import { useCallback, useState } from "react";
|
import { Button } from "@humansignal/ui";
|
import { useAPI } from "../../../providers/ApiProvider";
|
import { Typography } from "@humansignal/ui";
|
|
export const StartModelTraining = ({ backend }) => {
|
const api = useAPI();
|
const [response, setResponse] = useState(null);
|
|
const onStartTraining = useCallback(
|
async (backend) => {
|
const res = await api.callApi("trainMLBackend", {
|
params: {
|
pk: backend.id,
|
},
|
});
|
|
setResponse(res.response || {});
|
},
|
[api],
|
);
|
|
return (
|
<div className="max-w-[680px]">
|
<Typography size="small" className="text-neutral-content-subtler">
|
您即将手动触发模型的训练过程。此操作将根据 ML 后端中训练方法的实现方式开始学习阶段。继续以开始此过程。
|
</Typography>
|
<Typography size="small" className="text-neutral-content-subtler mt-base mb-wide">
|
*注意：目前，此界面中没有用于跟踪训练进度的内置反馈循环。您需要通过模型自己的工具和环境直接监控模型的训练步骤。
|
</Typography>
|
|
{!response && (
|
<Button
|
onClick={() => {
|
onStartTraining(backend);
|
}}
|
>
|
开始训练
|
</Button>
|
)}
|
|
{!!response && (
|
<>
|
<pre>请求已发送！</pre>
|
<pre>响应：{JSON.stringify(response, null, 2)}</pre>
|
</>
|
)}
|
</div>
|
);
|
};
|
