Tall penguins", "text": "Emperor penguins are the tallest growing up to 122 cm in height." },
{ "title": "Penguin habitats", "text": "Emperor penguins only live in Antarctica."}
]
# render the tool use prompt as a string:
grounded_generation_prompt = tokenizer.apply_grounded_generation_template(
conversation,
documents=documents,
citation_mode="accurate", # or "fast"
tokenize=False,
add_generation_prompt=True,
)
print(grounded_generation_prompt)
````
</details>
<details>
<summary><b>Example Rendered Grounded Generation Prompt [CLICK TO EXPAND]</b></summary>
````<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
# System Preamble
## Basic Rules
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
# User Preamble
## Task and Context
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
## Style Guide
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
Document: 0
title: Tall penguins
text: Emperor penguins are the tallest growing up to 122 cm in height.
Document: 1
title: Penguin habitats
text: Emperor penguins only live in Antarctica.
</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
Thirdly, Write 'Answer:' followed by a response to the user's last input in high quality natural english. Use the retrieved documents to help you. Do not insert any citations or grounding markup.
Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
````
</details>
<details>
<summary><b>Example Rendered Grounded Generation Completion [CLICK TO EXPAND]</b></summary>
````
Relevant Documents: 0,1
Cited Documents: 0,1
Answer: The Emperor Penguin is the tallest or biggest penguin in the world. It is a bird that lives only in Antarctica and grows to a height of around 122 centimetres.
Grounded answer: The <co: 0>Emperor Penguin</co: 0> is the <co: 0>tallest</co: 0> or biggest penguin in the world. It is a bird that <co: 1>lives only in Antarctica</co: 1> and <co: 0>grows to a height of around 122 centimetres.</co: 0>
````
</details>
### Code Capabilities:
Command-R has been optimized to interact with your code, by requesting code snippets, code explanations, or code rewrites. It might not perform well out-of-the-box for pure code completion. For better performance, we also recommend using a low temperature (and even greedy decoding) for code-generation related instructions.
### Model Card Contact
For errors or additional questions about details in this model card, contact [[email protected]](mailto:[email protected]).
### Terms of Use:
We hope that the release of this model will make community-based research efforts more accessible, by releasing the weights of a highly performant 35 billion parameter model to researchers all over the world. This model is governed by a [CC-BY-NC](https://cohere.com/c4ai-cc-by-nc-license) License with an acceptable use addendum, and also requires adhering to [C4AI's Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).
### Try Chat:
You can try Command-R chat in the playground [here](https://dashboard.cohere.com/playground/chat).
| null |
Non_BioNLP
|
Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
c4ai-command-r-v01 - bnb 4bits
- Model creator: https://huggingface.co/CohereForAI/
- Original model: https://huggingface.co/CohereForAI/c4ai-command-r-v01/
Original model description:
---
library_name: transformers
language:
- en
- fr
- de
- es
- it
- pt
- ja
- ko
- zh
- ar
license: cc-by-nc-4.0
---
# Model Card for C4AI Command-R
🚨 **This model is non-quantized version of C4AI Command-R. You can find the quantized version of C4AI Command-R using bitsandbytes [here](https://huggingface.co/CohereForAI/c4ai-command-r-v01-4bit)**.
## Model Summary
C4AI Command-R is a research release of a 35 billion parameter highly performant generative model. Command-R is a large language model with open weights optimized for a variety of use cases including reasoning, summarization, and question answering. Command-R has the capability for multilingual generation evaluated in 10 languages and highly performant RAG capabilities.
Developed by: Cohere and [Cohere For AI](https://cohere.for.ai)
- Point of Contact: Cohere For AI: [cohere.for.ai](https://cohere.for.ai/)
- License: [CC-BY-NC](https://cohere.com/c4ai-cc-by-nc-license), requires also adhering to [C4AI's Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy)
- Model: c4ai-command-r-v01
- Model Size: 35 billion parameters
- Context length: 128K
**Try C4AI Command R**
If you want to try Command R before downloading the weights, the model is hosted in a hugging face space [here](https://huggingface.co/spaces/CohereForAI/c4ai-command-r-v01).
**Usage**
Please use `transformers` version 4.39.1 or higher
```python
# pip install 'transformers>=4.39.1'
from transformers import AutoTokenizer, AutoModelForCausalLM
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
# Format message with the command-r chat template
messages = [{"role": "user", "content": "Hello, how are you?"}]
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
gen_tokens = model.generate(
input_ids,
max_new_tokens=100,
do_sample=True,
temperature=0.3,
)
gen_text = tokenizer.decode(gen_tokens[0])
print(gen_text)
```
**Quantized model through bitsandbytes, 8-bit precision**
```python
# pip install 'transformers>=4.39.1' bitsandbytes accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
bnb_config = BitsAndBytesConfig(load_in_8bit=True)
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=bnb_config)
# Format message with the command-r chat template
messages = [{"role": "user", "content": "Hello, how are you?"}]
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
gen_tokens = model.generate(
input_ids,
max_new_tokens=100,
do_sample=True,
temperature=0.3,
)
gen_text = tokenizer.decode(gen_tokens[0])
print(gen_text)
```
**Quantized model through bitsandbytes, 4-bit precision**
You can find a quantized version of this model to 4-bit precision [here](https://huggingface.co/CohereForAI/c4ai-command-r-v01-4bit).
## Model Details
**Input**: Models input text only.
**Output**: Models generate text only.
**Model Architecture**: This is an auto-regressive language model that uses an optimized transformer architecture. After pretraining, this model uses supervised fine-tuning (SFT) and preference training to align model behavior to human preferences for helpfulness and safety.
**Languages covered**: The model is optimized to perform well in the following languages: English, French, Spanish, Italian, German, Brazilian Portuguese, Japanese, Korean, Simplified Chinese, and Arabic.
Pre-training data additionally included the following 13 languages: Russian, Polish, Turkish, Vietnamese, Dutch, Czech, Indonesian, Ukrainian, Romanian, Greek, Hindi, Hebrew, Persian.
**Context length**: Command-R supports a context length of 128K.
### Tool use capabilities:
Command-R has been specifically trained with conversational tool use capabilities. These have been trained into the model via a mixture of supervised fine-tuning and preference fine-tuning, using a specific prompt template. Deviating from this prompt template will likely reduce performance, but we encourage experimentation.
Command-R’s tool use functionality takes a conversation as input (with an optional user-system preamble), along with a list of available tools. The model will then generate a json-formatted list of actions to execute on a subset of those tools. Command-R may use one of its supplied tools more than once.
The model has been trained to recognise a special `directly_answer` tool, which it uses to indicate that it doesn’t want to use any of its other tools. The ability to abstain from calling a specific tool can be useful in a range of situations, such as greeting a user, or asking clarifying questions.
We recommend including the `directly_answer` tool, but it can be removed or renamed if required.
Comprehensive documentation for working with command-R's tool use prompt template can be found [here](https://docs.cohere.com/docs/prompting-command-r).
The code snippet below shows a minimal working example on how to render a prompt.
<details>
<summary><b>Usage: Rendering Tool Use Prompts [CLICK TO EXPAND]</b> </summary>
```python
from transformers import AutoTokenizer
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
# define conversation input:
conversation = [
{"role": "user", "content": "Whats the biggest penguin in the world?"}
]
# Define tools available for the model to use:
tools = [
{
"name": "internet_search",
"description": "Returns a list of relevant document snippets for a textual query retrieved from the internet",
"parameter_definitions": {
"query": {
"description": "Query to search the internet with",
"type": 'str',
"required": True
}
}
},
{
'name': "directly_answer",
"description": "Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history",
'parameter_definitions': {}
}
]
# render the tool use prompt as a string:
tool_use_prompt = tokenizer.apply_tool_use_template(
conversation,
tools=tools,
tokenize=False,
add_generation_prompt=True,
)
print(tool_use_prompt)
```
</details>
<details>
<summary><b>Example Rendered Tool Use Prompt [CLICK TO EXPAND]</b></summary>
````
<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
# System Preamble
## Basic Rules
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
# User Preamble
## Task and Context
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
## Style Guide
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.
## Available Tools
Here is a list of tools that you have available to you:
```python
def internet_search(query: str) -> List[Dict]:
"""Returns a list of relevant document snippets for a textual query retrieved from the internet
Args:
query (str): Query to search the internet with
"""
pass
```
```python
def directly_answer() -> List[Dict]:
"""Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history
"""
pass
```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
```json
[
{
"tool_name": title of the tool in the specification,
"parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
}
]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
````
</details>
<details>
<summary><b>Example Rendered Tool Use Completion [CLICK TO EXPAND]</b></summary>
````
Action: ```json
[
{
"tool_name": "internet_search",
"parameters": {
"query": "biggest penguin in the world"
}
}
]
```
````
</details>
### Grounded Generation and RAG Capabilities:
Command-R has been specifically trained with grounded generation capabilities. This means that it can generate responses based on a list of supplied document snippets, and it will include grounding spans (citations) in its response indicating the source of the information.
This can be used to enable behaviors such as grounded summarization and the final step of Retrieval Augmented Generation (RAG).This behavior has been trained into the model via a mixture of supervised fine-tuning and preference fine-tuning, using a specific prompt template.
Deviating from this prompt template may reduce performance, but we encourage experimentation.
Command-R’s grounded generation behavior takes a conversation as input (with an optional user-supplied system preamble, indicating task, context and desired output style), along with a list of retrieved document snippets.
The document snippets should be chunks, rather than long documents, typically around 100-400 words per chunk. Document snippets consist of key-value pairs. The keys should be short descriptive strings, the values can be text or semi-structured.
By default, Command-R will generate grounded responses by first predicting which documents are relevant, then predicting which ones it will cite, then generating an answer.
Finally, it will then insert grounding spans into the answer. See below for an example. This is referred to as `accurate` grounded generation.
The model is trained with a number of other answering modes, which can be selected by prompt changes . A `fast` citation mode is supported in the tokenizer, which will directly generate an answer with grounding spans in it, without first writing the answer out in full. This sacrifices some grounding accuracy in favor of generating fewer tokens.
Comprehensive documentation for working with command-R's grounded generation prompt template can be found [here](https://docs.cohere.com/docs/prompting-command-r).
The code snippet below shows a minimal working example on how to render a prompt.
<details>
<summary> <b>Usage: Rendering Grounded Generation prompts [CLICK TO EXPAND]</b> </summary>
````python
from transformers import AutoTokenizer
model_id = "CohereForAI/c4ai-command-r-v01"
tokenizer = AutoTokenizer.from_pretrained(model_id)
# define conversation input:
conversation = [
{"role": "user", "content": "Whats the biggest penguin in the world?"}
]
# define documents to ground on:
documents = [
{ "title": "Tall penguins", "text": "Emperor penguins are the tallest growing up to 122 cm in height." },
{ "title": "Penguin habitats", "text": "Emperor penguins only live in Antarctica."}
]
# render the tool use prompt as a string:
grounded_generation_prompt = tokenizer.apply_grounded_generation_template(
conversation,
documents=documents,
citation_mode="accurate", # or "fast"
tokenize=False,
add_generation_prompt=True,
)
print(grounded_generation_prompt)
````
</details>
<details>
<summary><b>Example Rendered Grounded Generation Prompt [CLICK TO EXPAND]</b></summary>
````<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
# System Preamble
## Basic Rules
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
# User Preamble
## Task and Context
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
## Style Guide
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
Document: 0
title: Tall penguins
text: Emperor penguins are the tallest growing up to 122 cm in height.
Document: 1
title: Penguin habitats
text: Emperor penguins only live in Antarctica.
</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
Thirdly, Write 'Answer:' followed by a response to the user's last input in high quality natural english. Use the retrieved documents to help you. Do not insert any citations or grounding markup.
Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
````
</details>
<details>
<summary><b>Example Rendered Grounded Generation Completion [CLICK TO EXPAND]</b></summary>
````
Relevant Documents: 0,1
Cited Documents: 0,1
Answer: The Emperor Penguin is the tallest or biggest penguin in the world. It is a bird that lives only in Antarctica and grows to a height of around 122 centimetres.
Grounded answer: The <co: 0>Emperor Penguin</co: 0> is the <co: 0>tallest</co: 0> or biggest penguin in the world. It is a bird that <co: 1>lives only in Antarctica</co: 1> and <co: 0>grows to a height of around 122 centimetres.</co: 0>
````
</details>
### Code Capabilities:
Command-R has been optimized to interact with your code, by requesting code snippets, code explanations, or code rewrites. It might not perform well out-of-the-box for pure code completion. For better performance, we also recommend using a low temperature (and even greedy decoding) for code-generation related instructions.
### Model Card Contact
For errors or additional questions about details in this model card, contact [[email protected]](mailto:[email protected]).
### Terms of Use:
We hope that the release of this model will make community-based research efforts more accessible, by releasing the weights of a highly performant 35 billion parameter model to researchers all over the world. This model is governed by a [CC-BY-NC](https://cohere.com/c4ai-cc-by-nc-license) License with an acceptable use addendum, and also requires adhering to [C4AI's Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).
### Try Chat:
You can try Command-R chat in the playground [here](https://dashboard.cohere.com/playground/chat).
|
{}
|
task
|
[
"QUESTION_ANSWERING",
"SUMMARIZATION"
] | 39,906 |
CATIE-AQ/NERmembert-large-4entities
|
CATIE-AQ
|
token-classification
|
[
"transformers",
"safetensors",
"camembert",
"token-classification",
"fr",
"dataset:CATIE-AQ/frenchNER_4entities",
"arxiv:1910.09700",
"base_model:almanach/camembert-large",
"base_model:finetune:almanach/camembert-large",
"doi:10.57967/hf/1752",
"license:mit",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2024-01-10T14:29:16Z |
2024-11-26T14:09:30+00:00
| 164 | 3 |
---
base_model: camembert/camembert-large
datasets:
- CATIE-AQ/frenchNER_4entities
language:
- fr
library_name: transformers
license: mit
metrics:
- precision
- recall
- f1
- accuracy
pipeline_tag: token-classification
widget:
- text: 'Le dévoilement du logo officiel des JO s''est déroulé le 21 octobre 2019
au Grand Rex. Ce nouvel emblème et cette nouvelle typographie ont été conçus par
le designer Sylvain Boyer avec les agences Royalties & Ecobranding. Rond, il rassemble
trois symboles : une médaille d''or, la flamme olympique et Marianne, symbolisée
par un visage de femme mais privée de son bonnet phrygien caractéristique. La
typographie dessinée fait référence à l''Art déco, mouvement artistique des années
1920, décennie pendant laquelle ont eu lieu pour la dernière fois les Jeux olympiques
à Paris en 1924. Pour la première fois, ce logo sera unique pour les Jeux olympiques
et les Jeux paralympiques.'
co2_eq_emissions: 80
new_version: CATIE-AQ/NERmemberta-4entities
model-index:
- name: NERmembert-large-4entities
results: []
---
# NERmembert-large-4entities
## Model Description
We present **NERmembert-large-4entities**, which is a [CamemBERT large](https://huggingface.co/camembert/camembert-large) fine-tuned for the Name Entity Recognition task for the French language on four French NER datasets for 4 entities (LOC, PER, ORG, MISC).
All these datasets were concatenated and cleaned into a single dataset that we called [frenchNER_4entities](https://huggingface.co/datasets/CATIE-AQ/frenchNER_4entities).
There are a total of **384,773** rows, of which **328,757** are for training, **24,131** for validation and **31,885** for testing.
Our methodology is described in a blog post available in [English](https://blog.vaniila.ai/en/NER_en/) or [French](https://blog.vaniila.ai/NER/).
## Dataset
The dataset used is [frenchNER_4entities](https://huggingface.co/datasets/CATIE-AQ/frenchNER_4entities), which represents ~385k sentences labeled in 4 categories:
| Label | Examples |
|:------|:-----------------------------------------------------------|
| PER | "La Bruyère", "Gaspard de Coligny", "Wittgenstein" |
| ORG | "UTBM", "American Airlines", "id Software" |
| LOC | "République du Cap-Vert", "Créteil", "Bordeaux" |
| MISC | "Wolfenstein 3D", "Révolution française", "Coupe du monde de football" |
The distribution of the entities is as follows:
<table>
<thead>
<tr>
<th><br>Splits</th>
<th><br>O</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<td><br>train</td>
<td><br>7,539,692</td>
<td><br>307,144</td>
<td><br>286,746</td>
<td><br>127,089</td>
<td><br>799,494</td>
</tr>
<tr>
<td><br>validation</td>
<td><br>544,580</td>
<td><br>24,034</td>
<td><br>21,585</td>
<td><br>5,927</td>
<td><br>18,221</td>
</tr>
<tr>
<td><br>test</td>
<td><br>720,623</td>
<td><br>32,870</td>
<td><br>29,683</td>
<td><br>7,911</td>
<td><br>21,760</td>
</tr>
</tbody>
</table>
## Evaluation results
The evaluation was carried out using the [**evaluate**](https://pypi.org/project/evaluate/) python package.
### frenchNER_4entities
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>0.971</td>
<td><br>0.947</td>
<td><br>0.902</td>
<td><br>0.663</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.974</td>
<td><br>0.948</td>
<td><br>0.892</td>
<td><br>0.658</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.978</td>
<td><br>0.957</td>
<td><br>0.904</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>0.980</td>
<td><br>0.963</td>
<td><br>0.919</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.978</td>
<td><br>0.958</td>
<td><br>0.903</td>
<td><br>0.814</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br><b>0.982</b></td>
<td><br><b>0.964</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.834</b></td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.952</td>
<td><br>0.924</td>
<td><br>0.870</td>
<td><br>0.845</td>
<td><br>0.986</td>
<td><br>0.976</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.990</td>
<td><br>0.972</td>
<td><br>0.938</td>
<td><br>0.546</td>
<td><br>0.992</td>
<td><br>0.976</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.971</td>
<td><br>0.947</td>
<td><br>0.902</td>
<td><br>0.663</td>
<td><br>0.989</td>
<td><br>0.976</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.962</td>
<td><br>0.933</td>
<td><br>0.857</td>
<td><br>0.830</td>
<td><br>0.985</td>
<td><br>0.976</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.987</td>
<td><br>0.963</td>
<td><br>0.930</td>
<td><br>0.545</td>
<td><br>0.993</td>
<td><br>0.976</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.974</td>
<td><br>0.948</td>
<td><br>0.892</td>
<td><br>0.658</td>
<td><br>0.989</td>
<td><br>0.976</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.973</td>
<td><br>0.955</td>
<td><br>0.886</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.983</td>
<td><br>0.960</td>
<td><br>0.923</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.978</td>
<td><br>0.957</td>
<td><br>0.904</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.978</td>
<td><br>0.960</td>
<td><br>0.899</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.985</td>
<td><br>0.966</td>
<td><br>0.940</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.980</td>
<td><br>0.963</td>
<td><br>0.919</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.973</td>
<td><br>0.951</td>
<td><br>0.888</td>
<td><br>0.850</td>
<td><br>0.993</td>
<td><br>0.984</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.983</td>
<td><br>0.964</td>
<td><br>0.918</td>
<td><br>0.781</td>
<td><br>0.993</td>
<td><br>0.984</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.978</td>
<td><br>0.958</td>
<td><br>0.903</td>
<td><br>0.814</td>
<td><br>0.993</td>
<td><br>0.984</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.977</td>
<td><br>0.961</td>
<td><br>0.896</td>
<td><br>0.872</td>
<td><br>0.993</td>
<td><br>0.986</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.987</td>
<td><br>0.966</td>
<td><br>0.943</td>
<td><br>0.798</td>
<td><br>0.995</td>
<td><br>0.986</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.982</b></td>
<td><br><b>0.964</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.834</b></td>
<td><br><b>0.994</b></td>
<td><br><b>0.986</b></td>
</tr>
</tbody>
</table>
</details>
In detail:
### multiconer
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>0.940</td>
<td><br>0.761</td>
<td><br>0.723</td>
<td><br>0.560</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.921</td>
<td><br>0.748</td>
<td><br>0.694</td>
<td><br>0.530</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.960</td>
<td><br>0.887</td>
<td><br>0.877</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>0.965</td>
<td><br>0.902</td>
<td><br>0.896</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.960</td>
<td><br>0.890</td>
<td><br>0.867</td>
<td><br>0.852</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br><b>0.969</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.904</b></td>
<td><br><b>0.864</b></td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.908</td>
<td><br>0.717</td>
<td><br>0.753</td>
<td><br>0.620</td>
<td><br>0.936</td>
<td><br>0.889</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.975</td>
<td><br>0.811</td>
<td><br>0.696</td>
<td><br>0.511</td>
<td><br>0.938</td>
<td><br>0.889</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.940</td>
<td><br>0.761</td>
<td><br>0.723</td>
<td><br>0.560</td>
<td><br>0.937</td>
<td><br>0.889</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.885</td>
<td><br>0.738</td>
<td><br>0.737</td>
<td><br>0.589</td>
<td><br>0.928</td>
<td><br>0.881</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.960</td>
<td><br>0.759</td>
<td><br>0.655</td>
<td><br>0.482</td>
<td><br>0.939</td>
<td><br>0.881</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.921</td>
<td><br>0.748</td>
<td><br>0.694</td>
<td><br>0.530</td>
<td><br>0.934</td>
<td><br>0.881</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.957</td>
<td><br>0.894</td>
<td><br>0.876</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.962</td>
<td><br>0.880</td>
<td><br>0.878</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.960</td>
<td><br>0.887</td>
<td><br>0.877</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.960</td>
<td><br>0.903</td>
<td><br>0.916</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.970</td>
<td><br>0.900</td>
<td><br>0.877</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.965</td>
<td><br>0.902</td>
<td><br>0.896</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.954</td>
<td><br>0.893</td>
<td><br>0.851</td>
<td><br>0.849</td>
<td><br>0.979</td>
<td><br>0.954</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.967</td>
<td><br>0.887</td>
<td><br>0.883</td>
<td><br>0.855</td>
<td><br>0.974</td>
<td><br>0.954</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.960</td>
<td><br>0.890</td>
<td><br>0.867</td>
<td><br>0.852</td>
<td><br>0.977</td>
<td><br>0.954</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.964</td>
<td><br>0.922</td>
<td><br>0.904</td>
<td><br>0.856</td>
<td><br>0.981</td>
<td><br>0.961</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.975</td>
<td><br>0.917</td>
<td><br>0.904</td>
<td><br>0.872</td>
<td><br>0.976</td>
<td><br>0.961</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.969</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.904</b></td>
<td><br><b>0.864</b></td>
<td><br><b>0.978</b></td>
<td><br><b>0.961</b></td>
</tr>
</tbody>
</table>
</details>
### multinerd
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>0.962</td>
<td><br>0.934</td>
<td><br>0.888</td>
<td><br>0.419</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.972</td>
<td><br>0.938</td>
<td><br>0.884</td>
<td><br>0.430</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br><b>0.987</b></td>
<td><br><b>0.979</b></td>
<td><br><b>0.953</b></td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0.770</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br><b>0.987</b></td>
<td><br>0.976</td>
<td><br>0.948</td>
<td><br><b>0.790</b></td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.931</td>
<td><br>0.893</td>
<td><br>0.827</td>
<td><br>0.725</td>
<td><br>0.979</td>
<td><br>0.966</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.994</td>
<td><br>0.980</td>
<td><br>0.959</td>
<td><br>0.295</td>
<td><br>0.990</td>
<td><br>0.966</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.962</td>
<td><br>0.934</td>
<td><br>0.888</td>
<td><br>0.419</td>
<td><br>0.984</td>
<td><br>0.966</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.954</td>
<td><br>0.908</td>
<td><br>0.817</td>
<td><br>0.705</td>
<td><br>0.977</td>
<td><br>0.967</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.991</td>
<td><br>0.969</td>
<td><br>0.963</td>
<td><br>0.310</td>
<td><br>0.990</td>
<td><br>0.967</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.972</td>
<td><br>0.938</td>
<td><br>0.884</td>
<td><br>0.430</td>
<td><br>0.984</td>
<td><br>0.967</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.974</td>
<td><br>0.965</td>
<td><br>0.910</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.995</td>
<td><br>0.981</td>
<td><br>0.968</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.979</td>
<td><br>0.970</td>
<td><br>0.927</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.996</td>
<td><br>0.987</td>
<td><br>0.980</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.987</b></td>
<td><br><b>0.979</b></td>
<td><br><b>0.953</b></td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.976</td>
<td><br>0.961</td>
<td><br>0.911</td>
<td><br>0.829</td>
<td><br>0.991</td>
<td><br>0.983</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.994</td>
<td><br>0.985</td>
<td><br>0.967</td>
<td><br>0.719</td>
<td><br>0.993</td>
<td><br>0.983</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0.770</td>
<td><br>0.992</td>
<td><br>0.983</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.979</td>
<td><br>0.967</td>
<td><br>0.922</td>
<td><br>0.852</td>
<td><br>0.991</td>
<td><br>0.985</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.996</td>
<td><br>0.986</td>
<td><br>0.974</td>
<td><br>0.736</td>
<td><br>0.994</td>
<td><br>0.985</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.987</b></td>
<td><br>0.976</td>
<td><br>0.948</td>
<td><br><b>0.790</b></td>
<td><br>0.993</td>
<td><br>0.985</td>
</tr>
</tbody>
</table>
</details>
### wikiner
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br><b>0.986</b></td>
<td><br><b>0.966</b></td>
<td><br><b>0.938</b></td>
<td><br><b>0.938</b></td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.983</td>
<td><br>0.964</td>
<td><br>0.925</td>
<td><br>0.926</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.878</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>0.972</td>
<td><br>0.953</td>
<td><br>0.893</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.876</td>
<td><br>0.872</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br>0.975</td>
<td><br>0.953</td>
<td><br>0.896</td>
<td><br>0.893</td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.986</td>
<td><br>0.962</td>
<td><br>0.925</td>
<td><br>0.943</td>
<td><br>0.998</td>
<td><br>0.992</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.987</td>
<td><br>0.969</td>
<td><br>0.951</td>
<td><br>0.933</td>
<td><br>0.997</td>
<td><br>0.992</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.986</b></td>
<td><br><b>0.966</b></td>
<td><br><b>0.938</b></td>
<td><br><b>0.938</b></td>
<td><br><b>0.998</b></td>
<td><br><b>0.992</b></td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.982</td>
<td><br>0.964</td>
<td><br>0.910</td>
<td><br>0.942</td>
<td><br>0.997</td>
<td><br>0.991</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.985</td>
<td><br>0.963</td>
<td><br>0.940</td>
<td><br>0.910</td>
<td><br>0.998</td>
<td><br>0.991</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.983</td>
<td><br>0.964</td>
<td><br>0.925</td>
<td><br>0.926</td>
<td><br>0.997</td>
<td><br>0.991</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.971</td>
<td><br>0.947</td>
<td><br>0.866</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.969</td>
<td><br>0.943</td>
<td><br>0.891</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.878</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.973</td>
<td><br>0.953</td>
<td><br>0.873</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.971</td>
<td><br>0.948</td>
<td><br>0.913</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.972</td>
<td><br>0.953</td>
<td><br>0.893</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.970</td>
<td><br>0.944</td>
<td><br>0.872</td>
<td><br>0.878</td>
<td><br>0.996</td>
<td><br>0.986</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.969</td>
<td><br>0.947</td>
<td><br>0.880</td>
<td><br>0.866</td>
<td><br>0.996</td>
<td><br>0.986</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.876</td>
<td><br>0.872</td>
<td><br>0.996</td>
<td><br>0.986</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.975</td>
<td><br>0.957</td>
<td><br>0.872</td>
<td><br>0.901</td>
<td><br>0.997</td>
<td><br>0.989</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.975</td>
<td><br>0.949</td>
<td><br>0.922</td>
<td><br>0.884</td>
<td><br>0.997</td>
<td><br>0.989</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.975</td>
<td><br>0.953</td>
<td><br>0.896</td>
<td><br>0.893</td>
<td><br>0.997</td>
<td><br>0.989</td>
</tr>
</tbody>
</table>
</details>
## Usage
### Code
```python
from transformers import pipeline
ner = pipeline('token-classification', model='CATIE-AQ/NERmembert-large-4entities', tokenizer='CATIE-AQ/NERmembert-large-4entities', aggregation_strategy="simple")
results = ner(
"Le dévoilement du logo officiel des JO s'est déroulé le 21 octobre 2019 au Grand Rex. Ce nouvel emblème et cette nouvelle typographie ont été conçus par le designer Sylvain Boyer avec les agences Royalties & Ecobranding. Rond, il rassemble trois symboles : une médaille d'or, la flamme olympique et Marianne, symbolisée par un visage de femme mais privée de son bonnet phrygien caractéristique. La typographie dessinée fait référence à l'Art déco, mouvement artistique des années 1920, décennie pendant laquelle ont eu lieu pour la dernière fois les Jeux olympiques à Paris en 1924. Pour la première fois, ce logo sera unique pour les Jeux olympiques et les Jeux paralympiques."
)
print(result)
```
```python
[{'entity_group': 'MISC', 'score': 0.9922348, 'word': 'JO', 'start': 35, 'end': 38},
{'entity_group': 'LOC', 'score': 0.9995632, 'word': 'Grand Rex', 'start': 74, 'end': 84},
{'entity_group': 'PER', 'score': 0.99612623, 'word': 'Sylvain Boyer', 'start': 164, 'end': 178},
{'entity_group': 'ORG', 'score': 0.87376696, 'word': 'Royalties & Ecobranding', 'start': 195, 'end': 219},
{'entity_group': 'PER', 'score': 0.8121169, 'word': 'Marianne', 'start': 298, 'end': 307},
{'entity_group': 'MISC', 'score': 0.7016645, 'word': 'Art déco', 'start': 438, 'end': 446},
{'entity_group': 'MISC', 'score': 0.99880475, 'word': 'Jeux olympiques', 'start': 549, 'end': 565},
{'entity_group': 'LOC', 'score': 0.98473144, 'word': 'Paris', 'start': 567, 'end': 573},
{'entity_group': 'MISC', 'score': 0.99421215, 'word': 'Jeux olympiques', 'start': 634, 'end': 650},
{'entity_group': 'MISC', 'score': 0.9936283, 'word': 'Jeux paralympiques', 'start': 657, 'end': 676}]
```
### Try it through Space
A Space has been created to test the model. It is available [here](https://huggingface.co/spaces/CATIE-AQ/NERmembert).
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|:-------------:|:-----:|:------:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.0347 | 1.0 | 41095 | 0.0537 | 0.9832 | 0.9832 | 0.9832 | 0.9832 |
| 0.0237 | 2.0 | 82190 | 0.0448 | 0.9858 | 0.9858 | 0.9858 | 0.9858 |
| 0.0119 | 3.0 | 123285 | 0.0532 | 0.9860 | 0.9860 | 0.9860 | 0.9860 |
### Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.16.1
- Tokenizers 0.15.0
## Environmental Impact
*Carbon emissions were estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). The hardware, runtime, cloud provider, and compute region were utilized to estimate the carbon impact.*
- **Hardware Type:** A100 PCIe 40/80GB
- **Hours used:** 4h17min
- **Cloud Provider:** Private Infrastructure
- **Carbon Efficiency (kg/kWh):** 0.078 (estimated from [electricitymaps](https://app.electricitymaps.com/zone/FR) for the day of January 10, 2024.)
- **Carbon Emitted** *(Power consumption x Time x Carbon produced based on location of power grid)*: 0.08 kg eq. CO2
## Citations
### NERmembert-large-4entities
```
@misc {NERmembert2024,
author = { {BOURDOIS, Loïck} },
organization = { {Centre Aquitain des Technologies de l'Information et Electroniques} },
title = { NERmembert-large-4entities (Revision 1cd8be5) },
year = 2024,
url = { https://huggingface.co/CATIE-AQ/NERmembert-large-4entities },
doi = { 10.57967/hf/1752 },
publisher = { Hugging Face }
}
```
### multiconer
```
@inproceedings{multiconer2-report,
title={{SemEval-2023 Task 2: Fine-grained Multilingual Named Entity Recognition (MultiCoNER 2)}},
author={Fetahu, Besnik and Kar, Sudipta and Chen, Zhiyu and Rokhlenko, Oleg and Malmasi, Shervin},
booktitle={Proceedings of the 17th International Workshop on Semantic Evaluation (SemEval-2023)},
year={2023},
publisher={Association for Computational Linguistics}}
@article{multiconer2-data,
title={{MultiCoNER v2: a Large Multilingual dataset for Fine-grained and Noisy Named Entity Recognition}},
author={Fetahu, Besnik and Chen, Zhiyu and Kar, Sudipta and Rokhlenko, Oleg and Malmasi, Shervin},
year={2023}}
```
### multinerd
```
@inproceedings{tedeschi-navigli-2022-multinerd,
title = "{M}ulti{NERD}: A Multilingual, Multi-Genre and Fine-Grained Dataset for Named Entity Recognition (and Disambiguation)",
author = "Tedeschi, Simone and Navigli, Roberto",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2022",
month = jul,
year = "2022",
address = "Seattle, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.findings-naacl.60",
doi = "10.18653/v1/2022.findings-naacl.60",
pages = "801--812"}
```
### pii-masking-200k
```
@misc {ai4privacy_2023,
author = { {ai4Privacy} },
title = { pii-masking-200k (Revision 1d4c0a1) },
year = 2023,
url = { https://huggingface.co/datasets/ai4privacy/pii-masking-200k },
doi = { 10.57967/hf/1532 },
publisher = { Hugging Face }}
```
### wikiann
```
@inproceedings{rahimi-etal-2019-massively,
title = "Massively Multilingual Transfer for {NER}",
author = "Rahimi, Afshin and Li, Yuan and Cohn, Trevor",
booktitle = "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
month = jul,
year = "2019",
address = "Florence, Italy",
publisher = "Association for Computational Linguistics",
url = "https://www.aclweb.org/anthology/P19-1015",
pages = "151--164"}
```
### wikiner
```
@article{NOTHMAN2013151,
title = {Learning multilingual named entity recognition from Wikipedia},
journal = {Artificial Intelligence},
volume = {194},
pages = {151-175},
year = {2013},
note = {Artificial Intelligence, Wikipedia and Semi-Structured Resources},
issn = {0004-3702},
doi = {https://doi.org/10.1016/j.artint.2012.03.006},
url = {https://www.sciencedirect.com/science/article/pii/S0004370212000276},
author = {Joel Nothman and Nicky Ringland and Will Radford and Tara Murphy and James R. Curran}}
```
### frenchNER_4entities
```
@misc {frenchNER2024,
author = { {BOURDOIS, Loïck} },
organization = { {Centre Aquitain des Technologies de l'Information et Electroniques} },
title = { frenchNER_4entities },
year = 2024,
url = { https://huggingface.co/CATIE-AQ/frenchNER_4entities },
doi = { 10.57967/hf/1751 },
publisher = { Hugging Face }
}
```
### CamemBERT
```
@inproceedings{martin2020camembert,
title={CamemBERT: a Tasty French Language Model},
author={Martin, Louis and Muller, Benjamin and Su{\'a}rez, Pedro Javier Ortiz and Dupont, Yoann and Romary, Laurent and de la Clergerie, {\'E}ric Villemonte and Seddah, Djam{\'e} and Sagot, Beno{\^\i}t},
booktitle={Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
year={2020}}
```
## License
MIT
| null |
Non_BioNLP
|
# NERmembert-large-4entities
## Model Description
We present **NERmembert-large-4entities**, which is a [CamemBERT large](https://huggingface.co/camembert/camembert-large) fine-tuned for the Name Entity Recognition task for the French language on four French NER datasets for 4 entities (LOC, PER, ORG, MISC).
All these datasets were concatenated and cleaned into a single dataset that we called [frenchNER_4entities](https://huggingface.co/datasets/CATIE-AQ/frenchNER_4entities).
There are a total of **384,773** rows, of which **328,757** are for training, **24,131** for validation and **31,885** for testing.
Our methodology is described in a blog post available in [English](https://blog.vaniila.ai/en/NER_en/) or [French](https://blog.vaniila.ai/NER/).
