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: 768 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': 768, 'do_lower_case': False, 'architecture': 'PeftModelForFeatureExtraction'})
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': False, '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': True, '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("sentence_transformers_model_id")
# Run inference
queries = [
    "What actions did George M. Harris commit to regarding the crash cart inspection documentation discrepancy identified by Patricia A. Johnson in October 2025?",
]
documents = [
    'George M. Harris agreed to review the crash cart inspection logs for missing or incomplete entries, assess for process gaps or staff confusion, and coordinate with the training department for a refresher if needed, with an update to Patricia by the end of the week.',
    'George M. Harris stated he would review the updated Dietary Services menu, cross-reference it with current billing codes and payer documentation requirements, and provide recommendations for coding updates and additional documentation to ensure compliance and prevent reimbursement issues.',
    'Dr. Collins agreed to assemble a cross-departmental task force within the week to review the credentialing workflow and explore automation or process improvements, aiming to address the delays and maintain accreditation timelines.',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 1024] [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[0.7461, 0.3457, 0.0243]], dtype=torch.bfloat16)

Evaluation

Metrics

Information Retrieval

Dataset: val_evaluation
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.7978
cosine_accuracy@3	0.9233
cosine_accuracy@5	0.9511
cosine_accuracy@10	0.9844
cosine_precision@1	0.7978
cosine_precision@3	0.3078
cosine_precision@5	0.1902
cosine_precision@10	0.0984
cosine_recall@1	0.7978
cosine_recall@3	0.9233
cosine_recall@5	0.9511
cosine_recall@10	0.9844
cosine_ndcg@10	0.8951
cosine_mrr@10	0.866
cosine_map@100	0.8669

Training Details

Training Dataset

Unnamed Dataset

Size: 2,100 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: 18 tokens
mean: 32.19 tokens
max: 55 tokens
min: 13 tokens
mean: 39.75 tokens
max: 110 tokens

Samples:

sentence_0	sentence_1
`What recent improvements did Paul R. Nelson mention regarding the hospital's safety management in the October 2025 email thread about the 安全建议箱反馈公告?`	`Paul R. Nelson stated that the hospital has optimized the access control system, visitor procedures, and night patrols, and these improvements have already shown initial results.`
`What days did Ethan L. Foster suggest for meeting Nicole R. Barnes for coffee during the week of September 5, 2025?`	`Ethan L. Foster suggested Thursday and Friday afternoon as possible times to meet Nicole R. Barnes for coffee during the week of September 5, 2025.`
`What actions did Xavier D. Brooks commit to regarding the increased turnaround time for surgical site infection tracking as discussed on November 10, 2025?`	`Xavier D. Brooks stated he would coordinate with nursing and records teams to review recent workflow changes and identify bottlenecks, and promised to share findings and propose improvements by the end of the week.`

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

per_device_train_batch_size: 64
per_device_eval_batch_size: 64
num_train_epochs: 10
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 64
per_device_eval_batch_size: 64
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: 10
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: None
warmup_ratio: None
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
enable_jit_checkpoint: False
save_on_each_node: False
save_only_model: False
restore_callback_states_from_checkpoint: False
use_cpu: False
seed: 42
data_seed: None
bf16: False
fp16: False
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: -1
ddp_backend: None
debug: []
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: False
ignore_data_skip: False
fsdp: []
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
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
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
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_for_metrics: []
eval_do_concat_batches: True
auto_find_batch_size: False
full_determinism: False
ddp_timeout: 1800
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
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
use_cache: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	val_evaluation_cosine_ndcg@10
1.0	33	0.8290
2.0	66	0.8660
3.0	99	0.8781
4.0	132	0.8848
5.0	165	0.8905
6.0	198	0.8939
7.0	231	0.8956
8.0	264	0.8959
9.0	297	0.8969
10.0	330	0.8951

Framework Versions

Python: 3.12.12
Sentence Transformers: 5.2.2
Transformers: 5.0.0
PyTorch: 2.10.0+cu128
Accelerate: 1.12.0
Datasets: 4.0.0
Tokenizers: 0.22.2

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