zemushka / coursera-aml-docker Goto Github PK

Greetings

I am in a Windows 10 Professional PC using docker.

When I try to mount the image, run the docker and open Jupyter; I found myself in an empty repository.

I am taking the course of NPL too, that course's repository works well doing the same steps.

What can I do to expand the workaround of the bug? Maybe a Docker diagnose will share more information.

I am getting the following error message while running
'import download_utils' on Jupytor note book. But I could see the file "tqdm.py" in intro-to-dl folder. Unable to procede further. Any guidance will be highly appreciated. Thanks

ModuleNotFoundError Traceback (most recent call last)
in
----> 1 import download_utils

~\intro-to-dl\download_utils.py in
7 from functools import wraps
8 import traceback
----> 9 import tqdm_utils
10
11

~\intro-to-dl\tqdm_utils.py in
2 # -- coding: utf-8 --
3 from future import print_function
----> 4 import tqdm
5 tqdm.monitor_interval = 0 # workaround for tqdm/tqdm#481
6

ModuleNotFoundError: No module named 'tqdm'

I followed the instruction, I downloaded the image and then when I ran:
docker run -it -p 127.0.0.1:8080:8080 --name coursera-aml-1 zimovnov/coursera-aml-docker.
the jupyter tree was empty.
Is it the complete and right procedure, or I forgot some steps?

There is a conflict in the conda requirements. Keras 2.0.6 needs tensorflow <= 1.0.1 but tensorflow=1.2.1 is specified.

amls) Quasar:~ subhachandra$ conda install --yes --file conda_requirements.txt
Fetching package metadata .............
Solving package specifications: .

UnsatisfiableError: The following specifications were found to be in conflict:

• keras ==2.0.6 -> tensorflow <=1.0.1 -> protobuf ==3.0.0b2
• tensorflow ==1.2.1
  Use "conda info " to see the dependencies for each package.

Hello,

I've followed the instructions for running on GPU, done tons of research and I've gotten GPU support to work in other containers just not this one. Any assistance? (Week 4 takes a really long time, and I would like to see how it fares with my GPU)

Hey,

I can run the Jupyter notebook from within the Docker container, but in the 2nd week (v2) lessons, it says to go to run the tensorboard command in bash. I run this within the Jupyter terminal, and I can't figure out where to find the webpage...

Link https://www.tensorflow.org/versions/r1.2/install/install_linux#nvidia_requirements_to_run_tensorflow_with_gpu_support mentioned in README is broken.