Penerapan model

Pengantar MLflow

Weston Bassler

Senior MLOps Engineer

Siklus ML

Tahapan Siklus ML

1 datacamp.com
Pengantar MLflow

Versi dan tahap model

  • Versi model

    • Versi +1

  • Tahap model

    • Staging
    • Production
    • Archived

Versi dan tahap model

Pengantar MLflow

Cara menerapkan model

Muat model

# MLflow flavor
mlflow.FLAVOR.load_model()

Layani model

# Perintah MLflow serve
mlflow models serve
Pengantar MLflow

URI model

Konvensi

models:/

Versi model

models:/model_name/version

Tahap model

models:/model_name/stage
Pengantar MLflow

Memuat model

# Impor flavor
import mlflow.FLAVOR


# Muat versi
mlflow.FLAVOR.load_model("models:/model_name/version")


# Muat tahap
mlflow.FLAVOR.load_model("models:/model_name/stage")
Pengantar MLflow

Contoh memuat model

# Impor flavor
import mlflow.sklearn


# Muat model Unicorn di Staging
model = mlflow.sklearn.load_model("models:/Unicorn/Staging")

# Cetak model
model

LogisticRegression()

# Inferensi
model.predict(data)
Pengantar MLflow

Melayani model

# Layani model Unicorn di tahap Production
mlflow models serve -m "models:/Unicorn/Production"

2023/03/26 15:07:00 INFO mlflow.models.flavor_backend_registry: 
Selected backend for flavor 'python_function'
2023/03/26 15:07:00 INFO mlflow.pyfunc.backend: === Running command 'exec gunicorn 
--timeout=60 -b 127.0.0.1:5000 -w 1 ${GUNICORN_CMD_ARGS} -- 
mlflow.pyfunc.scoring_server.wsgi:app'
[2023-03-26 15:07:00 -0400] [86409] [INFO] Starting gunicorn 20.1.0
[2023-03-26 15:07:00 -0400] [86409] [INFO] Listening at: http://127.0.0.1:5000
[2023-03-26 15:07:00 -0400] [86409] [INFO] Using worker: sync
[2023-03-26 15:07:00 -0400] [86410] [INFO] Booting worker with pid: 86410
Pengantar MLflow

Endpoint invocations

api

http://localhost:5000/invocations

  • Format
    • CSV
    • JSON
1 Flaticon.com
Pengantar MLflow

Format CSV

pandas_df.to_csv()

Format JSON

{
  "dataframe_split": {
      "columns": ["R&D Spend", "Administration", "Marketing Spend", "State"],
      "data": [["165349.20", 136897.80, 471784.10, 1]]
  }
}
Pengantar MLflow

Prediksi model

# Kirim payload ke endpoint invocations
curl http://127.0.0.1:5000/invocations -H 'Content-Type: application/json' -d
{
  "dataframe_split": {
      "columns": ["R&D Spend", "Administration", "Marketing Spend", "State"],
      "data": [["165349.20", 136897.80, 471784.10, 1]]
  }
}

[[104055.1842384]]
Pengantar MLflow

Ayo berlatih!

Pengantar MLflow

Preparing Video For Download...