Upload 12 files

2dcef2a verified 3 months ago

23.4 kB

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:3763
  - loss:MultipleNegativesRankingLoss
base_model: sentence-transformers/all-mpnet-base-v2
widget:
  - source_sentence: >-
      request increase server capacity dear customer support ask upgrade server
      capacity optimize database improve scalability performance saas project
      management platform current setup limit increase user volume lead long
      response time reduced productivity ensure smooth user experience remain
      competitive extension infrastructure necessary ask urgent review request
      timely solution please inform u detail need process start
    sentences:
      - >-
        customer service update please write request update integration enhance
        compatibility across multiple product within scalable saas project
        management platform aim improve user experience increase efficiency
      - >-
        system failure data synchronization problem detect incident impact
        several product result system crash data synchronization error root
        cause suspect server overload lead integration failure effort resolve
        include restarting service clear cache review log problem persist
        scalability challenge likely underlying issue
      - ' customer support inquire updating security protocol software integration within hospital system objective improve data protection compliance ensure confidentiality integrity patient information would like know solution available process implement update could please provide detail matter thank assistance look forward prompt response best regard thank support'
  - source_sentence: >-
      medical data security problem dear customer service would like contact
      medical data loss problem ubuntu server loss problem may occur due weak
      password policy outdated cisco io try solve problem change password update
      joomla plugins however need additional support secure security system
      could please give u instruction follow investigation recommendation avoid
      future loss problem would appreciate help provide matter thank understand
      support sincerely name
    sentences:
      - >-
        need technical assistance digital campaign halt due technical problem
        might relate old software malware already reboot system verify update
        perform antivirus scan yet issue remain unresolved
      - >-
        inquiry clickup feature financial firm hello customer support write
        inquire clickup feature optimize investment analytics financial firm
        representative financial firm interest learn clickup help u streamline
        investment analysis process enhance performance specifically would like
        know follow feature 1 customizable dashboard clickup provide
        customizable dashboard allow u track key performance indicator kpis
        metric relevant investment analytics 2 automate workflow clickup
        automate workflow task related investment analysis data collection
        processing report 3 integration tool clickup integrate tool platform use
        investment analysis data provider risk management system portfolio
        management software 4 collaboration communication clickup facilitate
        collaboration communication among team member stakeholder involved
        investment analysis portfolio manager analyst risk manager 5 data
        visualization clickup provide data visualization capability enable u
        easily interpret understand complex investment data analytics would
        appreciate information feature benefit financial firm additionally
        request demo trial clickup see firsthand feature apply specific need
        thank time assistance look forward hear back soon
      - >-
        multiple equipment failure dear customer support n ni encounter
        concurrent failure across various office gadget include soundbar ring
        light surface pro smart doorbell hdmi cable thinkpad usb drive desktop
        computer air purifier vr headset issue significantly hamper workflow
        suspect recent power surge network outage might cause despite restart
        device inspect cable thoroughly problem remain unresolved unaltered
  - source_sentence: >-
      digital tool operational digital strategy tool use marketing agency
      experience malfunction restart device update application resolve problem
    sentences:
      - >-
        immediate attention need zoom screen share issue dear customer support
        write report high priority technical issue zoom version 5 11 0 screen
        share feature not work video conference affect team productivity require
        urgent resolution please address matter early convenience thank name
        company
      - >-
        problem website integration not work website social medium integration
        cease function might due api connection problem restart server verified
        configuration setting issue remain unresolved
      - ' data analytics tool occasionally fail process investment data expect might due recent software update increase data volume restart system check basic configuration issue still persist assistance need resolve problem'
  - source_sentence: >-
      request additional server administration hello customer support hope
      message find well currently partnership continuous solution require
      additional support server management expand operation require improve
      supervision server system maintain efficiency security please let u know
      available option associate cost change current agreement look forward
      continue productive partnership thank attention request
    sentences:
      - >-
        issue access notion today employee report difficulty access notion
        microsoft dynamic 365 try resetting password clear cache instal late
        software update without success
      - >-
        warranty request dear customer service write inquire option available
        extended warranty dell xps 13 9310 ultrabook recently purchase model
        high performance specification want make sure remain protect could
        please tell detail plan cost runtimes thank support matter sincerely
        name tel num acc num
      - >-
        integration problem report dear support team experience integration
        issue impact system functionality specifically several tool fail
        synchronize data essential daily operation believe problem may stem api
        authentication difficulty face similar issue despite attempt restart
        service verify credential issue still persist kindly request address
        matter promptly provide u solution please inform u require additional
        information u thank assistance
  - source_sentence: >-
      difference invoice new tariff change marketing agency complain billing
      difference multiple subscription due new service change overlap fee change
      price level could behind try contact customer service check account
      statement hop problem solve please help u solve problem
    sentences:
      - >-
        payment problem identify recently encounter issue subscription payment
        decline problem might due insufficient fund card expiration verify card
        detail ensure sufficient balance issue still persist
      - >-
        problem digital strategy digital strategy tool provide marketing agency
        malfunction unexpectedly may software compatibility configuration
        problem attempt resolve include restart system update software verify
        connection kindly assist resolve issue promptly reduce disruption
      - ' deploy advance security measure secure medical information across interconnect hospital device system'
pipeline_tag: sentence-similarity
library_name: sentence-transformers

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 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-mpnet-base-v2
Maximum Sequence Length: 384 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': 384, 'do_lower_case': False, 'architecture': '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})
  (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 = [
    'difference invoice new tariff change marketing agency complain billing difference multiple subscription due new service change overlap fee change price level could behind try contact customer service check account statement hop problem solve please help u solve problem',
    'payment problem identify recently encounter issue subscription payment decline problem might due insufficient fund card expiration verify card detail ensure sufficient balance issue still persist',
    'problem digital strategy digital strategy tool provide marketing agency malfunction unexpectedly may software compatibility configuration problem attempt resolve include restart system update software verify connection kindly assist resolve issue promptly reduce disruption',
]
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, 0.7374, 0.5017],
#         [0.7374, 1.0000, 0.3963],
#         [0.5017, 0.3963, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

Size: 3,763 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 float
details
min: 5 tokens
mean: 52.91 tokens
max: 267 tokens

min: 8 tokens
mean: 53.77 tokens
max: 267 tokens

min: 1.0
mean: 1.0
max: 1.0

	sentence_0	sentence_1	label
type	string	string	float
details	min: 5 tokens mean: 52.91 tokens max: 267 tokens	min: 8 tokens mean: 53.77 tokens max: 267 tokens	min: 1.0 mean: 1.0 max: 1.0

Samples:

sentence_0	sentence_1	label
`intern network issue saas application experience temporary connectivity issue saas application network instability misconfigurations might cause please restart tp link switch netgear router`	`problem application crash peak usage application crash unexpectedly peak usage hour may due database overload resource constraint although attempt make optimize sql query reduce load step not successful would greatly appreciate assistance resolve issue prevent future crash ensure good user experience`	`1.0`
`request update discord drupal hello customer support contact request update discord drupal integration digital marketing effort heavily depend platform think improve integration greatly enhance track performance give fast paced environment digital marketing essential stay forefront optimize tool increase reach engagement could please examine provide solution fit requirement additional step need undertake detail need please let know thank time assistance look forward response`	`update require immediately please integration must update good compatibility`	`1.0`
`dear customer support contact address problem integration multiple application unexpectedly stop work believe issue might due recent api modification excessive server load despite effort reset service review logs test application independently problem remain unresolved kindly request examine situation offer prompt solution please inform information end necessary address issue thank attention cooperation sincerely name`	`integration problem jira clickup dear support team would like report integration issue jira clickup synchronization error suddenly appear night suspect might relate api change try restart server review log issue persists would appreciate could look matter offer solution soon possible please let know need additional information resolve issue thank time assistance look forward prompt response`	`1.0`

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim",
    "gather_across_devices": false
}

Training Hyperparameters

Non-Default Hyperparameters

num_train_epochs: 5
fp16: True
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: 5
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: True
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
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss
1.0616	500	1.8284
2.1231	1000	1.5296
3.1847	1500	1.3349
4.2463	2000	1.1681

Framework Versions

Python: 3.12.7
Sentence Transformers: 5.2.3
Transformers: 4.49.0
PyTorch: 2.6.0+cu124
Accelerate: 1.13.0
Datasets: 3.3.2
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",
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}