Views
No views yet
1# Install required packages
2!pip install -U torch==2.5.1 transformers==4.44.2 sentence-transformers==2.7.0 xformers==0.0.28.post3
3
4from sentence_transformers import SentenceTransformer
5
6# Initialize the model
7model = SentenceTransformer('vectopath/SearchMap_Preview', trust_remote_code=True)
8
9# Encode queries
10query = "A treat my dog and I can eat together"
11query_embedding = model.encode(query)
12
13# Encode products
14product_description = "Organic peanut butter dog treats, safe for human consumption..."
15product_embedding = model.encode(product_description)1import numpy as np
2import faiss
3
4# Create FAISS index
5embedding_dimension = 1024 # or your chosen dimension
6index = faiss.IndexFlatL2(embedding_dimension)
7
8# Add product embeddings
9product_embeddings = model.encode(product_descriptions, show_progress_bar=True)
10index.add(np.array(product_embeddings).astype('float32'))
11
12# Search
13query_embedding = model.encode([query])
14distances, indices = index.search(
15 np.array(query_embedding).astype('float32'),
16 k=10
17)
1@misc{vectorpath2025searchmap,
2 title={SearchMap: Conversational E-commerce Search Embedding Model},
3 author={VectorPath Research Team},
4 year={2025},
5 publisher={Hugging Face},
6 journal={HuggingFace Model Hub},
7}