SentenceTransformer

This is a sentence-transformers model trained for semantic text understanding. It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Maximum Sequence Length: 128 tokens
Output Dimensionality: 384 dimensions
Similarity Function: Cosine Similarity
Supported Modality: Text

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'transformer_task': 'feature-extraction', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'last_hidden_state'}}, 'module_output_name': 'token_embeddings', 'architecture': 'BertModel'})
  (1): Pooling({'embedding_dimension': 384, 'pooling_mode': 'mean', '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("HarishMaths/Hotel-Policy-Embedding")
# Run inference
sentences = [
    'Smoking or vaping is allowed in designated rooms.',
    'Marriott is committed to providing its guests and associates with a smoke-free environment, and is proud to boast one of the most comprehensive smoke-free hotel policies in the industry.',
    'Some upgrades to Premium Rooms require payment in local currency and cannot be purchased with Points.',
]
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, 0.9151, 0.2602],
#         [0.9151, 1.0000, 0.3726],
#         [0.2602, 0.3726, 1.0000]])

Evaluation

Metrics

Semantic Similarity

Dataset: val
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.6244
spearman_cosine	0.6463

Training Details

Training Dataset

Unnamed Dataset

Size: 3,872 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 100 samples:
sentence_0 sentence_1
type string string

modality text text
details
min: 6 tokens
mean: 24.39 tokens
max: 117 tokens

min: 9 tokens
mean: 27.12 tokens
max: 103 tokens

	sentence_0	sentence_1
type	string	string
modality	text	text
details	min: 6 tokens mean: 24.39 tokens max: 117 tokens	min: 9 tokens mean: 27.12 tokens max: 103 tokens

Samples:

sentence_0	sentence_1
`Three months after booking and 20 days after purchasing insurance, the tour operator files bankruptcy and ceases all operations.`	`California SB 644 requires hotels and third-party booking sites to give a full refund when a guest cancels within 24 hours of booking, as long as the reservation was made at least 72 hours before check-in.`
`One of the easiest ways to avoid resort fees is by booking an award stay.`	`For a typical domestic Hilton hotel, it will show you the points options and cash rates all on one screen.`
`The General Contractor went out to several Home Depot locations around the City to find over 250 battery-operated smoke detectors.`	`Modern sensors detect vapour, and the resulting charge is identical to a cigarette violation.`

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim",
    "gather_across_devices": false,
    "directions": [
        "query_to_doc"
    ],
    "partition_mode": "joint",
    "hardness_mode": null,
    "hardness_strength": 0.0
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 32
num_train_epochs: 25
fp16: True
per_device_eval_batch_size: 32
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

per_device_train_batch_size: 32
num_train_epochs: 25
max_steps: -1
learning_rate: 5e-05
lr_scheduler_type: linear
lr_scheduler_kwargs: None
warmup_steps: 0
optim: adamw_torch_fused
optim_args: None
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
optim_target_modules: None
gradient_accumulation_steps: 1
average_tokens_across_devices: True
max_grad_norm: 1
label_smoothing_factor: 0.0
bf16: False
fp16: True
bf16_full_eval: False
fp16_full_eval: False
tf32: None
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
use_liger_kernel: False
liger_kernel_config: None
use_cache: False
neftune_noise_alpha: None
torch_empty_cache_steps: None
auto_find_batch_size: False
log_on_each_node: True
logging_nan_inf_filter: True
include_num_input_tokens_seen: no
log_level: passive
log_level_replica: warning
disable_tqdm: False
project: huggingface
trackio_space_id: None
trackio_bucket_id: None
trackio_static_space_id: None
per_device_eval_batch_size: 32
prediction_loss_only: True
eval_on_start: False
eval_do_concat_batches: True
eval_use_gather_object: False
eval_accumulation_steps: None
include_for_metrics: []
batch_eval_metrics: False
save_only_model: False
save_on_each_node: False
enable_jit_checkpoint: False
push_to_hub: False
hub_private_repo: None
hub_model_id: None
hub_strategy: every_save
hub_always_push: False
hub_revision: None
load_best_model_at_end: False
ignore_data_skip: False
restore_callback_states_from_checkpoint: False
full_determinism: False
seed: 42
data_seed: None
use_cpu: False
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
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_pin_memory: True
dataloader_persistent_workers: False
dataloader_prefetch_factor: None
dataloader_multiprocessing_context: None
dataloader_in_order: True
remove_unused_columns: True
label_names: None
train_sampling_strategy: random
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
ddp_static_graph: None
ddp_backend: None
ddp_timeout: 1800
fsdp: None
fsdp_config: None
deepspeed: None
debug: []
skip_memory_metrics: True
do_predict: False
resume_from_checkpoint: None
local_rank: -1
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}
warmup_ratio: None

Training Logs

Epoch	Step	Training Loss	val_spearman_cosine
-1	-1	-	0.1722
0.8264	100	-	0.2700
1.0	121	-	0.2967
1.6529	200	-	0.4105
2.0	242	-	0.4552
2.4793	300	-	0.4991
3.0	363	-	0.5376
3.3058	400	-	0.5566
4.0	484	-	0.5790
4.1322	500	3.5721	0.5825
4.9587	600	-	0.6156
5.0	605	-	0.6162
5.7851	700	-	0.6144
6.0	726	-	0.6166
6.6116	800	-	0.6266
7.0	847	-	0.6269
7.4380	900	-	0.6348
8.0	968	-	0.6276
8.2645	1000	2.4308	0.6329
9.0	1089	-	0.6336
9.0909	1100	-	0.6332
9.9174	1200	-	0.6399
10.0	1210	-	0.6381
10.7438	1300	-	0.6390
11.0	1331	-	0.6397
11.5702	1400	-	0.6446
12.0	1452	-	0.6418
12.3967	1500	2.1294	0.6463

Framework Versions

Python: 3.13.15
Sentence Transformers: 5.7.0
Transformers: 5.16.1
PyTorch: 2.11.0+cu128
Accelerate: 1.14.0
Datasets: 4.8.5
Tokenizers: 0.23.1

Downloads last month: 294

Safetensors

Model size

22.7M params

Tensor type

F32

Evaluation results

Pearson Cosine on val
self-reported

0.624
Spearman Cosine on val
self-reported

0.646