Exempel på grundläggande inbäddningsmodell för vektorsökning (GTE)

Den här notebook-filen visar hur du använder SDK för vektorsökning Python, vilket ger en VectorSearchClient som ett primärt API för att arbeta med vektorsökning.

Den här notebook-filen använder API:er för Databricks Foundation-modell för att komma åt GTE-inbäddningsmodellen för att generera inbäddningar.

%pip install --upgrade --force-reinstall databricks-vectorsearch
dbutils.library.restartPython()

from databricks.vector_search.client import VectorSearchClient

vsc = VectorSearchClient(disable_notice=True)

help(VectorSearchClient)

Ladda leksaksuppsättning i käll-Delta-tabellen

Följande skapar deltatabellen för källan.

# Specify the catalog and schema to use. You must have USE_CATALOG privilege on the catalog and USE_SCHEMA and CREATE_TABLE privileges on the schema.
# Change the catalog and schema here if necessary.

catalog_name = "main"
schema_name = "default"

source_table_name = "wiki_articles_demo"
source_table_fullname = f"{catalog_name}.{schema_name}.{source_table_name}"

# Uncomment if you want to start from scratch.

# spark.sql(f"DROP TABLE {source_table_fullname}")

source_df = spark.read.parquet("/databricks-datasets/wikipedia-datasets/data-001/en_wikipedia/articles-only-parquet").limit(10)
display(source_df)

Segmentexempeldatauppsättning

Genom att segmentera exempeldatamängden kan du undvika att överskrida kontextgränsen för inbäddningsmodellen. GTE-modellen stöder upp till 8 192 token. Databricks rekommenderar dock att du delar upp data i mindre kontextsegment så att du kan mata in en bredare mängd exempel i resonemangsmodellen för ditt RAG-program.

import tiktoken
import pandas as pd

# The GTE model has been trained on a max context lenth of 8192 tokens.
max_chunk_tokens = 8192
encoding = tiktoken.get_encoding("cl100k_base")

def chunk_text(text):
    # Encode and then decode within the UDF
    tokens = encoding.encode(text)
    chunks = []
    while tokens:
        chunk_tokens = tokens[:max_chunk_tokens]
        chunk_text = encoding.decode(chunk_tokens)
        chunks.append(chunk_text)
        tokens = tokens[max_chunk_tokens:]
    return chunks

# Process the data and store in a new list
pandas_df = source_df.toPandas()
processed_data = []
for index, row in pandas_df.iterrows():
    text_chunks = chunk_text(row['text'])
    chunk_no = 0
    for chunk in text_chunks:
        row_data = row.to_dict()

        # replace the id column with a new unique chunk id
        # and the text column with the text chunk
        row_data['id'] = f"{row['id']}_{chunk_no}"
        row_data['text'] = chunk

        processed_data.append(row_data)
        chunk_no += 1

chunked_pandas_df = pd.DataFrame(processed_data)
chunked_spark_df = spark.createDataFrame(chunked_pandas_df)

# Write the chunked DataFrame to a Delta table
spark.sql(f"DROP TABLE IF EXISTS {source_table_fullname}")
chunked_spark_df.write.format("delta") \
    .option("delta.enableChangeDataFeed", "true") \
    .saveAsTable(source_table_fullname)

display(spark.sql(f"SELECT * FROM {source_table_fullname}"))

Skapa slutpunkt för vektorsökning

vector_search_endpoint_name = "vector-search-demo-endpoint"

vsc.create_endpoint(
    name=vector_search_endpoint_name,
    endpoint_type="STANDARD" # or "STORAGE_OPTIMIZED"
)

vsc.get_endpoint(
  name=vector_search_endpoint_name
)

Skapa vektorindex

# Vector index
vs_index = f"{source_table_name}_gte_index"
vs_index_fullname = f"{catalog_name}.{schema_name}.{vs_index}"

embedding_model_endpoint = "databricks-gte-large-en"

index = vsc.create_delta_sync_index(
  endpoint_name=vector_search_endpoint_name,
  source_table_name=source_table_fullname,
  index_name=vs_index_fullname,
  pipeline_type='TRIGGERED',
  primary_key="id",
  embedding_source_column="text",
  embedding_model_endpoint_name=embedding_model_endpoint
)
index.describe()['status']['message']

# Wait for index to come online. Expect this command to take several minutes.
# You can also track the status of the index build in Catalog Explorer in the
# Overview tab for the vector index.
import time
index = vsc.get_index(endpoint_name=vector_search_endpoint_name,index_name=vs_index_fullname)
while not index.describe().get('status')['ready']:
  print("Waiting for index to be ready...")
  time.sleep(30)
print("Index is ready!")
index.describe()

Likhetssökning

Följande celler visar hur du frågar vektorindexet för att hitta liknande dokument.

results = index.similarity_search(
  query_text="Greek myths",
  columns=["id", "text", "title"],
  num_results=5
  )
rows = results['result']['data_array']
for (id, text, title, score) in rows:
  if len(text) > 32:
    # trim text output for readability
    text = text[0:32] + "..."
  print(f"id: {id}  title: {title} text: '{text}' score: {score}")

# Search with a filter. Note that the syntax depends on the endpoint type.

# Standard endpoint syntax
results = index.similarity_search(
  query_text="Greek myths",
  columns=["id", "text", "title"],
  num_results=5,
  filters={"title NOT": "Hercules"}
)

# Storage-optimized endpoint syntax
# results = index.similarity_search(
#   query_text="Greek myths",
#   columns=["id", "text", "title"],
#   num_results=5,
#   filters='title != "Hercules"'
#   )


rows = results['result']['data_array']
for (id, text, title, score) in rows:
  if len(text) > 32:
    # trim text output for readability
    text = text[0:32] + "..."
  print(f"id: {id}  title: {title} text: '{text}' score: {score}")

Ta bort vektorindex

vsc.delete_index(
  endpoint_name=vector_search_endpoint_name,
  index_name=vs_index_fullname
)

Exempelanteckningsbok

Exempel på grundläggande inbäddningsmodell för vektorsökning (GTE)

Hämta anteckningsbok