New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
TensorFlow version 1.0.0-rc2 on Windows: “OpKernel ('op: ”BestSplits“ device_type: ”CPU“') for unknown op: BestSplits” with test code #7500
Comments
@mrry might have a clue. |
As far as I can tell this is fixed at HEAD, but didn't make it into the release candidate. Fortunately you can ignore this message (unless you want to use |
Could you verify that this works @Franck-Dernoncourt and close it if so? |
I just ran into this error on the 1.0 release linked to on the new install guide. Too soon for fix to get there? |
I tested and can confirm, as @mrry pointed out, that today's nightly build worked. I just received those SSE warnings which are unrelated.
|
Hi, I am also facing similar type of issue. I am running tensorflow 1.0 on windows 10. When I run the following program **import numpy as np #one real values column estimator = tf.contrib.learn.LinearRegressor(feature_columns=features) dataSet = tf.contrib.learn.datasets.base.Dataset( estimator.fit(x=dataSet.data, y=dataSet.target, steps=1000) estimator.evaluate(x=dataSet.data, y=dataSet.target)** I get the following error messages. WARNING:tensorflow:Using temporary folder as model directory: C:\Users\CRCV\AppData\Local\Temp\tmp2rbt5gl8 any solution?? " |
Running the same code as @sajjo79 and seeing the same issue using tf 1.0.0 on Windows 10/Python 3.5.2/CUDA 8.0/cuDNN 5.1 Going to try installing the nightly EDIT: Nightly build has the same issue |
@sajjo79 It looks like most of the issues in your post are due to 2 things. First is the display issues that opkernal.cc is outputting. The second is deprecated features being used. For the deprecated features the the warning message shows the code changes you will need to make. And as @mrry mentions upgrading to nightly should fixed the opkernel issues until the update gets pushed to the main. Also, https://www.tensorflow.org/install/migration contains a list of breaking changes in 1.0. I had to update several of my notebooks because things have moved around, got renamed, or args were changed. |
@JerryKurata I agree with you on the possible fixes. This is tutorial code though. It's strange it doesn't "just work" |
Seeing the exact same issue as @Franck-Dernoncourt with the Windows installation test code using the 1.0 release on Windows 10 Anniversary Edition. |
@maxamante Any time your write a tutorial it is the best your can do at the time. But, over time things change and there is only so much time to keep up the documentation. FWIW, does anyone know if the TensorFlow team is looking for help with the documentation, tutorials, etc on tensorflow.org? |
For everybody replying with this same issue, the fix was already provided with @maxamante Could you please give more details on what you did because the nightly build (85 with GPU support) worked for me. @JerryKurata Not a TensorFlow member but I think PRs usually are welcome. If there is something you want to address open an issue and feel free to submit a PR. [0] http://ci.tensorflow.org/view/Nightly/job/nightly-win/85/ |
@maxamante This got a little confusing here. I think this issue could be closed when @Franck-Dernoncourt gives a follow up of his side and folks having other problems could open new issues. |
Thanks, installing today's nightly build (CPU version):
fixed the issue (no more But as @Carmezim remarked there are now some SSE warnings:
@aselle Should I close this issue and reopen one about the SSE warnings? |
@gunan fyi
…On Feb 15, 2017 5:03 PM, "Franck Dernoncourt" ***@***.***> wrote:
Installing today's nightly build
<http://ci.tensorflow.org/view/Nightly/job/nightly-win/85/> (CPU version)
fixed the issue (no more “OpKernel ('op: ”BestSplits“ device_type:
”CPU“') for unknown op: BestSplits” etc.).
But as @Carmezim <https://github.com/Carmezim> remarked there are now
some SSE warnings:
TensorFlow version: 1.0.0-rc2
b'Hello, TensorFlow!'
2017-02-15 19:56:22.688266: W c:\tf_jenkins\home\workspace\nightly-win\device\cpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE instructions, but these are available on your machine and could speed up CPU computations.
2017-02-15 19:56:22.688266: W c:\tf_jenkins\home\workspace\nightly-win\device\cpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE2 instructions, but these are available on your machine and could speed up CPU computations.
2017-02-15 19:56:22.689266: W c:\tf_jenkins\home\workspace\nightly-win\device\cpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations.
2017-02-15 19:56:22.689266: W c:\tf_jenkins\home\workspace\nightly-win\device\cpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2017-02-15 19:56:22.689266: W c:\tf_jenkins\home\workspace\nightly-win\device\cpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2017-02-15 19:56:22.689266: W c:\tf_jenkins\home\workspace\nightly-win\device\cpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#7500 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AT_Sbd3iNXDX7RJrYJ2Cx4ukIo-PkI3yks5rc6BYgaJpZM4MAjbp>
.
|
I see the same issue using the raw win32 interpreter (cmd). However, it's working for me in python IDLE (shell) |
I too faced the same issue. After installing nightly build, error gone. Now getting below warnings:-
|
Validate tf installationWindows CMDC:\Windows\system32>python
Python 3.5.3 Shell
|
Thanks @Carmezim :), and @Franck-Dernoncourt for confirming that this is fixed in the nightlies. I'm going to close this issue because the original problem has already been solved, and lock it for new discussion. Please open a new issue if you still have problems! |
I installed TensorFlow version 1.0.0-rc2 on Windows 7 SP1 x64 Ultimate (Python 3.5.2 |Anaconda custom (64-bit)) using:
When I try running the test script from https://web.archive.org/web/20170214034751/https://www.tensorflow.org/get_started/os_setup#test_the_tensorflow_installation in Eclipse 4.5 or in the console:
I obtain some error message:
Why?
I didn't have such issues with TensorFlow 0.12.1 (installed with
pip install tensorflow==0.12.1
):Stack Exchange thread: TensorFlow version 1.0.0-rc2 on Windows: "OpKernel ('op: "BestSplits" device_type: "CPU"') for unknown op: BestSplits" with test code
@drpngx
The text was updated successfully, but these errors were encountered: