Dense model parameters and detached metrics
Configure initialization, optimizer, dropout and held out validation using explicit native model settings.
// Explicit model/configuration variables; no files are created.
import aner.dataset
import aner.nn
import aner.ml
let parts = dataset.iris().partition(validation: 0.2, test: 0.2, seed: 81)
let network = nn.sequential(seed: 42)
.dense(inputs: 4, outputs: 16, initialization: "he").relu()
.dropout(probability: 0.2)
.dense(inputs: 16, outputs: 3, initialization: "xavier")
let plan = ml.classifier(network, standardize: true,
loss: "cross_entropy", optimizer: "adam")
.on(device: "cpu", precision: "Float64", policy: "preserve")
.dropout_seed(seed: 2026)
let config = ml.training(epochs: 20, batch: 16, rate: 0.01, seed: 91)
.validation(parts.validation()).select_best(false)
let fitted = plan.fit(parts.train(), config: config)
print(fitted.summary())
print(fitted.evaluate(parts.test()).summary())
Make it your experiment.
Use Aner 0.0.5 or newer. Save this program as examples/neural_model_reference.aner inside a folder for your experiment. Open a terminal in that folder, then check and run the program.
aner check examples/neural_model_reference.aner
aner run examples/neural_model_reference.anerThe aner command must be on your PATH. Follow the installation guide if your terminal cannot find it.
Requires Aner 0.0.5 or newer. Uses complete synthetic tensors and the native classical ML runtime. No dataset download, credentials or external service is required. The guide documents exact expected results and all input constraints.
aner check examples/neural_model_reference.aner
aner run examples/neural_model_reference.anerBundled datasets retain their own source attribution and license notices. Example outcomes are not comparative benchmarks or evidence of clinical validity.
Dataset sources & attribution