Generate text embeddings for semantic tasks
sentence-transformersskillsetup L2★9,423
Orchestra-Research/AI-Research-SKILLs ↗What it does
Generate dense embeddings from text using pre-trained transformer models
Best for
Text embedding when you need semantic vectors that are domain-tuned or when you want pure open-source.
Inputs
- · text strings
- · optional list of texts
- · optional batch size
Outputs
- · embeddings (768/384/1536-dim)
Requires
- · sentence-transformers
- · torch
- · huggingface_hub
Preconditions
Python 3.7+; model name valid (e.g., 'all-MiniLM-L6-v2'); optional GPU for speed
Failure modes
Out-of-memory on large batches without batch size tuning; model download hangs on no internet
Trust signals
- · 50k+ GitHub stars
- · Apache 2.0 license
- · SBERT architecture pre-trained on 1B+ sentence pairs