Source code for app.train

import torch
import qumphy
import yaml

torch.set_float32_matmul_precision("medium")




def load_config(args):
    config = {}
    for config_file in args.config:
        with open(config_file, "r") as stream:
            qumphy.misc.update_dictionary(config, yaml.safe_load(stream))

    for parameter in args.parameters:
        qumphy.misc.update_dictionary(config, parameter)

    return config





def train(args, config):
    trainer = qumphy.trainer.Trainer(config)

    if args.fit:
        trainer.fit()
    if args.test:
        trainer.test()
    if args.predict:
        trainer.predict()





def main():
    args = qumphy.misc.train_argument_parser()

    config = load_config(args)

    train(args, config)



if __name__ == "__main__":
    main()