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 dimensions
Similarity Function: Cosine Similarity

Language: en
License: apache-2.0

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("Mohamed-Gamil/multilingual-e5-small-JapaneseTeacher")
# Run inference
sentences = [
    "page_content='Do native speakers correctly use counters every time? The answer is no. We often make mistakes or intentionally use wrong ones for the sake of simplicity.  \nCounter: 膳  \n<!-- 🖼️❌ Image not available. Please use `PdfPipelineOptions(generate_picture_images=True)` -->  \nFor example, when you count chopsticks, the counter: 膳 ぜん is right. However, a lot of people count chopsticks with 本 ほん . Although it's not right, 本 ほん is applicable because of the form of chopsticks. What we wanted to say here is that you **don't need be a perfectionist** . Of course, it's better that you can use every counter correctly, but actually, it is a fact that most native speakers can't do it themselves.' metadata={'h3': 'Practical Usages in Reality'}",
    'Are there situations where using a wrong counter is acceptable?',
    'Can you give an example of a casual sentence versus a polite one in Japanese?',
]
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

Information Retrieval

Dataset: ir_evaluator
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6637
cosine_accuracy@3	0.8561
cosine_accuracy@5	0.9103
cosine_accuracy@10	0.9475
cosine_precision@1	0.6637
cosine_precision@3	0.2854
cosine_precision@5	0.1822
cosine_precision@10	0.0949
cosine_recall@1	0.6633
cosine_recall@3	0.8557
cosine_recall@5	0.9099
cosine_recall@10	0.9475
cosine_ndcg@10	0.8122
cosine_mrr@10	0.7681
cosine_map@100	0.7702

Training Details

Training Dataset

Unnamed Dataset

Size: 4,961 training samples
Columns: positive and anchor
Approximate statistics based on the first 1000 samples:
positive anchor
type string string
details
min: 50 tokens
mean: 327.75 tokens
max: 512 tokens
min: 10 tokens
mean: 21.79 tokens
max: 40 tokens
Samples:
positive anchor
page_content=' 私わたしが困こまっていたとき、
page_content=' 私わたしが困こまっていたとき、
page_content=' 私わたしが困こまっていたとき、
Loss: MultipleNegativesRankingLoss with these parameters:
json
1{ 2 "scale": 20.0, 3 "similarity_fct": "cos_sim" 4}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: epoch
per_device_train_batch_size: 16
gradient_accumulation_steps: 32
learning_rate: 2e-05
num_train_epochs: 20
lr_scheduler_type: cosine
warmup_ratio: 0.1
bf16: True
dataloader_num_workers: 2
load_best_model_at_end: True
optim: adamw_torch_fused
dataloader_pin_memory: False
gradient_checkpointing: True
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: epoch
prediction_loss_only: True
per_device_train_batch_size: 16
per_device_eval_batch_size: 8
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 32
eval_accumulation_steps: None
torch_empty_cache_steps: None
learning_rate: 2e-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: 20
max_steps: -1
lr_scheduler_type: cosine
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: True
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: 2
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: False
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: True
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
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: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	ir_evaluator_cosine_ndcg@10
-1	-1	-	0.5500
1.0	10	67.2044	0.6343
2.0	20	49.2443	0.6757
3.0	30	21.3377	0.7179
4.0	40	8.6437	0.7687
5.0	50	5.8509	0.7862
6.0	60	5.0683	0.7905
7.0	70	3.6658	0.8006
8.0	80	3.5062	0.8011
9.0	90	3.0544	0.8055
10.0	100	2.7832	0.8060
11.0	110	2.743	0.8090
12.0	120	2.3785	0.8056
13.0	130	2.3046	0.8069
14.0	140	2.4136	0.8119
15.0	150	2.3528	0.8119
16.0	160	2.032	0.8115
17.0	170	2.1875	0.8115
18.0	180	2.0299	0.8124
19.0	190	2.1747	0.8126
20.0	200	2.1729	0.8122

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.11.13
Sentence Transformers: 4.1.0
Transformers: 4.52.4
PyTorch: 2.6.0+cu124
Accelerate: 1.8.1
Datasets: 3.6.0
Tokenizers: 0.21.2

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",
}

MultipleNegativesRankingLoss

bibtex

@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}
}

	positive	anchor
type	string	string
details	min: 50 tokens mean: 327.75 tokens max: 512 tokens	min: 10 tokens mean: 21.79 tokens max: 40 tokens

positive	anchor
`page_content='`	私わたしが困こまっていたとき、
`page_content='`	私わたしが困こまっていたとき、
`page_content='`	私わたしが困こまっていたとき、