metadata

base_model: sentence-transformers/paraphrase-multilingual-mpnet-base-v2
datasets: []
language: []
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:2481
  - loss:BatchAllTripletLoss
  - loss:MultipleNegativesRankingLoss
  - loss:CachedMultipleNegativesRankingLoss
widget:
  - source_sentence: 'Territorial Scope: Worldwide'
    sentences:
      - 'La definizione di Rischio è riportata in APPENDICE. '
      - >-
        The total sum actually paid or payable by the [MASK] in settlement of
        [MASK] (excluding [MASK] in [MASK] of [MASK] [MASK] Extra Contractual
        Obligations [MASK] Loss Adjustment [MASK] as [MASK] defined) paid or
        payable by [MASK] Company in settlement of claims and in satisfaction of
        [MASK] [MASK] on account of such claims [MASK] deduction of all [MASK]
        [MASK] recoveries and [MASK] [MASK] on inuring reinsurance; whether
        [MASK] or [MASK] Nothing herein shall be construed to mean that [MASK]
        under this Contract are not recoverable until the Ultimate [MASK] Loss
        has been ascertained. 
      - 'Territorial Scope: Worldwide '
  - source_sentence: >-
      ARTICLE XXIV RESERVES AND FUNDING – CREDIT FOR REINSURANCE A. The
      Reinsurer shall provide funding under the terms of this Article only if
      the Company will be denied statutory credit for reinsurance ceded to the
      Reinsurer pursuant to the credit for reinsurance law or regulations of the
      regulatory authority having jurisdiction over the Company’s reserves. B.
      As regards Policies issued by the Company coming within the scope of this
      Contract; the Company agrees that when it shall file with the insurance
      regulatory authority or set up on its books reserves for unearned premium
      and losses covered hereunder which it shall be required by law to set up;
      it shall forward to the Reinsurer a statement showing the proportion of
      such reserves which is applicable to the Reinsurer. The Reinsurer hereby
      agrees to fund such reserves in respect of unearned premium; known
      outstanding losses that have been reported to the Reinsurer and allocated
      Loss Adjustment Expense relating thereto; losses and allocated Loss
      Adjustment Expense paid by the Company but not recovered from the
      Reinsurer; plus reserves for losses incurred but not reported; as shown in
      the statement prepared by the Company (hereinafter referred to as the
      “Reinsurer’s Obligations”) by funds withheld; cash advances or a Letter of
      Credit (hereinafter referred to as “LOC”). The Reinsurer shall have the
      option of determining the method of funding provided it is acceptable to
      the Company and each applicable insurance regulatory authority.
    sentences:
      - >-
        1. The term 'Ultimate Net Loss'; as used herein; shall mean the sum
        actually paid or payable (including any amounts paid or payable under
        the Excess of Original Policy Limits and Extra Contractual Obligations
        NMX100 (Amended) Articles contained herein) by the Reinsured (including
        any Retention Vehicle(s) used by the Reinsured) in settlement of losses;
        including legal costs and expenses of litigation; if any; after making
        deductions for all recoveries; all salvages and all claims upon other
        inuring reinsurances; whether collected or not; and shall include all
        Loss Adjustment Expenses arising from the settlement of claims other
        than the salaries of employees and office expenses of the Reinsured.
      - >-
        Excluding:

        (a) All Business derived directly or indirectly from any Pool;
        Association or Syndicate

        which maintains its own reinsurance facilities.  Captive Insurers or
        Risk Retention Groups shall not be considered Pools; Associations or
        Syndicates.

        (b) Any Pool or Scheme (whether voluntary or mandatory) formed after
        March 1; 1968 for

        the purpose of insuring Property whether on a country- wide basis or in
        respect of

        designated areas.  This exclusion shall not apply to so- called
        Automobile Insurance

        Plans or other Pools formed to provide coverage for Automobile Physical
        Damage;

        however; any coverage provided by the Ontario; Alberta; New Brunswick
        and Nova

