SentenceTransformer

This is a sentence-transformers model. It maps sentences & paragraphs to a dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more. In the context of MatchMiner-AI, this model is trained to embed patient summaries and clinical trial "space" (target population summaries) to enable rapid retrieval of trial options for patients and vice versa.

Model Details

Model Description

Model Type: Sentence Transformer
Maximum Sequence Length: 2500 tokens
Output Dimensionality: 1024 dimensions
Similarity Function: Cosine Similarity
Training Datasets:
- mnri_dataset
- contrastive_dataset

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': 2500, 'do_lower_case': False, 'architecture': 'Qwen3Model'})
  (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.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
queries = [
    "Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 63  \nSex: Male  \nCancer type: Diffuse large B\u2011cell lymphoma (DLBCL)  \nHistology: Activated B\u2011cell (non\u2011germinal\u2011center) phenotype, CD20\u202fpositive, MUM1\u202fpositive, BCL2\u202fpositive, Ki\u201167\u224895\u202f%  \nCurrent extent: Limited residual disease \u2013 residual metabolic activity in the original right posterior thigh soft\u2011tissue mass (partial metabolic response) and a solitary right iliac lymph node with stable low\u2011volume uptake (stable disease); no other sites of disease evident.  \n\nBiomarkers:  \n- Immunohistochemistry: CD20+, MUM1+, BCL2+, CD10\u2011, BCL6\u2011, Ki\u201167\u224895%  \n- FISH: No MYC, BCL2, or BCL6 rearrangements (triple\u2011hit excluded)  \n- NGS (original thigh lesion): MYD88 L265P (VAF\u202f\u224835\u202f%), CD79B Y196F (VAF\u202f\u224830\u202f%), CARD11 S328F (VAF\u202f\u224828\u202f%), REL copy\u2011number gain (\u22483\u20134 copies) \u2013 low tumor mutational burden (5\u202fmut/Mb), microsatellite stable.  \n- Additional alterations identified in iliac node (September\u00a02013): TP53 p.R248Q (VAF\u202f\u224822\u202f%), CDKN2A homozygous deletion.  \n\nTreatment history:  \n# Early\u202f2011\u2013mid\u202f2011: R\u2011CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone) \u2013 6 cycles, 21\u2011day schedule. Best response: partial metabolic response (Deauville\u202f4).  \n# Mid\u202f2011 (following R\u2011CHOP): High\u2011dose BEAM conditioning with autologous stem\u2011cell transplantation \u2013 achieved only persistent residual disease (no complete remission).  \n# Early\u202f2012 (exact infusion date unspecified, Day\u202f0 of CAR\u2011T): Axicabtagene\u202fciloleucel (axi\u2011cel) CAR\u2011T after fludarabine/cyclophosphamide lymphodepletion. Course complicated by Grade\u202f3 cytokine\u2011release syndrome (tocilizumab, brief vasopressors) and Grade\u202f2 immune effector cell\u2011associated neurotoxicity syndrome (resolved). Day\u202f60 restaging (July\u202f2012) showed partial metabolic response of thigh lesion (\u2193SUV\u202ffrom\u202f\u224813\u202fto\u202f4.2) but emergence of a new right iliac node (SUV\u202f\u22489.3) consistent with progressive disease.  \n# Late\u202f2012 (December\u202f2012): Palliative external\u2011beam radiation to right iliac node, 30\u202fGy in 15 fractions (2\u202fGy per fraction). Resulted in symptom relief and reduction of nodal size/metabolic activity (node size \u2193 to 2.5\u202fcm, SUV\u202f\u22485.8).  \n# 2013\u202fMarch onward: Observation/off\u2011therapy; serial imaging through 2014 demonstrates stable disease (persistent low\u2011grade uptake in thigh lesion, SUV\u202f\u22484.2; iliac node stable, SUV\u202f\u22485.8). No further systemic therapy administered.",
]
documents = [
    'Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: ≥18 years. Sex allowed: both. Cancer type allowed: diffuse large b cell lymphoma. Histology allowed: diffuse large b cell lymphoma–not otherwise specified. Cancer burden allowed: evaluable/measurable disease; absent massive uncontrolled pleural, pericardial, or peritoneal effusions, pulmonary lymphangitis, and liver involvement exceeding 50\u202f%. Prior treatment required: relapsed or refractory disease after prior systemic anti‑lymphoma therapy (including patients ineligible for car t‑cell therapy). Prior treatment excluded: any concurrent antitumor‑directed drug therapy. Biomarkers required: NA. Biomarkers excluded: NA.',
    'Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: 18-75 years. Sex allowed: Male and Female. Cancer type allowed: Any malignant solid tumor. Histology allowed: NA. Cancer burden allowed: Stage\u202fIIIB‑IV disease that has progressed after prior standard therapy and for whom immune‐monotherapy is planned, regardless of PD‑L1 result. Prior treatment required: Progression after receipt of standard therapy. Prior treatment excluded: NA. Biomarkers required: Presence of at least one positively expressed tumor neoantigen; PD‑L1 testing permitted but not mandatory. Biomarkers excluded: NA.',
    'Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: >=18. Sex allowed: female. Cancer type allowed: ovarian carcinosarcoma. Histology allowed: mixed carcinoma. Cancer burden allowed: relapsed or refractory disease after at least one prior platinum‑based regimen (platinum‑resistant <183\xa0days) with measurable disease. Prior treatment required: at least one prior line of platinum chemotherapy. Prior treatment excluded: prior MUC1‑directed therapy. Biomarkers required: MUC1 expression positive (>0 IHC) assessed at screening. Biomarkers excluded: NA',
]
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([[0.8415, 0.3722, 0.0634]])

