Pokemons
Collection
5 items
•
Updated
•
2
Squirtle
Squirtle is a distill of bge-base-en-v1.5.
Intended purpose
This model is designed for use in semantic-autocomplete (click here for demo).
Make sure you also pass pipelineParams={{ pooling: "cls", normalize: true }} since the default pooling in the component is mean.
Usage
Other than within semantic-autocomplete, you can use this model same as bge-base-en-v1.5.
- Downloads last month
- 1,571
This model does not have enough activity to be deployed to Inference API (serverless) yet.
Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Dataset used to train Mihaiii/Squirtle
Space using Mihaiii/Squirtle 1
Collection including Mihaiii/Squirtle
Evaluation results
- accuracy on MTEB AmazonCounterfactualClassification (en)test set self-reported69.597
- ap on MTEB AmazonCounterfactualClassification (en)test set self-reported31.808
- f1 on MTEB AmazonCounterfactualClassification (en)test set self-reported63.432
- accuracy on MTEB AmazonPolarityClassificationtest set self-reported82.090
- ap on MTEB AmazonPolarityClassificationtest set self-reported76.950
- f1 on MTEB AmazonPolarityClassificationtest set self-reported82.045
- accuracy on MTEB AmazonReviewsClassification (en)test set self-reported41.944
- f1 on MTEB AmazonReviewsClassification (en)test set self-reported40.410
- map_at_1 on MTEB ArguAnatest set self-reported13.869
- map_at_10 on MTEB ArguAnatest set self-reported24.631