## Dataset
The dataset used is [frenchNER_4entities](https://huggingface.co/datasets/CATIE-AQ/frenchNER_4entities), which represents ~385k sentences labeled in 4 categories:
| Label | Examples |
|:------|:-----------------------------------------------------------|
| PER | "La Bruyère", "Gaspard de Coligny", "Wittgenstein" |
| ORG | "UTBM", "American Airlines", "id Software" |
| LOC | "République du Cap-Vert", "Créteil", "Bordeaux" |
| MISC | "Wolfenstein 3D", "Révolution française", "Coupe du monde de football" |
The distribution of the entities is as follows:
<table>
<thead>
<tr>
<th><br>Splits</th>
<th><br>O</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<td><br>train</td>
<td><br>7,539,692</td>
<td><br>307,144</td>
<td><br>286,746</td>
<td><br>127,089</td>
<td><br>799,494</td>
</tr>
<tr>
<td><br>validation</td>
<td><br>544,580</td>
<td><br>24,034</td>
<td><br>21,585</td>
<td><br>5,927</td>
<td><br>18,221</td>
</tr>
<tr>
<td><br>test</td>
<td><br>720,623</td>
<td><br>32,870</td>
<td><br>29,683</td>
<td><br>7,911</td>
<td><br>21,760</td>
</tr>
</tbody>
</table>
## Evaluation results
The evaluation was carried out using the [**evaluate**](https://pypi.org/project/evaluate/) python package.
### frenchNER_4entities
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>0.971</td>
<td><br>0.947</td>
<td><br>0.902</td>
<td><br>0.663</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.974</td>
<td><br>0.948</td>
<td><br>0.892</td>
<td><br>0.658</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.978</td>
<td><br>0.957</td>
<td><br>0.904</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>0.980</td>
<td><br>0.963</td>
<td><br>0.919</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.978</td>
<td><br>0.958</td>
<td><br>0.903</td>
<td><br>0.814</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br><b>0.982</b></td>
<td><br><b>0.964</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.834</b></td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.952</td>
<td><br>0.924</td>
<td><br>0.870</td>
<td><br>0.845</td>
<td><br>0.986</td>
<td><br>0.976</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.990</td>
<td><br>0.972</td>
<td><br>0.938</td>
<td><br>0.546</td>
<td><br>0.992</td>
<td><br>0.976</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.971</td>
<td><br>0.947</td>
<td><br>0.902</td>
<td><br>0.663</td>
<td><br>0.989</td>
<td><br>0.976</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.962</td>
<td><br>0.933</td>
<td><br>0.857</td>
<td><br>0.830</td>
<td><br>0.985</td>
<td><br>0.976</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.987</td>
<td><br>0.963</td>
<td><br>0.930</td>
<td><br>0.545</td>
<td><br>0.993</td>
<td><br>0.976</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.974</td>
<td><br>0.948</td>
<td><br>0.892</td>
<td><br>0.658</td>
<td><br>0.989</td>
<td><br>0.976</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.973</td>
<td><br>0.955</td>
<td><br>0.886</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.983</td>
<td><br>0.960</td>
<td><br>0.923</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.978</td>
<td><br>0.957</td>
<td><br>0.904</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.978</td>
<td><br>0.960</td>
<td><br>0.899</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.985</td>
<td><br>0.966</td>
<td><br>0.940</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.980</td>
<td><br>0.963</td>
<td><br>0.919</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.973</td>
<td><br>0.951</td>
<td><br>0.888</td>
<td><br>0.850</td>
<td><br>0.993</td>
<td><br>0.984</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.983</td>
<td><br>0.964</td>
<td><br>0.918</td>
<td><br>0.781</td>
<td><br>0.993</td>
<td><br>0.984</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.978</td>
<td><br>0.958</td>
<td><br>0.903</td>
<td><br>0.814</td>
<td><br>0.993</td>
<td><br>0.984</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.977</td>
<td><br>0.961</td>
<td><br>0.896</td>
<td><br>0.872</td>
<td><br>0.993</td>
<td><br>0.986</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.987</td>
<td><br>0.966</td>
<td><br>0.943</td>
<td><br>0.798</td>
<td><br>0.995</td>
<td><br>0.986</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.982</b></td>
<td><br><b>0.964</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.834</b></td>
<td><br><b>0.994</b></td>
<td><br><b>0.986</b></td>
</tr>
</tbody>
</table>
</details>
In detail:
### multiconer
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>0.940</td>
<td><br>0.761</td>
<td><br>0.723</td>
<td><br>0.560</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.921</td>
<td><br>0.748</td>
<td><br>0.694</td>
<td><br>0.530</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.960</td>
<td><br>0.887</td>
<td><br>0.877</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>0.965</td>
<td><br>0.902</td>
<td><br>0.896</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.960</td>
<td><br>0.890</td>
<td><br>0.867</td>
<td><br>0.852</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br><b>0.969</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.904</b></td>
<td><br><b>0.864</b></td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.908</td>
<td><br>0.717</td>
<td><br>0.753</td>
<td><br>0.620</td>
<td><br>0.936</td>
<td><br>0.889</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.975</td>
<td><br>0.811</td>
<td><br>0.696</td>
<td><br>0.511</td>
<td><br>0.938</td>
<td><br>0.889</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.940</td>
<td><br>0.761</td>
<td><br>0.723</td>
<td><br>0.560</td>
<td><br>0.937</td>
<td><br>0.889</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.885</td>
<td><br>0.738</td>
<td><br>0.737</td>
<td><br>0.589</td>
<td><br>0.928</td>
<td><br>0.881</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.960</td>
<td><br>0.759</td>
<td><br>0.655</td>
<td><br>0.482</td>
<td><br>0.939</td>
<td><br>0.881</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.921</td>
<td><br>0.748</td>
<td><br>0.694</td>
<td><br>0.530</td>
<td><br>0.934</td>
<td><br>0.881</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.957</td>
<td><br>0.894</td>
<td><br>0.876</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.962</td>
<td><br>0.880</td>
<td><br>0.878</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.960</td>
<td><br>0.887</td>
<td><br>0.877</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.960</td>
<td><br>0.903</td>
<td><br>0.916</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.970</td>
<td><br>0.900</td>
<td><br>0.877</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.965</td>
<td><br>0.902</td>
<td><br>0.896</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.954</td>
<td><br>0.893</td>
<td><br>0.851</td>
<td><br>0.849</td>
<td><br>0.979</td>
<td><br>0.954</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.967</td>
<td><br>0.887</td>
<td><br>0.883</td>
<td><br>0.855</td>
<td><br>0.974</td>
<td><br>0.954</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.960</td>
<td><br>0.890</td>
<td><br>0.867</td>
<td><br>0.852</td>
<td><br>0.977</td>
<td><br>0.954</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.964</td>
<td><br>0.922</td>
<td><br>0.904</td>
<td><br>0.856</td>
<td><br>0.981</td>
<td><br>0.961</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.975</td>
<td><br>0.917</td>
<td><br>0.904</td>
<td><br>0.872</td>
<td><br>0.976</td>
<td><br>0.961</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.969</b></td>
<td><br><b>0.919</b></td>
<td><br><b>0.904</b></td>
<td><br><b>0.864</b></td>
<td><br><b>0.978</b></td>
<td><br><b>0.961</b></td>
</tr>
</tbody>
</table>
</details>
### multinerd
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>0.962</td>
<td><br>0.934</td>
<td><br>0.888</td>
<td><br>0.419</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.972</td>
<td><br>0.938</td>
<td><br>0.884</td>
<td><br>0.430</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br><b>0.987</b></td>
<td><br><b>0.979</b></td>
<td><br><b>0.953</b></td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0.770</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br><b>0.987</b></td>
<td><br>0.976</td>
<td><br>0.948</td>
<td><br><b>0.790</b></td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.931</td>
<td><br>0.893</td>
<td><br>0.827</td>
<td><br>0.725</td>
<td><br>0.979</td>
<td><br>0.966</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.994</td>
<td><br>0.980</td>
<td><br>0.959</td>
<td><br>0.295</td>
<td><br>0.990</td>
<td><br>0.966</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.962</td>
<td><br>0.934</td>
<td><br>0.888</td>
<td><br>0.419</td>
<td><br>0.984</td>
<td><br>0.966</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.954</td>
<td><br>0.908</td>
<td><br>0.817</td>
<td><br>0.705</td>
<td><br>0.977</td>
<td><br>0.967</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.991</td>
<td><br>0.969</td>
<td><br>0.963</td>
<td><br>0.310</td>
<td><br>0.990</td>
<td><br>0.967</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.972</td>
<td><br>0.938</td>
<td><br>0.884</td>
<td><br>0.430</td>
<td><br>0.984</td>
<td><br>0.967</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.974</td>
<td><br>0.965</td>
<td><br>0.910</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.995</td>
<td><br>0.981</td>
<td><br>0.968</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.979</td>
<td><br>0.970</td>
<td><br>0.927</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.996</td>
<td><br>0.987</td>
<td><br>0.980</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.987</b></td>
<td><br><b>0.979</b></td>
<td><br><b>0.953</b></td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.976</td>
<td><br>0.961</td>
<td><br>0.911</td>
<td><br>0.829</td>
<td><br>0.991</td>
<td><br>0.983</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.994</td>
<td><br>0.985</td>
<td><br>0.967</td>
<td><br>0.719</td>
<td><br>0.993</td>
<td><br>0.983</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.985</td>
<td><br>0.973</td>
<td><br>0.938</td>
<td><br>0.770</td>
<td><br>0.992</td>
<td><br>0.983</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.979</td>
<td><br>0.967</td>
<td><br>0.922</td>
<td><br>0.852</td>
<td><br>0.991</td>
<td><br>0.985</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.996</td>
<td><br>0.986</td>
<td><br>0.974</td>
<td><br>0.736</td>
<td><br>0.994</td>
<td><br>0.985</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.987</b></td>
<td><br>0.976</td>
<td><br>0.948</td>
<td><br><b>0.790</b></td>
<td><br>0.993</td>
<td><br>0.985</td>
</tr>
</tbody>
</table>
</details>
### wikiner
For space reasons, we show only the F1 of the different models. You can see the full results below the table.
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="1"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br><b>0.986</b></td>
<td><br><b>0.966</b></td>
<td><br><b>0.938</b></td>
<td><br><b>0.938</b></td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>0.983</td>
<td><br>0.964</td>
<td><br>0.925</td>
<td><br>0.926</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.878</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>0.972</td>
<td><br>0.953</td>
<td><br>0.893</td>
<td><br>0</td>
</tr>
<tr>
<td rowspan="1"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.876</td>
<td><br>0.872</td>
</tr>
<tr>
<td rowspan="1"><br>NERmembert-large-4entities (this model)</td>
<td><br>0.975</td>
<td><br>0.953</td>
<td><br>0.896</td>
<td><br>0.893</td>
</tr>
</tbody>
</table>
<details>
<summary>Full results</summary>
<table>
<thead>
<tr>
<th><br>Model</th>
<th><br>Metrics</th>
<th><br>PER</th>
<th><br>LOC</th>
<th><br>ORG</th>
<th><br>MISC</th>
<th><br>O</th>
<th><br>Overall</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3"><br><a href="https://hf.co/Jean-Baptiste/camembert-ner">Jean-Baptiste/camembert-ner</a></td>
<td><br>Precision</td>
<td><br>0.986</td>
<td><br>0.962</td>
<td><br>0.925</td>
<td><br>0.943</td>
<td><br>0.998</td>
<td><br>0.992</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.987</td>
<td><br>0.969</td>
<td><br>0.951</td>
<td><br>0.933</td>
<td><br>0.997</td>
<td><br>0.992</td>
</tr>
<tr>
<td>F1</td>
<td><br><b>0.986</b></td>
<td><br><b>0.966</b></td>
<td><br><b>0.938</b></td>
<td><br><b>0.938</b></td>
<td><br><b>0.998</b></td>
<td><br><b>0.992</b></td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/cmarkea/distilcamembert-base-ner">cmarkea/distilcamembert-base-ner</a></td>
<td><br>Precision</td>
<td><br>0.982</td>
<td><br>0.964</td>
<td><br>0.910</td>
<td><br>0.942</td>
<td><br>0.997</td>
<td><br>0.991</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.985</td>
<td><br>0.963</td>
<td><br>0.940</td>
<td><br>0.910</td>
<td><br>0.998</td>
<td><br>0.991</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.983</td>
<td><br>0.964</td>
<td><br>0.925</td>
<td><br>0.926</td>
<td><br>0.997</td>
<td><br>0.991</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-3entities">NERmembert-base-3entities</a></td>
<td><br>Precision</td>
<td><br>0.971</td>
<td><br>0.947</td>
<td><br>0.866</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.969</td>
<td><br>0.943</td>
<td><br>0.891</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.878</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-large-3entities">NERmembert-large-3entities</a></td>
<td><br>Precision</td>
<td><br>0.973</td>
<td><br>0.953</td>
<td><br>0.873</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.971</td>
<td><br>0.948</td>
<td><br>0.913</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.972</td>
<td><br>0.953</td>
<td><br>0.893</td>
<td><br>0</td>
<td><br>X</td>
<td><br>X</td>
</tr>
<tr>
<td rowspan="3"><br><a href="https://hf.co/CATIE-AQ/NERmembert-base-4entities">NERmembert-base-4entities</a></td>
<td><br>Precision</td>
<td><br>0.970</td>
<td><br>0.944</td>
<td><br>0.872</td>
<td><br>0.878</td>
<td><br>0.996</td>
<td><br>0.986</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.969</td>
<td><br>0.947</td>
<td><br>0.880</td>
<td><br>0.866</td>
<td><br>0.996</td>
<td><br>0.986</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.970</td>
<td><br>0.945</td>
<td><br>0.876</td>
<td><br>0.872</td>
<td><br>0.996</td>
<td><br>0.986</td>
</tr>
<tr>
<td rowspan="3"><br>NERmembert-large-4entities (this model)</td>
<td><br>Precision</td>
<td><br>0.975</td>
<td><br>0.957</td>
<td><br>0.872</td>
<td><br>0.901</td>
<td><br>0.997</td>
<td><br>0.989</td>
</tr>
<tr>
<td><br>Recall</td>
<td><br>0.975</td>
<td><br>0.949</td>
<td><br>0.922</td>
<td><br>0.884</td>
<td><br>0.997</td>
<td><br>0.989</td>
</tr>
<tr>
<td>F1</td>
<td><br>0.975</td>
<td><br>0.953</td>
<td><br>0.896</td>
<td><br>0.893</td>
<td><br>0.997</td>
<td><br>0.989</td>
</tr>
</tbody>
</table>
</details>
## Usage
### Code
```python
from transformers import pipeline
ner = pipeline('token-classification', model='CATIE-AQ/NERmembert-large-4entities', tokenizer='CATIE-AQ/NERmembert-large-4entities', aggregation_strategy="simple")
results = ner(
"Le dévoilement du logo officiel des JO s'est déroulé le 21 octobre 2019 au Grand Rex. Ce nouvel emblème et cette nouvelle typographie ont été conçus par le designer Sylvain Boyer avec les agences Royalties & Ecobranding. Rond, il rassemble trois symboles : une médaille d'or, la flamme olympique et Marianne, symbolisée par un visage de femme mais privée de son bonnet phrygien caractéristique. La typographie dessinée fait référence à l'Art déco, mouvement artistique des années 1920, décennie pendant laquelle ont eu lieu pour la dernière fois les Jeux olympiques à Paris en 1924. Pour la première fois, ce logo sera unique pour les Jeux olympiques et les Jeux paralympiques."
)
print(result)
```
```python
[{'entity_group': 'MISC', 'score': 0.9922348, 'word': 'JO', 'start': 35, 'end': 38},
{'entity_group': 'LOC', 'score': 0.9995632, 'word': 'Grand Rex', 'start': 74, 'end': 84},
{'entity_group': 'PER', 'score': 0.99612623, 'word': 'Sylvain Boyer', 'start': 164, 'end': 178},
{'entity_group': 'ORG', 'score': 0.87376696, 'word': 'Royalties & Ecobranding', 'start': 195, 'end': 219},
{'entity_group': 'PER', 'score': 0.8121169, 'word': 'Marianne', 'start': 298, 'end': 307},
{'entity_group': 'MISC', 'score': 0.7016645, 'word': 'Art déco', 'start': 438, 'end': 446},
{'entity_group': 'MISC', 'score': 0.99880475, 'word': 'Jeux olympiques', 'start': 549, 'end': 565},
{'entity_group': 'LOC', 'score': 0.98473144, 'word': 'Paris', 'start': 567, 'end': 573},
{'entity_group': 'MISC', 'score': 0.99421215, 'word': 'Jeux olympiques', 'start': 634, 'end': 650},
{'entity_group': 'MISC', 'score': 0.9936283, 'word': 'Jeux paralympiques', 'start': 657, 'end': 676}]
```
### Try it through Space
A Space has been created to test the model. It is available [here](https://huggingface.co/spaces/CATIE-AQ/NERmembert).
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|:-------------:|:-----:|:------:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.0347 | 1.0 | 41095 | 0.0537 | 0.9832 | 0.9832 | 0.9832 | 0.9832 |
| 0.0237 | 2.0 | 82190 | 0.0448 | 0.9858 | 0.9858 | 0.9858 | 0.9858 |
| 0.0119 | 3.0 | 123285 | 0.0532 | 0.9860 | 0.9860 | 0.9860 | 0.9860 |
### Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2
- Datasets 2.16.1
- Tokenizers 0.15.0
## Environmental Impact
*Carbon emissions were estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). The hardware, runtime, cloud provider, and compute region were utilized to estimate the carbon impact.*
- **Hardware Type:** A100 PCIe 40/80GB
- **Hours used:** 4h17min
- **Cloud Provider:** Private Infrastructure
- **Carbon Efficiency (kg/kWh):** 0.078 (estimated from [electricitymaps](https://app.electricitymaps.com/zone/FR) for the day of January 10, 2024.)
- **Carbon Emitted** *(Power consumption x Time x Carbon produced based on location of power grid)*: 0.08 kg eq. CO2
## Citations
### NERmembert-large-4entities
```
@misc {NERmembert2024,
author = { {BOURDOIS, Loïck} },
organization = { {Centre Aquitain des Technologies de l'Information et Electroniques} },
title = { NERmembert-large-4entities (Revision 1cd8be5) },
year = 2024,
url = { https://huggingface.co/CATIE-AQ/NERmembert-large-4entities },
doi = { 10.57967/hf/1752 },
publisher = { Hugging Face }
}
```
### multiconer
```
@inproceedings{multiconer2-report,
title={{SemEval-2023 Task 2: Fine-grained Multilingual Named Entity Recognition (MultiCoNER 2)}},
author={Fetahu, Besnik and Kar, Sudipta and Chen, Zhiyu and Rokhlenko, Oleg and Malmasi, Shervin},
booktitle={Proceedings of the 17th International Workshop on Semantic Evaluation (SemEval-2023)},
year={2023},
publisher={Association for Computational Linguistics}}
@article{multiconer2-data,
title={{MultiCoNER v2: a Large Multilingual dataset for Fine-grained and Noisy Named Entity Recognition}},
author={Fetahu, Besnik and Chen, Zhiyu and Kar, Sudipta and Rokhlenko, Oleg and Malmasi, Shervin},
year={2023}}
```
### multinerd
```
@inproceedings{tedeschi-navigli-2022-multinerd,
title = "{M}ulti{NERD}: A Multilingual, Multi-Genre and Fine-Grained Dataset for Named Entity Recognition (and Disambiguation)",
author = "Tedeschi, Simone and Navigli, Roberto",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2022",
month = jul,
year = "2022",
address = "Seattle, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.findings-naacl.60",
doi = "10.18653/v1/2022.findings-naacl.60",
pages = "801--812"}
```
### pii-masking-200k
```
@misc {ai4privacy_2023,
author = { {ai4Privacy} },
title = { pii-masking-200k (Revision 1d4c0a1) },
year = 2023,
url = { https://huggingface.co/datasets/ai4privacy/pii-masking-200k },
doi = { 10.57967/hf/1532 },
publisher = { Hugging Face }}
```
### wikiann
```
@inproceedings{rahimi-etal-2019-massively,
title = "Massively Multilingual Transfer for {NER}",
author = "Rahimi, Afshin and Li, Yuan and Cohn, Trevor",
booktitle = "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
month = jul,
year = "2019",
address = "Florence, Italy",
publisher = "Association for Computational Linguistics",
url = "https://www.aclweb.org/anthology/P19-1015",
pages = "151--164"}
```
### wikiner
```
@article{NOTHMAN2013151,
title = {Learning multilingual named entity recognition from Wikipedia},
journal = {Artificial Intelligence},
volume = {194},
pages = {151-175},
year = {2013},
note = {Artificial Intelligence, Wikipedia and Semi-Structured Resources},
issn = {0004-3702},
doi = {https://doi.org/10.1016/j.artint.2012.03.006},
url = {https://www.sciencedirect.com/science/article/pii/S0004370212000276},
author = {Joel Nothman and Nicky Ringland and Will Radford and Tara Murphy and James R. Curran}}
```
### frenchNER_4entities
```
@misc {frenchNER2024,
author = { {BOURDOIS, Loïck} },
organization = { {Centre Aquitain des Technologies de l'Information et Electroniques} },
title = { frenchNER_4entities },
year = 2024,
url = { https://huggingface.co/CATIE-AQ/frenchNER_4entities },
doi = { 10.57967/hf/1751 },
publisher = { Hugging Face }
}
```
### CamemBERT
```
@inproceedings{martin2020camembert,
title={CamemBERT: a Tasty French Language Model},
author={Martin, Louis and Muller, Benjamin and Su{\'a}rez, Pedro Javier Ortiz and Dupont, Yoann and Romary, Laurent and de la Clergerie, {\'E}ric Villemonte and Seddah, Djam{\'e} and Sagot, Beno{\^\i}t},
booktitle={Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
year={2020}}
```
## License
MIT
|
{"base_model": "camembert/camembert-large", "datasets": ["CATIE-AQ/frenchNER_4entities"], "language": ["fr"], "library_name": "transformers", "license": "mit", "metrics": ["precision", "recall", "f1", "accuracy"], "pipeline_tag": "token-classification", "widget": [{"text": "Le dévoilement du logo officiel des JO s'est déroulé le 21 octobre 2019 au Grand Rex. Ce nouvel emblème et cette nouvelle typographie ont été conçus par le designer Sylvain Boyer avec les agences Royalties & Ecobranding. Rond, il rassemble trois symboles : une médaille d'or, la flamme olympique et Marianne, symbolisée par un visage de femme mais privée de son bonnet phrygien caractéristique. La typographie dessinée fait référence à l'Art déco, mouvement artistique des années 1920, décennie pendant laquelle ont eu lieu pour la dernière fois les Jeux olympiques à Paris en 1924. Pour la première fois, ce logo sera unique pour les Jeux olympiques et les Jeux paralympiques."}], "co2_eq_emissions": 80, "new_version": "CATIE-AQ/NERmemberta-4entities", "model-index": [{"name": "NERmembert-large-4entities", "results": []}]}
|
task
|
[
"NAMED_ENTITY_RECOGNITION"
] | 39,907 |
ReDiX/bge-base-it-v1.5
|
ReDiX
|
sentence-similarity
|
[
"sentence-transformers",
"safetensors",
"bert",
"sentence-similarity",
"feature-extraction",
"dataset_size:10K<n<100K",
"loss:MatryoshkaLoss",
"loss:MultipleNegativesRankingLoss",
"it",
"dataset:ReDiX/wikipediaQA-ita",
"arxiv:1908.10084",
"arxiv:2205.13147",
"arxiv:1705.00652",
"base_model:BAAI/bge-base-en-v1.5",
"base_model:finetune:BAAI/bge-base-en-v1.5",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] | 2024-06-06T20:28:25Z |
2024-06-07T07:22:24+00:00
| 11 | 0 |
---
base_model: BAAI/bge-base-en-v1.5
datasets:
- ReDiX/wikipediaQA-ita
language:
- it
library_name: sentence-transformers
license: apache-2.0
metrics:
- cosine_accuracy@1
- cosine_accuracy@3
- cosine_accuracy@5
- cosine_accuracy@10
- cosine_precision@1
- cosine_precision@3
- cosine_precision@5
- cosine_precision@10
- cosine_recall@1
- cosine_recall@3
- cosine_recall@5
- cosine_recall@10
- cosine_ndcg@10
- cosine_mrr@10
- cosine_map@100
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- dataset_size:10K<n<100K
- loss:MatryoshkaLoss
- loss:MultipleNegativesRankingLoss
widget:
- source_sentence: Informazioni sul Duomo di Milano
sentences:
- 'Il Duomo di Milano, ufficialmente Cattedrale Metropolitana della Natività della
Beata Vergine Maria (Dòmm de Milan in milanese, IPA: [ˈdɔm de miˈlãː]), è la cattedrale
dell''arcidiocesi di Milano e monumento nazionale italiano[2]. Simbolo del capoluogo
lombardo, e situato nell''omonima piazza al centro della metropoli, è dedicata
a santa Maria Nascente. È la chiesa più grande d''Italia[3][4] (la più grande
della Repubblica Italiana, giacché la basilica di San Pietro, più grande, è nel
territorio della Città del Vaticano; la seconda più grande considerando, invece,
tutta la penisola italiana), la terza nel mondo per superficie, la sesta per volume[5].
È sede della parrocchia di Santa Tecla nel duomo di Milano.[6].'
- 'Principale centro economico e finanziario della Repubblica Italiana, Milano ne
guidò lo sviluppo industriale, costituendo con Torino e Genova il cosiddetto Triangolo
industriale, in particolar modo durante gli anni del boom economico, quando la
crescita industriale e urbanistica coinvolse anche le città limitrofe, creando
la vasta area metropolitana milanese. In ambito culturale, Milano è il principale
centro italiano dell''editoria ed è ai vertici del circuito musicale mondiale
grazie alla stagione lirica del Teatro alla Scala e alla sua lunga tradizione
operistica. È, inoltre, tra i principali poli fieristici europei (con due esposizioni
universali ospitate: Expo 1906 e Expo 2015) e del disegno industriale, ed è considerata
una delle capitali mondiali della moda.'
- 'Il Duomo di Milano è il principale luogo di culto cattolico della città di Milano
e cattedrale dell''omonima arcidiocesi. Altri edifici di valore sono le basiliche
paleocristiane di Milano: la basilica di Sant''Ambrogio, considerata da sempre
il massimo esempio dell''architettura romanica lombarda,[100] nonché uno dei più
antichi monumenti dell''arte cristiana, la basilica di San Lorenzo, edificio a
pianta centrale bizantino conosciuto anche per la vicinanza delle omonime colonne,
la basilica di San Nazaro in Brolo e la basilica di San Simpliciano, legate come
le precedenti alla figura di sant''Ambrogio.'
- source_sentence: Quando è avvenuto lo sbarco in Normandia?
sentences:
- Lo sbarco in Normandia (nome in codice operazione Neptune, parte marittima della
più ampia operazione Overlord) fu una delle più grandi invasioni anfibie della
storia, messa in atto dalle forze alleate durante la seconda guerra mondiale per
aprire un secondo fronte in Europa, dirigersi verso la Germania nazista e allo
stesso tempo alleggerire il fronte orientale, sul quale da tre anni l'Armata Rossa
stava sostenendo un aspro confronto contro i tedeschi.
- La Normandia è una regione storica ed amministrativa della Francia. Si affaccia
a nord sul canale della Manica e confina a sud-ovest con la Bretagna, a sud con
i Paesi della Loira e con il Centro-Valle della Loira, a est con l'Île-de-France
e a nord-est con l'Alta Francia. Occupa la bassa vallata della Senna (Alta Normandia)
e si estende verso ovest fino alla penisola del Cotentin (Bassa Normandia).
- L'invasione iniziò nelle prime ore di martedì 6 giugno 1944 (data conosciuta come
D-Day), quando toccarono terra nella penisola del Cotentin e nella zona di Caen
le truppe alleate aviotrasportate, che aprirono la strada alle forze terrestri.
All'alba del 6 giugno, precedute da un imponente bombardamento aeronavale, le
fanterie sbarcarono su cinque spiagge.
- source_sentence: Quando è nato Luca?
