def test_train_eval():
    """
    test_train_eval
    """
    config = WideDeepConfig()
    data_path = config.data_path
    batch_size = config.batch_size
    epochs = config.epochs
    print("epochs is {}".format(epochs))
    ds_train = create_dataset(data_path, train_mode=True, epochs=epochs, batch_size=batch_size,
                              data_type=DataType.MINDRECORD, rank_id=get_rank(), rank_size=get_group_size())
    ds_eval = create_dataset(data_path, train_mode=False, epochs=epochs + 1, batch_size=batch_size,
                             data_type=DataType.MINDRECORD, rank_id=get_rank(), rank_size=get_group_size())
    print("ds_train.size: {}".format(ds_train.get_dataset_size()))
    print("ds_eval.size: {}".format(ds_eval.get_dataset_size()))

    net_builder = ModelBuilder()

    train_net, eval_net = net_builder.get_net(config)
    train_net.set_train()
    auc_metric = AUCMetric()

    model = Model(train_net, eval_network=eval_net, metrics={"auc": auc_metric})

    eval_callback = EvalCallBack(model, ds_eval, auc_metric, config)

    callback = LossCallBack(config=config)
    context.set_auto_parallel_context(strategy_ckpt_save_file="./strategy_train.ckpt")
    model.train(epochs, ds_train,
                callbacks=[TimeMonitor(ds_train.get_dataset_size()), eval_callback, callback])
    eval_values = list(eval_callback.eval_values)
    assert eval_values[0] > 0.78
Exemple #2
0
def test_train_eval():
    """
    test_train_eval
    """
    np.random.seed(1000)
    config = WideDeepConfig()
    data_path = config.data_path
    batch_size = config.batch_size
    epochs = config.epochs
    print("epochs is {}".format(epochs))
    ds_train = create_dataset(data_path, train_mode=True, epochs=epochs,
                              batch_size=batch_size, rank_id=get_rank(), rank_size=get_group_size())
    ds_eval = create_dataset(data_path, train_mode=False, epochs=epochs + 1,
                             batch_size=batch_size, rank_id=get_rank(), rank_size=get_group_size())
    print("ds_train.size: {}".format(ds_train.get_dataset_size()))
    print("ds_eval.size: {}".format(ds_eval.get_dataset_size()))

    net_builder = ModelBuilder()

    train_net, eval_net = net_builder.get_net(config)
    train_net.set_train()
    auc_metric = AUCMetric()

    model = Model(train_net, eval_network=eval_net, metrics={"auc": auc_metric})

    eval_callback = EvalCallBack(model, ds_eval, auc_metric, config)

    callback = LossCallBack(config=config)
    ckptconfig = CheckpointConfig(save_checkpoint_steps=ds_train.get_dataset_size(), keep_checkpoint_max=5)
    ckpoint_cb = ModelCheckpoint(prefix='widedeep_train',
                                 directory=config.ckpt_path, config=ckptconfig)
    out = model.eval(ds_eval)
    print("=====" * 5 + "model.eval() initialized: {}".format(out))
    model.train(epochs, ds_train,
                callbacks=[TimeMonitor(ds_train.get_dataset_size()), eval_callback, callback, ckpoint_cb])
Exemple #3
0
def test_eval(config):
    """
    test evaluate
    """
    data_path = config.data_path
    batch_size = config.batch_size
    ds_eval = create_dataset(data_path, train_mode=False, epochs=2,
                             batch_size=batch_size)
    print("ds_eval.size: {}".format(ds_eval.get_dataset_size()))

    net_builder = ModelBuilder()
    train_net, eval_net = net_builder.get_net(config)

    param_dict = load_checkpoint(config.ckpt_path)
    load_param_into_net(eval_net, param_dict)

    auc_metric = AUCMetric()
    model = Model(train_net, eval_network=eval_net, metrics={"auc": auc_metric})

    eval_callback = EvalCallBack(model, ds_eval, auc_metric, config)

    model.eval(ds_eval, callbacks=eval_callback)


if __name__ == "__main__":
    widedeep_config = WideDeepConfig()
    widedeep_config.argparse_init()

    test_eval(widedeep_config)
Exemple #4
0
                              batch_size=batch_size)
    print("ds_train.size: {}".format(ds_train.get_dataset_size()))

    net_builder = ModelBuilder()
    train_net, _ = net_builder.get_net(configure)
    train_net.set_train()

    model = Model(train_net)
    callback = LossCallBack(config=configure)
    ckptconfig = CheckpointConfig(
        save_checkpoint_steps=ds_train.get_dataset_size(),
        keep_checkpoint_max=5)
    ckpoint_cb = ModelCheckpoint(prefix='widedeep_train',
                                 directory=configure.ckpt_path,
                                 config=ckptconfig)
    model.train(epochs,
                ds_train,
                callbacks=[
                    TimeMonitor(ds_train.get_dataset_size()), callback,
                    ckpoint_cb
                ])


if __name__ == "__main__":
    config = WideDeepConfig()
    config.argparse_init()

    context.set_context(mode=context.GRAPH_MODE,
                        device_target=config.device_target)
    test_train(config)
Exemple #5
0
                                 directory=config.ckpt_path + '/ckpt_' +
                                 str(get_rank()) + '/',
                                 config=ckptconfig)
    callback_list = [
        TimeMonitor(ds_train.get_dataset_size()), eval_callback, callback
    ]
    if get_rank() == 0:
        callback_list.append(ckpoint_cb)
    model.train(epochs,
                ds_train,
                callbacks=callback_list,
                dataset_sink_mode=(parameter_server and cache_enable))


if __name__ == "__main__":
    wide_deep_config = WideDeepConfig()
    wide_deep_config.argparse_init()
    context.set_context(mode=context.GRAPH_MODE,
                        device_target=wide_deep_config.device_target,
                        save_graphs=True)
    cache_enable = wide_deep_config.vocab_cache_size > 0
    if cache_enable and wide_deep_config.device_target != "GPU":
        context.set_context(variable_memory_max_size="24GB")
    context.set_ps_context(enable_ps=True)
    init()
    context.set_context(save_graphs_path='./graphs_of_device_id_' +
                        str(get_rank()))

    if cache_enable:
        context.set_auto_parallel_context(
            parallel_mode=ParallelMode.AUTO_PARALLEL, gradients_mean=True)
Exemple #6
0
def create_network(name, *args, **kwargs):
    if name == 'wide_and_deep_multitable':
        wide_deep_config = WideDeepConfig()
        eval_net = get_WideDeep_net(wide_deep_config)
        return eval_net
    raise NotImplementedError(f"{name} is not implemented in the repo")