Static Embeddings with BEE-spoke-data/wordpiece-tokenizer-32k-en_code-msp tokenizer finetuned on GooAQ pairs

This is a sentence-transformers model trained on the gooaq dataset. 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
Maximum Sequence Length: inf tokens
Output Dimensionality: 1024 dimensions
Similarity Function: Cosine Similarity
Training Dataset:
- gooaq
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): StaticEmbedding(
    (embedding): EmbeddingBag(31999, 1024, mode='mean')
  )
)

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("tomaarsen/static-BEE-spoke-data-tokenizer-v1-gooaq")
# Run inference
sentences = [
    "how to reverse a video on tiktok that's not yours?",
    '[\'Tap "Effects" at the bottom of your screen — it\\\'s an icon that looks like a clock. Open the Effects menu. ... \', \'At the end of the new list that appears, tap "Time." Select "Time" at the end. ... \', \'Select "Reverse" — you\\\'ll then see a preview of your new, reversed video appear on the screen.\']',
    'Relative age is the age of a rock layer (or the fossils it contains) compared to other layers. It can be determined by looking at the position of rock layers. Absolute age is the numeric age of a layer of rocks or fossils. Absolute age can be determined by using radiometric dating.',
]
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

Dataset: gooaq-1024-dev
Evaluated with InformationRetrievalEvaluator with these parameters:
```
{
    "truncate_dim": 1024
}
```

Metric	Value
cosine_accuracy@1	0.6374
cosine_accuracy@3	0.8431
cosine_accuracy@5	0.9006
cosine_accuracy@10	0.9474
cosine_precision@1	0.6374
cosine_precision@3	0.281
cosine_precision@5	0.1801
cosine_precision@10	0.0947
cosine_recall@1	0.6374
cosine_recall@3	0.8431
cosine_recall@5	0.9006
cosine_recall@10	0.9474
cosine_ndcg@10	0.7979
cosine_mrr@10	0.7492
cosine_map@100	0.7516

Information Retrieval

Dataset: gooaq-512-dev
Evaluated with InformationRetrievalEvaluator with these parameters:
```
{
    "truncate_dim": 512
}
```

Metric	Value
cosine_accuracy@1	0.6324
cosine_accuracy@3	0.8382
cosine_accuracy@5	0.899
cosine_accuracy@10	0.9452
cosine_precision@1	0.6324
cosine_precision@3	0.2794
cosine_precision@5	0.1798
cosine_precision@10	0.0945
cosine_recall@1	0.6324
cosine_recall@3	0.8382
cosine_recall@5	0.899
cosine_recall@10	0.9452
cosine_ndcg@10	0.7945
cosine_mrr@10	0.7455
cosine_map@100	0.748

Information Retrieval

Dataset: gooaq-256-dev
Evaluated with InformationRetrievalEvaluator with these parameters:
```
{
    "truncate_dim": 256
}
```

Metric	Value
cosine_accuracy@1	0.6273
cosine_accuracy@3	0.8299
cosine_accuracy@5	0.8919
cosine_accuracy@10	0.9414
cosine_precision@1	0.6273
cosine_precision@3	0.2766
cosine_precision@5	0.1784
cosine_precision@10	0.0941
cosine_recall@1	0.6273
cosine_recall@3	0.8299
cosine_recall@5	0.8919
cosine_recall@10	0.9414
cosine_ndcg@10	0.7888
cosine_mrr@10	0.7392
cosine_map@100	0.7418

Information Retrieval

Dataset: gooaq-128-dev
Evaluated with InformationRetrievalEvaluator with these parameters:
```
{
    "truncate_dim": 128
}
```

Metric	Value
cosine_accuracy@1	0.6009
cosine_accuracy@3	0.8112
cosine_accuracy@5	0.874
cosine_accuracy@10	0.9296
cosine_precision@1	0.6009
cosine_precision@3	0.2704
cosine_precision@5	0.1748
cosine_precision@10	0.093
cosine_recall@1	0.6009
cosine_recall@3	0.8112
cosine_recall@5	0.874
cosine_recall@10	0.9296
cosine_ndcg@10	0.7688
cosine_mrr@10	0.7167
cosine_map@100	0.7196

