SentenceTransformer based on Qwen/Qwen3-0.6B

This is a sentence-transformers model finetuned from Qwen/Qwen3-0.6B. It maps sentences & paragraphs to a 1024-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: Qwen/Qwen3-0.6B
Maximum Sequence Length: 512 tokens
Output Dimensionality: 1024 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': 512, 'do_lower_case': False}) with Transformer model: Qwen3Model 
  (1): Pooling({'word_embedding_dimension': 1024, '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.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
    'what county is neptune city nj',
    'Neptune City, NJ. Neptune City is a borough in Monmouth County, New Jersey, United States. As of the 2010 United States Census, the borough population was 4,869. The Borough of Neptune City was incorporated on October 4, 1881, based on a referendum held on March 19, 1881.',
    'Neptune City, NJ. Sponsored Topics. Neptune City is a borough in Monmouth County, New Jersey, United States. As of the 2010 United States Census, the borough population was 4,869. The Borough of Neptune City was incorporated on October 4, 1881, based on a referendum held on March 19, 1881.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

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

Training Details

Training Dataset

Unnamed Dataset

Size: 17,048,032 training samples
Columns: sentence_0, sentence_1, and sentence_2

Approximate statistics based on the first 1000 samples:

	sentence_0	sentence_1	sentence_2
type	string	string	string
details	min: 2 tokens mean: 7.16 tokens max: 33 tokens	min: 25 tokens mean: 84.34 tokens max: 243 tokens	min: 16 tokens mean: 81.56 tokens max: 300 tokens

Samples:

sentence_0	sentence_1	sentence_2
`what county is nettles island fl`	`Nettles Island. Nettles Island in Hutchinson Island Florida. A development of close to 1300 lots with anything from trailer pads to updated concrete block homes on a mostly man made island that juts out into the Indian River on Hutchinson Island in Saint Lucie County FL. Though, the official address for Nettles Island is in Jensen Beach.`	`Fleming Island is an unincorporated community and census-designated place in Clay County, Florida, United States. It is located 21 miles southwest of downtown Jacksonville, on the western side of the St. Johns River, off US 17. As of the 2010 census the Fleming Island CDP had a population of 27,126. Fleming Island's ZIP code became 32003 in 2004, giving it a different code from Orange Park, the incorporated town to the north.`
`what time of day to take estrogen`	`Time of day to take Estrogen. Hi. I think we all may find different times of day are better for each of our needs. I actually feel much better using my estrogen twice a day. I use half in the morning and half in the evening. I am using a different estrogen than you and am able to split my dose. I'm glad to hear that you have been feeling very good on your current hormone therapy Hopefully just a small adjustment may be needed as our estrogen needs can change overtime.`	`Eating fresh carrots or drinking a cup of fresh carrot juice 2-3 times a day is a wonderful way to bring on your period sooner than expected. Carrots contain high amounts of carotene, which encourages the production of estrogen. The more estrogen you have in your body, the more your period desires to arrive.`
`what effects does nicotine have on your body`	`Nicotine also activates areas of the brain that are involved in producing feelings of pleasure and reward. Recently, scientists discovered that nicotine raises the levels of a neurotransmitter called dopamine in the parts of the brain that produce feelings of pleasure and reward.`	The action of nicotine in the body is very complicated. It is a mild stimulant which has an effect upon the heart and brain. It stimulates the central nervous system causing irregular heartbeat and blood pressure, induces vomiting and diarrhea, and first stimulates, then inhibits glandular secretions.icotine seems to provide both a stimulant and a depressant effect, and it is likely that the effect it has at any time is determined by the mood of the user, the environment and the circumstances of use.

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

Training Hyperparameters

Non-Default Hyperparameters

num_train_epochs: 1
max_steps: 10000
fp16: True
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 8
per_device_eval_batch_size: 8
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: 1
max_steps: 10000
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
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: 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}
tp_size: 0
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: None
hub_always_push: False
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: 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: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss
0.0002	500	1.7342
0.0005	1000	1.7194
0.0007	1500	1.6713
0.0009	2000	1.5885
0.0012	2500	1.4152
0.0014	3000	1.3052
0.0016	3500	1.1763
0.0019	4000	1.0714
0.0021	4500	1.0235
0.0023	5000	0.9484
0.0026	5500	0.9207
0.0028	6000	0.9076
0.0031	6500	0.8736
0.0033	7000	0.8671
0.0035	7500	0.8621
0.0038	8000	0.8414
0.0040	8500	0.8228
0.0042	9000	0.8101
0.0045	9500	0.8339
0.0047	10000	0.7968

Framework Versions

Python: 3.10.14
Sentence Transformers: 4.0.1
Transformers: 4.51.3
PyTorch: 2.6.0+cu124
Accelerate: 1.6.0
Datasets: 3.2.0
Tokenizers: 0.21.1

Citation

BibTeX

Sentence Transformers

@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

@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}
}

chungimungi
/

Qwen3-0.6B-ms-marco