sentences:
- Luca è nato nel 1990
- Luca ha un fratello di nome Lorenzo e una sorella di nome Chiara
- Luca lavora come sviluppatore
model-index:
- name: BGE base WikipediaQA-ita
results:
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: dim 768
type: dim_768
metrics:
- type: cosine_accuracy@1
value: 0.7534636553425698
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.8732207249952553
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.8883089770354906
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.9052002277472007
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.7534636553425698
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.2910735749984184
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.17766179540709814
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.09052002277472007
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.7534636553425698
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.8732207249952553
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.8883089770354906
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.9052002277472007
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.8384852729647283
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.816234142799904
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.8179246209727115
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: dim 512
type: dim_512
metrics:
- type: cosine_accuracy@1
value: 0.7517555513380147
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.8708483583222623
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.8864110836970962
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.9031125450749667
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.7517555513380147
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.29028278610742075
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.17728221673941924
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.09031125450749669
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.7517555513380147
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.8708483583222623
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.8864110836970962
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.9031125450749667
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.8364916692827067
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.8142769402946844
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.8159454559810521
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: dim 256
type: dim_256
metrics:
- type: cosine_accuracy@1
value: 0.7410324539760865
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.8637312583032833
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.878724615676599
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.8953311823875498
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.7410324539760865
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.2879104194344278
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.1757449231353198
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.089533118238755
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.7410324539760865
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.8637312583032833
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.878724615676599
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.8953311823875498
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.8275013364480452
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.8048458790108056
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.8065838557898369
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: dim 128
type: dim_128
metrics:
- type: cosine_accuracy@1
value: 0.7194913645853103
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.844942114253179
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.8621180489656481
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.8800531410134751
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.7194913645853103
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.28164737141772633
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.17242360979312965
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.08800531410134751
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.7194913645853103
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.844942114253179
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.8621180489656481
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.8800531410134751
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.8089255548002888
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.7852223020542415
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.7872945920851742
name: Cosine Map@100
- task:
type: information-retrieval
name: Information Retrieval
dataset:
name: dim 64
type: dim_64
metrics:
- type: cosine_accuracy@1
value: 0.6789742360017088
name: Cosine Accuracy@1
- type: cosine_accuracy@3
value: 0.811169140387136
name: Cosine Accuracy@3
- type: cosine_accuracy@5
value: 0.8296035881283715
name: Cosine Accuracy@5
- type: cosine_accuracy@10
value: 0.8477945467750865
name: Cosine Accuracy@10
- type: cosine_precision@1
value: 0.6789742360017088
name: Cosine Precision@1
- type: cosine_precision@3
value: 0.27038971346237867
name: Cosine Precision@3
- type: cosine_precision@5
value: 0.16592071762567434
name: Cosine Precision@5
- type: cosine_precision@10
value: 0.08477945467750865
name: Cosine Precision@10
- type: cosine_recall@1
value: 0.6789742360017088
name: Cosine Recall@1
- type: cosine_recall@3
value: 0.811169140387136
name: Cosine Recall@3
- type: cosine_recall@5
value: 0.8296035881283715
name: Cosine Recall@5
- type: cosine_recall@10
value: 0.8477945467750865
name: Cosine Recall@10
- type: cosine_ndcg@10
value: 0.774416217728066
name: Cosine Ndcg@10
- type: cosine_mrr@10
value: 0.7482063280040872
name: Cosine Mrr@10
- type: cosine_map@100
value: 0.7505568607057078
name: Cosine Map@100
---
# BGE base WikipediaQA-ita
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) <!-- at revision a5beb1e3e68b9ab74eb54cfd186867f64f240e1a -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **Similarity Function:** Cosine Similarity
<!-- - **Training Dataset:** Unknown -->
- **Language:** it
- **License:** apache-2.0
### Evaluation
| Dimension | Baseline | Fine-tuned | Improvement |
|:---------:|:--------:|:----------:|:-----------:|
| 768 | 0.7245 | 0.8385 | 15.75% |
| 512 | 0.7028 | 0.8365 | 19.02% |
| 256 | 0.6484 | 0.8275 | 27.60% |
| 128 | 0.5337 | 0.8089 | 51.57% |
| 64 | 0.3630 | 0.7697 | 112.00% |
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': True}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
(2): Normalize()
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
"Cos'è il volume?",
'Cosa si intende per singolarità in fisica?',
'Cosa fece durante il suo insegnamento?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Information Retrieval
* Dataset: `dim_768`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7534636553 | 0.6136838110 |
| cosine_accuracy@3 | 0.8732207250 | 0.7589675460 |
| cosine_accuracy@5 | 0.8883089770 | 0.7898083128 |
| cosine_accuracy@10 | 0.9052002277 | 0.8223571835 |
| cosine_precision@1 | 0.7534636553 | 0.6136838110 |
| cosine_precision@3 | 0.2910735750 | 0.2529891820 |
| cosine_precision@5 | 0.1776617954 | 0.1579616626 |
| cosine_precision@10 | 0.0905200228 | 0.0822357184 |
| cosine_recall@1 | 0.7534636553 | 0.6136838110 |
| cosine_recall@3 | 0.8732207250 | 0.7589675460 |
| cosine_recall@5 | 0.8883089770 | 0.7898083128 |
| cosine_recall@10 | 0.9052002277 | 0.8223571835 |
| cosine_ndcg@10 | 0.8384852730 | 0.7244706923 |
| cosine_mrr@10 | 0.8162341428 | 0.6924611082 |
| **cosine_map@100** | **0.8179246210** | **0.6959149352** |
#### Information Retrieval
* Dataset: `dim_512`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7517555513 | 0.5930916682 |
| cosine_accuracy@3 | 0.8708483583 | 0.7353387740 |
| cosine_accuracy@5 | 0.8864110837 | 0.7657050674 |
| cosine_accuracy@10 | 0.9031125451 | 0.8025241981 |
| cosine_precision@1 | 0.7517555513 | 0.5930916682 |
| cosine_precision@3 | 0.2902827861 | 0.2451129247 |
| cosine_precision@5 | 0.1772822167 | 0.1531410135 |
| cosine_precision@10 | 0.0903112545 | 0.0802524198 |
| cosine_recall@1 | 0.7517555513 | 0.5930916682 |
| cosine_recall@3 | 0.8708483583 | 0.7353387740 |
| cosine_recall@5 | 0.8864110837 | 0.7657050674 |
| cosine_recall@10 | 0.9031125451 | 0.8025241981 |
| cosine_ndcg@10 | 0.8364916693 | 0.7027860074 |
| cosine_mrr@10 | 0.8142769403 | 0.6703873661 |
| **cosine_map@100** | **0.8159454560** | **0.6741256048** |
#### Information Retrieval
* Dataset: `dim_256`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7410324540 | 0.5345416588 |
| cosine_accuracy@3 | 0.8637312583 | 0.6777377111 |
| cosine_accuracy@5 | 0.8787246157 | 0.7135130006 |
| cosine_accuracy@10 | 0.8953311824 | 0.7551717593 |
| cosine_precision@1 | 0.7410324540 | 0.5345416588 |
| cosine_precision@3 | 0.2879104194 | 0.2259125704 |
| cosine_precision@5 | 0.1757449231 | 0.1427026001 |
| cosine_precision@10 | 0.0895331182 | 0.0755171759 |
| cosine_recall@1 | 0.7410324540 | 0.5345416588 |
| cosine_recall@3 | 0.8637312583 | 0.6777377111 |
| cosine_recall@5 | 0.8787246157 | 0.7135130006 |
| cosine_recall@10 | 0.8953311824 | 0.7551717593 |
| cosine_ndcg@10 | 0.8275013364 | 0.6484387526 |
| cosine_mrr@10 | 0.8048458790 | 0.6139468424 |
| **cosine_map@100** | **0.8065838558** | **0.6183180421** |
#### Information Retrieval
* Dataset: `dim_128`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7194913646 | 0.4141203264 |
| cosine_accuracy@3 | 0.8449421143 | 0.5562725375 |
| cosine_accuracy@5 | 0.8621180490 | 0.5971721389 |
| cosine_accuracy@10 | 0.8800531410 | 0.6548680964 |
| cosine_precision@1 | 0.7194913646 | 0.4141203264 |
| cosine_precision@3 | 0.2816473714 | 0.1854241792 |
| cosine_precision@5 | 0.1724236098 | 0.1194344278 |
| cosine_precision@10 | 0.0880053141 | 0.0654868096 |
| cosine_recall@1 | 0.7194913646 | 0.4141203264 |
| cosine_recall@3 | 0.8449421143 | 0.5562725375 |
| cosine_recall@5 | 0.8621180490 | 0.5971721389 |
| cosine_recall@10 | 0.8800531410 | 0.6548680964 |
| cosine_ndcg@10 | 0.8089255548 | 0.5337452948 |
| cosine_mrr@10 | 0.7852223021 | 0.4951471545 |
| **cosine_map@100** | **0.7872570794** | **0.5011613720** |
#### Information Retrieval
* Dataset: `dim_64`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.6734674511 | 0.2564053900 |
| cosine_accuracy@3 | 0.8043271968 | 0.3720819890 |
| cosine_accuracy@5 | 0.8261529702 | 0.4202884798 |
| cosine_accuracy@10 | 0.8500664263 | 0.4790282786 |
| cosine_precision@1 | 0.6734674511 | 0.2564053900 |
| cosine_precision@3 | 0.2681090656 | 0.1240273297 |
| cosine_precision@5 | 0.1652305940 | 0.0840576960 |
| cosine_precision@10 | 0.0850066426 | 0.0479028279 |
| cosine_recall@1 | 0.6734674511 | 0.2564053900 |
| cosine_recall@3 | 0.8043271968 | 0.3720819890 |
| cosine_recall@5 | 0.8261529702 | 0.4202884798 |
| cosine_recall@10 | 0.8500664263 | 0.4790282786 |
| cosine_ndcg@10 | 0.7672278878 | 0.3629672889 |
| cosine_mrr@10 | 0.7425634870 | 0.3263833534 |
| **cosine_map@100** | **0.7445402342** | **0.3343989330** |
#### Information Retrieval
* Dataset: `sequential_score`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| **sequential_score** | **0.6734674511** | **0.2564053900** |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 94,833 training samples
* Columns: <code>positive</code> and <code>anchor</code>
* Approximate statistics based on the first 1000 samples:
| | positive | anchor |
|:--------|:-------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 60 tokens</li><li>mean: 505.04 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 19.99 tokens</li><li>max: 72 tokens</li></ul> |
* Samples:
| positive | anchor |
|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
| <code>Cleopatra (nome)<br>. Varianti Ipocoristici: Clea, Cleo Maschili: Cleopatro Varianti in altre lingue Catalano: Cleòpatra, Cleopatra Ceco: Kleopatra Croato: Kleopatra Francese: Cléopâtre Ipocoristici: Cléa, Cléo Greco antico: Κλεοπάτρα (Kleopátra) Maschili: Κλεόπατρος (Kleópatros) Ipocoristici maschili: Κλωπᾶς (Klopâs), Κλεόπας (Kleópas) Inglese: Cleopatra Ipocoristici: Cleo Latino: Cleopatra Olandese: Cleopatra Polacco: Kleopatra Portoghese: Cleópatra Russo: Клеопатра (Kleopatra) Spagnolo: Cleopatra Tedesco: Kleopatra Ucraino: Клеопатра (Kleopatra) Origine e diffusione Deriva, tramite il latino Cleopatra, dal nome greco Κλεοπατρα (Kleopatra); si tratta di un composto di κλέος (kléos, "fama", "gloria") e πατρός (patrós), genitivo di πατήρ (patḗr, "padre"), quindi il significato complessivo può essere interpretato come "di padri gloriosi", "che ha gloria per la sua stirpe", "gloria del padre". Entrambi gli elementi sono comuni nell'onomastica greca: il primo è presente in Pericle, Empedocle, Temistocle, Tecla ed Ercole, il secondo in Antipatro, ed entrambi, disposti nell'ordine inverso, formano il nome Patroclo. Va però notato che alcune fonti riconducono il primo elemento a κλείς (kleís, "chiave"), quindi "chiave per la patria (terra dei padri)". Il nome era comune nell'antica Grecia ed è presente anche nella mitologia con alcune figure secondarie, fra cui Cleopatra, moglie di Fineo re dei Traci, ripudiata e imprigionata dal marito e poi liberata dagli Argonauti, e Cleopatra Alcione, figlia di Idas e moglie di Meleagro. Era inoltre tradizionale in varie dinastie reali macedoni ed elleniche, specie in quella tolemaica, nella quale le donne con questo nome sono numerosissime: tra queste va citata in particolare Cleopatra, l'ultima regina a governare l'Egitto, amante di Giulio Cesare e poi di Marco Antonio; la sua figura è presente in numerose opere artistiche, letterarie, musicali e cinematografiche, che l'hanno resa talmente famosa che l'associazione tra nome e regina è quasi automatica. Proprio a questa notorietà è dovuto l'utilizzo del nome in Italia, che è scarso ma non nullo: è attestato per un</code> | <code>Chi è stata l'ultima regina a governare l'Egitto con il nome Cleopatra?</code> |
| <code>Caravino<br>Caravino (Caravin in piemontese) è un comune italiano di abitanti della città metropolitana di Torino in Piemonte. Geografia fisica Il paese è adagiato tra le colline del Canavese eporediese orientale, confinando a ovest con Strambino ed Ivrea, a est con Cossano Canavese e Settimo Rottaro, a sud con Vestignè e Borgomasino, a nord con Azeglio e Albiano d'Ivrea. Non è attraversato da nessun fiume; solo il piccolo Naviglio di Ivrea scorre 2 km verso ovest, mentre la Dora Baltea a circa 6 km più a ovest. Storia Sono stati ritrovati resti romani del III secolo. Il suo toponimo potrebbe derivare dal termine cavrinum, ovvero "caprino", indicando la locale pastorizia ovina; oppure da quadratum, essendo al centro geografico degli allora poteri tra le marche di Ivrea, Torino e la diocesi di Vercelli. Ultima ipotesi, da carra - vinum, un'antica unità di misura torinese per immagazzinare delle grandi quantità vino, corrispondente a 493,069 litri. Il toponimo comparve ufficialmente dall'XI secolo, come feudo eporediese dei conti Masino-Valperga, che si definirono sia discendenti diretti di Arduino, che fu a capo della tutta la Marca d'Ivrea, sia dei Conti di Pombia, già padroni di Borgomasino. Inoltre, fu proprio a partire dall'XI secolo che i benedettini portarono qui culto di San Giacomo, divenuto poi il patrono del paese, di cui si attesta anche la chiesetta in località Carpaneto, poi affrescata dal pittore Giacomino da Ivrea (del 1465), dipendente dalla Cella di Vestignè e dall'Abbazia di Fruttuaria di San Benigno Canavese. Il prestigio del borgo al di sotto del Castello seguì le stesse sorti di quest'ultimo, almeno fino alla prima metà del XIV secolo, allorquando i Valperga non riuscirono più a contrastare l'espansione del monferrino Facino Cane, che saccheggiò barbaramente il paese e altri comuni limitrofi nell'estate del 1397. A quest'ultimo succedettero i francesi del XV secolo</code> | <code>Il toponimo di Caravino potrebbe derivare da quale termine latino?</code> |
| <code>Peć<br>Peć ( o Peja; ; ) è una città del Kosovo occidentale. Nel 2011 la città aveva una popolazione di abitanti, mentre l'intera municipalità aveva una popolazione di abitanti. In epoca medioevale, nel 1346, fu sede del Patriarcato della Chiesa ortodossa serba. Secoli dopo, nel 1899 vi fu fondata dalla popolazione locale albanese la Lega di Peja. Il monastero patriarcale di Peć è patrimonio dell'umanità dell'UNESCO, incluso nei Monumenti medievali in Kosovo. Geografia fisica Peć sorge nell'estremità nord-occidentale della piana di Metochia, ad est del gruppo montuoso delle Prokletije e della gola di Rugova. La città è attraversata da ovest verso est dal torrente Bistrica, affluente di destra del Drin Bianco. È situata ad 85 km ad ovest di Pristina e a 250 km a nord di Tirana. Etimologia In serbo peć significa letteralmente forno. Nei documenti della Repubblica di Ragusa la città viene talvolta menzionata come Forno. Durante la dominazione ottomana la città era conosciuta come İpek (ايپك). Altre denominazioni storiche sono in latino Pescium e in greco Episkion (Επισκιον). Storia Tra il 1180 ed il 1190 il principe Stefano Nemanja liberò l'area di Peć dalla presenza bizantina, dando il via così al dominio serbo sulla regione. Suo figlio Stefano Nemanjić cedette la città e il contado al monastero di Žiča nel 1220. Peć passò così sotto il diretto controllo degli arcivescovi serbi, che costruirono le loro residenze e le loro chiese. Nel 1346 il re serbo Stefan Dusan fece di Peć il centro della chiesa ortodossa serba. Peć fu conquistata dagli Ottomani nel 1455 e ribattezzata İpek. Nel 1513 divenne capoluogo del sangiaccato omonimo a sua volta suddiviso in quattro kaza. Nel corso dei quattro secoli di dominazione ottomana affluirono nella città numerosi immigrati turchi che le conferirono un aspetto tipicamente levantino. Nel 1899, su iniziativa di Haxhi</code> | <code>In quale regione del Kosovo si trova la città di Peć?</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "MultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64
],
"matryoshka_weights": [
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: epoch
- `per_device_train_batch_size`: 32
- `per_device_eval_batch_size`: 16
- `gradient_accumulation_steps`: 16
- `learning_rate`: 2e-05
- `num_train_epochs`: 4
- `lr_scheduler_type`: cosine
- `warmup_ratio`: 0.1
- `bf16`: True
- `tf32`: True
- `load_best_model_at_end`: True
- `optim`: adamw_torch_fused
- `batch_sampler`: no_duplicates
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: epoch
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 32
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 16
- `eval_accumulation_steps`: None
- `learning_rate`: 2e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 4
- `max_steps`: -1
- `lr_scheduler_type`: cosine
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: True
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: True
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: True
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch_fused
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: False
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
</details>
### Framework Versions
- Python: 3.10.12
- Sentence Transformers: 3.0.0
- Transformers: 4.41.2
- PyTorch: 2.3.0+cu121
- Accelerate: 0.30.1
- Datasets: 2.19.2
- Tokenizers: 0.19.1
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MatryoshkaLoss
```bibtex
@misc{kusupati2024matryoshka,
title={Matryoshka Representation Learning},
author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
year={2024},
eprint={2205.13147},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
title={Efficient Natural Language Response Suggestion for Smart Reply},
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},
year={2017},
eprint={1705.00652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# BGE base WikipediaQA-ita
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
## Model Details
### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5) <!-- at revision a5beb1e3e68b9ab74eb54cfd186867f64f240e1a -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **Similarity Function:** Cosine Similarity
<!-- - **Training Dataset:** Unknown -->
- **Language:** it
- **License:** apache-2.0
### Evaluation
| Dimension | Baseline | Fine-tuned | Improvement |
|:---------:|:--------:|:----------:|:-----------:|
| 768 | 0.7245 | 0.8385 | 15.75% |
| 512 | 0.7028 | 0.8365 | 19.02% |
| 256 | 0.6484 | 0.8275 | 27.60% |
| 128 | 0.5337 | 0.8089 | 51.57% |
| 64 | 0.3630 | 0.7697 | 112.00% |
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
### Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': True}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
(2): Normalize()
)
```
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
"Cos'è il volume?",
'Cosa si intende per singolarità in fisica?',
'Cosa fece durante il suo insegnamento?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Information Retrieval
* Dataset: `dim_768`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7534636553 | 0.6136838110 |
| cosine_accuracy@3 | 0.8732207250 | 0.7589675460 |
| cosine_accuracy@5 | 0.8883089770 | 0.7898083128 |
| cosine_accuracy@10 | 0.9052002277 | 0.8223571835 |
| cosine_precision@1 | 0.7534636553 | 0.6136838110 |
| cosine_precision@3 | 0.2910735750 | 0.2529891820 |
| cosine_precision@5 | 0.1776617954 | 0.1579616626 |
| cosine_precision@10 | 0.0905200228 | 0.0822357184 |
| cosine_recall@1 | 0.7534636553 | 0.6136838110 |
| cosine_recall@3 | 0.8732207250 | 0.7589675460 |
| cosine_recall@5 | 0.8883089770 | 0.7898083128 |
| cosine_recall@10 | 0.9052002277 | 0.8223571835 |
| cosine_ndcg@10 | 0.8384852730 | 0.7244706923 |
| cosine_mrr@10 | 0.8162341428 | 0.6924611082 |
| **cosine_map@100** | **0.8179246210** | **0.6959149352** |
#### Information Retrieval
* Dataset: `dim_512`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7517555513 | 0.5930916682 |
| cosine_accuracy@3 | 0.8708483583 | 0.7353387740 |
| cosine_accuracy@5 | 0.8864110837 | 0.7657050674 |
| cosine_accuracy@10 | 0.9031125451 | 0.8025241981 |
| cosine_precision@1 | 0.7517555513 | 0.5930916682 |
| cosine_precision@3 | 0.2902827861 | 0.2451129247 |
| cosine_precision@5 | 0.1772822167 | 0.1531410135 |
| cosine_precision@10 | 0.0903112545 | 0.0802524198 |
| cosine_recall@1 | 0.7517555513 | 0.5930916682 |
| cosine_recall@3 | 0.8708483583 | 0.7353387740 |
| cosine_recall@5 | 0.8864110837 | 0.7657050674 |
| cosine_recall@10 | 0.9031125451 | 0.8025241981 |
| cosine_ndcg@10 | 0.8364916693 | 0.7027860074 |
| cosine_mrr@10 | 0.8142769403 | 0.6703873661 |
| **cosine_map@100** | **0.8159454560** | **0.6741256048** |
#### Information Retrieval
* Dataset: `dim_256`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7410324540 | 0.5345416588 |
| cosine_accuracy@3 | 0.8637312583 | 0.6777377111 |
| cosine_accuracy@5 | 0.8787246157 | 0.7135130006 |
| cosine_accuracy@10 | 0.8953311824 | 0.7551717593 |
| cosine_precision@1 | 0.7410324540 | 0.5345416588 |
| cosine_precision@3 | 0.2879104194 | 0.2259125704 |
| cosine_precision@5 | 0.1757449231 | 0.1427026001 |
| cosine_precision@10 | 0.0895331182 | 0.0755171759 |
| cosine_recall@1 | 0.7410324540 | 0.5345416588 |
| cosine_recall@3 | 0.8637312583 | 0.6777377111 |
| cosine_recall@5 | 0.8787246157 | 0.7135130006 |
| cosine_recall@10 | 0.8953311824 | 0.7551717593 |
| cosine_ndcg@10 | 0.8275013364 | 0.6484387526 |
| cosine_mrr@10 | 0.8048458790 | 0.6139468424 |
| **cosine_map@100** | **0.8065838558** | **0.6183180421** |
#### Information Retrieval
* Dataset: `dim_128`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.7194913646 | 0.4141203264 |
| cosine_accuracy@3 | 0.8449421143 | 0.5562725375 |
| cosine_accuracy@5 | 0.8621180490 | 0.5971721389 |
| cosine_accuracy@10 | 0.8800531410 | 0.6548680964 |
| cosine_precision@1 | 0.7194913646 | 0.4141203264 |
| cosine_precision@3 | 0.2816473714 | 0.1854241792 |
| cosine_precision@5 | 0.1724236098 | 0.1194344278 |
| cosine_precision@10 | 0.0880053141 | 0.0654868096 |
| cosine_recall@1 | 0.7194913646 | 0.4141203264 |
| cosine_recall@3 | 0.8449421143 | 0.5562725375 |
| cosine_recall@5 | 0.8621180490 | 0.5971721389 |
| cosine_recall@10 | 0.8800531410 | 0.6548680964 |
| cosine_ndcg@10 | 0.8089255548 | 0.5337452948 |
| cosine_mrr@10 | 0.7852223021 | 0.4951471545 |
| **cosine_map@100** | **0.7872570794** | **0.5011613720** |
#### Information Retrieval
* Dataset: `dim_64`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| cosine_accuracy@1 | 0.6734674511 | 0.2564053900 |
| cosine_accuracy@3 | 0.8043271968 | 0.3720819890 |
| cosine_accuracy@5 | 0.8261529702 | 0.4202884798 |
| cosine_accuracy@10 | 0.8500664263 | 0.4790282786 |
| cosine_precision@1 | 0.6734674511 | 0.2564053900 |
| cosine_precision@3 | 0.2681090656 | 0.1240273297 |
| cosine_precision@5 | 0.1652305940 | 0.0840576960 |
| cosine_precision@10 | 0.0850066426 | 0.0479028279 |
| cosine_recall@1 | 0.6734674511 | 0.2564053900 |
| cosine_recall@3 | 0.8043271968 | 0.3720819890 |
| cosine_recall@5 | 0.8261529702 | 0.4202884798 |
| cosine_recall@10 | 0.8500664263 | 0.4790282786 |
| cosine_ndcg@10 | 0.7672278878 | 0.3629672889 |
| cosine_mrr@10 | 0.7425634870 | 0.3263833534 |
| **cosine_map@100** | **0.7445402342** | **0.3343989330** |
#### Information Retrieval
* Dataset: `sequential_score`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)
| Metric | This Model | Base Model |
|:--------------------|:-------------|:-------------|
| **sequential_score** | **0.6734674511** | **0.2564053900** |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 94,833 training samples
* Columns: <code>positive</code> and <code>anchor</code>
* Approximate statistics based on the first 1000 samples:
| | positive | anchor |
|:--------|:-------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
| type | string | string |
| details | <ul><li>min: 60 tokens</li><li>mean: 505.04 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 19.99 tokens</li><li>max: 72 tokens</li></ul> |
* Samples:
| positive | anchor |
|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
| <code>Cleopatra (nome)<br>. Varianti Ipocoristici: Clea, Cleo Maschili: Cleopatro Varianti in altre lingue Catalano: Cleòpatra, Cleopatra Ceco: Kleopatra Croato: Kleopatra Francese: Cléopâtre Ipocoristici: Cléa, Cléo Greco antico: Κλεοπάτρα (Kleopátra) Maschili: Κλεόπατρος (Kleópatros) Ipocoristici maschili: Κλωπᾶς (Klopâs), Κλεόπας (Kleópas) Inglese: Cleopatra Ipocoristici: Cleo Latino: Cleopatra Olandese: Cleopatra Polacco: Kleopatra Portoghese: Cleópatra Russo: Клеопатра (Kleopatra) Spagnolo: Cleopatra Tedesco: Kleopatra Ucraino: Клеопатра (Kleopatra) Origine e diffusione Deriva, tramite il latino Cleopatra, dal nome greco Κλεοπατρα (Kleopatra); si tratta di un composto di κλέος (kléos, "fama", "gloria") e πατρός (patrós), genitivo di πατήρ (patḗr, "padre"), quindi il significato complessivo può essere interpretato come "di padri gloriosi", "che ha gloria per la sua stirpe", "gloria del padre". Entrambi gli elementi sono comuni nell'onomastica greca: il primo è presente in Pericle, Empedocle, Temistocle, Tecla ed Ercole, il secondo in Antipatro, ed entrambi, disposti nell'ordine inverso, formano il nome Patroclo. Va però notato che alcune fonti riconducono il primo elemento a κλείς (kleís, "chiave"), quindi "chiave per la patria (terra dei padri)". Il nome era comune nell'antica Grecia ed è presente anche nella mitologia con alcune figure secondarie, fra cui Cleopatra, moglie di Fineo re dei Traci, ripudiata e imprigionata dal marito e poi liberata dagli Argonauti, e Cleopatra Alcione, figlia di Idas e moglie di Meleagro. Era inoltre tradizionale in varie dinastie reali macedoni ed elleniche, specie in quella tolemaica, nella quale le donne con questo nome sono numerosissime: tra queste va citata in particolare Cleopatra, l'ultima regina a governare l'Egitto, amante di Giulio Cesare e poi di Marco Antonio; la sua figura è presente in numerose opere artistiche, letterarie, musicali e cinematografiche, che l'hanno resa talmente famosa che l'associazione tra nome e regina è quasi automatica. Proprio a questa notorietà è dovuto l'utilizzo del nome in Italia, che è scarso ma non nullo: è attestato per un</code> | <code>Chi è stata l'ultima regina a governare l'Egitto con il nome Cleopatra?</code> |
| <code>Caravino<br>Caravino (Caravin in piemontese) è un comune italiano di abitanti della città metropolitana di Torino in Piemonte. Geografia fisica Il paese è adagiato tra le colline del Canavese eporediese orientale, confinando a ovest con Strambino ed Ivrea, a est con Cossano Canavese e Settimo Rottaro, a sud con Vestignè e Borgomasino, a nord con Azeglio e Albiano d'Ivrea. Non è attraversato da nessun fiume; solo il piccolo Naviglio di Ivrea scorre 2 km verso ovest, mentre la Dora Baltea a circa 6 km più a ovest. Storia Sono stati ritrovati resti romani del III secolo. Il suo toponimo potrebbe derivare dal termine cavrinum, ovvero "caprino", indicando la locale pastorizia ovina; oppure da quadratum, essendo al centro geografico degli allora poteri tra le marche di Ivrea, Torino e la diocesi di Vercelli. Ultima ipotesi, da carra - vinum, un'antica unità di misura torinese per immagazzinare delle grandi quantità vino, corrispondente a 493,069 litri. Il toponimo comparve ufficialmente dall'XI secolo, come feudo eporediese dei conti Masino-Valperga, che si definirono sia discendenti diretti di Arduino, che fu a capo della tutta la Marca d'Ivrea, sia dei Conti di Pombia, già padroni di Borgomasino. Inoltre, fu proprio a partire dall'XI secolo che i benedettini portarono qui culto di San Giacomo, divenuto poi il patrono del paese, di cui si attesta anche la chiesetta in località Carpaneto, poi affrescata dal pittore Giacomino da Ivrea (del 1465), dipendente dalla Cella di Vestignè e dall'Abbazia di Fruttuaria di San Benigno Canavese. Il prestigio del borgo al di sotto del Castello seguì le stesse sorti di quest'ultimo, almeno fino alla prima metà del XIV secolo, allorquando i Valperga non riuscirono più a contrastare l'espansione del monferrino Facino Cane, che saccheggiò barbaramente il paese e altri comuni limitrofi nell'estate del 1397. A quest'ultimo succedettero i francesi del XV secolo</code> | <code>Il toponimo di Caravino potrebbe derivare da quale termine latino?</code> |
| <code>Peć<br>Peć ( o Peja; ; ) è una città del Kosovo occidentale. Nel 2011 la città aveva una popolazione di abitanti, mentre l'intera municipalità aveva una popolazione di abitanti. In epoca medioevale, nel 1346, fu sede del Patriarcato della Chiesa ortodossa serba. Secoli dopo, nel 1899 vi fu fondata dalla popolazione locale albanese la Lega di Peja. Il monastero patriarcale di Peć è patrimonio dell'umanità dell'UNESCO, incluso nei Monumenti medievali in Kosovo. Geografia fisica Peć sorge nell'estremità nord-occidentale della piana di Metochia, ad est del gruppo montuoso delle Prokletije e della gola di Rugova. La città è attraversata da ovest verso est dal torrente Bistrica, affluente di destra del Drin Bianco. È situata ad 85 km ad ovest di Pristina e a 250 km a nord di Tirana. Etimologia In serbo peć significa letteralmente forno. Nei documenti della Repubblica di Ragusa la città viene talvolta menzionata come Forno. Durante la dominazione ottomana la città era conosciuta come İpek (ايپك). Altre denominazioni storiche sono in latino Pescium e in greco Episkion (Επισκιον). Storia Tra il 1180 ed il 1190 il principe Stefano Nemanja liberò l'area di Peć dalla presenza bizantina, dando il via così al dominio serbo sulla regione. Suo figlio Stefano Nemanjić cedette la città e il contado al monastero di Žiča nel 1220. Peć passò così sotto il diretto controllo degli arcivescovi serbi, che costruirono le loro residenze e le loro chiese. Nel 1346 il re serbo Stefan Dusan fece di Peć il centro della chiesa ortodossa serba. Peć fu conquistata dagli Ottomani nel 1455 e ribattezzata İpek. Nel 1513 divenne capoluogo del sangiaccato omonimo a sua volta suddiviso in quattro kaza. Nel corso dei quattro secoli di dominazione ottomana affluirono nella città numerosi immigrati turchi che le conferirono un aspetto tipicamente levantino. Nel 1899, su iniziativa di Haxhi</code> | <code>In quale regione del Kosovo si trova la città di Peć?</code> |
* Loss: [<code>MatryoshkaLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters:
```json
{
"loss": "MultipleNegativesRankingLoss",
"matryoshka_dims": [
768,
512,
256,
128,
64
],
"matryoshka_weights": [
1,
1,
1,
1,
1
],
"n_dims_per_step": -1
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: epoch
- `per_device_train_batch_size`: 32
- `per_device_eval_batch_size`: 16
- `gradient_accumulation_steps`: 16
- `learning_rate`: 2e-05
- `num_train_epochs`: 4
- `lr_scheduler_type`: cosine
- `warmup_ratio`: 0.1
- `bf16`: True
- `tf32`: True
- `load_best_model_at_end`: True
- `optim`: adamw_torch_fused
- `batch_sampler`: no_duplicates
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: epoch
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 32
- `per_device_eval_batch_size`: 16
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 16
- `eval_accumulation_steps`: None
- `learning_rate`: 2e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 4
- `max_steps`: -1
- `lr_scheduler_type`: cosine
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: True
- `fp16`: False
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: True
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: True
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch_fused
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: False
- `hub_always_push`: False
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`:
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `dispatch_batches`: None
- `split_batches`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
</details>
### Framework Versions
- Python: 3.10.12
- Sentence Transformers: 3.0.0
- Transformers: 4.41.2
- PyTorch: 2.3.0+cu121
- Accelerate: 0.30.1
- Datasets: 2.19.2
- Tokenizers: 0.19.1
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
#### MatryoshkaLoss
```bibtex
@misc{kusupati2024matryoshka,
title={Matryoshka Representation Learning},
author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
year={2024},
eprint={2205.13147},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```
#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
title={Efficient Natural Language Response Suggestion for Smart Reply},
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},
year={2017},
eprint={1705.00652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "BAAI/bge-base-en-v1.5", "datasets": ["ReDiX/wikipediaQA-ita"], "language": ["it"], "library_name": "sentence-transformers", "license": "apache-2.0", "metrics": ["cosine_accuracy@1", "cosine_accuracy@3", "cosine_accuracy@5", "cosine_accuracy@10", "cosine_precision@1", "cosine_precision@3", "cosine_precision@5", "cosine_precision@10", "cosine_recall@1", "cosine_recall@3", "cosine_recall@5", "cosine_recall@10", "cosine_ndcg@10", "cosine_mrr@10", "cosine_map@100"], "pipeline_tag": "sentence-similarity", "tags": ["sentence-transformers", "sentence-similarity", "feature-extraction", "dataset_size:10K<n<100K", "loss:MatryoshkaLoss", "loss:MultipleNegativesRankingLoss"], "widget": [{"source_sentence": "Informazioni sul Duomo di Milano", "sentences": ["Il Duomo di Milano, ufficialmente Cattedrale Metropolitana della Natività della Beata Vergine Maria (Dòmm de Milan in milanese, IPA: [ˈdɔm de miˈlãː]), è la cattedrale dell'arcidiocesi di Milano e monumento nazionale italiano[2]. Simbolo del capoluogo lombardo, e situato nell'omonima piazza al centro della metropoli, è dedicata a santa Maria Nascente. È la chiesa più grande d'Italia[3][4] (la più grande della Repubblica Italiana, giacché la basilica di San Pietro, più grande, è nel territorio della Città del Vaticano; la seconda più grande considerando, invece, tutta la penisola italiana), la terza nel mondo per superficie, la sesta per volume[5]. È sede della parrocchia di Santa Tecla nel duomo di Milano.[6].", "Principale centro economico e finanziario della Repubblica Italiana, Milano ne guidò lo sviluppo industriale, costituendo con Torino e Genova il cosiddetto Triangolo industriale, in particolar modo durante gli anni del boom economico, quando la crescita industriale e urbanistica coinvolse anche le città limitrofe, creando la vasta area metropolitana milanese. In ambito culturale, Milano è il principale centro italiano dell'editoria ed è ai vertici del circuito musicale mondiale grazie alla stagione lirica del Teatro alla Scala e alla sua lunga tradizione operistica. È, inoltre, tra i principali poli fieristici europei (con due esposizioni universali ospitate: Expo 1906 e Expo 2015) e del disegno industriale, ed è considerata una delle capitali mondiali della moda.", "Il Duomo di Milano è il principale luogo di culto cattolico della città di Milano e cattedrale dell'omonima arcidiocesi. Altri edifici di valore sono le basiliche paleocristiane di Milano: la basilica di Sant'Ambrogio, considerata da sempre il massimo esempio dell'architettura romanica lombarda,[100] nonché uno dei più antichi monumenti dell'arte cristiana, la basilica di San Lorenzo, edificio a pianta centrale bizantino conosciuto anche per la vicinanza delle omonime colonne, la basilica di San Nazaro in Brolo e la basilica di San Simpliciano, legate come le precedenti alla figura di sant'Ambrogio."]}, {"source_sentence": "Quando è avvenuto lo sbarco in Normandia?", "sentences": ["Lo sbarco in Normandia (nome in codice operazione Neptune, parte marittima della più ampia operazione Overlord) fu una delle più grandi invasioni anfibie della storia, messa in atto dalle forze alleate durante la seconda guerra mondiale per aprire un secondo fronte in Europa, dirigersi verso la Germania nazista e allo stesso tempo alleggerire il fronte orientale, sul quale da tre anni l'Armata Rossa stava sostenendo un aspro confronto contro i tedeschi.", "La Normandia è una regione storica ed amministrativa della Francia. Si affaccia a nord sul canale della Manica e confina a sud-ovest con la Bretagna, a sud con i Paesi della Loira e con il Centro-Valle della Loira, a est con l'Île-de-France e a nord-est con l'Alta Francia. Occupa la bassa vallata della Senna (Alta Normandia) e si estende verso ovest fino alla penisola del Cotentin (Bassa Normandia).", "L'invasione iniziò nelle prime ore di martedì 6 giugno 1944 (data conosciuta come D-Day), quando toccarono terra nella penisola del Cotentin e nella zona di Caen le truppe alleate aviotrasportate, che aprirono la strada alle forze terrestri. All'alba del 6 giugno, precedute da un imponente bombardamento aeronavale, le fanterie sbarcarono su cinque spiagge."]}, {"source_sentence": "Quando è nato Luca?", "sentences": ["Luca è nato nel 1990", "Luca ha un fratello di nome Lorenzo e una sorella di nome Chiara", "Luca lavora come sviluppatore"]}], "model-index": [{"name": "BGE base WikipediaQA-ita", "results": [{"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "dim 768", "type": "dim_768"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.7534636553425698, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.8732207249952553, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.8883089770354906, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.9052002277472007, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.7534636553425698, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.2910735749984184, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.17766179540709814, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.09052002277472007, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.7534636553425698, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.8732207249952553, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.8883089770354906, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.9052002277472007, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.8384852729647283, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.816234142799904, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.8179246209727115, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "dim 512", "type": "dim_512"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.7517555513380147, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.8708483583222623, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.8864110836970962, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.9031125450749667, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.7517555513380147, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.29028278610742075, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.17728221673941924, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.09031125450749669, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.7517555513380147, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.8708483583222623, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.8864110836970962, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.9031125450749667, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.8364916692827067, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.8142769402946844, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.8159454559810521, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "dim 256", "type": "dim_256"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.7410324539760865, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.8637312583032833, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.878724615676599, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.8953311823875498, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.7410324539760865, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.2879104194344278, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.1757449231353198, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.089533118238755, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.7410324539760865, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.8637312583032833, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.878724615676599, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.8953311823875498, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.8275013364480452, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.8048458790108056, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.8065838557898369, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "dim 128", "type": "dim_128"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.7194913645853103, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.844942114253179, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.8621180489656481, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.8800531410134751, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.7194913645853103, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.28164737141772633, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.17242360979312965, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.08800531410134751, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.7194913645853103, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.844942114253179, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.8621180489656481, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.8800531410134751, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.8089255548002888, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.7852223020542415, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.7872945920851742, "name": "Cosine Map@100"}]}, {"task": {"type": "information-retrieval", "name": "Information Retrieval"}, "dataset": {"name": "dim 64", "type": "dim_64"}, "metrics": [{"type": "cosine_accuracy@1", "value": 0.6789742360017088, "name": "Cosine Accuracy@1"}, {"type": "cosine_accuracy@3", "value": 0.811169140387136, "name": "Cosine Accuracy@3"}, {"type": "cosine_accuracy@5", "value": 0.8296035881283715, "name": "Cosine Accuracy@5"}, {"type": "cosine_accuracy@10", "value": 0.8477945467750865, "name": "Cosine Accuracy@10"}, {"type": "cosine_precision@1", "value": 0.6789742360017088, "name": "Cosine Precision@1"}, {"type": "cosine_precision@3", "value": 0.27038971346237867, "name": "Cosine Precision@3"}, {"type": "cosine_precision@5", "value": 0.16592071762567434, "name": "Cosine Precision@5"}, {"type": "cosine_precision@10", "value": 0.08477945467750865, "name": "Cosine Precision@10"}, {"type": "cosine_recall@1", "value": 0.6789742360017088, "name": "Cosine Recall@1"}, {"type": "cosine_recall@3", "value": 0.811169140387136, "name": "Cosine Recall@3"}, {"type": "cosine_recall@5", "value": 0.8296035881283715, "name": "Cosine Recall@5"}, {"type": "cosine_recall@10", "value": 0.8477945467750865, "name": "Cosine Recall@10"}, {"type": "cosine_ndcg@10", "value": 0.774416217728066, "name": "Cosine Ndcg@10"}, {"type": "cosine_mrr@10", "value": 0.7482063280040872, "name": "Cosine Mrr@10"}, {"type": "cosine_map@100", "value": 0.7505568607057078, "name": "Cosine Map@100"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,908 |
aroot/eng-guj-simcse_nbrs_r
|
aroot
|
translation
|
[
"transformers",
"pytorch",
"tensorboard",
"mbart",
"text2text-generation",
"translation",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-07-18T19:48:02Z |
2023-07-18T22:12:52+00:00
| 15 | 0 |
---
metrics:
- bleu
tags:
- translation
- generated_from_trainer
model-index:
- name: eng-guj-simcse_nbrs_r
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# eng-guj-simcse_nbrs_r
This model is a fine-tuned version of [facebook/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3536
- Bleu: 2.6731
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# eng-guj-simcse_nbrs_r
This model is a fine-tuned version of [facebook/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3536
- Bleu: 2.6731
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3
|
{"metrics": ["bleu"], "tags": ["translation", "generated_from_trainer"], "model-index": [{"name": "eng-guj-simcse_nbrs_r", "results": []}]}
|
task
|
[
"TRANSLATION"
] | 39,909 |
Iker/TowerInstruct-13B-v0.1-EN2ES
|
Iker
|
text-generation
|
[
"transformers",
"safetensors",
"llama",
"text-generation",
"translation",
"conversational",
"en",
"es",
"dataset:Iker/InstructTranslation-EN-ES",
"base_model:Unbabel/TowerInstruct-13B-v0.1",
"base_model:finetune:Unbabel/TowerInstruct-13B-v0.1",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2024-04-02T09:39:03Z |
2024-04-16T14:40:36+00:00
| 16 | 0 |
---
base_model: Unbabel/TowerInstruct-13B-v0.1
datasets:
- Iker/InstructTranslation-EN-ES
language:
- en
- es
library_name: transformers
license: apache-2.0
pipeline_tag: text-generation
tags:
- translation
---
This is a [TowerInstruct-13B](https://huggingface.co/Unbabel/TowerInstruct-13B-v0.1) model fine-tuned for translating instructions datasets from English into Spanish. This model has GPT4 translation quality, but you can run it on your own machine for free 🎉
The model has been finetuned using ~1.500 prompts and answers from [teknium/OpenHermes-2.5](teknium/OpenHermes-2.5) translated to Spanish using GPT-4-0125-preview. The dataset is available here: https://huggingface.co/datasets/Iker/InstructTranslation-EN-ES/
This model was finetuned using [axolotl](https://github.com/OpenAccess-AI-Collective/axolotl), the training config is available here: https://huggingface.co/Iker/TowerInstruct-13B-v0.1-EN2ES/blob/main/Tower13B.yml
# Demo
```python
import torch
from transformers import pipeline
og = pipeline("text-generation", model="Unbabel/TowerInstruct-13B-v0.1", torch_dtype=torch.bfloat16, device_map=0)
fn7 = pipeline("text-generation", model="Iker/TowerInstruct-7B-v0.2-EN2ES", torch_dtype=torch.bfloat16, device_map=1)
fn = pipeline("text-generation", model="Iker/TowerInstruct-13B-v0.1-EN2ES", torch_dtype=torch.bfloat16, device_map=2)
msg = """
Let's use Bayes' theorem again to solve this problem:\n\nLet A represent the event that the man actually has the ability to predict dice rolls with 90% accuracy, and C represent the event of predicting correctly on the first attempt.\n\nWe want to find P(A|C), the probability that the man actually has the ability given that he predicted correctly on his first attempt.\n\nBayes' theorem states that P(A|C) = P(C|A) * P(A) / P(C)\n\nFirst, let's find P(C|A): the probability of predicting correctly on the first attempt if the man actually has the ability. Since he claims 90% accuracy, this probability is 0.9.\n\nNext, let's find P(A): the probability that someone actually has the ability to predict dice rolls with 90% accuracy. We are told this is 1%, so P(A) = 0.01.\n\nNow we need to find P(C): the overall probability of predicting correctly on the first attempt. This can be calculated as the sum of probabilities for each case: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), where ¬A represents not having the ability and P(¬A) = 1 - P(A) = 0.99.\n\nTo find P(C|¬A), the probability of predicting correctly on the first attempt without the ability, we use the fact that there's a 1/6 chance of guessing correctly by random chance: P(C|¬A) = 1/6.\n\nSo, P(C) = (0.9)*(0.01) + (1/6)*(0.99) = 0.009 + 0.165 = 0.174.\n\nFinally, we can calculate P(A|C) using Bayes' theorem:\n\nP(A|C) = P(C|A) * P(A) / P(C) = (0.9)*(0.01) / (0.174) ≈ 0.0517.\n\nTherefore, the probability that the man actually has the ability to predict dice rolls with 90% accuracy is approximately 5.17%.
""".strip()
messages = [
{"role": "user",
"content": f"Translate the following text from English into Spanish.\n{msg}\nSpanish:"},
]
prompt = og.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = og(prompt, max_new_tokens=1024, do_sample=False)
print(outputs[0]["generated_text"])
prompt = fn7.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = fn7(prompt, max_new_tokens=1024, do_sample=False)
print(outputs[0]["generated_text"])
prompt = fn.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = fn(prompt, max_new_tokens=1024, do_sample=False)
print(outputs[0]["generated_text"])
```
### Unbabel/TowerInstruct-13B-v0.1
This model fails with very large inputs, the answer is in Russian instead of Spanish 🥴
```
Воспроизвем теорему Байеса для решения этой задачи:
Предположим, что A означает событие, когда человек действительно обладает способностью предсказывать результаты броска кубика с точностью 90%, а C означает событие правильного предсказания на первой попытке.
Мы хотим найти P(A|C), вероятность того, что у человека действительно есть способность, если он правильно предсказал на первой попытке.
Теорема Байеса утверждает, что P(A|C) = P(C|A) * P(A) / P(C)
Сначала определим P(C|A): вероятность правильного предсказания на первой попытке, если у человека действительно есть способность. Поскольку он утверждает, что его точность составляет 90%, эта вероятность равна 0,9.
Далее определим P(A): вероятность того, что у кого-то действительно есть способность предсказывать результаты броска кубика с точностью 90%. Нам сказано, что она составляет 1%, поэтому P(A) = 0,01.
Теперь нам нужно найти P(C): общую вероятность правильного предсказания на первой попытке. Это можно рассчитать как сумму вероятностей для каждого случая: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), где ¬A означает отсутствие способности, и P(¬A) = 1 - P(A) = 0,99.
Чтобы найти P(C|¬A), вероятность правильного предсказания на первой попытке без способности, мы используем тот факт, что существует 1/6 шанс сделать правильный выбор случайно: P(C|¬A) = 1/6.
Таким образом, P(C) = (0,9)*(0,01) / (0,174) ≈ 0,009 + 0,165 = 0,174.
Наконец, мы можем рассчитать P(A|C) с помощью теоремы Байеса:
P(A|C) = P(C|A) * P(A) / P(C) = (0,9)*(0,01) / (0,174) ≈ 0,0517.
Таким образом, вероятность того, что у человека действительно есть способность предсказывать результаты броска кубика с точностью 90%, составляет примерно 5,17%.
```
### Iker/TowerInstruct-7B-v0.2-EN2ES
```
Vamos a usar de nuevo el teorema de Bayes para resolver este problema:
A representa el evento de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión, y C representa el evento de predecir correctamente en el primer intento.
Queremos encontrar P(A|C), la probabilidad de que el hombre realmente tenga la capacidad dado que predecía correctamente en su primer intento.
El teorema de Bayes establece que P(A|C) = P(C|A) * P(A) / P(C)
Primero, vamos a encontrar P(C|A): la probabilidad de predecir correctamente en el primer intento si el hombre realmente tiene la capacidad. Dado que afirma un 90% de precisión, esta probabilidad es 0.9.
A continuación, vamos a encontrar P(A): la probabilidad de que alguien realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión. Nos dicen que esto es del 1%, así que P(A) = 0.01.
Ahora necesitamos encontrar P(C): la probabilidad total de predecir correctamente en el primer intento. Esto se puede calcular como la suma de probabilidades para cada caso: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), donde ¬A representa no tener la capacidad y P(¬A) = 1 - P(A) = 0.99.
Para encontrar P(C|¬A), la probabilidad de predecir correctamente en el primer intento sin la capacidad, usamos el hecho de que hay una probabilidad del 1/6 de adivinar correctamente por azar: P(C|¬A) = 1/6.
Así, P(C) = (0.9)*(0.01) + (1/6)*(0.99) = 0.009 + 0.165 = 0.174.
Finalmente, podemos calcular P(A|C) usando el teorema de Bayes:
P(A|C) = P(C|A) * P(A) / P(C) = (0.9)*(0.01) / (0.174) ≈ 0.0517.
Por lo tanto, la probabilidad de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión es aproximadamente del 5.17%.
```
### Iker/TowerInstruct-13B-v0.1-EN2ES
```
Vamos a usar de nuevo el teorema de Bayes para resolver este problema:
Sea A el evento de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión, y C el evento de predecir correctamente en el primer intento.
Queremos encontrar P(A|C), la probabilidad de que el hombre realmente tenga la capacidad dada que predijo correctamente en su primer intento.
El teorema de Bayes establece que P(A|C) = P(C|A) * P(A) / P(C)
Primero, vamos a encontrar P(C|A): la probabilidad de predecir correctamente en el primer intento si el hombre realmente tiene la capacidad. Dado que afirma un 90% de precisión, esta probabilidad es 0.9.
A continuación, vamos a encontrar P(A): la probabilidad de que alguien realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión. Se nos dice que este es 1%, así que P(A) = 0.01.
Ahora necesitamos encontrar P(C): la probabilidad general de predecir correctamente en el primer intento. Esto puede calcularse como la suma de probabilidades para cada caso: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), donde ¬A representa no tener la capacidad y P(¬A) = 1 - P(A) = 0.99.
Para encontrar P(C|¬A), la probabilidad de predecir correctamente en el primer intento sin la capacidad, utilizamos el hecho de que hay una probabilidad de 1/6 de adivinar correctamente por casualidad: P(C|¬A) = 1/6.
Así que, P(C) = (0.9)*(0.01) + (1/6)*(0.99) = 0.009 + 0.165 = 0.174.
Finalmente, podemos calcular P(A|C) usando el teorema de Bayes:
P(A|C) = P(C|A) * P(A) / P(C) = (0.9)*(0.01) / (0.174) ≈ 0.0517.
Por lo tanto, la probabilidad de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión es aproximadamente 5.17%.
```
| null |
Non_BioNLP
|
This is a [TowerInstruct-13B](https://huggingface.co/Unbabel/TowerInstruct-13B-v0.1) model fine-tuned for translating instructions datasets from English into Spanish. This model has GPT4 translation quality, but you can run it on your own machine for free 🎉
The model has been finetuned using ~1.500 prompts and answers from [teknium/OpenHermes-2.5](teknium/OpenHermes-2.5) translated to Spanish using GPT-4-0125-preview. The dataset is available here: https://huggingface.co/datasets/Iker/InstructTranslation-EN-ES/
This model was finetuned using [axolotl](https://github.com/OpenAccess-AI-Collective/axolotl), the training config is available here: https://huggingface.co/Iker/TowerInstruct-13B-v0.1-EN2ES/blob/main/Tower13B.yml
# Demo
```python
import torch
from transformers import pipeline
og = pipeline("text-generation", model="Unbabel/TowerInstruct-13B-v0.1", torch_dtype=torch.bfloat16, device_map=0)
fn7 = pipeline("text-generation", model="Iker/TowerInstruct-7B-v0.2-EN2ES", torch_dtype=torch.bfloat16, device_map=1)
fn = pipeline("text-generation", model="Iker/TowerInstruct-13B-v0.1-EN2ES", torch_dtype=torch.bfloat16, device_map=2)
msg = """
Let's use Bayes' theorem again to solve this problem:\n\nLet A represent the event that the man actually has the ability to predict dice rolls with 90% accuracy, and C represent the event of predicting correctly on the first attempt.\n\nWe want to find P(A|C), the probability that the man actually has the ability given that he predicted correctly on his first attempt.\n\nBayes' theorem states that P(A|C) = P(C|A) * P(A) / P(C)\n\nFirst, let's find P(C|A): the probability of predicting correctly on the first attempt if the man actually has the ability. Since he claims 90% accuracy, this probability is 0.9.\n\nNext, let's find P(A): the probability that someone actually has the ability to predict dice rolls with 90% accuracy. We are told this is 1%, so P(A) = 0.01.\n\nNow we need to find P(C): the overall probability of predicting correctly on the first attempt. This can be calculated as the sum of probabilities for each case: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), where ¬A represents not having the ability and P(¬A) = 1 - P(A) = 0.99.\n\nTo find P(C|¬A), the probability of predicting correctly on the first attempt without the ability, we use the fact that there's a 1/6 chance of guessing correctly by random chance: P(C|¬A) = 1/6.\n\nSo, P(C) = (0.9)*(0.01) + (1/6)*(0.99) = 0.009 + 0.165 = 0.174.\n\nFinally, we can calculate P(A|C) using Bayes' theorem:\n\nP(A|C) = P(C|A) * P(A) / P(C) = (0.9)*(0.01) / (0.174) ≈ 0.0517.\n\nTherefore, the probability that the man actually has the ability to predict dice rolls with 90% accuracy is approximately 5.17%.
""".strip()
messages = [
{"role": "user",
"content": f"Translate the following text from English into Spanish.\n{msg}\nSpanish:"},
]
prompt = og.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = og(prompt, max_new_tokens=1024, do_sample=False)
print(outputs[0]["generated_text"])
prompt = fn7.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = fn7(prompt, max_new_tokens=1024, do_sample=False)
print(outputs[0]["generated_text"])
prompt = fn.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = fn(prompt, max_new_tokens=1024, do_sample=False)
print(outputs[0]["generated_text"])
```
### Unbabel/TowerInstruct-13B-v0.1
This model fails with very large inputs, the answer is in Russian instead of Spanish 🥴
```
Воспроизвем теорему Байеса для решения этой задачи:
Предположим, что A означает событие, когда человек действительно обладает способностью предсказывать результаты броска кубика с точностью 90%, а C означает событие правильного предсказания на первой попытке.
Мы хотим найти P(A|C), вероятность того, что у человека действительно есть способность, если он правильно предсказал на первой попытке.
Теорема Байеса утверждает, что P(A|C) = P(C|A) * P(A) / P(C)
Сначала определим P(C|A): вероятность правильного предсказания на первой попытке, если у человека действительно есть способность. Поскольку он утверждает, что его точность составляет 90%, эта вероятность равна 0,9.
Далее определим P(A): вероятность того, что у кого-то действительно есть способность предсказывать результаты броска кубика с точностью 90%. Нам сказано, что она составляет 1%, поэтому P(A) = 0,01.
Теперь нам нужно найти P(C): общую вероятность правильного предсказания на первой попытке. Это можно рассчитать как сумму вероятностей для каждого случая: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), где ¬A означает отсутствие способности, и P(¬A) = 1 - P(A) = 0,99.
Чтобы найти P(C|¬A), вероятность правильного предсказания на первой попытке без способности, мы используем тот факт, что существует 1/6 шанс сделать правильный выбор случайно: P(C|¬A) = 1/6.
Таким образом, P(C) = (0,9)*(0,01) / (0,174) ≈ 0,009 + 0,165 = 0,174.
Наконец, мы можем рассчитать P(A|C) с помощью теоремы Байеса:
P(A|C) = P(C|A) * P(A) / P(C) = (0,9)*(0,01) / (0,174) ≈ 0,0517.
Таким образом, вероятность того, что у человека действительно есть способность предсказывать результаты броска кубика с точностью 90%, составляет примерно 5,17%.
```
### Iker/TowerInstruct-7B-v0.2-EN2ES
```
Vamos a usar de nuevo el teorema de Bayes para resolver este problema:
A representa el evento de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión, y C representa el evento de predecir correctamente en el primer intento.
Queremos encontrar P(A|C), la probabilidad de que el hombre realmente tenga la capacidad dado que predecía correctamente en su primer intento.
El teorema de Bayes establece que P(A|C) = P(C|A) * P(A) / P(C)
Primero, vamos a encontrar P(C|A): la probabilidad de predecir correctamente en el primer intento si el hombre realmente tiene la capacidad. Dado que afirma un 90% de precisión, esta probabilidad es 0.9.
A continuación, vamos a encontrar P(A): la probabilidad de que alguien realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión. Nos dicen que esto es del 1%, así que P(A) = 0.01.
Ahora necesitamos encontrar P(C): la probabilidad total de predecir correctamente en el primer intento. Esto se puede calcular como la suma de probabilidades para cada caso: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), donde ¬A representa no tener la capacidad y P(¬A) = 1 - P(A) = 0.99.
Para encontrar P(C|¬A), la probabilidad de predecir correctamente en el primer intento sin la capacidad, usamos el hecho de que hay una probabilidad del 1/6 de adivinar correctamente por azar: P(C|¬A) = 1/6.
Así, P(C) = (0.9)*(0.01) + (1/6)*(0.99) = 0.009 + 0.165 = 0.174.
Finalmente, podemos calcular P(A|C) usando el teorema de Bayes:
P(A|C) = P(C|A) * P(A) / P(C) = (0.9)*(0.01) / (0.174) ≈ 0.0517.
Por lo tanto, la probabilidad de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión es aproximadamente del 5.17%.
```
### Iker/TowerInstruct-13B-v0.1-EN2ES
```
Vamos a usar de nuevo el teorema de Bayes para resolver este problema:
Sea A el evento de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión, y C el evento de predecir correctamente en el primer intento.
Queremos encontrar P(A|C), la probabilidad de que el hombre realmente tenga la capacidad dada que predijo correctamente en su primer intento.
El teorema de Bayes establece que P(A|C) = P(C|A) * P(A) / P(C)
Primero, vamos a encontrar P(C|A): la probabilidad de predecir correctamente en el primer intento si el hombre realmente tiene la capacidad. Dado que afirma un 90% de precisión, esta probabilidad es 0.9.
A continuación, vamos a encontrar P(A): la probabilidad de que alguien realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión. Se nos dice que este es 1%, así que P(A) = 0.01.
Ahora necesitamos encontrar P(C): la probabilidad general de predecir correctamente en el primer intento. Esto puede calcularse como la suma de probabilidades para cada caso: P(C) = P(C|A) * P(A) + P(C|¬A) * P(¬A), donde ¬A representa no tener la capacidad y P(¬A) = 1 - P(A) = 0.99.
Para encontrar P(C|¬A), la probabilidad de predecir correctamente en el primer intento sin la capacidad, utilizamos el hecho de que hay una probabilidad de 1/6 de adivinar correctamente por casualidad: P(C|¬A) = 1/6.
Así que, P(C) = (0.9)*(0.01) + (1/6)*(0.99) = 0.009 + 0.165 = 0.174.
Finalmente, podemos calcular P(A|C) usando el teorema de Bayes:
P(A|C) = P(C|A) * P(A) / P(C) = (0.9)*(0.01) / (0.174) ≈ 0.0517.
Por lo tanto, la probabilidad de que el hombre realmente tenga la capacidad de predecir lanzamientos de dados con un 90% de precisión es aproximadamente 5.17%.
```
|
{"base_model": "Unbabel/TowerInstruct-13B-v0.1", "datasets": ["Iker/InstructTranslation-EN-ES"], "language": ["en", "es"], "library_name": "transformers", "license": "apache-2.0", "pipeline_tag": "text-generation", "tags": ["translation"]}
|
task
|
[
"TRANSLATION"
] | 39,910 |
gsasikiran/bart-base-finetuned-cnn
|
gsasikiran
|
summarization
|
[
"transformers",
"tensorboard",
"safetensors",
"bart",
"text2text-generation",
"summarization",
"newsarticles",
"en",
"dataset:abisee/cnn_dailymail",
"base_model:facebook/bart-base",
"base_model:finetune:facebook/bart-base",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2025-02-13T14:10:50Z |
2025-03-02T23:45:35+00:00
| 63 | 0 |
---
base_model:
- facebook/bart-base
datasets:
- abisee/cnn_dailymail
language:
- en
library_name: transformers
license: apache-2.0
metrics:
- rouge
pipeline_tag: summarization
tags:
- summarization
- newsarticles
---
| null |
Non_BioNLP
|
{"base_model": ["facebook/bart-base"], "datasets": ["abisee/cnn_dailymail"], "language": ["en"], "library_name": "transformers", "license": "apache-2.0", "metrics": ["rouge"], "pipeline_tag": "summarization", "tags": ["summarization", "newsarticles"]}
|
task
|
[
"SUMMARIZATION"
] | 39,911 |
|
gaudi/opus-mt-fr-tl-ctranslate2
|
gaudi
|
translation
|
[
"transformers",
"marian",
"ctranslate2",
"translation",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | 2024-07-25T15:15:01Z |
2024-10-19T04:49:56+00:00
| 6 | 0 |
---
license: apache-2.0
tags:
- ctranslate2
- translation
---
# Repository General Information
## Inspired by and derived from the work of [Helsinki-NLP](https://huggingface.co/Helsinki-NLP), [CTranslate2](https://github.com/OpenNMT/CTranslate2), and [michaelfeil](https://huggingface.co/michaelfeil)!
- Link to Original Model ([Helsinki-NLP](https://huggingface.co/Helsinki-NLP)): [Model Link](https://huggingface.co/Helsinki-NLP/opus-mt-fr-tl)
- This respository was based on the work of [CTranslate2](https://github.com/OpenNMT/CTranslate2).
- This repository was based on the work of [michaelfeil](https://huggingface.co/michaelfeil).
# What is CTranslate2?
[CTranslate2](https://opennmt.net/CTranslate2/) is a C++ and Python library for efficient inference with Transformer models.
CTranslate2 implements a custom runtime that applies many performance optimization techniques such as weights quantization, layers fusion, batch reordering, etc., to accelerate and reduce the memory usage of Transformer models on CPU and GPU.
CTranslate2 is one of the most performant ways of hosting translation models at scale. Current supported models include:
- Encoder-decoder models: Transformer base/big, M2M-100, NLLB, BART, mBART, Pegasus, T5, Whisper
- Decoder-only models: GPT-2, GPT-J, GPT-NeoX, OPT, BLOOM, MPT, Llama, Mistral, Gemma, CodeGen, GPTBigCode, Falcon
- Encoder-only models: BERT, DistilBERT, XLM-RoBERTa
The project is production-oriented and comes with backward compatibility guarantees, but it also includes experimental features related to model compression and inference acceleration.
# CTranslate2 Benchmarks
Please note that the results presented below are only valid for the configuration used during this benchmark: absolute and relative performance may change with different settings. Tested against `newstest2014` (En -> De) dataset.
The benchmark reports the number of target tokens generated per second (higher is better). The results are aggregated over multiple runs. See the benchmark scripts for more details and reproduce these numbers.
Please note that the results presented below are only valid for the configuration used during this benchmark: absolute and relative performance may change with different settings.
## CPU Benchmarks for Generic Opus-MT Models
| Library | Tokens per Second | Max Memory Usage | BLEU |
| :----: | :----: | :----: | :----: |
| Transformers 4.26.1 (with PyTorch 1.13.1) | 147.3 | 2332MB | 27.90 |
| Marian 1.11.0 (int16) | 330.2 | 5901MB | 27.65 |
| Marian 1.11.0 (int8) | 355.8 | 4763MB | 27.27 |
| CTranslate2 3.6.0 (int16) | 596.1 | 660MB | 27.53 |
| CTranslate2 3.6.0 (int8) | 696.1 | 516MB | 27.65 |
## GPU Benchmarks for Generic Opus-MT Models
| Library | Tokens per Second | Max GPU Memory Usage | Max Memory Usage | BLEU |
| :----: | :----: | :----: | :----: | :----: |
| Transformers 4.26.1 (with PyTorch 1.13.1) | 1022.9 | 4097MB | 2109MB | 27.90 |
| Marian 1.11.0 (float16) | 3962.4 | 3239MB | 1976MB | 27.94 |
| CTranslate2 3.6.0 (float16) | 9296.7 | 909MB | 814MB | 27.9 |
| CTranslate2 3.6.0 (int8 + float16) | 8362.7 | 813MB | 766MB | 27.9 |
`Executed with 4 threads on a c5.2xlarge Amazon EC2 instance equipped with an Intel(R) Xeon(R) Platinum 8275CL CPU.`
**Source to benchmark information can be found [here](https://github.com/OpenNMT/CTranslate2).**<br />
**Original model BLEU scores can be found [here](https://huggingface.co/Helsinki-NLP/opus-mt-fr-tl).**
## Internal Benchmarks
Internal testing on our end showed **inference times reduced by 6x-10x** on average compared the vanilla checkpoints using the *transformers* library. A **slight reduction on BLEU scores (~5%)** was also identified in comparison to the vanilla checkpoints with a few exceptions. This is likely due to several factors, one being the quantization applied. Further testing is needed from our end to better assess the reduction in translation quality. The command used to compile the vanilla checkpoint into a CTranslate2 model can be found below. Modifying this command can yield differing balances between inferencing performance and translation quality.
# CTranslate2 Installation
```bash
pip install hf-hub-ctranslate2>=1.0.0 ctranslate2>=3.13.0
```
### ct2-transformers-converter Command Used:
```bash
ct2-transformers-converter --model Helsinki-NLP/opus-mt-fr-tl --output_dir ./ctranslate2/opus-mt-fr-tl-ctranslate2 --force --copy_files README.md generation_config.json tokenizer_config.json vocab.json source.spm .gitattributes target.spm --quantization float16
```
# CTranslate2 Converted Checkpoint Information:
**Compatible With:**
- [ctranslate2](https://github.com/OpenNMT/CTranslate2)
- [hf-hub-ctranslate2](https://github.com/michaelfeil/hf-hub-ctranslate2)
**Compute Type:**
- `compute_type=int8_float16` for `device="cuda"`
- `compute_type=int8` for `device="cpu"`
# Sample Code - ctranslate2
#### Clone the repository to the working directory or wherever you wish to store the model artifacts. ####
```bash
git clone https://huggingface.co/gaudi/opus-mt-fr-tl-ctranslate2
```
#### Take the python code below and update the 'model_dir' variable to the location of the cloned repository. ####
```python
from ctranslate2 import Translator
import transformers
model_dir = "./opus-mt-fr-tl-ctranslate2" # Path to model directory.
translator = Translator(
model_path=model_dir,
device="cuda", # cpu, cuda, or auto.
inter_threads=1, # Maximum number of parallel translations.
intra_threads=4, # Number of OpenMP threads per translator.
compute_type="int8_float16", # int8 for cpu or int8_float16 for cuda.
)
tokenizer = transformers.AutoTokenizer.from_pretrained(model_dir)
source = tokenizer.convert_ids_to_tokens(tokenizer.encode("XXXXXX, XXX XX XXXXXX."))
results = translator.translate_batch([source])
target = results[0].hypotheses[0]
print(tokenizer.decode(tokenizer.convert_tokens_to_ids(target)))
```
# Sample Code - hf-hub-ctranslate2
**Derived From [michaelfeil](https://huggingface.co/michaelfeil):**
```python
from hf_hub_ctranslate2 import TranslatorCT2fromHfHub, GeneratorCT2fromHfHub
from transformers import AutoTokenizer
model_name = "gaudi/opus-mt-fr-tl-ctranslate2"
model = TranslatorCT2fromHfHub(
model_name_or_path=model_name,
device="cuda",
compute_type="int8_float16",
tokenizer=AutoTokenizer.from_pretrained(model_name)
)
outputs = model.generate(
text=["XXX XX XXX XXXXXXX XXXX?", "XX XX XXXX XX XXX!"],
)
print(outputs)
```
# License and other remarks:
License conditions are intended to be idential to [original huggingface repository](https://huggingface.co/Helsinki-NLP/opus-mt-fr-tl) by Helsinki-NLP.
| null |
Non_BioNLP
|
# Repository General Information
## Inspired by and derived from the work of [Helsinki-NLP](https://huggingface.co/Helsinki-NLP), [CTranslate2](https://github.com/OpenNMT/CTranslate2), and [michaelfeil](https://huggingface.co/michaelfeil)!
- Link to Original Model ([Helsinki-NLP](https://huggingface.co/Helsinki-NLP)): [Model Link](https://huggingface.co/Helsinki-NLP/opus-mt-fr-tl)
- This respository was based on the work of [CTranslate2](https://github.com/OpenNMT/CTranslate2).
- This repository was based on the work of [michaelfeil](https://huggingface.co/michaelfeil).
# What is CTranslate2?
[CTranslate2](https://opennmt.net/CTranslate2/) is a C++ and Python library for efficient inference with Transformer models.
CTranslate2 implements a custom runtime that applies many performance optimization techniques such as weights quantization, layers fusion, batch reordering, etc., to accelerate and reduce the memory usage of Transformer models on CPU and GPU.
CTranslate2 is one of the most performant ways of hosting translation models at scale. Current supported models include:
- Encoder-decoder models: Transformer base/big, M2M-100, NLLB, BART, mBART, Pegasus, T5, Whisper
- Decoder-only models: GPT-2, GPT-J, GPT-NeoX, OPT, BLOOM, MPT, Llama, Mistral, Gemma, CodeGen, GPTBigCode, Falcon
- Encoder-only models: BERT, DistilBERT, XLM-RoBERTa
The project is production-oriented and comes with backward compatibility guarantees, but it also includes experimental features related to model compression and inference acceleration.
# CTranslate2 Benchmarks
Please note that the results presented below are only valid for the configuration used during this benchmark: absolute and relative performance may change with different settings. Tested against `newstest2014` (En -> De) dataset.
The benchmark reports the number of target tokens generated per second (higher is better). The results are aggregated over multiple runs. See the benchmark scripts for more details and reproduce these numbers.
Please note that the results presented below are only valid for the configuration used during this benchmark: absolute and relative performance may change with different settings.
## CPU Benchmarks for Generic Opus-MT Models
| Library | Tokens per Second | Max Memory Usage | BLEU |
| :----: | :----: | :----: | :----: |
| Transformers 4.26.1 (with PyTorch 1.13.1) | 147.3 | 2332MB | 27.90 |
| Marian 1.11.0 (int16) | 330.2 | 5901MB | 27.65 |
| Marian 1.11.0 (int8) | 355.8 | 4763MB | 27.27 |
| CTranslate2 3.6.0 (int16) | 596.1 | 660MB | 27.53 |
| CTranslate2 3.6.0 (int8) | 696.1 | 516MB | 27.65 |
## GPU Benchmarks for Generic Opus-MT Models
| Library | Tokens per Second | Max GPU Memory Usage | Max Memory Usage | BLEU |
| :----: | :----: | :----: | :----: | :----: |
| Transformers 4.26.1 (with PyTorch 1.13.1) | 1022.9 | 4097MB | 2109MB | 27.90 |
| Marian 1.11.0 (float16) | 3962.4 | 3239MB | 1976MB | 27.94 |
| CTranslate2 3.6.0 (float16) | 9296.7 | 909MB | 814MB | 27.9 |
| CTranslate2 3.6.0 (int8 + float16) | 8362.7 | 813MB | 766MB | 27.9 |
`Executed with 4 threads on a c5.2xlarge Amazon EC2 instance equipped with an Intel(R) Xeon(R) Platinum 8275CL CPU.`
**Source to benchmark information can be found [here](https://github.com/OpenNMT/CTranslate2).**<br />
**Original model BLEU scores can be found [here](https://huggingface.co/Helsinki-NLP/opus-mt-fr-tl).**
## Internal Benchmarks
Internal testing on our end showed **inference times reduced by 6x-10x** on average compared the vanilla checkpoints using the *transformers* library. A **slight reduction on BLEU scores (~5%)** was also identified in comparison to the vanilla checkpoints with a few exceptions. This is likely due to several factors, one being the quantization applied. Further testing is needed from our end to better assess the reduction in translation quality. The command used to compile the vanilla checkpoint into a CTranslate2 model can be found below. Modifying this command can yield differing balances between inferencing performance and translation quality.
# CTranslate2 Installation
```bash
pip install hf-hub-ctranslate2>=1.0.0 ctranslate2>=3.13.0
```
### ct2-transformers-converter Command Used:
```bash
ct2-transformers-converter --model Helsinki-NLP/opus-mt-fr-tl --output_dir ./ctranslate2/opus-mt-fr-tl-ctranslate2 --force --copy_files README.md generation_config.json tokenizer_config.json vocab.json source.spm .gitattributes target.spm --quantization float16
```
# CTranslate2 Converted Checkpoint Information:
**Compatible With:**
- [ctranslate2](https://github.com/OpenNMT/CTranslate2)
- [hf-hub-ctranslate2](https://github.com/michaelfeil/hf-hub-ctranslate2)
**Compute Type:**
- `compute_type=int8_float16` for `device="cuda"`
- `compute_type=int8` for `device="cpu"`
# Sample Code - ctranslate2
#### Clone the repository to the working directory or wherever you wish to store the model artifacts. ####
```bash
git clone https://huggingface.co/gaudi/opus-mt-fr-tl-ctranslate2
```
#### Take the python code below and update the 'model_dir' variable to the location of the cloned repository. ####
```python
from ctranslate2 import Translator
import transformers
model_dir = "./opus-mt-fr-tl-ctranslate2" # Path to model directory.
translator = Translator(
model_path=model_dir,
device="cuda", # cpu, cuda, or auto.
inter_threads=1, # Maximum number of parallel translations.
intra_threads=4, # Number of OpenMP threads per translator.
compute_type="int8_float16", # int8 for cpu or int8_float16 for cuda.
)
tokenizer = transformers.AutoTokenizer.from_pretrained(model_dir)
source = tokenizer.convert_ids_to_tokens(tokenizer.encode("XXXXXX, XXX XX XXXXXX."))
results = translator.translate_batch([source])
target = results[0].hypotheses[0]
print(tokenizer.decode(tokenizer.convert_tokens_to_ids(target)))
```
# Sample Code - hf-hub-ctranslate2
**Derived From [michaelfeil](https://huggingface.co/michaelfeil):**
```python
from hf_hub_ctranslate2 import TranslatorCT2fromHfHub, GeneratorCT2fromHfHub
from transformers import AutoTokenizer
model_name = "gaudi/opus-mt-fr-tl-ctranslate2"
model = TranslatorCT2fromHfHub(
model_name_or_path=model_name,
device="cuda",
compute_type="int8_float16",
tokenizer=AutoTokenizer.from_pretrained(model_name)
)
outputs = model.generate(
text=["XXX XX XXX XXXXXXX XXXX?", "XX XX XXXX XX XXX!"],
)
print(outputs)
```
# License and other remarks:
License conditions are intended to be idential to [original huggingface repository](https://huggingface.co/Helsinki-NLP/opus-mt-fr-tl) by Helsinki-NLP.
|
{"license": "apache-2.0", "tags": ["ctranslate2", "translation"]}
|
task
|
[
"TRANSLATION"
] | 39,912 |
ritheshwar/autotrain-codet5_base_cpsl-3727399186
|
ritheshwar
|
translation
|
[
"transformers",
"pytorch",
"t5",
"text2text-generation",
"autotrain",
"translation",
"unk",
"dataset:ritheshwar/autotrain-data-codet5_base_cpsl",
"co2_eq_emissions",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2023-02-25T09:17:08Z |
2023-02-25T09:19:59+00:00
| 14 | 0 |
---
datasets:
- ritheshwar/autotrain-data-codet5_base_cpsl
language:
- unk
- unk
tags:
- autotrain
- translation
co2_eq_emissions:
emissions: 3.846331276578152
---
# Model Trained Using AutoTrain
- Problem type: Translation
- Model ID: 3727399186
- CO2 Emissions (in grams): 3.8463
## Validation Metrics
- Loss: 0.223
- SacreBLEU: 2.566
- Gen len: 19.000
| null |
Non_BioNLP
|
# Model Trained Using AutoTrain
- Problem type: Translation
- Model ID: 3727399186
- CO2 Emissions (in grams): 3.8463
## Validation Metrics
- Loss: 0.223
- SacreBLEU: 2.566
- Gen len: 19.000
|
{"datasets": ["ritheshwar/autotrain-data-codet5_base_cpsl"], "language": ["unk", "unk"], "tags": ["autotrain", "translation"], "co2_eq_emissions": {"emissions": 3.846331276578152}}
|
task
|
[
"TRANSLATION"
] | 39,913 |
HPLT/translate-nn-en-v1.0-hplt_opus
|
HPLT
|
translation
|
[
"transformers",
"pytorch",
"marian",
"text2text-generation",
"translation",
"nn",
"en",
"license:cc-by-4.0",
"autotrain_compatible",
"region:us"
] | 2024-02-27T11:14:54Z |
2024-03-14T00:56:53+00:00
| 14 | 0 |
---
language:
- nn
- en
license: cc-by-4.0
tags:
- translation
inference: false
---
## HPLT MT release v1.0
This repository contains the translation model for Norwegian-English trained with OPUS and HPLT data. The model is available in both Marian and Hugging Face formats.
### Model Info
* Source language: Norwegian
* Target language: English
* Dataset: OPUS and HPLT data
* Model architecture: Transformer-base
* Tokenizer: SentencePiece (Unigram)
* Cleaning: We used [OpusCleaner](https://github.com/hplt-project/OpusCleaner) with a set of basic rules. Details can be found in the filter files [here](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/en-nn/raw/v2).
You can check out our [deliverable report](https://hplt-project.org/HPLT_D5_1___Translation_models_for_select_language_pairs.pdf), [GitHub repository](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0), and [website](https://hplt-project.org) for more details.
### Usage
The model has been trained with [MarianNMT](https://github.com/marian-nmt/marian) and the weights are in the Marian format. We have also converted the model into the Hugging Face format so it is compatible with `transformers`.
#### Using Marian
To run inference with MarianNMT, refer to the [Inference/Decoding/Translation](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0#inferencedecodingtranslation) section of our GitHub repository. You will need the model file `model.npz.best-chrf.npz` and the vocabulary file `model.nn-en.spm` from this repository.
#### Using transformers
We have also converted this model to the Hugging Face format and you can get started with the script below. **Note** that due a [known issue](https://github.com/huggingface/transformers/issues/26216) in weight conversion, the checkpoint cannot work with transformer versions <4.26 or >4.30. We tested and suggest `pip install transformers==4.28`.