Training Details

Training Datasets

mnri_dataset

Dataset: mnri_dataset
Size: 756,840 training samples
Columns: patient_summary_trunc and this_space_trunc
Approximate statistics based on the first 1000 samples:
patient_summary_trunc this_space_trunc
type string string
details
min: 386 tokens
mean: 911.75 tokens
max: 1827 tokens

min: 112 tokens
mean: 168.64 tokens
max: 388 tokens

	patient_summary_trunc	this_space_trunc
type	string	string
details	min: 386 tokens mean: 911.75 tokens max: 1827 tokens	min: 112 tokens mean: 168.64 tokens max: 388 tokens

Samples:

patient_summary_trunc	this_space_trunc
Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 32 Sex: Male Cancer type: B‑cell precursor acute lymphoblastic leukemia (ALL) Histology: B‑cell lineage lymphoblasts (CD19⁺, CD20⁺, CD10⁺, TdT⁺) Current extent: Systemic marrow disease (diffusely involved bone marrow, no detectable extramedullary or CNS disease) – considered advanced/systemic Biomarkers: - PAX5 p.Arg31Ser (likely pathogenic missense) – VAF ≈ 48% - NRAS p.Gln61Lys (activating missense) – VAF ≈ 35‑41% - FLT3 internal tandem duplication (30‑bp juxtamembrane) – pathogenic, allelic ratio ≈ 30‑36% - CDKN2A homozygous deletion (exons 1‑2) – pathogenic - Low‑level JAK2 V617F (p.Val617Phe) – subclonal, VAF ≈ 3% - CREBBP p.Arg1664His – variant of uncertain significance, VAF ≈ 7% - No BCR‑ABL1 fusion, no TEL‑AML1, no MLL/KMT2A rearrange...	Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: 0 through 45 years. Sex allowed: both sexes. Cancer type allowed: acute lymphoblastic leukaemia. Histology allowed: B‑cell precursor acute lymphoblastic leukaemia. Cancer burden allowed: newly diagnosed, never relapsed, de novo disease. Prior treatment required: NA. Prior treatment excluded: systemic corticosteroids ≥10 mg/m²/day prednisone equivalents for longer than one week before diagnosis; any chemotherapeutic agent administered within four weeks before diagnosis. Biomarkers required: surface immunoglobulin negative phenotype; IG::MYC rearrangement accepted only when BCL2 and BCL6 rearrangements are absent. Biomarkers excluded: KMT2A‑rearranged B‑cell precursor ALL in patients younger than 1 year; Philadelphia chromosome‑positive (t[9;22]/BCR‑ABL) ...
Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 66 years (born 1959, most recent entry dated September 2017) Sex: Female Cancer type: Pancreatic ductal adenocarcinoma (PDAC) Histology: Poorly differentiated invasive ductal adenocarcinoma Current extent: Metastatic (Stage IV) – hepatic dominant disease with two liver deposits (segment VI ≈ 1.4 cm, segment VII ≈ 1.8 cm), solitary osteoblastic metastasis to L3 vertebral body, and persistent peripancreatic soft‑tissue thickening (likely post‑surgical fibrosis vs low‑volume nodal disease). No CNS disease documented. Biomarkers (most recent specimen, September 2017 hepatic core biopsy): • KRAS G12D point mutation (VAF 18%) – activating hotspot, different from earlier KRAS G12R clone • KRAS focal amplification (estimated copy ≈ 7) • TP53 R273H missense...	Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: 18-75 years. Sex allowed: Male and Female. Cancer type allowed: Advanced solid tumor. Histology allowed: Any. Cancer burden allowed: Locally advanced or metastatic disease refractory to standard therapy or lacking effective treatment. Prior treatment required: Progressive disease after standard systemic therapy. Prior treatment excluded: Prior receipt of targeted ROR1 inhibitor therapy. Biomarkers required: ROR1 positive (assessment planned during screening). Biomarkers excluded: NA.
Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 66 Sex: Female Cancer type: Ovarian carcinosarcoma (malignant mixed Müllerian tumor) Histology: Biphasic high‑grade serous carcinoma component (WT1⁺, PAX8⁺, p53 overexpression) with undifferentiated spindle‑cell sarcomatous component (vimentin⁺, desmin‑) Current extent: Metastatic disease (peritoneal carcinomatosis with implants up to ≈2 cm, right pleural effusion, bilateral subcentimeter pulmonary nodules) – clinically stage IV, stable on current therapy Biomarkers: - TP53 p.Y220C (Likely pathogenic, somatic, VAF 68%) - KRAS p.G12D (Likely pathogenic, somatic, VAF 9%) - BRCA2 splice‑site alteration c.5946+2T>A (Likely pathogenic, somatic, VAF 12%) – HRD score negative, microsatellite stable, TMB 6 mut/Mb (low) - CCNE1 copy‑number gain (~5‑fold am...	Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: >=18. Sex allowed: female. Cancer type allowed: ovarian carcinosarcoma. Histology allowed: mixed carcinoma. Cancer burden allowed: relapsed or refractory disease after at least one prior platinum‑based regimen (platinum‑resistant <183 days) with measurable disease. Prior treatment required: at least one prior line of platinum chemotherapy. Prior treatment excluded: prior MUC1‑directed therapy. Biomarkers required: MUC1 expression positive (>0 IHC) assessed at screening. Biomarkers excluded: NA

Loss: MultipleNegativesRankingLoss with these parameters:

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

contrastive_dataset

Dataset: contrastive_dataset
Size: 1,530,582 training samples
Columns: patient_summary_trunc, this_space_trunc, and label

Approximate statistics based on the first 1000 samples:

	patient_summary_trunc	this_space_trunc	label
type	string	string	float
details	min: 202 tokens mean: 906.99 tokens max: 1827 tokens	min: 121 tokens mean: 185.89 tokens max: 388 tokens	min: 0.0 mean: 0.39 max: 1.0

Samples:

patient_summary_trunc	this_space_trunc	label
Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 32 Sex: Male Cancer type: B‑cell precursor acute lymphoblastic leukemia (ALL) Histology: B‑cell lineage lymphoblasts (CD19⁺, CD20⁺, CD10⁺, TdT⁺) Current extent: Systemic marrow disease (diffusely involved bone marrow, no detectable extramedullary or CNS disease) – considered advanced/systemic Biomarkers: - PAX5 p.Arg31Ser (likely pathogenic missense) – VAF ≈ 48% - NRAS p.Gln61Lys (activating missense) – VAF ≈ 35‑41% - FLT3 internal tandem duplication (30‑bp juxtamembrane) – pathogenic, allelic ratio ≈ 30‑36% - CDKN2A homozygous deletion (exons 1‑2) – pathogenic - Low‑level JAK2 V617F (p.Val617Phe) – subclonal, VAF ≈ 3% - CREBBP p.Arg1664His – variant of uncertain significance, VAF ≈ 7% - No BCR‑ABL1 fusion, no TEL‑AML1, no MLL/KMT2A rearrange...	Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: 0 through 45 years. Sex allowed: both sexes. Cancer type allowed: acute lymphoblastic leukaemia. Histology allowed: B‑cell precursor acute lymphoblastic leukaemia. Cancer burden allowed: newly diagnosed, never relapsed, de novo disease. Prior treatment required: NA. Prior treatment excluded: systemic corticosteroids ≥10 mg/m²/day prednisone equivalents for longer than one week before diagnosis; any chemotherapeutic agent administered within four weeks before diagnosis. Biomarkers required: surface immunoglobulin negative phenotype; IG::MYC rearrangement accepted only when BCL2 and BCL6 rearrangements are absent. Biomarkers excluded: KMT2A‑rearranged B‑cell precursor ALL in patients younger than 1 year; Philadelphia chromosome‑positive (t[9;22]/BCR‑ABL) ...	`1.0`
Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 66 years (born 1959, most recent entry dated September 2017) Sex: Female Cancer type: Pancreatic ductal adenocarcinoma (PDAC) Histology: Poorly differentiated invasive ductal adenocarcinoma Current extent: Metastatic (Stage IV) – hepatic dominant disease with two liver deposits (segment VI ≈ 1.4 cm, segment VII ≈ 1.8 cm), solitary osteoblastic metastasis to L3 vertebral body, and persistent peripancreatic soft‑tissue thickening (likely post‑surgical fibrosis vs low‑volume nodal disease). No CNS disease documented. Biomarkers (most recent specimen, September 2017 hepatic core biopsy): • KRAS G12D point mutation (VAF 18%) – activating hotspot, different from earlier KRAS G12R clone • KRAS focal amplification (estimated copy ≈ 7) • TP53 R273H missense...	Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: 18-75 years. Sex allowed: Male and Female. Cancer type allowed: Advanced solid tumor. Histology allowed: Any. Cancer burden allowed: Locally advanced or metastatic disease refractory to standard therapy or lacking effective treatment. Prior treatment required: Progressive disease after standard systemic therapy. Prior treatment excluded: Prior receipt of targeted ROR1 inhibitor therapy. Biomarkers required: ROR1 positive (assessment planned during screening). Biomarkers excluded: NA.	`1.0`
Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age: 18 Sex: Female Cancer type: B‑myeloid mixed‑phenotype acute leukemia (MPAL) Histology: Biphenotypic blasts expressing B‑lineage (CD19+) and myeloid (CD33+, MPO+) markers; Philadelphia‑positive (BCR‑ABL1 p190) Current extent: Relapsed/refractory disease with low‑burden molecular persistence (BCR‑ABL1 transcript ~0.02 % IS); no morphologic blasts in marrow, no extramedullary disease evident on imaging Biomarkers: - BCR‑ABL1 p190 fusion (RNA‑based detection, pathogenic) – present throughout disease course - IKZF1 deletion, KRAS Q61H, PTPN11 N308S, MYC copy‑number gain – identified on initial NGS (2018‑12) and persistently reported - MRD quantitative PCR: 0.05 % (day 30 post‑transplant, 2019‑03‑08); 0.02 % (Sept 2019); 0.03 % (Mar 2020); 0.022 % (June ...	Instruct: Given a cancer patient summary, retrieve clinical trial options that are reasonable for that patient; or, given a clinical trial option, retrieve cancer patients who are reasonable candidates for that trial. Age range allowed: 1 Year to 39 Years. Sex allowed: Male and Female. Cancer type allowed: Mixed phenotype acute leukemia, B‑myeloid. Histology allowed: B‑myeloid mixed phenotype acute leukemia. Cancer burden allowed: Relapsed or refractory disease. Prior treatment required: Full recovery from prior hematopoietic stem cell transplantation or anthracycline exposure. Prior treatment excluded: Current administration of anticancer agents (except intrathecal agents or hydroxyurea). Biomarkers required: NA. Biomarkers excluded: KMT2A rearrangement, Philadelphia chromosome/BCR‑ABL1 fusion.	`0.0`

