768-dimensional semantic search
According to the model card, the model maps sentences and paragraphs to a 768-dimensional dense vector space and was designed for semantic search.
Open Source Model Profile · sentence-transformers
msmarco-distilbert-dot-v5
msmarco-distilbert-dot-v5 is a 66.36M-parameter DistilBERT sentence-similarity model from sentence-transformers for semantic search over 768-dimensional vectors.
- Publisher
- sentence-transformers
- Task
- sentence-similarity
- Model type
- distilbert
- License
- apache-2.0
- Library
- sentence-transformers
- Publication status
- Accepted · not indexed
Model overview
msmarco-distilbert-dot-v5 is published by sentence-transformers as a sentence-similarity model. Captured configuration identifies DistilBertModel with a distilbert model type, and Safetensors metadata reports 66,362,880 parameters under Apache-2.0. According to the model card, it was trained on MS MARCO query-answer pairs for semantic search with 768-dimensional embeddings.
Recorded capabilities
MS MARCO dot-score training
The model card says it was trained on 500K MS MARCO query-answer pairs, with usage examples scoring queries against documents using dot score.
DistilBERT with mean pooling
Captured configuration records DistilBertModel with a distilbert model type, and the card documents mean pooling over a 768-dimensional word-embedding space with 512 maximum sequence length.
Documented MarginMSELoss setup
According to the model card, training used MarginMSELoss with 30 epochs, AdamW, WarmupLinear scheduling, and weight decay 0.01.
Use cases in the source record
- Semantic-search workflows that encode queries and passages into dense vectors and rank them by dot score.
- Passage-retrieval experiments using the card's documented SentenceTransformer and Transformers encoding examples.
Limitations and unknowns
- Provider state is historical snapshot data and should be refreshed before being presented as current availability.
- No evaluation results were extracted from this record.
- No context-window value was extracted beyond the card's 512 maximum sequence length for the Transformer component.
- Dataset reuse is subject to MS MARCO terms in addition to the Apache 2 model license, according to the model card.
Source and provenance
Source: sentence-transformers/msmarco-distilbert-dot-v5
Captured: Unknown. Processed: 2026-09-07T19:34:57.698734+00:00.
msmarco-distilbert-dot-v5 This is a sentence-transformers model: It maps sentences & paragraphs to a 768 dimensional dense vector space and was designed for semantic search . It has been trained on 500K (query, answer) pairs from the MS MARCO dataset . For an introduction to semantic search, have a look at: SBERT.net - Semantic Search Usage (Sentence-Transformers) Using this model becomes easy when you have sentence-transformers installed: pip install -U sentence-transformers Then you can use the model like this: from sentence_transformers import SentenceTransformer, util query = "How many people live in London?" docs = [ "Around 9…
F001F002F003F004F005F006F007F010F011F012F017F018F019F020