thebajajra's picture
Upload folder using huggingface_hub
29e288a verified
metadata
language:
  - en
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:221599363
  - loss:MultipleNegativesRankingLoss
base_model: thebajajra/RexBERT-base
widget:
  - source_sentence: I found out that this novel was based on real ...
    sentences:
      - >-
        I found out that this novel was based on real people only by reading the
        afterword. This is a tremendously important piece of information about
        the book.
      - >-
        I recently got a mbp 16 and although I’m very impressed by the speakers
        I still wanted to purchase a set of external speakers for the desk
        setup. The thing is since these are so good I don’t even know at which
        price point I should be shopping to get something better. 


        The other day a youtuber I watch said that he has been using the mbp 16
        speakers instead of his $200 speakers because he doesn’t feel the need
        to anymore.


        So, is a pair of $60 speakers going to be better or do I need to go
        higher in price to really hear a difference?
      - >-
        Larry A Winters is a real good story teller.  His use and knowledge of
        Jessie Black as the heroine indicates a familiarity that makes the
        reader wonder if Larry and Jessie are one and the same.  A real page
        turner but not quite in the cant put it down stage.
  - source_sentence: Excellent seller but product did not work and returned for ...
    sentences:
      - >-
        Summary The K-II continuous-sample-flow-with-banding-9 zonal centrifuge
        hasbeen developed for large-scale virus isolation. The cylindrical
        aluminum rotor (capacity 3600 ml) contains a 3 liter gradient and has a
        700 ml stream volume. Fluid line seals are located on opposite ends of
        the rotor, eliminating the possibility of cross-leakage. An air turbine
        drive is used to accelerate the rotor to 35,000 rpm (83,440 g ,
        maximum). The development of safe armor and control systems is
        described.
      - >-
        I have a 2015 model and both front turn signals are busted after a
        wheelie mistake.  

        The left one doesn't light up and the right one is missing.  

        Where can I get the original stock turn signals?
      - >-
        Excellent seller but product did not work and returned for a full
        refund.  Did not pair up with my ATT box.  Was with a tech from ATT but
        we both could not get it to pair up.
  - source_sentence: where will the men's rings be held in 2016 olympics
    sentences:
      - >-
        the event of there being more than two gymnasts from same NOC, only the
        first two ranked among them would qualify for the final, with the next
        best ranked gymnast qualifying instead. Gymnastics at the 2016 Summer
        Olympics – Men's rings The Men's rings competition at the 2016 Summer
        Olympics was held at the HSBC Arena on 15 August. The medals were
        presented by Bernard Rajzman IOC member, Brazil and Wolfgang Willam, FIG
        Executive Committee Member. The top 8 qualifiers in the qualification
        phase (limit two per NOC), based on combined score of each apparatus,
        advanced to the individual all-around
      - >-
        Russians marked the anniversary of the 1917 Bolshevik Revolution on
        Sunday with marches, Communist rallies and protests against a
        parliamentary proposal to scrap what was once the most sacred Soviet
        holiday.
      - >-
        The only reason I could think of is the system has to log each
        individual painting as it’s own but since the same asset doesn’t require
        any new space and all the strokes are memorized anyway as you can move
        them individually, is there any real difference aside from perhaps the
        mild tedium of one way or the other?


        Just one of the small questions that keep me up at night.
  - source_sentence: So back to a semi-normal profile.
    sentences:
      - >-
        but suffers from uneven pacing. It's dark in tone, but not as arresting
        as, say, 3 AM or Damiano's Devil in Miss Jones. I'm glad I saw it, but
        there's little chance I'll ever revisit (except, maybe, to check out
        that Graham/Colt scene...)." The Story of Joanna The Story of Joanna is
        a 1975 pornographic film directed by Gerard Damiano and starring Jamie
        Gillis and Terri Hall. The film has a sado-masochism theme influenced by
        "Story of O" (1954). It is considered one of the classics of the Golden
        Age of Porn (1969–1984). It has been inducted into the XRCO
      - >


        Introduction


        Most medical schools across the globe use academic achievement as the
        primary selection criteria for admission into medical school. 1 This
        also applies to all medical schools in Nepal where entrance examinations
        conducted by the universities or Academies are based on general science
        subjects. 2,3 However, academic achievement alone as the predictor of
        someone becoming a 'good' physician has been questioned by many.
        [3][4][5][6][7][8] Certainly, personal qualities play an immense role in
        medical practice, which in itself is a complex phenomenon. [9][10] Patan
        Academy of Health Sciences (PAHS) is an autonomous, health sciences
        institute established in 2008 in Nepal with a mandate to improve the
        health of people in rural Nepal by producing health professionals who
        were competent, compassionate, and willing to serve in rural Nepal. It
        was clear from the outset that academic attainment alone among the
        aspirants for medical school was not going to be enough as admission
        criteria for the School of Medicine of PAHS. Further, PAHS also
        determined the desired characteristics/attributes of its graduates
        involving all the stakeholders a priori. 11 In this respect, a
        psychometric test battery (Personal Qualities Assessment, PQA 9-10 ) and
        an Admission OSCE [12][13] procedures were explored to see if they could
        be used to select medical students for PAHS. This paper reports the
        validation of the PQA test battery using science and health sciences
        students as they represented the majority of prospective applicants for
        the PAHS undergraduate medical education program commonly known as
        Bachelor in Medicine and Bachelor in Surgery (MBBS) in Nepal/South Asia,
        as well as non-science students of public/community schools. 14



        Method


        The Personal Qualities Assessment (PQA) test battery is used
        commercially to select health science students in many countries around
        the world (http://www.pqa.net.au/research.html) and found to be valid,
        reliable and predictive across different population. [9][10] The PQA
        test battery tests the cognitive ability through PQA Test A1 or Mental
        Agility Test (MAT) and a range of non-cognitive qualities though PQA
        Test A2 (Moral Orientation for Justice and Care: MOJAC) and PQA Test A3
        (Empathy, Confidence, Aloofness and Narcissism: ECAN). 15 PAHS, Nepal
        and PQA Innovation, Australia collaborated to locally validate and use
        PQA test for selecting medical students of School of Medicine, PAHS in
        early 2008. PAHS conducted the pilot tests of PQA test batteries using
        Optical Mark Reader (OMR) sheets, scanned them, created raw file and,
        sent it as secured spreadsheet file to PQA team after the test. The PQA
        team then scored the tests using pre-defined keys and rules and, send
        them back as secured spreadsheet and report files to PAHS for further
        processing.


        In order to validate the PQA tools in local context, they were
        forward-translated to Nepali by a professional bilingual person and was
        back-translated to English by another bilingual professional under the
        aegis of PAHS Admission team formed in 2008. The original PQA and
        back-translated PQA tools were then discussed iteratively among PAHS
        Admission and PQA teams before finalizing the Nepali version with
        consensus.


        The Personal Qualities Assessment tests in Nepali language was pretested
        with volunteer 10+2 non-science students of public/community schools
        located inside (n=75) and outside of Kathmandu valley (n=95) and
        volunteer 10+2 science students of a public school located outskirts of
        Kathmandu (n=35). As per PQA norm, only the volunteer students
        completing 80% and above items on Test A2 and Test A3 were included in
        the final analysis, whereas data of all the volunteer students on Test
        A1 was included in the final analysis. These tests were scored using the
        pre-defined keys and rules in Australia by the PQA team. Ethical
        approval was obtained from the Institutional Review Committee of PAHS
        (Ref: phs2204081608).


        Journal of Patan Academy of Health Sciences. 2022Aug;9(2):95-101.


        The Personal Qualities Assessment Test A1 (48 items with complex verbal,
        numerical, spatial, and abstract reasoning) and PQA Test A4 (90 items
        with simple verbal, numerical, spatial, and abstract reasoning) in
        English and Nepali languages were trailed again with the larger pool of
        volunteer 10+2 science students (n=131) of two community colleges and
        10+3 health science students (n=56) of a Government College, both
        located outside of Kathmandu valley. Descriptive statistics were used to
        describe the test scores whereas the ttest was used to compare the test
        scores between groups. A p-value less than 0.05 was taken as a
        statistically significant result.



        Result


        The Personal Qualities Assessment tests in Nepali were pre-tested first
        with 205 (110 males and 95 females) 10+2 non-science students in 2008,
        and Test A1 and Test A4 in English and Nepali were pre-tested again with
        187(141 males and 46 females) 10+2 science and 10+3 health science
        students in 2009. The mean±standard deviation (range) of students' age
        (in years) in the first and second samples were 20.3±1.3 (17-24) and
        18.5±3.4 .


