ium_444409/train_model.py

import torch
import mlflow
import argparse
import numpy as np
from torch import nn
from torch.utils.data import DataLoader
from urllib.parse import urlparse

from model import PlantsDataset, MLP, train, test

default_batch_size = 64
default_epochs = 5

device = "cuda" if torch.cuda.is_available() else "cpu"

mlflow.set_tracking_uri("http://172.17.0.1:5000")
mlflow.set_experiment("s444409")


def setup_args():
    args_parser = argparse.ArgumentParser(prefix_chars='-')
    args_parser.add_argument('-b', '--batchSize', type=int, default=default_batch_size)
    args_parser.add_argument('-e', '--epochs', type=int, default=default_epochs)

    return args_parser.parse_args()


if __name__ == "__main__":
    args = setup_args()
    batch_size = args.batchSize
    epochs = args.epochs

    print(f"Using {device} device")

    plant_test = PlantsDataset('data/Plant_1_Generation_Data.csv.test')
    plant_train = PlantsDataset('data/Plant_1_Generation_Data.csv.train')

    input_example = np.array([plant_test.x_train.numpy()[0]])

    train_dataloader = DataLoader(plant_train, batch_size=batch_size)
    test_dataloader = DataLoader(plant_test, batch_size=batch_size)

    for i, (data, labels) in enumerate(train_dataloader):
        print(data.shape, labels.shape)
        print(data, labels)
        break

    model = MLP()
    print(model)

    loss_fn = nn.MSELoss()
    optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)
    
    print("Done!")

    torch.save(model.state_dict(), './model_out')
    print("Model saved in ./model_out file.")
    with mlflow.start_run() as run:
        print("MLflow run experiment_id: {0}".format(run.info.experiment_id))
        print("MLflow run artifact_uri: {0}".format(run.info.artifact_uri))

        mlflow.log_param("batch_size", batch_size)
        mlflow.log_param("epochs", epochs)
        for t in range(epochs):
            print(f"Epoch {t + 1}\n-------------------------------")
            train(train_dataloader, model, loss_fn, optimizer)
            last_loss = test(test_dataloader, model, loss_fn)
            mlflow.log_metric("rmse", last_loss)

        with torch.no_grad():
            preds = model(plant_test.x_train)
            signature = mlflow.models.signature.infer_signature(plant_test.x_train.numpy(), preds.numpy())
        tracking_url_type_store = urlparse(mlflow.get_tracking_uri()).scheme
        if tracking_url_type_store != "file":
            mlflow.pytorch.log_model(
                model,
                "s444409",
                registered_model_name="s444409",
                signature=signature,
                input_example=input_example
            )
        else:
            mlflow.pytorch.log_model(model, "s444409", signature=signature, input_example=input_example)