Loss: OnlineContrastiveLoss

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 18
learning_rate: 2e-05
warmup_ratio: 0.01
bf16: True

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 18
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: 2e-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: 3
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.01
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: True
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: 1
ddp_backend: None
tpu_num_cores: None
tpu_metrics_debug: False
debug: []
dataloader_drop_last: True
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
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: proportional
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Click to expand

Epoch	Step	Training Loss
0.0063	100	0.6184
0.0126	200	0.6423
0.0189	300	0.597
0.0252	400	0.582
0.0315	500	0.6083
0.0378	600	0.6311
0.0441	700	0.644
0.0504	800	0.6166
0.0567	900	0.597
0.0630	1000	0.663
0.0693	1100	0.6215
0.0755	1200	0.6366
0.0818	1300	0.6275
0.0881	1400	0.6278
0.0944	1500	0.6346
0.1007	1600	0.6317
0.1070	1700	0.6328
0.1133	1800	0.6139
0.1196	1900	0.656
0.1259	2000	0.6238
0.1322	2100	0.6208
0.1385	2200	0.6551
0.1448	2300	0.6257
0.1511	2400	0.6162
0.1574	2500	0.6386
0.1637	2600	0.65
0.1700	2700	0.6196
0.1763	2800	0.6263
0.1826	2900	0.6327
0.1889	3000	0.6268
0.1952	3100	0.6456
0.2015	3200	0.613
0.2078	3300	0.6455
0.2141	3400	0.602
0.2203	3500	0.6435
0.2266	3600	0.6329
0.2329	3700	0.625
0.2392	3800	0.6222
0.2455	3900	0.6177
0.2518	4000	0.6229
0.2581	4100	0.6502
0.2644	4200	0.6275
0.2707	4300	0.6424
0.2770	4400	0.6112
0.2833	4500	0.6298
0.2896	4600	0.6337
0.2959	4700	0.6148
0.3022	4800	0.625
0.3085	4900	0.6295
0.3148	5000	0.6326
0.3211	5100	0.6522
0.3274	5200	0.618
0.3337	5300	0.6631
0.3400	5400	0.6295
0.3463	5500	0.6388
0.3526	5600	0.6279
0.3589	5700	0.649
0.3651	5800	0.6289
0.3714	5900	0.638
0.3777	6000	0.6162
0.3840	6100	0.6324
0.3903	6200	0.6164
0.3966	6300	0.6086
0.4029	6400	0.613
0.4092	6500	0.6214
0.4155	6600	0.6106
0.4218	6700	0.6312
0.4281	6800	0.588
0.4344	6900	0.6144
0.4407	7000	0.6084
0.4470	7100	0.602
0.4533	7200	0.6224
0.4596	7300	0.6029
0.4659	7400	0.5989
0.4722	7500	0.6427
0.4785	7600	0.6246
0.4848	7700	0.5989
0.4911	7800	0.5956
0.4974	7900	0.6298
0.5037	8000	0.6092
0.5099	8100	0.6203
0.5162	8200	0.6239
0.5225	8300	0.6211
0.5288	8400	0.6102
0.5351	8500	0.5934
0.5414	8600	0.6367
0.5477	8700	0.6424
0.5540	8800	0.6014
0.5603	8900	0.6181
0.5666	9000	0.6012
0.5729	9100	0.6292
0.5792	9200	0.6316
0.5855	9300	0.6294
0.5918	9400	0.6294
0.5981	9500	0.6207
0.6044	9600	0.6119
