fix: 修复评估时模型输出json格式不对导致读取错误的问题 (#133)

* feature: add cot data evaluation function * fix: added verification to evaluation results * fix: fix the prompt for evaluating * fix: 修复当评估结果为空导致读取失败的问题
2025-12-04 18:49:50 +08:00
parent 31c4966608
commit 744d15ba24
14 changed files with 373 additions and 219 deletions
--- a/frontend/src/pages/DataEvaluation/Create/CreateTask.tsx
+++ b/frontend/src/pages/DataEvaluation/Create/CreateTask.tsx
@@ -1,3 +1,4 @@
+// TypeScript
 import React, { useState, useEffect } from 'react';
 import { Button, Form, Input, Select, message, Modal, Row, Col, Table, Space } from 'antd';
 import { EyeOutlined } from '@ant-design/icons';
@@ -36,6 +37,7 @@ interface CreateTaskModalProps {

 const TASK_TYPES = [
  { label: 'QA评估', value: 'QA' },
+  { label: 'COT评估', value: 'COT' },
 ];

 const EVAL_METHODS = [
@@ -55,7 +57,7 @@ const CreateTaskModal: React.FC<CreateTaskModalProps> = ({ visible, onCancel, on
    dimension: '',
    description: ''
  });
-  const [taskType, setTaskType] = useState<string>("QA");
+  const [taskType, setTaskType] = useState<string>(DEFAULT_TASK_TYPE);
  const [promptTemplates, setPromptTemplates] = useState<PromptTemplate[]>([]);
  const [previewVisible, setPreviewVisible] = useState(false);
  const [evaluationPrompt, setEvaluationPrompt] = useState('');
@@ -82,9 +84,24 @@ const CreateTaskModal: React.FC<CreateTaskModalProps> = ({ visible, onCancel, on
      fetchDatasets().then();
      fetchModels().then();
      fetchPromptTemplates().then();
+      // sync form with local taskType default
+      form.setFieldsValue({ taskType: DEFAULT_TASK_TYPE });
    }
  }, [visible]);

+  // when promptTemplates or taskType change, switch dimensions to template defaults (COT/QA)
+  useEffect(() => {
+    if (!promptTemplates || promptTemplates.length === 0) return;
+    const template = promptTemplates.find(t => t.evalType === taskType);
+    if (template && template.defaultDimensions) {
+      setDimensions(template.defaultDimensions.map((dim: any, index: number) => ({
+        key: `dim-${index}`,
+        dimension: dim.dimension,
+        description: dim.description
+      })));
+    }
+  }, [taskType, promptTemplates]);
+
  const fetchDatasets = async () => {
    try {
      const { data } = await queryDatasetsUsingGet({ page: 1, size: 1000 });
@@ -106,31 +123,46 @@ const CreateTaskModal: React.FC<CreateTaskModalProps> = ({ visible, onCancel, on
  };

  const formatDimensionsForPrompt = (dimensions: Dimension[]) => {
-    let result = "\n";
+    let result = "";
    dimensions.forEach((dim, index) => {
-      result += `### ${index + 1}. ${dim.dimension}\n**评估标准：**\n${dim.description}\n\n`;
+      if (index > 0) {
+        result += "\n";
+      }
+      result += `### ${index + 1}. ${dim.dimension}\n**评估标准：**\n${dim.description}`;
+      if (index < dimensions.length - 1) {
+        result += "\n";
+      }
    });
    return result;
  };

  const formatResultExample = (dimensions: Dimension[]) => {
-    return dimensions.map(dim => `\n    "${dim.dimension}": "Y",`).join('');
+    let result = "";
+    dimensions.forEach((dim, index) => {
+      if (index > 0) {
+        result += "\n    ";
+      }
+      result += `"${dim.dimension}": "Y"`;
+      if (index < dimensions.length - 1) {
+        result += ",";
+      }
+    });
+    return result;
  };

