---
license: mit
language:
- en
---
# Caption Embeddings

This dataset contains the embeddings for the captions of the Flickr8k, MSCOCO and Pascal50s datasets by different embedding models.

The embeddings are split by embedding model / endpoint where the following embedding models were chosen:

cohere -> embed-english-v3.0

openai -> text-embedding-3-small

vertexai(google) -> textembedding-gecko@003

voyageai(anthropic) -> voyage-large-2-instruct

To view the creation process and the usage for the project they were create for, please view https://github.com/OliverGrein/Image-Captioning-6Evaluation/tree/main