LOL. That sounds like a good guess.
Hi Cedric,
If I want to build a language model at character level as I need to classify url to malicious or non-malicious hw to do with fast ai language modelling technique
what’s the apps name?
Well, you know it. 今日头条(Toutiao)
video and audio is clear
Has text classification been extended to text regression in the fastai library like was done with images?
What does this error mean?
SyntaxError: unexpected EOF while parsing
I get it when i try
def show_batch(data, rows:int=10):
I talked about the YouTube conspiracy theory example (as well as several other related issues) in my PyBay keynote.
haha. sounds familiar havent used yet.

Are there any recommendations or papers on how to counter this? I can think of using a model trained for one purpose to do predictions on other non related purposes, but that’s a convoluted way to observe and find biases. I’m wondering if more tools can be or are already built into the process of training a model. Thanks!
Stream is offline now
Use it if you want to read news for entertainment. Definitely don’t use it for information retrieval.
We just had elections here in Brazil and it was hugely influenced by Fake news spread on Whatsapp.
Also, I would say that recommendation algorithms of Facebook have created a terrible social experiment in global scale and we are paying a price for that.
@rachel when I click the live stream link above it take me to the recorded lecture and not the current live stream
how to use Nlp with transfer learning at charecter level instead of word level?
Can local “live” students come to the daily sessions, or is it only for USF students?
After the Windows update another restart might help!!
For people that want to learn more about this kind of bias and AI ethics, what sources would you recommend?