Instructions to use Chandar/sv-subject-based-matryoshka-mpnet-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Chandar/sv-subject-based-matryoshka-mpnet-base with sentence-transformers:

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("Chandar/sv-subject-based-matryoshka-mpnet-base")

sentences = [
    "channels. Cl- ions enter the cell and hyperpolarizes the membrane, making the neuron less likely to fire\nan action potential.\nOnce neurotransmission has occurred, the neurotransmitter must be removed from the synaptic\ncleft so the postsynaptic membrane can “reset” and be ready to receive another signal. This can be\naccomplished in three ways: the neurotransmitter can diffuse away from the synaptic cleft, it can be\ndegraded by enzymes in the synaptic cleft, or it can be recycled (sometimes called reuptake) by the\npresynaptic neuron. Several drugs act at this step of neurotransmission. For example, some drugs that\nare given to Alzheimer’s patients work by inhibiting acetylcholinesterase, the enzyme that degrades\nacetylcholine. This inhibition of the enzyme essentially increases neurotransmission at synapses that\nrelease acetylcholine. Once released, the acetylcholine stays in the cleft and can continually bind and\nunbind to postsynaptic receptors.\nNeurotransmitter Function and Location\nNeurotransmitter Example Location\nAcetylcholine — CNS and/or\nPNS\nBiogenic amine Dopamine, serotonin, norepinephrine CNS and/or\nPNS\nAmino acid Glycine, glutamate, aspartate, gamma aminobutyric\nacidCNS\nNeuropeptide Substance P, endorphins CNS and/or\nPNS\nTable 35.2\nElectrical Synapse\nWhile electrical synapses are fewer in number than chemical synapses, they are found in all nervous\nsystems and play important and unique roles. The mode of neurotransmission in electrical synapses is\nquite different from that in chemical synapses. In an electrical synapse, the presynaptic and postsynaptic\nmembranes are very close together and are actually physically connected by channel proteins forming\ngap junctions. Gap junctions allow current to pass directly from one cell to the next. In addition to the\nions that carry this current, other molecules, such as ATP, can diffuse through the large gap junction\npores.\nThere are key differences between chemical and electrical synapses. Because chemical synapses\ndepend on the release of neurotransmitter molecules from synaptic vesicles to pass on their signal, there\nis an approximately one millisecond delay between when the axon potential reaches the presynaptic\nterminal and when the neurotransmitter leads to opening of postsynaptic ion channels. Additionally, this\nsignaling is unidirectional. Signaling in electrical synapses, in contrast, is virtually instantaneous (which\nis important for synapses involved in key reflexes), and some electrical synapses are bidirectional.\nElectrical synapses are also more reliable as they are less likely to be blocked, and they are important\nfor synchronizing the electrical activity of a group of neurons. For example, electrical synapses in the\nthalamus are thought to regulate slow-wave sleep, and disruption of these synapses can cause seizures.\nSignal Summation\nSometimes a single EPSP is strong enough to induce an action potential in the postsynaptic neuron,\nbut often multiple presynaptic inputs must create EPSPs around the same time for the postsynaptic\nneuron to be sufficiently depolarized to fire an action potential. This process is calledsummation and\noccurs at the axon hillock, as illustrated inFigure 35.16. Additionally, one neuron often has inputs\nfrom many presynaptic neurons—some excitatory and some inhibitory—so IPSPs can cancel out EPSPs\nand vice versa. It is the net change in postsynaptic membrane voltage that determines whether the\npostsynaptic cell has reached its threshold of excitation needed to fire an action potential. Together,\nsynaptic summation and the threshold for excitation act as a filter so that random “noise” in the system\nis not transmitted as important information.\n1004 CHAPTER 35 | THE NERVOUS SYSTEM\nThis content is available for free at http://cnx.org/content/col11448/1.9",
    "7.2 \nCHAPTER \nMoray Offshore Renewables Limited - Environmental Statement  \nTelford, Stevenson and MacColl Offshore Wind Farms and Transmission Infrastructure \n \n \nSection 3 – Offshore Generating Station Impact Assessment 7-51 \nTable 7.2-10 Qualitative Assessment for Species not Modelled and Without Defined Surrogates Based \non Potential Magnitude of Effects and Receptor Sensitivities \nSpecies \nPotential \nMagnitude of \nEffect \nSensitivity of Receptor Magnitude \nof Effect \nSandeels Small \n Important prey species; and \n Known to be present in the Moray Firth area.  The results \nof the site specific survey undertaken, however, \nsuggest that within the three proposed wind farm sites \nthere are not extensive areas supporting important \nsandeel populations.  Substrate specific. \nMedium \nElasmobranchs Small \n Most species are of conservation Importance; \n Generally more prevalent in the north and west of \nScotland than in the Moray Firth; and \n Some with nursery grounds defined in the proposed \nsites (spurdog, spotted ray and thornback ray). \nLow-\nMedium \nRiver and Sea \nLamprey Small \n Conservation importance; and \n Potentially transiting the site during migration (lack of \ndetailed information on migration). \nMedium \nAnglerfish Small \n Commercially important; and \n High intensity nursery area in the sites. \nMedium \nHaddock Small to \nMedium \n Commercially important; and \n Nursery grounds in the area and spawning grounds in \nthe proximity of the proposed sites, however \ncomparatively large. \nLow \nEuropean Eel Small to \nMedium \n Conservation importance; and \n Potentially transiting the site during migration (lack of \ndetailed information on migration). \nMedium \nSprat Medium \n Important as prey species; and \n Spawning and nursery grounds in the area, however \nthese are comparatively large. \nLow- \nMedium \nLife Stages of Limited Mobility \n7.2.5.47 Life stages of limited mobility such as larvae, and in the case of European eel, their \njuvenile form (glass eels), will not be able to avoid areas where the highest noise \nlevels are reached during construction, assum ing they drift through the proposed \nwind farm sites .  Although there is limited information on the effect of piling noise to \ndate on early life stages of fish, research recently carried out by the Institute for \nMarine Resources and Ecosystem Studies (IMARES) (Bolle et al., 2011) suggests that \nthe assumption of 100 % of larvae mortality within a radius of 1,000  m around a piling \nsite (used in the Appropriate Assessment of Dutch offshore wind farms) is too \nconservative.  Bolle et al ., (2011) found no significant effects in the larval stages \nanalysed at the highe st exposure level (cumulative SEL = 206  dB re 1µPa 2s) which \nrepresented 100 pulses at a distance of 100  m from piling.  It is recognised that the \nresults, based on sole ( Solea solea) larvae, should not be extrapolated to fish larvae \nin general as inter -specific differences in vulnerability to sound exposure may exist .  \nThe findings, however suggest that larval mortality would only occur within a few",
