Batch upsert

Database Vektor untuk Embeddings dengan Pinecone

James Chapman

Curriculum Manager, DataCamp

Batasan upsert

Laju permintaan
Ukuran permintaan

Batching: membagi permintaan menjadi chunk kecil

Batas laju Pinecone

¹ https://docs.pinecone.io/reference/quotas-and-limits#rate-limits

Mendefinisikan fungsi chunking

def chunks(iterable, batch_size=100):

    it = iter(iterable)

    chunk = tuple(itertools.islice(it, batch_size))

    while chunk:

        yield chunk

        chunk = tuple(itertools.islice(it, batch_size))

Batching sekuensial

Membagi permintaan dan mengirimkannya berurutan satu per satu

pc.Pinecone(api_key="YOUR API KEY")
index = pc.Index('datacamp-index')


for chunk in chunks(vectors):
    index.upsert(vectors=chunk)

Kelebihan:

Mengatasi batas laju dan ukuran

Kekurangan:

Sangat lambat!

Batching paralel

Membagi permintaan dan mengirimkannya secara paralel

pc = Pinecone(api_key="YOUR_API_KEY", pool_threads=30)


with pc.Index('datacamp-index', pool_threads=30) as index:

    async_results = [index.upsert(vectors=chunk, async_req=True)
        for chunk in chunks(vectors, batch_size=100)]


    [async_result.get() for async_result in async_results]

Ayo berlatih!

Database Vektor untuk Embeddings dengan Pinecone