MLOps完全ガイド: MLパイプラインから本番デプロイまで

データバージョン管理: DVC, LakeFS, Delta Lake
実験トラッキング: MLflow, W&B, Neptune, Comet ML
パイプライン: Airflow, Prefect, Metaflow, Kubeflow Pipelines
モデルレジストリ: MLflow Registry, W&B Artifacts, Vertex AI
コンテナ化: Docker, Podman
オーケストレーション: Kubernetes, ECS, GKE
モデルサービング: Triton, TorchServe, BentoML, KServe
モニタリング: Evidently, WhyLogs, Arize, Fiddler
フィーチャーストア: Feast, Tecton, Vertex AI Feature Store

# DVCのインストール
pip install dvc

# S3サポート付きインストール
pip install "dvc[s3]"

# GCSサポート付きインストール
pip install "dvc[gs]"

# 全リモートサポート付きインストール
pip install "dvc[all]"

# Gitリポジトリの初期化（必要な場合）
git init

# DVCの初期化
dvc init

# 作成されたファイルの確認
ls .dvc/
# config  .gitignore  tmp/

# データのトラッキング開始
dvc add data/train.csv

# .dvcファイルが作成される（Gitでトラッキング）
git add data/train.csv.dvc data/.gitignore
git commit -m "Add training data"

# S3リモートストレージの設定
dvc remote add -d myremote s3://my-bucket/dvc-store

# AWS認証情報の設定（環境変数または~/.aws/credentials）
dvc remote modify myremote access_key_id YOUR_ACCESS_KEY
dvc remote modify myremote secret_access_key YOUR_SECRET_KEY

# GCSリモートストレージ
dvc remote add -d gcsstorage gs://my-bucket/dvc-store

# ローカルリモートストレージ（テスト用）
dvc remote add -d localremote /tmp/dvc-storage

# データのプッシュ
dvc push

# データのプル
dvc pull

# dvc.yaml
stages:
  prepare:
    cmd: python src/prepare.py
    deps:
      - src/prepare.py
      - data/raw
    outs:
      - data/prepared

  featurize:
    cmd: python src/featurize.py
    deps:
      - src/featurize.py
      - data/prepared
    outs:
      - data/features

  train:
    cmd: python src/train.py
    deps:
      - src/train.py
      - data/features
    params:
      - params.yaml:
          - train.lr
          - train.n_estimators
    outs:
      - models/model.pkl
    metrics:
      - metrics/scores.json:
          cache: false

  evaluate:
    cmd: python src/evaluate.py
    deps:
      - src/evaluate.py
      - models/model.pkl
      - data/features
    metrics:
      - metrics/eval.json:
          cache: false

# パイプラインの実行
dvc repro

# 特定のステージまで実行
dvc repro train

# パイプラインDAGの可視化
dvc dag

# 実験結果の比較
dvc metrics show
dvc metrics diff HEAD~1

# params.yaml
train:
  lr: 0.001
  n_estimators: 100
  max_depth: 5
  batch_size: 32
  epochs: 10

data:
  test_size: 0.2
  random_state: 42

# src/train.py
import yaml

# パラメータの読み込み
with open("params.yaml") as f:
    params = yaml.safe_load(f)

lr = params["train"]["lr"]
n_estimators = params["train"]["n_estimators"]

# モデルの学習
from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(
    n_estimators=n_estimators,
    max_depth=params["train"]["max_depth"],
    random_state=params["data"]["random_state"]
)

# 新しい実験ブランチの作成
git checkout -b experiment/increase-lr

# パラメータの変更
# params.yamlで lr: 0.001 を lr: 0.01 に変更

# パイプラインの再実行
dvc repro

# 結果の確認
dvc metrics show

# コミット
git add dvc.lock params.yaml
git commit -m "Experiment: increase learning rate to 0.01"

# 実験の比較
git checkout main
dvc metrics diff experiment/increase-lr

# MLflowのインストール
pip install mlflow

# MLflow UIの起動
mlflow ui

# ホストとポートの指定
mlflow ui --host 0.0.0.0 --port 5001

# リモートトラッキングサーバーの設定
export MLFLOW_TRACKING_URI=http://mlflow-server:5000

import mlflow
import mlflow.sklearn
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, f1_score
import numpy as np

# トラッキングURIの設定（デフォルト: ./mlruns）
mlflow.set_tracking_uri("http://localhost:5000")

# 実験の設定
mlflow.set_experiment("my-classification-experiment")

# 実験の実行
with mlflow.start_run(run_name="random-forest-v1"):
    # パラメータのログ
    n_estimators = 100
    max_depth = 5
    lr = 0.001

    mlflow.log_param("n_estimators", n_estimators)
    mlflow.log_param("max_depth", max_depth)
    mlflow.log_param("learning_rate", lr)

    # モデルの学習
    X_train, X_test, y_train, y_test = train_test_split(
        X, y, test_size=0.2, random_state=42
    )

    model = RandomForestClassifier(
        n_estimators=n_estimators,
        max_depth=max_depth,
        random_state=42
    )
    model.fit(X_train, y_train)

    # メトリクスのログ
    y_pred = model.predict(X_test)
    accuracy = accuracy_score(y_test, y_pred)
    f1 = f1_score(y_test, y_pred, average='weighted')

    mlflow.log_metric("accuracy", accuracy)
    mlflow.log_metric("f1_score", f1)

