Tejasw1/bge-base-legal-matryoshka-v1

Jina-v3 Legal Matryoshka

This is a sentence-transformers model finetuned from BAAI/bge-base-en-v1.5 on the json 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: BAAI/bge-base-en-v1.5
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 tokens
• Similarity Function: Cosine Similarity
• Training Dataset:
  • json
• Language: en
• License: apache-2.0

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': 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:

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("Tejasw1/bge-base-legal-matryoshka-v1")
# Run inference
sentences = [
    'The Board may, for ensuring that an information utility takes into account the objectives sought to be achieved under this Code, require every information utility to set up a governing board, with such number of independent members, as may be specified by regulations.',
    'How many independent members are required on the governing board of an information utility according to Section 212 of The Insolvency and Bankruptcy Code, 2016?',
    'How can the government clarify the scope or applicability of a notification or order issued under Section 11(1) or 11(2) of the Central Goods and Services Tax Act, 2017?',
]
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]

Evaluation

Metrics

Information Retrieval

• Dataset: dim_768
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.0101
cosine_accuracy@3	0.1241
cosine_accuracy@5	0.3419
cosine_accuracy@10	0.6103
cosine_precision@1	0.0101
cosine_precision@3	0.0414
cosine_precision@5	0.0684
cosine_precision@10	0.061
cosine_recall@1	0.0101
cosine_recall@3	0.1241
cosine_recall@5	0.3419
cosine_recall@10	0.6103
cosine_ndcg@10	0.2494
cosine_mrr@10	0.141
cosine_map@100	0.1568

Information Retrieval

• Dataset: dim_512
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.0129
cosine_accuracy@3	0.1204
cosine_accuracy@5	0.3318
cosine_accuracy@10	0.602
cosine_precision@1	0.0129
cosine_precision@3	0.0401
cosine_precision@5	0.0664
cosine_precision@10	0.0602
cosine_recall@1	0.0129
cosine_recall@3	0.1204
cosine_recall@5	0.3318
cosine_recall@10	0.602
cosine_ndcg@10	0.2466
cosine_mrr@10	0.1401
cosine_map@100	0.1563

Training Details

Training Dataset

json

• Dataset: json
• Size: 9,791 training samples
• Columns: positive and anchor
• Approximate statistics based on the first 1000 samples:

  	positive	anchor
  type	string	string
  details	min: 2 tokens mean: 320.43 tokens max: 512 tokens	min: 14 tokens mean: 33.12 tokens max: 66 tokens

• Samples:

  positive	anchor
  Show Related Subordinates (1) The Central Government may, by notification in the Official Gazette and in the Electronic Gazette, make rules to carry out the provisions of this Act. (2) In particular, and without prejudice to the generality of the foregoing power, such rules may provide for all or any of the following matters, namely:-- 1 [(a) the conditions for considering reliability of electronic signature or electronic authentication technique under sub-section (2) of section 3A; (aa) the procedure for ascertaining electronic signature or authentication under sub-section (3) of section 3A; (ab) the manner in which any information or matter may be authenticated by means of electronic signature under section 5;] (b) the electronic form in which filing, issue, grant or payment shall be effected under sub-section (1) of section 6; (c) the manner and format in which electronic records shall be filed, or issued and the method of payment under sub-section (2) of section 6; 2 [(ca) the manner in which the authorised service provider may collect, retain and appropriate service charges under sub-section (2) of section 6A;] (d) the matters relating to the type of 3 [electronic signature], manner and format in which it may be affixed under section 10; 4 [(e) the manner of storing and affixing electronic signature creation data under section 15; (ea) the security procedures and practices under section 16;] (f) the qualifications, experience and terms and conditions of service of Controller, Deputy Controllers 5 [, Assistant Controllers, other officers and employees] under section 17; 6 * * * * * (h) the requirements which an applicant must fulfil under sub-section (2) of section 21; (i) the period of validity of licence granted under clause (a) of sub-section (3) of section 21; (j) the form in which an application for licence may be made under sub-section (1) of section 22; (k) the amount of fees payable under clause (c) of sub-section (2) of section 22; (l) such other documents which shall accompany an application for licence under clause (d) of subsection (2) of section 22; (m) the form and the fee for renewal of a licence and the fee payable thereof under section 23; 7 [(ma) the form of application and fee for issue of Electronic Signature Certificate under section 35;] (n) the form in which application for issue of a 3 [electronic signature] Certificate may be made under sub-section (1) of section 35; (o) the fee to be paid to the Certifying Authority for issue of a 3 [electronic signature] Certificate under sub-section (2) of section 35; 2 [(oa) the duties of subscribers under section 40A; (ob) the reasonable security practices and procedures and sensitive personal data or information under section 43A;] (p) the manner in which the adjudicating officer shall hold inquiry under sub-section (1) of section 46; (q) the qualification and experience which the adjudicating officer shall possess under sub-section (3) of section 46; 8 * * * * (u) the form in which appeal may be filed and the fee thereof under sub-section (3) of section 57; (v) any other power of a civil court required to be prescribed under clause (g) of sub-section (2) of section 58; and 9 [(w) the powers and functions of the Chairperson of the 10 [Appellate Tribunal] under section 52A; (x) the information, duration, manner and form of such information to be retained and preserved under section 67C; (y) the procedures and safeguards for interception, monitoring or decryption under sub-section (2) of section 69; (z) the procedures and safeguards for blocking for access by the public under sub-section (3) of section 69 A; (za) the procedure and safeguards for monitoring and collecting traffic data or information under sub-section (3) of section 69B; (zb) the information security practices and procedures for protected system under section 70; (zc) manner of performing functions and duties of the agency under sub-section (3) of section 70A; (zd) the officers and employees under sub-section (2) of section 70B; (ze) salaries and allowances and terms and conditions of service of the Director General and other officers and employees under sub-section (3) of section 70B; (zf) the manner in which the functions and duties of agency shall be performed under sub-section (5) of section 70B; (zg) the guidelines to be observed by the intermediaries under sub-section (2) of section 79; (zh) the modes or methods for encryption under section 84A.] (3) 11 [Every notification made by the Central Government under sub-section (1) of section 70A and every rule made by it] shall be laid, as soon as may be after it is made, before each House of Parliament, while it is in session, for a total period of thirty days which may be comprised in one session or in two or more successive sessions, and if, before the expiry of the session immediately following the session or the successive sessions aforesaid, both Houses agree in making any modification in 12 *** the rule or both Houses agree that 12 *** the rule should not be made, 12 *** the rule shall thereafter have effect only in such modified form or be of no effect, as the case may be; so, however, that any such modification or annulment shall be without prejudice to the validity of anything previously done under that notification or rule. 1 Subs. by s. 46, ibid., for clause (a) (w.e.f. 27-10-2009). 