SentenceTransformer based on Snowflake/snowflake-arctic-embed-l

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.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("Gonalb/flucold-ft-v0")
# Run inference
sentences = [
    'QUESTION #1: How might changes in posture from sitting to supine affect sinus pain according to the context?',
    'gas absorption from the sinus and “vacuum maxillary\nsinusitis”.37 However, sinuses with patent ostia may also\nbe painful, indicating that the generation of\ninﬂammatory mediators within the sinus may be\nsufﬁcient to trigger the sensation of pain either by direct\nstimulation of pain nerve ﬁbres or via distension of blood\nvessels that are also served by sensory nerves.36 Changes\nin posture from sitting to supine cause an increase in\nsinus pain that may be related to dilation of the blood\nvessels draining the sinus caused by an increase in\nvenous pressure. Pressure changes in the sinus may also\ncause pain by stimulation of branches of the trigeminal\nnerve that course in and around the sinuses.37\nWatery eyes',
    'American Indians and Alaska Natives\nChildren younger than 5 years (particularly those younger \nthan 2 years)\nInstitutionalized adults (e.g., residents of nursing homes or \nchronic care facilities)\nPregnant and postpartum women (up to 2 weeks postpartum, \nincluding pregnancy loss)\nAdapted with permission from Erlikh IV, Abraham S, Kondamudi VK. \nManagement of influenza. Am Fam Physician. 2010; 82(9): 1088, with \nadditional information from references 18 and 19.',
]
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	0.75
cosine_accuracy@3	0.9167
cosine_accuracy@5	1.0
cosine_accuracy@10	1.0
cosine_precision@1	0.75
cosine_precision@3	0.3056
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.75
cosine_recall@3	0.9167
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@10	0.8865
cosine_mrr@10	0.8486
cosine_map@100	0.8486

Training Details

Training Dataset

Unnamed Dataset

Size: 334 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 334 samples:
sentence_0 sentence_1
type string string
details
min: 12 tokens
mean: 24.85 tokens
max: 61 tokens

min: 61 tokens
mean: 159.74 tokens
max: 248 tokens

	sentence_0	sentence_1
type	string	string
details	min: 12 tokens mean: 24.85 tokens max: 61 tokens	min: 61 tokens mean: 159.74 tokens max: 248 tokens

Samples:

sentence_0	sentence_1
`QUESTION #1: What is the source website from which the document was downloaded?`	`Downloaded from the American Family Physician website at www.aafp.org/afp. Copyright © 2019 American Academy of Family Physicians. For the private, noncom - mercial use of one individual user of the website. All other rights reserved. Contact [email protected] for copyright questions and/or permission requests.`
`QUESTION #2: Who should be contacted for copyright questions and/or permission requests regarding the document?`	`Downloaded from the American Family Physician website at www.aafp.org/afp. Copyright © 2019 American Academy of Family Physicians. For the private, noncom - mercial use of one individual user of the website. All other rights reserved. Contact [email protected] for copyright questions and/or permission requests.`
`QUESTION #1: Why is early diagnosis essential for antiviral therapy and public-health measures in the community?`	syndrome (SARS) 3 because early diagnosis is essential for any antiviral therapy and for the initiation of public- health measures in the community (eg, isolation of infected cases). Here, I discuss the mechanisms that generate symptoms associated with URTIs, especially common cold and ﬂu, but will not review virology in any detail except as regards relevance to symptoms. Is it a cold or ﬂu? The clinical expression of URTIs is variable and is partly inﬂuenced by the nature of the infecting virus but to a greater extent is modulated by the age, physiological state, and immunological experience of the host. 4 Depending on these factors, URTIs may occur without symptoms, may kill, or most commonly

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "MultipleNegativesRankingLoss",
    "matryoshka_dims": [
        768,
        512,
        256,
        128,
        64
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

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	34	0.9108
1.4706	50	0.9098
2.0	68	0.8834
2.9412	100	0.9051
3.0	102	0.9066
4.0	136	0.9205
4.4118	150	0.9019
5.0	170	0.9156
5.8824	200	0.9247
6.0	204	0.9238
7.0	238	0.9019
7.3529	250	0.8856
8.0	272	0.8856
8.8235	300	0.8879
9.0	306	0.8879
10.0	340	0.8865

Framework Versions

Python: 3.11.11
Sentence Transformers: 3.4.1
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

@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

@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

@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}
}

Gonalb
/

flucold-ft-v0