kubernetes
/
examples
mirror of https://github.com/kubernetes/examples.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #5 from volatilemolotov/ai-starter-kit-kuberay 

change kuberay docker image
This commit is contained in:
Drogovoz Dima 2025-08-21 12:28:05 +01:00 committed by GitHub
parent e8a311cec5 9026beab73
commit ecbe922183
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 533 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

474
ai/ai-starter-kit/helm-chart/ai-starter-kit/files/ray.ipynb Normal file
View File

@ -0,0 +1,474 @@
{
"cells": [
{
"cell_type": "markdown",
"id": "1c5f9954-1e94-45f4-a4a0-a02d05bd268f",
"metadata": {},
"source": [
"# MLFlow and Ray example\n",
"\n",
"In this notebook we will train an ML model and deploy it in to Ray cluster."
]
},
{
"cell_type": "markdown",
"id": "a4f320db-fb65-40b3-89b1-789711133e3d",
"metadata": {},
"source": [
"## MLFlow experiment tracking\n",
"\n",
"Here we train the ML model and log metrics using MLFlow tracking server."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "d17448f3-dc2f-4fa7-8c6c-58ec4e0f8f02",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"2025/08/12 15:00:18 INFO mlflow.tracking.fluent: Experiment with name 'Diabetes_Prediction_PyTorch' does not exist. Creating a new experiment.\n",
"2025/08/12 15:00:18 WARNING mlflow.tracking.context.registry: Encountered unexpected error during resolving tags: 'getpwuid(): uid not found: 1001'\n",
"2025/08/12 15:00:18 WARNING mlflow.utils.git_utils: Failed to import Git (the Git executable is probably not on your PATH), so Git SHA is not available. Error: Failed to initialize: Bad git executable.\n",
"The git executable must be specified in one of the following ways:\n",
" - be included in your $PATH\n",
" - be set via $GIT_PYTHON_GIT_EXECUTABLE\n",
" - explicitly set via git.refresh(<full-path-to-git-executable>)\n",
"\n",
"All git commands will error until this is rectified.\n",
"\n",
"This initial message can be silenced or aggravated in the future by setting the\n",
"$GIT_PYTHON_REFRESH environment variable. Use one of the following values:\n",
" - quiet|q|silence|s|silent|none|n|0: for no message or exception\n",
" - warn|w|warning|log|l|1: for a warning message (logging level CRITICAL, displayed by default)\n",
" - error|e|exception|raise|r|2: for a raised exception\n",
"\n",
"Example:\n",
" export GIT_PYTHON_REFRESH=quiet\n",
"\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"🏃 View run stately-kite-741 at: http://ai-starter-kit-mlflow:5000/#/experiments/1/runs/872da23f5a5541a39c0f893adbe53466\n",
"🧪 View experiment at: http://ai-starter-kit-mlflow:5000/#/experiments/1\n"
]
}
],
"source": [
"import torch\n",
"import torch.nn as nn\n",
"import torch.optim as optim\n",
"from sklearn.datasets import load_diabetes\n",
"from sklearn.model_selection import train_test_split\n",
"import numpy as np\n",
"import mlflow\n",
"import mlflow.pytorch\n",
"\n",
"# -------------------\n",
"# Prepare Data\n",
"# -------------------\n",
"mlflow.set_tracking_uri(uri=\"http://ai-starter-kit-mlflow:5000\")\n",
"data = load_diabetes()\n",
"X = data.data\n",
"y = data.target.reshape(-1, 1)\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(\n",
" X, y, test_size=0.2, random_state=42\n",
")\n",
"\n",
"X_train_tensor = torch.tensor(X_train, dtype=torch.float32)\n",
"y_train_tensor = torch.tensor(y_train, dtype=torch.float32)\n",
"X_test_tensor = torch.tensor(X_test, dtype=torch.float32)\n",
"y_test_tensor = torch.tensor(y_test, dtype=torch.float32)\n",
"\n",
"# -------------------\n",
"# Define Model\n",
"# -------------------\n",
"class RegressionModel(nn.Module):\n",
" def __init__(self, input_dim):\n",
" super().__init__()\n",
" self.linear = nn.Linear(input_dim, 1)\n",
"\n",
" def forward(self, x):\n",
" return self.linear(x)\n",
"\n",
"input_dim = X_train.shape[1]\n",
"model = RegressionModel(input_dim)\n",
"\n",
"# -------------------\n",
"# Training\n",
"# -------------------\n",
"epochs = 100\n",
"lr = 0.01\n",
"\n",
"criterion = nn.MSELoss()\n",
"optimizer = optim.Adam(model.parameters(), lr=lr)\n",
"\n",
"mlflow.set_experiment(\"Diabetes_Prediction_PyTorch\")\n",
"\n",
"with mlflow.start_run():\n",
" mlflow.log_param(\"epochs\", epochs)\n",
" mlflow.log_param(\"learning_rate\", lr)\n",
" mlflow.log_param(\"optimizer\", \"Adam\")\n",
" mlflow.log_param(\"loss_fn\", \"MSELoss\")\n",
" mlflow.log_param(\"input_features\", input_dim)\n",
"\n",
" for epoch in range(epochs):\n",
" model.train()\n",
" optimizer.zero_grad()\n",
" outputs = model(X_train_tensor)\n",
" loss = criterion(outputs, y_train_tensor)\n",
" loss.backward()\n",
" optimizer.step()\n",
"\n",
" # -------------------\n",
" # Evaluation\n",
" # -------------------\n",
" model.eval()\n",
" with torch.no_grad():\n",
" preds = model(X_test_tensor)\n",
" mse = criterion(preds, y_test_tensor).item()\n",
" rmse = np.sqrt(mse)\n",
"\n",
" mlflow.log_metric(\"mse\", mse)\n",
" mlflow.log_metric(\"rmse\", rmse)\n",
"\n",
" # # Log model to MLflow\n",
" # mlflow.pytorch.log_model(model, \"pytorch_model\")\n"
]
},
{
"cell_type": "markdown",
"id": "208f1ba2-3ea2-4487-a18d-61778d83f5ae",
"metadata": {},
"source": [
"## Ray deployment\n",
"\n",
"In this step we will use the model from the previous step to deploy it to our Ray cluster."
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "d2c6009a-991a-49e7-b345-e3e6d8cea648",
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Defaulting to user installation because normal site-packages is not writeable\n",
"Collecting ray[client,default,serve]\n",
" Downloading ray-2.48.0-cp312-cp312-manylinux2014_aarch64.whl.metadata (19 kB)\n",
"Requirement already satisfied: click>=7.0 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (8.2.1)\n",
"Requirement already satisfied: filelock in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (3.18.0)\n",
"Requirement already satisfied: jsonschema in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (4.24.0)\n",
"Collecting msgpack<2.0.0,>=1.0.0 (from ray[client,default,serve])\n",
" Downloading msgpack-1.1.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl.metadata (8.4 kB)\n",
"Requirement already satisfied: packaging in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (25.0)\n",
"Requirement already satisfied: protobuf!=3.19.5,>=3.15.3 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (6.31.1)\n",
"Requirement already satisfied: pyyaml in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (6.0.2)\n",
"Requirement already satisfied: requests in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (2.32.3)\n",
"Collecting grpcio (from ray[client,default,serve])\n",
" Downloading grpcio-1.74.0-cp312-cp312-manylinux_2_17_aarch64.whl.metadata (3.8 kB)\n",
"Requirement already satisfied: aiohttp>=3.7 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (3.12.12)\n",
"Collecting aiohttp_cors (from ray[client,default,serve])\n",
" Downloading aiohttp_cors-0.8.1-py3-none-any.whl.metadata (20 kB)\n",
"Collecting colorful (from ray[client,default,serve])\n",
" Downloading colorful-0.5.7-py2.py3-none-any.whl.metadata (16 kB)\n",
"Collecting py-spy>=0.4.0 (from ray[client,default,serve])\n",
" Downloading py_spy-0.4.1-py2.py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl.metadata (510 bytes)\n",
"Collecting opencensus (from ray[client,default,serve])\n",
" Downloading opencensus-0.11.4-py2.py3-none-any.whl.metadata (12 kB)\n",
"Requirement already satisfied: opentelemetry-sdk>=1.30.0 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (1.36.0)\n",
"Collecting opentelemetry-exporter-prometheus (from ray[client,default,serve])\n",
" Downloading opentelemetry_exporter_prometheus-0.57b0-py3-none-any.whl.metadata (1.8 kB)\n",
"Collecting opentelemetry-proto (from ray[client,default,serve])\n",
" Downloading opentelemetry_proto-1.36.0-py3-none-any.whl.metadata (2.3 kB)\n",
"Requirement already satisfied: pydantic!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,<3 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (2.11.5)\n",
"Requirement already satisfied: prometheus_client>=0.7.1 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from ray[client,default,serve]) (0.22.1)\n",
"Collecting smart_open (from ray[client,default,serve])\n",
" Downloading smart_open-7.3.0.post1-py3-none-any.whl.metadata (24 kB)\n",
"Collecting virtualenv!=20.21.1,>=20.0.24 (from ray[client,default,serve])\n",
" Downloading virtualenv-20.33.1-py3-none-any.whl.metadata (4.5 kB)\n",
"Collecting watchfiles (from ray[client,default,serve])\n",
" Downloading watchfiles-1.1.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl.metadata (4.9 kB)\n",
"Requirement already satisfied: uvicorn[standard] in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (0.35.0)\n",
"Requirement already satisfied: fastapi in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (0.116.1)\n",
"Requirement already satisfied: starlette in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from ray[client,default,serve]) (0.47.2)\n",
"Requirement already satisfied: annotated-types>=0.6.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from pydantic!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,<3->ray[client,default,serve]) (0.7.0)\n",
"Requirement already satisfied: pydantic-core==2.33.2 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from pydantic!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,<3->ray[client,default,serve]) (2.33.2)\n",
"Requirement already satisfied: typing-extensions>=4.12.2 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from pydantic!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,<3->ray[client,default,serve]) (4.14.0)\n",
"Requirement already satisfied: typing-inspection>=0.4.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from pydantic!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.*,!=2.4.*,<3->ray[client,default,serve]) (0.4.1)\n",
"Requirement already satisfied: aiohappyeyeballs>=2.5.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (2.6.1)\n",
"Requirement already satisfied: aiosignal>=1.1.2 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (1.3.2)\n",
"Requirement already satisfied: attrs>=17.3.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (25.3.0)\n",
"Requirement already satisfied: frozenlist>=1.1.1 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (1.7.0)\n",
"Requirement already satisfied: multidict<7.0,>=4.5 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (6.4.4)\n",
"Requirement already satisfied: propcache>=0.2.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (0.3.2)\n",
"Requirement already satisfied: yarl<2.0,>=1.17.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from aiohttp>=3.7->ray[client,default,serve]) (1.20.1)\n",
"Requirement already satisfied: idna>=2.0 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from yarl<2.0,>=1.17.0->aiohttp>=3.7->ray[client,default,serve]) (3.10)\n",
"Requirement already satisfied: opentelemetry-api==1.36.0 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from opentelemetry-sdk>=1.30.0->ray[client,default,serve]) (1.36.0)\n",
"Requirement already satisfied: opentelemetry-semantic-conventions==0.57b0 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from opentelemetry-sdk>=1.30.0->ray[client,default,serve]) (0.57b0)\n",
"Requirement already satisfied: importlib-metadata<8.8.0,>=6.0 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from opentelemetry-api==1.36.0->opentelemetry-sdk>=1.30.0->ray[client,default,serve]) (8.7.0)\n",
"Requirement already satisfied: zipp>=3.20 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from importlib-metadata<8.8.0,>=6.0->opentelemetry-api==1.36.0->opentelemetry-sdk>=1.30.0->ray[client,default,serve]) (3.23.0)\n",