2. Ins. by Act 10 of 2009, s. 46 (w.e.f. 27-10-2009). 3. Subs. by s. 5, ibid., for "digital signature" (w.e.f. 27-10-2009). 4. Subs. by s. 46, ibid., for clause (e) (w.e.f. 27-10-2009). 5. Subs. by s. 46, ibid., for "and Assistant Controllers" (w.e.f. 27-10-2009). 6. Clause (g) omitted by s. 46, ibid. (w.e.f. 27-10-2009). 7. Ins. by s. 46, ibid. (w.e.f. 27-10-2009). 8. Clauses (r), (s) and (t) omitted by Act 7 of 2017, s. 169 (w.e.f. 26-5-2017). 9. Subs. by Act 10 of 2009, s. 46, for clause (w) (w.e.f. 27-10-2009). 10. Subs. by Act 7 of 2017, s. 169, for "Cyber Appellate Tribunal"(w.e.f. 26-5-2017). 11. Subs. by Act 10 of 2009, s. 46, for certain words, brackets, letter and figures (w.e.f. 27-10-2009). 12. The words "the notification or" omitted by s. 46, ibid. (w.e.f. 27-10-2009). Rules Regulations Notifications YearDescriptionHindi DescriptionFiles(Eng)Files(Hindi)17-10-2000 Information Technology (Certifying Authorities) Rules, 2000. 22-10-2002 The Cyber Regulations Appellate Tribunal (Salaries, Allowances and other condition of Service of other Officers and Employees) Rules, 2002. 17-03-2003 Cyber Regulations Appellate Tribunal (Salary, Allowances and other terms and conditions of service of Presiding Officer) Rules, 2003. 17-03-2003 Information Technology (Qualification and Experience of Adjudicating Officers and Manner of Holding Enquiry) Rules, 2003. 21-03-2003 Information Technology (Other Standards) Rules, 2003 21-11-2003 Cyber Regulations Appellate Tribunal (Procedure for Investigation of misbehaviour or Incapacity of Presiding Officer) Rules, 2003. 21-11-2003 The Information Technology (Other Powers of Civil Court vested in cyber appellate tribunal) Rules, 2003 21-11-2003 (Certifying Authorities) Amendment Rules, 2003 - Rules 8, 14 20-08-2004 The Information Technology (Certifying Authorities) (Amendment) Rules, 2004 06-09-2004 Information Technology (Use of Electronic Records and Digital Signature) Rules, 2004 29-10-2004 Information Technology (Security Procedure) Amendment Rules, 2004 06-08-2009 Certifying Authorities) Amendment Rules, 2009 -Rule 6 08-09-2009 Controller of Certifying Authorities Gazetted & Non-Gazetted posts Recruitment Rules, 2009 27-10-2009 Information Technology (Procedures and Safeguards for Interception, Monitoring or Decryption of Information) Rules, 2009. 27-10-2009 Information Technology (Procedure and Safeguards for Blocking for Access of Information by Public) Rules, 2009 27-10-2009 Cyber Appellate Tribunal (Salary, Allowances and Other Terms and Conditions of Service of Chairperson and Members) Rules, 2009 27-10-2009 Cyber Appellate Tribunal (Procedure for Investigation of Misbehaviour or Incapacity of Chairperson and Members) Rules, 2009 27-10-2009 Information Technology (Procedure and Safeguards for Monitoring and Collecting Traffic Data or Information) Rules, 2009 11-04-2011 Information Technology(Guidelines for Cyber Cafe Rules) Rules, 2011 11-04-2011 Information Technology (Electronic Service Delivery) Rules 2011 11-04-2011 The Information Technology (Reasonable Security Practices and Procedures and Sensitive Personal Data or Information) Rules, 2011. 11-04-2011 Information Technology (Intermediaries Guidelines) Rules, 2011. 25-10-2011 (Certifying Authorities) Amendment Rules, 2011 -Rules 4,5A and 6 24-01-2013 The salary, allowances and the other terms and conditions of service of the Director General, CERT In, Rules 2012 24-01-2013 Cyber Appellate Tribunal (Salary, Allowances and Other Terms and Conditions of Service of Chairperson and Members) Amendment Rules, 2012 16-08-2013 Controller of Certifying Authorities Director (Finance and Administrator) Recruitment Rules 2013 16-01-2014 Information Technology (National Critical Information Infrastructure Protection Centre and Manner of Performing Functions and Duties) Rules, 2013 16-01-2014 Information Technology (The Indian Computer Emergency Response Team and Manner of Performing Functions and Duties) Rules, 2013. 25-08-2015 Digital Signature (End Entity) Rules, 2015 25-08-2015 Information Technology (Security Procedures) Amendment Rules, 2015 25-08-2015 Information Technology (Certifying Authorities) Amendment Rules, 2015 16-05-2016 Cyber Appellate Tribunal (Powers and Functions of the Chairperson) Rules, 2016. 21-07-2016 Information Technology (Preservation and Retention of Information by Intermediaries Providing Digital Locker Facilities) Rules, 2016 08-02-2017 Information Technology (Preservation and Retention of Information by Intermediaries Providing Digital Locker Facilities) Amendment Rules, 2017 20-11-2017 Recruitment Rules for Group A and Group B posts of CERT-In, MeitY 22-05-2018 The Information Security Practices and Procedures for Protected System Rules, 2018 under section 70 of the IT Act 2000 25-02-2021 Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. 28-10-2022 Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2022. 06-04-2023 Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 updated as on 6.4.2023 YearDescriptionHindi DescriptionFiles(Eng)Files(Hindi)09-07-2001 INFORMATION TECHNOLOGY (CERTIFYING AUTHORITY) REGULATIONS, 2001 YearDescriptionHindi DescriptionFiles(Eng)Files(Hindi)16-05-2016 Powers and Functions of Chairperson, CyAT 26-02-2021 Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021.	What procedures and safeguards should be implemented for blocking access by the public under sub-section (3) of Section 69A of the Information Technology Act, 2000?
