chat-notifier 0.1.2

Google Chat notifier for MLflow and Ray jobs

Chat Notifier

Simple Python library to send Google Chat alerts for:

• MLflow best metric updates
• Ray training job status

Installation

pip install chat-notifier

or for local development

pip install -e .

---

MLflow Notifier Usage

Used to notify when a new best metric is achieved in MLflow.

import mlflow from notifier import MLflowNotifier

notifier = MLflowNotifier()

with mlflow.start_run() as run:

mlflow.log_metric("mAP50", 0.91)

notifier.check_new_best(
    experiment_name="ppe-detection",
    run_id=run.info.run_id
)

Required environment variables:

MLFLOW_TRACKING_URI=http://localhost:5000 GOOGLE_CHAT_WEBHOOK_URL=your_google_chat_webhook METRIC_NAME=mAP50 METRIC_MODE=max

---

Ray Job Notifier Usage

Used to notify training success or failure.

from notifier import RayJobNotifier from datetime import datetime

notifier = RayJobNotifier()

start_time = datetime.utcnow()

status = 0 logs = ""

end_time = datetime.utcnow()

notifier.send_notification( status=status, logs=logs, start_time=start_time, end_time=end_time )

Required environment variables:

GOOGLE_CHAT_RAY_WEBHOOK_URL=your_google_chat_webhook SERVER=gpu-server MODEL_NAME=my-model