Question

我的模型是2个不同的keras模型的集合，这些模型连接到相同的输入层，合并后具有2个输出层。两种模型都是经过预训练的，我正在尝试创建并行架构。我的架构是： `

define('PACKAGE_PRODUCT_ID', 1010);

    add_action('woocommerce_payment_complete', 'order_splitter', 100, 1);
    function order_splitter($order_id){
        $completed_order = new WC_Order($order_id);
        $item_splitted = false;

        $address = array(
            'first_name' => $completed_order->get_billing_first_name(),
            'last_name'  => $completed_order->get_billing_last_name(),
            'company'    => '',
            'email'      => $completed_order->get_billing_email(),
            'phone'      => $completed_order->get_billing_phone(),
            'address_1'  => $completed_order->get_billing_address_1(),
            'address_2'  => $completed_order->get_billing_address_2(),
            'city'       => $completed_order->get_billing_city(),
            'state'      => $completed_order->get_billing_state(),
            'postcode'   => $completed_order->get_billing_postcode(),
            'country'    => $completed_order->get_billing_country()
        );

        foreach($completed_order->get_items() as $item){

                if (!$item_splitted && $item->get_product_id() === PACKAGE_PRODUCT_ID) {

                    //create new order
                    $new_order_args = array(
                        'customer_id' => $completed_order->get_customer_id(),
                        'status' => 'wc-pending',
                    );
                    $new_order = wc_create_order($new_order_args);
                    $product_to_add = wc_get_product(PACKAGE_PRODUCT_ID);
                    $new_order->add_product($product_to_add, 1, array());
                    $new_order->set_address($address, 'billing');
                    $new_order->set_address($address, 'shipping');
                    $new_order->update_status('wc-processing');
                    $new_order->add_order_note('This order created automatically');
                    $new_order->save();
                    $completed_order->remove_item($item->get_id());
                    $item_splitted = true;
                } else  if ($item_splitted && $item['product_id'] === PACKAGE_PRODUCT_ID){
                    # This will ensure every 2 products are splitted (skipping the 2nd one)
                    $item_splitted = false;
                    continue;
                }
        }
    }

`

模型被编译，我也可以做出预测，但是当我保存并尝试重新加载它时，我得到了：

model_input = Input(shape=(224,224,3), name="model_input") gender_model_copy.layers.pop(0) color_model_copy.layers.pop(0) color_model_ens1 = color_model_copy(model_input) gender_model_ens1 = gender_model_copy(model_input) model_f = Model(input=[model_input], output=[color_model_ens1,gender_model_ens1]) model_f.save('path') 完整跟踪：Github gist link。

我有一个使用ValueError: Invalid input_shape argument (None, 224, 224, 3): model has 0 tensor inputs.添加的自定义图层 custom_objects={'Scale':Scale()}中的参数我的keras版本是2.2.5，而tensorflow版本是1.15

Answer 1

如果您重命名其中一个图层，则可能会出现问题。停止重命名。

加载整体keras模型会给出ValueError：无效的input_shape参数（None，224、224、3）：模型具有0张量输入

1 个答案: