ray.rllib.policy.sample_batch.SampleBatch.set_training# SampleBatch.set_training(training: bool | tensorflow.compat.v1.placeholder = True)[源代码]# 设置此 SampleBatch 的 is_training 标志。