模型评价指标问题咨询

快速开始中模型评估以及效果测试中对比模型的评测指标均为mirco-F1 score，请问可以设置输出其他的评测指标吗，诸如：precision、recall、accuracy？对于监督式的graphsage_supervised在评估时能将模型给测试数据集预测的标签输出吗，从而便于人为计算各项指标（准确率等）？

Review the quick-start model evaluation and effect-testing paths, including the supervised GraphSAGE evaluation flow mentioned in the issue. Determine whether precision, recall, and accuracy can be selected and whether predicted labels can be emitted; done means those outputs are available for the supervised test data.

Written by the indexing model from the issue text.