feat: callback training data

src/api/model.ts

@@ -52,7 +52,10 @@ export const getExportDataList = (modelId: string) =>
  * 获取模型正在拆分数据的数量
  */
 export const getModelSplitDataListLen = (modelId: string) =>
-  GET<number>(`/model/data/getSplitData?modelId=${modelId}`);
+  GET<{
+    splitDataQueue: number;
+    embeddingQueue: number;
+  }>(`/model/data/getTrainingData?modelId=${modelId}`);
 
 /**
  * 获取 web 页面内容

src/pages/api/model/data/getTrainingData.ts

@@ -2,6 +2,8 @@ import type { NextApiRequest, NextApiResponse } from 'next';
 import { jsonRes } from '@/service/response';
 import { connectToDatabase, SplitData, Model } from '@/service/mongo';
 import { authToken } from '@/service/utils/auth';
+import { ModelDataStatusEnum } from '@/constants/model';
+import { PgClient } from '@/service/pg';
 
 /* 拆分数据成QA */
 export default async function handler(req: NextApiRequest, res: NextApiResponse) {
@@ -14,15 +16,29 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse)
 
     const userId = await authToken(req);
 
-    // 找到长度大于0的数据
+    // split queue data
     const data = await SplitData.find({
       userId,
       modelId,
       textList: { $exists: true, $not: { $size: 0 } }
     });
 
+    // embedding queue data
+    const where: any = [
+      ['user_id', userId],
+      'AND',
+      ['model_id', modelId],
+      'AND',
+      ['status', ModelDataStatusEnum.waiting]
+    ];
+
     jsonRes(res, {
-      data: data.map((item) => item.textList).flat().length
+      data: {
+        splitDataQueue: data.map((item) => item.textList).flat().length,
+        embeddingQueue: await PgClient.count('modelData', {
+          where
+        })
+      }
     });
   } catch (err) {
     jsonRes(res, {

src/pages/api/model/data/splitData.ts

@@ -1,6 +1,6 @@
 import type { NextApiRequest, NextApiResponse } from 'next';
 import { jsonRes } from '@/service/response';
-import { connectToDatabase, SplitData, Model } from '@/service/mongo';
+import { connectToDatabase, SplitData } from '@/service/mongo';
 import { authModel, authToken } from '@/service/utils/auth';
 import { generateVector } from '@/service/events/generateVector';
 import { generateQA } from '@/service/events/generateQA';
@@ -65,7 +65,7 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse)
 export const config = {
   api: {
     bodyParser: {
-      sizeLimit: '10mb'
+      sizeLimit: '100mb'
     }
   }
 };

src/pages/model/components/detail/components/ModelDataCard.tsx

@@ -88,7 +88,7 @@ const ModelDataCard = ({ modelId, isOwner }: { modelId: string; isOwner: boolean
     onClose: onCloseSelectCsvModal
   } = useDisclosure();
 
-  const { data: splitDataLen = 0, refetch } = useQuery(
+  const { data: { splitDataQueue = 0, embeddingQueue = 0 } = {}, refetch } = useQuery(
     ['getModelSplitDataList'],
     () => getModelSplitDataListLen(modelId),
     {
@@ -109,7 +109,7 @@ const ModelDataCard = ({ modelId, isOwner }: { modelId: string; isOwner: boolean
 
   useQuery(['refetchData'], () => refetchData(pageNum), {
     refetchInterval: 5000,
-    enabled: splitDataLen > 0
+    enabled: splitDataQueue > 0 || embeddingQueue > 0
   });
 
   // 获取所有的数据，并导出 json
@@ -186,8 +186,12 @@ const ModelDataCard = ({ modelId, isOwner }: { modelId: string; isOwner: boolean
         )}
       </Flex>
       <Flex mt={4}>
-        {isOwner && splitDataLen > 0 && (
+        {isOwner && (splitDataQueue > 0 || embeddingQueue > 0) && (
-          <Box fontSize={'xs'}>{splitDataLen}条数据正在拆分，请耐心等待...</Box>
+          <Box fontSize={'xs'}>
+            {splitDataQueue > 0 ? `${splitDataQueue}条数据正在拆分，` : ''}
+            {embeddingQueue > 0 ? `${embeddingQueue}条数据正在生成索引，` : ''}
+            请耐心等待...
+          </Box>
         )}
         <Box flex={1} />
         <Input