from datasets import load_dataset
import json
def save_to_json(data, filename):
    """保存数据到 JSON 文件"""
    with open(filename, 'w', encoding='utf-8') as f:
        json.dump(data, f, ensure_ascii=False, indent=4)
    print(f"Saved to {filename}, data length: {len(data)}")

# 定义文件路径
file_path = "/opt/aps/workdir/sunshuang/deep_search/math_data/math_qwq_4524_add_prompt_token_4524.json"

# 加载数据集
dataset = load_dataset('json', data_files=file_path)['train']

# 按 seq_token_len 字段升序排序
sorted_dataset = dataset.sort("seq_token_len")

# 选择最小的 871 条数据
selected_dataset = sorted_dataset.select(range(871))

# 获取最后一条数据的 seq_token_len 值
last_seq_token_len = selected_dataset[-1]['seq_token_len']

# 输出结果
print(f"The seq_token_len of the last selected data is: {last_seq_token_len}")

# 保存筛选后的数据集为 JSON 文件
output_file_path = "/opt/aps/workdir/sunshuang/deep_search/math_data/selected_data_871.json"

# 将筛选得到的数据保存为json
selected_dataset = selected_dataset.to_list()

# 保存为json
save_to_json(selected_dataset, output_file_path)

print(f"Selected data has been saved to {output_file_path}")