[09x09] Bayesian Deep Learning through Laplace's Approximation using LaplaceRedux.jl and Flux.jl

[@patrickaltmeyer4339]

I love this ❤ thank you doggo!

[@doggodotjl]

Thank you for making this package and sharing it with the community! It's awesome!⭐⭐⭐⭐⭐

[@taraskywalker453]

omg your videos are not just simple, but incredibly impressive too! The sheer extent of your academic coverage is so appreciated! Thank you for providing all of these resources

[@doggodotjl]

You're welcome! Thanks for watching!

[@LasmaHan]

THX for always. Nowadays, I am frustrated with Julia due to its slooow debugging speed. Especially debugging the massive iteration code makes me crazy. Could you handle this topic once?

[@doggodotjl]

Sorry for the frustration. I am planning on covering debugging at some point, but it probably won't be for a while. If you don't mind me asking, is the slow debugging in VS Code?

[@LasmaHan]

Oh, I have not realized what I didn't response to you. As you guessed it, I'm using VScode as a IDF for JULIA!

[@TobeFreeman]

Dear Doggo.jl, I love these presentations. I only wanted to say you are using a legacy pattern in Flux to set the optimisation parameters. Optimisations.jl has moved out of FluxML and I wondered if you could update the repos to reflect this?

[@doggodotjl]

Thanks for letting me know. I didn't realize that Flux had changed. I'm happy to update my code in GitHub but I need some help since I can't get the code to work. Here's what I did. I added OptimizationOptimisers in the Package Manager. I added the lines "using OptimizationOptimisers" and "using OptimizationOptimisers.Optimisers: Adam" to my code and I removed the related Flux lines. But when I try to train the model I get the following error message: "ERROR: Optimisers.jl cannot be used with Zygote.jl's implicit gradients, `Params` & `Grads`". If you'd prefer to submit a PR, I'd be happy to take a look at it. Thanks!