"Collecting distlib<1,>=0.3.7 (from virtualenv!=20.21.1,>=20.0.24->ray[client,default,serve])\n",
" Downloading distlib-0.4.0-py2.py3-none-any.whl.metadata (5.2 kB)\n",
"Requirement already satisfied: platformdirs<5,>=3.9.1 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from virtualenv!=20.21.1,>=20.0.24->ray[client,default,serve]) (4.3.8)\n",
"Requirement already satisfied: anyio<5,>=3.6.2 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from starlette->ray[client,default,serve]) (4.9.0)\n",
"Requirement already satisfied: sniffio>=1.1 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from anyio<5,>=3.6.2->starlette->ray[client,default,serve]) (1.3.1)\n",
"Requirement already satisfied: jsonschema-specifications>=2023.03.6 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from jsonschema->ray[client,default,serve]) (2025.4.1)\n",
"Requirement already satisfied: referencing>=0.28.4 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from jsonschema->ray[client,default,serve]) (0.36.2)\n",
"Requirement already satisfied: rpds-py>=0.7.1 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from jsonschema->ray[client,default,serve]) (0.25.1)\n",
"Collecting opencensus-context>=0.1.3 (from opencensus->ray[client,default,serve])\n",
" Downloading opencensus_context-0.1.3-py2.py3-none-any.whl.metadata (3.3 kB)\n",
"Requirement already satisfied: six~=1.16 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from opencensus->ray[client,default,serve]) (1.17.0)\n",
"Collecting google-api-core<3.0.0,>=1.0.0 (from opencensus->ray[client,default,serve])\n",
" Downloading google_api_core-2.25.1-py3-none-any.whl.metadata (3.0 kB)\n",
"Collecting googleapis-common-protos<2.0.0,>=1.56.2 (from google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve])\n",
" Downloading googleapis_common_protos-1.70.0-py3-none-any.whl.metadata (9.3 kB)\n",
"Collecting proto-plus<2.0.0,>=1.22.3 (from google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve])\n",
" Downloading proto_plus-1.26.1-py3-none-any.whl.metadata (2.2 kB)\n",
"Requirement already satisfied: google-auth<3.0.0,>=2.14.1 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve]) (2.40.3)\n",
"Requirement already satisfied: cachetools<6.0,>=2.0.0 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from google-auth<3.0.0,>=2.14.1->google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve]) (5.5.2)\n",
"Requirement already satisfied: pyasn1-modules>=0.2.1 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from google-auth<3.0.0,>=2.14.1->google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve]) (0.4.2)\n",
"Requirement already satisfied: rsa<5,>=3.1.4 in /opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages (from google-auth<3.0.0,>=2.14.1->google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve]) (4.9.1)\n",
"Requirement already satisfied: charset_normalizer<4,>=2 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from requests->ray[client,default,serve]) (3.4.2)\n",
"Requirement already satisfied: urllib3<3,>=1.21.1 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from requests->ray[client,default,serve]) (2.4.0)\n",
"Requirement already satisfied: certifi>=2017.4.17 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from requests->ray[client,default,serve]) (2025.4.26)\n",
"Requirement already satisfied: pyasn1>=0.1.3 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from rsa<5,>=3.1.4->google-auth<3.0.0,>=2.14.1->google-api-core<3.0.0,>=1.0.0->opencensus->ray[client,default,serve]) (0.6.1)\n",
"Collecting wrapt (from smart_open->ray[client,default,serve])\n",
" Downloading wrapt-1.17.3-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl.metadata (6.4 kB)\n",
"Requirement already satisfied: h11>=0.8 in /opt/bitnami/miniforge/lib/python3.12/site-packages (from uvicorn[standard]; extra == \"serve\"->ray[client,default,serve]) (0.16.0)\n",
