finetunesentence

SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2. It maps sentences & paragraphs to a 384-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: sentence-transformers/all-MiniLM-L6-v2 <!-- at revision fa97f6e7cb1a59073dff9e6b13e2715cf7475ac9 -->
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity

<!-- - **Training Dataset:** Unknown --> <!-- - **Language:** Unknown --> <!-- - **License:** Unknown -->

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': 256, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, '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})
  (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("shivamgoel97/finetunesentence")
# Run inference
sentences = [
    'What steps should the Fulfillment SPC Leadership take when an Internal Partner forwards an escalation request?',
    'Context: team (a selected group of offline individuals) completes all escalated requests. Escalations from a previous representative will not be reassigned to that representative rather assigned and handled through the Escalation team. The representative\'s supervisor receives real-time feedback to provide the representative immediate coaching and corrective action. Do not assign escalated requests to the Outbound team. The Escalations team owns all escalated requests. If a call to PACT is needed for credit approval, the Outbound team can assist but will not own the request. If more information is needed to complete the request, follow the AIR guidelines.Ways to Receive Escalated RequestsTable: <table border="1" cellpadding="5" cellspacing="0" width="100%"><tbody><tr><th style="text-align: left;" valign="top">If ...</th><th style="text-align: left;" valign="top">Then ...</th></tr><tr><td style="text-align: left;" valign="top">the reason for the escalation is due to a mistake on the original COE-processed application</td><td style="text-align: left;" valign="top"><p>do not reassign to the original coordinator.<br/><br/>Note:\xa0<span style="background-color: transparent;">The Escalation Representative assigned the request makes the necessary changes/corrections and completes the request.</span></p></td></tr><tr><td style="text-align: left;" valign="top">the reason for the escalation is a Missed SLA / Revenue Impacting Order<br/></td><td style="text-align: left;" valign="top"><ul><li>Do not reassign the escalation email request to the coordinator who has the related email if the related WFM is with a coordinator who does not belong to the Escalation team and is in\xa0<b>Wait</b>\xa0status.<br/></li><li>Own the escalated WFM and monitor the progress of the original request until completion.</li><li>Create a case to track completion of the request.</li><li>Be responsible for providing the updates to the Sales partner by sending an email with the status and timelines as possible.</li><li>Complete WFM as (Request Type) Correspondence.</li></ul></td></tr><tr><td style="text-align: left;" valign="top">the related WFM is with another Center/Team<br/></td><td style="text-align: left;" valign="top"><p>reassign any follow-up emails related to that request in order for that team to continue to own it.<br/></p><ul><li>Do not respond to provide an update.\xa0</li><li>Submit a reassignment request if needed.</li></ul></td></tr><tr><td style="text-align: left;" valign="top">Internal Partners forward the escalation request<br/></td><td style="text-align: left;" valign="top"><p>the Fulfillment SPC Leadership does the following:<br/></p><ol><li>Acknowledges the email.</li><li>Provides an update within 1 hour after acknowledgement.</li><li>Tracks completion until closure and provides updates as needed.<ul><li>If the assigned lead is out of the office or unavailable, another leader will cover on their behalf.</li></ul></li></ol></td></tr></tbody></table>',
    'Context: Program provides VZW B2B Data Sales Teams with a selection of key M2M certified devices for use in customer pilots pursuant to a sale. This program is intended to motivate and enable the VZW B2B Data Sales Teams to generate M2M sales. The program is available for Data Sales Teams residing in the 4 VZW B2B Sales Areas and the Enterprise and Government group. It contains a portfolio of key M2M devices that may provide a pilot at no cost to the customer when connected on area-supplied concession lines or test lines.Contact Nishant Shah [email protected] to request these solution-based devices:Table: <table border="1" cellpadding="5" width="100%"><tbody><tr><td style="text-align: left;" valign="top" width="30%"><b>Manufacturer</b></td><td style="text-align: left;" valign="top" width="70%"><b>Model</b></td></tr><tr><td style="text-align: left;" valign="top" width="30%">Cisco</td><td style="text-align: left;" valign="top" width="70%">Demo</td></tr></tbody></table>',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

<!-- ### Direct Usage (Transformers) <details><summary>Click to see the direct usage in Transformers</summary> </details> --> <!-- ### Downstream Usage (Sentence Transformers) You can finetune this model on your own dataset. <details><summary>Click to expand</summary> </details> --> <!-- ### Out-of-Scope Use *List how the model may foreseeably be misused and address what users ought not to do with the model.* --> <!-- ## Bias, Risks and Limitations *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* --> <!-- ### Recommendations *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* -->

