Sentence Similarity
sentence-transformers
PyTorch
Transformers
Italian
distilbert
feature-extraction
text-embeddings-inference
Instructions to use efederici/sentence-BERTino with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- sentence-transformers
How to use efederici/sentence-BERTino with sentence-transformers:
from sentence_transformers import SentenceTransformer model = SentenceTransformer("efederici/sentence-BERTino") sentences = [ "Questa è una persona felice", "Questo è un cane felice", "Questa è una persona molto felice", "Oggi è una giornata di sole" ] embeddings = model.encode(sentences) similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [4, 4] - Transformers
How to use efederici/sentence-BERTino with Transformers:
# Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("efederici/sentence-BERTino") model = AutoModel.from_pretrained("efederici/sentence-BERTino") - Notebooks
- Google Colab
- Kaggle
Embedding normalisation
#1
by ancc - opened
Hi,
I'd like to know whether the model was trained with normalised embeddings as to use dot-product in evaluation or the usual cosine similarity should be used.
Kind regards and thanks for this fantastic model
Hi! No, you have to use cosine similarity, or you can normalize your embeddings once computed :)