metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:25743
  - loss:MultipleNegativesRankingLoss
base_model: WhereIsAI/UAE-Large-V1
widget:
  - source_sentence: >-
      2:00 PM Facebook ... 0.0KB/sill Arief Smansa Fadhillah Jun 9 at 9:44 am. 
      89 111 60 = If in the next 2 weeks the people America who violates PSBB
      will not happen corpses scattered on the streets, then I sure that the
      fear of Corona is just a scam created by WHO and in Support by Mass Media.
    sentences:
      - MPs are entitled to a full pension after six months in office
      - Photos of anti-racism demonstrations in the United States
      - Wisconsin has more votes cast than registered voters.
  - source_sentence: >-
      A religious festival in Jaffna... Radical Otulabban, who opposes the
      ordination of children, has nothing to do with this...
    sentences:
      - >-
        A genuine article on Olympic female weightlifter suffering testicle
        injury?
      - This video shows pilots demonstrating against Covid vaccines
      - Photo shows distressed children at a religious ritual in Sri Lanka
  - source_sentence: >-
      ← 42 CHANNEL Markus Hain... * 107.4K subscribers Pinned message If you
      like my work for our freedom... 74% 22:32 KANAL Markus Haintz, Lawyer &
      Fre... forwarded message By Vicky_TheRedSparrow BREAKING NEWS: The Supreme
      Court of Justice in the United States decided that the Covid vaccination
      no vaccine is unsafe and um must be avoided at all costs - Big Pharma and
      Anthony Fauci have lost a lawsuit by Robert F. Kennedy Jr. and a group of
      scientists has been submitted!  /breaking-news-the-supreme-court
      -in-the-us-has-ruled-that-the-covid -pathogen-is-not-a-vaccine-is-unsafe
      -and-must-be-avoided-at-all-costs-big -pharma-and-anthony-fauci-have-lost
      -a-lawsuit-filed-by-r/ Truth To Power BREAKING NEWS: The Supreme Court In
      The US Has Ruled That The Covid Dathanen in Distress & Vanaina la Llunafn
      MUTE OFF X 138
    sentences:
      - 'USA: Supreme Court rules against corona vaccinations'
      - >-
        Pakistani government appoints former army general to head medical
        regulatory body
      - >-
        "In Denmark, the law obliges owners of large agricultural land to plant
        5% of their land flowers for bees. In Portugal?"
  - source_sentence: MEXICO, Failed extortion in Celaya… and he came back to throw a grenade ….
    sentences:
      - Attack on people in a cafe in Celaya, Mexico
      - UNICEF issued guidelines for the prevention of coronavirus infections
      - Image shows a road in Sri Lanka
  - source_sentence: >-
      The ELN movement supported with 80 thousand dollars! That is little money.
      What's wrong with that? For us, nor the FARC nor the ELN they are groups
      terrorists ” revores Arauz PRISI ANDRES ARAUZLela campaign with funds from
      drug traffickers and terrorists
    sentences:
      - >-
        Andrés Arauz said that he accepted financing from the ELN and that
        neither the ELN nor the FARC are armed groups
      - Holy communion banned in Toronto
      - >-
        Myanmar leader gives three-fingered salute in support of Thai
        protesters?
pipeline_tag: sentence-similarity
library_name: sentence-transformers

SentenceTransformer based on WhereIsAI/UAE-Large-V1

This is a sentence-transformers model finetuned from WhereIsAI/UAE-Large-V1. It maps sentences & paragraphs to a 1024-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: WhereIsAI/UAE-Large-V1
Maximum Sequence Length: 512 tokens
Output Dimensionality: 1024 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': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)

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 = [
    "The ELN movement supported with 80 thousand dollars! That is little money. What's wrong with that? For us, nor the FARC nor the ELN they are groups terrorists ” revores Arauz PRISI ANDRES ARAUZLela campaign with funds from drug traffickers and terrorists",
    'Andrés Arauz said that he accepted financing from the ELN and that neither the ELN nor the FARC are armed groups',
    'Holy communion banned in Toronto',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

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

Training Details

Training Dataset

Unnamed Dataset

Size: 25,743 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: 2 tokens
mean: 109.01 tokens
max: 512 tokens

min: 5 tokens
mean: 18.19 tokens
max: 131 tokens

min: 1.0
mean: 1.0
max: 1.0

	sentence_0	sentence_1	label
type	string	string	float
details	min: 2 tokens mean: 109.01 tokens max: 512 tokens	min: 5 tokens mean: 18.19 tokens max: 131 tokens	min: 1.0 mean: 1.0 max: 1.0

Samples:

sentence_0	sentence_1	label
`In the coming weeks and months, You will see the bananas with more pints of normal, due to the effect of the ashes of the volcano! Don't stop buying them! It only affects the image not the taste! Crops need to be harvested so that the banana trees can come out ahead! alamy a a alam alamy`	`Canary bananas are going to have more spots than normal due to the effect of the ashes of the volcano`	`1.0`
`Are they canceling Title of those who are over 70 years old!? Negative certificate Electoral registry office, says I owe nothing. But at the bottom of the page. it says "unsubscribed"! Over 70s must check that everything is in order with their title. Millions of retirees can vote for Bolsonaro.`	`Population over 70 is having the voter registration canceled in 2022`	`1.0`
`VIN dti PHILIPPINES FDA APPROVED Honey-C H52% 18:43 itine Appemess Vinity Resistance Bus KONTRA CORONA VIRUS Let's boost our immune system!`	`Government-approved immunity booster for COVID-19 sold online`	`1.0`

Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 2
per_device_eval_batch_size: 2
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: 2
per_device_eval_batch_size: 2
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
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
0.0388	500	0.0473
0.0777	1000	0.0264
0.1165	1500	0.0258
0.1554	2000	0.0322
0.1942	2500	0.0225
0.2331	3000	0.0318
0.2719	3500	0.036
0.3108	4000	0.0254
0.3496	4500	0.0166
0.3884	5000	0.0231
0.4273	5500	0.0268
0.4661	6000	0.0293
0.5050	6500	0.0315
0.5438	7000	0.0292
0.5827	7500	0.0308
0.6215	8000	0.0206
0.6603	8500	0.0329
0.6992	9000	0.0379
0.7380	9500	0.0133
0.7769	10000	0.0255
0.8157	10500	0.0138
0.8546	11000	0.0414
0.8934	11500	0.015
0.9323	12000	0.0234
0.9711	12500	0.0274

Framework Versions

Python: 3.11.11
Sentence Transformers: 3.4.1
Transformers: 4.48.3
PyTorch: 2.5.1+cu124
Accelerate: 1.3.0
Datasets: 3.3.1
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}
}