"Collecting httptools>=0.6.3 (from uvicorn[standard]; extra == \"serve\"->ray[client,default,serve])\n",
" Downloading httptools-0.6.4-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl.metadata (3.6 kB)\n",
"Collecting python-dotenv>=0.13 (from uvicorn[standard]; extra == \"serve\"->ray[client,default,serve])\n",
" Downloading python_dotenv-1.1.1-py3-none-any.whl.metadata (24 kB)\n",
"Collecting uvloop>=0.15.1 (from uvicorn[standard]; extra == \"serve\"->ray[client,default,serve])\n",
" Downloading uvloop-0.21.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl.metadata (4.9 kB)\n",
"Collecting websockets>=10.4 (from uvicorn[standard]; extra == \"serve\"->ray[client,default,serve])\n",
" Downloading websockets-15.0.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl.metadata (6.8 kB)\n",
"Downloading ray-2.48.0-cp312-cp312-manylinux2014_aarch64.whl (69.2 MB)\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m69.2/69.2 MB\u001b[0m \u001b[31m36.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
"Downloading msgpack-1.1.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl (425 kB)\n",
"Downloading grpcio-1.74.0-cp312-cp312-manylinux_2_17_aarch64.whl (5.9 MB)\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m5.9/5.9 MB\u001b[0m \u001b[31m38.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
"\u001b[?25hDownloading py_spy-0.4.1-py2.py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl (2.0 MB)\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.0/2.0 MB\u001b[0m \u001b[31m37.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
"\u001b[?25hDownloading virtualenv-20.33.1-py3-none-any.whl (6.1 MB)\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m6.1/6.1 MB\u001b[0m \u001b[31m40.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
"Downloading distlib-0.4.0-py2.py3-none-any.whl (469 kB)\n",
"Downloading aiohttp_cors-0.8.1-py3-none-any.whl (25 kB)\n",
"Downloading colorful-0.5.7-py2.py3-none-any.whl (201 kB)\n",
"Downloading opencensus-0.11.4-py2.py3-none-any.whl (128 kB)\n",
"Downloading google_api_core-2.25.1-py3-none-any.whl (160 kB)\n",
"Downloading googleapis_common_protos-1.70.0-py3-none-any.whl (294 kB)\n",
"Downloading proto_plus-1.26.1-py3-none-any.whl (50 kB)\n",
"Downloading opencensus_context-0.1.3-py2.py3-none-any.whl (5.1 kB)\n",
"Downloading opentelemetry_exporter_prometheus-0.57b0-py3-none-any.whl (12 kB)\n",
"Downloading opentelemetry_proto-1.36.0-py3-none-any.whl (72 kB)\n",
"Downloading smart_open-7.3.0.post1-py3-none-any.whl (61 kB)\n",
"Downloading httptools-0.6.4-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl (508 kB)\n",
"Downloading python_dotenv-1.1.1-py3-none-any.whl (20 kB)\n",
"Downloading uvloop-0.21.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl (4.6 MB)\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m4.6/4.6 MB\u001b[0m \u001b[31m32.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
"\u001b[?25hDownloading watchfiles-1.1.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl (450 kB)\n",
"Downloading websockets-15.0.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl (183 kB)\n",
"Downloading wrapt-1.17.3-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl (88 kB)\n",
"Installing collected packages: py-spy, opencensus-context, distlib, colorful, wrapt, websockets, virtualenv, uvloop, python-dotenv, proto-plus, opentelemetry-proto, msgpack, httptools, grpcio, googleapis-common-protos, watchfiles, smart_open, google-api-core, aiohttp_cors, ray, opencensus, opentelemetry-exporter-prometheus\n",
"\u001b[33m WARNING: The script websockets is installed in '/opt/bitnami/jupyterhub-singleuser/.local/bin' which is not on PATH.\n",
" Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\u001b[33m\n",
"\u001b[33m WARNING: The script virtualenv is installed in '/opt/bitnami/jupyterhub-singleuser/.local/bin' which is not on PATH.\n",
" Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\u001b[33m\n",
"\u001b[33m WARNING: The script dotenv is installed in '/opt/bitnami/jupyterhub-singleuser/.local/bin' which is not on PATH.\n",
" Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\u001b[33m\n",
"\u001b[33m WARNING: The script watchfiles is installed in '/opt/bitnami/jupyterhub-singleuser/.local/bin' which is not on PATH.\n",
" Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\u001b[33m\n",
"\u001b[33m WARNING: The scripts ray, serve and tune are installed in '/opt/bitnami/jupyterhub-singleuser/.local/bin' which is not on PATH.\n",
" Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\u001b[33m\n",
"\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m22/22\u001b[0m [opentelemetry-exporter-prometheus]\n",
"\u001b[1A\u001b[2KSuccessfully installed aiohttp_cors-0.8.1 colorful-0.5.7 distlib-0.4.0 google-api-core-2.25.1 googleapis-common-protos-1.70.0 grpcio-1.74.0 httptools-0.6.4 msgpack-1.1.1 opencensus-0.11.4 opencensus-context-0.1.3 opentelemetry-exporter-prometheus-0.57b0 opentelemetry-proto-1.36.0 proto-plus-1.26.1 py-spy-0.4.1 python-dotenv-1.1.1 ray-2.48.0 smart_open-7.3.0.post1 uvloop-0.21.0 virtualenv-20.33.1 watchfiles-1.1.0 websockets-15.0.1 wrapt-1.17.3\n"
]
}
],
"source": [
"!pip install \"ray[serve,client,default]\""
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "5be0b7fa-0815-4224-a4e6-ad5b74647e20",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"2025-08-12 15:00:37,849\tINFO worker.py:1606 -- Using address ray://ai-starter-kit-kuberay-head-svc:10001 set in the environment variable RAY_ADDRESS\n",
"2025-08-12 15:00:37,850\tINFO client_builder.py:242 -- Passing the following kwargs to ray.init() on the server: log_to_driver\n",
"SIGTERM handler is not set because current thread is not the main thread.\n",
"2025-08-12 15:00:39,818\tWARNING utils.py:1280 -- Python patch version mismatch: The cluster was started with:\n",
" Ray: 2.48.0\n",
" Python: 3.12.9\n",
"This process on Ray Client was started with:\n",
" Ray: 2.48.0\n",
" Python: 3.12.10\n",
"\n",
"\u001b[36m(ProxyActor pid=3272)\u001b[0m INFO 2025-08-12 08:00:44,231 proxy 10.244.0.9 -- Proxy starting on node 66724a5e2332cd618965646d0b7ab0d4d89622990053ab677dc8f588 (HTTP port: 8000).\n",
"\u001b[36m(ProxyActor pid=3272)\u001b[0m INFO 2025-08-12 08:00:44,333 proxy 10.244.0.9 -- Got updated endpoints: {}.\n",
"INFO 2025-08-12 15:00:44,830 serve 72 -- Started Serve in namespace \"serve\".\n",
"\u001b[36m(ServeController pid=3174)\u001b[0m INFO 2025-08-12 08:00:46,387 controller 3174 -- Deploying new version of Deployment(name='PyTorchMLflowDeployment', app='default') (initial target replicas: 1).\n",
"\u001b[36m(ProxyActor pid=3272)\u001b[0m INFO 2025-08-12 08:00:46,396 proxy 10.244.0.9 -- Got updated endpoints: {Deployment(name='PyTorchMLflowDeployment', app='default'): EndpointInfo(route='/predict', app_is_cross_language=False)}.\n",
"\u001b[36m(ProxyActor pid=3272)\u001b[0m INFO 2025-08-12 08:00:46,423 proxy 10.244.0.9 -- Started <ray.serve._private.router.SharedRouterLongPollClient object at 0xffff4c01e4e0>.\n",
"\u001b[36m(ServeController pid=3174)\u001b[0m INFO 2025-08-12 08:00:46,503 controller 3174 -- Adding 1 replica to Deployment(name='PyTorchMLflowDeployment', app='default').\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[36m(ServeReplica:default:PyTorchMLflowDeployment pid=1940, ip=10.244.0.10)\u001b[0m Loading model from MLflow...\n",
"\u001b[36m(ServeReplica:default:PyTorchMLflowDeployment pid=1940, ip=10.244.0.10)\u001b[0m Model loaded successfully.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"INFO 2025-08-12 15:00:50,456 serve 72 -- Application 'default' is ready at http://127.0.0.1:8000/predict.\n",
"INFO 2025-08-12 15:00:50,519 serve 72 -- Started <ray.serve._private.router.SharedRouterLongPollClient object at 0xffff704638f0>.\n"
]
},
{
"data": {
"text/plain": [
