Thread by @SoyGema

𝐕𝐞𝐫𝐬𝐢𝐨𝐧𝐢𝐧𝐠 ML experiments is key because experiments can grow rapidly:visualize them in a structured manner allows us to compare & analyze results🧠@DVCorg @code extension transforms IDE into a code-native machine learning experimentation platform! A thread 👇(1/4 )

1 ) 𝐑𝐮𝐧 𝐚𝐧𝐝 𝐜𝐨𝐦𝐩𝐚𝐫𝐞 . Define the metrics that will impact your machine learning model performance and business problem in evaluate.py very easily 🧙‍♂️

2) 𝐕𝐢𝐬𝐮𝐚𝐥𝐢𝐳𝐞 & 𝐟𝐢𝐥𝐭𝐞𝐫𝐢𝐧𝐠 : compare confusion matrix of all models and model performance. Filter from thresholds that you can choose to select the best machine learning model. 👀

3) 𝐏𝐞𝐫𝐬𝐢𝐬𝐭.Bring the desired models to the workspace or commit them into your github repository
Take a look at the extension, the blogpost -> dvc.org/blog/DVC-VS-Co… and feel free to reach out to us in our discord channel -> discord.gg/W49xzNmycw 😀😊