        Scotia Risk Sharing Pools shall be excluded hereunder.
      - >-
        Notwithstanding any provision to the contrary within this Reinsurance
        Contract; this Reinsurance Contract excludes any loss; damage;
        liability; claim; cost or expense of whatsoever nature; directly or
        indirectly caused by; contributed to by; resulting from; arising out of;
        or in connection with a Communicable Disease or the fear or threat
        (whether actual or perceived) of a Communicable Disease regardless of
        any other cause or event contributing concurrently or in any other
        sequence thereto.
  - source_sentence: >-
      "Special Circumstance Subscribing Reinsurer" as used herein will mean a
      Subscribing Reinsurer that experiences one or more of the circumstances
      set forth in paragraph A of the Special Termination Article.
    sentences:
      - >-
        “Net Loss” as used in this Agreement shall mean (1) the sum or sums paid
        or scheduled to be paid by the Company for which it is liable; under
        Policies reinsured hereunder; including; (2) any Loss Adjustment Expense
        which erodes part of the Policy limits under the Company’s Policies;
        after deduction of all salvage and all recoveries (including recoveries
        of all claims on inuring insurance or reinsurance; whether collectible
        or not). In the event that the Company has no obligation to make an
        indemnity payment in connection with a declaratory judgment action; the
        Company’s Declaratory Judgment Expense as defined herein shall be
        treated as Net Loss and shall be recoverable hereunder subject to the
        applicable limits of this Agreement. All recoveries; payments and
        reversals or reductions of verdicts or judgments whether recovered;
        received or obtained prior or subsequent to loss settlement under this
        Agreement; including amounts recoverable under other reinsurance whether
        collected or not; shall be applied as if recovered; received or obtained
        prior to the aforesaid settlement and shall be deducted from the actual
        losses sustained to arrive at the amount of the Net Loss. Nothing herein
        shall be construed to mean that losses under this Agreement are not
        recoverable until the Company’s final Net Loss has been ascertained.
      - >-
        The term “Total Insurable Value” shall mean the overall combined amount
        composed of the total full values (whether insured or not) for all
        buildings and contents and total sums insured for Business Interruption
        and/or Extra Expenses for an individual company or group of companies
        for exposures in which the Reinsured is involved or whose existence
        and/or whose insurable values are known to the Reinsured at the
        inception of such Policies.
      - NAIC Code 10759
  - source_sentence: Custom Wholesale Property; except for Large Property U.S.
    sentences:
      - >-
        exceeds the amount of the Affiliated Companies’ aggregate Insured Loss;
        this excess shall be the “Excess Amount(s)”.
      - >-
        In lieu of the provisions of subparagraph 16 of paragraph A of Article 5
        - Exclusions - the following shall apply: '16. Loss or liability
        excluded by the Limited communicable disease exclusion(Property Treaty
        Reinsurance) attached to and forming part of this Contract.'
      - La definizione di Rischio è riportata in APPENDICE.
  - source_sentence: >-
      ARTICLE 33 – FOREIGN ACCOUNT TAX COMPLIANCE ACT (FATCA) A. At inception of
      this Contract; but in no event later than five (5) business days prior to
      the first premium payment hereunder; the Reinsurer shall provide to the
      Company or its Intermediary such documentation required under FATCA that
      confirms that the Reinsurer is not subject to FATCA withholding.
    sentences:
      - >-
        A. This Article applies only to those Subscribing Reinsurers not
        domiciled in the United States of America; and/or not authorized in any
        state; territory and/or district of the United States of America where
        authorization is required by insurance regulatory authorities.
      - >-
        This Article applies: only to the extent a Subscribing Reinsurer does
        not qualify for credit with any insurance regulatory authority having
        jurisdiction over the Company’s reserves; or. The Subscribing
        Reinsurer(s) shall promptly notify the Company of any event or change or
        condition that will reasonably result in any event that may result in
        the Company being unable to obtain full statutory financial statement
        credit for the reinsurance provided under this Contract in any
        applicable United States jurisdiction at any point in time during the
        term of this Contract.
      - >-
        ARTICLE 33 – FOREIGN ACCOUNT TAX COMPLIANCE ACT (FATCA) A. At inception
        of this Contract; but in no event later than five (5) business days
        prior to the first premium payment hereunder; the Reinsurer shall
        provide to the Company or its Intermediary such documentation required
        under FATCA that confirms that the Reinsurer is not subject to FATCA
        withholding.

SentenceTransformer based on sentence-transformers/paraphrase-multilingual-mpnet-base-v2

This is a sentence-transformers model finetuned from sentence-transformers/paraphrase-multilingual-mpnet-base-v2 on the apnc and apnd datasets. 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/paraphrase-multilingual-mpnet-base-v2
Maximum Sequence Length: 128 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity
Training Datasets:
- apnc
- apnd

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (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:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("Anakeen/datasets_apn_sts_0.8")
# Run inference
sentences = [
    'ARTICLE 33 – FOREIGN ACCOUNT TAX COMPLIANCE ACT (FATCA) A. At inception of this Contract; but in no event later than five (5) business days prior to the first premium payment hereunder; the Reinsurer shall provide to the Company or its Intermediary such documentation required under FATCA that confirms that the Reinsurer is not subject to FATCA withholding.',
    'ARTICLE 33 – FOREIGN ACCOUNT TAX COMPLIANCE ACT (FATCA) A. At inception of this Contract; but in no event later than five (5) business days prior to the first premium payment hereunder; the Reinsurer shall provide to the Company or its Intermediary such documentation required under FATCA that confirms that the Reinsurer is not subject to FATCA withholding.',
    'A. This Article applies only to those Subscribing Reinsurers not domiciled in the United States of America; and/or not authorized in any state; territory and/or district of the United States of America where authorization is required by insurance regulatory authorities.',
]
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]

Training Details

Training Datasets

apnc

Dataset: apnc
Size: 635 training samples
Columns: anchor, positive, and negative

Approximate statistics based on the first 1000 samples:

	anchor	positive	negative
type	string	string	string
details	min: 4 tokens mean: 94.86 tokens max: 128 tokens	min: 4 tokens mean: 94.86 tokens max: 128 tokens	min: 8 tokens mean: 105.14 tokens max: 128 tokens

Samples:

