The next online study group meetup is scheduled for Tuesday, 31.03., 18:00-20:00, via Skype.
I am looking forward to talk to you soon!
2 Likes
Is there a kind of agenda for the meetup tomorrow?
We will cover the lesson topics and will discuss everything else the fellows want to discuss, i.e., a fluid agenda depending on where the discussion goes.
If you have topics you want to discuss prepare something and we can look into that too.
Couldn’t join the meetup on 31st. I was let go by my employer yesterday (startups in the travel industry are getting massively hit by the COVID-19 situation). Haven’t seen the 2nd lecture yet. Looking forward to it 
Sorry to hear about that!
You can still easily catch up with our learning group, as we were only going through the basics during first two meetups. Just join us the next meetups if you find time for it. 
The email about the next meetup I will send out later.
1 Like
Hi Guys, I’m organizing an virtual study group from Amsterdam. I was just wondering whether everyone in your study group has access to the closed forum topics? In my case, only I have access and not the other participants, which makes it harder to colab on the forum.
Hi Bauke, we have the same situation in our study group.
Only the people with the invite can access the 2020 course forum threads.
The next online meetup is scheduled for Tuesday, 7.4.2020, 18:00-20:00.
Could we talk about deployment? I have tried binder and heroku so far but was not really successful in deploying the apps.
I could run my jupyter notebook from a github repo on binder but not the web-app via voila.
On Heroku I had already successfully deployed python-webapps with tha fastai1 framework but it does not work for me with version 2.
If you are curious these are the repos:
binder-voila: https://github.com/we-make-ai/saab-model-classifier-voila
starlette app: https://github.com/we-make-ai/saab-model-classifier
It’s a classifier for different models of the former Swedish car manufacturer Saab.
Sure, if you have time please feel free to prepare a notebook/slides etc. for our meetup tomorrow. Anyway, we can discuss this topic.
Well I can present my attempt but I was not successful
Yeah, sounds interesting and then we discuss it in the group.
Hello everybody,
I am not sure that I can join the learning group meetup today at 18:00 sharp but latest at 19:00 - can therefore somebody lead the learning group until then?
Please let me know & kind regards
Michael
The next meetup will be on Monday, 13.04., 18:00-20:00. 
Unfortunately I’ve not been able to make it to the Skype meetings, since my 4 month old son has been quite demanding of late 
Anyway, since we’re talking about deployment, I’ve just made a Bärlauch vs Maiglöckchen classifier here :
https://forums.fast.ai/t/share-your-v2-projects-here/65757/126?u=gautam_e
The notebook itself was extremely easy since I used the bear classifier code. Getting it to work on Binder was a totally different ball game. Perhaps we can put it on the agenda next time if there’s any interest. We can also just see how this weeks lecture goes and then take a call. I’m open either way.
2 Likes
The next online meetup via Skype will be on Wednesday, 22.04., 18:00-20:00!
Looking forward to meet you again!
1 Like
The next meetup will be on Wednesday, 29.04., 18:00-20:00.
We will start with a presentation on segmentation and then we will go through the lesson material.
1 Like
The next meetup for lesson 7 will be on Wednesday, 06.05., 18:00-20:00.
The last meetup for lesson 8 will be on Wednesday, 13.05., 18:00-20:00, via Skype.
Note: I had to change the meetup to Wednesday.
Hi!
Has anyone worked with German language models by now? I have trained a model but I have difficulties with creating meaningful texts by the help of this model.
https://colab.research.google.com/drive/1B6PZ5gf9bw9sDqzniAAxrdiyPjIQdhMH?usp=sharing
I publish the texts on the topic “Kurkuma / Curcurmin” on this fastpages blog: https://kurkuma.online/
Does anyone have ideas on how to improve the model’s quality?
Another thing is to create texts based on keywords. E.g.: create a text on these keywords [“Kurkuma”, “Gallensteine”, “Behandlung”]
By now the model creates rather funny texts ;=)