Beginner · Text as numbers
Embeddings and search Search the shop's products with a small embedding model: compare queries and titles by cosine similarity, find each query's nearest products, score the results against people's judgements with recall, MRR and nDCG, and ship an index that searches a fraction of the catalogue.
You build search.py, then evaluate.py
Stages 3 Build, 1 Catch, 1 Ship
Time About 3.5 hours Loading the workspace…
Loading the workspace…
Data: Shopping Queries Dataset , by Amazon.com, Inc. or its affiliates, Apache-2.0 . 100 queries from the US shop, chosen at random among its test queries, half of them for validation and half for testing, each with the first 256 numbers of its embedding by Snowflake's arctic-embed-m-v1.5.
Data: Shopping Queries Dataset , by Amazon.com, Inc. or its affiliates, Apache-2.0 . The 2,083 products people judged for those queries, each with the first 256 numbers of its title's embedding by Snowflake's arctic-embed-m-v1.5.
Data: Shopping Queries Dataset , by Amazon.com, Inc. or its affiliates, Apache-2.0 . The label people gave each of those products for each query.