清理张量流摘要

作者: 哎？小查查
发布时间: 2024-12-17 02:05:29 (1月前)
转自：

2 条回复

0#
回复此人
v-star*위위 | 2019-08-31 10-32

<div class =“post-text”itemprop =“text”> <P> 允许您读取事件文件（存储摘要的位置）的功能是 <a href="https://www.tensorflow.org/api_docs/python/tf/train/summary_iterator" rel="nofollow noreferrer"> <code> tf.train.summary_iterator </code> </A> 。你可以尝试这样的事情： </p> <pre> <code> import tensorflow as tf tfevents_logdir = path_to_existing_event_logdir tfevents_filepath_new = path_to_new_event_file writer = tf.summary.FileWriter(tfevents_logdir) for e in tf.train.summary_iterator(tfevents_filepath_new): if e.step == 0 or e.step % 50 == 0: # or any other criterion but make sure # you keep events from step 0 writer.add_event(e) writer.close() </code> </pre> </DIV>

编辑

登录后才能参与评论