关于finetune后输出结果的困惑
Created by: mrhan36
对以下代码存在疑问： 在fintune and eval后，通过以下代码获取训练集，验证集和测试集的指标，但发现以下三段代码输出结果一样，请帮忙解答一下，谢谢！
第一段：
with task.phase_guard("train"): #train:训练 val, test:评估
print(task.best_score)
task.metrics_choices=["acc","f1","matthews"]
eval_avg_score, eval_avg_loss, eval_run_speed = task._calculate_metrics(run_states)
第2段：
with task.phase_guard("val"): #train:训练 val, test:评估
print(task.best_score)
task.metrics_choices=["acc","f1","matthews"]
eval_avg_score, eval_avg_loss, eval_run_speed = task._calculate_metrics(run_states)
第3段：
with task.phase_guard("test"): #train:训练 val, test:评估
print(task.best_score)
task.metrics_choices=["acc","f1","matthews"]
eval_avg_score, eval_avg_loss, eval_run_speed = task._calculate_metrics(run_states)
非常感谢！