This is a sentence-transformers model finetuned from intfloat/multilingual-e5-small. It maps sentences & paragraphs to a 384-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: intfloat/multilingual-e5-small
Maximum Sequence Length: 512 tokens
Output Dimensionality: 384 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': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, '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})
  (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.

python

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("srikarvar/multilingual-e5-small-pairclass-1")
# Run inference
sentences = [
    'What is the boiling point of water at sea level?',
    'What is the melting point of ice at sea level?',
    'Can you recommend a good hotel nearby?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

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

Evaluation

Metrics

Binary Classification

Dataset: pair-class-dev
Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.8683
cosine_accuracy_threshold	0.8612
cosine_f1	0.8621
cosine_f1_threshold	0.8612
cosine_precision	0.8065
cosine_recall	0.9259
cosine_ap	0.9228
dot_accuracy	0.8683
dot_accuracy_threshold	0.8612
dot_f1	0.8621
dot_f1_threshold	0.8612
dot_precision	0.8065
dot_recall	0.9259
dot_ap	0.9228
manhattan_accuracy	0.8642
manhattan_accuracy_threshold	7.6678
manhattan_f1	0.8559
manhattan_f1_threshold	8.1834
manhattan_precision	0.8099
manhattan_recall	0.9074
manhattan_ap	0.9202
euclidean_accuracy	0.8683
euclidean_accuracy_threshold	0.5269
euclidean_f1	0.8621
euclidean_f1_threshold	0.5269
euclidean_precision	0.8065
euclidean_recall	0.9259
euclidean_ap	0.9228
max_accuracy	0.8683
max_accuracy_threshold	7.6678
max_f1	0.8621
max_f1_threshold	8.1834
max_precision	0.8099
max_recall	0.9259
max_ap	0.9228

Binary Classification

Dataset: pair-class-test
Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.8683
cosine_accuracy_threshold	0.8612
cosine_f1	0.8621
cosine_f1_threshold	0.8612
cosine_precision	0.8065
cosine_recall	0.9259
cosine_ap	0.9228
dot_accuracy	0.8683
dot_accuracy_threshold	0.8612
dot_f1	0.8621
dot_f1_threshold	0.8612
dot_precision	0.8065
dot_recall	0.9259
dot_ap	0.9228
manhattan_accuracy	0.8642
manhattan_accuracy_threshold	7.6678
manhattan_f1	0.8559
manhattan_f1_threshold	8.1834
manhattan_precision	0.8099
manhattan_recall	0.9074
manhattan_ap	0.9202
euclidean_accuracy	0.8683
euclidean_accuracy_threshold	0.5269
euclidean_f1	0.8621
euclidean_f1_threshold	0.5269
euclidean_precision	0.8065
euclidean_recall	0.9259
euclidean_ap	0.9228
max_accuracy	0.8683
max_accuracy_threshold	7.6678
max_f1	0.8621
max_f1_threshold	8.1834
max_precision	0.8099
max_recall	0.9259
max_ap	0.9228

Training Details

Training Dataset

Unnamed Dataset

Size: 971 training samples
Columns: label, sentence1, and sentence2
Approximate statistics based on the first 1000 samples:
label sentence1 sentence2
type int string string
details
0: ~48.61%
1: ~51.39%
min: 6 tokens
mean: 10.82 tokens
max: 22 tokens
min: 4 tokens
mean: 10.12 tokens
max: 22 tokens

Samples:

label	sentence1	sentence2
`1`	`How many bones are in the human body?`	`Total number of bones in an adult human body`
`0`	`What is the largest lake in North America?`	`What is the largest river in North America?`
`0`	`What is the capital of New Zealand?`	`What is the capital of Australia?`

Loss: OnlineContrastiveLoss

Evaluation Dataset

Unnamed Dataset

Size: 243 evaluation samples
Columns: label, sentence1, and sentence2
Approximate statistics based on the first 1000 samples:
label sentence1 sentence2
type int string string
details
0: ~55.56%
1: ~44.44%
min: 6 tokens
mean: 10.55 tokens
max: 22 tokens
min: 4 tokens
mean: 10.09 tokens
max: 20 tokens

Samples:

label	sentence1	sentence2
`1`	`What are the different types of renewable energy?`	`What are the various forms of renewable energy?`
`1`	`Who discovered gravity?`	`Gravity discoverer`
`0`	`Can you help me understand this report?`	`Can you help me write this report?`

Loss: OnlineContrastiveLoss

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: epoch
per_device_train_batch_size: 32
per_device_eval_batch_size: 32
gradient_accumulation_steps: 2
learning_rate: 1e-06
weight_decay: 0.01
num_train_epochs: 12
lr_scheduler_type: reduce_lr_on_plateau
warmup_ratio: 0.1
load_best_model_at_end: True
optim: adamw_torch_fused

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: epoch
prediction_loss_only: True
per_device_train_batch_size: 32
per_device_eval_batch_size: 32
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 2
eval_accumulation_steps: None
learning_rate: 1e-06
weight_decay: 0.01
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 12
max_steps: -1
lr_scheduler_type: reduce_lr_on_plateau
lr_scheduler_kwargs: {}
warmup_ratio: 0.1
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: True
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_fused
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
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	loss	pair-class-dev_max_ap	pair-class-test_max_ap
0	0	-	0.6426	-
0.9677	15	4.5769	0.6975	-
2.0	31	3.8280	0.7466	-
2.9677	46	3.1501	0.7848	-
4.0	62	2.8302	0.8220	-
4.9677	77	2.4840	0.8469	-
6.0	93	2.2746	0.8692	-
6.9677	108	2.0923	0.8835	-
8.0	124	1.9265	0.8962	-
8.9677	139	1.8076	0.9048	-
10.0	155	1.7673	0.9130	-
10.9677	170	1.6653	0.9201	-
11.6129	180	1.5428	0.9228	0.9228

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.1.2+cu121
Accelerate: 0.32.1
Datasets: 2.19.1
Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

bibtex

@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",
}

	label	sentence1	sentence2
type	int	string	string
details	0: ~48.61% 1: ~51.39%	min: 6 tokens mean: 10.82 tokens max: 22 tokens	min: 4 tokens mean: 10.12 tokens max: 22 tokens