Helping data science teams deploy their algorithms.
Joined December 2020
- Tweets191
- Following272
- Followers263
- Likes336
jbencook retweeted
Check out our latest podcast episode. #DataScience #Bioinformatics with hosts @jillianerowe @jbencook @donnywinston podbean.com/eau/pb-u63bf-113… on #Podbean
jbencook retweeted
It was , as always, great fun hanging out with @donnywinston @jbencook on the @dsdeployed podcast!
Livestream - youtu.be/SAQmOENaO78
Or you can subscribe where the podcasts are - datasciencedeployed.io/e/epi…
jbencook retweeted
Check out our latest podcast episode. #DataScience #Bioinformatics with hosts @jillianerowe @jbencook @donnywinston podbean.com/eau/pb-p3kxp-113… on #Podbean
jbencook retweeted
Check out my last (short) article on @sicara_fr blog:
How to Automate @DVCorg Pipelines with Makefile?
sicara.ai/blog/automate-dvc-…
jbencook retweeted
Software engineering for data science is such a glaring, unsolved problem. I've worked for years at the intersection of the two and am still amazed by how difficult it is to write good, maintainable code for data science and how no one really even knows what that means yet.
jbencook retweeted
Migrated from Git LFS to DVC from @DVCorg a few months back. Model/dataset version control has never been smoother. We now spend way less time uploading and downloading large model/dataset versions manually.
#Data #MLOps #DeepLearning #ArtificialIntelligence
@Iterativeai
jbencook retweeted
Check out our latest podcast episode. #DataScience #Bioinformatics with hosts @jillianerowe @jbencook @donnywinston podbean.com/eau/pb-jadj8-112… on #Podbean
jbencook retweeted
Useful tips and cool tech if you are thinking if there is a better way to version your data in #datascience by @jillianerowe and the @dsdeployed team
xitter.cf/jillianerowe/status/14…
I built a toy ML pipeline to illuminate the most important ML Engineering best practices. You should check it out: github.com/sparrowml/mnist-p….
jbencook retweeted
x100 - argparse has that vintage, je ne sais quoi that suggests you’d train artisanal models with a vanilla Bourne Shell if only Torch had the right “shellTorch” bindings.
I nominate Fire for most underrated Python package — I don't understand why people still use argparse
github.com/google/python-fir…
3
6
Remote containers are the killer feature of VS Code because you can debug your Python code on a remote machine. That means you can run code that touches CUDA from your laptop. sparrow.dev/devcontainers-qu…
VS Code is better than PyCharm: 1) it’s an open source project with a huge community 2) features get added faster (because of 1) and 3) it works well for languages other than Python.
1
4
