Hi,
I have the fastai installed on my Windows laptop. For deploying with huggingface, I need to install the fastai again under the ubuntu using fastsetup isn’t it?
Personally, I would want to use it on cloud. If you are a beginner, time spent configuring the laptop with dependencies and support libraries is better used in creating models.
But if you insists installing on laptop, you could follow the steps documented here.
1 Like
You just need to export the .pkl file from your model, once you’re happy that the model works correctly.
The Hugging Face part is from the internet. So, create a space on Hugging Face, clone the repo from there onto your local machine, make sure the .pkl file is included and the app.py file you also need to write and reference the .pkl file, and push back to the Hugging Face space.
Perhaps double-check that:
-
The filenames of the figures (including the extensions of the figures), matches your reference of them on the app.py file
-
Ensure that the figures are located in the same directory as the app.py file
2 Likes
From what I’ve learnt so far, there’s a host of tricks that can be employed to get your network to train better. It’s not necessarily a standard recipe that can be applied to all the tasks, after the initial training cycles it really depends on what the task is, what other tricks you’ve already applied and so on.
The book covers these methods in details over a bunch of chapters. So, that would really be my first recommendation, to go through these lessons, understand them and apply them while tracking the results.
1 Like
If you want to post what went wrong, I am aure there is a ‘legit’ way to fix it.
Happy to take a look
i love to use a utility called terminator. not the same as tmux but pretty convenient when I want multiple terminal sessions
1 Like
I installed the windows terminal and the wsl – install as explained in the video. However after rebooting I an not able to launch ubuntu terminal and getting the following message:
PS C:\Windows\system32>
wsl.exe
Windows Subsystem for Linux has no installed distributions.
Distributions can be installed by visiting the Microsoft Store:
https://aka.ms/wslstore
Have gone wrong in installation?
PS C:\Windows\system32> wsl --install
Installing: Virtual Machine Platform
Virtual Machine Platform has been installed.
Installing: Windows Subsystem for Linux
Windows Subsystem for Linux has been installed.
Downloading: WSL Kernel
Installing: WSL Kernel
WSL Kernel has been installed.
Downloading: GUI App Support
Installing: GUI App Support
GUI App Support has been installed.
Downloading: Ubuntu
The requested operation is successful. Changes will not be effective until the system is rebooted.
PS C:\Windows\system32>
Installing WSL2 is just like installing “Virtual Box” … so after installing this, you still need to pick a linux distro to install on top of it. If you go to Microsoft Store and search for Ubuntu you can see some distros that you can install. When it installs, it automatically uses the already installed WSL2 infrastructre.
HTH.
Thanks I did that and it is working now
1 Like
I am not familiar with Linux. So having some issues sorted out. I installed VSCode under Linux subsystem for windows. On running code . I got the following prompt:
To use Visual Studio Code with the Windows Subsystem for Linux, please install Visual Studio Code in Windows and uninstall the Linux version in WSL. You can then use the code command in a WSL terminal just as you would in a normal command prompt.
I uninstalled VSCode from linux and installed on Windows. But now code . is not being recognized by Linux.
Hi @tapashettisr , if you post help request in the relevant thread as this is not specific to Lesson 2 , it would be useful for others who come looking for help in those threads.
Also, can you try restarting your linux VM under WSL ?
Personally, my path of least resistance would be to just install everything including jupyter in your Linux instance, clone the fastbook repo, then start jupyter notbook from there. I’d be able to connect to it from windows side. But then again, I’m lazy by nature and try to minimize excitement in life that comes from having to mess around with hundreds of moving parts and rather get straight to training models which is the point of the course. 
This topic was covered in lesson 2.
It comes with Ubuntu - no need to separately install a distro.
Hi All,
I came across this, which is apt for the last lesson (lesson 2) of this past week. So, I thought I’d share it.
Thanks
Kind Regards,
Zakia
Ah Ok, thanks Jeremy, I didn’t know that. I installed this long time ago and I was running multiple distros so I recall going to Microsoft store to get different “apps” as Microsoft Store calls them. I had installed Ubuntu and Kali linux at the time.
I made the same mistake a few days ago. I managed to get it part fixed, but sorry I can’t remember exactly. Just can you confirm on Linux that the following returns nothing…
$ which code
My jupiter install was a bit screwed up, and easier to reinstall from scratch than troubleshoot. See my install log.
It gives /usr/bin/code. I am able to access VSCode from wsl terminal but not through Linux. That is not a major problem.
Now I am installing fastsetup and following the steps given in the GitHub - fastai/fastsetup: Setup all the things.
I am stuck at sudo ./ubuntu-initial.sh
The prompt says Enter hostname to set:
What does it mean and what should I enter here?
@suvash As you mentioned if error_rate once starts improved, and then goes on decreasing is that an indication of model over fitting as shown in the image below?
Full code:
1 Like