shivamgoel97/mediumfinetunesentence

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
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 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': 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/mediumfinetunesentence")
# Run inference
sentences = [
    'What should a 5G/LTE Home Internet Messaging Support Agent do if a customer with Fixed Wireless Access (FWA) needs assistance and the dedicated FWA Support team is occupied?',
    '5G/LTE Home Internet Messaging Support Agents in Messaging provide assistance to 5G and LTE Home Internet customers. Customers with Fixed Wireless Access (FWA) and who indicate they need assistance with their FWA device route to a dedicated FWA Support team. However, volume may overflow if those agents are occupied. Assist these customers BAU by providing billing and technical support. Refer to these resources if skilled for Tier 2 and receive a 5G or LTE Home Internet customer: Verizon 5G Internet Gateway - Install / Set Up 5G Home - Window Mount Verizon 5G Internet Gateway - Install / Set Up 5G Home - Wall Mount Verizon 5G Internet Gateway - Install / Set Up 5G Home - Speed Test 5G Home Internet Extender Setup Router Setup Verizon 5G Internet Gateway - Install/Setup 5G Home - Signal Test 5G Resource Center YouTube Video - How to: Self setup your Verizon 5G Internet Gateway - Internal Only)',
    'Context: form and return it to [email protected] after purchasing an eFax license Porting may take up to 5 weeks to be completed.Government Process to Enable Ordering eFax Billing eFax licenses are billed at the account level and reflected on the wireless bill. The number of licenses does not have to match the number of wireless lines. Customers may distribute licenses across multiple carriers. After order submission and POC approval, customer billing begins. Cancellation Policy All sales are final for the subscription term/contract initially purchased month-to-month contract (no commitment). Customers who purchase month-to-month licenses can cancel at the end of the month.Table: <table border="1" cellpadding="5" cellspacing="0" width="100%"><tbody><tr><th style="text-align: center;" valign="top" width="5%">Stage</th><th style="text-align: left;" valign="top" width="15%">Who</th><th style="text-align: left;" valign="top">Does What</th></tr><tr><td style="text-align: center;" valign="top" width="5%">1</td><td style="text-align: left;" valign="top" width="15%"><p>SLED Sales\xa0<span style="background-color: transparent;">Representative</span></p></td><td style="text-align: left;" valign="top"><p>Confirms the customer\'s contract is amended to include eFax.<br/></p><ul><li>If the Contract includes eFax:<br/><ol><li>Access the\xa0<a href="https://vzweb.verizon.com/state-local-government" target="_blank">State and Local Government Contracts</a> web page.<br/></li><li>Select the black oval to request a profile update for BuSS product eFax.</li></ol></li><li>If the agreement does not include eFax:<br/><ol><li>Access the\xa0<a href="https://vzweb.verizon.com/state-local-government" target="_blank">State and Local Government Contracts</a>\xa0web page.</li><li>Select the red oval to request an amendment to add eFax to the agreement.</li><li>Select the green oval to upload the amendments for implementation (once the amendment is executed).<br/></li></ol></li></ul></td></tr><tr><td style="text-align: center;" valign="top" width="5%">2</td><td style="text-align: left;" valign="top" width="15%"><p>Contract\xa0<span style="background-color: transparent;">Management Team</span></p></td><td style="text-align: left;" valign="top"><p>Reviews BuSS request, and if approved forwards the request to the BFO to enable\xa0<span style="background-color: transparent;">the customer profile for eFAX (based on contract availability).</span></p></td></tr><tr><td style="text-align: center;" valign="top" width="5%">3</td><td style="text-align: left;" valign="top" width="15%">BFO</td><td style="text-align: left;" valign="top">Completes request, and replies all upon completion.<br/></td></tr><tr><td style="text-align: center;" valign="top" width="5%">4</td><td style="text-align: left;" valign="top" width="15%"><p>SLED Sales\xa0<span style="background-color: transparent;">Representative</span></p></td><td style="text-align: left;" valign="top">Works with the customer to complete the license order.<br/></td></tr><tr><td style="text-align: center;" valign="top" width="5%">5</td><td style="text-align: left;" valign="top" width="15%">eFax<br/></td><td style="text-align: left;" valign="top"><p>Activates the licenses and sends a welcome / get started email to the customer’s email sales\xa0<span style="background-color: transparent;">entered in POS.</span></p></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]

Training Details

Training Dataset

Unnamed Dataset

• Size: 10,500 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	label
  type	string	string	int
  details	min: 11 tokens mean: 24.76 tokens max: 61 tokens	min: 55 tokens mean: 211.78 tokens max: 256 tokens	0: ~24.00% 1: ~76.00%

• Samples:

  sentence_0	sentence_1	label
  What is the internal transfer number for Spanish HSI Tech Support for Consumer HSI?	Context: CRTC. See Ordering a Return Label (ACSS) for more information.Refer to the return and exchange periods below (by purchase location): Returns Exchanges Note: For any exceptions to the return or exchange period, use the Blind Return flow process and remark the orders with: Original order number Location Reason for the exceptionDevice returns (not exchanges) reset the customer's contract and upgrade dates. Verizon requires a contract correction to: Verify all device returns. Reset the customer's contract and upgrade dates. When the date does not systematically correct, validate the: Device was returned. Customer requested a disconnect and returned to a previous device.Table: Purchase Location Return Period Verizon stores and Indirect locations 0 What is the time frame within which customers must call and request a return for a Stream TV device to be eligible for a refund from Verizon? Review the following guidelines before processing a device payment agreement: Canceling a device payment agreement erases all of the monthly installments the customer has paid towards a device. Example: If the customer has paid 3 payments, the bill is refunded for all 3 payments made. If the customer has made no payments, canceling a device payment agreement makes the agreement act like it never existed and the customer does not see it on the bill. In order to cancel a device payment agreement, verify the following: The status of the device payment agreement prior to requesting an agreement cancellation. Refer to the Loan Status Codes Grid for additional information. Device payment agreements in Paid status cannot be canceled. The device has been returned to a store or warehouse depending upon where the order was processed. If a store processed the order, verbally confirm the store has the device at the store location and attempted a return. If a direct fulfillment order, the customer ... 0 Which SKU must be added to the order when placing an order for BuSS solutions? Context: multiple licenses, the page limits are aggregated. Faxing to international phone numbers is not available at this time.Ordering All solutions in the BuSS are billed on the Account Level. Customers may purchase BuSS solutions through these Verizon POS systems: B360 ACSS OMNI Retail OMNI Telesales Customers can mix and match licenses. When placing the order, the EFAX_OVERAGE SKU must be added to the order. Only 1 EFAX_OVERAGE SKU is required per account.Customer Payment Methods Retail, Retail SMB and Telesales B2B can only sell through Bill To Account (BTA). B2B and ACSS may sell through BTA and Purchasing Order (PO).Standard PricingTable: Stage Who Does What 1 1 Loss: CosineSimilarityLoss with these parameters: { "loss_fct": "torch.nn.modules.loss.MSELoss" } Training Hyperparameters Non-Default Hyperparameters multi_dataset_batch_sampler: round_robin All Hyperparameters Click to expand 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 Training Logs Epoch Step Training Loss 0.3808 500 0.1354 0.7616 1000 0.1184 1.1424 1500 0.1072 1.5232 2000 0.0968 1.9040 2500 0.095 2.2848 3000 0.0871 2.6657 3500 0.0832 Framework Versions Python: 3.11.11 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", }