SentenceTransformer based on sentence-transformers/all-mpnet-base-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-mpnet-base-v2. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for retrieval.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: sentence-transformers/all-mpnet-base-v2
• Maximum Sequence Length: 384 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity
• Supported Modality: Text

Model Sources

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

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'transformer_task': 'feature-extraction', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'last_hidden_state'}}, 'module_output_name': 'token_embeddings', 'architecture': 'MPNetModel'})
  (1): Pooling({'embedding_dimension': 768, 'pooling_mode': 'mean', 'include_prompt': True})
  (2): Normalize({})
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
    'Name: Yash Banerjee || Email: yash.banerjee14@gmail.com | Phone: +91-8711810227 | LinkedIn: linkedin.com/in/yash-banerjee9 || Location: Kolkata, India | Open to: Remote / Hybrid || --- || PROFESSIONAL SUMMARY || Brand and content strategist with 8 years at Oracle. Deep expertise in GAAP and building engaged communities. Led campaigns reaching 36M+ users. || --- || SKILLS || CA, Auditing, GAAP, CFA, Oracle Financials, Equity Research, FP&A || --- || WORK EXPERIENCE || Brand Manager | Oracle | Kolkata | 2013 – Present || • Launched influencer programme with 23+ creators, generating 41M+ impressions and 65% engagement rate || • Reduced paid CAC by 68% by implementing Oracle Financials attribution model and reallocating budget || • Managed ₹48Cr monthly GAAP budget, achieving 67% lower CAC than industry benchmark || --- || Junior Brand Manager | Cognizant | Kolkata | 2012 – 2013 || • Led rebranding campaign that improved brand recall by 41% in post-campaign surveys || • Produced 5+ pieces of long-form content that each ranked on Page 1 within 4 months || --- || EDUCATION || PGDM in Marketing | NMIMS Jaipur | 2012 | CGPA: 7.7 || --- || CERTIFICATIONS || Meta Blueprint',
    'Job Title: DevOps Engineer || Company: Swiggy | Location: Chennai / Remote | Type: Full-Time || Experience: 2+ years | Salary: ₹9–22 LPA || Team Size: ~17 | Interview Rounds: 4 || --- || ABOUT THE ROLE || We are looking for a passionate DevOps Engineer to join our DevOps team at Swiggy. You will work on high-impact problems affecting millions of users and collaborate with some of the best minds in the industry. || --- || RESPONSIBILITIES || • Manage and scale kubernetes clusters across multiple cloud regions || • Design and maintain ci/cd pipelines enabling fast, reliable software delivery || • Participate in on-call rotation, respond to incidents, and drive rca to prevent recurrence || • Collaborate with development teams to embed devops practices and platform tooling || • Define and enforce slos, slis, and error budgets for critical services || --- || REQUIRED SKILLS: Bash, Kubernetes, Loki, ArgoCD, GKE, AWS, GitLab CI || NICE TO HAVE: New Relic, Pulumi, Istio || --- || PERKS: ESOPs, Annual performance bonus, ₹50K learning budget, Quarterly offsites || Domain: DevOps',
    'Job Title: Computer Vision Engineer || Company: TCS | Location: Kochi / Remote | Type: Full-Time || Experience: 5+ years | Salary: ₹27–46 LPA || Team Size: ~30 | Interview Rounds: 6 || --- || ABOUT THE ROLE || We are looking for a passionate Computer Vision Engineer to join our Data Science team at TCS. You will work on high-impact problems affecting millions of users and collaborate with some of the best minds in the industry. || --- || RESPONSIBILITIES || • Own model performance in production — monitor drift, retrain models, and improve accuracy || • Build end-to-end ml pipelines from data ingestion to model deployment and monitoring || • Research and prototype state-of-the-art ml techniques from recent literature || • Collaborate with engineering and product to identify ml opportunities and define success metrics || • Work closely with data engineering to ensure data quality and feature availability || --- || REQUIRED SKILLS: Pandas, Data Modeling, Feature Engineering, Airflow, Kubeflow, TensorFlow, Statistics, Time Series, Hypothesis Testing, Hadoop || NICE TO HAVE: R, Power BI, PostgreSQL || --- || PERKS: Mental health support, ESOPs, Flexible work hours, Maternity/paternity leave, Quarterly offsites, Gym membership || Domain: Data Science',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.1557, 0.2299],
#         [0.1557, 1.0000, 0.2397],
#         [0.2299, 0.2397, 1.0000]])

Evaluation

Metrics

Semantic Similarity

• Dataset: ats-val
• Evaluated with EmbeddingSimilarityEvaluator

Training Details

Training Dataset

Unnamed Dataset

• Size: 350 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 100 samples:

  	sentence_0	sentence_1	label
  type	string	string	float
  modality	text	text
  details	min: 308 tokensmean: 345.37 tokensmax: 384 tokens	min: 261 tokensmean: 286.24 tokensmax: 314 tokens	min: 0.05mean: 0.53max: 0.98

• Samples:

  sentence_0	sentence_1	label
  Name: Shruti Rao || Email: shruti.rao44@gmail.com | Phone: +91-7125381387 | LinkedIn: linkedin.com/in/shruti-rao7 || Location: Noida, India | Open to: Remote / Hybrid || --- || PROFESSIONAL SUMMARY || 3+ years designing intuitive, accessible products at Freshworks used by 15M+ users. Expert in MEDDIC and SPIN Selling. Known for translating complex problems into elegant solutions. || --- || SKILLS || Figma, Interaction Design, Motion Design, Sketch, User Research, Negotiation, Solution Selling, SPIN Selling, MEDDIC || --- || WORK EXPERIENCE || UX Researcher | Freshworks | Noida | 2013 – Present || • Improved accessibility score from 17 to 38/100 ensuring WCAG 2.2 compliance for 39M+ users || • Redesigned MEDDIC checkout flow reducing drop-off by 74% and increasing conversion by ₹17Cr monthly || • Led end-to-end redesign of Negotiation app achieving 39% improvement in task completion rate in usability testing || --- || Junior UX Researcher | Byju's | Noida | 2010 – 2012 || • Conducted 45...	Job Title: Product Designer || Company: Capgemini | Location: Gurgaon / Remote | Type: Full-Time || Experience: 4+ years | Salary: ₹15–44 LPA || Team Size: ~28 | Interview Rounds: 6 || --- || ABOUT THE ROLE || We are looking for a passionate Product Designer to join our UI/UX Design team at Capgemini. You will work on high-impact problems affecting millions of users and collaborate with some of the best minds in the industry. || --- || RESPONSIBILITIES || • Own the end-to-end ux process from research and ideation to prototyping and final handoff || • Conduct user research including interviews, surveys, and usability testing || • Maintain and evolve the product's design system and component library || • Present design decisions with clear rationale to stakeholders and leadership || • Use data and user feedback to continuously iterate and improve the product experience || --- || REQUIRED SKILLS: Design Thinking, Information Architecture, Adobe XD, Design Systems, Sketch, Figma, Wireframi...	0.51
  Name: Siddharth Chopra || Email: siddharth.chopra50@gmail.com | Phone: +91-8707729724 | LinkedIn: linkedin.com/in/siddharth-chopra7 || Location: Delhi, India | Open to: Remote / Hybrid || --- || PROFESSIONAL SUMMARY || Enterprise sales professional with 11 years at Airbnb selling REST APIs solutions to C-suite buyers. Built and managed pipeline of ₹36Cr+ across 8 industry verticals. || --- || SKILLS || FastAPI, REST APIs, MongoDB, Microservices || --- || WORK EXPERIENCE || Enterprise AE | Airbnb | Delhi | 2012 – Present || • Closed ₹32Cr ARR in FY23, exceeding annual quota by 52% and ranking #5 in team of 26 || • Expanded 31 existing accounts through upselling, generating ₹5Cr incremental ARR || • Negotiated and closed 28+ multi-year enterprise contracts totalling ₹3Cr with 21% above floor pricing || --- || Junior Enterprise AE | SAP | Delhi | 2011 – 2013 || • Built SDR team of 23 reps from scratch, establishing outreach playbooks and achieving 69% meeting conversion || • Won ₹9Cr comp...	Job Title: Senior PM || Company: Paytm | Location: Delhi / Remote | Type: Full-Time || Experience: 4+ years | Salary: ₹22–54 LPA || Team Size: ~35 | Interview Rounds: 6 || --- || ABOUT THE ROLE || We are looking for a passionate Senior PM to join our Product Management team at Paytm. You will work on high-impact problems affecting millions of users and collaborate with some of the best minds in the industry. || --- || RESPONSIBILITIES || • Represent the voice of the customer in all product decisions || • Write detailed prds, user stories, and acceptance criteria || • Drive customer discovery through interviews, surveys, and usability studies || • Define and own the product roadmap aligned with company okrs and user needs || • Prioritise backlog using data-driven frameworks and communicate trade-offs clearly || --- || REQUIRED SKILLS: JIRA, SQL, A/B Testing, Scrum, Customer Discovery, Stakeholder Management, Competitive Analysis, User Research, Agile, Mixpanel || NICE TO HAVE: Growth Ha...	0.37
  Name: Kabir Pillai || Email: kabir.pillai80@gmail.com | Phone: +91-8986468570 | LinkedIn: linkedin.com/in/kabir-pillai6 || Location: Kochi, India | Open to: Remote / Hybrid || --- || PROFESSIONAL SUMMARY || UX researcher and designer with 11 years. Conducted 41+ user studies that directly shaped roadmap decisions. Proficient in PPC, Information Architecture and Storyboarding. || --- || SKILLS || Information Architecture, Sketch, A/B Testing, Zeplin, Storyboarding, PPC, Market Research, Affiliate Marketing || --- || WORK EXPERIENCE || UI Designer | HDFC Bank | Kochi | 2019 – Present || • Redesigned A/B Testing checkout flow reducing drop-off by 26% and increasing conversion by ₹39Cr monthly || • Led end-to-end redesign of A/B Testing app achieving 74% improvement in task completion rate in usability testing || • Ran 21+ A/B design experiments resulting in 76% lift in key engagement metric || --- || Junior UI Designer | Microsoft | Kochi | 2017 – 2018 || • Improved accessibility score fr...	Job Title: Design Systems Engineer || Company: Salesforce | Location: Pune / Remote | Type: Full-Time || Experience: 7+ years | Salary: ₹9–42 LPA || Team Size: ~29 | Interview Rounds: 6 || --- || ABOUT THE ROLE || We are looking for a passionate Design Systems Engineer to join our UI/UX Design team at Salesforce. You will work on high-impact problems affecting millions of users and collaborate with some of the best minds in the industry. || --- || RESPONSIBILITIES || • Advocate for accessibility and inclusive design across all platforms || • Present design decisions with clear rationale to stakeholders and leadership || • Use data and user feedback to continuously iterate and improve the product experience || • Conduct user research including interviews, surveys, and usability testing || • Collaborate closely with product managers and engineers throughout the design-build cycle || --- || REQUIRED SKILLS: Figma, Heuristic Evaluation, Storyboarding, User Research, Wireframing, Lottie, HT...	0.48

