As described in jypyter notebook I tried:
import tensorflow
then my kernel is killed, like "The kernel appears to have died. It will restart automatically."
Tried to reinstall anaconda, numpy, tensorflow, didn't work.
well I also experienced this normally this can't work because tensorflow is too big for jupyter notebook though you may try downloading the anaconda navigator if you don't have it already though this may not work
Related
I bought a new Mac version 12.2 (OS Monterey) and installed Anaconda. Most of the Python packages can be installed correctly. However, Tensorflow (ver 2.8.0) and Keras (ver 2.8.0) have major issues. The Jupyter notebook kernel gets killed when tensorflow and/or keras get imported. I looked up various posts on Stackoverflow and Medium, however, nothing seems helpful. I even tried to convert the .ipynb to .py script, however, the same error occurs.
Is there anything that can be done to resolve this?
I am using Macbook Air with M1 chip. When trying to import tensorflow in Jupyter notebook, the kernel dies and displays a prompt that "Kernel has died and will restart in sometime". Could someone help me fix this?
Tensorflow version - 2.5.0
Python version - 3.8.8
Try running the notebook file within VS Code, there are extensions to help with that. Also check this article on how to install tf on M1 https://towardsdatascience.com/installing-tensorflow-on-the-m1-mac-410bb36b776
It seems this is a recurring issue with multiple people with the m1 macs. Since it is still fairly new, it is possible that Jupiter notebook still doesn't fully support it. Try using anaconda navigator with the windows emulator. Here is a link to a forum post with people having the same problem.
https://github.com/apple/tensorflow_macos/issues/45
Anaconda and upgrading to new M1 Mac
I installed tensorflow v2.1.0. The Jupyter notebook on my Mac showed lots of warnings already when I just imported tensorflow. See the screenshot below:
How do I fix the compatibility issue?
To start out I had a anaconda installation with python 3.5.2. After doing a pip tensorflow installation I got an error something like "nosetest...". I searched stackoverflow and got a suggestion to move to python 3.6.
So I uninstalled the whole of anaconda and python. Then did complete fresh install of anaconda with python 3.6.10. Did a pip install of tensforflow. launched spyder and jupyter and could not import tensorflow. However, when I issue commands from python command prompt things work fine.
So can't seem to get tensorflow to import in spypder or jupyter.
Lastly following another suggest I ran the c:/from/my/anaconda/Scripts/activate base. But that also didn't work in spyder. However in jupyter it seemed to import but then crashed on tf.version
with access violation.
Can you please help, I've tried many things and suggestions, can't seem to get tensorflow to work with python 3.5 or 3.6.
Thanks.
P.S. I remember using python3.5.2 sometime ago with the then tensorflow, I think version 1.1x and everything went smoothly, everything installed and imported worked beautifully.
Is it possible to run Tensorboard on a machine without CUDA support?
I'm working at a computation center (via ssh) which has two major clusters:
CPU-Cluster which is a general workhorse without CUDA support (no dedicated GPU)
GPU-Cluster with dedicated GPUs e.g. for running neural networks with tensorflow-gpu.
The access to the GPU-cluster is limited to Training etc. such that I can't afford to run Tensorboard on a machine with CUDA-support. Instead, I'd like to run Tensorboard on the CPU-Cluster.
With the TF bundled Tensorboard I get import errors due to missing CUDA support.
It seems reasonable that the official Tensorboard should have a mode for running with CPU-only. Is this true?
I've also found an inofficial standalone Tensorboard version (github.com/dmlc/tensorboard), does this work without CUDA-support?
Solved my problem: just install tensorflow instead of tensorflow-gpu.
Didn't work for me for a while due to my virtual environment (conda), which didn't properly remove tensorflow-gpu.
Tensorboard is not limited by whether a machine has GPU or not.
And as far as I know, what Tensorboard do is parsing events pb files and display them on web. There is not computing, so it doesn't need GPU.