model2vec
Model2Vec is a technique to turn any sentence transformer into a really small static modelmodel2vec can be used to generate embeddings.
Setup
pip install -U langchain-community
Instantiation
Ensure thatmodel2vec is installed
pip install -U model2vec
Indexing and Retrieval
from langchain_community.embeddingsimport Model2vecEmbeddings
API Reference:Model2vecEmbeddings
embeddings= Model2vecEmbeddings("minishlab/potion-base-8M")
query_text="This is a test query."
query_result= embeddings.embed_query(query_text)
document_text="This is a test document."
document_result= embeddings.embed_documents([document_text])
Direct Usage
Here's how you would directly make use ofmodel2vec
from model2vecimport StaticModel
# Load a model from the HuggingFace hub (in this case the potion-base-8M model)
model= StaticModel.from_pretrained("minishlab/potion-base-8M")
# Make embeddings
embeddings= model.encode(["It's dangerous to go alone!","It's a secret to everybody."])
# Make sequences of token embeddings
token_embeddings= model.encode_as_sequence(["It's dangerous to go alone!","It's a secret to everybody."])
API Reference
For more information check out the model2vec githubrepo
Related
- Embedding modelconceptual guide
- Embedding modelhow-to guides