|
Daniël de Kok 🦀 ❄️
@
danieldekok
Groningen, Nederland
|
|
Researcher in computational linguistics, c(rust)acean 🦀, Nix/NixOS ❄️, occasional tinkerer with electronics. Dad of a Lego queen 👸.
Opinions are my own.
|
|
|
4.403
Tweetovi
|
129
Pratim
|
269
Osobe koje vas prate
|
| Tweetovi |
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
31. sij |
|
Just pushed the version 0.1.0 of the `sentencepiece` @rustlang crate, which provides a rustic interface to the sentencepiece unsupervised text tokenizer. crates.io/crates/sentenc… rustdoc.danieldk.eu/sentencepiece/
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
30. sij |
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
26. sij |
|
|
||
| Daniël de Kok 🦀 ❄️ proslijedio/la je tweet | ||
|
$ zimbatm
@zimbatm
|
25. sij |
|
direnv 2.21.0 has been released. This is a massive release!
* `.envrc` files are now loaded with `set -euo pipefail` which is probably going to expose issues in existing scripts.
* Ctrl-C now actually works during reload in bash and zsh.
And more!
github.com/direnv/direnv/…
|
||
|
|
||
| Daniël de Kok 🦀 ❄️ proslijedio/la je tweet | ||
|
Laurent Mazare
@lmazare
|
25. sij |
|
Our #rust and #ocaml bindings for #PyTorch are now compatible with PyTorch 1.4, updated opam package/crate are available! A bunch of examples can be found on GitHub, RNNs, GAN, Reinforcement Learning...
github.com/LaurentMazare/…
github.com/LaurentMazare/…
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
25. sij |
|
Keep it short, keep it sweet. The goals is to convince an interviewer to do a follow-up, not to tell your life story.
I agree with @agonzalezro that burial in resumees is often not really a thing in IT outside FANG-like companies.
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
25. sij |
|
Dear lazyweb, I mean fellow @rustlang crustaceans. I am working on a crate that binds a C library, but can only reasonably unit-test it by providing a ~500KB data file. Would you find this acceptable for a tiny binding? Alternative: feature-gate those tests.
|
||
|
|
||
| Daniël de Kok 🦀 ❄️ proslijedio/la je tweet | ||
|
Ed Hawkins
@ed_hawkins
|
6. sij |
|
Australia: you have just experienced the future. pic.twitter.com/5H0ab4tMzJ
|
||
|
|
||
| Daniël de Kok 🦀 ❄️ proslijedio/la je tweet | ||
|
Rutger Bregman
@rcbregman
|
20. sij |
|
For some reason, I haven't been invited to Davos this year. Was it something I said? 🤔
pic.twitter.com/6Qmrup7HrB
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
18. sij |
|
Nix, magit, Emacs evil, direnv, tmux, fzf
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
18. sij |
|
0.45 dependency parsing LAS improvement on Dutch, with only half the parameters. That's what I call a productive week 🥂. Enjoy the weekend!
|
||
|
|
||
| Daniël de Kok 🦀 ❄️ proslijedio/la je tweet | ||
|
Saoirse Shipwreckt
@withoutboats
|
15. sij |
|
Firefox market share vs Mozilla Foundation chair salary (2.5 million/year in 2018) pic.twitter.com/htEKDNFMDH
|
||
|
|
||
| Daniël de Kok 🦀 ❄️ proslijedio/la je tweet | ||
|
Michal Strehovský
@MStrehovsky
|
9. sij |
|
1/7 Did you ever need to run a piece of C# code on Windows 3.11? Me neither, but I did it anyway. A thread. pic.twitter.com/ZW0p29d9Ef
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
9. sij |
|
Finally made the switch from Tensorflow to PyTorch. The tch-rs @rustlang bindings for libtorch are really nice. Learning the libtorch API and porting the BERT model from Hugging Face Transformers to Rust was really quick: crates.io/crates/sticker…
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
8. sij |
|
I am very happy that we use far less gas for heating/hot water than the average of all house types. But that bar chart is weird. The deltas are probably proportional, but the bars themselves… pic.twitter.com/HbIQ2IwG1I
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
7. sij |
|
They rolled out a large fleet here in Groningen as well. They are awesome for cyclists! No more burnt diesel exhaust when you are cycling behind a bus!
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
6. sij |
|
Why do you commit Cargo.lock for library crates, given that they are ignored by Cargo when the library is used as a dependency? Documentation?
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
6. sij |
|
The final, finetuned models, use each layer almost equally. The interesting patterns here seem to be: for lemmatization, almost all layers are used equally. For the other tasks, use increases per layer. 4/4 pic.twitter.com/LqoAQDaYXM
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
6. sij |
|
However, for morphological and POS tagging, it seems that for ML BERT the classifiers rely most on the initial/middle layers, whereas for BERTje more uniformly on all the layers. 3/4
|
||
|
|
||
|
Daniël de Kok 🦀 ❄️
@danieldekok
|
6. sij |
|
This can give hints where information is represented in the initial encoder. Interestingly, the information seems to be distributed quite differently between ML BERT and BERTje. In both, the last layers are used most for dependencies and the initial layers for lemmatization. 2/4 pic.twitter.com/KJKaVYklei
|
||
|
|
||