Maybe I'm obsessed with @DVCorg, but in each paragraph of this white paper cloud.google.com/resources/mlops-whitepaper, I think of DVC as a so good solution. What a wonderful framework. In my particular case, is a very good guide to putting all in order.
This! Literally hundreds of #MLOps popping everyone promising data scientists an experience based on limited choices and inflexible and repetitive GUIs. Thank you @Iterativeai for letting me write code and configuration instead ๐ง twitter.com/DVCorg/status/1532358000546942976
๐ Haven't hit the "enroll" button that hard in a while. Beautiful course, am excited to explore the materials: twitter.com/DVCorg/status/1529907752545439754
Having fun with @DVCorg and @jarvislabsai.
Currently wrote a script to launch VM instance, run a set of predefined experiments, make branches, visualize the results in @Iterativeai Studio automatically.
Next step is to write GitHub Actions for more automations on given PRs
At DeGould weโre big fans of @Iterativeaiโs tools and itโs great to see them progressing. See my quote in their recent @BusinessWire article businesswire.com/news/home/20220317005882/en/MLOps-Company-Iterative-Achieves-Significant-Customer-and-Company-Growth-in-2021 #MLOps
Swag received :) Thanks @DVCorg !! We love DVC at @sicara_fr ๐ keep up the great work ๐ @_Okamille @e_vignon @cpierrehenri @JPro20 @SoulMathieu @Arnault_Chaz
๐ฆ Am very much enjoying the modularity of @DVCorg's open-source products (particularly CML and DVC). The tools feel so much more composable than other MLOps systems...
"Want to use MLFlow for a portion of the process, instead? Go ahead!" Libraries = โฅ๏ธ .
dvc.org/doc/use-cases/ci-cd-for-machine-learning#continuous-integration-and-deployment-for-machine-learning
Interesting blog post from @DVCorg with their philosophy for managing workflows with ML.
I've enjoyed using DVC for the last year - for the first time I've had a process that allows for easy collaboration on data analysis with colleagues
dvc.org/blog/road-to-hell via @DVCorg
Startups I'm *incredibly* bullish about: @Stripe, @IterativeAI, @HuggingFace, and @Explosion_AI.
If you're an engineer/PM considering a career change (and it's that time of the year again, no? ๐)โbut want to opt away from FAAMG, definitely consider one of the companies above.
Reading up on @DVCorg and drooling at the potential for repeatable data science experiments with versioned code, data *and* metrics, AND the ability to do comparative plots!
If only I knew about @DVCorg during my MSc. thesis ๐คฆโโ๏ธI would not have had to come up with my โhome-grownโ version.
@DynamicWebPaige @DVCorg @github They are an exceptionally nice crew. Glad you are thinking about them. A very worthy direction for machine learning tooling.
๐ฆ I'm really enjoying reading through @DVCorg's documentation and use cases for operationalizing machine learning models.
dvc.org/doc/use-cases/versioning-data-and-model-files
If you've ever tried to put a model in production, these concepts will resonate. Check out their open-source project on @Github! โจ
DVC is an amazing tool. Great milestone.
It already removes a lot of headaches in my #MachineLearning work.
But with new features, I will be even more productive :) twitter.com/DVCorg/status/1367117270946746374
O @DVCorg seguramente estรก entre as melhores ferramentas do ecossistema dos รบltimos 3 anos. Nรฃo vai demorar para o DVC ser tรฃo comum quanto o Scikit-Learn em projetos de ML/DS com alta maturidade. ๐๐๐ twitter.com/DVCorg/status/1367117270946746374
Companies such as @astronomerio, @HashiCorp, @supabase_io, @Iterativeai are excellent examples of companies with a relentless focus on building for developer love.
This looks extremely useful. Full CI/CD for ML projects, baked directly into GitHub using DVC. Any PR that changes training code auto-triggers a new training run, and new metrics / image samples are auto-commented on the PR. Very cool! twitter.com/DVCorg/status/1280505393496629249