Define the task. Build your first model.
Dawn is about clarity. Start with a simple dataset and a strong baseline. Establish a reliable benchmark you can build on.
from breezeml import datasets, fit, predict
# 1. Load a DataFrame (the one you picked, live)
df = datasets.load(dataset)
# 2. Name the target - that's the whole config
model = fit(df, target) # RandomForest, 5-fold CV
# 3. Read honest, cross-validated scores
model.evaluate() # accuracy, F1, ROC AUC
model.feature_importances_ # what mattered most