anchor	positive	negative
1. This Contract does not cover any loss or liability accruing to the Reassured; directly or indirectly and whether as Insurer or Reinsurer; from any Pool of Insurers or Reinsurers formed for the purpose of covering Atomic or Nuclear Energy risks. 2. Without in any way restricting the operation of paragraph (1) of this Clause; this Contract does not cover any loss or liability accruing to the Reassured; directly or indirectly and whether as Insurer or Reinsurer; from any insurance against Physical Damage (including business interruption or consequential loss arising out of such Physical Damage) to: I. Nuclear reactor power plants including all auxiliary property on the site; or II. Any other nuclear reactor installation; including laboratories handling radioactive materials in connection with reactor installations; and “critical facilities” as such; or III. Installations for fabricating complete fuel elements or for processing substantial quantities of “special nuclear material;” and for reprocessing; salvaging; chemically separating; storing or disposing of “spent” nuclear fuel or waste materials; or IV. Installations other than those listed in paragraph (2) III above using substantial quantities of radioactive isotopes or other products of nuclear fission. 3. Without in any way restricting the operations of paragraphs (1) and (2) hereof; this Contract does not cover any loss or liability by radioactive contamination accruing to the Reassured; directly or indirectly; and whether as Insurer or Reinsurer; from any insurance on property which is on the same site as a nuclear reactor power plant or other nuclear installation and which normally would be insured therewith except that this paragraph (3) shall not operate (a) where the Reassured does not have knowledge of such nuclear reactor power plant or nuclear installation; or (b) where said insurance contains a provision excluding coverage for damage to property caused by or resulting from radioactive contamination; however caused. However on and after 1st January 1960 this sub-paragraph (b) shall only apply provided the said radioactive contamination exclusion provision has been approved by the Governmental Authority having jurisdiction thereof. 4. Without in any way restricting the operations of paragraphs (1); (2) and (3) hereof; this Contract does not cover any loss or liability by radioactive contamination accruing to the Reassured; directly or indirectly; and whether as Insurer or Reinsurer; when such radioactive contamination is a named hazard specifically insured against. 5. It is understood and agreed that this Clause shall not extend to risks using radioactive isotopes in any form where the nuclear exposure is not considered by the Reassured to be the primary hazard. 6. The term “special nuclear material” shall have the meaning given it in the Atomic Energy Act of 1954 or by any law amendatory thereof. 7. The Reassured to be sole judge of what constitutes: (a) substantial quantities; and (b) the extent of installation; plant or site. Note. Without in any way restricting the operation of paragraph (1) hereof; it is understood and agreed that (a) all policies issued by the Reassured on or before 31st December 1957 shall be free from the application of the other provisions of this Clause until expiry date or 31st December 1960 whichever first occurs whereupon all the provisions of this Clause shall apply; (b) with respect to any risk located in Canada policies issued by the Reassured on or before 31st December 1958 shall be free from the application of the other provisions of this Clause until expiry date or 31st December 1960 whichever first occurs whereupon all the provisions of this Clause shall apply.	1. This Contract does not cover any loss or liability accruing to the Reassured; directly or indirectly and whether as Insurer or Reinsurer; from any Pool of Insurers or Reinsurers formed for the purpose of covering Atomic or Nuclear Energy risks. 2. Without in any way restricting the operation of paragraph (1) of this Clause; this Contract does not cover any loss or liability accruing to the Reassured; directly or indirectly and whether as Insurer or Reinsurer; from any insurance against Physical Damage (including business interruption or consequential loss arising out of such Physical Damage) to: I. Nuclear reactor power plants including all auxiliary property on the site; or II. Any other nuclear reactor installation; including laboratories handling radioactive materials in connection with reactor installations; and “critical facilities” as such; or III. Installations for fabricating complete fuel elements or for processing substantial quantities of “special nuclear material;” and for reprocessing; salvaging; chemically separating; storing or disposing of “spent” nuclear fuel or waste materials; or IV. Installations other than those listed in paragraph (2) III above using substantial quantities of radioactive isotopes or other products of nuclear fission. 3. Without in any way restricting the operations of paragraphs (1) and (2) hereof; this Contract does not cover any loss or liability by radioactive contamination accruing to the Reassured; directly or indirectly; and whether as Insurer or Reinsurer; from any insurance on property which is on the same site as a nuclear reactor power plant or other nuclear installation and which normally would be insured therewith except that this paragraph (3) shall not operate (a) where the Reassured does not have knowledge of such nuclear reactor power plant or nuclear installation; or (b) where said insurance contains a provision excluding coverage for damage to property caused by or resulting from radioactive contamination; however caused. However on and after 1st January 1960 this sub-paragraph (b) shall only apply provided the said radioactive contamination exclusion provision has been approved by the Governmental Authority having jurisdiction thereof. 4. Without in any way restricting the operations of paragraphs (1); (2) and (3) hereof; this Contract does not cover any loss or liability by radioactive contamination accruing to the Reassured; directly or indirectly; and whether as Insurer or Reinsurer; when such radioactive contamination is a named hazard specifically insured against. 5. It is understood and agreed that this Clause shall not extend to risks using radioactive isotopes in any form where the nuclear exposure is not considered by the Reassured to be the primary hazard. 6. The term “special nuclear material” shall have the meaning given it in the Atomic Energy Act of 1954 or by any law amendatory thereof. 7. The Reassured to be sole judge of what constitutes: (a) substantial quantities; and (b) the extent of installation; plant or site. Note. Without in any way restricting the operation of paragraph (1) hereof; it is understood and agreed that (a) all policies issued by the Reassured on or before 31st December 1957 shall be free from the application of the other provisions of this Clause until expiry date or 31st December 1960 whichever first occurs whereupon all the provisions of this Clause shall apply; (b) with respect to any risk located in Canada policies issued by the Reassured on or before 31st December 1958 shall be free from the application of the other provisions of this Clause until expiry date or 31st December 1960 whichever first occurs whereupon all the provisions of this Clause shall apply.	`This Contract shall exclude: a) Business defined by the Reinsured as Liability Business (unless included in Cargo or Engineering All Risks/Contractors All Risks Business). b) Space and related risks. c) Marine business; but not applying to pleasure craft. d) Disease losses in respect of Fish Farm. This Contract shall also be subject to the following exclusion clauses: a) War and Civil War Exclusion NMA 464.`
`Downgrading clause ~ ABR1001 (Amended) Reinsurer with an S&P Rating Unless otherwise agreed by the Reinsured; the Reinsurer shall at all times during the Period of this Contract maintain an Insurer Financial Strength (IFS) rating from Standard & Poor's Rating Group of 55 Water Street; New York; NY 10041; USA ("S&P") equal to or greater than a rating of A minus as applied by S&P to that Reinsurer.`	`Downgrading clause ~ ABR1001 (Amended) Reinsurer with an S&P Rating Unless otherwise agreed by the Reinsured; the Reinsurer shall at all times during the Period of this Contract maintain an Insurer Financial Strength (IFS) rating from Standard & Poor's Rating Group of 55 Water Street; New York; NY 10041; USA ("S&P") equal to or greater than a rating of A minus as applied by S&P to that Reinsurer.`	Communicable disease clause - LMA 5394 1. Notwithstanding any provision to the contrary within this reinsurance agreement; this reinsurance agreement excludes any loss; damage; liability; claim; cost or expense of whatsoever nature; directly or indirectly caused by; contributed to by; resulting from; arising out of; or in connection with a Communicable Disease or the fear or threat (whether actual or perceived) of a Communicable Disease regardless of any other cause or event contributing concurrently or in any other sequence thereto. 2. As used herein; a Communicable Disease means any disease which can be transmitted by means of any substance or agent from any organism to another organism where: 2.1. the substance or agent includes; but is not limited to; a virus; bacterium; parasite or other organism or any variation thereof; whether deemed living or not; and 2.2. the method of transmission; whether direct or indirect; includes but is not limited to; airborne transmission; bodily fluid transmission; transmission from or to any surface or object; solid; liquid or gas or between organisms; and 2.3. the disease; substance or agent can cause or threaten damage to human health or human welfare or can cause or threaten damage to; deterioration of; loss of value of; marketability of or loss of use of property.
`Dispute Resolution ~ ABR1004 Where any dispute or difference between the parties arising out of or in connection with this Contract; including formation and validity and whether arising during or after the period of this Contract; has not been settled through negotiation; both parties agree to try in good faith to settle such dispute by non- binding mediation; before resorting to arbitration in the manner set out below.`	`Dispute Resolution ~ ABR1004 Where any dispute or difference between the parties arising out of or in connection with this Contract; including formation and validity and whether arising during or after the period of this Contract; has not been settled through negotiation; both parties agree to try in good faith to settle such dispute by non- binding mediation; before resorting to arbitration in the manner set out below.`	`Brokerage for this Contract is 15.00% of gross ceded premium. No brokerage will be paid on reinstatement premium.`

Loss: CachedMultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

apnd

Dataset: apnd
Size: 1,846 training samples
Columns: anchor, positive, and negative

Approximate statistics based on the first 1000 samples:

	anchor	positive	negative
type	string	string	string
details	min: 3 tokens mean: 62.84 tokens max: 128 tokens	min: 3 tokens mean: 62.84 tokens max: 128 tokens	min: 5 tokens mean: 73.05 tokens max: 128 tokens

Samples:

anchor	positive	negative
`“North American CAT Perils” means certain Named Storms and Earthquake; each as defined below; in respect of that portion of losses which occur in the United States and Canada and their possessions and territories; excluding the Territory of Guam; the Territory of American Samoa; the Commonwealth of the Northern Mariana Islands; Wake Island; Johnston Atoll; Palmyra Atoll; and the State of Hawaiiterritory of Guam.`	`“North American CAT Perils” means certain Named Storms and Earthquake; each as defined below; in respect of that portion of losses which occur in the United States and Canada and their possessions and territories; excluding the Territory of Guam; the Territory of American Samoa; the Commonwealth of the Northern Mariana Islands; Wake Island; Johnston Atoll; Palmyra Atoll; and the State of Hawaiiterritory of Guam.`	`'Insurance Compensation' shall mean any compensation; interest or Allocated Expenses paid or payable by the Reinsured in respect of any loss occurrence under Policies covered under this Agreement.`
`For the purposes of this Paragraph A.; “Named Storm” means any windstorm or windstorm system that has been named by a Reporting Agency at any time in its lifecycle and ensuing losses therefrom.`	`For the purposes of this Paragraph A.; “Named Storm” means any windstorm or windstorm system that has been named by a Reporting Agency at any time in its lifecycle and ensuing losses therefrom.`	`‘Contingency policies’ means contracts of contingency insurance unless: a) written as an integral component of General Cover or b) the subject of a binding written commitment on or before 31 December 2018 and incepting or renewing on or before 31 March 2019.`
`For the purposes of this Paragraph A.; “Earthquake” means earthquake shake and ensuing losses therefrom.`	`For the purposes of this Paragraph A.; “Earthquake” means earthquake shake and ensuing losses therefrom.`	`Means any programme code; programming instruction or other set of instructions intentionally constructed with the ability to damage; interfere with or otherwise adversely affect computer programmes; data files or operations (whether involving self-replication or not); including but not limited to “Virus;” “Trojan Horses;” “Worms;” “Logic Bombs;” or “Denial of Service Attack.”`

Loss: CachedMultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

Evaluation Datasets

sscc

Dataset: sscc
Size: 100 evaluation samples
Columns: sentence and label

Approximate statistics based on the first 1000 samples:

	sentence	label
type	string	int
details	min: 20 tokens mean: 104.63 tokens max: 128 tokens	1: ~1.00% 4: ~2.00% 5: ~3.00% 11: ~1.00% 12: ~1.00% 18: ~1.00% 19: ~6.00% 20: ~8.00% 21: ~3.00% 30: ~1.00% 32: ~1.00% 34: ~1.00% 45: ~1.00% 46: ~2.00% 47: ~2.00% 50: ~1.00% 51: ~1.00% 53: ~4.00% 65: ~1.00% 68: ~2.00% 69: ~1.00% 74: ~1.00% 79: ~5.00% 88: ~1.00% 89: ~1.00% 93: ~1.00% 107: ~3.00% 115: ~1.00% 126: ~1.00% 143: ~1.00% 150: ~1.00% 161: ~1.00% 175: ~1.00% 202: ~1.00% 220: ~1.00% 222: ~1.00% 227: ~1.00% 229: ~1.00% 231: ~1.00% 232: ~1.00% 235: ~1.00% 236: ~1.00% 251: ~2.00% 275: ~1.00% 276: ~1.00% 296: ~1.00% 305: ~1.00% 309: ~1.00% 314: ~1.00% 334: ~1.00% 342: ~1.00% 368: ~1.00% 376: ~1.00% 381: ~1.00% 400: ~1.00% 402: ~1.00% 404: ~1.00% 419: ~1.00% 473: ~1.00% 496: ~1.00% 547: ~1.00% 548: ~1.00% 553: ~1.00% 585: ~1.00% 594: ~2.00% 605: ~1.00% 606: ~1.00% 662: ~1.00% 798: ~1.00%