Information Retrieval

Dataset: gooaq-64-dev
Evaluated with InformationRetrievalEvaluator with these parameters:
```
{
    "truncate_dim": 64
}
```

Metric	Value
cosine_accuracy@1	0.5589
cosine_accuracy@3	0.7641
cosine_accuracy@5	0.8367
cosine_accuracy@10	0.8995
cosine_precision@1	0.5589
cosine_precision@3	0.2547
cosine_precision@5	0.1673
cosine_precision@10	0.09
cosine_recall@1	0.5589
cosine_recall@3	0.7641
cosine_recall@5	0.8367
cosine_recall@10	0.8995
cosine_ndcg@10	0.7291
cosine_mrr@10	0.6743
cosine_map@100	0.6782

Information Retrieval

Dataset: gooaq-32-dev
Evaluated with InformationRetrievalEvaluator with these parameters:
```
{
    "truncate_dim": 32
}
```

Metric	Value
cosine_accuracy@1	0.4745
cosine_accuracy@3	0.6696
cosine_accuracy@5	0.7508
cosine_accuracy@10	0.8302
cosine_precision@1	0.4745
cosine_precision@3	0.2232
cosine_precision@5	0.1502
cosine_precision@10	0.083
cosine_recall@1	0.4745
cosine_recall@3	0.6696
cosine_recall@5	0.7508
cosine_recall@10	0.8302
cosine_ndcg@10	0.6483
cosine_mrr@10	0.5904
cosine_map@100	0.5962

Training Details

Training Dataset

gooaq

Dataset: gooaq at b089f72
Size: 3,002,496 training samples
Columns: question and answer
Approximate statistics based on the first 1000 samples:
question answer
type string string
details
min: 18 characters
mean: 43.23 characters
max: 96 characters

min: 55 characters
mean: 253.36 characters
max: 371 characters

	question	answer
type	string	string
details	min: 18 characters mean: 43.23 characters max: 96 characters	min: 55 characters mean: 253.36 characters max: 371 characters

Samples:

question	answer
`what is the difference between broilers and layers?`	`An egg laying poultry is called egger or layer whereas broilers are reared for obtaining meat. So a layer should be able to produce more number of large sized eggs, without growing too much. On the other hand, a broiler should yield more meat and hence should be able to grow well.`
`what is the difference between chronological order and spatial order?`	`As a writer, you should always remember that unlike chronological order and the other organizational methods for data, spatial order does not take into account the time. Spatial order is primarily focused on the location. All it does is take into account the location of objects and not the time.`
`is kamagra same as viagra?`	`Kamagra is thought to contain the same active ingredient as Viagra, sildenafil citrate. In theory, it should work in much the same way as Viagra, taking about 45 minutes to take effect, and lasting for around 4-6 hours. However, this will vary from person to person.`

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "MultipleNegativesRankingLoss",
    "matryoshka_dims": [
        1024,
        512,
        256,
        128,
        64,
        32
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

Evaluation Dataset

gooaq

Dataset: gooaq at b089f72
Size: 10,000 evaluation samples
Columns: question and answer
Approximate statistics based on the first 1000 samples:
question answer
type string string
details
min: 18 characters
mean: 43.17 characters
max: 98 characters

min: 51 characters
mean: 254.12 characters
max: 360 characters

	question	answer
type	string	string
details	min: 18 characters mean: 43.17 characters max: 98 characters	min: 51 characters mean: 254.12 characters max: 360 characters

Samples:

question	answer
`how do i program my directv remote with my tv?`	`['Press MENU on your remote.', 'Select Settings & Help > Settings > Remote Control > Program Remote.', 'Choose the device (TV, audio, DVD) you wish to program. ... ', 'Follow the on-screen prompts to complete programming.']`
`are rodrigues fruit bats nocturnal?`	`Before its numbers were threatened by habitat destruction, storms, and hunting, some of those groups could number 500 or more members. Sunrise, sunset. Rodrigues fruit bats are most active at dawn, at dusk, and at night.`
`why does your heart rate increase during exercise bbc bitesize?`	`During exercise there is an increase in physical activity and muscle cells respire more than they do when the body is at rest. The heart rate increases during exercise. The rate and depth of breathing increases - this makes sure that more oxygen is absorbed into the blood, and more carbon dioxide is removed from it.`