    # ステップごとのメトリクス（学習曲線）
    for epoch in range(10):
        train_loss = np.random.random() * 0.5 / (epoch + 1)
        mlflow.log_metric("train_loss", train_loss, step=epoch)

    # アーティファクトのログ
    mlflow.log_artifact("data/train.csv")

    # モデルのログ
    mlflow.sklearn.log_model(
        model,
        "model",
        registered_model_name="RandomForestClassifier"
    )

    print(f"Accuracy: {accuracy:.4f}, F1: {f1:.4f}")
    print(f"Run ID: {mlflow.active_run().info.run_id}")

# sklearnの自動ログを有効化
mlflow.sklearn.autolog()

with mlflow.start_run():
    model = RandomForestClassifier(n_estimators=100)
    model.fit(X_train, y_train)
    # パラメータ、メトリクス、モデルが自動的にログされる

# PyTorchの自動ログ
mlflow.pytorch.autolog()

# XGBoostの自動ログ
mlflow.xgboost.autolog()

import torch
import torch.nn as nn
import torch.optim as optim
import mlflow
import mlflow.pytorch

class SimpleNet(nn.Module):
    def __init__(self, input_dim, hidden_dim, output_dim):
        super().__init__()
        self.fc1 = nn.Linear(input_dim, hidden_dim)
        self.fc2 = nn.Linear(hidden_dim, output_dim)
        self.relu = nn.ReLU()
        self.dropout = nn.Dropout(0.3)

    def forward(self, x):
        x = self.relu(self.fc1(x))
        x = self.dropout(x)
        return self.fc2(x)

def train_with_mlflow(
    model, train_loader, val_loader,
    epochs=10, lr=0.001, experiment_name="pytorch-training"
):
    mlflow.set_experiment(experiment_name)

    with mlflow.start_run():
        # ハイパーパラメータのログ
        mlflow.log_params({
            "epochs": epochs,
            "learning_rate": lr,
            "model_type": "SimpleNet",
            "optimizer": "Adam",
            "batch_size": train_loader.batch_size,
        })

        optimizer = optim.Adam(model.parameters(), lr=lr)
        criterion = nn.CrossEntropyLoss()
        scheduler = optim.lr_scheduler.StepLR(optimizer, step_size=5, gamma=0.5)

        for epoch in range(epochs):
            # 学習
            model.train()
            train_loss = 0.0
            for batch_X, batch_y in train_loader:
                optimizer.zero_grad()
                outputs = model(batch_X)
                loss = criterion(outputs, batch_y)
                loss.backward()
                optimizer.step()
                train_loss += loss.item()

            train_loss /= len(train_loader)

            # 検証
            model.eval()
            val_loss = 0.0
            correct = 0
            total = 0

            with torch.no_grad():
                for batch_X, batch_y in val_loader:
                    outputs = model(batch_X)
                    loss = criterion(outputs, batch_y)
                    val_loss += loss.item()

                    _, predicted = outputs.max(1)
                    total += batch_y.size(0)
                    correct += predicted.eq(batch_y).sum().item()

            val_loss /= len(val_loader)
            val_accuracy = correct / total

            # ステップごとのメトリクスログ
            mlflow.log_metrics({
                "train_loss": train_loss,
                "val_loss": val_loss,
                "val_accuracy": val_accuracy,
                "learning_rate": scheduler.get_last_lr()[0],
            }, step=epoch)

            scheduler.step()

            print(f"Epoch {epoch+1}/{epochs} | "
                  f"Train Loss: {train_loss:.4f} | "
                  f"Val Loss: {val_loss:.4f} | "
                  f"Val Acc: {val_accuracy:.4f}")

        # 最終モデルの保存
        mlflow.pytorch.log_model(model, "model")

        return mlflow.active_run().info.run_id

# W&Bのインストール
pip install wandb

# ログイン
wandb login

import wandb
import torch

# W&Bの初期化
wandb.init(
    project="my-ml-project",
    name="experiment-001",
    config={
        "learning_rate": 0.001,
        "epochs": 100,
        "batch_size": 64,
        "architecture": "ResNet50",
        "dataset": "ImageNet",
    }
)

# 設定値へのアクセス
lr = wandb.config.learning_rate

# 学習ループ
for epoch in range(wandb.config.epochs):
    train_loss, train_acc = train_epoch(model, train_loader)
    val_loss, val_acc = eval_epoch(model, val_loader)

    # メトリクスのログ
    wandb.log({
        "epoch": epoch,
        "train_loss": train_loss,
        "train_accuracy": train_acc,
        "val_loss": val_loss,
        "val_accuracy": val_acc,
        "learning_rate": optimizer.param_groups[0]['lr'],
    })

# モデルの保存
wandb.save("model.pth")

# 画像のログ
wandb.log({
    "predictions": [
        wandb.Image(img, caption=f"Pred: {pred}, True: {true}")
        for img, pred, true in sample_predictions
    ]
})