  Show Related Subordinates (1) The Commission may, by notification, make regulations consistent with this Act and the rules made thereunder to carry out the purposes of this Act. (2) In particular, and without prejudice to the generality of the foregoing provisions, such regulations may provide for all or any of the following matters, namely:-- (a) the cost of production to be determined under clause (b) of the Explanation to section 4; (b) the form of notice as may be specified and the fee which may be determined under sub-section (2) of section 6; 1[(c) the manner of determination of substantial business operations in India under clause (d) of section 5; (ca) the form and fee for notice for combination under sub-section (4) of section 6; (cb) the time and manner for filing notice of acquisition under clause (a) of section 6A; (cc) the manner and circumstance in which the acquirer may exercise the ownership or beneficial right or interest in shares or convertible securities including voting right and receipt of dividends or any other distributions as an exception under clause (b) of section 6A;] 2[(d) the procedures to be followed for engaging the experts and professionals under sub-section (3) of section 17; (e) the fee which may be determined under clause (a) of sub-section (1) of section 19; (f) the rules of procedure in regard to the transaction of business at the meetings of the Commission under sub-section (1) of section 22; 1[(fa) other details to be indicated in the show-cause notice under sub-section (9) of section 26; (fb) the manner of determining turnover or income under theExplanation I to clause (b) of section 27; (fc) the manner in which modification may be proposed by parties to the combination to the Commission under sub-section (2) of section 29A;] (g) the manner in which penalty shall be recovered under sub-section (1) of section 39; 1[(ga) the lesser penalty to be imposed on producer, seller, distributor, trader or service provider under sub-section (1) of section 46; (gb) the manner and time for withdrawal of application for lesser penalty under sub-section (2) of section 46; (gc) the lesser penalty to be imposed on producer, seller, distributor, trader or service provider under sub-section (4) of section 46; (gd) the manner of determining income under clause (c) of Explanation to section 48; (ge) the form of application and fee under sub-section (1), the time under sub-section (2), the terms and manner of implementations and monitoring under sub-section (3) and the procedure for conducting settlement proceedings under sub-section (6) of section 48A; (gf) the form of application and fee under sub-section (1), the time under sub-section (2), the terms and manner of implementations and monitoring under sub-section (3) and the procedure for commitments offered under sub-section (6) of section 48B; (gg) the other details to be published along with draft regulations and the period for inviting public comments under clause (a) of section 64A;] (h) any other matter in respect of which provision is to be, or may be, made by regulations.] (3) Every regulation made under this Act shall be laid, as soon as may be after it is made, before each House of Parliament, while it is in session, for a total period of thirty days which may be comprised in one session or in two or more successive sessions, and if, before the expiry of the session immediately following the session or the successive sessions aforesaid, both Houses agree in making any modification in the regulation, or both Houses agree that the regulation should not be made, the regulation shall thereafter have effect only in such modified form or be of no effect, as the case may be; so, however, that any such modification or annulment shall be without prejudice to the validity of anything previously done under that regulation. 1. Subs. by Act 9 of 2023, s. 44, for clause (c) (w.e.f. 18-5-2023). 2. Subs. by Act 39 of 2007, s. 49, for clauses (d) and (e) (w.e.f. 12-10-2007). Notifications YearDescriptionHindi DescriptionFiles(Eng)Files(Hindi)11-12-2013 SO.3641(E) dated December 11,2013	How is income determined under Explanation to Section 48(c) of the Competition Act, 2002?
  (1) The Board may, by notification, make regulations consistent with this Code and the rules made thereunder, to carry out the provisions of this Code. (2) In particular, and without prejudice to the generality of the foregoing power, such regulations may provide for all or any of the following matters, namely:-- (a) the form and the manner of accepting electronic submission of financial information under sub-clause (a) of clause (9) of section 3; (b) the persons to whom access to information stored with the information utility may be provided under sub-clause (d) of clause (9) of section 3; (c) the other information under sub-clause (f) of clause (13) of section 3; (d) the other costs under clause (e) of sub-section (13) of section 5; (e) the cost incurred by the liquidator during the period of liquidation which shall be liquidation cost under sub-section (16) of section 5; 1[(ea) the other costs under sub-clause (e) of clause (23C) of section 5;] (f) the other record or evidence of default under clause (a), and any other information under clause (c), of sub-section (3) of section 7; 2* * * * * (h) the period under clause (a) of sub-section (3) of section 10; (i) the supply of essential goods or services to the corporate debtor under sub-section (2) of section 14; 3[(ia) circumstances in which supply of critical goods or services may be terminated, suspended or interrupted during the period of moratorium under sub-section (2A) of section 14;] (j) the manner of making public announcement under sub-section (2) of section 15; 4[(ja) the last date for submission of claims under clause (c) of sub-section (1) of section 15;] (k) the manner of taking action and the restrictions thereof under clause (b) of sub-section (2) of section 17; (l) the other persons under clause (d) of sub-section (2) of section 17; (m) the other matters under clause (d) of sub-section (2) of section 17; (n) the other matters under sub-clause (iv) of clause (a), and the other duties to be performed by the interim resolution professional under clause (g), of section 18; 4[(na) the number of creditors within a class of creditors under clause (b) of sub-section (6A) of section 21; (nb) the remuneration payable to authorised representative under clause (ii) of the proviso to sub-section (6B) of section 21; (nc) the manner of voting and determining the voting share in respect of financial debts under sub-section (7) of section 21;] (o) the persons who shall comprise the committee of creditors, the functions to be exercised such committee and the manner in which functions shall be exercised under the proviso to sub-section (8) of section 21; (p) the other electronic means by which the members of the committee of creditors may meet under sub-section (1) of section 24; (q) the manner of assigning voting share to each creditor under sub-section (7) of section 24; (r) the manner of conducting the meetings of the committee of creditors under sub-section (8) of section 24; (s) the manner of appointing accountants, lawyers and other advisors under clause (d) of sub-section (2) of section 25; 5[(sa) other conditions under clause (h) of sub-section (2) of section 25;] (t) the other actions under clause (k) of sub-section (2) of section 25; (u) the form and the manner in which an information memorandum shall be prepared by the resolution professional sub-section (1) of section 29; (v) the other matter pertaining to the corporate debtor under the Explanation to sub-section (2) of section 29; (w) the manner of making payment of insolvency resolution process costs under clause (a), the manner of 6[payment of debts] under clause (b), and the other requirements to which a resolution plan shall conform to under clause (d), of sub-section (2) of section 30; 7[(wa) other requirements under sub-section (4) of section 30;]. (x) the fee for the conduct of the liquidation proceedings and proportion to the value of the liquidation estate assets under sub-section (8) of section 34; (y) the manner of evaluating the assets and property of the corporate debtor under clause (c), the manner of selling property in parcels under clause (f), the manner of reporting progress of the liquidation process under clause (n), and the other functions to be performed under clause (o), of sub-section (1) of section 35; (z) the manner of making the records available to other stakeholders under sub-section (2) of section 35; (za) the other means under clause (a) of sub-section (3) of section 36; (zb) the other assets under clause (e) of sub-section (4) of section 36; (zc) the other source under clause (g) of sub-section (1) of section 37; (zd) the manner of providing financial information relating to the corporate debtor under sub-section (2) of section 37; (ze) the form, the manner and the supporting documents to be submitted by operational creditor to prove the claim under sub-section (3) of section 38; (zf) the time within which the liquidator shall verify the claims under sub-section (1) of section 39; (zg) the manner of determining the value of claims under section 41; (zh) the manner of relinquishing security interest to the liquidation estate and receiving proceeds from the sale of assets by the liquidator under clause (a), and the manner of realising security interest under clause (b) of sub-section (1) of section 52; (zi) the other means under clause (b) of sub-section (3) of section 52; (zj) the manner in which secured creditor shall be paid by the liquidator under sub-section (9) of section 52; (zk) the period and the manner of distribution of proceeds of sale under sub-section (1) of section 53; 8[(zka) such number of financial creditors and the manner of proposing the insolvency professional, and the form for approving such insolvency professional by the financial creditors under clause (e), the persons who shall provide approval under the proviso to clause (e), the form for making a declaration under clause (f) of sub-section (2) of section 54A; (zkb) the form for obtaining approval from financial creditors under sub-section (3), and the persons who shall provide approval under the proviso to sub-section (3) of section 54A; (zkc) the other conditions for the base resolution plan under clause (c), and such information and documents under clause (d) of sub-section (4) of section 54A; (zkd) the form in which the report is to be prepared under clause (a), such reports and other documents under clause (b), and such other duties under clause (c) of sub-section (1), and the manner of determining and bearing the fees in sub-section (3) of section 54B; (zke) the form for providing written consent of the insolvency professional under clause (b), the form for declaration under clause (c), the information relating to books of account and such other documents relating to such period under clause (d) of sub-section (3) of section 54C; (zkf) the form and manner for making application for termination of the pre-packaged insolvency resolution process under sub-section (3) of section 54D; (zkg) the form and manner of making public announcement under clause (c) of sub-section (1) of section 54E; (zkh) the manner of confirming the list of claims under clause (a), the manner of informing creditors under clause (b), the manner of maintaining an updated list of claims under clause (c), the form and manner of preparing the information memorandum under clause (g), and such other duties under clause (i) of sub-section (2) of section 54F; (zki) such other persons under clause (c), the manner of appointing accountants, legal or other professionals under clause (e), such other matters under sub-clause (iv) of clause (f) and the manner of taking other actions under clause (g) of sub-section (3) of section 54F; (zkj) the manner of determination of fees and expenses as may be incurred by the resolution professional under sub-section (6) of section 54F; (zkk) the manner of bearing fees and expenses under sub-section (7) of section 54F; (zkl) the form and manner of list of claims and preliminary information memorandum under sub-section (1) of section 54G; (zkm) the conditions under clause (a) of section 54H; (zkn) the manner of alteration of the composition of the committee of creditors under the proviso to sub-section (1) of section 54-I; (zko) the form and manner of making application under sub-section (1) of section 54J; (zkp) the manner of inviting prospective resolution applicants under sub-section (5) of section 54K; (zkq) the other conditions under sub-section (6) of section 54K; (zkr) the conditions under clause (a) and the manner of providing the basis for evaluation of resolution plans and the information referred to in section 29 under sub-section (7) of section 54K; (zks) the conditions under the proviso to sub-section (10) of section 54K; (zkt) the manner and conditions under sub-section (11) of section 54K; (zku) the form and manner of filing application under the proviso to sub-section (12) of section 54K; (zkv) the other requirements under sub-section (13) of section 54K; (zkw) the form for submission of written consent under clause (b) of sub-section (2) of section 54-O;] (zl) the other means under clause (a) and the other information under clause (b) of section 57; (zm) the conditions and procedural requirements under sub-section (2) of section 59; (zn) the details and the documents required to be submitted under sub-section (7) of section 95; (zo) the other matters under clause (c) of sub-section (3) of section 105; (zp) the manner and form of proxy voting under sub-section (4) of section 107; (zq) the manner of assigning voting share to creditor under sub-section (2) of section 109; (zr) the manner and form of proxy voting under sub-section (3) of section 133; (zs) the fee to be charged under sub-section (1) of section 144; (zt) the appointment of other officers and employees under sub-section (2), and the salaries and allowances payable to, and other terms and conditions of service of, such officers and employees of the Board under sub-section (3), of section 194; (zu) the other information under clause (i) of sub-section (1) of section 196; (zv) the intervals in which the periodic study, research and audit of the functioning and performance of the insolvency professional agencies, insolvency professionals and information utilities under clause (r), and mechanism for disposal of assets under clause (t), of sub-section (1) of section 196; (zw) the place and the time for discovery and production of books of account and other documents under clause (i) of sub-section (3) of section 196; (zx) the other committees to be constituted by the Board and the other members of such committees under section 197; (zy) the other persons under clause (b) and clause (d) of section 200; (zz) the form and the manner of application for registration, the particulars to be contained therein and the fee it shall accompany under sub-section (1) of section 201; (zza) the form and manner of issuing a certificate