        The cognitive ability test (PQA Test A1/MAT) had lower mean±SD scores
        15.3±3.7 than the norm 27.6±5.6 (multinational pool of 1187 applicants
        to medical schools). The range revealed the minimum and maximum scores
        as 7 and 25 with a median of 15. The internal consistency reliability
        coefficient (Cronbach's Alpha) was very low (0.27) for the first
        pre-test samples.


        On the other hand, the non-cognitive personality tests had comparable
        mean±SD scores of 109.4±13.9 for Test A2 and 259.5± 20.1 for Test A3
        with the multinational norms. Most importantly, the Coefficient alpha or
        the internal consistency reliability of Test A2 and Test A3 were greater
        than 0.80 (higher than the minimum value of 0.70) for both the tests.
        Further, a significant and low degree of negative correlation (r=-0.153,
        p=0.028) was found between Test A1 and Test A3 whereas a non-significant
        low negative correlation was observed between Test A1 and Test A2
        (r=-0.125, p=0.074). As Personal Qualities Assessment Test A1/MAT score
        in Pre-Test I followed a normal distribution (Shapiro-Wilk=0.987,
        p-value=0.068) and both science and non-science groups had equal
        variance (Levene's F=0.421, p-value=0.517), independent samples t-test
        was used to compare Test A1 score between science and non-science
        students, Table 2. Test A1 scores were found to be higher for science
        students and the result was highly significant statistically
        (t-test=-3.963, p-value<0.0001). The scatterplot of the standardized
        scores (zscores) of Test A2/MOJAC and Test A3/ECAN in Nepal language
        from non-science students shows that most of these students' LibCom
        (total of MOJAC) and ECAN z-scores lie between -2 and +2 SD and few
        students' scores were outside of this range, Figure 1. The Test A1
        (English language) scores for 10+3 health science students and 10+2
        science students were not significantly different but Test A4 (English
        language) scores for 10+3 health science students and 10+2 science
        students were statistically different in the second pre-test, Table
        4.   



        Discussion


        The MAT (Test A1) score was found to be lower than the international
        norms for both 10+2 science as well as non-science students, which
        suggests unfamiliarity with this form of test, differences in schooling,
        general cultural differences in approach to tests, etc. among these
        public/community school students. As the Test A1 questions were based on
        complex verbal, numerical, spatial, and abstract reasoning dimensions,
        it suggests that higher school students of Nepal require more exposure
        and practice on these types of aptitude tests as they are used widely to
        select students, screen recruits for military/police forces, and test
        job applicants. 16,17 The MAT (Test A1) in Nepali language scores were
        found to be significantly higher among higher secondary level science
        students compared to non-science students, possibly due to mathematical
        intuition leading to plausible numerical and abstract reasoning as part
        of their courses rather than higher verbal and spatial reasoning
        abilities. The MAT in the Nepali language had a low internal consistency
        reliability coefficient in the first pre-test, indicating that the
        different types of items i.e., verbal, numerical, spatial, and abstract
        reasoning items included in the test were of differing difficulties for
        this group, who may have guessed many of their answers. Also, possibly
        the volunteers felt that since their future was not at stake, they did
        not feel the need to fully exercise their intellectual ability in
        answering the questions, as most of them (170 out of 205) were
        non-science students. It may also be true that Test A1 in the Nepali
        language is not a suitable cognitive ability test for the 10+2 science
        as well as 10+2 nonscience students.


        On the other hand, MOJAC (Test A2) and NACE (Test A3) scores in Pre-Test
        I were similar to the international norm and had very high internal
        construct reliability (>0.80) suggesting they are satisfactory tests for
        Nepali applicants at 10+2 level or equivalent in both science and
        nonscience streams. These tests could be used to deselect outliers,
        i.e., students with potential behavior problems where outliers were
        Journal of Patan Academy of Health Sciences. 2022Aug;9(2):95-101.


        defined statistically as below -2 SD and above +2 SD for the
        standardized total MOJAC (LibCom) and ECAN scores. 15 Tests A2 and Test
        A3 had small negative but statistically insignificant correlations with
        Test A1, showing that they measured different traits (Test A1 measuring
        cognitive abilities and Test A2 and A3 measure non-cognitive traits) and
        thus are both potentially useful for selecting students for all the
        undergraduate level health-related programs in Nepal.


        Ironically, the small negative correlation further indicates that there
        is a slight tendency for those who are stronger in cognitive skills to
        be weaker in interpersonal skills, but there were still a substantial
        proportion of applicants who are strong in both.


        During the second pre-test done to check the consistency of the first
        pre-test results, Test A1 in Nepali was again found to have a slightly
        low internal reliability coefficient (Alpha < 0.50) for a larger pool of
        n=89 of volunteer science and health science students whereas the
        original MAT/Test A1 in the English language had a slightly more
        acceptable internal construct reliability (Alpha>0.60) for n=98
        volunteer science and health science students. This result is similar to
        2003 Scottish medical school applicants 10 though it is lower than the
        PQA international norm student sample average of 0.73. 15 Coefficient
        alpha of 0.60 and above is considered good and 0.70 and above is
        considered very good for tests with complex items i.e., MAT (Test A1) in
        the English language used in Nepal. 18 Further, Test A4 in the English
        language showed statistically different and higher results for science
        and health sciences students compared to Test A4 in the Nepali language
        whereas Test A1 in the English language showed higher but statistically
        insignificant results compared to Test A1 in the Nepali language. So,
        Test A4 containing simple verbal, numerical, spatial, and abstract
        reasoning items is found to be easy whereas Test A1 containing complex
        verbal, numerical, spatial, and abstract reasoning items is found to be
        difficult for both groups of students, Figure 1.


        When the Test A1 and Test A4 test scores in the English language were
        analyzed separately for the science and health science students, Test A1
        scores were not found to be statistically different indicating a fair
        test to select undergraduate medical students compared to Test A4 which
        produced a statistically different score. Thus, MAT (Test A1) in the
        English version was chosen to select MBBS students of the School of
        Medicine, Patan Academy of Health Sciences as it had sufficient internal
        consistency reliability and was fair to both science and health science
        students, despite being a bit difficult test of verbal, numerical,
        spatial and abstract reasoning aptitude required for the course. Recent
        studies confirmed the predictive validity of PQA tests among medical
        students in the UK, which remains to be done at PAHS. [18][19][20] 



        Conclusion


        The MAT (PQA Test A1) in English was found to be a reliable test to
        select medical students for PAHS and similar institutions in Nepal as it
        was also found to be fair among 10+2 science/10+3 health science
        students. Also, PQA Test A2 and Test A3 in Nepali were found to be fair
        and reliable tests to identify unusual personality traits and to
        deselect such candidates for all the undergraduate level health science
        programs in Nepal and beyond.



        PAHS Admission committee conducted the second Pre-Test of Test A1 and
        Test A4 in Nepali and English languages with the large (187) volunteer
        10+2 science students and 10+3 health science students in public
        school/college outside of Kathmandu valley as Nepali Test A1 results
        were not promising with the non-science and the science students. The
        Mental Agility Test (Test A1) had lower mean±SD scores for English
        18.4±5.0 language and Nepali 16.8±4.3 language than the multinational
        norm of 27.6±5.6. Yet, Test A1 in the English language's mean score of
        18.4 was different from Test A1 in Nepali's mean score of 16.8, which
        was also statistically significant (t=2.3348, p-value=0.0206). Although
        the mean of Test A1 in the English language (18.4) done in 2009 was
        found to be higher than Test A1 in the Nepali language (17.51) for the
        10+2 Test A1 in the English language and 0.49 for the Nepali language
        for the second pre-test samples. On the other hand, the mean percentage
        score of the General Ability Test (Test A4) in English language and
        Nepali Journal of Patan Academy of Health Sciences. 2022Aug;9(2):95-101.
        language were 50.0 and 47.6 respectively, which was higher than the mean
        percentage score of Test A1 in the English language (38.3) and Nepali
        language (35.0). The mean percentage scores were statistically different
        for English and Nepali versions of Test A4 (t=2.310, p-value=0.022) but
        it was not statistically different for English and Nepali versions of
        Test A1 (t=1.367, p-value=0.171). Further analysis of Test A1 and Test
        A4 scores in English and Nepali languages with the sex of the applicants
        was not found to be statistically different. However, Test A1 and Test
        A4 showed statistically significant negativescience students in 2008,
        they were not 

        significantly 

        different 

        (t=0.9769, 

        p-

        value=0.3304), Table 2 & 3. The internal 

        consistency reliability (coefficient alpha), a 

        proxy for construct validity, was found to be 

        0.63 for correlations with the age of the students for 

        English (r=-0.226, p-value=0.025; r=-0.395, p-

        value<0.001) and Nepali (r=-0.261, p-

        value=0.011; 

        r=-0.211, 

        p-value=0.047) 

        languages. 




        Table 1 .

        1Personal qualities assessment pre-test I with non-science major
        students, 2008, NepalTable 2. Test A1 (Nepali) score of 10+2 science and
        non-science students, 2008Test (Language) 

        Nepali candidates 

        International candidates 

        N Mean 

        SD 

        Median 

        Range Alpha 

        N Mean 

        SD 

        Test A1-NEPALI 

        205 

        15.3 

        3.7 

        15.0 

        7 -25 

        0.27 

        1811 

        27.6 

        5.6 

        Test A2-NEPALI 

        205 109.4 

        13.9 

        108.0 

        79 -147 

        0.82 

        9762 116.0 

        15.3 

        Test A3-NEPALI 

        205 259.6 

        20.1 

        259.0 

        216 -320 

        0.81 

        7032 283.0 

        22.8 


        Stream 

        N Mean 

        SD Median 

        Min 

        Max 

        t 

        p-value 

        Non-sciences -NEPALI (48 items) 

        170 14.87 3.643 

        15.00 

        7 

        24 

        3.963 

        <0.0001 

        Science -NEPALI (48 items) 

        35 17.51 3.338 

        17.00 

        12 

        25 

        Total 

        205 15.32 3.721 

        15.00 

        7 

        25 




        Table 3 .

        3Personal qualities assessment pre-test II with science and health
        science students, 2009, NepalTest (Language) 

        Nepali candidates 

        International candidates 

        N 

        Mean 

        SD 

        Median 

        Range Alpha 

        N Mean 

        SD 

        Test A1-ENGLISH 

        (48 items) 


        98 

        18.4 

        5.0 

        18.0 

        8 -30 

        0.63 

        1811 

        27.6 

        5.6 


        Test A1-NEPALI 

        (48 items) 


        89 

        16.8 

        4.3 

        17.0 

        8 -25 

        0.49 

        NA 


        Test A4-ENGLISH 

        (90 items) 


        98 

        45.0 

        11.3 

        44.5 

        18 -77 

        0.88 

        NA 


        Test A4-NEPALI 

        (90 items) 


        89 

        42.8 

        10.1 

        43.0 

        12 -60 

        NA 

        NA 


        NA=Not Available 




        Table 4 .

        4Comparison of mental agility test (test A1) and general ability test
        (test A4) (English version) scores 

        among science and health sciences students, pre-test II, 2009, Nepal 


        Test and Stream 

        (Test: Discipline) 


        Nepal Candidates 

        N 

        Mean 

        SD 

        SEM 

        Median 

        Range 

        T 

        p-value 

        Test A1-ENGLISH: 

        10+3 Health Sciences 


        56 

        18.3 

        4.5 

        0.61 

        17.0 

        12 -29 


        0.342 

        0.772 

        Test A1-ENGLISH: 

        10+2 Sciences 


        42 

        18.6 

        5.6 

        0.87 

        18.5 

        8 -30 


        Test A4-ENGLISH: 

        10+3 Health Sciences 


        56 

        41.77 

        9.3 

        1.25 

        42.5 

        18 -60 


        3.292 

        0.002 

        Test A4-ENGLISH: 

        10+2 Sciences 


        42 

        49.24 

        12.3 

        1.90 

        51.5 

        24 -77 


        SEM = Standard error of the measurement 



        AcknowledgmentWe are grateful to the officials of PAHS and the two
        Admission Teams for their help during this study. We are thankful to all
        the faculty and fellows at PSG FAIMER Regional Institute, Coimbatore,
        India for their inputs and feedback as this study was conducted as part
        of the educational innovation project of the first author to fulfill the
        partial requirement for the FAIMER fellowship in medical education in
        2008-2010.Conflict of Interest None

        How effective are selection methods in medical education? A systematic
        review. F Patterson, A Knight, J Dowell, S Nicholson, F Cousans, Cleland
        J , Patterson F, Knight A, Dowell J, Nicholson S, Cousans F, Cleland J.
        How effective are selection methods in medical education? A systematic
        review. Med Educ. 2016


        . 10.1111/medu.12817| DOI | PubMed | Google Scholar | Weblink
        |50Jan;50(1):36-60. | DOI | PubMed | Google Scholar | Weblink |


        MBBS student selection: search for proper criteria. S Maharjan, H Dixit,
        Kathmandu Univ Med J (KUMJ). 23| PubMed | Google Scholar | Full Text |
        WeblinkMaharjan S, Dixit H. MBBS student selection: search for proper
        criteria. Kathmandu Univ Med J (KUMJ). 2003;2(3):252-259. | PubMed |
        Google Scholar | Full Text | Weblink |


        Critical Analysis of Performance of Medical Students. S R Niraula, S S
        Khanal, 10.1080/13576280500534578Educ Health (Abingdon). 191| DOI |
        PubMed | Google Scholar | Full Text | WeblinkNiraula SR, Khanal SS.
        Critical Analysis of Performance of Medical Students. Educ Health
        (Abingdon). 2006;19(1):5-13. | DOI | PubMed | Google Scholar | Full Text
        | Weblink |


        Testing medical school selection tests. C Mcmanus, D Powis,
        10.5694/j.1326-5377.2007.tb00832.xMed J Aust. 118. | DOI | PubMed |
        Google Scholar | Full Text | Weblink |1863McManus C, Powis D. Testing
        medical school selection tests. Med J Aust. 2007;186(3):118. | DOI |
        PubMed | Google Scholar | Full Text | Weblink |


        Health inequities: the need for action by schools of medicine. R W
        Sanson-Fisher, N Williams, S Outram, Med Teach. 30389Sanson-Fisher RW,
        Williams N, Outram S. Health inequities: the need for action by schools
        of medicine. Med Teach. 2008;30:389-


        . | Doi | Pubmed, | Google,
        https:/www.tandfonline.com/doi/abs/10.1080/01421590801948042| DOI |
        PubMed | Google Scholar |Weblink |


        Selecting medical students. D Powis,
        https:/onlinelibrary.wiley.com/doi/abs/10.1111/j.1365-2923.1994.tb02555.xMed
        Educ. 28| DOI | PubMed | Google Scholar | WeblinkPowis D. Selecting
        medical students. Med Educ. 1994;28:443-69. | DOI | PubMed | Google
        Scholar | Weblink |


        How to do it: select medical students. D Powis, BMJ. 317| DOI | PubMed |
        Google Scholar | Full Text | WeblinkPowis D. How to do it: select
        medical students. BMJ. 1998;317:1149-50. | DOI | PubMed | Google Scholar
        | Full Text | Weblink |


        Widening access by changing the criteria for selecting medical students.
        D Powis, J Hamilton, Ic ; | Mcmanus, | Doi | Google Scholar, Weblink,
        https:/psycnet.apa.org/doi/10.1016/j.tate.2007.06.001Teaching Teach
        Educ. 23Powis D, Hamilton J, McManus IC. Widening access by changing the
        criteria for selecting medical students. Teaching Teach Educ.
        2007;23:1235-45. | DOI | Google Scholar | Weblink |


        Development of the personal qualities assessment as a tool for selecting
        medical students. D Powis, M Bore, D Munro, Ma ; | Lumsden, | Doi |
        Google Scholar, Weblink,
        https:/journals.sagepub.com/doi/10.7227/JACE.11.1.2J Adult Continuing
        Education. 111Powis D, Bore M, Munro D, Lumsden MA. Development of the
        personal qualities assessment as a tool for selecting medical students.
        J Adult Continuing Education. 2005;11(1):3-14. | DOI | Google Scholar |
        Weblink |


        Assessment of personal qualities in relation to medical school. M A
        Lumsden, M Bore, K Millar, R Jack, D Powis,
        https:/onlinelibrary.wiley.com/doi/abs/10.1111/j.1365-2929.2005.02087.xMed
        Educ. 39| DOI | PubMed | Google Scholar | WeblinkLumsden MA, Bore M,
        Millar K, Jack R, Powis D. Assessment of personal qualities in relation
        to medical school. Med Educ. 2005;39:258-65. | DOI | PubMed | Google
        Scholar | Weblink |


        Designing an assessment tool for professional attributes of medical
        graduates from a new medical school in Nepal. Jhc Morgan, Google Scholar
        | Full Text |. 3Morgan JHC. Designing an assessment tool for
        professional attributes of medical graduates from a new medical school
        in Nepal. SEAJME. 2009;3(1):2-7. | Google Scholar | Full Text |


        The objective structured interview for medical student selection. D
        Powis, Rlb Neame, T Bristow, L B Murphy, BMJ. 296| DOI | PubMed | Google
        Scholar | Full Text | WeblinkPowis D, Neame RLB, Bristow T, Murphy LB.
        The objective structured interview for medical student selection. BMJ.
        1988;296:765-8. | DOI | PubMed | Google Scholar | Full Text | Weblink |


        An admissions OSCE: the multiple mini-interview. K W Eva, J Rosenfield,
        H I Reiter, G R Norman, Eva KW, Rosenfield J, Reiter HI, Norman GR. An
        admissions OSCE: the multiple mini-interview.