Samples:

sentence	label
`Article 26 - Federal Excise Tax (BRMA 17D) A.The Reinsurer has agreed to allow for the purpose of paying the Federal Excise Tax the applicable percentage of the premium payable hereon (as imposed under Section 4371 of the Internal Revenue Code) to the extent such premium is subject to the Federal Excise Tax.`	`32`
`Notwithstanding any provision to the contrary within this Reinsurance Contract; this Reinsurance Contract excludes any loss; damage; liability; claim; cost or expense of whatsoever nature; directly or indirectly caused by; contributed to by; resulting from; arising out of; or in connection with a Communicable Disease or the fear or threat (whether actual or perceived) of a Communicable Disease regardless of any other cause or event contributing concurrently or in any other sequence thereto.`	`79`
CYBER LOSS LIMITED EXCLUSION CLAUSE (PROPERTY TREATY REINSURANCE) Based on LMA 5410 - Amended to clarify consistency of coverage in the write-back 1. Notwithstanding any provision to the contrary within this reinsurance agreement or any endorsement thereto; this reinsurance agreement excludes all loss; damage; liability; cost or expense of whatsoever nature directly or indirectly caused by; contributed to by; resulting from; arising out of or in connection with: 1.1 any loss of; alteration of; or damage to or a reduction in the functionality; availability or operation of a Computer System; unless subject to the provisions of paragraph 2; 1.2 any loss of use; reduction in functionality; repair; replacement; restoration or reproduction of any Data; including any amount pertaining to the value of such Data. 2. Subject to the other terms; conditions and exclusions contained in this reinsurance agreement; this reinsurance agreement will cover physical damage to property insured under the original policies and any Time Element Loss directly resulting therefrom where such physical damage is directly occasioned by a peril otherwise covered hereunder.	`88`

Loss: BatchAllTripletLoss

sscd

Dataset: sscd
Size: 100 evaluation samples
Columns: sentence and label

Approximate statistics based on the first 1000 samples:

	sentence	label
type	string	int
details	min: 5 tokens mean: 70.36 tokens max: 128 tokens	1: ~1.00% 3: ~8.00% 4: ~1.00% 5: ~1.00% 7: ~1.00% 8: ~2.00% 19: ~4.00% 25: ~1.00% 26: ~2.00% 29: ~2.00% 32: ~2.00% 33: ~2.00% 34: ~1.00% 38: ~1.00% 39: ~1.00% 54: ~3.00% 55: ~1.00% 68: ~1.00% 78: ~2.00% 80: ~1.00% 82: ~1.00% 84: ~1.00% 93: ~1.00% 98: ~1.00% 120: ~1.00% 134: ~1.00% 135: ~1.00% 143: ~1.00% 144: ~2.00% 149: ~1.00% 154: ~1.00% 161: ~1.00% 173: ~1.00% 180: ~1.00% 181: ~1.00% 183: ~2.00% 206: ~1.00% 236: ~1.00% 238: ~1.00% 239: ~1.00% 243: ~1.00% 244: ~1.00% 256: ~1.00% 264: ~1.00% 326: ~1.00% 361: ~1.00% 367: ~1.00% 374: ~1.00% 377: ~1.00% 429: ~1.00% 433: ~1.00% 443: ~1.00% 448: ~1.00% 473: ~1.00% 488: ~1.00% 521: ~1.00% 535: ~1.00% 556: ~1.00% 557: ~1.00% 580: ~1.00% 589: ~1.00% 679: ~1.00% 693: ~1.00% 797: ~1.00% 857: ~1.00% 859: ~1.00% 871: ~1.00% 873: ~1.00% 960: ~1.00% 979: ~1.00% 1028: ~1.00% 1155: ~1.00% 1209: ~1.00% 1213: ~1.00% 1256: ~1.00% 1297: ~1.00% 1331: ~1.00% 1481: ~1.00% 1528: ~1.00% 1541: ~1.00%

Samples:

sentence	label
“Communicable Disease” means any disease which can be transmitted by means of any substance or agent from any organism to another organism where: a. the substance or agent includes; but is not limited to; a virus; bacterium; parasite or other organism or any variation thereof; whether deemed living or not; and b. the method of transmission; whether direct or indirect; includes but is not limited to; airborne transmission; bodily fluid transmission; transmission from or to any surface or object; solid; liquid or gas or between organisms; and c. the disease; substance or agent can cause or threaten damage to human health or human welfare or can cause or threaten damage to; deterioration of; loss of value of; marketability of or loss of use of property.	`4`
`“Production; Use or Storage of Nuclear Material” means the production; manufacture; enrichment; conditioning; processing; reprocessing; use; storage; handling and disposal of Nuclear Material.`	`25`
`means information; facts; concepts; code or any other information of any kind that is recorded or transmitted in a form to be used; accessed; processed; transmitted or stored by a Computer System.`	`7`

Loss: BatchAllTripletLoss

mlmc

Dataset: mlmc
Size: 100 evaluation samples
Columns: anchor and positive
Approximate statistics based on the first 1000 samples:
anchor positive
type string string
details
min: 20 tokens
mean: 104.63 tokens
max: 128 tokens

min: 22 tokens
mean: 112.7 tokens
max: 128 tokens

	anchor	positive
type	string	string
details	min: 20 tokens mean: 104.63 tokens max: 128 tokens	min: 22 tokens mean: 112.7 tokens max: 128 tokens