0.6107	9700	0.5915
0.6170	9800	0.6153
0.6233	9900	0.626
0.6296	10000	0.6628
0.6359	10100	0.6043
0.6422	10200	0.5846
0.6485	10300	0.6261
0.6547	10400	0.616
0.6610	10500	0.622
0.6673	10600	0.5993
0.6736	10700	0.6219
0.6799	10800	0.6078
0.6862	10900	0.6102
0.6925	11000	0.6266
0.6988	11100	0.6058
0.7051	11200	0.617
0.7114	11300	0.6091
0.7177	11400	0.5837
0.7240	11500	0.6037
0.7303	11600	0.5996
0.7366	11700	0.6304
0.7429	11800	0.6428
0.7492	11900	0.6289
0.7555	12000	0.617
0.7618	12100	0.6209
0.7681	12200	0.6064
0.7744	12300	0.6168
0.7807	12400	0.6027
0.7870	12500	0.5924
0.7933	12600	0.624
0.7995	12700	0.6058
0.8058	12800	0.5708
0.8121	12900	0.6067
0.8184	13000	0.6194
0.8247	13100	0.6379
0.8310	13200	0.5997
0.8373	13300	0.5787
0.8436	13400	0.6229
0.8499	13500	0.621
0.8562	13600	0.6089
0.8625	13700	0.5786
0.8688	13800	0.5904
0.8751	13900	0.6216
0.8814	14000	0.5873
0.8877	14100	0.5898
0.8940	14200	0.5901
0.9003	14300	0.6105
0.9066	14400	0.5703
0.9129	14500	0.6118
0.9192	14600	0.5709
0.9255	14700	0.5768
0.9318	14800	0.5901
0.9381	14900	0.5881
0.9443	15000	0.6011
0.9506	15100	0.6083
0.9569	15200	0.6213
0.9632	15300	0.5922
0.9695	15400	0.618
0.9758	15500	0.6276
0.9821	15600	0.5851
0.9884	15700	0.589
0.9947	15800	0.5988
1.0010	15900	0.6132
1.0073	16000	0.5756
1.0136	16100	0.6208
1.0199	16200	0.5767
1.0262	16300	0.5605
1.0325	16400	0.5581
1.0388	16500	0.5591
1.0451	16600	0.5734
1.0514	16700	0.5361
1.0577	16800	0.5334
1.0640	16900	0.5841
1.0703	17000	0.5455
1.0766	17100	0.5494
1.0829	17200	0.5511
1.0891	17300	0.5654
1.0954	17400	0.5305
1.1017	17500	0.5502
1.1080	17600	0.5496
1.1143	17700	0.526
1.1206	17800	0.5682
1.1269	17900	0.5422
1.1332	18000	0.5551
1.1395	18100	0.5577
1.1458	18200	0.5257
1.1521	18300	0.5467
1.1584	18400	0.5525
1.1647	18500	0.554
1.1710	18600	0.5185
1.1773	18700	0.5414
1.1836	18800	0.5539
1.1899	18900	0.5262
1.1962	19000	0.5687
1.2025	19100	0.5165
1.2088	19200	0.5575
1.2151	19300	0.516
1.2214	19400	0.5466
1.2277	19500	0.5321
1.2339	19600	0.5363
1.2402	19700	0.5487
1.2465	19800	0.5188
1.2528	19900	0.541
1.2591	20000	0.5655
1.2654	20100	0.5349
1.2717	20200	0.5451
1.2780	20300	0.5095
1.2843	20400	0.5542
1.2906	20500	0.5448
1.2969	20600	0.5097
1.3032	20700	0.5333
1.3095	20800	0.5425
1.3158	20900	0.5459
1.3221	21000	0.5527
1.3284	21100	0.5284
1.3347	21200	0.5664
1.3410	21300	0.544
1.3473	21400	0.5398
1.3536	21500	0.5399
1.3599	21600	0.5351
1.3662	21700	0.5463
1.3725	21800	0.5413
1.3787	21900	0.5372
1.3850	22000	0.5283
1.3913	22100	0.5219
1.3976	22200	0.5232
1.4039	22300	0.5185
1.4102	22400	0.5418
1.4165	22500	0.5167
1.4228	22600	0.5306