    "168 7.17. Ferromagnetism\n0\n0.5\n1\n1.5\n10−6 ρ/χ (kg m−3)\n0 300 600 900 1200 1500\nT(K)\nFigure 7.11: Plot ofρ/χ(where ρis the mass density) versus temperature for gadolinium above\nits Curie temperature. The curve is (apart from some slight departures at high temperatures) a\nstraight-line, in accordance with the Curie-Weiss law. The intercept of the line with the temperature\naxis givesTc =310 K. The metal becomes ferromagneticbelow 289 K. Data from S. Arajs, and\nR.V . Colvin, J. appl. Phys.32, 336S (1961).\nHere, r /nequal1. Hence, deduce that\nS ∞= a\n1 −r,\nassuming that 0<r <1.\n7.2 Let\nI(n) =\n∫ ∞\n0\nxn e−αx2\ndx.\nDemonstrate that\nI(n) =−∂I(n −2)\n∂α .\nFurthermore, show that\nI(0) =\n√\nπ\n2 α1/2\n(see Exercise 2.2), and\nI(1) = 1\n2 α.",
    "416 10 Spontaneous Symmetry Breaking\n10.9 The Goldstone Theorem\nSo far, we have only discussed the simplest of all symmetry groups, namelyG = Z2,\nwhich is both ﬁnite and abelian. Although it will not change our picture of SSB ,\nfor the sake of completeness (and interest to foundations) we also present a brief\nintroduction to continuous symmetries, culminating in the Goldstone Theorem and\nthe Higgs mechanism (which at ﬁrst sight contradict each other and hence require a\nvery careful treatment). The former results when the broken symmetry groupG is a\nLie group, whereas the latter arises when it is an inﬁnite-dimensional gauge group.\nLet us start with the simple case G = SO(2), acting on R2 by rotation. This\ninduces the obvious action on the classical phase space T ∗R2, i.e.,\nR(p,q)=( Rp,Rq), (10.300)\ncf. (3.94), as well as on the quantum Hilbert space H = L2(R2), that is,\nuRψ(x)= ψ(R−1x). (10.301)\nLet us see what changes with respect to the action of Z2 on R considered in §10.1.\nWe now regard the double-well potentialV in (10.11) as anSO(2)-invariant function\non R2 through the reinterpretation ofx2 as x2\n1 +x2\n2. This is theMexican hat potential.\nThus the classical Hamiltonianh(p,q)= p2/2m+V (q), similarly with p2 = p2\n1 + p2\n2,\nis SO(2)-invariant, and the set of classical ground states\nE0 = {(p,q) ∈T ∗R2 |p = 0,q2 = a2} (10.302)\nis the SO(2)-orbit through e.g. the point (p1 = p2 = 0,q1 = a,q2 = 0). Unlike the\none-dimensional case, the set of ground states is now connected and forms a cir-\ncle in phase space, on which the symmetry group SO(2) acts. The intuition behind\nthe Goldstone Theorem is that a particle can freely move in this circle at no cost\nof energy. If we look at mass as inertia, such motion is “massless”, as there is no\nobstruction. However, this intuition is only realized in quantum ﬁeld theory. In quan-\ntum mechanics, the ground state of the Hamiltonian (10.6) (now acting onL2(R2))\nremains unique, as in the one-dimensional case. In polar coordinates (r,φ) we have\nh¯h = −¯h2\n2m\n(∂2\n∂r2 + 1\nr\n∂\n∂r + 1\nr2\n∂2\n∂φ2\n)\n+V (r), (10.303)\nwith V (r)= 1\n4 λ(r2 −a2)2. With\nL2(R2) ∼= L2(R+) ⊗ℓ2(Z) (10.304)\nunder Fourier transformation in the angle variable, this becomes\nh¯hψ(r,n)=\n(\n−¯h2\n2m\n(∂2\n∂r2 + 1\nr\n∂\n∂r −n2\nr2\n)\n+V (r)\n)\nψ(r,n). (10.305)"
]
embeddings = model.encode(sentences)

similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [4, 4]

Notebooks
Google Colab
Kaggle

SentenceTransformer based on microsoft/mpnet-base

This is a sentence-transformers model finetuned from microsoft/mpnet-base. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: microsoft/mpnet-base
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 dimensions
Similarity Function: Cosine Similarity

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': False}) with Transformer model: MPNetModel 
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

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("Chandar/sv-subject-based-matryoshka-mpnet-base")
# Run inference
sentences = [
    'channels. Cl- ions enter the cell and hyperpolarizes the membrane, making the neuron less likely to fire\nan action potential.\nOnce neurotransmission has occurred, the neurotransmitter must be removed from the synaptic\ncleft so the postsynaptic membrane can “reset” and be ready to receive another signal. This can be\naccomplished in three ways: the neurotransmitter can diffuse away from the synaptic cleft, it can be\ndegraded by enzymes in the synaptic cleft, or it can be recycled (sometimes called reuptake) by the\npresynaptic neuron. Several drugs act at this step of neurotransmission. For example, some drugs that\nare given to Alzheimer’s patients work by inhibiting acetylcholinesterase, the enzyme that degrades\nacetylcholine. This inhibition of the enzyme essentially increases neurotransmission at synapses that\nrelease acetylcholine. Once released, the acetylcholine stays in the cleft and can continually bind and\nunbind to postsynaptic receptors.\nNeurotransmitter Function and Location\nNeurotransmitter Example Location\nAcetylcholine — CNS and/or\nPNS\nBiogenic amine Dopamine, serotonin, norepinephrine CNS and/or\nPNS\nAmino acid Glycine, glutamate, aspartate, gamma aminobutyric\nacidCNS\nNeuropeptide Substance P, endorphins CNS and/or\nPNS\nTable 35.2\nElectrical Synapse\nWhile electrical synapses are fewer in number than chemical synapses, they are found in all nervous\nsystems and play important and unique roles. The mode of neurotransmission in electrical synapses is\nquite different from that in chemical synapses. In an electrical synapse, the presynaptic and postsynaptic\nmembranes are very close together and are actually physically connected by channel proteins forming\ngap junctions. Gap junctions allow current to pass directly from one cell to the next. In addition to the\nions that carry this current, other molecules, such as ATP, can diffuse through the large gap junction\npores.\nThere are key differences between chemical and electrical synapses. Because chemical synapses\ndepend on the release of neurotransmitter molecules from synaptic vesicles to pass on their signal, there\nis an approximately one millisecond delay between when the axon potential reaches the presynaptic\nterminal and when the neurotransmitter leads to opening of postsynaptic ion channels. Additionally, this\nsignaling is unidirectional. Signaling in electrical synapses, in contrast, is virtually instantaneous (which\nis important for synapses involved in key reflexes), and some electrical synapses are bidirectional.\nElectrical synapses are also more reliable as they are less likely to be blocked, and they are important\nfor synchronizing the electrical activity of a group of neurons. For example, electrical synapses in the\nthalamus are thought to regulate slow-wave sleep, and disruption of these synapses can cause seizures.\nSignal Summation\nSometimes a single EPSP is strong enough to induce an action potential in the postsynaptic neuron,\nbut often multiple presynaptic inputs must create EPSPs around the same time for the postsynaptic\nneuron to be sufficiently depolarized to fire an action potential. This process is calledsummation and\noccurs at the axon hillock, as illustrated inFigure 35.16. Additionally, one neuron often has inputs\nfrom many presynaptic neurons—some excitatory and some inhibitory—so IPSPs can cancel out EPSPs\nand vice versa. It is the net change in postsynaptic membrane voltage that determines whether the\npostsynaptic cell has reached its threshold of excitation needed to fire an action potential. Together,\nsynaptic summation and the threshold for excitation act as a filter so that random “noise” in the system\nis not transmitted as important information.\n1004 CHAPTER 35 | THE NERVOUS SYSTEM\nThis content is available for free at http://cnx.org/content/col11448/1.9',
    'WHERESELENIUMPOISONING\nIncertainportionsoftheNorthCentralgreatplains,plantsabsorbenoughsele-niumfromthesoiltoinjureanimalsthatfeeduponthem.Thepoisoningmayresultinaslowdiseaseknownas"blindstaggers"oras"alkalidisease",oritmaybequicklyfatal.Asaresultoftheselenium,thejointsoftheleg-bonesbecomebadlyeroded.Thehoofsdevelopabnormalitiesordropoff.Locomotionisimpaired.Theeffectoftheseleniumper-sists,fortheanimalsdonotusuallyrecoverevenifre-movedfromsucharegionandfedagoodrationOCCURSironisanessentialconstituentofthehemoglobin(seepage205).Coppercompoundsaregenerallypoisonoustomostkindsofprotoplasm;yetforsomespeciescopperisnecessaryinsmallamounts.Copperisanessentialelementinthebluishoxygen-carrierhemocyaninofthekingcrabandthelobster.InsomeoftheWesternstatesthesoilcontainstheelementselenium.Thiselementispresentalsoinplantsgrowinginsuchsoil,althoughitdoesnotappeartoaffecttheminanyway.Butanimalsthatfeeduponsuchplantsareoftenseriouslypoisoned(seeillustrationopposite).Inotherregionsvariationintheamountoffluorineinthesoilmaybeimportanttous.Theelementfluorine,whichisverywidelybutunevenlydistributed,seemstoplayaroleintheassimilationofcalciumandphosphorus,andsoaffectstheformationoftheteeth.Astudyof7000girlsandboysofhigh-schoolageinvariousmiddleandsouthwesternstatesbroughtoutthefactthattherewasmuchmoretoothdecay,orcaries^incommunitieswhosewatersupplieswerefreeoffluorinethanincommunitiesusingwaterwith0.5ormorepartsfluorinepermillionpartswater.Thus,thepopulationofacertainpartofTexas,DeafSmithCounty,wasfoundtohaveanexceptionallylownumberofdecayedteeth;andthisrelativefreedomfromdentalcariesisasso-ciatedwithmorethanusualamountsoffluorineinthelocalwaters.Inotherregionsunusualamountsoffluorineinthesoilandsoilwatersapparentlybringaboutthedevelopmentof"mottledteeth"amongthechildrenlivingthere.Nobodywantsblotchyteeth,butnobodywantscaries102',
    "INDEX 275\nSenescence\n1\n, 27-30,46,7078\ntheoiietiof,43f>0\nSenilechangesinneivecells,27-29\nSenilityat,causeofdeath,1011\ninplants,44,71,75\nSepiK.rniia,231,282\nSeitna,252,258\nMeium,influenceontissuecultme,\n70,77\nSe\\organs,107,IOS,111,121-125,\n217219\nSexualiepioduction,3741\nShell,J, 20,27\nSkeletalsyhtem,107,108,112,127,\n128\nskin,107,H)8,no,112,331,132\nSlonakei,,T I?,212,213,218,228,\n2U7\nSloiopolhki,B, 33,2(>7\nSnow,12 C,179-383,225,267\nSofteningofthe-hiain,231,2.12\nSoma,40\nSomaticcella,ranmntalityof,5878\nSpan,174,175\nSpiegelberg,W, 87\nSpmtuahsm,18-20\nSpleen,61\nSponges,62\nStatine,174,175\nStomach,M,217-210,207\nStenuntomum,35,36\nSievenflon,THC.,206-208,2(>7\nStillbirths,205\n$ ongyloocnt')otuspitrpmnIus,55,\n56\nHummaiyofreaultH,223-227\nf-hxi'vivorfiluplinesofDrosophilo,\n188,192,195\nSyphihH,123\nTable,life,70-82\nTempeiatuic,208-217\nTethehn,70,220222\nTbeoiiesofdeath,4350\nTheoiyofpopulationgiowtli,249\nThyioulgland,01\nTissuecnltmet Ditto,5878\nTianaplantationoftumois,04,(51\nTubeicuIomH,101,204,208,230,2!1,\n238\nTuiuoiti.uihpliinlation,61, 65\nTyphoidfcvei,230,2,11,2.J5,2,'iO\nUnitedStates,giowtliol,2502,12,\n254-257\nLi'iostylayiandis,72\nVanBuien,GII, 11J,2UO\nVariation,genetic,190\nVeneiwildiseases,123,124\nVeilmlHt,PF,249,2(57\nVerwoin,M,,44,207\nVienna,245,246\nVoumoll,217\nWallei,AD,216,207\nWalwoith,BH,152,267\nWar,243\nWedekmd,33,267\nWeiimann,A, 26,43,65,207\nWlialc,longevityof,22\nWilaon,HV,62,267\nWittalom,99\nWoodruff,LL,30,33,72,73,267,\n268\nWoodn,FA, 38,3,208\nYellowfever,240,21-2\nYoung,TTO,,23-25,268",
]
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

