盘古大模型 PANGULARGEMODELS-排列特征重要性:请求示例

时间：2025-05-23 14:25:55

请求示例

{
  "data": [
    {
      "feature_1": xx,
      "feature_2": xx,
      ...
      "feature_n": xx,
      "target": 0,
    },
    ...
    {
      "feature_1": xx,
      "feature_2": xx,
      ...
      "feature_n": xx,
      "target": 1
    }
  ],
  "fi_args":{
    
"num_shuffle_sets": 1,
    "sample_num": 1000
  }
}

此处data列表的长度至少为1000，建议通过以下代码根据csv文件生成请求体（DATA_NUM为从csv文件中采样数据的数量）：

import random
import json
import pandas as pd

ENCODING_TYPE = "utf-8"
DATA_NUM = 1200

data_path = "path/to/xxx.csv"
df = pd.read_csv(data_path, encoding=ENCODING_TYPE)
data_list = df.to_dict("records")
data_list = random.sample(data_list, DATA_NUM)
request_data = {
    "data": data_list,
    "fi_args": {
        
"num_shuffle_sets": 1,
        "sample_num": 1000,
    }
}
request_str = json.dumps(request_data)
print(request_str)

上一篇：盘古大模型 PANGULARGEMODELS-排列特征重要性:URI

下一篇：盘古大模型 PANGULARGEMODELS-排列特征重要性:URI