Pretrained Models
OCI Search with OpenSearch hosts various OpenSearch pretrained models. The built in support for these models means that you don't need to import the model, you only need to register and deploy the model for the cluster.
Supported OpenSearch Pretrained Models
OpenSearch provides several open source pretrained models for a range of machine learning (ML) search and analytics use cases. OCI Search with OpenSearch hosts the OpenSearch pretrained models listed in this section.
Sentence Transformers
Sentence transformer models map sentences and paragraphs across a dimensional dense vector space. The number of vectors depends on the type of model. You can use these models for clustering or semantic search.
| Model name | Version | Vector dimensions | Auto-truncation | Script format |
|---|---|---|---|---|
huggingface/sentence-transformers/all-distilroberta-v1 |
1.0.1 | 768-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/all-MiniLM-L6-v2 |
1.0.1 | 384-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/all-MiniLM-L12-v2 |
1.0.1 | 384-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/all-mpnet-base-v2 |
1.0.1 | 768-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/msmarco-distilbert-base-tas-b |
1.0.2 | 768-dimensional dense vector space. Optimized for semantic search. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/multi-qa-MiniLM-L6-cos-v1 |
1.0.1 | 384-dimensional dense vector space. Designed for semantic search and trained on 215 million question/answer pairs. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/multi-qa-mpnet-base-dot-v1 |
1.0.1 | 384-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/paraphrase-MiniLM-L3-v2 |
1.0.1 | 384-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |
huggingface/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2 |
1.0.1 | 384-dimensional dense vector space. | Yes |
TORCH_SCRIPT ONNX |