# W&B Artifacts（データセットのバージョン管理）
artifact = wandb.Artifact("training-data", type="dataset")
artifact.add_dir("data/train")
wandb.log_artifact(artifact)

wandb.finish()

import wandb

# スイープ設定
sweep_config = {
    "method": "bayes",  # random, grid, bayes
    "metric": {
        "name": "val_accuracy",
        "goal": "maximize"
    },
    "parameters": {
        "learning_rate": {
            "distribution": "log_uniform_values",
            "min": 1e-5,
            "max": 1e-1,
        },
        "batch_size": {
            "values": [16, 32, 64, 128]
        },
        "hidden_dim": {
            "values": [64, 128, 256, 512]
        },
        "dropout": {
            "distribution": "uniform",
            "min": 0.1,
            "max": 0.5,
        }
    }
}

def train_sweep():
    with wandb.init() as run:
        config = wandb.config

        model = SimpleNet(
            input_dim=784,
            hidden_dim=config.hidden_dim,
            output_dim=10
        )

        optimizer = torch.optim.Adam(
            model.parameters(),
            lr=config.learning_rate
        )

        for epoch in range(10):
            train_loss, val_loss, val_acc = run_epoch(
                model, optimizer, config.batch_size
            )
            wandb.log({
                "train_loss": train_loss,
                "val_loss": val_loss,
                "val_accuracy": val_acc,
            })

# スイープの作成と実行
sweep_id = wandb.sweep(sweep_config, project="my-project")
wandb.agent(sweep_id, function=train_sweep, count=50)

# dags/ml_pipeline.py
from airflow import DAG
from airflow.operators.python import PythonOperator
from datetime import datetime, timedelta

default_args = {
    'owner': 'mlops-team',
    'depends_on_past': False,
    'start_date': datetime(2026, 1, 1),
    'email_on_failure': True,
    'retries': 1,
    'retry_delay': timedelta(minutes=5),
}

dag = DAG(
    'ml_training_pipeline',
    default_args=default_args,
    description='ML model training pipeline',
    schedule_interval='0 2 * * *',  # 毎日午前2時
    catchup=False,
)

def extract_data(**kwargs):
    """データの抽出"""
    import pandas as pd
    from sqlalchemy import create_engine

    engine = create_engine('postgresql://user:pass@db:5432/mldb')
    df = pd.read_sql(
        "SELECT * FROM features WHERE date >= CURRENT_DATE - 7",
        engine
    )

    output_path = '/tmp/raw_data.parquet'
    df.to_parquet(output_path)
    kwargs['ti'].xcom_push(key='data_path', value=output_path)
    return output_path

def preprocess_data(**kwargs):
    """データの前処理"""
    ti = kwargs['ti']
    data_path = ti.xcom_pull(task_ids='extract_data', key='data_path')

    import pandas as pd
    from sklearn.preprocessing import StandardScaler

    df = pd.read_parquet(data_path)
    scaler = StandardScaler()
    features = df.drop('target', axis=1)
    scaled_features = scaler.fit_transform(features)

    processed_path = '/tmp/processed_data.parquet'
    pd.DataFrame(scaled_features).to_parquet(processed_path)
    ti.xcom_push(key='processed_path', value=processed_path)

def train_model(**kwargs):
    """モデルの学習"""
    ti = kwargs['ti']
    processed_path = ti.xcom_pull(
        task_ids='preprocess_data',
        key='processed_path'
    )

    import mlflow
    import pandas as pd
    from sklearn.ensemble import GradientBoostingClassifier
    from sklearn.model_selection import train_test_split

    mlflow.set_experiment("production-training")

    with mlflow.start_run():
        data = pd.read_parquet(processed_path)
        X = data.drop('target', axis=1)
        y = data['target']
        X_train, X_test, y_train, y_test = train_test_split(X, y)

        model = GradientBoostingClassifier(n_estimators=200)
        model.fit(X_train, y_train)

        accuracy = model.score(X_test, y_test)
        mlflow.log_metric("accuracy", accuracy)
        mlflow.sklearn.log_model(model, "model")

        run_id = mlflow.active_run().info.run_id
        ti.xcom_push(key='run_id', value=run_id)

def evaluate_and_deploy(**kwargs):
    """モデルの評価とデプロイ"""
    ti = kwargs['ti']
    run_id = ti.xcom_pull(task_ids='train_model', key='run_id')

    import mlflow
    client = mlflow.tracking.MlflowClient()

    run = client.get_run(run_id)
    accuracy = run.data.metrics['accuracy']

    if accuracy >= 0.90:
        model_uri = f"runs:/{run_id}/model"
        mlflow.register_model(model_uri, "ProductionModel")
        print(f"Model deployed with accuracy: {accuracy:.4f}")
    else:
        raise ValueError(f"Model accuracy {accuracy:.4f} below threshold 0.90")

# タスクの定義
extract_task = PythonOperator(
    task_id='extract_data',
    python_callable=extract_data,
    dag=dag,
)

preprocess_task = PythonOperator(
    task_id='preprocess_data',
    python_callable=preprocess_data,
    dag=dag,
)

train_task = PythonOperator(
    task_id='train_model',
    python_callable=train_model,
    dag=dag,
)

evaluate_task = PythonOperator(
    task_id='evaluate_and_deploy',
    python_callable=evaluate_and_deploy,
    dag=dag,
)

# 依存関係の設定
extract_task >> preprocess_task >> train_task >> evaluate_task

from prefect import flow, task
from prefect.tasks import task_input_hash
from datetime import timedelta
import pandas as pd
import mlflow

@task(cache_key_fn=task_input_hash, cache_expiration=timedelta(hours=1))
def load_data(data_path: str) -> pd.DataFrame:
    """データの読み込み（キャッシュ付き）"""
    return pd.read_parquet(data_path)

