[x] Add models to the configuration generator once they are published
Models
Add the following models. I have added a one-sentence description for each model that you can use as the description in the config generator:
sentence-transformers/gtr-t5-base
(New: February 2022) Built on Google's T5 Model. Optimized for Semantic Search. The smallest of the T5 models, however also the weakest in performance.
sentence-transformers/gtr-t5-large
(New: February 2022) Built on Google's T5 Model. Optimized for Semantic Search. Good trade-off between size/accuracy among the T5 models.
sentence-transformers/gtr-t5-xl
(New: February 2022) Built on Google's T5 Model. Optimized for Semantic Search. Best T5-based model for semantic-search. However, beaten by other SBERT models. See benchmarks on sbert.net for details.
Notes
I've decided to skip the sentence-transformers/gtr-t5-xxl model is it's massively big (almost 10 GB) and does not perform any better than xl according to the benchmarks.
Several new SBERT models were released.
Goals
Models
Add the following models. I have added a one-sentence description for each model that you can use as the description in the config generator:
sentence-transformers/gtr-t5-base
sentence-transformers/gtr-t5-large
sentence-transformers/gtr-t5-xl
Notes
I've decided to skip the
sentence-transformers/gtr-t5-xxl
model is it's massively big (almost 10 GB) and does not perform any better thanxl
according to the benchmarks.