helene-rousset/enigma_lora_qwen_0_5b_TEST_evaluation_sample

该数据集主要包含与商户交易相关的数据，涵盖了商户ID、时间戳、贷款ID、标签、批次、借款金额、支付金额、商户开户时间、文档类型、类别、子类描述、最近30天和7天的批准交易金额、最近90天的交易次数、平均批准交易金额、最高批准交易金额、拒绝交易百分比、欺诈交易次数、收入增长率、收入稳定性指数、最高和最低收入月份、平均余额、最小余额、转账次数和金额、账单支出百分比、授权码欺诈交易次数、资金不足交易次数、交易增长率、账户开户时间、活跃天数、平均分期付款金额、聊天对话次数、平均首次回复时间、会话时间、常客百分比、最近15天交易时间百分位数、不同卡令牌数量、批准交易百分比、平均批准交易金额、不同时间段的交易次数比例等。数据集分为训练集，包含1000个样本。

This dataset primarily contains data related to merchant transactions, including merchant ID, timestamp, loan ID, label, batch, borrowed amount, paid amount, months since merchant opening, document type, category, subclass description, approved transaction amounts in the last 30 and 7 days, transaction counts in the last 90 days, average approved transaction amount, maximum approved transaction amount, percentage of denied transactions, fraud transaction counts, revenue growth rate, revenue stability index, highest and lowest revenue months, average balance, minimum balance, transfer counts and amounts, percentage of revenue spent on bills, authorization code fraud transaction counts, insufficient funds transaction counts, transaction growth rate, account opening time, active days, average installment amount, chat conversation counts, average first reply time, session time, percentage of recurrent customers, percentile time since last transaction in the last 15 days, count of distinct card tokens, percentage of approved transactions, average approved transaction amount, and transaction count ratios in different time periods. The dataset is divided into a training set containing 1000 samples.