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss",
      "cos_score_transformation": "torch.nn.modules.linear.Identity"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 16
• num_train_epochs: 10
• per_device_eval_batch_size: 16
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

• per_device_train_batch_size: 16
• num_train_epochs: 10
• max_steps: -1
• learning_rate: 5e-05
• lr_scheduler_type: linear
• lr_scheduler_kwargs: None
• warmup_steps: 0
• optim: adamw_torch_fused
• optim_args: None
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• optim_target_modules: None
• gradient_accumulation_steps: 1
• average_tokens_across_devices: True
• max_grad_norm: 1
• label_smoothing_factor: 0.0
• bf16: False
• fp16: False
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• use_liger_kernel: False
• liger_kernel_config: None
• use_cache: False
• neftune_noise_alpha: None
• torch_empty_cache_steps: None
• auto_find_batch_size: False
• log_on_each_node: True
• logging_nan_inf_filter: True
• include_num_input_tokens_seen: no
• log_level: passive
• log_level_replica: warning
• disable_tqdm: False
• project: huggingface
• trackio_space_id: None
• trackio_bucket_id: None
• trackio_static_space_id: None
• per_device_eval_batch_size: 16
• prediction_loss_only: True
• eval_on_start: False
• eval_do_concat_batches: True
• eval_use_gather_object: False
• eval_accumulation_steps: None
• include_for_metrics: []
• batch_eval_metrics: False
• save_only_model: False
• save_on_each_node: False
• enable_jit_checkpoint: False
• push_to_hub: False
• hub_private_repo: None
• hub_model_id: None
• hub_strategy: every_save
• hub_always_push: False
• hub_revision: None
• load_best_model_at_end: False
• ignore_data_skip: False
• restore_callback_states_from_checkpoint: False
• full_determinism: False
• seed: 42
• data_seed: None
• use_cpu: False
• 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
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• dataloader_prefetch_factor: None
• remove_unused_columns: True
• label_names: None
• train_sampling_strategy: random
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• ddp_static_graph: None
• ddp_backend: None
• ddp_timeout: 1800
• fsdp: None
• fsdp_config: None
• deepspeed: None
• debug: []
• skip_memory_metrics: True
• do_predict: False
• resume_from_checkpoint: None
• warmup_ratio: None
• local_rank: -1
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Training Time

• Training: 6.2 minutes

Framework Versions

• Python: 3.12.13
• Sentence Transformers: 5.5.1
• Transformers: 5.10.1
• PyTorch: 2.11.0+cu128
• Accelerate: 1.13.0
• Datasets: 4.0.0
• Tokenizers: 0.22.2

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",
}