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 768
}
```

Metric	Value
cosine_accuracy	0.7608
cosine_accuracy_threshold	0.956
cosine_f1	0.5962
cosine_f1_threshold	0.9436
cosine_precision	0.5779
cosine_recall	0.6158
cosine_ap	0.6558
cosine_mcc	0.3892

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 512
}
```

Metric	Value
cosine_accuracy	0.7512
cosine_accuracy_threshold	0.9695
cosine_f1	0.5758
cosine_f1_threshold	0.96
cosine_precision	0.579
cosine_recall	0.5726
cosine_ap	0.6315
cosine_mcc	0.3695

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 256
}
```

Metric	Value
cosine_accuracy	0.7688
cosine_accuracy_threshold	0.8354
cosine_f1	0.6109
cosine_f1_threshold	0.7678
cosine_precision	0.5523
cosine_recall	0.6833
cosine_ap	0.6883
cosine_mcc	0.3938

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 128
}
```

Metric	Value
cosine_accuracy	0.755
cosine_accuracy_threshold	0.8873
cosine_f1	0.5923
cosine_f1_threshold	0.8024
cosine_precision	0.5241
cosine_recall	0.6809
cosine_ap	0.6656
cosine_mcc	0.3587

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 64
}
```

Metric	Value
cosine_accuracy	0.7398
cosine_accuracy_threshold	0.9206
cosine_f1	0.5858
cosine_f1_threshold	0.8388
cosine_precision	0.5022
cosine_recall	0.7027
cosine_ap	0.6336
cosine_mcc	0.3404

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 32
}
```

Metric	Value
cosine_accuracy	0.7362
cosine_accuracy_threshold	0.9587
cosine_f1	0.5954
cosine_f1_threshold	0.8968
cosine_precision	0.5004
cosine_recall	0.735
cosine_ap	0.6264
cosine_mcc	0.3528

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 16
}
```

Metric	Value
cosine_accuracy	0.7298
cosine_accuracy_threshold	0.9745
cosine_f1	0.5948
cosine_f1_threshold	0.929
cosine_precision	0.5095
cosine_recall	0.7143
cosine_ap	0.6023
cosine_mcc	0.3555

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 8
}
```

Metric	Value
cosine_accuracy	0.708
cosine_accuracy_threshold	0.9946
cosine_f1	0.518
cosine_f1_threshold	0.9767
cosine_precision	0.4194
cosine_recall	0.6772
cosine_ap	0.5203
cosine_mcc	0.2049

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 4
}
```

Metric	Value
cosine_accuracy	0.6796
cosine_accuracy_threshold	0.9989
cosine_f1	0.4953
cosine_f1_threshold	-0.7454
cosine_precision	0.3291
cosine_recall	1.0
cosine_ap	0.4414
cosine_mcc	0.014

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 2
}
```

Metric	Value
cosine_accuracy	0.6728
cosine_accuracy_threshold	1.0
cosine_f1	0.4953
cosine_f1_threshold	-0.7795
cosine_precision	0.3291
cosine_recall	1.0
cosine_ap	0.3823
cosine_mcc	0.014

Binary Classification

Evaluated with BinaryClassificationEvaluator with these parameters:
```
{
    "truncate_dim": 1
}
```

Metric	Value
cosine_accuracy	0.6708
cosine_accuracy_threshold	1.0
cosine_f1	0.4953
cosine_f1_threshold	-1.0
cosine_precision	0.3293
cosine_recall	0.9994
cosine_ap	0.3352
cosine_mcc	0.0

Training Details

Training Dataset

Unnamed Dataset

Size: 1,124,250 training samples
Columns: sentence1, sentence2, and label
Approximate statistics based on the first 1000 samples:
sentence1 sentence2 label
type string string int
details
min: 512 tokens
mean: 512.0 tokens
max: 512 tokens

min: 23 tokens
mean: 432.19 tokens
max: 512 tokens

0: ~50.10%
1: ~49.90%

	sentence1	sentence2	label
type	string	string	int
details	min: 512 tokens mean: 512.0 tokens max: 512 tokens	min: 23 tokens mean: 432.19 tokens max: 512 tokens	0: ~50.10% 1: ~49.90%

