最新消息:雨落星辰是一个专注网站SEO优化、网站SEO诊断、搜索引擎研究、网络营销推广、网站策划运营及站长类的自媒体原创博客

tensorflow - Tensorboard callback on model.evaluate() - Stack Overflow

programmeradmin12浏览0评论

I'm working with Tensorflow and Keras and trying to get the output of my model into Tensorboard for easy graph generation and visualization. This has worked fine so far for model.fit's callback, but model.evaluate does not have any sort of writing to Tensorboard. I looked on StackOverflow and I found one other person with this issue, but they never got a response.

This is the relevant part of my code, and I can give more as needed.


log_dir1 = 'C:/Users/myname/Desktop/DATASET/logs'

tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir1, histogram_freq=1)

history = model.fit(train_data_keras, epochs=1, validation_data=valid_data_keras, callbacks= [tensorboard_callback])


model.evaluate(test_data_keras, callbacks = [tensorboard_callback])
发布评论

评论列表(0)

  1. 暂无评论