Samples:

anchor	positive
`Article 26 - Federal Excise Tax (BRMA 17D) A.The Reinsurer has agreed to allow for the purpose of paying the Federal Excise Tax the applicable percentage of the premium payable hereon (as imposed under Section 4371 of the Internal Revenue Code) to the extent such premium is subject to the Federal Excise Tax.`	`Article 26 - [MASK] Excise Tax (BRMA 17D) A.The Reinsurer has agreed to allow for the purpose of paying the Federal Excise Tax the applicable percentage of the [MASK] payable hereon (as [MASK] under [MASK] 4371 of the [MASK] Revenue Code) to [MASK] extent such premium is subject to the Federal Excise Tax.`
`Notwithstanding any provision to the contrary within this Reinsurance Contract; this Reinsurance Contract excludes any loss; damage; liability; claim; cost or expense of whatsoever nature; directly or indirectly caused by; contributed to by; resulting from; arising out of; or in connection with a Communicable Disease or the fear or threat (whether actual or perceived) of a Communicable Disease regardless of any other cause or event contributing concurrently or in any other sequence thereto.`	`Notwithstanding any provision to the contrary within [MASK] Reinsurance Contract; this Reinsurance Contract excludes [MASK] loss; damage; liability; [MASK] cost or expense of whatsoever nature; directly or indirectly [MASK] by; [MASK] to by; [MASK] from; arising out of; or in connection with a [MASK] Disease or the fear or threat (whether actual or perceived) of a Communicable Disease regardless of any other cause or event contributing [MASK] or in any [MASK] sequence thereto.`
`CYBER LOSS LIMITED EXCLUSION CLAUSE (PROPERTY TREATY REINSURANCE)`
Based on LMA 5410 - Amended to clarify consistency of coverage in the write-back
1. Notwithstanding any provision to the contrary within this reinsurance agreement or any endorsement thereto; this reinsurance agreement excludes all loss; damage; liability; cost or expense of whatsoever nature directly or indirectly caused by; contributed to by; resulting from; arising out of or in connection with:
1.1 any loss of; alteration of; or damage to or a reduction in the functionality; availability or operation of a Computer System; unless subject to the provisions of paragraph 2;
1.2 any loss of use; reduction in functionality; repair; replacement; restoration or reproduction of any Data; including any amount pertaining to the value of such Data.
2. Subject to the other terms; conditions and exclusions contained in this reinsurance agreement; this reinsurance agreement will cover physical damage to property insured under the original policies and any Time Element Loss directly resulting therefrom where such physical damage is directly occasioned by a peril otherwise covered hereunder.	CYBER LOSS LIMITED EXCLUSION [MASK] (PROPERTY TREATY REINSURANCE) Based on [MASK] [MASK] - Amended to clarify [MASK] of coverage in the [MASK] 1. [MASK] any provision to the contrary within this reinsurance agreement or any endorsement thereto; this reinsurance agreement excludes all loss; damage; [MASK] cost or expense of whatsoever nature [MASK] or indirectly caused by; contributed to by; resulting from; arising out of or in connection with: 1.1 any loss of; alteration of; or damage to or a reduction in [MASK] functionality; availability or operation of a Computer System; unless subject to the provisions of paragraph 2; [MASK] any loss of use; reduction in functionality; repair; replacement; restoration or reproduction of any Data; including any amount pertaining to the [MASK] of [MASK] [MASK] 2. Subject to the other terms; conditions and exclusions contained in this reinsurance agreement; this reinsurance agreement [MASK] cover physical [MASK] to property insured under the original policies and any Time Element [MASK] directly resulting therefrom where such physical damage is directly occasioned by a [MASK] [MASK] covered hereunder.

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

mlmd

Dataset: mlmd
Size: 100 evaluation samples
Columns: anchor and positive
Approximate statistics based on the first 1000 samples:
anchor positive
type string string
details
min: 5 tokens
mean: 70.36 tokens
max: 128 tokens

min: 5 tokens
mean: 78.55 tokens
max: 128 tokens

	anchor	positive
type	string	string
details	min: 5 tokens mean: 70.36 tokens max: 128 tokens	min: 5 tokens mean: 78.55 tokens max: 128 tokens

Samples:

anchor	positive
“Communicable Disease” means any disease which can be transmitted by means of any substance or agent from any organism to another organism where: a. the substance or agent includes; but is not limited to; a virus; bacterium; parasite or other organism or any variation thereof; whether deemed living or not; and b. the method of transmission; whether direct or indirect; includes but is not limited to; airborne transmission; bodily fluid transmission; transmission from or to any surface or object; solid; liquid or gas or between organisms; and c. the disease; substance or agent can cause or threaten damage to human health or human welfare or can cause or threaten damage to; deterioration of; loss of value of; marketability of or loss of use of property.	“Communicable Disease” means any disease which can be transmitted by [MASK] of [MASK] [MASK] or agent from any organism to another organism where: a. the substance or agent includes; [MASK] is not [MASK] to; a virus; bacterium; parasite or other [MASK] or [MASK] variation thereof; whether deemed living or [MASK] and b. the method of transmission; [MASK] direct or indirect; includes [MASK] is not [MASK] to; airborne transmission; bodily fluid transmission; transmission from or to any [MASK] or object; solid; liquid or gas or between organisms; [MASK] c. the [MASK] substance or agent can cause or threaten damage to human health or human [MASK] or can cause or [MASK] damage to; deterioration of; loss of value of; marketability of or [MASK] of use of property.
`“Production; Use or Storage of Nuclear Material” means the production; manufacture; enrichment; conditioning; processing; reprocessing; use; storage; handling and disposal of Nuclear Material.`	`“Production; Use or Storage of Nuclear Material” means the production; manufacture; [MASK] conditioning; processing; reprocessing; use; [MASK] [MASK] and disposal of Nuclear Material.`
`means information; facts; concepts; code or any other information of any kind that is recorded or transmitted in a form to be used; accessed; processed; transmitted or stored by a Computer System.`	`[MASK] information; facts; concepts; [MASK] or any other information of any kind [MASK] is recorded or transmitted in a [MASK] to be used; accessed; processed; transmitted or stored by a Computer System.`

