dBeker / Faster-RCNN-TensorFlow-Python3

Tensorflow Faster R-CNN for Windows/Linux and Python 3 (3.5/3.6/3.7)
MIT License
613 stars 329 forks source link

How to change the number of iterations during training? #102

Closed 18342960922 closed 5 years ago

18342960922 commented 5 years ago

I only saw" max_iters" in config.py.what dose it mean?

morpheusthewhite commented 5 years ago

Stop spamming issues.

18342960922 commented 5 years ago

sorry I really don't know how to change the number of iterations. Neural networks have only recently been learned

morpheusthewhite commented 5 years ago

It's the number of times the neural net is trained on a batch

18342960922 commented 5 years ago

I am sorry to bother you.I know that every time I train a picture, I want to know if the data in the training set has been trained.

------------------ 原始邮件 ------------------ 发件人: "morpheusthewhite"notifications@github.com; 发送时间: 2019年9月6日(星期五) 下午4:33 收件人: "dBeker/Faster-RCNN-TensorFlow-Python3"Faster-RCNN-TensorFlow-Python3@noreply.github.com; 抄送: "没时间当网红"2283934965@qq.com; "Author"author@noreply.github.com; 主题: Re: [dBeker/Faster-RCNN-TensorFlow-Python3] How to change the numberof iterations during training? (#102)

It's the number of times the neural net is trained on a batch

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

morpheusthewhite commented 5 years ago

Explain yourself better

18342960922 commented 5 years ago

I don't see the epoch parameter.I don't know what epoch is equal to.How many 1000 pictures in the training set can be trained?

------------------ 原始邮件 ------------------ 发件人: "morpheusthewhite"notifications@github.com; 发送时间: 2019年9月6日(星期五) 下午5:05 收件人: "dBeker/Faster-RCNN-TensorFlow-Python3"Faster-RCNN-TensorFlow-Python3@noreply.github.com; 抄送: "没时间当网红"2283934965@qq.com; "Author"author@noreply.github.com; 主题: Re: [dBeker/Faster-RCNN-TensorFlow-Python3] How to change the numberof iterations during training? (#102)

Explain yourself better

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

18342960922 commented 5 years ago

batch = ? I didn't see this parameter.

------------------ 原始邮件 ------------------ 发件人: "morpheusthewhite"notifications@github.com; 发送时间: 2019年9月6日(星期五) 下午4:33 收件人: "dBeker/Faster-RCNN-TensorFlow-Python3"Faster-RCNN-TensorFlow-Python3@noreply.github.com; 抄送: "没时间当网红"2283934965@qq.com; "Author"author@noreply.github.com; 主题: Re: [dBeker/Faster-RCNN-TensorFlow-Python3] How to change the numberof iterations during training? (#102)

It's the number of times the neural net is trained on a batch

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

morpheusthewhite commented 5 years ago

Looking at config.py in lib/config/ you can see

tf.app.flags.DEFINE_integer('batch_size', 256, "Network batch size during training")
tf.app.flags.DEFINE_integer('max_iters', 40000, "Max iteration")  # the number of epochs