        .
        https:/onlinelibrary.wiley.com/doi/full/10.1046/j.1365-2923.2004.01776.xMed
        Educ. | DOI | PubMed | Google Scholar| Weblink |38Med Educ.
        2004;38:314-26. | DOI | PubMed | Google Scholar| Weblink |


        Reliability and validity of admissions tools used to select students for
        the health professions. P Salvatori,
        https:/link.springer.com/article/10.1023/A:1011489618208Adv Health Sci
        Educ. 62| DOI | PubMed | Google Scholar | WeblinkSalvatori P.
        Reliability and validity of admissions tools used to select students for
        the health professions. Adv Health Sci Educ. 2001;6(2):159-75. | DOI |
        PubMed | Google Scholar | Weblink |


        A comprehensive model for the selection of medical students. M Bore, D
        Munro, D Powis,
        https:/www.tandfonline.com/doi/abs/10.3109/01421590903095510Med Teach.
        3112| DOI | PubMed | Google Scholar | WeblinkBore M, Munro D, Powis D. A
        comprehensive model for the selection of medical students. Med Teach.
        2009;31(12):1066-72. | DOI | PubMed | Google Scholar | Weblink |


        Role of public and private schools in developing of IQ among elementary
        students. M Batool, M A Dahar, M I Ali, | Weblink | Full Text. 82Batool
        M, Dahar MA, Ali MI. Role of public and private schools in developing of
        IQ among elementary students. IJSER. 2017;8(2):248-62. | Weblink | Full
        Text |


        Development and validation of measures of noncognitive college student
        potential. The College Board. N Schmitt, A Billington, J Keeney, M
        Reeder, T J Pleskac, R Sinha, Research Report. 2011;1. | Full Text |
        WeblinkSchmitt N, Billington A, Keeney J, Reeder M, Pleskac TJ, Sinha R,
        et al. Development and validation of measures of noncognitive college
        student potential. The College Board: Research Report. 2011;1. | Full
        Text | Weblink |


        Making sense of Cronbach's alpha. M Tavakol, R Dennick, Int J Med Educ.
        2Tavakol M, Dennick R. Making sense of Cronbach's alpha. Int J Med Educ.
        2011;2:53-5.


        | Doi | Pubmed, Google Scholar | Full Text | Weblink |. | DOI | PubMed |
        Google Scholar | Full Text | Weblink |


        Predictors of professional behavior and academic outcomes in a UK
        medical school: A longitudinal cohort study. J Adam, M Bore, R Child, J
        Dunn, J Mckendree, D Munro,
        https:/www.tandfonline.com/doi/abs/10.3109/0142159X.2015.1009023?journalCode=imte20Med
        Teach. 379| DOI | PubMed | Google Scholar | Full Text | WeblinkAdam J,
        Bore M, Child R, Dunn J, Mckendree J, Munro D, et al. Predictors of
        professional behavior and academic outcomes in a UK medical school: A
        longitudinal cohort study. Med Teach. 2015;37(9):868-80. | DOI | PubMed
        | Google Scholar | Full Text | Weblink |


        Assessment for selection for the health care professions and specialty
        training: consensus statement and recommendations from the Ottawa. D
        Prideaux, C Roberts, K Eva, A Centeno, P Mccrorie, C Mcmanus, Prideaux
        D, Roberts C, Eva K, Centeno A, Mccrorie P, Mcmanus C, et al. Assessment
        for selection for the health care professions and specialty training:
        consensus statement and recommendations from the Ottawa 2010


        .
        https:/www.tandfonline.com/doi/abs/10.3109/0142159X.2011.551560?journalCode=imte20Conference.
        Med Teach. 333| DOI | PubMed | Google Scholar | WeblinkConference. Med
        Teach. 2011;33(3):215-23. | DOI | PubMed | Google Scholar | Weblink |


        The change from UMAT to UCAT for undergraduate medical school
        applicants: impact on selection outcomes. B Griffin, G L Horton, L
        Lampe, B Shulruf, W Hu, 10.5694/mja2.50877Med J Aust. 2142DOI | PubMed |
        Google Scholar | Full Text | WeblinkGriffin B, Horton GL, Lampe L,
        Shulruf B, Hu W. The change from UMAT to UCAT for undergraduate medical
        school applicants: impact on selection outcomes. Med J Aust.
        2021;214(2):84-9. | DOI | PubMed | Google Scholar | Full Text | Weblink
        |
      - >-
        Profile


        After the winking / rating / favoriting experiment, I figured it was
        time to go back to my normally scheduled profile with a few of the
        undertones left over from the old one. Old Profile


        What do you guys think? 


        Also it's amazing what a picture of a pet can do. :|
  - source_sentence: Corners are still lifting.
    sentences:
      - >-
        Great product. I purchased this item becuase my wrists would ache after
        triceps day at the gym. I would never be able to straighten my wrist and
        this helped in fixing that issue.
      - >-
        These are awesome quart jars. They have a beautiful color, and I use
        them for storing soups, nuts and homemade nut milk. I would purchase
        them again.
      - >-
        Hello, I got my ender 3 a little over a year ago and have gotten many
        successful prints off of my machine. 


        I have always had a problem with the corners of my prints lifting. I
        originally used a glass plate. That by itself was horrible, but then I
        added hairspray, and that worked. The problem was that on long prints,
        corners still lifted.


        After doing this for around 5 months I switched to a PEI sheet.


        This worked comparably as well as the glass/hairspray combo, except the
        corners STILL LIFT on long prints.


        Now I have a PEI sheet on boro glass with an EZABL attached and the
        corners of my prints are STILL LIFTING.


        I don't know what i could possibly be doing wrong. The bed must be
        level. I get beautiful first layers, which I have tried to "smudge"
        around during printing and I can confirm that the plastic is being layed
        down solidly.


        If anyone could enlighten me as to what is going on I would be thrilled.


        I do have my first layer printing at 30% speed with 150% layer width
        with the print cooling fan off as well. Printing PLA at 200C tool temp,
        60C bed.
datasets:
  - nomic-ai/nomic-embed-unsupervised-data
pipeline_tag: sentence-similarity
library_name: sentence-transformers

SentenceTransformer based on thebajajra/RexBERT-base

This is a sentence-transformers model finetuned from thebajajra/RexBERT-base on the nomic-embed-unsupervised-data 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 Sources

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 1024, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
  (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("sentence_transformers_model_id")
# Run inference
queries = [
    "Corners are still lifting.",
]
documents = [
    'Hello, I got my ender 3 a little over a year ago and have gotten many successful prints off of my machine. \n\nI have always had a problem with the corners of my prints lifting. I originally used a glass plate. That by itself was horrible, but then I added hairspray, and that worked. The problem was that on long prints, corners still lifted.\n\nAfter doing this for around 5 months I switched to a PEI sheet.\n\nThis worked comparably as well as the glass/hairspray combo, except the corners STILL LIFT on long prints.\n\nNow I have a PEI sheet on boro glass with an EZABL attached and the corners of my prints are STILL LIFTING.\n\nI don\'t know what i could possibly be doing wrong. The bed must be level. I get beautiful first layers, which I have tried to "smudge" around during printing and I can confirm that the plastic is being layed down solidly.\n\nIf anyone could enlighten me as to what is going on I would be thrilled.\n\nI do have my first layer printing at 30% speed with 150% layer width with the print cooling fan off as well. Printing PLA at 200C tool temp, 60C bed.',
    'These are awesome quart jars. They have a beautiful color, and I use them for storing soups, nuts and homemade nut milk. I would purchase them again.',
    'Great product. I purchased this item becuase my wrists would ache after triceps day at the gym. I would never be able to straighten my wrist and this helped in fixing that issue.',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[0.5487, 0.0457, 0.1961]])