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

apnc

Dataset: apnc
Size: 100 evaluation samples
Columns: anchor, positive, and negative

Approximate statistics based on the first 1000 samples:

	anchor	positive	negative
type	string	string	string
details	min: 20 tokens mean: 104.63 tokens max: 128 tokens	min: 20 tokens mean: 104.63 tokens max: 128 tokens	min: 20 tokens mean: 104.63 tokens max: 128 tokens

Samples:

anchor	positive	negative
`Article 26 - Federal Excise Tax (BRMA 17D) A.The Reinsurer has agreed to allow for the purpose of paying the Federal Excise Tax the applicable percentage of the premium payable hereon (as imposed under Section 4371 of the Internal Revenue Code) to the extent such premium is subject to the Federal Excise Tax.`	`Article 26 - Federal Excise Tax (BRMA 17D) A.The Reinsurer has agreed to allow for the purpose of paying the Federal Excise Tax the applicable percentage of the premium payable hereon (as imposed under Section 4371 of the Internal Revenue Code) to the extent such premium is subject to the Federal Excise Tax.`	ARTICLE XXXII NON-ASSIGNABILITY A. The Reinsurer shall not reinsure or otherwise assign or transfer its entire liability or obligations under this Contract without the Company’s prior written consent. B. The Reinsurer shall not transfer its claims-paying authority under this Contract to an unaffiliated entity or in any other way assign its interests or delegate its obligations under this Contract to an unaffiliated entity without the Company’s prior written consent. Notwithstanding the foregoing; the transfer of claims-paying authority or administration to a third party; where the subscribing reinsurer maintains control over claims settlement decisions; will not constitute a transfer of its claims-paying authority for purposes of this subparagraph.
`Notwithstanding any provision to the contrary within this Reinsurance Contract; this Reinsurance Contract excludes any loss; damage; liability; claim; cost or expense of whatsoever nature; directly or indirectly caused by; contributed to by; resulting from; arising out of; or in connection with a Communicable Disease or the fear or threat (whether actual or perceived) of a Communicable Disease regardless of any other cause or event contributing concurrently or in any other sequence thereto.`	`Notwithstanding any provision to the contrary within this Reinsurance Contract; this Reinsurance Contract excludes any loss; damage; liability; claim; cost or expense of whatsoever nature; directly or indirectly caused by; contributed to by; resulting from; arising out of; or in connection with a Communicable Disease or the fear or threat (whether actual or perceived) of a Communicable Disease regardless of any other cause or event contributing concurrently or in any other sequence thereto.`	`The obligations and duties of a Subscribing Reinsurer under this Contract shall not be assigned to or assumed by another reinsurer without the prior written consent of the Company.`
CYBER LOSS LIMITED EXCLUSION CLAUSE (PROPERTY TREATY REINSURANCE) Based on LMA 5410 - Amended to clarify consistency of coverage in the write-back 1. Notwithstanding any provision to the contrary within this reinsurance agreement or any endorsement thereto; this reinsurance agreement excludes all loss; damage; liability; cost or expense of whatsoever nature directly or indirectly caused by; contributed to by; resulting from; arising out of or in connection with: 1.1 any loss of; alteration of; or damage to or a reduction in the functionality; availability or operation of a Computer System; unless subject to the provisions of paragraph 2; 1.2 any loss of use; reduction in functionality; repair; replacement; restoration or reproduction of any Data; including any amount pertaining to the value of such Data. 2. Subject to the other terms; conditions and exclusions contained in this reinsurance agreement; this reinsurance agreement will cover physical damage to property insured under the original policies and any Time Element Loss directly resulting therefrom where such physical damage is directly occasioned by a peril otherwise covered hereunder.	CYBER LOSS LIMITED EXCLUSION CLAUSE (PROPERTY TREATY REINSURANCE) Based on LMA 5410 - Amended to clarify consistency of coverage in the write-back 1. Notwithstanding any provision to the contrary within this reinsurance agreement or any endorsement thereto; this reinsurance agreement excludes all loss; damage; liability; cost or expense of whatsoever nature directly or indirectly caused by; contributed to by; resulting from; arising out of or in connection with: 1.1 any loss of; alteration of; or damage to or a reduction in the functionality; availability or operation of a Computer System; unless subject to the provisions of paragraph 2; 1.2 any loss of use; reduction in functionality; repair; replacement; restoration or reproduction of any Data; including any amount pertaining to the value of such Data. 2. Subject to the other terms; conditions and exclusions contained in this reinsurance agreement; this reinsurance agreement will cover physical damage to property insured under the original policies and any Time Element Loss directly resulting therefrom where such physical damage is directly occasioned by a peril otherwise covered hereunder.	The Reinsurer agrees not to disclose any Confidential Information which it may acquire in connection with this Contract; except: 1. To its professional advisors; auditors; attorneys; and other consultants on a need-to-know basis; 2. To any of its affiliates and to the directors; officers; employees; professional advisors; auditors; attorneys; and other consultants of such affiliates on a need-to-know basis; 3. To any other party to whom such disclosure is necessary for the Reinsurer to enforce its rights hereunder; 4. To any party from whom the Reinsurer is seeking or from whom the Reinsurer has obtained reinsurance as long as the disclosing party is under a similar obligation of confidentiality as the Reinsurer is under this Article; 5. When required for the Reinsurer's internal operations. Further; the Reinsurer agrees not to use any Confidential Information for any purpose not related to the performance of its obligations or enforcement of its rights under this Contract. Notwithstanding the above; in the event that the Reinsurer is required by court order; other legal process or any regulatory authority to release or disclose any or all of the Confidential Information; the Reinsurer agrees to provide the Company with written notice of same at least 10 days prior to such release or disclosure and to use its best efforts to assist the Company in maintaining the confidentiality provided for in this Article. The provisions of this Article shall apply to renewal information provided to the Reinsurer by the Company prior to or upon the expiration or termination of this Contract.