@task(retries=3, retry_delay_seconds=10)
def preprocess(df: pd.DataFrame) -> pd.DataFrame:
    """データの前処理"""
    df = df.dropna()
    df = df.drop_duplicates()
    return df

@task
def train(df: pd.DataFrame, params: dict) -> str:
    """モデルの学習"""
    from sklearn.ensemble import RandomForestClassifier

    mlflow.set_experiment("prefect-ml-pipeline")

    with mlflow.start_run():
        model = RandomForestClassifier(**params)
        X = df.drop('target', axis=1)
        y = df['target']
        model.fit(X, y)

        mlflow.sklearn.log_model(model, "model")
        mlflow.log_params(params)

        return mlflow.active_run().info.run_id

@task
def evaluate(run_id: str) -> float:
    """モデルの評価"""
    client = mlflow.tracking.MlflowClient()
    run = client.get_run(run_id)
    return run.data.metrics.get('accuracy', 0.0)

@flow(name="ml-training-pipeline")
def ml_pipeline(data_path: str, params: dict = None):
    """メインパイプライン"""
    if params is None:
        params = {"n_estimators": 100, "max_depth": 5}

    df = load_data(data_path)
    processed_df = preprocess(df)
    run_id = train(processed_df, params)
    accuracy = evaluate(run_id)

    print(f"Pipeline complete. Accuracy: {accuracy:.4f}")
    return accuracy

if __name__ == "__main__":
    ml_pipeline(
        data_path="data/train.parquet",
        params={"n_estimators": 200, "max_depth": 7}
    )

import mlflow
from mlflow.tracking import MlflowClient

client = MlflowClient()

# モデルの登録
model_uri = "runs:/abc123def456/model"
registered_model = mlflow.register_model(model_uri, "MyClassifier")

# バージョン情報の確認
print(f"Version: {registered_model.version}")
print(f"Status: {registered_model.status}")

# モデルバージョンのメタデータ更新
client.update_model_version(
    name="MyClassifier",
    version=registered_model.version,
    description="Random Forest with 200 estimators, accuracy 0.94"
)

# タグの追加
client.set_model_version_tag(
    name="MyClassifier",
    version=registered_model.version,
    key="validated_by",
    value="data-science-team"
)

# Stagingへの移行
client.transition_model_version_stage(
    name="MyClassifier",
    version=1,
    stage="Staging",
    archive_existing_versions=False
)

# 検証後にProductionへ昇格
def promote_to_production(model_name: str, version: int, min_accuracy: float = 0.90):
    """モデルをProductionに昇格"""
    model_version = client.get_model_version(model_name, version)
    run_id = model_version.run_id
    run = client.get_run(run_id)

    accuracy = run.data.metrics.get('accuracy', 0)

    if accuracy < min_accuracy:
        raise ValueError(
            f"Model accuracy {accuracy:.4f} is below minimum {min_accuracy}"
        )

    # Productionへの昇格（既存のProductionバージョンをアーカイブ）
    client.transition_model_version_stage(
        name=model_name,
        version=version,
        stage="Production",
        archive_existing_versions=True
    )

    print(f"Model {model_name} v{version} promoted to Production!")
    print(f"Accuracy: {accuracy:.4f}")

# Productionモデルの読み込み
production_model = mlflow.pyfunc.load_model(
    model_uri=f"models:/MyClassifier/Production"
)
predictions = production_model.predict(X_test)

# MLflowモデルのサービング
mlflow models serve \
  -m "models:/MyClassifier/Production" \
  --host 0.0.0.0 \
  --port 5001

# 予測リクエスト
curl -X POST http://localhost:5001/invocations \
  -H "Content-Type: application/json" \
  -d '{"dataframe_split": {"columns": ["feature1", "feature2"], "data": [[1.0, 2.0]]}}'

# Dockerfile.train - 学習用イメージ
FROM python:3.11-slim

WORKDIR /app

RUN apt-get update && apt-get install -y \
    gcc \
    g++ \
    git \
    && rm -rf /var/lib/apt/lists/*

COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

COPY src/ ./src/
COPY params.yaml .
COPY dvc.yaml .

CMD ["python", "src/train.py"]

# Dockerfile.serve - サービング用マルチステージビルド
# ステージ1: ビルダー
FROM python:3.11 AS builder

WORKDIR /build

COPY requirements.txt .
RUN pip install --no-cache-dir --target /install -r requirements.txt

# ステージ2: 最終スリムイメージ
FROM python:3.11-slim AS runtime

WORKDIR /app

COPY --from=builder /install /usr/local/lib/python3.11/site-packages

COPY src/serve.py .
COPY models/ ./models/

EXPOSE 8080

RUN useradd -m -u 1000 mluser
USER mluser

CMD ["python", "serve.py"]

# Dockerfile.gpu - GPU対応イメージ
FROM nvidia/cuda:12.1.0-cudnn8-devel-ubuntu22.04

ENV DEBIAN_FRONTEND=noninteractive

RUN apt-get update && apt-get install -y \
    python3.11 \
    python3-pip \
    git \
    && rm -rf /var/lib/apt/lists/*

RUN pip3 install torch torchvision torchaudio \
    --index-url https://download.pytorch.org/whl/cu121

