Coding environment (Colab tips)

I didn’t get your question. Which course, week, and assignment/lab it is?

this is " Supervised Machine Learning: Regression and Classification" course. Week 3, optional Lab: Classification…

Okay. I tried and found the same thing. This is related to Colab and as the message says, “Support for third party widgets (widgets outside of the ipywidgets package) needs to be enabled separately. Support for these widgets will be loaded from a CDN external from Colab.”

The simple solution is to add this code before the cell where you see this weird message.

from google.colab import output
output.enable_custom_widget_manager()
3 Likes

It worked, thank you very much!

Once you ownloaded the images to google drive, you can run some python code like this:

#Choose your desired working directory (full path)
import os
os.chdir(my_working_directory)

#to show your image:
from PIL import Image
import matplotlib.pyplot as plt
#put all your images in a subdirectory called image
im = Image.open(“./image/your image.jpg”)
#to size the image as you choose
plt.figure(figsize=(10, 10))
plt.imshow(im)
plt.axis(‘off’) # Hide the axis
plt.show()

The image will not be in a markdown window, but you will see it.

Hello, Mr Simon do you mean I can run programs related to tensorflow having a CPU without any problem with jetbrain?
Anaconda is so frustrating

I will give it a try to see if it works for me all I want is to build good project :innocent:

I would like to know where you run your own programs sir.
:grin:

Hi @Ebong,

I generally use Conda to set up my python environments. You can read about it here.

Please be warned, it can vary from OS to OS.

1 Like

Do you mean I can install Anaconda from crash following the link above without any errors?

because I’ve gone through a lot of challenges to setup labs to run programs like TensorFlow without any success

It is not as easy, hence I cannot promise it will work, but it is worth a shot. Like I said, it varies from OS to OS.

And especially if you are using a Mac with silicon chip, then there’s no easy way to install previous versions of tensoflow like 2.3 or 2.2 easily. All latest versions of tensorflow are compatible with Macs with silicon chips.

I’d suggest to search online, YouTube or even LLMs like chatgpt to help you with this task. Mention the specifics of the OS you have, its version and what you are trying to install, and try doing so by creating python environments.

P.S I failed at this many times myself until I got familiar with it. And still, there are times when I fail, but that’s how you learn.

1 Like

Me too.

Good morning, TMosh
Has your solution worked?

No, it has not.

+1

@Ebong
Over the years I have completed DLS, MLS, NLP, Tensorflow, and AI for Medicine specializations, and run pretty much every single exercise locally on a Mac using custom environments created using Conda and pip. I don’t install anything natively directly onto the OS. I use pip list to find out what the exercise environment is using, then replicate that locally. Sometimes requires some finesse. Sometimes requires a mallet. Never couldn’t eventually get the environment sorted, though a few times could not get the data (some of the MRI data was too big to download) or an archive file (the trained YOLO model comes to mind)

There are threads on the forum where people have worked through these questions and issues to get environments established. Try search.

That said, I think cloud is a good place to start, unless you run into a particular reason you have to run local. More time doing AI, less time doing system admin.

2 Likes

well I’m not using a Mac, and sure that’s where my problems are from, HP etc just make things difficult for people to work with.
I just need someone who can show me how to run Anaconda to a Windows operating system where I can work with TensorFlow etc :face_exhaling:

Please does it give you the possibility to compute labs like tensorflow etc?

Notice both Mubsi and I refer to Conda, not Anaconda. They are adjacent, but not identical. Conda / Miniconda has a much smaller footprint. Windows install instructions here:

https://docs.conda.io/projects/conda/en/latest/user-guide/install/windows.html

You can read about installing TensorFlow for example here:

I think there are also threads in this forum that discuss the process on Windows, but I only use Mac OS so can’t advise on the details. One such thread is…

Well i thought Anaconda and Conda were the same thing :face_with_open_eyes_and_hand_over_mouth:
so you are advising me to go for Miniconda right?