This is a sentence-transformers model finetuned from google/embeddinggemma-300m. 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: google/embeddinggemma-300m
Maximum Sequence Length: 2048 tokens
Output Dimensionality: 768 dimensions
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': 2048, 'do_lower_case': False, 'architecture': 'Gemma3TextModel'})
  (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})
  (2): Dense({'in_features': 768, 'out_features': 3072, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (3): Dense({'in_features': 3072, 'out_features': 768, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (4): 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("sentence_transformers_model_id")
# Run inference
queries = [
    "task: search result | query: \u0935\u0943\u0937\u094d\u091f\u093f-\u0935\u093f\u0926\u094d\u092f\u0941\u0924\u094d-\u0938\u0926\u0943\u0936\u0902 \u0926\u0948\u0935\u093f\u0915\u0902 \u0906\u0917\u092e\u0928\u092e\u094d",
]
documents = [
    'title: none | text: असामि हि प्रयज्यवः कण्वं दद प्रचेतसः\nअसामिभिर् मरुत आ न ऊतिभिर् गन्ता वृष्टिं न विद्युतः',
    'title: none | text: उत द्वार उशतीर् वि श्रयन्ताम् उत देवाṁ उशत आ वहेह',
    'title: none | text: प्राग्नये बृहते यज्ञियाय ऋतस्य वृष्णे असुराय मन्म\nघृतं न यज्ञ आस्ये सुपूतं गिरम् भरे वृषभाय प्रतीचीम्',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[0.5082, 0.0894, 0.0246]])

Evaluation

Metrics

Triplet

Dataset: test
Evaluated with TripletEvaluator

Metric	Value
cosine_accuracy	0.9553

Training Details

Training Dataset

Unnamed Dataset

Size: 51,368 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1
type string string
details
min: 13 tokens
mean: 20.48 tokens
max: 41 tokens
min: 25 tokens
mean: 48.95 tokens
max: 100 tokens

Samples:

sentence_0	sentence_1
`task: search result \| query: Passage describing multiple entities undergoing purification or being purifying`	`title: none \| text: सम् उ प्रिया अनूषत गावो मदाय घृष्वयः[object Object]सोमासः कृण्वते पथः पवमानास इन्दवः`
`task: search result \| query: Which verse describes adversaries or those who attempt to inflict injury, but are ultimately rendered ineffective?`	`title: none \| text: न यं रिपवो न रिषण्यवो गर्भे सन्तं रेषणा रेषयन्ति[object Object]अन्धा अपश्या न दभन्न् अभिख्या नित्यास ईम् प्रेतारो अरक्षन्`
`task: search result \| query: A hymn requesting blessings for both human and animal residents of a dwelling.`	`title: none \| text: वास्तोष्पते प्रति जानीह्य् अस्मान् स्वावेशो अनमीवो भवा नः[object Object]यत् त्वेमहे प्रति तन् नो जुषस्व शं नो भव द्विपदे शं चतुष्पदे`

Loss: MultipleNegativesRankingLoss with these parameters:
json
1{ 2 "scale": 20.0, 3 "similarity_fct": "cos_sim", 4 "gather_across_devices": false 5}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 32
per_device_eval_batch_size: 32
fp16: True
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
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: 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
num_train_epochs: 3
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
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: 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}
parallelism_config: None
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
project: huggingface
trackio_space_id: trackio
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
hub_revision: None
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
include_tokens_per_second: False
include_num_input_tokens_seen: no
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
eval_on_start: False
use_liger_kernel: False
liger_kernel_config: None
eval_use_gather_object: False
average_tokens_across_devices: True
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	test_cosine_accuracy
0.3113	500	2.092	0.8555
0.6227	1000	1.5127	0.8821
0.9340	1500	1.1996	0.9108
1.0	1606	-	0.9135
1.2453	2000	0.783	0.9205
1.5567	2500	0.6516	0.9266
1.8680	3000	0.5921	0.9415
2.0	3212	-	0.9439
2.1793	3500	0.3662	0.9487
2.4907	4000	0.2307	0.9497
2.8020	4500	0.2088	0.9541
3.0	4818	-	0.9553

Framework Versions

Python: 3.12.9
Sentence Transformers: 5.1.2
Transformers: 4.57.1
PyTorch: 2.7.1+cu118
Accelerate: 1.11.0
Datasets: 4.3.0
Tokenizers: 0.22.1

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