WORKDIR /app
COPY requirements-gpu.txt .
RUN pip3 install --no-cache-dir -r requirements-gpu.txt

COPY src/ ./src/

CMD ["python3", "src/train_gpu.py"]

# GPUコンテナの実行
docker run --gpus all \
  -v /data:/data \
  -v /models:/models \
  --shm-size=8gb \
  my-ml-gpu:latest

# docker-compose.yml
version: '3.8'

services:
  mlflow:
    image: ghcr.io/mlflow/mlflow:latest
    command: >
      mlflow server
      --backend-store-uri postgresql://mlflow:password@postgres:5432/mlflow
      --default-artifact-root s3://my-bucket/mlflow
      --host 0.0.0.0
      --port 5000
    ports:
      - '5000:5000'
    depends_on:
      - postgres

  postgres:
    image: postgres:15
    environment:
      POSTGRES_DB: mlflow
      POSTGRES_USER: mlflow
      POSTGRES_PASSWORD: password
    volumes:
      - postgres_data:/var/lib/postgresql/data

  minio:
    image: minio/minio:latest
    command: server /data --console-address ":9001"
    environment:
      MINIO_ROOT_USER: minioadmin
      MINIO_ROOT_PASSWORD: minioadmin
    ports:
      - '9000:9000'
      - '9001:9001'
    volumes:
      - minio_data:/data

  trainer:
    build:
      context: .
      dockerfile: Dockerfile.train
    environment:
      MLFLOW_TRACKING_URI: http://mlflow:5000
    volumes:
      - ./data:/app/data
      - ./models:/app/models
    depends_on:
      - mlflow
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]

volumes:
  postgres_data:
  minio_data:

# Dockerfile.pytorch-serve
FROM python:3.11-slim

WORKDIR /app

RUN pip install --no-cache-dir \
    torch==2.2.0+cpu \
    torchvision \
    fastapi \
    uvicorn \
    pydantic \
    numpy \
    pillow \
    --index-url https://download.pytorch.org/whl/cpu

COPY src/server.py .
COPY models/model.pt .

EXPOSE 8000

CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "8000"]

# src/server.py
from fastapi import FastAPI, HTTPException
from pydantic import BaseModel
import torch
import numpy as np
from typing import List

app = FastAPI(title="ML Model Server")

# サーバー起動時にモデルを一度読み込む
model = torch.jit.load("model.pt")
model.eval()

class PredictionRequest(BaseModel):
    features: List[List[float]]

class PredictionResponse(BaseModel):
    predictions: List[int]
    probabilities: List[List[float]]

@app.get("/health")
async def health_check():
    return {"status": "healthy"}

@app.post("/predict", response_model=PredictionResponse)
async def predict(request: PredictionRequest):
    try:
        tensor = torch.tensor(request.features, dtype=torch.float32)

        with torch.no_grad():
            outputs = model(tensor)
            probabilities = torch.softmax(outputs, dim=1)
            predictions = torch.argmax(probabilities, dim=1)

        return PredictionResponse(
            predictions=predictions.tolist(),
            probabilities=probabilities.tolist()
        )
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

# k8s/namespace.yaml
apiVersion: v1
kind: Namespace
metadata:
  name: ml-platform
  labels:
    app: ml-platform
    environment: production

# k8s/training-job.yaml
apiVersion: batch/v1
kind: Job
metadata:
  name: ml-training-job
  namespace: ml-platform
spec:
  backoffLimit: 3
  template:
    spec:
      restartPolicy: OnFailure

      nodeSelector:
        accelerator: nvidia-t4

      tolerations:
        - key: nvidia.com/gpu
          operator: Exists
          effect: NoSchedule

      containers:
        - name: trainer
          image: my-registry/ml-trainer:v1.2.0

          resources:
            requests:
              memory: '8Gi'
              cpu: '4'
              nvidia.com/gpu: '1'
            limits:
              memory: '16Gi'
              cpu: '8'
              nvidia.com/gpu: '1'

          env:
            - name: MLFLOW_TRACKING_URI
              value: 'http://mlflow-service.ml-platform:5000'
            - name: AWS_ACCESS_KEY_ID
              valueFrom:
                secretKeyRef:
                  name: aws-credentials
                  key: access_key_id

          volumeMounts:
            - name: training-data
              mountPath: /data
            - name: dshm
              mountPath: /dev/shm

      volumes:
        - name: training-data
          persistentVolumeClaim:
            claimName: training-data-pvc
        - name: dshm
          emptyDir:
            medium: Memory
            sizeLimit: 8Gi

# k8s/training-cronjob.yaml
apiVersion: batch/v1
kind: CronJob
metadata:
  name: daily-retraining
  namespace: ml-platform
spec:
  schedule: '0 2 * * *' # 毎日午前2時
  concurrencyPolicy: Forbid
  successfulJobsHistoryLimit: 3
  failedJobsHistoryLimit: 3

  jobTemplate:
    spec:
      template:
        spec:
          restartPolicy: OnFailure
          containers:
            - name: retrainer
              image: my-registry/ml-trainer:latest
              command: ['python', 'src/retrain.py']
              resources:
                requests:
                  memory: '4Gi'
                  cpu: '2'
                limits:
                  memory: '8Gi'
                  cpu: '4'

