Quickstart¶
Welcome! This quickstart guide will convey the basics needed to use nnbench. You will define a benchmark, initialize a runner and reporter, and execute the benchmark, obtaining the results in the console in tabular format.
A short scikit-learn model benchmark¶
In the following simple example, we put the training and benchmarking logic in the same file. For more complex workloads, we recommend structuring your code into multiple files to improve project organization, similarly to unit tests. Check the user guide on structuring your benchmarks for inspiration.
from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
data = load_iris()
X, y = data.data, data.target
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)
model = RandomForestClassifier()
model.fit(X_train, y_train)
To benchmark your model, you encapsulate the benchmark code into a function and apply the @benchmark decorator.
This marks the function for collection to our benchmark runner later.
import nnbench
import numpy as np
from sklearn import base, metrics
@nnbench.benchmark()
def accuracy(model: base.BaseEstimator, X_test: np.ndarray, y_test: np.ndarray) -> float:
y_pred = model.predict(X_test)
accuracy = metrics.accuracy_score(y_test, y_pred)
return accuracy
Now we can instantiate a benchmark runner to collect and run the accuracy benchmark.
Then, using the BenchmarkReporter we report the resulting accuracy metric by printing it to the terminal in a table.
import nnbench
r = nnbench.BenchmarkRunner()
reporter = nnbench.BenchmarkReporter()
# To collect in the current file, pass "__main__" as module name.
result = r.run("__main__", params={"model": model, "X_test": X_test, "y_test": y_test})
reporter.display(result)
The resulting output might look like this: