metadata

language:
  - en
license: apache-2.0
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:72
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
base_model: nomic-ai/modernbert-embed-base
widget:
  - source_sentence: What do the packets contain that is essential for their travel?
    sentences:
      - >-
        Packet switching breaks data into small packets, each containing a
        destination address. These packets travel independently across the
        network, taking different paths if necessary, and reassemble at the
        destination. This method proved to be more efficient and resilient,
        making it the backbone of modern internet communication.


        The Birth of ARPANET
      - Early Concepts of Networking
      - >-
        . Researchers such as Paul Baran at RAND Corporation and Donald Davies
        at the National Physical Laboratory in the UK independently developed
        the concept of packet switching.
  - source_sentence: >-
      Which laboratory was Donald Davies associated with when he developed
      packet switching?
    sentences:
      - >-
        . Understanding the beginning of the internet requires an exploration of
        the early concepts of networking, the establishment of ARPANET, and the
        development of key protocols that laid the foundation for the modern
        internet.
      - >-
        . Researchers such as Paul Baran at RAND Corporation and Donald Davies
        at the National Physical Laboratory in the UK independently developed
        the concept of packet switching.
      - |-
        The Beginning of the Internet: A Journey Through Innovation

        Introduction
  - source_sentence: What role did commercial networking play in relation to the internet?
    sentences:
      - >-
        Beyond ARPANET, various institutions contributed to the internet’s
        expansion. The emergence of local area networks (LANs), the Domain Name
        System (DNS), and the rise of commercial networking played significant
        roles in shaping the internet.
      - >-
        Beyond ARPANET, various institutions contributed to the internet’s
        expansion. The emergence of local area networks (LANs), the Domain Name
        System (DNS), and the rise of commercial networking played significant
        roles in shaping the internet.
      - >-
        The TCP/IP Protocol SuiteIn 1973, Vinton Cerf and Robert Kahn developed
        the Transmission Control Protocol (TCP) and later, the Internet Protocol
        (IP), collectively known as TCP/IP. This protocol suite allowed networks
        of different architectures to communicate, forming the foundation of the
        modern internet.
  - source_sentence: What type of communication system did the United States government seek?
    sentences:
      - >-
        Packet Switching and Its RoleTraditional telephone networks relied on
        circuit-switching, which established a direct connection between two
        parties. However, circuit-switching was inefficient for data
        communication, as it required dedicated lines for each connection
      - >-
        Packet switching breaks data into small packets, each containing a
        destination address. These packets travel independently across the
        network, taking different paths if necessary, and reassemble at the
        destination. This method proved to be more efficient and resilient,
        making it the backbone of modern internet communication.


        The Birth of ARPANET
      - >-
        The idea of interconnected networks dates back to the 1950s and 1960s,
        during the height of the Cold War. The United States government,
        concerned with maintaining communication in the event of a nuclear
        attack, sought a decentralized communication system that could withstand
        disruptions. This vision led to research in packet-switching technology
        and distributed networks.
  - source_sentence: What are the origins of the internet said to be rooted in?
    sentences:
      - >-
        The internet is one of the most transformative technological
        advancements in human history, shaping the way we communicate, work, and
        interact with the world. However, its origins are rooted in decades of
        research, experimentation, and collaboration among scientists,
        engineers, and visionaries
      - |-
        The Beginning of the Internet: A Journey Through Innovation

        Introduction
      - >-
        The first successful ARPANET message was sent on October 29, 1969, from
        UCLA to SRI. The intended message was “LOGIN,” but the system crashed
        after transmitting only “LO.” This marked the first instance of
        networked digital communication, paving the way for the modern internet.


        Expansion and Development of Protocols
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@3
  - cosine_accuracy@5
  - cosine_accuracy@10
  - cosine_precision@1
  - cosine_precision@3
  - cosine_precision@5
  - cosine_precision@10
  - cosine_recall@1
  - cosine_recall@3
  - cosine_recall@5
  - cosine_recall@10
  - cosine_ndcg@10
  - cosine_mrr@10
  - cosine_map@100
model-index:
  - name: Fine-tuned with [QuicKB](https://github.com/ALucek/QuicKB)
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 768
          type: dim_768
        metrics:
          - type: cosine_accuracy@1
            value: 1
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 1
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 1
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 1
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.3333333333333333
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.2
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.1
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 1
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 1
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 1
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 1
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 1
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 1
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 512
          type: dim_512
        metrics:
          - type: cosine_accuracy@1
            value: 1
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 1
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 1
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 1
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.3333333333333333
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.2
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.1
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 1
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 1
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 1
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 1
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 1
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 1
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 256
          type: dim_256
        metrics:
          - type: cosine_accuracy@1
            value: 1
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 1
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 1
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 1
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.3333333333333333
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.2
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.1
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 1
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 1
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 1
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 1
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 1
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 1
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 128
          type: dim_128
        metrics:
          - type: cosine_accuracy@1
            value: 0.875
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.875
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.875
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.875
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.29166666666666663
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.17500000000000002
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.1
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.875
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.875
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.875
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.9195258983885027
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.8958333333333334
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.8958333333333334
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 64
          type: dim_64
        metrics:
          - type: cosine_accuracy@1
            value: 0.625
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.625
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.875
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.625
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.20833333333333331
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.17500000000000002
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.1
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.625
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.625
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.875
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.7662391001971381
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.6958333333333334
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.6958333333333333
            name: Cosine Map@100

Fine-tuned with QuicKB

This is a sentence-transformers model finetuned from nomic-ai/modernbert-embed-base. It maps sentences & paragraphs to a 768-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: nomic-ai/modernbert-embed-base
Maximum Sequence Length: 1024 tokens
Output Dimensionality: 768 dimensions
Similarity Function: Cosine Similarity
Language: en
License: apache-2.0

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': 1024, 'do_lower_case': False}) with Transformer model: ModernBertModel 
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, '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("onecd2000/modernbert-embed-test")
# Run inference
sentences = [
    'What are the origins of the internet said to be rooted in?',
    'The internet is one of the most transformative technological advancements in human history, shaping the way we communicate, work, and interact with the world. However, its origins are rooted in decades of research, experimentation, and collaboration among scientists, engineers, and visionaries',
    'The first successful ARPANET message was sent on October 29, 1969, from UCLA to SRI. The intended message was “LOGIN,” but the system crashed after transmitting only “LO.” This marked the first instance of networked digital communication, paving the way for the modern internet.\n\nExpansion and Development of Protocols',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

Datasets: dim_768, dim_512, dim_256, dim_128 and dim_64
Evaluated with InformationRetrievalEvaluator

Metric	dim_768	dim_512	dim_256	dim_128	dim_64
cosine_accuracy@1	1.0	1.0	1.0	0.875	0.625
cosine_accuracy@3	1.0	1.0	1.0	0.875	0.625
cosine_accuracy@5	1.0	1.0	1.0	0.875	0.875
cosine_accuracy@10	1.0	1.0	1.0	1.0	1.0
cosine_precision@1	1.0	1.0	1.0	0.875	0.625
cosine_precision@3	0.3333	0.3333	0.3333	0.2917	0.2083
cosine_precision@5	0.2	0.2	0.2	0.175	0.175
cosine_precision@10	0.1	0.1	0.1	0.1	0.1
cosine_recall@1	1.0	1.0	1.0	0.875	0.625
cosine_recall@3	1.0	1.0	1.0	0.875	0.625
cosine_recall@5	1.0	1.0	1.0	0.875	0.875
cosine_recall@10	1.0	1.0	1.0	1.0	1.0
cosine_ndcg@10	1.0	1.0	1.0	0.9195	0.7662
cosine_mrr@10	1.0	1.0	1.0	0.8958	0.6958
cosine_map@100	1.0	1.0	1.0	0.8958	0.6958

Training Details

Training Dataset

Unnamed Dataset

Size: 72 training samples
Columns: anchor and positive
Approximate statistics based on the first 72 samples:
anchor positive
type string string
details
min: 9 tokens
mean: 14.6 tokens
max: 20 tokens

min: 8 tokens
mean: 57.92 tokens
max: 89 tokens

	anchor	positive
type	string	string
details	min: 9 tokens mean: 14.6 tokens max: 20 tokens	min: 8 tokens mean: 57.92 tokens max: 89 tokens

Samples:

anchor	positive
`What does the text focus on regarding the Internet?`	`The Beginning of the Internet: A Journey Through Innovation Introduction`
`What was the primary purpose of the first web browser?`	`The First Web Browser – A tool for accessing and navigating websites. The World Wide Web revolutionized internet usage, making it more accessible and appealing to the general public. By the mid-1990s, web browsers like Netscape Navigator and Microsoft Internet Explorer fueled rapid internet adoption, leading to the digital age we live in today. Conclusion`
`What system contributed to the organization of internet addresses?`	`Beyond ARPANET, various institutions contributed to the internet’s expansion. The emergence of local area networks (LANs), the Domain Name System (DNS), and the rise of commercial networking played significant roles in shaping the internet.`

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: epoch
per_device_train_batch_size: 32
gradient_accumulation_steps: 16
learning_rate: 2e-05
num_train_epochs: 4
lr_scheduler_type: cosine
warmup_ratio: 0.1
bf16: True
tf32: True
load_best_model_at_end: True
optim: adamw_torch_fused
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: epoch
prediction_loss_only: True
per_device_train_batch_size: 32
per_device_eval_batch_size: 8
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 16
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: 4
max_steps: -1
lr_scheduler_type: cosine
lr_scheduler_kwargs: {}
warmup_ratio: 0.1
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: True
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: True
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: True
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}
tp_size: 0
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_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
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: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	dim_768_cosine_ndcg@10	dim_512_cosine_ndcg@10	dim_256_cosine_ndcg@10	dim_128_cosine_ndcg@10	dim_64_cosine_ndcg@10
1.0	1	0.8827	0.8827	0.8827	0.7413	0.6383
2.0	2	0.9288	0.9539	1.0	0.8289	0.7611
3.0	3	1.0	1.0	1.0	0.9167	0.7634
4.0	4	1.0	1.0	1.0	0.9195	0.7662

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.12.9
Sentence Transformers: 3.4.1
Transformers: 4.50.0
PyTorch: 2.6.0+cu126
Accelerate: 1.3.0
Datasets: 3.2.0
Tokenizers: 0.21.1

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",
}

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}
}