```
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("HPLT/translate-nn-en-v1.0-hplt_opus")
model = AutoModelForSeq2SeqLM.from_pretrained("HPLT/translate-nn-en-v1.0-hplt_opus")
inputs = ["Input goes here.", "Make sure the language is right."]
batch_tokenized = tokenizer(inputs, return_tensors="pt", padding=True)
model_output = model.generate(
**batch_tokenized, num_beams=6, max_new_tokens=512
)
batch_detokenized = tokenizer.batch_decode(
model_output,
skip_special_tokens=True,
)
print(batch_detokenized)
```
### Benchmarks
When decoded using Marian, the model has the following test scores.
| Test set | BLEU | chrF++ | COMET22 |
| -------------------------------------- | ---- | ----- | ----- |
| FLORES200 | 23.1 | 47.3 | 0.7042 |
| NTREX | 21.1 | 44.6 | 0.669 |
### Acknowledgements
This project has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No 101070350 and from UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee [grant number 10052546]
Brought to you by researchers from the University of Edinburgh and Charles University in Prague with support from the whole HPLT consortium.
| null |
Non_BioNLP
|
## HPLT MT release v1.0
This repository contains the translation model for Norwegian-English trained with OPUS and HPLT data. The model is available in both Marian and Hugging Face formats.
### Model Info
* Source language: Norwegian
* Target language: English
* Dataset: OPUS and HPLT data
* Model architecture: Transformer-base
* Tokenizer: SentencePiece (Unigram)
* Cleaning: We used [OpusCleaner](https://github.com/hplt-project/OpusCleaner) with a set of basic rules. Details can be found in the filter files [here](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/en-nn/raw/v2).
You can check out our [deliverable report](https://hplt-project.org/HPLT_D5_1___Translation_models_for_select_language_pairs.pdf), [GitHub repository](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0), and [website](https://hplt-project.org) for more details.
### Usage
The model has been trained with [MarianNMT](https://github.com/marian-nmt/marian) and the weights are in the Marian format. We have also converted the model into the Hugging Face format so it is compatible with `transformers`.
#### Using Marian
To run inference with MarianNMT, refer to the [Inference/Decoding/Translation](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0#inferencedecodingtranslation) section of our GitHub repository. You will need the model file `model.npz.best-chrf.npz` and the vocabulary file `model.nn-en.spm` from this repository.
#### Using transformers
We have also converted this model to the Hugging Face format and you can get started with the script below. **Note** that due a [known issue](https://github.com/huggingface/transformers/issues/26216) in weight conversion, the checkpoint cannot work with transformer versions <4.26 or >4.30. We tested and suggest `pip install transformers==4.28`.
```
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("HPLT/translate-nn-en-v1.0-hplt_opus")
model = AutoModelForSeq2SeqLM.from_pretrained("HPLT/translate-nn-en-v1.0-hplt_opus")
inputs = ["Input goes here.", "Make sure the language is right."]
batch_tokenized = tokenizer(inputs, return_tensors="pt", padding=True)
model_output = model.generate(
**batch_tokenized, num_beams=6, max_new_tokens=512
)
batch_detokenized = tokenizer.batch_decode(
model_output,
skip_special_tokens=True,
)
print(batch_detokenized)
```
### Benchmarks
When decoded using Marian, the model has the following test scores.
| Test set | BLEU | chrF++ | COMET22 |
| -------------------------------------- | ---- | ----- | ----- |
| FLORES200 | 23.1 | 47.3 | 0.7042 |
| NTREX | 21.1 | 44.6 | 0.669 |
### Acknowledgements
This project has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No 101070350 and from UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee [grant number 10052546]
Brought to you by researchers from the University of Edinburgh and Charles University in Prague with support from the whole HPLT consortium.
|
{"language": ["nn", "en"], "license": "cc-by-4.0", "tags": ["translation"], "inference": false}
|
task
|
[
"TRANSLATION"
] | 39,914 |
platzi/platzi-distilroberta-base-mrpc-glue-paola-daft
|
platzi
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"roberta",
"text-classification",
"generated_from_trainer",
"dataset:glue",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-05-05T14:50:13Z |
2023-05-07T14:00:42+00:00
| 11 | 0 |
---
datasets:
- glue
license: apache-2.0
metrics:
- accuracy
- f1
tags:
- text-classification
- generated_from_trainer
widget:
- text:
- Yucaipa owned Dominick 's before selling the chain to Safeway in 1998 for $ 2.5
billion.
- Yucaipa bought Dominick's in 1995 for $ 693 million and sold it to Safeway for
$ 1.8 billion in 1998.
example_title: Not Equivalent
- text:
- Revenue in the first quarter of the year dropped 15 percent from the same period
a year earlier.
- With the scandal hanging over Stewart's company revenue the first quarter of the
year dropped 15 percent from the same period a year earlier.
example_title: Equivalent
model-index:
- name: platzi-distilroberta-base-mrpc-glue-paola-daft
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: datasetX
type: glue
config: mrpc
split: validation
args: mrpc
metrics:
- type: accuracy
value: 0.8382352941176471
name: Accuracy
- type: f1
value: 0.8749999999999999
name: F1
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# platzi-distilroberta-base-mrpc-glue-paola-daft
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the datasetX dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4514
- Accuracy: 0.8382
- F1: 0.8750
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.5291 | 1.09 | 500 | 0.4514 | 0.8382 | 0.8750 |
| 0.3759 | 2.18 | 1000 | 0.6055 | 0.8382 | 0.8740 |
### Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# platzi-distilroberta-base-mrpc-glue-paola-daft
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the datasetX dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4514
- Accuracy: 0.8382
- F1: 0.8750
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.5291 | 1.09 | 500 | 0.4514 | 0.8382 | 0.8750 |
| 0.3759 | 2.18 | 1000 | 0.6055 | 0.8382 | 0.8740 |
### Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
|
{"datasets": ["glue"], "license": "apache-2.0", "metrics": ["accuracy", "f1"], "tags": ["text-classification", "generated_from_trainer"], "widget": [{"text": ["Yucaipa owned Dominick 's before selling the chain to Safeway in 1998 for $ 2.5 billion.", "Yucaipa bought Dominick's in 1995 for $ 693 million and sold it to Safeway for $ 1.8 billion in 1998."], "example_title": "Not Equivalent"}, {"text": ["Revenue in the first quarter of the year dropped 15 percent from the same period a year earlier.", "With the scandal hanging over Stewart's company revenue the first quarter of the year dropped 15 percent from the same period a year earlier."], "example_title": "Equivalent"}], "model-index": [{"name": "platzi-distilroberta-base-mrpc-glue-paola-daft", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "datasetX", "type": "glue", "config": "mrpc", "split": "validation", "args": "mrpc"}, "metrics": [{"type": "accuracy", "value": 0.8382352941176471, "name": "Accuracy"}, {"type": "f1", "value": 0.8749999999999999, "name": "F1"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,915 |
KoalaAI/HateSpeechDetector
|
KoalaAI
|
text-classification
|
[
"transformers",
"pytorch",
"safetensors",
"deberta",
"text-classification",
"autotrain",
"social",
"offensive speech detection",
"moderation",
"en",
"dataset:tweet_eval",
"license:openrail",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-08-21T15:55:43Z |
2023-08-21T16:11:54+00:00
| 32 | 1 |
---
datasets:
- tweet_eval
language:
- en
license: openrail
tags:
- autotrain
- text-classification
- social
- offensive speech detection
- moderation
widget:
- text: I love cake!
- text: lets build the wall and deport illegals "they walk across the border like
this is Central park"
- text: EU offers to pay countries 6,000 euros per person to take in migrants
co2_eq_emissions:
emissions: 0.8636047103139626
---
# Hate Speech Detector
"Hate Speech Detector" is a text classification model based on Deberta that predicts whether a text contains hate speech or not.
The model is fine-tuned on the tweet_eval dataset, which consists of seven heterogeneous tasks in Twitter, all framed as multi-class tweet classification. The 'hate' subset is used for this task.
This model is part of our series in moderation models, which includes the following other models that may be of interest to you:
* [Offensive Speech Detector](https://huggingface.co/KoalaAI/OffensiveSpeechDetector)
We believe these models can be used in tandem to support one another and thus build a more robust moderation tool, for example.
## Intended uses & limitations
Offensive Speech Detector is intended to be used as a tool for detecting hate speech in texts, which can be useful for applications such as content moderation, sentiment analysis, or social media analysis. The model can be used to filter out or flag tweets that contain hate speech, or to analyze the prevalence and patterns of hate speech.
However, the model has some limitations that users should be aware of:
- The model is only trained and evaluated on tweets, which are short and informal texts that may contain slang, abbreviations, emojis, hashtags, or user mentions. The model may not perform well on other types of texts, such as news articles, essays, or books.
- The model is only trained and evaluated on English tweets. The model may not generalize well to other languages or dialects.
- The model is based on the tweet_eval dataset, which may have some biases or errors in the annotation process. The labels are assigned by human annotators, who may have different opinions or criteria for what constitutes hate speech. The dataset may also not cover all possible forms or contexts, such as sarcasm, irony, humor, or euphemism.
- The model is a statistical classifier that outputs a probability score for each label. The model does not provide any explanation or justification for its predictions. The model may also make mistakes or produce false positives or false negatives. Users should not blindly trust the model's predictions without further verification or human oversight.
## Ethical Considerations
This is a model that deals with sensitive and potentially harmful language. Users should consider the ethical implications and potential risks of using or deploying this model in their applications or contexts. Some of the ethical issues that may arise are:
- The model may reinforce or amplify existing biases or stereotypes in the data or in the society. For example, the model may associate certain words or topics with offensive language based on the frequency or co-occurrence in the data, without considering the meaning or intent behind them. This may result in unfair or inaccurate predictions for some groups or individuals.
Users should carefully consider the purpose, context, and impact of using this model, and take appropriate measures to prevent or mitigate any potential harm. Users should also respect the privacy and consent of the data subjects, and adhere to the relevant laws and regulations in their jurisdictions.
## License
This model is licensed under the CodeML OpenRAIL-M 0.1 license, which is a variant of the BigCode OpenRAIL-M license. This license allows you to freely access, use, modify, and distribute this model and its derivatives, for research, commercial or non-commercial purposes, as long as you comply with the following conditions:
- You must include a copy of the license and the original source of the model in any copies or derivatives of the model that you distribute.
- You must not use the model or its derivatives for any unlawful, harmful, abusive, discriminatory, or offensive purposes, or to cause or contribute to any social or environmental harm.
- You must respect the privacy and consent of the data subjects whose data was used to train or evaluate the model, and adhere to the relevant laws and regulations in your jurisdiction.
- You must acknowledge that the model and its derivatives are provided "as is", without any warranties or guarantees of any kind, and that the licensor is not liable for any damages or losses arising from your use of the model or its derivatives.
By accessing or using this model, you agree to be bound by the terms of this license. If you do not agree with the terms of this license, you must not access or use this model.
## Model Training Info
- Problem type: Multi-class Classification
- CO2 Emissions (in grams): 0.8636
## Validation Metrics
- Loss: 0.500
- Accuracy: 0.763
- Macro F1: 0.761
- Micro F1: 0.763
- Weighted F1: 0.764
- Macro Precision: 0.763
- Micro Precision: 0.763
- Weighted Precision: 0.775
- Macro Recall: 0.769
- Micro Recall: 0.763
- Weighted Recall: 0.763
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/models/KoalaAI/HateSpeechDetector
```
Or Python API:
```
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("KoalaAI/HateSpeechDetector", use_auth_token=True)
tokenizer = AutoTokenizer.from_pretrained("KoalaAI/HateSpeechDetector", use_auth_token=True)
inputs = tokenizer("I love AutoTrain", return_tensors="pt")
outputs = model(**inputs)
```
| null |
Non_BioNLP
|
# Hate Speech Detector
"Hate Speech Detector" is a text classification model based on Deberta that predicts whether a text contains hate speech or not.
The model is fine-tuned on the tweet_eval dataset, which consists of seven heterogeneous tasks in Twitter, all framed as multi-class tweet classification. The 'hate' subset is used for this task.
This model is part of our series in moderation models, which includes the following other models that may be of interest to you:
* [Offensive Speech Detector](https://huggingface.co/KoalaAI/OffensiveSpeechDetector)
We believe these models can be used in tandem to support one another and thus build a more robust moderation tool, for example.
## Intended uses & limitations
Offensive Speech Detector is intended to be used as a tool for detecting hate speech in texts, which can be useful for applications such as content moderation, sentiment analysis, or social media analysis. The model can be used to filter out or flag tweets that contain hate speech, or to analyze the prevalence and patterns of hate speech.
However, the model has some limitations that users should be aware of:
- The model is only trained and evaluated on tweets, which are short and informal texts that may contain slang, abbreviations, emojis, hashtags, or user mentions. The model may not perform well on other types of texts, such as news articles, essays, or books.
- The model is only trained and evaluated on English tweets. The model may not generalize well to other languages or dialects.
- The model is based on the tweet_eval dataset, which may have some biases or errors in the annotation process. The labels are assigned by human annotators, who may have different opinions or criteria for what constitutes hate speech. The dataset may also not cover all possible forms or contexts, such as sarcasm, irony, humor, or euphemism.
- The model is a statistical classifier that outputs a probability score for each label. The model does not provide any explanation or justification for its predictions. The model may also make mistakes or produce false positives or false negatives. Users should not blindly trust the model's predictions without further verification or human oversight.
## Ethical Considerations
This is a model that deals with sensitive and potentially harmful language. Users should consider the ethical implications and potential risks of using or deploying this model in their applications or contexts. Some of the ethical issues that may arise are:
- The model may reinforce or amplify existing biases or stereotypes in the data or in the society. For example, the model may associate certain words or topics with offensive language based on the frequency or co-occurrence in the data, without considering the meaning or intent behind them. This may result in unfair or inaccurate predictions for some groups or individuals.
Users should carefully consider the purpose, context, and impact of using this model, and take appropriate measures to prevent or mitigate any potential harm. Users should also respect the privacy and consent of the data subjects, and adhere to the relevant laws and regulations in their jurisdictions.
## License
This model is licensed under the CodeML OpenRAIL-M 0.1 license, which is a variant of the BigCode OpenRAIL-M license. This license allows you to freely access, use, modify, and distribute this model and its derivatives, for research, commercial or non-commercial purposes, as long as you comply with the following conditions:
- You must include a copy of the license and the original source of the model in any copies or derivatives of the model that you distribute.
- You must not use the model or its derivatives for any unlawful, harmful, abusive, discriminatory, or offensive purposes, or to cause or contribute to any social or environmental harm.
- You must respect the privacy and consent of the data subjects whose data was used to train or evaluate the model, and adhere to the relevant laws and regulations in your jurisdiction.
- You must acknowledge that the model and its derivatives are provided "as is", without any warranties or guarantees of any kind, and that the licensor is not liable for any damages or losses arising from your use of the model or its derivatives.
By accessing or using this model, you agree to be bound by the terms of this license. If you do not agree with the terms of this license, you must not access or use this model.
## Model Training Info
- Problem type: Multi-class Classification
- CO2 Emissions (in grams): 0.8636
## Validation Metrics
- Loss: 0.500
- Accuracy: 0.763
- Macro F1: 0.761
- Micro F1: 0.763
- Weighted F1: 0.764
- Macro Precision: 0.763
- Micro Precision: 0.763
- Weighted Precision: 0.775
- Macro Recall: 0.769
- Micro Recall: 0.763
- Weighted Recall: 0.763
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/models/KoalaAI/HateSpeechDetector
```
Or Python API:
```
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("KoalaAI/HateSpeechDetector", use_auth_token=True)
tokenizer = AutoTokenizer.from_pretrained("KoalaAI/HateSpeechDetector", use_auth_token=True)
inputs = tokenizer("I love AutoTrain", return_tensors="pt")
outputs = model(**inputs)
```
|
{"datasets": ["tweet_eval"], "language": ["en"], "license": "openrail", "tags": ["autotrain", "text-classification", "social", "offensive speech detection", "moderation"], "widget": [{"text": "I love cake!"}, {"text": "lets build the wall and deport illegals \"they walk across the border like this is Central park\""}, {"text": "EU offers to pay countries 6,000 euros per person to take in migrants"}], "co2_eq_emissions": {"emissions": 0.8636047103139626}}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,916 |
alelov/test-model-label1-MiniLMVERSION2
|
alelov
|
text-classification
|
[
"setfit",
"safetensors",
"bert",
"sentence-transformers",
"text-classification",
"generated_from_setfit_trainer",
"arxiv:2209.11055",
"base_model:sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
"base_model:finetune:sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
"model-index",
"region:us"
] | 2024-05-10T14:07:04Z |
2024-05-10T14:07:22+00:00
| 5 | 0 |
---
base_model: sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
library_name: setfit
metrics:
- accuracy
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
widget:
- text: Não apenas isso. A bola de neve do endividamento
- text: ' Bueno, yo lo que espero es que se traten con respeto, que se quieran. '
- text: ' SÃ, pues pedirle a MarÃa Luisa que le dé seguimiento y que siga atendiendo
las demandas de los ciudadanos de Vallarta, si te parece. Ya ella seguramente
nos está viendo y está tomando nota para darle continuidad a las demandas de
ambientalistas de Vallarta. '
- text: A confiança na economia despertou o apetite pelo risco, criando instrumentos
financeiros indispensáveis à captação de novos recursos para a expansão produtiva.
- text: " A ver, pon la carta de Elba Esther. Es que luego la borró. Fue en mayo\
\ del 23, 2 de mayo: ‘Ahà le espero con el Ejército —supuestamente esto\
\ es lo que le dijo Calderón a la maestra Elba Esther, ahà la espero con el\
\ Ejército— esa fue la respuesta del entonces presidente de México, Felipe\
\ Calderón, cuando le dije —según la maestra— que las y los maestros de\
\ México nos oponÃamos a que Miguel Ã\x81ngel Yunes continuara como titular\
\ del Issste, dadas las malversaciones de fondos financieros que con tanto trabajo\
\ las los trabajadores al servicio del Estado logramos con la reforma a dicha\
\ institución. ‘Cuando me comentó que Yunes estaba haciendo bien su trabajo,\
\ no me dejó más alternativa —dice la maestra— que advertirle que tomarÃamos\
\ las instalaciones del Issste y justo esa fue su respuesta: Ahà la espero con\
\ el Ejército. Esto sucedió en el marco de un evento público en una escuela\
\ secundaria técnica de la ahora Ciudad de México. Ante su respuesta, me levanté\
\ y me retiré. ‘Recordemos que la elección y remoción del director del Issste\
\ compete única y exclusivamente al titular del Ejecutivo federal y no a una\
\ servidora.’ Aquà me está contestando a mÃ, porque yo dije que a ella le\
\ habÃan entregado por ayudar en el fraude, que no me dirÃa la maestra que no\
\ ayudó en el fraude del 2006, y a cambio yo sostengo que le entregaron el Issste,\
\ la SubsecretarÃa de Educación Pública y la LoterÃa Nacional. ‘Por ello,\
\ en relación a las declaraciones hechas por el presidente Andrés Manuel López\
\ Obrador el pasado 29 de abril del presente año, sobre mi persona y la gestión\
\ del señor Miguel Ã\x81ngel Yunes al frente del Issste, le digo categóricamente\
\ que no participé el acto ilÃcito alguno, como me acusa desde su tribuna’.\
\ Yo no estoy acusando más que de haberse aliado con Calderón y ayudarle en\
\ el fraude electoral. ‘Siempre me he conducido conforme a derecho, de respeto\
\ a las instituciones de este paÃs y, desde luego, a la investidura presidencial.\
\ Por ello, señor presidente, basta de falsas acusaciones a mi persona’. No\
\ es nada personal, maestra, es que estamos viviendo un momento importantÃsimo\
\ de transformación. Entonces, como el compañero que viene a hacernos preguntas\
\ sobre salud, ayuda a recordar, porque es como si padecieran amnesia, ya se olvidó\
\ cómo era. Y antes esto no lo tocaban, era silencio, como vasallos, obedecer\
\ y callar, siempre y cuando hubiese dinero de por medio, porque lo que no suena\
\ lógico suena metálico. Entonces, hay que ir aclarando todo, seguir purificando\
\ la vida pública del paÃs y por eso son muy buenas estas mañaneras. Pero,\
\ bueno, eso es lo que querÃamos decir. ¿Qué se está haciendo? Procurar, ya\
\ es un compromiso, garantizar el derecho a la salud. Y vaya que ha costado, por\
\ estos intereses. ImagÃnense, no se podÃan comprar medicinas en el extranjero\
\ porque la ley lo prohibÃa, lo impedÃa; tuvimos que reformar la ley. ¿Y quiénes\
\ votaron en contra de que se pudiera comprar la medicina en el extranjero? El\
\ bloque conservador. ¿Qué son entonces? Representantes de minorÃas, no representantes\
\ del pueblo, esa es nuestra diferencia de fondo. No es nada personal, pero sÃ\
\ es importante el darle su sitio que le corresponde a lo público. República\
\ es, res publica, cosa pública. Si vivimos en una república, tenemos que pensar\
\ en eso, en lo público. Eso ya se habÃa olvidado. Entonces, vamos a continuar\
\ con lo mismo y va adelante todo el plan de transformación. El viernes vamos\
\ a informar sobre salud y luego vamos a informar en especÃfico sobre el Issste,\
\ porque ya llevamos… ¿Cuánto tiempo llevamos? "
inference: true
model-index:
- name: SetFit with sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: accuracy
value: 0.7727272727272727
name: Accuracy
---
# SetFit with sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
- **Sentence Transformer body:** [sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2)
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 128 tokens
- **Number of Classes:** 3 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
### Model Labels
| Label | Examples |
|:------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 0 | <ul><li>' No, si te estamos diciendo todo, nada más pregúntanos; ahora sí que como diría el clásico, tú pregúntame. '</li><li>' Pues nos vamos a poner de acuerdo, pero sí vamos a acatar ese resolutivo. '</li><li>' \xa0Lo más pronto posible, mañana. '</li></ul> |
| 1 | <ul><li>' Mujer y luego el compañero. '</li><li>' Bueno, que ayer fue un dÃ\xada muy importante, histórico, porque se ofreció perdón a los pueblos yaquis, que han sido vÃ\xadctimas de la represión durante mucho tiempo. Sin duda, es la etnia, es la cultura más reprimida en la historia, los yaquis. Y durante el porfiriato padecieron como una segunda conquista, fue una represión brutal. Se habla, ayer lo mencionaba, lo declara un gobernador de ese entonces de Sonora, que en la guerra contra los yaquis asesinaron a 15 mil yaquis, más los deportados; separaban a hombres y mujeres, los mandaban al sureste a trabajar como esclavos a Valle Nacional a la producción del tabaco, a las haciendas azucareras, al henequén en Yucatán. Una situación bárbara que avergüenza. Como a los mayas y a los mayos, y a los seris y a los totonacos, y a todos los pueblos originarios que padecieron de la represión durante la Colonia, y en el México independiente y en especial durante el porfiriato para quitarles sus tierras, sus bosques, sus aguas, era muy común en ese entonces que en el discurso oficial se hablara de que eran unos salvajes, un poco lo que todavÃ\xada se escucha, de que los invasores europeos vinieron a civilizarnos. Por eso este año de conmemoraciones el tÃ\xadtulo de todos los eventos, el distintivo es ‘La grandeza cultural de México’, porque esta es una gran nación, donde florecieron civilizaciones extraordinarias, espléndidas: los teotihuacanos, los toltecas, los olmecas, la cultura madre. Estamos hablando de que México se fundó hace miles de años. Entonces, ese argumento, que tiene que ver mucho con la opresión y con quien invade tierras o despoja, se usa para cometer todas esas injusticias, hay que descalificar al vencido, hay que justificar la represión. Entonces, eran bárbaros, salvajes.Â\xa0 Todos los generales del porfiriato más sanguinarios se formaron reprimiendo a indÃ\xadgenas. Huerta estuvo allá en Quintana Roo reprimiendo pueblos indÃ\xadgenas, Bernardo Reyes estuvo reprimiendo a los yaquis, y todos asÃ\xad se formaron. En ese entonces el Ejército adquirió armamento de lo más sofisticado, fÃ\xadjense, armas usadas por un ejército formado también por indÃ\xadgenas para reprimir a sus hermanos indÃ\xadgenas. Entonces, por eso fue el acto de ayer con los yaquis. Y no sólo fue ir a ofrecerles perdón y reconocerles por su resistencia, sino fue un acto para reafirmar nuestros compromisos de justicia, porque se está llevando a cabo un plan de justicia a los pueblos yaquis, que tiene tres vertientes: Primero, es el que tengan la tierra que les dotó el general Lázaro Cárdenas. Hubo resoluciones, acuerdos desde la época del general Cárdenas, pero no se cumplieron o empezaron a revertir esos acuerdos y se les quitaron las tierras, no todas, pero no tienen las tierras que se entregaron por resolución presidencial. Entonces, ayer ya entregamos dos mil 900 hectáreas y tenemos el compromiso de entregarles 20 mil. Lo segundo es el agua, porque también en la resolución del general Cárdenas se establecÃ\xada que las aguas del rÃ\xado Yaqui tenÃ\xadan que ser para los yaquis en un 50 por ciento. Está establecido y esto no sucede; al contrario, los pueblos yaquis no tienen agua ni siquiera para su uso doméstico. Entonces, ahora el compromiso es darles el agua que requieren los pueblos y se está creando con infraestructura moderna, canales, un nuevo distrito de riego que van a manejar ellos, esto es muy importante, por primera vez los pueblos originarios van a manejar un distrito de riego en el paÃ\xads, y van a ser los yaquis, y estamos contemplando una inversión total de cerca de 12 mil millones de pesos. Y tenemos trabajando servidores públicos, a los que les agradezco mucho, porque se han ido allá, Aarón Mastache, que está a cargo del plan hidráulico, porque tenemos 27 meses nada más, tenemos que terminar a finales del 23 para hacer los canales y poder regar hasta 60 mil hectáreas. Lo que el ingeniero Cárdenas planteó ayer es que se hizo en el gobierno anterior pues un acueducto hacia Hermosillo y esto generó una oposición de los pueblos, por lo mismo. Pero nosotros consideramos que con esta decisión que estamos tomando de entregar agua suficiente a los pueblos se puede resolver esta situación, porque también ahora cancelar eso nos va a resultar problemático, ya hay una inversión; además, es el agua para Hermosillo, que también se requiere.Â\xa0 Entonces, pensamos que nos va a alcanzar, porque también hay mucho desperdicio de agua, porque como no se invierte en infraestructura hidráulica, no se están recubriendo los canales para que el agua no se desperdicie. Y ahora la inversión que vamos a hacer va en ese sentido, o sea, no sólo de extraer más agua, de obtener más agua, que haya más abasto de agua, sino también vamos a rehabilitar todos los sistemas de riego, esto nos va a ayudar a resolver el problema.Â\xa0 Ah, y lo tercero es un plan integral para el bienestar, que tiene que ver con vivienda, que tiene que ver con atención médica, escuelas, se está creando la universidad de los pueblos yaquis, una nueva universidad pública, también estamos llegando a acuerdos. Todo esto consultado, aprobado por los gobernadores yaquis, no se impone nada, todo lo que hemos avanzado ha sido con el acuerdo de ellos. Se han llevado a cabo más de 100 asambleas, todo se les presenta, ellos son los que autorizan. Y ayer pues nos reunimos para firmar nuevos acuerdos y fue un hecho histórico muy importante. Y le agradecemos al ingeniero Cárdenas que nos acompañó, y fueron también representantes de otras etnias y de otras culturas del paÃ\xads. '</li><li>' Bueno, en cuanto al documental, no lo he visto, pero si existe una denuncia estoy seguro que la FiscalÃ\xada de la Ciudad de México le va a dar seguimiento, y lo he dicho aquÃ\xad en otras ocasiones, le tengo confianza a la fiscal, a la procuradora Ernestina Godoy. Es una mujer Ã\xadntegra, honesta, incapaz de encubrir o de ser cómplice de una injusticia. Entonces, es cosa de recurrir a ella. Acerca de los medios, sin enfadarnos, tenemos posturas distintas. Ellos formaron parte de un régimen que, yo sostengo, se dedicó a saquear a México. El distintivo de lo que llamaron neoliberalismo, que yo llamo neoporfirismo, fue el saqueo, el transferir los bienes públicos a particulares, los bienes del pueblo, de la nación a un grupo que le llamaron en su momento ‘grupo compacto’. Es una nueva oligarquÃ\xada que se creó son Salinas de Gortari. Entonces, a partir de entonces, para poder saquear a sus anchas, pues se creó una red de componendas y complicidades en donde no sólo eran los traficantes de influencia los beneficiados, sino eran dueños de medios de comunicación, periodistas. No todos, desde luego, ni todos los medios de comunicación, pero pues es de dominio público qué medios de comunicación, qué periodistas guardaron silencio cuando estaba saqueando a México como nunca en la historia.Â\xa0 Nunca hablaron de corrupción. Además, encubrieron delitos que se cometÃ\xadan como lo que significó la guerra contra el narcotráfico. Es también sabido que Calderón da a conocer que va a iniciar esa guerra y se atreve a decir que va a costar vidas, pero que no importa, y habla incluso de vidas de civiles, de inocentes. Y como fue una guerra con altos niveles de letalidad donde, si habÃ\xada un enfrentamiento, lo hemos visto muchas veces, entre el Ejército, las Fuerzas Armadas y la delincuencia, a los delincuentes, integrantes de bandas que quedaban heridos los remataban, eran masacres. Y estos medios de información, los que están ahora en contra de nosotros, hablando de que no se permite la libertad de expresión, fueron convocados por Calderón y les pidió que guardaran silencio sobre estas atrocidades. Ahora que se está llevando a cabo un cambio, una transformación, pues no sólo están ofendidos los que gozaban de privilegios en lo económico, los traficantes de influencia, sino también los dueños de medios de comunicación y hasta periodistas, desde luego intelectuales orgánicos y hasta gente de la academia, porque asÃ\xad es un proceso de transformación. Entonces, no hay que asombrarse, son dos proyectos distintos, contrapuestos de nación. Antes era saquear y repartirse el botÃ\xadn, y a unos les tocaba más, a otros menos, a quien no le tocaba nada era al pueblo raso, porque esta polÃ\xadtica consiste en que todos los intereses cuentan, menos el interés del pueblo. Entonces, ¿qué significa la transformación?, hacer a un lado a los privilegiados, acabar con la corrupción y todo lo ahorrado, todo lo que se robaban entregarlo a los más necesitados, a la mayorÃ\xada del pueblo, que es lo que estamos haciendo. Porque lo he dicho en otras ocasiones, no se conformaban con quedarse con una empresa, con un banco, con una mina, además saqueaban también el presupuesto público.Â\xa0 Como dicen los abogados, aceptando sin conceder, de que se quedaran con los bienes, con el sofisma, la mentira de que es mejor el sector privado que el sector público y que no hace falta el Estado, que basta con el mercado y que hay que diluir al Estado, desaparecerlo; que tampoco es del todo cierto porque es muy hipócrita ese discurso o esa pedacerÃ\xada de discurso, porque cuando necesitan al Estado lo usan, ¿para qué?, para rescatar a los bancos, para beneficiarse como lo hicieron con el Fobaproa. Si no se necesitara al Estado, ¿para que se rescató a empresarios y a banqueros?, ¿por qué no se dejó al mercado? Si en el libre mercado un comerciante, como sucedió y como sucede, pequeño le va mal, ni modo que el Estado lo rescate. Pero aquÃ\xad sÃ\xad lo consideraron válido rescatar a los de arriba y convertir las deudas privadas de unos cuantos en deuda pública con el famoso Fobaproa, que todavÃ\xada está ahÃ\xad como una deuda onerosa. Entonces, son dos proyectos distintos, contrapuestos de nación. Y baste… El otro dÃ\xada estaba yo viendo, yo creo que te mandé una foto de cuando Calderón llama… ¿O fue Peña? Creo que Calderón, porque fue cuando estaba la guerra. A ver, busquen eso, es una foto en donde convoca a todos los medios de información para guardar silencio. Creo que fue en el 11, que fue el año de más enfrentamientos y del mayor Ã\xadndice de letalidad, 2011, van a haber ahorita la foto. Son exactamente los que ahora están en contra de nosotros. Es normal y no hay ningún problema.Â\xa0 Tú me preguntas: ‘¿Cómo enfrentar eso?’ Pues contextualizando. Repito, no hay texto sin contexto, nada más dando los antecedentes, informando y ya la gente va a tener una idea de las cosas. Antes era muy difÃ\xadcil porque pues eran todos ellos y nosotros no tenÃ\xadamos manera. Ah sÃ\xad, es Calderón: ‘Celebra Calderón el acuerdo para la cobertura informativa de la violencia’. Pero ¿no puedes abrir la foto? SÃ\xad. Y mira, aunque no les vaya, no les guste, tú que los conoces más, ve diciendo quiénes son los que están ahÃ\xad. O sea, porque, ¿quiénes estaban?, ¿quiénes asistieron?, es un pacto de silencio. A lo mejor sÃ\xad se distinguen. '</li></ul> |
| 2 | <ul><li>' Setenta y cuatro por ciento; y en jóvenes también. Entonces, sí procuramos que se atienda a los más pobre y hay una política también de apoyo a mujeres. Vamos, precisamente mujeres, vamos a darle la palabra a una mujer, allá. '</li><li>'Un grupo pequeño, minoritario, no la mayorÃ\xada de las mujeres, se fue sobre la valla al quererla tirar con martillos, sopletes, y no pudieron porque es una valla alta y fuerte, pero eso evitó el choque y desgracias. Entonces, eso querÃ\xadamos informarlo y decirle al pueblo de México que todo terminó en santa paz. Y de nuevo felicitar las mujeres que participaron, por su actitud responsable y combativa al mismo tiempo. Pero se optó por la paz, no por la violencia. Y yo estoy seguro que si se sigue actuando de esa manera cada vez va a haber más mujeres manifestándose, con la seguridad de que no hay violencia y no hay espacio para la provocación y se va a desbordar hacia adelante, y qué bueno que esto suceda sin violencia. Pues ahora sÃ\xad, vamos Elizabeth. '</li><li>' SÃ\xad, pero está bien, esto es el Metro, miren, aquÃ\xad se les cayó el… Miren, esa es la valla, pero hay… debe haber otras. AhÃ\xad está, miren. Se quitaron como 60 martillos, sopletes, marros. Pero a ver si no hay una imagen. AhÃ\xad está, ahÃ\xad está el Palacio atrás. Eso es humo de extintores para pagar el fuego. Esta es una franja, la de enfrente, del Palacio y de la Catedral. En la plancha del Zócalo, la gente, mujeres manifestándose, que no participaron en esto, aquÃ\xad estamos hablando de 500 gentes más o menos, 500 personas, sobre todo mujeres, pero también ya al final hombres que se metieron, pero afortunadamente resisten las barreras. Ah, bueno, ese es un… Eso ya no es la…pero bueno, ese es un decomiso de bombas molotov y de martillos y… Pero ya. Prevenir, o sea, por lo que dices del fútbol. Comentaba yo que cuando —antes, yo creo que sigue haciéndose— cuando hay estos juegos de Pumas-América, América-Chivas, salen primero los de la porra de un equipo y luego al final los otros, se ordena para que no se enfrenten y se cuida la entrada. '</li></ul> |
## Evaluation
### Metrics
| Label | Accuracy |
|:--------|:---------|
| **all** | 0.7727 |
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("alelov/test-model-label1-MiniLMVERSION2")
# Run inference
preds = model("Não apenas isso. A bola de neve do endividamento")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Set Metrics
| Training set | Min | Median | Max |
|:-------------|:----|:---------|:-----|
| Word count | 1 | 103.4095 | 2340 |
| Label | Training Sample Count |
|:------|:----------------------|
| 0 | 311 |
| 1 | 27 |
| 2 | 21 |
### Training Hyperparameters
- batch_size: (16, 16)
- num_epochs: (4, 4)
- max_steps: -1
- sampling_strategy: oversampling
- body_learning_rate: (2e-05, 1e-05)
- head_learning_rate: 0.01
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: True
### Training Results
| Epoch | Step | Training Loss | Validation Loss |
|:-------:|:---------:|:-------------:|:---------------:|
| 0.0002 | 1 | 0.2894 | - |
| 0.0081 | 50 | 0.2562 | - |
| 0.0163 | 100 | 0.3346 | - |
| 0.0244 | 150 | 0.3106 | - |
| 0.0326 | 200 | 0.2452 | - |
| 0.0407 | 250 | 0.2848 | - |
| 0.0489 | 300 | 0.188 | - |
| 0.0570 | 350 | 0.1865 | - |
| 0.0651 | 400 | 0.1345 | - |
| 0.0733 | 450 | 0.1494 | - |
| 0.0814 | 500 | 0.1723 | - |
| 0.0896 | 550 | 0.0241 | - |
| 0.0977 | 600 | 0.0298 | - |
| 0.1058 | 650 | 0.01 | - |
| 0.1140 | 700 | 0.0354 | - |
| 0.1221 | 750 | 0.004 | - |
| 0.1303 | 800 | 0.0016 | - |
| 0.1384 | 850 | 0.0022 | - |
| 0.1466 | 900 | 0.0032 | - |
| 0.1547 | 950 | 0.0029 | - |
| 0.1628 | 1000 | 0.0009 | - |
| 0.1710 | 1050 | 0.0031 | - |
| 0.1791 | 1100 | 0.0525 | - |
| 0.1873 | 1150 | 0.0006 | - |
| 0.1954 | 1200 | 0.0007 | - |
| 0.2035 | 1250 | 0.0007 | - |
| 0.2117 | 1300 | 0.0014 | - |
| 0.2198 | 1350 | 0.0006 | - |
| 0.2280 | 1400 | 0.0071 | - |
| 0.2361 | 1450 | 0.0004 | - |
| 0.2443 | 1500 | 0.0003 | - |
| 0.2524 | 1550 | 0.0004 | - |
| 0.2605 | 1600 | 0.0019 | - |
| 0.2687 | 1650 | 0.0499 | - |
| 0.2768 | 1700 | 0.0004 | - |
| 0.2850 | 1750 | 0.0259 | - |
| 0.2931 | 1800 | 0.0002 | - |
| 0.3013 | 1850 | 0.0001 | - |
| 0.3094 | 1900 | 0.0003 | - |
| 0.3175 | 1950 | 0.0002 | - |
| 0.3257 | 2000 | 0.0003 | - |
| 0.3338 | 2050 | 0.0038 | - |
| 0.3420 | 2100 | 0.0001 | - |
| 0.3501 | 2150 | 0.0002 | - |
| 0.3582 | 2200 | 0.0002 | - |
| 0.3664 | 2250 | 0.0001 | - |
| 0.3745 | 2300 | 0.0001 | - |
| 0.3827 | 2350 | 0.0001 | - |
| 0.3908 | 2400 | 0.0044 | - |
| 0.3990 | 2450 | 0.0436 | - |
| 0.4071 | 2500 | 0.0002 | - |
| 0.4152 | 2550 | 0.0007 | - |
| 0.4234 | 2600 | 0.0001 | - |
| 0.4315 | 2650 | 0.0001 | - |
| 0.4397 | 2700 | 0.0001 | - |
| 0.4478 | 2750 | 0.0023 | - |
| 0.4560 | 2800 | 0.0001 | - |
| 0.4641 | 2850 | 0.0009 | - |
| 0.4722 | 2900 | 0.0001 | - |
| 0.4804 | 2950 | 0.0001 | - |
| 0.4885 | 3000 | 0.003 | - |
| 0.4967 | 3050 | 0.0001 | - |
| 0.5048 | 3100 | 0.0004 | - |
| 0.5129 | 3150 | 0.0 | - |
| 0.5211 | 3200 | 0.0001 | - |
| 0.5292 | 3250 | 0.0001 | - |
| 0.5374 | 3300 | 0.0 | - |
| 0.5455 | 3350 | 0.0 | - |
| 0.5537 | 3400 | 0.0001 | - |
| 0.5618 | 3450 | 0.0 | - |
| 0.5699 | 3500 | 0.0001 | - |
| 0.5781 | 3550 | 0.0 | - |
| 0.5862 | 3600 | 0.0 | - |
| 0.5944 | 3650 | 0.0 | - |
| 0.6025 | 3700 | 0.0 | - |
| 0.6106 | 3750 | 0.0 | - |
| 0.6188 | 3800 | 0.0001 | - |
| 0.6269 | 3850 | 0.0 | - |
| 0.6351 | 3900 | 0.0 | - |
| 0.6432 | 3950 | 0.0004 | - |
| 0.6514 | 4000 | 0.0004 | - |
| 0.6595 | 4050 | 0.0 | - |
| 0.6676 | 4100 | 0.0 | - |
| 0.6758 | 4150 | 0.0 | - |
| 0.6839 | 4200 | 0.0011 | - |
| 0.6921 | 4250 | 0.0006 | - |
| 0.7002 | 4300 | 0.0001 | - |
| 0.7084 | 4350 | 0.0 | - |
| 0.7165 | 4400 | 0.0 | - |
| 0.7246 | 4450 | 0.0 | - |
| 0.7328 | 4500 | 0.0 | - |
| 0.7409 | 4550 | 0.0 | - |
| 0.7491 | 4600 | 0.0 | - |
| 0.7572 | 4650 | 0.0 | - |
| 0.7653 | 4700 | 0.0 | - |
| 0.7735 | 4750 | 0.0041 | - |
| 0.7816 | 4800 | 0.0004 | - |
| 0.7898 | 4850 | 0.0006 | - |
| 0.7979 | 4900 | 0.0 | - |
| 0.8061 | 4950 | 0.0 | - |
| 0.8142 | 5000 | 0.0 | - |
| 0.8223 | 5050 | 0.0 | - |
| 0.8305 | 5100 | 0.0 | - |
| 0.8386 | 5150 | 0.0 | - |
| 0.8468 | 5200 | 0.0 | - |
| 0.8549 | 5250 | 0.0 | - |
| 0.8631 | 5300 | 0.0 | - |
| 0.8712 | 5350 | 0.0 | - |
| 0.8793 | 5400 | 0.0 | - |
| 0.8875 | 5450 | 0.0 | - |
| 0.8956 | 5500 | 0.0 | - |
| 0.9038 | 5550 | 0.0 | - |
| 0.9119 | 5600 | 0.0 | - |
| 0.9200 | 5650 | 0.0 | - |
| 0.9282 | 5700 | 0.0 | - |
| 0.9363 | 5750 | 0.0 | - |
| 0.9445 | 5800 | 0.0 | - |
| 0.9526 | 5850 | 0.0 | - |
| 0.9608 | 5900 | 0.0 | - |
| 0.9689 | 5950 | 0.0 | - |
| 0.9770 | 6000 | 0.0 | - |
| 0.9852 | 6050 | 0.0595 | - |
| 0.9933 | 6100 | 0.0001 | - |
| 1.0 | 6141 | - | 0.2767 |
| 1.0015 | 6150 | 0.0 | - |
| 1.0096 | 6200 | 0.0 | - |
| 1.0177 | 6250 | 0.0 | - |
| 1.0259 | 6300 | 0.0 | - |
| 1.0340 | 6350 | 0.0 | - |
| 1.0422 | 6400 | 0.0 | - |
| 1.0503 | 6450 | 0.0 | - |
| 1.0585 | 6500 | 0.0 | - |
| 1.0666 | 6550 | 0.0 | - |
| 1.0747 | 6600 | 0.0 | - |
| 1.0829 | 6650 | 0.0 | - |
| 1.0910 | 6700 | 0.0 | - |
| 1.0992 | 6750 | 0.0 | - |
| 1.1073 | 6800 | 0.0 | - |
| 1.1155 | 6850 | 0.0 | - |
| 1.1236 | 6900 | 0.0 | - |
| 1.1317 | 6950 | 0.0 | - |
| 1.1399 | 7000 | 0.0 | - |
| 1.1480 | 7050 | 0.0 | - |
| 1.1562 | 7100 | 0.0 | - |
| 1.1643 | 7150 | 0.0 | - |
| 1.1724 | 7200 | 0.0 | - |
| 1.1806 | 7250 | 0.0 | - |
| 1.1887 | 7300 | 0.0 | - |
| 1.1969 | 7350 | 0.0 | - |
| 1.2050 | 7400 | 0.0 | - |
| 1.2132 | 7450 | 0.0 | - |
| 1.2213 | 7500 | 0.0001 | - |
| 1.2294 | 7550 | 0.0 | - |
| 1.2376 | 7600 | 0.0 | - |
| 1.2457 | 7650 | 0.0 | - |
| 1.2539 | 7700 | 0.0 | - |
| 1.2620 | 7750 | 0.0 | - |
| 1.2702 | 7800 | 0.0001 | - |
| 1.2783 | 7850 | 0.0 | - |
| 1.2864 | 7900 | 0.0 | - |
| 1.2946 | 7950 | 0.0002 | - |
| 1.3027 | 8000 | 0.0 | - |
| 1.3109 | 8050 | 0.0003 | - |
| 1.3190 | 8100 | 0.0588 | - |
| 1.3271 | 8150 | 0.0 | - |
| 1.3353 | 8200 | 0.0002 | - |
| 1.3434 | 8250 | 0.0 | - |
| 1.3516 | 8300 | 0.0 | - |
| 1.3597 | 8350 | 0.0 | - |
| 1.3679 | 8400 | 0.0261 | - |
| 1.3760 | 8450 | 0.0 | - |
| 1.3841 | 8500 | 0.0 | - |
| 1.3923 | 8550 | 0.0 | - |
| 1.4004 | 8600 | 0.0 | - |
| 1.4086 | 8650 | 0.0 | - |
| 1.4167 | 8700 | 0.0 | - |
| 1.4248 | 8750 | 0.0 | - |
| 1.4330 | 8800 | 0.0 | - |
| 1.4411 | 8850 | 0.0 | - |
| 1.4493 | 8900 | 0.0 | - |
| 1.4574 | 8950 | 0.0 | - |
| 1.4656 | 9000 | 0.0 | - |
| 1.4737 | 9050 | 0.0 | - |
| 1.4818 | 9100 | 0.0 | - |
| 1.4900 | 9150 | 0.0153 | - |
| 1.4981 | 9200 | 0.0 | - |
| 1.5063 | 9250 | 0.0 | - |
| 1.5144 | 9300 | 0.0 | - |
| 1.5226 | 9350 | 0.0 | - |
| 1.5307 | 9400 | 0.0 | - |
| 1.5388 | 9450 | 0.0003 | - |
| 1.5470 | 9500 | 0.0 | - |
| 1.5551 | 9550 | 0.0003 | - |
| 1.5633 | 9600 | 0.0 | - |
| 1.5714 | 9650 | 0.0 | - |
| 1.5795 | 9700 | 0.0 | - |
| 1.5877 | 9750 | 0.0 | - |
| 1.5958 | 9800 | 0.0 | - |
| 1.6040 | 9850 | 0.0 | - |
| 1.6121 | 9900 | 0.0 | - |
| 1.6203 | 9950 | 0.0 | - |
| 1.6284 | 10000 | 0.0 | - |
| 1.6365 | 10050 | 0.0 | - |
| 1.6447 | 10100 | 0.0 | - |
| 1.6528 | 10150 | 0.0 | - |
| 1.6610 | 10200 | 0.0 | - |
| 1.6691 | 10250 | 0.0 | - |
| 1.6773 | 10300 | 0.0 | - |
| 1.6854 | 10350 | 0.0 | - |
| 1.6935 | 10400 | 0.0 | - |
| 1.7017 | 10450 | 0.0 | - |
| 1.7098 | 10500 | 0.0 | - |
| 1.7180 | 10550 | 0.0 | - |
| 1.7261 | 10600 | 0.0 | - |
| 1.7342 | 10650 | 0.0 | - |
| 1.7424 | 10700 | 0.0 | - |
| 1.7505 | 10750 | 0.0 | - |
| 1.7587 | 10800 | 0.0 | - |
| 1.7668 | 10850 | 0.0 | - |
| 1.7750 | 10900 | 0.0 | - |
| 1.7831 | 10950 | 0.0 | - |
| 1.7912 | 11000 | 0.0 | - |
| 1.7994 | 11050 | 0.0 | - |
| 1.8075 | 11100 | 0.0 | - |
| 1.8157 | 11150 | 0.0 | - |
| 1.8238 | 11200 | 0.0 | - |
| 1.8319 | 11250 | 0.0 | - |
| 1.8401 | 11300 | 0.0 | - |
| 1.8482 | 11350 | 0.0 | - |
| 1.8564 | 11400 | 0.0 | - |
| 1.8645 | 11450 | 0.0 | - |
| 1.8727 | 11500 | 0.0 | - |
| 1.8808 | 11550 | 0.0 | - |
| 1.8889 | 11600 | 0.0 | - |
| 1.8971 | 11650 | 0.0 | - |
| 1.9052 | 11700 | 0.0 | - |
| 1.9134 | 11750 | 0.0 | - |
| 1.9215 | 11800 | 0.0 | - |
| 1.9297 | 11850 | 0.0 | - |
| 1.9378 | 11900 | 0.0006 | - |
| 1.9459 | 11950 | 0.0 | - |
| 1.9541 | 12000 | 0.0 | - |
| 1.9622 | 12050 | 0.0 | - |
| 1.9704 | 12100 | 0.0 | - |
| 1.9785 | 12150 | 0.0 | - |
| 1.9866 | 12200 | 0.0 | - |
| 1.9948 | 12250 | 0.0 | - |
| 2.0 | 12282 | - | 0.2742 |
| 2.0029 | 12300 | 0.0 | - |
| 2.0111 | 12350 | 0.0 | - |
| 2.0192 | 12400 | 0.0 | - |
| 2.0274 | 12450 | 0.0 | - |
| 2.0355 | 12500 | 0.0 | - |
| 2.0436 | 12550 | 0.0 | - |
| 2.0518 | 12600 | 0.0 | - |
| 2.0599 | 12650 | 0.0 | - |
| 2.0681 | 12700 | 0.0 | - |
| 2.0762 | 12750 | 0.0 | - |
| 2.0844 | 12800 | 0.0 | - |
| 2.0925 | 12850 | 0.0 | - |
| 2.1006 | 12900 | 0.0 | - |
| 2.1088 | 12950 | 0.0 | - |
| 2.1169 | 13000 | 0.0 | - |
| 2.1251 | 13050 | 0.0 | - |
| 2.1332 | 13100 | 0.0 | - |
| 2.1413 | 13150 | 0.0 | - |
| 2.1495 | 13200 | 0.0 | - |
| 2.1576 | 13250 | 0.0 | - |
| 2.1658 | 13300 | 0.0 | - |
| 2.1739 | 13350 | 0.0 | - |
| 2.1821 | 13400 | 0.0 | - |
| 2.1902 | 13450 | 0.0 | - |
| 2.1983 | 13500 | 0.0 | - |
| 2.2065 | 13550 | 0.0 | - |
| 2.2146 | 13600 | 0.0 | - |
| 2.2228 | 13650 | 0.0 | - |
| 2.2309 | 13700 | 0.0 | - |
| 2.2390 | 13750 | 0.0 | - |
| 2.2472 | 13800 | 0.0 | - |
| 2.2553 | 13850 | 0.0 | - |
| 2.2635 | 13900 | 0.0 | - |
| 2.2716 | 13950 | 0.0 | - |
| 2.2798 | 14000 | 0.0 | - |
| 2.2879 | 14050 | 0.0013 | - |
| 2.2960 | 14100 | 0.0 | - |
| 2.3042 | 14150 | 0.0 | - |
| 2.3123 | 14200 | 0.0 | - |
| 2.3205 | 14250 | 0.0 | - |
| 2.3286 | 14300 | 0.0 | - |
| 2.3368 | 14350 | 0.0 | - |
| 2.3449 | 14400 | 0.0 | - |
| 2.3530 | 14450 | 0.0019 | - |
| 2.3612 | 14500 | 0.0 | - |
| 2.3693 | 14550 | 0.0 | - |
| 2.3775 | 14600 | 0.0 | - |
| 2.3856 | 14650 | 0.0 | - |
| 2.3937 | 14700 | 0.0 | - |
| 2.4019 | 14750 | 0.0 | - |
| 2.4100 | 14800 | 0.0 | - |
| 2.4182 | 14850 | 0.0 | - |
| 2.4263 | 14900 | 0.0 | - |
| 2.4345 | 14950 | 0.0 | - |
| 2.4426 | 15000 | 0.0 | - |
| 2.4507 | 15050 | 0.0 | - |
| 2.4589 | 15100 | 0.0 | - |
| 2.4670 | 15150 | 0.0 | - |
| 2.4752 | 15200 | 0.0 | - |
| 2.4833 | 15250 | 0.0 | - |
| 2.4915 | 15300 | 0.0 | - |
| 2.4996 | 15350 | 0.0 | - |
| 2.5077 | 15400 | 0.0 | - |
| 2.5159 | 15450 | 0.0 | - |
| 2.5240 | 15500 | 0.0 | - |
| 2.5322 | 15550 | 0.0 | - |
| 2.5403 | 15600 | 0.0 | - |
| 2.5484 | 15650 | 0.0 | - |
| 2.5566 | 15700 | 0.0 | - |
| 2.5647 | 15750 | 0.0 | - |
| 2.5729 | 15800 | 0.0 | - |
| 2.5810 | 15850 | 0.0 | - |
| 2.5892 | 15900 | 0.0001 | - |
| 2.5973 | 15950 | 0.0 | - |
| 2.6054 | 16000 | 0.0 | - |
| 2.6136 | 16050 | 0.0 | - |
| 2.6217 | 16100 | 0.0 | - |
| 2.6299 | 16150 | 0.0 | - |
| 2.6380 | 16200 | 0.0 | - |
| 2.6461 | 16250 | 0.0 | - |
| 2.6543 | 16300 | 0.0 | - |
| 2.6624 | 16350 | 0.0 | - |
| 2.6706 | 16400 | 0.0 | - |
| 2.6787 | 16450 | 0.0 | - |
| 2.6869 | 16500 | 0.0 | - |
| 2.6950 | 16550 | 0.0 | - |
| 2.7031 | 16600 | 0.0 | - |
| 2.7113 | 16650 | 0.0002 | - |
| 2.7194 | 16700 | 0.0 | - |
| 2.7276 | 16750 | 0.0 | - |
| 2.7357 | 16800 | 0.0 | - |
| 2.7439 | 16850 | 0.0 | - |
| 2.7520 | 16900 | 0.0 | - |
| 2.7601 | 16950 | 0.0 | - |
| 2.7683 | 17000 | 0.0291 | - |
| 2.7764 | 17050 | 0.0 | - |
| 2.7846 | 17100 | 0.0 | - |
| 2.7927 | 17150 | 0.0 | - |
| 2.8008 | 17200 | 0.0 | - |
| 2.8090 | 17250 | 0.0 | - |
| 2.8171 | 17300 | 0.0 | - |
| 2.8253 | 17350 | 0.0 | - |
| 2.8334 | 17400 | 0.0 | - |
| 2.8416 | 17450 | 0.0 | - |
| 2.8497 | 17500 | 0.0 | - |
| 2.8578 | 17550 | 0.0 | - |
| 2.8660 | 17600 | 0.0 | - |
| 2.8741 | 17650 | 0.0 | - |
| 2.8823 | 17700 | 0.0 | - |
| 2.8904 | 17750 | 0.0 | - |
| 2.8986 | 17800 | 0.0 | - |
| 2.9067 | 17850 | 0.0 | - |
| 2.9148 | 17900 | 0.0 | - |
| 2.9230 | 17950 | 0.0 | - |
| 2.9311 | 18000 | 0.0 | - |
| 2.9393 | 18050 | 0.0 | - |
| 2.9474 | 18100 | 0.0 | - |
| 2.9555 | 18150 | 0.0 | - |
| 2.9637 | 18200 | 0.0 | - |
| 2.9718 | 18250 | 0.0 | - |
| 2.9800 | 18300 | 0.0 | - |
| 2.9881 | 18350 | 0.0 | - |
| 2.9963 | 18400 | 0.0 | - |
| **3.0** | **18423** | **-** | **0.2642** |
| 3.0044 | 18450 | 0.0012 | - |
| 3.0125 | 18500 | 0.0 | - |
| 3.0207 | 18550 | 0.0 | - |
| 3.0288 | 18600 | 0.0 | - |
| 3.0370 | 18650 | 0.0 | - |
| 3.0451 | 18700 | 0.0041 | - |
| 3.0532 | 18750 | 0.0 | - |
| 3.0614 | 18800 | 0.0 | - |
| 3.0695 | 18850 | 0.0 | - |
| 3.0777 | 18900 | 0.0 | - |
| 3.0858 | 18950 | 0.0 | - |
| 3.0940 | 19000 | 0.0 | - |
| 3.1021 | 19050 | 0.0 | - |
| 3.1102 | 19100 | 0.0 | - |
| 3.1184 | 19150 | 0.0 | - |
| 3.1265 | 19200 | 0.0 | - |
| 3.1347 | 19250 | 0.0 | - |
| 3.1428 | 19300 | 0.0 | - |
| 3.1510 | 19350 | 0.0 | - |
| 3.1591 | 19400 | 0.0 | - |
| 3.1672 | 19450 | 0.0 | - |
| 3.1754 | 19500 | 0.0014 | - |
| 3.1835 | 19550 | 0.0 | - |
| 3.1917 | 19600 | 0.0 | - |
| 3.1998 | 19650 | 0.0 | - |
| 3.2079 | 19700 | 0.0 | - |
| 3.2161 | 19750 | 0.0 | - |
| 3.2242 | 19800 | 0.0 | - |
| 3.2324 | 19850 | 0.0 | - |
| 3.2405 | 19900 | 0.0 | - |
| 3.2487 | 19950 | 0.0 | - |
| 3.2568 | 20000 | 0.0 | - |
| 3.2649 | 20050 | 0.0 | - |
| 3.2731 | 20100 | 0.0 | - |
| 3.2812 | 20150 | 0.0 | - |
| 3.2894 | 20200 | 0.0453 | - |
| 3.2975 | 20250 | 0.0 | - |
| 3.3057 | 20300 | 0.0 | - |
| 3.3138 | 20350 | 0.0 | - |
| 3.3219 | 20400 | 0.0 | - |
| 3.3301 | 20450 | 0.0 | - |
| 3.3382 | 20500 | 0.0 | - |
| 3.3464 | 20550 | 0.0 | - |
| 3.3545 | 20600 | 0.0 | - |
| 3.3626 | 20650 | 0.0 | - |
| 3.3708 | 20700 | 0.0 | - |
| 3.3789 | 20750 | 0.0 | - |
| 3.3871 | 20800 | 0.0 | - |
| 3.3952 | 20850 | 0.0 | - |
| 3.4034 | 20900 | 0.0 | - |
| 3.4115 | 20950 | 0.0 | - |
| 3.4196 | 21000 | 0.0 | - |
| 3.4278 | 21050 | 0.0 | - |
| 3.4359 | 21100 | 0.0 | - |
| 3.4441 | 21150 | 0.0 | - |
| 3.4522 | 21200 | 0.0 | - |
| 3.4603 | 21250 | 0.0 | - |
| 3.4685 | 21300 | 0.0 | - |
| 3.4766 | 21350 | 0.0 | - |
| 3.4848 | 21400 | 0.0 | - |
| 3.4929 | 21450 | 0.0 | - |
| 3.5011 | 21500 | 0.0 | - |
| 3.5092 | 21550 | 0.0 | - |
| 3.5173 | 21600 | 0.0 | - |
| 3.5255 | 21650 | 0.0 | - |
| 3.5336 | 21700 | 0.0 | - |
| 3.5418 | 21750 | 0.0 | - |
| 3.5499 | 21800 | 0.0 | - |
| 3.5581 | 21850 | 0.0 | - |
| 3.5662 | 21900 | 0.0 | - |
| 3.5743 | 21950 | 0.0 | - |
| 3.5825 | 22000 | 0.0 | - |
| 3.5906 | 22050 | 0.0 | - |
| 3.5988 | 22100 | 0.0 | - |
| 3.6069 | 22150 | 0.0 | - |
| 3.6150 | 22200 | 0.0 | - |
| 3.6232 | 22250 | 0.0 | - |
| 3.6313 | 22300 | 0.0 | - |
| 3.6395 | 22350 | 0.0 | - |
| 3.6476 | 22400 | 0.0 | - |
| 3.6558 | 22450 | 0.0 | - |
| 3.6639 | 22500 | 0.0 | - |
| 3.6720 | 22550 | 0.0 | - |
| 3.6802 | 22600 | 0.0 | - |
| 3.6883 | 22650 | 0.0 | - |
| 3.6965 | 22700 | 0.0 | - |
| 3.7046 | 22750 | 0.0 | - |
| 3.7128 | 22800 | 0.0 | - |
| 3.7209 | 22850 | 0.0 | - |
| 3.7290 | 22900 | 0.0 | - |
| 3.7372 | 22950 | 0.0 | - |
| 3.7453 | 23000 | 0.0 | - |
| 3.7535 | 23050 | 0.0 | - |
| 3.7616 | 23100 | 0.0 | - |
| 3.7697 | 23150 | 0.0 | - |
| 3.7779 | 23200 | 0.0 | - |
| 3.7860 | 23250 | 0.0 | - |
| 3.7942 | 23300 | 0.0 | - |
| 3.8023 | 23350 | 0.0 | - |
| 3.8105 | 23400 | 0.0 | - |
| 3.8186 | 23450 | 0.0 | - |
| 3.8267 | 23500 | 0.0 | - |
| 3.8349 | 23550 | 0.0 | - |
| 3.8430 | 23600 | 0.0 | - |
| 3.8512 | 23650 | 0.0 | - |
| 3.8593 | 23700 | 0.0 | - |
| 3.8674 | 23750 | 0.0 | - |
| 3.8756 | 23800 | 0.0 | - |
| 3.8837 | 23850 | 0.0 | - |
| 3.8919 | 23900 | 0.0 | - |
| 3.9000 | 23950 | 0.0 | - |
| 3.9082 | 24000 | 0.0 | - |
| 3.9163 | 24050 | 0.0 | - |
| 3.9244 | 24100 | 0.0 | - |
| 3.9326 | 24150 | 0.0 | - |
| 3.9407 | 24200 | 0.0 | - |
| 3.9489 | 24250 | 0.0 | - |
| 3.9570 | 24300 | 0.0 | - |
| 3.9652 | 24350 | 0.0 | - |
| 3.9733 | 24400 | 0.0 | - |
| 3.9814 | 24450 | 0.0 | - |
| 3.9896 | 24500 | 0.0 | - |
| 3.9977 | 24550 | 0.0 | - |
| 4.0 | 24564 | - | 0.2671 |
* The bold row denotes the saved checkpoint.
### Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.7.0
- Transformers: 4.40.2
- PyTorch: 2.2.1+cu121
- Datasets: 2.19.1
- Tokenizers: 0.19.1
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SetFit with sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
- **Sentence Transformer body:** [sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2)
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 128 tokens
- **Number of Classes:** 3 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
### Model Labels
| Label | Examples |
|:------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 0 | <ul><li>' No, si te estamos diciendo todo, nada más pregúntanos; ahora sí que como diría el clásico, tú pregúntame. '</li><li>' Pues nos vamos a poner de acuerdo, pero sí vamos a acatar ese resolutivo. '</li><li>' \xa0Lo más pronto posible, mañana. '</li></ul> |
| 1 | <ul><li>' Mujer y luego el compañero. '</li><li>' Bueno, que ayer fue un dÃ\xada muy importante, histórico, porque se ofreció perdón a los pueblos yaquis, que han sido vÃ\xadctimas de la represión durante mucho tiempo. Sin duda, es la etnia, es la cultura más reprimida en la historia, los yaquis. Y durante el porfiriato padecieron como una segunda conquista, fue una represión brutal. Se habla, ayer lo mencionaba, lo declara un gobernador de ese entonces de Sonora, que en la guerra contra los yaquis asesinaron a 15 mil yaquis, más los deportados; separaban a hombres y mujeres, los mandaban al sureste a trabajar como esclavos a Valle Nacional a la producción del tabaco, a las haciendas azucareras, al henequén en Yucatán. Una situación bárbara que avergüenza. Como a los mayas y a los mayos, y a los seris y a los totonacos, y a todos los pueblos originarios que padecieron de la represión durante la Colonia, y en el México independiente y en especial durante el porfiriato para quitarles sus tierras, sus bosques, sus aguas, era muy común en ese entonces que en el discurso oficial se hablara de que eran unos salvajes, un poco lo que todavÃ\xada se escucha, de que los invasores europeos vinieron a civilizarnos. Por eso este año de conmemoraciones el tÃ\xadtulo de todos los eventos, el distintivo es ‘La grandeza cultural de México’, porque esta es una gran nación, donde florecieron civilizaciones extraordinarias, espléndidas: los teotihuacanos, los toltecas, los olmecas, la cultura madre. Estamos hablando de que México se fundó hace miles de años. Entonces, ese argumento, que tiene que ver mucho con la opresión y con quien invade tierras o despoja, se usa para cometer todas esas injusticias, hay que descalificar al vencido, hay que justificar la represión. Entonces, eran bárbaros, salvajes.Â\xa0 Todos los generales del porfiriato más sanguinarios se formaron reprimiendo a indÃ\xadgenas. Huerta estuvo allá en Quintana Roo reprimiendo pueblos indÃ\xadgenas, Bernardo Reyes estuvo reprimiendo a los yaquis, y todos asÃ\xad se formaron. En ese entonces el Ejército adquirió armamento de lo más sofisticado, fÃ\xadjense, armas usadas por un ejército formado también por indÃ\xadgenas para reprimir a sus hermanos indÃ\xadgenas. Entonces, por eso fue el acto de ayer con los yaquis. Y no sólo fue ir a ofrecerles perdón y reconocerles por su resistencia, sino fue un acto para reafirmar nuestros compromisos de justicia, porque se está llevando a cabo un plan de justicia a los pueblos yaquis, que tiene tres vertientes: Primero, es el que tengan la tierra que les dotó el general Lázaro Cárdenas. Hubo resoluciones, acuerdos desde la época del general Cárdenas, pero no se cumplieron o empezaron a revertir esos acuerdos y se les quitaron las tierras, no todas, pero no tienen las tierras que se entregaron por resolución presidencial. Entonces, ayer ya entregamos dos mil 900 hectáreas y tenemos el compromiso de entregarles 20 mil. Lo segundo es el agua, porque también en la resolución del general Cárdenas se establecÃ\xada que las aguas del rÃ\xado Yaqui tenÃ\xadan que ser para los yaquis en un 50 por ciento. Está establecido y esto no sucede; al contrario, los pueblos yaquis no tienen agua ni siquiera para su uso doméstico. Entonces, ahora el compromiso es darles el agua que requieren los pueblos y se está creando con infraestructura moderna, canales, un nuevo distrito de riego que van a manejar ellos, esto es muy importante, por primera vez los pueblos originarios van a manejar un distrito de riego en el paÃ\xads, y van a ser los yaquis, y estamos contemplando una inversión total de cerca de 12 mil millones de pesos. Y tenemos trabajando servidores públicos, a los que les agradezco mucho, porque se han ido allá, Aarón Mastache, que está a cargo del plan hidráulico, porque tenemos 27 meses nada más, tenemos que terminar a finales del 23 para hacer los canales y poder regar hasta 60 mil hectáreas. Lo que el ingeniero Cárdenas planteó ayer es que se hizo en el gobierno anterior pues un acueducto hacia Hermosillo y esto generó una oposición de los pueblos, por lo mismo. Pero nosotros consideramos que con esta decisión que estamos tomando de entregar agua suficiente a los pueblos se puede resolver esta situación, porque también ahora cancelar eso nos va a resultar problemático, ya hay una inversión; además, es el agua para Hermosillo, que también se requiere.Â\xa0 Entonces, pensamos que nos va a alcanzar, porque también hay mucho desperdicio de agua, porque como no se invierte en infraestructura hidráulica, no se están recubriendo los canales para que el agua no se desperdicie. Y ahora la inversión que vamos a hacer va en ese sentido, o sea, no sólo de extraer más agua, de obtener más agua, que haya más abasto de agua, sino también vamos a rehabilitar todos los sistemas de riego, esto nos va a ayudar a resolver el problema.Â\xa0 Ah, y lo tercero es un plan integral para el bienestar, que tiene que ver con vivienda, que tiene que ver con atención médica, escuelas, se está creando la universidad de los pueblos yaquis, una nueva universidad pública, también estamos llegando a acuerdos. Todo esto consultado, aprobado por los gobernadores yaquis, no se impone nada, todo lo que hemos avanzado ha sido con el acuerdo de ellos. Se han llevado a cabo más de 100 asambleas, todo se les presenta, ellos son los que autorizan. Y ayer pues nos reunimos para firmar nuevos acuerdos y fue un hecho histórico muy importante. Y le agradecemos al ingeniero Cárdenas que nos acompañó, y fueron también representantes de otras etnias y de otras culturas del paÃ\xads. '</li><li>' Bueno, en cuanto al documental, no lo he visto, pero si existe una denuncia estoy seguro que la FiscalÃ\xada de la Ciudad de México le va a dar seguimiento, y lo he dicho aquÃ\xad en otras ocasiones, le tengo confianza a la fiscal, a la procuradora Ernestina Godoy. Es una mujer Ã\xadntegra, honesta, incapaz de encubrir o de ser cómplice de una injusticia. Entonces, es cosa de recurrir a ella. Acerca de los medios, sin enfadarnos, tenemos posturas distintas. Ellos formaron parte de un régimen que, yo sostengo, se dedicó a saquear a México. El distintivo de lo que llamaron neoliberalismo, que yo llamo neoporfirismo, fue el saqueo, el transferir los bienes públicos a particulares, los bienes del pueblo, de la nación a un grupo que le llamaron en su momento ‘grupo compacto’. Es una nueva oligarquÃ\xada que se creó son Salinas de Gortari. Entonces, a partir de entonces, para poder saquear a sus anchas, pues se creó una red de componendas y complicidades en donde no sólo eran los traficantes de influencia los beneficiados, sino eran dueños de medios de comunicación, periodistas. No todos, desde luego, ni todos los medios de comunicación, pero pues es de dominio público qué medios de comunicación, qué periodistas guardaron silencio cuando estaba saqueando a México como nunca en la historia.Â\xa0 Nunca hablaron de corrupción. Además, encubrieron delitos que se cometÃ\xadan como lo que significó la guerra contra el narcotráfico. Es también sabido que Calderón da a conocer que va a iniciar esa guerra y se atreve a decir que va a costar vidas, pero que no importa, y habla incluso de vidas de civiles, de inocentes. Y como fue una guerra con altos niveles de letalidad donde, si habÃ\xada un enfrentamiento, lo hemos visto muchas veces, entre el Ejército, las Fuerzas Armadas y la delincuencia, a los delincuentes, integrantes de bandas que quedaban heridos los remataban, eran masacres. Y estos medios de información, los que están ahora en contra de nosotros, hablando de que no se permite la libertad de expresión, fueron convocados por Calderón y les pidió que guardaran silencio sobre estas atrocidades. Ahora que se está llevando a cabo un cambio, una transformación, pues no sólo están ofendidos los que gozaban de privilegios en lo económico, los traficantes de influencia, sino también los dueños de medios de comunicación y hasta periodistas, desde luego intelectuales orgánicos y hasta gente de la academia, porque asÃ\xad es un proceso de transformación. Entonces, no hay que asombrarse, son dos proyectos distintos, contrapuestos de nación. Antes era saquear y repartirse el botÃ\xadn, y a unos les tocaba más, a otros menos, a quien no le tocaba nada era al pueblo raso, porque esta polÃ\xadtica consiste en que todos los intereses cuentan, menos el interés del pueblo. Entonces, ¿qué significa la transformación?, hacer a un lado a los privilegiados, acabar con la corrupción y todo lo ahorrado, todo lo que se robaban entregarlo a los más necesitados, a la mayorÃ\xada del pueblo, que es lo que estamos haciendo. Porque lo he dicho en otras ocasiones, no se conformaban con quedarse con una empresa, con un banco, con una mina, además saqueaban también el presupuesto público.Â\xa0 Como dicen los abogados, aceptando sin conceder, de que se quedaran con los bienes, con el sofisma, la mentira de que es mejor el sector privado que el sector público y que no hace falta el Estado, que basta con el mercado y que hay que diluir al Estado, desaparecerlo; que tampoco es del todo cierto porque es muy hipócrita ese discurso o esa pedacerÃ\xada de discurso, porque cuando necesitan al Estado lo usan, ¿para qué?, para rescatar a los bancos, para beneficiarse como lo hicieron con el Fobaproa. Si no se necesitara al Estado, ¿para que se rescató a empresarios y a banqueros?, ¿por qué no se dejó al mercado? Si en el libre mercado un comerciante, como sucedió y como sucede, pequeño le va mal, ni modo que el Estado lo rescate. Pero aquÃ\xad sÃ\xad lo consideraron válido rescatar a los de arriba y convertir las deudas privadas de unos cuantos en deuda pública con el famoso Fobaproa, que todavÃ\xada está ahÃ\xad como una deuda onerosa. Entonces, son dos proyectos distintos, contrapuestos de nación. Y baste… El otro dÃ\xada estaba yo viendo, yo creo que te mandé una foto de cuando Calderón llama… ¿O fue Peña? Creo que Calderón, porque fue cuando estaba la guerra. A ver, busquen eso, es una foto en donde convoca a todos los medios de información para guardar silencio. Creo que fue en el 11, que fue el año de más enfrentamientos y del mayor Ã\xadndice de letalidad, 2011, van a haber ahorita la foto. Son exactamente los que ahora están en contra de nosotros. Es normal y no hay ningún problema.Â\xa0 Tú me preguntas: ‘¿Cómo enfrentar eso?’ Pues contextualizando. Repito, no hay texto sin contexto, nada más dando los antecedentes, informando y ya la gente va a tener una idea de las cosas. Antes era muy difÃ\xadcil porque pues eran todos ellos y nosotros no tenÃ\xadamos manera. Ah sÃ\xad, es Calderón: ‘Celebra Calderón el acuerdo para la cobertura informativa de la violencia’. Pero ¿no puedes abrir la foto? SÃ\xad. Y mira, aunque no les vaya, no les guste, tú que los conoces más, ve diciendo quiénes son los que están ahÃ\xad. O sea, porque, ¿quiénes estaban?, ¿quiénes asistieron?, es un pacto de silencio. A lo mejor sÃ\xad se distinguen. '</li></ul> |
| 2 | <ul><li>' Setenta y cuatro por ciento; y en jóvenes también. Entonces, sí procuramos que se atienda a los más pobre y hay una política también de apoyo a mujeres. Vamos, precisamente mujeres, vamos a darle la palabra a una mujer, allá. '</li><li>'Un grupo pequeño, minoritario, no la mayorÃ\xada de las mujeres, se fue sobre la valla al quererla tirar con martillos, sopletes, y no pudieron porque es una valla alta y fuerte, pero eso evitó el choque y desgracias. Entonces, eso querÃ\xadamos informarlo y decirle al pueblo de México que todo terminó en santa paz. Y de nuevo felicitar las mujeres que participaron, por su actitud responsable y combativa al mismo tiempo. Pero se optó por la paz, no por la violencia. Y yo estoy seguro que si se sigue actuando de esa manera cada vez va a haber más mujeres manifestándose, con la seguridad de que no hay violencia y no hay espacio para la provocación y se va a desbordar hacia adelante, y qué bueno que esto suceda sin violencia. Pues ahora sÃ\xad, vamos Elizabeth. '</li><li>' SÃ\xad, pero está bien, esto es el Metro, miren, aquÃ\xad se les cayó el… Miren, esa es la valla, pero hay… debe haber otras. AhÃ\xad está, miren. Se quitaron como 60 martillos, sopletes, marros. Pero a ver si no hay una imagen. AhÃ\xad está, ahÃ\xad está el Palacio atrás. Eso es humo de extintores para pagar el fuego. Esta es una franja, la de enfrente, del Palacio y de la Catedral. En la plancha del Zócalo, la gente, mujeres manifestándose, que no participaron en esto, aquÃ\xad estamos hablando de 500 gentes más o menos, 500 personas, sobre todo mujeres, pero también ya al final hombres que se metieron, pero afortunadamente resisten las barreras. Ah, bueno, ese es un… Eso ya no es la…pero bueno, ese es un decomiso de bombas molotov y de martillos y… Pero ya. Prevenir, o sea, por lo que dices del fútbol. Comentaba yo que cuando —antes, yo creo que sigue haciéndose— cuando hay estos juegos de Pumas-América, América-Chivas, salen primero los de la porra de un equipo y luego al final los otros, se ordena para que no se enfrenten y se cuida la entrada. '</li></ul> |
## Evaluation
### Metrics
| Label | Accuracy |
|:--------|:---------|
| **all** | 0.7727 |
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("alelov/test-model-label1-MiniLMVERSION2")
# Run inference
preds = model("Não apenas isso. A bola de neve do endividamento")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Set Metrics
| Training set | Min | Median | Max |
|:-------------|:----|:---------|:-----|
| Word count | 1 | 103.4095 | 2340 |
| Label | Training Sample Count |
|:------|:----------------------|
| 0 | 311 |
| 1 | 27 |
| 2 | 21 |
### Training Hyperparameters
- batch_size: (16, 16)
- num_epochs: (4, 4)
- max_steps: -1
- sampling_strategy: oversampling
- body_learning_rate: (2e-05, 1e-05)
- head_learning_rate: 0.01
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: True
### Training Results
| Epoch | Step | Training Loss | Validation Loss |
|:-------:|:---------:|:-------------:|:---------------:|
| 0.0002 | 1 | 0.2894 | - |
| 0.0081 | 50 | 0.2562 | - |
| 0.0163 | 100 | 0.3346 | - |
| 0.0244 | 150 | 0.3106 | - |
| 0.0326 | 200 | 0.2452 | - |
| 0.0407 | 250 | 0.2848 | - |
| 0.0489 | 300 | 0.188 | - |
| 0.0570 | 350 | 0.1865 | - |
| 0.0651 | 400 | 0.1345 | - |
| 0.0733 | 450 | 0.1494 | - |
| 0.0814 | 500 | 0.1723 | - |
| 0.0896 | 550 | 0.0241 | - |
| 0.0977 | 600 | 0.0298 | - |
| 0.1058 | 650 | 0.01 | - |
| 0.1140 | 700 | 0.0354 | - |
| 0.1221 | 750 | 0.004 | - |
| 0.1303 | 800 | 0.0016 | - |
| 0.1384 | 850 | 0.0022 | - |
| 0.1466 | 900 | 0.0032 | - |
| 0.1547 | 950 | 0.0029 | - |
| 0.1628 | 1000 | 0.0009 | - |
| 0.1710 | 1050 | 0.0031 | - |
| 0.1791 | 1100 | 0.0525 | - |
| 0.1873 | 1150 | 0.0006 | - |
| 0.1954 | 1200 | 0.0007 | - |
| 0.2035 | 1250 | 0.0007 | - |
| 0.2117 | 1300 | 0.0014 | - |
| 0.2198 | 1350 | 0.0006 | - |
| 0.2280 | 1400 | 0.0071 | - |
| 0.2361 | 1450 | 0.0004 | - |
| 0.2443 | 1500 | 0.0003 | - |
| 0.2524 | 1550 | 0.0004 | - |
| 0.2605 | 1600 | 0.0019 | - |
| 0.2687 | 1650 | 0.0499 | - |
| 0.2768 | 1700 | 0.0004 | - |
| 0.2850 | 1750 | 0.0259 | - |
| 0.2931 | 1800 | 0.0002 | - |
| 0.3013 | 1850 | 0.0001 | - |
| 0.3094 | 1900 | 0.0003 | - |
| 0.3175 | 1950 | 0.0002 | - |
| 0.3257 | 2000 | 0.0003 | - |
| 0.3338 | 2050 | 0.0038 | - |
| 0.3420 | 2100 | 0.0001 | - |
| 0.3501 | 2150 | 0.0002 | - |
| 0.3582 | 2200 | 0.0002 | - |
| 0.3664 | 2250 | 0.0001 | - |
| 0.3745 | 2300 | 0.0001 | - |
| 0.3827 | 2350 | 0.0001 | - |
| 0.3908 | 2400 | 0.0044 | - |
| 0.3990 | 2450 | 0.0436 | - |
| 0.4071 | 2500 | 0.0002 | - |
| 0.4152 | 2550 | 0.0007 | - |
| 0.4234 | 2600 | 0.0001 | - |
| 0.4315 | 2650 | 0.0001 | - |
| 0.4397 | 2700 | 0.0001 | - |
| 0.4478 | 2750 | 0.0023 | - |
| 0.4560 | 2800 | 0.0001 | - |
| 0.4641 | 2850 | 0.0009 | - |
| 0.4722 | 2900 | 0.0001 | - |
| 0.4804 | 2950 | 0.0001 | - |
| 0.4885 | 3000 | 0.003 | - |
| 0.4967 | 3050 | 0.0001 | - |
| 0.5048 | 3100 | 0.0004 | - |
| 0.5129 | 3150 | 0.0 | - |
| 0.5211 | 3200 | 0.0001 | - |
| 0.5292 | 3250 | 0.0001 | - |
| 0.5374 | 3300 | 0.0 | - |
| 0.5455 | 3350 | 0.0 | - |
| 0.5537 | 3400 | 0.0001 | - |
| 0.5618 | 3450 | 0.0 | - |
| 0.5699 | 3500 | 0.0001 | - |
| 0.5781 | 3550 | 0.0 | - |
| 0.5862 | 3600 | 0.0 | - |
| 0.5944 | 3650 | 0.0 | - |
| 0.6025 | 3700 | 0.0 | - |
| 0.6106 | 3750 | 0.0 | - |
| 0.6188 | 3800 | 0.0001 | - |
| 0.6269 | 3850 | 0.0 | - |
| 0.6351 | 3900 | 0.0 | - |
| 0.6432 | 3950 | 0.0004 | - |
| 0.6514 | 4000 | 0.0004 | - |
| 0.6595 | 4050 | 0.0 | - |
| 0.6676 | 4100 | 0.0 | - |
| 0.6758 | 4150 | 0.0 | - |
| 0.6839 | 4200 | 0.0011 | - |
| 0.6921 | 4250 | 0.0006 | - |
| 0.7002 | 4300 | 0.0001 | - |
| 0.7084 | 4350 | 0.0 | - |
| 0.7165 | 4400 | 0.0 | - |
| 0.7246 | 4450 | 0.0 | - |
| 0.7328 | 4500 | 0.0 | - |
| 0.7409 | 4550 | 0.0 | - |
| 0.7491 | 4600 | 0.0 | - |
| 0.7572 | 4650 | 0.0 | - |
| 0.7653 | 4700 | 0.0 | - |
| 0.7735 | 4750 | 0.0041 | - |
| 0.7816 | 4800 | 0.0004 | - |
| 0.7898 | 4850 | 0.0006 | - |
| 0.7979 | 4900 | 0.0 | - |
| 0.8061 | 4950 | 0.0 | - |
| 0.8142 | 5000 | 0.0 | - |
| 0.8223 | 5050 | 0.0 | - |
| 0.8305 | 5100 | 0.0 | - |
| 0.8386 | 5150 | 0.0 | - |
| 0.8468 | 5200 | 0.0 | - |
| 0.8549 | 5250 | 0.0 | - |
| 0.8631 | 5300 | 0.0 | - |
| 0.8712 | 5350 | 0.0 | - |
| 0.8793 | 5400 | 0.0 | - |
| 0.8875 | 5450 | 0.0 | - |
| 0.8956 | 5500 | 0.0 | - |
| 0.9038 | 5550 | 0.0 | - |
| 0.9119 | 5600 | 0.0 | - |
| 0.9200 | 5650 | 0.0 | - |
| 0.9282 | 5700 | 0.0 | - |
| 0.9363 | 5750 | 0.0 | - |
| 0.9445 | 5800 | 0.0 | - |
| 0.9526 | 5850 | 0.0 | - |
| 0.9608 | 5900 | 0.0 | - |
| 0.9689 | 5950 | 0.0 | - |
| 0.9770 | 6000 | 0.0 | - |
| 0.9852 | 6050 | 0.0595 | - |
| 0.9933 | 6100 | 0.0001 | - |
| 1.0 | 6141 | - | 0.2767 |
| 1.0015 | 6150 | 0.0 | - |
| 1.0096 | 6200 | 0.0 | - |
| 1.0177 | 6250 | 0.0 | - |
| 1.0259 | 6300 | 0.0 | - |
| 1.0340 | 6350 | 0.0 | - |
| 1.0422 | 6400 | 0.0 | - |
| 1.0503 | 6450 | 0.0 | - |
| 1.0585 | 6500 | 0.0 | - |
| 1.0666 | 6550 | 0.0 | - |
| 1.0747 | 6600 | 0.0 | - |
| 1.0829 | 6650 | 0.0 | - |
| 1.0910 | 6700 | 0.0 | - |
| 1.0992 | 6750 | 0.0 | - |
| 1.1073 | 6800 | 0.0 | - |
| 1.1155 | 6850 | 0.0 | - |
| 1.1236 | 6900 | 0.0 | - |
| 1.1317 | 6950 | 0.0 | - |
| 1.1399 | 7000 | 0.0 | - |
| 1.1480 | 7050 | 0.0 | - |
| 1.1562 | 7100 | 0.0 | - |
| 1.1643 | 7150 | 0.0 | - |
| 1.1724 | 7200 | 0.0 | - |
| 1.1806 | 7250 | 0.0 | - |
| 1.1887 | 7300 | 0.0 | - |
| 1.1969 | 7350 | 0.0 | - |
| 1.2050 | 7400 | 0.0 | - |
| 1.2132 | 7450 | 0.0 | - |
| 1.2213 | 7500 | 0.0001 | - |
| 1.2294 | 7550 | 0.0 | - |
| 1.2376 | 7600 | 0.0 | - |
| 1.2457 | 7650 | 0.0 | - |
| 1.2539 | 7700 | 0.0 | - |
| 1.2620 | 7750 | 0.0 | - |
| 1.2702 | 7800 | 0.0001 | - |
| 1.2783 | 7850 | 0.0 | - |
| 1.2864 | 7900 | 0.0 | - |
| 1.2946 | 7950 | 0.0002 | - |
| 1.3027 | 8000 | 0.0 | - |
| 1.3109 | 8050 | 0.0003 | - |
| 1.3190 | 8100 | 0.0588 | - |
| 1.3271 | 8150 | 0.0 | - |
| 1.3353 | 8200 | 0.0002 | - |
| 1.3434 | 8250 | 0.0 | - |
| 1.3516 | 8300 | 0.0 | - |
| 1.3597 | 8350 | 0.0 | - |
| 1.3679 | 8400 | 0.0261 | - |
| 1.3760 | 8450 | 0.0 | - |
| 1.3841 | 8500 | 0.0 | - |
| 1.3923 | 8550 | 0.0 | - |
| 1.4004 | 8600 | 0.0 | - |
| 1.4086 | 8650 | 0.0 | - |
| 1.4167 | 8700 | 0.0 | - |
| 1.4248 | 8750 | 0.0 | - |
| 1.4330 | 8800 | 0.0 | - |
| 1.4411 | 8850 | 0.0 | - |
| 1.4493 | 8900 | 0.0 | - |
| 1.4574 | 8950 | 0.0 | - |
| 1.4656 | 9000 | 0.0 | - |
| 1.4737 | 9050 | 0.0 | - |
| 1.4818 | 9100 | 0.0 | - |
| 1.4900 | 9150 | 0.0153 | - |
| 1.4981 | 9200 | 0.0 | - |
| 1.5063 | 9250 | 0.0 | - |
| 1.5144 | 9300 | 0.0 | - |
| 1.5226 | 9350 | 0.0 | - |
| 1.5307 | 9400 | 0.0 | - |
| 1.5388 | 9450 | 0.0003 | - |
| 1.5470 | 9500 | 0.0 | - |
| 1.5551 | 9550 | 0.0003 | - |
| 1.5633 | 9600 | 0.0 | - |
| 1.5714 | 9650 | 0.0 | - |
| 1.5795 | 9700 | 0.0 | - |
| 1.5877 | 9750 | 0.0 | - |
| 1.5958 | 9800 | 0.0 | - |
| 1.6040 | 9850 | 0.0 | - |
| 1.6121 | 9900 | 0.0 | - |
| 1.6203 | 9950 | 0.0 | - |
| 1.6284 | 10000 | 0.0 | - |
| 1.6365 | 10050 | 0.0 | - |
| 1.6447 | 10100 | 0.0 | - |
| 1.6528 | 10150 | 0.0 | - |
| 1.6610 | 10200 | 0.0 | - |
| 1.6691 | 10250 | 0.0 | - |
| 1.6773 | 10300 | 0.0 | - |
| 1.6854 | 10350 | 0.0 | - |
| 1.6935 | 10400 | 0.0 | - |
| 1.7017 | 10450 | 0.0 | - |
| 1.7098 | 10500 | 0.0 | - |
| 1.7180 | 10550 | 0.0 | - |
| 1.7261 | 10600 | 0.0 | - |
| 1.7342 | 10650 | 0.0 | - |
| 1.7424 | 10700 | 0.0 | - |
| 1.7505 | 10750 | 0.0 | - |
| 1.7587 | 10800 | 0.0 | - |
| 1.7668 | 10850 | 0.0 | - |
| 1.7750 | 10900 | 0.0 | - |
| 1.7831 | 10950 | 0.0 | - |
| 1.7912 | 11000 | 0.0 | - |
| 1.7994 | 11050 | 0.0 | - |
| 1.8075 | 11100 | 0.0 | - |
| 1.8157 | 11150 | 0.0 | - |
| 1.8238 | 11200 | 0.0 | - |
| 1.8319 | 11250 | 0.0 | - |
| 1.8401 | 11300 | 0.0 | - |
| 1.8482 | 11350 | 0.0 | - |
| 1.8564 | 11400 | 0.0 | - |
| 1.8645 | 11450 | 0.0 | - |
| 1.8727 | 11500 | 0.0 | - |
| 1.8808 | 11550 | 0.0 | - |
| 1.8889 | 11600 | 0.0 | - |
| 1.8971 | 11650 | 0.0 | - |
| 1.9052 | 11700 | 0.0 | - |
| 1.9134 | 11750 | 0.0 | - |
| 1.9215 | 11800 | 0.0 | - |
| 1.9297 | 11850 | 0.0 | - |
| 1.9378 | 11900 | 0.0006 | - |
| 1.9459 | 11950 | 0.0 | - |
| 1.9541 | 12000 | 0.0 | - |
| 1.9622 | 12050 | 0.0 | - |
| 1.9704 | 12100 | 0.0 | - |
| 1.9785 | 12150 | 0.0 | - |
| 1.9866 | 12200 | 0.0 | - |
| 1.9948 | 12250 | 0.0 | - |
| 2.0 | 12282 | - | 0.2742 |
| 2.0029 | 12300 | 0.0 | - |
| 2.0111 | 12350 | 0.0 | - |
| 2.0192 | 12400 | 0.0 | - |
| 2.0274 | 12450 | 0.0 | - |
| 2.0355 | 12500 | 0.0 | - |
| 2.0436 | 12550 | 0.0 | - |
| 2.0518 | 12600 | 0.0 | - |
| 2.0599 | 12650 | 0.0 | - |
| 2.0681 | 12700 | 0.0 | - |
| 2.0762 | 12750 | 0.0 | - |
| 2.0844 | 12800 | 0.0 | - |
| 2.0925 | 12850 | 0.0 | - |
| 2.1006 | 12900 | 0.0 | - |
| 2.1088 | 12950 | 0.0 | - |
| 2.1169 | 13000 | 0.0 | - |
| 2.1251 | 13050 | 0.0 | - |
| 2.1332 | 13100 | 0.0 | - |
| 2.1413 | 13150 | 0.0 | - |
| 2.1495 | 13200 | 0.0 | - |
| 2.1576 | 13250 | 0.0 | - |
| 2.1658 | 13300 | 0.0 | - |
| 2.1739 | 13350 | 0.0 | - |
| 2.1821 | 13400 | 0.0 | - |
| 2.1902 | 13450 | 0.0 | - |
| 2.1983 | 13500 | 0.0 | - |
| 2.2065 | 13550 | 0.0 | - |
| 2.2146 | 13600 | 0.0 | - |
| 2.2228 | 13650 | 0.0 | - |
| 2.2309 | 13700 | 0.0 | - |
| 2.2390 | 13750 | 0.0 | - |
| 2.2472 | 13800 | 0.0 | - |
| 2.2553 | 13850 | 0.0 | - |
| 2.2635 | 13900 | 0.0 | - |
| 2.2716 | 13950 | 0.0 | - |
| 2.2798 | 14000 | 0.0 | - |
| 2.2879 | 14050 | 0.0013 | - |
| 2.2960 | 14100 | 0.0 | - |
| 2.3042 | 14150 | 0.0 | - |
| 2.3123 | 14200 | 0.0 | - |
| 2.3205 | 14250 | 0.0 | - |
| 2.3286 | 14300 | 0.0 | - |
| 2.3368 | 14350 | 0.0 | - |
| 2.3449 | 14400 | 0.0 | - |
| 2.3530 | 14450 | 0.0019 | - |
| 2.3612 | 14500 | 0.0 | - |
| 2.3693 | 14550 | 0.0 | - |
| 2.3775 | 14600 | 0.0 | - |
| 2.3856 | 14650 | 0.0 | - |
| 2.3937 | 14700 | 0.0 | - |
| 2.4019 | 14750 | 0.0 | - |
| 2.4100 | 14800 | 0.0 | - |
| 2.4182 | 14850 | 0.0 | - |
| 2.4263 | 14900 | 0.0 | - |
| 2.4345 | 14950 | 0.0 | - |
| 2.4426 | 15000 | 0.0 | - |
| 2.4507 | 15050 | 0.0 | - |
| 2.4589 | 15100 | 0.0 | - |
| 2.4670 | 15150 | 0.0 | - |
| 2.4752 | 15200 | 0.0 | - |
| 2.4833 | 15250 | 0.0 | - |
| 2.4915 | 15300 | 0.0 | - |
| 2.4996 | 15350 | 0.0 | - |
| 2.5077 | 15400 | 0.0 | - |
| 2.5159 | 15450 | 0.0 | - |
| 2.5240 | 15500 | 0.0 | - |
| 2.5322 | 15550 | 0.0 | - |
| 2.5403 | 15600 | 0.0 | - |
| 2.5484 | 15650 | 0.0 | - |
| 2.5566 | 15700 | 0.0 | - |
| 2.5647 | 15750 | 0.0 | - |
| 2.5729 | 15800 | 0.0 | - |
| 2.5810 | 15850 | 0.0 | - |
| 2.5892 | 15900 | 0.0001 | - |
| 2.5973 | 15950 | 0.0 | - |
| 2.6054 | 16000 | 0.0 | - |
| 2.6136 | 16050 | 0.0 | - |
| 2.6217 | 16100 | 0.0 | - |
| 2.6299 | 16150 | 0.0 | - |
| 2.6380 | 16200 | 0.0 | - |
| 2.6461 | 16250 | 0.0 | - |
| 2.6543 | 16300 | 0.0 | - |
| 2.6624 | 16350 | 0.0 | - |
| 2.6706 | 16400 | 0.0 | - |
| 2.6787 | 16450 | 0.0 | - |
| 2.6869 | 16500 | 0.0 | - |
| 2.6950 | 16550 | 0.0 | - |
| 2.7031 | 16600 | 0.0 | - |
| 2.7113 | 16650 | 0.0002 | - |
| 2.7194 | 16700 | 0.0 | - |
| 2.7276 | 16750 | 0.0 | - |
| 2.7357 | 16800 | 0.0 | - |
| 2.7439 | 16850 | 0.0 | - |
| 2.7520 | 16900 | 0.0 | - |
| 2.7601 | 16950 | 0.0 | - |
| 2.7683 | 17000 | 0.0291 | - |
| 2.7764 | 17050 | 0.0 | - |
| 2.7846 | 17100 | 0.0 | - |
| 2.7927 | 17150 | 0.0 | - |
| 2.8008 | 17200 | 0.0 | - |
| 2.8090 | 17250 | 0.0 | - |
| 2.8171 | 17300 | 0.0 | - |
| 2.8253 | 17350 | 0.0 | - |
| 2.8334 | 17400 | 0.0 | - |
| 2.8416 | 17450 | 0.0 | - |
| 2.8497 | 17500 | 0.0 | - |
| 2.8578 | 17550 | 0.0 | - |
| 2.8660 | 17600 | 0.0 | - |
| 2.8741 | 17650 | 0.0 | - |
| 2.8823 | 17700 | 0.0 | - |
| 2.8904 | 17750 | 0.0 | - |
| 2.8986 | 17800 | 0.0 | - |
| 2.9067 | 17850 | 0.0 | - |
| 2.9148 | 17900 | 0.0 | - |
| 2.9230 | 17950 | 0.0 | - |
| 2.9311 | 18000 | 0.0 | - |
| 2.9393 | 18050 | 0.0 | - |
| 2.9474 | 18100 | 0.0 | - |
| 2.9555 | 18150 | 0.0 | - |
| 2.9637 | 18200 | 0.0 | - |
| 2.9718 | 18250 | 0.0 | - |
| 2.9800 | 18300 | 0.0 | - |
| 2.9881 | 18350 | 0.0 | - |
| 2.9963 | 18400 | 0.0 | - |
| **3.0** | **18423** | **-** | **0.2642** |
| 3.0044 | 18450 | 0.0012 | - |
| 3.0125 | 18500 | 0.0 | - |
| 3.0207 | 18550 | 0.0 | - |
| 3.0288 | 18600 | 0.0 | - |
| 3.0370 | 18650 | 0.0 | - |
| 3.0451 | 18700 | 0.0041 | - |
| 3.0532 | 18750 | 0.0 | - |
| 3.0614 | 18800 | 0.0 | - |
| 3.0695 | 18850 | 0.0 | - |
| 3.0777 | 18900 | 0.0 | - |
| 3.0858 | 18950 | 0.0 | - |
| 3.0940 | 19000 | 0.0 | - |
| 3.1021 | 19050 | 0.0 | - |
| 3.1102 | 19100 | 0.0 | - |
| 3.1184 | 19150 | 0.0 | - |
| 3.1265 | 19200 | 0.0 | - |
| 3.1347 | 19250 | 0.0 | - |
| 3.1428 | 19300 | 0.0 | - |
| 3.1510 | 19350 | 0.0 | - |
| 3.1591 | 19400 | 0.0 | - |
| 3.1672 | 19450 | 0.0 | - |
| 3.1754 | 19500 | 0.0014 | - |
| 3.1835 | 19550 | 0.0 | - |
| 3.1917 | 19600 | 0.0 | - |
| 3.1998 | 19650 | 0.0 | - |
| 3.2079 | 19700 | 0.0 | - |
| 3.2161 | 19750 | 0.0 | - |
| 3.2242 | 19800 | 0.0 | - |
| 3.2324 | 19850 | 0.0 | - |
| 3.2405 | 19900 | 0.0 | - |
| 3.2487 | 19950 | 0.0 | - |
| 3.2568 | 20000 | 0.0 | - |
| 3.2649 | 20050 | 0.0 | - |
| 3.2731 | 20100 | 0.0 | - |
| 3.2812 | 20150 | 0.0 | - |
| 3.2894 | 20200 | 0.0453 | - |
| 3.2975 | 20250 | 0.0 | - |
| 3.3057 | 20300 | 0.0 | - |
| 3.3138 | 20350 | 0.0 | - |
| 3.3219 | 20400 | 0.0 | - |
| 3.3301 | 20450 | 0.0 | - |
| 3.3382 | 20500 | 0.0 | - |
| 3.3464 | 20550 | 0.0 | - |
| 3.3545 | 20600 | 0.0 | - |
| 3.3626 | 20650 | 0.0 | - |
| 3.3708 | 20700 | 0.0 | - |
| 3.3789 | 20750 | 0.0 | - |
| 3.3871 | 20800 | 0.0 | - |
| 3.3952 | 20850 | 0.0 | - |
| 3.4034 | 20900 | 0.0 | - |
| 3.4115 | 20950 | 0.0 | - |
| 3.4196 | 21000 | 0.0 | - |
| 3.4278 | 21050 | 0.0 | - |
| 3.4359 | 21100 | 0.0 | - |
| 3.4441 | 21150 | 0.0 | - |
| 3.4522 | 21200 | 0.0 | - |
| 3.4603 | 21250 | 0.0 | - |
| 3.4685 | 21300 | 0.0 | - |
| 3.4766 | 21350 | 0.0 | - |
| 3.4848 | 21400 | 0.0 | - |
| 3.4929 | 21450 | 0.0 | - |
| 3.5011 | 21500 | 0.0 | - |
| 3.5092 | 21550 | 0.0 | - |
| 3.5173 | 21600 | 0.0 | - |
| 3.5255 | 21650 | 0.0 | - |
| 3.5336 | 21700 | 0.0 | - |
| 3.5418 | 21750 | 0.0 | - |
| 3.5499 | 21800 | 0.0 | - |
| 3.5581 | 21850 | 0.0 | - |
| 3.5662 | 21900 | 0.0 | - |
| 3.5743 | 21950 | 0.0 | - |
| 3.5825 | 22000 | 0.0 | - |
| 3.5906 | 22050 | 0.0 | - |
| 3.5988 | 22100 | 0.0 | - |
| 3.6069 | 22150 | 0.0 | - |
| 3.6150 | 22200 | 0.0 | - |
| 3.6232 | 22250 | 0.0 | - |
| 3.6313 | 22300 | 0.0 | - |
| 3.6395 | 22350 | 0.0 | - |
| 3.6476 | 22400 | 0.0 | - |
| 3.6558 | 22450 | 0.0 | - |
| 3.6639 | 22500 | 0.0 | - |
| 3.6720 | 22550 | 0.0 | - |
| 3.6802 | 22600 | 0.0 | - |
| 3.6883 | 22650 | 0.0 | - |
| 3.6965 | 22700 | 0.0 | - |
| 3.7046 | 22750 | 0.0 | - |
| 3.7128 | 22800 | 0.0 | - |
| 3.7209 | 22850 | 0.0 | - |
| 3.7290 | 22900 | 0.0 | - |
| 3.7372 | 22950 | 0.0 | - |
| 3.7453 | 23000 | 0.0 | - |
| 3.7535 | 23050 | 0.0 | - |
| 3.7616 | 23100 | 0.0 | - |
| 3.7697 | 23150 | 0.0 | - |
| 3.7779 | 23200 | 0.0 | - |
| 3.7860 | 23250 | 0.0 | - |
| 3.7942 | 23300 | 0.0 | - |
| 3.8023 | 23350 | 0.0 | - |
| 3.8105 | 23400 | 0.0 | - |
| 3.8186 | 23450 | 0.0 | - |
| 3.8267 | 23500 | 0.0 | - |
| 3.8349 | 23550 | 0.0 | - |
| 3.8430 | 23600 | 0.0 | - |
| 3.8512 | 23650 | 0.0 | - |
| 3.8593 | 23700 | 0.0 | - |
| 3.8674 | 23750 | 0.0 | - |
| 3.8756 | 23800 | 0.0 | - |
| 3.8837 | 23850 | 0.0 | - |
| 3.8919 | 23900 | 0.0 | - |
| 3.9000 | 23950 | 0.0 | - |
| 3.9082 | 24000 | 0.0 | - |
| 3.9163 | 24050 | 0.0 | - |
| 3.9244 | 24100 | 0.0 | - |
| 3.9326 | 24150 | 0.0 | - |
| 3.9407 | 24200 | 0.0 | - |
| 3.9489 | 24250 | 0.0 | - |
| 3.9570 | 24300 | 0.0 | - |
| 3.9652 | 24350 | 0.0 | - |
| 3.9733 | 24400 | 0.0 | - |
| 3.9814 | 24450 | 0.0 | - |
| 3.9896 | 24500 | 0.0 | - |
| 3.9977 | 24550 | 0.0 | - |
| 4.0 | 24564 | - | 0.2671 |
* The bold row denotes the saved checkpoint.
### Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.7.0
- Transformers: 4.40.2
- PyTorch: 2.2.1+cu121
- Datasets: 2.19.1
- Tokenizers: 0.19.1
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"base_model": "sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2", "library_name": "setfit", "metrics": ["accuracy"], "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification", "generated_from_setfit_trainer"], "widget": [{"text": "Não apenas isso. A bola de neve do endividamento"}, {"text": " Bueno, yo lo que espero es que se traten con respeto, que se quieran. "}, {"text": " SÃ, pues pedirle a MarÃa Luisa que le dé seguimiento y que siga atendiendo las demandas de los ciudadanos de Vallarta, si te parece. Ya ella seguramente nos está viendo y está tomando nota para darle continuidad a las demandas de ambientalistas de Vallarta. "}, {"text": "A confiança na economia despertou o apetite pelo risco, criando instrumentos financeiros indispensáveis à captação de novos recursos para a expansão produtiva."}, {"text": " A ver, pon la carta de Elba Esther. Es que luego la borró. Fue en mayo del 23, 2 de mayo: ‘Ahà le espero con el Ejército —supuestamente esto es lo que le dijo Calderón a la maestra Elba Esther, ahà la espero con el Ejército— esa fue la respuesta del entonces presidente de México, Felipe Calderón, cuando le dije —según la maestra— que las y los maestros de México nos oponÃamos a que Miguel Ãngel Yunes continuara como titular del Issste, dadas las malversaciones de fondos financieros que con tanto trabajo las los trabajadores al servicio del Estado logramos con la reforma a dicha institución. ‘Cuando me comentó que Yunes estaba haciendo bien su trabajo, no me dejó más alternativa —dice la maestra— que advertirle que tomarÃamos las instalaciones del Issste y justo esa fue su respuesta: Ahà la espero con el Ejército. Esto sucedió en el marco de un evento público en una escuela secundaria técnica de la ahora Ciudad de México. Ante su respuesta, me levanté y me retiré. ‘Recordemos que la elección y remoción del director del Issste compete única y exclusivamente al titular del Ejecutivo federal y no a una servidora.’ Aquà me está contestando a mÃ, porque yo dije que a ella le habÃan entregado por ayudar en el fraude, que no me dirÃa la maestra que no ayudó en el fraude del 2006, y a cambio yo sostengo que le entregaron el Issste, la SubsecretarÃa de Educación Pública y la LoterÃa Nacional. ‘Por ello, en relación a las declaraciones hechas por el presidente Andrés Manuel López Obrador el pasado 29 de abril del presente año, sobre mi persona y la gestión del señor Miguel Ãngel Yunes al frente del Issste, le digo categóricamente que no participé el acto ilÃcito alguno, como me acusa desde su tribuna’. Yo no estoy acusando más que de haberse aliado con Calderón y ayudarle en el fraude electoral. ‘Siempre me he conducido conforme a derecho, de respeto a las instituciones de este paÃs y, desde luego, a la investidura presidencial. Por ello, señor presidente, basta de falsas acusaciones a mi persona’. No es nada personal, maestra, es que estamos viviendo un momento importantÃsimo de transformación. Entonces, como el compañero que viene a hacernos preguntas sobre salud, ayuda a recordar, porque es como si padecieran amnesia, ya se olvidó cómo era. Y antes esto no lo tocaban, era silencio, como vasallos, obedecer y callar, siempre y cuando hubiese dinero de por medio, porque lo que no suena lógico suena metálico. Entonces, hay que ir aclarando todo, seguir purificando la vida pública del paÃs y por eso son muy buenas estas mañaneras. Pero, bueno, eso es lo que querÃamos decir. ¿Qué se está haciendo? Procurar, ya es un compromiso, garantizar el derecho a la salud. Y vaya que ha costado, por estos intereses. ImagÃnense, no se podÃan comprar medicinas en el extranjero porque la ley lo prohibÃa, lo impedÃa; tuvimos que reformar la ley. ¿Y quiénes votaron en contra de que se pudiera comprar la medicina en el extranjero? El bloque conservador. ¿Qué son entonces? Representantes de minorÃas, no representantes del pueblo, esa es nuestra diferencia de fondo. No es nada personal, pero sà es importante el darle su sitio que le corresponde a lo público. República es, res publica, cosa pública. Si vivimos en una república, tenemos que pensar en eso, en lo público. Eso ya se habÃa olvidado. Entonces, vamos a continuar con lo mismo y va adelante todo el plan de transformación. El viernes vamos a informar sobre salud y luego vamos a informar en especÃfico sobre el Issste, porque ya llevamos… ¿Cuánto tiempo llevamos? "}], "inference": true, "model-index": [{"name": "SetFit with sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "Unknown", "type": "unknown", "split": "test"}, "metrics": [{"type": "accuracy", "value": 0.7727272727272727, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,917 |
Helsinki-NLP/opus-mt-efi-fr
|
Helsinki-NLP
|
translation
|
[
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"efi",
"fr",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2022-03-02T23:29:04Z |
2023-08-16T11:28:44+00:00
| 57 | 0 |
---
license: apache-2.0
tags:
- translation
---
### opus-mt-efi-fr
* source languages: efi
* target languages: fr
* OPUS readme: [efi-fr](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/efi-fr/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-20.zip](https://object.pouta.csc.fi/OPUS-MT-models/efi-fr/opus-2020-01-20.zip)
* test set translations: [opus-2020-01-20.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/efi-fr/opus-2020-01-20.test.txt)
* test set scores: [opus-2020-01-20.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/efi-fr/opus-2020-01-20.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| JW300.efi.fr | 25.1 | 0.419 |
| null |
Non_BioNLP
|
### opus-mt-efi-fr
* source languages: efi
* target languages: fr
* OPUS readme: [efi-fr](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/efi-fr/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2020-01-20.zip](https://object.pouta.csc.fi/OPUS-MT-models/efi-fr/opus-2020-01-20.zip)
* test set translations: [opus-2020-01-20.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/efi-fr/opus-2020-01-20.test.txt)
* test set scores: [opus-2020-01-20.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/efi-fr/opus-2020-01-20.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| JW300.efi.fr | 25.1 | 0.419 |
|
{"license": "apache-2.0", "tags": ["translation"]}
|
task
|
[
"TRANSLATION"
] | 39,918 |
PulseWave/PAYMENT-TERMS
|
PulseWave
|
text-classification
|
[
"setfit",
"safetensors",
"mpnet",
"sentence-transformers",
"text-classification",
"generated_from_setfit_trainer",
"arxiv:2209.11055",
"region:us"
] | 2024-03-01T18:59:45Z |
2024-03-01T19:02:33+00:00
| 8 | 0 |
---
library_name: setfit
metrics:
- accuracy
pipeline_tag: text-classification
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
widget: []
inference: true
---
# SetFit
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
<!-- - **Sentence Transformer:** [Unknown](https://huggingface.co/unknown) -->
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 512 tokens
- **Number of Classes:** 2 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("setfit_model_id")
# Run inference
preds = model("I loved the spiderman movie!")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Framework Versions
- Python: 3.11.7
- SetFit: 1.0.3
- Sentence Transformers: 2.3.1
- Transformers: 4.37.2
- PyTorch: 2.2.0
- Datasets: 2.16.1
- Tokenizers: 0.15.1
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
| null |
Non_BioNLP
|
# SetFit
This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification.