Training Details

Training Dataset

nomic-embed-unsupervised-data

  • Dataset: nomic-embed-unsupervised-data at 917bae6
  • Size: 221,599,363 training samples
  • Columns: query and document
  • Approximate statistics based on the first 1000 samples:
    query document
    type string string
    details
    • min: 6 tokens
    • mean: 34.17 tokens
    • max: 1024 tokens
    • min: 8 tokens
    • mean: 166.07 tokens
    • max: 1024 tokens
  • Samples:
    query document
    Effect of steam reforming on methane-fueled chemical looping combustion with Cu-based oxygen carrier Abstract The reduction characteristics of Cu-based oxygen carrier with H 2 , CO and CH 4 were investigated using a fixed bed reactor, TPR and TGA. Results showed that temperatures for the complete reduction of Cu-based oxygen carrier with H 2 and CO are 300 °C and 225 °C, respectively, while the corresponding temperature with CH 4 is 650 °C. The carbon deposition from CH 4 occurred at over 550 °C. CO-chemisorption experiments were also conducted on the oxygen carrier, and it was indicated that Cu-based oxygen carrier sinter seriously at 700 °C. In order to lower the required reduction temperature of oxygen carriers, a new chemical looping combustion (CLC) process with CH 4 steam reforming has been presented in this paper. The basic feasibility of the process was illustrated using CuO–SiO 2 . The new CLC process has the potential to replace the conventional gas-fired middle- and low-pressure steam and hot water boilers.
    who appointed onesicritus as chief pilot of the fleet by the king to hold a conference with the Indian philosophers or Gymnosophists, the details of which have been transmitted to us from his own account of the interview. It was Onesicritus, whom Alexander first sent to summon Dandamis to his court. When later Onesicritus returned empty-handed with the reply of Dandamis, the King went to forest to visit Dandamis. When Alexander constructed his fleet on the Hydaspes, he appointed Onesicritus to the important position of pilot of the king's ship, or chief pilot of the fleet (). Onesicritus held this position not only during the descent of the Indus,
    when did the madonna of foligno go to paris Madonna of Foligno hence the name. In 1799 it was carried to Paris, France by Napoleon. There, in 1802, the painting was transferred from panel to canvas by Hacquin and restored by Roser of Heidelberg. A note was made by the restorer: "Rapporto dei cittadini Guijon Vincent Tannay e Berthollet sul ristauro dei quadri di Raffaello conosciuto sotto il nome di Madonna di Foligno." In 1815, after the Battle of Waterloo, it was returned to Italy, where it was placed in the room with the Transfiguration in the Pinacoteca Vaticana of the Vatican Museum in the Vatican City. The painting is a "sacra
  • Loss: MultipleNegativesRankingLoss with these parameters:
    {
        "scale": 20.0,
        "similarity_fct": "cos_sim",
        "gather_across_devices": false
    }
    

Evaluation Dataset

nomic-embed-unsupervised-data

  • Dataset: nomic-embed-unsupervised-data at 917bae6
  • Size: 1,113,579 evaluation samples
  • Columns: query and document
  • Approximate statistics based on the first 1000 samples:
    query document
    type string string
    details
    • min: 5 tokens
    • mean: 31.98 tokens
    • max: 1024 tokens
    • min: 6 tokens
    • mean: 161.48 tokens
    • max: 1024 tokens
  • Samples:
    query document
    Concise methods for the synthesis of chiral polyoxazolines and their application in asymmetric hydrosilylation Seven polyoxazoline ligands were synthesized in high yield in a one-pot reaction by heating polycarboxylic acids or their esters and chiral β-amino alcohols under reflux with concomitant removal of water or the alcohol produced in the reaction. The method is much simpler and more efficient in comparison to those methods reported in the literature.The compounds were used as chiral ligands in the rhodium-catalyzed asymmetric hydrosilylation of aromatic ketones, and the effects of the linkers and the substituents present on the oxazoline rings on the yield and enantioselectivity investigated. Compound 2 was identified as the best ligand of this family for the hydrosilylation of aromatic ketones.
    On the road to a stronger public health workforce: visual tools to address complex challenges. The Public Health Workforce Taxonomy: Revisions and Recommendations for Implementation
    140mm Jetflo fan availability? I recently purchased a Nepton 280L, and would like to install an additional pair of 140mm Jetflo fans. Unfortunately they don't seem to be currently available, will they be in the future?

    Thank you so much!

    PS - I'm loving the cooling system!
  • Loss: MultipleNegativesRankingLoss with these parameters:
    {
        "scale": 20.0,
        "similarity_fct": "cos_sim",
        "gather_across_devices": false
    }
    

Training Hyperparameters

Non-Default Hyperparameters

  • eval_strategy: steps
  • per_device_train_batch_size: 384
  • per_device_eval_batch_size: 128
  • learning_rate: 1e-05
  • num_train_epochs: 4
  • warmup_steps: 1000
  • bf16: True
  • dataloader_num_workers: 20
  • dataloader_prefetch_factor: 4
  • ddp_find_unused_parameters: False

All Hyperparameters

Click to expand
  • overwrite_output_dir: False
  • do_predict: False
  • eval_strategy: steps
  • prediction_loss_only: True
  • per_device_train_batch_size: 384
  • per_device_eval_batch_size: 128
  • per_gpu_train_batch_size: None
  • per_gpu_eval_batch_size: None
  • gradient_accumulation_steps: 1
  • eval_accumulation_steps: None
  • torch_empty_cache_steps: None
  • learning_rate: 1e-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: linear
  • lr_scheduler_kwargs: {}
  • warmup_ratio: 0.0
  • warmup_steps: 1000
  • 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
  • bf16: True
  • fp16: False
  • fp16_opt_level: O1
  • half_precision_backend: auto
  • bf16_full_eval: False
  • fp16_full_eval: False
  • tf32: None
  • local_rank: 0
  • ddp_backend: None
  • tpu_num_cores: None
  • tpu_metrics_debug: False
  • debug: []
  • dataloader_drop_last: True
  • dataloader_num_workers: 20
  • dataloader_prefetch_factor: 4
  • 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}
  • parallelism_config: 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
  • project: huggingface
  • trackio_space_id: trackio
  • ddp_find_unused_parameters: False
  • ddp_bucket_cap_mb: None
  • ddp_broadcast_buffers: False
  • dataloader_pin_memory: True
  • dataloader_persistent_workers: False
  • skip_memory_metrics: True
  • use_legacy_prediction_loop: False
  • push_to_hub: False
  • resume_from_checkpoint: None
  • hub_model_id: None
  • hub_strategy: every_save
  • hub_private_repo: None
  • hub_always_push: False
  • hub_revision: None
  • gradient_checkpointing: False
  • gradient_checkpointing_kwargs: None
  • include_inputs_for_metrics: False
  • include_for_metrics: []
  • eval_do_concat_batches: True
  • fp16_backend: auto
  • push_to_hub_model_id: None
  • push_to_hub_organization: None
  • mp_parameters:
  • auto_find_batch_size: False
  • full_determinism: False
  • torchdynamo: None
  • ray_scope: last
  • ddp_timeout: 1800
  • torch_compile: False
  • torch_compile_backend: None
  • torch_compile_mode: None
  • include_tokens_per_second: False
  • include_num_input_tokens_seen: no
  • neftune_noise_alpha: None
  • optim_target_modules: None
  • batch_eval_metrics: False
  • eval_on_start: False
  • use_liger_kernel: False
  • liger_kernel_config: None
  • eval_use_gather_object: False
  • average_tokens_across_devices: True
  • prompts: None
  • batch_sampler: batch_sampler
  • multi_dataset_batch_sampler: proportional
  • router_mapping: {}
  • learning_rate_mapping: {}