# k8s/model-serving-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: model-server
  namespace: ml-platform
spec:
  replicas: 3
  selector:
    matchLabels:
      app: model-server

  template:
    metadata:
      labels:
        app: model-server
    spec:
      containers:
        - name: model-server
          image: my-registry/model-server:v1.0.0
          ports:
            - containerPort: 8000

          resources:
            requests:
              memory: '2Gi'
              cpu: '1'
            limits:
              memory: '4Gi'
              cpu: '2'

          livenessProbe:
            httpGet:
              path: /health
              port: 8000
            initialDelaySeconds: 30
            periodSeconds: 10

          readinessProbe:
            httpGet:
              path: /health
              port: 8000
            initialDelaySeconds: 15
            periodSeconds: 5

---
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: model-server-hpa
  namespace: ml-platform
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: model-server
  minReplicas: 2
  maxReplicas: 20
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 70

# kubeflow_pipeline.py
import kfp
from kfp import dsl

@dsl.component(base_image="python:3.11", packages_to_install=["pandas", "scikit-learn"])
def prepare_data(data_path: str, output_path: kfp.dsl.OutputPath(str)):
    import pandas as pd
    from sklearn.model_selection import train_test_split

    df = pd.read_parquet(data_path)
    train, test = train_test_split(df, test_size=0.2)
    train.to_parquet(f"{output_path}/train.parquet")
    test.to_parquet(f"{output_path}/test.parquet")

@dsl.component(
    base_image="python:3.11",
    packages_to_install=["scikit-learn", "mlflow", "pandas"]
)
def train_model(
    data_path: str,
    n_estimators: int,
    max_depth: int,
    model_output: kfp.dsl.OutputPath(str),
    mlflow_uri: str
):
    import mlflow
    import pandas as pd
    from sklearn.ensemble import RandomForestClassifier

    mlflow.set_tracking_uri(mlflow_uri)

    df = pd.read_parquet(f"{data_path}/train.parquet")
    X = df.drop('target', axis=1)
    y = df['target']

    with mlflow.start_run():
        model = RandomForestClassifier(
            n_estimators=n_estimators,
            max_depth=max_depth
        )
        model.fit(X, y)

        run_id = mlflow.active_run().info.run_id
        mlflow.sklearn.log_model(model, "model")

    with open(model_output, 'w') as f:
        f.write(run_id)

@dsl.pipeline(
    name="ML Training Pipeline",
    description="End-to-end ML training pipeline"
)
def ml_pipeline(
    data_path: str = "gs://my-bucket/data/train.parquet",
    n_estimators: int = 100,
    max_depth: int = 5,
    mlflow_uri: str = "http://mlflow:5000"
):
    prepare_task = prepare_data(data_path=data_path)

    train_task = train_model(
        data_path=prepare_task.output,
        n_estimators=n_estimators,
        max_depth=max_depth,
        mlflow_uri=mlflow_uri
    )

if __name__ == "__main__":
    kfp.compiler.Compiler().compile(
        pipeline_func=ml_pipeline,
        package_path="ml_pipeline.yaml"
    )

    client = kfp.Client(host="http://kubeflow-host/pipeline")
    run = client.create_run_from_pipeline_package(
        pipeline_file="ml_pipeline.yaml",
        arguments={"n_estimators": 200, "max_depth": 7},
        run_name="training-run-001"
    )

# .github/workflows/ml-pipeline.yml
name: ML CI/CD Pipeline

on:
  push:
    branches: [main, develop]
    paths:
      - 'src/**'
      - 'params.yaml'
      - 'requirements.txt'
  pull_request:
    branches: [main]

env:
  MLFLOW_TRACKING_URI: ${{ secrets.MLFLOW_TRACKING_URI }}

jobs:
  test:
    name: Unit Tests
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'
          cache: 'pip'
      - name: Install dependencies
        run: |
          pip install -r requirements.txt
          pip install pytest pytest-cov
      - name: Run unit tests
        run: pytest tests/unit/ -v --cov=src --cov-report=xml

  train-and-evaluate:
    name: Train and Evaluate Model
    runs-on: ubuntu-latest
    needs: test
    if: github.ref == 'refs/heads/main'
    steps:
      - uses: actions/checkout@v4
      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'
      - name: Install dependencies
        run: pip install -r requirements.txt
      - name: Run DVC pipeline
        run: dvc repro
      - name: Check model performance
        run: |
          python scripts/check_metrics.py \
            --min-accuracy 0.90 \
            --metrics-file metrics/scores.json
      - name: Register model
        run: python scripts/register_model.py

  build-and-push:
    name: Build Docker Image
    runs-on: ubuntu-latest
    needs: train-and-evaluate
    steps:
      - uses: actions/checkout@v4
      - name: Login to Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Build and push
        uses: docker/build-push-action@v5
        with:
          context: .
          file: Dockerfile.serve
          push: true
          tags: |
            ghcr.io/${{ github.repository }}/model-server:latest
            ghcr.io/${{ github.repository }}/model-server:${{ github.sha }}

  deploy:
    name: Deploy to Kubernetes
    runs-on: ubuntu-latest
    needs: build-and-push
    environment: production
    steps:
      - uses: actions/checkout@v4
      - name: Configure kubectl
        uses: azure/k8s-set-context@v3
        with:
          kubeconfig: ${{ secrets.KUBECONFIG }}
      - name: Deploy
        run: |
          kubectl set image deployment/model-server \
            model-server=ghcr.io/${{ github.repository }}/model-server:${{ github.sha }} \
            -n ml-platform
          kubectl rollout status deployment/model-server -n ml-platform

pip install evidently

import pandas as pd
from evidently.report import Report
from evidently.metric_preset import DataDriftPreset, DataQualityPreset
from evidently.metrics import ColumnDriftMetric

