dnth/setfit-riasec-classifier-subset

SetFit with sentence-transformers/paraphrase-mpnet-base-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-mpnet-base-v2 as the Sentence Transformer embedding model. A LogisticRegression instance is used for classification.

The model has been trained using an efficient few-shot learning technique that involves:

1. Fine-tuning a Sentence Transformer with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.

Model Details

Model Description

• Model Type: SetFit
• Sentence Transformer body: sentence-transformers/paraphrase-mpnet-base-v2
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 512 tokens
• Number of Classes: 4 classes

Model Sources

• Repository: SetFit on GitHub
• Paper: Efficient Few-Shot Learning Without Prompts
• Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts

Model Labels

Label	Examples
IRC	'The Industrial Process Optimization Specialist investigates inefficiencies in Singapore’s chemical processing plants by analyzing real-time sensor data from reactors, pumps, and flow meters. He/She conducts on-site inspections of pipelines, valves, and safety interlocks, identifying mechanical wear or misalignments that reduce yield. He develops statistical models to correlate process parameters with product quality deviations, then tests hypotheses through controlled operational adjustments. He recalibrates control system sensors, replaces faulty transmitters, and documents every modification in compliance with PEHS and ISO 14001 standards. He reviews batch production records and identifies procedural deviations that introduce variability into output. He balances engineering intuition with algorithmic analysis, preferring evidence derived from physical observation and data triangulation. He is mechanically skilled, logically systematic, and unwavering in his adherence to documented protocols and safety benchmarks.' 'The Supply Chain Integrity Inspector evaluates logistics operations across Singapore’s port and warehousing hubs to ensure adherence to safety and documentation protocols. He/She systematically inspects cargo handling equipment, verifies container seals, and records physical conditions of stored goods using digital inspection checklists. He analyzes shipment delay trends, traces logistical anomalies to operational bottlenecks, and recommends mechanical or procedural fixes based on forensic review of workflow data. He operates forklifts and material handling devices to conduct on-ground assessments and collaborates with engineering teams to maintain calibration of weighing and scanning systems. He ensures all records—paper and digital—reflect accurate inventory counts, customs declarations, and import/export certifications. He applies logical deduction to isolate systemic failures and uses statistical tools to forecast risk areas in supply chain flows. He is dexterous in physical task execution, observant in detail, and meticulous in maintaining structured documentation under stringent trade compliance standards.' 'The Quality Assurance Analyst is responsible for designing and executing test protocols to validate the integrity of manufacturing systems used in Singapore’s pharmaceutical and biomedical production facilities. He/She analyzes process deviation data, identifies root causes of equipment inconsistencies, and develops corrective action models based on empirical observations. He works hands-on with automated production lines, calibrates precision instrumentation, and conducts on-site inspections to ensure compliance with Health Sciences Authority regulations. He maintains rigorous documentation of test results, batch records, and equipment logs using standardized digital systems. He applies statistical methods to detect anomalies in production outputs and collaborates with engineering teams to refine workflow sequences. He is methodical in his approach, demonstrates precision in mechanical adjustments, and thrives in environments where procedural accuracy and data-driven problem-solving converge. He remains current with Singapore’s evolving Good Manufacturing Practice standards and integrates digital monitoring tools to enhance traceability and reliability.'
EIC	'The Workforce Innovation Lead identifies emerging skill needs in Singapore’s digital economy by analysing labour market signals, industry trends, and skills gap reports. He/She designs and pilots scalable upskilling pilots, secures partnerships with edtech providers, and negotiates industry-wide certification standards to enhance workforce mobility. He translates predictive analytics on job role evolution into actionable learning pathways, ensuring alignment with SkillsFuture credit uptake and employer-driven micro-credentialing. He enforces rigorous data entry protocols for participation tracking, maintains compliance with MEA’s training reporting requirements, and structures reporting dashboards for stakeholder review. He is a forward-thinking influencer with a sharp analytical edge, skilled at convincing businesses to invest in future-proofing their teams. He operates in a rapidly shifting landscape where regulatory compliance, data integrity, and persuasive strategy must converge seamlessly.' 'The Organisational Development Director spearheads enterprise-wide culture and capability initiatives, aligning internal structures with business transformation goals. He/She investigates workforce engagement data, uncovers latent leadership gaps through behavioural analytics, and designs intervention frameworks that drive accountability and performance. He persuades C-suite stakeholders to fund high-impact learning ecosystems and reallocates budget across departments to maximise ROI on talent development. He insists on meticulous documentation of competency progressions, ensures audit-ready records of training outcomes, and maintains compliance with MOM’s guidelines on continuous learning. He is a decisive leader with a methodical analytical mindset, able to turn complex HR datasets into compelling business cases. He navigates Singapore’s competitive labour market by strategically blending data insight with executive influence to shape sustainable organisational change.' 'The HR Strategy & Analytics Lead designs and drives talent initiatives that align with business growth objectives, leveraging data-driven insights to influence executive decision-making. He/She identifies workforce trends through advanced analytics of HR metrics, models future competency gaps, and recommends scalable interventions that enhance productivity and retention. He leads cross-functional stakeholder engagements to secure buy-in for talent programs, negotiates budgets for HR technology adoption, and ensures alignment with Singapore’s SkillsFuture framework and tripartite labour standards. He applies structured data governance protocols to maintain compliance in employee reporting and proficiency tracking. He is a persuasive communicator with strong logical reasoning abilities, adept at transforming complex analytics into actionable executive narratives. He operates in a dynamic environment where regulatory shifts and digital transformation require continuous adaptation, and he consistently influences organisational priorities through evidence-based persuasion.'