Samples:

sentence1	sentence2	label
channels. Cl- ions enter the cell and hyperpolarizes the membrane, making the neuron less likely to fire an action potential. Once neurotransmission has occurred, the neurotransmitter must be removed from the synaptic cleft so the postsynaptic membrane can “reset” and be ready to receive another signal. This can be accomplished in three ways: the neurotransmitter can diffuse away from the synaptic cleft, it can be degraded by enzymes in the synaptic cleft, or it can be recycled (sometimes called reuptake) by the presynaptic neuron. Several drugs act at this step of neurotransmission. For example, some drugs that are given to Alzheimer’s patients work by inhibiting acetylcholinesterase, the enzyme that degrades acetylcholine. This inhibition of the enzyme essentially increases neurotransmission at synapses that release acetylcholine. Once released, the acetylcholine stays in the cleft and can continually bind and unbind to postsynaptic receptors. Neurotransmitter Function and Location N...	`Figure 25.6 This table shows the major divisions of green plants. Which of the following statements about plant divisions is false? a. Lycophytes and pterophytes are seedless vascular plants. b. All vascular plants produce seeds. c. All nonvascular embryophytes are bryophytes. d. Seed plants include angiosperms and gymnosperms. 25.2`	Green Algae: Precursors of Land Plants By the end of this section, you will be able to: • Describe the traits shared by green algae and land plants • Explain the reasons why Charales are considered the closest relative to land plants • Understand that current phylogenetic relationships are reshaped by comparative analysis of DNA sequences Streptophytes Until recently, all photosynthetic eukaryotes were considered members of the kingdom Plantae. The brown, red, and gold algae, however, have been reassigned to the Protista kingdom. This is because apart from their ability to capture light energy and fix CO2, they lack many structural and biochemical traits ...
channels. Cl- ions enter the cell and hyperpolarizes the membrane, making the neuron less likely to fire an action potential. Once neurotransmission has occurred, the neurotransmitter must be removed from the synaptic cleft so the postsynaptic membrane can “reset” and be ready to receive another signal. This can be accomplished in three ways: the neurotransmitter can diffuse away from the synaptic cleft, it can be degraded by enzymes in the synaptic cleft, or it can be recycled (sometimes called reuptake) by the presynaptic neuron. Several drugs act at this step of neurotransmission. For example, some drugs that are given to Alzheimer’s patients work by inhibiting acetylcholinesterase, the enzyme that degrades acetylcholine. This inhibition of the enzyme essentially increases neurotransmission at synapses that release acetylcholine. Once released, the acetylcholine stays in the cleft and can continually bind and unbind to postsynaptic receptors. Neurotransmitter Function and Location N...	NATURALDEATH,PUBLICHEALTH229 andpublichealthhave,byallodds,thebestcasewhen measuredintermsofaccomplishment.Man'sexpecta- tionoflifehasincreasedashehascomedownthrough thecenturies(cf.PearsonandMacdonell.)Alarge partofthisimprovementmustsurelybecreditedtohis improvedunderstandingofhowtocopewithanalways moreorlessinimicalenvironmentandassuageitsasper- itiestohisgreatercomfortandwell-being.Tqfailto givethiscreditwouldbemanifestlyabsurd. Butitwouldbeequallyabsurdtoattempttomain- tainthatalldeclineinthedeath-ratewhichhasoccurred liasbeenduetotheeffortsofhealthofficials,whether consciousorunconscious,asisoftenassertedandstill moreoftenimpliedintheimpassionedoutpouringsof zealouspropagandists.Theopen-mindedstudentofthe naturalhistoryofdiseaseknqwsperfectlywellthata largepartoftheimprovementintherateofmortality cannotpossiblyhavebeenduetoanysuchefforts.To illustratethepoint,Ihavepreparedaseriesofillustra- tionsdealingwithconditionsintheEegistrationArea oftheUnitedStatesintheimmediatepast.Allthe...	`1`
channels. Cl- ions enter the cell and hyperpolarizes the membrane, making the neuron less likely to fire an action potential. Once neurotransmission has occurred, the neurotransmitter must be removed from the synaptic cleft so the postsynaptic membrane can “reset” and be ready to receive another signal. This can be accomplished in three ways: the neurotransmitter can diffuse away from the synaptic cleft, it can be degraded by enzymes in the synaptic cleft, or it can be recycled (sometimes called reuptake) by the presynaptic neuron. Several drugs act at this step of neurotransmission. For example, some drugs that are given to Alzheimer’s patients work by inhibiting acetylcholinesterase, the enzyme that degrades acetylcholine. This inhibition of the enzyme essentially increases neurotransmission at synapses that release acetylcholine. Once released, the acetylcholine stays in the cleft and can continually bind and unbind to postsynaptic receptors. Neurotransmitter Function and Location N...	through a mass which is now formed out of sugar and is now dissolved again into sugar. For it appears that this soluble form of carbo-hydrate, taken into the liver from the intestine, is there, when not immediately needed, stored up in the form of glycogen, ready to be re-dissolved and carried into the system either for immediate use or for re-deposit as glycogen at the places where it is presently to be consumed: the great deposit in the liver and the minor deposits in the muscles being, to use the simile of Prof. Michael Foster, analogous in their functions to a central bank and branch banks. An instructive parallelism may be noted between these processes carried on in the animal organism and those carried on in the vegetal organism. For the carbo- hydrates named, easily made to assume the soluble or the insoluble form by the addition or subtraction of a molecule of water, and thus fitted sometimes for distribution and sometimes for accumulation, are similarly dealt with in the two c...	`1`

