Registar e servir um modelo de incorporação OSS

Abrir a versão desta página do bloco de notas

Este caderno configura o modelo de embedding de texto de código aberto e5-small-v2 num endpoint de Model Serving utilizável para Pesquisa Vetorial.

  • Descarregue o modelo do Hugging Face Hub.
  • Registe-o no MLflow Model Registry.
  • Inicie um endpoint de Model Serving para servir o modelo.

O modelo e5-small-v2 está disponível em https://huggingface.co/intfloat/e5-small-v2.

Para uma lista de versões de biblioteca incluídas no Databricks Runtime, consulte as notas de lançamento da sua versão Databricks Runtime.

Instalar Databricks Python SDK

Este notebook utiliza o seu cliente Python para trabalhar com endpoints de serviço.

%pip install -U databricks-sdk python-snappy
%pip install sentence-transformers
dbutils.library.restartPython()

Descarregar modelo

# Download model using the sentence_transformers library.
from sentence_transformers import SentenceTransformer

source_model_name = 'intfloat/e5-small-v2'  # model name on Hugging Face Hub
model = SentenceTransformer(source_model_name)

# Test the model, just to show it works.
sentences = ["This is an example sentence", "Each sentence is converted"]
embeddings = model.encode(sentences)
print(embeddings)

Modelo de registo para MLflow

import mlflow
mlflow.set_registry_uri("databricks-uc")


# 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 = "main"
schema = "default"
model_name = "e5-small-v2"

# MLflow model name. The Model Registry uses this name for the model.
registered_model_name = f"{catalog}.{schema}.{model_name}"

# Compute input and output schema.
signature = mlflow.models.signature.infer_signature(sentences, embeddings)
print(signature)

model_info = mlflow.sentence_transformers.log_model(
  model,
  artifact_path="model",
  signature=signature,
  input_example=sentences,
  registered_model_name=registered_model_name)

inference_test = ["I enjoy pies of both apple and cherry.", "I prefer cookies."]

# Load the custom model by providing the URI for where the model was logged.
loaded_model_pyfunc = mlflow.pyfunc.load_model(model_info.model_uri)

# Perform a quick test to ensure that the loaded model generates the correct output.
embeddings_test = loaded_model_pyfunc.predict(inference_test)
embeddings_test

# Extract the version of the model you just registered.
mlflow_client = mlflow.MlflowClient()

def get_latest_model_version(model_name):
  client = mlflow_client
  model_version_infos = client.search_model_versions("name = '%s'" % model_name)
  return max([int(model_version_info.version) for model_version_info in model_version_infos])

model_version = get_latest_model_version(registered_model_name)
model_version

Criar endpoint de disponibilização de modelo

Para mais detalhes, veja Criar endpoints de serviço para modelo fundamental.

Nota: Este exemplo cria um pequeno endpoint de CPU que reduz para 0. Isto é para testes rápidos e pequenos. Para casos de uso mais realistas, considere usar endpoints GPU para cálculos de embedding mais rápidos e não reduzir para 0 se esperar consultas frequentes, pois os endpoints Model Serving têm alguma sobrecarga de cold start.

endpoint_name = "e5-small-v2"  # Name of endpoint to create

from databricks.sdk import WorkspaceClient
from databricks.sdk.service.serving import EndpointCoreConfigInput

w = WorkspaceClient()

endpoint_config_dict = {
    "served_entities": [
        {
            "name": f'{registered_model_name.replace(".", "_")}_{1}',
            "entity_name": registered_model_name,
            "entity_version": model_version,
            "workload_type": "CPU",
            "workload_size": "Small",
            "scale_to_zero_enabled": True,
        }
    ]
}

endpoint_config = EndpointCoreConfigInput.from_dict(endpoint_config_dict)

# The endpoint may take several minutes to get ready.
w.serving_endpoints.create_and_wait(name=endpoint_name, config=endpoint_config)

Endpoint de consulta

O comando acima create_and_wait espera até que o endpoint esteja pronto. Também pode verificar o estado do endpoint de serviço na interface do Databricks.

Para mais informações, consulte Consultar modelos de fundação.

# Only run this command after the Model Serving endpoint is in the Ready state.
import time

start = time.time()

# If the endpoint is not yet ready, you might get a timeout error. If so, wait and then rerun the command.
endpoint_response = w.serving_endpoints.query(name=endpoint_name, dataframe_records=['Hello world', 'Good morning'])

end = time.time()

print(endpoint_response)
print(f'Time taken for querying endpoint in seconds: {end-start}')

Bloco de notas de exemplo

Registrar e disponibilizar um modelo de embedding OSS

Obter caderno

  • Last updated on