This is a sentence-transformers model finetuned from Snowflake/snowflake-arctic-embed-l. 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: Snowflake/snowflake-arctic-embed-l
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: BertModel 
  (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})
  (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("christinemahler/aie5-midter-new")
# Run inference
sentences = [
    'What is the purpose of the funding opportunity RFA-DK-26-007 titled "Collaborative Research Using Biosamples and/or Data from Type 1 Diabetes Clinical Studies"?',
    'RFA-DK-26-007: Collaborative Research Using Biosamples and/or Data from Type 1 Diabetes Clinical Studies (R01 - Clinical Trial Not Allowed) Part 1. Overview Information\n    \n \n \n   Participating Organization(s)\n    \n    National Institutes of Health (\n     \n     NIH\n     \n    )\n    \n \n \n   Components of Participating Organizations\n    \n    National Institute of Diabetes and Digestive and Kidney Diseases (\n     \n     NIDDK\n     \n    )\n     \n \n     Office of The Director, National Institutes of Health (\n      \n      OD\n      \n     )\n     \n \n \n   Funding Opportunity Title\n    \n    Collaborative Research Using Biosamples and/or Data from Type 1 Diabetes Clinical Studies (R01 - Clinical Trial Not Allowed)\n    \n \n \n   Activity Code\n    \n \n     R01\n     \n    Research Project Grant\n    \n \n \n   Announcement Type\n   Reissue of\n    \n    RFA-DK-22-021\n    \n \n \n \n   Related Notices\n    \n \n \n      April 4, 2024\n      \n     - Overview of Grant Application and Review Changes for Due Dates on or after January 25, 2025. See Notice\n      \n      NOT-OD-24-084\n      \n     .\n     \n \n \n      August 31, 2022\n      \n     - Implementation Changes for Genomic Data Sharing Plans Included with Applications Due on or after January 25, 2023. See Notice\n      \n      NOT-OD-22-198\n      \n     .\n     \n \n \n      August 5, 2022\n      \n     - Implementation Details for the NIH Data Management and Sharing Policy. See Notice\n      \n      NOT-OD-22-189\n      \n     .\n     \n \n \n \n   Funding Opportunity Number (FON)\n    \n    RFA-DK-26-007\n    \n \n \n   Companion Funding Opportunity\n   None\n    \n \n   Number of Applications\n    \n    See\n     \n     Section III. 3. Additional Information on Eligibility\n     \n    .\n    \n \n \n   Assistance Listing Number(s)\n   93.847\n    \n \n   Funding Opportunity Purpose\n    \n    This Notice of Funding Opportunity (NOFO) invites applications for studies of type 1 diabetes etiology and pathogenesis using data and samples from clinical trials and studies. This opportunity is intended to fund investigative teams collaborating to answer important questions about disease mechanisms leading to improved delay and durable prevention of type 1 diabetes. This NOFO is associated with the Special Diabetes Program (\n     \n     https://www.niddk.nih.gov/about-niddk/research-areas/diabetes/type-1-diabetes-special-statutory-funding-program/about-special-diabetes-program\n     \n    ) which funds research on the prevention, treatment, and cure of type 1 diabetes and its complications, including unique, innovative, and collaborative research consortia and clinical trials networks.\n    \n \n \n \n \n \n \n   Funding Opportunity Goal(s)\n    \n    To promote extramural basic and clinical biomedical research that improves the understanding of the mechanisms underlying disease and leads to improved preventions, diagnosis, and treatment of diabetes, digestive, and kidney diseases. Programmatic areas within the National Institute of Diabetes and Digestive and Kidney Diseases include diabetes, digestive, endocrine, hematologic, liver, metabolic, nephrologic, nutrition, obesity, and urologic diseases.\n    \n This variable defines that we need to start a new row.',
    'RFA-DK-26-007: Collaborative Research Using Biosamples and/or Data from Type 1 Diabetes Clinical Studies (R01 - Clinical Trial Not Allowed) Part 1. Overview Information\n    \n \n \n   Participating Organization(s)\n    \n    National Institutes of Health (\n     \n     NIH\n     \n    )\n    \n \n \n   Components of Participating Organizations\n    \n    National Institute of Diabetes and Digestive and Kidney Diseases (\n     \n     NIDDK\n     \n    )\n     \n \n     Office of The Director, National Institutes of Health (\n      \n      OD\n      \n     )\n     \n \n \n   Funding Opportunity Title\n    \n    Collaborative Research Using Biosamples and/or Data from Type 1 Diabetes Clinical Studies (R01 - Clinical Trial Not Allowed)\n    \n \n \n   Activity Code\n    \n \n     R01\n     \n    Research Project Grant\n    \n \n \n   Announcement Type\n   Reissue of\n    \n    RFA-DK-22-021\n    \n \n \n \n   Related Notices\n    \n \n \n      April 4, 2024\n      \n     - Overview of Grant Application and Review Changes for Due Dates on or after January 25, 2025. See Notice\n      \n      NOT-OD-24-084\n      \n     .\n     \n \n \n      August 31, 2022\n      \n     - Implementation Changes for Genomic Data Sharing Plans Included with Applications Due on or after January 25, 2023. See Notice\n      \n      NOT-OD-22-198\n      \n     .\n     \n \n \n      August 5, 2022\n      \n     - Implementation Details for the NIH Data Management and Sharing Policy. See Notice\n      \n      NOT-OD-22-189\n      \n     .\n     \n \n \n \n   Funding Opportunity Number (FON)\n    \n    RFA-DK-26-007\n    \n \n \n   Companion Funding Opportunity\n   None\n    \n \n   Number of Applications\n    \n    See\n     \n     Section III. 3. Additional Information on Eligibility\n     \n    .\n    \n \n \n   Assistance Listing Number(s)\n   93.847\n    \n \n   Funding Opportunity Purpose\n    \n    This Notice of Funding Opportunity (NOFO) invites applications for studies of type 1 diabetes etiology and pathogenesis using data and samples from clinical trials and studies. This opportunity is intended to fund investigative teams collaborating to answer important questions about disease mechanisms leading to improved delay and durable prevention of type 1 diabetes. This NOFO is associated with the Special Diabetes Program (\n     \n     https://www.niddk.nih.gov/about-niddk/research-areas/diabetes/type-1-diabetes-special-statutory-funding-program/about-special-diabetes-program\n     \n    ) which funds research on the prevention, treatment, and cure of type 1 diabetes and its complications, including unique, innovative, and collaborative research consortia and clinical trials networks.\n    \n \n \n \n \n \n \n   Funding Opportunity Goal(s)\n    \n    To promote extramural basic and clinical biomedical research that improves the understanding of the mechanisms underlying disease and leads to improved preventions, diagnosis, and treatment of diabetes, digestive, and kidney diseases. Programmatic areas within the National Institute of Diabetes and Digestive and Kidney Diseases include diabetes, digestive, endocrine, hematologic, liver, metabolic, nephrologic, nutrition, obesity, and urologic diseases.\n    \n This variable defines that we need to start a new row.',
]
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]

Evaluation

Metrics

Information Retrieval

Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.875
cosine_accuracy@3	1.0
cosine_accuracy@5	1.0
cosine_accuracy@10	1.0
cosine_precision@1	0.875
cosine_precision@3	0.3333
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.875
cosine_recall@3	1.0
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@10	0.9484
cosine_mrr@10	0.9306
cosine_map@100	0.9306

Training Details

Training Dataset

Unnamed Dataset

Size: 216 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 216 samples:
sentence_0 sentence_1
type string string
details
min: 19 tokens
mean: 34.9 tokens
max: 59 tokens
min: 8 tokens
mean: 377.44 tokens
max: 512 tokens

Samples:

sentence_0	sentence_1
`What initiatives is the Department of Health and Human Services pursuing under opportunity ID [insert ID] to improve public health outcomes?`	`Department of Health and Human Services`
`How does the title of opportunity ID [insert ID] align with the strategic goals of the Department of Health and Human Services?`	`Department of Health and Human Services`
`What are the main goals of the funding opportunity titled "Laboratory Flexible Funding Model (LFFM)" under opportunity ID RFA-FD-25-007?`	RFA-FD-25-007: Laboratory Flexible Funding Model (LFFM) Part 1. Overview Information[object Object] [object Object] [object Object] [object Object] Participating Organization(s)[object Object] [object Object] U.S. Food and Drug Administration ([object Object] [object Object] FDA[object Object] [object Object] )[object Object] [object Object] [object Object] [object Object] [object Object] NOTE: The policies, guidelines, terms, and conditions stated in this Notice of Funding Opportunity (NOFO) may differ from those used by the NIH. Where this NOFO provides specific written guidance that may differ from the general guidance provided in the grant application form, please follow the instructions given in this NOFO.[object Object] [object Object] [object Object] The FDA does not follow the NIH Page Limitation Guidelines or the NIH Review Criteria. Applicants are encouraged to consult with FDA Agency Contacts for additional information regarding page limits and the FDA Objective Review Process.[object Object] [object Object] [object Object] [object Object] Components of Participating Organizations[object Object] [object Object] FOOD AND DRUG ADMINISTRATION ([object Object] [object Object] FDA[object Object] [object Object] )[object Object] [object Object] [object Object] [object Object] Funding Opportunity Title[object Object] [object Object] Laboratory Flexible Funding Model (LFFM)[object Object]...

Loss: MatryoshkaLoss with these parameters:
json
1{ 2 "loss": "MultipleNegativesRankingLoss", 3 "matryoshka_dims": [ 4 768, 5 512, 6 256, 7 128, 8 64 9 ], 10 "matryoshka_weights": [ 11 1, 12 1, 13 1, 14 1, 15 1 16 ], 17 "n_dims_per_step": -1 18}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 10
per_device_eval_batch_size: 10
num_train_epochs: 10
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: 10
per_device_eval_batch_size: 10
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: 10
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}
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
dispatch_batches: None
split_batches: 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	cosine_ndcg@10
1.0	22	0.8768
2.0	44	0.9484
2.2727	50	0.9330
3.0	66	0.9276
4.0	88	0.9484
4.5455	100	0.9330
5.0	110	0.9638
6.0	132	0.9638
6.8182	150	0.9638
7.0	154	0.9638
8.0	176	0.9484
9.0	198	0.9484
9.0909	200	0.9484
10.0	220	0.9484

Framework Versions

Python: 3.11.11
Sentence Transformers: 3.3.1
Transformers: 4.48.3
PyTorch: 2.5.1+cu124
Accelerate: 1.3.0
Datasets: 3.3.2
Tokenizers: 0.21.0

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",
}

MatryoshkaLoss

bibtex

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

MultipleNegativesRankingLoss

bibtex

@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}
}

	sentence_0	sentence_1
type	string	string
details	min: 19 tokens mean: 34.9 tokens max: 59 tokens	min: 8 tokens mean: 377.44 tokens max: 512 tokens