Whenever I execute TensorBoard I just get:
Starting TensorBoard 54 at http://localhost:6006
(Press CTRL+C to quit)
and then nothing happens. Any advice on how to get the graph to show?
EDIT: Sorry I meant to clarify that I copy and paste "http://localhost:6006" into my browser and "No scalar data was found" appears.
FIXED: I was not typing the correct log directory. For anyone in the future who has this problem, don't be like me and assume that TensorBoard automatically reads through the /tmp directory.
One possible reason might be not giving the correct log directory, and at other times people often forget to write back summaries.
Related
I am trying to use Python to write an automation bot and need to continuously take screenshots of a certain area.
from PIL import ImageGrab
import time
def screenGrab():
box = (1, 1, 2174, 1248)
im = ImageGrab.grab(box)
im.save("current_screenshot.png", 'PNG')
# time.sleep(2)
Then in another module I will continuously call this screenGrab() function and use PIL to analyze the screenshot, until certain conditions are triggered and corresponding actions are taken. Then the loop of "taking screenshots - analyzing - waiting for a certain condition to be satisfied - take actions" starts again.
It can work pretty well for a while (sometimes a whole day, sometimes only three minutes, depending on how lucky I am), but when I am unlucky, I see the following error:
PermissionError: [Errno 13] Permission denied: 'current_screenshot.png'
Why does this happen? What should I do to prevent it?
I have done some research online and think it may be that the file hasn't been properly closed when I call another module to open this image and analyze it. However, I am still confused why it should happen at all. Shouldn't Python completely finish every command before it goes to the next? Before it runs whatever command to open the current_screenshot.png, it must have finished the im.save("current_screenshot.png", 'PNG') command, which should have properly closed the file, right? Then how can this error pop up?
By the way, it does happen less frequently when I un-comment the time.sleep(2) command to give my laptop more time to save the image, but doing this significantly reduces the efficiency of my codes.
I am trying to visualize a training session I trained in a remote server. I used scp to copy the file in my local iMac. I tried to visualize the data by running tensorboard. It runs the tensorboard site but I can't get the visualization. Every chart has a single dot at zero. I get this warning on the terminal.
WARNING:tensorflow:Unable to get first event timestamp for run
470_313_0.0001_2500_200/train
WARNING:tensorflow:Unable to get first event timestamp for run
470_313_0.0001_2500_200/train
WARNING:tensorflow:Unable to get first event timestamp for run
470_313_0.0001_2500_200/val
WARNING:tensorflow:Unable to get first event timestamp for run
470_313_0.0001_2500_50/train
WARNING:tensorflow:Unable to get first event timestamp for run
470_313_0.0001_2500_50/val
Any idea what is going on?
I ran into the same problem (with a TensorFlow 1.4 trainer running in the cloud with Google Cloud ML Engine).
Explicitly closing tf.summary.FileWriters with close() solved it in my case.
I ran into a similar problem. There are 2 solutions to it -
Delete all past tfevents files from the directory and keep the latest one (temporary solution).
Create a new directory for building your logs (permanent solution).
In given below picture, first I build logs in directory 2 which resulted in same error/warnings. Later I changed the directory to 3 and build logs there and tensorboard ran successfully.
In my case, the problem was the directory names for the runs were too long. After I manually renamed them, the problem is solved.
I've got a graph that isn't behaving as it should in CloudConnect.
I'm running it locally, and it's completing, but not doing its work.
In an effort to figure out why this is, I've added printLog calls in many places, like the following
printLog(warn, 'transfrom from file ' + $in.0.fileName);
printLog(debug, 'joining etc');
The Phase consists of a FileList into a SimpleCopy, into a LookupJoin, a Reformat (produce SQL) and a DBInsert.
However, while I see logs for phases above, I'm not seeing anything produced in the log for any part of my phase. All parts of the phase do report running successfully in log. I've also done Enable Debugging on all connections in this phase.
Am I missing something to enable logging? Is there a better way to debug processing in CloudConnect?
Discovered the problem - the FileList will succeed if the source file cannot be found, but none of the subsequent steps will then fire. It's somewhat unintuitive, since the log files says 'succeeded'.
For debugging, after run you can access the data by right clicking on the connection, and selecting "View Data"
Sorry for the elementary question, but documentation didn't seem to cover this clearly, at least for a GoodData noob. I'll leave it up for anyone with the same problem!
Ok, so in VB, I create and run a batch file. Pretty simple.
The problem is, the batch file starts another program that runs through cmd.
I want to get the output from the cmd program the batch file starts, any ideas?
Someone earlier said that i could do somthing with command parameters,
could you explain that to me if that is right?
I ask that you are not particularly harsh in giving a downvote becasue the question is worded very vaugue. To me it sounds like you are looking for something like this.
#echo hello >> log.txt
What this is going to do is force the output of whatever command you have typed to be sent into a text file named log.txt.
I have created a very useful application where I record network activity on my home network and save it into a log file using the same overall idea so I know it can be useful.
This isn't strictly a programming question, but hopefully someone here has the knowledge to help me.
Running Vista x64, the "Open File With" dialog takes forever to appear when trying to open a file with unknown type.
To be specific, the prompt giving the choice between "Use the web service to find the correct program" and "Select a program from a list of installed programs" appears immediately, but there is about a 10 second delay between pressing OK here and the next window (with a list of installed programs) appearing.
The same delay is present when I run "rundll32 shell32.dll,OpenAs_RunDLL" from a command line.
I'm guessing that this is not common behaviour, because I find it incredibly annoying, and I couldn't find any mention of it anywhere.
So..does anyone have any ideas what it might be? Or what I could do to try and find out?
VonC could be correct.
Read this article:
http://blogs.technet.com/markrussinovich/archive/2005/08/28/the-case-of-the-intermittent-and-annoying-explorer-hangs.aspx
You should use process explorer to watch the stack when you initiate the action. You should be able to tell what sort of operation is timing out.
Download process explorer here:
http://technet.microsoft.com/en-us/sysinternals/bb896653.aspx
You could check if you do not have have network connection declared (net use), or more generally any network paths which may be used by your list of program listed to "Open File With".
If a network path is unavailable, a 'timeout' delay can affect the GUI responsiveness, especially if that GUI must explore/list those paths.