You've already forked DataMate
210 lines
5.6 KiB
TypeScript
210 lines
5.6 KiB
TypeScript
// Add mock files data
|
|
export const mockFiles = [
|
|
{ id: "file1", name: "dataset_part_001.jsonl", size: "2.5MB", type: "JSONL" },
|
|
{ id: "file2", name: "dataset_part_002.jsonl", size: "2.3MB", type: "JSONL" },
|
|
{ id: "file3", name: "dataset_part_003.jsonl", size: "2.7MB", type: "JSONL" },
|
|
{ id: "file4", name: "training_data.txt", size: "1.8MB", type: "TXT" },
|
|
{ id: "file5", name: "validation_set.csv", size: "856KB", type: "CSV" },
|
|
{ id: "file6", name: "test_samples.json", size: "1.2MB", type: "JSON" },
|
|
{ id: "file7", name: "raw_text_001.txt", size: "3.1MB", type: "TXT" },
|
|
{ id: "file8", name: "raw_text_002.txt", size: "2.9MB", type: "TXT" },
|
|
];
|
|
|
|
export const mockSynthesisTasks: SynthesisTask[] = [
|
|
{
|
|
id: 1,
|
|
name: "文字生成问答对_判断题",
|
|
type: "qa",
|
|
status: "completed",
|
|
progress: 100,
|
|
sourceDataset: "orig_20250724_64082",
|
|
targetCount: 1000,
|
|
generatedCount: 1000,
|
|
createdAt: "2025-01-20",
|
|
template: "判断题生成模板",
|
|
estimatedTime: "已完成",
|
|
quality: 95,
|
|
},
|
|
{
|
|
id: 2,
|
|
name: "知识蒸馏数据集",
|
|
type: "distillation",
|
|
status: "running",
|
|
progress: 65,
|
|
sourceDataset: "teacher_model_outputs",
|
|
targetCount: 5000,
|
|
generatedCount: 3250,
|
|
createdAt: "2025-01-22",
|
|
template: "蒸馏模板v2",
|
|
estimatedTime: "剩余 15 分钟",
|
|
quality: 88,
|
|
},
|
|
{
|
|
id: 3,
|
|
name: "多模态对话生成",
|
|
type: "multimodal",
|
|
status: "failed",
|
|
progress: 25,
|
|
sourceDataset: "image_text_pairs",
|
|
targetCount: 2000,
|
|
generatedCount: 500,
|
|
createdAt: "2025-01-23",
|
|
template: "多模态对话模板",
|
|
errorMessage: "模型API调用失败，请检查配置",
|
|
},
|
|
{
|
|
id: 4,
|
|
name: "金融问答数据生成",
|
|
type: "qa",
|
|
status: "pending",
|
|
progress: 0,
|
|
sourceDataset: "financial_qa_dataset",
|
|
targetCount: 800,
|
|
generatedCount: 0,
|
|
createdAt: "2025-01-24",
|
|
template: "金融问答模板",
|
|
estimatedTime: "等待开始",
|
|
quality: 0,
|
|
},
|
|
{
|
|
id: 5,
|
|
name: "医疗文本蒸馏",
|
|
type: "distillation",
|
|
status: "paused",
|
|
progress: 45,
|
|
sourceDataset: "medical_corpus",
|
|
targetCount: 3000,
|
|
generatedCount: 1350,
|
|
createdAt: "2025-01-21",
|
|
template: "医疗蒸馏模板",
|