The model has been trained using an efficient few-shot learning technique that involves:
1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.
## Model Details
### Model Description
- **Model Type:** SetFit
<!-- - **Sentence Transformer:** [Unknown](https://huggingface.co/unknown) -->
- **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance
- **Maximum Sequence Length:** 512 tokens
- **Number of Classes:** 2 classes
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit)
- **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
- **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)
## Uses
### Direct Use for Inference
First install the SetFit library:
```bash
pip install setfit
```
Then you can load this model and run inference.
```python
from setfit import SetFitModel
# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("setfit_model_id")
# Run inference
preds = model("I loved the spiderman movie!")
```
<!--
### Downstream Use
*List how someone could finetune this model on their own dataset.*
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Framework Versions
- Python: 3.11.7
- SetFit: 1.0.3
- Sentence Transformers: 2.3.1
- Transformers: 4.37.2
- PyTorch: 2.2.0
- Datasets: 2.16.1
- Tokenizers: 0.15.1
## Citation
### BibTeX
```bibtex
@article{https://doi.org/10.48550/arxiv.2209.11055,
doi = {10.48550/ARXIV.2209.11055},
url = {https://arxiv.org/abs/2209.11055},
author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Efficient Few-Shot Learning Without Prompts},
publisher = {arXiv},
year = {2022},
copyright = {Creative Commons Attribution 4.0 International}
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->
|
{"library_name": "setfit", "metrics": ["accuracy"], "pipeline_tag": "text-classification", "tags": ["setfit", "sentence-transformers", "text-classification", "generated_from_setfit_trainer"], "widget": [], "inference": true}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,919 |
BAAI/Emu2
|
BAAI
|
text-generation
|
[
"transformers",
"pytorch",
"text-generation",
"custom_code",
"en",
"arxiv:2312.13286",
"autotrain_compatible",
"region:us"
] | 2023-12-19T13:47:42Z |
2023-12-21T12:30:50+00:00
| 301 | 88 |
---
language:
- en
---
<div align='center'>
<h1>Generative Multimodal Models are In-Context Learners</h1h1>
<h3><a href="">Generative Multimodal Models are In-Context Learners</a></h3>
[Quan Sun](https://github.com/Quan-Sun)<sup>1*</sup>, [Yufeng Cui](https://scholar.google.com/citations?hl=en&user=5Ydha2EAAAAJ)<sup>1*</sup>, [Xiaosong Zhang](https://zhangxiaosong18.github.io)<sup>1*</sup>, [Fan Zhang](https://scholar.google.com/citations?user=VsJ39HMAAAAJ)<sup>1*</sup>, [Qiying Yu](https://yqy2001.github.io)<sup>2,1*</sup>, [Zhengxiong Luo](https://greatlog.github.io)<sup>1</sup>, [Yueze Wang]()<sup>1</sup>, [Yongming Rao](https://raoyongming.github.io)<sup>1</sup>,<br>[Jingjing Liu](https://air.tsinghua.edu.cn/en/info/1046/1194.htm)<sup>2</sup>, [Tiejun Huang](https://scholar.google.com/citations?user=knvEK4AAAAAJ&hl=en)<sup>1,3</sup>, [Xinlong Wang](https://www.xloong.wang/)<sup>1†</sup>
<sup>1</sup> [BAAI](https://www.baai.ac.cn/english.html), <sup>2</sup> [THU](https://air.tsinghua.edu.cn), <sup>3</sup> [PKU](https://english.pku.edu.cn/) <br><sup>*</sup> equal contribution <sup>†</sup> project lead
| [Paper](https://arxiv.org/abs/2312.13286) | [🤗HF Demo](https://huggingface.co/spaces/BAAI/Emu2) | [Demo](https://emu.ssi.plus) | [Project Page](https://baaivision.github.io/emu2/) | [Github](https://github.com/baaivision/Emu)
</div>
The human ability to easily solve multimodal tasks in context (i.e., with only a few demonstrations or simple instructions), is what current multimodal systems have largely struggled to imitate.
In this work, we demonstrate that the task-agnostic in-context learning capabilities of large multimodal models can be significantly enhanced by effective scaling-up.
We introduce **Emu2**, a generative multimodal model with 37 billion parameters, trained on large-scale multimodal sequences with a unified autoregressive objective.
**Emu2** exhibits strong multimodal in-context learning abilities, even emerging to solve tasks that require on-the-fly reasoning, such as visual prompting and object-grounded generation.
The model sets a new record on multiple multimodal understanding tasks in few-shot settings.
When instruction-tuned to follow specific instructions, **Emu2** further achieves new state-of-the-art on challenging tasks such as question answering benchmarks for large multimodal models and open-ended subject-driven generation.
These achievements demonstrate that **Emu2** can serve as a base model and general-purpose interface for a wide range of multimodal tasks.
Code and models are publicly available to facilitate future research.
## Model Weights
| Model name | Weight |
| ------------------ | ------------------------------------------------------- |
| **Emu2** | [🤗 HF link](https://huggingface.co/BAAI/Emu2) |
| **Emu2-Chat** | [🤗 HF link](https://huggingface.co/BAAI/Emu2-Chat) |
| **Emu2-Gen** | [🤗 HF link](https://huggingface.co/BAAI/Emu2-Gen) |
## Inference (Huggingface Version)
#### Single GPU
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).to('cuda').eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = '[<IMG_PLH>]Describe the image in details:'
image = Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB')
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=[image]
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
Interleaved image and text
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).to('cuda').eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = "[<IMG_PLH>][red, white, 3, bottom left].[<IMG_PLH>][yellow, white, 2, top left].[<IMG_PLH>][green, black, 4, bottom right][<IMG_PLH>]"
images = [
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/red_white_3_bottom_left.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/yellow_white_2_top_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/green_black_4_bottom_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB'),
]
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=images
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
#### Multi GPU
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from accelerate import init_empty_weights, infer_auto_device_map, load_checkpoint_and_dispatch
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
with init_empty_weights():
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True)
device_map = infer_auto_device_map(model, max_memory={0:'38GiB',1:'38GiB',}, no_split_module_classes=['Block','LlamaDecoderLayer'])
# input and output logits should be on same device
device_map["model.decoder.lm.lm_head"] = 0
model = load_checkpoint_and_dispatch(
model,
'local/path/to/hf/version/Emu2/model',
device_map=device_map).eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = '[<IMG_PLH>]Describe the image in details:'
image = Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB')
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=[image]
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
Interleaved image and text
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from accelerate import init_empty_weights, infer_auto_device_map, load_checkpoint_and_dispatch
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
with init_empty_weights():
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True)
device_map = infer_auto_device_map(model, max_memory={0:'38GiB',1:'38GiB',}, no_split_module_classes=['Block','LlamaDecoderLayer'])
# input and output logits should be on same device
device_map["model.decoder.lm.lm_head"] = 0
model = load_checkpoint_and_dispatch(
model,
'local/path/to/hf/version/Emu2/model',
device_map=device_map).eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = "[<IMG_PLH>][red, white, 3, bottom left].[<IMG_PLH>][yellow, white, 2, top left].[<IMG_PLH>][green, black, 4, bottom right][<IMG_PLH>]"
images = [
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/red_white_3_bottom_left.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/yellow_white_2_top_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/green_black_4_bottom_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB'),
]
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=images
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
#### Quantization
Check quantization guidance at [transformers](https://huggingface.co/docs/transformers/v4.28.0/main_classes/quantization)
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
load_in_4bit=True,
trust_remote_code=True,
bnb_4bit_compute_dtype=torch.float16).eval()
query = '[<IMG_PLH>]Describe the image in details:'
image = Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB')
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=[image]
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.float16), # should be torch.float16
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
## Citation
If you find Emu2 useful for your research and applications, please consider starring this repository and citing:
```
@article{Emu2,
title={Generative Multimodal Models are In-Context Learners},
author={Quan Sun and Yufeng Cui and Xiaosong Zhang and Fan Zhang and Qiying Yu and Zhengxiong Luo and Yueze Wang and Yongming Rao and Jingjing Liu and Tiejun Huang and Xinlong Wang},
publisher={arXiv preprint arXiv:2312.13286},
year={2023},
}
```
| null |
Non_BioNLP
|
<div align='center'>
<h1>Generative Multimodal Models are In-Context Learners</h1h1>
<h3><a href="">Generative Multimodal Models are In-Context Learners</a></h3>
[Quan Sun](https://github.com/Quan-Sun)<sup>1*</sup>, [Yufeng Cui](https://scholar.google.com/citations?hl=en&user=5Ydha2EAAAAJ)<sup>1*</sup>, [Xiaosong Zhang](https://zhangxiaosong18.github.io)<sup>1*</sup>, [Fan Zhang](https://scholar.google.com/citations?user=VsJ39HMAAAAJ)<sup>1*</sup>, [Qiying Yu](https://yqy2001.github.io)<sup>2,1*</sup>, [Zhengxiong Luo](https://greatlog.github.io)<sup>1</sup>, [Yueze Wang]()<sup>1</sup>, [Yongming Rao](https://raoyongming.github.io)<sup>1</sup>,<br>[Jingjing Liu](https://air.tsinghua.edu.cn/en/info/1046/1194.htm)<sup>2</sup>, [Tiejun Huang](https://scholar.google.com/citations?user=knvEK4AAAAAJ&hl=en)<sup>1,3</sup>, [Xinlong Wang](https://www.xloong.wang/)<sup>1†</sup>
<sup>1</sup> [BAAI](https://www.baai.ac.cn/english.html), <sup>2</sup> [THU](https://air.tsinghua.edu.cn), <sup>3</sup> [PKU](https://english.pku.edu.cn/) <br><sup>*</sup> equal contribution <sup>†</sup> project lead
| [Paper](https://arxiv.org/abs/2312.13286) | [🤗HF Demo](https://huggingface.co/spaces/BAAI/Emu2) | [Demo](https://emu.ssi.plus) | [Project Page](https://baaivision.github.io/emu2/) | [Github](https://github.com/baaivision/Emu)
</div>
The human ability to easily solve multimodal tasks in context (i.e., with only a few demonstrations or simple instructions), is what current multimodal systems have largely struggled to imitate.
In this work, we demonstrate that the task-agnostic in-context learning capabilities of large multimodal models can be significantly enhanced by effective scaling-up.
We introduce **Emu2**, a generative multimodal model with 37 billion parameters, trained on large-scale multimodal sequences with a unified autoregressive objective.
**Emu2** exhibits strong multimodal in-context learning abilities, even emerging to solve tasks that require on-the-fly reasoning, such as visual prompting and object-grounded generation.
The model sets a new record on multiple multimodal understanding tasks in few-shot settings.
When instruction-tuned to follow specific instructions, **Emu2** further achieves new state-of-the-art on challenging tasks such as question answering benchmarks for large multimodal models and open-ended subject-driven generation.
These achievements demonstrate that **Emu2** can serve as a base model and general-purpose interface for a wide range of multimodal tasks.
Code and models are publicly available to facilitate future research.
## Model Weights
| Model name | Weight |
| ------------------ | ------------------------------------------------------- |
| **Emu2** | [🤗 HF link](https://huggingface.co/BAAI/Emu2) |
| **Emu2-Chat** | [🤗 HF link](https://huggingface.co/BAAI/Emu2-Chat) |
| **Emu2-Gen** | [🤗 HF link](https://huggingface.co/BAAI/Emu2-Gen) |
## Inference (Huggingface Version)
#### Single GPU
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).to('cuda').eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = '[<IMG_PLH>]Describe the image in details:'
image = Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB')
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=[image]
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
Interleaved image and text
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).to('cuda').eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = "[<IMG_PLH>][red, white, 3, bottom left].[<IMG_PLH>][yellow, white, 2, top left].[<IMG_PLH>][green, black, 4, bottom right][<IMG_PLH>]"
images = [
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/red_white_3_bottom_left.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/yellow_white_2_top_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/green_black_4_bottom_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB'),
]
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=images
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
#### Multi GPU
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from accelerate import init_empty_weights, infer_auto_device_map, load_checkpoint_and_dispatch
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
with init_empty_weights():
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True)
device_map = infer_auto_device_map(model, max_memory={0:'38GiB',1:'38GiB',}, no_split_module_classes=['Block','LlamaDecoderLayer'])
# input and output logits should be on same device
device_map["model.decoder.lm.lm_head"] = 0
model = load_checkpoint_and_dispatch(
model,
'local/path/to/hf/version/Emu2/model',
device_map=device_map).eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = '[<IMG_PLH>]Describe the image in details:'
image = Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB')
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=[image]
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
Interleaved image and text
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from accelerate import init_empty_weights, infer_auto_device_map, load_checkpoint_and_dispatch
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
with init_empty_weights():
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True)
device_map = infer_auto_device_map(model, max_memory={0:'38GiB',1:'38GiB',}, no_split_module_classes=['Block','LlamaDecoderLayer'])
# input and output logits should be on same device
device_map["model.decoder.lm.lm_head"] = 0
model = load_checkpoint_and_dispatch(
model,
'local/path/to/hf/version/Emu2/model',
device_map=device_map).eval()
# `[<IMG_PLH>]` is the image placeholder which will be replaced by image embeddings.
# the number of `[<IMG_PLH>]` should be equal to the number of input images
query = "[<IMG_PLH>][red, white, 3, bottom left].[<IMG_PLH>][yellow, white, 2, top left].[<IMG_PLH>][green, black, 4, bottom right][<IMG_PLH>]"
images = [
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/red_white_3_bottom_left.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/yellow_white_2_top_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/green_black_4_bottom_right.jpg?raw=true',stream=True).raw).convert('RGB'),
Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB'),
]
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=images
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.bfloat16),
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
#### Quantization
Check quantization guidance at [transformers](https://huggingface.co/docs/transformers/v4.28.0/main_classes/quantization)
```python
from PIL import Image
import requests
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("BAAI/Emu2")
model = AutoModelForCausalLM.from_pretrained(
"BAAI/Emu2",
load_in_4bit=True,
trust_remote_code=True,
bnb_4bit_compute_dtype=torch.float16).eval()
query = '[<IMG_PLH>]Describe the image in details:'
image = Image.open(requests.get('https://github.com/baaivision/Emu/Emu2/examples/blue_black_1_top_left.jpg?raw=true',stream=True).raw).convert('RGB')
inputs = model.build_input_ids(
text=[query],
tokenizer=tokenizer,
image=[image]
)
with torch.no_grad():
outputs = model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
image=inputs["image"].to(torch.float16), # should be torch.float16
max_new_tokens=64,
length_penalty=-1)
output_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)
```
## Citation
If you find Emu2 useful for your research and applications, please consider starring this repository and citing:
```
@article{Emu2,
title={Generative Multimodal Models are In-Context Learners},
author={Quan Sun and Yufeng Cui and Xiaosong Zhang and Fan Zhang and Qiying Yu and Zhengxiong Luo and Yueze Wang and Yongming Rao and Jingjing Liu and Tiejun Huang and Xinlong Wang},
publisher={arXiv preprint arXiv:2312.13286},
year={2023},
}
```
|
{"language": ["en"]}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,920 |
platzi/platzi-distilroberta-base-mrpc-glue-joselier
|
platzi
|
text-classification
|
[
"transformers",
"pytorch",
"roberta",
"text-classification",
"generated_from_trainer",
"dataset:glue",
"base_model:distilbert/distilroberta-base",
"base_model:finetune:distilbert/distilroberta-base",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-08-29T00:21:54Z |
2023-08-29T02:41:23+00:00
| 13 | 0 |
---
base_model: distilroberta-base
datasets:
- glue
license: apache-2.0
metrics:
- accuracy
- f1
tags:
- text-classification
- generated_from_trainer
widget:
- text:
- Yucaipa owned Dominick 's before selling the chain to Safeway in 1998 for $ 2.5
billion.
- Yucaipa bought Dominick's in 1995 for $ 693 million and sold it to Safeway for
$ 1.8 billion in 1998.
example_title: Not Equivalent
- text:
- Revenue in the first quarter of the year dropped 15 percent from the same period
a year earlier.
- With the scandal hanging over Stewart's company revenue the first quarter of the
year dropped 15 percent from the same period a year earlier.
example_title: Equivalent
model-index:
- name: platzi-distilroberta-base-mrpc-glue-joselier
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: glue
type: glue
config: mrpc
split: validation
args: mrpc
metrics:
- type: accuracy
value: 0.821078431372549
name: Accuracy
- type: f1
value: 0.8809135399673736
name: F1
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# platzi-distilroberta-base-mrpc-glue-joselier
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the glue and the mrpc datasets.
It achieves the following results on the evaluation set:
- Loss: 0.5993
- Accuracy: 0.8211
- F1: 0.8809
## Model description
This model uses transfer learning to classify 2 sentences (a string of 2 sentences separated by a comma) in "Equivalent" or "Not Equivalent".
The model platzi-distilroberta-base-mrpc-glue-joselier was programmed as part of a class from Platzi's course ["Curso de Transfer Learning con Hugging Face"](https://platzi.com/cursos/hugging-face/)
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.5365 | 1.09 | 500 | 0.5993 | 0.8211 | 0.8809 |
| 0.3458 | 2.18 | 1000 | 0.8336 | 0.8235 | 0.8767 |
### Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# platzi-distilroberta-base-mrpc-glue-joselier
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the glue and the mrpc datasets.
It achieves the following results on the evaluation set:
- Loss: 0.5993
- Accuracy: 0.8211
- F1: 0.8809
## Model description
This model uses transfer learning to classify 2 sentences (a string of 2 sentences separated by a comma) in "Equivalent" or "Not Equivalent".
The model platzi-distilroberta-base-mrpc-glue-joselier was programmed as part of a class from Platzi's course ["Curso de Transfer Learning con Hugging Face"](https://platzi.com/cursos/hugging-face/)
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.5365 | 1.09 | 500 | 0.5993 | 0.8211 | 0.8809 |
| 0.3458 | 2.18 | 1000 | 0.8336 | 0.8235 | 0.8767 |
### Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
|
{"base_model": "distilroberta-base", "datasets": ["glue"], "license": "apache-2.0", "metrics": ["accuracy", "f1"], "tags": ["text-classification", "generated_from_trainer"], "widget": [{"text": ["Yucaipa owned Dominick 's before selling the chain to Safeway in 1998 for $ 2.5 billion.", "Yucaipa bought Dominick's in 1995 for $ 693 million and sold it to Safeway for $ 1.8 billion in 1998."], "example_title": "Not Equivalent"}, {"text": ["Revenue in the first quarter of the year dropped 15 percent from the same period a year earlier.", "With the scandal hanging over Stewart's company revenue the first quarter of the year dropped 15 percent from the same period a year earlier."], "example_title": "Equivalent"}], "model-index": [{"name": "platzi-distilroberta-base-mrpc-glue-joselier", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "glue", "type": "glue", "config": "mrpc", "split": "validation", "args": "mrpc"}, "metrics": [{"type": "accuracy", "value": 0.821078431372549, "name": "Accuracy"}, {"type": "f1", "value": 0.8809135399673736, "name": "F1"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,921 |
pinzhenchen/sft-lora-es-bloom-1b7
|
pinzhenchen
| null |
[
"generation",
"question answering",
"instruction tuning",
"es",
"arxiv:2309.08958",
"license:cc-by-nc-4.0",
"region:us"
] | 2024-03-05T23:46:50Z |
2024-03-05T23:46:53+00:00
| 0 | 0 |
---
language:
- es
license: cc-by-nc-4.0
tags:
- generation
- question answering
- instruction tuning
---
### Model Description
This HF repository contains base LLMs instruction tuned (SFT) with LoRA and then used to study whether monolingual or multilingual instruction tuning is more favourable.
* [GitHub](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main)
* [Paper](https://arxiv.org/abs/2309.08958)
#### Instruction tuning details
* Base model: [bigscience/bloom-1b7](https://huggingface.co/bigscience/bloom-1b7)
* Instruction tuning language: Spanish
* Training method: LoRA.
* LoRA details: rank=8, alpha=16, target modules={key, query, value}.
* Best checkpoint: best cross-entropy on a validation set, trained for 5 epochs.
* Dataset: machine-translated from [yahma/alpaca-cleaned](https://huggingface.co/datasets/yahma/alpaca-cleaned). You can download our data [HERE](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main/training-data).
#### Usage
The model checkpoint should be loaded with the base model together using `transformers` and `peft` libraries.
Please refer to our Github repository [HERE](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main/loraft) for inference and training instructions.
#### Citation
```
@inproceedings{chen-etal-2024-monolingual,
title="Monolingual or multilingual instruction tuning: Which makes a better {Alpaca}",
author="Pinzhen Chen and Shaoxiong Ji and Nikolay Bogoychev and Andrey Kutuzov and Barry Haddow and Kenneth Heafield",
year="2024",
booktitle = "Findings of the Association for Computational Linguistics: EACL 2024",
}
```
| null |
Non_BioNLP
|
### Model Description
This HF repository contains base LLMs instruction tuned (SFT) with LoRA and then used to study whether monolingual or multilingual instruction tuning is more favourable.
* [GitHub](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main)
* [Paper](https://arxiv.org/abs/2309.08958)
#### Instruction tuning details
* Base model: [bigscience/bloom-1b7](https://huggingface.co/bigscience/bloom-1b7)
* Instruction tuning language: Spanish
* Training method: LoRA.
* LoRA details: rank=8, alpha=16, target modules={key, query, value}.
* Best checkpoint: best cross-entropy on a validation set, trained for 5 epochs.
* Dataset: machine-translated from [yahma/alpaca-cleaned](https://huggingface.co/datasets/yahma/alpaca-cleaned). You can download our data [HERE](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main/training-data).
#### Usage
The model checkpoint should be loaded with the base model together using `transformers` and `peft` libraries.
Please refer to our Github repository [HERE](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main/loraft) for inference and training instructions.
#### Citation
```
@inproceedings{chen-etal-2024-monolingual,
title="Monolingual or multilingual instruction tuning: Which makes a better {Alpaca}",
author="Pinzhen Chen and Shaoxiong Ji and Nikolay Bogoychev and Andrey Kutuzov and Barry Haddow and Kenneth Heafield",
year="2024",
booktitle = "Findings of the Association for Computational Linguistics: EACL 2024",
}
```
|
{"language": ["es"], "license": "cc-by-nc-4.0", "tags": ["generation", "question answering", "instruction tuning"]}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,922 |
valenaparicio16/bioBERT-finetuned-financial-phrasebank
|
valenaparicio16
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"safetensors",
"bert",
"text-classification",
"generated_from_trainer",
"dataset:financial_phrasebank",
"base_model:dmis-lab/biobert-v1.1",
"base_model:finetune:dmis-lab/biobert-v1.1",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-11-02T02:10:25Z |
2023-11-21T03:57:45+00:00
| 24 | 0 |
---
base_model: dmis-lab/biobert-v1.1
datasets:
- financial_phrasebank
metrics:
- f1
tags:
- generated_from_trainer
model-index:
- name: bioBERT-finetuned-financial-phrasebank
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: financial_phrasebank
type: financial_phrasebank
config: sentences_66agree
split: train
args: sentences_66agree
metrics:
- type: f1
value: 0.9153044889504942
name: F1
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bioBERT-finetuned-financial-phrasebank
This model is a fine-tuned version of [dmis-lab/biobert-v1.1](https://huggingface.co/dmis-lab/biobert-v1.1) on the financial_phrasebank dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3305
- F1: 0.9153
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log | 1.0 | 238 | 0.3082 | 0.8843 |
| No log | 2.0 | 476 | 0.3200 | 0.9001 |
| 0.3619 | 3.0 | 714 | 0.3305 | 0.9153 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bioBERT-finetuned-financial-phrasebank
This model is a fine-tuned version of [dmis-lab/biobert-v1.1](https://huggingface.co/dmis-lab/biobert-v1.1) on the financial_phrasebank dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3305
- F1: 0.9153
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log | 1.0 | 238 | 0.3082 | 0.8843 |
| No log | 2.0 | 476 | 0.3200 | 0.9001 |
| 0.3619 | 3.0 | 714 | 0.3305 | 0.9153 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0
|
{"base_model": "dmis-lab/biobert-v1.1", "datasets": ["financial_phrasebank"], "metrics": ["f1"], "tags": ["generated_from_trainer"], "model-index": [{"name": "bioBERT-finetuned-financial-phrasebank", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "financial_phrasebank", "type": "financial_phrasebank", "config": "sentences_66agree", "split": "train", "args": "sentences_66agree"}, "metrics": [{"type": "f1", "value": 0.9153044889504942, "name": "F1"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,923 |
gokuls/distilbert_add_GLUE_Experiment_logit_kd_qnli_192
|
gokuls
|
text-classification
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"en",
"dataset:glue",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | 2023-01-28T23:21:00Z |
2023-01-28T23:39:57+00:00
| 138 | 0 |
---
datasets:
- glue
language:
- en
license: apache-2.0
metrics:
- accuracy
tags:
- generated_from_trainer
model-index:
- name: distilbert_add_GLUE_Experiment_logit_kd_qnli_192
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: GLUE QNLI
type: glue
config: qnli
split: validation
args: qnli
metrics:
- type: accuracy
value: 0.5830129965220575
name: Accuracy
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert_add_GLUE_Experiment_logit_kd_qnli_192
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the GLUE QNLI dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3981
- Accuracy: 0.5830
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.4154 | 1.0 | 410 | 0.4115 | 0.5054 |
| 0.4103 | 2.0 | 820 | 0.4001 | 0.5826 |
| 0.3967 | 3.0 | 1230 | 0.3981 | 0.5830 |
| 0.3897 | 4.0 | 1640 | 0.3995 | 0.5942 |
| 0.3849 | 5.0 | 2050 | 0.4017 | 0.5885 |
| 0.3804 | 6.0 | 2460 | 0.4072 | 0.5836 |
| 0.3763 | 7.0 | 2870 | 0.4096 | 0.5751 |
| 0.3717 | 8.0 | 3280 | 0.4092 | 0.5773 |
### Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert_add_GLUE_Experiment_logit_kd_qnli_192
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the GLUE QNLI dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3981
- Accuracy: 0.5830
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.4154 | 1.0 | 410 | 0.4115 | 0.5054 |
| 0.4103 | 2.0 | 820 | 0.4001 | 0.5826 |
| 0.3967 | 3.0 | 1230 | 0.3981 | 0.5830 |
| 0.3897 | 4.0 | 1640 | 0.3995 | 0.5942 |
| 0.3849 | 5.0 | 2050 | 0.4017 | 0.5885 |
| 0.3804 | 6.0 | 2460 | 0.4072 | 0.5836 |
| 0.3763 | 7.0 | 2870 | 0.4096 | 0.5751 |
| 0.3717 | 8.0 | 3280 | 0.4092 | 0.5773 |
### Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2
|
{"datasets": ["glue"], "language": ["en"], "license": "apache-2.0", "metrics": ["accuracy"], "tags": ["generated_from_trainer"], "model-index": [{"name": "distilbert_add_GLUE_Experiment_logit_kd_qnli_192", "results": [{"task": {"type": "text-classification", "name": "Text Classification"}, "dataset": {"name": "GLUE QNLI", "type": "glue", "config": "qnli", "split": "validation", "args": "qnli"}, "metrics": [{"type": "accuracy", "value": 0.5830129965220575, "name": "Accuracy"}]}]}]}
|
task
|
[
"TEXT_CLASSIFICATION"
] | 39,924 |
valhalla/t5-small-qa-qg-hl
|
valhalla
|
text2text-generation
|
[
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"question-generation",
"dataset:squad",
"arxiv:1910.10683",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2022-03-02T23:29:05Z |
2021-06-23T14:42:41+00:00
| 12,515 | 12 |
---
datasets:
- squad
license: mit
tags:
- question-generation
widget:
- text: 'generate question: <hl> 42 <hl> is the answer to life, the universe and everything.
</s>'
- text: 'question: What is 42 context: 42 is the answer to life, the universe and
everything. </s>'
---
## T5 for multi-task QA and QG
This is multi-task [t5-small](https://arxiv.org/abs/1910.10683) model trained for question answering and answer aware question generation tasks.
For question generation the answer spans are highlighted within the text with special highlight tokens (`<hl>`) and prefixed with 'generate question: '. For QA the input is processed like this `question: question_text context: context_text </s>`
You can play with the model using the inference API. Here's how you can use it
For QG
`generate question: <hl> 42 <hl> is the answer to life, the universe and everything. </s>`
For QA
`question: What is 42 context: 42 is the answer to life, the universe and everything. </s>`
For more deatils see [this](https://github.com/patil-suraj/question_generation) repo.
### Model in action 🚀
You'll need to clone the [repo](https://github.com/patil-suraj/question_generation).
[](https://colab.research.google.com/github/patil-suraj/question_generation/blob/master/question_generation.ipynb)
```python3
from pipelines import pipeline
nlp = pipeline("multitask-qa-qg")
# to generate questions simply pass the text
nlp("42 is the answer to life, the universe and everything.")
=> [{'answer': '42', 'question': 'What is the answer to life, the universe and everything?'}]
# for qa pass a dict with "question" and "context"
nlp({
"question": "What is 42 ?",
"context": "42 is the answer to life, the universe and everything."
})
=> 'the answer to life, the universe and everything'
```
| null |
Non_BioNLP
|
## T5 for multi-task QA and QG
This is multi-task [t5-small](https://arxiv.org/abs/1910.10683) model trained for question answering and answer aware question generation tasks.
For question generation the answer spans are highlighted within the text with special highlight tokens (`<hl>`) and prefixed with 'generate question: '. For QA the input is processed like this `question: question_text context: context_text </s>`
You can play with the model using the inference API. Here's how you can use it
For QG
`generate question: <hl> 42 <hl> is the answer to life, the universe and everything. </s>`
For QA
`question: What is 42 context: 42 is the answer to life, the universe and everything. </s>`
For more deatils see [this](https://github.com/patil-suraj/question_generation) repo.
### Model in action 🚀
You'll need to clone the [repo](https://github.com/patil-suraj/question_generation).
[](https://colab.research.google.com/github/patil-suraj/question_generation/blob/master/question_generation.ipynb)
```python3
from pipelines import pipeline
nlp = pipeline("multitask-qa-qg")
# to generate questions simply pass the text
nlp("42 is the answer to life, the universe and everything.")
=> [{'answer': '42', 'question': 'What is the answer to life, the universe and everything?'}]
# for qa pass a dict with "question" and "context"
nlp({
"question": "What is 42 ?",
"context": "42 is the answer to life, the universe and everything."
})
=> 'the answer to life, the universe and everything'
```
|
{"datasets": ["squad"], "license": "mit", "tags": ["question-generation"], "widget": [{"text": "generate question: <hl> 42 <hl> is the answer to life, the universe and everything. </s>"}, {"text": "question: What is 42 context: 42 is the answer to life, the universe and everything. </s>"}]}
|
task
|
[
"QUESTION_ANSWERING"
] | 39,925 |
antonkurylo/t5-base-billsum
|
antonkurylo
|
summarization
|
[
"transformers",
"tensorboard",
"safetensors",
"t5",
"text2text-generation",
"summarization",
"generated_from_trainer",
"base_model:google-t5/t5-base",
"base_model:finetune:google-t5/t5-base",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | 2024-10-22T16:00:04Z |
2024-10-23T20:33:04+00:00
| 29 | 0 |
---
base_model: t5-base
library_name: transformers
license: apache-2.0
metrics:
- rouge
tags:
- summarization
- generated_from_trainer
model-index:
- name: t5-base-billsum
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-base-billsum
This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6188
- Rouge1: 51.4114
- Rouge2: 30.6521
- Rougel: 40.9417
- Rougelsum: 44.6839
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|
| 1.9236 | 1.0 | 1185 | 1.5895 | 52.5513 | 32.239 | 42.0215 | 45.9665 |
| 1.7231 | 2.0 | 2370 | 1.5380 | 53.3168 | 33.2784 | 42.9286 | 46.7854 |
| 1.6708 | 3.0 | 3555 | 1.5187 | 53.2982 | 33.3262 | 42.979 | 46.8863 |
| 1.7884 | 4.0 | 4740 | 1.6197 | 51.4854 | 30.768 | 41.0231 | 44.7727 |
| 1.8212 | 5.0 | 5925 | 1.6188 | 51.4114 | 30.6521 | 40.9417 | 44.6839 |
### Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.19.1
| null |
Non_BioNLP
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-base-billsum
This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6188
- Rouge1: 51.4114
- Rouge2: 30.6521
- Rougel: 40.9417
- Rougelsum: 44.6839
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|
| 1.9236 | 1.0 | 1185 | 1.5895 | 52.5513 | 32.239 | 42.0215 | 45.9665 |
| 1.7231 | 2.0 | 2370 | 1.5380 | 53.3168 | 33.2784 | 42.9286 | 46.7854 |
| 1.6708 | 3.0 | 3555 | 1.5187 | 53.2982 | 33.3262 | 42.979 | 46.8863 |
| 1.7884 | 4.0 | 4740 | 1.6197 | 51.4854 | 30.768 | 41.0231 | 44.7727 |
| 1.8212 | 5.0 | 5925 | 1.6188 | 51.4114 | 30.6521 | 40.9417 | 44.6839 |
### Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.19.1
|
{"base_model": "t5-base", "library_name": "transformers", "license": "apache-2.0", "metrics": ["rouge"], "tags": ["summarization", "generated_from_trainer"], "model-index": [{"name": "t5-base-billsum", "results": []}]}
|
task
|
[
"SUMMARIZATION"
] | 39,926 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.