Training seems to be very slow on Tesla K40 about fast-style-transfer

[lengstrom]

checkpoint_iterations=1 forces your program to save a checkpoint every iteration, so you should make that parameter larger if you want to train in a reasonable amount of time.

[michaelhuang74]

@lengstrom Thanks. I will change checkpoint_iterations to a larger number.

Additionally, I have a question regarding when the training is supposed to finish.
(1) Does the training process go through a fixed number of iterations, say, 10,000? If the training goes through a fixed number of iterations, where is the number of iterations defined? I went through style.py, but could not find this information. I am new to tensorflow and might overlook the code.
(2) If the above is not true, under what kind of condition does the training process finish?

Thanks.

[lengstrom]

Look at the epochs flag under docs.md, you should find the information you're looking for there.

[jinaviva]

Amazing project! Must use the Unix like OS（ Linux operating system） to running tensorflow?Can i use VMware Linux?Is there any suggestion. sorry,im a noob!

[lengstrom]

I don't know, I've never used VMWare, sorry!
ᐧ

On Mon, Nov 14, 2016 at 4:57 AM, jinaviva [email protected] wrote:

Amazing project! Must use the Unix like OS（ Linux operating system） to
running tensorflow?Can i use VMware Linux?Is there any suggestion. sorry,im
a noob!

—
You are receiving this because you modified the open/close state.
Reply to this email directly, view it on GitHub
#22 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAucIdCnoU0IEhRuvfV2FB16ZK29QpFhks5q-DCjgaJpZM4KqqCh
.

[Zulqurnain24]

Hello,
I have mac with following specifications:

MacBook Air (13-inch, Early 2015)
1.6 GHz Intel Core i5
4 GB 1600 MHz DDR3
Intel HD Graphics 6000 1536 MB

I have executed it but the process is really slow even for lesser than 1000 iterations, does Intel HD Graphics 6000 1536 MB supports cuda tensorflow or i need to buy a MacBook pro with radeon 450 or above?