  const fetchPromptTemplates = async () => {
    try {
      const response = await queryPromptTemplatesUsingGet();
-      const templates: PromptTemplate[] = response.data?.templates
-      setPromptTemplates(templates)
-      if (taskType) {
-        const template = templates.find(t => t.evalType === taskType);
-        if (template) {
-          setDimensions(template.defaultDimensions.map((dim: any, index: number) => ({
-            key: `dim-${index}`,
-            dimension: dim.dimension,
-            description: dim.description
-          })));
-        }
+      const templates: PromptTemplate[] = response.data?.templates || [];
+      setPromptTemplates(templates);
+      // if a template exists for current taskType, initialize dimensions (handled also by useEffect)
+      const template = templates.find(t => t.evalType === taskType);
+      if (template) {
+        setDimensions(template.defaultDimensions.map((dim: any, index: number) => ({
+          key: `dim-${index}`,
+          dimension: dim.dimension,
+          description: dim.description
+        })));
      }
    } catch (error) {
      console.error('Error fetching prompt templates:', error);
@@ -144,8 +176,11 @@ const CreateTaskModal: React.FC<CreateTaskModalProps> = ({ visible, onCancel, on
      return;
    }
    const template = promptTemplates.find(t => t.evalType === taskType);
-    setEvaluationPrompt(template?.prompt.replace("{dimensions}", formatDimensionsForPrompt(dimensions))
-      .replace('{result_example}', formatResultExample(dimensions)));
+    const basePrompt = template?.prompt || '';
+    const filled = basePrompt
+      .replace('{dimensions}', formatDimensionsForPrompt(dimensions))
+      .replace('{result_example}', formatResultExample(dimensions));
+    setEvaluationPrompt(filled);
    setPreviewVisible(true);
  };

@@ -243,6 +278,13 @@ const CreateTaskModal: React.FC<CreateTaskModalProps> = ({ visible, onCancel, on
          evalMethod: DEFAULT_EVAL_METHOD,
          taskType: DEFAULT_TASK_TYPE,
        }}
+        onValuesChange={(changed) => {
+          if (changed.taskType) {
+            setTaskType(changed.taskType);
+            setEvaluationPrompt('');
+            setPreviewVisible(false);
+          }
+        }}
      >
        <Row gutter={16}>
          <Col span={12}>
--- a/frontend/src/pages/DataEvaluation/Detail/components/EvaluationItems.tsx
+++ b/frontend/src/pages/DataEvaluation/Detail/components/EvaluationItems.tsx
@@ -1,7 +1,8 @@
 import { useEffect, useState } from 'react';
-import { Table, Typography, Button, Space, Spin, Empty, message, Tooltip } from 'antd';
+import { Table, Typography, Button, Space, Empty, Tooltip } from 'antd';
 import { FolderOpen, FileText, ArrowLeft } from 'lucide-react';
 import { queryEvaluationFilesUsingGet, queryEvaluationItemsUsingGet } from '../../evaluation.api';
+import useFetchData from '@/hooks/useFetchData';

 const { Text } = Typography;

@@ -39,63 +40,52 @@ type EvalItem = {
 };

 export default function EvaluationItems({ task }: { task: any }) {
-  const [loadingFiles, setLoadingFiles] = useState<boolean>(false);
-  const [files, setFiles] = useState<EvalFile[]>([]);
-  const [filePagination, setFilePagination] = useState({ current: 1, pageSize: 10, total: 0 });
-
  const [selectedFile, setSelectedFile] = useState<{ fileId: string; fileName: string } | null>(null);
-  const [loadingItems, setLoadingItems] = useState<boolean>(false);
-  const [items, setItems] = useState<EvalItem[]>([]);
-  const [itemPagination, setItemPagination] = useState({ current: 1, pageSize: 10, total: 0 });

-  // Fetch files list
-  useEffect(() => {
-    if (!task?.id || selectedFile) return;
-    const fetchFiles = async () => {
-      setLoadingFiles(true);
-      try {
-        const res = await queryEvaluationFilesUsingGet({ taskId: task.id, page: filePagination.current, size: filePagination.pageSize });
-        const data = res?.data;
-        const list: EvalFile[] = data?.content || [];
-        setFiles(list);
-        setFilePagination((p) => ({ ...p, total: data?.totalElements || 0 }));
-      } catch (e) {
-        message.error('加载评估文件失败');
-        console.error(e);
-      } finally {
-        setLoadingFiles(false);
-      }
-    };
-    fetchFiles();
-    // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [task?.id, filePagination.current, filePagination.pageSize, selectedFile]);
+  // 文件列表数据（使用 useFetchData），pageOffset=0 表示后端分页为 1 基
+  const {
+    loading: loadingFiles,
+    tableData: files,
+    pagination: filePagination,
+    setSearchParams: setFileSearchParams,
+  } = useFetchData<EvalFile>(
+    (params) => queryEvaluationFilesUsingGet({ taskId: task?.id, ...params }),
+    (d) => d as unknown as EvalFile,
+    30000,
+    false,
+    [],
+    0
+  );