# 参照データ（学習時のデータ）
reference_data = pd.read_parquet("data/reference.parquet")

# 現在のデータ（本番環境での予測データ）
current_data = pd.read_parquet("data/current.parquet")

# データドリフトレポート
drift_report = Report(metrics=[
    DataDriftPreset(),
    DataQualityPreset(),
    ColumnDriftMetric(column_name="age"),
    ColumnDriftMetric(column_name="income"),
])

drift_report.run(
    reference_data=reference_data,
    current_data=current_data
)

drift_report.save_html("drift_report.html")

result = drift_report.as_dict()
drift_detected = result['metrics'][0]['result']['dataset_drift']

if drift_detected:
    print("Data drift detected! Consider retraining.")
    send_alert("Data drift detected in production model")

from evidently.report import Report
from evidently.metric_preset import ClassificationPreset

prediction_data = pd.DataFrame({
    'target': y_true,
    'prediction': y_pred,
    'prediction_proba': y_proba,
    'feature1': X_test['feature1'],
    'feature2': X_test['feature2'],
})

performance_report = Report(metrics=[ClassificationPreset()])
performance_report.run(
    reference_data=reference_predictions,
    current_data=prediction_data
)

metrics = performance_report.as_dict()
current_accuracy = metrics['metrics'][0]['result']['current']['accuracy']
reference_accuracy = metrics['metrics'][0]['result']['reference']['accuracy']

degradation = reference_accuracy - current_accuracy
if degradation > 0.05:
    trigger_retraining(f"Model degraded by {degradation:.2%}")

from prometheus_client import Counter, Histogram, Gauge, start_http_server

prediction_counter = Counter('model_predictions_total', 'Total predictions')
prediction_latency = Histogram('model_prediction_duration_seconds', 'Prediction latency')
model_accuracy_gauge = Gauge('model_accuracy', 'Current model accuracy')
drift_score_gauge = Gauge('data_drift_score', 'Data drift score')

class MonitoredModelServer:
    def __init__(self, model, reference_data):
        self.model = model
        self.reference_data = reference_data
        self.predictions_buffer = []
        start_http_server(8001)

    def predict(self, features):
        with prediction_latency.time():
            prediction = self.model.predict(features)

        prediction_counter.inc()
        self.predictions_buffer.append(prediction)

        if len(self.predictions_buffer) >= 1000:
            self._run_monitoring()

        return prediction

    def _run_monitoring(self):
        current_df = pd.DataFrame(self.predictions_buffer)
        drift_report = Report(metrics=[DataDriftPreset()])
        drift_report.run(
            reference_data=self.reference_data,
            current_data=current_df
        )

        result = drift_report.as_dict()
        drift_share = result['metrics'][0]['result']['share_of_drifted_columns']
        drift_score_gauge.set(drift_share)

        self.predictions_buffer = []

pip install feast
feast init my-feature-store
cd my-feature-store

# features.py
from feast import Entity, FeatureView, FileSource, ValueType, Field
from feast.types import Float64, Int64
from datetime import timedelta

user = Entity(
    name="user_id",
    value_type=ValueType.INT64,
    description="User ID"
)

user_stats_source = FileSource(
    path="data/user_stats.parquet",
    timestamp_field="event_timestamp",
)

user_stats_fv = FeatureView(
    name="user_statistics",
    entities=[user],
    ttl=timedelta(days=30),
    schema=[
        Field(name="purchase_count_7d", dtype=Int64),
        Field(name="purchase_amount_7d", dtype=Float64),
        Field(name="avg_session_duration", dtype=Float64),
        Field(name="last_purchase_days_ago", dtype=Int64),
    ],
    online=True,
    source=user_stats_source,
)

# feast_usage.py
from feast import FeatureStore
import pandas as pd
from datetime import datetime

store = FeatureStore(repo_path=".")

# オフライン特徴量の取得（学習用）
entity_df = pd.DataFrame({
    "user_id": [1001, 1002, 1003],
    "event_timestamp": pd.to_datetime(["2026-01-01", "2026-01-02", "2026-01-03"])
})

training_df = store.get_historical_features(
    entity_df=entity_df,
    features=[
        "user_statistics:purchase_count_7d",
        "user_statistics:purchase_amount_7d",
        "user_statistics:avg_session_duration",
    ]
).to_df()

# オンラインストアへのマテリアライズ
store.materialize_incremental(end_date=datetime.now())

# オンライン特徴量の取得（リアルタイム推論用）
online_features = store.get_online_features(
    features=[
        "user_statistics:purchase_count_7d",
        "user_statistics:purchase_amount_7d",
    ],
    entity_rows=[
        {"user_id": 1001},
        {"user_id": 1002},
    ]
).to_dict()

