Optimized for code retrieval (17% better than alternatives) / Previous generation of code embeddings. See blog post for details. Visit the Voyage documentation for an overview of all Voyage embedding models and rerankers.Access to models is through the Voyage Python client. You must register for Voyage API keys to access.
# Embed datadata = [ {"id": "code1", "text": "def factorial(n):\n if n == 0:\n return 1\n return n * factorial(n-1)"}, {"id": "code2", "text": "Sort a list using the quicksort algorithm.\nSteps:\n1. Select a pivot.\n2. Partition the list into elements smaller and larger than the pivot.\n3. Recursively apply the process to partitions.\n4. Combine results."}, {"id": "code3", "text": "def reverse_string(s):\n return s[::-1]"}, {"id": "code4", "text": "Determine if a number is prime - iterate from 2 to sqrt(num) to check divisibility."}, {"id": "code5", "text": "def fibonacci(n):\n if n <= 0:\n return 0\n elif n == 1:\n return 1\n return fibonacci(n-1) + fibonacci(n-2)"},]import voyageaivo = voyageai.Client(api_key=VOYAGE_API_KEY)model_id = "voyage-code-2"def embed(docs: list[str], input_type: str) -> list[list[float]]: embeddings = vo.embed( docs, model=model_id, input_type=input_type ).embeddings return embeddings# Use "document" input type for documentsembeddings = embed([d["text"] for d in data], input_type="document")vectors = []for d, e in zip(data, embeddings): vectors.append({ "id": d['id'], "values": e, "metadata": {'text': d['text']} })index.upsert( vectors=vectors, namespace="ns1")