This is a sentence-transformers model finetuned from distilbert/distilbert-base-multilingual-cased. 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: distilbert/distilbert-base-multilingual-cased
Maximum Sequence Length: 512 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': 512, 'do_lower_case': False}) with Transformer model: DistilBertModel 
  (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.

python

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("pritamdeka/distilbert-base-multilingual-cased-indicxnli-random-negatives-v1")
# Run inference
sentences = [
    'মই ভালদৰে জানিব নোৱাৰোঁ আপোনালোকৰ সৈতে কথা বতৰা আৰু এক ভাল সন্ধ্যা আছিল',
    'মই নিশ্চিত নহয় কিন্তু মই অলপ ভাল, আজি ৰাতি আপোনালোকৰ সৈতে কথা পাতিবলৈ পাই ভাল লাগিল।',
    'Shannon এ বাৰ্তা উপেক্ষা কৰিছে।',
]
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]

Evaluation

Metrics

Semantic Similarity

Dataset: pritamdeka/stsb-assamese-translated-dev
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.717
spearman_cosine	0.7221
pearson_manhattan	0.738
spearman_manhattan	0.7452
pearson_euclidean	0.7387
spearman_euclidean	0.7459
pearson_dot	0.6481
spearman_dot	0.6478
pearson_max	0.7387
spearman_max	0.7459

Semantic Similarity

Dataset: pritamdeka/stsb-assamese-translated-test
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.6568
spearman_cosine	0.6622
pearson_manhattan	0.6675
spearman_manhattan	0.6722
pearson_euclidean	0.6682
spearman_euclidean	0.6727
pearson_dot	0.5692
spearman_dot	0.5709
pearson_max	0.6682
spearman_max	0.6727

Training Details

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 64
per_device_eval_batch_size: 64
num_train_epochs: 1
warmup_ratio: 0.1
fp16: True
load_best_model_at_end: True
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 64
per_device_eval_batch_size: 64
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_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: 1
max_steps: -1
lr_scheduler_type: linear
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: True
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
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
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	pritamdeka/stsb-assamese-translated-dev_spearman_cosine	pritamdeka/stsb-assamese-translated-test_spearman_cosine
0	0	-	-	0.5489	-
0.0489	500	1.9387	1.7308	0.6808	-
0.0978	1000	1.0503	1.7373	0.6689	-
0.1467	1500	0.92	1.5838	0.6761	-
0.1956	2000	0.8754	1.4807	0.6518	-
0.2445	2500	0.7988	1.3797	0.6853	-
0.2933	3000	0.7606	1.3713	0.7108	-
0.3422	3500	0.7228	1.2510	0.6677	-
0.3911	4000	0.688	1.2374	0.6734	-
0.4400	4500	0.6992	1.2173	0.6891	-
0.4889	5000	0.6108	1.1638	0.7017	-
0.5378	5500	0.612	1.0815	0.7102	-
0.5867	6000	0.6259	1.0664	0.7202	-
0.6356	6500	0.5863	1.0464	0.7047	-
0.6845	7000	0.5941	1.0111	0.7101	-
0.7334	7500	0.5436	1.0023	0.7171	-
0.7822	8000	0.555	0.9633	0.7202	-
0.8311	8500	0.5466	0.9651	0.7279	-
0.8800	9000	0.5326	0.9611	0.7262	-
0.9289	9500	0.5055	0.9313	0.7276	-
0.9778	10000	0.4828	0.9172	0.7221	-
1.0	10227	-	-	-	0.6622

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.42.4
PyTorch: 2.3.1+cu121
Accelerate: 0.32.1
Datasets: 2.20.0
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",
}

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}
}