ESC	'The Talent Acquisition Manager owns the end-to-end recruitment strategy for high-growth divisions in Singapore’s financial services and tech sectors. He/She identifies talent needs, negotiates competitive compensation packages with hiring managers, and leads employer branding campaigns to attract top-tier candidates in a tight labour market. He/She mentors recruiters, conducts de-briefs with new hires to enhance onboarding quality, and champions inclusive hiring practices aligned to Singapore’s Multi-Racial and Inclusive Workplace framework. He/She maintains candidate databases, ensures compliance with the Personal Data Protection Act, and reports on recruitment KPIs with precision using HRIS platforms. He is a results-driven negotiator with deep empathy for candidate experiences, capable of balancing commercial urgency with human-centred processes. He thrives in fast-paced environments where persuasion, relationship-building, and data accuracy are critical to sustaining workforce pipelines.' 'The HR Operations & Compliance Officer leads the implementation of HR policies that streamline workforce processes while enhancing employee experience across Singapore’s service industry. He/She negotiates with vendors for HR tech solutions, champions process automation to reduce manual errors, and motivates cross-functional teams to adopt new HR systems aligned with SkillsFuture digital priorities. He/She conducts regular staff briefings on leave, attendance, and benefits, provides coaching on HR protocols, and personally resolves sensitive employee queries with discretion and empathy. He/She ensures flawless maintenance of personnel records, timely submissions to MOM and CPF Board, and audit readiness through rigorous adherence to data governance standards. He is a confident influencer who bridges frontline support with operational rigour, combining persuasive leadership with meticulous attention to regulatory detail in Singapore’s highly regulated workplace environment.' 'The Rewards & Benefits Specialist designs and launches incentive structures that boost employee motivation in Singapore’s competitive IT and aerospace industries. He/She negotiates with external providers for health, wellness, and fintech-linked benefits, persuades budgets to align compensation with market benchmarks, and presents value propositions to leadership to secure investment in talent retention. He/She conducts benefit orientation sessions, counsels staff on retirement planning and insurance options, and responds to individual queries with patience and clarity. He/She maintains precise records of eligibility, entitlements, and claims across integrated HR systems, ensuring full compliance with CPF and tax regulations. He is a persuasive strategist with a service-oriented approach, combining commercial acumen with compassionate communication and unwavering accuracy in administrative execution.'
SEC	'The Learning Experience Designer develops training curricula focused on soft skills development—communication, empathy, and collaboration—for Singapore’s public sector and healthcare institutions. He/She delivers interactive modules that equip frontline staff with tools to navigate difficult conversations and serve vulnerable populations with dignity. He persuades department leads to embed recurring learning into workflow cycles, demonstrating value through pre- and post-assessment data. He maintains structured learning records, accreditation logs, and compliance reports to meet MOH and SG-Safety standards. He is a patient, intuitive educator who adapts content to varied literacy levels and cultural backgrounds. His work supports Singapore’s national push for humane, patient-centred care, where emotional intelligence is as vital as clinical competence. He thrives in environments that demand both regulatory precision and heartfelt human engagement.' 'The Workplace Learning Coordinator facilitates skill-building interventions for Singapore’s ageing workforce in sectors like healthcare and eldercare, helping employees adapt to new technologies and evolving care standards. He/She designs accessible training materials and conducts small-group coaching sessions that prioritise dignity, patience, and peer learning. He works with union representatives and management to secure buy-in for continuous learning, often balancing budget constraints with human development goals. He tracks attendance, competency assessments, and certification renewals using government-linked platforms like mySkillsFuture, ensuring audit compliance. He is a principled, empathetic guide who builds trust through consistency and genuine care. His approach reflects Singapore’s commitment to lifelong employability, where tailored support enables older workers to remain valued, capable contributors in a rapidly changing economy.' 'The People Culture Coordinator leads employee engagement campaigns that promote mental wellness, inclusion, and recognition across Singapore’s tech and professional services firms. He/She organises team-building activities, employee resource groups, and well-being weeks tailored to local customs and multiracial sensitivities. He champions cross-departmental collaboration by influencing managers to allocate time and resources for non-performance-related activities that build trust. He maintains structured calendars, budget trackers, and vendor procurement records, ensuring all initiatives are transparently funded and compliant with corporate policies. He is a compelling storyteller who inspires participation through relatable messaging and authentic recognition. He operates in a fast-evolving workplace landscape where cultural inclusion and psychological safety are key differentiators, and aligns all programmes with Singapore’s Tripartite Standards on Fair Employment and Mental Well-being.'

Evaluation

Metrics

Label	Accuracy
all	0.25

Uses

Direct Use for Inference

First install the SetFit library:

pip install setfit

Then you can load this model and run inference.

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("dnth/setfit-riasec-classifier-subset")
# Run inference
preds = model("The HR Learning & Development Specialist designs and delivers workforce upskilling programmes that align with Singapore’s SkillsFuture initiatives, focusing on employee growth, well-being, and career progression. He/She facilitates training workshops, coaches managers on supportive leadership, and counsels staff on learning pathways, ensuring inclusive and accessible development opportunities across diverse teams. He/She collaborates with department heads to identify skill gaps and mobilises internal and external resources to deliver targeted interventions, including digital learning platforms and micro-credentialing solutions. He/She maintains accurate records of training participation and compliance with MOM’s Tripartite Guidelines, applying structured processes to track outcomes and report impact. He is empathetic, a natural communicator, and adept at building trust with employees at all levels. He thrives in Singapore’s dynamic labour landscape, where continuous learning is prioritised and HR practices must reflect both human-centric values and regulatory precision.")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	114	126.625	146

Label	Training Sample Count
IRC	10
EIC	10
ESC	10
SEC	10

Training Hyperparameters

• batch_size: (8, 8)
• num_epochs: (5, 5)
• max_steps: -1
• sampling_strategy: oversampling
• body_learning_rate: (2e-05, 1e-05)
• head_learning_rate: 0.01
• loss: CosineSimilarityLoss
• distance_metric: cosine_distance
• margin: 0.25
• end_to_end: False
• use_amp: False
• warmup_proportion: 0.1
• l2_weight: 0.01
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0067	1	0.1555	-
0.3333	50	0.1366	-
0.6667	100	0.0117	-
1.0	150	0.001	0.0772
1.3333	200	0.0006	-
1.6667	250	0.0003	-
2.0	300	0.0003	0.0802
2.3333	350	0.0002	-
2.6667	400	0.0002	-
3.0	450	0.0001	0.0805
3.3333	500	0.0001	-
3.6667	550	0.0001	-
4.0	600	0.0001	0.0810
4.3333	650	0.0001	-
4.6667	700	0.0001	-
5.0	750	0.0001	0.0814

Framework Versions

• Python: 3.12.8
• SetFit: 1.1.3
• Sentence Transformers: 5.1.1
• Transformers: 4.56.2
• PyTorch: 2.8.0+cu128
• Datasets: 4.1.1
• Tokenizers: 0.22.1

Citation

BibTeX

@article{https://doi.org/10.48550/arxiv.2209.11055,
    doi = {10.48550/ARXIV.2209.11055},
    url = {https://arxiv.org/abs/2209.11055},
    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
    title = {Efficient Few-Shot Learning Without Prompts},
    publisher = {arXiv},
    year = {2022},
    copyright = {Creative Commons Attribution 4.0 International}
}