Triton Examples

Prerequisites

• For the test data you will need to install torch, torchvision and tensorflow

• For the visualization matplotlib

• For calling the service curl

Setup Seldon Core

Follow the instructions to Setup Cluster with Ambassador Ingress and Install Seldon Core.

Then port-forward to that ingress on localhost:8003 in a separate terminal either with:

• Ambassador: kubectl port-forward $(kubectl get pods -n seldon -l app.kubernetes.io/name=ambassador -o jsonpath='{.items[0].metadata.name}') -n seldon 8003:8080

• Istio: kubectl port-forward $(kubectl get pods -l istio=ingressgateway -n istio-system -o jsonpath='{.items[0].metadata.name}') -n istio-system 8003:8080

Create Namespace for experimentation

We will first set up the namespace of Seldon where we will be deploying all our models

!kubectl create namespace seldon

namespace/seldon created

And then we will set the current workspace to use the seldon namespace so all our commands are run there by default (instead of running everything in the default namespace.)

!kubectl config set-context $(kubectl config current-context) --namespace=seldon

Context "kind-kind" modified.

import os

import matplotlib.pyplot as plt
import tensorflow as tf

train, test = tf.keras.datasets.cifar10.load_data()
X_test, y_test = test
X_test = X_test.astype("float32") / 255
print(X_test.shape, y_test.shape)
class_names = [
    "airplane",
    "automobile",
    "bird",
    "cat",
    "deer",
    "dog",
    "frog",
    "horse",
    "ship",
    "truck",
]

2022-02-01 16:53:52.029992: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
2022-02-01 16:53:52.030017: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.


(10000, 32, 32, 3) (10000, 1)

Triton Model Naming

You need to name the model in the graph with the same name as the triton model loaded as this name will be used in the path to triton.

Tensorflow CIFAR10 Model

%%writefile resources/triton_tf_cifar10.yaml
apiVersion: machinelearning.seldon.io/v1
kind: SeldonDeployment
metadata:
  name: cifar10
  namespace: seldon
spec:
  annotations:
    seldon.io/engine-seldon-log-messages-externally: "true"
  name: resnet32
  predictors:
  - graph:
      implementation: TRITON_SERVER
      logger:
        mode: all
      modelUri: gs://seldon-models/triton/tf_cifar10
      name: cifar10
      type: MODEL
    name: default
    replicas: 1
  protocol: v2

Writing resources/triton_tf_cifar10.yaml

!kubectl apply -f resources/triton_tf_cifar10.yaml

seldondeployment.machinelearning.seldon.io/cifar10 created

!kubectl rollout status -n seldon deploy/$(kubectl get deploy -l seldon-deployment-id=cifar10 -n seldon -o jsonpath='{.items[0].metadata.name}')

Waiting for deployment "cifar10-default-0-cifar10" rollout to finish: 0 of 1 updated replicas are available...
deployment "cifar10-default-0-cifar10" successfully rolled out

import json
from subprocess import PIPE, Popen, run

import numpy as np

idx = 1
test_example = X_test[idx : idx + 1].tolist()
payload = (
    '{"inputs":[{"name":"input_1","datatype":"FP32","shape":[1, 32, 32, 3],"data":'
    + f"{test_example}"
    + "}]}"
)
cmd = f"""curl -d '{payload}' \
   http://localhost:8003/seldon/seldon/cifar10/v2/models/cifar10/infer \
   -H "Content-Type: application/json"
"""
ret = Popen(cmd, shell=True, stdout=PIPE)
raw = ret.stdout.read().decode("utf-8")
res = json.loads(raw)
arr = np.array(res["outputs"][0]["data"])
X = X_test[idx].reshape(1, 32, 32, 3)
plt.imshow(X.reshape(32, 32, 3))
plt.axis("off")
plt.show()
print("class:", class_names[y_test[idx][0]])
print("prediction:", class_names[arr.argmax()])

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 63906  100   339  100 63567    696   127k --:--:-- --:--:-- --:--:--  127k

png

class: ship
prediction: ship

!kubectl delete -f resources/triton_tf_cifar10.yaml

seldondeployment.machinelearning.seldon.io "cifar10" deleted

ONNX CIFAR10 Model

%%writefile resources/triton_onnx_cifar10.yaml
apiVersion: machinelearning.seldon.io/v1
kind: SeldonDeployment
metadata:
  name: cifar10
  namespace: seldon
spec:
  annotations:
    seldon.io/engine-seldon-log-messages-externally: "true"
  name: resnet32
  predictors:
  - graph:
      implementation: TRITON_SERVER
      logger:
        mode: all
      modelUri: gs://seldon-models/triton/onnx_cifar10
      name: cifar10
      type: MODEL
    name: default
    replicas: 1
  protocol: v2

Writing resources/triton_onnx_cifar10.yaml

!kubectl apply -f resources/triton_onnx_cifar10.yaml

seldondeployment.machinelearning.seldon.io/cifar10 created

!kubectl rollout status -n seldon deploy/$(kubectl get deploy -l seldon-deployment-id=cifar10 -n seldon -o jsonpath='{.items[0].metadata.name}')

Waiting for deployment "cifar10-default-0-cifar10" rollout to finish: 0 of 1 updated replicas are available...
deployment "cifar10-default-0-cifar10" successfully rolled out

import json
from subprocess import PIPE, Popen, run

import numpy as np

idx = 1
test_example = X_test[idx : idx + 1].tolist()
payload = (
    '{"inputs":[{"name":"input_1:0","datatype":"FP32","shape":[1, 32, 32, 3],"data":'
    + f"{test_example}"
    + "}]}"
)
cmd = f"""curl -d '{payload}' \
   http://localhost:8003/seldon/seldon/cifar10/v2/models/cifar10/infer \
   -H "Content-Type: application/json"
"""
ret = Popen(cmd, shell=True, stdout=PIPE)
raw = ret.stdout.read().decode("utf-8")
res = json.loads(raw)
arr = np.array(res["outputs"][0]["data"])
X = X_test[idx].reshape(1, 32, 32, 3)
plt.imshow(X.reshape(32, 32, 3))
plt.axis("off")
plt.show()
print("class:", class_names[y_test[idx][0]])
print("prediction:", class_names[arr.argmax()])

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 100 63913 100 344 100 63569 20081 3623k --:--:-- --:--:-- --:--:-- 3651k

class: ship
prediction: ship

!kubectl delete -f resources/triton_onnx_cifar10.yaml

seldondeployment.machinelearning.seldon.io "cifar10" deleted

TorchScript CIFAR10 Model

%%writefile resources/triton_pt_cifar10.yaml
apiVersion: machinelearning.seldon.io/v1
kind: SeldonDeployment
metadata:
  name: cifar10
  namespace: seldon
spec:
  annotations:
    seldon.io/engine-seldon-log-messages-externally: "true"
  name: resnet32
  predictors:
  - graph:
      implementation: TRITON_SERVER
      logger:
        mode: all
      modelUri: gs://seldon-models/triton/pytorch_cifar10
      name: cifar10
      type: MODEL
    name: default
    replicas: 1
  protocol: v2

Writing resources/triton_pt_cifar10.yaml

!kubectl apply -f resources/triton_pt_cifar10.yaml

seldondeployment.machinelearning.seldon.io/cifar10 created

!kubectl rollout status -n seldon deploy/$(kubectl get deploy -l seldon-deployment-id=cifar10 -n seldon -o jsonpath='{.items[0].metadata.name}')

Waiting for deployment "cifar10-default-0-cifar10" rollout to finish: 0 of 1 updated replicas are available...
deployment "cifar10-default-0-cifar10" successfully rolled out

import torch
import torchvision
import torchvision.transforms as transforms

transform = transforms.Compose(
    [transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]
)
testset = torchvision.datasets.CIFAR10(
    root="./data", train=False, download=True, transform=transform
)

testloader = torch.utils.data.DataLoader(
    testset, batch_size=10, shuffle=False, num_workers=2
)
for data in testloader:
    images, labels = data
    break

Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz



  0%|          | 0/170498071 [00:00<?, ?it/s]


Extracting ./data/cifar-10-python.tar.gz to ./data

import json
from subprocess import PIPE, Popen, run

import numpy as np

idx = 3
test_example = images[idx : idx + 1]
test_example = test_example.tolist()
payload = (
    '{"inputs":[{"name":"input__0","datatype":"FP32","shape":[1, 3, 32, 32],"data":'
    + f"{test_example}"
    + "}]}"
)
cmd = f"""curl -d '{payload}' \
   http://localhost:8003/seldon/seldon/cifar10/v2/models/cifar10/infer \
   -H "Content-Type: application/json"
"""
ret = Popen(cmd, shell=True, stdout=PIPE)
raw = ret.stdout.read().decode("utf-8")
res = json.loads(raw)
arr = np.array(res["outputs"][0]["data"])
X = X_test[idx].reshape(1, 32, 32, 3)
plt.imshow(X.reshape(32, 32, 3))
plt.axis("off")
plt.show()

print("class:", class_names[labels[idx]])
print("prediction:", class_names[arr.argmax()])

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 100 63672 100 309 100 63363 2175 435k --:--:-- --:--:-- --:--:-- 435k 100 63672 100 309 100 63363 2174 435k --:--:-- --:--:-- --:--:-- 432k

png

class: airplane
prediction: airplane

!kubectl delete -f resources/triton_pt_cifar10.yaml

seldondeployment.machinelearning.seldon.io "cifar10" deleted

Multi-Model Serving

%%writefile resources/triton_multi.yaml
apiVersion: machinelearning.seldon.io/v1
kind: SeldonDeployment
metadata:
  name: multi
  namespace: seldon
spec:
  predictors:
  - graph:
      implementation: TRITON_SERVER
      logger:
        mode: all
      modelUri: gs://seldon-models/triton/multi
      name: multi
      type: MODEL
    name: default
    replicas: 1
  protocol: v2

Writing resources/triton_multi.yaml

!kubectl apply -f resources/triton_multi.yaml

seldondeployment.machinelearning.seldon.io/multi created

!kubectl rollout status -n seldon deploy/$(kubectl get deploy -l seldon-deployment-id=multi -n seldon -o jsonpath='{.items[0].metadata.name}')

Waiting for deployment "multi-default-0-multi" rollout to finish: 0 of 1 updated replicas are available...
deployment "multi-default-0-multi" successfully rolled out

import json
from subprocess import PIPE, Popen, run

import numpy as np

idx = 1
test_example = X_test[idx : idx + 1].tolist()
payload = (
    '{"inputs":[{"name":"input_1","datatype":"FP32","shape":[1, 32, 32, 3],"data":'
    + f"{test_example}"
    + "}]}"
)
cmd = f"""curl -d '{payload}' \
   http://localhost:8003/seldon/seldon/multi/v2/models/cifar10/infer \
   -H "Content-Type: application/json"
"""
ret = Popen(cmd, shell=True, stdout=PIPE)
raw = ret.stdout.read().decode("utf-8")
res = json.loads(raw)
arr = np.array(res["outputs"][0]["data"])
X = X_test[idx].reshape(1, 32, 32, 3)
plt.imshow(X.reshape(32, 32, 3))
plt.axis("off")
plt.show()
print("class:", class_names[y_test[idx][0]])
print("prediction:", class_names[arr.argmax()])

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 63906  100   339  100 63567    752   137k --:--:-- --:--:-- --:--:--  137k

class: ship
prediction: ship

X=!curl -s -d '{"inputs":[{"name":"INPUT0","data":[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16],"datatype":"INT32","shape":[1,16]},{"name":"INPUT1","data":[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16],"datatype":"INT32","shape":[1,16]}]}'  \
        -X POST http://0.0.0.0:8003/seldon/seldon/multi/v2/models/simple/infer \
        -H "Content-Type: application/json"
d=json.loads(X[0])
print(d)
assert(d["outputs"][0]["data"][0]==2)

{'model_name': 'simple', 'model_version': '1', 'outputs': [{'name': 'OUTPUT0', 'datatype': 'INT32', 'shape': [1, 16], 'data': [2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32]}, {'name': 'OUTPUT1', 'datatype': 'INT32', 'shape': [1, 16], 'data': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]}]}

!kubectl delete -f resources/triton_multi.yaml

seldondeployment.machinelearning.seldon.io "multi" deleted