1.4291	22700	0.4993
1.4354	22800	0.5316
1.4417	22900	0.5042
1.4480	23000	0.501
1.4543	23100	0.5345
1.4606	23200	0.5263
1.4669	23300	0.5057
1.4732	23400	0.5392
1.4795	23500	0.5352
1.4858	23600	0.5091
1.4921	23700	0.5028
1.4984	23800	0.5318
1.5047	23900	0.5285
1.5110	24000	0.5247
1.5173	24100	0.5375
1.5235	24200	0.5253
1.5298	24300	0.5191
1.5361	24400	0.5107
1.5424	24500	0.5553
1.5487	24600	0.5345
1.5550	24700	0.5059
1.5613	24800	0.5165
1.5676	24900	0.5209
1.5739	25000	0.5393
1.5802	25100	0.5339
1.5865	25200	0.5448
1.5928	25300	0.5236
1.5991	25400	0.5329
1.6054	25500	0.5056
1.6117	25600	0.5136
1.6180	25700	0.5265
1.6243	25800	0.541
1.6306	25900	0.5556
1.6369	26000	0.508
1.6432	26100	0.5146
1.6495	26200	0.5323
1.6558	26300	0.5395
1.6620	26400	0.5185
1.6683	26500	0.5197
1.6746	26600	0.5137
1.6809	26700	0.5265
1.6872	26800	0.5256
1.6935	26900	0.5253
1.6998	27000	0.5158
1.7061	27100	0.5293
1.7124	27200	0.4998
1.7187	27300	0.5057
1.7250	27400	0.5114
1.7313	27500	0.5037
1.7376	27600	0.5432
1.7439	27700	0.5529
1.7502	27800	0.5301
1.7565	27900	0.5329
1.7628	28000	0.524
1.7691	28100	0.5289
1.7754	28200	0.5178
1.7817	28300	0.5089
1.7880	28400	0.5116
1.7943	28500	0.5219
1.8006	28600	0.5241
1.8068	28700	0.4835
1.8131	28800	0.5066
1.8194	28900	0.5424
1.8257	29000	0.5409
1.8320	29100	0.4875
1.8383	29200	0.5012
1.8446	29300	0.5239
1.8509	29400	0.5553
1.8572	29500	0.5021
1.8635	29600	0.4949
1.8698	29700	0.4939
1.8761	29800	0.5311
1.8824	29900	0.5083
1.8887	30000	0.4924
1.8950	30100	0.5095
1.9013	30200	0.5125
1.9076	30300	0.4866
1.9139	30400	0.511
1.9202	30500	0.4887
1.9265	30600	0.4958
1.9328	30700	0.5005
1.9391	30800	0.4888
1.9454	30900	0.5259
1.9516	31000	0.5166
1.9579	31100	0.5219
1.9642	31200	0.5173
1.9705	31300	0.5248
1.9768	31400	0.5387
1.9831	31500	0.4897
1.9894	31600	0.5028
1.9957	31700	0.5272
2.0020	31800	0.5201
2.0083	31900	0.479
2.0146	32000	0.5329
2.0209	32100	0.4709
2.0272	32200	0.4628
2.0335	32300	0.4479
2.0398	32400	0.4616
2.0461	32500	0.4806
2.0524	32600	0.434
2.0587	32700	0.4469
2.0650	32800	0.483
2.0713	32900	0.4439
2.0776	33000	0.4457
2.0839	33100	0.4516
2.0902	33200	0.464
2.0964	33300	0.4403
2.1027	33400	0.4579
2.1090	33500	0.4603
2.1153	33600	0.4454
2.1216	33700	0.4727
2.1279	33800	0.4453
2.1342	33900	0.4698
2.1405	34000	0.456
2.1468	34100	0.4414
2.1531	34200	0.4608
2.1594	34300	0.473
2.1657	34400	0.4579
2.1720	34500	0.4574
2.1783	34600	0.4483
2.1846	34700	0.4641
2.1909	34800	0.4397
2.1972	34900	0.47
2.2035	35000	0.4493
2.2098	35100	0.4673
2.2161	35200	0.4298
2.2224	35300	0.4629
2.2287	35400	0.4571
2.2350	35500	0.4531
2.2412	35600	0.4595
2.2475	35700	0.4294
2.2538	35800	0.4765
2.2601	35900	0.4744
2.2664	36000	0.4651
2.2727	36100	0.451
2.2790	36200	0.4391
2.2853	36300	0.461
2.2916	36400	0.4595
2.2979	36500	0.4329
2.3042	36600	0.4547
2.3105	36700	0.4731
2.3168	36800	0.4666
2.3231	36900	0.4581
2.3294	37000	0.4633
2.3357	37100	0.4755
2.3420	37200	0.4712
2.3483	37300	0.4478
2.3546	37400	0.4519
2.3609	37500	0.4486
2.3672	37600	0.4795
2.3735	37700	0.4525
2.3798	37800	0.4664
2.3860	37900	0.4523
2.3923	38000	0.4481
2.3986	38100	0.4487
2.4049	38200	0.4496
2.4112	38300	0.4609
2.4175	38400	0.4344
2.4238	38500	0.4464
2.4301	38600	0.4308
2.4364	38700	0.4461
2.4427	38800	0.4323
2.4490	38900	0.4279
2.4553	39000	0.4662
2.4616	39100	0.4537
2.4679	39200	0.4344
2.4742	39300	0.4732
2.4805	39400	0.4616
2.4868	39500	0.4439
2.4931	39600	0.4331
2.4994	39700	0.4486
2.5057	39800	0.4696
2.5120	39900	0.4562
2.5183	40000	0.4669
2.5246	40100	0.4554
2.5308	40200	0.4449
2.5371	40300	0.447
2.5434	40400	0.4963
2.5497	40500	0.4613
2.5560	40600	0.4411
2.5623	40700	0.4549
2.5686	40800	0.4557
2.5749	40900	0.4752
2.5812	41000	0.4617
2.5875	41100	0.4686
2.5938	41200	0.4548
2.6001	41300	0.4683
2.6064	41400	0.4393
2.6127	41500	0.4574
2.6190	41600	0.4602
2.6253	41700	0.4835
2.6316	41800	0.4856
2.6379	41900	0.4442
2.6442	42000	0.4581
2.6505	42100	0.4754
2.6568	42200	0.4749
2.6631	42300	0.4529
2.6694	42400	0.4524
2.6756	42500	0.45
2.6819	42600	0.4567
2.6882	42700	0.4586
2.6945	42800	0.4562
2.7008	42900	0.4526
2.7071	43000	0.484
2.7134	43100	0.4359
2.7197	43200	0.4642
2.7260	43300	0.4446
2.7323	43400	0.4356
2.7386	43500	0.4909
2.7449	43600	0.4916
2.7512	43700	0.4761
2.7575	43800	0.4666
2.7638	43900	0.4763
2.7701	44000	0.4745
2.7764	44100	0.4637
2.7827	44200	0.4437
2.7890	44300	0.4724
2.7953	44400	0.4692
2.8016	44500	0.4569
2.8079	44600	0.4533
2.8142	44700	0.4621
2.8204	44800	0.4744
2.8267	44900	0.4951
2.8330	45000	0.4443
2.8393	45100	0.4514
2.8456	45200	0.4848
2.8519	45300	0.4941
2.8582	45400	0.4596
2.8645	45500	0.4481
2.8708	45600	0.454
2.8771	45700	0.4732
2.8834	45800	0.4711
2.8897	45900	0.4582
2.8960	46000	0.4792
2.9023	46100	0.4634
2.9086	46200	0.441
2.9149	46300	0.4675
2.9212	46400	0.4513
2.9275	46500	0.4505
2.9338	46600	0.4592
2.9401	46700	0.4657
2.9464	46800	0.4891
2.9527	46900	0.4693
2.9590	47000	0.4814
2.9652	47100	0.4828
2.9715	47200	0.4899
2.9778	47300	0.4939
2.9841	47400	0.4507
2.9904	47500	0.4679
2.9967	47600	0.49

Framework Versions

Python: 3.12.12
Sentence Transformers: 5.1.2
Transformers: 4.57.1
PyTorch: 2.9.0+cu128
Accelerate: 1.12.0
Datasets: 4.4.1
Tokenizers: 0.22.1

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}
}