javascript - Currently trying to use my MobileNet in a web application, nothing seems to work
问题描述
EDIT: The issue was the h5 file I saved before converting to tfjs. If you try something similar use the method save instead of save_weights when saving your keras model.
My project is supposed to identify me or my friends in any given image. I have trained a Mobile CNN made with keras and transformed it with tensorflowjs like so
I have followed a few tutorials and I still can't understand why my model won't load. This is part of my script:
let model;
async function loadMyModel() {
model = undefined;
model = await tf.loadLayersModel("https://consumingtoast.github.io/SelfieTime/model/model.json");
$('.progress-bar').hide();
alert("model loaded..");
}
try {
loadMyModel();
} catch {
alert("Something went wrong");
}
解决方案
推荐阅读
- r - AI平台R笔记本
- c - 创建 n 大小字节的数据类型
- reactjs - Google 地方信息(多个)
- powershell - Powershell 测试网站页面可用性和 HTTP 响应代码状态 try/catch
- python - 在仿射中使用极小的数字。在python中仿射
- php - php文件内容第一次不正确
- excel - Excel 表格硬编码结构化引用的公式错误
- sql - 如何在 bigquery 中旋转我的 sql 表?
- python - 2020 - Windows 10 - 带有 Tor 的 Python Selenium(错误:Tor 无法启动。)
- python - 如何将键发送到元素,如