|
estimatedTime: "已暂停",
|
|
quality: 92,
|
|
},
|
|
];
|
|
|
|
export const mockTemplates: Template[] = [
|
|
{
|
|
id: 1,
|
|
name: "判断题生成模板",
|
|
type: "preset",
|
|
category: "问答对生成",
|
|
prompt: `根据给定的文本内容，生成一个判断题。
|
|
|
|
文本内容：{text}
|
|
|
|
请按照以下格式生成：
|
|
1. 判断题：[基于文本内容的判断题]
|
|
2. 答案：[对/错]
|
|
3. 解释：[简要解释为什么这个答案是正确的]
|
|
|
|
要求：
|
|
- 判断题应该基于文本的核心内容
|
|
- 答案必须明确且有依据
|
|
- 解释要简洁清晰`,
|
|
variables: ["text"],
|
|
description: "根据文本内容生成判断题，适用于教育和培训场景",
|
|
usageCount: 156,
|
|
lastUsed: "2025-01-20",
|
|
quality: 95,
|
|
},
|
|
{
|
|
id: 2,
|
|
name: "选择题生成模板",
|
|
type: "preset",
|
|
category: "问答对生成",
|
|
prompt: `基于以下文本，创建一个多选题：
|
|
|
|
{text}
|
|
|
|
请按照以下格式生成：
|
|
问题：[基于文本的问题]
|
|
A. [选项A]
|
|
B. [选项B]
|
|
C. [选项C]
|
|
D. [选项D]
|
|
正确答案：[A/B/C/D]
|
|
解析：[详细解释]
|
|
|
|
要求：
|
|
- 问题要有一定难度
|
|
- 选项要有迷惑性
|
|
- 正确答案要有充分依据`,
|
|
variables: ["text"],
|
|
description: "生成多选题的标准模板，适用于考试和评估",
|
|
usageCount: 89,
|
|
lastUsed: "2025-01-19",
|
|
quality: 92,
|
|
},
|
|
{
|
|
id: 3,
|
|
name: "知识蒸馏模板",
|
|
type: "preset",
|
|
category: "蒸馏数据集",
|
|
prompt: `作为学生模型，学习教师模型的输出：
|
|
|
|
输入：{input}
|
|
教师输出：{teacher_output}
|
|
|
|
请模仿教师模型的推理过程和输出格式，生成相似质量的回答。
|
|
|
|
要求：
|
|
- 保持教师模型的推理逻辑
|
|
- 输出格式要一致
|
|
- 质量要接近教师模型水平`,
|
|
variables: ["input", "teacher_output"],
|
|
description: "用于知识蒸馏的模板，帮助小模型学习大模型的能力",
|
|
usageCount: 234,
|
|
lastUsed: "2025-01-22",
|
|
quality: 88,
|
|
},
|
|
{
|
|
id: 4,
|
|
name: "金融问答模板",
|
|
type: "custom",
|
|
category: "问答对生成",
|
|
prompt: `基于金融领域知识，生成专业问答对：
|
|
|
|
参考内容：{content}
|
|
|
|
生成格式：
|
|
问题：[专业的金融问题]
|
|
答案：[准确的专业回答]
|
|
关键词：[相关金融术语]
|
|
|
|
要求：
|
|
- 问题具有实用性
|
|
- 答案准确专业
|
|
- 符合金融行业标准`,
|
|
variables: ["content"],
|
|
description: "专门用于金融领域的问答对生成",
|
|
usageCount: 45,
|
|
lastUsed: "2025-01-18",
|
|
quality: 89,
|
|
},
|
|
{
|
|
id: 5,
|
|
name: "医疗蒸馏模板",
|
|
type: "custom",
|
|
category: "蒸馏数据集",
|
|
prompt: `医疗知识蒸馏模板：
|
|
|
|
原始医疗文本：{medical_text}
|
|
专家标注：{expert_annotation}
|
|
|
|
生成医疗知识点：
|
|
1. 核心概念：[提取关键医疗概念]
|
|
2. 临床意义：[说明临床应用价值]
|
|
3. 注意事项：[重要提醒和禁忌]
|
|
|
|
要求：
|
|
- 确保医疗信息准确性
|
|
- 遵循医疗伦理规范
|
|
- 适合医学教育使用`,
|
|
variables: ["medical_text", "expert_annotation"],
|
|
description: "医疗领域专用的知识蒸馏模板",
|
|
usageCount: 67,
|
|
lastUsed: "2025-01-21",
|
|
quality: 94,
|
|
},
|
|
];
|