Training Logs

Click to expand
Epoch Step Training Loss Validation Loss
0.0014 100 4.5665 -
0.0028 200 2.223 -
0.0042 300 0.3767 -
0.0055 400 0.1622 -
0.0069 500 0.1154 -
0.0083 600 0.0934 -
0.0097 700 0.0797 -
0.0111 800 0.0704 -
0.0125 900 0.0625 -
0.0139 1000 0.0582 -
0.0152 1100 0.0535 -
0.0166 1200 0.0492 -
0.0180 1300 0.0463 -
0.0194 1400 0.044 -
0.0208 1500 0.0416 -
0.0222 1600 0.0395 -
0.0236 1700 0.0381 -
0.0250 1800 0.0367 -
0.0263 1900 0.0358 -
0.0277 2000 0.0345 -
0.0291 2100 0.0335 -
0.0305 2200 0.0319 -
0.0319 2300 0.0318 -
0.0333 2400 0.0304 -
0.0347 2500 0.0301 -
0.0360 2600 0.0291 -
0.0374 2700 0.0293 -
0.0388 2800 0.0281 -
0.0402 2900 0.0277 -
0.0416 3000 0.0266 -
0.0430 3100 0.0265 -
0.0444 3200 0.0261 -
0.0457 3300 0.0253 -
0.0471 3400 0.0256 -
0.0485 3500 0.0247 -
0.0499 3600 0.0239 -
0.0513 3700 0.0239 -
0.0527 3800 0.0235 -
0.0541 3900 0.0233 -
0.0555 4000 0.0229 -
0.0568 4100 0.0227 -
0.0582 4200 0.0226 -
0.0596 4300 0.0221 -
0.0610 4400 0.0219 -
0.0624 4500 0.0211 -
0.0638 4600 0.0212 -
0.0652 4700 0.021 -
0.0665 4800 0.0205 -
0.0679 4900 0.0202 -
0.0693 5000 0.0206 -
0.0707 5100 0.0199 -
0.0721 5200 0.0202 -
0.0735 5300 0.0194 -
0.0749 5400 0.0195 -
0.0762 5500 0.0189 -
0.0776 5600 0.0194 -
0.0790 5700 0.0189 -
0.0804 5800 0.0183 -
0.0818 5900 0.0184 -
0.0832 6000 0.0183 -
0.0846 6100 0.018 -
0.0859 6200 0.0178 -
0.0873 6300 0.018 -
0.0887 6400 0.0174 -
0.0901 6500 0.0175 -
0.0915 6600 0.0176 -
0.0929 6700 0.0171 -
0.0943 6800 0.0168 -
0.0957 6900 0.0174 -
0.0970 7000 0.0171 -
0.0984 7100 0.0169 -
0.0998 7200 0.0167 -
0.1012 7300 0.0165 -
0.1026 7400 0.0166 -
0.1040 7500 0.0162 -
0.1054 7600 0.0164 -
0.1067 7700 0.0159 -
0.1081 7800 0.0159 -
0.1095 7900 0.0162 -
0.1109 8000 0.0157 -
0.1123 8100 0.0157 -
0.1137 8200 0.0155 -
0.1151 8300 0.0154 -
0.1164 8400 0.0155 -
0.1178 8500 0.0154 -
0.1192 8600 0.015 -
0.1206 8700 0.0151 -
0.1220 8800 0.0149 -
0.1234 8900 0.015 -
0.1248 9000 0.0146 -
0.1262 9100 0.015 -
0.1275 9200 0.0148 -
0.1289 9300 0.0145 -
0.1303 9400 0.0146 -
0.1317 9500 0.0148 -
0.1331 9600 0.0143 -
0.1345 9700 0.0144 -
0.1359 9800 0.0142 -
0.1372 9900 0.0142 -
0.1386 10000 0.0141 -
0.1400 10100 0.0139 -
0.1414 10200 0.0141 -
0.1428 10300 0.0139 -
0.1442 10400 0.0136 -
0.1456 10500 0.0135 -
0.1469 10600 0.0135 -
0.1483 10700 0.0134 -
0.1497 10800 0.0136 -
0.1511 10900 0.0133 -
0.1525 11000 0.0135 -
0.1539 11100 0.0133 -
0.1553 11200 0.0134 -
0.1567 11300 0.0133 -
0.1580 11400 0.0134 -
0.1594 11500 0.013 -
0.1608 11600 0.0131 -
0.1622 11700 0.0129 -
0.1636 11800 0.0127 -
0.1650 11900 0.0129 -
0.1664 12000 0.0125 -
0.1677 12100 0.0129 -
0.1691 12200 0.013 -
0.1705 12300 0.013 -
0.1719 12400 0.013 -
0.1733 12500 0.0125 -
0.1747 12600 0.0125 -
0.1761 12700 0.0122 -
0.1774 12800 0.0124 -
0.1788 12900 0.0124 -
0.1802 13000 0.0121 -
0.1816 13100 0.0124 -
0.1830 13200 0.0122 -
0.1844 13300 0.0123 -
0.1858 13400 0.0121 -
0.1871 13500 0.012 -
0.1885 13600 0.0118 -
0.1899 13700 0.0119 -
0.1913 13800 0.0117 -
0.1927 13900 0.0119 -
0.1941 14000 0.0119 -
0.1955 14100 0.0117 -
0.1969 14200 0.0119 -
0.1982 14300 0.0116 -
0.1996 14400 0.0116 -
0.2 14427 - 0.0044
0.2010 14500 0.012 -
0.2024 14600 0.0116 -
0.2038 14700 0.0118 -
0.2052 14800 0.0116 -
0.2066 14900 0.0118 -
0.2079 15000 0.0118 -
0.2093 15100 0.0113 -
0.2107 15200 0.0114 -
0.2121 15300 0.0115 -
0.2135 15400 0.0116 -
0.2149 15500 0.0113 -
0.2163 15600 0.0115 -
0.2176 15700 0.0112 -
0.2190 15800 0.0112 -
0.2204 15900 0.0114 -
0.2218 16000 0.0111 -
0.2232 16100 0.0112 -
0.2246 16200 0.0111 -
0.2260 16300 0.011 -
0.2274 16400 0.011 -
0.2287 16500 0.0109 -
0.2301 16600 0.0106 -
0.2315 16700 0.011 -
0.2329 16800 0.011 -
0.2343 16900 0.0108 -
0.2357 17000 0.0106 -
0.2371 17100 0.0108 -
0.2384 17200 0.0107 -
0.2398 17300 0.0105 -
0.2412 17400 0.0107 -
0.2426 17500 0.011 -
0.2440 17600 0.0105 -
0.2454 17700 0.0107 -
0.2468 17800 0.0106 -
0.2481 17900 0.0108 -
0.2495 18000 0.0106 -
0.2509 18100 0.0105 -
0.2523 18200 0.0103 -
0.2537 18300 0.0104 -
0.2551 18400 0.0105 -
0.2565 18500 0.0103 -
0.2578 18600 0.0104 -
0.2592 18700 0.0103 -
0.2606 18800 0.0102 -
0.2620 18900 0.0101 -
0.2634 19000 0.0102 -
0.2648 19100 0.0103 -
0.2662 19200 0.01 -
0.2676 19300 0.0103 -
0.2689 19400 0.0101 -
0.2703 19500 0.0103 -
0.2717 19600 0.0101 -
0.2731 19700 0.0103 -
0.2745 19800 0.0102 -
0.2759 19900 0.0102 -
0.2773 20000 0.0103 -
0.2786 20100 0.0101 -
0.2800 20200 0.0102 -
0.2814 20300 0.0099 -
0.2828 20400 0.0099 -
0.2842 20500 0.0099 -
0.2856 20600 0.0098 -
0.2870 20700 0.0099 -
0.2883 20800 0.0097 -
0.2897 20900 0.0101 -
0.2911 21000 0.0098 -
0.2925 21100 0.0099 -
0.2939 21200 0.0099 -
0.2953 21300 0.0098 -
0.2967 21400 0.0096 -
0.2981 21500 0.0097 -
0.2994 21600 0.0097 -
0.3008 21700 0.0099 -
0.3022 21800 0.0098 -
0.3036 21900 0.0096 -
0.3050 22000 0.0097 -
0.3064 22100 0.0098 -
0.3078 22200 0.0094 -
0.3091 22300 0.0096 -
0.3105 22400 0.0095 -
0.3119 22500 0.0098 -
0.3133 22600 0.0096 -
0.3147 22700 0.0094 -
0.3161 22800 0.0095 -
0.3175 22900 0.0093 -
0.3188 23000 0.0093 -
0.3202 23100 0.0093 -
0.3216 23200 0.0094 -
0.3230 23300 0.0094 -
0.3244 23400 0.0093 -
0.3258 23500 0.0091 -
0.3272 23600 0.0093 -
0.3286 23700 0.0093 -
0.3299 23800 0.0093 -
0.3313 23900 0.0093 -
0.3327 24000 0.0093 -
0.3341 24100 0.009 -
0.3355 24200 0.0093 -
0.3369 24300 0.0089 -
0.3383 24400 0.0089 -
0.3396 24500 0.0092 -
0.3410 24600 0.009 -
0.3424 24700 0.0092 -
0.3438 24800 0.009 -
0.3452 24900 0.0091 -
0.3466 25000 0.0088 -
0.3480 25100 0.009 -
0.3493 25200 0.0089 -
0.3507 25300 0.0088 -
0.3521 25400 0.0089 -
0.3535 25500 0.0089 -
0.3549 25600 0.009 -
0.3563 25700 0.0092 -
0.3577 25800 0.0089 -
0.3590 25900 0.0089 -
0.3604 26000 0.009 -
0.3618 26100 0.0088 -
0.3632 26200 0.0088 -
0.3646 26300 0.0091 -
0.3660 26400 0.0088 -
0.3674 26500 0.0089 -
0.3688 26600 0.0087 -
0.3701 26700 0.0089 -
0.3715 26800 0.0087 -
0.3729 26900 0.0088 -
0.3743 27000 0.0086 -
0.3757 27100 0.0088 -
0.3771 27200 0.0087 -
0.3785 27300 0.0085 -
0.3798 27400 0.0085 -
0.3812 27500 0.0086 -
0.3826 27600 0.0088 -
0.3840 27700 0.0084 -
0.3854 27800 0.0086 -
0.3868 27900 0.0085 -
0.3882 28000 0.0085 -
0.3895 28100 0.0086 -
0.3909 28200 0.0085 -
0.3923 28300 0.0086 -
0.3937 28400 0.0088 -
0.3951 28500 0.0086 -
0.3965 28600 0.0085 -
0.3979 28700 0.0086 -
0.3993 28800 0.0085 -
0.4 28854 - 0.0031
0.4006 28900 0.0084 -
0.4020 29000 0.0084 -
0.4034 29100 0.0085 -
0.4048 29200 0.0083 -
0.4062 29300 0.0084 -
0.4076 29400 0.0084 -
0.4090 29500 0.0084 -
0.4103 29600 0.0082 -
0.4117 29700 0.0085 -
0.4131 29800 0.0083 -
0.4145 29900 0.0081 -
0.4159 30000 0.0084 -
0.4173 30100 0.0085 -
0.4187 30200 0.0081 -
0.4200 30300 0.0084 -
0.4214 30400 0.0084 -
0.4228 30500 0.0082 -
0.4242 30600 0.0084 -
0.4256 30700 0.0084 -
0.4270 30800 0.0082 -
0.4284 30900 0.0081 -
0.4297 31000 0.0081 -
0.4311 31100 0.0079 -
0.4325 31200 0.0082 -
0.4339 31300 0.0082 -
0.4353 31400 0.0082 -
0.4367 31500 0.0079 -
0.4381 31600 0.0079 -
0.4395 31700 0.0081 -
0.4408 31800 0.008 -
0.4422 31900 0.0081 -
0.4436 32000 0.0081 -
0.4450 32100 0.0081 -
0.4464 32200 0.0078 -
0.4478 32300 0.0079 -
0.4492 32400 0.0081 -
0.4505 32500 0.0081 -
0.4519 32600 0.0081 -
0.4533 32700 0.0079 -
0.4547 32800 0.0079 -
0.4561 32900 0.0079 -
0.4575 33000 0.0079 -
0.4589 33100 0.0079 -
0.4602 33200 0.0078 -
0.4616 33300 0.0077 -
0.4630 33400 0.008 -
0.4644 33500 0.0079 -
0.4658 33600 0.008 -
0.4672 33700 0.0079 -
0.4686 33800 0.0078 -
0.4700 33900 0.008 -
0.4713 34000 0.0077 -
0.4727 34100 0.0077 -
0.4741 34200 0.0078 -
0.4755 34300 0.0076 -
0.4769 34400 0.0078 -
0.4783 34500 0.0078 -
0.4797 34600 0.0078 -
0.4810 34700 0.0079 -
0.4824 34800 0.0078 -
0.4838 34900 0.0077 -
0.4852 35000 0.0075 -
0.4866 35100 0.0076 -
0.4880 35200 0.0078 -
0.4894 35300 0.0076 -
0.4907 35400 0.0078 -
0.4921 35500 0.0077 -
0.4935 35600 0.0076 -
0.4949 35700 0.0076 -
0.4963 35800 0.0077 -
0.4977 35900 0.0076 -
0.4991 36000 0.0077 -
0.5005 36100 0.0077 -
0.5018 36200 0.0077 -
0.5032 36300 0.0077 -
0.5046 36400 0.0076 -
0.5060 36500 0.0076 -
0.5074 36600 0.0077 -
0.5088 36700 0.0076 -
0.5102 36800 0.0075 -
0.5115 36900 0.0077 -
0.5129 37000 0.0076 -
0.5143 37100 0.0075 -
0.5157 37200 0.0074 -
0.5171 37300 0.0074 -
0.5185 37400 0.0075 -
0.5199 37500 0.0075 -
0.5212 37600 0.0074 -
0.5226 37700 0.0074 -
0.5240 37800 0.0072 -
0.5254 37900 0.0076 -
0.5268 38000 0.0075 -
0.5282 38100 0.0072 -
0.5296 38200 0.0074 -
0.5309 38300 0.0073 -
0.5323 38400 0.0073 -
0.5337 38500 0.0074 -
0.5351 38600 0.0073 -
0.5365 38700 0.0073 -
0.5379 38800 0.0074 -
0.5393 38900 0.0072 -
0.5407 39000 0.0076 -
0.5420 39100 0.0072 -
0.5434 39200 0.0073 -
0.5448 39300 0.0071 -
0.5462 39400 0.0072 -
0.5476 39500 0.0073 -
0.5490 39600 0.0074 -
0.5504 39700 0.0072 -
0.5517 39800 0.0072 -
0.5531 39900 0.0073 -
0.5545 40000 0.0071 -
0.5559 40100 0.0072 -
0.5573 40200 0.0072 -
0.5587 40300 0.0071 -
0.5601 40400 0.0072 -
0.5614 40500 0.0071 -
0.5628 40600 0.0073 -
0.5642 40700 0.0073 -
0.5656 40800 0.0072 -
0.5670 40900 0.0071 -
0.5684 41000 0.0073 -
0.5698 41100 0.0072 -
0.5712 41200 0.0071 -
0.5725 41300 0.0074 -
0.5739 41400 0.0072 -
0.5753 41500 0.0071 -
0.5767 41600 0.0071 -
0.5781 41700 0.007 -
0.5795 41800 0.0071 -
0.5809 41900 0.0071 -
0.5822 42000 0.0073 -
0.5836 42100 0.0071 -
0.5850 42200 0.0069 -
0.5864 42300 0.0071 -
0.5878 42400 0.0072 -
0.5892 42500 0.0073 -
0.5906 42600 0.0071 -
0.5919 42700 0.0071 -
0.5933 42800 0.0072 -
0.5947 42900 0.0071 -
0.5961 43000 0.0072 -
0.5975 43100 0.007 -
0.5989 43200 0.0072 -
0.6 43281 - 0.0026
0.6003 43300 0.0071 -
0.6016 43400 0.0069 -
0.6030 43500 0.007 -
0.6044 43600 0.0069 -
0.6058 43700 0.007 -
0.6072 43800 0.0068 -
0.6086 43900 0.0071 -
0.6100 44000 0.0069 -
0.6114 44100 0.0069 -
0.6127 44200 0.0069 -
0.6141 44300 0.0071 -
0.6155 44400 0.0071 -
0.6169 44500 0.007 -
0.6183 44600 0.0069 -
0.6197 44700 0.0069 -
0.6211 44800 0.007 -
0.6224 44900 0.0068 -
0.6238 45000 0.0069 -
0.6252 45100 0.0069 -
0.6266 45200 0.0069 -
0.6280 45300 0.0068 -
0.6294 45400 0.0069 -
0.6308 45500 0.007 -
0.6321 45600 0.0068 -
0.6335 45700 0.0068 -
0.6349 45800 0.0068 -
0.6363 45900 0.0069 -
0.6377 46000 0.007 -
0.6391 46100 0.0067 -
0.6405 46200 0.0066 -
0.6419 46300 0.0069 -
0.6432 46400 0.0068 -
0.6446 46500 0.007 -
0.6460 46600 0.0069 -
0.6474 46700 0.0069 -
0.6488 46800 0.0068 -
0.6502 46900 0.007 -
0.6516 47000 0.0069 -
0.6529 47100 0.0067 -
0.6543 47200 0.0068 -
0.6557 47300 0.0065 -
0.6571 47400 0.0067 -
0.6585 47500 0.007 -
0.6599 47600 0.0067 -
0.6613 47700 0.0067 -
0.6626 47800 0.0068 -
0.6640 47900 0.0067 -
0.6654 48000 0.0066 -
0.6668 48100 0.0069 -
0.6682 48200 0.0067 -
0.6696 48300 0.0067 -
0.6710 48400 0.0067 -
0.6724 48500 0.0069 -
0.6737 48600 0.0066 -
0.6751 48700 0.0066 -
0.6765 48800 0.0068 -
0.6779 48900 0.0067 -
0.6793 49000 0.0067 -
0.6807 49100 0.0068 -