-  // Fetch items of selected file
-  useEffect(() => {
-    if (!task?.id || !selectedFile) return;
-    const fetchItems = async () => {
-      setLoadingItems(true);
-      try {
-        const res = await queryEvaluationItemsUsingGet({
-          taskId: task.id,
-          page: itemPagination.current,
-          size: itemPagination.pageSize,
-          file_id: selectedFile.fileId,
-        });
-        const data = res?.data;
-        const list: EvalItem[] = data?.content || [];
-        setItems(list);
-        setItemPagination((p) => ({ ...p, total: data?.totalElements || 0 }));
-      } catch (e) {
-        message.error('加载评估条目失败');
-        console.error(e);
-      } finally {
-        setLoadingItems(false);
+  // 评估条目数据（使用 useFetchData），依赖选中文件
+  const {
+    loading: loadingItems,
+    tableData: items,
+    pagination: itemPagination,
+    setSearchParams: setItemSearchParams,
+    fetchData: fetchItems,
+  } = useFetchData<EvalItem>(
+    (params) => {
+      if (!task?.id || !selectedFile?.fileId) {
+        return Promise.resolve({ data: { content: [], totalElements: 0 } });
      }
-    };
-    fetchItems();
-    // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [task?.id, selectedFile?.fileId, itemPagination.current, itemPagination.pageSize]);
+      return queryEvaluationItemsUsingGet({ taskId: task.id, file_id: selectedFile.fileId, ...params });
+    },
+    (d) => d as unknown as EvalItem,
+    30000,
+    false,
+    [],
+    0
+  );
+
+  // 当选择文件变化时，主动触发一次条目查询，避免仅依赖 searchParams 变更导致未触发
+  useEffect(() => {
+    if (task?.id && selectedFile?.fileId) {
+      setItemSearchParams((prev: any) => ({ ...prev, current: 1 }));
+      // 立即拉取一次，保证点击后立刻出现数据
+      fetchItems();
+    }
+  }, [task?.id, selectedFile?.fileId]);

  const fileColumns = [
    {
@@ -228,19 +218,20 @@ export default function EvaluationItems({ task }: { task: any }) {
          dataSource={files}
          loading={loadingFiles}
          size="middle"
-          onRow={(record) => ({ onClick: () => setSelectedFile({ fileId: record.fileId, fileName: record.fileName }) })}
-          pagination={{
-            current: filePagination.current,
-            pageSize: filePagination.pageSize,
-            total: filePagination.total,
-            onChange: (current, pageSize) => setFilePagination({ current, pageSize, total: filePagination.total }),
-          }}
+          onRow={(record) => ({
+            onClick: () => {
+              setSelectedFile({ fileId: record.fileId, fileName: record.fileName });
+              // 切换文件时，重置条目表到第一页
+              setItemSearchParams((prev: any) => ({ ...prev, current: 1 }));
+            },
+          })}
+          pagination={filePagination}
        />
      ) : (
        <div className="flex flex-col gap-3">
          <div className="sticky top-0 z-10 bg-white py-2" style={{ borderBottom: '1px solid #f0f0f0' }}>
            <Space wrap>
-              <Button icon={<ArrowLeft size={16} />} onClick={() => { setSelectedFile(null); setItems([]); }}>
+              <Button icon={<ArrowLeft size={16} />} onClick={() => { setSelectedFile(null); }}>
                返回文件列表
              </Button>
              <Space>
@@ -257,12 +248,7 @@ export default function EvaluationItems({ task }: { task: any }) {
            dataSource={items}
            loading={loadingItems}
            size="middle"
-            pagination={{
-              current: itemPagination.current,
-              pageSize: itemPagination.pageSize,
-              total: itemPagination.total,
-              onChange: (current, pageSize) => setItemPagination({ current, pageSize, total: itemPagination.total }),
-            }}
+            pagination={itemPagination}
          />
        </div>
      )}
--- a/frontend/src/pages/DataEvaluation/Home/DataEvaluation.tsx
+++ b/frontend/src/pages/DataEvaluation/Home/DataEvaluation.tsx
@@ -82,6 +82,7 @@ export default function DataEvaluationPage() {
      label: '任务类型',
      options: [
        { value: 'QA', label: 'QA评估' },
+        { value: 'COT', label: 'COPT评估' },
      ],
    },
    {
@@ -89,7 +90,6 @@ export default function DataEvaluationPage() {
      label: '评估方式',
      options: [
        { value: 'AUTO', label: '自动评估' },
-        { value: 'MANUAL', label: '人工评估' },
      ],
    },
  ];