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "MultipleNegativesRankingLoss",
    "matryoshka_dims": [
        1024,
        512,
        256,
        128,
        64,
        32
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 2048
per_device_eval_batch_size: 2048
learning_rate: 0.2
num_train_epochs: 1
warmup_ratio: 0.1
bf16: True
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 2048
per_device_eval_batch_size: 2048
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: 0.2
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 1
max_steps: -1
lr_scheduler_type: linear
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: 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: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	Validation Loss	gooaq-1024-dev_cosine_ndcg@10	gooaq-512-dev_cosine_ndcg@10	gooaq-256-dev_cosine_ndcg@10	gooaq-128-dev_cosine_ndcg@10	gooaq-64-dev_cosine_ndcg@10	gooaq-32-dev_cosine_ndcg@10
-1	-1	-	-	0.2283	0.2131	0.1847	0.1395	0.0746	0.0334
0.0007	1	44.3995	-	-	-	-	-	-	-
0.0682	100	20.4944	-	-	-	-	-	-	-
0.1363	200	5.7432	-	-	-	-	-	-	-
0.1704	250	-	1.6135	0.7337	0.7307	0.7204	0.7006	0.6527	0.5522
0.2045	300	4.6818	-	-	-	-	-	-	-
0.2727	400	4.237	-	-	-	-	-	-	-
0.3408	500	3.9465	1.3375	0.7628	0.7601	0.7544	0.7340	0.6917	0.6024
0.4090	600	3.724	-	-	-	-	-	-	-
0.4772	700	3.5496	-	-	-	-	-	-	-
0.5112	750	-	1.2214	0.7782	0.7764	0.7676	0.7492	0.7075	0.6208
0.5453	800	3.4443	-	-	-	-	-	-	-
0.6135	900	3.3312	-	-	-	-	-	-	-
0.6817	1000	3.2537	1.1280	0.7877	0.7841	0.7768	0.7582	0.7195	0.6336
0.7498	1100	3.1613	-	-	-	-	-	-	-
0.8180	1200	3.0985	-	-	-	-	-	-	-
0.8521	1250	-	1.0693	0.7955	0.7922	0.7858	0.7663	0.7267	0.6434
0.8862	1300	3.0416	-	-	-	-	-	-	-
0.9543	1400	3.0249	-	-	-	-	-	-	-
-1	-1	-	-	0.7979	0.7945	0.7888	0.7688	0.7291	0.6483

Environmental Impact

Carbon emissions were measured using CodeCarbon.

Energy Consumed: 0.026 kWh
Carbon Emitted: 0.010 kg of CO2
Hours Used: 0.173 hours

Training Hardware

On Cloud: No
GPU Model: 1 x NVIDIA GeForce RTX 3090
CPU Model: 13th Gen Intel(R) Core(TM) i7-13700K
RAM Size: 31.78 GB

Framework Versions

Python: 3.11.6
Sentence Transformers: 4.1.0.dev0
Transformers: 4.49.0
PyTorch: 2.6.0+cu124
Accelerate: 1.5.1
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}
}

Downloads last month: -; Downloads are not tracked for this model. How to track

Dataset used to train tomaarsen/static-BEE-spoke-data-tokenizer-v1-gooaq

Evaluation results

Cosine Accuracy@1 on gooaq 1024 dev
self-reported

0.637
Cosine Accuracy@3 on gooaq 1024 dev
self-reported

0.843
Cosine Accuracy@5 on gooaq 1024 dev
self-reported

0.901
Cosine Accuracy@10 on gooaq 1024 dev
self-reported

0.947
Cosine Precision@1 on gooaq 1024 dev
self-reported

0.637
Cosine Precision@3 on gooaq 1024 dev
self-reported

0.281
Cosine Precision@5 on gooaq 1024 dev
self-reported

0.180
Cosine Precision@10 on gooaq 1024 dev
self-reported

0.095
Cosine Recall@1 on gooaq 1024 dev
self-reported

0.637
Cosine Recall@3 on gooaq 1024 dev
self-reported

0.843

View on Papers With Code