Loss: CachedMultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

apnd

Dataset: apnd
Size: 100 evaluation samples
Columns: anchor, positive, and negative

Approximate statistics based on the first 1000 samples:

	anchor	positive	negative
type	string	string	string
details	min: 5 tokens mean: 70.36 tokens max: 128 tokens	min: 5 tokens mean: 70.36 tokens max: 128 tokens	min: 5 tokens mean: 71.3 tokens max: 128 tokens

Samples:

anchor	positive	negative
“Communicable Disease” means any disease which can be transmitted by means of any substance or agent from any organism to another organism where: a. the substance or agent includes; but is not limited to; a virus; bacterium; parasite or other organism or any variation thereof; whether deemed living or not; and b. the method of transmission; whether direct or indirect; includes but is not limited to; airborne transmission; bodily fluid transmission; transmission from or to any surface or object; solid; liquid or gas or between organisms; and c. the disease; substance or agent can cause or threaten damage to human health or human welfare or can cause or threaten damage to; deterioration of; loss of value of; marketability of or loss of use of property.	“Communicable Disease” means any disease which can be transmitted by means of any substance or agent from any organism to another organism where: a. the substance or agent includes; but is not limited to; a virus; bacterium; parasite or other organism or any variation thereof; whether deemed living or not; and b. the method of transmission; whether direct or indirect; includes but is not limited to; airborne transmission; bodily fluid transmission; transmission from or to any surface or object; solid; liquid or gas or between organisms; and c. the disease; substance or agent can cause or threaten damage to human health or human welfare or can cause or threaten damage to; deterioration of; loss of value of; marketability of or loss of use of property.	shall be defined as the sum of all losses directly occasioned by any one disaster; accident or loss or series of disasters; accidents or losses arising out of one event. The duration and extent of any one Loss Occurrence will be limited to all individual losses sustained by the Company occurring during any period of 168 consecutive hours (except as otherwise provided below) arising out of and directly occasioned by the same event except that the term Loss Occurrence will be further defined as follows: 1. As regards windstorm; hail; tornado; including ensuing collapse and water damage; all individual losses sustained by the Company occurring during any period of 168 consecutive hours arising out of and directly occasioned by the same event. Notwithstanding the foregoing; as respects Named Storm only; the period of consecutive hours applicable to such Named Storm may be extended beyond 168 hours in accordance with the provisions of paragraph E below. 2. As regards riot; riot attending a strike; civil commotion; vandalism and malicious mischief; all individual losses sustained by the Company occurring during any period of 96 consecutive hours arising out of and directly occasioned by the same event. The maximum duration of 96 consecutive hours may be extended in respect of individual losses which occur beyond such 96 consecutive hours during the continued occupation of an insured’s premises by strikers or locked-out workers; provided such occupation commenced during the aforesaid period. 3. As regards earthquake (the epicenter of which need not necessarily be within the territorial confines referred to in the Territory Article) and fire following directly occasioned by the earthquake; only those earthquake losses and individual fire losses which commence during the period of 168 consecutive hours may be included in any one Loss Occurrence. 4. As regards freezing; frost; ice; snow; sleet; including weight of snow; ice or sleet; collapse of buildings; breakage of glass and water damage (caused by bursting of frozen pipes and tanks or freezing and/or melting snow or sleet; including but not limited to ice dams); as well as other perils; all individual losses sustained by the Company occurring during any period of 72 consecutive hours arising out of and directly occasioned by the same event.
`“Production; Use or Storage of Nuclear Material” means the production; manufacture; enrichment; conditioning; processing; reprocessing; use; storage; handling and disposal of Nuclear Material.`	`“Production; Use or Storage of Nuclear Material” means the production; manufacture; enrichment; conditioning; processing; reprocessing; use; storage; handling and disposal of Nuclear Material.`	`“Contract” shall be understood to mean “Contract;” “Policy” or whatever other term is used to designate the attached reinsurance document.`
`means information; facts; concepts; code or any other information of any kind that is recorded or transmitted in a form to be used; accessed; processed; transmitted or stored by a Computer System.`	`means information; facts; concepts; code or any other information of any kind that is recorded or transmitted in a form to be used; accessed; processed; transmitted or stored by a Computer System.`	`13. Self-Insurance applicable to section D (Eureko Sigorta A.Ş;) of the Risk details`

Loss: CachedMultipleNegativesRankingLoss with these parameters:

{
    "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
learning_rate: 3e-05
num_train_epochs: 10
warmup_ratio: 0.1

All Hyperparameters

Click to expand

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: 3e-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: 10
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

Training Logs

Epoch	Step	Training Loss	mlmd loss	sscc loss	apnd loss	mlmc loss	apnc loss	sscd loss
3.2051	500	0.0016	0.0480	3.8732	0.0902	0.0331	0.0847	3.0775
6.4103	1000	0.0013	0.0411	3.7545	0.0902	0.0346	0.0862	3.0353
9.6154	1500	0.001	0.0390	3.8221	0.0902	0.0162	0.0839	3.0140

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.3.0+cu121
Accelerate: 0.32.1
Datasets: 2.20.0
Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

@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",
}

BatchAllTripletLoss

@misc{hermans2017defense,
    title={In Defense of the Triplet Loss for Person Re-Identification}, 
    author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
    year={2017},
    eprint={1703.07737},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

MultipleNegativesRankingLoss

@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}
}

CachedMultipleNegativesRankingLoss

@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}
}