mirror of
https://github.com/hanzo-ml/kserve.git
synced 2026-10-30 20:53:49 +00:00
SivananthamandGitHub
2a534af9fd
* Add client sdk for localmodelcache, localmodelnodegroup Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Add e2e test for modelcache Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Use docker driver and minikube tunnel Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Merge "Allow multiple node groups in the model cache CR (#4134)" Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Try mounting image dir Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Add local model agent to image scan Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Debug Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Create model root directory beforehand Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Restart kserve controller after patch Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Enablepvc direct mount in e2e test Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Reduce pv storage to 1GB Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Update modelcache test Signed-off-by: Sivanantham Chinnaiyan <[email protected]> * Update status-check to include modelcache logs Signed-off-by: Sivanantham Chinnaiyan <[email protected]> --------- Signed-off-by: Sivanantham Chinnaiyan <[email protected]>
82 lines
2.6 KiB
Python
82 lines
2.6 KiB
Python
# Copyright 2022 The KServe Authors.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
import os
|
|
from kubernetes import client
|
|
from kserve import (
|
|
constants,
|
|
KServeClient,
|
|
V1beta1InferenceService,
|
|
V1beta1InferenceServiceSpec,
|
|
V1beta1PredictorSpec,
|
|
V1beta1ModelSpec,
|
|
V1beta1ModelFormat,
|
|
)
|
|
from kubernetes.client import V1ResourceRequirements
|
|
import pytest
|
|
|
|
from ..common.utils import predict_isvc
|
|
from ..common.utils import KSERVE_TEST_NAMESPACE
|
|
|
|
|
|
@pytest.mark.predictor
|
|
@pytest.mark.asyncio(scope="session")
|
|
async def test_mlflow_v2_runtime_kserve(rest_v2_client):
|
|
service_name = "isvc-mlflow-v2-runtime"
|
|
protocol_version = "v2"
|
|
|
|
predictor = V1beta1PredictorSpec(
|
|
min_replicas=1,
|
|
model=V1beta1ModelSpec(
|
|
model_format=V1beta1ModelFormat(
|
|
name="mlflow",
|
|
),
|
|
storage_uri="gs://kfserving-examples/models/mlflow/wine",
|
|
protocol_version=protocol_version,
|
|
resources=V1ResourceRequirements(
|
|
requests={"cpu": "50m", "memory": "128Mi"},
|
|
limits={"cpu": "1", "memory": "1Gi"},
|
|
),
|
|
readiness_probe=client.V1Probe(
|
|
http_get=client.V1HTTPGetAction(
|
|
path=f"/v2/models/{service_name}/ready", port=8080
|
|
),
|
|
initial_delay_seconds=30,
|
|
),
|
|
),
|
|
)
|
|
|
|
isvc = V1beta1InferenceService(
|
|
api_version=constants.KSERVE_V1BETA1,
|
|
kind=constants.KSERVE_KIND_INFERENCESERVICE,
|
|
metadata=client.V1ObjectMeta(
|
|
name=service_name, namespace=KSERVE_TEST_NAMESPACE
|
|
),
|
|
spec=V1beta1InferenceServiceSpec(predictor=predictor),
|
|
)
|
|
|
|
kserve_client = KServeClient(
|
|
config_file=os.environ.get("KUBECONFIG", "~/.kube/config")
|
|
)
|
|
kserve_client.create(isvc)
|
|
kserve_client.wait_isvc_ready(service_name, namespace=KSERVE_TEST_NAMESPACE)
|
|
res = await predict_isvc(
|
|
rest_v2_client,
|
|
service_name,
|
|
"./data/mlflow_input_v2.json",
|
|
)
|
|
assert res.outputs[0].data == [5.576883936610762]
|
|
|
|
kserve_client.delete(service_name, KSERVE_TEST_NAMESPACE)
|