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: 128 tokens
Output Dimensionality: 768 tokens
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}) with Transformer model: 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})
)

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("knguyennguyen/mpnet_20k")
# Run inference
sentences = [
    "I'm looking for a casual short-sleeve top with a fun and stylish design for women. It should have a round neck and a playful message, perfect for everyday wear. I'd prefer something that feels comfortable and has a unique print.",
    'Title: HONTOUTE Women Mom Boss T Shirt Funny Leopard Letter Print Shirts with Saying Vintage O Neck Short Sleeve Tees Casual Tops Descripion: [\'Women Mom Boss T-Shirt Funny Leopard Letters Printed Shirts with Saying Vintage Round Neck Short Sleeve Tees Cute Casual Tops\'\n \'Size Chart:(1inch=2.54cm)\'\n \'Size S: Length 66cm/25.98" Bust 94cm/37.01" Size M: Length 67cm/26.38" Bust 98cm/38.58" Size L: Length 68cm/26.77" Bust 102cm/40.16" Size XL: Length 69cm/27.17" Bust 110cm/43.31" Please allow slight (±3cm)manual measurement deviation for the data The real color of the item may be slightly different from the pictures shown on website，caused by many factors such as brightness of your monitor and light brightness\'\n \'Two Ways About Delivery:\' \'FBM:\'\n \'Ship from China,88% customers will receive within 2 weeks,9.9% lucky dog will receive within 1 week,and others will receive within 3-4 weeks\'\n \'FBA:\' \'Customers will receive within 1-3 days\' \'Service Guarantee:\'\n \'We endeavors 100% customer satisfaction service and experience If you receive damaged or wrong items Please contact us with attached pictures about the problem We will provide you a satisfactory solution within 24 hours You may find that someone sells at a lower price than us But they cannot guarantee the same quality and service as we do If you are satisfied with our product or service Hope you can leave your positive feedback\']',
    'Title: Batman: Gotham By Gaslight Descripion: ["It\'s Batman vs. Jack the Ripper in an Elseworld\'s adventure that imagines the Dark Knight over a hundred years ago in a turn-of-the-century Gotham."]',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

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

Training Details

Training Dataset

Unnamed Dataset

Size: 20,108 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: 11 tokens
mean: 36.23 tokens
max: 73 tokens

min: 13 tokens
mean: 88.42 tokens
max: 128 tokens

	sentence_0	sentence_1
type	string	string
details	min: 11 tokens mean: 36.23 tokens max: 73 tokens	min: 13 tokens mean: 88.42 tokens max: 128 tokens

Samples:

sentence_0	sentence_1
`I'm looking for a stylish pair of eyewear with a luxurious touch. They should have a unique color combination and come with a special case and cleaning accessory.`	`Title: Sunglasses Gucci GG 0528 S- 008 GOLD/BROWN CRYSTAL, 63-14-150 Descripion: ['Authentic Gucci GG0528 S 008 Gold Crystal/Brown Sunglasses. Comes with a matching satin flannel pouch and ivory microfiber cloth and Authenticity card.']`
`I'm looking for comfortable and stylish capri pants for girls that are easy to wear and have a stretchy fit.`	`Title: French Toast Girls' Stretch Skinny Pull-on Capri Pant Descripion: ["Easy to wear and even easier to love! French Toast's classroom capri features a simple navy and white elastic stripe on the waistband, functional front and back pockets and pull-on styling, making it even to easier to get her dressed and out the door."]`
`I'm in need of a replacement screen for a laptop that offers clear visuals and fits a specific model. It should provide high-definition quality for general use.`	`Title: BRIGHTFOCAL New Screen Replacement for HP 14-CF0006DX HD 1366x768 LCD LED Display Panel Descripion: ['BRIGHTFOCAL New Screen Replacement for HP 14-CF0006DX HD 1366x768 LCD LED Display Panel']`

Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 128
per_device_eval_batch_size: 128
num_train_epochs: 5
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: 128
per_device_eval_batch_size: 128
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: 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: False
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
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
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss
3.1646	500	0.493

Framework Versions

Python: 3.11.11
Sentence Transformers: 3.1.1
Transformers: 4.45.2
PyTorch: 2.5.1+cu121
Accelerate: 1.2.1
Datasets: 3.2.0
Tokenizers: 0.20.3

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}
}

knguyennguyen
/

mpnet_20k