metadata

base_model: sentence-transformers/msmarco-distilbert-base-tas-b
datasets: []
language: []
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:6192
  - loss:MultipleNegativesRankingLoss
widget:
  - source_sentence: how to calculate a service load
    sentences:
      - what is the height of a lead in antenna
      - types se cable
      - what is the purpose of a circuit breaker
  - source_sentence: minimum ampacity for ungrounded conductors
    sentences:
      - types of mv cables
      - can optical fiber cables be installed in raceway
      - what is a motor and motor operated equipment
  - source_sentence: what is the code for a circuit breaker
    sentences:
      - what color insulation is required to be grounded
      - what is a suitable marker for antflix
      - what conductors are permitted to originate in auxiliary gutter
  - source_sentence: what is plfa cable
    sentences:
      - what is a noncombustible surface
      - how much liquid can be enclosed in a capacitor
      - what is flammable gas in a busway
  - source_sentence: how many volts to ground a transformer
    sentences:
      - what is a grounded conductor
      - how long is a plenum cable
      - what is the operating voltage of a transformer

SentenceTransformer based on sentence-transformers/msmarco-distilbert-base-tas-b

This is a sentence-transformers model finetuned from sentence-transformers/msmarco-distilbert-base-tas-b. 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: sentence-transformers/msmarco-distilbert-base-tas-b
Maximum Sequence Length: 1024 tokens
Output Dimensionality: 768 tokens
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': 1024, 'do_lower_case': False}) with Transformer model: DistilBertModel 
  (1): Pooling({'word_embedding_dimension': 768, '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})
)

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
sentences = [
    'how many volts to ground a transformer',
    'how long is a plenum cable',
    'what is the operating voltage of a transformer',
]
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]

Training Details

Training Dataset

Unnamed Dataset

Size: 6,192 training samples
Columns: sentence_0, sentence_1, and sentence_2
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1 sentence_2
type string dict dict
details
min: 5 tokens
mean: 9.71 tokens
max: 33 tokens

	sentence_0	sentence_1	sentence_2
type	string	dict	dict
details	min: 5 tokens mean: 9.71 tokens max: 33 tokens

Samples:

sentence_0	sentence_1	sentence_2
`what is a metal water piping system`	{'content': 'Metal water piping system(s) installed in or attached to a building or structure shall be bonded to any of the following: Service equipment enclosureGrounded conductor at the serviceGrounding electrode conductor, if of sufficient sizeOne or more grounding electrodes used, if the grounding electrode conductor or bonding jumper to the grounding electrode is of sufficient sizeThe bonding jumper(s) shall be installed in accordance with 250.64(A), (B), and (E). The points of attachment of the bonding jumper(s) s'}	`{'content': 'Metal fences enclosing, and other metal structures in or surrounding, a substation with exposed electrical conductors and equipment shall be grounded and bonded to limit step, touch, and transfer voltages. 250.194'}`
`how many amperes should a circuit breaker be`	`{'content': '10 amperes, provided all the following conditions are met: Continuous loads do not exceed 8 amperesOvercurrent protection is provided by one of the following:Branch-circuit-rated circuit breakers are listed and marked for use with 14 AWG copper-clad aluminum conductor.Branch-circuit-rated fuses are listed and marked for use with 14 AWG copper-clad aluminum conductor. 240.4(D)(3)'}`	`{'content': 'For installations to supply only limited loads of a single branch circuit, the branch circuit disconnecting means shall have a rating of not less than 15 amperes. 225.39(A)'}`
`phase converter installation`	`{'content': 'This article covers the installation and use of phase converters. 455.1'}`	{'content': 'The 120-volt ac side of the voltage converter shall be wired in full conformity with the requirements of Parts I, II, and IV of this article for 120-volt electrical systems. Exception: Converters supplied as an integral part of a listed appliance shall not be subject to 551.20(B). All converters and transformers shall be listed for use in recreational vehicles and designed or equipped to provide overtemperature protection. To determine the converter rating, the following percentages shall be applied to the '}

Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 16
per_device_eval_batch_size: 16
num_train_epochs: 10
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
prediction_loss_only: True
per_device_train_batch_size: 16
per_device_eval_batch_size: 16
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_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
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}
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: False
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
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_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss
1.2920	500	0.2747
2.5840	1000	0.0887
3.8760	1500	0.0512
5.1680	2000	0.0344
6.4599	2500	0.0279
7.7519	3000	0.0213
9.0439	3500	0.02

Framework Versions

Python: 3.10.13
Sentence Transformers: 3.0.1
Transformers: 4.39.3
PyTorch: 2.1.2
Accelerate: 0.32.1
Datasets: 2.20.0
Tokenizers: 0.15.2

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}
}