Training Details

Training Dataset

Unnamed Dataset

• Size: 7,057 training samples
• Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	label
  type	string	string	int
  details	<ul><li>min: 11 tokens</li><li>mean: 24.8 tokens</li><li>max: 49 tokens</li></ul>	<ul><li>min: 91 tokens</li><li>mean: 230.28 tokens</li><li>max: 256 tokens</li></ul>	<ul><li>0: ~29.30%</li><li>1: ~70.70%</li></ul>

• Samples:

  sentence_0	sentence_1	label
  <code>What steps should be taken if the appropriate ticket is not visible?</code>	<code>Context: appropriate ticket may not be visible. Click View Ticket Details for more information. Type in Ticket Number and click Search. Scroll to the bottom of the screen to view the Comments Log, which contains all the information about the ticket's creation and resolution. Tickets can also be located by picking a date range (Start Date and End Date), selecting Me as the Assigned To value and clicking Search to pull up all tickets created by the Representative doing the searching.Email the escalations mailbox and request the ticket be closed. Note: Cancel, Fraud Cancel and Address Change tickets cannot be undone.Table: <table border="1" cellpadding="5" cellspacing="0" width="100%"><tbody><tr><th style="text-align: left;" valign="top">Ticket rejection reason</th><th style="text-align: left;" valign="top">Solution</th></tr><tr><td style="text-align: left;" valign="top">Missing or invalid information</td><td style="text-align: left;" valign="top"><p>After utilizing all available resource...</code>	<code>1</code>
  <code>What are the steps to assign an Account Manager in My Verizon when the Account Manager does not have an MDN on the account?</code>	<code>Context: and password The Account Manager: Is listed as an Authorized User in the Billing System and in My Verizon Can transact account business on My Verizon, My Verizon Express, in-store and by calling CS. Assign Account Manager In-Store or CS Access Only: Account Owner follows these steps: Note: This process can be used when the Account Manager does not have an MDN on the account. The Account Manager is listed as Authorized User in the billing system and has access to transact account business in-store or with CS. Adding an Account Manager is not restricted to a Digital Only Transaction.Table: <table border="1" cellpadding="5" cellspacing="0" width="100%"><tbody><tr><th style="text-align: center;" valign="top" width="5%">Step</th><th style="text-align: left;" valign="top">Action</th></tr><tr><td style="text-align: center;" valign="top" width="5%">1</td><td style="text-align: left;" valign="top">Sign in to My Verizon.<br/></td></tr><tr><td style="text-align: center;" valign="top" wid...</code>	<code>1</code>
  <code>What should an agent do if a case is received with an executive handle tagged?</code>	<code>Context: to send the form). W-9 addresses must always be the corporate address of record for the entity servicing the customer. W-9s are issued to claim exemption from withholding (i.e., 1099) and provide the name and address for information reporting. Payment Questions The EPO Support team is unable to answer any questions about payments and applications. For wireless customers, open a Payment Investigation Ticket through ACSS. For further assistance, search for “Accessing Payment Hub in ACSS” through InfoManager. For wireline customers, open a Payment Investigation Ticket through the Payment Hub. For further assistance, search for “Payment Hub Participants Guide” through VZKnowledge.Table: <table border="1" cellpadding="5" cellspacing="0" width="100%"><tbody><tr><th style="text-align: center;" valign="top" width="5%">Step</th><th style="text-align: left;" valign="top">Action</th></tr><tr><td style="text-align: center;" valign="top" width="5%">1</td><td style="text-align: left;" valig...</code>	<code>0</code>

• Loss: <code>CosineSimilarityLoss</code> with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• multi_dataset_batch_sampler: round_robin

All Hyperparameters

<details><summary>Click to expand</summary>

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: no
• prediction_loss_only: True
• per_device_train_batch_size: 8
• per_device_eval_batch_size: 8
• 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: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1
• num_train_epochs: 3
• max_steps: -1
• 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
• 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
• dispatch_batches: None
• split_batches: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

</details>

Training Logs

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.3.1
• Transformers: 4.47.1
• PyTorch: 2.5.1+cu121
• Accelerate: 1.2.1
• Datasets: 3.2.0
• Tokenizers: 0.21.0

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

<!-- ## Glossary *Clearly define terms in order to be accessible across audiences.* --> <!-- ## Model Card Authors *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.* --> <!-- ## Model Card Contact *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.* -->