metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:72
  - loss:ContrastiveLoss
base_model: sentence-transformers/multi-qa-mpnet-base-dot-v1
widget:
  - source_sentence: >-
      What was the original purpose of the Basilica di San Lorenzo's design by
      Filippo Brunelleschi in 1419?
    sentences:
      - ' It is one of several churches that claim to be the oldest in Florence, having been consecrated in 393 AD, at which time it stood outside the city walls.'
      - >-
        The Palazzo Pitti, in English sometimes called the Pitti Palace, is a
        vast, mainly Renaissance, palace in Florence, Italy. It is situated on
        the south side of the River Arno in Pitti Square, a short distance from
        the Ponte Vecchio.
      - ' The architects were Mariano Falcini, Professor Vincenzo Micheli, and Marco Treves, who was Jewish. '
  - source_sentence: >-
      What is the name of the architect who expanded the façade and the rear
      section of the Palazzo Pitti in 1549?
    sentences:
      - ' The palace was left incomplete by Simone del Pollaiolo (il Cronaca), who was in charge of the construction of the palace until 1504. '
      - >-
        The Palazzo Pitti, in English sometimes called the Pitti Palace, is a
        vast, mainly Renaissance, palace in Florence, Italy. It is situated on
        the south side of the River Arno in Pitti Square, a short distance from
        the Ponte Vecchio.
      - ' In 1939, these were joined by the Palestrina Pietà, discovered in the Barberini chapel in Palestrina, though experts now consider its attribution to Michelangelo to be dubious. '
  - source_sentence: When did the Uffizi Gallery officially open to the public?
    sentences:
      - ' The project was intended to display prime artworks of the Medici collections on the piano nobile; the plan was carried out by his son, Grand Duke Francesco I.'
      - ' The gallery had been open to visitors by request since the sixteenth century, and in 1769 it was officially opened to the public, formally becoming a museum in 1865.'
      - ' In 1939, these were joined by the Palestrina Pietà, discovered in the Barberini chapel in Palestrina, though experts now consider its attribution to Michelangelo to be dubious. '
  - source_sentence: >-
      When was the first church on the site of the current Santa Felicita church
      in Florence probably built?
    sentences:
      - ' The project was intended to display prime artworks of the Medici collections on the piano nobile; the plan was carried out by his son, Grand Duke Francesco I.'
      - ' It was employed as a prison; executions took place in the Bargello''s yard until they were abolished by Grand Duke Peter Leopold in 1786, but it remained the headquarters of the Florentine police until 1859.'
      - >-
        Santa Felicita (Church of St Felicity) is a Roman Catholic church in
        Florence, region of Tuscany, Italy, probably the oldest in the city
        after San Lorenzo. 
  - source_sentence: What was the original purpose of the building in 1255?
    sentences:
      - ' The palace was built to house first the Capitano del Popolo and later, in 1261, the ''podestà'', the highest magistrate of the Florence City Council.'
      - >-
        The Ponte Vecchio is a medieval stone closed-spandrel segmental arch
        bridge over the Arno, in Florence, Italy. It is the only bridge in
        Florence spared from destruction during World War II and is noted for
        the shops built along it, a practice that was once common on bridges.
        Initially, these shops were occupied by butchers, tanners, and farmers,
        but today they are home to jewellers, art dealers, and souvenir sellers.
      - ' The door retains its original massive, iron-clad doors. '
pipeline_tag: sentence-similarity
library_name: sentence-transformers

SentenceTransformer based on sentence-transformers/multi-qa-mpnet-base-dot-v1

This is a sentence-transformers model finetuned from sentence-transformers/multi-qa-mpnet-base-dot-v1. 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/multi-qa-mpnet-base-dot-v1
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 dimensions
Similarity Function: Dot Product

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: MPNetModel 
  (1): Pooling({'word_embedding_dimension': 768, '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("Marco127/D1_finetuned_2_test_1")
# Run inference
sentences = [
    'What was the original purpose of the building in 1255?',
    " The palace was built to house first the Capitano del Popolo and later, in 1261, the 'podestà', the highest magistrate of the Florence City Council.",
    ' The door retains its original massive, iron-clad doors. ',
]
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: 72 training samples
Columns: sentence1, sentence2, and label
Approximate statistics based on the first 72 samples:
sentence1 sentence2 label
type string string int
details
min: 14 tokens
mean: 24.06 tokens
max: 36 tokens

min: 2 tokens
mean: 36.08 tokens
max: 98 tokens

0: ~50.00%
1: ~50.00%

	sentence1	sentence2	label
type	string	string	int
details	min: 14 tokens mean: 24.06 tokens max: 36 tokens	min: 2 tokens mean: 36.08 tokens max: 98 tokens	0: ~50.00% 1: ~50.00%

Samples:

sentence1	sentence2	label
`What was the name of the first owner of the Palazzo Pitti, and in which year did he die?`	`The Palazzo Pitti, in English sometimes called the Pitti Palace, is a vast, mainly Renaissance, palace in Florence, Italy. It is situated on the south side of the River Arno in Pitti Square, a short distance from the Ponte Vecchio.`	`1`
`What is the name of the architect who expanded the façade and the rear section of the Palazzo Pitti in 1549?`	`The palace became a great treasure house as generations of the Medici and subsequent dynasties amassed paintings, plates, jewelry, and luxurious possessions. Today, the Palazzo Pitti is the largest museum complex in Florence, divided into several principal galleries or museums.`	`1`
`What was the name of the first owner of the Palazzo Pitti, and in which year did he die?`	`The palace became a great treasure house as generations of the Medici and subsequent dynasties amassed paintings, plates, jewelry, and luxurious possessions. Today, the Palazzo Pitti is the largest museum complex in Florence, divided into several principal galleries or museums.`	`0`

Loss: ContrastiveLoss with these parameters:

{
    "distance_metric": "SiameseDistanceMetric.COSINE_DISTANCE",
    "margin": 0.5,
    "size_average": true
}

Training Hyperparameters

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 8
per_device_eval_batch_size: 8
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: 1.0
num_train_epochs: 3.0
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: proportional

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.3.1
Transformers: 4.47.1
PyTorch: 2.5.1+cu121
Accelerate: 1.2.1
Datasets: 3.2.0
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",
}

ContrastiveLoss

@inproceedings{hadsell2006dimensionality,
    author={Hadsell, R. and Chopra, S. and LeCun, Y.},
    booktitle={2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)},
    title={Dimensionality Reduction by Learning an Invariant Mapping},
    year={2006},
    volume={2},
    number={},
    pages={1735-1742},
    doi={10.1109/CVPR.2006.100}
}