of registration and the terms and conditions thereof, under sub-section (3) of section 201; (zzb) the manner of renewal of the certificate of registration and the fee therefor, under sub-section (4) of section 201; (zzc) the other ground under clause (d) of sub-section (5) of section 201; (zzd) the form of appeal to the National Company Law Appellate Tribunal, the period within which it shall be filed under section 202; (zze) the other information under clause (g) of section 204; (zzf) the other grounds under Explanation to section 196; (zzg) the setting up of a governing board for its internal governance and management under clause (e), the curriculum under clause (l), the manner of conducting examination under clause (m) , of section 196; (zzh) the time within which, the manner in which, and the fee for registration of insolvency professional under sub-section (1) of section 207; (zzi) the categories of professionals or persons, the qualifications and experience and the fields under sub-section (2) of section 207; (zzj) the manner and the conditions subject to which the insolvency professional shall perform his function under clause (f) of sub-section (2) of section 208; (zzk) the form and manner in which, and the fee for registration of information utility under sub-section (1) of section 210; (zzl) the form and manner for issuing certificate of registration and the terms and conditions thereof, under sub-section (3) of section 210; (zzm) the manner of renewal of the certificate of registration and the fee therefor, under sub-section (4) of section 210; (zzn) the other ground under clause (d) of sub-section (5) of section 210; (zzo) the form, the period and the manner of filing appeal to the National Company Law Appellate Tribunal under section 211; (zzp) the number of independent members under section 212; (zzq) the services to be provided by information utility and the terms and conditions under section 213; (zzr) the form and manner of accepting electronic submissions of financial information under clause (b) and clause (c) of section 214; (zzs) the minimum service quality standards under clause (d) of section 214; (zzt) the information to be accessed and the manner of accessing such information under clause (f) of section 214; (zzu) the statistical information to be published under clause (g) of section 214; (zzv) the form, the fee and the manner for submitting or accessing information under sub-section (1) of section 215; (zzw) the form and manner for submitting financial information and information relating to assets under sub-section (2) of section 215; (zzx) the manner and the time within which financial information may be updated or modified or rectified under section 216; (zzy) the form, manner and time of filing complaint under section 217; (zzz) the time and manner of carrying out inspection or investigation under sub-section (2) of section 218; (zzza) the manner of carrying out inspection of insolvency professional agency or insolvency professional or information utility and the time for giving reply under section 219; (zzzb) the procedure for claiming restitution under sub-section (6), the period within which such restitution may be claimed and the manner in which restitution of amount may be made under sub-section (7) of section 220; (zzzc) the other funds of clause (c) of sub-section (1) of section 222. 1.Ins. by Act 26 of 2021, s. 16 (w.e.f. 4-4-2021). 2. Clause (g) omitted by Act 26 of 2018, s. 36 (w.e.f. 6-6-2018). 3. Ins. by Act 1 of 2020, s. 13 (w.e.f. 28-12-2019). 4. Ins. by s. 36, ibid. (w.e.f. 6-6-2018). 5. Ins. by Act 8 of 2018, s. 9 (w.e.f 23-11-2017). 6. Subs. by Act 26 of 2019, s. 9, for repayment of debts of operational creditors (w.e.f. 16-08-2019). 7. Ins. by Act 8 of 2018, s. 9 (w.e.f. 23-11-2017). 8. Ins. by Act 26 of 2021, s. 16 (w.e.f. 4-4-2021).	How should the value of claims be determined under section 41 of The Insolvency and Bankruptcy Code, 2016?

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512
      ],
      "matryoshka_weights": [
          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

Click to expand

• 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
• 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: 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
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	dim_512_cosine_map@100	dim_768_cosine_map@100
0.5229	10	2.0356	-	-
0.9935	19	-	0.1430	0.1489
1.0425	20	0.9141	-	-
1.5654	30	0.6628	-	-
1.9837	38	-	0.1485	0.1504
2.0850	40	0.5076	-	-
2.6078	50	0.4656	-	-
2.9739	57	-	-	0.1549
0.5229	10	0.3808	-	-
0.9935	19	-	0.1508	0.1575
1.0425	20	0.3038	-	-
1.5654	30	0.2403	-	-
1.9837	38	-	0.1539	0.1521
2.0850	40	0.2263	-	-
2.6078	50	0.2144	-	-
2.9739	57	-	0.1557	0.1586
3.1275	60	0.2121	-	-
3.6503	70	0.2114	-	-
3.9641	76	-	0.1563	0.1568

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.11.5
• Sentence Transformers: 3.1.1
• Transformers: 4.45.0
• PyTorch: 2.3.0+cu121
• Accelerate: 0.34.2
• Datasets: 3.0.0
• Tokenizers: 0.20.0

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",
}

MatryoshkaLoss

@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

@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}
}