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "CoSENTLoss",
    "matryoshka_dims": [
        768,
        512,
        256,
        128,
        64,
        32,
        16,
        8,
        4,
        2,
        1
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1,
        1,
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 16
per_device_eval_batch_size: 32
learning_rate: 2e-05
weight_decay: 0.01
max_steps: 2000

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 16
per_device_eval_batch_size: 32
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_steps: None
torch_empty_cache_steps: None
learning_rate: 2e-05
weight_decay: 0.01
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 3.0
max_steps: 2000
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.0
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
save_safetensors: True
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
no_cuda: False
use_cpu: False
use_mps_device: False
seed: 42
data_seed: None
jit_mode_eval: False
use_ipex: False
bf16: False
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: 0
ddp_backend: None
tpu_num_cores: None
tpu_metrics_debug: False
debug: []
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
past_index: -1
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: False
ignore_data_skip: False
fsdp: []
fsdp_min_num_params: 0
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
fsdp_transformer_layer_cls_to_wrap: None
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
dataloader_pin_memory: True
dataloader_persistent_workers: False
skip_memory_metrics: True
use_legacy_prediction_loop: False
push_to_hub: False
resume_from_checkpoint: None
hub_model_id: None
hub_strategy: every_save
hub_private_repo: None
hub_always_push: False
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: False
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: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	cosine_ap
-1	-1	-	0.3352
0.0071	500	13.3939	-
0.0142	1000	3.2648	-
0.0213	1500	2.8893	-
0.0285	2000	2.9935	-

Framework Versions

Python: 3.12.9
Sentence Transformers: 4.1.0
Transformers: 4.53.0
PyTorch: 2.7.1
Accelerate: 1.8.1
Datasets: 3.6.0
Tokenizers: 0.21.2

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}
}

CoSENTLoss

@online{kexuefm-8847,
    title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
    author={Su Jianlin},
    year={2022},
    month={Jan},
    url={https://kexue.fm/archives/8847},
}

Downloads last month: 58

Safetensors

Model size

0.1B params

Tensor type

F32

Model tree for Chandar/sv-subject-based-matryoshka-mpnet-base

Base model

microsoft/mpnet-base

Finetuned

(134)

this model

Papers for Chandar/sv-subject-based-matryoshka-mpnet-base

Matryoshka Representation Learning

Paper • 2205.13147 • Published May 26, 2022 • 28

Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks

Paper • 1908.10084 • Published Aug 27, 2019 • 16

Evaluation results

Cosine Accuracy on Unknown
self-reported

0.761
Cosine Accuracy Threshold on Unknown
self-reported

0.956
Cosine F1 on Unknown
self-reported

0.596
Cosine F1 Threshold on Unknown
self-reported

0.944
Cosine Precision on Unknown
self-reported

0.578
Cosine Recall on Unknown
self-reported

0.616
Cosine Ap on Unknown
self-reported

0.656
Cosine Mcc on Unknown
self-reported

0.389