Google Cloud & Vertex AI

from google.cloud import aiplatform

aiplatform.init(project='my-project', location='us-central1')

job = aiplatform.CustomTrainingJob(
    display_name="my-pytorch-training",
    script_path="train.py",
    container_uri="us-docker.pkg.dev/vertex-ai/training/pytorch-gpu.2-1:latest",
    requirements=["transformers", "datasets"],
    model_serving_container_image_uri="us-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-1:latest"
)

model = job.run(
    machine_type="n1-standard-4",
    accelerator_type="NVIDIA_TESLA_A100",
    accelerator_count=1,
    replica_count=1,
    args=["--epochs=20", "--learning-rate=0.001"]
)

import google.generativeai as genai

genai.configure(api_key="YOUR_API_KEY")

model = genai.GenerativeModel('gemini-1.5-pro')

# Text + image multimodal
import PIL.Image
img = PIL.Image.open("chart.png")
response = model.generate_content(["Analyse this chart:", img])
print(response.text)

# Streaming
for chunk in model.generate_content("Explain LLMs", stream=True):
    print(chunk.text, end="")

-- Train a logistic regression model directly in SQL
CREATE OR REPLACE MODEL `myproject.dataset.churn_model`
OPTIONS(
    model_type = 'logistic_reg',
    input_label_cols = ['churned'],
    data_split_method = 'auto_split'
) AS
SELECT
    days_since_last_purchase,
    total_orders,
    avg_order_value,
    customer_age,
    churned
FROM `myproject.dataset.customer_features`;

-- Evaluate
SELECT * FROM ML.EVALUATE(MODEL `myproject.dataset.churn_model`);

-- Predict
SELECT * FROM ML.PREDICT(MODEL `myproject.dataset.churn_model`,
    TABLE `myproject.dataset.new_customers`);