"DeploymentHandle(deployment='PyTorchMLflowDeployment')"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"WARNING 2025-08-12 15:00:50,695 serve 72 -- Failed to get queue length from Replica(id='7iflc5wc', deployment='PyTorchMLflowDeployment', app='default') within 0.1s. If this happens repeatedly it's likely caused by high network latency in the cluster. You can configure the deadline using the `RAY_SERVE_QUEUE_LENGTH_RESPONSE_DEADLINE_S` environment variable.\n",
"2025-08-12 15:00:50,725\tERROR dataclient.py:312 -- Callback error:\n",
"Traceback (most recent call last):\n",
" File \"/opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages/ray/util/client/dataclient.py\", line 301, in _process_response\n",
" can_remove = callback(response)\n",
" ^^^^^^^^^^^^^^^^^^\n",
" File \"/opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages/ray/util/client/dataclient.py\", line 179, in __call__\n",
" self.callback(self.data)\n",
" File \"/opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages/ray/util/client/common.py\", line 179, in deserialize_obj\n",
" py_callback(data)\n",
" File \"/opt/bitnami/jupyterhub-singleuser/.local/lib/python3.12/site-packages/ray/util/client/common.py\", line 147, in set_future\n",
" fut.set_result(data)\n",
" File \"/opt/bitnami/miniforge/lib/python3.12/concurrent/futures/_base.py\", line 544, in set_result\n",
" raise InvalidStateError('{}: {!r}'.format(self._state, self))\n",
"concurrent.futures._base.InvalidStateError: CANCELLED: <Future at 0xffff7065c740 state=cancelled>\n",
"\u001b[36m(ProxyActor pid=1993, ip=10.244.0.10)\u001b[0m INFO 2025-08-12 08:00:51,185 proxy 10.244.0.10 -- Proxy starting on node 1d5e87519ede5427735866024d720fbcd019925255178b592dd245a3 (HTTP port: 8000).\n",
"\u001b[36m(ProxyActor pid=1993, ip=10.244.0.10)\u001b[0m INFO 2025-08-12 08:00:51,223 proxy 10.244.0.10 -- Got updated endpoints: {Deployment(name='PyTorchMLflowDeployment', app='default'): EndpointInfo(route='/predict', app_is_cross_language=False)}.\n",
"\u001b[36m(ProxyActor pid=1993, ip=10.244.0.10)\u001b[0m INFO 2025-08-12 08:00:51,230 proxy 10.244.0.10 -- Started <ray.serve._private.router.SharedRouterLongPollClient object at 0xffff782785c0>.\n",
"\u001b[36m(ServeReplica:default:PyTorchMLflowDeployment pid=1940, ip=10.244.0.10)\u001b[0m INFO 2025-08-12 08:01:29,315 default_PyTorchMLflowDeployment 7iflc5wc 4d7cb9a3-7b88-40e3-ab03-d933717e195a -- POST /predict 200 40.5ms\n",
"\u001b[36m(ServeController pid=3174)\u001b[0m INFO 2025-08-12 08:01:38,376 controller 3174 -- Removing 1 replica from Deployment(name='PyTorchMLflowDeployment', app='default').\n",
"\u001b[36m(ServeController pid=3174)\u001b[0m INFO 2025-08-12 08:01:40,427 controller 3174 -- Replica(id='7iflc5wc', deployment='PyTorchMLflowDeployment', app='default') is stopped.\n"
]
}
],
"source": [
"import torch\n",
"import mlflow.pytorch\n",
"import numpy as np\n",
"from starlette.requests import Request\n",
"from typing import Dict\n",
"\n",
"from ray import serve\n",
"import ray\n",
"\n",
"# ray.init(\"ray://ai-starter-kit-kuberay-head-svc:10001\", namespace=\"my_new_namespace\")\n",
"\n",
"# MLFLOW_MODEL_URI = \"mlruns/0/<RUN_ID>/artifacts/pytorch_model\" # Change to your run path\n",
"\n",
"@serve.deployment\n",
"class PyTorchMLflowDeployment:\n",
" def __init__(self):\n",
" print(\"Loading model from MLflow...\")\n",
" # # self.model = mlflow.pytorch.load_model(MLFLOW_MODEL_URI)\n",
" self.model = model\n",
" self.model.eval()\n",
" print(\"Model loaded successfully.\")\n",
"\n",
" async def __call__(self, request: Request) -> Dict:\n",
" try:\n",
" data = await request.json()\n",
" features = data.get(\"features\", None)\n",
" if features is None:\n",
" return {\"error\": \"Missing 'features' in request\"}\n",
"\n",
" X = np.array(features).reshape(1, -1)\n",
" X_tensor = torch.tensor(X, dtype=torch.float32)\n",
"\n",
" with torch.no_grad():\n",
" prediction = self.model(X_tensor).numpy().tolist()\n",
