Hi, everyone! We just released Rustuna, a Rust implementation of Optuna with a focus on making large-scale hyperparameter optimization faster and more memory-efficient.
Rustuna keeps the familiar Optuna Python API while moving the optimization engine to Rust. Rustuna is not intended to replace Optuna. It currently supports a smaller feature set, and we plan to make the two work together more seamlessly. Looking ahead, we’re also excited about expanding support to other languages like Node.js, Go, and beyond.
Thank you! I updated the title.
I think type annotation in Python3 is a great.
And I thought there is a usage other than typing checking using mypy, so I try to convert url variables using the type information of function annotations.
Hi! I created and released a package named Feedy for collecting and processing the data from RSS feed. For example, If you want to collect images on CNN Website, creating main.py:
from feedy import Feedy
from bs4 import BeautifulSoup
app = Feedy('feedy.dat') # store last fetched time
@app.add('http://rss.cnn.com/rss/edition.rss')
def cnn(info, body):
soup = BeautifulSoup(body, "html.parser")
for x in soup.find_all('img'):
print(x['src'])
if __name__ == '__main__':
app.run()
If you are interested, please read README.md. And If you have some requests for this package, please tell me. I'll improve this package as soon as I can. (◕‿◕)
Rustuna keeps the familiar Optuna Python API while moving the optimization engine to Rust. Rustuna is not intended to replace Optuna. It currently supports a smaller feature set, and we plan to make the two work together more seamlessly. Looking ahead, we’re also excited about expanding support to other languages like Node.js, Go, and beyond.
GitHub: https://github.com/optuna/rustuna/
Blog post: https://medium.com/optuna/announcing-rustuna-cc82a6815bf7
We’d love to hear your feedback, or any ideas!
reply