import weaviate
import numpy as np
import weaviate.classes.config as wc

client = weaviate.connect_to_local()
foo = client.collections.create(name="Foo")
with foo.batch.dynamic() as batch:
    for i in range(2000):
        batch.add_object(vector=np.random.rand(100).tolist())

foo = client.collections.get("Foo")
foo.config.update(
    vectorizer_config=wc.Reconfigure.VectorIndex.hnsw(
        quantizer=wc.Reconfigure.VectorIndex.Quantizer.sq(rescore_limit=31, training_limit=2001)
    )
)

# After training complete
foo.config.get().vector_index_config.to_dict()

{'cleanupIntervalSeconds': 300,
 'distanceMetric': 'cosine',
 'dynamicEfMin': 100,
 'dynamicEfMax': 500,
 'dynamicEfFactor': 8,
 'ef': -1,
 'efConstruction': 128,
 'flatSearchCutoff': 40000,
 'maxConnections': 64,
 'skip': False,
 'vectorCacheMaxObjects': 1000000000000,
 'sq': {'rescoreLimit': 31, 'trainingLimit': 2001, 'enabled': True}}

client.collections.create(
   name="Bar",
   vector_index_config=wc.Configure.VectorIndex.hnsw(
       quantizer=wc.Configure.VectorIndex.Quantizer.sq(training_limit=1000)
   ),
   properties=[
       wc.Property(name="question", data_type=wc.DataType.TEXT),
       wc.Property(name="answer", data_type=wc.DataType.TEXT),
   ],
)