ridhanhimmee/model-2-selector-stability-bge

SentenceTransformer based on BAAI/bge-large-en-v1.5

This is a sentence-transformers model finetuned from BAAI/bge-large-en-v1.5. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: BAAI/bge-large-en-v1.5
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 dimensions
• Similarity Function: Cosine Similarity
• Supported Modality: Text

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'transformer_task': 'feature-extraction', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'last_hidden_state'}}, 'module_output_name': 'token_embeddings', 'architecture': 'BertModel'})
  (1): Pooling({'embedding_dimension': 1024, 'pooling_mode': 'cls', '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("ridhanhimmee/model-2-selector-stability-bge")
# Run inference
sentences = [
    'CSS selector: .wixui-footer',
    '<!--/$--></div></div></section><!--/$--></div></div></div><!--/$--></div></div></div></div></main><!--$--><div id="soapAfterPagesContainer" class="soapAfterPagesContainer page-without-sosp"><div data-mesh-id="soapAfterPagesContainerinlineContent" data-testid="inline-content" class=""><div data-mesh-id="soapAfterPagesContainerinlineContent-gridContainer" data-testid="mesh-container-content"><!--$--><div id="CONTROLLER_COMP_CUSTOM_ID" style="display:none"></div><!--/$--></div></div></div><!--/$--><footer id="SITE_FOOTER" class="CDNoF4 SITE_FOOTER wixui-footer" tabindex="-1"><div class="kBvIHj"><div class="EaoiSu" data-testid="screenWidthContainerBg"></div></div><div class="YmWgf4"><div class="H2MmBb"><div class="CmHRsG" data-testid="screenWidthContainerBgCenter"></div></div><div class="XgJ1FR"><div data-mesh-id="SITE_FOOTERinlineContent" data-testid="inline-content" class=""><div data-mesh-id="SITE_FOOTERinlineContent-gridContainer" data-testid="mesh-container-content"><!--$--><div id="comp-jpbunzyw1" class="N8MGzv _v6ohL AWXfZq PO9MfV comp-jpbunzyw1 wixui-rich-text" data-testid="richTextElement"><p class="font_8 wixui-rich-text__text" style="line-height:1.5em; font-size:20px;"><span style="letter-spacing:normal;" class="wixui-rich-text__text"><span class="color_15 wixui-rich-text__text"><a data-auto-recognition="true" href="mailto:hello@tessascott.co.nz" class="wixui-rich-text__text">hello@tessascott.co.nz</a></span></span></p></div><!--/$--><!--$--><div id="comp-jvh5mptc" cla',
    'e5120b" data-element_type="column" data-e-type="column">\n\t\t\t<div class="elementor-widget-wrap elementor-element-populated">\n\t\t\t\t\t\t<div class="elementor-element elementor-element-0c03939 elementor-hidden-tablet elementor-hidden-mobile elementor-widget-mobile__width-initial elementor-widget__width-initial elementor-widget elementor-widget-heading" data-id="0c03939" data-element_type="widget" data-e-type="widget" data-widget_type="heading.default">\n\t\t\t\t<div class="elementor-widget-container">\n\t\t\t\t\t<h2 class="elementor-heading-title elementor-size-xxl">Elevate Your Game</h2>\t\t\t\t</div>\n\t\t\t\t</div>\n\t\t\t\t<div class="elementor-element elementor-element-49d62b8 elementor-widget-mobile__width-initial elementor-hidden-tablet elementor-hidden-mobile elementor-nav-menu__align-center elementor-widget-tablet__width-initial elementor-nav-menu--dropdown-tablet elementor-nav-menu__text-align-aside elementor-nav-menu--toggle elementor-nav-menu--burger elementor-widget elementor-widget-nav-menu" data-id="49d62b8" data-element_type="widget" data-e-type="widget" data-settings="{&quot;layout&quot;:&quot;horizontal&quot;,&quot;submenu_icon&quot;:{&quot;value&quot;:&quot;&lt;i class=\\&quot;fas fa-caret-down\\&quot; aria-hidden=\\&quot;true\\&quot;&gt;&lt;\\/i&gt;&quot;,&quot;library&quot;:&quot;fa-solid&quot;},&quot;toggle&quot;:&quot;burger&quot;}" data-widget_type="nav-menu.default">\n\t\t\t\t<div class="elementor-widget-container">\n\t\t\t\t\t\t\t\t<nav aria-label="Menu" class="elementor-nav-menu--main elementor-nav-',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.8597, 0.0826],
#         [0.8597, 1.0000, 0.0473],
#         [0.0826, 0.0473, 1.0000]])

Evaluation

Metrics

Binary Classification

• Datasets: val and test
• Evaluated with BinaryClassificationEvaluator

Metric	val	test
cosine_accuracy	0.7368	0.7236
cosine_accuracy_threshold	0.6691	0.6662
cosine_f1	0.752	0.7384
cosine_f1_threshold	0.6515	0.6594
cosine_precision	0.701	0.6856
cosine_recall	0.811	0.8
cosine_ap	0.8127	0.7843
cosine_mcc	0.4655	0.4551

Training Details

Training Dataset

Unnamed Dataset

• Size: 10,912 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 100 samples:

  	sentence_0	sentence_1	label
  type	string	string	float
  modality	text	text
  details	min: 8 tokens mean: 12.15 tokens max: 22 tokens	min: 355 tokens mean: 495.17 tokens max: 512 tokens	min: 0.0 mean: 0.56 max: 1.0

• Samples:

  sentence_0	sentence_1	label
  CSS selector: .css-editing-scope	ot;}}" data-motion-part="BG_IMG pageBackground_c1dmp" data-bg-effect-name="" data-has-ssr-src="true"> 1.0
  CSS selector: .elementskit-menu-offcanvas-elements	ontainer"> Emergency Sprinkler Repair We’re here to Help! Accidents happen, so if you break a head or spot a gusher DON’T PANIC! Check out what you should do immediately and give us a call. 303-646-2323