データソース（DB、S3、API）
    ↓
データ収集（Airflow DAG）
    ↓
データ検証（Great Expectations）
    ↓
特徴量エンジニアリング（Feast）
    ↓
モデル学習（MLflowトラッキング）
    ↓
モデル評価（自動検証）
    ↓
モデルレジストリ（MLflow Registry）
    ↓
CI/CD（GitHub Actions）
    ↓
コンテナビルド（Docker）
    ↓
K8sデプロイ（Kubernetes）
    ↓
サービング（FastAPI + Triton）
    ↓
モニタリング（Evidently + Prometheus + Grafana）
    ↓
アラート（ドリフト検出時に再学習をトリガー）

# scripts/full_pipeline.py
import mlflow
import pandas as pd
import numpy as np
from sklearn.ensemble import GradientBoostingClassifier
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.metrics import accuracy_score, roc_auc_score
from sklearn.preprocessing import StandardScaler
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

class ChurnPredictionPipeline:
    """顧客チャーン予測パイプライン"""

    def __init__(self, mlflow_uri: str = "http://localhost:5000"):
        mlflow.set_tracking_uri(mlflow_uri)
        mlflow.set_experiment("churn-prediction")
        self.scaler = StandardScaler()

    def load_data(self, data_path: str) -> pd.DataFrame:
        logger.info(f"Loading data from {data_path}")
        df = pd.read_parquet(data_path)

        assert df.shape[0] > 1000, "Too few samples"
        assert 'churn' in df.columns, "Target column 'churn' missing"

        missing_rate = df.isnull().mean()
        high_missing = missing_rate[missing_rate > 0.5].index.tolist()
        if high_missing:
            logger.warning(f"Columns with >50% missing: {high_missing}")
            df = df.drop(columns=high_missing)

        return df

    def engineer_features(self, df: pd.DataFrame) -> pd.DataFrame:
        logger.info("Engineering features...")

        numeric_cols = df.select_dtypes(include=[np.number]).columns
        df[numeric_cols] = df[numeric_cols].fillna(df[numeric_cols].median())

        if 'tenure_months' in df.columns and 'monthly_charges' in df.columns:
            df['total_value'] = df['tenure_months'] * df['monthly_charges']

        if 'num_support_tickets' in df.columns and 'tenure_months' in df.columns:
            df['tickets_per_month'] = (
                df['num_support_tickets'] / (df['tenure_months'] + 1)
            )

        return df

    def train(self, df: pd.DataFrame, params: dict = None) -> str:
        if params is None:
            params = {
                "n_estimators": 200,
                "max_depth": 5,
                "learning_rate": 0.05,
                "subsample": 0.8,
                "random_state": 42,
            }

        feature_cols = [c for c in df.columns if c != 'churn']
        X = df[feature_cols]
        y = df['churn']

        X_train, X_test, y_train, y_test = train_test_split(
            X, y, test_size=0.2, random_state=42, stratify=y
        )

        X_train_scaled = self.scaler.fit_transform(X_train)
        X_test_scaled = self.scaler.transform(X_test)

        with mlflow.start_run(run_name="churn-gbt"):
            mlflow.log_params(params)
            mlflow.log_param("train_size", len(X_train))
            mlflow.log_param("n_features", len(feature_cols))

            model = GradientBoostingClassifier(**params)
            model.fit(X_train_scaled, y_train)

            y_pred = model.predict(X_test_scaled)
            y_proba = model.predict_proba(X_test_scaled)[:, 1]

            accuracy = accuracy_score(y_test, y_pred)
            auc_roc = roc_auc_score(y_test, y_proba)

            mlflow.log_metrics({
                "accuracy": accuracy,
                "auc_roc": auc_roc,
            })

            cv_scores = cross_val_score(
                model, X_train_scaled, y_train, cv=5, scoring='roc_auc'
            )
            mlflow.log_metric("cv_auc_mean", cv_scores.mean())
            mlflow.log_metric("cv_auc_std", cv_scores.std())

            mlflow.sklearn.log_model(
                model,
                "model",
                registered_model_name="ChurnPredictor"
            )

            run_id = mlflow.active_run().info.run_id
            logger.info(f"Training complete. AUC-ROC: {auc_roc:.4f}")
            return run_id

    def promote_best_model(self, min_auc: float = 0.85):
        client = mlflow.tracking.MlflowClient()

        experiment = client.get_experiment_by_name("churn-prediction")
        runs = client.search_runs(
            experiment_ids=[experiment.experiment_id],
            filter_string="metrics.auc_roc > 0.80",
            order_by=["metrics.auc_roc DESC"],
            max_results=1
        )

        if not runs:
            raise ValueError("No runs found meeting criteria")

        best_run = runs[0]
        auc = best_run.data.metrics['auc_roc']

        if auc < min_auc:
            raise ValueError(f"Best AUC {auc:.4f} below minimum {min_auc}")

        model_uri = f"runs:/{best_run.info.run_id}/model"
        mv = mlflow.register_model(model_uri, "ChurnPredictor")

        client.transition_model_version_stage(
            name="ChurnPredictor",
            version=mv.version,
            stage="Production",
            archive_existing_versions=True
        )

        logger.info(f"Model v{mv.version} promoted to Production! AUC: {auc:.4f}")
        return mv.version

if __name__ == "__main__":
    pipeline = ChurnPredictionPipeline()
    df = pipeline.load_data("data/customers.parquet")
    df = pipeline.engineer_features(df)
    run_id = pipeline.train(df)
    pipeline.promote_best_model(min_auc=0.85)