SentenceTransformer based on Qwen/Qwen3-Embedding-0.6B

This is a sentence-transformers model finetuned from Qwen/Qwen3-Embedding-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-Embedding-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, 'architecture': 'Qwen3Model'})
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, '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 = [
    'to general appeal . 2 the inpatient in determination determination is subject to the appeal',
    'Right to pursue the general claims appeal process. (2) If the beneficiary is no longer an inpatient in the hospital and is dissatisfied with this determination, the determination is subject to the general claims appeal process.',
    "Column (1) Part A Cases- Use Column 1 to record information on reconsiderations of redeterminations for Part A services processed by the A/B MAC (A) Column (2) Part B of A Cases Use Column 2 to record information on reconsiderations of redeterminations for Part B services processed by the A/B MAC (A) Column (3) Part B Cases- Use Column 3 to record information on reconsiderations of redeterminations for Part B services processed by the A/B MAC (B) or DME MAC Opening Pending - Show the number of closing pending reconsiderations reported on Line 47 on the previous month's report Adjustments to Pending - If it is necessary to revise the pending figure for the close of the previous month because of inventories or reporting errors, enter the adjustment",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.7645, 0.6362],
#         [0.7645, 1.0000, 0.4567],
#         [0.6362, 0.4567, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

Size: 60,936 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: 9 tokens
mean: 59.85 tokens
max: 285 tokens

min: 38 tokens
mean: 148.86 tokens
max: 512 tokens

	sentence_0	sentence_1
type	string	string
details	min: 9 tokens mean: 59.85 tokens max: 285 tokens	min: 38 tokens mean: 148.86 tokens max: 512 tokens

Samples:

sentence_0	sentence_1
`B 60 percent for calendar CY C quarters CY) in (90 percent calendar quarters in F percent, for quarters CY and all subsequent calendar.`	`(B) 60 percent, for calendar quarters in CY 2015 (C) 70 percent, for calendar quarters in CY 2016 (D) 80 percent, for calendar quarters in CY 2017 (E) 90 percent, for calendar quarters in CY 2018 and (F) 100 percent, for calendar quarters in CY 2019 and all subsequent calendar years.`
`To verify's, A/B (A submit a status query prospective are from the and procedure where utilization affected A hospital is 60 days of A/B MAC notice (advice) for or was resulting weighted reported QIO limit adjusted (A hospital`	To verify CMS's acceptance, the A/B MAC (A) can submit a status query Under inpatient hospital prospective payment, adjustment requests are required from the hospital where errors occur in diagnosis and procedure coding that changes the DRG, or where the deductible or utilization is affected A hospital is allowed 60 days from the date of the A/B MAC (A) payment notice (remittance advice) for adjustment requests where diagnostic or procedure coding was in error resulting in a change to a higher weighted DRG Adjustments reported by the QIO have no corresponding time limit and are adjusted automatically by the A/B MAC (A) without requiring the hospital to submit an adjustment request
`blend alternative diem portion determined in) and portion in (6,671.84 + $\ *, was in the denominator the 5/6th of (28.0 25 days`	`Compute the blend alternative Add the LTC-DRG per diem portion determined in step (1-c) and the IPPS comparable per diem portion determined in step (2-d) $6,671.84 + $4,491.10 $11,162.94 * In this example, 25 days was used in the denominator since the 5/6th ALOS of LTC DRG XYZ (28.0 days) is greater than 25 days`

Loss: DenoisingAutoEncoderLoss

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 24
per_device_eval_batch_size: 24
num_train_epochs: 2
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: 24
per_device_eval_batch_size: 24
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: 2
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
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: 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_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: 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: False
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: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss
0.1969	500	2.3686
0.3939	1000	2.1397
0.5908	1500	2.0367
0.7877	2000	1.9555
0.9846	2500	1.9082
1.1816	3000	1.6029
1.3785	3500	1.5672
1.5754	4000	1.5639
1.7724	4500	1.5663
1.9693	5000	1.5593

Framework Versions

Python: 3.12.6
Sentence Transformers: 5.2.0
Transformers: 4.56.0
PyTorch: 2.8.0+cu129
Accelerate: 1.10.1
Datasets: 4.4.1
Tokenizers: 0.22.0

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",
}

DenoisingAutoEncoderLoss

@inproceedings{wang-2021-TSDAE,
    title = "TSDAE: Using Transformer-based Sequential Denoising Auto-Encoderfor Unsupervised Sentence Embedding Learning",
    author = "Wang, Kexin and Reimers, Nils and Gurevych, Iryna",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2021",
    month = nov,
    year = "2021",
    address = "Punta Cana, Dominican Republic",
    publisher = "Association for Computational Linguistics",
    pages = "671--688",
    url = "https://arxiv.org/abs/2104.06979",
}