0.6821 49200 0.0066 -
0.6834 49300 0.0067 -
0.6848 49400 0.0065 -
0.6862 49500 0.0067 -
0.6876 49600 0.0066 -
0.6890 49700 0.0065 -
0.6904 49800 0.0067 -
0.6918 49900 0.0066 -
0.6931 50000 0.0066 -
0.6945 50100 0.0066 -
0.6959 50200 0.0065 -
0.6973 50300 0.0068 -
0.6987 50400 0.0068 -
0.7001 50500 0.0066 -
0.7015 50600 0.0067 -
0.7028 50700 0.0068 -
0.7042 50800 0.0066 -
0.7056 50900 0.0065 -
0.7070 51000 0.0065 -
0.7084 51100 0.0065 -
0.7098 51200 0.0066 -
0.7112 51300 0.0065 -
0.7126 51400 0.0064 -
0.7139 51500 0.0063 -
0.7153 51600 0.0064 -
0.7167 51700 0.0063 -
0.7181 51800 0.0064 -
0.7195 51900 0.0065 -
0.7209 52000 0.0065 -
0.7223 52100 0.0065 -
0.7236 52200 0.0065 -
0.7250 52300 0.0065 -
0.7264 52400 0.0065 -
0.7278 52500 0.0065 -
0.7292 52600 0.0064 -
0.7306 52700 0.0065 -
0.7320 52800 0.0064 -
0.7333 52900 0.0064 -
0.7347 53000 0.0065 -
0.7361 53100 0.0063 -
0.7375 53200 0.0063 -
0.7389 53300 0.0064 -
0.7403 53400 0.0064 -
0.7417 53500 0.0064 -
0.7431 53600 0.0066 -
0.7444 53700 0.0064 -
0.7458 53800 0.0063 -
0.7472 53900 0.0064 -
0.7486 54000 0.0063 -
0.7500 54100 0.0063 -
0.7514 54200 0.0062 -
0.7528 54300 0.0064 -
0.7541 54400 0.0063 -
0.7555 54500 0.0063 -
0.7569 54600 0.0062 -
0.7583 54700 0.0063 -
0.7597 54800 0.0062 -
0.7611 54900 0.0062 -
0.7625 55000 0.0063 -
0.7638 55100 0.0065 -
0.7652 55200 0.0064 -
0.7666 55300 0.0062 -
0.7680 55400 0.0064 -
0.7694 55500 0.0063 -
0.7708 55600 0.0063 -
0.7722 55700 0.0062 -
0.7735 55800 0.0063 -
0.7749 55900 0.0062 -
0.7763 56000 0.0063 -
0.7777 56100 0.0064 -
0.7791 56200 0.0062 -
0.7805 56300 0.0065 -
0.7819 56400 0.006 -
0.7833 56500 0.0065 -
0.7846 56600 0.006 -
0.7860 56700 0.0062 -
0.7874 56800 0.0064 -
0.7888 56900 0.0061 -
0.7902 57000 0.0063 -
0.7916 57100 0.0062 -
0.7930 57200 0.0062 -
0.7943 57300 0.0062 -
0.7957 57400 0.0062 -
0.7971 57500 0.0062 -
0.7985 57600 0.0061 -
0.7999 57700 0.0061 -
0.8 57708 - 0.0022
0.8013 57800 0.0064 -
0.8027 57900 0.0062 -
0.8040 58000 0.0063 -
0.8054 58100 0.0061 -
0.8068 58200 0.0061 -
0.8082 58300 0.0063 -
0.8096 58400 0.0062 -
0.8110 58500 0.0062 -
0.8124 58600 0.0061 -
0.8138 58700 0.0062 -
0.8151 58800 0.0061 -
0.8165 58900 0.0061 -
0.8179 59000 0.0062 -
0.8193 59100 0.0062 -
0.8207 59200 0.0061 -
0.8221 59300 0.006 -
0.8235 59400 0.0061 -
0.8248 59500 0.006 -
0.8262 59600 0.006 -
0.8276 59700 0.0061 -
0.8290 59800 0.0062 -
0.8304 59900 0.0059 -
0.8318 60000 0.006 -
0.8332 60100 0.006 -
0.8345 60200 0.0061 -
0.8359 60300 0.006 -
0.8373 60400 0.0059 -
0.8387 60500 0.0061 -
0.8401 60600 0.006 -
0.8415 60700 0.0059 -
0.8429 60800 0.006 -
0.8443 60900 0.0061 -
0.8456 61000 0.0062 -
0.8470 61100 0.006 -
0.8484 61200 0.006 -
0.8498 61300 0.0059 -
0.8512 61400 0.0059 -
0.8526 61500 0.006 -
0.8540 61600 0.006 -
0.8553 61700 0.0059 -
0.8567 61800 0.006 -
0.8581 61900 0.0059 -
0.8595 62000 0.0059 -
0.8609 62100 0.0059 -
0.8623 62200 0.0059 -
0.8637 62300 0.0062 -
0.8650 62400 0.0061 -
0.8664 62500 0.0059 -
0.8678 62600 0.006 -
0.8692 62700 0.0061 -
0.8706 62800 0.0059 -
0.8720 62900 0.0061 -
0.8734 63000 0.006 -
0.8747 63100 0.0059 -
0.8761 63200 0.0059 -
0.8775 63300 0.0057 -
0.8789 63400 0.006 -
0.8803 63500 0.0058 -
0.8817 63600 0.0059 -
0.8831 63700 0.0058 -
0.8845 63800 0.0058 -
0.8858 63900 0.0059 -
0.8872 64000 0.0059 -
0.8886 64100 0.0059 -
0.8900 64200 0.0058 -
0.8914 64300 0.0058 -
0.8928 64400 0.006 -
0.8942 64500 0.0059 -
0.8955 64600 0.0059 -
0.8969 64700 0.0059 -
0.8983 64800 0.0058 -
0.8997 64900 0.0059 -
0.9011 65000 0.0059 -
0.9025 65100 0.0058 -
0.9039 65200 0.0058 -
0.9052 65300 0.0058 -
0.9066 65400 0.0059 -
0.9080 65500 0.0057 -
0.9094 65600 0.0057 -
0.9108 65700 0.0059 -
0.9122 65800 0.0059 -
0.9136 65900 0.0058 -
0.9150 66000 0.0058 -
0.9163 66100 0.0058 -
0.9177 66200 0.0057 -
0.9191 66300 0.0057 -
0.9205 66400 0.0059 -
0.9219 66500 0.0056 -
0.9233 66600 0.0058 -
0.9247 66700 0.0057 -
0.9260 66800 0.0058 -
0.9274 66900 0.0056 -
0.9288 67000 0.0057 -
0.9302 67100 0.0057 -
0.9316 67200 0.0055 -
0.9330 67300 0.0058 -
0.9344 67400 0.0058 -
0.9357 67500 0.0058 -
0.9371 67600 0.0057 -
0.9385 67700 0.0058 -
0.9399 67800 0.0056 -
0.9413 67900 0.0057 -
0.9427 68000 0.0058 -
0.9441 68100 0.0058 -
0.9454 68200 0.0057 -
0.9468 68300 0.0057 -
0.9482 68400 0.0057 -
0.9496 68500 0.0057 -
0.9510 68600 0.0057 -
0.9524 68700 0.0057 -
0.9538 68800 0.0059 -
0.9552 68900 0.0058 -
0.9565 69000 0.0058 -
0.9579 69100 0.0056 -
0.9593 69200 0.0057 -
0.9607 69300 0.0057 -
0.9621 69400 0.0057 -
0.9635 69500 0.0058 -
0.9649 69600 0.0056 -
0.9662 69700 0.0059 -
0.9676 69800 0.0055 -
0.9690 69900 0.0057 -
0.9704 70000 0.0054 -
0.9718 70100 0.0055 -
0.9732 70200 0.0055 -
0.9746 70300 0.0057 -
0.9759 70400 0.0057 -
0.9773 70500 0.0057 -
0.9787 70600 0.0056 -
0.9801 70700 0.0058 -
0.9815 70800 0.0054 -
0.9829 70900 0.0057 -
0.9843 71000 0.0056 -
0.9857 71100 0.0057 -
0.9870 71200 0.0057 -
0.9884 71300 0.0056 -
0.9898 71400 0.0057 -
0.9912 71500 0.0055 -
0.9926 71600 0.0055 -
0.9940 71700 0.0057 -
0.9954 71800 0.0057 -
0.9967 71900 0.0056 -
0.9981 72000 0.0058 -
0.9995 72100 0.0056 -
1.0 72135 - 0.0020

Framework Versions

  • Python: 3.11.13
  • Sentence Transformers: 5.1.2
  • Transformers: 4.57.1
  • PyTorch: 2.8.0+cu129
  • Accelerate: 1.11.0
  • Datasets: 4.3.0
  • Tokenizers: 0.22.1

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",
}

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}
}