"\n",
" return {\"prediction\": prediction}\n",
" except Exception as e:\n",
" return {\"error\": str(e)}\n",
"\n",
"app = PyTorchMLflowDeployment.bind()\n",
"serve.run(app, route_prefix=\"/predict\")\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "1e29a0ae-7514-4881-ae81-99b4e01e4e1f",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"INFO 2025-08-12 15:01:38,215 serve 72 -- Deleting app ['PyTorchMLflowDeployment']\n"
]
}
],
"source": [
"serve.delete(\"PyTorchMLflowDeployment\")\n",
"serve.shutdown()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.10"
}
},
"nbformat": 4,
"nbformat_minor": 5
}

14
ai/ai-starter-kit/helm-chart/ai-starter-kit/templates/configmaps.yaml
View File

@ -38,4 +38,16 @@ metadata:
helm.sh/chart: "{{ .Chart.Name }}-{{ .Chart.Version }}" helm.sh/chart: "{{ .Chart.Name }}-{{ .Chart.Version }}"
data: data:
welcome.ipynb: |- welcome.ipynb: |-
{{ .Files.Get "files/welcome.ipynb" | nindent 4 }} {{ .Files.Get "files/welcome.ipynb" | nindent 4 }}
---
apiVersion: v1
kind: ConfigMap
metadata:
name: {{ .Release.Name }}-ray-notebook
labels:
app.kubernetes.io/managed-by: {{ .Release.Service | quote }}
app.kubernetes.io/instance: {{ .Release.Name | quote }}
helm.sh/chart: "{{ .Chart.Name }}-{{ .Chart.Version }}"
data:
ray.ipynb: |-
{{ .Files.Get "files/ray.ipynb" | nindent 4 }}

56
ai/ai-starter-kit/helm-chart/ai-starter-kit/values.yaml
View File

@ -55,6 +55,9 @@ jupyterhub:
- name: welcome-notebook - name: welcome-notebook
configMap: configMap:
name: "{{ .Release.Name }}-welcome-notebook" name: "{{ .Release.Name }}-welcome-notebook"
- name: ray-notebook
configMap:
name: "{{ .Release.Name }}-ray-notebook"
- name: hf-token-secret - name: hf-token-secret
secret: secret:
secretName: "{{ .Release.Name }}-hf-token-secret" secretName: "{{ .Release.Name }}-hf-token-secret"
@ -71,11 +74,14 @@ jupyterhub:
- name: welcome-notebook - name: welcome-notebook
mountPath: /tmp/welcome.ipynb mountPath: /tmp/welcome.ipynb
subPath: welcome.ipynb subPath: welcome.ipynb
- name: ray-notebook
mountPath: /tmp/ray.ipynb
subPath: ray.ipynb
- name: hf-token-secret - name: hf-token-secret
mountPath: "/etc/secrets/huggingface" mountPath: "/etc/secrets/huggingface"
readOnly: true readOnly: true
extraEnvVars: extraEnvVars:
RAY_ADDRESS: "{{ .Release.Name }}-kuberay-head-svc:6379" RAY_ADDRESS: "ray://{{ .Release.Name }}-kuberay-head-svc:10001"
MLFLOW_TRACKING_URI: "http://{{ .Release.Name }}-mlflow-tracking" MLFLOW_TRACKING_URI: "http://{{ .Release.Name }}-mlflow-tracking"
# JUPYTERLAB_DIR: "/opt" # JUPYTERLAB_DIR: "/opt"
resources: resources:
@ -85,18 +91,48 @@ jupyterhub:
memory: 4Gi memory: 4Gi
hub: hub:
password: "sneakypass" password: "sneakypass"
extraEnvVars:
- name: "RAY_ADDRESS"
value: "{{ .Release.Name }}-kuberay-head-svc"
- name: "MLFLOW_TRACKING_URI"
value: "http://{{ .Release.Name }}-mlflow-tracking"
ray-cluster:
head:
serviceType: ClusterIP
resources: resources:
limits: limits:
memory: 512Mi
requests:
memory: 512Mi
# extraEnvVars:
# - name: "RAY_ADDRESS"
# value: "{{ .Release.Name }}-kuberay-head-svc"
# - name: "MLFLOW_TRACKING_URI"
# value: "http://{{ .Release.Name }}-mlflow-tracking"
ray-cluster:
image:
tag: 2.48.0.2c63f6-py312-cpu-aarch64
head:
serviceType: ClusterIP
command: ["pip", "install", "ray[serve,default,client]", "transformers", "torch", "torchvision", "accelerate", "scikit-learn"]
resources:
requests:
cpu: "1"
memory: "2G"
limits:
cpu: "4"
memory: "8G" memory: "8G"
worker:
command: ["pip", "install", "ray[serve,default,client]", "transformers", "torch", "torchvision", "accelerate", "scikit-learn"]
resources:
requests:
cpu: "1"
memory: "2G"
limits:
cpu: "4"
memory: "8G"
mlflow:
resources:
requests:
cpu: 100m
memory: 512Mi
limits:
cpu: 1000m
memory: 1Gi
huggingface: huggingface:
# Provide your Hugging Face token here to download gated or private models. # Provide your Hugging Face token here to download gated or private models.