开源代码的文件格式和星火文件格式有差异,使用python将开源格式文件转化成星火文件:
import json import jsonlines x=[] with open('SFT_智能家居_small.jsonl', 'r',encoding='utf-8') as f: for line in f: data = json.loads(line) if 'instruction' in data: del data['instruction'] if 'output' in data: data['target']=data.pop('output') x.append(data) with jsonlines.open('outputxml0408.jsonl','w')as file: file.write_all(x) import pandas as pd # 读取数据 data = pd.DataFrame(x) # 将数据保存为Excel文件 data.to_excel('outputxml0408.xlsx', index=False)
import json import jsonlines x=[] with open('SFT_智能家居_small.jsonl', 'r',encoding='utf-8') as f: for line in f: data = json.loads(line) if 'instruction' in data: del data['instruction'] if 'output' in data: data['target']=data.pop('output') x.append(data) with jsonlines.open('outputxml0408.jsonl','w')as file: file.write_all(x) import pandas as pd # 读取数据 data = pd.DataFrame(x) # 将数据保存为Excel文件 data.to_excel('outputxml0408.xlsx', index=False)