SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L12-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-L12-v2
Maximum Sequence Length: 128 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': 128, '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-l12-v2-civic")
# Run inference
sentences = [
    "County signed off on it. &gt;&gt; All set. &gt;&gt; All set. &gt;&gt; All right. I'd like to call the regular meeting of the village planning commission of Elmont for November 6, 2025 to order, please. May I have a roll &gt;&gt; He must have heard his calling. Um, the next item on the agenda is the approval of the agenda. Unless there are any changes or alterations or corrections, I accept a motion to approve the agenda as is presented. &gt;&gt; I have Steve as a move and Tim as a support. All those in favor? &gt;&gt; All those opposed? Motion carried. The next item is the approval of minutes for the regular meeting of October 2nd, um they've been included in your packet. Rather lengthy, by the way, we must have had a great conversation. Um anyway, unless there are any changes or alterations, I would accept a motion to approve the minutes as presented. &gt;&gt; I wasn't here at the last meeting, but it looked okay to me. Anybody &gt;&gt; Glenn moved? Do I have a second? &gt;&gt; Gary s Regular Village Planning Commission Meeting 11/06/2025",
    "West Carolin City Council meeting will come to order. Everyone, please stand and join City Council for Pledge Allegiance. I pledge it to the Prince. 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. Mrs. Moore, can we have the roll call, please? Mr. Tilton? Here. Mrs. Tomlin? Here. Mrs. Williams? Here. Mr. Boyle? Here. Mr. Dobson? Here. Mayor Barnhart? He's excused. Deputy Mayor Fryman. Here. Council, I'd like to make a motion to amend the agenda this evening by adding an executive session pursuant to Ohio Revised Code 121.22G1 to consider the employment and compensation of a public employee or official. May I give a second? Second. Deputy Mayor Freyman moved, seconded by Mr. Dobson and Mr. Boyle to amend the agenda and add an executive session pursuant to Ohio Revised Code 121.22 G1 to consider the employment and compensation of a public employee or official. Mr. T 11/25/25 West Carrollton City Council Meeting",
    "Good evening, everyone. We will call this City Council meeting to order for Monday, November 3rd. Please rise for the 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. Thank you. The clerk will take the role. Councilmember Fleming, here. Ford here. Larson? Here. Massolia? Here. Newland? Here. Robertson? Here. Mayor Sanders. Here. All right. There are no awards or presentations here this evening, so we'll start with communications. An opportunity for members or staff to talk about any announcements that they might have. Are there any communications this evening? Councilmember Robertson. Thank you, Mayor. Really, it's just a quick thank you to staff. We just wrapped up Citizens Academy. Congratulations to Councilmember Larson for graduating. To Councilmember Larson for graduating from Citizens Academy was the largest class they've ever had, whi City Council Meeting - Media",
]
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, 0.9994, 0.9995],
#         [0.9994, 1.0000, 0.9998],
#         [0.9995, 0.9998, 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: 22 tokens
mean: 127.47 tokens
max: 128 tokens

min: 16 tokens
mean: 127.26 tokens
max: 128 tokens

	sentence_0	sentence_1
type	string	string
details	min: 22 tokens mean: 127.47 tokens max: 128 tokens	min: 16 tokens mean: 127.26 tokens max: 128 tokens

Samples:

sentence_0	sentence_1
We can admit you Zane your uh your business travel or something. >> All right. I'll call the meeting to order. This is the Dwage County Zoning Board of Appeals November 20, 2025 uh meeting. We have uh I'll indicate that all of the members are present with the exception of Mr. Raman who u wishes to be admitted remotely. Um we uh need to take a motion to admit him remotely. Reason for uh his uh well has a business conflict. Uh so >> we have a motion >> to admit. >> Motion by Mr. Moran. >> Second by uh Mr. Schultz. Uh is there any discussion on the motion? >> All in favor of the motion signify by saying I. Any opposed Mr. Raman will be admitted. We have no minutes to approve. We have uh one continued case. This is a continuation of the case first heard on October 30, 2025, which is a variation to allow video gaming cafe restaurant within a th00and ft of an existing video gaming cafe restaurant video gaming terminal establishment. And Mr. uh Armstrong is...	All right, we'll call this meeting to water. First thing >> is right here. >> You got it? >> I am. >> Okay, that's good. >> What makes you want to do it? >> Couple couple different things. uh grew up on the river. I mean, it's been an integral part of my life for ever. Um I care about it. I hate to see it turn into something it shouldn't be. Um I mean, I love it. I want to take care of it. I want to help take care of it. I want to do everything I can for it. seem like the Right. Well, it's a good chance to contact some >> Appreciate you. >> Get out there, do some of the water, Um well that you had already approved but there's one for All right. Tacos circular circular tacos the boil room. My interpretation is that this is needed whether we no matter what because it circulates the water through the heating system. >> So the furnace just is something else. But right now we're operating on the This is a backup. >> Yeah, I guess w...
The rain coming down, so we're getting some storm out. Yeah, no, you don't need 8: 30 to begin our meeting of the Public Works and Community Development Committee. Um, join me in the Pledge of Allegiance. I pledge allegiance to the flag of the United States of America, to the Republic for which it stands, one nation under God, indivisible, with the justice for all. Okay, um, on the agenda, we have unfinished business starting with Blank Pole Park proposal. Um, not sure we have any pre-word. Um, Harry had said he would be here, but uh, okay, well, we can we can just wait a little bit, so we'll come back to it. Okay, okay, that's it. Um, well, let's go on to the agreement with uh Hittertas County Youth Baseball and Softball Association. Got an update, okay. Um, just uh, just yesterday, uh, oh, there we go. I have an update here. So, um, if you recall, we did our work here, we've got a good document. We sent it back to our attorney because the attorneys were talking. The UKC is attorney w...	SPECIAL AND REGULAR MEETINGS OF THE RAVALLI COUNTY OPEN LANDS BOARD Thursday, November 20, 2025 7:00 P.M. Commissioners Meeting Room County Administration Building, 215 S. 4th St, Hamilton, MT 59840. Chair Paul Moore Vice Chair Kent Myers Member Alan Maki Member Sharon Schroeder Member Penney Howe-Canton Member Bob Cron Member Katherine Stone Member Albert Babel Member Tony Hudson Member Katrina Mendrey Member Tonia Bloom (EO) Member Kimberly Richardson (EO) Member Howard Lyons Member Cherie Fisher Member Gil Gale COMMENTS FROM THE PUBLIC Members of the audience will be provided an opportunity to address the Board concerning each item on the agenda, and will be afforded the opportunity to comment on items not on the agenda at the end of the meeting or as solicited by the Chairperson. (See protocol for addressing the Board or submitting written comments below.) If you have a petition or other information pertaining to your subject, please present it to the clerk. Please sign the sign-in...
GENERAL GOVERNMENT MAYOR George D. Alpogianis TRUSTEES Susan Debartolo Morgan Dubiel Ajay Mody Craig Niedermaier Dean Strzelecki Marryann Warda VILLAGE MANAGER Frank Lindbloom VILLAGE CLERK Denise McCreery BICYCLE AND PEDESTRIAN PLAN ADVISORY GROUP IN-PERSON MEETING AGENDA Wednesday, November 19, 2025 4:00 p.m. Second Floor Conference Room #218 Niles Village Hall, 1000 Civic Center Dr, Niles IL 60714 Regular Meeting I. Call to Order II. Approval of Minutes a. October 20, 2025 Minutes III. New Business a. Police Accidents Report b. Engineering Updates c. Follow-up on Fatality at 5600 Touhy d. Waukegan/Shermer Intersection IV. Next Meeting a. TBD V. Comments from the Public VI. Adjournment 1000 Civic Center Drive Niles, IL 60714 Phone (847) 588-8000 Fax (847) 588-8051 WWW.VNILES.COM Regular Meeting	We're ready. I'm going to call the meeting to order. It's about um 6:31. So um let's go ahead and get ready to get started. Looking at our uh welcome to everyone. Let me just say welcome first to our uh November meeting for the design review board. Um we did delay the meeting by two days because of the election. So I appreciate everybody's patience on that. And um I wanted to say that um the purpose of the board is to review applications for certificate of appropriateness for construction, alteration, de demolition, relocation of structures and historic districts in an effort to help Achen retain its historic Um we're going to start by reviewing the meetings minutes from the previous meetings. In a minute I'm going to ask the um board for a vote on that. And um when we do start calling applicants to the uh podium, we would like for you to give us your name and address before you speak. Um and also that uh for members of the public to keep your comments to no more than five minutes. So,...

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	3.636
0.5333	1000	3.4658
0.8	1500	3.4657

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}
}