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_laptopjacke")
# Run inference
sentences = [
    "girls' winter jacket with a waterproof outer layer, a warm inner fleece, and adjustable features for comfort and fit.. girls' winter jacket with a waterproof outer layer, a warm inner fleece, and adjustable features for comfort and fit.",
    "Title: Columbia Girls' Bugaboo Ii Fleece Interchange Jacket Descripion: ['Finding an all-inclusive girls winter jacket that can be used in different weather conditions can be a challenge. Fortunately, our Bugaboo II Fleece Interchange Winter Jacket is the perfect multiple-use all-weather coat – utilizing our classic three-in-one design. It features a waterproof and breathable outer shell and an inner fleece layer that can be worn separately or zipped together for extra protection against wet and cold weather. The waterproof outer shell features an inner layer of our thermal heat reflective secret sauce we call Omni-HEAT. The engineered silver dots are designed to capture natural body heat, reducing weight while increasing comfort. The warm inner fleece jacket can be worn separately or zipped into the waterproof and breathable outer layer. This three-in-one jacket system features zippered hand pockets and adjustable cuffs, a taffeta lined removable storm hood, media and goggle pocket, fleece lined zippered hand pockets, and adjustable cuffs for added warmth control in evolving weather conditions. Available in a range of colors and youth sizes.']",
    'Title: ANOKA Yellowstone Jacket for Women Green XXL Descripion: [\'size details\'\n "Size:S----US:6----Bust:106cm/41.73\'\'----Sleeve:69cm/27.17\'\'----Length:71cm/27.95\'\' Size:M----US:8----Bust:110cm/43.31\'\'----Sleeve:70cm/27.56\'\'----Length:72cm/28.35\'\' Size:L----US:10----Bust:116cm/45.67\'\'----Sleeve:71cm/27.95\'\'----Length:73cm/28.74\'\' Size:XL----US:12----Bust:122cm/48.03\'\'----Sleeve:72cm/28.35\'\'----Length:74cm/29.13\'\' Size:XXL----US:14----Bust:128cm/50.39\'\'----Sleeve:73cm/28.74\'\'----Length:75cm/29.53\'\' Size:XXXL----US:16----Bust:134cm/52.76\'\'----Sleeve:74cm/29.13\'\'----Length:76cm/29.92\'\' Size:XXXXL----US:18----Bust:140cm/55.12\'\'----Sleeve:75cm/29.53\'\'----Length:77cm/30.31\'\' Size:XXXXXL----US:20----Bust:146cm/57.48\'\'----Sleeve:76cm/29.92\'\'----Length:78cm/30.71\'\'"]',
]
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: 15,123 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: 5 tokens
mean: 27.46 tokens
max: 101 tokens

min: 30 tokens
mean: 110.08 tokens
max: 128 tokens

	sentence_0	sentence_1
type	string	string
details	min: 5 tokens mean: 27.46 tokens max: 101 tokens	min: 30 tokens mean: 110.08 tokens max: 128 tokens

Samples:

sentence_0	sentence_1
`boys' winter jacket with a water-resistant exterior, thermal insulation, and an adjustable hood.`	Title: Columbia Boys' Powder Lite Hooded Winter Jacket Descripion: ["Our Powder Lite cold-weather jacket combines a classic fit with technology to keep boys warm and dry. Crafted of a water resistant shell, lined with Omni-HEAT reflective system, and packed with our Thermarator insulation, this coat will keep kids warm and comfortable when the weather turns blustery and cold… ready to help them take winter by storm. \u2028\u2028Complete with hood and a soft chin guard, zipped hand pockets to keep items secure, while a draw cord adjustable hem keeps the cold locked out, making for the perfect fit for active youths. This boy's winter jacket is available in many accommodating colors and boy sizes. To ensure the size you choose is right, utilize our sizing chart and the following measurement instructions: For the sleeves, start at the center back of your neck and measure across the shoulder and down to the sleeve. If you come up with a partial number, round up to the next even number. For the chest, measure at the fullest part of the chest, under the armpits and over the shoulder blades, keeping the tape measure firm and level.\u2028 Imported. \u2028Made from 100% polyester. \u2028Zippered closure. \u2028Machine Wash."]
`laptop with a large display, efficient processor, ample memory, and multiple connectivity options.`	Title: ASUS New VivoBook 15 15.6 Inch FHD 1080P Laptop (AMD Ryzen 3 3250U up to 3.5GHz, 12GB DDR4 RAM, 256GB SSD, AMD Radeon Vega 3, WiFi, Bluetooth, HDMI, Windows 10) (Grey) Descripion: ['XM sells computers with upgraded configurations. If the computer has modifications (listed above), then the manufacturer box is opened for it to be tested and inspected and to install the upgrades to achieve the specifications as advertised. Operating System: Windows 10 Home 64-bitDisplay: 15.6 inch FHD(1920 x 1080) with four-sided wider NanoEdge bezel displayProcessor: AMD Ryzen 3 3250U Processor (2.6 GHz base frequency up to 3.5 GHz, 2 Cores, 1MB Cache)Memory: Up to 16GB DDR4 RAMHard Drive: Up to 1TB SSDGraphics: AMD Radeon Vega 3Wireless: 802.11ac, Bluetooth 4.1Webcam: YESAudio features: Stereo speakersPorts: 1 x COMBO audio jack1 x Type-A USB 3.0 (USB 3.1 Gen 1)1 x Type-C USB 3.0 (USB 3.1 Gen 1)2 x USB 2.0 port(s)1 x HDMIBattery Type: 2 -Cell 37 Wh BatteryWeight: 3.75lbsDimensions: 14.4 x 9.1 x 0.8 inchesColor: Gray']
`men's parka with a hood, featuring a waterproof design, secure storage options, and insulation for warmth.. men's parka with a hood, featuring a waterproof design, secure storage options, and insulation for warmth.`	`Title: Under Armour Men's Unstoppable Waterproof Hooded Down Parka Project Rock Long STORM Jacket Descripion: ['Under Armour Project Rock Hooded Down Parka Jacket Black 1346093-001 Storm technology: breathable and waterproof, Secure pockets. Zip/snap closure. Down fill.']`

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
4.2017	500	1.9644

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_laptopjacke