tensorflow-wavenet
tensorflow-wavenet copied to clipboard
ibab
Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error.
Using default logdir: ./logdir\train\2017-04-16T14-17-39
Traceback (most recent call last):
File "E:/Spring 2017/CS 523/project 3/part-3/tensorflow-wavenet-master/tensorflow-wavenet-master/train.py", line 337, in
I'm not a Windows user, but maybe you need to use .\logdir\train\2017-04-16T14-17-39 instead of ./logdir\train\2017-04-16T14-17-39 ? It looks like this is the default logdir though, so try specifying the logdir directly as a command line argument.
I am also having a similar issue: but with ./logdir
Using default logdir: ./logdir/train/2017-12-03T23-24-51
Traceback (most recent call last):
File "train.py", line 337, in
I had that problem. Can't remember the solution. Try running the program with sudo.
Devin Roth
On Dec 3, 2017, at 3:25 PM, racheljchen [email protected] wrote:
I am also having a similar issue:\ but with ./logdir
Using default logdir: ./logdir/train/2017-12-03T23-24-51 Traceback (most recent call last): File "train.py", line 337, in main() File "train.py", line 263, in main writer = tf.summary.FileWriter(logdir) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/summary/writer/writer.py", line 336, in init filename_suffix) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/summary/writer/event_file_writer.py", line 67, in init gfile.MakeDirs(self._logdir) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/lib/io/file_io.py", line 367, in recursive_create_dir pywrap_tensorflow.RecursivelyCreateDir(compat.as_bytes(dirname), status) File "/usr/lib/python2.7/contextlib.py", line 24, in exit self.gen.next() File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/errors_impl.py", line 466, in raise_exception_on_not_ok_status pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.PermissionDeniedError: ./logdir
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.
realized I was in the wrong directory for tensorflow-wavenet (was using another implementation of it). switched to the folder pulled directly from cloning this project and it didn't give me this problem anymore.
I also have such issue ,but resolved by below command python -m retrain --bottleneck_dir=bottlenecks --how_many_training_steps=500 --model_dir=models --summaries_dir=tf_files --output_graph=retrained.pb --output_labels=retrained_labels.txt --architecture="mobilenet_0.50_224" --image_dir=flower_photos
above is my command run successfully in windows 10
I think this is simply a 'path' issue. In Windows, just modify the Log directory ROOT from LOGDIR_ROOT = ./logdir to LOGDIR_ROOT = 'logdir'
In Windows, create folder manually.
please can you explain further, do you mean the logs folder should be created manually
@hallibay Yes, I had this issue too on my Windows 10 machine. My error looked like this:
NotFoundError: Failed to create a directory: ./23_logs\train\plugins\profile\2019-11-11_13-02-07; No such file or directory
So, for me, within the directory of my project's .py file, or whichever folder I'm running the program from, I needed to manually create a folder called 23_logs, then within that folder, manually create a new folder called train, and then within the train folder, create a new folder called plugins, and again for profile. Then I just re-ran the program and it worked. Not sure why TF doesn't just make the directory's as it needs but whatever. Hope that helps?
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error. ... pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
Fix that worked for me in Windows10
Instead of path suggested by TF tutorial:
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
Use this:
log_dir="logs\\fit\\" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
I found out the solution.
just replace the path with "r" in the beginning in the code wherever the path strings are mentioned.
ex.
instead of :
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
try this:
log_dir=r"logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error. ... pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
Fix that worked for me in Windows10 Instead of path suggested by TF tutorial:
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")Use this:
log_dir="logs\\fit\\" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
This worked for me..thanks@StenkinVlad
Failed to create a directory: Logs\20200303/27/20-013517\train; No such file or directory [Op:CreateSummaryFileWriter]
I am using windows 10. Found the solution for this: log_dir=os.path.join('logs',datetime.datetime.now().strftime("%Y%m%D")) change it to single quotes.
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error. ... pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
Fix that worked for me in Windows10 Instead of path suggested by TF tutorial:
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")Use this:
log_dir="logs\\fit\\" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
this worked for me.
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error. ... pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
Fix that worked for me in Windows10 Instead of path suggested by TF tutorial:
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")Use this:
log_dir="logs\\fit\\" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
I had the same issue with the log file location for tensorboard. The following quick fix worked for me on Win10:
- use pathlib to define your paths: from pathlib import Path
- provide Path object for tensorboard:
target_dir_tb = Path.cwd() / "logs" / ... # specify location for tensorboard files in a Path object
tb = TensorBoard(log_dir=target_dir_tb, # this must be a Path object!
histogram_freq=15,
batch_size=batch_size, write_graph=True, write_grads=True)
Doesn't it work if i have used os instead of path ? I have used all these things to set path but it still seems to be taking the wrong path
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error. ... pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
Fix that worked for me in Windows10 Instead of path suggested by TF tutorial:
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")Use this:
log_dir="logs\\fit\\" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
@NEMO-WXL replacing the slashes worked. What is the purpose behind the replacement? Is this a windows specific fix?
I am having the following error when trying to train the VCTK-Corpus. I am using tensorflow version "0.12.1". Does anybody knows how to resolve this error. ... pywrap_tensorflow.TF_GetCode(status)) tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./logdir\train\2017-04-16T14-17-39
Fix that worked for me in Windows10 Instead of path suggested by TF tutorial:
log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")Use this:log_dir="logs\\fit\\" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")This worked for me..thanks@StenkinVlad
I think this is simply a 'path' issue. In Windows, just modify the Log directory ROOT from LOGDIR_ROOT = ./logdir to LOGDIR_ROOT = 'logdir'
Thanks :)
This bug still exists on Windows:
tbCall = TensorBoard(log_dir='./log/1000')
turns out to be :
tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./log/1000\train; No such file or directory
Anyone can help? Thanks!
What helped me, is wrapping the path with Path object (from pathlib module) and converting back to str
