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, 'architecture': '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("ar9av/all-minilm-l6-v2-civic")
# Run inference
sentences = [
    "Take it. Oh, okay. hey. Oh. I wanted Heat. Heat. Good evening, Vakville. This is the the open session portion of our regular meeting that was supposed to start at 6. We went into close session uh shortly after 5:00 and we will start this meeting at 6:37. We have a roll call. &gt;&gt; Council member Stockton &gt;&gt; Council member Richie &gt;&gt; Council member Silva &gt;&gt; Council member Freemo &gt;&gt; Council member Wy &gt;&gt; Vice Mayor Chapman &gt;&gt; Mayor Carly &gt;&gt; here. If you're able, please rise for a moment of silence. &gt;&gt; Please join me in the pledge of to the flag of the United States of America and to the republic for it stands. One nation under God, indivisible, with liberty and justice for all. &gt;&gt; So, there are four items on close session. We completed item A. Um, there's nothing to report out and at the conclusion of this meeting, we have three other items that we will go into close session. I don't anticipate anything to report out tonight. Are the Vacaville City Council Special Meeting - November 3, 2025",
    'WC Planning Department Aaron Holmes, Director Rogers C. Anderson 1320 W Main Street, Suite 400 Williamson County Mayor Franklin, TN 37064 WILLIAMSON COUNTY REGIONAL PLANNING COMMISSION SCHEDULE OF MEETINGS 2026 All meetings are scheduled to commence at 5:30 p.m. in the Main Auditorium of the Williamson County Administrative Complex. Schedule is subject to change. Pre-Application Deadline Submittal Deadline at 4:30 PM WCRPC Meeting Date Friday, November 14, 2025 Wednesday, November 26, 2025 Thursday, January 8, 2026 Friday, December 19, 2025 Wednesday, December 31, 2025 Thursday, February 12, 2026 Friday, January 16, 2026 Wednesday, January 28, 2026 Thursday, March 19, 2026 Friday, February 13, 2026 Wednesday, February 25, 2026 Thursday, April 9, 2026 Friday, March 20, 2026 Wednesday, April 1, 2026 Thursday, May 14, 2026 Thursday, April 16, 2026 Wednesday, April 29, 2026 T Planning Commission Schedule of meetings 2026',
    "Can I have a motion to return to an executive session? Motion. Second. Properly moved and second. Question or remark. Question being called, all in favor? Aye. Opposed? Motion carries. Unfinished business. I need a motion for borough council to approve the revised settlement agreement for the case of Christina Fulton versus Claude Co Borough. If the motion is not approved, council may further make a motion to modify the agreement and return it to the solicitor. So would somebody like to make a motion to approve the revised settlement agreement for the case of Christina Fulton versus Clay College? I'll make that motion. Second. Properly moved and seconded. Question or remark. Question. Question being called, all in favor? Aye. Opposed. Motion carries. Do you have anything else? Nope. Special meeting. That's all we have. Okay. Motion to adjourn. Second. Properly moved and seconded. Question or mark. Question being called. All in favor? Aye. Opposed. Thank you. Motion is adjourned at 4: 4 November 11, 2025 Special Meeting part 2",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.9987, 0.9991],
#         [0.9987, 1.0000, 0.9983],
#         [0.9991, 0.9983, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

Size: 60,000 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1
type string string
details
min: 66 tokens
mean: 236.91 tokens
max: 256 tokens

min: 28 tokens
mean: 237.37 tokens
max: 256 tokens

	sentence_0	sentence_1
type	string	string
details	min: 66 tokens mean: 236.91 tokens max: 256 tokens	min: 28 tokens mean: 237.37 tokens max: 256 tokens

Samples:

sentence_0	sentence_1
The December 1st, 2025 legislative session of the Anaonda County Council is closed to order. Called to order. Um, please ensure all phones and electronic devices are set to silent. Um, please rise for the invocation and the pledge of allegiance. Please pray according to your custom as I pray according to mine. Dear God, as we go through this holiday season, open our hearts to those most in need and assistance. May we provide food for those who are hungry, shelter for those who are cold, and serve others in this time of giving. Guide our actions tonight that we may best use our resources to serve the most vulnerable in our county with compassion and that we may bring joy to others now and throughout the year. Amen. >> To the flag of the United States of America, to the republic for which it stands, one nation under God, indivisible, with liberty and justice for all. >> Madam Secretary, please read the ethics >> The ethics commission has asked that I advise you that und...	Morning everyone. Welcome to the including our AI soon to be overlords. Um, happy Thanksgiving week. It is great to have you all here. Thank you everyone for great attendance. I want to give a shout out to the teachers and principles and DOE staff who came in person tonight and will be presenting in a couple minutes. I'm really thankful for you guys coming during a holiday week. So, thank you for coming and I'm looking forward to your presentation soon. Um, I know everyone has places to go and things to do this week. So, let's go ahead and start our meeting. Um, we're going to do roll call. Go ahead, Curtis. absent. Kachchuck. >> here online. >> here. Young >> here online >> and Lee Char of Quorum with 11 members >> All right, let's start by hearing from the Honolulu Fire Department if they are All right, let's go down to HPD district >> Hello, I'm Sergeant Sai. Hey, >> Sergeant Sai. >> Hello. Hi, everyone. I'm just going to read uh the s...
>> going to open this meeting at 6:01 >> And we're all here. Reading on the >> Make a motion. We accept the minutes. >> No, I make a >> Oh, you got to change. >> I want to amend it. >> Okay, go ahead. the motion by the department of the following. The motion I thought I said but I didn't. I'd like to amend it to say that uh accept the town manager's recommendation to accept JB Liby and Sun's bid of $4,800 for one year. >> I have made that change and email copy out everything. >> I was going to say it's already in my email. So with that amendment I would you go make a motion? >> No. You make the amendment motion. >> I make a motion we accept the minutes as >> Second that. >> Any discussion? >> in favor? Announcements. town office will be closed November 27th and 28th for the Thanksgiving holiday and 2026 dog registrations are available. >> Uh, one item missing from that is budget committee meetings...	10 o'clock, actually 10.05. We'll call this meeting of the Jefferson County Water Authority to order. Roll call, please. Logan Jaskowitz? Here. Mark Greg Camp? Here. Michael Christopher? Here. Kevin Dennis? Here. We do have a quorum pledge of allegiance. I pledge allegiance to the flag of the United States of America and to the republic for which it stands, one nation under God, indivisible, with liberty and justice for all. All right. Thank you all. We have the minutes from September 18th, 2025. I would entertain a motion to approve the minutes unless there's questions. Motion. Motion and a second. All those in favor, say aye. Aye. Any opposed? All right. The bills, we did send out an amended agenda that did include miscellaneous vendors. It was an additional 68, 430.57 cents. So the total bills is $727, 764.35. Who would like to pay the bills? I'll make a motion. We've got a motion and a second from Kevin. All those in favor? I'm sorry, roll call, please. Michael Christopher? Aye. Gr...
Period. End of story. It has been a delightful two years. None. Not. It is 6: 30. Thank you. All right, good evening, everybody. I would like to call the November 10th, 2025 Board of School Committee meeting to order. I would welcome Principal Bolak and the students from Green Acres Elementary School who are present to lead us in the Pledge of Allegiance. Come on forward. I would ask Committee Member Bergeron and Committee Member Spillers to please introduce our students. I just wanted to thank the students for coming out tonight and to introduce them to all of us from Green Acres. Addison Dougherty, James Dougherty, Darus Mula. Oh, I'm going to butcher it. I apologize. Mula Bogovic, Syra Mula Bogovic, Kensley Hines, and Grayson Hines, and Teddy Lombardo and Penny Lombardo. And to the republic for which it stands, one nation under God, indivisible, with liberty and justice for all. Please remain standing for a moment of silence. Would the clerk please call the roll? Committee Member Tu...	Alam alaikum everyone. Welcome to another Human Rights Commission meeting. Today is November 6, 2025 and it's currently 5:02 p.m. Um we are located at the city council chambers and I am calling a meeting the meeting to order. Um, I'll start with roll call. Um, Commissioner Kuru Pitch >> here. Commissioner Albert Tedy >> Commissioner Shayla Johnson >> Um, I would like to put on record that um, Commissioners Dr. Garcia, Commissioners Gene Baptist, and Commissioner Victor Pelletary are not currently in our meeting today. And of course, uh, Vice Chair Shima is also here. Um, I would like to provide the opportunity to, um, for our new commissioners to introduce themselves. Um, I would like, I'll have, um, Chief Maria Lvita start off first. Please introduce yourself. >> Good evening. Chief Maria Lvita. I'm with the River Police Department. >> And Chief James Cullen. >> Good evening. Uh, Chief James Cullen. recently appointed as the uh chief of the rea fire...

Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 32
per_device_eval_batch_size: 32
num_train_epochs: 1
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: 32
per_device_eval_batch_size: 32
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: 1
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
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}
parallelism_config: None
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
project: huggingface
trackio_space_id: trackio
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
hub_revision: None
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
include_tokens_per_second: False
include_num_input_tokens_seen: no
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
eval_on_start: False
use_liger_kernel: False
liger_kernel_config: None
eval_use_gather_object: False
average_tokens_across_devices: True
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss
0.2667	500	3.635
0.5333	1000	3.4659
0.8	1500	3.4658

Framework Versions

Python: 3.11.10
Sentence Transformers: 5.2.0
Transformers: 4.57.3
PyTorch: 2.4.1+cu124
Accelerate: 1.12.0
Datasets: 4.4.2
Tokenizers: 0.22.1

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}
}